Anti-Angiogenic Therapy in Ophthalmology

Essentials in Ophthalmology
Series Editor: Arun D. Singh
Andreas Stahl Editor
Anti-
Angiogenic
Therapy in
Ophthalmology
Series Editors
Arun D. Singh
More information about this series at http://www.springer.com/series/5332

Andreas Stahl
Editor
Anti-Angiogenic Therapy
in Ophthalmology
Editor
Andreas Stahl, MD
Eye Center at the Medical Center
University of Freiburg
Freiburg, Germany
ISSN 1612-3212 ISSN 2196-890X (electronic)

ISBN 978-3-319-24095-4 ISBN 978-3-319-24097-8 (eBook)
DOI 10.1007/978-3-319-24097-8
Library of Congress Control Number: 2016937532
Springer Cham Heidelberg New York Dordrecht London

© Springer International Publishing Switzerland 2016
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made.
Printed on acid-free paper
Springer International Publishing AG Switzerland is part of Springer Science+Business Media

(www.springer.com)
Preface
In 2006, intraocular anti-VEGF therapy for exudative age-related macular

degeneration (AMD) was ranked among the top 10 breakthroughs of the year
by Science Magazine. Since then, antiangiogenic therapy has broadened its
impact from AMD treatment to various other diseases of the eye like macular
oedema in diabetic retinopathy or retinal vein occlusion. In other areas, for
example, retinopathy of prematurity (ROP), antiangiogenic therapy is just
beginning to find its place and is currently being evaluated in clinical studies
that weigh its benefit against potential risks. As a third category, there are
indications like macular telangiectasia where antiangiogenic therapy has
after initial hopeful use become to be seen as potentially unfavourable in the
long run.
Due to the broad use of antiangiogenic therapies in these fundamentally
different ocular diseases, it is crucial for the treating physician to understand
both the underlying principles of angiogenic eye diseases and the available
clinical data on therapies and outcome. This book therefore combines an
overview over retinal vascular physiology with a detailed analysis of the
available clinical data on antiangiogenic therapy in various ocular disorders.
The authors are all experts in their respective fields and have achieved to
combine concise but crucial pathophysiologic background information with
detailed clinical data reflecting our current state of knowledge on antiangio-
genic therapy in ophthalmology.
Freiburg, Germany Andreas Stahl, MD
v
Contents
1 Retinal Vascular Development ..................................................... 1

Jing Chen, Chi-Hsiu Liu, and Przemyslaw Sapieha
2 Retinopathy of Prematurity ......................................................... 21
Andreas Stahl, Ann Hellström, and Lois E.H. Smith
3 Anti-vascular Endothelial Growth Factor (VEGF)
Treatment in Neovascular Age-Related Macular
Degeneration: Outcomes and Outcome Predictors.................... 31
Dujon Fuzzard, Robyn H. Guymer, and Robert P. Finger
4 AREDS Supplementation and the Progression
Towards Exudative AMD ............................................................. 67
David J. Valent and Emily Y. Chew
5 VEGF-Inhibition in Macular Telangiectasia Type 2 ................. 79
Peter Charbel Issa and Frank G. Holz
6 Diabetic Retinopathy .................................................................... 89
Focke Ziemssen and Hansjürgen T. Agostini
7 Retinal Vein Occlusion.................................................................. 131
Amelie Pielen, Bernd Junker, and Nicolas Feltgen
8 Pharmacokinetics of Intravitreally Applied
VEGF Inhibitors ........................................................................... 139
Tim U. Krohne, Frank G. Holz, and Carsten H. Meyer
9 Neovascular Glaucoma ................................................................. 149
Julia Lüke, Matthias Lüke, and Salvatore Grisanti
10 Corneal Neovascular Diseases...................................................... 159
Deniz Hos, Felix Bock, Björn Bachmann, and Claus Cursiefen
11 Anti-Angiogenic Gene Therapy: Basic Science
and Challenges for Translation into the Clinic .......................... 173
Clemens Lange and James Bainbridge
Index ....................................................................................................... 189
vii
Contributors
Hansjürgen T. Agostini, MD Eye Center, Albert-Ludwigs-University

Freiburg, Freiburg im Breisgau, Germany
Björn Bachmann, MD Department of Ophthalmology, University of
Cologne, Cologne, Germany
James Bainbridge, PhD, FRCOphth Department of Genetics, UCL
Institute of Ophthalmology, London, UK
Felix Bock, PhD Department of Ophthalmology, University of Cologne,
Jing Chen, PhD Department of Ophthalmology, Boston Children’s Hospital,
Harvard Medical School, Boston, MA, USA
Emily Y. Chew, MD National Institutes of Health, National Eye Institute,
Bethesda, MD, USA
Claus Cursiefen, MD Department of Ophthalmology, University of
Nicolas Feltgen Universitätsmedizin Göttingen, University Eye Hospital,
Göttingen, Germany
Robert P. Finger, MD, MIH, PhD Department of Ophthalmology, Centre for
Eye Research Australia, Royal Victorian Eye and Ear Hospital, University of
Melbourne, East Melbourne, VIC, Australia
Dujon Fuzzard, MBBS/BMedSc Department of Ophthalmology, Centre
for Eye Research Australia, Royal Victorian Eye and Ear Hospital, University
of Melbourne, East Melbourne, VIC, Australia
Salvatore Grisanti, MD Department of Ophthalmology, University of
Lübeck, Lübeck, Schleswig-Holstein, Germany
Robyn H. Guymer, MBBS, PhD, FRANZCO Department of
Ophthalmology, Centre for Eye Research Australia, Royal Victorian Eye and
Ear Hospital, University of Melbourne, East Melbourne, VIC, Australia
ix
x Contributors
Ann Hellström, MD, PhD Department of Ophthalmology, Institute of

Neuroscience and Physiology, The Queen Silvia Children’s Hospital, The
Sahlgrenska Academy at University of Gothenburg, Gothenburg, Sweden
Frank G. Holz, MD Department of Ophthalmology, University of Bonn,
Bonn, Germany
Deniz Hos, MD Department of Ophthalmology, University of Cologne,
Peter Charbel Issa, MD, DPhil Department of Ophthalmology, University
of Bonn, Bonn, Germany
Bernd Junker, MD Hannover Medical School, University Eye Hospital,
Hannover, Germany
Tim U. Krohne, MD, FEBO Department of Ophthalmology, University of
Bonn, Bonn, Germany
Clemens Lange, MD, PhD Eye Center, University Hospital Freiburg,
Freiburg, Germany
Chi-Hsiu Liu, PhD Department of Ophthalmology, Boston Children’s
Hospital, Harvard Medical School, Boston, MA, USA
Julia Lüke, MD, PD Department of Ophthalmology, University of Lübeck,
Lübeck, Schleswig-Holstein, Germany
Matthias Lüke, MD, PD Department of Ophthalmology, University of
Lübeck, Lübeck, Schleswig-Holstein, Germany
Carsten H. Meyer, MD, FEBO, FMH Department of Ophthalmology,
Pallas Clinic, Aargau, Switzerland
Amelie Pielen, MD Hannover Medical School, University Eye Hospital,
Hannover, Germany
Przemyslaw Sapieha, PhD Departments of Ophthalmology and
Biochemistry, Maisonneuve-Rosemont Hospital Research Center,
Maisonneuve-Rosemont Hospital, University of Montreal, Montreal, QC,
Canada
Lois E.H. Smith, MD, PhD Department of Ophthalmology, Harvard
Medical School, Boston Children’s Hospital, Boston, MA, USA
Andreas Stahl, MD Eye Center at the Medical Center, University of
Freiburg, Freiburg, Germany
David J. Valent, DO National Institutes of Health, National Eye Institute,
Bethesda, MD, USA
Focke Ziemssen, MD Center for Ophthalmology, Eberhard-Karl University
Tübingen, Tübingen, Germany
Retinal Vascular Development
1
Jing Chen, Chi-Hsiu Liu, and Przemyslaw Sapieha
rior ciliary arteries, serve the optic nerve head,

1.1 Anatomy of Blood Vessel choroid, ciliary body, and iris (Gray 2008; Paul
Networks in the Eye Riordan-Eva 2011; Hayreh 2006).
To aid in understanding retinal vascular develop-

ment, we will first describe the origins of ocular 1.1.1 Retinal Vessels
blood vessels. The orbital vascular anatomy is
highly complex in human. The ophthalmic artery, The retina is one of the most structurally intricate
the first major branch of the internal carotid artery, and metabolically active tissues in the body. It
is the main source of the arterial supply to the orbit receives its blood supply from two sources: (1)
and its derived arterial structures. It passes beneath the central retinal artery and its three branched
the optic nerve and accompanies the nerve through plexi, which supplies the inner two-thirds of the
the optic canal into the inner wall of the orbit. The retina; and (2) the choriocapillaris (choriocapil-
central retinal artery, the first branch of the oph- lary layer) adjacent to the Bruch’s membrane
thalmic artery, pierces the optic nerve sheath infe- which supplies the outer retina. The central reti-
riorly about 8–15 mm (in humans) behind the nal artery and its accompanying vein run along
globe, and occupies a central position within the the inferior margin of the optic nerve sheath and
optic nerve when entering the retina. Other enter the eye through the optic disk. The vessel
branches of ophthalmic artery, including the poste- branches then immediately bifurcate into the
superior nasal and temporal, or the inferior nasal
and temporal branches, each supplying a distinct
J. Chen (*) • C.-H. Liu
quadrant of the retina. The branching pattern of
Department of Ophthalmology, Boston Children’s the vessels is either dichotomous or at right
Hospital, Harvard Medical School, 300 Longwood angles to the original vessel (Gray 2008; Paul
Avenue, Boston, Boston, MA 02115, USA Riordan-Eva 2011; Netter 2006). Branches from
e-mail: Jing.Chen@childrens.harvard.edu;
Chi-Hsiu.Liu@childrens.harvard.edu
the central retinal artery then dive into the retina
towards photoreceptors forming a capillary
P. Sapieha
Departments of Ophthalmology and Biochemistry,
plexus which provides nutrients to the inner reti-
Maisonneuve-Rosemont Hospital Research Centre, nal layers. The overall structure of retinal vessels
Maisonneuve-Rosemont Hospital, University of is composed of three distinct capillary layers, one
Montreal, 5415 Assomption Boulvard, Montreal, in the nerve fiber layer and the other two along
QC, Canada, H1T 2M4
e-mail: Mike.Sapieha@umontreal.ca
each sides of the inner nuclear layer (Fig. 1.1).
© Springer International Publishing Switzerland 2016 1

A. Stahl (ed.), Anti-Angiogenic Therapy in Ophthalmology,
Essentials in Ophthalmology, DOI 10.1007/978-3-319-24097-8_1
2 J. Chen et al.
Fig. 1.1 A schematic illustration of the ocular vascula- the intermediate and deep retinal vascular networks align
ture. Left: A schematic cross-section through an eye along each sides of the INL. The choroidal vessels
showing the retinal vasculature lining the inner surface of between RPE and sclera serve to supply blood to the outer
the retina and the choroid vessels. Right: An enlarged portion of the retina. GCL ganglion cell layer, INL inner
cross-sectional illustration of the eye showing detailed nuclear layer, NFL nerve fiber layer, ONL outer nuclear
structure of the retinal and choroidal vasculature. Three layer, RPE retinal pigment epithelium. Enlarged images
layers of retinal vessels are embedded among retinal neu- on the right depict retinal and choroidal vascular cast from
rons: the superficial retinal vasculature lies in the NFL; mouse eyes
Retinal vessels have a non-fenestrated endothe- layer, and the outermost Haller’s layer (Hartnett
lium forming the inner blood–retinal barrier. In 2013). The more outer the vessels are located in
addition, branches of the central retinal artery are the choroid, the bigger the size of their lumens.
terminal arteries that do not anastomose with While the outermost choroidal layer is com-
each other (Gray 2008; Netter 2006). Contrary to posed mainly of small arteries and veins, the
the inner retinal layers, the photoreceptor layer is innermost choriocapillaris is characterized by an
avascular without blood vessels from the central exceedingly fine capillary plexus adjacent to the
retinal artery. Thus, it relies on the choriocapil- Bruch’s membrane (Fig. 1.1) (Paul Riordan-Eva
laris to supply oxygen and nutrient by diffusion 2011; Ross and Pawlina 2005). In humans, the
from choroidal vessels. capillaries of the choriocapillaris are approxi-
mately 3–18 μm in diameter and oval shaped in
the posterior eye, becoming gradually wider
1.1.2 Choroidal Vessels (approximately 6–36 μm in diameter) and longer
(36–400 μm in length) as they move towards the
The choroid, a thin highly vascular membrane, equatorial region. The choriocapillaris is a sinu-
lies between the retina and the sclera and invests soidal vascular plexus with highly fenestrated
the posterior five-sixths of the globe in human. endothelium, and as the site of the greatest blood
The choroid vessels originate from two groups of flow in the body (Henkind et al. 1979), it pro-
branches of the ophthalmic artery: (1) the short vides 65–85 % of the blood volume in the eye.
posterior ciliary arteries, which supply the poste- Through diffusion, it nourishes the cells in the
rior portion of choroid; (2) the long posterior cili- outer portion of the retina (Bela et al. 2011),
ary arteries, which supply the anterior choroid, including the retinal pigment epithelium (RPE)
ciliary body, and iris. They are distinguished in and photoreceptors, as well as the fovea, which
three layers of choroidal vasculature: the inner- contains only photoreceptors for high acuity cen-
most choriocapillaris, the intermediate Sattler’s tral vision and is devoid of other retinal neurons.
1 Retinal Vascular Development 3
Interestingly, some mammalian species such eling, and maturation. Studies on retinal vascula-
as echidnas, guinea pigs, and rabbits lack retinal ture over the past several decades have greatly
vasculature, with their oxygen and nutrient sup- expanded our understanding of the fundamental
ply to the retina being solely provided by diffu- processes governing normal and pathologic vas-
sion from the choriocapillaris. It appears that the cularization including the relationship between
thickness of the retina is directly related to their hypoxia and vessel growth, as well as the contri-
evolutionary vascularization state. These avascu- bution of neurovascular interaction in vascular
lar retinas are typically thinner than the theoreti- homeostasis.
cal oxygen diffusion maximum of 143 μm,
whereas vascularized retinas are approximately
twice as thick, yet their avascular portion are still 1.2.1 Angiogenesis Is the Dominant
within the oxygen diffusion limit (Chase 1982; Process in Retinal Vascular
Buttery et al. 1991; Dreher et al. 1992). Development
Blood vessels are generally composed of sev-

1.1.3 Hyaloidal Vessels eral distinct cell layers with a single layer of
endothelial cells forming the lumen in the
The hyaloid vasculature is a transient embryonic innermost part of the vessel. In large macroves-
vascular bed which develops during embryonic sels such as aortae, the inner endothelial cell
and fetal stages to provide blood supply to the layer is covered by a central layer of mural
developing eye. The hyaloid artery originates cells/smooth muscle cells, and usually an exter-
from the ophthalmic artery. It enters the embry- nal layer consisting of connective tissue lined
onic fissure and extends through the vitreous to with small vessels and nerves. In microvessels
the lens. In the developing eye, the hyaloid and capillaries, which constitute most of the
vasculature plays an important role in many retinal vessels, the endothelial cell layer is cov-
aspects. It supplies the inner part of the retina ered externally by a noncontiguous single layer
with oxygen and nutrients; it is also involved in of pericytes/mural cells, allowing close interac-
the development and maturation of the lens and tion of vascular endothelial cells with surrounding
makes up the primary vitreous (Hartnett 2013; neurons, glia, and inflammatory cells to coordi-
Fruttiger 2007). During human fetal develop- nate the process of vascular growth, remodeling,
ment, the hyaloid vasculature is first seen at the and repair.
fourth week of gestation and reaches its maxi- The developmental vascularization process in
mum prominence during the ninth week. During the retina is mediated primarily via angiogenesis
mid-gestation, the hyaloid vasculature regresses (Fruttiger 2002), similarly as some other tissues
and the retinal vasculature contemporaneously such as the kidney and the brain. In angiogenesis,
develops. Regression of the hyaloid artery leaves vascular endothelial cells sprout and proliferate
a central extension from the optic disk to the pos- from preexisting blood vessels, usually venules,
terior lens surface, called the hyaloid canal or and develop into new vessels with fully func-
Cloquet’s canal (Hartnett 2013). tional lumen. During this process, local increases
in growth factors destabilize a portion of the pre-
existing vessels, allowing the activation of peri-
1.2 Development of Retinal cytes and remodeling of extracellular matrix.
Vasculature Endothelial cell migration and proliferation
subsequently occurs to form new vessels.
Among the three vascular beds in the eye, the Angiogenesis is also considered the dominant
retinal vasculature is the most extensively stud- process governing new blood vessel growth dur-
ied. The development of the retinal vasculature ing the wound healing process and in pathologic
has served as an excellent model for elucidating retinal vessel growth such as in tumors and reti-
the mechanisms of vascular development, remod- nopathies (Saint-Geniez and D’Amore 2004).
4 J. Chen et al.
This mechanism of angiogenic development is in layer in the inner plexiform layer, and a deep
contrast with vasculogenesis where dispersed layer in the outer plexiform layer (Dorrell et al.
primitive vascular precursor cells or hemangio- 2002) (Fig. 1.1).
blasts cluster together and form into tube-like In other mammals such as primates and
endothelial structures, in the absence of existing rodents, the conserved pattern of three retinal
vessels. Vasculogenesis occurs during the embry- vascular layers forms over varying timescales
onic development of the circulatory system and (Gariano and Gardner 2005). In mice, one of the
gives rise to the heart and the first primitive vas- most studied model systems of retinal vascular
cular plexus such as the yolk sac circulation. It development, the superficial vascular plexus
was suggested that the very initial process of vas- starts to develop during the first week after birth,
cular development in the retina results from vas- with radial growth as seen similarly in humans
culogenesis from resident angioblasts (McLeod (Fig. 1.2). During the second week, angiogenic
et al. 2006), then angiogenesis becomes domi- sprouts start to form from the superficial layer and
nant to form the rest of retinal vasculature. Yet grow perpendicular to the primary vascular plexus
with increasing evidences of circulating endothe- into the retina to create two deeper layers of capil-
lial precursor cells from bone marrow modifying lary networks. A complete vascular system is
developing and injured retinal blood vessels formed by the end of three weeks after birth
(Grant et al. 2002; Sengupta et al. 2003; Dorrell (Stahl et al. 2010a). Studies in the mouse retina
et al. 2004), the precise distinction between have shed light on the cross talk among multiple
angiogenesis and vasculogenesis is becoming cell types that function together to direct vascular
blurry. growth in the retina, and identified the important
roles of oxygen and oxygen-mediated growth
factors in this process.
1.2.2 Temporal and Spatial
Development of Three Layers
of Retinal Vasculature 1.2.3 Oxygen and VEGF in Retinal
Vascular Development
In humans, retinal vascularization starts in utero
at about 16 weeks of gestational age and is com- 1.2.3.1 Lack of Oxygen Drives Blood
pleted at approximately 40 weeks of gestation, Vessel Growth in the Eye
right before birth. Developmental retinal vascu- A hypothesized role of oxygen in retinal vascular
larization occurs concurrently as the hyaloid development originates from early observations
vessels regress. The retina is vascularized first that capillaries grow more profusely near venules
in the most superficial (i.e., innermost) layer on than around arteries (Michaelson 1948; Ashton
the vitreous side, starting from the optic nerve 1966; Wise 1961). Observed retinal vascular pat-
head and then progressing centrifugally out- terns from some eye diseases also support this
wards towards the ora serrata, the peripheral notion. In retinopathy of prematurity and diabetic
edge of the retina. This superficial primary retinopathy, an initial lack of retinal vessels with
plexus reaches the nasal side of the ora serrata at resulting retinal ischemia precedes pathologic
about 36 weeks gestational age, and the tempo- vessel growth, supporting the idea of hypoxia as
ral retina at approximately 40 weeks gestational a critical stimulator of new blood vessel growth
age. As the superficial layer is nearing comple- (Gariano and Gardner 2005). During develop-
tion, retinal vessels dive into the retina to form ment, as retinal neurons and glial cells differenti-
first a deep and then an intermediate layer along ate and mature, their metabolic demands increase,
with a well-organized network of inter-connecting creating a radial wave of hypothesized “physio-
vessels to complete three vascular layers: a logic hypoxia” that leads the development of new
superficial vascular layer which lies in the inner vessels from the center towards the periphery of
part of the nerve fiber layer, an intermediate the retina (Chan-Ling et al. 1995) (Fig. 1.2).
Fig. 1.2 A schematic illustration of the retinal vascular cytic template (green). Bottom panel: After completion
development in mice. Top panel: A scheme of retinal of the superficial vascular layer, blood vessels dive down
vascular growth that originates from the optic nerve towards the outer retina during the second week after
head after birth and grows radially towards the periph- birth, forming a deeper layer of vessels first, and an
eral of the retina, and reaches the edge of the retina intermediate vascular layer last. The whole retinal vas-
around postnatal day (P) 8. The growth of the superficial cular growth process is completed approximately three
vascular plexus of retinal vessels (red) follows a hypoth- weeks after birth in mice. In humans, this process occurs
esized physiologic hypoxia wave (blue) and an astro- prenatally during in utero development
After new vessels have formed, bringing oxygen Under normoxic conditions, HIF1α or HIF2α
and alleviating hypoxia, the vascular growth con- are hydroxylated on a proline residue by the
tinues radially towards the peripheral retina. prolyl hydroxylases, enabling its interaction
with von Hippel–Lindau protein, resulting in
1.2.3.2 VEGF Is A Dominant Growth its ubiquitylation and proteasomal degradation
Factor Governing Retinal (Jaakkola et al. 2001; Ohh et al. 2000; Epstein
Vascular Growth et al. 2001). Under hypoxic conditions, HIF1α or
The effect of oxygen on vascular growth is medi- HIF2α escapes prolyl hydroxylation and interacts
ated in large part by vascular endothelial growth with HIF1β to form a heterodimer, which then
factor (VEGF), a hypoxia-induced growth factor. translocates to the nucleus and activates target
VEGF is expressed in the developing retina in genes such as VEGF by binding to a hypoxia-
a pattern that coincides with retinal blood ves- responsive element in their promoter region
sel development both temporally and spatially (Wang et al. 1995). While a gradient of VEGF is
(Stone et al. 1995; Pierce et al. 1995, 1996). found preceding the superficial vascular plexus
VEGF and other hypoxia-induced growth fac- in the retina, it is also postulated that the rela-
tors are induced through the transcription fac- tive hypoxia of the deeper retinal layers during
tor hypoxia-inducible factor (HIF), a global development also results in a VEGF gradient that
regulator of O2 homeostasis (Wang et al. 1995; favors sprouting from the superficial layer down-
Wang and Semenza 1993a, b), composed of an wards, resulting in the formation of first the deep
oxygen-sensitive α unit and a constitutively and then the intermediate layers of capillary net-
expressed transcription activating β subunit. works (Stone et al. 1995; Pierce et al. 1995).
6 J. Chen et al.
In addition to VEGF, erythropoietin (Epo) is 1.2.3.3 Retinal Neuroglia Serves

another hypoxia-induced growth factor that plays as the Template of Retinal
an important role during retinal blood vessel Vascular Growth
homeostasis (Chen et al. 2008, 2009; Watanabe In order to form the stereotyped and precisely
et al. 2005). As a growth hormone that is best organized pattern of retinal vasculature, the dif-
known for its role in stimulating erythropoiesis, fused gradients of oxygen and hypoxia alone
Epo also has potent neuroprotective and pro- would be insufficient. Striking structural align-
angiogenic effects (Caprara and Grimm 2012), ment exists among retinal vessels, astrocytes, and
and was suggested important for the formation of neurons (Fig. 1.3a), suggesting the contribution of
the intermediate plexus of retina vessels in an cellular guidance mechanisms during retinal
HIF-dependent manner (Caprara et al. 2011). In angiogenesis. Astrocytes are a type of glial cells
the vitreous of patients with diabetic retinopathy formed from precursor cells that migrated into the
(Watanabe et al. 2005; Katsura et al. 2005), retina from the optic nerve head during embry-
retinopathy of prematurity (Sato et al. 2009) or onic development (Chu et al. 2001), prior to the
retinal vein occlusion (Stahl et al. 2010b), Epo is formation of retinal blood vessels. Astrocyte
upregulated along with VEGF. These oxygen- growth in the retina follows the radially oriented
regulated hypoxia-responsive factors are there- ganglion cell axons (Dorrell et al. 2002; Gariano
fore considered as some of the master regulators et al. 1996), which synthesize platelet-derived
of both normal retinal vessel development and growth factor (PDGF) to stimulate the growth and
proliferative retinopathies (Stone et al. 1995; alignment of astrocytes (Fruttiger et al. 1996),
Aiello et al. 1995; Smith et al. 1997). that express PDGF receptor alpha (PDGFR-α). A
Besides these hypoxia-regulated growth fac- complex interconnected astrocytic network forms
tors, some other non-oxygen-regulated growth at the inner surface of the retina after birth just
factors are also important for retinal vascular preceding the vessels. The spatially and tempo-
development, in part through modulation of the rally overlapping pattern of astrocyte network
VEGF response. For example, the Tie1–Tie2 preceding vascular growth led to the hypothesis
receptors are related tyrosine kinase receptors that astrocytes secrete VEGF and other growth
selectively expressed on vascular endothelial factors to guide the filopodia of sprouting endo-
cells and required for embryonic vascular devel- thelial cells towards the right direction in an
opment (Dumont et al. 1994; Sato et al. 1995). R-cadherin-dependent manner (Dorrell et al.
Angiopoietin2, a Tie2 ligand, promotes retinal 2002; Gariano et al. 1996; Watanabe and Raff
angiogenesis through increasing sensitivity to the 1988; Gerhardt et al. 2003) (Fig. 1.3b). However,
angiogenic effects of VEGF in retinal vessels some recent findings challenge this indispensable
(Hackett et al. 2002; Oshima et al. 2004), and the role of astrocytes in retinal vascular development
balance between the counteracting effects of (Weidemann et al. 2010; Scott et al. 2010), sug-
angiopoietin1 and angiopoietin2 dictates the con- gesting that astrocytic VEGF is perhaps more
sequent vascular response. In addition, the sig- critical for vascular maintenance in pathologic
nificant role of insulin-like growth factor-1 conditions. This indicates the possibility that
(IGF-1) in retinal vascular development was sup- other cells surrounding blood vessels, such as reti-
ported by the finding that IGF-1 is required for nal neurons and inflammatory cells, may also be
maximum VEGF activation of vascular endothe- crucial in governing retinal vascular development.
lial cell proliferation and survival pathways Both cell types are capable of secreting angio-
(Smith et al. 1997, 1999). In premature infants, genic factors including VEGF and will be dis-
low IGF-1 levels are associated with an increased cussed in the next sections.
risk of retinopathy of prematurity (Hellstrom
et al. 2001, 2002, 2003; Smith 2005; Lofqvist 1.2.3.4 Dynamics of Vascular Growth
et al. 2006), characterized by slowed, deficient As described above, retinal vessels elaborate
development of retinal vessels in its initial stage. their networks throughout the retina by angio-
Fig. 1.3 Structural alignment of retinal neuroglia and express platelet-derived growth factor alpha-receptor
vessels. (a) Three-dimensional reconstruction of an (PDGFR-α) and invade the developing retina from the
adult mouse retinal flatmount. The image depicts the optic nerve head, ahead of the vascular front. They travel
inner retinal vascular plexus (red; stained with Isolectin on top of PDGF-α-expressing retinal ganglion cells
B4) interwoven with retinal ganglion neurons and their (RGCs). Nascent vessels follow the astrocytic template
axons (green; stained with β-III tubulin). (b) Model of and form R-cadherin junctions with proximal astrocytes.
vascular growth during retinal development. Astrocytes, Adapted from Sapieha, Blood, 2012
genesis, a coordinated process involving endo- mokines, through a process modulated in part by
thelial cell proliferation, migration, and assembly microglia (Fantin et al. 2010; Rymo et al. 2011).
into tube-like structures containing a lumen As a consequence of angiogenic signaling, angio-
(Provis et al. 1997; Dorrell and Friedlander poietin2 stimulates pericyte detachment from the
2006). Typically, vessel growth is stimulated by vessel walls, leading to basement membrane deg-
angiogenic factors such as VEGF, fibroblast radation by matrix metalloproteinases (MMPs).
growth factors, angiopoietin2, and/or other che- Endothelial cells then slacken their VE-cadherin
8 J. Chen et al.
Fig. 1.4 Graphic depiction of the cellular mechanisms by stalk cells and antagonizes Dll4. This promotes
governing retinal angiogenesis. (a) A stable, quiescent VEGFR2 expression and renders ECs more responsive to
vessel with aligned endothelial cells (ECs), united by VEGF and thus more susceptible to form tip cells. (c)
VE-cadherin-rich junctions, and covered by pericytes. (b) Once the vessel has sprouted, stalk cells behind the tip cell
Upon stimulation by angiogenic factors, a cascade of divide and assemble to form the lumen of the neovessel.
events ensues leading to pericyte detachment, basement Pericytes are then recruited and basement membrane is
membrane degradation and endothelial junction loosen- laid down. (d) To reach its final destination, the growing
ing. Determination of stalk versus tip cell phenotypes is neovessel must navigate through the tissue by responding
achieved through Notch-dependent signaling. VEGF to a series of diffusible and membrane-bound attractive
through VEGFR2 induces the Notch ligand Dll4 on tip and repellent guidance cues. (e) Confocal image of sprout-
cells which subsequently activates Notch in adjacent ing retinal vessels with filopodia-rich tip cells (stained
endothelial cells and specifies their stalk cell phenotype. with isolectin B4) from P4 mouse retinas. Scale bar:
Conversely, the Notch ligand Jagged1 is highly expressed 10 μm. Adapted from Sapieha, Blood, 2012
and claudin-rich junctions and a leading tip cell nal guidance receptors include the neuropilins
protrudes and advances towards the attractant and plexins (for semaphorins), Unc5b, neogenin
angiogenic gradients (reviewed in (Carmeliet and and DCC (for netrins), the Eph receptors (for
Jain 2011)) (Fig. 1.4a–c). ephrins), and roundabouts (for slits) (Adams
Growth of nascent blood vessels occurs via et al. 1999; Klagsbrun and Eichmann 2005;
the advancement of leader tip cells (Gerhardt Wilson et al. 2006) (Fig. 1.4d–e). Once the tip
et al. 2003), with the leading tip position cells come in contact with a given cue, they will
sometimes shared by several endothelial cells. respond by either advancing, stalling, turning, or
Endothelial tip cells, which are induced by retracting depending on which cell surface recep-
VEGF, are non-mitotic and enriched in both tor predominates and the overall intracellular
receptors for angiogenic factor (such as VEGFR) environment of the tip cells (Larrivee et al. 2009).
and other receptors that were initially described The role of neuronal guidance cues in vascular
to respond to neuronal guidance cues (Huber growth has been comprehensively reviewed
et al. 2003). The physiological role of tip cells is (Larrivee et al. 2009; Carmeliet and Tessier-
to probe the tissue environment through their Lavigne 2005; De Smet et al. 2009; Gelfand et al.
projecting filopodias and guide the nascent vessel 2009). Conversely, the stalk cells, stimulated by
towards growth factor gradients or guidance cues Dll4/Jagged1/Notch-mediated later cellular inhi-
to its appropriate final destination. These neuro- bition (Hellstrom et al. 2007; Benedito et al.
2009), are located behind tip cells in the wake of metabolic activity from the newly operating neu-
the vascular front, and divide in response to rons and their need for fuel (Cringle et al. 2006).
VEGF to form a lumen. In addition, cell-specific ablation of VEGF,
HIF-1α, or HIF-2α in astrocytes has no overt
defects on developmental retinal vascularization
1.2.4 Neurovascular Cross Talk (Weidemann et al. 2010; Scott et al. 2010), while
in Retinal Vascular neuroretinal-specific knockout of HIF-1α sub-
Development stantially perturbs retinal vascular development
(Caprara et al. 2011; Nakamura-Ishizu et al.
The retina is the most metabolically active tissue 2012) suggesting that a neuronal cell population,
of the body, per weight. It is also the most exter- rather than astrocytes, likely provides angiogenic
nal and accessible portion of the central nervous factors during retinal development. One such
system (CNS). Central neurons such as those that neuronal cell population is retinal ganglion cells
populate the retina require a steady supply of (RGCs). Among retinal neuron cell types, RGCs
nutrients and oxygen to ensure appropriate neu- are the most anatomically coupled with the
ronal function and sensory transmission. superficial retinal vascular plexus. A specific role
Consequently, nervous and vascular systems for RGCs in vascular development was estab-
must be adequately paired. In recent years, it has lished using mouse models of genetic ablation of
become clear that neurons play an important role RGCs (Edwards et al. 2011; Sapieha et al. 2008).
in instigating, promoting, and steering angiogen- In transgenic mice which express a toxin in newly
esis within nervous tissue and specifically in the formed RGCs and hence eliminate RGCs as they
retina (Edwards et al. 2011; Fukushima et al. form (Mu et al. 2005), astrocytic networks remain
2011; Kim et al. 2011; Nakamura-Ishizu et al. largely intact, yet these mice are completely
2012; Sapieha et al. 2008). devoid of a retinal vascular plexus and show per-
sistent hyaloid vasculature (Sapieha et al. 2008).
1.2.4.1 Influence of Retinal Ganglion Similarly, Math 5−/− mice which lack 95 % of
Cells on Vascular Growth RGCs do not form a functional retinal vascular
There is a current debate on whether astrocytes or layer (Edwards et al. 2011).
alternatively, retinal neurons such as retinal gan-
glion cells are primary drivers of retinal develop- 1.2.4.2 Photoreceptors and Retinal
mental vascular growth, since evidence has been Vascular Growth
provided for both. One possible interpretation to In addition to RGCs, photoreceptors may also play
reconcile the contribution of both cell popula- a significant role in determining retinal energy
tions during retinal vascular plexus formation demand and hence influence vascular growth.
suggests astrocytes contribute to trophic support Photoreceptors are the most abundant neuronal
of vessels through providing a template for cell population in the retina and also require very
growth (Fruttiger et al. 1996; Gerhardt et al. high rates of oxygenation (Wangsa-Wirawan and
2003; Uemura et al. 2006) as described above, Linsenmeier 2003). While the direct impact of
while retinal neurons may drive vascular growth photoreceptors on retinal vascular development
to ensure their own metabolic support (Sapieha is unclear given their contemporaneous matura-
2012). This view is supported by findings in the tion and migration towards the outer retina dur-
human fetus during the formation of the outer ing retinal vascular formation, evidence for a role
vascular plexus of the retina at 25–26 weeks of of photoreceptor in vascular maintenance comes
gestation (Hughes et al. 2000). Interestingly, this directly from clinical observations. Patients suf-
time frame coincides with the first appearance of fering from both proliferative diabetic retinopathy
visually evoked potentials and thus functional and retinitis pigmentosa with progressive photo-
neurons (Dreher and Robinson 1988) and likely receptors loss show considerably less pathologi-
reflects an increase in oxygen consumption and cal retinal angiogenesis than diabetic patients
10 J. Chen et al.
with healthy photoreceptors (Sternberg et al. Monori et al. 1985). When succinate levels rise,
1984). Furthermore, preretinal neovasculariza- they activate GPR91 with a half-maximal
tion associated with late-stage diabetes mellitus response of 28–25 μM (He et al. 2004) and
has been reported to spontaneously regress with prompt release of VEGF and angiopoietins 1 and
the onset of retinitis pigmentosa (Lahdenranta 2. When succinate levels rise substantially, they
et al. 2001). Similarly, mice with genetically can stabilize HIF with a Ki between 350 and
ablated photoreceptors failed to mount reactive 460 μM, further promoting secretion of HIF-
retinal neovascularization in a model of oxygen- dependent angiogenic growth factors. Together
induced proliferative retinopathy (Lahdenranta this suggests that the succinate-GPR91 axis can
et al. 2001), and diabetic mice that had photo- act as an early sensor of hypoxic stress and work
receptor degeneration showed lower levels of to enhance regional circulation, and once
retinal pro-angiogenic VEGF (de Gooyer et al. hypoxia is sustained, succinate levels rise further
2006). These observations likely stem from the and stabilize HIF, which then activates a collec-
high energetic requirements of photoreceptor tion of pathways to non-discriminately provoke
neurons (Linsenmeier and Padnick-Silver 2000) angiogenesis.
and hence loss of photoreceptors substantially Another potential pathway that retinal neu-
reduces the energy demand of the retina, reduc- rons may employ to reinstate local circulation is
ing the stimuli for angiogenic factors and conse- via the metabolically regulated protein deacety-
quent pathologic neovascularization. lase sirtuin-1 (Sirt-1). Sirt1 is an NAD+-dependent
protein deacetylase critical for neuronal function
1.2.4.3 Neuronal Energy Metabolism in the CNS (Michan 2013; Michan et al. 2010),
May Underlie Retinal and can be activated under stress condition to
Angiogenesis regulate cell cycle and longevity (Brunet et al.
One potential mechanism by which the energy 2004). When retinal ganglion neurons sense tis-
status in retinal neuron may modulate vascular sue ischemia and hypoxia such as during an isch-
network development is likely evolutionarily emic retinopathy, they induce increased levels of
conserved and directly employs intermediates of Sirt1 (Chen et al. 2013). In mice with neuronal
energy metabolism, such as those from glucose, deficiency of Sirt1, the reparative angiogenesis
the main energy source of CNS including reti- that takes place during the mouse model of isch-
nas. By activating cognate G-protein-coupled emic retinopathies is significantly attenuated.
receptors, energy metabolites such as lactate These effects are likely mediated through Sirt1-
(Ahmed et al. 2010), α-ketoglutarate and succi- mediated deacetylation of HIF which helps main-
nate (He et al. 2004) have documented physio- tain elevated levels of HIF and its consequent
logical roles beyond ATP production. In this reparative angiogenic growth factors (Chen et al.
regard, cellular signaling events triggered via 2013). Together, these studies, underscore the
energy metabolites in response to a compro- propensity of retinal neurons to govern their vas-
mised energy status in retinal neurons have been cular environment and reciprocally regulate their
proposed as a contributor to both physiological metabolic supply.
and pathological retinal vascularization (Sapieha
et al. 2008; Grant et al. 1999, 2001; Mino et al.
2001). This concept has been elucidated for the 1.2.5 Immune Influence on Retinal
Krebs cycle intermediate succinate and its recep- Vessel Development
tor GPR91 (Sapieha et al. 2008). Succinate accu-
mulates under conditions of hypoxic stress by a As blood vessels develop throughout the inner
mechanism involving feedback inhibition of retina, they are accompanied by microglia, the
succinate dehydrogenase by nonoxidized flavin primary resident immune cells of the retina
and nicotinamide nucleotides and by reactive (Fig. 1.5). Microglia are myeloid-derived tissue
oxygen species (Gutman et al. 1980; Meixner- macrophages of the CNS which populate the
Fig. 1.5 Microglia and retinal vascular development. between vessels (stained in red) and microglia lasts
Immunohistochemistry on developing mouse retinas throughout development of the various vascular layers
demonstrates that microglial cells (stained in green) are that make up the retina. P postnatal day
associated with nascent vasculature. This relationship
mammalian retina before it becomes vascular- mation in preterm infants (Tremblay et al.
ized and become activated in response to injury 2013). In this regard, neonatal sepsis is being
(Dejda et al. 2014). Retinal microglia are rapidly recognized as a major risk factor for developing
activated after an inflammatory insult (Santos severe retinopathy of prematurity (Klinger et al.
et al. 2008; Chen et al. 2002), yet during develop- 2010; Tolsma et al. 2011; Lee and Dammann
ment, there is increasing evidence that they asso- 2012). Experimentally induced systemic inflam-
ciate with growing vessels and play a key role in mation mimicking a sepsis-like state in mice
promoting normal vascularization. provokes a profound increase in the number of
Evidence for the major contribution of microg- activated microglia in intimate association with
lia to retinal vascular growth is severalfold. the nascent vascular plexuses of the retina, and
Microglia associate directly with nascent vessels leads to compromised vascular development
at the vascular front and modulate angiogenesis and formation of dense, almost neovascular
(Checchin et al. 2006; Kubota et al. 2009; Stefater membrane-like structures on the retina
Iii et al. 2011). They are found apposed to devel- (Tremblay et al. 2013). This aberrant vascular
oping vessels during human fetal development, development leads to compromised retinal func-
and pharmacologic depletion of retinal microglia tion in later life in mice as determined by elec-
slowed retinal vessel development (Checchin troretinograms (Tremblay et al. 2013). Hence,
et al. 2006). Similarly, osteopetrotic csf-1op/op while required for normal stereotyped vascular
mice, which harbor an inactivation mutation in development, over-activation of retinal microglia
the csf-1 gene and are deficient in retinal microg- may perturb retinal angiogenesis and lead to
lia in the first week of postnatal development, vascular sequelae similar to those observed in
show significantly sparser retinal vessel networks patients with ROP.
when compared to wild-type controls (Rymo
et al. 2011). This may be partially attributed to
the vascular anastomosis-promoting properties of 1.2.6 Remodeling of Retinal
microglia (Smith et al. 1997) and their ability to Vascular Growth
regulate vessel branching (Stefater Iii et al. 2011).
Given the inherent role of microglia in Once fully formed, the superficial layer of retinal
defending the retina against foreign intrusion, vessels is remodeled before sprouting to the
they may also partake in perturbing vascular deeper layer (Ishida et al. 2003) as regulated by
development during bouts of systemic inflam- the specific needs of the local tissue for oxygen
12 J. Chen et al.
and nutrients. Maturation and pruning of the maturation of the choroidal vasculature appears
superficial plexus leads to a less dense network to be tightly associated with signals produced by
with finer vessels compared to the structures RPE cells (Zhao and Overbeek 2001). In particu-
formed during early development. Relative lar, VEGF and its receptor are highly expressed
hyperoxia around arteries is considered an impor- by RPE cells at the time of choriocapillaris for-
tant factor for pruning of existing vasculature. mation (Gogat et al. 2004; Yi et al. 1998), sug-
Local hyperoxia would lead to local suppression gesting involvement of RPE-derived VEGF in
of VEGF, which is an endothelial survival factor the choroidal vascular development. Furthermore,
for immature capillaries. Reduction of VEGF in colobomas, failure of RPE differentiation
leads to regression of these immature capillaries results in defective development of the choroid
in areas with excess vasculature (Claxton and and sclera (Torczynski 1982). In addition, in
Fruttiger 2003; Alon et al. 1995). Leukocytes patients with exudative age-related macular
from the circulation also participate in vessel degeneration, degeneration of RPE cells and
remodeling through induction of endothelial cell associated overproduction of VEGF leads to
apoptosis, a process considered most relevant in pathological activation and sprouting of the nor-
the context of early vessel loss in diabetic reti- mally quiescent choriocapillaris; all pointing
nopathy (Ishida et al. 2003). towards an essential role of VEGF in mediating
choroidal vessels growth during development
and in pathologies.
1.3 Development of Choroid
and Hyaloid Vessels
1.3.2 Hyaloidal Regression
1.3.1 Choroidal Development
During early eye development, the hyaloid artery
In humans, a definitive choriocapillaris layer penetrates into the optic cup and extends through
appears at around 8–12 weeks of gestational age, the vitreous to form branches which envelop the
prior to the formation of retinal vessels. The developing lens. As retinal vasculature forms,
primitive choroidal vascular system arises from the hyaloid vessel regresses (Saint-Geniez and
the surrounding mesoderm during optic cup D’Amore 2004; Zhu et al. 2000), and eventually
development. Following the development of disappears upon completion of eye develop-
RPE, this primitive choroidal vascular layer con- ment, usually before 34 weeks of gestation in
tinues to expand (Saint-Geniez and D’Amore humans.
2004), until the capillaries completely encircle The mechanisms governing both formation
the optic cup. and regression of hyaloid vessels are still incom-
The mechanisms of choroidal vascular devel- pletely defined. VEGF was suggested to trigger
opment overall are less well studied and under- the growth of hyaloid vessels because VEGF is
stood than those related to the development of the expressed in the portion of the lens closest to the
retinal vasculature. Clinically, the retinal vascula- forming vessels (Gogat et al. 2004; Mitchell et al.
ture can be clearly observed funduscopically in 1998; Shui et al. 2003). In addition, patients with
humans, while visualization of the choroidal VEGF-induced endothelial cell hyperplasia also
plexus is blocked by a dense layer of RPE. In ani- have disorganized and persistent hyaloid vessels,
mal models, retinal vasculature development can further supporting a role of VEGF in hyaloid for-
be easily studied due to the postnatal develop- mation (Ash and Overbeek 2000).
ment of retinal vessels in rodents and advanced Recently, an entirely different process has
imaging techniques on retinal flat mounts which been uncovered for hyaloid regression and retinal
expose retinal vessels. In contrast, choroidal vas- vascular formation from studies investigating a
culature is less readily quantifiable during devel- group of rare hereditary eye diseases: Norrie dis-
opment. Nevertheless, the formation and ease and familial exudative vitreoretinopathy
(FEVR). A common feature in these diseases is receptors constitute the outer BRB to block
that hyaloid vasculature fails to regress, resulting leakage from the choroid into the retina
in persistent fetal vasculature. This is associated (Fig. 1.6a). The inner BRB is formed by tight/
with incomplete formation of retinal vasculature adherens junctions among non-fenestrated retinal
with a lack of deep retinal plexus growth (Rehm vascular endothelial cells and the surrounding
et al. 2002), as well as a poorly developed neural pericytes and glia end feet (Fig. 1.6a, b). Similar
retina and impaired vision. These genetic dis- to the outer BRB, the inner BRB maintains bar-
eases arise from a number of mutations in an rier integrity (Janzer and Raff 1987) to prevent
interrelated Wnt/β-catenin signaling pathway: leakage from retinal blood vessels.
Wnt ligands Norrin or Wnt receptors Frizzled4 Two types of junctions, tight and adherens
and Lrp5. These works suggest a critical role for junctions, maintain the proper function of BRB
Wnt signaling in hyaloid regression in part (Fig. 1.6c). The tight junctions are the most api-
through macrophage-mediated endothelial cell cally located complexes that restrict intracellular
apoptosis (Xu et al. 2004; Niehrs 2004; Lobov movement of solute and fluid from entering the
et al. 2005). Yet development of retinal vessels retina between the vascular endothelial cells in
also depends in part on Wnt signaling as lack of inner BRB, or RPE cells in outer BRB. The three
Wnt activation results in delayed development of major transmembrane proteins involved in the
primary vascular plexus and deficient intermediate formation of tight junctions are occludin, clau-
and deep layer vessel formation in the retina (Xu din, and junctional adhesion molecule (JAM).
et al. 2004; Ye et al. 2011; Chen et al. 2011, The extracellular domains of these transmem-
2012). Beyond development, a pathological role brane proteins cross-interact to seal the intercel-
for Wnt signaling is suggested in destructive neo- lular space between neighboring cells while the
vascular growth in retinopathies (Chen et al. cytoplasmic domains of these proteins are linked
2011; Ohlmann et al. 2010). In addition, a critical to actin filaments via zonula occludens proteins
role for Wnt signaling in maintaining the integ- (ZOs) (Ross and Pawlina 2005; Runkle and
rity of blood–brain barrier, and potentially blood– Antonetti 2011) for stable anchoring with the
retinal barrier was proposed, likely through cytoskeleton. The more basally located adherens
mediation of tight junction proteins (Chen et al. junctions are composed of cadherins and provide
2011, 2012; Liebner et al. 2008; Daneman et al. lateral adhesion between cells. The extracellular
2009; Wang et al. 2012). portions of cadherins interact with similar
domains from the neighboring cells; on the cyto-
plasmic side, cadherins are bound to α, β, or
1.4 Formation of Blood–Retinal γ-catenins resulting in the formation of cadherin-
Barrier catenin complexes which are bound to actin fila-
ments by vinculin (Ross and Pawlina 2005;
As blood vessels mature in the retina, a func- Runkle and Antonetti 2011). Together, the tight
tional blood–retinal barrier (BRB) starts to form and adherens junctions restrict the passage of
at 26–34 weeks gestation to prevent toxins, molecules of a certain size through the cells to
pathogens, and other large molecules from reduce permeability and maintain BRB. VEGF is
entering and harming the retina (Choi et al. 2007). considered a main inducer of vascular permeabil-
Similar to the blood–brain barrier, the BRB is ity, since it was initially identified as a vascular
vital to preserve the integrity and function of the permeability factor (Senger et al. 1983). VEGF
retina. In the outer retina, the choriocapillaris is may promote breakdown of BRB through down-
highly fenestrated to allow blood flow beneath regulating tight junction proteins essential for
the basal surface of the RPE, which digests shed cellular adhesion (Argaw et al. 2009).
photoreceptor outer segments and clears other Alteration of BRB plays a crucial role in
photoreceptor waste. The RPE with tight junc- many vascular eye diseases, where breakdown
tions on the apical surface adjacent to the photo- of BRB can lead to exudate, hemorrhage, and
14 J. Chen et al.
Fig. 1.6 Compositions of blood–retinal barrier (BRB). (a) in the inner BRB, or between RPE cells in the outer BRB.
Schematic illustration of the localization of inner and outer The apical side illustrates the tight junction, including
BRB in the eye. The inner BRB is mainly composed of junction adhesion molecules (JAMs), occludin, and clau-
retinal vascular endothelial cells (ECs) and the end feet of din, all transmembrane proteins with their extracellular
surrounding pericytes and glia. The outer BRB is estab- domains on adjacent cells joining each other to form a
lished by retinal pigment epithelial (RPE) cells. The BRB sealing barrier. The intracellular domain of these tight
plays a fundamental role in maintaining the specialized junction molecules are complexed to the scaffolding adap-
environment in the neural retina. (b) A magnified view tor protein zonula occludens (ZO), which is in turn
showing detailed composition of the inner BRB. The reti- anchored to the actin cytoskeleton. The adherens junction,
nal vascular ECs with junction complexes between neigh- usually on the more basal side than tight junctions, is
boring ECs form the main structure of the inner BRB, mainly composed of cadherin, transmembrane proteins
resting on the basement membrane which is surrounded by that form dimers on adjacent cells. Cadherin is partnered
pericytes and Müller glial cell foot processes. (c) An over- with α, β, or γ-catenins, the complex of which is again con-
view of the junction complexes between endothelial cells nected to the actin cytoskeleton network via vinculin
retinal detachment. The inner BRB is compro- common retinal vascular disease, is also associ-
mised in diabetic retinopathy, leading to retinal ated with plasma leakage and hemorrhage in its
edema and often severe visual impairment. In initial stage due to increased intraluminal pres-
exudative age-related macular degeneration, in sure in the obstructed vein; and in later stages
contrast, it is the outer BRB that is compromised due to VEGF-induced proliferation of leaky
when choroidal blood vessels (fenestrated without neovessels and VEGF-induced increased perme-
tight junctions) grow through Bruch’s membrane ability of normal vessels.
and into the subretinal space, breaking through While breakdown of the BRB poses a signifi-
the RPE’s outer BRB. Increased levels of vitre- cant problem in eye diseases, on the other hand,
ous VEGF in both diseases may play a signifi- the presence of BRB also represents a unique chal-
cant role in inducing vascular permeability in lenge in ocular drug delivery, as it blocks systemic
these conditions. Retinal vein occlusion, another delivery of large molecular therapeutic agents to
the eye. Identifying ways to modulate BRB both

ways will promote not only better treatment of Compliance with Ethical Requirements
ocular vascular leakage but also discovery of new No human studies were carried out by the
methods for improved ocular drug delivery. authors for this article. For animal studies,
all institutional and national guidelines for
the care and use of laboratory animals were
1.5 Summary followed. The authors received research
grants from NIH/NEI (R01 EY024963),
In summary, studies conducted over the last sev- BrightFocus Foundation, Boston Children’s
eral decades on retinal vascular development Hospital (BCH) Ophthalmology Foundation,
have uncovered important insights into the fun- BCH career development award, Mass
damental cellular and molecular processes that Lions Eye Research Fund Inc., and Alcon
govern blood vessel growth in the eye. These dis- Research Institute (J.C.). The authors
coveries provided the much needed basis for declare no conflict of interest.
understanding the pathogenesis of neovascular
eye diseases and the development of antiangio-
genic therapies. In particular, identification of the
role of VEGF in retinal vascular development
greatly facilitated the development of the first References
pharmacological treatment for exudative age-
related macular degeneration—anti-VEGF ther- Adams RH, et al. Roles of ephrinB ligands and EphB
apy. Currently, there are many active areas of receptors in cardiovascular development: demarcation
of arterial/venous domains, vascular morphogenesis,
research on retinal vascular development and and sprouting angiogenesis. Genes Dev. 1999;13:
pathologic ocular angiogenesis, which will con- 295–306.
tinue to expand our knowledge on the roles of Ahmed K, et al. An autocrine lactate loop mediates
cellular interaction within the eye and the signal- insulin-dependent inhibition of lipolysis through
GPR81. Cell Metab. 2010;11:311–9.
ing pathways and cellular mechanisms involved. Aiello LP, et al. Suppression of retinal neovascularization
Moreover, there is increasing interest towards in vivo by inhibition of vascular endothelial growth
exploring the impact of cellular energy metabo- factor (VEGF) using soluble VEGF-receptor chimeric
lism, both in endothelial cells (De Bock et al. proteins. Proc Natl Acad Sci U S A. 1995;92:
10457–61.
2013) and in retinal neurons, as one of the funda- Alon T, et al. Vascular endothelial growth factor acts as a
mental driving forces of vascular growth. survival factor for newly formed retinal vessels and
Repercussions of elucidating neurovascular ener- has implications for retinopathy of prematurity. Nat
getics within the eye may have broad implica- Med. 1995;1:1024–8.
Argaw AT, Gurfein BT, Zhang Y, Zameer A, John
tions for a range of neurodegenerative diseases GR. VEGF-mediated disruption of endothelial CLN-5
beyond the retina. In addition, valuable biologi- promotes blood-brain barrier breakdown. Proc Natl
cal insights can be gained through studies of rare Acad Sci U S A. 2009;106:1977–82.
inherited vascular eye diseases with defective Ash JD, Overbeek PA. Lens-specific VEGF-A expression
induces angioblast migration and proliferation and
retinal vascular development, which may also stimulates angiogenic remodeling. Dev Biol.
generate important implication for other more 2000;223:383–98.
prevalent vascular eye disorders such as ROP, Ashton N. Oxygen and the growth and development of
diabetic retinopathy, and age-related macular retinal vessels. In vivo and in vitro studies The XX
Francis I Proctor Lecture. Am J Ophthalmol. 1966;
degeneration. Together, these ongoing and future 62:412–35.
studies will provide new directions in the devel- Bela Anand-Apte JGH. Developmental anatomy of the
opment of novel antiangiogenic or angiomodula- retinal and choroidal vasculature. In: Joseph Besharse
tory therapies to prevent or treat pathologic DB, editor. The retina and its disorders. Oxford:
Academic Press; 2011.
ocular angiogenesis and the breakdown of the Benedito R, et al. The notch ligands Dll4 and Jagged1
blood–retinal barrier that is central to vision loss have opposing effects on angiogenesis. Cell. 2009;
in children and adults. 137:1124–35.
16 J. Chen et al.
Brunet A, et al. Stress-dependent regulation of FOXO Claxton S, Fruttiger M. Role of arteries in oxygen induced
transcription factors by the SIRT1 deacetylase. vaso-obliteration. Exp Eye Res. 2003;77:305–11.
Science. 2004;303:2011–5. Cringle SJ, Yu PK, Su EN, Yu DY. Oxygen distribution
Buttery RG, Hinrichsen CF, Weller WL, Haight JR. How and consumption in the developing rat retina. Invest
thick should a retina be? A comparative study of mam- Ophthalmol Vis Sci. 2006;47:4072–6.
malian species with and without intraretinal vascula- Daneman R, et al. Wnt/beta-catenin signaling is required
ture. Vision Res. 1991;31:169–87. for CNS, but not non-CNS, angiogenesis. Proc Natl
Caprara C, Grimm C. From oxygen to erythropoietin: rel- Acad Sci U S A. 2009;106:641–6.
evance of hypoxia for retinal development, health and De Bock K, Georgiadou M, Carmeliet P. Role of endothe-
disease. Prog Retin Eye Res. 2012;31:89–119. lial cell metabolism in vessel sprouting. Cell Metab.
Caprara C, et al. HIF1A is essential for the development 2013;18:634–47.
of the intermediate plexus of the retinal vasculature. de Gooyer TE, et al. Retinopathy is reduced during exper-
Invest Ophthalmol Vis Sci. 2011;52:2109–17. imental diabetes in a mouse model of outer retinal
Carmeliet P, Jain RK. Molecular mechanisms and clinical degeneration. Invest Ophthalmol Vis Sci. 2006;47:
applications of angiogenesis. Nature. 2011;473: 5561–8.
298–307. De Smet F, Segura I, De Bock K, Hohensinner PJ,
Carmeliet P, Tessier-Lavigne M. Common mechanisms of Carmeliet P. Mechanisms of vessel branching: filopo-
nerve and blood vessel wiring. Nature. 2005;436: dia on endothelial tip cells lead the way. Arterioscler
193–200. Thromb Vasc Biol. 2009;29:639–49.
Chan-Ling T, Gock B, Stone J. The effect of oxygen on Dejda A, et al. Neuropilin-1 mediates myeloid cell che-
vasoformative cell division. Evidence that ‘physiolog- moattraction and influences retinal neuroimmune
ical hypoxia’ is the stimulus for normal retinal vascu- crosstalk. J Clin Invest. 2014.
logenesis. Invest Ophthalmol Vis Sci. 1995;36: Dorrell M, Friedlander M. Mechanisms of endothelial cell
1201–14. guidance and vascular patterning in the developing
Chase J. The evolution of retinal vascularization in mouse retina. Prog Retin Eye Res. 2006;25:277–95.
mammals. A comparison of vascular and avascular Dorrell MI, Aguilar E, Friedlander M. Retinal vascular
retinae. Ophthalmology. 1982;89:1518–25. development is mediated by endothelial filopodia, a
Checchin D, Sennlaub F, Levavasseur E, Leduc M, preexisting astrocytic template and specific R-cadherin
Chemtob S. Potential role of microglia in retinal blood adhesion. Invest Ophthalmol Vis Sci. 2002;43:
vessel formation. Invest Ophthalmol Vis Sci. 3500–10.
2006;47:3595–602. Dorrell MI, Otani A, Aguilar E, Moreno SK, Friedlander
Chen L, Yang P, Kijlstra A. Distribution, markers, and M. Adult bone marrow-derived stem cells use
functions of retinal microglia. Ocul Immunol Inflamm. R-cadherin to target sites of neovascularization in the
2002;10:27–39. developing retina. Blood. 2004;103:3420–7.
Chen J, Connor KM, Aderman CM, Smith LE. Dreher B, Robinson SR. Development of the retinofugal
Erythropoietin deficiency decreases vascular stability pathway in birds and mammals: evidence for a com-
in mice. J Clin Invest. 2008;118:526–33. mon ‘timetable’. Brain Behav Evol. 1988;31:369–90.
Chen J, et al. Suppression of retinal neovascularization by Dreher Z, Robinson SR, Distler C. Muller cells in vascu-
erythropoietin siRNA in a mouse model of prolifera- lar and avascular retinae: a survey of seven mammals.
tive retinopathy. Invest Ophthalmol Vis Sci. J Comp Neurol. 1992;323:59–80.
2009;50:1329–35. Dumont DJ, et al. Dominant-negative and targeted null
Chen J, et al. Wnt signaling mediates pathological vascu- mutations in the endothelial receptor tyrosine kinase,
lar growth in proliferative retinopathy. Circulation. tek, reveal a critical role in vasculogenesis of the
2011;124:1871–81. embryo. Genes Dev. 1994;8:1897–909.
Chen J, et al. Retinal expression of Wnt-pathway medi- Edwards MM, et al. The deletion of Math5 disrupts retinal
ated genes in low-density lipoprotein receptor-related blood vessel and glial development in mice. Exp Eye
protein 5 (Lrp5) knockout mice. PLoS One. 2012;7, Res. 2011.
e30203. Epstein AC, et al. C. elegans EGL-9 and mammalian homo-
Chen J, et al. Neuronal sirtuin1 mediates retinal vascular logs define a family of dioxygenases that regulate HIF
regeneration in oxygen-induced ischemic retinopathy. by prolyl hydroxylation. Cell. 2001;107:43–54.
Angiogenesis. 2013. Fantin A, et al. Tissue macrophages act as cellular chaper-
Choi YK, et al. AKAP12 regulates human blood-retinal ones for vascular anastomosis downstream of VEGF-
barrier formation by downregulation of hypoxia- mediated endothelial tip cell induction. Blood.
inducible factor-1alpha. J Neurosci. 2007;27: 2010;116:829–40.
4472–81. Fruttiger M. Development of the mouse retinal vascula-
Chu Y, Hughes S, Chan-Ling T. Differentiation and ture: angiogenesis versus vasculogenesis. Invest
migration of astrocyte precursor cells and astrocytes in Ophthalmol Vis Sci. 2002;43:522–7.
human fetal retina: relevance to optic nerve coloboma. Fruttiger M. Development of the retinal vasculature.
FASEB J. 2001;15:2013–5. Angiogenesis. 2007;10:77–88.
Fruttiger M, et al. PDGF mediates a neuron-astrocyte Hellstrom M, et al. Dll4 signalling through Notch1 regu-
interaction in the developing retina. Neuron. 1996;17: lates formation of tip cells during angiogenesis.
1117–31. Nature. 2007;445:776–80.
Fukushima Y, et al. Sema3E-PlexinD1 signaling selec- Henkind P, Hansen RI, Szalay J, Editors. Ocular circula-
tively suppresses disoriented angiogenesis in ischemic tion. In: Physiology of the human eye and visual sys-
retinopathy in mice. J Clin Invest. 2011;121:1974–85. tem. New York: Harper & Row; 1979, pp. 98–155.
Gariano RF, Gardner TW. Retinal angiogenesis in devel- Huber AB, Kolodkin AL, Ginty DD, Cloutier JF. Signaling
opment and disease. Nature. 2005;438:960–6. at the growth cone: ligand-receptor complexes and the
Gariano RF, Sage EH, Kaplan HJ, Hendrickson control of axon growth and guidance. Annu Rev
AE. Development of astrocytes and their relation to Neurosci. 2003;26:509–63.
blood vessels in fetal monkey retina. Invest Ophthalmol Hughes S, Yang H, Chan-Ling T. Vascularization of the
Vis Sci. 1996;37:2367–75. human fetal retina: roles of vasculogenesis and angio-
Gelfand MV, Hong S, Gu C. Guidance from above: com- genesis. Invest Ophthalmol Vis Sci. 2000;41:1217–28.
mon cues direct distinct signaling outcomes in vascu- Ishida S, et al. Leukocytes mediate retinal vascular remod-
lar and neural patterning. Trends Cell Biol. eling during development and vaso-obliteration in dis-
2009;19:99–110. ease. Nat Med. 2003;9:781–8.
Gerhardt H, et al. VEGF guides angiogenic sprouting uti- Jaakkola P, et al. Targeting of HIF-alpha to the von Hippel-
lizing endothelial tip cell filopodia. J Cell Biol. Lindau ubiquitylation complex by O2-regulated prolyl
2003;161:1163–77. hydroxylation. Science. 2001;292:468–72.
Gogat K, et al. VEGF and KDR gene expression during Janzer RC, Raff MC. Astrocytes induce blood-brain bar-
human embryonic and fetal eye development. Invest rier properties in endothelial cells. Nature. 1987;
Ophthalmol Vis Sci. 2004;45:7–14. 325:253–7.
Grant MB, et al. Adenosine receptor activation induces Katsura Y, et al. Erythropoietin is highly elevated in vitre-
vascular endothelial growth factor in human retinal ous fluid of patients with proliferative diabetic reti-
endothelial cells. Circ Res. 1999;85:699–706. nopathy. Diabetes Care. 2005;28:2252–4.
Grant MB, et al. Proliferation, migration, and ERK activa- Kim J, Oh WJ, Gaiano N, Yoshida Y, Gu C. Semaphorin
tion in human retinal endothelial cells through A (2B) 3E-Plexin-D1 signaling regulates VEGF function in
adenosine receptor stimulation. Invest Ophthalmol Vis developmental angiogenesis via a feedback mecha-
Sci. 2001;42:2068–73. nism. Genes Dev. 2011;25:1399–411.
Grant MB, et al. Adult hematopoietic stem cells provide Klagsbrun M, Eichmann A. A role for axon guidance
functional hemangioblast activity during retinal neo- receptors and ligands in blood vessel development and
vascularization. Nat Med. 2002;8:607–12. tumor angiogenesis. Cytokine Growth Factor Rev.
Gray H. Gray’s anatomy. Philadelphia: Elsevier; 2008. 2005;16:535–48.
Gutman M, Bonomi F, Pagani S, Cerletti P, Kroneck Klinger G, et al. Outcome of early-onset sepsis in a
P. Modulation of the flavin redox potential as mode of national cohort of very low birth weight infants.
regulation of succinate dehydrogenase activity. Pediatrics. 2010;125:e736–40.
Biochim Biophys Acta. 1980;591:400–8. Kubota Y, et al. M-CSF inhibition selectively targets path-
Hackett SF, Wiegand S, Yancopoulos G, Campochiaro ological angiogenesis and lymphangiogenesis. J Exp
PA. Angiopoietin-2 plays an important role in retinal Med. 2009;206:1089–102.
angiogenesis. J Cell Physiol. 2002;192:182–7. Lahdenranta J, et al. An anti-angiogenic state in mice and
Hartnett ME. Pediatric Retina. Philadelphia, Lippincott humans with retinal photoreceptor cell degeneration.
Williams & Wilkins. 2013. Proc Natl Acad Sci U S A. 2001;98:10368–73.
Hayreh SS. Orbital vascular anatomy. Eye (Lond). Larrivee B, Freitas C, Suchting S, Brunet I, Eichmann
2006;20:1130–44. A. Guidance of vascular development: lessons from
He W, et al. Citric acid cycle intermediates as ligands for the nervous system. Circ Res. 2009;104:428–41.
orphan G-protein-coupled receptors. Nature. Lee J, Dammann O. Perinatal infection, inflammation,
2004;429:188–93. and retinopathy of prematurity. Semin Fetal Neonatal
Hellstrom A, et al. Low IGF-I suppresses VEGF-survival Med. 2012;17:26–9.
signaling in retinal endothelial cells: direct correlation Liebner S, et al. Wnt/beta-catenin signaling controls
with clinical retinopathy of prematurity. Proc Natl development of the blood-brain barrier. J Cell Biol.
Acad Sci U S A. 2001;98:5804–8. 2008;183:409–17.
Hellstrom A, et al. IGF-I is critical for normal vasculariza- Linsenmeier RA, Padnick-Silver L. Metabolic depen-
tion of the human retina. J Clin Endocrinol Metab. dence of photoreceptors on the choroid in the normal
2002;87:3413–6. and detached retina. Invest Ophthalmol Vis Sci.
Hellstrom A, et al. Postnatal serum insulin-like growth 2000;41:3117–23.
factor I deficiency is associated with retinopathy of Lobov IB, et al. WNT7b mediates macrophage-induced
prematurity and other complications of premature programmed cell death in patterning of the vascula-
birth. Pediatrics. 2003;112:1016–20. ture. Nature. 2005;437:417–21.
18 J. Chen et al.
Lofqvist C, et al. Postnatal head growth deficit among pre- Pierce EA, Foley ED, Smith LE. Regulation of vascular
mature infants parallels retinopathy of prematurity and endothelial growth factor by oxygen in a model of reti-
insulin-like growth factor-1 deficit. Pediatrics. nopathy of prematurity. Arch Ophthalmol. 1996;114:
2006;117:1930–8. 1219–28.
McLeod DS, Hasegawa T, Prow T, Merges C, Lutty Provis JM, et al. Development of the human retinal vascu-
G. The initial fetal human retinal vasculature develops lature: cellular relations and VEGF expression. Exp
by vasculogenesis. Dev Dyn. 2006;235:3336–47. Eye Res. 1997;65:555–68.
Meixner-Monori B, Kubicek CP, Habison A, Kubicek- Rehm HL, et al. Vascular defects and sensorineural deaf-
Pranz EM, Rohr M. Presence and regulation of the ness in a mouse model of Norrie disease. J Neurosci.
alpha-ketoglutarate dehydrogenase multienzyme com- 2002;22:4286–92.
plex in the filamentous fungus Aspergillus niger. Ross MH, Pawlina W. Histology: a text and atlas with
J Bacteriol. 1985;161:265–71. correlated cell and molecular biology. Philadelphia:
Michaelson IC. The mode of development of the vascular Lippincott Williams & Wilkins; 2005.
system of the retina, with some observations on its sig- Runkle EA, Antonetti DA. The blood-retinal barrier:
nificance for certain retinal diseases. Trans structure and functional significance. Methods Mol
Ophthalomol Soc UK. 1948;68:137–81. Biol. 2011;686:133–48.
Michan S. Acetylome regulation by sirtuins in the brain: Rymo SF, et al. A two-way communication between
from normal physiology to aging and pathology. Curr microglial cells and angiogenic sprouts regulates
Pharm Des. 2013. angiogenesis in aortic ring cultures. PLoS One.
Michan S, et al. SIRT1 is essential for normal cognitive 2011;6, e15846.
function and synaptic plasticity. J Neurosci. 2010; Saint-Geniez M, D’Amore PA. Development and pathol-
30:9695–707. ogy of the hyaloid, choroidal and retinal vasculature.
Mino RP, et al. Adenosine receptor antagonists and retinal Int J Dev Biol. 2004;48:1045–58.
neovascularization in vivo. Invest Ophthalmol Vis Sci. Santos AM, et al. Embryonic and postnatal development
2001;42:3320–4. of microglial cells in the mouse retina. J Comp Neurol.
Mitchell CA, Risau W, Drexler HC. Regression of vessels 2008;506:224–39.
in the tunica vasculosa lentis is initiated by coordi- Sapieha P. Eyeing central neurons in vascular growth and
nated endothelial apoptosis: a role for vascular endo- reparative angiogenesis. Blood. 2012;120:2182–94.
thelial growth factor as a survival factor for Sapieha P, et al. The succinate receptor GPR91 in neurons
endothelium. Dev Dyn. 1998;213:322–33. has a major role in retinal angiogenesis. Nat Med.
Mu X, et al. Ganglion cells are required for normal 2008;14:1067–76.
progenitor-cell proliferation but not cell-fate determi- Sato TN, et al. Distinct roles of the receptor tyrosine
nation or patterning in the developing mouse retina. kinases Tie-1 and Tie-2 in blood vessel formation.
Curr Biol. 2005;15:525–30. Nature. 1995;376:70–4.
Nakamura-Ishizu A, et al. The formation of an angiogenic Sato T, Kusaka S, Shimojo H, Fujikado T. Vitreous levels
astrocyte template is regulated by the neuroretina in a of erythropoietin and vascular endothelial growth fac-
HIF-1-dependent manner. Dev Biol. 2012;363:106–14. tor in eyes with retinopathy of prematurity.
Netter FH. Atlas of human anatomy. Philadelphia: Ophthalmology. 2009;116:1599–603.
Elsevier Health Sciences; 2006. Scott A, et al. Astrocyte-derived vascular endothelial
Niehrs C. Norrin and frizzled; a new vein for the eye. Dev growth factor stabilizes vessels in the developing reti-
Cell. 2004;6:453–4. nal vasculature. PLoS One. 2010;5, e11863.
Ohh M, et al. Ubiquitination of hypoxia-inducible factor Senger DR, et al. Tumor cells secrete a vascular permea-
requires direct binding to the beta-domain of the von bility factor that promotes accumulation of ascites
Hippel-Lindau protein. Nat Cell Biol. 2000;2:423–7. fluid. Science. 1983;219:983–5.
Ohlmann A, et al. Norrin promotes vascular regrowth Sengupta N, et al. The role of adult bone marrow-derived
after oxygen-induced retinal vessel loss and sup- stem cells in choroidal neovascularization. Invest
presses retinopathy in mice. J Neurosci. 2010;30: Ophthalmol Vis Sci. 2003;44:4908–13.
183–93. Shui YB, et al. Vascular endothelial growth factor expres-
Oshima Y, et al. Angiopoietin-2 enhances retinal vessel sion and signaling in the lens. Invest Ophthalmol Vis
sensitivity to vascular endothelial growth factor. J Cell Sci. 2003;44:3911–9.
Physiol. 2004;199:412–7. Smith LE. IGF-1 and retinopathy of prematurity in the
Paul Riordan-Eva ETC. Vaughan & Asbury’s general preterm infant. Biol Neonate. 2005;88:237–44.
ophthalmology. New York: McGraw-Hill Smith LE, et al. Essential role of growth hormone in
Professional. 2011. ischemia-induced retinal neovascularization. Science.
Pierce EA, Avery RL, Foley ED, Aiello LP, Smith 1997;276:1706–9.
LE. Vascular endothelial growth factor/vascular per- Smith LE, et al. Regulation of vascular endothelial growth
meability factor expression in a mouse model of reti- factor-dependent retinal neovascularization by insulin-
nal neovascularization. Proc Natl Acad Sci U S A. like growth factor-1 receptor. Nat Med. 1999;5:
1995;92:905–9. 1390–5.
Stahl A, et al. The mouse retina as an angiogenesis model. Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-
Invest Ophthalmol Vis Sci. 2010a;51:2813–26. inducible factor 1 is a basic-helix-loop-helix-PAS het-
Stahl A, et al. Vitreal levels of erythropoietin are increased erodimer regulated by cellular O2 tension. Proc Natl
in patients with retinal vein occlusion and correlate Acad Sci U S A. 1995;92:5510–4.
with vitreal VEGF and the extent of macular edema. Wang Y, et al. Norrin/Frizzled4 signaling in retinal vascu-
Retina. 2010b. lar development and blood brain barrier plasticity.
Stefater Iii JA, et al. Regulation of angiogenesis by a non- Cell. 2012;151:1332–44.
canonical Wnt-Flt1 pathway in myeloid cells. Nature. Wangsa-Wirawan ND, Linsenmeier RA. Retinal oxygen:
2011;474:511–5. fundamental and clinical aspects. Arch Ophthalmol.
Sternberg Jr P, Landers 3rd MB, Wolbarsht M. The nega- 2003;121:547–57.
tive coincidence of retinitis pigmentosa and prolifera- Watanabe T, Raff MC. Retinal astrocytes are immigrants
tive diabetic retinopathy. Am J Ophthalmol. 1984; from the optic nerve. Nature. 1988;332:834–7.
97:788–9. Watanabe D, et al. Erythropoietin as a retinal angiogenic
Stone J, et al. Development of retinal vasculature is medi- factor in proliferative diabetic retinopathy. N Engl
ated by hypoxia-induced vascular endothelial growth J Med. 2005;353:782–92.
factor (VEGF) expression by neuroglia. J Neurosci. Weidemann A, et al. Astrocyte hypoxic response is essen-
1995;15:4738–47. tial for pathological but not developmental angiogen-
Tolsma KW, et al. Neonatal bacteremia and retinopathy of esis of the retina. Glia. 2010;58:1177–85.
prematurity: the ELGAN study. Arch Ophthalmol. Wilson BD, et al. Netrins promote developmental and
2011;129:1555–63. therapeutic angiogenesis. Science. 2006;313:640–4.
Torczynski E. Choroid and suprachoroid. Ocular anatomy, Wise GN. Factors influencing retinal new vessel forma-
embryology and teratology. Philadelphia: Harper & tion. Am J Ophthalmol. 1961;52:637–50.
Row; 1982. Xu Q, et al. Vascular development in the retina and inner
Tremblay S, et al. Systemic inflammation perturbs devel- ear: control by Norrin and Frizzled-4, a high-affinity
opmental retinal angiogenesis and neuroretinal function. ligand-receptor pair. Cell. 2004;116:883–95.
Invest Ophthalmol Vis Sci. 2013;54:8125–39. Ye X, Smallwood P, Nathans J. Expression of the Norrie
Uemura A, Kusuhara S, Wiegand SJ, Yu RT, Nishikawa disease gene (Ndp) in developing and adult mouse eye,
S. Tlx acts as a proangiogenic switch by regulating ear, and brain. Gene Expr Patterns. 2011;11:151–5.
extracellular assembly of fibronectin matrices in retinal Yi X, Mai LC, Uyama M, Yew DT. Time-course expres-
astrocytes. J Clin Invest. 2006;116:369–77. sion of vascular endothelial growth factor as related to
Wang GL, Semenza GL. Characterization of hypoxia- the development of the retinochoroidal vasculature in
inducible factor 1 and regulation of DNA binding activ- rats. Exp Brain Res. 1998;118:155–60.
ity by hypoxia. J Biol Chem. 1993a;268:21513–8. Zhao S, Overbeek PA. Regulation of choroid development by
Wang GL, Semenza GL. General involvement of hypoxia- the retinal pigment epithelium. Mol Vis. 2001;7:277–82.
inducible factor 1 in transcriptional response to hypoxia. Zhu M, et al. The human hyaloid system: cell death and
Proc Natl Acad Sci U S A. 1993b;90:4304–8. vascular regression. Exp Eye Res. 2000;70:767–76.
Retinopathy of Prematurity
2
Andreas Stahl, Ann Hellström, and Lois E.H. Smith
25 weeks gestational age) or the ones that are

2.1 Natural History and Grading born small for their respective gestational age or
of ROP show a reduced postnatal growth rate often com-
plicated by accompanying problems like sepsis,
In order to understand when and how treatment anemia, or lung disorders requiring prolonged
strategies for ROP can be delivered in a sensible oxygen supplementation.
and timely manner, a thorough understanding of The reason that very prematurely born infants
ROP disease characteristics and development is in particular may progress to severe ROP stages
essential. ROP develops predominantly in infants lies in the nature of physiologic retinal vascular-
born before 32 weeks of gestational age. Among ization as it is described in detail in the previous
these, most infants will experience only mild chapter by Chen, Liu, and Sapieha. This physio-
forms of ROP that will regress and do not require logic growth of retinal vessels originates from the
treatment. Our challenge as ophthalmologists is optic nerve head and proceeds to the ora serrata
to identify the few infants that will progress to in a centrifugal pattern. If it is disrupted by pre-
ROP stages that require treatment. Those most at mature birth, the peripheral retina remains void
risk are infants that are either born very prema- of blood vessels in the early postnatal period. The
turely (in developed countries mainly below earlier the child is born, the greater the area of
avascular retina at birth. Due to the fact that oxy-
gen saturations in the retinal vasculature increase
A. Stahl (*) significantly after switching from placental sup-
Eye Center at the Medical Center, University of ply to breathing room air, the retinal vessels will
Freiburg, Freiburg 79106, Germany sense a hyperoxic environment during the first
e-mail: andreas.stahl@uniklinik-freiburg.de
weeks after premature birth, even without addi-
A. Hellström tional iatrogenic oxygen supply. This hyperoxic
Department of Ophthalmology, Institute of
Neuroscience and Physiology, The Queen Silvia
retinal environment leads to slowing or even
Children’s Hospital, The Sahlgrenska Academy at stalling of physiologic vessel growth. This hyper-
University of Gothenburg, Gothenburg 416-85, oxic phase of slowed or stalled vessel growth in
Sweden the first weeks of life is often referred to as Phase
e-mail: ann.hellstrom@medfak.gu.se
I of ROP.
L.E.H. Smith In Phase I ROP, the ophthalmologist finds a
Department of Ophthalmology, Harvard Medical
School, Boston Children’s Hospital, 300 Longwood
vascularized central retina and avascular periph-
Avenue, Boston, MA 02465, USA eral retina without a clear demarcation between
e-mail: lois.smith@childrens.harvard.edu the two and without signs of pathologic vessel

22 A. Stahl et al.
activation. The retinal vessel growth is stalled but an increasing area of the retina becoming vascu-
the nonvascularized peripheral retina appears to larized from visit to visit. However, in some
receive sufficient nutrients and oxygen from the infants the hypoxic environment in the peripheral
increased oxygen tension after birth. This over- retina is so severe that pathologically high levels
supply of oxygen that is due to switching from of the angiogenic growth factor VEGF are pro-
placental to pulmonary supply after premature duced (Sonmez et al. 2008; Velez-Montoya et al.
birth is further increased if supplemental oxygen 2010). This leads to a breakdown of natural
is given (which is often necessary to ensure the growth factor gradients in the retina and to aber-
development of other organs or even to guarantee rant and uncontrolled growth of retinal vessels.
survival of the infant (Stenson 2013; Chen et al. Several stages of ROP are distinguished dur-
2010)). After usually 6 or more weeks of life, the ing the course of disease development. In stage 1
peripheral avascular retina, which has by that of ROP, there is only a non-prominent demarca-
time grown in thickness and has matured to some tion line between vascularized and nonvascular-
degree, begins to experience an undersupply of ized retina. In stage 2, the line becomes a
oxygen for its now increased metabolic demand. prominent demarcation ridge. Stage 3 is charac-
As a consequence, the retinal microenvironment terized by extraretinal proliferations on the ridge
switches from hyperoxic to hypoxic. This that can be identified on fundoscopy. This is a
hypoxic phase of ROP is often referred to as stage that will develop only in a minority of
Phase II and is characterized by a reactivation of infants (approximately 5 %), but if it is present,
vessel growth. most cases will require treatment, in particular
Phase II ROP needs to be monitored closely when it is associated with increased tortuosity
by strict adherence to the established screening and dilation of central retinal vessels (referred to
protocols in order to treat if neovascularization as Plus disease). An example of ROP stage 3+ is
progresses to the point of high risk of retinal given in Fig. 2.1.
detachment. During Phase II ROP, the vessels In addition to identifying the correct ROP
start to resume their centrifugal growth and pro- stages, it is essential for the ophthalmologist to
ceed towards the ora serrata. In most infants, this document the zone where the demarcation line or
centrifugal growth progresses without significant ridge is seen between vascularized and nonvascu-
disturbance and the ophthalmologist can identify larized retina. The more central this border lies,
Fig. 2.1 Exemplary images of bilateral ROP stage 3 with The prominent ridge between vascularized and nonvascu-
plus disease (ROP 3+). The optic nerve head is located on larized retina with pathologic vessel proliferations
the central margin of each image. Vessel dilation and tor- defines stage 3 of ROP
tuosity are the defining characteristics of plus disease.
2 Retinopathy of Prematurity 23
subgroup of aggressive posterior ROP (AP-ROP)

which is characterized by pronounced angiogenic
activation and often very fast progression of
pathologic vascular changes that do not in all
cases follow the stage-to-stage progression
described above but may show variable progres-
sion patterns that are challenging to reign in at
early stages. AP-ROP is mainly seen in very sick
and extremely preterm infants.
2.2 Treatment of ROP
Published in 1990, the CRYO-ROP study defined

the underlying principle for ROP treatment for
the next decades (Multicenter trial of cryotherapy
Fig. 2.2 Schematic representation of the three zones of
for retinopathy of prematurity). Three-month
ROP. Zone I is defined as a circle around the optic nerve outcome. Cryotherapy for Retinopathy of
head (ONH) with a radius of twice the distance ONH- Prematurity Cooperative Group 1990b;
fovea. Central zone II is defined as a circle around the Multicenter trial of cryotherapy for retinopathy
ONH with a radius of three times the distance ONH-fovea
(dashed line in zone II). Peripheral zone II is a circle
of prematurity. One-year outcome—structure
around the ONH reaching the ora serrata nasally. Zone III and function. Cryotherapy for Retinopathy of
is the remaining crescent-shaped temporal retina Prematurity Cooperative Group 1990a). The
main finding was that when treated at a point
when about 50 % would regress and 50 % would
the larger the avascular retinal area and the more go on to retinal detachment (called “threshold”),
problematic the potential course of the disease. those infants treated with cryoablation of avascular
For the purpose of using standardized descriptions peripheral retina had fewer unfavorable struc-
or ROP, the retina is divided into different zones. tural outcomes compared to non-treated infants.
Zone I is defined by a circle around the optic Building on this same principle, laser photoco-
nerve head (ONH) with twice the diameter agulation was introduced in the 1990s providing
ONH—fovea. Central zone II (or posterior zone a treatment alternative to cryotherapy (McNamara
II) is defined by a circle around the ONH with et al. 1991; Landers et al. 1992; Fleming et al.
three times the diameter ONH—fovea. Peripheral 1992). Today, laser photocoagulation of periph-
zone II (or anterior zone II) is a circle around the eral avascular retina has replaced cryotherapy in
ONH reaching the ora serrata nasally. Zone III most countries in the treatment of ROP (Simpson
refers to the crescent-shaped remaining retina on et al. 2012). Both cryo and laser therapy, how-
the temporal side. The ROP zones are schemati- ever, follow the same principle: avascular retina
cally depicted in Fig. 2.2. is destroyed in order to decrease the production
Stage 4 and 5 of ROP are characterized by of pro-angiogenic growth factors like VEGF
partial or complete retinal detachment and can be from hypoxic retinal cells. The treatment is safe
avoided in most cases by timely treatment at ear- and, when performed properly, will in most cases
lier stages. Stage 4 and 5 of ROP are often associ- stop disease progression before stage 4 or 5 of
ated with poor functional long-term outcomes. ROP develops. The downside of laser therapy,
However, in some cases ROP can progress so however, is that potentially viable retinal tissue is
aggressively that even with best-possible screen- replaced by functionless scar tissue and that laser
ing and treatment protocols in place stage 4 or 5 therapy may increase the extent of myopia
will develop. This is particularly the case in the (Geloneck et al. 2014).
24 A. Stahl et al.
In the recent years, two mutually synergistic and inhibition of VEGF targets Phase II of ROP
medical approaches have been evaluated that when overexpression of angiogenic growth fac-
have the potential to significantly broaden our tors and activation of pathologic vessel growth
range of therapeutic options in ROP. The first of has already occurred. It is therefore potentially
these medical approaches is supplementation of complementary to IGF-1 supplementation and
the fetal growth factor insulin-like growth factor could be administered to infants proceeding to
1 (IGF-1) during the first weeks of life. This Phase II ROP despite a best-possible preventive
treatment targets Phase I of ROP and has the management in Phase I. Once pathologic vessel
potential of providing a preventive intervention growth has occurred in the retina, IGF-1 supple-
by aiming at normalizing retinal vascular devel- mentation is no longer helpful and potentially
opment during the first weeks of life. IGF-1 is could even aggravate disease progression by fur-
essential for the physiologic vascular growth in ther stimulating vessel growth. During Phase II,
the retina (Hellstrom et al. 2002; Smith et al. an antiangiogenic treatment approach is appeal-
1999; Hellstrom et al. 2001). In addition, IGF-1 ing if found to be safe. And with VEGF being the
is also an important inducer of overall growth and main angiogenic molecule in the vitreous of ROP
development of all tissues (Netchine et al. 2011). infants (Sonmez et al. 2008), it appears sensible
The role of IGF-1 in ROP may therefore be two- to target VEGF for this purpose. The first effect
fold: First, IGF-1 may directly promote normal of anti-VEGF therapy is a reduction of plus dis-
retinal development during Phase I of ROP when ease which can often be observed in the first days
vascular growth in the retina is stalled. Augmented after treatment (Fig. 2.3).
retinal vascular growth during Phase I in turn After several smaller studies were published,
would lead to a reduced size of peripheral avas- the BEAT-ROP study was the first larger trial
cular retina in later weeks of life and would comparing the effect of standard laser treatment
therefore limit overexpression of angiogenic in ROP to the effect of an intravitreal injection of
growth factors, thereby preventing or curtailing the anti-VEGF antibody bevacizumab (Mintz-
Phase II. The second effect of IGF-1 on ROP Hittner et al. 2011). The primary end point of the
might be more indirect but nevertheless similarly BEAT-ROP study was recurrence of active ROP
important. It is known that serum IGF-1 is sub- up to 54 weeks postmenstrual age (PMA). The
stantially reduced after preterm birth (Engstrom study found that for zone I disease there were
et al. 2005) and that low IGF-1 levels are associ- fewer recurrences in the bevacizumab group
ated with poor postnatal weight gain and the compared to laser and for central zone II disease
development of more severe forms of ROP both treatments were equal with regard to recur-
(Hellstrom et al. 2003; Perez-Munuzuri et al. rence rates. One caveat when interpreting these
2010). Increasing systemic IGF-1 to levels that results is that the recurrence rate in the zone I
would be normal during that developmental win- laser group of the BEAT-ROP study was rela-
dow could therefore have a beneficial effect on tively high compared to other published studies.
overall postnatal growth and improve the general Another limitation is the relatively short follow-
health and development of a preterm infant— up window that did not examine if avascular ret-
with associated beneficial effects on his or her ina persisted long term. Different from laser
ROP risk. A phase I study administering IGF-I to therapy, recurrences of proliferative disease after
preterm infants has been performed (Hellstrom bevacizumab therapy may occur later due to the
et al. 2003), and a multicenter, international persistence of peripheral avascular retinal tissue.
phase II clinical trial is currently ongoing to These late recurrences could fall outside the
investigate whether IGF-1 supplementation dur- observational window of 54 weeks PMA investi-
ing Phase I of ROP is safe and well tolerated gated in the published BEAT-ROP results (Hu
(clinical trials identifier NCT01096784). et al. 2012; Lee et al. 2012). Systemic effects of
The second medical approach to treat anti-VEGF treatment were not evaluated in the
advanced ROP is anti-VEGF treatment. Binding BEAT-ROP study.
Fig. 2.3 ROP stage 3+ before and three days after treatment with intravitreal anti-VEGF therapy. Note the rapid reduction
in plus disease visible by reduced dilatation and reduced tortuosity of the retinal vessels
Despite these caveats, the BEAT-ROP results zumab is the ideal drug in this vulnerable popula-
raised some important points: tion (Avery 2012). Bevacizumab has a systemic
(1) Bevacizumab is effective in halting neo- half-life of several days. So even if lower doses of
vascular disease progression in ROP in most bevacizumab were used intravitreally, bevaci-
cases; (2) unlike laser therapy, the avascular reti- zumab would be secreted from the vitreous into
nal areas are not destroyed but may become (par- the circulation (possibly via active transport
tially) vascularized over time. Whether the through the RPE (Powner et al. 2014)) and would
vessels and underlying neural retina are fully nor- therefore accumulate over time in the systemic cir-
mal is unknown at this point. Recent data, how- culation (Krohne et al. 2014). Whether and how
ever, point towards comparable visual acuity this systemic exposure to an anti-VEGF antibody
results in infants treated with bevacizumab com- could affect developmental processes in, for exam-
pared to laser-treated infants—with the potential ple, brain or lung maturation is currently not
advantage of fewer cases of high myopia in known. Ranibizumab, however, is an alternative
infants with bevacizumab treatment over laser anti-VEGF agent with comparable efficacy at
(Harder et al. 2013; Geloneck et al. 2014; treating retinal vascular proliferations in adults
Martinez-Castellanos et al. 2013). (Martin et al. 2012; Chakravarthy et al. 2013;
Some important questions remain with regard Chakravarthy et al. 2012; Martin et al. 2011) and
to anti-VEGF treatment in ROP. In particular, only with the possible advantage of much faster sys-
limited data exist on potential systemic effects of temic clearance rates (Krohne et al. 2014).
intravitreal anti-VEGF treatment. It is known from Ranibizumab is therefore currently being investi-
adult data that after one single bevacizumab injec- gated in the CARE-ROP study, a multicenter ran-
tion systemic VEGF levels in the circulation domized controlled trial for ROP (clinical trials
remain suppressed below the detection levels of identifier NCT02134457).
the most prevalently used ELISA assays over sev- The CARE-ROP study investigates two perti-
eral weeks (Zehetner et al. 2013; Matsuyama et al. nent questions:
2010). The same appears to be true in infants (Sato (1) Is ranibizumab with its advantageous sys-
et al. 2012) where the currently used dose of beva- temic clearance rate effective in treating Phase II
cizumab (0.625 mg per eye) by far exceeds the ROP; and (2) are lower doses than the currently
standard adult exposure when calculated per body used 50 % adult dose sufficient to treat ROP with
weight or per body surface area. It is currently not the potential benefit of reduced local and
known what the ideal dosing for bevacizumab in systemic drug exposure in a developing infant.
ROP infants is—nor is it clear whether bevaci- In addition to its faster systemic clearance rate,
26 A. Stahl et al.
ranibizumab has been developed explicitly for 2.3 Summary

use in the eye and is approved for intravitreal use
in adults for several indications (Rosenfeld et al. ROP therapy is currently at a crucial crossroads:
2006; Massin et al. 2010; Brown et al. 2010). The If the currently ongoing clinical trials on IGF-1
two doses compared in the CARE-ROP study are supplementation and anti-VEGF treatment pro-
24 % vs. 40 % of the standard adult dose. Both vide sufficient evidence for their safety and effi-
these doses are lower than the 50 % adult dose of cacy, then the current laser therapy for ROP may
bevacizumab used in the BEAT-ROP study. be supplemented by two potentially complemen-
The use of lower doses may have an additional tary medical approaches:
beneficial effect. The physiologic vascularization (1) In Phase I, promotion of physiologic retinal
of the peripheral avascular retina may progress vascularization by normalization of IGF-1; and
better than occurs in infants treated with higher (2) in Phase II inhibition of VEGF with lowest
doses. Animal studies found that lower doses of possible anti-VEGF dosing to inhibit pathologic
anti-VEGF agents were comparable to high doses angiogenesis but to allow further peripheral
in controlling pathologic angiogenic activity but vascularization.
superior in allowing physiologic peripheral Ideally, ROP management would consist of
vascularization at the same time (Lutty et al. 2011). two steps. In a first step aimed at preventive
It is important to emphasize the significance treatment, all at-risk infants would receive sup-
of sufficient peripheral retinal vascularization plemental systemic IGF-1 from birth to normal-
after anti-VEGF treatment. Different from laser ize their physiologic retinal vascular
treatment, the peripheral retina is not rendered a development. In those infants who despite this
scar tissue by anti-VEGF treatment. If neural preventive treatment proceed to Phase II ROP
retinal development is normal, then these retinal with aberrant vessel proliferation in relatively
areas retain their potential to contribute to visual central retinal areas (where laser treatment
function. Further, no scar tissue is induced that would induce large areas of scarring), anti-
might produce traction on the more central retina VEGF treatment could be applied—in the lowest
and macula. possible dose and using compounds that are
The downside of not using laser is that the cleared fast from the systemic circulation. After
peripheral retina may not function properly but will anti-VEGF treatment, all infants would then be
continue to produce VEGF. By not destroying this followed-up regularly to monitor first the regres-
dysfunctional tissue, there may be a late recurrence sion of pathologic vessel growth and then (over a
of ROP—up to months after an initially successful longer period of time) the centrifugal progres-
treatment (Hu et al. 2012; Ittiara et al. 2013). In sion of physiologic vessel growth in the periph-
infants where no or incomplete vascularization can ery. If infants develop recurrent ROP, then laser
be obtained, a prolonged period of frequent follow- therapy may be warranted to permanently stop
up examinations after anti-VEGF treatment is war- the production of angiogenic factors from
ranted. In some cases, a sequential approach of first remaining peripheral avascular areas—ideally at
anti-VEGF therapy for central ROP (in zone I or a time when some peripheral vascularization has
central zone II) followed by laser therapy at a later already occurred due to the prior IGF-1 and anti-
time point for remaining peripheral avascular retina VEGF treatments. This sequential approach
with signs of recurring ROP activity may be sensi- could in future ROP treatment protocols poten-
ble. With lower doses of anti-VEGF agents as they tially reduce the number of infants progressing
are investigated in the CARE-ROP study, there to ROP stages with pathologic vessel prolifera-
may be superior peripheral vascularization after tion (by IGF-1 supplementation) and for those
anti-VEGF treatment which would reduce the risk infants that do progress, reduce the laser-induced
for late recurrences while at the same time control- scar areas to a minimum (by low-dose anti-
ling excessive aberrant vessel growth. VEGF treatment).
2010;125(6):e1483–1492. doi:10.1542/peds.2009-
Compliance with Ethical Requirements 2218. peds.2009-2218 [pii].
Cryotherapy for Retinopathy of Prematurity Cooperative
Informed Consent and Animal Studies Group. Multicenter trial of cryotherapy for retinopa-
disclosures are not applicable to this review. thy of prematurity. Arch Ophthalmol. 1990a;
The authors received research grants from 108(10):1408–16.
Deutsche Forschungsgemeinschaft (DFG Cryotherapy for Retinopathy of Prematurity Cooperative
Group. Multicenter trial of cryotherapy for retinopathy
STA 1102/5-1), Deutsche Ophthalmologische of prematurity. Three-month outcome. Arch
Gesellschaft (DOG), Novartis Pharma (AS), Ophthalmol. 1990b;108(2):195–204.
European Commission FP7 project 305485 Engstrom E, Niklasson A, Wikland KA, Ewald U,
PREVENT-ROP (AH), NIH/NEI (EY022275, Hellstrom A. The role of maternal factors, postnatal
nutrition, weight gain, and gender in regulation of
EY017017, P01 HD18655, RPB Sr. serum IGF-I among preterm infants. Pediatr Res.
Investigator Award, Lowy Medical 2005;57(4):605–10. doi:10.1203/01.PDR.0000
Foundation, (LEHS). AS is consultant for 155950.67503.BC. 01.PDR.0000155950.67503.BC
Novartis Pharma and has received speaker [pii].
Fleming TN, Runge PE, Charles ST. Diode laser photo-
honoraria from Zeiss, Böhringer-Ingelheim, coagulation for prethreshold, posterior retinopathy of
Bausch&Lomb. LEHS and AH are prematurity. Am J Ophthalmol. 1992;114(5):
consultants for Shire Pharmaceuticals. 589–92.
Geloneck MM, Chuang AZ, Clark WL, Hunt MG,
Norman AA, Packwood EA, Tawansy KA, Mintz-
Hittner HA. Refractive outcomes following bevaci-
zumab monotherapy compared with conventional
laser treatment: a randomized clinical trial. JAMA
References Ophthalmol. 2014;132(11):1327–33. doi:10.1001/
jamaophthalmol.2014.2772. 1893536 [pii].
Avery RL. Bevacizumab (Avastin) for retinopathy of pre- Harder BC, Schlichtenbrede FC, von Baltz S, Jendritza W,
maturity: wrong dose, wrong drug, or both? J AAPOS. Jendritza B, Jonas JB. Intravitreal bevacizumab for
2012;16(1):2–4. doi:10.1016/j.jaapos.2011.11.002. retinopathy of prematurity: refractive error results. Am
S1091-8531(11)00612-4 [pii]. J Ophthalmol. 2013;155(6):1119–24. doi:10.1016/j.
Brown DM, Campochiaro PA, Singh RP, Li Z, Gray S, ajo.2013.01.014. e1111S0002-9394(13)00070-6 [pii].
Saroj N, Rundle AC, Rubio RG, Murahashi Hellstrom A, Perruzzi C, Ju M, Engstrom E, Hard AL, Liu
WY. Ranibizumab for macular edema following JL, Albertsson-Wikland K, Carlsson B, Niklasson A,
central retinal vein occlusion: six-month primary Sjodell L, LeRoith D, Senger DR, Smith LE. Low
end point results of a phase III study. Ophthalmology. IGF-I suppresses VEGF-survival signaling in retinal
2010;117(6):1124–33. doi:10.1016/j.oph- endothelial cells: direct correlation with clinical reti-
tha.2010.02.022. e1121S0161-6420(10)00186-7 nopathy of prematurity. Proc Natl Acad Sci U S A.
[pii]. 2001;98(10):5804–8. doi:10.1073/pnas.101113998.
Chakravarthy U, Harding SP, Rogers CA, Downes SM, 101113998 [pii].
Lotery AJ, Wordsworth S, Reeves BC. Ranibizumab Hellstrom A, Carlsson B, Niklasson A, Segnestam K,
versus bevacizumab to treat neovascular age-related Boguszewski M, de Lacerda L, Savage M, Svensson
macular degeneration: one-year findings from the E, Smith L, Weinberger D, Albertsson Wikland K,
IVAN randomized trial. Ophthalmology. Laron Z. IGF-I is critical for normal vascularization of
2012;119(7):1399–411. doi:10.1016/j.oph- the human retina. J Clin Endocrinol Metab.
tha.2012.04.015. S0161-6420(12)00358-2 [pii]. 2002;87(7):3413–6. doi:10.1210/jcem.87.7.8629.
Chakravarthy U, Harding SP, Rogers CA, Downes SM, Hellstrom A, Engstrom E, Hard AL, Albertsson-Wikland
Lotery AJ, Culliford LA, Reeves BC. Alternative treat- K, Carlsson B, Niklasson A, Lofqvist C, Svensson E,
ments to inhibit VEGF in age-related choroidal neovas- Holm S, Ewald U, Holmstrom G, Smith LE. Postnatal
cularisation: 2-year findings of the IVAN randomised serum insulin-like growth factor I deficiency is associ-
controlled trial. Lancet. 2013;382(9900):1258–67. ated with retinopathy of prematurity and other compli-
doi:10.1016/S0140-6736(13)61501-9. S0140-6736(13) cations of premature birth. Pediatrics.
61501-9 [pii]. 2003;112(5):1016–20.
Chen ML, Guo L, Smith LE, Dammann CE, Dammann Hu J, Blair MP, Shapiro MJ, Lichtenstein SJ, Galasso JM,
O. High or low oxygen saturation and severe retinopa- Kapur R. Reactivation of retinopathy of prematurity
thy of prematurity: a meta-analysis. Pediatrics. after bevacizumab injection. Arch Ophthalmol.
28 A. Stahl et al.
2012;130(8):1000–6. doi:10.1001/archophthal- McNamara JA, Tasman W, Brown GC, Federman

mol.2012.592. archophthalmol.2012.592 [pii]. JL. Laser photocoagulation for stage 3+ retinopathy of
Ittiara S, Blair MP, Shapiro MJ, Lichtenstein SJ. Exudative prematurity. Ophthalmology. 1991;98(5):576–80.
retinopathy and detachment: a late reactivation of reti- Mintz-Hittner HA, Kennedy KA, Chuang AZ. Efficacy of
nopathy of prematurity after intravitreal bevacizumab. intravitreal bevacizumab for stage 3+ retinopathy of
J AAPOS. 2013;17(3):323–5. doi:10.1016/j.jaa- prematurity. N Engl J Med. 2011;364(7):603–15.
pos.2013.01.004. S1091-8531(13)00083-9 [pii]. doi:10.1056/NEJMoa1007374.
Krohne TU, Holz FG, Meyer CH. Pharmacokinetics of Netchine I, Azzi S, Le Bouc Y, Savage MO. IGF1 molecu-
intravitreally administered VEGF inhibitors. lar anomalies demonstrate its critical role in fetal,
Ophthalmologe. 2014;111(2):113–20. doi:10.1007/ postnatal growth and brain development. Best Pract
s00347-013-2932-9. Res Clin Endocrinol Metab. 2011;25(1):181–90.
Landers 3rd MB, Toth CA, Semple HC, Morse LS. doi:10.1016/j.beem.2010.08.005. S1521-
Treatment of retinopathy of prematurity with argon 690X(10)00090-4 [pii].
laser photocoagulation. Arch Ophthalmol. Perez-Munuzuri A, Fernandez-Lorenzo JR, Couce-Pico
1992;110(1):44–7. ML, Blanco-Teijeiro MJ, Fraga-Bermudez JM. Serum
Lee BJ, Kim JH, Heo H, Yu YS. Delayed onset atypical levels of IGF1 are a useful predictor of retinopathy of
vitreoretinal traction band formation after an intravit- prematurity. Acta Paediatr. 2010;99(4):519–25.
real injection of bevacizumab in stage 3 retinopathy of doi:10.1111/j.1651-2227.2009.01677.x. APA1677
prematurity. Eye (Lond). 2012;26(7):903–9. [pii].
doi:10.1038/eye.2012.111. quiz 910eye2012111 [pii]. Powner MB, McKenzie JA, Christianson GJ, Roopenian
Lutty GA, McLeod DS, Bhutto I, Wiegand SJ. Effect of DC, Fruttiger M. Expression of neonatal Fc receptor in
VEGF trap on normal retinal vascular development the eye. Invest Ophthalmol Vis Sci. 2014;55(3):1607–
and oxygen-induced retinopathy in the dog. Invest 15. doi:10.1167/iovs.13-12574. iovs.13-12574 [pii].
Ophthalmol Vis Sci. 2011;52(7):4039–47. Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser
doi:10.1167/iovs.10-6798. iovs.10-6798 [pii]. PK, Chung CY, Kim RY. Ranibizumab for neovascular
Martin DF, Maguire MG, Ying GS, Grunwald JE, Fine age-related macular degeneration. N Engl J Med.
SL, Jaffe GJ. Ranibizumab and bevacizumab for neo- 2006;355(14):1419–31.
vascular age-related macular degeneration. N Engl Sato T, Wada K, Arahori H, Kuno N, Imoto K, Iwahashi-
J Med. 2011;364(20):1897–908. doi:10.1056/ Shima C, Kusaka S. Serum concentrations of bevaci-
NEJMoa1102673. zumab (avastin) and vascular endothelial growth
Martin DF, Maguire MG, Fine SL, Ying GS, Jaffe GJ, factor in infants with retinopathy of prematurity. Am
Grunwald JE, Toth C, Redford M, Ferris 3rd J Ophthalmol. 2012;153(2):327–33. doi:10.1016/j.
FL. Ranibizumab and bevacizumab for treatment of ajo.2011.07.005. e321S0002-9394(11)00582-4
neovascular age-related macular degeneration: two- [pii].
year results. Ophthalmology. 2012;119(7):1388–98. Simpson JL, Melia M, Yang MB, Buffenn AN, Chiang
doi:10.1016/j.ophtha.2012.03.053. MF, Lambert SR. Current role of cryotherapy in reti-
S0161-6420(12)00321-1 [pii]. nopathy of prematurity: a report by the American
Martinez-Castellanos MA, Schwartz S, Hernandez-Rojas Academy of Ophthalmology. Ophthalmology.
ML, Kon-Jara VA, Garcia-Aguirre G, Guerrero-Naranjo 2012;119(4):873–7. doi:10.1016/j.oph-
JL, Chan RV, Quiroz-Mercado H. Long-term effect of tha.2012.01.003. S0161-6420(12)00005-X [pii].
antiangiogenic therapy for retinopathy of prematurity up Smith LE, Shen W, Perruzzi C, Soker S, Kinose F, Xu X,
to 5 years of follow-up. Retina. 2013;33(2):329–38. Robinson G, Driver S, Bischoff J, Zhang B, Schaeffer
doi:10.1097/IAE.0b013e318275394a. JM, Senger DR. Regulation of vascular endothelial
Massin P, Bandello F, Garweg JG, Hansen LL, Harding growth factor-dependent retinal neovascularization by
SP, Larsen M, Mitchell P, Sharp D, Wolf-Schnurrbusch insulin-like growth factor-1 receptor. Nat Med.
UE, Gekkieva M, Weichselberger A, Wolf S. Safety 1999;5(12):1390–5. doi:10.1038/70963.
and efficacy of ranibizumab in diabetic macular edema Sonmez K, Drenser KA, Capone Jr A, Trese MT. Vitreous
(RESOLVE Study): a 12-month, randomized, con- levels of stromal cell-derived factor 1 and vascular
trolled, double-masked, multicenter phase II study. endothelial growth factor in patients with retinopathy of
Diabetes Care. 2010;33(11):2399–405. doi:10.2337/ prematurity. Ophthalmology. 2008;115(6):1065–70.
dc10-0493. dc10-0493 [pii]. doi:10.1016/j.ophtha.2007.08.050. e1061S0161-
Matsuyama K, Ogata N, Matsuoka M, Wada M, Takahashi 6420(07)00976-1 [pii].
K, Nishimura T. Plasma levels of vascular endothelial Stenson BJ. Oxygen targets for preterm infants.
growth factor and pigment epithelium-derived factor Neonatology. 2013;103(4):341–5. doi:10.1159/
before and after intravitreal injection of bevacizumab. 000349936. 000349936 [pii].
Br J Ophthalmol. 2010;94(9):1215–8. doi:10.1136/ Terry TL. Retrolental fibroplasia in the premature infant:
bjo.2008.156810. bjo.2008.156810 [pii]. V. Further studies on fibroblastic overgrowth of the
persistent tunica vasculosa lentis. Trans Am Zehetner C, Kirchmair R, Huber S, Kralinger MT,
Ophthalmol Soc. 1944;42:383–96. Kieselbach GF. Plasma levels of vascular endothelial
Velez-Montoya R, Clapp C, Rivera JC, Garcia-Aguirre G, growth factor before and after intravitreal injection of
Morales-Canton V, Fromow-Guerra J, Guerrero- bevacizumab, ranibizumab and pegaptanib in patients
Naranjo JL, Quiroz-Mercado H. Intraocular and sys- with age-related macular degeneration, and in patients
temic levels of vascular endothelial growth factor in with diabetic macular oedema. Br J Ophthalmol.
advanced cases of retinopathy of prematurity. Clin 2013;97(4):454–9. doi:10.1136/bjophthalmol-
Ophthalmol. 2010;4:947–53. 2012-302451. bjophthalmol-2012-302451 [pii].
Anti-vascular Endothelial Growth
Factor (VEGF) Treatment 3
in Neovascular Age-Related
Macular Degeneration: Outcomes
and Outcome Predictors
Dujon Fuzzard, Robyn H. Guymer,

and Robert P. Finger
BVZ Bevacizumab (Avastin®)

Abbreviations C2 Complement factor 2
C3 Complement factor 3
ABC Aflibercept (Eylea®)
C5 Complement factor 5
ABC-Trial Bevacizumab for neovas-
CATT Comparisons of Age-Related
cular age-related macular
Macular Degeneration
degeneration—multicenter ran-
Treatments Trial
domized double-masked study
CFB Complement factor B
AMD Age-related macular
CFT Central foveal thickness
degeneration
CFH Complement factor H
ANCHOR Anti-VEGF antibody for the
CFHR Complement factor H receptor
treatment of predominantly
CI Confidence interval
classic choroidal neovascular-
CMT Central macular thickness
ization in AMD—phase III
CNV Choroidal neovascularization
clinical trial
CRT Central retinal thickness
APOE Apolipoprotein E
CSMT Central subfield macular
AREDS Age-Related Eye Disease
thickness
study
ETDRS Early Treatment Diabetic
ARMS2 Age-related maculopathy sus-
Retinopathy Study
ceptibility 2 gene
EXCITE Efficacy and safety of monthly
BCVA Best corrected visual acuity
versus quarterly ranibizumab
BMES Blue Mountains Eye Study
treatment in neovascular age-
related macular degenera-
tion—phase IIIb clinical trial
D. Fuzzard (*) • R.H. Guymer • R.P. Finger
Department of Ophthalmology, Centre for Eye EXTEND-I Long-term efficacy and safety
Research Australia, Royal Victorian Eye and Ear of ranibizumab administered
Hospital, University of Melbourne, East Melbourne, pro re nata in Japanese patients
VIC 3002, Australia with neovascular age-related
e-mail: dujonfuzzard@gmail.com;
robyn.guymer@unimelb.edu.au; macular degeneration—open
robert.finger@unimelb.edu.au label phase I/II study

32 D. Fuzzard et al.
FA Fluorescein angiography Ranibizumab—phase III clini-

FLT1 fms-related tyrosine kinase 1 cal trial
(vascular endothelial growth MedDRA SOC Medical Dictionary for
factor/vascular permeability Regulatory Activities System
factor receptor) Organ Classes
FOCUS Ranibizumab combined with MT Macular thickness. In the con-
verteporfin photodynamic ther- text of this chapter, this encom-
apy in neovascular age-related passes a variety of anatomical
macular degeneration—phase measures, including CFT,
I/II randomized-controlled trial CMT, CRT, CSMT, and FT
FT Foveal thickness NR Not reported
GEFAL Ranibizumab versus bevaci- nv Neovascular
zumab for neovascular age- OCT Optical coherence tomography
related macular degeneration—a PCV Polypoidal choroidal
non-inferiority randomized trial vasculopathy
HELIOS Real-world variability in ranibi- PDT Photodynamic therapy
zumab treatment and associated PED Pigment epithelial detachment
clinical quality of life and safety PIER Randomized double-masked,
outcomes over 24 months— sham-controlled trial of ranibi-
prospective, observation open- zumab for neovascular age-
label study with arms based in related macular degeneration
Belgium and the Netherlands PLA2G12A Group 12 secretory phospholi-
HORIZON Open label extension trial of pase A2 gene
ranibizumab for choroidal neo- PRN Pro re nata, i.e., as required
vascularization secondary to age- PrONTO Prospective OCT Study with
related macular degeneration Lucentis for Neovascular AMD
HTRA1 High-temperature requirement RAP Retinal angiomatous
A-1 proliferations
IL Interleukin RBZ Ranibizumab (Lucentis®)
IQR Inter-quartile range RCT Randomized-controlled trial
IVAN Randomized-Controlled Trial RPE Retinal pigment epithelium
of Alternative Treatments to SD Standard deviation
Inhibit VEGF in Age-related SAILOR A phase IIb study to evaluate
Choroidal Neovascularization the safety of ranibizumab in
LogMAR Logarithm of the minimum subjects with neovascular age-
angle of resolution related macular degeneration
LUMINOUS Safety of ranibizumab in rou- SE Standard error
tine clinical practice: 1-year SECURE Long-term safety of ranibi-
retrospective pooled analysis zumab 0.5 mg in neovascular
of four European neovascular age-related macular degenera-
AMD registries. tion—24-month phase IV
MANTA A randomized double-masked extension study
trial comparing the visual out- SEVEN-UP Seven-year outcome in
come after treatment with ranibizumab-treated patients
ranibizumab or bevacizumab in in ANCHOR MARINA and
patients with neovascular age- HORIZON—a multicenter
related macular degeneration cohort study
MARINA Minimally Classic/Occult SNP Single nucleotide
Trial of Anti-VEGF antibody polymorphisms
3 Anti-vascular Endothelial Growth Factor (VEGF) Treatment in Neovascular… 33
SUSTAIN Safety and efficacy of a flexible would be legally blind due to their nvAMD
dosing regimen of ranibizumab within 2 years (Bressler et al. 2011). This high-
in neovascular age-related mac- lights not only the devastating natural history of
ular degeneration—12-month the disease, but also the importance of data on
phase III single arm study long-term outcomes and the predictors of out-
UK United Kingdom comes for this potentially lifelong treatment.
US United States Several anti-VEGF agents are available,
VA Visual acuity including aflibercept (ABC; Eylea®, Regeneron,
VEGF Vascular endothelial growth Tarrytown, New York, approved use), ranibi-
factor zumab (RBZ, Lucentis®, Genetech, San
VIEW VEGF Trap-Eye: Investigation Francisco, approved use), and bevacizumab
of efficacy and safety in wet (BVZ, Avastin®, Roche, Basel, Switzerland, off-
AMD—non-inferiority trial label) (Wickremasinghe et al. 2011). BVZ is
comparing aflibercept with approved for intravenous use in treatment of
ranibizumab some cancers; however, due to its similar phar-
WAVE Lucentis in wet AMD: macological profile to RBZ (Braithwaite et al.
Evaluation of visual acuity and 2014), it is used off-label as a less-expensive
quality of life—German non- alternative agent (Miller et al. 2013). Also avail-
interventional clinical practice able, but less commonly used due to its lower
study clinical effectiveness compared with RBZ, ABC,
and BVZ is pegaptanib (Macugen®, Pfizer,
New York), the first anti-VEGF treatment regis-
tered for treatment in the eye (Takeda et al. 2007).
3.1 Introduction Currently, available data on treatment outcomes
and their predictors for RBZ, BVZ, and ABC are
The late-stage neovascular form of age-related summarized in this chapter.
macular degeneration (nvAMD) is the largest
single cause of irreversible, severe vision loss in
all developed countries (Lim et al. 2012b). 3.2 Overview of Anti-VEGF
Current treatment options that preserve sight and Treatment Outcomes
lead to a considerable improvement in some
patients are based on the inhibition of vascular Establishment of optimized treatment regimes in
endothelial growth factor (anti-VEGF) and are terms of agent choice and frequency of injec-
delivered as regular intravitreal injections (Brown tions, based on patient factors at presentation,
et al. 2006; Rosenfeld et al. 2006). As anti-VEGF remains an important focus of current research,
treatment does not cure nvAMD but only con- to achieve optimally effective and safe anti-
trols it, treatment is for the long term and is often VEGF treatment outcomes (Finger et al. 2014).
required until either the eye worsens to levels, In this chapter, treatment outcomes will be pre-
where treatment is no longer indicated or the sented by visual acuity (VA) outcomes, anatomi-
patient dies. To date, however, data on treatment cal outcomes as assessed by optical coherence
outcomes and their predictors beyond 2 years are tomography (OCT), the incidence of legal blind-
sparse (Rasmussen and Sander 2014). In just 1 year ness and safety.
in the USA, an estimated 151,340 non-Hispanic
whites develop nvAMD in one eye, of whom one
third have preexisting nvAMD in the other eye 3.2.1 Visual Acuity Outcomes
(Bressler et al. 2011). Out of these, 103,582 new
cases are in need of anti-VEGF treatment. There are considerable differences between the
Without treatment, it is estimated that 16 % patient populations and treatment outcomes of
phase III clinical trials published to date and Long-term follow-up of patients treated with
those encountered in routine clinical practice RBZ was reported in the SEVEN-UP study
(Table 3.1) (Finger et al. 2014). To illustrate this (Rofagha et al. 2013), a multicenter cohort study
point, available information from each setting is which reviewed 65 study eyes, which had suc-
discussed separately. Whilst it exists as a treat- cessfully completed ANCHOR or MARINA pro-
ment option, pegaptanib has not been included in tocols within the RBZ arms (which received
this work, due to the cessation of its use in clini- 4-weekly RBZ for 2 years) followed by comple-
cal practice. The volume of high-level evidence tion of the HORIZON study protocol (Singer
available varies considerably between agents. et al. 2012), which administered RBZ to partici-
pants on a PRN basis for a further 2 years. On
3.2.1.1 Phase III Clinical Trials average, SEVEN-UP participants were followed
up 7.3 years (range 6.33–8.49 years) after enter-
Ranibizumab ing either the ANCHOR or MARINA trials
The efficacy of RBZ in treatment of AMD was (Rofagha et al. 2013). Vision outcomes varied
demonstrated to be superior to no treatment or widely; whilst 37 % of eyes had VA of 20/70 or
photodynamic therapy (PDT) over a 2-year better and 43 % of eyes had a stable VA compared
period in the ANCHOR and MARINA phase III with baseline, 37 % of eyes were legally blind,
clinical trials (Brown et al. 2009; Rosenfeld et al. with a VA of 20/200 or worse in the study eye.
2006). Where different doses of RBZ have been This corresponded to a mean decrease in VA of
studied, 0.5 mg has appeared superior to 0.3 mg 8.6 ETDRS letters when compared with baseline
in terms of the mean numbers of letters gained measures. Despite being the longest period of
(Mitchell 2011). Outcomes in terms of VA and follow-up available in a cohort of patients treated
central retinal thickness (CRT) have been best for nvAMD with anti-VEGF agents, this study
with regular 4-weekly treatment regimes com- had multiple limitations. The low enrollment rate
pared to regular 3-monthly or as required (pro re of eligible participants (42 %) that were not ran-
nata (PRN)) treatment strategies (Lanzetta et al. domly sampled may have resulted in selection
2013). Quarterly RBZ dosing has been demon- bias. Furthermore, the absence of a treatment pro-
strated not to maintain initial treatment gains at tocol during the 3–4-year period following com-
12 months after an initial loading phase or pletion of the HORIZON trial resulted in wide
monthly injections for 3 months in the EXCITE differences in the number of injections given to
and PIER studies (Schmidt-Erfurth et al. 2011; different patients, with 41 % receiving no further
Regillo et al. 2008). A similar pattern has been anti-VEGF therapy at all. Like the PIER study
observed with PRN treatment regimes in the (Regillo et al. 2008), SEVEN-UP demonstrated
SUSTAIN and SAILOR trials, where the full that under-treatment of nvAMD with anti-VEGF
extent of treatment gains after a loading phase is agents would lead to worsening of visual out-
not maintained to the 12-month mark (Holz et al. comes (Rofagha et al. 2013).
2011; Boyer et al. 2009). Where PRN regimes
are more aggressive (i.e., averaging greater than Aflibercept
six injections in 12 months as in the CATT and The efficacy of ABC 2.0 mg administered
HARBOR studies) initial gains are maintained to 8-weekly was compared to that of RBZ 0.5 mg
the 12-month mark (Martin et al. 2011; Busbee administered 4-weekly in the VIEW studies
et al. 2013), although regular 4-weekly treatment (Heier et al. 2012; Schmidt-Erfurth et al. 2014).
regimes yielded overall better results (Martin These comprised two multinational, double-
et al. 2012; Busbee et al. 2013). The mean gain in blinded randomized-controlled trials (RCTs).
VA reported with regular 4-weekly regimes VIEW-1 was conducted in the USA and Canada
across phase III clinical trials has varied between with 1217 patients, whilst VIEW-2 was per-
7 and 11 Early Treatment Diabetic Retinopathy formed in Europe with 1240 patients across
Study (EDTRS) letters (Lanzetta et al. 2013). Europe, Asia Pacific, and Latin America (Heier
Table 3.1 Visual acuity (VA) outcomes of anti-VEGF treatment reported in clinical trials and clinical practice settings
Setting (clinical Treatment regime VA at >2 years Injections
practice or Drug and dose (n per regime at Mean age Mean VA VA at 3 months (n Injections VA at 1 year Injections VA at 2 years Injections at (Time point) (n at at specified
Study clinical trial) (if reported) baseline) at baseline at baseline at 3 months) at 3 months (n at 1 year) at 1 year (n at 2 years) 2 years given time point) time point
CATT (Martin Clinical trial RBZ 0.5 mg Monthly (n = 146) 79.5 (SD 59.9 (SD 65.0 (SD 12.7) NR 68.8 (SD 17.7) 11.7 68.5 (SD 18.9) 22.4 (SD NR NR
et al. 2011, 2012) 7.4) 12.2) (n = 273) (n = 284) (SD 1.5) (n = 134) 3.9)
Monthly (1 year) 78.8 (SD 60.9 (SD 67.7 (SD 18.5) NR (5.0 in
then PRN (n = 138) 7.5) 14.3) (n = 130) second year)
PRN (n = 298) 78.3 (SD 61.6 67.2 (SD 13.8) 68.4 (SD 16.4) 6.9 (SD 3.0) 68.5 (SD 15.3) 12.6 (SD
7.8) (SD 13.1) (n = 277) (n = 285) (n = 264) 6.6)
66.3 (SD 14.3)
(n = 260)
BVZ 1.25 mg Monthly (n = 135) 79.7 (SD 60.2 (SD 66.3 (SD 14.3) 68.4 (SD 18.2) 11.9 (SD 1.2) 68.2 (SD 16.1) 23.4 (SD
7.5) 13.6) (n = 260) (n = 265) (n = 129) 2.8)
Monthly (1 year) 80.4 (SD 60.4 (SD 7.7 (SD 3.5) 65.0 (SD 21.8) NR (5.8 in
then PRN (n = 131) 7.1) 12.4) (n = 122) second year)
PRN (n = 300) 78.9 (SD 60.6 (SD 66.2 (SD 13.7) 66.5 (SD 19.0) 66.0 (SD 19.9) 14.1 (SD
7.4) 13.0) (n = 275) (n = 271) (n = 251) 7.0)
a
IVAN Clinical trial RBZ 0.5 mg Monthly (n = 157) 77.8 (SD 61.8 (SD NR NR 69.0 (SD 16.0) NR 67.8 (SD 17.0) NR NR NR
(Chakravarthy 7.6) 15.0) (n = 287) (n = 271)
et al. 2012, 2013)
3-dose loading phase,
then PRN (n = 155)
BVZ 1.25 mg Monthly (n = 149) 77.8 (SD 61.1 (SD 66.1 (SD 17.4) 66.1 (SD 18.4)
8.0) 15.6) (n = 274) (n = 254)
3-dose loading phase,
then PRN (n = 145)
Analysis Monthly (agents 77.8 (SD 60.0 (SD 66.8 (SD 17.4) 66.6 (SD 17.9)
combining agents combined) (n = 308) 8.0) 15.5) (n = 277) (n = 261)
3-dose loading phase, 77.6 (SD 62.9 (SD 68.4 (SD 16.1) 67.3 (SD 17.5)
then PRN (agents 6.8) 15.0) (n = 284 (n = 264)
combined) (n = 302)
GEFAL Clinical trial RBZ 0.5 mg 3-dose loading phase, 78.68 (SD 55.78 (SD 59.9 (SD 11.9) NR 59.4 (SD 14.2) 6.5 (SD 2.4) NR NR NR NR
(Kodjikian et al. then PRN (n = 183) 7.27) 13.99) (n = 176) (n = 176)
2013)
Data for cohort BVZ 1.25 mg 3-dose loading phase, 79.62 (SD 54.62 (SD 59.9 (SD 10.6) 60.0 (SD 14.4) 6.8 (SD 2.7)
which completed then PRN (n = 191) 6.90) 14.07) (n = 187) (n = 187)
study protocol
MANTA (Krebs Clinical trial BVZ 1.25 mg 3-dose loading phase, 76.7 (SD 57.0 (SD NR NR 62.2 (CI 60.1 9.1 (SD 2.8) NR NR NR NR
et al. 2013) then PRN (n = 154) 7.8) 13.0) –64.3) (n = 121)
RBZ 0.5 mg 3-dose loading phase, 77.6 (SD 56.4 (SD 60.7 (CI 58.7 8.8 (SD 2.7)
then PRN (n = 163) 8.1) 13.5) –62.8) (n = 127)
(continued)
Table 3.1 (continued)
VIEW 1 (Heier Clinical trial RBZ 0.5 mg 4-weekly (n = 304) 78.2 (SD 54.0 (SD NR NR 62.1 (SD 15.3) 12.1–12.5 for all NR NR NR NR
et al. 2012) 7.6) 13.4) (n = 269) monthly treatment
arms (SD NR)
ABC 0.5 mg 4-weekly (n = 301) 78.4 (SD 55.6 (SD 62.5 (SD 13.4)
8.1) 13.1) (n = 270)
ABC 2 mg 4-weekly (n = 304) 77.7 (SD 55.2 (SD 66.1 (SD 13.8)
7.9) 13.2) (n = 285)
3-month loading 77.9 (SD 55.7 (SD 63.6 (SD 15.0) 7.5 (SD NR)
phase, then 8-weekly 8.4) 12.8) (n = 265)
(n = 301)
VIEW 2 (Heier Clinical trial RBZ 0.5 mg 4-weekly (n = 291) 73.0 (SD 53.8 (SD NR NR 63.4 (SD 13.5) 12.2–12.4 for all NR NR NR NR
et al. 2012) 9.0) 13.5) (n = 269) monthly treatment
arms (SD NR)
ABC 0.5 mg 4-weekly (n = 296) 74.7 (SD 51.6 (SD 61.3 (SD 14.1)
8.6) 14.2) (n = 268)
ABC 2 mg 4-weekly (n = 309) 74.1 (SD 52.8 (SD 60.4 (SD 12.6)
8.5) 13.9) (n = 274)
3-month loading 73.8 (SD 51.6 (SD 60.5 (SD 14.4) 7.5 (SD NR)
phase, then 8-weekly 8.6) 13.9) (n = 270)
(n = 306)
VIEW studies Clinical trial RBZ 0.5 mg 4-weekly (n = 595) 75.6 (SD 53.9 (SD NR NR 62.6 (SD NR) 12.1–12.5 for all 61.8 (n = 519) 16.5 (SD NR NR
combined (Heier 8.7) 13.4) (n = 560) monthly treatment 3.7)
et al. 2012; arms (SD NR)
Schmidt-Erfurth ABC 0.5 mg 4-weekly (n = 597) 76.5 (SD 53.6 61.9 (n = 551) 60.2 (n = 502) 16.2 (SD
et al. 2014) 8.5) (13.8) 4.0)
ABC 2 mg 4-weekly (n = 613) 75.9 (SD 54.0 63.3 (n = 574) 61.6 (n = 529) 16.0 (SD
8.4) (13.6) 3.2)
3-month loading 75.8 (SD 53.6 62 (n = 560) 7.5 (SD NR) 61.2 (n = 513) 11.2 (SD
phase, then 8-weekly 8.8) (13.5) 2.9)
(n = 607)
PIER (Regillo Clinical trial RBZ 0.3 mg 3-month loading 78.7 (SD 55.8 (SD NR NR 54.2 (SD 15.1) NR 53.6 (SD 15.6) NR NR NR
et al. 2008; phase, then quarterly. 6.3) 12.2) (n = 59) (n = 53)
Abraham et al. Switched to monthly
2010) at month19 (n = 60)
RBZ 0.5 mg 3-month loading 78.8 (SD 53.7 (SD 53.5 (SD 13.1) 51.4 (SD 14.4)
phase, then quarterly. 7.9) 15.5) (n = 58) (n = 54)
Switched to monthly
at month19 (n = 61)
EXCITE Clinical trial RBZ 0.3 mg Monthly (0.3 mg) 75 (SD 56.5 (SD 64.0 (SD NR) NR 64.5 (SD 16.27) 11.4 (SD 1.69) NR NR NR NR
(Schmidt-Erfurth (n = 115) 8.26) 12.19) (n = NR (n = 101)
et al. 2011)
3-dose loading phase, 75.1 (SD 55.8 (SD 62.6 (SD NR) 60.2 (SD 16.01) 5.7 (SD 0.80)
then quarterly 7.45) 11.81) (n = NR) (n = 104)
(0.3 mg) (n = 120)
RBZ 0.5 mg 3-dose loading phase, 75.8 (SD 57.7 (SD 64.3 (SD NR) 61.3 (SD 16.32) 5.5 (SD 1.05)
then quarterly 6.96) 13.06) (n = NR) (n = 88)
(0.5 mg) (n = 118)
SUSTAIN (Holz Clinical trial RBZ 3-dose loading phase, 75.1 (SD 56.1 (SD 61.9 (SD 11.12) 2.9 (SD 59.7 (SD 13.89) 5.6 (SD 2.37) NR NR NR NR
et al. 2011) 0.3 mg/0.5 mg then PRN (n = 513) 8.06) 12.19) (n = 509) 0.35) (n = 509)
SECURE (Silva Clinical trial RBZ Completion of 74.5 (SD NR NR NR 60.7 (SD 16.14) NR 58.7 (SD NR 56.4 (SD 13.04) 6.1 (SD
et al. 2013) 0.3 mg/0.5 mg EXCITE or 7.62) (n = 231) 10.44) (n = 230) 5.67)
SUSTAIN protocol (n = 230)
Number of
(1 year), then PRN
injections
(n = 234)
from end
of Year 1
to
36 months
SAILOR (Boyer Clinical trial RBZ 0.3 mg 3-dose loading phase, 79.9 (SD 55.0 (SD 60.8 (SD NR) 96 % of 55.5 (SD NR) 4.6 (combined figure NR NR NR NR
et al. 2009) then PRN (n = 462) 7.9) 12.5) cohort for both groups) (SD
(Note: only received 3 NR) (81.7 %
RBZ 0.5 mg 3-dose loading phase, 75.8 (SD 48.9 (SD 55.9 (SD NR) 51.3 (SD NR)
treatment naïve doses in follow-up reported)
then PRN (n = 490) 8.0) 13.8)
cohorts are first 3
included) months
ANCHOR Clinical trial RBZ 0.3 mg Monthly (n = 140) 77.4 (SD 47.0 (SD 53.8 (SD NR) NR 55.5 (SD 14.6) 11.0 (SD NR) 55.1 (SD 16.2) 21.5 (SD NR NR
(Brown et al. 7.5) 13.1) (n = NR) (n > 126) (n = 117) NR)
2006, 2009)
RBZ 0.5 mg Monthly (n = 139) 76.0 (SD 47.1 (SD 57.1 (SD NR) 58.4 (SD 14.6) 11.2 (SD NR) 57.8 (SD 16.5) 21.3 (SD
8.6) 13.2) (n = NR) (n > 126) (n = 116) NR)
MARINA Clinical trial RBZ 0.3 mg Monthly (n = 238) 77 (SD 8) 53.1 (SD 58.2 (SD NR) NR 59.6 (SD) NR 58.5 (SD) NR NR NR
(Rosenfeld et al. 12.9) (n = NR) (n = 226) (n = 210)
2006)
RBZ 0.5 mg Monthly (n = 240) 77 (SD 8) 53.7 (SD 59.6 (SD NR) 60.9 (SD) 60.3 (SD)
12.8) (n = NR) (n = 226) (n = 215)
HORIZON Clinical trial RBZ 0.5 mg Monthly for 2 years 76.1 (SD 51.6 (SD NR NR NR NR 60.5 (SD 17.9) NR 55.7 (SD NR) NR
(Singer et al. (in ANCHOR, 7.6) 13.0) (n = 600) (n = 481)
2012) MARINA or 36 months
FOCUS), then PRN
53.6 (SD NR)
(n = 600)
(n = 388)
48 months
51.5 (SD NR)
(n = 73)
60 months
(continued)
SEVEN-UP Clinical practice RBZ 0.5 mg Monthly for 2 years 74.9 (SD 54.3 (SD NR NR NR NR NR NR 45.7 (SE 2.8) 6.8 (Range
(Rofagha et al. follow-up of a (n = 36) and (in ANCHOR or 7.7) 12.4) (n = 65) 0–46)
2013) Clinical trial 0.3 mg (n = 29). MARINA), then 7 year follow-up Since exit
sample Data pooled PRN for 2 years in from
HORIZON, then HORIZON
PRN (n = 65) (from 4
years
onwards)
EXTEND-I Clinical trial RBZ 0.3 mg Monthly (n = 35) 70.7 (SD 46.7 (SD NR NR 57.2 (SD NR) 11.2 (Range 3–12) NR NR 55.4 (SD 17.14) 4.1 (SD
(Tano and Ohji NR) 11.8) (n = 31) (n = 28) 4.12)
2010, 2011) 2.7 years (SD No. of
0.35) injections/
year after
12 months
RBZ 0.5 mg Monthly (n = 41) 71.6 (SD 48.1 (SD 57.6 (SD NR) 11.1 (Range 3–12) 57.6 (SD 15.36) 3.9 (SD
NR) 10.8) (n = 37) (n = 33) 4.63)
2.93 years (SD No. of
0.09) injections/
year after
12 months
PrONTO (Fung Clinical trial RBZ 0.5 mg 3-dose loading phase, 83.5 (SD 56.2 (SD 67.0 (SD NR) 3 (SD NR) 65.5 (SD NR) 5.6 (SD 2.3) 67.0 (SD 12.2) 9.9 (SD 5.3) NR NR
et al. 2007; then PRN (n = 40) 7.2) NR) (n = 40) (n = 40) (n = 37)
Lalwani et al.
2009)
ABC-Trial Clinical trial BVZ 1.25 mg 3-dose loading phase 79 (SD 50 (IQR 56.7 (SD NR) NR 57 (SD NR) 7.1 (Range 3–9) NR NR NR NR
(Tufail et al. (6-weekly intervals) NR) 43–61) (n = NR) (n = 64)
2010) then PRN (n = 65)
Sacu et al. (2009) Clinical trial BVZ 1 mg 3-dose loading phase, 78 (SD 8) 50 (SD 62 (SD NR) NR 58 (SD NR) 6.8 (SD NR) NR NR NR NR
then PRN (n = 14) NR) (n = 14) (n = 14)
(n = 14)
Kruger Falk et al. Clinical RBZ 3-month loading 77.3 53.2 NR NR NR NR NR NR 50.5 (Variable 8.7 (Range
(2013) Practice phase, then variable (Range (Range follow-up time. 1–35)
regimes (n = 855) 54–98) 1–85) Mean
23.3 months,
range 4–48
months) (n = 855)
Rasmussen et al. Clinical RBZ 3-dose loading phase, 79.9 (SD 0.24 (SD 0.26 (SD NR) NR NR NR NR NR 10.2 (SD 0.18 (SD NR) NR
(2013) Practice then PRN (n = 600) 9.9) NR) NR)
Note: Visual 4 years – —last
acuity recorded visit carried
as Snellen forward (n = 600)
equivalent Subgroup who 78.0 (SD 0.30 0.38 (95%CI 0.35 (95%CI 5.2 (SD NR) 0.32 (95%CI 22.1 (SD
completed 4 years of 11.5) (95%CI 0.35–0.42) NR) 0.29–0.35) NR)
screening and 0.28–0.33) (n = 192) (n = 192)
re-treatment (n = 192) (n = 192)
Muniraju et al. Clinical RBZ 3-dose loading phase, 82.7 48.2 (SD NR NR 51.2 (SD 18.7) 4.8 (SD 2.2) 50.4 (SD 20.8) 7.8 (SD 4.2) 49.1 (SD 21.7) 10.2 (SD
(Muniraju et al. Practice then PRN (n = 292) (Range 16.9) (n = NR) (n = NR) (n = 192) 6.2)
2013) 55–97) 36 months
Marques et al. Clinical RBZ PRN (n = 84) 77.39 49.3 (SD 52.6 (SD 15.3) 1.53 (SD 50.7 (SD 15.9) 3.75 (SD 1.20) 48.8 (SD 18.8) 6.35 (SD 47.7 (SD 18.7) 8.67 (SD
(Marques et al. Practice (Range 15.2) (n = 67) 0.53) (n = 84) (n = 77) 2.3) (n = 52) 3.3)
2013) 61–94) 34.3 months (SD
6.9)
Rung and Clinical RBZ 3-dose loading phase, 76 (SD 7) 53 (SD 61 (SD 14) NR NR NR NR NR 44 (SD 24) 7.8 (SD
Lovestam-Adrian Practice then PRN (n = 66) 14) (n = 51) 5.0)
(Rung and 37-month
Lovestam-Adrian follow-up
2013)
WAVE Germany Clinical RBZ 3-month loading 77.6 (SD 48.8 (SD 0.64 (SE 0.01) 2.95 (SE 48.0 (SD 11.7) 4.34 (SD 1.9) NR NR NR NR
(Finger et al. Practice phase, then PRN 7.8) 18.7) (n = 3124) 0.01) (n = 2587)
2013; Holz et al. (n = 3470)
2013)
Note: Visual
acuity at 3
months recorded
using LogMAR
scale
HELIOS Clinical RBZ 3-dose loading phase, 77.9 (SD 45.1 (SD NR NR 50.7 (SD 24.0) 5.1 (SD 2.4) NR NR NR NR
Netherlands Practice then PRN (n = 243) 8.0) 21.5) (n = 208)
(Holz et al. 2013)
HELIOS Clinical RBZ 3-dose loading phase, 78.5 (SD 56.3 (SD 61.7 (SD 14.9) 2.5 (SD 58.5 (SD 17.8) 5.0 (SD 2.1) 53.3 (SD 19.3) 7.6 (SD 4.1) NR NR
Belgium (Rakic Practice then PRN (n = 267) 7.3) 14.3) (n = NR) 0.7) (n = 206) (n = 184)
et al. 2013) (2.5-month
follow-up)
Sweden Clinical RBZ 3-dose loading phase, 77.7 (SD 58.3 (SD 63.3 (SD 12.5) 95 % of 59.3 (SD 16.2) 4.7 (SD 1.6) NR NR NR NR
ranibizumab Practice then PRN (n = 370) 8.0) 12.2) patients (n = 370)
registry had 3
(Hjelmqvist et al. injections
2011) at 3 months
Data reported for
patients who
remained in
treatment for 12
months (370/471)
(continued)
NvAMD Clinical RBZ 3-month loading 79.1 (IQR 55 NR NR 57 (n = 8598) 5.7 (Range 1–13) 55 (n = 4990) 9.4 (mean 53 (n = 2470) 13.1 (mean
database, UK Practice phase, then PRN 75–85) 3.7 in second (3 years) 3.7 in third
(The neovascular (n = 12,951) year, range year, range
age-related 0–13) 0–12)
macular
degeneration
database:
multicenter study
of 92 976
ranibizumab
injections: report
1: visual acuity
2014)
Only studies reporting outcomes to 1 year or greater have been included. All VA measurements were reported using the Early Treatment Diabetic Retinopathy Study (ETDRS) measurement
unless otherwise stated. Loading doses were spaced in 4-weekly intervals unless otherwise stated
NR not reported, IQR inter-quartile range, SD standard deviation, SE standard error
a
IVAN trial reported findings according to drug allocation and regime allocation, but did not report findings for individual subgroups.
et al. 2012). This work demonstrated 8-weekly BVZ and RBZ. Like the CATT study, this work
ABC 2.0 mg to be non-inferior to 4-weekly RBZ demonstrated a similar efficacy of the two agents;
0.5 mg in terms of vision and anatomical out- however, a difference in efficacy between regular
comes over 52 weeks (Heier et al. 2012). Both and PRN treatment regimes was not found
cohorts were subsequently followed up to 96 (Chakravarthy et al. 2013). The MANTA (Krebs
weeks (Schmidt-Erfurth et al. 2014). For weeks et al. 2013) and GEFAL (Kodjikian et al. 2013)
52–96, patients were treated on a PRN basis, with studies have contributed further comparative data
ongoing monthly monitoring and a 12-weekly between BVZ and RBZ over a 1-year treatment
minimum injection frequency. Small losses from duration, utilizing a protocol of three initial treat-
gains achieved at 52 weeks were noted in both ments spaced monthly followed by further treat-
cohorts; however, equal efficacy between ment as required. Participants in the MANTA
4-weekly RBZ 0.5 mg and 8-weekly ABC 2.0 mg study received more treatments across both treat-
was demonstrated (Schmidt-Erfurth et al. 2014). ment arms than the GEFAL cohorts; however, VA
outcomes were comparable. Both studies found
Bevacizumab BVZ to be non-inferior to RBZ for VA outcomes
Due to its off-label use, highly standardized clini- after 1 year (Krebs et al. 2013; Kodjikian et al.
cal trials testing BVZ for nvAMD have only 2013). A subsequent meta-analysis, including all
recently been completed. A review published in above-mentioned comparison studies performed
2011 (Mitchell 2011) identified only one study by the research group who conducted the GEFAL
constituting Level I evidence (Tufail et al. 2010), study, concluded that the BVZ and RBZ treat-
which demonstrated a positive effect of BVZ ments achieved similar VA outcomes (Kodjikian
compared to pegaptanib, PDT, and sham injec- et al. 2014).
tions. In 2009, a small study compared the effi-
cacy of BVZ with RBZ over a 6-month period for 3.2.1.2 Outcomes Achieved
the first time (Subramanian et al. 2009). Since in Routine Medical Practice
then, four large comparative studies have reported Treatment outcomes of nvAMD using anti-
results (Martin et al. 2011; 2012; Chakravarthy VEGF agents appear to be worse in routine medi-
et al. 2012, 2013; Krebs et al. 2013; Kodjikian cal practice, when compared with data generated
et al. 2013). In 2012, 2-year results were pub- from phase III clinical trials (Finger et al. 2014;
lished for the CATT study, which compared RBZ Rasmussen and Sander 2014; the neovascular
0.5 mg and BVZ 1.25 mg across regular 4-weekly age-related macular degeneration database: mul-
and PRN injection regimes, which did not include ticenter study of 92,976 ranibizumab injections:
a loading phase (Martin et al. 2012). Of the 1185 report 1: visual acuity 2014). In much of the pub-
patients enrolled in the study, 1107 (93.4 %) were lished work from the clinical trial setting, a stable
followed up during the second year. This trial VA was defined as a loss of less than 15 EDTRS
found similar efficacy of the two agents on VA letters (Martin et al. 2011; Brown et al. 2009;
over a 2-year period, despite a significant Rosenfeld et al. 2006; Krebs et al. 2013; Tufail
difference in the proportion of participants with- et al. 2010). In the trial setting, approximately
out the presence of fluid on OCT in favor of RBZ 90 % of participants achieved stable VA after 2
(Martin et al. 2012). Groups receiving regular years of treatment according to this definition
4-weekly injections with either agent achieved (Martin et al. 2011; Brown et al. 2009; Rosenfeld
higher visual gains than those receiving PRN et al. 2006; Krebs et al. 2013; Tufail et al. 2010).
regimes. In the case of BVZ, the mean reported In comparison, research in clinical practice has
gains after 2 years were 7.8 letters for 4-weekly tended to find 70–80 % of patients maintained
dosing and 5.0 letters for PRN regimes. A second stable vision with treatment (Rasmussen and
multicenter, non-inferiority factorial trial (IVAN) Sander 2014). Clinical practice studies have
studied 610 patients with nvAMD over 2 years, focused almost exclusively on RBZ, with a
comparing regular and PRN treatment regimes of regime consisting of three initial monthly doses
followed by monthly reviews and treatment on a with 10.2 (SD 6.2) injections (Muniraju et al.
PRN basis. Fewer studies have reported out- 2013), the other studies recorded losses of 1.6 and
comes for a Treat & Extend regimen, in which 7 ETDRS letters with 8.67 (SD 3.3) and 7.8 (SD
intravitreal injections are given monthly until dis- 5.0) injections, respectively (Marques et al. 2013;
ease inactivity has been achieved, defined as sta- Rung and Lovestam-Adrian 2013). A very large
ble VA, absence of fluid on OCT, and no new study utilized the United Kingdom national
hemorrhage. Following this, patient visits are nvAMD database to examine VA outcomes of
extended by a 2-week interval from 4 to 6 to 8, 12,951 eyes of 11,135 individuals over a treat-
etc., weeks with an intravitreal injection given at ment period of up to 3 years (The neovascular
each visit as long as there is no sign of disease age-related macular degeneration database: multi-
activity (no loss of VA, a dry OCT, and no new center study of 92,976 ranibizumab injections:
hemorrhage). In case of a reactivation (loss of VA report 1: visual acuity 2014). The 2470 eyes that
or increase/presence of fluid on OCT, or fresh completed 3 years of follow-up had a loss of 2
hemorrhage), intervals are reduced again. ETDRS letters on average, with a mean of 13.1
Several studies have reported treatment out- injections over this time period. Whilst longer
comes of Treat & Extend regimes for nvAMD, term follow-up has demonstrated a further decline
which achieved improvements in VA ranging in vision over time, the relatively low number of
from an 11 letter gain at 1 year to a 15 letter gain injections administered to patients in clinical
at 2 years with a very similar mean number of practice studies may indicate under-treatment in
injections (7–8 in year 1 and 5–6 in year 2) this setting (Rasmussen and Sander 2014).
(Oubraham et al. 2011; Abedi et al. 2014). The The need for ongoing follow-up of patients
OCT-guided Treat & Extend treatment protocol with nvAMD was highlighted in a Danish longi-
has been shown to lead to treatment outcomes tudinal study over 4 years, where 192 of 600
comparable to outcomes in phase III clinical tri- eyes (32 %) followed were still receiving active
als with fewer injections over the first 2 years treatment (Rasmussen et al. 2013). Vision had
(Lalwani et al. 2009; Gupta et al. 2010; Abedi been maintained over 4 years compared with
et al. 2014). baseline readings with a mean of 22.1 injections
Populations that received fewer injections over this period and was significantly better than
have not fared as well. A large study produced in those who had discontinued treatment.
Demark followed a cohort of 855 patients under- Furthermore, of 120 eyes (20 %) that had ceased
going PRN treatment with RBZ for AMD over a treatment as a result of apparent disease inactiv-
4-year period, with a mean follow-up time of ity, 25 were subsequently referred for further
23.3 months (Kruger Falk et al. 2013). A mean VA treatment (Rasmussen et al. 2013).
loss of 3 ETDRS letters was recorded, with a Patient registries have been examined in
mean of 8.7 injections over the time frame stud- Germany, the Netherlands, Belgium, and Sweden
ied. During the study period, 399 patients (46.7 %) through the LUMINOUS program to assess the
discontinued treatment for several different rea- efficacy of RBZ for nvAMD in clinical practice
sons. Chiefly, this was due to the absence of fur- over a 1-year treatment period (Holz et al. 2013).
ther disease activity in 181 cases, treatment being Patient age was similar across the four registries,
deemed futile in 113 cases and 36 cases where between 77.6 (SD 7.8) in Germany and 78.7 (SD
patients declined further injections (Kruger Falk 6.8) in Belgium; however, the mean baseline VAs
et al. 2013). Three clinical practice studies have were higher in the Belgian (56.3, SD 14.2) and
documented VA outcomes after approximately 3 Swedish (58.3, SD 12.2) registries compared
years (34–37 months) from commencing treat- with those of the Netherlands (45.1, SD 21.5) and
ment (Muniraju et al. 2013; Marques et al. 2013; Germany (48.8, SD 18.7). Over 1 year of treat-
Rung and Lovestam-Adrian 2013). Whilst ment, three populations gained between 1.0 and
patients in the largest of these studies maintained 5.6 letters with between 4.7 and 5.7 injections,
a modest gain from baseline (0.9 ETDRS letters) however the German cohort lost 0.73 letters dur-
ing the same time period with a mean of 4.3 (Subramanian et al. 2009; Martin et al. 2011;
injections in patients who completed a year of Heier et al. 2012; Schmidt-Erfurth et al. 2011;
follow-up (Holz et al. 2013; Finger et al. 2013). Holz et al. 2011; Boyer et al. 2009; Fung et al.
These results further demonstrate that VA out- 2007). Reductions in mean MT at 3 months are
comes in clinical practice are inferior to those in often not maintained to the 12-month mark
phase III clinical trials, with under-treatment of (Kodjikian et al. 2013; Holz et al. 2011; Boyer
patients in the clinical practice setting being a et al. 2009; Fung et al. 2007; Finger et al. 2013);
likely contributing factor. however, as most studies reporting 3-month fig-
ures involve a loading phase of three 4-weekly
injections, followed by PRN treatment, this may
3.2.2 Anatomical Outcomes Based reflect under-treatment post-loading phase.
on Optical Coherence Reductions in mean MT seen after 1 year are
Tomography maintained at 2 years (Martin et al. 2012;
Chakravarthy et al. 2013; Lalwani et al. 2009;
The response of macular tissue to anti-VEGF Marques et al. 2013), regardless of treatment
therapy has been examined with OCT in a num- regime (4-weekly or PRN). However, the CATT
ber of studies. Changes in macular thickness study found regular 4-weekly treatment to be
(MT) viewed with OCT have been the most com- more effective than PRN therapy for reducing
monly reported outcome (see Table 3.2) (Finger foveal thickness, and a switch from 4-weekly
et al. 2014). Extensive variation of thickness therapy to PRN after 1 year resulted in an increase
measurements exists between studies, due to dif- in MT to the level of participants that received
ferences in methods of measurement as well as PRN treatment from the commencement of ther-
different OCT machines being used. Terminology apy (Martin et al. 2012).
is similarly variable between papers, with foveal Minimal evidence is available on long-term
thickness (FT), CRT, central macular thickness changes in MT from anti-VEGF therapy for
(CMT), central subfield macular thickness nvAMD. However a clinical practice-based
(CSMT), and central foveal thickness (CFT) all 36-month follow-up of 52 patients in Portugal
used in various clinical trials (Mitchell 2011). Of found reductions in CMT to be maintained,
these, CRT is the most often reported, however despite a relatively low injection-rate throughout
variation in reported measurements exists even the study (Marques et al. 2013). The SEVEN-UP
within this subgroup. Despite this high level of study performed retinal imaging to assess the
variance, trends observed in macular thickness anatomical status of previous participants in the
changes with anti-VEGF therapy are apparent ANCHOR or MARINA trials (Brown et al. 2006;
when comparing agents and treatment regimes Rosenfeld et al. 2006), followed by the HORIZON
(see Table 3.2). study (Singer et al. 2012). As baseline measure-
ments of MT were not published in these trials,
3.2.2.1 Ranibizumab figures published in SEVEN-UP cannot shed
RBZ has been extensively demonstrated to light on long-term changes in MT with RBZ ther-
reduce the thickness of macular tissue in the apy. However, no association was found between
treatment of nvAMD (Martin et al. 2011, 2012; OCT findings (including measures of retinal
Chakravarthy et al. 2012, 2013; Kodjikian et al. thickness) in SEVEN-UP and the visual acuity
2013; Krebs et al. 2013; Heier et al. 2012; outcomes of its participants. The authors dis-
Schmidt-Erfurth et al. 2011; Holz et al. 2011; cussed the complexity of the relationship between
Boyer et al. 2009; Lalwani et al. 2009; Fung et al. anatomical outcomes within the macula and VA
2007; Marques et al. 2013; Finger et al. 2013). (Rofagha et al. 2013), implying that MT may be
This reduction in thickness is mostly due to reso- an overly simplistic outcome measure. Almost all
lution of edema and is evident in most patients of the participants followed up at this late time
followed up 4 weeks after their first injection point (7 years post entering ANCHOR or
Table 3.2 Anatomical outcomes in terms of macular thickness (MT) changes from anti-VEGF therapy reported in clinical trials and clinical practice settings. Only studies reporting outcomes
to 1 year or greater have been included
Setting (clinical MT at >2 years
practice or Treatment regime Mean MT MT at 3 MT (in μm) MT (in μm) (in μm) (Time Injections at
clinical trial) and Drug and dose (n per regime at Mean age at (in μm) at months (n at Injections at 1 year (n at Injections at 2 years (n Injections point) (n at given specified time
Study measure used (if reported) baseline) baseline baseline 3 months) at 3 months 1 year) at 1 year at 2 years) at 2 years time point) point
CATT (Martin Clinical trial RBZ 0.5 mg Monthly (n = 146) 79.5 (SD 7.4) 460 (SD 194) NR NR 266 (SD 125) 11.7 (SD 1.5) 267 (SD 143) 22.4 (SD 3.9) NR NR
et al. 2011, 2012) (n = 280) (n = 134)
FT Monthly (1 year) 78.8 (SD 7.5) 462 (SD 184) 295 (SD 135) NR (5.0 in
then PRN (n = 138) (n = 130) second year)
PRN (n = 298) 78.3 (SD 7.8) 462 (SD 195) 294 (SD 139) 6.9 (SD 3.0) 293 (SD 129) 12.6 (SD 6.6)
(n = 281) (n = 264)
BVZ 1.25 mg Monthly (n = 135) 79.7 (SD 7.5) 462 (SD 205) 300 (SD 149) 11.9 (SD 1.2) 274 (SD 137) 23.4 (SD 2.8)
(n = 261) (n = 129)
Monthly (1 year) 80.4 (SD 7.1) 471 (SD 185) 334 (SD 190) NR (5.8 in
then PRN (n = 131) (n = 122) second year)
PRN (n = 300) 78.9 (SD 7.4) 459 (SD 173) 308 (SD 127) 7.7 (SD 3.5) 306 (SD 134) 14.1 (SD 7.0)
(n = 266) (n = 251)
IVANa (Chakravarthy Clinical trial RBZ 0.5 mg Monthly (n = 157) 77.8 (SD 7.6) 468 (SD 187) NR NR 322 (SD 139) NR 322.4 (SD NR NR NR
et al. 2012, 2013) (n = 302) (n = 287) 137.3)
(n = 271)
FT 3-dose loading
phase, then PRN
(n = 155)
BVZ 1.25 mg Monthly (n = 149) 77.8 (SD 8.0) 465 (SD 184) 325 (SD 134) 331.0 (SD
(n = 279) (n = 274) 144.2)
(n = 254)
3-dose loading
phase, then PRN
(n = 145)
Analysis Monthly (agents 77.8 (SD 8.0) 474 (SD 188) 311 (SD 126) 314.7 (SD
combining agents combined) (n = 308) (n = 293) (n = 277) 137.1)
(n=261)
(n = 261)
3-dose loading 77.6 (SD 6.8) 459 (SD 182) 459 (SD 182) 338.5 (SD
phase, then PRN (n = 289) (n = 289) 143.3)
(agents combined) (n = 264)
(n = 302)
GEFAL (Kodjikian Clinical trial RBZ 0.5 mg 3-dose loading 78.68 (SD 7.27) 354.75 (SD 239 (SD NR 248 (SD 103) 6.5 (SD 2.4) NR NR NR NR
et al. 2013) phase, then PRN 109.90) NR) (n = 183)
(n = 183) (n = 183)
Data for cohort which CSMT BVZ 1.25 mg 3-dose loading 79.62 (SD 6.90) 359.21 (SD 258 (SD 264 (SD 133) 6.8 (SD 2.7)
completed study phase, then PRN 120.72) NR) (n = 187)
protocol (n = 187) (n = 187)
MANTA Clinical trial BVZ 1.25 mg 3-dose loading 76.7 (SD 7.8) 374.6 (SD NR NR 288.3 (SD 9.1 (SD 2.8) NR NR NR NR
(Krebs et al. 2013) phase, then PRN 8.4) 8.0) (n = 121)
(n = 154)
CRT RBZ 0.5 mg 3-dose loading 77.6 (SD 8.1) 365.0 (SD 275.1 (SD 8.8 (SD 2.7)
phase, then PRN 8.1) 6.8) (n = 127)
(n = 163)
VIEW 1 (Heier et al. Clinical trial RBZ 0.5 mg 4-weekly (n = 304) 78.2 (SD 7.6) 315.3 (SD NR NR 198.5 (SD 12.1–12.5 for NR NR NR NR
2012) 108.3) 109.0) all monthly
(n = 269) treatment
arms (SD
NR)
CRT ABC 0.5 mg 4-weekly (n = 301) 78.4 (SD 8.1) 313.2 (SD 197.6 (SD
106.0) 104.1)
(n = 270)
ABC 2 mg 4-weekly (n = 304) 77.7 (SD 7.9) 313.6 (SD 197.1 (SD
103.4) 98.4)
(n = 285)
3-month loading 77.9 (SD 8.4) 324.4 (SD 195.9 (SD 7.5 (SD NR)
phase, then 111.2) 108.5)
8-weekly (n = 301) (n = 265)
VIEW 2 (Heier et al. Clinical trial RBZ 0.5 mg 4-weekly (n = 291) 73.0 (SD 9.0) 325.9 (SD NR NR 187.4 (SD 12.2–12.4 for NR NR NR NR
2012) 110.9) 122.2) all monthly
(n = 269) treatment
arms (SD
NR)
CRT ABC 0.5 mg 4-weekly (n = 296) 74.7 (SD 8.6) 326.5 (SD 196.7 (SD
116.5) 114.8)
(n = 268)
ABC 2 mg 4-weekly (n = 309) 74.1 (SD 8.5) 334.6 (SD 177.8 (SD
119.8) 122.8)
(n = 274)
3-month loading 73.8 (SD 8.6) 342.6 (SD 193.4 (SD 7.5 (SD NR)
phase, then 124.0) 119.7)
8-weekly (n = 306) (n = 270)
VIEW studies Clinical trial RBZ 0.5 mg 4-weekly (n = 594) 75.6 (SD 8.7) 296 (SD 123) NR NR NR 12.1–12.5 for +10 μm from 16.5 (SD 3.7) NR NR
combined (Heier et al. all monthly 1 year
2012; Schmidt- treatment
Erfurth et al. 2014) arms (SD
NR)
CRT ABC 0.5 mg 4-weekly (n = 594) 76.5 (SD 8.5) 296 (SD 132) +10 μm from 16.2 (SD 4.0)
1 year
ABC 2 mg 4-weekly (n = 611) 75.9 (SD 8.4) 299 (SD 126) +10 μm from 16.0 (SD 3.2)
1 year
3-month loading 75.8 (SD 8.8) 306 (SD 134) 7.5 (SD NR) +6 μm from 11.2 (SD 2.9)
phase, then 1 year
8-weekly (n = 603)
(continued)
Setting (clinical MT at >2 years
practice or Treatment regime Mean MT MT at 3 MT (in μm) MT (in μm) (in μm) (Time Injections at
clinical trial) and Drug and dose (n per regime at Mean age at (in μm) at months (n at Injections at 1 year (n at Injections at 2 years (n Injections point) (n at given specified time
Study measure used (if reported) baseline) baseline baseline 3 months) at 3 months 1 year) at 1 year at 2 years) at 2 years time point) point
EXCITE Clinical trial RBZ 0.3 mg Monthly (0.3 mg) 75 (SD = 8.26) 320.6 NR NR 215.3 (SD 11.4 NR NR NR NR
(Schmidt-Erfurth (n = 95) (SD = 118.55) NR) (n NR) (SD = 1.69)
et al. 2011)
CRT 3-dose loading 75.1 (SD = 7.45) 313.6 217.6 5.7
phase, then (SD = 85.05) (SD = 0.80)
quarterly (0.3 mg)
(n = 100)
RBZ 0.5 mg 3-loading phase, 75.8 (SD = 6.96) 324.5 218.9 5.5
then quarterly (SD = 115.94) (SD = 1.05)
(0.5 mg) (n = 100)
SUSTAIN (Holz et al. Clinical trial RBZ 3-dose loading 75.1 (SD = 8.06) 340.5 (SD 239.4 (SD 2.9 (SD 249.0 (SD 5.6 (SD 2.37) NR NR NR NR
2011) CRT 0.3 mg/0.5 mg phase, then PRN 113.19) NR) 0.35) NR) (n = 509)
(n = 512) (n = 509)
SAILOR (Boyer et al. Clinical trial RBZ 0.3 mg 3-dose loading 79.9 (SD 7.9) 312 (SD 104) 205 (SD 96 % of 240 (SD NR) 4.6 NR NR NR NR
2009) (Note: only phase, then PRN NR) total cohort (n = NR) (combined
treatment naïve (n = 462) (n = NR) received 3 figure for
cohorts are included) CFT RBZ 0.5 mg 3-dose loading 75.8 (SD 8.0) 322 (SD 116) 200 (SD doses in 230 (SD NR) both groups)
phase, then PRN NR) first 3 (n = NR) (SD NR)
(n = 490) (n = NR) months (81.7 %
follow-up
reported)
PrONTO (Fung et al. Clinical trial RBZ 0.5 mg 3-dose loading 83.5 (SD 7.2) 393.9 (SD 204.3 (SD 3 (SD NR) 216.1 (SD 5.6 (SD 2.3) 179.3 (SD 9.9 (SD 5.3) NR NR
2007; Lalwani et al. CRT phase, then PRN NR) NR) NR) NR) (n = 37)
2009) (n = 40)
ABC-Trial (Tufail Clinical trial BVZ 1.25 mg 3-dose loading 79 (SD NR) 328 (IQR NR NR 239 (IQR 7.1 (Range NR NR NR NR
et al. 2010) phase (6-weekly 271–376) 127–350) 3–9)
CMT intervals) then PRN (n = 64)
(n = 65)
Sacu et al. (2009) Clinical trial BVZ 1 mg 3-dose loading 78 (SD 8) 357 (SD NR) 230 (SD NR 244 (SD NR) 6.8 (SD NR) NR NR NR NR
phase, then PRN (n = 14) NR) (n = 14) (n = 14)
CRT (n = 14)
Marques et al. Clinical Practice RBZ PRN (n = 84) 77.39 (Range 373.3 (SD 314.2 (SD 1.53 (SD 296.3 (SD 3.75 (SD 259.4 (SD 6.35 (SD 2.3) 264.3 (SD 67.7) 8.67 (SD 3.3)
(Marques et al. 2013) 61–94) 102.7) 102.7) 0.53) 68.6) (n = 84) 1.20) 67.9) (n = 77) (n = 52)
CMT (n = 67) 36 months
WAVE Germany Clinical Practice RBZ 3-month loading NR 349.4 (SE 250.8 (SE 2.95 (SE 270.5 (SE 4.34 (SE NR NR NR NR
(Finger et al. 2013) CRT phase, then PRN 4.31) 5.42) 0.01) 7.31) 0.05)
(n = 871) (n = 679) (n = 545)
NR Not reported, IQR inter-quartile range, SD standard deviation, SE standard error, FT foveal thickness, CRT central retinal thickness, CFT central foveal thickness, CMT central macular
thickness, CSMT central subfield macular thickness
a
IVAN trial reported findings according to drug allocation and regime allocation, but did not report findings for individual sub-groups
MARINA) showed some degree of macular atro- have consistently found the magnitude of MT
phy (Rofagha et al. 2013). Indeed, the interplay reduction to be greater with RBZ 0.5 mg than
between the variable and concurrent effects of BVZ 1.25 mg, when continuous and PRN
nvAMD, atrophic AMD, and intravitreal RBZ regimes are compared.
may limit the value of MT as a predictor of long-
term treatment outcome.
3.2.3 Incidence of Legal Blindness
3.2.2.2 Aflibercept
The VIEW studies compared 8-weekly ABC The impact of RBZ on the incidence of legal
2.0 mg to 4-weekly RBZ 0.5 mg over 96 weeks blindness has been modeled in the United States
(Heier et al. 2012; Schmidt-Erfurth et al. 2014). (Bressler et al. 2011) and Australia (Mitchell
In those receiving 8-weekly ABC treatment, a et al. 2014). The US model (applied to the non-
small degree of retinal thickening was regularly Hispanic white population only) combined
observed 8-week post injection (Heier et al. national population data with the incidence of
2012). However, overall comparable reductions nvAMD found in the Beaver Dam Eye Study
in CRT were recorded in all treatment groups at (Klein et al. 2007) to derive an estimate of dis-
all time points to 96 weeks. It is noteworthy that ease incidence and progression across the coun-
the reported baseline CRT values in the 96-week try (Bressler et al. 2011). The effectiveness of
follow-up paper (Schmidt-Erfurth et al. 2014) RBZ was based on treatment outcomes in the
were different to those reported in the 52-week Age-related Eye Disease Study (ARED)
paper (Heier et al. 2012) (see Table 3.2). No rea- (Bressler et al. 2003), ANCHOR, and MARINA
son for this difference was cited in the 96-week phase III clinical trials (Brown et al. 2009;
paper; however, given that this subsequent paper Rosenfeld et al. 2006). This model found that
had baseline CRT values consistently 20–30 μm regular 4-weekly treatment with RBZ over 2
thinner, it is plausible that a different anatomical years would reduce the incidence of legal blind-
landmark was used for measurement in the sec- ness (defined as BCVA ≤ 20/200 in both eyes) by
ond analysis. This discrepancy highlights the dif- 72 % (95%CI: 70–74 %) from 16,268 to 4484
ficulty of comparing MT measurements between individuals, of the 103,582 modeled to develop
different studies, or even amongst subsamples of nvAMD. The same model estimated this treat-
the same study. However, although far less evi- ment regime would reduce development of
dence exists for ABC than for RBZ, available visual impairment (BCVA ≤ 20/40 in both eyes)
data suggests they are similarly efficacious in by 37 % (95%CI: 35–39 %) from 34,702 to
reducing MT. 21,919 cases (Bressler et al. 2011). The
Australian model used the same treatment out-
3.2.2.3 Bevacizumab come data (Brown et al. 2009; Rosenfeld et al.
BVZ therapy reduces MT in nvAMD compared 2006; Bressler et al. 2003), combined with popu-
with PDT (Tufail et al. 2010). Results beyond 2 lation data from 2010 and the 10-year cumula-
years of treatment are yet to be reported. As tive incidence of AMD in the Blue Mountains
with RBZ, continuous 4-weekly treatment has a Eye Study (BMES) (Wang et al. 2007). In addi-
greater magnitude of effect than PRN treatment, tion to studying the effect of 4-weekly RBZ, this
and reductions seen after a year of continuous model estimated the effectiveness of PRN treat-
treatment were not maintained with a second ment, based on results of the CATT study (Martin
year of PRN treatment (Martin et al. 2012). et al. 2012). Using the same criteria to define
Most available evidence is from head-to-head legal blindness and visual impairment, the esti-
comparisons with RBZ 0.5 mg under highly mated effect of 4-weekly RBZ over 2 years was
standardized clinical trial conditions (Martin most similar to that of the US model, estimating
et al. 2012; Chakravarthy et al. 2013; Kodjikian a 72 % (95%CI: 70–74 %) reduced incidence of
et al. 2013; Krebs et al. 2013). These results legal blindness and a 37 % (95%CI: 34–39 %)
decrease in the development of visual impair- their target profiles and binding affinity for
ment (Mitchell et al. 2014). The modeled effec- VEGF, available data for the safety of one agent
tiveness of PRN RBZ was relatively poorer, should not be extrapolated to others (Mitchell
estimating a 68 % (95%CI: 64–71 %) reduction 2011). Some of the available data are summa-
in incidence of legal blindness and a 28 % rized in Table 3.3.
(95%CI: 23–33 %) decrease in visual impair-
ment development (Mitchell et al. 2014). 3.2.4.1 Ranibizumab
Decreases in the incidence of blindness from The safety of intravitreal RBZ is supported by
AMD have been observed in population studies extensive, robust evidence generated by large pro-
based on national registers of blind persons since spective clinical studies (Mitchell 2011; Holz
the introduction of anti-VEGF therapy. In et al. 2013). Evidence has been collected from
Denmark, a nation with a similar population inci- over 12,500 patients enrolled in RCTs for multi-
dence of AMD to Australia and the United States ple treatment indications, including nvAMD
(Buch et al. 2005; Wang et al. 2007; Klein et al. (Brown et al. 2009; Rosenfeld et al. 2006;
2007), a dramatic decrease in the incidence of Abraham et al. 2010; Boyer et al. 2009; Holz et al.
registered blindness (defined in terms of visual 2011), diabetic macular edema (Mitchell et al.
acuity as BCVA ≤ 20/200 in both eyes) from 2011) as well as macular edema following branch
AMD was observed over a decade, from 52.2 per (Brown et al. 2011) and central retinal vein occlu-
100,000 people in 2000 to 25.7 per 100,000 peo- sions (Campochiaro et al. 2011). Following the
ple in 2010 (Bloch et al. 2012). The rate of conclusion of phase III clinical trials involving
decline was greatest in the years following the RBZ, long-term safety outcomes have been
introduction of anti-VEGF therapy in 2006 recorded in extension studies. The SECURE
(Bloch et al. 2012). Another population register study followed 234 patients that had been treated
study from Israel also reported a declining inci- with RBZ for 12 months in the EXCITE and
dence of blindness from AMD (Skaat et al. 2012). SUSTAIN studies for an additional 2 years (Silva
Blindness was defined differently in terms of et al. 2013). The HORIZON study followed 600
visual acuity in this study, as BCVA < 1/60 in patients who had received RBZ for 2 years in the
both eyes; however, from 1999 to 2008 the inci- ANCHOR, MARINA, and FOCUS studies for an
dence decreased from 511 cases out of a popula- additional 2 years (Singer et al. 2012). These two
tion of 6.13 million (8.34 per 100,000) to 440 out extension studies reported similar rates of adverse
of a population of 7.31 million (6.02 per 100,000) ocular and systemic events. In SECURE, the most
(Skaat et al. 2012). Anti-VEGF agents were commonly reported adverse ocular events were
introduced into Israeli clinical practice in 2004 retinal hemorrhage (12.8 %), cataract formation
(Skaat et al. 2012). Neither population study was (11.5 %), and increased intra-ocular pressure
able to separate AMD patients into wet and dry (6.4 %) (Silva et al. 2013). In HORIZON, the
subgroups, nor did they find the same magnitude most common adverse ocular events were cataract
effect as models based on phase III clinical trials. (12.5 %) and post-dose IOP rises to ≥30 mmHg
However, these results point to a significant ben- (9.2 %) (Singer et al. 2012). The incidence of
efit of anti-VEGF therapy for prevention of legal endophthalmitis was low; 0.2 % in HORIZON
blindness in the clinical practice setting. and 0.9 % in SECURE. The most common non-
ocular events in both studies were nasopharyngi-
tis (9.5 % in HORIZON and 9.0 % in SECURE)
3.2.4 Safety and hypertension, affecting 8.7 % and 9.0 % of
the HORIZON and SECURE populations, respec-
Variable degrees of evidence exist for the safety tively (Singer et al. 2012; Silva et al. 2013). The
of the available anti-VEGF agents used for treat- incidence of arterial thromboembolic events
ment of nvAMD. Due to differences between (including hemorrhagic and ischemic cerebrovas-
agents at the molecular level (Chen et al. 1999), cular conditions, myocardial infarction as well as
Table 3.3 RCTs comparing the systemic safety of BVZ and RBZ, included in a Cochrane meta-analysis, 2014 (Moja et al. 2014)
BVZ RBZ Risk ratio for all
All serious All serious serious systemic
Total number Length of systemic Total systemic Total Risk ratio for adverse events,
Study name of participants follow-up Deaths adverse events number Deaths adverse events number death, 95%CIa 95%CI
Biswas 2011 120 18 months 0 0 60 0 0 60 Not estimable Not estimable
(Biswas et al. 2011)
BRAMD (unpublished) 327 1 year Not available 34 161 Not available 37 166 Not estimable 0.95 [0.63–1.43]
(Schauwvlieghe et al. 2014)
CATT (Martin et al. 2012) 1185 2 years 36 234 586 32 190 599 1.15 1.26 [1.08–1.47]
[0.72–1.83]
GEFAL (Kodjikian 485 1 year 2 30 246 3 24 239 0.65 1.21 [0.73–2.02]
et al. 2013) [0.11–3.84]
IVAN (Chakravarthy 610 2 years 15 80 296 15 81 314 1.06 1.05 [0.80–1.37]
et al. 2013) [0.53–2.13]
LUCAS (unpublished) 432 2 years 3 33 214 3 51 218 1.02 0.66 [0.44–0.98]
(Berg 2013) [0.21–4.99]
MANTA (Krebs et al. 2013) 317 1 year 3 18 154 2 15 163 1.59 1.27 [0.66–2.43]
[0.27–9.37]
Subramanian et al. (2010) 28 1 year 2 2 20 0 0 8 2.14 2.14
[0.11–40.30] [0.11–40.30]
VIBERA (unpublished) 161 1 year 1 22 107 1 6 54 0.50 1.85 [0.80–4.29]
(NCT00559715 2007) [0.03–7.91]
Total 3665 62 453 1844 56 404 1821 1.10 1.08 [0.90–1.31]
[0.78–1.57]
Data reproduced from Moja L, Lucenteforte E, Kwag KH, Bertele V, Campomori A, Chakravarthy U, D’Amico R, Dickersin K, Kodjikian L, Lindsley K, Loke Y, Maguire M, Martin DF,
Mugelli A, Muhlbauer B, Puntmann I, Reeves B, Rogers C, Schmucker C, Subramanian ML, Virgili G (2014) Systemic safety of bevacizumab versus ranibizumab for neovascular age-related
macular degeneration. The Cochrane Database of Systematic Reviews 9:Cd011230. doi:10.1002/14651858.CD011230.pub2, with permission of Wiley
a
Risk ratio above one favors events in the BVZ group, risk ratio below one favors events in the RBZ group
other embolic and thrombotic events) was 5.6 % in rare but serious intraocular or systemic compli-
in SECURE and 5.3 % in HORIZON (Singer cations (Heier et al. 2012). Nevertheless, a reduc-
et al. 2012; Silva et al. 2013). In the PIER study, a tion in the number of injections required with
similar incidence of death, arterial thromboem- regular 8-weekly treatment with ABC compared
bolic events, and other adverse events was found with monthly RBZ is likely to cause a substantial
in the RBZ 0.5 mg and sham injection groups reduction in cumulative risk of adverse events
(Abraham et al. 2010). related to intraocular injections. Whether ABC
Safety data for RBZ has also been collected in can be extended further using an Inject & Extend
the clinical practice setting. The LUMINOUS regime compared to RBZ or BVZ in clinical rou-
program combined data from treatment registries tine practice is yet to be established. If not, side
in Germany, Sweden, Belgium, and the effects related to the intraocular injection proce-
Netherlands to study the “real-world safety” of 1 dure are likely to be similar between all agents.
year of treatment with RBZ (Holz et al. 2013).
Adverse events of interest to the reviewers were 3.2.4.3 Bevacizumab
those thoughts that pertain to the mechanism of Sufficient safety data for the intravitreal use of
action of RBZ and the intravitreal injection pro- BVZ for nvAMD has been much slower to
cedure. Of the 4444 patients included in the anal- emerge compared to RBZ. This is due to the fact
ysis, 73 (1.64 %) experienced ocular that BVZ was initially developed and approved
complications of particular interest, the most by drug regulatory authorities for treatment of
common of these being retinal pigment epithelial cancers via systemic intravenous administration
tear (27, 0.61 %), intraocular pressure-related (Moja et al. 2014), whereas RBZ was bespoke
events (12, 0.27 %), and traumatic cataract (10, designed for intravitreal administration and has
0.23 %). 55 (1.24 %) suffered non-ocular adverse been approved for use accordingly (Moja et al.
events during data collection, the most common 2014). Publication of the CATT study provided
of which comprised stroke (19, 0.43 %), hyper- important data on the safety of BVZ compared to
sensitivity (8, 0.18 %; the specifics of such hyper- RBZ (Martin et al. 2012). Prior to this, only one
sensitivity were not specified), and hypertension study (the ABC-trial) had provided high-level
(7, 0.16 %). The reviewers concluded that RBZ evidence for use of BVZ in nvAMD (Tufail et al.
was associated with a low rate of adverse events 2010), with its relatively small size (n = 131) lim-
(Holz et al. 2013). However, potential under- iting its findings with regard to safety. CATT
treatment of patients, particularly in the German study findings raised concerns about the systemic
WAVE study data (Finger et al. 2013), may safety of BVZ, with a significantly increased
explain the low number of adverse events (Holz incidence of participants in this group developing
et al. 2013). Similarly, treatment registers tend to one or more serious systemic adverse events
not follow patients up who are lost to follow-up; (adjusted risk ratio 1.30; 95%CI: 1.07–1.57)
thus, information on particular severe adverse (Martin et al. 2012). Both agents have been
events which will stop patients attending treating shown to reach the systemic circulation post
ophthalmologists’ practices are likely to be con- intravitreal injections; however, BVZ has a longer
siderably under-reported. systemic half-life, due to its larger molecular
structure (Avery et al. 2014). This has been dis-
3.2.4.2 Aflibercept cussed as a possible mechanism for potential dif-
The VIEW studies have compared the safety of ferences in the risk profiles of the two agents,
ABC 2 mg with RBZ 0.5 mg over 96 weeks should a difference exist.
(Schmidt-Erfurth et al. 2014). No significant dif- In the subsequent 2 years following publica-
ference in rates of death, adverse ocular, thrombo- tion of the CATT study findings, a number of
embolic, or other systemic events were found other RCTs were conducted comparing BVZ
(Schmidt-Erfurth et al. 2014), although the trial with RBZ whilst recording adverse events.
was insufficiently powered to identify differences In September 2014, the Cochrane Collaboration
published a review directly comparing the sys- lished data from three RCTs will provide addi-
temic safety of the two agents (Moja et al. 2014). tional useful evidence on the subject. The authors
Results were obtained from a random-effects also proposed that the safety of both drugs might
meta-analysis involving nine nonindustry spon- vary between patients and recommended an indi-
sored RCTs (see Table 3.3), which included 3665 vidual patient data meta-analysis to explore the
participants, who were followed up for up to 2 impact of predisposing risk factors and the effect
years after commencing treatment. All studies of different treatment regimes.
directly compared the two agents; however, vari-
able treatment regimes were included. Three of
the nine trials were unpublished at this time, 3.3 Predictors of Outcomes
comprising 482 individuals who received BVZ
and 438 that were treated with RBZ, which lim- The response to anti-VEGF therapy in nvAMD
ited assessment of the quality of evidence in this varies greatly amongst patients. Whilst patients
subgroup. However, for the two primary out- treated monthly for 2 years in phase III clinical
comes of all-cause deaths and all serious sys- trials showed an average gain of 7–15 ETDRS let-
temic adverse events, the review did not find a ters, a subgroup 10–15 % continued to lose vision
significant difference between the two agents on despite treatment. In contrast, a second subgroup
either outcome (Moja et al. 2014). A number of of approximately 30 % shows significant visual
secondary outcomes were also assessed, includ- improvement (Brown et al. 2009; Rosenfeld et al.
ing myocardial infarction, stroke, arteriothrom- 2006, 2011; Martin et al. 2012). Most studies
botic events, serious hemorrhage, serious have used a change in vision threshold of 15
neutropenia, gastrointestinal perforation, serious ETDRS letters (Rosenfeld et al. 2011). In the clin-
infection, treatment related drug discontinuation, ical practice setting, where treatment tends to
serious systemic adverse events classified as per involve fewer injections, the rate of poor response
the Medical Dictionary for Regulatory Activities to treatment has been reported to be as high as
System Organ Classes (MedDRA SOC) (ICH 20–30 % (Rasmussen and Sander 2014). To date,
2014), and serious adverse events previously the wide variance in treatment outcomes is not
associated with drugs affecting the VEGF path- well understood; however, a number of genetic
way (Moja et al. 2014). Variable levels of evi- and clinical factors have been suggested as poten-
dence existed for various secondary outcomes, tial predictors for treatment outcomes, with vari-
however only one positive association was found able levels of proposed effect (Finger et al. 2014).
across these. A higher incidence of MedDRA Whilst evidence for single nucleotide polymor-
classified gastrointestinal disorders was identi- phisms (SNPs) in the complement factor H (CFH)
fied in study participants who had received BVZ, and VEGF-A genes has been found to be some-
at a rate of 2.9 % compared to 1.6 % in the RBZ what associated with treatment outcomes, clinical
cohorts (risk ratio 1.82, 95%CI: 1.04–3.19) factors recorded at baseline, including age, visual
(Moja et al. 2014). Gastrointestinal disorders acuity, choroidal neovascularization (CNV) lesion
included abdominal pain, vomiting, dyspepsia, size, and a delay in treatment of more than 3
duodenal ulcer, pancreatitis, intestinal obstruc- weeks from symptom onset appear to be of greater
tion, intestinal perforation, faecaloma, colitis, importance in determining anti-VEGF treatment
and Crohn’s disease (ICH 2014). In summary, outcomes (Finger et al. 2014).
this review drew on nine nonindustry sponsored
head-to-head RCTs and was comprehensive in its
acquisition and analysis of available data. Its 3.3.1 Genetic Factors
authors concluded that systemic safety data does
not provide significant evidence to support pref- A number of genes have been linked to an elevated
erential use of either RBZ or BVZ for nvAMD risk of developing AMD, including CFH, age-
(Moja et al. 2014). The release of yet to be pub- related maculopathy susceptibility 2 (ARMS2)/
high-temperature requirement A-1 (HTRA1), initiation of treatment (Menghini et al. 2012).

complement factor 3 (C3), complement factor B However, in complete contrast, a different paper
(CFB)/complement factor 2 (C2), and apolipopro- previously found the CT genotype of this SNP to
tein E (APOE) genes (Chamberlain et al. 2006). be linked to the worst outcomes (Imai et al. 2010).
However, the role these genetic polymorphisms A greater number of injections in the first 12
may play in the response of nvAMD to anti- months of therapy were found for the AG geno-
VEGF therapy remains to be well understood, types of SNP rs194918455 when compared to the
with inconsistencies in the nature and extent of AA genotype (Francis 2011). Whilst evidence has
their associations in the literature published to accrued over time for an effect of the CFH gene on
date (Finger et al. 2014). The most evidence cur- nvAMD treatment outcomes, the picture painted
rently available is for the CFH gene, where almost by research to-date is far from clear. A number of
half of all SNPs assessed have been found to be factors affect visual acuity measured by com-
associated with visual outcomes and the number monly used ETDRS or Snellen distance visual
of injections required. Approximately, 15 % of acuity letter charts, and thus determining the effect
SNPs assessed in the VEGF gene have also been of genetic factors on visual anti-VEGF treatment
linked to treatment outcomes (Finger et al. 2014). outcomes is per se very challenging.
3.3.1.1 Complement Factor H 3.3.1.2 Age-Related Maculopathy

(CFH) Gene Susceptibility 2 (ARMS2)/
The CFH gene complex has been investigated High-Temperature
with mixed results. Several studies have identified Requirement A-1
associations pertaining to VA (Imai et al. 2010; (HTRA1) Genes
Tian et al. 2012; McKibbin et al. 2012; Nischler Papers examining ARMS2/HTRA1 for associa-
et al. 2011; Brantley et al. 2007; Lee et al. 2009; tions with treatment outcomes have reported
Francis 2011; Kloeckener-Gruissem et al. 2011), variable results. Homozygous carriers of the SNP
for as long as 12 months after commencing treat- rs10490924 (A69 S) TT genotype of the ARMS2
ment (Francis 2011; Kloeckener-Gruissem et al. gene have an increased risk of not improving or
2011). Others, however, did not find any associa- continuing to lose vision (Teper et al. 2010). The
tion with treatment (Inglehearn et al. 2012; Orlin presence of the TT genotype (independent of
et al. 2012; Teper et al. 2010; Wickremasinghe homozygosity) has also been associated with a
et al. 2011), including the largest study to date lack of treatment response (Kitchens et al. 2013;
(n = 834) which was based on the CATT study Tian et al. 2012). Congruently, the CC genotype
cohort (Hagstrom et al. 2013). This study also is associated with a better response to treatment
returned negative findings for associations with (Abedi et al. 2013a). A higher risk of AMD
lesion size, leakage seen on fluorescein angiogra- development has been shown, where the T-allele
phy (FA), OCT findings of mean foveal thickness at SNP rs10490924 is present (Rivera et al. 2005).
changes and the presence of fluid and also the In the HTRA1 gene, homozygous carriers of the
mean of number of injections required in the first AA genotype at SNP rs11200638 are at an
12 months of treatment (Hagstrom et al. 2013). increased risk of developing AMD and (confus-
Where associations have been identified with CFH ingly) have been associated with improved
and treatment outcomes, the CC risk genotype of (Abedi et al. 2013a) and worsened (Tian et al.
SNP rs1061170 (Y402 H) has been linked to 2012) vision outcomes. Despite the presence of
poorer treatment outcomes including reduced VA several papers that identify associations between
improvement (Brantley et al. 2007; Kloeckener- ARMS2/HTRA1 and treatment outcomes, a
Gruissem et al. 2011) and a greater number of number of negative studies have also been pub-
injections required (Lee et al. 2009). Congruently, lished (Hagstrom et al. 2013; Inglehearn et al.
the CT genotype has been identified as predictive 2012; Kloeckener-Gruissem et al. 2011, 2012;
of favorable visual outcomes up to 24 months after Yamashiro et al. 2012).
3.3.1.3 Vascular Endothelial Growth 3.3.1.5 Complement System-Related

Factor (VEGF) Gene Genes
The VEGF gene has been examined at length for Genes comprising the complement system,
an effect on treatment outcomes with anti-VEGF including complement factors two (C2), three
therapy for nvAMD. Approximately, 15 % of the (C3), five (C5), B (CFB), and complement factor
SNPs studied within this gene have been linked H receptor (CFHR) genes, have been examined
to treatment outcomes (Finger et al. 2014). A for associations with treatment outcomes. One
small proportion of studies have identified asso- study linked C3 to reduced CRT on OCT, as well
ciations of VEGF-A SNPs rs699946 and as reduced leakage on FA (Francis 2011).
rs699947 with VA outcomes (Imai et al. 2010; However, the bulk of published work did not find
Nakata et al. 2011; Park et al. 2014) whilst links between these genes and treatment out-
another found the SNP rs3025000 to be associ- comes (Abedi et al. 2013a; Hagstrom et al. 2013;
ated with improved visual outcome 6 months Kloeckener-Gruissem et al. 2011; Tian et al.
after initiation of treatment (Abedi et al. 2013b). 2012; Kloeckener-Gruissem et al. 2012).
In addition, the VEGF-A gene has been linked to
fewer injections over a 12-month period (Francis 3.3.1.6 Other Genes
2011). The SNP rs833069 of VEGF-A has been The influence of other genes on nvAMD treat-
associated with reductions in CSMT over 6 ment outcomes with anti-VEGF agents has been
months of treatment (Chang et al. 2013), whilst assessed in variable degrees of detail. One study
SNP rs943080 has also been associated with has suggested that various polymorphisms within
anatomical outcomes over 12 months (Zhao the C-reactive protein gene might be associated
et al. 2013). Positive associations have been with poorer vision outcomes after treatment
identified in studies that comprised different eth- (Brantley et al. 2007). An interesting association
nic groups, further complicating analysis of was reported for patients heterozygous for two
these results (Finger et al. 2014). The majority of SNPs; rs1061170 within the CFH gene and
research into the effect of the VEGF gene has rs10896563 in the frizzled family receptor 4
not found an association with treatment out- gene. These patients had an improved response to
comes (Inglehearn et al. 2012; Kitchens et al. treatment after 12 months (Kloeckener-Gruissem
2013; Kloeckener-Gruissem et al. 2011; et al. 2011); however, the same association was
McKibbin et al. 2012; Tian et al. 2012; Wang not found at 24 months (Kloeckener-Gruissem
et al. 2012), or only found an association that did et al. 2012). A study focused primarily on the
not remain when factors such as age and baseline interleukin (IL) 23 R gene identified an associa-
BCVA had been controlled for (Boltz et al. tion between the Group 12 secretory phospholi-
2012). pase A2 (PLA2G12A) gene; however, after
adjustment for multiple testing by the false dis-
3.3.1.4 Apolipoprotein E (APOE) covery rate had been made the association was no
Gene longer present (Wang et al. 2012). The study did
Available evidence suggests that variants of the not find any association between the 11 tested
APOE gene can affect both the risk of develop- SNPs of the IL-23 R gene and treatment out-
ment of AMD, as well the response to anti-VEGF comes (Wang et al. 2012). Several genes known
treatment. The ε2 variant appears to increase the to increase patient risk of developing AMD have
risk of AMD development compared to ε3 and ε4 not been associated with treatment outcomes
alleles (Wickremasinghe et al. 2011). from anti-VEGF agents, including complement
Furthermore, a comparison between the ε2 and factor I gene, the tissue inhibitor of metallopro-
ε4 polymorphisms found ε4 to be associated with teinase 3 gene as well as genes involved in cho-
superior treatment outcomes (Wickremasinghe lesterol metabolism in addition to APOE (Finger
et al. 2011). et al. 2014).
3.3.1.7 Genome Wide Association (Boyer et al. 2007; Kaiser et al. 2007; Rosenfeld
Studies et al. 2011; Ying et al. 2013). Much of this work
A genome-wide association study comprising 65 has focused on functional outcomes, identifying
patients identified a number of genetic associa- predictive factors for improvements in BCVA
tions relating to various treatment outcomes. (Finger et al. 2014). However, other outcome
Within the CFH gene, the AA genotype in the measures such as anatomical changes, adverse
SNP rs1065489 was associated with less events, and the number of injections required by
improvement, whilst comparison of the AG and patients have also been studied.
AA genotypes of the SNP rs3766404 found the
former to be associated with a higher number of 3.3.2.1 Vision Outcomes
required injections over the first 12 months of Factors at presentation that predispose individu-
treatment (Francis 2011). Two genotypes within als to poorer vision outcomes include reduced
the VEGF-A gene also impacted the number of VA, increasing age and a larger CNV lesion,
required injections. The AG genotypes within whilst a higher baseline VA places at increased
SNPs rs833068 and rs833069 both predicted the risk of losing vision (Boyer et al. 2007; Kaiser
need for more intensive treatment than the AA et al. 2007; Rosenfeld et al. 2011; Ying et al.
and GG genotypes, respectively (Francis 2011). 2013). These findings are supported by studies
With regard to the C3 gene, the SNP rs6660704 performed in the clinical practice setting (Kang
contained a genotype (AA) that was associated and Roh 2009; Lim et al. 2012a; Wickremasinghe
with reduced CRT on OCT examination after 1 et al. 2011; Yamashiro et al. 2012).
year of anti-VEGF treatment (Francis 2011). A past history of other treatments for AMD
This work yielded the discovery of the fms- has been implicated in predicting future treat-
related tyrosine kinase 1 (FLT1) gene’s involvement outcomes with anti-VEGF therapy. Previous
ment in treatment outcomes. It was associated intravitreal triamcinolone or PDT has been
with the persistence of leakage on FA as well as reported to limit achievable visual outcomes
the number of required injections over 1 year (Jyothi et al. 2010; Levy et al. 2009; Lux et al.
(Francis 2011). 2007). An elevated intraocular pressure has also
In 2012 a large pharmacogenetic study exam- been associated with reduced visual gains (Ying
ined a number of risk alleles within the CFH, et al. 2013).
ARMS2/HTRA1, and VEGF-A genes and their Different types of CNV lesion have been
cumulative impact on treatment outcomes with linked to vision outcomes with a variety of con-
RBZ (Smailhodzic et al. 2012). This work found flicting results. Classical CNV lesions (both
that patients with higher numbers of risk alleles minimally and predominantly classic) have been
required treatment earlier; with a mean increase linked with reduced visual improvement (Ying
of up to 10 years earlier when those with four et al. 2013) and a greater number of recurrences
high-risk alleles in the VEGF gene were requiring further injections (Horster et al. 2011).
compared to those with zero (Smailhodzic et al. Other studies have not replicated these findings
2012). In addition, an increased number of high- (Lalwani et al. 2009; Lux et al. 2007) whilst
risk alleles corresponded to poorer treatment out- some research have found minimally and
comes (Smailhodzic et al. 2012). predominantly classical CNVs to gain more
vision (Heimes et al. 2011; Jyothi et al. 2010).
Eyes with retinal angiomatous proliferations
3.3.2 Clinical Factors (RAP) made favorable vision gains compared
with other lesions in the CATT study cohort
Clinical factors predictive of treatment outcomes (Ying et al. 2013).
have been searched for and examined within the Various OCT findings have been linked to
cohorts of several large phase III clinical trials vision outcomes. Atrophic changes in the outer
retina at baseline are predictive of poor VA out- 3.3.2.2 Anatomical Outcomes

comes after 12 months of treatment (Ristau et al. The type of macular edema viewed on OCT has
2014). An intact external limiting membrane and been studied for its predictive value. Cystoid
ellipsoid zone have been independently associ- edema (intraretinal cystic spaces visible on OCT)
ated with good prognosis for VA after three con- has been linked to the largest reductions in CRT,
secutive injections (Kwon et al. 2014) to as long when compared with a PED, subretinal fluid, or
as 12 months after commencement of treatment other types of macular edema (Guber et al. 2014).
(Mathew et al. 2013). The anatomical integrity of Other work, however, reported cystoid macular
the retina viewed on OCT, looking at features edema to have a higher risk of nonresponse after
including the continuity of the external limiting 12 months of treatment and follow-up (Byun
membrane and inner/outer segment band were et al. 2010). CRT reductions have been found to
linked to superior visual acuity outcomes follow- occur during the initial six injections, but not
ing more than 6 months of anti-VEGF therapy beyond this volume of treatment (Guber et al.
(Oishi et al. 2013). Despite this, the same study 2014). Age and baseline BCVA have not been
identified baseline VA to be of stronger predictive shown to be predictive for CRT response to treat-
value for long-term vision outcomes (Oishi et al. ment (Guber et al. 2014).
2013). Although such results are encouraging Utilization of indocyanine green angiography
that OCT characteristics may hold predictive has found persistent lesion activity to be associ-
value for vision outcomes, other research has ated with the presence of an arteriolarized vascu-
been unable to link OCT baseline features to lar complex or a polypoidal choroidal
12-month visual outcomes (Kolb et al. 2012). vasculopathy (PCV) in serous PEDs (Mettu et al.
Features of retinal pigment epithelium (RPE) 2012). The capillary subtype in classic mem-
dysfunction have been linked to vision loss after branes has indicated a greater chance of resolu-
2 years of treatment. A pooled analysis of indi- tion of activity after 3–5 monthly injections (Lad
viduals with a gain or loss of ≥15 EDTRS let- et al. 2012).
ters in the ANCHOR and MARINA trials found
a greater number of signs including pigmentary 3.3.2.3 Adverse Events
abnormalities and atrophic scars (Rosenfeld Minimal research has been published on base-
et al. 2011). Conversely, two smaller studies line factors that might predict adverse events
found RPE abnormalities viewed on fundus (Finger et al. 2014). One study found that indi-
autofluorescence to be the only factors linked to viduals with a higher baseline BCVA and the
better treatment outcomes (Heimes et al. 2008, presence of a CNV lesion, instead of an RAP
2011). However, the strength of these findings lesion, predicted a higher risk of developing an
was diminished by not controlling for confound- RPE tear (Introini et al. 2012). However, possi-
ers in multivariate testing (Finger et al. 2014). ble confounders were not adjusted for in this
The presence of a pigment epithelial detach- work (Finger et al. 2014). There is a paucity of
ment (PED) seen at baseline might suggest a evidence for predictors of adverse outcomes
higher probability of secondary vision loss after such as retinal atrophy or hemorrhage (Finger
three initial monthly injections of RBZ (Mariani et al. 2014).
et al. 2011).
The interval between symptom onset and 3.3.2.4 Predictors of Required
treatment has been identified as an important pre- Injections
dictive factor for treatment outcomes. A delay of Both predominantly and minimally classic CNV
over 21 days between first symptoms and admin- lesions have been linked to a greater number of
istration of treatment was shown to predict an required injections (Horster et al. 2011). Baseline
increased risk for vision loss or lack of improve- subretinal fluid on OCT has been associated
ment after 12 months (Lim et al. 2012a). with a greater likelihood of ongoing treatment
requirement over 20 months (Tannan et al. cal features that are chiefly involved are age, VA,
2010); however, other research has not identified the size of the CNV lesion, and the delay to ini-
any link between OCT characteristics and the tiation of treatment. A higher risk of vision loss
number of required injections in the first 12 or reduced vision gains has been associated with
months of therapy (Jeng and Sadda 2010). increased age, a better baseline VA, a larger CNV
Studies commonly report the number of injec- at presentation, and delay of greater than 21 days
tions administered to participants, but seldom from symptom onset to commencement of treat-
report the intervals between injections in PRN ment. Patients with better baseline VA are more
treatment regimes. Consequently, predictive fac- likely to have good vision outcomes; however,
tors to assess the possibility of extending an indi- they are less likely to gain vision than their coun-
vidual’s retreatment interval have not been terparts with relatively poorer vision at baseline.
assessed (Finger et al. 2014). This is the result of a ceiling effect (Finger et al.
2014). Available evidence pertaining to baseline
CNV characteristics observed using FA and OCT
3.3.3 Behavioral Factors remains unclear, with variable evidence for and
against predictive usefulness for VA outcomes or
Little research has been published on the effect of the frequency of injections required. Evidence
behavioral and lifestyle risk factors on the treat- for an effect from smoking is equally conflicted,
ment outcomes with anti-VEGF therapy (Finger with different studies pointing to both better and
et al. 2014). Conflicting evidence exists for the worse treatment outcomes in smokers, whilst
impact of smoking. A Korean study of 125 eyes another paper reported no association (Finger
found current smoking to be independently asso- et al. 2014).
ciated with poor visual acuity outcomes after 3 A large number of studies have assessed a
months of treatment (Lee et al. 2013). Another range of SNPs and their association with treat-
study found that an increasing number of pack ment outcomes, with the highest yield of evidence
years correlated with less gain in vision after 2 being for the CFH and VEGF genes. The out-
years of follow-up in both current and former comes that these have most often been associated
smokers (Menger et al. 2012). Other research with are VA or the number of injections required
found the opposite, where current and former by an individual during follow-up. Assessment of
smokers gained vision (Inglehearn et al. 2012), genetic factors has varied between studies in
whilst no relationship either way has also been terms of the length of follow-up, time points, and
reported (Naj et al. 2013). Individuals receiving racial characteristics of samples (Finger et al.
antihypertensive therapy gained less vision in 2014). Despite the large number of papers on this
one study (Menger et al. 2012). The use of topic, most of these are uncontrolled, prospective
anticoagulants may predispose those with or retrospective case series. Data is available from
nvAMD to intraocular hemorrhage more so than only one large randomized-controlled trial (the
patient age and disease duration (Kiernan et al. CATT study) for assessment of pharmacogenetic
2010); however, anticoagulant use has not been associations in treatment of nvAMD with anti-
assessed as a potential predictive factor in larger VEGF drugs (Hagstrom et al. 2013). The pres-
studies (Finger et al. 2014). ence of an increasing number of known risk
alleles in higher risk genes (namely, CFH,
ARMS2/HTRA1, and polymorphisms in the
3.3.4 Summary of Evidence VEGF gene) has been associated with a younger
age of nvAMD onset and poorer treatment out-
To date, clinical factors have been implicated to a comes (Smailhodzic et al. 2012). Overall, genetic
much greater degree in predicting treatment out- factors appear to have an association with treat-
comes than genetic factors. Those baseline clini- ment outcomes that is weak at best (Gorin 2012).
3.4 Need for Further Research Informing patients of their individual prognosis
and allowing treatment to be planned in the most
Despite the unrivalled dominance of anti-VEGF cost-effective way are certainly valuable; how-
as the most effective available treatment option ever, predictors of treatment outcome can also
for nvAMD, a number of questions regarding its guide selection of the most appropriate therapy
use remain to be answered. Given the absence of and treatment protocol (Finger et al. 2014).
a comparably effective alternative treatment, Whilst a comparative alternative therapy does
studies must remain uncontrolled for now, as it not exist (Gorin 2012), the relevance of this point
is not ethical to leave nvAMD untreated. This is diminished. However, whilst currently avail-
creates challenges when attempting to deter- able anti-VEGF options produce comparable
mine useful prognostic factors of treatment out- outcomes, subgroups may respond better to dif-
comes. A greater uniformity of study design ferent treatment selections. One US study
between projects would allow results to be syn- explored the effect of switching patients refrac-
thesized, which may provide an improved tory to or relapsing with BVZ or RBZ (with a
understanding of the predictive value of base- mean of 20.4 previous RBZ/BVZ injections) to
line factors (Finger et al. 2014). Studies pro- ABC in 94 patients (Yonekawa et al. 2013).
duced to date often fail to separate AMD into its After a mean follow-up of 18 weeks and 3.4
various subtypes. Separation of subtypes (i.e., ABC injections, vision had stabilized, and CMT
PCV and RAP) will produce data of greater had improved significantly. This study has sev-
comparative value with other work (Finger et al. eral limitations, including the pooling of eyes
2014). Improved consistency of treatment that had relapsed whilst not on treatment with
regimes and follow-up time points between refractory cases and also the grouping together
studies would improve the comparability of of patients treated with RBZ and BVZ. However,
studies further (Finger et al. 2014). it raises the possibility that some subgroups may
Variable amounts of data exist for the different be better suited to particular anti-VEGF agents.
anti-VEGF agents available. There is robust evi- An improved understanding of predictive factors
dence from multiple randomized clinical trials (genetic, clinical, behavioral, or otherwise)
for the efficacy and safety of RBZ (Mitchell might aid in selection of the best appropriate
2011). Extensive data has been produced com- agent selection and treatment protocol for a
paring BVZ with RBZ, with further trials yet to given individual.
be published (Moja et al. 2014), which has found In the current climate of nvAMD manage-
the two agents to be comparable in terms of effi- ment, clinicians can control several factors,
cacy and systemic safety. Similarly, ABC has including the frequency of injections and fol-
been compared to RBZ and found to be low-up interval. Furthermore, the standard rec-
non-inferior (Heier et al. 2012; Schmidt-Erfurth ommended dose of RBZ (0.5 mg) may be
et al. 2014). Despite the number of clinical trials increased to 2 mg in poor responders (Brown
produced on treatment outcomes, only one of et al. 2013). Whilst this does not appear to have
these (the CATT study) produced data relating benefit in treatment naive patients (Busbee et al.
outcomes to AMD susceptibility genes (Hagstrom 2013), one small study found improved func-
et al. 2013). As new anti-VEGF treatment options tional and anatomical outcomes in cases refrac-
become available (through development of new tory to ongoing conventional 0.5 mg RBZ with
agents and/or drug combinations) pharmacoge- an increased RBZ dose (Brown et al. 2013).
netics may aid in identifying certain subgroups Such treatment decisions might be guided by
likely to respond better to particular drugs (Finger further knowledge of predictive factors for treat-
et al. 2014). ment outcomes.
Knowledge of reliable predictors of treatment A current discussion as to whether long-term
outcomes serves patient care in numerous ways. anti-VEGF treatment leads to chorioretinal
atrophy and/or the development of GA in a

dose–response manner, i.e., the probability of Compliance with Ethical Requirements
atrophy increases with the number of injections, Conflict of Interest: No authors have any
will need to be answered with more longer term conflicts of interests related to this book
follow-up data of large cohorts of patients with chapter.
detailed imaging, in particular OCTs. Most
studies today cannot control for small areas of Informed Consent: No human or animal
atrophy present at the initiation of anti-VEGF studies were carried out by the authors for
treatment, or preclinical areas of atrophy, with this book chapter.
only loss of certain bands on OCT, which needs
to be incorporated in future studies to accurately
assess the effect of anti-VEGF on the develop-
ment of atrophy. Furthermore, clinicians need to
be more specific and potentially develop new References
language to assess changes on OCT, as not every
small “black” spot (or “white” spot, depending Abedi F, Wickremasinghe S, Richardson AJ, Islam AF,
Guymer RH, Baird PN. Genetic influences on the out-
on standard image settings) indicates fluid in the
come of anti-vascular endothelial growth factor treat-
sense of CNV activity. Some may be chronic ment in neovascular age-related macular degeneration.
cystic changes, others preceding atrophy, and Ophthalmology. 2013a;120(8):1641–8. doi:10.1016/j.
yet others a “looseness” in the tissue following ophtha.2013.01.014.
Abedi F, Wickremasinghe S, Richardson AJ, Makalic E,
retinal cell death. All this will aid in more accu-
Schmidt DF, Sandhu SS, Baird PN, Guymer RH. Variants
rately determining not only the need for ongo- in the VEGFA gene and treatment outcome after anti-
ing anti-VEGF, but also the risk of atrophy VEGF treatment for neovascular age-related macular
development associated with long-term anti- degeneration. Ophthalmology. 2013b;120(1):115–21.
doi:10.1016/j.ophtha.2012.10.006.
VEGF treatment.
Abedi F, Wickremasinghe S, Islam AF, Inglis KM,
Guymer RH. Anti-VEGF treatment in neovascular
age-related macular degeneration: a treat-and-extend
3.5 Conclusions protocol over 2 years. Retina. 2014;34(8):1531–8.
doi:10.1097/iae.0000000000000134.
Abraham P, Yue H, Wilson L. Randomized, double-
Anti-VEGF treatment maintains vision in most masked, sham-controlled trial of ranibizumab for neo-
and improves vision in a considerable proportion vascular age-related macular degeneration: PIER
of patients with nvAMD. Several clinical factors study year 2. Am J Ophthalmol. 2010;150(3):315–24.
doi:10.1016/j.ajo.2010.04.011. e311S0002-
can predict anti-VEGF treatment outcomes.
9394(10)00267-9 [pii].
Those who present at a younger age, with better Avery RL, Castellarin AA, Steinle NC, Dhoot DS,
baseline vision and smaller CNV lesions are Pieramici DJ, See R, Couvillion S, Nasir MA, Rabena
more likely to have a better final vision outcome, MD, Le K, Maia M, Visich JE. Systemic pharmacoki-
netics following intravitreal injections of ranibizumab,
although individuals with a better visual acuity at
bevacizumab or aflibercept in patients with neovascu-
baseline stand to gain fewer letters due to a ceil- lar AMD. Br J Ophthalmol. 2014. doi:10.1136/
ing effect. Genetic factors implicated in the bjophthalmol-2014-305252.
development of nvAMD, predominantly related Berg K. Lucentis Compared to Avastin Study (LUCAS).
AAO annual meeting, 2013 subspecialty day. 2013.
to CFH and VEGF-A gene polymorphisms, may
Biswas P, Sengupta S, Choudhary R, Home S, Paul A, Sinha
also be associated with anti-VEGF treatment out- S. Comparative role of intravitreal ranibizumab versus
comes. However, the overall effect of genetic fac- bevacizumab in choroidal neovascular membrane in
tors implicated to date appears to be small. age-related macular degeneration. Indian J Ophthalmol.
2011;59(3):191–6. doi:10.4103/0301-4738.81023.
Clinical factors at presentation as well as inten-
IndianJOphthalmol_2011_59_3_191_81023 [pii].
sity of treatment largely determine current anti- Bloch SB, Larsen M, Munch IC. Incidence of legal blind-
VEGF treatment outcomes. ness from age-related macular degeneration in Denmark:
year 2000 to 2010. Am J Ophthalmol. 2012;153(2):209– Brown DM, Campochiaro PA, Bhisitkul RB, Ho AC, Gray S,
13. doi:10.1016/j.ajo.2011.10.016. e202. Saroj N, Adamis AP, Rubio RG, Murahashi WY. Sustained
Boltz A, Ruiss M, Jonas JB, Tao Y, Rensch F, Weger M, benefits from ranibizumab for macular edema following
Garhofer G, Frantal S, El-Shabrawi Y, Schmetterer branch retinal vein occlusion: 12-month outcomes of a
L. Role of vascular endothelial growth factor polymor- phase III study. Ophthalmology. 2011;118(8):1594–602.
phisms in the treatment success in patients with wet doi:10.1016/j.ophtha.2011.02.022.
age-related macular degeneration. Ophthalmology. Brown DM, Chen E, Mariani A, Major Jr JC. Super-dose
2012;119(8):1615–20. doi:10.1016/j. anti-VEGF (SAVE) trial: 2.0 mg intravitreal ranibi-
ophtha.2012.02.001. zumab for recalcitrant neovascular macular degeneration-
Boyer DS, Antoszyk AN, Awh CC, Bhisitkul RB, Shapiro primary end point. Ophthalmology. 2013;120(2):349–54.
H, Acharya NR. Subgroup analysis of the MARINA doi:10.1016/j.ophtha.2012.08.008.
study of ranibizumab in neovascular age-related macu- Buch H, Nielsen NV, Vinding T, Jensen GB, Prause JU, la
lar degeneration. Ophthalmology. 2007;114(2):246– Cour M. 14-Year incidence, progression, and visual
52. doi:10.1016/j.ophtha.2006.10.045. S0161-6420(06) morbidity of age-related maculopathy: the Copenhagen
01484-9 [pii]. City Eye Study. Ophthalmology. 2005;112(5):787–98.
Boyer DS, Heier JS, Brown DM, Francom SF, Ianchulev doi:10.1016/j.ophtha.2004.11.040.
T, Rubio RG. A Phase IIIb study to evaluate the Busbee BG, Ho AC, Brown DM, Heier JS, Suner IJ, Li Z,
safety of ranibizumab in subjects with neovascular Rubio RG, Lai P. Twelve-month efficacy and safety of
age-related macular degeneration. Ophthalmology. 0.5 mg or 2.0 mg ranibizumab in patients with subfo-
2009;116(9):1731–9.doi:10.1016/j.ophtha.2009.05.024. veal neovascular age-related macular degeneration.
S0161-6420(09)00543-0 [pii]. Ophthalmology. 2013;120(5):1046–56. doi:10.1016/j.
Braithwaite T, Nanji AA, Lindsley K, Greenberg ophtha.2012.10.014.
PB. Anti-vascular endothelial growth factor for macu- Byun YJ, Lee SJ, Koh HJ. Predictors of response after
lar oedema secondary to central retinal vein occlusion. intravitreal bevacizumab injection for neovascular
Cochrane Database Syst Rev. 2014;5:Cd007325. age-related macular degeneration. Jpn J Ophthalmol.
doi:10.1002/14651858.CD007325.pub3. 2010;54(6):571–7. doi:10.1007/s10384-010-0866-1.
Brantley Jr MA, Fang AM, King JM, Tewari A, Kymes Campochiaro PA, Brown DM, Awh CC, Lee SY, Gray S,
SM, Shiels A. Association of complement factor H and Saroj N, Murahashi WY, Rubio RG. Sustained benefits
LOC387715 genotypes with response of exudative from ranibizumab for macular edema following central
age-related macular degeneration to intravitreal beva- retinal vein occlusion: twelve-month outcomes of a
cizumab. Ophthalmology. 2007;114(12):2168–73. phase III study. Ophthalmology. 2011;118(10):2041–9.
doi:10.1016/j.ophtha.2007.09.008. S0161-6420(07) doi:10.1016/j.ophtha.2011.02.038.
01026-3 [pii]. Chakravarthy U, Harding SP, Rogers CA, Downes SM,
Bressler NM, Bressler SB, Congdon NG, Ferris 3rd FL, Lotery AJ, Wordsworth S, Reeves BC. Ranibizumab
Friedman DS, Klein R, Lindblad AS, Milton RC, versus bevacizumab to treat neovascular age-related
Seddon JM. Potential public health impact of Age- macular degeneration: one-year findings from the IVAN
Related Eye Disease Study results: AREDS report no. randomized trial. Ophthalmology. 2012;119(7):1399–
11. Arch Ophthalmol. 2003;121(11):1621–4. 411. doi:10.1016/j.ophtha.2012.04.015.
doi:10.1001/archopht.121.11.1621. Chakravarthy U, Harding SP, Rogers CA, Downes SM,
Bressler NM, Doan QV, Varma R, Lee PP, Suner IJ, Dolan Lotery AJ, Culliford LA, Reeves BC. Alternative treat-
C, Danese MD, Yu E, Tran I, Colman S. Estimated ments to inhibit VEGF in age-related choroidal neovas-
cases of legal blindness and visual impairment avoided cularisation: 2-year findings of the IVAN randomised
using ranibizumab for choroidal neovascularization: controlled trial. Lancet. 2013;382(9900):1258–67.
non-Hispanic white population in the United States doi:10.1016/s0140-6736(13)61501-9.
with age-related macular degeneration. Arch Chamberlain M, Baird P, Dirani M, Guymer R. Unraveling
Ophthalmol. 2011;129(6):709–17. doi:10.1001/ a complex genetic disease: age-related macular degen-
archophthalmol.2011.140. 129/6/709 [pii]. eration. Surv Ophthalmol. 2006;51(6):576–86.
Brown DM, Kaiser PK, Michels M, Soubrane G, Heier doi:10.1016/j.survophthal.2006.08.003.
JS, Kim RY, Sy JP, Schneider S. Ranibizumab versus Chang W, Noh DH, Sagong M, Kim IT. Pharmacogenetic
verteporfin for neovascular age-related macular association with early response to intravitreal ranibi-
degeneration. N Engl J Med. 2006;355(14):1432–44. zumab for age-related macular degeneration in a
doi:10.1056/NEJMoa062655. 355/14/1432 [pii]. Korean population. Mol Vis. 2013;19:702–9.
Brown DM, Michels M, Kaiser PK, Heier JS, Sy JP, Chen Y, Wiesmann C, Fuh G, Li B, Christinger HW,
Ianchulev T. Ranibizumab versus verteporfin photody- McKay P, de Vos AM, Lowman HB. Selection and
namic therapy for neovascular age-related macular analysis of an optimized anti-VEGF antibody: crystal
degeneration: two-year results of the ANCHOR study. structure of an affinity-matured Fab in complex with
Ophthalmology. 2009;116(1):57–65. doi:10.1016/j.oph- antigen. J Mol Biol. 1999;293(4):865–81. doi:10.1006/
tha.2008.10.018. e55S0161-6420(08)01075-0 [pii]. jmbi.1999.3192.
Finger RP, Wiedemann P, Blumhagen F, Pohl K, Holz Ophthalmol. 2008;246(9):1229–34. doi:10.1007/

FG. Treatment patterns, visual acuity and quality-of-life s00417-008-0854-z.
outcomes of the WAVE study—a noninterventional Heimes B, Lommatzsch A, Zeimer M, Gutfleisch M,
study of ranibizumab treatment for neovascular age- Spital G, Pauleikhoff D. Anti-VEGF therapy of exuda-
related macular degeneration in Germany. Acta tive AMD: Prognostic factors for therapy success.
Ophthalmol.2013;91(6):540–6.doi:10.1111/j.1755-3768. Ophthalmologe. 2011;108(2):124–31. doi:10.1007/
2012.02493.x. s00347-010-2210-z.
Finger RP, Wickremasinghe SS, Baird PN, Guymer Hjelmqvist L, Lindberg C, Kanulf P, Dahlgren H,
RH. Predictors of anti-VEGF treatment response in Johansson I, Siewert A. One-year outcomes using
neovascular age-related macular degeneration. Surv ranibizumab for neovascular age-related macular
Ophthalmol. 2014;59(1):1–18. doi:10.1016/j. degeneration: results of a prospective and retrospec-
survophthal.2013.03.009. tive observational multicentre study. J Ophthalmol.
Francis PJ. The influence of genetics on response to treat- 2011;2011:405724. doi:10.1155/2011/405724.
ment with ranibizumab (Lucentis) for age-related Holz FG, Amoaku W, Donate J, Guymer RH, Kellner U,
macular degeneration: the Lucentis Genotype Study Schlingemann RO, Weichselberger A, Staurenghi
(an American Ophthalmological Society thesis). Trans G. Safety and efficacy of a flexible dosing regimen of
Am Ophthalmol Soc. 2011;109:115–56. ranibizumab in neovascular age-related macular
Fung AE, Lalwani GA, Rosenfeld PJ, Dubovy SR, degeneration: the SUSTAIN study. Ophthalmology.
Michels S, Feuer WJ, Puliafito CA, Davis JL, Flynn Jr 2011;118(4):663–71. doi:10.1016/j.oph-
HW, Esquiabro M. An optical coherence tomography- tha.2010.12.019. S0161-6420(10)01372-2 [pii].
guided, variable dosing regimen with intravitreal Holz FG, Bandello F, Gillies M, Mitchell P, Osborne A,
ranibizumab (Lucentis) for neovascular age-related Sheidow T, Souied E, Figueroa MS. Safety of ranibi-
macular degeneration. Am J Ophthalmol. zumab in routine clinical practice: 1-year retrospective
2007;143(4):566–83. doi:10.1016/j.ajo.2007.01.028. pooled analysis of four European neovascular AMD
S0002-9394(07)00068-2 [pii]. registries within the LUMINOUS programme. Br
Gorin MB. Genetic insights into age-related macular J Ophthalmol. 2013;97(9):1161–7. doi:10.1136/
degeneration: controversies addressing risk, causality, bjophthalmol-2013-303232.
and therapeutics. Mol Aspects Med. 2012;33(4):467– Horster R, Ristau T, Sadda SR, Liakopoulos S. Individual
86. doi:10.1016/j.mam.2012.04.004. recurrence intervals after anti-VEGF therapy for age-
Guber J, Josifova T, Henrich PB, Guber I. Clinical risk related macular degeneration. Graefes Arch Clin Exp
factors for poor anatomic response to ranibizumab in Ophthalmol. 2011;249(5):645–52. doi:10.1007/
neovascular age-related macular degeneration. Open s00417-010-1588-2.
Ophthalmol J. 2014;8:3–6. doi:10.2174/18743641014 ICH (2014) International conference on harmonisation of
08010003. technical requirements for registration of pharmaceu-
Gupta OP, Shienbaum G, Patel AH, Fecarotta C, Kaiser ticals for human use (ICH). Introductory guide
RS, Regillo CD. A treat and extend regimen using MedDRA version 17.0. http://www.meddra.org/sites/
ranibizumab for neovascular age-related macular default/files/guidance/file/intguide_17_0_english.pdf.
degeneration clinical and economic impact. Accessed 21 September 2014.
Ophthalmology. 2010;117(11):2134–40. doi:10.1016/j. Imai D, Mori K, Horie-Inoue K, Gehlbach PL, Awata T,
ophtha.2010.02.032. S0161-6420(10)00244-7 [pii]. Inoue S, Yoneya S. CFH, VEGF, and PEDF genotypes
Hagstrom SA, Ying GS, Pauer GJ, Sturgill-Short GM, and the response to intravitreous injection of bevaci-
Huang J, Callanan DG, Kim IK, Klein ML, Maguire zumab for the treatment of age-related macular degen-
MG, Martin DF. Pharmacogenetics for genes associ- eration. J Ocul Biol Dis Infor. 2010;3(2):53–9.
ated with age-related macular degeneration in the doi:10.1007/s12177-010-9055-1.
Comparison of AMD Treatments Trials (CATT). Inglehearn CF, Ali M, Gale R, Cassidy F, Varma D,
Ophthalmology. 2013;120(3):593–9. doi:10.1016/j. Downey LM, Baxter PD, McKibbin M. Improved
ophtha.2012.11.037. response to ranibizumab in ex and current smokers
Heier JS, Brown DM, Chong V, Korobelnik JF, Kaiser with Age-related Macular Degeneration (AMD), but
PK, Nguyen QD, Kirchhof B, Ho A, Ogura Y, no evidence that CFH, ARMS2/HTRA1 or VEGF
Yancopoulos GD, Stahl N, Vitti R, Berliner AJ, Soo Y, Genotypes Predict Treatment Outcome. Invest
Anderesi M, Groetzbach G, Sommerauer B, Sandbrink Ophthalmol Vis Sci. 2012;53(6):3325.
R, Simader C, Schmidt-Erfurth U. Intravitreal afliber- Introini U, Torres Gimeno A, Scotti F, Setaccioli M,
cept (VEGF trap-eye) in wet age-related macular Giatsidis S, Bandello F. Vascularized retinal pigment
degeneration. Ophthalmology. 2012;119(12):2537– epithelial detachment in age-related macular degener-
48. doi:10.1016/j.ophtha.2012.09.006. ation: treatment and RPE tear incidence. Graefes Arch
Heimes B, Lommatzsch A, Zeimer M, Gutfleisch M, Clin Exp Ophthalmol. 2012;250(9):1283–92.
Spital G, Bird AC, Pauleikhoff D. Foveal RPE auto- doi:10.1007/s00417-012-1955-2.
fluorescence as a prognostic factor for anti-VEGF Jeng DJ, Sadda SR. Commonly evaluated optical coher-
therapy in exudative AMD. Graefes Arch Clin Exp ence tomography features are not predictive of number
of anti-VEGF injections in neovascular age-related tion: an updated meta-analysis of randomised clinical

macular degeneration. Invest Ophthalmol Vis Sci. trials. Graefes Arch Clin Exp Ophthalmol.
2010;51. 2014;252(10):1529–37. doi:10.1007/s00417-014-
Jyothi S, Chowdhury H, Elagouz M, Sivaprasad 2764-6.
S. Intravitreal bevacizumab (Avastin) for age-related Kolb S, Menghini M, Barthelmes D, Sutter F, Kurz-Levin
macular degeneration: a critical analysis of literature. M. The predictive value of OCT characteristics for the
Eye (Lond). 2010;24(5):816–24. doi:10.1038/ visual outcome in neovascular AMD. Klin Monbl
eye.2009.219. Augenheilkd. 2012;229(4):343–7. doi:10.1055/s-0031-
Kaiser PK, Brown DM, Zhang K, Hudson HL, Holz FG, 1299250.
Shapiro H, Schneider S, Acharya NR. Ranibizumab Krebs I, Schmetterer L, Boltz A, Told R, Vecsei-Marlovits
for predominantly classic neovascular age-related V, Egger S, Schonherr U, Haas A, Ansari-Shahrezaei S,
macular degeneration: subgroup analysis of first-year Binder S. A randomised double-masked trial comparing
ANCHOR results. Am J Ophthalmol. the visual outcome after treatment with ranibizumab or
2007;144(6):850–7. doi:10.1016/j.ajo.2007.08.012. bevacizumab in patients with neovascular age-related
S0002-9394(07)00718-0 [pii]. macular degeneration. Br J Ophthalmol. 2013;97(3):266–
Kang S, Roh YJ. One-year results of intravitreal ranibi- 71. doi:10.1136/bjophthalmol-2012-302391.
zumab for neovascular age-related macular degenera- Kruger Falk M, Kemp H, Sorensen TL. Four-year treat-
tion and clinical responses of various subgroups. Jpn ment results of neovascular age-related macular degen-
J Ophthalmol. 2009;53(4):389–95. doi:10.1007/ eration with ranibizumab and causes for discontinuation
s10384-009-0670-y. of treatment. Am J Ophthalmol. 2013;155(1):89–95.
Kiernan DF, Hariprasad SM, Rusu IM, Mehta SV, Mieler doi:10.1016/j.ajo.2012.06.031. e83.
WF, Jager RD. Epidemiology of the association Kwon YH, Lee DK, Kim HE, Kwon OW. Predictive find-
between anticoagulants and intraocular hemorrhage in ings of visual outcome in spectral domain optical coher-
patients with neovascular age-related macular degen- ence tomography after ranibizumab treatment in
eration. Retina. 2010;30(10):1573–8. doi:10.1097/ age-related macular degeneration. Korean J Ophthalmol.
IAE.0b013e3181e2266d. 2014;28(5):386–92. doi:10.3341/kjo.2014.28.5.386.
Kitchens JW, Kassem N, Wood W, Stone TW, Isernhagen Lad EM, Grunwald L, Mettu PS, Serrano NP, Crowell S,
R, Wood E, Hancock BA, Radovich M, Waymire J, Li Cousins SW. Lesion morphology on indocyanine
L, Schneider BP. A pharmacogenetics study to predict green angiography in age-related macular degenera-
outcome in patients receiving anti-VEGF therapy in tion with classic choroidal neovascular membrane:
age related macular degeneration. Clin Ophthalmol. implications for response to anti-VEGF treatment.
2013;7:1987–93. doi:10.2147/opth.s39635. Invest Ophthalmol Vis Sci. 2012;53: 5161.
Klein R, Klein BE, Knudtson MD, Meuer SM, Swift M, Lalwani GA, Rosenfeld PJ, Fung AE, Dubovy SR,
Gangnon RE. Fifteen-year cumulative incidence of Michels S, Feuer W, Davis JL, Flynn Jr HW, Esquiabro
age-related macular degeneration: the Beaver Dam M. A variable-dosing regimen with intravitreal ranibi-
Eye Study. Ophthalmology. 2007;114(2):253–62. zumab for neovascular age-related macular degenera-
doi:10.1016/j.ophtha.2006.10.040. tion: year 2 of the PrONTO Study. Am J Ophthalmol.
Kloeckener-Gruissem B, Barthelmes D, Labs S, Schindler 2009;148(1):43–58. doi:10.1016/j.ajo.2009.01.024.
C, Kurz-Levin M, Michels S, Fleischhauer J, Berger e41S0002-9394(09)00102-0 [pii].
W, Sutter F, Menghini M. Genetic association with Lanzetta P, Mitchell P, Wolf S, Veritti D. Different anti-
response to intravitreal ranibizumab in patients with vascular endothelial growth factor treatments and regi-
neovascular AMD. Invest Ophthalmol Vis Sci. mens and their outcomes in neovascular age-related
2011;52(7):4694–702. doi:10.1167/iovs.10-6080. macular degeneration: a literature review. Br
iovs.10-6080 [pii]. J Ophthalmol. 2013;97(12):1497–507. doi:10.1136/
Kloeckener-Gruissem B, Menghini M, Sutter F, bjophthalmol-2013-303394.
Fleischhauer J, Kurz-Levin MM, Berger W, Lee AY, Raya AK, Kymes SM, Shiels A, Brantley Jr
Barthelmes D. Long-term studies on the anti-VEGF MA. Pharmacogenetics of complement factor H
treatment success in age-related macular degeneration (Y402H) and treatment of exudative age-related mac-
(AMD). Invest Ophthalmol Vis Sci. 2012;53:5168. ular degeneration with ranibizumab. Br J Ophthalmol.
Kodjikian L, Souied EH, Mimoun G, Mauget-Faysse M, 2009;93(5):610–3. doi:10.1136/bjo.2008.150995.
Behar-Cohen F, Decullier E, Huot L, Aulagner bjo.2008.150995 [pii].
G. Ranibizumab versus bevacizumab for neovascular Lee S, Song SJ, Yu HG. Current smoking is associated
age-related macular degeneration: results from the with a poor visual acuity improvement after intravit-
GEFAL Noninferiority Randomized Trial. real ranibizumab therapy in patients with exudative
Ophthalmology. 2013;120(11):2300–9. doi:10.1016/j. age-related macular degeneration. J Korean Med Sci.
ophtha.2013.06.020. 2013;28(5):769–74. doi:10.3346/jkms.2013.28.5.769.
Kodjikian L, Decullier E, Souied EH, Girmens JF, Durand Levy J, Shneck M, Rosen S, Klemperer I, Rand D,
EE, Chapuis FR, Huot L. Bevacizumab and ranibi- Weinstein O, Pitchkhadze A, Belfair N, Lifshitz
zumab for neovascular age-related macular degenera- T. Intravitreal bevacizumab (Avastin) for subfoveal
neovascular age-related macular degeneration. Int Menghini M, Kloeckener-Gruissem B, Fleischhauer J,

Ophthalmol. 2009;29(5):349–57. doi:10.1007/ Kurz-Levin MM, Sutter FK, Berger W, Barthelmes
s10792-008-9243-1. D. Impact of loading phase, initial response and CFH
Lim JH, Wickremasinghe SS, Xie J, Chauhan DS, Baird PN, genotype on the long-term outcome of treatment for
Robman LD, Hageman G, Guymer RH (2012a) Delay to neovascular age-related macular degeneration.
treatment and visual outcomes in patients treated with PLoS One. 2012;7(7), e42014. doi:10.1371/journal.
anti-vascular endothelial growth factor for age-related pone.0042014.
macular degeneration. Am J Ophthalmol. 153(4):678– Mettu PS, Crowell S, Shaw J, Grunwald L, Lad EM,
686, 686.e671-672;2012a. doi:10.1016/j.ajo.2011.09.013 Serrano N, Cousins SW. Neovascular morphology on
Lim LS, Mitchell P, Seddon JM, Holz FG, Wong TY. Age- ICG angiography predicts response to anti-VEGF
related macular degeneration. Lancet. therapy in eyes with serous pigment epithelial detach-
2012b;379(9827):1728–38. doi:10.1016/ ments and age related macular degeneration. Invest
S0140-6736(12)60282-7. Ophtholmol Vis Sci. 2012;53:2654.
Lux A, Llacer H, Heussen FM, Joussen AM. Non- Miller JW, Le Couter J, Strauss EC, Ferrara N. Vascular
responders to bevacizumab (Avastin) therapy of cho- endothelial growth factor a in intraocular vascular dis-
roidal neovascular lesions. Br J Ophthalmol. ease. Ophthalmology. 2013;120(1):106–14. doi:10.1016/
2007;91(10):1318–22. doi:10.1136/bjo.2006.113902. j.ophtha.2012.07.038.
Mariani A, Deli A, Ambresin A, Mantel I. Characteristics Mitchell P. A systematic review of the efficacy and safety
of eyes with secondary loss of visual acuity receiving outcomes of anti-VEGF agents used for treating neo-
variable dosing ranibizumab for neovascular age- vascular age-related macular degeneration: compari-
related macular degeneration. Graefes Arch Clin Exp son of ranibizumab and bevacizumab. Curr Med Res
Ophthalmol. 2011;249(11):1635–42. doi:10.1007/ Opin. 2011;27(7):1465–75. doi:10.1185/03007995.20
s00417-011-1734-5. 11.585394.
Marques IP, Fonseca P, Luz Cachulo M, Pires I, Figueira Mitchell P, Bandello F, Schmidt-Erfurth U, Lang GE,
J, Faria de Abreu JR, Silva R. Treatment of exudative Massin P, Schlingemann RO, Sutter F, Simader C,
age-related macular degeneration with intravitreal Burian G, Gerstner O, Weichselberger A. The
ranibizumab in clinical practice: a 3-year follow-up. RESTORE study: ranibizumab monotherapy or com-
Ophthalmologica. 2013;229(3):158–67. doi:10.1159/ bined with laser versus laser monotherapy for diabetic
000343709. macular edema. Ophthalmology. 2011;118(4):615–
Martin DF, Maguire MG, Ying GS, Grunwald JE, Fine 25. doi:10.1016/j.ophtha.2011.01.031.
SL, Jaffe GJ. Ranibizumab and bevacizumab for neo- Mitchell P, Bressler N, Doan QV, Dolan C, Ferreira A,
vascular age-related macular degeneration. N Engl Osborne A, Rochtchina E, Danese M, Colman S,
J Med. 2011;364(20):1897–908. doi:10.1056/ Wong TY. Estimated cases of blindness and visual
NEJMoa1102673. impairment from neovascular age-related macular
Martin DF, Maguire MG, Fine SL, Ying GS, Jaffe GJ, degeneration avoided in Australia by ranibizumab
Grunwald JE, Toth C, Redford M, Ferris 3rd treatment. PLoS One. 2014;9(6), e101072.
FL. Ranibizumab and bevacizumab for treatment of doi:10.1371/journal.pone.0101072.
neovascular age-related macular degeneration: two- Moja L, Lucenteforte E, Kwag KH, Bertele V, Campomori
year results. Ophthalmology. 2012;119(7):1388–98. A, Chakravarthy U, D’Amico R, Dickersin K,
doi:10.1016/j.ophtha.2012.03.053. Kodjikian L, Lindsley K, Loke Y, Maguire M, Martin
Mathew R, Richardson M, Sivaprasad S (2013) Predictive DF, Mugelli A, Muhlbauer B, Puntmann I, Reeves B,
value of spectral-domain optical coherence tomography Rogers C, Schmucker C, Subramanian ML, Virgili
features in assessment of visual prognosis in eyes with G. Systemic safety of bevacizumab versus ranibi-
neovascular age-related macular degeneration treated zumab for neovascular age-related macular degenera-
with ranibizumab. Am J Ophthalmol. 155(4):720–726, tion. Cochrane Database Syst Rev. 2014;9:Cd011230.
726.e721. doi:10.1016/j.ajo.2012.11.003 doi:10.1002/14651858.CD011230.pub2.
McKibbin M, Ali M, Bansal S, Baxter PD, West K, Muniraju R, Ramu J, Sivaprasad S. Three-year visual out-
Williams G, Cassidy F, Inglehearn CF. CFH, VEGF come and injection frequency of intravitreal ranibi-
and HTRA1 promoter genotype may influence the zumab therapy for neovascular age-related macular
response to intravitreal ranibizumab therapy for neo- degeneration. Ophthalmologica. 2013;230(1):27–33.
vascular age-related macular degeneration. Br doi:10.1159/000350238.
J Ophthalmol. 2012;96(2):208–12. doi:10.1136/ Naj AC, Scott WK, Courtenay MD, Cade WH, Schwartz
bjo.2010.193680. SG, Kovach JL, Agarwal A, Wang G, Haines JL,
Menger JF, Haubitz I, Keilhauer-Strachwitz CN. Influence Pericak-Vance MA. Genetic factors in nonsmokers
Of AMD-risk factors on the effectiveness of anti- with age-related macular degeneration revealed
VEGF therapy in neovascular age-related macular through genome-wide gene-environment interaction
degeneration. Invest Ophthalmol Vis Sci. 2012; analysis. Ann Hum Genet. 2013;77(3):215–31.
53(6):857. doi:10.1111/ahg.12011.
Nakata I, Yamashiro K, Nakanishi H, Tsujikawa A, Otani Regillo CD, Brown DM, Abraham P, Yue H, Ianchulev T,
A, Yoshimura N. VEGF gene polymorphism and Schneider S, Shams N. Randomized, double-masked,
response to intravitreal bevacizumab and triple therapy sham-controlled trial of ranibizumab for neovascular
in age-related macular degeneration. Jpn J Ophthalmol. age-related macular degeneration: PIER Study year 1.
2011;55(5):435–43. doi:10.1007/s10384-011-0061-z. Am J Ophthalmol. 2008;145(2):239–48. doi:10.1016/j.
NCT00559715 (2007) Prevention of Vision Loss in ajo.2007.10.004. S0002-9394(07)00881-1 [pii].
Patients With Age-Related Macular Degeneration Ristau T, Keane PA, Walsh AC, Engin A, Mokwa N,
(AMD) by Intravitreal Injection of Bevacizumab and Kirchhof B, Sadda SR, Liakopoulos S. Relationship
Ranibizumab (VIBERA). http://clinicaltrials.gov/ between visual acuity and spectral domain optical
show/NCT00559715. coherence tomography retinal parameters in neovascu-
Nischler C, Oberkofler H, Ortner C, Paikl D, Riha W, lar age-related macular degeneration. Ophthalmologica.
Lang N, Patsch W, Egger SF. Complement factor H 2014;231(1):37–44. doi:10.1159/000354551.
Y402H gene polymorphism and response to intravit- Rivera A, Fisher SA, Fritsche LG, Keilhauer CN, Lichtner
real bevacizumab in exudative age-related macular P, Meitinger T, Weber BH. Hypothetical LOC387715
degeneration. Acta Ophthalmol. 2011;89(4):e344–9. is a second major susceptibility gene for age-related
doi:10.1111/j.1755-3768.2010.02080.x. macular degeneration, contributing independently of
Oishi A, Shimozono M, Mandai M, Hata M, Nishida A, complement factor H to disease risk. Hum Mol Genet.
Kurimoto Y. Recovery of photoreceptor outer segments 2005;14(21):3227–36. doi:10.1093/hmg/ddi353.
after anti-VEGF therapy for age-related macular degen- Rofagha S, Bhisitkul RB, Boyer DS, Sadda SR,
eration. Graefes Arch Clin Exp Ophthalmol. Zhang K. Seven-year outcomes in ranibizumab-
2013;251(2):435–40. doi:10.1007/s00417-012-2034-4. treated patients in ANCHOR, MARINA, and
Orlin A, Hadley D, Chang W, Ho AC, Brown G, Kaiser HORIZON: a multicenter cohort study (SEVEN-UP).
RS, Regillo CD, Godshalk AN, Lier A, Kaderli B, Ophthalmology. 2013;120(11):2292–9. doi:10.1016/j.
Stambolian D. Association between high-risk disease ophtha.2013.03.046.
loci and response to anti-vascular endothelial growth Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser
factor treatment for wet age-related macular degenera- PK, Chung CY, Kim RY. Ranibizumab for neovascu-
tion. Retina. 2012;32(1):4–9. doi:10.1097/ lar age-related macular degeneration. N Engl J Med.
IAE.0b013e31822a2c7c. 2006;355(14):1419–31. doi:10.1056/NEJMoa054481.
Oubraham H, Cohen SY, Samimi S, Marotte D, Bouzaher 355/14/1419 [pii].
I, Bonicel P, Fajnkuchen F, Tadayoni R. Inject and Rosenfeld PJ, Shapiro H, Tuomi L, Webster M, Elledge J,
extend dosing versus dosing as needed: a comparative Blodi B. Characteristics of patients losing vision after
retrospective study of ranibizumab in exudative age- 2 years of monthly dosing in the phase III ranibizumab
related macular degeneration. Retina. 2011;31(1):26– clinical trials. Ophthalmology. 2011;118(3):523–30.
30. doi:10.1097/IAE.0b013e3181de5609. doi:10.1016/j.ophtha.2010.07.011. S0161-6420(10)
Park UC, Shin JY, Kim SJ, Shin ES, Lee JE, McCarthy 00747-5 [pii].
LC, Newcombe PJ, Xu CF, Chung H, Yu HG. Genetic Rung L, Lovestam-Adrian M. Three-year follow-up of
factors associated with response to intravitreal ranibi- visual outcome and quality of life in patients with age-
zumab in Korean patients with neovascular age-related related macular degeneration. Clin Ophthalmol.
macular degeneration. Retina. 2014;34(2):288–97. 2013;7:395–401. doi:10.2147/opth.s41585.
doi:10.1097/IAE.0b013e3182979e1e. Sacu S, Michels S, Prager F, Weigert G, Dunavoelgyi R,
Rakic JM, Leys A, Brie H, Denhaerynck K, Pacheco C, Geitzenauer W, Pruente C, Schmidt-Erfurth
Vancayzeele S, Hermans C, Macdonald K, Abraham U. Randomised clinical trial of intravitreal Avastin vs
I. Real-world variability in ranibizumab treatment and photodynamic therapy and intravitreal triamcinolone:
associated clinical, quality of life, and safety outcomes long-term results. Eye (Lond). 2009;23(12):2223–7.
over 24 months in patients with neovascular age-related doi:10.1038/eye.2008.423.
macular degeneration: the HELIOS study. Clin Schauwvlieghe A-SM, Dijkman G, Hooymans JM,
Ophthalmo. 2013;7:1849–58. doi:10.2147/opth.s49385. Verbraak FD, Dijkgraaf MG, Peto T, Vingerling JR,
Rasmussen A, Sander B. Long-term longitudinal study of Hoyng C, Schlingemann RO (2014) Comparing the
patients treated with ranibizumab for neovascular age- effectiveness of bevacizumab to ranibizumab in
related macular degeneration. Curr Opin Ophthalmol. patients with exudative age-related macular degenera-
2014. doi:10.1097/ICU.0000000000000050. tion. BRAMD. Invest Ophthalmol Vis Sci 55 (ARVO
Rasmussen A, Bloch SB, Fuchs J, Hansen LH, Larsen M, E-Abstract 870).
Lacour M, Lund-Andersen H, Sander B. A 4-year lon- Schmidt-Erfurth U, Eldem B, Guymer R, Korobelnik JF,
gitudinal study of 555 patients treated with ranibi- Schlingemann RO, Axer-Siegel R, Wiedemann P,
zumab for neovascular age-related macular Simader C, Gekkieva M, Weichselberger A. Efficacy
degeneration. Ophthalmology. 2013;120(12):2630–6. and safety of monthly versus quarterly ranibizumab
doi:10.1016/j.ophtha.2013.05.018. treatment in neovascular age-related macular degen-
eration: the EXCITE study. Ophthalmology. Tano Y, Ohji M. EXTEND-I: safety and efficacy of
2011;118(5):831–9. doi:10.1016/j.oph- ranibizumab in Japanese patients with subfoveal cho-
tha.2010.09.004. S0161-6420(10)00972-3 [pii]. roidal neovascularization secondary to age-related
Schmidt-Erfurth U, Kaiser PK, Korobelnik JF, Brown DM, macular degeneration. Acta Ophthalmol. 2010;88(3):
Chong V, Nguyen QD, Ho AC, Ogura Y, Simader C, 309–16. doi:10.1111/j.1755-3768.2009.01843.x.
Jaffe GJ, Slakter JS, Yancopoulos GD, Stahl N, Vitti R, AOS1843 [pii].
Berliner AJ, Soo Y, Anderesi M, Sowade O, Zeitz O, Tano Y, Ohji M. Long-term efficacy and safety of ranibi-
Norenberg C, Sandbrink R, Heier JS. Intravitreal zumab administered pro re nata in Japanese patients with
aflibercept injection for neovascular age-related macu- neovascular age-related macular degeneration in the
lar degeneration: ninety-six-week results of the VIEW EXTEND-I study. Acta Ophthalmol. 2011;89(3):208–
studies. Ophthalmology. 2014;121(1):193–201. 17. doi:10.1111/j.1755-3768.2010.02065.x.
doi:10.1016/j.ophtha.2013.08.011. Teper SJ, Nowinska A, Pilat J, Palucha A, Wylegala
Silva R, Axer-Siegel R, Eldem B, Guymer R, Kirchhof B, E. Involvement of genetic factors in the response to a
Papp A, Seres A, Gekkieva M, Nieweg A, Pilz S. The variable-dosing ranibizumab treatment regimen for
SECURE study: long-term safety of ranibizumab age-related macular degeneration. Mol Vis.
0.5 mg in neovascular age-related macular degenera- 2010;16:2598–604.
tion. Ophthalmology. 2013;120(1):130–9. The Neovascular Age-Related Macular Degeneration
doi:10.1016/j.ophtha.2012.07.026. Database. Multicenter study of 92 976 ranibizumab
Singer MA, Awh CC, Sadda S, Freeman WR, Antoszyk injections: report 1: visual acuity. Ophthalmology.
AN, Wong P, Tuomi L. HORIZON: an open-label 2014;121(5):1092–101. doi:10.1016/j.ophtha.2013.
extension trial of ranibizumab for choroidal neovascu- 11.031.
larization secondary to age-related macular degenera- Tian J, Qin X, Fang K, Chen Q, Hou J, Li J, Yu W, Chen
tion. Ophthalmology. 2012;119(6):1175–83. D, Hu Y, Li X. Association of genetic polymor-
doi:10.1016/j.ophtha.2011.12.016. phisms with response to bevacizumab for neovascu-
Skaat A, Chetrit A, Belkin M, Kinori M, Kalter-Leibovici lar age-related macular degeneration in the Chinese
O. Time trends in the incidence and causes of blind- population. Pharmacogenomics. 2012;13(7):779–
ness in Israel. Am J Ophthalmol. 2012;153(2):214–21. 87. doi:10.2217/pgs.12.53.
doi:10.1016/j.ajo.2011.08.035. e211. Tufail A, Patel PJ, Egan C, Hykin P, da Cruz L, Gregor Z,
Smailhodzic D, Muether PS, Chen J, Kwestro A, Zhang Dowler J, Majid MA, Bailey C, Mohamed Q, Johnston
AY, Omar A, Van de Ven JP, Keunen JE, Kirchhof B, R, Bunce C, Xing W. Bevacizumab for neovascular
Hoyng CB, Klevering BJ, Koenekoop RK, Fauser S, age related macular degeneration (ABC Trial): multi-
den Hollander AI. Cumulative effect of risk alleles in centre randomised double masked study. BMJ.
CFH, ARMS2, and VEGFA on the response to ranibi- 2010;340:c2459. doi:10.1136/bmj.c2459.
zumab treatment in age-related macular degeneration. Wang JJ, Rochtchina E, Lee AJ, Chia EM, Smith W,
Ophthalmology. 2012;119(11):2304–11. doi:10.1016/j. Cumming RG, Mitchell P. Ten-year incidence and pro-
ophtha.2012.05.040. gression of age-related maculopathy: the Blue Mountains
Subramanian ML, Ness S, Abedi G, Ahmed E, Daly M, Eye Study. Ophthalmology. 2007;114(1):92–8.
Feinberg E, Bhatia S, Patel P, Nguyen M, Houranieh doi:10.1016/j.ophtha.2006.07.017.
A. Bevacizumab vs ranibizumab for age-related macu- Wang VM, Rosen RB, Meyerle CB, Kurup SK, Ardeljan
lar degeneration: early results of a prospective double- D, Agron E, Tai K, Pomykala M, Chew EY, Chan CC,
masked, randomized clinical trial. Am J Ophthalmol. Tuo J. Suggestive association between PLA2G12A
2009;148(6):875–82. doi:10.1016/j.ajo.2009.07.009. single nucleotide polymorphism rs2285714 and
e871S0002-9394(09)00505-4 [pii]. response to anti-vascular endothelial growth factor
Subramanian ML, Abedi G, Ness S, Ahmed E, Fenberg therapy in patients with exudative age-related macular
M, Daly MK, Houranieh A, Feinberg EB. Bevacizumab degeneration. Mol Vis. 2012;18:2578–85.
vs ranibizumab for age-related macular degeneration: Wickremasinghe SS, Xie J, Lim J, Chauhan DS, Robman
1-year outcomes of a prospective, double-masked ran- L, Richardson AJ, Hageman G, Baird PN, Guymer
domised clinical trial. Eye (Lond). 2010;24(11):1708– R. Variants in the APOE gene are associated with
15. doi:10.1038/eye.2010.147. eye2010147 [pii]. improved outcome after anti-VEGF treatment for neo-
Takeda AL, Colquitt J, Clegg AJ, Jones J. Pegaptanib and vascular AMD. Invest Ophthalmol Vis Sci.
ranibizumab for neovascular age-related macular 2011;52(7):4072–9. doi:10.1167/iovs.10-6550. iovs.
degeneration: a systematic review. Br J Ophthalmol. 10-6550 [pii].
2007;91(9):1177–82. doi:10.1136/bjo.2007.118562. Yamashiro K, Tomita K, Tsujikawa A, Nakata I, Akagi-
bjo.2007.118562 [pii]. Kurashige Y, Miyake M, Ooto S, Tamura H, Yoshimura
Tannan A, Li C, Stevens TS. Predicting treatment out- N. Factors associated with the response of age-related
comes in patients with newly diagnosed exudative macular degeneration to intravitreal ranibizumab
age-related macular degeneration based on initial pretreatment. Am J Ophthalmol. 2012;154(1):125–36.
sentation. Invest Ophtholmol Vis Sci. 2010;51. doi:10.1016/j.ajo.2012.01.010.
Ying GS, Huang J, Maguire MG, Jaffe GJ, Grunwald JE, recurrent neovascular age-related macular degenera-
Toth C, Daniel E, Klein M, Pieramici D, Wells J, tion. Am J Ophthalmol. 2013;156(1):29–35.
Martin DF. Baseline predictors for one-year visual doi:10.1016/j.ajo.2013.03.030.
outcomes with ranibizumab or bevacizumab for neo- Zhao L, Grob S, Avery R, Kimura A, Pieramici D, Lee J,
vascular age-related macular degeneration. Rabena M, Ortiz S, Quach J, Cao G, Luo H, Zhang M,
Ophthalmology. 2013;120(1):122–9. doi:10.1016/j. Pei M, Song Y, Tornambe P, Goldbaum M, Ferreyra H,
ophtha.2012.07.042. Kozak I, Zhang K. Common variant in VEGFA and
Yonekawa Y, Andreoli C, Miller JB, Loewenstein JI, response to anti-VEGF therapy for neovascular age-
Sobrin L, Eliott D, Vavvas DG, Miller JW, Kim related macular degeneration. Curr Mol Med.
IK. Conversion to aflibercept for chronic refractory or 2013;13(6):929–34.
AREDS Supplementation
and the Progression Towards
4
Exudative AMD
David J. Valent and Emily Y. Chew
With the retina’s high concentration of polyunsat-

4.1 Introduction urated fatty acids and exposure to both sunlight
and high oxygen content, the retina is particularly
Age-related macular degeneration (AMD) is a susceptible to production of oxidative radicals.
neurodegenerative disease of the eye and, accord- This hypothesis, along with other supportive
ing to the World Health Organization, is the third studies, led to the advent of Age-Related Eye
leading cause of blindness in the world (WHO Disease Study (AREDS).
2012). AMD affects between 25 and 30 million
people worldwide (Chopdar et al. 2003; Friedman
et al. 2004), and with an aging population, this 4.2 Age-Related Eye
number is projected to increase markedly. The Disease Study
neovascular or “wet” form of the disease has effec-
tive therapies with intravitreal injections of anti- The AREDS was a large, multicentered, phase III,
vascular endothelial growth factor (VEGF) agents randomized, double-masked placebo-controlled
(Brown et al. 2006; Heier et al. 2012; Kaiser et al. clinical trial (Age-Related Eye Disease Study
2007; Rosenfeld et al. 2006). There are no treatments Research Group 1999). The primary goal of this
currently for the atrophic or “dry” form of the study was to assess the clinical course, prognosis,
disease leading to several studies investigating and risk factors including the role of nutrition for
treatments that attempt to prevent the progression the development and progression of both AMD
of geographic atrophy associated with AMD and cataracts. The rational for this study was mul-
(Borrelli et al. 2012; Studnicka et al. 2013; tifactorial. With limited data regarding the natural
Yehoshua et al. 2014). This idea of preventing history of both cataract formation and AMD, fur-
progression of disease is paramount and can signifi- ther studies were deemed necessary. Additionally,
cantly reduce morbidity. One hypothesis regarding study results suggested that antioxidant vitamins
the pathogenesis and progression of AMD involves and supplements may reduce the risk of cardio-
damage to the retina secondary to oxidative stress. vascular disease (Stampfer et al. 1993) and eye
diseases including the development of cataracts
(Gerster 1989; Robertson et al. 1989) and macular
D.J. Valent • E.Y. Chew (*) degeneration (Newsome et al. 1988). A random-
National Institutes of Health, National Eye Institute,
10 Center Drive, Building 10, 10D45, Bethesda, MD
ized double-masked placebo-controlled trial
20892-1166, USA involving 151 patients with the diagnosis of mac-
e-mail: dv299204@ohio.edu; echew@nei.nih.gov ular degeneration, defined as the presence of

68 D.J. Valent and E.Y. Chew
drusen and pigmentary changes, showed that Table 4.1 Baseline ocular characteristics from partici-
pants from the Age-Related Eye Disease Study (1999)
those participants who received zinc supplemen-
tation of 100 mg, had less visual acuity loss than Number of
Category Definition participants
those in the placebo group at 12 and 24 months
1 Few if any drusen 1117
(Newsome et al. 1988). In addition, when examin-
2 Extensive small drusen, 1063
ing color fundus photographs, those participants pigment abnormalities, or at
in the zinc group were more likely to remain sta- least 1 intermediate size druse
ble or accumulate less drusen than those in the 3 Extensive intermediate drusen, 1621
placebo group (Newsome et al. 1988). It has been GA not involving the center of
the macula, or at least 1 large
shown in the pigmented rat model that zinc defi-
druse
ciency leads to an increase of lipofuscin accumu- 4 Advanced AMD or visual 956
lation in the RPE when compared with controls acuity less than 20/32 due to
(Julien et al. 2011). However, findings from pop- AMD in 1 eye
ulation-based trials and observational data investi-
gating antioxidant use in AMD were inconsistent
and consisted of varying types and doses of anti- Table 4.2 Study formulation of the antioxidants used in
oxidants and other micronutrients (Christen et al. the Age-Related Eye Disease Study (1999)
1999; Eye Diseases Prevalence Research Group Study formulation Daily dose
1993; Goldberg et al. 1988; Seddon et al. 1994; Antioxidants
VandenLangenberg et al. 1998). With at least Vitamin C 500 mg
some supportive trials on nutritional supplemen- Vitamin E 400 IU
tation and eye health, commercially available Beta-carotene 15 mg
vitamins were being used by patients both with Zinc (zinc oxide) 80 mg
Copper (cupric oxide) 2 mg
and without macular degeneration without strong
efficacy (Sperduto et al. 1990). In addition to pro-
viding valuable information regarding the natural ture (Age-Related Eye Disease Study Research
history of two very common eye conditions, Group 1999). Lutein and zeaxanthin were known
AREDS attempted to establish efficacy and safety to be present in the retina however, were not
for antioxidant vitamins in decreasing the pro- included in the AREDS supplements because
gression of AMD and/or cataracts. they were not commercially available at the time
AREDS recruited 4757 participants aged of study. The primary outcome measure was pro-
55–80 years, between 1992 and 1998, and these gression to or treatment for advanced AMD and
participants were enrolled into different AMD moderate visual acuity loss (>15 letters) from
severity categories based on their fundus find- baseline. Secondary outcomes included the
ings. Table 4.1 describes the ocular findings and development of neovascular macular degenera-
the number of participants in each category. tion, incidence of geographic atrophy, progres-
These participants were randomly assigned to sion to advanced AMD with an associated 15
receive either daily oral tablets of (1) high-dose letter decrease in visual acuity, and worsening of
antioxidants, (2) zinc and copper, (3) high-dose AMD classification category 2 participants to
antioxidants plus zinc, or (4) placebo (Age- category 3 or 4 during follow-up.
Related Eye Disease Study Research Group Advanced AMD was defined as either photo-
1999). The formulation for the AREDS vitamins coagulation or other treatment for choroidal neo-
can be seen in Table 4.2 (Age-Related Eye vascularization (CNV) or photographic
Disease Study Research Group 1999). The dos- documentation of any of the following: non-
ages of the different vitamins and supplements drusenoid retinal pigment epithelial detachment,
were determined by a panel of expert biochem- serous or hemorrhagic retinal detachment, hem-
ists, nutritionists, and ophthalmologists after orrhage under the retina or the retinal pigment
reviewing the basic science and published litera- epithelium and/or subretinal fibrosis (Age-
4 AREDS Supplementation and the Progression Towards Exudative AMD 69
Related Eye Disease Study Research Group At 5 years in those with more severe AMD
2001). Another definition of advanced AMD was (categories 3 and 4), there was a 25 % reduction
the development of geographic atrophy involving in the risk of progressing to advanced disease in
the center of the macula. the antioxidants plus zinc group when compared
These participants were followed every 6 with the placebo group (Odds Ratio (OR) 0.66;
months during the course of the clinical trial and 99 % confidence interval (CI), 0.47–0.91). As for
then annually. Each study visit included a com- visual acuity in those in categories 3 and 4, there
plete ophthalmic examination. Best corrected was a statistically significant reduction in the risk
visual acuity was obtained using the Early of 15 or more letter loss in the antioxidant plus
Treatment Diabetic Retinopathy Study protocol zinc group when compared with the placebo
at every annual visit. Stereoscopic fundus photo- group (OR 0.73; 99 % CI, 0.54–0.99) (Age-
graphs of the macula were taken both at baseline Related Eye Disease Study Research Group
and then annually starting at year 2. These photo- 2001).
graphs were evaluated by certified personnel One secondary outcome measure was the like-
using standardized grading protocols at the lihood of progression to neovascular disease.
Fundus Photograph Reading center. Fundus pho- There were 592 participants that developed neo-
tographs were also taken if there was a decrease vascular disease in AREDS. There was a statisti-
in visual acuity of 10 or more letter from base- cally significant benefit for those participants in
line. Demographic information, smoking history, category 3 and 4 in the antioxidant plus zinc
medical history, nutrition, and medication use group when compared to placebo for the devel-
were also obtained at baseline (Age-Related Eye opment of CNV (OR 0.62 99 % CI, 0.43–0.90).
Disease Study Research Group 1999). The effect on the development of CNV of zinc
vs. no zinc was statistically significant as well
(OR 0.76; 99 % CI, 0.58–0.98) and was sugges-
4.3 Outcomes tive for the zinc arm when compared with pla-
cebo (OR 0.73; 99 % CI, 0.51–1.04) (Age-Related
Four thousand seven hundred fifty seven partici- Eye Disease Study Research Group 2001).
pants were evaluated for a median duration of However, this finding was not significant for the
6.5 years. Only 2.4 % of the AREDS participants antioxidant alone group when compared with
were lost to follow-up (Age-Related Eye Disease placebo (OR, 0.79; 99 % CI, 0.56–1.13) (Age-
Study Research Group 2001). Of the over 4700 Related Eye Disease Study Research Group
AREDS participants, 1117 participants had few if 2001).
any drusen. In this group, only 5 (0.004 %) Risk factors for the development for advanced
progressed to advanced AMD during the study disease in AREDS based off the 5 year data has
while 1.3 % in category 2 progressed to advanced been evaluated (Clemons et al. 2005). Using
disease at 5 years. Those with eyes assigned to multivariable models while controlling for age,
category 3, the probability to progression to gender, and treatment group, risk factors associ-
advanced AMD was 18 %. However, when looking ated with advancement to neovascular AMD
exclusively at those patients with either non- included race (OR, white vs. black, 6.77; 95 %
central GA in at least 1 eye or those that had CI, 1.24–36.9) and larger amount smoked (OR,
bilateral large drusen, these participants were >10 vs. ≤10 pack years, 1.55; 95 % CI, 1.15–
four times as likely to progress to advanced AMD 2.09). Smoking was also a statistically significant
when compared to the rest of the category three risk factor for the development of geographic
participants. Those with advanced AMD in 1 eye, atrophy. In those persons at risk of developing
defined as category 4, had a 43 % expected prob- advanced AMD in one eye, as defined as those
ability of progression to advanced AMD in the patients having unilateral advanced AMD, diabe-
fellow eye at 5 years (Age-Related Eye Disease tes was associated with an increased incidence of
Study Research Group 2001). neovascular AMD (Clemons et al. 2005).
Fig. 4.1 Odds ratio (central dot) and 95 % confidence 3 and 4; and C: Categories 4. GA geographic atrophy.
intervals for each original treatment assignment compared Modified from Chew EY et al. Long-term effects of vita-
with placebo for participants in the following Age-Related mins C and E, β-carotene, and zinc on age-related macular
Eye Disease Study age-related macular degeneration degeneration: AREDS report no. 35. Ophthalmology. 2013,
(AMD) categories. A: Categories 2, 3, and 4; B: Categories Aug; 120 (8): 1604-11. With permission from Elsevier
Ten year data regarding risk factors for pro- follow-up study was approximately 70 % in the
gression to neovascular disease has also been last year of participation. Approximately 5 years
published on this population (Chew et al. 2014). after the clinical trial was completed, those in the
Increasing age, female sex (Hazard ratio (HR), antioxidant plus zinc formulation continued to
1.23; p = 0.005) and current smoking (HR 1.56; have statistically significant reduced odds of
p < 0.001) were all associated with development developing advanced AMD when compared with
of neovascular AMD (Chew et al. 2014). Forty- those in the placebo group for those in categories
eight percent of those participants between the 2, 3, or 4 (OR, 0.69; 95 % CI 0.56–0.86 P = 0.001)
ages of 75 and 80 and in categories 3 or 4 at base- (Chew et al. 2013). There was also a statistically
line went on to develop neovascular AMD at 10 significant reduction in the odds for the develop-
years. As Fig. 4.1 shows, assignment to antioxi- ment of neovascular disease (OR, 0.64; 95 % CI
dants plus zinc formulation in the original clini- 0.50–0.82 P < 0.001) for those in the antioxidant
cal trial appeared to decrease the probability of plus zinc group when compared with the placebo
developing neovascular AMD but not central arm. Those participants in categories 2, 3, and 4
geographic atrophy (Chew et al. 2013, 2014). who were randomly assigned to the antioxidant at
After the clinical trial had been completed, baseline, also showed statistically significant
participants were invited to continue in a long- reduction in the likelihood for developing
term observational study (Chew et al. 2013). This advanced AMD, especially the development on
constituted a comprehensive eye examination neovascular disease (Chew et al. 2013).
including assessment of visual acuity using the When looking exclusively at those with cate-
EDTRS standardized protocol and color fundus gory 4 disease, there was a significant reduction in
photographs. In addition, when the AREDS for- the odds of developing both advanced AMD (OR
mulation became available, participants in 0.56; 95 % CI, 0.40–0.79; P < 0.001) and neovas-
AREDS category 3 or 4 were offered the antioxi- cular AMD (OR 0.44; 95 % CI 0.30–0.65;
dant plus zinc formulation. The observational p < 0.001) in those assigned to the antioxidant plus
study enrolled 3549 of the 4203 surviving par- zinc formulation when compared with controls.
ticipants (Chew et al. 2013). Estimates when examining exclusively category 3
The proportion of the participants in categories disease trended in the beneficial direction but were
3 and 4 taking the AREDS formulation during the not statistically significant (Chew et al. 2013).
The impact of the AREDS supplementation At baseline, the original AREDS participants
on public health is quite significant. Statistical took a semiquantitative food frequency question-
models estimate that if all high-risk patients were naire and those reporting the highest intake of
able to receive AREDS supplementation, it could omega-3 LCPUFAs (Sangiovanni et al. 2009b;
potentially prevent 300,000 patients from devel- SanGiovanni et al. 2007) and lutein/zeaxanthin
oping advanced AMD at 5 years (Bressler et al. (Age-Related Eye Disease Study Research Group
2003). This would have a huge impact on health 2007) were less likely to progress to advanced
care costs. This includes both direct costs, which disease when compared to those with lower
include doctor visits and medications, and indi- intake of these compounds. Those in the group
rect costs, such as caregiver-related costs. Direct with the highest intake of omega 3 LCPUFA
costs for just the treatment for wet macular (Sangiovanni et al. 2009b) or lutein and zeaxan-
degeneration can range between $385 and thin (Age-Related Eye Disease Study Research
$23,400 per year depending on the medication Group et al. 2007) were less likely to develop
and frequency of injections (CATT Research neovascular AMD when compared with those
Group et al. 2011). with the lowest intake. In addition, higher fish
consumption was also associated with lower risk
of developing neovascular AMD (Sangiovanni
4.4 Age-Related Eye Disease et al. 2009b). Additional observational studies
Study 2 also demonstrated a similar correlation between
higher dietary consumption of lutein/zeaxanthin
With the significant cost savings in the preven- and omega-3 LCPUFAs (Augood et al. 2008;
tion of advanced macular degeneration, addi- Chua et al. 2006; Moeller et al. 2006; SanGiovanni
tional studies provided rational for investigating et al. 2008, 2009a; Seddon et al. 2006; Swenor
additional supplements for the prevention of et al. 2010; Tan et al. 2008) and decreased pro-
advanced AMD. One area of focus was carot- gression to advanced disease. In a prospective
enoids. Carotenoids, specifically the xantho- study involving fifteen patients with AMD treated
phylls, lutein, and zeaxanthin are the major with lutein/zeaxanthin, functional improvement
component of the macular pigment (Landrum was seen on electroretinograms at 1 year when
and Bone 2001). As stated previously, these were compared with age-matched controls (Parisi
not available as supplements for use at the start of et al. 2008). Because of these supportive findings,
AREDS, but became commercially available and the Age-Related Eye Disease Study 2 (AREDS2)
sparked much interest. These xanthophylls are was performed to investigate the safety and effi-
thought to provide some protection against light- cacy of adding lutein/zeaxanthin and omega 3
induced damage and possibly AMD (Krinsky LCPUFAs to the AREDS supplements for per-
et al. 2003; Landrum and Bone 2001). The pro- sons at high risk for progression to advanced dis-
posed mechanism for protection by lutein and ease. In addition, AREDS2 also evaluated the
zeaxanthin include the ability to absorb blue light outcome of reducing the concentration of zinc
and the elimination of free radicals (Schalch and omitting beta-carotene.
1992; Snodderly 1995). Omega-3 long chain AREDS2 was a large, phase III, randomized,
poly-unsaturated fatty acids (LCPUFAs) are also double-masked, placebo-controlled clinical trial
thought to play an important role within the ret- involving 82 clinical sites across the United
ina, including a role in ganglion cell function States (AREDS2 Research Group 2012). For the
(Nguyen et al. 2008), making up some of the con- primary analysis, there were 4203 participants
stituents of the photoreceptor outer segments enrolled between the ages of 50–85 years. All
(Bazan et al. 1992; Litman and Mitchell 1996) participants were considered at risk for the devel-
modulation of rhodopsin (Bush et al. 1994) and opment of advanced AMD, with the majority of
may have a neuroprotective effect (Bazan et al. the patients having bilateral large drusen (65 %)
2013). at baseline. The remaining participants had
advanced AMD in one eye and large drusen in Table 4.3 Four treatment formulations in the second ran-
domization in AREDS2 (2012)
the fellow eye. The primary outcome for
AREDS2 was the comparison of the three active Cupric
Vitamin C Vitamin E Beta Zinc oxide
treatment arms to placebo on progression to
(mg) (IU) carotene (mg) (mg)
advanced AMD. Advanced AMD was defined as 1 500 400 15 mg 80 2
atrophic or neovascular changes of AMD that 2 500 400 0 80 2
include one or more of the following: definite 3 500 400 15 mg 25 2
geographic atrophy involving the center of the 4 500 400 0 25 2
macula or evidence suggesting CNV. Secondary
analysis included progression to moderate vision
loss (≥15 letter loss), progression of lens opacity information regarding any additional therapies
and moderate vision loss, or improvement in par- for their AMD.
ticipants with advanced AMD (AREDS2
Research Group 2012).
The AREDS2 participants were randomly 4.5 Results
assigned, with equal probability, to one of four
study formulations daily: lutein (10 mg)/zeaxan- The primary analysis of AREDS2 demonstrated
thin (2 mg), omega-3 LCPUFAs in the form of no beneficial or harmful effects of adding lutein/
docosahexaenoic acid (DHA) 350 mg and eicos- zeaxanthin (HR 0.90 98.7 % CI 0.76–1.07;
apentaenoic acid (EPA) 650 mg, both lutein/zea- P = 0.12), omega-3 LCPUFAs (HR 0.97; 98.7 %
xanthin and DHA/EPA or placebo. Of the 4203 CI 0.82–1.16; P = 0.70), or both (HR 0.89, 95 %
participants, 1167 (28 %) opted to take the CI 0.75–1.05, P = 0.10) to the original AREDS
AREDS supplement, while 3036 (72 %) agreed formulation on the progression to advanced
to the secondary randomization, which was per- AMD when compared with the placebo (Age-
formed to evaluate the effect of removing beta- Related Eye Disease Study 2 Research Group
carotene and/or lowering the level of zinc from 2013). However, the pre-specified analyses of the
the original AREDS formulation. The reasoning main effects (taking accounts of all those taking
for eliminating beta-carotene is the increased risk lutein/zeaxanthin vs. no lutein/zeaxanthin), there
for developing lung cancer in those persons who was an incremental increase in the effect of
are smokers and taking beta-carotene (Albanes reducing the risk of progression to advanced
et al. 1995; Omenn et al. 1996). The interest in AMD. When examining those participants in the
lower the levels of zinc was that evidence sug- lowest quintile of dietary intake of lutein and zea-
gested that the maximal level of zinc absorbed xanthin, comparing those that had received
was closer to 25 mg than 80 mg (Hambidge 2003; lutein/zeaxanthin to those that did not, those
Newsome et al. 1988). The four treatment formu- receiving lutein/zeaxanthin were less likely to
lations are shown in Table 4.3. In those progress to advanced disease (HR: 0.74; 95 % CI
participants that were current or former smokers 0.59–0.94; P = 0.01) (Age-Related Eye Disease
who discontinued their tobacco use within the Study 2 Research Group 2013) Secondary analy-
previous year were assigned to one of the two sis was performed comparing the lutein/zeaxan-
arms that excluded beta-carotene. Participants thin group vs. the beta-carotene group, the hazard
were followed at annual study visits, which ratios were 0.82 (95 % CI, 0.64–0.94; P = 0.02)
included a complete eye examination with best for progression to advanced disease and 0.78
corrected visual acuity using an electronic version (95 % CI, 0.64–0.94; p = 0.01) for developing
of the ETDRS technique. Standardized fundus neovascular AMD (Age-Related Eye Disease
photographs were obtained at each visit and eval- Study AREDS2 Research Group 2014). When
uated at a central reading center (AREDS2 comparing those assigned to the lutein/zeaxan-
Research Group 2012). Participants were con- thin and beta-carotene vs. beta-carotene only, the
tacted by telephone 3 months after randomization results were: significant for the development of
and at 6 months between study visits to acquire advanced AMD (HR 0.82; 95 % CI, 0.69–0.97;
P = 0.02) and for the progression to neovascular however, there was no treatment effect on the
AMD (HR 0.72; 95 % CI, 0.59–0.89; P = 0.002) development of geographic atrophy. The exact
(Age-Related Eye Disease Study AREDS2 mechanism for this beneficial effect is unknown.
Research Group 2014). It may be that the two processes have different
An important adverse effect emerged when cellular signaling which, in the formation of
evaluating the treatment with beta-carotene in CNV, may more responsive to the potential anti-
those AREDS2 participants who were former oxidative or immune-modulatory effects of sup-
smokers and those who had never smoked. There plements. This is purely speculation and this
were more participants that were diagnosed with question cannot be answered based on the clini-
lung cancer in the beta-carotene group when cal or basic science data available at this time. In
compared with the no beta-carotene group (23 addition, as 34 % of patients with GA in one eye
(2.0 %) vs. 11 (0.9 %), P = 0.04) (Age-Related and neovascular AMD in the fellow eye will
Eye Disease Study 2 Research Group 2013). Of develop neovascularization in the eye with GA at
those who developed lung cancer, 91 % were for- 4 years (Sunness et al. 1999), treatment with
mer smokers (Age-Related Eye Disease Study 2 AREDS/AREDS2 supplementation in this group
Research Group 2013). should be considered.
In regard to visual acuity, lutein/zeaxanthin or Could the addition of other supplements or
omega 3 LCPUFAs had no effects on moderate vitamins be protective against the progression to
visual acuity loss, defined as a loss of 15 or more advanced AMD? Perhaps AREDS2 supplemen-
letters from baseline (Age-Related Eye Disease tation did not include every known retinal carot-
Study 2 Research Group 2013). Lowering zinc enoid. To date, there are as many as 25 dietary
and eliminating beta-carotene had no significant carotenoids identified; some of these could pro-
effect on the progression to advanced AMD vide added benefit to the current AREDS supple-
(Age-Related Eye Disease Study 2 Research mentation. As shown by Bernstein et al. (2001),
Group 2013). These findings, in addition to the the RPE and choroid have not only lutein and
potential risks in smokers, have led to the recom- zeaxanthin, but several other carotenoids which
mendation of substituting lutein/zeaxanthin for could assist in protecting against oxidative dam-
beta-carotene in the AREDS formulation. age. There are ongoing studies attempting to
Additionally, when analysis was limited to those evaluate three carotenoids in a double-blinded
just with bilateral large drusen at baseline, the randomized placebo-controlled trial (Akuffo
comparison of supplements containing lutein/ et al. 2014). In addition, the Women’s Antioxidant
zeaxanthin vs. beta-carotene produced a hazard and Folic Acid Cardiovascular Study (Christen
ratio of 0.76 (95 % CI 0.61–0.96; P = 0.02) for the et al. 2009) showed daily supplementation of B
progression to advanced AMD and 0.65 (95 % vitamins, folic acid, pyridoxine, and cyanoco-
CI, 0.49–0.85; P = 0.002) for the progression to balamin may decrease the risk of AMD in a sec-
neovascular AMD (Age-Related Eye Disease ondary analysis. Vitamin D may also play a role
Study AREDS2 Research Group 2014). A sig- as there is supportive evidence that it too may
nificant reduction in the risk of geographic atro- protect against the AMD (Millen et al. 2011;
phy was not identified. Montgomery et al. 2010).
Could the supplementation of lutein and zea-
xanthin given at an earlier time point in the
4.6 Discussion patient’s care have an effect on the development
of AMD? One study which was an off shoot of
Both AREDS and AREDS2 clearly demonstrated the Woman’s Health Initiative evaluated 1787
the role of antioxidants for the progression of women aged 50–79 years of age and examined
advanced disease in those with intermediate their dietary intake of lutein plus zeaxanthin
AMD. Both AREDS and in subgroup analysis in (Moeller et al. 2006). There was no statistically
AREDS2, the supplements appear to be protec- difference between those with low and high
tive against the formation of neovascular disease; intake of lutein and zeaxanthin in the develop-
ment of intermediate or advanced AMD. However, References

when evaluating the participants less than 75
years of age, without a history chronic disease Age-Related Eye Disease Study (AREDS2) Research
Group. Antioxidant status and neovascular age-related
and stable dietary intake of lutein and zeaxanthin, macular degeneration. Eye Disease Case-Control
this did become statistically significant. In addi- Study Group. Arch Ophthalmol. 1993;111:104–9.
tion, there was a trend that higher intake of lutein Age-Related Eye Disease Study (AREDS2) Research
and zeaxanthin was protective against advanced Group, Chew EY, Clemons TE, Sangiovanni JP, Danis
RP, Ferris 3rd FL, Elman MJ, Antoszyk AN, Ruby AJ,
disease (Moeller et al. 2006). Another study Orth D, Bressler SB, Fish GE, Hubbard GB, Klein ML,
investigated a link, independent of dietary con- Chandra SR, Blodi BA, Domalpally A, Friberg T,
sumption, between genetic determinates for mac- Wong WT, Rosenfeld PJ, Agrón E, Toth CA, Bernstein
ular xanthophylls and AMD (Meyers et al. 2014). PS, Sperduto RD. Secondary analyses of the effects of
lutein/zeaxanthin on age-related macular degeneration
How these genes influence the onset and severity progression: AREDS2 report No. 3. JAMA Ophthalmol.
of AMD is yet to be determined. 2014;132:142–9.
Age-Related Eye Disease Study 2 Research Group. Lutein
+ zeaxanthin and omega-3 fatty acids for age-related
macular degeneration: the Age-Related Eye Disease
4.7 Conclusions Study 2 (AREDS2) randomized clinical trial. JAMA.
2013;309:2005–15.
AREDS and AREDS2 have shown that supple- Age-Related Eye Disease Study Research Group. The
mentation with antioxidants and zinc may decrease Age-Related Eye Disease Study (AREDS): design
implications. AREDS report no. 1. Control Clin Trials.
the risk of developing advanced AMD, especially 1999;20:573–600.
neovascular AMD. This risk reduction appears Age-Related Eye Disease Study Research Group. A ran-
greatest for those at highest risk for advanced dis- domized, placebo-controlled, clinical trial of high-
ease, categories 3 and 4. The totality of evidence dose supplementation with vitamins C and E, beta
carotene, and zinc for age-related macular degenera-
from the pre-planned secondary analyses as well tion and vision loss: AREDS report no. 8. Arch
as post hoc AREDS2 did demonstrate an incre- Ophthalmol. 2001;119:1417–36.
mental benefit of lutein/zeaxanthin on progression Age-Related Eye Disease Study Research Group,
to advanced AMD, especially for those in the low- SanGiovanni JP, Chew EY, Clemons TE, Ferris 3rd
FL, Gensler G, Lindblad AS, Milton RC, Seddon JM,
est quintile of lutein dietary intake, those with Sperduto RD. The relationship of dietary carotenoid
bilateral large drusen and those advancing to neo- and vitamin A, E, and C intake with age-related macu-
vascular AMD. Given this data, and the concerns lar degeneration in a case-control study: AREDS
associated with beta-carotene, consideration for Report No. 22. Arch Ophthalmol. 2007;125:1225–32.
Akuffo KO, Beatty S, Stack J, Dennison J, O'Regan S,
substituting lutein/zeaxanthin in place of beta-car- Meagher KA, Peto T, Nolan J. Central Retinal
otene in AREDS supplements was recommended. Enrichment Supplementation Trials (CREST): design
Future considerations include the addition of other and methodology of the CREST randomized controlled
vitamins or supplements to further decrease the trials. Ophthalmic Epidemiol. 2014;21:111–23.
Albanes D, Heinonen OP, Huttunen JK, Taylor PR,
risk of progression to advanced disease. Virtamo J, Edwards BK, Haapakoski J, Rautalahti M,
Hartman AM, Palmgren J, et al. Effects of alpha-
tocopherol and beta-carotene supplements on cancer
Compliance with Ethical Requirements incidence in the Alpha-Tocopherol Beta-Carotene
Cancer Prevention Study. Am J Clin Nutr. 1995;62:
David Valent declares that he has no con- 1427S–30S.
flict of interest. AREDS2 Research Group, Chew EY, Clemons T,
Emily Chew declares that she has no SanGiovanni JP, Danis R, Domalpally A, McBee W,
conflict of interest. Sperduto R, Ferris FL. The Age-Related Eye Disease
Study 2 (AREDS2): study design and baseline charac-
Informed Consent and Animal Studies teristics (AREDS2 report number 1). Ophthalmology.
disclosures are not applicable to this 2012;119:2282–9.
review. Augood C, Chakravarthy U, Young I, Vioque J, de Jong
PT, Bentham G, Rahu M, Seland J, Soubrane G,
Tomazzoli L, Topouzis F, Vingerling JR, Fletcher supplement use and the risk of age-related maculopa-
AE. Oily fish consumption, dietary docosahexaenoic thy. Am J Epidemiol. 1999;149:476–84.
acid and eicosapentaenoic acid intakes, and associa- Christen WG, Glynn RJ, Chew EY, Albert CM, Manson
tions with neovascular age-related macular degenera- JE. Folic acid, pyridoxine, and cyanocobalamin combi-
tion. Am J Clin Nutr. 2008;88:398–406. nation treatment and age-related macular degeneration
Bazan NG, Gordon WC, Rodriguez de Turco in women: the Women's Antioxidant and Folic Acid
EB. Docosahexaenoic acid uptake and metabolism in Cardiovascular Study. Arch Intern Med. 2009;169:
photoreceptors: retinal conservation by an efficient 335–41.
retinal pigment epithelial cell-mediated recycling pro- Chua B, Flood V, Rochtchina E, Wang JJ, Smith W,
cess. Adv Exp Med Biol. 1992;318:295–306. Mitchell P. Dietary fatty acids and the 5-year inci-
Bazan NG, Calandria JM, Gordon WC. Docosahexaenoic dence of age-related maculopathy. Arch Ophthalmol.
acid and its derivative neuroprotectin D1 display neu- 2006;124:981–6.
roprotective properties in the retina, brain and central Clemons TE, Milton RC, Klein R, Seddon JM, Ferris 3rd
nervous system. Nestle Nutr Inst Workshop Ser. FL, Age-Related Eye Disease Study Research Group.
2013;77:121–31. Risk factors for the incidence of Advanced Age-
Bernstein PS, Khachik F, Carvalho LS, Muir GJ, Zhao Related Macular Degeneration in the Age-Related Eye
DY, Katz NB. Identification and quantitation of carot- Disease Study (AREDS) AREDS report no. 19.
enoids and their metabolites in the tissues of the Ophthalmology. 2005;112:533–9.
human eye. Exp Eye Res. 2001;72:215–23. Friedman DS, O'Colmain BJ, Munoz B, Tomany SC,
Borrelli E, Diadori A, Zalaffi A, Bocci V. Effects of major McCarty C, de Jong PT, Nemesure B, Mitchell P,
ozonated autohemotherapy in the treatment of dry age Kempen J, Eye Diseases Prevalence Research Group.
related macular degeneration: a randomized controlled Prevalence of age-related macular degeneration in the
clinical study. Int J Ophthalmol. 2012;5:708–13. United States. Arch Ophthalmol. 2004;122:564–72.
Bressler NM, Bressler SB, Congdon NG, Ferris 3rd FL, Gerster H. Antioxidant vitamins in cataract prevention. Z
Friedman DS, Klein R, Lindblad AS, Milton RC, Ernahrungswiss. 1989;28:56–75.
Seddon JM, Age-Related Eye Disease Study Research Goldberg J, Flowerdew G, Smith E, Brody JA, Tso
Group. Potential public health impact of Age-Related MO. Factors associated with age-related macular
Eye Disease Study results: AREDS report no. 11. degeneration. An analysis of data from the first
Arch Ophthalmol. 2003;121:1621–4. National Health and Nutrition Examination Survey.
Brown DM, Michels M, Kaiser PK, Heier JS, Sy JP, Am J Epidemiol. 1988;128:700–10.
Ianchulev T, ANCHOR Study Group. Ranibizumab Hambidge M. Underwood Memorial Lecture: human zinc
versus verteporfin for neovascular age-related macular homeostasis: good but not perfect. J Nutr. 2003;133:
degeneration. N Engl J Med. 2006;355:1432–44. 1438S–42S.
Bush RA, Malnoe A, Reme CE, Williams TP. Dietary Heier JS, Brown DM, Chong V, Korobelnik JF, Kaiser PK,
deficiency of N-3 fatty acids alters rhodopsin content Nguyen QD, Kirchhof B, Ho A, Ogura Y, Yancopoulos
and function in the rat retina. Invest Ophthalmol Vis GD, Stahl N, Vitti R, Berliner AJ, Soo Y, Anderesi M,
Sci. 1994;35:91–100. Groetzbach G, Sommerauer B, Sandbrink R, Simader
CATT Research Group, Martin DF, Maguire MG, Ying C, Schmidt-Erfurth U, VIEW 1 and VIEW 2 Study
GS, Grunwald JE, Fine SL, Jaffe GJ. Ranibizumab Groups. Intravitreal aflibercept (VEGF trap-eye) in wet
and bevacizumab for neovascular age-related macular age-related macular degeneration. Ophthalmology.
degeneration. N Engl J Med. 2011;364:1897–908. 2012;119:2537–48.
Chew EY, Clemons TE, Agrón E, Sperduto RD, Julien S, Biesemeier A, Kokkinou D, Eibl O, Schraermeyer
Sangiovanni JP, Kurinij N, Davis MD; Age-Related U. Zinc deficiency leads to lipofuscin accumulation in
Eye Disease Study Research Group. Long-term effects the retinal pigment epithelium of pigmented rats.
of vitamins C and E, beta-carotene, and zinc on age- PLoS One. 2011;6, e29245.
related macular degeneration: AREDS report no. 35. Kaiser PK, Brown DM, Zhang K, Hudson HL, Holz FG,
Ophthalmology 2013; 120: 1604–11 e1604. Shapiro H, Schneider S, Acharya NR. Ranibizumab for
Chew EY, Clemons TE, Agrón E, Sperduto RD, Sangiovanni predominantly classic neovascular age-related macular
JP, Davis MD, Ferris 3rd FL, Age-Related Eye Disease degeneration: subgroup analysis of first-year ANCHOR
Study Research Group. Ten-year follow-up of age- results. Am J Ophthalmol. 2007;144:850–7.
related macular degeneration in the age-related eye Krinsky NI, Landrum JT, Bone RA. Biologic mechanisms
disease study: AREDS report no. 36. JAMA Ophthalmol. of the protective role of lutein and zeaxanthin in the
2014;132:272–7. eye. Annu Rev Nutr. 2003;23:171–201.
Chopdar A, Chakravarthy U, Verma D. Age related macu- Landrum JT, Bone RA. Lutein, zeaxanthin, and the macular
lar degeneration. BMJ. 2003;326:485–8. pigment. Arch Biochem Biophys. 2001;385:28–40.
Christen WG, Ajani UA, Glynn RJ, Manson JE, Litman BJ, Mitchell DC. A role for phospholipid polyun-
Schaumberg DA, Chew EC, Buring JE, Hennekens saturation in modulating membrane protein function.
CH. Prospective cohort study of antioxidant vitamin Lipids. 1996;31(Suppl):S193–197.
Meyers KJ, Mares JA, Igo Jr RP, Truitt B, Liu Z, Millen SanGiovanni JP, Chew EY, Agrón E, Clemons TE, Ferris
AE, Klein M, Johnson EJ, Engelman CD, Karki CK, 3rd FL, Gensler G, Lindblad AS, Milton RC, Seddon
Blodi B, Gehrs K, Tinker L, Wallace R, Robinson J, JM, Klein R, Sperduto RD, Age-Related Eye Disease
LeBlanc ES, Sarto G, Bernstein PS, SanGiovanni JP, Study Research Group. The relationship of dietary
Iyengar SK. Genetic evidence for role of carotenoids omega-3 long-chain polyunsaturated fatty acid intake
in age-related macular degeneration in the Carotenoids with incident age-related macular degeneration: AREDS
in Age-Related Eye Disease Study (CAREDS). Invest report no. 23. Arch Ophthalmol. 2008;126:1274–9.
Ophthalmol Vis Sci. 2014;55:587–99. SanGiovanni JP, Agron E, Clemons TE, Chew
Millen AE, Voland R, Sondel SA, Parekh N, Horst RL, EY. Omega-3 long-chain polyunsaturated fatty acid
Wallace RB, Hageman GS, Chappell R, Blodi BA, intake inversely associated with 12-year progression
Klein ML, Gehrs KM, Sarto GE, Mares JA, CAREDS to advanced age-related macular degeneration. Arch
Study Group. Vitamin D status and early age-related Ophthalmol. 2009a;127:110–2.
macular degeneration in postmenopausal women. Sangiovanni JP, Agrón E, Meleth AD, Reed GF, Sperduto
Arch Ophthalmol. 2011;129:481–9. RD, Clemons TE, Chew EY, Age-Related Eye
Moeller SM, Parekh N, Tinker L, Ritenbaugh C, Blodi B, Disease Study Research Group. {Omega}-3 Long-
Wallace RB, Mares JA, CAREDS Research Study chain polyunsaturated fatty acid intake and 12-y inci-
Group. Associations between intermediate age-related dence of neovascular age-related macular
macular degeneration and lutein and zeaxanthin in the degeneration and central geographic atrophy: AREDS
Carotenoids in Age-related Eye Disease Study report 30, a prospective cohort study from the Age-
(CAREDS): ancillary study of the Women's Health Related Eye Disease Study. Am J Clin Nutr.
Initiative. Arch Ophthalmol. 2006;124:1151–62. 2009b;90:1601–7.
Montgomery MP, Kamel F, Pericak-Vance MA, Haines JL, Schalch W. Carotenoids in the retina—a review of their
Postel EA, Agarwal A, Richards M, Scott WK, Schmidt possible role in preventing or limiting damage caused
S. Overall diet quality and age-related macular degen- by light and oxygen. EXS. 1992;62:280–98.
eration. Ophthalmic Epidemiol. 2010;17:58–65. Seddon JM, Ajani UA, Sperduto RD, Hiller R, Blair N, Burton
Newsome DA, Swartz M, Leone NC, Elston RC, Miller TC, Farber MD, Gragoudas ES, Haller J, Miller DT, et al.
E. Oral zinc in macular degeneration. Arch Ophthalmol. Dietary carotenoids, vitamins A, C, and E, and advanced
1988;106:192–8. age-related macular degeneration. Eye Disease Case-
Nguyen CT, Vingrys AJ, Bui BV. Dietary omega-3 fatty Control Study Group. JAMA. 1994;272:1413–20.
acids and ganglion cell function. Invest Ophthalmol Seddon JM, George S, Rosner B. Cigarette smoking, fish
Vis Sci. 2008;49:3586–94. consumption, omega-3 fatty acid intake, and associa-
Omenn GS, Goodman GE, Thornquist MD, Balmes J, tions with age-related macular degeneration: the US
Cullen MR, Glass A, Keogh JP, Meyskens Jr FL, Twin Study of Age-Related Macular Degeneration.
Valanis B, Williams Jr JH, Barnhart S, Cherniack MG, Arch Ophthalmol. 2006;124:995–1001.
Brodkin CA, Hammar S. Risk factors for lung cancer Snodderly DM. Evidence for protection against age-related
and for intervention effects in CARET, the macular degeneration by carotenoids and antioxidant
Beta-Carotene and Retinol Efficacy Trial. J Natl vitamins. Am J Clin Nutr. 1995;62:1448S–61S.
Cancer Inst. 1996;88:1550–9. Sperduto RD, Ferris 3rd FL, Kurinij N. Do we have a nutri-
Parisi V, Tedeschi M, Gallinaro G, Varano M, Saviano S, tional treatment for age-related cataract or macular
Piermarocchi S, Group CS. Carotenoids and antioxi- degeneration? Arch Ophthalmol. 1990;108:1403–5.
dants in age-related maculopathy Italian study: multi- Stampfer MJ, Hennekens CH, Manson JE, Colditz GA,
focal electroretinogram modifications after 1 year. Rosner B, Willett WC. Vitamin E consumption and
Ophthalmology. 2008;115(324–333), e322. the risk of coronary disease in women. N Engl J Med.
Robertson JM, Donner AP, Trevithick JR. Vitamin E 1993;328:1444–9.
intake and risk of cataracts in humans. Ann N Y Acad Studnicka J, Rencova E, Blaha M, Rozsival P, Lanska M,
Sci. 1989;570:372–82. Blaha V, Nemcansky J, Langrova H. Long-term out-
Rosenfeld PJ, Brown DM, Heier JS, Boyer DS, Kaiser comes of rheohaemapheresis in the treatment of dry
PK, Chung CY, Kim RY, MARINA Study Group. form of age-related macular degeneration. J Ophthalmol.
Ranibizumab for neovascular age-related macular 2013;2013:135798.
degeneration. N Engl J Med. 2006;355:1419–31. Sunness JS, Gonzalez-Baron J, Bressler NM, Hawkins B,
SanGiovanni JP, Chew EY, Clemons TE, Davis MD, Ferris Applegate CA. The development of choroidal neovas-
FL, Gensler GR, Kurinij N, Lindblad AS, Milton RC, cularization in eyes with the geographic atrophy form
Seddon JM, Sperduto RD, Age-Related Eye Disease of age-related macular degeneration. Ophthalmology.
Study Research Group. The relationship of dietary lipid 1999;106:910–9.
intake and age-related macular degeneration in a case- Swenor BK, Bressler S, Caulfield L, West SK. The impact
control study: AREDS Report No. 20. Arch Ophthalmol. of fish and shellfish consumption on age-related macular
2007;125:671–9. degeneration. Ophthalmology. 2010;117:2395–401.
Tan JS, Wang JJ, Flood V, Rochtchina E, Smith W, Mitchell WHO. Global data on visual impairments 2010. Geneva:
P. Dietary antioxidants and the long-term incidence of WHO; 2012.
age-related macular degeneration: the Blue Mountains Yehoshua Z, de Amorim Garcia CA, Filho RP, Nunes G,
Eye Study. Ophthalmology. 2008;115:334–41. Gregori FM, Penha AA, Moshfeghi K, Zhang S,
VandenLangenberg GM, Mares-Perlman JA, Klein R, Sadda WF, Rosenfeld PJ. Systemic complement
Klein BE, Brady WE, Palta M. Associations between inhibition with eculizumab for geographic atrophy
antioxidant and zinc intake and the 5-year incidence of in age-related macular degeneration: the
early age-related maculopathy in the Beaver Dam Eye COMPLETE study. Ophthalmology. 2014;121:
Study. Am J Epidemiol. 1998;148:204–14. 693–701.
VEGF-Inhibition in Macular
Telangiectasia Type 2
5
Peter Charbel Issa and Frank G. Holz
(“pigment-plaques”), probably due to intrareti-

5.1 Introduction nal proliferation of pigment epithelial cells
along right-angled vessels. Stage 5 is character-
Macular telangiectasia (MacTel) type 2 is a mac- ized by secondary neovascular membranes, the
ular degenerative disease characterized by slowly so-called “proliferative MacTel type 2.”
progressing photoreceptor atrophy and peculiar However, it should be noted that neovascular
vascular alterations within the central retina membranes may develop at any time point and
(Charbel Issa et al. 2013). The disease usually should rather be seen as a complication during
presents bilaterally and has no gender predilec- the disease course than as the natural end point
tion. A genetic component has been suggested of the disease process (Charbel Issa et al. 2013).
due to occurrence in monozygotic twins and Neovascular membranes develop mostly from
other family members (Gillies et al. 2009; retinal vessels, i.e., not from the choroid as in
Charbel Issa et al. 2013). However, the exact age-related macular degeneration and many
genetic cause and the pathophysiology of MacTel other retinal diseases with secondary choroidal
type 2 currently remain unknown. neovascularization.
Gass and Blodi (1993) proposed five consec- In all disease stages, visualization of the vascu-
utive disease stages: stage 1 is characterized by lar alterations is enhanced on fluorescein angiog-
a diffuse paracentral (juxtafoveolar) hyperfluo- raphy, which may show ectatic capillaries in early
rescence in late phase fluorescein angiography. frames and leakage from these vessels in later
In stage 2, a reduced parafoveolar retinal trans- angiographic phases. The functionally relevant
parency becomes evident ophthalmoscopically. natural end point of the disease is an atrophy of the
The hallmark of stage 3 are dilated right-angled photoreceptor layer in the macular area (Charbel
venules and of stage 4 plaque-like pigmentation Issa et al. 2013) which is best visualized on spec-
tral-domain optical coherence tomography
Parts of the text of this chapter have been modified from (SD-OCT). All morphological changes predomi-
Charbel Issa P, Gillies MC. Macular telangiectasia type 2. nantly manifest temporal to the foveal center, but
Prog Retin Eye Res. 2013;34:49–77. doi:10.1016/j.pretey- an area extending by about 2 disc diameters across
eres.2012.11.002.Epub2012Dec3.
the central retina may eventually be affected.
P. Charbel Issa, M.D., D.Phil. (*) • F.G. Holz, M.D. First symptoms experienced by patients with
Department of Ophthalmology, University of Bonn,
MacTel type 2 commonly are reading problems
Ernst Abbe Strasse 2, Bonn 53127, Germany
e-mail: peter.issa@ukb.uni-bonn.de; and/or metamorphopsia (Heeren et al. 2013).
frank.holz@ukb.uni-bonn.de Reading difficulties have been explained by the

80 P. Charbel Issa and F.G. Holz
presence of paracentral scotomata, which repre- Therapeutic effects for MacTel type 2 may
sent the characteristic functional defect in this depend upon the presence or absence of a prolif-
disease (Finger et al. 2009). Initially, such scoto- erative disease stage. Thus, we separately report
mata develop in the nasal paracentral visual field, potential therapeutic options in non-proliferative
in correspondence to the predominant localiza- and proliferative (i.e., with a neovascular mem-
tion of morphological alterations in the temporal brane) disease. In any case, the potential benefit
paracentral area. Because the scotomata initially of any therapy will depend upon the presence of
are very small, they may be missed on conven- atrophic changes, irreversible cell death, in the
tional functional testing using static perimetry photoreceptor layer at the time of intervention
and visual acuity testing. Due to their sharp and during follow-up.
delineation from normally functioning retina,
visual acuity may be unimpaired even in the pres-
ence of an absolute scotoma. Functional deterio- 5.2 Non-proliferative Stage
ration beyond those related to degenerative of Macular Telangiectasia
changes may occur due to the development of Type 2
secondary neovascular membranes as well as
partial- or full-thickness macular holes (Charbel In disease stages without neovascular membranes
Issa et al. 2009, 2013). (i.e., stages 1–4), argon laser photocoagulation
Through increased research efforts stimulated (Chopdar 1978; Hutton et al. 1978; Friedman
by the Macular Telangiectasia Project (“MacTel et al. 1993; Gass and Blodi 1993; Park et al.
Project”; http://www.mactelresearch.org), it has 1997), photodynamic therapy (De Lahitte et al.
become evident that certain novel imaging 2004), intravitreal injection of steroids alone
modalities have a high diagnostic value and (Alldredge and Garretson 2003; Martinez 2003;
might also be relevant for longitudinal disease Cakir et al. 2006), or in combination with indo-
monitoring. Specifically, optical coherence cyanine green-mediated photothrombosis
tomography (OCT) commonly shows alterations (Arevalo et al. 2007) and posterior juxtascleral
within the central retina and may be the most sen- administration of steroids (Eandi et al. 2006)
sitive noninvasive imaging technology to diag- have been tried. All these therapeutic approaches
nose MacTel type 2. Characteristic alterations on have been shown to be ineffective or even to
OCT imaging include a disruption of the ellip- potentially accelerate disease progression. For
soid zone, hyporeflective cavities at the level of instance, there may be an increased risk for the
the inner or outer retina, and atrophy of the pho- development of new fibrovascular membranes
toreceptor layer in later disease stages (Charbel after focal argon laser photocoagulation in
Issa et al. 2013; Gaudric et al. 2006). Although MacTel type 2 (Gass and Blodi 1993; Park et al.
OCT findings are not included in the classifica- 1996, 1997; Engelbrecht et al. 2002). For other
tion by Gass and Blodi, these atrophic changes therapies, only very few cases have been reported
appear to be the functionally limiting natural end so that no conclusion can be drawn with regard to
point of the disease. effectiveness or safety.
Fundus autofluorescence is another imaging The eponymous ectatic and leaking capillaries
modality that reveals characteristic alterations in within the macular area have led to the hypothe-
patients with MacTel type 2. Typical changes sis that vascular endothelial growth factor
include a relatively increased signal within the (VEGF) might play a role in the pathophysiology
affected area due to loss of macular pigment of this disease. In a first report on intravitreal
(Charbel Issa et al. 2013). Such new noninvasive inhibition of VEGF, it was shown that intravitreal
imaging techniques may aid as an objective mea- injection of bevacizumab resulted in a marked
sure to assess therapeutic effects and to comple- decrease of vascular leakage; however, without
ment fluorescein angiography. significant improvement in visual acuity (Charbel
5 VEGF-Inhibition in Macular Telangiectasia Type 2 81
Issa et al. 2007). This observation confirmed the really administered ranibizumab was performed
hypothesis on a role of VEGF in the pathophysi- monthly for 1 year. In both studies, there was no
ology of MacTel type 2 and stimulated further significant change in visual acuity. Of note, para-
therapeutic investigations in small case series. A central photoreceptor atrophy and corresponding
report on short-term effects in six non- scotomata progressed slowly despite a marked
proliferative eyes revealed that intravitreal injec- effect on the vascular pathological changes (Figs.
tions of bevacizumab with an interval of 4 weeks 5.1, 5.2 and 5.3) (Charbel Issa et al. 2010, 2011;
resulted in a decrease in macular thickness on Toy et al. 2012). These findings prompted us to
SD-OCT imaging and a reduction of angiograph- conclude that regular intravitreal injections of
ically visible vascular changes in all eyes VEGF-inhibitors should not be recommended for
(Charbel Issa et al. 2007). However, visual acuity patients with non-proliferative MacTel type 2.
improved in only a subset of patients. Two subse- Nine of the ten patients included in one of these
quent studies reported a longer follow-up (18 and studies were examined again after a mean follow-
32 months, respectively) after intravitreal bevaci- up of 6.0 ± 0.4 years (Kupitz et al. 2015). More
zumab injections (Charbel Issa et al. 2008; Matt eyes of the treatment group had lost two or more
et al. 2010). These studies showed that the effect lines on BCVA testing (4 versus 1), although
of anti-VEGF treatment, monitored by fluores- mean visual acuity at baseline was similar in
cein angiography and OCT imaging, abated after treated and control eyes. In addition, more eyes
3–4 months. Additional treatments within the had developed an absolute paracentral scotoma (7
reported review period showed similar morpho- versus 2; Fig. 5.4) and a secondary neovascular
logical effects as the initial injections. membrane had formed in four of the treated and in
In two prospective studies including ten and none of the untreated fellow eyes (Kupitz et al.
five patients, respectively, treatment with intravit- 2015). Thus, although the worse outcome in
Fig. 5.1 Early fluorescein angiography in two different after therapy. Reproduced from Charbel Issa, P., R. P.
patients with non-proliferative macular telangiectasia type Finger, et al. Monthly ranibizumab for nonproliferative
2 before (left column) and after (right column) three intra- macular telangiectasia type 2: a 12-month prospective
vitreal injections of ranibizumab in monthly intervals. study. Am J Ophthalmol (2011) 151(5): 876-886 e871,
The telangiectatic capillaries show a marked reduction with permission of Elsevier
Fig. 5.2 Reduction of late phase fluorescein angiographic localization of retinal thinning is also shown. Reproduced
leakage in a patient with non-proliferative macular telangi- from Charbel Issa, P., R. P. Finger, et al. Monthly ranibi-
ectasia type 2 before (left column) and after (right column) zumab for nonproliferative macular telangiectasia type 2:
three intravitreal injections of ranibizumab in monthly a 12-month prospective study. Am J Ophthalmol (2011)
intervals. Topographic (green coding) and cross-sectional 151(5): 876-886 e871, with permission of Elsevier
Fig. 5.3 Longitudinal structure–function correlation in (enlarged cut-outs of the middle row) show corresponding
macular telangiectasia type 2 treated with monthly injec- SD-OCT images superimposed on back-tilted infrared
tions of ranibizumab over 1 year. Retinal sensitivity of the cSLO images and the same functional map as shown in the
individual testing points is color coded. Shown are images upper row. A defect in the photoreceptor layer developed
obtained at baseline (left column), after 6 (middle column) over time (panels e, h) despite normalization of the vascu-
and 12 monthly treatments (right column). The first row lar changes visible on angiography. Damage of the outer
shows functional maps superimposed on late-phase fluo- retina was associated with a strong loss of retinal sensitiv-
rescein angiography images. Treatment considerably ity (f, i). Visual acuity at baseline was 20/40 and remained
reduced angiographic leakage (a–c). However, an absolute unchanged over the study period despite the pronounced
scotoma (open red rectangles) developed at the comple- loss of paracentral visual function (reprint from Charbel
tion of the study period (c). The second and third row Issa et al. 2010 under a Creative Commons license)
Fig. 5.4 Long-term microperimetric follow-up in patients follow-up. An absolute scotoma developed in only 2/8 of
with macular telangiectasia type 2 who received monthly the fellow eyes. Visual acuity at each time point is provided
injections of ranibizumab in one eye during the first year of in the right lower corner of each microperimetry image.
the observational period. At baseline, no patient presented Time between baseline and last follow-up is shown on the
with an absolute scotoma. In the treated eye, an absolute right. One patient was excluded from functional assess-
scotoma (red) was present in 4/8 patients after 12 months ment due to psychogenic vision loss (reprint from Kupitz
(i.e., 1 month after 12 injections) and in 7/8 patients at last et al. 2015, with permission of Wolters Kluwer)
treated eyes may have been due to selection bias Visual stabilization as well as deterioration
or the small sample size, these results suggest that has been reported after argon laser
VEGF-inhibition has no long-term benefit in photocoagulation in stage 5 MacTel type 2 (Lee
patients with non-proliferative MacTel type 2. 1996; Watzke et al. 2005). Although this treat-
There might be exceptions from this conclu- ment approach seems to be relatively safe in
sion for well-controlled and short-term use of terms of recurrence of the membranes, large
VEGF-inhibition for foveal detachments associ- subsequent parafoveal scars may severely inter-
ated with MacTel type 2. Maia Jr. and coworkers fere with reading ability. However, this functional
published an observation after intravitreal beva- outcome measure was not reported.
cizumab injection in three non-proliferative eyes Photodynamic therapy (PDT) with vertepor-
with foveal detachment on OCT analysis (Maia fin has been shown to be beneficial for the treat-
et al. 2007). The foveal anatomy regained a nor- ment of subfoveal choroidal neovascularization
mal configuration and visual acuity significantly secondary to age-related macular degeneration
increased after a single intravitreal injection of and other diseases. The largest series, a retro-
bevacizumab. spective analysis, encompassed seven eyes of
six patients. Patients received on average 2.4
treatments and mean follow-up after the last
5.3 Proliferative Stage treatment was 21 months. Median initial and
of Macular Telangiectasia final visual acuity was 20/80. More than two
Type 2 lines decrease or increase in visual acuity was
observed in one eye each while the other five
Development of secondary neovascularizations eyes remained stable. There is the theoretical
represents a major cause for severe vision loss concern that the photosensitizing drug may
in MacTel type 2 (Gass and Blodi 1993; leak out of the retinal vessels in the macula and
Engelbrecht et al. 2002). Therapeutic approaches is potentially associated with adverse effects
including focal laser photocoagulation (Lee (Potter et al. 2002; Hussain et al. 2005). Indeed,
1996; Watzke et al. 2005), photodynamic ther- Shanmugam et al. showed atrophy of the retinal
apy alone (Potter et al. 2002, 2006; Hershberger pigment epithelium corresponding to the size
et al. 2003; Snyers et al. 2004; Hussain et al. of the laser spot (Shanmugam and Agarwal
2005; Shanmugam and Agarwal 2005), or com- 2005). However, others did not observe compli-
bined with intravitreal injection of triamcino- cations attributed to the treatment (Potter et al.
lone (Smithen and Spaide 2004), transpupillary 2002, 2006).
thermotherapy (Shukla et al. 2004), posterior The use of intravitreal applicationofbevaci-
juxtascleral administration of steroids (Eandi zumab for proliferative MacTel type 2 was first
et al. 2006), and recently, intravitreal injection described by Jorge and coworkers (2007). A
of VEGF-inhibitors (Charbel Issa et al. 2007; single injection in their patient presenting with
Jorge et al. 2007; Maia et al. 2007; Mandal et al. an extrafoveal membrane resulted in the absence
2007; Shanmugam et al. 2007) have been tried of signs of activity and significant improvement
to limit the consequences of this complicated of visual acuity within a follow-up period of 6
disease course. Park and coworkers found little months. Similar observations of significant
change in the size of the fibrovascular tissue functional and anatomical improvement in a
over time and consequently questioned the use- proliferative disease stage were reported in indi-
fulness for treatment in stage 5 eyes (Park et al. vidual patients and case series with or without
1996). Therefore, interventions appear to be accompanying foveal detachment (Roller et al.
most beneficial in early and active proliferative 2011; Mandal et al. 2007; Shanmugam et al.
disease stages (i.e., while a neovascular mem- 2007). In a patient with a larger subfoveal mem-
brane grows) before the development of fibrotic brane, no increase in visual acuity despite ana-
membranes. tomical improvement was achieved, again
Fig. 5.5 Effects of intravitreal bevacizumab therapy in real bevacizumab, the hemorrhage had been resorbed and
neovascular MacTel type 2. (a) Findings at baseline visit: leakage was reduced. The foveal center after treatment
Foveal hemorrhage (left) from an actively leaking neovas- shows only a relative defect and visual acuity has
cular membrane (middle; top, early phase; bottom, late improved from 20/80 to 20/50. Reproduced from Charbel
phase). On microperimetry testing, there is a large abso- Issa, P., M. C. Gillies, et al. (2013). “Macular telangiecta-
lute scotoma which encompasses the foveal center (right). sia type 2.” Prog Retin Eye Res 34: 49-77, with permis-
(b) Four months after one single treatment with intravit- sion of Elsevier
suggesting that early intervention seems to be

essential (Charbel Issa et al. 2007). In another 5.4 Conclusions from Clinical
retrospective case series, 16 treatment-naïve Observations
eyes of 16 patients were treated with intravitreal
ranibizumab or bevacizumab monotherapy In summary, patients with MacTel type 2 who
(Narayanan et al. 2012). A mean of 1.9 injec- develop an active neovascular membrane and
tions (range, 1–3 injections) were needed during recent loss in visual function appear to benefit
a mean follow-up time of 12 months (range, from timely therapy with intravitreally applied
3–43 months). Mean visual acuity improved sig- VEGF-antagonists. There is yet insufficient evi-
nificantly from 20/120 to 20/70. Figure 5.5 dence for following specific treatment regimens.
shows the functional and morphological disease However, a single initial injection with subse-
course under anti-VEGF treatment in a right eye quent treatment on a pro re nata basis may be suf-
of a patient with proliferative MacTel type 2 ficient, similar to the current approach in myopic
over a period of 2 months. choroidal neovascularization.
There are two reports on subretinal surgery The situation appears different in patients with
with removal of subfoveal vascular membranes non-proliferative disease stages. If MacTel type 2
in two eyes with MacTel type 2 (Berger et al. is a progressive neurodegenerative disease with
1997; Davidorf et al. 2004). Due to the adherence secondary changes of macular vessels due to
of the membranes to the neurosensory retina, altered VEGF levels, the long-term success of
removal was obviously difficult and subsequent treatments targeting VEGF may be limited. Indeed,
visual outcome was poor. long-term observations after monthly intravitreal
VEGF-inhibition over one year were unfavorable, References

with an even worse functional and morphological
outcome in treated versus untreated fellow eyes. Alldredge CD, Garretson BR. Intravitreal triamcinolone
for the treatment of idiopathic juxtafoveal telangiecta-
Müller cells have been implicated in the sis. Retina. 2003;23(1):113–6.
pathophysiology of MacTel type 2. The involve- Arevalo JF, Sanchez JG, et al. Indocyanine-green-
ment of Müller cells in supporting both neurons mediated photothrombosis (IMP) with intravitreal tri-
and vasculature suggests that their dysfunction amcinolone acetonide for macular edema secondary to
group 2A idiopathic parafoveal telangiectasis without
may be a central, although not necessarily pri- choroidal neovascularization: a pilot study. Graefes
mary, feature of MacTel type 2, which is charac- Arch Clin Exp Ophthalmol. 2007;245(11):1673–80.
terized mainly by dysfunction of these two retinal Berger AS, McCuen 2nd BW, et al. Surgical removal of
elements. Two clinicopathological studies of subfoveal neovascularization in idiopathic juxtafoveo-
lar retinal telangiectasis. Retina. 1997;17(2):94–8.
eyes from patients with MacTel type 2 suggested Cakir M, Kapran Z, et al. Optical coherence tomography
a depletion of Müller cells (Powner et al. 2010, evaluation of macular edema after intravitreal triam-
2013) and, thus, support this hypothesis. cinolone acetonide in patients with parafoveal telangi-
Disease characteristics similar to MacTel type 2 ectasis. Eur J Ophthalmol. 2006;16(5):711–7.
Charbel Issa P, Holz FG, et al. Findings in fluorescein angi-
have recently been shown in an animal model with ography and optical coherence tomography after intra-
conditional Müller cell ablation (Shen et al. 2012). vitreal Bevacizumab in Type 2 idiopathic macular
Impairment of Müller cells resulted in upregula- telangiectasia. Ophthalmology. 2007a;114(9):1736–42.
tion of VEGF and photoreceptor degeneration. Charbel Issa P, Scholl HPN, et al. Short-term effects of
intravitreal bevacizumab in Type II idiopathic macular
VEGF-inhibition attenuated vascular alterations telangiectasia. Retin Cases Brief Rep. 2007b;1(4):189–
that occurred after Müller cell death, but had no 91. doi:10.1097/ICB.0b013e31802efef2.
effect on photoreceptor death. Because VEGF has Charbel Issa P, Finger RP, et al. Eighteen-month follow-up
a role in the maintenance of cones (Kurihara et al. of intravitreal bevacizumab in type 2 idiopathic macu-
lar telangiectasia. Br J Ophthalmol. 2008;92(7):941–5.
2012) and Müller cells (Wurm et al. 2008), there is Charbel Issa P, Scholl HP, et al. Macular full-thickness and
the possibility of a protective upregulation of lamellar holes in association with type 2 idiopathic
VEGF in the retina of patients with MacTel type 2. macular telangiectasia. Eye. 2009;23(2):435–41.
Moreover, the rather high rate of incident neovas- Charbel Issa P, Troeger E, et al. Structure-function correlation
of the human central retina. PLoS One. 2010;5(9):e12864.
cular membranes (four out of nine) might also sug- Charbel Issa P, Finger RP, et al. Monthly ranibizumab for
gest a rebound effect with higher VEGF levels after nonproliferative macular telangiectasia type 2: a
VEGF suppression with subsequent development 12-month prospective study. Am J Ophthalmol.
of a neovascular membrane (Kupitz et al. 2015). 2011;151(5):876–86. e871.
Charbel Issa P, Gillies MC, et al. Macular telangiectasia
Hence, prolonged VEGF-inhibition in non-prolif- type 2. Prog Retin Eye Res. 2013;34:49–77.
erative disease stages might predispose to a worse Chew EY, Clemons TE, et al. Ciliary neurotrophic factor
functional and morphological outcome and neuro- for macular telangiectasia type 2: results from a phase
protective treatment strategies might be a more 1 safety trial. Am J Ophthalmol. 2015;159(4):659–66.
e651.
promising treatment approach (Chew et al. 2015). Chopdar A. Retinal telangiectasis in adults: fluorescein
angiographic findings and treatment by argon laser. Br
J Ophthalmol. 1978;62(4):243–50.
Davidorf FH, Pressman MD, et al. Juxtafoveal telangiec-
Compliance with Ethical Requirements
tasis-a name change? Retina. 2004;24(3):474–8.
Informed Consent and Animal Studies dis- De Lahitte GD, Cohen SY, et al. Lack of apparent short-term
benefit of photodynamic therapy in bilateral, acquired,
closures are not applicable to this review. parafoveal telangiectasis without subretinal neovascular-
The authors have received research grants ization. Am J Ophthalmol. 2004;138(5):892–4.
from the Lowy Medical Research Institute Eandi CM, Ober MD, et al. Anecortave acetate for the
as a clinical center of the MacTel research treatment of idiopathic perifoveal telangiectasia: a
pilot study. Retina. 2006;26(7):780–5.
consortium, and a research grant from Engelbrecht NE, Aaberg Jr TM, et al. Neovascular mem-
Novartis for an investigator initiated trial branes associated with idiopathic juxtafoveolar telan-
on anti-VGEF therapy in MacTel type 2. giectasis. Arch Ophthalmol. 2002;120(3):320–4.
Finger RP, Charbel Issa P, et al. Reading performance is neovascularization secondary to macular telangiecta-
reduced due to parafoveal scotomas in patients with sia type 2. Retina. 2012;32(10):2001–5.
macular telangiectasia type 2. Invest Ophthalmol Vis Park D, Schatz H, et al. Fibrovascular tissue in bilateral
Sci. 2009;50(3):1366–70. juxtafoveal telangiectasis. Arch Ophthalmol. 1996;
Friedman SM, Mames RN, et al. Subretinal hemorrhage 114(9):1092–6.
after grid laser photocoagulation for idiopathic juxta- Park DW, Schatz H, et al. Grid laser photocoagulation for
foveolar retinal telangiectasis. Ophthalmic Surg. macular edema in bilateral juxtafoveal telangiectasis.
1993;24(8):551–3. Ophthalmology. 1997;104(11):1838–46.
Gass JD, Blodi BA. Idiopathic juxtafoveolar retinal telan- Potter MJ, Szabo SM, et al. Photodynamic therapy of a
giectasis. Update of classification and follow-up study. subretinal neovascular membrane in type 2A idio-
Ophthalmology. 1993;100(10):1536–46. pathic juxtafoveolar retinal telangiectasis. Am
Gaudric A, Ducos de Lahitte G, et al. Optical coherence J Ophthalmol. 2002;133(1):149–51.
tomography in group 2A idiopathic juxtafoveolar retinal Potter MJ, Szabo SM, et al. Photodynamic therapy for
telangiectasis. Arch Ophthalmol. 2006;124(10):1410–9. subretinal neovascularization in type 2A idiopathic
Gillies MC, Zhu M, et al. Familial asymptomatic macular juxtafoveolar telangiectasis. Can J Ophthalmol. 2006;
telangiectasia type 2. Ophthalmology. 2009;116(12): 41(1):34–7.
2422–9. Powner MB, Gillies MC, et al. Perifoveal Müller cell
Heeren TF, Holz FG, et al. First symptoms and their age depletion in a case of macular telangiectasia type 2.
of onset in macular telangiectasia Type 2. Retina. Ophthalmology. 2010;117(12):2407–16.
2013;34(5):916–9. Powner MB, Gillies MC, et al. Loss of Muller’s cells and
Hershberger VS, Hutchins RK, et al. Photodynamic ther- photoreceptors in macular telangiectasia type 2.
apy with verteporfin for subretinal neovascularization Ophthalmology. 2013;120(11):2344–52.
secondary to bilateral idiopathic acquired juxtafoveo- Roller AB, Folk JC, et al. Intravitreal bevacizumab for
lar telangiectasis. Ophthalmic Surg Lasers Imaging. treatment of proliferative and nonproliferative type 2
2003;34(4):318–20. idiopathic macular telangiectasia. Retina. 2011;31(9):
Hussain N, Das T, et al. Bilateral sequential photody- 1848–55.
namic therapy for sub-retinal neovascularization with Shanmugam MP, Agarwal M. RPE atrophy following
type 2A parafoveal telangiectasis. Am J Ophthalmol. photodynamic therapy in type 2A idiopathic parafo-
2005;140(2):333–5. veal telangiectasis. Indian J Ophthalmol. 2005;
Hutton WL, Snyder WB, et al. Focal parafoveal retinal 53(1):61–3.
telangiectasis. Arch Ophthalmol. 1978;96(8):1362–7. Shanmugam MP, Mythri HM, et al. Intravitreal
Jorge R, Costa RA, et al. Intravitreal bevacizumab bevacizumab for parafoveal telangiectasia-associated
(Avastin) associated with the regression of subretinal choroidal neovascular membrane. Indian
neovascularization in idiopathic juxtafoveolar retinal J Ophthalmol. 2007;55(6):490–1.
telangiectasis. Graefes Arch Clin Exp Ophthalmol. Shen W, Fruttiger M, et al. Conditional Müller cell abla-
2007;245(7):1045–8. tion causes independent neuronal and vascular pathol-
Kupitz EH, Heeren TFC et al. (2015). Long-term poor ogies in a novel transgenic model. J Neurosci.
outcome of anti-VEGF therapy in non-proliferative 2012;32(45):15715–27.
macular telangiectasia type 2. Retina [in press]. Shukla D, Singh J, et al. Transpupillary thermotherapy for
Kurihara T, Westenskow PD, et al. Targeted deletion of subfoveal neovascularization secondary to group 2A
Vegfa in adult mice induces vision loss. J Clin Invest. idiopathic juxtafoveolar telangiectasis. Am
2012;122(11):4213–7. J Ophthalmol. 2004;138(1):147–9.
Lee BL. Bilateral subretinal neovascular membrane in idio- Smithen LM, Spaide RF. Photodynamic therapy and intra-
pathic juxtafoveolar telangiectasis. Retina. 1996;16(4): vitreal triamcinolone for a subretinal neovasculariza-
344–6. tion in bilateral idiopathic juxtafoveal telangiectasis.
Maia Jr OO, Bonanomi MT, et al. Intravitreal bevaci- Am J Ophthalmol. 2004;138(5):884–5.
zumab for foveal detachment in idiopathic perifoveal Snyers B, Verougstraete C, et al. Photodynamic therapy of
telangiectasia. Am J Ophthalmol. 2007;144(2):296–9. subfoveal neovascular membrane in type 2A idio-
Mandal S, Venkatesh P, et al. Intravitreal bevacizumab pathic juxtafoveolar retinal telangiectasis. Am
(Avastin) for subretinal neovascularization secondary to J Ophthalmol. 2004;137(5):812–9.
type 2A idiopathic juxtafoveal telangiectasia. Graefes Toy BC, Koo E, et al. Treatment of nonneovascular idio-
Arch Clin Exp Ophthalmol. 2007;245(12):1825–9. pathic macular telangiectasia type 2 with intravitreal
Martinez JA. Intravitreal triamcinolone acetonide for ranibizumab: results of a phase II clinical trial. Retina.
bilateral acquired parafoveal telangiectasis. Arch 2012;32(5):996–1006.
Ophthalmol. 2003;121(11):1658–9. Watzke RC, Klein ML, et al. Long-term juxtafoveal reti-
Matt G, Sacu S, et al. Thirty-month follow-up after intravit- nal telangiectasia. Retina. 2005;25(6):727–35.
real bevacizumab in progressive idiopathic macular tel- Wurm A, Pannicke T, et al. Glial cell-derived glutamate
angiectasia type 2. Eye (Lond). 2010;24(10):1535–41. mediates autocrine cell volume regulation in the retina:
Narayanan R, Chhablani J, et al. Efficacy of anti-vascular activation by VEGF. J Neurochem. 2008;104(2):
endothelial growth factor therapy in subretinal 386–99.
Diabetic Retinopathy
6
Focke Ziemssen and Hansjürgen T. Agostini
Nevertheless, blindness is not only the fear No.

6.1 Nomenclature, 1 in persons affected. Diabetic retinopathy still is a
Classification, growing disease. The current epidemic of diabetes
and Epidemiology and growing numbers in the world are linked to a
changing lifestyle—in the Western world as well as
6.1.1 Incidence and Prevalence in the developing countries (Scanlon et al. 2013).
The International Diabetes Federation (IDF) pre-
Diabetic retinopathy (DR) is still the leading dicts a further rise from 382 million in 2013 to 552
cause of vision loss in the working-aged adults million people affected in 2030. Diabetes accord-
worldwide (Yau et al. 2012), thereby causing a ing to most recent estimates affects 8.3 % of adults.
significant impairment in utility and quality of Yet, with 175 million persons (45.8 %) currently
life. Forty years ago, persons with diabetes were undiagnosed, a large amount of people are pro-
25 times more likely to become blind than people gressing globally towards complications while
without diabetes (Kahn and Hiller 1974). being unaware of their disease (Beagley et al.
However, over the last decades, lower incidences 2014). An estimated 83.8 % of all undiagnosed
of diabetic retinopathy (DR) and rates of progres- cases are in low- and middle-income countries. The
sion to proliferative retinopathy (PRD) were seen lifetime risk of developing diabetes is approx.
than historically (Ding and Wong 2012). The risk 30 %, but decisively influenced by the body weight
of severe visual loss has slightly decreased in the (Narayan et al. 2003, 2007).
Western countries, which may reflect better con-
trol of glucose, blood pressure, and serum lipids
as well as earlier diagnosis (Kempen et al. 2004). 6.1.2 Worldwide Incidence
and Prevalence
F. Ziemssen (*) Type 1 diabetes shows a heterogeneous distribu-

Center for Ophthalmology, Eberhard-Karl tion, being more common in the Northern coun-
University Tübingen, Schleichstr. 12, tries. Globally, the age-standardized prevalence of
Tübingen 72076, Germany
e-mail: focke.ziemssen@med.uni-tuebingen.de
DR is assumed to be 10 % for the vision-threatening
disease, with diabetic macular edema and/or pro-
H.T. Agostini
Eye Center, Albert-Ludwigs-University Freiburg,
liferative DR each accounting for approx. 7 %
Killianstr. 5, Freiburg im Breisgau 79106, Germany (Yau et al. 2012). Diabetes still is a common cause
e-mail: hansjuergen.agostini@uniklinik-freiburg.de of blindness, especially in the working age adults.

90 F. Ziemssen and H.T. Agostini
Even in developed countries, DR is responsible for nopathy increased to 60 % after 10 years and
10 % of new blindness (Congdon et al. 2004; even 80 % after 15 years. It is noteworthy that the
Finger et al. 2011), in dependence of race and risk of severe retinopathy seems to be reduced in
living conditions. One in 11 adults has diabetes, more recent studies (LeCaire et al. 2013).
one of two adults with diabetes is undiagnosed The Los Angeles Latino Eye Study (LALES)
(IDF 2015). Conservative estimations lead to the described a prevalence of 18 % of proliferative
rate of one person becoming blind every minute. diabetic retinopathy, with no difference between
A genetic predisposition has been shown to those with type 1 vs. type 2 diabetes (Varma et al.
affect the lifetime risk of retinopathy. After con- 2004). A pooled analysis in China reported a
trolling for other independent risk factors some prevalence of DR of 29 % for the rural areas. The
polymorphisms have been found to be either pro- frequency was slightly higher in the Northern
tective or deleterious (Grassi et al. 2012; Schwartz region (Liu et al. 2012). In the Middle East, most
et al. 2013). patients presented with mild nonproliferative DR
The Wisconsin Epidemiologic Study of (up to 65 %), but the proliferative disease varied
Diabetic Retinopathy (WESDR) underlined the between 2.3 and 10 % among affected persons
influence of the duration of disease on the devel- (Zabetian et al. 2013). These discrepancies point
opment of DR (Klein et al. 1984, 2009). The inci- also to the higher risk of some racial groups.
dence of DR increases with the duration of The Gutenberg Health study found a total
diabetes (Fig. 6.1). While almost none of the prevalence of DR of 24 % in a German cross-
affected persons with type 1 diabetes had macu- sectional cohort (Raum et al. 2014). The rate of
lar edema within the first 5 years of onset, reti- diabetes was 7.5 % within the age between 35
Fig. 6.1 Influence of diabetes duration on frequency of diabetic eye disease (Yau et al. 2012). VTDR = vision-threatening
diabetic retinopathy
6 Diabetic Retinopathy 91
and 74 years of age. Only 5 % of these suffered Retinal capillary microaneurysms (MAs) are
from sight-threatening DR. Other studies indi- usually the first visible sign of diabetic retinopa-
cate the frequent unawareness of disease, in early thy. The characteristics of MAs are the appear-
phases similarly to vision-threatening stages ance of round deep-red dots, a size of 15–60 μm
(Huang et al. 2009; Silva et al. 2010). and a spatial preference for the macula, located in
Persistent severe visual loss was an infrequent the inner nuclear layer and the capillary network
observation in the Early Treatment of Diabetic (Byeon et al. 2012). The saccular outpouchings
Retinopathy Study (ETDRS) and most com- of the capillary wall are a consequence of a loss
monly caused by vitreous or preretinal hemor- of pericytes and precede capillary closure
rhage, followed by macular edema or macular (Fig. 6.2). There is recent evidence that the struc-
pigmentary changes (Fong et al. 1999). The influ- tural variability is higher when examined by
ence of modifiable risk factors for progression adaptive optics imaging (Dubow et al. 2014).
makes it difficult to project the exact incidences Mechanistically, a weakness of the capillary wall
of the advanced stages or blindness for the future. from the early loss of pericytes and changes of
the vascular endothelium or the intraluminal
pressure are discussed as contributing factors.
6.1.3 Prevention and Health Costs Rupture of larger aneurysms can lead to extensive
intraretinal or vitreous hemorrhage.
Even in the absence of other major complications In fluorescein angiography (FLA), MAs
DR has a major impact on the quality of life exhibit bright hyperfluorescence, distinguishable
(Alcubierre et al. 2014). Early detection and treat- from small hemorrhages (Hellstedt et al. 1996;
ment of DR can avoid and considerably reduce the Ito et al. 2013; Jalli et al. 1997; Kohner and
economic costs of the most frequent diabetic com- Henkind 1970; Schiffman et al. 2005; Wang et al.
plication. Reducing the burden and follow-up 2012a). Although increased density is associated
costs, the effectiveness of screening and treatment with an area of leakage, the presence or absence
is dependent on the subtype of DR (Javitt and of MAs alone is not of clinical significance and
Aiello 1996). Therefore, prevention programs therefore does not in itself warrant the need for
aimed at improving eye care for diabetic persons the invasive FLA procedure (Fig. 6.3).
result in substantial federal budgetary savings and Nevertheless, it is important to be aware of the
are highly cost-effective health investments for prognostic value of the turnover occurrence of
society. Besides the improvement of quality of life, MAs, respectively (Nunes et al. 2009). Regarding
the main benefit is the prevention of further medi- the dynamic nature of MAs, it is noteworthy that
cal costs (Pelletier et al. 2009). A Swedish study older or regressing MAs are not necessarily
calculated annual DR-related costs of 106.000€ identified in FLA (Goatman et al. 2003; Leicht
per 100.000 inhabitants (Heintz et al. 2010). et al. 2014). When compared to color photogra-
phy, FLA examinations allow a modest increase
in sensitivity of DR (DCCT Research Group
6.2 Clinical Entities 1987; Agardh and Cavallin-Sjoberg 1998)
(Table 6.1). Therefore, an implementation of the
6.2.1 Nonproliferative Diabetic procedure in addition to biomicroscopy remains
Retinopathy reserved for predefined scenarios (Banerjee et al.
2007; Khalaf et al. 2007; Olsen et al. 2014).
Vascular changes and exudation are the hallmark Hemorrhages within the macular area usu-
of early diabetic retinopathy. On ophthalmo- ally originate from ruptured small vessels and
scopic examination, the characteristic features of MAs. Small pinpoint bleedings occur early in
nonproliferative diabetic retinopathy (NPDR) are maculopathy. In contrast, hemorrhages in the
microaneurysms (MAs), intraretinal hemor- periphery are more frequently an indication of
rhages, and hard exudates. other concomitant diseases or extensive isch-
Fig. 6.2 Retinal microaneurysms appear as round red dots (Dubow et al. 2014)
Fig. 6.3 Clinical image of microaneurysms (right: arrow in fluorescein angiogram)
Table 6.1 Rational for the performance of FLA

Scenarios to perform fluorescein angiography (FLA)
Essential Helpful Not appropriate
• Clarification of unexplained vision loss • Differential diagnosis in • Screening at no or minimal DR
unusual findings
• Identification of capillary nonperfusion • Support of laser treatment
emia. In cases of preretinal bleeding, PRD 2014; Lombardo et al. 2013). On histopathologi-
should be excluded. cal examination, endothelial proliferation and
Hard exudates can represent the extracellular pericyte damage in the small vessels (Ashton
accumulation of lipids, proteins, and lipopro- 1958) are more and more accompanied by vaso-
teins. There is a huge variation in size and struc- regression (Cunha-Vaz et al. 2014b).
ture, from tiny precipitates and circinate atolls to Initially smaller patches of acellular capillaries
confluent arrangements and large lipid plaques. can become confluent over time. Tortuous cluster
Hard exudates can indicate the extent and loca- of MAs and intraretinal microvascular abnormali-
tion of vascular leakage and transiently increase, ties (IRMA) develop, often coinciding. The areas
if edema diminishes. of capillary nonperfusion enlarge as the disease
In the course of the disease, the closure of progresses (Sim et al. 2013a), while associated
retinal capillaries is a major feature and reflects with retinal thinning (Sim et al. 2014). Although
retinal ischemia. While there are early preclinical visual acuity can be preserved during mild grades
signs such as dysfunction of photoreceptors of ischemia, more severe stages are associated
(Verma et al. 2012) and/or vascular inflammation with a loss of visual acuity, especially if the papil-
(Daley et al. 1987; Lung et al. 2012), capillary lomacular nerve fiber bundle is affected (Arden
obliteration and nonperfusion are the hallmarks and Sivaprasad 2011; Sim et al. 2013b) (Fig. 6.4).
of the advanced diabetic retinopathy (Li et al. Rarefication of vascular branches leads to the
Fig. 6.4 Progression of foveal ischemia (first row: perifoveal capillary in green after 2 years, second row: perifoveal
capillaries in red, and third row: merged angiograms) shown by Sim et al. (2013a)
Fig. 6.5 Differential layers of the perifoveal morphology can be detected by OCT Angiography (Courtesy: Dr. Scott
Lee, Oakland; Prototype Zeiss SD-Angio-OCT)
typical appearance of pruning (ETDRS Research et al. 2012). Proliferative diabetic retinopathy
Group 1991b). (PDR) is much more frequently seen in younger
Venous beading and loops, intraretinal micro- persons with type 1 diabetes (Klein et al. 1992).
vascular changes and areas of nonperfusion indi- Hyperglycemia, longer duration of diabetes,
cate an increased risk of progression of the and more severe retinopathy at baseline were
retinopathy to more severe stages (Benson et al. associated with an increased 4-year risk of
1988) (Fig. 6.5). developing PDR.
The accompanying inflammation and chronic
disruption of the blood–retina barrier contribute
6.2.2 Proliferative Diabetic to the reduced perfusion (Klaassen et al. 2013;
Retinopathy Lang 2013; Silva et al. 2007) (Fig. 6.6). Retinal
nonperfusion was reported to affect predomi-
Proliferative changes, affecting approximately nantly the midperiphery (Cardillo et al. 1987).
7 % of persons with DR, are considered to be However, it has to be mentioned that wide-angle
mainly induced by ischemia (Liu et al. 2013; imaging nowadays does allow a much better
Miller et al. 2013; Rodrigues et al. 2013; Yau assessment of the far periphery. There have been
Fig. 6.6 Subtle NVD can be delineated by fluorescein angiography
different hypothesis that a disc-centered pattern Initially, new vessels may be very subtle (Li
might be a sign of a more important role of hyper- et al. 2010). Wheel-like networks, an irregularity
tension, while peripheral patterns might be attrib- in shape or adjacent hemorrhages are common.
uted to the presence of additional systemic The growth rate of neovessels is very variable.
diseases. Chronic ischemia contributes in turn to Regarding concomitant proliferative vitreopathy
a localized low-grade inflammatory response or fibrovascular proliferations, early membranes
within the vessels, subsequent migration, and can sometimes be overlooked or underestimated
stimulation of immunogenic cells in the tissue due to their translucency in the early phase. OCT
(Sebag 1993). imaging has a high sensitivity to better detect a
While VEGF is the most prominent proangio- fibrovascular reaction of the posterior pole.
genic factor (Ishida et al. 2003), other growth fac- If in doubt, FLA should be performed to con-
tors like TGFβ are important for the clinically firm the diagnosis and determine the proliferative
relevant processes of membrane formation and activity. FLA is also superior to fundoscopy dif-
fibrovascular contraction (Deissler et al. 2006; ferentiating between IRMAs and a neovascular-
Nawaz et al. 2013; Sohn et al. 2012). ization that is located outside the disc area,
The risk for conversion to PDR increases with classified as neovascularization elsewhere (NVE).
the severity of NPDR. Peripheral lesions seen by If fibrovascular membranes contract and/or
200° ultrawide field imaging might be predictive vitreous detachment occurs, hemorrhages are
(Silva et al. 2015). However, due to the transient more frequently seen, otherwise important indi-
nature of hemorrhages and vessel abnormalities, cators of local ischemia. Vitreous hemorrhage is
it is not possible in all cases to correctly identify an important sight-threatening complication in
the severity of NPDR based on fundoscopy alone DR. Depending on the extent and duration, it
(Danis and Davis 2008). does not only cause serious loss of vision, but can
Two out of three subjects with PDR, the neo- also restrict the diagnostic possibilities (fundus
vascularization is first detected on or adjacent to details) and delay initiation of panretinal photo-
the optic disc (DRS Group 1981b; Taylor and coagulation (PRP).
Dobree 1970). Such neovascularization of the The contraction of fibrovascular proliferations
optic disc (NVD) starts as fine loops or networks can displace the retina or cause tractional detach-
of vessels on the surface of the disc or bridging ment (Bresnick et al. 1979; Kampik et al. 1981).
across the physiologic cup (Fig. 6.6). The extent of subretinal fluid under the detached
retina and macular involvement are the most

important predictors of function (Panozzo et al.
2004). However, subsequent thinning of the ret-
ina due to malperfusion of the detached retina as
well as the high risk of re-proliferation of epireti-
nal membranes limits the functional and anatom-
ical outcome, even after vitreoretinal surgery
with silicon tamponade (Boynton et al. 2015).
6.2.3 Macular Disease
Macular edema (DME) is defined as retinal thick-

ening due to an accumulation of fluid. The leak-
age of fluid is caused by a disruption of the
blood–retinal barrier.
In former times, edema was detected with slit-
lamp microscopy and/or stereoscopic photogra-
phy. Nowadays, OCT is the diagnostic gold
standard, not only to test for the presence of
thickening, but to quantify retinal thickness and
allow follow-up examinations in the same retinal
location (Chalam et al. 2012; De et al. 2015;
Dmuchowska et al. 2014; Fiore et al. 2013;
Vujosevic et al. 2013) (Fig. 6.7). The introduc-
tion of OCT has reduced the need of the invasive
FLA examination, as it can detect already sub-
clinical macular edema, a precursor to significant
DME (Bressler et al. 2012; Pires et al. 2013).
Although retinal thickness does not directly
correlate with function, changes in retinal thick-
ness are of importance when dosing medical
treatment (Browning et al. 2007). The cautious
prognosis of functional outcome after therapy is
better delivered by evaluating the integrity of the
outer retinal layers (i.e., the photoreceptors)
(Maheshwary et al. 2010). In addition, spectral-
domain OCT imaging is able to visualize the vit-
reoretinal interface or indicate inflammatory
disease activity such as neurosensory detachment
(Murakami et al. 2013; Sonoda et al. 2013).
Fig. 6.7 OCT imaging allows classifying for different

morphology patterns: (a) focal or multifocal cystic edema
(b) non-focal capillary leakage (c) subretinal fluid (arrow)
(d) lipid plaques (e) atrophic edema (f) diffuse thickening
(g) ischemic atrophy
Although a consistent relationship between • Retinal thickening of an area of more than one
FLA and OCT has been described (e.g., petaloid disc area within a radius of one disc diameter
pattern and large pseudocysts within the outer around the fovea
nuclear and plexiform layers, honeycomb-like
pattern with involvement of the inner layers) The differentiation of focal and diffuse edema
(Bolz et al. 2009a; Byeon et al. 2012), the clinical is not without problems, as the terms are not used
relevance of the OCT findings has not been fully in a consistent way (Browning et al. 2008).
established in prospective trials, including such Physicians made use of different methods and
as subretinal fluid or hyperreflective dots. definitions. Often the angiographic pattern and
Previous analyses indicate that edema showing source of fluorescein leakage are described. Others
subretinal fluid have favorable outcome under grade on the base of extent and location of macu-
anti-VEGF therapy (Sophie et al. 2015). lar thickening, involvement of the center of the
The concept of clinically significant macular macula, and quantity and pattern of lipid exudates
edema (CSME) has been introduced by the or grade the edema with respect to the modality of
ETDRS Research Group (1985). The term com- the planned photocoagulation technique (Callanan
prises clinical findings associated with an et al. 2013; Funatsu et al. 2009; Ophir 2014).
increased risk of visual loss. Therefore, it is Although some scientists found a difference in
important to identify the area and localization of treatment efficacy in dependence of the chronicity
retinal edema and lipid deposits with respect to (Lee and Olk 1991), a general significance of the
the fovea. It should be remembered that the diag- duration is questionable. In accordance to most
nosis of CSME was based on binocular ophthal- definitions, every diffuse edema is a CSME, but in
moscopy or stereoscopic photographs. In the era contrast a CSME can be located either extrafove-
preceeding the resolution of SD-OCT, the images ally or be of minimum size. With regard to the
had to be stereoscopically recorded in order not treatment options, it is useful to distinguish edema
to overlook retinal thickening. Future studies with and without involvement of the fovea.
have to assess how OCT imaging changes the The amount of foveal ischemia is of great rele-
phenotyping (Bolz et al. 2009b). vance for clinical evaluation and treatment deci-
The definition of CSME included the follow- sion. Nevertheless, there is no uniform definition of
ing findings (Fig. 6.8): ischemic maculopathy. The classic diagnosis is
made based upon FLA examination. Most fre-
• Retinal thickening within a radius of 500 μm quently, an outage of the perifoveolar capillary
around the foveal center arcade and/or areas of capillary closure are consid-
• Lipid deposits within a radius of 500 μm ered to delimit the damage (Sim et al. 2013a; Zheng
around the fovea with an accompanying et al. 2014). The diameter and/or the area of the
thickening foveal avascular zone (FAZ) were often calculated
Fig. 6.8 Definition of clinically significant macular edema (CSME)

to quantify the extent of nonperfusion and the lim- velocity at which worsening occurs are both cor-
ited prognosis. The diameter of the regular FAZ is related to the risk of severe vision loss. Therefore,
often defined to lie between 400 and 500 μm, while exact fundoscopic examination in mydriasis and
DR can lead to diameters of more than 1000 μm correct classification of DR severity is clinically
(Bresnick 1983). In future, the loss of ganglion important (Fig. 6.10).
cells or the reduced integrity of photoreceptors Future classification systems should also include
might be more easily detected by OCT imaging the manifestations of DR that are visible on
(Byeon et al. 2009; Sim et al. 2014); the size of the SD-OCT imaging. The more accurately DR sever-
avascular zone may be imaged by Angio-OCT. ity is classified, the better the risk of subsequent
Vitreomacular traction reflects the involve- worsening can be predicted and treatment and
ment of the vitreous interface in DR (Haller et al. follow-up exams can be scheduled accordingly.
2010). These alterations of the vitreomacular
interface as well as epiretinal membranes can
induce retinal thickening and complicate the 6.2.5 Screening for Diabetic
treatment of DME (Buabbud et al. 2010; Retinopathy
Giovannini et al. 1999; Ophir et al. 2010)
(Fig. 6.9). As changes of the vitreomacular inter- As sight-threatening retinopathy can be asymp-
face are frequent after chronic edema as detected tomatic, screening has to be conducted in regu-
by SD-OCT, it is sometimes difficult to judge lar time intervals before the onset of symptoms
their respective contribution to visual deteriora- (Klein 2003). Young patients are able to com-
tion, respectively, the potential for visual acuity pensate for an edema by accommodation; subtle
improvement through surgical membrane peeling and thus subclinical DME is apparently not all
(Ghazi et al. 2007). that uncommon (Braun et al. 1995). The
unknown date of onset is an important reason
not to forget regular screening intervals in type
6.2.4 ETDRS Severity Score 2 diabetes. In prediabetes or even “nondiabetes”
DR was seen in up to 8 % of patients (DPP
DR often worsens in unequal steps described by Research Group 2007). Treatment for DR may
the ETDRS severity scale (ETDRS Research be 90 % effective in preventing severe vision
Group 1991a). The stage itself as well as the loss using current strategies, but fewer persons
Fig. 6.9 Retinal folds and focal detachment due to vitreoretinal traction by epiretinal membrane
ETDRS
severity
score 10 20 35 43 47 53 60 89
no DR DR questionable MA mild NPDR moderate NPDR moderate-to- severe NPDR PDR Diabetic Macular pregnancy
severe NPDR Edema
hard exudates hard exudates retinal
cotton-wool spots cotton-wool spots neovascularization
³ 1 venous loop
RH IRMA RH RH present severe ³ 1 quadrant severe ³ 2 quadrants severe ³ 4 quadrants

no MA
moderate ³ 4 quadrants
MA no MA MA severe ³ 1 quadrant severe ³ 4 quadrants

moderate ³ 4 quadrants
IRMA IRMA IRMA IRMA IRMA ³ 2 quadrants

questionable
VB VB questionable VB VB ³ 2 quadrants
Fig. 6.10 ETDRS severity scale. RH retinal hemorrhage, IRMA intraretinal microvascular abnormalities, VB venous
beading, adapted from the ETDRS Research Group (1991d). Future classifications will also include ocular coherence
tomography (OCT)
are referred for ophthalmic care than necessary initially based on observations from epidemio-
in accordance to current guidelines (Ferris 1993; logical studies. Following cost-utility analyses,
Paz et al. 2006). the rate of deterioration might not justify the
Different factors are discussed that could costs and efforts of yearly examinations (Vijan
explain the insufficient disease awareness of per- et al. 2000). Due to interindividual differences,
sons with diabetes that leads to an underuse of however, the time course of DR progression can
the healthcare system (Bressler et al. 2014; be fast, even within these first years of
Maberley et al. 2002; Maclennan et al. 2014). DR. Essential framework conditions include the
Primary care physicians are not always familiar duration of disease, the type of diabetes, and all
with the increasing incidence of DR during the cardiovascular risk factors for the individual
course of metabolic changes and therefore pre- person (Mehlsen et al. 2012). Type 1 diabetes
clude the implementation of treatment guidelines must not be mixed up with type 2 diabetes
(Looker et al. 2014; Witkin and Klein 1984). (Klonoff and Schwartz 2000). To achieve better
30–50 % of persons with diabetes still do not knowledge of the individual risk factors, a good
receive timely examinations, even within the communication between primary physicians and
framework of cost-free programs of NEI or NHS ophthalmologists is essential.
and despite extensive education of patients and Accurate and early diagnosis is of great impor-
physicians (Brechner et al. 1993). tance for the successful treatment of DR (Fong
The low utilization keeps most guideline et al. 2001). The stereoscopic capabilities of digi-
commissions from extending the screening inter- tal cameras are less important, since SD-OCT is
vals (Hutchins et al. 2012; Nam et al. 2011; much more sensitive to detect retinal thickening
Romero-Aroca et al. 2010; Yuen et al. 2010). (Nisic et al. 2014). Imaging and discussing the
Longer screening intervals might increase the obtained images with the persons screened might
risk of an irreversible loss of vision due to nonap- enhance the adherence to future examinations
pearance (Kristinsson et al. 1995). Recent experi- (Fonda et al. 2007). Some authors even argue in
ences showed that adherence did not improve favor of a smartphone ophthalmoscopy though
with longer control intervals (Misra et al. 2009). missing 5 % of the patients (Russo et al. 2015).
Sociocultural differences and individual risk pro- However, newer devices allow wide-angle
files should be considered when discussing the imaging under non-mydriatic conditions. Further
best interval for the next visit in the individual health care research studies have to prove whether
patient (Do and Eggleston 2011; Garay-Sevilla these methods allow a sensitive detection of DR,
et al. 2011; Gulliford et al. 2010). thus equivalent efficacy at reasonable costs (Nam
Nevertheless, some economists find it accept- et al. 2011). Telemedicine using wide-angle cam-
able to screen less and less frequently. The rec- eras might be of even greater use if the access to
ommended interval for eye examinations was ophthalmic care is limited.
6.3 Current Knowledge models (Kern and Engerman 1996; Ly et al.

on Pathophysiology 2014; Robinson et al. 2012). Several biochemical
pathways have been discussed to link hypergly-
Although hyperglycemia remains the key charac- cemia and microvascular complications (Ahsan
teristic of DR, the huge differences between type 2015; Kamoi et al. 2013; Ola et al. 2012; Tarr
1 and type 2 diabetes with regard to DR progres- et al. 2013; Zhang et al. 2012).
sion underline the multiple factors involved The structural features of the central retina, in
(Campos 2012) (Fig. 6.11). Increasing the serum particular loose intercellular contacts and the
level of glucose has been shown to be sufficient absence of Mueller cells are considered to explain
to initiate the early retinopathy changes in animal the predilection of the disease for the macular
Fig. 6.11 Effects of hyperglycemia (adopted from Lutz Hansen and Klaus Lemmen)
area (Cunha-Vaz 2004). At an early stage, the Among all factors involved, VEGF A is the
crosstalk of glial cells, pericytes, and endothelial most investigated and certainly an important
cells seems to be impaired (Antonetti et al. 2006). mediator of vascular permeability and prolifera-
Later, a thickening of the basement membrane, tive endothelial activity (Klaassen et al. 2013;
loss of pericytes, and an altered endothelium con- Pfeiffer et al. 1997). VEGF-induced breakdown
tribute to focal areas of hypoxia (Ashton 1974; of the blood–retinal barrier is mediated by a
Ejaz et al. 2008; Lutty 2013). The exact relation- change of occluding and claudin-1, suggesting
ship between vasoregression, primary neurode- the latter to be potentially the most relevant tight
generation and the impairment of neurovascular junction protein (Lang 2012). Other factors than
coupling, however, still needs to be clarified VEGF did not show such a strong effect on tran-
(Clermont and Bursell 2007; Hammes et al. sendothelial resistance (Deissler et al. 2013).
2011a). Mechanisms of neurodegeneration have This does, of course, not preclude any other—to
been identified, such as a loss of astrocytes, date uninvestigated factors—from being impor-
Müller cells, and retinal ganglion cells. tant additional effectors of DME formation.
The distinct inflammatory component of DR
is most clearly illustrated by the increased leuko-
cytes adhesion and emigration (Matsuoka et al. 6.4 Systemic Risk Factors
2007; Tang and Kern 2011). In addition, a marked in Diabetes
increase in inflammatory cytokines can be mea-
sured (Rangasamy et al. 2014). It is very likely 6.4.1 Risk Factors and Metabolic
that the proinflammatory steps are not only coin- Control
ciding phenomena, but part of sequential, inter-
dependent pathways. In this context, it is Intensive lifestyle interventions can not only pre-
noteworthy that Müller cells were found to vent or delay the onset of diabetes in high-risk
show inflammation-linked responses if exposed individuals (Karam and McFarlane 2011;
to diabetic retinal milieu (Kern 2007; Zhong Palermo et al. 2014; Tahrani et al. 2011), but have
et al. 2012). The intercellular adhesion molecule an important influence on the course of the dis-
ICAM-1, a transmembrane protein binding to ease and the incidence of complications (Klein
integrins, was attributed to increased leukostasis et al. 2002). Therefore, knowledge and manage-
(Ugurlu et al. 2013). ment of medical and vascular parameters in
Long-lived glycosylated proteins (AGE) con- persons with diabetes is of upmost importance
tribute to the generalized endothelial dysfunction (Kiire et al. 2013).
(Kandarakis et al. 2014; Stitt 2010; Yamagishi In type 1 diabetes, convincing data suggest a
et al. 2008). An increased transvascular passage direct relationship between the occurrence of DR
of macromolecules has been documented and and the severity as well as the duration of diabe-
resembles the pathogenic mechanism of diabetic tes (Hietala et al. 2010; Moss et al. 1994). Type 1
nephropathy and DR (Sander et al. 2003; diabetes is characterized by an autoimmune
Zandbergen et al. 2007). The oxygen consump- destruction of beta cells leading to a deficiency of
tion in the retina is very demanding. Toxic end- insulin. Very rarely DR becomes apparent before
products as peroxides, superoxides, and reactive 6 years after diagnosis of diabetes, but increases
oxygen species are released secondary to hyper- fast with longer diabetes duration (Klein et al.
glycaemia and the activation of alternative 1984). Puberty can act as an accelerator of com-
pathways (Madsen-Bouterse and Kowluru 2008; plications (Cho et al. 2014). Most guidelines rec-
Santos et al. 2011). The activation of protein ommend ophthalmic examination 5 years after
kinase C (PKC-β) and increased diacylglycerol the onset of type 1 diabetes.
are associated with vascular changes of different The precise time of onset is often much more
organs (Brownlee 2001). difficult to assess in type 2 diabetes, when an
asymptomatic phase precedes diagnosis. The rate 6.4.3 Control of Blood Sugar
of preexisting DR is not negligible (DPP Research
Group 2007). The variability of the type 2 diabe- Glycemic control is an important effector when
tes is much larger ranging from slight insulin thinking of microvascular complications (Sander
resistance to an absolute insulin deficiency. The et al. 1994). The evidence of its impact is sup-
majority of persons with type 2 diabetes are ported by epidemiologic studies and prospective
obese and the prevalence of type 2 diabetes clinical trials (Zhang et al. 2015).
increases with age. Increased glucose levels are not only risk
Gestational diabetes is a special form seen in factors for developing DR, but also relevant for
women at a preexisting predisposition or so far the progression to more advanced stages.
unknown diabetes of earlier onset. If it is the for- Traditionally, it was recommended to gener-
mer (and previous disease is unlikely) the risk of ally aim for a low target of glycemic control, thus
retinopathy does not seem to be significantly preferring a target of 6.5 % instead of 7 % HbA1c
increased (Gunderson et al. 2007). If DR is known, (ADA 2013). However, the ACCORD study
there is a high risk of worsening during pregnancy raised concerns that a very aggressive HbA1c
(DCCT Research Group 2000; Chew et al. 1995b; management might increase the risk of dementia
Hellstedt et al. 1997; Klein et al. 1990): persons and macrovascular complications in older
with diabetes who plan to become pregnant or are patients (Chew et al. 2014; Frank 2014). This is
in early pregnancy, should therefore undergo a the more important as patients with type 2 diabe-
comprehensive eye exam and be counseled about tes and retinopathy represent a subgroup at higher
the risk of progression. Follow-up visits should be risk for future cognitive decline (Hugenschmidt
arranged depending on the severity of the retinop- et al. 2014).
athy (Vestgaard et al. 2010). The DCCT and UKPDS showed that the
development and progression of DR can be
delayed when HbA1c is optimized (DCCT
6.4.2 Disease Duration Research Group 1995). Several studies indicated
that each 1 % increase in hemoglobin above 7 %
When assessing the risk of DR development, the raises the incidence of DR by 50 % and the pro-
duration of diabetes is an important parameter. gression by 50 % (LeCaire et al. 2013). A 35 %
The differences between type 1 and type 2 diabe- risk reduction can be achieved with every per-
tes have to be considered. centage point of HbA1c decreased (The Diabetes
In historic trials, DR was seen in 25 % of persons Control and Complications Trial/Epidemiology
with type 1 diabetes after 5 years. The rate further of Diabetes Interventions and Complications
increased to 60 % at 10 years and 80 % at 15 years Research Group 2000; Fullerton et al. 2014).
(Hammes et al. 2011b; LeCaire et al. 2013). In type There is no glycemic threshold when metabolic
2 diabetes, the rates of DR also rose in dependence control becomes irrelevant. Furthermore, there is
of the duration of diabetes. The percentages of DR evidence of an “HbA1c memory effect” meaning
in those patients taking insulin were found consid- that persons with good HbA1c management ben-
erably higher than in those on diet or oral drugs efit not only during the times of tight glycemic
(5 years: 40 % and 19 years: 84 % vs. 5 years: 24 % control, but also over the following decades (The
and 19 years: 53 %, respectively). Diabetes Control and Complications Trial/
There is only limited data available regarding Epidemiology of Diabetes Interventions and
the effects of changed treatment patterns. Complications Research Group 2000;
However, when appraising the different risk Hemmingsen et al. 2011; Kohner et al. 2001). It
factors of each population the background of has to be considered, however, that after initia-
each ethnic group has to be considered (Mazhar tion of strict glycemic control a transient worsen-
et al. 2011; West et al. 2001). ing has to be anticipated in particular in patients
with preexisting advanced DR and high HbA1c blood pressure above 140/90 mmHg is considered
levels before therapy is initiated (Chantelau and uncontrolled. The achieved values in the UKPDS
Meyer-Schwickerath 2003). The overall effect of were 144/82 mmHg and 154/87 mmHg, respec-
slowing the progression of DR by intensive treat- tively. Substantial reductions in risk for any diabe-
ment of glycemia alone was observed to be stron- tes end point, deaths related to diabetes, and stroke
ger in patients with mild DR (Chew et al. 2014). were seen in the “tight” control group.
Nevertheless, it has to be mentioned that the Each reduction of 10 mmHg in blood pressure
HbA1c gives only a rough estimate of the indi- can lead to a 10 % reduction of DR incidence
vidual patient´s extreme values, more precisely (UKPDS Group 1998). Even after 9 years of fol-
representing the average peak values of the last low-up, the patients assigned to “tight” blood
120 days (Inchiostro et al. 2013). Therefore, the pressure control had a 34 % risk reduction with
value has its limitations with regard to interpre- regard to the proportion of patients with deterio-
tation of 24 h fluctuations and is influenced by ration of retinopathy by two steps and a 47 % risk
other diseases such as anaemia. A lower HbA1c reduction with regard to deterioration in visual
can also indicate the risk of hypoglycemia acuity (three ETDRS-lines). Intervening on blood
(Tamborlane et al. 2008). The occurrence of pressure solely to prevent diabetic retinopathy is
extreme hypoglycemia is not only detrimental not necessarily beneficial. Other comorbidities
to DR, but increases the risk of falls and prohib- have to be considered, when the target pressure is
its driving (Frier 2014). defined (usual goal ≤130/80 mmHg).
In the clinical setting, a close partnership with
the primary care physician is important in order
to enhance the exchange of knowledge. Analyses 6.4.5 Dyslipidemia
showed that blood sugar and duration of diabetes
explain only 11 % of the risk to develop retinopa- Epidemiologic data showed only a very little
thy (Lachin et al. 2008). If the stage of DR is effect of serum lipids or lipid-lowering statins on
known, an individualized education of patients the incidence of PDR or DME (Klein et al. 2015).
can be ensured. Newer drugs as inhibitors of the Some interventional studies failed to prove any
sodium/glucose cotransporter 2 (SGLT2) might significant influence of statins on microvascular
provide another risk-benefit ratio than Insulin complications. This does not question the role of
alone (Matsuda et al. 2015). statins in the context of macrovascular events.
However, the decision of their use should not be
made in dependence of DR.
6.4.4 Blood Pressure In contrast, there is growing evidence that the
(additional) intake of fibrates has a moderate
A tight blood pressure control was proven to impact on DR progression. The FIELD study
lower the incidence of DR as well as the progres- showed a 30 % reduction for the need of laser
sion rate of already existing DR (Do et al. 2015; treatment and progression of albuminuria (Keech
Ferrannini and Cushman 2012; Snow et al. 2003). et al. 2007). In ACCORD-Eye, the addition of
The UKPDS study randomly assigned patients fenofibrate to a basal statin therapy resulted in a
to a “tight” blood pressure control group with a significant decrease in the progression of dia-
target of less than 150/85 mmHg or a “less tight” betic retinopathy, in a similar manner as that
control group with a target of less than observed with intensifying blood glucose control
180/105 mmHg (UKPDS Group 1998). It is (Chew et al. 2010, 2014). However, a strong
important to know that these targets, described as degree of retardation of DR progression was
“tight” and “less tight” control, do not conform only seen in study participants with mild reti-
with current standards from the Joint National nopathy at baseline. Only very limited conclu-
Committee on Prevention, Detection, Evaluation, sions were possible regarding the incidence of
and Treatment of High Blood Pressure, in which a DME. The small MacuFen study showed minor
effects on the macular volume in DME (Massin In some patients, the eye can be an indicator
et al. 2014). of an imminent kidney disease. Similarly, exten-
The mode of action of fenofibrate might be sive capillary nonperfusion was associated with
based more on its activation of the peroxisome an increased risk of chronic kidney disease (Lee
proliferator-activated receptor (PPAR), indepen- et al. 2014). Besides the function of retinal
dently of any lipid-lowering effects (Simo et al. vascular calibers as early sign of preclinical tar-
2013). get organ damage (Daien et al. 2013; Edwards
Concerning the traditional lipid markers as et al. 2005), association between nephropathy
total cholesterol and low-density lipoprotein cho- (defined by microalbuminuria) and retinopathy
lesterol, these parameters might be associated have been well recorded (Wong et al. 2014). The
with the presence of hard exudates in patients bidirectional relationship let suggest a common
with DR (Chang and Wu 2013). The level of apo- but still poorly understood pathogenic pathway
lipoprotein A1 (ApoA1) might be negatively (Grunwald et al. 2014).
related to DR. It has also been speculated that the diabetic
nephropathy may lead to an increased vascular
damage and might provoke/intensify DME as a
6.4.6 Nephropathy and Retinopathy result of a change in osmotic balance. That pos-
sibly is the reason why microalbuminuria was
Nephropathy is the other significant microvascu- identified as an independent risk factor for
lar complication of diabetes (Kaul et al. 2010; DME. The WESDR described an increased risk
Wong et al. 2014). The renal damage is character- for developing DME with gross albuminuria
ized by a decreased glomerular filtration rate (Klein et al. 2009). Within type 1 diabetes, micro-
(GFR) and microalbuminuria and frequently albuminuria and macroalbuminuria were found to
coincides with DR (Fig. 6.12). be associated with an increased risk of developing
Fig. 6.12 Strong link between the microvascular complications, nephropathy, and DR (adopted from Wong et al.
(2014))
DME (Roy and Klein 2001). However, smaller prior to conception (Rahman et al. 2007;
studies did not find a significant correlation when Rasmussen et al. 2010; Schultz et al. 2005;
adjusting for confounding factors. Vestgaard et al. 2010).
There are reports on a potential clinical Preferably, women who are planning to
improvement of DME under dialysis (Matsuo become pregnant should have their eyes exam-
2006; Tokuyama et al. 2000). However, these ined before they attempt to conceive and be coun-
reports are contradictory and even when using selled about the risks of worsening (Ringholm
OCT imaging to quantify retinal thickness, the et al. 2012). PRP for severe NPDR or PDR should
findings of DME patients undergoing dialysis be considered. In unstable disease, patients
are not consistent (Theodossiadis et al. 2012; should be advised to postpone conception until
Tokuyama et al. 2000). these conditions are properly treated.
An exam in early pregnancy has been recom-
mended anyway, at the latest soon after concep-
6.4.7 Mortality tion. The findings of the first trimester grading
determine the time of the next eye examination:
Several studies suggest that the presence of DR is Pregnant women with less than severe NPDR
an indicator of increased mortality risk (Hayes should be examined every 3 months, whereas
et al. 2013). Already incidental findings of mild those with more severe stages should be seen
DR have to remind the physicians that the micro- more frequently, e.g., every 1–3 months. Intensive
vascular alterations point towards the risk profile glycemic control before and during pregnancy
of severely ill patients, being at an increased risk gives considerable benefits to pregnant women
of many other serious events (Stamler et al. and their offspring. Severe hypoglycemia occurs
1993). In comparison to healthy subjects, people most frequently in the first trimester, but should
with diabetes have a greater than fourfold risk of be possibly avoided (Nielsen et al. 2008). Early
cardiovascular diseases and a life expectancy worsening has been reported after intensification
shortened by up to 8 years (Gu et al. 1998). The of insulin. The therapy should be managed as a
presence of DME increases the risk of being hos- collaborative effort between obstetricians, endo-
pitalized by 98 % after adjustment for other crinologists, dieticians, and ophthalmologists.
comorbidities (Nguyen-Khoa et al. 2012). Thy typical gestational diabetes is not consid-
ered to be associated with DR and does not require
regular check-up examinations (Gunderson et al.
6.4.8 Pregnancy 2007). According to current guidelines, women
also have to be followed up after delivery (Chen
Pregnancy is a known risk factor for progression et al. 2004). During the postnatal period, the
of diabetic retinopathy (DCCT Research Group mother´s need for insulin declines to approxi-
2000; Klein et al. 1990; Phelps et al. 1986). The mately 60 % of the pregnancy dose, owing to the
responsible mechanisms discussed to lead to dis- lack of placental hormonal influence, and the risk
ease acceleration are hemodynamic changes, of fluctuating glucose levels are associated with an
increased release of growth factors (e.g., IGF1), increased risk of maternal hypoglycemia during
and hormonal influences (Lauszus et al. 2003; breastfeeding. Facing a high risk of progression,
Loukovaara et al. 2003). meticulous retinal surveillance is mandatory.
The stage of disease prior to conception is an
important predictor of the disease progression,
including the time after delivery (Chew et al. 6.4.9 Systemic Treatment
1995b). That explains why patients with
moderate-to-severe NPDR have to be watched Therapeutic exercise, weight loss, and smoking
very closely. Additional risk factors of progres- cessation are important lifestyle changes influ-
sion are hypertension and poor glycemic control encing DR risk (Klein et al. 2014; Loprinzi et al.
2014; Perreault et al. 2012). Newer drugs such as to be considered. Some pioneers have suggested
injectable incretin mimetics, Glucagon-like pep- that, following the reduction in tissue associated
tide analogs/agonists, and dipeptidyl peptidase-4 with photocoagulation, the retinal autoregula-
Inhibitors may have their role, but currently there tion might decrease the retinal blood flow to the
is not sufficient data to evaluate their effect on macula, attributable to improvements in the
DR. The theoretical benefits of glucosurics like oxygenation of the outer retina (Stefansson
SGLT-2 inhibitors similarly still need to be veri- 2006; Wilson et al. 1988). Changes of the retinal
fied (Dziuba et al. 2014). Previous expectations pigment epithelium (RPE) characterize the
in relation to the efficacy of PKCβ inhibitors or response mechanisms, concomitant to the reso-
somatostatin replacement therapy ended disap- lution of the DME (Ogata et al. 2001).
pointingly (Shah et al. 2010; Sheetz et al. 2013). Restoration of the outer blood–retina barrier and
At this point, we want to remind of the previ- a better pump function of the RPE are also dis-
ous experiences proving that aspirin and rheo- cussed (Klaassen et al. 2013).
logical therapy is ineffective (ETDRS Research The ETDRS was the first trial, describing the
Group 1991c). Aspirin therapy at a dose of various manifestations of CSME (ETDRS
650 mg per day was shown to be ineffective by Research Group 1985). All these constellations are
the ETDRS, although not causing more severe or associated with an increased risk of visual loss.
frequent hemorrhages in PDR (Chew et al. The effect of focal laser photocoagulation on DME
1995a). Data of epidemiologic studies and small was analyzed in eyes with a broad range of edema
case series is the only hint that anemia might be severity at baseline (ETDRS Research Group
associated with the progression of DR (Davis 1995). The extent and intensity of the effect were
et al. 1998; Qiao et al. 1997; Shorb 1985). independent of the size of edema, the type of leak-
age pattern (focal, diffuse) or the extent of con-
comitant ischemia. Prompt treatment was advised
6.5 Treatment of DR in eye with edema involving the center of the mac-
ular or large plaques of hard exudate threatening
The treatment of DR depends on the presence or the center. The main benefit of the central photo-
absence of clinically significant or center— coagulation as identified in ETDRS was the
involving macular edema and peripheral disease. reduction of visual loss (Fong et al. 1999).
The current treatment recommendations are A later modification of the initial laser proto-
based on the results of two major randomized col was evaluated in 840 eyes with a retinal thick-
landmark trials, the Diabetic Retinopathy Study ness of more than 250 μm by the DRCR.net
(DRS), and the ETDRS. While the treatment (2008). The laser spot size was limited to
algorithm of severe NPDR accords with that of 50–80 μm (Fong et al. 2007) (see below table).
PDR regarding their “high-risk profile”, the Besides the “focal” sources of leakage such as
approach will be presented together following microaneurysms, every retinal thickening was
the strategies in DME. targeted by an additional “grid” with a pattern of
laser spots applied at a distance of two laser spot
diameters to each other. An FLA was taken at
6.6 Treatment of DME baseline and follow-up treatments. The parame-
ters of laser spots (energy, exposure) were chosen
6.6.1 Focal/Grid Laser Coagulation in order to generate very faint, but still visible
spots. The wavelength of the lasers was green or
The efficacy of laser treatment was discussed to yellow. The focal/grid laser resulted in a modest
be partly due to its ability to occlude leaking stabilization of visual acuity which was superior
microaneurysms. But as grid treatment alone to repeated injections of triamcinolone (Fong
was shown to be effective, indirect effects have et al. 2007).
Settings of focal/grid laser fluorescence, OCT or FLA. The use of the 532 nm-
Spot size 50–80 μm wavelength promises less pain during the treatment,
Laser energy/exposure (Barely) visible spots a wavelength of the diode laser of 810 nm might
Distance of laser spots (grid) 2 spots have a better penetration (blood, cataract).
Retreatment interval 4 months Some authors reported a good response to
subthreshold micropulse laser treatment, mea-
sured by visual acuity or electrophysiology
Drawbacks of the focal/grid laser are a (Friberg and Karatza 1997; Venkatesh et al.
decrease in retinal sensitivity. This has an impact 2011). Prospective trials showed at least a non-
on the reading ability or reading speed of treated inferiority to conventional lasers, including the
patients (Comyn et al. 2014; Pearce et al. 2014). promise of a better safety profile (Figueira et al.
Regarding its impact on reading, the central area 2009; Pei-Pei et al. 2015).
to the left should be spared if possible. Older
studies did only include a limited documentation
of visual fields following laser treatment. This is 6.6.3 Selective Laser Therapy
important, since later laser spots can enlarge and
become confluent over time (Kang et al. 2010; By using a wavelength of 527 nm and even
Lovestam-Adrian and Agardh 2000; Maeshima shorter pulses than the micropulse lasers (1.7 μs),
et al. 2004). When comparing the amount and the selective retina therapy (SRT) was described
intensity of the laser treatment used in various to be absorbed to a greater extent by the melano-
studies, a large variety has to be noted. The somes of the RPE. High peak temperatures that
number of laser treatments ranges even within develop around melanosomes during irradiation
controlled DME studies from 1.1x to 3.1x create short-lived microbubbles that can mechan-
(DRCR.net 2008; Mitchell et al. 2011). ically disrupt RPE cells as the cell volume rises
Today, monitoring noncentral edema with (Brinkmann et al. 2000). SRT showed promising
SD-OCT and having alternative drug treatments results in early pilot and phase II studies, provid-
at hand, the observation—instead of laser treating follow-up periods of 6 months (Roider et al.
ment—has to be reevaluated. Therefore, some 2000, 2010). As the effects are invisible, the
patients with noncentral DME might be spared implementation of optoacoustic systems and
the downsides of laser treatment and for those reflectometry was used to allow dosimetry and
that do expand into the fovea; drug therapy (when ensure the safety of the applied energy.
delivered timely) can still prevent visual deterio-
ration or achieve improvement.
6.6.4 Navigated or Targeted Laser
6.6.2 Subthreshold Micropulse Newer laser machines implement technical

Laser enhancements, such as the possibility of noncon-
tact imaging or eye-tracking (Kozak et al. 2011;
In contrast to conventional lasers, devices deliver- Liegl et al. 2014) (Fig. 6.13). The procedure still
ing short pulses are intended to cause less thermal has to be conducted by an attentive user, as varia-
damage to the tissue (Park et al. 2014). Shorter tions in fundus pigmentation require adjustment
exposure times are considered to cause less harm to of the laser energy during the treatment, but the
the neural retina and the choriocapillaries than the strategy for the laser treatment becomes more
“continuous-wave” lasers, while still reaching the independent on the experience of the treating
RPE (Framme et al. 2009). One practical conse- physicians. At present, the applied parameters as
quence is that the laser spots of subthreshold micro- retreatment intervals still need to be established
pulse diode laser remain invisible, using ophthalmic in larger conclusive trials (Neubauer et al. 2013).
imaging methods as biomicroscopy, fundus auto- A standardized procedure and the complete
Fig. 6.13 Navigated laser treatment is easily possible progression is indicated during the implementation ((c),
without contact lens (a) Location and density of the Courtesy: Michael Ulbig, wearing reading glasses)
laser pattern can be preplanned in the photograph (b) The
documentation of target localization and energy 6.6.5 Intravitreal Steroids

used are important advantages, especially when
considering the large variability of older trials. As many inflammatory cytokines were identified
This might be more important for newer laser in the vitreous of patients with DME, the pleio-
modalities, invisible to the eye of the examiners. tropic mechanisms of corticosteroids have been
It has been recommended not only to laser on discussed to be beneficial for DME treatment
the basis of biomicroscopy or angiograms, but to (Zhang et al. 2011).
use the information of OCT examinations to plan The intravitreous route of steroid administra-
the laser protocol (Kozak et al. 2014). The question is thought to minimize systemic side effects.
tion of combination therapy is addressed in the The modalities of the different steroid formulations,
abstract of the particular drug. however, have not been directly compared in a
clinical trial (Al Dhibi and Arevalo 2013).
Important to remind
Focal/grid laser can deliver a considerable 6.6.6 Triamcinolone Acetate
risk reduction of a visual loss in CSME.
By affecting the central visual field, focal/ Triamcinolone therapy has been reported to be
grid laser may cause (growing) retinal quite effective in the short term, but to be inferior
sensitivity defects and reading problems. to focal/grid laser therapy over 3 years in the
Micropulse lasers have shown equal DRCR.net study protocol B (Elman et al. 2011).
efficacy to conventional lasers in small, but Glaucoma surgery was described to be necessary
prospective trials. However, longer in 1.6 % of treated patients. A subgroup analysis
follow-up data is needed. of the protocol I study showed a much greater
amount of visual improvement, when combining
laser and triamcinolone in pseudophakic eyes (Elman et al. 2010). Thus, the long-term benefits
(Elman et al. 2015). However, in the overall of triamcinolone treatment do not seem to persist,
group the probability of anatomic restoration of even after repeated injections (Table 6.2).
retinal thickness after 2 years was only 16 % with Throughout all published studies, the cumula-
triamcinolone, compared to 34 % by ranibizumab tive risk of subsequent cataract surgery (83 %) or
Table 6.2 Important landmark studies of DR and DME treatment

Abbreviation Study name Key findings
DRS Diabetic Retinopathy Study Risk reduction for loss of vision by laser photocoagulation in
severe NPDR (13–4 % < 5/200) and PDR (44–20 % < 5/200)
ETDRS Early Treatment of Diabetic Risk reduction for loss of vision by laser photocoagulation in
Retinopathy Study clinically significant macular edema (33–14 %)
DRVS Diabetic Vitrectomy Study Early vitrectomy beneficial in eyes <5/200 and severe
hemorrhage >4 weeks with type 1 diabetes (≥20/40: 36 vs.
12 %)
Vitrectomy beneficial in eyes ≥20/400 if severe
neovascularization, fibrous proliferation, and moderate
vitreous hemorrhage or moderate neovascularization with
severe fibrous proliferation (≥20/40: 44 vs. 28 %)
FIELD Fenofibrate Intervention and In spite of a baseline imbalances (statins) addition of
Event Lowering in Diabetes fenofibrate positive impact on progression of DR
ACCORD-EYE Action to Control Cardiovascular Retardation of DR by fenofibrate seen in mild and early DR
Risk in Diabetes stages from 10 to 6 %
Tight blood sugar control with HbA1c ≤6.4 decreased
progression of DR from 10 to 7 %
Drug therapy DME
RESTORE 0.5 mg ranibizumab as 0.5 mg ranibizumab monotherapy (similar to adjunctive
monotherapy and/or adjunctive to therapy to laser photocoagulation) provided superior benefits
laser treatment vs. laser therapy as compared to laser monotherapy
37–43 % of 0.5 mg ranibizumab-treated patients improved
vision by ≥10 letters compared to 16 % with laser therapy
REVEAL Ranibizumab in the TrEatment of More Asian patients gained 15 letters with 0.5 mg
Visual Impairment in DiabEtic ranibizumab (18.8 %) and combined laser (17.8 %) compared
MAcuLar Edema with laser (7.8 %)
RIDE and RISE Monthly intravitreal ranibizumab Mean visual acuity shows an increase beyond 12 months
(0.5 or 0.3 mg) vs. sham Significant differences between the twin studies indicate the
need on confirmatory studies
BOLT Prospective randomized trial of Superiority of bevacizumab (n = 42) over laser (n = 38) in
intravitreal Bevacizumab Or pretreated patients with CSME
Laser Therapy
VISTA and Aflibercept 2 mg every 4 weeks More patients gained 15 letters with 2q4 (VISTA: 41.6 %,
VIVID (2q4), aflibercept 2 mg every VIVID: 32.4 %) and 2q8 (VISTA: 31.1 %, VIVD: 33.3 %)
8 weeks after 5 initial monthly compared with laser (VISTA: 7.8 %, VIVID: 9.1 %)
doses (2q8) vs. macular laser
MEAD Dexamethasone implant More patients gained 15 letters with the 0.7 mg implant
(0.7 mg or 0.35 mg) vs. sham (22.2 %, 4.1 tx) and 0.35 mg implant (18.4 %, 4.4 tx)
compared with sham (12.0 %, 3.3tx) over 3 years
Indications of undertreatment with 6-months retreatment
intervals of dexamethasone implants
Better functional outcomes of pseudophakic patients
High rate of study discontinuation due to the study protocol
(continued)

Abbreviation Study name Key findings
DRCR.net The Diabetic Retinopathy Clinical Research Network
Protocol B Randomized Trial Comparing Intravitreal Triamcinolone Acetonide and Laser Photocoagulation
for DME
Protocol H Subclinical Diabetic Macular Edema
Protocol I Randomized Trial Evaluating Ranibizumab Plus Prompt or Deferred Laser or Triamcinolone Plus
Prompt Laser for Diabetic Macular Edema
Protocol M Effects of Diabetes Education during Retinal Ophthalmology Visits on Diabetes Control
Protocol R A Phase II Evaluation of Topical NSAIDs in Eyes with Noncentral-Involved DME
Protocol S Prompt panretinal photocoagulation vs. intravitreal ranibizumab with Deferred panretinal
photocoagulation for PDR
Protocol T A comparative effectiveness study of intravitreal aflibercept, bevacizumab, and ranibizumab for
DME
Protocol V Treatment for Central-involved DME in eyes with very good visual acuity
increased intraocular pressure (IOP, >10 mmHg: tion system, before releasing 700 μg of dexa-
33 %) after triamcinolone have been reported methasone as the polymer degrades (Chang-Lin
invariably high (Kiddee et al. 2013; Yilmaz et al. et al. 2011; Fialho et al. 2006). Single case series
2009). There are some indications that the risk of showed efficacy in eyes more resistant to other
an increase in intraocular pressure depends on treatment modalities (Kuppermann et al. 2007;
the applied dosage (2 mg < 4 mg < 10 mg). Zucchiatti et al. 2012) (Fig. 6.14).
None of the available triamcinolone drugs has The approval of the dexamethasone implant is
been approved for the use in DME. The crystal- based on a prospective randomized comparison
loid triamcinolone has large variability in resorp- with sham injections (including rescue laser)
tion, though the intravitreal administration seems (Boyer et al. 2014). The MEAD study design was
to be superior to subtenonal injections (Qi et al. influenced by the hope that the effect of the deliv-
2012). After intravitreal injection, however, ery device always lasts for 6 months. Fixed
crystalloid particles or remnants of conserva- retreatments were planned twice a year, unless
tives can induce the unpleasant complication of resolution was seen or one of the many discon-
pseudo-endophthalmitis which can be confused tinuation criteria (rise of IOP, lack of efficacy,
with infectious endophthalmitis and sometimes etc.) was fulfilled. Less than 60 % of patients
needs removal of the drug by vitrectomy (607 of 1048) completed the study. This constel-
(Marticorena et al. 2012). In addition, intravitreal lation limits the evaluation of the ITT-analysis.
triamcinolone should be used with caution in The main study publication described a mean VA
eyes after removal of the inner limiting mem- increase of 6.5 ETDRS letters for pseudophakic
brane (ILM) due to potential toxicity (Jaissle patients after 3 years. Study participants received
et al. 2004). a mean number of 4.1 treatments with the 700 μg
implant. The saw tooth pattern of central retinal
thickness seen on OCT examinations suggests
6.6.7 Dexamethasone Implants that the treatment interval was too long for most
of the patients. Mean visual acuity started to
The dexamethasone drug delivery system worsen after the 6-week examination. The maxi-
(Ozurdex®, Allergan, Irvine, California, USA) mum gain in visual acuity was achieved after the
consists of dexamethasone bound with a biode- first injection and was stabilized by the following
gradable copolymer of lactic and glycolic acids treatments. There is room for speculation about
(Nehme and Edelman 2008). The implant is the results if an individualized PRN dosing would
injected through the pars plana using a 22G injec- have been used.
Fig. 6.14 While the initial gain in best-corrected visual methasone implants, the subgroup of pseudophakic
acuity from baseline could not be maintained for the total patients (b) achieved a mean improvement of 6.1 ETDRS
study population (a) by 6-monthly retreatment with dexa- letters (Boyer et al. 2014)
An IOP of ≥ 25 mmHg was found to occur in One study found comparable treatment effica-
32 % of treated patients. 5 out of 347 patients had cies when comparing bevacizumab and the
undergone glaucoma surgery. Meanwhile, there 22-Gauge dexamethasone implant in a random-
are many case series reported, showing shorter ized trial over 12 months (Gillies et al. 2014).
retreatment intervals and a better efficacy of The implant exhibits some variability in dissolv-
naïve patients when compared with pretreated ing; sometimes, the biologic activity is decreas-
DME (Escobar-Barranco et al. 2015; Guigou ing though patient and physician still clearly
et al. 2014; Scaramuzzi et al. 2015). notice the “bar” consisting of the remaining
Fig. 6.15 The resorption of the intravitreal PLGA Dislocation in the anterior chamber in pseudophakic and
implant into a translucent matrix usually takes 6–8 weeks in particular aphakic patients can lead to corneal decom-
and can be monitored in the inferior vitreous (a). pensation (b)
PLGA shell (Fig. 6.15). Corneal problems have drug per day. The prospective design allowed
been reported, when displacement of the implant concomitant treatments with anti-VEGF drugs;
into the anterior chamber occurs in aphakic or over a quarter of recruited patients received off-
pseudophakic patients (Voykov and Bartz- protocol treatment and 40 % additional laser
Schmidt 2012). therapy (Campochiaro et al. 2012). Therefore, it
is not possible to directly assign the change of
visual acuity to the implant alone. Both fluocino-
6.6.8 Fluocinolone Implants lone dosages showed similar numbers of patients
with 15-letter improvements at month 24
Fluocinolone was first evaluated in a study with (28.7 %, 28.6 %) sham-treated patients achieved
surgically implanted, nonbiodegradable intravit- a 15-letter VA improvement only in 15 % of the
real insert (Messenger et al. 2013). The break- cases. Glaucoma requiring surgery was neces-
down of the central polymer–drug matrix leads to sary in 3.7 and 7.6 % of patients in the two fluo-
a release of 0.59 μg fluocinolone per day over a cinolone groups. In addition, almost 90 % of
time period of approximately 3 years. The phakic patients required cataract surgery after
Retisert® (Bausch & Lomb) implant which fluocinolone implant treatment. Further post hoc
received initial FDA approval for the treatment of analyses found a more pronounced difference
chronic, posterior uveitis was also investigated between chronic edema (defined as edema of at
for its efficacy in DME (Pearson et al. 2011). least 3 years duration) and non-chronic DME
Two trials showed an increase in visual acuity of (Cunha-Vaz et al. 2014a). A similar stratification
more than three lines in about 30 % of patients was not investigated for any other treatment
after 3 years, however also a considerable amount modality.
of side effects. After the approval of the 190 μg implant
The other delivery system was analyzed in (0.2 μg/d over 18 months) head-to-head compari-
two double-masked trials (FAME A, FAME B) sons to other treatment modalities or laser treat-
that recruited 951 participants with persistent ment alone are still missing (Kane et al. 2008). The
DME (Campochiaro et al. 2011). Two different cylindrical tubes (3.5 × 0.37 mm), injected through
dosages were tested, releasing 0.2 or 0.5 μg of a transconjunctival self-sealing wound with a
25-gauge needle, have the disadvantage to be non- focal/grid laser treatment, regarding the primary
biodegradable. Facing the safety profile discussed outcome of visual acuity (Table 6.3).
above, fluocinolone implants might be considered A recent head-to-head study described a better
if prior treatments have failed and long duration of response to aflibercept in comparison to bevaci-
persistent DME has been verified. zumab, when picking out those patients with a
visual acuity of 20/50 or worse (Wells et al. 2015).
The difference might be more pronounced, if treat-
Important to remind ing a macular edema with retinal thickness ≥ 400
In the absence of direct head-to-head µm (Wells et al. 2015). Most studies focused on
comparisons, only indirect conclusions can DME involving the center of the macula by defin-
be drawn regarding potentially different ing inclusion criteria, either a decrease of visual
efficacies between different steroid agents. acuity related to center involvement or a minimum
The broadest evidence is available for thickness of the fovea when measured with OCT
dexamethasone implants, although open (Aiello et al. 2011). Therefore, it is still unclear,
questions exist regarding retreatment whether watch and wait (for later anti-VEGF treat-
intervals, the need of long-term treatment ment) might be an alternative to focal/grid laser in
as well as complications. CSME fovea-sparing.
The rapid and profound steroid action is The evidence is low regarding the best retreat-
limited by the complications of cataract ment approach (Treat&Extend vs. Pro re nata).
formation and steroid-induced glaucoma, There is only one unpublished study comparing
which lead to their attribution in most two different regimens for the same drug
guidelines as a second-line treatment. (RETAIN, NCT01171976): After an uniform
The need of regular control examinations upload phase ranibizumab 0.5 mg was adminis-
(in order not to miss any significant rise tered either in a T&E regimen or an as-needed
in IOP) limits the advantage of longer retreatment (PRN). Non-inferior results were
duration and treatment intervals. reported for T&E to PRN over 24 months. The
Very limited experience exists regarding T&E regimen might lead to a reduction in the
the switching between different steroid number of treatment visits, although having some
drugs and the efficacy of combination issues related to its implementation in a bilateral
therapy of steroids with other modalities. disease.
A marked improvement of the DR severity
has been reported in most of the prospective tri-
als (Korobelnik et al. 2014; Stewart 2014).
Moreover, eyes receiving anti-VEGF drugs were
6.6.9 Intravitreal Anti-VEGF significantly more likely to improve by ≥2 or ≥3
Therapy on the ETDRS severity scale. However, caution
seems to be necessary regarding the end or ces-
The inhibitors of the vascular endothelial growth sation of treatment (Ip et al. 2015). The short-
factor (VEGF) have brought the treatment of ness of follow-up might not allow completely
DME to a new era (Rosberger 2013). Since a assessing the possibility of reverting worsening.
meta-analysis in 2009 questioning the evidence As vasoconstriction of retinal vessels was
large RCTs have not only proven the treatment described to be induced by the inhibition of
efficacy but also its long-term safety (Abouammoh VEGF (Sacu et al. 2011), a change of the perfu-
2013; Parravano et al. 2009; Regnier et al. 2014; sion and vascular remodeling have been dis-
Schwartz et al. 2014; Virgili et al. 2014). Anti- cussed to cause a reduction of treatment burden
VEGF treatment was shown to be superior to the over time (Elman et al. 2015). Although pro-
Table 6.3 Visual improvement as seen after the first year of treatment (Ishibashi et al. 2015; Korobelnik et al. 2014;
Mitchell et al. 2011; Nguyen et al. 2012; Rajendram et al. 2012; Wells et al. 2015)
Increase in
Mean initial Prior BCVA ≥ 15
BCVA anti-VEGF Mean number of treatments ETDRS letters
1 year 2 years 1 year 2 years
Ranibizumab 0.5 mg
PRN + prompt laser Median 66 13 % Median 8 Median 2 30 26
Prot I (n = 187)
PRN + deferred laser Median 66 11 % Median 9 Median 3 28 29
Prot I (n = 188)
Sham + prompt laser Median 65 9% 15 17
(n = 293)
PRN RESTORE 64.8 (10.1) – 7.0 10.0 22 28.9
(n = 116)
Laser/sham (n = 110) 62.4 (11.1) – 2.1 (+1.1) 8 18.9
PRN REVEAL 58.6 –
(n = 133)
Laser/sham (n = 131) 56.6 –
FIXED RISE (n = 125) 56.9 (11.6) 17 % 39.2
Sham (n = 127) 57.2 (11.1) 17 % 18.1
FIXED RIDE (n = 125) 56.9 (11.8) 20 % 45.7
Sham (n = 130) 57.3 (11.3) 16 % 12.3
Aflibercept 2 mg
2q8 VISTA (n = 151) 59.4 (10.9) 45 % 8.4 31
Laser/sham (n = 154) 59.7 (10.9) 41 % 31.2 %* 8
2q8 VIVID (n = 135) 60.8 (10.6) 11 % 8.7 33
Laser/sham (n = 132) 58.8 (11.2) 10 % 24.1 %* 9
PRN Protocol T 65.0 (11.8) 11 % 9.2 42
(n = 224)
Bevacizumab 1.25 mg
PRN BOLT (n = 37) 55.8 (9.7) n.d. 9 13 12 32
Laser (n = 28) 55.4 (7.9) n.d. 5 4
PRN Protocol T 64.8 (11.2) 13 % 9.4 32
(n = 206)
spective studies were running at longest over 5 years; however, such eyes may need slightly
years, the marked decrease of necessary treat- more injections.
ments does further reinforce the cost efficacy of Retinal thinning remains an imminent part of
the expensive drugs (Pershing et al. 2014). While the underlying disease (Douvali et al. 2014; Shin
more than eight treatments were considered dur- et al. 2014). In spite of the limited experience, the
ing the first year for the majority of patients ocular safety of intravitreal injections seems to be
(>2/3), the reduction of intraretinal fluid is convincing (Virgili et al. 2014; Zechmeister-Koss
accompanied by only very little treatments after and Huic 2012).
year 3, even though a regular combination with a Facing the high comorbidity and reduced life
deferred focal/grid laser was applied (Fig. 6.16). expectancy due to the cardiovascular risk pro-
More than half of the eyes in which laser treat- file of DME patients, there have been valid con-
ment is deferred may avoid laser for at least 5 cerns about the systemic safety of anti-VEGF
Fig. 6.16 Injection frequency during the first year of anti-VEGF treatment (DRCR.net PRN retreatment algorithm)
(Wells et al. 2015)
treatment (Lim et al. 2011). So far, there are no 2014; Liegl et al. 2014). Especially, the combina-
threatening safety signals in the prospective tion of focal/grid laser has extensively been stud-
studies which are different to the sham groups. ied (Schmidt-Erfurth et al. 2014). Furthermore,
The phase III trials were not powered to detect most of the studies investigating monotherapy
small differences. Therefore, more investiga- have included patients with a considerable
tions are necessary, while some patients with a amount of previous laser treatment. A sequential
recent cardiovascular event had been excluded combination frequently occurs in the daily
and study cohorts are known to be healthier routine.
than the general population in real life (Thulliez A substantial difference in the number of laser
et al. 2014). Due to the different duration and treatments was found as difference of a prompt
extent of use, the body of evidence regarding against a deferred laser combination (Elman
the clinical safety differs between the available et al. 2015). Nevertheless, the need or burden of
drugs. anti-VEGF could not be significantly reduced. It
is unclear whether the absence of any additional
improvement is influenced by the low standard-
6.6.10 Combination ization of laser. One smaller trial used a navi-
gated laser and reported a potential reduction of
So far, no convincing evidence exists that a fixed injection number (Liegl et al. 2014). Before this
combination of different treatment modalities experience of a single center has not been con-
(anti-VEGF and laser, anti-VEGF and steroids) firmed, a routine co-administration cannot be
might further increase the efficacy (Chen et al. recommended.
There have been some randomized studies

Important to remind verifying a reduction of postoperative hemor-
Frequent anti-VEGF treatment is necessary rhages for vitrectomy in proliferative vitreoreti-
during the first year of initiation. The need nopathy. The improvement of DR severity seen
for retreatment considerably decreases in DME trials does not yet justify leaving the
during the following years. panretinal scatter laser. Trials (NCT01594281,
One recent head-to-head trial showed a NCT01489189) addressing this issue are either
more pronounced response to aflibercept in still recruiting or ongoing.
a subgroup of patients with visual acuity of Vitreous hemorrhage accumulates depending
20/50 or worse. on the position and gravity, patients with dense
The usefulness of a combined laser vitreous hemorrhage can be advised to sleep with
treatment with anti-VEGF therapy has not the head elevated so that the blood can settle in
yet been fully established. Though the inferior periphery and not restrict the optical
implemented in some studies, a marked axis by deposition in the macula.
reduction of treatment burden or other
advantages have not been found.
Do not forget to evaluate the peripheral 6.7.1 Panretinal Photocoagulation
retina when treating DME patients.
Progression of DR might occur, especially The development of retinal laser photocoagulation
during longer therapy-free intervals. by Meyer-Schwickerath did achieve a remarkable
Before appraising nonresponders (and decrease in visual loss by PDR (Meyer-
switching drugs), careful follow-up Schwickerath 1977). Successfully at preventing
examinations and adherence to predefined blindness, PRP still is the most widespread treat-
retreatment algorithms allow a better evaluation ment for PDR (DRS Group 1981a). Recent analy-
of the action in the individual patient. ses of visual outcome suggest that the
In DME with foveal involvement anti- recommendation to consider PRP in severe NPDR
VEGF therapy should be preferred over or non-high-risk PDR before the development of
focal/grid laser therapy. high-risk PDR is particularly appropriate for per-
sons with type 2 diabetes and older age (Ferris
1996). However, the clinical application and its
reported efficacy vary markedly (Luo et al. 2015).
PRP leads to a regression of neovascularization
6.7 Antiangiogenic Treatment within several weeks after treatment, presumably
of PDR due to a reduction of the metabolic demand of the
lasered retina (Doft and Blankenship 1984).
While new vessels were found to respond to While the initial pathomechanistic concept of
anti-VEGF drugs (Avery 2006), their temporary PRP was based on destruction of poorly perfused
use have been discussed as adjunct or therapeu- cells in the neurosensory retina, the retinal pig-
tic agents to achieve regression of retinal neo- ment epithelium and the photoreceptor layers of
vascularizations or at least decrease the risk of the peripheral retina are now thought the key tar-
later vitreous hemorrhage during vitreoretinal gets in order to reduce the angiogenic signaling.
surgery (Gunduz and Bakri 2007; Martinez- Monitoring of the choroidal thickness demon-
Zapata et al. 2014). In contrast to case series and strates a redistribution of choroidal blood flow
single reports, the evidence—apart from the after PRP indicating an altered metabolic demand
effects on DME—is still limited (Martinez- of the lasered retina in the outer retinal layers
Zapata et al. 2014). and/or RPE (Zhu et al. 2015).
It has to be kept in mind that PRP treatment treatment in PDR. Ischemic retinal area, the key
causes collateral damage to the retina, associated trigger of proliferations, are unlikely to disappear
with a constriction of visual fields, impaired dark (within a short time).
adaption, and a decrease in contrast sensitivity
(Boynton et al. 2015). A focal retinochoroidal
atrophy at the laser spots and a progressive scar 6.7.3 Vitreomacular Traction
formation is seen. There are attempts to reduce
the visual field defects by the spatial localization The high incidence of vitreomacular traction in
of laser burns (Wang et al. 2014). Patients treated DR is caused by an increased glycosylation of
with PRP demonstrate increased photostress collagen fibers and the accumulation of prolifera-
recovery time and dark adaptation speed. After tive factors in the vitreous causing vascular
PRP, a diffuse thickening of the retinal nerve hyperpermeability and augmented cell migration
fiber layers and thinning of the retinal pigment (Khan and Haller 2015). Traction at the vitreo-
epithelium layers is observed. Ultimately, the macular interface has been considered to be an
unwanted side effects can lead to reduced driving important contributing factor for DME. The
abilities, especially under low light conditions attachment of the vitreous might increase the fre-
(Gardner et al. 2002). quency of barrier disturbance (Jackson et al.
Another caveat is that following PRP an 2013). Similarly, epiretinal membranes might
increase in macular edema was noticed in some weaken the efficacy of laser and medical treat-
patients. No difference of the risk, however, was ment (Snead et al. 2008).
found whether the PRP was conducted in one or There have been reports about eyes with DME
four sittings (Brucker et al. 2009). Panretinal scat- and vitreomacular traction showing benefits from
ter coagulation remains the first-line treatment in the removal of epiretinal membranes (Laidlaw
severe NPDR and PDR, as the evidence of medi- 2008; Simunovic et al. 2014). Regarding pharma-
cal treatment has not been provided with regard to cological vitreolysis, no data is available beyond
longevity of the effect (Evans et al. 2014). a theoretic concept (de and Castilla 2013).
Spontaneous resolution of vitreomacular traction
seems to occur very rarely in the context of DR.
6.7.2 Intravitreal Anti-VGEF Therapy In a prospective study, a marked reduction of
retinal thickness was achieved by vitrectomy
There is a growing body of evidence that intravit- with removal of the posterior vitreous and peel-
real drugs might be - at least in the short term - an ing of epiretinal membranes (if present).
alternative to panretinal lasercoagulation. The However, the functional results were less con-
DRCR.net provided data of the protocol S study vincing: Visual acuity after 6 months was
indicating non-inferior visual acuity of ranibi- improved in 28–49 % of cases, but worsened in
zumab (DRCR.net 2015). However, during the 13–31 % of patients (Haller et al. 2010).
follow-up of two years, a large portion of the Inconsistent results are available regarding
laser arm (53 %) did also receive ranibizumab vitrectomy with or without peeling of the ILM in
treatment. Without laser, a larger extent of the DME devoid of vitreomacular traction. Some
visual fields was found to be preserved. The dif- authors report on functional improvement, others
ferent rate of vitrectomy (laser: 15 %, ranibi- on a reduction/normalization of retinal thickness
zumab: 4 %) might be influenced by the reserved only (Kumagai et al. 2009; Patel et al. 2006). The
attitude of surgeons to perform surgery in eyes low quality of report leads to a low level of evi-
without any laser pretreatment. Regardless of dence, not justifying a general recommendation
costs, the short follow-up of the study feeds the (Yamamoto et al. 2007). A multivariate analysis
doubts about the sustainability of anti-VEGF of the DRCR.net described a significant decrease
of edema, but without functional benefit in 241 If the visual acuity is good and stable in spite
eyes. Worse visual acuity and the removal of of a localized traction, an observant approach
epiretinal membranes increased the probability to including regular control examinations is possible
gain (Flaxel et al. 2010). after a sufficient laser treatment. But at latest in
The known risks of vitrectomy have to be the case of a progression, tractional forces have
reminded, in spite of theoretical thoughts about a been removed during vitrectomy (Helbig 2007).
better oxygenation (Navarro et al. 2010; Park The dissection of tractional membranes has to be
et al. 2009). Besides the formation of cataract and done very carefully, as retinal breaks may deter-
retinal holes, the development of secondary pro- mine the need of permanent endotamponade
liferative vitreoretinopathy (PVR) and an (Fortun and Hubbard 2011; Wang et al. 2012b).
increased risk of endophthalmitis are relevant Good outcomes have been reported, if the indica-
postoperative problems. In addition, it has to be tion has taken into account visual prognosis and
kept in mind, that vitrectomy might change the systemic treatment (Gupta et al. 2012; Kumar
pharmakokinetics of intravitreally injected drugs et al. 2014; Ostri et al. 2014; Zenoni et al. 2010).
Anti-VEGF drugs prior to vitrectomy might not
necessarily facilitate the surgery, but prevent
6.7.4 Tractional Retinal Detachment intra- and postoperative bleeding (Pokroy et al.
2011; Romano et al. 2009).
Fibrovascular tissue-induced retinal elevation is
an important cause of visual loss in PDR (Imai
et al. 2001). Membranes frequently develop 6.7.5 Neovascular Glaucoma
along the temporal vascular arc, while the inter-
face of the attached vitreous might serve as Neovascular glaucoma is a blinding, intractable
“guide rail”. The whole complex can cause a complication of late PDR (Hayreh 2007). The
localized distance. Macular elevation is seen in glaucoma may occur without retinal or optic disc
tractional retinal detachment and tractional reti- neovascularization, however it is more com-
noschisis (Su et al. 2014). monly seen in association with PDR (Fig. 6.17).
Fig. 6.17 It has to be reminded that the first changes of by angiography far earlier than additional vessel trunks
iris neovascularization can be seen in the iridocorneal have been formed
angle. Leaking vessels can be detected around the pupil
In the patients affected, differentiation of the ment protocol for center-involved diabetic macular
edema. Ophthalmology. 2011;118(12):e5–14.
various diseases (retinal vascular occlusive dis-
Al Dhibi HA, Arevalo JF. Clinical trials on corticosteroids
eases, ocular ischemic syndrome) might be diffi- for diabetic macular edema. World J Diabetes.
cult, but important. Early on, the disease can 2013;4(6):295–302.
present with subtle findings. A thorough exami- Alcubierre N, Rubinat E, Traveset A, et al. A prospective
cross-sectional study on quality of life and treatment
nation of the iris and angle are essential before
satisfaction in type 2 diabetic patients with retinopathy
pupil dilation, since up to 30 % of the neovascu- without other major late diabetic complications.
lar changes are limited to the angle. Health Qual Life Outcomes. 2014;12:131.
Once neovascular glaucoma develops and the American Diabetes Association (ADA). Standards of
medical care in diabetes—2013. Diabetes Care.
intraocular pressure is high, the major aspect of
2013;36 Suppl 1:S11–66.
management is to control the high intraocular Andreoli CM, Miller JW. Anti-vascular endothelial
pressure, which is almost invariably the main fac- growth factor therapy for ocular neovascular disease.
tor in irreversible and massive visual loss in addi- Curr Opin Ophthalmol. 2007;18(6):502–8.
Antonetti DA, Barber AJ, Bronson SK, et al. Diabetic
tion to PDR. The care includes retinal ablation by
retinopathy: seeing beyond glucose-induced micro-
laser or cryocoagulation and control of the vascular disease. Diabetes. 2006;55(9):2401–11.
increased intraocular pressure with medical and Arden GB, Sivaprasad S. Hypoxia and oxidative stress in
surgical therapy (DRS Group 1976; Sivak- the causation of diabetic retinopathy. Curr Diabetes
Rev. 2011;7(5):291–304.
Callcott et al. 2001). Besides topical therapy
Ashton N. Diabetic microangiopathy. Bibl Ophthalmol.
cycloablative procedures promise to be more 1958;12(52):1–84.
successful (Bloom et al. 1997; Delgado et al. Ashton N. Vascular basement membrane changes in
2003). If filtering surgery becomes necessary, the diabetic retinopathy. Montgomery lecture, 1973. Br
J Ophthalmol. 1974;58(4):344–66.
high risk of failure might be reduced by the use of
Avery RL. Regression of retinal and iris neovascularization
drainage devices (Yalvac et al. 2007). after intravitreal bevacizumab (Avastin) treatment.
If the neovascularizations of the angle are Retina. 2006;26(3):352–4.
detected early, anti-VEGF treatment might achieve Banerjee S, Ghosh US, Biswas G, et al. Comparative eval-
uation of ophthalmoscopy and angiography for the
some temporary regression (Luke et al. 2013).
assessment of retinopathy in type 2 diabetes mellitus.
But even in the presence of synechiae and pretra- J Indian Med Assoc. 2007;105(1):33–6.
becular fibrovascular membranes, the drugs might Beagley J, Guariguata L, Weil C, et al. Global estimates of
offer a decrease in barrier disturbance (Andreoli undiagnosed diabetes in adults. Diabetes Res Clin
Pract. 2014;103(2):150–60.
and Miller 2007). The limited short-term experi-
Benson WE, Brown GC, Tasman W. Diabetes and its ocu-
ence does not justify the usage of VEGF-inhibitors lar complications. Philadelphia: Saunders; 1988.
only (Simha et al. 2013), but to try a combination Bloom PA, Tsai JC, Sharma K, et al. “Cyclodiode”.
treatment with conventional measures of pressure Trans-scleral diode laser cyclophotocoagulation in
the treatment of advanced refractory glaucoma.
release (Jacobson et al. 1979).
Ophthalmology. 1997;104(9):1508–19.
Bolz M, Ritter M, Schneider M, et al. A systematic cor-
relation of angiography and high-resolution optical
References coherence tomography in diabetic macular edema.
Ophthalmology. 2009a;116(1):66–72.
Bolz M, Schmidt-Erfurth U, Deak G, et al. Optical coher-
Abouammoh MA. Ranibizumab injection for diabetic mac-
ence tomographic hyperreflective foci: a morphologic
ular edema: meta-analysis of systemic safety and sys-
sign of lipid extravasation in diabetic macular edema.
tematic review. Can J Ophthalmol. 2013;48(4):317–23.
Ophthalmology. 2009b;116(5):914–20.
Agardh E, Cavallin-Sjoberg U. Peripheral retinal evalua-
Boyer DS, Yoon YH, Belfort Jr R, et al. Three-year,
tion comparing fundus photographs with fluorescein
randomized, sham-controlled trial of dexamethasone
angiograms in patients with diabetes mellitus. Retina.
intravitreal implant in patients with diabetic macular
1998;18(5):420–3.
edema. Ophthalmology. 2014;121(10):1904–14.
Ahsan H. Diabetic retinopathy—biomolecules and multi-
Boynton GE, Stem MS, Kwark L, et al. Multimodal char-
ple pathophysiology. Diabetes Metab Syndr.
acterization of proliferative diabetic retinopathy
2015;9(1):51–4.
reveals alterations in outer retinal function and struc-
Aiello LP, Beck RW, Bressler NM, et al. Rationale for the
ture. Ophthalmology. 2015;122:957–67.
diabetic retinopathy clinical research network treat-
Braun CI, Benson WE, Remaley NA, et al. Accommodative Campos C. Chronic hyperglycemia and glucose toxicity:
amplitudes in the Early Treatment Diabetic pathology and clinical sequelae. Postgrad Med.
Retinopathy Study. Retina. 1995;15(4):275–81. 2012;124(6):90–7.
Brechner RJ, Cowie CC, Howie LJ, et al. Ophthalmic Cardillo PF, Zingirian M, Mosci C. Classification of pro-
examination among adults with diagnosed diabetes liferative diabetic retinopathy. Graefes Arch Clin Exp
mellitus. JAMA. 1993;270(14):1714–8. Ophthalmol. 1987;225(4):245–50.
Bresnick GH. Diabetic maculopathy. A critical review Chalam KV, Bressler SB, Edwards AR, et al. Retinal
highlighting diffuse macular edema. Ophthalmology. thickness in people with diabetes and minimal or no
1983;90(11):1301–17. diabetic retinopathy: Heidelberg Spectralis optical
Bresnick GH, Haight B, de VG. Retinal wrinkling and coherence tomography. Invest Ophthalmol Vis Sci.
macular heterotopia in diabetic retinopathy. Arch 2012;53(13):8154–61.
Ophthalmol. 1979;97(10):1890–95. Chang YC, Wu WC. Dyslipidemia and diabetic retinopathy.
Bressler NM, Miller KM, Beck RW, et al. Observational Rev Diabet Stud. 2013;10(2–3):121–32.
study of subclinical diabetic macular edema. Eye Chang-Lin JE, Burke JA, Peng Q, et al. Pharmacokinetics
(Lond). 2012;26(6):833–40. of a sustained-release dexamethasone intravitreal
Bressler NM, Varma R, Doan QV, et al. Underuse of the implant in vitrectomized and nonvitrectomized eyes.
health care system by persons with diabetes mellitus Invest Ophthalmol Vis Sci. 2011;52(7):4605–9.
and diabetic macular edema in the United States. Chantelau E, Meyer-Schwickerath R. Reversion of ‘early
JAMA Ophthalmol. 2014;132(2):168–73. worsening’ of diabetic retinopathy by deliberate resto-
Brinkmann R, Huttmann G, Rogener J, et al. Origin of ration of poor metabolic control. Ophthalmologica.
retinal pigment epithelium cell damage by pulsed laser 2003;217(5):373–7.
irradiance in the nanosecond to microsecond time Chen YJ, Kuo HK, Huang HW. Retinal outcomes in prolif-
regimen. Lasers Surg Med. 2000;27(5):451–64. erative diabetic retinopathy presenting during and after
Browning DJ, Glassman AR, Aiello LP, et al. Relationship pregnancy. Chang Gung Med J. 2004;27(9):678–84.
between optical coherence tomography-measured cen- Chen G, Li W, Tzekov R, et al. Ranibizumab monother-
tral retinal thickness and visual acuity in diabetic mac- apy or combined with laser versus laser monotherapy
ular edema. Ophthalmology. 2007;114(3):525–36. for diabetic macular edema: a meta-analysis of ran-
Browning DJ, Altaweel MM, Bressler NM, et al. Diabetic domized controlled trials. PLoS One. 2014;9(12):
macular edema: what is focal and what is diffuse? Am e115797.
J Ophthalmol. 2008;146(5):649–55. Chew EY, Klein ML, Murphy RP, et al. Effects of aspirin
Brownlee M. Biochemistry and molecular cell biology of on vitreous/preretinal hemorrhage in patients with
diabetic complications. Nature. 2001;414(6865):813–20. diabetes mellitus. Early Treatment Diabetic
Brucker AJ, Qin H, Antoszyk AN, et al. Observational study Retinopathy Study report no. 20. Arch Ophthalmol.
of the development of diabetic macular edema following 1995a;113(1):52–5.
panretinal (scatter) photocoagulation given in 1 or 4 sit- Chew EY, Mills JL, Metzger BE, et al. Metabolic control
tings. Arch Ophthalmol. 2009;127(2):132–40. and progression of retinopathy. The Diabetes in Early
Buabbud JC, Al-latayfeh MM, Sun JK. Optical coherence Pregnancy Study. National Institute of Child Health
tomography imaging for diabetic retinopathy and and Human Development Diabetes in Early Pregnancy
macular edema. Curr Diab Rep. 2010;10(4):264–9. Study. Diabetes Care. 1995b;18(5):631–7.
Byeon SH, Chu YK, Lee H, et al. Foveal ganglion cell layer Chew EY, Ambrosius WT, Davis MD, et al. Effects of
damage in ischemic diabetic maculopathy: correlation medical therapies on retinopathy progression in type 2
of optical coherence tomographic and anatomic changes. diabetes. N Engl J Med. 2010;363(3):233–44.
Ophthalmology. 2009;116(10):|1949–59. Chew EY, Davis MD, Danis RP, et al. The effects of medi-
Byeon SH, Chu YK, Hong YT, et al. New insights into the cal management on the progression of diabetic reti-
pathoanatomy of diabetic macular edema: angio- nopathy in persons with type 2 diabetes: the Action to
graphic patterns and optical coherence tomography. Control Cardiovascular Risk in Diabetes (ACCORD)
Retina. 2012;32(6):1087–99. Eye Study. Ophthalmology. 2014;121(12):2443–51.
Callanan DG, Gupta S, Boyer DS, et al. Dexamethasone Cho YH, Craig ME, Donaghue KC. Puberty as an accel-
intravitreal implant in combination with laser photo- erator for diabetes complications. Pediatr Diabetes.
coagulation for the treatment of diffuse diabetic macu- 2014;15(1):18–26.
lar edema. Ophthalmology. 2013;120(9):1843–51. Clermont AC, Bursell SE. Retinal blood flow in diabetes.
Campochiaro PA, Brown DM, Pearson A, et al. Long- Microcirculation. 2007;14(1):49–61.
term benefit of sustained-delivery fluocinolone ace- Comyn O, Sivaprasad S, Peto T, et al. A randomized trial to
tonide vitreous inserts for diabetic macular edema. assess functional and structural effects of ranibizumab
Ophthalmology. 2011;118(4):626–35. versus laser in diabetic macular edema (the LUCIDATE
Campochiaro PA, Brown DM, Pearson A, et al. Sustained study). Am J Ophthalmol. 2014;157(5):960–70.
delivery fluocinolone acetonide vitreous inserts provide Congdon N, O’Colmain B, Klaver CC, et al. Causes and
benefit for at least 3 years in patients with diabetic mac- prevalence of visual impairment among adults in the
ular edema. Ophthalmology. 2012;119(10):2125–32. United States. Arch Ophthalmol. 2004;122(4):477–85.
Cunha-Vaz JG. The blood-retinal barriers system. Basic for diabetic macular edema. Ophthalmology. 2008;
concepts and clinical evaluation. Exp Eye Res. 115(9):1447–9.
2004;78(3):715–21. Diabetic Retinopathy Study (DRS) Group. Preliminary
Cunha-Vaz J, Ashton P, Iezzi R, et al. Sustained delivery report on effects of photocoagulation therapy. Am
fluocinolone acetonide vitreous implants: long-term J Ophthalmol. 1976;81(4):383–96.
benefit in patients with chronic diabetic macular DRS Group. Photocoagulation treatment of proliferative
edema. Ophthalmology. 2014a;121(10):1892–903. diabetic retinopathy. Clinical application of Diabetic
Cunha-Vaz J, Ribeiro L, Lobo C. Phenotypes and bio- Retinopathy Study (DRS) findings, DRS Report
markers of diabetic retinopathy. Prog Retin Eye Res. Number 8. Ophthalmology. 1981a;88(7):583–600.
2014b;41:90–111. DRS Group. Report Number 6. Design, methods, and
Daien V, Kawasaki R, Villain M, et al. Retinal vascular cali- baseline results. Report Number 7. A modification of
ber is associated with renal function in apparently the Airlie House classification of diabetic retinopathy.
healthy subjects. Acta Ophthalmol. 2013;91(4):e283–8. Prepared by the Diabetic Retinopathy. Invest
Daley ML, Watzke RC, Riddle MC. Early loss of blue- Ophthalmol Vis Sci. 1981b;21(1 Pt 2):1–226.
sensitive color vision in patients with type I diabetes. Ding J, Wong TY. Current epidemiology of diabetic
Diabetes Care. 1987;10(6):777–81. retinopathy and diabetic macular edema. Curr Diab
Danis R, Davis M. Proliferative diabetic retinopathy. In: Rep. 2012;12(4):346–54.
Duh E, editor. Diabetic retinopathy. Totowa: Humana Dmuchowska DA, Krasnicki P, Mariak Z. Can optical
Press; 2008. p. 29–65. coherence tomography replace fluorescein angiogra-
Davis MD, Fisher MR, Gangnon RE, et al. Risk factors phy in detection of ischemic diabetic maculopathy?
for high-risk proliferative diabetic retinopathy and Graefes Arch Clin Exp Ophthalmol. 2014;252(5):
severe visual loss: Early Treatment Diabetic 731–8.
Retinopathy Study Report #18. Invest Ophthalmol Vis Do YK, Eggleston KN. Educational disparities in quality
Sci. 1998;39(2):233–52. of diabetes care in a universal health insurance system:
Diabetes Control and Complications Trial (DCCT) evidence from the 2005 Korea National Health and
Research Group. Color photography vs fluorescein Nutrition Examination Survey. Int J Qual Health Care.
angiography in the detection of diabetic retinopathy in 2011;23(4):397–404.
the diabetes control and complications trial. Arch Do DV, Wang X, Vedula SS, et al. Blood pressure control
Ophthalmol. 1987;105(10):1344–51. for diabetic retinopathy. Cochrane Database Syst Rev.
DCCT Research Group. The relationship of glycemic 2015;1:CD006127.
exposure (HbA1c) to the risk of development and pro- Doft BH, Blankenship G. Retinopathy risk factor regres-
gression of retinopathy in the diabetes control and sion after laser panretinal photocoagulation for prolif-
complications trial. Diabetes. 1995;44(8):968–83. erative diabetic retinopathy. Ophthalmology.
DCCT Research Group. Effect of pregnancy on microvas- 1984;91(12):1453–7.
cular complications in the diabetes control and com- Douvali M, Chatziralli IP, Theodossiadis PG, et al. Effect
plications trial. Diabetes Care. 2000;23(8):1084–91. of macular ischemia on intravitreal ranibizumab treat-
de S, Castilla M. Ocriplasmin for diabetic retinopathy. ment for diabetic macular edema. Ophthalmologica.
Expert Opin Biol Ther. 2013;13(12):1741–47. 2014;232(3):136–43.
De BU, Sacconi R, Pierro L, et al. Optical coherence Dubow M, Pinhas A, Shah N, et al. Classification of human
tomographic hyperreflective foci in early stages retinal microaneurysms using adaptive optics scanning
of diabetic retinopathy. Retina. 2015;35(3): light ophthalmoscope fluorescein angiography. Invest
449–53. Ophthalmol Vis Sci. 2014;55(3):1299–309.
Deissler H, Deissler H, Lang GK, et al. TGFbeta induces Dziuba J, Alperin P, Racketa J, et al. Modeling effects of
transdifferentiation of iBREC to alpha SMA- SGLT-2 inhibitor dapagliflozin treatment versus
expressing cells. Int J Mol Med. 2006;18(4):577–82. standard diabetes therapy on cardiovascular and
Deissler HL, Deissler H, Lang GK, et al. VEGF but not microvascular outcomes. Diabetes Obes Metab. 2014;
PlGF disturbs the barrier of retinal endothelial cells. 16(7):628–35.
Exp Eye Res. 2013;115:162–71. Edwards MS, Wilson DB, Craven TE, et al. Associations
Delgado MF, Dickens CJ, Iwach AG, et al. Long-term between retinal microvascular abnormalities and
results of noncontact neodymium:yttrium-aluminum- declining renal function in the elderly population: the
garnet cyclophotocoagulation in neovascular glau- Cardiovascular Health Study. Am J Kidney Dis.
coma. Ophthalmology. 2003;110(5):895–9. 2005;46(2):214–24.
Diabetes Prevention Program (DPP) Research Group. The Ejaz S, Chekarova I, Ejaz A, et al. Importance of pericytes
prevalence of retinopathy in impaired glucose tolerance and mechanisms of pericyte loss during diabetes reti-
and recent-onset diabetes in the Diabetes Prevention nopathy. Diabetes Obes Metab. 2008;10(1):53–63.
Program. Diabet Med. 2007;24(2):137–44. Elman MJ, Aiello LP, Beck RW, et al. Randomized trial
Diabetic Retinopathy Clinical Research Network (DRCR. evaluating ranibizumab plus prompt or deferred laser
net), A randomized trial comparing intravitreal triam- or triamcinolone plus prompt laser for diabetic macu-
cinolone acetonide and focal/grid photocoagulation lar edema. Ophthalmology. 2010;117(6):1064–77.
Elman MJ, Bressler NM, Qin H, et al. Expanded 2-year jections for 2030. Invest Ophthalmol Vis Sci.
follow-up of ranibizumab plus prompt or deferred laser 2011;52(7):4381–9.
or triamcinolone plus prompt laser for diabetic macular Fiore T, Androudi S, Iaccheri B, et al. Repeatability and
edema. Ophthalmology. 2011;118(4):609–14. reproducibility of retinal thickness measurements in
Elman MJ, Ayala A, Bressler NM, et al. Intravitreal diabetic patients with spectral domain optical coher-
Ranibizumab for diabetic macular edema with prompt ence tomography. Curr Eye Res. 2013;38(6):
versus deferred laser treatment: 5-year randomized 674–9.
trial results. Ophthalmology. 2015;122(2):375–81. Flaxel CJ, Edwards AR, Aiello LP, et al. Factors associ-
Escobar-Barranco JJ, Pina-Marin B, Fernandez-Bonet M. ated with visual acuity outcomes after vitrectomy for
Dexamethasone implants in patients with naive or diabetic macular edema: diabetic retinopathy clinical
refractory diffuse diabetic macular edema. research network. Retina. 2010;30(9):1488–95.
Ophthalmologica. 2015;233:176–85. Fonda SJ, Bursell SE, Lewis DG, et al. The relationship of
Early Treatment Diabetic Retinopathy Study (ETDRS) a diabetes telehealth eye care program to standard eye
Research Group. Photocoagulation for diabetic macu- care and change in diabetes health outcomes. Telemed
lar edema. Early Treatment Diabetic Retinopathy JE Health. 2007;13(6):635–44.
Study report number 1. Arch Ophthalmol. 1985; Fong DS, Ferris III FL, Davis MD, et al. Causes of severe
103(12):1796–806. visual loss in the early treatment diabetic retinopathy
ETDRS Research Group. Grading diabetic retinopathy study: ETDRS report no. 24. Early Treatment Diabetic
from stereoscopic color fundus photographs—an Retinopathy Study Research Group. Am J Ophthalmol.
extension of the modified Airlie House classification 1999;127(2):137–41.
ETDRS report number 10. Ophthalmology. Fong DS, Gottlieb J, Ferris III FL, et al. Understanding
1991a;98(5 Suppl):786–806. the value of diabetic retinopathy screening. Arch
ETDRS Research Group. Classification of diabetic retinop- Ophthalmol. 2001;119(5):758–60.
athy from fluorescein angiograms. ETDRS report num- Fong DS, Strauber SF, Aiello LP, et al. Comparison of the
ber 11. Ophthalmology. 1991b;98(5 Suppl):807–22. modified Early Treatment Diabetic Retinopathy Study
ETDRS Research Group. Effects of aspirin treatment on and mild macular grid laser photocoagulation strate-
diabetic retinopathy. ETDRS report number 8. gies for diabetic macular edema. Arch Ophthalmol.
Ophthalmology. 1991c;98(5 Suppl):757–65. 2007;125(4):469–80.
ETDRS Research Group . Fundus photographic risk Fortun JA, Hubbard III GB. New viscodissection instru-
factors for progression of diabetic retinopathy. ment for use with microincisional vitrectomy in the
ETDRS report number 12. Ophthalmology. treatment of diabetic tractional retinal detachments.
1991;98(5 Suppl):823–33. Arch Ophthalmol. 2011;129(3):352–5.
ETDRS Research Group. Focal photocoagulation treat- Framme C, Walter A, Prahs P, et al. Structural changes of
ment of diabetic macular edema. Relationship of treat- the retina after conventional laser photocoagulation
ment effect to fluorescein angiographic and other and selective retina treatment (SRT) in spectral domain
retinal characteristics at baseline: ETDRS report no. OCT. Curr Eye Res. 2009;34(7):568–79.
19. Arch Ophthalmol. 1995;113(9):1144–55. Frank RN. Systemic therapies for diabetic retinopathy:
Evans JR, Michelessi M, Virgili G. Laser photocoagula- the accord eye study. Ophthalmology. 2014;121(12):
tion for proliferative diabetic retinopathy. Cochrane 2295–6.
Database Syst Rev. 2014;11:CD011234. Friberg TR, Karatza EC. The treatment of macular disease
Ferrannini E, Cushman WC. Diabetes and hypertension: the using a micropulsed and continuous wave 810-nm
bad companions. Lancet. 2012;380(9841):601–10. diode laser. Ophthalmology. 1997;104(12):2030–8.
Ferris III FL. How effective are treatments for diabetic Frier BM. Hypoglycaemia in diabetes mellitus: epidemi-
retinopathy? JAMA. 1993;269(10):1290–1. ology and clinical implications. Nat Rev Endocrinol.
Ferris F. Early photocoagulation in patients with either 2014;10(12):711–22.
type I or type II diabetes. Trans Am Ophthalmol Soc. Fullerton B, Jeitler K, Seitz M, et al. Intensive glucose
1996;94:505–37. control versus conventional glucose control for type 1
Fialho SL, Rego MB, Siqueira RC, et al. Safety and phar- diabetes mellitus. Cochrane Database Syst Rev.
macokinetics of an intravitreal biodegradable implant 2014;2:CD009122.
of dexamethasone acetate in rabbit eyes. Curr Eye Funatsu H, Noma H, Mimura T, et al. Association of vitre-
Res. 2006;31(6):525–34. ous inflammatory factors with diabetic macular
Figueira J, Khan J, Nunes S, et al. Prospective ran- edema. Ophthalmology. 2009;116(1):73–9.
domised controlled trial comparing sub-threshold Garay-Sevilla ME, Porras JS, Malacara JM. Coping strat-
micropulse diode laser photocoagulation and conven- egies and adherence to treatment in patients with type
tional green laser for clinically significant diabetic 2 diabetes mellitus. Rev Invest Clin. 2011;63(2):
macular oedema. Br J Ophthalmol. 2009;93(10): 155–61.
1341–4. Gardner TW, Antonetti DA, Barber AJ, et al. Diabetic
Finger RP, Fimmers R, Holz FG, et al. Incidence of blind- retinopathy: more than meets the eye. Surv
ness and severe visual impairment in Germany: pro- Ophthalmol. 2002;47 Suppl 2:S253–62.
Ghazi NG, Ciralsky JB, Shah SM, et al. Optical coherence Heintz E, Wirehn AB, Peebo BB, et al. Prevalence and
tomography findings in persistent diabetic macular healthcare costs of diabetic retinopathy: a population-
edema: the vitreomacular interface. Am J Ophthalmol. based register study in Sweden. Diabetologia.
2007;144(5):747–54. 2010;53(10):2147–54.
Gillies MC, Lim LL, Campain A, et al. A randomized clinical Helbig H. Surgery for diabetic retinopathy.
trial of intravitreal bevacizumab versus intravitreal dexa- Ophthalmologica. 2007;221(2):103–11.
methasone for diabetic macular edema: the BEVORDEX Hellstedt T, Vesti E, Immonen I. Identification of individ-
study. Ophthalmology. 2014;121(12):2473–81. ual microaneurysms: a comparison between fluores-
Giovannini A, Amato GP, Mariotti C, et al. Diabetic mac- cein angiograms and red-free and colour photographs.
ulopathy induced by vitreo-macular traction: evalua- Graefes Arch Clin Exp Ophthalmol. 1996;234 Suppl
tion by optical coherence tomography (OCT). Doc 1:S13–7.
Ophthalmol. 1999;97(3–4):361–6. Hellstedt T, Kaaja R, Teramo K, et al. The effect of preg-
Goatman KA, Cree MJ, Olson JA, et al. Automated mea- nancy on mild diabetic retinopathy. Graefes Arch Clin
surement of microaneurysm turnover. Invest Exp Ophthalmol. 1997;235(7):437–41.
Ophthalmol Vis Sci. 2003;44(12):5335–41. Hemmingsen B, Lund SS, Gluud C, et al. Intensive glycae-
Grassi MA, Tikhomirov A, Ramalingam S, et al. mic control for patients with type 2 diabetes: systematic
Replication analysis for severe diabetic retinopathy. review with meta-analysis and trial sequential analysis
Invest Ophthalmol Vis Sci. 2012;53(4):2377–81. of randomised clinical trials. BMJ. 2011;343:d6898.
Grunwald JE, Pistilli M, Ying GS, et al. Retinopathy and Hietala K, Harjutsalo V, Forsblom C, et al. Age at onset
progression of CKD: The CRIC study. Clin J Am Soc and the risk of proliferative retinopathy in type 1 dia-
Nephrol. 2014;9(7):1217–24. betes. Diabetes Care. 2010;33(6):1315–9.
Gu K, Cowie CC, Harris MI. Mortality in adults with and Huang OS, Tay WT, Tai ES, et al. Lack of awareness
without diabetes in a national cohort of the U.S. popula- amongst community patients with diabetes and dia-
tion, 1971–1993. Diabetes Care. 1998;21(7):1138–45. betic retinopathy: the Singapore Malay eye study. Ann
Guigou S, Hajjar C, Parrat E, et al. Multicenter Ozurdex(R) Acad Med Singapore. 2009;38(12):1048–55.
assessment for diabetic macular edema: MOZART Hugenschmidt CE, Lovato JF, Ambrosius WT, et al. The
study. J Fr Ophtalmol. 2014;37(6):480–5. cross-sectional and longitudinal associations of dia-
Gulliford MC, Dodhia H, Chamley M, et al. Socio- betic retinopathy with cognitive function and brain
economic and ethnic inequalities in diabetes retinal MRI findings: the Action to Control Cardiovascular
screening. Diabet Med. 2010;27(3):282–8. Risk in Diabetes (ACCORD) trial. Diabetes Care.
Gunderson EP, Lewis CE, Tsai AL, et al. A 20-year prospective 2014;37(12):3244–52.
study of childbearing and incidence of diabetes in young Hutchins E, Coppell KJ, Morris A, et al. Diabetic reti-
women, controlling for glycemia before conception: the nopathy screening in New Zealand requires improve-
Coronary Artery Risk Development in Young Adults ment: results from a multi-centre audit. Aust N Z
(CARDIA) Study. Diabetes. 2007;56(12):2990–6. J Public Health. 2012;36(3):257–62.
Gunduz K, Bakri SJ. Management of proliferative dia- Imai M, Iijima H, Hanada N. Optical coherence tomogra-
betic retinopathy. Compr Ophthalmol Update. phy of tractional macular elevations in eyes with pro-
2007;8(5):245–56. liferative diabetic retinopathy. Am J Ophthalmol.
Gupta B, Sivaprasad S, Wong R, et al. Visual and ana- 2001;132(3):458–61.
tomical outcomes following vitrectomy for complica- Inchiostro S, Candido R, Cavalot F. How can we monitor
tions of diabetic retinopathy: the DRIVE UK study. glycaemic variability in the clinical setting? Diabetes
Eye (Lond). 2012;26(4):510–6. Obes Metab. 2013;15 Suppl 2:13–6.
Haller JA, Qin H, Apte RS, et al. Vitrectomy outcomes in Ip MS, Domalpally A, Sun JK, et al. Long-term effects of
eyes with diabetic macular edema and vitreomacular therapy with ranibizumab on diabetic retinopathy
traction. Ophthalmology. 2010;117(6):1087–93. severity and baseline risk factors for worsening reti-
Hammes HP, Feng Y, Pfister F, et al. Diabetic retinopathy: nopathy. Ophthalmology. 2015;122(2):367–74.
targeting vasoregression. Diabetes. 2011a;60(1):| Ishibashi T, Li X, Koh A, et al. The REVEAL Study:
9–16. ranibizumab monotherapy or combined with laser ver-
Hammes HP, Kerner W, Hofer S, et al. Diabetic retinopa- sus laser monotherapy in Asian patients with diabetic
thy in type 1 diabetes-a contemporary analysis of macular edema. Ophthalmology. 2015;122:1402–15.
8,784 patients. Diabetologia. 2011b;54(8):1977–84. Ishida S, Usui T, Yamashiro K, et al. VEGF164 is proin-
Hayes AJ, Leal J, Gray AM, et al. UKPDS outcomes flammatory in the diabetic retina. Invest Ophthalmol
model 2: a new version of a model to simulate lifetime Vis Sci. 2003;44(5):2155–62.
health outcomes of patients with type 2 diabetes mel- Ito H, Horii T, Nishijima K, et al. Association between
litus using data from the 30 year United Kingdom fluorescein leakage and optical coherence tomo-
Prospective Diabetes Study: UKPDS 82. Diabetologia. graphic characteristics of microaneurysms in diabetic
2013;56(9):1925–33. retinopathy. Retina. 2013;33(4):732–9.
Hayreh SS. Neovascular glaucoma. Prog Retin Eye Res. Jackson TL, Nicod E, Angelis A, et al. Vitreous attach-
2007;26(5):470–85. ment in age-related macular degeneration, diabetic
macular edema, and retinal vein occlusion: a system- Khan MA, Haller JA. Clinical management of vitreomac-
atic review and metaanalysis. Retina. 2013; ular traction. Curr Opin Ophthalmol. 2015;26(3):
33(6):1099–108. 143–8.
Jacobson DR, Murphy RP, Rosenthal AR. The treatment Kiddee W, Trope GE, Sheng L, et al. Intraocular pressure
of angle neovascularization with panretinal photoco- monitoring post intravitreal steroids: a systematic
agulation. Ophthalmology. 1979;86(7):1270–7. review. Surv Ophthalmol. 2013;58(4):291–310.
Jaissle GB, Szurman P, Bartz-Schmidt KU. Ocular side Kiire CA, Porta M, Chong V. Medical management for
effects and complications of intravitreal triamcinolone the prevention and treatment of diabetic macular
acetonide injection. Ophthalmologe. 2004;101(2): edema. Surv Ophthalmol. 2013;58(5):459–65.
121–8. Klaassen I, Van Noorden CJ, Schlingemann RO. Molecular
Jalli PY, Hellstedt TJ, Immonen IJ. Early versus late stain- basis of the inner blood-retinal barrier and its break-
ing of microaneurysms in fluorescein angiography. down in diabetic macular edema and other pathologi-
Retina. 1997;17(3):211–5. cal conditions. Prog Retin Eye Res. 2013;34:19–48.
Javitt JC, Aiello LP. Cost-effectiveness of detecting and Klein R. Screening interval for retinopathy in type 2 dia-
treating diabetic retinopathy. Ann Intern Med. betes. Lancet. 2003;361(9353):190–1.
1996;124(1 Pt 2):164–9. Klein R, Klein BE, Moss SE, et al. The Wisconsin epide-
Kahn HA, Hiller R. Blindness caused by diabetic retinop- miologic study of diabetic retinopathy. II. Prevalence
athy. Am J Ophthalmol. 1974;78(1):58–67. and risk of diabetic retinopathy when age at diagnosis
Kamoi K, Takeda K, Hashimoto K, et al. Identifying risk is less than 30 years. Arch Ophthalmol. 1984;102(4):
factors for clinically significant diabetic macula edema 520–6.
in patients with type 2 diabetes mellitus. Curr Diabetes Klein BE, Moss SE, Klein R. Effect of pregnancy on pro-
Rev. 2013;9(3):209–17. gression of diabetic retinopathy. Diabetes Care.
Kampik A, Kenyon KR, Michels RG, et al. Epiretinal and 1990;13(1):34–40.
vitreous membranes. Comparative study of 56 cases. Klein R, Klein BE, Moss SE. Epidemiology of prolifera-
Arch Ophthalmol. 1981;99(8):1445–54. tive diabetic retinopathy. Diabetes Care. 1992;15(12):
Kandarakis SA, Piperi C, Topouzis F, et al. Emerging role 1875–91.
of advanced glycation-end products (AGEs) in the Klein BE, Klein R, Lee KE. Components of the metabolic
pathobiology of eye diseases. Prog Retin Eye Res. syndrome and risk of cardiovascular disease and dia-
2014;42:85–102. betes in Beaver Dam. Diabetes Care. 2002;25(10):
Kane FE, Burdan J, Cutino A, et al. Iluvien: a new sus- 1790–4.
tained delivery technology for posterior eye disease. Klein R, Knudtson MD, Lee KE, et al. The Wisconsin
Expert Opin Drug Deliv. 2008;5(9):1039–46. Epidemiologic Study of Diabetic Retinopathy XXIII:
Kang H, Su L, Zhang H, et al. Early histological alteration the twenty-five-year incidence of macular edema in
of the retina following photocoagulation treatment in persons with type 1 diabetes. Ophthalmology.
diabetic retinopathy as measured by spectral domain 2009;116(3):497–503.
optical coherence tomography. Graefes Arch Clin Exp Klein R, Lee KE, Gangnon RE, et al. Relation of smok-
Ophthalmol. 2010;248(12):1705–11. ing, drinking, and physical activity to changes in
Karam JG, McFarlane SI. Update on the prevention of vision over a 20-year period: the Beaver Dam Eye
type 2 diabetes. Curr Diab Rep. 2011;11(1):56–63. Study. Ophthalmology. 2014;121(6):1220–8.
Kaul K, Hodgkinson A, Tarr JM, et al. Is inflammation a Klein BE, Myers CE, Howard KP, et al. Serum lipids and
common retinal-renal-nerve pathogenic link in diabe- proliferative diabetic retinopathy and macular edema
tes? Curr Diabetes Rev. 2010;6(5):294–303. in persons with long-term type 1 diabetes mellitus: the
Keech AC, Mitchell P, Summanen PA, et al. Effect of Wisconsin Epidemiologic Study of Diabetic
fenofibrate on the need for laser treatment for diabetic Retinopathy. JAMA Ophthalmol. 2015;133:503–10.
retinopathy (FIELD study): a randomised controlled Klonoff DC, Schwartz DM. An economic analysis of
trial. Lancet. 2007;370(9600):1687–97. interventions for diabetes. Diabetes Care. 2000;23(3):
Kempen JH, O’Colmain BJ, Leske MC, et al. The preva- 390–404.
lence of diabetic retinopathy among adults in the United Kohner EM, Henkind P. Correlation of fluorescein angio-
States. Arch Ophthalmol. 2004;122(4):552–63. gram and retinal digest in diabetic retinopathy. Am
Kern TS. Contributions of inflammatory processes to the J Ophthalmol. 1970;69(3):403–14.
development of the early stages of diabetic retinopa- Kohner EM, Stratton IM, Aldington SJ, et al. Relationship
thy. Exp Diabetes Res. 2007;2007:95103. between the severity of retinopathy and progression to
Kern TS, Engerman RL. A mouse model of diabetic reti- photocoagulation in patients with type 2 diabetes mel-
nopathy. Arch Ophthalmol. 1996;114(8):986–90. litus in the UKPDS (UKPDS 52). Diabet Med.
Khalaf SS, Al-Bdour MD, Al-Till MI. Clinical biomicros- 2001;18(3):178–84.
copy versus fluorescein angiography: effectiveness Korobelnik JF, Do DV, Schmidt-Erfurth U, et al.
and sensitivity in detecting diabetic retinopathy. Eur Intravitreal aflibercept for diabetic macular edema.
J Ophthalmol. 2007;17(1):84–8. Ophthalmology. 2014;121(11):2247–54.
Kozak I, Oster SF, Cortes MA, et al. Clinical evaluation Liegl R, Langer J, Seidensticker F, et al. Comparative
and treatment accuracy in diabetic macular edema evaluation of combined navigated laser photocoagula-
using navigated laser photocoagulator NAVILAS. tion and intravitreal ranibizumab in the treatment of
Ophthalmology. 2011;118(6):1119–24. diabetic macular edema. PLoS One. 2014;9(12):
Kozak I, El-Emam SY, Cheng L, et al. Fluorescein angi- e113981.
ography versus optical coherence tomography- Lim LS, Cheung CM, Mitchell P, et al. Emerging evi-
guided planning for macular laser photocoagulation dence concerning systemic safety of anti-VEGF
in diabetic macular edema. Retina. 2014;34(8): agents—should ophthalmologists be concerned? Am
1600–5. J Ophthalmol. 2011;152(3):329–31.
Kristinsson JK, Gudmundsson JR, Stefansson E, et al. Liu L, Wu X, Liu L, et al. Prevalence of diabetic retinopa-
Screening for diabetic retinopathy. Initiation and fre- thy in mainland China: a meta-analysis. PLoS One.
quency. Acta Ophthalmol Scand. 1995;73(6):525–8. 2012;7(9):e45264.
Kumagai K, Furukawa M, Ogino N, et al. Long-term fol- Liu H, Zhang W, Xu Z, et al. Hyperoxia causes regression
low-up of vitrectomy for diffuse nontractional diabetic of vitreous neovascularization by downregulating
macular edema. Retina. 2009;29(4):464–72. VEGF/VEGFR2 pathway. Invest Ophthalmol Vis Sci.
Kumar A, Duraipandi K, Gogia V, et al. Comparative 2013;54(2):918–31.
evaluation of 23- and 25-gauge microincision vitrec- Lombardo M, Parravano M, Serrao S, et al. Analysis of
tomy surgery in management of diabetic macular trac- retinal capillaries in patients with type 1 diabetes and
tion retinal detachment. Eur J Ophthalmol. 2014; nonproliferative diabetic retinopathy using adaptive
24(1):107–13. optics imaging. Retina. 2013;33(8):1630–9.
Kuppermann BD, Blumenkranz MS, Haller JA, et al. Looker HC, Nyangoma SO, Cromie DT, et al. Rates of
Randomized controlled study of an intravitreous dexa- referable eye disease in the Scottish National Diabetic
methasone drug delivery system in patients with per- Retinopathy Screening Programme. Br J Ophthalmol.
sistent macular edema. Arch Ophthalmol. 2014;98(6):790–5.
2007;125(3):309–17. Loprinzi PD, Brodowicz GR, Sengupta S, et al.
Laidlaw DA. Vitrectomy for diabetic macular oedema. Accelerometer-assessed physical activity and diabetic
Eye (Lond). 2008;22(10):1337–41. retinopathy in the United States. JAMA Ophthalmol.
Lang GE. Diabetic macular edema. Ophthalmologica. 2014;132(8):1017–9.
2012;227 Suppl 1:21–9. Loukovaara S, Harju M, Kaaja R, et al. Retinal capillary
Lang GE. Mechanisms of retinal neurodegeneration as a blood flow in diabetic and nondiabetic women during
result of diabetes mellitus. Klin Monbl Augenheilkd. pregnancy and postpartum period. Invest Ophthalmol
2013;230(9):929–31. Vis Sci. 2003;44(4):1486–91.
Lauszus FF, Klebe JG, Bek T, et al. Increased serum IGF-I Lovestam-Adrian M, Agardh E. Photocoagulation of
during pregnancy is associated with progression of diabetic macular oedema—complications and visual
diabetic retinopathy. Diabetes. 2003;52(3):852–6. outcome. Acta Ophthalmol Scand. 2000;78(6):
LeCaire TJ, Palta M, Klein R, et al. Assessing progress in 667–71.
retinopathy outcomes in type 1 diabetes: comparing Luke J, Nassar K, Luke M, et al. Ranibizumab as adjuvant
findings from the Wisconsin Diabetes Registry Study in the treatment of rubeosis iridis and neovascular
and the Wisconsin Epidemiologic Study of Diabetic glaucoma—results from a prospective interventional
Retinopathy. Diabetes Care. 2013;36(3):631–7. case series. Graefes Arch Clin Exp Ophthalmol.
Lee CM, Olk RJ. Modified grid laser photocoagulation for 2013;251(10):2403–13.
diffuse diabetic macular edema. Long-term visual Lung JC, Swann PG, Chan HH. Early local functional
results. Ophthalmology. 1991;98(10):1594–602. changes in the human diabetic retina: a global flash
Lee WJ, Sobrin L, Kang MH, et al. Ischemic diabetic retinopa- multifocal electroretinogram study. Graefes Arch Clin
thy as a possible prognostic factor for chronic kidney dis- Exp Ophthalmol. 2012;250(12):1745–54.
ease progression. Eye (Lond). 2014;28(9):1119–25. Luo D, Zheng Z, Xu X, et al. Systematic review of various
Leicht SF, Kernt M, Neubauer A, et al. Microaneurysm laser intervention strategies for proliferative diabetic
turnover in diabetic retinopathy assessed by automated retinopathy. Expert Rev Med Devices. 2015;12(1):
RetmarkerDR image analysis—potential role as bio- 83–91.
marker of response to ranibizumab treatment. Lutty GA. Effects of diabetes on the eye. Invest
Ophthalmologica. 2014;231(4):198–203. Ophthalmol Vis Sci. 2013;54(14):ORSF81–7.
Li CS, Neu MB, Shaw LC, et al. EPCs and pathological Ly A, Scheerer MF, Zukunft S, et al. Retinal proteome
angiogenesis: when good cells go bad. Microvasc Res. alterations in a mouse model of type 2 diabetes.
2010;79(3):207–16. Diabetologia. 2014;57(1):192–203.
Li LX, Li MF, Lu JX, et al. Retinal microvascular abnor- Maberley DA, Koushik A, Cruess AF. Factors associated
malities are associated with early carotid atheroscle- with missed eye examinations in a cohort with diabe-
rotic lesions in hospitalized Chinese patients with type tes. Can J Public Health. 2002;93(3):229–32.
2 diabetes mellitus. J Diabetes Complications. Maclennan PA, McGwin Jr G, Heckemeyer C, et al. Eye
2014;28(3):378–85. care use among a high-risk diabetic population seen in
a public hospital’s clinics. JAMA Ophthalmol. Moss SE, Klein R, Klein BE. Ten-year incidence of visual
2014;132(2):162–7. loss in a diabetic population. Ophthalmology.
Madsen-Bouterse SA, Kowluru RA. Oxidative stress and 1994;101(6):1061–70.
diabetic retinopathy: pathophysiological mechanisms Murakami T, Uji A, Ogino K, et al. Association between
and treatment perspectives. Rev Endocr Metab Disord. perifoveal hyperfluorescence and serous retinal
2008;9(4):315–27. detachment in diabetic macular edema.
Maeshima K, Utsugi-Sutoh N, Otani T, et al. Progressive Ophthalmology. 2013;120(12):2596–603.
enlargement of scattered photocoagulation scars in Nam S, Chesla C, Stotts NA, et al. Barriers to diabetes
diabetic retinopathy. Retina. 2004;24(4):507–11. management: patient and provider factors. Diabetes
Maheshwary AS, Oster SF, Yuson RM, et al. The associa- Res Clin Pract. 2011;93(1):1–9.
tion between percent disruption of the photoreceptor Narayan KM, Boyle JP, Thompson TJ, et al. Lifetime risk
inner segment-outer segment junction and visual acu- for diabetes mellitus in the United States. JAMA.
ity in diabetic macular edema. Am J Ophthalmol. 2003;290(14):1884–90.
2010;150(1):63–7. Narayan KM, Boyle JP, Thompson TJ, et al. Effect of
Marticorena J, Romano V, Gomez-Ulla F. Sterile endo- BMI on lifetime risk for diabetes in the U.S. Diabetes
phthalmitis after intravitreal injections. Mediators Care. 2007;30(6):1562–6.
Inflamm. 2012;2012:928123. Navarro A, Pournaras JA, Hoffart L, et al. Vitrectomy may
Martinez-Zapata MJ, Marti-Carvajal AJ, Sola I, et al. prevent the occurrence of diabetic macular oedema.
Anti-vascular endothelial growth factor for prolifera- Acta Ophthalmol. 2010;88(4):483–5.
tive diabetic retinopathy. Cochrane Database Syst Rev. Nawaz MI, Van RK, Mohammad G, et al. Autocrine
2014;11:CD008721. CCL2, CXCL4, CXCL9 and CXCL10 signal in retinal
Massin P, Peto T, Ansquer JC, et al. Effects of fenofibric endothelial cells and are enhanced in diabetic retinop-
acid on diabetic macular edema: the MacuFen study. athy. Exp Eye Res. 2013;109:67–76.
Ophthalmic Epidemiol. 2014;21(5):307–17. Nehme A, Edelman J. Dexamethasone inhibits high glu-
Matsuda S, Tam T, Singh RP, et al. Impact of insulin treat- cose-, TNF-alpha-, and IL-1beta-induced secretion of
ment in diabetic macular edema therapy in type 2 dia- inflammatory and angiogenic mediators from retinal
betes. Can J Diabetes. 2015;39(1):73–7. microvascular pericytes. Invest Ophthalmol Vis Sci.
Matsuo T. Disappearance of diabetic macular hard 2008;49(5):2030–8.
exudates after hemodialysis introduction. Acta Med Neubauer AS, Langer J, Liegl R, et al. Navigated macular
Okayama. 2006;60(3):201–5. laser decreases retreatment rate for diabetic macular
Matsuoka M, Ogata N, Minamino K, et al. Leukostasis edema: a comparison with conventional macular laser.
and pigment epithelium-derived factor in rat models of Clin Ophthalmol. 2013;7:121–8.
diabetic retinopathy. Mol Vis. 2007;13:1058–65. Nguyen QD, Brown DM, Marcus DM, et al. Ranibizumab
Mazhar K, Varma R, Choudhury F, et al. Severity of dia- for diabetic macular edema: results from 2 phase III
betic retinopathy and health-related quality of life: the randomized trials: RISE and RIDE. Ophthalmology.
Los Angeles Latino Eye Study. Ophthalmology. 2012;119(4):789–801.
2011;118(4):649–55. Nguyen-Khoa BA, Goehring EL, Werther W, et al.
Mehlsen J, Erlandsen M, Poulsen PL, et al. Individualized Hospitalized cardiovascular events in patients with dia-
optimization of the screening interval for diabetic betic macular edema. BMC Ophthalmol. 2012;12:11.
retinopathy: a new model. Acta Ophthalmol. Nielsen LR, Pedersen-Bjergaard U, Thorsteinsson B,
2012;90(2):109–14. et al. Hypoglycemia in pregnant women with type 1
Messenger WB, Beardsley RM, Flaxel CJ. Fluocinolone ace- diabetes: predictors and role of metabolic control.
tonide intravitreal implant for the treatment of diabetic Diabetes Care. 2008;31(1):9–14.
macular edema. Drug Des Devel Ther. 2013;7:425–34. Nisic F, Turkovic S, Mavija M, et al. Correlation between
Meyer-Schwickerath G. Historical perspective of photo- the findings of optical coherent retinal tomography
coagulation (in retinal vascular diseases). Doc (OCT), stereo biomicroscopic images from fundus of
Ophthalmol. 1977;44(1):77–9. an eye and values from visual acuity of diabetic macu-
Miller JW, Le CJ, Strauss EC, et al. Vascular endothelial lar edema. Acta Inform Med. 2014;22(4):232–6.
growth factor a in intraocular vascular disease. Nunes S, Pires I, Rosa A, et al. Microaneurysm turnover
Ophthalmology. 2013;120(1):106–14. is a biomarker for diabetic retinopathy progression to
Misra A, Bachmann MO, Greenwood RH, et al. Trends in clinically significant macular edema: findings for
yield and effects of screening intervals during 17 years type 2 diabetics with nonproliferative retinopathy.
of a large UK community-based diabetic retinopathy Ophthalmologica. 2009;223(5):292–7.
screening programme. Diabet Med. 2009;26(10): Ogata N, Tombran-Tink J, Jo N, et al. Upregulation of
1040–7. pigment epithelium-derived factor after laser photoco-
Mitchell P, Bandello F, Schmidt-Erfurth U, et al. The agulation. Am J Ophthalmol. 2001;132(3):427–9.
RESTORE study: ranibizumab monotherapy or com- Ola MS, Nawaz MI, Siddiquei MM, et al. Recent advances
bined with laser versus laser monotherapy for diabetic in understanding the biochemical and molecular
macular edema. Ophthalmology. 2011;118(4): mechanism of diabetic retinopathy. J Diabetes
615–25. Complications. 2012;26(1):56–64.
Olsen TW, Adelman RA, Flaxel CJ, et al. Preferred prac- Diabetes Prevention Program Outcomes Study.
tice pattern diabetic retinopathy. Am Acad Ophthalmol. Lancet. 2012;379(9833):2243–51.
2014. Pershing S, Enns EA, Matesic B, et al. Cost-effectiveness
Ophir A. Full-field 3-D optical coherence tomography of treatment of diabetic macular edema. Ann Intern
imaging and treatment decision in diffuse diabetic Med. 2014;160(1):18–29.
macular edema. Invest Ophthalmol Vis Sci. Pfeiffer A, Spranger J, Meyer-Schwickerath R, et al.
2014;55(5):3052–3. Growth factor alterations in advanced diabetic reti-
Ophir A, Trevino A, Fatum S. Extrafoveal vitreous trac- nopathy: a possible role of blood retina barrier
tion associated with diabetic diffuse macular oedema. breakdown. Diabetes. 1997;46 Suppl 2:S26–30.
Eye (Lond). 2010;24(2):347–53. Phelps RL, Sakol P, Metzger BE, et al. Changes in dia-
Ostri C, Lux A, Lund-Andersen H, et al. Long-term betic retinopathy during pregnancy. Correlations with
results, prognostic factors and cataract surgery after regulation of hyperglycemia. Arch Ophthalmol.
diabetic vitrectomy: a 10-year follow-up study. Acta 1986;104(12):1806–10.
Ophthalmol. 2014;92(6):571–6. Pires I, Santos AR, Nunes S, et al. Subclinical macular
Palermo A, Maggi D, Maurizi AR, et al. Prevention of edema as a predictor of progression to clinically signifi-
type 2 diabetes mellitus: is it feasible? Diabetes Metab cant macular edema in type 2 diabetes. Ophthalmologica.
Res Rev. 2014;30 Suppl 1:4–12. 2013;230(4):201–6.
Panozzo G, Parolini B, Gusson E, et al. Diabetic macular Pokroy R, Desai UR, Du E, et al. Bevacizumab prior to
edema: an OCT-based classification. Semin vitrectomy for diabetic traction retinal detachment.
Ophthalmol. 2004;19(1–2):13–20. Eye (Lond). 2011;25(8):989–97.
Park JH, Woo SJ, Ha YJ, et al. Effect of vitrectomy on Qi HP, Bi S, Wei SQ, et al. Intravitreal versus subtenon
macular microcirculation in patients with diffuse triamcinolone acetonide injection for diabetic macular
diabetic macular edema. Graefes Arch Clin Exp edema: a systematic review and meta-analysis. Curr
Ophthalmol. 2009;247(8):1009–17. Eye Res. 2012;37(12):1136–47.
Park YG, Kim EY, Roh YJ. Laser-based strategies to Qiao Q, Keinanen-Kiukaanniemi S, Laara E. The rela-
treat diabetic macular edema: history and new promis- tionship between hemoglobin levels and diabetic
ing therapies. J Ophthalmol. 2014;2014:769213. retinopathy. J Clin Epidemiol. 1997;50(2):153–8.
Parravano M, Menchini F, Virgili G. Antiangiogenic ther- Rahman W, Rahman FZ, Yassin S, et al. Progression of
apy with anti-vascular endothelial growth factor retinopathy during pregnancy in type 1 diabetes mel-
modalities for diabetic macular oedema. Cochrane litus. Clin Experiment Ophthalmol. 2007;35(3):231–6.
Database Syst Rev. 2009;4:CD007419. Rajendram R, Fraser-Bell S, Kaines A, et al. A 2-year pro-
Patel JI, Hykin PG, Schadt M, et al. Pars plana vitrectomy spective randomized controlled trial of intravitreal
with and without peeling of the inner limiting mem- bevacizumab or laser therapy (BOLT) in the manage-
brane for diabetic macular edema. Retina. 2006; ment of diabetic macular edema: 24-month data:
26(1):5–13. report 3. Arch Ophthalmol. 2012;130(8):972–9.
Paz SH, Varma R, Klein R, et al. Noncompliance with Rangasamy S, McGuire PG, Franco NC, et al. Chemokine
vision care guidelines in Latinos with type 2 diabetes mediated monocyte trafficking into the retina: role of
mellitus: the Los Angeles Latino Eye Study. inflammation in alteration of the blood-retinal barrier
Ophthalmology. 2006;113(8):1372–7. in diabetic retinopathy. PLoS One. 2014;9(10):
Pearce E, Sivaprasad S, Chong NV. Factors affecting e108508.
reading speed in patients with diabetic macular edema Rasmussen KL, Laugesen CS, Ringholm L, et al.
treated with laser photocoagulation. PLoS One. 2014; Progression of diabetic retinopathy during pregnancy
9(9):e105696. in women with type 2 diabetes. Diabetologia.
Pearson PA, Comstock TL, Ip M, et al. Fluocinolone ace- 2010;53(6):1076–83.
tonide intravitreal implant for diabetic macular edema: Raum P, Peto T, Pfeiffer N, Wild P, Hoehn R, Ponto K,
a 3-year multicenter, randomized, controlled clinical Schneider A, Schulz A, Lamparter J, Mirshahi
trial. Ophthalmology. 2011;118(8):1580–7. A. Prevalence and cardiovascular risk factors of dia-
Pei-Pei W, Shi-Zhou H, Zhen T, et al. Randomised clini- betic retinopathy: results from the Gutenberg Health
cal trial evaluating best-corrected visual acuity and Study. Invest Ophthalmol Vis Sci. 2014;55:5339
central macular thickness after 532-nm subthreshold (Abstract).
laser grid photocoagulation treatment in diabetic mac- Regnier S, Malcolm W, Allen F, et al. Efficacy of anti-
ular oedema. Eye (Lond). 2015;29(3):313–22. VEGF and laser photocoagulation in the treatment of
Pelletier EM, Shim B, Ben-Joseph R, et al. Economic out- visual impairment due to diabetic macular edema: a
comes associated with microvascular complications of systematic review and network meta-analysis. PLoS
type 2 diabetes mellitus: results from a US claims data One. 2014;9(7):e102309.
analysis. Pharmacoeconomics. 2009;27(6):479–90. Ringholm L, Mathiesen ER, Kelstrup L, et al. Managing
Perreault L, Pan Q, Mather KJ, et al. Effect of regression type 1 diabetes mellitus in pregnancy—from planning
from prediabetes to normal glucose regulation on to breastfeeding. Nat Rev Endocrinol. 2012;8(11):
long-term reduction in diabetes risk: results from the 659–67.
Robinson R, Barathi VA, Chaurasia SS, et al. Update on diabetic retinopathy in the primary care environment.
animal models of diabetic retinopathy: from molecular Ophthalmic Surg Lasers Imaging. 2005;36(1):46–56.
approaches to mice and higher mammals. Dis Model Schmidt-Erfurth U, Lang GE, Holz FG, et al. Three-year
Mech. 2012;5(4):444–56. outcomes of individualized ranibizumab treatment in
Rodrigues M, Xin X, Jee K, et al. VEGF secreted by patients with diabetic macular edema: the RESTORE
hypoxic Muller cells induces MMP-2 expression and extension study. Ophthalmology. 2014;121(5):
activity in endothelial cells to promote retinal neovas- 1045–53.
cularization in proliferative diabetic retinopathy. Schultz KL, Birnbaum AD, Goldstein DA. Ocular disease
Diabetes. 2013;62(11):3863–73. in pregnancy. Curr Opin Ophthalmol. 2005;16(5):
Roider J, Brinkmann R, Wirbelauer C, et al. Subthreshold 308–14.
(retinal pigment epithelium) photocoagulation in mac- Schwartz SG, Brantley Jr MA, Flynn Jr HW. Genetics and dia-
ular diseases: a pilot study. Br J Ophthalmol. 2000; betic retinopathy. Curr Diabetes Rev. 2013;9(1):86–92.
84(1):40–7. Schwartz SG, Flynn Jr HW, Scott IU. Emerging drugs for
Roider J, Liew SH, Klatt C, et al. Selective retina therapy diabetic macular edema. Expert Opin Emerg Drugs.
(SRT) for clinically significant diabetic macular 2014;19(3):397–405.
edema. Graefes Arch Clin Exp Ophthalmol. 2010; Sebag J. Abnormalities of human vitreous structure in
248(9):1263–72. diabetes. Graefes Arch Clin Exp Ophthalmol.
Romano MR, Gibran SK, Marticorena J, et al. Can a pre- 1993;231(5):257–60.
operative bevacizumab injection prevent recurrent Shah SM, Nguyen QD, Mir HS, et al. A randomized,
postvitrectomy diabetic vitreous haemorrhage? Eye double-masked controlled clinical trial of Sandostatin
(Lond). 2009;23(8):1698–701. long-acting release depot in patients with postsurgical
Romero-Aroca P, Sagarra-Alamo R, Basora-Gallisa J, cystoid macular edema. Retina. 2010;30(1):160–6.
et al. Prospective comparison of two methods of Sheetz MJ, Aiello LP, Davis MD, et al. The effect of the
screening for diabetic retinopathy by nonmydriatic oral PKC beta inhibitor ruboxistaurin on vision loss in
fundus camera. Clin Ophthalmol. 2010;4:1481–8. two phase 3 studies. Invest Ophthalmol Vis Sci.
Rosberger DF. Diabetic retinopathy: current concepts and 2013;54(3):1750–7.
emerging therapy. Endocrinol Metab Clin North Am. Shin HJ, Shin KC, Chung H, et al. Change of retinal nerve
2013;42(4):721–45. fiber layer thickness in various retinal diseases treated
Roy MS, Klein R. Macular edema and retinal hard exudates with multiple intravitreal antivascular endothelial
in African Americans with type 1 diabetes: the New growth factor. Invest Ophthalmol Vis Sci. 2014;
Jersey 725. Arch Ophthalmol. 2001;119(2):251–9. 55(4):2403–11.
Russo A, Morescalchi F, Costagliola C, et al. Comparison Shorb SR. Anemia and diabetic retinopathy. Am
of smartphone ophthalmoscopy with slit-lamp biomi- J Ophthalmol. 1985;100(3):434–6.
croscopy for grading diabetic retinopathy. Am Silva KC, Pinto CC, Biswas SK, et al. Hypertension
J Ophthalmol. 2015;159(2):360–4. increases retinal inflammation in experimental diabe-
Sacu S, Pemp B, Weigert G, et al. Response of retinal vessels tes: a possible mechanism for aggravation of diabetic
and retrobulbar hemodynamics to intravitreal anti-VEGF retinopathy by hypertension. Curr Eye Res.
treatment in eyes with branch retinal vein occlusion. 2007;32(6):533–41.
Invest Ophthalmol Vis Sci. 2011;52(6):3046–50. Silva PS, Sun JK, Al-Latayfeh MM, Sanchez CR, Walia
Sander B, Larsen M, Engler C, et al. Early changes in dia- S, Tolson AM, Cavallerano JD, Aiello LP. Unawareness
betic retinopathy: capillary loss and blood-retina bar- of vision threatening diabetic retinopathy and appro-
rier permeability in relation to metabolic control. Acta priate followup in a tertiary telemedicine eye care pro-
Ophthalmol (Copenh). 1994;72(5):553–9. gram. Invest Ophthalmol Vis Sci. 2010;51:5059
Sander B, Best J, Johansen S, et al. Fluorescein transport (Abstract).
through the blood-aqueous and blood-retinal barriers Silva PS, Cavallerano JD, Haddad NM, et al. Peripheral
in diabetic macular edema. Curr Eye Res. lesions identified on ultrawide field imaging predict
2003;27(4):247–52. increased risk of diabetic retinopathy progression over
Santos JM, Mohammad G, Zhong Q, et al. Diabetic reti- 4 years. Ophthalmology. 2015;122(5):949–56.
nopathy, superoxide damage and antioxidants. Curr Sim DA, Keane PA, Zarranz-Ventura J, et al. Predictive
Pharm Biotechnol. 2011;12(3):352–61. factors for the progression of diabetic macular isch-
Scanlon PH, Aldington SJ, Stratton IM. Epidemiological emia. Am J Ophthalmol. 2013a;156(4):684–92.
issues in diabetic retinopathy. Middle East Afr Sim DA, Keane PA, Zarranz-Ventura J, et al. The effects
J Ophthalmol. 2013;20(4):293–300. of macular ischemia on visual acuity in diabetic reti-
Scaramuzzi M, Querques G, Spina C, et al. Repeated nopathy. Invest Ophthalmol Vis Sci. 2013b;54(3):
intravitreal dexamethasone implant for diabetic macu- 2353–60.
lar edema. Retina. 2015;35:1216–22. Sim DA, Keane PA, Fung S, et al. Quantitative analysis of
Schiffman RM, Jacobsen G, Nussbaum JJ, et al. diabetic macular ischemia using optical coherence
Comparison of a digital retinal imaging system and tomography. Invest Ophthalmol Vis Sci. 2014;
seven-field stereo color fundus photography to detect 55(1):417–23.
Simha A, Braganza A, Abraham L, et al. Anti-vascular Taylor E, Dobree JH. Proliferative diabetic retinopathy.
endothelial growth factor for neovascular glaucoma. Site and size of initial lesions. Br J Ophthalmol.
Cochrane Database Syst Rev. 2013;10:CD007920. 1970;54(1):11–8.
Simo R, Roy S, Behar-Cohen F, et al. Fenofibrate: a new The Diabetes Control and Complications Trial/
treatment for diabetic retinopathy. Molecular mecha- Epidemiology of Diabetes Interventions and
nisms and future perspectives. Curr Med Chem. Complications Research Group. Retinopathy and
2013;20(26):3258–66. nephropathy in patients with type 1 diabetes four years
Simunovic MP, Hunyor AP, Ho IV. Vitrectomy for dia- after a trial of intensive therapy. N Engl J Med.
betic macular edema: a systematic review and meta- 2000;342(6):381–9.
analysis. Can J Ophthalmol. 2014;49(2):188–95. The Royal College of Ophthalmologists. Diabetic
Sivak-Callcott JA, O’Day DM, Gass JD, et al. Evidence- Retinopathy Guidelines. London: Royal College of
based recommendations for the diagnosis and treatment Ophthalmologists; 2012.
of neovascular glaucoma. Ophthalmology. 2001;108(10): Theodossiadis PG, Theodoropoulou S, Neamonitou G,
1767–76. et al. Hemodialysis-induced alterations in macular
Snead DR, James S, Snead MP. Pathological changes in thickness measured by optical coherence tomography
the vitreoretinal junction 1: epiretinal membrane for- in diabetic patients with end-stage renal disease.
mation. Eye (Lond). 2008;22(10):1310–7. Ophthalmologica. 2012;227(2):90–4.
Snow V, Weiss KB, Mottur-Pilson C. The evidence base Thulliez M, Angoulvant D, Le Lez ML, et al.
for tight blood pressure control in the management of Cardiovascular events and bleeding risk associated
type 2 diabetes mellitus. Ann Intern Med. with intravitreal antivascular endothelial growth fac-
2003;138(7):587–92. tor monoclonal antibodies: systematic review and
Sohn EH, He S, Kim LA, et al. Angiofibrotic response to meta-analysis. JAMA Ophthalmol. 2014;132(11):
vascular endothelial growth factor inhibition in dia- 1317–26.
betic retinal detachment: report no. 1. Arch Tokuyama T, Ikeda T, Sato K. Effects of haemodialysis on
Ophthalmol. 2012;130(9):1127–34. diabetic macular leakage. Br J Ophthalmol. 2000;
Sonoda S, Sakamoto T, Shirasawa M, et al. Correlation 84(12):1397–400.
between reflectivity of subretinal fluid in OCT images Ugurlu N, Gerceker S, Yulek F, et al. The levels of the
and concentration of intravitreal VEGF in eyes with circulating cellular adhesion molecules ICAM-1,
diabetic macular edema. Invest Ophthalmol Vis Sci. VCAM-1 and endothelin-1 and the flow-mediated
2013;54(8):5367–74. vasodilatation values in patients with type 1 diabetes
Sophie R, Lu N, Campochiaro PA. Predictors of func- mellitus with early-stage diabetic retinopathy. Intern
tional and anatomic outcomes in patients with diabetic Med. 2013;52(19):2173–8.
macular edema treated with Ranibizumab. UK Prospective Diabetes Study (UKPDS) Group. Tight
Ophthalmology. 2015;122:1395–401. blood pressure control and risk of macrovascular and
Stamler J, Vaccaro O, Neaton JD, et al. Diabetes, other microvascular complications in type 2 diabetes:
risk factors, and 12-yr cardiovascular mortality for UKPDS 38. BMJ. 1998;317(7160):703–13.
men screened in the multiple risk factor intervention Varma R, Torres M, Pena F, et al. Prevalence of diabetic
trial. Diabetes Care. 1993;16(2):434–44. retinopathy in adult Latinos: the Los Angeles Latino
Stefansson E. Ocular oxygenation and the treatment of eye study. Ophthalmology. 2004;111(7):1298–306.
diabetic retinopathy. Surv Ophthalmol. 2006;51(4): Venkatesh P, Ramanjulu R, Azad R, et al. Subthreshold
364–80. micropulse diode laser and double frequency neodym-
Stewart MW. Anti-VEGF therapy for diabetic macular ium: YAG laser in treatment of diabetic macular
edema. Curr Diab Rep. 2014;14(8):510. edema: a prospective, randomized study using multi-
Stitt AW. AGEs and diabetic retinopathy. Invest focal electroretinography. Photomed Laser Surg.
Ophthalmol Vis Sci. 2010;51(10):4867–74. 2011;29(11):727–33.
Su CC, Yang CH, Yeh PT, et al. Macular tractional reti- Verma A, Raman R, Vaitheeswaran K, et al. Does neuro-
noschisis in proliferative diabetic retinopathy: clinical nal damage precede vascular damage in subjects with
characteristics and surgical outcome. Ophthalmologica. type 2 diabetes mellitus and having no clinical diabetic
2014;231(1):23–30. retinopathy? Ophthalmic Res. 2012;47(4):202–7.
Tahrani AA, Bailey CJ, Del PS, et al. Management of type Vestgaard M, Ringholm L, Laugesen CS, et al. Pregnancy-
2 diabetes: new and future developments in treatment. induced sight-threatening diabetic retinopathy in
Lancet. 2011;378(9786):182–97. women with type 1 diabetes. Diabet Med. 2010;
Tamborlane WV, Beck RW, Bode BW, et al. Continuous 27(4):431–5.
glucose monitoring and intensive treatment of type 1 Vijan S, Hofer TP, Hayward RA. Cost-utility analysis of
diabetes. N Engl J Med. 2008;359(14):1464–76. screening intervals for diabetic retinopathy in patients
Tang J, Kern TS. Inflammation in diabetic retinopathy. with type 2 diabetes mellitus. JAMA. 2000;283(7):
Prog Retin Eye Res. 2011;30(5):343–58. 889–96.
Tarr JM, Kaul K, Chopra M, et al. Pathophysiology of dia- Virgili G, Parravano M, Menchini F, et al. Anti-vascular
betic retinopathy. ISRN Ophthalmol. 2013;2013: endothelial growth factor for diabetic macular oedema.
343560. Cochrane Database Syst Rev. 2014;10:CD007419.
Voykov B, Bartz-Schmidt KU. Dislocation of dexametha- Yilmaz T, Weaver CD, Gallagher MJ, et al. Intravitreal
sone intravitreous implant. Arch Ophthalmol. triamcinolone acetonide injection for treatment of
2012;130(6):706. refractory diabetic macular edema: a systematic
Vujosevic S, Bini S, Midena G, et al. Hyperreflective review. Ophthalmology. 2009;116(5):902–11.
intraretinal spots in diabetics without and with non- Yuen J, Clark A, Ng JQ, et al. Further survey of Australian
proliferative diabetic retinopathy: an in vivo study ophthalmologist’s diabetic retinopathy management:
using spectral domain OCT. J Diabetes Res. did practice adhere to National Health and Medical
2013;2013:491835. Research Council guidelines? Clin Exp Ophthalmol.
Wang H, Chhablani J, Freeman WR, et al. Characterization 2010;38(6):613–9.
of diabetic microaneurysms by simultaneous fluores- Zabetian A, Keli HM, Echouffo-Tcheugui JB, et al.
cein angiography and spectral-domain optical coher- Diabetes in the Middle East and North Africa. Diabetes
ence tomography. Am J Ophthalmol. 2012a; Res Clin Pract. 2013;101(2):106–22.
153(5):861–7. Zandbergen AA, Vogt L, de ZD, et al. Change in albuminuria
Wang ZY, Zhao KK, Zhao DS, et al. Dissection under per- is predictive of cardiovascular outcome in normotensive
fluorocarbon liquid: a modified vitrectomy technique patients with type 2 diabetes and microalbuminuria.
for diabetic tractional retinal detachment. Retina. Diabetes Care. 2007;30(12):3119–21.
2012b;32(4):848–52. Zechmeister-Koss I, Huic M. Vascular endothelial growth
Wang Y, Muqit MM, Stanga PE, et al. Spatial changes of factor inhibitors (anti-VEGF) in the management of
central field loss in diabetic retinopathy after laser. diabetic macular oedema: a systematic review. Br
Optom Vis Sci. 2014;91(1):111–20. J Ophthalmol. 2012;96(2):167–78.
Wells JA, Glassman AR, Ayala AR, et al. Aflibercept, Zenoni S, Comi N, Fontana P. Individualised treatment of
bevacizumab, or ranibizumab for diabetic macular proliferative diabetic retinopathy: optimal surgical
edema. N Engl J Med. 2015;372(13):1193–203. timing improves long-term outcomes. EPMA J.
West SK, Klein R, Rodriguez J, et al. Diabetes and dia- 2010;1(1):78–81.
betic retinopathy in a Mexican-American popula- Zhang W, Liu H, Rojas M, et al. Anti-inflammatory ther-
tion: Proyecto VER. Diabetes Care. 2001; apy for diabetic retinopathy. Immunotherapy.
24(7):1204–9. 2011;3(5):609–28.
Wilson DJ, Finkelstein D, Quigley HA, et al. Macular grid Zhang L, Chen B, Tang L. Metabolic memory: mecha-
photocoagulation. An experimental study on the pri- nisms and implications for diabetic retinopathy.
mate retina. Arch Ophthalmol. 1988;106(1):100–5. Diabetes Res Clin Pract. 2012;96(3):286–93.
Witkin SR, Klein R. Ophthalmologic care for persons Zhang X, Zhao J, Zhao T, et al. Effects of intensive glyce-
with diabetes. JAMA. 1984;251(19):2534–7. mic control in ocular complications in patients with
Wong CW, Wong TY, Cheng CY, et al. Kidney and eye type 2 diabetes: a meta-analysis of randomized clini-
diseases: common risk factors, etiological mecha- cal trials. Endocrine. 2015;49:78–89.
nisms, and pathways. Kidney Int. 2014;85(6): Zheng Y, Kwong MT, Maccormick IJ, et al. A compre-
1290–302. hensive texture segmentation framework for segmen-
Yalvac IS, Eksioglu U, Satana B, et al. Long-term results tation of capillary non-perfusion regions in fundus
of Ahmed glaucoma valve and Molteno implant in fluorescein angiograms. PLoS One. 2014;9(4):
neovascular glaucoma. Eye (Lond). 2007;21(1): e93624.
65–70. Zhong Y, Li J, Chen Y, et al. Activation of endoplasmic
Yamagishi S, Ueda S, Matsui T, et al. Role of advanced reticulum stress by hyperglycemia is essential for
glycation end products (AGEs) and oxidative stress in Muller cell-derived inflammatory cytokine production
diabetic retinopathy. Curr Pharm Des. 2008;14(10): in diabetes. Diabetes. 2012;61(2):492–504.
962–8. Zhu Y, Zhang T, Wang K, et al. Changes in choroidal
Yamamoto T, Takeuchi S, Sato Y, et al. Long-term follow- thickness after panretinal photocoagulation in patients
up results of pars plana vitrectomy for diabetic macu- with type 2 diabetes. Retina. 2015;35(4):695–703.
lar edema. Jpn J Ophthalmol. 2007;51(4):285–91. Zucchiatti I, Lattanzio R, Querques G, et al. Intravitreal
Yau JW, Rogers SL, Kawasaki R, et al. Global prevalence dexamethasone implant in patients with persistent
and major risk factors of diabetic retinopathy. Diabetes diabetic macular edema. Ophthalmologica. 2012;
Care. 2012;35(3):556–64. 228(2):117–22.
Retinal Vein Occlusion
7
Amelie Pielen, Bernd Junker, and Nicolas Feltgen
neovascularization of the iris and of the retina.

7.1 Epidemiology and Clinical Untreated ischemic RVO may result in vitreous
Features hemorrhage and secondary glaucoma. The lon-
ger the duration of macular edema, the more
Retinal vein occlusion (RVO) is the second most damage is done to photoreceptors and retinal
common vascular eye disorder after diabetic reti- pigment epithelium (Lardenoye et al. 2000).
nopathy. The prevalence of RVO is 5.2 per 1000 Anti-vascular endothelial growth factor (VEGF)
estimated from studies in the United States, therapy leads to prompt resolution of macular
Europe, Asia, and Australia. Branch retinal vein edema and early treatment may therefore result
occlusion (BRVO) occurs more often than central in significantly better results than delayed treat-
retinal vein occlusion (CRVO) (4.42 versus 0.80 ment (Pielen et al. 2013).
per 1000, respectively) (Rogers et al. 2010). The
incidence rises with age and was reported to be
0.7 % under the age of 60 years compared to 7.2 Pathophysiology
4.6 % over 80 years (Rogers et al. 2010).
Clinical characteristics are dilated and The exact etiology of RVO is still unknown.
engorged retinal veins, macular edema, retinal Patients with RVO often present with risk factors
hemorrhages, retinal ischemia manifested as associated with arterial vascular diseases, and
cotton-wool spots and peripheral ischemia visi- pathophysiology differs from systemic vein
ble in fluorescein angiography as capillary non- occlusions such as deep vein thrombosis (Hayreh
perfusion. Macular edema results in impairment 1994; Cugati et al. 2006). It is supposed that the
of visual acuity (VA) while ischemia leads to process starts with a compression of the vein
wall. BRVO typically occurs at arteriovenous
crossings (Feltgen et al. 2010) and CRVO at the
level of the lamina cribrosa (Hayreh et al. 1994).
A. Pielen (*) • B. Junker The compression causes turbulence and reduc-
Hannover Medical School, University Eye Hospital, tion of the venous blood flow which leads to par-
Carl-Neuberg-Str.1, Hannover 30625, Germany
e-mail: Pielen.Amelie@mh-hannover.de
tial thrombus formation. As a consequence, flow
is further reduced, venous pressure rises and
N. Feltgen
Universitätsmedizin Göttingen, University Eye
blood and plasma exudate into the surrounding
Hospital, Robert-Koch-Str. 40, Göttingen 37075, tissue causing edema. In the later stages, reduced
Germany retinal perfusion causes hypoxia that triggers the

132 A. Pielen et al.
transcription of multiple factors, among them drawbacks are cataract progression and rise in
VEGF and inflammatory cytokines (Noma et al. intraocular pressure (IOP).
2006; Ehlken et al. 2011). VEGF further increases
macular edema and contributes to the formation 7.3.1.1 Dexamethasone
of neovascularization. Dexamethasone is a potent soluble corticoste-
Risk factors for RVO are glaucoma, metabolic roid, but half-life after intravitreal injection is
syndrome, and its components diabetes mellitus, very short. The development of a sustained-
arterial hypertension, hyperlipidemia, and ciga- release biodegradable implant (Ozurdex®, Pharm
rette smoke (Stem et al. 2013). A higher inci- Allergan GmbH) overcame this deficit. Results
dence of CRVO is found in patients with history of the randomized controlled clinical trial (RCT)
of stroke and peripheral artery disease (Stem GENEVA (Global Evaluation of Implantable
et al. 2013; Zhou et al. 2013). Increased levels of Dexamethasone in Retinal Vein Occlusion with
atherosclerotic and thrombophilic factors (factor Macular Edema) led to approval of Ozurdex
VIII) were found to increase the risk of develop- 0.7 mg by FDA and EMA for the treatment of
ing an ischemic CRVO (Sodi et al. 2011). macular edema due to RVO (Haller et al. 2010,
Therapy in RVO patients should therefore not 2011).
only focus on ocular treatment but also include a
systematic work-up to identify any underlying Results of GENEVA: A single application of
systemic disease and initiate or optimize treat- Ozurdex 0.7 mg was compared to Ozurdex
ment of systemic risk factors. 0.35 mg versus sham injection over 6 months.
During the second 6 months, Ozurdex 0.7 mg
was given in an open extension of the study.
7.3 Intravitreal Therapy Ozurdex 0.7 mg reduced macular edema and led
to a three-line improvement in best-corrected VA
Visual impairment due to macular edema is (BCVA) in 29.6 % of patients at day 60 compared
treated with anti-inflammatory and antiangio- to 12.5 % in the sham group (p < 0.001). The
genic intravitreal agents targeting vascular maximum effect as well as the maximum rise in
permeability and leakage. Substances currently IOP were seen at 2 months after injection and
in clinical use are either corticosteroids or decreased thereafter. Consequently, evaluation of
anti-VEGF. 12-month data showed no significant difference
in BCVA (Pielen et al. 2013). Regarding adverse
events, cataract progression was observed in
7.3.1 Corticosteroids 10.7 % of patients receiving one implant
(sham/0.7 mg Ozurdex) compared to 29.8 % of
Triamcinolone acetonide, dexamethasone, and patients after two Ozurdex 0.7 mg implants ver-
fluocinolone have been investigated in eyes sus 5.7 % of sham patients (Haller et al. 2011;
affected with macular edema due to RVO (Ip Pielen et al. 2013). IOP rise occurred in every
et al. 2009; Scott et al. 2009; Haller et al. 2010; third patient after Ozurdex implant at day 60 and
Jain et al. 2012). Corticosteroids reduce the was most often treated with local anti-
expression of VEGF, act as anti-inflammatory glaucomatous therapy, but 1.3 % of patients
agents, inhibit vascular permeability, leakage, underwent glaucoma surgery (versus none in
and the breakdown of the blood–retinal barrier sham-treated patients).
(Zhang et al. 2008; Wang et al. 2008; McAllister Duration of macular edema has a direct
et al. 2009; Kunikata et al. 2012). A direct neuro- impact on the results of intravitreal treatment.
protective effect was reported (Jeanneteau et al. In GENEVA, only 16.4 % of patients were
2008). Intravitreal treatment with corticosteroids treated after a disease duration <3 months, the
improves VA by reducing macular edema. Main majority presented with macular edema dura-
7 Retinal Vein Occlusion 133
tion between 3 and 6 months (51.3 %) or longer led to a significantly higher percentage of visual
(32.3 %). In trials investigating triamcinolone gain (≥15 letters) in 26.5 % (1 mg), 25.6 %
(SCORE BRVO/CRVO) and anti-VEGF agents (4 mg) compared to sham treatment (6.8 %,
(BRAVO, CRUISE, GALILEO, COPERNICUS) p = 0.001). BCVA did not increase but was stabi-
the proportion of patients with a duration of lized at baseline levels in triamcinolone groups,
macular edema <3 months was significantly while sham patients lost BCVA. SCORE BRVO
higher: SCORE BRVO > 50 %, SCORE CRVO did not show a significant difference in VA gain
36 %, BRAVO 51.5–53.8 %, CRUISE 51.5– at 12 months (26 % 1 mg, 27 % 4 mg, 29 % grid
61.5 %. Secondary analysis of all trials shows laser coagulation). Regarding adverse events,
that early treatment of macular edema second- SCORE investigators found IOP rise and cataract
ary to RVO is more effective than delayed treat- progression as most frequent events. The adverse
ment (Coscas et al. 2011; Pielen et al. 2013). effects seem to be dose dependent, 20 % of
Therefore, intravitreal agents should be com- patients who received 1 mg triamcinolone needed
pared in head-to-head trials guaranteeing com- antiglaucomatous medication versus 35 % of
parable baseline characteristics of patient patients in the 4 mg group (8 % observation).
populations and results of different RCTs Glaucoma surgery was performed in 2.2 %
should be compared with caution. (1 mg) and 4 % (4 mg).
7.3.1.2 Fluocinolone
A fluocinolone implant is under investigation that 7.3.2 VEGF-Inhibitors
showed efficacy up to 36 months after single
application (Jain et al. 2012). Results in 24 eyes Anti-VEGF agents act specifically: Bevacizumab
of 23 patients with chronic macular edema due to (Avastin®, Roche) and ranibizumab (Lucentis®,
CRVO showed improvement of BCVA of 4.5 let- Novartis) inhibit VEGF-A, pegaptanib sodium
ters at 12 months, 8.2 at 24 months, and 3.4 at (Macugen®, Pfizer) binds to the isoform 165 of
36 months. Reduction of central retinal thickness VEGF-A, aflibercept (Eylea®, Bayer Health Care)
was sustained. During the 3 years, all phakic eyes is a receptor fusion protein which binds all VEGF-A
underwent cataract surgery and 5/24 eyes isoforms and placental growth factor (PIGF). The
received glaucoma surgery. The fluocinolone primary effect of anti-VEGF in RVO is not antian-
implant is approved for second-line treatment in giogenic but agents reduce vessel permeability and
diabetic macular edema but not in RVO. macular edema leading to improvement of VA.
VEGF levels are increased after CRVO and
7.3.1.3 Triamcinolone Acetonide BRVO and vitreous levels of VEGF as well as
Intravitreal injection of triamcinolone acetonide interleukin-1 correlate with the amount of mac-
is off-label use. Triamcinolone is crystalline and ular edema (Noma et al. 2005, 2006, 2008).
is either injected in formulations with There is evidence that different splicing variants
METHOCEL or as commercially available of VEGF 165 might act controversial as either
Kenalog® (Bristol-Myers Squibb) or Trivaris® pro-angiogenic isoform (VEGF 165) or antian-
(Pharm Allergan, Inc.). giogenic isoform (VEGF 165b) (Harper and
The Standard Care versus Corticosteroid for Bates 2008). In vitreous samples of RVO
Retinal Vein Occlusion (SCORE) study investi- patients, the ratio of VEGF 165b/VEGF 165
gated intravitreal triamcinolone 1 mg versus was altered towards a pro-angiogenic shift in
4 mg versus observation in CRVO and versus correlation to the severity of the occlusion
grid laser coagulation in BRVO (Ip et al. 2009; (Ehlken et al. 2011). Targeting different iso-
Scott et al. 2009). Retreatment could be given forms of VEGF might be a feasible approach in
every 4 months. In SCORE CRVO, triamcinolone the future.
7.3.2.1 Ranibizumab events was low, especially no signs of an

Safety and efficacy of ranibizumab were investi- increased incidence of arteriothrombotic events
gated in the CRUISE (Ranibizumab for the were noted.
Treatment of Macular Edema After Central Ranibizumab patients were followed up until
Retinal Vein Occlusion Study) and BRAVO 2 years in the HORIZON trial and showed sus-
(Ranibizumab for the Treatment of Macular tained visual improvement with a relatively low
Edema Following Branch Retinal Vein mean number of ranibizumab injection in year 2
Occlusion) RCTs (Brown et al. 2010; (CRVO 3.5/0.5 mg and 2.9/sham + 0.5 mg; BRVO
Campochiaro et al. 2010). In both pivotal trials, 2.1/0.5 mg and 2/sham + 0.5 mg) (Heier et al.
6 monthly intravitreal injections of ranibizumab 2012). The longest follow-up currently is
0.5 mg (or 0.3 mg) were superior to sham injec- 48 months in the RETAIN study (Campochiaro
tions and led to improvement of BCVA of +14.9 et al. 2014). Visual improvement at 6 months
letters (CRVO 0.5 mg) and +18.3 letters (BRVO could be maintained until 48 months in both
0.5 mg) compared to sham (CRVO 0.8 letters, CRVO (+14.0 letters) and BRVO (+20.1 letters).
BRVO +7.3 letters, Table 7.1). Improvement The percentage of patients with complete resolu-
could be stabilized over a second period of tion of macular edema over ≥6 months was 50 %
6 months by pro re nata injections (CRVO +13.9 (BRVO) and 44 % (CRVO), of these patients,
letters, BRVO +18.3 letters) (Brown et al. 2011; 76 % (BRVO) and 71 % (CRVO) had received
Campochiaro et al. 2011). Patients in sham their last ranibizumab injection within ≤2 years.
groups received intravitreal ranibizumab 0.5 mg The response to intravitreal ranibizumab might
from month 6 onward. This delay in treatment of hint at the prognosis for VA development:
macular edema led to a significantly reduced Improvement of VA was significantly higher in
response at month 12 (CRVO sham/ranibizumab patients presenting with completely resolved
0.5 mg +7.3 letters, BRVO +12.1 letters sham/ macular edema versus unresolved edema over 48
ranibizumab 0.5 mg) compared to 6-month data months (BRVO +25.9/+17.1 letters resolved/
supporting the hypothesis that long-standing unresolved, CRVO +25.2/+4.3 letters, respec-
macular edema leads to irreversible structural tively). Mean number of injections was 3.2
damages. The rate of ocular and systemic adverse (BRVO) and 5.9 (CRVO) in year 4.
Table 7.1 Change in best-corrected visual acuity in letters in randomized controlled trials (RCT) investigating anti-
VEGF agents for macular edema secondary to central retinal vein occlusion
12 24 48
RCT Agent 6 months months months months
CRUISE (1) Ranibizumab +14.9 +13.9 HORIZON +12.0 RETAIN +14.0
(2) (3)
Sham +0.8 Sham +7.3 +7.6
+ Ran.
Epstein (4) Bevacizumab +14.1 +16.0
Sham −2.0 Sham +4.6
+ Bev.
COPERNICUS (5) Aflibercept +17.3 +16.2 +13.0
sham −4.0 Sham +3.8 +1.5
+ Afl.
GALILEO (6) Aflibercept +18.0 +16.9 +13.1
+3.8
Sham +3.3 Sham + Afl. +6.2
(1) Brown et al. (2010), Campochiaro et al. (2011), (2) Heier et al. (2012), (3) Campochiaro et al. (2014), (4) Epstein
et al. (2012a, b), (5) Boyer et al. (2012), Brown et al. (2013), Heier et al. (2014), (6) Holz et al. (2013), Korobelnik et al.
(2014), Ogura et al. (2014)
Comparison to Grid Laser Treatment or EMA (Wroblewski et al. 2009, 2010). In

in BRVO CRVO, VA gain ≥15 letters at month 6 was not
BRAVO did not compare intravitreal ranibi- significantly higher in pegaptanib groups
zumab to the gold standard of grid laser photoco- compared to sham (36 %/0.3 mg versus
agulation set forth by the Branch Vein Occlusion 39 %/1 mg versus 28 %/sham) (Wroblewski
Study (1984). Two investigator-initiated RCTs et al. 2009). In BRVO, pegaptanib 0.3 mg and
overcame this gap of knowledge (Tan et al. 2014; 1 mg led to BCVA improvement of +14 letters,
Pielen et al. 2014). Tan et al. found that BCVA at but the study lacked comparison to sham
12 months was significantly higher after ranibi- (Wroblewski et al. 2010).
zumab (6 monthly + PRN, +12.5 letters) com-
pared to standard-of-care grid laser treatment 7.3.2.4 Aflibercept
(−1.6 letters). Rate of additional grid laser treat- Monthly VEGF Trap-Eye (aflibercept) for macu-
ment was low in the ranibizumab group at 13 and lar edema secondary to CRVO was investigated
25 weeks (6.7 %, 8.3 % versus 68.4 %, 50 % in over 6 months compared to sham treatment in the
controls). Ranibizumab for Branch Retinal Vein parallel RCTs COPERNICUS and GALILEO
Occlusion Associated Macular Edema Study (Boyer et al. 2012; Holz et al. 2013). In
(RABAMES) compared ranibizumab (3 monthly COPERNICUS, patients from both groups
injections + observation) with grid laser (at base- received intravitreal aflibercept on PRN basis
line and 8 weeks optional) versus a combination from week 24 onward with monthly (weeks
(Pielen et al. 2014). Best results were found in 24–52) and quarterly evaluation (weeks 52–100).
anti-VEGF-treated patients. The combination of In GALILEO, sham patients could receive active
ranibizumab and grid laser did not lead to an treatment from week 52 onward, while afliber-
additional effect and did not protect from recur- cept patients received further injections PRN
rence of macular edema. from week 24 onward. Evaluation intervals were
monthly (weeks 24–52) and every other month
7.3.2.2 Bevacizumab (weeks 52–76).
There are many publications on off-label bevaci- COPERNICUS: At 6 months, the proportion
zumab in RVO indicating similar results com- of patients with VA gain ≥15 letters was 56.1 %
pared to ranibizumab in RVO, but the evidence for aflibercept 2 mg compared to 12.3 % sham,
level is low due to predominantly retrospective BCVA improvement was +17.3 letters (afliber-
clinical observations. There is one RCT investi- cept) versus loss of 4.0 letters (sham, Table 7.1)
gating bevacizumab every 6 weeks versus sham (Boyer et al. 2012). At 1 year, after all patients
until month 6 in macular edema due to CRVO could receive intravitreal aflibercept PRN, 55.3 %
with an open label extension until month 12 of aflibercept patients and 30.1 % of sham/
(Epstein et al. 2012a, b). BCVA significantly aflibercept showed VA gain ≥15 letters (Brown
increased in the bevacizumab group versus sham et al. 2013). Again, sham patients did not profit
(6 months: +14.1 versus −2.0 letters, Table 7.1) from delayed anti-VEGF treatment as much as
(Epstein et al. 2012b). Previously sham-treated patients with prompt treatment, BCVA change at
patients did not improve as pronounced during month 12 was +16.2 letters (aflibercept) versus
the 6-month extension (12 months: +16.0 versus +3.8 letters (sham/aflibercept). At 2 years, VA
+4.6 letters), mimicking CRUISE results. gain was 49.1 % (aflibercept) versus 23.3 %
(sham/aflibercept) and change in BCVA was
7.3.2.3 Pegaptanib +13.0 versus +1.5 letters (Heier et al. 2014).
Pegaptanib sodium was investigated for macu- Even over the period of 18 months active afliber-
lar edema secondary to CRVO and BRVO in cept treatment, patients from the sham/afliber-
phase II dose-ranging RCTs but was not inves- cept group did not catch up with those treated
tigated further and was not approved by FDA from baseline on.
GALILEO: At month 6, the proportion of response to treatment before the first injection
patients gaining ≥15 letters was 60.2 % (afliber- remain to be determined and results of head-to-
cept 2 mg monthly) versus 22.1 % (sham) (Holz head studies may provide conclusive results for
et al. 2013). At month 12, the ratio was similarly comparison of the various agents.
pronounced (60.2 %/aflibercept versus 32.4 %/
sham) because sham patients received sham
injections until week 52 (Korobelnik et al. 2014).
From weeks 52 to 76, patients were evaluated Compliance with Ethical Requirements
every 8 weeks and both groups received afliber- Author Amelie Pielen has received a speaker
cept 2 mg as needed (Ogura et al. 2014). After honorarium from Pharm Allergan GmbH,
12 months of sham treatment, the ratio of patients Novartis Pharma GmbH and Bayer Health
who gained ≥15 letters remained limited despite Care. Bernd Junker has received a speaker
aflibercept treatment (29.4 %/sham + aflibercept honorarium from Novartis Pharma GmbH
versus 57.3 % aflibercept), while macular edema and Bayer Health Care. Nicolas Feltgen
reduction was pronounced. One more clinical received a speaker honorarium from
fact to stress the hypothesis that long-standing Novartis Pharma GmbH, Pharm Allergan
edema damages macular function and limits GmbH, Bayer Health Care, and research
treatment outcome. grants from Novartis Pharma GmbH.
No human or animal studies were car-
ried out by the authors of this chapter.
7.4 Implications for Clinical
Practice
Results from RCTs demonstrate the efficacy and

safety of both corticosteroids and anti-VEGF
References
agents for macular edema secondary to Boyer D, Heier J, Brown DM, et al. Vascular endothelial
RVO. Comparison of 12-month data suggests growth factor Trap-Eye for macular edema secondary to
that VEGF-inhibitors might be superior to corti- central retinal vein occlusion: six-month results of the
costeroids, but comparison of results is limited phase 3 COPERNICUS study. Ophthalmology.
2012;119:1024–32. doi:10.1016/j.ophtha.2012.01.042.
because of the lack of head-to-head trials, differ- Branch Vein Occlusion Study Group. Argon laser photo-
ences in duration of macular edema at baseline coagulation for macular edema in branch vein occlu-
and different treatment regimens (Pielen et al. sion. Am J Ophthalmol. 1984;98:271–82.
2013). Corticosteroids, especially slow-release Brown DM, Campochiaro PA, Singh RP, et al. Ranibizumab
for macular edema following central retinal vein occlu-
devices such as the dexamethasone implant allow sion: six-month primary end point results of a phase III
elongated injection intervals, but they are accom- study. Ophthalmology. 2010;117:1124–33.e1.
panied by a relatively high rate of the ocular doi:10.1016/j.ophtha.2010.02.022.
adverse events cataract progression and glau- Brown DM, Campochiaro PA, Bhisitkul RB, et al. Sustained
benefits from ranibizumab for macular edema following
coma. Intravitreal anti-VEGF agents come with branch retinal vein occlusion: 12-month outcomes of a
little ocular and systemic events, but are accom- phase III study. Ophthalmology. 2011;118:1594–602.
panied by the burden of considerably higher doi:10.1016/j.ophtha.2011.02.022.
injection frequency and the need for monthly Brown DM, Heier JS, Clark WL, et al. Intravitreal afliber-
cept injection for macular edema secondary to central
controls in a pro re nata regimen. Long-term retinal vein occlusion: 1-year results from the Phase 3
results, however, demonstrate that this commit- COPERNICUS Study. Am J Ophthalmol.
ment is worth the effort: the sooner the macular 2013;155:429–7.e7. doi:10.1016/j.ajo.2012.09.026.
edema is treated, the better the structural response Campochiaro PA, Heier JS, Feiner L, et al. Ranibizumab
for macular edema following branch retinal vein
(resolved edema) and the longer the consequent occlusion: six-month primary end point results of a
retreatment the better is the prognosis for sus- phase III study. Ophthalmology. 2010;117:1102–12.
tained visual acuity. Factors to predict patients’ e1. doi:10.1016/j.ophtha.2010.02.021.
Campochiaro PA, Brown DM, Awh CC, et al. Sustained study. Ophthalmology. 2014;121:1414–20.e1.
benefits from ranibizumab for macular edema follow- doi:10.1016/j.ophtha.2014.01.027.
ing central retinal vein occlusion: twelve-month out- Holz FG, Roider J, Ogura Y, et al. VEGF Trap-Eye for
comes of a phase III study. Ophthalmology. macular oedema secondary to central retinal vein
2011;118:2041–9. doi:10.1016/j.ophtha.2011.02.038. occlusion: 6-month results of the phase III GALILEO
Campochiaro PA, Sophie R, Pearlman J, et al. Long-term study. Br J Ophthalmol. 2013;97:278–84. doi:10.1136/
outcomes in patients with retinal vein occlusion treated bjophthalmol-2012-301504.
with ranibizumab: the RETAIN study. Ophthalmology. Ip MS, Scott IU, VanVeldhuisen PC, et al. A randomized
2014;121:209–19. doi:10.1016/j.ophtha.2013.08.038. trial comparing the efficacy and safety of intravitreal
Coscas G, Loewenstein A, Augustin A, et al. Management triamcinolone with observation to treat vision loss
of retinal vein occlusion—consensus document. associated with macular edema secondary to central
Ophthalmologica. 2011;226:4–28. doi:10.1159/000327391. retinal vein occlusion: the Standard Care vs
Cugati S, Wang JJ, Rochtchina E, Mitchell P. Ten-year Corticosteroid for Retinal Vein Occlusion (SCORE)
incidence of retinal vein occlusion in an older popula- study report 5. Arch Ophthalmol. 2009;127:1101–14.
tion: the Blue Mountains Eye Study. Arch Ophthalmol. doi:10.1001/archophthalmol.2009.234.
2006;124:726–32. doi:10.1001/archopht.124.5.726. Jain N, Stinnett SS, Jaffe GJ. Prospective study of a fluo-
Ehlken C, Rennel ES, Michels D, et al. Levels of VEGF cinolone acetonide implant for chronic macular edema
but not VEGF(165b) are increased in the vitreous of from central retinal vein occlusion: thirty-six-month
patients with retinal vein occlusion. Am J Ophthalmol. results. Ophthalmology. 2012;119:132–7.
2011;152:298–303.e1. doi:10.1016/j.ajo.2011.01.040. doi:10.1016/j.ophtha.2011.06.019.
Epstein DL, Algvere PV, von Wendt G, et al. Benefit from Jeanneteau F, Garabedian MJ, Chao MV. Activation of
bevacizumab for macular edema in central retinal vein Trk neurotrophin receptors by glucocorticoids pro-
occlusion: twelve-month results of a prospective, ran- vides a neuroprotective effect. Proc Natl Acad Sci U S
domized study. Ophthalmology. 2012a;119:2587–91. A. 2008;105:4862–7. doi:10.1073/pnas.0709102105.
doi:10.1016/j.ophtha.2012.06.037. Korobelnik J-F, Holz FG, Roider J, et al. Intravitreal
Epstein DLJ, Algvere PV, von Wendt G, et al. Bevacizumab aflibercept injection for macular edema resulting from
for macular edema in central retinal vein occlusion: a central retinal vein occlusion: One-year results of the
prospective, randomized, double-masked clinical Phase 3 GALILEO study. Ophthalmology.
study. Ophthalmology. 2012b;119:1184–9. 2014;121:202–8. doi:10.1016/j.ophtha.2013.08.012.
doi:10.1016/j.ophtha.2012.01.022. Kunikata H, Shimura M, Nakazawa T, et al. Chemokines
Feltgen N, Pielen A, Hansen L, et al. Intravitreal drug therapy in aqueous humour before and after intravitreal triam-
for retinal vein occlusion—pathophysiological mecha- cinolone acetonide in eyes with macular oedema asso-
nisms and routinely used drugs. Klin Monbl Augenheilkd. ciated with branch retinal vein occlusion. Acta
2010;227:681–93. doi:10.1055/s-0029-1245606. Ophthalmol. 2012;90:162–7. doi:10.1111/j.1755-3768.
Haller JA, Bandello F, Belfort Jr R, et al. Randomized, 2010.01892.x.
sham-controlled trial of dexamethasone intravitreal Lardenoye CW, Probst K, DeLint PJ, Rothova
implant in patients with macular edema due to retinal A. Photoreceptor function in eyes with macular
vein occlusion. Ophthalmology. 2010;117:1134–46. edema. Invest Ophthalmol Vis Sci. 2000;41:4048–53.
e1. doi:10.1016/j.ophtha.2010.03.032. McAllister IL, Vijayasekaran S, Chen SD, Yu D-Y. Effect
Haller JA, Bandello F, Belfort Jr R, et al. Dexamethasone of triamcinolone acetonide on vascular endothelial
intravitreal implant in patients with macular edema growth factor and occludin levels in branch retinal
related to branch or central retinal vein occlusion vein occlusion. Am J Ophthalmol. 2009;147:838–46.
twelve-month study results. Ophthalmology. e1–2. doi:10.1016/j.ajo.2008.12.006.
2011;118:2453–60. doi:10.1016/j.ophtha.2011.05.014. Noma H, Funatsu H, Yamasaki M, et al. Pathogenesis of
Harper SJ, Bates DO. VEGF-A splicing: the key to anti- macular edema with branch retinal vein occlusion and
angiogenic therapeutics? Nat Rev Cancer. 2008;8:880– intraocular levels of vascular endothelial growth factor
7. doi:10.1038/nrc2505. and interleukin-6. Am J Ophthalmol. 2005;140:256–
Hayreh SS. Retinal vein occlusion. Indian J Ophthalmol. 61. doi:10.1016/j.ajo.2005.03.003.
1994;42:109–32. Noma H, Minamoto A, Funatsu H, et al. Intravitreal lev-
Hayreh SS, Zimmerman MB, Podhajsky P. Incidence of els of vascular endothelial growth factor and interleu-
various types of retinal vein occlusion and their recur- kin-6 are correlated with macular edema in branch
rence and demographic characteristics. Am retinal vein occlusion. Graefes Arch Clin Exp
J Ophthalmol. 1994;117:429–41. Ophthalmol. 2006;244:309–15. doi:10.1007/s00417-
Heier JS, Campochiaro PA, Yau L, et al. Ranibizumab for 004-1087-4.
macular edema due to retinal vein occlusions: long- Noma H, Funatsu H, Mimura T, Hori S. Changes of vas-
term follow-up in the HORIZON trial. Ophthalmology. cular endothelial growth factor after vitrectomy for
2012;119:802–9. doi:10.1016/j.ophtha.2011.12.005. macular edema secondary to retinal vein occlusion.
Heier JS, Clark WL, Boyer DS, et al. Intravitreal aflibercept Eur J Ophthalmol. 2008;18:1017–9.
injection for macular edema due to central retinal vein Ogura Y, Roider J, Korobelnik J-F, et al. Intravitreal
occlusion: two-year results from the COPERNICUS aflibercept for macular edema secondary to central
retinal vein occlusion: 18-month results of the Phase 3 vein occlusion. Ophthalmology. 2013;120:362–70.
GALILEO Study. Am J Ophthalmol. 2014. doi:10.1016/j.ophtha.2012.07.080.
doi:10.1016/j.ajo.2014.07.027. Tan MH, McAllister IL, Gillies ME, et al. Randomized
Pielen A, Feltgen N, Isserstedt C, et al. Efficacy and safety of controlled trial of intravitreal ranibizumab versus stan-
intravitreal therapy in macular edema due to branch and dard grid laser for macular edema following branch
central retinal vein occlusion: a systematic review. PLoS retinal vein occlusion. Am J Ophthalmol.
One. 2013;8, e78538. doi:10.1371/journal.pone.0078538. 2014;157:237–47.e1. doi:10.1016/j.ajo.2013.08.013.
Pielen A, Mirshahi A, Feltgen N, et al. Ranibizumab for Wang K, Wang Y, Gao L, et al. Dexamethasone inhibits
Branch Retinal Vein Occlusion Associated Macular leukocyte accumulation and vascular permeability in
Edema Study (RABAMES): six-month results of a retina of streptozotocin-induced diabetic rats via
prospective randomized clinical trial. Acta reducing vascular endothelial growth factor and inter-
Ophthalmol. 2014. doi:10.1111/aos.12488. cellular adhesion molecule-1 expression. Biol Pharm
Rogers S, McIntosh RL, Cheung N, et al. The prevalence Bull. 2008;31:1541–6.
of retinal vein occlusion: pooled data from population Wroblewski JJ, Wells 3rd JA, Adamis AP, et al. Pegaptanib
studies from the United States, Europe, Asia, and sodium for macular edema secondary to central retinal
Australia. Ophthalmology. 2010;117:313–9.e1. vein occlusion. Arch Ophthalmol. 2009;127:374–80.
doi:10.1016/j.ophtha.2009.07.017. doi:10.1001/archophthalmol.2009.14.
Scott IU, Ip MS, VanVeldhuisen PC, et al. A randomized Wroblewski JJ, Wells 3rd JA, Gonzales CR. Pegaptanib
trial comparing the efficacy and safety of intravitreal sodium for macular edema secondary to branch retinal
triamcinolone with standard care to treat vision loss vein occlusion. Am J Ophthalmol. 2010;149:147–54.
associated with macular Edema secondary to branch doi:10.1016/j.ajo.2009.08.005.
retinal vein occlusion: the Standard Care vs Zhang X, Bao S, Lai D, et al. Intravitreal triamcinolone
Corticosteroid for Retinal Vein Occlusion (SCORE) acetonide inhibits breakdown of the blood-retinal
study report 6. Arch Ophthalmol. 2009;127:1115–28. barrier through differential regulation of
doi:10.1001/archophthalmol.2009.233. VEGF-A and its receptors in early diabetic rat reti-
Sodi A, Giambene B, Marcucci R, et al. Atherosclerotic nas. Diabetes. 2008;57:1026–33. doi:10.2337/
and thrombophilic risk factors in patients with isch- db07-0982.
emic central retinal vein occlusion. Retina. Zhou JQ, Xu L, Wang S, et al. The 10-year incidence and
2011;31:724–9. doi:10.1097/IAE.0b013e3181eef419. risk factors of retinal vein occlusion: the Beijing eye
Stem MS, Talwar N, Comer GM, Stein JD. A longitudinal study. Ophthalmology. 2013;120:803–8. doi:10.1016/j.
analysis of risk factors associated with central retinal ophtha.2012.09.033.
Pharmacokinetics of Intravitreally
Applied VEGF Inhibitors
8
Tim U. Krohne, Frank G. Holz,
and Carsten H. Meyer
differ considerably in terms of their molecular

8.1 Ocular Pharmacokinetics structure and consequently also their pharmaco-
following Intravitreal kinetic properties.
Injection Pharmacokinetic studies on intravitreal
VEGF inhibitors have so far been conducted pri-
The introduction of inhibitors of vascular endo- marily on rabbits and rhesus monkeys. Both ani-
thelial growth factor (VEGF) in ophthalmology mal species have a distinctly smaller ocular
represents a breakthrough in the treatment of volume than humans so that the pharmacokinetic
numerous retinal diseases, some of which have parameters determined usually differ consider-
been untreatable before. Four medicines of this ably from those of humans. The ocular half-lives
substance class have so far been used by ophthal- of VEGF inhibitors derived from animal experi-
mologists: pegaptanib (Macugen), bevacizumab ments are usually around half of the respective
(Avastin), ranibizumab (Lucentis), and afliber- human values. The pharmacokinetic values
cept (Eylea). Although these substances are determined in animal models cannot therefore be
directed against the same target molecule, they directly transferred to humans. However, they
are suitable for comparing the pharmacokinetic
properties of different substances if they have
Original Publication (in German):
been determined in the same animal model under
Tim U. Krohne, Frank G. Holz, Carsten H. Meyer. identical conditions.
Pharmakokinetik intravitreal applizierter VEGF- The ocular half-lives measured in rabbits and
Inhibitoren. Der Ophthalmologe. February 2014, Volume humans for the different VEGF inhibitors and
111, Issue 2, pp 113-120. doi: 10.1007/s00347-013-
2932-9. © Springer-Verlag Berlin Heidelberg 2013.
other selected intravitreally administered sub-
Republication with kind permission of Springer stances are summarized in Tables 8.1 and 8.2. The
Science+Business Media ocular half-life is usually calculated directly using
T.U. Krohne, M.D., F.E.B.O. (*) • F.G. Holz, M.D. the values measured from aqueous humor or vit-
Department of Ophthalmology, University of Bonn, reous samples, whereby comparative measure-
Ernst Abbe-Str. 2, Bonn 53127, Germany ments from animal experiments have shown that
e-mail: krohne@uni-bonn.de;
both types of sample provide virtually identical
frank.holz@ukb.uni-bonn.de
values (Bakri et al. 2007a; Gaudreault et al. 2005,
C.H. Meyer, M.D.
2007). Occasionally, however, the ocular half-life
Department of Ophthalmology, Pallas Clinics,
Bahnhofplatz 4, Aarau, Aargau, Switzerland is also estimated using measurements from
e-mail: carsten.meyer@klinik-pallas.ch plasma samples (Basile et al. 2012; Xu et al.

140 T.U. Krohne et al.
Table 8.1 Ocular half-lives after intravitreal injection in rabbits (EMA, European Medicines Agency)
Molecular Ocular half-life in
Molecular class Substance (trade name) weight (kDa) rabbits (days)] Reference
RNA aptamer Pegaptanib (Macugen) 50 3.46 Eyetech Study Group (2002)
Fab fragment Ranibizumab (Lucentis) 48 2.88 Bakri et al. (2007a)
2.9 Gaudreault et al. (2007)
IgG antibody Bevacizumab (Avastin) 149 4.32 Bakri et al. (2007b)
5.95 Nomoto et al. (2009)
Rituximab (MabThera)a 144 4.7 Kim et al. (2006)
Trastuzumab (Herceptin)b 146 5.6 Mordenti et al. (1999)
Fc fragment Aflibercept (Eylea) 115 4.79 Bayer/EMA (European
fusion protein Medicines 2013)
Conberceptc 143 4.24 Li et al. (2012)
a
Anti-CD20 antibody
b
Anti-HER2 antibody
c
Fusion protein of VEGF receptor binding domains and Fc fragment, similar to aflibercept
Table 8.2 Ocular half-lives after intravitreal injection in For better comparability of the results, we
humans measured the ocular half-lives of bevacizumab
Ocular and ranibizumab in humans using identical meth-
half-life in Eyes ods. We determined a value of 9.82 days for the
Substance humans examined IgG antibody bevacizumab (149 kDa) and a value
(trade name) (days) (number) Reference
of 7.19 days for the Fab fragment ranibizumab (48
Pegaptanib 8a –a Basile et al.
(Macugen) (2012) kDa), corresponding to a 1.4 times longer half-life
Ranibizumab 7.19 18 Krohne of bevacizumab compared to ranibizumab
(Lucentis) et al. (2012) (Krohne et al. 2008, 2012). This is in line with the
9a –a Xu et al. results of a clinical trial in which a 1.4 times lon-
(2013) ger period of action of bevacizumab compared
Bevacizumab 9.82 30 Krohne with ranibizumab was measured in patients with
(Avastin) et al. (2008)
age-related macular degeneration (AMD) based
10 18 Csaky et al.
(2007) on the retinal thickness analysis by optical coher-
6.7 11 Zhu et al. ence tomography (OCT) (Shah and Del Priore
(2008) 2009). The value furthermore agrees with animal
Aflibercept No published data studies which describe a 1.5 times longer half-life
(Eylea) of bevacizumab compared to ranibizumab (Bakri
a
Values estimated using plasma concentration measurements et al. 2007a, b). Since the molecular weight of a
substance and the associated diffusion speed
through the vitreous usually constitutes a main
2013) provided that the examined substance is not factor of the ocular elimination speed (Durairaj
intraocularly broken down but is eliminated from et al. 2009), it is notable that bevacizumab and
the eye by release into the systemic circulation, other IgG antibodies have around three times the
and that the systemic half-life can be assumed to molecular weight compared to ranibizumab and
be negligible compared to the ocular half-life. the RNA aptamer pegaptanib (50 kDa), but only a
Under these conditions, the time of elimination slightly higher ocular half-life by comparison. A
from the bloodstream essentially corresponds to possible explanation for this discrepancy is the
that of ocular elimination, as is the case for ranibi- activity of the neonatal Fc receptor, as will be
zumab and pegaptanib for example. explained in the following section.
8 Pharmacokinetics of Intravitreally Applied VEGF Inhibitors 141
The VEGF inhibitor aflibercept (115 kDa), IgG antibodies are bound to the FcRn via
which is a fusion protein from the binding their Fc fragment. Therefore, only complete anti-
domains of VEGF receptors 1 and 2 and an Fc bodies such as bevacizumab are bound by the
antibody fragment, as well as the similarly struc- receptor while Fab fragments such as ranibi-
tured VEGF inhibitor conbercept (Li et al. 2012) zumab are not bound in the absence of the Fc
both show ocular half-lives in rabbits which are fragment. An active FcRn-mediated outward
within the range of bevacizumab and other IgG transportation of bevacizumab but not of ranibi-
antibodies. This could indicate that the ocular zumab from the eye could therefore explain why
half-life of aflibercept is similar to the known bevacizumab has only a slightly higher ocular
half-life of bevacizumab also in humans. Human half-life than ranibizumab despite a molecular
data for aflibercept, however, are not available weight that is three times higher. The recycling
to date. function of the FcRn with respect to complete
antibodies may also lead to the plasma half-life
of bevacizumab of 20 days being distinctly
8.2 Pharmacokinetic extended compared to 2 h for ranibizumab (Lu
Significance of the Neonatal et al. 2008; Xu et al. 2013). As an aptamer,
Fc Receptor pegaptanib is not bound by the FcRn, while
aflibercept as a fusion protein with Fc fragment
The neonatal Fc receptor (FcRn) can bind IgG is assumed to bind to the FcRn in a similar way
antibodies in order to transport them unidirec- as IgG antibodies.
tionally through various body barriers. This is
important, for example, for the transportation of
maternal antibodies via the placental barrier to 8.3 Clinical Relevance of Ocular
the bloodstream of the unborn child. In the Pharmacokinetics
blood–brain barrier, the FcRn transports antibod-
ies from cerebral tissues into the bloodstream and The ocular elimination rate is an important factor
therefore possibly plays a role in the containment for the effect duration of intravitreal active phar-
of intracerebral inflammatory reactions. An maceutical compounds and therefore co-
expression of FcRn has also been detected in the determines reinjection intervals and frequency. It
inner and outer blood–retina barrier (van Bilsen was originally assumed that substances such as
et al. 2011; Kim et al. 2008) so that here too an bevacizumab with a higher molecular weight
FcRn-mediated outward transportation of anti- compared to ranibizumab would demonstrate a
bodies from the retina to the bloodstream is longer effect and therefore a lower re-treatment
assumed (Fig. 8.1). frequency. However, as already explained, beva-
In addition to its transport function, the FcRn is cizumab and ranibizumab have only compara-
also expressed in numerous other tissues where it tively little diverging ocular half-lives so that
prevents the degradation or elimination of IgG little differences in terms of their reinjection fre-
antibodies. For example, the antibodies absorbed quency are to be expected. The injection num-
by vascular endothelial cells are transported back bers per year for different VEGF inhibitors
into the blood via the FcRn so that they are not determined in clinical studies for pro re nata
broken down intracellularly by lysosomes like treatment regimens are summarized in Table 8.3.
other serum proteins. This recycling function of In agreement with the pharmacokinetic data,
the FcRn means that antibodies have a much essentially equal injection frequencies for beva-
higher half-life in the systemic circulation than cizumab and ranibizumab were found. For
other comparable serum proteins and can therefore aflibercept, data is so far only available from the
satisfy their immunological function for longer. marketing authorization studies which currently
Fig. 8.1 Possible role of the neonatal Fc receptor (FcRn) ates elimination from the eye by outward transportation
in the transportation of bevacizumab (Avastin) and ranibi- via the blood–retina barrier (top) and increases the dura-
zumab (Lucentis). Bevacizumab, but not ranibizumab, tion of retention in the systemic circulation through pro-
binds to the FcRn via the Fc fragment. The FcRn acceler- tection from lysosomal breakdown (bottom)
Table 8.3 Average number of injections per year with a treatment regimen including an initial fixed upload phase and
subsequent monthly treatments as needed (pro re nata, PRN) in neovascular age-related macular degeneration
Substance (trade name) Study Injections in first year Injections in second year
Pegaptanib (Macugen) No published data
Bevacizumab (Avastin) CATT 7.7 6.4
Ranibizumab (Lucentis) CATT 6.9 5.7
SUSTAIN 5.7 No second year
HARBOR 7.7 5.6
PrONTO 5.6 4.2
VIEW 1/2 No PRN arm 5.6a
Aflibercept (Eylea) VIEW 1/2 No PRN arm 4.8b
CLEAR-IT 2 5.6c No second year
a
Corresponds to 4.7 injections in 44 study weeks with capped PRN treatment regimen
b
Corresponds to 4.1 injections in 44 study weeks in 2q4 study arm with “capped PRN” treatment regimen
c
In “2q4” study arm including initial fixed upload phase of 4 monthly injections
do not allow any conclusion as to whether the following intravitreal administration can be
number of injections differs significantly from modelled using the Bateman function (Fig. 8.2).
other VEGF inhibitors in a pro re nata treatment The model indicates a distinctly higher systemic
regimen. exposure to bevacizumab compared to ranibi-
zumab so that a stronger effect of bevacizumab
on the systemic VEGF activity would be
8.4 Systemic Pharmacokinetics expected. Indeed, clinical studies on patients
following Intravitreal with AMD, diabetic macular edema, and retinop-
Injection athy of prematurity indicate that intravitreally
injected bevacizumab can lead to suppression of
VEGF inhibitors are eliminated from the eye not systemic VEGF activity over a period of several
through intraocular breakdown but by release weeks after injection (Carneiro et al. 2012;
into the systemic circulation so that following Matsuyama et al. 2010; Sato et al. 2011; Zehetner
intravitreal administration a suppression of sys- et al. 2013). In agreement with the pharmacoki-
temic VEGF activity with potentially systemic netic data, two comparative investigations into
adverse events is possible. The systemic expo- bevacizumab and other VEGF inhibitors could
sure depends on the systemic half-lives of the not detect such a systemic effect for ranibizumab
substances. Data is available on the systemic and pegaptanib (Carneiro et al. 2012; Zehetner
half-lives particularly for bevacizumab and et al. 2013). Therefore, the pharmacokinetic
aflibercept used systemically for cancer treat- properties of ranibizumab would appear more
ment, and also for other VEGF inhibitors. An advantageous compared to bevacizumab for areas
overview of the values in humans is summarized of application which are particularly sensitive in
in Table 8.4. A distinct difference is shown here terms of a systemic VEGF suppression, such as
between substances with and without Fc frag- the treatment of retinopathy of prematurity
ment which is most likely attributable to the (Krohne et al. 2012). An effect on the untreated
above described different binding to the neonatal fellow eye was also determined in a study on dia-
Fc receptor and the associated protection from a betic macular edema only after the injection of
rapid systemic breakdown of the substances. bevacizumab, but not of ranibizumab (Bakbak
Using the ocular and systemic half-lives et al. 2013). However, other studies were unable
measured, the systemic level of active substances to detect either a measurable substance level of
bevacizumab or a reduction in the VEGF activity anti-VEGF treatment including VISION,

in the untreated fellow eye (Meyer et al. 2012; ANCHOR/MARINA, CATT, IVAN, and VIEW
Sawada et al. 2008). So far, there is no published 1/2, the rates of systemic serious adverse events
data on the effect of aflibercept on the systemic (SAE) were low in all study groups. Moreover,
VEGF level or contralateral eye exposure. SAE rates in studies such as VISION and
In cancer treatment, VEGF inhibitors such as MARINA were not significantly different in
bevacizumab are used in distinctly higher dos- treatment groups compared with the untreated
ages and can lead to known adverse events such control groups. Even the fixed monthly adminis-
as bleeding, neutropenia, and intestinal perfora- tration of ranibizumab with four times the usual
tions. By comparison, the serum levels achieved dose over a period of 12 months as part of the
after intravitreal treatment are about 100 times HABOR study did not lead to a significant rise in
lower. In all larger ophthalmological studies on the SAE rate (Busbee et al. 2013). However, with
the low SAE incidence, patient numbers in all
studies have so far been insufficient for a statisti-
Table 8.4 Systemic half-lives in humans (EMA,
European Medicines Agency)
cally relevant statement to be made on systemic
safety. A current meta-analysis of several clinical
Systemic
Substance (trade half-life in
studies and an analysis of the data from the CATT
name) humans (days) Reference and IVAN studies both come to the conclusion
Pegaptanib No published human data; 0.4a in that there are indications of an increased general
(Macugen) monkey (Tucker et al. 1999) SAE rate under bevacizumab compared with
Ranibizumab 0.08b Xu et al. (2013) ranibizumab, without it being possible, however,
(Lucentis) to determine higher incidences in a specific organ
Bevacizumab 20 Lu et al. (2008)
system (Chakravarthy et al. 2013; Schmucker
(Avastin)
Aflibercept (Eylea) 2–6 Bayer/EMA
et al. 2012). These results on the SAE rate corre-
(2013) late with the stated pharmacokinetic differences
a
Corresponds to 9.3 h of the VEGF inhibitors with respect to their sys-
b
Corresponds to 2 h temic exposure.
Fig. 8.2 Pharmacokinetic model calculation of the intra- tions of 1.5 mg bevacizumab (Avastin) or 0.5 mg ranibi-
ocular and systemic substance levels after a single (left) zumab (Lucentis) using the ocular and systemic half-lives
and three consecutive monthly (right) intravitreal injec- measured in humans
8.5 Refraction and Ocular altered pharmacokinetics of intravitreal substances

Pharmacokinetics are to be assumed. Experimental studies on animals
have actually detected a significantly accelerated
Refraction differences may be accompanied by elimination of intravitreally injected substances
significant axial length differences and therefore a from the vitreous after enzymatic liquefaction of the
different ocular distribution volume for intravit- vitreous (Tan et al. 2011; Wu et al. 2011). However,
really injected medicines. If by way of simplifica- to what extent the duration of effect of VEGF inhib-
tion the eye is viewed as a sphere, then itors in humans decreases with increasing age and
arithmetically one eye with an axial length of 27 degree of vitreous liquefaction and whether this
mm, for example, will have around double the vol- influences the therapeutic effect in a clinically rele-
ume of an eye with an axial length of 22 mm. The vant manner has not so far been examined.
standard dosage of an intravitreal medicine in Following a vitrectomy, the vitreous is replaced
such an eye would therefore achieve only half the by the distinctly less viscous aqueous humor. Data
substance concentration, and the duration of from animal experiments provide evidence of an
action of the medicine would arithmetically be accelerated elimination from the vitrectomized
reduced by 1 half-life (in the case of bevacizumab, eye for different intravitreal substances (Chin
for example, around 10 days). This effect may et al. 2005; Doft et al. 1985; Jarus et al. 1985). The
possibly be enhanced by a faster elimination of the molecular weight of many of the active substances
active substance from larger eyes as a result of the investigated in these studies such as triamcinolone
larger retinal surface. Even if in reality the myopic (0.4 kDa), amphotericin B (0.9 kDa), and 5-fluo-
eye in particular can deviate substantially from the rouracil (0.1 kDa) is, however, considerably lower
spherical shape so that the true differences in vol- than that of VEGF inhibitors such as bevacizumab
umes will probably be smaller, this arithmetical (149 kDa) so that the pharmacokinetic results may
example does show the pronounced effects that possibly be transferred to the VEGF inhibitors only
axial length differences may possibly have on the to a limited extent. Due to the anatomical features
pharmacokinetics of intravitreal medicine. of the animal models used here, the vitrectomy
Although the arithmetical effect of ocular volume was also combined with a lensectomy in many
on intraocular drug concentration and duration of studies which results in the absence of a barrier to
action could not yet be confirmed in experimental the anterior chamber with possibly easier diffu-
studies (Krohne et al. 2015), individual authors sion of the substances from the vitreous, therefore
recommend to adjust the dosage of intravitreal making comparability with the situation in the vit-
medicines in eyes with axial lengths which differ rectomized but phakic/pseudophakic human eye
extremely from the norm (Teichmann 2002). difficult.
Another possibility would be to reduce the control So far, only very restricted data on the phar-
and reinjection intervals of anti-VEGF treatment macokinetics in the vitrectomized eye exist for
for highly myopic eyes, particularly if an inade- VEGF inhibitors. A retrospective case series on
quate therapeutic effect is observed. However, no 11 vitrectomized eyes which were treated with
studies have so far investigated the effect of axial bevacizumab for diabetic macular edema reports
length on duration and extent of the clinical effect. on the absence of a clinical effect on retinal thick-
ness and vision, possibly due to an accelerated
ocular elimination of the substance (Yanyali et al.
8.6 Effects of Vitreous 2007). Contrary to this, another retrospective
Liquefaction, Vitrectomy, study on diabetic macular edema describes an
and Silicone Oil effect of bevacizumab in vitrectomized eyes
Endotamponade which, however, is smaller than in non-
vitrectomized eyes (Mehta et al. 2010). In mon-
The liquefaction of the vitreous which increases keys, a 54 % lower ocular half-life of bevacizumab
with age causes the viscosity of the vitreous to was measured after combined lensectomy and
decrease so that a faster diffusion and therefore vitrectomy (Kakinoki et al. 2012). In rabbits, a 46
% reduction was demonstrated for bevacizumab clinical testing, but still as intravitreal injection
and a 24 % decrease for ranibizumab following for the time being. Sustained release drug carriers
vitrectomy alone, while the reduction after are already being used with steroids (e.g.,
lensectomy alone was even more pronounced Ozurdex, Iluvien, Retisert) and antiviral drugs
(Christoforidis et al. 2013). By contrast, a differ- (e.g., Vitrasert) for ophthalmic treatment. They
ent study on rabbits came to the conclusion that are able to evenly release the active substance
vitrectomy alone has no influence on ocular half- contained in them over months or years after
life (Ahn et al. 2013). The data on the effect of intravitreal implantation. Therefore with these
VEGF inhibitors in vitrectomized eyes is there- devices, it is less the pharmacokinetic properties
fore not uniform so far. Even after vitrectomy, an of the active substances but more the release
intravitreal anti-VEGF treatment may be effec- characteristics of the carrier that determines
tive. However, in view of the possibly faster elim- ocular substance levels. The use of sustained
ination of the active pharmaceutical ingredients release drug technology with proteins such as the
from the eye, it may be expedient to shorten the currently available VEGF inhibitors is proving to
standard check-up and reinjection intervals. be difficult, however, due to the lower stability of
The injection into silicone oil-filled eyes is a these substances compared to, for example, ste-
rather rare application of VEGF inhibitors. roids. Sustained release drug carriers charged
Experimental animal studies, however, exist and with VEGF inhibitors are therefore currently
show that the transition from the vitreous to the only being investigated at a preclinical level.
retinal tissue and into the anterior chamber in sili-
cone oil-filled eyes is delayed, but that the subse-
quent elimination rate in retinal tissue and 8.8 Conclusions for Clinicians
anterior chamber does not differ from that of
non-vitrectomized eyes (Xu et al. 2012). This • The ocular and systemic pharmacokinetics of
indicates that a treatment with intravitreal VEGF intravitreally injected substances determine
inhibitors could be effective also in silicone oil- the ocular effect duration and the systemic
filled eyes. Clinical data on the treatment of reti- exposure and are therefore of clinical rele-
nal diseases are not available, however. Only for vance with respect to reinjection frequency
diabetic rubeosis iridis is a decline of the disease and systemic safety.
following intravitreal injection of bevacizumab • Despite distinct differences in molecular
in silicone oil-filled eyes reported in a case series weight, the currently available VEGF inhibi-
of five patients (Falavarjani et al. 2010). tors show only comparatively small differ-
ences in terms of ocular half-lives (Tables 8.1
and 8.2), possibly because the slower diffu-
8.7 Outlook on Future sion of larger, Fc fragment-containing sub-
Developments stances from the eye is compensated by an
active outward transportation via the neonatal
In order to reduce the necessity of intravitreal Fc receptor.
injections and therefore the burden on patients as • VEGF inhibitors containing an Fc fragment
well as the endophthalmitis risk, new forms of can have a distinctly longer systemic half-life
administration such as topical VEGF inhibitors compared with Fc-free substances (Table 8.4)
and sustained release drug carriers are desirable. because binding to the neonatal Fc receptor on
Topical administration could be achieved by a extraocular cells such as vascular endothelial
lower molecular weight and improved corneal cells may slow its systemic breakdown.
penetrability. VEGF inhibitors from new sub- • The pharmacokinetics of intravitreal VEGF
stance classes such as single-chain antibody frag- inhibitors can also be influenced by properties
ments (ESBA1008, MW 26 kDa) and DARPins of the individual patient or eye, such as refrac-
(AGN-150998, MW 34 kDa) could possess prop- tion (ocular volume), age (vitreous liquefac-
erties of this nature and are already undergoing tion), and previous surgeries (vitrectomy).
Chakravarthy U, Harding SP, Rogers CA, et al. Alternative

Compliance with Ethical Requirements treatments to inhibit VEGF in age-related choroidal
neovascularisation: 2-year findings of the IVAN ran-
T.U. Krohne—support of research projects domised controlled trial. Lancet. 2013;382(9900):
and clinical studies: Alcon, Novartis; con- 1258–67.
sultancy honoraria, lecture fees, travel Chin HS, Park TS, Moon YS, et al. Difference in clear-
expenses: Bayer, Heidelberg Engineering, ance of intravitreal triamcinolone acetonide between
vitrectomized and nonvitrectomized eyes. Retina.
Novartis. F.G. Holz—support of research 2005;25:556–60.
projects and clinical studies: Acucela, Christoforidis JB, Williams MM, Wang J, et al. Anatomic
Alcon, Allergan, Bayer, Carl Zeiss Meditec, and pharmacokinetic properties of intravitreal bevaci-
Genentech, Heidelberg Engineering, zumab and ranibizumab after vitrectomy and lensec-
tomy. Retina. 2013;33:946–52.
Novartis, Optos; consultancy honoraria, Csaky KG, Gordiyenko N, Rabena MG et al.
lecture fees, travel expenses: Acucela, Pharmacokinetics of intravitreal bevacizumab in
Alcon, Allergan, Bayer, Genentech, humans. Invest Ophthalmol Vis Sci 2007;48:
Heidelberg Engineering, Novartis, Roche. E-Abstract 4936.
Doft BH, Weiskopf J, Nilsson-Ehle I, et al. Amphotericin
C.H. Meyer—support of research projects clearance in vitrectomized versus nonvitrectomized
and clinical studies: Novartis, Allergan; eyes. Ophthalmology. 1985;92:1601–5.
consultancy honoraria, lecture fees, travel Durairaj C, Shah JC, Senapati S, et al. Prediction of vitreal
expenses: Bayer, GSK, Novartis. half-life based on drug physicochemical properties:
quantitative structure-pharmacokinetic relationships
No human or animal studies were car- (QSPKR). Pharm Res. 2009;26:1236–60.
ried out by the authors for this chapter. European Medicines Agency. Eylea: EPAR—European
public assessment report (September 20, 2012). http://
www.ema.europa.eu/docs/en_GB/document_library/
EPAR_-_Public_assessment_report/human/002392/
WC500135744.pdf. 2013. Accessed: January 23,
References 2015.
Eyetech Study Group. Preclinical and phase 1A clinical
Ahn J, Kim H, Woo SJ, et al. Pharmacokinetics of intravit- evaluation of an anti-VEGF pegylated aptamer
really injected bevacizumab in vitrectomized eyes. (EYE001) for the treatment of exudative age-related
J Ocul Pharmacol Ther. 2013;29(7):612–8. macular degeneration. Retina. 2002;22:143–52.
Bakbak B, Ozturk BT, Gonul S, et al. Comparison of the Falavarjani KG, Modarres M, Nazari H. Therapeutic
effect of unilateral intravitreal bevacizumab and effect of bevacizumab injected into the silicone oil in
ranibizumab injection on diabetic macular edema of eyes with neovascular glaucoma after vitrectomy for
the fellow eye. J Ocul Pharmacol Ther. 2013;29(8): advanced diabetic retinopathy. Eye (Lond). 2010;
728–32. 24:717–9.
Bakri SJ, Snyder MR, Reid JM, et al. Pharmacokinetics of Gaudreault J, Fei D, Rusit J, et al. Preclinical pharmacoki-
intravitreal ranibizumab (Lucentis). Ophthalmology. netics of Ranibizumab (rhuFabV2) after a single intra-
2007a;114:2179–82. vitreal administration. Invest Ophthalmol Vis Sci.
Bakri SJ, Snyder MR, Reid JM, et al. Pharmacokinetics of 2005;46:726–33.
intravitreal ranibizumab (Lucentis). Ophthalmology. Gaudreault J, Fei D, Beyer JC, et al. Pharmacokinetics
2007b;114:2179–82. and retinal distribution of ranibizumab, a humanized
Basile AS, Hutmacher M, Nickens D, et al. Population antibody fragment directed against VEGF-A, follow-
pharmacokinetics of pegaptanib in patients with neo- ing intravitreal administration in rabbits. Retina.
vascular, age-related macular degeneration. J Clin 2007;27:1260–6.
Pharmacol. 2012;52:1186–99. Jarus G, Blumenkranz M, Hernandez E, et al. Clearance
Busbee BG, Ho AC, Brown DM, et al. Twelve-month effi- of intravitreal fluorouracil. Normal and aphakic vitrec-
cacy and safety of 0.5 mg or 2.0 mg ranibizumab in tomized eyes. Ophthalmology. 1985;92:91–6.
patients with subfoveal neovascular age-related macu- Kakinoki M, Sawada O, Sawada T, et al. Effect of vitrec-
lar degeneration. Ophthalmology. 2013;120:1046–56. tomy on aqueous VEGF concentration and pharmaco-
Carneiro AM, Costa R, Falcao MS, et al. Vascular endo- kinetics of bevacizumab in macaque monkeys. Invest
thelial growth factor plasma levels before and after Ophthalmol Vis Sci. 2012;53:5877–80.
treatment of neovascular age-related macular degen- Kim H, Csaky KG, Chan CC, et al. The pharmacokinetics
eration with bevacizumab or ranibizumab. Acta of rituximab following an intravitreal injection. Exp
Ophthalmol. 2012;90:e25–30. Eye Res. 2006;82:760–6.
Kim H, Fariss RN, Zhang C, et al. Mapping of the neona- injected with intravitreal bevacizumab. Graefes Arch
tal Fc receptor in the rodent eye. Invest Ophthalmol Clin Exp Ophthalmol. 2008;246:1379–81.
Vis Sci. 2008;49:2025–9. Schmucker C, Ehlken C, Agostini HT, et al. A safety
Krohne TU, Eter N, Holz FG, et al. Intraocular pharmaco- review and meta-analyses of bevacizumab and ranibi-
kinetics of bevacizumab after a single intravitreal zumab: off-label versus gold standard. PLoS One.
injection in humans. Am J Ophthalmol. 2012;7, e42701.
2008;146:508–12. Shah AR, Del Priore LV. Duration of action of intravitreal
Krohne TU, Aisenbrey S, Holz FG. Current therapeutic ranibizumab and bevacizumab in exudative AMD eyes
options in retinopathy of prematurity. Ophthalmologe. based on macular volume measurements. Br
2012a;109:1189–97. J Ophthalmol. 2009;93:1027–32.
Krohne TU, Liu Z, Holz FG, et al. Intraocular pharmaco- Tan LE, Orilla W, Hughes PM, et al. Effects of vitreous
kinetics of ranibizumab following a single intravitreal liquefaction on the intravitreal distribution of sodium
injection in humans. Am J Ophthalmol. 2012b; fluorescein, fluorescein dextran, and fluorescent mic-
154(682-686), e682. roparticles. Invest Ophthalmol Vis Sci. 2011;52:
Krohne TU, Muether PS, Stratmann NK, et al. Influence 1111–8.
of ocular volume and lens status on pharmacokinetics Teichmann KD. Intravitreal injections: does globe size
and duration of action of intravitreal vascular endothe- matter? J Cataract Refract Surg. 2002;28:1886–9.
lial growth factor inhibitors. Retina. 2015;35(1): Tucker CE, Chen LS, Judkins MB, et al. Detection and
69–74. plasma pharmacokinetics of an anti-vascular endothe-
Li H, Lei N, Zhang M, et al. Pharmacokinetics of a long- lial growth factor oligonucleotide-aptamer (NX1838)
lasting anti-VEGF fusion protein in rabbit. Exp Eye in rhesus monkeys. J Chromatogr B Biomed Sci Appl.
Res. 2012;97:154–9. 1999;732:203–12.
Lu JF, Bruno R, Eppler S, et al. Clinical pharmacokinetics van Bilsen K, van Hagen PM, Bastiaans J, et al. The neo-
of bevacizumab in patients with solid tumors. Cancer natal Fc receptor is expressed by human retinal pig-
Chemother Pharmacol. 2008;62:779–86. ment epithelial cells and is downregulated by tumour
Matsuyama K, Ogata N, Matsuoka M, et al. Plasma levels necrosis factor-alpha. Br J Ophthalmol. 2011;95:
of vascular endothelial growth factor and pigment 864–8.
epithelium-derived factor before and after intravitreal Wu WC, Chen CC, Liu CH, et al. Plasmin treatment
injection of bevacizumab. Br J Ophthalmol. 2010;94: accelerates vascular endothelial growth factor clear-
1215–8. ance from rabbit eyes. Invest Ophthalmol Vis Sci.
Mehta S, Blinder KJ, Shah GK, et al. Intravitreal bevaci- 2011;52:6162–7.
zumab for the treatment of refractory diabetic macular Xu Y, You Y, Du W, et al. Ocular pharmacokinetics of
edema. Ophthalmic Surg Lasers Imaging. bevacizumab in vitrectomized eyes with silicone oil
2010;41:323–9. tamponade. Invest Ophthalmol Vis Sci. 2012;53:
Meyer CH, Krohne TU, Holz FG. Concentrations of 5221–6.
unbound bevacizumab in the aqueous of untreated fel- Xu L, Lu T, Tuomi L, et al. Pharmacokinetics of ranibi-
low eyes after a single intravitreal injection in humans. zumab in patients with neovascular age-related
Acta Ophthalmol. 2012;90:68–70. macular degeneration: a population approach. Invest
Mordenti J, Cuthbertson RA, Ferrara N, et al. Comparisons Ophthalmol Vis Sci. 2013;54:1616–24.
of the intraocular tissue distribution, pharmacokinet- Yanyali A, Aytug B, Horozoglu F, et al. Bevacizumab
ics, and safety of 125I-labeled full-length and Fab (Avastin) for diabetic macular edema in previously
antibodies in rhesus monkeys following intravitreal vitrectomized eyes. Am J Ophthalmol. 2007;144:
administration. Toxicol Pathol. 1999;27:536–44. 124–6.
Nomoto H, Shiraga F, Kuno N, et al. Pharmacokinetics of Zehetner C, Kirchmair R, Huber S, et al. Plasma levels of
bevacizumab after topical, subconjunctival, and intra- vascular endothelial growth factor before and after
vitreal administration in rabbits. Invest Ophthalmol intravitreal injection of bevacizumab, ranibizumab
Vis Sci. 2009;50:4807–13. and pegaptanib in patients with age-related macular
Sato T, Wada K, Arahori H, et al. Serum concentrations of degeneration, and in patients with diabetic macular
bevacizumab (Avastin) and vascular endothelial oedema. Br J Ophthalmol. 2013;97:454–9.
growth factor in infants with retinopathy of prematu- Zhu Q, Ziemssen F, Henke-Fahle S et al. Vitreous levels
rity. Am J Ophthalmol. 2011;153(2):327–33.e1. of bevacizumab and vascular endothelial growth fac-
Sawada O, Kawamura H, Kakinoki M, et al. Vascular tor-A in patients with choroidal neovascularization.
endothelial growth factor in fellow eyes of eyes Ophthalmology. 2008;115:1750–5, 1755e1751.
Neovascular Glaucoma
9
Julia Lüke, Matthias Lüke, and Salvatore Grisanti
Similarly, intravitreal injections of recombinant

9.1 Pathogenesis of Neovascular human VEGF in amounts comparable to those
Glaucoma measured in eyes with active neovascularization
induce noninflammatory iris neovascularization
Ischemic retinopathies and the concomitant in nonhuman primates. Prolonged exposure to
increased level of vascular endothelial growth VEGF can therefore produce neovascular glau-
factor (VEGF) are the leading causes of neovas- coma (Tolentino et al. 1996).
cularization of the iris surface. In progressive
stages, fibrovascular membranes occlude the
anterior chamber angle and inhibit aqueous out- 9.2 Conventional Therapeutic
flow resulting in elevated intraocular pressure Approach in Neovascular
(IOP). This is often difficult to control and fre- Glaucoma
quently results in loss of vision. Rubeosis iridis
was initially described by Coats in 1906. Since There are two key aspects to the management of
his original description, neovascularization of neovascular glaucoma: treatment of the underly-
the iris and anterior chamber angle has been ing angiogenic disease process and treatment of
described in many diseases, 97 % of which are the increased IOP. Previous treatment modalities
associated with retinal ischemia (Brown et al. include (a) panretinal photocoagulation (PRP) or
1984). The most common conditions are diabetes cryocoagulation to ablate the source of VEGF
mellitus, central retinal vein occlusion (CRVO), and (b) cyclodestructive or drainage procedures
and ocular ischemic syndrome. Diseases associ- to counteract the angular obstruction and to
ated with rubeosis and neovascular glaucoma reduce IOP.
(NVG) are listed in Table 9.1. The reason for this PRP is considered the standard treatment of
process is elevated intraocular VEGF levels. ischemic retinal diseases. Ohnishi and colleagues
documented regression of rubeosis in 68 % of
patients and normalization of IOP in 42 % of
J. Lüke, M.D., P.D. (*) • M. Lüke, M.D., P.D. patients treated with PRP. The efficacy of PRP
S. Grisanti, M.D. has been studied in specific disease processes,
Department of Ophthalmology, University of Lübeck, including diabetic retinopathy, CRVO, radiation
Ratzeburger Allee 160, Lübeck, Schleswig-Holstein
retinopathy following plaque treatment, and cen-
23538, Germany
e-mail: julia.lueke@uksh.de; tral retinal artery occlusion (Wand et al. 1978).
Matthias.lueke@uksh.de; Salvatore.grisanti@uksh.de PRP results in new vessel regression in all of

150 J. Lüke et al.
Table 9.1 Disease in which rubeosis iridis has been these conditions. Extensive cryotherapy of the
reported (Sivak-Callcott et al. 2001)
anterior retina has been shown to be effective in
Retinal ischemic diseases causing regression of rubeosis. Fernandez-Vigo
Diabetic retinopathy et al. did not find any difference in the regression
Central retinal vein occlusion of rubeosis with cryotherapy when compared
Ocular ischemic syndrome/carotid occlusive disease with PRP (Fernández-Vigo et al. 1997).
Central retinal artery occlusion
Medical management of neovascular glau-
Retinal detachment
coma is based on IOP-lowering agents, includ-
Leber’s congenital amaurosis
ing topical beta-adrenergic antagonists, alpha-2
Coat’s disease
agonists, and topical and oral carbonic anhy-
Eales disease
Sickle cell retinopathy
drase inhibitors. Partial destruction of the ciliary
Retinal hemangioma body lowers aqueous humor production and
Persistent hyperplastic primary vitreous therefore reduces IOP. Although diode laser
Norrie’s disease cyclodestruction seems to show promise for suc-
Wyburn Mason cessful IOP control in patients with neovascular
Carotid-cavernous fistula glaucoma, its long-term success rate has not
Dural shunt been well described. Moreover, standardization
Stickler’s syndrome of a formal treatment protocol is required.
X-linked retinoschisis Although cryotherapy of the ciliary body may be
Takayasu’s aortitis highly effective in lowering IOP, its rate of com-
Juxtafoveal telangiectasis plications including progression to phthisis
Surgically induced seems to be higher than that associated with
Carotid endarterectomy cyclophotocoagulation. In 1972, Feibl and
Cataract extraction
Bigger reported a 50 % reduction in IOP treated
Pars plana vitrectomy/lensectomy
with cyclocryotherapy (Feibel and Bigger 1972).
Silicone oil
Despite adequate IOP control, up to 70 % of
Scleral buckle
patients treated with cyclocryotherapy for refrac-
Neodymium-doped yttrium aluminium garnet
capsulotomy tory neovascular glaucoma have been reported
Laser coreoplasty to lose vision (Caprioli et al. 1985). Although
Tumors laser cyclophotocoagulation seems to have a
Iris: melanoma, hemangioma, metastatic lesion lower complication rate, the percentage of
Ciliary body: ring melanoma patients with neovascular glaucoma who lose
Retina: retinoblastoma, large cell lymphoma vision with this modality remains high, with
Choroid: melanoma long-term vision loss of 46.6 % reported by
Conjunctiva: squamous cell melanoma Shields and Shields (1994).
Radiation Filtering surgery for neovascular glaucoma is
External beam initially effective but often unsuccessful at the
Charged particle: proton, helium
end due to the wound-healing process. Many
Plaques
modifications have been tried to improve the out-
Photoradiation
come of these procedures (Brouillette and Chebil
Inflammatory diseases
1987; Kahook et al. 2006b; L’Esperance et al.
Uveitis: chronic iridocyclitis, Behcet’s disease
Vogt–Koyanagi–Harada syndrome
1983; Tripathi et al. 1998). A different approach
Syphilitic retinitis of aqueous drainage has been described by
Sympathetic ophthalmia Kirchhof (1994). He performed pars plana vitrec-
Endophthalmitis tomy, lensectomy, and retinectomy to re-route
Miscellaneous aqueous drainage through the choroidal circula-
Vitreous wick syndrome tion in nine eyes with neovascular glaucoma. An
Interferon alpha average decrease of 20–50 mmHg in IOP was
9 Neovascular Glaucoma 151
noted. Phthisis occurred in two eyes, hypotony in injection a reduction of leakage was observed
three eyes, and a retinal detachment in one eye. that persisted throughout the follow-up time of 1
This treatment, however, was often associated month (Fig. 9.1) (Grisanti et al. 2006). Three
with cyclophotocoagulation ab interno. The IOP- years later, a report described a decrease in aque-
lowering effect could therefore not be clearly ous VEGF concentration in NVG patients who
associated with the retinectomy alone. received intracameral bevacizumab (Grover et al.
Aqueous tube shunts have reported success 2009). Since then, multiple case series reported
rates of 22 % up to 97 % for patients with neovas- on the regression of neovascularization of both
cular glaucoma (Assaad et al. 1999). However, the the iris and the angle with the use of VEGF inhib-
authors cautioned that the visual prognosis might itors. In one case report, a patient with neovascu-
still be poor because of the severe underlying dis- lar glaucoma who failed IOP control with
ease process and postoperative complications. transscleral cyclophotocoagulation and PRP,
showed an immediate decrease in IOP after injec-
tion of bevacizumab (1 mg in 0.04 mL) (Kahook
9.3 Anti-VEGF Drugs et al. 2006a). Six consecutive patients with NVG
in Neovascular Glaucoma described by Iliev et al. showed a marked regres-
sion of anterior segment neovascularization. Half
In 2006, the first short-term case reports described of the patients showed a significant reduction of
the effects of intravitreal bevacizumab on the IOP after bevacizumab, and the remaining
regression of anterior segment neovasculariza- patients were controlled after additional cyclo-
tion (Avery 2006; Davidorf et al. 2006). At the photocoagulation and PRP (Iliev et al. 2006).
same time, we presented iris fluorescein angiog- VEGF-inhibition and regression of neovascular-
raphy in a case series of six eyes of three patients izations may therefore have an additional IOP-
receiving a bevacizumab injection (1.0 mg) into lowering effect that can develop rapidly within a
the anterior chamber. As early as 1 day after the few days.
Fig. 9.1 At baseline, one day before intracameral Avastin iris fluorescein angiography (c: 30 s, d: 5 min). The stabile
injection (1.0 mg bevacizumab, Genentech, USA) iris regression of rubeosis was observed 30 days (e: 30 s, f: 5
fluorescein angiography revealed an advanced rubeosis min) as well as 170 days (g: 30 s, h: 5 min) after initial
(a: 30 s, b: 5 min). As early as 5 days after bevacizumab bevacizumab injection
injection a massive reduction of rubeosis was evident in
152 J. Lüke et al.
9.4 Pharmacokinetics of Anti- in the vitreous or in the anterior chamber. Osamu

VEGF Drugs in NVG et al. found that the VEGF concentration in the
aqueous humor averaged 326 ± 125 pg/mL before
The degree and duration of the antiproliferative intravitreal injection of bevacizumab and
effect induced by anti-VEGF drugs is still a mat- decreased to less than 31 pg/mL in all eyes 1
ter of debate. Both depend on the area of isch- week after injection (Sawada et al. 2007). Lim
emic retina and the associated intravitreal VEGF et al. reported on the VEGF levels after intracam-
level. Tripathi et al. demonstrated significantly eral bevacizumab injection. In this case series of
increased levels of VEGF in the aqueous humor 5 NVG patients, the intracameral VEGF levels
of patients with NVG that were 40- and 113-fold were remarkably lowered from 1181.8 ± 1248.3
higher than in patients with primary open angle pg/mL to 33.2 pg/mL, which was much higher
glaucoma or cataract, respectively (Tripathi et al. than that seen after intravitreal injection (Lim
1998). Comparable findings were reported for et al. 2009). The application of anti-VEFG in the
ischemic retinal disease such as diabetic retinop- anterior chamber may approach the anterior seg-
athy and central retinal vein occlusion (Shinoda ment pathology more directly, but the injection of
et al. 2000). A direct blockade of VEGF by, e.g., anti-VEGF in the vitreous seems to achieve a
bevacizumab interrupts this signaling cascade. comparable clinical regression of NVI and may
The clinical effect can be observed as early as 1 have even a longer lasting effect in view of the
day after injection (Grisanti et al. 2006). depot effect of the vitreous.
Therefore, the application of anti-VEGF drugs in
patients with NVG allows a rapid control of the
neovascularizations. In a rabbit model, the vitre- 9.5 Treatment Regime in NVG
ous half-life of bevacizumab (1.25 mg) was Patients
shown to be 4.32 days. Small amounts of bevaci-
zumab were also found in the serum and in the Treatment of neovascular glaucoma needs a com-
fellow uninjected eye (Bakri et al. 2007). prehensive approach. In view of the heterogene-
Gheith et al. analyzed the length of the thera- ity of the underlying diseases, the varying degree
peutic benefit of bevacizumab in a case series of of ischemia and differences in anterior chamber
6 patients with an average follow-up of 9.7 angle involvement, the selection of therapeutic
months. They applied 1.25 mg/0.05 mL of beva- instruments must be carefully and individually
cizumab intravitreally followed by PRP 1 week assessed for every patient. First of all, the involve-
later which led to a complete regression of iris ment of the anterior chamber angle has to be con-
and angle neovascularizations in all patients. sidered for treatment planning. The study group
After 3 and 5 months, respectively, 2 patients had of Whakayabashi et al. analyzed NVG patients
recurrence of rubeosis but displayed regression divided into different subgroups depending on
after another injection of bevacizumab (Gheith the degree of involvement of the anterior cham-
et al. 2007). Based on the pathomechanisms, a ber angle. In patients with NVG and an open
recurrence of iris neovascularizations must be angle, intravitreal bevacizumab injection (IVB)
expected if the underlying ischemic disease is not seems to be effective in stabilizing the IOP. In
sufficiently treated. The date of recurrence varied these cases, the aqueous outflow is not inhibited
depending on the amount of untreated retina. In by neovascular membranes and one injection of
our own series of 20 eyes (of 18 patients) with IVB reduced the elevated IOP into the normal
neovascular glaucoma. 14 (70 %) received addi- range in approximately 70 % of the study eyes.
tional to a mean of 2.75 bevacizumab injections Rubeosis, however, recurred in 44 % by 6 months
(1.25/0.05 mL) an ablation of the ischemic retina of follow-up. In contrast, the recurrence rate in
(laser photoablation: n = 13, cryoablation: n = 6) the NVG group with a closed anterior chamber
during a follow-up of 1 year (Beutel et al. 2010). angle was even higher (71 %), indicating that the
Clinical studies addressed the question if the burden of VEGF in this advanced stage is more
injection of anti-VEGF drugs should be applied prominent. Nevertheless, the IOP decreased rap-
idly to a near-normal level by IVB monotherapy early surgical intervention. In total, 14 eyes (93
or IVB with topical antiglaucoma medication in %) underwent surgery to stabilize the markedly
9 eyes (53 %). Therefore, IVB seems to be effec- elevated IOP by the first 2 months of follow-up.
tive at least in selected cases also in this group. In these advanced stages of NVG, the role of
Nevertheless, a high proportion of patients (41 anti-VEGF drugs is limited to halting the neovas-
%) needed surgical intervention. This indicates cular activities (Wakabayashi et al. 2008).
that established fibrous membranes in the ante- However, this adjuvant allows reducing the risk
rior chamber angle that are not sensitive to of bleeding and inflammation which are well-
VEGF-inhibition were present. The increased known risks of trabeculectomy or shunt-tube
IOP was controlled successfully by filtering sur- drainage procedures for NVG independent of the
gery after repeated IVB without perioperative degree of anterior chamber angle involvement
bleedings. In the last group with synechiae and (Elgin et al. 2006; Parrish and Herschler 1983;
closure of the anterior chamber angle, the effi- Tsai et al. 1995).
cacy of bevacizumab for achieving rapid and Therefore, a situation with early neovascular-
marked regression of iris and angle neovascularization and an open anterior chamber angle
ization in patients was similar to that in the other requires a less aggressive treatment to lower the
groups. Nevertheless, IVB was not able to reduce IOP than a situation with an advanced involve-
the elevated IOP in the majority of patients in this ment and closure of the anterior chamber angle.
group with closed anterior chamber angle and Regarding the combination with other treat-
progressive synechiae. Within 1 week after the ment modalities different therapeutic options
initial IVB injection, 11 eyes (73 %) required are available (see Fig. 9.2). Treatment of the
Fig. 9.2 A 84-year-female patient with a rubeotic sec- 60 s, b; 3 min, c) revealed a stage 4 rubeosis. After 3 days
ondary glaucoma based on a central retinal vein occlusion of intravitreal injection of ranibizumab, a rapid regression
presented initially with an intraocular pressure of 48 of rubeosis occurred (early phase, d; 60 s, e; 3 min, f) as
mmHg. The iris fluorescein angiography (early phase, a; well as a reduction of intraocular pressure (11 mmHg)
154 J. Lüke et al.
Treatment regime
Ischemic ± Rubeosis ± IOP

retinopathy
Local anti-
Anti-VEGF
glaucomatous
therapy
Media clear Media opaque

monthly IOP
control
No Cyclo-
Panretinal destructive
photocoagulation Intravitreal procedure
Cataract Persisting
Bleeding
Iris NV?
IOP
Retinal
ppV
cryocoagulation,
Yes
Cataract surgery, Filtrating
in the further surgery
course
Fig. 9.3 Flow-chart demonstrating the different interven- injected in every case with active rubeosis according to
tions chosen depending on the stage of ischemic retinopa- this treatment regime (Lüke et al. 2013)
thy and the intraocular pressure. Anti-VEGF should be
underlying ischemic disease remains fundamen- regime which is based on the IOP level, the
tal in all cases. PRP is the treatment of choice in degree of untreated retinal ischemia, the status
case of clear optical media. Ehlers et al. com- of the optical media, and the involvement of the
pared the effect of PRP monotherapy with a anterior chamber angle (see Fig. 9.3).
combined approach of PRP and IVB. In this ret- So far ranibizumab (Lüke et al. 2013, Fig. 9.2)
rospective comparative analysis, the patients and bevacizumab (Fig. 9.1) have been tested suc-
that received adjuvant bevacizumab injection cessfully for this indication. A study comparing
had neovascular regression, while only 17 % of the effect of different VEGF inhibitors has not
patients who underwent PRP monotherapy had been performed yet. In view of the off-label char-
regression. The high response rate in the combi- acter of all anti-VEGF drugs for NVG, the choice
nation treatment group supports the critical role will depend on underlying disease and
that VEGF plays in anterior segment neovascu- comorbidities.
larization (Ehlers et al. 2008). In a prospective In terms of safety, concerns may arise regard-
series of 10 patients with NVG and 10 patients ing the VEGF-inhibition of all VEGF isoforms
with rubeosis of the iris, all patients presented possibly leading to side effects in normal retinal
with a rapid regression of rubeosis (see Fig. 9.2) tissues and circulation (Inan et al. 2007; Ishida
and a fast and a significant reduction of IOP in et al. 2003). In our case series of severe ischemic
the NVG group. All patients required treatment retinal diseases, no increase of retinal ischemia
of the underlying ischemic disease (Lüke et al. was observed during a follow-up of 12 month
2013). In one case, a vitrectomy was performed after repeated ranibizumab injection for rubeosis
combined with PRP in one case. In three and NVG. Additionally, no severe systemic side
patients, cataract surgery was necessary to allow effects were observed. Nevertheless, 2 out of 18
further laser coagulation of the retina. In the patients (11 %) had a Serious Adverse Event per
NVG group, the vast majority (90 %) received FDA definition that is in the range of previous
an IOP-lowering operative procedure. Based on ranibizumab studies especially of those with a
these findings, we recommend a treatment comparable patients’ selection (Massin et al.
2010). In view of the results of the CATT 2 study, intravitreal application may additionally prevent
which reported on higher rates of serious adverse intraoperative bleedings (Cornish et al. 2009;
events with bevacizumab, ranibizumab should be Li et al. 2009). The subconjunctival injection in
preferentially selected in those patients with car- contrast may have a better local antifibrotic effect
diovascular risk factors (Martin et al. 2012). (Grewal et al. 2008).
Additionally, it has to be considered that only
ranibizumab is an approved therapy for CRVO or
diabetic macular edema which can be coexistent 9.7 Summary
in rubeotic secondary glaucoma (Massin et al.
2010; Mitchell et al. 2011). Neovascular glaucoma may be the cause of
severe loss in visual function and has mainly a
bad prognosis as it is highly resistant to different
9.6 Adjuvant Treatment treatment modalities. In this severe disease, adju-
in Filtering Surgery vant application of anti-VEGF drugs allows a
for Neovascular Glaucoma rapid regression of rubeosis. Dependent on the
stage of the disease, a fast IOP reduction will
Angiogenesis is part of the wound-healing pro- help to improve prognosis. Anti-VEGF treatment
cess that may lead to scar formation after glau- is meaningful for short-term control, but it
coma filtration surgery. The risk for this unwanted achieves also a long-term effect by inhibiting the
side effect is augmented in neovascular glaucoma progression and establishment of angle obstruc-
patients. VEGF promotes the early phase of tion, which is a severe complication of rubeosis.
angiogenesis especially endothelial cell migra- In the treatment of NVG, conventional therapeu-
tion and proliferation. The strongest stimulus for tic procedures addressing the retinal ischemia
VEGF secretion is hypoxia. But also other para- and the source of the angiogenic stimulus should
crine acting factors such as TGF-ß1, EGF, TGF- always be performed, but the adjuvant applica-
α, bFGF, PDGF-BB, and IL-1b stimulate tion of anti-VEGF therapy is a necessary instru-
VEGF-expression (Barrientos et al. 2008). ment in a modern treatment approach in patients
The inhibition of VEGF to prevent angiogen- with rubeosis and neovascular glaucoma. The
esis during wound healing on the one hand and need for continued monitoring of NVI has to be
the reduction of fibroblast proliferation on the considered in this complex disease.
other hand by influencing the supply of FGF are
synergistic effects of VEGF blockade and inhibit
therefore both wound healing and angiogenesis
(Corral et al. 1999; Jonas et al. 2007; Seghezzi Compliance with Ethical Requirements
et al. 1998). Preclinical studies, which compared Compliance with Ethical Requirements:
anti-VEGF (bevacizumab) with 5-FU or BSS in a Julia Lüke, Matthias Lüke, and Salvatore
rabbit model revealed a significant delayed scar- Grisanti declare that they have no conflict
ring after bevacizumab injection (Memarzadeh of interest. All procedures followed were in
et al. 2009). Several case reports describe a posi- accordance with the ethical standards of
tive effect of anti-VEGF as adjuvant in glaucoma the responsible committee on human
filtration surgery in neovascular glaucoma experimentation (institutional and national)
patients (Katz et al. 1995; Kahook et al. 2006a; and with the Helsinki Declaration of 1975,
Kitnarong et al. 2008). A positive effect was as revised in 2000. Informed consent was
reported after intravitreal injection (Jonas et al. obtained from all patients for being
2007; Kitnarong et al. 2008) and after subcon- included in the study.
junctival injection during (Caprioli et al. 1985; The authors were supported by Novartis
Gheith et al. 2007) or after glaucoma filtration Pharma to perform an IIT.
surgery (Kahook et al. 2006a). The preoperative
156 J. Lüke et al.
References Grewal DS, Jain R, Kumar H, Grewal SP. Evaluation of

subconjunctival bevacizumab as an adjunct to trabecu-
lectomy a pilot study. Ophthalmology. 2008;
Assaad MH, Baerveldt G, Rockwood EJ. Glaucoma
115:2141–5.
drainage devices: pros and cons. Curr Opin
Grisanti S, Biester S, Peters S, Tatar O, Ziemssen F, Bartz-
Ophthalmol. 1999;10:147–53.
Schmidt KU, Tuebingen Bevacizumab Study Group.
Avery RL. Regression of retinal and iris neovasculariza-
Intracameral bevacizumab for iris rubeosis. Am
tion after intravitreal bevacizumab (Avastin) treat-
J Ophthalmol. 2006;142:158–60.
ment. Retina. 2006;26:352–4.
Grover S, Gupta S, Sharma R, Brar VS, Chalam
Bakri SJ, Snyder MR, Reid JM, Pulido JS, Singh
KV. Intracameral bevacizumab effectively reduces
RJ. Pharmacokinetics of intravitreal bevacizumab
aqueous vascular endothelial growth factor concentra-
(Avastin). Ophthalmology. 2007;114:855–9.
tions in neovascular glaucoma. Br J Ophthalmol.
Barrientos S, Stojadinovic O, Golinko MS, Brem H,
2009;93:273–4.
Tomic-Canic M. Growth factors and cytokines in
Iliev ME, Domig D, Wolf-Schnurrbursch U, Wolf S, Sarra
wound healing. Wound Repair Regen.
GM. Intravitreal bevacizumab (Avastin) in the treat-
2008;16:585–601.
ment of neovascular glaucoma. Am J Ophthalmol.
Beutel J, Peters S, Lüke M, Aisenbrey S, Szurman P,
2006;142:1054–6.
Spitzer MS, Yoeruek E, Bevacizumab Study Group,
Inan UU, Avci B, Kusbeci T, Kaderli B, Avci R, Temel
Grisanti S. Bevacizumab as adjuvant for neovascular
SG. Preclinical evaluation of intravitreal injection of
glaucoma. Acta Ophthalmol. 2010;88:103–9.
full-length humanized vascular endothelial growth
Brouillette G, Chebil A. Long-term results of modified
factor antibody in rabbit eyes. Invest Ophthalmol Vis
trabeculectomy with supramid implant for neovascu-
Sci. 2007;48:1773–81.
lar glaucoma. Can J Ophthalmol. 1987;22:254–6.
Ishida S, Usui T, Yamashiro K, Kaji Y, Amano S, Ogura
Brown GC, Magargal LE, Schachat A, Shah
Y, Hida T, Oguchi Y, Ambati J, Miller JW, Gragoudas
H. Neovascular glaucoma. Etiologic considerations.
ES, Ng YS, D’Amore PA, Shima DT, Adamis
Ophthalmology. 1984;91:315–20.
AP. VEGF164-mediated inflammation is required for
Caprioli J, Strang SL, Spaeth GL, Poryzees
pathological, but not physiological, ischemia-induced
EH. Cyclocryotherapy in the treatment of advanced
retinal neovascularization. J Exp Med.
glaucoma. Ophthalmology. 1985;92:947–54.
2003;198:483–9.
Cornish KS, Ramamurthi S, Saidkasimova S, Ramaesh
Jonas JB, Spandau UH, Schlichtenbrede F. Intravitreal
K. Intravitreal bevacizumab and augmented trabecu-
bevacizumab for filtering surgery. Ophthalmic Res.
lectomy for neovascular glaucoma in young diabetic
2007;39:121–2.
patients. Eye (Lond). 2009;23:979–81.
Kahook MY, Schuman JS, Noecker RJ. Intravitreal beva-
Corral CJ, Siddiqui A, Wu L, Farrell CL, Lyons D, Mustoe
cizumab in a patient with neovascular glaucoma.
TA. Vascular endothelial growth factor is more impor-
Ophthalmic Surg Lasers Imaging. 2006a;37:144–6.
tant than basic fibroblastic growth factor during isch-
Kahook MY, Schuman JS, Noecker RJ. Needle bleb revi-
emic wound healing. Arch Surg. 1999;134:200–5.
sion of encapsulated filtering bleb with bevacizumab.
Davidorf FH, Mouser JG, Derick RJ. Rapid improvement
Ophthalmic Surg Lasers Imaging. 2006b;37:148–50.
of rubeosis iridis from a single bevacizumab (Avastin)
Katz GJ, Higginbotham EJ, Lichter PR, Skuta GL, Musch
injection. Retina. 2006;26:354–6.
DC, Bergstrom TJ, Johnson AT. Mitomycin C versus
Ehlers JP, Spirn MJ, Lam A, Sivalingam A, Samuel MA,
5-fluorouracil in high-risk glaucoma filtering surgery.
Tasman W. Combination intravitreal bevacizumab/
Extended follow-up. Ophthalmology. 1995;102:
panretinal photocoagulation versus panretinal photo-
1263–9.
coagulation alone in the treatment of neovascular
Kirchhof B. Retinectomy lowers intraocular pressure in
glaucoma. Retina. 2008;28:696–702.
otherwise intractable glaucoma: preliminary results.
Elgin U, Berker N, Batman A, Simsek T, Cankaya
Ophthalmic Surg. 1994;25:262–7.
B. Trabeculectomy with mitomycin C combined with
Kitnarong N, Chindasub P, Metheetrairut A. Surgical out-
direct cauterization of peripheral iris in the manage-
come of intravitreal bevacizumab and filtration sur-
ment of neovascular glaucoma. J Glaucoma.
gery in neovascular glaucoma. Adv Ther. 2008;25:
2006;15:466–70.
438–43.
Feibel RM, Bigger JF. Rubeosis iridis and neovascular
L’Esperance Jr FA, Mittl RN, James Jr WA. Carbon diox-
glaucoma. Evaluation of cyclocryotherapy. Am
ide laser trabeculostomy for the treatment of neovas-
J Ophthalmol. 1972;74:862–7.
cular glaucoma. Ophthalmology. 1983;90:821–9.
Fernández-Vigo J, Castro J, Macarro A. Diabetic iris neo-
Li D, Zhang C, Song F, Lubenec I, Tian Y, Song QH. VEGF
vascularization. Natural history and treatment. Acta
regulates FGF-2 and TGF-beta1 expression in injury
Ophthalmol Scand. 1997;75:89–93.
endothelial cells and mediates smooth muscle cells
Gheith ME, Siam GA, de Barros DS, Garg SJ, Moster
proliferation and migration. VEGF regulates FGF-2
MR. Role of intravitreal bevacizumab in neovascular
and TGF-beta1 expression in injury endothelial cells
glaucoma. J Ocul Pharmacol Ther. 2007;23:487–91.
and mediates smooth muscle cells proliferation and Bevacizumab in eyes with diabetic retinopathy. Arch
migration. Microvasc Res. 2009;77:134–42. Ophthalmol. 2007;125:1363–6.
Lim TH, Bae SH, Cho YJ, Lee JH, Kim HK, Sohn Seghezzi G, Patel S, Ren CJ, Gualandris A, Pintucci G,
YH. Concentration of vascular endothelial growth fac- Robbins ES, Shapiro RL, Galloway AC, Rifkin DB,
tor after intracameral bevacizumab injection in eyes Mignatti P. Fibroblast growth factor-2 (FGF-2)
with neovascular glaucoma. Korean J Ophthalmol. induces vascular endothelial growth factor (VEGF)
2009;23(3):188–92. expression in the endothelial cells of forming capillar-
Lüke J, Nassar K, Lüke M, Grisanti S. Ranibizumab as ies: an autocrine mechanism contributing to angiogen-
adjuvant in the treatment of rubeosis iridis and neovas- esis. J Cell Biol. 1998;141(7):1659–73.
cular glaucoma—results from a prospective interven- Shields MB, Shields SE. Noncontact transscleal Nd:YAG
tional case series. Graefes Arch Clin Exp Ophthalmol. cyclophotocoagulation: a long-term follow-up of 500
2013;251:2403–13. patients. Trans Am Ophthalmol Soc. 1994;92:271–83.
Martin DF, Maguire MG, Fine SL, Fine SL, Ying GS, discussion 283-287.
Jaffe GJ, Grunwald JE, Toth C, Redford M, Ferris 3rd Shinoda K, Ishida S, Kawashima S, Wakabayashi T,
FL. Comparison of Age-related Macular Degeneration Uchita M, Matsuzaki T, Takayama M, Shinmura K,
Treatments Trials (CATT) Research Group Writing Yamada M. Clinical factors related to the aqueous lev-
Committee: Ranibizumab and Bevacizumab for treat- els of vascular endothelial growth factor and hepato-
ment of neovascular age-related macular degenera- cyte growth factor in proliferative diabetic retinopathy.
tion: two-year results. Ophthalmology. 2012;119: Curr Eye Res. 2000;21:655–61.
1388–98. Sivak-Callcott JA, O’Day DM, Gass JDM, Tsai
Massin P, Bandello F, Garweg JG, Hansen LL, Harding JC. Evidence-based recommendations for the diagnosis
SP, Larsen M, Mitchell P, Sharp D, Wolf-Schnurrbusch and treatment of neovascular glaucoma. Ophthalmology.
UE, Gekkieva M, Weichselberger A, Wolf S. Safety 2001;108:1767–78.
and efficacy of ranibizumab in diabetic macular edema Tolentino MJ, Miller JW, Gragoudas ES, Chatzistefanou
(RESOLVE Study): a 12-month, randomized, con- K, Ferrara N, Adamis AP. Vascular endothelial growth
trolled, double-masked, multicenter phase II study. factor is sufficient to produce iris neovascularization
Diabetes Care. 2010;33:2399–405. and neovascular glaucoma in a nonhuman primate.
Memarzadeh F, Varma R, Lin LT, Parikh JG, Dustin L, Arch Ophthalmol. 1996;114(8):964–70.
Alcaraz A, Eliott D. Postoperative use of bevacizumab Tripathi RC, Li J, Tripathi BJ, Chalam KV, Adamis
as an antifibrotic agent in glaucoma filtration surgery AP. Increased level of vascular endothelial growth fac-
in the rabbit. Invest Ophthalmol Vis Sci. tor in aqueous humor of patients with neovascular
2009;50:3233–7. glaucoma. Ophthalmology. 1998;105:232–7.
Mitchell P, Bandello F, Schmidt-Erfurth U, Lang GE, Tsai JC, Feuer WJ, Parrish 2nd RK, Grajewski AL.
Massin P, Schlingemann RO, Sutter F, Simader C, 5-Fluorouracil filtering surgery and neovascular glau-
Burian G, Gerstner O, Weichselberger A, RESTORE coma. Long-term follow-up of the original pilot study.
Study Group. The RESTORE study: ranibizumab Ophthalmology. 1995;102:887–92.
monotherapy or combined with laser versus laser Wakabayashi T, Oshima Y, Sakaguchi H, Ikuno Y, Miki
monotherapy for diabetic macular edema. A, Gomi F, Otori Y, Kamei M, Kusaka S, Tano
Ophthalmology. 2011;118:615–62. Y. Intravitreal bevacizumab to treat iris neovascular-
Parrish R, Herschler J. Eyes with end-stage neovascular ization and neovascular glaucoma secondary to isch-
glaucoma: natural history following successful modi- emic retinal diseases in 41 consecutive cases.
fied filtering operation. Arch Ophthalmol. 1983; Ophthalmology. 2008;115:1571–80.
101:745–6. Wand M, Dueker DK, Aiello LM, Grant WM. Effects of
Sawada O, Kawamura H, Kakinoki M, Sawada T, Ohji panretinal photocoagulation on rubeosis iridis, angle
M. Vascular endothelial growth factor in aqueous neovascularization, and neovascular glaucoma. Am
humor before and after intravitreal injection of J Ophthalmol. 1978;86:332–9.
Corneal Neovascular Diseases
10
Deniz Hos, Felix Bock, Björn Bachmann,
and Claus Cursiefen
the cornea (corneal neovascularization)

10.1 Introduction (Cursiefen 2007). This corneal neovasculariza-
tion (1) results in reduced corneal transparency
The healthy cornea is one of the very few avascu- and visual acuity; (2) is the most important risk
lar tissues of the body that contains no blood and factor for graft rejection after keratoplasty; and
lymphatic vessels. This avascularity, which is (3) has recently been shown to contribute to the
also termed the corneal “(lymph)angiogenic priv- development of several corneal pathologies such
ilege”, is evolutionary highly conserved and is as dry eye disease (Bock et al. 2013; Hos and
essential for corneal transparency and proper Cursiefen 2014; Hos et al. 2014; Goyal et al.
vision (Cursiefen 2007). Thus, minor inflamma- 2010). The inhibition of corneal neovasculariza-
tory and angiogenic stimuli, to which the cornea tion as a therapeutic approach has recently been
is constantly exposed due to its anatomical posi- translated into the clinic and shows promising
tion, generally do not result in corneal neovascu- results (Bock et al. 2013).
larization, as the cornea expresses potent This review will recapitulate recent findings
antiangiogenic molecules to actively maintain its in the field of corneal neovascular disease. We
avascular state (Albuquerque et al. 2009; Ambati will first give an overview of recently discovered
et al. 2006; Cursiefen et al. 2006, 2011; Singh mechanisms that contribute to the corneal angio-
et al. 2013). However, severe corneal damage can genic privilege. Second, we will present diseases
overcome this corneal angiogenic privilege and that are associated with corneal neovasculariza-
lead to the pathological ingrowths of blood as tion and review the impact of corneal neovascu-
well as lymphatic vessels from the limbus into larization on visual acuity and on corneal graft
survival after transplantation. Third, we will
present how corneal neovessels can be morpho-
metrically analyzed. Finally, we will specify cur-
D. Hos, M.D. • F. Bock, Ph.D.
B. Bachmann, M.D. (*) • C. Cursiefen M.D. rent clinically applied antiangiogenic treatment
Department of Ophthalmology, University strategies against corneal neovascularization. As
of Cologne, Kerpener Strasse 62, 50924 was declared by a recent consensus meeting,
Cologne, Germany
there is a huge unmet need for topical inhibitors
e-mail: deniz.hos@uk-koeln.de; felix.bock@
uk-koeln.de; bjoern.bachmann@uk-koeln.de; of angiogenesis at the ocular surface (Cursiefen
claus.cursiefen@uk-koeln.de et al. 2012).

160 D. Hos et al.
10.2 Corneal (Lymph)angiogenic

Privilege
Recently, several molecular mechanisms that

contribute to the corneal angiogenic privilege
could be identified. In this context, especially the
corneal epithelium seems to be crucial for the
maintenance of corneal avascularity, as it
expresses several potent antiangiogenic factors
such as endostatin, angiostatin, thrombospondin-
1, thrombospondin-2, pigment epithelium-
derived factor, and others (Armstrong and
Bornstein 2003; Cursiefen et al. 2004c, 2011;
Lin et al. 2001). These molecules are able to
directly inhibit vascular endothelial cell prolif-
eration and migration, and it has also been shown Fig. 10.1 Corneal neovascularization in a patient with
that these factors can interfere with growth fac- recurrent herpetic ulcerative keratitis. Blood vessels origi-
nate from the limbus and grow towards the center of
tor mobilization and binding. Additionally, the the cornea. Note secondary central corneal scar
corneal epithelium expresses soluble VEGF
receptor-1 (sVEGFR-1), soluble VEGFR-2
(sVEGFR-2), and soluble VEGFR-3 10.3 Corneal Neovascular Disease
(sVEGFR-3) which all act as decoy receptors
and sequester (lymph)angiogenic VEGFs 10.3.1 Corneal Diseases Resulting
(Albuquerque et al. 2009; Ambati et al. 2006; in Corneal
Singh et al. 2013). In addition, we could previ- Neovascularization
ously demonstrate that the corneal epithelium
also ectopically expresses membrane bound As mentioned in the previous chapter, severe
VEGFR-3, which is able to bind angiogenic pathological processes can lead to a breakdown of
ligands and therefore reduce binding of these to the corneal angiogenic privilege and result in sec-
VEGFR-3 expressed on lymphatic vascular ondary ingrowths of blood and lymphatic vessels
endothelial cells (Cursiefen et al. 2006). from the limbus into the cornea. These processes
Furthermore, the cornea is also able to actively include inflammatory (e.g., ocular pemphigoid,
suppress neovascularization under hypoxic con- Lyell-syndrome, Stevens–Johnson syndrome,
ditions by the expression of inhibitory PAS graft-versus-host disease, corneal graft rejection,
domain protein (IPAS), which negatively regu- dry eye disease), infectious (e.g., viral, bacterial,
lates hypoxia-induced upregulation of VEGF or fungal dermatitis), degenerative (e.g., Terrien
(Makino et al. 2001). marginal degeneration, pterygium), hypoxic (e.g.,
Due to these antiangiogenic mechanisms, cor- extended contact lens wear), and traumatic dis-
neal neovascularization usually does not occur as eases (e.g., chemical burns). Ocular pathologies
the cornea is able to buffer most minor angio- that lead to a loss of limbal barrier function (e.g.,
genic stimuli. However, severe and eye- limbal stem cell deficiency) are also frequently
threatening inflammation can overcome the accompanied by corneal neovessels. In this con-
corneal antiangiogenic mechanisms and result in text, corneal inflammation and ulceration accom-
the pathological ingrowths of blood and clini- panied by corneal neovascularization are amongst
cally invisible lymphatic vessels from the limbus the leading causes of severe vision loss world-
into the cornea (Fig. 10.1) (Cursiefen 2007). wide (Whither et al. 2002).
10 Corneal Neovascular Diseases 161
10.3.2 Reduction of Visual Acuity

and Promotion of Immune
Responses by Corneal Blood
and Lymphatic Vessels
Independent of the underlying disease, corneal

blood vessels can lead to corneal scarring and
significantly impair corneal transparency and
vision. Visual acuity can become directly
impaired by the ingrowths of corneal blood ves-
sels into the optical zone leading to light scatter-
ing or obscuration. Also secondary effects such
as corneal edema, lipid deposition, and hemor-
rhage through immature and leaky capillaries can
further reduce visual acuity. In addition, immune
effector cells such as macrophages and T lym-
phocytes can easily reach the cornea via patho-
logical blood vessels and can damage the
structural integrity of the cornea.
Clinically, improvements in visual acuity fre-
quently occur when central corneal neovascular-
ization or central lipid keratopathy decrease
(Fig. 10.2). However, peripheral corneal edema Fig. 10.2 (a) Dense corneal neovascularization associ-
or corneal obscuration outside the optic zone ated with secondary lipid keratopathy. The patient was
only able to see hand movements. (b) Improvement of
might still allow for good visual function. In con- corneal neovascularization and lipid keratopathy 1 month
trast to the clinical experience, only little evi- after fine needle diathermy of corneal vessels combined
dence in the literature suggests a positive effect with topical application of Bevacizumab. Visual acuity
of the reduction of corneal neovascularization on also increased to 20/100. The aim of this measure was to
improve graft survival after subsequent penetrating
visual acuity (Bachmann et al. 2013). So far, keratoplasty
studies are missing that investigate the space-
resolved change of corneal neovascularization
and its influence on visual acuity. However, the serve as conduits that enable accelerated migra-
localization of corneal neovascularization is tion of antigen-presenting cells from the ocular
decisive whether a successful antiangiogenic surface to the lymph nodes and contribute to—as
treatment results in improved visual acuity. in the case of these diseases—undesired immune
Future studies are needed that correlate the responses (Hos et al. 2014).
regression of centrally located corneal vessels
with visual acuity.
Unlike blood vessels, clinically invisible lym- 10.3.3 Corneal Neovascularization
phatic vessels do not lead to a significant reduc- as a Risk Factor for Corneal
tion of corneal transparency per se, but have been Transplant Rejection
shown to substantially contribute to several
corneal pathologies such as corneal transplant Corneal neovascularization has long been accepted
rejection, dry eye disease, and ocular allergy as a risk factor for graft rejection and graft failure
(Lee et al. 2015 Bock et al. 2013; Dietrich et al. after perforating corneal transplantation (Dana and
2010; Goyal et al. 2010; Hos and Cursiefen Streilein 1996; Sano et al. 1995). The cornea is
2014). Here, corneal lymphatic vessels seem to considered to be an immune privileged anatomical
162 D. Hos et al.
structure not only due to its angiogenic privilege to increase the risk for subsequent immune reac-
but also because of the absence of MHCII positive tions (Jonas et al. 2002).
antigen-presenting cells in the central part of the After corneal grafting, preexisting corneal
cornea and the expression of Fas ligand on corneal vessels can further increase. Whether this affects
endothelial cells which induces apoptosis in graft rejection has so far only been evaluated
immune competent cells expressing the Fas recep- experimentally. In mice, the additional ingrowths
tor (Niederkorn 1999, 2010). Moreover, certain of blood and lymphatic vessels after corneal
neuropeptides secreted by sensory corneal nerves transplantation leads to higher rejection rates not
seem to contribute to the corneal immune privi- only in normal corneas, but also in high-risk cor-
lege (Niederkorn 2010). Thus, graft rejection after neas (Bachmann et al. 2008, 2009; Cursiefen
keratoplasty performed in avascular normal-risk et al. 2004a). This suggests that the inhibition of
corneas occurs in less than 20 % within 5 years corneal neovascularization after keratoplasty
after transplantation even without systemic immu- might also have a protective effect against graft
nosuppressive therapy (Williams et al. 2012). rejection in humans (Fig. 10.3).
However, in prevascularized, high-risk corneas
the rejection rate is higher than 50 %, even with
systemic immunosuppressive therapy. 10.3.4 Corneal Lymphangiogenesis
The definition of high-risk corneal transplan- in Dry Eye Disease
tation is primarily based on the degree of corneal and Ocular Allergy
neovessels in the host’s cornea. Corneas are
thought to be on a higher risk for graft rejection, Studies analyzing the time course of pathologic
which is the most frequent reason for graft fail- corneal hem- and lymphangiogenesis during cor-
ure, when the host’s deep corneal stoma is vascu- neal inflammation have revealed that blood and
larized in a minimum of two quadrants. Compared lymphatic vessels usually grow in parallel into
with avascular recipients the relative risk for the cornea (Cursiefen et al. 2004b, 2006b).
graft rejection is 2.67 if more than two quadrants Furthermore, it is known that after brief inflam-
are vascularized (Bachmann et al. 2010). Based mation corneal lymphatic vessels regress earlier
on these findings, the concept of preconditioning and more completely in comparison to corneal
high-risk grafts aims on a reduction of the preop- blood vessels (Cursiefen et al. 2002b, 2006b).
erative corneal vessel load by adopting different Accordingly, vascularized corneas may contain
antiangiogenic therapies (see Sect. 10.5). only blood but no more lymphatic vessels, when
In contrast to many clinical studies evaluating inflammation is no longer present and had
the effect of preexisting corneal neovessels occurred more than 6 months ago. However, sev-
before corneal transplantation so far only limited eral experimental studies could demonstrate that
study evidence describes the influence of postop- there are also diseases of the ocular surface that
erative corneal neovascularization on graft rejec- result in selective corneal lymphangiogenesis
tion or graft failure (Altenburger et al. 2012; Lam without hemangiogenesis, namely dry eye and
et al. 2006; Cursiefen et al. 2001b, 2002a; Jonas ocular allergy.
et al. 2002). In fact, about ¾ of patients undergo- It is widely accepted that dry eye is not only a
ing penetrating keratoplasty develop post- condition with alterations in tear quantity and
keratoplasty corneal neovascularization composition, but a complex disease, where
(Altenburger et al. 2012; Cursiefen et al. 2001b, chronic inflammation is maintained by adaptive
2002a; Jonas et al. 2002; Lam et al. 2006). In immune responses (Niederkorn et al. 2006;
about 10 %, these new vessels even reach the Schaumburg et al. 2011). Recently, our group
host–graft interface (Altenburger et al. 2012; showed that in thrombospondin-1 deficient mice,
Cursiefen et al. 2001b; Lam et al. 2006). This which develop dry eye similar to Sjogren's syn-
post-keratoplasty neovascularization also seems drome, an isolated growth of corneal lymphatic
hemangiogenesis (Goyal et al. 2010). These cor-

neal lymphatics seem to facilitate trafficking of
antigen-presenting cells from the ocular surface
to the lymph nodes, where accelerated
autosensitization occurs. Importantly, selective
inhibition of lymphangiogenesis can significantly
improve the clinical course of dry eye disease, at
least in the experimental setting (Goyal et al.
2012).
Very recently, the contribution of lymphangio-
genesis to ocular allergy has also been deter-
mined. It was shown that mice with severe forms
of experimentally induced ocular allergy develop
corneal lymphatic, but not blood vessels (Lee
et al. 2015). Further analysis of allergic mice
revealed that these pathologic lymphatic vessels
enable egress of activated antigen-presenting
cells from the cornea to the regional lymph nodes
and contribute to Th-2 responses. Notably, inhi-
bition of corneal lymphangiogenesis in these
mice reduced allergy-associated Th-2 responses
and significantly ameliorated clinical disease
(Lee et al. 2015).
Taken together, corneal lymphangiogenesis
seems to play an essential role in the develop-
ment of dry eye disease and ocular allergy. This
offers an exciting and promising therapeutic tar-
get to treat these very frequently occurring
diseases.
10.4 Morphological
Characterization
and Quantification
Fig. 10.3 (a) Corneal blood vessels reaching the graft– of Corneal
host interface after perforating keratoplasty. (b) One day Neovascularization
after fine needle diathermy combined with topical appli-
cation of Bevacizumab the site of coagulation is visible. As described in the above chapters, blood and
(c) Two months after fine needle diathermy, no blood ves-
sels are visible in the graft or in the host’s cornea. This lymphatic vessels play a crucial role in several
shall reduce the risk of graft rejection after keratoplasty corneal diseases. To analyze the morphology of
corneal blood and lymphatic vessels, the cornea
provides unique properties in several ways: the
vessels occurs (Cursiefen et al. 2011). cornea is physiologically transparent and is also
Furthermore, mice that are exposed to desiccat- easily accessible for in vivo imaging. This chap-
ing stress with high air flow and low humidity ter will summarize current imaging and quantifi-
also develop corneal inflammation accompanied cation methods to characterize corneal hem- and
by lymphangiogenesis without concurrent lymphangiogenesis.
164 D. Hos et al.
10.4.1 Morphology of Corneal The most common method to quantify corneal

Vessels neovascularization is the assessment of numbers
of corneal quadrants which show vascularization
In 1998, Cursiefen et al. performed a histopatho- (Bachmann et al. 2010; Niederkorn 2003). A
logical study to analyze the incidence and local- more sophisticated method is the digital image
ization of vessels in diseased corneas. Within over analysis. This method was first used in our labo-
1200 corneal buttons obtained by keratoplasty, the ratories for the quantification of blood and lym-
group detected corneal neovascularization in phatic vessels in a murine model of corneal
about 20 % of the buttons with different disease neovascularization, where the vascularized cor-
associations. The occurrence of vessels was neas were stained immunohistochemically with
accompanied by corneal edema and an inflamma- CD31 as blood endothelial marker and LYVE-1
tory cell infiltrate. Most of the vessels were as lymph endothelial marker (Bock et al. 2008b).
located in the upper and middle third of the cor- The morphometry of corneal hem- and lymphan-
neal stroma, whereas only few vessels were giogenesis is afterwards performed semiautomat-
located in the deep stromal layers (Cursiefen et al. ically on digital images using image analysis
1998). Within the vascularized corneas in this software. The semiautomatic method is mainly
study, 87 % of the vessels were covered by peri- based on a threshold analysis. After the use of
cytes, which support the long-term, growth factor different digital filters, the blood and lymphatic
independent survival of vascular endothelial cells. vessels appear bright and are represented by gray
The coverage of corneal blood vessels by peri- values which are very well distinguishable from
cytes starts early and increases with the duration the background gray values (Fig. 10.4). We could
of neovascularization, until almost all vessels show that the semiautomatic quantification of
have recruited pericytes (Cursiefen et al. 2003). corneal neovascularization is very valid in mea-
By ultrastructural analysis it was shown that, suring the area covered by blood or lymphatic
in addition to corneal blood vessels, thin-walled, vessels and has a good reproducibility with
erythrocyte-free lymphatic vessels could be respect to both vessel types. This new semiauto-
found in human corneal buttons obtained from matic morphometry method based on threshold
keratoplasty. These lymphatics can be stained analysis therefore provides a high accuracy.
with the lymphatic markers LYVE-1 and podo- Using this method, we analyzed experimental
planin and represented about 8 % of all vessels. corneal hem- and lymphangiogenesis in various
Lymphatic vessels were mainly present in the scenarios and were able to even detect small dif-
early phase of corneal neovascularization and ferences within the analyzed groups (Bock et al.
were always accompanied by blood vessels as 2007, 2008a; Bucher et al. 2012; Dietrich et al.
well as by stromal inflammatory cells (Cursiefen 2007; Hos et al. 2008a, b, 2011a, b, 2013;
et al. 2002b). Regenfuss et al. 2010; Lipp et al. 2014). Based on
this method, further characteristics of the vessel
network like vessel endpoints, branching points,
10.4.2 Quantification of Corneal vessel length, or vessel diameter can now be
Blood and Lymphatic Vessels assessed (Blacher et al. 2009).
The method of semiautomatic quantification
As described in the previous chapters, corneal neo- of corneal vessels is also already transferred into
vascularization serves as an indicator of corneal clinical studies (Cursiefen et al. 2009, 2014). In
inflammation and is a major risk factor for corneal this context, we performed a multicenter, ran-
graft rejection (Bachmann et al. 2010). Therefore, domized, double-blind study in which we ana-
the quantification of the corneal vessel load is an lyzed the area covered by vessels in a three-month
important tool to monitor the course of ongoing follow-up time period (Fig. 10.5) (Cursiefen
inflammatory processes at the cornea as well as for et al. 2009). The extent of corneal neovascular-
the follow-up of antiangiogenic therapies. ization was determined by repeatedly performed
Fig. 10.4 Digital quantification of corneal blood and lym- vessels in red; (c) original flat mount specifically stained
phatic vessels in a mouse model of corneal vascularization. for lymphatic vessels (LYVE-1); (d) modified image of (c)
(a) Original flat mount specifically stained for blood ves- with detected lymphatic vessels in red. Modified from
sels (CD31); (b) modified image of (a) with detected blood Bock et al. 2008b, with permission of Elsevier
standardized digital slit-lamp photographs, which larization. Although several specific angiogene-
then were analyzed morphometrically using the sis inhibitors such as Ranibizumab, Aflibercept,
image analysis software based on gray filter sam- and Pegaptanib have been approved by the US
pling. The method made it possible to detect sig- Food and Drug Administration for the treatment
nificant changes of neovascularization in a range of pathologic neovascularization at the posterior
of 2 % of the total corneal area. pole of the eye, no specific angiogenesis inhibitor
against corneal neovascularization has been
approved so far. This part of the review will focus
10.5 Therapeutic Inhibition on antiangiogenic therapies at the cornea with
of Corneal already existing clinical experience.
Neovascularization
Numerous potential antiangiogenic treatment 10.5.1 Corticosteroids

strategies for corneal neovascular diseases have
already been tested in preclinical models with Corticosteroids are potent anti-inflammatory
promising results. However, only few have so far drugs which are still the clinical standard of anti-
reached the way into clinical trials or are used angiogenic therapy at the cornea (Bock et al.
off-label to treat patients with corneal neovascu- 2013; Hos et al. 2011b). These drugs are known
166 D. Hos et al.
Fig. 10.5 Digital quantification of corneal blood vessels blood vessels (red) in the modified corneal image.
in the clinical setting. (a) Original image of a vascularized Modified from Cursiefen et al. 2009, with permission of
cornea; (b) shading correction of (a); (c) gray value image Elsevier
of (b) with defined region of interest; (d) detection of
to reduce corneal hemangiogenesis and, as inflamed settings such as after penetrating kera-
recently demonstrated, also corneal lymphangio- toplasty, steroids even at high dosage are not suf-
genesis (Hos et al. 2011b). The antiangiogenic ficient to block corneal neovascularization
effect of corticosteroids is mainly indirect due to (Cursiefen et al. 2001a).
the reduction of inflammation in the cornea. In
addition, corticosteroids also show direct inhibi-
tory effects on vascular endothelial cells (Hos 10.5.2 Anti-VEGF Therapy
et al. 2011b). The anti(lymph)angiogenic potency
of corticosteroids varies with prednisolone hav- Preclinically, several specific antiangiogenic
ing the strongest effects. Nonetheless, although treatment strategies targeting VEGF have already
corticosteroids are effective antiangiogenic been tested in the past. In 2004, the VEGFR1R2
agents, the prolonged use of these drugs may trap (Aflibercept) showed an almost complete
cause several side effects such as delayed epithe- blockade of both hem- and lymphangiogenesis in
lial wound healing, elevated intraocular pressure, the inflamed mouse cornea (Cursiefen et al.
or cataract. Furthermore, although corticoste- 2004a). Further anti-VEGF therapeutic strategies
roids suppress the formation of new vessels in that were experimentally tested so far are antibod-
progressive corneal neovascular diseases, they ies directed against VEGF (Bevacizumab,
are less effective in regressing already present, Ranibizumab) (Bock et al. 2007; Bucher et al.
mature vessels. Furthermore, especially in highly 2012), aptamers binding VEGF (Pegaptanib)
(Lipp et al. 2014), VEGF receptor blocking anti- thereby contributing to intracellular signal cas-
bodies (anti-VEGFR3) (Bock et al. 2008a), cades. IRS-1 has been shown to assist in multiple
VEGF receptor tyrosine kinases (Detry et al. growth hormone and cytokine receptor signalling
2013; Hos et al. 2008b), and several others. pathways, including the VEGF/VEGFR pathway
Despite the progress in the experimental setting, (White 1998). It was shown that IRS-1 is
none of these specific inhibitors has an FDA expressed in the cornea (Andrieu-Soler et al.
approval for the use at the cornea, as already men- 2005). The blockade of IRS-1 signalling inhibits
tioned above. Nonetheless, Bevacizumab is fre- corneal hemangiogenesis and, as recently shown,
quently used off-label to treat patients with also lymphangiogenesis (Hos et al. 2011a). An
corneal neovascularization also in the context of antisense oligonucleotide against IRS-1 (GS-
corneal transplantation, and several groups have 101, Aganirsen) has successfully been tested as
already proven the effectiveness and safety of eyedrops in phase II and phase III clinical trials
Bevacizumab in inhibiting corneal hemangiogen- (Cursiefen et al. 2009, 2014) (Fig. 10.6). The
esis (Koenig et al. 2012). In addition, recent stud- phase II/III data showed that Aganirsen eyedrops
ies have also shown the effectiveness of are safe, were well tolerated and effective in
Ranibizumab in inhibiting corneal neovascular- inhibiting progressive corneal neovascularization
ization in patients (Ferrari et al. 2013). when applied twice daily (Fig. 10.6) (Cursiefen
Bevacizumab and Ranibizumab can penetrate et al. 2009, 2014). Furthermore, Aganirsen eye-
well into the cornea and reach sufficient concen- drops were able to reduce the need for corneal
trations especially in case of an absent or altered transplantation in patients with viral keratitis-
corneal epithelial barrier (as is the case in vascu- associated central corneal neovascularization
larized corneas). In this context, Dastjerdi et al. (Cursiefen et al. 2014). Aganirsen therefore holds
analyzed if Bevacizumab as a full length antibody great promise to be the first therapeutic that will
can penetrate the cornea. They found that indeed be approved for the topical treatment of corneal
the intact corneal epithelium almost completely neovascularization by the FDA. Since Aganirsen
avoids the penetration of this antibody. As soon as eyedrops in primate models also reached suffi-
the cornea is vascularized the penetration is much cient tissue levels at the retina, Aganirsen may be
better. The subconjunctival application of this a novel topical treatment option against AMD
antibody also resulted in strong deposition of the and retinal vascular diseases without the need for
drug in the corneal stroma even in healthy eyes intravitreal injections (Cloutier et al. 2012).
(Dastjerdi et al. 2011). Ranibizumab and
Bevacizumab show comparable antiangiogenic
results at the cornea although Ranibizumab seems 10.5.4 Fine Needle Diathermy
to be slightly more effective. In addition, both of Corneal Vessels
have recently been shown to block also corneal
lymphangiogenesis, at least in the experimental In mature blood vessels, a pure pharmaceutical
setting (Bock et al. 2007; Bucher et al. 2012). antiangiogenic approach is not effective in reduc-
ing the degree of corneal neovascularization, as
these vessels have pericyte covered walls and are
10.5.3 GS-101: Antisense less dependent on growth levels for the mainte-
Oligonucleotide nance of their anatomical structure. Fine needle
against IRS-1 diathermy is a method to physically occlude
mature vessels by cauterization (Koenig et al.
Insulin receptor substrate-1 (IRS-1) has recently 2012; Pillai et al. 2000). This is achieved by
been shown to be an important intracellular sig- introducing a 10-0 stainless steel side cutting nee-
nalling molecule in the angiogenic cascade. The dle into the lumen of a corneal vessel. Thereafter,
main function of IRS-1 is the recruitment of cyto- electric current is applied to the needle by the use
solic proteins to the corresponding receptors and of a diathermy device. This approach can be
168 D. Hos et al.
Fig. 10.6 Aganirsen antisense oligonucleotide eyedrops treatment. Bottom: Corresponding morphometric analy-
inhibit and regress corneal neovascularization in a patient ses. Note the significant reduction in area covered by
with herpetic keratitis. Top: photographs before treatment pathologic corneal blood vessels. Modified from Cursiefen
(a), after 90 days (b) and 180 days (c) of topical Aganirsen et al. 2014, with permission of Elsevier
combined with topical adjunct anti-VEGF treat- widely used for the treatment of patients with
ment (Koenig et al. 2009, 2012) which seems to chronic or recurrent central serous chorioretinop-
reduce the rate of reperfusion of cauterized ves- athy. Further indications for photodynamic ther-
sels and thereby the rate of retreatments (Koenig apy with Verteporfin are subfoveal choroidal
et al. 2012). In case of intense corneal neovascu- neovascularization in age-related macular degen-
larization, repeated vessel cauterizations might eration, pathologic myopia, choroidal hemangi-
be necessary (Faraj et al. 2014). Sufficient cor- oma, and polypoidal choroidal vasculopathy.
neal vessel reduction can lead to improvements Several studies have analyzed whether PDT
of lipid keratopathy and increase visual acuity with Verteporfin is an effective treatment for
(Wertheim et al. 2007). Moreover, improvements patients with corneal neovascularization and
in corneal neovascularization are associated with have demonstrated that corneal blood vessels can
a reversal of graft rejection episodes. be regressed by this approach (Brooks et al.
2004; Yoon et al. 2007; Al-Abdullah and
Al-Assiri 2011). This is particularly of interest,
10.5.5 Photodynamic Therapy since many patients present with already mature
with Verteporfin and stable corneal vessels that are refractory to
conservative treatment. These studies that aimed
Verteporfin, a benzoporphyrin derivative, is a to regress corneal blood vessels applied
photosensitizer, which selectively binds to the Verteporfin systemically. In addition, we have
endothelium of immature, actively proliferating previously shown that it is also possible to selec-
vessels. After activation of this compound by tively regress corneal lymphatic vessels with
nonthermal red light in the presence of oxygen, PDT, when Verteporfin is applied locally by
highly reactive short-lived singlet oxygen and intrastromal injection. After local application,
other reactive oxygen radicals are produced and Verteporfin seems to selectively bind to the drain-
cause damage to the proliferating endothelium ing lymphatics, which can then be selectively
(photodynamic therapy, PDT). Verteporfin is regressed by PDT (Bucher et al. 2014).
10.6 Concluding Remarks Funding/Support German Research Foundation: DFG

Cu 47/4-1 (CC), DFG Cu 47/6-1 (CC), DFG
Forschergruppe FOR 2240 “(Lymph)Angiogenesis and
Although actively maintained by a variety of Cellular Immunity in Inflammatory Diseases of the Eye”
anti(lymph)angiogenic mechanisms, the (lymph) (DH, FB, CC); EU COST BM1302 (DH, FB, BB, CC);
angiogenic privilege of the cornea is not invinci- GEROK-Programme, University of Cologne (DH)
ble and can be overcome by severe diseases of the
ocular surface. As corneal blood and lymphatic
vessels reduce visual acuity, increase the risk for References
graft rejection after corneal transplantation and
Al-Abdullah AA, Al-Assiri A. Resolution of bilateral cor-
contribute to the development and maintenance
neal neovascularization and lipid keratopathy after
of a variety of corneal diseases such as dry eye photodynamic therapy with verteporfin. Optometry.
and ocular allergy, anti(lymph)angiogenic ther- 2011;82(4):212–4. doi:10.1016/j.optm.2010.09.012.
apy has emerged as a novel therapeutic concept Albuquerque RJ, Hayashi T, Cho WG, Kleinman ME,
Dridi S, Takeda A, Baffi JZ, Yamada K, Kaneko H,
to treat patients with these diseases. Several
Green MG, Chappell J, Wilting J, Weich HA,
anti(lymph)angiogenic treatment strategies have Yamagami S, Amano S, Mizuki N, Alexander JS,
been established so far and include conservative Peterson ML, Brekken RA, Hirashima M, Capoor S,
therapy with anti-inflammatory or specific Usui T, Ambati BK, Ambati J. Alternatively spliced
vascular endothelial growth factor receptor-2 is an
anti(lymph)angiogenic pharmaceuticals as well
essential endogenous inhibitor of lymphatic vessel
as invasive surgical and laser-assisted approaches. growth. Nat Med. 2009;15(9):1023–30. doi:10.1038/
Although no specific (lymph)angiogenesis inhib- nm.2018.
itor against corneal neovascularization has been Altenburger AE, Bachmann B, Seitz B, Cursiefen
C. Morphometric analysis of postoperative corneal
approved so far, several promising candidates
neovascularization after high-risk keratoplasty: her-
have already reached the clinic, with others likely petic versus non-herpetic disease. Graefes Arch Clin
to follow. Exp Ophthalmol. 2012;250(11):1663–71. doi:10.1007/
s00417-012-1988-6.
Ambati BK, Nozaki M, Singh N, Takeda A, Jani PD,
Suthar T, Albuquerque RJ, Richter E, Sakurai E,
Newcomb MT, Kleinman ME, Caldwell RB, Lin Q,
Compliance with Ethical Requirements Ogura Y, Orecchia A, Samuelson DA, Agnew DW, St
Leger J, Green WR, Mahasreshti PJ, Curiel DT, Kwan
Authors DH, FB, and BB declare that they
D, Marsh H, Ikeda S, Leiper LJ, Collinson JM,
have no conflict of interest. Bogdanovich S, Khurana TS, Shibuya M, Baldwin
Author CC is a consultant for Gene ME, Ferrara N, Gerber HP, De Falco S, Witta J, Baffi
Signal, France JZ, Raisler BJ, Ambati J. Corneal avascularity is due
to soluble VEGF receptor-1. Nature.
All procedures followed were in accor-
2006;443(7114):993–7. doi:10.1038/nature05249.
dance with the ethical standards of the Andrieu-Soler C, Berdugo M, Doat M, Courtois Y,
responsible committee on human experi- BenEzra D, Behar-Cohen F. Downregulation of IRS-1
mentation (institutional and national) and expression causes inhibition of corneal angiogenesis.
Invest Ophthalmol Vis Sci. 2005;46(11):4072–8.
with the Helsinki Declaration of 1975, as
doi:10.1167/iovs.05-0105.
revised in 2000. Informed consent was Armstrong LC, Bornstein P. Thrombospondins 1 and 2
obtained from all patients for being function as inhibitors of angiogenesis. Matrix Biol.
included in the study. 2003;22(1):63–71.
Bachmann B, Taylor RS, Cursiefen C. Corneal neovascu-
All institutional and national guidelines
larization as a risk factor for graft failure and rejection
for the care and use of laboratory animals after keratoplasty: an evidence-based meta-analysis.
were followed. Ophthalmology. 2010;117(7):1300–5. doi:10.1016/j.
ophtha.2010.01.039. e1307.
170 D. Hos et al.
Bachmann B, Taylor RS, Cursiefen C. The association Hadri A, Cursiefen C, Udaondo P, Viaud E, Thorin E,
between corneal neovascularization and visual acuity: Chemtob S. Antiangiogenic activity of aganirsen in
a systematic review. Acta Ophthalmol. 2013;91(1):12– nonhuman primate and rodent models of retinal neo-
9. doi:10.1111/j.1755-3768.2011.02312.x. vascular disease after topical administration. Invest
Bachmann BO, Bock F, Wiegand SJ, Maruyama K, Dana Ophthalmol Vis Sci. 2012;53(3):1195–203.
MR, Kruse FE, Luetjen-Drecoll E, Cursiefen doi:10.1167/iovs.11-9064.
C. Promotion of graft survival by vascular endothelial Cursiefen C. Immune privilege and angiogenic privilege
growth factor a neutralization after high-risk corneal of the cornea. Chem Immunol Allergy. 2007;92:50–7.
transplantation. Arch Ophthalmol. 2008;126(1):71–7. doi:10.1159/000099253.
doi:10.1001/archopht.126.1.71. Cursiefen C, Bock F, Horn FK, Kruse FE, Seitz B,
Bachmann BO, Luetjen-Drecoll E, Bock F, Wiegand SJ, Borderie V, Fruh B, Thiel MA, Wilhelm F, Geudelin
Hos D, Dana R, Kruse FE, Cursiefen C. Transient B, Descohand I, Steuhl KP, Hahn A, Meller D. GS-101
postoperative vascular endothelial growth factor antisense oligonucleotide eye drops inhibit corneal
(VEGF)-neutralisation improves graft survival in cor- neovascularization: interim results of a randomized
neas with partly regressed inflammatory neovasculari- phase II trial. Ophthalmology. 2009;116(9):1630–7.
sation. Br J Ophthalmol. 2009;93(8):1075–80. doi:10.1016/j.ophtha.2009.04.016.
doi:10.1136/bjo.2008.145128. Cursiefen C, Cao J, Chen L, Liu Y, Maruyama K, Jackson
Blacher S, Detry B, Bruyere F, Foidart JM, Noel D, Kruse FE, Wiegand SJ, Dana MR, Sterile
A. Additional parameters for the morphometry of JW. Inhibition of hemangiogenesis and lymphangio-
angiogenesis and lymphangiogenesis in corneal flat genesis after normal-risk corneal transplantation by
mounts. Exp Eye Res. 2009;89(2):274–6. neutralizing VEGF promotes graft survival. Invest
doi:10.1016/j.exer.2009.02.021. Ophthalmol Vis Sci. 2004a;45(8):2666–73.
Bock F, Maruyama K, Regenfuss B, Hos D, Steven P, doi:10.1167/iovs.03-1380.
Heindl LM, Cursiefen C. Novel anti(lymph)angio- Cursiefen C, Chen L, Borges LP, Jackson D, Cao J,
genic treatment strategies for corneal and ocular sur- Radziejewski C, D’Amore PA, Dana MR, Wiegand
face diseases. Prog Retin Eye Res. 2013;34:89–124. SJ, Sterile JW. VEGF-A stimulates lymph angiogene-
doi:10.1016/j.preteyeres.2013.01.001. sis and hemangiogenesis in inflammatory neovascu-
Bock F, Onderka J, Dietrich T, Bachmann B, Kruse FE, larization via macrophage recruitment. J Clin Invest.
Paschke M, Zahn G, Cursiefen C. Bevacizumab as a 2004b;113(7):1040–50. doi:10.1172/JCI20465.
potent inhibitor of inflammatory corneal angiogenesis Cursiefen C, Chen L, Saint-Geniez M, Hamrah P, Jin Y,
and lymphangiogenesis. Invest Ophthalmol Vis Sci. Rashid S, Pytowski B, Persaud K, Wu Y, Sterile JW,
2007;48(6):2545–52. doi:10.1167/iovs.06-0570. Dana R. Nonvascular VEGF receptor 3 expression by
Bock F, Onderka J, Dietrich T, Bachmann B, Pytowski B, corneal epithelium maintains avascularity and vision.
Cursiefen C. Blockade of VEGFR3-signalling specifi- Proc Natl Acad Sci U S A. 2006a;103(30):11405–10.
cally inhibits lymphangiogenesis in inflammatory cor- doi:10.1073/pnas.0506112103.
neal neovascularisation. Graefes Arch Clin Exp Cursiefen C, Colin J, Dana R, Diaz-Llopis M, Faraj LA,
Ophthalmol. 2008a;246(1):115–9. doi:10.1007/ Garcia-Delpech S, Geerling G, Price FW, Remeijer L,
s00417-007-0683-5. Rouse BT, Seitz B, Udaondo P, Meller D, Dua
Bock F, Onderka J, Hos D, Horn F, Martus P, Cursiefen H. Consensus statement on indications for anti-
C. Improved semiautomatic method for morphometry angiogenic therapy in the management of corneal dis-
of angiogenesis and lymphangiogenesis in corneal eases associated with neovascularisation: outcome of
flatmounts. Exp Eye Res. 2008b;87(5):462–70. an expert roundtable. Br J Ophthalmol. 2012;96(1):3–
doi:10.1016/j.exer.2008.08.007. 9. doi:10.1136/bjo.2011.204701.
Brooks BJ, Ambati BK, Marcus DM, Ratanasit Cursiefen C, Hofmann-Rummelt C, Kuchle M, Schlotzer-
A. Photodynamic therapy for corneal neovascularisa- Schrehardt U. Pericyte recruitment in human corneal
tion and lipid degeneration. Br J Ophthalmol. angiogenesis: an ultrastructural study with clinico-
2004;88(6):840. pathological correlation. Br J Ophthalmol. 2003;
Bucher F, Bi Y, Gehlsen U, Hos D, Cursiefen C, Bock 87(1):101–6.
F. Regression of mature lymphatic vessels in the cor- Cursiefen C, Kuchle M, Naumann GO. Angiogenesis in
nea by photodynamic therapy. Br J Ophthalmol. corneal diseases: histopathologic evaluation of 254
2014;98(3):391–5. doi:10.1136/bjophthalmol-2013- human corneal buttons with neovascularization.
303887. Cornea. 1998;17(6):611–3.
Bucher F, Parthasarathy A, Bergua A, Onderka J, Cursiefen C, Martus P, Nguyen NX, Langenbucher A,
Regenfuss B, Cursiefen C, Bock F. Topical Seitz B, Kuchle M. Corneal neovascularization after
Ranibizumab inhibits inflammatory corneal hem- and nonmechanical versus mechanical corneal trephina-
lymphangiogenesis. Acta Ophthalmol. 2012. tion for non-high-risk keratoplasty. Cornea. 2002a;
doi:10.1111/j.1755-3768.2012.02525.x. 21(7):648–52.
Cloutier F, Lawrence M, Goody R, Lamoureux S, Cursiefen C, Maruyama K, Bock F, Saban D, Sadrai Z,
Al-Mahmood S, Colin S, Ferry A, Conduzorgues JP, Lawler J, Dana R, Masli S. Thrombospondin 1 inhibits
inflammatory lymph angiogenesis by CD36 ligation Dietrich T, Onderka J, Bock F, Kruse FE, Vossmeyer D,
on monocytes. J Exp Med. 2011;208(5):1083–92. Stragies R, Zahn G, Cursiefen C. Inhibition of inflam-
doi:10.1084/jem.20092277. matory lymphangiogenesis by integrin alpha5 block-
Cursiefen C, Maruyama K, Jackson DG, Sterile JW, ade. Am J Pathol. 2007;171(1):361–72.
Kruse FE. Time course of angiogenesis and lymph Faraj LA, Elalfy MS, Said DG, Dua HS. Fine needle dia-
angiogenesis after brief corneal inflammation. Cornea. thermy occlusion of corneal vessels. Br J Ophthalmol.
2006b;25(4):443–7. doi:10.1097/01.ico.0000183485. 2014;98(9):1287–90. doi:10.1136/bjophthalmol-2014-
85636.ff. 304891.
Cursiefen C, Masli S, Ng TF, Dana MR, Bornstein P, Ferrari G, Dastjerdi MH, Okanobo A, Cheng SF, Amparo
Lawler J, Sterile JW. Roles of thrombospondin-1 and F, Nallasamy N, Dana R. Topical Ranibizumab as a
-2 in regulating corneal and iris angiogenesis. Invest treatment of corneal neovascularization. Cornea.
Ophthalmol Vis Sci. 2004c;45(4):1117–24. 2013;32(7):992–7.doi:10.1097/ICO.0b013e3182775f8d.
Cursiefen C, Schlotzer-Schrehardt U, Kuchle M, Sorokin Goyal S, Chauhan SK, Dana R. Blockade of prolymphan-
L, Breiteneder-Geleff S, Alitalo K, Jackson giogenic vascular endothelial growth factor C in dry
D. Lymphatic vessels in vascular zed human corneas: eye disease. Arch Ophthalmol. 2012;130(1):84–9.
immunohistochemical investigation using LYVE-1 doi:10.1001/archophthalmol.2011.266.
and podoplanin. Invest Ophthalmol Vis Sci. Goyal S, Chauhan SK, El Annan J, Nallasamy N, Zhang
2002b;43(7):2127–35. Q, Dana R. Evidence of corneal lymphangiogenesis in
Cursiefen C, Viaud E, Bock F, Geudelin B, Ferry A, dry eye disease: a potential link to adaptive immunity?
Kadlecova P, Levy M, Al Mahmood S, Colin S, Thorin Arch Ophthalmol. 2010;128(7):819–24. doi:10.1001/
E, Majo F, Frueh B, Wilhelm F, Meyer-Ter-Vehn T, archophthalmol.2010.124.
Geerling G, Bohringer D, Reinhard T, Meller D, Hos D, Bachmann B, Bock F, Onderka J, Cursiefen
Pleyer U, Bachmann B, Seitz B. Aganirsen antisense C. Age-related changes in murine limbal lymphatic
oligonucleotide eye drops inhibit dermatitis-induced vessels and corneal lymphangiogenesis. Exp Eye Res.
corneal neovascularization and reduce need for trans- 2008a;87(5):427–32. doi:10.1016/j.exer.2008.07.013.
plantation: The I-CAN Study. Ophthalmology. 2014. Hos D, Bock F, Dietrich T, Onderka J, Kruse FE, Thierauch
doi:10.1016/j.ophtha.2014.03.038. KH, Cursiefen C. Inflammatory corneal (lymph)angio-
Cursiefen C, Wenkel H, Martus P, Langenbucher A, genesis is blocked by VEGFR-tyrosine kinase inhibi-
Nguyen NX, Seitz B, Kuchle M, Naumann GO. Impact tor ZK 261991, resulting in improved graft survival
of short-term versus long-term topical steroids on cor- after corneal transplantation. Invest Ophthalmol Vis
neal neovascularization after non-high-risk kerato- Sci. 2008b;49(5):1836–42. doi:10.1167/iovs.07-1314.
plasty. Graefes Arch Clin Exp Ophthalmol. Hos D, Cursiefen C. Lymphatic vessels in the develop-
2001a;239(7):514–21. ment of tissue and organ rejection. Adv Anat Embryol
Cursiefen C, Wenkel H, Martus P, Langenbucher A, Seitz Cell Biol. 2014;214:119–41. doi:10.1007/978-3-7091-
B, Kuchle M. Standardized semiquantitative analysis 1646-3_10.
of corneal neovascularization using projected corneal Hos D, Koch KR, Bucher F, Bock F, Cursiefen C, Heindl
photographs--pilot study after perforating corneal ker- LM. Serum eyedrops antagonize the anti(lymph)
atoplasty before immune reaction. Klin Monbl angiogenic effects of Bevacizumab in vitro and
Augenheilkd. 2001b;218(7):484–91. doi:10.1055/ in vivo. Invest Ophthalmol Vis Sci. 2013;54(9):6133–
s-2001-16291. 42. doi:10.1167/iovs.13-12460.
Dana MR, Streilein JW. Loss and restoration of immune Hos D, Regenfuss B, Bock F, Onderka J, Cursiefen
privilege in eyes with corneal neovascularization. C. Blockade of insulin receptor substrate-1 inhibits
Invest Ophthalmol Vis Sci. 1996;37(12):2485–94. corneal lymphangiogenesis. Invest Ophthalmol Vis
Dastjerdi MH, Sadrai Z, Saban DR, Zhang Q, Dana Sci. 2011a;52(8):5778–85. doi:10.1167/iovs.10-6816.
R. Corneal penetration of topical and subconjunctival Hos D, Saban DR, Bock F, Regenfuss B, Onderka J, Masli
bevacizumab. Invest Ophthalmol Vis Sci. S, Cursiefen C. Suppression of inflammatory corneal
2011;52(12):8718–23. doi:10.1167/iovs.11-7871. lymphangiogenesis by application of topical cortico-
Detry B, Blacher S, Erpicum C, Paupert J, Maertens L, steroids. Arch Ophthalmol. 2011b;129(4):445–52.
Maillard C, Munaut C, Sounni NE, Lambert V, Foidart doi:10.1001/archophthalmol.2011.42.
JM, Rakic JM, Cataldo D, Noel A. Sunitinib inhibits Hos D, Schlereth SL, Bock F, Heindl LM, Cursiefen
inflammatory corneal lymph angiogenesis. Invest C. Antilymphangiogenic therapy to promote trans-
Ophthalmol Vis Sci. 2013;54(5):3082–93. doi:10.1167/ plant survival and to reduce cancer metastasis: what
iovs.12-10856. can we learn from the eye? Semin Cell Dev Biol.
Dietrich T, Bock F, Yuen D, Hos D, Bachmann BO, Zahn 2014;38:117–30. doi:10.1016/j.semcdb.2014.11.003.
G, Wiegand S, Chen L, Cursiefen C. Cutting edge: Jonas JB, Rank RM, Budde WM. Immunologic graft
lymphatic vessels, not blood vessels, primarily medi- reactions after allogenic penetrating keratoplasty. Am
ate immune rejections after transplantation. J Ophthalmol. 2002;133(4):437–43.
J Immunol. 2010;184(2):535–9. doi:10.4049/ Koenig Y, Bock F, Horn F, Kruse F, Straub K, Cursiefen
jimmunol.0903180. C. Short- and long-term safety profile and efficacy of
172 D. Hos et al.
topical bevacizumab (Avastin) eyedrops against cor- induces T cell-mediated Sjogren’s syndrome-like
neal neovascularization. Graefes Arch Clin Exp lacrimal keratoconjunctivitis. J Immunol. 2006;
Ophthalmol. 2009;247(10):1375–82. doi:10.1007/ 176(7):3950–7.
s00417-009-1099-1. Pillai CT, Dua HS, Hossain P. Fine needle diathermy
Koenig Y, Bock F, Kruse FE, Stock K, Cursiefen occlusion of corneal vessels. Invest Ophthalmol Vis
C. Angioregressive pretreatment of mature corneal Sci. 2000;41(8):2148–53.
blood vessels before keratoplasty: fine-needle vessel Regenfuss B, Onderka J, Bock F, Hos D, Maruyama K,
coagulation combined with anti-VEGFs. Cornea. Cursiefen C. Genetic heterogeneity of lymphangiogen-
2012;31(8): 887–92. doi:10.1097/ICO.0b013e esis in different mouse strains. Am J Pathol.
31823f8f7a. 2010;177(1):501–10. doi:10.2353/ajpath.2010.090794.
Lam VM, Nguyen NX, Martus P, Seitz B, Kruse FE, Sano Y, Ksander BR, Streilein JW. Fate of orthotopic cor-
Cursiefen C. Surgery-related factors influencing corneal allografts in eyes that cannot support anterior
neal neovascularization after low-risk keratoplasty. chamber-associated immune deviation induction.
Am J Ophthalmol. 2006;141(2):260–6. doi:10.1016/j. Invest Ophthalmol Vis Sci. 1995;36(11):2176–85.
ajo.2005.08.080. Schaumburg CS, Siemasko KF, De Paiva CS, Wheeler LA,
Lee HS, Hos D, Blanco T, Bock F, Reyes NJ, Mathew R, Niederkorn JY, Pflugfelder SC, Stern ME. Ocular sur-
Cursiefen C, Dana R, Saban DR. Involvement of face APCs are necessary for autoreactive T cell-mediated
corneal lymphangiogenesis in a mouse model of experimental autoimmune lacrimal keratoconjunctivitis.
allergic eye disease. Invest Ophthalmol Vis Sci. J Immunol. 2011;187(7):3653–62. doi:10.4049/
2015;56(5):3140–8. jimmunol.1101442.
Lin HC, Chang JH, Jain S, Gabison EE, Kure T, Kato T, Singh N, Tiem M, Watkins R, Cho YK, Wang Y, Olsen T,
Fukai N, Azar DT. Matrilysin cleavage of corneal col- Uehara H, Mamalis C, Luo L, Oakey Z, Ambati
lagen type XVIII NC1 domain and generation of a BK. Soluble vascular endothelial growth factor recep-
28-kDa fragment. Invest Ophthalmol Vis Sci. tor-3 is essential for corneal alymphaticity. Blood.
2001;42(11):2517–24. 2013;121(20):4242–9. doi:10.1182/blood-2012-
Lipp M, Bucher F, Parthasarathy A, Hos D, Onderka J, 08-453043.
Cursiefen C, Bock F. Blockade of the VEGF isoforms Wertheim MS, Cook SD, Knox-Cartwright NE, Van DL,
in inflammatory corneal hemangiogenesis and lym- Tole DM. Electrolysis-needle cauterization of corneal
phangiogenesis. Graefes Arch Clin Exp Ophthalmol. vessels in patients with lipid keratopathy. Cornea.
2014;252(6):943–9. doi:10.1007/s00417-014-2626-2. 2007;26(2):230–1. doi:10.1097/01.ico.0000248383.
Makino Y, Cao R, Svensson K, Bertilsson G, Asman M, 09272.ee.
Tanaka H, Cao Y, Berkenstam A, Poellinger Whitcher JP, Srinivasan M, Upadhyay MP. Prevention of
L. Inhibitory PAS domain protein is a negative regula- corneal ulceration in the developing world. Int
tor of hypoxia-inducible gene expression. Nature. Ophthalmol Clin. 2002;42(1):71–7.
2001;414(6863):550–4. doi:10.1038/35107085. White MF. The IRS-signaling system: a network of dock-
Niederkorn JY. The immune privilege of corneal ing proteins that mediate insulin and cytokine action.
allografts. Transplantation. 1999;67(12):1503–8. Recent Prog Horm Res. 1998;53:119–38.
Niederkorn JY. The immune privilege of corneal grafts. Williams KA, Lowe MT, Keane MC, Jones VJ, Loh RS,
J Leukoc Biol. 2003;74(2):167–71. Coster DJ. The Australian Corneal Graft Registry
Niederkorn JY. High-risk corneal allografts and why they 2012 Report. Adelaide, Australia: SnapPrinting; 2012,
lose their immune privilege. Curr Opin Allergy Clin p. 235–42. Accessed 1 October 2014.
Immunol. 2010;10(5):493–7. doi:10.1097/ACI.0b013 Yoon KC, You IC, Kang IS, Im SK, Ahn JK, Park YG,
e32833dfa11. Ahn KY. Photodynamic therapy with verteporfin for
Niederkorn JY, Stern ME, Pflugfelder SC, De Paiva CS, corneal neovascularization. Am J Ophthalmol.
Corrales RM, Gao J, Siemasko K. Desiccating stress 2007;144(3):390–5. doi:10.1016/j.ajo.2007.05.028.
Anti-Angiogenic Gene Therapy:
Basic Science and Challenges 11
for Translation into the Clinic
Clemens Lange and James Bainbridge
tive diabetic retinopathy (Rofagha et al. 2013;

11.1 Introduction Salam et al. 2001). The success of this treatment,
however, is limited by the relative short half-life
The growth of abnormal blood vessels (neovas- of antibodies in the eye and depends on fre-
cularisation) is a central feature of retinal disor- quently repeated injections that pose a significant
ders including retinopathy of prematurity (ROP), cumulative risk of local complications including
retinal vein occlusions (RVO), proliferative dia- intraocular infection, vitreous haemorrhage, and
betic retinopathy (PDR), and neovascular age- retinal detachment.
related macular degeneration (nAMD), which are Ocular gene therapy is an emerging therapeu-
the leading causes of blindness in infants, indi- tic approach that offers the potential for targeted,
viduals of working age and the elderly in the sustained, and regulatable delivery of angiostatic
Western world, respectively. Although neovascu- proteins to the retina after a single procedure.
larisation occurs at a relatively late stage of dis- The eye has major advantages as a target organ
ease, it is nonetheless an attractive target for for gene therapy. It is a well-confined and highly
therapeutic intervention, since it represents a compartmentalised organ that is readily accessi-
pathway common to many different diseases and ble for vector administration by microsurgical
typically leads directly to visual loss. Intraocular techniques. Vector suspensions can be delivered
delivery of antibodies directed against VEGF has precisely under direct visualisation into the
become the conventional treatment modality for desired compartment, including the vitreous
neovascular AMD and increasingly for prolifera- body and the subretinal or suprachoroidal spaces,
enabling specific ocular cell targeting with mini-
mal risk of systemic dissemination. The blood–
retina barrier further limits the systemic spread
of locally administered vectors and reduces the
C. Lange, M.D., Ph.D.
Eye Center, University Hospital Freiburg, possibility of systemic adverse events and expo-
Killianstrasse 5, Freiburg 79106, Germany sure to vector antigens that might cause inflam-
e-mail: Clemens.lange@uniklinik-freiburg.de mation and limit transgene expression. Because
J. Bainbridge, Ph.D., F.R.C.Ophth. (*) of its small size, only small volumes of vector
Department of Genetics, UCL Institute of suspensions are required to reach a significant
Ophthalmology, 11-43 Bath Street, London
proportion of the desired target cells.
EC1V 9EL, UK
Furthermore, the eye’s unique immune environ-
Moorfields Eye Hospital, London, UK
ment confers additional protection against
e-mail: j.bainbridge@ucl.ac.uk

174 C. Lange and J. Bainbridge
immune responses to foreign antigens (Forrester term transgene expression in tissues of the ante-
and Xu 2012) that could otherwise limit long- rior segment and the retinal pigment epithelium,
term transgene expression (Bennett et al. 1996). but transduce photoreceptor cells with only low
Finally, therapeutic effects on structure and efficiency (Bainbridge et al. 2001; Balaggan
function can be easily observed, recorded, and et al. 2006a). Lentiviral vectors promote integra-
quantified using a variety of in vivo techniques tion of the transferred gene into the host chromo-
both experimentally and clinically. These advan- some, posing a risk of oncogenesis as a result of
tages have led to numerous preclinical studies insertional mutagenesis. For this reason, self-
and the first clinical gene therapy trials to treat inactivating and non-integrating lentiviral vec-
patients with Leber congenital amaurosis tors are being developed as potentially safer
(Bainbridge et al. 2008; Hauswirth et al. 2008; alternatives.
Maguire et al. 2008) choroideremia (MacLaren Adenoviral vectors are non-integrating viruses
et al. 2014), and exudative age-related macular that efficiently target both dividing and non-
degeneration (Campochiaro et al. 2006). dividing cells in the outer retina (Ali et al. 1998;
Bennett et al. 1994). However, the duration of
gene expression in the targeted cell is limited by
11.2 Concepts of Ocular Gene a significant T-cell-mediated immune response
Therapy (Reichel et al. 1998). This limitation might be
overcome by the use of helper-dependent Ad vec-
The fundamental concepts of gene therapy were tors, which lengthen the duration of transgene
developed in the 1960s following the identifica- expression from shorter than 3 months to up to 1
tion of nucleic acids and the discovery of viral year (Lamartina et al. 2007).
transduction. Recombinant DNA techniques Recombinant adeno-associated virus (rAAV)
enabled the cloning of genes that could be used to vectors are currently the vector of choice for
correct genetic defects and disease phenotypes in ocular gene therapy. They typically do not inte-
mammalian cells in vitro (Friedmann 1992). grate into the genome but persist mainly as
Since then, significant advances in vector design extragenomic circular episomes, with only low
and gene transfer methods have been made in the risk of insertional oncogenesis. rAAV vectors
field of human gene therapy. In the eye, gene mediate stable, long-term transgene expression
transfer strategies can be used for gene supple- in a variety of retinal cells, including photore-
mentation in recessive disease, gene inactivation ceptors, RPE cells, ganglion cells, and Müller
in dominant disease, expression of neuroprotec- cells, with minimal immune response. Most
tive mediators in degenerative disease, “suicide recombinant AAV vectors are composed of the
genes” for example in tumour diseases, and genomic component of AAV serotype 2 pack-
expression of immunomodulatory and angio- aged using capsids derived from other AAV
static factors in immunological and neovascular serotypes. The capsid serotype determines the
eye disease. tropism and efficacy of the vector. Whereas
Many different viral and non-viral vector sys- AAV2/2 (i.e., the genome based on AAV-2 and
tems, designed to improve the efficiency of gene packaged in AAV-2 capsid), 2/5, and 2/8 trans-
delivery to target cells, have been evaluated and duce the RPE and photoreceptors following sub-
optimised for use in the eye (Reichel et al. 1999; retinal delivery. AAV2/1, AAV2/6, and AAV2/4
Buch et al. 2008). Among these vectors, lentivi- exclusively target the RPE, and AAV2/9 effi-
rus (LV), adenovirus (AV), and adeno-associated ciently transduces Müller cells (for a detailed
virus (AAV) are the most frequently used viral review, see Buch et al. 2008). The AAV variant
vector systems for gene therapy. Lentiviral vec- ShH10 efficiently and selectively transduces
tors such as HIV, SIV, or EIAV are retroviruses Müller cells by about 94 % following intravitreal
and able to transduce non-dividing cells. In the injection (Klimczak et al. 2009), and other AAV
eye, these vector systems can induce stable, long- variants have been reported to efficiently trans-
11 Anti-Angiogenic Gene Therapy: Basic Science and Challenges for Translation into the Clinic 175
duce human and rat astrocytes (Koerber et al. long-term VEGF blockade could be highly effec-
2009). A major limitation of AAV vectors, how- tive. Anti-VEGF gene therapy can be achieved by
ever, is their small payload capacity, which is inducing the expression of molecules that bind
limited to 5 KB irrespective of the capsid type extracellular VEGF, such as the VEGF receptor
(Wu et al. 2010). Therefore, large genes or sFlt1 (see Fig. 11.1) or VEGF antibodies, or by
extensive promoter/regulatory elements cannot inhibiting intracellular VEGF protein expression
be introduced. This limitation may be addressed using sh- and siRNA. Soluble fms-like tyrosine
by a trans-splicing technique in which a large kinase-1 (sFlt1 or sVEGFR-1) is a splice variant
coding sequence is divided into smaller of the VEGF receptor (Flt1) and a potent inhibi-
sequences for delivery, and subsequently recom- tor of extracellular VEGF. Numerous reports
bined intracellularly. confirm the efficiency of sFlt1 gene therapy in
reducing the development of ocular neovascu-
larisation (see Tables 11.1 and 11.2). Systemic,
11.3 Gene Therapy intravitreal or subretinal injection of Ad or AAV
for Neovascular Eye Disease vectors carrying the sFlt1 gene suppress the
development of retinal (Lamartina et al. 2007;
The unique potential of gene therapy is that a Rota et al. 2004; Bainbridge et al. 2002; Lai et al.
single intraocular injection of a vector can lead to 2009) and CNV in rodents (Lai et al. 2001, 2009;
the sustained expression of an angiostatic protein Honda et al. 2000; Igarashi et al. 2010), and
to address the chronic pro-angiogenic drive in AAV-mediated sFlt-1 expression reduces the
neovascular eye disease. The principle of angio- development of laser-induced CNV in non-
static gene therapy has been intensively exam- human primates (Lai et al. 2005). Furthermore,
ined in various animal models of ocular the intraocular injection of AAV2 viruses encod-
neovascularisation. Ocular neovascularisation ing sFLT01, a chimeric VEGF-binding molecule
can be induced experimentally via focal laser consisting of domain 2 of Flt-1 linked to a human
photocoagulation of Bruch’s membrane, which IgG1 heavy-chain fragment, suppresses RNV
causes choroidal neovascularisation (CNV) to and CNV in mice and primates without histologi-
develop, or by exposing rodents to hyperoxia cal evidence of toxicity for 12 months (Pechan
which causes retinal ischaemia and the develop- et al. 2009; Lukason et al. 2011). Similarly, a
ment of retinal neovascularisation (RNV). single intravitreal administration of AAV vectors
Furthermore, the effect of ocular gene therapy expressing the VEGF antibody bevacizumab was
can be assessed in various transgenic mice which shown to cause long-term expression of bevaci-
develop spontaneous RNV and CNV. Although zumab in the RPE which was associated with
none of these models recapitulates the full suppression of neovascularisation in transgenic
spectrum of pathology of human diseases, they mice overexpressing human VEGF165 in photo-
model some features in the development of RNV receptors (Mao et al. 2011). Moreover, subretinal
and CNV and have proven invaluable in proof-of- delivery of AAV2 vectors encoding anti-VEGF
concept experiments for anti-angiogenic drug siRNA and shRNA or the VEGF-binding intra-
development. ceptor Flt23k reduced both VEGF protein levels
and laser-induced CNV in mice (Igarashi et al.
2014; Askou et al. 2012; Zhang et al. 2015).
11.3.1 Candidate Genes Notably, a safety study in cynomolgus monkeys
for Angiostatic Gene Therapy using AAV2.sFLT01 demonstrated no cell-
mediated immune reaction or adverse effects on
11.3.1.1 VEGF-Inhibiting Molecules retinal vascular or electroretinal function for up
Because of the convincing clinical success of to 1 year after intraocular administration
anti-VEGF therapy in neovascular eye disease, a (Maclachlan et al. 2011). These encouraging pre-
gene therapy approach that enables sustained clinical data in non-human primates led to two
1 2 3
Promoter Transgene
4 5 Expression 6 Expression
of sFIt Sequestration of sFlt
VEGF-A
of VEGF-A
VEGF
receptor
Fig. 11.1 Schematic drawing of anti-VEGF (sFLT1) angi- The transgene is then transported into the nucleus, integrates
ostatic gene therapy. The angiostatic transgene (e.g. sFLT1) in the genome, or rests episomal depending on the virus
is introduced into viral vectors (1). Upon intravitreal or sub- (4) and is continuously expressed by the host cells (5).
retinal delivery of the viral vector in the eye (2) the viruses These angiostatic proteins are secreted and antagonise
bind to cellular surface antigens (3) and enter the target cells. VEGF action (6). Illustrated by Scott Robbie, London
phase I/II clinical trials currently investigating the vitreous of eyes with PDR and nAMD com-
the safety and tolerability of AAV2-SFlt01 pared with healthy eyes (Holekamp et al.
(NCT01024998, sponsored by Genzyme) and 2002; Spranger et al. 2000). As PEDF inhibits
rAAV.sFlt-1 (NCT01494805, sponsored by the hypoxia-induced VEGF expression at the tran-
Lions Eye Institute, Perth, Australia) in neovas- scriptional level, it has been postulated that the
cular AMD. These studies are expected to com- decrease in PEDF is at least partially responsible
plete in July 2018 and May 2017 respectively for increased VEGF levels and the associated
(see Table 11.3). vascular leakage and neovascularisation in neo-
vascular eye disease (Zhang et al. 2006).
11.3.1.2 Pigment Epithelium-Derived Furthermore, PEDF promotes the survival of
Factor neurons in vitro and protects photoreceptors from
Pigment epithelium-derived factor (PEDF) is a excessive light exposure (Cao et al. 2001; Steele
multifunctional secreted protein possessing both et al. 1993). Vector-mediated overexpression of
anti-angiogenic and neuroprotective properties. PEDF therefore offers the attractive possibility of
It is constitutively secreted by Müller and RPE both countering harmful VEGF-induced
cells and acts as a potent endogenous angiostatic angiogenesis and preventing retinal cell death in
molecule which contributes to the avascularity of neovascular eye disease. Intravitreal or subretinal
the outer retina in the healthy state. Reciprocal to injection of adenoviruses encoding human PEDF
VEGF, PEDF levels are significantly reduced in (AdPEDF.11) suppresses the development of
Table 11.1 Summary of preclinical angiostatic gene therapy studies for retinal neovascularisation
Gene therapy in experimental models of ocular neovascularisation
Retinal neovascularisation
Gene Vector Delivery Experimental model Result Reference
Angiostatin HIV Intravitreal OIR mouse model Reduced RNV in 90 % Igarashi et al. (2003)
of animals
CD59 AAV2/8 Intravitreal Streptozotocin Reduced leakage and Adhi et al. (2013)
diabetes non-perfusion
Endostatin Ad-Tx Subretinal Transgenic mice Reduced Takahashi et al. (2003)
vasopermeability and
RNV
Endostatin rAAV2/1 Subretinal OIR mouse model Reduced RNV Auricchio et al.
(2002b)
PEDF Ad Intravitreal OIR mouse model Reduced RNV Mori et al. (2001)
PEDF rAAV2 Intravitreal OIR mouse model Reduced RNV by 74 % Raisler et al. (2002)
PEDF AAV2 Intravitreal Transgenic mice Reduced RNV and Haurigot et al. (2012)
capillary dropout
TIMP3 rAAV2/1 Subretinal OIR mouse model Reduced RNV Auricchio et al.
(2002b)
Vasohibin Ad Intravitreal OIR mouse model Reduced RNV Shen et al. (2006)
Vasoinhibin AAV2 Intravitreal VEGF-induced VP Reduced Ramírez et al. (2011)
vasopermeability
Anti-VEGF
Anti-VEGF Lentivirus Subretinal Transfected Müller
OIR rat model Wang et al. (2013)
shRNA cells, reduced RNV
Anti-VEGF164 Lentivirus Subretinal OIR rat model Transfection of Müller Jiang et al. (2014)
shRNA cells, reduced RNV
Bevacizumab AAVrh.10 Intravitreal Transgenic mice Reduced RNV by 90 % Mao et al. (2011)
for 168 days
sFLT01 AAV2 Intravitreal Non-human primate Reduced CNV, Lukason et al. (2011)
transfected ganglion
cells
sFLT1 HD-Ad/dox Intravitreal OIR rat model transfected Müller cells Lamartina et al. (2007)
reduced RNV
sFLT1 Ad Intravitreal OIR rat model Reduced RNV by 97 % Rota et al. (2004)
sFLT1 Ad or AAV Intravitreal OIR mouse model Reduced RNV by more Bainbridge et al.
than 50 % (2002)
sFLT1 rAAV Intravitreal Transgenic mice Reduced RNV, Lai et al. (2009)
leukocyte accumulation
sFLT1 chimaera AAV2 Intravitreal OIR mouse model Reduced RNV, no Pechan et al. (2009)
toxicity for 12 months
VEGF peptides rAAV Intravitreal OIR mouse model Reduced RNV by Deng et al. (2005)
71–83 %
HD-Ad helper-dependent adenovirus, dox doxycycline-inducible, VP vasopermeability. Updated November 2014
RNV and CNV in mice. Interestingly, when Furthermore, periocular injection of the same
given after neovascularisation has become estab- vector results in an efficient transduction of the
lished, overexpression of PEDF is associated episclera, PEDF penetration into the choroid, and
with the apoptosis of activated, but not quiescent reduced laser-induced CNV in mice and pigs
endothelial cells, and the regression of pathologi- (Gehlbach et al. 2003; Saishin et al. 2005).
cal neovascularisation (Mori et al. 2001, 2002). Intraocular delivery of AAV-encoding PEDF
Table 11.2 Summary of preclinical angiostatic gene therapy studies for choroidal neovascularisation
Choroidal neovascularisatio
Gene Vector Delivery Experimental model Result Reference
Angiostatin rAAV Subretinal Laser-induced CNV, rat Reduced CNV area for Lai et al. (2001)
>150 days
Angiostatin EIAV Subretinal Laser-induced CNV, Reduced CNV area by Balaggan et al.
mouse 50 % (2006b)
CD59 Ad Subretinal Laser-induced CNV, Reduced MAC and Cashman et al.
mouse CNV formation (2011)
Endostatin EIAV Subretinal Laser-induced CNV, Reduce CNV area by Balaggan et al.
mouse 60 % (2006b)
PEDF AAV Subretinal Laser-induced CNV, Reduced CNV Mori et al. (2002)
mouse
PEDF Ad Subretinal Laser-induced CNV, Reduced CNV Mori et al. (2001)
mouse
PEDF Ad Periocular Laser-induced CNV, pig Reduced CNV Saishin et al. (2005)
PEDF Ad Intravitreal Cynomolgus monkey Dose-related Rasmussen et al.
inflammation (2001)
PEDF Ad Periocular Laser-induced CNV, Increased choroidal Gehlbach et al.
mouse PEDF, reduced CNV (2003)
TIMP3 Nonviral HVJ Subretinal Laser-induced CNV, rat Reduced CNV Takahashi et al.
(2000)
Anti-VEGF
Anti-VEGF AAV2 Subretinal Laser-induced CNV, Expression for >6 Zhang et al. (2015)
intraceptor mouse months, CNV
(Flt23k) reduction by 62 %, no
AE observed
Anti-VEGF AAV Subretinal Laser-induced CNV, Reduced CNV by 84 % Cashman et al.
shRNA mouse (2006)
Anti-VEGF scAAV2/8 Subretinal Laser-induced CNV, Reduced CNV by 48 % Askou et al. (2012)
shRNA mouse
Anti-VEGF AAV2/8 Subretinal Laser-induced CNV, VEGF reduced by 50 Igarashi et al.
siRNA mouse %, CNV reduced (2014)
sFLT1 AAV8 Subretinal Laser-induced CNV, Reduced CNV Igarashi et al.
mouse (2010)
sFLT1 AAV Subretinal Laser-induced CNV, Long-term expression Lai et al. (2005)
mouse and monkeys of Flt1, CNV
suppression in 85 %
sFLT1 (3 Nonviral/EP Ciliary Laser-induced CNV, rat Reduced CNV, VEGF El Sanharawi et al.
variants) muscle inhibition for 6M (2013)
sFLT1 Ad Systemic Laser-induced CNV, rat Reduced fibroblast Honda et al. (2000)
proliferation and
inflammatory cell
infiltration in CNV
HVJ hemagglutinating virus of Japan liposomes, scAAV2/8-hU6-sh9 self-complementary AAV vectors were packaged
in serotype 8 capsids, Updated November 2014
leads to long-term PEDF production, suppressed progression of human DR (Haurigot et al. 2012).
retinal and choroidal NV (Mori et al. 2002; Toxicology studies performed in non-human pri-
Raisler et al. 2002), and is associated with mates demonstrate a dose-related inflammatory
reduced RNV, normal retinal capillary density, response which is minimal and fully reversible at
and reduced incidence of retinal detachment in a doses below 1 × 109 (Rasmussen et al. 2001).
transgenic mouse model that mimics the chronic These encouraging preclinical results led to a
Table 11.3 Summary of human trials of angiostatic retinal gene therapy (updated November 2014)
Ongoing human trials of retinal gene therapy for neovascular eye disease
Neovascular AMD
Estimated
Gene Vector Delivery Description Status completion Country Reference
sFLT-01 AAV2 Intravitreal Phase I, open-label, ongoing July 2018 USA NCT01024998
multi-centre,
dose-escalating,
safety and tolerability
study sponsored by
Genzyme
sFLT-1 rAAV Subretinal Phase I/II, controlled Ongoing May 2017 Australia NCT01494805
dose-escalating trial,
safety, and efficacy
study
Endostatin/ EIAV, Subretinal Phase I, dose- Ongoing March USA NCT01301443
Angiostatin lentivirus escalating safety 2015
study sponsored by
Oxford BioMedica
PEDF Ad Intravitreal Phase I, open-label, Completed N/A USA NCT00109499
single-administration, Campochiaro
dose-escalation study et al. (2006)
sponsored by GenVec
phase-I clinical trial of ocular gene therapy inves- 11.3.1.3 Endostatin and Angiostatin
tigating the safety of subretinal delivery of ade- Endostatin is a naturally occurring protein pro-
noviruses encoding PEDF (AdPEDF) in 28 duced by proteolytic cleavage of collagen type
patients with neovascular AMD. Following a 18, which acts as an endogenous inhibitor of
single intravitreal delivery of AdPEDF, no seri- angiogenesis. The subretinal injection of lentivi-
ous adverse events related to the virus occurred, ral vectors encoding for Endostatin was associ-
nor were any dose-limiting toxicities docu- ated with an increased expression of Endostatin
mented. Mild, transient intraocular inflammation in the RPE and reduced vasopermeability and
was evident in seven patients, and raised intra- RNV in transgenic mice (Takahashi et al. 2003).
ocular pressure in six patients was controlled by The intravitreal or subretinal administration of
topical medication alone. Sputum and urine cul- lentiviruses or AAV containing the Endostatin
tures were negative for replicating adenovirus, gene is associated with attenuation of laser-
though, serum-neutralising antibodies directed induced CNV (Mori et al. 2002; Balaggan et al.
against adenovirus were raised in several patients 2006b) and of RNV in the OIR mouse model
indicating a systemic response following intravit- (Auricchio et al. 2002a) indicating that overex-
real vector delivery. Following administration of pression of Endostatin restricts the development
low-dose vector, the mean CNV lesion size dou- of ocular neovascularisation. Angiostatin, on the
bled from 6 to 12 months, whereas following other hand, is a naturally occurring fragment of
administration of a higher dose the CNV size plasmin, which inhibits endothelial cell migra-
remained stable. This study demonstrated an tion and induces apoptosis (O’Reilly et al. 1997).
encouraging safety profile and suggested angio- Intravitreal injection of either adeno-associated
static activity following a single intravitreal or lentiviral vectors encoding Angiostatin sup-
injection of AdPEDF in AMD (Campochiaro presses RNV and CNV in mice (Lai et al. 2001;
et al. 2006). Despite these findings, no phase-II Raisler et al. 2002; Balaggan et al. 2006b;
trial has been reported to date. Igarashi et al. 2003). Based on these preclinical
data, a phase-I clinical trial is currently ongoing (Hauswirth et al. 2008; Maguire et al. 2008;
to determine the bioactivity and safety of subreti- Banin et al. 2010). Follow-up has surpassed 5
nal delivery of EIAV lentiviral vectors encoding years, with no adverse sight-threatening inflam-
both Endostatin and Angiostatin in neovascular mation reported to date (Simonelli et al. 2010;
AMD (Retinostat, Oxford BioMedica UK Ltd., Testa et al. 2013; Jacobson et al. 2012).
NCT01301443). This trial is expected to have Furthermore bio-distribution studies have
completed by March 2015. detected no systemic spread of AAV vector in the
tears, saliva, serum, or semen, except for tran-
11.3.1.4 Miscellaneous Factors sient AAV detected in one subject in serum and
In addition to the factors previously described, tears for a few days after surgery. Furthermore,
several other angiostatic and neuroprotective neutralising antibodies against rAAV2/2 were
molecules are the subject of investigation for undetectable in the serum with the exception of
potential delivery in neovascular eye disease by two subjects with transient marginally increased
gene therapy (see Tables 11.1 and 11.2). These titres (Hauswirth et al. 2008; Maguire et al. 2008)
factors include the Tissue Inhibitor (for a detailed review, see Willett and Bennett
Metalloproteinase 3 (Auricchio et al. 2002b; 2013). Taken together, these data indicate a mild-
Takahashi et al. 2000), Vasohibin (Wakusawa adaptive humoural immune responses and an
et al. 2011; Shen et al. 2006; Ramírez et al. 2011), acceptable safety profile of AAV ocular gene
and CD59 (Adhi et al. 2013; Cashman et al. therapy in humans.
2006). Owing to the limited data available, these
factors are not discussed in detail. 11.4.1.2 Oncogenesis
Since the first gene therapy trial for x-linked
severe combined immunodeficiency (X-SCID)
11.4 Challenges for Translation reported iatrogenic acute lymphoblastic leukae-
into the Clinic mia after the ex vivo retroviral transduction of
bone marrow cells (Hacein-Bey-Abina et al.
11.4.1 Potential Side Effects 2003), oncogenesis has been a major concern in
of Ocular Gene Therapy the field. Ectopic integration of vector DNA into
the host genome has the theoretical potential to
The duration of vector-mediated transgene promote oncogenesis by upregulating endoge-
expression in the eye is not yet known. AAV- nous proto-oncogenes as a consequence of pro-
mediated transgene expression appears to last the moter/enhancer sequences in the vector genes, or
entire life in rodents and for more than 2½ years by disrupting tumour-suppressor genes. In con-
in non-human primates (Lebherz et al. 2005). trast to retroviral vectors, which promote chromo-
Although such long-term expression is an advan- somal integration, rAAV vectors pose minimal
tage from one viewpoint, it also raises potential risk of mutagenic effects. The vast majority of
concerns. rAAV vector genomes remain episomal in the
host cell nucleus (Schnepp et al. 2003) and inte-
11.4.1.1 Immune Response gration, if it occurs at all, is typically passive,
Uncontrolled immune responses can both limit largely random and of low frequency (Flotte et al.
the duration of expression of the transferred gene 1994; Ponnazhagan et al. 1997). rAAV integration
and cause direct injury to intraocular tissues. For studies in the mouse demonstrate that vector DNA
these reasons, the benefit of gene therapy depends integration occurs mainly in transcription units
on immune tolerance to the vector and the gene (53–72 %) and rarely (3.5 %) near or within can-
transferred. Four independent clinical gene ther- cer-related genes (Nakai et al. 2003; 2005). In
apy trials for LCA have assessed cellular and human cell culture models, rAAV integrated with
humoural immune responses and safety in human a slight preference to transcription units, but no
subjects following subretinal delivery of AAV2 integration near cancer-related genes has been
reported (Miller et al. 2005). The oncogenic 2008; Pellissier et al. 2014). Furthermore, anti-
potential of rAAV2 in non-dividing cells, such as VEGF gene therapy using AAV2.sFLT01 in cyno-
the naturally quiescent non-dividing cells in the molgus monkeys caused no adverse effects on
eye, is likely to be lower still. The intraocular retinal vascular or electroretinal function for up to
delivery of rAAV2 in p53 tumour-suppressor- 1 year after intraocular administration (Maclachlan
gene knockout mice, which are highly susceptible et al. 2011). Long-term follow-up data of the 2
to intraocular tumour formation, results in no ongoing anti-VEGF gene therapy trials in exuda-
increased risk of intraocular tumours (Balaggan tive AMD is expected to provide further data on
et al. 2012). Furthermore, clinical trials of ocular the safety of sustained anti-VEGF therapy.
gene therapy have reported no safety concerns of
tumourigenesis up to 5 years following subretinal
rAAV vector administration. Taken together, 11.5 Future Directions
these findings are evidence of the highly limited
oncogenic potential of rAAV vectors following 11.5.1 Choosing the Optimal Vector
intraocular administration. However, long-term System
safety data from clinical trials still needs to be
awaited to exclude with confidence the risk of 11.5.1.1 Cell-Specific Vector Systems
rAAV-induced ocular malignant transformation. The optimal choice of a vector system depends
on the desired target cells, the vector’s safety pro-
11.4.1.3 Inference file, transfection efficiency, and the longevity of
with Vasoprotection transgene expression. In the last few years, rAAV
and Neuroprotection viral vectors have become the vector of choice
Long-term suppression of VEGF in the treatment because of their long-term expression (Lebherz
of neovascular eye disease may risk compromis- et al. 2005), minimal immune response, and
ing its role in maintaining the health of normal favourable safety profile (Simonelli et al. 2010).
choroidal and retinal vasculature, as well as that of Initial studies have used mainly AAV2 serotypes,
retinal neurones. Experimental studies in mice but ongoing research has identified new variants
demonstrate that VEGF secreted by the RPE is of AAV based on engineered variants or naturally
essential to maintain the choriocapillaris and for occurring AAV serotypes (Klimczak et al. 2009).
the neuroprotection of photoreceptors. Genetic These refined AAV variants are expected to
inactivation of the soluble VEGF isoforms enable more specific targeting of angiogenic
VEGF120 and VEGF164 in mice is associated molecules to the cell population of interest,
with progressive choroidal atrophy followed by thereby reducing the risks for adverse effects.
retinal degeneration (Saint-Geniez et al. 2009). Transgene delivery can be made more specific by
Similarly, repeated systemic or intraocular injec- cell-specific promoter systems which restrict
tions of neutralising VEGF antibodies causes reti- transgene expression to the cell population of
nal ganglion cell degeneration in the adult rat interest. Inner retinal vascular disease, such as
(Nishijima et al. 2007) and systemic administra- PDR and RVO, require transgene delivery to
tion of adenoviruses containing sFLt1 (Ad-sFlt1) inner retinal cells, whereas outer retinal vascular
cause inner and outer nuclear layers degeneration disease such as neovascular AMD may benefit
(Saint-Geniez et al. 2008). These findings high- primarily from transgene expression in the pho-
light the important role of endogenous VEGF in toreceptors and RPE. The former may be achieved
the maintenance and function of the adult retina by introducing the GFAP, CD44, or RLBP1 pro-
and raise safety concerns of long-term VEGF sup- moter in the vector, which have been reported to
pression by gene therapy. Other studies, however, be cell specific for Müller cells (Prentice et al.
have challenged these findings and suggest that 2011; Greenberg et al. 2007; Esumi et al. 2004).
prolonged blockade of VEGF does not damage As Müller cells are one of the major sources for
retinal neurons or vasculature in mice (Ueno et al. VEGF production upon inner retinal hypoxia
(Watkins et al. 2013), this approach may enable sion systems have been applied to various isch-
specific suppression of VEGF production in the aemic disease models including tumours, the
affected cells and thereby reduce any potential ischaemic myocardium, stroke, and injured spi-
side-effects associated with VEGF inhibition in nal cord (Kim et al. 2009). In the eye, delivery of
other cell types. On the other hand, cell-specific rAAV expressing GFP under an HRE promoter
gene therapy to suppress CNV may be achieved results in transgene expression at sites of laser-
by using photoreceptor-specific promoters such induced CNV in mice, but not elsewhere
as human rhodopsin kinase (Khani et al. 2007) or (Bainbridge et al. 2003). Recently, this approach
RPE-specific promoters such as RPE65 or VMD2 was combined with a cell-specific promoter sys-
(Esumi et al. 2004). tem, demonstrating that an HRE-RPE65-GFP
vector system led to strong transgene expression
11.5.1.2 Regulation of Transgene in hypoxic RPE cells, but it elicited no transgene
Expression expression in other hypoxic cell types or in nor-
Long-term uncontrolled constitutive transgene moxic cultured RPE cells (Dougherty et al.
expression may in certain circumstances be unde- 2008). Similarly, a hypoxia-regulated, retinal
sirable, for example, where this might undermine glial cell-specific AAV vector was used success-
physiological functions essential for retinal and fully in the OIR mouse model leading to cell-
vascular homeostasis in the mature retina. specific transgene expression in hypoxic Müller
Pharmacologically inducible promoter systems cells in vivo (Prentice et al. 2011). These auto-
offer an attractive strategy to regulate transgene initiated gene therapy approaches offer an attrac-
expression temporally in response to the systemic tive and intelligent way to achieve expression of
delivery of an appropriate drug, i.e. mifepristone, angiostatic proteins for early intervention in neo-
ecdysone, doxycycline, or rapamycin. Preclinical vascularisation, when it is most sensitive to
studies have demonstrated promising results of inhibition.
inducible and long-term transgene expression in
the retina following subretinal delivery of doxy- 11.5.1.3 Non-viral Gene Therapy
cyclin- or rapamycin-inducible AAV2 vector sys- Because of the limited payload capacity and
tems (Auricchio et al. 2002a; Lebherz et al. 2005; potential immunogenicity and oncogenicity of
Folliot et al. 2003; McGee Sanftner et al. 2001). viral vector systems, there is a persistent need to
However, this approach depends on accurate dos- refine and develop alternative gene delivery strat-
ing and regular clinical observation for adverse egies. Non-viral gene transfer has historically
effects and is limited by inefficient drug penetra- been limited by short-lived and inefficient gene
tion across the blood–retina barrier. expression compared with viral vectors.
An alternative attractive approach is to couple Liposomes delivered topically and intravitreally
transgene expression with a tissue-responsive have displayed poor transfection efficacy in the
promoter which drives transgene expression retina and RPE, attributed to the inner limiting
dependent on the local tissue environment, for membrane and blood–retina barrier that restricts
example, tissue hypoxia, a central factor in the passage of liposomes (Masuda et al. 1996).
development of neovascular eye disease (Lange Subretinal delivery of liposomes demonstrated
et al. 2011; Stefansson et al. 2011). This can be better results with varying degrees of expression
achieved by utilising the hypoxia-response ele- and longevity (Hangai et al. 1996). Recent
ment (HRE), a specific enhancer present in a advances however have overcome some of these
number of angiogenic genes. Under hypoxic con- limitations through the use of refined liposomes,
ditions, the activated hypoxia-inducible factor-1 microparticles, or nanoparticles, which enhance
binds to the HRE sequence and drives the expres- nucleic acid stability and promote their cellular
sion of VEGF, EPO, and other hypoxia-regulated internalisation. Subretinal delivery of compacted
molecules. Hypoxia-inducible transgene expres- DNA nanoparticles was reported to transfect
nearly all RPE- and photoreceptor cells and pro- static gene therapy studies, further clinical trials
duced expression levels almost equal to those of are currently investigating the safety and efficacy
rhodopsin without provoking any immune of vectors expressing molecules that target the
responses in rodents (Farjo et al. 2006). VEGF pathway and others that overexpress
Furthermore, adjunctive techniques such as elec- Endostatin and Angiostatin in neovascular
tric current (electroporation or iontophoresis), AMD. These clinical trials are expected to pro-
high hydrostatic pressure, or ultrasound (sono- vide valuable information regarding the efficacy
phoresis) are now being evaluated to enhance and safety of angiostatic gene therapy in neovas-
nucleic-acid delivery to cells (Andrieu-Soler cular eye disease. Although gene therapy appears
et al. 2006). Electroporation of plasmids after acceptably safe in preclinical models, care needs
subretinal administration in neonatal mice is to be taken with regard to long-term safety and
associated with efficient transfection of RPE and efficacy of ocular gene therapy. Potential adverse
photoreceptor cells and can mediate protection effects such as oncogenesis, neuronal toxicity,
against retinal degeneration (Chen and Cepko and immune responses need to be excluded
2009). Similarly, efficient and prolonged RPE before gene therapy can become a routine ther-
transfection can be achieved in the adult rat after apy for neovascular eye disease. To address the
electroporation of subretinal plasmids containing possibility of local or systemic adverse effects,
specific RPE promoters (Kachi et al. 2006). novel strategies can restrict transgene expression
Electroporation of sFlt-1 into the ciliary muscle to the desired cell population by cell-specific pro-
can mediate a sustained reduction in VEGF lev- moters and viral tropisms. The incorporation of
els and reduced vascular leakage in laser-induced regulatable, pharmacological, or tissue-
CNV in rats (El Sanharawi et al. 2013). Taken responsive elements into the vector construct can
together, these non-viral vector systems and provide additional safety by preventing or reduc-
adjunct techniques may offer efficient, long- ing uncontrolled transgene expression. In this
term, and safe transfection of the retina and RPE way, sites of active neovascularisation may be
that can be considered for future clinical effectively targeted while minimising inappropri-
application. ate expression elsewhere, or during periods of
angiogenic inactivity. Finally, next-generation
AAV vectors are expected to enable more rapid
11.6 Conclusions onset of expression and more efficient transduc-
tion of retinal cells following intravitreal vector
Although considerable progress has been made delivery, which is potentially safer than subretinal
in our understanding and treatment of neovascu- delivery. These developments may be combined
lar eye disease, there is a significant unmet need to offer the possibility of a cell-specific, regulat-
for the means to deliver angiostatic molecules to able, efficient, and safe transgene expression sys-
the retina in a sustained way. Gene therapy for tem that can provide lasting protection of sight in
neovascular eye disease is an emerging strategy neovascular eye disease.
that may prove to be safe, efficacious, fast-acting,
and long-lasting following a single intraocular
injection. Vector-mediated expression offers the Compliance with Ethical Requirements
additional potential advantage of regulated The authors declare no competing financial
expression in response to exogenous or endoge- interests.
nous cues. Given the impact of anti-VEGF anti- No human or animal studies were per-
body treatment in neovascular eye disease and formed by the authors for this chapter.
the encouraging findings of preclinical angio-
References Balaggan KS, Binley K, Esapa M, MacLaren RE, Iqball

S, Duran Y, Pearson RA, Kan O, Barker SE, Smith
AJ, Bainbridge JW, Naylor S, Ali RR. EIAV vector-
Adhi M, Cashman SM, Kumar-Singh R. Adeno-
mediated delivery of endostatin or angiostatin inhib-
associated virus mediated delivery of a non-membrane
its angiogenesis and vascular hyperpermeability in
targeted human soluble CD59 attenuates some aspects
experimental CNV. Gene Ther. 2006b;
of diabetic retinopathy in mice. PLoS One. 2013;8(10):
13(0969–7128):1153–65.
79661.
Balaggan KS, Duran Y, Georgiadis A, Thaung C, Barker
Ali RR, Reichel MB, Byrnes AP, Stephens CJ, Thrasher
SE, Buch PK, MacNeil A, Robbie S, Bainbridge JWB,
AJ, Baker D, Hunt DM, Bhattacharya SS. Co-injection
Smith AJ, Ali RR. Absence of ocular malignant trans-
of adenovirus expressing CTLA4-Ig prolongs adeno-
formation after sub-retinal delivery of rAAV2/2 or
virally mediated lacZ reporter gene expression in the
integrating lentiviral vectors in p53-deficient mice.
mouse retina. Gene Ther. 1998;5(11):1561–5.
Gene Ther. 2012;19(2):182–8.
Andrieu-Soler C, Bejjani R-A, de Bizemont T, Normand
Banin E, Bandah-Rozenfeld D, Obolensky A, Cideciyan
N, BenEzra D, Behar-Cohen F. Ocular gene therapy: a
AV, Aleman TS, Marks-Ohana D, Sela M, Boye S,
review of nonviral strategies. Mol Vis. 2006;12:
Sumaroka A, Roman AJ, Schwartz SB, Hauswirth
1334–47.
WW, Jacobson SG, Hemo I, Sharon D. Molecular
Askou AL, Pournaras J-AC, Pihlmann M, Svalgaard JD,
anthropology meets genetic medicine to treat blind-
Arsenijevic Y, Kostic C, Bek T, Dagnaes-Hansen F,
ness in the North African Jewish population: human
Mikkelsen JG, Jensen TG, Corydon TJ. Reduction of
gene therapy initiated in Israel. Hum Gene Ther.
choroidal neovascularization in mice by adeno-
2010;21(12):1749–57.
associated virus-delivered anti-vascular endothelial
Bennett J, Wilson J, Sun D, Forbes B, Maguire A.
growth factor short hairpin RNA. J Gene Med.
Adenovirus vector-mediated in vivo gene transfer into
2012;14(11):632–41.
adult murine retina. Invest Ophthalmol Vis Sci.
Auricchio A, Rivera VM, Clackson T, O’Connor EE,
1994;35(5):2535–42.
Maguire AM, Tolentino MJ, Bennett J, Wilson
Bennett J, Pakola S, Zeng Y, Maguire A. Humoral
JM. Pharmacological regulation of protein expression
response after administration of E1-deleted adenovi-
from adeno-associated viral vectors in the eye. Mol
ruses: immune privilege of the subretinal space. Hum
Ther. 2002a;6(2):238–42.
Gene Ther. 1996;7(14):1763–9.
Auricchio A, Behling KC, Maguire AM, O’Connor EM,
Buch PK, Bainbridge JW, Ali RR. AAV-mediated gene
Bennett J, Wilson JM, Tolentino MJ. Inhibition of reti-
therapy for retinal disorders: from mouse to man.
nal neovascularization by intraocular viral-mediated
Gene Ther. 2008;15(11):849–57.
delivery of anti-angiogenic agents. Mol Ther. 2002b;
Campochiaro PA, Nguyen QD, Shah SM, Klein ML, Holz
6(4):490–4.
E, Frank RN, Saperstein DA, Gupta A, Stout JT,
Bainbridge JW, Stephens C, Parsley K, Demaison C,
Macko J, DiBartolomeo R, Wei LL. Adenoviral
Halfyard A, Thrasher AJ, Ali RR. In vivo gene trans-
vector-delivered pigment epithelium-derived factor
fer to the mouse eye using an HIV-based lentiviral vec-
for neovascular age-related macular degeneration:
tor; efficient long-term transduction of corneal
results of a phase I clinical trial. Hum Gene Ther.
endothelium and retinal pigment epithelium. Gene
2006;17(2):167–76.
Ther. 2001;8(21):1665–8.
Cao W, Tombran-Tink J, Elias R, Sezate S, Mrazek D,
Bainbridge JWB, Mistry A, De Alwis M, Paleolog E,
McGinnis JF. In vivo protection of photoreceptors
Baker A, Thrasher AJ, Ali RR. Inhibition of retinal
from light damage by pigment epithelium-derived fac-
neovascularisation by gene transfer of soluble VEGF
tor. Invest Ophthalmol Vis Sci. 2001;42(7):1646–52.
receptor sFlt-1. Gene Ther. 2002;9(5):320–6.
Cashman SM, Bowman L, Christofferson J, Kumar-Singh
Bainbridge JW, Mistry A, Binley K, De AM, Thrasher AJ,
R. Inhibition of choroidal neovascularization by
Naylor S, Ali RR. Hypoxia-regulated transgene expres-
adenovirus-mediated delivery of short hairpin RNAs
sion in experimental retinal and choroidal neovascular-
targeting VEGF as a potential therapy for AMD. Invest
ization. Gene Ther. 2003;10(0969–7128):1049–54.
Ophthalmol Vis Sci. 2006;47(8):3496–504.
Bainbridge JWB, Smith AJ, Barker SS, Robbie S,
Cashman SM, Ramo K, Kumar-Singh R. A non
Henderson R, Balaggan K, Viswanathan A, Holder
membrane-targeted human soluble CD59 attenuates
GE, Stockman A, Tyler N, Petersen-Jones S,
choroidal neovascularization in a model of age
Bhattacharya SS, Thrasher AJ, Fitzke FW, Carter BJ,
related macular degeneration. PLoS One. 2011;6(4):
Rubin GS, Moore AT, Ali RR. Effect of gene therapy
e19078.
on visual function in Leber’s congenital amaurosis. N
Chen B, Cepko CL. HDAC4 regulates neuronal survival
Engl J Med. 2008;358(21):2231–9.
in normal and diseased retinas. Science. 2009;
Balaggan KS, Binley K, Esapa M, Iqball S, Askham Z,
323(5911):256–9.
Kan O, Tschernutter M, Bainbridge JWB, Naylor S,
Deng W-T, Yan Z, Dinculescu A, Pang J, Teusner JT,
Ali RR. Stable and efficient intraocular gene transfer
Cortez NG, Berns KI, Hauswirth WW. Adeno-
using pseudotyped EIAV lentiviral vectors. J Gene
associated virus-mediated expression of vascular
Med. 2006a;8(3):275–85.
endothelial growth factor peptides inhibits retinal neo-
vascularization in a mouse model of oxygen-induced pression prevents neovascularization in a murine adult

retinopathy. Hum Gene Ther. 2005;16(11):1247–54. model of retinopathy. PLoS One. 2012;7(7):e41511.
Dougherty CJ, Smith GW, Dorey CK, Prentice HM, Hauswirth WW, Aleman TS, Kaushal S, Cideciyan AV,
Webster KA, Blanks JC. Robust hypoxia-selective Schwartz SB, Wang L, Conlon TJ, Boye SL, Flotte
regulation of a retinal pigment epithelium-specific TR, Byrne BJ, Jacobson SG. Treatment of Leber con-
adeno-associated virus vector. Mol Vis. genital amaurosis due to RPE65 mutations by ocular
2008;14:471–80. subretinal injection of adeno-associated virus gene
El Sanharawi M, Touchard E, Benard R, Bigey P, Escriou vector: short-term results of a phase I trial. Hum Gene
V, Mehanna C, Naud M-C, Berdugo M, Jeanny J-C, Ther. 2008;19(10):979–90.
Behar-Cohen F. Long-term efficacy of ciliary muscle Holekamp NM, Bouck N, Volpert O. Pigment epithelium-
gene transfer of three sFlt-1 variants in a rat model of derived factor is deficient in the vitreous of patients
laser-induced choroidal neovascularization. Gene with choroidal neovascularization due to age-related
Ther. 2013;20(11):1093–103. macular degeneration. Am J Ophthalmol. 2002;
Esumi N, Oshima Y, Li Y, Campochiaro PA, Zack 134(2):220–7.
DJ. Analysis of the VMD2 promoter and implication Honda M, Sakamoto T, Ishibashi T, Inomata H, Ueno
of E-box binding factors in its regulation. J Biol Chem. H. Experimental subretinal neovascularization is
2004;279(18):19064–73. inhibited by adenovirus-mediated soluble VEGF/flt-1
Farjo R, Skaggs J, Quiambao AB, Cooper MJ, Naash receptor gene transfection: a role of VEGF and possible
MI. Efficient non-viral ocular gene transfer with com- treatment for SRN in age-related macular degenera-
pacted DNA nanoparticles. PLoS One. 2006;1(1):e38. tion. Gene Ther. 2000;7(11):978–85.
Flotte TR, Afione SA, Zeitlin PL. Adeno-associated virus Igarashi T, Miyake K, Kato K, Watanabe A, Ishizaki M,
vector gene expression occurs in nondividing cells in Ohara K, Shimada T. Lentivirus-mediated expression
the absence of vector DNA integration. Am J Respir of angiostatin efficiently inhibits neovascularization in
Cell Mol Biol. 1994;11(5):517–21. a murine proliferative retinopathy model. Gene Ther.
Folliot S, Briot D, Conrath H, Provost N, Cherel Y, 2003;10(3):219–26.
Moullier P, Rolling F. Sustained tetracycline-regulated Igarashi T, Miyake K, Masuda I, Takahashi H, Shimada
transgene expression in vivo in rat retinal ganglion T. Adeno-associated vector (type 8)-mediated expres-
cells using a single type 2 adeno-associated viral vec- sion of soluble Flt-1 efficiently inhibits neovascular-
tor. J Gene Med. 2003;5(6):493–501. ization in a murine choroidal neovascularization
Forrester JV, Xu H. Good news-bad news: the Yin and model. Hum Gene Ther. 2010;21(5):631–7.
Yang of immune privilege in the eye. Front Immunol. Igarashi T, Miyake N, Fujimoto C, Yaguchi C, Iijima O,
2012;3:338. Shimada T, Takahashi H, Miyake K. Adeno-associated
Friedmann T. A brief history of gene therapy. Nat Genet. virus type 8 vector-mediated expression of siRNA tar-
1992;2(2):93–8. geting vascular endothelial growth factor efficiently
Gehlbach P, Demetriades AM, Yamamoto S, Deering T, inhibits neovascularization in a murine choroidal neo-
Duh EJ, Yang HS, Cingolani C, Lai H, Wei L, vascularization model. Mol Vis. 2014;20:488–96.
Campochiaro PA. Periocular injection of an adenovi- Jacobson SG, Cideciyan AV, Ratnakaram R, Heon E,
ral vector encoding pigment epithelium-derived factor Schwartz SB, Roman AJ, Peden MC, Aleman TS, Boye
inhibits choroidal neovascularization. Gene Ther. SL, Sumaroka A, Conlon TJ, Calcedo R, Pang J-J, Erger
2003;10(8):637–46. KE, Olivares MB, Mullins CL, Swider M, Kaushal S,
Greenberg KP, Geller SF, Schaffer DV, Flannery Feuer WJ, Iannaccone A, Fishman GA, Stone EM,
JG. Targeted transgene expression in Muller glia of Byrne BJ, Hauswirth WW. Gene therapy for Leber con-
normal and diseased retinas using lentiviral vectors. genital amaurosis caused by RPE65 mutations: safety
Invest Ophthalmol Vis Sci. 2007;48(4):1844–52. and efficacy in 15 children and adults followed up to 3
Hacein-Bey-Abina S, Von Kalle C, Schmidt M, McCormack years. Arch Ophthalmol. 2012;130(1):9–24.
MP, Wulffraat N, Leboulch P, Lim A, Osborne CS, Jiang Y, Wang H, Culp D, Yang Z, Fotheringham L,
Pawliuk R, Morillon E, Sorensen R, Forster A, Fraser P, Flannery J, Hammond S, Kafri T, Hartnett
Cohen JI, de Saint Basile G, Alexander I, Wintergerst ME. Targeting Müller cell-derived VEGF164 to
U, Frebourg T, Aurias A, Stoppa-Lyonnet D, Romana S, reduce intravitreal neovascularization in the rat model
Radford-Weiss I, Gross F, Valensi F, Delabesse E, of retinopathy of prematurity. Invest Ophthalmol Vis
Macintyre E, Sigaux F, Soulier J, Leiva LE, Wissler M, Sci. 2014;55(2):824–31.
Prinz C, Rabbitts TH, Le Deist F, Fischer A, Cavazzana- Kachi S, Esumi N, Zack DJ, Campochiaro PA. Sustained
Calvo M. LMO2-associated clonal T cell proliferation expression after nonviral ocular gene transfer using
in two patients after gene therapy for SCID-X1. Science. mammalian promoters. Gene Ther. 2006;13(9):
2003;302(5644):415–9. 798–804.
Hangai M, Kaneda Y, Tanihara H, Honda Y. In vivo gene Khani SC, Pawlyk BS, Bulgakov OV, Kasperek E, Young
transfer into the retina mediated by a novel liposome sys- JE, Adamian M, Sun X, Smith AJ, Ali RR, Li T. AAV-
tem. Invest Ophthalmol Vis Sci. 1996;37(13):2678–85. mediated expression targeting of rod and cone photo-
Haurigot V, Villacampa P, Ribera A, Bosch A, Ramos D, receptors with a human rhodopsin kinase promoter.
Ruberte J, Bosch F. Long-term retinal PEDF overex- Invest Ophthalmol Vis Sci. 2007;48(9):3954–61.
Kim HA, Mahato RI, Lee M. Hypoxia-specific gene patients with choroideremia: initial findings from a
expression for ischemic disease gene therapy. Adv phase 1/2 clinical trial. Lancet. 2014;383(9923):
Drug Deliv Rev. 2009;61(7–8):614–22. 1129–37.
Klimczak RR, Koerber JT, Dalkara D, Flannery JG, Maguire AM, Simonelli F, Pierce EA, Pugh EN,
Schaffer DV. A novel adeno-associated viral variant Mingozzi F, Bennicelli J, Banfi S, Marshall KA,
for efficient and selective intravitreal transduction of Testa F, Surace EM, Rossi S, Lyubarsky A, Arruda
rat Müller cells. PLoS One. 2009;4(10):e7467. VR, Konkle B, Stone E, Sun J, Jacobs J, Dell’Osso L,
Koerber JT, Klimczak R, Jang J-H, Dalkara D, Flannery Hertle R, Ma J, Redmond TM, Zhu X, Hauck B,
JG, Schaffer DV. Molecular evolution of adeno- Zelenaia O, Shindler KS, Maguire MG, Wright JF,
associated virus for enhanced glial gene delivery. Mol Volpe NJ, McDonnell JW, Auricchio A, High KA,
Ther. 2009;17(12):2088–95. Bennett J. Safety and efficacy of gene transfer for
Lai CC, Wu WC, Chen SL, Xiao X, Tsai TC, Huan SJ, Leber’s congenital amaurosis. N Engl J Med.
Chen TL, Tsai RJ, Tsao YP. Suppression of choroidal 2008;358(21):2240–8.
neovascularization by adeno-associated virus vector Mao Y, Kiss S, Boyer JL, Hackett NR, Qiu J, Carbone A,
expressing angiostatin. Invest Ophthalmol Vis Sci. Mezey JG, Kaminsky SM, D’Amico DJ, Crystal
2001;42(10):2401–7. RG. Persistent suppression of ocular neovascularization
Lai C-M, Shen W-Y, Brankov M, Lai YKY, Barnett NL, with intravitreal administration of AAVrh.10 coding for
Lee S-Y, Yeo IYS, Mathur R, Ho JES, Pineda P, bevacizumab. Hum Gene Ther. 2011;22(12):1525–35.
Barathi A, Ang C-L, Constable IJ, Rakoczy EP. Long- Masuda I, Matsuo T, Yasuda T, Matsuo N. Gene transfer
term evaluation of AAV-mediated sFlt-1 gene therapy with liposomes to the intraocular tissues by different
for ocular neovascularization in mice and monkeys. routes of administration. Invest Ophthalmol Vis Sci.
Mol Ther. 2005;12(4):659–68. 1996;37(9):1914–20.
Lai C-M, Estcourt MJ, Wikstrom M, Himbeck RP, Barnett McGee Sanftner LH, Rendahl KG, Quiroz D, Coyne M,
NL, Brankov M, Tee LBG, Dunlop SA, Degli-Esposti Ladner M, Manning WC, Flannery JG. Recombinant
MA, Rakoczy EP. rAAV.sFlt-1 gene therapy achieves AAV-mediated delivery of a tet-inducible reporter
lasting reversal of retinal neovascularization in the gene to the rat retina. Mol Ther. 2001;3(5 Pt 1):
absence of a strong immune response to the viral vec- 688–96.
tor. Invest Ophthalmol Vis Sci. 2009;50(9):4279–87. Miller DG, Trobridge GD, Petek LM, Jacobs MA, Kaul
Lamartina S, Cimino M, Roscilli G, Dammassa E, R, Russell DW. Large-scale analysis of
Lazzaro D, Rota R, Ciliberto G, Toniatti C. Helper- adeno-associated virus vector integration sites in nor-
dependent adenovirus for the gene therapy of prolif- mal human cells. J Virol. 2005;79(17):11434–42.
erative retinopathies: stable gene transfer, regulated Mori K, Duh E, Gehlbach P, Ando A, Takahashi K,
gene expression and therapeutic efficacy. J Gene Med. Pearlman J, Mori K, Yang HS, Zack DJ, Ettyreddy D,
2007;9(10):862–74. Brough DE, Wei LL, Campochiaro PA. Pigment
Lange CA, Stavrakas P, Luhmann UF, de Silva DJ, Ali RR, epithelium-derived factor inhibits retinal and choroi-
Gregor ZJ, Bainbridge JW. Intraocular oxygen distribu- dal neovascularization. J Cell Physiol. 2001;188(2):
tion in advanced proliferative diabetic retinopathy. Am 253–63.
J Ophthalmol. 2011;152(1879–1891):406–12. Mori K, Gehlbach P, Yamamoto S, Duh E, Zack DJ, Li Q,
Lebherz C, Auricchio A, Maguire AM, Rivera VM, Tang Berns KI, Raisler BJ, Hauswirth WW, Campochiaro
W, Grant RL, Clackson T, Bennett J, Wilson JM. Long- PA. AAV-mediated gene transfer of pigment epithelium-
term inducible gene expression in the eye via adeno- derived factor inhibits choroidal neovascularization.
associated virus gene transfer in nonhuman primates. Invest Ophthalmol Vis Sci. 2002;43(6):1994–2000.
Hum Gene Ther. 2005;16(2):178–86. Nakai H, Montini E, Fuess S, Storm TA, Grompe M, Kay
Lukason M, DuFresne E, Rubin H, Pechan P, Li Q, Kim I, MA. AAV serotype 2 vectors preferentially integrate into
Kiss S, Flaxel C, Collins M, Miller J, Hauswirth W, active genes in mice. Nat Genet. 2003;34(3):297–302.
Maclachlan T, Wadsworth S, Scaria A. Inhibition of Nakai H, Wu X, Fuess S, Storm TA, Munroe D, Montini
choroidal neovascularization in a nonhuman primate E, Burgess SM, Grompe M, Kay MA. Large-scale
model by intravitreal administration of an AAV2 vec- molecular characterization of adeno-associated virus
tor expressing a novel anti-VEGF molecule. Mol Ther. vector integration in mouse liver. J Virol.
2011;19(2):260–5. 2005;79(6):3606–14.
Maclachlan TK, Lukason M, Collins M, Munger R, Nishijima K, Ng YS, Zhong L, Bradley J, Schubert W, Jo
Isenberger E, Rogers C, Malatos S, Dufresne E, N, Akita J, Samuelsson SJ, Robinson GS, Adamis AP,
Morris J, Calcedo R, Veres G, Scaria A, Andrews L, Shima DT. Vascular endothelial growth factor-A is a
Wadsworth S. Preclinical safety evaluation of AAV2- survival factor for retinal neurons and a critical neuro-
sFLT01—a gene therapy for age-related macular protectant during the adaptive response to ischemic
degeneration. Mol Ther. 2011;19(2):326–34. injury. Am J Pathol. 2007;171(0002–9440):53–67.
MacLaren RE, Groppe M, Barnard AR, Cottriall CL, O’Reilly MS, Boehm T, Shing Y, Fukai N, Vasios G, Lane
Tolmachova T, Seymour L, Clark KR, During MJ, WS, Flynn E, Birkhead JR, Olsen BR, Folkman
Cremers FPM, Black GCM, Lotery AJ, Downes SM, J. Endostatin: an endogenous inhibitor of angiogenesis
Webster AR, Seabra MC. Retinal gene therapy in and tumor growth. Cell. 1997;88(2):277–85.
Pechan P, Rubin H, Lukason M, Ardinger J, DuFresne E, soluble VEGF in the maintenance of the choriocapil-
Hauswirth WW, Wadsworth SC, Scaria A. Novel anti- laris. Proc Natl Acad Sci U S A.
VEGF chimeric molecules delivered by AAV vectors 2009;106(1091–6490):18751–6.
for inhibition of retinal neovascularization. Gene Ther. Saishin Y, Silva RL, Saishin Y, Kachi S, Aslam S, Gong
2009;16(1):10–6. YY, Lai H, Carrion M, Harris B, Hamilton M, Wei L,
Pellissier LP, Hoek RM, Vos RM, Aartsen WM, Klimczak Campochiaro PA. Periocular gene transfer of pigment
RR, Hoyng SA, Flannery JG, Wijnholds J. Specific epithelium-derived factor inhibits choroidal neovascu-
tools for targeting and expression in Müller glial cells. larization in a human-sized eye. Hum Gene Ther.
Mol Ther Methods Clin Dev. 2014;1:14009. 2005;16(4):473–8.
Ponnazhagan S, Erikson D, Kearns WG, Zhou SZ, Salam A, Mathew R, Sivaprasad S. Treatment of prolif-
Nahreini P, Wang XS, Srivastava A. Lack of site- erative diabetic retinopathy with anti-VEGF agents.
specific integration of the recombinant adeno- Acta Ophthalmol. 2001;89(1755–3768):405–11.
associated virus 2 genomes in human cells. Hum Gene Schnepp BC, Clark KR, Klemanski DL, Pacak CA,
Ther. 1997;8(3):275–84. Johnson PR. Genetic fate of recombinant adeno-
Prentice HM, Biswal MR, Dorey CK, Blanks JC. Hypoxia- associated virus vector genomes in muscle. J Virol.
regulated retinal glial cell-specific promoter for poten- 2003;77(6):3495–504.
tial gene therapy in disease. Invest Ophthalmol Vis Shen J, Yang X, Xiao W-H, Hackett SF, Sato Y,
Sci. 2011;52(12):8562–70. Campochiaro PA. Vasohibin is up-regulated by
Raisler BJ, Berns KI, Grant MB, Beliaev D, Hauswirth VEGF in the retina and suppresses VEGF receptor 2
WW. Adeno-associated virus type-2 expression of and retinal neovascularization. FASEB J. 2006;20(6):
pigmented epithelium-derived factor or Kringles 1-3 723–5.
of angiostatin reduce retinal neovascularization. Proc Simonelli F, Maguire AM, Testa F, Pierce EA, Mingozzi
Natl Acad Sci U S A. 2002;99(13):8909–14. F, Bennicelli JL, Rossi S, Marshall K, Banfi S, Surace
Ramírez M, Wu Z, Moreno-Carranza B, Jeziorski MC, EM, Sun J, Redmond TM, Zhu X, Shindler KS, Ying
Arnold E, Díaz-Lezama N, Martínez de la Escalera G, G-S, Ziviello C, Acerra C, Wright JF, McDonnell JW,
Colosi P, Clapp C. Vasoinhibin gene transfer by High KA, Bennett J, Auricchio A. Gene therapy for
adeno-associated virus type 2 protects against VEGF- Leber’s congenital amaurosis is safe and effective
and diabetes-induced retinal vasopermeability. Invest through 1.5 years after vector administration. Mol
Ophthalmol Vis Sci. 2011;52(12):8944–50. Ther. 2010;18(3):643–50.
Rasmussen H, Chu KW, Campochiaro P, Gehlbach PL, Spranger J, Hammes HP, Preissner KT, Schatz H,
Haller JA, Handa JT, Nguyen QD, Sung JU. Clinical Pfeiffer AF. Release of the angiogenesis inhibitor
protocol. An open-label, phase I, single administra- angiostatin in patients with proliferative diabetic reti-
tion, dose-escalation study of ADGVPEDF.11D nopathy: association with retinal photocoagulation.
(ADPEDF) in neovascular age-related macular Diabetologia. 2000;43(0012–186X):1404–7.
degeneration (AMD). Hum Gene Ther. 2001; Steele FR, Chader GJ, Johnson LV, Tombran-Tink
12(16):2029–32. J. Pigment epithelium-derived factor: neurotrophic
Reichel MB, Ali RR, Thrasher AJ, Hunt DM, Bhattacharya activity and identification as a member of the serine
SS, Baker D. Immune responses limit adenovirally protease inhibitor gene family. Proc Natl Acad Sci U S
mediated gene expression in the adult mouse eye. A. 1993;90(4):1526–30.
Gene Ther. 1998;5(8):1038–46. Stefansson E, Geirsdottir A, Sigurdsson H. Metabolic
Reichel MB, Hudde T, Ali RR, Wiedemann P. Gene physiology in age related macular degeneration. Prog
transfer in ophthalmology. Ophthalmologe. 1999; Retin Eye Res. 2011;30(1):72–80.
96(9):570–7. Takahashi T, Nakamura T, Hayashi A, Kamei M,
Rofagha S, Bhisitkul RB, Boyer DS, Sadda SR, Zhang K, Nakabayashi M, Okada AA, Tomita N, Kaneda Y,
SEVEN-UP Study Group. Seven-year outcomes in Tano Y. Inhibition of experimental choroidal neovas-
ranibizumab-treated patients in ANCHOR, MARINA, cularization by overexpression of tissue inhibitor of
and HORIZON: a multicenter cohort study metalloproteinases-3 in retinal pigment epithelium
(SEVEN-UP). Ophthalmology. 2013;120(11):2292–9. cells. Am J Ophthalmol. 2000;130(6):774–81.
Rota R, Riccioni T, Zaccarini M, Lamartina S, Gallo AD, Takahashi K, Saishin Y, Saishin Y, Silva RL, Oshima Y,
Fusco A, Kovesdi I, Balestrazzi E, Abeni DC, Ali RR, Oshima S, Melia M, Paszkiet B, Zerby D, Kadan MJ,
Capogrossi MC. Marked inhibition of retinal neovas- Liau G, Kaleko M, Connelly S, Luo T, Campochiaro
cularization in rats following soluble-flt-1 gene trans- PA. Intraocular expression of endostatin reduces
fer. J Gene Med. 2004;6(9):992–1002. VEGF-induced retinal vascular permeability, neovas-
Saint-Geniez M, Maharaj ASR, Walshe TE, Tucker BA, cularization, and retinal detachment. FASEB
Sekiyama E, Kurihara T, Darland DC, Young MJ, J. 2003;17(8):896–8.
D’Amore PA. Endogenous VEGF is required for Testa F, Maguire AM, Rossi S, Pierce EA, Melillo P,
visual function: evidence for a survival role on Müller Marshall K, Banfi S, Surace EM, Sun J, Acerra C,
cells and photoreceptors. PLoS One. 2008;3(11):e3554. Wright JF, Wellman J, High KA, Auricchio A, Bennett
Saint-Geniez M, Kurihara T, Sekiyama E, Maldonado J, Simonelli F. Three-year follow-up after unilateral
AE, D’Amore PA. An essential role for RPE-derived subretinal delivery of adeno-associated virus in
patients with Leber congenital Amaurosis type 2. of retinopathy of prematurity. Am J Pathol.

Ophthalmology. 2013;120(6):1283–91. 2013;183(3):964–74.
Ueno S, Pease ME, Wersinger DMB, Masuda T, Vinores Watkins WM, McCollum GW, Savage SR, Capozzi ME,
SA, Licht T, Zack DJ, Quigley H, Keshet E, Penn JS, Morrison DG. Hypoxia-induced expression
Campochiaro PA. Prolonged blockade of VEGF fam- of VEGF splice variants and protein in four retinal cell
ily members does not cause identifiable damage to types. Exp Eye Res. 2013;116:240–6.
retinal neurons or vessels. J Cell Physiol. Willett K, Bennett J. Immunology of AAV-mediated gene
2008;217(1):13–22. transfer in the eye. Front Immunol. 2013;4:261.
Wakusawa R, Abe T, Sato H, Sonoda H, Sato M, Mitsuda Wu Z, Yang H, Colosi P. Effect of genome size on AAV
Y, Takakura T, Fukushima T, Onami H, Nagai N, vector packaging. Mol Ther. 2010;18(1):80–6.
Ishikawa Y, Nishida K, Sato Y. Suppression of choroi- Zhang SX, Wang JJ, Gao G, Parke K, Ma J. Pigment
dal neovascularization by vasohibin-1, a vascular epithelium-derived factor downregulates vascular
endothelium-derived angiogenic inhibitor. Invest endothelial growth factor (VEGF) expression and
Ophthalmol Vis Sci. 2011;52(6):3272–80. inhibits VEGF-VEGF receptor 2 binding in diabetic
Wang H, Smith GW, Yang Z, Jiang Y, McCloskey M, retinopathy. J Mol Endocrinol. 2006;37(1):1–12.
Greenberg K, Geisen P, Culp WD, Flannery J, Kafri T, Zhang X, Das SK, Passi SF, Uehara H, Bohner A, Chen
Hammond S, Hartnett ME. Short hairpin RNA- M, Tiem M, Archer B, Ambati BK. AAV2 delivery of
mediated knockdown of VEGFA in Müller cells Flt23k intraceptors inhibits murine choroidal neovas-
reduces intravitreal neovascularization in a rat model cularization. Mol Ther. 2015;123(2):226–34.
Index
A study formulations, 72
Adenoviruses containing sFLt1 (Ad-sFlt1), 181 treatment formulations, 72
Aflibercept (ABC) Aggressive posterior ROP (AP-ROP), 23
anatomical outcomes, 47 Anterior zone II, 23
safety, 50 Anti-angiogenic gene therapy
visual acuity outcomes, 34 adenoviral vectors, 174
Age-related eye disease study (AREDS) blood–retina barrier, 173
CNV/photographic documentation, 68 clinical translation
dosages, 68 immune responses, 180
Early Treatment Diabetic Retinopathy Study protocol, 69 neuroprotection, 181
formulation, 68 oncogenesis, 180–181
lutein and zeaxanthin, 68 potential side effects, 180–181
ocular findings and participants number, 68 vasoprotection, 181
outcomes fundamental concepts, 174
antioxidant plus zinc formulation, 70 lentiviral vectors, 174
color fundus photograph, 70 neovascular eye disease
direct costs, 71 angiostatin, 179
EDTRS standardized protocol, 70 CD59, 180
impact of, 71 endostatin, 179
primary outcome measure, 68 PEDF, 176–179
risk factors, 70 tissue inhibitor metalloproteinase 3, 180
secondary outcome measure, 69 VEGF-inhibiting molecules, 175–176
pigmented rat model, 68 neovascularisation, 173
randomized double-masked placebo-controlled trial, 67 optimal vector system
secondary outcomes, 68 cell-specific vector systems, 181–182
Age-related eye disease study 2 (AREDS2) non-viral gene therapy, 182–183
adverse effect, 73 transgene expression, regulation of, 182
beta-carotene, risk of, 72 rAAV vectors, 174
clinical sites, 71 vector suspensions, 173
geographic atrophy, 72 viral and non-viral vector systems, 174
mechanism, 71 Antiangiogenic treatment
observational studies, 71 PRP
omega LCPUFAs, 71 choroidal thickness, 116
omega 3 LCPUFA, 71 clinical application and efficacy, 116
primary analysis, 72 focal retino-choroidal atrophy, 117
primary outcome, 72 neovascular glaucoma, 118, 119
safety and efficacy, 71 tractional retinal detachment, 118
secondary analysis, 72 vitreomacular traction, 117
standardized fundus photographs, 72 vitreous hemorrhage, 116

Essentials in Ophthalmology, DOI 10.1007/978-3-319-24097-8
190 Index
Anti-vascular endothelial growth factor (VEGF) morphological characterization and quantification

treatment blood and lymphatic vessels, 164–165
anatomical outcomes corneal vessels, 164
ABC, 47 neovascularization, 159
BVZ, 47 therapeutic inhibition
clinical factors, 55 anti-VEGF therapy, 166–167
RBZ, 43–47 corticosteroids, 165–166
behavioral factors, 56 fine needle diathermy, 167–168
clinical factors IRS-1, 167
anatomical outcomes, 55 PDT, with verteporfin, 168
injections administration, 56 transplant rejection, 161–162
vision outcomes, 54–55 traumatic diseases, 160
genetic factors visual acuity, 161
APOE gene, 53 Corneal transplant rejection, 161–162
ARMS2/HTRA1, 52 Corticosteroids, 165–166
CFH gene, 52 CRYO-ROP study, 23
complement system-related genes, 53
genome-wide association study, 54
IL-23 R gene, 53 D
PLA2G12A gene, 53 Diabetic macular edema (DME)
VEGF gene, 53 clinical improvement, 105
legal blindness, incidence, 47–48 combination treatment, 115
safety definition, 96
ABC, 50 focal/grid laser photocoagulation
intravitreal BVZ, 50–51 disadvantages, 107
intravitreal RBZ, 48–50 ETDRS, 106
therapy, 166–167 monitoring noncentral edema, 107
visual acuity outcomes (see Phase III clinical trials) navigated/targeted laser, 107, 108
Argon laser photocoagulation, 84 RPE, 106
selective laser therapy, 107
subthreshold micropulse laser, 107
B intravitreal anti-VEGF therapy, 113–115
BEAT-ROP study, 24, 26 intravitreal steroids
Bevacizumab (BVZ) dexamethasone drug delivery system, 110
anatomical outcomes, 47 fluocinolone, 112, 113
MacTel type 2, 84 triamcinolone acetate, 108, 110
safety, 50–51 Diabetic retinopathy (DR)
visual acuity outcomes, 41 DME (see Diabetic macular edema (DME))
Blood–retinal barrier (BRB) ETDRS severity score, 98
alteration of, 13 ETDRS studies, 91
extracellular portions, 13 genetic predisposition, 90
exudative age-related macular degeneration, 14 Gutenberg Health study, 90
functional, 13 incidence, 90
inner BRB, 13 LALES, 90
ocular vascular leakage, 15 macular disease
outer BRB, 13 CSME, 97
tight and adherens junctions, 13 definition, 96
transmembrane proteins, 13 foveal ischemia, 97
OCT, 96
vitreomacular traction, 98
C NPDR
Choroidal neovascularization (CNV), 68 FLA, 91
Clinically significant macular edema (CSME), 97 hard exudates, 93
Corneal lymphangiogenesis, 162–163 hemorrhages, 91
Corneal neovascular diseases microaneurysms, 91
degenerative, 160 pathophysiology, 100–101
infectious, 160 PDR, 116
inflammatory syndrome, 160 antiangiogenic treatment (see Antiangiogenic
lymphangiogenesis, 162–163 treatment)
(lymph)angiogenic privilege, 159, 160 fibrovascular proliferations, 96
Index 191
inflammation and chronic disruption, 94 I

neovascularization, 95 Insulin-like growth factor-1 (IGF-1), 6, 24
peripheral lesions, 95 Insulin receptor substrate-1 (IRS-1), 167
retinal nonperfusion, 94 Interleukin (IL) 23 R gene, 53
with type 1 diabetes, 94 International Diabetes Federation (IDF), 89
VEGF, 95 Intravitreal bevacizumab (IVB) injection (IVB), 152
vitreous hemorrhage, 95 Intravitreal therapy
prevention and health costs, 91 corticosteroids, 132–133
screening, 98–99 dexamethasone, 132–133
systemic risk factors fluocinolone, 133
blood pressure, 103 triamcinolone acetonide, 133
blood sugar, 102
disease duration, 102
dyslipidemia, 103, 104 L
gestational diabetes, 102 Laser photocoagulation, 23
mortality, 105 Los Angeles Latino Eye Study (LALES), 90
nephropathy, 104
pregnancy, 105
systemic treatment, 105 M
type 1 diabetes, 101 Macular telangiectasia (MacTel) type 2
type 2 diabetes, 101 causes, 79
treatment, 106 clinical observations, 85–86
type 1 diabetes, 89 definition, 79
visual loss, 89 fundus autofluorescence, 80
Dyslipidemia, 103, 104 genetic component, 79
morphological changes, 79
neovascular membranes, 79
E non-proliferative stages, 80–84
Early Treatment Diabetic Retinopathy Study (ETDRS) OCT, 80
protocol, 34, 40, 69, 98, 99, 106, 109 paracentral scotomata, 80
Erythropoietin (Epo), 6 pathophysiology, 79
proliferative stages
argon laser photocoagulation, 84
F intravitreal application, 84
Familial exudative vitreoretinopathy (FEVR), 12–13 PDT, 84
Fine needle diathermy, 167 subretinal surgery, 85
Fluorescein angiography (FLA), 91 therapeutic approaches, 84
vision loss, 84
SD-OCT, 79
G stages, 79
Genetic factors therapeutic effects, 80
APOE gene, 53 vascular alterations, 79
ARMS2/HTRA1, 52 Microglia, 10
CFH gene, 52
complement system-related genes, 53
genome-wide association study, 54 N
IL-23 R gene, 53 Neonatal Fc receptor (FcRn), 141
PLA2G12A gene, 53 Neovascular eye disease
VEGF gene, 53 angiostatin, 179
Geographic atrophy, 72 CD59, 180
Grid laser photocoagulation, 106–107 endostatin, 179
Group 12 secretory phospholipase A2 (PLA2G12A), 53 PEDF, 176–179
Gutenberg Health study, 90 tissue inhibitor metalloproteinase 3, 180
VEGF-inhibiting molecules, 175–176
Neovascular glaucoma
H adjuvant treatment, 155
HbA1c memory effect, 102 anti-VEGF drugs
Hyperoxic phase, 21 bevacizumab, 151
192 Index
Neovascular glaucoma (cont.) vitreous liquefaction, 145

CATT 2 study, 155 Phase III clinical trials
in NVG, 152 ABC, 34
ranibizumab, 154, 155 BVZ, 41
aqueous tube shunts, 151 clinical practice studies, 41, 42
bevacizumab, 154 Danish longitudinal study, 42
combination treatment, 153 nvAMD, 41, 42
diode laser cyclodestruction, 150 OCT-guided Treat & Extend treatment protocol, 42
filtering surgery, 150 RBZ, 34
IVB, 152, 153 Photodynamic therapy (PDT), 84, 168
laser cyclophotocoagulation, 150 Pigment epithelium-derived factor (PEDF), 176–179
neovascularization and open anterior chamber Post-keratoplasty neovascularization, 162
angle, 153 Proliferative diabetic retinopathy (PDR)
partial destruction, 150 antiangiogenic treatment (see Antiangiogenic
pathogenesis, 149 treatment)
PRP, 149 fibrovascular proliferations, 96
Nephropathy, 104 inflammation and chronic disruption, 94
Non-proliferative diabetic retinopathy (NPDR) neovascularization, 95
FLA, 91 panretinal photocoagulation, 116–117
hard exudates, 93 peripheral lesions, 95
hemorrhages, 91 retinal nonperfusion, 94
microaneurysms, 91 with type 1 diabetes, 94
prognostic value, 91 VEGF, 95
retinal capillaries, 93 vitreous hemorrhage, 95
Norrie disease, 12 Proliferative MacTel type 2, 79
O R
Ocular coherence tomography (OCT), 96 Ranibizumab (RBZ)
Omega-3 long chain poly-unsaturated fatty acids anatomical outcomes, 43–47
(LCPUFAs), 71 CARE-ROP study, 25
Oncogenesis, 180–181 safety, 48–50
Ophthalmic artery, 2 visual acuity outcomes, 34
Optical coherence tomography (OCT), 80 Recombinant adeno-associated virus (rAAV) vectors, 174
Optimal vector system Retinal capillary microaneurysms, 91
cell-specific vector systems, 181–182 Retinal vascular development
non-viral gene therapy, 182–183 angiogenesis, 3–4
transgene expression, regulation, 182 blood vessel networks, anatomy
choroidal vessels, 2–3
hyaloidal vessels, 3
P retinal vessels, 1–2
Panretinal photocoagulation (PRP) BRB
choroidal thickness, 116 alteration of, 13
clinical application and efficacy, 116 extracellular portions, 13
focal retino-choroidal atrophy, 117 exudative age-related macular degeneration, 14
neovascular glaucoma, 118, 119, 149 inner BRB, 13
tractional retinal detachment, 118 ocular vascular leakage, 15
vitreomacular traction, 117 outer BRB, 13
Pharmacokinetics tight and adherens junctions, 13
aflibercept, 139, 141 transmembrane proteins, 13
bevacizumab, 139, 140 choroidal development, 12
clinical relevance, 141–143 dynamics of, 6–9
FcRn, 141 hyaloidal regression, 12–13
pegaptanib, 139, 140 neurovascular cross talk
ranibizumab, 139, 140 central neurons, 9
refraction differences, 145 immune cells, 10, 11
silicone oil-filled eyes, 146 neuronal energy metabolism, 10
systemic, 143–144 photoreceptors, 9, 10
vitrectomy, 145 remodeling, 11–12
Index 193
retinal ganglion cells, 9 cryotherapy, 23

oxygen, lack of, 4–5 IGF-1, 24
retinal neuroglia, 6 laser photocoagulation, 23
temporal and spatial development, 4 laser therapy, 23
VEGF, 5–6 zone I, 23
Retinal vein occlusion (RVO) zone III, 23
clinical characteristics, 131
etiology, 131
implications, 136 S
intravitreal therapy Sirtuin-1 (Sirt-1), 10
corticosteroids, 132–133 Sodium/glucose cotransporter 2 (SGLT2), 103
dexamethasone, 132–133 Spectral-domain optical coherence tomography
fluocinolone, 133 (SD-OCT), 79
triamcinolone acetonide, 133 Subretinal surgery, 85
prevalence, 131 Succinate-GPR91 axis, 10
risk factors, 132
VEGF-inhibitors
aflibercept, 135–136 T
bevacizumab, 135 Triamcinolone acetonide, 133
pegaptanib, 135
ranibizumab, 134–135
Retinopathy of prematurity (ROP) V
AP-ROP, 23 Vascular endothelial growth factor (VEGF)-inhibitors
central zone II, 23 Aflibercept, 135–136
history, 21 bevacizumab, 135
peripheral zone II, 23 MacTel type 2, 80
phase I, 21 pegaptanib, 135
phase II, 22 Vascularization process, 3
physiologic growth, 21 Vascular layers, 4
stage 1, 22 Vasculogenesis, 4
stage 2, 22
stage 3, 22
stage 3+, 22 W
stage 4 and 5, 23 Wisconsin Epidemiologic Study of Diabetic Retinopathy
treatment (WESDR), 90
antiangiogenic treatment approach, 24 Wnt/β-catenin signaling pathway, 13
anti-VEGF treatment, 24, 25
BEAT-ROP study, 24, 25
CARE-ROP study, 25, 26 X
cryoablation, 23 X-linked severe combined immunodeficiency
CRYO-ROP study, 23 (X-SCID), 180

Anti-Angiogenic Therapy in Ophthalmology

Uploaded by

Copyright:

Available Formats

Anti-Angiogenic Therapy in Ophthalmology

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

Anti-Angiogenic Therapy in Ophthalmology

Uploaded by

Copyright:

Available Formats

Essentials in Ophthalmology

Series Editor: Arun D. Singh

Andreas Stahl Editor

More information about this series at http://www.springer.com/series/5332

ISSN 1612-3212 ISSN 2196-890X (electronic)

Library of Congress Control Number: 2016937532

Springer Cham Heidelberg New York Dordrecht London

Printed on acid-free paper

Springer International Publishing AG Switzerland is part of Springer Science+Business Media

In 2006, intraocular anti-VEGF therapy for exudative age-related macular

Freiburg, Germany Andreas Stahl, MD

1 Retinal Vascular Development ..................................................... 1

Index ....................................................................................................... 189

Hansjürgen T. Agostini, MD Eye Center, Albert-Ludwigs-University

Ann Hellström, MD, PhD Department of Ophthalmology, Institute of

rior ciliary arteries, serve the optic nerve head,

To aid in understanding retinal vascular develop-

© Springer International Publishing Switzerland 2016 1

Blood vessels are generally composed of sev-

In addition to VEGF, erythropoietin (Epo) is 1.2.3.3 Retinal Neuroglia Serves

the eye. Identifying ways to modulate BRB both

25 weeks gestational age) or the ones that are

© Springer International Publishing Switzerland 2016 21

subgroup of aggressive posterior ROP (AP-ROP)

2.2 Treatment of ROP

Published in 1990, the CRYO-ROP study defined

ranibizumab has been developed explicitly for 2.3 Summary

2012;130(8):1000–6. doi:10.1001/archophthal- McNamara JA, Tasman W, Brown GC, Federman

Dujon Fuzzard, Robyn H. Guymer,

BVZ Bevacizumab (Avastin®)

© Springer International Publishing Switzerland 2016 31

FA Fluorescein angiography Ranibizumab—phase III clini-

high-temperature requirement A-1 (HTRA1), initiation of treatment (Menghini et al. 2012).

3.3.1.1 Complement Factor H 3.3.1.2 Age-Related Maculopathy

3.3.1.3 Vascular Endothelial Growth 3.3.1.5 Complement System-Related

retina at baseline are predictive of poor VA out- 3.3.2.2 Anatomical Outcomes

atrophy and/or the development of GA in a

Finger RP, Wiedemann P, Blumhagen F, Pohl K, Holz Ophthalmol. 2008;246(9):1229–34. doi:10.1007/

of anti-VEGF injections in neovascular age-related tion: an updated meta-analysis of randomised clinical

neovascular age-related macular degeneration. Int Menghini M, Kloeckener-Gruissem B, Fleischhauer J,

David J. Valent and Emily Y. Chew

With the retina’s high concentration of polyunsat-

© Springer International Publishing Switzerland 2016 67

ment of intermediate or advanced AMD. However, References

(“pigment-plaques”), probably due to intrareti-

© Springer International Publishing Switzerland 2016 79

suggesting that early intervention seems to be

VEGF-inhibition over one year were unfavorable, References

Nevertheless, blindness is not only the fear No.

F. Ziemssen (*) Type 1 diabetes shows a heterogeneous distribu-

© Springer International Publishing Switzerland 2016 89

Fig. 6.3 Clinical image of microaneurysms (right: arrow in fluorescein angiogram)

Table 6.1 Rational for the performance of FLA

Fig. 6.6 Subtle NVD can be delineated by fluorescein angiography

retina and macular involvement are the most

6.2.3 Macular Disease

Macular edema (DME) is defined as retinal thick-

Fig. 6.7 OCT imaging allows classifying for different

Fig. 6.8 Definition of clinically significant macular edema (CSME)

RH IRMA RH RH present severe ³ 1 quadrant severe ³ 2 quadrants severe ³ 4 quadrants