2017 WHO TRS 1004 Biological Standardization

This report presents the recommendations of a WHO Expert
Committee commissioned to coordinate activities leading to the

adoption of international recommendations for the production
1004
W H O Te c h n i c a l R e p o r t S e r i e s
and control of vaccines and other biological substances, and the
establishment of international biological reference materials.
Following a brief introduction, the report summarizes a number
1004
of general issues brought to the attention of the Committee. The
WHO Expert Committee on Biological Standardization

next part of the report, of particular relevance to manufacturers
and national regulatory authorities, outlines the discussions
held on the development and revision of WHO Guidelines for
a number of vaccines, blood products and related substances.
Specific discussion areas included WHO guidance on the
production and evaluation of the quality, safety and efficacy
of monoclonal antibodies as similar biotherapeutic products
(SBPs); blood and blood components as essential medicines;
estimation of residual risk of HIV, HBV or HCV infections
via cellular blood components and plasma; snake antivenom
immunoglobulins; human pandemic influenza vaccines in
non-vaccine-producing countries; and clinical evaluation of
vaccines: regulatory expectations. In addition, the following
WHO guidance documents were also adopted: WHO manual
WHO Expert Committee
for the preparation of secondary reference materials for in vitro
diagnostic assays designed for infectious disease nucleic acid or on Biological
Standardization
antigen detection: calibration to WHO International Standards;
and Human challenge trials for vaccine development:
regulatory considerations. One WHO addendum document –
Labelling information of inactivated influenza vaccines for use
in pregnant women – was also adopted.
Subsequent sections of the report provide information on the
current status, proposed development and establishment of
international reference materials in the areas of: biotherapeutics Sixty-seventh report
other than blood products; blood products and related
substances; cellular and gene therapies; in vitro diagnostics;
and vaccines and related substances.
A series of annexes are then presented which include an
WHO Technical Report Series

updated list of all WHO Recommendations, Guidelines and
other documents on biological substances used in medicine
(Annex 1). The above nine WHO documents adopted on
the advice of the Committee are then published as part
of this report (Annexes 2–10). Finally, all additions and
discontinuations made during the 2016 meeting to the list of
International Standards, Reference Reagents and Reference
Panels for biological substances maintained by WHO are
summarized in Annex 11. The updated full catalogue of
WHO International Reference Preparations is available at:
http://www.who.int/bloodproducts/catalogue/en/.
ISBN 978 92 4 121013 3

SELECTED WHO PUBLICATIONS OF RELATED INTEREST

The World Health Organization was established in 1948 as a specialized agency of the
Sixty-sixth report.
United Nations serving as the directing and coordinating authority for international
WHO Technical Report Series, No. 999, 2016 (xix + 267 pages)
health matters and public health. One of WHO’s constitutional functions is to
provide objective and reliable information and advice in the field of human health, a WHO Expert Committee on Biological Standardization
responsibility that it fulfils in part through its extensive programme of publications. Sixty-fifth report.
WHO Technical Report Series, No. 993, 2015 (xvi + 262 pages)
The Organization seeks through its publications to support national health strategies
and address the most pressing public health concerns of populations around the world. WHO Expert Committee on Biological Standardization
To respond to the needs of Member States at all levels of development, WHO publishes Sixty-fourth report.
practical manuals, handbooks and training material for specific categories of health WHO Technical Report Series, No. 987, 2014 (xviii + 266 pages)
workers; internationally applicable guidelines and standards; reviews and analyses of
health policies, programmes and research; and state-of-the-art consensus reports that WHO Expert Committee on Biological Standardization
offer technical advice and recommendations for decision-makers. These books are Sixty-third report.
closely tied to the Organization’s priority activities, encompassing disease prevention WHO Technical Report Series, No. 980, 2014 (xv + 489 pages)
and control, the development of equitable health systems based on primary health
care, and health promotion for individuals and communities. Progress towards better WHO Expert Committee on Biological Standardization
health for all also demands the global dissemination and exchange of information Sixty-second report.
that draws on the knowledge and experience of all WHO’s Member countries and the WHO Technical Report Series, No. 979, 2013 (xiii + 366 pages)
collaboration of world leaders in public health and the biomedical sciences. WHO Expert Committee on Biological Standardization
To ensure the widest possible availability of authoritative information and guidance on Sixty-first report.
health matters, WHO secures the broad international distribution of its publications WHO Technical Report Series, No. 978, 2013 (xi + 384 pages)
and encourages their translation and adaptation. By helping to promote and protect WHO Expert Committee on Biological Standardization
health and prevent and control disease throughout the world, WHO’s books contribute Sixtieth report.
to achieving the Organization’s principal objective – the attainment by all people of the WHO Technical Report Series, No. 977, 2013 (viii + 231 pages)
highest possible level of health.
The WHO Technical Report Series makes available the findings of various international Fifty-ninth report.
groups of experts that provide WHO with the latest scientific and technical advice on WHO Technical Report Series, No. 964, 2012 (viii + 228 pages)
a broad range of medical and public health subjects. Members of such expert groups
serve without remuneration in their personal capacities rather than as representatives WHO Expert Committee on Biological Standardization
of governments or other bodies; their views do not necessarily reflect the decisions or Fifty-eighth report.
the stated policy of WHO. WHO Technical Report Series, No. 963, 2011 (viii + 244 pages)
For further information, please contact: WHO Press, World Health Organization, Website: http://www.who.int/biologicals
20 avenue Appia, 1211 Geneva 27, Switzerland (tel. +41 22 791 3264; fax: +41 22 791 4857;
email: bookorders@who.int; order on line: www.who.int/bookorders).
Further information on these and other WHO publications can be obtained from
WHO Press, World Health Organization, 1211 Geneva 27, Switzerland
(tel.: +41 22 791 3264; fax: + 41 22 791 4857; email: bookorders@who.int;
order online: www.who.int/bookorders)
1 0 0 4

on Biological
Standardization
Sixty-seventh report
This report contains the collective views of an international group of experts and
does not necessarily represent the decisions or the stated policy of the World Health Organization
WHO Library Cataloguing-in-Publication Data
WHO Expert Committee on Biological Standardization, sixty-seventh report
(WHO technical report series ; no. 1004)
ISBN 978-92-4-121013-3
ISBN (PDF) 978-92-4-069645-7
ISSN 0512-3054
© World Health Organization 2017

Some rights reserved. This work is available under the Creative Commons Attribution-NonCommercial-
ShareAlike 3.0 IGO licence (CC BY-NC-SA 3.0 IGO; https://creativecommons.org/licenses/by-nc-sa/3.0/
igo).
Under the terms of this licence, you may copy, redistribute and adapt the work for non-commercial
purposes, provided the work is appropriately cited, as indicated below. In any use of this work, there
should be no suggestion that WHO endorses any specific organization, products or services. The use of
the WHO logo is not permitted. If you adapt the work, then you must license your work under the same
or equivalent Creative Commons licence. If you create a translation of this work, you should add the
following disclaimer along with the suggested citation: “This translation was not created by the World
Health Organization (WHO). WHO is not responsible for the content or accuracy of this translation. The
original English edition shall be the binding and authentic edition”.
Any mediation relating to disputes arising under the licence shall be conducted in accordance with
the mediation rules of the World Intellectual Property Organization.
Suggested citation. WHO Expert Committee on Biological Standardization, sixty-seventh report.
Geneva: World Health Organization; 2017 (WHO technical report series ; no. 1004). Licence:
CC BY‑NC-SA 3.0 IGO.
Cataloguing-in-Publication (CIP) data. CIP data are available at http://apps.who.int/iris.
Sales, rights and licensing. To purchase WHO publications, see http://apps.who.int/bookorders.
To submit requests for commercial use and queries on rights and licensing, see http://www.who.int/
about/licensing.
Third-party materials. If you wish to reuse material from this work that is attributed to a third party,
such as tables, figures or images, it is your responsibility to determine whether permission is needed
for that reuse and to obtain permission from the copyright holder. The risk of claims resulting from
infringement of any third-party-owned component in the work rests solely with the user.
General disclaimers. The designations employed and the presentation of the material in this
publication do not imply the expression of any opinion whatsoever on the part of WHO concerning
the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation
of its frontiers or boundaries. Dotted and dashed lines on maps represent approximate border lines
for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they
are endorsed or recommended by WHO in preference to others of a similar nature that are not
mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by
initial capital letters.
All reasonable precautions have been taken by WHO to verify the information contained in this
publication. However, the published material is being distributed without warranty of any kind, either
expressed or implied. The responsibility for the interpretation and use of the material lies with the
reader. In no event shall WHO be liable for damages arising from its use.
This publication contains the collective views of an international group of experts and does not
necessarily represent the decisions or the policies of WHO.
Printed in Italy
Contents
Abbreviations xv
1. Introduction 1
2. General 4
2.1 Current directions 4
2.1.1 Strategic directions in biological standardization: WHO priorities 4
2.1.2 Vaccines and biotherapeutics: recent and planned activities in biological
standardization 5
2.1.3 Blood products and in vitro diagnostics: recent and planned activities in
biological standardization 8
2.2 Reports 10
2.2.1 Report from the WHO Blood Regulators Network 10
2.2.2 Report from the WHO network of collaborating centres on standardization
and regulatory evaluation of vaccines 12
2.2.3 Report from the WHO network of collaborating centres for blood products
and in vitro diagnostics 13
2.3 Feedback from custodian laboratories 14
2.3.1 Developments and scientific issues highlighted by custodians of WHO
biological reference preparations 14
2.4 Cross-cutting activities of other WHO committees and groups 20
2.4.1 WHO Global Model Regulatory Framework for Medical Devices
including IVD medical devices 20
2.4.2 Report of a WHO ad hoc consultation on International Nonproprietary
Names for biological substances 21
2.4.3 Monographs on capreomycin sulfate and capreomycin for injection in
The International Pharmacopoeia 22
2.4.4 Yellow fever vaccine shortages and outbreak response 24
2.4.5 Vaccines for public health emergencies 25
3. International Recommendations, Guidelines and other matters
related to the manufacture, quality control and evaluation of
biological substances 28
3.1 Biotherapeutics other than blood products 28
3.1.1 Guidelines on evaluation of monoclonal antibodies as similar
biotherapeutic products (SBPs) 28
3.2 Blood products and related substances 29
3.2.1 Blood regulation activities 29
3.2.2 Africa Society for Blood Transfusion 30
3.2.3 Guidelines on management of blood and blood components as
essential medicines 32
3.2.4 Guidelines on estimation of residual risk of HIV, HBV or HCV infections
via cellular blood components and plasma 33
3.2.5 WHO assessment of antivenoms 34
3.2.6 Guidelines for the production, control and regulation of snake
antivenom immunoglobulins 35
3.3 Cellular and gene therapies 36
3.3.1 Regulation of cell therapy products 36
iii
3.3.2 Reference preparations for gene therapy products 38
3.4 In vitro diagnostics 40
3.4.1 Preparation of secondary reference materials for in vitro diagnostic
assays designed for infectious disease nucleic acid or antigen detection:
calibration to WHO International Standards 40
3.4.2 WHO prequalification of in vitro diagnostic devices 41
3.5 Vaccines and related substances 43
3.5.1 Revision of WHO Guidelines for the safe production and quality control of
inactivated poliomyelitis vaccines manufactured from wild polioviruses 43
3.5.2 Guidelines on regulatory preparedness for provision of marketing
authorization of human pandemic influenza vaccines in non-vaccine-
producing countries 44
3.5.3 Labelling information of inactivated influenza vaccines for use in
pregnant women 46
3.5.4 Revision of the WHO Guidelines on clinical evaluation of vaccines:
regulatory expectations 47
3.5.5 Human challenge trials for vaccine development: regulatory considerations 48
3.5.6 Guidelines on the quality, safety and efficacy of Ebola vaccines 49
4. International reference materials – biotherapeutics other than
blood products 51
4.1 Proposed new projects and updates – biotherapeutics other than blood products 51
4.1.1 Proposed Second WHO International Standard for parathyroid
hormone 1-34 (recombinant, human) 51
4.1.2 Proposed First WHO international standards for vascular endothelial
growth factor antagonists 52
4.1.3 Proposed First WHO international standards (or reference panels)
for antibodies for use in immunogenicity assessments of biotherapeutic
products 53
4.1.4 Proposed First WHO international standards (or reference reagents)
for monoclonal antibodies to the ErbB/HER receptor family 54
5. International reference materials – blood products and related
substances 56
5.1 WHO International Standards and Reference Reagents – blood products and
related substances 56
5.1.1 Second WHO International Standard for ancrod 56
5.1.2 First WHO Reference Reagent for batroxobin 57
5.1.3 Second WHO International Standard for blood coagulation factor XI
(plasma, human) 57
5.1.4 Fifth WHO International Standard for thromboplastin (recombinant,
human, plain) 58
5.1.5 Fifth WHO International Standard for thromboplastin (rabbit, plain) 59
5.2 Proposed new projects and updates – blood products and related substances 60
5.2.1 Proposed Second WHO International Standard for blood coagulation
factor V (plasma, human) 60
5.2.2 Proposed First WHO International Standard for blood coagulation
factor XII (plasma, human) 60
5.2.3 Proposed assignment of factor XIII-B subunit (total and free) values to
the First WHO International Standard for factor XIII plasma 61
iv
5.2.4 Proposed First WHO International Standard for thrombin-activatable
fibrinolysis inhibitor (plasma, human) 62
5.2.5 Proposed Third WHO International Standard for anti-D immunoglobulin 63
5.2.6 Update on the use of WHO reference materials in assays to detect
activated blood coagulation factor XI in immunoglobulins 63
6. International reference materials – cellular and gene therapies 65
6.1 Proposed new projects and updates – cellular and gene therapy 65
6.1.1 Proposed First WHO International Standard for lentiviral vector copy
number quantitation 65
7. International reference materials – in vitro diagnostics 67
7.1 WHO International Standards and Reference Reagents – in vitro diagnostics 67
7.1.1 First WHO International Standard for Zika virus RNA for NAT-based assays 67
7.1.2 First WHO Reference Panel for Ebola virus VP40 antigen 68
7.1.3 First WHO reference reagents for dengue virus serotypes 1–4 RNA for
NAT-based assays 69
7.1.4 Fourth WHO International Standard for hepatitis B virus DNA for
NAT-based assays 70
7.1.5 Fourth WHO International Standard for prolactin (pituitary, human) 72
7.1.6 First WHO Reference Panel for the Janus kinase 2 V617F gene mutation 73
7.2 Proposed new projects and updates – in vitro diagnostics 74
7.2.1 Update on the proposed First WHO International Standard for Ebola virus
antibodies (plasma, human) 74
7.2.2 Proposed First WHO International Standard for Zika virus antibodies
(immunoglobulin G and immunoglobulin M) (human) 75
7.2.3 Proposed First WHO International Standard for chikungunya virus
antibodies (immunoglobulin G and immunoglobulin M) (human) 75
7.2.4 Proposed Second WHO International Standard for syphilitic plasma
(immunoglobulin G and immunoglobulin M) (human) 76
7.2.5 Proposed Second WHO International Standard for syphilitic plasma
(immunoglobulin G) (human) 77
7.2.6 Proposed Second WHO International Standard for human
immunodeficiency virus type 2 RNA for NAT-based assays 77
7.2.7 Proposed First WHO International Standard for respiratory syncytial virus
RNA for NAT-based assays 78
7.2.8 Proposed First WHO International Standard for influenza virus type A RNA
for NAT-based assays 79
7.2.9 Proposed First WHO International Standard for influenza virus type B RNA
for NAT-based assays 79
7.2.10 Proposed First WHO Reference Panel for the BRAF V600E gene mutation 79
7.2.11 Proposed First WHO Reference Panel for ErbB2 copy number and
mRNA expression 80
8. International reference materials – vaccines and related substances 81
8.1 Proposed new projects and updates – vaccines and related substances 81
8.1.1 Proposed First WHO International Standard for Sabin inactivated
poliomyelitis vaccine 81
8.1.2 Proposed First WHO international standards for Group B streptococcus
(polysaccharide and antiserum) 81
v
8.1.3 Proposed Second WHO International Standard for diphtheria
antitoxin (equine) 82
8.1.4 Proposed First WHO international standards for antibodies against
human papillomavirus types 6, 11, 31, 33, 45, 52 and 58 83
8.1.5 Proposed Second WHO International Standard for hepatitis A vaccine 84
8.1.6 Proposed WHO international standards for oral poliomyelitis vaccine 85
Annex 1
WHO Recommendations, Guidelines and other documents related to the manufacture,
quality control and evaluation of biological substances used in medicine 87
Annex 2
Guidelines on evaluation of monoclonal antibodies as similar biotherapeutic
products (SBPs) 93
Annex 3
Guidelines on management of blood and blood components as essential medicines 129
Annex 4
Guidelines on estimation of residual risk of HIV, HBV or HCV infections via cellular
blood components and plasma 163
Annex 5
Guidelines for the production, control and regulation of snake antivenom
immunoglobulins
Replacement of Annex 2 of WHO Technical Report Series, No. 964 197
Annex 6
WHO manual for the preparation of secondary reference materials for in vitro
diagnostic assays designed for infectious disease nucleic acid or antigen detection:
calibration to WHO International Standards 389
Annex 7
Guidelines on regulatory preparedness for provision of marketing authorization of
human pandemic influenza vaccines in non-vaccine-producing countries 457
Annex 8
Labelling information of inactivated influenza vaccines for use in pregnant women
Addendum to Annex 3 of WHO Technical Report Series, No. 927 487
Annex 9
Guidelines on clinical evaluation of vaccines: regulatory expectations
Replacement of Annex 1 of WHO Technical Report Series, No. 924 503
Annex 10
Human challenge trials for vaccine development: regulatory considerations 575
Annex 11
Biological substances: WHO International Standards, Reference Reagents and
Reference Panels 589
vi
17 to 21 October 2016
Committee members1
Professor K. Cichutek, Paul-Ehrlich-Institut, Langen, Germany (Chair)
Dr J. Epstein, Center for Biologics Evaluation and Research, Food and Drug Administration,
Silver Spring, MD, United States of America (USA) (also Blood Regulators Network
(BRN) representative)
Dr E. Griffiths, Kingston-upon-Thames, England (Rapporteur)
Dr S. Hindawi, Blood Transfusion Services, Jeddah, Saudi Arabia
Mrs T. Jivapaisarnpong, Department of Medical Sciences, Ministry of Public Health,
Nonthaburi, Thailand
Dr H. Klein, National Institutes of Health, Bethesda, MD, the USA (Vice-Chair)
Dr P. Minor, National Institute for Biological Standards and Control, Potters Bar, England
Dr F. Moftah, Ministry of Health and Population, Cairo, Egypt
Mr V.R. Reddy,2 South African National Blood Service, Weltevreden Park, South Africa
Dr L.S. Slamet, Technical Adviser and Consultant to the National Agency of Drug and
Food Control, Jakarta Selatan, Indonesia
Dr P. Strengers, Sanquin, Amsterdam, the Netherlands
Dr Y. Sohn, Ministry of Food and Drug Safety, Chungcheongbuk-do, Republic of Korea
Dr J. Wang, National Institutes for Food and Drug Control, Beijing, China
Temporary advisers
Dr S. Baylis,3 Paul-Ehrlich-Institut, Langen, Germany
Dr M. Gruber, Center for Biologics Evaluation and Research, Food and Drug Administration,
Rockville, MD, the USA
Dr C. Morris, National Institute for Biological Standards and Control, Potters Bar, England
(Rapporteur for the blood products and in vitro diagnostics track)
1
The decisions of the Committee were taken in closed session with only members of the Committee
present. Each Committee member had completed a Declaration of Interests form prior to the meeting.
These were assessed by the WHO Secretariat and no declared interests were considered to be in conflict
with full meeting participation.
2
Unable to attend.
3
Participated via teleconference.
vii
WHO Expert Committee on Biological Standardization Sixty-seventh report
Dr M. Powell, Medicines and Healthcare Products Regulatory Agency, London, England

Dr J. Reinhardt, Paul-Ehrlich-Institut, Langen, Germany (Rapporteur for the blood products
and in vitro diagnostics track)
Dr R. Sheets,4 Grimalkin Partners, Silver Spring, MD, the USA
Dr J. Southern, Adviser to the Medicines Control Council of South Africa, Cape Town,
South Africa
Dr Y. Sun, Paul-Ehrlich-Institut, Langen, Germany
Dr M. Udell, Medicines and Healthcare Products Regulatory Agency, London, England
Dr A.M.H.P. van den Besselaar, Leiden University Medical Centre, Leiden, the Netherlands
Dr A.L. Waddell, Stanley, England (Freelance writer)
Dr J. Wang, Biologics and Genetic Therapies Directorate, Health Canada, Ottawa, Canada
Dr D. Williams, University of Melbourne, Australia
Participants
Dr F. Agbanyo, Biologics and Genetic Therapies Directorate, Health Canada, Ottawa,
Canada (also BRN representative)
Dr N. Almond, National Institute for Biological Standards and Control, Potters Bar, England
Dr L Amsler,5 Swiss Agency for Therapeutic Products, Bern, Switzerland
Dr J. Boyle,6 National Institute for Biological Standards and Control, Potters Bar, England
Dr A. Bristow, National Institute for Biological Standards and Control, Potters Bar, England
Professor D. Calam, Pewsey, England (International Nonproprietary Names (INN) expert)
Dr F. Cano, Agence nationale de sécurité du médicament et des produits de santé,
Lyons, France
WHO Technical Report Series, No. 1004, 2017
Dr H. Castro, Ministry of Health, Bogota, Colombia

Dr R. Chhabra, Ministry of Health and Family Welfare, New Delhi, India
Dr M. Chudy, Paul-Ehrlich-Institut, Langen, Germany
Dr G. Cooper,7 National Institute for Biological Standards and Control, Potters Bar, England
4
5
Participated only in closed BRN session (observer).
6
7
viii
Dr B. Cowper,8 National Institute for Biological Standards and Control, Potters Bar, England
Dr R. Dominguez Morales, Ministerio de Salud Pública, Havana, Cuba
Professor S. Efstahiou, National Institute for Biological Standards and Control, Potters Bar,
England
Dr L. Elmgren, Biologics and Genetic Therapies Directorate, Health Canada, Ottawa,
Canada (also BRN representative)
Dr S. Fakhrzadeh, Ministry of Health and Medical Education, Tehran, the Islamic Republic
of Iran
Dr J. Ferguson,9 National Institute for Biological Standards and Control, Potters Bar, England
Dr K. Grant, Therapeutic Goods Administration, Woden, ACT, Australia
Dr E. Gray,10 National Institute for Biological Standards and Control, Potters Bar, England
Dr I. Hamaguchi, National Institute of Infectious Diseases, Tokyo, Japan (also BRN
representative)
Mr J. Hou, National Institutes for Food and Drug Control, Beijing, China
Dr A. Hilger,11 Paul-Ehrlich-Institut, Langen, Germany
Dr C.C. Ilonze, National Agency for Food and Drug Administration and Control, Lagos,
Nigeria
Ms S. Jadoonkittinan, Ministry of Public Health, Nonthaburi, Thailand
Mrs W. Jariyapan, Ministry of Public Health, Nonthaburi, Thailand
Dr H. Jia,12 National Institute for Biological Standards and Control, Potters Bar, England
Dr A. Kato, National Institute of Infectious Diseases, Tokyo, Japan
Dr B. Kim, Ministry of Food and Drug Safety, Chungcheongbuk-do, Republic of Korea
Dr J. Kim, Ministry of Food and Drug Safety, Chungcheongbuk-do, Republic of Korea
Dr T. Kondo,13 Ministry of Health, Labour and Welfare, Tokyo, Japan
Dr C. Lee, Ministry of Food and Drug Safety, Chungcheongbuk-do, Republic of Korea
Dr N. Lelie,14 Scientific Affairs Consultant, Paris, France
8
9
10
11
Participated only in closed BRN session.
12
13
Participated only in closed BRN session via teleconference.
14
ix
Dr C. Longstaff,15 National Institute for Biological Standards and Control, Potters Bar,
England
Dr J. Martin,16 National Institute for Biological Standards and Control, Potters Bar, England
Dr F. Mawas,17 National Institute for Biological Standards and Control, Potters Bar, England
Dr H. Meyer, Paul-Ehrlich-Institut, Langen, Germany
Mr H. Mogtari, Food and Drugs Authority Ghana, Accra, Ghana
Ms N.A.M. Nur, National Pharmaceutical Regulatory Agency, Selangor, Malaysia
Dr M. Ochiai, National Institute of Infectious Diseases, Tokyo, Japan
Dr H. Oh, Ministry of Food and Drug Safety, Chungcheongbuk-do, Republic of Korea
Dr I. Osipova, Ministry of Health of the Russian Federation, Moscow, Russian Federation
Dr W. Oualikene-Gonin,18 Agence nationale de sécurité du médicament et des produits
de santé, Saint Denis, France
Dr D. Padley,19 National Institute for Biological Standards and Control, Potters Bar, England
Dr M. Page,20 National Institute for Biological Standards and Control, Potters Bar, England
Dr S. Phumiamorn, Ministry of Public Health, Nonthaburi, Thailand
Ms E.I. Prawahju, National Agency of Drug and Food Control, Jakarta Pusat, Indonesia
Dr S. Prior,21 National Institute for Biological Standards and Control, Potters Bar, England
Dr I. Prosser,22 Therapeutic Goods Administration, Woden, ACT, Australia
Dr S. Raut,23 National Institute for Biological Standards and Control, Potters Bar, England
Dr A. Reissinger,24 Paul-Ehrlich-Institut, Langen, Germany
Dr M. Rios, Center for Biologics Evaluation and Research, Food and Drug Administration,
Silver Spring, MD, the USA25
Ms J. Rodgers, Food and Drugs Authority Ghana, Accra, Ghana
15
16
17
18
19
20
21
22
23
24
25
x
Dr I. Sainte-Marie,26 Agence nationale de sécurité du médicament et des produits de

santé, Saint Denis, France
Dr C. Schaerer, Swiss Agency for Therapeutic Products, Bern, Switzerland (also BRN
representative)
Dr C. Schneider, National Institute for Biological Standards and Control, Potters Bar,
England
Dr G. Smith,27 Therapeutic Goods Administration, Woden, ACT, Australia
Dr D. Stahl, Paul-Ehrlich-Institut, Langen, Germany (also BRN representative)
Dr P. Stickings, National Institute for Biological Standards and Control, Potters Bar, England
Dr C. Thelwell,28 National Institute for Biological Standards and Control, Potters Bar,
England
Dr R. Thorpe, Consultant, Welwyn, England
Dr A. Tsindimeev, Ministry of Health of the Russian Federation, Moscow, Russian
Federation
Dr G. Unger, Paul-Ehrlich-Institut, Langen, Germany
Dr A. Vasheghani Farahani, Food and Drug Administration, Tehran, the Islamic Republic
of Iran
Mrs M. Visagie, National Control Laboratory for Biological Products, Health Sciences,
Bloemfontein, South Africa
Dr M. Wadhwa,29 National Institute for Biological Standards and Control, Potters Bar,
England
Dr J. Weir, Center for Biologics Evaluation and Research, Food and Drug Administration,
Silver Spring, MD, the USA
Dr D. Wilkinson,30 National Institute for Biological Standards and Control, Potters Bar,
England
Dr M. Xu, National Institutes for Food and Drug Control, Beijing, China
Dr D. Yakunin, Federal Service on Surveillance in Healthcare and Social Development,
Moscow, Russian Federation
26
27
28
29
30
xi
Representation from non-state actors

Africa Society for Blood Transfusion
Mrs B. Armstrong, Pinetown, South Africa
Bureau International des Poids et Mesures
Dr R. Wielgosz, Director of Chemistry Department, Sevres, France
Council of Europe, European Directorate for the Quality of Medicines & HealthCare
Dr K-H. Buchheit, Department of Biological Standardisation, OMCL Network & HealthCare,
Strasbourg, France
Dr E. Charton, European Pharmacopoeia Department, Strasbourg, France
Dr A. Lodi, Laboratory Department, Strasbourg, France
Dr M. Wierer, Department of Biological Standardisation, OMCL Network & HealthCare,
Strasbourg, France
Developing Countries Vaccine Manufacturers Network 31
Dr N. Dellepiane, Serum Institute of India Pvt. Ltd, Nyon, Switzerland
Dr S. Gairola, Serum Institute of India Pvt. Ltd, Pune, India
Dr S. Pagluisi, Executive Secretary, Nyon, Switzerland
Dr V. Paradkar, Biological E Ltd, Hyderabad, India
International Alliance of Biological Standardization
Dr P. Neels, University of Namur, Zoersel, Belgium
International Federation of Clinical Chemistry and Laboratory Medicine
Professor P. Gillery, Hôpital Maison Blanche, Reims, France
International Federation of Pharmaceutical Manufacturers & Associations 33
Dr M. Downham,32 MedImmune, the USA
Dr L. Mallet, Sanofi Pasteur, Marcy L’Etoile, France
Dr S. Ramanan,33 Amgen, Thousand Oaks, CA, the USA

Dr C. Saillez, GlaxoSmithKline Vaccines, Wavre, Belgium
Dr M. van Ooij, Janssen Vaccines & Prevention B.V., Leiden, the Netherlands
Dr J. Wolf,34 Merck & Co., Inc., Kenilworth, NJ, the USA
31
A maximum of two representatives of the Developing Countries Vaccine Manufacturers Network and
two representatives of the International Federation of Pharmaceutical Manufacturers & Associations were
present in the meeting room during discussion of any one agenda item.
32
33
34
xii
International Generic and Biosimilars Medicines Association (IGBA)

Dr S. Kox, Medicines for Europe, Brussels, Belgium
International Society of Blood Transfusion
Dr M. Olsson, Amsterdam, the Netherlands
Medicines for Europe
Dr M. Schiestl, Sandoz GmbH, Kundl, Austria
Plasma Protein Therapeutics Association
Dr D. Misztela, Brussels, Belgium
United States Pharmacopeial Convention
Dr K. Carrick, Science-Global Biologics, Global Science and Standards Division, Rockville,
MD, the USA
Dr T. Morris, Science-Global Biologics, Global Science and Standards Division, Rockville,
MD, the USA
World Federation of Hemophilia
Dr M. El-Ekiaby, Shabrawishi Hospital Blood Transfusion and Hemophilia Treatment
Center, Giza, Egypt
WHO Secretariat
Regional Offices
Regional Office for Africa – Dr B. Akanmori
Regional Office for the Eastern Mediterranean – Dr Y. Abdella and Dr H. Langar
Regional Office for South-East Asia – Mr M. Eisenhawer
Regional Office for the Western Pacific – Dr J. Shin
Headquarters
Dr S. Hill, Director, WHO Department of Essential Medicines and Health Products (EMP)
Dr D.J. Wood, Coordinator, Technologies, Standards and Norms (TSN)/EMP (Committee
Secretary)
Dr J. Fournier-Caruana, Prequalification Team (PQT)/EMP
Dr K. Gao, TSN/EMP
Dr J. Hansen, TSN/EMP
Dr J. Hombach, Department of Immunization, Vaccines & Biologicals (IVB)/Initiative for
Vaccine Research (IVR)
Dr H-N Kang TSN/EMP
Dr I. Knezevic TSN/EMP (Lead for the vaccines and biotherapeutics track)
xiii
Dr D. Lei TSN/EMP
Dr R. Meurant, PQT/EMP
Dr M. Nuebling, TSN/EMP (Lead for the blood products and in vitro diagnostics track)
Ms I. Prat, PQT/EMP
Ms C.A. Rodriguez-Hernandez, PQT/EMP
Dr H. Schmidt, TSN/EMP
Dr I. Shin, TSN/EMP
Dr T. Zhou, TSN/EMP
xiv
Abbreviations
Ab antibody
Ag antigen
anti-HBc antibodies to hepatitis B core protein
anti-HBs antibodies to hepatitis B surface antigen
ASfBT Africa Society for Blood Transfusion
BGTD Biologics and Genetic Therapies Directorate
BRN WHO Blood Regulators Network
CBER Center for Biologics Evaluation and Research
CEG Core Expert Group
CHIKV chikungunya virus
CTP cell therapy product
DENV dengue virus
DNA deoxyribonucleic acid
DTP diphtheria, tetanus and pertussis
EBOV Ebola virus
ECSPP WHO Expert Committee on Specifications for Pharmaceutical
Preparations
EDQM European Directorate for the Quality of Medicines & HealthCare
EGFR epidermal growth factor receptor
ELISA enzyme-linked immunosorbent assay
EMA European Medicines Agency
EMP WHO Department of Essential Medicines and Health Products
EUAL WHO emergency use assessment and listing (procedure)
EV enterovirus
EVD Ebola virus disease
EQA external quality assurance
FDA Food and Drugs Authority (Ghana)
FV blood coagulation factor V
xv
FXI blood coagulation factor XI

FXIa activated blood coagulation factor XI
FXI:Ag blood coagulation factor XI (antigen value)
FXI:C blood coagulation factor XI (functional activity)
FXII blood coagulation factor XII
FXIII blood coagulation factor XIII
GACVS WHO Global Advisory Committee on Vaccine Safety
GAPIII WHO Global Action Plan to minimize poliovirus facility-
associated risk after type-specific eradication of wild
polioviruses and sequential cessation of oral polio vaccine use
GBS Group B streptococcus
GCV geometric coefficient of variation
GMP good manufacturing practice(s)
GPP good preparation practice(s)
HAV hepatitis A virus
HBsAg hepatitis B surface antigen
HBV hepatitis B virus
HCV hepatitis C virus
HER human epidermal growth factor receptor
HI haemagglutination inhibition
HIV human immunodeficiency virus
HPLC high-performance liquid chromatography

HPV human papillomavirus
ICDRA International Conference of Drug Regulatory Authorities
INN International Nonproprietary Name(s)
INR international normalized ratio
IPV inactivated poliomyelitis vaccine
ISI International Sensitivity Index
ISTH International Society on Thrombosis and Haemostasis
IU International Unit(s)
xvi
Abbreviations
IVD in vitro diagnostic

mAb monoclonal antibody
MERS-CoV Middle East respiratory syndrome coronavirus
MFDS Ministry of Food and Drug Safety
MN microneutralization
MPN myeloproliferative neoplasm
MSF Médecins Sans Frontières
NAPTT non-activated partial thromboplastin time
NAT nucleic acid amplification technique
NBS National Blood Service (Ghana)
NCL national control laboratory
NDU NAT-detectable unit
NIBSC National Institute for Biological Standards and Control
NIFDC National Institutes for Food and Drug Control
NIFDS National Institute of Food and Drug Safety Evaluation
NIID National Institute of Infectious Diseases
NIST National Institute for Standards and Technologies
NRA national regulatory authority
OMCL Official Medicines Control Laboratory (network)
OPV oral poliomyelitis vaccine
PCR polymerase chain reaction
PDMP plasma-derived medicinal product
PEESP WHO Polio Eradication and Endgame Strategic Plan 2013–2018
PEI Paul-Ehrlich-Institut
PT prothrombin time
rDNA recombinant DNA
RDT rapid diagnostic test
rhPTH1-34 parathyroid hormone 1-34 (recombinant, human)
RNA ribonucleic acid
xvii
RSV respiratory syncytial virus

SAGE WHO Strategic Advisory Group of Experts
SBP similar biotherapeutic product
sIPV Sabin inactivated poliomyelitis vaccine
SoGAT Standardisation of Genome Amplification Techniques (group)
SSC Scientific and Standardization Committee (of ISTH)
TAFI Thrombin-activatable fibrinolysis inhibitor
TGA Therapeutic Goods Administration
VEGF vascular endothelial growth factor
VIFFP virus-inactivated fresh frozen plasma
WHO CC WHO collaborating centre
WNV West Nile virus
YF yellow fever
YFV yellow fever virus
ZIKV Zika virus
xviii
1. Introduction
The WHO Expert Committee on Biological Standardization met in Geneva
from 17 to 21 October 2016. The meeting was opened on behalf of the Director-
General of WHO by Dr Suzanne Hill, the recently appointed Director of Essential
Medicines and Health Products (EMP). Dr Hill welcomed the Committee,
meeting participants and observers. She informed the meeting of several staff
changes in the Department over the previous year, with a number of further
senior WHO staff changes expected in the near future. This would include
the Secretary to the WHO Expert Committee on Biological Standardization,
Dr David Wood, who would be retiring in early 2017.
Dr Hill referred to the United Nations overarching strategic
direction entitled “Transforming our world: The 2030 Agenda for Sustainable
Development” 1 and indicated that this was being translated into a new vision for
EMP. One new initiative would be a greater focus on access to biotherapeutics.
It is envisaged that by 2030, biological substances will be used more widely than
at present and ensuring sustainable access to biotherapeutics of assured quality
for public health-care systems will be a key challenge for all countries, rich and
poor alike. Dr Hill emphasized that the development and adoption of norms
and standards to regulate the quality, safety and efficacy of medical products
and guide their cost-effective use would be a critical foundation on which future
aspirations would be built. A global approach to this core normative work is
facilitated by the coordinated efforts of WHO Expert Committees and the vital
support of WHO collaborating centres (WHO CCs) and partner organizations.
Dr Hill reminded meeting participants that the Committee had a
mandate, enunciated in the WHO Constitution, to develop, establish and
promote international standards for biological products. In addition to
supporting the development and use of biological medicines, other goals now
needed to be considered – specifically, promoting access to essential medicines
and regulatory strengthening. Furthermore, the world of regulatory science was
changing, particularly as new products came to the market, and there was an
expectation concerning the role of WHO in providing regulatory guidance and
promoting regulatory strengthening. Norms and standards underpin and reflect
these expectations but need to be regularly reviewed to ensure they reflect the
best regulatory science. The convergence of norms and standards internationally
is recognized as a key driver in addressing these needs.
After norms and standards have been established there was then a
need for proactive technical support from WHO to its Member States in order
For further details see: https://sustainabledevelopment.un.org/post2015/transformingourworld (accessed

1
17 February 2017).
1
to obtain maximum understanding and impact, and to facilitate consistent

application. There were many emerging issues to be dealt with, including
learning the lessons of the recent Ebola public health emergency, particularly
in relation to the rolling out of candidate vaccines and other health products
during an epidemic. Regulatory preparedness for future emergency situations
was crucial. There was also a need to consider the issue of access to new products
in a timely way.
After touching on the extremely heavy workload of the Committee and
acknowledging the difficulty of running a two-track meeting – with the blood
products and in vitro diagnostics track and the vaccines and biotherapeutics
track running in parallel – Dr Hill moved on to the election of meeting officials.
In the absence of dissent, Professor Klaus Cichutek was elected as Chair and
Dr Elwyn Griffiths as Rapporteur for the plenary sessions, and for the track
considering vaccines and biotherapeutics. Dr Harvey Klein was elected as
Chair and Dr Clare Morris and Dr Jens Reinhardt as Rapporteurs for the track
considering blood products and in vitro diagnostics. Dr Klein was also elected
as Vice-Chair for the plenary sessions of the Committee.
Finally, Dr Hill expressed her thanks on behalf of WHO to the
Committee, to WHO CCs, and to all the experts, institutions and professional
societies working in this area whose efforts provided vital support to WHO
programmes in global public health. She concluded by reminding participants
that Committee members acted in their personal capacities as experts and not
on behalf of their organizations or countries.
Dr David Wood then gave a brief overview of WHO Expert Committees
and of their important and greatly valued role in providing assistance to WHO
Member States. He noted that two Expert Committees were meeting during the
week – the Expert Committee on Specifications for Pharmaceutical Preparations
and this Expert Committee on Biological Standardization – and that the 63rd
Consultation on International Nonproprietary Names for Pharmaceutical
Substances was also taking place. Dr Wood then introduced the members of the
2016 Expert Committee on Biological Standardization, and highlighted the new
requirement initiated in 2015 that biographical summaries of all the members
must be posted for public review and comment prior to the meeting. All
biographical summaries had been posted and no comments had been received.
Dr Wood then outlined the organization of the meeting and the major issues
to be discussed. Declarations of Interests made by members of the Committee,
Temporary Advisers and participants were then presented.2 Following prior
Dr E. Griffiths (consulting); Dr P. Minor (public statements; research support); Professor S. Efstathiou

2
(research support); Dr J. Ferguson (investments); Dr R. Sheets (consulting); Dr J. Southern (consulting);

and Dr D. Williams (public statements).
2
Introduction
evaluation, WHO had concluded that none of the declarations made constituted
a significant conflict of interest, and that the individuals concerned would be
allowed to participate fully in the meeting.
Following participant introductions, the Committee adopted the
proposed agenda (WHO/BS/2016.2303 Add.1).
3
2. General
2.1 Current directions
2.1.1 Strategic directions in biological standardization: WHO priorities
Dr Wood reminded the Committee of the core activities of WHO and that its
Constitution required it ...to develop, establish and promote international standards
with respect to biological and pharmaceutical products. WHO had been doing
this for over 60 years through a programme which included the development
of global written standards, global measurement standards and International
Nonproprietary Names (INN). In the field of biological substances there were
now over 70 WHO written standards and 300 reference preparations, all of
which make a significant contribution to global public health. Indeed, the first
international biological reference preparation, for insulin, had been established
in 1925 under the auspices of the League of Nations. Current global public health
priorities include responding to public health emergencies of international
concern, access to biotherapeutics and the strengthening of regulatory systems
– the latter two being supported by two World Health Assembly resolutions:
one on biotherapeutic products (WHA67.21, 2014) and the other on
regulatory systems strengthening (WHA67.20, 2014). Resolution WHA67.21
requests WHO to support Member States in the regulation of biotherapeutic
products, including similar biotherapeutic products (SBPs). In particular, the
Resolution requested WHO to convene the Expert Committee on Biological
Standardization in order to update its 2009 Guidelines in this area, taking
into account technological advances in the characterization of biotherapeutic
products and considering national regulatory needs and capacities. WHO had
recently reported on progress in this area to its Executive Board and to the
sixty-ninth World Health Assembly in May 2016.
As part of this programme, the Committee had adopted key written
guidelines on the regulatory assessment of approved recombinant DNA (rDNA)

biotherapeutics, and a series of WHO implementation workshops had been
held over a period of 5 years on current WHO Guidelines for biotherapeutic
products and SBPs. These workshops had involved participants from over
50 countries. The workshops had included case studies, with the outcomes
published in the scientific literature. Dr Wood highlighted the continuing and
growing need for international measurement standards for the calibration
of bioassays, especially for biotherapeutics. He noted that the number of
INN applications for biological and biotherapeutic substances had increased
enormously in recent years and now accounted for over 50% of all such
applications. The nature of the standards required was also evolving and WHO
would need to adapt its standardization programme for biotherapeutics as these
gained market authorization through the biosimilars route. There was thus a
4
General
need to consider very carefully the potential use and extent of applicability of
such reference preparations.
In considering the workload in the biologicals field, Dr Wood raised
the issues of the increasingly packed agendas of meetings of the Committee, the
difficulties in minimizing scheduling conflicts in the two-track system and the
insufficient time allocated for discussing strategic issues. Various proposals
were being considered to optimize the use of Committee time during face-to-
face meetings, such as holding meetings biannually, and the introduction of
more pre-meeting technical discussion of proposals for reference preparations.
Discussion at the face-to-face meetings could then be reserved for the more
complex proposals or those that were precedent-setting. Increasing the
inputs from WHO CC networks (see sections 2.2.2 and 2.2.3 below) was also
under consideration. Currently, WHO CC networks exist for vaccines and for
blood products and in vitro diagnostics (IVDs) but not for biotherapeutics.
Dr Wood also added that the WHO Secretariat is currently understaffed but
that additional staffing was being sought through secondments. Linkages with
other Expert Committees and Expert Groups were also considered important,
for example in developing both biological and chemical reference preparations
for biotherapeutic medicines, along with efforts to raise the visibility of the
work carried out in the area of biological standardization.
The Committee thanked Dr Wood for his overview and expressed
support for the focus and priorities outlined. The Committee drew attention to
standardization needs in potential new areas of work, such as vaccine platforms,
companion diagnostics and cell therapies, and highlighted the need to consider
where such activities might fit into the WHO programme of work. Such
consideration should include a careful review of the scope of the Committee,
which currently includes vaccines, biotherapeutics, blood products and IVDs.
The Committee also agreed that there was a need to improve the visibility of
the WHO biological standardization programme and its key role in supporting
global public health. This should involve not only reporting on the work of the
Committee in the scientific media but also keeping medical prescribers and
practitioners aware of the WHO biological standardization programme and its
significance in assuring the quality, safety and efficacy of biological medicines.
Indeed, the Committee considered that improving visibility and explaining the
crucial need for biological standardization should start during the education of
medical and pharmaceutical science students.
2.1.2 Vaccines and biotherapeutics: recent and planned

activities in biological standardization
Dr Ivana Knezevic reported on activities relating to the standardization and
regulatory evaluation of vaccines and biotherapeutics and discussed several
5
strategic issues. During the period 2013–2016 seven measurement standards

had been established for biotherapeutics and 11 for vaccines.
Vaccine development work was being carefully monitored to ensure
that standardization needs were being met in a timely way. This was achieved
by ensuring links between the Committee, the WHO Strategic Advisory Group
of Experts (SAGE) on Immunization and the relatively new WHO Product
Development for Vaccines Advisory Committee. There had been significant
recent developments in the infectious diseases area, including: (a) regulatory
approval of the first malaria and dengue vaccines; (b) licensure in China of
an enterovirus 71 (EV71) vaccine; (c) the first Phase III trials of a respiratory
syncytial virus (RSV) vaccine in pregnant women and the elderly; and (d)
the continuing problems of Middle East respiratory syndrome coronavirus
(MERS-CoV) and Zika viruses (ZIKVs), and associated vaccine development
work. Seven written standards related to vaccine regulation and evaluation
were under development, with five being submitted to the Committee for
consideration in this meeting (see sections 3.5.2–3.5.6 below), one planned
for 2017 (see section 3.5.1 below) and another on RSV vaccines scheduled for
consideration by the Committee in 2018. The development of three new written
standards (for meningitis B vaccines, EV71 vaccines and hepatitis E vaccines)
was being considered.
Two written standards for biotherapeutic products and SBPs were
also under development – with guidelines on the evaluation of monoclonal
antibodies as SBPs being submitted to the Committee in this meeting (see
section 3.1.1 below), and guidelines on the regulatory expectations for post-
approval changes for biotherapeutic products expected to be submitted for
adoption in 2017. A revision of the 2009 WHO Guidelines on evaluation of
similar biotherapeutic products (SBPs) was also under consideration, and the
possibility of developing guidelines for cellular and gene therapies would be
discussed separately (see section 3.3 below).

Dr Knezevic reported that between 2010 and 2015 a number of WHO
implementation workshops on standards for biotherapeutics, including SBPs,
had been held (in Canada, China, Colombia, Ghana and the Republic of
Korea) and these had been considered to be extremely useful to both national
regulatory authorities (NRAs) and manufacturers. Thanks to the support of the
Ministry of Food and Drugs, Republic of Korea, three workshops had been
held in Seoul. In 2016, implementation workshops on standards for human
papillomavirus (HPV) had been held in China and Thailand, and on typhoid
conjugate vaccines in Indonesia – with plans in place for a further four
workshops during 2017–2018.
The work of WHO CCs was becoming increasingly important in
supporting the WHO biological standardization programme, and the third
6
General
meeting of the WHO network of collaborating centres on standardization

and regulatory evaluation of vaccines (see section 2.2.2 below) was held in the
Republic of Korea in July 2016.
Improvements in testing methods are considered to be crucial for the
effective regulation of biologicals. Efforts had previously been made to improve
neurovirulence safety tests used in the quality control of oral poliomyelitis
vaccine (OPV) and yellow fever (YF) vaccines, and the potency tests used
for the lot release of diphtheria, tetanus and pertussis (DTP), rabies and YF
vaccines. However, more remains to be done in the case of rabies vaccines. One
particularly active area of investigation concerns the possibility of replacing the
NIH potency test for rabies vaccines – which is a test in animals established in
the 1950s – with alternative or complementary tests based on enzyme-linked
immunosorbent assays (ELISAs) or serological assays. These replacement
efforts are based both on a sound scientific basis and on increasing recognition
of the 3Rs concept (Replacement, Reduction, Refinement), which stresses the
importance of minimizing the use of animal testing in research.
A number of international and regional initiatives for promoting
regulatory convergence were also highlighted by Dr Knezevic. This is an area
in which many stakeholders recognized the unique role of WHO and had
called for action. Although there was a need to prioritize WHO activities, it
was considered important to assist Member States in making their national
regulatory requirements more consistent with each other by using WHO
standards as common ground. The International Conference of Drug Regulatory
Authorities (ICDRA) provides a forum to meet and discuss ways to strengthen
collaboration and Dr Knezevic reported that several relevant topics had been
selected for discussion at the 17th ICDRA to be held in Cape Town, South
Africa. These included SBPs, good regulatory practices, regulatory convergence
initiatives, maternal immunization and regulatory responses to shortages of
medicines and vaccines.
The Committee thanked Dr Knezevic for her overview and supported
the proposed initiatives. It considered that EV71 was of major regional
significance and the joint effort of the National Institutes for Food and Drug
Control (NIFDC) and the National Institute for Biological Standards and
Control (NIBSC) in developing the First International Standard for anti-EV71
serum (human) 3 had been timely and very successful. Consideration should
now be given to the development of a written standard for EV71 vaccine.
In: WHO Expert Committee on Biological Standardization: sixty-sixth report. Geneva: World Health
3
Organization; 2016: section 8.1.1 (WHO Technical Report Series, No. 999; http://apps.who.int/iris/bitstre
am/10665/208900/1/9789240695634-eng.pdf?ua=1, accessed 17 February 2017).
7
Developments in methodology were recognized as an area where WHO

had a comparative advantage. In that context, the Committee urged WHO
to continue to support activities aimed at replacing the NIH in vivo assay for
rabies vaccine potency with an alternative in vitro assay. In line with this
development, WHO was also strongly encouraged to consider deleting the
abnormal toxicity test (sometimes called General Safety or Innocuity test)
from all WHO Recommendations, Guidelines and other guidance documents
(see also section 2.3.1; EDQM). It was noted that this is particularly important
for global manufacturers who would greatly welcome WHO initiatives in
this direction.
The future development of cancer treatments based on combination
products consisting of active biological and chemical components (as well as
nanomedicines) was also highlighted as an area to be monitored. The Committee
agreed and noted that any future regulatory guidance for such products would
need both biological and pharmaceutical input and supported the idea of
increased involvement of academia in this area of work.
2.1.3 Blood products and in vitro diagnostics: recent and

planned activities in biological standardization
Dr Micha Nuebling reviewed recent WHO activities in blood products and in
vitro diagnostics, highlighting the following four activity areas: (a) antivenoms;
(b) ZIKV; (c) the Achilles project; and (d) standardization issues.
Dr Nuebling reported that each year there are more than 100 000
fatalities from snake-bites caused by difficulties in accessing antivenoms. WHO
had established a database showing the distribution of venomous snakes,
their respective antivenoms and the relevant manufacturers. However, the
manufacturing and distribution of antivenoms remains widely unregulated
and many products are of unknown quality. Moreover, in 2014 there had been
a production stop for “Fav-Afrique” by Sanofi Pasteur. In order to address

the need for antivenoms of assured quality, safety and efficacy, WHO had
initiated an assessment procedure for products intended for use in sub-Saharan
Africa. Applications for eight products, including polyvalent and monovalent
preparations, were received. For the assessment exercise, a panel of experts
(which included two regulators from Africa) was asked to evaluate the dossiers
at WHO headquarters in Geneva. Differences in the quality of the products were
obvious and plans are in place to supplement the evaluation of at least some
products by laboratory testing and/or inspection.
During the evaluation process, it became apparent that there was a
need to revise the 2008 WHO Guidelines for the production, control and
regulation of snake antivenom immunoglobulins. Expert revision was promptly
undertaken and, following public consultation, the revised guidelines were
8
General
now being presented to the Committee for adoption (see section 3.2.6 below).
Dr Nuebling indicated that if antivenom assessment is to become a sustainable
process in the future then funding issues will need to be addressed. Planned
next steps include the completion of first-round assessment based on laboratory
testing of promising products, funded by Médecins Sans Frontières (MSF), and
inspections of manufacturing sites. A side event on antivenoms took place at
the 2016 World Health Assembly, initiated by Costa Rica and supported by
17 Member States, with the intention of raising awareness and potential donor
interest in this important public health issue. Moreover, snake-bites will be
proposed for re-entry into the WHO List of Neglected Tropical Diseases and
the topic has been proposed as an agenda item of the WHO Executive Board.
There are also plans to promote snake-bite prevention and treatment initiatives,
including antivenom technology transfer, early case management and research.
Due to the 2015–2016 ZIKV epidemic in the Americas, and the
suspected association in Brazil between ZIKV infection in pregnant women and
microcephaly in newborns, WHO guidance on blood donations was urgently
required. Following close cooperation between EMP, the WHO Department of
Service Delivery Systems and the WHO Regional Office for the Americas, with
support from the WHO Blood Regulators Network (BRN), a WHO guidance
document on blood collection was published in February 2016. This document
advises on situations with and without active ZIKV transmission and evaluates
potential measures such as donor deferral, quarantining of blood components,
pathogen inactivation and testing. There was also an urgent need for reference
materials in this area, and projects to develop reference preparations for ZIKV
IVDs had therefore been initiated. This resulted in the rapid development of a
candidate WHO standard for ZIKV RNA to be considered for establishment
by the Committee this year (see section 7.1.1 below). Problems had been
encountered however in obtaining materials for serology standards.
Dr Nuebling also reported on the progress of the Achilles project for
improving access to safe blood products through local production and technology
transfer in blood establishments. Indonesia had been chosen as a pilot country
for the Achilles project. Project activities included the evaluation of IVDs related
to blood safety since many tests were available but their performance data were
not usually assessed. A workshop was held in Jakarta in August 2016 to review
current practices. Workshop participants included representatives of Ministry of
Health departments, the hepatitis surveillance programme, the Indonesian Red
Cross and the IVD laboratory of the Indonesian Centers for Disease Control.
The main aim was to evaluate > 40 different rapid diagnostic tests (RDTs) for
the detection of hepatitis B virus (HBV) and hepatitis C virus (HCV) infections.
RDTs are still used for blood screening for 15% of the national blood supply and
are the main surveillance tool in the country. Protocols for evaluating all HBV
and HCV RDTs were reviewed with an initial focus on sensitivity.
9
Aspects of blood regulation were also covered in WHO workshops

conducted in the WHO African Region and WHO Eastern Mediterranean
Region, resulting in respective regional strategies. Following a request from
Ghana, an in-country expert assessment of its new blood regulations was
undertaken involving WHO headquarters, the WHO Regional Office for Africa
and the BRN, using the new BRN Assessment Criteria for Blood Regulatory
Systems. This initial field test of the suitability of the assessment tool resulted
in the identification of gaps in the national blood regulatory system and the
generating of a number of recommendations (see section 3.2.1 below).
In the field of IVD standardization, concerns had been raised regarding
the suitability of the current First WHO International Standard for Anti-Rubella
Immunoglobulin (established in 1996) for use with more recent IVDs. There
appeared to be some uncertainty over whether the associated recommendation
of using 10 IU/ml as a general threshold for the decision to vaccinate young
women was still valid. Recent anti-rubella IVDs can show quite inconsistent
quantitative results in the low-titre range of anti-rubella antibodies, leading
to potentially discrepant decisions on the vaccination or re-vaccination of
individuals. A meeting was planned for June 2017 to review the current situation
and develop appropriate recommendations. Dr Nuebling concluded by reporting
that several new reference preparations in the area of blood products and IVDs
were to be considered for establishment by the Committee in 2016 (see sections
5.1.1–5.1.5 and sections 7.1.1–7.1.6 below) and a number of new projects
considered for endorsement.
The Committee thanked Dr Nuebling for his report and looked forward
to hearing of further progress at its next meeting.
2.2 Reports
2.2.1 Report from the WHO Blood Regulators Network
Dr Christian Schaerer reported on the activities of the BRN during the past year.
Following its face-to-face meeting during the previous Committee meeting in
October 2015, five teleconferences had been held. There had also been several
changes in BRN member representatives as well as a minor revision of the
BRN Terms of Reference. This now allowed “alternate” representatives to be
eligible candidates for the BRN Chair. Dr Schaerer himself had been elected
BRN Chair for a second term, until October 2017.
BRN activities during 2015–2016 had included discussion of the ongoing
revision of the WHO NRA assessment tools, and of the draft WHO Guidelines
on estimation of residual risk of HIV, HBV or HCV infections via cellular blood
components and plasma, proposed for adoption at this meeting (see section
3.2.4 below). Discussions were also held on the first results of clinical trials
10
General
performed in Guinea using convalescent plasma from Ebola virus (EBOV)

survivors, particularly in light of the infrastructure conditions under which
the trial was performed. During the above teleconferences, regular updates
on the Ebola disease situation and the activities under way to strengthen the
blood systems in Ebola-affected countries were provided. Revision of the BRN
position paper on Collection and use of convalescent plasma or serum as an
element in filovirus outbreak response was still pending. It was expected that data
and results on the antibody titres and other characteristics of the convalescent
plasmas used would soon become available and would be reflected in the revised
position paper.
Discussions also continued to be held on: (a) national decision-
making in relation to donor deferral for men who have sex with men; (b) the
Alliance of Blood Operators project on risk-based decision-making for blood
safety; and (c) the WHO expert assessment of antivenoms. One additional
item on the BRN agenda was the development of the draft WHO Guidelines
on management of blood and blood components as essential medicines, in
response to recommendations made at the 16th ICDRA. Following an extensive
development process, the final draft was now being proposed for adoption
at this meeting (see section 3.2.3 below). Other work products in 2015–2016
included: (a) BRN consultation with WHO on the development of emergency
guidance on yellow fever virus (YFV) vaccination and blood donation following
outbreaks in Angola, the Democratic Republic of the Congo and Uganda and;
(b) the sharing of information on the ZIKV situation, including discussion
of potential measures to protect the blood supply; and (c) discussing and
sharing information on the hepatitis E virus situation based on the results of a
standardized survey of BRN members.
BRN activities also included participation in the WHO in-country
assessment of the blood regulatory system in Ghana – representing the first real-
life application of the BRN assessment criteria to an external NRA assessment
process. The BRN was also involved in the WHO Regional training workshop
on regulatory systems for blood and blood products held in Benin in July 2016,
and would be assisting in a workshop on blood products at the upcoming 17th
ICDRA in South Africa.
The Committee thanked Dr Schaerer for his report and raised the issue
of BRN membership. While the Committee recognized the valuable support
provided by the BRN to WHO there was an impression that the composition
of the BRN did not take into account the experience of developing countries.
Regarding the criteria for BRN membership, Dr Schaerer pointed out that in
principle any country could join provided it met the conditions laid out in
the BRN Terms of Reference, and provided the example of Japan as a non-
founding member of the BRN. However, it was pointed out by the Committee
11
that the Terms of Reference make it clear that candidates should be experienced
blood regulating authorities. It was therefore suggested that the BRN develop
mechanisms for observing and taking into consideration the perspectives of
other countries. Dr Schaerer referred to already existing processes for such
interactions, such as the consultation phases for guidance documents.
It was also suggested that the BRN, in addition to responding to new
demands or emergencies, should also strengthen its support for the educating
and training of regulators in NRAs less experienced in blood regulation. It was
again emphasized that BRN members, despite limited capacities, were already
active in this endeavour through workshops in developing countries, for example
in Africa.
2.2.2 Report from the WHO network of collaborating centres on

standardization and regulatory evaluation of vaccines
Dr Yeowon Sohn and Dr Paul Stickings reported on the third meeting of the
network.4 One of the main objectives of this meeting was to discuss how the
network might best support the streamlining of the work of the Committee and
contribute to the priority-setting process for both written and measurement
standards. The possibility of setting up a working group to provide advice to WHO
on norms and standards priorities had been discussed. It was acknowledged
however that long-term planning was not easy, and had to reflect a range
of drivers both internal and external to WHO. The possibility of the network
playing a role in pre-reviewing selected proposals for measurement standards
intended for submission to the Committee was also considered. At present, the
review of large numbers of proposals for new or replacement measurement
standards consumed a significant amount of Committee time. It may be that
relevant expertise within the network could be utilized to support the Committee
by focusing on selected proposals and reducing some of the workload. It had
therefore been proposed that a “Core Expert Group” (CEG) from the network
be formed. As new measurement standards have strategic considerations (in
addition to scientific issues) it was felt that these should remain the responsibility
The network currently consists of the following eight WHOCCs: (a) National Institute for Biological
4
Standards and Control (NIBSC), Medicines and Healthcare products Regulatory Agency, England;
(b) Center for Biologics Evaluation and Research (CBER), Food and Drug Administration, the USA;
(c) Department of Bacterial Pathogenesis and Infection Control, National Institute of Infectious Diseases
(NIID), Japan; (d) Immunobiology and Biochemistry Group, Office of Laboratories & Scientific Services,
Therapeutic Goods Administration (TGA), Australia; (e) National Institute of Food and Drug Safety
Evaluation (NIFDS), Ministry of Food and Drug Safety (MFDS), Republic of Korea; (f ) Biologics and Genetic
Therapies Directorate (BGTD), Health Canada, Canada; (g) Institute for Biological Product Control of the
National Institutes for Food and Drug Control (NIFDC), China; and (h) Division of Virology, Paul-Ehrlich-
Institut (PEI), Germany.
12
General
of the Committee itself. However, proposals for replacement measurement

standards were likely to be more straightforward, with few strategic or scientific
issues, and could be the initial focus of a proposed CEG. It was further suggested
that this process might initially cover measurement standards for vaccines, with
potential expansion to other standards categories in the future. As any CEG pre-
review would be likely to affect the current timelines for submission of proposals
this would need to be taken into account. A potential process overview plan had
been suggested which would involve the submission of a one-page summary
for CEG review of a replacement standard. Where no issues were raised, the
standard would be recommended to the Committee for endorsement without
further review. Submissions for new standards would then be reviewed directly
by the Committee in the usual way.
The Committee thanked Dr Sohn and Dr Stickings for their report,
and welcomed the idea of creating a CEG to support its work. As a pilot study,
the Committee agreed that the CEG could pre-review selected measurement
standards in the vaccines area in order to streamline the current review process.
However, it was proposed and agreed that the CEG would include two or three
Committee members. The CEG would then prepare a one-page summary for
each of the recommended and rejected proposals for which there were no issues
to be resolved, for consideration by the Committee. A more detailed summary
would be needed where issues were identified that required further discussion
by the Committee. For measurement standards for which no issues were
noted, and for which the advice was simply to endorse or reject the proposal,
it was proposed that the one-page summary be considered by the Committee
in its closed meeting, without discussion in the open sessions. Proposals for
which issues had been identified would be considered in the usual way. The
public posting of all proposals (in the form of WHO/BS documents) would be
maintained. If successful, the pilot study could be expanded to cover all tracks,
and the Committee was informed that similar proposals were being explored
by the WHO network of collaborating centres for blood products and in vitro
diagnostics (see section 2.2.3 below).
2.2.3 Report from the WHO network of collaborating centres

for blood products and in vitro diagnostics
Since 2007 the network has held biennial 2-day meetings, with core delegates
attending in person and other stakeholders joining via teleconference facilities.
At the last meeting, held at the NIBSC in July 2015, it had been decided that
more frequent meetings were required and it was agreed that in the years
between the biennial meetings, two half-day WebEx meetings might be helpful.
Two such meetings were therefore held in 2016 to review progress on current
projects, discuss potential issues in advance of making presentations to the
13
Committee and to discuss new project areas. Dr Morris summarized the agenda
topics discussed, which in the case of two items – namely the dengue virus
(DENV) RNA reference reagent and the anti-CMV collaborative studies – had
led to a re‑evaluation of data and alteration of the final proposals made to
the Committee.
Comments from users of current standards, which are not normally
considered at routine meetings of the Committee, had also been discussed.
Dr Morris gave the example of the B19 DNA standard, explaining that some
users of one specific test (Roche DPX) had reported under-quantification of the
International Standard. On further investigation a single base-pair mismatch
was identified in the current standard compared to that of the manufacturer’s
primers/probes. Despite calls for the revision of the assigned IU of the standard,
it was agreed by the network, and by the delegates of the annual Standardisation
of Genome Amplification Techniques (group) (SoGAT) meeting, that the unit
should remain. Difficulties were also reported in reconstituting the current HCV
standard. This standard had been produced from a donation that on freeze-
thawing showed signs of insoluble lipid particles. These also appeared following
reconstitution. Users have now been advised to ensure thorough mixing
immediately prior to use to ensure that particulate matter is also incorporated
into the extraction.
Dr Morris reported that the WebEx meetings were considered overall
to have been a valuable addition to the network meeting calendar, allowing
pertinent issues to drive the agenda in a timely way while assisting in the
streamlining of the main process of standards approval by the Committee.
During discussion, it was suggested that where a project is considered
straightforward, a one-page summary could be submitted to the Committee for
its review and endorsement without requirement for a presentation of the project
itself. Where a project was considered more complex and issues remained to be
addressed, a two-page proposal could be submitted and the Committee could
also seek further clarification in the form of a presentation. However, it was also
made clear that the work of the WHO CCs could not supplant the role of the
Committee in establishing standards and endorsing new work. The Committee
thanked Dr Morris for her presentation and agreed that network meetings of the
kind outlined should be pursued and further explored as a means of facilitating
the work of the Committee.
2.3 Feedback from custodian laboratories

2.3.1 Developments and scientific issues highlighted by
custodians of WHO biological reference preparations
The Committee was informed of recent developments and issues identified by
the following custodians of WHO biological reference preparations.
14
General
National Institute for Biological Standards and Control

(NIBSC), Potters Bar, England
Dr Christian Schneider, the new Director of the NIBSC, presented a brief
overview of the scientific activities of the NIBSC and then drew attention to a
number of current issues in the field of biological standards.
He first drew attention to an issue raised at the last meeting of the
Committee, namely the increasing difficulty experienced in publishing data
from collaborative studies of candidate biological reference materials in peer-
reviewed journals in cases where those data had already been presented to the
Committee, incorporated into its report and published as part of the WHO
Technical Report Series. Dr Schneider considered that publishing data in
the public domain was crucial in both promoting the visibility of individual
studies and enhancing appreciation of the highly technical nature of biological
standardization in general. He urged WHO and the Committee to explore
ways of ensuring that data are published in both domains in order to convey
standardization messages and their importance to as broad an audience as
possible. Dr Schneider gave the example of the European Medicines Agency
(EMA) which had published a reflection paper on the management of clinical
risks deriving from insertional mutagenesis both in a peer-reviewed scientific
journal and on the EMA website. This had been achieved by the two parties
working together within specified timelines.
Dr Schneider then raised the issue of biotherapeutics, which he
considered had revolutionized modern medicine in numerous ways. The
field was now a major component of pharmaceutical business and was still
expanding on a global scale, especially with regard to SBPs. There was an
essential need to maintain the high standards of quality, safety and efficacy to
which these complex medicines are produced and licensed. Such standards
support innovation and facilitate global access to new medicines by supporting
the introduction of high-quality SBPs. As biotherapeutics are currently not
well covered by existing WHO CC networks there was a case to be made for
establishing a new dedicated group.
Dr Schneider also addressed the issue of alternative fill formats for
biological standards; in particular, alternatives to the existing formats for
freeze-dried preparations which had several recognized limitations. Flame-
sealed glass ampoules have been the preferred format for many years but these
are ill-suited for some materials – such as low fill volumes of DNA and viral
marker standards for nucleic acid amplification technique (NAT)-based assays.
Alternative approaches that do not use freeze-drying (as well as smaller freeze-
drying formats using 96-well plates with lyocluster caps) are being explored at
the NIBSC to deal with smaller volume formats and could be suitable for wider
use if successfully established. Work on seal integrity and moisture content was
ongoing but not straightforward.
15
The Committee noted Dr Schneider’s comments and thanked him for

an interesting, practical and forward-looking overview.
European Directorate for the Quality of Medicines &

HealthCare (EDQM), Strasbourg, France
Dr Karl-Heinz Buchheit outlined a number of recent EDQM activity areas in
biological standardization, including the European Pharmacopoeia, international
standards for antibiotics and the biological standardization programme, in
which WHO has Observer status.
The Committee was reminded that EDQM is the custodian centre for
international standards for antibiotics – a responsibility it assumed from the
NIBSC in 2006. Twenty-three international standards for old antibiotics are
available and there had been eight replacement batches established between 2006
and 2015. There was no requirement for any replacement batches this year and
no issues had been identified since the last meeting of the Committee. Eight of
the above antibiotics are on the WHO Model List of Essential Medicines and
the international standards are indispensable in the calibration of regional and
in‑house standards. Although there is no cost recovery for these standards
(and an average of 10–20 vials being distributed annually) the work is in line
with WHA67.25 and EDQM is happy to continue acting in this capacity.
Dr Buchheit then discussed a number of recent activities of the EDQM
biological standardization programme – the goal of which is to establish
European Pharmacopoeia biological reference preparations and to standardize
methods. The programme of work is established by a Steering Committee
and, whenever possible, collaboration and common projects were undertaken
with WHO and non-European partners. Current EDQM projects on human
and veterinary vaccines, blood-derived products and contaminants, and
biotechnology products were briefly described. Dr Buchheit reiterated that across
its programme of work, the development of alternatives to animal experiments

remained a major EDQM commitment in line with European Union directives.
WHO was once again strongly urged to consider the incorporation of the 3Rs
principles (Replacement, Reduction, Refinement) into its written standards and
other guidance where appropriate.
Specific projects of potential interest to the Committee included work
on rabies vaccines in which the intention was to replace the current in vivo NIH
potency test with an ELISA assay (see also section 2.1.2 above). A G-protein-
based ELISA for non-adjuvanted rabies vaccines for human use was reported to be
suitable for all European and some non-European vaccines. This project, initiated
by the European Partnership for Alternative Approaches to Animal Testing, had
now been handed over to the EDQM biological standardization programme
for further validation and inclusion into the European Pharmacopoeia. As the
16
General
assay is based on two commercial monoclonal antibodies a number of issues

concerning licensing agreements will need to be resolved with manufacturers.
Dr Buchheit reminded the Committee that one of the main outcomes
of a conference held in 2015 by the International Alliance for Biological
Standardization on the 3Rs concept was a formal request to WHO to initiate
steps to delete the abnormal toxicity test (sometimes called General Safety
or Innocuity test) from all WHO Recommendations, Guidelines and other
guidance documents. Recently, the European Pharmacopoeia Expert Group on
Vaccines agreed to the deletion of the abnormal toxicity test from all European
Pharmacopoeia monographs.
The Committee noted that deletion of the abnormal toxicity test from
the European Pharmacopoeia will provide additional impetus for the global
elimination of this test, although this may take some time to achieve. It was
pointed out that the US Code of Federal Register had already taken this step
and no longer required the General Safety test. Furthermore, the recently
revised WHO Recommendations for HPV vaccines adopted by the Committee
in 2015 note in small print that some countries no longer require this test.
Further consideration of this issue will be required by the Committee at its
future meetings.
The Committee also encouraged WHO to reflect on how to maintain
the IU for international standards currently calibrated using in vivo assays once
these assays are replaced by quite different in vitro assays. Examples would
include the assigned IU for the potency of diphtheria and tetanus vaccines. As
the number of laboratories performing in vivo assays is expected to be reduced,
technical competency in performing the in vivo test may be lost. Any associated
need for new reference standards should be considered in good time in light of
the time required to calibrate and establish international reference preparations.
Paul-Ehrlich-lnstitut (PEl), Langen, Germany

Dr Heidi Meyer reported on the activities of the two PEI WHO CCs and
discussed scientific issues of interest. There had also been two organizational
changes with Dr Dorothea Stahl succeeding Professor Rainer Seitz as Head of
the WHO CC for Quality Assurance of Blood Products and In Vitro Diagnostic
Devices, and Dr Meyer succeeding Dr Michal Pfleiderer as Head of the WHO
Collaborating Centre for the Standardization and Evaluation of Vaccines.
PEI WHO CC activities of interest included supporting the establishment
of WHO reference materials through participation in collaborative studies,
developing new standards – such as a new chikungunya virus (CHIKV)
international standard – and contributing to the development of WHO technical
documents. The PEI also provided support for the WHO prequalification
programme for IVDs and vaccines through dossier review, quality control
17
testing and participation in on-site inspections. The PEI was also involved in
activities related to: (a) the ongoing ZIKV public health emergency, including
the development of a WHO international standard for ZIKV RNA for NAT-
based assays (see section 7.1.1 below); (b) the WHO emergency use assessment
and listing (EUAL) procedure for ZIKV diagnostics; and (c) the safe use of
plasma-derived medicinal products (PDMPs). With regard to the latter, the
bloodborne transmission of ZIKV had raised the issue of PDMP safety and
PEI researchers therefore investigated the effectiveness of the most commonly
used virus-reduction/inactivation methodologies. Pasteurization and solvent/
detergent treatment both led to the rapid inactivation of ZIKV. Retentive virus
filtration was also shown to be effective, with ZIKV infectivity removed by filters
with nominal pore sizes ≤ 40 nm. Dr Meyer also reported that immunoglobulins
sourced in Europe or the USA had failed to neutralize ZIKV.
Dr Meyer acknowledged the excellent cooperation of its partners in the
two WHO CC networks in which it was involved, and highlighted PEI support
for the implementation of documents issued by the BRN. Topics highlighted for
future consideration by the Committee included: (a) considerations in advancing
the scientific evaluation of convalescent plasma collection and use beyond the
Ebola outbreak; (b) implementation of the assessment criteria for national blood
regulatory systems (which had already been applied in Ghana); (c) assessment
of the significance of hepatitis E virus for blood safety; (d) the quality and safety
standards for haematopoietic stem cells; and (e) the role of the Committee in
ensuring sustainable vaccine supply (see section 2.4.4 below). The PEI also
supported a proposal to facilitate the work of the Committee by streamlining the
process for evaluating new projects in the area of vaccines and biotherapeutics,
including the establishment of a process for project prioritization.
It was also reported that the PEI was providing support for two health-
system strengthening and capacity-building efforts. The first involved the
establishment of bilateral interactions with Ghana and Liberia to strengthen their
blood regulatory systems, which had included the development of a partnership

with the Ghanaian Food and Drugs Authority and the initiation of a twinning
project with Liberia. The second activity had been developed in light of the
recent Ebola epidemic and as a consequence of the G7 summit in June 2015.
The German Ministry of Health had agreed to fund two PEI projects to facilitate
access to medical countermeasures in low- and middle-income countries. These
focused on the availability, safety and quality of blood and blood products, and
on regulatory training in the evaluation and approval of clinical trials of vaccines
and biomedicines. Dedicated PEI personnel would be assigned to these projects,
which are expected to deliver short-, medium- and long-term outcomes.
The Committee thanked Dr Meyer for her presentation and raised the
question of the suitability of current ZIKV polymerase chain reaction (PCR)
assays for Asian strains. Dr Meyer referred this point to the later presentation on
18
General
the proposed establishment of a WHO international standard for ZIKV RNA for
NAT-based assays.
Center for Biologics Evaluation and Research (CBER), Silver Spring, MD, the USA
Dr Jay Epstein informed the Committee of recent and current developments at
CBER, including the appointment of Dr Peter Marks as Director following the
retirement of Dr Karen Midthun. Organizational restructuring had resulted in
three product line offices: the Office of Blood Research and Review, the Office of
Vaccines Research and Review and the Office of Tissues and Advanced Therapies.
Products regulated by the latter will include all purified and recombinant versions
of therapeutic proteins for use in haematology, as well as antivenins. CBER had
also been redesignated as a WHO CC for the period 2016–2020 and an umbrella
agreement had been put in place with WHO on vaccines, blood and blood
products, relevant IVDs and cell and tissue therapies. This agreement will support
the development of norms and standards, regulatory systems strengthening, the
WHO prequalification programme, product safety and vigilance, and regulatory
science in order to increase access to safe and effective biological products.
Dr Epstein outlined new United States Food and Drug Administration
NAT-testing policies with regards to blood donations (which are intended
to reduce the transfusion risk from ZIKV) as well as revised donor-deferral
criteria for HIV risk which permit men who have sex with men to donate
blood under certain circumstances. A Transfusion Transmissible Infections
Monitoring System had also been established to assess infectious disease risks
based on marker rates, incidence and risk factors in blood donors. Dr Epstein
also described CBER’s global involvement with other WHO CCs in the
standardization of plasma-derived coagulation factors, and in the development
of reference reagents and panels for the standardization of assays for transfusion-
transmitted infectious agents such as DENV, CHIKV, West Nile virus, HIV and
ZIKV, as well as for Babesia microti antibodies.
Recent CBER activities related to vaccines included participation in
an international study to evaluate new methods for measuring the potency
of inactivated influenza virus vaccines compared with the traditional single
radial immunodiffusion assay. Following completion, study results indicated
that despite the general feasibility of all alternative assays, additional studies
would be needed to identify the most promising ones for further development
and possible implementation. A second collaborative study was now under
development and scheduled for early 2017. An international study to evaluate
the inter-laboratory variability of influenza virus serological assays had also
been conducted. Coordinated by the Consortium for the Standardization of
Influenza Seroepidemiology, this study compared the inter-laboratory variability
of the standardized haemagglutination inhibition (HI) and microneutralization
19
(MN) assay protocols, with CBER performing the HI, MN and pseudotype
neutralization assays. Twenty laboratories provided data, and statistical analysis
of the results was ongoing at the NIBSC. Preparations for dealing with new
inactivated poliomyelitis vaccines (IPVs) based on Sabin vaccine strains instead
of conventional IPVs were also under way. Conventional IPV and Sabin IPV
differ antigenically and the current D-antigen potency ELISAs for conventional
IPV may not be suitable for Sabin IPV vaccines. A PATH-sponsored collaboration
with the Lankenau Institute of Medical Research to evaluate human monoclonal
antibodies in new D-antigen potency ELISAs has therefore been initiated. In
addition, CBER is supporting the development of new vaccine technologies
by participating in studies led by the Advanced Virus Detection Technologies
Interest Group (which is supported by the Parenteral Drug Association) to
develop next-generation sequencing controls for detecting adventitious viruses
in biological substances. CBER was responsible for preparing the large well-
characterized virus stocks to be used as controls in spiking studies for the
evaluation and standardization of next-generation sequencing platforms.
The Committee thanked Dr Epstein for his presentation and asked
whether the standardization materials being developed for next-generation
sequencing would be applicable across all available platforms. Dr Epstein
confirmed that this was the intention.
2.4 Cross-cutting activities of other WHO committees and groups

2.4.1 WHO Global Model Regulatory Framework for Medical
Devices including IVD medical devices
The Committee was informed of progress in the development of this framework
and invited to comment on the current draft (QAS/16.664/rev2), particularly
on those aspects concerning biological substances. This framework was
regarded as part of the response to resolution WHA67.20 on regulatory systems
strengthening for medical products in general, including medical devices and

diagnostics. A survey of the regulation of medical devices globally had shown
that only 58% of WHO Member States had any such regulations in place, with
low-income countries often having no regulations at all. The target audience of
the proposed document was countries with no or limited regulatory frameworks
in place but that wished to improve upon the situation. Since the implementation
of regulations would require political commitment, the document was also
specifically aimed at the legislative, executive and regulatory branches of
government. Two rounds of public consultations had resulted in over 900
comments being received, indicating a high level of engagement with the aims
of the document and the need for such guidance.
The Committee raised a number of points concerning the draft document,
including the need for a more appropriate classification of medical devices,
20
General
and highlighted the importance of emphasizing the responsibilities of device

manufacturers and users. In addition, medical devices that affected the safe and
effective use of a medicine (such as a companion diagnostic) or those used in the
manufacture of a biological substance (such as an apheresis machine or pathogen-
inactivation technology) require special considerations. These include careful
assessment of the level of risk of the device and of the potential need for clinical
studies of the related drug or biological substance as part of device validation.
The difference between IVDs and other medical devices was debated extensively,
with the view of some Committee members being that examples of each type
of device should be provided in the document. However, it was brought to
the Committee’s attention that country-specific requirements may contradict
each other and that too many examples would lead to confusion. Variations
in terminology were also discussed and it was pointed out that, in the field of
blood regulation, the terms validation and verification have precise meanings
and dictate specific courses of action. Instances where these terms had been
used interchangeably in the draft document should be addressed.
The Committee was informed that document QAS/16.664/rev2
was currently under consideration for endorsement by the WHO Expert
Committee on Specifications for Pharmaceutical Preparations (ECSPP). The
Committee asked that the above points and others raised during its discussions
be transmitted to the ECSPP and taken into consideration during its review.
The Committee was subsequently informed that the ECSPP had taken note of
the points raised and amendments to the text made. Following review of the
proposed amendments and further minor alterations, the Committee indicated
its agreement with the revised text. The Committee also took note of plans
to organize regional training workshops to promote the implementation of
this guidance.
The ECSPP adopted the amended WHO Global Model Regulatory
Framework for Medical Devices including IVD medical devices, and agreed that
this much needed guidance be annexed to its report.5
2.4.2 Report of a WHO ad hoc consultation on International

Nonproprietary Names for biological substances
The Committee was informed that in 2002 only 18% of all INN applications
were for biological substances. By 2016, this figure had risen to 50% of all
applications – with 45% of such applications being for monoclonal antibodies.
WHO Global Model Regulatory Framework for Medical Devices including IVD medical devices. In: WHO
5
Expert Committee on Specifications for Pharmaceutical Preparations: fifty-first report. Geneva: World
Health Organization; 2017: Annex 4 (WHO Technical Report Series, No. 1003).
21
This significantly increased level of activity in the biological/biotechnological

sector of the pharmaceutical industry was likely to be a pointer to future trends
in the work of the Committee.
The Committee was further informed that meetings have been held at
intervals since 2002 to address the general and specific aspects of the nomenclature
used for biological substances, which now included gene therapies, cell
therapies, monoclonal antibodies and SBPs. A review of this area: International
nonproprietary names (INN) for biological and biotechnological substances was
now available as a WHO working document (http://www.who.int/medicines/
services/inn/BioReview2016.pdf).
The objectives of the 2016 ad hoc consultation were to review the
current INN approach to naming specific classes of biological substances and to
discuss whether existing policies and established nomenclature were applicable
to emerging biological medicines. Recommendations arising from this ad hoc
consultation would be considered at the 63rd INN Consultation. Areas of
discussion included cell and advanced biotherapies, vaccine-like substances,
monoclonal antibodies and therapeutic proteins – the latter including fused and
conjugated substances as well as glycosylated biotherapeutics and monoclonal
antibodies. There are currently three sub-schemes for the nomenclature of
advanced therapies: (a) two-word names for existing gene therapies; (b) one-word
names for cell therapies; and (c) two-word names for genetically modified cells.
It was further noted that vaccines had generally been excluded from
INN assignment and that the Committee had a system in place for assigning
international names to prophylactic vaccines for infectious diseases. Immunization
communities were familiar with the international names so assigned and this was
an important consideration since the major use of prophylactic vaccines was by
public health bodies. The INN programme had also not assigned INN to defined
recombinant proteins used as active substances in vaccines but could do so on
request. Substances for anticancer immunotherapy (so-called cancer vaccines)

could be handled within existing INN policies, and INN could also be assigned
to engineered live viruses and bacteria.
2.4.3 Monographs on capreomycin sulfate and capreomycin

for injection in The International Pharmacopoeia
Antibiotics produced by fermentation often consist of complex mixtures of
structurally related components with different activities. Microbiological methods
have historically been used to quantify the total activity of these mixtures. As
knowledge of their structure and composition increased, a transition from
microbiological to physicochemical assays became possible. The latter assays are
considered to be more discriminative and easier to perform in quality control
22
General
laboratories. Microbiological assays, on the other hand, measure the total activity
of antibiotics, integrating all moieties that contribute to the antibiotic effect.
At the 2009 meeting of the ECSPP the decision was taken to replace
microbiological assays, where possible and appropriate, by physicochemical
methods (in particular, chromatographic methods) in The International
Pharmacopoeia. The Committee was informed that this transition from
microbiological to physicochemical assays had been largely completed for single-
component antibiotics. However, for multicomponent compounds the use of
physicochemical methods often poses a challenge as the total antimicrobiological
response of these substances is not only a function of their concentration but
also of their composition.
Capreomycin is a mixture of four structurally related compounds
(capreomycin IA, IB, IIA and IIB) with different individual activities. In The
International Pharmacopoeia monograph on capreomycin sulfate the active
substance is defined on a mass basis, whereas capreomycin is assayed using
high-performance liquid chromatography which discriminates between IA,
IB, IIA and IIB. In 1969, the activities of these four main components were
determined by a manufacturer of capreomycin. The results indicated that there
was a significant difference between the activities of components IA and IB
and between I and II. The United States, Indian and Chinese pharmacopoeias
all have similar requirements regarding the composition of capreomycin
sulfate but in these pharmacopoeias the capreomycin content is determined
using microbiological methods. Although The International Pharmacopoeia
limits capreomycin II content to a maximum of 10%, the ratio between
capreomycin IA and IB is not defined. There is therefore at least a theoretical
possibility that capreomycin samples with significant differences in their IA
and IB concentrations comply with the requirements of The International
Pharmacopoeia, but may be determined as sub- or super-potent when analysed
according the United States, Indian or Chinese pharmacopoeias.
The issue had been referred to the Committee and to the ECSPP for
advice as to whether a revision of the current International Pharmacopoeia
monograph is recommended. Prior to such an amendment, it was proposed that
the secretariat of The International Pharmacopoeia should contact manufacturers
of capreomycin sulfate and powders for injection to request information
regarding the composition of their products – as determined for example by
the chromatographic methods described in The International Pharmacopoeia.
This information would then be evaluated by a group of experts to see whether
amendments to the monograph were appropriate. Such amendments could
include the provision of a specification for the ratio between capreomycin IA
and IB and/or the establishment of a correlation between the results of the
23
chromatographic and microbiological assays described in the above-mentioned

pharmacopoeias.
Following discussion and clarification of various points, the Committee
agreed that more information was needed before any decision could be made
on revising The International Pharmacopoeia.
2.4.4 Yellow fever vaccine shortages and outbreak response

The Committee was informed that a recent YF outbreak in Africa had greatly
increased the demand for YF vaccine, exhausting the global stockpile and
putting routine immunization in endemic regions at risk. There was now a
shortage of such vaccine which could worsen if additional immunization
campaigns were required on a large scale. The risk of YF is also increasing,
especially in metropolitan areas with growing human population densities and
urban environments that provide the mosquito vector, Aedes aegypti, with an
ideal habitat. Increased urbanization, in particular among poorer groups of the
population without access to a clean water supply, and increased international
travel, could also potentially lead to the increased spread of the disease.
Potential strategies to improve YF vaccine supply include extension
of vaccine shelf-life, increasing the number of available vaccine doses by
immunizing with a fractional volume (0.1 ml instead of 0.5 ml) as an emergency
response and, in the longer term, the development of more flexible production
technologies based on cell cultures.
In response to the serious YF outbreak and acute vaccine shortages, an
emergency immunization campaign was conducted in Kinshasa, Democratic
Republic of the Congo, and along the border with Angola in August 2016 using
the fractional volume given by the usual route. This campaign was conducted
following a recommendation from WHO, based on WHO SAGE advice on the
use of the fractional volume as a dose-sparing measure. It was recognized that the
recommendation constituted an off-label use of the vaccine and that a fractional
volume should be used only as an exceptional response where there was a large
disease outbreak and vaccine shortage – and not in routine immunization.
The WHO SAGE advice was based on limited clinical studies of
fractional dose administration. While the data support the recommendation,
important data gaps remained – such as on vaccination responses among
children and immunocompromised populations to fractional doses and on the
duration of immunity. The Committee was reminded that there were currently
only four WHO-prequalified YF vaccines (all live-attenuated) which differ in
their properties. As batch records for these vaccines have shown excess potency
(albeit with some variability) there was potential scope for reducing the dose
required to achieve an acceptable seroconversion threshold. However, the
available data are based mainly on product from one manufacturer and the
long-term duration of immunity beyond one year following a fractional dose
24
General
of vaccine is unknown. There was therefore a need to follow up on the recent

emergency campaign in Africa.
The Committee supported the use of fractional volumes of YF vaccine in
emergency situations. However, discussion highlighted the need for caution in
extrapolating data on one vaccine to all YF vaccines given the known differences
in the manufacturing and release titres of the different vaccines, and the need
for good supporting clinical data. The importance of assay standardization
when comparing products from different manufacturers was emphasized.
The Committee also noted the need to address relevant regulatory aspects,
which could include updating the vaccine potency recommendations given in
current WHO guidance.6 As a medium-term strategy to increase vaccine supply,
exploration of the introduction of an upper potency limit might be considered by
manufacturers and NRAs. It was recalled that previous Committee discussions
had taken place on an upper potency specification from a safety perspective but
such a specification had not been introduced. The Committee was informed that
the issues of YF vaccine supply and the use of fractional volumes and doses were
also due to be considered at a concurrent WHO SAGE meeting.
2.4.5 Vaccines for public health emergencies

The Committee was informed of a new WHO initiative – the Blueprint for
Research and Development: Responding to Public Health Emergencies of
International Concern (R&D Blueprint) – developed in light of lessons learnt
from the Ebola epidemic of 2014–2016 and similar previous public health
emergencies. Following a request from its Member States, WHO had convened
a broad coalition to develop the R&D Blueprint as a sustainable platform for
accelerating research and development “in epidemics or health emergency
situations where there are no, or insufficient, preventive, and curative solutions,
taking into account other relevant work streams within WHO”.7 The R&D
Blueprint was subsequently endorsed by the sixty-ninth World Health Assembly
in May 2016.
The overarching vision of the R&D Blueprint is that the research and
development response to public health emergencies of international concern
caused by emerging pathogens will be faster and more effective than ever before,
and that continuous efforts will be made to accelerate the results of research
6
Recommendations to assure the quality, safety and efficacy of live attenuated yellow fever vaccines. In:
WHO Expert Committee on Biological Standardization: sixty-first report. Geneva: World Health Organization;
2013: Annex 5 (WHO Technical Report Series, No. 978).
7
2014 Ebola virus disease outbreak and follow-up to the Special Session of the Executive Board on Ebola.
Sixty-eighth World Health Assembly, Geneva, 18–26 May 2015. Geneva: World Health Organization; 2015.
Agenda item 16.1 (http://apps.who.int/gb/ebwha/pdf_files/WHA68/A68_ACONF5-en.pdf, accessed 19
February 2017).
25
adapted to the scientific, logistic and social challenges specific to epidemics. Four
principles underpin the project: (a) an inclusive process with a clear mandate and
milestones; (b) building on the efforts of others; (c) a collaborative effort with
affected Member States at the core of public health emergency responses; and
(d) a process driven by scientific knowledge. The R&D Blueprint aims to reduce
the time between the declaration of an international public health emergency
and the availability of effective tests, vaccines and medicines that can be used to
save lives and avert a crisis.
The R&D Blueprint has three work streams: (a) improving coordination
and fostering an enabling environment – for example, by building effective
governance and coordination frameworks; (b) accelerating the research and
development processes – for example, by assessing epidemic threats and
defining priority pathogens – and then in a second step by developing R&D
roadmaps to accelerate evaluation of diagnostics, therapeutics and vaccines;
and (c) developing new norms and standards adapted to the epidemic context
– for example, supporting expansion of capacity to implement adequate study
designs, developing guidance and tools to frame collaborations and exchanges,
and anticipating the evidence needed to inform regulatory review and policy
development. With regard to research strategy aspects, the Committee was
informed that a recent call for platform technologies for vaccine production,
medicines and diagnostics had been made by WHO. Forty-five proposals
were subsequently submitted, of which six were selected for presentation to
potential funders.
Priority pathogens and diseases listed in 2016 were: Crimean Congo
haemorrhagic fever, filovirus disease (EBOV and Marburg virus diseases), Lassa
fever, highly pathogenic emerging coronaviruses relevant to humans (MERS-
CoV and severe acute respiratory syndrome), Nipah, Rift Valley fever and
ZIKV. Two other diseases listed as serious are chikungunya and severe fever
with thrombocytopenia syndrome. The list will be reviewed annually or when
a new disease emerges. It was intended that R&D roadmaps for all priority
pathogens would be developed to guide the R&D response to large-scale public
health challenges, with such a roadmap having been developed for MERS-
CoV as a test case. The R&D Blueprint was also intended to link together other
international efforts, such as The Coalition for Epidemic Product Innovation
established following the Annual Meeting of the World Economic Forum in
Davos in January 2016.
The Committee welcomed this WHO initiative and noted that
considerable efforts were already under way in the area of norms and standards.
Guidelines on the quality, safety and efficacy of Ebola vaccines were being
developed (see section 3.5.6 below) – the principles of which could inform the
evaluation of similar vaccines against other priority pathogens. In addition, the
WHO CCs at NIBSC and PEI had developed international reference reagents
26
General
for pathogens with epidemic potential, including standards for ZIKV RNA and
for EBOV antigen and antibodies. However, the Committee also agreed that
more remained to be done and there were still many challenges to be resolved.
One problem was that of obtaining clinical material from priority pathogens
with which to produce international reference materials. Although efforts were
now under way to facilitate the sharing of clinical material, sending biological
materials across national and state borders, especially convalescent sera and
infectious materials, was often problematic. The Committee recommended
that WHO should play a role in facilitating procedures and ensuring that the
pathways for providing material to the relevant WHO CC were made clear.
The Committee also raised the possibility of using blood donations
as a source material for research. Despite challenges in ensuring informed
consent, this had been shown to be a viable option – for example, for following
virus evolution using next-generation sequencing technologies or for better
understanding epidemics with respect to pre-symptomatic patients via screening
programmes. The possibility of using material from deferred blood donations
to prepare reagents might also be considered.
The question was raised as to whether the use of convalescent plasma
as a treatment option could realistically benefit from a platform approach given
that in the case of the Ebola epidemic several issues could not be addressed due
to a lack of time and infrastructure. Such issues included determining the ideal
time point after infection or during convalescence to procure plasma with high
neutralizing titres.
27
3. International Recommendations, Guidelines and
other matters related to the manufacture, quality
control and evaluation of biological substances
All WHO Recommendations, Guidelines and guidance documents adopted at
the meeting are included in Annex 1, which provides an updated listing of all
current WHO Recommendations, Guidelines and other documents related to
the manufacture, quality control and evaluation of biological substances used
in medicine.
3.1 Biotherapeutics other than blood products

3.1.1 Guidelines on evaluation of monoclonal antibodies
as similar biotherapeutic products (SBPs)
Monoclonal antibodies (mAbs) are a major class of rDNA technology-derived
biotherapeutic products that have achieved outstanding success in treating many
life-threatening and chronic diseases. Some of these targeted therapy products
are ranked in the top-10 lists of annual global pharmaceutical revenue successes.
As patents and data-protection measures on mAb products have expired, or are
nearing expiry, considerable attention has turned towards producing SBPs (also
termed “biosimilars”) based upon the approved mAb innovator products with a
view to making these products more affordable and globally accessible.
WHO Guidelines on evaluation of similar biotherapeutic products
(SBPs) were adopted by the Committee in 2009 and have served well as a basis for
setting national requirements for SBPs. These Guidelines provide the scientific
principles, including the stepwise approach, for evaluating similarity between
an SBP and its reference biotherapeutic product. High similarity at the quality
level is regarded as a prerequisite for the use of a tailored nonclinical and clinical
data set for licensure. However, because of the structural complexity of mAbs,
comparability studies between a candidate biosimilar mAb and a reference
product mAb are challenging for both developers and regulators.
In 2014, the World Health Assembly adopted a resolution (WHA67.12)
on Access to biotherapeutic products, including similar biotherapeutic products,
and ensuring their quality, safety and efficacy. This resolution requested WHO
through its Expert Committee on Biological Standardization to update the 2009
Guidelines, taking into account technological advances in the characterization
of biotherapeutic products, and considering national regulatory needs and
capacities. A corresponding request was then made at the 16th ICDRA. In
response, WHO organized an informal consultation in 2015 to review the 2009
Guidelines and to consider ways of improving its guidance in this important
pharmaceutical sector. Participants included NRAs and national control
28
International Recommendations, Guidelines and other matters
laboratories (NCLs) from 26 countries in the six WHO regions, together with
developed and developing country manufacturers’ associations and individual
manufacturers. It was concluded there was no need to revise the overarching
2009 WHO Guidelines since the evaluation principles described still applied
and were valuable in facilitating the convergence of SBP requirements globally.
However, it was also agreed that, because of their complexity, there was a need for
additional WHO guidance on the evaluation of biosimilar mAbs.
Consequently, class-specific guidance on special considerations for the
evaluation of mAbs developed as SBPs was prepared and subjected to global
public consultation. The resulting draft Guidelines (WHO/BS/2016.2290)
covered rDNA-derived biosimilar mAbs, as well as mAb-derived fragments and
Fc fusion proteins, used in the treatment of human diseases. These Guidelines
are intended to be read in conjunction with the existing WHO Guidelines on
the quality, safety and efficacy of biotherapeutic protein products prepared
by recombinant DNA technology and the WHO Guidelines on evaluation of
similar biotherapeutic products (SBPs), and are intended to complement existing
relevant regulatory documents from other bodies.
The Committee reviewed the small number of comments that had
been received during a final round of public consultation. After agreeing
upon a number of further amendments to improve the clarity of the text, the
Committee recommended that the WHO Guidelines be adopted and annexed
to its report (Annex 2).
3.2 Blood products and related substances

3.2.1 Blood regulation activities
The Committee was informed of the outcome of a WHO BRN in-country
assessment of the Ghana Food and Drugs Authority (FDA) performed on
21–22 July 2016 using the BRN NRA assessment criteria compared against a
WHO Regional Office for Africa assessment tool. Ghana has a population of
25 million people and a national blood policy has been in place since 2006, with
the National Blood Service (NBS) Ghana being the responsible entity for blood
donation and transfusion. In 2015, 155 250 blood units were collected; with
54% of the units collected by the NBS coming from voluntary non-remunerated
blood donations, compared with 23% of private blood facility collections. More
than 150 blood facilities exist in Ghana (with a possible total number in excess
of 400) and all but one are hospital based.
The Ghana FDA regulates food, medical devices, medicines and
biologicals (including blood and blood components), with blood regulation
having been incorporated into this portfolio only recently. Blood products and
PDMPs have been covered to date by the biologicals regulations, whereas the
regulation of blood itself remains at an early interim stage. Due to the small
29
workforce currently dedicated to blood at the Ghana FDA (two full-time staff)
synergies achieved through interaction with other relevant Ghana FDA units
are imperative. In order to understand the status quo, the Ghana FDA had
previously visited more than 50 blood facilities across the country.
The in-country assessment identified areas for improvement and led to
several recommendations. These included making good use of already existing
structures for pharmaceutical regulation and generating synergies with additional
departments, such as those responsible for enforcement. However, blood-
specific training is also necessary, especially in relation to facility inspection and
haemovigilance. For blood-screening IVDs, a reliance on approvals made by
other NRAs or via WHO prequalification should be considered. If IVD testing
is to be performed at the Ghana FDA then the use of regional specimens for
test panels is recommended, rather than simply repeating the tests performed
by manufacturers.
The assessment report, and associated recommendations, were well
received by the Ghana FDA and were used during negotiations with the Ghana
Ministry of Health for further support. Both training (potentially in the form
of twinning projects) and funding will be essential in supporting the currently
understaffed but highly motivated team at Ghana FDA. WHO plans to follow up
with a reassessment after a certain time, but this was dependent upon funding.
Other countries that had expressed an interest in undergoing an external WHO
BRN assessment were Kenya and Zambia.
The Committee noted the assessment report and requested to be kept
informed of future assessments and their outcomes.
3.2.2 Africa Society for Blood Transfusion

The Committee was informed that at present only 20 of 46 WHO African Region
countries include PDMPs in their Essential Medicines List, with almost all
products currently being imported from abroad. Whole blood or fresh frozen
plasma is mostly used in place of PDMPs, which both increases the risk of
transfusion-transmitted infections and prolongs treatment. Only South Africa
manufactures products through the fractionation of plasma collected in the
country, and also exports surplus products to neighbouring countries.
During a presentation made to the Committee on behalf of the Africa
Society for Blood Transfusion (ASfBT), an action plan was outlined for possible
adoption by WHO as a set of concrete next steps to further advance the WHO
Achilles project on improving access to safe blood products through local
production and technology transfer in blood establishments. As proposed by
the ASfBT, this action plan for Southern Africa aimed to increase the plasma
supply and to improve access to PDMPs in Angola, Botswana, Lesotho, Malawi,
Mozambique, Namibia, South Africa, Swaziland, Zambia and Zimbabwe.
30
Although not distributed to the Committee, a specific proposal to WHO

for support in this activity had been submitted jointly by the ASfBT, National
Bioproducts Institute of South Africa, South African National Blood Service,
Western Province Blood Transfusion Service, Medicines Control Council of
South Africa and the International Plasma Fractionation Association. The
essential elements of the proposed action plan were:
■■ that WHO should encourage countries in Africa to participate in the
ASfBT Step-Wise Accreditation Programme, as a measurable means
of helping countries to achieve standards consistent with WHO good
manufacturing practice (GMP) standards for blood establishments;
■■ that the WHO Guidelines now being prepared on the management
of blood and blood components as essential medicines should be
used by the ASfBT to help educate and train blood establishment
personnel and blood regulators in Africa;
■■ that WHO should provide the ASfBT with full access to its training
materials on the quality production of blood;
■■ that WHO should routinely consider the ASfBT in global and
regional meetings in which the agenda impacts on blood safety and
availability in Africa.
The Committee was asked to provide a statement endorsing all of these
initiatives as actions appropriately designed to help meet the long-term goals
of self-sufficiency and regional collaboration towards ensuring the availability of
plasma derivatives as essential medicines.
The Committee acknowledged that this ASfBT initiative was clearly
constructive and accorded very well with the objectives of the Achilles project.
However, it was not clear what it was that was specifically to be endorsed,
as normally there would be a specific plan proposed, which included cost
calculations. In addition, the element of accreditation had previously been
discussed by the Committee and was considered to be controversial, particularly
because of the different levels to be evaluated. Nevertheless, a consensus was
reached on the following statement of support for this initiative:
The WHO Expert Committee on Biological Standardization agrees that
there is a need for WHO to establish an Action Plan to assure continued
progress of the Achilles Project. The Committee finds that the elements
of the ASfBT proposed Action Plan to define next steps for the Achilles
Project are well aligned with the goals of that project and encourages WHO
to adopt the plan subject to its normal review of external cooperative
engagements. The Committee encourages WHO to support development
of, and engagement with, similar initiatives in other WHO regions.
31
3.2.3 Guidelines on management of blood and blood

components as essential medicines
The development history of these WHO Guidelines was briefly summarized
for the Committee. Following its initial drafting, early consultations with
BRN members and subsequently with several members of the Committee
were held. After consolidation of the comments received, a revised draft
(WHO/BS/2016.2285) was released for public consultation between July and
September 2016. Final consolidation of the document was then conducted via
teleconferences and also at the BRN face-to-face meeting in October 2016.
The topics covered in the proposed Guidelines were presented and
the list of organizations and individuals who had submitted comments was
provided. Of the 236 comments received, almost 50% had been submitted by
the European Blood Alliance, with other comments being received from a range
of different parties. Some of these comments reflected perceived differences
between the contents of the Guidelines and blood regulation practices in Europe.
Clarification was given that the Guidelines were primarily intended to support
countries lacking a highly developed blood regulatory system.
A number of comments received during the public consultation had also
led to changes in the text. For example, the inclusion of explicit reference to the
principles of biomedical ethics, and an exclusive focus on the requirements for
blood and blood components only, and not for the plasma collected in excess of
transfusion needs and used as a source material for fractionation. As the latter
is not considered an essential medicine, such considerations were perceived to
lie outside the scope of the Guidelines. Other comments were not accepted, for
example in relation to making a recommendation that NRAs should be tasked
with monitoring the supply of blood and dealing with shortages.
During discussion, the distinction between current good manufacturing
practice (cGMP) as used for small-molecule drugs and good preparation practice
(GPP) as used for blood components was raised. As use of the latter term in
this context was in line with other relevant resources and with European Union
terminology, its addition to the Guidelines was accepted. Other discussion topics
raised included the requirement for obtaining specific consent for research, and
the possibility of using accreditation and monitoring systems instead of NRAs
for blood systems. With respect to the latter, it was noted that governmental
NRAs are considered essential for the oversight of blood establishments,
whereas accreditation, despite its potential in enhancing the quality of blood
establishments, was generally voluntary and should only be viewed as a
supplemental system. Although there was also some discussion of the concept
of national self-sufficiency margins defined in terms of annual blood units per
population size, it was decided that all issues pertaining to blood supply lie
outside the scope of the Guidelines.
32
Following incorporation of the comments made the Committee

recommended that the WHO Guidelines be adopted and annexed to its report
(Annex 3).
3.2.4 Guidelines on estimation of residual risk of HIV, HBV or HCV

infections via cellular blood components and plasma
During the initial phase of the Achilles Project the development of WHO
guidance on residual risk estimation had been requested by blood transfusion
services in a number of low- and middle-income countries, and a proposal to
develop such guidance had subsequently been endorsed by the Committee in
2012. The goal of these WHO Guidelines is to define the impact of screening
algorithms on blood safety in order to be able to analyse the cost–benefit
characteristics of different testing algorithms, primarily in low- and middle-
income countries. The Guidelines should also allow for risk estimations
to be made on less-detailed databases than are usually available in more
highly developed blood regulatory systems that employ computerized data-
management systems.
Following initial development of a first draft by an international working
group and subsequent additional inputs by experts in the field, the document
was presented and discussed at the 2015 meetings of the Committee and the
BRN. Further inputs were also obtained during workshops held in the WHO
African Region and the WHO Eastern Mediterranean Region, and during several
scientific conferences. A revised document (WHO/BS/2016.2283) was posted
on the WHO Biologicals website for public consultation between August and
September 2016.
The Committee was presented with the current version of the Guidelines
and informed of the comments received from a variety of stakeholders in
the blood field, including CBER, EDQM, the European Blood Alliance, the
International Plasma Fractionation Association, the International Society of
Blood Transfusion, PEI, the Plasma Protein Therapeutics Association and the
WHO IVD prequalification programme. In addition, the Recipient Epidemiology
and Donor Evaluation Study (REDS) III group had comprehensively commented
upon the proposed estimation model and provided comparisons with other
estimation methods. The comparison results confirmed that the model proposed
in the Guidelines can provide similar results to the other systems, especially in
settings in which less detailed data were available. However, where more detailed
data were available, some of the other methods would be more accurate. The
comments received led to a number of significant changes to the guideline text,
which were presented to the Committee. Such changes included the addition of
text on the risk of HBV transmission via cellular blood components arising from
occult HBV infection, a statement on the limitations of the approach taken and
provision of simplified formulae for calculation.
33
The Committee discussed and accepted the proposed revisions to the

document, and recommended that the WHO Guidelines be adopted and annexed
to its report (Annex 4).
3.2.5 WHO assessment of antivenoms

Snake-bites have a significant global impact on health with an estimated
5 million people bitten by snakes every year, resulting in around 100 000 deaths
and leaving 400 000 people permanently disabled or disfigured. Currently,
antivenoms are the only effective therapy. These immunoglobulin preparations
are manufactured from equine (or ovine) plasma and can be either monospecific
to individual snake species or polyspecific. Against a backdrop of regulatory
deficiencies in affected regions, falling numbers of antivenom producers,
increasing fragility of the remaining production systems and the 2014
announcement by Sanofi Pasteur of a production stop for “Fav-Afrique” due to
economic reasons, the inappropriate marketing of products is taking place.
In a 2015 editorial The Lancet highlighted snake-bite envenoming as a
neglected tropical disease and urged the world to increase antivenom production.
WHO took the lead by calling on manufacturers to propose antivenoms for
WHO assessment – the first round of which took place in 2016. This initiative
built on previous WHO efforts, such as the inclusion of antivenoms in its 2007
Model List of Essential Medicines, and adoption of the WHO Guidelines for
the production, control and regulation of snake antivenom immunoglobulins by
the Committee in 2008. In addition, WHO developed a database, summarizing
the distribution of snake species and listing corresponding antivenoms and
their manufacturers. Intended to allow, in the case of a snake-bite, for the rapid
identification of snakes endemic to that region and of the respective antivenoms
and their manufacturers, this database is outdated and requires revision.
The minimal data requirements for the 2016 WHO antivenom assessment
were based on the 2008 WHO Guidelines for the production, control and
regulation of snake antivenom immunoglobulins. These requirements ranged

from aspects of manufacturing and quality control, to nonclinical and clinical
data requirements, and post-marketing activities. The assessment was performed
by external experts and regulators from Australia, Brazil, Cuba, France, Kenya,
Nigeria and South Africa, as well as WHO staff members, based on a risk–
benefit evaluation, and was therefore different to the prequalification scheme
for procurement decisions. During the assessment it became clear that several
chapters of the 2008 WHO Guidelines on which the assessment requirements
had been based required urgent revision (see section 3.2.6 below).
In total, nine product applications were received, with one that targeted
Northern Africa not being further considered. The remaining eight products
came from manufacturers from Costa Rica, Egypt, India, Mexico, South
Africa, Spain and the United Kingdom. It was pointed out that although some
34
manufacturers appeared to be based in countries with sufficient regulatory

experience to handle market authorization without WHO assistance, the target
regions for which the products were proposed lacked the necessary regulatory
structures, and products may therefore be unregulated. The eight product
dossiers were highly variable in terms of depth of information provided and
on occasion included less than reliable data. Following assessment, questions
were sent to the manufacturers, with the responses received to date now
under evaluation. Initial desk review indicates that three to five products may
be suitable for listing by WHO. Next steps will include laboratory testing to
confirm product features, and possible inspection visits to manufacturers.
Inspection decisions will be based upon the respective regulatory environment,
for example, on whether prior inspection by a stringent regulatory authority has
already been conducted or not.
The Committee discussed the issues raised and acknowledged that
WHO capacity to deal with antivenoms and their assessment is constrained as
this activity area is significantly under-resourced.
3.2.6 Guidelines for the production, control and regulation

of snake antivenom immunoglobulins
The first version of these WHO Guidelines was established in 2008. However,
in 2016 it was proposed that, in light of changes in technology, the identification
of new snake species, taxonomic name changes and the need to ensure relevance
and accuracy, the document be reviewed and revised. A proposed revised
document (WHO/BS/2016.2300) was therefore presented to the Committee
which included the following major changes:
■■ updates to the lists of medically important snakes to reflect new
species discoveries and nomenclature changes;
■■ revision of methodologies for the production of venoms to ensure
traceability and quality control;
■■ specific mention of the need for national reference venom collections
that are independent of manufacturers;
■■ recommending of research into new adjuvants;
■■ updating of text relating to equine viruses;
■■ greater emphasis on controlling the health of donor animals prior to
and during bleeding sessions;
■■ a strong emphasis placed on the importance of addressing animal
welfare and associated ethical issues, including adoption of the 3Rs
concept and principles in relation to all animals used in antivenom
production and quality control processes;
35
■■ redrafted sections on quality control and preclinical testing, and

updating of recommendations on stability studies;
■■ inclusion of antivenomics as an additional preclinical testing
methodology that can supplement conventional approaches;
■■ updated section on clinical assessment, with expanded information
on the role of regulatory authorities in antivenom production.
The use of poorly produced low-quality medicines and the distribution
of products with inappropriate specificities for a given region are among the
major reasons why many snake-bite victims do not seek medical treatment with
antivenom. It is intended that the improvements that could be achieved through
application of the guidance set out in these WHO Guidelines would address this
issue and lead to both the increased use of antivenoms and improved outcomes.
The Committee discussed the issue of the under-utilization of automated
equine plasmapheresis and the basis for this. Although cost is an issue, some
manufacturers also believe that animals may not tolerate the longer duration of
this procedure. On the other hand, improved horse husbandry and increased
training of animal handlers in appropriate methods of control could potentially
solve this issue, and lead to improvements in husbandry, welfare and plasma
production. It was suggested that this topic could be addressed in the section
referring to the 3Rs concept and principles. Although the increased use of
automated plasmapheresis could prompt a rise in the cost of antivenoms, this
was considered to be unlikely given the expected productivity increases and
herd size reductions that might follow its implementation.
Improving the quality and safety of antivenoms by addressing current
production and quality control weaknesses in relation to the key starting materials
would be a crucial step in restoring confidence in antivenom immunotherapy,
particularly in resource-poor settings. It was hoped that these WHO Guidelines
would contribute to the dissemination and adoption of best practices in the
manufacture of snake antivenoms and hyperimmune plasma.

After making a number of minor changes to the text, the Committee
recommended that the revised WHO Guidelines be adopted and annexed to its
report (Annex 5).
3.3 Cellular and gene therapies

3.3.1 Regulation of cell therapy products
The Committee was provided with an outline of current activities in the
development and regulation of cell therapy products (CTPs). Despite being at
an early stage of development, the field was now very active worldwide, with
many products at different stages of clinical evaluation in a number of countries.
Technological breakthroughs and research advances had led to increasing
36
expectations that novel cell-based investigational products will become useful

new therapies. In addition to the regulatory oversight of clinical studies and
licensing, there was also a demand from medical researchers and manufacturers
in this area for proactive scientific and regulatory advice.
The need for WHO international guidelines on CTPs had already been
raised in various forums, such as the 16th ICDRA. ICDRA participants had
recommended the development of regulatory expertise for CTPs appropriate
to the specific nature of these products. As part of this, ICDRA urged WHO
to consider developing guidance on the manufacture, and nonclinical and
clinical development of CTPs, taking into account existing guidelines and in
collaboration with established regulatory authorities. In 2014, the World Health
Assembly resolution WHA67.20 on regulatory system strengthening also
recognized the need for increased support and guidance in strengthening the
capacity to regulate increasingly complex biological products, including somatic-
cell therapies. Furthermore, a proposal that WHO define the scientific and
regulatory considerations for CTPs as part of promoting their standardization and
regulatory convergence had emerged following several international conferences
organized by the International Alliance for Biological Standardization. Such
issues had also been considered at various international forums, including a Cell
Therapy Working Group initiated in 2011 by the International Pharmaceutical
Regulatory Forum to address issues related to this emerging product class. As
the principal technical agency in the area of biological standardization, WHO
is well placed to provide the leadership now required to address a range of key
needs in this area.
The Committee recognized that new cell-based medicinal products had
great potential in the treatment of various diseases, and that CTPs would become
important future public health interventions. Action was therefore needed to
promote global-level standardization of both technical and regulatory approaches
to these novel biotherapies. However, it was noted that the application of cell
therapy is generally undertaken in hospital settings where over-regulation at
this stage of development could potentially overburden those developing novel
products. The issue of stem cell tourism was also brought up as an issue requiring
raised awareness since there was little or no evidence of efficacy but numerous
safety concerns associated with such ventures.
It was pointed out that WHO involvement in the early stages of other
emerging technologies – for example, rDNA-derived biotherapeutics, including
mAbs – had in the past been helpful in their development and regulation.
Early guidelines and points to consider produced by the EMA, the United
States Food and Drug Administration and WHO on rDNA-derived products
had been instrumental in establishing expectations for their quality, safety and
efficacy. Although several of the organizations mentioned above were active in
developing guidelines on various aspects of cell therapies there was a need for
37
further harmonization at the global level. It was considered that WHO is in a

unique position to provide such guidance, especially to developing countries.
There was a clear consensus within the Committee that global
harmonization in the cell therapy field is needed and that WHO should become
engaged in this area. The Committee recommended that WHO collaborate
with the range of international groups active in cell therapy with the goal of
providing a common guideline. A focus should be placed on somatic and
not stem cell therapy and should include quality considerations. An agreed
definition of cell therapy would also be helpful, along with clarification of
whether genetically modified cells should be included or considered under
gene therapy. Harmonized definitions and terminology would be particularly
helpful for countries that are now setting their own national requirements. It
was however considered premature to consider measurement standards for
CTPs at this time, and the issue of WHO resources for engaging in the cell
therapy field should not be overlooked.
3.3.2 Reference preparations for gene therapy products

The Committee was provided with an outline of the current situation concerning
gene therapy, and their advice sought on proposed new work in this area,
including the need, if any, for international biological reference preparations. The
scope of the discussion was confined to reference preparations for gene therapy
products as the possibility of WHO developing written standards to guide
regulatory activities in this area would be the subject of a future Committee
discussion paper.
The technical aim of initiating work on such reference preparations –
which would be developed as WHO biological standards – would be to provide
global quality assurance tools to help regulatory authorities and public health
authorities make decisions on approving and monitoring gene therapy clinical
trials and products both now in high-income countries and, in the future, in
low- and middle-income countries. The long-term objective would then be
to help improve access to gene therapies of assured quality, safety and efficacy
by 2030.
The clinical development of gene therapies – which are at a more
advanced stage than cellular therapies – began over 20 years ago. Some of the
early clinical trials involved gene therapy vectors which caused leukaemia in
some immunodeficient children through unwanted insertional mutagenesis.
More recent trials, which used different vectors considered to carry less risk of
such adverse events, have led to more encouraging results and to a resurgence of
optimism about these products. The first retroviral gene therapy to treat the rare
inherited disease adenosine deaminase-severe combined immunodeficiency
(ADA-SCID) is now licensed in Europe. Tools to minimize future risk and
38
maximize potential benefits would be beneficial and would help realize the
public health benefits of this technology.
Reference preparations would potentially be used by product developers
and manufacturers of gene therapies to calibrate the gene dosage in clinical
trials and thus assist NRAs in making decisions on clinical trial (and eventually
product) approvals. It was also foreseen that hospital diagnostic laboratories
would require tools to help monitor gene therapy patients and the outcomes
of the therapies. The impact of standardization could be the assurance that the
same virus vector is applied under standardized conditions and/or that potential
integration into chromosomes is determined in a standardized way. The use of
reference preparations would also be expected to improve comparability between
clinical studies.
In 2002, WHO established a Clinical Gene Transfer Monitoring Group
which met in 2002 and 2003 and then reported to the Committee. At that time
there was no consensus on the most appropriate approach to standardizing
gene therapy vectors and as a result no WHO reference preparations have yet
been established. However, WHO reference preparations have been established
for genetic tests and the methodology proposed to develop standards for gene
therapy vectors would build upon experience with the genomic reference
materials. In 2005, the WHO INN Expert Group established a naming policy
for gene therapy products and has, to date, provided names for 24 candidate
products. Analysis of the applications being made to the WHO INN programme
provides a means of identifying trends in the field and the potential need for
biological standards. NIBSC has identified the standardization of gene copy
number for candidate products being tested in human clinical trials as a
sufficiently mature area of work to potentially benefit from the establishment
and maintenance of WHO biological standards. Based on a scientific workshop
convened by NIBSC in June 2016 a proposal has been made to develop a
lentiviral vector copy number standard (see section 6.1.1 below). Lentiviral
vectors are emerging as a platform technology for gene therapies. If other such
platform technologies are developed, and if WHO initiates work on reference
preparations, then WHO would need to consider establishing appropriate
additional biological standards for these other platform technologies. The
National Institute for Standards and Technologies (NIST) has also organized
workshops on standardization in the field of advanced therapy medicinal
products but there is no overlap with potential WHO projects on gene therapy.
The Committee agreed that the development of reference preparations
for gene therapy products should be explored further and considered the
proposed development of a lentivirus reference to be a good model exercise.
It also noted that the need for a corresponding written standard would be
considered at a future meeting.
39
3.4 In vitro diagnostics

3.4.1 Preparation of secondary reference materials for in vitro diagnostic
assays designed for infectious disease nucleic acid or antigen
detection: calibration to WHO International Standards
It is a principle of standardization that biological samples often cannot be fully
characterized by a physicochemical reference method. Moreover, biological
assays are heterogeneous and the lack of a reference method does not permit
their results to be expressed in absolute values according to the SI system.
Instead, WHO International Standards are defined that are expressed in arbitrary
International Units (IU). These are the highest order reference materials as
defined in ISO 17511: international convention calibrators, and WHO guidance
documents limit the use of WHO International Standards to the calibration of
other biological materials to minimize the need for their regular replacement.
The calibration of a secondary reference material is a complex process and the
effort involved in setting up such standards should not be underestimated. In
light of the current lack of published guidance on the production and evaluation
of secondary standards for use in IVD assays, WHO guidance has been developed
on the preparation and calibration of secondary reference materials against
WHO International Standards in this area, with a specific focus on in vitro
measurement procedures used for the diagnosis, detection and management of
infectious diseases.
It was expected that due to the complexity of secondary standard
preparation in this field such practical guidance would facilitate appropriate
standard design, manufacture and use, and would contribute to the global
harmonization and quality assurance of IVDs. The proposal to develop the
WHO guidance had been endorsed by the Committee in 2012. Following a series
of presentations at a number of SoGAT and WHO CC meetings, a draft text was
developed and sent out for comments. Based on the feedback received a second
draft was prepared and more widely circulated among relevant parties. Following
consideration of this document by the Committee in 2015, incorporation of the
points raised, and additional stakeholder discussions the proposed document
(WHO/BS/2016.2284) was submitted to the Committee for adoption.
The Committee was informed that the document was intended to provide
practical guidance on the preparation and calibration of secondary standards,
and would be of use to bodies that prepare and establish secondary standards,
IVD manufacturers, and providers of external quality assurance or proficiency
testing programmes, as well as to other laboratories using reference materials
for NAT-based and serological infectious disease assays. Due to their inherent
complexity – for example, inter-individual differences in antibody response and
inter-assay design differences – the scope of the document did not cover materials
40
used for antibody detection methods. The Committee was then provided with an
overview of the structure and content of the proposed document.
The Committee considered that the document as a whole provided
a better understanding of calibration than part B of the current WHO
Recommendations in this area.8 It was also agreed that the document should
be established as a stand-alone WHO manual instead of being integrated into
the current WHO Recommendations to allow for more rapid and flexible
adaptations to real-world changes than would otherwise be possible in the
context of a full revision of the latter.
The Committee recommended that the WHO manual be adopted and
annexed to its report (Annex 6). Additionally, the Committee also recommended
that part B of the current WHO Recommendations for the preparation,
characterization and establishment of international and other biological reference
standards be revised so that it referred to the new guidance.
3.4.2 WHO prequalification of in vitro diagnostic devices

The WHO prequalification programme for IVD devices aims to promote and
facilitate access to safe, appropriate and affordable IVDs of good quality in an
equitable manner. IVD prequalification is coordinated through the WHO
Prequalification (PQ) Team which undertakes the comprehensive assessment
of individual IVDs using a standardized procedure to determine if products
meet prequalification requirements. Currently the focus is placed on IVDs for
priority diseases (including HIV/AIDS, malaria, hepatitis B, hepatitis C, syphilis
and HPV). In future, and dependent upon funding, additional diseases such
as cholera, dengue, and possibly tuberculosis will be added to the scope of the
programme. The provision of an essential diagnostics list is also under discussion
which would serve in prioritizing fields of work.
The aim of IVD prequalification is to provide independent technical
information on the quality, safety and performance of IVDs, principally to
other United Nations agencies but also to WHO Member States and other
interested organizations such as MSF. In conjunction with other procurement
criteria WHO IVD prequalification status is used to guide IVD procurement.
IVD prequalification consists of three components, namely: (a) the review
of a product dossier; (b) a performance evaluation, including operational
characteristics; and (c) inspection of manufacturing site(s).
Recommendations for the preparation, characterization and establishment of international and other
8
biological reference standards (revised 2004). In: WHO Expert Committee on Biological Standardization:
fifty-fifth report. Geneva: World Health Organization; 2006: Annex 2 (WHO Technical Report Series, No. 932;
http://www.who.int/immunization_standards/vaccine_reference_preparations/TRS932Annex%202_
Inter%20_biol%20ef%20standards%20rev2004.pdf?ua=1, accessed 23 February 2017).
41
This prequalification process is complex and time intensive, and

in a best-case scenario may take 5 months from reception of the dossier to
issuing of eligibility – more typically, the procedure takes around one year.
Prequalification is then followed by post-qualification activities, including post-
marketing surveillance, change reporting and annual reporting. In addition to
the prequalification scheme, the WHO EUAL procedure had been established
to allow for a rapid response during an emergency situation – where demands
cannot be dealt with in the time frame of the usual prequalification procedure.
This slimmed-down approach was introduced during the Ebola crisis and
involves risk-based decision-making. The EUAL procedure lacks the inspection
component and has a shortened laboratory evaluation programme.
The WHO PQ Team also formulates technical specification documents,
and has prepared EUAL requirements and additional guidance documents.
As discussed at the previous meeting of the Committee, biological reference
materials will also be needed for IVD prequalification. These will include both
international standards and international reference panels for the purpose
of IVD verification and validation studies, performance evaluations and
post-marketing surveillance in countries. The materials most needed are for
quantitative and qualitative PCR – for example, for detecting HIV, HCV and
HPV. As appropriate, international reference preparations would need to cover
all major subtypes and mutants, and their commutability for use in whole
blood would be important. Moreover, reference standards are needed for rapid
diagnostic tests – for example, for detecting HIV Ab, HIV Ab/Ag, HIV/syphilis,
HCV Ab, HCV Ab/Ag, HBsAg, malaria Pf Ag, malaria Pf Pv Ag, malaria Pf Pan
Ag, EBOV (Zaire), ZIKV Ab, ZIKV Ag, and G6PD. Materials are also needed
for use in enzyme immunoassays to detect HIV and HCV. Previous experience
in this field has demonstrated the need for clear requirements and appropriate
standards.
The Committee asked how often IVD manufacturers are inspected and
were informed that inspections take place following every new application.
However, if a production line of interest had been inspected recently, an
inspection lasting 1 day (instead of 3 days) may be sufficient, depending on the
risk-based decision made. Further discussion included the observation that for
the EUAL procedure, post-marketing follow-up is currently not funded making
it difficult to perform, and the clarification that quality control laboratories can
apply to perform appropriate evaluations should they wish.
The Committee agreed that the relevant Technical Specification
Series documents being produced by the WHO PQ Team in the area of IVD
prequalification should be sent to the Committee for review and endorsement.
42
3.5 Vaccines and related substances

3.5.1 Revision of WHO Guidelines for the safe production
and quality control of inactivated poliomyelitis
vaccines manufactured from wild polioviruses
The Committee was reminded that the WHO Polio Eradication and Endgame
Strategic Plan 2013–2018 (PEESP) published by the Global Polio Eradication
Initiative sets the goal of achieving a polio-free world by 2018. One important
component of this goal is to minimize the risk of facility-associated reintroduction
of wild or attenuated oral poliomyelitis vaccine (OPV, Sabin) polioviruses
following eradication. In the third edition of the WHO Global Action Plan to
minimize poliovirus facility-associated risk after type-specific eradication of wild
polioviruses and sequential cessation of oral polio vaccine use (GAPIII) published
in May 2015, the safe handling and containment of poliovirus infectious and
potentially infectious materials has been aligned with the PEESP.
Noting these developments, the Committee at its previous meeting had
expressed its support for the revision of the above WHO Guidelines, which had
been published as a 2003 Addendum to the earlier WHO Recommendations.9
The Committee was informed that this revision process was now under way and
that a WHO working group meeting had been held in September 2016. Meeting
participants had agreed that the current Guidelines should be revised to provide
concise information on the safe production and quality control of poliomyelitis
vaccines, and should be aligned with other relevant WHO documents, particularly
GAPIII. The importance of the revised document was emphasized since many
WHO Member States will largely depend on it to ensure the safe production
of poliomyelitis vaccines. However, the full and immediate implementation of
GAPIII is difficult since it requires major changes in facilities and operating
procedures. Certification through the GAPIII Containment Certification Scheme
has three levels indicating increasing degrees of compliance, namely: Certificate
of participation; Interim certification; and Full certification. Working group
meeting participants had agreed that the issues identified should be resolved on
the basis of scientific evidence and risk-management principles.
Guidelines for the safe production and quality control of inactivated poliomyelitis vaccine manufactured
9
from wild polioviruses (Addendum, 2003, to the Recommendations for the production and quality
control of poliomyelitis vaccine (inactivated)). In: WHO Expert Committee on Biological Standardization:
fifty-third report. Geneva: World Health Organization; 2004: Annex 2 (WHO Technical Report Series,
No. 926; http://www.who.int/biologicals/publications/trs/areas/vaccines/polio/Annex%202%20(65-89)
TRS926Polio2003.pdf?ua=1, accessed 2 February 2016).
43
There was also agreement that the proposed scope of the revised
document should encompass: (a) containment of the manufacturing and quality
control facilities for inactivated poliomyelitis vaccine (IPV) derived from both
Sabin and wild-type poliovirus; and (b) a risk assessment of new safer strains of
poliovirus to be used in manufacturing or quality control testing. The routine
production of OPV using Sabin vaccine strains would be outside the scope of
the document because as long as it is being used for routine immunization
it does not have to be contained. When it is not in routine use it will not be
made, leaving only the problem of manufacturing a post-eradication OPV for
emergency response – assuming that the ability to make OPV is retained. It
should also be possible to store contained poliovirus, such as monovalent OPV,
outside GAPIII-compliant facilities which would greatly simplify future work
for manufacturers and other stakeholders. Several other issues had yet to be
resolved, such as the need for routine showering upon exit from the containment
area as required by GAPIII.
The Committee noted with interest the information provided and,
following discussion and clarification of a number of points, expressed its
support for the developments to date and looked forward to hearing of further
progress at its next meeting.
3.5.2 Guidelines on regulatory preparedness for provision of

marketing authorization of human pandemic influenza
vaccines in non-vaccine-producing countries
Strategies to shorten the time between the emergence of a human pandemic
influenza virus and the availability of safe and effective pandemic influenza
vaccines are among the highest priorities in global health security due to the
urgent public health need for vaccine in such situations. The WHO Guidelines on
regulatory preparedness for human pandemic influenza vaccines were adopted
by the Committee in 2007 and provide NRAs and vaccine manufacturers
with: (a) guidance on regulatory pathways for approving pandemic influenza
vaccines; (b) regulatory considerations to be taken into account in evaluating
the quality, safety and efficacy of candidate pandemic influenza vaccines; and
(c) guidance on effective post-marketing surveillance of pandemic influenza
vaccines. However, these Guidelines apply mainly to countries in which
influenza vaccine production takes place and where pandemic influenza
vaccines are likely to be given market authorization first in the event of a
pandemic. Consultation with stakeholders following the 2009 H1N1 influenza
pandemic identified regulatory delays due to a lack of regulatory preparedness
as a significant factor in delaying or preventing the deployment of pandemic
vaccine in non-vaccine-producing countries. This was especially the case for
44
vaccine destined for donation or deployment by United Nations agencies in

response to the pandemic emergency.
The Committee was informed that proposed WHO Guidelines had now
been drafted in response to requests from non-vaccine-producing countries
for guidance on the identification of appropriate regulatory approaches to the
marketing authorization of pandemic influenza vaccines, and on arrangements
for lot release of such vaccines during a public health emergency. The Guidelines
were developed in the context of the Pandemic Influenza Preparedness (PIP)
Framework’s Partnership Contribution Implementation Plan 2013–2016
for supporting regulatory capacity-building and strengthening pandemic
preparedness and response activities. The document had then undergone
considerable public consultation.
The resulting draft Guidelines (WHO/BS/2016.2289) provide guidance
to the NRAs of non-vaccine-producing countries on the regulatory oversight
of pandemic influenza vaccines for use in public health emergencies. The
document is aimed to assist such countries in preparing and putting in place, in
advance of a pandemic influenza emergency, a regulatory process for pandemic
influenza vaccines. Such a process should enable countries to expedite the
provision of marketing authorization and lot release of influenza vaccines in
response to a pandemic emergency. It is acknowledged that each country will
have national legislation and policies on the regulation of medicines, vaccines
and other health products. Some countries may also have regulations in place
on accepting donations of vaccines and ancillary products. The proposed
document is intended to provide additional and specific guidance to the NRAs
of non-vaccine-producing countries and emphasizes the need to put in place
appropriate decision-making processes that minimize duplication and make
much-needed vaccines available for use without unnecessary delay during
pandemic emergencies. In particular, it recommends that these processes be
established during the interpandemic phase.
The Committee reviewed the document WHO/BS/2016.2289 and made
a number of important changes. One table that had caused some difficulties
with respect to clarity was replaced by a simpler diagram and parts of the text,
including the title, were amended to improve the guidance given. Following these
changes the Committee recommended that the WHO Guidelines be adopted
and annexed to its report (Annex 7). The Committee additionally highlighted
the need for NRAs to interact closely with other stakeholders when addressing
issues related to pandemic influenza vaccines. It would also be important
that WHO follows up the adoption of these Guidelines with implementation
meetings in countries which most require assistance in this area, including
countries without a functional NRA.
45
3.5.3 Labelling information of inactivated influenza

vaccines for use in pregnant women
Enhancing the uptake of vaccines during pregnancy is an important element in
ongoing WHO efforts to improve maternal and child health. Levels of morbidity
and mortality due to seasonal influenza are considered to be substantial
worldwide, and pregnant women are especially vulnerable. Such women also have
an increased risk of severe disease and death from influenza and the infection
may also increase fetal complications. In 2012, the WHO position paper on
vaccines against influenza, endorsed by the WHO SAGE, recommended the
immunization of pregnant women with trivalent inactivated influenza vaccine
(IIV) at any stage of pregnancy. WHO SAGE also recommended that pregnant
women should have the highest priority in countries considering the initiation or
expansion of seasonal influenza vaccination programmes. In addition, the WHO
Global Advisory Committee on Vaccine Safety (GACVS), following careful
analysis of data worldwide, concluded that there was no evidence of adverse
pregnancy outcomes associated with the vaccination of pregnant women with
several inactivated viral or bacterial vaccines, including IIVs.
Nevertheless, the implementation of influenza immunization during
pregnancy remains suboptimal. One reason for this is the ongoing perceived
risk of administering influenza vaccine – or any vaccine – to this population
group, especially in view of the precautionary language used on some product
labels, which is open to misinterpretation. Such product labels also carry no
explicit indication for vaccine use during pregnancy. Furthermore, as pregnant
women are usually excluded from clinical studies during vaccine development,
licensure dossiers generally do not include information on their safety and
efficacy in this group. A recent survey had indicated that health-care providers
perceived package insert information as contradicting national immunization
recommendations, and that this affected their decision on whether or not to use
IIVs in pregnant women.
A proposal had therefore been made to develop an explanatory

addendum to the current WHO Recommendations for the production and
control of influenza vaccine (inactivated) 10 to clarify and interpret the labelling
information provided in the product insert of IIVs. This proposal had arisen
from the 2012 WHO SAGE recommendations and from the discussions at
several WHO consultations and meetings, including the previous meeting
of the Committee. The addendum was intended to be a clear statement
WHO Recommendations for the production and control of influenza vaccine (inactivated). In: WHO Expert
10
Committee on Biological Standardization: fifty-fourth report. Geneva: World Health Organization; 2005:
Annex 3 (WHO Technical Report Series, No. 927; http://apps.who.int/iris/bitstream/10665/43094/1/WHO_
TRS_927_eng.pdf, accessed 23 February 2017).
46
indicating that, on the basis of current evidence, the use of IIV in pregnant
women is not contraindicated. It was expected that this would in turn facilitate
maternal immunization programmes by raising awareness of the convergence
of regulatory positions on this matter. The Committee was informed that the
proposed document (WHO/BS/2016.2280) had been the subject of extensive
international public consultation with regulators, manufacturers, academia and
vaccine users, which had resulted in editorial changes to improve clarity.
Following review of the document, the Committee considered it to
be a suitable explanatory document without further revision, and therefore
recommended that the proposed addendum be adopted and annexed to its
report (Annex 8).
3.5.4 Revision of the WHO Guidelines on clinical evaluation

of vaccines: regulatory expectations
The current WHO Guidelines on clinical evaluation of vaccines: regulatory
expectations were adopted by the Committee in 200111 and since then have
provided guidance to NRAs and manufacturers, and informed WHO vaccine
prequalification activities. More than 20 vaccine-specific documents that
include a section on clinical evaluation have subsequently been adopted by the
Committee – all of which were intended to be read in conjunction with the 2001
WHO Guidelines. The Committee was informed, however, that the current text
was now outdated and did not address issues of vaccine development that have
emerged since that time, and there was also no dedicated safety section in the
current document. Furthermore, the nonclinical evaluation section had become
redundant with the subsequent adoption in 2003 of new WHO Guidelines
specifically addressing this aspect. As a result of these and other considerations,
the Guidelines had now been revised and updated to reflect the scientific and
regulatory experience gained from vaccine clinical development programmes
since 2001, and to take into account the content of clinical development
programmes, clinical trial designs, the interpretation of trial results and post-
licensing activities.
The main changes in this revision included more information on the
general principles of comparative immunogenicity studies, more details on
trial designs and analysis, a clarification of terminology (such as the distinction
between vaccine effectiveness and efficacy) and a new section on vaccine safety
evaluation. The structure of the document had also changed, with a number of
Guidelines on clinical evaluation of vaccines: regulatory expectations. In: WHO Expert Committee on
11
Biological Standardization: fifty-second report. Geneva: World Health Organization; 2004: Annex 1 (WHO
Technical Report Series, No. 924; http://www.who.int/biologicals/publications/trs/areas/vaccines/clinical_
evaluation/035-101.pdf?ua=1, accessed 23 February 2017).
47
methodological considerations having been incorporated into relevant sections

and subsections rather than being described in a separate section. In particular,
the revised document was not organized along the lines of the traditional pre-
licensure phases of drug clinical development (Phases I, II and III) given that
vaccine clinical development programmes are very variable and depend upon:
(a) what is already known about the antigen and adjuvant content; (b) the
epidemiological situation regarding the disease to be prevented; (c) any prior
vaccine efficacy studies; and (d) the existence of immune correlates of protection.
There was however a section on the different phases of pre-licensure development,
and one on post-licensure clinical evaluations.
The Committee was informed that the revised document had undergone
three rounds of public consultation, including discussion at consultation meetings,
and had been well received. The Committee reviewed the proposed document
(WHO/BS/2016.2287) and despite some initial concerns regarding a document
structure not based on the traditional phases of clinical development, recognized
the variable nature of modern clinical development programmes for vaccines
and the advantages of the new document structure. After making a number of
amendments to the text, including clarification of what was absolutely necessary
for licensing as opposed to what might be desirable to know, the Committee
recommended that the revised WHO Guidelines be adopted and annexed to
its report (Annex 9). The Committee once again emphasized the importance of
WHO implementation workshops in ensuring that NRAs, researchers, public
health officials and others involved in vaccine development were made aware of
and understood the new Guidelines and modern approaches to vaccine clinical
development programmes. Strong support was also expressed by representatives
of manufacturer associations for the running of such workshops.
3.5.5 Human challenge trials for vaccine development:

regulatory considerations
During the course of the revision process for the WHO Guidelines on clinical
evaluation of vaccines: regulatory expectations (see section 3.5.4 above) it
became clear that one subject area that was not covered by the original 2001
document, and which might be considered for addition during revision, was
human challenge trials. Human challenge trials are studies in which immunized
and non-immunized volunteers are intentionally challenged with an infectious
disease-causing organism to see whether immunization affords any protection
against the challenge strain. The challenge organism may be close to wild-type
and pathogenic (but adapted and/or attenuated from wild-type, with possibly
less or no pathogenicity) or genetically modified in some way. During subsequent
public consultations, it was agreed that since human challenge trials are not
required for licensing, any guidance on this matter should be developed as a
48
separate WHO guidance document to be read in conjunction with the updated

WHO Guidelines.
A document on human challenge trials (WHO/BS/2016.2288) was
therefore developed to highlight the standardization, regulatory and ethical
aspects of such trials, and to make clear that such studies, although not essential,
could nevertheless be helpful for many reasons during vaccine development. All
principles relating to the conduct of clinical evaluations of vaccines in humans
should apply, including appropriate study design, approval by an NRA and
ethical committees, and compliance with good clinical practice.
The Committee reviewed the document WHO/BS/2016.2288 and,
following discussion, requested that a number of clarifications be made to the
text and title. The Committee then recommended that the WHO guidance
document be adopted and annexed to its report (Annex 10). The Committee also
recommended that the topic of human challenge trials in clinical development
programmes for vaccines should be included in the WHO implementation
workshops on the revised WHO Guidelines on clinical evaluation of vaccines:
regulatory expectations.
3.5.6 Guidelines on the quality, safety and efficacy of Ebola vaccines

The unprecedented scale and severity of the Ebola virus disease (EVD) epidemic
in West Africa in 2014–2016 had led to calls for the urgent development and
licensing of an Ebola vaccine, and considerable efforts had been made towards
achieving this goal over a very short time. The Committee was reminded that as
part of ongoing WHO measures to support the development of Ebola vaccines,
draft WHO Guidelines had been prepared on the scientific and regulatory
considerations relating to their quality, safety and efficacy. In March 2015, WHO
convened an informal consultation, attended by scientific experts, regulatory
professionals and other stakeholders involved in Ebola vaccine development,
production, evaluation and licensure, to review the draft Guidelines and seek
consensus on key technical and regulatory issues. The draft document was
then revised in the light of comments made and underwent a round of public
consultation, resulting in a large number of further comments and suggestions.
At its 2015 meeting, the Committee reviewed the comments received and
agreed upon the proposed content, scope and style of the proposed document.
Following another round of consultation and discussion a revised draft
document (WHO/BS/2016.2279) was again subjected to public consultation,
resulting in only a few new comments and suggestions.
One major challenge in the development of the Ebola vaccine Guidelines
was that they had initially been prepared during the rapidly evolving epidemic
situation when the need for a vaccine was most urgent. However, the epidemic
was later brought under control by infection-control measures not involving
49
vaccines. The first drafts of the Guidelines were thus concerned with an epidemic
situation with a focus on accelerating product development and availability
during a public health emergency. The end of the large-scale EVD outbreak in
Africa and the post-emergency epidemiological situation made the assessment of
Ebola vaccine efficacy challenging due to the now sporadic nature of the disease.
The Committee was informed that document WHO/BS/2016.2279
provides scientific and regulatory guidance for NRAs and vaccine manufacturers
on the quality, and nonclinical and clinical aspects of Ebola vaccines relevant to
marketing authorization applications. The Guidelines particularly apply to Ebola
vaccines based on viral vectors as these are the vaccines currently at the most
advanced stages of development and for which no specific WHO guidance is
available. For the first time in any WHO Guidelines of this type, opportunities
to accelerate product development and availability during a public health
emergency are discussed.
Although the Committee recognized that ongoing clinical studies may
generate more data on vaccine performance, it agreed that any new data were
unlikely to significantly alter the technical guidance given. Furthermore, the
early availability of such WHO Guidelines in the public domain would aid the
evaluation of applications for clinical trials and for the licensure of Ebola vaccines
in the near future. The Guidelines may also be of benefit in the evaluation of
other viral vectored vaccines and could help guide future activities in outbreak
situations, especially those which constituted a Public Health Emergency of
International Concern. At the same time, specific outbreaks have unique features
and there would always be a need for flexibility of response.
The Committee reviewed the draft document WHO/BS/2016.2279
and after extensive discussion agreed that the guidance given on multivalent
Ebola vaccines and on the clinical evaluation of candidate vaccines using newer
clinical trial designs should be expanded. A revised document will therefore be
submitted to the Committee in 2017.
50
4. International reference materials –
biotherapeutics other than blood products
All reference materials established at the meeting are listed in Annex 11.
4.1 Proposed new projects and updates –

biotherapeutics other than blood products
4.1.1 Proposed Second WHO International Standard for
parathyroid hormone 1-34 (recombinant, human)
The Committee was informed that stocks of the First WHO International
Standard for parathyroid hormone 1-34 (recombinant, human) (rhPTH 1-34)
were likely to be exhausted within 2 years. The International Standard defines
the mass unit for rhPTH 1-34 and is essential for the correct potency labelling of
therapeutic rhPTH 1-34 products used to treat osteoporosis. The establishment
of a replacement International Standard should therefore commence as soon
as possible to ensure the continuous availability of this WHO International
Standard.
The standard is used by manufacturers, and regulatory, quality assurance
and academic laboratories for the calibration of therapeutic rhPTH 1-34 potency
assays, with an anticipated demand of 250–300 ampoules per year. Currently
there is one licensed rhPTH 1 34 product in the United States and Europe but
several SBPs have already been licensed in India and it is anticipated that the
expiry of the United States/European patent in 2018 will lead to the global
emergence of more.
It is expected that rhPTH 1-34 (expressed in Escherichia coli) will be
donated by the United States/European manufacturer and assigned a mass value
via a collaborative study using a primary calibrant approach. NIBSC intends to
dispatch ampouled material and the primary calibrant (previously mass value
assigned) to more than 10 collaborating laboratories for mass value assignment
using the reversed-phase high-performance liquid chromatography (RP-HPLC)
assay. Bioassays will also be performed where available to confirm bioactivity
according to a given study protocol. The accelerated thermal degradation of
samples will be analysed in order to predict long-term stability.
The Committee recognized the importance of replacing the current
International Standard and endorsed the proposal (WHO/BS/2016.2296 Rev.1)
to establish a Second International Standard for parathyroid hormone 1-34
(recombinant, human).
51
4.1.2 Proposed First WHO international standards for

vascular endothelial growth factor antagonists
Vascular endothelial growth factor-A (VEGF-A) is the most important angiogenic
growth factor of the VEGF family and has several isoforms (VEGF-A121,
VEGF-A145, VEGF-A165, VEGF-A183, VEGF-A189 and VEGF-A201).
VEGF-A165 is the major isoform required for angiogenesis through binding
to its signalling receptor, VEGF receptor-2 (VEGFR2), to promote the growth
and differentiation of the endothelial cells which compose the inner lining of
the vasculature. Over-expression of VEGF is associated with pathogenesis in
the form of neovascular disorders (including solid neoplasms and intraocular
diseases). Consequently, several VEGF antagonists are approved as therapeutics
in various cancers and/or eye disorders, including neovascular age-related
macular degeneration. Such antagonists include mAbs, for example, bevacizumab
(a humanized mAb) and ranibizumab (a Fab fragment derived from the same
parent antibody as bevacizumab). Both these mAbs bind with high affinity to
all VEGF-A isoforms and prevent the binding of VEGF specifically to VEGFR1
and VEGFR2. Bevacizumab is approved for various cancers while ranibizumab
is indicated for ocular diseases. Also approved are the Fc-fusion proteins
aflibercept and ziv-aflibercept which consist of the extracellular domain 2 of
VEGFR1 and the extracellular domain 3 of VEGFR2 fused to the Fc portion of
human immunoglobulin G1, which act as soluble decoy receptors to VEGF-A,
VEGF-B and placenta growth factor. Aflibercept and ziv-aflibercept are identical
in structure but differ in their purification process and formulation. The former
is indicated for treating ocular diseases whereas the latter is approved as an
anticancer drug.
As the patent expiry of originator products is imminent, biosimilar
VEGF antagonists (particularly bevacizumab and ranibizumab) are currently
under development worldwide. There is therefore an urgent need for standards
to control the performance of assays for bioactivity evaluations of these
products to ensure their clinical safety and efficacy. The intended users of such
standards would be manufacturers and regulatory authorities. Ideally, candidate
materials from both originator and SBP manufacturers would be included in
a collaborative study. The proposed NIBSC plan is to initiate activities with
bevacizumab or ranibizumab (depending on the availability of materials) and
to follow on with other molecules as available. Pilot lyophilizations would be
conducted to determine a suitable formulation for the molecule of interest.
Following definitive fills, the different candidate preparations would then be
included in an international collaborative study. The range of assays likely to be
used would include binding assays and bioassays involving both primary and
engineered cell lines.
The Committee regarded this as a longer term project with somewhat
unclear timelines, and it was envisaged that securing candidate materials for the
52
collaborative study might be an issue. Nevertheless, the Committee considered

this to be a worthwhile project and endorsed the proposal (WHO/BS/2016.2296
Rev.1) to develop standards in this area.
4.1.3 Proposed First WHO international standards (or reference

panels) for antibodies for use in immunogenicity
assessments of biotherapeutic products
Testing the potential immunogenicity of a biotherapeutic product is a regulatory
expectation for product approval. The bioanalytical testing strategy adopted for
antibody testing requires a multi-tiered approach which includes a screening
assay for the detection of binding antibodies, a confirmatory step and subsequent
analysis of positive samples for the presence of neutralizing antibodies. This
approach has been universally adopted and used in the immunogenicity
assessment of most biotherapeutic products. Testing for antibodies is however
a specialized field. Despite recent advances in new and evolving technologies,
antibody assays remain very challenging and antibody detection difficult, with
low-affinity antibodies having the potential to be missed. With the sole exception
of the First WHO Reference Panel for antibodies to erythropoietin (human),
there are currently no reference materials for antibody assays for biotherapeutic
products, and it is difficult for laboratories involved in this work to determine
the performance of their assays in the clinical setting.
There is thus a need for reference antibody standards/panels for use
as positive controls to standardize antibody testing across different assay
platforms and different laboratories, as this would provide a consistent basis for
antibody detection and measurement. This requirement was also emphasized at
a recent EMA workshop on immunogenicity assessment. The provision of such
antibody-positive controls would also facilitate the immunogenicity assessment
of emerging SBPs and thus promote wider access to safe and effective medicines,
while potentially also improving clinical decision-making to the benefit of
the patient.
This NIBSC-proposed collaborative project would aim to make available
either reference antibodies or antibody reference panels as positive control(s)
to standardize antibody testing across assay platforms and laboratories,
where possible for a range of biotherapeutic products. Purified human mAbs,
originating from isolated B-cells from patients treated with an innovator product
(such as adalimumab, infliximab, natalizumab, rituximab or interferon-β)
would be sourced and lyophilized. In some cases there was likely to be only one
antibody, whereas in others panels of antibodies with different characteristics
(for example, low or high affinity, or non-neutralizing/neutralizing) would
need to be procured and made available. Following pilot fills and selection of
an appropriate formulation, definitive fills would be conducted and multicentre
collaborative studies undertaken. Participants would be required to test the
53
antibody/panel in different assay platforms for evaluation of both binding

activity and neutralizing activity. It was anticipated that some patient samples
(subject to availability) would be included in the studies. The expected users
of the reference antibodies or antibody panels would be regulatory agencies,
industry, clinical laboratories, independent investigators and researchers in
academic and scientific organizations worldwide.
Despite appreciating that the timelines of the proposed project were
somewhat uncertain, the Committee recognized the importance of this work
and endorsed the proposal (WHO/BS/2016.2296 Rev.1) to develop international
standards and/or reference panels for antibodies for use in this area.
4.1.4 Proposed First WHO international standards (or reference reagents)

for monoclonal antibodies to the ErbB/HER receptor family
The ErbB/HER family of receptor tyrosine kinases (ErbB-1/HER-1, ErbB-2/
HER-2, ErbB-3/HER-3 and ErbB-4/HER-4) are structurally related to epidermal
growth factor receptors and are important therapeutic targets. Increased
expression of members of the ErbB/HER family occurs in a number of solid
tumours, and mAbs such as trastuzumab (directed to ErbB2/HER-2) and
cetuximab (targeting ErbB1/HER-1) are currently used as mono- or combined
therapies in a number of oncology indications. Their mechanism of action
and clinical efficacy rely upon the inhibition of ErbB signalling by preventing
ligand binding and promoting receptor internalization. In addition, Fc-effector
functions such as antibody-dependent cellular cytotoxicity often contribute to
their therapeutic activity.
Consisting of structurally complex molecules comprising several
functional domains, mAbs are sensitive to small quality differences that may
arise through changes in the manufacturing process. Bioassays that measure
the direct or indirect mAb-induced cytotoxic effects are generally used for
the potency testing and lot release of these products. To date, developers,
manufacturers and regulators have relied solely on the use of in-house qualified
reference standards and the reference clinical product during development
and manufacturing, and no higher-order of mAb standard for bioassays is
currently available.
Given that mAbs are the fastest-growing class of biotherapeutic product,
and with increasing numbers of SBPs now in development, there is a widely
recognized need for the global standardization of biotechnology products in this
field. Preliminary data from ongoing collaborative studies for the development
of WHO international standards and reference reagents for mAbs such as
rituximab are encouraging, and highlight the value of such reagents in assessing,
and ensuring the consistency of, bioassay performance by different stakeholders.
The proposal to develop WHO international standards or reference reagents for
54
use in the calibration, evaluation and validation of bioassays for measuring the
biological activity of mAbs to the ErbB/HER receptor family is aligned with
these recognized needs.
NIBSC therefore proposed an international multicentre collaborative
study to evaluate the suitability of candidate preparations to act as biological
standards/reference reagents in one or more relevant direct or indirect functional
assays. Depending on the specific mAb under study, these could include
bioassays for measuring cell binding, inhibition of receptor signalling, inhibition
of proliferation of target cells and/or antibody-dependent cytotoxicity.
Candidate reference materials would now be sought from an innovator
or SBP manufacturer. As this may be challenging, the first study to be carried
out will be determined by the ability to source suitable candidate material.
Other preparations may also be tested in parallel for comparison. It was
expected that an international collaborative study would be likely to take place
in 2017 and 2018.
The Committee agreed that there was a need for the standardization of
biotechnology products globally and endorsed the proposal (WHO/BS/2016.2296
Rev.1) to develop international standards and/or reference reagents for use in
this area. It noted that, ideally, the relevant reference materials should become
available as SBPs were being developed and licensed, and that there was now a
need for a catch-up programme of work.
55
5. International reference materials – blood
products and related substances
5.1 WHO International Standards and Reference Reagents –

blood products and related substances
5.1.1 Second WHO International Standard for ancrod
Ancrod and batroxobin are snake venom thrombin-like serine proteases
produced by Calloselasma rhodostoma and Bothrops atrox moojeni, respectively.
Both ancrod and batroxobin clot fibrinogen but lack the other functions
of thrombin so do not activate or cleave other proteins or present a target
for inhibitors. Because of the nature of the fibrin produced by ancrod and
batroxobin they physiologically promote fibrinogen depletion and have been
investigated as potential treatments to reduce clot formation under various
circumstances. Ancrod has been demonstrated to be effective in the treatment
of sudden sensorineural hearing loss and is also believed to have anticoagulant
properties, whereas batroxobin is primarily used clinically as a diagnostic reagent
for reptilase time. Despite being used in different clinical applications the two
materials are derived in similar ways and assayed in the same tests.
The Committee was informed that there was now a need to replace both
standards, and in the interests of efficiency it was decided that this would best
be achieved by conducting a combined international collaborative study which
included both the current First WHO International Standard for ancrod and the
current Second British Standard for batroxobin. This study had now assessed
the suitability of: (a) the proposed candidate material (NIBSC code 15/106) for
use as the Second WHO International Standard for ancrod; and (b) a proposed
reference reagent (NIBSC code 15/140) for use as the First WHO Reference
Reagent for batroxobin.
A total of 17 laboratories returned data, with each asked to perform four
independent assays. A range of assay methods were encouraged and suitable
protocols provided. Potency estimates for ancrod were made relative to the
current standard using parallel-line analysis and an unweighted geometric mean
potency of 53.9 IU/ampoule obtained, consistent with the expected potency of
55 IU/ampoule. Greater variability was observed in the batroxobin values, with
additional data from assays independent of the study confirming an apparent
10% higher potency when using plasma as the substrate compared to fibrinogen,
thus highlighting a small but significant discrepancy between the proposed
and current materials – potentially due to degradation of the current standard.
56
International reference materials – blood products and related substances
Overall, study data indicated a geometric mean value of 49.8 U/ampoule –

which it was proposed be rounded up to 50 U/ampoule.
The Committee considered the report of the study (WHO/BS/2016.2282)
and as no issues were raised recommended that the candidate material 15/106
be established as the Second WHO International Standard for ancrod with a
potency of 54 IU/ampoule.
5.1.2 First WHO Reference Reagent for batroxobin

The Committee considered the report of the study (WHO/BS/2016.2282) and
as no issues were raised recommended that the candidate material 15/140 be
established as the First WHO Reference Reagent for batroxobin with a potency
of 50 U/ampoule (see section 5.1.1 above).
5.1.3 Second WHO International Standard for blood

coagulation factor XI (plasma, human)
Blood coagulation factor XI (FXI) deficiency is generally mild and bleeding
is most often associated with surgery or trauma. Although deficiency is most
common among Ashkenazi Jews (with approximately 1 in 190 homozygous for
an FXI gene mutation and around 1 in 8 heterozygous) the condition has now
been identified in a wide variety of populations.
The First WHO International Standard for blood coagulation factor XI
(plasma, human) was established in 2005. This standard is used to aid the
diagnosis of FXI deficiency and to assign potency to licensed FXI concentrates
and to virus-inactivated plasma products, both of which are used as treatments.
As stock levels of this standard were now nearing depletion a replacement
standard was required. For continuity of the IU, the replacement standard
for FXI functional activity (FXI:C) was value assigned relative to the current
international standard. In addition, this study had also aimed to establish an
antigen value for FXI (FXI:Ag) for the same candidate material relative to the
local normal pooled plasma of participating laboratories.
The candidate material (NIBSC code 15/180) was frozen normal plasma
purchased from the the UK NHS Blood and Transplant. Following thawing at
37 °C and filling at room temperature (avoiding contact with glass) the material
was lyophilized into siliconized glass ampoules (with around 6000 ampoules
being available). For the value assignment of FXI:C, potency was assigned relative
to the current international standard, and the relationship with local normal
pooled plasma assessed. The assignment of FXI:Ag – to help assess patient
antigen value in addition to functional activity – was performed relative to local
normal plasma pools as no former value existed. For FXI:C value assignment,
29 laboratories from 11 countries participated, with 11 laboratories from eight
57
countries taking part in the assignment of FXI:Ag values. In all cases, the use
of four independent assays was requested, each with four or more dilutions in
duplicate. All data were analysed by parallel-line assay, for which only assays
valid for linearity were used (p > 0.01). Parallelism was assessed as the ratio of
standard and test slopes. Potency estimates were calculated using the unweighted
geometric mean.
Excellent inter-laboratory agreement was observed for the assignment
of FXI:C value (geometric coefficient of variation (GCV) = 1.8%) and no
discrepancy was observed between the IU and the plasma unit (0.71 versus
0.72 IU/ampoule). For FXI:Ag value assignment, there was also good
agreement between laboratories, albeit with wider variation due to differences
between plasma pools. A small method bias was also seen, which is unlikely to
be significant.
During discussion, the Committee noted that although enzyme activity
had remained stable during stability studies, the antigenicity had dropped.
Although this was the opposite of what would be expected, this may have
resulted from the different reference products used or to the higher variability of
the antigenicity test results. It was also commented that in future such validation
studies of international reference preparations should involve laboratories from
a wider selection of WHO regions.
and recommended that the candidate material 15/180 (6000 ampoules) be
established as the Second WHO International Standard for blood coagulation
factor XI (plasma, human) with an FXI:C value of 0.71 IU/ampoule, and an
FXI:Ag value of 0.78 IU/ampoule.
5.1.4 Fifth WHO International Standard for thromboplastin

(recombinant, human, plain)
International standards for thromboplastins are used to determine the

International Sensitivity Index (ISI) for commercial or local prothrombin
time (PT) test reagents and instruments, which is needed to determine the
international normalized ratio (INR) in patients receiving vitamin K antagonist.
Stocks of the current Fourth WHO International Standard for thromboplastin
(recombinant, human, plain) and the Fourth WHO International Standard
for thromboplastin (rabbit, plain) were now running low. An international
collaborative study involving 20 laboratories was therefore conducted to
assign ISI values to two candidate materials (NIBSC codes 14/001 and 15/001,
respectively) and to assess their overall suitability as replacement standards.
PT determination was performed by manual tilt tube technique on 10
different days using serum from two healthy subjects and six patients per day
– with the patients selected to have an INR between 1.5 and 4.5. In addition,
58
four freeze-dried control plasmas were used. To avoid bias in the testing system,
the order of testing was different from day to day. The mean ISIs obtained were
1.11 for candidate 14/001 (inter-laboratory coefficient of variation = 5.7%) and
1.21 for candidate 15/001 (inter-laboratory coefficient of variation = 4.6%).
Accelerated degradation studies were also performed to assess the stability of the
freeze-dried candidate materials. There was no significant change in PT following
storage of the candidate materials at 5 °C for 56 days. After storage at elevated
temperatures (31–42 °C) there was a slight but significant change in PT. Stability
after reconstitution was assessed at room temperature, over a time interval of
0.5–4 hours. Between 1 and 4 hours after reconstitution there was no change
in PT for candidate material 14/001 and a slight change (−1.5%) for candidate
material 15/001. Should these materials be established for use as international
standards, it was proposed that the provisional codings used above be replaced
by rTF/16 and RBT/16 respectively.
The results revealed significant variation in ISI results between
laboratories, indicating that further standardization of the manual tilt tube
technique is needed. It was intended that a document would be prepared
that provided full details of the technique for approval by the Scientific and
Standardization Committee of the International Society on Thrombosis and
Haemostasis (SSC/ISTH). The document was then to be submitted as an
approved reference method to the Joint Committee for Traceability in Laboratory
Medicine of the Bureau International des Poids et Mesures.
The Committee noted that an automated method would be a good way
of standardizing results as the manual tilt tube technique was apparently prone
to variability. Despite a number of potential methodological challenges, a strong
recommendation was made by the Committee to include automated methods
in future validation studies of reference preparations for thromboplastins.
Furthermore, given the observed instability of the candidate materials at
temperatures > 30 °C, it was proposed that written advice to ship them with
cooling packs to maintain temperatures at or below controlled room temperature
be added to the product Instructions for Use.
and recommended that the candidate material 14/001 (to be re-coded as rTF/16)
be established as the Fifth WHO International Standard for thromboplastin
(recombinant, human, plain) with an assigned value of 1.11 IU/ml.
5.1.5 Fifth WHO International Standard for thromboplastin (rabbit, plain)

The Committee considered the report of the study (WHO/BS/2016.2294) and
recommended that the candidate material 15/001 (to be re-coded as RBT/16)
be established as the Fifth WHO International Standard for thromboplastin
(rabbit, plain) with an assigned value of 1.21 IU/ml (see section 5.1.4 above).
59
5.2 Proposed new projects and updates – blood

products and related substances
blood coagulation factor V (plasma, human)
The measurement of blood coagulation factor V (FV) in human plasma is
important for the diagnosis of FV and combined FV/FVIII deficiency, as well as
for the quality control of virus-inactivated fresh frozen plasma (VIFFP). Users
include manufacturers of VIFFP and commercial plasma calibrators, clinical
laboratories and regulatory laboratories. Stocks of the First WHO International
Standard for blood coagulation factor V (plasma, human) were now running
low and a replacement was needed.
The proposed source material would be normal human plasma
donations purchased from the UK NHS Blood and Transplant. The material
will be pooled and buffered with HEPES (final concentration 40 mmol/l).
Aliquots of 1 ml pooled plasma filled into 5 ml DIN ampoules will yield 10 000
ampoules. For value assignment a multicentre international collaborative study
will be conducted, involving approximately 20 laboratories using one-stage
clotting assays relative to the current international standard and local normal
plasma pools. Raw data analysis will then be performed at NIBSC and a
consensus mean value calculated relative to the current international standard.
It was also planned to incorporate a “commutability” sample (for example,
virus‑inactivated plasma) into the study and to include the calibration of the
SSC/ISTH Secondary Coagulation Standard Lot #5.
The Committee endorsed the proposal (WHO/BS/2016.2297) to
develop a Second WHO International Standard for blood coagulation factor V
(plasma, human).
5.2.2 Proposed First WHO International Standard for blood

coagulation factor XII (plasma, human)

In vitro, blood coagulation factor XII (FXII) is involved in the contact phase of
the coagulation cascade and has also been shown to influence the complement
system. For many years, FXII was not considered to be important for haemostasis
because its deficiency is not associated with excessive bleeding – with some
patients paradoxically predisposing to thrombosis. Recent advances indicate
that in vivo FXII does have a role to play in thrombosis, especially in events that
are driven through interactions with endogenous FXII activators such as mast
cell heparin, RNA and platelet polyphosphate. In addition, a missense mutation
of the gene for FXII has been associated with hereditary angioedema, a life-
threatening disorder.
The lack of certified reference materials currently hinders both the
diagnosis of FXII deficiency and the comparison of research results in this
60
emerging area. In addition, the development of WHO international standards

encompassing both FXII:C and FXII:Ag was now essential to aid patient
diagnosis and the development of assay methods for FXII. It is intended that
normal plasma will be sourced from the UK NHS Blood and Transplant. Since
the candidate material has the same processing requirement as the recently
established Second WHO International Standard for blood coagulation
factor XI (plasma, human) (see section 5.1.3 above) it is further proposed to
assign FXII:C and FXII:Ag values as additional analyte values to the proposed
factor XI international standard. Stocks of the replacement international
standard for FXI are estimated to be sufficient for approximately 15 years
(6000 ampoules) – and are thus sufficient to allow for the assignment of FXII
values as additional analytes.
The Committee discussed whether potential variability in antigenicity,
which had been an issue in the FXI collaborative study, may also occur during
the evaluation of FXII antigenicity. However, the proposed use of local plasma
pools that overall contain > 20 000 donations should mitigate this risk. The
Committee then endorsed the proposal (WHO/BS/2016.2297) to develop a First
WHO International Standard for blood coagulation factor XII (plasma, human)
and to assign FXII:C and FXII:Ag values as additional analyte values.
5.2.3 Proposed assignment of factor XIII-B subunit (total and free) values
to the First WHO International Standard for factor XIII plasma
The First WHO International Standard for factor XIII plasma is currently used to
measure the potency (functional activity and antigen value) of blood coagulating
factor XIII (FXIII) in patient plasma during the diagnosis of FXIII deficiencies,
and also to evaluate FXIII therapeutic concentrates. FXIII circulates in plasma
as a heterotetramer of two A and two B subunits (FXIII-A2B2) in a 1:1 complex.
The active A subunit functions by cross-linking fibrin and stabilizing the fibrin
structure. The B subunit is a carrier protein without activity. However, FXIII-B is
in excess over FXIII-A with around 50% of total FXIII-B existing in complex with
FXIII-A and around 50% in free form. The half-life of FXIII-A depends upon
the amount of available FXIII-B. Congenital and acquired FXIII deficiencies are
severe bleeding disorders, and FXIII-B measurements are crucial in correctly
diagnosing and characterizing the type of FXIII deficiency. Furthermore, therapy
for FXIII-A deficiency with recombinant FXIII-A relies on available FXIII-B.
Free FXIII-B may also have other so far unknown functions.
It was proposed that an international multicentre study involving
manufacturers, clinical and research laboratories, and regulatory authorities be
conducted to calibrate both the total and free levels of the FXIII-B subunit in
the First WHO International Standard for factor XIII plasma relative to levels in
locally collected and pooled normal plasma. Calibration will be performed by
specific ELISAs with value assignment for the B subunit made in IUs. In order to
61
determine both total and free FXIII-B levels, mAb to the free FXIII-B subunit is
required and has recently been developed – although availability was restricted
to a single supplier. These antibodies have been made available free of charge for
the proposed value-assignment study but have also been commercialized and
will thereafter be available for purchase. This proposal could also be viewed as
part of the standardization of a companion diagnostic for informing the use of
a recombinant FXIII-A product (Novo XIII) which should not be administered
without assessing the patient’s FXIII-B subunit levels.
The Committee noted that approximately 150–200 ampoules are
currently dispatched each year and that this level of demand was likely to increase
with the new proposed value assignment. The Committee also emphasized that
FXIII deficiency can also be treated with plasma. The Committee then endorsed
the proposal (WHO/BS/2016.2297) to assign additional analyte potencies to the
current First WHO International Standard for factor XIII plasma.
5.2.4 Proposed First WHO International Standard for thrombin-

activatable fibrinolysis inhibitor (plasma, human)
Thrombin-activatable fibrinolysis inhibitor (TAFI – also proCPU or CPB2) is a
human plasma protein circulating as a zymogen. TAFI is activated by thrombin
to TAFIa, a basic carboxypeptidase that cleaves C-terminal lysines from partially
degraded fibrin, thereby inhibiting fibrinolysis. High and low levels of TAFI
and high and low levels of TAFIa have been shown to correlate with thrombotic
diseases and inflammation. However, the results from a large number of studies
investigating TAFI as a disease marker are confusing and often contradictory.
In addition, estimates of plasma TAFI concentration vary widely (4–15 µg/ml
in normal plasma) due in part to the use of different calibrators across the
range of commercial and in-house methods for measuring TAFIa in plasma
currently in use. ELISA-based assays show differences in detection between
some antibodies and do not allow for the measurement of TAFIa. Assays based
on functional activity measurement are available – either direct-activity-based
assays (complicated in plasma by carboxypeptidase N activity) or indirect
measurements that require the quantitative activation of TAFI. Both these assay
types present challenges, including lower specificity when using direct assays.
To obtain reliable results from these different assays, standardized
methods are required, and a common reference material for TAFI is now needed
to harmonize global measurements. Such a material would be intended for use
by manufacturers of commercial kits and by academic and clinical laboratories
investigating TAFI as a disease marker. It was therefore proposed that a study
be conducted involving all available antigen and activity assay methods. It was
recognized that such an approach may yield highly variable results, particularly
as ELISA-based assays measure concentration and not activity, and that the
62
value-assignment strategy would be challenging across the different assay

platforms. Initially, a single candidate material (pooled human plasma) will be
calibrated relative to local plasma pools. Eventually, the calibration would be
performed in IUs (1 IU = amount of TAFI in 1 ml plasma) where a specific
concentration would correlate with a given level of activity.
The Committee enquired whether the labelling of antigen content
of the candidate material was envisaged to be in ng, consensus ng or IUs. The
Committee was informed that the intention was to label the product in IUs both
for functional activity and antigen content. The Committee then endorsed the
proposal (WHO/BS/2016.2296 Rev.1) to develop a First WHO International
Standard for thrombin-activatable fibrinolysis inhibitor (plasma, human) with
assignment of both functional and antigenic units.
5.2.5 Proposed Third WHO International Standard

for anti-D immunoglobulin
The Committee was informed that stocks of the Second WHO International
Standard for anti D immunoglobulin were nearing exhaustion. This international
standard is used in potency assays of anti-D immunoglobulin products by
manufacturers of blood products, the Official Medicines Control Laboratory
(OMCL) network and blood transfusion laboratories.
In light of the frequent global demand for the current international
standard, and the need to ensure both continuity of supply and ongoing
comparability of test results across different laboratories, an international
collaborative study had been proposed to develop a replacement international
standard. The source material to be used would be purified anti-D immunoglobulin
donated by manufacturers. The candidate replacement preparation would then
be calibrated against the current international standard by end users.
The Committee endorsed the proposal (WHO/BS/2016.2297) to develop
a Third WHO International Standard for anti-D immunoglobulin.
5.2.6 Update on the use of WHO reference materials in assays to detect

activated blood coagulation factor XI in immunoglobulins
In 2010, a cluster of thrombotic events associated with intravenous
immunoglobulin administration was reported. Root-cause analysis indicated
that the thrombotic component was activated FXI (FXIa) which was a process-
related impurity. Regulators and manufacturers therefore began to develop and
validate procedures for removing FXIa and other procoagulant components
from intravenous immunoglobulins.
Two panels of immunoglobulins – each with three different levels
of procoagulant activity – were produced for use in the development of assay
methods and for the investigation of assay discrepancies. These panels were
63
made available to stakeholders such as immunoglobulin manufacturers, kit

manufacturers and regulators. In 2012, the First WHO Reference Reagent for
activated blood coagulation factor XI (human) was established, followed in
2014 by the First WHO International Standard for activated blood coagulation
factor XI.
There are a number of challenges associated with the three assay methods
currently used, with each exhibiting different degrees of sensitivity: (a) the FXIa
functional activity assay; (b) the non-activated partial thromboplastin time
(NAPTT); and (c) the thrombin generation assay. In addition, inter-laboratory
agreement continues to be poor, especially for NAPTT and the thrombin
generation assay. This is mainly due to the non-usage or misuse of the current
International Standard for FXIa. Moreover, there are local variations in assay
procedures and a lack of system-suitability samples for the validation of
methods, as well as product-specific matrix and/or excipient interference.
Moreover, the inconsistent handling of pre- and post-analytical variables (for
example, pre-dilution steps) and a lack of agreement on acceptable readouts
and on the statistical analysis of data add to the observed inconsistencies.
To address these challenges, a global working group consisting of CBER,
EDQM, NIBSC and the US Pharmacopeial Convention agreed to develop
and provide guidance on the harmonization of test methods, and to generate
appropriate reference materials. The first step will be to conduct a survey of
manufacturers and the OMCL network on current practices for measuring the
procoagulant activity of immunoglobulin products. It is then intended that 2–3
laboratories will further develop and refine assay procedures within the time
frame Q4–2016 to Q2–2017. This will then be followed by large collaborative
studies (from Q2–2017 to Q4–2018) with the aim of testing the agreed assay
procedures and identifying appropriate reference materials and/or system-
suitability samples for subsequent production.
The Committee enquired whether the only options available to measure

FXIa were functional assays or if, for example, antibodies with the ability to
distinguish FXI from FXIa were available. The Committee was informed that
although such antibodies do exist, reliable results in comparison to functional
assays have not yet been obtained. Nor are FXI antigen assays sufficiently sensitive
to be used as routine control assays in the immunoglobulin manufacturing
process. The Committee looked forward to receiving an update in due course
on the reference material selected for development and on whether this material
would be a suitable replacement for the current reference materials.
64
cellular and gene therapies

cellular and gene therapy
6.1.1 Proposed First WHO International Standard for
lentiviral vector copy number quantitation
The first retroviral gene therapy to treat the rare inherited disease ADA-SCID
is now licensed in Europe. The clinical trials pipeline for other rare inherited
diseases such as X-linked severe combined immunodeficiency (X-SCID),
Wiskott-Aldrich Syndrome (WAS), X-linked chronic granulomatous disease
(X-CGD), beta-thalassemia, adreno-leukodystrophy (ALD) and metachromatic
leukodystrophy (MLD) includes the use of lentiviral vectors. As the diseases
treated are rare, patients need to be followed up worldwide. The same platform
technology has also been used in clinical trials using genetically modified T-cells
such as chimeric antigen receptor T-cells (CAR-T-cells) to treat cancer. Here
there is potential for much larger numbers of patients. Regulatory authorities
generally require that a minimum gene copy number of approximately one copy
per cell should be used in gene transfer to ensure efficacy, with an upper limit of
four copies per cell to minimize the risk of toxicity.
The proposed development of a First WHO International Standard for
lentiviral vector copy number quantitation would build upon the experience
of NIBSC in developing genomic reference materials. Genomic DNA from
lentiviral vector transduced cells would be produced from a cell clone with a
single lentivirus integrant which would contain sequences essential for vector
function and which could be detected by various methods. It was expected that
candidate integration cell lines would be characterized by the end of 2017, with
the filling of candidate material and a collaborative study involving 10 or more
hospitals and manufacturers being initiated in early 2018. It was expected that
results would be presented to the Committee in October 2019.
The Committee heard that the anticipated use of this reference material
would be as a primary standard for the quantitation of lentiviral vector
integration copy number – an issue related to both product safety and efficacy.
It would also be used as a calibrant in assays and for the calibration of secondary
genomic DNA, plasmid or synthetic DNA standards. It was expected that
the reference material would be future-proof with regard to novel lentivirus
designs, and disease-independent. Potential users were likely to include vector
manufacturers (potency), manufacturers of genetically modified cells (potency
65
and safety), hospital diagnostic laboratories following patients (clinical efficacy)

and NCLs worldwide.
Following discussion, the Committee agreed that the availability
of a First WHO International Standard for lentiviral vector copy number
quantitation would be useful for the standardization of vector copy number – an
issue with both safety and efficacy implications for gene therapy – and endorsed
the proposal made (see also section 3.3.2 above).
66
7. International reference materials – in vitro diagnostics
7.1 WHO International Standards and Reference

Reagents – in vitro diagnostics
7.1.1 First WHO International Standard for Zika
virus RNA for NAT-based assays
ZIKV is a mosquito-borne flavivirus that was first identified in Rhesus monkeys
in the Ugandan Zika forest in 1947. The first human cases were reported
in Nigeria in 1954 and up until 2007 only 14 cases were known. In 2007, an
outbreak occurred in Yap Island, Micronesia, followed by a new outbreak in
French Polynesia in 2013 which was associated with an approximately 20-fold
increase in reported Guillain-Barré Syndrome cases. Between 2014 and 2016 a
widening outbreak occurred, first in the Pacific and then in central and South
America. Although associated with mainly mild symptoms, increased numbers
of Guillain-Barré Syndrome cases were again observed, along with cases of
microcephaly and other congenital abnormalities, and fetal loss. On 1 February
2016, WHO declared a Public Health Emergency of International Concern
following the recognition of clusters of microcephaly cases and other neurological
disorders associated with ZIKV infection.
In 2016, an international collaborative study had been conducted to
determine the suitability of a candidate material (PEI code 11468/16) for use as
an international standard. The candidate material consisted of a heat-inactivated,
lyophilized ZIKV preparation (4092 vials) formulated in a stabilizing neutral
solution, and was intended for dilution using a range of different types of sample
matrix. The virus strain used originated from a ZIKV-infected patient in 2013
from French Polynesia, closely related to the ZIKV strains currently circulating
in the Asia-Pacific region and in central and South America.
Test materials for the collaborative study were made available from
different outbreak locations, representing different subclades of all ZIKV
sequences available at the time of the study, and obtained from different body
fluids. In addition, two virus samples from previous outbreaks underwent
extensive laboratory cultivation and were also included in the panel.
Results were obtained by 24 laboratories in 11 countries using a range of
assays (in-house and commercial). Three sets of study materials were sent out,
with laboratories performing three independent assay runs. Harmonization of
results was observed for all biological reference materials and clinical samples
when potencies were expressed relative to 11468/16 with the candidate material
being detected by all assays evaluated in the study. However, such harmonization
67
for in vitro transcripts only occurred when expressed relative to other in vitro
transcripts, with no harmonization observed when the candidate material was
expressed relative to the universal in vitro transcript.
Discussion then took place of possible explanations for an observed
variability between two laboratories in the degree of harmonization of results
against the candidate material when material of African lineage was assayed.
The Committee considered that the discrepant results might have been due
to the different methods of virus inactivation used or to the matrix used for
lyophilization, rather than the African origin of the isolates per se. In relation to
the results observed for in vitro transcripts, it was noted that armored RNA was
not available at the time of the study, but may provide a solution for improving
harmonization in this regard. It was agreed that the continued fitness for
purpose of the candidate material should be monitored against new isolates as
they occur and that further data should be generated using NAT-based assays
of African isolates.
and recommended that the candidate material 11468/16 be established as the
First WHO International Standard for Zika virus RNA for NAT-based assays
with an assigned content of 50 000 000 IU/ml.
7.1.2 First WHO Reference Panel for Ebola virus VP40 antigen
Although the EBOV outbreak in West Africa has now ended, WHO preparedness
activities are continuing to ensure that all countries are operationally ready to
effectively and safely detect, investigate and report potential EVD cases, and to
mount effective responses. There is thus a need for the development of point-
of-care rapid tests for the specific and accurate EVD diagnosis, which do not
require complex laboratory equipment or highly trained staff.
An international collaborative study was therefore conducted to evaluate
the suitability of candidate recombinant protein preparations for use as WHO

reference materials. Three potential candidate materials were available: (a) a full-
length VP40 based upon EBOV Mayinga sequences and expressed in E. coli
(the preferred candidate); (b) a full-length native viral protein (VP40) based on
2014 Kissidougou-C15 and expressed in E. coli; and (c) EBOV virus-like particles
produced by co-transfection of three plasmids (pCMV3-codon optimized
VP40, pCMV3-codon optimized nucleoprotein and pCAGGS-codon-optimized
glycoprotein into HEK293T cells using polyethylenimine. A novel lyophilization
method was used for the preferred candidate. Due to the low fill volume it was
not possible to use conventional flame-sealed 5 ml ampoules and the material
was therefore freeze-dried in 1 ml plastic screw-cap vials containing volumes
of 120 µl. A negative sample was also included in the collaborative study along
with other antigen-positive materials of differing strengths. A range of assays
68
International reference materials – in vitro diagnostics
were used, which included point-of-care tests, research use only tests and tests
approved for use in the WHO EUAL procedure.
Eight laboratories from four countries each received nine blinded study
samples and were asked to reconstitute these and test them without dilution.
Each sample was assessed in three different runs and the results reported as either
positive or negative. In general, results were consistent across the different VP
assays with few exceptions. Assays of nucleoprotein and glycoprotein contained
in the preferred candidate material almost invariably produced negative results
for samples demonstrated as positive for VP40. As only limited stability data
were available, it was acknowledged that additional studies would be required
before any long-term stability predictions could be made. It was also noted that
the moisture content in the novel vials was higher than is usually observed in
ampoules. It was proposed that the preferred candidate medium- and low-titre
VP40 samples plus a negative sample should be used to form a reference panel
for use by laboratories in the qualitative assessment of assay performance in
detecting VP40, with the potential limitations of the materials to be made clear
to end users. The panel would not be suitable for assessing assays used to detect
EBOV antigens other than VP40, and further work would be needed to produce
reference materials for glycoprotein and nucleoprotein assays.
The Committee discussed the suitability of VP40 expressed in bacterial
cells as a reference material for assays for point-of-care use. It was recommended
that this aspect should be highlighted in the Instructions for Use of the proposed
reference panel, and that the future inclusion of reference reagents for VP40
expressed in mammalian cells should be considered.
and recommended that the candidate materials be established as the First WHO
Reference Panel for Ebola virus VP40 antigen with the provision that it is clearly
indicated that the material was only suitable for VP40 assays and that some
assays might fail to detect a non-envelope-associated antigen.
7.1.3 First WHO reference reagents for dengue virus

serotypes 1–4 RNA for NAT-based assays
Dengue is a mosquito-borne disease that affects more than 100 tropical and
subtropical countries. An estimated 390 million infections occur each year caused
by any one of four closely related flaviviruses (DENV serotypes 1–4). Infection
with any of the four serotypes can be asymptomatic in approximately 80% of
infected individuals, or can result in dengue fever, an influenza-like illness that
may progress to severe dengue, a potentially life-threatening condition. DENVs
are also transmissible through the transfusion of blood and blood components,
and by solid organ transplant, thus posing a risk to recipients. NAT-based assays
are considered to be the most appropriate approach for blood donor screening
69
for recent DENV infection, and a proposal by the United States Food and Drug
Administration to prepare standards for use in this area had been endorsed by
the Committee in 2009.
Four candidate materials (one for each of the serotypes 1–4) had been
prepared, heat-inactivated and diluted into a base matrix at a concentration
of approximately 6 log10 PCR-detectable units/ml. Both liquid frozen and
lyophilized materials were prepared for each serotype (2000 and 8000 vials
respectively). The materials were then evaluated in an international collaborative
study during which 21 laboratories in 15 countries returned data. A wide range
of extraction and amplification methods were used, with most assays using a
different target region for amplification. Observed variations in intra-laboratory
estimates indicated a lack of reproducibility even within the same test, and
considerable variability was also observed in data sets produced by two of the
four laboratories that had used quantitative methods. However, despite these
issues, all assays were able to detect the viral RNA of each serotype, with negative
samples also correctly identified. Proposed potency estimates were assigned
to the proposed reference materials for each serotype with values reported as
NAT-detectable units (NDUs) and a 95% confidence interval established for each
data value.
The Committee agreed that there was a need for these materials. During
discussion, the subject of value assignment was raised, and in particular the
potential for confusion which may arise from the use of the term “NDU”. There
was also much discussion as to whether the materials should be established
as an international reference panel, as four separate reference reagents or as
international standards for each serotype, with a case for each option being
made. Following further deliberation, the Committee concluded that the
materials should be established as four separate WHO reference reagents, each
with an assigned value (“unit”) without a cited confidence interval. It was noted
that these reference materials could be further developed as DENV serotype-
specific international standards based on future commutability studies.

and recommended that the candidate materials be established as four separate
WHO reference reagents for DENV RNA for NAT-based assays with the
following assigned values: (a) DENV-1 RNA – 13 500 units/ml; (b) DENV-2
RNA – 69 200 units/ml; (c) DENV-3 RNA – 23 400 units/ml; and (d) DENV-4
RNA – 33 900 units/ml.
7.1.4 Fourth WHO International Standard for hepatitis B

virus DNA for NAT-based assays
HBV infection remains a major public health problem worldwide despite the
availability of an effective vaccine and antiviral therapies. More than 240 million
70
people worldwide are estimated to be chronically infected, with 0.5–1 million

people dying annually as a result of serious liver disease. The virus is transmitted
in blood and body fluids, perinatally and through close person-to-person contact
in early childhood (in regions with high HBV prevalence), as well as through
infected needles and sexual contact (in regions with low HBV prevalence).
NAT-based assays for HBV were first introduced for blood screening in 1997,
and are now implemented in at least 30 countries worldwide. The current WHO
international standard is used by in vitro diagnostic device manufacturers, blood
transfusion centres, clinical laboratories and regulatory authorities to calibrate
secondary reference materials and to validate HBV NAT-based assays.
During production of the current Third WHO International Standard
for hepatitis B virus for NAT-based assays, established in 2011, a second batch
of identical bulk material was produced but was lyophilized at a different time
point. This material (NIBSC code 10/266) was assessed as part of the 2011 study
and was deemed at the time to be a suitable candidate material for developing a
replacement standard when required.
The candidate material was therefore included – along with the current
international standard, three commutability samples and three materials for
secondary reference material calibration – in an international collaborative study.
Thirteen laboratories participated in the study with 14 data sets returned (13
from quantitative assays and one from a qualitative assay). Potency was assessed
against the current international standard, with the overall mean estimates
derived for both it and the candidate material being very similar to those seen
in the 2011 study. Ongoing stability studies suggested that the proposed material
remains stable. Given the close alignment of data between the two studies it was
proposed that the unit assigned to the replacement material should be derived
from the study data of 2011 – namely, 5.98 log 10 IU/ml – to reduce any potential
drift in the IU value.
The Committee noted that this was a good example of a project that could
be assessed in principle during meetings of the WHO network of collaborating
centres for blood products and in vitro diagnostics (see section 2.2.3 above)
with the aim of submitting a short one-page proposal to the Committee for final
approval. The WHO Secretariat commented that the study had indeed been
presented at the network meeting held earlier in the year. As no issues had
been raised at that time, a reduced presentation time had been allocated for the
topic during the current Committee meeting, thus demonstrating the potential
efficiency gains of such an approach.
Fourth WHO International Standard for hepatitis B virus DNA for NAT-based
assays with an assigned value of 5.98 log 10 IU/ml.
71
7.1.5 Fourth WHO International Standard for

prolactin (pituitary, human)
Prolactin immunoassays are used to evaluate pituitary gland function and to
monitor prolactinomas. The majority of immunoassays are calibrated with respect
to the current Third WHO International Standard for prolactin, human, stocks
of which were now running low. A candidate material (NIBSC code 83/573) was
therefore prepared under similar conditions to the current international standard
but using a different donated source of pituitary prolactin. Following its initial
formulation in the mid 1980s this candidate material had previously been assessed
twice in different collaborative studies – once in 1986 during evaluation of both
the second and third international standards by radioimmunoassay, and once in
2001 during the assessment of a recombinant reference reagent immunoassay.
In the current international collaborative study, 10 laboratories from
seven countries provided data using 11 different methods. In addition to the
candidate study samples received by all laboratories, which included the
First WHO Reference Reagent for prolactin (recombinant, human), some
laboratories also received human serum samples with normal and high prolactin
concentrations. Study results indicated that the evaluated candidate material
gave a value of 67.2 mIU/ampoule with a GCV of 8.1%. The observed stability of
the material was also good with a predicted loss of immunoreactivity of 0.007%
per year at −20 °C.
The issue of commutability was addressed by assessing the inter-
laboratory variability of results obtained for serum standards reported in terms
of the kit standards and by evaluating sources of potential bias. This analysis
was performed for both the candidate material and the recombinant reference
material. The outcomes of both assessments showed that the candidate material
was commutable – as was the recombinant material, which could act as a
replacement preparation in the future.
The Committee noted that one laboratory had reported only limited
commutability and suggested that further work should be undertaken with
the assay manufacturer to understand the reason for this. The Committee also
questioned whether manufacturers use recombinant materials for their kit
controls and were informed that end users are not provided with this information.
The Committee commended the commutability work undertaken for this
study but noted that problems in sourcing suitable materials for commutability
assessments in this area should not be underestimated.
Fourth WHO International Standard for prolactin (pituitary, human) with an
assigned value of 67 mIU/ampoule.
72
7.1.6 First WHO Reference Panel for the Janus kinase

2 V617F gene mutation
Janus kinase 2 is a non-receptor tyrosine kinase encoded by the JAK2 gene
and is involved in cytokine receptor signalling in the JAK/STAT pathway –
particularly in relation to the production of blood cells from haematopoietic
stem cells. Chronic myeloproliferative neoplasms (MPNs) are associated with
malfunctioning blood cell production and include polycythaemia vera (PV),
essential thrombocythaemia (ET), idiopathic myelofibrosis (IM) and chronic
myeloid leukaemia (CML). The JAK2 V617F mutation was discovered in 2005
and is present in > 95% of PV patients and in 50–60% of patients with either
ET or IM. The mutation results in constitutive activation of JAK2 and increased
blood cell production. WHO classification criteria for MPNs include JAK2
analysis as a requirement. Furthermore, ruxolitinib – a small-molecule JAK2
V617F inhibitor – is approved by several NRAs, and other drugs and therapies
are in development. The clinically actionable nature of such genotyping
highlights the need to identify JAK2 V617F-positive patients, and to monitor
treatment response and remission via JAK2 V617F quantitation. To achieve this
goal, accurate and sensitive testing is essential.
During an international collaborative study involving 29 laboratories,
a panel of materials (NIBSC panel code 16/120) comprising seven members
of differing wild-type/mutant ratio was formulated from a lymphoblastoid cell
line (wild-type) and a UKE-1 cell line established from the peripheral blood
of an acute myeloid leukaemia patient. The final formulation of each panel
member was approximately 5 µg of genomic DNA freeze-dried into glass
ampoules, with each panel member assigned a separate code number and
the study conducted blind. The panel was assessed using a range of assays,
where possible using quantitative methods. All data sets returned accurately
quantified both extremities of the panel (that is, 100% and 0%) with small
deviations from the expected values observed for other ratios. The final panel,
of which 1300 were produced, was proposed for establishment with genomic
JAK2 V617F DNA concentrations of 0, 0.03, 1.0, 10.8, 29.6, 89.5 and 100%. The
stability of all panel members was assessed for up to 8 months and shown to
be acceptable.
The Committee queried the apparent large gap in the proposed
concentration range of the panel – with the range between 30% and 90% lacking
representation. It was clarified that this range only accounted for 0.5 log 10,
which equated to only a very small factorial difference. It was also highlighted
that the panel was intended for use in the calibration of secondary materials,
and therefore such a range should allow for a suitable calibration curve to
be generated.
73

and recommended that the proposed materials be established as the First WHO
Reference Panel for the Janus kinase 2 V617F gene mutation with the genomic
DNA concentrations proposed above.
7.2 Proposed new projects and updates – in vitro diagnostics

7.2.1 Update on the proposed First WHO International Standard
for Ebola virus antibodies (plasma, human)
The Committee was provided with an update on recent efforts to develop a
First WHO International Standard for Ebola virus antibody (plasma, human)
and associated reference panel (WHO/BS/2016.2301). The Committee was
reminded that a convalescent plasma from the American Red Cross had
been established as an interim First WHO Reference Reagent for Ebola virus
antibodies in 2015. However, the proposed candidate material for establishment
as a first international standard, subject to satisfactory outcomes from a recent
international collaborative study, was pooled patient plasma from recovered
Ebola patients in Sierra Leone, where treatment history is unknown but where
a greater volume of material was available. Other study materials had included
a series of convalescent plasmas from four patients who had received different
drug therapies, and two purified monoclonal antibodies.
The study involved 17 laboratories from four countries with a total
of 26 data sets being returned using a range of different assay methods. Each
laboratory had received nine blinded study samples and was asked to perform
serial dilutions of these in three independent assays. As not all studies had yet
been completed, additional data would soon become available for analysis. In
the meantime, results collated to date indicate that most assays targeted the
glycoprotein region of the virus genome, with significant variability in the results
obtained. This was unexpected based on the initial in-house validation of the
materials used. Additional data also need to be generated on the stability of
the candidate reference preparation. Once all analysis is completed, a report will
be prepared.
The Committee suggested that in the final report potency should be
assigned relative to the interim standard. It was then proposed that, as the
finalization of the report would occur well in advance of the 2017 meeting
of the Committee, the material could be made available before its formal
establishment as an international standard. Following some discussion around
this point, it was agreed that the material should remain labelled as a “candidate
standard” with a value assigned in units and not IUs – which could be changed
if the Committee formally established the material in 2017. Accordingly, the
Committee recommended that, following the completed analysis, the candidate
74
material be made available to the community in advance of its meeting in 2017.

The material should be labelled as a candidate standard with a value assigned in
units until its formal establishment.
7.2.2 Proposed First WHO International Standard for Zika virus antibodies
(immunoglobulin G and immunoglobulin M) (human)
The accurate diagnosis of ZIKV infection, particularly in pregnant women, is
a crucial step in making appropriate health-care decisions. However, the virus
has a short period of detection in plasma, and PCR-based tests of plasma are
thus only useful during this period (approximately 9 days following the onset
of symptoms). After a negative PCR result, diagnosis is made on the basis of
serological testing that measures antibody responses to the virus. Current
serological tests are prone to high cross-reactivity against other flaviviruses
(particularly DENVs) spread by the same mosquito vector. To ensure accurate
diagnosis, the standardization of tests is required to improve both their sensitivity
and specificity. Both immunoglobulin G and immunoglobulin M are currently
measured to determine prior exposure, where immunoglobulin M would indicate
a recent infection. To support the WHO response to the ZVD outbreak NIBSC
had initiated development of an international standard for use in the calibration
and control of ZIKV antibody assays.
The intended use of such a material would be the calibration
of ELISAs and neutralization assays used to measure ZIKV antibody
levels (immunoglobulin G and immunoglobulin M) in human serum. The
recommendation to use the candidate material for this purpose will be subject
to the satisfactory demonstration of commutability. Primary users will be public
health and other clinical laboratories, kit manufacturers and ZIKV vaccine
manufacturers (for clinical trials).
The Committee expressed concern over reported delays and problems
encountered in receiving positive plasma donations from countries with infected
individuals. It was noted that difficulties were also encountered in sourcing ZIKV
antibody-positive plasma that was negative for antibodies to other flaviviruses.
The Committee endorsed the proposal (WHO/BS/2016.2298 Rev.1) to develop a
First WHO International Standard for Zika virus antibodies (immunoglobulin G
and immunoglobulin M) (human).
7.2.3 Proposed First WHO International Standard for chikungunya virus

antibodies (immunoglobulin G and immunoglobulin M) (human)
The mosquito-borne CHIKV is a member of the Alphavirus genus in the
Togaviridae family. Chikungunya was first identified in the United Republic of
Tanzania in the early 1950s. The virus is present not only in Africa but also in
75
Asia and the Indian subcontinent and, since 2013, has spread to the Americas,
particularly central and southern areas. Small outbreaks have also occurred
recently in Europe. The diagnosis of CHIKV infection requires a variety of
tests, including the detection of immunoglobulin M and immunoglobulin G
antibodies. The co-circulation of CHIKV with DENV and ZIKV frequently
occurs and infections caused by these viruses share common signs and symptoms
in infected patients.
The accurate diagnosis and discrimination of CHIKV infection from
other virus infections is thus vital for patient care. Analysis of immunoglobulin M
antibodies is particularly useful for the confirmation of acute infection. Anti-
CHIKV immunoglobulin G may also be detectable during acute infection, but
is also a marker of past CHIKV infection and of seroprevalence. Anti-CHIKV
assays vary in their performance, and Alphavirus serological cross-reactivity is
known to exist with members of the Semliki Forest serocomplex.
A collaborative study was therefore proposed to evaluate sera from
CHIKV-infected patients (and potentially from blood donors) for their
suitability for use as an international standard. The samples would include both
immunoglobulin G and immunoglobulin M reactive sera to distinguish between
recent and past infections.
The Committee was informed that sera would be sourced through national
and international collaborations, with the proposed project complementing
a previously endorsed and ongoing CHIKV RNA project. The Committee
endorsed the proposal (WHO/BS/2016.2298 Rev.1) to develop a First WHO
International Standard for chikungunya virus antibodies (immunoglobulin G
and immunoglobulin M) (human).
7.2.4 Proposed Second WHO International Standard for syphilitic

plasma (immunoglobulin G and immunoglobulin M) (human)
Syphilis is a sexually transmitted disease caused by spirochetes of the species

Treponema pallidum subsp. pallidum. The Committee was informed that recent
reports indicate that the incidence of syphilis had risen in both developed and
developing countries in recent years. This fact, plus the projected imminent
depletion of both the current First International Standard for human syphilitic
plasma (immunoglobulin G and immunoglobulin M) and the First International
Standard for human syphilitic plasma (immunoglobulin G), necessitated the
replacement of reference materials in this area.
A number of candidate materials were therefore currently being
evaluated at NIBSC for their suitability as replacement standards. A proposed
international collaborative study was anticipated to take place in 2017, with
submission of its report to the Committee in 2018. Anticipated users of these
76
reference materials include hospital and reference laboratories, sexual health

clinics and manufacturers.
The Committee recognized the need for the proposed replacement
standards and enquired whether the proposed study would include specific and
nonspecific assays, and if so whether any difference between the two was expected.
The Committee was informed that as was usual practice a wide range of assays
would be included in the collaborative study and any differences between them
highlighted in the outcome report. The Committee then endorsed the proposal
(WHO/BS/2016.2298 Rev.1) to develop a Second WHO International Standard
for syphilitic plasma (immunoglobulin G and immunoglobulin M) (human).

syphilitic plasma (immunoglobulin G) (human)
The Committee endorsed the proposal (WHO/BS/2016.2298 Rev.1) to develop a
Second WHO International Standard for syphilitic plasma (immunoglobulin G)
(human) (see section 7.2.4 above).
7.2.6 Proposed Second WHO International Standard for human

immunodeficiency virus type 2 RNA for NAT-based assays
The incidence of HIV type 2 (HIV-2) is lower than that of HIV-1 and typically
follows a different course of infection, with lower viral loads and a higher
percentage of longer term non-progressors. As not all treatments established for
HIV-1 are effective against HIV-2, the accurate and timely diagnosis of HIV‑2
infection is crucial in ensuring use of the correct therapy and in informing
individuals of their infection status to prevent the risk of unknown transmissions.
In addition, although screening for HIV-2 RNA is not mandated in as many
countries as for HIV-1 due to the lower incidence of infection and lower titres
observed in infected individuals, accurate and sensitive assays are vital in
ensuring the safety of the blood supply in countries where HIV-2 is prevalent.
Although supplies of the First International Standard for HIV-2 RNA
established in 2009 remain plentiful, reports from users have suggested that
the assigned unitage of 103 IU/ml is too low to create a calibration curve. This
standard is required to calibrate HIV-2 NAT-based assays and secondary HIV‑2
NAT standards and is used by test kit manufacturers, blood fractionators,
reference and diagnostic laboratories, external quality assurance (EQA) providers
and the OMCL network.
An international collaborative study was therefore proposed in order to
formulate a replacement standard at a high titre (in the order of 5 × 10 6 log 10 ).
Where possible the collaborative study would also include a clinical sample
for commutability assessment, although sourcing such material may prove
problematic. The stock material to be used would be derived from the same
77
batch of tissue-culture-grown virus (HIV-2 CAM-2) as was used for the current
international standard. Approximately 500 vials of lyophilized material would
be produced.
The Committee endorsed the proposal (WHO/BS/2016.2298 Rev.1) to
develop a Second WHO International Standard for human immunodeficiency
virus type 2 RNA for NAT-based assays.
7.2.7 Proposed First WHO International Standard for respiratory

syncytial virus RNA for NAT-based assays
Respiratory syncytial virus (RSV) and influenza virus types A and B cause
respiratory tract infections which are often more severe in vulnerable populations
such as the elderly or the very young. In both these cases, the diagnostic field is
moving towards molecular methods of detection due to their greater sensitivity
compared to other methods. However, EQA data have indicated large variations
in reported titres, highlighting the need for improved standardization in this area.
The Committee was informed that preliminary work was required to
determine the most suitable influenza A variants for inclusion as candidate
materials in the proposed parallel collaborative study. As the conserved regions
of the genome appear to be used for molecular detection, a panel of variants could
be used to assess this in a small pilot study. There was also a need to determine
whether a single RSV type (A or B) could be used for the harmonization of
both types. In addition, the sourcing of suitable clinical samples in relevant
matrix types in sufficient volumes to assess commutability may be problematic
given the limited supply of such materials. The candidate reference materials
themselves would be formulated from tissue-culture-grown virus, spiked into a
universal buffer.
The Committee discussed the impact of the genetic diversity of these
viruses on detection efficiency, and a panel-based approach was proposed as one
way of addressing this issue. However, the longevity of such a panel for influenza
was questioned should conserved regions of virus variants exhibit differing

amplification abilities. Variants chosen for a standard may not in this instance
be future-proof. This point was accepted and it was suggested that producing
batches to last up to 5 years (instead of > 10 years) may be prudent to allow for
reformulation. In addition, a different form of continued analysis may be required
to determine whether the selected strain remained suitable for standardization
purposes. It was suggested that this may be achieved by annual inclusion in an
EQA scheme. It was further proposed that egg-grown influenza viruses be used
as candidate materials instead of the originally proposed laboratory-adapted
tissue-culture-grown viruses.
develop a First WHO International Standard for respiratory syncytial virus RNA
for NAT-based assays.
78

influenza virus type A RNA for NAT-based assays
The Committee endorsed the proposal (WHO/BS/2016.2298 Rev.1) to develop
a First WHO International Standard for influenza virus type A RNA for NAT-
based assays (see section 7.2.7 above).

influenza virus type B RNA for NAT-based assays
The Committee endorsed the proposal (WHO/BS/2016.2298 Rev.1) to develop
a First WHO International Standard for influenza virus type B RNA for NAT-
based assays (see section 7.2.7 above).
7.2.10 Proposed First WHO Reference Panel for

the BRAF V600E gene mutation
BRAF V600E is a mutation of the BRAF gene known to be responsible for a
high percentage of malignant melanomas as well as other solid tumours. As
several drugs are approved for the treatment of metastatic melanoma patients
with known BRAF V600 mutations, accurate and sensitive assays for screening
melanoma cases are required. As the incidence and mortality rates of melanoma
have risen sharply throughout the world over the past few decades – with
> 130 000 new cases of melanoma diagnosed globally each year – there is a
strong public health need to improve the quality of diagnosis and treatment.
It is proposed that two genomic DNA materials are produced – the BRAF
wild-type and the BRAF V600E – with a view to evaluating the suitability of
a candidate panel based on ratio dilutions of the two materials as a reference
material. By providing purified genomic DNA prepared from cell lines, a
reproducible replacement strategy would appear to be possible. Envisaged users
include manufacturers (for the calibration of diagnostic kits) and clinical and
reference laboratories (for the calibration of secondary standards used in multiple
routine diagnostic assays for BRAF V600E detection). The project is predicted to
progress such that the results for the panel to be established would be submitted
to the Committee in 2019.
The Committee queried the suitability of the proposed format of the
study as a mimic of the way in which material would normally be assessed in the
laboratory – with users typically extracting diagnostic material from formalin-
fixed tissue blocks. In response, it was pointed out that this method of tissue
preparation is known to be highly variable and therefore introducing such a
step would be likely to add undesirable variability to the proposed evaluation.
It was therefore agreed that the candidate standardized DNA material would be
provided as initially proposed and then used in assays with material obtained
post-extraction from formalin-fixed blocks. The inclusion of an additional panel
79
member at a 10–20% dilution ratio was suggested. There was also some discussion
concerning the decision not to include lyophilized cells, with the exclusive use of
extracted DNA viewed as a positive step in assay standardization. It was clarified
that the cell line used in the development of the BRAF material had been derived
from an Epstein-Barr virus-positive cell line and it would not be appropriate to
supply laboratories with a class-2 pathogen in addition to the genetic material,
with variability in DNA extraction methods additionally compromising the
reference value of the DNA standard.
develop a First WHO Reference Panel for the BRAF V600E gene mutation.
7.2.11 Proposed First WHO Reference Panel for ErbB2

copy number and mRNA expression
Human epidermal growth factor receptor 2 (HER2) is a member of the epidermal
growth factor receptor family of receptor tyrosine kinases and is coded for by the
ErbB2 gene. This gene is amplified (and thus the HER2 protein over-expressed) in
18–20% of breast cancers, and is associated with increased tumour aggressiveness.
HER2 over-expression is also associated with more aggressive forms of ovarian,
stomach and uterine cancer. The patent for the biotherapeutic product Herceptin
(the approved treatment for HER2 over-expression) has now expired in Europe
and will expire in the United States in 2019. It is therefore likely that SBPs will now
become available. Nucleic-acid-based screening methods are increasingly used
for diagnosis and have proved to be more sensitive than immunohistochemistry
and in situ hybridization.
The proposed study will investigate several candidate HER2-positive cells
lines as well as a wild-type cell line. The intention is that one HER2-positive high
mRNA expression cell line will be selected and then serially diluted in the wild-
type cell line prior to lyophilization. This approach will allow for analysis of both
genomic DNA copy number and mRNA expression. Although another agency
(NIST) has prepared a genomic standard for HER2 this is for genomic DNA
measurement only. Envisaged users include manufacturers (for the calibration
of diagnostic kits) and clinical and reference laboratories (for the calibration
of secondary standards used in multiple routine diagnostic assays for ErbB2
characterization). The project is predicted to progress such that the results for
the panel to be established would be submitted to the Committee in 2020.
The Committee endorsed the proposal (WHO/BS/2016.2298 Rev.1)
to develop a First WHO Reference Panel for ErbB2 copy number and mRNA
expression.
80
vaccines and related substances

vaccines and related substances
Sabin inactivated poliomyelitis vaccine
The production of IPV using Sabin live-attenuated strains (sIPV) instead of wild-
type poliovirus strains is considered to be safer in the context of containment
requirements for the endgame of polio eradication. However, in contrast to
traditional IPV, the human dose for sIPV products is significantly different
between manufacturers. There is no current international standard for sIPV –
complicating the standardization of both in vitro and in vivo potency assays and
making the comparison of different sIPV products difficult.
A preliminary collaborative study was undertaken to assess whether the
current Third WHO International Standard for inactivated poliomyelitis vaccine
would be suitable for use with sIPV. Results showed that this current international
standard was not universally suitable for determining the D-antigen potency of
sIPV products as consistency between laboratories was only achieved when using
common reagents in the ELISA method. It was concluded that an international
standard specific for sIPV would be more appropriate.
The sIPV vaccine plays a key role in the global eradication of polio
and there are numerous sIPV manufacturers worldwide. The provision of an
international standard will support the standardization of potency assays of both
licensed and newly developed sIPVs and assist manufacturers, NCLs and the
OMCL network in evaluating this crucial vaccine. Ensuring the availability of
safe and effective vaccines during the endgame of polio eradication, and in the
post-eradication era, is a recognized key public health goal.
Following discussion and clarifications, the Committee endorsed the
proposal (WHO/BS/2016.2295) to develop a First WHO International Standard
for Sabin inactivated poliomyelitis vaccine, and agreed that NIBSC should seek
donations of sIPV concentrate from manufacturers worldwide as soon as possible.
8.1.2 Proposed First WHO international standards for Group B

streptococcus (polysaccharide and antiserum)
Two Group B streptococcus (GBS) polysaccharide-conjugate vaccines based on
the five most common serotypes (Ia, Ib, II, III and V) were now being developed
by two different manufacturers. Given the very low incidence of this disease
81
(around 3 per 1000 live births) these vaccines were likely to be licensed on the
basis of surrogates of protection (immunoglobulin G level and assessment of
functional antibody by opsonophagocytosis assay) rather than through very
large clinical protection studies. There is, therefore, an urgent need to establish
standardized assays for the quality control of such vaccines and to measure the
concentration and functional activity of antibodies against GBS in the sera of
immunized individuals.
The development of several standards and reference reagents
(polysaccharide standards and human reference serum) is required. The
polysaccharide standards would be used to calibrate internal controls in various
physicochemical and immune assays used in the quality control of GBS vaccines
(identity, total and free polysaccharide) and as antigens to measure antibody
responses. Human reference sera will be used as standards in assays for the
quantification of antibody response and for measurement of functional activity
in clinical trials materials. It was expected that vaccine manufacturers would
donate the polysaccharides and that human serum reference materials would be
obtained from volunteers immunized with the conjugate vaccine. It was proposed
that monovalent sera already available at NIBSC should be used to calibrate
the human reference serum. These monovalent sera have been calibrated by
radio-immunoassays for antibody concentration and have been looked upon
as gold standards in various studies to establish correlates of protection, and by
manufacturers in evaluating immune responses during clinical trials.
It was expected that the availability of relevant international standards
would expedite the development and licensing of GBS vaccines for immunizing
pregnant women in order to prevent GBS in newborns, as well as facilitate
ongoing quality control testing should such vaccines be licensed. Envisaged
users include vaccine and diagnostic kit manufacturers, public and national
health authorities, academic researchers and NCLs.
proposal (WHO/BS/2016.2295) to develop First WHO international standards

for Group B streptococcus (polysaccharide and antiserum).
8.1.3 Proposed Second WHO International Standard

for diphtheria antitoxin (equine)
Diphtheria antitoxin is an essential medicine for diphtheria therapy and is
used in countries where disease is endemic. In countries with good vaccination
coverage, it is stockpiled for emergency use. In some parts of the world equine
diphtheria antitoxin is used for diphtheria therapy and prophylaxis against
suspected cases. It is licensed nationally in some countries and imported by
others for stockpiling for emergency use. Stockpiled antitoxin is periodically
assessed for potency by a number of national regulatory authorities and an
antitoxin standard is essential for this purpose.
82
International reference materials – vaccines and related substances
The current First International Standard for diphtheria antitoxin (equine)

is a dried hyperimmune horse serum standard that was prepared in 1934 at the
Statens Serum Institut. The IU is defined as the activity contained in 0.0628
mg of the dried serum. Approximately every 2 years this dried serum is used
to prepare a standard preparation in 66% v/v glycerol/saline. The standard is
then distributed in vials containing 10 ml at 10 IU/ml. The stock of the original
dried serum is now running low. Despite being an old standard, a replacement
is needed as it is still used in some parts of the world. Although a First WHO
International Standard for human diphtheria antitoxin was established in 2012
there is a need for an equine standard for the calibration of equine products. As
minimum requirements for antitoxin potency (expressed in IU) exist in many
national and regional pharmacopoeias it is important to ensure continuity of
the IU for equine diphtheria antitoxin through the establishment of a second
international standard.
NIBSC proposes to prepare a freeze-dried standard to provide a single
homogeneous batch that can be used for a number of years. The standard is
intended to be used to calibrate potency assays for diphtheria antitoxin. These
assays include lethal and non-lethal toxin neutralization tests performed in
guinea-pigs and are used to determine the antitoxin potency of therapeutic
antitoxin products produced from equine serum. A global shortage of equine
diphtheria antitoxin has meant that source material has been difficult to identify
and cannot be freely donated. It was now expected to be sourced from an Indian
manufacturer. The proposed batch size was ≥ 2500 vials, which at the current
rate of use (approximately 50–100 vials per year) would be sufficient for more
than 20 years.
The Committee endorsed the proposal (WHO/BS/2016.2295) to develop
a Second WHO International Standard for diphtheria antitoxin (equine).
8.1.4 Proposed First WHO international standards for antibodies against

human papillomavirus types 6, 11, 31, 33, 45, 52 and 58
A second-generation vaccine against human papillomavirus (HPV), the cause of
cervical cancer, containing nine different HPV types had now been licensed and
other second-generation vaccines were under clinical development. Assessing
their immunogenicity will be crucial in defining a correlate of protection and in
monitoring their quality and performance in different populations. There were
also increasing demands to standardize HPV serological methods for measuring
past or present HPV infection in epidemiological studies as this is a key element
in the planning and follow-up of optimal HPV-control programmes.
WHO international standards for antibodies against the high-risk types
HPV16 and HPV18 were established in 2009 and 2012, respectively, and their use
has been shown to improve inter-laboratory comparisons. The development and
establishment of international standards for antibodies against low-risk types
83
HPV6 and HPV11 and against high-risk types HPV31, HPV33, HPV45, HPV52
and HPV58 would provide the full complement of international standards for
use in standardizing HPV serology assays. The users of such reference materials
would include vaccine developers and manufacturers, epidemiologists, research
laboratories, public health laboratories and developers of assay kits.
For each anti-HPV type, donations will be obtained from at least two
individuals naturally infected with the HPV type of interest. Sera would preferably
be reactive with only one genital HPV type, with donation suitability assessed by
testing for antibodies against multiple genital HPV types.
The International HPV Reference Center, Karolinska Institute will be
involved in enrolling and selecting suitable donors, with candidate sera having
also been obtained from collaborators at the National Institutes for Food and
Drug Control, China. NIBSC will undertake filling and freeze-drying operations
for each candidate standard according to standard operating procedures. The
freeze-dried candidates will then be assessed in an international collaborative
study involving 10–15 laboratories using a range of assays. Assay data will be
analysed at NIBSC using standard statistical techniques. Stability studies on the
candidate standards will also be carried out in the usual way. It was expected
that sourcing suitable samples from naturally infected individuals to produce
the seven international standards may be the rate-limiting step for this project.
proposal (WHO/BS/2016.2295) to develop WHO international standards for
antibodies against human papillomavirus types 6, 11, 31, 33, 45, 52 and 58.
8.1.5 Proposed Second WHO International Standard for hepatitis A vaccine

Vaccines against the hepatitis A virus (HAV) play a crucial role in the prevention
of HAV infection and are used globally, predominantly as travel vaccines. The
potencies of these vaccines are expressed in units unique to each manufacturer
and standardization requires the availability of an appropriate reference material.

The First WHO International Standard for inactivated hepatitis A vaccine
established in 1999 is used by manufacturers and control laboratories for the
calibration and validation of in-house standards for use in the determination
of the HAV antigen content of commercial vaccines. Potency determination is
usually carried out by ELISA or the mouse potency immunogenicity test. As
stocks of the current international standard were now running low there was a
need to consider a replacement strategy.
The collaborative study proposed by NIBSC would assign potency
in IU to the replacement standard using in vitro ELISAs and the mouse
immunogenicity assay. One issue to be addressed is the sourcing of the
candidate material. Originally, the only suppliers were European, but since
there were now several global vaccine manufacturers it will be important to see
84
International reference materials – vaccines and related substances
whether a candidate material can be used to evaluate vaccine produced using

different HAV strains, different cell lines and different manufacturing processes.
The stability profile of candidate bulk material obtained from new producers
during storage at −70 °C will also need to be studied since concentrated bulk is
normally stored at 4 °C.
The Committee agreed that there was a need to start the replacement
process for this important international standard before stocks of the current
international standard are exhausted, and endorsed the proposal (WHO/BS/
2016.2295) to develop a Second WHO International Standard for hepatitis A
vaccine.
8.1.6 Proposed WHO international standards for oral poliomyelitis vaccine

Following the significant progress made towards eliminating wild-type poliovirus
transmission, and the global eradication of serotype 2 polio, OPV2 has now been
withdrawn and the trivalent OPV (a blend of poliovirus serotypes 1, 2 and 3)
replaced by monovalent and bivalent vaccines without the serotype 2 component.
In addition, containment requirements for serotype 2 poliovirus mean that
the current trivalent WHO international standard for the assay of OPV can no
longer be used. In order to provide suitable OPV reference materials for current
vaccines, vaccine stockpiles and vaccines in development a number of new OPV
international standards are now required.
NIBSC intends to produce international standards for bivalent OPV
(containing serotypes 1 and 3) and for monovalent OPV1, 2 and 3 vaccines.
It was expected that these new international standards would be used: (a) in
potency assays of current OPV; (b) in the assessment of monovalent bulk titres;
(c) for the titration of dose preparations for neurovirulence testing; (d) for the
validation of in-house reference preparations; and (e) as an important tool in
the preparation of new sIPV bulks. Expected users of the new OPV international
standards included manufacturers, NCLs and consortia developing new and safer
vaccine strains for use following polio eradication. The proposed collaborative
study would assign potency in log 10 /CCID50 for each of the serotypes based on
infectivity assays using the Hep2C cell system.
Provision of these international standards will therefore support the
standardization of potency assays of new OPVs globally, aid NCLs in the control
of OPV and facilitate the maintaining of vaccine stockpiles, all of which will help
to ensure that safe and effective vaccines are available for disease prevention and
control during the endgame of polio eradication and in the post-eradication era.
The Committee was informed that monovalent poliovirus serotypes 1, 2 and 3
are already available at NIBSC and it was proposed that this project should take a
fast-track approach, while recognizing that containment issues linked to GAPIII
would be a potential issue, particularly for the polio serotype 2 standards.
85
The Committee fully supported the adoption of a fast-track approach

bearing in mind the urgent need for such standards and their importance to
the global polio eradication programme, and endorsed the proposal (WHO/
BS/2016.2295) to develop WHO international standards for oral poliomyelitis
vaccine.
86
Annex 1
WHO Recommendations, Guidelines and other documents
related to the manufacture, quality control and evaluation
of biological substances used in medicine
WHO Recommendations, Guidelines and other documents are intended to
provide guidance to those responsible for the production of biological substances
as well as to others who may have to decide upon appropriate methods of
assay and control to ensure that products are safe, reliable and potent. WHO
Recommendations (previously called Requirements) and Guidelines are scientific
and advisory in nature but may be adopted by a national regulatory authority
(NRA) as national requirements or used as the basis of such requirements.
Recommendations concerned with biological substances used in
medicine are formulated by international groups of experts and are published
in the WHO Technical Report Series 1 as listed below. A historical list of
Requirements and other sets of Recommendations is available on request from
the World Health Organization, 20 avenue Appia, 1211 Geneva 27, Switzerland.
Reports of the WHO Expert Committee on Biological Standardization
published in the WHO Technical Report Series can be purchased from:
WHO Press
World Health Organization
20 avenue Appia
1211 Geneva 27
Switzerland
Telephone: + 41 22 791 3246
Fax: +41 22 791 4857
Email: bookorders@who.int
Website: www.who.int/bookorders
Individual Recommendations and Guidelines may be obtained free of
charge as offprints by writing to:
Technologies Standards and Norms
Department of Essential Medicines and Health Products
World Health Organization
20 avenue Appia
1211 Geneva 27
Switzerland
Abbreviated in the following pages to TRS.

1
87
Recommendations, Guidelines and other Reference

documents
Animal cells, use of, as in vitro substrates for the Revised 2010, TRS 978 (2013)
production of biologicals
BCG vaccines (dried) Revised 2011, TRS 979 (2013)
Biological products: good manufacturing Revised 2015, TRS 999 (2016)
practices
Biological standardization and control: Unpublished document
a scientific review commissioned by the UK WHO/BLG/97.1
National Biological Standards Board (1997)
Biological substances: International Standards Revised 2004, TRS 932 (2006)
and Reference Reagents
Biotherapeutic protein products prepared by Revised 2013, TRS 987 (2014);
recombinant DNA technology Addendum 2015, TRS 999 (2016)
Biotherapeutic products, similar Adopted 2009, TRS 977 (2013)
Blood, blood components and plasma Revised 1992, TRS 840 (1994)
derivatives: collection, processing and quality
control
Blood and blood components: management Adopted 2016, TRS 1004 (2017)
as essential medicines
Blood components and plasma: estimation of Adopted 2016, TRS 1004 (2017)
residual risk of HIV, HBV or HCV infections
Blood establishments: good manufacturing Adopted 2010, TRS 961 (2011)
practices
Blood plasma (human) for fractionation Adopted 2005, TRS 941 (2007)
Blood plasma products (human): viral Adopted 2001, TRS 924 (2004)
inactivation and removal procedures
Blood regulatory systems, assessment criteria Adopted 2011, TRS 979 (2013)
for national
Cholera vaccines (inactivated, oral) Adopted 2001, TRS 924 (2004)
Dengue tetravalent vaccines (live, attenuated) Revised 2011, TRS 979 (2013)
Diphtheria, tetanus, pertussis (whole cell), and Revised 2012, TRS 980 (2014)
combined (DTwP) vaccines
Diphtheria vaccines (adsorbed) Revised 2012, TRS 980 (2014)
DNA vaccines: assuring quality and nonclinical Revised 2005, TRS 941 (2007)
safety
88
Annex 1

documents
Haemophilus influenzae type b conjugate Revised 1998, TRS 897 (2000)
vaccines
Haemorrhagic fever with renal syndrome (HFRS) Adopted 1993, TRS 848 (1994)
vaccines (inactivated)
Hepatitis A vaccines (inactivated) Adopted 1994, TRS 858 (1995)
Hepatitis B vaccines prepared from plasma Revised 1987, TRS 771 (1988)
Hepatitis B vaccines made by recombinant DNA Revised 2010, TRS 978 (2013)
techniques
Human interferons prepared from Adopted 1988, TRS 786 (1989)
lymphoblastoid cells
Influenza, biosafety risk assessment and safe Adopted 2005, TRS 941 (2007)
production and control for (human) pandemic
vaccines
Influenza vaccines (inactivated) Revised 2003, TRS 927 (2005)
Influenza vaccines (inactivated): labelling Addendum to TRS 927;
information for use in pregnant women TRS 1004 (2017)
Influenza vaccines (live) Revised 2009, TRS 977 (2013)
Influenza vaccines, human, pandemic, Adopted 2007, TRS 963 (2011)
regulatory preparedness
Influenza vaccines, human, pandemic: Adopted 2016, TRS 1004 (2017)
regulatory preparedness in non-vaccine-
producing countries
Japanese encephalitis vaccines (inactivated) Revised 2007, TRS 963 (2011)
for human use
Japanese encephalitis vaccines (live, attenuated) Adopted 2016, TRS 1004 (2017)
for human use
Louse-borne human typhus vaccines (live) Adopted 1982, TRS 687 (1983)
Malaria vaccines (recombinant) Adopted 2012, TRS 980 (2014)
Measles, mumps and rubella vaccines and Adopted 1992, TRS 848 (1994);
combined vaccines (live) Note TRS 848 (1994)
Meningococcal polysaccharide vaccines Adopted 1975, TRS 594 (1976);
Addendum 1980, TRS 658 (1981);
Amendment 1999, TRS 904 (2002)
89

documents
Meningococcal A conjugate vaccines Adopted 2006, TRS 962 (2011)
Meningococcal C conjugate vaccines Adopted 2001, TRS 924 (2004);
Addendum (revised) 2007,
TRS 963 (2011)
Monoclonal antibodies Adopted 1991, TRS 822 (1992)
Monoclonal antibodies as similar biotherapeutic Adopted 2016, TRS 1004 (2017)
products
Papillomavirus vaccines (human, recombinant, Revised 2015, TRS 999 (2016)
virus-like particle)
Pertussis vaccines (acellular) Revised 2011, TRS 979 (2013)
Pertussis vaccines (whole-cell) Revised 2005, TRS 941 (2007)
Pharmaceutical products, storage and transport Adopted 2010, TRS 961 (2011)
of time- and temperature-sensitive
Pneumococcal conjugate vaccines Revised 2009, TRS 977 (2013)
Poliomyelitis vaccines (inactivated) Revised 2014, TRS 993 (2015)
Poliomyelitis vaccines (inactivated): guidelines Adopted 2003, TRS 926 (2004)
for the safe production and quality control of
inactivated poliomyelitis vaccine manufactured
from wild polioviruses
Poliomyelitis vaccines (oral) Revised 2012, TRS 980 (2014)
Quality assurance for biological products, Adopted 1991, TRS 822 (1992)
guidelines for national authorities
Rabies vaccines for human use (inactivated) Revised 2005, TRS 941 (2007)
produced in cell substrates and embryonated
eggs
Reference materials, secondary: for NAT-based Adopted 2016, TRS 1004 (2017)
and antigen assays: calibration against WHO
International Standards
Regulation and licensing of biological products Adopted 1994, TRS 858 (1995)
in countries with newly developing regulatory
authorities
Regulatory risk evaluation on finding an Adopted 2014, TRS 993 (2015)
adventitious agent in a marketed vaccine:
scientific principles
90
Annex 1

documents
Rotavirus vaccines (live, attenuated, oral) Adopted 2005, TRS 941 (2007)
Smallpox vaccines Revised 2003, TRS 926 (2004)
Snake antivenom immunoglobulins Revised 2016, TRS 1004 (2017)
Sterility of biological substances Revised 1973, TRS 530 (1973);
Amendment 1995, TRS 872 (1998)
Synthetic peptide vaccines Adopted 1997, TRS 889 (1999)
Tetanus vaccines (adsorbed) Revised 2012, TRS 980 (2014)
Thiomersal for vaccines: regulatory expectations Adopted 2003, TRS 926 (2004)
for elimination, reduction or removal
Thromboplastins and plasma used to control Revised 2011, TRS 979 (2013)
oral anticoagulant therapy
Tick-borne encephalitis vaccines (inactivated) Adopted 1997, TRS 889 (1999)
Transmissible spongiform encephalopathies Revised 2005, WHO (2006)
in relation to biological and pharmaceutical http://www.who.int/biologicals/
products, guidelines publications/en/whotse2003.pdf
Tuberculins Revised 1985, TRS 745 (1987)
Typhoid vaccines, conjugated Adopted 2013, TRS 987 (2014)
Typhoid vaccines (live, attenuated, Ty21a, oral) Adopted 1983, TRS 700 (1984)
Typhoid vaccines, Vi polysaccharide Adopted 1992, TRS 840 (1994)
Vaccines, changes to approved vaccines: Adopted 2014, TRS 993 (2015)
procedures and data requirements
Vaccines, clinical evaluation: regulatory Revised 2016, TRS 1004 (2017)
expectations
Vaccines, clinical evaluation: use of human Adopted 2016, TRS 1004 (2017)
challenge trials
Vaccines, lot release Adopted 2010, TRS 978 (2013)
Vaccines, nonclinical evaluation Adopted 2003, TRS 926 (2004)
Vaccines, nonclinical evaluation of vaccine Adopted 2013, TRS 987 (2014)
adjuvants and adjuvanted vaccines
Vaccines, prequalification procedure Adopted 2010, TRS 978 (2013)
Vaccines, stability evaluation Adopted 2006, TRS 962 (2011)
91

documents
Vaccines, stability evaluation for use under Adopted 2015, TRS 999 (2016)
extended controlled temperature conditions
Varicella vaccines (live) Revised 1993, TRS 848 (1994)
Yellow fever vaccines Revised 2010, TRS 978 (2013)
Yellow fever vaccines, laboratories approved Revised 1995, TRS 872 (1998)
by WHO for the production of
Yellow fever virus, production and testing of Adopted 1985, TRS 745 (1987)
WHO primary seed lot 213–77 and reference
batch 168–736
92
Annex 2
Guidelines on evaluation of monoclonal antibodies as
similar biotherapeutic products (SBPs)
1. Introduction 97
2. Purpose and scope 98
3. Terminology 99
4. Special considerations for characterization and quality assessment 101
4.1 Strategy for assessment of mAb biological activity 101
4.2 Considerations for selection of the expression system 103
4.3 International standards for biological assays used in the characterization 104
5. Special considerations for nonclinical evaluation 104
5.1 In vitro studies 104
5.2 In vivo studies 106
6. Special considerations for clinical evaluation 109
6.1 Pharmacokinetics studies 109
6.2 Pharmacodynamics studies 112
6.3 Comparative clinical efficacy study 113
6.4 Indication extrapolation 122
6.5 Pharmacovigilance and post-approval consideration 123
Authors and acknowledgements 123
References 125
93
Guidelines published by the World Health Organization (WHO) are

intended to be scientific and advisory in nature. Each of the following
sections constitutes guidance for national regulatory authorities
(NRAs) and for manufacturers of biological products. If an NRA so
desires, these WHO Guidelines may be adopted as definitive national
requirements, or modifications may be justified and made by the NRA.
94
Annex 2
Abbreviations
ACR20 American College of Rheumatology 20% improvement criteria
ADA anti-drug antibody
ADCC antibody-dependent cellular cytotoxicity
ADCP antibody-dependent cellular phagocytosis
AUC area under the curve
CDC complement-dependent cytotoxicity
CHO Chinese hamster ovary
CLB competitive ligand-binding (assay)
CR complete response
CRP C-reactive protein
DAS28 disease activity score in 28 joints
DCVMN Developing Countries Vaccine Manufacturers Network
EGFR epidermal growth factor receptor
ESR erythrocyte sedimentation rate
IFPMA International Federation of Pharmaceutical Manufacturers
& Associations
IgE immunoglobulin E
IgG immunoglobulin G
IGPA International Generic Pharmaceutical Alliance
mAb monoclonal antibody
MOA mechanism of action
ORR overall response rate
pCR pathological complete response
PD pharmacodynamics
PK pharmacokinetics
RBP reference biotherapeutic product
rDNA recombinant deoxyribonucleic acid
95
SBP similar biotherapeutic product

TK toxicokinetics
TMD target-mediated disposition
TNF tumour necrosis factor
TOST two one-sided test
VEGF vascular endothelial growth factor
96
Annex 2
1. Introduction
Monoclonal antibodies (mAbs) are a major class of recombinant deoxyribonucleic
acid (rDNA) technology-derived biotherapeutic products that have achieved
outstanding success in treating many life-threatening and chronic diseases. Some
of these targeted therapy products are ranked in the top-10 lists of annual global
pharmaceutical revenue sources. As patents and data-protection measures on
mAb products have expired, or are nearing expiry, considerable attention has
turned towards producing similar biotherapeutic products (SBPs, also termed
“biosimilars”) based upon the approved mAb innovator products, with a view
to making more affordable products that could improve global access to these
so‑called blockbusters.
Therapeutic mAbs are preparations of an immunoglobulin or a fragment
of an immunoglobulin with specificity for a target ligand and are derived from
a single clone of cells. Each full-length molecule of a mAb consists of two heavy
and two light polypeptide chains which are linked by disulfide bonds. MAbs
have several possible functional domains within a single molecule. The defined
specificity of a mAb is based on the binding region for an antigen that is located
in the antigen-binding fragment (Fab) region. For full-length mAbs, their
crystallizable fragment (Fc) region has the ability to bind to specific receptors,
potentially leading to immune effector functions such as antibody-dependent
cellular cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC)
and antibody-dependent cellular phagocytosis (ADCP). Full-length mAbs
are glycoproteins with glycosylation sites in the Fc region of the heavy chains,
with further possible glycosylation sites depending on the type of molecule.
Therefore, mAbs are highly complex biological macromolecules with size and
charge variants, various post-translational modifications including different
glycosylation patterns and N- and C-terminal heterogeneity, long half-lives and
the potential to induce immunogenicity. Each individual mAb may therefore
present a unique profile, which should be taken into consideration during the
evaluation of such products as SBPs.
The WHO Guidelines on evaluation of similar biotherapeutic products
(SBPs) were adopted by the WHO Expert Committee on Biological
Standardization in 2009 (1). This document set out the scientific principles,
including the stepwise approach, which should be applied for the demonstration
of similarity between an SBP and the reference biotherapeutic product (RBP).
High similarity at the quality level is regarded as a prerequisite for enabling the
use of a tailored nonclinical and clinical programme for licensure. The goal of
the clinical comparability exercise is to confirm the similarity established at
previous stages of development and to demonstrate that there are no clinically
meaningful differences between the SBP and the RBP – and not to re-establish
safety and efficacy, as this has been done already for the RBP. The decision on
97
licensure of the SBP should be based upon evaluation of the totality of evidence
from quality, nonclinical and clinical parameters. It should be noted that clinical
studies cannot be used to resolve substantial differences in physicochemical
characteristics and biological activity between the RBP and the SBP. If substantial
differences in quality attributes are present, a stand-alone licensing approach
may be considered.
The set of globally acceptable key principles outlined above for the
regulatory evaluation and licensing of SBPs has served well as a basis for setting
national requirements for SBPs. However, because of the structural complexity
and heterogeneity of mAbs, their quality attributes can vary from product
to product. Furthermore, one mAb product may have multiple indications.
Therefore, biosimilar comparability studies between a candidate biosimilar mAb
and a reference product mAb are challenging for both developers and regulators.
Consequently, in 2014, WHO was requested to update its 2009 SBP Guidelines
to take into account technological advances in the characterization of rDNA-
derived products, and particularly mAbs. In response, WHO organized an
informal consultation in 2015 on the possible amendment of the Guidelines, with
an additional focus placed on SBPs containing mAbs. All participants, including
national regulatory authorities (NRAs) and industry, recognized and agreed
that the evaluation principles described in the WHO Guidelines were still valid,
valuable and applicable in facilitating the harmonization of SBP requirements
globally. It was therefore concluded that there was no need to revise the main
body of the existing WHO Guidelines on SBPs. However, it was also agreed that,
rather than an amendment, there was a need for additional guidance on the
evaluation of biosimilar mAbs.
2. Purpose and scope

The intention of this class-specific document is to set out the specific

considerations involved in the evaluation of mAbs developed as SBPs. These
WHO Guidelines cover rDNA-derived biosimilar mAbs used in the treatment
of human diseases. The principles discussed in this document also apply to
mAb‑derived proteins – for example, mAb fragments and Fc fusion proteins.
From a regulatory perspective, mAb assessment is based on the same
principles as those used for the evaluation of other rDNA-derived biotherapeutic
proteins. On the other hand, biosimilar mAbs should also comply with the
criteria established for demonstration of similarity. Therefore this document
should be read in conjunction with both the WHO Guidelines on evaluation of
similar biotherapeutic products (SBPs) and the WHO Guidelines on the quality,
safety and efficacy of biotherapeutic protein products prepared by recombinant
DNA technology (1, 2).
98
Annex 2
Guidance on various aspects of rDNA-derived medicines, SBPs and

mAbs is also available from several other bodies. These WHO Guidelines
are not intended to conflict with, but rather to complement, existing relevant
regulatory documents.
3. Terminology
The definitions given below apply to the terms as used in these WHO Guidelines.
These terms may have different meanings in other contexts.
American College of Rheumatology 20% improvement criteria
(ACR20): a combined index that measures disease activity in patients with
rheumatoid arthritis, and which corresponds to at least a 20% improvement
in both the tender joint count and the swollen joint count, and at least a 20%
improvement in 3 of 5 other score-set measures.
Antibody-dependent cellular cytotoxicity (ADCC): an immune
mechanism through which Fc receptor-bearing effector cells can recognize
and kill antibody-coated target cells expressing tumour- or pathogen-derived
antigens on their surface.
Antibody-dependent cellular phagocytosis (ADCP): an immune
mechanism which relies on Fc receptors, especially FcγRIIa, on macrophages
or other phagocytic cells which bind to antibodies that are attached to target
cells, followed by the phagocytosis and destruction of target cells, including
tumour cells.
Anti-drug antibodies (ADAs): host antibodies that are capable of
binding to a therapeutic antigen (recombinant protein or mAb). This may or may
not inactivate the therapeutic effects of the treatment and, in rare cases, induce
serious adverse effects.
Area under the curve (AUC): the area under the curve in a plot of
concentration of drug in serum or plasma against time.
AUC t : the area under the concentration-time curve of drug in serum
or plasma from zero up to a definite time t.
AUC tau : the area under the concentration-time curve of drug in serum
or plasma during a dosage interval.
Biological activity: the specific ability or capacity of a product to achieve
a defined biological effect.
Biosimilar mAb: a mAb product that is similar in terms of quality, safety
and efficacy to an already licensed reference product.
C max : the maximum (peak) serum or plasma concentration observed
that a drug achieves in a tested area after the drug has been administered and
prior to the administration of a second dose.
99
C min : the minimum serum or plasma concentration observed that a drug

achieves in a tested area after the drug has been administered and prior to the
administration of a second dose.
C trough : the measured serum or plasma concentration of a drug in a tested
area at the end of a dosing interval prior to the administration of the next dose.
Complement-dependent cytotoxicity (CDC): the immune process by
which an antibody–antigen complex activates complement that ultimately results
in the formation of a terminal lytic complex that is inserted into a cell membrane,
resulting in lysis and cell death.
Complete response (CR): the disappearance of all signs of cancer in
response to treatment. This does not always mean the cancer has been cured.
Disease activity score in 28 joints (DAS28): a combined index that
measures disease activity in patients with rheumatoid arthritis, which assesses
the number of swollen and tender joints, and the erythrocyte sedimentation rate
(ESR) or C-reactive protein (CRP) levels indicating how active the rheumatoid
arthritis is, along with a patient’s global assessment of their health.
Equivalence margin: a pre-specified value in the equivalence trials,
which is the largest difference that can be judged as being clinically acceptable
and which should be smaller than differences observed in superiority trials of
the active comparator.
Equivalence trial: a trial with the primary objective of showing that
the response to two or more treatments differs by an amount which is clinically
unimportant. This is usually demonstrated by showing that the true treatment
difference is likely to lie between a lower and an upper equivalence margin of
clinically acceptable differences.
Mechanism of action (MOA): the specific biochemical interaction
through which a product produces its pharmacological effect.
Monoclonal antibody (mAb): antibody derived from a single clone
of cells.
Non-inferiority trial: a trial with the primary objective of showing that
the response to the investigational product is not clinically inferior to that of a
comparative agent.
Overall response rate (ORR): the overall percentage of patients whose
cancer shrinks or disappears after treatment; this includes the rate of complete
response (CR) and partial response (PR).
Potency: the quantitative measure of biological activity based on the
attribute of the product which is linked to the relevant biological properties and
is expressed in units.
Similarity: absence of a relevant difference in the parameter of interest.
100
Annex 2
4. Special considerations for characterization

and quality assessment
The WHO Guidelines on evaluation of similar biotherapeutic products (SBPs)
set out the principle that demonstrating the similarity of a candidate SBP with
respect to the RBP in terms of quality is a prerequisite for moving forward to
comparative nonclinical and clinical studies (1). In particular, studies should be
comparative in nature and should be performed with an appropriate number
of batches of the reference product and of the SBP that is representative of
the material intended for clinical use. The RBP should be extensively tested
by analysing multiple batches, preferably over an extended period, in order
to detect possible changes in the quality profile of the RBP over time. The
minimum number of batches that should be tested will depend on the extent
of the variability of the reference product and on assay variability. The number
tested should be sufficient for drawing meaningful conclusions on the variability
of a given parameter for both the SBP and the RBP, and on the similarity of
both products. To obtain unambiguous results, the methods used should be
sufficiently sensitive, scientifically valid and suitable for their purpose.
In comparison to many other proteins, mAbs are complex glycoproteins
with distinct structural features which contribute to their diverse and variable
biological functions. Specific carbohydrates can also have an impact on the
biological activity of mAbs. For example, fucose bound by an α1–6 linkage to
the core portion of N-linked carbohydrate chains interferes with the ability
of the antibody to bind well to certain Fc receptors, resulting in diminished
Fc-mediated activities, including ADCC, whereas an increase in non-reduced
terminal galactose can enhance FcγRIIIa binding and ADCC activity.
Consequently, the assessment of biological activity of biosimilar mAbs is
particularly important and has some unique characteristics. The expression
system used for the production of mAbs can, in some cases, considerably affect
the structure and function of the mAb product. The general principles for quality
assessment of biosimilar mAbs, including physicochemical characterization, are
already described in the WHO SBP and rDNA guidelines (1, 2). Thus the quality
aspects covered in this document will focus only on specific considerations for
the assessment of mAb biological activity and on the impact of the expression
system selected for production.
4.1 Strategy for assessment of mAb biological activity

Biological activity of mAb products is an important parameter and should be
appropriately assessed. Since changes of higher-order structure could alter the
biological activity of the mAb and may not be detected by physicochemical
101
methods, the analysis of bioactivity is additionally useful for confirming the

comparability of higher-order structure.
An understanding of the mechanism of action (MOA) and receptor
interactions of the mAb is important when considering the strategy for biological
activity assessment in both the characterization study and the comparability
study. MAbs exert their action by various mechanisms ranging from simple
binding to antigen (which alone mediates the clinical effect) to binding antigen
and mediating one or more immunobiological mechanisms that combine
to produce the overall clinical response. These properties may play a role in
the MOA and/or have an impact on product safety and efficacy. Therefore, a
detailed analysis of the biological activity of the mAb – demonstrating the MOA
(including Fab- and/or Fc-mediated functions) and ability to bind to Fcγ and
neonatal Fc receptors (as well as to complement C1q) – should be provided (see
section 5.1.2 below).
Although simple antigen binding may seem to be the only mechanism
operating to achieve clinical efficacy, other effects may also play a role. In
some cases multiple functions of the mAb may be involved in an additive or
synergistic manner to produce an overall combined clinical effect. This combined
effect may be hard to dissect experimentally when seeking a clear understanding
of how the mAb mediates its clinical efficacy. Therefore, if intact mAbs are used,
care should be taken not to assume that the Fc-mediated immunobiological
effects of the product are not involved in clinical efficacy, even in situations
where simple antigen binding is considered to be the primary MOA. For
example, rituximab (a chimeric mAb specific for CD20) requires Fc function,
including ADCC, for its clinical efficacy. Assessment of Fc functions is therefore
paramount for this mAb. For infliximab (a tumour necrosis factor alpha (TNFα)
antagonist) the neutralization of soluble TNF is the primary MOA while Fc
function seems less important. However, ADCC along with other Fc- and Fab-
related functions (for example, reverse signalling) also need to be considered as
potential secondary MOAs.

Assays for measuring Fc functions can be technically demanding.
Differences in both assay formats and cell combinations have significant
impact on assay sensitivities. Assays for investigating ADCC activity require
appropriately responsive target cells and efficient effector cells. Although the use
of primary cells may provide a more physiologically relevant model, the criteria
of low assay variability and robustness may not be satisfied. Continuously
growing cell lines may overcome these limitations in some cases provided they
are more sensitive and more capable of detecting minor differences between
the RBP and the SBP. However, identifying or producing a suitable cell line
can be difficult and arduous. Furthermore, the clinical relevance of data
generated by engineered/artificial cell lines may also be challenged because of
the use of a homogeneous cell population over-expressing the targets/receptors.
102
Annex 2
Therefore, selection of an appropriate assay for the intended purpose should

always be considered as a priority in developing the strategy for assessing mAb
biological activity. Additional data may be generated by the use of different assay
formats and cell combinations to obtain results that are more relevant to the
physiological/pathophysiological conditions in patients. Although biological
assays used in characterization or for demonstrating similarity may not be as
robust as release assays, the assays should be qualified for the intended use and
should be sufficiently sensitive to detect minor differences between the RBP
and the SBP.
4.2 Considerations for selection of the expression system

The WHO SBP Guidelines (1) allow for the use of different expression systems
for production of the SBP compared to the reference product, as long as the
manufacturer can convincingly demonstrate that the structure of the molecule is
not affected or that the clinical profile of the product will not change. However,
this may pose a challenge in the context of biosimilar mAb development.
Therefore, the expression system should be carefully selected, taking into account
expression system differences that may result in undesired consequences such as
an atypical glycosylation pattern or a different impurity profile when compared
to the RBP.
Differences in glycoforms present on products may or may not have
clinical consequences. For example, production cells based on mouse cell lines
(such as SP2/0 and NS0) secrete mAbs with the carbohydrate structure alpha-
gal-1,3-gal present on the carbohydrate moiety. Humans cannot produce the
alpha-gal-1,3-gal structure as they lack the necessary enzyme for its synthesis;
however, many humans produce antibodies against this. In a proportion of
these individuals the antibodies are of the immunoglobulin E (IgE) class and
this sensitization can result in anaphylactic reactions (often serious) if they
are treated with mouse-cell-line-derived mAbs containing alpha-gal-1,3‑gal.
Such pre-existing antibodies are particularly evident for cetuximab – an
inhibitor of epidermal growth factor receptor (EGFR), which contains an
additional glycosylation site on the Fab region that is accessible for IgE binding.
Anaphylactic responses may potentially be avoided by using cell substrates
of human origin or selected clones of Chinese hamster ovary (CHO) cells for
mAb production since these cells normally cannot synthesize alpha-gal-1,3 gal.
This type of phenomenon can have important implications for biosimilar mAb
development. For example, producing an SBP of cetuximab in mouse cells
would probably show the same alpha-gal-1,3-gal-related anaphylaxis problems
as the reference product. Although production of the mAb in CHO cells may
avoid the anaphylaxis problem (since the alpha-gal-1,3-gal structure would not
be likely to be present on the mAb) the differences in glycosylation, and possibly
103
other modifications, could have an impact on the extent of studies needed for
demonstration of biosimilarity. Therefore, the selection of an expression system
for a biosimilar mAb requires careful consideration, with various potential
issues needing to be thoroughly assessed to ensure that an expression system
difference does not result in changes to critical quality attributes.
4.3 International standards for biological

assays used in the characterization
The development of assays for the determination of biological activity of
mAbs will be facilitated by WHO International Standards or WHO Reference
Reagents when available. Importantly, a clear distinction exists between reference
products and WHO International Standards or Reference Reagents since they
serve different purposes and cannot be used interchangeably. The key difference
between their uses reflects the fact that the RBP is used for all the comparability
studies, whereas WHO International Standards and Reference Reagents are used
for calibrating procedures, particularly bioassays, and cannot be used as RBPs.
The distinct roles of reference products and international standards are described
elsewhere (1, 3).
5. Special considerations for nonclinical evaluation

As with all SBPs undergoing nonclinical evaluation, a stepwise approach should
be applied to evaluate the similarity of biosimilar and reference mAbs. In vitro
studies should be conducted first and a decision then made regarding the extent
to which, if necessary, in vivo studies will be required. When deemed necessary,
in vivo nonclinical studies should be performed before initiating clinical trials.
The following approach may be considered and should be tailored
on a case-by-case basis to the SBP concerned. The approach used should be
scientifically justified in the nonclinical overview.
5.1 In vitro studies

5.1.1 SBP – general aspects
In order to assess any difference in biological activity between the SBP and the
RBP, data from a number of in vitro studies, some of which may already be
available from quality-related assays, should be provided.
As for all SBPs, the following general principles apply to biosimilar mAbs:
■■ The studies should be sensitive, specific and sufficiently
discriminatory to provide evidence that observed differences in
quality attributes, as well as possible differences that may not have
104
Annex 2
been detected during the comparative analytical assessment, are not

clinically relevant. Functional studies should be comparative and
should be designed to be sufficiently sensitive to detect differences in
the concentration–activity relationship between the SBP and the RBP.
■■ Together, these assays should cover the whole spectrum of
pharmacological/toxicological aspects with potential clinical
relevance for the reference product and for the product class.
■■ The manufacturer should discuss to what degree the in vitro assays
used are representative/predictive of the clinical situation according
to current scientific knowledge.
Since in vitro assays may often be more specific and sensitive for detecting
differences between the SBP and the reference product than studies in animals,
these assays can be considered as paramount for the nonclinical biosimilar
comparability exercise.
5.1.2 Biosimilar mAbs – specific aspects

For biosimilar mAbs, the nonclinical in vitro programme should usually include
relevant assays for the following specific evaluations:
■■ Binding studies:
(a) binding to soluble and/or membrane-bound target antigen(s);
and
(b) binding to representative isoforms of the relevant Fc receptors
(that is, for immunoglobulin G (IgG)-based mAbs to FcγRI,
FcγRII and FcγRIII), FcRn and complement (C1q).
■■ Functional studies/biological activities:

(a) Fab-associated functions (for example, neutralization of a
soluble ligand, receptor activation or blockade, reverse signalling
via activation of membrane-bound antigen); and
(b) Fc-associated functions (for example, ADCC, ADCP, CDC),
as applicable.
These assays are often technically demanding and the models chosen
should be appropriately justified by the applicant (see section 4.1 above). Together,
these assays should broadly cover the functional aspects of the mAb even though
some may not be considered essential for the therapeutic MOA. However, an
evaluation of ADCC, ADCP and CDC may be waived for mAbs directed against
non-membrane-bound targets if appropriately justified.
105
Additional note: as indicated in the ICH Guideline S6(R1) (4), tissue

cross-reactivity studies with mAbs are not suitable for detecting subtle changes
in critical quality attributes and are thus not recommended for assessing
biosimilar comparability.
5.2 In vivo studies

5.2.1 Determination of the need for in vivo studies
■■ As for SBPs in general, on the basis of the totality of quality and
nonclinical in vitro data available and the extent to which there is
residual uncertainty about the similarity of the test mAb and the
reference mAb, it is at the discretion of NRAs to waive or not to
waive a requirement for nonclinical in vivo studies. If the biosimilar
quality-comparability exercise and nonclinical in vitro studies are
considered satisfactory, and no issues are identified that would block
direct entrance into humans, an in vivo animal study may not be
considered necessary.
5.2.2 General aspects to be considered for all SBPs, including biosimilar mAbs
■■ If there is a need for additional in vivo information, the availability
of a relevant animal species or other relevant models (for example,
transgenic animals or transplant models) should be considered.
■■ If a relevant in vivo animal model is not available the manufacturer
may choose to proceed to human studies, taking into account the
principles for mitigating any potential risk.
■■ When the need for additional in vivo nonclinical studies is evaluated,
the factors to be considered include but are not restricted to:
(a) (the presence of potentially relevant quality attributes that have
not been detected in the reference product (for example, new

post-translational modification structures);
(b) the presence of potentially relevant quantitative differences in
quality attributes between the SBP and the RBP; and
(c) relevant differences in formulation (for example, use of
excipients not widely used for mAbs).
Although not all of the factors mentioned here necessarily warrant in
vivo testing, these factors should be considered together to assess
the level of concern and to determine whether or not there is a need
for in vivo testing.
■■ If product-inherent factors that have an impact on pharmacokinetics
(PK) and/or biodistribution (such as glycosylation) cannot
106
Annex 2
sufficiently be characterized on a quality and in vitro level, the

manufacturer should carefully consider if in vivo animal PK and/or
pharmacodynamics (PD) studies should be performed in advance
of clinical PK/PD testing.
5.2.3 Performance of in vivo studies

The following guidance applies to all SBPs, including biosimilar mAbs.
5.2.3.1 General aspects

If an in vivo evaluation is deemed necessary, the focus of the study/studies (PK
and/or PD, and/or safety) depends on the need for additional information to
address residual uncertainty from the quality and in vitro nonclinical evaluation.
Animal studies should be designed to maximize the information
obtained. The duration of the study (including observation period) should be
justified, taking into consideration the PK behaviour of the reference mAb, the
time to onset of formation of anti-drug antibodies (ADAs) in the test species and
the clinical use of the reference mAb.
The effects of SBPs are often species-specific. In accordance with the
WHO Guidelines on the quality, safety and efficacy of biotherapeutic protein
products prepared by recombinant DNA technology (2) and ICH Guideline S6(R1)
(4), in vivo studies should be performed only in relevant species – that is, species
which are pharmacologically and/or toxicologically responsive to the SBP.
5.2.3.2 PK and/or PD studies

When the model allows, the PK and/or PD of the SBP and the RBP should
be compared quantitatively, including, if feasible, through a dose–response
assessment that includes the intended exposure in humans.
In vivo assays may include the use of animal models of disease to evaluate
functional effects on PD markers or efficacy measures.
5.2.3.3 Safety studies

If in vivo safety studies are deemed necessary on the basis of a need for additional
information, a flexible approach should be considered. The conduct of repeated
dose-toxicity studies in non-human primates is usually not recommended. If
appropriately justified, a repeated dose-toxicity study with refined design (for
example, using just one dose level of SBP and RBP, and/or just one biological sex
and/or no recovery animals) and/or an in-life evaluation of safety parameters
(such as clinical signs, body weight and vital functions) may be considered.
Depending on the end-points needed it may not be necessary to kill the animals
at the end of the study.
107
For repeated dose-toxicity studies, where only one dose is evaluated and
the focus of the study is an evaluation of potential qualitative differences in the
toxicity profile between RBP and SBP, the dose would usually be selected at
the high end of the known dosing range of the RBP. Where the focus of the study
is an evaluation of potential quantitative differences with regard to the known
toxicity profile of the RBP, the dose level most likely to reveal differences between
the RBP and SBP should be chosen as justified on the basis of the known toxicity
and/or pharmacodynamic response of the RBP.
The conduct of toxicity studies in non-relevant species (that is, to assess
nonspecific toxicity only, based on impurities) is not recommended. Because
of the different production processes used by the SBP and reference product
manufacturers, qualitative differences in process-related impurities will occur
(for example, host cell proteins). Such impurities should be kept to a minimum
in order to minimize any associated risk.
5.2.3.4 Immunogenicity studies

Qualitative or quantitative difference(s) in product-related variants (for example,
glycosylation patterns, charge variants, aggregates and impurities such as host-
cell proteins) may have an effect on immunogenic potential and on the potential
to cause hypersensitivity. These effects are usually difficult to predict from animal
studies and should be further assessed in clinical studies.
However, while immunogenicity assessment in animals is generally
not predictive of immunogenicity in humans, it may be needed for the PK/
toxicokinetics (TK) interpretation of in vivo animal studies. Therefore, adequate
blood samples should be identified and stored for future evaluations if needed.
5.2.3.5 Local tolerance studies

Studies on local tolerance are usually not required. If excipients are introduced
for which there is little or no experience with the intended clinical route, local
tolerance may need to be evaluated. If other in vivo studies are performed the
evaluation of local tolerance may be part of the design of those studies to avoid
the need for separate local tolerance studies.
5.2.3.6 Other studies

In general, safety pharmacology and reproductive and development toxicity
studies are not warranted in the nonclinical testing of biosimilar mAbs.
In accordance with the WHO Guidelines on the quality, safety and efficacy
of biotherapeutic protein products prepared by recombinant DNA technology
(2) and ICH Guideline S6(R1) (4), genotoxicity and (rodent) carcinogenicity
studies are not required for (similar) biotherapeutic products. This guidance also
applies to biosimilar mAbs.
108
Annex 2
6. Special considerations for clinical evaluation

In general, the goal of the clinical evaluation programme is to confirm that
any residual uncertainty about quality attributes or related to the preclinical
assessment would not result in clinically meaningful differences – and not to
establish the product’s efficacy and safety in a particular indication. The clinical
comparability exercise is a stepwise procedure that should begin with PK/
PD studies and usually continues with one controlled clinical trial addressing
comparative safety and efficacy. In exceptional circumstances, data obtained in
clinical PK/PD studies may suffice in confirming biosimilarity established in
preceding steps (see section 6.2.1 below). If relevant differences between the SBP
and the RBP are detected at any stage, the reasons for the differences should be
explored and justified. In reaching a conclusion as to whether a product qualifies
as an SBP, the totality of evidence should be considered.
If the original development programme demonstrated that the
reference product performed the same in different ethnic groups then there is
no scientific rationale for conducting a comparative clinical study of the SBP in
each ethnic group.
6.1 Pharmacokinetics studies

6.1.1 Aim of comparative PK studies
Comparative clinical PK studies are always required and should be used to
further confirm the similarity of a biosimilar mAb to the RBP already established
through comparative structural, functional and nonclinical studies. In general,
factors to consider include whether the mAb is targeting a soluble antigen or a
membrane-bound antigen, and whether it is dependent on FcRn binding and/or
dependent on target-mediated clearance or non-target-mediated clearance. For
example, a biosimilar mAb may differ in its affinity for FcRn receptors from its
RBP which may lead to either a shorter or longer half-life. As a consequence of
a shorter half-life, drug exposure would be reduced, which may lead to lower
efficacy (5). Comparative PK studies may be useful in monitoring the impact of the
formation of ADAs on efficacy and safety – while exploring the impact of ADAs
on PK is also necessary. Both approaches contribute to establishing evidence in
support of extrapolation. It is not necessary to study the PK of the biosimilar
mAb in every indication that is being sought. In general, one comparative PK
study under sensitive conditions that allow any potential differences between the
SBP candidate and the RBP to be detected should be sufficient to bridge across
the indications for which the reference mAb has been authorized. The design of
comparative PK studies depends on various factors, including clinical context,
safety profiles and PK characteristics of the reference product (for example,
target-mediated disposition (TMD), linear or nonlinear PK, time-dependency
and half-life).
109
6.1.2 Study design and population

A single-dose PK study in healthy volunteers is generally recommended as they
can be considered a sensitive and homogeneous study population (6). A parallel-
group design, which generally requires a higher number of subjects, is usually
required for mAbs since a single-dose cross-over design may not be appropriate
due to the long half-lives of mAbs and the potential influence of immunogenicity
on the PK profile. However, for mAb fragments or mAbs that are not administered
systemically, alternative approaches may also be applied.
A number of key issues should be taken into account regarding the use of
healthy volunteers to study the PK of mAbs. First, healthy subjects are generally
preferred, if possible, because of their higher sensitivity and homogeneity as
compared to patient populations. Second, administering a clinically relevant
dose of some mAbs (for example, bevacizumab) may not be considered ethical
in healthy volunteers because of safety concerns, and in these cases a sub-
therapeutic dose on the linear part of the dose–response curve may be required.
Third, it may be necessary to perform the PK study for some biosimilar mAbs
(for example, rituximab) in a sensitive patient population rather than in healthy
volunteers for safety reasons. Unnecessary exposure to risk (because of safety
or medical reasons) would be viewed as unethical. Fourth, it may sometimes
be necessary to perform the PK study in a different population to that selected
for the comparative clinical efficacy study in order to establish similar clinical
efficacy. In such scenarios, population PK measurements should be collected
during the clinical efficacy trial since such data may add relevant information on
similarity. Measurement of PK parameters (especially trough levels, along with
sampling for immunogenicity) is also recommended for the evaluation of clinical
correlates of possible ADAs. Furthermore, the choice of a particular population
for PK analysis also depends on the range of therapeutic indications of the mAb
under development. For example, if a reference mAb is authorized both as an
anti-inflammatory agent and as an anticancer antibody (as for example with

rituximab) then PK data in one therapeutic area may complement clinical data
obtained in another therapeutic area and thus can also strengthen the evidence
for indication extrapolation.
6.1.3 Regimen
MAbs are often indicated both for monotherapy and as a part of combination
regimens that incorporate immunosuppressants or chemotherapeutics. It may
be sensible to study the comparative PK in the monotherapy setting in order to
minimize sources of variability. When concomitant therapy alters PK, it may
be appropriate to study comparative PK both in the monotherapy setting and in
combination, particularly where differences cannot be excluded with regard to
110
Annex 2
quality attributes that might specifically have an impact on how the drug was
cleared when used in combination.
6.1.4 PK characteristics of the reference mAb

The PK of the mAb may be affected by factors such as the antigen/receptor
level (for example, related to tumour burden in oncology), the existence of
target-mediated clearance, and/or receptor shedding which has an impact
on the variability of PK measurements. These factors should be considered
when selecting the population in which to compare the PK of the SBP to the
reference product.
6.1.5 Doses
A dose should be selected that will enable detection of potential PK differences
between the biosimilar mAb and the reference mAb. MAbs generally possess a
high degree of target selectivity, with many exhibiting nonlinear distribution and
elimination, influenced by binding to their target. In general, it is recommended
that the PK profiles should be compared using the lowest recommended
therapeutic dose. A higher (or the highest) therapeutic dose may be required
where the nonspecific clearance mechanism dominates. For mAbs that are
eliminated by TMD, a low dose (that is, one at which TMD is not saturated) may
be particularly useful for detecting differences in PK (7).
6.1.6 Routes of administration

Administration via a route that requires an absorption step is preferred unless
intravenous administration only is intended. Where the route of administration
requires an absorption step, such as the subcutaneous route, standard comparisons
of C max , AUC t and AUC 0–inf may be used to assess PK comparability.
6.1.7 Sampling times and parameters

Primary PK comparability studies should include early time points to accurately
measure C max and should also include sufficient sampling time points in the
later phases to adequately characterize the late elimination phase. This will allow
for reliable estimation of the terminal disposition rate constant and sufficient
characterization of any ADA response. In single-dose studies, optimal sampling
should continue past the expected last quantifiable concentration (AUC t ), and
the concentration–time curve should cover at least 80% of AUC 0–inf .
If a multiple-dose study is performed in patients, sampling should be
carried out at first dose and at steady state. Steady state is typically reached
after five half-lives of the mAb. PK parameters that should be evaluated include
111
AUC 0–t , AUC tau , C max and C trough , clearance and half-life. For mAbs that are
administered only intravenously, the aforementioned parameters should be
compared, as should parameters that reflect the clearance of the product.
6.1.8 Specific assays for serum drug concentration

It is preferable to have a single, validated bioanalytical assay to detect both
the biosimilar mAb and the reference mAb. The bioanalytical assay should
be appropriate for the detection and quantification of mAbs, and should be
demonstrated to be bioanalytically comparable with respect to its ability to
quantify precisely and accurately both the biosimilar mAb and the reference
mAb (8). The production of ADAs may interfere with assays for test products.
Therefore, ADAs should be measured in parallel with PK assessment, using the
most appropriate sampling time points and a subgroup analysis by ADA status
should be performed. PK analysis on the ADA-negative samples is of particular
interest as it provides the clearest picture of PK similarity.
6.1.9 Equivalence margin

In general, a comparability margin of 80–125% for the primary parameters may
be acceptable but should be justified. In some circumstances, narrowing or
widening of this margin may be required and this too should be justified.
6.2 Pharmacodynamics studies

In general, it is advisable to include PD markers as part of the clinical
comparability exercise.
6.2.1 PD markers and PD assay

For some mAbs it may even be possible to perform confirmatory PD studies
instead of controlled clinical safety and efficacy studies with conventional

clinical outcome measures. When clinical studies using PD markers are planned
to provide the main clinical evidence to establish similarity, it is recommended
that such an approach is discussed with the regulatory authorities.
The characteristics of PD markers that would support clinical efficacy,
and that manufacturers should pay attention to, are (6):
■■ The PD marker should be sufficiently sensitive to detect relevant
differences, and should be measurable with sufficient precision.
■■ The use of multiple PD markers, if they exist, is recommended.
■■ The study dose–concentration–response relationships or time–
response relationships of the selected doses should be within the
linear part of the established dose–response curve of the RBP.
112
Annex 2
■■ A clear dose–response relationship is shown.

■■ The PD marker is an acceptable surrogate marker and is related to
patient outcome.
■■ An equivalence margin should be predefined and justified.
■■ The PD assay should at least be relevant to a pharmacological effect
of the biological product (PD assay is highly dependent on the
pharmacological activity of the product; the approach for assay
validation and the characteristics of the assay performance may
differ depending on the specific PD assay).
In general, the principles regarding study design, conduct, analysis and
interpretation that are relevant to equivalence trials with a clinical outcome as
the primary end-point are applicable to equivalence trials with a PD marker
as the primary outcome.
6.3 Comparative clinical efficacy study

The confirmatory efficacy trial is the last step of the comparability exercise, thus
confirming that the clinical performance of the SBP and the RBP are comparable.
Typically, one randomized, adequately powered and preferably double-blinded
clinical efficacy study should be performed.
The manufacturer of a biosimilar mAb should perform a thorough
analysis of the publicly available clinical data for the reference product to
determine the most appropriate study population and primary end-point
combination likely to provide a relevant and sensitive model for detecting
clinically meaningful differences in efficacy and safety – and for extrapolating
efficacy and safety to therapeutic indications that are not investigated. The type
of comparative clinical trial required for the proposed biosimilar mAb could be
influenced by several factors, including:
■■ the nature and complexity of the mAb and derived products;
■■ the behaviour of the reference product in the clinic;
■■ the degree of understanding of the MOA of the mAb and disease
pathology, and the extent to which these vary in different indications
– including MOA, site of action, antigen load, drug administration
(dose, route, regimen and duration), concomitant medications, and
target population sensitivity to drug effects.
The clinical data obtained in a sensitive model can also be used to support
extrapolation to other indications of the RBP for which the proposed biosimilar
mAb has not been tested.
113
6.3.1 Clinical trial design

Clinical trial design and statistical analysis of equivalence and non-inferiority
trials that are already addressed in the WHO Guidelines on evaluation of
similar biotherapeutic products (SBPs) also apply to biosimilar mAbs (1). The
Guidelines stress the importance of clearly stating the specific design selected
for a given study and include details on the determination of the equivalence/
non-inferiority margin, sample size determination and statistical analyses. For
biosimilar mAbs, extrapolation to other indications is especially important, and
additional considerations are required in order to design a meaningful trial to
support additional indications.
Although equivalence or non-inferiority studies would be acceptable for
the comparative clinical studies of a biosimilar mAb against the RBP, equivalence
trials are generally preferred. Detailed explanations of the advantages and
disadvantages of equivalence and non-inferiority trial designs for SBPs are
provided in the WHO Guidelines and in guidance developed by other agencies
(1, 9–11). Special considerations for clinical trial design in relation to biosimilar
mAbs are explained below.
A demonstration of equivalence, as opposed to non-inferiority, is
especially important given that extrapolation to other indications may be one of
the goals of the development programme for the biosimilar mAb. Non-inferiority
trials are one-sided and hence do not exclude the possibility that the biosimilar
mAb could be found to be superior to the RBP. Such a finding would create
challenges in justifying extrapolation to other indications of the RBP. From a
statistical perspective, assay sensitivity is important to provide some confidence
that the trial, as planned and designed, will be able to detect differences between
the biosimilar mAb and the RBP, if such differences exist (12). A trial that lacks
sensitivity could lead to the erroneous conclusion of equivalence of the biosimilar
mAb to the RBP. The selected study population should not only be representative
of the approved therapeutic indications of the RBP, but should also be sufficiently
sensitive to detect potential differences between the biosimilar mAb and the
RBP. Hence, historical scientific evidence should be provided which shows that
appropriately designed and conducted trials with the RBP against placebo for
the approved indication have reliably demonstrated the superiority of the RBP
over placebo.
Study population or study end-points may deviate from those which led
to approval of the RBP for the specific indication as long as the primary end-
points are sensitive to the detection of clinically meaningful differences between
the biosimilar mAb and the RBP. Whatever approach is taken, applicants should
always justify their selection of end-points, time points for analysis and the
predefined margin, irrespective of whether this follows the RBP approach or not.
If in doubt, applicants may wish to consult relevant regulatory authorities during
the planning and design stage of the trial.
114
Annex 2
The efficacy of the RBP compared to placebo will have been demonstrated
previously. Therefore, it is considered clinically important to ensure that the
biosimilar mAb retains a substantial fraction of the effect of the RBP. As a
consequence, an equivalence margin that preserves a fraction of the smallest
effect size that the RBP can be expected to have relative to a placebo control
is the most suitable. The fraction of the effect size of the RBP that should be
retained by the biosimilar mAb should be clearly justified in each case, and
should take into account the smallest clinically important difference in a given
setting. Once the margin has been selected, determination of the required
sample size should be based on methods specifically designed for equivalence
and non-inferiority trials.
Statistical analysis of data from equivalence trials is typically based on
the indirect confidence interval comparison which requires specification of the
equivalence limits (13). Equivalence is demonstrated when the confidence
interval for the selected metric of the treatment effect falls entirely within
the lower and upper equivalence limits. If a P-value approach is used then the
P-values should be computed on the basis of the two one-sided test (TOST)
procedure, testing simultaneously the null hypotheses of inferiority and
superiority. When using the TOST procedure, equivalence is demonstrated when
the P-values obtained are less than the significance level used.
6.3.2 Study population

In order to detect differences between the biosimilar mAb and the reference
mAb, clinical trials of the biosimilar mAb should be carried out in an
appropriately sensitive patient population using end-points that are sensitive
to the detection of clinically meaningful differences between the SBP and the
reference product for the indication (see section 6.3.3 below). The rationale
for the study population selected should be provided. In general, using a
homogeneous population of patients (for example, same line(s) of therapy,
severity or stage of disease progression) will minimize inter-patient variability
and thus increase the likelihood of detecting differences between the biosimilar
mAb and the reference mAb, if such differences exist. Patients who have not
received previous treatment (for example, first-line therapy) are considered to
be more homogeneous than patients who have previously received several or
different lines of therapy. Ideally, the observed clinical effects should be triggered
by the direct action of the biosimilar mAb/reference mAb without interference
by other medications, as concomitant medications may affect or mask differences
in the PK/PD, efficacy, safety and/or immunogenicity of the tested products. To
validate the effect of the reference mAb and the sensitivity of the chosen study
population, historical data should be used to justify the selection of the study
population and equivalence margin. This could generally be done through a
systematic review and/or meta-analysis of the relevant studies.
115
MAbs can function through various MOAs, including agonist activity

or receptor blockade (for example, of vascular endothelial growth factor (VEGF)
and EGFR), induction of apoptosis, delivery of a drug or cytotoxic agent, and
immune-mediated mechanisms (for example, CDC, ADCC and regulation of
T-cell function). Because the mechanisms involved in one disease may differ
from those involved in another, extensive consideration should be given to the
setting in which clinical comparability is to be tested, particularly if functional
differences are identified in sensitive assays, and especially where it is known
that extrapolation to other indications and uses will be sought.
Clinical studies in an unauthorized population (for example, with respect
to line of therapy, combination therapy, disease severity or indication authorized
in some but not all jurisdictions) may be acceptable for demonstrating “no
clinically meaningful differences” for biosimilar mAbs. However, manufacturers
of biosimilar mAbs should consult the relevant regulatory authorities prior to
conducting such studies.
6.3.3 Primary study end-point

Clinically relevant and sensitive study end-points within a sensitive population
should be selected to improve the likelihood of detection of potential differences
between the biosimilar mAb and the reference product. In general, clinical
outcomes, surrogate outcomes or a combination of both can be used as primary
end-points in biosimilar mAb trials. The same study end-points used for
the innovator products may be used because a large body of historical data is
generally available in the public domain for setting the equivalence margin and
calculating the sample size. Alternatively, the study end-points used may be
different from those traditionally used or from the end-points recommended
by study guidelines for innovator products, as more sensitive end-points and/
or time points may exist for detecting clinically meaningful differences in an
equivalent trial setting where the objective is assessing similarity of efficacy,
safety and immunogenicity, and not re-establishing the clinical benefit already
demonstrated by the originator. A surrogate end-point can be used as the primary
end-point when surrogacy to the clinical outcome is well established or generally
accepted, as is the case, for example, with pathological complete response (pCR)
in neoadjuvant treatment of breast cancer. The choice of study end-point should
always be scientifically justified. More sensitive clinical end-points could be used
as secondary end-points for the innovator product, primary or secondary end-
points for the innovator products at different time points of analysis, and/or new
surrogates. For example, overall response rate (ORR) or complete response (CR)
rate can be considered as end-points for clinical efficacy studies of biosimilar
mAbs in oncology trials because these end-points may be more sensitive and
are not time related. However, if progression-free survival (which is one of the
end-points frequently used for clinical efficacy testing for innovator products)
116
Annex 2
is considered more sensitive than ORR, then this may be the preferred option.
Likewise, both continuous outcomes (for example, changes in DAS28 from
baseline) and dichotomous outcomes (for example, ACR20) are considered in
rheumatoid arthritis trials for determining clinical comparability (14).
When the primary efficacy end-points that were used for the RBP cannot
be used for the SBP it is advisable to include some common end-points as
secondary end-points to facilitate comparisons between the SBP and the RBP.
The role of these secondary end-points in the overall interpretation of the study
results should be clearly defined, particularly in terms of whether the secondary
end-points are used to support or to confirm equivalency or similarity.
NRAs may not always agree on the choice of study end-points. For an SBP
manufacturer with a global development programme that is guided or required
by various NRAs to fulfil local regulatory or clinical practice requirements it
may be possible to pre-specify different primary study end-points with the
statistical power in the same trial to comply with various regulatory requirements.
6.3.4 Safety
6.3.4.1 General considerations
Comparative safety data should normally be collected pre-authorization. The
extent of data collection depends on the type and severity of known safety
issues for the reference product. The SBP study population should be followed
to provide information on safety events of interest according to experiences
with the reference mAb. Care should be taken to compare the nature, severity
and frequency of adverse events between the biosimilar mAb and the reference
product in clinical trials that enrolled a sufficient number of patients treated
for an acceptable period. Clinical safety issues should be captured throughout
clinical development during initial PK and/or PD evaluations and also in the
primary clinical study establishing comparability.
6.3.4.2 Immunogenicity of a biosimilar mAb

Therapeutic mAbs, like other rDNA-derived biotherapeutics, may be recognized
by the human immune system leading to an unwanted immune response. As
mAbs may often be immunogenic in patients, the goal during development of
a biosimilar mAb is to demonstrate similar immunogenicity to the reference
product. There are some special considerations regarding the immunogenicity
of mAbs as compared to other biotherapeutics. For example, mAbs do not evoke
cross-reacting antibodies against the body’s endogenous proteins, as some
growth factors and proteins for replacement therapy do. However, developing
assays to test for anti-mAb-antibodies can be challenging.
From the regulatory point of view, animal data are not sufficiently
predictive of the human immune response against a therapeutic protein. Thus,
117
immunogenicity generally needs to be investigated as part of the clinical trial

programme of a biosimilar mAb. The analysis of the immunogenicity of DNA-
derived biotherapeutics is outlined in the WHO Guidelines on evaluation of
similar biotherapeutic products (SBPs) and the WHO Guidelines on the quality,
safety and efficacy of biotherapeutic protein products prepared by recombinant
DNA technology (1, 2). These general guidance resources should be taken
into account when assessing biosimilar mAbs. In addition, further details
regarding the advantages and disadvantages of particular assays, as well as some
considerations on the interpretation of the results and on the decision-making
process, are provided in several review articles (15–18).
The basic data package contains the incidence, titre, neutralization
ability and persistence of antibodies against the product/mAb, determined by
appropriate assays, as well as their pharmacokinetic and clinical correlations.
The immunogenicity programme needs to be tailored to each product. Thus,
the evaluation of immunogenicity requires a multidisciplinary approach,
including considerations of product-, process-, patient- and disease-related
factors that will form the basis of a risk-based immunogenicity programme. It is
recommended that the application for marketing authorization of a mAb includes
a summary of the immunogenicity programme in support of the selected
approach to immunogenicity. This summary should address the following topics
as appropriate:
■■ risk assessment
■■ risk-based immunogenicity programme
■■ comparative immunogenicity
■■ assays and mAb characterization
■■ clinical immunogenicity assessment.
6.3.4.2.1 Risk assessment
■■ previous knowledge of the immunogenicity of the reference

product, such as the presence of immunogenic structures in the
active substance as well as the incidence, type, persistence and
clinical correlations of the antibodies;
■■ findings of the physicochemical and structural comparisons
between the biosimilar mAb and its reference product, including
process-related impurities and aggregates;
■■ differences in formulation and packaging (for example, potential
impurities and leachables);
■■ route and/or mode of administration of the product;
118
Annex 2
■■ patient- and disease-related factors such as the state of the immune

system, concomitant immunomodulatory therapy and potential
pre-existing immunity, antigenicity and sensitivity.
6.3.4.2.2 Risk-based immunogenicity programme

The manufacturer should present a risk-based immunogenicity assessment
programme.
■■ The basis of the immunogenicity assessment is the testing of
patient samples pretreatment, during treatment and, if needed,
post-treatment in an appropriate set of assays that are suitable for
the product in question. The measurement of antibodies to mAbs
is methodologically challenging since standard assay formats
involving anti-immunoglobulin reagents are inappropriate for this
product class; therefore alternative methods should be used. As with
other biotherapeutic products a multi-tiered assessment approach
is needed. The developer has to validate assays for screening,
confirmation and neutralization ability. Special attention should be
paid to the choice of the control matrix, determination of cut-off
points and the estimation of interference by matrix components,
including the drug target and the residual drug in the sample. To
mitigate this potential interference, corrective measures should be
implemented. For example, drug interference may be overcome by
allowing time for clearance of the drug from the circulation prior to
sampling, or by dissociating immune complexes, and/or removal of
the drug. Inclusion of any of these measures should not compromise
the detection of antibodies or patient treatment.
■■ With regard to the integration of the product antibody testing
into the comparative clinical trials, it is particularly important to
synchronize the sampling schedule and duration of the follow-up
for product antibody determination and PK measurements, as well
as for assessments of safety and efficacy.
■■ Special emphasis should be placed on the potential association of
product antibodies with loss of efficacy, with infusion reactions, and
with acute and delayed hypersensitivity. The manufacturer should
systematically use terminology and definitions to characterize
potentially immune-mediated symptoms, in accordance with
relevant publications (19, 20).
■■ The manufacturer should take into account the dose and dosing
schedule, including re-administration, after discontinuation of
treatment.
119
■■ The vulnerability of the patient population(s) and the expected risks

of immunogenicity should be taken into account in planning for the
intensity of monitoring.
■■ The manufacturer should provide a description and analysis of the
use of pre-medication or de-immunization measures to mitigate
acute infusion/injection-related reactions and other possibly
immune-mediated reactions.
After discontinuation of the therapy, it is important to investigate the
persistence of product antibodies formed during drug administration, as well as
the emergence of product antibodies that may have escaped detection because
of the immunosuppressive action of the product or because of technical
problems (notably drug interference). The timing of the post-treatment samples
should be justified.
6.3.4.2.3 Comparative immunogenicity

The lack of standardization and rapid evolution of the assay methodology
makes it difficult to compare immunogenicity studies. Therefore, pre-licensing
comparative immunogenicity data are generally needed in the development of
SPBs (1, 11). Immunogenicity testing of the SBP and the reference product
should be conducted within the biosimilar comparability exercise by using
the same assay format and sampling schedule. A parallel-group design is
recommended because of the long half-life of antibodies and because it may be
difficult to interpret immunogenicity after a switch.
6.3.4.2.4 Assays and mAb characterization

ADA assays should ideally be capable of detecting all antibodies against both
the reference and biosimilar molecule. Thus assays can be performed with both
the reference and biosimilar molecule as the antigen/capture agent in parallel

in order to measure the immune response against the product received by each
patient. The challenge is to develop two assays with similar sensitivity. Cross-
testing all serum samples by both tests is useful for exploring assay performance
and antigen epitopes. The use of a single assay with the active substance of the
SBP as the antigen/capturing agent for evaluation of all samples (including those
from reference-product-treated patients) will be able to detect all antibodies
developed against the biosimilar molecule (that is, the conservative approach).
In general, the manufacturer should justify the chosen assay approach and
should demonstrate the suitability of the method(s) used to measure similarly
the immune response against the product received by each patient, irrespective
of whether the patient was treated with the RBP or the SBP.
120
Annex 2
Following identification of confirmed antibody-positive samples,

characterization of the antibodies is required. Determination of their neutralizing
potential is essential and deviation from this requires justification. Although
a functional (usually cell-based) bioassay or a binding assay (for example, a
competitive ligand-binding (CLB) assay) can be used alone, the latter should be
used only if relevant to the MOA of the product. For example, a CLB assay is
appropriate in a scenario where a therapeutic mAb acts by binding to a soluble
ligand, thereby blocking it from interacting with its receptor and thus inhibiting
the biological action of the ligand. Since the assay procedure measures binding
to the target and inhibition of the binding activity if neutralizing antibodies are
present, it is reflective of the MOA of the therapeutic mAb. For intact mAbs
where effector functions are likely to contribute to the clinical effect, functional
cell-based bioassays are recommended because the MOA cannot be reflected
adequately in a CLB assay. Nevertheless, such cell-based assays may not be
sufficiently sensitive and a CLB assay may give a more accurate assessment of the
incidence of neutralizing antibody induction.
Additional studies beyond the standard data package, such as
immunoglobulin class, epitope mapping and IgG subclass, may be useful in
specific situations (for example, occurrence of anaphylaxis or use of certain
assay formats). It may also be necessary to locate the antigenic sites (for example,
antigen-binding region versus constant region of the antibody molecule). The
banking of patient samples is necessary in order to retain the possibility for
retesting in case of technical problems in the original assay.
6.3.4.2.5 Clinical immunogenicity assessment

The selected patient population should be sensitive for the detection of
differences in immunogenicity. It is also important that the controlled safety and
efficacy study will include both immunogenicity and PK measurements (especially
C trough levels) in order to establish the clinical impact of immunogenicity. If the
study includes patients previously treated with the reference mAb, a subgroup
analysis of previously treated patients should be performed. The sampling
schedule should be optimized for the demonstration of similar onset and
persistency of antibodies to the test and reference products.
The duration of follow-up of immunogenicity depends on the duration
of exposure and should be sufficient to demonstrate similar persistence and
clinical impact of the antibodies. In chronic administration, the minimum
follow-up is 6 months.
Immunogenicity should be followed after licensing by monitoring
possible immune-mediated adverse effects. Special immunogenicity studies may
be necessary in high-risk situations (for example, when the reference product is
known to have serious but rare immune-mediated effects, such as anaphylaxis).
121
Evaluation of immunogenicity includes antibody incidence, titre,

neutralization capacity and persistency, as well as correlations to exposure, safety
and efficacy. Currently, there is no generally accepted statistical methodology
that could be used to define the limits of comparable immunogenicity. In
general, an increase in immunogenicity of an SBP when compared with the RBP
is incompatible with the biosimilarity principle unless the sponsor can show
that the product antibodies have no clinical relevance and that the underlying
difference between the SBP and the reference product does not signal an
otherwise important problem.
6.4 Indication extrapolation

Indication extrapolation is the regulatory and scientific process of extending
information and conclusions available from one patient population in order
to make inferences for other populations. In the context of SBPs, it refers to
granting a clinical indication to an SBP for which the reference product is
authorized, without conducting clinical efficacy and safety studies to support
that indication. Extrapolation cannot be claimed automatically for all
indications of the reference product and requires sound scientific justification
based on the totality of evidence. The starting point for extrapolation is that the
physicochemical and structural analyses, nonclinical tests and clinical studies
have demonstrated comparability. Thus, extrapolation should be considered in
the light of the totality of evidence of biosimilarity. Current WHO guidance on
SBP evaluation (1) sets out a number of recommended principles regarding the
extrapolation of clinical data across indications which also apply to biosimilar
mAbs. Extrapolation is possible when the following requirements are fulfilled:
■■ a sensitive clinical test model has been used that is able to detect
potential differences between the SBP and the RBP;
■■ the clinically relevant MOA and/or involved receptor(s) are the same;
■■ safety and immunogenicity of the SBP have been sufficiently
characterized and there are no unique or additional safety issues
expected for the extrapolated indication(s).
MAbs have both Fab and Fc-effector functions and may exert their
clinical effect through a variety of mechanisms – for example, ligand blockade,
receptor blockade, receptor down-regulation, cell depletion (via ADCC, CDC
or apoptosis) and signalling induction. A particular mAb may act through one
or a combination of these or other mechanisms. Where a therapeutic mAb is
indicated for a variety of diseases, various MOAs may be important, depending
on the indication in question. In order to support extrapolation, the mechanisms
that contribute to the efficacy of the mAb in each indication should ideally be well
122
Annex 2
understood and clearly defined. In practice, this is often not the case. Therefore,
extrapolation may pose additional challenges when a mAb is indicated for a
variety of diseases in which the MOAs are not the same or are not well understood
for each indication. In this situation, it is important to explore the comparability
of in vitro functions of the mAb. In cases where significant functional differences
exist, further nonclinical or clinical data are needed to support extrapolation.
Therefore it is essential that the basic functions of the antibody are considered
when relevant. The tests should be selected according to their relevance for a
particular product and therapeutic indication and, if possible, tailored accordingly
(for example, ADCC assays under different conditions). If minor quality
differences are found, and the affected mechanism is not considered active in the
studied indication, additional steps may be necessary to reach a conclusion on
biosimilarity. Additional data, with appropriate supporting scientific rationale,
could include quality, preclinical and/or PK/PD data and might impact on the
selection of the final clinical, safety and efficacy study. Special post-marketing
measures may be used to monitor aspects of safety and/or immunogenicity in the
extrapolated therapeutic indications.
6.5 Pharmacovigilance and post-approval consideration

A risk-management plan should be put in place once a biosimilar mAb is approved,
in order to ensure its long-term safety and efficacy. The general requirements
for pharmacovigilance are the same as for any approved new drug. As described
in WHO guidelines (1, 2) it is essential to record the product brand name,
batch number and manufacturers’ name, and, where it exists, the International
Nonproprietary Name (INN). In many cases, clinically important adverse events
occur at a relatively low frequency and the probability of them occurring during
the time frame of the clinical trial is also low. Additionally, because of their
relatively small sample size, biosimilar mAb clinical trials may have the statistical
power to detect only common adverse events. Thus, as for any biological
medicine, pharmacovigilance is essential in order to detect potential overt new
or rare biosimilar mAb-specific safety issues and to allow for the identification
and assessment of potential post-marketing risks.
Authors and acknowledgements

The first draft of these WHO Guidelines was prepared by Dr K. Gao, World
Health Organization, Switzerland; Dr E. Griffiths, Consultant, Kingston-upon-
Thames, England; Dr H-K. Heim, Federal Institute for Drugs and Medical
Devices, Germany; Dr I. Knezevic, World Health Organization, Switzerland; Dr P.
Kurki, Finnish Medicines Agency, Finland; Dr N. Ekman, Finnish Medicines
123
Agency, Finland; Dr C. Njue, Health Canada, Canada; Dr R. Thorpe, Consultant,

Welwyn, England; Dr M. Wadhwa, National Institute for Biological Standards
and Control, England; Dr J. Wang, Health Canada, Canada; and Dr E. Wolff-
Holz, Paul-Ehrlich-Institut, Germany.
The drafting group took into consideration proposals made during a
WHO informal consultation on monoclonal antibodies as similar biotherapeutic
products (SBPs) held in Geneva, Switzerland, 27–28 April 2015 and attended by:
Mrs A. Abas, Ministry of Health Malaysia, Malaysia; Dr A. Abdelaziz, Jordan
Food and Drug Administration, Jordan; Dr K. Bangarurajan, Central Drug
Standards Control Organization, India; Mrs J. Bernat (International Federation of
Pharmaceutical Manufacturers & Associations (IFPMA) representative), IFPMA,
Switzerland; Dr M.C. Bielsky, Medicines and Healthcare products Regulatory
Agency, England; Dr B. Boonyapiwat, Ministry of Public Health, Thailand; Dr S.
Bukofzer (International Generic Pharmaceutical Alliance (IGPA) representative),
Hospira, the USA; Ms J. Dahlan, National Agency of Drug and Food Control,
Indonesia; Mrs D. Darko, Food and Drug Authority, Ghana; Mr D. Goryachev,
Ministry of Health, Russian Federation; Dr E. Griffiths, Consultant, Kingston-
upon-Thames, England; Dr K. Heidenreich (IFPMA representative), Novartis
International AG, Switzerland; Dr H-K. Heim, Federal Institute for Drugs and
Medical Devices, Germany; Dr C. Ilonze, National Agency for Food and Drug
Administration and Control, Nigeria; Mr R. Ivanov, Biocad, Russian Federation;
Mrs W. Jariyapan, Ministry of Public Health, Thailand; Dr J. Joung, Ministry
of Food and Drug Safety, Republic of Korea; Dr D. Khokal, Health Sciences
Authority, Singapore; Mr B. Kim, Celltrion, Republic of Korea; Mrs S. Kox (IGPA
representative), Medicines for Europe, Belgium; Dr A. Kudrin, Celltrion, Republic
of Korea; Dr P. Kurki (on behalf of the Biosimilar Medicinal Products Working
Party, European Medicines Agency), Finnish Medicines Agency, Finland; Ms I.
Lyadova, Ministry of Health, Russian Federation; Dr J. Macdonald (IFPMA
representative), Pfizer, the United Kingdom; Mrs S. Marlina, National Agency of

Drug and Food Control, Indonesia; Dr J. Meriakol; Ministry of Health, Kenya;
Mr K. Nam, Ministry of Food and Drug Safety, Republic of Korea; Dr C. Njue,
Health Canada, Canada; Mrs Y.H. Nunez, Centro para el Control Estatal de la
Calidad de los Medicamentos, Cuba; Ms H.B. Pedersen, WHO Regional Office
for Europe, Denmark; Dr M.L. Pombo, Pan American Health Organization, the
USA; Dr A. Qu (Developing Countries Vaccine Manufacturers Network (DCVMN)
representative), Shanghai Institute of Biological Products Co., Ltd, China; Dr S.
Ramanan (IFPMA representative), Amgen Inc., the USA; Dr T. Schreitmueller
(IFPMA representative), F. Hoffmann–La Roche Ltd, Switzerland; Dr M. Schiestl
(Medicines for Europe representative), Sandoz GmbH, Austria; Dr J. Siegel
(IFPMA representative), Johnson & Johnson, the USA; Dr E. Spitzer (Asociación
Latinoamericana de Industrias Farmacéuticas (ALIFAR) representative), Argentina;
124
Annex 2
Ms I. Susanti (DCVMN representative), PT Bio Farma, Indonesia; Dr R. Thorpe,

Consultant, Welwyn, England; Dr C.P. Vaca Gonzalez, Ministry of Health,
Colombia; Ms B. Valente, Agência Nacional de Vigilância Sanitária, Brazil;
Ms N. Vergel, Instituto Nacional de Vigilancia de Medicamentos y Alimentos,
Colombia; Dr M. Wadhwa, National Institute for Biological Standards and
Control, England; Dr J. Wang, Health Canada, Canada; Dr J. Wang, National
Institutes for Food and Drug Control, China; Dr L. Wang, National Institutes for
Food and Drug Control, China; Dr K. Watson (IFPMA representative), AbbVie
Ltd, the United Kingdom; Dr C. Webster (IGPA representative), Baxter, the USA;
Dr E. Wolff-Holz, Paul-Ehrlich-Institut, Germany; Dr S. Xie, China Food and
Drug Administration, China; Dr T. Yamaguchi, Pharmaceuticals and Medical
Devices Agency, Japan; Dr A. Zhang (DCVMN representative), China National
Biotec Group Co. Ltd, China; and Dr K. Gao, Dr H. Kang, Dr I. Knezevic and
Dr D. Wood, World Health Organization, Switzerland.
The document WHO/BS/2016.2290 was prepared by the drafting group
and posted on the WHO Biologicals website for a round of public consultation
from 22 July to 16 September 2016, with feedback received from both regulators
and industry.
Further changes were subsequently made to document WHO/BS/
2016.2290 by the WHO Expert Committee on Biological Standardization.
References
1. Guidelines on evaluation of similar biotherapeutic products (SBPs). In: WHO Expert Committee
on Biological Standardization: sixtieth report. Geneva: World Health Organization; 2013:
Annex 2 (WHO Technical Report Series, No. 977; http://who.int/biologicals/publications/trs/areas/
biological_therapeutics/TRS_977_Annex_2.pdf, accessed 8 December 2016).
2. Guidelines on the quality, safety and efficacy of biotherapeutic protein products prepared by
recombinant DNA technology. In: WHO Expert Committee on Biological Standardization: sixty-
fourth report. Geneva: World Health Organization; 2014: Annex 4 (WHO Technical Report Series,
No. 987; http://www.who.int/biologicals/biotherapeutics/TRS_987_Annex4.pdf?ua=1, accessed
8 December 2016).
3. Thorpe R, Wadhwa M. Intended use of reference products & WHO international standards/
reference reagents in the development of similar biological products (biosimilars). Biologicals.
2011;39(5):262–5 (http://www.sciencedirect.com/science/article/pii/S1045105611000704, accessed
8 December 2016).
4. Preclinical safety evaluation of biotechnology-derived pharmaceuticals. ICH Guideline S6(R1).
Geneva: International Conference on Harmonisation of Technical Requirements for Registration
of Pharmaceuticals for Human Use; 2011 (http://www.ich.org/fileadmin/Public_Web_Site/ICH_
Products/Guidelines/Safety/S6_R1/Step4/S6_R1_Guideline.pdf, accessed 8 December 2016).
5. Wang W, Lu P, Fang Y, Hamuro L, Pittman T, Carr B et al. Monoclonal antibodies with identical
Fc sequences can bind to FcRn differentially with pharmacokinetic consequences. Drug
Metab Dispos. 2011;39(9):1469–77 (http://dmd.aspetjournals.org/content/39/9/1469, accessed
8 December 2016).
125
6. Committee for Medicinal Products for Human Use. Guideline on similar biological medicinal
products containing monoclonal antibodies – non-clinical and clinical issues. London: European
Medicines Agency; 2012 (EMA/CHMP/BMWP/403543/2010; http://www.ema.europa.eu/docs/en_
GB/document_library/Scientific_guideline/2012/06/WC500128686.pdf, accessed 8 December
2016).
7. Pippig SD, Brockmeyer C, Zoubek RE. Biosimilar monoclonal antibodies. In: Dübel S, Reichert JM,
editors. Handbook of therapeutic antibodies. Second edition; pages 681–704. Weinheim: Wiley-
VCH Verlag; 2014; doi: 10.1002/9783527682423.
8. Marini JC, Anderson M, Cai X-Y, Chappell J, Coffey T, Gouty D et al. Systemic verification of
bioanalytical similarity between a biosimilar and a reference biotherapeutic: committee
recommendations for the development and validation of a single ligand-binding assay to
support pharmacokinetic assessments. AAPS J. 2014;16(6):1149–58 (https://www.ncbi.nlm.nih.
gov/pmc/articles/PMC4389757/, accessed 8 December 2016).
9. Guidance for sponsors: information and submission requirements for subsequent entry biologics
(SEBs) [website]. Ottawa: Health Canada; 2010 (http://www.hc-sc.gc.ca/dhp-mps/brgtherap/
applic-demande/guides/seb-pbu/seb-pbu_2010-eng.php, accessed 8 December 2016).
10. Scientific considerations in demonstrating biosimilarity to a reference product. Guidance
for industry. Silver Spring: Food and Drug Administration; 2015 (http://www.fda.gov/
downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM291128.pdf,
accessed 8 December 2016).
11. Committee for Medicinal Products for Human Use. Guideline on similar biological medicinal
products containing biotechnology-derived proteins as active substance: non-clinical and
clinical issues. London: European Medicines Agency; 2015 (EMEA/CHMP/BMWP/42832/2005
Rev1; http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2015/01/
WC500180219.pdf, accessed 9 December 2016).
12. Choice of control group and related issues in clinical trials. E10 (Current Step 4 version). Geneva:
ICH Harmonised Tripartite Guideline. Geneva: International Conference on Harmonisation of
Technical Requirements for Registration of Pharmaceuticals for Human Use; 2000 (http://www.
ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E10/Step4/E10_Guideline.
pdf, accessed 9 December 2016).
13. Chow S-C, Liu J-P. Design and analysis of clinical trials: concepts and methodologies. Second
edition. Hoboken: John Wiley & Sons, Inc.; 2004.
14. Kudrin A, Knezevic I, Joung J, Kang H-N. Case studies on clinical evaluation of biosimilar
monoclonal antibody: scientific considerations for regulatory approval. Biologicals. 2015;43(1):
1–10 (abstract: http://www.sciencedirect.com/science/article/pii/S1045105614001158, accessed
9 December 2016).
15. Wadhwa M, Knezevic I, Kang H-N, Thorpe R. Immunogenicity assessment of biotherapeutic
products: an overview of assays and their utility. Biologicals. 2015;43(5):298–306 (http://www.
sciencedirect.com/science/article/pii/S1045105615000627, accessed 9 December 2016).
16. Knezevic I, Kang H-N, Thorpe R. Immunogenicity assessment of monoclonal antibody products:
a simulated case study correlating antibody induction with clinical outcomes. Biologicals.
2015;43(5):307–17 (http://www.sciencedirect.com/science/article/pii/S1045105615000780,
17. Gupta S, Devanarayan V, Finco D, Gunn GR, Kirshner S, Richards S et al. Recommendations for the
validation of cell-based assays used for the detection of neutralizing antibody immune responses
elicited against biological therapeutics. J Pharm Biomed Anal. 2011;55(5):878–88 (abstract: http://
www.sciencedirect.com/science/article/pii/S0731708511001907, accessed 9 December 2016).
126
Annex 2
18. Shankar G, Devanarayan V, Amaravadi L, Barrett YC, Bowsher R, Finco-Kent D et al.

Recommendations for the validation of immunoassays used for detection of host antibodies
against biotechnology products. J Pharm Biomed Anal. 2008;48(5):1267–81 (abstract: http://
www.sciencedirect.com/science/article/pii/S0731708508005001, accessed 9 December 2016).
19. Kang SP, Saif MW. Infusion-related and hypersensitivity reactions of monoclonal antibodies
used to treat colorectal cancer – identification, prevention, and management. J Support Oncol.
2007;5(9):451–7 (https://www.researchgate.net/publication/5826910_Infusion-Related_and_
Hypersensitivity_Reactions_of_Monoclonal_Antibodies_Used_to_Treat_Colorectal_Cancer-
Identification_Prevention_and_Management, accessed 9 December 2016).
20. Rup B, Pallardy M, Sikkema D, Albert T, Allez M, Broet P et al. Standardizing terms, definitions
and concepts for describing and interpreting unwanted immunogenicity of biopharmaceuticals:
recommendations of the Innovative Medicines Initiative ABIRISK consortium. Clin Exp
Immunol. 2015;181(3):385–400 (http://www.abirisk.eu/documents/publications/Rup_A_et_al_
Standardizing_Terms.pdf, accessed 9 December 2016).
127
Annex 3
Guidelines on management of blood and blood
components as essential medicines
1. Introduction 132
3. Blood and blood components as biological therapeutic products 135
3.1 Historical background of blood transfusion 135
3.2 Indications for essential blood and blood component therapy 136
3.3 Risks of blood and blood components 138
4. Preparation of blood and blood components 139
4.1 Ethical aspects of blood donation 139
4.2 Description of key product preparation steps 141
4.3 Associated substances and equipment 145
5. Comparison of blood components with PDMPs 146
5.1 General 146
5.2 Product safety and quality 147
5.3 Product efficacy 148
6. The blood regulatory system 149
6.1 Guiding principles 149
6.2 Regulatory framework 149
6.3 The regulatory authority 153
7. The blood supply system 154
7.1 Organization of the blood supply system 154
7.2 Functions of blood establishments/banks 154
8 The blood transfusion system 154
9. Stepwise implementation of a nationally regulated blood system 155
References 158
Appendix Examples of existing legislation, regulations and guidance 160
129

(NRAs) and for blood establishments/banks that prepare blood and
blood components intended for transfusion. If an NRA so desires, these
WHO Guidelines may be adopted as definitive national requirements,
or modifications may be justified and made by the NRA.
130
Annex 3
Abbreviations
EM essential medicine
GPP good preparation practice(s)
GvHD graft versus host disease(s)
HLA human leukocyte antigen
PRP platelet-rich plasma
RBC red blood cell
RTTI relevant transfusion-transmitted infection(s)
SOP standard operating procedure
131
1. Introduction
Essential medicines (EMs) are defined by WHO as those medicinal products
that satisfy the health-care needs of the majority of the population. They should
therefore be available at all times, in adequate amounts and in appropriate
dosage forms, with assured quality and affordability. The WHO Model List of
Essential Medicines 1 was first generated in 1977 and has been updated every
2 years since then.
This list of EMs includes a core list of minimum medicine needs for a
basic health-care system (that is, the most efficacious, safe and cost-effective
medicines for priority conditions that are selected based on current and estimated
future public health relevance), as well as a complementary list of medicines
for priority diseases for which specialized diagnostic or monitoring facilities,
specialist medical care and/or specialist training are needed. A number of human
plasma-derived medicinal products (PDMPs) – namely, factor VIII concentrate
and factor IX complex concentrate (coagulation factors II, VII, IX and X) – were
added to the 2nd edition of the complementary list of EMs in 1979, followed
by the addition of human normal immunoglobulin to the 15th edition in 2007.
In the 18th edition of the complementary list published in 2013, factor VIII
concentrate and factor IX complex concentrate were replaced with coagulation
factor VIII and coagulation factor IX respectively. Furthermore, anti-D, anti-
rabies and anti-tetanus immunoglobulins were added to the 19th edition of the
core list of EMs in 2015.
In the 2010 World Health Assembly resolution WHA63.12 concern
was expressed about the unequal levels of access globally to blood products,2
particularly PDMPs (also called plasma derivatives). Such inequality of access
left many patients without needed treatment, and many of those with severe
congenital and acquired disorders without adequate plasma-derivative
treatments. In this resolution, the World Health Assembly urged WHO Member
States:
...to take all the necessary steps to update their national regulations on
donor assessment and deferral, the collection, testing, processing, storage,
transportation and use of blood products, and operation of regulatory
authorities in order to ensure that regulatory control in the area of quality
and safety of blood products across the entire transfusion chain meets
internationally recognized standards.
See: http://www.who.int/medicines/publications/essentialmedicines/en/
1
The term “blood products” used in resolution WHA63.12 means blood, blood components and plasma
2
derivatives/PDMPs.
132
Annex 3
Requirements for implementing effective national blood regulation are described

in the WHO Assessment criteria for national blood regulatory systems (1).
In accordance with resolution WHA63.12, and in recognition of the fact
that achieving self-sufficiency in the supply of safe blood is an important national
goal in preventing blood shortages and meeting the transfusion needs of the
patient population, blood and blood components (whole blood, red blood cells
(RBCs), platelets and fresh frozen plasma) were added to the 18th edition of the
core list of the WHO Model List of Essential Medicines in 2013. Self-sufficiency
in this context means that the national needs of patients for safe blood and blood
components, as assessed within the framework of the national health system, are
met in a timely manner, and that patients have equitable access to safe blood
for transfusion, and that this can be accomplished by promoting voluntary non-
remunerated blood donation. Defining blood and blood components as EMs
(that is, as biological therapeutic products or simply “therapeutics”) could also
contribute to self-sufficiency by: (a) drawing attention to the role of national
governments in providing the necessary organizational and other support
required for assuring a safe and adequate blood supply; and (b) encouraging
countries to develop and ensure compliance with safety and quality standards,
as well as good practices, in product preparation for transfusion.
Assuring the quality, safety and availability of blood and blood
components is additionally linked to promoting self-sufficiency in essential
PDMPs. If more plasma is collected by apheresis or recovered from whole
blood than is needed for transfusion, it may be used as a starting material for
fractionation and thereby support self-sufficiency in PDMPs – provided that
the plasma meets required quality standards. As noted above, PDMPs such as
coagulation factors and human immunoglobulins have been recognized as EMs
since 1979 and 2007 respectively, and have been regulated in several countries as
biological therapeutic products for decades to ensure they meet internationally
recognized standards for safety, quality and efficacy. However, given the unequal
access globally to PDMPs, some countries still rely primarily on the use of whole
blood and plasma for various diseases and conditions that could be treated
with PDMPs (for example, fresh frozen plasma used instead of factor VIII and
factor IX for the treatment of patients with haemophilia A and B respectively),
contributing to an essential need for plasma components. Furthermore, plasma
components are used for the treatment of several plasma protein deficiency
diseases that are not treated with PDMPs.
Effective blood regulation is crucial for the establishment of blood
components as EMs. However, blood and blood components may not meet the
legal definition of medicines in all countries and this could have an impact on
the approach that must be taken to assure their quality, safety and availability
(compared to the approach employed for conventional medicines). Consequently,
in 2014, the International Conference of Drug Regulatory Authorities (ICDRA)
133
recommended that WHO undertake a project to provide guidance on the

management of blood and blood components as EMs. This project involved the
WHO Blood Regulators Network (BRN) in cooperation with the WHO Expert
Committee on Biological Standardization.
Blood and blood components are either prepared by blood establishments
and distributed to hospitals and other facilities, or prepared by hospital blood
banks for use in the treatment of various diseases – with the latter in some cases
being perceived as part of medical practice rather than the preparation of a
biological therapeutic product. There is therefore concern that blood and blood
components could be prepared in facilities (including hospitals) that are not
subject to appropriate regulatory oversight. Consequently, the regulatory system
should apply to all facilities.
In the context of blood and blood components for transfusion, quality
requirements for the preparation of blood components may in some jurisdictions
not be called “good manufacturing practices” (GMP) – for example in Europe
they are called “good practices”. However, in general, WHO recognizes, and
has developed, specific GMP for the preparation of blood components (2).
In this GMP document WHO defines the relevant aspects of quality system
requirements for blood establishments, including the relevant aspects of GMP
that are applicable and necessary for the preparation of blood components
for transfusion. In order to support the implementation of comparable
regulatory systems for blood components for transfusion, the alternative term
“good preparation practices” (GPP) will be used in the current document.
The implementation of GPP that are equivalent to the WHO GMP for blood
establishments (2) will ensure that blood components have similar safety and
quality profiles regardless of where they are prepared.

These WHO Guidelines are intended to provide a framework for establishing

regulatory oversight of blood and blood components for use in transfusion as
EMs. The underlying concept is that blood and blood components are biological
therapeutic products of human origin whose preparation should be subject
to regulatory standardization and oversight to assure their quality, safety and
efficacy. The framework provided in these Guidelines is similar to that which is
widely applied to the regulation of drugs produced under current GMP (cGMP)
but is adapted to address the specific attributes of blood and blood components
for transfusion that distinguish them from PDMPs and from pharmaceutical
medicines (drugs) in general. In jurisdictions where the legal frameworks in
place for medicines manufactured under cGMP for pharmaceuticals would not
apply to blood and blood components, parallel regulation based on the model
134
Annex 3
provided in these Guidelines would involve application of the analogous “GPP”

for such products.
The scope of these Guidelines includes elements that:
■■ make reference to resolution WHA63.12 (2010) regarding the
approach that must be taken to assure the quality, safety and
availability of blood and blood components for transfusion (see
section 1 above);
■■ clarify the specific nature of blood and blood components as
biological therapeutic products of human origin (see section 3 below);
■■ focus on the ethical aspects of blood donation, such as the need
to protect donors against exploitation, and to establish voluntary
non-remunerated donations of blood and blood components for
transfusion (see section 4.1 below);
■■ recognize the necessity to implement standards and controls, a
quality assurance system and good practices for blood and blood
component preparation (see sections 4.2 and 4.3 below);
■■ highlight the similarities and differences between blood and
blood components and conventional biological medicines and
biopharmaceuticals (see section 5 below);
■■ focus on the need to sustain nationally regulated blood systems (see
sections 6 and 7 below).
3. Blood and blood components as

biological therapeutic products
3.1 Historical background of blood transfusion
The first successful transfusion of human blood, as a treatment for postpartum
haemorrhage, was performed in 1818 by a British obstetrician, Dr James
Blundell, who drew blood from the patient’s husband and, to prevent the
blood from coagulating ex vivo, infused it directly into the patient. This was
followed by several technological advancements in transfusion medicine (3, 4),
which included:
■■ A number of discoveries in the early 1900s that led to the introduction
of blood typing, cross-matching and antibody identification in order
to prevent the immunological risks associated with blood transfusion.
■■ The development of blood banks in the early to mid 20th century
following the discovery that blood collected in anticoagulant
solution can be stored for several days when refrigerated.
135
■■ Developments in blood component manufacturing between the mid

and late 20th century, which included the use of interconnected,
sterile, disposable plastic containers for collection and preparation
of blood and blood components, collection by apheresis and the
storage of platelets at 22 ± 2 °C.
■■ The implementation of specific serological and nucleic-acid-based
tests for various infectious disease pathogens, such as human
immunodeficiency virus (HIV), hepatitis B virus (HBV), hepatitis C
virus (HCV) and syphilis in the mid to late 20th century, to reduce
transfusion-transmitted diseases.
Over time, blood collection and component preparation have become
increasingly complex, and currently include: (a) donor selection using
questionnaires to elicit risk factors for relevant transfusion-transmitted infections
(RTTI) (5); (b) aseptic collection (6); (c) laboratory testing, and quarantine
measures (2); (d) bacterial detection in platelets and pathogen reduction (7, 8);
and (e) the use of data-management software. Further information on these
aspects is provided below in section 4.
3.2 Indications for essential blood and blood component therapy

Human blood is a complex fluid which circulates in the vascular system and is
composed of plasma (the liquid portion which contains proteins and a variety
of small molecules) and cellular elements that include RBCs, white blood cells
and platelets.
Blood and blood components perform numerous vital functions in the
body (9, 10). Consequently, severe blood loss could result in life-threatening
conditions such as hypovolaemic/haemorrhagic shock, which requires immediate
blood transfusion in order to prevent organ failure and death. Blood transfusion
is also used as a supportive therapy for surgery, chemotherapy, and stem cell
and organ transplantation, as well as the treatment of serious acute and chronic
diseases caused by deficiencies or defects in plasma proteins or cellular blood
components, in order to avoid complications such as life-threatening haemorrhage
or to improve quality of life by reducing anaemia-related symptoms. As blood
systems developed, transfusion evolved from whole blood transfusion to targeted
therapy with specific blood components. This is because several of these diseases
are due to deficiencies or defects in a single blood component or plasma protein
(for example, abnormal or low RBC counts for anaemia (including abnormal
haemoglobin for thalassaemia); low platelet counts for thrombocytopenia; and
clotting factor deficiency for haemophilia). Plasma derived from whole blood or
apheresis can also serve as the starting material for the manufacturing of PDMPs.
In this regard, the transfusion of cellular blood components instead of whole
136
Annex 3
blood could serve to generate additional plasma for further manufacturing of

PDMPs, thereby providing one of the possible pathways towards self-sufficiency
in PDMPs. Examples of diseases and conditions that are treated with blood or
blood component transfusion are listed in Table A3.1 (11, 12).
The increasing global demand for access to safe blood and blood
components for transfusion has led to the manufacturing or development
of various types of equipment and tests used for their preparation. These
technological advancements, along with the large number of components
prepared annually, have resulted in a significant increase in the complexity of
blood and blood component preparation, which further underlines the need for
the development of standards for blood banking, and for the inspection of blood
establishments/banks to verify compliance with these standards. However, the
implementation of internationally accepted standards such as the WHO GMP
Guidelines (2) is currently not mandatory in all countries. Regulatory controls
should be established worldwide in order to enhance the safety and quality of
blood and blood components intended for transfusion.
Table A3.1
Examples of indications for use of essential blood and blood components
Blood/blood component Examples of indications a

Whole blood RBC replacement in acute active blood loss
with hypovolaemia.
Therapy in the indications below when no
specific blood components are available.
RBCs Supplement oxygen-carrying capacity (for
example, RBC replacement to treat symptomatic
anaemia; blood loss during surgical
intervention, trauma and haemolysis; and bone
marrow failure; and to support patients with
haemoglobinopathies).
Platelets Prevention or treatment of bleeding due
to platelet deficiency or dysfunction, or
massive blood loss (for example, in patients
with decreased platelet production due to
congenital or acquired bone marrow failure;
platelet-destructive conditions; dilutional
thrombocytopenia; and functionally abnormal
platelets).
137
Table A3.1 continued

Blood/blood component Examples of indications a
Plasma Management of patients who require multiple
coagulation factors (for example, bleeding
patients; and patients undergoing invasive
procedures).
Treatment of patients with clinically significant
coagulation abnormalities.
Treatment of patients with selected coagulation
factor or rare specific plasma protein
deficiencies for which a more appropriate
alternative therapy such as specific coagulation
concentrate or recombinant products is not
available.
Plasma exchange in patients with thrombotic
thrombocytopenic purpura.
In order to preserve factor VIII, plasma frozen
within 8 hours of collection is preferable for
indications requiring labile coagulations, or for
the preparation of cryoprecipitate for use in the
correction of factor VIII deficiency. Plasma frozen
within 24 hours could also be used.
a
Additional specific medical indications apply to further processed blood components such as washed and
irradiated components and cryoprecipitate.
3.3 Risks of blood and blood components

Blood transfusion carries the risk of transmitting infections if the donated blood
contains pathogens.
■■ As the collection of blood requires a venepuncture to be performed,
pathogenic bacteria could be transferred into the donation from a
contaminated skin area and subsequently proliferate (particularly in
platelets) to clinically significant numbers capable of causing an RTTI.
This risk must be minimized by the use of standardized and validated
techniques and disinfectants for aseptic venepuncture. Moreover,
thorough adherence to aseptic technique with closed systems and
appropriate microbiological sterility testing should be implemented.
■■ In the case of several pathogens causing severe disease (including
HIV, HBV and HCV) an exposed donor harbouring an RTTI may
feel well and wish to donate despite being at risk of transmitting
138
Annex 3
infections to patients. Therefore, it is crucial to: (a) collect blood

from voluntary non-remunerated donors, as they are known to have
lower rates of RTTI; (b) exclude from donating, through enquiry,
any person who has been at increased risk of acquiring such an
infection; and (c) test all donors for RTTI using validated assays that
have been approved by relevant regulatory authorities.
■■ Donors who have tested positive on a first or previous donation
must be systematically deferred – that is, the donation must
not enter the processing and testing cycle. To achieve this it is
recommended that: (a) a national blood donor registry (for example,
as part of the blood management system) is maintained at all points
of donation; and (b) donors are registered using a unique-identifier
system. Conditions for the potential re-entry of donors (for
example, after proven clearance of the infection or demonstration of
a false-positive test) may be defined.
There are also a number of adverse reactions due to immunological
mechanisms; the most relevant of these is blood group incompatibility. Therefore,
careful blood group typing and documentation is essential to avoid errors (for
example, giving the wrong blood to the patient).
The risks associated with blood transfusion necessitate traceability
from donor to patient and vice versa, and a system of haemovigilance – that
is, documenting and reporting adverse events and reactions, and initiating
corrective actions where necessary. The management of risks associated with
blood transfusion needs to be part of the quality management system developed
by the blood establishment.
4. Preparation of blood and blood components

4.1 Ethical aspects of blood donation
To ensure the safety of both donors and patients, transactions of human blood
and blood components should comply with the well-acknowledged principles of
biomedical ethics – namely, autonomy, beneficence, non-maleficence and justice.
Dignity also applies to donors in the sense of prohibiting the use of the human
body as a source of financial gain (13). Respecting the rights and ensuring the
safety and well-being of both donors and patients is fundamental. Consequently,
blood donation should begin with a consideration of a number of ethical issues,
which include:
■■ Encouraging voluntary non-remunerated blood donation which
serves as an important foundation for a safe and sustainable
blood supply.
139
■■ Providing information to the donor regarding the potential risks

associated with the donation, the risk of donating infected blood,
and the donor’s responsibility with respect to patient safety and
well‑being.
■■ Obtaining the donor’s consent to the donation and to the use of
the donation either for transfusion or for further manufacturing of
PDMPs. The donor must be mentally competent and the consent
given voluntarily. Collection of plasma for PDMPs should be
undertaken only after ensuring sufficient plasma for transfusion.
The use of blood and blood components for other purposes
should only be allowed when self-sufficiency in blood and blood
components for transfusion is already ensured. Their use in research
requires ethics approval and a separate and specific informed
consent. Under national laws, exceptions may apply in situations
where the donation is anonymized.
■■ Encouraging “non-directed donations” (that is, donations made
independently of the needs of a particular patient) in order to prevent
coercion by known donors/family members, as such coercion could
result in a reluctance to disclose behaviours associated with infectious
risks. An exception could be made for designated donations based
on medical reasons (for example, for patients with rare blood types
where no compatible non-directed donations are available).
■■ Minimizing the impact of deferrals on donors (for example, health
concerns, or feelings of rejection or discrimination) by educating
staff on donor-deferral criteria and communication to ensure they
are able to explain the reasons for deferral to donors and to follow
up with deferred donors as appropriate.
■■ Protecting donor health and safety during the collection of
blood and blood components and, if needed, dealing with donor

adverse reactions and obtaining medical care for the donor for an
appropriate period of time after the collection.
■■ Informing donors of abnormal test results and ensuring that
reactive infectious disease test results are confirmed and the donors
counselled with respect to further investigation and management by
an appropriately specialized physician.
■■ Protecting donors against exploitation.
■■ Avoiding incentives that could influence an individual’s decision
to donate.
■■ Protecting personal data and making them accessible only to
authorized personnel such as physicians or the responsible person.
140
Annex 3
4.2 Description of key product preparation steps

4.2.1 Donor suitability assessment
Blood and blood component preparation begins with the health screening of
carefully recruited donors. Risk-based health criteria and acceptable limits
should be taken into consideration during the donor-selection phase of the
donation procedure. The measurement of haemoglobin is essential and a
limited physical examination, including vital signs (for example, pulse, blood
pressure and temperature), may be performed either routinely or where the
donor’s condition raises suspicions of any possible anomaly, in accordance with
national standards. Each time donors donate, standardized donor-screening
questionnaires should be used to elicit information on their medical and
social history in order to determine that: (a) they are in good health and will
not be harmed by donating blood; and (b) they are not at increased risk of
infection with communicable bloodborne diseases. A confidential interview
should be conducted by trained personnel to clarify the answers obtained in
the questionnaires.
The standard operating procedures (SOPs) of the blood establishment
should specify the donor-exclusion criteria as well as the donor-deferral time
frames, taking into account both the specific local epidemiology and internationally
accepted standards and guidance, such as the Pan American Health Organization
guidelines on prospective donor education and selection (14). Donors should
also be informed about the necessity to provide post-donation information to the
collection facility on any illness or any other information relevant to the safety of
donated blood that was unknown prior to donating.
Donors should be tested for selected transfusion-transmissible infectious
agents such as HIV, HBV, HCV and syphilis to prevent the use of blood and
blood components from infected donors. The testing requirements for additional
infectious disease agents should be based on epidemiological data for the
geographical region in which the donations are made.
It should be noted that while donor suitability assessment significantly
reduces the risk of disease transmission to recipients, there are still concerns
about the residual risks that can result from: (a) limitations associated with
the donor-screening process (for example, inaccurate responses to screening
questions); (b) recent (“window-period”) infections (15); (c) assay failures;
(d) known pathogens for which testing is not performed; and (e) unknown
pathogens. Essential measures for maintaining and/or enhancing the safety of the
blood supply should be implemented, and should include quality management
as well as the continuous monitoring of new infectious disease threats and
timely implementation of appropriate risk-mitigation strategies involving donor
screening and/or infectious-disease testing.
141
4.2.2 Collection and component preparation

After local skin disinfection using a defined and validated disinfection procedure,
blood should be collected aseptically into single-use blood bags that meet a
suitable regulatory standard. The blood bags, which contain anticoagulant
solutions (and preservative/additive solutions where applicable), constitute a
closed system. The use of blood bags with diversion pouches can further reduce
the risk of contamination with skin microbiota by preventing the initial blood
flow from entering the blood bags.
Blood components may be prepared using either a manual or automated
procedure. The manual method involves the centrifugation of a unit of whole
blood at low speed to obtain RBCs and platelet-rich plasma (PRP), the transfer
of the PRP into a satellite blood bag and centrifugation of the PRP at high speed
to obtain the platelets and plasma. Alternatively, whole blood can be centrifuged
at high speed to obtain three layers consisting of RBCs, plasma and a buffy coat
containing platelets and leukocytes. The buffy coats derived from approximately
4–6 units are then pooled and centrifuged at low speed to separate the platelets
from the leukocytes. The RBCs, whole blood and platelet components should be
leukocyte reduced by the use of pre-storage filters. Leukocyte reduction is needed
to reduce the risk of:
■■ platelet refractoriness due to alloimmunization against human
leukocyte antigen (HLA) and platelet-specific antigens in multiply
transfused patients;
■■ febrile non-haemolytic transfusion reactions (FNTRs);
■■ transmission of leukocyte intracellular pathogens such as human
cytomegalovirus (HCMV);
■■ transmission of variant Creutzfeldt-Jakob disease.
The second method of component preparation is an automated

procedure that involves the use of apheresis machines that separate whole
blood into its components, transfer the desired components into containers and
return the remaining components to the donor. Some apheresis machines have
built-in leukocyte-reduction mechanisms.
Blood and blood components may also be subject to additional
processing steps such as: (a) pooling; (b) irradiation for the prevention of graft
versus host diseases (GvHD); (c) the use of filter systems for the reduction of
micro-aggregates; (d) washing to remove plasma; and (e) applying pathogen
reduction (currently using photochemical methods) to enhance safety from
infections. Donations from family members should be leukocyte-reduced and
irradiated to prevent transfusion-associated GvHD.
142
Annex 3
The SOPs used by blood establishments/banks should specify limits for

the volumes collected at each donation, as well as the frequency of donation, in
order to protect donor health and safety.
4.2.3 Additional testing

In addition to testing donors for RTTI, blood and blood components should
also be subject to the following testing:
■■ Each donation intended for transfusion should be tested for ABO

and RhD blood groups. Testing for red cell antibodies of potential
clinical significance is also recommended, particularly for first-time
donors and donors with a history of pregnancy or transfusion since
their last donation. Additional testing is required for specialized
products such as HLA-matched and phenotyped components.
■■ Quality control testing should be performed on a statistically
based proportion of components to ensure ongoing assessment of
the quality of the procedures used for product preparation. The
frequency of quality control testing, the test parameters (for example,
haemoglobin, haematocrit, platelet count, factor VIII concentration
and sterility) and acceptance criteria should all be established for
each type of component. Test results should be analysed on an
ongoing basis and appropriate corrective action taken when values
deviate from acceptable limits. Bacterial detection in platelets may
also be performed. Note: in some countries each platelet component
is subject to bacteriological testing for the detection of bacterial
contamination.
4.2.4 Labelling
The “labelling” of blood and blood components refers to both information
appearing on the direct product label and/or contained in accompanying
documentation. More specifically, the product label should include the product
type, blood groups, a unique donation code that is traceable to the donor, the
site of product preparation, the list of pathogens for which discretionary testing
is performed (for example, cytomegalovirus or hepatitis E virus), the storage
conditions and expiry date (and time of day if applicable). Standardized labels
that can be universally read (such as those printed using machine-readable ISBT
128 standard terminology) should be used. The list of pathogens for which testing
was performed and found to be negative should appear either on the product
label or in accompanying documentation.
143
4.2.5 Storage
Blood and blood components should be stored under specified conditions
in order to maintain their safety and quality. Units determined to be safe and
released for transfusion should be segregated from untested units, and access to
storage areas should be restricted to designated personnel. Plasma components
should be frozen within a specified period after collection (preferably within
8 hours for fresh frozen plasma, or within 24 hours). Whole blood and RBCs
should be refrigerated (at 1–6 °C) and platelets should be stored at 20–24 °C
under agitation.
4.2.6 Distribution and shipping

To ensure the safety and quality of blood and blood components they should be
formally released to hospitals for further storage in hospital blood banks, or for
transfusion, after verifying that they meet all safety and quality standards, and are
appropriately packaged prior to transportation. The shipping containers should
be validated to maintain acceptable storage conditions for the blood and blood
components.
4.2.7 Haemovigilance
4.2.7.1 Documentation
There should be a documentation system that assures bidirectional traceability
of blood components between donors and patients as a foundation of
haemovigilance.
4.2.7.2 Adverse reaction reporting and investigating

There should be a system in place for reporting and investigating serious donor
reactions, and serious or unexpected adverse recipient reactions reported by
hospitals. In the case of recipient reactions, measures should be taken to notify
those in possession of co-components when applicable, and to quarantine and/

or recall the co-components. NRAs should also be notified as required. System-
wide corrective actions should be implemented where feasible and appropriate.
4.2.7.3 Look-back and trace-back

Blood establishments should have a look-back procedure in place in order
to identify previous donations (and related blood components) from a
donor who, on subsequent testing, is confirmed positive for a transfusion-
transmissible infectious agent – and to identify recipients who received blood
or blood components from a donor who is later confirmed positive for such an
infectious agent.
Trace-back procedures should also be established to investigate any
report of a suspected RTTI in order to: (a) identify a potential implicated donor;
144
Annex 3
(b) determine whether any donor who contributed to the transfusion is infected
with (or positive for serological markers of) the implicated infectious agent;
(c) trigger a recall of in-date blood or blood components contributed by that
donor; and (d) notify consignees and recipients of components collected from
that donor. The NRA should also be notified as required.
4.2.8 Good preparation practices/quality systems

4.2.8.1 Key requirements
It is recommended that blood establishments/banks comply with relevant
elements of GPP to assure the quality and safety of blood and blood components
(2). These elements include:
■■ organization and personnel (including training);
■■ maintenance of facilities/premises;
■■ equipment qualification, calibration and maintenance;
■■ quality control programme for products, supplies and services;
■■ donor selection, blood collection, testing, processing, storage and
distribution, and record-keeping;
■■ SOPs containing step-by-step instructions for all activities
undertaken during product preparation, as well as specifications
for the resulting blood components;
■■ process validation;
■■ change control;
■■ corrective and preventive measures;
■■ quality monitoring;
■■ management of risks, documentation, nonconformities, audits
and contracts.
4.2.8.2 Nonconformity and deviation reporting and investigating

There should be a system in place to ensure that any nonconformities and
deviations that occur during blood and blood component preparation are
documented, investigated for their causative factors and followed up by corrective
actions. This should include a system for notifying those in possession of the
implicated products (and the NRA, if applicable), and for quarantining and/or
recalling products whose safety may have been compromised.
4.3 Associated substances and equipment

Associated substances and equipment used during blood and blood component
preparation include:
145
■■ anticoagulant solutions and additive solutions for RBCs and platelets;

■■ blood pressure and pulse monitors, thermometers, haemoglobin
analysers, etc. that are used to assess donor health;
■■ apheresis equipment, automated blood processors, blood bag
collection systems, centrifuges, automated red cell washers, gamma
and X-ray irradiators, sterile connection devices, automated blood
extractors, plasma freezers, etc. that are used for blood and blood
component collection and/or processing;
■■ in vitro screening test kits used for donor testing, and systems for
microbial detection, compatibility testing and quality control testing
(including automated systems);
■■ pathogen-reduction technology systems;
■■ computerized blood management systems, specifically systems that
analyse data regarding the suitability of blood and blood components
for transfusion (note: the classification of blood management systems
as medical devices depends on the specifications of the product and
the national medical devices legislation).
These substances and equipment are generally regulated as medical
devices, except for the anticoagulant and additive solutions, which may be
regulated as either drugs or devices. Blood establishments/banks need to ensure
that the materials and devices being used for the preparation of blood and blood
components are approved by their regulatory authorities. Furthermore, even
though device manufacturers are responsible for the validation of the software
in automated devices, in some cases additional validation is required prior to
implementation – particularly when the equipment needs to be programmed
according to the specific needs of the blood establishment/bank. This further
complicates the preparation process for blood and blood components and
underscores the need to comply with internationally recognized standards.
5. Comparison of blood components with PDMPs

5.1 General
These WHO Guidelines propose the regulation of blood and blood components
under GPP, which consist of cGMP that have been adapted to address the
attributes of blood and blood components that distinguish them from PDMPs.
PDMPs may already be regulated as medicines under an existing framework.
The following sections highlight both the similarities and the differences
between blood and blood components and PDMPs to assist in determining
quality requirements that could be applied to blood and blood components.
146
Annex 3
5.2 Product safety and quality

Conventional biological medicines are typically manufactured on an industrial
scale using complex proprietary processes that vary between manufacturers.
One example is the manufacturing of PDMPs, which may involve, among
other steps: (a) the pooling of thousands of plasma units; (b) the concentration
and/or purification of one or more plasma proteins using methods such
as cryoprecipitation and various fractionation procedures that utilize
chromatographic, precipitation and filtration techniques; (c) viral inactivation/
removal techniques to enhance product safety; and (d) formulation, filling and
lyophilization. In-process testing is performed at various steps to monitor the
manufacturing process, and final product testing of each lot is performed to
ensure that product specifications are met.
The manufacturing of PDMPs is similar to that of other
biopharmaceuticals with respect to the complexity of the manufacturing process
and its potential impact on the biological characteristics of the final products.
Thus, like other biopharmaceuticals, PDMPs are subject to GMP regulations to
ensure the products are consistently safe, efficacious and of high quality.
The preparation of blood and blood components differs from PDMP
manufacturing in that: (a) closed single-use systems are used for product
preparation to reduce the risk of contamination/cross-contamination; (b) each
component is derived from one donation or from a limited number of
donations; and (c) in some cases, the preparation techniques employed are
limited to mechanical or physical methods such as centrifugation, separation
and cryoprecipitation (for cryoprecipitates). Additional methods such as
leukocyte reduction, pooling, washing, irradiation and photochemical methods
for pathogen inactivation are also employed. Consequently, blood and blood
components can be produced at various facilities, ranging from large blood
establishments to small hospital blood banks. There are concerns that not all
blood establishments/banks are regulated by an NRA. Even in settings where
unregulated blood establishments/banks have adopted manufacturing standards
developed by professional organizations, there is no mechanism for verifying
compliance with these standards.
Notwithstanding the differences in the complexity of the processes used
for the manufacturing of PDMPs and those used for the preparation of blood
and blood components, there are also similarities with respect to the following:
■■ a reliance upon the availability of healthy donors and the need to

protect donors;
■■ the risks associated with RTTI and the donor screening and testing
measures required to mitigate these risks;
147
■■ the importance of linking the donor with each lot of product

manufactured or prepared through appropriate labelling and record-
keeping to facilitate recalls and, where applicable, look-backs and
trace-backs;
■■ the need to validate new or modified procedures employed for
product manufacturing or preparation;
■■ the use of appropriately validated automated systems, particularly
when there is a need to track a large number of donors/donations
and the results of their screening and infectious-disease tests;
■■ the need for segregation and holding (that is, quarantine) of
donations/products until they are released for distribution to prevent
the release of potentially unsafe products;
■■ the need for product storage and transportation at appropriate
temperatures and conditions.
These similarities lead to the underlying concept that blood and blood
components should be prepared within a quality management system based
on the principles of GMP (adapted to blood and blood components) when
relevant and appropriate – and which includes elements such as the testing of
starting materials, in-process quality testing and controls (for example, bacterial
detection and other quality control tests), labelling that reflects product
identity and assures traceability, and adverse event reporting (see section
4.2.8 above). Consequently, the regulation of blood and blood components as
biological therapeutic products would ensure the consistent implementation of
appropriate standards for product quality, safety and efficacy. Such regulation
would apply to all blood establishments/banks involved in the preparation of
these products.
5.3 Product efficacy

As with other biopharmaceuticals, PDMPs are subjected to clinical trials in the
target population to establish their safety and efficacy before approval for clinical
use. Such trials are typically not required for conventional blood and blood
components because: (a) their efficacy has been established through historical
use; and (b) they are prepared and stored using established procedures that
are published in standards developed by professional organizations. However,
clinical trials are currently required for blood and blood components when
they are prepared using new technologies or processing steps (for example,
pathogen-reduction technologies) as these could potentially alter their biological
characteristics.
148
Annex 3
6. The blood regulatory system

6.1 Guiding principles
The management of blood and blood components as EMs should take into
consideration the need to:
■■ sustain nationally regulated self-sufficient blood systems;
■■ protect donors against exploitation and prohibit financial gain;
■■ base blood and blood component standards and controls on a
quality management system derived from GPP in order to assure the
quality, safety and availability of these products;
■■ ensure that the regulations for blood and blood components and for
PDMPs are complementary, and incorporate the essential elements
and core functions specified in the WHO Assessment criteria for
national blood regulatory systems (1).
6.2 Regulatory framework

6.2.1 General
Blood and blood components should be controlled under an appropriate
regulatory system in order to promote and enhance their quality, safety and
availability. The elements and functions of an effective national blood regulatory
system have been described by WHO and are applicable both in developed and
developing countries (1).
The regulatory system should consist of a regulatory framework
administered by an NRA that is responsible for regulating the activities associated
with the preparation of these products. Regulatory frameworks consist primarily
of legislative instruments such as legislation (or act) and regulations that can be
supplemented by non-legislative instruments such as policies, guidelines and
guidance documents. Collectively, these instruments allow for the categorization
of risk to an appropriate level of control and the capacity to respond quickly to
rapid technological advances, while providing the required authority and capacity
to take immediate action during crises and emergencies.
6.2.2 Legislation
The legislation or law serves as the first level of a comprehensive regulatory
framework and provides a legal basis for the establishment of a regulatory system.
A law is needed that governs the preparation of blood and blood components,
as well as the use of associated substances and relevant medical devices. The
law should define the scope of regulations and provide the legal authority for
149
their development. The following are examples of the kinds of provisions that
could be included in legislation:
■■ Definition of the therapeutic products and devices to be regulated.
■■ Prohibitions that prevent the preparation or sale of potentially
unsafe products (for example, products that are adulterated or
prepared under unsanitary conditions).
■■ Assignment of an NRA with legal powers to administer, enforce
and verify compliance with the legislation and regulations (for
example, powers for inspection, seizure and forfeiture and for the
establishment of a list that sets out the classes of products to be
regulated).
■■ The offences and punishment of persons who deliberately
contravene the legislation or regulations.
■■ Definition of the areas for which regulations should be developed
and granting of the authority to develop the regulations necessary
for carrying the purposes and provisions of the legislation into effect.
These areas should include those covered below in section 6.2.3.
Detailed guidance regarding provisions that could be included in
national acts or legislation may be found in the list of documents provided in
the Appendix to these Guidelines.
6.2.3 Regulations
Regulations form the second level of the regulatory framework. They are
developed under the authority of the legislation and serve to interpret the
legislation and to provide policies and standards/technical requirements that are
legally binding.
Regulations can be developed using different approaches. In the

traditional risk-management approach, good practices and standards are written
directly into regulations. The process for developing and amending regulations
can be lengthy and can take up to several years in some jurisdictions, thus
making it difficult to ensure that they remain current with regard to technological
advances and emerging threats.
An alternative and more flexible approach is the development and use
of standards that are not directly incorporated into regulations, but can be
referenced in the regulations. For example, instead of specifying the requirements
for donor screening and infectious-disease testing in regulations, the sections
of voluntary or mandatory national or internationally recognized standards
containing these requirements could be referenced in regulations to give them
the force of law. Since the standards are a stand-alone document, they can be
150
Annex 3
amended rapidly when required without amending the regulations themselves.

This approach is particularly useful for standards/technical requirements that
are likely to require frequent amendments in response to rapid technological
advances and emerging threats.
The incorporation of standards into regulations by reference may be
achieved using one of the following approaches:
■■ Static or fixed – this approach references requirements in a specific
version of a standard at a defined date to ensure that amendments to
the standard do not automatically become part of the regulations. In
this approach, a regulatory amendment will be required to reference
subsequent versions of the standard.
■■ Ambulatory or flexible – this approach references requirements
in the standard as amended from time to time to automatically
make any amendments part of the regulations. In this approach,
the regulations do not need to be amended to reference subsequent
versions of the standard.
This standards-based approach to regulation could be adopted, at least
in part, for blood and blood components for which the procedures used for their
preparation and storage are well established.
Regulations for blood and blood components should focus on managing
risk in four key areas:
■■ protection of donor health and safety;
■■ prevention of infectious-disease transmission from donors
to recipients;
■■ prevention of adverse reactions due to immunological mechanisms
in transfusion recipients;
■■ prevention of improper handling or processing that could affect
product safety, efficacy and quality.
This can be accomplished by including requirements for the following
elements in blood regulations:
■■ Standards for the collection and processing of blood and blood
components, which include the methods used for their preparation
(see section 4 above).
■■ GPP (consistent with GMP in some jurisdictions) to assure the
quality, safety and availability of these products.
■■ The use of test kits, blood-collection sets, anticoagulant/additive
solutions and other collection equipment that have been approved
by the NRA.
151
■■ Importation and exportation of blood and blood components

– although self-sufficiency in blood and blood components is a
basic principle.
■■ The definition of clinical trials and the requirement for clinical trials
of blood and blood components prepared using new technologies
(for example, pathogen-reduction technologies) that could
potentially alter their biological characteristics.
■■ Pre-approval of applications/submissions to determine if the data
submitted support the claims made for product safety and quality
and, if applicable, for efficacy (this may include on-site evaluations
of the facility and processes used for product preparation).
■■ The issuance of authorization by the NRA to carry out product
preparation activities.
■■ The review of applications/submissions for post-approval changes.
■■ The submission of applications for (or amendments to) registration,
accreditation or blood establishment licensing.
■■ The registration of blood establishments/banks and importers
or the issuance of licences to such facilities based on evidence of
compliance with GPP.
■■ The authority of the NRA to issue, refuse, suspend, reinstate or cancel
an authorization, registration or accreditation, or facility licence.
■■ The provision of information to the NRA regarding serious
reactions in donors and recipients by the holders of blood and blood
component registration or authorization and licences.
■■ The performance of risk–benefit evaluation and investigation of the
root cause of nonconformity, deviation and adverse events reports.
■■ Powers of inspectors, which allows for the performance of
compliance and enforcement activities such as inspection of blood

establishments/banks to assess compliance with regulatory
requirements, investigation of nonconformities and follow-up
of corrective actions.
Consideration should be given to the adoption of internationally
recognized standards, such as the examples cited in these Guidelines (2, 5,
6, 14), that set out detailed requirements for the activities described in section 4
above. Regardless of the approach taken, all stakeholders should be given an
opportunity to comment on the regulations before they are finalized.
There should also be regulations that define “investigational test” and
that require the investigational testing and pre-approval of applications for
associated substances (such as anticoagulant, additive and preservative solutions)
152
Annex 3
and relevant medical devices (such as in vitro screening and diagnostic test kits,
blood-collection equipment and blood bag systems) that are used during the
preparation of blood and blood components. Systems should be put in place to
ensure compliance with these regulations.
6.2.4 Non-binding instruments

The third level of the regulatory framework consists of policies, guidance
documents/guidelines and voluntary standards that can be used to supplement
regulations. Typically, these documents may be simpler and faster to introduce
than regulations, and can be used to interpret regulations and/or provide details to
blood establishments/banks on how to meet regulatory requirements. Since they
are not legally binding, they allow flexibility with respect to their interpretation
and are adaptable to change. However, if a failure to implement these non-
binding instruments were to result in a serious adverse event, the blood facility
in question would need to explain why the guidance was not followed.
6.3 The regulatory authority

6.3.1 Organization of the regulatory authority
The regulatory authorities in different countries may currently be organized at a
local, regional or national level. The establishment of regulatory authorities at the
local or regional level could lead to differences in the standards and regulatory
requirements applied to blood and blood components, as well as in the level
of regulatory oversight. While recognizing that huge difficulties exist in some
regions, it is recommended that countries move towards the establishment of
an NRA in order to ensure consistency across the country in both regulatory
requirements and oversight.
6.3.2 Functions of the NRA

The key functions of the NRA with respect to blood and blood components are
described in the WHO Assessment criteria for national blood regulatory systems
(1). Such functions include the development of regulations and standards for
the preparation of blood and blood components, and the provision of regulatory
oversight to verify compliance with regulatory requirements (for the individual
elements of such requirements see section 6.2.3 above).
The WHO assessment criteria document also contains additional
information on: (a) the essential elements necessary to establish the legal basis,
authority and general characteristics of an NRA; (b) the core functions of an
NRA required for the comprehensive oversight of blood and blood components;
and (c) major criteria, indicators and associated ratings to assist NRAs in
assessing their performance and identifying areas for improvement.
153
7. The blood supply system

7.1 Organization of the blood supply system
The blood supply systems in different countries may currently be organized at
a local, regional or national level with respect to blood collection, testing and
processing. The establishment of blood systems at the local or regional level
could lead to differences in the implementation of standards and regulatory
requirements, and consequently to blood and blood components with different
safety and quality profiles. Where possible, it is recommended that countries
move towards a nationally regulated and coordinated blood supply system in
order to: (a) harmonize procedures and best practices at the national level; and
(b) provide assurance that blood and blood components from different areas
are of equivalent safety and quality and thereby facilitate the exchange of these
products across the country.
7.2 Functions of blood establishments/banks

The blood supply system consists of blood establishments/banks that collect,
test, process (including washing, pooling and irradiation) and distribute whole
blood and blood components intended for transfusions, as well as plasma
intended for further manufacturing into PDMPs. Such facilities are responsible
for: (a) performing the activities described in sections 4.1, 4.2 and 4.3 above; and
(b) implementing the regulations and standards developed by NRAs for these
activities. All facilities that perform these activities (including hospital blood
banks that prepare blood and blood components for use within their hospitals)
should implement the regulations and standards developed by the NRA.
8. The blood transfusion system

The blood transfusion system consists of care centres (hospitals, surgical centres
and outpatient facilities; and sometimes ambulances) that utilize blood and
blood components for the treatment of patients. Such centres are responsible for
carrying out the following activities:
■■ storing blood and blood components at appropriate temperatures
and conditions;
■■ developing appropriate procedures for further processing of the
blood and blood components prior to transfusion – for example,
pooling, washing and irradiation, where applicable;
■■ appropriate pre-transfusion testing of patients and cross-matching
to ensure compatibility of the blood component to be transfused;
154
Annex 3
■■ maintaining appropriate records to ensure that blood components

can be traced to their recipients and from recipients back to
their donors;
■■ documenting and investigating nonconformities and deviations
related to the handling of blood and blood components;
■■ quarantining of blood and blood components that are under
investigation by the blood establishments/banks and hospitals;
■■ reporting adverse events and reactions that are related to the quality
of blood components to the blood establishments/banks;
■■ investigating, evaluating and documenting all adverse transfusion
reactions;
■■ ensuring the appropriate use of blood and blood components
by clinicians.
9. Stepwise implementation of a nationally

regulated blood system
The implementation of a nationally regulated blood system is fundamental to
assuring the quality, safety and availability of blood and blood components in
accordance with their listing as EMs. A risk-based strategy is recommended
when considering the development of a regulatory model for the blood system
and a national roadmap for its implementation.
It is recognized that, when implementing a nationally regulated blood
system, the starting situation may vary considerably from one country to
another. In some countries, the blood system may be fragmented and central
coordination completely lacking, whereas, in other countries, national or
regional bodies may perform a coordinating function within the blood services.
In any case, the political commitment of the ministry of health is necessary
to establish a roadmap for implementing a nationally regulated blood system.
The main elements of this roadmap should be developed and agreed upon in
cooperation with the key stakeholders. This implementation plan may also
incorporate an initial review and improvement of the existing structure of the
blood system in a country. The assignment of the main tasks for implementing
this roadmap and the role and mandate of key personnel should be defined and
agreed upon as part of the political process. In any case, a cooperative and step-
by-step process to restructure the blood system in a country (if needed) and to
implement a nationally regulated blood system is encouraged in order to foster
more success over time.
The key stakeholders in the blood system (that is, the blood regulatory
system, the blood supply system and the blood transfusion system) should
155
be involved from the beginning in order to understand and define their
individual responsibilities and expected contributions. Regular interaction
among stakeholders is essential.
The legislative body should define a legal framework (regulation)
applicable to blood and blood components. This would include assigning the
NRA to oversee all institutions and health-care professionals supplying blood
and blood components.
An NRA is an essential element of a regulatory system. The decision to
adopt a particular regulatory model should take into account existing regulatory
structures, capacities and expertise. Establishing the regulation of blood and
blood components under the NRA for medicines may be the most effective
and rapid way to accomplish this in settings where blood regulation is otherwise
lacking. Regulatory frameworks for blood and blood components and for PDMPs
should be complementary.
Blood establishments, other related health institutions, and health-care
professionals supplying blood and blood components for transfusion should
be engaged and their experience used to inform the establishment of standards
and procedures. This may result in improving the existing structure of the blood
system. The initial use of existing standards as a starting point for establishing a
common language between all key players may provide an acceptable approach
for all parties.
Where applicable, representatives of the plasma fractionation industry
should be invited to participate as additional key players to support this process.
This should ensure that appropriate standards are implemented and that the
quality of surplus plasma as a starting material for further manufacturing will
meet the necessary requirements.
The development of national blood standards covering donor-selection
criteria, infectious disease marker testing strategies, quality system requirements
and standards for the final products (specifications and/or monographs) will
be an essential step. Both during and after the legislative process, these initial
standards may continuously be improved and implemented on a national
basis. The establishment of such standards should take into account existing
national and international guidelines such as the WHO guidelines on good
manufacturing practices for blood establishments (2). As soon as possible,
blood establishments should apply these standards in a consistent manner by
implementing appropriate procedures (including SOPs and training) within
their quality system.
A parallel national implementation process and regular interaction
among the key players will be essential in accelerating the implementation
of standards and regulatory functions. Since the process of implementing a
regulatory system and reaching an acceptable compliance status may take
several years, a political decision will be required to define the time frames
156
Annex 3
for reaching full compliance with standards and for effective enforcement
of regulations by the NRA. A possible stepwise implementation plan for a
nationally regulated blood system is outlined in Fig. A3.1.
Fig. A3.1
Stepwise implementation plan for a nationally regulated blood system
• Development of a legal framework and

characterization of an NRA
System design • Development and/or adoption of blood standards
and development • Review/improve organization, infrastructure and
sustainable funding mechanisms
• Interaction among stakeholders
• Establishment of regulatory functions including

empowerment of the NRA
Implementation
• Capacity-building and training
and validation
• Interaction among key players
• Implementation of blood standards
• Achievement of compliance including enforcement

by the NRA
Performance
• Full performance of regulatory functions
and enforcement
• Maintaining blood standards
• Increasing availability and ensuring supply

The development of these WHO Guidelines was initiated in 2014 following a
recommendation made by the International Conference of Drug Regulatory
Authorities (ICDRA) to the WHO Blood Regulators Network (BRN). The BRN
(Chair: Dr C. Schaerer) then undertook the project in cooperation with the
WHO Expert Committee on Biological Standardization.
The BRN members involved in this project were Dr F. Agbanyo and
Dr L. Elmgren, Health Canada, Canada; Dr J. Epstein and Dr G. Michaud,
United States Food and Drug Administration, the USA; Professor R. Seitz, Dr M.
Heiden and Dr A. Hilger, Paul-Ehrlich-Institut, Germany; Dr C. Schaerer and
Dr M. Jutzi, Swissmedic, Switzerland; Dr G. Smith and Dr I. Prosser, Therapeutic
157
Goods Administration, Australia; Dr I. Sainte-Marie and Dr W. Oualikene-

Gonin, Agence Nationale de Sécurité du Médicament et des Produits de Santé,
France; and Dr I. Hamaguchi and Dr T. Kondo, Ministry of Health, Labour and
Welfare, Japan. The BRN secretariat at WHO consisted of Dr C.M. Nuebling
and C. Pasztor.
Further input was provided by Dr P. Strengers, International Plasma
Fractionation Association; and Dr H. Klein, National Institutes of Health, the
USA, prior to public consultation.
During the public consultation phase (11 July–26 September 2016)
comments on the draft WHO Guidelines were received from the following
organizations: European Blood Alliance (EBA); International Council for
Commonality in Blood Bank Automation (ICCBBA); International Plasma
Fractionation Association (IPFA); International Society of Blood Transfusion
(ISBT); Plasma Protein Therapeutics Association (PPTA). In addition, comments
were received from the following individuals: M d P Alvarez Castello, CECMED,
Cuba; Dr S. Hindawi, Saudi Society of Transfusion Medicine, Saudi Arabia;
Dr M-L. Hecquet and Dr G. Rautmann, European Directorate for the Quality
of Medicines & HealthCare, France; Dr F. Moftah, Egyptian Society of Blood
Services, Egypt; Dr G. Praefcke, Paul-Ehrlich-Institut, Germany; Dr N. Prunier,
Dr T. Schneider and Dr G. Folléa, Etablissement Français du Sang, France; Dr M.
Ruta, Dr B. Peoples, Dr W. Paul, Dr O. Illoh and Dr E. Storch, United States Food
and Drug Administration, the USA; Dr B. Sorensen, Danish Society of Clinical
Immunology, Denmark; Dr K. Tadokoro, Japanese Red Cross Society, Japan;
J. Yu, World Health Organization, Switzerland.
The BRN reviewed all comments received for incorporation into the
revised final draft document WHO/BS/2016.2285.
References
1. Assessment criteria for national blood regulatory systems. Geneva: World Health Organization;
2012 (http://www.who.int/bloodproducts/NationalBloodRegSystems.pdf, accessed 26 December
2016).
2. WHO guidelines on good manufacturing practices for blood establishments. In: WHO Expert
Committee on Specifications for Pharmaceutical Preparations: forty-fifth report. Geneva: World
Health Organization; 2011: Annex 4 (WHO Technical Report Series, No. 961; http://www.who.int/
bloodproducts/publications/GMP_Bloodestablishments.pdf?ua=1, accessed 26 December 2016).
3. Highlights of transfusion medicine history [website]. Bethesda (MD): American Association of
Blood Banks (http://www.aabb.org/tm/Pages/highlights.aspx, accessed 26 December 2016).
4. History of blood transfusion [website]. Washington (DC): American Red Cross (http://www.
redcrossblood.org/learn-about-blood/blood-transfusions/history-blood-transfusions, accessed
26 December 2016).
158
Annex 3
5. Blood donor selection: guidelines on assessing donor suitability for blood donation. Geneva:
World Health Organization; 2012; chapter 7 (http://www.who.int/bloodsafety/publications/
guide_selection_assessing_suitability.pdf, accessed 26 December 2016).
6. WHO guidelines on drawing blood: best practices in phlebotomy. Geneva: World Health
Organization; 2010 (http://www.who.int/injection_safety/phleb_final_screen_ready.pdf, accessed
26 December 2016).
7. ‘Sterility testing of blood components and advanced therapy medicinal products’ (Munich,
April 29, 2010) organized by the DGTI section ‘Safety in hemotherapy’ – Abstracts. Transfus
Med Hemother. 2011;38(5):337–40 (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3364038,
8. Bacterial risk control strategies for blood collection establishments and transfusion services to
enhance the safety and availability of platelets for transfusion. Silver Spring (MD): US Food and
Drug Administration; 2016 (http://www.fda.gov/downloads/Guidances/Blood/UCM425952.pdf,
9. Blood basics [website]: Washington (DC): American Society of Hematology (http://www.
hematology.org/Patients/Basics/, accessed 26 December 2016).
10. What does blood do? [website]. Bethesda (MD): PubMed Health (http://www.ncbi.nlm.nih.gov/
pubmedhealth/PMH0072576/, accessed 26 December 2016).
11. Circular of Information for the use of human blood and blood components. Revised November
2013. Bethesda (MD): American Association of Blood Banks (https://www.aabb.org/tm/coi/
Documents/coi1113.pdf, accessed 26 December 2016).
12. Circular of Information for the use of human blood components. Ottawa: Canadian Blood
Services (https://www.blood.ca/en/hospitals/circular-information, accessed 26 December 2016).
13. Folléa G. Donor compensation and remuneration – is there really a difference? ISBT Science
Series. 2016;11(S1):3–9 (http://onlinelibrary.wiley.com/doi/10.1111/voxs.12188/full, accessed
26 December 2016).
14. Eligibility for blood donation: recommendations for education and selection of prospective
blood donors. Washington (DC): Pan American Health Organization; 2009 (http://www1.paho.
org/hq/dmdocuments/2009/EligiBlood09EN.pdf?ua=1, accessed 26 December 2016).
15. Guidelines on estimation of residual risk of HIV, HBV or HCV infections via cellular blood
components and plasma. In: WHO Expert Committee on Biological Standardization: sixty-seventh
report. Geneva: World Health Organization; 2017: Annex 4 (WHO Technical Report Series, 1004).
159
Appendix
Examples of existing legislation, regulations and guidance
The documents listed here are provided as examples of existing legislation,
regulations and guidance that may be helpful in establishing a national
regulatory framework.
Country Relevant legislation, regulations and guidance

or region
Canada 1. Food and Drugs Act (1985): http://laws-lois.justice.gc.ca/eng/
acts/f-27/ (accessed 26 December 2016).
• This Act applies to food, drugs (including blood and blood

components), cosmetics and devices.
2. Blood Regulations: http://laws-lois.justice.gc.ca/eng/regulations/

SOR-2013-178/index.html (accessed 26 December 2016).
• These are stand-alone regulations for blood and blood

components intended for transfusion and further manufacturing,
and were developed under the authority of the Food and
Drugs Act.
3. Guidance Document: Blood Regulations: http://www.hc-sc.gc.ca/

dhp-mps/brgtherap/applic-demande/guides/blood-reg-sang/
blood-guid-sang-ligne-eng.php (accessed 26 December 2016).
4. Food and Drug Regulations:

http://laws-lois.justice.gc.ca/eng/regulations/C.R.C.,_c._870/
index.html (accessed 26 December 2016).
• These regulations apply to food and drugs. The requirements

for drugs (which may apply to associated substances such
as anticoagulants/additive solutions) can be found in Part C,
Divisions 1, 1A, 2, 5 and 8 of these regulations.
5. Medical Devices Regulations:

http://laws-lois.justice.gc.ca/eng/regulations/SOR-98-282/index.
html (accessed 26 December 2016).
• These regulations apply to medical devices, for example,

infectious disease test kits, blood-collection sets and apheresis
equipment.
160
Annex 3
Table continued
Country Relevant legislation, regulations and guidance
or region
Europe 1. Directive 2002/98/EC of the European Parliament and of the
Council of 27 January 2003 – setting standards of quality
and safety for the collection, testing, processing, storage and
distribution of human blood and blood components and
amending Directive 2001/83/EC: https://ec.europa.eu/health/sites/
health/files/files/eudralex/vol-1/dir_2002_98/dir_2002_98_en.pdf
(accessed 26 December 2016).
2. Commission Directive 2004/33/EC of 22 March 2004 –
implementing Directive 2002/98/EC of the European Parliament
and of the Council as regards certain technical requirements
for blood and blood components: http://www.ema.europa.eu/
docs/en_GB/document_library/Regulatory_and_procedural_
guideline/2009/10/WC500004484.pdf (accessed 26 December
2016).
3. Commission Directive 2005/62/EC of 30 September 2005 –
and of the Council as regards Community standards and
specifications relating to a quality system for blood establishments:
http://www.ema.europa.eu/docs/en_GB/document_library/
Regulatory_and_procedural_guideline/2009/10/WC500004486.pdf
(accessed 26 December 2016).
4. Commission Directive 2005/61/EC of 30 September 2005 –
and of the Council as regards traceability requirements and
notification of serious adverse reactions and events: http://www.
ema.europa.eu/docs/en_GB/document_library/Regulatory_and_
procedural_guideline/2009/10/WC500004485.pdf (accessed
26 December 2016).
5. Council of Europe (EDQM) Recommendations & Resolutions:
https://www.edqm.eu/en/blood-transfusion-recommendations-
resolutions-71.html (accessed 26 December 2016).
The USA 1. United States Food and Drug Administration; subchapter F –
Biologics: http://www.ecfr.gov/cgi-bin/text-idx?SID=7f01bf
1d25c364e2d287f227cd6833c8&mc=true&tpl=/ecfrbrowse/
Title21/21CIsubchapF.tpl (accessed 26 December 2016).
2. United States Food and Drug Administration; Blood
Guidances: http://www.fda.gov/BiologicsBloodVaccines/
GuidanceComplianceRegulatoryInformation/Guidances/Blood/
default.htm (accessed 26 December 2016).
161
Annex 4
Guidelines on estimation of residual risk of HIV, HBV or
HCV infections via cellular blood components and plasma
1. Introduction 166
3. Terminology 167
4. Course of HIV, HBV and HCV infections 170
4.1 Acute infection 170
4.2 Chronic persistent infection 170
5. Residual risk origins 171
5.1 Assay failures 171
5.2 Diagnostic window periods 172
6. Screening assay categories and diagnostic window periods 173
6.1 Screening assay categories 173
6.2 Diagnostic window periods 175
7. Virus concentrations during diagnostic window period 178
8. Confirmation of reactive screening results 178
9. Virus epidemiology of donor populations 179
9.1 First-time donors 179
9.2 Repeat donors 180
10. Estimation of incidence and window period modelling of risks 180
10.1 Incidence 180
10.2 Residual risk per blood donation in repeat donors 181
11. Residual risks 183
11.1 Infection of recipients of non-pathogen-inactivated blood components 184
11.2 Contamination of plasma pools 184
References 186
Appendix 1 Evaluation of new blood-screening assays 190
Appendix 2 Examples for estimation of residual risks 194
163

164
Annex 4
Abbreviations
anti-HBc antibodies to hepatitis B core protein
anti-HBs antibodies to hepatitis B surface antigen
CE Conformité Européenne (conforms to European requirements)
CLIA chemiluminescence immunoassay
EIA enzyme immunoassay
FDA Food and Drug Administration
HBsAg hepatitis B surface antigen
ID-NAT individual donation nucleic acid amplification technique
IDI interdonation interval
MP-NAT minipool nucleic acid amplification technique
OBI occult hepatitis B infection
P probability
PCR polymerase chain reaction
RDT rapid diagnostic test
RR residual risk (used in mathematical formulae)
vDWP viraemic phase of the diagnostic window period
165
1. Introduction
The course that a viral infection may take in an individual and the different
phases of viral infections are described in the following sections – together with
the advantages and limitations of using different blood-screening assays for the
different infection phases. Blood-screening assays are differentiated by distinct
categories. The residual risk of missing viral infections using any screening
assay is mainly due to the viraemic phase of the diagnostic window period
(vDWP) for each assay – the mean size of which varies between different assay
categories. Another component of the residual risk is the virus epidemiology
of the donor population (consisting of repeat and first-time donors) with the
rate of new infections (incidence) in donors determining the probability of
window-period donations. The residual risk per donation from the repeat-
donor subpopulation may be used to extrapolate the respective risk for the first-
time donor subpopulation, for which incidence data are often unavailable. The
residual risk affects recipients of non-pathogen-inactivated blood components to
whom viruses may be transmitted. It also determines the potential viral load of
plasma pools used for the manufacturing of plasma-derived medicinal products
(PDMPs); this potential contamination level needs to be assessed against the
viral inactivation or reduction strategies in the manufacturing process.

These WHO Guidelines provide advice on estimating the residual risk of
human immunodeficiency virus (HIV), hepatitis B virus (HBV) or hepatitis C
virus (HCV) being present in cellular blood components and plasma. This
estimation has implications for the safety of non- (or incompletely) pathogen-
inactivated blood components or plasma products. There are large differences
in the prevalence and incidence of viral infections in blood donors around the
world. The impact of such epidemiological differences on blood safety needs
to be assessed together with the sensitivity of the testing strategy applied. Such
assessments may be used to guide strategic decisions on the choice of assays to
detect virus-positive blood donations and as a basis for cost–benefit analysis of
the different testing scenarios most suitable in the region. The factors influencing
the risk of virus transmission by blood components are described, as well as
simple mathematical formulae to calculate its probability. These estimates may
also be used to counsel recipients on the risks of transfusion. Similarly, the
probability and potential level of viral contamination of plasma pools used
for the manufacture of PDMPs can be calculated. The infectivity risk of plasma
products can then be estimated in relation to the inactivation and reduction
capacity of the manufacturing process.
166
Annex 4
Currently, recovered plasma from whole blood donations is often not

used for plasma fractionation because of perceived potential virus risks and
quality concerns. This is true for (but not limited to) many blood establishments
in low- and middle-income countries, where specific data (for example, on
interdonation periods of individual donors) are often not available due to a lack
of computerized systems. These WHO Guidelines therefore aim to enable the
approximate estimation of residual risks based on limited data, while recognizing
that more precise models have been published in the scientific literature.
Nevertheless, it is hoped that this document can help in rationalizing decision-
making on the use of plasma units for fractionation.
Since the performance of screening assays is one of the key elements in
minimizing the residual risk of blood components and guaranteeing the safety of
plasma products, these WHO Guidelines also contain advice on the assessment
of in vitro diagnostics (IVDs) in studies using specimen panels from the region
(Appendix 1). Such targeted performance evaluations for new assays may be
performed prior to the acceptance of a new blood-screening assay in a country.
3. Terminology
Analytical sensitivity: the smallest amount of the target marker that
can be precisely detected by an IVD assay; it may be expressed as the limit of
detection and is often determined by testing limiting dilutions of a biological
reference preparation.
Apheresis: the process by which one or more blood components are
selectively obtained from a donor by withdrawing whole blood, separating it
by centrifugation and/or filtration into its components, and returning those
not required to the donor. The term “plasmapheresis” is used for a procedure
dedicated specifically to the collection of plasma.
Blood collection: a procedure whereby a single donation of blood is
collected in a sterile receptacle containing anticoagulant and/or stabilizing
solution, under conditions designed to minimize microbiological contamination,
cellular damage and/or coagulation activation.
Blood component: a constituent of blood that can be used directly or
after further processing for therapeutic applications. The main therapeutic blood
components are red blood cell concentrates, platelet concentrates, plasma for
transfusion and cryoprecipitate.
Blood establishment: any structure, facility or body that is responsible
for any aspect of the collection, testing, processing, storage, release and/or
distribution of human blood or blood components when intended for transfusion
167
or further industrial manufacturing. It encompasses the terms “blood bank”,

“blood centre”, “blood transfusion unit”, “blood service” and “blood transfusion
service”. The definition of this term may differ between legislations.
Blood product: any therapeutic substance derived from human blood,
including whole blood, blood components and PDMPs.
Diagnostic sensitivity: the probability that an assay gives a positive result
in human specimens containing the target marker (that is, being true positive).
Diagnostic window period: the time interval from infection to the
time point when a blood specimen from that infected person first yields a
positive result in a diagnostic or screening assay for that agent (for example,
for specific antibodies); in the context of residual risk this is often simply
called the “diagnostic window” or “window period”. The diagnostic window
period consists of two phases – the first period of viral replication in the target
tissue without presence in peripheral blood is called the eclipse period; the
eclipse period is then followed by the ramp-up phase during which the virus
concentration increases exponentially in the blood (viraemic phase). Blood
components prepared from a blood donation made during the viraemic phase
of the diagnostic window period (vDWP) (the potentially infectious window
period) can transmit infection to the transfusion recipient, or the respective
plasma may contaminate the plasma pool used for manufacturing PDMPs.
Donor: a person in defined good health conditions who voluntarily
donates blood or blood components.
First-time (tested) donor: a donor whose blood or plasma is tested for
the first time for infectious disease markers in a blood establishment.
Fractionation: the (large-scale) process by which plasma is separated
into individual protein fractions that are further purified for medicinal use. The
term “fractionation” is usually used to describe a sequence of processes, including
plasma protein separation steps (typically precipitation and/or chromatography)
and purification steps (typically ion-exchange or affinity chromatography). These

steps may also contribute to the inactivation or removal of bloodborne infectious
agents (most specifically viruses and, possibly, prions).
Hepatitis B virus (HBV): an enveloped double-stranded DNA virus;
causative agent of hepatitis B.
Hepatitis C virus (HCV): an enveloped single-stranded RNA virus;
causative agent of hepatitis C.
Human immunodeficiency virus (HIV): an enveloped diploid single-
stranded RNA virus; causative agent of acquired immunodeficiency syndrome
(AIDS).
Incidence: the number of newly acquired infections per unit of time in
a defined population.
168
Annex 4
NAT conversion: the time period during which specific nucleic acids
(for example, viral nucleic acids after a recent virus infection) become detectable
by a nucleic acid amplification technique.
Nucleic acid amplification technique (NAT): a testing method to detect
the presence of a targeted area of a defined nucleic acid sequence (for example,
viral genome) using amplification techniques such as polymerase chain reaction
(PCR) or transcription mediated amplification (TMA).
Plasma: the liquid portion remaining after separation of the cellular
elements from blood – collected in a receptacle containing an anticoagulant, or
separated by the continuous filtration or centrifugation of anticoagulated blood.
Plasma for fractionation: plasma (from whole blood or apheresis) used
for the production of PDMPs.
Plasma for transfusion: plasma (from whole blood or apheresis) used
for direct infusion into patients without a prior fractionation step. It can be
subjected to treatment for inactivating a broad range of pathogens.
Plasma-derived medicinal products (PDMPs): a range of medicinal
products obtained by the fractionation of human plasma. Also called plasma
derivatives, plasma products or fractionated plasma products.
Plasmapheresis: see “Apheresis” above.
Prevalence: the proportion of past infections identified over a specified
period in a defined population.
Recovered plasma: plasma recovered from a whole blood donation and
used for transfusion or for fractionation into PDMPs.
Repeat donor: a person who has donated blood/plasma previously in the
blood establishment. The definition of this term may differ between legislations.
Sensitivity: see “Analytical sensitivity” and “Diagnostic sensitivity” above.
Seroconversion: the time period during which specific antibodies
develop (for example, after a recent virus infection) and become detectable in
the blood; this term is sometimes also used for the time period during which
viral antigens, such as hepatitis B surface antigen (HBsAg), or viral nucleic
acids become detectable in the blood after recent infection. See also “NAT
conversion” above.
Source plasma: plasma obtained by apheresis for further fractionation
into PDMPs.
Viraemic phase of the diagnostic window period (vDWP): the part of
the diagnostic window period during which viruses are present in the blood; the
beginning of the viraemic phase is defined by the putative presence of one virus
particle in a blood component (20 mL plasma for packed red blood cells) and
can be extrapolated using viral replication kinetics (viral doubling time).
Window period: see “Diagnostic window period” above.
169
4. Course of HIV, HBV and HCV infections

The course of infection in humans differs for HIV, HBV and HCV depending
on the biological features of the virus and on the individual immunological
response to the infection. In principle, chronically persistent virus infections
can be distinguished from infection courses leading to clearance of the virus.
Both courses have in common an acute phase which is associated with viral
replication, detectable viraemia and sometimes with clinical symptoms. A
chronically persisting infection without viral clearance almost always occurs
with HIV, frequently with HCV and sometimes with HBV.
4.1 Acute infection

The acute viraemic phase of infection is followed by the humoral and cellular
immune responses, resulting in seroconversion and potential clearance of
the virus. For some infections the immunity also protects against reinfection.
The acute viraemic phase of virus infection in blood donors may be detected
by antigen assays or, more sensitively, by assays based upon the nucleic acid
amplification technique (NAT). Antibody assays are not useful for the detection
of acute infections, but have long been used for the detection of persistent
infection (HIV, HCV). Usually there is an overlap of immunoglobulin detection
(for example, of immunoglobulin M) and the declining phase of viraemia.
For HBV, both acute resolving and chronic persistent infection courses
occur. The frequencies of either are dependent upon different factors (such as
the age of the individual becoming infected). It has been estimated that in 70%
of HBV-infected donors hepatitis B surface antigen (HBsAg) may be detected
transiently in blood, 5% develop chronic HBV infection with continuous
antigenaemia and 25% do not show detectable antigenaemia. In principle the
marker HBV DNA follows the same transient pattern as HBsAg but the median
length of viraemia detection is longer. The transient nature of these HBV blood-
screening markers requires the use of an adjustment factor when calculating
rates of new infections (1).
4.2 Chronic persistent infection

HIV causes persistent infection in almost all infected individuals, while HCV
infection becomes chronic in approximately 70% of cases (2). A minority of
HBV-infected adults (around 5%) become chronic carriers, depending on the
age and immune status of the infected subjects. These chronic infections of
HIV, HBV and HCV are usually lifelong active infections associated with viral
replication characterized by continuous or reappearing (undulating) phases of
viraemia, despite the presence of specific antibodies.
170
Annex 4
Persistent viraemic infections are usually detectable by both serology

and NAT-based assays. An exception is HBV where low-level HBV-DNA-positive
carriers (HBsAg negative; antibodies to hepatitis B core protein (anti-HBc)
positive) have been associated with so-called occult hepatitis B infection (OBI)
(3, 4). In some low-prevalence countries the potential OBI transmission risk has
been greatly reduced by the introduction of anti-HBc testing. However, in large
parts of the world where HBV is endemic, screening for this marker would lead
to the loss of an unacceptable proportion of donors. Blood components from
donors with OBI have transmitted HBV at a low frequency (approximately
3%), while the presence of detectable levels of antibody against HBsAg (anti-
HBs) has been found to protect against infection, with few exceptions (5–9).
The OBI-associated risk for HBV transmission via cellular blood components
may be reduced by sensitive NAT-based screening assays. The OBI-associated
input of HBV into plasma pools used for the manufacture of PDMPs appears
negligible when compared to the potential viral loads in diagnostic window
period donations.
5. Residual risk origins

Predominantly, the residual risk of HIV, HBV or HCV infections in blood or
plasma donations is defined as the probability of collecting a donation from an
asymptomatic viraemic donor infected with one of these bloodborne viruses,
and this not being detected by the routine screening assays.
Such an undetected blood donation may transmit the infection to a
recipient if the blood components are not pathogen inactivated. If the pathogen
inactivation and removal capacity of the production process is not sufficient an
infectious unit of plasma may also contaminate a manufacturing plasma pool
and pose a risk to the recipients of the plasma-derived products.
The non-detection of virus infection in blood or plasma donors may be
caused by assay failures or by donors being in the diagnostic window period.
5.1 Assay failures

Assay failures in blood screening can occur due to viral variants escaping
detection (for example by oligonucleotide mismatches in NAT-based methods,
monoclonal antibodies not detecting the antigen of a mutant virus, or
recombinant antigens/peptides not detecting antiviral antibodies) (10–12).
The contribution of assay failures to the residual risk is considered negligible
for “state-of-the-art” assays and will not be factored into the residual risk
calculation suggested by these Guidelines. Nevertheless, it is important to
continuously survey the quality features of screening assays and to identify
potential causes of false test results. Post-marketing surveillance of assay safety,
171
quality and performance is a mechanism for detecting, investigating and acting

on any issues and failures identified, and for addressing the need for continuous
assay improvement (13).
Another potential root cause of assay failure is an inadequate
quality management system in place within the testing laboratory. Quality
assurance aspects include: (a) participation in external quality assessment
(proficiency testing and on-site supervision); (b) the conduct of process (quality)
control; (c) maintaining adequate documentation (through standard operating
procedures) and record-keeping (testing logbooks, registers); (d) maintaining
proper inventory and purchasing systems; (e) equipment maintenance; (f) safe
facilities; (g) appropriate organization; and (h) measures to ensure adequately
trained and competent testing personnel.
5.2 Diagnostic window periods

Historically, the phase elapsing between the time point of infection and the
point of first detectability of the viral marker by the screening assay has been
called the diagnostic window period. All types of screening assays are associated
with a diagnostic window, the length of which is dependent upon the screening
marker, the screening assay category, the sensitivity of the assay used and the
replication kinetics of the virus during early infection.
The diagnostic window of HIV, HBV and HCV infections begins with
the eclipse phase during which the virus is not yet detectable in blood (even
by highly sensitive NAT-based assays). This non-viraemic phase is followed by
the viraemic ramp-up phase during which virus concentration in the plasma
increases in an exponential fashion. For each of the three bloodborne viruses
covered in these Guidelines (HIV, HBV and HCV) a specific constant replication
rate is apparent until a peak or plateau phase of maximal viral concentration
is reached.
In the context of blood safety, the viraemic phase within the diagnostic
window period is relevant. The start of the potentially infectious window period
during the early ramp-up phase of viraemia can be defined as the point at
which one virus particle is present in a blood component. A generally accepted
worst-case assumption for cellular components is to define the start of the
infectious window period as the point at which the concentration reaches one
virus particle in 20 mL of plasma (the volume co-transfused with a red blood
cell unit suspended in additive solution) (14). Viral replication characteristics
in the early phase of infection are rather consistent among recently infected
individuals. This phenomenon results in characteristic doubling times of plasma
viral concentration for HIV, HBV and HCV. By knowing the viral replication
kinetics of HIV, HBV or HCV in the early infection phase, along with the
diagnostic sensitivity of the screening assay, the length of the viraemic phase can
be extrapolated for each screening assay.
172
Annex 4
5.2.1 HIV
HIV replicates with an average doubling time of 20 hours (0.83 days) to
reach a peak level of viraemia of up to 10 7 IU HIV RNA/mL (15). This virus
concentration decreases in parallel with the development of specific antibodies
detectable by anti-HIV assays. The currently most sensitive antigen assays
can detect HIV p24 antigen at a level corresponding to 10 4 IU HIV RNA/mL.
Most HIV antigen‑antibody combination (“combo”) assays are less sensitive
in their detection of p24 antigen when compared to antigen assays – with
the corresponding HIV RNA concentration for detection by state-of-the-art
combo assays being around 10 5 IU/mL (15, 16). Attention should be paid to
donors having taken early antiretroviral treatment or pre-exposure antiretroviral
treatment which could reverse seroconversion and lower viral load (17).
5.2.2 HBV
The replication rate of HBV in the early infection phase as determined by the
increase in viraemia is significantly lower when compared to HIV or HCV,
with an HBV average doubling time of 2.6 days (18, 19). HBV viraemia in the
early infection phase is detected earlier by NAT-based assays than by HBsAg
assays. In the absence of NAT-based assays the use of HBsAg assays with a high
analytical sensitivity is key for the detection of early infection.
5.2.3 HCV
For HCV an average doubling time of 10.8 hours (0.45 days) during the ramp‑up
phase has been determined, followed by an anti-HCV-negative plateau phase
of several weeks characterized by high-level viraemia of up to 10 8 IU HCV
RNA/mL (20, 21). HCV core antigen appears to be detectable by core antigen
assays during the major part of this anti-HCV-negative phase, namely the
entire plateau phase and the last part of the ramp-up phase. Similar to HIV, the
antigen detection efficiency of current HCV combo assays is less than that of
the antigen assays. Combo assays have an overall detection rate of approximately
40% of anti-HCV-negative window period specimens, and preferentially detect
those with virus concentrations above 10 6 IU HCV RNA/mL (22).
6. Screening assay categories and

diagnostic window periods
6.1 Screening assay categories
In these Guidelines screening assays are discussed according to the following
categories:
■■ NAT-based
173
■■ antigen
■■ combo
■■ antibody
■■ rapid diagnostic test (RDT).
While antibody assays are designed to detect both recent and chronic
persistent infections, the additional benefit of antigen or viral genome detection
lies mainly in further reducing the diagnostic window period. The length of the
diagnostic window period varies greatly between the different assay categories.
6.1.1 NAT-based assays

NAT-based assays detect viral nucleic acids after in vitro amplification of a target
region of the viral genome. Such assays are performed on individual donations
(ID-NAT) or in small minipools of donations (MP-NAT). A true infection may
not be detectable by NAT-based assays if the concentration of viral genomes
is below the detection limit of the assay. Without virus-enrichment steps (for
example, ultracentrifugation) in pooled specimens the length of the window
period increases with the minipool size and is shortest with ID-NAT. At low
virus concentrations in the early ramp-up phase of the window period the
amount of virus in a defined volume follows a Poisson distribution, with higher
virus concentrations associated with increasing detection probabilities by NAT-
based assay. The concentration range between a 5% and a 95% probability of
detection may be 100-fold, and this complicates the estimation of window-period
reduction that can be achieved by the use of NAT-based assays. In these WHO
Guidelines the three-fold concentration of the 95% detection probability has
been taken as worst-case assumption for reliable NAT detection for estimating
virus concentration in a potentially contaminated plasma pool (Table A4.1;
normal font). However, NAT-based assay window periods may be significantly
shorter at the lower bound of uncertainty range. The vDWP corresponding to

the 50% NAT-detection probability is considered a more accurate estimate for
virus transmission risk by blood components without pathogen inactivation
(Table A4.1; bold italic font) (23, 24).
6.1.2 Antigen assays

Antigen assays have been optimized for the detection of viral proteins (antigens),
which are part of the virus particle, such as viral capsids (for example, HIV p24
or HCV core) or virus envelopes; or are subviral particles (for example, HBsAg).
For recently infected individuals, non-reactive test results using antigen assays
are caused by an absence of viral proteins, the presence of mutated antigen or the
presence of antigens with concentrations below the detection limit of the assay.
174
Annex 4
6.1.3 Combo assays

Combo assays are designed to simultaneously detect specific antibodies and viral
proteins; non-reactive combo assay test results for a true infection may be caused
by the absence (or too low a concentration) of antibodies and/or viral antigens
in the test sample, or by hidden epitopes in the immune complexes. The antigen-
detection potency of combo assays is often lower than that of assays optimized
for exclusive antigen detection.
6.1.4 Antibody assays

Antibody assays report infection through the detection of specific antibodies
against the pathogen; for recently infected individuals, non-reactive test results
using antibody assays can be caused by the absence of specific antibodies,
an antibody concentration that is insufficient for obtaining a signal in the
immunoassay or low binding strength (avidity) of antibodies. The design of
the antibody assay determines its sensitivity and capacity to detect low-avidity
antibodies.
6.1.5 RDTs
RDTs are diagnostic devices of simple design, often based on
immunochromatographic (lateral flow) or immunofiltration (flow-through)
technologies. RDTs do not require complex equipment and provide the test
result within a short time (15–30 minutes). Although often not claimed by the
manufacturer a suitable for use in blood screening, these devices are sometimes
used for blood-safety testing in resource-limited settings or in emergency
situations. RDT technology is associated with a lower sensitivity than that of
more sophisticated immunoassays developed specifically for blood screening
(25, 26).
6.2 Diagnostic window periods

NAT-based assays are generally able to detect a recent infection sooner than
antigen assays, followed by combo assays and antibody assays. These differential
capacities for detecting recent infections result in different lengths of the
diagnostic window period for different assay categories. Within each of the assay
categories, individual assays from different manufacturers may have different
sensitivities. These differences sometimes result in overlapping diagnostic
sensitivities in detecting early infection when less sensitive assays of one category
are compared with the more sensitive assays in another category. For example,
currently the most sensitive HIV1/2 antibody assay provides a shorter diagnostic
window period than the least sensitive HIV1/2 combo assay. This is true both for
assays prequalified by WHO and for CE-marked assays. Furthermore, assays may
175
have differing sensitivities for different viral genotypes and/or for viral subtypes.
The vast majority of commercial seroconversion panels used for diagnostic
sensitivity studies originate from regular plasma donors, and mainly represent
viral genotypes and subtypes prevalent in Europe and the United States (namely
HIV subtype B, HCV genotypes 1–3 and HBV genotype A). However, the
sensitivity of assays observed with these seroconversion panels may not always
be representative for early infection with viral genotypes prevalent elsewhere in
the world (27). Further details on this and other considerations in the evaluation
of new blood-screening assays are provided in Appendix 1.
Mean estimates of the length of the vDWP for so-called state-of-the-
art assays are presented by assay category in Table A4.1. These estimates should
be used for risk calculation unless more detailed information is available on
the sensitivity and corresponding window period of the assay used for blood
screening. Hence, if comparative data obtained with multiple seroconversion
panels indicate that the sensitivity of a specific assay is clearly different from the
mean value shown in Table A4.1, the more accurate data for this assay should be
used for the estimation of residual risk.
176
Table A4.1
Length of the vDWP for different assay categories (days) a
ID-NAT b MP16-NAT b, c Antigen Combo Antibody Antigen Combo Antibody

EIA/CLIAd EIA/CLIAd EIA/CLIAd RDT e RDT e RDT e
8 11
HIV 14 16 21 20 28
4 7 –
27 37 –
HBV 42 – 55 – –
17 27
5 7 –
HCV 9 38 60 – 80
3 5
a
vDWP: defined here as the period with a virus concentration of ≥ 1 virus particle in 20 mL plasma. 1 virus particle has been assumed to correspond to 1 (HCV, HBV) or 2 (HIV)
viral genome copies. 1 IU HCV RNA has been assumed to correspond to 4 genome copies HCV RNA; 1 IU HBV DNA to 5 genome copies HBV DNA; and 1 IU HIV-1 RNA to 0.5
genome copies HIV-1 RNA.
b
NAT-based assays: to date, only a limited number of NAT-based assays claiming blood screening as an intended use have been CE-marked or approved by the United States
Food and Drug Administration (FDA). For a worst-case scenario, diagnostic window periods of less sensitive NAT-based assay versions have been taken as examples in
Table A4.1.
Plasma pool contamination: for estimating the maximal virus concentration in a contaminated plasma pool the three-fold concentration of the 95% detection probability
has been taken as a worst-case assumption for reliable and consistent (“100%”) NAT detection. This approach is analogous to the determination of the whole system failure
rate in the European Commission’s common technical specifications for in vitro diagnostic medical devices (28). The respective sizes of the vDWPs in days are indicated in
normal font.
Transmission risk by non-pathogen-inactivated blood components: the Poisson distribution property of analyte detection by NAT-based assay is considered suitable for
more accurate estimation of virus transmission risk by blood components without pathogen inactivation. NAT-based assay window periods may be significantly shorter at
lower bound of uncertainty range. The probability of 50% detection in the early ramp-up phase of viraemia may be taken as the basis for the respective vDWPs (indicated
in bold italic font) (23, 24).
c
MP16-NAT = MP-NAT of 16 donations.
d
EIA/CLIA: for these assay types United States FDA-approved, CE-marked and/or WHO-prequalified assays of medium sensitivity have been selected as examples (20, 22, 25,
26, 29, 30).
e
RDT: there is a wide range of sensitivity among different RDT assays; values for medium-sensitivity RDTs have been used for Table A4.1 (25, 26).
Annex 4
177
7. Virus concentrations during diagnostic window period

For risk modelling of plasma pool contamination the maximal virus
concentrations that can be found during the respective window period are
relevant. Viral loads in viraemic plasma units undetected by screening assays
define the extent of initial contamination of the plasma pool. Other parameters
for calculation of potential contamination of plasma pools are the number of
viraemic donations expected per pool and the individual plasma unit volume
relative to the pool size. The maximal viral loads of window-period donations
are listed in Table A4.2 as worst-case scenarios for each of the different assay
categories correspondingly shown in Table A4.1.
Table A4.2
Maximal concentration of viral genomes in the vDWP (IU/mL) a
ID- MP16- Antigen Combo Antibody Antigen Combo Antibody

NAT NAT EIA/CLIA EIA/ EIA/CLIA RDT RDT RDT
CLIA
HIV 150 2400 2 ×10 4 10 5 10 7 10 7 10 7
HBV 24 384 10 3 3 × 10 4
HCV 30 480 10 4 5 × 10 6 10 8 10 8
IU/mL = International Units per millilitre
a
8. Confirmation of reactive screening results

The residual risk estimations rely on reactive screening assay results representing
true infection events. Initially reactive screening results obtained by antibody or

antigen tests should be checked by repeat testing in duplicate in the same assay.
Even when reactivity is repeatedly obtained in the routine screening assay, the
test result should still be checked by a confirmation strategy (31).
Confirmation strategies may include the use of more specific assays (for
example, HIV Western blot or immunoblot, HCV immunoblot and HBsAg
neutralization assay) or another screening or diagnostic assay for the same
marker, but of different design.
NAT results should be checked by testing an independent aliquot of the
donation to exclude contamination and/or by testing replicates to overcome
potential Poisson distribution of the analyte present at low concentration.
Follow-up investigations of the donor may further assist in differentiating false-
positive from true-positive test results.
178
Annex 4
Only reactive screening test results subsequently confirmed as true

positive should be taken for the estimation of residual risk. If no confirmation
is performed, residual risk estimations based on reactive test results represent a
worst-case scenario and may considerably overestimate the risks.
9. Virus epidemiology of donor populations

Donor populations consist of first-time donors (individuals donating for the
first time) and repeat donors (donors with previous donation(s) having tested
negative). Blood systems aim towards having an established population of repeat
donors undergoing constant selection for absence of infectious markers.
9.1 First-time donors

Positive screening test results in first-time donors may be an indication of
infections that occurred either a longer time ago (prevalent infections) or more
recently (incident infections). Prevalent infections in first-time donors are
expected to be easily detected by high-quality screening assay(s) without assay
failures; in contrast, incident infections represent the major contribution to
the residual risk of window-period infections. Making the distinction between
prevalent and incident infections will require more detailed investigation –
recently infected donors may be identified by NAT-only or antigen-only positive
results. Furthermore, for antibody-positive donors, modified antibody assays
(“detuned” or “recency” assays) can be used to determine the antibody binding
strength (avidity). As antibody avidity increases with maturation of the humoral
immune response it is possible to differentiate first-time donors with more recent
(incident) infections (low-avidity antibodies) from donors with past (prevalent)
infections (high-avidity antibodies) and thus determine the specific incidence
of infection in this subpopulation (14, 32). If results from these investigations
are not available for a specific first-time donor population, the incidence of
infection in these donors can be derived from the rate among repeat donors by
applying an adjustment factor. A number of scientific studies on HIV, HBV and
HCV infections in different donor populations have investigated their incidence
among both first-time and repeat donors. Although some of these studies found
a two- to three-fold higher rate of recent infections among first-time donors
(compared to the corresponding repeat donors) other studies have not found
such a difference between the two donor subpopulations (33–38). In the absence
of incidence data specific to the first-time donor population, one option is to
assume a three-fold higher incidence of virus infections as the worst-case
scenario for this subpopulation when compared to the corresponding repeat-
donor subpopulation of the same blood establishment. This factor will be referred
to as the “first-time donor incidence adjustment factor”.
179
First-time donor incidence (and corresponding adjustment factor) does

not have to be calculated for blood establishments in which newly registered
donors are routinely tested for bloodborne infections prior to their first donation
of blood or blood components.
9.2 Repeat donors

For repeat donors any confirmed positive screening test result indicates a new
infection, which is likely to have occurred during the interdonation interval
(IDI) – defined as the time period between the most recent donation (which in
this case will have tested positive) and the previous donation (which will have
tested negative). However, it is also possible that the previous donation (tested
negative) was drawn during the diagnostic window period of the screening
assay. The relative frequency of this possibility depends on the length of the IDI,
with shorter IDIs increasing the probability of a vDWP donation that tested
negative in the screening assay.
10. Estimation of incidence and window

period modelling of risks
10.1 Incidence
The rate of new infections of repeat donors (incidence) is defined as the number
of NAT conversions or seroconversions (number of infected donors) divided
by the total number of person years of observation of all donors during the
study period (14, 39, 40). Determining the person years of observation requires
a computer system that records the follow-up periods for each individual
donation. This kind of information management system is often not available in
resource-limited blood establishments.
For the purpose of these Guidelines, both the estimation of incidence

and the estimation of the residual risk per blood donation are derived from
data from the repeat-donor population for the period of one calendar year (365
days). Incidence is calculated by dividing the number of newly infected repeat
donors by the total number of repeat donors, usually expressed as the number
of new infection cases per 100 000 repeat donors. If one calendar year is taken
as the observation period then the incidence is expressed as per 100 000 person
years. This simplification assumes that each repeat donor has been followed for
one year during the calendar year and that differences in follow-up periods for
individual donors will average out at one person year of observation per donor.
In low-incidence regions the number of positive donors may show strong
year-to-year variation. For these situations longer periods may be chosen for the
calculation of residual risks.
180
Annex 4
Screening-positive donations that were excluded for other reasons (for

example, donor self-exclusion) may be excluded from the calculation (adjusted
incidence).
Formula 1: Incidence (per 100 000 person years)
number of repeat donors tested positive during one year
Incidence = × 100 000
total number of repeat donors in the year
10.2 Residual risk per blood donation in repeat donors

For calculating the probability that a blood donation has been collected during
the vDWP different factors are involved:
■■ the rate of new infections (incidence) in the repeat-donor
population
■■ the length of the vDWP for the assay used (Table A4.1).
The residual risk of a blood donation from a repeat donor having
been collected during the vDWP of the screening assay used can be calculated
as follows:
Formula 2: Residual risk (RR) per donation
RR per donation = vDWP × incidence
RR is usually expressed as per million donations (for which one has to multiply
the RR figure calculated above by 1 million.
Formula 2 can be directly used to calculate RR for HIV and HCV
infections in repeat donors; for HBV infections RR calculated by this formula
has to be multiplied by an HBV incidence adjustment factor.
10.2.1 HBV incidence adjustment factor

An adjustment factor of ≥ 1 is necessary because HBV (sero)conversions
in repeat donors may be missed due to the transient nature of viraemia and
antigenaemia in HBV infections that resolve after the acute phase. Such a
transient infection course is seen in adults for the majority of HBV infections
(95%) while 5% become chronic carriers. The probability of missing transiently
detectable HBsAg or HBV DNA in repeat donors by respective screening
assays depends on the length of the IDIs and on assay sensitivity. The donation
frequency of repeat donors (average number of donations per repeat donor)
determines the average length of the IDI. The average IDI (in days) can be
calculated by dividing the observation period of one calendar year (365 days)
by the average number of donations per repeat donor. For each assay category
181
a mean detection period for the transient HBV marker (HBsAg or HBV DNA)
can be factored into the adjustment. Further contributions to the adjustment
factor originate from HBV infections without detectable antigenaemia (assumed
to be 25%; transiently picked up by sensitive HBV NAT-based assays) (1). The
scientific literature provides several different estimates for the length of transient
antigenaemia (1, 19, 41). The differences observed between the underlying
studies may be explained by different infection routes, different inoculum,
different HBV genotypes, and HBsAg or HBV-DNA assays of different sensitivity.
The lengths of the HBV marker detection periods have been estimated
from the available data for the different assay categories and are listed in
Table A4.3.
Table A4.3
HBV DNA and HBsAg detection period (days) for different assay categories
NAT ID MP16-NAT HBsAg EIA/CLIA HBsAg RDT

90 70 60 44
The probability P (in %) of detection by HBsAg assays (Table A4.3) may be

calculated as:
HBsAg detection period
P = 70% × + 5%
IDI
The probability P (in %) of detection by NAT-based testing (Table A4.3) may

be calculated as:
HBV DNA detection period
P = 95% × + 5%
IDI
The HBV incidence adjustment factor is calculated as 100/P. For results where
P ≥ 100% no adjustment is necessary.
To determine the RR per donation for HBV infection, the figure obtained
for HBV using Formula 2 in section 10.2 above is then multiplied by the
adjustment factor for the specific assay category used.
10.2.2 Adjustment for IDIs

The incidence/window period modelling of residual risk, as described above,
assumes that donation behaviour with regard to donation timing and frequency
is the same for both infected and non-infected donors. However, evidence
can be found in the scientific literature indicating that seroconverting or
182
Annex 4
NAT-converting donors sometimes delay their return to blood donation, and

therefore have larger average IDIs than non-infected donors, resulting in a lower
residual risk (42). Mathematical models are available to reflect this difference
in donor behaviour (43). For high-incidence settings (that is, settings in which
a higher number of repeat donors have tested positive (seroconverters or NAT
converters) for HIV, HBV or HCV infection) the harmonic mean of individual
IDIs (in days) of the converting repeat donors (that is, the period between the
last negative donation and the first positive donation after infection with the
respective virus) may be compared with the mean IDI of non-infected repeat
donors (36). Respective functions for calculating mean or harmonic mean values
are part of commonly used statistical software (for example, Excel). The residual
risk calculation may then include the IDI adjustment factor S.
mean IDI of all donors

S=
harmonic mean IDI of converters for virus X
If, however, only a few acute infections are found it is advised to take the average
IDI of all repeat donors.
10.2.3 First-time donor incidence adjustment factor

In the absence of specific incidence data for first-time donors, a three-fold
higher residual risk may be assumed for blood donations from such donors
when compared to repeat donors of the same donor population.
Accordingly, the residual risk of a blood donation from a first-time donor
having been collected during the vDWP of the screening assay may be assumed
to be three-fold higher than the risk calculated for a blood donation obtained
from the corresponding repeat donors of the same blood establishment.
11. Residual risks

The approach to residual risk estimation proposed by these Guidelines requires
less detailed data on individual donors when compared to other models published
in the scientific literature. A recent comparison of seven different models for
estimating HIV incidence was performed by simulating donor populations
with different donation frequencies combined with different incidence rates
(44). The approach proposed by these Guidelines was retrospectively included
in the same simulation scenarios. In summary, this exercise revealed a slight
overestimation of incidence (by up to 20%) in the scenarios with low donation
frequency. This finding confirms the validity of the approach proposed in these
Guidelines and is in line with the worst-case scenarios chosen for the different
parameters, for example: (a) the proposed lengths of the vDWP (Table A4.1);
183
(b) the assumption of one virus particle in 20 mL plasma being infectious; or

(c) the use of the maximal viral concentration for all vDWP donations for the
calculation of potential plasma pool contamination (Table A4.2).
11.1 Infection of recipients of non-pathogen-

inactivated blood components
The actual infection risk in recipients of non-pathogen-inactivated blood
components is dependent on factors such as the amount of intact viruses
transmitted, the presence of potentially neutralizing antibodies in the donation
or recipient, virus properties and recipient immunological factors (30). Using
worst-case scenarios, the probability of viraemic donations escaping screening
can be estimated using Formula 2 in section 10.2 above. For whole blood
donations, different blood components (red cells, platelets and plasma) may be
obtained from the same donation and transfused to recipients, each contributing
to the residual risk. The amount of plasma in the blood component, the
probability of non-detection by the screening assay(s) and the infectivity of the
virus after storage of the blood component are all important factors influencing
the infection risk but are beyond the scope of these Guidelines (24, 30).
11.2 Contamination of plasma pools

Plasma prepared from whole blood donations (recovered plasma) or obtained
by plasmapheresis may be used as source material for plasma-derived products
manufactured from plasma pools (such as immunoglobulins, albumin and
clotting factors). These pools may be contaminated with HIV, HBV or HCV
as a result of the inclusion of plasma units originating from window-period
donations not detected by the screening assays. The extent of potential plasma-
pool contamination depends upon a number of factors:
■■ the expected probability of obtaining donations during the vDWP

of the screening assay used;
■■ the (maximal) amount of virus contamination in vDWP
plasma units;
■■ the volume of contaminated plasma unit(s) relative to pool size.
The proportion of viraemic plasma units is estimated by the residual risk
calculation. The (maximal) level of virus contamination in respective plasma
units can be calculated from the individual plasma volume and its virus
concentration. For these calculations, the maximal viral load of window-period
donations (shown above in Table A4.2 for the different assay categories) should
be taken as the worst-case scenario, even though only a minority of window-
period plasma units will reach the maximal viral load.
184
Annex 4

The development of these WHO Guidelines was initiated by Dr A. Padilla,
World Health Organization, Switzerland, and continued by Dr C.M. Nuebling,
World Health Organization, Switzerland. The members of the WHO working
group on residual risk were: Dr S Laperche, Institut National de la Transfusion
Sanguine, France; Dr N. Lelie, Consultant, Amsterdam, the Netherlands; Dr S.
Nick, Paul-Ehrlich-Institut, Germany; K. Preussel, Robert Koch Institute,
Germany; Dr Y. Soedermono, Ministry of Health, Indonesia; H. Yang, United
States Food and Drug Administration, the USA; and J. Yu, World Health
Organization, Switzerland.
Substantial inputs to the first draft of the document were provided by:
Dr J. Epstein, United States Food and Drug Administration, the USA; Dr S.
Kleinman, Centre for Blood Research, Canada; Dr R. Offergeld, Robert Koch
Institute, Germany; M. Perez Gonzalez, World Health Organization, Switzerland;
Dr R. Reddy, South African National Blood Service, South Africa; and A. Sands,
World Health Organization, Switzerland.
Between September 2015 and May 2016 the draft document was
presented and extensively discussed at a range of international workshops and
other forums. The resulting revised draft document was then published on
the WHO Biologicals website during a round of public consultation (11 July–
26 September 2016). During this phase further comments were received from the
following organizations, institutions and individuals: European Blood Alliance
(EBA); International Plasma Fractionation Association (IPFA); International
Society of Blood Transfusion (ISBT); Permanent United Nations Representation
of France; Plasma Protein Therapeutics Association (PPTA); Dr P. Akolkar and
Dr I. Hewlett, United States Food and Drug Administration, the USA; Dr D.
Brambilla, Dr S. Kleinman, Dr M. Busch, Dr R. Dodd, and S. Glynn for the
Recipient Epidemiology and Donor Evaluation Study (REDS)-III; Dr N. Lelie,
Consultant, Amsterdam, the Netherlands; Dr F. Moftah, Egyptian Society
for Blood Services, and Arab Transfusion Medicine Forum, Egypt; Dr M-L.
Hecquet, European Directorate for the Quality of Medicines & HealthCare,
France; Dr E. Lindberg, Medical Products Agency, Sweden; Dr M. Jannssen,
University Medical Center Utrecht, the Netherlands; Dr G. Praefcke, Paul-
Ehrlich-Institut, Germany; A. Sands, World Health Organization, Switzerland;
M. Vermeulen, South African National Blood Service, South Africa;
In October 2016, following review and incorporation of all comments
received, the final draft document WHO/BS/2016.2283 was prepared.
185
References
1. Korelitz JJ, Busch MP, Kleinman SH, Williams AE, Gilcher RO, Ownby HE et al. A method for estimating
hepatitis B virus incidence rates in volunteer blood donors. Transfusion. 1997;37(6):634–40
(https://www.researchgate.net/publication/14027062_A_method_for_estimating_hepatitis_B_
virus_incidence_rates_in_volunteer_blood_donors_National_Heart_Lung_and_Blood_
Institute_Retrovirus_Epidemiology_Donor_Study, accessed 18 December 2016).
2. Guidelines for the screening, care and treatment of persons with chronic hepatitis C infection.
Updated version April 2016. Geneva: World Health Organization; 2016 (http://apps.who.int/iris/
bitstream/10665/205035/1/9789241549615_eng.pdf?ua=1, accessed 18 December 2016).
3. Hollinger FB. Hepatitis B virus infection and transfusion medicine: science and the occult.
Transfusion. 2008;48(5):1001–26.
4. Allain JP. Occult hepatitis B virus infection: implications in transfusion. Vox Sang. 2004;86(2):83–91
(abstract: https://www.ncbi.nlm.nih.gov/pubmed/15023176/, accessed 18 December 2016).
5. Seed CR, Maloney R, Kiely P, Bell B, Keller AJ, Pink J. Infectivity of blood components from
donors with occult hepatitis B infection – results from an Australian lookback programme. Vox
Sang. 2015;108(2):113–22 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/25234417, accessed
18 December 2016).
6. Allain JP, Mihaljevic I, Gonzalez-Fraile MI, Gubbe K, Holm-Harritshøj L, Garcia JM et al. Infectivity of
blood products from donors with occult hepatitis B virus infection. Transfusion. 2013;53(7):1405–
15 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/23362802, accessed 18 December 2016).
7. Satake M, Taira R, Yugi H, Hino S, Kanemitsu K, Ikeda H et al. Infectivity of blood components with
low hepatitis B virus DNA levels identified in a lookback program. Transfusion. 2007;47(7):1197–
205 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/17581154, accessed 18 December 2016).
8. Levicnik-Stezinar S, Rahne-Potokar U, Candotti D, Lelie N, Allain JP. Anti-HBs positive occult
hepatitis B virus carrier blood infectious in two transfusion recipients. J Hepatol. 2008;48(6):1022–5
(https://www.researchgate.net/publication/5417215_Anti-HBs_positive_occult_hepatitis_B_
virus_carrier_blood_infectious_in_two_transfusion_recipients, accessed 18 December 2016).
9. Spreafico M, Berzuini A, Foglieni B, Candotti D, Raffaele L, Guarnori I et al. Poor efficacy of
nucleic acid testing in identifying occult HBV infection and consequences for safety of blood
supply in Italy. J Hepatol. 2015;63(5):1068–76 (abstract: https://www.researchgate.net/
publication/279300519_Poor_efficacy_of_nucleic_acid_testing_in_identifying_occult_HBV_
infection_and_consequences_for_safety_of_blood_supply_in_Italy, accessed 18 December
2016).
10. Humpe A, Legler TJ, Nübling CM, Riggert J, Unger G, Wolf C et al. Hepatitis C virus transmission
through quarantine fresh-frozen plasma. Thromb Haemost. 2000;84(5):784–8 (abstract: https://
www.researchgate.net/publication/12200759_Hepatitis_C_virus_transmission_through_
quarantine_fresh-frozen_plasma, accessed 18 December 2016).
11. Schmidt M, Korn K, Nübling CM, Chudy M, Kress J, Horst HA et al. First transmission of human
immunodeficiency virus Type 1 by a cellular blood product after mandatory nucleic acid
screening in Germany. Transfusion. 2009;49(9):1836–44 (abstract: https://www.researchgate.
net/publication/24438716_First_transmission_of_HIV-1_by_a_cellular_blood_product_after_
mandatory_NAT_blood_screening_in_Germany, accessed 19 December 2016).
12. Chudy M, Weber-Schehl M, Pichl L, Jork C, Kress J, Heiden M et al. Blood screening nucleic
acid amplification tests for human immunodeficiency virus Type 1 may require two different
amplification targets. Transfusion. 2012;52(2):431–9 (abstract: https://www.researchgate.net/
186
Annex 4
publication/51541171_Blood_screening_nucleic_acid_amplification_tests_for_human_
immunodeficiency_virus_Type_1_may_require_two_different_amplification_targets, accessed
19 December 2016).
13. Post-market surveillance for in vitro diagnostics (IVDs) [website]. Geneva: World Health
Organization (http://www.who.int/diagnostics_laboratory/postmarket/en/, accessed 19 December
2016).
14. Busch MP, Glynn SA, Stramer SL, Strong DM, Caglioti S, Wright DJ et al. A new strategy for estimating
risks of transfusion-transmitted viral infections based on rates of detection of recently infected
donors. Transfusion. 2005;45(2):254–64 (https://www.researchgate.net/publication/8070296_A_
new_strategy_for_estimating_risk_of_transfusion-transmitted_viral_infections_based_on_
rates_of_detection_of_recently_infected_donors, accessed 19 December 2016).
15. Fiebig EW, Wright DJ, Rawal BD, Garrett PE, Schumacher RT, Peddada L et al. Dynamics of
HIV viremia and antibody seroconversion in plasma donors: implications for diagnosis and
staging of primary HIV infection. AIDS. 2003;17(13):1871–9 (https://www.researchgate.net/
publication/271733355_Dynamics_of_HIV_viremia_and_antibody_seroconversion_in_plasma_
donors, accessed 20 December 2016).
16. Consolidated guidelines on HIV testing services. Geneva: World Health Organization; 2015
(http://apps.who.int/iris/bitstream/10665/179870/1/9789241508926_eng.pdf?ua=1&ua=1,
17. de Souza MS, Pinyakorn S, Akapirat S, Pattanachaiwit S, Fletcher JLK, Chomchey N et al.
Initiation of antiretroviral therapy during acute HIV-1 infection leads to a high rate of
nonreactive HIV serology. Clin Infect Dis. 2016;63(4):555–61 (https://www.researchgate.net/
publication/304070497_Initiation_of_Antiretroviral_Therapy_during_Acute_HIV-1_Infection_
leads_to_A_High_Rate_of_Non-Reactive_HIV_Serology, accessed 20 December 2016).
18. Biswas R, Tabor E, Hsia CC, Wright DJ, Laycock ME, Fiebig EW et al. Comparative sensitivity of HBV
NATs and HBsAg assays for detection of acute HBV infection. Transfusion. 2003;43(6):788–98
(https://www.researchgate.net/publication/10750139_Comparative_sensitivity_of_HBVNATs_
and_HBsAg_assays_for_detection_of_acute_HBV_infection, accessed 20 December 2016).
19. Yoshikawa A, Gotanda Y, Itabashi M, Minegishi K, Kanemitsu K, Nishioka K. Hepatitis B NAT virus-
positive blood donors in the early and late stages of HBV infection: analyses of the window
period and kinetics of HBV DNA. Vox Sang. 2005;88(2):77–86. Erratum in Vox Sang. 2005;88(3):223
(https://www.ncbi.nlm.nih.gov/pubmed/15720604/, accessed 20 December 2016).
20. Nübling CM, Unger G, Chudy M, Raia S, Löwer J. Sensitivity of HCV core antigen and HCV RNA
detection in the early infection phase. Transfusion. 2002;42(8):1037–45 (abstract: https://www.
researchgate.net/publication/11076628_Sensitivity_of_HCV_core_antigen_and_HCV_RNA_in_
early_infection_phase, accessed 20 December 2016).
21. Glynn SA, Wright DJ, Kleinman SH, Hirschkorn D, Tu Y, Heldebrant C et al. Dynamics of viremia
in early hepatitis C virus infection. Transfusion. 2005;45(6):994–1002 (https://www.researchgate.
net/publication/7807301_Dynamics_of_viremia_in_early_hepatitis_C_virus_infection, accessed
20 December 2016).
22. Laperche S, Nübling CM, Stramer SL, Brojer E, Grabarczyk P, Yoshizawa H et al. Sensitivity
of hepatitis C virus core antigen and antibody combination assays in a global panel of
window period samples. Transfusion. 2015;55(10):2489–98 (https://www.researchgate.net/
publication/277412688_Sensitivity_of_hepatitis_C_virus_core_antigen_and_antibody_
combination_assays_in_a_global_panel_of_window_period_samples_HCV_Core_Antigen_
Detection_in_Window_Period, accessed 20 December 2016).
187
23. Weusten JJ, van Drimmelen HA, Lelie N. Mathematic modeling of the risk of HBV, HCV, and HIV
transmission by window-phase donations not detected by NAT. Transfusion. 2002;42(5):537–48
(abstract: https://www.researchgate.net/publication/11291469_Mathematic_modeling_of_the_
risk_of_HBV_HCV_and_HIV_transmission_by_window-phase_donations_not_detected_by_
NAT, accessed 20 December 2016).
24. Weusten J, Vermeulen M, van Drimmelen H, Lelie N. Refinement of a viral transmission risk
model for blood donations in seroconversion window phase screened by nucleic acid testing in
different pool sizes and repeat test algorithms. Transfusion. 2011;51(1):203–15 (abstract: https://
www.ncbi.nlm.nih.gov/pubmed/20707858, accessed 20 December 2016).
25. Scheiblauer H, El-Nageh M, Diaz S, Nick S, Zeichhardt H, Grunert H-P et al. Performance
evaluation of 70 hepatitis B virus (HBV) surface antigen (HBsAg) assays from around the world by
a geographically diverse panel with an array of HBV genotypes and HBsAg subtypes. Vox Sang.
2010;98(3p2):403–14 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2860763/, accessed 20
December 2016).
26. Scheiblauer H, El-Nageh M, Nick S, Fields H, Prince A, Diaz S. Evaluation of the performance
of 44 assays used in countries with limited resources for the detection of antibodies to
hepatitis C virus. Transfusion. 2006;46(5):708–18 (abstract: https://www.ncbi.nlm.nih.gov/
pubmed/16686838, accessed 20 December 2016).
27. Apetrei C, Loussert-Ajaka I, Descamps D, Damond F, Saragosti S, Brun-Vézinet F et al. Lack of
screening test sensitivity during HIV-1 non-subtype B seroconversions. AIDS. 1996;10(14):F57–60
(abstract: https://www.ncbi.nlm.nih.gov/pubmed/8970678, accessed 20 December 2016).
28. 2009/108/EC: Commission Decision of 3 February 2009 amending Decision 2002/364/EC on
common technical specifications for in vitro-diagnostic medical devices. European Commission.
Official Journal of the European Union. 2009;L39/34–L39/49 (http://eur-lex.europa.eu/legal-
content/EN/TXT/?uri=uriserv:OJ.L_.2009.039.01.0034.01.ENG&toc=OJ:L:2009:039:TOC, accessed
18 December 2016).
29. Assal A, Barlet V, Deschaseaux M, Dupont I, Gallian P, Guitton C et al. Sensitivity of two hepatitis B
virus, hepatitis C virus (HCV), and human immunodeficiency virus (HIV) nucleic acid test systems
relative to hepatitis B surface antigen, anti-HCV, anti-HIV, and p24/anti-HIV combination assays in
seroconversion panels. Transfusion. 2009;49(2):301–10 (abstract: https://www.ncbi.nlm.nih.gov/
30. Kleinman SH, Lelie N, Busch MP. Infectivity of human immunodeficiency virus-1, hepatitis C
virus, and hepatitis B virus and risk of transmission by transfusion. Transfusion. 2009;49(11):
2454–89 (available at: https://www.researchgate.net/publication/26743554_Infectivity_of_
human_immunodeficiency_virus-1_hepatitis_C_virus_and_hepatitis_B_virus_and_risk_of_
transmission_by_transfusion, accessed 20 December 2016).
31. Screening donated blood for transfusion-transmissible infections. Recommendations. Geneva:
World Health Organization; 2010 (http://www.who.int/bloodsafety/ScreeningDonatedBloodfor
Transfusion.pdf, accessed 20 December 2016).
32. Murphy G, Parry JV. Assays for the detection of recent infections with human immunodeficiency
virus type 1. Euro Surveill. 2008;13(36):4–10 (http://www.eurosurveillance.org/ViewArticle.
aspx?ArticleId=18966, accessed 20 December 2016).
33. Schreiber GB, Glynn SA, Busch MP, Sharma UK, Wright DJ, Kleinman SH. Incidence rates of viral
infections among repeat donors: are frequent donors safer? Transfusion. 2001;41(6):730–5
(https://www.researchgate.net/publication/11939481_Incidence_rates_of_viral_infections_
among_repeat_donors_Are_frequent_donors_safer, accessed 20 December 2016).
188
Annex 4
34. Zou S, Stramer SL, Dodd RY. Donor testing and risk: current prevalence, incidence, and
residual risk of transfusion-transmissible agents in US allogeneic donations. Transfus Med Rev.
2012;26(2):119–28 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/21871776, accessed 20
December 2016).
35. O’Brien SF, Yi QL, Fan W, Scalia V, Fearon MA, Allain JP. Current incidence and residual risk of HIV,
HBV and HCV at Canadian Blood Services. Vox Sang. 2012;103(1):83–6 (abstract: https://www.
ncbi.nlm.nih.gov/pubmed/22289147, accessed 20 December 2016).
36. Bruhn R, Lelie N, Custer B, Busch M, Kleinman S. Prevalence of human immunodeficiency virus
RNA and antibody in first-time, lapsed, and repeat blood donations across five international
regions and relative efficacy of alternative screening scenarios. Transfusion. 2013;53(10 Pt 2):
2399–412 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/23782054, accessed 20 December
2016).
37. Bruhn R, Lelie N, Busch M, Kleinman S. Relative efficacy of nucleic acid amplification testing and
serologic screening in preventing hepatitis C virus transmission risk in seven international regions.
Transfusion. 2015;55(6):1195–205 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/25727549/,
38. Offergeld R, Ritter S, Hamouda O. HIV, HCV, HBV and syphilis surveillance among blood donors
in Germany 2008–2010. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz.
2012;55(8):907–13 (article in German; English abstract: https://www.ncbi.nlm.nih.gov/
39. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The risk of transfusion-transmitted
viral infections. N Engl J Med. 1996;334:1685–90 (http://www.nejm.org/doi/full/10.1056/
NEJM199606273342601#t=article, accessed 20 December 2016).
40. Glynn SA, Kleinman SH, Wright DJ, Busch MP. International application of the incidence
rate/window period model. Transfusion. 2002;42(8):966–72 (https://www.researchgate.
net/publication/11076616_NHLBI_Retrovirus_Epidemiology_Donor_Study_International_
application_of_the_incidence_ratewindow_period_model, accessed 20 December 2016).
41. Yoshikawa A, Gotanda Y, Minegishi K, Taira R, Hino S, Tadokoro K et al. Lengths of hepatitis B
viremia and antigenemia in blood donors: preliminary evidence of occult (hepatitis B surface
antigen-negative) infection in the acute stage. Transfusion. 2007;47(7):1162–71 (abstract:
https://www.ncbi.nlm.nih.gov/pubmed/17581150, accessed 20 December 2016).
42. Schreiber GB, Glynn SA, Satten GA, Kong F, Wright D, Busch MP et al. HIV seroconverting donors
delay their return: screening test implications. Transfusion. 2002;42(4):414–21 (https://www.
researchgate.net/publication/11299285_HIV_seroconverting_donors_delay_their_return_
screening_test_implications, accessed 20 December 2016).
43. an der Heiden M, Ritter S, Hamouda O, Offergeld R. Estimating the residual risk for HIV, HCV
and HBV in different types of platelet concentrates in Germany. Vox Sang. 2015;108(2):123–30
(abstract: https://www.ncbi.nlm.nih.gov/pubmed/25335096, accessed 20 December 2016).
44. Brambilla DJ, Busch MP, Dodd RY, Glynn SA, Kleinman SH. A comparison of methods for estimating
the incidence of human immunodeficiency virus infection in repeat blood donors. Transfusion.
2017;57(3pt2):823–31 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/27910095, accessed
20 April 2017).
189
Appendix 1
Evaluation of new blood-screening assays
Depending on the legal structure in a country, a regulatory body or the national
blood system itself may be responsible for decisions on the acceptability of new
blood-screening assays. It is recommended that previous assessments of quality
features of the assay performed by experienced regulatory authorities (for
example, United States FDA approval, European CE certification, and Australian
Therapeutic Goods Administration (TGA) or Health Canada marketing
authorizations) or by the WHO Prequalification Programme for IVDs should be
taken into account. Previous assessments by such stringent regulatory bodies will
have included the review of analytical and clinical performance data submitted
by the manufacturer, and of the manufacturer’s quality management system
and batch-to-batch consistency – and in the case of WHO prequalification, an
independent performance evaluation.
As a result, a country’s assessment of manufacturer documentation, with
a focus on the specific regional situation and needs, may be sufficient for assays
already approved elsewhere under stringent regulation.
If local regulation requires a performance evaluation of new assays (for
example, by a national reference laboratory) prior to their implementation, it is
recommended that the evaluation focuses on essential assay features through a
targeted performance evaluation.
Assessment of documents
Documents provided by the IVD manufacturer may be assessed, with a
special focus placed on the specific regional situation and needs. Such a focus
may include assessing whether or not the stability studies performed by the
manufacturer cover the regional environmental conditions (for example, with
regard to temperature and humidity) or whether the Instructions for Use are
appropriate for the target users.
In addition, performance evaluation studies documented by the IVD
manufacturer may be reviewed to evaluate the extent of representation of
specimens reflecting the regional situation (for example, with regard to viral
genotypes or variants) or to assess potential interference with the test result by
other regionally more prevalent infections.
190
Annex 4
Targeted performance evaluation of new

assays used for blood screening
If laboratory testing of a new IVD is a component of the national or regional
evaluation and approval scheme, it is advisable not to repeat evaluation
elements already performed by other bodies, but to focus instead on regionally
important quality aspects. This would involve, for example, a focused assessment
of performance data with respect to viral variants or genotypes prevalent in
the region.
Well-characterized specimen panels representing the regional
epidemiological situation with regard to viral variants/genotypes of HIV, HBV
or HCV may be helpful for comparative independent evaluation of new assays.
A comparative database obtained using a number of assays may then be the
scientific basis for the definition of acceptance criteria for new assays and for the
identification of less suitable assays.
The preconditions for the suitability of such panels are the inclusion of
specimens differentiating between different assays (for example, low-positive
specimens or positive specimens previously tested discrepantly by different
assays) and the availability of sufficient volumes to allow a number of evaluations
to obtain comparative data. The recommended size of such a panel strongly
depends on its composition, with more critical panel members (for example,
low-positive or early infection specimens) able to differentiate between assays
being more important than a high number of strong positive specimens. Panels
used for this type of exercise typically comprise 20–50 members collected from
different phases of the infection. A strategy for the replacement of panel members
should be in place.
Furthermore, WHO offers through its IVD standardization programme a
range of biological reference preparations that may be useful in the confirmation
of basic assay features. WHO International Standards (expressed in IU) are
available for the confirmation of analytical sensitivity, while WHO Reference
Panels representing the major viral genotypes could be used to check genotype-
detection efficiency.
These WHO reference preparations are usually lyophilized to facilitate
worldwide shipping and are listed in the WHO online catalogue (http://www.
who.int/bloodproducts/catalogue/en/). They can be obtained from the WHO
Collaborating Centres which act as WHO custodians in this field – namely, the
National Institute for Biological Standards and Control (NIBSC), England, or
the Paul-Ehrlich-Institut (PEI), Germany.
191
Table A4A1.1 summarizes the most important WHO reference

preparations currently available in the field of blood screening.
Table A4A1.1
WHO reference preparations in the field of blood screening
Marker Preparation Details Custodian

anti-HIV-1/2 1st International Reference Panel HIV-1 subtypes A, NIBSC
Lyophilized B, C, CRF01_AE;
No unitage group O; HIV-2
HIV-1 p24 1st International Reference – NIBSC
Reagent
Lyophilized
1000 IU/ampoule
HIV-1 RNA 3rd International Standard – NIBSC
Lyophilized
185 000 IU/mL
1st International Reference HIV-1 CRFs 11GJ, NIBSC
Panel HIV-1 circulating 02AG, 01AE,
recombinant forms (CRFs) 01AGJU,BG24;
Lyophilized subtypes J, G, C;
No unitage group O
2nd International Reference HIV-1 subtypes NIBSC
Panel HIV-1 subtypes A, B, C, D, AE, F, G,
Lyophilized AG–GH;
No unitage groups N and O
HIV-2 RNA 1st International Standard – NIBSC
Lyophilized
1000 IU/vial
HBsAg 3rd International Standard – NIBSC
Lyophilized
50 IU/mL
Dilutional panel – NIBSC
8.25; 2.06; 0.52; 0.13 IU/vial
1st International Reference HBV genotypes PEI
Panel HBV genotypes A–F, H
Lyophilized
No unitage
192
Annex 4
Table A4A1.1 continued

Marker Preparation Details Custodian
HBV DNA 4th International Standard – NIBSC
Lyophilized
955 000 IU/mL
1st International Reference HBV genotypes PEI
Panel HBV genotypes A–G
Lyophilized
No unitage
anti-HBc 1st International Standard – NIBSC
Lyophilized
50 IU/vial
HCV core 1st International Standard – PEI
Lyophilized
3200 IU/mL
HCV RNA 5th International Standard – NIBSC
Lyophilized
100 000 IU/mL
193
Appendix 2
Examples for estimation of residual risks
Example 1: HCV screening by anti-HCV EIA
Centre A; observation period 01.06.2011–31.05.2012
49 660 repeat donors; 100 313 donations; 45 anti-HCV pos (EIA)
11 452 first-time donors; 11 452 donations; 89 anti-HCV pos (EIA)
Table A4.1 (see section 6.2 main text) – anti-HCV EIA: vDWP = 60 days =
0.164 years
Table A4.2 (see section 7 main text) – anti-HCV EIA: maximal virus
concentration: 10 8 IU HCV RNA/mL
plasma of vDWP donation
A. Residual risk (RR) per blood donation from repeat donors
45
= × 100 000
49 660
= 90.61 HCV infections per 100 000 donor years
RR per blood donation = vDWP × incidence
= 0.164 × 0.000 906 1 = 0.000 148 600
= 148.60 per million donations

Number (N) of vDWP blood donations from repeat donors
148.60
N = 100 313 × = 14.90
1 000 000
B. Residual risk (RR) per blood donation from first-time donors
Positive screening test results for first-time donors represent mainly old
(prevalent) infections. The rate of recent infections can be determined by specific
investigations (for example, recency assays or NAT-only positive results).
In the absence of incidence data, the worst-case assumption is a three-
fold incidence in first-time donors compared to the corresponding repeat donors.
194
Annex 4
RR = 0.000 148 61 × 3 = 0.000 445 = 445 per million donations

Number (N) of vDWP blood donations from first-time donors:
445
N = 11 452 × = 5.10
1 000 000
C. Expected number (N) and risk of window-phase donations for repeat and first-
time donors combined (Centre A; observation period of 1 year)
N = 14.90 + 5.10 = 20.00
20
RR = = 0.000 179 = 179 per million donations
100 313 + 11 452
Example 2: HBV screening by HBsAg RDT; HBV adjustment factor

Centre A; observation period 01.06.2011–31.05.2012
49 660 repeat donors; 100 313 donations; 184 HBsAg RDT pos
11 452 first-time donors; 11 452 donations; 291 HBsAg RDT pos
Table A4.1 (section 6.2 main text) – HBsAg RDT: vDWP = 55 days = 0.15 years
Table A4.3 (section 10.2.1 main text) – HBsAg RDTs: HBV marker detection
period = 44 days
Average number of donations per repeat donor: 100 313/49 660 = 2.02
Interdonation interval (IDI)
365 days
IDI = = 180.69 days
average number of donations per repeat donor
A. Residual risk (RR) per blood donation from repeat donors (without adjustment
for transient HBsAg)
184
= × 100 000
49 660
= 370.52 HBV infections per 100 000 donor years
195
vDWP = 55 days = 0.15 years

RR = vDWP × incidence
= 0.15 × 0.003 705 2 = 0.000 555 78
= 555.78 per million donations
B. HBV incidence adjustment factor
Probability (P) for HBsAg detection
HBV marker detection period
P = 70% × + 5%
IDI

44 days
= 70% × + 5% = 70% × 0.24 + 5% = 21.8%
180.69 days
HBV incidence adjustment factor =
100% 100%
= = 4.58
P 21.8%
C. Residual risk (RR) per blood donation from repeat donors (with adjustment for
transient HBsAg)
Adjusted RR = 4.58 × 0.000 555 78 = 0.002 545 = 2545 per million donations.
196
Annex 5
Guidelines for the production, control and regulation of
snake antivenom immunoglobulins
Replacement of Annex 2 of WHO Technical Report Series, No. 964
1. Introduction 203
3. Terminology 205
4. The ethical use of animals 211
4.1 Ethical considerations for the use of venomous snakes in the production
of snake venoms 212
4.2 Ethical considerations for the use of large animals in the production
of hyperimmune plasma 212
4.3 Ethical considerations for the use of animals in preclinical testing of antivenoms 213
4.4 Development of alternative assays to replace murine lethality testing 214
4.5 Refinement of the preclinical assay protocols to reduce pain, harm and distress
to experimental animals 214
4.6 Main recommendations 215
5. General considerations 215
5.1 Historical background 215
5.2 The use of serum versus plasma as source material 216
5.3 Antivenom purification methods and product safety 216
5.4 Pharmacokinetics and pharmacodynamics of antivenoms 217
5.5 Need for national and regional reference venom preparations 217
6. Epidemiological background 218
6.1 Global burden of snake-bites 218
7. Worldwide distribution of venomous snakes 220
7.1 Taxonomy of venomous snakes 220
7.2 Medically important venomous snakes 224
7.3 Minor venomous snake species 228
7.4 Sea snake venoms 229
8. Antivenoms design: selection of snake venoms 232
8.1 Selection and preparation of representative venom mixtures 232
8.2 Manufacture of monospecific or polyspecific antivenoms 232
197
9. Preparation and storage of snake venom 235

9.1 Production of snake venoms for immunization 236
9.2 Staff responsible for handling snakes 244
10. Quality control of venoms 247
10.1 Records and traceability 247
10.2 National reference materials 248
10.3 Characterization of venom batches 249
11. Overview of the production process of antivenoms 250
12. Selection and veterinary health care of animals used for
production of antivenoms 253
12.1 Selection and quarantine period 253
12.2 Veterinary care, monitoring and vaccinations 253
12.3 Animal health and welfare after inclusion in the herd 254
13. Immunization regimens and use of adjuvant 257
13.1 Animals used in antivenom production 257
13.2 Venoms used for immunization 258
13.3 Preparation of venom doses 258
13.4 Detoxification of venom 259
13.5 Immunological adjuvants 259
13.6 Preparation of immunogen in adjuvants 260
13.7 Immunization of animals 260
13.8 Traceability of the immunization process 263
14. Collection and control of animal plasma for fractionation 265
14.1 Health control of the animal prior to and during bleeding sessions 265
14.2 Premises for blood or plasma collection 266
266
14.3 Blood or plasma collection session

14.4 Labelling and identification 267
14.5 Pooling 270
14.6 Control of plasma prior to fractionation 271
15. Purification of immunoglobulins and immunoglobulin
fragments in the production of antivenoms 272
15.1 Good manufacturing practices 272
15.2 Purification of the active substance 273
15.3 Pharmacokinetic and pharmacodynamic properties of IgG, F(abʹ)2 and Fab 285
16. Control of infectious risks 288
16.1 Background 288
16.2 Risk of viral contamination of the starting plasma 288
198
Annex 5
16.3 Viral validation of manufacturing processes 289

16.4 Viral validation studies of antivenom immunoglobulins 297
16.5 Production-scale implementation of process steps contributing to viral safety 302
16.6 Transmissible spongiform encephalopathy 303
17. Quality control of antivenoms 305
17.1 Standard quality assays 306
17.2 Antivenom reference preparations 311
18. Stability, storage and distribution of antivenoms 312
18.1 Stability 312
18.2 Storage 313
18.3 Distribution 313
19. Preclinical assessment of antivenom efficacy 314
19.1 Preliminary steps that may limit the need for animal experimentation 315
19.2 Essential preclinical assays to measure antivenom neutralization of
venom-induced lethality 316
19.3 Supplementary preclinical assays to measure antivenom neutralization
of specific venom-induced pathologies 320
19.4 Limitations of preclinical assays 325
20. Clinical assessment of antivenoms 326
20.1 Introduction 326
20.2 Clinical studies of antivenom 329
20.3 Post-marketing surveillance 332
21. Role of national regulatory authorities 335
21.1 Regulatory evaluation of antivenoms 336
21.2 Establishment licensing and site inspections 336
21.3 Impact of good manufacturing practices 337
21.4 Inspections and audit systems in the production of antivenoms 338
21.5 Antivenom licensing 340
21.6 National reference venoms 341
References 343
Appendix 1 Worldwide distribution of medically important venomous snakes 354
Appendix 2 Model protocol for the production and testing of snake antivenom
immunoglobulins 385
199

200
Annex 5
Abbreviations
ASV anti-snake venom
BVDV bovine viral diarrhoea virus
CK creatine kinase
CPD citrate phosphate dextrose solution
CTD Common Technical Document
ds-DNA double-stranded deoxyribonucleic acid
ds-RNA double-stranded ribonucleic acid
ED50 effective dose 50%
EIA enzyme immunoassay
EMCV encephalomyocarditis virus
FCA Freund’s complete adjuvant
FIA Freund’s incomplete adjuvant
GCP good clinical practice
Hb haemoglobin
HPLC high-performance liquid chromatography
ICH International Conference on Harmonisation of Technical
Requirements for Registration of Pharmaceuticals for
Human Use
IgM immunoglobulin M
LD50 lethal dose 50%
MCD minimum coagulant dose
MDD minimum defibrinogenating dose
MHD minimum haemorrhagic dose
MHD50 MHD-median effective dose
MMD minimum myotoxic dose
201
MMD50 MMD-median effective dose

MND minimum necrotizing dose
MND50 MND-median effective dose
Mr relative molecular mass
PCV packed cell volume
RCT randomized controlled trial
SDS-PAGE sodium dodecyl sulfate–polyacrylamide gel electrophoresis
SOP standard operating procedure
ss-DNA single-stranded deoxyribonucleic acid
ss-RNA single-stranded ribonucleic acid
TPP total plasma protein
TSE transmissible spongiform encephalopathy
VSV vesicular stomatitis virus
WNV West Nile virus
202
Annex 5
1. Introduction
Snake antivenom immunoglobulins (antivenoms) are the only therapeutic
products for the treatment of snake-bite envenoming. The lack of availability of
effective snake antivenom immunoglobulins to treat envenoming by medically
important venomous snakes encountered in various regions of the world has
become a critical health issue at global level. The crisis has reached its greatest
intensity in sub-Saharan Africa, but other regions, such as South and South-East
Asia, are also suffering from a lack of effective and affordable products.
The complexity of the production of efficient antivenoms, in particular the
importance of preparing appropriate snake venom mixtures for the production
of hyperimmune plasma (the source of antivenom immunoglobulins), the
decreasing number of producers, and the fragility of the production systems in
developing countries further jeopardize the availability of effective antivenoms
in Africa, Asia, the Middle East and South America. Most of the remaining
current producers are located in countries where the application of quality and
safety standards needs to be improved.
In October 2005, the WHO Expert Committee on Biological
Standardization recognized the extent of the problem and asked the WHO
Secretariat to support and strengthen world capacity to ensure the long-term
and sufficient supply of safe and efficient antivenoms. In March 2007, snake
antivenom immunoglobulins were included in the WHO Model List of Essential
Medicines (1), acknowledging their role in a primary health-care system.
WHO recognizes that urgent measures are needed to support the design
of immunizing snake venom mixtures that can be used to make appropriate
antivenoms for various geographical areas of the world. Sustainable availability
of effective and safe antivenom immunoglobulins must be ensured and
production systems for these effective treatments must be strengthened at global
level. Meaningful preclinical assessment of the neutralizing capacity of snake
antivenom immunoglobulins needs to be done before these products are used
in humans and medicines regulatory authorities should enforce the licensing of
these products in all countries, before they are used in the population.
The first edition of the WHO Guidelines for the production, control and
regulation of snake antivenom immunoglobulins was developed in response to
the above-mentioned needs and approved by the WHO Expert Committee on
Biological Standardization in October 2008. These Guidelines covered all the steps
involved in the production, control and regulation of venoms and antivenoms.
The Guidelines are supported by a WHO antivenoms database website1 that
See: http://apps.who.int/bloodproducts/snakeantivenoms/database/ (accessed 15 February 2017).

1
203
features information on all the venomous snakes listed in Appendix 1, including

distributions and photographs, as well as information on available antivenoms.
It is intended that these updated Guidelines, by comprehensively
describing the current existing experience in the manufacture, preclinical and
clinical assessment of these products, will serve as a guide to national regulatory
authorities (NRAs) and manufacturers in support of worldwide production of
these essential medicines. The production of snake antivenoms following good
manufacturing practices (GMP) should be the aim of all countries involved in
the manufacture of these life-saving biological products.
In addition to the need to produce appropriate antivenoms, there is a
need to ensure that antivenoms are appropriately used and that outcomes
for envenomed patients are improved. This entails improving availability and
access to antivenoms, appropriate distribution policies, antivenom affordability,
and training of health workers to allow safe, selective and effective use
of antivenoms and effective management of snake-bite envenoming. These
important issues are beyond the scope of this document and will not be further
addressed specifically here, but should be considered as vital components in the
care pathway for envenoming.
This second edition of the Guidelines was prepared in 2016 in order to
ensure that the information contained in these sections remains relevant to the
current production of snake antivenom immunoglobulins and their subsequent
control and regulation.
Major updates in this second edition include:
■■ inclusion of stronger animal welfare and ethical compliance
messages (section 4) to reinforce the importance of humane use of
animals in the production of antivenoms;
■■ updates to lists of medically important snakes to reflect new species
discoveries and recent nomenclatural changes (section 7; and

Appendix 1);
■■ revision of methodologies for serpentariums that produce venoms
to emphasize traceability and quality control, including the
recommendation to discontinue use of wild-capture/release strategies
for ethical and quality control reasons (section 9);
■■ increased emphasis on the specific health control of plasma donor
animals, particularly prior to, and during plasma collection session
(sections 12 and 14);
■■ updated lists of known potential equine virus contaminants
(section 16);
204
Annex 5
■■ redrafting and reorganization of sections on the quality control

(section 17), stability studies (section 18) and preclinical assessment
(section 19) of antivenoms, to incorporate new approaches and
technologies, and eliminate repetition;
■■ revised information on the clinical assessment of antivenoms
(section 20), as well as an expanded and strengthened discussion
on the role of NRAs and the need for national reference venom
collections independent from antivenom manufacturers (section 21;
see also section 10.3).

These WHO Guidelines provide guidance to NRAs and manufacturers on the
production, control and regulation of snake antivenom immunoglobulins. It
should however be recognized that some sections, such as: those dealing with
immunogen quality control, reference materials, and the production, purification
and testing of antibodies (sections 10–19); as well as most of the guidance which
deals with regulatory oversight (section 21); and the ethical use of laboratory
animals and plasma donor animals (section 4); may also apply to other types of
antivenoms, such as those produced for the treatment of envenoming caused by
spiders, scorpions and other organisms. There are also other immunoglobulin
products of animal origin for which some of the production methodologies
described here may be similar or identical – for example, the selection and
veterinary health care of animals; immunization regimens and use of adjuvants;
collection and control of animal plasma for fractionation; purification of
immunoglobulins; and control of infectious risks. These WHO Guidelines may
therefore have application beyond providing information for the production
of snake antivenom immunoglobulins, and may be applicable also to other
antivenoms or animal-derived immunoglobulin products (for example, equine-
derived botulism antitoxins).
3. Terminology
Antivenom – also called antivenin or anti-snake venom (ASV): a purified
fraction of immunoglobulins or immunoglobulin fragments fractionated from
the plasma of animals that have been immunized against one or more snake
venoms.
205
Apheresis: procedure whereby blood is removed from the donor,

separated by physical means into components and one or more of them returned
to the donor.
Batch: a defined quantity of starting material or product manufactured in
a single process or series of processes so that it is expected to be homogeneous.
Batch records: all documents associated with the manufacture of a batch
of bulk product or finished product. They provide a history of each batch of
product and of all circumstances pertinent to the quality of the final product.
Blood collection: a procedure whereby a single donation of blood is
collected in an anticoagulant and/or stabilizing solution, under conditions
designed to minimize microbiological contamination of the resulting donation.
Bulk product: any product that has completed all processing stages up to,
but not including, aseptic filling and final packaging.
Clean area: an area with defined environmental control of particulate
and microbial contamination constructed and used in such a way as to reduce
the introduction, generation, and retention of contaminants within the area.
Contamination: the undesired introduction of impurities of a
microbiological or chemical nature, or of foreign matter, into or on to a
starting material or intermediate during production, sampling, packaging, or
repackaging, storage or transport.
Convention on International Trade in Endangered Species of Wild
Fauna and Flora (CITES): an international agreement between governments
that aims to ensure that international trade in specimens of wild animals and
plants does not threaten their survival.
Cross-contamination: contamination of a starting material, intermediate
product or finished product with another starting material or product during
production.
Cross-neutralization: the ability of an antivenom raised against a venom,
or a group of venoms, to react and neutralize the toxic effects of the venom of a
related species not included in the immunizing venom mixture.
Common Technical Document (CTD) format: a specific format for
product dossier preparation recommended by WHO and the International
Conference on Harmonisation of Technical Requirements for Registration of
Pharmaceuticals for Human Use (ICH).
Desiccation: a storage process where venoms are dehydrated under
vacuum in the presence of calcium salts or phosphoric acid.
Effectiveness: the effectiveness of an antivenom is a measure of its
ability to produce a clinically effective outcome when used to treat snake-bite
envenoming.
Efficacy: the efficacy of an antivenom is a measure of the in vivo or in
vitro neutralizing potency against a specific activity of a venom or venoms.
206
Annex 5
Envenoming: injection of venom by an organism (for example, venomous

snake) into another organism, leading to pathological manifestations (also called
envenomation).
Fab: an antigen-binding fragment (Fab) of an immunoglobulin
comprising a heavy chain and a light chain that each have a single constant
domain and a single variable domain. Fab fragments result from the proteolytic
digestion of immunoglobulins by papain (or pepsin after F(abʹ)2 digestion).
F(abʹ)2: an immunoglobulin fragment comprising a pair of Fab fragments
connected by a protein hinge, and produced by proteolytic digestion of whole
immunoglobulins with pepsin.
Fractionation: large-scale process by which animal plasma is separated
to isolate the immunoglobulin fraction that is further processed for therapeutic
use or may be subjected to digestion with pepsin or papain to generate
immunoglobulin fragments. The term fractionation is generally used to describe
a sequence of processes, usually including plasma protein precipitation and/or
chromatography, ultrafiltration and filtration steps.
Freund’s complete adjuvant (FCA): an adjuvant that may be used in the
immunization process of animals to enhance the immune response to venoms.
It is composed of mineral oil, an emulsifier and inactivated Mycobacterium
tuberculosis.
Freund’s incomplete adjuvant (FIA): an adjuvant that may be used
in the immunization process of animals to enhance the immune response to
venoms. It is composed of mineral oil and an emulsifier.
Good clinical practice (GCP): an international standard for rigorous,
ethical and high quality conduct in clinical research, particularly in relation to all
aspects of the design, conduct, analysis, record-keeping, auditing and reporting
of clinical trials involving human subjects. GCP standards are established by the
ICH under Topic E 6 (R1).
Good manufacturing practice (GMP): that part of quality assurance
which ensures that products are consistently produced and controlled to the
quality standards appropriate to their intended use and as required by the
marketing authorization or product specification. It is concerned with both
production and quality control.
Immunization process: a process by which an animal is injected with
venom(s) to produce a high-titre antibody response against the lethal and other
deleterious components in the immunogen.
Immunoglobulin: immune system forming protein produced by B-cells
in plasma that can recognize specific antigens. These can be generated by
immunizing an animal (most often a horse) against a snake venom or a snake
venom mixture. Immunoglobulin G (IgG) is the most abundant immunoglobulin
fraction.
207
Immunoglobulin G (IgG): one of the five classes of antibodies produced

by the B-cells. It is synthesized in response to invasions by bacteria, fungi and
viruses. IgG crosses the placenta and protects the fetus. It is a complex protein
composed of four peptide chains – two identical heavy chains and two identical
light chains arranged in a typical Y shape of antibody monomers. Representing
approximately 75% of serum antibodies in humans, IgG has a molecular mass
of approximately 150 kDa.
Immunoglobulin M (IgM): another type of antibody. It is an
immunoglobulin of high molecular weight that is released into the blood early
in the immune response to be replaced later by IgG and is highly efficient in
binding complement. IgM antibodies make up about 5 to 10% of all the
antibodies in the body; they have a polymeric form, mostly as pentamers. IgM
has a molecular mass of approximately 970 kDa.
In-process control: checks performed during production to monitor
and, if necessary, to adjust the process to ensure that the antivenom conforms
to specifications. The control of the environment or equipment may also be
regarded as part of in-process control.
Manufacture: all operations of purchase of materials and products,
production, quality control, release, storage and distribution of snake antivenom
immunoglobulins, and the related controls.
Median effective dose – or effective dose 50% (ED50 ): the quantity of
antivenom that protects 50% of test animals injected with a median lethal dose
of venom.
Median lethal dose – or lethal dose 50% (LD50 ): the quantity of snake
venoms, injected intravenously or intraperitoneally, that leads to the death
of 50% of the animals in a group after an established period of time (usually
24–48 hours).
Minimum coagulant dose (MCD): the minimum amount of venom
(in mg/L or µg/mL) that clots either a solution of bovine fibrinogen (2.0 g/L) in
60 seconds at 37 °C (MCD-F) and/or a standard citrated solution of human
plasma (2.8 g/L fibrinogen) under the same conditions (MCD-P).
Minimum coagulant dose-F-effective dose (MCD-F100 ) and MCD‑P-
effective dose (MCD-P100 ): the minimum volume of antivenom or venom/
antivenom ratio, which completely prevents clotting induced by either one
MCD-F or MCD-P dose of venom.
Minimum defibrinogenating dose (MDD): the minimum amount of
venom that produces incoagulable blood in all mice tested within one hour
of intravenous injection.
Minimum defibrinogenating dose-effective dose (MDD100 ): the
minimum volume of antivenom or venom/antivenom ratio, at which the blood
208
Annex 5
samples of all injected mice show clot formation after administration of one or
more MDD doses of venom.
Minimum haemorrhagic dose (MHD): The minimum amount of
venom (in µg) that when injected intradermally in mice causes a 10 mm
haemorrhagic lesion within a predefined time interval (for example, 2–3 hours).
Minimum haemorrhagic dose-median effective dose (MHD50 ): the
minimum volume of antivenom (in µL) that reduces the diameter of haemorrhagic
lesions by 50% compared to those induced in animals who receive a control
solution of venom/saline.
Minimum myotoxic dose (MMD): the minimum amount of venom that
produces a four-fold increase in serum or plasma creatine kinase (CK) activity
above that of control animals.
Minimum myotoxic dose-median effective dose (MMD50 ): the
minimum amount of antivenom (in µL or the venom/antivenom ratio) that
reduces the serum or plasma CK activity by 50% compared to those induced in
animals who receive a control solution of venom/saline.
Minimum necrotizing dose (MND): the minimum amount of venom
(in µg) that when injected intradermally in groups of lightly anaesthetized mice
results in a necrotic lesion 5 mm in diameter within 72 hours.
Minimum necrotizing dose-median effective dose (MND50 ): the
minimum amount of antivenom (in µL or the venom/antivenom ratio) that
reduces the diameter of necrotic lesions by 50% compared to those induced in
animals who receive a control solution of venom/saline.
Monospecific antivenom: antivenoms that are raised from venom of a
single species, and are limited in use to that species or to a few closely related
species (typically from the same genus) whose venoms show clinically effective
cross-neutralization with the antivenom. The term “monovalent” is often used
and has the same meaning.
Nanofilter: filters, most typically with effective pore sizes of 50 nm or
below, designed to remove viruses from protein solutions.
National regulatory authority (NRA): WHO terminology to refer to
national medicines regulatory authorities. Such authorities promulgate medicine
regulations and enforce them.
Plasma: the liquid portion remaining after separation of the cellular
elements from blood collected in a receptacle containing an anticoagulant, or
separated by continuous filtration or centrifugation of anticoagulated blood in
an apheresis procedure.
Plasmapheresis: procedure in which whole blood is removed from the
donor, the plasma is separated from the cellular elements by sedimentation,
filtration, or centrifugation, and at least the red blood cells are returned to
the donor.
209
Polyspecific antivenom: antivenoms that are obtained by fractionating

the plasma from animals immunized with a mixture of venoms from more than
one species of venomous snake. The term “polyvalent” is often used and has the
same meaning.
Prion: a particle of protein that is thought to be able to self-replicate and
to be the agent of infection in a variety of diseases of the nervous system, such as
scrapie, mad cow disease and other transmissible spongiform encephalopathies
(TSEs). It is generally believed not to contain nucleic acid.
Production: all operations involved in the preparation of snake
antivenom immunoglobulins, from preparation of venoms, immunization of
animals, collection of blood or plasma, processing, packaging and labelling, to
its completion as a finished product.
Quality manual: an authorized, written controlled document that defines
and describes the quality system, the scope and operations of the quality system
throughout all levels of production, management responsibilities, key quality
systems processes and safeguards.
Quarantine: a period of enforced isolation and observation typically
to contain the spread of an infectious disease among animals. The same
terminology applies to the period of isolation used to perform quality control
of plasma prior to fractionation, or of antivenom immunoglobulins prior to
release and distribution.
Randomized controlled trial (RCT): randomized controlled trial of a
pharmaceutical substance or medical device.
Serpentarium: a place where snakes are kept, for example, for exhibition
and/or for collection of venoms.
Serum: a liquid portion remaining after clotting of the blood. Serum has
a composition similar to plasma (including the immunoglobulins) apart from
fibrinogen and other coagulation factors which constitute the fibrin clot.
Site Master File: an authorized, written controlled document containing
specific factual details of the GMP production and quality control manufacturing
activities that are undertaken at every site of operations linked to products that
a company produces.
Standard operating procedure (SOP): an authorized written procedure
giving instructions for performing operations not necessarily specific to a given
product or material (for example, equipment operation, maintenance and
cleaning; validation; cleaning of premises and environmental control; sampling
and inspection). Certain SOPs may be used to supplement product-specific
master and batch production documentation.
Toxin: a toxic substance, especially a peptide or protein, which is
produced by living cells or organisms and is capable of causing disease when
introduced into the body tissues. It is often also capable of inducing neutralizing
antibodies or antitoxins.
210
Annex 5
Traceability: ability to trace each individual snake, venom, immunized

animal, or unit of blood or plasma used in the production of an antivenom
immunoglobulin with each batch of the final product. The term is used to describe
forward and reverse tracing.
Validation: action of proving, in accordance with the principles of GMP,
that any procedure, process, equipment, material, activity, or system actually
leads to the expected results.
Venom: the toxic secretion of a specialized venom gland which, in the
case of snakes, is delivered through the fangs and provokes deleterious effects.
Venoms usually comprise many different protein components of variable
structure and toxicity.
Venom extraction – or venom collection or “milking”: The process of
collecting venom from live snakes.
Viral inactivation: a process of enhancing viral safety in which viruses
are intentionally “killed”.
Viral reduction: a process of enhancing viral safety in which viruses are
inactivated and/or removed.
Viral removal: a process of enhancing viral safety by partitioning viruses
from the components of interest.
4. The ethical use of animals

Current methods of antivenom production rely on the use of animals to
manufacture these life-saving products. For all animals, whether they are
venomous snakes from which venom is obtained for use as an immunogen;
the horses, sheep or other large animals that are injected with the venom, and
serve as living antibody factories, producing hyperimmune plasma from which
antivenom is derived; or the small laboratory animals sacrificed in order to test
the preclinical efficacy and safety of antivenoms, there is an absolute necessity
for all manufacturers to use animals humanely and ethically.
It is imperative that venom producers, antivenom manufacturers and
quality control laboratories that make use of animals in venom or antivenom
research, production, or in the preclinical evaluation of antivenoms adhere to
the highest ethical standards. The International guiding principles for biomedical
research involving animals (2012) developed by the International Council for
Laboratory Animal Science and the Council for International Organization of
Medical Sciences provide an international benchmark for the use of animals
in research. Compliance with national guidelines, laws and regulations is also
essential. All animal experimentation should be subject to regulatory oversight
at an institutional and national level. In many jurisdictions, the 3R principles of
Replacement, Reduction and Refinement have been adopted as cornerstones
211
of ethical use of animals, and WHO strongly recommends that every effort be
made to reduce pain, distress and discomfort to experimental animals – for
example, by the routine use of analgesia in mice used in these assays.
4.1 Ethical considerations for the use of venomous

snakes in the production of snake venoms
Venomous snakes kept in serpentariums for use in venom production should
be maintained according to nationally and internationally accepted ethical
standards. All relevant local regulations should be strictly adhered to, and where
required the use of venomous snakes in venom production should be conducted
in accordance with ethics approvals obtained from responsible authorities in
the jurisdiction. This particularly applies to the collection of wild specimens
and their transportation to serpentariums. It is important that specimens be
sourced from legal suppliers, and venom producers should ensure that the
collection localities of all specimens are known, and that evidence of legal
collection is supplied. As discussed in section 9.1.4.2 the practice of capturing
wild venomous snakes, extracting venom and releasing the snakes after
translocation into new habitat must be discontinued. This is not just because
of issues relating to traceability and quality control, which are fundamental to
production of antivenoms in accordance with GMP, but also because mounting
evidence demonstrates unacceptably high mortality among translocated
venomous snakes. Compliance with local ethical requirements for the keeping
of venomous snakes in captivity, the humane handling of specimens, veterinary
care and supervision, and euthanasia (when necessary for humane reasons)
should be maintained. Another important consideration for serpentariums
is the necessity to use other animals as food sources for venomous snakes.
The types of animals used as food, their production, humane euthanasia, or
in some cases, presentation to snakes as live prey, require appropriate ethical

considerations, and specific licences and ethics approvals may be required
to keep, breed and use some animals as sources of food for venomous snakes.
Venom producers must ensure that their operations comply with all necessary
regulations and requirements in this regard.
4.2 Ethical considerations for the use of large animals

in the production of hyperimmune plasma
The use of large animals (for example, horses, ponies, mules and sheep) in the
production of hyperimmune plasma requires constant veterinary supervision
and strict adherence to approved national and international ethical standards
for these animals. Equines are the most commonly used for production of
212
Annex 5
hyperimmune plasma in antivenom production and have specific physiological

and psychological requirements for good health and the minimization of pain
and distress. Manufacturers must recognize these needs and structure their use
of animals to ensure that their social, physical and environmental needs are
appropriately met. Relevant guidelines and regulations established by competent
authorities should be implemented. Veterinary care of animals should meet
the highest standards, and the health and welfare of individual animals used
for plasma production should be closely monitored at all times. The process
of immunizing donor animals with snake venoms raises important ethical
considerations, particularly because of the potential harm that can be caused by
some venoms (for example, neurotoxins, necrotic or cytotoxic venoms) and by
the adjuvants that are used in most immunization protocols, particularly Freund’s
complete adjuvant (FCA) or Freund’s incomplete adjuvant (FIA). Animals used
in plasma production may suffer considerable distress, pain or discomfort as a
result of the immunization process and all manufacturers have an obligation to
strictly comply with animal welfare and ethical use requirements and actively
work to minimize these deleterious effects. Similarly, the bleeding of animals to
collect hyperimmune plasma can be traumatic for donor animals if appropriate
techniques are not used to minimize negative effects, including fear, pain, distress
and physical harm. Manufacturers are encouraged very strongly to proactively
improve the welfare of large animals used in plasma production, and to develop
protocols that reduce suffering and improve the health of animals.
4.3 Ethical considerations for the use of animals

in preclinical testing of antivenoms
The preclinical testing of new or existing antivenoms necessitates the use of
experimental animals, typically rodents, particularly for essential median lethal
venom dose (LD50 ) and median effective antivenom dose (ED50 ) determination.
Despite reservations about the physiological relevance of these animal models to
human envenoming and the harm that these in vivo assays cause to the animals
(sections 19.2 and 19.3), they are used by both manufacturers and regulatory
authorities worldwide for determining venom lethality (LD50 ) and antivenom
neutralizing capacity (ED50 ) as these are currently the only validated means of
assessing venom toxicity and antivenom neutralizing potency. Non-sentient
or in vitro assays as alternatives to the standard venom LD50 and antivenom
ED50 in vivo tests have been promoted (2). Unfortunately, such systems have not
been developed to the point where they can fully replace the above-mentioned
preclinical assays. In the absence of effective alternatives, the continued use of
experimental animals is still justified by the considerable benefits to human
health of these preclinical assays.
213
4.4 Development of alternative assays to

replace murine lethality testing
In vivo murine assays cause considerable suffering and a 3R approach involving
innovation and validation should be applied in the development of standardized
LD50 and ED50 test protocols. Designing protocols that use the minimal
number of animals necessary and introducing procedures to minimize pain
and suffering is essential. The development of alternative methods to replace
animal testing in the preclinical evaluation of antivenoms should be encouraged.
When tests on live animals are absolutely necessary, anaesthesia or analgesia
should be considered and evaluated to ensure that the humane benefits of these
interventions to the experimental animals do not invalidate the objectives of
the assay by altering relevant physiological processes (3). In particular, the use
of analgesia is recommended when working with venoms that induce tissue
damage, and experimental evidence demonstrates convincingly that opioid drugs
relieve suffering without altering critical end-points such as LD50 and ED50 (4).
The establishment of humane end-points to reduce suffering and limiting the
duration of the assays to reduce the period of animal suffering is encouraged;
this requires appropriate standardization and validation within a quality
assurance framework.
4.5 Refinement of the preclinical assay protocols to reduce

pain, harm and distress to experimental animals
The substantial suffering caused to small animals by the preclinical assays is
outweighed by the considerable benefits to human health. Nevertheless, WHO
strongly encourages that opportunities to implement alternatives to the essential
and supplementary tests, according to the 3R, to reduce pain, harm and distress
be tested. Thus, designing protocols that use the minimum number of animals
necessary and introducing procedures to minimize pain and suffering is essential.
Analgesia should be considered, and evaluated to ensure that the humane benefits
of analgesia to the experimental animals do not invalidate the objectives of the
assay by altering relevant physiological processes (3). In particular, the use of
analgesia is recommended when working with venoms that induce tissue damage
(4). The establishment of humane end-points, instead of using survival/death as
the assay metric, is encouraged to reduce suffering and limit the duration of the
assays. The use of humane end-points also offers the opportunity to introduce
‘dose-staging’ into the experimental design (in which multiple doses are prepared
for the assays, one dose given and the next dose(s) selected based on the results of
giving the previous dose) to reduce the number of mice required for these assays.
All such efforts towards 3R require appropriate standardization and validation
within a quality assurance framework.
214
Annex 5
4.6 Main recommendations

■■ It is imperative that venom producers, antivenom manufacturers
and quality control laboratories that use animals in venom or
antivenom production, in research or in the preclinical evaluation
of antivenoms adhere to the highest ethical standards.
■■ Relevant national and international animal welfare and ethical use
guidelines and regulations should be adhered to.
■■ Wherever possible, alternative protocols and procedures that
minimize pain, suffering and physical or psychological distress to
animals should be developed and validated.
■■ The 3R approach should be applied in the development of
standardized and validated LD50 and ED50 test protocols.
5. General considerations
Snake antivenom immunoglobulins – antivenoms, antivenins, anti-snake-
bite serum and anti-snake venom (ASV) – are the only specific treatment for
envenoming by snake-bites. They are produced by the fractionation of plasma
that is usually obtained from large domestic animals hyperimmunized against
relevant venoms. Important but seldom used antivenoms may be prepared in
smaller animals. In general, when injected into an envenomed human patient,
an effective antivenom will neutralize toxins in any of the venoms used in its
production, and in some instances, will also neutralize venoms from closely
related species.
5.1 Historical background

Shortly after the identification of diphtheria and tetanus toxins, von Behring and
Kitasato reported the antitoxic properties of the serum of animals immunized
against diphtheria or tetanus toxins and suggested the use of antisera for the
treatment of these diseases (5). In 1894, von Behring diphtheria antitoxin was
first successfully administered by Roux to save children suffering from severe
diphtheria. Thus, serum therapy was born and the antitoxin was manufactured
by Burroughs Wellcome in the United Kingdom. The same year, Phisalix and
Bertrand (6) and Calmette (7) simultaneously, but independently, presented
during the same session of the same meeting their observations on the antitoxic
properties of the serum of rabbits and guinea-pigs immunized against cobra and
viper venoms, respectively. Immediately after his discovery of “antivenin serum
therapy”, Calmette became actively involved in proving its effectiveness in the
215
treatment of human envenoming. The first horse-derived antivenom sera that

he prepared were in clinical use in 1895 by Haffkine in India and by Lépinay in
Viet Nam. The latter reported the first successful use of antivenin serum therapy
in patients in 1896 (8).
5.2 The use of serum versus plasma as source material

Historically, the pioneers, Calmette, Vital Brazil and others, used serum
separated from the blood of hyperimmunized horses for the preparation of
antivenom (“antivenin serum therapy”). Later, antibodies (immunoglobulins) were
demonstrated to be the active molecules responsible for the therapeutic action
of “antivenom serum”. Subsequently, immunoglobulins, or immunoglobulin
fragments (F(abʹ)2, Fab), purified from serum were used instead of crude serum
(9, 10). Nowadays, plasmapheresis, whereby red blood cells are re-injected into
the donor animal within 24 hours of blood collection, is commonly employed
to reduce anaemia in the hyperimmunized animal that donates the plasma.
Accordingly, it is almost exclusively plasma rather than serum, which is used as
the starting material for the extraction of the immunoglobulin or its fragments
(11–13). Thus “snake antivenom immunoglobulin” is the preferred term, rather
than “anti-snake-bite serum” or “antiserum” which are no longer accurate.
5.3 Antivenom purification methods and product safety

The recognition of their role, and the purification of immunoglobulins from
other components of the serum or plasma of donor animals, was pioneered in
the earliest years of the last century using simple chemical reactions (14–18).
The subsequent discovery, more than half a century later, of the structure of
antibodies opened new doors to the fractionation of immunoglobulins. It
became possible to produce antibody fractions (F(abʹ)2 or Fab) that were believed
to potentially reduce the frequency of early and late antivenom reactions by

removing the Fc fragment from IgG (19). This was subsequently believed to
prevent complement activation and perhaps reduce the intensity of immune-
complex formation responsible for late antivenom reactions (serum sickness).
For 60–70 years, immunoglobulin F(abʹ)2 fragments have been widely used.
However, antivenom protein aggregation, and not Fc-mediated complement
activation, was increasingly identified as a major cause of antivenom reactions.
Thus, a critical issue in antivenom safety probably lies in the physicochemical
characteristics of antivenoms and not exclusively in the type of neutralizing
molecules constituting the active substance. It is also important to ensure that
the current methods of producing antivenoms provide a sufficient margin of
safety with regard to the potential risk of transmission of zoonoses.
216
Annex 5
5.4 Pharmacokinetics and pharmacodynamics of antivenoms

Rapid elimination of some therapeutic antivenoms (for example, when Fab
fragments are used) has led to recurrence of envenoming in patients. However,
the choice of preparing specific IgG or fragments appears to depend on the
size and toxicokinetics of the principal toxin(s) of the venoms. Large relative
molecular mass (Mr) bivalent antibodies (IgG and F(abʹ)2 fragments) may
be effective for the complete and prolonged neutralization of intravascular
toxins (for example, procoagulant enzymes), which have a long half-life in
envenomed patients. Low Mr and monovalent IgG fragments, such as Fab, may
be more appropriate against low-molecular-mass neurotoxins which are rapidly
distributed to their tissue targets and are rapidly eliminated from the patient’s
body, for example, scorpion and spider toxins (20).
5.5 Need for national and regional reference venom preparations

Antivenom production is technically demanding. The need to design appropriate
monospecific or polyspecific antivenoms (depending on the composition of
the snake fauna) is supported by the difference in venom composition among
venomous animals, in particular bearing in mind that:
■■ many countries can be inhabited by several medically important

species;
■■ there may be wide variation in venom composition (and hence
antigenicity) through the geographical range of a single species;
■■ in some circumstances there is no distinctive clinical syndrome to
direct the use of monospecific antivenoms.
However, similarities in the venom toxins of closely related venomous

species may result in cross-neutralization (paraspecific neutralization), thus
reducing the number of venoms required for the preparation of polyspecific
antivenoms. Cross-neutralization should be tested in animal models and ideally
by clinical studies in envenomed patients. Preclinical testing of antivenoms against
medically important venoms present in each geographical region or country is a
prerequisite for product licences and batch approval, and should always precede
clinical use in envenomed patients. This requires efforts by manufacturers and/or
regulators to establish regional or national reference venom preparations that can
be used to test the neutralization capacity of antivenoms. The national control
laboratory of the country where the antivenom will be used, or the manufacturer
seeking a licence for the antivenom, should perform such preclinical testing using
reference venom preparations relevant to the country or the geographical area.
217
6. Epidemiological background
The incidence of medically important snake-bites in different parts of the world
and the recognition of the species of greatest medical importance is fundamental
to the appropriate design of monospecific and polyspecific antivenoms in
countries and regions. Up-to-date epidemiological and herpetological information
is therefore highly relevant to antivenom manufacturers and regulators, especially
for the selection of the most appropriate venoms or venom mixtures to be used
in the production and quality control of antivenoms.
6.1 Global burden of snake-bites

Envenoming and deaths resulting from snake-bites are a particularly important
public health problem in rural tropical areas of Africa, Asia, Latin America and
Papua New Guinea (21). Agricultural workers and children are the most affected
groups. Epidemiological assessment of the true incidence of global mortality
and morbidity from snake-bite envenoming has been hindered by several well
recognized problems (22, 23). Snake-bite envenoming and associated mortality
are underreported because many victims (20–70% in some studies) do not seek
treatment in government dispensaries or hospitals and hence are not recorded.
This is compounded by the fact that medical posts in regions of high incidence
are unable to keep accurate records of the patients who do present for treatment,
and because death certification of snake-bite is often imprecise (24, 25).
Correctly designed population surveys, in which questionnaires are
distributed to randomly selected households in demographically well-defined
areas, are the only reliable method for estimating the true burden of snake-bites
in rural areas. The results of the few such surveys that have been performed
have shown surprisingly high rates of bites, deaths and permanent sequelae
of envenoming (25–29). However, because of the heterogeneity of snake-
bite incidence within countries, the results of surveys of local areas cannot
be extrapolated to give total national values. Most of the available data suffer
from these deficiencies and, in general, should be regarded as underestimates
and approximations.
However, the true burden of national snake-bite morbidity and mortality
in three South Asian countries has recently been revealed by the results of three
well-designed community-based studies. In India, a direct estimate of 46 000
snake-bite deaths in 2005 was derived from the Million Death Study (30), in
Bangladesh there were an estimated 589 919 snake-bites resulting in 6,041 deaths
in 2009 (31), and in Sri Lanka in 2012–2013, 80 000 bites, 30 000 envenomings
and 400 deaths in one year (32). Published estimates of global burden, employing
highly controversial methodologies, suggest a range from a minimum of 421 000
218
Annex 5
envenomings and 20 000 deaths up to as many as 2.5 million cases and more than
100 000 deaths each year (23, 33). In view of the recent data from South Asia,
these figures would seem to be underestimates. In addition, the number of people
left with permanent sequelae as a result of envenoming is likely to be higher
than the number of fatalities (21). As already identified, most of the estimated
burden of snake-bite is in sub-Saharan Africa, Central and South America and
South and South-East Asia.
The current literature on snake-bite epidemiology highlights the
inadequacy of the available data on this neglected tropical disease. There is
clearly a need to improve reporting and record-keeping of venomous bites in
health facilities, to support high-quality epidemiological studies of snake-bite in
different regions, and to improve the training of medical personnel. Wherever
possible, recording the species that caused the bite as well as death or injury
would greatly assist in documenting which species are of clinical significance in
individual countries. Making venomous bites notifiable and fully implementing
the use of the International Statistical Classification of Diseases and Related
Health Problems 10th Revision (34) in official death certification (for example,
T 63.0 snake venom) would further help to determine the burden of snake-bite
more accurately.

■■ In most parts of the world, snake-bites are underreported and
in some parts are completely unreported. This deficiency in
surveillance and the paucity of properly designed epidemiological
studies explain why the impact of this important public health
problem has remained for so long unrecognized and neglected.
■■ National health authorities should be encouraged to improve the
scope and precision of their epidemiological surveillance of this
disease by:
–– improving the training of all medical personnel so that they are
more aware of the local causes, manifestations and treatment of
venomous bites;
–– making venomous bites notifiable;
–– setting up standardized and consistent epidemiological
surveys;
–– improving the reporting and record-keeping of venomous bites
by hospitals, clinics, dispensaries and primary health-care posts,
and relating the bites to the species of venomous snake that
caused the bite wherever possible; and
219
–– fully implementing the use of the International Statistical

Classification of Diseases and Related Health Problems 10th
Revision (2007) (22) in official death certification (for example,
T 63.0 snake venom).2
7. Worldwide distribution of venomous snakes

7.1 Taxonomy of venomous snakes
Recognizing the species causing the greatest public health burden, designing and
manufacturing antivenoms and optimizing patient treatment are all critically
dependent on a correct understanding of the taxonomy of venomous snakes.
Like other sciences, the field of taxonomy is constantly developing. New species
are still being discovered, and many species formerly recognized as being
widespread have been found to comprise multiple separate species as scientists
obtain better information, often with new technologies. As the understanding of
the relationships between species is still developing, the classification of species
into genera is also subject to change. The names of venomous species used in
these guidelines conform to the taxonomic nomenclature that was current at
the time of publication. Some groups of venomous snakes remain under-studied
and poorly known. In these cases, the classification best supported by what
evidence exists is presented with the limitation that new studies may result in
changes to the nomenclature.
Clinicians, toxinologists, venom producers and antivenom manufacturers
should endeavour to remain abreast of these nomenclatural changes. These
changes often reflect improved knowledge of the heterogeneity of snake
populations, and may have implications for venom producers, researchers and
antivenom manufacturers. Although taxonomic changes do not necessarily
indicate the presence of “new” venoms, they strongly suggest that toxinological
and epidemiological research into these “new” taxa may be required to establish
their medical relevance, if any.

Since some of the names of medically important species have changed
in recent years, the following points are intended to enable readers to relate the
current nomenclature to information in the older literature:
■■ The large group of Asian arboreal pit vipers, which in recent years
had been split from a single genus (Trimeresurus), into a number
of new genera (for example, Cryptelytrops, Parias, Peltopelor,
Himalayophis, Popeia, Viridovipera, Ovophis and Protobothrops,
with a few species retained in Trimeresurus) based on prevailing
http://www.who.int/classifications/apps/icd/icd10online/ (accessed 15 February 2017).

2
220
Annex 5
views of the interrelationships between these groups, have now

largely been returned to Trimeresurus. There are divergent views on
this approach to the taxonomy of these snakes, and interested parties
should consult the literature. Some changes made in the early 1980s
have gained acceptance and been retained (that is, Protobothrops).
Medically important species formerly classified in Cryptelytrops
include Trimeresurus albolabris, T. erythrurus and T. insularis.
Viridovipera stejnegeri has been returned to Trimeresurus.
■■ It is likely that new species of cobra (Naja spp.) will be identified
within existing taxa in both Africa and Asia. Three new species
(N. ashei, N. mandalayensis and N. nubiae) have been described and
several subspecies elevated to specific status since 2000 (for example,
N. annulifera and N. anchietae, from being subspecies of N. haje),
in addition to the synonymization of the genera Boulengerina and
Paranaja within the Naja genus. Such changes may hold significance
for antivenom manufacturers and should stimulate further research
to test whether existing antivenoms cover all target snake populations.
■■ Several medically important vipers have been reclassified:
Daboia siamensis has been recognized as a separate species from
Daboia russelii; Macrovipera mauritanica and M. deserti have
been transferred to Daboia; the Central American rattlesnakes,
formerly classified with Crotalus durissus, are now Crotalus simus;
and Bothrops neuwiedi has been found to consist of a number
of different species, three of which (B. neuwiedi, B. diporus and
B. mattogrossensis) may be of public health importance.
It is recognized that there have been many accepted revisions of taxonomy
over the past few decades. These WHO Guidelines are aimed at a very wide
range of readers, and to assist in matching some old and familiar names with the
current nomenclature, Tables A5.1 and A5.2 summarize the major changes made
between 1999 and 2016. A list of relevant herpetological references is provided at
the end of Appendix 1 of these Guidelines.
Table A5.1
Genus-level name changes (1999–2016)
Currently accepted name Previous name(s)

Bothrocophias hyoprora Bothrops hyoprora
Bothrocophias microphthalmus Bothrops microphthalmus
Trimeresurus albolabris Cryptelytrops albolabris
221

Trimeresurus erythrurus Cryptelytrops erythrurus
Trimeresurus insularis Cryptelytrops insularis,
Trimeresurus albolabris insularis
Trimeresurus macrops Cryptelytrops macrops
Trimeresurus purpureomaculatus Cryptelytrops purpureomaculatus
Trimeresurus septentrionalis Cryptelytrops septentrionalis,
Trimeresurus albolabris septentrionalis
Daboia deserti Macrovipera deserti, Vipera mauritanica deserti,
Vipera lebetina deserti
Daboia mauritanica Macrovipera mauritanica,
Vipera lebetina mauritanica
Daboia palaestinae Vipera palaestinae
Daboia russelii Vipera russelii
Himalayophis tibetanus Trimeresurus tibetanus
Montivipera raddei Vipera raddei
Montivipera xanthina Vipera xanthina
Naja annulata Boulengerina annulata
Naja christyi Boulengerina christyi
Trimeresurus flavomaculatus Parias flavomaculatus
Trimeresurus sumatranus Parias sumatranus
Protobothrops mangshanensis Zhaoermia mangshanensis, Ermia mangshanensis,

Trimeresurus mangshanensis
Trimeresurus stejnegeri Viridovipera stejnegeri
Table A5.2
Changes resulting from new species descriptions or redefinitions (1999–2016)

Acanthophis crytamydros Previously part of Acanthophis rugosus
Acanthophis laevis Acanthophis antarcticus laevis, confused with
A. antarcticus or A. praelongus
222
Annex 5

Acanthophis rugosus (New Guinea) Acanthophis antarcticus rugosus, confused with
A. antarcticus or A. praelongus
Agkistrodon howardgloydi Agkistrodon bilineatus howardgloydi
Agkistrodon russeolus Agkistrodon bilineatus russeolus
Agkistrodon taylori Agkistrodon bilineatus taylori
Bitis gabonica Bitis gabonica gabonica
Bitis harenna New species
Bitis rhinoceros Bitis gabonica rhinoceros
Bothrops diporus Bothrops neuwiedi diporus
Bothrops mattogrossensis Bothrops neuwiedi mattogrossensis,
B.n. bolivianus
Bothrops pubescens Bothrops neuwiedi pubescens
Bungarus persicus New species
Cerrophidion sasai Previously part of Cerrophdion godmani
Cerrophidion wilsoni Previously part of Cerrophidion godmani
Crotalus oreganus Previously considered part of Crotalus viridis
Crotalus ornatus Previously considered part of Crotalus molossus
Crotalus simus Crotalus durissus durissus (Central American
populations of C. durissus complex)
Crotalus totonacus Crotalus durissus totonacus
Crotalus tzabcan Crotalus simus tzabcan, Crotalus durissus tzabcan
Daboia russelii Daboia russelii russelii, Daboia r. pulchella
Daboia siamensis Daboia russelii siamensis, D.r. limitis, D.r. sublimitis,
D.r. formosensis
Echis borkini Previously part of Echis pyramidum
Echis omanensis Previously known as NE population of Echis
coloratus
Gloydius intermedius Previously named Gloydius saxatilis
Hypnale zara New species
223

Lachesis acrochorda Previously part of Lachesis stenophrys
Naja arabica Previously part of Naja haje
Naja anchietae Naja annulifera anchietae, Naja haje anchietae
Naja ashei Previously part of Naja nigricollis
Naja nigricincta Naja nigricollis nigricincta, Naja nigricollis woodi
Naja nubiae Previously part of Naja pallida
Naja senegalensis Previously part of Naja haje
Pseudechis rossignolii Pailsus rossignolii, previously part of Pseudechis
australis
Pseudonaja aspidorhyncha Previously part of Pseudonaja nuchalis
Pseudonaja mengdeni Previously part of Pseudonaja nuchalis
Thelotornis mossambicanus Thelotornis capensis mossambicanus
Thelotornis usambaricus Thelotornis capensis mossambicanus
Trimeresurus cardamomensis Previously part of Trimeresurus macrops
Trimeresurus rubeus Previously part of Trimeresurus macrops
Tropidolaemus philippensis Previously part of Tropidolaemus wagleri
Tropidolaemus subannulatus Previously part of Tropidolaemus wagleri
Vipera renardi Previously part of V. ursinii
Walterinnesia morgani Previously part of Walterinnesia aegyptia
7.2 Medically important venomous snakes

Based on current herpetological and medical literature, it is possible to partially
prioritize the species of snakes that are of greatest medical importance in different
regions. Detailed statistics on the species of snakes responsible for morbidity and
mortality throughout the world are lacking, except for a few epidemiological
studies which include rigorous identification of the biting snake in a few scattered
localities. Thus, establishing a list of medically important species for different
countries, territories and other areas relies, at least in part, on extrapolation from
the few known studies, as well as on the biology of the snake species concerned:
for example, where species of a group of snakes are known to be of public health
224
Annex 5
importance, based on epidemiological studies, it seems reasonable to deduce

that closely related species with similar natural history occurring in hitherto
unstudied regions are also likely to be medically important. Examples include
Asian cobras in several under-studied regions of Asia, lowland Bungarus species
in Asia, and spitting cobras in Africa.
Tables A5.3–A5.6 list the species of venomous snakes of greatest medical
importance in each of four broad geographical regions. Species listed in these
tables are either:
■■ those that are common or widespread in areas with large human
populations and which cause numerous snake-bites, resulting in high
levels of morbidity, disability or mortality among victims; or
■■ poorly known species that are strongly suspected of falling into this
category; or
■■ species that cause major and life-threatening envenoming responsive
to antivenom, but are not common causes of bites.
The venoms of these species should be considered a starting point for
establishing the most important targets for antivenom production. The need for
additional epidemiological and toxinological research to better define which
venoms to include and exclude for antivenom production in various regions,
territories and countries around the world is emphasized. Detailed data from
countries, territories and other areas on species believed to contribute most to the
global burden of injury, and/or which pose the most significant risk of morbidity
or mortality are provided in Appendix 1 of these Guidelines. Illustrations of
some important venomous snakes of Africa and the Middle East are shown in
Figs A5.1 and A5.2.
Table A5.3
Medically important venomous snakes: Africa and the Middle East
North Africa/Middle East

Atractaspididae: Atractaspis andersonii; Elapidae: Naja arabica, Naja haje, Naja oxiana;
Viperidae: Bitis arietans; Cerastes cerastes, Cerastes gasperettii; Daboia mauritanica,
Daboia palaestinae; Echis borkini, Echis carinatus, Echis coloratus, Echis omanensis,
Echis pyramidum; Macrovipera lebetina; Montivipera xanthina; Pseudocerastes persicus
Central sub-Saharan Africa
Elapidae: Dendroaspis jamesoni, Dendroaspis polylepis; Naja anchietae, Naja haje,
Naja melanoleuca, Naja nigricollis; Viperidae: Bitis arietans, Bitis gabonica, Bitis nasicornis;
Echis leucogaster, Echis ocellatus, Echis pyramidum
225

Eastern sub-Saharan Africa
Elapidae: Dendroaspis angusticeps, Dendroaspis jamesoni, Dendroaspis polylepis;
Naja anchietae, Naja annulifera, Naja ashei, Naja haje, Naja melanoleuca,
Naja mossambica, Naja nigricollis; Viperidae: Bitis arietans, Bitis gabonica, Bitis nasicornis;
Echis pyramidum
Southern sub-Saharan Africa
Elapidae: Dendroaspis angusticeps, Dendroaspis polylepis; Naja anchietae,
Naja annulifera, Naja mossambica, Naja nigricincta, Naja nivea; Viperidae: Bitis arietans
Western sub-Saharan Africa
Elapidae: Dendroaspis jamesoni, Dendroaspis polylepis, Dendroaspis viridis; Naja haje,
Naja katiensis, Naja melanoleuca, Naja nigricollis, Naja senegalensis;
Viperidae: Bitis arietans, Bitis gabonica, Bitis nasicornis, Bitis rhinoceros; Cerastes cerastes;
Echis jogeri, Echis leucogaster, Echis ocellatus
Table A5.4
Medically important venomous snakes: Asia and Australasia
Central Asia
Elapidae: Naja oxiana; Viperidae: Echis carinatus; Gloydius halys; Macrovipera lebetina
East Asia
Elapidae: Bungarus multicinctus; Naja atra; Viperidae: Trimeresurus albolabris;
Daboia russelii; Deinagkistrodon acutus; Gloydius blomhoffii, Gloydius brevicaudus;
Protobothrops flavoviridis, Protobothrops mucrosquamatus; Trimeresurus stejnegeri
South Asia
Elapidae: Bungarus caeruleus, Bungarus ceylonicus, Bungarus niger, Bungarus sindanus,
Bungarus walli; Naja kaouthia, Naja naja, Naja oxiana; Viperidae: Trimeresurus erythrurus;
Daboia russelii; Echis carinatus; Hypnale hypnale; Macrovipera lebetina
South-East Asia (excluding Indonesian West Papua)
Elapidae: Bungarus candidus, Bungarus magnimaculatus, Bungarus multicinctus,
Bungarus slowinskii; Naja atra, Naja kaouthia, Naja mandalayensis, Naja philippinensis,
Naja samarensis, Naja siamensis, Naja sputatrix, Naja sumatrana;
Viperidae: Calloselasma rhodostoma; Trimeresurus albolabris, Trimeresurus erythrurus,
Trimeresurus insularis; Daboia siamensis; Deinagkistrodon acutus
226
Annex 5

Australo-Papua (includes Indonesian West Papua)
Elapidae: Acanthophis laevis; Notechis scutatus; Oxyuranus scutellatus;
Pseudechis australis; 3 Pseudonaja affinis, Pseudonaja mengdeni, Pseudonaja nuchalis,
Pseudonaja textilis
3
Table A5.5
Medically important venomous snakes: Europe
Central Europe
Viperidae: Vipera ammodytes
Eastern Europe
Viperidae: Vipera berus
Western Europe
Viperidae: Vipera aspis, Vipera berus
Table A5.6
Medically important venomous snakes: the Americas
North America
Viperidae: Agkistrodon bilineatus, Agkistrodon contortrix, Agkistrodon piscivorus,
Agkistrodon taylori; Bothrops asper; Crotalus adamanteus, Crotalus atrox,
Crotalus horridus, Crotalus oreganus, Crotalus simus, Crotalus scutulatus,
Crotalus molossus, Crotalus viridis
Caribbean
Viperidae: Bothrops cf. atrox (Trinidad), Bothrops caribbaeus (St Lucia),
Bothrops lanceolatus (Martinique); Crotalus durissus (Aruba)
Central America
Viperidae: Bothrops asper; Crotalus simus
Pseudechis australis is common and widespread and causes numerous snake-bites; bites may be severe,
3
although this species has not caused a death in Australia since 1968.
227

South America
Viperidae: Bothrops alternatus, Bothrops asper, Bothrops atrox, Bothrops brazili,
Bothrops bilineatus, Bothrops diporus, Bothrops jararaca, Bothrops jararacussu,
Bothrops leucurus, Bothrops matogrossensis, Bothrops moojeni, Bothrops pictus,
Bothrops venezuelensis; Crotalus durissus; Lachesis muta
7.3 Minor venomous snake species

In many countries, territories and other areas there are species of snakes that
rarely bite humans but are capable of causing severe or fatal envenoming. Their
medical importance may not justify inclusion of their venoms in the immunizing
mixture for production of polyspecific antivenoms but the need to make
antivenoms against these species needs to be carefully analysed.
In some cases, such as with some Central American pit vipers
(genera Agkistrodon, Porthidium, Bothriechis, Atropoides among others), there
is clinically effective cross-neutralization of venoms by standard national
polyspecific antivenoms (35).
In other cases, there is no effective cross-neutralization and manufacturers
may therefore consider that the production of a monospecific antivenom is
justified for use in potentially fatal cases of envenoming, provided that such cases
can be identified. Such antivenoms are currently available for envenoming by
the boomslang (Dispholidus typus), desert black snake (Walterinnesia aegyptia),
Arabian burrowing asp (Atractaspis andersonii) (36), king cobra (Ophiophagus
hannah), Malayan krait (Bungarus candidus) (36) “yamakagashi” (Rhabdophis
tigrinus) and red-necked keelback (R. subminiatus), Martinique’s “Fer-de-lance”
(Bothrops lanceolatus), St Lucia’s B. caribbaeus, and some species of American
coral snake (Micrurus).

No antivenoms are currently available for envenoming by species such
as African bush vipers (for example, Atheris, Proatheris), berg adder (Bitis
atropos) and several other small southern African Bitis spp. (for example,
B. peringueyi), Sri Lankan and south-west Indian hump-nosed vipers (Hypnale
spp.) (37, 38), many Asian pit vipers (“Trimeresurus” sensu lato), some species
of kraits (for example, B. niger) and all but one species of burrowing asp (genus
Atractaspis).
An alternative to antivenom production against species that cause few,
but potentially severe envenomings, is to manufacture polyspecific antivenoms
for broadly distributed groups that have similar venom compositions (for
example, African Dendroaspis and Atractaspis; Asian “green pit vipers”; American
Micrurus). This may result in antivenoms that offer broad protection against
228
Annex 5
venoms from minor species within genera, or species whose bites are less
frequent than those of others in the same taxonomic groups (that is, genus,
subfamily or family).
7.4 Sea snake venoms

Although venomous marine sea snakes have not been included in the tables of
medically important venomous snakes, it should be recognized that there are a
number of species of marine snakes with potent venoms that can cause illness
or death. Available evidence, particularly clinical experience, indicates that the
major sea snake antivenom that is currently commercially available, which
uses venom of a single sea snake, Hydrophis schistosus (previously known as
Enhydrina schistosa), in the immunizing venoms mixture, is effective against
envenoming by other sea snake species for which there are clinical data. Further
research would be needed to better define the full extent of cross-neutralization
offered by this antivenom against other sea snake species.

■■ Clinicians, toxinologists, poison centres, regulators, venom
producers and antivenom manufacturers should be well informed
about current nomenclature and new changes to taxonomy, so as
to ensure the currency of information, correct identification of
species in their countries, and correct selection and sourcing of
venoms used in the manufacture of antivenoms.
■■ Identification of the medically important venomous snakes
(those that cause the greatest burden of injury, disability and/
or mortality) is a critical prerequisite to meeting the need for
efficacious antivenom. Improving the quality of the available data
and correcting and amplifying the level of geographical detail and
precision of attribution should be important priorities.
■■ Support for establishment of local capacity for venom production
as a means of ensuring that venom immunogens from
geographically representative populations of medically important
snake species are used in antivenom production would improve
antivenom specificity.
229
Fig. A5.1
Medically important North African and Middle Eastern venomous snakes:
(A) Egyptian cobra (Naja haje), (B) East Africa carpet viper (Echis pyramidum), (C) puff
adder (Bitis arietans), (D) Saharan horned viper (Cerastes cerastes) and (E) Levant viper
(Macrovipera lebetina)
A B
D E
230
Annex 5
Fig. A5.2
Medically important sub-Saharan African venomous snakes: (A) West African carpet
viper (Echis ocellatus), (B) Gaboon viper (Bitis gabonica), (C) Black mamba (Dendroaspis
polylepis), (D) Black-necked spitting cobra (Naja nigricollis), (E) Mozambique spitting
cobra (Naja mossambica)
A B
D E
231
8. Antivenoms design: selection of snake venoms

Venomous snakes exhibit significant species- and genus-specific variation
in venom protein composition (39). The clinical effectiveness of antivenom is
therefore largely restricted to the venom(s) used in its manufacture. It is therefore
imperative that antivenom manufacturers carefully consider the venoms used
in antivenom manufacture by first defining the geographical area where the
antivenom will be deployed, and sequentially:
■■ identifying the most medically important snakes in that region;

■■ examining the venom protein composition of the snakes, including
information from relevant literature;
■■ conducting antivenom preclinical efficacy tests on venoms of all the
most medically important snakes in that region.
8.1 Selection and preparation of representative venom mixtures

Appendix 1 presents an up-to-date list of the most medically important venomous
snake species by country, region and continent. The venoms from Category 1
snakes must be included for antivenom production and venoms from Category 2
snakes only excluded after careful risk–benefit assessment.
It is important to appreciate that there are variations in venom
composition and antigenicity: (a) within the geographical range of a single
species; and (b) between snakes of different ages (40, 41). Therefore, venom
should be collected from specimens of different geographical origins
and ages, and mixed before being used for immunization (see section 9 on
venom preparation). The greater the intra-specific variation, the more snake
specimens of distinct origin and age are required to create an adequate venom
immunization mixture.
Cross-neutralization of venoms with similar protein composition

profiles to the venoms used for immunization may extend the effectiveness of
some antivenoms, but requires, minimally, preclinical efficacy testing to identify
the potential cross-neutralization capacity of an antivenom. In vitro preclinical
immunological cross-reactivity testing alone is NOT an adequate measure of
antivenom efficacy.
8.2 Manufacture of monospecific or polyspecific antivenoms

Antivenom manufacturers face an early, critical decision as to whether the
antivenom should possess monospecific or polyspecific effectiveness.
232
Annex 5
8.2.1 Monospecific antivenoms

Monospecific antivenoms are manufactured with venoms from a single venomous
snake species, and their effectiveness is largely restricted to that snake species.
These conditions apply in areas where:
■■ there is only one medically important species (for example,
Vipera berus in Scandinavia and the United Kingdom ) or where
one species is responsible for the majority of cases (for example,
Oxyuranus scutellatus in southern Papua New Guinea);
■■ a simple blood test, suitable for use even in under-resourced
health‑care centres, can define the biting species (for example,
detection of incoagulable blood by the 20-minute whole blood
clotting test in the northern third of Africa, where only Echis spp.
cause coagulopathy);
■■ a simple algorithmic approach allows the species to be inferred
from the pattern of clinical and biological features;
■■ there is a reliable and affordable rapid immunodiagnostic test
readily available allowing the toxins to be identified unambiguously
(currently only available in Australia).
Monospecific antivenoms can be effective in treating envenoming by
a few closely related species whose venoms show clinically effective cross-
neutralization – but this requires preclinical and clinical confirmation.
8.2.2 Polyspecific antivenoms

Most tropical countries are inhabited by several medically important snake
species, and it is commercially unrealistic to develop multiple monospecific
antivenoms. In these cases, the manufacture of polyspecific antivenoms is highly
recommended. Polyspecific antivenoms are designed to contain IgG effective
against venoms from multiple species or genera of venomous snakes in a defined
region. Manufacturing protocols of polyspecific antivenom include:
1. Mixing venoms from multiple snake species or genera (sometimes
in amounts quantitatively associated with medical importance,
immunogenicity etc.) and immunizing donor animals with
this mixture. Immunizing an animal with venoms from several
taxonomically related snakes (for example, different vipers) can
have the advantage over monospecific antivenom of increasing the
titre of neutralizing IgG to any one snake venom (42).
233
2. Immunizing groups of donor animals with distinct venom

mixtures and then mixing the hyperimmune plasma from each
group of animals.
3. Immunizing groups of donor animals with distinct venom
mixtures and then mixing the monospecific antivenom IgGs to
formulate the final polyspecific antivenom.
When using options 2. and 3. it is important to monitor the efficacy
for each monospecific antivenom to ensure that the efficacy of the mixed final
product is consistent, reproducible and in line with the product specification for
each individual venom. This “combined monospecific antivenoms” approach
anticipates that the amount of neutralizing IgG targeting each individual venom
will be proportionally diluted – necessitating administration of more vials to
reverse venom pathology, which in turn increases the risks of adverse reactions.
In some regions, it is possible to differentiate envenoming by detecting
distinct clinical syndromes: neurotoxicity, haematological disturbances
(haemorrhage or coagulopathy) and/or local tissue damage. Such situations justify
the preparation of syndrome-specific polyspecific antivenoms by immunizing
donor animals with mixtures of either neurotoxic venoms or venoms causing
haemorrhage and/or coagulopathy and local tissue damage.
In most tropical regions where snake-bite is a significant medical burden,
polyspecific antivenoms offer significant clinical advantages and their production
should be encouraged. They can also offer greater commercial manufacturing
incentives (economies of scale) than monospecific antivenoms because of
their significantly greater geographical and snake-species cover – increasing
the likelihood of their delivery to victims residing in regions where antivenom
manufacture is not government subsidized.
Main recommendations
8.3
■■ Prior to importing antivenoms, national health authorities should

carefully consider their regional threat from venomous snakes to
guide their antivenom requirements.
■■ The design of the venom mixture used in immunization, and the
decision to prepare monospecific or polyspecific antivenoms, must
be guided by the epidemiological and clinical information on
snake-bites in the defined country, region or continent.
■■ In most tropical countries polyspecific antivenoms are likely to
have significant clinical and logistical advantages over monospecific
antivenoms, particularly in the absence of rapid, affordable snake
venom diagnosis.
234
Annex 5
■■ Polyspecific antivenom may be prepared from IgG of donor animals

immunized with a mixture of venoms, or by mixing monospecific
antivenoms.
■■ Manufacturers seeking marketing authorization for antivenoms
in a given country should provide experimental evidence from
preclinical testing that the product exhibits a neutralization
capacity against different local venoms (see section 19).
■■ National health authorities should organize independent preclinical
efficacy testing prior to importation of any antivenom to avoid
national distribution of dangerously ineffective products.
9. Preparation and storage of snake venom

Venom preparations are used both to hyperimmunize animals as part of
antivenom production, and to provide reference venom samples for routine
and/or preclinical potency assessment of antivenoms. According to GMP for
pharmaceutical products, snake venoms are starting materials, and therefore
ensuring their quality is critical, and their preparation should follow the principles
and recommendations stated below. The essential principles of quality systems
should be applied to venom production including traceability, reproducibility,
taxonomic accuracy and hygiene control. Manufacturers of snake venoms used
in antivenom production should strive to comply with WHO’s Guidelines on
GMP for biological products and Guidelines for good manufacturing practices
for pharmaceutical products.4
Venoms used for antivenom manufacture should be representative of
the snake population living in the area where the antivenom is to be used. To
take account of the variability in venom composition within a species (43–47),
it is imperative that the venom of an adequate number of individual snakes
(generally no fewer than 20 specimens, including males and females) collected
from various regions covering the entire geographical distribution of the
particular venomous snake species should be collected together. Consideration
should also be given to including venom from juvenile or sub-adult snakes in
these venom pools as there is strong evidence of age-related venom variation
within individual specimens and populations (48). A similar approach should
be used in the preparation of Standard Reference Venoms (national or regional)
for use in the validation of antivenom products by reference laboratories and
WHO Good Manufacturing Practices for Biological Products. WHO Technical Report Series, No. 996, 2016,
4
Annex 3. (Replacement of Annex 1 of WHO Technical Report Series, No. 822).

235
regulatory agencies (see section 10) or in preclinical testing of antivenoms by

manufacturers (see section 19).
Venom producers should ensure that they fully document, and can
provide evidence of:
■■ geographical origin and the length or age (juvenile or adult) of each
individual snake used for venom production;
■■ taxonomic details of each snake species used;
■■ correct implementation of compliance with local wildlife legislation,
and the Convention on International Trade in Endangered Species
(CITES) documents in the case of endangered species;
■■ application of appropriate withholding rules (for example, not
collecting venom from animals under quarantine, or which are
gravid, injured, sick or in poor condition);
■■ individual identification of snake specimens contributing to each
venom batch;
■■ traceability of each venom batch;
■■ appropriate handling and stabilization of venoms (for example,
rapid freezing of the venom after collection and lyophilization for
long-term stable storage);5
■■ quality control confirmation of batch-to-batch consistency of
venoms of each species and country of origin (for example, sodium
dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
or high-performance liquid chromatography (HPLC) profiling of
venoms, measurement of residual moisture in lyophilized venom);
■■ confirmation of batch-to-batch similarity of venom of the
same origin.
9.1 Production of snake venoms for immunization

The maintenance of a serpentarium and the handling of snakes used for
antivenom production should comply with quality systems principles.
9.1.1 Quarantine of snakes

All new accessions should be quarantined for at least 2 months in a special
“quarantine room” which should be located as far as possible from the “production
rooms” where snakes qualified for venom production are kept.
Desiccation or vacuum-drying may be acceptable if proven to ensure stability of the preparation.

5
236
Annex 5
On arrival, snakes should be examined by a specialized veterinary surgeon

(or experienced person) for ectoparasites, wounds and fractures. Endoparasites
(nematodes, cestodes, trematodes and pentastomids) should be eliminated using
broad-spectrum antiparasitic drugs and any injury must be adequately treated
by a veterinarian (49–51). Some viruses can be transmitted between different
species, and between different families of snakes. Therefore, different families
should be kept in different rooms.
Sick snakes should be treated and their quarantine extended for 1–2
months after complete clinical recovery. Sick animals found in “production
rooms” may be treated in situ (although quarantine is preferable) but they cannot
be used for venom production. If an antibiotic treatment is given, the snake should
not be used to obtain venom for 4 weeks following the end of the treatment.
When housed in good conditions, adult snakes collected from the wild can live
in captivity for 10 years or more. When handling snakes, the risk of infection
with human mosquito-borne viruses such as Japanese encephalitis should be
prevented, since arbovirus infections have been reported in some snakes (52).
9.1.2 Maintenance of captive snakes for venom production

Individual snakes should preferably be housed in separate cages large enough to
allow them to move about, according to local and international standards. There
are several acceptable options for the design of the cages. Transparent or black
(for burrowing snakes) plastic boxes are recommended. Cage materials should
be impermeable, free from fissures, and inert to disinfectants, cleaning chemicals
and common solvents. Cleaning and disinfecting agents should be carefully
selected to ensure they do not have adverse effects on the snakes. Cages should be
adequately ventilated but perforations or mesh must be small enough to prevent
escape. Ventilation holes should be clearly marked as hazard areas since there is
a risk of accidental envenoming (for example, spitting cobras have been known
to spray venom through such openings, and large vipers have fangs which can
extend through a small hole if the snake strikes). In the case of gravid viviparous
snakes, the ventilation holes or mesh should be sufficiently fine to prevent escape
of their tiny, liveborn offspring. The cage interior should be visible from the
outside to allow safe maintenance and handling. Access to cages through doors,
lids or sliding panels should facilitate management without compromising safety
or allowing snakes to escape. Be wary of cages with internal ledges or lips above
doors, as some snakes can conceal themselves above them out of sight of the
keepers. A disposable floor covering (for example, newspaper) is recommended.
Cryptic and nocturnal species should be provided with a small shelter where
they can hide.
The use of “hide boxes” is increasingly common as these provide both
a more reassuring environment for the snake, and increased safety for keepers.
237
Hide boxes should be designed to be slightly larger than the curled snake, with
an entrance/exit hole large enough to allow a recently fed snake easy access, plus
some simple closure device to lock the snake in the hide box. This will allow
removal of the snake from the cage without hazard to the keeper, making routine
cage maintenance simpler and safer. Hide boxes can be made from plastic, wood
or even cardboard (which is inexpensive and can be discarded and replaced
regularly). Permanent hide boxes should be readily cleanable or autoclavable. The
roof or side of the hide box should be removable, to allow easy and safe extraction
of the snake when required.
Cages should be thoroughly cleaned and disinfected when soiled (daily
if necessary). Faeces and uneaten or regurgitated food items should be removed
as soon as possible. To avoid misidentification of the snake, a microchip should
be implanted in the hypodermal layer of the snake’s posterior region and a label
bearing its individual data should be attached to the cage and transferred with
the snake when it is moved to another cage. Water should be freely available,
and for species from humid climates more frequent watering or misting may
be required, particularly when sloughing. Water should be changed regularly
and as soon as it becomes contaminated. Water treatment by ultraviolet (UV)
sterilization or acidification may be considered.
Tens of cages may be accommodated in the same “production room”,
provided that there is enough space for maintenance and venom extraction. This
room should be kept as clean as possible at all times, and thoroughly cleaned at
least weekly. Measures should be taken to minimize or eliminate contamination
or spread of diseases. The use of antiseptic hand washes, disposable over-clothing,
antiseptic foot wash trays at entry and exit points, and other measures should be
routine. The temperature and humidity of the snake room should be controlled
according to the climatic requirements of the particular snake species. Ventilation
should be ensured using fans, air-conditioning, or air renewing systems.
Access to snake rooms should be restricted to personnel responsible

for their maintenance. The rooms should be kept locked, with any windows
permanently closed or protected by bars and mosquito proofing. Access
should be via a safety porch not allowing simultaneous door opening and
with a transparent panel allowing a view of the entire snake room for pre-
entry safety inspections. The spaces below the doors should be less than 3 mm
and all openings to the exterior (for example, water pipes, drainage conduits,
ventilation entrances and exits) should be protected by grilles having holes
smaller than 3 mm. Natural light is often used; however, when not available,
artificial light should be turned on for 12 hours during the day and turned off
during the night for tropical species, but species from temperate zones may have
different requirements. Snakes of the same species, collected at the same time in
the same area should be placed in the same racks. The same “production room”
238
Annex 5
can contain snakes of different species, provided that they require similar living
conditions (that is, temperature and humidity).
When kept under favourable housing and climatic conditions, and if left
undisturbed, snakes will reproduce in captivity (53). Animals should be mated
only with specimens from the same species, subspecies and local origin (54,
55). Sexing can be difficult, but is helped by the use of intra-cloacal probes. The
male and the female should be individually identified and separated soon after
copulation. The female should be kept under careful surveillance. Eggs from
oviparous snakes and neonates from viviparous snakes should be removed from
the female’s cage as soon as possible. Differences in the venom composition of
adult and juvenile snakes have been reported in some species (43, 48, 56–58), and
where this is known to occur or is suspected, the venom of a certain proportion
of juvenile snakes might be mixed with that of adults during the production of
venom batches.
The ideal frequency of feeding captive snakes depends on the species and
age of the snake, varying from twice per week to once per month. Snakes are
usually fed after venom extraction, ideally with dead mice or other appropriate
prey according to the snake species. Animals such as rats and mice that are raised
to feed snakes should be produced under appropriate quarantine standards in
facilities designed for this task. Humane euthanasia should be employed in the
killing of food animals, and ideally these food animals should be frozen for at
least 7 days before being thawed for use. Some snakes will only accept living prey,
but attempts should be made to wean them onto dead prey, and all local ethical
standards should be followed in the production and use of food animals. Snake-
eating species, such as kraits, coral snakes and king cobras, can be enticed to
take dead mice if the prey is first flavoured with snake tissue fluids, although any
such material should be frozen first for at least 7 days to kill parasites, before it is
thawed for use. Some coral species can be fed with fish strips (59). Living, dead or
regurgitated prey should not be left in the cage for more than a few hours. Force
feeding may be necessary for neonates and snakes that persistently refuse to feed.
Feeding time affords an opportunity to carefully check the snake for abnormal
behaviour, wounds, and possible infections and to give dietary supplements when
necessary. Individual feeding records are crucial. They should include details of
what, when and how prey was offered, when it was consumed and whether it
was regurgitated. The health of captive snakes can be estimated and recorded
by observing regular feeding and by measuring their weight and length. These
data are best stored on a computer system, using a “barcode” for each snake,
or, alternatively, using a reliable manual recording system, and constitute useful
records related to the venom batches produced. Venom extraction rooms should
be equipped with emergency eyewash stations and safety showers as is the case in
laboratories where there is a risk of chemical contact hazards.
239
9.1.3 General maintenance of a serpentarium

Serpentariums should be designed to comply with appropriate GMP principles.
Quarantine facilities should be isolated in all respects from the main animal
housing area, and should have separate air-handling systems, or be in a separate
building. Maintenance areas such as storerooms, rooms for cleaning and
sanitizing cages and racks, animal houses for production of food animals (for
example, rodents or invertebrates), and rooms used for administration or for
venom processing, venom quality control and secure storage of venoms, should
also be separated by appropriate barrier systems from the main snake housing
and venom extraction rooms. The main housing rooms for snakes used in venom
production should be designed with security, hygiene and disease control needs
in mind. Separate rooms for accommodation of snake egg incubators and both
neonates and juvenile snakes should be included in the design of the serpentarium.
The cage cleaning rooms should be large enough to hold all the cages
that are being cleaned and sanitized. Dirty cages and other items should be kept
separate from clean cages and equipment being stored ready for use. Furthermore
it is desirable to have two sets of washing and sanitizing rooms, a larger one for
equipment from the venom production room and a smaller one for equipment
from the quarantine area. These rooms should be secure in case a snake is
inadvertently left in its cage when the container is placed in the cleaning room.
The cleaning procedures for production rooms and for cages in which snakes are
kept, and the cleaning schedule, should be established and documented.
Food animals, usually rodents, should be purpose bred in clean
conventional animal houses, and kept, handled and killed in accordance with
ethical principles. The rooms, exclusively used for rodent production, should be
large enough to provide sufficient numbers of rats or mice to feed the snakes.
Ideally, rodent production should be performed in facilities meeting the
corresponding international guidelines. Alternatively, rodents can be purchased
from qualified commercial sources. Breeding of rats and mice cannot take place
in the same room, because of the stress induced by the rats in the mice. The diets
required by young snakes may differ from those of adults (for instance, frogs and
tadpoles are preferred to rodents by some species), and facilities for producing
these food animals may also be required.
When possible, it is useful to have a small laboratory for performing
quality control on the venoms. All serpentariums need to be designed with
separate laboratories where venom can be processed after extraction and quality
control performed (see section 10). An area for repairing broken equipment
and for other miscellaneous purposes is also required. The administrative area
should be sufficiently large and adequately equipped with computer facilities so
that the traceability requirements needed for venom production can be met. The
whole venom production facility should be made secure against unauthorized
intrusion.
240
Annex 5
9.1.4 Snake venom production

The collection of venom is an inherently dangerous task; therefore specific
safety protocols for operators must be applied and rigidly enforced (see section
9.2). All operations should be fully described in written procedures and SOPs,
which should be checked and revised periodically according to a written
master document. Pools of venom require unique batch numbers, and should
be traceable to the individual specimens from which venom was collected for
that batch.
9.1.4.1 Venom collection in serpentariums

Venom can be extracted from snakes according to a regular schedule, depending
on the species. The interval between extractions varies among producers and
ranges from every 2 or 3 weeks to every 3 months. Specimens that are quarantined,
are gravid, are undergoing treatment for sickness or injury, or in the process of
sloughing their skins should not be used for venom production.
Handling equipment must be appropriate for the particular species of
snake to minimize risk of stress, discomfort and injury to both the snake and the
operator. Staff must be familiar with the equipment and properly trained in its
use. Common methods of restraint include gently removing the snake from its
cage with a hook and either placing it on a foam rubber pad before being pinned
behind the head, or encouraging the snake to crawl into a transparent plastic tube
in which it can be restrained. Developing innovative methods that enable safe
restraint of venomous snakes and minimize the risk of injury both to operators
and snakes is strongly recommended. For very dangerous species, the use of
short-acting general anaesthesia or moderate cooling (15 °C) during venom
extraction can be considered (for example, inhaled isoflurane or sevoflurane or
even carbon dioxide) as it reduces the risk of accidents both to the snake and to
the snake handler. Excessive cooling of the snake in a refrigerator is potentially
harmful and is not recommended. For the collection of venom, the snake’s head
is grasped in one hand just behind the angle of the jaw, while the snake’s body is
held with the other hand, or by an assistant snake handler. Individual techniques
for holding the head of the snake vary and each operator should use the method
that works best for them. An assistant should gently occlude the snake’s cloaca to
prevent messy contamination of the locality by spraying of faeces.
Different techniques are used to collect venom. Many rely on encouraging
the snake to open its mouth and either bite through a plastic- or parafilm-covered
membrane, which provides a barrier to contaminants such as saliva and blood
(from minor oral trauma), or to release venom into a container over which the
fangs have been hooked by the operator. In the case of large vipers, the dental
sheath may be retracted when necessary with sterile forceps. Although it is
common practice to squeeze the sides of the snake’s head to try to force venom
241
from the glands, this may cause traumatic bruising to the animal and should be
avoided. The use of brief electrical impulses of moderate intensity to stimulate
venom secretion is not recommended. Any venom sample contaminated with
blood should be centrifuged. After venom extraction, the fangs are carefully
withdrawn from the collection vessel, while preventing damage to the mouth
and dentition and avoiding the snake impaling itself with its own fangs. Then, the
oral cavity should be sprayed with an antiseptic solution to avoid stomatitis. After
each venom extraction, all materials used in the process should be sterilized.
Peptides and proteins in venom are amphiphatic and will adhere to most
common surfaces including glass and plastic (60) resulting in the potential loss
of toxins from the venom used to produce hyperimmune plasma. The use of
polypropylene vessels and the addition of 1% bovine serum albumin can help
reduce such losses, but different peptides may have variable affinity for being
retained on vessel surfaces regardless of the approach taken to minimize loss.
Special procedures that avoid direct handling should be employed in
the case of burrowing asps (genus Atractaspis) because they cannot be held
safely in the way described above (61). For some species with small fangs and
small venom yields, the use of sterile pipette tips or capillary tubes which are
slipped over each fang one at a time, and pressure applied to the base of the
fang to stimulate venom release into the tube, is recommended. In the case of
colubrid snakes, special techniques are required, such as application of foam
rubber pads (from which venom is recovered in the laboratory) or pipette tips/
capillary tubes to the posteriorly placed fangs and the use of secretagogue drugs.
Similarly, some elapid snakes have small fangs and the pipette tip or capillary
tube technique is required to collect venom. At the time of venom extraction,
there is an opportunity to remove broken or diseased fangs and to examine the
snake for ectoparasites (for example, ticks and mites), wounds, dermatitis, areas
of adherent dead skin and retained spectacles over the eyes. The snake can be
treated with drugs and/or vitamins at the same time and, if necessary, can be
force fed. When force fed with rodents, the rodent’s incisors must be cut out
so as not to cause any injury to the snake’s oesophagus. The process of venom
extraction is often combined with cage cleaning and disinfection and the feeding
of the snake. Avoiding trauma to the snake’s mouth and dentition is critical to
prevent infection and “mouth rot” and the venom extraction process should be
performed in accordance with clean practices.
Several snakes from the same group (same species and subspecies
collected at the same time in the same area) can have their venom collected
into the same venom collection vessel. The vessel should be kept in an ice bath
between individual extractions, and the venom aliquoted into labelled storage
tubes or vials and snap-frozen at −20 °C or colder within 1 hour. For venoms
with high proteolytic activity, the collected venom pool should be transferred
into a vial maintained at ultra-low temperature (−70 to −80 °C) or at least
242
Annex 5
−20 °C, every 10–30 minutes, before continuing extractions from that group
of specimens. Another method is to transfer the collected venom into a vial
maintained in an ice bath. Refrigerated centrifugation of freshly collected
venom is recommended, for instance at 1000 g for 5 minutes (4 °C), to remove
cellular debris.
It is important to identify the vial into which the venom has been
collected or transferred for storage, with an appropriate reference number.
Primary indelible identification must be on the vial. This allows the identification
of all the snakes used during venom extraction, the name of the operator and any
other relevant information. To obtain large venom batches for the preparation
of antivenom, especially from species with low yields, one approach is to use
the same vial over several months for extractions performed with the same
specimens, providing the cold chain is never broken. Pools of venom require
unique batch numbers, and the individual venom extractions contributing to the
pool must be traceable. When a pool is sufficient in volume, the venom should
be either freeze- or vacuum-dried and kept in the dark at a low temperature
(either −20 °C or 4 °C) in a well-sealed flask, precisely identified with a number,
up to the time of delivery. Some producers use an alternative system, keeping
dried venom at 20–25 °C in a desiccator. Regardless of the method used, the
procedures for drying venom should be well established, documented, validated
and incorporate appropriate quality control steps (for example, periodic
determination of residual moisture against established standards). The potency
of venom stored for considerable periods of time must be tested at least annually
to ensure that no degradation or loss of activity has occurred (see section 10),
and if a loss of potency is observed the batch must be replaced.
The equipment used for storage of frozen venom (freezers) and for
venom drying, should be cleaned using established procedures, and the cleaning
documented, in order to minimize cross-contamination. Likewise, equipment
requiring calibration, such as freezers, balances and freeze-driers, should be
calibrated according to a defined schedule.
9.1.4.2 Venom collection from wild snakes

The practice of collecting venoms from wild-caught snakes that are subsequently
released in either the same or a different location should be discontinued, and
is not recommended due to the lack of traceability and difficulties in ensuring
effective quality control of venoms. There is also evidence of high levels of
mortality among relocated snake species particularly if they are released at a
distance from the capture site (62–65). In jurisdictions where it is current practice
for collectors to go to designated localities in the wild, catch snakes and collect
venom before releasing them elsewhere, strong efforts must be made to replace
this approach with regulated production using captive snakes maintained in well-
designed serpentariums.
243
9.2 Staff responsible for handling snakes

9.2.1 Safety and health considerations
Handling and extracting venom from snakes is a dangerous operation. One
envenoming occurred every two years in each of the 15 extraction facilities
reviewed by Powell et al. (66). At a commercial venom production plant in
Uberlândia, Brazil between 1981 and 1999, 25 technicians performed 370 768
venom extractions from Bothrops moojeni. Twelve bites were recorded, 10 with
envenoming, and one case of venom squirted into the eye of a worker (67).
Venom extractions should be performed according to well-designed
and documented SOPs by well-trained snake handlers. All personnel involved
in snake handling and venom collection should be fully informed about the
potential dangers of being bitten and envenomed. They should be thoroughly
trained, and the training procedures must be documented and specific protocols
practised as a team. A minimum of two people should be present during snake
handling for venom collection. For safety reasons, it is recommended that venom
extraction sessions should be interrupted at least every 2 hours for a rest period,
before restarting the process.
Personnel involved in snake handling and venom extraction should
observe established hygiene standards (see below) to minimize the impact on
snakes and the potential transfer of pathogens between snakes.
9.2.2 Personal protective equipment (PPE) for snake or venom handling

Protective clothing should include appropriate eye protection (safety glasses
or face shields), face masks, nitrile gloves and a laboratory coat or gown. Eye
protection is especially important when handling spitting elapids capable of
squirting their venom. The wearing of puncture-resistant gloves designed to
prevent an effective bite is unpopular among many keepers who fear that it
impairs manual dexterity and sense of touch, but the use of nitrile gloves is
advisable to prevent cross-contamination. Puncture-resistant gloves should be
mandatory as protective equipment for assistants helping to restrain or handle
snakes during procedures such as venom extraction.
When lyophilized or desiccated venom is being handled, the safety of
operators is paramount, since dried venom can easily be aerosolized and affect
people through skin breaks, eyes or mucous membranes, or by sensitizing them
to the venom (68). Appropriate gowning is necessary when handling dried or
liquid venom, to prevent contact of the venom with skin or mucous membranes.
It is highly recommended that a biological safety cabinet (for example, Class II,
B2), be used while handling lyophilized or desiccated venom.
244
Annex 5
9.2.3 Procedures to be followed if a bite occurs

There are several important measures to be put in place for dealing with a bite
(69), as described below.
9.2.3.1 Procedures and alarms

Clearly defined, prominently displayed, well-understood and regularly rehearsed
procedures should be in place in case of a bite. An alarm should be sounded to
summon help, the snake returned safely to its cage or box and the victim should
withdraw to an area designated for first aid.
9.2.3.2 First aid protocols

Clearly understandable first aid protocols should be established for each species.
These should be available in printed form adjacent to each cage. Immediate
application of pressure-immobilization may be appropriate for treating the bites
of rapidly neurotoxic elapids. However, the technique is not easy and, if they are
to use the method properly, staff will need extensive training and regular practice,
and must be provided with the necessary materials (a number of crepe bandages,
10 cm wide × 4.5 m long, and splints). Analgesia should only be provided for
pain during the pre-hospital period upon the advice of an attending physician.
Provision of appropriate analgesia for first aid should be considered. If venom
enters the eyes, immediate irrigation with generous volumes of clean water is an
urgent necessity.
9.2.3.3 Hospital admission

As a precaution, all victims of bites, scratches by snakes’ fangs or teeth, and those
in whom venom has entered the eye, or anyone else suspected of a snake-bite
or venom exposure injury (for example, through aerosolized dried venom),
should be transferred as quickly as possible to the designated local hospital, by
prearranged transport, for medical assessment. It may be helpful to remove from
the cage, and take to the hospital with the victim, the label identifying the snake
responsible for the bite, so that accurate identification of the snake species and
of the antivenom to be administered is ensured. Staff members should wear,
or carry a card detailing their personal medical information (including drug
allergies) at all times and the card should be taken with them to hospital in the
event of an injury. The contact details of a recognized clinical toxinology expert
should be included on this card.
It is highly recommended that all serpentariums stock in-date supplies of
antivenom appropriate to the species of snakes being held, so that an adequate
245
supply of the correct antivenom can accompany the victim to hospital. Hospital
staff should be warned in advance by telephone of the arrival of the casualty and
informed about the species responsible and any background medical problems
and relevant medical history, such as past reactions to antivenom or other equine
sera (for example, anti-tetanus serum), and known allergies.
9.2.3.4 Snake venom hypersensitivity

Snake venom hypersensitivity is an occupational hazard of snake handlers that
occurs due to sensitization to venom proteins. Two out of 12 snake-bites in a
commercial venom production plant in Brazil resulted in venom anaphylaxis
(67). Hypersensitivity is usually acquired by mucosal contact with aerosolized
dried venom. Important early evidence of evolving sensitization is sneezing,
coughing, wheezing, itching of the eyes or weeping when working around snakes
and snake enclosures, or even upon entering the snake room. No one with
established venom allergy should be permitted to continue working with snakes.
Venom-induced anaphylaxis should be treated with self-injectable adrenaline
(epinephrine) 0.5 ml of 0.1% solution by intramuscular injection (adult dose),
which should be stocked in adequate doses in each room holding snakes, or
where snakes are used for procedures such as venom extraction.
9.2.3.5 Medico-legal and health insurance aspects

The occupational exposure to venomous snake-bites in commercial venom
production units is the responsibility of the employers and requires their
formal attention.

■■ Well-managed serpentariums are a key element in the production
of venom preparations meeting the quality requirements for the
production of effective antivenoms.

■■ The quality of snake venoms used for animal immunization, as
material for both preclinical and batch release assessment of
neutralization efficacy, or for the development of national or
regional reference preparations is of critical importance.
■■ The procedures used in snake maintenance, handling and venom
extraction, as well as in all aspects of venom collection should be
properly documented and scheduled.
■■ Venoms used for antivenom preparations should be representative
of the entire snake population living in the area for which the
polyspecific and/or monospecific antivenoms are intended
to be used. Because of regional and individual variations in
246
Annex 5
venom composition within snake species, the venoms used for

immunization should be collected from a large number of
individuals (generally at least 20, including males and females of
different ages) collected from various regions covering the entire
geographical distribution of the particular venomous snake species.
■■ Venom producers should adhere rigorously to the following
recommendations and should be able to demonstrate their
application:
–– Taxonomic identity and geographical origin of each individual
animal used for venom production should be known and
recorded.
–– Housing, feeding and handling of snakes should meet
the highest veterinary and ethical standards, and follow
documented protocols.
–– Adequate training should be provided to personnel involved
in venom production in all procedures, and implementation of
health and safety measures.
–– Formal guidelines and procedures for emergency response in
the event of any suspected snake-bite or venom exposure should
be established and well documented.
–– Venom should not be collected from animals under quarantine,
or which are gravid, injured, sick or in poor condition.
–– Full traceability of each venom batch should be ensured.
–– Venoms should be frozen as soon as possible after collection,
and at least within 1 hour.
–– Freeze-drying or desiccation of the venoms should be done
under conditions that ensure stability for long-term storage.
–– Batch-to-batch consistency of venoms of the same origin
should be confirmed.
10. Quality control of venoms

10.1 Records and traceability
It is critical to accurately identify the species (and subspecies, if any) of each
individual snake used for venom production and the taxonomic status should be
validated by a competent herpetologist. Increasingly, DNA taxonomy is replacing
conventional morphological methods, but this technique is impracticable in most
venom production units which will continue to rely on well-established physical
features such as colour pattern and scale count to distinguish the principal
medically important species.
247
Internationally recognized scientific names should be used and the

biogeographical origin of each snake should be specified, since differences
in venom composition may occur between different populations of the same
species or subspecies (43–48, 70). Venom producers can consult academic
zoologists who have appropriate skill and experience.
Data pertaining to each numbered venom batch should include
the information considered to be key for traceability, quality and specific
characteristics of the venom (for example, identification of all the snakes used,
the species, subspecies and biogeographical origin, feeding, health care, date of
each extraction, number of specimens used to prepare the batch, and quantity of
venom produced). This information should be made available upon request to
any auditor or control authority.
For long-term storage, venom should be appropriately aliquoted to
minimize wastage and must then be stored in sealed vials until use. Liquid
venoms should be stored frozen at −80 °C, while lyophilized or dried venoms
may be stored at −20 °C. After opening the vial, the venom required should
be used and any surplus product discarded. Unused venom should not be
re‑lyophilized, re‑dried or refrozen (in the case of liquid venom).
10.2 National reference materials

The quality of snake venoms used as a reference standard by quality control
laboratories and NRAs is crucial. WHO recommends that national reference
venom collections be established and that these cover each medically important
snake species used in antivenom production. Such reference venoms should be
prepared as described elsewhere in these Guidelines (see sections 9 and 21),
and must be tested for potency at least annually to ensure they comply with the
original specifications.
Owing to the large variations in venom composition even within a
single species it is recommended that national reference venom collections

should be established, which cover the entire interspecies variability. Regional
reference materials could be used when countries within a region share a similar
distribution of venomous snakes.
Establishing a collection of reference venoms ensures that the
antivenoms produced will be tested against the same relevant venoms in the
specific countries or regions. The characterization and maintenance of reference
venom collections should be performed under oversight from NRAs and other
competent agencies with technical expertise to ensure that reference venoms are
produced to international reference material standards.
Venom batches may be prepared following the procedure outlined in
section 9. Whatever their origin, the snakes used for these reference standards
should be accurately authenticated (species, subspecies) by a qualified person and
248
Annex 5
the place of capture recorded. Genetic samples (for example, tissue and blood)
should be routinely collected from all specimens for DNA analysis if questions
arise regarding the validity of the identification of specimens. Photographs of
individual specimens may also have value.
10.3 Characterization of venom batches

It is the responsibility of the venom producer to provide clear information
pertaining to the species, the subspecies and the geographical origin of the snakes
used for the production of the venoms supplied for antivenom production,
quality control and preclinical studies. This information should be included in
the technical dossier supporting the marketing authorization of any antivenom.
In addition to the certificate which details the scientific name of the snake
species (and subspecies if any), the geographical origin and the number of
animals used for preparing the batch, and the date of collection of the venom as
well as additional biochemical and biological information may be provided for
each venom batch as evidence of consistency.
This information may include analysis of:
■■ biochemical characteristics of the venom;
–– protein concentration per gram,
–– scans or pictures of SDS-PAGE (in reducing and non-reducing
conditions), and
–– size-exclusion or reverse-phase chromatographic profiles (for
example, reverse-phase HPLC);
■■ enzymatic and toxicological activities of the venoms;
–– for example, LD50 and, depending on the particular venom, in
vitro procoagulant activity, proteinase activity or phospholipase
A2 activity;
■■ for lyophilized, vacuum-dried or desiccated venoms, analysis of
residual moisture.
If the venom producer is not able to perform these determinations they
can be subcontracted. Alternatively (depending on the agreement), the antivenom
manufacturer can perform relevant assays to confirm compliance of venoms
with specifications as part of the quality control of the raw material.

■■ Quality control of snake venoms is essential to provide assurance
that the venoms are representative of venomous snakes inhabiting
the region for which the antivenoms are manufactured.
249
■■ National reference venom collections should be established

covering each of the medically important snake species for which
antivenoms are produced.
■■ Traceability of each venom batch is important for rapid detection
of any errors that might occur during the preparation process.
■■ For each venom batch, a certificate stating the scientific names of
the snake species (and subspecies if any), their geographical origin
and the number of animals used in collecting the batch, the date of
collection of the venom, and any other relevant information, must
be provided by the venom supplier to the antivenom manufacturer
and also to the regulatory authority if required.
■■ Consistency (within established limits of composition and quality)
of venom batches produced over time for the same venomous
species of the same origin should be guaranteed. Specific tests
should be performed on each venom sample, and data recorded for
traceability, including: the protein concentration per g (or mg), an
assessment of biochemical and biological activity, scans or pictures
from SDS-PAGE (in reducing and non-reducing conditions), and/
or size-exclusion or reverse-phase HPLC chromatographic profiles
of the venom sample. This information has proved useful to
confirm the origin and the integrity of the venom.
11. Overview of the production process of antivenoms

Antivenoms are obtained following a complex production process (Fig. A5.3),
which involves several steps critical to their effectiveness, quality and safety (71).
These steps are summarized below:
■■ Collection of representative venom pools from correctly identified

individual venomous snakes which have been confirmed to be in
good health. They should be representative of the snake populations
(for example, males/females, adults/juveniles) and region(s) where
the resulting antivenom immunoglobulins are intended to be used.
■■ Preparation of the venom(s) mixture(s) used for the programme of
immunization of animals.
■■ Immunization regimens of animals (most often horses). Animals
should be selected and controlled carefully, and subjected to
continuous health surveillance.
250
Annex 5
■■ Collection of blood or plasma from the immunized animals, once

the immune response to the immunizing venom mixture has yielded
satisfactory antibody levels.
■■ Preparation of the pool of plasma for fractionation.
■■ Fractionation of the plasma to extract and to purify the antivenom
immunoglobulins if applicable.
■■ Formulation of the bulk antivenom immunoglobulins and aseptic
filling.
■■ Quality control tests, including potency assessment by in vivo assay.
■■ Labelling, packaging, boxing and release.
■■ Distribution within the region(s) where the snakes used to prepare
the venoms to immunize the animals are prevalent.
251
Fig. A5.3
General manufacturing process of antivenoms
Veterinary surveillance
Collection of venoms Selection of animals (horses, sheep)

(venom extraction)
S E R P E N TA R I U M
ANIMAL PLASMA PRODUCTION

Quarantine, vaccination
Preparation of venom mixtures and veterinary
Quality control of venom mixtures Inclusion in the herd

Continuous veterinary surveillance
Preparation of immunizing
doses of venoms
Immunization programme
for each animal
Control of animal immune responses
Collection of blood or plasma

Storage and pooling of plasma for

fractionation to isolate immunoglobulins
QUALITY CONTROL
Quality control of plasma

F R AC T I O N AT I O N
for fractionation
Fractionation of plasma
Formulation and filling
Quality control of antivenom
Labelling, packaging, boxing and release
252
Annex 5
12. Selection and veterinary health care of animals

used for production of antivenoms
12.1 Selection and quarantine period
Animals selected for antivenom production should comply with specific
selection criteria relating to breed, size, age, health status and history, and should
preferably be purchased from known accredited suppliers. The use of animals in
the production of hyperimmune plasma should follow strict ethical standards in
accordance with national and international conventions on the use and welfare
of animals. Animals must be transported according to local transport standards.
Before an animal is introduced into the herd used for a production programme,
it should be subjected to a period of quarantine (which, in most countries, lasts
from 6 to 12 weeks), depending upon the source of the animal. During the
quarantine period an appropriate veterinary assessment should be performed to
ensure the animal’s suitability for the programme. The quarantine facility should
be separate from the main animal housing facility or farm and a biosecurity plan
for all animal premises is recommended. Each animal should have an individual
monitoring record system created on its entry into the quarantine facility, which
will remain with the animal throughout its life at the facility or farm. All activities
and information on all aspects of its life, including husbandry, health, antivenom
immunization, bleeding and emergency care must be recorded in this file, which
should be accessible for external review.
When an animal is imported from a country or region with different
ecological characteristics, a period of acclimatization to the local environment
of about 3 months is needed. Each individual animal should be unambiguously
identified using, for example, a microchip, branding or ear clipping.
In the case of horses and other equines, animals aged between 3 and
10 years are usually included in an immunization programme, but in some cases
older animals may also be suitable as long as they exhibit a satisfactory immune
response to the immunization programme. In the case of sheep, animals retired
from wool production have proved capable of useful antibody production for a
number of years (beyond the age of 10 years). No particular breed is preferred, but
in general large horses or sheep are chosen because they yield larger individual
volumes of blood.
12.2 Veterinary care, monitoring and vaccinations

The veterinary examination will include a complete physical examination and
blood tests including serological testing for the most prevalent infectious diseases
for that type of animal in that particular geographical location (for example,
equine infectious anaemia).
253
Depending upon the local epidemiological situation, animals should be

vaccinated against tetanus and, possibly other endemic diseases, such as rabies,
equine influenza, anthrax, brucellosis, glanders, African horse sickness and
equine encephalitides. Animals should go through a treatment programme to
eliminate gut helminths and other locally prevalent parasites. All vaccinations
and health information should be recorded on the animal’s individual record.
Staff in contact with the animals should be vaccinated against tetanus
and rabies.
12.3 Animal health and welfare after inclusion in the herd

After the quarantine period, if the animal is in good health according to a
veterinary examination and blood parameters and body condition score, and the
results of relevant serological tests are negative, the animal may be incorporated
into the herd of animals used for immunization.
An individual record should be kept for each animal being used in an
immunization programme for antivenom production. In addition to surveillance
by a veterinary professional, the staff in charge of the animals should be well
trained, and the operations related to animal care, emergency care and use should
be clearly specified in the standard operating procedure (SOP).
Throughout the time an animal is used for immunization aimed at
antivenom production, careful veterinary monitoring should be maintained,
including continued vaccination regimes, and the performance of regular clinical
examinations, together with clinical laboratory tests such as packed cell volume,
haemogram, clotting tests and other tests associated with the possible clinical
effects of venoms (72) and of successive large-volume blood collection (73).
Possible anaemia, resulting from excessive volume or frequency of bleeding
(when red blood cells are not re-infused into the animals after the whole
blood bleeding session) or from the deleterious action of venoms should also be
tested for.
The immune response against venom components should, when
feasible, be followed throughout the immunization schedule, in order to detect
when animals reach an acceptable antivenom titre. However, the monitoring
of the immune response can be done on a pool of sera from various animals.
This response may be followed by in vivo potency assays of neutralization of
lethality or by in vitro tests, such as enzyme immunoassays (EIAs) (provided
that a correlation has been demonstrated between these tests and the in vivo
potency tests).
Whenever an animal develops any manifestation of sickness, it must be
temporarily withdrawn from the immunization programme to allow it to receive
appropriate veterinary examination and treatment. If the disease is controlled,
the animal may return to the immunization programme after a suitable length
254
Annex 5
of time, usually 4 weeks. If an animal is receiving any type of antibiotic or drug,

it should be withdrawn from the immunization programme for a period that
would depend on the elimination kinetics of the particular drug(s) concerned.
In the case of vaccination, this withdrawal period should not be shorter than
1 month.6 Any blood, plasma or serum obtained from the animal during the
incubation period of any contracted disease should be excluded from use for the
production of antivenoms. Animals should have appropriate physical exercise
and routine husbandry (hoof care, teeth rasping etc.). Their feed should originate
from a controlled source and should be free of ruminant-derived material.
Ideally, the diet should include both hay and grass, or alternative plant material,
and concentrated food preparations containing vitamins including folic acid,
iron and other mineral supplements. The routine quality control of the food
and water is recommended, in order to assure a consistent composition and an
adequate level of nutrients.
As a consequence of immunization with venoms (see section 13) a
common problem in antivenom-producing animals is the development of local
ulcers or abscesses (sterile and infected) at sites of venom injection. This is a
particular problem when necrotic venoms and FCA are used. All injections
should be given under aseptic conditions and administered subcutaneously.
There should be a limit to the total volume and dose of venom injected at a single
site. Infected or ulcerated areas should be treated appropriately. Abscesses should
be lanced and drained and the affected skin site should not be used again. In the
event of the death of an animal being used for antivenom production, a careful
analysis of the causes of death should be performed, including, when necessary,
the performance of a necropsy and histopathology. All deaths should be recorded
together with the necropsy report and made available for external review.
Some animals show declining titres of specific venom antibodies over
time, despite rest or increasing doses of immunizing venoms. Such animals
should be retired from the immunization programme. In agreement with the
principles of GMP and to avoid impact on the composition and consistency
of the antivenom produced, it is, in principle, not considered good practice to
move animals from a given venom immunization programme to another one.
If, however, the animal has been used in the preparation of a monospecific
antivenom that is included in a polyspecific preparation, or if it was used for
the production of other animal-derived antisera (for example, anti-rabies, anti-
tetanus or anti-botulism), moving it to another programme may be acceptable.
When an animal is withdrawn from the herd, it may be kept on the horse
farm. If it is sold, continued good care should be ensured.
In some areas, legislation stipulates that animals used for production of plasma cannot be treated with
6
penicillin or streptomycin.
255

■■ A thorough biosecurity plan should be developed and implemented
for each farm and facility.
■■ All staff working with the animals should be trained and qualified
to care for them. Staff training records and history should be
available for review.
■■ An emergency care protocol is essential especially during
procedures – for example, during sensitization to venoms, blood
collection and post-plasmapheresis. Adverse events must be
reported and tracked appropriately.
■■ Animals intended for antivenom production programmes should
be identified to ensure full traceability and health monitoring.
■■ Animals should go through a quarantine period of 6–12 weeks
during which they are submitted to veterinary scrutiny and are
vaccinated against specific diseases and treated for internal and
external parasites.
■■ Following the quarantine period, animals may be introduced into
the immunization programme. Animals should be appropriately
housed, fed and managed according to best practice in veterinary,
animal welfare and ethical standards.
■■ During immunization, the clinical status of each animal must
be followed by a veterinarian through clinical and laboratory
assessments which are recorded on the animals’ records. If an
animal develops clinical signs of disease, it should be temporarily
separated from the immunization programme to receive
appropriate care and treatment. Particular care must be paid to
local lesions that develop at the site of venom injections and to

development of anaemia.
■■ The immune response of each animal to venoms should, when
possible, be monitored during the immunization schedule
(alternatively, the antivenom titres can be monitored indirectly by
testing the plasma pool).
■■ An animal receiving an antibiotic or drug should be withdrawn
from the immunization programme for a period depending on
the elimination kinetics of the drug concerned. In the case of
vaccination, this withdrawal period should not be shorter than
1 month.
256
Annex 5
13. Immunization regimens and use of adjuvant

One of the most crucial steps in antivenom production involves the immunization
of animals with venom(s) to produce a long-lasting and high-titre antibody
response against the lethal and other deleterious effects of the immunogenic
toxins. To achieve this goal, the following considerations are important:
■■ Venom(s) used should be prepared as described in section 9,
and should be in an optimal condition for inducing specific and
neutralizing antibodies.
■■ Immunogen and the immunization regimens used should not
seriously affect the health of the animal.
■■ Preparation of immunogens and the immunization protocol should
be technically simple and economical and use a minimal amount of
venom. The procedures followed must be included in a protocol and
their performance must be documented.
The antivenom manufacturer is responsible for defining the appropriate
immunization programme (choice of doses, selection of adjuvants, sites of
immunization, and bleeding schedule) able to generate the best immune response
and plasma production, while also ensuring optimal animal care. GMP principles
should be applied in the preparation of the immunizing doses as well as in the
immunization process.
13.1 Animals used in antivenom production

Numerous animal species have been used on various scales in antivenom
production (horse, sheep, donkey, goat and rabbit) or for experimental purposes
(camel, llama, dog and hen) (74, 75). However, the production of large volumes
of antivenom from large animals such as equines is an advantage compared to the
smaller species. The selection of the animal species should be based on several
considerations, such as locally prevalent diseases, availability in the region,
adaptation to the local environment, and cost of maintenance. The information
in these Guidelines refers mostly to horse-derived immunoglobulins.
The horse is the animal of choice for commercial antivenom production.
Horses are docile, thrive in most climates and yield a large volume of plasma.
Antivenoms made from horse plasma have proven over time to have a
satisfactory safety and efficacy profile (3). Sheep have also been used as an
alternative source for antivenom production because they are cheaper, easier to
raise, can better tolerate oil-based adjuvant than horses, and their antibodies
may be useful in patients who are hypersensitive to equine proteins (75, 76).
However, increasing concern about prion diseases may limit the use of sheep
257
for commercial antivenom production. Larger animals are preferable to smaller

ones because of their greater blood volume, but breed and age are less important.
Any animals used should be under veterinary supervision (see section 12).
When sheep or goats are to be used, manufacturers should comply with
regulations to minimize the risk of transmissible spongiform encephalopathies
(TSEs) to humans, such as the WHO Guidelines on tissue infectivity distribution
in transmissible spongiform encephalopathies (77).
13.2 Venoms used for immunization

Venoms used as immunogens in antivenom production are chosen based on
criteria discussed in section 9. Priority should be given to venoms from snakes
responsible for frequent and severe envenomings. The quality, quantity and
biological variation of venoms are important considerations (see sections 9
and 10).
13.3 Preparation of venom doses

Venom doses used for the immunization of animals should be prepared carefully
in a clean environment, maintained according to an established, scheduled and
documented cleaning regime. All venom manipulations should be performed
using aseptic techniques under a hood; for highly toxic venoms, a cytotoxic
cabinet may be used. Batch process records should be completed for each dose
preparation session. The venom batches used and the animals to be immunized
should be recorded and the containers in which the venom is dissolved should
be appropriately identified. Ideally, the calculations and operations related
to the dose of venom to be used, as well as dilutions, require verification by a
second person to ensure accuracy and to prevent errors that may lead to animals
receiving overdoses.
Venoms, when freeze-dried, are highly hygroscopic and allergenic, thus
care should be taken when manipulating them. When taken out of the refrigerator
or freezer, the venom should be allowed to warm up to room temperature before
the bottle is opened. Otherwise, condensation may occur causing inaccuracy in
weighing and, more seriously, proteolytic degradation of the venom proteins by
venom enzymes. Venom should be dissolved in distilled water or buffer, but care
should be taken not to shake the solution too vigorously since excessive foaming
may cause protein denaturation.
The solvents used to dissolve venoms should be sterile and should be
used before the established expiry dates. A stock solution of each venom should
be prepared separately, rather than being mixed with other venoms. This is to
allow flexibility of dosage and to avoid proteolytic degradation by one venom
component of other venom proteins. Venom solutions can be sterile filtered
(where this is known not to affect the potency of the preparation), aliquoted
258
Annex 5
and labelled. It is recommended that venoms used for immunization be freshly

prepared at the time of use. The storage conditions of venom solutions before
immunization must be validated by properly conducted stability tests.
All the equipment used for venom storage (freezers and refrigerators)
and preparation (for example, balances) should be calibrated and validated for
their intended purpose. Balances should be calibrated at least annually and
calibration should be checked daily. Where possible, laboratory items used in
venom preparation, that is, pipettes, syringes and other such items should be
pre-sterilized, single-use, disposable items. The siliconization of venom solution
containers may be considered to avoid the adherence of venom components
to the surfaces of containers. Venom solutions/suspensions must be safely
transported to the facilities where animals are going to be injected and the
venom solutions/suspensions should be kept cold at about 2–8 °C.
Care should be taken to avoid accidents that may result in envenoming
of the personnel preparing the venom solutions. Protective equipment (for
example, eyewear, gloves and gowns) should be worn. Procedures for cleaning up
broken glass or plastic containers that have held venom should be established
and the personnel should be trained to follow them.
13.4 Detoxification of venom

Some snake venoms can cause local and/or systemic toxicity when injected into
naive horses at the beginning of an immunization course. Various physical or
chemical means have been adopted to decrease venom toxicity, for example,
treatment with aldehydes (formaldehyde or glutaraldehyde), hypochlorite,
ultraviolet or gamma radiation, and heat, among others. However, in most cases,
not only the toxic sites, but also the antigenic sites of the toxins are destroyed
after these treatments (78). For example, when glutaraldehyde is used, the protein
polymerization is often extensive and is difficult to control and reproduce. Thus,
although the detoxified toxin (toxoid or venoid) induces vigorous antibody
response, the antibodies usually fail to neutralize the native toxin. In fact, no
detoxification is usually necessary if inoculation is made with a small dose of
venom that is well emulsified in an adjuvant such as FCA or FIA. Furthermore,
traces of chemicals, especially aldehyde, can have deleterious effects on animals’
vital organs, for example, the liver.
13.5 Immunological adjuvants

Various types of immunological adjuvants have been tested, for example, FCA
and FIA, aluminium salts (hydroxide and phosphate), bentonite and liposomes
(79). The choice of adjuvant is determined by its effectiveness, side-effects, ease
of preparation (especially on a large scale) and cost. It may vary depending
upon the type of venoms and following manufacturers’ experience. FIA contains
259
mineral oil and an emulsifier. FCA, which contains mineral oil, an emulsifier
and inactivated Mycobacterium tuberculosis, has been shown in experimental
animals to be one of the most potent adjuvants known. However, horses are quite
sensitive to FCA which tends to cause granuloma formation. For this reason,
some producers prefer to use other adjuvants. It is recommended that when
using FCA and FIA, they be utilized only at the beginning of the immunization
schedule, and not during the rest of the immunization, nor during booster
injections of venom; this significantly reduces the formation of granulomas
in horses.
It has been noted that the granuloma caused by FCA is caused by the
injection of a large volume (5–10 mL) of the emulsified immunogen at one or
two sites. The large granuloma formed usually ruptures, resulting in a large
infected wound. If the emulsified immunogen is injected subcutaneously in small
volumes (50–200 µL/site) at multiple sites, granuloma formation may be avoided.
Manufacturers are also encouraged to adopt an innovative approach with regard
to adjuvants used for antivenom production, and should strive to replace FCA
and FIA with new compounds of low toxicity and high adjuvant effect. The
advances in the vaccine field concerning new adjuvants should be transferred to
the antivenom field; for example, the use of microbial-derived products of low
toxicity or of Toll-like receptor 4 (TLR4) ligand-based adjuvants (80).
13.6 Preparation of immunogen in adjuvants

To minimize infection at the immunization sites, all procedures should be
carried out under aseptic conditions. Venom solutions are prepared in distilled
water or phosphate-buffered saline solution and filtered through a 0.22 µm
membrane. The venom solution is then mixed and/or emulsified with adjuvant,
according to the instructions of the supplier. An example for the preparation of
venom immunogen in FCA, FIA and aluminium salts is described in Box A5.1.
To facilitate the injections, the immunogen suspension is filled in tuberculin

syringes (Fig. A5.4).
13.7 Immunization of animals

The areas to be immunized should be thoroughly scrubbed with a disinfectant,
shaved and rubbed with 70% ethanol before venom immunogen injection.
In general, the sites of immunization (Fig. A5.5) should be in areas close
to major lymph nodes, preferably on the animal’s neck and back. The route of
injection should be subcutaneous so as to recruit a large number of antigen-
presenting cells resulting in a high antibody response. Some procedures call
for a small volume of injection at each site (50–200 µL) so that the total surface
area of the immunogen droplets is maximized, enhancing the interaction with
the antigen-presenting cells and the immune response (81, 82). An example
260
Annex 5
of immunization of a horse using venom emulsified in FCA is described in

Box A5.2.
Other immunization protocols, using larger amounts of venoms devoid
of local-tissue-damaging activity (such as those of some elapids) and/or adjuvants
other than FCA may be used with satisfactory results, as long as the schedule does
not compromise the health of the animals. In situations where the main toxins
of a given venom have a low molecular mass and would not induce a sufficient
immune response if injected together with the other venom components,
isolating such toxins using mild chromatographic procedures or ultrafiltration
can be useful. Such isolated fractions can then be used for immunization.
Fig. A5.4
Tuberculin syringes are filled with immunogen suspension and used for the
subcutaneous injection of the horse
Box A5.1
Example of preparation of venom immunogen in FCA, FIA and aluminium salts
Since FCA can cause severe irritation, precautions should be taken to avoid contact with
the eyes, and protective eyewear and gloves are recommended. The vial containing
FCA is shaken to disperse the insoluble Mycobacterium tuberculosis. The venom
solution is mixed in a stainless steel container with an equal volume of FCA at 4 ºC.
261
Box A5.1 continued

The emulsification is achieved by vigorous blending in a high-speed blender at a
speed of approximately 3000 rpm for 15 minutes. The container is put in ice-water to
dissipate the heat generated. The resultant emulsion should be quite thick and remain
stable when dropped on the surface of cold water. The highly viscous emulsion is then
transferred into a sterile 50 ml syringe with the plunger removed. The plunger is then
put into the syringe to expel any air pocket inside. By means of a three-way stopcock,
the emulsion from the 50 ml syringe is then transferred into tuberculin syringes to
give a volume of 0.1–0.2 ml/syringe. After the tuberculin syringe is fitted with a 38 mm
no. 21 gauge disposable needle, the needle cover with its end cut off is attached so that
only 2–3 mm of the needle tip is exposed and penetrates the horse’s skin (Fig. A5.5).
With each filled tuberculin syringe, immunization at a particular site can be performed
by injection and expulsion of the immunogen almost simultaneously in one single
step. This immunization procedure makes multiple subcutaneous injections with small
immunogen volume easier, faster and requires minimal restraint of the horse.
Immunogen in FIA is prepared by a process similar to that described above except
that FIA is used in place of FCA. Both the FCA and FIA emulsified immunogens may, if
necessary, be stored at 4 ºC, preferably for a maximum of 2 weeks, but re-emulsification
is necessary before their injection. When the immunogen is prepared in aluminium
hydroxide (Al(OH)3 ) or aluminium phosphate (Al(PO)4 ), a sterile venom solution and a
suspension of aluminium salts are mixed in a ratio of 1:3 (v/v) and homogenized. When
using other adjuvants, the preparation of the solution or emulsion should follow the
manufacturer’s instructions for that type of adjuvant.
Fig. A5.5
Recommended areas of immunization in horses
262
Annex 5
Box A5.2
Example of immunization of horses using FCA, FIA and aluminium salts
The primary immunization could be done with venom(s) mixed with FCA as described
in Box A5.1. The initial dose of each venom could be as low as 1–4 mg/horse with a
total combined volume of injection of about 2 ml. The immunogen is filled in several
1 ml tuberculin syringes with 21G needles as described in Box A5.1 and Fig. A4.5.
Subcutaneous injections of 100–200 µL of immunogen are made at each site, up to
as many as 8–12 sites, although some producers may use only 3–4 injection sites.
The neck of the horse, supplied with extensive lymphatic vessels and large lymph
nodes, is a preferred area for immunization. If inoculation is made on the lateral
sides of the neck, the animal tends to rub itself causing skin blisters. Thus, injections
should be made to the upper (dorsal) part of the neck, close to the mane. About 4–6
injections can be made at each side of the neck. If injection in the rump is possible,
1–2 injections can be made in the area between the outer hip bone and the top of
the thigh bone. The scratching of injected sites by animals can be partially alleviated
by massaging the injection site after venom injection to disperse the dose material.
Immunization using FCA is usually done only once; in most cases, repeated use of this
adjuvant can cause serious reactions which affect the horse’s health. After 2 weeks,
the horses should receive a booster injection with the same venom(s) well emulsified
in FIA. Similar volumes and areas of injection to those described above can be used.
Subsequent booster immunizations at 2-week intervals can be administered, with
higher doses (5–10 mg) of venom(s) in saline or mixed with aluminium salts or any other
suitable adjuvant. In this case, subcutaneous injections of 1 ml of immunogen at each
site in a total of 4 sites are recommended. Blood (10–20 ml) should be drawn before
each immunization. Serum or plasma is prepared and EIA titres and/or lethality potency
are determined. When the EIA titres reach a plateau, usually about 8–10 weeks after the
primary immunization, an in vivo potency assay may be performed to confirm that the
horse could be bled. After bleeding for antivenom production, the horses are allowed
4–8 weeks rest, depending on their physical condition. After the rest period, a new round
of immunization can be performed as described above, but without the use of FCA.
13.8 Traceability of the immunization process

The traceability of the immunization process is critical for the quality control of
the antivenoms produced and the steps to ensure traceability should be performed
very accurately. Each immunized animal should be identified by its code number
(see section 12) and the details of each immunization step should be recorded
precisely. The details to be recorded include:
■■ date of immunization;
■■ batch(es) of venom(s) used with its (their) reference number(s)
(see section 10);
■■ venom dose(s);
263
■■ adjuvant and/or salt used;

■■ names of the veterinary and supporting staff in charge of the
immunization;
■■ occurrence of reaction and/or sickness.
The antivenom titre of the immunized animals should be monitored
throughout the immunization procedure either in vitro, using EIA, during the
immunization phase, or in vivo, by neutralization potency assays of lethality
when the immunization plateau is reached or before each blood collection.
Each plasma batch should be assigned a unique reference number (for
example, a barcode), which should allow complete traceability to the donor
animal. Information (such as the date of collection, the unique identification
number of the immunized donor animal, and the reference number of the
venom(s) used for immunization) should be recorded to allow traceability to all
venoms. Computer-based databases are very useful for properly recording these
data, which are crucial for the traceability of the antivenoms produced. Standard
procedures should be used to protect the integrity of data stored on a computer,
including regular, frequent backup, protection against unauthorized access and
storage of backup copies securely off-site.

■■ Venom solutions should be prepared in such a way as to minimize
proteolytic digestion and denaturation of the venom proteins.
Venom solution should be prepared under aseptic conditions to
avoid infection at the injection sites.
■■ The type of adjuvant used is selected on the basis of its effectiveness,
side-effects, ease of preparation and cost.
■■ Primary immunization should be done by subcutaneous injections

of small volumes at multiple sites close to the animal’s lymphatic
system to favour the recruitment of antigen-presenting cells and
involvement of anatomically different groups of lymph nodes for
antibody production.
■■ Subsequent booster injections can be made using venom
immunogen doses, at volumes and intervals depending on the type
of adjuvant used, until the antivenom titre reaches a plateau or a
pre-established minimum accepted titre.
■■ After collection of blood for antivenom production, animals should
have a resting period of 4–8 weeks. After this, a new round of
immunization can be performed as above without the use of FCA.
264
Annex 5
■■ All steps in the immunization of the donor animal, as well as the

collection of blood or plasma, should be traceable.
14. Collection and control of animal

plasma for fractionation
Historically, serum separated from the blood of hyperimmunized horses was the
basis of “antivenin serum therapy”, but today plasma is used, almost exclusively,
as the starting material and undergoes a fractionation process for the separation
of purified antivenoms. Thus “antivenom immunoglobulins” is the preferred
term, rather than “anti-snake-bite serum” or “antiserum”, which are imprecise
and confusing terms that refer to a crude therapeutic preparation.
Plasma as a starting material is preferred to serum largely because
red blood cells can be returned to the animal, thus preventing anaemia and
hypovolaemia in the donor animal and allowing more frequent bleeding. Some
laboratories have found that using plasma enables higher recovery of antibodies
per donation and it is less contaminated with haemoglobin (Hb) than serum.
Separation of plasma from anticoagulated blood is much faster than separation
of serum from clotted blood. Plasma for fractionation can be obtained either
from the collection of whole blood or by the apheresis procedure.
14.1 Health control of the animal prior to

and during bleeding sessions
When an immunized animal has developed an antivenom antibody titre that
meets the necessary specifications it can be bled. The animal should be in a
satisfactory clinical condition and blood parameters and biochemistry need to
be within the normal range for the animal type and breed. Before bleeding is
performed, the animals should be evaluated by a veterinarian or other qualified
person and declared healthy. Individual blood chemistry parameters – packed
cell volume (PCV); Hb and total plasma protein (TPP) – must be within
specified parameters. Animals showing evidence of clinical deterioration, such
as weight loss, altered horse body condition score, a drop in Hb or serum protein
concentration below a critical predefined value for the animal type and breed,
or evidence of infection, should not be bled. It is recommended that animals
to be bled have no contact with potentially infectious animals. Human beings
can be a potential source of fomite infection in horses and therefore a biosecurity
plan is essential.
265
14.2 Premises for blood or plasma collection

The bleeding of animals should be performed in designated rooms or areas
dedicated to this activity and equipped with appropriate restraining devices. Some
producers may design the bleeding rooms so that they can be closed, if needed,
during the bleeding sessions, but this is not general practice. The rooms or areas
should be thoroughly washed and cleaned before and after each bleeding session
and their design should facilitate such cleaning procedures, which should be
clearly established. The room or area should be inspected before the confinement
of the animal. Animals need to be made as safe and comfortable as possible, in
a quiet environment, during bleeding to minimize the chance of injury to the
animal or its handlers. Individual animals should be confined in circumstances
that reduce the potential for stress as much as possible. It is recommended that
these rooms allow the simultaneous bleeding of several animals to reduce the
time required for this operation as well as the stress.
To avoid bacterial and fungal contamination, animals should be
cleaned and injection sites and jugular catheter sites clipped in a separate room
before bleeding. Humidity control of the surrounding bleeding area should
be ensured.
14.3 Blood or plasma collection session

Animals are bled by venepuncture from the external jugular vein. The area
surrounding the venepuncture site should be clipped before bleeding and
thoroughly cleaned and disinfected, using a disinfectant that has not reached the
end of its recommended shelf-life, and, depending on the type of disinfectant, it
should be allowed to dry. The disinfected area should not be touched or palpated
before the needle has been inserted.
Before venepuncture, all containers and tubing should be inspected for
defects (for example, abnormal moisture or discoloration as these may suggest
a defect). There should be means to determine the volume of blood or plasma

collected (such as a weighing machine).
When using disposable plastic bags to collect blood (which may
take about one hour) it is recommended that the blood should be gently and
continuously mixed to ensure a homogeneous distribution of the anticoagulant
and avoid formation of clots.
Horses should be weighed before bleeding, if possible, and their weight
recorded on their record. The clinical condition of the animals being bled
should be closely monitored at the time of bleeding and during the days that
follow, and bleeding should be suspended in the event of any adverse effect
on the animal. If an animal shows signs of distress during the operation, the
collection procedure should be terminated. In addition, animals should be kept
under observation for at least 1 hour after the bleeding to allow any evidence of
266
Annex 5
physical alterations to be detected. Horses can be fed during blood collection

depending on the horse crush set-up.
After bleeding for antivenom production and depending on whether red
blood cells are re-transfused, the horses should not be re-bled for 4–8 weeks.
Horses should have moderate (5 out of 9) body condition score and normal
range haemogram (PCV, TPP and Hb) and be normal on clinical examination
by the antivenom production veterinarian.
14.4 Labelling and identification

The identity of the animal should be recorded immediately before venepuncture.
Labels on all bottles or bags of blood or plasma should be marked with the
animal’s unique identification number. The label should be waterproof and heat
resistant, and contain the following information: specificity of antivenom, plasma
unit number and date of collection.
A document to register all steps of the production of the plasma lot
should be maintained to guarantee traceability of the process.
14.4.1 Collection and storage of whole blood

14.4.1.1 Collection
The volume of blood to be obtained depends on the species and size of the
immunized animal. It is recommended that approximately 13–15 mL of blood
per kilogram body weight are collected in one bleeding session, or 1.5 to 2%
of the weight of the animal. For sheep, 0.5 L is a typical yield, whereas for
horses, the volume of blood may range between 3 and 6 L, depending on the
size of the animal. The use of automated plasmapheresis may enable larger
volumes of plasma to be collected, and has benefits for animal health and plasma
quality. Manufacturers are encouraged to evaluate and implement automated
plasmapheresis subject to approval from local regulators and in accordance with
local regulations and standards.
Blood is collected, ideally, in disposable plastic bags containing sterile
citrate anticoagulant or other preparations containing citrate phosphate dextrose
solution (CPD), to prolong the durability of red blood cells. Usually, the volume
ratio of anticoagulant to blood is 1:9 to 1:15, depending on the anticoagulant.
Use of double plastic bags containing anticoagulant is recommended to avoid
bacterial contamination and for ease of use. When plastic bags are not available,
disposable polypropylene plastic bottles, or sterilized glass bottles containing
anticoagulant may be considered.
While the bleeding is taking place, a constant flow of blood should be
ensured. Blood should be gently and continuously mixed with the anticoagulant
solution to ensure a homogeneous distribution of the anticoagulant, to avoid the
risks of activation of the coagulation cascade and, therefore, avoid the formation
267
of clots. The duration of a bleeding session per animal is usually between 30

and 45 minutes depending upon the weight of the animal and the total volume
collected. Care should be taken to avoid contamination of the blood by exposing
the needle to contaminated surfaces. It is recommended to seal or occlude the
device before removing the needle from the animal.
14.4.1.2 Storage
The bags or bottles in which the whole blood has been collected should be
appropriately cleaned and sanitized on their external surfaces. They should
be put into a refrigerated room (2–8 °C) for the plasma and blood cell separation
procedure. They should be stored for no more than 24 hours before the
reinfusion of the red cells, unless CPD is used. In this case, blood cells may be
stored for up to 72 hours. Alternatively, aseptically collected blood can be stored
for a maximum of 7 hours at 20–25 °C to allow for sedimentation. Under such
circumstances, great care should be taken to avoid bacterial contamination.
14.4.1.3 Separation of plasma from whole blood

Hyperimmune plasma should be separated from blood cells under aseptic
conditions and should be transferred into sterile containers (plastic bags, bottles
or stainless steel containers). A designated room, designed to allow proper
cleaning and sanitization, should be used for separation. When bottles are used,
separation of plasma from blood cells should be performed in a laminar flow
cabinet located in a room separated from the plasma fractionation area.
14.4.1.4 Reinfusion of the red blood cells

Reinfusion of the red blood cells after whole blood collection is recommended.
Blood cells, most specifically red blood cells, should be separated from plasma by
validated centrifugation or sedimentation procedures. Red blood cell reinfusion
should take place within 24 hours after blood collection (or 72 hours if CPD
anticoagulant is used), and after being suspended in sterile saline solution at
room temperature for 1 hour (or 32–37 °C for less than 1 hour) prior to infusion.
This procedure in which whole blood is collected and red blood cells are
re‑infused into the animal is commonly referred to as “manual apheresis”.
14.4.2 Plasma collection by automatic apheresis and storage

14.4.2.1 Plasma collection
In some laboratories, plasmapheresis machines are used to perform automatic
plasma collection. This has proved a useful investment in some facilities; it
ensures that the donor animal does not become hypovolaemic, increases
plasma yield and purity, and reduces the risks of handling errors, in particular
268
Annex 5
during reinfusion of the red blood cells to the donor. Plasma from automatic
apheresis tends to be less contaminated by blood cells (red blood cells, leukocytes
and platelets). In the experience of some laboratories the plasma is easier to
fractionate, as the filtration steps, in particular, are more readily performed,
resulting in higher yields.
In such procedures, whole blood is collected from the animal, mixed
with anticoagulant, and passed through an automated cell separator. The
plasma is separated from the cellular components of the blood, which are
returned to the animal in a series of collection/separation and return cycles.
The plasma is separated from the red blood cells by centrifugation or filtration,
or a combination of the two. The operational parameters of the plasmapheresis
equipment are provided by the manufacturers of the equipment. In general, the
anticoagulant is delivered at a rate yielding a specified ratio of anticoagulant to
blood. The anticoagulant solutions used include AB16 (35.6 g sodium citrate,
12.6 g citric acid monohydrate, 51.0 g glucose monohydrate per litre using water
for injection) and anticoagulant citrate dextrose formula A (ACDA) (22.0 g
sodium citrate, 8.0 g citric acid, 24.5 g dextrose monohydrate, per litre using
water for injection). The number of collection/separation and return cycles
for each donor animal depends on the total volume of plasma that is to be
harvested. For horses, the average volume of plasma collected may be about 6
litres per session. The number of cycles ranges from 10 to 20 depending upon
the haematocrit of the horses. The collection process lasts for 1–4 hours. The
apheresis equipment and apheresis procedures should be validated, maintained
and serviced. Machine plasmapheresis can take several hours and animals can
be fed during the operation.
14.4.2.2 Plasma storage

Bags or bottles containing apheresis plasma should be stored in a refrigerated
room (2–8 °C) in the dark until the fractionation process starts. Individual or
pooled plasma should be stored at 2–8 °C in a cold room dedicated for this
purpose. This refrigerated storage room should be designed to allow proper
cleaning and sanitization. To prevent microbial contamination of plasma,
preservatives (phenol or cresols 7 ) can be added at a dose of less than 3 g/L at
this stage and kept during storage of plasma. Care should be taken to dilute the
phenol or cresols with water or saline solution before they are added to plasma
with gentle stirring, to avoid denaturation of plasma proteins. The transportation
of containers or bottles containing pooled plasma within the production facility
or between facilities should be performed in such a way that contamination is
In these Guidelines cresol isomers are referred to as “cresols”.

7
269
avoided and the cold chain is maintained. To avoid the risk of contamination,
it is recommended that individual or pooled plasma is not stored for too long
before fractionation, that is, the plasma should be fractionated as soon as possible
after pooling.
If plasma is stored for prolonged periods, the storage time and conditions
should be validated to ensure that there is no detrimental impact on the quality
of the plasma material, on the fractionation process, or on the quality, efficacy
and stability of the antivenoms.
Manufacturers of human plasma have found that plasma can be stored
frozen at −20 °C or colder for 2 years without addition of a preservative, and
with no observed detrimental effects on the fractionated plasma products.
14.5 Pooling
Plasma from individual animals should be pooled into sterile and sanitized
containers before fractionation. For traceability purposes each plasma pool
should be identified with a unique number. The number of plasma units collected
from individual animals and used in the pool should be recorded. Before the
large pool of plasma is prepared, it is recommended to prepare a small-volume
pool and to test it for microbial contamination. If there is no contamination,
the large pool can be prepared. If microbial contamination is detected, the
plasma from the individual animals should be checked, and the contaminated
plasma should be discarded to ensure that the pool is prepared with plasma free
of microbial contamination. It may also be advisable to test small pools using
a cytotoxicity assay, which can reveal the presence of unanticipated viruses
or toxins (for example, following the US Code of Federal Regulations, 9 CFR
113.53 “Requirements for ingredients of animal origin used for production
of biologics”).8
Such pooling should be performed in an environment suitable to prevent
microbial contamination, for example, classified areas (class D (83)) and pools
should be adequately identified. The room should be designed to allow for
appropriate cleaning and sanitization of all surfaces. Individual or pooled plasma
should be stored at 2–8 °C in a room dedicated for this purpose. To ensure the
prevention of microbial contamination of plasma, follow the recommendation
given in section 14.4.2.2.
https://www.gpo.gov/fdsys/search/pagedetails.action?packageId=CFR-2005-title9-vol1&granuleId=CFR-
8
2005-title9-vol1-sec113-53&bread=true&collectionCode=CFR&browsePath=Title+9%2FChapter+I%2F
Subchapter+E%2FPart+113%2FSubjgrp%2FSection+113.53&collapse=true&fromBrowse=true, accessed
24 April 2017.
270
Annex 5
14.6 Control of plasma prior to fractionation

Before fractionation, pools of plasma should be checked for macroscopically
evident precipitates, gross haemolysis and bacterial contamination (bioburden
assay). The neutralizing potency of the starting plasma should be ensured so
that the resulting antivenoms will be within potency specifications. Additional
checks may include, when relevant, a test for pyrogenic substances and total
protein content.
Plasma pools should be discarded if the bioburden exceeds a defined
limit stipulated in the marketing dossier or if the neutralizing potency is below a
minimum limit established by the producer. Cloudy plasma, below this defined
bioburden limit, may still be used for fractionation provided the fractionation
process and product quality has been proven not to be impaired. Grossly
haemolysed plasma should not be used for fractionation. Cloudy plasma may
also reflect the increased level of chylomicrons in the animal plasma; therefore it
is recommended to fast the animal before bleeding for a few hours (for example,
8 hours).

■■ When animals have developed an adequate immune response
against venoms, and if they are in good health, they can be bled for
antivenom production. Bleeding should be performed in enclosed
rooms which should be kept scrupulously clean. Traceability of the
donations should be ensured.
■■ Plasma is preferred to serum as a source material. Animals should
be bled from the external jugular vein. Plasma can be obtained
either from whole blood or by automated plasmapheresis and
using approved anticoagulants. Blood or plasma should ideally be
collected into closed plastic bags. When this is not possible, glass
or plastic bottles can be used, if they can be readily cleaned and
sterilized.
■■ Plasmapheresis is recommended using either automatic or manual
procedures. When manual apheresis is used, blood cells should
be sedimented, separated from the plasma, re-suspended in
saline solution and returned to the animals within 24 to 72 hours.
Plasma separation should be performed in a designated room with
a controlled environment.
■■ Plasma containers should be thoroughly cleaned on their external
surfaces, adequately identified and stored in refrigerated rooms
until further fractionation.
271
■■ Plasma should be checked prior to fractionation to establish

compliance with relevant acceptance criteria for fractionation,
in particular the neutralizing potency and lack of bacterial
contamination.
■■ Special attention should be paid to ensuring traceability between
individual animal donors and the plasma pool.
■■ A certificate from a veterinarian or other qualified person should
be issued stating that the donor animals were checked periodically
to ensure that they were in good health at the time of plasma
collection and during the follow-up observation period.
15. Purification of immunoglobulins and immunoglobulin

fragments in the production of antivenoms
15.1 Good manufacturing practices
The purification of immunoglobulins and immunoglobulin fragments for the
production of antivenoms should aim at obtaining products of consistent
quality, safety and clinical effectiveness. The fractionation processes used should
adhere to the GMP principles developed for medicinal products. All operations
should therefore be carried out in accordance with an appropriate system
of quality assurance and GMP. This covers all stages leading to the finished
antivenoms, including the production of water, the production of plasma (animal
selection and health control, production of venoms and immunization protocols,
containers used for blood and plasma collection, anticoagulant solutions and
quality control methods) and the purification, storage, transport, processing,
quality control and delivery of the finished product. Of particular relevance is the
control of microbiological risks, contamination with particulates and pyrogens,
and the existence of a documentation system that ensures the traceability of

all production steps. To establish satisfactory traceability of the antivenom
produced, all the steps of the purification procedure used for the preparation
of the antivenom batch should be recorded carefully in pre-established and
approved batch record documents, and sampling should be done at established
critical steps for in-process quality control tests.
WHO Guidelines on good manufacturing practices for medicinal
products are available (83) and the main principles of GMP for the manufacture
of blood plasma products of human origin have also been published (84, 85).
These Guidelines can serve as a general guide for manufacturing practices in
the production of antivenoms. A useful reference in the field of antivenoms is the
Note for guidance on production and quality control of animal immunoglobulins
and immunosera for human use (CPMP/BWP/3354/99) (86).
272
Annex 5
15.2 Purification of the active substance

Antivenoms are prepared from the starting plasma pool using diverse methods
to obtain one of the following active substances:
■■ intact IgG molecules
■■ F(abʹ)2 fragments
■■ Fab fragments.
In general, fractionation procedures should not impair the neutralizing
activity of antibodies; they should yield a product of acceptable physicochemical
characteristics and purity with a low content of protein aggregates, which is
non-pyrogenic and which should provide good recovery of antibody activity.
If possible, the process should be simple (with few steps) and economical.
The characteristics of a batch of plasma to be fractionated should be
clearly established. The methods used to purify the active substance and the
in-process controls should be described in detail in the relevant SOPs. In the
following sections, examples of basic protocols used for the production of IgG,
F(abʹ)2 and Fab antivenoms are described. Some additional methodologies
introduced to further purify the active substance of antivenoms are also
discussed. Variations in these manufacturing procedures have often been
developed by individual fractionators and should be considered as acceptable
when shown to yield consistently safe and effective preparations of antivenoms.
15.2.1 Purification of intact IgG antivenoms

15.2.1.1 Ammonium sulfate precipitation
In the past, most laboratories that produced whole IgG antivenoms have used
fractionation protocols based on salting-out procedures employing ammonium
sulfate or sodium sulfate (87). Two precipitation steps are included using two
different salt concentrations in addition to the elimination of “euglobulins” by
precipitation in a diluted acidic solution.
Such fractionation protocols generally lead to a recovery of antibodies
of between 40 and 50% and to the formation of protein aggregates. The final
product of this procedure used to contain a relatively high proportion of
contaminating proteins, such as albumin (88). This compromised the safety of the
product, since a high incidence of early adverse reactions has been described in
response to protein aggregates (89).
15.2.1.2 Caprylic acid precipitation

The use of caprylic acid (octanoic acid) as an agent for precipitating proteins
from animal plasma has been described in the literature (90). Several procedures
for the purification of whole IgG antivenoms with a good physicochemical
273
profile and purity using caprylic acid precipitation of non-immunoglobulin

proteins have been developed (88, 91, 92) and are now used for the production
of licensed antivenoms.
Fig. A4.6 illustrates a particular process in which caprylic acid is added
slowly to undiluted plasma, with constant stirring, to reach a concentration
of 5% (v/v) and pH 5.5. The mixture is stirred at 22–25 °C for a minimum of
1 hour. The precipitated proteins are removed by filtration or centrifugation
and discarded. The filtrate or the supernatant containing the immunoglobulins
is then submitted to tangential flow filtration to remove residual caprylic acid
and low‑molecular‑mass proteins, depending on the molecular cut-off of the
ultrafiltration membranes, and to concentrate the proteins. The immunoglobulin
solution is then formulated by adding sodium chloride solution (NaCl), an
antimicrobial agent and any other excipient(s) needed, such as stabilizers. The
pH is then adjusted to a neutral value and finally subjected to sterile filtration
through a filter of pore size 0.22 µm, and dispensed into final containers (vials
or ampoules). Variations of this procedure have been introduced by various
manufacturers, and include dilution of plasma, changes in caprylic acid
concentration, pH, and temperature among others.
Caprylic acid fractionation allows the production of antivenoms of
relatively high purity and with a low protein aggregate content, because the
immunoglobulins are not precipitated during the process. The yield may
reach up to 60–75% of the activity in the starting plasma, depending upon the
particular procedure and/or the equipment used. The effectiveness and safety
profiles of caprylic acid-fractionated antivenom immunoglobulins have been
demonstrated in clinical trials (89, 93, 94).
15.2.2 Purification of F(abʹ)2 antivenoms

Many manufacturers follow the classical protocol for F(abʹ)2 antivenom
production developed by Pope (9, 10), with a number of recent modifications
(12, 13, 95).

The method of pepsin digestion (see Fig. A4.7) involves the digestion of
horse plasma proteins by pepsin, leading to the degradation of many non-IgG
proteins, and to the cleavage of IgG into bivalent F(abʹ)2 fragments by removal
and digestion of the Fc fragment into small peptides. A heating step and the
purification of F(abʹ)2 fragments by salting out using ammonium sulfate are
also key elements of this methodology. Some procedures involve performing the
pepsin digestion step on a pre-purified IgG fraction that is obtained by treatment
of plasma with ammonium sulfate to obtain an IgG-enriched precipitate, whereas
albumin is not precipitated.
Pepsin digestion is accomplished at a pH of 3.0–3.5. A typical protocol is
based on incubation at pH 3.3 for 1 hour, at 30–37 °C in a jacketed tank, with a
pepsin concentration of 1.0 g/L. Other procedures can be used which give similar
274
Annex 5
results. Each manufacturer should adjust the pepsin concentration to achieve the
required enzymatic activity.
15.2.2.1 Downstream processing using ammonium sulfate

After pepsin digestion, the pH is adjusted to 4.5–5.0 by adding sodium
hydroxide (or a weak alkaline buffer; then ammonium sulfate is added with
stirring to a final concentration usually close to 12% (w/v). The precipitate is
eliminated by filtration or centrifugation, and the filtrate, or supernatant, is
heat treated (usually at 56 °C for 1 hour; this is known as “thermocoagulation”).
After thermocoagulation, the preparation is cooled down to less than 30 °C, for
example, by passing cold water through a jacketed vessel. The resulting fraction
is filtered or centrifuged to remove the precipitate. The pH is then adjusted to
7.0–7.2 with sodium hydroxide, and a solution of ammonium sulfate is added
with stirring to a final concentration high enough to precipitate the F(abʹ)2
fragments (usually 23% (w/v) or higher). After an additional filtration step, or
following centrifugation, the F(abʹ)2 precipitate is dissolved, and then desalted
(to remove the ammonium sulfate) and concentrated preferentially by tangential
flow diafiltration. Care should be taken to avoid aggregate formation by ensuring
gentle mixing and rapid dissolving of the precipitate. Alternatively, the 23%
(w/v) step is bypassed by some manufacturers and, directly after the heating
step, the filtrate obtained is subjected to ultrafiltration. Additional precipitation
may also be applied on the starting material at low ionic strength and acid pH to
remove “euglobulins” (10).
The F(abʹ)2 solution is then formulated by adding sodium chloride
(NaCl), an antimicrobial agent, and any other excipient needed for formulation,
such as protein stabilizers, and the pH is adjusted, generally to a neutral value.
Finally, the preparation is sterilized by filtration through 0.22 µm filters, and
dispensed into final containers (vials or ampoules). Such a process, or similar
ones developed by other manufacturers, using pepsin digestion, ammonium
sulfate precipitation and tangential diafiltration, is the most often used for the
manufacture of F(abʹ)2 fragments. The yield of this fractionation protocol usually
ranges between 30% and 40%.
15.2.2.2 Downstream processing using caprylic acid

Purification of F(abʹ)2 has also been shown, on an experimental scale, to be
achievable by caprylic acid precipitation of non-F(abʹ)2 proteins after pepsin
digestion, with an improved yield (~60%) (96). However, the yield obtained
on a large scale has not been reported. Fig. A4.8 shows a fractionation scheme
for F(abʹ)2 using caprylic acid. F(abʹ)2 is not precipitated, therefore reducing
the formation of aggregates. Some manufacturers have introduced additional
or alternative processing steps such as ion-exchange chromatography or
ultrafiltration to eliminate low-molecular-mass contaminants.
275
15.2.3 Purification of Fab antivenoms

Production of monovalent Fab fragments is performed by some manufacturers
(97), currently using hyperimmunized sheep plasma. Papain is used to carry
out the enzymatic digestion in the presence of L-cysteine as a promoter, and
the process of preparation of the fragment may use ammonium sulfate, sodium
sulfate or caprylic acid. Fig. A5.9 shows a process in which immunoglobulins
are precipitated from plasma by adding ammonium sulfate or sodium sulfate
to a concentration of 23%. After filtration the filtrate is discarded and the
immunoglobulin precipitate is dissolved in a sodium chloride solution at
pH 7.4. Papain is added and digestion performed at 37 °C for 18–20 hours in
a jacketed tank. Reaction is stopped by adding iodoacetamide. The product is
then applied to a diafiltration system to remove iodoacetamide, salts and low-
molecular-mass peptides and equilibrated with a buffered isotonic NaCl solution.
The preparation is then chromatographed on an anion exchanger (usually
in quaternary aminoethyl (QAE)-based or diethylaminoethyl (DEAE)-based
media). Fc fragments and other impurities are bound on the column, whereas
Fab fragments pass through. After an additional diafiltration/dialysis step, the
product is formulated by adding NaCl, antimicrobial agents (when used) and
any other excipients needed, and the pH is adjusted. Finally, the preparation is
sterile filtered and dispensed into the final containers.
276
Annex 5
Fig. A5.6
Example of a fractionation process in which intact IgG is prepared by caprylic acid
precipitation of non-immunoglobulin proteins
Fractionation of plasma for purification of IgG
Hyperimmune plasma
Acidification at pH 5.5, addition

of caprylic acid to a concentration
of 5% with stirring for 1 hr
Filtration or centrifugation
Discard the precipitate
Filtrate or supernatant
Adjust the pH to 7.0
Tangential diafiltration
and concentration
Concentrated IgG solution
Formulation and sterile

filtration
Bulk preparation
Dispensing in final
container
Final product
277
Fig. A5.7
Example of a fractionation process in which F(abʹ)2 fragments are prepared by pepsin
digestion and ammonium sulfate precipitation
Fractionation of plasma for purification of F(abʹ)2 fragments
Hyperimmune plasma
Acidification at pH 3.0–3.5, digestion with

pepsin at 30 °C for 1 hr
Adjust pH to 4.5–5.0
Addition of 12% ammonium sulfate,
with stirring for 1 hr
Filtration or centrifugation Discard the precipitate
Heating to 56 °C for 1 hr with stirring

Filtration or centrifugation Discard the precipitate
Addition of ammonium sulfate

to reach 23% concentration;
stirring for 1 hr
Filtration or centrifugation Discard the supernatant
F(abʹ)2 paste
Solubilization of precipitate
Tangential flow diafiltration and
concentration
Concentrated F(abʹ)2 solution

filtration
Bulk preparation
Dispensing in final
container
Final product
278
Annex 5
Fig. A5.8
Example of a fractionation process in which F(abʹ)2 fragments are prepared by pepsin
digestion and caprylic acid precipitation
Fractionation of plasma for purification of F(abʹ)2 fragments
Hyperimmune plasma
Acidification at pH 3.0–3.5, digestion with

pepsin at 30 °C for 1 hr
Adjust pH to 5.5. Addition of caprylic

acid to a concentration of 2 to 5% with
stirring for 1 hr
Discard the precipitate
Adjust pH to 7.0
Tangential flow diafiltration and

concentration
Concentrated F(abʹ)2 solution

filtration
Bulk preparation
Dispensing in final
container
Final product
279
Fig. A5.9
Example of a fractionation process in which Fab fragments are prepared by papain
digestion and ammonium sulfate precipitation
Fractionation of plasma for purification of Fab fragments
Hyperimmune plasma
Addition of ammonium or sodium

sulfate to 23% concentration, with
stirring for 1 hr
Discard the supernatant
IgG-rich precipitate
Solubilization of precipitate in
buffered saline solution
Digestion with papain at 37 °C
Addition of iodoacetamide to stop

the reaction
Fab solution
Tangential flow diafiltration
Anion-exchange chromatography
Tangential flow diafiltration

and concentration

filtration
Bulk preparation
Dispensing in final
container
Final product
280
Annex 5
15.2.4 Optional additional or alternative steps used by some manufacturers

When performed following GMP and using validated fractionation protocols, the
basic methodologies described above for the manufacture of IgG, F(abʹ)2 and Fab
antivenoms allow the production of antivenoms of adequate purity, safety and
preclinical efficacy. Nevertheless, some manufacturers include additional steps
to enhance product purity. The methodologies include those described below.
15.2.4.1 Ion-exchange chromatography

Ion-exchange chromatography can be successfully used for antivenom
purification based on charge differential of the contaminants. Anion-exchange
columns of DEAE or QAE gels or membranes, such as quaternary ammonium
cellulose microporous membranes, can be used at neutral pH to adsorb protein
contaminants and endotoxins (13, 95, 98). Alternatively, cation-exchange columns,
for example, carboxymethyl or sulfopropyl gels, have been used for purification of
IgG or F(abʹ)2 fragments (96). The column is equilibrated at acid pH, for example,
pH 4.5, to bind the antivenom IgG or its fragments, whereas protein contaminants
are eluted in the break-through.
Chromatographic procedures should follow GMP. Columns should be
adequately regenerated, sanitized and stored to prolong their useful lifetime.
The reproducibility of columns over cycles should be validated. Measures to
avoid batch-to-batch contamination should be in place. Specific SOPs should be
developed and followed.
15.2.4.2 Affinity chromatography

Affinity chromatography using either immobilized venom or other ligands can
be designed to bind immunoglobulins or their fragments (99). However, columns
usually deteriorate rather rapidly, and meticulous care should be taken to wash,
sanitize and store them under appropriate conditions. Procedures should be
followed to ensure that any substances leaching from the columns do not affect
the quality and safety of the product or else are completely removed during
downstream processing; this is especially critical in affinity chromatography
using immobilized venom. Affinity processes may affect recovery and high-
affinity antibodies may be lost and/or denatured as a result of the harsh elution
conditions needed to elute them from the chromatographic material.
15.2.4.3 Process improvement

Some manufacturers have introduced process improvements to enhance the
quality or the yield of antivenoms. These include the use of a depth-filtration
system combined with filter aids to facilitate filtration steps and improve
antivenom recovery. Other additional manufacturing steps may be introduced
to ensure inactivation or removal of infectious agents (see section 16).
281
15.2.5 Formulation
During formulation of antivenoms after diafiltration steps one should consider
the addition of salts to adjust the osmolality, addition of preservatives, other
excipients, if needed for protein stability, and the adjustment of pH.
In general, antivenoms are formulated at neutral pH (pH 7.0 ± 0.5)
although some manufacturers are exploring the feasibility of formulation at more
acidic pHs to improve stability and/or to reduce aggregate formation.
Formulation at a pH higher than 7.5 may not be recommended, since the
stability of immunoglobulins and their fragments at alkaline pH may be poor,
and the formation of aggregates may be favoured.
15.2.6 Analysis of bulk product before dispensing

The biological, physical and chemical characteristics of the final bulk product
should meet pre-established specifications before dispensing. Analysis may
include tests required to demonstrate:
■■ the purity and potency of the product
■■ product sterility
■■ compliance with the specifications for the aggregate content
■■ the pyrogen limit and/or the bacterial endotoxin content
■■ the formulation, that is, the concentration of excipients and the pH.
When the product is a stored liquid, some of these tests (such as the
potency assay) may not need to be duplicated on the final container if the
processing after the bulk preparation has been validated and shown not to alter
this activity.
The sterilization equipment and the integrity of the membrane should
be guaranteed before and verified after sterilization; the aseptic filling should
also be validated.
15.2.7 Dispensing and labelling of final product

Once compliance of the final bulk product with the quality control specifications
is established, the final product is bottled. For this, final glass containers (vials
or ampoules) should be used. General principles pertaining to the dispensing
of parenteral medicinal products should be applied. The dispensing should be
performed in class A (83) clean room conditions, usually under a laminar flow
hood. The equipment used for dispensing should be calibrated beforehand to
ensure that the correct volume is delivered. European GMPs now recommend
that sterile filtration be carried out at the closest point immediately before filling.
In the case of ampoules, the dispensing system should ensure an aseptic
closure and the sealing of the ampoule should prevent risk of protein denaturation
282
Annex 5
due to heat. For vials, insertion of rubber stoppers should be done inside this
clean dispensing area. The quality of the rubber stoppers should be such as to
guarantee inertness and to prevent leaching. Thereafter, aluminium seals should
be placed on each vial in a clean area outside the class A area. Ampoules or vials
containing the final product should then be properly identified and stored in
a quarantine area maintained under proper storage conditions. Samples of the
antivenoms should be sent to the quality control laboratory for analysis.
When an antivenom complies with all the quality control tests established
for the final product, it should be properly labelled and identified.
■■ The vial or ampoule should be labelled with, at least, the following
information:
(a) name of the product and of the producer;
(b) animal species used to produce the antivenom;
(c) batch number;
(d) pharmaceutical presentation (liquid or freeze-dried);
(e) volume content;
(f ) administration route;
(g) specificity – venoms neutralized by the antivenom, including
both the common and the scientific name of the snake(s);9
(h) neutralizing potency;
( i ) storage conditions; and
( j ) expiry date.
Additional information may be requested by NRAs.
■■ The package, which is usually a cardboard box, in which the vials
or ampoules are packed, should include the same information as is
given on the primary container.
■■ The package insert should include all the information relating to the
product, as established by NRAs, including:
(a) the neutralizing potency;
(b) the recommended dosage;
(c) reconstitution procedure, if lyophilized;
(d) the mode of administration (for example, the dilution of
antivenom in a carrier fluid such as saline);
Special care should be taken to consider potential changes in snake species taxonomy.
9
283
(e) the rate of administration;

(f ) details on the symptoms and treatment of early and delayed
adverse reactions;
(g) snake species against which the antivenom is effective;
(h) recommended storage conditions; and
( i ) an indication that the product is for single use.
15.2.8 Use of preservatives

The addition of preservatives to prevent bacterial and fungal contamination
should be kept to a minimum during plasma storage and during fractionation.
Their inclusion during the manufacturing process should be clearly justified, and
should never substitute for any aspect of GMP. Preservatives can be considered
in the final product, especially if it is manufactured in liquid form. Antimicrobial
agents currently used in antivenom formulation include phenol and cresols.
In general, phenol concentration is adjusted to 2.5 g/L, and concentration of
cresols should be less than 3.5 g/L. The concentration of preservatives should be
validated by each production laboratory on the basis of assays to test their efficacy
and keeping in mind that they may degrade with time and cease to be effective.
It is necessary to ascertain that any agent used has no potential detrimental
interaction with the active substance and excipients of antivenoms. Any change
in the formulation involving preservatives, or the elimination of preservatives
from the final product, requires a very careful risk–benefit assessment on various
microbial safety aspects, as well as a detailed validation procedure. Mercury-
containing preservatives are not recommended in antivenom manufacture.
The volume of antivenom required for the treatment of envenoming (in excess
of 50 mL) might lead to an exposure to mercury far higher than the amounts
currently used for other biological preparations and the levels at which they are
toxic, especially in young children, are not known (100, 101).
15.2.9 Freeze-drying
Antivenoms are available either as liquid or as freeze-dried preparations. Freeze-
dried antivenoms, which may usually be stored at a temperature not exceeding
25 °C, are generally distributed to markets where the cold chain cannot be
guaranteed, such as in many tropical regions of the world. The absence of guarantee
of a cold chain during distribution highlights the need for manufacturers to
demonstrate the stability of the antivenoms under the high temperatures found
in tropical climates.
Freeze-drying is a critical operation. Careful attention should be paid to
the rate of freezing as well as to the protocol used for the primary and secondary
drying cycles (102). The details of the freeze-drying protocols are product-
284
Annex 5
specific and should be adjusted according to the particular formulation of each

antivenom. Inadequate freeze-drying protocols may affect the physicochemical
quality of the product, inducing protein precipitation and denaturation, as well as
aggregate formation, and altering stability and reconstitution. Specific stabilizers,
such as sugars or polyols, aimed at protecting proteins from denaturation and
aggregation, may be added to the final formulation of the antivenom (103).
Bulking agents, frequently used for some biological products, are generally not
required in antivenoms owing to their relatively high protein concentration;
however, they are sometimes used for high-titre monospecific antivenoms.
15.2.10 Inspection of final container

All of the vials or ampoules in each batch of liquid antivenoms should be inspected,
either visually or using a mechanical device. Any vial or ampoule presenting
turbidity, abnormal coloration, presence of particulate matter, or defects of the
vial, stopper or capsule should be discarded. In the case of freeze-dried products,
a representative sample of the whole batch should be first tested dry for meltback
and contamination with foreign matter, then dissolved in the solvent and
inspected further as described. Turbidity can be assessed quantitatively using
a turbidimeter.
15.2.11 Archive samples of antivenoms

In compliance with GMP, manufacturing laboratories should archive a number
of vials of each antivenom batch, under the recommended storage conditions,
in an amount that would enable the repetition of all quality control tests,
when required.
15.3 Pharmacokinetic and pharmacodynamic

properties of IgG, F(abʹ)2 and Fab
Owing to their different molecular mass, the pharmacokinetics of heterologous
IgG molecules (approximately 150 kDa) and F(abʹ)2 (approximately 100 kDa)
and Fab (approximately 50 kDa) fragments differ significantly. In envenomed
patients, Fab fragments have the largest volume of distribution and readily reach
extravascular compartments. Fab fragments are, however, rapidly eliminated,
mainly by renal excretion, thus having a short elimination half-life (from
4–24 hours) (104, 105). In contrast, F(abʹ)2 fragments and intact IgG molecules
are not eliminated by the renal route (they are eliminated by phagocytosis
and opsonized by the reticuloendothelial system) and therefore have a longer
elimination half-life (between 2 and 4 days) (20, 106, 107). Such different
pharmacokinetic profiles have important pharmacodynamic implications, and
the selection of the ideal type of active substance in an antivenom should rely on
a careful analysis of the venom toxicokinetics and antivenom pharmacokinetics.
285
Another difference between low-molecular-mass fragments, such as Fab

and those with a higher molecular mass, such as F(abʹ)2 and IgG, is the number
of paratopes of each molecule: Fab has one antigen-binding site whereas IgG
and F(abʹ)2 each have two binding sites. Thus they will be able to form large and
stable complexes or precipitates with antigens carrying several epitopes, while
Fab will form small, reversible non-precipitable complexes.
Ideally, the volume of distribution of an antivenom should be as similar
as possible to the volume of distribution of the main toxins in a particular venom;
however, this is rarely the case. In venoms composed of low-molecular-mass
toxins, such as some elapid snake venoms, low-molecular-mass neurotoxins
are rapidly absorbed into the bloodstream and are rapidly distributed to
the extravascular spaces where toxin targets are located. Furthermore, low-
molecular-mass toxins are eliminated from the body in a few hours. In these
cases, an antivenom of high distribution volume that readily reaches extravascular
spaces, such as Fab, might be convenient, although its action is then eliminated
within a few hours. It should be noted, however, that a number of elapid venoms
contain some high-molecular-mass toxins of great clinical significance, such as
procoagulants and pre-synaptic phospholipase A2 neurotoxins.
In contrast, in the case of viperid snake venoms and other venoms made
up of toxins of larger molecular mass, including a number of elapid venoms,
many of which act intravascularly to provoke bleeding and coagulopathy, the
situation is different. The time required for toxins to distribute to extravascular
spaces is longer than in the case of low-molecular-mass neurotoxins, and the
targets of some of these toxins are present in the vascular compartment. In
addition, the toxins of viperid venoms have a long half-life in vivo and can
remain in the body for several days (108, 109). In this case, an antivenom
made by Fab fragments neutralizes the toxins that reach the circulation but,
after a certain time has elapsed, the Fab fragments are eliminated and non-
neutralized toxins reach the circulation. This gives rise to the well-known
phenomenon of recurrent envenoming, that is, the reappearance of signs and
symptoms of envenoming at later time intervals after the initial control of
envenoming. This situation demands repeated administration of antivenom to
maintain therapeutic levels of Fab in the circulation (110). Therefore, in such
envenomings, antivenoms made of IgG or F(abʹ)2 may be more appropriate
because of their longer elimination half-lives. Moreover, it has been proposed
that formation of venom–antivenom complexes in the circulation results in the
redistribution of venom components from the extravascular space to the blood
compartment, where they are bound and neutralized by circulating antivenom,
provided that the dose of antivenom is sufficient (111, 112). Consequently, the
maintenance of a high concentration of specific antivenom antibodies in the
circulation for many hours is required for complete neutralization of toxins
286
Annex 5
reaching the bloodstream during both early and late phases of envenoming
(redistribution of toxins) present in the extravascular space. In conclusion, IgG
and F(abʹ)2 antivenoms have a pharmacokinetic profile that makes them more
effective in many types of snake-bite envenoming.

■■ Antivenoms should be manufactured using fractionation
procedures that are well established, validated, and shown to
yield products with proven safety and effectiveness. Fractionation
processes used for the manufacture of antivenoms should adhere
to the principles of GMP for parenteral medicinal products.
■■ Antivenoms can be composed of intact IgG molecules, F(abʹ)2
fragments or Fab fragments. Intact IgG antivenoms are mainly
produced by caprylic acid precipitation of non-IgG plasma
proteins, leaving a highly purified IgG preparation in the
supernatant or filtrate.
■■ F(abʹ)2 fragment antivenoms are produced by pepsin digestion
of plasma proteins, at acidic pH, usually followed by F(abʹ)2
purification by salting out with ammonium sulfate solutions or
by caprylic acid precipitation. Fab monovalent fragments are
obtained by papain digestion of IgG at neutral pH.
■■ Further to ultrafiltration to remove low-molecular-mass
contaminants, preparations are formulated, sterilized by filtration
and dispensed in the final containers. Formulations of antivenoms
may include preservative agents. Additional steps, such as
chromatography, can be added to the fractionation protocols to
enhance purity.
■■ Antivenoms can be presented as liquid or freeze-dried
preparations. Freeze-drying of antivenoms should be performed
in conditions that ensure no denaturation of proteins and no
formation of protein aggregates.
■■ IgG, F(abʹ)2 and Fab antivenoms exhibit different pharmacokinetic
profiles: Fab fragments have a larger distribution volume and a
much shorter elimination half-life. Thus, for viperid envenomings,
IgG or F(abʹ)2 antivenoms have a more suitable pharmacokinetic
profile, whereas Fab fragments may be useful for the neutralization
of venoms rich in low-molecular-mass neurotoxins which are
rapidly distributed to the tissues. However, in general terms, IgG
and F(abʹ)2 antivenoms have shown a better pharmacokinetic
profile than Fab antivenoms.
287
16. Control of infectious risks

16.1 Background
The viral safety of any biological product results from a combination of measures
to ensure a minimal risk of viral contamination in the starting raw material
(plasma), together with steps to inactivate or remove potential contaminating
viruses during processing.
There are currently several recognized complementary approaches used
for virus risk reduction for biological products. These are:
■■ minimizing the potential initial virus content by implementing a
quality system for the production of the raw material;
■■ contribution of the manufacturing processes to inactivating and/
or removing residual viruses during manufacture of the biological
product; such a contribution can be inherent to the existing
production technology or may result from the introduction of
dedicated viral reduction steps;
■■ adherence to GMP at all steps of the manufacturing process;
■■ appropriate and timely response to any infectious events recognized
during the clinical use of the product.
Production steps to inactivate and/or remove viruses have long been
shown to play a powerful role in ensuring the safety of biologicals (84). Similarly,
keeping to a minimum the potential viral load at the stage of the plasma pool,
through appropriate epidemiological surveillance and health control of the
donor animals, is also an important safety measure (see section 12).
Based on experience with human plasma products, a production process
for antivenoms that includes two robust steps for viral reduction (preferably
comprising at least one viral inactivation step) should provide a satisfactory level
of viral safety. However, it should be kept in mind that non-enveloped viruses
are more difficult to inactivate or remove than lipid-enveloped viruses.
16.2 Risk of viral contamination of the starting plasma

The main structural characteristics of viruses reported to possibly infect horses,
sheep and goats are presented in Tables A5.7 and A5.8. They include viruses with
a DNA or RNA genome, with and without a lipid envelope, and vary widely in
size (22 to 300 nm).
A few of these viruses have been identified as possibly present, at least at
some stages of the infection cycle, in the blood, or are considered pathogenic to
humans. Special attention should be paid to these viruses.
288
Annex 5
16.3 Viral validation of manufacturing processes

An understanding of how much a manufacturing process may contribute to the
viral safety of antivenoms is fundamental to both manufacturers and regulators.
Such an understanding can only be achieved by viral validation studies.
These studies are complex and require well-established virology laboratory
infrastructure and cell culture methodologies. They are usually carried out by
specialized laboratories, outside the manufacturing facilities. The principles
guiding such studies have been described in WHO Guidelines (84) and are
summarized below.
Table A5.7
Viruses identified in horses (86, 113)
Virus Family Size Genome a Presence Classified as

(nm) in blood pathogenic
reported b to humans
(86)
Lipid-enveloped viruses
Borna virus c Bornaviridae 70–130 ss-RNA Yes
Equine Arteriviridae 50–60 ss-RNA
arteritis virus
Equine Togaviridae 40–70 ss-RNA Yes
encephalitis
virus, Eastern
and Western
Equine Coronaviridae 75–160 ss-RNA
coronavirus
Equine Retroviridae 80–100 ss-RNA Yes
foamy virus
Equine Herpesviridae 125–150 ds-DNA Yes
herpesvirus
1–5
Equine Lentiviridae 80–100 ss-RNA Yes
infectious
anaemia
virus
Equine Orthomyxoviridae 80–120 ss-RNA Yes
influenza
virus
289

(86)
Equine Paramyxoviridae 150 ss-RNA Yes
morbillivirus
(Hendra virus)
Japanese Flaviviridae 40–70 ss-RNA Yes
encephalitis
virus
Equine Flaviviridae 40–70 ss-RNA Yes (114)
hepacivirus
Equine Flaviviridae 40–70 ss-RNA Yes (115)
pegivirus
Nipah virus Paramyxoviridae 150–300 ss-RNA Yes
Rabies virus Rhabdoviridae 75–180 ss-RNA Yes
Salem virus Paramyxoviridae 150–300 ss-RNA
St Louis Flaviviridae 40–70 ss-RNA Yes
encephalitis
virus
Theiler’s Flaviviridae 40–70 ss-RNA No (114)
disease-
associated
virus
Tick-borne Flaviviridae 40–70 ss-RNA Yes
encephalitis
virus
(116, 117)
Venezuelan Togaviridae 40–70 ss-RNA Yes Yes
equine
encephalitis
virus
Vesicular Rhabdoviridae 50–80 ss-RNA Yes Yes
stomatitis
virus
West Nile Flaviviridae 40–70 ss-RNA Yes Yes
virus
290
Annex 5

(86)
Non-lipid-enveloped viruses
Equine Reoviridae 80 ds-RNA
encephalosis
viruses
Equine Picornaviridae 22–30 ss-RNA
rhinitis A
and B viruses
Equine Reoviridae 60–80 ds-RNA
rotavirus
a
ss-RNA, single-stranded ribonucleic acid; ds-DNA, double-stranded deoxyribonucleic acid; ds-RNA, double-
stranded ribonucleic acid.
b
Absence of a report does not mean that the virus may not be found in the blood at certain stages of the cycle
of infection.
c
Recent studies have suggested that Borna virus is non-pathogenic to humans (118).
Table A5.8
Viruses identified in sheep and goats (86)
Virus Family Size Genome a Reported Classified as

(nm) presence pathogenic
in blood b to humans
(86)
Lipid-enveloped viruses
Adenovirus Adenoviridae 80–110 ds-DNA
Akabane Bunyaviridae 80–120 ss-RNA
virus
Bluetongue Reoviridae 80 ds-RNA Yes
virus
Border Flaviviridae 40–70 ss-RNA
disease virus
Borna virus c Bornaviridae 70–130 ss-RNA Yes
Bovine Herpesviridae 120–200 ds-DNA
herpesvirus
types 1, 2, 4
291

(86)
Bovine viral Togaviridae 40–60 ss-RNA
diarrhoea
virus
Loiping ill Flaviviridae 40–50 ss-RNA Yes
virus
Nairobi sheep Bunyaviridae 80–120 ss-RNA
disease
Ovine/bovine Papillomaviridae 40–55 ds-DNA
papillomavirus
Ovine Herpesviridae 120–200 ds-DNA
herpesvirus 2
Parainfluenza Paramyxoviridae 150–300 ss-RNA Yes
virus type 3
Peste des Paramyxoviridae 150–300 ss-RNA
petits
ruminants
(Morbillivirus)
Poxviruses Poxviridae 140–260 ds-DNA Yes
(Parapox,
Capripox,
Cowpox)
Respiratory Paramyxoviridae 150–300 ss-RNA

syncytial
virus
Retroviruses Retroviridae 80–100 ss-RNA
(Caprine
arthritis
encephalitis
virus, Maedi-
Visna virus,
Jaagsiekte
virus, Bovine
leukaemia
virus
292
Annex 5

(86)
Rift Valley Bunyaviridae 80–120 ss-RNA Yes
fever
complex
Ross river Togaviridae 70 ss-RNA
virus
Rotavirus Reoviridae 80 ds-RNA
Tick-borne Flaviviridae 40–50 ss-RNA Yes
encephalitis
virus
Vesicular Rhabdoviridae 50–380 ss-RNA Yes Yes
stomatitis
virus
Wesselbron Flaviviridae 40–50 ss-RNA Yes
virus
Non-lipid-enveloped viruses
Epizootic Reoviridae 80 ds-RNA
haemorrhagic
disease virus
Foot and Picornaviridae 27–30 ss-RNA Yes
mouth
disease virus
Reovirus 1-3 Reoviridae 60–80 ds-RNA
a
ds-DNA, double-stranded deoxyribonucleic acid; ss-RNA, single-stranded ribonucleic acid; ds-RNA, double-
stranded ribonucleic acid.
b
Absence of a report does not mean that the virus may not be found in the blood at certain stages of the cycle
of infection.
c
Recent studies have suggested that Borna virus is non-pathogenic to humans (118).
293
16.3.1 Down-scale experiments

The contribution of manufacturing processes to inactivation and/or removal of
potential viral contamination should be demonstrated. For this purpose, viral
validation studies should be performed using at least three viruses exhibiting
different structural characteristics. The antivenom manufacturer should first
identify the steps that, based on the existing literature, are likely to remove or
inactivate viruses and, then, provide evidence and quantitative assessment of the
extent of virus reduction achieved for the specific process evaluated.
Validation should be done by down-scale experiments. The accuracy
of the down-scale process should be assessed by comparing the characteristics
of the starting intermediate and the fraction resulting from that step, for both
the laboratory and the production scales. It may be more appropriate to use
manufacturing intermediates for spiking in viral validation studies. Selected
physical factors (for example, temperature, stirring or filtration conditions) and
chemical factors (for example, pH or concentration of precipitating agents such as
caprylic acid) should be as close as possible to those used at manufacturing scale.
Once the step is accurately modelled, the antivenom fraction derived
from the fractionation process just prior to the step being evaluated (for example,
the starting plasma to be subjected to a low pH treatment, or to caprylic acid
precipitation, or a F(abʹ)2 fragment fraction to be subjected to ammonium sulfate
heat treatment) should be spiked with one of the model viruses selected. Viral
infectivity, most often determined using cell culture assays (less frequently animal
models), should be quantified before (for example, prior to pH adjustment
and addition of pepsin) and immediately after (for example, following low pH
adjustment and incubation at that pH for a known period of time in the presence
of pepsin) the steps evaluated to determine the viral clearance achieved. The
results are conventionally expressed as the logarithm (log) of the reduction in
infectivity that is observed. Total infectivity or viral load is calculated as the
infectious titre (infectious units per mL) multiplied by the volume. For a viral
inactivation step, it is highly recommended that the kinetics of the virus kill
be evaluated. Such inactivation kinetics of the infectivity provide an important
indication of the virucidal potential of the step and enables comparison of the
data obtained to those from published studies.
Typically, a viral reduction of 4 logs or more is considered to represent
an effective and reliable viral safety step.
Establishing the relative insensitivity of a manufacturing step to changes
or deviations in process conditions is also important in evaluating its robustness,
in addition to adding to the level of understanding of its contribution to the
overall viral safety of the preparation. This can be achieved by validating the
same step using a range of conditions deviating from those used in production
(such as an upper pH limit applied to a pepsin digestion or to a caprylic acid
precipitation step).
294
Annex 5
Virus validation studies are subject to a number of limitations (84),

which should be considered when interpreting the results.
16.3.2 Selection of viruses for the validation of

antivenom production processes
Viruses selected for viral validation studies should resemble as closely as possible
those which may be present in the starting animal plasma material (Tables A5.7
and A5.8). It is usual to select a wide variety of viruses, some enveloped and some
non-enveloped. At least 1–2 non-enveloped, relatively small viruses should be
selected for validation. When possible, viruses known to potentially contaminate
animal plasma (called “relevant viruses”) should be used.
Table A5.9 gives examples of a few viruses that have been used for the
validation of animal-derived immunoglobulins. Vesicular stomatitis virus (VSV)
and West Nile virus (WNV) are relevant lipid-enveloped horse plasma-borne
viruses. Bovine viral diarrhoea virus (BVDV), a lipid-enveloped flavivirus, can
be used as a model for WNV, tick-borne encephalitis virus, and for the Eastern,
Western, and Venezuelan equine encephalitis togaviruses. Pseudorabies virus is a
lipid-enveloped virus that can serve as a model for pathogenic equine herpesvirus.
Encephalomyocarditis virus (EMCV), a picornavirus, can serve as a model for
non-lipid-enveloped viruses. Porcine parvovirus can also be selected as a model
for small resistant non-lipid-enveloped viruses or as a relevant virus when pepsin
of porcine origin is used in the manufacture of F(abʹ)2 fragments.
This list is not exhaustive and other model viruses can be used for
validation studies of animal-derived antivenoms, in particular taking into account
the characteristics of the viruses that may be present in the animal species used
to generate antivenoms.
295
296
Table A5.9
Examples of laboratory model viruses that can be used for validation studies of horse-derived antivenoms
Virus Family Lipid- Size Genome a Resistance Model for

enveloped (nm)
Animal parvovirus (for Parvoviridae No 18–26 ss-DNA High Relevant virus (when pepsin of
example, porcine) porcine origin is used)
Bovine viral diarrhoea Togaviridae Yes 40–60 ss-RNA Low Eastern, Western and Venezuelan
virus equine encephalitis virus;
tick‑borne encephalitis virus
Parainfluenza virus Paramyxoviridae Yes 100–200 ss-RNA Low Hendra virus; Nipah virus;
Salem virus
Poliovirus; Picornaviridae No 25–30 ss-RNA Medium- Equine rotavirus
encephalomyocarditis high
virus; hepatitis A virus
Pseudorabies virus Herpes Yes 100–200 ds-DNA Medium Equine herpesvirus
Reovirus type 3 Reoviridae No 60–80 ds-RNA Medium Equine encephalosis virus

Sindbis virus Togaviridae Yes 60–70 ss-RNA Low Eastern, Western and
Venezuelan equine
encephalitis virus
Vesicular stomatitis Rhabdoviridae Yes 50–200 ss-RNA Low Relevant virus
virus
West Nile virus Flaviridae Yes 40–70 ss-RNA Low Relevant virus and model for
Eastern equine encephalitis virus
a
ss-DNA, single-stranded deoxyribonucleic acid; ss-RNA, single-stranded ribonucleic acid; ds-DNA, double-stranded deoxyribonucleic acid; ds-RNA, double-stranded
ribonucleic acid.
Annex 5
16.4 Viral validation studies of antivenom immunoglobulins

There is no documented case of transmission of zoonotic infections, including
viral diseases, by antivenom immunoglobulins, or any other animal-derived
immunoglobulins. Absence of reports of viral transmission may result from a
lack of long-term surveillance of the patients receiving antivenoms. Alternatively,
this may reveal that current processes for the manufacturing of antivenoms
include processing steps that contribute to viral safety.
Among the various processing steps used in the production of antivenoms,
caprylic acid and low pH treatments are known to contribute to safety against
lipid-enveloped viruses. This information is based on well-established experience
in the fractionation of human plasma with a production step comprising caprylic
acid (119–121) or low pH treatment (84, 122–124).
Although information is still limited, there is growing evidence that
similar steps used in the production of animal-derived immunoglobulins may also
inactivate or remove viruses. In addition, some manufacturers have implemented
dedicated viral reduction procedures. After the introduction of a new step in the
process, the stability of the product must be verified.
16.4.1 Caprylic acid treatment

The conditions used for caprylic acid treatment of antivenoms (88, 113) and
of human immunoglobulins (119–121) are similar, in particular the pH range,
duration of treatment, temperature, and the caprylic acid/protein ratio, as
summarized in Table A5.10.
Table A5.10
Comparison of conditions for caprylic acid treatment used for human immunoglobulin
preparations and antivenoms (113)
Product Protein Caprylic acid pH Temperature Duration

concentration (g/kg solution) (°C) (hours)
(g/L)
Human IgG 35 7.45 5.5 22 1
Human IgM 43 15 4.8 20 1
enriched
Human IgM 25 20 5.0 20 1
Antivenoms 60–90 50 5.5–5.8 18–22 1
297
16.4.1.1 Validation studies with human immunoglobulins

Unsaturated fatty acids, most specifically caprylic acid, have long been known
to have the capacity to inactivate lipid-enveloped viruses in human plasma
protein fractions (125, 126). The non-ionized form of caprylic acid is thought to
disrupt the lipid bilayer and membrane-associated proteins of enveloped viruses.
Utilizing the dissociation reaction and varying the concentration of the ionized
form of caprylate, a specific amount of the non-ionized form of caprylate can be
maintained over a wide pH range.
The robustness of caprylic acid treatment applied to human IgG,
human IgM and IgM-enriched preparations has been investigated using various
enveloped viruses (human immunodeficiency virus (HIV), BVDV, Sindbis virus
and Pseudorabies virus (120). Under the conditions applied during manufacture,
caprylic acid leads to robust inactivation of lipid-enveloped viruses; pH is a
particularly critical parameter and should be less than 6.
Another investigation studied the viral reduction achieved during
treatment by caprylate of a human IgG product (119). At pH 5.1, 23 °C, in the
presence of 9 mM caprylate, ≥ 4.7 and ≥ 4.2 log of HIV and pseudorabies virus,
respectively, were inactivated during the 1 hour treatment, but only 1.5 log for
BVDV was inactivated. At 12 mM caprylate, ≥ 4.4 log of BVDV were inactivated
within this period. At pH 5.1, 24 °C, and 19 mM caprylate, and pH 5.1, 24 °C, and
12 mM caprylate, complete inactivation of BVDV and of HIV and pseudorabies
virus was achieved in less than 3 minutes.
Treatment of cryoprecipitate-poor plasma with 5% caprylic acid/pH
5.5 at 31 ± 0.5 °C (a condition close to that used to prepare antivenoms) was
shown to inactivate ≥ 5 log of HIV, BVDV and pseudorabies virus in less than
5 min (127).
16.4.1.2 Validation studies with antivenom immunoglobulins

Virus inactivation studies have been carried out on an F(abʹ)2 fraction obtained
from pepsin-digested plasma subjected to ammonium sulfate precipitation. The
F(abʹ)2 fraction was subjected to precipitation by drop-wise addition of caprylic
acid to 0.5% (final concentration) and the mixture was maintained under
vigorous stirring for 1 hour at 18 °C. Rapid and complete reduction of BVDV,
pseudorabies virus and VSV (> 6.6 log 10 , > 6.6 log 10 , and > 7.0 log 10 , respectively)
was observed. No significant reduction (0.7 log 10 ) of the non-enveloped EMCV
(126) was observed.
In another process used to prepare equine immunoglobulins, serum
is thawed at 4 °C, subjected to heating at 56 °C for 90 minutes, brought to
20 ± 5 °C, adjusted to pH 5.5 and subjected to 5% caprylic acid treatment
for 1 hour. This process led to fast reduction of infectivity of > 4.32 and
298
Annex 5
> 4.65 log 10 for pseudorabies virus and BVDV, respectively. The caprylic acid
step was confirmed to have only limited impact on the infectivity of EMCV
and minute virus of mice (MVM) non-lipid-enveloped viruses (128). Data
suggest that significant reduction in the infectivity of lipid-enveloped viruses
can be obtained during caprylic acid treatment of antivenoms. The reduction of
viral infectivity may result from both viral inactivation and partitioning during
the precipitation step. No significant inactivation of non-enveloped viruses
is expected.
16.4.1.3 Recommended actions

Further studies of the viral reduction achieved during caprylic acid treatment
of antivenoms are recommended; in particular, robustness studies to define the
impact on process variations should also be performed.
16.4.2 Acid pH treatment

The conditions used for low pH treatment of equine-derived antivenom
immunoglobulins and of human immunoglobulins are summarized in
Table A5.11.
Table A5.11
Typical conditions for acid pH treatment of human IgG preparations and equine
antivenoms (113)
Product Protein pH Temperature Duration

concentration (g/L) (°C) (hours)
Human IgG 40–60 4.0 30–37 20–30
Antivenoms 60–90 3.1–3.3 30–37 0.6–24
16.4.2.1 Validation studies with human immunoglobulins

Many studies have demonstrated that the low pH 4 treatment used in the
manufacture of human intravenous IgG has the capacity to inactivate lipid-
enveloped viruses (122–124). The rate and extent of inactivation may differ
depending upon the virus. Inactivation is temperature dependent, and is
influenced by the formulation of the IgG solution. Pepsin is sometimes added in
trace amounts (to reduce anticomplementary activity and content of aggregates)
but, at this low concentration, contributes little to virus kill (86). Most non-lipid-
enveloped viruses are resistant to acid pH treatment.
299
16.4.2.2 Virus inactivation studies performed with antivenom immunoglobulins

As described in section 15, peptic cleavage of horse plasma IgG at pH 3.0–3.3
for 60 minutes is a common procedure for the preparation of F(abʹ)2 . More
than 4 logs of inactivation of WNV and of Sindbis virus has been observed
in horse plasma subjected to peptic digestion at pH 3.2 for 60 minutes (129).
WNV was very sensitive whether pepsin was added or not, whereas the rate and
extent of inactivation of Sindbis virus was higher in the presence of pepsin. This
suggests that pH 3.2 alone inactivates WNV, while other phenomena involving
the action of pepsin contribute to Sindbis virus inactivation at low pH.
Confirmation of the significant inactivation of lipid-enveloped viruses
during peptic cleavage of plasma at pH 3.2 was obtained by another study (126).
In this study process, plasma was diluted with two volumes of saline, pH was
adjusted to 3.3, and pepsin was added to a final concentration of 1 g/L. The
mixture was incubated at pH 3.3 for 1 hour. Inactivation of pseudorabies virus
> 5.1 log 10 occurred in less than 6 minutes and > 7.0 log 10 in 60 minutes. There
was > 3.1 log 10 and > 4.5 log 10 inactivation of VSV after 6 and 20 minutes,
respectively. The reduction of infectivity of BVDV was less: 1.7 log 10 after
60 minutes. Inactivation of EMC, a non-enveloped virus, was relatively slow but
reached between 2.5 and 5.7 log 10 after 60 minutes of pepsin incubation. This
showed that reduction of infectivity of at least some non-lipid-enveloped viruses
may take place during peptic digestion of diluted horse plasma. This does not
mean, however, that other non-lipid-enveloped viruses would be inactivated to
the same extent under such conditions.
16.4.2.3 Recommended actions

Manufacturers of F(abʹ)2 antivenoms must validate the pepsin digestion process
since virus inactivation is likely to be influenced by pH, time, temperature,
pepsin content and protein content. Robustness studies to define the impact on
process variations are also recommended.
16.4.3 Filtration steps

Other steps used in antivenom production may contribute to viral safety
through nonspecific virus removal. The virus removal capacity of two depth-
filtration steps performed in the presence of filter aids and used in the
production of equine-derived immunoglobulins prepared by ammonium sulfate
precipitation of pepsin-digested IgG has been evaluated (130). Clearance factors
of 5.7 and 4.0 log 10 have been found for two lipid-enveloped viruses (infectious
bovine rhinotracheitis virus and canine distemper virus, respectively) and of
5.3 and 4.2 log 10 for two non-lipid-enveloped viruses (canine adenovirus virus
and poliovirus type I, respectively). However, it should be kept in mind that
viral reductions obtained by non-dedicated removal steps are usually regarded
300
Annex 5
as less robust than those resulting from dedicated viral inactivation or removal
steps (84).
16.4.4 Validation of dedicated viral reduction treatments

16.4.4.1 Pasteurization
Pasteurization is defined as the treatment of a liquid protein fraction for
10 hours, usually at 60 °C. It is a well-established viral inactivation treatment of
human plasma products, such as IgG (84). It is being used in the production
process of a few equine-derived immunoglobulins (13).
Validation studies have shown that heating a purified equine
immunoglobulin at 58 °C ± 0.1 °C without stabilizers inactivates ≥ 4.8 log 10
of pseudorabies virus and ≥ 4.3 log 10 of BVDV in less than 30 minutes, and
> 4.7 log of EMCV in less than 1 hour. In contrast, infectivity of MVM, a non-
enveloped virus, was still detected after 9 hours and 30 minutes of treatment;
only 1.59 log 10 were inactivated (128).
16.4.4.2 Nanofiltration
Nanofiltration is a technique of filtration specifically designed to remove viruses,
based on size, while permitting flow-through of the desired protein (131).
Effective virus removal requires, in principle, that the pore size of the filter be
smaller than the effective diameter of the virus particles.
16.4.5 Other viral inactivation treatments currently

not used in antivenom manufacture
Other methods of viral inactivation have been developed to ensure the viral
safety of biological products. These include, in particular, a treatment with
a combination of an organic solvent (tri-n-butyl phosphate or TnBP) at
concentrations between 0.3 and 1%, and detergents such as Triton™ X-100
or Tween® 80, also at concentrations generally between 0.3% and 1%. Such
solvent–detergent (S/D) procedures have proven very efficient and robust in
the inactivation of lipid-enveloped viruses in human plasma products (84).
However, use of this method for antivenoms has not been reported.
Implementation of dedicated viral inactivation treatments, such as
S/D or other methods, should be encouraged for processes which, based on
risk assessment, would offer an insufficient margin of viral safety. Process
changes associated with the introduction of new viral reduction steps, and
the subsequent removal of any toxic compounds needed for viral inactivation,
should be demonstrated not to affect the quality and stability of antivenoms, and
most particularly the neutralization efficacy of venoms. Preclinical assessment
of the possible impact of newly introduced viral inactivation treatments should
be mandatory.
301
16.4.6 Possible contribution of phenol and cresols

The antibacterial agents, phenol and cresols, and more rarely formaldehyde,
are added by most manufacturers to the starting plasma donations as well as
to the final liquid antivenom preparations, at a maximum final concentration
of 0.25–0.35%. Compounds like phenol are known to be very lipid soluble and
lipophilic.
Addition of 0.25% (final concentration) phenol to concentrated
antivenom bulk at pH 6.5–6.7, prior to its dilution and formulation, was found
to inactivate eight enveloped and non-enveloped viruses very efficiently within
30 minutes (132).
Performing additional validations of the virucidal effect of antimicrobial
agents as added to the starting hyperimmune plasma and to the final antivenom
preparations is encouraged. More information is needed on the potential impact
of these antimicrobial agents on the viral safety of antivenoms.
16.5 Production-scale implementation of process

steps contributing to viral safety
As there is increasing, although preliminary, evidence that at least some of the
existing steps in the production of antivenoms contribute to viral reduction,
it is already recommended that specific care should be taken to ensure their
appropriate industrial implementation so as not to compromise any possible
benefits they provide in terms of viral safety.
Measures should therefore be taken to ensure that such steps are correctly
carried out in a manufacturing environment and that cross-contamination and
downstream contamination are avoided. Such important aspects of product
safety have been highlighted in WHO Guidelines (84) and should also be taken
into consideration for large-scale manufacture of antivenoms. Specific attention
should be paid to:

■■ Process design and layout: in particular the production floor area
needed to carry out such treatment safely, minimizing the chance of
cross-contamination between pre-viral-treatment steps product and
post-treatment product; the justification for creating a safety zone to
avoid risk of downstream contamination, and the procedures used
for cleaning and sanitization of the equipment to avoid batch-to-
batch cross-contamination.
■■ Equipment specifications: having in mind the potential contribution
to viral reduction. For instance, vessels used for low pH incubation
or caprylic acid treatment should be fully enclosed and temperature-
controlled. There should be no “dead points” where the temperature
302
Annex 5
defined in the specification or the homogeneity of mixing cannot

be ensured. A poor equipment design could compromise the viral
safety potentially afforded by a given production step.
■■ Qualification and validation: should verify that the equipment
conforms to predefined technical specifications and relevant GMP.
■■ Process implementation: production steps contributing to viral
safety such as low pH treatment and caprylic acid treatments could
be implemented in two stages performed in two distinct enclosed
tanks. Care should be taken to ensure complete process segregation
before and after the completion of these treatments to avoid risks of
downstream contamination.
■■ Process control: is a critical part of the manufacturing process since
completion of viral inactivation and removal cannot be guaranteed
by testing the final product. Samples should be taken to confirm
that the process conditions of claimed inactivation steps meet the
specified limits (for example, for pH, concentration of stabilizers
and concentration of virus inactivating agent, such as caprylate).
When this is technically feasible and intermediates are stable,
samples can be kept frozen for possible additional analysis prior to
the release of the batch. It is the responsibility of the manufacturer
to ensure that the execution of steps contributing to virus inactivation
and removal in a production setting conforms to the conditions that
contribute to such virus reduction.
■■ SOPs: steps contributing to viral reduction should be described in
approved SOPs. These should contain critical process limits for the
viral inactivation and removal methods.
■■ Role of the quality assurance department: because of the critical
nature of the viral inactivation and removal steps, quality assurance
personnel should review and approve the recorded conditions for
viral inactivation and removal while the batch is being processed;
that is, not just as part of the final overall review of the batch file.
16.6 Transmissible spongiform encephalopathy

Transmissible spongiform encephalopathy (TSE) has not been identified in
any equine species. There has been no case of transmission of TSE linked to
antivenoms or other equine-derived blood products.
Of particular concern, however, is that TSEs include scrapie in sheep, a
ruminant species that is used, although much less frequently than horses, in the
manufacture of antivenoms. Scrapie is a disease similar to bovine spongiform
encephalopathy (BSE or “mad cow disease”), but is not known to infect humans.
303
However, the blood of sheep with experimental BSE or natural scrapie can be
infectious and, because scrapie and BSE prion agents behave similarly in sheep
and goats, the use of the blood of small ruminants in preparing biologicals should
either be avoided or the animals should be selected very carefully from sources
known to be free of TSEs. The findings of disease-associated proteins in muscle
tissue of sheep with scrapie and the recognition of BSE itself in a goat, reinforce
the need for manufacturers of biologicals, including antivenoms, to maintain the
precautionary safety measures recommended in the WHO guidelines on TSE
tissue infectivity (77).
According to these recommendations, the use of tissues or body fluids
of ruminant origin should be avoided in the preparation of biological and
pharmaceutical products. When sheep-derived materials must be used, they
should therefore be obtained from sources assessed to have negligible risk
from the infectious agent of scrapie. Documented surveillance records should
be available. The feed of animals used for production of hyperimmune plasma
should be free of ruminant-derived material.
The infectious agent is thought to be a misfolded, abnormal, prion protein
(PrPTSE). It is not yet known whether manufacturing processes used to produce
antivenoms from sheep plasma include steps that can contribute to the removal
of PrPTSE. Experimental prion clearance studies, based on spiking experiments,
can be performed to assess the capacity of the process to remove prions. However,
there is still uncertainty about the validity of such experimental studies since the
biochemical features of PrPTSE in blood and plasma are not known.

■■ The viral safety of antivenoms results from a combination of
measures:
–– to ensure satisfactory health status of the animals;

–– to reduce the risk of contamination in the starting raw
material;
–– to ensure the contribution of the manufacturing process
towards inactivation and/or removal of viruses; and
–– to ensure compliance with GMP along the entire chain of
production.
■■ Manufacturing processes should include at least two steps
contributing to robust viral reduction. A virus inactivation step
that can be easily monitored is usually preferred to other means
of viral reduction, such as nonspecific removal.
304
Annex 5
■■ Manufacturers must evaluate the capacity of their current

manufacturing processes (in particular low pH pepsin digestion,
caprylic acid treatment, ammonium sulfate or heat precipitation,
and possibly other steps) to inactivate or remove viruses and
validate them, if necessary. These studies should be done following
existing international guidelines and using relevant and/or
model viruses that are representative of the viruses that could
affect the animals used for the production of the antivenom
immunoglobulins.
■■ The removal of antimicrobial agents from the final formulation of
antivenoms should be carefully weighed against the likely benefits
these agents may have on the viral safety.
■■ Should the viral reduction processes used be found to be insufficient
to ensure a margin of safety, the introduction of dedicated viral
reduction methods should be considered. The impact of such
process changes on product efficacy and safety should be carefully
analysed in vitro as well as in preclinical studies before performing
clinical evaluations in humans.
■■ Great attention should be paid to the production-scale
implementation of all steps contributing to viral safety to ensure
a consistent and reproducible batch-to-batch viral reduction
and an absence of risks of cross-contamination and downstream
recontamination that would jeopardize the viral safety of
the product.
■■ When materials originating from sheep must be used for the
production of plasma, they should be obtained from sources
assessed to have negligible risk from the infectious agent of scrapie.
17. Quality control of antivenoms

Quality control of the final product is a key element in the quality assurance
of antivenoms. Quality control tests should be performed by the manufacturer
or under its responsibility before the product is released. In addition, relevant
analyses should be performed on any intermediate steps of the manufacturing
protocol as part of the in-process quality control system.
The results obtained should meet the specifications approved for
each antivenom product or its intermediates, and constitute part of the batch
record. For a liquid preparation, some quality control tests, such as the venom-
neutralizing efficacy test or the detection of residual reagents used during
fractionation, can be performed on the final bulk and may not need to be repeated
305
on the final bottled product if the processing after the bulk preparation has been
validated and shown not to have any impact. Quality control assessment of the
final antivenom product includes the tests described below.
17.1 Standard quality assays

17.1.1 Appearance
The appearance of the product (for example, colour and clarity of the liquid,
appearance of the powder) should comply with the description in the marketing
dossier.
17.1.2 Solubility (freeze-dried preparations)

The time from the addition of solvent to the complete dissolution of freeze-dried
antivenom, under gentle mixing, should be determined. Antivenoms should
be completely dissolved within 10 minutes at room temperature. The solution
should not be cloudy. Shaking of the container should be avoided to prevent the
formation of foam.
17.1.3 Extractable volume

The volume of product extractable from the container should be in compliance
with that indicated on the label.
17.1.4 Venom-neutralizing efficacy tests

These tests determine the capability of an antivenom to neutralize the lethal
effect of the snake venom(s) against which the antivenom is designed. It is first
necessary to determine the lethal potency of the venom, using the LD50 assay.
The exact volume of antivenom required to neutralize venom lethality can then
be determined using the antivenom effective dose (ED50 ) assay.
The outputs of these tests provide globally applicable standard metrics
of: venom lethality and antivenom efficacy, which enable internal monitoring
and external, independent auditing of antivenom efficacy – thereby preventing
the distribution of ineffective antivenom.
Consistent use of outbred strains of mice, of a defined weight range (for
example, 18–20 g) that receive a defined challenge dose, is recommended for all
the assays. Some producers use other test animals, such as guinea-pigs. While
weights will clearly vary between animal species, a series of principles, specified
for mice, will still apply to these alternative test animals. It should be borne in
mind that there are variations in the susceptibility of different strains of mice
to the lethal effect of venoms.
The venom-neutralizing potency tests are used for quality control and
preclinical assessment, so protocol details are described in section 19, while
ethical issues are discussed in section 4.
306
Annex 5
17.1.5 Osmolality
Osmolality is used to measure the tonicity of the antivenom solution, and should
be at least 240 mOsmol/kg. Determination of osmolality is also an indirect
means to determine the quantity of salts or excipients added for formulating
the batch.
17.1.6 Identity test

When several types of antivenoms are produced by a single production facility, a
system to identify each batch of antivenom should be established for monitoring
and auditing purposes. Identity tests may include biological assays as well as
physicochemical and immunological tests. Double immunodiffusion assays,
confronting the antivenom with the venoms against which the antivenom is
designed to act, are often used. In the case of laboratories that use various animal
species to raise antivenoms, that is, horses and sheep, an immunological identity
test should be used to identify the mammalian species in which the antivenoms
are produced.
17.1.7 Protein concentration

The total protein concentration of antivenoms can be determined using a variety
of approaches, including:
■■ the Kjeldahl method to determine nitrogen content;
■■ several colorimetric procedures; and
■■ spectrophotometric (280 nm) assays.
The presence of preservatives should be taken into account since they may
interfere with some protein determination methods (133).
The total protein concentration of antivenoms should preferably not
exceed 10 g/dL, since administration of higher amounts of protein may be
associated with higher adverse reaction rates, although some jurisdictions have
authorized higher concentrations.
17.1.8 Purity and integrity of the immunoglobulin

The purity and integrity of the active substance, that is, intact immunoglobulin
or immunoglobulin fragments, should be assessed to identify contaminants
and immunoglobulin degradation. Immunoglobulins or their fragments should
constitute the great majority of the preparation, ideally greater than 90%. Evidence
suggests, however, that although antivenoms may have physicochemical purity
> 90% (for example, immunoglobulins or their fragments), immunochemical
purity (for example, specificity for the snake venoms they are produced from) can
be lower than 40% (134). These findings have emphasized the need to incorporate
307
both physicochemical and immunochemical analyses in the assessment of

antivenom purity.
Electrophoretic methods in polyacrylamide gels (SDS-PAGE run under
reducing or non-reducing conditions) are suitable for this purpose, since
these techniques allow the detection and monitoring of IgG, F(abʹ)2 , Fab, non-
immunoglobulin plasma protein contaminants (in particular albumin) and
degradation products. The electrophoretic pattern should be compared to that of
a reference preparation. A semi-quantification can be performed by calibration of
the procedure. Of particular relevance is the assessment of the albumin content,
which ideally should not exceed 1% of total protein content. The following
approach can serve as a guide in assessing the purity of antivenoms:
■■ SDS-PAGE under non-reducing conditions – this analysis provides
qualitative (or, at best, semi-quantitative) information on the
amounts of intact immunoglobulins, digestion products and,
importantly, on the presence of high-molecular-mass oligomers
(soluble aggregates) and low-molecular-mass contaminants (which
are expected in the case of enzymatically digested antivenoms).
■■ SDS-PAGE under reducing conditions – analysis under these
conditions can provide information on the amount of
immunoglobulins and their fragments by direct visualization of
intact and/or digested immunoglobulin heavy chains.
17.1.9 Molecular-size distribution

The presence of aggregates and other components in antivenoms can be assessed
by size-exclusion liquid chromatography (gel filtration) in fast protein liquid
chromatography (FPLC) or HPLC systems.
Densitometric analyses of chromatographic profiles allow the
quantification of protein aggregates and of the relative abundances of: intact

immunoglobulins, divalent immunoglobulin fragments (F(abʹ)2 , monovalent
immunoglobulin fragments (Fab) and dimers, as well as low-molecular-mass
enzymatic digestion products.
In intact immunoglobulin-based antivenoms this method allows
quantitation of albumin, as its molecular mass (~66 kDa) can be resolved from
the ~160 kDa peak of intact immunoglobulins.
17.1.10 Test for pyrogen substances

Antivenoms should comply with the rabbit pyrogen test where required by the
local regulations. This test is based on intravenous injection of antivenoms in
the ear vein of rabbits. The dose of antivenom must be calculated by dividing the
threshold pyrogenic dose in rabbits by the endotoxin although other doses might
308
Annex 5
be used depending on the pharmacopoeia), followed by the measurement of rectal

temperature at various time intervals after injection. The detailed procedures
are described in various pharmacopoeias. Bacterial lipopolysaccharides can
also be detected by the Limulus amoebocyte lysate (LAL) test. The test should
be validated for each type of antivenom, since there have been reports of false-
positive and false-negative reactions when testing antivenoms and other plasma-
derived products. The sensitivity of this LAL test should be correlated with the
rabbit pyrogen test, and the endotoxin limits established (135). When regulation
allows, a validated LAL test is used in place of the rabbit pyrogen test.
17.1.11 Abnormal toxicity test

The abnormal toxicity test (7 day observation of the effects of intraperitoneal
injection of 0.2 mL and 0.5 mL antivenom into mice and guinea-pigs,
respectively), is still required by some pharmacopoeias and is performed at the
stage of product development.
However, because of the very limited quality control value of this assay,
it is increasingly being abandoned by most regulatory authorities. Correct
implementation of GMP should provide evidence that the product would
comply with the test for abnormal toxicity.
17.1.12 Sterility test

Antivenoms should be free of bacteria and fungi, that is, they should be sterile.
The sterility test is performed following methodologies specified in various
pharmacopoeias such as the European Pharmacopoeia.
Since antivenoms may contain preservatives in their formulation, it
is necessary to “neutralize” the preservatives before the samples are added to
culture media. This is usually done by filtering a volume of antivenom through
a 0.45 µm pore-size membrane, and then filtering through the same membrane
a solution that neutralizes the bacteriostatic and fungistatic effects of the
preservatives used in antivenom. The membrane is then aseptically removed
and cut into two halves. One half is added to trypticase soy broth and the other
is added to thioglycolate medium. Control culture flasks are included for each
medium. Flasks are incubated at 20–25 °C (trypticase soy broth) or at 30–35 °C
(thioglycolate) for 14 days. Culture flasks are examined daily for bacterial or
fungal growth. The number of vials tested per batch should be in compliance
with local regulations.
17.1.13 Concentration of sodium chloride and other excipients

The concentration of the various excipients or stabilizers added during
formulation should be determined using appropriate chemical methods.
309
17.1.14 Determination of pH
The pH of antivenom should be determined using a potentiometer.
17.1.15 Concentration of preservatives

Phenol concentration should not exceed 2.5 g/L and cresols 3.5 g/L.
Phenol concentration can be determined spectrophotometrically on
the basis of the reactivity of phenol with 4-aminoantipyrine, under alkaline
conditions (pH 9.0–9.2) in the presence of potassium ferrocyanide as oxidant.
Other methods are also available. Phenol and cresols can be determined by
HPLC methods.
17.1.16 Chemical agents used in plasma fractionation

The chemical reagents used in the precipitation and purification of antivenoms,
such as ammonium sulfate, caprylic acid and others, should be removed from
the final product during diafiltration or dialysis. Limits should be established and
their residual amount quantified in the final product. Likewise, the elimination
of pepsin or papain from the final preparations should be guaranteed, especially
for preparations that are maintained in liquid form, to avoid proteolytic activity
that may damage the antivenoms.
The determination of the residual amount of agents used in plasma
fractionation could be excluded from routine release testing if the process of
manufacturing has been validated to eliminate these reagents. The detection
of residual reagents can also be performed on the final bulk rather than in the
final product.
17.1.17 Residual moisture (freeze-dried preparations)

Residual moisture content can be determined by several methodologies, such as:
■■ a gravimetric method assessing the loss of weight on heating;

■■ the Karl-Fischer titration, based on the principle that iodine, together
with pyridine, sulfur dioxide and methanol from the reagent react
quantitatively with water;
■■ thermogravimetric methods.
The methodology most commonly recommended is the Karl-Fischer
titration. Every manufacturing and quality control laboratory must establish
the accepted maximum residual moisture for their antivenom ensuring the
stability of the product over its claimed shelf-life. A residual moisture content
of less than 3% is usually recommended for most freeze-dried therapeutic
biological products.
310
Annex 5
17.2 Antivenom reference preparations

The use by manufacturers of in-house reference preparations of antivenoms,
instead of national or regional standards, is recommended, since venom-
neutralizing efficacy, specificity, and purity can only be compared with antivenoms
of similar specificity and neutralizing profile. In-house reference antivenom
preparations should be obtained from a suitable batch of a product that has been
fully characterized and evaluated by the quality control laboratory. For assays not
related to venom-neutralizing efficacy or specificity, such as the quantification
of proteins, preservatives and excipients, national or regional standards can be
used. The preparation of national or regional reference antivenom preparations
should be undertaken by relevant national or regional drug control laboratories
and regulatory agencies. These also require comprehensive characterization
and evaluation by drug control laboratories and appropriate validation prior to
acceptance and establishment.

■■ Quality control of antivenom preparations, both for intermediate
and final products, as part of the batch release, must be performed
by the manufacturers and results disclosed in the documentation.
■■ Results from the following quality control tests need to be provided
by the manufacturer as part of the batch release documentation:
(a) venom-neutralizing efficacy test against the most relevant
venoms to be neutralized;
(b) identity test;
(c) protein concentration;
(d) purity of the active substance;
(e) content of protein aggregates and non-IgG contaminants;
(f ) pyrogen test;
(g) sterility test;
(h) concentration of excipients;
( i ) osmolality;
( j ) pH;
(k) concentration of preservatives;
( l ) determination of traces of agents used in plasma fractionation;
(m) appearance, and, for freeze-dried preparations, residual
moisture and solubility; and
(n) labelling validation and confirmation.
311
■■ Antivenom reference preparations reflecting specific characteristics

of antivenoms produced should be prepared by each manufacturer
to be used as standards in their laboratory settings, in particular to
measure neutralization capacity of their specific antivenom
products against targeted venoms. Relevant standards are also used
to establish conformity of purity and integrity. When possible, a
national reference antivenom should be established.
■■ It is the ethical responsibility of the manufacturer to use only
the minimum number of experimental animals to measure the
efficacy of an antivenom.
■■ The development of in vitro methods validated for replacing
animal experiments is strongly encouraged.
18. Stability, storage and distribution of antivenoms

18.1 Stability
Stability studies should be performed to determine the stability of antivenoms.
These studies should be done when a new product, a process change, or a new
formulation is developed. They are essential to define the shelf-life of the product
and are intended to prove that the antivenom remains stable and efficacious
until the expiry date. During the developmental stages, stability studies should
be included. Their design should take into consideration that they are a complex
set of procedures involving considerable cost, time and expertise necessary to
build in quality, efficacy and safety of a product.
Most liquid antivenom preparations have a shelf-life of up to 3 years
when stored refrigerated at 2–8 °C, whereas freeze-dried antivenoms have shelf-
lives of up to 5 years when kept in the dark at room temperature. It is essential,
however, that manufacturers determine the actual stability of each antivenom

formulation under appropriate conditions using validated methodologies. It
is also highly recommended that manufacturers perform stability studies to
evaluate the possibility that their preparations could be stored for a long period
without refrigeration (for instance at 30 °C).
Real-time stability tests should be performed under the expected storage
conditions of the antivenom. In addition, these tests could be performed under
worst-case storage conditions. Quality control parameters are determined at
regular pre-established time intervals, normally extending the test period in
order to allow significant product degradation under recommended storage
conditions. Essential parameters include venom neutralization potency, turbidity
and content of aggregates, among others, since these are especially prone to alter
upon storage.
312
Annex 5
Accelerated stability studies may be performed to provide early useful

information on the product stability profile, but are not a substitute for real-time
data. In accelerated studies, the antivenom is exposed to harsher conditions than
usual, such as a higher temperature, and the stability is assessed over a shorter
time span. This is done to assess the conditions that accelerate degradation of the
product and this information is then used to predict shelf-life.
Retained sample stability testing is usual practice for every product for
which stability data are required. In such a study, retained samples from at least
one batch a year are selected and tested at predetermined intervals – that is, if a
product has shelf-life of 5 years, it is conventional to test samples at 3, 6, 9, 12, 18,
24, 36, 48, and 60 months.
Cyclic temperature stress testing is not routinely performed, but it
may be useful since it is designed to mimic likely conditions in field place
storage. It is recommended that the minimum and maximum temperatures are
selected on a product-by-product basis and taking into account factors like
recommended storage temperatures as well as specific chemical and physical
degradation properties.
18.2 Storage
Antivenoms should be stored at a temperature within the range that assures
stability, as found by stability tests. This is particularly critical for liquid
formulations, which usually require storage at between 2 and 8 °C. Therefore,
deviations from this temperature range, due to interruptions in the cold chain
during transportation or storage, are likely to result in product deterioration. The
design of adequate cold chain programmes, as part of the public health systems
in every country, is critical, and national protocols should be developed. The
distribution policies for national vaccination programmes can be adopted for
the transportation and storage of antivenoms. The stability of liquid preparations
at temperatures higher than 2–8 °C should be evaluated and, if needed, new
formulations allowing such storage conditions should be developed.
18.3 Distribution
Adequate distribution of antivenoms is a matter of great concern in many
regions of the world. Since most of the antivenoms available are liquid
preparations, the maintenance of an adequate cold chain must be guaranteed,
despite the difficulties to be encountered in rural areas of some developing
countries. National and regional health authorities should develop distribution
strategies to ensure that antivenoms are allocated to the areas where they are
needed or use the distribution channels in place for other national primary
health-care programmes. Both the specificity of the antivenom and the number
313
of vials or ampoules to be distributed should be taken into consideration. This

is particularly relevant in countries that use monospecific antivenoms, since
distribution of these products should be guided by the known distribution of
the species and the epidemiological data. To ensure an appropriate supply for
clinical use, inventories should be in excess of the estimated number of cases, to
allow for unpredictable surges in local demand, accepting that some antivenoms
will not have been used before their expiry date.

■■ The quality control of each antivenom batch prepared by a
manufacturer should include the potency test for neutralization of
lethality (ED50 ).
■■ In general, liquid preparations require a cold chain, whereas freeze-
dried preparations do not. However, storage conditions are product-
or formulation-specific and may vary. Manufacturers should
therefore determine the stability of each antivenom pharmaceutical
preparation by conducting real-time stability studies.
■■ Manufacturers should study the stability of antivenoms at the
ambient temperatures in the areas where the product will be used.
■■ The distribution of antivenoms by health authorities should
rely on a proper assessment of the epidemiology of snake-bite
envenomings, and on the proper knowledge of the geographical
distribution of the most relevant venomous species. This is
particularly important for monospecific antivenoms.
■■ NRAs should ask manufacturers to provide information obtained
from the preclinical assessment of all antivenom used in their
territories against the venoms found in the region or country
where the product is intended to be used.
19. Preclinical assessment of antivenom efficacy

Efficacy testing of antivenoms is one of a suite of assessments required for the
quality control of antivenoms (see section 17 where further quality control tests
are described) performed for each new batch. Efficacy testing of antivenoms is
also part of the preclinical programme to be performed for new antivenoms,
where the respective data are used for licensing or registration of antivenoms
by regulatory agencies. The details of efficacy testing in the preclinical phase of
antivenoms and for quality control purposes are described below. The testing of
antivenoms on animals raises important ethical considerations (section 4) and it
is essential that manufacturers and others apply the highest standards of ethical
314
Annex 5
conduct, including appropriate 3R steps, and use of analgesia or anaesthesia for

the minimization of pain and discomfort.
It is a fundamental regulatory and ethical requirement that all new
therapeutic agents for human use are tested for their safety and efficacy – initially
by in vitro LD50 laboratory tests and then in vivo ED50 preclinical tests and, if
the results of these prove satisfactory, by clinical trials in human patients. The
preclinical efficacy tests must therefore be performed on new antivenoms, and
newly manufactured batches of existing antivenoms.
19.1 Preliminary steps that may limit the need

for animal experimentation
To prevent unnecessary animal use, careful perusal of existing literature for data
on venom lethality may help to refine the experimental design and thereby
reduce the number of experimental animals required.
Manufacturers may also investigate the immunological venom-binding
capability of an antivenom by performing immunological assays (for example,
ELISA, to identify, and exclude from experimentation, antivenoms that do not
possess the requisite titre of venom-binding immunoglobulins. It is crucial to
note, however, that: (a) a high venom-binding titre in an ELISA result for an
antivenom cannot be used to infer venom-neutralizing efficacy; and (b) the
failure of an antivenom to bind venom in an ELISA result suggests very strongly
that the antivenom should be considered ineffective at neutralizing the effects of
that venom – and withdrawn from ED50 testing. This step can further limit non-
productive animal experimentation. There is no single ELISA metric that enables
stop/go decisions to be made for all the possible snake venom and antivenom
combinations. These will therefore be in-house decisions.
An additional immunological cross-reactivity technology that can inform
the preclinical assessment process before animal experiments are undertaken
is the use of a proteomics-centred platform, termed antivenomics, which has
been developed to assess the immunological reactivity of antivenoms against
homologous and heterologous venoms (136–139). Antivenomics complements
the in vitro and in vivo venom activity neutralization assays and substitute for
the traditional, essentially qualitative, immunological methods, such as ELISA
and Western blotting. Antivenomics uses an affinity chromatography approach
to investigate the immuno-capturing ability of immobilized IgG, F(abʹ)2 , or
Fab antibody molecules followed by the proteomic identification of the venom
components recovered in both the retained and the non-bound fractions.
The fraction of non-immuno-captured protein “i” (%NRi) is estimated as the
relative ratio of the chromatographic areas of the same protein recovered in the
non‑retained (NRi) and retained (Ri) affinity chromatography fractions using
the equation:
315
%NRi = 100 − [(Ri/(Ri + NRi)) × 100]

The antivenomic analysis provides both qualitative and quantitative
information on the types of venom proteins presenting antivenom-recognized
epitopes and those exhibiting impaired immunoreactivity. Although the level
of immune recognition gathered from antivenomics should not be absolutely
relied upon to predict the in vivo neutralization capacity of an antivenom (since
both experiments involve radically different protocols), an immuno-capture
capability of ≥ 25% generally correlates with a good outcome in homologous in
vivo neutralization tests. If immuno-capture by this method is < 25% the further
testing of an antivenom using in vivo methods should be reconsidered.
As the degree of immuno-recognition of a given toxin by the
immunoglobulins present in antivenom represents a measure of the capability
of that particular antivenom to neutralize the toxic activity of that toxin, the
antivenomics analysis may assist in assessing the range of clinical applications
of current commercial or experimental antivenoms, and in the development of
improved antivenoms on an immunologically sound basis. Growing evidence
shows the potential of the combination of antivenomics and neutralization
assays for analysing at the molecular level the preclinical efficacy of antivenoms
against homologous and heterologous venoms. This is particularly so where
antivenoms from more than one manufacturer, or from more than one batch of
antivenom produced by a manufacturer require preclinical evaluation against
the same venoms (136, 140).
Manufacturers should take steps to incorporate these approaches into
their preliminary screening of antivenoms before in vivo animal testing is
conducted.
19.2 Essential preclinical assays to measure antivenom

neutralization of venom-induced lethality
These tests determine the capability of an antivenom to neutralize the lethal effect
of the snake venom(s). It is first necessary to determine the lethal potency of
the venom using the LD50 assay. The exact volume of antivenom, or the venom/
antivenom ratio, required to neutralize venom lethality can then be determined
using the antivenom effective dose (ED50 ) assay.
Preclinical testing of antivenom is required:
■■ for routine quality control of efficacy of each newly manufactured
batch of an existing antivenom;
■■ to test the ability of a new antivenom to neutralize the lethal effects
of venoms from snakes from the country or region where it is going
to be introduced;
316
Annex 5
■■ to test the ability of an existing antivenom to neutralize the lethal

effects of venoms for which no prior ED50 data exist (for example,
prior to introducing an antivenom to treat envenoming in a new
geographical region or country).
The outputs of these tests provide globally applicable standard metrics
of venom lethality and antivenom efficacy. This enables internal/manufacturer
monitoring and external/independent auditing of antivenom efficacy – thereby
providing manufacturers and NRAs worldwide with a standardized mechanism
for preventing the distribution of dangerously ineffective antivenom.
Consistent use of venom standards and outbred strains of mice of a
defined weight range (for example, 18–20 g), are recommended for all the assays.
Some producers use other test animals, such as guinea-pigs. While weights will
clearly vary between animal species, the following principles, specified for mice,
will still apply to these alternative test animals. It should be borne in mind that
there are variations in the susceptibility of various strains of mice to the lethal
effect of venoms.
19.2.1 LD50 range-finding test

For venoms whose LD50 is unknown, it is recommended that a range dose-
finding study, using one mouse per venom dose, is performed to set a narrow
range of dose parameters for the full LD50 test – reducing the total number
of animals required. Range-finding tests are not required for venoms from
validated suppliers where the lethal potency across successive batches of venom
is established. If venom is sourced from a new supplier, re-establishment of LD50
should be performed.
Various venom doses are prepared using saline as diluent, and aliquots
of a precise volume (maximum 0.2 mL) of each dose are injected, using one
mouse per dose, by the intravenous route, in the tail vein (or, alternatively, by
the intraperitoneal route (using injection volumes of maximum 0.5 mL)). Deaths
are recorded at 24 hours (intravenous test) or at 48 hours (intraperitoneal test).
On the basis of this preliminary dose-finding experiment, a range of venom
doses causing 0% to 100% lethality is established and thus narrows the range of
venom doses required for the full venom LD50 assay.
19.2.2 The LD50 assay

Venom doses are prepared in saline and intravenously injected (maximum
0.2 mL) into the tail vein of groups of 5–6 mice (of a defined weight range).
A group size of five mice is the smallest number recommended for obtaining
a statistically significant result. In some laboratories the LD50 is estimated by
the intraperitoneal route using an injection volume of a maximum of 0.5 mL.
317
Deaths are recorded at 24 hours (for assays involving intravenous injections) or

at 48 hours (when intraperitoneal injections are used), and LD50 is estimated by
Probit analysis (141), Spearman-Karber (11) or alternative procedures (such as
non-parametric methods). One venom LD50 dose is defined as the amount of
venom causing death in 50% of injected mice.
19.2.3 Antivenom efficacy assessment

The venom LD50 results provide the information necessary to test the venom-
neutralizing efficacy of an antivenom – using the median effective dose (ED50 )
assay. It is important that the venoms used in the ED50 assays are from the
same batch (lot) as that used to determine the venom LD50 result. It is equally
important that all the LD50 and ED50 assays utilize mice of identical strain
and weight.
19.2.3.1 ED50 range-finding test

For an antivenom whose ED50 against a specific venom is unknown, it is
recommended that a range dose-finding study, using one mouse per venom
dose, is performed to set a narrow range of dose parameters for the full ED50
test – reducing the total number of animals required. Range-finding tests are not
required for an antivenom whose venom-neutralizing efficacy is established.
The selected multiple of the venom LD50 (3–6 LD50) is mixed with
different doses of antivenom, incubated at 37 °C for 30 minutes, and each
mixture then intravenously injected into a single mouse. This preliminary test
establishes a range of antivenom volumes that result in 100% survival and 100%
death of the injected mice and thus narrows the range of doses required for the
formal antivenom ED50 test.
19.2.3.2 The median effective dose (ED50 ) assay

A fixed amount of venom (“challenge dose”, usually corresponding to 3–6 LD50 )

is mixed with various volumes of the antivenom and adjusted to a constant final
volume with saline (3, 142, 143). The mixtures are incubated for 30 minutes at
37 °C, and then aliquots of a precise volume (maximum 0.2 mL intravenously;
maximum 0.5 mL intraperitoneally) of each mixture are injected into groups
of 5–6 mice10 (of the same strain and weight range used for the LD 50 assay).
A control group injected with a mixture of the venom “challenge dose” with saline
solution alone (no antivenom) should be included to confirm that the venom
“challenge dose” induces 100% lethality. Centrifugation of the antivenom–venom
Ten mice may be needed for some venoms.

10
318
Annex 5
mixtures before use is not recommended because residual venom toxicity may
remain in the immuno-precipitate.
After injection, deaths are recorded at 24 hours (intravenous test) or at
48 hours (intraperitoneal test) and the results analysed using Probit analysis,
Spearman-Karber or alternative procedures (such as non-parametric methods).
One antivenom ED50 dose is defined as the amount of antivenom, or the venom/
antivenom ratio, resulting in the survival of 50% of mice injected with a mixture
of antivenom and a lethal quantity of venom.
The ED50 result can be expressed in various ways:
■■ mg of venom neutralized by mL of antivenom;
■■ µL antivenom required to neutralize the “challenge dose” of venom
used;
■■ µL of antivenom required to neutralize 1 mg of venom.
The practice of defining the ED50 by the number of murine LD50 s of
venom neutralized per mL of antivenom is inaccurate and has little clinical
usefulness. Since LD50 values for the same venom may vary from one
manufacturer to another, this representation of ED50 should be avoided in
favour of one of the approaches listed above.
19.2.4 General recommendations

Before any antivenom is used therapeutically in humans, its efficacy against the
relevant snake venoms should be confirmed in the essential preclinical LD50 and
ED50 assays. Where minimum standards for venom-neutralizing efficacy exist
in geographically relevant pharmacopoeias, or have been established by NRAs,
these requirements must be met.
In some regions, no minimum acceptable levels of therapeutic efficacy
that are clinically relevant to human envenoming have been established that take
into account the need to deliver a therapeutic dose of antivenom in a realistic
volume for administration. In such cases, NRAs in consultation with other
organizations should establish such standards as a matter of priority for the
various antivenoms produced or distributed in these jurisdictions.
The essential tests of preclinical efficacy, the venom LD50 and antivenom
ED50 , should be standardized by NRAs and national quality control laboratories,
and common protocols adopted to avoid variation in methodology between
production facilities. Therefore, manufacturers should disclose details of their
ED50 protocol to the corresponding NRA as part of the licensing or registration
application to demonstrate compliance.
Quality control laboratories need to establish national reference venom
collections (venoms representing the taxonomic and geographical range of
319
snake species in a country), and these must periodically be independently

evaluated to ensure that they have not deteriorated (see section 10 on quality
control of venoms).
19.3 Supplementary preclinical assays to measure antivenom

neutralization of specific venom-induced pathologies
Snake venoms generate a wide range of systemic pathologies, including a variety
of haemostasis-disruptive (haemorrhage, pro- and anti-coagulopathic effects),
neurotoxic, myotoxic, nephrotoxic and cardiac effects. Supplementary tests are
therefore recommended for new antivenoms and for new applications of existing
antivenoms to determine whether they are effective in eliminating the most
clinically relevant pathophysiological effects induced by the specific venom(s)
of interest.
For example, a new antivenom developed against Echis ocellatus
envenoming should, in addition to preclinical LD50 and ED50 testing, be tested
for its ability to eliminate venom-induced coagulopathy and haemorrhage – the
most medically important effects of envenoming by E. ocellatus.
In this context it may be useful to consider that postmortem observations
of mice from LD50 and associated ED50 experiments can provide a wealth of
pathophysiological information as to antivenom neutralization of venom-
induced pathology. Postmortem observations from LD50 and associated ED50
experiments may prove useful in reducing the need for and frequency of some
of the supplementary assays recommended here; however, their use requires
the same degree of scientific and procedural validation as other supplementary
preclinical assays.
These supplementary preclinical tests are outlined below.
Neutralization of venom haemorrhagic activity

19.3.1
Many venoms, especially those of vipers, exert powerful local and systemic
haemorrhagic activity effected primarily by snake venom zinc-dependent
metalloproteinases. These enzymes damage the basement membrane that
surrounds the endothelial cells of capillaries resulting in bleeding into the
tissues. Bleeding into the brain and other major organs is considered to be the
major lethal effect of envenoming by many viperid species (144). The minimum
haemorrhagic dose of a venom (MHD) quantifies this venom-induced pathology,
and is defined as the amount of venom (in µg dry weight) which, when injected
intradermally, induces in mice a 10 mm haemorrhagic lesion after a predefined
time interval, usually 2–3 hours, after injection (145, 146).
The venom MHD test is carried out by preparing aliquots of 50 µL of
physiological saline solution containing a range of venom doses. Mice (18–20 g
320
Annex 5
body weight; five mice per group) are held securely and the hair surrounding
the injection site is shaved. Venom solutions (50 µL) are injected intradermally
into the shaved skin of lightly anaesthetized mice. After a defined time interval
(usually 2–3 hours), mice are killed using an approved humane procedure, the
area of the injected skin is removed, and the size of the haemorrhagic lesion
in the inner side of the skin is measured using calipers in two directions with
background illumination. Care should be taken not to stretch the skin. The mean
diameter of the haemorrhagic lesion is calculated for each venom dose and the
MHD estimated by plotting mean lesion diameter against venom dose and
reading off the dose corresponding to a 10 mm diameter (145, 146).
The assay measuring the efficacy of antivenom to neutralize venom-
induced haemorrhage is termed the MHD-median effective dose (MHD50 ),
and is defined as the volume of antivenom, in microlitres, or the venom/
antivenom ratio, which reduces the diameter of haemorrhagic lesions by 50%
when compared with the diameter of the lesion in animals injected with the
control venom/saline mixture (146). A “challenge dose” of venom is selected –
between one and five venom MHDs have been used as the challenge dose by
different laboratories. The test is carried out as above, using five mice per group.
Mixtures of a fixed amount of venom and various dilutions of antivenom are
prepared so that the challenge dose of venom is contained in 50 µL. Controls
must include venom solutions incubated with physiological saline solution
alone. Mixtures are incubated at 37 °C for 30 minutes, and aliquots of 50 µL
are injected intradermally in lightly anaesthetized mice. The diameter of
haemorrhagic lesions is quantified as described above, and the neutralizing
ability of antivenom is expressed as the MHD50 .
19.3.2 Neutralization of venom necrotizing activity

Venom-induced local dermonecrosis is a major problem in humans bitten by
snakes and it has long been considered important to have an assay system to
evaluate the effect of an antivenom on this pathology. However, the value of
antivenoms in overcoming the cytolytic effects of venoms has not yet been
established; indeed, there is considerable doubt as to whether antivenom is
useful in obviating such effects in human victims of snake-bite. This is because
venom-induced dermonecrosis occurs quickly after a bite and there is usually
a considerable delay between the envenoming of a victim and his or her arrival
in hospital for treatment. Consequently, antivenom therapy can have little or
no effect in reversing the damage (147, 148). Animal experiments in which
the antivenom was administered to the test animal at different times after
the venom support this opinion (148–150). The minimum necrotizing dose
(MND) of a venom is defined as the smallest amount of venom (in µg dry weight)
which, when injected intradermally into groups of five lightly anaesthetized
321
mice (18–20 g body weight), results in a necrotic lesion of 5 mm diameter 3 days
later. The method used is the same as that for the MHD, except that the skin is
examined 3 days after the intradermal injection of the venom (145).
The assay measuring the ability of an antivenom to neutralize venom-
induced dermonecrosis is termed the MND-median effective dose (MND50 ), and
is defined as the volume of antivenom, in microlitres or the venom/antivenom
ratio, which reduces the diameter of necrotic lesions by 50% when compared
with the diameter of the lesion in mice injected with the control venom/saline
mixture. A challenge dose of venom is selected, usually between one and two
MNDs. The test is carried out as above, using five mice per group. Mixtures of a
fixed concentration of venom and various dilutions of antivenom are prepared
so that the venom challenge dose is contained in 50 µL. Controls include venom
solutions incubated with physiological saline solution alone. Mixtures are
incubated at 37 °C for 30 minutes, and aliquots of 50 µL are injected intradermally
in lightly anaesthetized mice (151, 152). The diameter of dermonecrotic lesions
is quantified 3 days post-injection, as described above, and the neutralization by
antivenom, expressed as the MND50 .
19.3.3 Neutralization of venom procoagulant effect

Many venoms, especially from some vipers, cause consumption of coagulation
factors, which results in incoagulable blood. This, combined with the
haemorrhagic nature of some of these venoms, can result in a very poor prognosis
for severely envenomed patients. Simple in vitro methods exist to measure this
venom-induced pathophysiological effect and the ability of an antivenom to
eliminate it. The minimum coagulant dose (MCD) of a venom is defined as the
smallest amount of venom (in mg dry weight per litre of test solution or µg/mL)
that clots either a solution of bovine fibrinogen (2.0 g/L) in 60 seconds at 37 °C
(MCD-F) and/or a standard citrated solution of human plasma (fibrinogen
content 2.8 g/L) under the same conditions (MCD-P).

For measurement of the MCD-F, 50 µL of physiological saline with
final venom concentrations ranging from 240 to 0.5 mg/L is added to 0.2 mL
of bovine fibrinogen solution (2.0 g/L) at 37 °C in new glass clotting tubes. The
solutions are mixed thoroughly and the clotting time recorded. The MCD-P is
estimated by adding the same venom concentrations to 0.2 mL of the standard
citrated human plasma solution under identical conditions and recording the
clotting time. In each case, the MCD is calculated by plotting clotting time
against venom concentration and reading off the level at the 60 second clotting
time (145).
To estimate the ability of an antivenom to neutralize venom procoagulant
activity, a challenge dose of venom is selected, which corresponds to one MCD-P
or one MCD-F. Mixtures of a fixed concentration of venom and various dilutions
322
Annex 5
of antivenom are prepared so that the challenge dose of venom is contained in

50 µL. Controls include venom solutions incubated with physiological saline
solution alone. Mixtures are incubated at 37 °C for 30 minutes, and aliquots
of 50 µL are added to 0.2 mL of plasma or fibrinogen solution, as described.
The formation or absence of clots is observed during a maximum period of
30 minutes. The minimum volume of antivenom, or venom/antivenom ratio that
completely prevents clotting is estimated and is known either as the MCD-F-
effective dose (MCD-F100 ) or MCD-P-effective dose (MCD-P100 ).
19.3.4 Neutralization of in vivo venom defibrinogenating activity

This test is a direct measure of the in vivo defibrinogenating effect of certain
venoms. To measure the minimum venom defibrinogenating dose (MDD), a
wide range of venom doses is selected and each dose, in a volume of 0.2 mL,
is injected intravenously into five mice (18–20 g body weight). One hour after
injection, the mice are placed under terminal general anaesthesia and bled by
cardiac puncture. The blood from each animal is placed in a new glass clotting
tube, left at room temperature for 1 hour and the presence or absence of a clot
recorded by gently tilting the tube. The MDD is defined as the minimum dose
of venom that produces incoagulable blood in all mice tested within 1 hour of
intravenous injection.
Antivenom neutralization of the venom component(s) responsible for in
vivo defibrinogenation is estimated by incubating a challenge dose of venom,
corresponding to one or more MDD, with different dilutions of the antivenom.
Controls should include venom solutions incubated with saline solution instead
of antivenom. Mixtures are incubated at 37 °C for 30 minutes before injection of
0.2 mL by the intravenous route in groups of five mice (18–20 g body weight).
After 1 hour, mice are bled as described above, and the blood is placed in new
glass clotting tubes and left undisturbed for 1 hour at room temperature, after
which the presence or absence of a clot is recorded. Neutralizing ability of
antivenoms is expressed as MDD-effective dose (MDD100 ), corresponding to the
minimum volume of antivenom, or venom/antivenom ratio at which the blood
samples of all injected mice showed clot formation (152, 153).
19.3.5 Neutralization of venom myotoxic activity

The presence of myotoxic components in a venom results in the degeneration of
skeletal muscle by myonecrosis of muscle fibres. Damage is characterized by the
disruption of the muscle cell plasma membranes, myofilament hypercontraction,
local infiltration of inflammatory cells and oedema. Myotoxicity is characterized
by the appearance of myoglobin in urine and by increments in the serum levels of
muscle-derived enzymes, such as creatine kinase (CK). Myotoxic phospholipase
A2 (PLA2) enzymes are found in a wide range of snake venoms. Some of these
323
PLA2s may be primarily myotoxic, or neurotoxic or both. Cytotoxic proteins

of the three-finger toxin family present in some elapid venoms also cause
myonecrosis. In addition, myotoxicity may occur as a consequence of ischaemia
induced in muscle fibres by the effect of haemorrhagic venom components in the
microvasculature (154).
Venom myotoxic activity is determined by injecting mice with various
doses of venom in a constant volume of 50 µL (using saline solution as diluent)
into the right gastrocnemius muscle. Groups of five animals of 18–20 g body
weight are used per dose. Control animals are injected with the same volume
of saline solution. Tail-snip blood samples are collected after a specified time
interval (3 hours in mice), and the CK activity of serum or plasma is determined
using commercially available diagnostic kits (155, 156). Myotoxic activity is
expressed as the minimum myotoxic dose (MMD), defined as the amount of
venom that induces an increment in serum or plasma CK activity corresponding
to four times the activity in serum or plasma of animals injected with saline
solution alone. Myotoxicity can also be assessed by histological evaluation of
muscle damage after venom injection, although this is a more expensive and
more time consuming method than the CK determination.
To estimate the ability of an antivenom to neutralize venom myotoxicity,
a challenge dose of venom is selected, which corresponds to 3 MMDs. The test is
carried out as above, using five mice per group. Mixtures of a fixed concentration
of venom and various dilutions of antivenom are prepared so that the challenge
dose of venom is contained in 50 µL. Controls include venom solutions incubated
with physiological saline solution alone. Mixtures are incubated at 37 °C for
30 minutes, and aliquots of 50 µL are injected into the gastrocnemius muscle, as
described above. Blood samples are collected 3 hours after injection (in the case
of mice) and serum or plasma CK activity is quantified. The neutralizing ability
of antivenom, expressed as MMD-median effective dose (MMD50 ) is estimated
as the volume of antivenom in microlitres, or the venom/antivenom ratio, which

reduces the serum or plasma CK activity by 50% when compared to the activity
in animals injected with venom incubated with saline solution only (143).
19.3.6 Neutralization of venom neurotoxic activity

Several laboratory methods for assessing venom-induced neurotoxicity have
been developed (for example, chick biventer cervicis nerve-muscle preparation
(157, 158) and the mouse hemidiaphragm phrenic nerve preparation (2,
159–162). However, they are difficult to perform, require costly equipment
and technological expertise and are unlikely to be practicable for most
antivenom producers. Mouse lethality tests are usually reliable in predicting the
neutralization of neurotoxic effects of venoms.
324
Annex 5
19.4 Limitations of preclinical assays

It is acknowledged that the in vivo and in vitro essential and supplementary
preclinical tests have physiological limitations. Venom and venom/antivenom
injection protocols do not represent the natural situation, and the physiological
responses of rodents to envenoming and treatment may differ from those
of humans. Even comparing the levels of immune recognition gathered from
antivenomic or ELISA data with the in vivo neutralization capacity of an
antivenom, is not straightforward. Such limitations make the rodent model of
human envenoming and treatment less than ideal. Care should therefore be
taken to avoid simplistic extrapolations from this assay to the clinical situation.
Nevertheless, the LD50 and ED50 tests represent the methods most widely used
for assessment of antivenom potency, and a number of clinical trials have
demonstrated that the ED50 test is useful (2, 163), but not infallible (164, 165), at
predicting the efficacy of antivenoms in the clinical setting. In some cases, it is
recommended to test the ability of antivenoms to neutralize pathophysiologically
relevant effects other than the lethal effect. Examples include the neutralization
of haemorrhagic and in vitro coagulant effects in the case of Echis sp. venoms,
and of dermonecrotizing effect in the case of cytotoxic Naja sp. venoms. An
additional value of these tests is the assurance that antivenoms are manufactured
with an accepted, quantifiable and uniform neutralizing potency.

■■ The estimation of the ability of an antivenom to neutralize the
lethal activity of venom(s) (LD50 and ED50 ) is a critical preclinical
assessment and should be performed by manufacturers for all
antivenoms, and enforced by the NRAs as part of the antivenom
licensing procedure.
■■ All practitioners of these preclinical tests must prioritize the
implementation of 3R to reduce the substantial number of mice
used, and their collective pain, harm and distress.
■■ In vitro methods such as ELISA, antivenomics or other emerging
technologies that enable antivenoms to be screened for immune
recognition of venom components prior to in vivo evaluation
should be adopted by manufacturers.
■■ The development of improved in vivo assay protocols to reduce
pain and suffering of animals, such as routine use of opioids or
other analgesics, and of in vitro alternatives to the in vivo assays
to reduce the number of animals used in preclinical testing, is
encouraged. The results of any modified in vivo, or new in vitro
325
protocols, should be rigorously compared with results from

existing protocols and validated to ensure statistical reliability
of the newly developed methods.
■■ All new antivenoms, as well as existing antivenoms to be used in
new geographical areas, should furthermore be assessed for their
ability to eliminate specific pathologies caused by the venoms of
the snakes for which the antivenom has been designed.
■■ The selection of which preclinical supplementary test(s) to
perform will depend on the predominant pathophysiological
effects induced by the specific snake venom and be appropriately
adapted for each antivenom. These supplementary tests are not
required for quality control assessment of subsequent batches
of antivenom.
20. Clinical assessment of antivenoms

20.1 Introduction
Antivenoms are unusual among pharmaceutical agents in that they have been
used in human patients since 1896 with little attention being paid to clinical
trials of their effectiveness and safety. However, since the 1970s it has been
clearly demonstrated that it is possible to carry out dose-finding and randomized
controlled trials (RCTs) in human victims of snake-bite envenoming. These
studies have yielded invaluable information, as in the case of clinical trials of
other therapeutic agents for which clinical trials are generally regarded as the
essential basis for regulatory approval.
The conduct of clinical studies is guided by the principles set down in
the international regulations governing good clinical practice,11 comprising
European Union, United Kingdom and USA regulations, summarized in ICH

Topic E 6 (R1) Guideline for good clinical practice (GCP), an international
ethical and scientific quality standard for designing, conducting, recording
and reporting trials that involve the participation of human subjects.12 These
principles emphasize the responsibilities of the researcher and of the organization
sponsoring the research to protect participants in the research and to ensure that
the conduct of the trial is likely to lead to reliable results.
11
See http://www.therqa.com/committees-working-parties/good-clinical-practice/regulations-and-
guidelines/, accessed 23 April 2017.
12
http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2009/09/WC500002874.
pdf, accessed 23 April 2017.
326
Annex 5
Clinical trials should be registered with an appropriate registration body,

before they commence.13
The conventional pathway for clinical evaluation of new therapeutic
products is:
■■ Phase I: healthy volunteer studies – detection of unanticipated
adverse events;
■■ Phase II: limited effectiveness and safety studies, often dose-finding;
■■ Phase III: full-scale clinical evaluation, often using blinded RCTs to
avoid potential introduction of bias;
■■ Phase IV: post-marketing surveillance.
The appropriateness of this pathway for antivenoms depends upon a
number of factors, including whether an antivenom is new or has been previously
used in human patients, the ethical basis for the study, the trial’s practicability as
well as ethical and national regulatory considerations. So far, most antivenoms
have been registered without prior clinical studies. This situation should not
persist: it is desirable, first, to collect the existing clinical data on antivenoms
already marketed, and second to promote Phase II or III clinical trials before
registering new antivenoms. In the absence of clinical data for antivenoms
already in use, appropriate clinical trials should be quickly implemented.
20.1.1 Identification of biting species in clinical studies of antivenoms

It is absolutely essential that all clinical studies of antivenom effectiveness or
safety, including clinical trials incorporate robust methodologies for ensuring
the identification of the biting species. This can be achieved through:
■■ expert identification of the dead snake responsible for the bite,
or of a photographic image of that snake; or
■■ the identification of specific venom components unique to
particular species (from bite site swabs, wound exudate, serum
or urine samples) through the use of EIAs (ELISA) or other
immunological methods.
Failure to properly identify the species of snakes that are responsible
for cases of envenoming included in clinical trials and other studies of snake
antivenoms significantly diminishes the value of the research, and renders the
https://clinicaltrials.gov/ct2/manage-recs/how-register; http://www.isrctn.com/;
13
https://www.clinicaltrialsregister.eu/; http://www.umin.ac.jp/ctr/; http://www.anzctr.org.au/

327
results unreliable. NRAs should be cautious about accepting the results of

clinical evaluations of antivenom where robust, reliable identification of the
biting species is not available.
20.1.2 Phase I studies

Conventional clinical studies using healthy volunteers are not appropriate in
the case of antivenoms14 because of the risk of anaphylactic and other reactions
(for example, pyrogenic or serum sickness and, rarely, hypersensitivity reactions
to equine or ovine plasma proteins) and the risk of sensitization to equine or
ovine plasma proteins in the volunteers. Phase I studies are primarily designed to
detect unanticipated adverse reactions. This can be done only in human subjects
as it is not possible in an animal model. Such studies are an essential protection
against severe and even fatal effects of a new medication, before it is tested in
the much larger numbers of subjects demanded for Phase II and III studies.
Recent disasters or near disasters during Phase I studies of new therapeutic
monoclonal antibodies have emphasized not only the need for such studies
but also their potential dangers. A similar situation exists in the early testing of
cytotoxic drugs and antibodies used in oncology. A preliminary open-label dose-
finding study can establish both the effectiveness and safety of an initial dose of
a new antivenom in small groups of adult, non-pregnant patients with systemic
envenoming, but excluding those with features of severe envenoming. The aim is
to assess clinical safety and effectiveness, as a prelude to full-scale Phase II or III
RCTs. This modified Phase I approach can be combined with a preliminary dose-
finding study. The “3 + 3” dose-escalation design (166) can be used to determine
the minimum dose capable of achieving a defined end-point in two-thirds or
more of a small group of patients. An additional stopping rule is added to ensure
that patients are not exposed to doses likely to cause reactions in more than
one‑third of them – for details see Abubakar et al., 2010 (94). Currently, there is
no alternative for ethical Phase I studies.
20.1.3 Phase II and III studies

Phase II studies are usually conducted to optimize doses, establish or confirm
the relative safety of a product and give an indication of effectiveness. Phase III
studies are normally used to establish effectiveness of a product, often in
comparison with an existing product, or occasionally a placebo. Since antivenoms
are so well established in the treatment of snake-bite envenoming, the use of
placebo controls is ethically acceptable only where there is genuine uncertainty
Immunoglobulins derived from animal plasma.

14
328
Annex 5
about whether the benefit (degree of clinical improvement) from the antivenom
outweighs the risk (potential incidence and severity of adverse events). Depending
on the speed of evolution of envenoming, immediate treatment might be
compared to delayed antivenom treatment. A new antivenom with demonstrable
preclinical potency (see above) can be compared with an established product, or
two markedly different initial doses or regimens of the same antivenom can be
compared. In RCTs, non-inferiority, rather than superiority of a new antivenom
or regimen, compared to an existing treatment, requires smaller numbers of trial
participants to achieve acceptable power (94). Basic requirements for any clinical
antivenom RCT are that the participants should be reasonably homogeneous as
far as the species of snake responsible and their pretreatment characteristics (for
example, interval between bite and treatment) are concerned, and that objective
clinical end-points should be selected to judge effectiveness, and measure rates
of adverse events.
20.1.4 Phase IV studies

Phase IV studies are clinical surveillance studies that occur after market
authorization of the product. In view of the difficulty in performing standard
clinical trials of antivenom in some situations, this may be the only way to
study safety and effectiveness of an antivenom in a large number of patients.
In practice, such studies have rarely if ever been attempted for antivenoms, but
they are strongly recommended for the future.
20.2 Clinical studies of antivenom

Although preclinical testing may be valuable in ensuring that antivenoms
neutralize the venoms of interest, the complex effects of venoms in humans
and the need to consider venom pharmacokinetics mean that, ultimately, the
effectiveness and safety of antivenoms for the treatment of human envenoming
can only be determined by well-designed clinical studies. Clinical studies of
antivenoms primarily address three main issues:
■■ assessment of the optimal initial dose of antivenom;
■■ assessment of effectiveness of the antivenom;
■■ assessment of the safety of an antivenom, particularly the incidence
and severity of early and late reactions.
Antivenom safety and tolerance depend on manufacturing factors
(immunoglobulins composition, purification of immunoglobulin fragments,
protein concentration, and presence of preservatives) (167). Consequently,
incidence and severity of adverse reactions to similar doses of a given batch of
antivenom are unlikely to vary in different geographical locations. Conversely,
329
the effectiveness depends on both manufacturing factors (choice of venoms,

immunological titre) and circumstantial factors (quality and quantity of
inoculated venom, patient’s physical condition, delay of treatment, etc.).
However, following initial preclinical testing, both effectiveness and dose-finding
studies may need to be repeated for a new geographical location, depending
upon the similarity of the snake species in the new location with those where
the antivenom was initially tested. If the species are similar, if preclinical testing
indicates good neutralization, and if evidence of clinical effectiveness has been
reported in other places, post-marketing surveillance studies may be adequate.
20.2.1 Dose-finding studies

Dose-finding studies seek to establish the optimum initial dose of an antivenom
required to control envenoming in patients with different severities of envenoming.
The therapeutic dose of an antivenom administered by the intravenous route
depends on:
■■ the quantity of venom injected (assessed by clinical and laboratory
outcomes);
■■ the neutralizing potency of the antivenom (given by preclinical
tests); and
■■ the dose regimen.
The dose is calculated to neutralize a certain amount of venom and
does not vary between adults and children. Preclinical testing may be used to
estimate starting doses and these dosage regimens may be evaluated in a number
of ways using standard effectiveness and safety end-points. Dose regimens can
be assessed approximately by using prospective observational studies (105).
High-quality observational studies may extend evidence over a wider population
and are particularly useful in defining safety or when RCTs are unethical or
impracticable. In these, the proportion of patients with good clinical outcomes
(for example, restoration of blood coagulability or failure to develop local wound
necrosis) can be observed with different, escalating or de-escalating doses of
antivenom. However, the many weaknesses of observational studies must always
be borne in mind.
As part of the design of the study, it is important to determine the
minimum number of patients required to establish meaningful results by using
sample size calculations (168). Results may sometimes be compared to those of
previous studies (historical controls) to determine how the effectiveness or safety
of a newly introduced antivenom compares with previously used antivenoms
(169). However, such comparisons are susceptible to many kinds of confounding
variables and are potentially unreliable. Subsequently, the minimum dose that
330
Annex 5
appears to be effective can be evaluated in larger Phase II trials or compared to

another antivenom or a different dose in Phase III RCTs.
20.2.2 Randomized controlled trials

Definitive Phase III RCTs may require large numbers of patients because of
considerable individual variation in the clinical manifestation of envenoming
(or the great variability in the quantity and quality of venom injected in different
patients). The new antivenom is compared with the existing standard antivenom
treatment or, if none exists, two different doses of the test antivenom may be
compared. Placebo controls are rarely justified unless there is genuine uncertainty
about the risk and benefits of antivenom treatment. In this situation, as a
safeguard against unnecessary morbidity in either treatment group, a restricted
sequential plan might be incorporated (170) which allows evaluation of results
as the trial progresses, as in the early trials of therapeutic tetanus antitoxin (171).
To avoid bias, patients should be randomly allocated to the groups and
the study should be blinded, at a minimum to those research personnel who are
assessing the clinical response and ideally to both investigators and participants.
The number of patients required in each trial arm should be calculated to
give the study sufficient statistical power. These power calculations are based
on the expected difference in outcome between the treatment groups if the
study is designed to demonstrate superiority of one treatment over another.
Alternatively, predefined limits of the acceptable performance compared to
an existing product are set if the trial is designed to demonstrate that the new
antivenom is not worse than existing products (non-inferiority). All patients
enrolled in an RCT and randomly allocated to treatment should be included
in the analysis of results according to the principle of “intention to treat”, so
that any deleterious effects of an antivenom are not concealed by the recipients
dropping out of the trial.
20.2.3 Effectiveness end-points for antivenom trials

The assessment criteria (end-points) used for antivenom studies should be
predefined a priori and objective. They may be clinical or assessed by laboratory
investigations. Common end-points include mortality, development of local tissue
effects of envenoming such as necrosis, time taken to restore blood coagulability
(assessed by the 20 minute whole blood clotting test) (172), other laboratory
parameters such as the prothrombin time, halting of bleeding or objective clinical
improvement in neurotoxicity. Surrogate markers such as platelet count are less
suitable as they may be affected by complement activation resulting from the
antivenom treatment itself. Patients should be observed carefully for long enough
to reveal evidence of recurrent envenoming (seen particularly with short half-life
Fab antivenoms) (173).
331
However, owing to the high variability of the mode of action of venoms

and of the individual patient’s responses, as well as the diagnostic capacity of
health centres, particularly in developing countries, it is necessary to promote
clinical research to identify appropriate clinical and laboratory criteria.
20.2.4 Safety end-points for antivenom trials

Because antivenoms consist of foreign proteins/fragments that are liable to
aggregation, adverse effects are an inevitable risk in therapy. Appropriate
manufacturing steps can reduce the rate of adverse reactions. Rates of reaction
are correlated with the purity of the antivenom product and the amount of
protein infused. Continuous clinical observation at the bedside is necessary
for several hours after treatment to detect acute reactions; late adverse
reactions may occur several weeks later. Accurate reaction rates can only
be assessed prospectively. Reaction rates may differ considerably between
different antivenoms, but in most cases only a small proportion of reactions
are life-threatening. Although there is no consensus on classifying or grading
early adverse reactions, studies should aim to detect both early adverse events
(anaphylaxis and pyrogenicity) occurring at the time of, or within 24 hours
of, antivenom administration (such as urticaria itching, fever, hypotension or
bronchospasm) and late reactions such as serum sickness occurring between
5 and 24 days after antivenom administration (for example, fever, urticaria,
arthralgia, lymphadenopathy, proteinuria or neuropathy).
20.2.5 Challenges in clinical testing of antivenoms

Several particular features of snake-bite make clinical testing of antivenoms
challenging. These features include the large variation in the consequences of
envenoming between individuals that make it necessary to study large numbers
of patients, difficulties in identification of the species responsible for envenoming
and the inaccessibility and logistical challenges of areas where snake-bite is

sufficiently common to provide sufficient numbers of patients to study. Clinical
studies may also be expensive, particularly multicentre studies, with the attendant
additional complexity and logistics of between-centre variations. However,
despite these difficulties, a number of RCTs have been undertaken and published
since 1974 (89, 93, 104, 172, 174–178).
20.3 Post-marketing surveillance

Phase IV studies may be of much greater importance for antivenoms than for
other products. A period of active post-licensing surveillance should follow:
■■ the introduction of a new antivenom (often a regulatory
requirement);
332
Annex 5
■■ the introduction of changes in manufacturing processes or in the

use of raw materials (for example, switching from the use of venoms
produced in wild-caught snakes to venoms from captive specimens
in a serpentarium), which may result in changes in the quality or
effectiveness of an established antivenom;15
■■ the introduction of an established antivenom into a new
geographical area.
Post-marketing studies of antivenoms examine effectiveness as well as the
frequency of immediate or delayed side-effects. The combination of preclinical
testing and post-marketing surveillance studies is a minimum acceptable clinical
evaluation when an existing antivenom is used in a new region.
20.3.1 Possible approaches

Passive surveillance is currently practised by some antivenom manufacturers.
However, approaches that rely upon voluntary return of questionnaires about
safety and effectiveness are unlikely to provide the high quality data that are
necessary. There are three potential approaches to obtaining such data as
outlined below.
20.3.1.1 National or regional system for post-marketing surveillance

Countries using antivenoms should establish a national or regional system for the
post-marketing surveillance of antivenoms. Clinicians and health workers (such
as those working in poison centres) should be encouraged to report actively to
national control authorities and manufacturers any unexpected lack of clinical
effectiveness and any adverse reactions. These should include both early adverse
events, occurring at the time of, or within 24 hours of, antivenom administration,
and late reactions occurring between 5 and 24 days later. The mechanism for
reporting (such as the use of standardized forms), the receiving body (for example,
the national control authority), the deadline for reporting and the type of adverse
events that are reportable need to be clearly defined by the authority and will
depend on its structure and resources. The manufacturer of the antivenom
and the authorities should assess these reports and, in consultation with one
another and with specialists in the field, attempt to evaluate their significance.
This assessment may require the testing of products already released and the
Major changes in the design, manufacturing process, or source of venoms used for production of
15
antivenom may necessitate new preclinical and clinical trials of a product. Such changes may also have
licensing implications depending on the legislated regulations in the country of manufacture, or the
countries where the product will be marketed and used.
333
inspection of production and control facilities and local distribution channels.

If an imported product is associated with adverse reactions, the manufacturer
and the national control authorities both in the country of distribution and the
country of origin should be notified.
20.3.1.2 Observational studies

In certain situations, for example, the first use of an established antivenom in
a new geographical area, or when routine surveillance has identified safety or
effectiveness concerns, there is a rationale for setting up observational studies to
ensure adequate effectiveness and safety. In the case of first use of an established
antivenom in a new geographical area, such studies should follow preclinical
testing that ensures neutralization of locally important venoms. Observational
studies should carefully document the clinical responses to antivenom, the
clinical outcomes and the frequency of reactions in a substantial cohort of
patients (179). However, owing to inherent weaknesses of these non-randomized
trials, results of observational studies may be misleading.
20.3.1.3 Sentinel sites

In some settings, where post-marketing surveillance of the whole of a country
may be problematic, the use of sentinel sites may allow focusing of limited
resources to maximize surveillance effectiveness.
20.3.2 Responses to results of post-marketing studies

High-quality post-marketing studies will allow clinicians, public health officials
and manufacturers to identify antivenoms with poor effectiveness (batch
variations in potency and safety), instances of incorrect use and dosage of
antivenoms, and serious safety issues arising from the use of antivenoms. In some
situations, these issues may be addressed by improving training of staff in the

management of snake-bite, but these studies may also allow identification of
the use of an inappropriate antivenom (180).

■■ Preclinical and clinical testing of antivenoms has been largely
neglected in the past. Despite challenges, clinical trials of
antivenoms in human patients have proved feasible and useful. As
far as possible, trials should adhere to the principles of WHO and
ICH GCP and should measure robust, objective end-points.
■■ NRAs should expect producers either to provide data confirming
the clinical effectiveness and safety of their antivenoms against
334
Annex 5
envenoming by local species of venomous snakes or, to support

in-country clinical testing of these products.
■■ Incorporating robust methodologies for reliable identification of
the biting snake species is absolutely essential to the design of all
clinical trials and other clinical studies of antivenoms.
■■ Prospective observational studies are of some use in monitoring
the effectiveness and safety of an antivenom when first used in a
new geographical region.
■■ Post-marketing surveillance studies should play a major role in
the evaluation of effectiveness and safety of antivenoms.
21. Role of national regulatory authorities

NRAs or medicines regulatory authorities play a crucial role in ensuring that
pharmaceuticals, vaccines, biological and other medicinal products that are
available for use in a country have been carefully and thoroughly evaluated
against internationally recognized standards of safety and quality. These
agencies of government are vital to the process of strengthening health systems
by providing regulatory controls based on legislative frameworks and technical
expertise. NRAs therefore have a pivotal role in ensuring the quality, safety and
efficacy of antivenoms.
WHO Guidelines for national authorities on quality assurance for
biological products and on regulation and licensing of biological products in
countries with newly developing regulatory authorities (181, 182) state that
NRAs should ensure that available biological products, whether imported or
manufactured locally, are of good quality, safe and efficacious, and should thus
ensure that manufacturers adhere to approved standards regarding quality
assurance and GMP. The responsibilities should also include the enforcement and
implementation of effective national regulations, and the setting of appropriate
standards and control measures. The evaluation and control of the quality, safety
and consistency of production of animal-derived blood products involve the
evaluation of the starting material, production processes and test methods to
characterize batches of the product.
This requires the regulatory authorities to have appropriate expertise.
WHO provides Member States with support in the establishment of NRAs and
with the development of regulatory functions, technical abilities and adoption
of standards and best practice guidelines, such as this document. A model
protocol for the production and testing of snake antivenom immunoglobulins
to assist NRAs in reviewing the quality of antivenom batches is provided in
Appendix 2.
335
21.1 Regulatory evaluation of antivenoms

The regulatory evaluation and control of the quality, safety and consistency
of production of antivenoms is summarized in Fig. A5.10 and involves the
evaluation and approval of:
■■ the preparation of the starting plasma material from immunized
animals (including the preparation of snake venom batches
representative of the venomous animals of the geographical region
the antivenom is made for), and the animal husbandry, control
and traceability of the immunized animals and of the
immunization process;
■■ the fractionation process used to produce the antivenoms;
■■ the test methods used to control batches of the product including
realistic and validated potency tests based on neutralization of likely
maximal envenomation;
■■ shelf-life and stability testing of intermediates and final product;
■■ the preclinical data supporting the expected effectiveness of the
products for treatment of local envenomings;
■■ the clinical effectiveness of locally manufactured or imported
antivenoms against the species of snakes found in the country,
through active marketing surveillance.
21.2 Establishment licensing and site inspections

Many NRAs implement control systems based on licensing manufacturing
establishments, inspecting them regularly, and enforcing the implementation of
the legal requirements and applicable standards. This applies to the preparation
of snake venoms, production of animal hyperimmune plasma for fractionation,
and the manufacturing process of the antivenoms. Establishments involved in

any or all stages of the manufacture of antivenoms need to have an establishment
licence and to be inspected by the competent NRA before operations commence.
To obtain the licence, the establishments have to fulfil a defined set of
requirements to guarantee that their operation ensures the safety, quality and
clinical effectiveness of the antivenoms.
336
Annex 5
Fig. A5.10
Schematic diagram for the regulatory evaluation and control of the quality, safety and
standards of production of antivenoms
Venom producer
specifications
AUDITS
Plasma producer
GMP INSPECTION
specifications
Fractionator of antivenoms
specifications
Antivenom
Post-marketing
surveillance
Antivenom marketing file evaluation and authorization
National regulatory authority
21.3 Impact of good manufacturing practices

Implementing the principles of GMP in the production of therapeutic products
is acknowledged as essential for assuring the quality and safety of biological
medicinal products. For antivenoms, GMP becomes even more important and
more complex due to the biological nature of the production process and the
complexity and local specificities of snake envenoming.
Therefore, taking into account the principles of GMP, and the existence
of an appropriate quality assurance system to address and implement these
requirements at all stages of manufacture, should be pivotal in ensuring the
quality and safety of antivenoms. The following benefits are expected:
■■ ensures the application of quality assurance principles at all steps
involved in the preparation of snake venoms, the production of
animal plasma and the fractionation process of antivenoms;
337
■■ reduces errors and technical problems at all stages of manufacture of

plasma for fractionation and antivenoms;
■■ contributes to the release of products that comply with quality and
safety requirements;
■■ ensures adequate documentation and full traceability of plasma for
fractionation and antivenom production stages;
■■ enables continuous improvement in production of plasma for
fractionation and antivenoms;
■■ provides suitable tools for NRAs to assess the compliance status of a
manufacturer of antivenoms, either local or abroad;
■■ supports regional cooperation networks that may result in the
formation of competence centres by centralizing activities to enable
compliance to be achieved at the required level.
An establishment licensing system for antivenom manufacturers
operated by the responsible and competent NRA should therefore exist. The
main requirements to be met to obtain an establishment licence may include
in particular:
■■ quality assurance system and GMP applied to all steps of venom
and antivenom production;
■■ personnel directly involved in collection, testing, processing, storage
and distribution of antivenoms are appropriately qualified and
provided with timely and relevant training;
■■ adequate premises and equipment are available;
■■ an adequate control system to ensure traceability of antivenoms
manufacture is to be enforced through accurate identification
procedures, record maintenance, and an appropriate labelling system;
■■ the existence of a post-marketing information system.
21.4 Inspections and audit systems in the

production of antivenoms
The ongoing operations of antivenom manufacturers need to be subject to
regulatory authority control and supervision in accordance with legislation.
Regulatory authorities can also make use of WHO technical support services,
monographs and guidelines to assist them in developing the capacity to
undertake ongoing inspection activities, audits and reviews of manufacturing,
quality control and other production process systems.
All manufacturers must have in place a quality assurance system for
manufacture of animal-derived plasma products that comprehensively covers
338
Annex 5
all stages leading to the finished product, from production of plasma (including
venom sourcing and preparation, production animal sourcing, selection,
immunization and animal health control) to the collection and fractionation
of the plasma into the finished products and their control. Manufacturers of
antivenoms must maintain complete Site Master Files containing specific, factual
details of the application of GMP to production and quality control activities that
are undertaken at every site of operations linked to the products they produce.
Manufacturers should also maintain quality manuals that define and describe
the quality system, the scope and operations of the quality system at all levels
of production, management responsibilities, key quality systems processes
and safeguards. For individual products a product dossier in CTD format as
recommended by WHO and ICH may also be required. NRAs should make
full use of these three forms of production documentation in preparing for and
conducting site inspections and audits.
For local producers, the NRA should enforce the implementation of
GMP with the aim of ensuring the compliance of the manufacturer with the
existing provisions. It is the responsibility of the NRA inspector to ensure that
manufacturers adhere to the approved standards of GMP and quality assurance.
The inspection and enforcement of control measures for venom producers,
immune plasma producers, fractionation facilities and final product producers
and distributors should be carried out by officials representing the competent
NRA. They should be familiar with biological product technologies and trained
in GMP inspections.
Inspections should follow common inspection procedures. These include:
■■ an opening meeting with management and key personnel;
■■ a tour of the facility, including inspection of the main areas and
activities, such as:
(a) serpentariums;
(b) animal husbandry practices;
(c) animal identification and suitability for blood or plasma
collection;
(d) process of collection and storage of blood or plasma;
(e) plasma fractionation process;
(f ) testing and availability of test results for venoms, antivenoms
and raw materials;
(g) storage, transportation and shipment; and
(h) quality assurance (including internal audits and change
control);
339
■■ documentation (SOPs, records, donor record files, log books);

personnel and organization;
■■ qualification and process validations;
■■ error and corrective action systems, recalls and complaints and
product quality controls;
■■ a final meeting summarizing the inspection outcome.
A thorough inspection includes the observation of staff during
performance of operations and comparison with established SOPs. The inspection
should not only be considered as checking compliance with GMP, but also as an
indirect product quality assessment by checking product-specific validation and
quality control data.
A written report should summarize the main findings of the inspection
including its scope, a description of the company, the deficiencies listed, specified
and classified (for example, as critical, major or minor), and a conclusion. The
written report is sent to the manufacturer. The manufacturers are requested to
notify the NRA of the specific steps being taken, or which are planned, to correct
the failures and to prevent their recurrence. If necessary, follow-up inspections
should be performed – for example, to check the successful implementation
of specific corrective actions. The NRA should have the mandate to withdraw
an establishment licence in a case where inspection results show critical
noncompliance with the requirements or product specifications. In the procedure
for granting marketing authorization for an antivenom, information on the
collection and control of the venoms and of the starting animal blood or plasma
needs to be documented as part of the dossier. In summary, the enforcement and
implementation of licensing and inspection regulatory systems for antivenoms
constitute fundamental tools to ensure the quality of antivenoms produced or
distributed to treat envenomings in a country.
21.5 Antivenom licensing

All antivenoms that are in use in a country must be approved and licensed by
the appropriate NRA or another competent authority with legal jurisdiction. The
process for applying for, considering and making a decision on the merits of an
application should follow established processes and be subject to transparency
and review. The process of product dossier assessment and review should be
defined by legislation and appropriate regulations. Marketing authorization
(licensing) of antivenoms should be subject to a thorough review of the product
dossier, Site Master File and quality management system. For a product that is to
be imported, NRAs should communicate directly with the licensing authority in
the country of manufacturer to ensure that claims in the documents are factual
and the product meets licensing requirements in its country of origin.
340
Annex 5
21.6 National reference venoms

As discussed in section 10.2 countries or regions should establish collections of
reference venoms (“standard venoms”) against which antivenom products and the
venoms used by manufacturers can be assessed. The establishment of reference
venoms for release control of final product should be reviewed and monitored
by the regulatory authority or by other competent authorities with technical
expertise in the production of international reference material standards.
Antivenom manufacturers should not be involved in the production of reference
standards in order to ensure transparency. The task should be assigned to a
central quality control laboratory, or to a third-party organization with specific
expertise. The potency of each batch of final product should be confirmed by
specific neutralization of a standard venom of each species of snake for which the
antivenom is indicated.
A system of control for the reference venoms and for the design of
the venom pools (for example, the geographical selection of animals) should
be in place as part of the procedures for the management of reference venom
collections.

■■ NRAs should regulate and supervise local antivenom
manufacturers.
■■ NRAs are responsible for market authorization of antivenoms
distributed in the country.
■■ Inspection and audit processes are fundamental to the effective
regulation and control of antivenoms and NRAs should seek
appropriate assistance to develop both the legislative and technical
expertise necessary to undertake these functions.
■■ Only antivenoms that pass stringent applications processes should
be granted market authorization.
■■ National reference collections of “standard venoms” should be
established according to accepted international reference material
standards, and used to independently assess antivenoms, or to
validate venoms used by manufacturers.

This second edition of these WHO Guidelines was revised and redrafted in
2016 under the coordination of Dr C.M. Nuebling, World Health Organization,
Switzerland. Dr D.J. Williams, University of Melbourne, Australia, led the editing
341
of the new draft. Document reviewers included: Dr G. Alcoba, Geneva University

Hospitals, Switzerland; Mrs D. Barr, University of Melbourne, Australia; Dr A.
Britton, Ultimate Efficacy Consulting Pty Ltd, Australia; Dr T. Burnouf, Human
Protein Process Sciences, France; Professor J.J. Calvete, Instituto de Biomedicinia
de Valencia, Spain; Dr F. Chappuis, Geneva University Hospitals, Switzerland;
Dr J-P. Chippaux, Institut de Recherche pour le Développement, Benin; Dr R.
Ferreira, Instituto Butantan, Brazil; Dr C. Graner, Instituto Butantan, Brazil;
Dr K. Grego, Instituto Butantan, Brazil; Professor J.M. Gutiérrez, Instituto
Clodomiro Picado, Costa Rica; Professor A. Habib, Bayero University, Nigeria;
Dr R. Harrision, Liverpool School of Tropical Medicine, England; Dr G. Leon,
Instituto Clodomiro Picado, Costa Rica; Dr J. Marcelino, Instituto Butantan,
Brazil; Dr F. Petitto de Assis, Instituto Butantan, Brazil; Dr K. Ragas, CSL Limited,
Australia; Professor K. Ratanabanangkoon, Mahidol University, Thailand; Dr C.
Rolls, Australia; Dr J. Southern, South Africa; Emeritus Professor D.A. Warrell,
Oxford University, England; Dr F.H. Wen, Instituto Butantan, Brazil; and Dr W.
Wüster, University of Bangor, Wales.
Prior to their presentation to the sixty-seventh meeting of the WHO
Expert Committee on Biological Standardization, held in Geneva, Switzerland,
from 17 to 21 October 2016, the draft Guidelines were published on the
WHO Biologicals website as part of a process of public consultation. The draft
document was also emailed to: all second edition section reviewers; antivenom
manufacturers; regulatory authorities in countries producing antivenoms; WHO
regional offices (for further distribution in each region); and all members of the
Committee, with a request to provide feedback and comments.
In response, submissions were received from: Emeritus Professor D.A.
Warrell, Oxford University, England; Dr L. Bowen, Sanofi US, the USA; Dr G.
Leon, Instituto Clodomiro Picado, Costa Rica; Professor J.M. Gutiérrez, Instituto
Clodomiro Picado, Costa Rica; Dr G. Habib, VACSERA, Egypt; Dr R.H. Harrison,
Liverpool School of Tropical Medicine, England; Dr M. del Pilar Álvarez, Centro

para el Control Estatal de Medicamentos, Equipos y Dispositivos Médicos
(CECMED), Cuba; Dr E. Griffiths, WHO Expert Committee on Biological
Standardization; Dr J. Southern, South Africa; Professor K. Ratanabanangkoon,
Mahidol University, Thailand; Dr D. Garcia, Agence Nationale de Sécurité du
Médicament et des Produits de Santé, France; Dr A. Britton, Ultimate Efficacy
Consulting Pty Ltd, Australia; Dr A. Fernandes, Bharat Serums and Vaccines
Ltd, India and Dr D. Scott and Dr O. Simakova, United States Food and
Drug Administration, Center for Biologics Evaluation and Research, the USA.
The resulting document WHO/BS/2016.2300 was then prepared in light of all
comments received.
342
Annex 5
References
1. WHO Model List of Essential Medicines (15th list). Geneva: World Health Organization; 2007
(http://www.who.int/medicines/publications/08_ENGLISH_indexFINAL_EML15.pdf, accessed 13
February 2017).
2. Sells PG. Animal experimentation in snake venom research and in vitro alternatives. Toxicon.
2003;42(2):115–33.
3. Theakston RD, Warrell DA, Griffiths E. Report of a WHO workshop on the standardization and
control of antivenoms. Toxicon. 2003;41(5):541–57.
4. Gutiérrez JM, Herrera C. The analgesics morphine and tramadol do not alter the acute toxicity
induced by Bothrops asper snake venom in mice. Toxicon. 2014;81:54–7.
5. Von Behring E, Kitasato S. Über das Zustande-kommen der Diphtherie-Immunität und der
Tetanus-Immunität bei Thieren. Deutsche Medizinische Wochenschrif. 1890;16:1113–45.
6. Phisalix C, Bertrand G. Sur la propriété antitoxique du sang des animaux vaccinés contre le venin
de vipère. Comptes Rendus de la Société de Biologie. 1894;46:111–3.
7. Calmette A. Contribution à l’étude du venin des serpents. Immunisation des animaux et
traitement de l’envenimation. Annales de l’Institut Pasteur, VIII, 1894:275–91.
8. Calmette, A. Sur le venin des serpents et sur l’emploi du sérum antivenimeux dans la
thérapeutique des morsures venimeuses chez l’homme et chez les animaux. Annales de l’Institut
Pasteur. 1897;XII:214–37.
9. Pope CG. The action of proteolytic enzymes on the antitoxins and proteins in immune sera. I.
True digestion of the proteins. Br J Exp Pathol. 1939;20:132–49.
10. Pope CG. The action of proteolytic enzymes on the antitoxins and proteins in immmune sera. II.
Heat denaturation after partial enzyme action. Br J Exp Pathol. 1939;20:201–12.
11. Progress in the characterization of venoms and standardization of antivenoms. Geneva: World
Health Organization; 1981.
12. Raw I, Guidolin R, Higashi HG, Kelen EMA. Antivenins in Brazil: Preparation. In: Tu AT, editor.
Handbook of natural toxins. Volume 5: Reptile venoms and toxins. New York: Marcel Dekker;
1991:557–81.
13. Grandgeorge M, Véron JL, Lutsch C, Makula MF, Riffard P, et al. Preparation of improved F(abʹ)2
antivenoms. An example: new polyvalent European viper antivenom (equine). In: Bon C, Goyffon
M, editors. Envenomings and their treatments. Lyons: Fondation Marcel Mérieux; 1996:161–72.
14. Atkinson JP. The fractional precipitation of the globulin and albumin of normal horse’s serum
and diphtheria antitoxic serum, and the antitoxic strength of the precipitates. J Exp Med.
1900;5(1):67–76.
15. Hiss PH, Atkinson JP. Serum-globulin and diphtheric antitoxin – a comparative study of the
amount of globulin in normal and antitoxic sera, and the relation of the globulins to the antitoxic
bodies. J Exp Med. 1900;5(1):47–66.
16. Atkinson JF. A preliminary note on the fractional precipitation of the globulin and albumin
of normal horse’s serum and diphtheric antitoxic serum, and the antitoxic strength of the
precipitates. J Exp Med. 1899;4(5–6):649–54.
17. Avery OT. The distribution of the immune bodies occurring in antipneumococcus serum. J Exp
Med. 1915;21(2):133–45.
18. Gibson RB. The concentration of antitoxin for therapeutic use. J Biol Chem. 1906;1:161–70.
343
19. Porter RR. The hydrolysis of rabbit y-globulin and antibodies with crystalline papain. Biochem J.
1959;73:119–26.
20. Gutiérrez JM, Leon G, Lomonte B. Pharmacokinetic-pharmacodynamic relationships of
immunoglobulin therapy for envenomation. Clin Pharmacokinet. 2003;42(8):721–41.
21. Rabies and envenomings: a neglected public health issue. Report of a consultative meeting.
Geneva: World Health Organization; 2007 (http://www.who.int/bloodproducts/animal_sera/
Rabies.pdf, accessed 13 February 2017).
22. Swaroop S, Grab B. Snakebite mortality in the world. Bull World Health Organ. 1954;10(1):35–76.
23. Chippaux JP. Snake-bites: appraisal of the global situation. Bull World Health Organ.
1998;76(5):515–24.
24. Snow RW, Bronzan R, Roques T, Nyamawi C, Murphy S, Marsh K. The prevalence and morbidity of
snake bite and treatment-seeking behaviour among a rural Kenyan population. Ann Trop Med
Parasitol. 1994;88(6):665–71.
25. Fox S, Rathuwithana AC, Kasturiratne A, Lalloo DG, de Silva HJ. Underestimation of snakebite
mortality by hospital statistics in the Monaragala District of Sri Lanka. Trans R Soc Trop Med Hyg.
2006;100(7):693–5.
26. Hati AK, Mandal M, De MK, Mukherjee H, Hati RN. Epidemiology of snake bite in the district of
Burdwan, West Bengal. J Indian Med Assoc. 1992;90(6):145–7.
27. Pugh RN, Theakston RD. Incidence and mortality on snake bite in savanna Nigeria. Lancet.
1980;2(8205):1181–3.
28. Sharma SK, Chappuis F, Jha N, Bovier PA, Loutan L, Koirala S. Impact of snake bites and determinants
of fatal outcomes in southeastern Nepal. Am J Trop Med Hyg. 2004;71(2):234–8.
29. Trape JF, Pison G, Guyavarch E, Mane Y. High mortality from snakebite in south-eastern Senegal.
Trans R Soc Trop Med Hyg. 2001;95(4):420–3.
30. Mohapatra B, Warrell DA, Suraweera W, Bhatia P, Dhingra N, Jotkar RM et al. Snakebite mortality in
India: a nationally representative mortality survey. PLoS Negl Trop Dis. 2011;5(4):e1018.
31. Rahman R, Abul Faiz M, Selim S, Rahman B, Basher A, Jones A et al. Annual incidence of snake
bite in rural Bangladesh. PLoS Negl Trop Dis. 2010;4(10):e860.
32. Ediriweera DS, Kasturiratne A, Pathmeswaran A, Gunawardena NK, Wijayawickrama BA,
Jayamanne SF et al. Mapping the risk of snakebite in Sri Lanka – a national survey with geospatial
analysis. PLoS Negl Trop Dis. 2016;10(7):e0004813.

33. Kasturiratne A, Wickremasinghe AR, de Silva N, Gunawardena NK, Pathmeswaran A, Premaratna
R et al. The global burden of snakebite: a literature analysis and modelling based on regional
estimates of envenoming and deaths. PLoS Med. 2008;5(11): e218.
34. International statistical classification of diseases and related health problems (ICD-10). Geneva:
World Health Organization; 2007 (http://www.who.int/classifications/icd/en/index.html, accessed
13 February 2017).
35. Theakston RD, Warrell DA. Antivenoms: a list of hyperimmune sera currently available for the
treatment of envenoming by bites and stings. Toxicon. 1991;29(12):1419–70.
36. Ismail M, Al-Ahaidib MS, Abdoon N, Abd-Elsalam MA. Preparation of a novel antivenom against
Atractaspis and Walterinnesia venoms. Toxicon. 2007;49(1):8–18.
37. Joseph JK, Simpson ID, Menon NCS, Jose MP, Kulkarni KJ, Raghavendra GB et al. First authenticated
cases of life-threatening envenoming by the hump-nosed pit viper (Hypnale hypnale) in India.
Trans R Soc Trop Med Hyg. 2007;101(1):85–90.
344
Annex 5
38. Ariaratnam CA, Thuraisingam V, Kularatne SAM, Sheriff MHR, Theakston RDG, Silva A et al. Frequent
and potentially fatal envenoming by hump-nosed pit vipers (Hypnale hypnale and H. nepa) in Sri
Lanka: lack of effective antivenom. Trans R Soc Trop Med Hyg. 2008;102(11):1120–6.
39. Chippaux JP, Williams V, White J. Snake venom variability: methods of study, results and
interpretation. Toxicon. 1991;29(11):1279–303.
40. Warrell DA. Geographical and intraspecies variation in the clinical manifestations of envenoming
by snakes. In: Thorpe RS, Wüster W, Malhotra A, editors. Venomous snakes: ecology, evolution and
snakebite. Oxford: Clarendon Press; 1997:189–203.
41. Fry BG, Winkel KD, Wickramaratna JC, Hodgson WC, Wüster W. Effectiveness of snake antivenom:
species and regional variation and its clinical impact. J Toxicol Toxin Reviews. 2003;22(1):23–34.
42. Raweerith R, Ratanabanangkoon K. Immunochemical and biochemical comparisons of equine
monovalent and polyvalent snake antivenoms. Toxicon. 2005;45(3):369–75.
43. Saravia P, Rojas E, Arce V, Guevara C, López JC, Chaves E et al. Geographic and ontogenic variability
in the venom of the neotropical rattlesnake Crotalus durissus: pathophysiological and therapeutic
implications. Rev Biol Trop. 2002;50(1):337–46.
44. Faure G, Bon C. Several isoforms of crotoxin are present in individual venoms from the South
American rattlesnake Crotalus durissus terrificus. Toxicon. 1987;25(2):229–34.
45. Creer S, Malhotra A, Thorpe RS, Stöcklin RS, Favreau PS, Chou W-H. Genetic and ecological
correlates of intraspecific variation in pitviper venom composition detected using matrix-assisted
laser desorption time-of-flight mass spectrometry (MALDI-TOF-MS) and isoelectric focusing.
J Mol Evol. 2003;56(3):317–29.
46. Tan KY, Tan CH, Sim SM, Fung SY, Tan NH. Geographical venom variations of the Southeast Asian
monocled cobra (Naja kaouthia): venom-induced neuromuscular depression and antivenom
neutralization. Comp Biochem Physiol C Toxicol Pharmacol. 2016;185–186:77–86.
47. Tan KY, Tan CH, Fung SY, Tan NH. Venomics, lethality and neutralization of Naja kaouthia (monocled
cobra) venoms from three different geographical regions of Southeast Asia. J Proteomics.
2015;120:105–25.
48. Calvete JJ, Sanz L, Pérez A, Borges A, Vargas AM, Lomonte B et al. Snake population venomics
and antivenomics of Bothrops atrox: Paedomorphism along its transamazonian dispersal
and implications of geographic venom variability on snakebite management. J Proteomics.
2011;74(4):510–27.
49. Reichenbach-Klinke H-H, Elkan E. Diseases of reptiles. Neptune (NJ): TFH Publications; 1965.
50. Frye FL. Reptile care. An atlas of diseases and treatments. Two volumes. Neptune (NJ): TFH
Publications; 1991.
51. Cooper JE, Jackson OF. Diseases of the reptilia. Two volumes. London: Academic Press; 1981.
52. Shortridge KF, Ng MH, Oya A, Kobayashi M, Munro R, Wong F et al. Arbovirus infections in reptiles:
immunological evidence for a high incidence of Japanese encephalitis virus in the cobra Naja
naja. Trans R Soc Trop Med Hyg. 1974;68(6):454–60.
53. Leloup P. Various aspects of venomous snake breeding on a large scale. Acta Zool Pathol Antverp.
1984;78:177–98.
54. Mitchell MA. Snake care and husbandry. Vet Clin Exot Anim. 2004;7:421–46.
55. Chanhome L, Jintakune P, Wilde H, Cox MJ. Venomous snake husbandry in Thailand. Wilderness
Environ Med. 2001;12(1):17–23.
56. Gutiérrez JM, Chaves F, Bolaños R. Comparative study of venoms of newborn and adult specimens
of Bothrops asper. Rev Biol Trop. 1981;28(2):341–51.
345
57. Furtado MFD, Maruyama M, Kamiguti AS, Antonio LC. Comparative study of nine Bothrops snake
venoms from adult female snakes and their offspring. Toxicon. 1991;29(2):219–26.
58. Alape-Girón A, Sanz L, Escolano J, Flores-Díaz M, Madrigal M, Sasa M et al. Snake venomics of
the lancehead pitviper Bothrops asper: geographic, individual, and ontogenetic variations.
J Proteome Res. 2008;7(8):3556–71.
59. Chacón D, Rodríguez S, Arias J, Solano G, Bonilla F, Gómez A. Maintaining coral snakes (Micrurus
nigrocinctus, Serpentes: Elapidae) for venom production on an alternative fish-based diet. Toxicon.
2012;60(3):249–53.
60. Goebel-Stengel M, Stengel A, Taché Y, Reeve Jr JR. The importance of using the optimal
plasticware and glassware in studies involving peptides. Anal Biochem. 2011;414(1):38–46.
61. Kurnik D, Haviv Y, Kochva E. A snake bite by the Burrowing Asp, Atractaspis engaddensis. Toxicon.
1999;37(1):223–7.
62. Harvey DS, Lentini AM, Cedar K, Weatherhead PJ. Moving massasaugas: insight into rattlesnake
relocation using Sistrurus c. catenatus. Herpetol Conserv Bio. 2014;9(1):67–75.
63. Roe JH, Frank MR, Gibson SE, Attum O, Kingsbury BA. No place like home: an experimental
comparison of reintroduction strategies using snakes. J Appl Ecol. 2010;47(6):1253–61.
64. Butler H, Malone B, Clemann N. The effects of translocation on the spatial ecology of tiger snakes
(Notechis scutatus) in a suburban landscape. Wildlife Res. 2005;32(2):165–71.
65. Reinert HK, Rupert RR. Impacts of translocation on behavior and survival of timber rattlesnakes,
Crotalus horridus. J Herpetol. 1999;33(1):45–61.
66. Powell RL, Sanchez EE, Pérez JC. Farming of venom: survey of snake venom extraction facilities
worldwide. Appl Herpetol. 2006;3(1):1–10.
67. Nishioka SA, Silveira PVP, Peixoto-Filho FM, Jorge MT, Sandoz A. Occupational injuries with
captive lance-headed vipers (Bothrops moojeni): experience from a snake farm in Brazil. Trop Med
Int Health. 2000;5(7):507–10.
68. de Medeiros CR, Barbaro KC, Lira MS, França FOS, Zaher VL, Kokron CM et al. Predictors of Bothrops
jararaca venom allergy in snake handlers and snake venom handlers. Toxicon. 2008;51(4):672–80.
69. Warrell DA. Envenoming, poisoning, hypersensitivity and injuries caused by animals. In: Warrell
DA, Cox TM, Firth JD, editors. Oxford textbook of medicine. Oxford: Oxford University Press; 2005.
70. Wüster W, McCarthy CJ. Venomous snake systematics: implications for snakebite treatment and
toxinology. In: Bon C, Goyffon M, editors. Envenomings and their treatments. Lyons: Fondation
Marcel Mérieux; 1996:13–23.
71. León G, Segura A, Gomez A, Hernández A, Navarro D, Villalta M et al. Industrial production
and quality control of snake antivenoms. In: Gopalakrishnakone P, Calvete JJ, editors. Venom
genomics and proteomics. Dordrecht: Springer; 2016:1–22.
72. Angulo Y, Estrada R, Gutiérrez JM. Clinical and laboratory alterations in horses during
immunization with snake venoms for the production of polyvalent (Crotalinae) antivenom.
Toxicon. 1997;35(1):81–90.
73. Malikides N, Hodgson JL, Rose RJ, Hodgson DR. Cardiovascular, haematological and biochemical
responses after large volume blood collection in horses. Vet J. 2001;162(1):44–55.
74. Landon J, Woolley JA, McLean C. Antibody production in the hen. In: Landon J, Chard T, editors.
Therapeutic antibodies. London: Springer-Verlag; 1995:47–68.
75. Landon J, Smith D. Merits of sheep antisera for antivenom manufacture. J Toxicol Toxin Reviews.
2003;22:15–22.
346
Annex 5
76. Hanly WC, Artwohl JE, Bennett BT. Review of polyclonal antibody production procedures in
mammals and poultry. ILAR J. 1995;37(3):93–118.
77. WHO Guidelines on tissue infectivity distribution in transmissible spongiform encephalopathies.
Geneva: World Health Organization; 2006 (http://www.who.int/bloodproducts/cs/TSEPUBLISHED
REPORT.pdf, accessed 14 February 2017).
78. Moroz-Perlmutter C, Goldblum N, de Vries A, Gitter S. Detoxification of snake venoms and venom
fractions by formaldehyde. Exp Biol Med (Maywood). 1963;112(3):595–8.
79. Freitas TV, Fortes-Dias CL, Diniz CR, Velarde DT, Freitas CF. Immunization of horses with Crotalus
durissus terrificus (South American rattlesnake) venom. A comparison of four different procedures.
Braz J Med Biol Res. 1991;24(3):281–90.
80. Reed SG, Hsu F-C, Carter D, Orr MT. The science of vaccine adjuvants: advances in TLR4 ligand
adjuvants. Curr Opin Immunol. 2016;41:85–90.
81. Pratanaphon R, Akesowan S, Khow O, Sriprapat S, Ratanabanangkoon K. Production of highly
potent horse antivenom against the Thai cobra (Naja kaouthia). Vaccine. 1997;15(14):1523–8.
82. Chotwiwatthanakun C, Pratanaphon R, Akesowan S, Sriprapat S, Ratanabanangkoon K. Production
of potent polyvalent antivenom against three elapid venoms using a low dose, low volume,
multi-site immunization protocol. Toxicon. 2001;39(10):1487–94.
83. Quality assurance of pharmaceuticals: a compendium of guidelines and related materials.
Volume 2, Second updated edition. Good manufacturing practices and inspection. Geneva:
World Health Organization; 2007 (http://www.who.int/medicines/areas/quality_safety/quality_
assurance/QualityAssurancePharmVol2.pdf, accessed 14 February 2017).
84. Guidelines on viral inactivation and removal procedures intended to assure the viral safety of
human blood plasma products. In: WHO Expert Committee on Biological Standardization: fifty-
second report. Geneva: World Health Organization; 2004: Annex 4 (WHO Technical Report Series,
No. 924; http://www.who.int/bloodproducts/publications/WHO_TRS_924_A4.pdf, accessed 14
February 2017).
85. Recommendations for the production, control and regulation of human plasma for fractionation.
In: WHO Expert Committee on Biological Standardization: fifty-sixth report. Geneva: World
bloodproducts/publications/TRS941Annex4blood.pdf, accessed 14 February 2017).
86. Note for guidance on production and quality control of animal immunoglobulins and immunsera
for human use. London: European Agency for the Evaluation of Medicinal Products (EMEA),
CPMP; 2002:1–14.
87. Bolanos R, Cerdas L. Production and control of antivenin sera in Costa Rica. Bol Oficina Sanit
Panam. 1980;88(3):189–96.
88. Rojas G, Jimenez JM, Gutiérrez JM. Caprylic acid fractionation of hyperimmune horse plasma:
description of a simple procedure for antivenom production. Toxicon. 1994;32(3):351–63.
89. Otero R, Gutiérrez JM, Rojas G, Núñez V, Díaz A, Miranda E et al. A randomized blinded clinical
trial of two antivenoms, prepared by caprylic acid or ammonium sulphate fractionation of IgG,
in Bothrops and Porthidium snake bites in Colombia: correlation between safety and biochemical
characteristics of antivenoms. Toxicon. 1999;37(6):895–908.
90. Steinbuch M, Audran R. The isolation of IgG from mammalian sera with the aid of caprylic acid.
Arch Biochem Biophys. 1969;134(2):279–84.
91. Gutiérrez JM, Rojas E, Quesada L, León G, Núñez J, Laing GD et al. Pan-African polyspecific
antivenom produced by caprylic acid purification of horse IgG: an alternative to the antivenom
crisis in Africa. Trans R Soc Trop Med Hyg. 2005;99(6):468–75.
347
92. dos Santos MC, D’Império Lima MR, Furtado GC, Colletto GM, Kipnis TL, Dias da Silva W. Purification
of F(abʹ)2 anti-snake venom by caprylic acid: a fast method for obtaining IgG fragments with high
neutralization activity, purity and yield. Toxicon. 1989;27(3):297–303.
93. Otero-Patiño R, Cardoso JL, Higashi HG, Nunez V, Diaz A, Toro MF et al. A randomized, blinded,
comparative trial of one pepsin-digested and two whole IgG antivenoms for Bothrops snake bites
in Uraba, Colombia. The Regional Group on Antivenom Therapy Research (REGATHER). Am J Trop
Med Hyg. 1998;58(2):183–9.
94. Abubakar IS, Abubakar SB, Habib AG, Nasidi A, Durfa N, Yusuf PO et al. Randomised controlled
double-blind non-inferiority trial of two antivenoms for saw-scaled or carpet viper (Echis
ocellatus) envenoming in Nigeria. PLoS Negl Trop Dis. 2010;4(7):e767.
95. Jones RG, Landon J. A protocol for “enhanced pepsin digestion”: a step by step method for
obtaining pure antibody fragments in high yield from serum. J Immunol Methods. 2003;275(1–
2):239–50.
96. Raweerith R, Ratanabanangkoon K. Fractionation of equine antivenom using caprylic acid
precipitation in combination with cationic ion-exchange chromatography. J Immunol Methods.
2003;282(1–2):63–72.
97. Al-Abdulla I, Garnvwa JM, Rawat S, Smith DS, Landon J, Nasidi A. Formulation of a liquid ovine
Fab-based antivenom for the treatment of envenomation by the Nigerian carpet viper (Echis
ocellatus). Toxicon. 2003;42(4):399–404.
98. Saetang T, Treamwattana N, Suttijitpaisal P, Ratanabanangkoon K. Quantitative comparison on
the refinement of horse antivenom by salt fractionation and ion-exchange chromatography. J
Chromatogr B Biomed Sci Appl. 1997;700(1–2):233–9.
99. Sullivan JB Jr, Russell FE. Isolation and purification of antibodies to rattlesnake venom by affinity
chromatography. Proc West Pharmacol Soc. 1982;25:185–92.
100. Magos L. Review on the toxicity of ethylmercury, including its presence as a preservative in
biological and pharmaceutical products. J Appl Toxicol. 2001;21(1):1–5.
101. Guidelines on regulatory expectations related to the elimination, reduction or replacement
of thiomersal in vaccines. In: WHO Expert Committee on Biological Standardization: fifty-third
report. Geneva: World Health Organization; 2004: Annex 4 (WHO Technical Report Series, No. 926;
http://www.who.int/biologicals/publications/trs/areas/vaccines/thiomersal/Annex%204%20(95-
102)TRS926thiomersal.pdf, accessed 14 February 2017).
102. Pikal MJ. Freeze drying. In: Encyclopedia of pharmaceutical technology. New York: Marcel Dekker;
2002:1299–326.
103. Herrera M, Tattini V Jr, Pitombo RNM, Gutiérrez JM, Borgognoni C, Vega-Baudrit J et al. Freeze-dried
snake antivenoms formulated with sorbitol, sucrose or mannitol: comparison of their stability in
an accelerated test. Toxicon. 2014;90:56–63.
104. Meyer WP, Habib AG, Onayade AA, Yakubu A, Smith DC, Nasidi A et al. First clinical experiences
with a new ovine Fab Echis ocellatus snake bite antivenom in Nigeria: randomized comparative
trial with Institute Pasteur Serum (IPSER) Africa antivenom. Am J Trop Med Hyg. 1997;56(3):
291–300.
105. Ariaratnam CA, Meyer WP, Perera G, Eddleston M, Kuleratne SA, Attapattu W et al. A new
monospecific ovine Fab fragment antivenom for treatment of envenoming by the Sri Lankan
Russell’s viper (Daboia Russelii Russelii): a preliminary dose-finding and pharmacokinetic study.
Am J Trop Med Hyg. 1999;61(2):259–65.
106. Scherrmann JM. Antibody treatment of toxin poisoning – recent advances. J Toxicol Clin Toxicol.
1994;32(4):363–75.
348
Annex 5
107. Ho M, Silamut K, White NJ, Karbwang J, Looareesuwan S, Phillips RE et al. Pharmacokinetics of

three commercial antivenoms in patients envenomed by the Malayan pit viper, Calloselasma
rhodostoma, in Thailand. Am J Trop Med Hyg. 1990;42(3):260–6.
108. Choumet V, Audebert F, Rivière G, Sorkine M, Urtizberea M, Sabouraud A et al. New approaches in
antivenom therapy. Adv Exp Med Biol. 1996;391:515–20.
109. Audebert F, Sorkine M, Robbe-Vincent A, Bon C. Viper bites in France: clinical and biological
evaluation; kinetics of envenomations. Hum Exp Toxicol. 1994;13(10):683–8.
110. Boyer LV, Seifert SA, Cain JS. Recurrence phenomena after immunoglobulin therapy for snake
envenomations: Part 2. Guidelines for clinical management with crotaline Fab antivenom. Ann
Emerg Med. 2001;37(2):196–201.
111. Riviere G, Choumet V, Saliou B, Debray M, Bon C. Absorption and elimination of viper venom
after antivenom administration. J Pharmacol Exp Ther. 1998;285(2):490–5.
112. Riviere G, Choumet V, Audebert F, Sabouraud A, Debray M, Scherrmann J-M et al. Effect of
antivenom on venom pharmacokinetics in experimentally envenomed rabbits: toward an
optimization of antivenom therapy. J Pharmacol Exp Ther. 1997;281(1):1–8.
113. Burnouf T, Griffiths E, Padilla A, Seddik S, Stephano MA, Gutiérrez JM. Assessment of the viral
safety of antivenoms fractionated from equine plasma. Biologicals. 2004; 32(3):115–28.
114. Lu G, Sun L, Xu T, He D, Wang Z, Ou S et al. First description of hepacivirus and pegivirus infection
in domestic horses in China: a study in Guangdong Province, Heilongjiang Province and Hong
Kong District. PLoS One. 2016;11(5):e0155662.
115. Kapoor A, Simmonds P, Cullen JM, Scheel TKH, Medina JL, Giannitti F et al. Identification of a
pegivirus (GB virus-like virus) that infects horses. J Virol. 2013;87(12):7185–90.
116. Rushton JO, Lecollinet S, Hubálek Z, Svobodová P, Lussy H, Nowotny N. Tick-borne encephalitis
virus in horses, Austria, 2011. Emerg Infect Dis. 2013;19(4):635–7.
117. Klaus C, Hörügel U, Hoffmann B, Beer M. Tick-borne encephalitis virus (TBEV) infection in horses:
clinical and laboratory findings and epidemiological investigations. Vet Microbiol. 2013;163
(3–4):368–72.
118. Dürrwald R, Kolodziejek J, Herzog S, Nowotny N. Meta-analysis of putative human bornavirus
sequences fails to provide evidence implicating Borna disease virus in mental illness. Rev Med
Virol. 2007;17(3):181–203.
119. Korneyeva M, Hotta J, Lebing W, Rosenthal RS, Franks L, Petteway SR Jr. Enveloped virus
inactivation by caprylate: a robust alternative to solvent-detergent treatment in plasma derived
intermediates. Biologicals. 2002;30(2):153–62.
120. Dichtelmuller H, Rudnick D, Kloft M. Inactivation of lipid enveloped viruses by octanoic acid
treatment of immunoglobulin solution. Biologicals. 2002;30(2):135–42.
121. Parkkinen J, Rahola A, Bonsdorff L, Tölö H, Törmä E. A modified caprylic acid method for
manufacturing immunoglobulin G from human plasma with high yield and efficient virus
clearance. Vox Sang. 2006;90(2):97–104.
122. Reid KG, Cuthbertson B, Jones ADL, McIntosh RV. Potential contribution of mild pepsin treatment
at pH4 to the viral safety of human immunoglobulin products. Vox Sang. 1988;55(2):75–80.
123. Omar A, Kempf C, Immelmann A, Rentsch M, Morgenthaler J-J. Virus inactivation by pepsin
treatment at pH 4 of IgG solutions: factors affecting the rate of virus inactivation. Transfusion.
1996;36(10):866–72.
124. Bos OJM, Sunyé DGJ, Nieuweboer CEF, van Engelenburg FAC, Schuitemaker H, Over J. Virus
validation of pH 4-treated human immunoglobulin products produced by the Cohn fractionation
process. Biologicals. 1998;26(4):267–76.
349
125. Lundblad JL, Seng RL. Inactivation of lipid-enveloped viruses in proteins by caprylate. Vox Sang.
1991;60(2):75–81.
126. Burnouf T, Terpstra F, Habib G, Seddik S. Assessment of viral inactivation during pH 3.3 pepsin
digestion and caprylic acid treatment of antivenoms. Biologicals. 2007;35(4):329–34.
127. El-Ekiaby M, Vargas M, Sayed M, Gorgy G, Goubran H, Radosevic M et al. Minipool caprylic
acid fractionation of plasma using disposable equipment: a practical method to enhance
immunoglobulin supply in developing countries. PLoS Negl Trop Dis. 2015;9(2):e0003501.
128. Mpandi M, Schmutz P, Legrand E, Duc R, Geinoz J, Henzelin-Nkubana C et al. Partitioning and
inactivation of viruses by the caprylic acid precipitation followed by a terminal pasteurization
in the manufacturing process of horse immunoglobulins. Biologicals. 2007;35(4):335–41.
129. Lazar A, Epstein E, Lustig S, Barnea A, Silberstein L, Reuveny S. Inactivation of West-Nile virus
during peptic cleavage of horse plasma IgG. Biologicals. 2002;30(2):163–5.
130. Cameron-Smith R, Miloradovic L, Cheyne I, Healy K. The removal of viruses during the purification
of equine antisera using filtration aids Hyflo Super-Cel™ and Fulmont™ Super A. Biologicals.
2000;28(3):169–74.
131. Burnouf T, Radosevich M, Goubran HA, Willkommen H. Place of nanofiltration for assuring viral
safety of biologicals. Curr Nanosci. 2005;1(3):189–201.
132. Caricati CP, Oliveira-Nascimento L, Yoshida JT, Stephano MA, Caricati ATP. Safety of snake
antivenom immunoglobulins: efficacy of viral inactivation in a complete downstream process.
Biotechnol Prog. 2013;29(4):972–9.
133. Tuck S, Ng R. Protein assays. In: Rodriguez-Diaz R, Wehr T, editors. Analytical techniques for
biopharmaceutical development. New York: Marcel Dekker; 2005:14–24.
134. Segura A, Herrera M, Villalta M, Vargas M, Gutiérrez JM, León G. Assessment of snake antivenom
purity by comparing physicochemical and immunochemical methods. Biologicals. 2013;41(2):
93–7.
135. Solano G, Gomez A, Leon G. Assessing endotoxins in equine-derived snake antivenoms:
comparison of the USP pyrogen test and the Limulus amoebocyte lysate assay (LAL). Toxicon.
2015;105:13–8.
136. Sanchez LV, Pla D, Herrera M, Chippaux JP, Calvete JJ, Gutiérrez JM. Evaluation of the preclinical
efficacy of four antivenoms, distributed in sub-Saharan Africa, to neutralize the venom of the
carpet viper, Echis ocellatus, from Mali, Cameroon, and Nigeria. Toxicon. 2015;106:97–107.
137. Pla D, Gutiérrez JM, Calvete JJ. Second generation snake antivenomics: comparing immunoaffinity
and immunodepletion protocols. Toxicon. 2012;60(4):688–99.

138. Gutiérrez JM, Lomonte B, Sanz L, Calvete JJ, Pla D. Immunological profile of antivenoms: preclinical
analysis of the efficacy of a polyspecific antivenom through antivenomics and neutralization
assays. J Proteomics. 2014;105:340–50.
139. Calvete JJ, Sanz L, Pla D, Lomonte B, Gutiérrez JM. Omics meets biology: application to the design
and preclinical assessment of antivenoms. Toxins (Basel). 2014;6(12):3388–405.
140. Calvete JJ, Arias AS, Rodríguez Y, Quesada-Bernat S, Sánchez LV, Chippaux JP et al. Preclinical
evaluation of three polyspecific antivenoms against the venom of Echis ocellatus: neutralization
of toxic activities and antivenomics. Toxicon. 2016;119:280–8.
141. Finney DJ. Probit analysis, third edition. Cambridge: Cambridge University Press; 1971.
142. Antivenoms – guidelines for potency assay. In: WHO Expert Committee on Biological
Standardization: forty-first report. Geneva: World Health Organization; 1991 (WHO Technical
Report Series, No. 814:4 (http://apps.who.int/iris/bitstream/10665/41100/1/WHO_TRS_814.pdf,
accessed 26 March 2017).
350
Annex 5
143. Rojas E, Quesada L, Arce V, Lomonte B, Rojas G, Gutiérrez JM. Neutralization of four Peruvian
Bothrops sp. snake venoms by polyvalent antivenoms produced in Perú and Costa Rica: preclinical
assessment. Acta Trop. 2005;93(1):85–95.
144. Reid HA, Theakston RD. The management of snake bite. Bull World Health Organ. 1983;61(6):
885–95.
145. Theakston RD, Reid HA. Development of simple standard assay procedures for the characterization
of snake venom. Bull World Health Organ. 1983;61(6):949–56.
146. Gutiérrez JM, Gené JA, Rojas G, Cerdas L. Neutralization of proteolytic and hemorrhagic activities
of Costa Rican snake venoms by a polyvalent antivenom. Toxicon. 1985;23(6):887–93.
147. Reid HA. Cobra-bites. Br Med J. 1964;2(5408):540–5.
148. Gutiérrez JM, León G, Rojas G, Lomonte B, Rucavado A, Chaves F. Neutralization of local tissue
damage induced by Bothrops asper (terciopelo) snake venom. Toxicon. 1998;36(11):1529–38.
149. Iddon D, Theakston RD, Ownby CL. A study of the pathogenesis of local skin necrosis induced
by Naja nigricollis (spitting cobra) venom using simple histological staining techniques. Toxicon.
1987;25(6):665–72.
150. Rivel M, Solano D, Herrera M, Vargas M, Villalta M, Segura A et al. Pathogenesis of dermonecrosis
induced by venom of the spitting cobra, Naja nigricollis: an experimental study in mice. Toxicon.
2016;119:171–9.
151. Laing GD, Theakston RDG, Leite RP, Dias da Silva WD, Warrell DA, BIASG (Butantan Institute
Antivenom Study Group). Comparison of the potency of three Brazilian Bothrops antivenoms
using in vivo rodent and in vitro assays. Toxicon. 1992;30(10):1219–25.
152. Theakston RDG. Characterization of venoms and standardization of antivenoms. In: Harris JB,
editor. Natural toxins, animal, plant and microbial. Oxford: Clarendon Press, Oxford Science
Publications; 1986:287–303.
153. Gene JA, Roy A, Rojas G, Gutiérrez JM, Cerdas L. Comparative study on coagulant, defibrinating,
fibrinolytic and fibrinogenolytic activities of Costa Rican crotaline snake venoms and their
neutralization by a polyvalent antivenom. Toxicon. 1989;27(8):841–8.
154. Gutiérrez JM, Romero M, Núñez J, Chaves F, Borkow G, Ovadia M. Skeletal muscle necrosis and
regeneration after injection of BaH1, a hemorrhagic metalloproteinase isolated from the venom
of the snake Bothrops asper (terciopelo). Exp Mol Pathol. 1995;62(1):28–41.
155. Gutiérrez JM, Rojas G, Jorge da Silva N, Núñez J. Experimental myonecrosis induced by the
venoms of South American Micrurus (coral snakes). Toxicon. 1992;30(10):1299–302.
156. Alam JM, Ali SA, Faridi N, Alam SM. Myotoxic characteristics of snake venoms: myonecrosis
induced in albino rats by crude venoms and the purified myotoxic component, phospholipase A2.
J Coll Physicians Surg Pak. 2002;12(7):398–402.
157. Ginsborg BL, Warriner J. The isolated chick biventer cervicis nerve-muscle preparation. Br J
Pharmacol Chemother. 1960;15:410–11.
158. Harvey AL, Barfaraz A, Thomson E, Faiz A, Preston S, Harris JB. Screening of snake venoms for
neurotoxic and myotoxic effects using simple in vitro preparations from rodents and chicks.
Toxicon. 1994;32(3):257–65.
159. Bulbring E. Observations on the isolated phrenic nerve diaphragm preparation of the rat. Br J
Pharmacol Chemother. 1946;1:38–61.
160. Jones RG, Lee L, Landon J. The effects of specific antibody fragments on the “irreversible”
neurotoxicity induced by Brown snake (Pseudonaja) venom. Br J Pharmacol. 1999;126(3):581–4.
161. Crachi MT, Hammer LW, Hodgson WC. A pharmacological examination of venom from the
Papuan taipan (Oxyuranus scutellatus canni). Toxicon. 1999;37(12):1721–34.
351
162. Kitchen I. Neuromuscular blocking drugs and the rat phrenic nerve hemidiaphragm preparation.
In: Textbook of in vitro pharmacology. Oxford: Blackwell Scientific; 2009:79–83.
163. Theakston RDG, Laing GD, Fielding CM, Freite Lascano A, Touzet J-M, Vallejo F et al. Treatment of
snake bites by Bothrops species and Lachesis muta in Ecuador: laboratory screening of candidate
antivenoms. Trans R Soc Trop Med Hyg. 1995;89(5):550–4.
164. Keegan HL, Weaver RE, Matsui T. Southeast Asian snakebite antivenin studies. Japan: 406th
Medical Laboratory, Control Department of Entomology, US Army Medical Command; 1964.
165. Warrell DA, Warrell MJ, Edgar W, Prentice CR, Mathison J, Mathison J. Comparison of Pasteur
and Behringwerke antivenoms in envenoming by the carpet viper (Echis carinatus). Br Med J.
1980;280(6214):607–9.
166. Rosenberger WF, Haines LM. Competing designs for phase I clinical trials: a review. Stat Med.
2002;21(18):2757–70.
167. Chippaux J-P, Stock RP, Massougbodji A. Antivenom safety and tolerance for the strategy of
snake envenomation management. In: Gopalakrishnakone P, Inagaki H, Mukherjee AK, Rahmy TR,
Vogel C-W, editors. Snake venoms. Dordrecht: Springer Science+Business Media; 2015:1–16.
168. Machin D, Campbell MJ, Tan SB, Tan SH. Sample size tables for clinical studies, third edition.
Chichester: Wiley Blackwell; 2009.
169. Visser LE, Kyei-Faried S, Belcher DW, Geelhoed DW, Schagen van Leeuwen J, van Roosmalen J.
Failure of a new antivenom to treat Echis ocellatus snake bite in rural Ghana: the importance of
quality surveillance. Trans R Soc Trop Med Hyg. 2008;102(5):445–50.
170. Armitage P. Sequential medical trials, second edition. New York: John Wiley; 1975.
171. Vakil BJ, Armitage P, Clifford RE, Laurence DR. Therapeutic trial of intracisternal human tetanus
immunoglobulin in clinical tetanus. Trans R Soc Trop Med Hyg. 1979;73(5):579–83.
172. Smalligan R, Cole J, Brito N, Laing GD, Mertz BL, Manock S et al. Crotaline snake bite in the
Ecuadorian Amazon: randomised double blind comparative trial of three South American
polyspecific antivenoms. BMJ. 2004;329:1129.
173. Ariaratnam CA, Sjöström L, Raziek Z, Abeyasinghe S, Kularatne M, Kodikara Arachchi RWK et al.
An open, randomized comparative trial of two antivenoms for the treatment of envenoming by
Sri Lankan Russell’s viper (Daboia russelii russelii). Trans R Soc Trop Med Hyg. 2001;95(1):74–80.
174. Warrell DA, Looareesuwan S, Theakston RD, Phillips RE, Chanthavanich P, Viravan C et al.
Randomized comparative trial of three monospecific antivenoms for bites by the Malayan pit
viper (Calloselasma rhodostoma) in southern Thailand: clinical and laboratory correlations. Am J

Trop Med Hyg. 1986;35(6):1235–47.
175. Warrell DA, Davidson NM, Omerod LD, Pope HM, Watkins BJ, Greenwood BM et al. Bites by
the saw‑scaled or carpet viper (Echis carinatus): trial of two specific antivenoms. Br Med J.
1974;4(5942):437–40.
176. Jorge MT, Cardoso JLC, Castro SCB, Ribeiro L, França FOS, Sbrogio de Almeida ME, Kamiguti
AS et al. A randomized “blinded” comparison of two doses of antivenom in the treatment of
Bothrops envenoming in São Paulo, Brazil. Trans R Soc Trop Med Hyg. 1995;89(1):111–4.
177. Cardoso JLC, Fan HW, França FOS, Jorge MT, Leite RP, Nishioka SA et al. Randomized comparative
trial of three antivenoms in the treatment of envenoming by lance-headed vipers (Bothrops
jararaca) in São Paulo, Brazil. QJM. 1993;86(5):315–25.
178. Otero R, León G, Gutiérrez JM, Rojas G, Toro MF, Barona J et al. Efficacy and safety of two whole IgG
polyvalent antivenoms, refined by caprylic acid fractionation with or without β-propiolactone,
in the treatment of Bothrops asper bites in Colombia. Trans R Soc Trop Med Hyg. 2006;100(12):
1173–82.
352
Annex 5
179. Bush SP, Green SM, Moynihan JA, Hayes WK, Cardwell MD. Crotalidae polyvalent immune Fab
(ovine) antivenom is efficacious for envenomations by Southern Pacific rattlesnakes (Crotalus
helleri). Ann Emerg Med. 2002;40(6):619–24.
180. Warrell DA. Unscrupulous marketing of snake bite antivenoms in Africa and Papua New Guinea:
choosing the right product – “what’s in a name?”. Trans R Soc Trop Med Hyg. 2008;102(5):397–9.
181. Guidelines for national authorities on quality assurance for biological products. In: WHO
Expert Committee on Biological Standardization: forty-second report. Geneva: World Health
Organization; 1992: Annex 2 (WHO Technical Report Series, No. 822; http://apps.who.int/iris/
bitstream/10665/39431/1/WHO_TRS_822.pdf, accessed 14 February 2017).
182. Regulation and licensing of biological products in countries with newly developing regulatory
authorities. In: WHO Expert Committee on Biological Standardization: forty-fifth report. Geneva:
World Health Organization; 1995: Annex 1 (WHO Technical Report Series, No. 858; http://apps.
who.int/iris/bitstream/10665/36999/1/WHO_TRS_858.pdf, accessed 14 February 2017).
353
Appendix 1
Worldwide distribution of medically important
venomous snakes
Venomous snakes are widely distributed, especially in tropical countries, from
sea level to altitudes of up to 4900 metres (Gloydius himalayanus). The European
adder (Vipera berus) enters the Arctic Circle, and the Argentine Yararanata
(Bothrops ammodytoides) occurs to 47 °S and is the most southerly occurring
venomous snake. No other venomous species occur in cold regions such as
the Arctic, Antarctic and north of around latitude 51 °N in North America
(Newfoundland, Nova Scotia).
This Appendix lists venomous snake species considered to represent the
greatest threat to public health in various countries, territories and other areas or
regions around the world. Only species which fall into one of the two categories
listed below are shown, and category listings are in alphabetical order according
to taxonomic family, genus and species. The intention in categorizing these
medically important snakes into two groups is to provide users of these WHO
Guidelines with a prioritized listing. Snakes in both Category 1 and Category 2
are species for which antivenom production is important; however species
listed in Category 1 within a country, territory or area should be considered as
being of highest priority for antivenom production on the basis that available
knowledge implicates them as being responsible for the greater burden in that
particular setting.
Definitions of the categories used in this listing are:
■■ CATEGORY 1 (CAT 1): Highest medical importance
Definition: highly venomous snakes which are common or widespread

and cause numerous snake-bites, resulting in high levels of morbidity,
disability or mortality.
■■ CATEGORY 2 (CAT 2): Secondary medical importance
Definition: highly venomous snakes capable of causing morbidity,
disability or death, but: (a) for which exact epidemiological or clinical
data may be lacking; and/or (b) are less frequently implicated (owing
to their activity cycles, behaviour, habitat preferences or occurrence in
areas remote from large human populations).
There are numerous other venomous species that rank as lesser threats
in countries territories and other areas listed here, and interested readers should
354
Annex 5
refer to the herpetological references provided in these Guidelines. It should be

noted that over time, as more information becomes available, new species will
doubtlessly be added to these lists, and some species, currently defined within
Category 1 or Category 2 will be re-ranked.
It should also be noted that the organization of countries, territories and
other areas in this Appendix does not follow the WHO regional organization, but
instead is arranged biogeographically in alphabetical order of country, territory
or geographical area. This approach was necessary to reflect the geographical
distribution of major groups of venomous snakes throughout the world. For
example, the venomous snakes of the eastern Indonesian Province of Papua have
biogeographical origins in Australo-Papua, and are evolutionarily distinct from
the venomous snakes of Asian origin that occur west of Wallace’s Line, which runs
south of the Philippines, between Borneo and Sulawesi, and between Bali and
Lombok, and which separates the zoogeographical regions of Asia and Australia.
For this reason, the medically important snakes of Indonesian Papua are listed in
the Australo-Papuan region, rather than the South-East Asian region.
Users of this Appendix should also recognize that the relative risk of
injury from a particular species may vary from one country, territory or area to
another. For this reason, some species that have been listed under Category 1
in one country, territory or area may have been listed under Category 2 in
another country, territory or area to reflect the different risk posed by that species
in different locations. Assignment to Category 1 or Category 2 was based in some
cases on the relative importance of a species as a cause of snake-bite. In Europe,
for example, the overall incidence of snake-bite is trivial compared to that in
West Africa or India, but where a European species (such as Vipera berus) is a
major (or sole) cause of envenoming where it occurs, this warrants ranking it as
a medically important species in that setting.
AFRICA AND THE MIDDLE EAST

Island populations
Off the coast of Africa, there are no medically important snakes in Mauritius,
Réunion, Rodrigues, the Comoros, the Canary Islands, the Cabo Verde Islands
or the Seychelles. The islands that do have venomous snakes include the Lamu
group, Zanzibar, Pemba and Mafia Islands, the Bazaruto Archipelago and Inhaca
Island, São Tomé, Principe, Bioko (Fernando Po) and Dahlak Islands. The
venomous snakes on these islands tend to be similar to those on the adjacent
mainland. A colubrid, Madagascarophis meridionalis, and perhaps other species
of the same genus, are the only terrestrial snakes of possible, if minimal, medical
importance found in Madagascar.
355
North Africa/Middle East

Algeria:
Cat 1: Elapidae: Naja haje; Viperidae: Cerastes cerastes; Daboia mauritanica
Cat 2: Viperidae: Daboia deserti; Echis leucogaster; Macrovipera lebetina;
Vipera latastei
Cyprus:
Cat 1: None
Cat 2: Viperidae: Macrovipera lebetina
Egypt:
Cat 1: Elapidae: Naja haje; Viperidae: Cerastes cerastes; Echis coloratus (east),
Echis pyramidum
Cat 2: Atractaspididae: Atractaspis engaddensis (Sinai); Elapidae: Naja nubiae;
Walterinnesia aegyptia (Sinai); Viperidae: Pseudocerastes fieldi
Iraq:
Cat 1: Viperidae: Echis carinatus; Macrovipera lebetina
Cat 2: Elapidae: Walterinnesia morgani; Viperidae: Cerastes gasperettii;
Pseudocerastes fieldi, Pseudocerastes persicus
Iran (Islamic Republic of):

Cat 1: Elapidae: Naja oxiana; Viperidae: Echis carinatus; Macrovipera lebetina;
Pseudocerastes persicus
Cat 2: Elapidae: Bungarus persicus (south-east); Walterinnesia morgani (west);
Viperidae: Eristicophis macmahonii (east); Gloydius halys caucasicus;
Montivipera raddei; Vipera spp.
Israel:
Cat 1: Viperidae: Daboia palaestinae; Echis coloratus
Cat 2: Atractaspididae: Atractaspis engaddensis; Elapidae: Walterinnesia aegyptia;
Viperidae: Cerastes cerastes, Cerastes gasperettii; Pseudocerastes fieldi
Jordan:
Viperidae: Cerastes gasperettii; Macrovipera lebetina; Pseudocerastes fieldi
356
Annex 5
Kuwait and Qatar:

Cat 1: Viperidae: Cerastes gasperettii
Cat 2: Elapidae: Walterinnesia morgani (Kuwait only)
Lebanon:
Cat 1: Viperidae: Daboia palaestinae; Macrovipera lebetina
Cat 2: None
Libya:
Cat 1: Elapidae: Naja haje; Viperidae: Cerastes cerastes; Echis pyramidum
Cat 2: Viperidae: Daboia deserti
Morocco:
Cat 1: Elapidae: Naja haje; Viperidae: Bitis arietans; Cerastes cerastes; Daboia
mauritanica
Cat 2: Viperidae: Echis leucogaster; Vipera latastei
Oman:
Cat 1: Atractaspididae: Atractaspis andersonii (south-west); Viperidae: Bitis arietans
(south-west); Echis coloratus (south-west), Echis carinatus, Echis omanensis
(north)
Cat 2: Elapidae: Naja arabica (south-west); Viperidae: Cerastes gasperettii;
Echis khosatzkii (south-west); Pseudocerastes persicus
Saudi Arabia:
Cat 1: Atractaspididae: Atractaspis andersonii (south-west);
Viperidae: Cerastes gasperettii; Echis coloratus, Echis borkini (south-west)
Cat 2: Atractaspididae: Atractaspis engaddensis (north-west); Elapidae: Naja arabica
(south-west); Walterinnesia aegyptia (west), Walterinnesia morgani (central
and south); Viperidae: Bitis arietans (south-west); Cerastes cerastes (south-
west); Pseudocerastes fieldi
Syrian Arab Republic:

Cat 1: Viperidae: Daboia palaestinae; Macrovipera lebetina
Cat 2: Viperidae: Pseudocerastes fieldi
357
Tunisia:
Cat 1: Viperidae: Daboia mauritanica
Cat 2: Elapidae: Naja haje; Viperidae: Cerastes cerastes; Daboia deserti; Echis
leucogaster; Macrovipera lebetina; Vipera latastei
Turkey:
Cat 1: Viperidae: Macrovipera lebetina; Montivipera xanthina
Cat 2: Elapidae: Walterinnesia morgani (south); Viperidae: Montivipera raddei;
Vipera ammodytes, Vipera eriwanensis, Vipera spp.
United Arab Emirates:

Cat 1: Viperidae: Echis carinatus (east); Echis omanensis
Cat 2: Viperidae: Cerastes gasperettii; Pseudocerastes persicus
West Bank and Gaza Strip:

Viperidae: Cerastes cerastes; Pseudocerastes fieldi
Western Sahara:
Cat 1: Viperidae: Cerastes cerastes
Cat 2: Elapidae: Naja haje; Viperidae: Bitis arietans
Yemen:
Cat 1: Atractaspididae: Atractaspis andersonii; Elapidae: Naja arabica;
Viperidae: Bitis arietans; Echis borkini, Echis coloratus
Cat 2: Viperidae: Cerastes cerastes, Cerastes gasperettii; Echis khosatzkii
Central sub-Saharan Africa

Angola:
Cat 1: Elapidae: Dendroaspis jamesoni, Dendroaspis polylepis; Naja anchietae,
Naja melanoleuca, Naja nigricollis; Viperidae: Bitis arietans, Bitis gabonica
Cat 2: Atractaspididae: Atractaspis bibronii, Atractaspis irregularis;
Colubridae: Dispholidus typus; Thelotornis capensis, Thelotornis kirtlandii
(north); Elapidae: Naja christyi (Cabinda), Naja mossambica (south),
Naja nigricincta (south-west); Pseudohaje goldii; Viperidae: Atheris squamigera;
Bitis nasicornis (Cabinda)
358
Annex 5
Burundi:
Cat 1: Elapidae: Naja nigricollis, Naja melanoleuca; Viperidae: Bitis arietans
Colubridae: Dispholidus typus; Thelotornis mossambicanus;
Elapidae: Dendroaspis jamesoni; Viperidae: Bitis gabonica, Bitis nasicornis
Central African Republic:

Cat 1: Elapidae: Dendroaspis jamesoni, Dendroaspis polylepis; Naja haje,
Naja nigricollis; Viperidae: Bitis arietans, Bitis gabonica; Echis ocellatus,
Echis pyramidum
Cat 2: Atractaspididae: Atractaspis irregularis; Colubridae: Dispholidus typus;
Thelotornis kirtlandii; Elapidae: Naja annulata, Naja melanoleuca;1
Pseudohaje goldii; Viperidae: Atheris broadleyi, Atheris squamigera;
Bitis nasicornis
Chad:
Cat 1: Elapidae: Naja haje, Naja nigricollis; Viperidae: Bitis arietans (south);
Echis ocellatus (south)
Cat 2: Colubridae: Dispholidus typus; Elapidae: Naja katiensis, Naja nubiae;
Viperidae: Cerastes cerastes
Republic of the Congo:

Cat 1: Elapidae: Dendroaspis jamesoni; Naja melanoleuca; Viperidae: Bitis gabonica,
Bitis nasicornis
Thelotornis kirtlandii; Elapidae: Naja annulata, Naja christyi, Naja nigricollis;
Pseudohaje goldii; Viperidae: Atheris squamigera; Bitis arietans
Democratic Republic of the Congo:

Cat 1: Elapidae: Dendroaspis jamesoni; Naja melanoleuca, Naja nigricollis;
Viperidae: Bitis arietans, Bitis gabonica, Bitis nasicornis
Colubridae: Dispholidus typus; Thelotornis capensis, Thelotornis kirtlandii;
Elapidae: Dendroaspis polylepis; Naja anchietae (Katanga pedicle),
Naja annulata, Naja christyi, Naja haje (north); Pseudohaje goldii;
Viperidae: Atheris squamigera
The medical importance of this species may be higher in the primary forest zone of the south-western
1
Central African Republic, and in some secondary forest mosaic zones elsewhere in the Central African
Republic.
359
Equatorial Guinea:
Cat 1: Elapidae: Dendroaspis jamesoni; Naja melanoleuca; Viperidae: Bitis gabonica,
Bitis nasicornis
Cat 2: Atractaspididae: Atractaspis irregularis; Colubridae: Thelotornis kirtlandii;
Elapidae: Naja annulata; Pseudohaje goldii; Viperidae: Atheris squamigera
Gabon:
Cat 1: Elapidae: Dendroaspis jamesoni; Naja melanoleuca, Naja nigricollis;
Viperidae: Bitis gabonica, Bitis nasicornis
Elapidae: Naja annulata; Pseudohaje goldii; Viperidae: Atheris squamigera;
Bitis arietans
Rwanda:
Cat 1: Elapidae: Dendroaspis jamesoni; Naja nigricollis; Viperidae: Bitis arietans
Colubridae: Dispholidus typus; Thelotornis kirtlandii;
Elapidae: Dendroaspis polylepis; Naja annulata, Naja melanoleuca;
Pseudohaje goldii; Viperidae: Bitis gabonica, Bitis nasicornis
East sub-Saharan Africa

Djibouti:
Cat 1: Viperidae: Echis pyramidum
Cat 2: Atractaspididae: Atractaspis fallax; Colubridae: Dispholidus typus;
Elapidae: Naja pallida; Viperidae: Bitis arietans
Eritrea:
Cat 1: Elapidae: Dendroaspis polylepis; Naja haje; Viperidae: Bitis arietans;

Echis pyramidum
Elapidae: Naja nubiae; Viperidae: Echis megalocephalus
Ethiopia:
Cat 1: Elapidae: Dendroaspis polylepis; Naja ashei (south-east), Naja haje,
Naja nigricollis; Viperidae: Bitis arietans; Echis pyramidum
Cat 2: Atractaspididae: Atractaspis fallax, Atractaspis irregularis (Mt Bizen);
Colubridae: Dispholidus typus; Elapidae: Naja melanoleuca, Naja pallida;
Viperidae: Bitis parviocula, Bitis harenna
360
Annex 5
Kenya:
Cat 1: Elapidae: Dendroaspis angusticeps, Dendroaspis polylepis; Naja ashei (north
& east), Naja haje, Naja nigricollis; Viperidae: Bitis arietans; Echis pyramidum
Cat 2: Atractaspididae: Atractaspis bibronii, Atractaspis fallax, Atractaspis irregularis;
Colubridae: Dispholidus typus; Thelotornis mossambicanus,
Thelotornis usambaricus (east coast); Elapidae: Dendroaspis jamesoni;
Naja melanoleuca (west and coastal forest), Naja pallida (north and east);
Pseudohaje goldii; Viperidae: Atheris squamigera; Bitis nasicornis, Bitis gabonica
(west)
Malawi:
Cat 1: Elapidae: Dendroaspis angusticeps, Dendroaspis polylepis; Naja annulifera,
Naja mossambica, Naja nigricollis; Viperidae: Bitis arietans
Cat 2: Atractaspididae: Atractaspis bibronii; Colubridae: Dispholidus typus;
Thelotornis capensis, Thelotornis mossambicanus; Elapidae: Naja melanoleuca;
Viperidae: Proatheris superciliaris
Mozambique:
Cat 1: Elapidae: Dendroaspis angusticeps, Dendroaspis polylepis; Naja annulifera,
Naja mossambica; Viperidae: Bitis arietans, Bitis gabonica
Thelotornis capensis, Thelotornis mossambicanus;
Elapidae: Hemachatus haemachatus; Naja melanoleuca;
Viperidae: Proatheris superciliaris
Somalia:
Cat 1: Elapidae: Dendroaspis polylepis; Naja ashei (south), Naja haje;
Viperidae: Bitis arietans; Echis pyramidum
Cat 2: Atractaspididae: Atractaspis fallax; Colubridae: Dispholidus typus;
Thelotornis mossambicanus; Elapidae: Naja pallida, Naja melanoleuca;
Viperidae: Echis hughesi (north)
South Sudan:
Cat 1: Elapidae: Naja haje, Naja nigricollis; Viperidae: Bitis arietans;
Echis pyramidum
Cat 2: Atractaspididae: Atractaspis fallax, Atractaspis irregularis;
Colubridae: Dispholidus typus; Elapidae: Dendroaspis jamesoni,
Dendroaspis polylepis; Naja melanoleuca, Naja nubiae, Naja pallida;
Viperidae: Bitis gabonica, Bitis nasicornis
361
Sudan:
Cat 1: Elapidae: Naja haje; Viperidae: Bitis arietans; Echis pyramidum
Cat 2: Colubridae: Dispholidus typus; Elapidae: Dendroaspis polylepis (east);
Naja nubiae; Viperidae: Cerastes cerastes, Echis coloratus (east)
United Republic of Tanzania:

Cat 1: Elapidae: Dendroaspis angusticeps, Dendroaspis polylepis; Naja mossambica
(including Pemba Island), Naja nigricollis; Viperidae: Bitis arietans
Cat 2: Atractaspididae: Atractaspis bibronii, Atractaspis fallax (north),
Atractaspis irregularis (north-east); Colubridae: Dispholidus typus;
Thelotornis capensis, Thelotornis kirtlandii (Mahali and Udzungwa Mountains),
Thelotornis mossambicanus, Thelotornis usambaricus (East Usambara
Mountains); Elapidae: Naja ashei (poss. north-east), Naja annulata, Naja haje
(north), Naja melanoleuca (west and coast, including Mafia Island),
Naja pallida; Viperidae: Atheris squamigera; Bitis gabonica (west and south-
east), Bitis nasicornis (north); Proatheris superciliaris
Uganda:
Cat 1: Elapidae: Naja ashei (north-east), Naja haje (north), Naja nigricollis;
Dendroaspis jamesoni, Dendroaspis polylepis; Viperidae: Bitis arietans,
Bitis gabonica
Thelotornis kirtlandii; Elapidae: Naja melanoleuca; Pseudohaje goldii;
Viperidae: Atheris squamigera; Bitis nasicornis
Zambia:
Cat 1: Elapidae: Dendroaspis polylepis; Naja anchietae, Naja annulifera,
Naja mossambica, Naja nigricollis; Viperidae: Bitis arietans, Bitis gabonica

Thelotornis capensis, Thelotornis kirtlandii, Thelotornis mossambicanus;
Elapidae: Naja annulata, Naja melanoleuca
South sub-Saharan Africa

Botswana:
Cat 1: Elapidae: Dendroaspis polylepis; Naja anchietae (west), Naja annulifera
(east), Naja mossambica, Naja nivea (south-west); Viperidae: Bitis arietans
Thelotornis capensis
362
Annex 5
Lesotho:
Cat 1: Elapidae: Naja nivea; Viperidae: Bitis arietans
Cat 2: Elapidae: Hemachatus haemachatus
Namibia:
Cat 1: Elapidae: Dendroaspis polylepis; Naja anchietae, Naja nivea (central and
southern), Naja mossambica (north-east), Naja nigricincta;
Viperidae: Bitis arietans
Thelotornis capensis; Elapidae: Naja nigricollis (Caprivi)
South Africa:
Cat 1: Elapidae: Dendroaspis angusticeps (Natal), Dendroaspis polylepis;
Naja annulifera (north-east), Naja nivea, Naja mossambica (north-east);
Thelotornis capensis; Elapidae: Hemachatus haemachatus; Naja melanoleuca
(KwaZulu-Natal), Naja nigricincta (north-west); Viperidae: Bitis gabonica
(KwaZulu-Natal)
Swaziland:
Cat 1: Elapidae: Dendroaspis polylepis; Naja annulifera, Naja mossambica;
Thelotornis capensis; Elapidae: Hemachatus haemachatus
Zimbabwe:
Cat 1: Elapidae: Dendroaspis polylepis; Naja anchietae (west), Naja annulifera,
Naja mossambica; Viperidae: Bitis arietans
Thelotornis capensis, Thelotornis mossambicanus;
Elapidae: Dendroaspis angusticeps (east); Hemachatus haemachatus (Nyanga
Mountains); Naja melanoleuca (east); Viperidae: Bitis gabonica (east)
363
West sub-Saharan Africa

Benin:
Cat 1: Elapidae: Naja nigricollis; Viperidae: Bitis arietans; Echis ocellatus
Elapidae: Dendroaspis jamesoni; Naja katiensis, Naja melanoleuca,
Naja senegalensis; Pseudohaje nigra; Viperidae: Bitis rhinoceros;
Echis leucogaster (far north)
Burkina Faso:
Cat 1: Elapidae: Naja nigricollis, Naja katiensis; Viperidae: Bitis arietans;
Echis ocellatus
Cat 2: Colubridae: Dispholidus typus; Elapidae: Dendroaspis polylepis;
Naja melanoleuca, Naja senegalensis; Viperidae: Echis leucogaster
Cameroon:
Cat 1: Elapidae: Dendroaspis jamesoni; Naja haje, Naja nigricollis, Naja
melanoleuca; 2 Viperidae: Bitis arietans, Bitis gabonica, Bitis nasicornis;
Echis ocellatus
Thelotornis kirtlandii; Elapidae: Dendroaspis polylepis; Naja annulata,
Naja katiensis; Pseudohaje goldii; Viperidae: Atheris broadleyi (East Province),
Atheris squamigera
Côte d’Ivoire:
Cat 1: Elapidae: Dendroaspis viridis; Naja nigricollis, Naja melanoleuca,
Naja senegalensis; Viperidae: Bitis arietans, Bitis nasicornis, Bitis rhinoceros;
Echis ocellatus

Thelotornis kirtlandii; Elapidae: Dendroaspis polylepis; Naja katiensis;
Pseudohaje goldii, Pseudohaje nigra; Viperidae: Atheris chlorechis
Gambia:
Cat 1: Elapidae: Dendroaspis viridis; Naja nigricollis; Viperidae: Bitis arietans;
Echis jogeri
Cat 2: Colubridae: Dispholidus typus; Elapidae: Naja katiensis, Naja melanoleuca,
Naja senegalensis
This large, highly venomous snake is common in forested areas of south-west Cameroon and a high
2
burden of injury may be expected, although clinical data with direct attribution are not yet available.
364
Annex 5
Ghana:
Cat 1: Elapidae: Dendroaspis viridis; Naja nigricollis, Naja senegalensis;
Viperidae: Bitis arietans; Echis ocellatus
Cat 2: Atractaspididae: Atractaspis irregularis; Colubridae: Dispholidus typus,
Thelotornis kirtlandii; Elapidae: Naja katiensis, Naja melanoleuca; 3
Pseudohaje goldii, Pseudohaje nigra; Viperidae: Atheris chlorechis;
Bitis nasicornis, Bitis rhinoceros
Guinea:
Cat 1: Elapidae: Dendroaspis polylepis, Dendroaspis viridis; Naja katiensis,
Naja nigricollis, Naja melanoleuca, Naja senegalensis;
Viperidae: Bitis arietans; Echis jogeri
Thelotornis kirtlandii; Elapidae: Pseudohaje nigra; Viperidae: Atheris chlorechis;
Bitis nasicornis, Bitis rhinoceros
Guinea-Bissau:
Cat 1: Elapidae: Dendroaspis viridis; Naja nigricollis, Naja melanoleuca,
Naja senegalensis; Viperidae: Bitis arietans; Echis jogeri
Cat 2: Colubridae: Dispholidus typus; Thelotornis kirtlandii; Viperidae: Bitis rhinoceros
Liberia:
Cat 1: Elapidae: Dendroaspis viridis; Naja melanoleuca, Naja nigricollis
Elapidae: Pseudohaje nigra; Viperidae: Atheris chlorechis; Bitis nasicornis,
Bitis rhinoceros
Mali:
Cat 1: Elapidae: Naja katiensis, Naja nigricollis, Naja senegalensis;
Viperidae: Bitis arietans; Echis jogeri (west), Echis leucogaster, Echis ocellatus
Cat 2: Colubridae: Dispholidus typus; Elapidae: Naja melanoleuca;
Mauritania:
Cat 1: Elapidae: Naja senegalensis (south-east); Viperidae: Cerastes cerastes;
Echis leucogaster
Cat 2: Viperidae: Bitis arietans
The medical importance of this species may be higher in the forested zone of southern Ghana.
3
365
Niger:
Cat 1: Elapidae: Naja nigricollis; Viperidae: Bitis arietans; Echis leucogaster,
Echis ocellatus
Cat 2: Colubridae: Dispholidus typus; Elapidae: Naja haje (south-central),
Naja katiensis, Naja nubiae; Naja senegelensis (south-west);
Nigeria:
Cat 1: Elapidae: Dendroaspis jamesoni; Naja haje (north-east), Naja nigricollis;
Viperidae: Bitis arietans, Bitis gabonica; Echis ocellatus
Thelotornis kirtlandii; Elapidae: Naja katiensis, Naja melanoleuca,4
Naja senegalensis (north-west); Pseudohaje goldii, Pseudohaje nigra;
Viperidae: Atheris squamigera; Bitis nasicornis; Echis leucogaster (north)
Sao Tome and Principe:

Cat 1: Elapidae: Dendroaspis jamesoni; Naja melanoleuca
Cat 2: None
Senegal:
Cat 1: Elapidae: Naja katiensis, Naja nigricollis; Viperidae: Bitis arietans;
Echis leucogaster, Echis jogeri
Cat 2: Colubridae: Dispholidus typus; Elapidae: Dendroaspis polylepis,
Dendroaspis viridis; Naja melanoleuca, Naja senegalensis
Sierra Leone:
Cat 1: Elapidae: Dendroaspis viridis; Naja nigricollis; Viperidae: Bitis arietans

Thelotornis kirtlandii; Elapidae: Naja melanoleuca; 5 Pseudohaje nigra;
Viperidae: Atheris chlorechis; Bitis nasicornis, Bitis rhinoceros
4
The medical importance of this species may be higher in the southern rainforest belt of Nigeria, from
Ibadan in the west to Oban and Eket in the east, and in the forested southern quarter of Sierra Leone.
5
The medical importance of this species may be higher in the forested southern quarter of Sierra Leone.
366
Annex 5
Togo:
Cat 1: Elapidae: Naja nigricollis, Naja senegalensis; Viperidae: Bitis arietans (south);
Echis ocellatus
Thelotornis kirtlandii; Elapidae: Dendroaspis jamesoni, Dendroaspis viridis;
Naja katiensis, Naja melanoleuca; Pseudohaje goldii, Pseudohaje nigra;
Viperidae: Atheris chlorechis; Bitis nasicornis, Bitis rhinoceros
ASIA AND AUSTRALASIA

Central Asia
Armenia:
Cat 2: Viperidae: Montivipera raddei; Vipera eriwanensis, Vipera spp.
Azerbaijan:
Cat 2: Viperidae: Gloydius halys; Vipera eriwanensis, Vipera spp.
Georgia:
Cat 1: Viperidae: Macrovipera lebetina; Vipera ammodytes
Cat 2: Viperidae: Vipera kaznakovi, Vipera renardi, Vipera spp.
Kazakhstan, Kyrgyzstan, Tajikistan, Uzbekistan and Turkmenistan:

Cat 1: Elapidae: Naja oxiana (except Kazakhstan and Kyrgyzstan);
Viperidae: Echis carinatus (except Kyrgyzstan); Macrovipera lebetina (except
Kazakhstan and Kyrgyzstan); Gloydius halys (throughout)
Cat 2: Viperidae: Vipera renardi (except Turkmenistan)
Mongolia:
Cat 1: Viperidae: Gloydius halys
Cat 2: Viperidae: Vipera berus, Vipera renardi
367
East Asia
China:
China mainland
Cat 1: Elapidae: Bungarus multicinctus; Naja atra;
Viperidae: Trimeresurus albolabris; Daboia siamensis;
Deinagkistrodon acutus; Gloydius brevicaudus; Protobothrops
mucrosquamatus
Cat 2: Colubridae: Rhabdophis tigrinus; Elapidae: Bungarus bungaroides (south-east
Tibet), Bungarus fasciatus; Naja kaouthia; Ophiophagus hannah;
Viperidae: Trimeresurus septentrionalis (south Tibet); Gloydius halys,
Gloydius intermedius, Gloydius ussuriensis; Himalayophis tibetanus
(south Tibet); Protobothrops jerdonii, Protobothrops kaulbacki,
Protobothrops mangshanensis; Vipera berus (Jilin, western Xinjiang),
Vipera renardi (western Xinjiang); Trimeresurus stejnegeri
Hong Kong, Special Administrative Region

Cat 1: Elapidae: Bungarus multicinctus; Naja atra; Viperidae: Trimeresurus albolabris
Cat 2: None
Taiwan Province
Cat 1: Elapidae: Bungarus multicinctus; Naja atra;
Viperidae: Protobothrops mucrosquamatus; Trimeresurus stejnegeri
Cat 2: Viperidae: Deinagkistrodon acutus; Daboia siamensis
Democratic People’s Republic of Korea:

Cat 1: Viperidae: Gloydius brevicaudus
Cat 2: Viperidae: Gloydius intermedius, Gloydius ussuriensis; Vipera berus
Japan (including Ryukyu Islands):

Cat 1: Viperidae: Gloydius blomhoffii (main islands); Protobothrops flavoviridis
(Ryukyu Islands)
Cat 2: Colubridae: Rhabdophis tigrinus; Viperidae: Gloydius tsushimaensis
(Tsushima); Protobothrops elegans
Republic of Korea:
Cat 1: Viperidae: Gloydius brevicaudus
Cat 2: Colubridae: Rhabdophis tigrinus; Viperidae: Gloydius intermedius,
Gloydius ussuriensis
368
Annex 5
South Asia
Afghanistan:
Cat 1: Elapidae: Naja oxiana; Viperidae: Echis carinatus; Macrovipera lebetina
Cat 2: Elapidae: Bungarus caeruleus (east), Bungarus sindanus (east); Naja naja (poss.
south-east); Viperidae: Eristicophis macmahonii (south-west); Gloydius halys
(north)
Bangladesh:
Cat 1: Elapidae: Bungarus caeruleus, Bungarus niger, Bungarus walli; Naja kaouthia;
Viperidae: Trimeresurus erythrurus
Cat 2: Elapidae: Bungarus bungaroides, Bungarus fasciatus, Bungarus lividus;
Naja naja; Ophiophagus hannah; Viperidae: Trimeresurus albolabris (far north-
west); Daboia russelii (west)
Bhutan:
Cat 1: Elapidae: Bungarus niger; Naja naja
Cat 2: Elapidae: Bungarus caeruleus, Bungarus fasciatus, Bungarus lividus;
Naja kaouthia; Ophiophagus hannah; Viperidae: Trimeresurus erythrurus;
Daboia russelii; Protobothrops jerdonii
India:
Cat 1: Elapidae: Bungarus caeruleus; Naja kaouthia (east), Naja naja (throughout);
Viperidae: Daboia russelii; Echis carinatus; Hypnale hypnale (south-west)
Cat 2: Bungarus bungaroides, Bungarus fasciatus, Bungarus lividus, Bungarus niger,
Bungarus sindanus, Bungarus walli; Naja oxiana (west), Naja sagittifera
(Andaman Islands); Ophiophagus hannah (south, north-east, Andaman
Islands); Viperidae: Trimeresurus albolabris, Trimeresurus erythrurus,
Trimeresurus septentrionalis; Gloydius himalayanus; Protobothrops jerdonii,
Protobothrops kaulbacki, Protobothrops mucrosquamatus;
Trimeresurus gramineus (south India), Trimeresurus malabaricus (south-west),
Nepal:
Cat 1: Elapidae: Bungarus caeruleus, Bungarus niger; Naja naja, Naja kaouthia;
Viperidae: Daboia russelii
Cat 2: Elapidae: Bungarus bungaroides, Bungarus fasciatus, Bungarus lividus,
Bungarus walli; Ophiophagus hannah; Viperidae: Trimeresurus septentrionalis;
Gloydius himalayanus; Himalayophis tibetanus; Protobothrops jerdonii
369
Pakistan:
Cat 1: Elapidae: Bungarus caeruleus, Bungarus sindanus; Naja naja, Naja oxiana;
Viperidae: Daboia russelii; Echis carinatus
Cat 2: Viperidae: Eristicophis macmahonii (west); Gloydius himalayanus (north);
Macrovipera lebetina (west)
Sri Lanka:
Cat 1: Elapidae: Bungarus caeruleus; Naja naja; Viperidae: Daboia russelii;
Hypnale hypnale
Cat 2: Elapidae: Bungarus ceylonicus; Viperidae: Echis carinatus; Hypnale nepa,
Hypnale zara; Trimeresurus trigonocephalus
South-East Asia
Brunei Darussalam:
Cat 1: Elapidae: Naja sumatrana
Cat 2: Elapidae: Bungarus fasciatus, Bungarus flaviceps; Calliophis bivirgatus,
Calliophis intestinalis; Ophiophagus hannah;
Viperidae: Trimeresurus sumatranus; Tropidolaemus subannulatus
Cambodia:
Cat 1: Elapidae: Bungarus candidus; Naja kaouthia, Naja siamensis;
Viperidae: Calloselasma rhodostoma; Trimeresurus albolabris;
Daboia siamensis
Cat 2: Elapidae: Bungarus fasciatus, Bungarus flaviceps; Ophiophagus hannah;
Viperidae: Trimeresurus cardamomensis
Indonesia (Sumatra, Java, Borneo, Sulawesi and Lesser Sunda Islands):

Cat 1: Elapidae: Bungarus candidus (Sumatra and Java); Naja sputatrix (Java and
Lesser Sunda Islands), Naja sumatrana (Sumatra and Borneo);
Viperidae: Calloselasma rhodostoma (Java); Trimeresurus albolabris;
Daboia siamensis
Cat 2: Elapidae: Bungarus fasciatus, Bungarus flaviceps (Sumatra and Borneo);
Calliophis bivirgatus, Calliophis intestinalis; Ophiophagus hannah
(Sumatra, Borneo & Java); Viperidae: Trimeresurus insularis,
Trimeresurus purpureomaculatus (Sumatra), Trimeresurus sumatranus;
Tropidolaemus subannulatus
370
Annex 5
Lao People’s Democratic Republic:

Cat 1: Elapidae: Bungarus candidus, Bungarus multicinctus; Naja atra (north),
Naja siamensis (south and east); Viperidae: Calloselasma rhodostoma;
Trimeresurus albolabris
Cat 2: Elapidae: Bungarus fasciatus; Naja kaouthia (south and east);
Ophiophagus hannah; Viperidae: Trimeresurus macrops; Protobothrops jerdonii,
Protobothrops mucrosquamatus
Malaysia:
Cat 1: Elapidae: Bungarus candidus (Peninsular Malaysia); Naja kaouthia (northern
Peninsular Malaysia), Naja sumatrana (Peninsular Malaysia, Sabah and
Sarawak); Viperidae: Calloselasma rhodostoma
Calliophis intestinalis; Ophiophagus hannah;
Viperidae: Trimeresurus purpureomaculatus, Trimeresurus hageni;
Myanmar:
Cat 1: Elapidae: Bungarus magnimaculatus, Bungarus multicinctus;
Naja kaouthia, Naja mandalayensis; Viperidae: Trimeresurus albolabris,
Trimeresurus erythrurus; Daboia siamensis
Cat 2: Elapidae: Bungarus bungaroides (Chin State), Bungarus candidus
(Thaninthayi Div.), Bungarus flaviceps (east Shan State), Bungarus niger;
Ophiophagus hannah; Viperidae: Calloselasma rhodostoma (Thaninthayi
Div.); Trimeresurus purpureomaculatus; Protobothrops jerdonii,
Protobothrops kaulbacki, Protobothrops mucrosquamatus (Kachin)
Philippines:
Cat 1: Elapidae: Naja philippinensis (Luzon), Naja samarensis (Mindanao),
Naja sumatrana (Palawan)
Cat 2: Elapidae: Calliophis intestinalis; Ophiophagus hannah;
Viperidae: Trimeresurus flavomaculatus; Tropidolaemus philippensis,
Singapore:
Cat 1: Elapidae: Bungarus candidus; Naja sumatrana
Cat 2: Elapidae: Bungarus fasciatus; Calliophis bivirgatus, Calliophis intestinalis;
Ophiopahgus hannah; Viperidae: Trimeresurus purpureomaculatus
371
Thailand:
Cat 1: Elapidae: Bungarus candidus; Naja kaouthia, Naja siamensis;
Viperidae: Calloselasma rhodostoma; Trimeresurus albolabris;
Daboia siamensis
Calliophis intestinalis; Naja sumatrana; Ophiophagus hannah;
Viperidae: Trimeresurus macrops, Trimeresurus hageni
Timor-Leste:
Cat 1: Viperidae: Trimeresurus insularis
Cat 2: None
Viet Nam:
Cat 1: Elapidae: Bungarus candidus, Bungarus multicinctus, Bungarus slowinskii
(north); Naja atra (north), Naja kaouthia (south);
Viperidae: Calloselasma rhodostoma; Trimeresurus albolabris (throughout)
Cat 2: Elapidae: Bungarus fasciatus, Bungarus flaviceps (south); Naja siamensis
(south); Ophiophagus hannah; Viperidae: Trimeresurus rubeus;
Protobothrops jerdonii, Protobothrops mucrosquamatus (north);
Trimeresurus stejnegeri; Deinagkistrodon acutus
Australo-Papua (including Pacific Islands):

There are no medically important land snakes in American Samoa, Cook Islands,
Fiji, French Polynesia, Guam, Kiribati, Marshall Islands, Nauru, New Caledonia,
New Zealand, Northern Mariana Islands, Pitcairn Island, Samoa, Tokelau, Tonga,
Tuvalu, or Wallis and Futuna Islands. Fiji possesses a single terrestrial venomous
snake species (Ogmodon vitianus) while the Solomon Islands possess three
terrestrial venomous species (Salomonelaps par; Loveridgelaps elapoides and

Parapistocalamus hedigeri) associated with no and few snake-bites, respectively.
Australia:
Cat 1: Elapidae: Notechis scutatus; Pseudechis australis; 6 Pseudonaja affinis,
Pseudonaja mengdeni, Pseudonaja nuchalis, Pseudonaja textilis
Cat 2: Elapidae: Acanthophis antarcticus, Acanthophis cryptamydros, Acanthophis spp.;
Austrelaps spp.; Hoplocephalus spp.; Oxyuranus microlepidotus, Oxyuranus
scutellatus, Oxyuranus temporalis; Pseudechis spp.; Pseudonaja aspidorhyncha,
Pseudonaja spp.; Tropidechis carinatus
Pseudechis australis is common and widespread and causes numerous snake-bites; bites may be severe,
6
although this species has not caused a fatality in Australia since 1968.
372
Annex 5
Indonesia (West Papua and Maluku):

Cat 1: Elapidae: Acanthophis laevis
Cat 2: Elapidae: Acanthophis rugosus; Micropechis ikaheka; Oxyuranus scutellatus;
Pseudechis papuanus, Pseudechis rossignolii; Pseudonaja textilis
Papua New Guinea:

Cat 1: Elapidae: Acanthophis laevis; Oxyuranus scutellatus
Cat 2: Elapidae: Acanthophis rugosus; Micropechis ikaheka; Pseudonaja textilis;
Pseudechis papuanus, Pseudechis rossignolii
EUROPE
There are no venomous snakes in Iceland, Ireland, Isle of Man, Outer Hebrides,
Orkney or the Shetland Islands. Crete and most of the islands of the western
Mediterranean are also free of venomous snakes.
Central Europe
Albania, Bulgaria, Romania, Serbia, Montenegro, Slovenia, The former Yugoslav
Republic of Macedonia:
Cat 1: Viperidae: Vipera ammodytes
Cat 2: Viperidae: Vipera berus
Bosnia and Herzegovina:

Croatia:
Czechia:
Cat 1: None
Greece:
Cat 2: Viperidae: Macrovipera schweizeri; Montivipera xanthina; Vipera berus
373
Hungary:
Cat 1: None
Poland:
Cat 1: None
Slovakia:
Cat 1: None
Eastern Europe
Belarus, Estonia, Latvia, Lithuania, Republic of Moldova:
Cat 1: None
Cat 2: Viperidae: Vipera berus, Vipera nikolskii (Moldova)
Russian Federation:
Cat 2: Viperidae: Gloydius halys, Gloydius intermedius, Gloydius ussuriensis (far-east
Russia); Macrovipera lebetina (Dagestan); Vipera nikolskii, Vipera renardi,
Vipera spp.
Ukraine:
Cat 1: None
Cat 2: Viperidae: Vipera berus, Vipera nikolskii, Vipera renardi
Western Europe
Austria:
Cat 1: None
Cat 2: Viperidae: Vipera ammodytes, Vipera berus
Belgium:
Cat 1: None
374
Annex 5
Denmark:
Cat 1: None
Finland:
Cat 1: None
France:
Cat 1: Viperidae: Vipera aspis
Germany:
Cat 1: None
Italy:
Cat 1: Viperidae: Vipera aspis
Cat 2: Viperidae: Vipera ammodytes, Vipera berus
the Netherlands:
Cat 1: None
Norway:
Cat 1: None
Portugal:
Cat 1: None
Cat 2: Viperidae: Vipera latastei, Vipera seoanei
Spain:
Cat 1: None
Cat 2: Viperidae: Vipera aspis, Vipera latastei, Vipera seoanei
375
Sweden:
Cat 2: None
Switzerland:
Cat 1: None
Cat 2: Viperidae: Vipera aspis, Vipera berus
United Kingdom of Great Britain and Northern Ireland:

Cat 1: Viperidae: Vipera berus (not Northern Ireland)
Cat 2: None
THE AMERICAS
North America
Canada:
Cat 1: None
Cat 2: Viperidae: Crotalus oreganus, Crotalus viridis; Sistrurus catenatus
Mexico:
Cat 1: Viperidae: Agkistrodon bilineatus, Agkistrodon taylori; Crotalus atrox,
Crotalus scutulatus, Crotalus simus, Crotalus molossus; Bothrops asper
Cat 2: Elapidae: Micruroides euryxanthus, Micrurus nigrocinctus, Micrurus tener,
Micrurus spp.; Viperidae: Agkistrodon contortrix, Agkistrodon russeolus;
Atropoides mexicanus, Atropoides occiduus, Atropoides spp.;
Bothriechis schlegelii, Bothriechis spp.; Cerrophidion godmani,
Cerrophidion spp.; Crotalus basiliscus, Crotalus totonacus, Crotalus oreganus,

Crotalus ruber, Crotalus tzabcan, Crotalus viridis, Crotalus spp.; Ophryacus spp.;
Porthidium nasutum, Porthidium spp.; Sistrurus catenatus
United States of America:

Cat 1: Viperidae: Agkistrodon contortrix, Agkistrodon piscivorus;
Crotalus adamanteus, Crotalus atrox, Crotalus horridus, Crotalus oreganus,
Crotalus scutulatus, Crotalus viridis
Cat 2: Elapidae: Micrurus fulvius, Micrurus tener; Viperidae: Crotalus molossus,
Crotalus ornatus, Crotalus ruber, Crotalus spp.; Sistrurus catenatus,
Sistrurus miliarius
376
Annex 5
Central America
The most medically important species are Crotalus simus and Bothrops asper.
Belize:
Cat 1: Viperidae: Bothrops asper
Cat 2: Elapidae: Micrurus spp.; Viperidae: Agkistrodon russeolus;
Atropoides mexicanus; Bothriechis schlegelii; Crotalus tzabcan;
Porthidium nasutum
Costa Rica:
Cat 1: Viperidae: Bothrops asper; Crotalus simus
Cat 2: Elapidae: Micrurus nigrocinctus, Micrurus spp.;
Viperidae: Agkistrodon howardgloydi; Atropoides mexicanus,
Atropoides picadoi; Bothriechis schlegelii, Bothriechis lateralis, Bothriechis spp.;
Cerrophidion sasai; Lachesis melanocephala, Lachesis stenophrys;
Porthidium nasutum, Porthidium ophrymegas, Porthidium spp.
El Salvador:
Cat 1: Viperidae: Crotalus simus
Cat 2: Elapidae: Micrurus nigrocinctus; Micrurus spp.;
Viperidae: Agkistrodon bilineatus; Atropoides occiduus; Bothriechis spp.;
Cerrophidion wilsoni; Porthidium ophryomegas
Guatemala:
Viperidae: Agkistrodon bilineatus, Agkistrodon russeolus; Atropoides mexicanus,
Atropoides occiduus; Bothriechis schlegelii, Bothriechis spp.;
Cerrophidion godmani; Crotalus tzabcan, Porthidium nasutum,
Porthidium ophryomegas
Honduras:
Viperidae: Agkistrodon howardgloydi; Atropoides mexicanus, Atropoides spp.;
Bothriechis marchi, Bothriechis schlegelii, Bothriechis spp.; Cerrophidion wilsoni;
Crotalus simus; Porthidium nasutum, Porthidium ophryomegas
377
Nicaragua:
Viperidae: Agkistrodon howardgloydi; Atropoides mexicanus;
Bothriechis schlegelii; Cerrophidion godmani; Lachesis stenophrys;
Porthidium nasutum, Porthidium ophryomegas
Panama:
Cat 2: Elapidae: Micrurus mipartitus, Micrurus nigrocinctus, Micrurus spp.;
Viperidae: Atropoides mexicanus, Atropoides spp.; Bothriechis lateralis,
Bothriechis schlegelii, Bothriechis spp.; Cerrophidion sasai; Lachesis acrochorda,
Lachesis stenophrys; Porthidium nasutum, Porthidium lansbergii,
Porthidium spp.
Caribbean
No medically important snakes occur naturally in Anguilla, Antigua and Barbuda,
the Bahamas, Barbados, Bermuda, The British Virgin Islands, Cayman Islands,
Cuba, Dominica, the Dominican Republic, Grenada, Guadeloupe, Haiti, Jamaica,
Montserrat, the Netherlands Antilles, Saint Kitts and Nevis, Saint Vincent and
the Grenadines, and Turks and Caicos Islands.
Aruba, Martinique, Saint Lucia, Trinidad and Tobago, and offshore islands:
Cat 1: Viperidae: Bothrops cf. atrox (Trinidad), Bothrops caribbaeus (St Lucia),
Bothrops lanceolatus (Martinique); Crotalus durissus (Aruba)
Cat 2: Elapidae: Micrurus circinalis (Trinidad), Micrurus lemniscatus (Trinidad);
Viperidae: Lachesis muta (Trinidad)
South America
No venomous snakes are occur naturally in the Falkland Islands and no
dangerously venomous snakes occur naturally in Chile.
Argentina:
Cat 1: Viperidae: Bothrops alternatus, Bothrops diporus; Crotalus durissus
Cat 2: Elapidae: Micrurus corallinus, Micrurus lemniscatus, Micrurus spp.;
Viperidae: Bothrops ammodytoides, Bothrops jararaca, Bothrops jararacussu,
Bothrops mattogrossensis, Bothrops neuwiedi, Bothrops pubescens
378
Annex 5
Bolivia (Plurinational State of):

Cat 1: Viperidae: Bothrops atrox, Bothrops mattogrossensis; Crotalus durissus
Cat 2: Elapidae: Micrurus lemniscatus, Micrurus spixii, Micrurus surinamensis,
Micrurus spp.; Viperidae: Bothrocophias hyoprora,
Bothrocophias microphthalmus; Bothrops bilineatus, Bothrops
brazili, Bothrops jararacussu, Bothrops jonathani, Bothrops moojeni,
Bothrops sanctaecrucis, Bothrops spp., Bothrops taeniatus; Lachesis muta
Brazil:
Cat 1: Viperidae: Bothrops atrox, Bothrops jararaca, Bothrops jararacussu,
Bothrops leucurus, Bothrops moojeni; Crotalus durissus
Cat 2: Elapidae: Micrurus corallinus, Micrurus lemniscatus, Micrurus spixii,
Micrurus surinamensis, Micrurus spp.; Viperidae: Bothrocophias hyoprora;
Bothrocophias microphthalmus; Bothrops alternatus, Bothrops bilineatus,
Bothrops brazili, Bothrops diporus, Bothrops mattogrossensis,
Bothrops neuwiedi, Bothrops pubescens, Bothrops taeniatus, Bothrops spp.;
Lachesis muta
Colombia:
Cat 1: Viperidae: Bothrops asper, Bothrops atrox, Bothrops bilineatus;
Crotalus durissus
Cat 2: Elapidae: Micrurus lemniscatus, Micrurus mipartitus, Micrurus nigrocinctus,
Micrurus spixii, Micrurus surinamensis, Micrurus spp.;
Viperidae: Bothriechis schlegelii; Bothrocophias hyoprora,
Bothrocophias microphthalmus, Bothrocophias spp.; Bothrops brazili,
Bothrops taeniatus, Bothrops spp.; Lachesis acrochorda, Lachesis muta;
Porthidium nasutum, Porthidium lansbergii
Ecuador:
Cat 1: Viperidae: Bothrops asper, Bothrops atrox, Bothrops bilineatus;
Lachesis muta
Cat 2: Elapidae: Micrurus lemniscatus, Micrurus mipartitus, Micrurus spixii,
Micrurus surinamensis, Micrurus spp.; Viperidae: Bothriechis schlegelii;
Bothrocophias hyoprora, Bothrocophias microphthalmus, Bothrocophias spp.;
Bothrops brazili, Bothrops taeniatus, Bothrops spp.; Lachesis acrochorda;
Porthidium nasutum, Porthidium spp.
French Guiana (France):

Cat 1: Viperidae: Bothrops atrox, Bothrops bilineatus; Crotalus durissus
Cat 2: Elapidae: Micrurus lemniscatus, Micrurus surinamensis, Micrurus spp.;
Viperidae: Bothrops brazili, Bothrops taeniatus; Lachesis muta
379
Guyana:
Paraguay:
Cat 1: Viperidae: Bothrops alternatus; Crotalus durissus
Cat 2: Elapidae: Micrurus corallinus, Micrurus lemniscatus, Micrurus spixii,
Micrurus spp.; Viperidae: Bothrops diporus, Bothrops jararaca,
Bothrops jararacussu, Bothrops mattogrossensis, Bothrops moojeni,
Bothrops neuwiedi, Bothrops spp.
Peru:
Cat 1: Viperidae: Bothrops atrox, Bothrops bilineatus, Bothrops pictus;
Crotalus durissus; Lachesis muta
Cat 2: Elapidae: Micrurus lemniscatus, Micrurus mipartitus, Micrurus spixii,
Micrurus surinamensis, Micrurus spp.; Viperidae: Bothriechis schlegelii;
Bothrocophias hyoprora, Bothrocophias microphthalmus; Bothrops asper;
Bothrops brazili, Bothrops mattogrossensis, Bothrops taeniatus, Bothrops spp.
Suriname:
Uruguay:
Cat 1: Viperidae: Bothrops alternatus; Bothrops pubescens
Cat 2: Elapidae: Micrurus corallinus, Micrurus spp.; Viperidae: Crotalus durissus
Venezuela (Bolivarian Republic of):

Cat 1: Viperidae: Bothrops atrox, Bothrops cf. atrox, Bothrops venezuelensis;
Crotalus durissus (including Isla de Margarita)
Cat 2: Elapidae: Micrurus circinalis, Micrurus lemniscatus, Micrurus mipartitus,
Micrurus spixii, Micrurus surinamensis, Micrurus spp.;
Viperidae: Bothriechis schlegelii; Bothrops asper, Bothrops brazili,
Bothrops bilineatus; Lachesis muta; Porthidium lansbergii
380
Annex 5
Herpetological references 7
Abtin E, Nilson G, Mobaraki A, Hosseini AA, Dehgannejhad M. A new species of Krait, Bungarus (Reptilia,
Elapidae, Bungarinae) and the first record of that genus in Iran. Russ J Herpetol. 2014;21(4):243–50.
Ananjeva NB, Orlov NL, Khalikov RG, Darevsky IS, Ryabov IS et al. The reptiles of Northern Eurasia. Sofia:
Pensoft; 2006.
Anderson CG, Greenbaum E. Phylogeography of northern populations of the Black-Tailed Rattlesnake
(Crotalus molossus Baird and Girard, 1853), with the revalidation of C. ornatus Hallowell, 1854. Herpetol
Monogr. 2012;26(1):19–57.
Ashton KG, de Querioz A. Molecular systematics of the western rattlesnake, Crotalus viridis (Viperidae),
with comments on the utility of the D-loop in phylogenetic studies of snakes. Mol Phylogenet Evol.
2001;21:176–89.
Babocsay G. A new species of saw-scaled viper of the Echis coloratus complex (Ophidia: Viperidae) from
Oman, Eastern Arabia. Syst Biodivers. 2004;1:503–14.
Basoglu M, Baran I. The reptiles of Turkey. Part II. The snakes. Izmir: University of Matbaasi; 1980.
Buys PJ, Buys PJC. Snakes of Namibia. Windhoek; Gamsberg Macmillan; 1983.
Bons J, Geniez P. Amphibiens et reptiles du Maroc. Barcelona; Asociación Herpetológica Española; 1996.
Boycott RC. A herpetological survey of Swaziland [MSc thesis]. Durban: University of Natal; 1992.
Branch B. Field guide to the snakes and other reptiles of Southern Africa. Cape Town: Sruik; 1988.
Broadley DG, Doria CT, Wigge J. Snakes of Zambia: An atlas and field guide. Frankfurt am Main: Edition
Chimair; 2003.
Broadley DG. FitzSimons snakes of Southern Africa. Cape Town: Delta Books; 1983.
Broadley DG. The herpetofaunas of the islands off the coast of south Moçambique. Arnoldia Zimbabwe.
1990;9:469–93.
Broadley DG. Reptiles and amphibians from the Bazaruto Archipelago, Mozambique. Arnoldia
Zimbabwe. 1992;9:539–48.
Broadley DG. Review of the Dispholidini, with the description of a new genus and species from Tanzania
(Serpentes, Colubridae). Bull Br Mus Nat Hist Zool. 2002;68:57–74.
Broadley DG. A review of the genus Thelotornis A. Smith in eastern Africa, with the description of a
new species from the Usambara Mountains (Serpentes: Colubridae: Dispholidini). Afr J Herpetol.
2001;50:53–70.
Broadley DG. The herpetofaunas of the islands off the coast of south Moçambique. Arnoldia Zimbabwe.
1990;9:469–493.
Campbell JA, Lamar WW. The venomous snakes of the western hemisphere. Vols. I and II. Ithaca, NY:
Comstock-Cornell; 2004.
Creer S, Malhotra A, Thorpe RS, Stöcklin RS, Favreau PS, Hao Chou WS. Genetic and ecological
correlates of intraspecific variation in pitviper venom composition detected using matrix-assisted
laser desorption time-of-flight mass spectrometry (MALDI-TOF-MS) and isoelectric focusing. J Mol
Evol. 2003;56:317–29.
da Silva VX. The Bothrops neuwiedi complex. In: Campbell JA, Lamar WW, editors. The venomous
reptiles of the Western Hemisphere. Vol. I. Ithaca (NY): Comstock Publishing; 2004:410–22.
Major regional guides have author names italicized.

7
381
David P, Ineich I. Les serpents venimeux du monde: systématique et répartition. Dumerilia. 1999;
3:3–499.
David P, Vogel G, Dubois A. On the need to follow rigorously the Rules of the Code for the subsequent
designation of a nucleospecies (type species) for a nominal genus which lacked one: the case of the
nominal genus Trimeresurus Lacépède, 1804 (Reptilia: Squamata: Viperidae). Zootaxa. 2011;2992:1–51.
De Smedt J. The vipers of Europe. Hablech: JDS Verlag; 2001.
Dobiey M, Vogel G. Venomous snakes of Africa. Frankfurt am Main: Terralog Edition Chimaira; 2006.
El Din SB. A guide to the reptiles and amphibians of Egypt. Cairo: American University of Cairo Press;
2006.
Gasperetti J. Snakes of Arabia. Fauna Saudi Arabia. 1988;9:169–450.
Geniez P, Mateo A, Geniez M, Pether J. The amphibians and reptiles of the Western Sahara. Frankfurt am
Main: Edition Chimaira; 2004.
Gloyd HK, Conant R. Snakes of the Agkistrodon complex. Oxford: OH, SSAR; 1990.
Gower DJ, Wade EOZ, Spaels S, Böhme W, Buechley ER, Sykes D, Colston TJ. A new large species of Bitis
Gray, 1842 (Serpentes: Viperidae) from the Bale Mountains of Ethiopia. Zootaxa. 2016;4093(1):41–63.
Guo P, Malhotra A, Li PP, Creer S. New evidence on the phylogenetic position of the poorly known Asian
pitviper Protobothrops kaulbacki (Serpentes: Viperidae: Crotalinae) with a redescription of the species
and a revision of the genus Protobothrops. Herpetol J. 2007;17:237–46.
Gumprecht A et al. Asian Pitvipers. Berlin: Geitje Books; 2004.
Gutberlet RL Jr, Campbell JA. Generic recognition of a neglected lineage of South American pitvipers
(Squamata: Viperidae: Crotalinae), with the description of a new species from the Colombian Chocó. Am
Museum Novitates. 2001;3316:1–15.
Jadin, RC, Townsend, JH, Castoe TA, Campbell JA. Cryptic diversity in disjunct populations of Middle
American Montane Pitvipers: a systematic reassessment of Cerrophidion godmani. Zool Scripta.
2012;41:455–70.
Kästle W, Rai K, Schleich H. Amphibians and reptiles of Nepal. Ruggall: Gantner Verlag; 2002.
Khan MS. A guide to the snakes of Pakistan. Khan MS. A guide to the snakes of Pakistan. Frankfurt am
Main: Edition Chimaira; 2002.
Kreiner G. The snakes of Europe. Frankfurt am Main: Edition Chimaira; 2007.
Latifi M. The snakes of Iran. Oxford: OH:SSAR; 1985.
Le Berre M. Faune du Sahara 1: Poissons-amphibiens-reptiles. Paris: Editions Raymond Chabaud; 1989.

Lenk P, Herrmann H-W, Joger U, Wink M. Phylogeny and taxonomic subdivision of Bitis (Reptilia:
Viperidae) based on molecular evidence. In: Joger U, editor. Phylogeny and systematics of the Viperidae.
Kaupia, Darmstädter Beiträge zur Naturgeschichte. 1999;8:31–8.
Lenk P, Kalyabina S, Wink M, Joger U. Evolutionary relationships among the true vipers (Reptilia:
Viperidae) inferred from mitochondrial DNA sequences. Mol Phylogenet Evol, 2001;19:94–104.
Leviton AE, Anderson SE, Adler K, Minton SA. Handbook to Middle East amphibians and reptiles. Oxford:
OH, SSAR; 1992.
Leviton AE, Wogan G, Koo M, Zug GR, Lucas R, Vindum J. The dangerously venomous snakes of Myanmar.
Illustrated checklist with keys. Proc Calif Acad Sci. 2003;54:407–62.
Maddock ST, Ellis RJ, Doughty P, Smith LA, Wuster W. A new species of death adder (Acanthophis:
Serpentes: Elapidae) from north-western Australia. Zootaxa. 2015;4007(3):301–26
Maduwage K, Silva A, Manamendra-Arachchi K, Pethiyagoda R. A taxonomic revision of the South Asian
hump-nosed pit vipers (Squamata: Viperidae: Hypnale). Zootaxa. 2009;2232:1–28.
382
Annex 5
Malhotra A, Thorpe RS. A phylogeny of four mitochondrial gene regions suggests a revised taxonomy
for Asian pitvipers (Trimeresurus and Ovophis). Mol Phylogenet Evol. 2004;32:83–100.
McDiarmid RW, Campbell JA, Toure T. Snake species of the world: A taxonomic and geographical
reference. Vol. 1. Wahington, DC: Herpetologists’ League: 1999.
Murphy JC. Amphibians and Reptiles of Trinidad and Tobago. Malabar, FL: Krieger; 1997.
Nilson G, Rastegar-Pouyani N. Walterinnesia aegyptia Lataste, 1887 (Ophidia: Elapidae) and the status of
Naja morgani Mocquard 1905. Russ J Herpetol. 2007;14:7–14.
Orlov NL, Barabanov AV. Analysis of nomenclature, classification, and distribution of the Agkistrodon
halys – intermedius complexes: a critical review. Russ J Herpetol. 1999;6:167–92.
O’Shea M. A guide to the snakes of Papua New Guinea. Port Moresby: Independent Publishing; 1996.
Parkinson CL Zamudio KR, Greene HW. Phylogeography of the pitviper clade Agkistrodon: historical
ecology, species status, and conservation of cantils. Mol Ecol. 2000;9:411–20.
Pitman CRS. A guide to the snakes of Uganda, second edition. Codicote, Wheldon & Wesley; 1974.
Porras LW, Wilson LD, Schuett GW, Reiserer R.S. A taxonomic reevaluation and conservation assessment
of the common cantil, Agkistrodon bilineatus (Squamata: Viperidae): a race against time. Amphib
Reptile Conserv. 2013;7(1):48–73.
Pyron RA, Burbrink FT, Colli GR, de Oca AN, Vitt LJ, Kuczynski CA et al. The phylogeny of advanced snakes
(Colubroidea), with discovery of a new subfamily and comparison of support methods for likelihood
trees. Mol Phylogenet Evol. 2011;58(2):329–42. doi:10.1016/j.ympev.2010.11.006.
Pyron RA, Dushantha Kandambi HK, Hendry CR, Pushpamal V, Burbrink FT, Somaweera R. Genus-
level phylogeny of snakes reveals the origins of species richness in Sri Lanka. Mol Phylogenet Evol.
2013;66(3):969–78
Pyron RA, Burbrink FT, Wiens JJ. A phylogeny and revised classification of Squamata, including 4161
species of lizards and snakes. BMC Evol Biol. 2013;13:93.
Quijada-Mascareñas JA Ferguson JE, Pook CE, Da Graca Saloma M, Thorpe RS, Wuester W. Phylogeographic
patterns of trans-Amazonian vicariants and Amazonian biogeography: the Neotropical rattlesnake
(Crotalus durissus complex) as an example. J Biogeogr 2007;34:1296–312.
Roman B. Serpents de Haute-Volta. Ouagadougou: CNRST; 1980.
Schleich HH, Kastle W, Kabish K. Amphibians and reptiles of North Africa. Koenigstein: Koeltz; 1996.
Spawls S, Branch B. The dangerous snakes of Africa. London: Blandford; 1995.
Spawls S, Howell A, Drewes R. A field guide to the reptiles of East Africa: Kenya, Tanzania, Uganda,
Rwanda and Burundi. London: Academic Press, Natural World; 2002.
Steward JW. The snakes of Europe. Newton Abbott: David & Charles; 1971.
Street D. Reptiles of northern and central Europe. London: Batsford; 1979.
Sweeney RCH. Snakes of Nyasaland. Amsterdam: Asher & Co; 1971.
Thorpe RS, Pook CE, Malhotra A. Phylogeography of the Russell’s viper (Daboia russelii) complex in
relation to variation in the colour pattern and symptoms of envenoming. Herpetol J. 2007;17:209–18.
Visser J, Chapman DS. Snakes and snakebite. Cape Town: Struik; 1978.
Vogel G. Venomous snakes of Asia. Frankfurt am Main: Terralog Edition Chimaira; 2006.
Vogel G, David P, Lutz M, van Rooijen, Vidal N. Revision of the Tropidolaemus wagleri-complex
(Serpentes: Viperidae: Crotalinae). I. Definition of included taxa and redescription of Tropidolaemus
wagleri (Boie, 1827). Zootaxa. 2007;1644:1–40.
Wallach V, Williams KL, Boundy J. Snakes of the world: A catalogue of living and extinct species. CRC
Press: London; 2014.
383
Whitaker R, Captain A. Snakes of India: The field guide. Chennai: Draco Books; 2004.
Williams DJ, O’Shea M, Daguerre RL. Origin of the eastern brownsnake, Pseudonaja textilis (Duméril,
Bibron and Duméril) (Serpentes: Elapidae: Hydrophiinae) in New Guinea: evidence of multiple
dispersals from Australia, and comments on the status of Pseudonaja textilis pughi Hoser 2003.
Zootaxa. 2008;1703:47–61.
Williams DJ, Wüster W. Snakes of Papua New Guinea. In: Willliams DJ et al., editors. Venomous bites and
stings in Papua New Guinea. Melbourne: AVRU University of Melbourne; 2005:33–64.
Wüster W et al. Snakes across the strait: trans-Torresian phylogenetic relationships in three genera
of Australo-Papuan snakes (Serpentes: Elapidae: Acanthophis, Oxyuranus and Pseudechis). Molecular
Phylogenetics and Evolution, 2005, 34:1–14.
Wuster W, Broadley DG. A new species of spitting cobra from north-eastern Africa (Serpentes: Elapidae:
Naja). J Zool London. 2003;259:345–59.
Wüster W, Broadley DG. Get an eyeful of this: a new species of giant spitting cobra from eastern and
north-eastern Africa (Squamata: Serpentes: Elapidae: Naja). Zootaxa. 2007;1532:51–68.
Wuster W, Wüster W, Salomao MG, Quijada-Mascarenas JA, Thorpe RS. Origin and evolution of the
South American pitviper fauna: evidence from mitochondrial DNA sequence analysis. In: Schuett
GW, Höggren M, Douglas ME, Greene HW, editors. Biology of the vipers. Eagle Mountain, Utah: Eagle
Mountain Publishing; 2002:111–128.
Wüster W, Thorpe RS, da Graça Salomão M, Thomas L. Origin and phylogenetic position of the Lesser
Antillean species of Bothrops (Serpentes: Viperidae): biogeographical and medical implications. Bull.
Nat. Hist. Mus. London, Zool. 2002;68:101–6.
Wüster W, Warrell DA, Cox MJ, Jintakune P, Nabhitabhata J. Redescription of Naja siamensis Laurenti,
(Serpentes: Elapidae), a widely overlooked spitting cobra from Southeast Asia: geographic variation,
medical importance and designation of a neotype. J Zool, 1997; 243:771–88.
Wüster W Salomão MG, Thorpe RS, Puorto G, Furtado MFD, Hoge SA, et al. Systematics of the
Bothrops atrox species complex: insights from multivariate analysis and mitochondrial DNA sequence
information. In: Thorpe RS, Wüster W, Malhotra A. editors. Venomous snakes: ecology, evolution and
snakebite. Oxford: Clarendon Press; 1997:99–113 (Symposia of the Zoological Society of London,
No. 70).
Wüster W, Crookes S, Ineich I, Mane Y, Pook CE, Trape J-F et al. The phylogeny of cobras inferred from
mitochondrial DNA sequences: evolution of venom spitting and the phylogeography of the African
spitting cobras (Serpentes: Elapidae: Naja nigricollis complex). Mol Phylogenet Evol. 2007;45:437–53.
Wüster W, Golay P, Warrell DA. Synopsis of recent developments in venomous snake systematics.
Toxicon. 1997;35:319–40.
Wüster W, Golay P, Warrell DA. Synopsis of recent developments in venomous snake systematics, No. 2.
Toxicon. 1998;36:299–307.
Wüster W, Golay P, Warrell DA. Synopsis of recent developments in venomous snake systematics, No. 3.
Toxicon. 1999;37:1123–29.
Wüster W, McCarthy CJ. Venomous snake systematics: Implications for snake bite treatment and
toxinology. In: Bon C, Goyffon M, editors. Envenomings and their treatments. Lyon: Fondation Mérieux;
1996:13–23.
Zhao E, Adler K. Herpetology of China. Oxford: OH, SSAR; 1993.
384
Annex 5
Appendix 2
Model protocol for the production and testing of snake
antivenom immunoglobulins
Identification of the lot
Name and address of manufacturer
Lot number of antivenom

Date of filling
Liquid or freeze-dried
Expiry date
Number of vials or ampoules
Temperature of storage
Control of the venom batch(es) used for animal immunization

Producer of venom and location
Information on the snake contributing to the
venom batch
Scientific names of the snake species
Number of snakes
Geographical origins of the snakes
Dates of collection of the venoms
Expiry date of the venoms preparation
Biochemical and biological characterization
of the venoms
–– Test performed
–– Results
Control of plasma donor animals

Location of the animal herd
Animal species used for immunization
Vaccinations performed on animals
Dates of immunization
Control of antivenom antibody titre of animal
Veterinary certificate of health of animal donor
385
Collection and storage of plasma

Method of collection
Date of collection
Date of storage
Type of containers
Temperature of storage
Type and content of preservatives added (if any)
Transport of plasma to fractionation facility

Date of transport
Temperature of transport
Date of arrival
Plasma pooling and fractionation

Temperature of plasma storage at fractionation facility
Volume of plasmas of different specificity pooled for the production
of polyspecific antivenoms (if applicable)
Date of plasma pooling
Volume of the manufacturing plasma pool
Number of animal donors contributing to the
manufacturing plasma pool
Quality control of manufacturing plasma pool
–– Results
Type of active substance (intact IgG, fragments)
Preparation and control of final bulk

Volume of bulk antivenoms of different specificity
pooled for the production of polyspecific antivenoms
(if applicable)
Concentration of preservatives (if used)
–– Type
–– Method
–– Result
Quality control of manufacturing plasma pool
–– Results
386
Annex 5
Filling and containers

Date of filling
Quantity of containers
Volume of antivenoms per container
Date of freeze-drying (if applicable)
Control tests on final product

Appearance
Solubility (freeze-dried product)
Extractable volume
Venom-neutralizing potency assay
–– Method
–– Venom used
–– Results
Osmolality
Identity test
–– Method
–– Result
Protein concentration
–– Method
–– Result
Purity
–– Method
–– Result
Molecular size distribution
–– Method
–– Result
Test for pyrogenic substances
–– Method
–– Result
Sterility test
No. of containers examined
–– Method
–– Date at start of test
–– Date at end of test
Concentration of sodium chloride and other excipients
–– Method
–– Result
387
Determination of pH
–– Result
Concentration of preservatives (if used)
–– Type
–– Method
–– Result
Chemical agents used in plasma fractionation
–– Type
–– Method
–– Result
Inspection of final containers
–– Results
Residual moisture in freeze-dried antivenoms
–– Method
–– Result
Internal certification
Certification by person taking overall responsibility for production of the antivenom
I certify that batch no. of
snake antivenom immunoglobulin meets all national requirements and/or
satisfies the 2016 WHO Guidelines for the production, control and regulation
of snake antivenom immunoglobulins.1
Signature
Name (typed)
Date
WHO Technical Report Series, No. 1004, Annex 5.

1
388
Annex 6
WHO manual for the preparation of secondary reference
materials for in vitro diagnostic assays designed for
infectious disease nucleic acid or antigen detection:
calibration to WHO International Standards
1. Background 392
3. Terminology 393
4. Principles of biological standardization 395
5. Calibration hierarchy of biological standards 395
5.1 WHO International Standards 398
5.2 Secondary standards 399
5.3 Tertiary standards 400
5.4 Other control material 400
6. Commutability of biological standards 400
7. Selection and characterization of materials for the preparation
of secondary standards – and application to tertiary standards 401
7.1 Analyte – biological versus synthetic material 401
7.2 Immunological and genetic diversity 402
7.3 Matrix 403
7.4 Concentration 403
7.5 Volume 403
7.6 Diagnostic specificity 404
7.7 Infectivity 404
7.8 Physical appearance 404
7.9 Homogeneity 405
7.10 Stability 405
7.11 Stability assessment during product lifetime 406
7.12 In-use stability 407
8. Calibration: testing and statistical analysis 408
8.1 Principles of calibration 408
8.2 Single assay calibration using qualitative tests 409
8.3 Single assay calibration using quantitative tests 410
8.4 Collaborative study calibration using multiple assays 412
8.5 Calculation of uncertainty of measurement 412
389
9. Establishment of reference materials in the absence of a WHO

International Standard 413
10. Post-production considerations 414
10.1 Storage of the material 414
10.2 Distribution of the material 414
10.3 Instructions for Use 415
10.4 Replacement batches 415
11. End user advice 415
References 416
Appendix 1 Example of the parallel calibration of a secondary standard in a
study to establish the International Standard 419
Appendix 2 Example of the calibration of a national standard (collaborative study
calibration using multiple assays) 444
Appendix 3 Example of the calibration of a reference preparation by a single
NAT assay 450
Guidance documents published by the World Health Organization

(WHO) are intended to be scientific and advisory in nature. Each of
the following sections constitutes guidance for national regulatory
authorities (NRAs) and for manufacturers of biological products.
390
Annex 6
Abbreviations
Ag antigen
% CV percentage coefficient of variation
Ct cycle threshold
EQA external quality assurance
IS International Standard(s)
MU measurement uncertainty
PT proficiency testing
S/CO sample-to-cut-off (ratio); also signal-to-cut-off (ratio)
Système international d’unités (measurement system using
SI
metric units)
SoGAT Standardisation of Genome Amplification Techniques (group)
391
1. Background
Through its Expert Committee on Biological Standardization, WHO developed
its Guidelines for the preparation and establishment of reference materials and
reference reagents for biological substances in 1978 (1). This document was
last revised in 2004 (2). In the revised WHO Recommendations document,
secondary standards are defined as reference preparations established by regional
or national authorities, or by other laboratories, that are calibrated against and
traceable to WHO reference materials. Part B of the 2004 Recommendations deals
with general considerations for the preparation, characterization and calibration
of regional or national biological reference standards.
Feedback from manufacturers and providers of secondary (for example,
regional) standards used for in vitro diagnostic (IVD) devices, and from
regulatory authorities, international trade organizations, IVD manufacturers,
providers of external quality assurance (EQA) or proficiency testing (PT)
programmes and laboratories using diagnostic assays, indicated a need for more
specific guidance on the preparation of secondary standards; it was therefore
concluded that a practical manual focusing on IVD needs would be helpful.
This topic was discussed at the 2012 meeting of the WHO Expert Committee on
Biological Standardization, at which the proposal to generate a WHO document
on secondary standards for use in the IVD field was endorsed.

This WHO document provides practical guidance on the preparation of
secondary biological reference materials and on their calibration to WHO
International Standards (IS) where available. The document focuses on the in
vitro measurement procedures used for diagnosis, detection and management
of infectious diseases where the typical analytes (measurands) are nucleic acid
or antigen (Ag). These IVD tests cover nucleic acid amplification technique
(NAT)-based assays for detecting the DNA or RNA of infectious agents and
immunological tests for the detection of Ag(s) of infectious agents. Currently,
there are only a small number of IS with an assigned unitage available where
the analyte is an antibody directed to an infectious agent. Due to their complexity
(that is, the epitope spectrum represented by polyclonal antibodies in the
serum of a patient) this document does not cover antibody-based secondary
standards. However, several principles outlined in this manual may also apply to
antibody assays. Where applicable, the document integrates existing guidance,
referenced accordingly.
The document is intended for use by manufacturers of secondary
reference materials, IVD manufacturers, providers of EQA or PT programmes
392
Annex 6
and other laboratories using reference materials for NAT-based and serological
infectious disease assays. Analogous guidance has already been issued by WHO
on secondary standards for vaccines (3) and chemical reference substances (4).
3. Terminology
The definitions given below apply to the terms as used in this WHO guidance
document. These terms may have different meanings in other contexts.
Accuracy: (measurement) closeness of agreement between a measured
quantity value and the true quantity value of a measurand (5).
Biological matrix: a discrete material of biological origin that can be
sampled and processed in a reproducible manner. Examples include blood,
serum, plasma, urine, faeces, saliva, sputum and various discrete tissues (6).
Calibration: a process that, under specified conditions, establishes as a
first step a relation between the quantity values with measurement uncertainties
provided by measurement standards and corresponding indications with
associated measurement uncertainties and, in a second step, uses this information
to establish a relation for obtaining a measurement result from an indication (5).
Calibration hierarchy: a sequence of calibrations from a reference to the
final measuring system, where the outcome of each calibration depends on
the outcome of the previous calibration (5).
Commutability (of a reference material): a property of a reference
material, demonstrated by the equivalence of the mathematical relationships
among the results of different measurement procedures for a reference material
and for representative samples of the type intended to be measured (7).
Control material: a substance, material or article used to verify the
performance characteristics of an in vitro diagnostic (IVD) medical device (8).
Diagnostic specificity: the probability that the device gives a negative
result in the absence of the target marker (9).
End-point titre: the reciprocal of the highest analyte dilution that gives a
reading above the assay cut-off (10).
International measurement standard: a measurement standard
recognized by signatories to an international agreement and intended to serve
worldwide, for example a WHO International Standard (IS) (5).
International conventional calibrator: a calibrator whose value of a
quantity is not metrologically traceable to SI units but is assigned by international
agreement (11).
International Unit(s) (IU): the unitage assigned by WHO to an
International Biological Standard (2).
Linearity (of a measuring system): the ability to provide measured
quantity values that are directly proportional to the value of the measurand in
the sample (12).
393
Potency: the specific ability or capacity of a product, as indicated by

appropriate laboratory tests or by adequately controlled clinical data obtained
through the administration of the product in the manner intended, to effect a
given result (13).
Precision: (measurement) closeness of agreement between indications
or measured quantity values obtained by replicate measurements on the same or
similar objects under specified conditions (5).
Relative potency: a measure obtained from the comparison of a test to a
standard on the basis of capacity to produce the expected potency (13).
Reference material: a material, sufficiently homogeneous and stable
with regard to specified properties, which has been established to be fit for its
intended use in measurement or in the examination of nominal properties (5).
Reference standard: a measurement standard designated for the
calibration of other measurement standards for quantities of a given kind in a
given organization or at a given location (5).
Sample-to-cut-off (S/CO) ratio (also signal-to-cut-off ratio): S/CO
ratios are calculated by dividing the signal value (for example, optical density or
relative light unit) of the sample being tested by the signal value of the enzyme
immunoassay or chemiluminescence assay cut-off for that run. If the signal
produced by a given test sample is equal to or greater than the calculated cut-off
value then the specimen is considered to be reactive in the test. In competitive
assays the relationship between the signal value of the sample and the signal
value of the cut-off is reversed (CO/S ratio) (14).
Secondary (reference) standards: reference standards established by
regional or national authorities, or by other laboratories, that are calibrated
against and traceable to WHO reference materials (2).
SI, International System of Units: a system of units – based on the
International System of Quantities – their names and symbols (including a series
of prefixes and their names and symbols), together with rules for their use,
adopted by the General Conference on Weights and Measures (5).
Traceability: (metrological) property of a measurement result whereby
the result can be related to a reference through a documented unbroken chain
of calibrations, each contributing to the measurement uncertainty (5).
Threshold cycle: the polymerase chain reaction (PCR) cycle at which
the gain in fluorescence generated by the accumulating amplicon exceeds a
threshold over baseline – for example, defined as 10 standard deviations of the
mean baseline fluorescence using data taken from cycles 3 to 15 (15).
Working standard: a measurement standard used routinely to calibrate
or verify measuring instruments or measuring systems (5).
394
Annex 6
Uncertainty: a parameter associated with the result of a measurement

that characterizes the dispersion of the values that could reasonably be attributed
to the measurand (16).
4. Principles of biological standardization

The aim of metrological traceability is to enable the results obtained by the
calibrated routine measurement procedure to be expressed in terms of the values
obtained at the highest available level of the calibration hierarchy (11). This is
usually achieved in the clinical chemistry field by physicochemical reference
methods, obtaining values in SI units.
The majority of biological samples, containing for example nucleic
acids, antigens or antibodies, are substances that cannot be fully characterized
by a physicochemical reference method. Instead, biological assays are used for
measurement of the potency or content of the analyte of interest. These methods
are heterogeneous and the lack of a reference method does not permit the results
to be expressed in absolute values according to the SI system.
The approach taken by the WHO Expert Committee on Biological
Standardization to quantify biological materials is to first establish a highest order
reference reagent – the IS. The procedure for the preparation, characterization
and establishment of WHO IS preparations is described in detail elsewhere (2).
Such material plays a crucial role in the standardization, harmonization and
quality control of IVD assays, as was demonstrated in the 1990s when WHO
IS were introduced for human immunodeficiency virus, hepatitis C virus and
hepatitis B virus. These reference materials were fundamental in the regulation
of IVD assays used for blood safety and for improving patient management in
the clinical setting (17).
5. Calibration hierarchy of biological standards

“Reference material” is a generic term which refers to a material or substance
whose property values are sufficiently homogeneous and stable, and whose
fitness for purpose is well established, for its intended use in a measurement
process (for example, the assessment of a measurement method or the assigning
of values to materials) (18, 19). Biological reference materials for a given
analyte can be related through a sequence of comparisons to create a calibration
hierarchy traceable to the highest order material – the WHO IS (Fig. A6.1).
395
Fig. A6.1
Calibration hierarchy and metrological traceability to the WHO IS
WHO International Standard (highest

order, international conventional
calibrator)
Uncertainty of measurement
Traceability
Secondary standard (calibrator for

assays and tertiary standards)
Tertiary standard (working reagent;

kit control; run control; calibrator)
Furthermore, all biological standards have a range of key properties as

summarized in Table A6.1.
Table A6.1
Key properties of WHO IS, secondary standards and tertiary standards
Property WHO IS Secondary standard Tertiary standard

Alternative Highest order, Regional or national Working reagents
names international reference materials, or standards,
conventional laboratory or manufacturer’s

calibrator manufacturer’s product calibrator,
working calibrator control material
Calibration Evaluated in an Calibrated against Calibrated against
international the WHO IS the secondary
collaborative standard
study, involving
laboratories
worldwide, different
assays and different
types of test
laboratories (usually
15–30 participants)
396
Annex 6

Unitage IU/mL IU/mL IU/mL
Traceability N/A Yes Yes
Uncertainty of No Yes (assay-specific) Yes (assay-specific)
measurement
Commutability Must be determined Should be Consideration
experimentally determined should be given
relative to clinical experimentally to experimentally
specimens relative to clinical determining relative
specimen to clinical specimen
Material Should resemble, Should resemble, as Should resemble, as
as closely and as closely as possible, closely as possible,
feasibly as possible, the analyte to be the analyte to be
the analyte being measured. However, measured. Biological
measured – for for assay-specific material similar to the
example, for assays secondary standards, tested sample, or non-
for viral nucleic synthetic materials biological materials,
acids the standard such as armored such as armored
will be the wild-type RNA, plasmids RNA, plasmids
patient-derived and recombinant and recombinant
virus in plasma (the proteins, may be proteins may be used,
normal sample type used and laboratories and laboratories
analysed) are encouraged are encouraged
to address to address
commutability commutability
Typical final Lyophilized Lyophilized or liquid Liquid
format of
standard
Usage Calibration Calibration of Working standards;
of secondary tertiary standards; run control;
standards; initial working standards; calibrator
validation of new run control;
assay/platform calibrator
397

Establishment International May be calibrated in 1. Assay-specific
of standard agreement through several ways: study, normally by a
a WHO international single laboratory for
1. In parallel with a
collaborative use with a specific
study to establish
study, proposal test/platform.
the IS
for adoption
Example: 2. Small study by a
and subsequent
Appendix 1 limited number of
establishment by
laboratories with
the WHO Expert 2. Regional
a single assay or a
Committee or national
limited number of
on Biological collaborative
different assays/
Standardization study similar
platforms
to the WHO
collaborative
study but with
fewer participants
Example:
Appendix 2
3. Small study by
one or a limited
number of
laboratories with
a single assay or a
limited number of
different assays/
platforms
Example:
Appendix 3
5.1 WHO International Standards

WHO IS are defined by ISO17511 as International Conventional Calibrators
and are the highest order of standard for biological references. They are solely
established by the WHO Expert Committee on Biological Standardization
following specific guidance (2).
Establishment of a WHO IS follows a collaborative study involving
various users of the material (including national control laboratories, IVD
manufacturers and other certified laboratories) and as many different, well-
established assays as feasible. The laboratories should be chosen to reflect the
global use of the standard and consideration should be given to the expertise
of laboratories with a proven track record (perhaps through EQA schemes).
398
Annex 6
As with all biological references, the material used should resemble as closely as
possible the natural analyte of the clinical sample to be measured. An assessment
of commutability should be performed as part of the collaborative study where
appropriate and feasible (2).
By definition, an IS has a specified value expressed in International Units
(IU). This value is arbitrarily assigned based on the results of the collaborative
study. The assigned IU value of each IS (new and replacement) does not carry an
uncertainty associated with calibration (see section 8.1 below). The uncertainty
is considered to be the variance of the vial weight determined during the filling
process (20, 21). The collaborative study design attempts to ensure continuity
of the IU as far as possible. As explained further in section 8.1 there is no
uncertainty value associated with replacements.
As the highest-order standard for biological material, the use of a
WHO IS should be limited to the calibration of secondary biological reference
materials in order to minimize the need to replace the IS on a regular basis.
Unfortunately, the limited availability of secondary calibrator standards and the
lack of specific guidance on the establishment and calibration of more readily
available standards have resulted in the overuse of WHO IS for more routine
procedures such as validation of assays and as run controls.
5.2 Secondary standards

A secondary standard is a material 1 that has been directly calibrated against the
IS. These preparations usually include regional or national reference preparations.
The titre, composition and method of production of secondary standards
will vary but should be suitable for obtaining sufficient measurements, when
dilution is needed, to achieve an accurate calibration. Regardless of the method
of production, each calibration will have a stated measurement uncertainty
(see section 8.1 below).
Secondary standards should be used for the calibration of tertiary
standards. They should also be used for the calibration and validation of
assay systems.
Biological stock materials assigned a potency based on calibration against the IS by using exclusively
1
one specific test, and used, for example, for the preparation of calibrators or run controls for this test,
are also considered as in-house secondary standards. Non-biological preparations such as synthetic
preparations (for example, plasmid preparations, transcripts, armored RNA and antigens produced by
recombinant DNA technology) are often used for test calibration. If this calibration is done against an
IS, these materials will also be considered as secondary standards under the scope of this document.
Nevertheless, these materials have a number of limitations compared to the biological preparations, for
example, in terms of commutability.
399
5.3 Tertiary standards

Tertiary standards are calibrated against secondary standards using the same
calibration procedure. These standards usually include working standards
or calibrators established for one specific assay used by a laboratory or
other institution.
The standard may be formulated from either biological (for example,
patient-derived) or non-biological material. However, regardless of the material
used, all references in the traceability chain should also demonstrate
commutability to the clinical sample of the tested analyte. Many of the principles
discussed in this document also apply to the development of tertiary standards.
Tertiary standards are typically formulated as a liquid preparation and
may comprise a concentration of the analyte that is detected without dilution
in the linear range of the assay it is intended for. They will often be used as
an external control material in addition to that normally supplied by the assay
manufacturer. Regular monitoring of such material may allow for the early
detection of problems with assay performance.
5.4 Other control material

Control material that does not follow the path of traceability back to the highest-
order reference material (that is, to the IS) may be produced by commercial and
in-house laboratories where no higher-order standard is available for the analyte.
The material may be used as a run control, whereby the unit of
measurement (for example, signal-to-cut-off ratio (S/CO) values, threshold
cycle values for real-time nucleic acid amplification technique (NAT)-based
assays, copies/mL or genome equivalents/mL) can be used for intra-laboratory
monitoring and may provide valuable trending data in a similar way to tertiary
standards. However, such material has not been designed to allow for the
comparison of results between different assays.
6. Commutability of biological standards

Commutability is a property of a reference material such that values measured
for that reference material and for representative clinical samples have the
same relationship between two or more measurement procedures for the same
measurand (10) – that is, reference materials should behave in the same way
as the native analyte itself. Producing biological reference materials that are
commutable can be challenging because the matrix of the analyte may vary in
different clinical conditions, or the analyte may be modified during preparation
or processing of the reference material. Commutability can therefore only
be demonstrated for particular combinations of assays, with particular
clinical samples. It is not a generic property of the reference materials. Thus,
400
Annex 6
demonstrating commutability for two assay methods does not guarantee that
there will be commutability with other methods. Similarly, if a set of samples
demonstrate commutability with each other for particular assay methods then
this does not guarantee that this will apply to all samples (22, 23).
However, ISO17511:2003 states that calibrators are to be commutable
at each step in the traceability chain (11). There are established recommendations
for the assessment of commutability of reference materials used in laboratory
medicine (7).
7. Selection and characterization of materials

for the preparation of secondary standards –
and application to tertiary standards
The basic parameters to be considered when selecting and characterizing a
material for the preparation of a secondary standard are described in this
section. In general, these considerations also apply to the process of developing
tertiary reference materials to be used daily as control materials. In both
scenarios, the selection of materials is dependent upon the type of technology
to which they are to be applied, for example, NAT-based or immunological test
systems. If necessary, expert scientific advice should be sought to support the
development of any secondary standard.
The following should be considered:
■■ analyte – type, source and specificity
■■ immunological and genetic diversity
■■ type of matrix
■■ target concentration
■■ volume of final aliquots and storage temperature
■■ diagnostic specificity
■■ infectivity/inactivation
■■ physical appearance
■■ homogeneity
■■ commutability of the material
■■ stability – real-time and accelerated degradation.
7.1 Analyte – biological versus synthetic material

Consideration must be given to the most suitable form of the analyte to be
established as a secondary standard – for example, whole organism, purified
nucleic acid, recombinant protein, or laboratory-derived or clinical isolate.
401
Ideally, the material selected should resemble the analyte in the IS and in
usual clinical specimens as closely as possible. The decision may be based on
the availability of a sufficient volume of the material to enable preparation of
a single batch of secondary standard that, when frozen, will last several years.
In most cases, laboratory strains of microorganisms are better characterized
and available in larger volumes than clinical strains. However, the latter may
better represent the samples that are routinely tested. Where a whole organism
is unavailable in sufficient quantities, a laboratory-derived material (such as
a purified nucleic acid preparation) may be the only option. Demonstrating
the commutability of such a material to different clinical samples may be
challenging. This will need to be addressed on a case-by-case basis and should
be done in cases where, through experimental assessment, it is proven that
the use of laboratory-derived material improves agreement between assays.
Analytes derived directly from human origin (such as a clinical sample) or
the matrix of a biological sample (such as plasma or whole blood) should be
tested and confirmed to be negative for the presence of pathogens other than
the analyte of interest. This should be done to exclude potential cross-reactivity
with the specific target analyte. If it is necessary to prepare a bulk material by
pooling from more than one source, each component of the bulk material must
be characterized and where possible all components should be identical – for
example, in molecular detection the sequence of the target regions should be
the same. All samples pooled must be mixed thoroughly and the pool should
be homogeneous. It should be noted that pooling may not be appropriate in
all circumstances. A biological bulk material with a high analyte concentration
could if needed be diluted in a suitable matrix.
7.2 Immunological and genetic diversity

The detectability of an analyte by a particular assay may vary due to the
immunological and genetic variability (serotypes, strains, variants, genotypes

or subtypes and so on) of the organism being tested, resulting in suboptimal
detection of particular variants. Therefore, the candidate material that best
reflects the samples being tested should be chosen for the preparation of the
secondary standard. Consideration should be given to the local geographical
patterns of genetic diversity.
Where a standard is being prepared for nucleic acid detection, a well-
characterized strain should be used for which the full genome sequence (or at
least the sequence covering the most frequently amplified regions) is available.
In principle, the same holds true for antigen standards where well-characterized
antigen variants should be chosen. In both cases consideration should be given
to the diagnostic implications of variant detection.
402
Annex 6
7.3 Matrix
The matrix in which the standard is formulated is crucial to creating a material
that is fit for purpose. For NAT-based assays, the matrix needs to be appropriate
for the assay or range of assays for which it is intended. Multiple sample types may
be considered. It follows that for commutability purposes, a biological matrix
would be preferred over a synthetic matrix. It should be taken into consideration
that some sample matrices include inhibitory factors that interfere with the
performance of specific types of assays (7). Furthermore, the chosen matrix of
the reference preparation should be compatible with further matrices into which
it may be spiked. For example, where a pathogen may be screened for in plasma,
whole blood or urine, it may not be appropriate to formulate a material in a
matrix that cannot be further diluted in clinical samples. In the case of plasma,
consideration should also be given to any anticoagulant treatment and to any
additional treatments such as cryoprecipitation or recalcification.
7.4 Concentration
Secondary standards will often be used in a quantitative capacity. Therefore, the
concentration used should be high enough to permit the preparation of dilutions
across the dynamic range of the assay and to allow for dilution into further
matrices. It must be noted that dilutions performed on a high-titre secondary
standard will contribute to the overall uncertainty arising from the dilution
process. This will be the case for most secondary standards.
The target concentration of a standard will be dependent upon its final
intended use and whether any clinical decision points exist when testing for the
analyte. The detectable/quantitative range of all well-established assays for that
pathogen must be taken into consideration. Reference materials that perform
within the dynamic range of an assay (where changes in signal correspond to
changes in analyte concentration) will typically be the norm.
In the case of tertiary standards where the material may be used as an
external run control in qualitative tests, the concentration should ideally be at the
lower end of the range of detection, at a concentration which will appropriately
challenge the assays (for example, three times the 95% limit of detection of a
NAT-based assay, or within the dynamic range of serological assays).
7.5 Volume
The aliquot volume may vary depending on the typical assay input volume for
that analyte and the final storage temperature. The suitability of the container
used for the filling of the aliquots should be validated in terms of the integrity
and stability of the analyte. Where the standard is intended for single use,
as defined in the Instructions for Use provided with the material, sufficient
403
volume should be provided for use in the assays for which it is intended and
any remaining material must be discarded by the laboratory as no in-use
stability testing will have been performed. If the standard is for multiple uses,
the volume will depend on the required number of tests, and on stability at the
recommended storage temperature. Where the manufacturer intends a material
to be used on multiple occasions, suitable stability testing should be carried out
to indicate the number of times the material can be removed from its storage
temperature, allowed to warm to ambient temperature and then be re-cooled
before the stability of the material is compromised.
7.6 Diagnostic specificity

Samples derived directly from human origin (for example, clinical samples or
where the matrix is a biological component such as plasma or whole blood)
should be tested and confirmed to be negative for the presence of pathogens
other than the analyte of interest in order to minimize risk and the potential for
cross-reactivity with the target analyte.
Where human material is used as the diluent, the diluent should be
tested and found to be negative for both the analyte and common high-risk
pathogens.
7.7 Infectivity
Use of infectious materials will impact on processing, handling and shipping of
the standard. Where samples are prepared from infectious material, it is important
to provide clear information to the end users about the exact nature of the
material and origin of the pathogens. Import regulations differ from country to
country and can vary depending on the origin of the pathogen. For example,
tissue-culture-derived viral specimens may be subjected to different shipping
regulations by some countries than a patient sample infected with this virus.
Where standards are prepared from inactivated materials it is important

to confirm the success of the inactivation procedure and to determine potential
effects of the inactivation on the performance of the final standard. The use
of established or proven inactivation methods (such as published methods) is
preferred. In addition, in cases where standards are prepared from inactivated
material but are diluted in a biological matrix (for example, human plasma) the
matrix should be screened by NAT-based and serological assays for the most
common bloodborne viruses.
7.8 Physical appearance

A number of factors will determine the most appropriate physical appearance
for the standard. Typically, IS are lyophilized preparations which have a better
stability and longer shelf-life than, for example, liquid preparations. They are
404
Annex 6
also required to be shipped around the world, preferably at ambient temperature.

This may also be the case for some secondary standards, especially national or
regional secondary standards, such as the European Pharmacopoeia Biological
Reference Preparations for NAT-based assays. A feasibility study should be
conducted to demonstrate that the freeze-drying procedure does not have an
adverse effect on the integrity of the target analyte. However, lyophilization is
costly and specialized, and, for some pathogens, maintaining a stable product is
problematic. In these cases, a liquid preparation stored and shipped at a suitable
temperature may be more appropriate. Lyophilization also requires additional
validation work to determine the potential impact of such a technique on the
biological activity of the standard and on the commutability of the standard
with clinical samples. In the case of nucleic acid extraction, lyophilization
can lead to the formation of aggregates which reduces extraction efficiency.
Lyophilized preparations should be evaluated against the liquid bulk preparation
in different assays as part of the commutability assessment. Where standards are
used frequently, such as tertiary control materials, a liquid or single-use frozen
preparation is probably more suitable so that the standard is ready for use
without the need for reconstitution.
7.9 Homogeneity
It is important to confirm the consistency of the filling procedure and to confirm
that the bulk was dispersed (for example, stirred) sufficiently throughout.
Homogeneity is assessed in two ways – by determining both the biological and
the physical content (weight or volume) of multiple vials across the batch. The
latter is particularly important prior to lyophilization, and can be addressed by
weighing a proportion of vials before and after filling and then calculating the
filled weight and associated coefficient of variation as a percentage (% CV). The
% CV will be higher for a more viscous matrix. It is also important to determine
the biological homogeneity by assessing the concentration of the analyte in
multiple vials across the batch. It is known that homogeneity may be impaired
by genetic quasi-species heterogeneity or antibody complexation. The number
of vials used for testing will depend on the batch size. As a minimum, typically
1–2% of the vials should be tested (17).
7.10 Stability
A stability testing programme should be implemented to monitor the potency of
the secondary standard over time. Stability monitoring can be based on real-time
data. However, additional data from accelerated thermal degradation studies
are helpful in characterizing the robustness of lyophilized reference materials.
Such data are also important for assessing the suitability of the material after
extreme shipping conditions.
405
Factors affecting stability will be dependent upon the physical

appearance and final storage temperature of the material. For example, for final
long-term storage at 2–8 °C, preservatives might need to be included in the
final formulation to prevent fungal or bacterial growth. Likewise, lyophilized
preparations may require different additives, such as inert sugars, which
might aid preservation and protect pathogen viability upon reconstitution. It
should be noted that the effect of any additives on downstream performance or
commutability should be evaluated. Factors affecting the stability of lyophilized
preparations also include residual moisture and oxygen, both of which can
compromise the integrity of the lyophilized product. It is advisable to assess
the levels of residual moisture and oxygen in such preparations following
lyophilization and if possible during storage of the material. This is particularly
important for stoppered vials which may permit the ingress of air during the
lifetime of the product, causing displacement of the vacuum or inert gas which
the material is held under. Loss of potency may occur as a result.
7.11 Stability assessment during product lifetime

7.11.1 Real-time stability
Materials should be periodically removed from their designated storage
temperatures for testing in the routine laboratory assay of choice. The required
frequency of testing will be dependent upon the physical appearance and final
storage temperature of the material. For example, a frozen liquid product stored
at −20 °C may require more frequent testing than a lyophilized product stored at
−20 °C. Real-time monitoring may be more frequent following production of
a new material, but by monitoring over time the frequency of testing could be
reduced. For example, a liquid-filled preparation stored at −20 °C could be tested
every 3 months for the first year following production. If the data suggest good
stability then the testing interval could be increased to 6 months. Likewise, a
lyophilized product may be tested every 6 months for the first year following
production, but further testing could be reduced to annually. Any assessment
of stability and associated outcome should be referenced on the Instructions for
Use distributed with the materials, for example where an acceptable number of
freeze–thaw cycles has been determined, this should be referenced.
7.11.2 Accelerated thermal degradation studies

Real-time stability studies may not demonstrate loss in analyte concentration
over the testing period. For lyophilized material, accelerated thermal degradation
studies can be used to predict the long-term stability of a product from its
performance at elevated temperatures. For example, data demonstrating a loss
in titre after 3 months at 37 °C can be used to predict the time it would take
406
Annex 6
for the same amount of degradation to occur at −20 °C or at another chosen

baseline storage temperature of the product. In addition, accelerated thermal
degradation studies cover the validation of the use of the reference material after
extreme shipment conditions.
A chosen number of samples should be stored in temperature-controlled
environments at for example, −20 °C, 4 °C, 20 °C and 37 °C. Studies at 45 °C and
56 °C may also be suitable for some pathogens and matrices but not all. For
example, a lyophilized plasma matrix will generally not reconstitute following
long-term storage at the higher temperatures indicated; however, this may
not fully reflect the degradation of the analyte. Vials should be periodically
removed and tested alongside vials that have been continuously stored at the
recommended storage temperature (reference point).
The Arrhenius equation can be applied to the resulting data in order
to predict the rate of degradation of the material at the recommended storage
temperature (24, 25).
While accelerated thermal degradation studies have considerable
usefulness in the production of secondary reference materials, they may be less
useful for tertiary standards, for which real-time monitoring will provide the
most valuable data set. Note that for tertiary standards used as external quality
control IVDs, accelerated stability testing is usually performed prior to release.
In all cases, stability testing protocols should be designed in conformity with
ISO 23640:2013.
7.12 In-use stability

In-use stability testing of the standard measures the stability of the standard once
it has been thawed, opened or reconstituted (depending on storage conditions
and physical appearance). It is important to establish in-use stability if the
material is intended to be used on multiple occasions and if it is stored under
different conditions during this period. Where multiple freeze–thaw events are
likely, the effect of these should be evaluated.
7.12.1 Lyophilized preparations

Materials should be reconstituted as defined in the Instructions for Use, aliquoted
and stored at an appropriate temperature (for example, −80 °C for materials
containing free RNA as the analyte).
Reconstituted materials should be subjected to freeze–thawing cycles
and tested at predetermined intervals – for example, weekly, monthly, annually
or biannually, and monitored by suitable assays (preferably quantitative assays if
available). Tests should be performed on at least three replicates per time point
to determine any loss in concentration.
407
7.12.2 Frozen preparations

If the total volume in each vial allows for multiple uses, at least three vials
should be freeze–thawed on multiple occasions. At each time point the samples
should be tested at least in duplicate in three replicate tests. If no degradation
is observed then further time points should be added up to the limit of volume
remaining in the vial.
7.12.3 Liquid preparations

Where materials are routinely held at 2–8 °C, exposure to short periods at
ambient temperature may occur when the product is in use. Where this is the
case multiple exposures to ambient temperatures should be assessed. Three
vials should be removed from storage, left at room temperature for up to 1 hour
and tested at least in duplicate in triplicate tests. These vials should then be
returned to 2–8 °C and the process repeated with the same vials at frequent
intervals. Interval frequency can be determined by the accumulation of data
points and may be reduced following an observation of good stability, as
discussed in section 7.11.1 above.
8. Calibration: testing and statistical analysis

8.1 Principles of calibration
Calibration is the process by which a concentration is assigned to a reference
by the direct comparison of measurements with a higher-order reference, and
represents one of the crucial stages of the establishment of a secondary standard.
Each calibration of a candidate secondary standard has to be performed in
parallel with the higher-order reference, in this case the WHO IS. The following
sections describe the minimum requirements for the calibration of secondary
standards intended for either one specific method in one laboratory (single assay
calibration) or for multiple methods (collaborative study calibration). In both

cases, several independent runs with the candidate standard and the IS in parallel
have to be performed (same assay using the same test conditions). For each run,
a new vial of each standard should be used and diluted in the matrix validated for
the respective assay (for example, negative human plasma).
This WHO guidance document reflects the common statistical methods
used for the calibration of reference materials. Any other statistical method
that has been demonstrated to be a reliable approach to the calibration of such
materials may also be applied. Appropriate software for the statistical analysis
should be available for evaluation of the data, and the statistical analysis should
be performed by staff with expertise in this field.
The calibration study data should be analysed using the relevant
statistical models for bioassays, for example, using the methods recommended
408
Annex 6
by the European Pharmacopoeia (26). The design of the study should take into
consideration that:
■■ each analyte, IS and the candidate secondary standard should be
tested with the same number of dilutions and replicates per dilution;
■■ the adjacent dilution steps should be equally spaced.
The statistical validity of the fitted model should be assessed for each
individual assay. For the parallel-line and Probit models – as the most appropriate
and proven statistical methods for this analysis – the linearity and parallelism
of the logarithmic dose–response relationships between the IS and secondary
standard should be evaluated. If the assay response is linear and the response
lines are parallel then the relative potency of the candidate secondary standard
against the IS can be calculated. Using the parallel-line model, validity criteria
for linearity could include the coefficient of determination (r 2) or a test for
nonlinearity (26). Parallelism could be demonstrated, for example, by means of a
test for non-parallelism or an equivalence approach for the difference or ratio of
slopes (that is, the 95% confidence interval for the ratio of slopes must lie entirely
between predefined equivalence margins). In addition, the precision with which
the potency has been estimated should be provided.
Each calibration will have a stated measurement uncertainty. This
estimate can be determined by identifying all sources of variation, calculating the
extent of variation and using established methods to combine the uncertainty.
The measurement uncertainty associated with assigning a value to the standard
is test-system specific (7, 8). It should be noted that an IS, by definition, has
a specified value which has been assigned and expressed in IU per millilitre
(IU/mL). As a consequence of defining the IU as a fraction of the contents of
the container of the current IS, and because the units defined by any previous IS
formally cease to exist, an uncertainty value is not given to the assigned IU (2).
The variability of the vial weight during filling for each IS is quoted in the study
report and in the Instructions for Use accompanying the standard.
8.2 Single assay calibration using qualitative tests

For qualitative NAT-based assays, four independent runs should be performed.
The first run involves testing serial ten-fold dilutions (until negative) of the IS
and the candidate secondary standard in duplicate, and is intended to determine
the end-point dilution of both standards. In the subsequent three assays, two
half-log10 dilutions either side of the predetermined end-point (5 dilutions in
total) of the standards should be tested. Each dilution (runs 2–4) should be
tested at least in duplicate, giving in total 6 replicates per dilution across all
runs. For each assay, data from all runs at each dilution step will be pooled to
409
give a number of positive results out of the total number of tests performed.
The Probit analysis will estimate the concentration at which 63% of the samples
tested were positive (that is, the dilution at which on average one single copy per
sample tested could be expected under the assumption of an underlying Poisson
distribution). The calculated end-point is used to give estimates expressed in
NAT-detectable units/mL after correcting for an equivalent volume of the test
sample. It should be noted that these estimates are not necessarily directly
equivalent to a genuine genome equivalent number or copy number per mL.
The software for the Poisson distribution will calculate the proportion of potency
of the test sample (candidate secondary standard) relative to the potency of the
standard sample (that is, the IS) so long as the dose–response curves fit within
the statistical model.
Using a real-time NAT-based assay, the calibration can be determined
from the cycle threshold (Ct) values by applying a parallel-line analysis,
conditional on the assumption that the slope fulfils the requirement of −1/log2.
The number of dilutions and the number of replicates per dilution should follow
the instructions given below in section 8.3.1.

The IS and the candidate secondary standard should be tested in three
independent assays. Both standards should be diluted, using serial half-log 10 or
two-fold dilutions, in the diluent appropriate for the assay. The dilution ranges
should be within the detection range of the assays used for the study and should
span the end-point titre (intercept with the cut-off of the assays). The analytical
sensitivity of each assay can be calculated by linear interpolation using the two
dilutions of the dilution series having values below and above the assay cut-off.
The (log-transformed) data should be evaluated against the results obtained
for the secondary standard using a parallel-line assay analysis to estimate the
potency (IU/mL) of the secondary standard relative to the IS. A logarithmic
transformation of the assay response may be necessary if the dilution range was
chosen around the sample cut-off rather than the dynamic range of the assay.
8.3 Single assay calibration using quantitative tests

For the calibration of secondary standards tested in a quantitative NAT-based
assay, the candidate material should be tested neat (where possible) and at two
or three further (for example, ten-fold) dilutions within the linear range of the
assay to obtain at least three concentrations giving quantitative values. The same
methodology applies to the IS, with the exception that this material should be
diluted starting from a concentration as close as possible to the estimated potency
of the secondary standard (as indicated by preliminary tests). All standards
410
Annex 6
should be tested in duplicate and at least three independent runs should be

performed. Where possible, multiple assay lots and reagents should be included
in this testing. The calculation of the potency of the secondary standard may be
performed in one of several ways:
■■ The assay output (for example, copies, genome equivalents2 or
IU/mL) should be analysed by the parallel-line method using, if
necessary, log-transformed data to obtain a “relative potency” in IU/
mL of the secondary standard against the IS (where the slope fulfils
the requirement of −1/log2 corresponding to a value of S = −3.322).
The parallel-line method should be the preferred option for the
data analysis.
■■ The difference in estimated potency (using the test software)
between the candidate secondary standard and the IS (log-
transformed data) can be used to determine the potency of the
secondary standard. The difference is then subtracted from the log-
transformed nominal IU/mL of the IS to obtain the potency of the
secondary standard.
■■ A standard curve generated by the instrument software using the
IS as the standard may be used to determine the potency of the
secondary standard.

As with the quantitative NAT-based assays, the candidate material and the IS
should be tested neat and at two or three further (for example, ten-fold) dilutions
within the linear range of the assay, using the dilution matrix appropriate for
the assay. All samples should be tested in duplicate in at least three independent
runs. The calculation of the potency of the secondary standard can be performed
in one of several ways:
■■ The results obtained with the parallel-line analysis (if necessary on
log-transformed data) should be used to give a “relative potency” of
the secondary standard against the IS in IU/mL. The parallel-line
analysis should be the preferred option for the data analysis.
■■ A standard curve generated by the instrument software using the
IS as the standard may be used to determine the unitage of the
secondary standard.
In the case of assays not yet calibrated against the IS.

2
411
8.4 Collaborative study calibration using multiple assays

Secondary standards that are intended to be used in different assays by multiple
end users should be calibrated using a collaborative study approach. The
amount of work and resources required to perform such a study should not
be underestimated. The collaborative study should be organized by/or with
advice from a body with experience in this field, such as a WHO collaborating
centre. If necessary, a scientific adviser from the field should be identified to
support the collaborative study, including the selection of study participants.
Owing to the complexity of the reported data, which typically include data from
many different types of assays, the statistical analysis should be performed by
a statistician. The general principles of planning and executing these types of
collaborative study are described in section 6 of the WHO Recommendations
for the preparation, characterization and establishment of international and other
biological reference standards (2).
Results from all participants should be analysed by statistical methods
described and considered appropriate by the responsible statistician. This analysis
typically requires access to suitable computing facilities and statistical software.
The testing requirements and protocol of each laboratory/test should follow the
protocol described for the single assay calibration depending on the assay type
(qualitative or quantitative). The results of each assay method should be analysed
separately and should provide an estimate of the relative potency and precision of
the candidate secondary standard against the IS.
The variation in results between test methods, and between laboratories,
should be described and assessed as part of the statistical analysis (precision and
consistency of the results). An assessment should be made of any factors causing
significant heterogeneity of the estimated potency, nonlinearity and differences
in slopes. Although there is no generic outlier detection rule from the statistical
point of view, the exclusion of data should be taken into account in subsequent
analysis where striking differences in results within assays, between assays, or

between participants or test methods are observed. All valid potency estimates
for the candidate secondary standard should be combined to produce a final
geometric mean potency/content with 95% confidence limits. The results of the
tests should be displayed graphically, for example, as histograms or scatter plots.
Tables of the quantitative and qualitative raw data should be included in the
report annexes.
8.5 Calculation of uncertainty of measurement

The assignment of an uncertainty value must be applied to secondary reference
materials. The uncertainty of an observed value is a property of the test
system and is not the effect of mistakes introduced through human error. The
412
Annex 6
measurement of uncertainty is complex and, where possible, advice should be

sought from a statistician. This section aims to give an overview of the area
and to highlight that consideration must be given to the assignment of an
uncertainty value to secondary references.
The uncertainty – often referred to as “measurement uncertainty”
(MU) – expresses the 95% confidence limits either side of the observed value
assigned to a product. By estimating the MU of a product the confidence in
the final value assigned is shown. Uncertainty should be calculated using
log‑transformed data. For example, a value of 3 log 10 IU/mL may be assigned
to a standard. Following estimation of uncertainty this value may be displayed
as 3 log 10 IU/mL ± 0.2 – that is, the value of the material could range from
2.8–3.2 log 10 IU/mL.
There are many aspects to uncertainty and well-documented examples
of how to estimate uncertainty (20, 27–29). However, one simple approach to
estimating MU for a secondary standard is to test the material multiple times on
different occasions (but always using the same test system) in parallel with the
WHO IS – that is, test the two standards under exactly the same conditions.
As a guide, a minimum of 30 test results for both the secondary standard
and the IS should be generated from a combination of at least three independent
tests. The more times the sample is tested the better. The test system used should
be of the highest order possible, that is, a commercial assay, or, in the absence of
such, a well-validated laboratory-developed test.
After determining the mean and standard deviation of the data points,
dividing the standard deviation by the square root of the number of samples
tested gives the standard uncertainty (or standard error) of measurement. This
approach demonstrates the imprecision involved and does not account for MU
derived from inherent bias.
9. Establishment of reference materials in the

absence of a WHO International Standard
In the absence of a WHO IS, an established regional or national reference
standard may become the standard of comparison for the candidate assay.
Current examples include: West Nile virus NAT reference reagent (Health
Canada), and Chikungunya virus standard (Center for Biologics Evaluation
and Research, United States Food and Drug Administration). These standards
should be characterized through extensive analysis. The methods for their
characterization, preparation and storage should ideally be published in peer-
reviewed journals.
The calibration of reference standards where no IS and secondary
standards are available should follow the WHO-recommended principles for
413
the preparation of international biological standards (2). Where no IS exists but

a regional standard is available the calibration of a reference standard of lower
order should follow the guidance provided in this document.
In the absence of an IS or a national reference standard, a candidate
reference material may be assigned a value using either a commercial assay or a
laboratory-developed test. Alternatively, the material may be calibrated against
other commercially available preparations or samples from EQA programmes
with assigned values, or a manufacturer’s working calibrator. For such reference
materials, reserving a separate proportion of vials (baseline samples) to be used
in the calibration of subsequent batches is strongly advised. This batch should be
stored at the lowest possible temperature validated for the vials, preferably at a
lower temperature than that at which the bulk of the reference material is stored
(for example, if a reference material is routinely stored at −70 °C, the baseline
samples should be stored under nitrogen vapour or liquid). Reference materials
may also be assigned a value based on a range of assays (collaborative study) as
described above rather than a single assay. In such cases, the assigned value will
be the mean of the results reported by all the assays. Whichever approach is used,
the method and assigned value need to be documented (20).
10. Post-production considerations

10.1 Storage of the material
Following the development and production of a batch of secondary reference
materials, the material should be stored at an appropriate temperature which
ensures stability throughout the lifetime of the product. The temperature of
the cold-storage unit should be monitored and recorded. A protocol should be
developed for the real-time monitoring of each product and should include
details of testing frequency, number of replicates, methodology used and
statistical analysis (see section 7.11 above).
10.2 Distribution of the material

Consideration should also be given to the method by which the material is to
be distributed. Where material is potentially infectious, specialist couriers may
be required. Some countries have import-permit requirements for infectious
materials. These requirements are country-specific and should be discussed with
the recipient in advance of the shipment.
Stability of the product during transportation can be addressed by
distributing the material at appropriate temperatures, employing where necessary
the use of dry ice or cold packs. The use of dry ice may be considered a hazard by
some couriers and may require the use of specialized companies.
414
Annex 6
10.3 Instructions for Use

Detailed Instructions for Use should be supplied with every shipment. These
should include the following information:
■■ characterization details of the analyte
■■ storage conditions
■■ procedures prior to use (for example, reconstitution)
■■ appropriate use
■■ stability information
■■ safety information
■■ references to any publications relating to the material (for example,
study report).
10.4 Replacement batches

The need for replacement batches should be addressed in the development
stages of the initial product. Consideration should be given to the acquisition
of material for future replacement of the standard. For example, if the analyte
comprises tissue-culture-derived material, large batches of stock material could
be cultured in the first instance. Detailed records of the production of the first
batch should also be documented to allow replication of the production method
at some point after the initial material was produced.
11. End user advice

Manufacturers are encouraged to include details of the production and
calibration process in the Instructions for Use provided with each material, or
to provide a reference to where this information can be found. Any additional
information not supplied could be requested from the manufacturer, and
could include:
■■ number of replicates and methods used to assess repeatability and
reproducibility;
■■ the metrological traceability of an assigned unit;
■■ whether a collaborative study was performed, and if so the number
of participants and range of assays evaluated;
■■ assessment of performance in a different matrix;
■■ stability assessments including of shelf-life and in-use stability;
■■ assessment of specificity;
■■ validation of limit of detection or cut-off.
415

This WHO guidance document was jointly drafted by Dr M. Chudy, Paul-
Ehrlich-Institut, Germany; Dr C. Morris, National Institute for Biological Standards
and Control, England; Dr J. Fryer National Institute for Biological Standards and
Control, England; Dr W. Dimech, National Reference Laboratory, Australia;
and Dr J. Saldanha, Immucor, Inc., the USA.
The first draft was then discussed at a Standardisation of Genome
Amplification Techniques (SoGAT) workshop in 2015 and presented to the
WHO Expert Committee on Biological Standardization in the same year. Further
comments on the draft were received from IVD manufacturer associations,
individual IVD manufacturers, regulatory bodies and experts in the field. A
revised draft was then discussed at a 2016 SoGAT workshop. Acknowledgement
is made to all delegates of the SoGAT 2015 and 2016 workshops for their critical
reviewing of the draft versions and other inputs.
The second draft version was published on the WHO Biologicals
website for a round of public consultation between 16 June and 16 September
2016, and the comments received were incorporated to produce the document
WHO/BS/2016.2284.
References
1. Guidelines for the preparation and establishment of reference materials and reference reagents
for biological substances. In: WHO Expert Committee on Biological Standardization: twenty-
ninth report. Geneva: World Health Organization; 1978: Annex 4 (WHO Technical Report
Series, No. 626; http://apps.who.int/iris/bitstream/10665/41323/1/WHO_TRS_626.pdf, accessed
5 February 2017).
2. Recommendations for the preparation, characterization and establishment of international

and other biological reference standards (revised 2004). In: WHO Expert Committee on
Biological Standardization: fifty-fifth report. Geneva: World Health Organization; 2006: Annex 2
(WHO Technical Report Series, No. 932; http://www.who.int/immunization_standards/vaccine_
reference_preparations/TRS932Annex%202_Inter%20_biol%20ef%20standards%20rev2004.
pdf?ua=1, accessed 5 February 2017).
3. WHO manual for the establishment of national and other secondary standards for vaccines.
Geneva: World Health Organization; 2011 (WHO/IVB/11.03; http://apps.who.int/iris/bitstream/
10665/70669/1/WHO_IVB_11.03_eng.pdf, accessed 5 February 2017).
4. General guidelines for the establishment, maintenance and distribution of chemical reference
substances. In: WHO Expert Committee on Specifications for Pharmaceutical Preparations: forty-
first report. Geneva: World Health Organization; 2007: Annex 3 (WHO Technical Report Series,
No. 943; http://apps.who.int/iris/bitstream/10665/43632/1/WHO_TRS_943_eng.pdf, accessed
5 February 2017).
416
Annex 6
5. International vocabulary of metrology – Basic and general concepts and associated terms
(VIM) third edition. JCGM 200:2012 (http://www.bipm.org/utils/common/documents/jcgm/
JCGM_200_2012.pdf, accessed 5 February 2017).
6. Guidance for industry. Bioanalytical method validation. Rockville (MD): Center for Veterinary
Medicine; 2013 (http://www.fda.gov/downloads/Drugs/Guidance/ucm070107.pdf, accessed
7 February 2017).
7. CLSI. Characterization and qualification of commutable reference materials for laboratory
medicine; approved guideline. CLSI document EP30-A. Wayne (PA): Clinical and Laboratory
Standards Institute; 2010 (sample: http://shop.clsi.org/site/Sample_pdf/EP30A_sample.pdf,
accessed 7 February 2017).
8. ISO 15198:2004. Clinical laboratory medicine – in vitro diagnostic medical devices – validation
of user quality control procedures by the manufacturer. Geneva: ISO; 2004 (abstract: http://www.
iso.org/iso/catalogue_detail.htm?csnumber=39751, accessed 7 February 2017).
9. 2009/886/EC: Commission Decision of 27 November 2009 amending Decision 2002/364/EC
on common technical specifications for in vitro diagnostic medical devices (notified under
document C (2009) 9464) (Text with EEA relevance). European Commission. Official Journal of the
European Union. 2009;L318/25–L318/40 (http://eur-lex.europa.eu/eli/dec/2009/886/oj, accessed
7 February 2017).
10. Frey A, Di Canzio J, Zurakowski D. A statistically defined endpoint titer determination method
for immunoassays. J Immunol Methods. 1998;221(1–2):35–41 (https://www.researchgate.net/
publication/13394025_A_statistically_defined_endpoint_titer_determination_method_for_
Immunoassays, accessed 7 February 2017).
11. ISO 17511:2003. In vitro diagnostic medical devices – measurement of quantities in biological
samples – metrological traceability of values assigned to calibrators and control materials.
Geneva: ISO; 2003 (abstract: http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.
htm?csnumber=30716, accessed 7 February 2017).
12. ISO 18113-1:2009. In vitro diagnostic medical devices – information supplied by the manufacturer
(labelling) – Part 1: Terms, definitions and general requirements. Geneva: ISO; 2009 (https://www.
iso.org/obp/ui/#iso:std:iso:18113:-1:ed-1:v1:en, accessed 7 February 2017).
13. United States Pharmacopeia 38 National Formulary 33 (USP 38-NF 33). United States
Pharmacopeial Convention; 2015.
14. Linden JV, Bianco C, editors. Blood safety and surveillance. New York/Basel: Marcel Dekker, Inc.;
2001.
15. Jung R, Soondrum K, Neumaier M. Quantitative PCR. Clin Chem Lab Med. 2000;38(9):833–6
(abstract: https://www.ncbi.nlm.nih.gov/pubmed/11097336, accessed 7 February 2017).
16. ISO/IEC Guide 98-3:2008. Uncertainty of measurement – Part 3: Guide to the expression of
uncertainty in measurement (GUM:1995). Geneva: ISO; 2009 (abstract: http://www.iso.org/iso/
catalogue_detail.htm?csnumber=50461, accessed 7 February 2017).
17. Resolution WHA63.12. Availability, safety and quality of blood products. Sixty-third World
Health Assembly, Geneva, 17–21 May 2010. Geneva: World Health Organization; 2010 (http://
apps.who.int/gb/ebwha/pdf_files/WHA63-REC1/WHA63_REC1-en.pdf, accessed 7 February 2017).
18. Analytical methods and reference materials program. AMRM Glossary of Terms [website]. NIH, US
Department of Health & Human Services (https://ods.od.nih.gov/Research/AMRMGlossary.aspx,
417
19. ISO 15194:2009. In vitro diagnostic medical devices – measurement of quantities in samples of
biological origin – requirements for certified reference materials and the content of supporting
documentation. Geneva: ISO; 2009 (https://www.iso.org/obp/ui/#iso:std:iso:15194:ed-2:v1:en,
20. Dimech W, Francis B, Kox J, Roberts G. Calculating uncertainty of measurement for serology
assays by use of precision and bias. Clin Chem. 2006;52(3):526–9 (https://www.researchgate.
net/publication/7362486_Calculating_Uncertainty_of_Measurement_for_Serology_Assays_by_
Use_of_Precision_and_Bias, accessed 7 February 2017).
21. Principles and methods of validation of diagnostic assays for infectious diseases. In: Manual of
diagnostic tests and vaccines for terrestrial animals [web version]. Paris: World Organisation for
Animal Health; 2016. Chapter 1.1.6, adopted in May 2013 (http://www.oie.int/fileadmin/Home/
eng/Health_standards/tahm/1.01.06_VALIDATION.pdf), accessed 7 February 2017).
22. Miller WG, Myers GL. Commutability still matters. Clin Chem. 2013;59(9):1291–3 (http://clinchem.
aaccjnls.org/content/59/9/1291, accessed 7 February 2017).
23. WHO Consultation on commutability of WHO Biological Reference Preparations for in vitro
detection of infectious markers. Meeting Report. Geneva: World Health Organization; 2013
(WHO/BS/2013.2230 Addendum 1; http://www.who.int/entity/bloodproducts/norms/BS_2230_
Addendum1_Commutability.pdf?ua=1, accessed 27 March 2017).
24. Kirkwood TB. Predicting the stability of biological standards and products. Biometrics. 1977;
33(4):736–42 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/588659, accessed 7 February
2017).
25. Egan W, Schofield T. Basic principles of stability. Biologicals. 2009;37(6):379–86; and discussion
421–3 (abstract: http://www.sciencedirect.com/science/article/pii/S1045105609001158, accessed
7 February 2017).
26. Statistical analysis of results of biological assays and tests. 01/2008:50300. European
Pharmacopoeia 6.0, section 5.3 (http://www.uspbpep.com/ep60/5.3.%20%20statistical%20
analysis%20of%20results%20of%20biological%20assays%20and%20tests%2050300e.pdf,
27. Konnert A, Berding C, Arends S. Uncertainty calculation for calibrators and controls of laboratory
diagnostic assays. Clin Chem Lab Med. 2006;44(10):1175–82 (abstract: https://www.researchgate.
net/publication/6763360_Uncertainty_calculation_for_calibrators_and_controls_of_laboratory_
diagnostic_assays, accessed 7 February 2017).
28. Bell S. A beginner’s guide to uncertainty of measurement. Measurement good practice guide
No. 11 (Issue 2 with amendments 2001). Teddington: National Physics Laboratory; 1999 (abstract:
http://www.npl.co.uk/publications/a-beginners-guide-to-uncertainty-in-measurement, accessed
7 February 2017).
29. Tang L, Sun Y, Buelow D, Gu Z, Caliendo AM, Pounds S et al. Quantitative assessment of
commutability for clinical viral load testing using a digital PCR-based reference standard.
J Clin Microbiol. 2016;54(6):1616–23 (http://jcm.asm.org/content/54/6/1616.long, accessed
7 February 2017).
418
Annex 6
Appendix 1
Example of the parallel calibration of a secondary standard
in a study to establish the International Standard
WHO/BS/2016.2291
Page 1
WHO/BS/2016.xxxx
ENGLISH ONLY
EXPERT COMMITTEE ON BIOLOGICAL STANDARDIZATION

Geneva, 17 to 21 October 2016
Collaborative Study to Evaluate the Proposed 4th WHO International

Standard for Hepatitis B Virus (HBV) DNA for Nucleic Acid Amplification
Techniques (NAT)
Jacqueline F. Fryer 1,3, Rehan Minhas 1, Thomas Dougall 2, Peter Rigsby 2, Clare L. Morris 1 and
the Collaborative Study Group *
1
Division of Virology and 2 Biostatistics, National Institute for Biological Standards and
Control, South Mimms, Potters Bar, Hertfordshire, EN6 3QG, UK; 3 Study coordinator, E-mail
Jacqueline.Fryer@nibsc.org; * Details provided in Appendix 1
NOTE:
This document has been prepared for the purpose of inviting comments and suggestions on the
proposals contained therein, which will then be considered by the Expert Committee on
Biological Standardization (ECBS). Comments MUST be received by 16 September 2016 and
should be addressed to the World Health Organization, 1211 Geneva 27, Switzerland, attention:
Technologies, Standards and Norms (TSN). Comments may also be submitted electronically to
the Responsible Officer: Dr C M Nübling at email: nueblingc@who.int.
© World Health Organization 2016
All rights reserved. Publications of the World Health Organization are available on the WHO web site (www.who.int) or can be
purchased from WHO Press, World Health Organization, 20 Avenue Appia, 1211 Geneva 27, Switzerland (tel.: +41 22 791
3264; fax: +41 22 791 4857; e-mail: bookorders@who.int).
Requests for permission to reproduce or translate WHO publications – whether for sale or for noncommercial distribution –
should be addressed to WHO Press through the WHO web site:
(http://www.who.int/about/licensing/copyright_form/en/index.html).
The designations employed and the presentation of the material in this publication do not imply the expression of any opinion
whatsoever on the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its
authorities, or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines
for which there may not yet be full agreement.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended
by the World Health Organization in preference to others of a similar nature that are not mentioned. Errors and omissions
excepted, the names of proprietary products are distinguished by initial capital letters.
All reasonable precautions have been taken by the World Health Organization to verify the information contained in this
publication. However, the published material is being distributed without warranty of any kind, either expressed or implied. The
responsibility for the interpretation and use of the material lies with the reader. In no event shall the World Health Organization 419
be liable for damages arising from its use. The named authors alone are responsible for the views expressed in this publication.
WHO/BS/2016.2291
Page 2
Summary
This report describes the collaborative study evaluation of the replacement 4th WHO
International Standard for hepatitis B virus (HBV) DNA for the calibration of secondary
reference preparations for HBV nucleic acid amplification techniques (NAT). The candidate 4th
HBV International Standard was prepared and evaluated as part of the collaborative study to
establish the 3rd HBV WHO International Standard in 2011. The lyophilized preparation
comprises a dilution of the same HBV DNA-positive plasma sample as used for the previous
HBV International Standards, in pooled human plasma. In this collaborative study, thirteen
laboratories from seven countries evaluated the suitability of the candidate using their routine
HBV NAT-based assay. The candidate NIBSC code 10/266 (sample 2) was evaluated alongside
the 3rd HBV WHO International Standard, NIBSC code 10/264 (sample 1), three HBV reference
preparations (samples 3-5) and three HBV-positive plasma samples comprising different HBV
genotypes (samples 6-8). A range of HBV NAT assays were used in the evaluation, the majority
of which were commercial quantitative assays based on real-time PCR technology.
The overall mean potency estimates for samples 1 and 2, were 5.94 and 5.97 log10 IU/mL
respectively. These values are very similar to the values obtained for these samples relative to
the pre-existing 2nd HBV WHO International Standard in the 2011 collaborative study (5.93 and
5.98 log10 IU/mL respectively). The standard deviation in individual laboratory mean estimates
for samples 1 and 2 was 0.13 log10 IU/mL. The overall mean potency estimate for sample 2,
relative to sample 1 was 5.96 log10 IU/mL. The results obtained from ongoing accelerated
thermal degradation studies indicate that the candidate sample 2 has remained stable over the 5
years post-manufacture.
The results of the study indicate the suitability of sample 2 as the replacement 4th HBV WHO
International Standard. Since the overall mean potency obtained for the candidate in this
collaborative study is very similar to the overall mean potency obtained in the 2011 collaborative
study, relative to the pre-existing 2nd HBV WHO International Standard, it is proposed that the
value assigned to the candidate sample 2 is that obtained in the 2011 collaborative study. This
approach would minimize any potential drift in the value of the IU during the replacement. It is
therefore proposed that candidate sample 2 (NIBSC code 10/266) is established as the 4th WHO
International Standard for HBV DNA for NAT with an assigned potency of 955,000 IU/mL
(~5.98 log10 IU/mL) when reconstituted in 0.5 mL of nuclease-free water.
Introduction
Hepatitis B virus (HBV) remains a major public health problem worldwide, despite the
availability of an effective vaccine and antiviral therapies. More than 240 million people
worldwide are chronically infected, with 0.5-1 million dying annually as a result of serious liver
disease 1. The virus is transmitted in blood and body fluids, perinatally and through close person-
to-person contact in early childhood (in regions with high HBV prevalence), and through
infected needles and sexual contact (in regions with low HBV prevalence) 1. Nucleic acid
amplification techniques (NAT) for HBV were first introduced for blood screening in 1997, and
are now implemented in at least 30 countries worldwide 2,3. However, there remains a residual
risk of transfusion-transmitted infection, through occult HBV infection and vaccine breakthrough
infections 4. NAT is routinely used in the diagnosis and management of HBV infections,
particularly, to guide the initiation of and monitor the response to antiviral therapy in
chronically-infected patients 5. A range of both commercial and laboratory-developed NAT-
based assays are currently in use. The WHO International Standard for HBV DNA was
established in 1999 6,7, and is used by manufacturers of in vitro diagnostic devices (IVDs), blood
transfusion centres, control authorities, and clinical laboratories, to calibrate secondary reference
materials for NAT in terms of the International Unit (IU).
420
Annex 6
WHO/BS/2016.2291
Page 3
The 1st and 2nd WHO International Standards for HBV were prepared by dilution of a Eurohep
R1 sample 8 (Genotype A2, HBsAg subtype adw2, derived from a single donor) in HBV-
negative pooled human plasma. Both materials were prepared from the same bulk (filled and
freeze-dried on two separate occasions) and evaluated in parallel in a worldwide collaborative
study using a range of NAT-based assays for HBV 6,7. The first candidate (NIBSC code 97/746)
was established as the 1st WHO International Standard for HBV DNA in 1999, with an assigned
potency of 1,000,000 IU/mL when reconstituted in 0.5 mL nuclease-free water. In 2006, the
WHO Expert Committee on Biological Standardization (ECBS) established the second candidate
(NIBSC code 97/750) as the replacement 2nd WHO International Standard for HBV DNA
following a smaller collaborative study 9,10. In 2011, two replacement batches (NIBSC codes
10/264 and 10/266) were prepared from the same original HBV Eurohep R1 stock as the 1st and
2nd WHO International Standards, diluted in pooled human plasma, and were evaluated in a
worldwide collaborative study in parallel with the 2nd WHO International Standard for HBV
(NIBSC code 97/750) 11. Mean relative potencies for 10/264 and 10/266 were 5.93 and 5.98
log10 IU/mL respectively, when compared against 97/750. The first candidate (10/264) was
established as the 3rd WHO International Standard for HBV DNA in October 2011 with a unitage
of 850,000 IU/mL. It was noted that the second candidate (10/266) would be a suitable
replacement HBV International Standard in due course, depending on ongoing stability
assessment.
The established use of the HBV IU as the unit of measurement for HBV DNA highlights the
importance of maintaining the availability of this International Standard. This report describes
the collaborative study evaluation of the second candidate 10/266 as the replacement 4th WHO
International Standard for HBV for NAT. The proposal to replace the 3rd WHO International
Standard for HBV DNA was endorsed by the WHO ECBS in October 2015. The collaborative
study results were presented to the Scientific Working Group on the Standardization of Genome
Amplification Techniques (SoGAT) in London in June 2016. The proposed standard is intended
to be used in the in vitro diagnostics field and relates to ISO 17511:2003 Section 5.5 12.
Aims of study
The aim of this collaborative study was to evaluate the suitability of the candidate lyophilized
preparation in parallel with the 3rd HBV WHO International Standard (NIBSC code 10/264)
using a range of NAT-based assays.
Materials
Candidate standard
The candidate preparation (NIBSC code 10/266) comprises lyophilized human plasma and HBV.
The HBV was sourced from a stock of the Eurohep R1 reference material stored at NIBSC and is
a genotype A2, HBsAg subtype adw2 virus from a single donor 8. The pooled human plasma
diluent was sourced from UK blood donations and had been tested and found negative for HIV
antibody, HCV antibody, HBsAg and syphilis. It was also tested at NIBSC and found negative
for HCV RNA by NAT. The preparation was lyophilized to ensure long-term stability.
The filling and lyophilization of the bulk material was performed under contract at eQAD, UK
NEQAS (Colindale, UK), in March 2011 and has been described previously 11. The bulk was
dispensed in 0.5 mL volumes into 3 mL screw-cap glass vials (Adelphi Tubes, Haywards Heath,
UK). The homogeneity of the fill was determined by performing check-weighing of
approximately every fiftieth vial, with vials outside the defined specification being discarded.
Filled vials were partially stoppered, lyophilized and then fully stoppered in the freeze dryer. A
total of 2700 vials were prepared for 10/266. The percentage coefficient of variation (%CV) of
421
WHO/BS/2016.2291
Page 4
the fill weight was 0.36%. The sealed vials were returned to NIBSC for storage at -20 °C under
continuous temperature monitoring for the lifetime of the product. Evaluation of multiple
aliquots of 10/266 (n=30) at NIBSC indicated that the HBV content was homogeneous (2SD of
0.12 log10 IU/mL). Comparison of the liquid bulk versus the lyophilized product indicated that
there was a minimal loss in potency of approximately 0.04 log10 IU/mL upon freeze-drying.
Assessments of residual moisture and oxygen content, as an indicator of vial integrity after
sealing, were determined for 20 vials of 10/266 as previously described 11, and were 0.29% (Karl
Fischer, 0.45% NIR units, CV=14.66%) and 0.17% (CV=63.81%) respectively.
Stability of the lyophilized candidate

Ongoing accelerated thermal degradation studies have been underway at NIBSC since 2011 in
order to predict the stability of 10/266 when stored at the recommended temperature of -20 °C.
Vials of lyophilized product have been held at -70 °C, -20 °C, +4 °C, +20 °C, +37 °C and
+45 °C. At 11 weeks, 12, 34 and 60 months post-manufacture, three vials have been removed
from storage at each temperature and HBV DNA quantified by NAT using the COBAS®
AmpliPrep/COBAS® TaqMan® HBV Test, version 2.0 (Roche Diagnostics GmbH, Mannheim,
Germany).
Study samples
The lyophilized candidate 10/266 was evaluated alongside the 3rd HBV WHO International
Standard (NIBSC code 10/264), a HBV Secondary Reference Reagent (SRR), a HBV National
Standard (NS) and Working Reagent (WR), and three individual HBV-positive plasma donations
comprising different genotypes. These plasma donations were sourced from HBV-positive
plasma packs rejected by the UK NHS Blood and Tranplant authority (Colindale, UK). The
HBV genotype was determined using a multiplex PCR assay with genotype-specific primers 13.
Lyophilized and liquid frozen study samples were stored at -20 °C and -70 °C, respectively, prior
to shipping to participants by courier on dry ice.
Study samples were coded as samples 1-8 and were as follows:

– Sample 1- Lyophilized 10/264 3rd HBV WHO International Standard in a 3 mL crimp-cap
glass vial.
– Sample 2 - Lyophilized candidate 10/266 in a 3 mL screw-cap glass vial.
– Sample 3 - Liquid frozen HBV Secondary Rreference Reagent (genotype A) in a 2 mL Sarstedt
tube
– Sample 4 - Liquid frozen HBV National Standard (genotype C) in a 2 mL Sarstedt tube.
– Sample 5 - Liquid frozen HBV Working Reagent (genotype C) in a 2 mL Sarstedt tube.
– Sample 6 - Liquid frozen HBV plasma (genotype D) in a 2 mL Sarstedt tube.
– Sample 7 - Liquid frozen HBV plasma (genotype E) in a 2 mL Sarstedt tube.
– Sample 8 - Liquid frozen HBV plasma (genotype A) in a 2 mL Sarstedt tube.
Study design
The aim of this collaborative study was to evaluate the suitability of the candidate 4th HBV
WHO International Standard in parallel with the 3rd WHO International Standard for HBV using
a range of NAT-based assays. Three HBV reference reagents were included in the study with the
intention of calibrating these reagents in IU. Three HBV plasma samples were included in the
study in order to provide a limited assessment of commutability 12,14. Three vials of each study
sample were sent to participating laboratories, with specific instructions for storage,
reconstitution and testing. Samples 6-8 were only sent to laboratories performing quantitative
HBV NAT.
422
Annex 6
WHO/BS/2016.2291
Page 5
Study protocol
Participants were requested to test dilutions of each sample using their routine HBV NAT assay
on three separate occasions, using a fresh vial of each sample in each independent assay. In
accordance with the study protocol (Appendix 2), the lyophilized samples were to be
reconstituted with 0.5 mL of deionized, nuclease-free molecular-grade water and left for a
minimum of 20 minutes with occasional agitation before use. Samples 6-8 were to be tested neat
and were therefore only evaluated by laboratories performing quantitative assays.
For quantitative assays, participants were requested to test samples 1-8 neat and to test samples
1-5 at a minimum of two serial ten-fold dilutions (10-1 and 10-2). For qualitative assays,
participants were requested to test ten-fold serial dilutions of samples 1-5, around the assay end-
point (in order to determine the actual assay end-point). For subsequent assays, participants were
asked to test the dilution at the predetermined end-point, and a minimum of two half-log10 serial
dilutions either side of the end-point (i.e., at least five dilutions in total). Participants were
requested to perform dilutions using the sample matrix specific to their individual assay (e.g.
HBV DNA-negative human plasma), and to extract samples prior to HBV DNA measurement.
Participants were requested to report the concentration of each sample in IU/mL

(positive/negative for qualitative assays) for each dilution of each sample and to return results,
including details of the methodology used, to NIBSC for analysis.
Participants
Study samples were sent to 13 participants representing 7 countries (Appendix 1). Participants
were selected for their experience in HBV NAT, geographic distribution and participation in
previous evaluation studies. They represented IVD manufacturers, Official Medicines Control
Laboratories (OMCLs) and WHO collaborating centres. All participating laboratories are
referred to by a code number, allocated at random, and not representing the order of listing in
Appendix 1. Where a laboratory returned data using different assay methods, the results were
analyzed separately, as if from different laboratories, and are referred to as, for example,
laboratory 01A, 01B etc.
Statistical methods
Qualitative and quantitative assay results were evaluated separately. In the case of qualitative
assays (from laboratory 12), data from all assays were pooled to give a number positive out of
number tested at each dilution step. A single ‘end-point’ for each dilution series was calculated,
to give an estimate of ‘NAT detectable units/mL’, as described previously 15. It should be noted
that these estimates are not necessarily directly equivalent to a genuine genome copy
number/mL. In the case of quantitative assays, analysis was based on the results supplied by the
participants. Results were reported as IU/mL. For each assay run, a single estimate of log10
IU/mL was obtained for each sample, by taking the mean of the log10 estimates of IU/mL across
replicates, after correcting for any dilution factor. A single estimate for the laboratory and assay
method was then calculated as the mean of the log10 estimates of IU/mL across assay runs.
All analysis was based on the log10 estimates of IU/mL or ‘NAT detectable units/mL’. Overall
mean estimates were calculated as the means of all individual laboratories. Variation between
laboratories (inter-laboratory) was expressed as standard deviation (SD) of the log10 estimates
and % geometric coefficient of variation (%GCV) 16 of the actual estimates. Potencies relative to
sample 1, the current HBV WHO International Standard (10/264), were calculated as the
difference in estimated log10 ‘units per mL’ (test sample – standard) plus the value in log10
IU/mL for the International Standard. Therefore for example, if in an individual assay, the test
423
WHO/BS/2016.2291
Page 6
sample is 0.5 log10 higher than the International Standard, assigned 5.93 log10 IU/mL, the relative
potency of the test sample is 6.43 log10 IU/mL.
For the quantitative assays, variation within laboratories, and between assays, (intra-laboratory)
was expressed as SDs and %GCVs of the individual assay mean log10 estimates. These estimates
were pooled across all samples. The significance of the inter-laboratory variation relative to the
intra-laboratory variation was assessed by an analysis of variance.
Results and data analysis

Validation of study samples and stability assessment
Production data for the candidate 10/266 from 2011 showed that the CV of the fill weight and
mean residual moisture were within acceptable limits for a WHO International Standard 17. The
residual oxygen content was within the NIBSC working limit of 1.1%.
Samples of 10/266 were stored at elevated temperatures, and assayed at NIBSC in parallel with
samples stored at -20 °C and -70 °C by HBV NAT (as described for the stability of the
lyophilized candidate). Three vials of each sample were evaluated after storage at each
temperature for 11 weeks, 12, 30 and 60 months. It was not possible to reconstitute the vials
stored at +37 °C and +45 °C for 12, 30 and 60 months. The mean estimated log10 IU/mL and
differences (log10 IU/mL) from the -20 °C baseline samples are shown in Table 1. A negative
value indicates a drop in potency relative to the -20 °C baseline. The 95% confidence intervals
for the differences are ±0.08 log10 based on a pooled estimate of the standard deviation between
individual vial test results. As there is no observed drop in potency it is not possible to fit the
usual Arrhenius model for accelerated degradation studies, or obtain any predictions for the
expected loss per year with long-term storage at -20 °C. All available data indicates adequate
stability. Stability testing of 10/266 will be ongoing.
The stability of 10/266 when reconstituted has not been specifically determined. Therefore, it is
recommended that the reconstituted material is for single use only.
Data received
Data were received from all 13 participating laboratories. Participants performed a variety of
different assay methods, with some laboratories performing more than one assay method. In
total, 15 datasets were received from 14 quantitative assays and 1 qualitative assay. Apart from
the cases noted below, there were no exclusions of data.
Quantitative Assays:
Data were returned with dilutions ranging from neat to 10-8 from different laboratories, although
only dilutions between neat and 10-2 were used in the analysis. For Sample 5, one or two
dilutions were removed from the following laboratories’ data, 01, 02B, 03, 05, 06, 07, 08, 10 and
13 because the results were below the limit of detection. Samples not demonstrating dilutional
linearity (i.e. a linear relationship between reported HBV content against log10 dilution with
fitted slope between 0.80 and 1.25) were excluded. Non-linearity was assessed visually and
determined in the following cases; laboratory 11A, Sample 5 on day 2, Samples 3 and 5 on day
3; laboratory 11B, Sample 5 on day 3; laboratory 13, Sample 3 on day 2. Laboratories 04, 05, 07,
09, 11B and 13 had one to three samples with slopes outside the range 0.80-1.25, with the
majority of these cases (7 out of 10) for Sample 5.
Qualitative Assays:
424
Annex 6
WHO/BS/2016.2291
Page 7
Laboratory 12 tested multiple dilutions, from 10-1 down to 10-8 for different samples. Only
dilutions that were tested in at least two of the three assays were used in the analysis.
Summary of assay methodologies

The majority of participants prepared dilutions of study samples 1-5 using negative human
plasma, however, Basematrix 53 (SeraCare Life Sciences, Inc., Milford MA, USA) was also
used. Assay methodologies for qualitative and quantitative assays are summarized in Table 2.
Ten different commercial HBV NAT assays were represented. Five new assays were included
that were not represented in the 2011 collaborative study (assay codes; c68, APT, VER, SKB
and SaB). Only one of the commercial assays was qualitative (assay code cTSM). No participant
used laboratory-developed tests.
Estimated IU/mL or ‘NAT detectable units/mL’

The laboratory mean estimates of IU/mL (log10) from the quantitative assays and ‘NAT
detectable units/mL’ (log10) from the qualitative assay (shaded in grey) are shown in Table 3.
The individual laboratory mean estimates are also shown in histogram form in Figure 1. Each
box represents the mean estimate from one laboratory, and the boxes are labeled with the
laboratory and assay code. They are also colour coded by assay. For samples 1, 4 and 5, there is
good agreement between qualitative and quantitative assays, however, for samples 2 and 3 there
is not. This may reflect variability in the dilution of study samples required by the study protocol
for qualitative assays rather than actual variability in the detection of HBV in the different study
samples. There was good agreement between the estimates from the quantitative assays,
particularly for samples 1-3 which all comprise the same genotype A virus. Laboratory 6 appears
to underquantify samples 4-8 compared to other quantitative assays. There also appears to be
variability in the quantification of samples 4-8 by the different assay methods, with some assays
showing either under or over-quantification compared to the overall mean estimate for each of
the samples. This may represent variability in the quantification of different HBV genotypes as
has been reported previously 18. However, for sample 8, the pattern of individual laboratory
mean estimates is different to that for samples 1-3 despite all comprising genotype A viruses.
Table 4 shows the overall mean estimates of log10 IU/mL from the quantitative assays, along
with the SD (of log10 estimates) and the %GCV (of actual estimates). The overall mean estimates
for samples 1 and 2 were 5.94 and 5.97 Log10 IU/mL respectively. These values are very similar
to the values obtained for these samples in the 2011 collaborative study 11 (5.93 and 5.98 Log10
IU/mL respectively), despite some differences in the participants and assays involved in each
study. For samples 1-3, the SDs and %GCVs are 0.11-0.13 log10 and 29-35% respectively. The
overall SDs for samples 1 and 2 are slightly higher than those reported in the 2011 collaborative
study. However, this still represents good agreement between laboratories and assay methods.
For samples 4 and 5, the SDs and %GCVs are 0.22-0.27 log10 and 65-86% respectively. For
samples 6-8, the SDs and %GCVs are 0.29-0.42 log10 and 95-160% respectively. The increased
SDs and %GCVs for samples 4-8 are principally due to the outlying results of laboratory 6 (see
Figure 1). The overall mean estimates, SDs and %GCVs for samples 1-8 excluding the dataset
for laboratory 6 are shown in Table 5. The SDs and %GCVs for samples 4-8 are all reduced (by
approximately 2-fold) when the results for laboratory 6 are excluded. Five laboratories reported
results using the cobas® AmpliPrep/cobas® TaqMan® HBV Test, v2.0 (assay code cTM). This
assay is over-represented in comparison to the other assays. However, removing datasets from 3
laboratories using this assay did not greatly alter the overall laboratory results (Table 6)
425
WHO/BS/2016.2291
Page 8
Potencies relative to the 3rd WHO International Standard for HBV (Sample 1)
The estimated concentrations of samples 2-8 were expressed in IU, by direct comparison
(relative potencies) to the current International Standard (10/264, sample 1), which has an
assigned unitage of 850,00 IU/mL (5.93 log10), as described in the statistical methods section.
The laboratory mean estimates are shown in Table 7 for the quantitative and qualitative assays.
Units are log10 IU/mL in both cases. The results are also shown in histogram form in Figure 2.
The overall mean relative potencies, along with SDs and %GCVs, are shown in Table 8. The
overall mean relative potency for the candidate sample 2 is 5.96 log10 IU/mL, based on the
quantitative assays. This value compares well to the direct estimate of 5.97 log10 IU/mL from the
quantitative assays which are all calibrated in IU/mL.
Figure 2 and Table 8, show an improvement in the agreement between laboratories for samples
2-5 for the quantitative assays. The SD and %GCV between laboratories has reduced for these
samples when compared to the values in Table 4. The reduction in these values for samples 4
and 5 (both genotype C) is less marked than for samples 1-3 (all genotype A), possibly because
of increased variability in the quantification of genotype C viruses between the assays. For
samples 6-8 there is no improvement in the agreement between laboratories when compared to
the values in Table 4. This may be due to variability in the quantification of different HBV
genotypes, although for sample 8 there is no improvement in the agreement between
laboratories, despite samples 1 and 8 comprising genotype A viruses. Again, the over-
representation of the cTM assay in the study did not greatly alter the overall laboratory results
(Table 9).
Potencies relative to the candidate 4th WHO International Standard for HBV
(Sample 2)
The estimated concentrations of samples 3-8 were expressed in IU, by direct comparison
(relative potencies) to the candidate International Standard (10/266, sample 2), based on a
candidate unitage of 955,000 IU/mL (5.98 log10), as described in the statistical methods section.
This candidate unitage was based on the overall mean potency obtained for 10/266, relative to
the 2nd HBV WHO International Standard (97/750), in the 2011 collaborative study.
Overall mean relative potencies, along with SDs and %GCVs, are shown in Table 10. Table 10
shows an improvement in the agreement between laboratories for samples 3-5 for the
quantitative assays. The SD and %GCV between laboratories has reduced for these samples
when compared to the values in Table 4. For samples 6-8 there is no improvement in the
agreement between laboratories when compared to the values in Table 4. Again, this may be due
to variability in the quantification of different HBV genotypes, although for sample 8 there is no
improvement in the agreement between laboratories, despite samples 2 and 8 comprising
genotype A viruses. The over-representation of the cTM assay in the study did not greatly alter
the overall laboratory results (Table 11).
Inter and intra-laboratory variation

For all samples, the inter-laboratory variation was greater than the intra-laboratory variation
(p<0.01). Table 12 shows the intra-laboratory SDs and %GCVs for each laboratory, calculated
by pooling the estimates for samples 1-8. There are differences between the repeatability of
laboratory estimates across assays. In general, the repeatability is good for assays of this type and
the average SD is 0.07 log10 or a %GCV of 18%. These values are slightly improved compared
to the equivalent values obtained in the 2011 collaborative study (average SD of 0.08
log10, %GCV of 20%) 11. These figures represent the variability between individual assay mean
estimates of IU/mL. Since each assay tested multiple replicates of samples at different dilutions,
426
Annex 6
WHO/BS/2016.2291
Page 9
the resulting between-assay variability is lower than would be expected if only a single replicate
was tested in each assay. The ‘NAT detectable units’ from the qualitative assays are obtained by
pooling all assay data to give a single series of number positive out of number tested at each
dilution. As a result, there is no comparable analysis of intra-assay variation for the qualitative
assay.
Conclusions
In this study, a range of NAT-based assays for HBV have been used to evaluate the suitability of
the candidate standard (NIBSC code 10/266) as the replacement 4th WHO International Standard
for HBV DNA for NAT-based assays. The candidate was prepared from the same virus stock
used for previous HBV WHO International Standards and was diluted in a similar pooled human
plasma material 6,7,9,10. Production data suggests that the batch is homogeneous and contains
residual moisture and oxygen levels that are within WHO and NIBSC limits for lyophilized
standards 11,17. Comparison of the liquid bulk versus the lyophilized product indicates that there
was minimal loss in potency upon freeze-drying (0.04 log10 IU/mL). The results of ongoing
accelerated thermal degradation studies at 60 months indicate that the candidate is stable and
suitable for long-term use.
In the collaborative study, the lyophilized candidate (sample 2) was evaluated alongside the 3rd
HBV WHO International Standard (sample 1). The overall mean estimates for samples 1 and 2
were 5.94 and 5.97 Log10 IU/mL, respectively, based on the calibration of quantitative assay kits
in IU/mL. These values are very similar to the values obtained for these samples relative to the
2nd HBV WHO International Standard in the 2011 collaborative study 11, despite laboratories
reporting results using different HBV NAT-based assays. In the present study, the agreement
between laboratories for sample 2 was improved when the potency was expressed relative to the
3rd HBV WHO International Standard (sample 1). There was some evidence for variability in the
quantification of different HBV genotypes present in the different study samples. This has been
reported previously 18. Inter-laboratory variability was higher than intra-laboratory variability for
the quantitative assays. This highlights the continued need for standardization of HBV NAT-
based assays, and the importance of accurate calibration to the WHO International Standard.
A full assessment of commutability of the candidate standard for HBV-positive samples has not
been possible in this study, due to the limited number of clinical samples that could be included.
Three HBV-positive plasma samples comprising HBV genotypes A, D and E, from rejected
blood donations were included in the study. There was no overall improvement in the agreement
between laboratories when the estimated concentrations of the three HBV plasma samples were
expressed relative to the candidate 4th HBV WHO International Standard (same formulation as
previous HBV WHO International Standards), compared to the uncorrected values. This is
principally due to variability in the quantification of different HBV genotypes between different
assays, and outlying results from one or two laboratories for samples 6-8. There is no evidence
of non-commutability with the three plasma samples that were included in the study.
In summary, the results of the study indicate the suitability of the candidate sample 2 as the
replacement 4th WHO International Standard for HBV DNA for NAT. Since the overall mean
potency obtained for the candidate in this collaborative study is very similar to the overall mean
potency obtained in the 2011 collaborative study, relative to the pre-existing 2nd HBV WHO
International Standard, it is proposed that the value assigned to the candidate sample 2 is that
obtained in the 2011 collaborative study. This approach would minimize any potential drift in the
value of the IU during the replacement.
427
WHO/BS/2016.2291
Page 10
Proposal
It is proposed that the candidate standard, NIBSC code 10/266, is established as the 4th WHO
International Standard for HBV DNA for use in NAT-based assays, with an assigned
potency of 955,000 IU/mL (~5.98 log10 IU/mL) when reconstituted in 0.5 mL of nuclease-free
water. This potency is based on the value that was assigned in the collaborative study in 2011
where the candidate was assessed alongside the 2nd HBV WHO International Standard. The
uncertainty can be derived from the variance of the fill weight and is 0.36%. The proposed
standard is intended to be used by IVD manufacturers, blood transfusion centres, control
authorities, and clinical laboratories, to calibrate secondary reference materials used in HBV
NAT assays. Proposed Instructions for Use (IFU) for the product are included in Appendix 3.
Comments from participants

8 of 13 participants responded to the report. There were no disagreements with the suitability of
the candidate standard (NIBSC code 10/266) to serve as the 4th WHO International Standard for
HBV DNA for NAT-based assays. Some comments suggested minor editorial changes and these
have been implemented.
Acknowledgements
We gratefully acknowledge the important contributions of the collaborative study participants.
References
1. World Health Organisation, Hepatitis B, Fact sheet No. 204, updated July 2014.
http://www.who.int/mediacentre/factsheets/fs204/en/
2. Roth WK1, Busch MP, Schuller A, Ismay S, Cheng A, Seed CR, Jungbauer C, Minsk
PM, Sondag-Thull D, Wendel S, Levi JE, Fearon M, Delage G, Xie Y, Jukic I, Turek P,
Ullum H, Tefanova V, Tilk M, Reimal R, Castren J, Naukkarinen M, Assal A, Jork C,
Hourfar MK, Michel P, Offergeld R, Pichl L, Schmidt M, Schottstedt V, Seifried E,
Wagner F, Weber-Schehl M, Politis C, Lin CK, Tsoi WC, O'Riordan J, Gottreich A,
Shinar E, Yahalom V, Velati C, Satake M, Sanad N, Sisene I, Bon AH, Koppelmann M,
Flanagan P, Flesland O, Brojer E, Ltowska M, Nascimento F, Zhiburt E, Chua SS, Teo
D, Stezinar SL, Vermeulen M, Reddy R, Park Q, Castro E, Eiras A, Gonzales Fraile I,
Torres P, Ekermo B, Niederhauser C, Chen H, Oota S, Brant LJ, Eglin R, Jarvis L,
Mohabir L, Brodsky J, Foster G, Jennings C, Notari E, Stramer S, Kessler D, Hillyer C,
Kamel H, Katz L, Taylor C, Panzer S, Reesink HW. International survey on NAT testing
of blood donations: expanding implementation and yield from 1999 to 2009. Vox Sang.
2012;102(1):82-90. doi: 10.1111/j.1423-0410.2011.01506.x.
3. Seo DH, Whang DH, Song EY, Han KS. Occult hepatitis B virus infection and blood
transfusion. World J Hepatol. 2015;7(3):600-6. doi: 10.4254/wjh.v7.i3.600.

4. Candotti D, Allain JP. Transfusion-transmitted hepatitis B virus infection. J Hepatol.
2009;51:798-809.
5. Sorrell MF, Belongia EA, Costa J, Gareen IF, Grem JL, Inadomi JM, Kern ER, McHugh
JA, Petersen GM, Rein MF, Sorrell MF, Strader DB, Trotter HT. National Institutes of
Health Consensus Development Conference Statement: management of hepatitis B. Ann
Intern Med. 2009;150:104-10.
6. Saldanha J, Gerlich W, Lelie N, Heermann K, Heath A. An international collaborative
study to establish a WHO International Standard for HBV DNA nucleic acid
amplification technology assay. WHO ECBS Report 1999; WHO/BS/99.1917.
http://www.who.int/bloodproducts/cs/en/991917.pdf
428
Annex 6
WHO/BS/2016.2291
Page 11
7. Saldanha J, Gerlich W, Lelie N, Dawson P, Heermann K, Heath A; WHO Collaborative
Study Group. An international collaborative study to establish a World Health
Organization international standard for hepatitis B virus DNA nucleic acid amplification
techniques. Vox Sang. 2001;80:63-71.
8. Heermann KH, Gerlich WH, Chudy M, Schaefer S, Thomssen R. Quantitative detection
of hepatitis B virus DNA in two international reference plasma preparations. Eurohep
Pathobiology Group. J Clin Microbiol. 1999;37:68-73.
9. Baylis SA, Gerlich W, Chudy M, Pisani G, Klotz A, Kerby S, Heath A. WHO
collaborative study to establish a replacement WHO International Standard for hepatitis
B virus DNA nucleic acid amplification technology (NAT) assays. WHO ECBS Report
2006; WHO/BS/06.2034.
https://extranet.who.int/iris/restricted/bitstream/10665/69957/1/WHO_BS_06.2034_eng.
pdf
10. Baylis SA, Heath AB, Chudy M, Pisani G, Klotz A, Kerby S, Gerlich W. An
international collaborative study to establish the 2nd World Health Organization
International Standard for hepatitis B virus DNA nucleic acid amplification technology-
based assays. Vox Sang. 2008;94:358-62.
11. Fryer JF, Heath AB, Wilkinson DE, Minor PD and the Collaborative Study Group.
Collaborative Study to Evaluate the Proposed 3rd WHO International Standard for
Hepatitis B Virus (HBV) for Nucleic Acid Amplification Technology (NAT)-Based
Assays. WHO ECBS Report 2011;WHO/BS/2011.2170.
http://apps.who.int/iris/bitstream/10665/70779/1/WHO_BS_2011.2170_eng.pdf
12. ISO 17511:2003. In vitro diagnostic medical devices  measurement of quantities in
biological samples metrological traceability of values assigned to calibrators and
control materials. Geneva: International Organization for Standardization (ISO); 2003.
13. Kirschberg O, Schüttler C, Repp R, Schaefer S. A multiplex-PCR to identify hepatitis B
virus-genotypes A-F. J Clin Virol. 2004;29:39-43.
14. WHO Consultation on Global Measurement Standards and their use in the in vitro
Biological Diagnostic Field, Switzerland, Geneva, Switzerland, 7-8 June 2004:
http://www.who.int/bloodproducts/publications/en/Minutes-220804.pdf
15. Saldanha J; Lelie N and Heath A B. Establishment of the First International Standard for
Nucleic Acid Amplification Technology (NAT) assays for HCV RNA. 1999. Vox
Sanquinis. 76:149-58.
16. Kirkwood TBL. Geometric means and measures of dispersion. Biometrics 1979;35:908-
9.
17. Recommendations for the preparation, characterization and establishment of international
and other biological reference standards (revised 2004). WHO Technical Report Series
2007. Geneva, Switzerland:WHO 2007; 932,73-131.
http://www.who.int/bloodproducts/publications/TRS932Annex2_Inter_biolefstandardsre
v2004.pdf
18. Chudy M, Hanschmann KM, Kress J, Gerlich W and Nübling CM. Collaborative study to
establish a World Health Organization International Genotype Panel for hepatitis B virus
nucleic acid amplification technique (NAT)-based assays. WHO ECBS Report
2009;WHO/BS/09.2121. http://whqlibdoc.who.int/hq/2009/WHO_BS_09.2121_eng.pdf
429
WHO/BS/2016.2291
Page 12
Table 1. Thermal stability of 10/266 at different storage temperatures. * Mean results from
single extractions from 3 vials at each time point and temperature.
Temperature (°C) Mean log10 IU/mL * (difference in log10 IU/mL from -

20 °C baseline sample)
11 weeks 12 months 34 months 60 months
-70 6.04 6.05 5.96 5.77
-20 6.03 6.02 5.92 5.77
4 6.07 (0.04) 6.01 (-0.01) 5.91 (-0.01) 5.80 (0.03)
20 6.06 (0.03) 6.01 (-0.01) 5.91 (-0.01) 5.89 (0.12)
37 6.09 (0.06) - -
45 6.13 (0.10) - -
Table 2. Collaborative study assay methods and codes.
Quantitative Assays
Assay Code Assay No. of datasets
kPCR VERSANT HBV DNA 1.0 Assay (kPCR) (Siemens 1
Healthcare Diagnostics)
cTM cobas® AmpliPrep/cobas® TaqMan® HBV Test, 5
v2.0 (Roche Molecular Systems, Inc.)
c68 cobas® HBV test for use on the cobas® 6800/8800 2
Systems (Roche Molecular Systems, Inc.)
AbRT Abbott RealTime HBV (Abbott Molecular, Inc.) 1
ArQS artus® HBV QS-RGQ Kit, Version 1 (QIAGEN) 1
APT Aptima HBV Quant assay on the Panther system 1
(Hologic, Inc.)
VER VERIS HBV Assay (Beckman Coulter, Inc.) 1
SKB HBV DNA real-time PCR detection kit (Shanghai 1
Kehua Bio-Engineering Co., Ltd.)
SaB Hepatitis B Viral DNA Quantitative Fluorescence 1
Diagnostic Kit (PCR Fluorescence Probing) Mag
(Sansure Biotech, Inc.)

Qualitative Assays
Assay Code Assay No. of datasets
cTSM cobas® TaqScreen MPX Test, v2.0 (Roche 1
Molecular Systems, Inc.)
430
Annex 6
WHO/BS/2016.2291
Page 13
Table 3. Laboratory mean estimates from quantitative assays (log10 IU/mL) and qualitative
assays (log10 ‘NAT detectable units/mL’). Qualitative results are shaded in grey. nd, not
determined (either not tested or data excluded).
Lab Assay Sample

S1 S2 S3 S4 S5 S6 S7 S8
1 kPCR 6.01 6.08 5.74 5.98 3.00 3.41 3.45 3.11
2A c68 6.00 6.01 5.68 6.05 3.06 3.28 3.24 3.54
2B cTM 5.95 5.98 5.67 6.16 3.19 3.54 3.51 3.78
3 AbRT 5.86 5.87 5.58 5.91 2.92 3.59 3.27 3.55
4 ArQS 6.15 6.19 5.88 6.14 nd 3.16 2.86 3.94
5 cTM 5.91 5.99 5.68 6.17 3.18 3.52 3.50 3.71
6 VER 5.79 5.81 5.61 5.45 2.33 1.98 2.50 2.39
7 APT 5.86 5.88 5.62 5.93 3.05 3.09 3.11 3.40
8 c68 5.92 5.96 5.63 6.06 3.12 3.40 3.30 3.61
9 cTM 5.95 5.96 5.71 6.15 3.25 3.45 3.47 3.67
10 cTM 5.94 5.95 5.69 6.21 3.20 3.47 3.54 3.87
11A SKB 6.23 6.26 5.91 6.30 3.10 3.43 3.40 3.44
11B SaB 5.75 5.78 5.48 5.78 nd 3.70 3.33 3.76
12 cTSM 6.06 6.57 6.3 6.16 2.78 nd nd nd
13 cTM 5.88 5.93 5.68 6.16 3.16 3.42 3.42 3.62
Table 4. Overall mean estimates and inter-laboratory variation for quantitative assays (log10
IU/mL).
Sample No. of Mean Min Max SD %GCV

datasets
S1: 10/264 (gt.A) 14 5.94 5.75 6.23 0.13 35
S2: 10/266 (gt.A) 14 5.97 5.78 6.26 0.13 35
S3: SRR (gt.A) 14 5.68 5.48 5.91 0.11 29
S4: NS (gt.C) 14 6.03 5.45 6.30 0.22 65
S5: WR (gt.C) 12 3.05 2.33 3.25 0.24 75
S6: HBVpl (gt.D) 14 3.32 1.98 3.70 0.42 160
S7: HBVpl (gt.E) 14 3.28 2.50 3.54 0.29 95
S8: HBVpl (gt.A) 14 3.53 2.39 3.94 0.39 145
431
WHO/BS/2016.2291
Page 14
IU/mL), excluding the dataset from laboratory 6.

datasets
S1: 10/264 (gt.A) 13 5.96 5.75 6.23 0.13 34
S2: 10/266 (gt.A) 13 5.99 5.78 6.26 0.13 34
S3: SRR (gt.A) 13 5.69 5.48 5.91 0.11 30
S4: NS (gt.C) 13 6.07 5.78 6.30 0.14 39
S5: WR (gt.C) 11 3.06 2.55 3.26 0.19 54
S6: HBVpl (gt.D) 13 3.42 3.09 3.70 0.17 47
S7: HBVpl (gt.E) 13 3.34 2.86 3.54 0.19 55
S8: HBVpl (gt.A) 13 3.61 3.11 3.94 0.22 66
IU/mL), excluding datasets from laboratories 9, 10 and 13 using the cTM assay.

datasets
S1: 10/264 (gt.A) 11 5.95 5.75 6.23 0.15 40
S2: 10/266 (gt.A) 11 5.98 5.78 6.26 0.15 41
S3: SRR (gt.A) 11 5.68 5.48 5.91 0.13 34
S4: NS (gt.C) 11 5.99 5.45 6.30 0.23 71
S5: WR (gt.C) 9 2.95 2.33 3.19 0.28 91
S6: HBVpl (gt.D) 11 3.28 1.98 3.70 0.47 193
S7: HBVpl (gt.E) 11 3.22 2.50 3.51 0.31 102
S8: HBVpl (gt.A) 11 3.47 2.39 3.94 0.42 165
432
Annex 6
WHO/BS/2016.2291
Page 15
Table 7. Laboratory estimates of potency relative to the 3rd WHO International Standard for
HBV (sample 1) from quantitative assays and qualitative assays (log10 IU/mL) - based on an
assigned unitage of the International Standard of 850,000 (5.93 log10) IU/mL. nd, not determined
(either not tested or data excluded).
Lab Assay Sample

S2 S3 S4 S5 S6 S7 S8
1 kPCR 6.00 5.67 5.90 2.92 3.33 3.38 3.03
2A c68 5.93 5.61 5.98 2.99 3.20 3.17 3.47
2B cTM 5.96 5.65 6.14 3.17 3.52 3.49 3.76
3 AbRT 5.93 5.64 5.98 2.99 3.65 3.34 3.62
4 ArQS 5.97 5.65 5.91 nd 2.94 2.63 3.72
5 cTM 6.01 5.70 6.19 3.19 3.54 3.52 3.73
6 VER 5.95 5.76 5.59 2.48 2.13 2.64 2.54
7 APT 5.95 5.69 6.00 3.12 3.16 3.19 3.47
8 c68 5.97 5.63 6.06 3.13 3.40 3.31 3.62
9 cTM 5.94 5.69 6.13 3.23 3.43 3.45 3.65
10 cTM 5.94 5.68 6.20 3.19 3.46 3.52 3.86
11A SKB 5.96 5.61 6.00 2.80 3.13 3.10 3.14
11B SaB 5.97 5.66 5.97 nd 3.89 3.51 3.94
12 cTSM 6.44 6.17 5.86 2.65 nd nd nd
13 cTM 5.98 5.73 6.21 3.22 3.47 3.47 3.67
Table 8. Overall mean estimates and inter-laboratory variation for potency relative to the 3rd
HBV WHO International Standard (sample 1) log10 IU/mL for quantitative assays - based on an
assigned unitage of the International Standard of 850,000 (5.93 log10) IU/mL.

datasets
S2: 10/266 (gt.A) 14 5.96 5.94 6.01 0.02 5
S3: SRR (gt.A) 14 5.67 5.61 5.76 0.04 10
S4: NS (gt.C) 14 6.02 5.59 6.21 0.16 46
S5: WR (gt.C) 12 3.03 2.48 3.23 0.22 67
S6: HBVpl (gt.D) 14 3.30 2.13 3.89 0.41 159
S7: HBVpl (gt.E) 14 3.27 2.63 3.52 0.30 100
S8: HBVpl (gt.A) 14 3.51 2.54 3.94 0.38 138
433
WHO/BS/2016.2291
Page 16
Table 9. Overall mean estimates and inter-laboratory variation for potency relative to the 3rd
HBV WHO International Standard (sample 1) log10 IU/mL for quantitative assays - based on an
assigned unitage of the International Standard of 850,000 (5.93 log10) IU/mL, excluding datasets
from laboratories 9, 10 and 13 using the cTM assay.

datasets
S2: 10/266 (gt.A) 11 5.96 5.93 6.01 0.02 5
S3: SRR (gt.A) 11 5.66 5.61 5.76 0.04 10
S4: NS (gt.C) 11 5.98 5.59 6.19 0.16 43
S5: WR (gt.C) 9 2.97 2.48 3.19 0.23 69
S6: HBVpl (gt.D) 11 3.26 2.13 3.89 0.46 189
S7: HBVpl (gt.E) 11 3.21 2.63 3.52 0.31 106
S8: HBVpl (gt.A) 11 3.46 2.54 3.94 0.41 155
Table 10. Overall mean estimates and inter-laboratory variation for potency relative to the
candidate 4th HBV WHO International Standard (sample 2) log10 IU/mL for quantitative assays -
based on a candidate unitage of 955,000 (5.98 log10) IU/mL.

datasets
S3: SRR (gt.A) 14 5.69 5.64 5.79 0.04 10
S4: NS (gt.C) 14 6.04 5.62 6.24 0.16 45
S5: WR (gt.C) 12 3.05 2.51 3.27 0.22 66
S6: HBVpl (gt.D) 14 3.32 2.16 3.90 0.41 157
S7: HBVpl (gt.E) 14 3.28 2.64 3.56 0.30 98
S8: HBVpl (gt.A) 14 3.53 2.57 3.95 0.38 138
434
Annex 6
WHO/BS/2016.2291
Page 17
Table 11. Overall mean estimates and inter-laboratory variation for potency relative to the
candidate 4th HBV WHO International Standard (sample 2) log10 IU/mL for quantitative assays -
based on a candidate unitage of 955,000 (5.98 log10) IU/mL, excluding datasets from laboratories
9, 10 and 13 using the cTM assay.

datasets
S3: SRR (gt.A) 11 5.68 5.64 5.79 0.04 10
S4: NS (gt.C) 11 5.99 5.62 6.16 0.15 41
S5: WR (gt.C) 9 2.99 2.51 3.18 0.22 66
S6: HBVpl (gt.D) 11 3.28 2.16 3.90 0.46 186
S7: HBVpl (gt.E) 11 3.22 2.64 3.52 0.31 103
S8: HBVpl (gt.A) 11 3.47 2.57 3.95 0.40 153
Table 12. Intra-laboratory SD of log10 IU/mL and %GCV for quantitative assays.
Lab Assay SD %GCV

1 kPCR 0.05 13
2A c68 0.03 7
2B cTM 0.05 12
3 AbRT 0.03 8
4 ArQS 0.17 47
5 cTM 0.04 11
6 VER 0.06 15
7 APT 0.08 21
8 c68 0.03 6
9 cTM 0.05 11
10 cTM 0.07 16
11A SKB 0.09 24
11B SaB 0.05 13
13 cTM 0.05 13
Overall 0.07 18
WHO/BS/2016.2291
Page 18
Figure legends
Figure 1. Individual laboratory mean estimates for study samples 1-8 (a-h respectively)
obtained using qualitative or quantitative NAT assays. Each box represents the mean estimate
from each laboratory assay and is labeled with the laboratory and assay code. Boxes are also
colour coded by assay.
Figure 2. Relative potencies of samples 2-8 against sample 1 (a-g respectively), for each
qualitative or quantitative assay. Units are expressed as candidate log10 IU/mL. Each box
represents the relative potency for each laboratory assay and is labeled with the laboratory and
assay code. Boxes are also colour coded by assay.
435
WHO/BS/2016.2291
Page 19
Figure 1
a Sample 1
WHO/BS/2016.2291
Page 20
b Sample 2
436
Annex 6
WHO/BS/2016.2291
Page 21
c Sample 3
WHO/BS/2016.2291
Page 22
d Sample 4
437
WHO/BS/2016.2291
Page 23
e Sample 5
WHO/BS/2016.2291
Page 24
f Sample 6
438
Annex 6
WHO/BS/2016.2291
Page 25
g Sample 7
WHO/BS/2016.2291
Page 26
h Sample 8
439
WHO/BS/2016.2291
Page 27
Figure 2
a Sample 2 relative to Sample 1
WHO/BS/2016.2291
Page 28
b Sample 3 relative to Sample 1

440
Annex 6
WHO/BS/2016.2291
Page 29
c Sample 4 relative to Sample 1
WHO/BS/2016.2291
Page 30
d Sample 5 relative to Sample 1
441
WHO/BS/2016.2291
Page 31
e Sample 6 relative to Sample 1
WHO/BS/2016.2291
Page 32
f Sample 7 relative to Sample 1

442
Annex 6
WHO/BS/2016.2291
Page 33
g Sample 8 relative to Sample 1
443
Appendix 2
Example of the calibration of a national standard
(collaborative study calibration using multiple assays)
Journal of Virological Methods 191 (2013) 122–127
Contents lists available at SciVerse ScienceDirect
Journal of Virological Methods

journal homepage: www.elsevier.com/locate/jviromet
A collaborative study to establish the first National Standard for HIV-1 RNA
nucleic acid amplification techniques (NAT) in Taiwan
Yi-Chen Yang a,b,∗ , Daniel Yang-Chih Shih a , Mong-Hsun Tsai b , Chia-Hung Cheng a , Hwei-Fang Cheng a ,
Chi-Fang Lo a , Der-Yuan Wang a
a
Food and Drug Administration, Department of Health, Executive Yuan, Taipei 11561, Taiwan
b
Institute of Biotechnology, National Taiwan University, Taipei 10672, Taiwan
a b s t r a c t
Article history: The World Health Organization (WHO) International Standard (IS) for human immunodeficiency virus
Received 5 January 2013 type 1 (HIV-1) RNA is only available in limited amounts. It is critical to use the most common HIV-1
Received in revised form 26 March 2013 genotype as source for secondary standards, e.g. a National Standard for Taiwan. The objective of this
Accepted 4 April 2013
study was to establish the first National Standard for HIV-1 RNA NAT assays in Taiwan. A collaborative
Available online 19 April 2013
study, including eleven laboratories from five different countries, was carried out to establish the HIV-1
RNA National Standard by calibration, in International Units (IU), against the WHO HIV-1 RNA IS. The
Keywords:
HIV-1 RNA content for the candidate was quantitated by each laboratory in three independent assays
Standard
Nucleic acid amplification techniques (NAT)
and the results were collected and analyzed statistically. Overall, a high level of agreement among results
Human immunodeficiency virus type 1 was achieved from different laboratories. In addition, the stability study indicated that the candidate was
(HIV-1) stable for 24 months at −80 ± 5 ◦ C. In conclusion, the candidate standard was established as the first
National Standard for HIV-1 RNA for use in NAT assays in Taiwan. The standard is intended to be used for
the quality control of HIV-1 NAT assays and as a quantitative reference material for HIV-1 NAT assays.
© 2013 Elsevier B.V. All rights reserved.
1. Introduction basis of differences in the envelope region. The M group is further

divided into 9 genetic subtypes (A–D, F–H, J, and K) and circulating
Human immunodeficiency virus type 1 (HIV-1) is the causative recombinant forms (CRFs). In Taiwan, subtype B was found to be
agent of acquired immunodeficiency syndrome, commonly known the predominant genotype in homosexual males and in the intra-
as AIDS (Weiss, 1993). According to the UNAIDS report on the venous drug abuser population. In recent years, subtype CRF 07 BC
global AIDS epidemic in 2012, an estimated 34 million people has been the major group in intravenous drug abuser population
are infected HIV (UNAIDS report, 2012) after 30 years of a very (Spira et al., 2003; Robertson et al., 2000; Simon et al., 1998; Lin
complex epidemic. Almost 30 million people have died from HIV- et al., 2007; Plantier et al., 2009).
related diseases so far. In Taiwan, the number of reported cases Screening of blood and plasma products for blood–borne viruses
of HIV/AIDS was approximately twenty four thousand from 1984 has usually been performed using sensitive antibody-detection
to 2012 December (CDC of Taiwan, 2012). Increasing sexual activ- assays. In recent years, the nucleic-acid amplification techniques
ity and needle sharing activity among drug abusers has resulted in (NAT) have been widely applied in blood safety screens to enhance
HIV/AIDS becoming a severe public health problem. the sensitivity of detection of HIV-1 when present in low concen-
HIV-1 is a member of the Retroviridae family and belongs to the trations and at earlier stages of infection (Piatak et al., 1993; Murthy
Lentivirus genus. The RNA genome of HIV-1 is approximately 9.7 kb, et al., 1999; Busch and Dodd, 2000).
containing three structural genes (gag, pol, and env) and six regu- To improve the safety of plasma products, a requirement that
lating genes (tat, rev, nef, vif, vpu, and vpr) (Pluta and Kacprzak, the plasma pools used to manufacture plasma products should be
2009; Bolinger and Boris-Lawrie, 2009; Karlsson Hedestam et al., screened for HIV RNA by NAT was announced by the Department
2008). HIV-1 strains are categorized as major group (M group), of Health in Taiwan on December 19, 2002. The development of
outlier group (O group), new group (N group), or P group on the a calibrated national reference standard that could be used rou-
tinely in assays would give assurance as to the validity of the test
results and therefore fulfillment with such regulations. In addi-
∗ Corresponding author at: 161-2, Kunyang St., Nangang, Taipei 11561, Taiwan. tion to this national requirement for plasma screening, the HIV
Tel.: +886 2 27877751. viral load assay is very critical in the management of antiretroviral
E-mail address: ycyang@fda.gov.tw (Y.-C. Yang). therapy. Recently, numerous in vitro diagnostic devices (IVDs)
444 0166-0934/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.jviromet.2013.04.002
Annex 6
Y.-C. Yang et al. / Journal of Virological Methods 191 (2013) 122–127 123
based on NAT technology have been developed to detect HIV proposed national candidate standard and WHO IS for HIV-1 RNA
qualitatively or quantitatively for blood screening or viral load mea- (97/650) and were requested to perform three independent assays
surement. Such NAT-based IVDs are highly related to blood safety, for HIV-1 RNA using the candidate standard and the WHO IS. For
the quality and performance of the IVDs are of great importance. In each assay, serial dilutions of the WHO IS were prepared using the
order to ensure the continued fitness for purpose of the IVD, both appropriate diluent. The recommended diluent for the study was
the pre-market approval testing and the performance evaluation HIV-1 RNA negative human plasma. The serially diluted IS were
are crucial in post-marketing surveillances. used to calibrate the candidate standard (Sample A) by creating a
The first and second World Health Organization (WHO) Inter- standard curve. If a commercial kit was used, the IS could be treated
national Standard (IS) for HIV-1 RNA was established by the WHO as a second unknown sample (sample B) and quantitated in parallel
Expert Committee on Biological Standardization (ECBS) in 1999 with sample A. A single estimate was obtained for each sample in
and 2008, and the National Institute for Biological Standards and each laboratory and for each assay method by calculating the geo-
Control (NIBSC) code numbers are 97/656 and 97/650, respectively metric mean of repeat data within a single assay. The overall mean
(Holmes et al., 2001; Davis et al., 2008). One of the main purposes and SD were then calculated from the results of all participants.
of an International Standard is to facilitate the calibration of sec-
ondary working reagents developed at a local level, i.e. by individual 2.3. Stability study on the candidate standard
laboratories. The calibration of secondary working reagents from
a higher order standard would help reduce test result variability Vials of the proposed national candidate standard were incu-
from different laboratories, aid in comparing the different com- bated at +4 ◦ C, +24 ◦ C, −20 ◦ C, and −80 ◦ C, three vials were removed
mercial and ‘in-house’ assays, and make it easier to compare the at regular intervals for three independent tests. Two different com-
proficiency of different laboratories (Revets et al., 1996; Schuurman mercial assays were used for quantitative analysis: the Abbott
et al., 1996). In order to ensure the correct use of the Interna- RealTime HIV-1 (Abbott Molecular Inc., Des Plaines, IL, USA) and
tional Standard by the end user, for example for secondary working the COBAS Ampliprep/COBAS TaqMan HIV-1 Test, v1.0 (Roche
reagent calibration and not for use as a routine run control, the Molecular Systems, Inc., Branchburg, NJ, USA). Here, 500 uL of the
WHO IS for HIV-1 RNA is available only in limited amounts, several HIV-1 RNA candidate standard was used in Abbott RealTime HIV-
control laboratories (such as National Institute for Biological Stan- 1 Kit and 1 mL 5× pre-diluted candidate standard was used in
dards and Control (NIBSC), Food and Drug Administration (FDA) COBAS Ampliprep/COBAS TaqMan HIV-1 Test Kit. Both real-time
and Istituto Superiore di Sanità (ISS)) have already prepared an PCR systems, the Abbott m2000 RealTime system and the COBAS
in house or national secondary HIV-1 NAT working reagent them- Ampliprep/TaqMan 48 system, were used according to the manu-
selves (Davis et al., 2003; Lee et al., 2006; Pisani et al., 2007). It facturer’s instructions.
is known that the distribution of HIV subtypes may differ by geo-
graphic region. It is therefore critical to use the major genotype of 3. Results
the HIV-1 as a source material for a National Standard. Since sub-
type B was found to be the predominant genotype in Taiwan, the 3.1. Assay methods
genotype of the National Standard would select to be subtype B, the
same as WHO IS. Therefore, the objective of this study was to estab- Ten of the participants performed quantitative assays: four
lish the first National Standard for HIV-1 RNA NAT assays and to laboratories used the Abbott RT HIV-1 (Abbott Molecular
calibrate the HIV-1 RNA content of the candidate standard against Inc., Des Plaines, IL, USA); two laboratories used the COBAS
the WHO IS for HIV-1 RNA NAT assays (97/650). The procedure for AmpliPrep/COBAS TaqMan HIV-1 Test, v1.0 (Roche Molecular Sys-
the development of a National Standard was based on the previous tems, Inc., Branchburg, NJ, USA); two laboratories used the COBAS
experience of the development of the National Standard for human AmpliPrep/COBAS TaqMan HIV-1 Test, v2.0 (Roche Molecular Sys-
parvovirus B19 DNA nucleic acid amplification techniques (NAT) in tems, Inc., Branchburg, NJ, USA); two laboratories used the COBAS
2008 (Yang et al., 2008). TaqMan HIV-1 Test, for use with High Pure System Viral Nucleic
Acid Kit, v1.0 (Roche Molecular Systems, Inc., Branchburg, NJ,
USA); two laboratories used the VERSANT HIV-1 RNA 3.0 (Siemens
2. Materials and methods
Healthcare Diagnostics Inc., Tarrytown, NY, USA); and one labo-
ratory used the COBAS Amplicor HIV-1 Monitor Test, v1.5 (Roche
2.1. Preparation of the candidate standard
Molecular Systems, Inc., Branchburg, NJ, USA). Between them, lab
code 1 used two different methods to detect HIV-1RNA and ana-
The candidate standard for HIV-1 RNA NAT assays was liquid
lyzed the data separately (1A, 1B), and lab code 2 used three
preparation and stored at or below −70 ◦ C. It was prepared by dilut-
different methods to detect HIV-1RNA and reported the results
ing HIV-1 RNA positive plasma in pooled human cryosupernatant.
separately (2A, 2B, 2C). The overall results were therefore based
The proposed titer was approximately 104 IU/mL. The original HIV-
on a maximum of 13 data sets. All these data sets were obtained
1 RNA positive plasma had a titer of HIV-1 RNA of approximately
by commercial assays. The quantitative methods used are summa-
4.7 × 104 IU/mL and was negative for HBsAg, HBV DNA, anti-HCV,
rized in Table 1. The other one participant used the COBAS HIV-1
HCV RNA, HAV RNA as well as B19V DNA. The genotype of the HIV-
AmpliScreen Test, v1.5 (Roche Molecular Systems, Inc., Branchburg,
1 RNA positive plasma was confirmed as subtype B by sequencing.
NJ, USA), which is a qualitative assay that was only give “positive”
The cryosupernatant was negative for HBsAg, HBV DNA, B19V DNA,
results (Detection limit: 78.4 IU/mL) and could not be calculated.
anti-HCV, HCV RNA, anti-HIV 1/2, HIV-1 RNA, and HAV RNA.
3.2. Estimated value of the HIV-1 RNA for the candidate standard
2.2. Design of the international collaborative study
The estimated values of HIV-1 RNA, relative to the International
The aim of the international collaborative study was to calibrate Standard, for the candidate standards from each laboratory are
the titers of the HIV-1 RNA National Standard that was prepared listed in Table 2 and shown in Fig. 1. All the values have shown a
by the Taiwan Food and Drug Administration (TFDA). Including good agreement with each other, except one laboratory has submit-
our laboratory, a total of eleven laboratories from five different ted an outlying result. The value of HIV-1 RNA estimate from each
countries have participated in this study. Participants received the laboratory is shown in Fig. 2. Each box represents the estimate from
445
124 Y.-C. Yang et al. / Journal of Virological Methods 191 (2013) 122–127
Table 1
An overview of the quantitative assays used in the collaborative study.
Lab codea Assay method Region for primer design
1B Abbott RT HIV-1 A highly conserved region in HIV-1

2A pol-int genesb
3
8
1A COBAS AmpliPrep/COBAS A highly conserved region in

6 TaqMan HIV-1 Test, v1.0 HIV-1 gag p41 gene
2C COBAS AmpliPrep/COBAS Two highly conserved regions of the HIV-1

5 TaqMan HIV-1 Test, v2.0 genome–gag and long terminal repeat (LTR)
2B COBAS TaqMan HIV-1 Test, for use with A highly conserved region in
10 High Pure System Viral Nucleic Acid Kit, v1.0 HIV-1 gag p41 gene
9 COBAS Amplicor HIV-1 Monitor Test, v1.5 A highly conserved regions in HIV-1 gag gene
4 VERSANT HIV-1 RNA 3.0 (b DNA) A highly conserved region in HIV-1

7 pol gene
a
Two laboratories (lab code 1and 2) returned data from two and three different assay methods, respectively. The results are reported separately.
b
pol-int gene is the integrase region of the polymerase gene.
A Table 2
8 The estimated values of HIV-1 RNA (Log IU/mL)a for candidate standard from 10
7 Mean(4.04) laboratories.
-2SD(3.66)
6 +2SD(4.42)
Lab code Mean Minimum Maximum CV (%)
Log IU/mL
5
4 1A 4.04 4.03 4.06 0.37
3 1B 3.98 3.96 4.01 0.58
2A 4.13 4.06 4.23 1.37
2
2B 4.11 4.01 4.18 1.63
1
2C 4.13 4.06 4.24 1.79
0 3 4.15 4.12 4.17 0.63
1A 1B 2A 2B 2C 3 4 5 6 7 8 9 10
4 3.73 3.67 3.79 1.04
Lab code 5 4.03 3.99 4.08 1.09
6 3.98 3.84 4.08 4.26
B 7 4.18 4.04 4.26 2.83
8 8 3.95 3.90 4.03 1.84
Mean(4.01)
7 9 – 4.42b 4.42b –
-2SD(3.69)
6 +2SD(4.33) 10 3.68 3.65 3.70 0.59
Log IU/mL
5 a
The measurements were performed using WHO International Standard for
4 human HIV-1 RNA (WHO IS, 97/650) as the standard.
b
3 Only one assay result was available from this laboratory, not three independent
2 assay results.
1
0
1A 1B 2A 2B 2C 3 4 5 6 7 8 9 10
one laboratory and/or assay method. All data were within a range of
Lab code 1.0 Log for each sample, indicating that all the laboratories were in
good agreement with the estimates. A comparison of the different
Fig. 1. The estimated values of HIV-1 RNA (Log IU/mL) for candidate standard as
determined by 10 laboratories (A) and by 9 laboratories (B). The results showed commercial kit results is shown in Table 3; the results showed that
that all the laboratories were in good agreement with the estimates, except one the COBAS Amplicor HIV-1 Monitor Test, v1.5 was significantly dif-
laboratory submitted an outlying result (A). The data generated from Lab code 9 ferent from other commercial assay kits. The data generated from
was excluded from the overall means for the candidate standard (B). the COBAS Amplicor HIV-1 Monitor Test was therefore excluded
from the overall means for the candidate standard. Therefore, the
Fig. 2. The histogram of estimated values of HIV-1 RNA (Log10 IU/mL) for candidate standard from 10 laboratories. The number labeled in the box represented the laboratory
code number.
446
Annex 6
A 4. Discussion
candidate 24°C
8 There was a clear variation between the results from earlier HIV-
candidate original value
7
6 1 viral load assays such as nucleic acid signal branch amplification
Log IU/mL
5 (NASBA), PCR end point detection and branched DNA (bDNA) sig-
4 nal amplification compared to more recent tests such as the Abbott
3
2 real-time assay. The limitation of these assays has previously been
1 reported (Church et al., 2011) and it is known that are optimized to
0 target subtype B group M viruses. New-generation real-time PCR
0 1 2 3 4 5 6 7 8 9
assays for HIV-1 RNA quantification include the Abbott RT HIV-1
Time (weeks)
assay and the Cobas Ampli-Prep/Cobas TaqMan HIV-1 assay (CAP-
B
CTM). These real-time PCR assays have been improved and are
candidate 4°C
8 candidate original value
now able to detect HIV-1 group M, non-B subtype viruses, group
7 N viruses and O viruses. In addition, the Abbott RealTime HIV-1
Log IU/mL
6
5 assay has been reported to successfully detect HIV-1 group P infec-
4 tion (Plantier et al., 2009). However, it has also been reported that
3
2 multiple mismatches in gag primers and probe binding regions for
1 the first version of the CTM assay (CTM1) exist, which can result in
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 an underestimation of the CTM1 values for some patients infected
Time (weeks) with HIV-1 group M, non-B subtypes. To overcome this problem,
Roche Diagnostics have already upgraded their test to CTM version
C
2.0 (CTM2), which uses a dual-target strategy (Damond et al., 2010;
candidate -20°C
Church et al., 2011; Wirden et al., 2011).
8 candidate original value
7 In this international collaborative study, most of the partici-
Log IU/mL
6 pants performed quantitative assays: four laboratories used the

5
4 Abbott RT HIV-1; six laboratories used the COBAS TaqMan HIV-1
3
2
Test (COBAS AmpliPrep/COBAS TaqMan HIV-1 Test, v1.0 & 2.0, and
1 COBAS TaqMan HIV-1 Test, for use with High Pure System Viral
0
0 2 4 6 8 10 12 14 16 18 20 22 24
Nucleic Acid Kit, v1.0); two laboratories used the VERSANT HIV-1
Time (months) RNA 3.0; and one laboratory used the COBAS Amplicor HIV-1 Mon-
itor Test, v1.5. The results showed that all the laboratories were in
D good agreement with the estimates, except one laboratory submit-
candidate -80°C ted an outlying result, which is generated from COBAS Amplicor
8 candidate original value HIV-1 Monitor Test, v1.5. By comparing the results from different
7 commercial kits, indicated that the COBAS Amplicor HIV-1 Monitor
Log IU/mL
6
5 Test, v1.5 was significantly different from other commercial assay
4 kits. The data generated from the COBAS Amplicor HIV-1 Monitor
3
2 Test was therefore excluded from the overall means for the candi-
1
0 date sample. Since there is only one assay result available in this
0 2 4 6 8 10 12 14 16 18 20 22 24 collaborative study, it does not represent the performance of the
Time (months) kit. Interestingly, a similar result was also shown in an earlier col-
laborative study to establish a replacement International Standard
Fig. 3. Stability analysis of HIV-1 RNA present in candidate standard after storage at
for HIV-1 RNA nucleic acid assays (Davis et al., 2008).
different temperatures: (A) +24 ◦ C for 8 weeks, (B) +4 ◦ C for 12 weeks, (C) −20 ◦ C for
24 months and (D) −80 ◦ C for 24 months. The solid line and the dotted line represent All data points received from laboratories were within a range
the tested values and the original values before storage, respectively. of 1.0 Log from this collaborative study, furthermore, most of the
data were within in a range from 3.9 Log IU/mL to 4.1 Log IU/mL. In
conclusion, a high level of agreement among the results obtained
overall mean for the candidate standard is 1.0 × 104 IU/mL, and the from the participating laboratories was observed. The first National
95% confidence intervals is 8.26 × 103 to 1.25 × 104 IU/mL (Table 4). Taiwan Standard for HIV-1 RNA NAT assays, with an assigned value
of 1.0 × 104 IU/mL, was recognized. In order to reflect the pre-
3.3. Stability analysis of the candidate standard dominant HIV-1 subtype found in Taiwan, this National Standard
was formulated from a subtype B plasma. The results of the sta-
Several vials of the candidate standard were stored at +24 ◦ C, bility study indicated that the HIV-1 RNA National Standard is
+4 ◦ C, −20 ◦ C, and −80 ◦ C, three vials were randomly selected for stable long-term when stored at −20 ◦ C and -80 ◦ C. Therefore, the
stability tests. Samples were taken after one-, two- or four-week first National Standard for HIV-1 RNA NAT assays in Taiwan was
intervals from candidate samples stored at +24 ◦ C and +4 ◦ C and established. This standard could be used for quality control of HIV-
after three- or six-month intervals from candidate samples stored 1 RNA assays and as a quantitative reference material for HIV-1
at −20 ◦ C and −80 ◦ C. Triplicate samples were assayed for each NAT assays. Moreover, the standard could be used nationally for
time point at different temperatures in three independent tests. pre-market approval testing and the performance evaluation in
The calculated mean concentration (IU/mL) for each time point post-marketing surveillances of NAT-based IVDs and facilitating to
and temperature is shown in Fig. 3. The results indicate that the ensure the continued fitness for purpose of the IVD, either imported
candidate samples were stable after storage at +24 ◦ C for 4 weeks, or domestic.
at +4 ◦ C for 8 weeks, at −20 ◦ C for 24 months, and at −80 ◦ C for In recent years, subtype CRF 07 BC has been the major group
24 months. The results suggest good long term stability for the of HIV-1 found in the intravenous drug abuser population in
proposed national candidate standard when stored at −20 ◦ C and Taiwan. As HIV-1 strain diversity and viral recombination events
−80 ◦ C. increase, the need for surveillance using commercial assays to
447
126 Y.-C. Yang et al. / Journal of Virological Methods 191 (2013) 122–127
Table 3
Comparison of different commercial kits.
Assay method Lab code Result Mean SD CV (%)
Abbott RT HIV-1 1B 3.98 4.05 0.10 2.52

2A 4.13
3 4.15
8 3.95
COBAS AmpliPrep/COBAS TaqMan 1A 4.04 4.01 0.04 1.06

HIV-1 test, v1.0 6 3.98
COBAS AmpliPrep/COBAS TaqMan 2C 4.13 4.08 0.07 1.73

HIV-1 test, v2.0 5 4.03
COBAS TaqMan HIV-1 test, for use with High 2B 4.11 3.90 0.30 7.81
Pure System Viral Nucleic Acid Kit, v1.0 10 3.68
a a
COBAS Amplicor HIV-1 Monitor Test, v1.5 9 4.42 4.42
VERSANT HIV-1 RNA 4 3.73 3.96 0.32 8.05

3.0 (b DNA) 7 4.18
a
p < 0.05. Please note that a single assay result does not represent the performance of the kit.
Table 4
Overall mean estimates of HIV-1 RNA (Log IU/mL) for candidate standard.
Sample Mean 95% confidence interval (95% CI)
Log IU/mL IU/mL Log IU/mL IU/mL
Candidate standard 4.01 1.01E+04 3.92–4.10 8.26E+03–1.25E+04
ensure detection ability and viral load monitoring accuracy in HIV- - Lena Panagiotopoulos/Stirling Dick, National Serology Reference
1-infected patients has increased. To fulfill this goal, an HIV-1 CRF Laboratory (NRL), Australia.
07 BC National standard will need to be established as the next - Micha Nübling/Michael Chudy, Paul-Ehrlich-Institute, Germany.
major step. - Yi-Li Shih, E-Da Hospital, Taiwan.
Disclaimer References
The findings and conclusions in this article have not been for- Bolinger, C., Boris-Lawrie, K., 2009. Mechanisms employed by retroviruses to exploit
mally disseminated by Taiwan FDA and should not be construed to host factors for translational control of a complicated proteome. Retrovirology
6, 8.
represent any agency determination or policy. Busch, M.P., Dodd, R.Y., 2000. NAT and blood safety: what is the paradigm? Trans-
fusion 40, 1157–1160.
Church, D., Gregson, D., Lloyd, T., Klein, M., Beckthold, B., Laupland, K., Gill, M.J., 2011.
Acknowledgements Comparison of the RealTime HIV-1, COBAS TaqMan 48 v1.0, Easy Q v1.2, and
Versant v3.0 assays for determination of HIV-1 viral loads in a cohort of Canadian
patients with diverse HIV subtype infections. J. Clin. Microbiol. 49, 118–124.
We would like to express our sincere thanks to all the par- Damond, F., Avettand-Fenoel, V., Collin, G., Roquebert, B., Plantier, J.C., Ganon, A.,
ticipants of the collaborative study group (Appendix A) for their Sizmann, D., Babiel, R., Glaubitz, J., Chaix, M.L., Brun-Vezinet, F., Descamps,
valuable contribution to this collaborative study and to Ai-Zhen D., Rouzioux, C., 2010. Evaluation of an upgraded version of the Roche Cobas
AmpliPrep/Cobas TaqMan HIV-1 test for HIV-1 load quantification. J. Clin. Micro-
Yang and Yu-Chen Pao for their technical assistance for formulat-
biol. 48, 1413–1416.
ing the candidate standard. We would like to thank Clare Morris Davis, C., Heath, A., Best, S., Hewlett, I., Lelie, N., Schuurman, R., Holmes, H., 2003.
for providing language help. In addition, we also acknowledge the Calibration of HIV-1 working reagents for nucleic acid amplification techniques
funding support from Taiwan FDA. against the 1st international standard for HIV-1 RNA. J. Virol. Methods 107,
37–44.
Davis, C., Berry, N., Heath, A., Holmes, H., 2008. An international collaborative study
to establish a replacement World Health Organization International Standard
Appendix A. for human immunodeficiency virus 1 RNA nucleic acid assays. Vox Sang. 95,
218–225.
List of participants: Holmes, H., Davis, C., Heath, A., Hewlett, I., Lelie, N., 2001. An international collabo-
rative study to establish the 1st international standard for HIV-1 RNA for use in
nucleic acid-based techniques. J. Virol. Methods 92, 141–150.
- Brian Erickson, Abbott Molecular Inc., USA. Karlsson Hedestam, G.B., Fouchier, R.A., Phogat, S., Burton, D.R., Sodroski, J., Wyatt,
R.T., 2008. The challenges of eliciting neutralizing antibodies to HIV-1 and to
- Chien-Feng Sun/Shuan Yang, Chang-Gung Memorial Hospital, influenza virus. Nat. Rev. Microbiol. 6, 143–155.
Taiwan. Lee, S., Wood, O., Taffs, R.E., Hu, J., Machuca, A., Vallejo, A., Hewlett, I., 2006. Devel-
- Clare Morris, National Institute for Biological Standards and Con- opment and evaluation of HIV-1 subtype RNA panels for the standardization of
HIV-1 NAT assays. J. Virol. Methods 137, 287–291.
trols (NIBSC), UK.
Lin, Y.T., Lan, Y.C., Chen, Y.J., Huang, Y.H., Lee, C.M., Liu, T.T., Wong, W.W., Yang,
- Chih-Yuan Yang/Cheng-Feng Kao, Centers for Disease Control J.Y., Wang, C.T., Chen, Y.M., 2007. Molecular epidemiology of HIV-1 infection
(TCDC), Taiwan. and full-length genomic analysis of circulating recombinant form 07 BC strains
- Der-Yuan Wang/Yi-Chen Yang, Food and Drug Administration from injection drug users in Taiwan. J. Infect. Dis. 195, 1283–1293.
Murthy, K.K., Henrard, D.R., Eichberg, J.W., Cobb, K.E., Busch, M.P., Allain, J.P.,
(TFDA), Taiwan. Alter, H.J., 1999. Redefining the HIV-infectious window period in the chim-
- Eric T. Natoli, Siemens Healthcare Diagnostics, USA. panzee model: evidence to suggest that viral nucleic acid testing can prevent
- Indira Hewlett/Sherwin Lee, Center for Biologics Evaluation and blood–borne transmission. Transfusion 39, 688–693.
Piatak Jr., M., Saag, M.S., Yang, L.C., Clark, S.J., Kappes, J.C., Luk, K.C., Hahn, B.H.,
Research/Food and Drug Administration (CBER/FDA), USA. Shaw, G.M., Lifson, J.D., 1993. High levels of HIV-1 in plasma during all stages
- John Saldanha/Matthew Lin, Roche Molecular Systems, USA. of infection determined by competitive PCR. Science 259, 1749–1754.
448
Annex 6
Pisani, G., Marino, F., Cristiano, K., Bisso, G.M., Mele, C., Luciani, F., Wirz, M., Gentili, Simon, F., Mauclere, P., Roques, P., Loussert-Ajaka, I., Muller-Trutwin, M.C., Saragosti,
G., 2007. Collaborative study for the calibration of HCV RNA, HBV DNA and HIV S., Georges-Courbot, M.C., Barre-Sinoussi, F., Brun-Vezinet, F., 1998. Identifica-
RNA reference preparations against the relative international standards. Ann. tion of a new human immunodeficiency virus type 1 distinct from group M and
Ist. Super. Sanita 43, 69–76. group O. Nat. Med. 4, 1032–1037.
Plantier, J.C., Leoz, M., Dickerson, J.E., De Oliveira, F., Cordonnier, F., Lemee, V., Spira, S., Wainberg, M.A., Loemba, H., Turner, D., Brenner, B.G., 2003. Impact of clade
Damond, F., Robertson, D.L., Simon, F., 2009. A new human immunodeficiency diversity on HIV-1 virulence, antiretroviral drug sensitivity and drug resistance.
virus derived from gorillas. Nat. Med. 15, 871–872. J. Antimicrob. Chemother. 51, 229–240.
Pluta, K., Kacprzak, M.M., 2009. Use of HIV as a gene transfer vector. Acta Biochim. Statistics of HIV/AIDS 12/31/2012. Center for Disease Control, ROC (Taiwan). 2012,
Pol. 56, 531–595. 1.
Revets, H., Marissens, D., de Wit, S., Lacor, P., Clumeck, N., Lauwers, S., Zissis, G., UNAIDS report on the global AIDS epidemic 2012. Joint United Nations Programme
1996. Comparative evaluation of NASBA HIV-1 RNA QT, AMPLICOR-HIV monitor, on HIV/AIDS, pp. 8–14.
and QUANTIPLEX HIV RNA assay, three methods for quantification of human Weiss, R.A., 1993. How does HIV cause AIDS? Science 260, 1273–1279.
immunodeficiency virus type 1 RNA in plasma. J. Clin. Microbiol. 34, 1058–1064. Wirden, M., Tubiana, R., Fourati, S., Thevenin, M., Simon, A., Canestri, A., Ait-
Robertson, D.L., Anderson, J.P., Bradac, J.A., Carr, J.K., Foley, B., Funkhouser, R.K., Gao, Arkoub, Z., Soulie, C., Marcelin, A.G., Katlama, C., Calvez, V., 2011. Upgraded
F., Hahn, B.H., Kalish, M.L., Kuiken, C., Learn, G.H., Leitner, T., McCutchan, F., Cobas Ampliprep-Cobas TaqMan version 2.0 HIV-1 RNA quantification assay
Osmanov, S., Peeters, M., Pieniazek, D., Salminen, M., Sharp, P.M., Wolinsky, S., versus first version: correction of underestimations. J. Clin. Microbiol. 49,
Korber, B., 2000. HIV-1 nomenclature proposal. Science 288, 55–56. 2700–2702.
Schuurman, R., Descamps, D., Weverling, G.J., Kaye, S., Tijnagel, J., Williams, I., Yang, Y.C., Chang, C.J., Tseng, C.I., Wang, D.Y., Cheng, H.F., Lin, C.P., 2008. Establish-
van Leeuwen, R., Tedder, R., Boucher, C.A., Brun-Vezinet, F., Loveday, C., 1996. ment of a National Taiwan Standard and working reagent for human parvovirus
Multicenter comparison of three commercial methods for quantification of B19 DNA nucleic acid amplification techniques (NAT). J. Food Drug Anal. 16,
human immunodeficiency virus type 1 RNA in plasma. J. Clin. Microbiol. 34, 9–14.
3016–3022.
449
Appendix 3
Example of the calibration of a reference preparation by a
single NAT assay
The reference preparation (RP) is calibrated against the WHO IS by testing
dilutions of both samples using a real-time PCR, collecting raw data (threshold
cycle values Ct) and performing a valid statistical analysis (e.g. parallel line).
Study performed under the supervision of G. Pisani, ISS, Rome, Italy.
Study samples:
■■ RP for HIV RNA with a presumptive titre of 15 000 IU/mL of
HIV RNA
■■ 3rd WHO IS HIV RNA batch 10/152 with a concentration of
185 000 IU/mL (5.26 log IU/mL).
Test the following dilutions of WHO IS and RP (in triplicate) on three separate
days. Collect the raw data.
Day 1
Sample Final dilution Concentration Raw data (Ct value)

(log)
Replica 1 Replica 2 Replica 3
WHO IS −1.09 15 000 IU/mL 28.5 28.5 28.3
−1.59 4 700 IU/mL 29.8 30.0 30.0
−2.09 1 500 IU/mL 31.6 31.5 31.6

−2.59 470 IU/mL 32.9 33.4 33.5
RP Not diluted – 28.4 28.9 28.7
−0.50 – 30.9 30.3 30.5
−1.00 – 32.1 31.9 32.5
−1.50 – 33.6 34.8 34.1
450
Annex 6
Day 2

(log)
WHO IS −1.09 15 000 IU/mL 28.7 28.4 28.4
−1.59 4 700 IU/mL 29.7 29.9 30.0
−2.09 1 500 IU/mL 31.5 31.4 31.7
−2.59 470 IU/mL 33.3 33.2 33.4
RP Not diluted – 29.1 28.6 29.1
−0.50 – 30.1 30.8 31.3
−1.00 – 32.9 32.6 31.5
−1.50 – 34.5 33.2 33.3
Day 3
(log)
WHO IS −1.09 15.000 IU/mL 28.6 28.3 28.2
−1.59 4.700 IU/mL 29.1 29.3 30.1
−2.09 1.500 IU/mL 31.1 31.2 31.5
−2.59 470 IU/mL 33.4 33.3 32.5
RP Not diluted – 28.8 28.9 28.7
−0.50 – 30.1 30.5 30.2
−1.00 – 32.1 32.2 31.9
−1.50 – 33.5 33.2 34.3
Perform statistical analysis: parallel-line assay.

Acceptance criteria: linearity and parallelism should be fulfilled.
It is possible that on each day (each experiment) one replicate or one dose may be
deleted in order to fulfil the acceptance criteria.
451
Example calibration: Combistat (EDQM).

452
Annex 6
453
454
Annex 6
455
Annex 7
Guidelines on regulatory preparedness for provision of
marketing authorization of human pandemic influenza
vaccines in non-vaccine-producing countries
1. Introduction 460
3. Terminology 462
4. General considerations for regulatory preparedness for pandemic
influenza vaccines 463
4.1 Acknowledgement of the role of the NRA in the national pandemic influenza
preparedness plan 464
4.2 Considerations for national regulatory preparedness 464
4.3 Reliance on the decisions and expertise of other regulatory authorities 466
4.4 Seasonal influenza vaccines and pandemic preparedness influenza vaccines 467
5. Regulatory evaluation processes 468
5.1 Expected basic documentation according to the source of pandemic
influenza vaccine 470
5.2 Possible regulatory review processes in a pandemic emergency 471
5.3 WHO collaborative procedure for prequalified vaccines 474
5.4 Final evaluation 475
5.5 Emergency approval 476
5.6 Post-marketing risk management and surveillance 476
6. Quality control preparedness 477
References 479
Appendix 1 Checklist of regulatory actions for pandemic influenza preparedness
and response 482
Appendix 2 Examples of information and documentation that may be required for
the evaluation of a seasonal influenza vaccine annual virus strain change 484
457

458
Annex 7
Abbreviations
CTD Common Technical Document
NCL national control laboratory
PIP Pandemic Influenza Preparedness (Framework)
459
1. Introduction
An influenza pandemic occurs when a novel influenza A virus emerges against
which most people do not have immunity, and spreads rapidly around the world.
A pandemic influenza A virus is significantly different from normally circulating
human influenza A viruses, with a widespread absence of immunity against the
virus observed in the population. As with seasonal influenza viruses, pandemic
influenza viruses have the ability to spread easily from human to human and
cause disease. This may result in several simultaneous epidemics worldwide with
high numbers of cases of clinical disease and deaths, leading to considerable
social disruption. Pandemic influenza viruses may evolve from subtypes that
previously only circulated in animals or from subtypes currently circulating
in humans but sufficiently different antigenically for pre-existing immunity in
the population to be low or minimal (an example of the latter case is the 2009
H1N1 influenza pandemic). Influenza viruses that have caused past pandemics
have typically originated from animals. Owing to the urgent public health need,
strategies to shorten the time between the emergence of a human pandemic
influenza virus and the availability of safe and effective pandemic influenza
vaccines are one of the highest priorities in global health security.
The WHO Guidelines on regulatory preparedness for human pandemic
influenza vaccines (1) were adopted by the WHO Expert Committee on
Biological Standardization in 2007. These Guidelines provide national regulatory
authorities (NRAs) and vaccine manufacturers with:
■■ guidance regarding regulatory pathways for approving pandemic
influenza vaccines;
■■ the regulatory considerations to take into account when evaluating
the quality, safety and efficacy of candidate vaccines;
■■ guidance on effective post-marketing surveillance of pandemic
influenza vaccines.
The Guidelines apply mainly to countries where influenza vaccine
production takes place, but also contain much information that can be useful
for countries in which vaccines are not produced (hereafter referred to as
non-vaccine-producing countries). However, consultations with stakeholders
following the 2009 H1N1 influenza pandemic identified lack of regulatory
preparedness as one of the factors that delayed or prevented the deployment
of pandemic influenza vaccine in non-vaccine-producing countries. This was
especially the case for vaccine destined for donation or deployed by United
Nations agencies in response to the pandemic emergency (2–4).
The present Guidelines were developed in response to requests
from non-vaccine-producing countries for guidance on the identification of
460
Annex 7
appropriate regulatory approaches to the marketing authorization of pandemic

influenza vaccines, and on arrangements for the lot release of these vaccines
in public health emergency conditions. The Guidelines were developed in the
context of the Pandemic Influenza Preparedness (PIP) Framework’s Partnership
Contribution Implementation Plan 2013–2016 for regulatory capacity-building
and strengthening of pandemic preparedness and response (5).

These WHO Guidelines provide guidance to NRAs of non-vaccine-producing
countries on the regulatory oversight of pandemic influenza vaccines for use
in public health emergencies. The document focuses in particular on the needs
of countries that are not producing influenza vaccines, including countries
supplied with vaccines through United Nations agencies and countries which
self-procure vaccines.
This guidance is aimed to aid such countries in preparing and putting
in place a regulatory process for pandemic influenza vaccines in advance of a
pandemic influenza emergency. Such a process should enable countries to
expedite the provision of marketing authorization and lot release of influenza
vaccines in response to a pandemic emergency. It is acknowledged that
each country will have national legislation and policies on the regulation of
medicines, vaccines and other health products. Some countries may also have
regulations in place on accepting donations of vaccines and ancillary products.
This document is intended to provide additional and specific guidance to
the NRAs of non-vaccine-producing countries when dealing with pandemic
influenza emergencies.
The document specifically provides NRAs of non-vaccine-producing
countries with the general principles for evaluating influenza vaccines and
establishing basic emergency procedures for regulating pandemic influenza
vaccines. A strong emphasis is placed on the need to prepare decision-making
processes which minimize duplication and make much-needed vaccines available
for use without unnecessary delay during pandemic emergencies. The need to
establish such appropriate regulatory processes during the interpandemic phase
is also emphasized.
These WHO Guidelines apply to all pandemic influenza vaccines. They
are intended for use by NRAs, but will also be of interest to national immunization
technical advisory groups (NITAGs), as well as manufacturers and authorities
in the private and public sectors responsible for planning and managing vaccine
deployment and vaccination operations at all levels.
Other relevant WHO guidelines should also be consulted as appropriate.
461
3. Terminology
Alert phase: the phase during which influenza caused by a new strain
is identified in humans. Increased vigilance and careful risk assessment at local,
national and global levels are characteristic of this phase (6).
Influenza pandemic: an influenza pandemic (or global epidemic) occurs
when a novel influenza virus strain appears which is significantly different from
circulating strains and against which almost no one is immune. The Director-
General of WHO may, as appropriate, declare a public health emergency of
international concern under the International Health Regulations (2005) following
the identification and determination of global spread of human influenza caused
by a new virus strain (6, 7).
Interpandemic phase: the period between influenza pandemics (6).
Marketing authorization: a formal authorization for a medicine to be
marketed. Once an NRA approves a marketing authorization application for a
new medicine, the medicine may be marketed and made available for physicians
to prescribe. Also referred to as “licensing” or “registration” in this and other
documents (8).
National pandemic influenza preparedness plan: a national plan that
aims to set out country-specific priorities and actions, and to identify the major
components that must be put in place (for example, coordination, resource
identification and allocation, and capacity-building) along with response actions
that should be strengthened to respond to a pandemic (9).
Non-vaccine-producing country: a country in which vaccines are not
produced.
Pandemic influenza vaccine: a monovalent vaccine containing the
human influenza A virus strain recommended by WHO for use either when a
pandemic is considered by WHO to be imminent or during a pandemic (1).
Pandemic phase: the period of global spread of human influenza
caused by a new virus strain. Progression from the interpandemic to the alert
and pandemic phases may occur quickly or gradually, as indicated by the global
risk assessment, principally based on virological, epidemiological and clinical
data (6).
Pandemic preparedness influenza vaccine: an influenza vaccine
developed and tested in anticipation of an influenza pandemic, and manufactured
using an influenza virus strain that is believed to have similar characteristics
to a potential pandemic virus strain (also referred to as “mock-up pandemic
influenza vaccine” or “vaccine against novel human influenza virus” in other
documents) (1, 10, 11).
462
Annex 7
Risk-management plan: a document submitted as part of the marketing

authorization dossier that is evaluated by regulatory authorities before a
medicine can be authorized and which is regularly updated as new information
becomes available. Risk-management plans include information on a medicine’s
safety profile and explain the measures that are taken in order to prevent or
minimize any risks associated with the use of the medicine in patients.
Seasonal influenza vaccine: a trivalent (or tetravalent) vaccine containing
the two influenza A virus strains and one (or two) influenza B virus strains
recommended by WHO at its biannual influenza vaccine composition meetings
(once for the northern hemisphere and once for the southern hemisphere) (1).
Supporting NRA: an NRA selected by the NRA of a receiving country
as suitable to support licensing decisions for pandemic influenza vaccines. The
eligibility of supporting NRAs could be decided upon after consultation with
WHO for guidance.
Transition phase: the phase during which the de-escalation of global
actions occurs as the assessed global risk of influenza reduces; a corresponding
reduction of response activities or movement towards recovery actions by
countries may be appropriate, according to their own risk assessments (6).
4. General considerations for regulatory

preparedness for pandemic influenza vaccines
Countries should have laws requiring that all medicinal products, including
influenza vaccines procured or donated in normal or emergency circumstances,
be licensed before being placed on the market.
All countries should prepare for public health emergency situations,
including influenza pandemics that may cause high morbidity and mortality
leading to considerable social disruption. In 2013, WHO revised and updated
its pandemic preparedness guidance to reflect experience gained from the
2009 H1N1 influenza pandemic and to support further efforts at national and
subnational levels. The updated guidance (6) provides for a risk-based approach
that: (a) enables a more flexible response to different scenarios; (b) emphasizes
reliance on multisectoral participation; and (c) uses a simplified pandemic
phase structure that includes the interpandemic and pandemic (alert and
transition) phases.
Regulatory preparations for an influenza pandemic should also be
undertaken in the interpandemic phase (6) in order to strengthen the legal and
regulatory requirements for importing and approving a vaccine in emergency
situations. This would include improving NRA capacity and clearly defining the
regulatory pathways for licensing the use of a new vaccine under emergency
conditions (12).
463
NRAs should review the options available to them during a public

health emergency and choose the appropriate procedures to fit the situation.
The emergency procedures should include processes for ensuring information
management, and effective communication and cooperation between different
branches of the NRA and relevant stakeholders such as public health authorities
(9, 13).
Plans should be developed to address the need for official communication
from the NRA relevant to specific audiences – such as the public, health-care
workers, national and subnational authorities and international collaborators
when needed. Principles set out in relevant WHO communication guidelines
(14, 15) should be followed. Communication and information-sharing systems
should be established and need to be implemented for all stakeholders (12).
NRAs together with the national immunization programme and other
stakeholders should develop post-marketing surveillance plans (including
consideration of a risk-management plan which is part of marketing authorization)
to monitor the safety and efficacy of pandemic influenza vaccines used during a
pandemic. For guidance on safety monitoring and post-marketing surveillance
plans, NRAs should refer to the WHO Guidelines on regulatory preparedness
for human pandemic influenza vaccines (1) and the WHO Global manual on
surveillance of adverse events following immunization (16).
4.1 Acknowledgement of the role of the NRA in the

national pandemic influenza preparedness plan
The national pandemic influenza preparedness plan should be established and
endorsed before a pandemic arises and should include acknowledgement of
the roles and responsibilities of the NRA in regulatory oversight of vaccines (9,
13, 17). The majority of WHO Member States developed and published their
national pandemic influenza preparedness plans in 2005 and 2006 and updated
them after the 2009 H1N1 influenza pandemic (9).
4.2 Considerations for national regulatory preparedness

During the interpandemic phase the NRA should be responsible for developing
the following procedures and plans to support the national pandemic influenza
preparedness plan and vaccine deployment plan (12):
■■ suitable regulatory pathways for pandemic influenza vaccines
during the emergency;
■■ appropriate vaccine lot release procedures for emergency use;
■■ post-marketing safety surveillance plans.
464
Annex 7
It is recommended that the NRA’s preparedness procedures for

facilitating the rapid availability of pandemic influenza vaccines should include:
■■ an NRA contact point for communications with WHO and other
stakeholders on public health/regulatory issues;
■■ allocation of resources to be used when a pandemic alert has been
declared by WHO (note that the national declaration of a pandemic
emergency would be made by the responsible national authority
following the declaration by WHO);
■■ a public risk-communications plan summarizing the basis for
decision-making;
■■ procedures for the timely appointment of an emergency evaluation
task team for pandemic influenza vaccines (and medicines) that will:
(a) include appropriate regulatory and programmatic expertise;
(b) prepare procedures for evaluation of applications for pandemic
influenza vaccine;
(c) define the dossier and supporting documents needed for NRA
evaluation;
(d) evaluate and recommend marketing authorization of suitable
vaccines to the NRA; and
(e) allow, during the interpandemic phase, for the regular review of
task team appointments and procedures;
■■ procedures for interactions (including discussion of options for
appropriate sources of vaccine) with the public health agencies that
will procure, deploy and administer the vaccines;
■■ a system to accelerate the licensure and lot release of pandemic
influenza vaccine including recognition of the decisions, or reliance
upon the expertise, of supporting NRAs, and the optimizing of
available resources in response to the pandemic;
■■ procedures and requirements for lot release of pandemic influenza
vaccines by the NRA during the pandemic phase (or emergency
situation).
The following steps should be included in the regulatory preparedness
procedures:
■■ a working procedure for marketing authorization of the seasonal
influenza vaccine annual virus strain change (this may be used
where the pandemic influenza vaccine involves a strain change from
a licensed seasonal influenza vaccine);
465
■■ preparation of a template emergency risk–benefit consideration and

assessment report;
■■ a procedure for emergency approval of the NRA recommendation,
as appropriate;
■■ a process to expedite marketing authorization through the WHO
collaborative procedure for prequalified vaccines, when appropriate;
■■ preparation of an outline post-marketing surveillance plan which
should include special provisions for post-marketing surveillance of
the pandemic influenza vaccine in use.
A checklist of regulatory actions for pandemic influenza preparedness

and response is provided in Appendix 1.
4.3 Reliance on the decisions and expertise

of other regulatory authorities
In the event of a pandemic emergency, the NRA of a non-vaccine-producing
country should consider reliance on the product evaluation decisions made by
other NRAs in vaccine-producing countries. Non-vaccine-producing countries
may select, and where possible establish links with, suitable supporting NRAs
during the interpandemic period. Reliance on the decisions or expertise of
supporting NRAs is highly encouraged.
The NRA of the non-vaccine-producing country should establish
mechanisms and procedures for recognizing the marketing authorization
decisions of the NRA of the country producing the vaccine, or of other
supporting NRAs as appropriate, when considering the licensing of a pandemic
influenza vaccine. Mechanisms and procedures may include the establishment
during the interpandemic phase of a memorandum of understanding or
recognition, including an information-sharing agreement between receiving and

selected supporting NRAs in the event of a pandemic.
The assessment reports (summary basis for decision) from other
NRAs may provide valuable information and insight into the decision-making
processes of these NRAs but may not be readily available in a public health
emergency. In this case communication with the relevant NRA regarding the
licensure is strongly encouraged.
In addition, a procedure for joint review of a pandemic influenza vaccine
dossier with neighbouring and supporting NRAs may be considered. This could
be facilitated by WHO.
The WHO collaborative procedure for marketing authorization of
prequalified vaccines (18, 19) could be used as a model.
466
Annex 7
It should be noted that both joint reviews and the WHO collaborative
procedure require advance planning so that agreements are brought into effect
at the earliest opportunity and that the vaccine product is already identified.
It is expected that future pandemic influenza vaccines prequalified
by WHO will include a summary assessment report outlining the basis for
prequalification that will be available to countries intending to import, grant
marketing authorization for and use these vaccines to mitigate an influenza
pandemic. Requests for more detailed information regarding prequalification
of a particular pandemic influenza vaccine should be addressed to the WHO
prequalification programme.
The NRAs of some vaccine-producing countries with considerable
experience in the evaluation of seasonal and pandemic influenza vaccines
supported WHO in expediting the prequalification of pandemic influenza
vaccines during the 2009 H1N1 influenza pandemic, and are encouraged to
support the NRAs of non-vaccine-producing countries in regulatory decision-
making and marketing authorization of pandemic influenza vaccines.
4.4 Seasonal influenza vaccines and pandemic

preparedness influenza vaccines
Seasonal influenza vaccines present many production and regulatory challenges
similar to those of pandemic influenza vaccines due to the need for an annual
change in formulation to reflect currently circulating virus strains, and very
short development timelines. Many countries have established accelerated
regulatory procedures for licensing seasonal influenza vaccines. Some non-
vaccine-producing countries may also have provisions in place for accelerated
regulatory approval of annual influenza virus strain changes in a seasonal vaccine
formulation. In all cases, the WHO recommendations on seasonal influenza
vaccine strain composition1 should be followed (8).
In appropriate circumstances, the NRA may decide that the procedure
for an annual seasonal vaccine strain change can be adapted to authorize
pandemic influenza vaccines. The combination of circumstances under which
the strain-change procedure can be adapted to license pandemic influenza
vaccines are:
■■ the candidate monovalent pandemic influenza vaccine has an
antigen content similar to that of the corresponding single
component in a licensed trivalent or tetravalent seasonal influenza
vaccine containing the same subtype; and
See: http://www.who.int/influenza/vaccines/virus/recommendations/en/, accessed 29 November 2016.

1
467
■■ the excipients in the candidate vaccine are the same as those in the
licensed vaccine; and
■■ the manufacturing technology (for example, eggs, inactivant,
purification process) and controls are the same as those of the
licensed vaccine.
Pandemic preparedness influenza vaccines are vaccines that have been
prepared using strains of influenza viruses that are considered to have pandemic
potential. These vaccines may be novel in formulation, antigen content and/or
adjuvant. Influenza vaccine manufacturers have been encouraged to develop
pandemic preparedness influenza vaccines and to conduct suitable nonclinical
and clinical testing to demonstrate their safety and immunogenicity.
The rationale for the decision to review pandemic preparedness influenza
vaccines should be made publicly available (10, 11).
Some countries may choose to make specific provisions for evaluating
pandemic preparedness influenza vaccine as a precautionary step so that the
strain-change policy and procedures used for seasonal influenza vaccine can be
adapted for suitable pandemic influenza vaccine applications. Once the pandemic
preparedness influenza vaccine has been evaluated and approved (although not
marketed for sale), the change to an appropriate pandemic virus strain – when
identified and formulated into a pandemic influenza vaccine – can be approved
using similar criteria to those used for an annual seasonal vaccine strain change.
This procedure may be implemented in countries with adequate regulatory
expertise and resources.
Some pandemic influenza vaccines or pandemic preparedness influenza
vaccines may be novel constructs or formulations requiring expert regulatory
evaluation. NRAs of non-vaccine-producing countries may request assistance in
such evaluations from WHO or other NRAs more experienced in the regulation
of both seasonal and pandemic influenza vaccines (see section 4.3 above).
5. Regulatory evaluation processes

The following elements are necessary to ensure an orderly and legal regulatory
marketing authorization or emergency approval and lot release of a pandemic
influenza vaccine in an emergency situation in the shortest possible time:
■■ an NRA or a regulatory system;
■■ a national pandemic preparedness plan that acknowledges that
pandemic influenza vaccines that are used shall be formally
licensed or granted emergency approval by the NRA and released
onto the market;
468
Annex 7
■■ NRA policies and procedures for:

(a) NRA evaluation of pandemic influenza vaccine applications;
(b) procedures and criteria for rapid identification of suitable
experts for regulatory evaluation of pandemic influenza vaccine
applications (task team);
(c) consideration of a joint review with neighbouring or
supporting NRAs; and
(d) recognition of the marketing authorization decisions of other
NRAs and the WHO prequalification decision;
■■ a procedure for emergency approval of the NRA’s pandemic
influenza vaccine recommendations (where higher authority
ratification is required);
■■ a collaborative procedure for expedited marketing authorization of
prequalified vaccines, when appropriate;
■■ a situation analysis of possible procedures for marketing
authorization of vaccines received through self-procurement,
donations and/or United Nations supply. The situation should also be
recognized whereby a pandemic preparedness influenza vaccine has
been evaluated and approved during the interpandemic period and
where the application can subsequently be approved for pandemic
use on the basis of the national seasonal influenza vaccine strain-
change procedure;
■■ recognition of lot release certificate of other responsible NRAs;
■■ plan for post-marketing surveillance of the pandemic influenza
vaccine in use.
Depending on the pandemic phase and the source of the vaccine,
the following regulatory approaches could be followed by an NRA (see section
5.2 below):
■■ Full review – a standard review process to authorize a product
licensure that can include fast-track review.
■■ Fast-track review of basic documentation – a fast-track review
process based on basic available information for emergency
authorization.
■■ Reliance – a process to review the marketing authorization report/
decision issued by a supporting NRA or WHO prequalification (19).
■■ Recognition – recognition of the marketing authorization decision of
another NRA or WHO prequalification without further evaluation.
469
■■ Strain-change procedure: a procedure for authorizing a seasonal

strain change for influenza vaccines:
(a) a procedure for the evaluation and approval of seasonal
influenza virus strain changes in an approved seasonal influenza
vaccine (see Appendix 2);
(b) the approved procedure to be used for pandemic preparedness
influenza vaccine evaluation and marketing authorization
following inclusion of the identified pandemic virus strain.
5.1 Expected basic documentation according to

the source of pandemic influenza vaccine
Non-vaccine-producing countries can access pandemic influenza vaccine from
different sources, including a United Nations agency, a donation from a company
or other source, or through national self-procurement. In general, full dossiers are
required for evaluation of the quality, safety and efficacy of vaccines – however,
in an emergency situation the accompanying documentation dossier may be
provided in sections as it becomes available.
Under these circumstances, at least the following documents should
be made available for evaluation to ensure the quality, safety and efficacy of
vaccines from each source:
United Nations agency supply (WHO-prequalified vaccines)
■■ Evidence/certificate of WHO prequalification with assessment
report (18, 19).
Donation from a company or other source
■■ Information on strain change of a licensed seasonal influenza
vaccine or pandemic preparedness influenza vaccine (if applicable).

■■ If the vaccine has been prequalified by WHO the Common
Technical Document (CTD) Module-2 and prequalification
assessment report should be provided.
■■ If the vaccine has been licensed by a supporting NRA the CTD
Module-2 and assessment report by the NRA, if available, should
be provided.
■■ Where the vaccine has been licensed by an NRA other than a
supporting NRA the full dossier for marketing authorization and
the assessment report by the NRA, if available, should be provided.
■■ In the case of a vaccine that has not previously been licensed a
full dossier for marketing authorization should be provided by
470
Annex 7
the manufacturer. The procedures and requirements in the WHO

Guidelines on regulatory preparedness for human pandemic
influenza vaccines should be followed (1).
National guidelines on donations of medicines should be followed. If
these do not exist the recommendations in the WHO Guidelines for medicine
donations (20) should be followed.
National self-procurement
■■ Information on strain change of a licensed seasonal influenza
vaccine or pandemic preparedness influenza vaccine (if applicable)
should be provided.
■■ If the vaccine has been prequalified by WHO the CTD Module-2
and prequalification assessment report should be provided.
■■ If the vaccine has been licensed by a supporting NRA the CTD
Module-2 and assessment report by the NRA, if available, should
be provided.
■■ Where the vaccine has been licensed by an NRA other than a
supporting NRA the full dossier for marketing authorization and
the assessment report by the NRA, if available, should be provided.
■■ In the case of a vaccine that has not previously been licensed a
full dossier for marketing authorization should be provided by
the manufacturer. The procedures and requirements in the WHO
Guidelines on regulatory preparedness for human pandemic influenza
vaccines should be followed (1). Seeking support from the NRA of
the producing country is strongly encouraged.
5.2 Possible regulatory review processes

in a pandemic emergency
Even in the midst of a pandemic emergency the NRA should conduct an
appropriate review of the documentation submitted that covers the components
set out below, and should document the extent of the available evidence on which
the recommendation to authorize, approve or reject had been based.
In a pandemic emergency it is possible that not all documentation for a
vaccine will be available at the time of application, and many NRAs have accepted
that applicants will submit the evidence as it becomes available. This approach is
generally known as a “rolling review” (21). It would be expected that the sections
on manufacturing, specifications and controls would be available, together
with evidence of consistency of manufacture. For nonclinical safety studies,
preliminary results should be available. The results of stability studies would be
delayed as would any results from clinical studies.
471
Where possible the NRA could make arrangements for the joint review
of pandemic influenza vaccine dossiers with neighbouring and/or supporting
NRAs. The possible parties involved in such an arrangement should establish
this agreement during the interpandemic phase.
Depending on the pandemic phase and the source of the vaccine,
review activities may include one or more of the following procedures (see also
Fig A7.1):
■■ full review;
■■ fast-track review of basic documentation;
■■ reliance;
■■ recognition;
■■ strain-change procedure.
5.2.1 Full review

This is the standard process of review of the full dossier in a fast-track review
process (as normally conducted in that country) for vaccines that are new
applications or previously licensed by NRAs other than a supporting NRA.
■■ Available documentation: the documentation should be complete, as
legally required in each country.
Applicability: this procedure would apply to licensed vaccines in the
interpandemic phase.
This would require evaluation of the documentation of product quality
and of the results of nonclinical and clinical studies to demonstrate safety and
efficacy in the target population. The documentation should be as legally required
in each country.
During the interpandemic phase the NRA of a non-vaccine-producing
country may conduct a full pandemic preparedness influenza vaccine dossier

review to ensure familiarity with the characteristics of such vaccines.
5.2.2 Fast-track review of basic documentation

This is a fast-track review process in which marketing authorization is based
upon the information available at the time. In the event that a fast-track review
is deemed appropriate (as defined in the approved NRA pandemic emergency
procedures) the following documents from the manufacturer and the responsible
NRA and/or WHO should be reviewed. The full application dossier may be
provided when available.
■■ Available documentation:
a) assessment reports of the responsible NRA;
472
Annex 7
b) evidence of quality (certificate of analysis or lot release) and good

manufacturing practices (GMP) compliance (GMP certificate);
c) CTD Module-2 quality, nonclinical and clinical overviews
(if available).
Applicability: this procedure would apply during the pandemic alert phase and
transition phase for a pandemic influenza vaccine licensed by an NRA other than
a supporting NRA.
5.2.3 Reliance
This is the process of reviewing the decisions of other competent NRAs with
which there has been an agreement for support. Where it has been agreed (as
defined in the approved NRA pandemic emergency procedures) that the decision
of another NRA can be considered and used as the basis of a recommendation
for marketing authorization, this approach would involve acceptance on the
basis of the already agreed conditions and limitations on the use of the vaccine,
and would require the following available documentation:
■■ certificate of the responsible NRA’s marketing authorization decision;
■■ assessment reports of the responsible NRA.
Applicability: this procedure would apply during the pandemic alert phase,
pandemic phase and transition phase for a pandemic influenza vaccine licensed
by an NRA other than a supporting NRA.
5.2.4 Recognition
This is the process of recognizing the WHO prequalification decision or
the decision of a supporting NRA. Where it has been agreed (as defined in
the approved NRA pandemic emergency procedures) that the decision of a
supporting NRA can be used as the basis for a recommendation for marketing
authorization, this approach would involve acceptance on the basis of the already
agreed conditions and limitations on the use of the vaccine, and would require
the following available documentation:
■■ certificate of the responsible NRA’s marketing authorization decision
or WHO prequalification assessment report.
Applicability: this procedure would apply during the pandemic alert phase,
pandemic phase and transition phase for a pandemic influenza vaccine licensed
by a supporting NRA or prequalified by WHO. It may also apply during the
pandemic phase for a pandemic influenza vaccine licensed by an NRA other
than a supporting NRA.
473
5.2.5 Strain-change procedure

This is a procedure for addressing a strain change in a licensed seasonal influenza
vaccine. Where it has been agreed (as defined in the approved NRA pandemic
emergency procedures) the dossier of a pandemic preparedness influenza vaccine
may be evaluated in this way, following the criteria set out for an annual strain
change, as applicable for pandemic use.
■■ Available documentation: as for an annual strain change.
The approved conditions and limitations on the use of the vaccine

should be accepted.
Applicability: this procedure would only apply to a pandemic influenza
vaccine licensed by a strain-change approach from a pandemic preparedness
influenza vaccine or seasonal influenza vaccine by the NRA of the producing
country.
If a pandemic influenza vaccine is licensed by a strain-change approach
from a pandemic preparedness influenza vaccine or seasonal influenza vaccine
then a receiving country NRA could use the strain-change procedure (or other
appropriate approach based on the source of vaccine).
If a vaccine has not been licensed by any NRA then the guidance
provided in the WHO Guidelines on regulatory preparedness for human
pandemic influenza vaccines should be followed (1).
5.3 WHO collaborative procedure for prequalified vaccines

Apart from the regulatory procedures for marketing authorization of pandemic
influenza vaccines, expedited licensure through the WHO collaborative
procedure for prequalified vaccines (18) may also be used for suitable pandemic
influenza vaccines as appropriate. An information-sharing agreement between
WHO, the receiving NRA and the manufacturer should be signed in the
interpandemic phase – particularly given that, during an emergency, time may
not allow for this step to occur prior to the decision to use the vaccine. For this
procedure the WHO prequalification assessment report should be provided
to the receiving NRA. The full dossier in the format of the CTD could also be
provided to the NRA.
It would be expected that, following the pandemic phase, the full dossier
as required by the relevant non-vaccine-producing country would be completed
and submitted for evaluation.
474
Annex 7
Fig. A7.1
Illustrative chart of regulatory approaches relative to the status of the vaccine and the
continuum of pandemic phases a
Licensed vaccines from any source
Prequalified Licensed by supporting NRA b Licensed by other NRA c
Recognition in the
pandemic phase
Recognition Recognition OR
procedure in all procedure in all Reliance or Fast-
phases phases track procedure in
the alert or transition
phases
a
Interpandemic phase, alert phase, pandemic phase or transition phase as defined by WHO (see section 3,
Terminology above).
b
Any NRA selected by the NRA of the receiving country as suitable in supporting pandemic influenza vaccine
licensing decisions; the eligibility of supporting NRAs could be decided upon after consultation with WHO.
c
Any NRA not designated as a “supporting NRA” by the NRA of the receiving country.
5.4 Final evaluation

Before a regulatory decision to recommend marketing authorization of a pandemic
influenza vaccine is taken a final evaluation of the available documentation
should be conducted to ensure that the pandemic influenza vaccine presentation
is suitable for use in the country (22, 23).
Provided the necessary procedures are in place this final evaluation can be
conducted rapidly (for example, in as little as one day depending on circumstances
and pandemic influenza vaccine marketing authorization status) with a risk–
benefit consideration and recommendation for marketing authorization.
The NRA should ensure that the following conditions are met:
■■ An adequate document package is provided. A post-marketing
commitment by the manufacturer to provide any outstanding
information should be considered.
475
■■ There is a local agency responsible for supply of the product (that is,
an “applicant” or state body that is a defined responsible legal entity).
■■ Packaging, label and package insert are nationally acceptable.
■■ The vaccine is compatible with the national pandemic influenza
preparedness plan.
■■ The vaccine is indicated for the circulating strain(s).
This evaluation may need to be based upon minimal and incomplete
documentation, and this should be acknowledged in the recommendation.
An evaluation report should be produced by the NRA.
5.5 Emergency approval

In some countries the NRA may have the authority to approve use of a medicine
or vaccine without reference to another authority, while in other countries a
final approval or directive is required. Thus reference can be made to either an
“approval” or “recommendation” process.
During the pandemic period, emergency approval procedures may
be used. Approval may be based upon limited clinical data or quality data (for
example, on stability) and upon expedited evaluation of the available evidence.
Therefore, the approval may include one or more special conditions for use.
These can include post-marketing safety reporting requirements, and limitations
such as:
■■ use only during the pandemic period

■■ use only by certain agencies
■■ use only in certain listed groups at high risk
■■ special conditions for post-marketing safety reporting.
5.6 Post-marketing risk management and surveillance

Each country should include post-marketing surveillance of adverse events in
the pandemic vaccine deployment plan. This should follow the WHO Guidelines
on regulatory preparedness for human pandemic influenza vaccines (1) and the
WHO Global manual on surveillance of adverse events following immunization
(16). The risk-management plan for pandemic influenza should be monitored by
the NRA and national immunization programme with input from the vaccine
manufacturer.
National systems for post-marketing surveillance and reporting of
adverse events following immunization should not be compromised by the
implementation of a pandemic influenza vaccination campaign.
476
Annex 7
6. Quality control preparedness

Lot release and quality control of pandemic influenza vaccines by the NRAs and/
or national control laboratories (NCLs) of non-vaccine-producing countries
should follow the guidance set out in relevant WHO documents (17, 18, 20,
22–24).
Vaccines received by procuring countries should be produced in
compliance with GMP, and tested for quality and safety by the vaccine
manufacturer. Typically, such vaccines should also be subjected to independent
quality control testing and released by the responsible NRA/NCL in accordance
with the WHO Guidelines for independent lot release of vaccines by regulatory
authorities (17). For vaccines supplied through United Nations agencies it is
recommended that further release by the NRA/NCL of receiving countries
should not be performed because such products are prequalified by WHO
and released by the responsible NRA/NCL. Likewise, self-procured WHO-
prequalified vaccines are normally released by the responsible NRA/NCL and, if
so, should not be subjected to further lot release by the importing country in the
event of an influenza pandemic. Recognition of the lot release certificate of the
responsible NRA/NCL of the producing country is recommended by WHO (17).
For self-procured non-WHO-prequalified pandemic influenza vaccines
the NRA/NCL of the procuring country may, in the event of an influenza
pandemic emergency, conduct lot release through review of the summary lot
protocol. Further laboratory testing by the NRA/NCL of the receiving country
may not be necessary, based on risk assessment. Part F of the WHO Guidelines
on regulatory preparedness for human pandemic influenza vaccines (1) should
be consulted.
The procedures adopted should ensure the deployment of vaccines
without undue delay.

Acknowledgement is due to Ms S. Ramirez for conducting an expert review
and analysis of available resources relevant to PIP Regulatory Capacity Building
Output 1 – “Develop guidelines on regulatory preparedness for non-vaccine-
producing countries that enable them to expedite approval of influenza vaccines
used in national immunization programs and/or deployed by United Nations
agencies in response to a pandemic emergency” – in January 2015.
The scientific basis for the development of these WHO Guidelines was
discussed at a working group meeting held in Tunis, Tunisia, 9–10 June 2015
and attended by: Dr M.E.M. Ahmed, National Medicines & Poisons Board,
Sudan; Mr S. Dorji, Drug Regulatory Authority of the Royal Government of
477
Bhutan, Bhutan; Mr M. Eisenhawer, WHO Regional Office for South-East

Asia, India; Dr L. Elmgren, Health Canada, Canada; Dr O.G. Engelhardt,
National Institute for Biological Standards and Control, England; Dr E. Griffiths,
Consultant, Kingston-upon-Thames, England; Mr S. Hiem, Registration Bureau
of Department of Drugs and Food, Cambodia; Mrs T. Jivapainsarnpong,
Ministry of Public Health, Thailand; Dr H. Langar, WHO Regional Office for
the Eastern Mediterranean, Egypt; Dr I.B. Mansour, Laboratoire National
Contrôle Médicaments, Tunisia; Ms M.L.L. Mendez, Comisión Federal para la
Protección contra Riesgos Sanitarios, Mexico; Ms E. Nantongo, National Drug
Authority, Uganda; Dr L. Oueslati, Laboratoire National Contrôle Médicaments,
Tunisia; Dr P. Palihawadana, Ministry of Healthcare and Indigenous Medicine,
Sri Lanka; Dr M. Pfleiderer, Paul-Ehrlich-Institut, Germany; Mr A.R.A. Rauf,
Drug Regulatory Authority of Pakistan, Pakistan; Ms J. Rodgers, Food and
Drugs Authority Ghana, Ghana; Dr S. Sebai, Laboratoire National Contrôle
Médicaments, Tunisia; Dr S.F. Shah, Consultant, WHO Regional Office for the
Western Pacific, Philippines; Dr J. Southern, Adviser to the Medicines Control
Council of South Africa, South Africa; Dr I. Tebib, Laboratoire National
Contrôle Médicaments, Tunisia; Ms E. Yonis, Food, Medicines and Health Care
Administration and Control Authority, Ethiopia; and Dr C.P. Alfonso, Ms D.
Decina, Dr R.O.A. Dehaghi, Ms L. Hedman, Dr D. Lei, Ms C.A. Rodriguez-
Hernandez and Dr T. Zhou, World Health Organization, Switzerland.
The first draft of these WHO Guidelines was prepared by Dr J. Southern,
Adviser to the Medicines Control Council of South Africa, South Africa; Dr E.
Griffiths, Consultant, Kingston-upon-Thames, England; and Dr D. Lei, World
Health Organization, Switzerland, with contributions received from the following
drafting group members: Dr L. Elmgren, Health Canada, Canada; Dr O.G.
Engelhardt, National Institute for Biological Standards and Control, England;
Mrs T. Jivapainsarnpong, Ministry of Public Health, Thailand; Ms M.L.L.
Mendez, Comisión Federal para la Protección contra Riesgos Sanitarios, Mexico;
and Dr M. Pfleiderer, Paul-Ehrlich-Institut, Germany.

The second and third drafts were prepared by Dr J. Southern, Dr E.
Griffiths and Dr D. Lei, taking into account comments made by other members
of the drafting group following a WHO informal consultation held in Geneva,
Switzerland, 6–7 April 2016 and attended by: Dr G.F.A. Ahmed, National
Medicines and Poisons Board, Sudan; Ms A. Bitegeko, Tanzania Food and Drugs
Authority, United Republic of Tanzania; Mr N. Dhakal, Ministry of Health, Nepal;
Dr M. Downham, MedImmune, England; Mr M. Eisenhawer, WHO Regional
Office for South-East Asia, India; Dr O.G. Engelhardt, National Institute for
Biological Standards and Control, England; Mrs S.S. Enyew, Food, Medicine and
Health Care Administration and Control Authority, Ethiopia; Mr D.C. Etuko,
National Drug Authority, Uganda; Dr E. Griffiths, Consultant, Kingston-upon-
Thames, England; Ms S. Hardy, Health Canada, Canada; Ms T. Jikia, Department
478
Annex 7
of Pharmaceutical Activities, Georgia; Mrs T. Jivapainsarnpong, Ministry of

Public Health, Thailand; Mr B. Khakurel, Ministry of Health, Nepal; Dr M. Lacey,
MedImmune, England; Dr J.N. Meriakol, Product Evaluation and Registration
Directorate, Kenya; Dr Obaidullah, Drug Regulatory Authority of Pakistan,
Pakistan; Dr K. Phounphenghack, Mother and Child Health Center, Lao People’s
Democratic Republic; Ms S. Ramirez, Consultant, World Health Organization,
Switzerland; Ms J. Rodgers, Food and Drugs Authority Ghana, Ghana; Ms N.S.
Sanchez, Comisión Federal para la Protección contra Riesgos Sanitarios, Mexico;
Dr J. Southern, Adviser to the Medicines Control Council of South Africa, South
Africa; Dr Y. Sun, Paul-Ehrlich-Institut, Germany; and Dr M. Alali, Dr C.P.
Alfonso, Ms D. Decina, Dr R.O.A. Dehaghi, Dr I. Knezevic, Dr D. Lei, Dr D.
Meek and Dr T. Zhou, World Health Organization, Switzerland.
The document WHO/BS/2016.2289, incorporating comments received
from regulators and industry following public consultation on the WHO
Biologicals website, was prepared by Dr J. Southern, Dr E. Griffiths and Dr D. Lei
with contributions from other drafting group members and from Dr D.
Akanmori, WHO Regional Office for Africa, Congo; Mr M. Eisenhawer, WHO
Regional Office for South-East Asia, India; Dr H. Langar, WHO Regional Office
for the Eastern Mediterranean, Egypt and Dr J. Shin, WHO Regional Office for
the Western Pacific, Philippines.
References
1. Guidelines on regulatory preparedness for human pandemic influenza vaccines. In: WHO Expert
Committee on Biological Standardization: fifty-eighth report. Geneva: World Health Organization;
2011: Annex 2 (WHO Technical Report Series, No. 963; http://www.who.int/biologicals/vaccines/
Annex_2_WHO_TRS_963-3.pdf?ua=1, accessed 28 November 2016).
2. Workshop on international regulatory capacity enhancement for influenza vaccines, 8–10 June
2011, São Paulo, Brazil. Meeting report. Geneva: World Health Organization; 2011 (http://www.
who.int/immunization_standards/national_regulatory_authorities/wirceiv_report_18jan2012.
pdf, accessed 28 November 2016).
3. Main operational lessons learnt from the WHO pandemic influenza A(H1N1) vaccine deployment
initiative. Report of a WHO meeting held in Geneva, Switzerland, 13–15 December 2010. Geneva:
World Health Organization; 2011 (http://www.who.int/influenza_vaccines_plan/resources/h1n1_
vaccine_deployment_initiaitve_moll.pdf, accessed 28 November 2016).
4. Pandemic influenza preparedness and response: a WHO guidance document. Geneva: World
Health Organization; 2009 (reprinted 2010) (http://apps.who.int/iris/bitstream/10665/44123/1/
9789241547680_eng.pdf, accessed 28 November 2016).
5. Pandemic Influenza Preparedness Framework. Partnership contribution implementation plan
2013–2016. Geneva: World Health Organization; 2013 (http://www.who.int/influenza/pip/pip_
pcimpplan_17jan2014.pdf, accessed 28 November 2016).
479
6. Pandemic Influenza Risk Management: WHO interim guidance. Geneva: World Health
Organization; 2013 (http://www.who.int/influenza/preparedness/pandemic/GIP_Pandemic
InfluenzaRiskManagementInterimGuidance_Jun2013.pdf?ua=1, accessed 28 November 2016).
7. International Health Regulations (2005). Second edition. Geneva: World Health Organization;
2008 (reprinted 2008) (http://apps.who.int/iris/bitstream/10665/43883/1/9789241580410_eng.
8. Guidelines on procedures and data requirements for changes to approved vaccines. In: WHO
Expert Committee on Biological Standardization: sixty-fifth report. Geneva: World Health
Organization; 2015: Annex 4 (WHO Technical Report Series, No. 993; http://www.who.int/
biologicals/vaccines/Annex4_Guidelines_changes_to_approved_vaccines_eng.pdf?ua=1,
accessed 28 November 2016).
9. Comparative analysis of national pandemic influenza preparedness plans. January 2011.
Geneva: World Health Organization; 2011 (http://www.who.int/influenza/resources/documents/
comparative_analysis_php_2011_en.pdf?ua=1, accessed 28 November 2016).
10. Pandemic influenza A(H1N1)v vaccines authorised via the core dossier procedure. Explanatory
note on scientific considerations regarding the licensing of pandemic A(H1N1)v vaccines. London:
European Medicines Agency; 2009 (EMEA/608259/2009 rev.; http://www.ema.europa.eu/docs/en_
GB/document_library/Medicine_QA/2009/11/WC500007567.pdf, accessed 28 November 2016).
11. Pandemic report and lessons learned. Outcome of the European Medicines Agency’s
activities during the 2009 (H1N1) flu pandemic. London: European Medicines Agency; 2011
(EMA/221017/2011; http://www.ema.europa.eu/docs/en_GB/document_library/Report/2011/04/
WC500105820.pdf, accessed 28 November 2016).
12. Guidance on development and implementation of a national deployment and vaccination plan
for pandemic influenza vaccines. Geneva: World Health Organization; 2012 (http://apps.who.int/
iris/bitstream/10665/75246/1/9789241503990_eng.pdf, accessed 28 November 2016).
13. WHO checklist for influenza pandemic preparedness planning. Geneva: World Health
Organization; 2005 (WHO/CDS/CSR/GIP/2005.4; http://www.who.int/influenza/resources/
documents/FluCheck6web.pdf?ua=1, accessed 28 November 2016).
14. WHO outbreak communication guidelines. Geneva: World Health Organization; 2005
(WHO/CDS/2005.28.) (http://apps.who.int/iris/bitstream/10665/69369/1/WHO_CDS_2005_28_
eng.pdf?ua=1, accessed 28 November 2016).
15. World Health Organization outbreak communication planning guide. 2008 edition. Geneva: World
Health Organization; 2008 (http://www.who.int/ihr/elibrary/WHOOutbreakCommsPlanngGuide.

16. Global manual on surveillance of adverse events following immunization. Geneva: World Health
Organization; 2014 (revised 2016) (http://www.who.int/vaccine_safety/publications/Global_
Manual_revised_12102015.pdf?ua=1, accessed 28 November 2016).
17. Guidelines for independent lot release of vaccines by regulatory authorities. In: WHO Expert
Committee on Biological Standardization: sixty-first report. Geneva: World Health Organization;
2013: Annex 2 (WHO Technical Report Series, No. 978; http://www.who.int/biologicals/TRS_978_
Annex_2.pdf?ua=1, accessed 28 November 2016).
18. Collaborative procedure between the World Health Organization (WHO) Prequalification Team and
national regulatory authorities in the assessment and accelerated national registration of WHO-
prequalified pharmaceutical products and vaccines. In: WHO Expert Committee on Specifications
for Pharmaceutical Preparations: fiftieth report. Geneva: World Health Organization; 2016:
Annex 8 (WHO Technical Report Series, No. 996; http://www.who.int/medicines/publications/
pharmprep/WHO_TRS_996_annex08.pdf?ua=1, accessed 28 November 2016).
480
Annex 7
19. Procedure for assessing the acceptability, in principle, of vaccines for purchase by United Nations
agencies. In: WHO Expert Committee on Biological Standardization: sixty-first report. Geneva:
World Health Organization; 2013: Annex 6 (WHO Technical Report Series, No. 978; http://www.
who.int/immunization_standards/vaccine_quality/TRS_978_61st_report_Annex_6_PQ_vaccine_
procedure.pdf?ua=1, accessed 27 March 2017).
20. Guidelines for medicine donations. Revised 2010. Third edition 2011. Geneva: World Health
Organization; 2011 (http://apps.who.int/iris/bitstream/10665/44647/1/9789241501989_eng.pdf,
21. Guidance for industry. Expedited programs for serious conditions – drugs and biologics. Silver
Spring (MD): Food and Drug Administration; 2014 (http://www.fda.gov/downloads/drugs/
guidancecomplianceregulatoryinformation/guidances/ucm358301.pdf, accessed 28 November
2016).
22. Regulation and licensing of biological products in countries with newly developing regulatory
authorities. In: WHO Expert Committee on Biological Standardization: forty-fifth report. Geneva:
who.int/biologicals/publications/trs/areas/biological_therapeutics/WHO_TRS_858_A1.pdf?ua=1,
23. Regulation of vaccines: building on existing drug regulatory authorities. Geneva: World
Health Organization; 1999 (WHO/V&B/99.10; http://apps.who.int/iris/bitstream/10665/65968/1/
WHO_V-B_99.10_eng.pdf, accessed 28 November 2016).
24. Guidelines for national authorities on quality assurance for biological products. In: WHO
Expert Committee on Biological Standardization: forty-second report. Geneva: World Health
biologicals/publications/trs/areas/biological_therapeutics/WHO_TRS_822_A2.pdf?ua=1, accessed
28 November 2016).
481
Appendix 1
Checklist of regulatory actions for pandemic influenza
preparedness and response
It is important to ensure that regulatory legislation is in place to enable the
various approaches listed below to be applied as needed in preparation for, or
during, a pandemic.
1. Prepare regulatory preparedness procedures compatible with the national
pandemic influenza preparedness plan during the interpandemic phase.
2. Appoint and maintain a pandemic task team (with staff, training, budget and
annual review).
3. In the interpandemic phase (provisionally) grant marketing authorization to
pandemic preparedness influenza vaccines.
4. Liaise with other national agencies on pandemic preparedness procedures.
5. Develop memoranda of understanding with potential supporting NRAs.
6. In the pandemic alert phase (or earlier, if possible):
(a) determine which vaccines may be sourced by national agencies;
(b) request data packages from potential vaccine suppliers;
(c) decide on appropriate evaluation procedures and evaluators;
(d) prepare a format for an assessment report and post-marketing
surveillance plan;
(e) make a recommendation for licensure or rejection that includes the

assessment report; and
(f ) alert the national control laboratory regarding potential vaccines that
may be granted marketing authorization and imported.
7. In the pandemic phase:
(a) complete the activities from the alert phase;
(b) conduct vaccine lot release procedures or, where appropriate, recognize
the lot release certificate issued by the national regulatory authority/
national control laboratory of the producing country;
(c) where possible, keep records of the vaccine lot deployment (consider
that there may be more than one vaccine approved for use);
482
Annex 7
(d) implement the national post-marketing surveillance plan;

(e) continue to update the data packages from the vaccine supplier(s); and
(f ) conduct regular reviews of activities and optimize where possible.
8. In the pandemic transition phase:
(a) complete the data package for the emergency-approved vaccine(s);
(b) collate and analyse the data from post-marketing surveillance activities;
(c) withdraw the licence of the emergency-approved pandemic influenza
vaccine(s) if appropriate;
(d) review the activities of the pandemic task team and propose
improvements; and
(e) review the reports from the pandemic surveillance plan.
483
Appendix 2
Examples of information and documentation that may
be required for the evaluation of a seasonal influenza
vaccine annual virus strain change
1. WHO-recommended strain list for the relevant hemisphere.
2. Manufacturer’s choice of strains for inclusion.
3. Details of manufacturing procedure (declaration if unchanged).
4. Validation of the inactivation and fragmentation.
5. Source, history and master/working seed characterization of each strain
included.
6. Egg or cell culture: safety specifications and tests (declaration if unchanged).
7. Qualification of potency test (single radial immunodiffusion – SRID)
reagents.
8. Final product release specifications and results (this must include endotoxin
release limit).
9. Retrospective data on the “efficacy or performance” of influenza vaccines
(preceding year or season).
10. Stability data (accelerated or from the most recent, or most similar, batch of
approved vaccine).
11. Copy of the approved package insert.

12. Copy of the proposed package insert, indicating:
(a) the year/season for which the vaccine will be used;
(b) WHO-recommended strains; and
(c) a statement that the vaccine complies with WHO recommendations
(southern or northern hemisphere) for the year/season.
13. Copy of the approved patient information leaflet.
484
Annex 7
14. Copy of the proposed patient information leaflet, indicating:

(a) the year/season for which the vaccine will be used; and
(b) WHO-recommended strains.
15. All labels and inner and outer containers must prominently indicate the
year/season for which the vaccine will be used, and a facsimile must be
submitted as proof.
16. International core data sheet or summary of product characteristics.
485
Annex 8
Labelling information of inactivated influenza vaccines
for use in pregnant women
Addendum to Annex 3 of WHO Technical Report Series, No. 927
1. Introduction 490
2. Background 490
4. Terminology 492
5. Labelling information 493
5.1 Indications and Usage 493
5.2 Warnings and Precautions 494
5.3 Contraindications 495
5.4 Use in Specific Populations 495
6. Summary 497
References 499
487

488
Annex 8
Abbreviations
EMA European Medicines Agency
GACVS WHO Global Advisory Committee on Vaccine Safety
IFPMA International Federation of Pharmaceutical Manufacturers
& Associations
IIV inactivated influenza vaccine
NITAG national immunization technical advisory group
SAGE WHO Strategic Advisory Group of Experts
SmPC summary of product characteristics
489
1. Introduction
Rates of morbidity and mortality resulting from seasonal influenza virus
infection are considered to be substantial worldwide (1, 2). Pregnant women
are especially vulnerable and have an increased risk of severe disease and
death from influenza. The infection may also lead to fetal complications such
as stillbirth, neonatal death, preterm delivery and decreased birthweight
(3, 4). For these reasons, the 2012 WHO position paper on vaccines against
influenza (3) – endorsed by the WHO Strategic Advisory Group of Experts on
Immunization (SAGE) – recommended the immunization of pregnant women
with trivalent inactivated influenza vaccine (IIV) at any stage of pregnancy.
SAGE also recommended that pregnant women should be given the highest
priority in countries considering the initiation or expansion of immunization
programmes for seasonal influenza vaccination (3, 5, 6). This recommendation
is based on evidence of a substantial risk of severe disease in this population
group and on evidence that the use of seasonal influenza vaccine is both safe
throughout pregnancy and effective in preventing influenza in women as well
as in their young infants in whom the disease burden is also high (3, 5). After
careful analysis of data worldwide, the WHO Global Advisory Committee
on Vaccine Safety (GACVS) concluded that there was no evidence of adverse
pregnancy outcomes associated with the vaccination of pregnant women with
several inactivated viral or bacterial vaccines, including IIVs (5, 6). However,
for various reasons, the implementation of influenza immunization during
pregnancy remains suboptimal (4). One reason for this has been the perceived
risk of administering influenza vaccine, or indeed any vaccine, to this population
group, particularly due to the precautionary language used in some product
labels and the likelihood of misinterpretation (7).
The development of this explanatory addendum arises from the
recommendations of SAGE regarding the immunization of pregnant women

with IIV and the resulting discussions at several WHO consultations (3, 6,
8–10), as well as discussions held during the 2015 meeting of the WHO Expert
Committee on Biological Standardization (11).
2. Background
Enhancing the uptake of vaccines during pregnancy is an important element
of WHO’s ongoing work to improve maternal and child health. As part of this
work, WHO held a consultation in July 2014 on influenza vaccines for pregnant
and lactating women which focused on the clinical data requirements for
product labelling information (8). The consultation was organized by the WHO
Technologies, Standards and Norms team and the WHO Initiative for Vaccine
490
Annex 8
Research and brought together regulators, manufacturers and researchers with

experience in vaccines. A further consultation was organized by WHO in 2015
to review existing guiding principles and regulations relating to product package
information for IIVs, and to explore the possibility of developing an addendum
to the existing WHO Recommendations for the production and control of
influenza vaccine (inactivated) (12) with the aim of clarifying and interpreting
the labelling information subsections to facilitate the appropriate use of IIVs in
pregnancy (9). Regulatory policy and requirements regarding permitted text in
the pregnancy and lactation subsections of product inserts were reviewed from
selected developed and developing countries (Brazil, Canada, Ghana, India,
Indonesia, Thailand and the United States of America) and from the European
Union. Also presented were the results of a 2014 Developing Country Vaccine
Regulators’ Network survey regarding regulatory policies and the interpretation
of the wording in the pregnancy and other subsections of IIV labelling. The
diversity of approach and understanding in different countries and regions was
evident. It also became clear that, in countries that import IIVs, the format, data
and language included in the product insert usually reflected the text approved
by the national regulatory authority (NRA) in the respective country of licensing.
Some developing countries require additional language that makes the perceived
cautionary message for use in pregnant women even stronger. The regulatory
position is based on the fact that licensing is product-specific, and reliant upon
data generated during the clinical evaluation of the vaccine and submitted by the
manufacturer. The European Medicines Agency (EMA) implemented a policy
based on an evaluation of all available evidence on the safety and effectiveness
of IIVs and expected all IIV licence holders in the European Union to amend
the pregnancy subsection of the labelling to include advice that IIVs can be used
during all stages of pregnancy (8, 13). However, this policy has recently changed
and the new guideline on influenza vaccines (clinical module) clarifies that a core
summary of product characteristics (SmPC) for IIV is no longer maintained but
individual SmPCs should be tailored to product-specific data (10, 14).
Some NRAs include recommendations made by national public health
advisory bodies on the use of IIVs in pregnancy to protect mother and infant
against influenza (15, 16), thus adding to the confusion regarding the meaning
of labelling information. Even though the wording in the package insert for IIVs
does not represent a contraindication to the use of the vaccines in pregnancy, the
particular wording employed is often misinterpreted to imply a contraindication.
Consequently it is interpreted as differing from statements made by advisory
bodies such as national immunization technical advisory groups (NITAGs) and
SAGE that recommend the use of IIVs in pregnancy to protect mother and infant
from the disease. Thus there is a perceived contradiction between the statements
of advisory bodies and the position of the NRA.
491
NITAG recommendations on the use of vaccines in pregnancy are

made on the basis that the benefit of vaccination in pregnant women usually
outweighs the risk of potential adverse effects in the mother or developing
offspring when: (a) the risk of disease exposure is high; (b) the infection poses
a special risk to mother and fetus; and (c) the vaccine is unlikely to cause harm.
In contrast, any statement in the “Indications and Usage” section of the labelling
that specifically addresses the use of the product in pregnancy can be approved
by an NRA only when supporting data from adequate and well-controlled studies
in pregnant women are available. As pregnant women are usually excluded from
clinical studies during vaccine development, licensure dossiers generally do not
include information on the safety and efficacy of a particular vaccine in pregnant
women. In the absence of such data, therefore, the “Indications and Usage”
section of the labelling will lack a statement that specifically describes the use of
the product in pregnancy.

The aim of this addendum is to provide clarification and interpretation of the
labelling information provided in the product insert of IIVs in order to facilitate
maternal immunization programmes. It is also intended to raise awareness of the
convergence of regulatory positions in spite of differing approaches to labelling
and regulatory language regarding the use of these vaccines in pregnant women.
On the basis of current evidence, the use of IIV in pregnant women is not
contraindicated.
This addendum applies to inactivated trivalent and quadrivalent
(tetravalent) influenza vaccines for which sufficient safety data are available. It is
intended for NRAs, manufacturers, end-user programme managers and NITAGs.
Liability issues are beyond the scope of this document.
4. Terminology
The definitions given below apply to the terms as used in this WHO guidance
document. These terms may have different meanings in other contexts.
Label(ling): all forms of product information – that is, container label,
SmPC, product/package insert, package leaflet and prescribing information.
Maternal immunization: frequently used to refer to vaccination prior
to, during or after pregnancy. For the purposes of this document the term refers
specifically to vaccination during pregnancy.
National Immunization Technical Advisory Group (NITAG): a national
expert advisory group that evaluates the available evidence on national disease
incidence, and available vaccines, in order to provide advice to the health ministry
492
Annex 8
on national immunization programme policies, and on priority vaccines and

target populations.
Summary of product characteristics (SmPC): the SmPC is the basis of
information for health-care professionals on how to use the medicinal product
safely and effectively. Product labelling should be drawn up in accordance with
the SmPC.
5. Labelling information
As with all prescription drugs and biological products, IIVs must be accompanied
by labelling that summarizes the scientific information concerning their safe and
effective use.
Labelling includes the package insert, which is also referred to as
prescribing information or the SmPC (17). This component of labelling is
the primary mechanism through which regulatory agencies and vaccine
manufacturers communicate essential, science-based information. This
information is then used by health-care professionals to make prescribing
decisions and to counsel patients about the risks and benefits of a product. The
content and format requirements for labelling are prescribed by regulations
specific to the country where the vaccine is licensed and may differ between
countries (17, 18). Nevertheless, common principles include that prescribing
information should be based on available data, that it must not be misleading
and that it must not contain implied claims or uses for which there is inadequate
evidence of safety or effectiveness (19).
The labelling sections relevant to the use of vaccines in pregnancy –
namely, “Indications and Usage”, “Warnings and Precautions”, “Contraindications”,
and “Use in Specific Populations” – are described below. Countries have
information on vaccination in pregnancy under various sections. Information
regarding the use of an IIV in pregnancy is typically found under the “Use in
Specific Populations” section. However, in some countries the NRA has required
that precautionary statements about the use of an IIV in pregnancy should be
included under the “Warnings and Precautions” and “Contraindications” sections
because safety data on use of the vaccine in pregnancy may be unavailable
or insufficient.
5.1 Indications and Usage

The “Indications and Usage” section of the product labelling communicates a
product’s approved indication(s) and should clearly convey the use(s) for which
the product has been shown to be safe and effective. Although pre-licensure
clinical trials are generally conducted in carefully selected populations, the
“Indications and Usage” statement(s) often reflect a broader population and take
493
into consideration the generalizability of the evidence. Typically, for preventive

vaccines, the “Indications and Usage” statement(s) state the disease being
prevented and the age range for which use is approved.
Specific regulatory requirements and standards for demonstrating that
a vaccine is safe and effective may vary between NRAs. However, in general, the
standards for demonstrating the safety of a vaccine for its intended indication
take into consideration the condition of the recipients and the characteristics of
the product. It is expected that pre-licensure data demonstrating that a vaccine
is effective for the intended indication and use are derived from adequate and
well-controlled clinical studies. Data from pregnancy exposure registries,
epidemiological studies and case-series are typically collected in the post-
marketing period and are used to inform the “Use in Specific Populations” section
of the labelling (see section 5.4 below).
While data from related, similar vaccines may be supportive of an
indication for use, it is typically expected that the specific vaccine is evaluated for
safe and effective use in the intended population. For most IIVs that are currently
licensed, data from adequate and well-controlled studies demonstrating that
vaccination during pregnancy is safe and effective for the pregnant woman or
newborn infant may not be available to support an indication in the labelling.
Data from studies published in the literature on the use of IIV in pregnancy may
not have been submitted to NRAs or may not meet regulatory requirements.
In such cases, product- (brand)-specific data demonstrating that the vaccine
is safe and effective may not be available. Consequently, the prescribing
information for IIVs will not include an “Indications and Usage” statement that
specifically addresses use in pregnancy. This does not mean, however, that IIVs
are contraindicated for use in pregnancy. IIVs are licensed for use in an age
range that includes women of childbearing age. In the absence of evidence that
the risk of use in pregnancy clearly outweighs any possible benefit, there is no
specific contraindication for use in pregnancy and, consequently, IIVs may be

administered to pregnant women. Available data specific to the use of IIVs in
pregnancy will be included in the “Use in Specific Populations” section of the
labelling (see section 5.4 below).
5.2 Warnings and Precautions

The “Warnings and Precautions” section of the product labelling is intended to
include, but is not limited to, a description of adverse events that are serious or
otherwise clinically significant because they have implications for prescribing
decisions or for patient management. For an adverse event to be included in this
section there should be reasonable evidence of a causal association between the
adverse event and the drug or biological product.
494
Annex 8
Clinically significant adverse reactions that have not been observed

following use of the specific drug or biological product, but which are anticipated
on the basis of data on another drug in the same class, or from animal data,
should be included under “Warnings and Precautions”. In addition, any
clinically significant interference with laboratory tests, clinically significant
drug interactions, and any special care or monitoring required to ensure safe
use, should also be included under “Warnings and Precautions”. The description
of each adverse reaction or topic included under “Warnings and Precautions”
is cross-referenced to a more detailed discussion of the risk elsewhere in the
labelling (for example, in “Adverse Reactions” and “Use in Specific Populations”).
Some NRAs require the inclusion of information on use of IIVs during pregnancy
in the “Warnings and Precautions” section of the labelling.
5.3 Contraindications
Although the specific wording used in the “Contraindications” section of
the product labelling may depend on the requirements of the NRA where
the vaccine has been licensed, there is a common requirement that drugs
or biological products, including vaccines, should be contraindicated only
in those situations where the known risk from use clearly outweighs any
possible benefit. Only known hazards, not theoretical possibilities, should
be the basis for contraindication. As an example relating to vaccine use in
pregnant women, evidence in humans or animals that a vaccine poses a serious
risk of developmental toxicity during pregnancy would usually warrant a
contraindication for use during pregnancy. However, for IIVs, if available
animal or human data do not indicate a risk in pregnancy that clearly outweighs
benefit, or if data are unavailable to inform risk in pregnancy, there should not
be a contraindication for use during pregnancy.
5.4 Use in Specific Populations

The “Use in Specific Populations” section of the product labelling summarizes
important differences in the response to the product, or in recommendations for
use, in specific populations. Information relevant to the use of a product during
pregnancy is generally found under this section and is sometimes referred to
as the “pregnancy subsection” of product labelling. However, depending on the
labelling requirements of the NRA where the vaccine was licensed, information
regarding the use of IIV in pregnancy may also be found in other sections of
product labelling, such as the “Warnings and Precautions” section (9).
The pregnancy subsection of the product labelling includes data, when
available, from reproductive-toxicity studies conducted in animal models
to assess the potential developmental and reproductive risks of the product.
495
Data that may be available concerning the safety of the product in pregnant
women are also described in this section. Sources of human data may include
pregnancy registries, pre-licensure clinical trials in which pregnant women were
inadvertently exposed to the product, large-scale epidemiological studies and
case-series reporting rare adverse events. In general, information regarding use of
IIVs during pregnancy is derived from post-marketing studies (for example, via
registries and/or from maternal immunization studies published in the literature).
The quality and quantity of data from specific sources will be evaluated by the
NRA to determine whether the data are scientifically acceptable for inclusion in
the pregnancy subsection of the product labelling. In some countries, the NITAG
recommendations are included in this section.
As with other sections of the product labelling, country-specific
requirements prescribe the information, and frequently the specific wording, to
be included in the pregnancy subsection in relation to what is known about the
risks of using the product in pregnancy.
WHO’s prequalification evaluation of the prescribing information
is evidence-based and takes into consideration the prescribing information
approved by the NRA of record for prequalification (generally the NRA in the
country of manufacture).
Required statements included in the pregnancy subsection of the
product labelling have often been precautionary (for example, Should be used
only following advice of a health-care professional; If pregnant, please inform your
doctor or pharmacist; Use only if clearly needed). The rationale for requiring
such language has largely stemmed from a lack of data from well-controlled
clinical trials rather than evidence suggesting specific risks of vaccination during
pregnancy. Such precautionary language has sometimes been misinterpreted to
mean that pregnancy is a contraindication for use.
Whereas many NRAs require that labelling includes such precautionary
language regarding use in pregnancy, some countries are considering ways to
improve the clarity of the information included in the pregnancy subsection

of the product labelling. For example, the United States Food and Drug
Administration recently revised its labelling regulations so that they no longer
require such precautionary language (18, 20). With the implementation of the
Pregnancy and Lactation Labeling Rule in the United States of America in June
2015 (18), the revised regulations now require that the pregnancy subsection of
product labelling includes narrative summaries of the risks of a product during
pregnancy and discussions of the data supporting those summaries. Under the
revised regulations, labelling will include relevant available clinical information
arising from the use of the product in pregnant and lactating women, as well as
relevant available animal and pharmacological data, to help inform prescribing
decisions and the counselling of women on the use of the product during
pregnancy and lactation.
496
Annex 8
6. Summary
IIVs are not contraindicated for use in pregnancy. The “Indications and Usage”
section for these vaccines specifies an age range that includes women of
childbearing age. Consequently, the lack of an “Indications and Usage” statement
specifically addressing use in pregnant women does not preclude use of these
vaccines during pregnancy. Certain countries include information on the use of
IIV in pregnancy under the “Warnings and Precautions” or “Contraindications”
sections of product labelling. However, this does not reflect a known or
suspected safety issue relating to the use of these vaccines during pregnancy
but rather a precautionary approach taken by certain NRAs. Programmatic
recommendations (such as those from SAGE and some NITAGs) on the use of
IIVs during pregnancy are consistent with labelling.

The first draft of this WHO guidance document was prepared by Dr E.
Griffiths, Consultant, Kingston-upon-Thames, England; Dr M. Gruber, United
States Food and Drug Administration, Center for Biologics Evaluation and
Research, the USA; Dr H. Kang, World Health Organization, Switzerland; Dr M.
Pfleiderer, Paul-Ehrlich-Institut, Germany; and Dr J. Southern, Consultant,
Cape Town, South Africa, taking into consideration comments received from
a WHO working group meeting on the development of guidelines on labelling
information of influenza vaccines intended to be used for pregnant women, held
in Geneva, Switzerland, 24–25 September 2015 and attended by: Dr D. Baswal,
Ministry of Health and Family Welfare, India; Dr G. Coleman, Health Canada,
Canada; Dr E. Griffiths. Consultant, Kingston-upon-Thames, England; Dr M.
Gruber, United States Food and Drug Administration, Center for Biologics
Evaluation and Research, the USA; Mrs N. Hidayati, National Agency of Drug
and Food Control, Indonesia; Mrs T. Jivapaisarnpong, Ministry of Public Health,
Thailand; Mr A. Kukrety, Central Drugs Standard Control Organization, India;
Dr P. Neels, Consultant, Zoersel, Belgium; Dr E. Nkansah, Food and Drugs
Authority, Ghana; Dr J. Southern, Consultant, Cape Town, South Africa; Ms M.F.
Thees, Brazilian Health Regulatory Agency, Brazil; Dr B. Voordouw, Medicines
Evaluation Board, the Netherlands; and Dr H. Kang, Dr P. Lambach, Mr O.C.
Lapujade, Dr A. Meek, Dr J. Ortiz and Ms C.A. Rodriguez-Hernandez, World
Health Organization, Switzerland.
The draft document was posted on the WHO Biologicals website
for a first round of public consultation from 12 January to 19 February 2016
and comments were received from the following reviewers: Dr N. Bachtiar,
BioFarma, Indonesia; Dr D. Brasseur (provided the consolidated comments of
the Vaccines Working Party, Committee for Medicinal Products for Human Use,
497
EMA), England; Dr C. Brades and Ms M.F. Thees, Brazilian Health Regulatory

Agency, Brazil; Dr G. Coleman, Health Canada, Canada; Ms J. Dahlan and Dr N.
Hidayati, National Agency of Drug and Food Control, Indonesia; Dr S. Darbooy,
Food and Drug Administration, the Islamic Republic of Iran; Dr C. Fu, Chinese
Center for Disease Control and Prevention, China; Dr M. Gruber (provided the
consolidated comments of the United States Food and Drug Administration, Center
for Biologics Evaluation and Research), the USA; Dr R. Hollingsworth (provided
the consolidated comments of the International Federation of Pharmaceutical
Manufacturers & Associations (IFPMA)), Sanofi Pasteur, the USA; Dr H. Jee,
Green Cross, Republic of Korea; Dr D. Kim, Ministry of Food and Drug Safety,
Republic of Korea; Dr M. Kucuku, National Agency for Medicines & Medical
Devices, Albania; Dr A. Meek (provided the consolidated comments of the
World Health Organization prequalification team), World Health Organization,
Switzerland; Dr S. Nishioka, Ministry of Health, Brazil; Dr J. Pathirana, Chris
Hani Baragwanath Academic Hospital, South Africa; Dr A. Weinberg, University
of Colorado Denver, the USA.
The document WHO/BS/2016.2280 was prepared by Dr E. Griffiths,
Consultant, Kingston-upon-Thames, England; Dr M. Gruber, United States
Food and Drug Administration, Center for Biologics Evaluation and Research,
the USA; Dr H. Kang, World Health Organization, Switzerland; Dr Y. Sun, Paul-
Ehrlich-Institut, Germany; and Dr J. Southern, Consultant, Cape Town, South
Africa, taking into consideration comments received from the first round of
public consultation as well as from a WHO informal consultation on labelling
information of influenza vaccines intended to be used for pregnant women, held
in Geneva, Switzerland, 4–5 April 2016 and attended by: Mr P. Akarapanon,
Ministry of Public Health, Thailand; Dr C. Blades, Brazilian Health Regulatory
Agency, Brazil; Dr G. Coleman, Health Canada, Canada; Ms J. Dahlan, National
Agency of Drug and Food Control, Indonesia; Dr S. Darbooy, Food and Drug
Administration, the Islamic Republic of Iran; Mrs M. Darko, Food and Drugs
Authority, Ghana; Dr M. Downham (IFPMA representative), AstraZeneca,

England; Dr V. Franck (IFPMA representative), GlaxoSmithKline Vaccines,
Belgium; Dr E. Griffiths, Consultant, Kingston-upon-Thames, England; Dr M.
Gruber, United States Food and Drug Administration, Center for Biologics
Evaluation and Research, the USA; Dr J. Hernandez, Centro para el Control
Estatal de la Calidad de los Medicamentos, Cuba; Dr R. Hollingsworth (IFPMA
representative), Sanofi Pasteur, the USA; Mrs T. Jivapaisarnpong, Ministry of
Public Health, Thailand; Dr D. Kim, Ministry of Food and Drug Safety, Republic
of Korea; Professor I. Krasilnikov, St. Petersburg Scientific Research Institute of
Vaccines and Sera, Russian Federation; Mr R. Shakhapure, Ministry of Health and
Family Welfare, India; Dr J. Southern, Consultant, Cape Town, South Africa; Dr Y.
Sun, Paul-Ehrlich-Institut, Germany; Mrs P. Thanaphollert, Ministry of Public
Health, Thailand; Dr F. Torkestani, Ministry of Health and Medical Education,
498
Annex 8
the Islamic Republic of Iran; Dr B. Voordouw, Medicines Evaluation Board, the

Netherlands; Dr N. Wairagkar, Bill & Melinda Gates Foundation, the USA; and
Mr G. Enwere, Dr I. Knezevic, Dr P. Lambach and Dr J. Ortiz, World Health
Organization, Switzerland.
The document WHO/BS/2016.2280 was then posted on the WHO
Biologicals website for a second round of public consultation from 27 July to
23 September 2016 and feedback was received from the following reviewers:
Dr C. Blades, Brazilian Health Regulatory Agency, Brazil; Dr G. Coleman, Health
Canada, Canada; Dr J. Hernandez, Centro para el Control Estatal de la Calidad
de los Medicamentos, Cuba; Dr M. Darko (provided the consolidated comments
of the Ghana Food and Drugs Authority), Ghana; Dr A. Hobbs (provided the
consolidated comments of the Department of Health), Australia; Dr M. Kucuku,
National Agency for Medicines & Medical Devices, Albania; Dr N. MacDonald,
Dalhousie University, Canada; Dr M. Mura (provided the consolidated comments
of the EMA); Dr S. Nishioka, Ministry of Health, Brazil; Dr E. Nkansah, Food
and Drugs Authority, Ghana; Dr G. Raychaudhuri (provided the consolidated
comments of the United States Food and Drug Administration, Center for Biologics
Evaluation and Research), the USA; Ms M. Xydia-Charmanta (provided the
consolidated comments of the IFPMA), Switzerland.
References
1. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V et al. Global and regional mortality
from 235 causes of death for 20 age groups in 1990 and 2010: a systematic analysis for the Global
Burden of Disease Study 2010. Lancet. 2012;380:2095–128 (abstract: http://www.thelancet.com/
journals/lancet/article/PIIS0140-6736(12)61728-0/abstract, accessed 5 December 2016).
2. Zhou H, Thompson WW, Viboud CG, Ringholz CM, Cheng PY, Steiner C et al. Hospitalizations
associated with influenza and respiratory syncytial virus in the United States, 1993–2008. Clin
Infect Dis. 2012;54(10):1427–36 (abstract: https://www.ncbi.nlm.nih.gov/pubmed/22495079,
3. Vaccines against influenza. WHO position paper – November 2012. Wkly Epidemiol Rec.
2012;87(47):461–76 (http://www.who.int/wer/2012/wer8747.pdf?ua=1, accessed 5 December
2016).
4. Meeting of the Strategic Advisory Group of Experts on immunization, November 2013 –
conclusions and recommendations. Wkly Epidemiol Rec. 2014;89(1):1–20 (http://www.who.int/
wer/2014/wer8901.pdf?ua=1, accessed 6 December 2016).
5. Report on safety of immunization during pregnancy. WHO Global Advisory Committee on
Vaccine Safety, 18 October 2013. Geneva: World Health Organization; 2013 (http://www.who.
int/immunization/sage/meetings/2013/november/2_GACVS_pregnancy_report.pdf, accessed 6
December 2016).
6. Global Advisory Committee on Vaccine Safety, 12–13 June 2013. Wkly Epidemiol Rec.
2013;88(29):301–12 (http://www.who.int/wer/2013/wer8829.pdf, accessed 6 December 2016).
499
7. Top KA, Arkell C, Scott H, McNeil SA, Mannerfeldt J, Ortiz JR et al. Effect of package insert
language on health-care providers’ perceptions of influenza vaccination safety during pregnancy.
Lancet. 2016;4(10):e690–1 (http://www.thelancet.com/journals/langlo/article/PIIS2214-109X(16)
30182-6/fulltext, accessed 6 December 2016).
8. WHO consultation on influenza vaccines for pregnant and lactating women: clinical data
requirements for product labelling, 15–16 July 2014, Geneva, Switzerland. Executive summary.
Geneva: World Health Organization; 2014 (http://www.who.int/biologicals/vaccines/INFLUENZA_
VACCINES_Executive_Summary_FINAL_29_Sept_R.pdf, accessed 6 December 2016).
9. WHO working group meeting on labelling information of influenza vaccines intended to be used
in pregnant women, 24–25 September 2015, Geneva, Switzerland. Executive summary. Geneva:
World Health Organization; 2015 (http://www.who.int/biologicals/areas/vaccines/Executive_
summary_FINAL_labelling_Sept_mtg_16_Dec_2015.pdf, accessed 6 December 2016).
10. WHO informal consultation on labelling information of influenza vaccines intended to be
used in pregnant women, 4–5 April 2016, Geneva, Switzerland. Executive summary. Geneva:
World Health Organization; 2016 (http://www.who.int/biologicals/areas/vaccines/INFLUENZA_
Executive_summary_Final_labelling_Apri_mgt_10_June_2016.pdf?ua=1, accessed 6 December
2016).
11. Labelling information for influenza vaccines intended for use in pregnant women. In: WHO
Expert Committee on Biological Standardization: sixty-sixth report. Geneva: World Health
Organization; 2016:25–26 (WHO Technical Report Series, No. 999; http://apps.who.int/iris/bitstr
eam/10665/208900/1/9789240695634-eng.pdf?ua=1, accessed 6 December 2016).
12. Recommendations for the production and control of influenza vaccine (inactivated). In: WHO
Expert Committee on Biological Standardization: fifty-fourth report. Geneva: World Health
biologicals/publications/trs/areas/vaccines/influenza/ANNEX%203%20InfluenzaP99-134.pdf,
13. Keller-Stanislawski B, Englund JA, Kang G, Mangtani P, Neuzil K, Nohynek H et al. Safety of
immunization during pregnancy: a review of the evidence of selected inactivated and live
attenuated vaccines. Vaccine. 2014;32(52):7057–64 (abstract: https://www.ncbi.nlm.nih.gov/
pubmed/25285883/, accessed 6 December 2016).
14. Guideline on influenza vaccines. Non-clinical and clinical module. London: European Medicines
Agency; 2014 (EMA/CHMP/VWP/457259/2014; http://www.ema.europa.eu/docs/en_GB/
document_library/Scientific_guideline/2014/07/WC500170300.pdf, accessed 6 December 2016).

15. ECDC scientific advice on seasonal influenza vaccination of children and pregnant women.
Stockholm: European Centre for Disease Prevention and Control; 2012 (http://ecdc.europa.eu/
en/publications/Publications/Seasonal%20influenza%20vaccination%20of%20children%20
and%20pregnant%20women.pdf, accessed 6 December 2016).
16. Immunization in pregnancy and breastfeeding. Canadian immunization guide: Part 3 –
Vaccination of specific populations. Ottawa: Public Health Agency of Canada; updated July 2015
(http://www.phac-aspc.gc.ca/publicat/cig-gci/p03-04-eng.php#a5, accessed 6 December 2016).
17. A guideline on Summary of Product Characteristics (SmPC), Revision 2. European Commission,
September 2009 (http://ec.europa.eu/health/files/eudralex/vol-2/c/smpc_guideline_rev2_en.pdf,
18. Content and format of labeling for human prescription drug and biological products; requirements
for pregnancy and lactation labeling. United States Food and Drug Administration. Federal
Register 79(233); 4 December 2014: effective June 30, 2015 (https://www.gpo.gov/fdsys/pkg/FR-
2014-12-04/pdf/2014-28241.pdf, accessed 6 December 2016).
500
Annex 8
19. Marshall V, Gruber M. Influenza immunization during pregnancy: US regulatory perspective. Am J
Obstet Gynecol. 2012;207(3 Suppl):S57–62 (http://www.ajog.org/article/S0002-9378(12)00729-
6/pdf, accessed 6 December 2016).
20. Roberts JN, Gruber MF. Regulatory considerations in the clinical development of vaccines
indicated for use during pregnancy. Vaccine. 2015;33(8):966–72 (abstract: http://www.
sciencedirect.com/science/article/pii/S0264410X1401723X, accessed 6 December 2016).
501
Annex 9
Guidelines on clinical evaluation of vaccines: regulatory
expectations
Replacement of Annex 1 of WHO Technical Report Series, No. 924
1. Introduction 506
3. Terminology 510
4. Vaccine clinical development programmes 514
4.1 General considerations 514
4.2 Pre-licensure clinical development programmes 516
4.3 Post-licensure clinical evaluations 517
5. Immunogenicity 518
5.2 Characterization of the immune response 518
5.3 Measuring the immune response 519
5.4 Identification and use of immune correlates of protection 524
5.5 Immunogenicity trials 526
5.6 Specific uses of immunogenicity trials 531
6. Efficacy and effectiveness 545
6.1 General considerations for efficacy trials 545
6.2 Types of efficacy trials 547
6.3 Design and conduct of efficacy trials 548
6.4 Approaches to determination of effectiveness 558
7. Safety 560
7.2 Assessment of safety in clinical trials 560
7.3 Size of the pre-licensure safety database 566
7.4 Post-licensure safety surveillance 567
References 571
503

It is recommended that modifications to these WHO Guidelines are
made only on condition that such modifications ensure that a vaccine
is at least as safe and efficacious as one evaluated in accordance with
the guidance set out below.
504
Annex 9
Abbreviations
AE adverse event
AEFI adverse event following immunization
AESI adverse event of special interest
AR attack rate
ARU attack rate in unvaccinated (control group)
ARV attack rate in vaccinated group
GMC geometric mean concentration
GMP good manufacturing practice
GMT geometric mean titre
HPV human papillomavirus
ICH International Conference on Harmonisation of Technical
Requirements for Registration of Pharmaceuticals for
Human Use
ICP immune correlate of protection
LLOD lower limit of detection
LLOQ lower limit of quantification
OPA opsonophagocytic antibody
RR relative risk
SAE serious adverse event
SBA serum bactericidal antibody
505
1. Introduction
These WHO Guidelines are intended to replace the WHO Guidelines on clinical
evaluation of vaccines: regulatory expectations, which were adopted by the Expert
Committee on Biological Standardization in 2001 (1). The document of 2001
provided guidance on the clinical evaluation of vaccines as well as on WHO
vaccine prequalification.
Since 2001, more than 20 vaccine-specific documents (each including a
section on clinical evaluation) have been adopted by the Committee. Originally
intended to be read in conjunction with the 2001 document, these documents
provide guidance on both oral and inactivated polio vaccines, whole cell pertussis
and acellular pertussis vaccines, meningococcal conjugate vaccines for serotypes
A and C, and pneumococcal conjugate vaccines, as well as on vaccines intended to
prevent diseases caused by rotaviruses, dengue viruses, human papillomaviruses
(HPVs) and malaria parasites.
These revised WHO Guidelines have been prepared to reflect the
scientific and regulatory experience that has been gained from vaccine clinical
development programmes since the adoption of the 2001 version. They are
intended for use by national regulatory authorities (NRAs), companies developing
and holding licences for vaccines, clinical researchers and investigators. The
document takes into account the content of clinical development programmes,
clinical trial designs, the interpretation of trial results and post-licensing activities.
The main content changes (modification or expansion of previous text
and additional issues covered) include, but are not limited to, the following:
Immunogenicity
■■ general principles for comparative immunogenicity studies,
including selection of the comparators, end-points and acceptance
criteria for concluding non-inferiority or superiority of immune

responses;
■■ situations in which age de-escalation studies are not necessary;
■■ assessment of the need for and timing of post-primary doses;
■■ use of different vaccines for priming and boosting;
■■ assessment of the ability of vaccines to elicit immune memory or to
cause hyporesponsiveness;
■■ use of immunogenicity data to predict vaccine efficacy, with or
without bridging to efficacy data;
■■ the derivation and uses of immune correlates of protection (ICPs);
■■ vaccination of pregnant women to protect them and/or their infants.
506
Annex 9
Efficacy and effectiveness

■■ the need for, and feasibility of, conducting vaccine efficacy studies;
■■ selection of appropriate control groups in different circumstances;
■■ comparison of new and licensed vaccines containing antigens from
different numbers of types or subtypes of the same organism;
■■ prediction of vaccine efficacy when there is no ICP and vaccine
efficacy studies are not feasible;
■■ preliminary and pivotal vaccine efficacy studies and their design;
■■ vaccines with modest efficacy and/or that provide a short duration
of protection;
■■ extrapolation of data between geographically or genetically diverse
populations;
■■ the role and potential value of human challenge studies;
■■ the role of sponsors and public health authorities in generating
vaccine-effectiveness data.
Safety
■■ detailed consideration of the collection and analysis of safety data
from clinical trials;
■■ consideration of size of the pre-licensure database by type of vaccine
and its novelty;
■■ consideration of the safety database by population subgroup;
■■ special safety considerations by vaccine construct;
■■ circumstances of limited pre-licensure safety data;
■■ use of registries;
■■ issues regarding vaccine pharmacovigilance activities.
Because a separate document on the nonclinical evaluation of vaccines
was established in 2003 (2), the corresponding section in the 2001 Guidelines has
been removed. Furthermore, the structure of the document has changed, with
a number of methodological considerations now incorporated into the relevant
sections and subsections rather than being described in a separate section. In
line with all the changes made in the document, the terminology and references
have been updated.
WHO has also made available several guidelines, manuals and reports
relevant to vaccine clinical development programmes. These should be consulted
as appropriate, and include:
507
■■ Guidelines for good clinical practice (GCP) for trials on pharmaceutical

products (3);
■■ WHO good manufacturing practices for pharmaceutical products:
main principles (4);
■■ WHO good manufacturing practices for biological products (5);
■■ Guidelines on nonclinical evaluation of vaccines (2);
■■ Guidelines on the nonclinical evaluation of vaccine adjuvants and
adjuvanted vaccines (6);
■■ Guidelines on procedures and data requirements for changes to
approved vaccines (7);
■■ Guidelines for independent lot release of vaccines by regulatory
authorities (8);
■■ Recommendations for the evaluation of animal cell cultures as
substrates for the manufacture of biological medicinal products and for
the characterization of cell banks (9);
■■ Clinical considerations for evaluation of vaccines for
prequalification (10);
■■ The WHO manual Immunization in practice: a practical guide for
health staff (11);
■■ Expert consultation on the use of placebos in vaccine trials (12).
Furthermore, guidance on various aspects of pre-licensure clinical

development programmes for vaccines and on post-licensure assessment is also
available from several other bodies, such as the International Conference on
Harmonisation of Technical Requirements for Registration of Pharmaceuticals
for Human Use (ICH), the European Medicines Agency (EMA), the United
States Food and Drug Administration and the United Kingdom Medical
Research Council. These WHO Guidelines are intended to complement these
other documents.

These WHO Guidelines consider clinical development programmes for vaccines
that are intended to prevent clinical disease in humans by eliciting protective
immune responses. The protective immune response to vaccination may be
directed against one or more specific antigenic components of microorganisms
or against substances produced and secreted by them (for example, toxins) that
are responsible for clinical disease. The clinical disease prevented by vaccination
508
Annex 9
may be an acute infectious disease and/or a disease that results from chronic
infection with an infectious agent.
These Guidelines are applicable to the clinical development of:
■■ new candidate vaccines;
■■ licensed vaccines;
■■ vaccines that are given by any route of administration;
■■ vaccines that may be given before exposure or shortly after known
or presumed exposure to an infectious agent to prevent the onset of
clinical disease.
The Guidelines are further applicable to vaccines that contain one or
more of the following:
■■ microorganisms that have been inactivated by chemical and/or
physical means;
■■ live microorganisms that are not virulent in humans as a result of
attenuation processes or specific genetic modification;
■■ antigenic substances that have been derived from microorganisms
(these may be purified from microorganisms and used in their
natural state, or they may be modified, for example, detoxified by
chemical or physical means, aggregated or polymerized);
■■ antigens that have been manufactured by synthetic processes or
produced by live organisms using recombinant RNA or DNA
technology;
■■ antigens (however manufactured) that have been chemically
conjugated to a carrier molecule to modify the interaction of the
antigen with the host immune system;
■■ antigens that are expressed by another microorganism which itself
does not cause clinical disease but acts as a live vector (for example,
live viral vectored vaccines and live-attenuated chimeric vaccines).
In addition, although naked DNA vaccines are not specifically discussed,
the principles and development programmes outlined are broadly applicable.
These Guidelines do not apply to:
■■ therapeutic vaccines (that is, those intended for treatment
of disease);
■■ vaccines intended for any purpose other than the prevention of
clinical disease caused by infectious agents.
509
3. Terminology
Adverse event (AE): any untoward medical occurrence in a participant
in a clinical trial. An AE does not necessarily have a causal relationship with
the vaccine.
Adverse event following immunization (AEFI): any untoward medical
occurrence that follows immunization and which does not necessarily have a
causal relationship with the use of the vaccine. The AEFI may be any unfavourable
or unintended sign, abnormal laboratory finding, symptom or disease. In clinical
trial documentation AEFI may often be shortened to AE.
Adverse event of special interest (AESI): a clinically important untoward
medical occurrence that is either known to occur following administration of
the type of vaccine under study (for example, hypotonic-hyporesponsive
episodes or febrile convulsions) or is considered to be a possible risk on the
basis of knowledge of the content of the vaccine and/or its interaction with the
host immune system (for example, autoimmune disease or antibody-dependent
enhanced clinical disease).
Attack rate (AR): the proportion of the population exposed to an
infectious agent that goes on to develop clinically manifest disease.
Blinding: a procedure by which one or more parties involved in a
clinical trial are kept unaware of the randomized intervention.
Booster dose: a dose that is given at a certain interval after completion of
the primary series that is intended to boost immunity to, and therefore prolong
protection against, the disease that is to be prevented.
Case ascertainment: the method adopted for detecting cases of the
disease targeted for prevention by vaccination in a vaccine efficacy trial or in a
study of vaccine effectiveness.
Case definition: the predefined clinical and/or laboratory criteria that

must be fulfilled to confirm a case of a clinically manifest disease in a vaccine
efficacy trial or in a study of vaccine effectiveness.
Cluster randomization: randomization of subjects by group (for example,
by household or by community) as opposed to randomization of individual
subjects within a clinical trial.
Geometric mean concentration (GMC): the average antibody
concentration for a group of subjects calculated by multiplying all values and
taking the nth root of this number, where n is the number of subjects with
available data.
Geometric mean titre (GMT): the average antibody titre for a group
of subjects calculated by multiplying all values and taking the nth root of this
number, where n is the number of subjects with available data.
510
Annex 9
Good clinical practice (GCP): GCP is a process that incorporates

established ethical and scientific quality standards for the design, conduct,
recording and reporting of clinical research that involves the participation of
human subjects. Compliance with GCP provides public assurance that the
rights, safety and well-being of research subjects are protected and respected,
consistent with the principles enunciated in the Declaration of Helsinki and
other internationally recognized ethical guidelines, and also ensures the integrity
of clinical research data.
Good manufacturing practice (GMP): GMP is the aspect of quality
assurance that ensures that medicinal products are consistently produced and
controlled to the quality standards appropriate to their intended use and as
required by the product specification.
Immune correlate of protection (ICP): an ICP is most commonly
defined as a type and amount of immunological response that correlates with
vaccine-induced protection against an infectious disease and that is considered
predictive of clinical efficacy (13).
Immune memory: an immunological phenomenon in which the
primary contact between the host immune system and an antigen results in a
T-cell-dependent immune response, often referred to as priming of the immune
system. Effective priming results in the development of antigen-specific memory
B-cells and an anamnestic (memory) immune response to post-primary doses,
which are commonly referred to as booster doses.
Immunogenicity: the capacity of a vaccine to elicit a measurable
immune response.
New candidate vaccine: a new candidate vaccine is a vaccine that is
regarded in national regulations as separate and distinct from other candidate
and licensed vaccines. Examples of new candidate vaccines include but are not
limited to:
■■ a vaccine that contains a new antigenic component (that is, one not
previously used in a licensed vaccine);
■■ a vaccine that contains a new adjuvant;
■■ a vaccine that contains antigen(s) ± adjuvant(s) not previously
combined together in a vaccine;
■■ a vaccine with the same antigenic component(s) ± adjuvant as
a licensed vaccine that is produced by a different manufacturer
(including situations in which seed lots or bulk antigenic
components used to make a licensed vaccine are supplied to other
manufacturers for their own vaccine production).
Non-inferiority trial: non-inferiority trials aim to demonstrate that the
test intervention is not worse than the reference intervention by more than
511
a small pre-specified amount known as the non-inferiority margin. In non-

inferiority trials it is assumed that the reference intervention has been established
to have a significant clinical effect (against placebo).
Pharmacovigilance: pharmacovigilance encompasses the science and
activities relating to the detection, assessment, understanding and prevention of
adverse effects or any other possible drug-related problems (14).
Pivotal trials: pivotal clinical trials provide the major evidence in
support of licensure.
Posology: the vaccine posology for a specific route of administration
and target population includes:
■■ the dose content and volume delivered per dose;
■■ the dose regimen (that is, the number of doses to be given in the
primary series and, if applicable, after the primary series);
■■ the dose schedule (that is, the dose intervals to be adhered to
within the primary series and between the primary series and any
further doses).
Post-licensure safety surveillance: a system for monitoring AEFIs in the
post-licensure period.
Post-primary doses: additional doses of vaccine given after a time
interval following the primary series of vaccination.
Preliminary trial: a clinical trial that is not intended to serve as a pivotal
trial. Preliminary trials are usually conducted to obtain information on the
safety and immunogenicity of candidate vaccine formulations and to select the
formulation(s) and regimen(s) for evaluation in pivotal trials. Preliminary trials
may also serve to inform the design of pivotal trials (for example, by identifying
the most appropriate populations and end-points for further study). On occasion,
a preliminary trial may provide an initial evaluation of vaccine efficacy.
Primary vaccination: the first vaccination or the initial series of

vaccinations intended to establish clinical protection.
Protocol: a document that states the background, rationale and objectives
of the clinical trial and describes its design, methodology and organization,
including statistical considerations and the conditions under which it is to
be performed and managed. The protocol should be signed and dated by the
investigator, the institution involved and the sponsor.
Randomization: in its simplest form, randomization is a process by
which n individuals are assigned to test (n T ) or control (nC ) treatment(s) so
that all possible groups of size n = n T + nC have equal probability of occurring.
Thus, randomization avoids systematic bias in the assignment of treatment.
Responder: a trial subject who develops an immune response (humoral
or cellular) that meets or exceeds a predefined threshold value using a specific
512
Annex 9
assay. This term may be applied whether or not there is an established ICP and
when the clinical relevance of achieving or exceeding the predefined response
is unknown.
Responder rate: the responder rate is the percentage of subjects in a
treatment group with immune responses that meet (or exceed) a predefined
immune response.
Serious adverse event (SAE): an AE is serious when it results in: (a)
death, admission to hospital, prolongation of a hospital stay, persistent or
significant disability or incapacity; (b) is otherwise life-threatening; or (c) results
in a congenital abnormality or birth defect. Some NRAs may have additional or
alternative criteria for defining SAEs.
Seroconversion: a predefined increase in serum antibody concentration
or titre. In subjects with no detectable antibody – below the lower limit of
detection (LLOD) – or no quantifiable antibody – below the lower limit of
quantification (LLOQ) – prior to vaccination, seroconversion is usually defined
as achieving a quantifiable antibody level post-vaccination. In subjects with
quantifiable antibody prior to vaccination, seroconversion is commonly defined
by a predefined fold-increase from pre- to post-vaccination.
Sponsor: the individual, company, institution or organization that takes
responsibility for the initiation, management and conduct of a clinical trial. The
sponsor of a clinical trial may not be the entity that applies for a licence to place
the same product on the market or the entity that holds the licence (that is, is
responsible for post-licensing safety reporting) in any one jurisdiction.
Superiority trial: a trial with the primary objective of demonstrating
that a test group is superior to a reference group on the basis of the primary
end-point. In the context of vaccine development the primary end-point may
be a safety parameter (for example, occurrence of a specific type of AE), a
clinical condition (for example, occurrence of a specific infectious disease) or an
immunological parameter (for example, a measure of the immune response to
one or more antigenic components of the vaccine).
Vaccine efficacy: vaccine efficacy measures direct protection (that is,
protection induced by vaccination in the vaccinated population sample). Vaccine
efficacy is most commonly a measure of the proportionate reduction in disease
attack rate (AR) between the control group that did not receive vaccination
against the infectious disease under study (ARU) and the vaccinated (ARV)
group(s). Vaccine efficacy can be calculated from the relative risk (RR) of disease
among the vaccinated group as (ARU − ARV/ARU) × 100 and (1 − RR) × 100.
This estimate may be referred to as absolute vaccine efficacy. Alternatively,
vaccine efficacy may be defined as a measure of the proportionate reduction
in disease AR between a control group that is vaccinated against the infectious
disease under study and the group vaccinated with the candidate vaccine. This
estimate may be referred to as relative vaccine efficacy.
513
Vaccine effectiveness: vaccine effectiveness is an estimate of the

protection conferred by vaccination. It is usually obtained by monitoring
the disease to be prevented by the vaccine during routine use in a specific
population. Vaccine effectiveness measures both direct and indirect protection
(that is, the estimate may in part reflect protection of unvaccinated people
secondary to the effect of use of the vaccine in the vaccinated population).
Vaccine vector: a vaccine vector is a genetically engineered
microorganism (which may be replication competent or incompetent) that
expresses one or more foreign antigen(s) (for example, antigens derived from a
different microorganism).
4. Vaccine clinical development programmes

4.1 General considerations
4.1.1 Consultation with national regulatory authorities
It is strongly recommended that dialogue with the appropriate NRAs occurs at
regular intervals during the pre-licensure clinical development programme to
allow for agreement to be reached on the content and extent of the application
dossier. This is especially important in the following cases:
■■ The clinical programme proposes a novel approach to any aspect of

development for which there is no precedent or guidance available.
■■ The proposed programme conflicts with existing guidance to
which the NRAs involved would usually refer when considering
programme suitability.
■■ Particular difficulties are foreseen in providing evidence to support
an expectation of vaccine efficacy (that is, there is no ICP and a
vaccine efficacy study is not feasible).

■■ There are other special considerations for the total content of the
pre-licensure programme (for example, when different vaccine
constructs are to be used for priming and boosting).
Appropriate NRAs should also be consulted when planning clinical

trials that are intended to support a revision of the prescribing information.
In addition, changes to the manufacturing process of a vaccine before or after
licensure should be discussed with NRAs to establish whether or not clinical
trials are required. When issues of vaccine safety or effectiveness arise in the
post-licensure period, consultation with NRAs is essential to determine any
actions that are needed.
514
Annex 9
4.1.2 Use of independent monitoring committees

The members of an independent monitoring committee should not include
persons who are employed by the sponsor of the clinical trial. The responsibilities
of an independent monitoring committee may include one or more of the
following:
■■ ongoing review of safety data;

■■ oversight of planned interim analyses of safety and/or efficacy, and
recommending to the sponsor that a trial is terminated early in
accordance with predefined stopping rules;
■■ determination of the eligibility of individual subjects for inclusion in
the primary analysis population or other analysis population(s), as
defined in the protocol;
■■ adjudication to determine whether cases of clinically apparent
infections meet the predefined case definition for inclusion in
analyses of efficacy;
■■ adjudication to determine whether reports of AEs meet the criteria
for specified types of AEs and AESIs and/or to determine causality.
The same or different independent monitoring committees may be

appointed to oversee one or more aspects of a clinical trial. Depending on their
role(s), independent monitoring committees may be referred to by specific terms
(for example, Data Monitoring Committee, Safety Data Monitoring Committee
and Independent Data Adjudication Committee).
4.1.3 Registering and reporting clinical trials

Before any clinical trial is initiated (that is, before the first subject receives the
first medical intervention in the trial) the details of the trial must be registered
in a clinical trial registry so that the information is publicly available, free to
access and can be searched. The registry should comply with individual NRA
requirements and, as a minimum, should comply with the WHO internationally
agreed standards.
The entry into the clinical trial registry site should be updated as
necessary to include final enrolment numbers achieved and the date of actual
study completion. A definition of study completion for this purpose should be
included in the protocol. For example, this may be defined as the point in time
when data analyses have been completed to address the major study objectives.
If a clinical trial is terminated prematurely the entry should be updated to reflect
this with a report of the numbers enrolled up to the point of termination.
515
The key outcomes of a clinical trial should be posted in the results section
of the entry in the clinical trial registry and/or posted on a publicly available,
free to access and searchable website (for example, that of the trial sponsor or
principal investigator). It is recommended that posting of these results should
usually occur within 12 months of completion or termination of the study, or in
accordance with the relevant NRA requirements.
Depending on individual NRA requirements, some or all regulatory
submissions may need to include a listing of all completed and ongoing trials
conducted with the vaccine by the sponsor. It is recommended that any trials
that are known to the sponsor (for example, from searching registries or from
publications) that were initiated by entities other than the sponsor (for example,
by a public health body, academic institution or another company that used the
product as a comparator) should be included.
4.2 Pre-licensure clinical development programmes

The main objective of the pre-licensure clinical development programme is
to accumulate adequate data to support licensure. The main elements of the
programme are:
■■ to describe the interaction between the vaccine and the host immune
response (see section 5 below);
■■ to identify safe and effective dose regimens and schedules (see
sections 5 and 6);
■■ to estimate vaccine efficacy by directly measuring efficacy and/or
to provide evidence of vaccine efficacy based on immune responses
(see sections 5 and 6);
■■ to describe the safety profile (see section 7);
■■ to assess co-administration with other vaccines if this is relevant
(see section 5.6.3).

Consideration of the content of pre-licensure clinical development
programmes is undertaken on a product-specific basis. Requirements may differ
depending on the type of vaccine, its manufacturing process, its mechanism of
action, the disease to be prevented and the target population.
4.2.1 Preliminary trials

The clinical programme for new candidate vaccines usually commences with
an exploration of the safety of different amounts of the antigen(s) in each dose
of candidate vaccine formulations, with or without an adjuvant. It is usual that
immune responses to the antigenic components are also explored. These are
commonly referred to as Phase I trials. In most cases the first clinical trials
516
Annex 9
are conducted in healthy adults. It may be appropriate, if feasible, that the first
trials are confined to subjects who have no history of previous exposure to the
organism(s) against which the candidate vaccine is intended to protect.
Further safety and immunogenicity trials that are conducted to build on
the Phase I trial results are commonly referred to as Phase II trials. In most cases
these trials are conducted in subjects who are representative of the intended
target population for the vaccine at the time of licensure. For vaccines intended
for a broad age range it may not be necessary in all instances to apply an age
de-escalation approach (for example, to move from adults to adolescents, then
to children aged 6–12 years, followed by younger children, toddlers and finally
infants) to sequential trials or to groups within trials. For example, if a vaccine
has negligible potential benefit for older children it may be acceptable in some
cases to proceed directly from trials in adults to trials in younger children,
including infants and toddlers.
These trials are usually designed to provide sufficient safety and
immunogenicity data to support the selection of one or more candidate
formulations for evaluation in pivotal trials (that is, to select the amount(s) of
antigenic component(s) and, where applicable, adjuvant in each dose).
4.2.2 Pivotal trials

Pivotal trials are intended to provide robust clinical evidence in support of
licensure. They are commonly referred to as Phase III trials. There may be
exceptional cases in which licensure is based on a Phase II trial that has
been designed to provide robust statistical conclusions. It is usual that the
investigational formulations used in pivotal trials are manufactured using
validated processes and undergo lot release in the same way as intended for the
commercial product.
Pivotal trials may be designed to provide an estimate of vaccine efficacy
or to provide an indication of the ability of the vaccine to prevent clinical
disease on the basis of immunogenicity data (see section 6.1 below). On
occasion, an assessment of a specific safety aspect may be the primary (or a
co‑primary) objective in a pivotal trial (see section 7.2.1 below).
4.3 Post-licensure clinical evaluations

After licensure:
■■ It is essential to monitor vaccine safety in routine use (see section 7
below).
■■ Studies designed to address specific safety issues that were
identified as potential concerns from pre-licensure trials may need
to be conducted.
517
■■ It may be appropriate to conduct studies specifically intended to

estimate vaccine effectiveness (see section 6.4 below).
Sponsors may choose to conduct additional trials that are intended to
extend or to otherwise modify the use of the vaccine through revision of the
prescribing information. In some jurisdictions, conducting one or more trials
after licensure to address specific issues may be a formal requirement.
5. Immunogenicity
Immunogenicity trials are conducted at all stages of pre-licensure vaccine
development and additional trials may be conducted in the post-licensure
period. The evaluation of immune responses relies upon the collection of
adequate specimens at appropriate time intervals and the measurement of
immune parameters most relevant to the vaccine.
Pre-licensure and post-licensure clinical trials commonly evaluate and
compare immune responses between trial groups to address a range of objectives.
In trials that are primarily intended to estimate vaccine efficacy and/or safety,
assessment of the immune response is usually a secondary objective but it is
important that data on immune responses are collected to support analyses of
the relationship between immunogenicity and efficacy, which may lead to the
identification of ICPs.
5.2 Characterization of the immune response

The appropriate range of investigations to be conducted should be discussed
with NRAs. As a general rule, for vaccines that contain microorganisms and
antigens that have not been used previously in human vaccines a thorough
investigation of their interaction with the human immune system should

usually be conducted as part of the overall clinical development programme.
For microorganisms and antigens that are already in licensed vaccines, it is
not usually necessary to repeat these types of investigations but consideration
should be given to conducting at least some trials in certain circumstances (for
example, when a new adjuvant is to be added to known antigens, a different
method of attenuation is used, a different carrier protein is used for antigen
conjugation or an antigen previously obtained by purification from cultures is to
be manufactured using recombinant technology).
In general the clinical development programme should include a
description of the magnitude of the immune response, including an assessment
of functional antibody (for example, antibody that neutralizes viruses or toxins,
or antibody that mediates bactericidal activity or opsonophagocytosis) if this can
518
Annex 9
be measured. Decisions on the range of additional investigations that may be

appropriate should take into account what is known about the immune response
resulting from natural exposure and whether or not this provides partial or
complete protection and, if so, whether it is temporary or lifelong. The range
of investigations chosen should also reflect the characteristics of the infecting
microorganism (for example, whether there are multiple subtypes that cause
human disease) and the content of the vaccine (15).
On a case-by-case basis, other investigations of the immune response
could possibly include some of the following:
■■ assessment of the ability of the vaccine to elicit a T-cell-dependent
primary immune response, with induction of immune memory
(that is, priming of the immune system) giving rise to anamnestic
responses to: (a) natural exposure following vaccination; (b) further
doses of the same vaccine; and/or (c) further doses of a vaccine
that contains closely related but non-identical microorganisms or
antigens (that is, cross-priming);
■■ assessment of the specificity and cross-reactivity of the immune
response;
■■ assessment of changes in antibody avidity with sequential doses,
which may be useful when investigating priming;
■■ evaluation of factors that could influence the immune responses,
such as the effect of maternal antibody on the infant immune
response to some antigens, pre-existing immunity to the same or
very similar organisms, and natural or vaccine-elicited antibody
against a live viral vector.
5.3 Measuring the immune response

5.3.1 Collection of specimens
Immune responses to vaccination are routinely measured in serum (humoral
immune responses) and blood (cellular immune responses). For some vaccines it
may be of interest to explore immune responses in other body fluids relevant to
the site at which the target microorganism infects and/or replicates (for example,
in nasal washes or cervical mucus), especially if it is known or suspected that the
systemic immune response does not show a strong correlation with protective
efficacy for the type of vaccine under trial (for example, intranasal vaccination
against influenza). Nevertheless, specimens other than sera have not to date
provided data that have been pivotal in regulatory decision-making processes
and have not resulted in the identification of ICPs. Therefore the rest of this
section focuses on the collection of blood samples.
519
Pre-vaccination samples should be collected from all subjects in early

preliminary immunogenicity trials, after which it may be justifiable to omit
these samples or to obtain them from subsets (for example, if antibody is rarely
detectable or quantifiable prior to vaccination in the target population). Pre-
vaccination sampling remains essential if it is expected that the target population
will have some degree of pre-existing immunity due to natural exposure and/
or vaccination history, since the assessment of the immune response will need
to take into account seroconversion rates and increments in geometric mean
titres (GMTs) or geometric mean concentrations (GMCs) from pre- to post-
vaccination. Pre-vaccination sampling is also necessary if it is known or suspected
that pre-existing immune status may have an impact on the magnitude of the
immune response to vaccination that is positive (for example, because pre-
existing antibody reflects past priming) or negative (for example, due to maternal
antibody interfering with primary vaccination with certain antigens in infants).
The timing of post-vaccination sampling should be based on what is
already known about the peak immune response after the first and, if applicable,
sequential doses (for example, for vaccines that elicit priming, the rise in
antibody after a booster dose is usually much more rapid than the rise after earlier
doses). For antigens not previously used in human vaccines, sampling times may
be based on nonclinical data and then adjusted when data that are specific to the
antigen(s) under trial have been generated. As information is accumulated, the
number and volume of samples taken from individual subjects may be reduced
to the minimum considered necessary to meet the trial objectives.
5.3.2 Immunological parameters

Immunological parameters are measures that describe the humoral immune
response (for example, antibody concentrations or antibody titres, depending
on the assay output) or the cell-mediated immune response (for example,
percentages of sensitized T-cells). To date, immunological parameters other than
those that measure the humoral immune response have not played a pivotal
or major role in vaccine licensure, so the focus is usually on determination of
antibody levels.
■■ For known microorganisms or antigens in a candidate vaccine the
range of parameters to be measured in clinical trials is usually
selected on the basis of prior experience and whether or not there is
an established ICP.
■■ For microorganisms or antigens not previously included in human
vaccines the selection of parameters to be measured should take into
account what is known about natural immunity. For some infectious
diseases the nature of the immune response to infection in animal
models may also be useful for parameter selection.
520
Annex 9
5.3.2.1 Humoral immune response

The humoral immune response is assessed from the post-vaccination appearance
of, or increase after vaccination in, antibody directed at specific microorganisms
or antigens in the vaccine.
■■ If data are available, most weight is usually placed on functional
antibody responses – for example, serum bactericidal antibody
(SBA), toxin- or virus-neutralizing antibody or opsonophagocytic
antibody (OPA). In some cases an appropriate assay for functional
antibody may not be available (for example, for typhoid vaccines
based on the Vi polysaccharide) or the only available assay may
have low feasibility for application to large numbers of samples (for
example, because it is very labour-intensive or requires high-level
biocontainment facilities).
■■ Alternatively, or in addition to the determination of functional
antibody, the immune response may be assessed by measuring total
antibody – for example, total immunoglobulin G (IgG) measured by
enzyme-linked immunosorbent assay (ELISA) that binds to selected
antigens (or, on occasion, to specific epitopes). Only a proportion of
the total antibody detected may be functional.
The following should be taken into consideration when deciding how
to measure the humoral immune response:
■■ If a correlation has already been established between total and
functional antibody responses to a specific microorganism
or antigen it may be acceptable to measure only total IgG in
further trials (for example, antibody to tetanus toxin). However,
determination of functional immune responses might be important
for specific age groups or target populations where it is known or
suspected that the binding and functional capacity of the antibodies
elicited differs (for example, pneumococcal conjugate vaccines in
older people).
■■ For antigens for which there is an established ICP it may suffice to
measure only the relevant functional antibody (for example, SBA for
meningococcal vaccines) or total IgG (for example, for antibody to
tetanus toxin) response.
■■ If the ICP is based on total IgG there may be instances where there
is still merit in measuring functional antibody (for example, for
antibody to diphtheria toxin for which a microneutralization assay
is available).
521
■■ If there is no ICP the functional antibody response should be

measured if this is feasible.
■■ Occasionally there may be more than one immunological parameter
that can measure functional antibody but one is considered to be a
more definitive measure than the other (for example, neutralizing
antibody to influenza virus versus antibody that inhibits
haemagglutination). In this case the more definitive parameter may
be determined, at least in a subset.
■■ For some vaccines against certain viruses there is a possibility that
some of the total antibody detected has no protective effect (for
example, is non-neutralizing) but could enhance cellular infection
by wild-type virus and result in an increased risk of severe disease
after vaccination (for example, this may apply to dengue vaccines).
To assess this possibility, the routine measurement of total antibody
to assess the humoral immune response to vaccination should be
supported by other detailed investigations.
5.3.2.2 Cell-mediated immune response

For some types of infectious disease (such as tuberculosis) assessment of the
cell-mediated immune response may have a role to play in the assessment of
the interaction between the vaccine and the human immune system. In other
cases, evaluation of the cellular immune response may serve to support findings
based on the humoral immune response (for example, when assessing the benefit
of adding an adjuvant or when evaluating the degree of cross-priming elicited by
a vaccine).
The cell-mediated immune response is most commonly assessed by
detecting and quantifying sensitized T-cells in blood from trial subjects. These
investigations may also serve to characterize the predominant cytokines released

and to detect differences in sensitization between T-cell subpopulations. Several
methods may be used. These are typically based on measuring the production of
a range of cytokines following in vitro stimulation of T-cells with individual or
pooled antigens.
The results may provide useful comparisons between treatment groups
within any one study (for example, they could describe the effect, if any, of
an adjuvant). If there are marked discrepancies in the patterns of responses
observed between cell-mediated and humoral responses (for example, if adding
an adjuvant has a major effect on antibody levels but does not increase the
percentages of sensitized cells in one or more T-cell subsets) the findings should
be carefully considered and discussed.
522
Annex 9
5.3.3 Assays
Assays of functional or total antibody that are used to report immune responses
to vaccination (whether to the candidate vaccine or to co-administered vaccines)
in trials intended to support licensure (that is, in pivotal trials) should be
acceptable to the relevant NRAs. They may be:
■■ commercially available assays specifically designed and intended
for quantification of antibody (that is, assays that have undergone a
robust regulatory review);
■■ assays that are not commercially available but have been validated
according to principles similar to those recommended for
quantitative lot release assays in the ICH Q2 (R1) document
Validation of analytical procedures: text and methodology (16);
■■ assays that are not commercially available but have been shown to be
comparable to a reference assay (for example, to an assay established
in a WHO reference laboratory or to an assay that is established in a
recognized public health laboratory and has been used previously to
support clinical trials that were pivotal for licensure).
It is expected that, if these exist, WHO International Standards and
Reference Reagents will be used in assay runs. Any omission of their use should
be adequately justified.
Clinical trial protocols should specify which assays will be used.
Clinical trial reports should include a summary of the assay methodology and
its commercial or other validation status. For assays that are not commercially
available any available validation reports should be provided.
The same assays should preferably be used in the same laboratories
throughout the clinical development programme (including pre- and post-
licensure trials) for an individual vaccine. It is also preferable that each assay
(whether it measures the response to the candidate vaccine or to a concomitant
vaccine) is run by one central laboratory. If this is not possible (for example,
because different laboratories have to be used, assays change over time, or a
switch is made to an improved and/or more suitable assay) the new and original
assays should be shown to give the same result or interpretation, or the impact
of any differences should be evaluated and the use of a new assay justified. It
is recommended that, as a minimum, a selection of stored sera (for example,
covering a range of low to high results when using the previous assay) should be
re-run using the previous and new assays in parallel. The number of sera retested
should be sufficient to support a statistical assessment of assay comparability
and/or reproducibility.
523
The microorganisms (for example, in assays of SBA, OPA and virus

neutralization) and antigens (for example, in ELISAs and for in vitro stimulation
of sensitized T-cells) used in the assay may affect both the result and the
interpretation of the result. For example:
■■ It is important to use purified antigen to avoid the possibility that
the assay detects and measures antibody to any extraneous antigenic
substances that may be in the vaccine.
■■ For vaccines that contain antigens from multiple strains of the same
pathogen (for example, multiple bacterial capsular types) the assays
selected (whether separate or multiplex) should determine the
immune response to each antigen.
■■ Although it is usually acceptable to conduct routine testing using the
same microorganisms or antigens as those present in the vaccine,
it may be very informative to perform additional testing, at least
in subsets of samples, using circulating wild-type organisms or
antigens derived from them in the assay. It is not expected that
these additional assays will necessarily be validated since they are
exploratory in nature. The results of additional testing can provide an
indication as to whether the results of routine testing could represent
an overestimate of the immune response to circulating strains.
This additional testing can also provide an assessment of the cross-
reactivity of the immune responses elicited by the vaccine to other
organisms of the same genus or species (for example, to different
flaviviruses, different clades of influenza virus or different HPV
types) and can guide decisions on the need to replace or add strains
or antigens in a vaccine to improve or maintain its protective effect.
Identification and use of immune correlates of protection

5.4
5.4.1 Immune correlates of protection and their uses
All established ICPs are based on humoral immune response parameters that
measure functional or total IgG antibody. Some examples of well-established
ICPs include those for antibody to diphtheria and tetanus toxoids, polioviruses,
hepatitis B virus and Haemophilus influenzae type b capsular polysaccharide (17).
In most cases established ICPs have been shown to correlate with prevention
of clinically apparent infectious disease, but for some pathogens the ICP \
correlates with prevention of documented infection (for example, hepatitis A and
hepatitis B).
Sections 5.5.2 and 5.5.3 below consider trial end-points and the approach
to analysis and interpretation of immunogenicity data in the presence or absence
of an ICP.
524
Annex 9
5.4.2 Establishing an ICP

Documentation of the immune response to natural infection, the duration
of protection after clinically apparent infection (that is, whether natural
protection is lifelong (solid immunity), temporary or absent) and the specificity
of protection (that is, whether the individual is protected only against specific
subtypes of a microorganism) should be taken into account when attempting to
establish an ICP from clinical data. For example, to date, widely accepted clinical
ICPs have been established on the basis of one or more of the following:
■■ serosurveillance and disease prevalence in specific populations;
■■ passive protection using antibody derived from immune humans or
manufactured using recombinant technology;
■■ efficacy trials;
■■ effectiveness trials;
■■ investigation of vaccine failure in immunosuppressed populations.
In the majority of cases clinical ICPs have been determined from
vaccine efficacy trials that were initiated pre-licensure, often with long-term
follow-up of subjects that extended into the post-licensure period. Efficacy trial
protocols should plan to collect sufficient information to allow for analyses of
the relationship between immune parameters and protection against clinically
apparent disease. At the minimum this requires the collection of post-
vaccination samples from all, or from a substantial subset of, the vaccinated
and control groups. Serial collection of samples over the longer term, along
with follow-up surveillance for vaccine breakthrough cases, has also served to
support identification of ICPs.
To investigate the predictive capacity of a putative ICP, protocols
should predefine the assessments to be applied to all cases of the disease to be
prevented that occur in the vaccinated and control groups. These assessments
should include investigation of the immune status of subjects as well as
microbiological studies with the infecting microorganisms whenever these have
been recovered. For breakthrough cases from which both post-vaccination sera
and organisms have been recovered it is recommended that, whenever feasible,
functional antibody (or, if not possible, total antibody) should be determined
for individuals against their own pathogen. An exploration of vaccine-elicited
cell-mediated responses in individuals against their own pathogen may also be
useful and, for some types of infectious disease (such as tuberculosis), may be
very important for further understanding vaccine-associated protection. These
data may be very important for investigating the broad applicability of the ICP,
depending on host and organism factors.
A single clinical ICP identified from a vaccine efficacy trial in a defined
population may not necessarily be applicable to other vaccine constructs
525
intended to prevent the same infectious disease. In addition, an ICP may not
be applicable to other populations and disease settings. For example, putative
ICPs have sometimes differed between populations of different ethnicities with
variable natural exposure histories for subtypes of a single microorganism. Thus,
the reliance that is placed on a clinical ICP, even if regarded as well supported
by the evidence, should take into account details of the efficacy trials from which
it was derived.
Clinical ICPs have also been derived from, or further supported by,
data collected during use of a vaccine to control an outbreak and from analyses
of effectiveness data. The methods used to derive ICPs from these types of
data have been very variable. The estimates may in part reflect the type of
immunization programme put in place and the extent to which the protection
of individuals relies on herd immunity rather than the initial and persisting
immune response in the individual. Therefore the wider applicability of ICPs
derived from interventional or routine use should be viewed in the light of how
and in what setting the estimates were obtained.
If it is not possible to derive a clinical ICP the interpretation of the
human immune response data may take into account what is known about
immunological parameters that correlate with protection in relevant animal
models and any nonclinical ICPs that have been identified (for example, from
trials that assess passive protection and active immunization). This approach
may be the only option available for interpreting immune responses to some
new candidate vaccines. Nevertheless, ICPs derived wholly from nonclinical
data should be viewed with caution and attempts should be made to obtain a
clinical ICP whenever the opportunity arises (for example, when the vaccine is
used in the context of an outbreak).
If conducted, human challenge trials may also provide preliminary
evidence supporting an ICP. If a human challenge trial suggests a correlation
between a specific immunological parameter and protection, this may be further
investigated during the clinical development programme.
5.5 Immunogenicity trials

5.5.1 Objectives
The objectives of immunogenicity trials include, but are not limited to, the
following:
■■ to select vaccine formulations and posologies (including primary
and booster doses) (see section 5.6.1 below);
■■ to compare immune responses documented in a specific population
and, using one vaccine formulation and posology, to immune
responses to the same vaccine when used in other settings (for
example, different populations) or with alternative posologies, or to
526
Annex 9
a different vaccine intended to protect against the same infectious

disease(s) (see section 5.6.2);
■■ to support co-administration with other vaccines (see section 5.6.3);
■■ to support maternal immunization (see section 5.6.4);
■■ to support major changes to the manufacturing process (see
section 5.6.5);
■■ to assess lot-to-lot consistency (7) (see section 5.6.6).
5.5.2 General considerations for trial designs

Immunogenicity trials are almost without exception comparative trials. For
candidate vaccines containing antigens for which there are well-established ICPs
that can be applied to interpret the results sponsors may sometimes question
the value of including a comparative arm. Nevertheless, there is great value in
conducting a randomized controlled trial. For example, the inclusion of a control
group that receives a licensed vaccine provides assurance of the adequacy of the
trial procedures and methods, including the assays, and facilitates interpretation
of data in circumstances in which unexpected results (for example, low immune
response to one or more antigens, high rates of specific AEs or unexpected AEs)
are observed.
Comparative trials include those in which all subjects receive the same
vaccine formulation but there are differences between groups in terms of how
or to whom the vaccine is administered (for example, using a different dose or
dose interval, or administering the vaccine to different age groups) as well as
trials in which one or more group(s) receive an alternative treatment, which may
be placebo and/or another licensed vaccine.
The design of comparative immunogenicity trials is driven by the
characteristics of the vaccine, the trial objectives, the stage of clinical
development, the trial population, the availability and acceptability of suitable
comparators, and what is known about immune parameters that correlate with
protection (including whether or not there is an established ICP).
In comparative immunogenicity trials, subjects should be randomized
to one of the trial groups at enrolment. This also applies to trials that enrol
sequential cohorts of subjects (as in ascending dose trials in which at least some
subjects are assigned to receive placebo or another vaccine). In some cases it may
be appropriate that subjects who meet certain criteria (for example, completed
all assigned doses in the initial part of the trial) are re-randomized at a later stage
of the trial to receive a further dose of a test or control treatment.
In all comparative trials the assays should be performed by laboratory
staff unaware of the treatment assignment. Whenever possible, comparative
immunogenicity trials should be of double-blind design. If the vaccines to be
compared are visually distinguishable it is preferable that designated individuals
527
at each trial site who are not otherwise involved in the trial should administer
the products. If this is not feasible, or if the vaccines to be compared are
given by different routes or according to different schedules, attempts should
be made to maintain blinding of the trial site staff conducting the study visits
and assessments.
In trials intended to provide only descriptive analyses of the
immunogenicity data the trial sample size is usually based on considerations
of feasibility and collection of sufficient safety data to support the design of
sequential trials. Trials that aim to assess superiority or non-inferiority between
vaccine groups should be sized according to the intended power and the
predefined margins. It is recommended that protocols and statistical analysis
plans for each trial should be developed in conjunction with an appropriately
experienced statistician.
5.5.2.1 End-points
The choice of the primary trial end-point and the range of other end-points for
immunogenicity trials should take into account sections 5.2, 5.3 and 5.4 above.
Protocols should predefine the primary, co-primary, secondary and any other
end-points (which may be designated tertiary or exploratory). Co-primary end-
points may be appropriate in some cases, namely:
■■ The vaccine is intended to protect against multiple subtypes of the
same microorganism (for example, HPV vaccines or pneumococcal
conjugate vaccines).
■■ The vaccine contains multiple microorganisms (such as measles,
mumps and rubella vaccine) or multiple antigens (such as
combination vaccines used for the primary immunization series
in infants).
The following should be taken into consideration when selecting the

primary end-point(s) following primary vaccination:
■■ When an ICP has been established the primary end-point is usually
the percentage of subjects that achieves an antibody level at or above
the ICP, which is sometimes referred to as the seroprotection rate.
■■ When there is no established ICP the primary end-point or the
co-primary end-points is/are usually based on a measure of the
humoral immune response.
(a) In some instances there may be evidence to support the
application of a threshold value (that is, the primary end-point
may be the percentage of subjects that achieves antibody levels
at or above the threshold value).
528
Annex 9
(b) If there is no threshold value that can be applied it may be

appropriate to base the primary end-point on the seroconversion
rate or on some other definition of the magnitude of the
immune response that differentiates responders from non-
responders. Comparisons of post-vaccination seropositivity rates
may also be informative if pre-vaccination rates are very low.
An anamnestic (memory) immune response is anticipated following
administration of a vaccine to subjects who are already primed (by natural
exposure or prior vaccination) against one or more microorganisms or antigens
in the vaccine. Thus the seroprotection, seroconversion (fold-rise from pre-
boost to post-boost) and seropositivity rates after the booster dose are likely to
be very high. In these cases, and in other situations in which post-vaccination
seroprotection and/or seroconversion rates are expected to be very high (that is,
the vaccine is very immunogenic) the most sensitive immunological parameter
for detecting differences between groups may be the GMC or GMT.
After primary vaccination and after any additional doses the results
for all measured immunological parameters should be presented in the clinical
trial report.
5.5.2.2 Trials designed to demonstrate superiority

Trials may assess whether a specific candidate vaccine formulation elicits
superior immune responses compared to no vaccination against the disease to
be prevented. In some cases trials may also assess whether immune responses
elicited by a specific formulation of a candidate vaccine are superior to those
elicited by other formulations.
An assessment of superiority may also be applicable when an adjuvant
is proposed for inclusion in the vaccine (for example, to demonstrate that the
immune response to at least one of the antigenic components in an adjuvanted
formulation is superior to the response in the absence of the adjuvant).
Protocols should predefine the magnitude of the difference between
vaccine groups or between vaccine and control groups that will be regarded as
evidence of superiority. This difference should be defined in such a way that it
provides some evidence of a potential clinical advantage.
5.5.2.3 Trials designed to demonstrate non-inferiority

Most comparative immunogenicity trials are intended to show that the test
vaccinated groups achieve comparable immune responses to the selected
reference groups. If these trials are intended to be pivotal they should be designed
and powered to demonstrate non-inferiority using a predefined and justifiable
non-inferiority margin.
529
Factors to consider with regard to the stringency of the non-inferiority

margin include the clinical relevance of the end-point, seriousness of the disease
to be prevented, vulnerability of the target population, availability of a well-
established ICP and the performance characteristics of the assay(s). A more
stringent margin may be appropriate when the vaccine is intended to prevent
severe or life-threatening diseases and/or will be used in particularly vulnerable
populations (for example, infants and pregnant women). A more stringent
margin could also be considered when there is potential for a downward drift
in immunogenicity such as that which could occur when a new candidate
vaccine can be compared only with vaccines that were approved on the basis
of non-inferiority trials. In contrast, if a new candidate vaccine is known to
offer substantial benefits in terms of safety or improved coverage then margins
that are less stringent may be considered. As a result of such considerations it is
possible that different non-inferiority margins may be considered appropriate in
different settings.
When it is proposed to demonstrate non-inferiority between vaccine
groups based on GMT or GMC ratios for antibody titres or concentrations it
is suggested that the lower bound of the 95% confidence interval around the
ratio (test versus reference vaccine) should not fall below 0.67. Under certain
circumstances NRAs may consider allowing a lower bound (for example, 0.5) or
alternative criteria. The selection of a criterion should take into account whether
or not an ICP has been identified. In addition, any marked separations between
the reverse cumulative distributions of antibody titres or concentrations should
be discussed in terms of potential clinical implications, including those which
occur at the lower or upper ends of the curves.
5.5.3 Analysis and interpretation

A statistical analysis plan should be finalized before closing the trial database
and unblinding treatment assignments (if these were blinded). This should
include any planned interim analyses, which should be adequately addressed in

terms of purpose, timing and any statistical adjustments required.
The immunogenicity data from all subjects with at least one result for any
immunological parameter measured in the trial should be included in the clinical
trial report. The analysis of the immune response based on any one parameter
is commonly restricted to all subjects with a pre-vaccination measurement
(if this is to be obtained from all subjects) and at least one post-vaccination
measurement. Protocols may also restrict the primary analysis population to
subjects with pre- and post-vaccination results, or to those with post-vaccination
results who received all the assigned doses within predefined windows of the
intended schedule and had no other major protocol violations. Other analysis
populations of interest may be predefined in accordance with the primary or
secondary objectives (for example, age subgroups or pre-vaccination serostatus
530
Annex 9
subgroups). Whatever the predefined primary analysis population, all available

immunogenicity data should be presented in the clinical trial report.
If a trial fails to meet the predefined criteria for superiority and/or non-
inferiority with respect to any of the antigenic components, the possible reasons
for the result and the clinical implications of it should be carefully considered
before proceeding with clinical development or licensure. The considerations
may take into account: (a) the basis for setting the predefined criteria (for
example, does failure to meet the criteria strongly imply that lower efficacy may
result?); (b) the comparisons made for all other immune parameters measured
(for example, were criteria not met for only one or several of many antigenic
components of the vaccine?); (c) any differences in composition between the
test and comparator vaccines that could explain the result; (d) the severity of
the disease(s) to be prevented; and (e) the overall anticipated benefits of the
vaccine, including its safety profile (17). Section 5.6 below provides further
examples and issues for consideration.
If additional analyses of the data that were not pre-specified in the
protocol and/or the statistical analysis plan (that is, post hoc analyses) are
conducted, they should usually be viewed with caution.
5.6 Specific uses of immunogenicity trials

5.6.1 Selection of formulation and posology
The vaccine formulation is determined by the numbers of microorganisms
or amounts of antigens and, if applicable, the amount of adjuvant that is to be
delivered in each dose, as well as by the route of administration.
The vaccine posology for a specific route of administration includes:
■■ the antigen content (as for formulation) and volume delivered
per dose;
■■ the dose regimen (number of doses to be given in the primary series
and, if applicable, after the primary series);
■■ the dose schedule (dose intervals within the primary series and
between the primary series and any further doses).
The posology for any one vaccine may vary between target populations
(for example, between age groups or according to prior vaccination history) in
one or more aspects (content, regimen or schedule).
The following sections outline the immunogenicity data that are usually
generated to support vaccine formulation and posology, and to assess the need
for, and immune response to, additional doses of the vaccine after completion of
the primary series. Section 7 below then addresses the importance of the safety
profile when selecting vaccine formulations and posologies.
531
5.6.1.1 Selecting the formulation and posology

The vaccine formulation and posology should be supported by safety and
immunogenicity data, with or without efficacy data, collected throughout the
pre-licensure clinical development programme. At the time of licensure the
data should at least support the formulation and posology for the primary series,
which may consist of one or more doses.
Depending on the intended formulation of the new candidate vaccine,
the following considerations may apply:
1. When a new candidate vaccine contains any microorganisms or
antigens not previously used in human vaccines, with or without
others already used in human vaccines, the preliminary trials may
explore the immune responses to different amounts of each of the
new microorganisms or antigens when given alone to non-immune
healthy adult subjects. These trials can be used to describe the
dose–response curve and may indicate a plateau for the immune
responses above a certain dose level. The next trials usually evaluate
immune responses to further doses at various dose intervals in
order to evaluate the kinetics of the immune response and any
increment in immune response achieved by further doses. The
transition from trials in healthy adults to trials in subjects in the
target age range at the time of licensure should occur as soon as
this can be supported, taking into account the safety profile.
However, evaluating the immune response to each of
the new microorganisms or antigens alone may not be a feasible
undertaking. For example, if the vaccine construct is manufactured
in such a way that production of individual antigens is not feasible
then evaluation of the appropriate vaccine dose may be based solely
on studies with the entire construct. Another example concerns
vaccines intended to protect against multiple subtypes of an

organism. In this case, the use of microorganisms or antigens that
could be regarded as broadly representative in the first trials may
provide some idea of the likely response to other subtypes. Further
trials may then explore formulations that contain increasing
numbers of the subtypes with the objective of assessing the effect
on the immune response of combining them into a single product.
2. For new candidate vaccines that contain known antigenic
components not previously combined in a single vaccine, the
preliminary trials are usually conducted in subjects within the age
ranges approved for licensed vaccines that contain some or all of
the same antigenic components. The aim is to demonstrate non-
532
Annex 9
inferiority of immune responses to each of the intended antigenic

components when combined in a candidate formulation compared
with co-administration of licensed vaccines that together provide
all of the same antigenic components. The same approach applies
whenever the antigenic components are not combined into a single
formulation but the contents of more than one product have to be
mixed immediately before administration to avoid a detrimental
physicochemical interaction.
3. For new candidate vaccines that contain both known and one or
more new antigenic components the preliminary trials may aim to
demonstrate non-inferiority of immune responses to each of the
known antigenic components when combined into a candidate
formulation compared with the separate administrations of known
and new antigenic components. It may also be informative to include
a control group that receives co-administration of the known and
new antigenic components. The exact trial design will depend
upon the availability of a single licensed vaccine that contains the
known antigenic components and whether more than one licensed
vaccine has to be given.
4. For vaccine formulations to which an adjuvant is to be added
there should be adequate data already available (known adjuvants)
or data should be generated (new adjuvants or when using any
adjuvant with a new antigenic component) to describe the effect
of the adjuvant on the immune responses. Some, or a major part,
of the evidence supporting the addition of an adjuvant may come
from nonclinical studies. The addition of an adjuvant, which
may or may not elicit superior immune responses to one or more
antigens, should not have a potentially detrimental effect on the
responses to any antigenic components. Addition of an adjuvant
may allow for the use of a much lower dose of an antigenic
component to achieve the desired level of immune response, and
it may also broaden the immune response (for example, it may
result in higher immune responses to antigens closely related to
those in the vaccine). Trials should evaluate a sufficient range of
combinations of antigenic components and adjuvants to support
the final selected formulation (that is, the ratio of adjuvant to
antigenic components).
5. The total data generated should be explored to identify the criteria
that should be applied to the release and stability specifications, and
to the determination of an appropriate shelf-life for the vaccine.
This is usually of particular importance for vaccines that contain
533
live microorganisms. Depending on data already generated, it may

be necessary to conduct additional trials with formulations known
to contain a range of microorganism numbers or antigen doses in
order to identify appropriate limits at the end of the shelf-life.
6. Comparative immunogenicity trials may be needed to determine
schedules that are appropriate for specific target populations, taking
into account the urgency to achieve protective immunity (that is,
trials based on diseases to be prevented and their epidemiology).
The data generated across all the trials should determine the
minimum period that should elapse between doses, as well as the
effects of delaying doses to support acceptable windows around
scheduled doses. Additionally, for some vaccines it may be useful
to explore the shortest time frame within which doses may be
completed without a detrimental effect on the final immune
response (for example, for vaccines for travellers who may need to
depart at short notice or for vaccines intended to provide post-
exposure prophylaxis).
Assessment of the effects of dose interval and the total
time taken to complete the primary series is a particular issue for
vaccines intended for use in infants as there is a very wide range
of schedules in use in different countries (for example, 3-dose
schedules include 6–10–14 weeks and 2–4–6 months). In general,
experience indicates that the magnitude of the post-primary series
immune responses broadly correlates with the age of infants at the
time of the final dose.
7. All data generated in accordance with points 1–6 above should
be taken into account when selecting the final formulation
and posology or posologies. The selection process is more
straightforward if there are established ICPs that can be applied to

the interpretation of the results for at least some of the antigenic
components. In the absence of an ICP the posology may be
selected on the basis of consideration of any plateau effects that are
observed and on the safety profile of various doses and regimens.
It is not unusual for the final selected formulation and
posology to represent, at least to some extent, a compromise
between immunogenicity and safety or, for combination vaccines,
a compromise between the potential benefits of a vaccine that
can protect against multiple types of infectious disease and some
negative effects on immune response that may occur. These negative
effects may result from a physicochemical interaction between
vaccine components and/or a negative immune interference
534
Annex 9
effect of some antigenic components. Such negative effects may

be accompanied by enhanced immune responses to other vaccine
components. The rationale for the final selection should be
carefully discussed in the application dossier.
5.6.1.2 Amending or adding posologies

Clinical trials may be considered necessary to address one or more of the
following situations:
■■ Change to the number of doses or dose intervals – in this case the
control group could be vaccinated using the licensed posology and
the trial could be conducted in a population for which the vaccine is
already licensed.
■■ Use of the licensed posology in a new population (for example, in
subjects who are younger or older than the currently licensed age
group, or in subjects with specific underlying conditions, such as
immunosuppression) – in this case the trial could compare use of
the licensed posology in the new target population with use in the
population for which the vaccine is already licensed.
■■ Use of an alternative to the licensed posology in a new population
– in this case the trial could compare the alternative posology
administered to the new population with the licensed posology in
the population for which the vaccine is already licensed.
■■ Support for alternative routes of administration for the licensed
formulation (for example, adding subcutaneous or intradermal
injection to intramuscular use).
Post-licensure clinical trials may also be conducted to support changes in
formulation. Formulation changes other than adding or removing a preservative
or removing thiomersal from the manufacturing process may or may not result
in a modified product that is considered to be a new candidate vaccine from
a regulatory standpoint (that is, would require a new application dossier and
adequate trials to support separate licensure).
5.6.1.3 Post-primary doses

5.6.1.3.1 Need for post-primary doses
The need to administer additional doses, and the timing of these doses, may be
determined before and/or after first licensure.
There may be experience with other similar vaccines indicating that
additional doses of a new candidate vaccine will be needed after completion of
the primary series (for example, after infant immunization with H. influenzae
535
type b and Neisseria meningitidis group C vaccines). In such cases the clinical
development programme should usually incorporate an assessment of immune
responses to a post-primary dose.
If it is not known whether post-primary doses of a new candidate
vaccine will be needed to maintain protection, it is preferable that this should
be determined from long-term follow-up of subjects who were enrolled in
efficacy trials and/or from post-licensure effectiveness studies. Although the
long-term monitoring of antibody persistence is important, these data alone
cannot determine if another dose is needed unless there is evidence, or a strong
reason to expect, that failure to maintain circulating antibody above a certain
level (for example, above the ICP if there is one) is associated with a risk of
breakthrough disease.
If it is unclear whether additional doses are needed it is prudent to
plan to obtain data on the immune response to doses administered at different
intervals after the last dose of the primary series so that such data are available
should it become clear that a further dose is required.
5.6.1.3.2 Assessment of prior priming

It is not always necessary to assess whether or not a vaccine elicits a T-cell-
dependent immune response that results in priming of the immune system and
an anamnestic (memory) response to further doses. However, for some new
candidate vaccines (for example, polysaccharide-protein conjugate vaccines
in which the polysaccharide and/or conjugate protein have not previously
been included in a licensed vaccine) there may be considerable interest in
understanding the ability of the vaccine to prime the immune system.
When assessing the immune response to additional doses and
determining whether or not the primary series elicited immune memory, the
following should be taken into account:
■■ Trials in which additional doses are administered may be extension

phases of primary series trials or new trials in subjects with
documented vaccine histories.
■■ When assessing whether the primary series elicited immune
memory the optimal design is to compare subjects who previously
completed a full primary series of the candidate vaccine with a
control group consisting of subjects not previously vaccinated.
Control subjects should be matched for age and for any host or
demographic factors that might have an impact on their immune
response (for example, they should be resident in similar areas so
that any natural exposure is likely to be similar).
■■ If the new candidate vaccine elicited immune memory in the primary
series the immune response to the additional (that is, booster) dose
536
Annex 9
should usually be superior (on the basis of comparisons of the GMCs

or GMTs of antibody) to that observed in individuals who have not
been vaccinated against the disease to be prevented. The percentages
that achieve seropositivity or seroprotection (as defined) may not
differ between the two groups if a single dose of the vaccine is highly
immunogenic even in unprimed individuals.
■■ The immune response to the additional dose in primed and
unprimed subjects may also be differentiated on the basis of the
rapidity of the rise in antibody levels (faster in primed) and in terms
of antibody avidity (greater in primed). Note that not all primed
individuals (whether priming results from natural exposure or from
previous vaccination) have detectable humoral immunity against the
relevant organism or the toxin that causes clinical disease.
■■ If the immune response as measured by geometric mean antibody
concentrations or titres in the vaccine-primed group is not superior
to that in controls this does not always mean that the primary
series did not elicit immune memory. For example, the immune
response in the vaccinated group may not be superior to the
immune response in the control group when natural priming has
occurred in a substantial proportion of subjects not previously
vaccinated against the disease to be prevented – in which case the
rapidity of response and measurements of avidity may also not be
distinguishable between groups. If natural priming has occurred
it may or may not be detectable from pre-vaccination antibody
levels in the control group.
■■ If an immune memory response is elicited in the primary series
it may be possible to achieve a robust anamnestic response using
a much lower dose of an antigenic component compared to the
primary series. A lower boosting dose may also provide a better
safety profile (for example, as occurs with diphtheria toxoid).
■■ For polysaccharide-protein conjugate vaccines that elicit immune
memory it may be informative to compare boosting with the same
type of conjugate used for priming with an alternative conjugate (for
example, to prime with a tetanus toxoid conjugate and boost with a
CRM197 conjugate and vice versa).
■■ It may also be informative to assess the ability of a candidate
vaccine to achieve cross-priming by using heterologous antigenic
components for priming and boosting. This may be assessed by
comparing boosting with the same vaccine used to prime with
administration of a formulation (which may be a licensed vaccine
537
or an unlicensed product manufactured specifically for the trial)

containing a different microorganism or antigen that is known to be
closely related but not identical to that in the vaccine (for example,
material derived from an influenza virus of a different clade).
■■ Elicitation of an immune memory response to a vector for an antigen
after the first dose(s) may sometimes interfere with or wholly prevent
the immune response to the antigen after subsequent doses (for
example, this may be observed when using certain adenoviruses
capable of infecting humans as live viral vectors). It is essential
to understand whether or not this occurs since it may necessitate
the use of a different vector for the antigen or an entirely different
vaccine construct to deliver subsequent doses.
■■ Some antigens elicit immune hyporesponsiveness to further doses.
The best known examples are some of the unconjugated
meningococcal and pneumococcal polysaccharides (18, 19). In
the past these were sometimes administered to assess whether
corresponding conjugated polysaccharides had elicited immune
memory in the primary series, based on the premise that this would
better mimic the immune response to natural exposure compared
to administration of a further dose of the conjugate. This practice
is not recommended since it is possible that a dose of unconjugated
polysaccharide could result in blunted immune responses to
further doses of the conjugate.
■■ Studies of cell-mediated immunity may provide supportive evidence
that the primary series elicited immune memory and may be
particularly useful for assessing cross-priming.
5.6.2 Using immunogenicity data to predict efficacy

5.6.2.1 Bridging to efficacy data
Immunogenicity data may be used to provide evidence of efficacy when:
■■ there is a well-established ICP that can be used to interpret the
immune responses to a specific antigenic component;
■■ it is possible to use immune responses to bridge to estimates of
vaccine efficacy obtained from prior well-designed clinical trials
(that is, to conduct bridging trials).
The following two main situations should be considered when using
immunogenicity data to bridge to estimates of vaccine efficacy obtained in prior
clinical trials. In both cases comparative immunogenicity trials designed to
538
Annex 9
demonstrate non-inferiority are recommended. The choice of comparator is a

critical factor in the interpretation of the results.
5.6.2.1.1 Modifying the use of the vaccine for which efficacy has been estimated
As described in section 6 below, vaccine efficacy trials are usually conducted in
specific target populations – characterized by factors such as age, region (which
may define the endemicity of some infectious diseases) and health status – using
the intended final vaccine posology. Before or after licensure, trials may be
conducted with the aim of extending the use of the vaccine to other populations
and/or to support alternative posologies.
When a different age group or posology is proposed it is usually very
clear that a bridging trial is necessary. A bridging trial may be required if there are
compelling scientific reasons to expect that the immune response to the vaccine,
and therefore its efficacy, could be significantly different to that documented
in a prior efficacy trial because of host factors (such as common underlying
conditions that may affect immune responses) and/or geographical factors (such
as distribution of subtypes of organisms and levels of natural exposure). In infants
there is also the possibility that very different levels of maternal antibody could
occur in different regions, resulting in variable interference with infant immune
responses to the primary series.
The trial design may involve a direct comparison between: (a) the new
posology and that used in the efficacy trial; or (b) the new intended population
and a control group consisting of subjects who are representative of the prior
efficacy trial population. It may also be acceptable to make an indirect (cross-
trial) comparison with the immunogenicity data that were obtained during the
efficacy trial.
The vaccine formulation and assay used should be the same as those
used in the efficacy trial whenever possible:
■■ If the exact vaccine used in the efficacy trial is no longer available
the comparator should be as similar as possible to the original
vaccine that was evaluated. Over time, it may be that the only bridge
back to the efficacy data is via a comparison with a licensed vaccine
that was itself licensed on the basis of a bridging efficacy trial. As the
number of bridging steps that have occurred between the original
efficacy data and the licensed comparator vaccine increases, the
reliance that may be placed on a demonstration of non-inferiority to
predict efficacy is weakened. This consideration also applies when
the vaccine for which efficacy was estimated contained a certain
number of subtypes but was later replaced by a vaccine containing
a larger number of subtypes on the basis of comparing immune
responses to the shared subtypes.
539
■■ If the assay has changed and has not been, or cannot be, directly
compared to the original assay used during the efficacy trial it may
be possible to re-assay stored sera collected during the prior efficacy
trial in parallel with the sera from the new trial population.
If it remains unknown which immunological parameter best correlates
with efficacy it is preferable that the primary comparison between vaccines is
based on functional antibody whenever this is feasible.
5.6.2.1.2 Inferring the efficacy of a new candidate vaccine

In this case the main evidence of efficacy for licensure comes from one or more
bridging efficacy trials. The same considerations described above regarding
primary comparison, choice of comparative vaccine and assay apply.
If the new candidate vaccine contains additional subtypes of an organism
compared to licensed products and/or it contains subtypes of an organism that
have not previously been included in any licensed vaccine then interpretation
of the immune responses to the added or new subtypes is not straightforward.
Approaches that could be considered include comparing immune responses
to each added or new subtype with the mean immune response to all subtypes
or with the lowest immune response to any individual subtype included in a
vaccine for which efficacy was demonstrated. Although these approaches may
provide a route to licensure, the limitations of these comparisons in predicting
efficacy should be taken into account when considering the overall risk–benefit
relationship for the new vaccine.
5.6.2.2 Other approaches

When there is no ICP and it is not possible to bridge to a prior demonstration
of efficacy the evidence that may be provided to support likely vaccine efficacy
must be considered and discussed with NRAs on a case-by-case basis. In each

case the strength of evidence that may be provided should be weighed against
the advantages of having a licensed vaccine – one that has been subjected to a
full review of quality and nonclinical data, and for which it is considered that
there are adequate clinical safety and immunogenicity data – available for use
when needed.
Potential approaches may include establishing a nonclinical model of
efficacy that is thought to be relevant to the human infection and identifying
which immunological parameter best correlates with protection (and, if
possible, a putative ICP). Data on immune responses that occur in response to
natural infection and the resulting protection against further disease may be
useful, as may any passive protection data that are available from nonclinical or
clinical trials.
540
Annex 9
5.6.3 Co-administration trials

Comparative immunogenicity trials that are intended to support co-
administration of a vaccine with one or more other vaccines should demonstrate
non-inferiority for immune responses to each of the co-administered antigenic
components in the group that receives co-administered vaccines compared with
the groups that receive each vaccine given alone.
When multiple licensed products contain the same antigenic components
that could be co-administered with the vaccine under trial (for example,
combination vaccines intended for the routine infant primary immunization
series) it is not feasible, nor is it usually necessary, to assess co-administration
with each licensed product.
A particular issue arises when there are several different types of
polysaccharide-protein conjugate vaccines available that may be co-administered
with the vaccine under trial. When the vaccine under trial contains protein that
is the same as, or similar to, that in available conjugate vaccines it is important to
appreciate that the results obtained with any one conjugate may not be applicable
to other types of conjugate (for example, lack of immune interference with a
tetanus toxoid conjugate does not rule out the possibility that this could occur
when a different protein is used in the conjugate). Therefore, if co-administration
with several different conjugate vaccines is foreseen the effects of representative
vaccines that contain different conjugative proteins should be evaluated.
If multiple doses of the co-administered vaccines are needed then it
is usual to make the comparison between groups only after completion of all
doses. The schedule at which the vaccines are co-administered may also be a
consideration if there are several possibilities (for example, as in the case of
vaccines for the primary immunization series in infants or for vaccines against
hepatitis A and B). Consideration may be given to using a schedule that is most
likely to detect an effect of co-administration on immune responses if there is one.
Trials that assess the effects of co-administration may randomize subjects
to receive only one or all of the vaccines proposed for co-administration.
Alternatively, all subjects may receive all vaccines proposed for co-administration
but are randomized to staggered administration or co-administration. Staggered
administration is necessary when it is not possible to withhold any antigenic
components to be co-administered (for example, during the infant primary
schedule). In staggered administration trials the final dose and post-dose
sampling occur later compared to the co-administration group, which in infants
could have some impact on the magnitude of the immune response.
5.6.4 Immunization of pregnant women

Whenever the target population for a vaccine includes women of childbearing
age there is a need to consider the importance of generating data in pregnant
women. These considerations should take into account the nature of the vaccine
541
construct (for example, whether the vaccine contains a live organism that is
replication competent), whether pregnant women can reasonably avoid exposure
to an infectious agent (for example, by not travelling) and whether they may
have the same risk of exposure but a greater risk of experiencing severe disease
compared to non-pregnant women of the same age.
Not all vaccines are, or need to be, evaluated in trials in pregnant women.
If there is no or very limited experience of the use of a vaccine in pregnant
women, NRAs may consider whether nonclinical data and any data available
from the clinical use of the vaccine and very similar vaccines could be provided
in the prescribing information.
5.6.4.1 Aims of immunization during pregnancy

The immunization of women during pregnancy may benefit the mother and, in
some cases, may also result in benefit to the infant for a limited postnatal period
by means of placental transfer of maternal antibody (for example, influenza,
acellular pertussis and tetanus vaccines). In other cases the immunization of
women during pregnancy may provide some benefit to the infant with no or
negligible benefit to the mother (for example, respiratory syncytial virus vaccine).
It is also possible that immunization during pregnancy could prevent an
infection occurring in the mother and so protect the fetus from the consequences
of infection in utero.
5.6.4.2 Safety and immunogenicity in pregnancy

Before conducting trials in pregnant women, safety and immunogenicity data
should be available from clinical trials conducted in non-pregnant women of
childbearing age (20). Once there are adequate relevant nonclinical data with
satisfactory findings and some clinical data on safety and immune responses in
non-pregnant women, data may be obtained from pregnant women covering
a representative age range, so that the effects of pregnancy on the immune

response can be evaluated. The doses tested in pregnant women should be
based on the non-pregnant adult data but may need to be adjusted (in terms of
antigen dose or dose regimen) if the results indicate an effect of pregnancy on
the immune response.
In all trials conducted in pregnant women adequate mechanisms should
be in place to document the outcome of the pregnancy, including the duration
of gestation at time of delivery, the condition of the infant at birth and the
presence of any congenital conditions (see section 7.4 below).
5.6.4.3 Passive protection of infants

If there is already evidence of humoral immunity in a substantial proportion of
pregnant women against the infectious disease to be prevented, such that that
542
Annex 9
the aim of vaccination during pregnancy is to increase the amount of antibody

transferred to the fetus, then the trials in pregnant women may need to include
exploration of maternal immune responses to vaccination in both seropositive
and seronegative subjects.
Dose-finding trials in pregnant women should include measurement of
antibody levels in cord blood samples taken at delivery. The number of samples
obtained should be sufficient to provide an estimate of inter-individual variability.
Additional investigations may include the collection of cord blood covering a
range of times between maternal vaccination and delivery. Cord blood antibody
levels in infants born to vaccinated mothers who received the final selected
vaccine posology should be superior to those in infants born to mothers who
were not vaccinated. Secondary analyses could examine whether this finding
also applies within subsets of mothers who were seronegative or seropositive
prior to vaccination.
To avoid multiple bleeds in individual infants when evaluating the
duration of detectable maternal antibody, mothers may be randomized so that
their infants are sampled once or a few times at defined intervals. The total
data collected can be used to describe the antibody decay curve. These data are
particularly important when it is planned that passive protection via maternal
antibody will be followed by active vaccination of infants against the same
antigen(s) because of the possibility that high levels of maternal antibody may
interfere with the infant immune response.
If an ICP is established for the infectious disease to be prevented then the
aim of the immunogenicity trials should be to identify a maternal vaccination
regimen that results in cord blood antibody levels that exceed the ICP in a high
proportion of newborn infants. If no ICP exists there should be discussion with
NRAs regarding whether vaccine efficacy should be estimated in a pre-licensure
efficacy trial or whether an evaluation of vaccine effectiveness may suffice.
5.6.5 Changes to the manufacturing process

Changes made to product composition (for example, adding, removing or
changing a preservative) or to product manufacture (such as changes to process,
site or scale of manufacture) during the pre-licensure clinical development
programme or after licensure do not always need to be supported by comparative
clinical immunogenicity trials between the prior and newer products.
For example, although it is common for the scale of manufacture to
change during the pre-licensure development programme, this step alone may
not necessarily have a clinically significant effect in the absence of other changes.
To avoid the need for additional clinical trials to address manufacturing changes
the pivotal trials should whenever possible be conducted using vaccine made
according to the final process. If this is not the case, and for all changes that are
543
made post-licensure, consideration must be given to whether a clinical trial is

required to compare vaccines manufactured using the prior and new processes.
This decision must be taken on a case-by-case basis after a full evaluation of
the in vitro data, and of any nonclinical in vivo data describing and supporting
the change. Although a single lot of vaccine made using each process may
typically be sufficient for the comparison, data may on occasion be required from
multiple lots.
In the post-licensure period there may be many changes to the
manufacturing process over time. Whereas each one of these changes may be
considered too minor to merit the conducting of a clinical trial, the product
that results from multiple minor changes could be substantially different from
that which was first licensed. Therefore, when considering the merit of a clinical
trial, it may be important to consider the full history of changes that have been
allowed without clinical data and whether the sum total of these changes could
have a clinical impact. In this situation, when many years have passed, a clinical
trial of the current vaccine compared to the original licensed vaccine will not be
possible. However, if disease surveillance suggests that there could be a problem
with vaccine effectiveness, a clinical trial that compares the current vaccine
against another licensed vaccine may be considered useful.
5.6.6 Clinical lot-to-lot consistency trials

Clinical lot-to-lot consistency trials are conducted to provide an assessment
of manufacturing consistency in addition to the information provided on the
manufacturing process. Clinical lot-to-lot consistency trials may or may not
be considered necessary. Such trials may be considered particularly useful for
certain types of vaccines where there is inherent variability in the manufacture
of the product or when manufacturing consistency cannot be characterized
adequately by bio-physicochemical methods.
If a clinical lot-to-lot consistency trial is conducted then the usual

expectation is that the 95% confidence interval around each pair-wise
comparison of the post-vaccination geometric mean antibody concentrations/
titres falls within predefined limits. The clinical implications of results that show
that one or more comparisons do not meet the predefined criteria set around
the ratios should be considered in light of all available clinical immune response
data and relevant product-characterization data.
Whether or not a clinical lot-to-lot consistency trial is conducted, the
consistency of manufacturing for the vaccine lots used in clinical trials should be
both demonstrated and well documented. The lots used in clinical trials should
also be adequately representative of the formulation intended for marketing.
544
Annex 9
6. Efficacy and effectiveness

6.1 General considerations for efficacy trials
The need for, and feasibility of, evaluating the protective efficacy of a candidate
vaccine should be considered at an early stage of vaccine development because
the decision made will determine the overall content of the pre-licensure clinical
programme and will impact on its duration. In all application dossiers that do
not include an evaluation of vaccine efficacy the sponsor should provide sound
justification for the lack of such data, taking into account the points raised in
the following sections 6.1.1–6.1.3.
6.1.1 Efficacy data are not required

Vaccine efficacy trials are not necessary if it is established that clinical
immunological data can be used to predict protection against disease. For
example, if there is an established ICP against a specific disease (for example,
antitoxin levels against diphtheria and tetanus toxins, or antibody against
hepatitis B surface antigen) the candidate vaccine should be shown to elicit
satisfactory responses based on the relevant correlate(s).
6.1.2 Efficacy data are usually required

Vaccine efficacy trials are usually required whenever a new candidate vaccine is
developed with intent to protect against an infectious disease and one or more of
the following apply:
■■ There is no established ICP that could be used to predict the efficacy
of the new candidate vaccine.
■■ There is no existing licensed vaccine with documented efficacy
against a specific infectious disease to allow for bridging to a new
candidate vaccine.
■■ Use of immune responses to bridge the documented efficacy of a
licensed vaccine to a new candidate vaccine is not considered to
be possible. For example, because there is no known relationship
between specific immune response parameters and efficacy or
because the new candidate vaccine does not elicit immune responses
to the same antigen(s) as the licensed vaccine.
■■ There are sound scientific reasons to expect that the efficacy of a
vaccine cannot be assumed to be similar between the population(s)
included in the prior efficacy trial(s) and one or more other
populations.
545
■■ It cannot be assumed that the vaccine efficacy demonstrated against

disease due to specific strains of a pathogen (for example, serotypes
or subtypes) would apply to other strains.
6.1.3 Efficacy data cannot be provided

It may not be feasible to conduct efficacy trials. For example, if the new candidate
vaccine is intended to prevent an infectious disease that:
■■ does not currently occur (for example, smallpox);

■■ occurs in unpredictable and short-lived outbreaks that do not allow
enough time for the conducting of appropriately designed trials to
provide a robust estimation of vaccine efficacy (for example, some
viral haemorrhagic fevers);
■■ occurs at a rate that is too low for vaccine efficacy to be evaluated in
a reasonably sized trial population and period of time. This situation
may apply:
(a) because of natural rarity of the infectious disease (for example,
plague, anthrax and meningitis due to N. meningitidis group B);
(b) because of rarity of the disease resulting from the widespread
use of effective vaccines.
If it is not feasible to perform vaccine efficacy trials and there is no ICP it

may be possible to obtain evidence in support of vaccine efficacy and/or to derive
an immunological marker of protection from one or more of the following:
■■ Nonclinical efficacy trials.

■■ Passive protection trials – that is, nonclinical or clinical trials which
assess the effects of administering normal or hyperimmune human

gamma globulin or convalescent sera. The results may point to
the sufficiency of humoral immunity for the prevention of clinical
disease and may suggest a minimum protective antibody level
that could be used to interpret data obtained in clinical trials with
candidate vaccines.
■■ Comparison of immunological responses with those seen in past
trials of similar vaccines with proven protective efficacy even if the
relationship between immune responses to one or more antigenic
components and efficacy remains unknown.
■■ Human challenge trials.
546
Annex 9
6.2 Types of efficacy trials

6.2.1 Human challenge trials
Human challenge trials, in which subjects are deliberately exposed to an infectious
agent in a controlled setting, are not always feasible or appropriate. However, in
some settings it may be useful and appropriate to obtain an assessment of vaccine
efficacy from human challenge trials. If conducted, human challenge trials may
be of particular use:
■■ when there is no appropriate nonclinical model (for example, when
a candidate vaccine is intended to protect against an infectious
disease that is confined to humans);
■■ when there is no known ICP;
■■ when vaccine efficacy trials are not feasible.
6.2.2 Preliminary efficacy trials

If conducted, preliminary vaccine efficacy trials may provide an estimate of the
magnitude of protection that can be achieved by the new candidate vaccine.
However, preliminary efficacy trials are not usually designed and powered to
provide robust estimates of vaccine efficacy. These trials may be used to inform
the design of pivotal trials. For example:
■■ by evaluating the efficacy of different doses and dose regimens;
■■ by estimating efficacy on the basis of a range of efficacy variables;
■■ by analysing efficacy on the basis of various case definitions in order
to identify or refine the most appropriate case definition;
■■ by exploring efficacy in specific subgroups in order to decide if
there is a need to design pivotal trials specifically to further evaluate
efficacy in such subgroups;
■■ by assessing the method of case ascertainment for feasibility in larger
and more geographically diverse trials;
■■ by using immunogenicity and efficacy data to support a provisional
assessment of potential ICPs.
If the candidate vaccine is intended to prevent a severe and/or life-
threatening infectious disease for which there is no vaccine, or no satisfactory
vaccine, already available then individual NRAs may agree to accept an
application for licensure based on one or more preliminary efficacy trial(s).
In these cases it is essential that sponsors and NRAs should discuss and agree
547
upon the main features of the design of the trials before initiation (including the
sample size) so that, subject to promising results, the data may be considered
robust and sufficient.
The availability of a licensed vaccine has potentially important
implications for the acceptability and feasibility of initiating or completing
additional efficacy trials that include a control group that does not receive active
vaccination. These issues should be discussed between NRAs and sponsors so
that expectations for the completion of additional efficacy trials are agreed upon
prior to the start of trials that could potentially support licensure.
6.2.3 Pivotal efficacy trials

Pivotal vaccine efficacy trials are designed and powered to provide statistically
robust estimates of vaccine efficacy to support licensure. Pivotal efficacy trials
may evaluate one or more vaccination regimen(s), and may or may not include
evaluations of efficacy before and after booster doses.
6.3 Design and conduct of efficacy trials

The protective efficacy of a vaccine against a specific infectious disease is
usually determined in randomized trials that compare the incidence of disease
after vaccination relative to the incidence of disease in the control group that
has not been vaccinated. Less frequently, vaccine efficacy may be determined
in a prospective randomized trial which compares the incidence of disease after
vaccination between the group that received the new candidate vaccine and a
control group that received a licensed vaccine intended to prevent the same
infectious disease.
The following sections (6.3.1–6.3.9) are applicable to both types of trial.
As the details of statistical methodologies are beyond the scope of these WHO
Guidelines only broad principles are described. It is recommended that an
appropriately experienced statistician should be consulted.
6.3.1 Selection of trial sites

Vaccine efficacy trials require the presence of a sufficient burden of clinical
disease to enable estimates to be obtained from feasible numbers of subjects
within a reasonable time frame. The infectious disease to be prevented may
occur at sufficiently high rates to enable efficacy trials to be conducted only
in certain geographical areas. Even when the disease to be prevented is more
widespread it may be necessary to confine efficacy trials to specific areas for
reasons that may include feasibility, the need to ensure adequacy of monitoring,
and a desire to accumulate representative numbers of cases due to specific
serotypes or subtypes of the relevant pathogen.
548
Annex 9
If adequate data are not already available from public health authorities
then sponsors may have to conduct feasibility assessments in order to accurately
ascertain the clinical disease rates in various age subgroups of populations
before selecting trial sites. Any nationally recommended non-vaccine-related
preventive measures that are in place (for example, prophylactic drug therapy
in high-risk settings or in individuals at high risk, or the use of insect repellents
and bednets) should be identified. Trials are usually conducted against a
background of such measures.
Trial sites need to be sufficiently accessible to allow regular visits for
monitoring. Prior to initiation of the trial, sponsors may have to engage in site
capacity-building exercises, including training of study personnel, and may need
to provide essential infrastructure to support the trial (for example, adequate
blood-collection and processing facilities, refrigeration facilities suitable for the
vaccine and/or sera, access to competent laboratories, data-handling capacity
and communication methods to allow for electronic randomization schemes,
rapid reporting of safety data and other trial issues to the sponsor).
6.3.2 Candidate (test) vaccine group(s)

If previous data do not support selection of a single dose level or regimen of the
candidate vaccine for assessment of efficacy then trials may include one or more
groups in which subjects receive the candidate vaccine (for example, more than
one dose or schedule may be evaluated). In some cases one or more placebo doses
may need to be interspersed with candidate vaccine doses to enable the matching
of all regimens under trial in a double-blind design (for example, if two or three
doses of the candidate vaccine are to be compared with the control group).
6.3.3 Control (reference) group(s)

Control groups comprise all subjects who do not receive the candidate vaccine.
Usually only one control group is enrolled in any one trial. Sometimes it may
be important to include more than one of the possible types of control groups
discussed below.
6.3.3.1 Control groups not vaccinated against the infectious disease to be prevented
Following consultation between the sponsor, NRA, ethics committees, local
public health authorities and investigators it may be appropriate to use a control
group that is not vaccinated against the disease to be prevented by the new
candidate vaccine. For example, this may be the case when the trial is to be
conducted in countries in which:
■■ no vaccine is yet licensed for prevention of the disease in question;
and/or
549
■■ no such vaccine is included in the routine immunization schedule;

and/or
■■ there are sound reasons to believe that no licensed vaccine is likely
to provide useful efficacy (for example, because the licensed vaccine
does not cover, or is known/expected to have poor efficacy against,
the pathogen types that are most prevalent in a specific region).
In these cases the control group may receive:
■■ A true placebo (that is, material without any pharmacological
activity, such as normal saline). This has the advantage of providing
safety data against a control that has no pharmacologically active
components. The use of an injectable placebo may not be acceptable
to all NRAs, ethics committees, investigators, trial subjects or their
caregivers in some age groups (for example, particular objections
may be raised against true placebo injections in infants). In contrast,
there is usually no objection to the use of a true placebo when the
candidate vaccine is administered orally or by nasal instillation.
■■ A licensed vaccine that does not prevent the infectious disease under
study but may have some benefit for recipients. In some cases both
licensed vaccine and placebo doses may have to be administered to
the control group to match the candidate vaccine regimen in order
to maintain blinding.
If there are major objections to the use of placebo injections but no
potentially beneficial licensed vaccine would be suitable for the target age
group, the control group may be randomized to receive no injection. This is an
undesirable situation and should be regarded as a last resort since it precludes the
blinding of trial personnel or subjects/caregivers.
6.3.3.2 Control groups vaccinated against the infectious disease to be prevented

In this case the control group receives a vaccine that is already licensed to prevent
the same infectious disease as the candidate vaccine.
In some instances the control group may receive a licensed vaccine
that prevents infectious disease due to some, but not all, types of the pathogen
responsible for the disease that is to be prevented – in which case the group that
receives the licensed vaccine may be regarded as an unvaccinated control group
for the types found only in the candidate vaccine.
It is important that selection of the control vaccine takes into account
the available evidence supporting its efficacy and, if relevant, whether it appears
to have similar efficacy against all types of the pathogen involved. When there is
more than one available licensed control vaccine, or the selected control vaccine
550
Annex 9
is unlicensed or is not the product in routine use in a particular jurisdiction(s),

sponsors are advised to discuss selection of the comparator with the relevant
NRA(s). If it is not possible to reach agreement on the use of the same control
vaccine in all regions where efficacy is to be evaluated, consideration should be
given to conducting more than one efficacy trial with a different vaccine used in
the control group in each trial.
6.3.4 Trial designs

6.3.4.1 Randomization
The unit of randomization is most usually the individual. Alternatives include
the household or the cluster under trial (for example, a school population or a
local community). Randomization of groups or clusters, rather than individuals,
may be preferred when it is logistically much easier to administer the vaccine
to groups than to individuals and when estimates of the indirect effects of
vaccination (for example, herd immunity) are of interest. When the trial aims to
vaccinate pregnant women to protect the infant during the early months of life
then the unit of randomization is the mother.
6.3.4.2 Types of trial design

The simplest design involves randomization of equal numbers of subjects to the
candidate vaccine and control groups (that is, 1:1). In trials that employ a control
group that is not vaccinated against the disease to be prevented, but some clinical
data are available to support the likely efficacy of the candidate vaccine, it may be
appropriate (subject to statistical considerations and an assessment of the impact
on the total trial sample size) to use unbalanced randomization (for example,
2:1 or 3:1) to reduce the chance that individual subjects will be randomized to
the control group, thus ensuring that the majority of trial subjects receive the
candidate vaccine.
Trials may be planned to follow trial subjects for a fixed period after
the last dose of the primary series. The time at which the primary analysis is
conducted should take into account the anticipated rates of the disease under
study in each treatment group, including the unvaccinated control group if
applicable. Other considerations regarding the timing of the primary analysis
may include the possible importance of having some information on the duration
of protection before licensure occurs, the feasibility of following up subjects for
prolonged periods, and whether or not the vaccine could address a pressing
unmet need (for example, in an outbreak situation where there is no approved
vaccine to prevent the disease).
Alternatively, a case-driven approach may be taken based on the
anticipated rates of the primary efficacy end-point in the control group and
the expected or minimum desirable level of efficacy of the candidate vaccine.
551
In this design the primary analysis is conducted once a pre-specified total
number of cases has been detected – based, in a double-blind setting, on the
anticipated numbers in test and control groups required to demonstrate the
projected vaccine effect.
Alternative designs that allow for comparison with a control group that
is not vaccinated against the disease to be prevented may, at least in the short
term, include the following:
■■ In a randomized stepped wedge trial, the candidate vaccine
is administered to predefined groups in a sequential fashion.
Each predefined group is a unit of randomization. These may be
geographical groups or groups defined by host factors (for example,
age) or other factors (for example, attendance at a specific school
or residence within a specific health-care facility catchment area).
Such a design may be chosen when there is good evidence to
indicate that the vaccine will do more good than harm (affecting the
equipoise associated with randomization to a control group that is
not vaccinated against the disease to be prevented) and/or when it is
impossible to deliver the intervention to all trial participants within
a short time frame.
■■ In a ring vaccination trial, the direct contacts (and sometimes
secondary contacts) of a case may be randomized to vaccine or
control or may be randomized to receive immediate vaccination or
vaccination after a period of delay (21). This type of post-exposure
cohort trial usually requires smaller sample sizes than prospective
randomized controlled trials.
Ring vaccination trials may be particularly applicable when
the infectious disease to be prevented is associated with a relatively
high incidence of secondary cases in susceptible populations.

Therefore the use of this trial design requires prior knowledge of the
infectivity of the infectious agent and of the proportion of infections
that are clinically apparent, as well as of the general susceptibility of
the trial population.
Ring vaccination trials may not be appropriate if the
vaccination regimen requires multiple doses over an extended period
to induce a protective immune response.
The follow-up period for subjects after contact with the index
case should extend to the upper limit of the incubation period, taking
into account both the period during which the index cases were
infectious and the contact period. The inclusion period for new cases
and controls and their contacts following the detection of the first
552
Annex 9
case should be stated in the protocol. The duration of the inclusion

period should take into account the potential for introducing bias if
the disease incidence changes over time.
6.3.5 Clinical end-points

6.3.5.1 Primary end-points
The primary end-point(s) in preliminary trials may be different from the primary
end-point(s) used in the pivotal trial(s).
In most cases the focus of vaccine efficacy trials is the prevention of
clinically apparent infections that fit the primary case definition based on clinical
and laboratory criteria.
If an organism causes a range of disease manifestations (for example,
from life-threatening invasive disease to disease that is not serious if adequately
treated or is self-limiting) the primary end-point in any one trial should be
carefully selected in accordance with the proposed indication(s) for use.
A candidate vaccine may contain antigens derived from one or several
types (serotypes, subtypes or genotypes) of the same organism. There may
also be some potential for cross-protection against types not included in the
vaccine (for example, as observed with rotavirus vaccines and HPV vaccines).
In such cases it is usual for the primary end-point to comprise cases due to any
of the types included in the vaccine, and the trial is powered for this composite
end-point. It is not usually possible to power the trial to assess efficacy against
individual types in the vaccine or to assess cross-protection against types not in
the vaccine.
Alternative primary end-points may include:
■■ clinical manifestations of reactivated latent infection (for example,
herpes zoster);
■■ established chronic infections that may be asymptomatic but
predispose to infection-related disease later in life (for example,
chronic hepatitis B infection or persistent infection with HPV);
■■ other markers that predict progression to clinically apparent disease
(for example, histological changes associated with HPV infection
that are established precursors of malignant neoplasia).
6.3.5.2 Secondary end-points

As applicable to the individual candidate vaccine, other important end-points
may include:
■■ cases that occur after each dose, when the vaccine schedule includes
multiple doses and/or a booster;
553
■■ cases due to each of the individual types of the organism included in

the vaccine;
■■ cases due to the organism, regardless of whether the cases are caused
by types that are or are not included in the candidate vaccine;
■■ cases due to non-vaccine types;
■■ cases occurring in groups with host factors of interest (for example,
age or region);
■■ cases meeting criteria for disease severity – if available, validated
measures of criteria for severity should be used to facilitate
interpretation of the results;
■■ duration and/or severity of the illness, which may include clinical
measurements (for example, duration of fever or rash) and
laboratory measurements (for example, duration of shedding).
Eradication of carriage and/or reduction in disease transmission that is
not directly linked to, and/or accompanied by, a clinical benefit of vaccination
to the individual are not usually considered to be sufficient to support licensure.
Sponsors contemplating trials with these as primary end-points are advised to
consult widely with NRAs.
6.3.6 Case definition

As part of the predefined primary efficacy end-point, the protocol should describe
the clinical and laboratory criteria that must be met to define a case.
■■ If an end-point is defined as the occurrence of an acute infectious
disease then the case definition should include the core clinical
features as well as laboratory confirmation of the presence of the
target pathogen.
■■ If the end-point is defined as a consequence of a persistent infection

then details of sampling (frequency and method) and grading (if
applicable) should be described.
All laboratory assays used to define a case should be validated to the
satisfaction of relevant NRAs prior to initiating pivotal clinical trials.
Adequate case definitions should also be provided for secondary end-
points.
6.3.7 Case ascertainment

It is critical that the same methodology for case detection should be applied
consistently at all clinical sites throughout the duration of the trial. Active case
ascertainment usually requires frequent monitoring and contact with trial
554
Annex 9
subjects/caregivers. Passive case ascertainment is usually based on trial subjects/

caregivers presenting to or otherwise contacting a local health-care facility due
to the onset of specific symptoms. In this case, contact is commonly triggered
by one or more of a list of signs or symptoms given to trial subjects/caregivers at
the time of randomization, when they may also have been instructed to contact
a specific health-care facility. Alternatively, or in parallel, cases may be detected
by monitoring all local clinics and hospitals.
For efficacy end-points based on clinically apparent disease the possible
range of clinical presentations will determine the mode of case ascertainment.
For example, this may be hospital based for cases of life-threatening infections, or
community-based for less severe infections. If community-based, case detection
may depend on family practitioners and on initial suspicion of infection
by vaccinated subjects or their caregivers. It is critically important that the
individuals who are most likely to initiate detection of a possible case should
have clear instructions. These may need to cover issues such as the criteria for
initiating contact with designated health-care professionals, telephone contacts,
first investigations and further investigations once a case is confirmed.
For efficacy end-points other than clinically apparent disease it is
essential for subjects to be monitored at regular intervals to detect clinically
non-apparent infections or changes in other selected markers (for example,
the appearance of histological changes). The frequency of these visits, and
acceptable windows around the visits, should be stated in the trial protocol and
carefully justified.
The appropriate period of case ascertainment during a trial should be
determined mainly by the characteristics of the disease to be prevented and the
claim of protection that is sought at the time of licensure. For infectious diseases
that have marked seasonality, at least in some geographical locations (for example,
influenza and respiratory syncytial virus), it is usual to follow subjects through
one or more seasons to accumulate sufficient cases to conduct the primary
analysis. In these settings it is usual to conduct an enrolment campaign over a
short period just before the expected onset of each season.
6.3.8 Duration of follow-up

At the time of conducting the primary analysis for the purposes of obtaining
licensure the duration of follow-up in vaccine efficacy trials may be relatively
short (for example, 6–12 months) and may be insufficient to detect waning
protection, if this occurs. If feasible, case ascertainment may continue in vaccine
efficacy trials with maintenance of the randomized populations for a sufficient
duration to assess waning protection over time. Alternatively, or in addition,
waning protection may be assessed during the post-licensure period. These
data may serve both to indicate the need for, and optimal timing of, booster
doses and to estimate efficacy after booster doses.
555
6.3.9 Analysis of efficacy

Detailed plans for the analysis of efficacy, including any interim analyses and/or
plans to adjust the sample size during the study on the basis of specific criteria,
should be developed in conjunction with appropriately experienced statisticians,
and should be discussed with the NRA(s) before the protocol is finalized (and/
or during the conducting of the study, as necessary).
6.3.9.1 Sample size calculation

The trial sample size should be calculated on the basis of:
■■ the selected primary efficacy end-point, which could be a composite

of cases due to any of the organism types included in the candidate
vaccine;
■■ the primary analysis population (see below);
■■ the primary hypothesis (that is, superiority or non-inferiority and
the predefined success criteria).
If the primary analysis population represents a subset of the total

randomized population then the sample size calculation should include
an adequate estimation of numbers likely to be excluded from the primary
analysis for various reasons. In addition, a blinded review (for example, using
an independent data adjudication committee) of total numbers of subjects
enrolled who are eligible for the primary analysis population may be conducted
after randomization of a predefined number so that the trial sample size can be
adjusted accordingly.
6.3.9.2 Analysis populations

Clinical efficacy is usually assessed in the total randomized trial population (that
is, those who are assigned to receive vaccine and/or control) and in predefined
subsets of the randomized population.
The predefined trial populations should include as a minimum:
■■ all randomized subjects (that is, the full analysis set);

■■ all vaccinated subjects regardless of the numbers of assigned doses
actually received and whether or not doses were administered
within predefined windows;
■■ subjects who have generally complied with the protocol and have
received all assigned doses within predefined windows.
556
Annex 9
Other populations may be appropriate for some predefined secondary or

exploratory analyses. These may include, for example:
■■ those who completed specific numbers of assigned doses or received
all doses within predefined windows around the scheduled trial
visits (that is, analyses of efficacy according to adherence to the
vaccination regimen);
■■ subsets of all vaccinated subjects separated according to baseline
seropositivity versus seronegativity;
■■ subgroups defined by demographic factors known or postulated to
have an impact on vaccine efficacy.
6.3.9.3 Primary analysis

The primary analysis may sometimes be based on estimating efficacy in the
“per protocol” population and on rates of true vaccine failures. In this case,
the calculation of efficacy takes into account only those cases with onset after
a minimum time has elapsed following completion of the assigned doses. For
example, depending on knowledge of the kinetics of the immune response,
true vaccine failures may be limited to cases with onset more than a specified
number of days or weeks after the final dose of the primary series. In addition,
for a vaccine that contains antigens from only certain serotypes or subtypes the
primary analysis may be based on cases due to vaccine types only. Alternative
primary analysis populations that may be preferred by NRAs in some cases
include the all-randomized or the all-treated populations.
In trials that compare a candidate vaccine group with a group that is not
vaccinated against the disease to be prevented, the aim is to demonstrate that
the lower bound of the 95% confidence interval around the estimate of vaccine
efficacy is above a predefined percentage (which will always be above zero).
The predefined percentage should be selected on the basis of the expectation
of the point estimate of vaccine efficacy, taking into account what might be seen
as the minimum level of efficacy that could be considered clinically important.
The sample size calculation is based on this objective.
In trials that compare a candidate vaccine with an active control the
aim is usually to demonstrate non-inferiority of the candidate vaccine against
a control vaccine with demonstrated efficacy. This requires a predefined non-
inferiority margin, which should be justified in accordance with prior estimates
of vaccine efficacy for the disease to be prevented and the level of alpha on
which the sample size calculation depends. If the sponsor also intends to assess
superiority of the candidate vaccine over the active control the statistical analysis
plan should predefine a hierarchical assessment so that superiority is assessed
only after establishing that non-inferiority has been demonstrated.
557
6.3.9.4 Other analyses

The full range of secondary and exploratory analyses will depend on the
predefined end-points. Some of these analyses may be conducted in specific
predefined trial populations. For example, important sensitivity analyses for
supporting the primary analysis include those based on all proven cases
whenever they occurred after randomization and in each analysis population.
If the schedule includes more than one dose, analyses should be conducted to
count cases from the time of each dose or from a specified number of days after
each dose for all subjects who were dosed up to that point.
Other analyses may be based on cases that meet some but not all of the
case definition criteria, cases that are severe and cases that require a medical
consultation or hospitalization.
6.3.9.5 Other issues

6.3.9.5.1 Vaccines that contain antigens derived from several serotypes, subtypes or genotypes
If the primary analysis was confined to cases due to organism types included in
the vaccine then additional analyses should be conducted to evaluate efficacy
on the basis of all cases, regardless of the organism type responsible. If there are
sufficient numbers of cases due to organism types not included in the vaccine
these analyses may provide some indication of cross-protection.
If the data suggest unusually low efficacy against one or more organism
types in the vaccine it may be necessary to explore this issue in further trials.
6.3.9.5.2 Magnitude of vaccine efficacy

The point estimate of vaccine efficacy and 95% confidence intervals that are
obtained may indicate that a relatively modest proportion of cases can be
prevented. This fact alone does not preclude licensure provided that the sponsor
can provide evidence that the vaccine efficacy observed represents an important
clinical benefit (for example, if the vaccine prevents life-threatening infections

for which there is no very effective specific therapy and for which no vaccine
is available).
6.4 Approaches to determination of effectiveness

Vaccine effectiveness reflects direct (vaccine-induced) and indirect (population-
related) protection during routine use. The information gained from assessments
of vaccine effectiveness may be particularly important to further knowledge
on the most appropriate mode of use of a vaccine (for example, the need for
booster doses to maintain adequate protection over time). Vaccine effectiveness
is influenced by a number of factors, including:
■■ vaccination coverage of the population;
558
Annex 9
■■ pre-existing immune status of the population;

■■ differences between organism types included in a vaccine and the
predominant circulating types;
■■ changes in circulating predominant types over time;
■■ transmissibility of the pathogen and any effect that the introduction
of routine vaccination may have had on transmission rates.
Vaccine effectiveness may be estimated in several ways, namely:
■■ In observational cohort studies that describe the occurrence of
the disease to be prevented in the target population over time.
However, there is no randomization step and there is a potential for
considerable biases to be introduced.
■■ During phased introduction (for example, in sequential age or risk
groups) of the vaccine into the target population in which the groups
might form the units of randomization (that is, using a stepped
wedge design).
■■ Using other designs such as a case test-negative study design. In
this modification of a case control study, subjects with symptoms
suggesting the infectious disease under trial and seeking medical
care are tested for the infectious agent of interest. The cases are those
who are positive and controls are those who are negative for the
pathogen of interest. Bias may occur if vaccinated cases are less or
more severely ill and seek care at different rates compared to cases
that occur in individuals who are not vaccinated against the disease
to be prevented (22).
It may not be possible or appropriate for sponsors to conduct studies
to estimate vaccine effectiveness themselves. For reasons of feasibility it may be
necessary to collect the data via regional or national networks. For some types of
disease the use of data collected by means of national or international registries
may be appropriate. In addition, in some jurisdictions the estimation of vaccine
effectiveness in the post-licensure period is not considered to fall within the remit
of the licence holder.
Whatever the local requirements and arrangements, sponsors should
discuss arrangements for ongoing disease surveillance and the potential for
estimating effectiveness with the public health authorities in countries where the
vaccine is to be used and where appropriate surveillance systems are in place.
The plans for estimation of effectiveness should also be agreed with NRAs
at the time of licensure and the requirements for reporting effectiveness data to
the NRA, either via the sponsor or directly from a public health authority, should
be clarified.
559
It may be that reliable estimates of effectiveness can be obtained only in

certain countries in which vaccination campaigns are initiated and where there
is already a suitable infrastructure in place to identify cases. In addition, it would
likely be inappropriate to extrapolate any estimates of effectiveness that are
obtained to other modes of use (such as introducing the same vaccine to different
or highly selected sectors of the population).
7. Safety
All clinical trials that are conducted pre-licensure or post-licensure should
include an exploration of safety.
The assessment of safety may be the primary objective, a co-primary
objective or a secondary objective in a clinical trial. Since the methods for
collection, analysis and interpretation of safety data during clinical trials contrast
with those applicable to post-licensure routine safety surveillance they are
considered separately below.
In principle, many of the approaches to documenting and reporting safety
data during vaccine clinical trials and conducting vaccine pharmacovigilance
activities are similar to those used for all medicinal products. The sections
that follow should be read in conjunction with the extensive guidance that
is available from numerous publications, and on the websites of WHO, the
Council for International Organizations of Medical Sciences (CIOMS), ICH
and individual regulatory bodies. The focus of the following sections is thus on
a number of methods and practices that are different for vaccines compared to
other medicinal products, and on issues that may need to be addressed because
of vaccine composition.
7.2 Assessment of safety in clinical trials

7.2.1 Safety outcomes as primary or secondary end-points
7.2.1.1 Safety outcomes as primary end-points
When the assessment of safety is a primary objective of a clinical trial it is
usual for the primary analysis to be based on a specific safety end-point (for
example, rates of a certain AE or rates of AEs that may be part of a clinical
syndrome of interest). The trial may or may not be powered to address the pre-
specified hypothesis.
7.2.1.2 Safety outcomes as secondary end-points

When the assessment of safety or specific aspects of the safety profile is a
secondary objective, trials are not usually powered a priori to support statistical
560
Annex 9
analyses of end-points such as rates of all, or of specific, AEs. Descriptive

comparisons are commonly used to screen for any differences in AE rates
between treatment groups. If statistical analyses of AE rates are conducted they
should be pre-specified in the protocol and in the statistical analysis plan. If any
findings indicate statistically significant differences in rates of AEs (overall or for
specific AEs) between treatments then they should be interpreted with caution
unless the trial was primarily designed to address pre-specified hypotheses
regarding safety end-points. The biological plausibility that AEs that occur more
frequently in the new candidate vaccine group may be related to vaccination
should be taken into consideration when deciding on the need for further pre-
or post-licensure clinical trials to investigate and quantify the potential risks.
7.2.2 Recording and reporting adverse events

7.2.2.1 Methods
AEs should be reported and recorded by investigators and sponsors according
to detailed procedures described in the trial protocol. AEs should be classified
according to a standardized terminology (such as ICH MedDRA) to enable their
categorization by System Organ Class (SOC) and Preferred Term (PT). If the
classification terminology is updated while the trial is being conducted then
the clinical trial report should indicate how the changes affect the tabulations.
Expedited reporting of AEs that meet specific criteria should take place
in accordance with the requirements of individual NRAs relevant to the location
of the trial sites.
It is standard practice for vaccinees to be observed immediately after
each dose (for example, for a defined period – commonly 20–60 minutes) for
any severe immediate reactions (for example, severe hypersensitivity reactions
requiring immediate medical attention).
It is usually expected that all AEs are collected from all randomized
subjects for defined periods after each dose:
■■ Solicited signs and symptoms are usually recorded daily for at least
4–7 days after each dose (see section 7.2.2.2 below). Longer periods
(for example, 10–14 days) may be appropriate for certain vaccines,
such as those that replicate in recipients.
■■ Unsolicited AE reports are usually collected for the entire period
between each dose or, for single doses or final doses of regimens, for
approximately 4 weeks post-dose (see section 7.2.2.3 below).
■■ Reports of serious adverse events (SAEs) and any pre-specified AEs
of special interest (AESIs) should be collected from all trial subjects
for at least 6 months after the last dose of assigned treatment.
561
■■ For vaccines that contain new adjuvants it is recommended that

there should be follow-up for at least 12 months after the last dose to
allow for the documentation of any autoimmune diseases or other
immune-mediated AEs.
In trials involving large numbers of subjects (for example, vaccine
efficacy trials) it may be acceptable for reports of non-serious AEs to be collected
from a representative (and preferably randomized) subset or, occasionally, not
at all, taking into account the safety profile observed in the previous trials
and the number of subjects from which detailed safety data have already been
obtained, In this case, reports of all SAEs and any pre-specified AESIs should
be collected from all randomized subjects. It may be acceptable that only SAE
and AESI reports are collected during long-term safety follow-up.
7.2.2.2 Solicited signs and symptoms

In most trials it is common practice for certain local and systemic AEs to be
documented for a predefined period after each dose of a vaccine or placebo. The
recording of AEs may be facilitated by the use of diary cards or other methods
to ensure that the information is captured. If diary cards are used they may be
completed by vaccinees, caregivers or by study staff who have questioned the
vaccinees or their caregivers. These AEs are commonly referred to as “solicited
signs and symptoms” since information on their occurrence is actively sought
and they should be listed in the trial protocol.
For injectable vaccines the local signs and symptoms to be documented
usually include, as a minimum, pain, redness and swelling at the injection site
in all age groups. Pain should be graded according to a scoring system and
preferably one that has been validated. Measuring devices of various types may
be used to record the extent of redness and swelling.
Consideration should be given to assessing whether reports of pain are

associated with immediate pain during and just after the injection or whether
the pain is of later onset. If there is frequent reporting of pain at or around the
injection site during the hours or days following vaccination this may suggest
that the overall tolerability of the vaccine could negatively impact on vaccine
uptake in routine immunization programmes. In these circumstances it may be
appropriate to consider whether an attempt should be made to reformulate the
vaccine to improve local tolerability.
When two or more vaccines are given by injection at the same time, the
diary card should ensure that separate data are recorded for the new candidate
vaccine injection site.
The systemic signs and symptoms to be collected and documented are
determined by the age range in the trial (for example, those appropriate for
562
Annex 9
infants will not be wholly applicable to toddlers and older subjects) and by the
route of administration (for example, nausea and vomiting could be solicited
symptoms for vaccines given orally). Fever should be documented using digital
thermometers and should be determined at a specific site (for example, rectal
or axillary in infants). Recordings of fever should be made at predefined times
and for a specified number of days after each dose. For subjective symptoms (for
example, fatigue and myalgia) a simple scoring system should be included in the
diaries to allow for the grading of severity.
Any self-administered treatments used to address signs or symptoms
(such as antipyretic and analgesic medicines) and any contact with – or treatment
administered by – a health-care professional should be captured. Instructions
on the use of antipyretics and analgesics should be stated in the clinical trial
protocol. If at the time of each dose a supply of a specific antipyretic or analgesic
was provided for use as needed, or as instructed in accordance with the protocol,
the post-dose usage recorded should be checked against returned supplies. If
prior safety data suggest that pre-vaccination antipyretic use is appropriate then
this can be administered and recorded by trial staff at the vaccination visit.
At each trial visit, whether involving face-to-face or telephone contact
between the trial subject/caregiver and site staff, all diary cards completed by
vaccinees or caregivers should be checked for level of completion and further
instructions given as needed to improve data recording after the next dose is
given. At face-to-face visits the prior vaccination site(s) should be inspected
for any remaining signs such as induration. Trial subjects or caregivers should
also be asked about the maximum extent of signs (for example, to determine
whether whole limb swelling occurred). Any unresolved local or systemic signs
and symptoms should be recorded and action taken as appropriate.
7.2.2.3 Unsolicited adverse events

Trial subjects/caregivers should be questioned at each visit on the occurrence of
any AEs since the last visit or for predefined periods following the last dose. For
each AE, the timing of onset in relation to vaccination should be captured, as
should any consultation with a health-care professional, whether hospitalization
occurred and any treatment that was given (prescribed or non-prescribed).
If the AE is not already resolved there should be further follow-up to document
the outcome.
It may be useful to pose specific questions to trial subjects/caregivers
at each visit to ensure that certain AEs or AESIs are captured in a systematic
fashion – for example, to determine whether persistent inconsolable crying or
hypotonic-hyporesponsive episodes occurred in infants. Where well-established
and widely applied definitions of these and other AEs are available, they should
be included in the protocol.
563
For all AEs that meet the criteria for classification as SAEs there should be
careful documentation of dates of onset, underlying conditions and concomitant
medications, and adequate follow-up to record the outcomes.
7.2.2.4 Other investigations

The collection of data on routine laboratory tests (haematology, chemistry and
urine analysis) is not necessary in many clinical trials of vaccines. If the sponsor
or NRA considers that there is a good rationale for obtaining such data then the
protocol should specify the tests to be performed at certain time points. The tests
should be conducted in appropriately certified laboratories and results reported
using well-established grading scales for laboratory abnormalities.
For vaccines that contain live organisms (including attenuated wild-types,
organisms that have been genetically engineered to render them non-virulent
and/or non-replicative, and live viral vector vaccines) additional investigations
related to safety may include the detection of viraemia and assessments of
shedding (quantity and duration) unless the omission of such studies can be
justified (for example, on the basis of prior experience with the same or very
similar strains and/or nonclinical data). Organisms recovered from vaccinees may
also be subject to genetic analyses to determine any instances of recombination
with wild-types and reversion to virulence and/or replication competency.
The release specifications for vaccines should take into account the
safety profile documented for the highest amount(s) of antigen(s) that have
been administered in the clinical trials. It may be necessary to support the final
proposed release specification by conducting a trial with the primary objective
of comparing safety between formulations that contain different numbers of live
organisms or amounts of antigen(s).
Categorization of adverse events

7.2.3
7.2.3.1 Causality
Section 8.5 of the WHO Global manual on surveillance of adverse events
following immunization (23) recommends that in clinical trials the investigator
should make a judgement on relatedness to vaccination for all solicited signs
and symptoms, and unsolicited AEs. The sponsor may have access to additional
information that is not available to investigators and should assess causality for
all SAEs. The assessment of relatedness to vaccination should take into account
factors such as:
■■ plausibility of relatedness, taking into account the vaccine construct

(for example, live-attenuated vaccines may be associated with
modified manifestations of natural infection, such as rashes);
564
Annex 9
■■ timing in relation to dosing (while most vaccine-related AEs occur

within 1–2 weeks of the dose, there may reasons to suspect that
illnesses with onset many months after the last dose could be related
to prior vaccination);
■■ concurrent illnesses, vaccines or other medications;
■■ the frequency with which any one AE occurred in groups that
received the candidate vaccine compared to groups that received
another vaccine or placebo;
■■ any correlation between rates of any one AE and dose of antigenic
components;
■■ changes in rates of any one AE with sequential doses;
■■ the results of medical investigations (for example, diagnostic tests
for concurrent illnesses) and of autopsies (for example, in cases of
sudden infant death).
7.2.3.2 Severity
Sufficient data should be collected for each solicited sign and symptom and
unsolicited AE in order to assess severity. Wherever possible, widely used grading
scales (including scales that may be age specific) should be used. The same scales
should be applied throughout the clinical development programme.
7.2.3.3 Other categorization

The classification of AEs as serious and the categorization of frequencies (that
is, very common, common, uncommon, rare and very rare) should follow
internationally accepted conventions, as described in section 3.1.2 of the WHO
Global manual on surveillance of adverse events following immunization (23).
Frequencies of solicited signs and symptoms by subject and of AEs in each
treatment group should be calculated on the basis of the denominator of all
vaccinated subjects in that group. Calculation of the frequencies of solicited
signs and symptoms after each dose should use as the denominator the number
of subjects who received each dose.
7.2.4 Adverse event reporting rates within and between trials

During any clinical development programme the reporting rates in clinical trials
for all AEs and/or for specific types of AEs, whether solicited or unsolicited, may
demonstrate the following:
■■ Differences between candidate vaccines and control groups within a
clinical trial. For example, differences in AE rates may be anticipated
between a candidate vaccine group and a placebo group or a
565
group that receives a licensed vaccine that does not have a similar
composition to the candidate vaccine. Any marked differences
between a candidate vaccine and a licensed vaccine that has the
same or very similar composition are generally not anticipated and
may require further investigation.
■■ Differences between clinical trials that may be observed in one or
both of the candidate vaccine and control groups for total or specific
AE reporting rates. It is important to consider possible explanations,
taking into account whether or not the same effect on the pattern
of reporting rates was observed in groups that received candidate
vaccines and licensed vaccines and whether the study was double-
blind or open-label. There may be real and anticipated differences
in vaccine reactogenicity between trial populations (for example,
age-related differences for specific AEs, such as higher fever rates in
trials conducted in infants and toddlers compared to trials in older
children and adults). When there is no clear explanation for the
differences observed, further investigation is merited. For example,
there may have been incomplete reporting of AEs or data-entry
errors, as well as cultural factors that lead to a greater reluctance to
report side-effects in some regions.
7.3 Size of the pre-licensure safety database

The size of the pre-licensure safety database must be considered on a case-
by-case basis and agreed with relevant NRAs. It is not possible to predefine
a minimum number of exposed subjects (usually confined to the number
exposed to the final dose and regimen appropriate for their age group and who
received the final vaccine formulation) that can be generally applied across
vaccine development programmes.
When considering the pre-licensure safety database the need for a

sufficient sample size to estimate AE rates with precision is an important factor.
For example, a total database of 3000 subjects across all trials and populations
provides a 95% chance of observing one instance of an AE that occurs on
average in 1 in 1000 subjects.1 Nevertheless, this figure should not be assumed
to be appropriate in all settings. In particular, this figure should not be applied
to application dossiers for any type of new candidate vaccine without further
consideration. When considering the size of the pre-licensure safety database,
factors to take into account include, but are not limited to, the following:
The number that would provide a 95% chance of observing one instance of an AE that occurs on average
1
in 1 in 10 000 subjects is 30 000.

566
Annex 9
■■ Fewer than 3000 subjects may be acceptable if the new candidate

vaccine consists only of antigenic components that are already
licensed in other vaccines with which there is considerable
experience in routine use. The method of manufacture should also
be taken into account.
■■ For specific types of vaccines (for example, new constructs or new
adjuvants) or specific modes of use (for example, in a population
considered to be vulnerable or otherwise at high risk that could
predispose it to certain AEs) individual NRAs may require that
considerably more than 3000 subjects are exposed prior to licensure.
■■ Additional considerations may apply to vaccines that contain
antigenic components not previously used in human vaccines but
for which efficacy trials are not possible. For example, the safety
profile documented in the preliminary trials may lead to reluctance
to expose large numbers of subjects unnecessarily in the absence of
an immediate threat and/or to expose large numbers in particular
population subsets.
■■ The acceptable size of the pre-licensure safety database should take
into account the actual safety profile observed in the clinical trials.
If there is concern regarding the occurrence and/or severity of a
particular AE and the available safety data do not allow for a clear
assessment of risk then, depending on the perceived benefit of the
vaccine, it may be appropriate to conduct further pre-licensure trials
and/or to conduct a post-licensure safety study to better estimate
the risk.
The total number of subjects exposed in clinical trials may cover many
age subgroups, or a single age group may predominate. In general there should
be adequate representation of all target age groups in the total safety database. In
some cases, and depending on the actual safety profile, it may be acceptable for
the majority of subjects included in the safety database to come from a specific
age range.
7.4 Post-licensure safety surveillance

The main purpose of post-licensure safety surveillance is to detect AEs that
occur too infrequently for detection in pre-licensure clinical trials.
The requirements of individual NRAs for reporting safety data collected
from post-licensure safety surveillance activities should be consulted along
with other guidance such as ICH E2E. NRAs should provide publicly available
guidance regarding their requirements for the content and timing of periodic
reports of safety data and for any expedited reporting considered necessary.
567
Licence holders should demonstrate that they have adequate capability and
appropriate staff to collect, interpret and act upon the safety data received. It
is important that efforts are made to accurately identify the vaccine(s) and lot
number(s) associated with each AEFI report.
It has become routine at the time of licensure for detailed proposals to
be in place for post-licensure safety surveillance activities, often in the form
of risk‑management plans. These documents and proposals are then routinely
updated at intervals in line with additional data that become available. The
plans usually outline the safety specification for the vaccine on the basis of all
available safety data at the time of submitting each version of the plan, along
with details of routine and proposed additional pharmacovigilance and risk-
minimization activities.
When planning pharmacovigilance activities for a vaccine it is
important to take into account that, in addition to routine pharmacovigilance
(that is, passive surveillance), important information may come from other
sources, namely:
■■ Data from active safety surveillance, which may be put in place by
public health bodies when a vaccine is introduced into a national
routine immunization programme, or when the use of a vaccine
within a programme changes significantly (for example, an
entirely different age group is vaccinated for the first time). Active
surveillance seeks to ascertain completely the number of AEs in
persons given a dose of a vaccine using a pre-organized process. It
may involve reviewing medical records or interviewing patients and/
or physicians in a sample of sentinel sites to ensure that complete
and accurate data are collected on reported AEs from those sites.
■■ Large databases that link information on vaccination history in
patient records with the occurrence of specific types of illness. These
databases can be searched to explore links between specific vaccines

and safety issues in the short and longer term.
■■ Various types of registries intended to capture details of vaccine use
in specific populations. For example, some registries collect
information on exposure of pregnant women to various types of
vaccines and indicate the outcome of the pregnancy (including
rates of spontaneous abortion, premature delivery and congenital
malformations in infants).
The limitations of each of these approaches are well known. Furthermore,
access to information from these other sources varies greatly between countries.
These factors underline the need to consider safety data from all sources along
with data that may come from post-licensure trials.
568
Annex 9
An additional consideration for vaccines is that when a safety signal is

identified for any one vaccine it may or may not be possible to ascribe the AEFIs
observed to any one antigenic component of the vaccine or to an adjuvant.
Furthermore, if there was concomitant administration of vaccines in some or
all cases generating the signal, it may not be possible to ascribe the AEFI to
only one of the products co-administered. The same or very similar antigenic
component(s) or adjuvant in the vaccine(s) from which the signal arose may
be present in several other licensed products marketed worldwide. Ultimately,
several different licence holders and NRAs without established data-sharing
agreements may need to be involved. As a result, the actions taken, if any,
and the speed at which action is taken are sometimes very variable between
countries. Such issues underscore the need for the efficient use of electronic
databases to facilitate rapid data sharing.

The first draft of this WHO Guidelines document was prepared by a WHO
drafting group comprising Dr M. Powell, Medicines and Healthcare products
Regulatory Agency, England; Dr R. Sheets, Consultant, Silver Spring (MD), the
USA; Dr J. McEwen, Therapeutic Goods Administration, Australia; and Dr I.
Knezevic and Dr V. Moorthy, World Health Organization, Switzerland, taking
into consideration the discussions and consensus reached during a WHO
Consultation on clinical evaluation of vaccines held in Geneva, Switzerland,
17–18 July 2014 and attended by the following participants: Dr P. Annunziato,
Merck & Co., the USA; Dr N. Bhat, Program for Appropriate Technology in
Health, the USA; Dr A. Chatterjee, Biological E Ltd, India; Dr K. Chirgwin, Bill
& Melinda Gates Foundation, the USA; Dr G. Coleman, Health Canada, Canada;
Dr D. Tuan Dat, The Company for Vaccines and Biological Production No. 1
(Vabiotech), Viet Nam; Dr P.E. Fast, International AIDS Vaccine Initiative, the
USA; Dr G. Foglia, Sanofi Pasteur, the USA; Dr M. Gruber, United States Food
and Drug Administration, Center for Biologics Evaluation and Research, the
USA; Dr P.M. Heaton, Bill & Melinda Gates Foundation, the USA; Dr D. Kaslow,
Program for Appropriate Technology in Health, the USA; Dr Y.H. Lee, Ministry
of Food and Drug Safety, Republic of Korea; Dr D.J.M. Lewis, University of
Surrey, England; Dr A. Lommel, Paul-Ehrlich-Institut, Germany; Dr J. McEwen,
Therapeutic Goods Administration, Australia; Dr P. Neels, Vaccine-Advice
BVBA, Belgium; Dr M. Nijs, GlaxoSmithKline Biologicals, Belgium; Dr S.A.
Nishioka, Ministry of Health, Brazil; Dr A. Podda, Novartis Vaccines Institute
for Global Health, Italy; Dr M. Powell, Medicines and Healthcare products
Regulatory Agency, England; Dr A. Ramkishan, Central Drugs Standard Control
Organization, India; Dr R. Sheets, Consultant, Silver Spring (MD), the USA; Dr J.
Shin WHO Regional Office for the Western Pacific, Philippines; Dr P. Smith,
569
London School of Hygiene and Tropical Medicine, England; Dr J. Southern,

Medicines Control Council, South Africa; Dr Y. Sun, Paul-Ehrlich-Institut,
Germany; Dr Z. Yang, Center for Drug Evaluation, China; and Dr U. Fruth, Dr I.
Knezevic, Mr O.C. Lapujade, Dr V. Moorthy, Dr K. Vannice and Dr D. Wood,
World Health Organization, Switzerland.
The resulting draft document was posted on the WHO Biologicals
website for the first round of public consultation from 30 October to 30
November 2015.
The second draft was prepared by a WHO drafting group, taking into
account comments received from: Dr B. Fritzell, BFL Conseils, France; Dr
G. Chen, National Institutes of Health, the USA; Dr G. Coleman, Health
Canada, Canada; Dr Z. Kusynová, The Hague, the Netherlands (provided the
consolidated comments of the International Pharmaceutical Federation); Dr M.
Nijs, GlaxoSmithKline Vaccines, Belgium; Clinical team of Novilia Sjafri and PT
Bio Farma, Indonesia; Dr Y. Sun, Paul-Ehrlich-Institut, Germany; Dr I. Uhnoo,
Uppsala Universitet, Sweden; Dr T. Yamaguchi, Pharmaceuticals and Medical
Devices Agency, Japan; and Dr K. Zoon, National Institutes of Health, the USA.
The draft document was then posted on the WHO Biologicals website
for a second round of public consultation from 1 February to 15 March 2016.
Comments were received from: Dr B. Brock, Sanofi Pasteur, the USA (provided
the consolidated comments of the International Federation of Pharmaceutical
Manufacturers & Associations (IFPMA)); Dr K. Farizo, United States Food and
Drug Administration, Center for Biologics Evaluation and Research, the USA;
Dr C. Meric, Lausanne University Hospital, Switzerland; Mr J.F. Modlin, Bill &
Melinda Gates Foundation, the USA; Dr D. Pratt, United States Food and Drug
Administration, Center for Biologics Evaluation and Research, the USA; Dr A.
Rinfret, Health Canada, Canada; and Dr K. Sohn, Ministry of Food and Drug
Safety, Republic of Korea.
A WHO meeting of the Working Group on clinical evaluation of vaccines
was then held in Geneva, Switzerland, 3 May 2016 and was attended by the
following participants: Dr G. Coleman, Health Canada, Canada; Dr M. Darko,
Food and Drugs Authority, Ghana; Dr D. Etuko, National Drug Authority,
Uganda; Dr E. Griffiths, Consultant, Kingston-upon-Thames, England; Dr S.
Kennedy, University of Liberia, Liberia; Dr J. McEwen, Therapeutic Goods
Administration, Australia; Dr M. Powell, Medicines and Healthcare products
USA; Dr J. Southern, Medicines Control Council, South Africa; Dr Y. Sun, Paul-
Ehrlich-Institut, Germany; Dr K. Zoon, National Institutes of Health, the USA;
and Dr I. Knezevic, World Health Organization, Switzerland.
Based on the comments received during the public consultation and
on the discussions of the above Working Group meeting, the document WHO/
BS/2016.2287 was prepared by the above-mentioned WHO drafting group.
570
Annex 9
The document was then posted on the WHO Biologicals website for
a third round of public consultation from 27 July to 16 September 2016 and
comments received from: Dr B. Brock, Sanofi Pasteur, the USA (provided the
consolidated comments of the IFPMA); Dr M. Cavaleri, European Medicines
Agency, England; Dr G. Coleman, Health Canada, Canada; Dr D. Kim and
Dr M. Jin, Ministry of Food and Drug Safety, Republic of Korea; Dr A.W. Lee,
Dr A. Sitlani and Dr W. Straus, Merck & Co., the USA; Dr T. Lu, Therapeutic
Goods Administration, Australia; Dr S.A. Nishioka, Ministry of Health, Brazil;
Office of International Affairs, Instituto Nacional de Vigilancia de Medicamento,
Colombia; Dr S-C Shin, Green Cross Corporation, Republic of Korea; and
Dr Y. Sun, Paul-Ehrlich-Institut, Germany. Dr Noni MacDonald, DalHousie
University, Halifax, Canada and Dr Anna Taddio, University of Toronto, Toronto,
Canada, provided comments on the safety evaluation, particularly on the pain
of injection.
References
1. Guidelines on clinical evaluation of vaccines: regulatory expectations. In: WHO Expert Committee
on Biological Standardization: fifty-second report. Geneva: World Health Organization; 2004:
Annex 1 (WHO Technical Report Series, No. 924; http://www.who.int/biologicals/publications/trs/
areas/vaccines/clinical_evaluation/035-101.pdf?ua=1, accessed 4 February 2017).
2. WHO guidelines on nonclinical evaluation of vaccines. In: WHO Expert Committee on Biological
Standardization: fifty-fourth report. Geneva: World Health Organization; 2005: Annex 1 (WHO
Technical Report Series, No. 927; http://www.who.int/biologicals/publications/trs/areas/vaccines/
nonclinical_evaluation/ANNEX%201Nonclinical.P31-63.pdf?ua=1, accessed 4 February 2017).
3. Guidelines for good clinical practice (GCP) for trials on pharmaceutical products. In: WHO Expert
Committee on the Use of Essential Drugs: sixth report. Geneva: World Health Organization; 1995:
Annex 3 (WHO Technical Report Series, No. 850; http://whqlibdoc.who.int/trs/WHO_TRS_850.
pdf?ua=1, accessed 4 February 2017).
4. WHO good manufacturing practices for pharmaceutical products: main principles. In: WHO Expert
Committee on Specifications for Pharmaceutical Preparations: forty-eighth report. Geneva:
who.int/medicines/areas/quality_safety/quality_assurance/TRS986annex2.pdf?ua=1, accessed
4 February 2017).
5. WHO good manufacturing practices for biological products. In: WHO Expert Committee on
Biological Standardization: sixty-sixth report. Geneva: World Health Organization; 2016: Annex 2
(WHO Technical Report Series, No. 999; http://www.who.int/biologicals/areas/vaccines/Annex_2_
WHO_Good_manufacturing_practices_for_biological_products.pdf?ua=1, accessed 27 March
2017).
6. Guidelines on the nonclinical evaluation of vaccine adjuvants and adjuvanted vaccines. In:
WHO Expert Committee on Biological Standardization: sixty-fourth report. Geneva: World
biologicals/areas/vaccines/TRS_987_Annex2.pdf?ua=1, accessed 4 February 2017).
571
7. Guidelines on procedures and data requirements for changes to approved vaccines. In: WHO Expert
Committee on Biological Standardization: sixty-fifth report. Geneva: World Health Organization;
2015: Annex 4 (WHO Technical Report Series, 993; http://www.who.int/biologicals/vaccines/
Annex4_Guidelines_changes_to_approved_vaccines_eng.pdf?ua=1, accessed 4 February 2017).
8. Guidelines for independent lot release of vaccines by regulatory authorities. In: WHO Expert
2013: Annex 2 (WHO Technical Report Series, No. 978; http://www.who.int/biologicals/TRS_978_
Annex_2.pdf?ua=1, accessed 4 February 2017).
9. Recommendations for the evaluation of animal cell cultures as substrates for the manufacture
of biological medicinal products and for the characterization of cell banks. In: WHO Expert
2013: Annex 3 (WHO Technical Report Series, No. 978; http://www.who.int/biologicals/vaccines/
TRS_978_Annex_3.pdf?ua=1, accessed 4 February 2017).
10. Clinical considerations for evaluation of vaccines for prequalification. Points to consider for
manufacturers of human vaccines. Geneva: World Health Organization; 2010 (http://www.
who.int/immunization_standards/vaccine_quality/clinical_considerations_oct10.pdf, accessed
4 February 2017).
11. Immunization in practice: a practical guide for health staff. 2015 update. Geneva: World
Health Organization; 2015 https://extranet.who.int/iris/restricted/bitstream/10665/193412/1/
9789241549097_eng.pdf, accessed 4 February 2017).
12. Expert consultation on the use of placebos in vaccine trials. Meeting report. Geneva: World Health
Organization; 2013 (http://apps.who.int/iris/bitstream/10665/94056/1/9789241506250_eng.pdf,
13. Plotkin SA, Gilbert PB. Nomenclature for immune correlates of protection after vaccination.
Clin Infect Dis. 2012;54(11):1615–7 (https://academic.oup.com/cid/article/54/11/1615/321882/
Nomenclature-for-Immune-Correlates-of-Protection?related-urls=yes&legid=cid;cis238v1,
accessed 27 March 2017).
14. The importance of pharmacovigilance: safety monitoring of medicinal products. Geneva: World
Health Organization; 2002 (http://apps.who.int/medicinedocs/pdf/s4893e/s4893e.pdf, accessed
4 February 2017).
15. Siegrist C-A. Vaccine immunology. In: Plotkin SA, Orenstein WA, Offit PA, editors. Vaccines, sixth
edition. Philadelphia (PA): Elsevier Saunders; 2012, chapter 2.
16. ICH tripartite guideline Q2(R1). Validation of analytical procedures: text and methodology (Step
4 version) Geneva: International Conference on Harmonisation of Technical Requirements for
Registration of Pharmaceuticals for Human Use; 1996; 6–13 (http://www.ich.org/fileadmin/
Public_Web_Site/ICH_Products/Guidelines/Quality/Q2_R1/Step4/Q2_R1__Guideline.pdf,
17. Recommendations to assure the quality, safety and efficacy of DT-based combined vaccines. In:
WHO Expert Committee on Biological Standardization: sixty-third report. Geneva: World Health
Organization; 2014: Annex 6; Table 6.2 (WHO Technical Report Series, No. 980; http://www.who.
int/biologicals/vaccines/Combined_Vaccines_TRS_980_Annex_6.pdf?ua=1, accessed 27 March
2017).
18. Buttery JP, Riddell A, McVernon J, Chantler T, Lane L, Bowen-Morris J et al. Immunogenicity
and safety of a combination pneumococcal-meningococcal vaccine in infants: a randomized
controlled trial. JAMA. 2005;293(14):1751–8 (http://jamanetwork.com/journals/jama/fullarticle/
200691, accessed 4 February 2017).
572
Annex 9
19. Richmond P, Kaczmarski E, Borrow R, Findlow J, Clark S, McCann R et al. Meningococcal C

polysaccharide vaccine induces immunologic hyporesponsiveness in adults that is overcome
by meningococcal C conjugate vaccine. J Infect Dis. 2000;181(2):761–4 (https://academic.oup.
com/jid/article/181/2/761/826799/Meningococcal-C-Polysaccharide-Vaccine-Induces, accessed
4 February 2017).
20. WHO consultation on influenza vaccines for pregnant and lactating women: clinical data
requirements for product labelling, Geneva, Switzerland, 15–16 July 2014. Meeting report
(http://www.who.int/biologicals/vaccines/INFLUENZA_VACCINES_Executive_Summary_
FINAL_29_Sept_R.pdf, accessed 4 February 2017).
21. Ebola ça suffit: Ring vaccination trial Consortium. The ring vaccination trial: a novel cluster
randomised controlled trial design to evaluate vaccine efficacy and effectiveness during
outbreaks, with special reference to Ebola. BMJ. 2015;351:h3740 (http://dx.doi.org/10.1136/bmj.
h3740, accessed 4 February 2017).
22. Foppa IM, Haber M, Ferdinands JM, Shay DK. The case test-negative design for studies of the
effectiveness of influenza vaccine. Vaccine. 2013;31(30):3104–9 (http://www.sciencedirect.com/
science/article/pii/S0264410X13004659, accessed 4 February 2017).
23. Global manual on surveillance of adverse events following immunization. Geneva: World
Health Organization; 2014 (http://www.who.int/vaccine_safety/publications/Global_Manual_on_
Surveillance_of_AEFI.pdf, accessed 4 February 2017).
573
Annex 10
Human challenge trials for vaccine development:
regulatory considerations
1. Introduction 578
2. Background 580
4. Purposes of human challenge trials in vaccine development 580
5. Study design of human challenge trials 582
6. Operational aspects 583
7. Some key ethical considerations 584
References 587
575

576
Annex 10
Abbreviations
GMO genetically modified organism
ICP immune correlate of protection
577
1. Introduction
Infectious human challenge trials involve the deliberate exposure of human
volunteers to infectious agents. Trial participants are intentionally challenged
(whether or not they have been vaccinated) with an infectious disease organism.
This challenge organism may be close to wild-type and pathogenic, adapted
and/or attenuated from wild-type with less or no pathogenicity, or genetically
modified in some manner.
Human challenge trials have been conducted over hundreds of years
and have contributed vital scientific knowledge that has led to advances in the
development of drugs and vaccines. Nevertheless, such research can appear to
be in conflict with the guiding principle in medicine to do no harm. A number
of well-documented historical examples of human exposure studies would be
considered unethical by current standards. It is essential that challenge trials be
conducted within an ethical framework in which truly informed consent is given.
When conducted, human challenge trials should be undertaken with abundant
forethought, caution and oversight. The value of the information to be gained
should clearly justify the risks to human subjects.
Although human challenge trials are not a required element of every
vaccine development programme, there are many reasons why a developer may
ask to conduct a “challenge-protection” study with humans, which might normally
be conducted in animals. Animal models are often quite imprecise in reflecting
human disease, and many infectious organisms against which a developer might
wish to develop a vaccine are species-specific for humans. Human challenge
trials may be safely and ethically performed in some cases, if properly designed
and conducted. Considerable insight can then be gained into the mode of action
and potential benefit of drugs and vaccines in humans. However, there are also
limitations on what challenge trials may be able to ascertain because, as with
animal-model challenge-protection studies, a human challenge trial represents a

model system. Nevertheless, because there are often such significant limitations
to animal models, the model system of a human challenge trial may significantly
advance, streamline and/or accelerate vaccine development (1).
It is important to note that not all diseases for which vaccines might be
developed are suitable for conducting human challenge trials. In many cases,
human challenge with a virulent or even attenuated organism would not be
considered ethical or safe. For example, if an organism causes a disease with a
high case-fatality rate (or there is a long and uncertain latency period) and there
are no existing therapies to prevent or ameliorate disease and preclude death,
then it would not be appropriate to consider human challenge trials with such
an organism. However, a human challenge trial might be considered when the
578
Annex 10
disease an organism causes has an acute onset, can be readily and objectively
detected, and existing efficacious treatments (whether curative or palliative) can
be administered at an appropriate juncture in disease development to prevent
significant morbidity and eliminate mortality.
It will also be important to consider the regulatory framework in which
the human challenge trial may be conducted. In some countries, challenge stocks
are expected to be regulated in the same manner as vaccines, and are expected
to be studied with authorization in accordance with clinical trial regulations,
whether or not an investigational vaccine is to be used in the same clinical
investigation protocol. For example, a challenge trial might be conducted to titrate
the challenge organism in humans (before using the challenge in a vaccine study)
in order to determine the proper dose of the challenge organism to administer,
and to characterize the symptoms, kinetics, shedding and transmissibility to be
expected from the challenge. The dose of challenge organism is usually titrated to
induce a relatively high attack rate while limiting disease severity. In cases where
the challenge should be studied in compliance with clinical trial regulations
there is greater clarity about regulatory expectations, including the quality of the
challenge stock to be used, because the clinical trial regulations or requirements
would apply. However, in many countries, because the challenge stock is not itself
considered to be a medicinal product, such characterization/model development
studies would not come under national regulatory authority (NRA) review and
authorization. Thus, much less clarity would exist on regulatory expectations and
issues of quality in such cases.
It should be understood that a pathogenic challenge strain will not have
the “safety” of an intended safe candidate vaccine. However, its quality should
be comparable to a candidate vaccine at the same clinical trial phase. Ideally,
a human challenge trial to establish the challenge model (that is, without use
of an investigational medicinal product) should also match the expectations
for conducting a vaccine study – that is, compliance with good clinical
practice (GCP) and subject to approval or concurrence under a Clinical Trial
Authorization by NRAs and ethics committees on the basis of requirements
appropriate for this type of study. If such a framework does not exist, countries
are encouraged to establish an appropriate regulatory and ethical framework for
challenge trials. However, there may be no regulatory framework to promulgate
such expectations in the country where the challenge study is to be conducted.
Trial sponsors, vaccine developers, researchers and other involved parties should
determine what regulatory expectations the relevant NRA may have when clarity
does not exist and when the human challenge study is intended to support the
development of a vaccine candidate they would ultimately like to license (that is,
obtain marketing authorization or registration).
579
2. Background
In July 2014, WHO held a consultation on Clinical evaluation of vaccines:
regulatory expectations (2). One area that was considered to be an important
element in facilitating vaccine development was human challenge trials. It was
recognized that the regulation of such trials needed to be well-defined by NRAs
and that vaccine developers and manufacturers needed to be aware of regulatory
expectations in this area.
This WHO guidance document on human challenge trials should be
read in conjunction with the updated WHO Guidelines on clinical evaluation of
vaccines: regulatory expectations (3) which were adopted, along with the current
document, by the WHO Expert Committee on Biological Standardization in
October 2016.

The purpose of this document is to provide guidance to NRAs, manufacturers,
vaccine developers, investigators and independent ethics committees – and
potentially to biosafety committees and national agencies that regulate genetically
modified organisms (GMOs) where separate from the NRA. The document only
covers issues specifically relevant to the design and conduct of clinical trials that
enrol healthy adult humans capable of truly informed consent, and that involve
the intentional exposure to, and potential infection with, an infectious disease
organism. All other issues common to the design, conduct and evaluation
(assessment) of vaccine clinical trials may be found in the updated WHO
Guidelines on clinical evaluation of vaccines: regulatory expectations (3).
4. Purposes of human challenge trials

in vaccine development
Human challenge trials are considered as a model by which challenge protection
can be evaluated and represent one possible approach for vaccine development.
Therefore, all principles for the clinical evaluation of vaccines should
apply, including the need for approval by the NRA and ethical committees as
well as compliance to GCP.
A vaccine developer may conduct human challenge trials to accomplish
one or more aims. The aims of the study determine the clinical phase in which the
study is conducted. Human challenge trials are often a type of efficacy-indicating
study, but most would not be considered to be pivotal efficacy studies. Almost all
would be pilot in nature and performed to gain useful information to aid in the
580
Annex 10
development of a vaccine. Several challenge trials might be performed during the

course of vaccine development.
Potential purposes of human challenge trials could include one or more
of the following:
■■ characterization of the challenge stock and model system in terms of
titration, symptoms, kinetics, shedding and transmissibility;
■■ clearer understanding of the pathogenesis of, and immunity to, the
organism in order to guide decisions on what immune responses
(type and/or quantity) a vaccine might need to elicit in order to
protect against that disease as part of gaining insight into vaccine
design – studies for this purpose may be referred to as experimental
medicine studies;
■■ identification of potential immune correlates of protection (ICPs)
which would then require validation in a traditional efficacy study;
■■ identification of the optimal design for traditional pivotal efficacy
trial(s) – for example, case definitions, end-points and other study
design aspects;
■■ generation of appropriate hypotheses to be formally tested in
traditional efficacy trials;
■■ proof of concept as to whether a particular vaccine candidate might
provide protection or not;
■■ down- or up-selection of various potential lead vaccine candidates
to advance only the best to large pilot or pivotal efficacy trials and to
eliminate those not worth advancement;
■■ de-risk or “left-shift” 1 risk of failure in a vaccine development
programme;
■■ comparison of vaccine performance in endemic settings versus
an efficacy trial population,2 including evaluating the impact of
prior immunity in the context of prevalent endemic diseases and
conditions;
1
When the timeline of vaccine development is viewed as a graph from early to the left to late to the
right, shifting the risk of failure earlier (left) in the timeline could: (a) minimize risk to human subjects by
avoiding large efficacy studies of vaccines that would not prove efficacious; (b) result in significant cost
and resource savings; and (c) minimize lost opportunity costs by abandoning an unpromising candidate
before committing greater expenditures to higher-phase clinical trials.
2
The target population in a particular country may have a higher rate of individuals with, for example,
sickle cell trait, poorer nutritional status or greater parasitic load in “normal” flora – any of which might
affect immune responsiveness in the endemic setting and thus efficacy (benefit) compared to the efficacy
trial population (ideal setting) or safety (greater risks). Either of these would have an impact on the risk–
benefit decision-making.
581
■■ support for emergency use of an investigational vaccine (for example,

during an influenza pandemic);
■■ provision of a basis for licensure (this purpose would be a rare
exception rather than routine);
■■ post-licensure exploration of whether immunity following
vaccination wanes, and if or when booster doses might be required
for durable protection; 3
■■ others.
No single study could accomplish all of the above aims. For example, if
the human challenge model system does not adequately mimic the wild-type
disease and the actual situation in which a vaccine would need to provide
protection, then a human challenge trial would not be usable as a primary basis
for licensure.
5. Study design of human challenge trials

As in all studies, the aim(s) of the human challenge trial guides the study design.
Consequently, even for the same disease, the challenge model may vary according
to the purposes and design of the study to be conducted. In some cases (for
example, to identify appropriate efficacy trial design and case definitions) the
challenge model may need to mimic wild-type disease as closely as feasible. In
other cases, consideration might be given to the use of an attenuated challenge
organism (for example, a previous vaccine candidate) or to a model system in
which objective early signs (for example, parasitaemia or viraemia) signal the
onset of disease. These signals could then trigger initiation of treatment to
prevent actual disease onset or morbidity. Such initiation of treatment should be
based on criteria pre-specified in the study protocol.
Another important consideration for a human challenge model system

would be its usefulness for positive or negative prediction. If used for down-
selection, de-risking or to identify vaccine candidates that would not warrant
advancement to large human efficacy studies, the degree of usefulness of the
model system for negative prediction should be high. If intended to be used
for evidence of vaccine efficacy, the degree of usefulness for positive prediction
might need to be almost as compelling and credible as for a traditional pivotal
efficacy trial. Whether the purpose of the study or studies is to provide supportive
evidence for licensure or to help inform and design traditional efficacy studies or
vaccine design, human challenge trials may contribute to the preponderance of
This might entail a challenge study in adults to extrapolate when children might need booster doses.
3
582
Annex 10
evidence upon which regulators could take a clinical trial or licensure decision.
Thus, the purpose of the study would influence its design, which would in
turn influence the conclusions and decisions that might be made by regulators
following consideration of the study results.
6. Operational aspects
In addition to general principles for all clinical trials in human subjects there are
some unique and important operational aspects to consider when conducting
a human challenge trial. Human challenge trials should be undertaken in
accordance with a protocol, and in special facilities that are designed and operated
in a manner that prevents the spread of the challenge organism to people outside
the study or to the environment. These clinical facilities should be capable of
providing continuous monitoring and medical attention at the appropriate
point(s) in time after the challenge is given. In addition to providing immediate
access to appropriate medical care and treatment, the facilities should be designed
to prevent the spread of disease, particularly when the challenge organism is a
GMO or an organism that is not endemic to the locality. These facilities may
need to be operated in a manner that permits all waste (including excrement) to
be collected and decontaminated before release. All staff, including janitorial and
administrative staff, might be required to work in personal protective equipment
appropriate for the pathogenicity of the challenge organism and its potential
hazard to the environment, and should be informed of the potential risks. It
should be noted that not all human challenge trials require such a high level of
control. When the challenge organism is attenuated and the wild-type organism
is likely to be present in the locality anyway, it may be adequate to conduct
human challenge trials in an outpatient setting or with appropriate procedures
to prevent spread. Examples of such approaches and procedures include the
use of BCG vaccine as a challenge organism, the use of bandaging to cover and
prevent spread from an intramuscular injection (assuming the organism is not
shed by other means) and the use of malaria challenge during winter months in a
temperate region. There may be other circumstances in which a human challenge
trial is undertaken, for example where the target organism of the vaccine to be
developed is not present in the location where the target group for its indication
lives (for example, in case of a traveller vaccine) – when the risk of spread of
the organism is low, human challenge trials using appropriate procedures could
be undertaken.
It may be necessary to ensure that controls and vaccinees are housed
together if an objective of the human challenge trial is to identify the potential
for transmissibility. In such a situation, only the vaccinees or unvaccinated
participants would be challenged, and the controls (who were not challenged)
583
would be monitored for evidence of acquiring the challenge organism through

contact with the challenged vaccinees. In this way, the transmissibility of the
challenge organism from challenged vaccinees may be determined. In order to
achieve the study objective of identifying transmissibility, it would be necessary
to conduct the study in-house even if the challenge organism was attenuated
and the wild-type organism was present in the locality.
It should be noted that human challenge trials have been, and can be,
successfully conducted in low- and middle-income settings. The same standards
would apply as in more developed countries. The investigators need to be
qualified, an independent ethics committee review is required, and assurance
of compliance with NRA requirements and regulations is needed. If relevant,
assurance of compliance with the national agency that regulates GMOs and/or
with local biosafety committees may also be needed. If a controlled inpatient
setting is required for the given study, this would also need to be in place.
7. Some key ethical considerations

Ethics in clinical trials include the precept of “minimizing risks to subjects and
maximizing benefits” and clinical trials should be designed and conducted
accordingly. Review of the proposed human challenge trial by an independent
ethics committee is essential. By their nature (that is, intentional infection of
humans with disease-causing organisms) human challenge trials would seem
to contradict this basic precept. Consideration must therefore be given to both
potential individual risks and benefits, as well as to potential societal risks and
benefits, such as the release into the environment of a pathogen that might not
otherwise be present. Provisions in clinical trial ethics are made for situations
in which there may be greater than minimal risk but no (or little) potential
for individual benefit when knowledge may be gained that benefits the larger
societal population with whom the potential trial participant shares significant
characteristics.
The ethical considerations concerning challenges in clinical trials
should be thoroughly evaluated. During a WHO Expert Consultation held in
January 2013 consideration was given to the way in which ethical principles
should be applied to vaccine trials. The main consultation topic concerned the
use of placebo in such trials, and a set of considerations for NRAs and ethics
committees was provided in the meeting report (4) and subsequently published
recommendations (5). Although specifically intended to facilitate review of
the proposed use of placebo in vaccine trials on a case-by-case basis these
considerations and recommendations are likely to have applicability to human
challenge trials.
It has to be acknowledged that in reality some individuals are greater
risk-takers than others, and that those who are risk averse would be unlikely to
584
Annex 10
accept the risk of receiving a challenge. The key to asking individuals to accept
the risk from a challenge study (in which they have little potential to receive
individual benefit) lies in the element of informed consent. Healthy adults may
consent when they are well informed and understand what the risks are that
they are agreeing to take – even if those risks may be considerably greater than
minimal (for example, accepting that they will develop an acute, but manageable,
disease that will resolve but in the meantime may cause considerable morbidity,
such as severe diarrhoea managed with fluid and electrolyte replacement). There
could be some potential for direct benefit should the trial participant become
immune to the disease caused by the challenge (or wild-type) organism but,
conversely, pre-existing immunity upon exposure to the wild-type organism in
the future may be harmful. Thus, in appropriate situations, it may be considered
ethical to ask healthy and informed adults to consent to volunteer and participate
in a human challenge trial whether they will receive an investigational vaccine
that may or may not protect them from the challenge organism, a placebo that
will not protect them or only the challenge organism itself. However, it is an
absolute requirement that accepting such risks and providing voluntary consent
are based upon being truly informed. For this reason (the absolute requirement
for truly informed consent) it is not deemed acceptable at this time to consider
conducting human challenge trials in children, or in any other vulnerable
population with diminished capacity to give informed consent. One possible
exception to this principle that might be considered would be a challenge model
that used a licensed live-attenuated vaccine as the challenge organism.
The need to minimize the risks to subjects in clinical trials calls for
investigators to give due consideration to whether the challenge organism needs
be pathogenic or not, or to what degree. As noted above, the aim or purpose of
the study may drive decisions on pathogenicity or attenuation, but the ethical
precept of minimizing risks to human subjects – to the maximum extent feasible
within the framework of sound science – should be given due consideration.
Key to such considerations is the credibility of the data to support regulatory
decision-making, which also needs to be taken into account when deciding how
pathogenic or attenuated the challenge organism needs to be.

The first draft of this WHO guidance document was prepared by Dr R. Sheets,
Consultant, Silver Spring (MD), the USA and Dr I. Knezevic, World Health
Organization, Switzerland, with inputs from: Dr J. McEwen, Therapeutic Goods
Regulatory Agency, England; and Dr V. Moorthy, World Health Organization,
Switzerland. The authors would like to acknowledge the publications of Dr M.T.
Aguado de Ros, ISGlobal, Spain; Dr B. Fritzel, BFL Conseils, France; Dr R. Sheets,
585
Grimalkin Partners, the USA; and particularly the report of the conference
organized by the International Alliance for Biological Standardization on
Human Challenge Trials in October 2014, which served as an important source
of information during the preparation of this document.
The proposal to prepare WHO guidance on human challenge trials was
developed during a WHO Consultation on clinical evaluation of vaccines held in
Geneva, Switzerland, 17–18 July 2014, and attended by the following participants:
Dr P. Annunziato, Merck & Co., the USA; Dr N. Bhat, Program for Appropriate
Technology in Health, the USA; Dr A. Chatterjee, Biological E Ltd, India; Dr K.
Chirgwin, Bill & Melinda Gates Foundation, the USA; Dr G. Coleman, Health
Canada, Canada; Dr D. Tuan Dat, The Company for Vaccines and Biological
Production No. 1 (Vabiotech), Viet Nam; Dr P.E. Fast, International AIDS
Vaccine Initiative, the USA; Dr G. Foglia, Sanofi Pasteur, the USA; Dr M. Gruber,
United States Food and Drug Administration, Center for Biologics Evaluation
and Research, the USA; Dr P.M. Heaton, Bill & Melinda Gates Foundation,
the USA; Dr D. Kaslow, Program for Appropriate Technology in Health,
the USA; Dr Y.H. Lee, Ministry of Food and Drug Safety, Republic of Korea;
Dr D.J.M. Lewis, University of Surrey, England; Dr A. Lommel, Paul-Ehrlich-
Institut, Germany; Dr J. McEwen, Therapeutic Goods Administration, Australia;
Dr P. Neels, Vaccine-Advice BVBA, Belgium; Dr M. Nijs, GlaxoSmithKline
Biologicals, Belgium; Dr S.A. Nishioka, Ministry of Health, Brazil; Dr A. Podda,
Novartis Vaccines Institute for Global Health, Italy; Dr M. Powell, Medicines and
Healthcare products Regulatory Agency, England; Dr A. Ramkishan, Central
Drugs Standard Control Organization, India; Dr R. Sheets, Consultant, Silver
Spring (MD), the USA; Dr J. Shin, WHO Regional Office for the Western Pacific,
Philippines; Dr P. Smith, London School of Hygiene and Tropical Medicine,
England; Dr J. Southern, Medicines Control Council, South Africa; Dr Y. Sun,
Paul-Ehrlich-Institut, Germany; Dr Z. Yang, Center for Drug Evaluation, China;
and Dr U. Fruth, Dr I. Knezevic, Mr O.C. Lapujade, Dr V. Moorthy, Dr K. Vannice
and Dr D. Wood, World Health Organization, Switzerland.

The resulting draft document was posted on the WHO Biologicals
website (as an appendix to the WHO Guidelines on clinical evaluation of
vaccines) for the first round of public consultation from 30 October to 30
November 2015.
The second draft was prepared by a WHO drafting group and posted
on the WHO Biologicals website (as an appendix to the WHO Guidelines on
clinical evaluation of vaccines) for a second round of public consultation from
1 February to 15 March 2016. Comments were received from: Dr B. Brock,
Sanofi Pasteur, the USA (provided the consolidated comments of the International
Federation of Pharmaceutical Manufacturers & Associations (IFPMA)); Dr K.
Farizo, United States Food and Drug Administration, Center for Biologics
Evaluation and Research, the USA; and Dr A. Rinfret, Health Canada, Canada.
586
Annex 10
At a WHO meeting of the Working Group on clinical evaluation of

vaccines held in Geneva, Switzerland, 3 May 2016 it was concluded that this
WHO guidance document should be provided as a separate document rather
than as an appendix to the WHO Guidelines on clinical evaluation of vaccines.
The meeting was attended by: Dr G. Coleman, Health Canada, Canada; Dr M.
Darko, Food and Drugs Authority, Ghana; Dr D. Etuko, National Drug Authority,
Uganda; Dr E. Griffiths, Consultant, Kingston-upon-Thames, England; Dr S.
Kennedy, University of Liberia, Liberia; Dr J. McEwen, Therapeutic Goods
USA; Dr J. Southern, Medicines Control Council, South Africa; Dr Y. Sun, Paul-
Ehrlich-Institut, Germany; Dr K. Zoon, National Institutes of Health, the USA;
and Dr I. Knezevic, World Health Organization, Switzerland.
Based on the comments received during the public consultation and on
the discussions of the above Working Group meeting, the document WHO/
BS/2016.2288 was prepared by Dr R. Sheets and Dr I. Knezevic.
The document was then posted on the WHO Biologicals website for
a third round of public consultation from 27 July to 16 September 2016 and
comments received from: Dr J. Auerbach and Dr A. Podda, GSK Vaccines
Institute for Global Health, Italy; Dr M. Gruber and Dr D. Pratt, United States
Food and Drug Administration, Center for Biologics Evaluation and Research,
the USA; Dr P. Njuguna, KEMRI Wellcome Trust Research Programme, Kenya;
and Dr P. Smith, London School of Hygiene and Tropical Medicine, England.
References
1. Sheets RL, Fritzell B, Aguado de Ros MT. Human challenge trials in vaccine development:
Strasbourg, September 29 – October 1, 2014. Biologicals. 2016;44(1):37–50 (abstract: http://www.
sciencedirect.com/science/article/pii/S104510561500113X, accessed 4 February 2017).
2. Knezevic I, Moorthy V, Sheets RL. WHO consultation on clinical evaluation of vaccines, 17–18 July
2014, WHO Headquarters, Geneva, Switzerland. Vaccine. 2015;33(17):1999–2003 (http://www.
sciencedirect.com/science/article/pii/S0264410X15001139, accessed 4 February 2017).
3. Guidelines on clinical evaluation of vaccines: regulatory expectations. In: WHO Expert Committee
on Biological Standardization: sixty-seventh report. Geneva: World Health Organization; 2017:
Annex 9 (WHO Technical Report Series, No. 1004).
4. Expert consultation on the use of placebos in vaccine trials. Meeting report. Geneva: World
Health Organization; 2013 (http://apps.who.int/iris/bitstream/10665/94056/1/9789241506250_
eng.pdf, accessed 4 February 2017).
5. Rid A, Saxena A, Baqui AH, Bhan A, Bines J, Bouesseau M-C et al. Placebo use in vaccine
trials: Recommendations of a WHO expert panel. Vaccine. 2014;32(37):4708–12 (http://www.
sciencedirect.com/science/article/pii/S0264410X14005374, accessed 4 February 2017).
587
Annex 11
Biological substances: WHO International Standards,
Reference Reagents and Reference Panels
The provision of global measurement standards is a core normative WHO
activity. WHO reference materials are widely used by manufacturers, regulatory
authorities and academic researchers in the development and evaluation of
biological products. The timely development of new reference materials is crucial
in harnessing the benefits of scientific advances in new biologicals and in vitro
diagnosis. At the same time, management of the existing inventory of reference
preparations requires an active and carefully planned programme of work to
replace established materials before existing stocks are exhausted.
The considerations and guiding principles used to assign priorities
and develop the programme of work in this area have previously been set out
as WHO Recommendations.1 In order to facilitate and improve transparency
in the priority-setting process, a simple tool was developed as Appendix 1 of
these WHO Recommendations. This tool describes the key considerations taken
into account when assigning priorities, and allows stakeholders to review and
comment on any new proposals being considered for endorsement by the WHO
Expert Committee on Biological Standardization.
A list of current WHO International Standards, Reference Reagents and
Reference Panels for biological substances is available at: http://www.who.int/
biologicals.
At its meeting in October 2016, the WHO Expert Committee on
Biological Standardization made the changes shown below to the previous list.
Recommendations for the preparation, characterization and establishment of international and other
1
biological reference standards (revised 2004). In: WHO Expert Committee on Biological Standardization:
fifty-fifth report. Geneva: World Health Organization; 2006: Annex 2 (WHO Technical Report Series, No. 932;
http://www.who.int/immunization_standards/vaccine_reference_preparations/TRS932Annex%202_
Inter%20_biol%20ef%20standards%20rev2004.pdf?ua=1, accessed 4 March 2017).
589
Additions2
Preparation Activity Status
Blood products and related substances
Ancrod 54 IU/ampoule Second WHO
International Standard
Batroxobin 50 U/ampoule First WHO Reference
Reagent
Blood coagulation factor FXI:C = 0.71 IU/ampoule Second WHO
XI (plasma, human) FXI:Ag = 0.78 IU/ampoule International Standard
Thromboplastin 1.11 IU/mL Fifth WHO International
(recombinant, human, Standard
plain)
Thromboplastin (rabbit, 1.21 IU/mL Fifth WHO International
plain) Standard
In vitro diagnostics
Zika virus RNA for 50 000 000 IU/mL First WHO International
NAT‑based assays* Standard
Ebola virus VP40 antigen Panel containing low- and First WHO Reference
medium-titre VP40 samples Panel
plus negative sample; no
unitage assigned
Dengue virus serotypes Four separate reference First WHO reference
1–4 RNA for NAT-based reagents with the following reagents
assays** assigned values:
DENV-1 RNA = 13 500 units/mL;
DENV-4 RNA = 33 900 units/mL
Hepatitis B virus DNA 5.98 log10 IU/mL Fourth WHO
for NAT-based assays International Standard
Unless otherwise indicated, all materials are held and distributed by the National Institute for Biological
2
Standards and Control, Potters Bar, Herts, EN6 3QG, England. Materials identified by an * in the above list
are held and distributed by the Paul-Ehrlich-Institut, 63225 Langen, Germany. Materials identified by an **
in the above list are held and distributed by the Center for Biologics Evaluation and Research, Food and
Drug Administration, Silver Spring, MD 20993, the USA.
590
Annex 11
Preparation Activity Status

Prolactin (pituitary, 67 mIU/ampoule Fourth WHO
human) International Standard
Janus kinase 2 V617F Panel of JAK2 V617F DNA First WHO Reference
gene mutation concentrations of 0, 0.03, 1.0, Panel
10.8, 29.6, 89.5 and 100%
591
SELECTED WHO PUBLICATIONS OF RELATED INTEREST

The World Health Organization was established in 1948 as a specialized agency of the
Sixty-sixth report.
United Nations serving as the directing and coordinating authority for international
WHO Technical Report Series, No. 999, 2016 (xix + 267 pages)
health matters and public health. One of WHO’s constitutional functions is to
provide objective and reliable information and advice in the field of human health, a WHO Expert Committee on Biological Standardization
responsibility that it fulfils in part through its extensive programme of publications. Sixty-fifth report.
WHO Technical Report Series, No. 993, 2015 (xvi + 262 pages)
The Organization seeks through its publications to support national health strategies
and address the most pressing public health concerns of populations around the world. WHO Expert Committee on Biological Standardization
To respond to the needs of Member States at all levels of development, WHO publishes Sixty-fourth report.
practical manuals, handbooks and training material for specific categories of health WHO Technical Report Series, No. 987, 2014 (xviii + 266 pages)
workers; internationally applicable guidelines and standards; reviews and analyses of
health policies, programmes and research; and state-of-the-art consensus reports that WHO Expert Committee on Biological Standardization
offer technical advice and recommendations for decision-makers. These books are Sixty-third report.
closely tied to the Organization’s priority activities, encompassing disease prevention WHO Technical Report Series, No. 980, 2014 (xv + 489 pages)
and control, the development of equitable health systems based on primary health
care, and health promotion for individuals and communities. Progress towards better WHO Expert Committee on Biological Standardization
health for all also demands the global dissemination and exchange of information Sixty-second report.
that draws on the knowledge and experience of all WHO’s Member countries and the WHO Technical Report Series, No. 979, 2013 (xiii + 366 pages)
collaboration of world leaders in public health and the biomedical sciences. WHO Expert Committee on Biological Standardization
To ensure the widest possible availability of authoritative information and guidance on Sixty-first report.
health matters, WHO secures the broad international distribution of its publications WHO Technical Report Series, No. 978, 2013 (xi + 384 pages)
and encourages their translation and adaptation. By helping to promote and protect WHO Expert Committee on Biological Standardization
health and prevent and control disease throughout the world, WHO’s books contribute Sixtieth report.
to achieving the Organization’s principal objective – the attainment by all people of the WHO Technical Report Series, No. 977, 2013 (viii + 231 pages)
highest possible level of health.
The WHO Technical Report Series makes available the findings of various international Fifty-ninth report.
groups of experts that provide WHO with the latest scientific and technical advice on WHO Technical Report Series, No. 964, 2012 (viii + 228 pages)
a broad range of medical and public health subjects. Members of such expert groups
serve without remuneration in their personal capacities rather than as representatives WHO Expert Committee on Biological Standardization
of governments or other bodies; their views do not necessarily reflect the decisions or Fifty-eighth report.
the stated policy of WHO. WHO Technical Report Series, No. 963, 2011 (viii + 244 pages)
For further information, please contact: WHO Press, World Health Organization, Website: http://www.who.int/biologicals
20 avenue Appia, 1211 Geneva 27, Switzerland (tel. +41 22 791 3264; fax: +41 22 791 4857;
email: bookorders@who.int; order on line: www.who.int/bookorders).
Further information on these and other WHO publications can be obtained from
WHO Press, World Health Organization, 1211 Geneva 27, Switzerland
(tel.: +41 22 791 3264; fax: + 41 22 791 4857; email: bookorders@who.int;
order online: www.who.int/bookorders)
This report presents the recommendations of a WHO Expert
Committee commissioned to coordinate activities leading to the
adoption of international recommendations for the production
1004
and control of vaccines and other biological substances, and the
establishment of international biological reference materials.
Following a brief introduction, the report summarizes a number
1004
of general issues brought to the attention of the Committee. The

next part of the report, of particular relevance to manufacturers
and national regulatory authorities, outlines the discussions
held on the development and revision of WHO Guidelines for
a number of vaccines, blood products and related substances.
Specific discussion areas included WHO guidance on the
production and evaluation of the quality, safety and efficacy
of monoclonal antibodies as similar biotherapeutic products
(SBPs); blood and blood components as essential medicines;
estimation of residual risk of HIV, HBV or HCV infections
via cellular blood components and plasma; snake antivenom
immunoglobulins; human pandemic influenza vaccines in
non-vaccine-producing countries; and clinical evaluation of
vaccines: regulatory expectations. In addition, the following
WHO guidance documents were also adopted: WHO manual
for the preparation of secondary reference materials for in vitro
diagnostic assays designed for infectious disease nucleic acid or on Biological
Standardization
antigen detection: calibration to WHO International Standards;
and Human challenge trials for vaccine development:
regulatory considerations. One WHO addendum document –
Labelling information of inactivated influenza vaccines for use
in pregnant women – was also adopted.
Subsequent sections of the report provide information on the
current status, proposed development and establishment of
international reference materials in the areas of: biotherapeutics Sixty-seventh report
other than blood products; blood products and related
substances; cellular and gene therapies; in vitro diagnostics;
and vaccines and related substances.
A series of annexes are then presented which include an
WHO Technical Report Series

updated list of all WHO Recommendations, Guidelines and
other documents on biological substances used in medicine
(Annex 1). The above nine WHO documents adopted on
the advice of the Committee are then published as part
of this report (Annexes 2–10). Finally, all additions and
discontinuations made during the 2016 meeting to the list of
International Standards, Reference Reagents and Reference
Panels for biological substances maintained by WHO are
summarized in Annex 11. The updated full catalogue of
WHO International Reference Preparations is available at:
http://www.who.int/bloodproducts/catalogue/en/.
ISBN 978 92 4 121013 3

2017 WHO TRS 1004 Biological Standardization

Uploaded by

Copyright:

Available Formats

2017 WHO TRS 1004 Biological Standardization

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2017 WHO TRS 1004 Biological Standardization

Uploaded by

Copyright:

Available Formats

This report presents the recommendations of a WHO Expert

Committee commissioned to coordinate activities leading to the

WHO Expert Committee on Biological Standardization

WHO Technical Report Series

ISBN 978 92 4 121013 3

WHO Expert Committee on Biological Standardization

WHO Expert Committee

© World Health Organization 2017

Dr M. Powell, Medicines and Healthcare Products Regulatory Agency, London, England

Dr H. Castro, Ministry of Health, Bogota, Colombia

Dr I. Sainte-Marie,26 Agence nationale de sécurité du médicament et des produits de

Representation from non-state actors

Dr S. Ramanan,33 Amgen, Thousand Oaks, CA, the USA

International Generic and Biosimilars Medicines Association (IGBA)

FXI blood coagulation factor XI

HPLC high-performance liquid chromatography

IVD in vitro diagnostic

RSV respiratory syncytial virus

For further details see: https://sustainabledevelopment.un.org/post2015/transformingourworld (accessed

to obtain maximum understanding and impact, and to facilitate consistent

Dr E. Griffiths (consulting); Dr P. Minor (public statements; research support); Professor S. Efstathiou

(research support); Dr J. Ferguson (investments); Dr R. Sheets (consulting); Dr J. Southern (consulting);

guidelines on the regulatory assessment of approved recombinant DNA (rDNA)

2.1.2 Vaccines and biotherapeutics: recent and planned

strategic issues. During the period 2013–2016 seven measurement standards

discussed separately (see section 3.3 below).

meeting of the WHO network of collaborating centres on standardization

Developments in methodology were recognized as an area where WHO

2.1.3 Blood products and in vitro diagnostics: recent and

a production stop for “Fav-Afrique” by Sanofi Pasteur. In order to address

Aspects of blood regulation were also covered in WHO workshops

performed in Guinea using convalescent plasma from Ebola virus (EBOV)

2.2.2 Report from the WHO network of collaborating centres on

of the Committee itself. However, proposals for replacement measurement

2.2.3 Report from the WHO network of collaborating centres

2.3 Feedback from custodian laboratories

National Institute for Biological Standards and Control

The Committee noted Dr Schneider’s comments and thanked him for

European Directorate for the Quality of Medicines &

its programme of work, the development of alternatives to animal experiments

assay is based on two commercial monoclonal antibodies a number of issues

Paul-Ehrlich-lnstitut (PEl), Langen, Germany

blood regulatory systems, which had included the development of a partnership

2.4 Cross-cutting activities of other WHO committees and groups

strengthening for medical products in general, including medical devices and

and highlighted the importance of emphasizing the responsibilities of device

2.4.2 Report of a WHO ad hoc consultation on International

This significantly increased level of activity in the biological/biotechnological

request. Substances for anticancer immunotherapy (so-called cancer vaccines)

2.4.3 Monographs on capreomycin sulfate and capreomycin

chromatographic and microbiological assays described in the above-mentioned

2.4.4 Yellow fever vaccine shortages and outbreak response

of vaccine is unknown. There was therefore a need to follow up on the recent

2.4.5 Vaccines for public health emergencies

3.1 Biotherapeutics other than blood products

3.2 Blood products and related substances

3.2.2 Africa Society for Blood Transfusion

Although not distributed to the Committee, a specific proposal to WHO

3.2.3 Guidelines on management of blood and blood

Following incorporation of the comments made the Committee