Dugernier et al. Ann.

Intensive Care (2016) 6:73

DOI 10.1186/s13613-016-0169-x

RESEARCH Open Access

Aerosol delivery with two ventilation

modes during mechanical ventilation: a
randomized study
Jonathan Dugernier1,3*, Gregory Reychler2,3, Xavier Wittebole4, Jean Roeseler4, Virginie Depoortere5,
Thierry Sottiaux6, Jean‑Bernard Michotte7, Rita Vanbever8, Thierry Dugernier9, Pierre Goffette10,
Marie‑Agnes Docquier11, Christian Raftopoulos12, Philippe Hantson4, François Jamar5
and Pierre‑François Laterre4

Abstract 
Background:  Volume-controlled ventilation has been suggested to optimize lung deposition during nebulization
although promoting spontaneous ventilation is targeted to avoid ventilator-induced diaphragmatic dysfunction.
Comparing topographic aerosol lung deposition during volume-controlled ventilation and spontaneous ventilation
in pressure support has never been performed. The aim of this study was to compare lung deposition of a radiola‑
beled aerosol generated with a vibrating-mesh nebulizer during invasive mechanical ventilation, with two modes:
pressure support ventilation and volume-controlled ventilation.
Methods:  Seventeen postoperative neurosurgery patients without pulmonary disease were randomly ventilated in

(2 mCi/3 mL) was administrated using a vibrating-mesh nebulizer (Aerogen Solo®, provided by Aerogen Ltd, Galway,
pressure support or volume-controlled ventilation. Diethylenetriaminepentaacetic acid labeled with technetium-99m

Ireland) connected to the endotracheal tube. Pulmonary and extrapulmonary particles deposition was analyzed using
planar scintigraphy.
Results:  Lung deposition was 10.5 ± 3.0 and 15.1 ± 5.0 % of the nominal dose during pressure support and volume-
controlled ventilation, respectively (p < 0.05). Higher endotracheal tube and tracheal deposition was observed during
pressure support ventilation (27.4 ± 6.6 vs. 20.7 ± 6.0 %, p < 0.05). A similar penetration index was observed for the
right (p = 0.210) and the left lung (p = 0.211) with both ventilation modes. A high intersubject variability of lung
deposition was observed with both modes regarding lung doses, aerosol penetration and distribution between the
right and the left lung.
Conclusions:  In the specific conditions of the study, volume-controlled ventilation was associated with higher lung
deposition of nebulized particles as compared to pressure support ventilation. The clinical benefit of this effect war‑
rants further studies.
Clinical trial registration NCT01879488
Keywords:  Aerosol delivery, Ventilation mode, Invasive mechanical ventilation, Vibrating-mesh nebulizer

*Correspondence: Jonathan.dugernier@uclouvain.be
3
Institut de Recherche Expérimentale et Clinique (IREC), Pneumologie,
ORL & Dermatologie, Université catholique de Louvain, 1200 Brussels,
Belgium
Full list of author information is available at the end of the article

© 2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made.
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 2 of 10

Background aged 18  years and older, admitted for brain neurosur-
Combining aerosol therapy with invasive mechanical ven- gery and the availability of krypton gas (81mKr). Patients
tilation is common in the intensive care unit (ICU) [1]. were included if the FEV1 to FVC ratio was superior to
In vitro and experimental studies have been performed 70  %. Exclusion criteria were spine neurosurgery, his-
to optimize aerosol lung deposition [2, 3]. Many factors tory of cardiovascular and pulmonary disease, extubation
related to the device, the artificial airways and the respira- immediately after surgery, the allocated ventilation mode
tory pattern inherent to invasive mechanical ventilation modification during the nebulization and inaccurate
influence lung deposition [4]. Most of them should be con- quantification of aerosol deposition (artifacts or overlap
trolled to deliver high doses of concentration-dependent of tracheal and pulmonary deposition).
drugs such as antibiotics. Vibrating-mesh nebulizers ensure The study protocol was approved by the Institutional
currently the best drug output, ergonomy and practical Medical Ethics Committee (B403201317342). Written
use in the ICU [5, 6]. Applying a constant inspiratory flow informed consent was obtained from all participants.
pattern with a slow inspiratory flow rate improved aerosol Preoperative pulmonary function testing was performed
delivery as compared to a decelerating inspiratory flow pat- in the Neurosurgery Department according to the Amer-
tern with a higher peak inspiratory flow rate [7]. This man- ican Thoracic Society guidelines using a MicroLoop
dates volume-controlled ventilation modality and a deep spirometer (CareFusion, San Diego, CA) [21]. The two
depression of the respiratory drive with sedatives [8]. Con- procedures were randomized by a computer-generated
versely, reducing sedation is promoted in ICU in order to random number list. Generation of the random alloca-
preserve the diaphragmatic function through a spontane- tion sequence, enrollment of the patients and allocation
ous breathing activity and hence to shorten the duration of to the assigned ventilation mode were performed by the
mechanical ventilation [9, 10]. The decelerating inspiratory same clinician author (JD). The double-blind design was
flow pattern and the spontaneous breathing activity of the related to the patients and the data analysis. The nebu-
patient (uncontrolled respiratory rate, inspiratory flow rate) lizer procedure and image acquisition were performed in
related to the pressure support modality may affect aerosol the same room in the Nuclear Medicine Department.
delivery to the lungs. The comparison of pressure support
ventilation and volume-controlled ventilation regarding Invasive mechanical ventilation and nebulization
aerosol administration has not been investigated. procedure
Recent guidelines focused on the standardization of Patients were randomly assigned to pressure support
radionuclide imaging methods validated to assess lung (PSV) or volume control (VCV) mode. Patients were ven-
deposition [11]. Imaging techniques are able to highlight tilated using an ICU turbine-driven ventilator (Bellavista
the aerosol distribution into the lungs [12]. Twenty years 1000e, Imtmedical, Buchs, Switzerland) (Fig. 1). Aerosol

mesh nebulizer (Aerogen Solo®, provided by Aerogen

ago, scintigraphic studies reported a low aerosol lung particles were generated continuously by a vibrating-
deposition during invasive mechanical ventilation with a
jet and ultrasonic nebulizer [13–16]. Since then, pharma- Ltd, Galway, Ireland). The nebulizer was placed between
cokinetic studies showed high concentrations of inhaled the endotracheal tube and the catheter mount because of
antibiotics through the lungs after dissection of sub- the presence of a proximal flow sensor. The nebulizer res-
pleural lung segments in piglets [17] and after endotra- ervoir was filled with technetium-99m-labeled diethylen-
cheal suctioning [18] or bronchoalveolar lavage [19, 20] etriaminepentaacetic acid (99mTc-DTPA, 2 mCi/3 mL).
in patients suffering from ventilator-associated pneumo- The sedative drug (propofol) was titrated in both
nia. However, topographic assessment of lung deposition groups to avoid patient’s movements, cough and patient-
by imaging technique during invasive mechanical ventila- ventilator asynchrony during the procedure, while keep-
tion has not been investigated. ing an exclusive spontaneous breathing activity during
The aim of this study was to compare in vivo the impact PSV. Patients in VCV did not trigger the ventilator
of two ventilation modalities, pressure support versus based on the absence of spontaneous respiratory rate as
volume-controlled ventilation, on lung dose of a radiola- assessed by continuous observation of the ventilator trac-
beled aerosol administered with a vibrating-mesh nebu- ings during the procedure. The ventilatory pattern during
lizer during invasive mechanical ventilation. VCV was: tidal volume of 8  mL/kg and respiratory rate
targeting a minute ventilation around 8 L/min during the
Methods postoperative awakening phase. Inspiratory time and res-
Study design and patient selection piratory rate were then adjusted to ensure an inspiratory/
This randomized, comparative, double-blind study expiratory ratio of 30 % with a constant inspiratory flow
included postoperative neurosurgery ventilated patients of 30 L/min. As there is no end-inspiratory pause in PSV,
with a healthy lung function. Eligibility criteria were none was imposed in VCV.
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 3 of 10

Fig. 1  Planar imaging to assess aerosol lung deposition during invasive mechanical ventilation. The ventilator was equipped with a 160-cm, 22-mm
inner-diameter ventilator circuit (IMMED, Bruxelles, Belgium) including a 7-cm proximal flow sensor (PFS, Hamilton Medical, Bonaduz, Switzerland)
positioned between the Y-piece and the catheter mount, a protection filter and the vibrating-mesh nebulizer. The patient was in semirecumbent
position at 15° with the head turned right to avoid the overlap of the thorax, the ventilator circuit and the gamma camera (at 10 cm of the sternum)

The inspiratory pressure level in PSV was set to reach Image acquisition and deposition analysis
a tidal volume of 8  mL/kg, and the expiratory trigger Image acquisitions were performed using a planar sin-
was set to obtain an inspiratory/expiratory ratio of 30 %. gle detector gamma camera (STARPORT 400 AC/T, GE,
Positive end-expiratory pressure was set at 5 cmH2O for Hørsholm, Denmark). Nine static anterior acquisitions
all patients. A bias flow of 10 L/min was imposed by the of 2  min were required for counts correction for back-
ventilator. Neither heating nor humidification systems ground, decay and attenuation and to analyze the pulmo-
were used to condition inspired air but a heat and mois- nary and extrapulmonary deposition according to recent
ture exchanger (HME) filter that was removed during the international recommendations (Fig. 2) [23, 24]. Delinea-
nebulization as suggested previously [8]. A HME filter tion of lung outlines, regions of interest and counts quan-
was placed on the expiratory limb during nebulization to tification were performed using the Odyssey program
measure the exhaled dose and to avoid radioactive parti- (LR Software (v7.0-1.7), Philips).
cles liberation. All data were anonymized by a code number. Data
analysis was blinded, i.e., performed by a physician of the
Particle size analysis Nuclear Medicine Department without knowledge nei-
Particle size analysis of the radiolabeled aerosol was ther of the ventilation mode used nor of any particular
assessed for a decelerating (peak inspiratory flow of 60 L/ clinical parameters.
min) and a constant inspiratory flow pattern (30 L/min) Right and left lung deposition was separately measured
that characterizes the PSV and VCV mode. We used an using a rectangular ROI dimensioned on lung outlines
eight-stage Andersen Cascade Impactor (Copley Scien- from the 15 % isocount contour of the 81mKr radioactive
tific Ltd, Nottingham, UK) based on pharmacopeia and gas scan [23]. Whole lung deposition was derived from
according to the bench model described by Miller et  al. the quantification of counts included in these both ROIs.
[22] (Additional file 1). The stage at the median amount Counts correction for background, decay and attenuation
of radioactivity defined the mass median aerodynamic was performed as described in previous studies [23–25]
diameter (MMAD). (Additional file  1). A right to the left lung deposition
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 4 of 10

Fig. 2  Summary of the protocol

ratio was calculated to compare both lung depositions. A from the 99mTc-DTPA acquisition normalized to the O/I
penetration index defined the penetration of aerosol par- ratio from the 81mKr gas acquisition as described previ-
ticles from an inner to an outer rectangular lung region ously [23]. Extrapulmonary deposition was determined
designed from the 81mKr ventilation scan and reported by the total count measured in the nebulizer reservoir
on the 99mTc-DTPA scan. The penetration index was minus whole lung count. Deposition analysis within the
calculated as the outer to inner lung region ratio (O/I) endotracheal tube and the trachea was determined by the
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 5 of 10

total extrapulmonary deposition minus the total count in Table 1  Patient characteristics
the ventilator circuit and retained in the expiratory filter PSV (n = 8) VCV (n = 9) p value
and the nebulizer reservoir. It was grouped due to the dif-
ficult differentiation of the deposition in the distal part Age (years) 56 ± 8 61 ± 11 0.254
of the endotracheal tube and the trachea. Radiolabeled Male, n (%) 5 (62.5) 4 (44.4)
deposition on each component of the ventilator circuit Height (cm) 169 ± 10 165 ± 11 0.441
and the expiratory filter was quantified using a ROI fitted Body weight (kg) 74 ± 14 68 ± 21 0.431
to their size. According to the ventilator circuit, the term Ideal body weight (kg) 65 ± 10 59 ± 10 0.516
“proximal” was used when the pieces were close to the Smoker, n (%) 3 (37.5) 3 (33.3)
patient, i.e., the proximal flow sensor, the catheter mount Surgery, n (%)
and the nebulizer T-piece.  Brain tumor resection 3 (37.5) 2 (22.2)
Pulmonary deposition and extrapulmonary deposition  Embolization of intracranial 3 (37.5) 0
unruptured aneurysm
of radiolabeled particles were expressed in counts and
 Neurosurgical clipping of a 1 (12.5) 6 (66.7)
as a percentage of the nominal dose (i.e., the amount of unruptured intracranial aneu‑
radioactivity placed in the nebulizer at the beginning of rysms
experiments).  Stereotactic brain biopsy 1 (12.5) 0
 Vestibular schwannoma resection 0 1 (11.1)
Statistical analysis Lung function
Statistical analysis was performed using SPSS software  FEV1 (% predicted value) 95 ± 16 97 ± 9 0.777
(version 20.0, IBM software). Sample size calculation was  FVC (% predicted value) 100 ± 19 101 ± 14 0.927
based on a 7  % expected difference in mean deposition  FEV1/FVC 77 ± 5 79 ± 6 0.509
for a statistical power of 80 % [3, 26]. Data are expressed
Quantitative variables are expressed as mean ± SD. Qualitative variables are
as mean ± standard deviation (SD) or median (25–75 % expressed as a proportion (%)
interquartile range (IQR)) depending on the normality FEV1 forced expiratory volume in 1 s, FVC forced vital capacity
tested using the Kolmogorov–Smirnov test. Compari-
son of both ventilation modes was conducted with inde-
pendent Student t tests or Mann–Whitney test. Both radioactivity quantified on the expiratory valve of the
lungs were compared using a Wilcoxon test for paired ventilator ensured the effective retention of exhaled par-
organs. Correlation between aerosol lung deposition and ticles in the HME filter. No ambient and surface contami-
the ventilatory pattern was conducted using the Pearson nation was detected after the procedure.
correlation coefficient (r). The intersubject variability of The MMAD at the distal tip of the endotracheal tube
whole lung deposition was determined by the calculation was 2.1  ±  0.07 and 1.95  ±  0.37  µm with a constant
of the coefficient of variation (CV). A p value lower than inspiratory flow pattern of 30  L/min and a decelerating
0.05 was considered significant. inspiratory flow pattern with a peak inspiratory flow of
60  L/min, respectively. Mechanical ventilation settings
Results and ventilatory pattern during inhalation are presented
Two hundred thirty-eight patients schedule for postoper- in Table 2. A similar inspiratory time was observed with
ative ICU admission were screened for eligibility between both ventilation modes. The decelerating inspiratory flow
July 2013 and June 2015 (completion date according pattern during PSV resulted in a higher inspiratory flow
to the sample size calculation). Among nineteen ran- rate. Patients had a lower respiratory rate during PSV,
domized patients, two patients allocated to PSV were resulting in a longer expiratory time.
excluded. One patient was excluded because mandatory Pulmonary and extrapulmonary deposition and the
sedation during the procedure implied to switch from penetration index according to both ventilation modes
PSV to pressure-controlled ventilation. The other patient are detailed in Table 3. The attenuation correction factor
was excluded because quantification of aerosol lung was 1.65 ± 0.19 and 1.56 ± 0.18 in PSV and VCV, respec-
deposition was impossible due to technical reasons. The tively. The deposited drug amount into the lungs was sig-
study included seventeen postoperative neurosurgery nificantly higher during VCV (p = 0.038), while a higher
ventilated patients, eight in PSV and nine in VCV (Fig. 2). aerosol deposition on the endotracheal tube, the trachea
Patient characteristics are depicted in Table 1. and main bronchi was observed during PSV (p = 0.043).
Nebulization lasted on average 9 ± 3 min. The dynamic We observed a similar aerosol deposition between both
acquisition revealed a linear increase in counts in lung ventilation modes in each lung analyzed separately
ROI that confirmed the absence of clearance of the (p = 0.057 and p = 0.885 for the right and the left lung,
aerosol during nebulization. The absence of potential respectively). The penetration index was also comparable
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 6 of 10

Table 2 Mechanical ventilation details and  ventilatory during PSV and VCV, respectively). The penetration
pattern during inhalation index of the right and the left lung was comparable dur-
PSV (n = 8) VCV (n = 9) p value ing PSV (p = 1.000) and during VCV (p = 0.066).
Pulmonary aerosol deposition was highly variable
Sedatives (propofol, mg/h) 190 (160–252) 200 (160–260) 0.622 among patients with a coefficient of variation of 28 % and
ETT diameter (mm) 8.0 (7.5–9.0) 7.0 (7.5–8.5) 0.252 33 % during PSV and VCV, respectively. A high variabil-
81m
Kr ventilation distribution ity of the penetration index and the right to the left lung
 Right/left lung ratio 1.13 ± 0.25 1.09 ± 0.32 0.755 deposition ratio was also found with the two ventilation
Ventilatory pattern during inhalation modes (Fig. 3a, b). Intersubject variability of aerosol lung
 Ppeak (cmH2O) 16 ± 3 20 ± 1 0.005 deposition whatever the ventilation mode is illustrated in
 VT insp (mL) 586 ± 117 530 ± 95 0.292 Fig. 3c.
 VT insp (mL/kg IBW) 8.65 (8.00–8.90) 8.70 (8.55–9.20) 0.439 No correlation between lung deposition and the inspir-
 RR (cycle/min) 14 ± 1 18 ± 2 <0.001 atory flow rate was found for the two modes. However,
 MVinsp (L/min) 8.01 ± 1.48 9.17 ± 0.92 0.070 lung deposition was correlated with the respiratory rate
 Flowpeak insp (L/min) 44 ± 8 32 ± 4 0.002 (r  =  0.741, p  =  0.036) and inversely correlated with the
 Tinsp (s) 1.15 ± 0.19 1.12 ± 0.12 0.664 expiratory time (r = −0.846, p = 0.008) during PSV.
 Texp (s) 3.72 (2.91–3.99) 2.24 (2.16–2.55) 0.001 Extrapulmonary aerosol deposition was higher dur-
 TinspTTot (%) 25 (22–34) 32 (32, 33) 0.072 ing PSV (p  =  0.038) due to a higher deposition on the
Data expressed as mean ± SD or median (25–75 % IQR) endotracheal tube, the trachea and main bronchi. A simi-
P-value in italic is considered significant (p < 0.05) lar aerosol deposition on the ventilator circuit (p = 0.486)
ETT endotracheal tube, MV minute ventilation, P pressure, RR respiratory rate, and the expiratory filter (p  =  0.710) was observed with
Tinsp inspiratory time, TinspT Tot inspiratory time to breathing cycle time ratio, Texp
expiratory time, V T tidal volume
both ventilation modes. Around 3 % of the nominal dose
was retained on the nebulizer reservoir at the end of
nebulization.
Table 3 Aerosol deposition in  seventeen postoperative
neurological patients Discussion
PSV (n = 8) VCV (n = 9) p value
This study showed that volume-controlled ventilation
increased aerosol delivery to the lungs as compared to
Pulmonary deposition (%) 10.5 ± 3.0 (28) 15.1 ± 5.0 (33) 0.038 pressure support ventilation with a vibrating-mesh nebu-
Right lung 6.1 ± 1.9 (31) 10.6 ± 5.8 (55) 0.057 lizer connected to the endotracheal tube. Lung dose and
 Penetration index 0.75 (0.30–0.94) 0.32 (0.16–0.77) 0.210 the site of deposition were highly variable among patients
Left lung 4.1 (3.8–4.6) 4.5 (2.2–5.6) 0.885 with both ventilation modes.
 Penetration index 0.67 (0.53–0.86) 0.74 (0.6–1.06) 0.211 This study demonstrated better aerosol delivery to the
Right/left lung ratio 1.39 (0.91–2.05) 3.33 (0.7–5.38) 0.336 lungs (between 10.5 and 15.1  % of the nominal dose)
Extrapulmonary deposition 89.5 ± 3.0 84.9 ± 5.0 0.038 through an endotracheal tube as compared to previous
(%) scintigraphic studies (2.1–5.3  % of the nominal dose)
ETT and tracheal area 27.4 ± 6.6 (24) 20.7 ± 6.0 (29) 0.043 [13, 14, 16]. High residual volume retained in the reser-
Expiratory filter 23.7 ± 5.3 (22) 22.5 ± 7.6 (34) 0.710 voir (29.6–51.5  % of the nominal dose) of their jet and
Ventilator circuit 34.7 ± 8.7 (25) 38.4 ± 12.3 (32) 0.486 ultrasonic nebulizers, delayed actuation of their breath-
  Proximal pieces 32.0 ± 7.4 (23) 35.9 ± 12.5 (35) 0.451 actuated jet nebulizer and deposition analysis performed
  Insp–expi tubing 2.7 ± 1.9 (70) 2.5 ± 1.7 (68) 0.833 on patients with respiratory failure (ARDS, COPD, neu-
Nebulizer retention 3.7 ± 0.9 (24) 3.3 ± 0.7 (21) 0.334 romuscular disorders, open-heart surgery) were factors
Data expressed as mean ± SD (coefficient of variation, %) or median (25–75 % pointed out by previous authors that affected lung deliv-
IQR). Proximal pieces of the ventilator circuit included the catheter mount, the
nebulizer T-piece and the proximal flow sensor
ery and hence explain this discrepancy [13, 14, 16].
P-value in italic is considered significant (p < 0.05) Lung deposition of 15  % reported during volume-
ETT endotracheal tube controlled ventilation is consistent with the amount
of amikacin measured in  vitro at the distal tip of the
between both modes for the right (p = 0.210) and the left endotracheal tube using the same nebulizer position and
lung (p = 0.211). ventilatory settings [7]. However, lung deposition is still
If we compare aerosol deposition between both lungs, low in comparison with other studies using breath-actu-
there was no significant difference of deposition between ated jet nebulizers (15.3 ± 9.5 % of radiotracer deposited
the right and the left lung (p  =  0.093 and p  =  0.066 in the lungs) [15], continuous vibrating-mesh nebulizers
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 7 of 10

Fig. 3  High intersubject variability of aerosol penetration through the lungs and its deposition between the right and the left lung during pressure
support ventilation and volume-controlled ventilation. a A penetration index equal to 1 indicated a linear aerosol penetration from the inner to
the outer part of the lungs. Particles deposition was limited to the central airways with both ventilation modes. b A right/left lung deposition ratio
equal to 1 indicated a similar aerosol deposition in both lungs. Right lung deposition was predominant with both ventilation modes, especially
during volume-controlled ventilation. c Scintigraphic images of aerosol lung deposition in two patients in volume-controlled ventilation (left) and
two patients in pressure support ventilation (right). With both ventilation modes, the first patient on the left benefits of a symmetrical aerosol lung
deposition while a predominant left lung or right lung deposition is depicted in the patient on the right. Aerosol penetration from the inner to the
outer lung region varies also among patients

(40–60  % of antibiotics deposited in piglets lungs) [26, nebulization. It led to a significant aerosol loss during the
27] and a recent breath-actuated vibrating-mesh nebu- expiratory time [5]. A more distal position was used in
lizer connected to the endotracheal tube (72  ±  11  % of other studies using continuous nebulization (from 30 cm
aerosol delivery to a lung model) [20]. The explanation of the Y-piece to the inspiratory inlet of the ventilator)
is the nebulizer configuration used in the present study, [3, 27]. At the time of the study, the proximal flow sen-
i.e., connected to the endotracheal tube for a continuous sor was considered as a potential barrier for nebulized
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 8 of 10

particles, implying to position the nebulizer after this higher variability of lung dose and lung penetration was
sensor just before the endotracheal tube. However, observed during volume-controlled ventilation as com-
we showed recently that the proximal flow sensor has pared to pressure support ventilation.
no impact when the nebulizer is placed at 45  cm of the The lung anatomy and the potential postoperative con-
Y-piece or connected to the inspiratory inlet of the venti- cerns (retention of bronchial secretions, minor atelec-
lator whatever the inspiratory flow pattern (decelerating tasis) more likely explain this high variability [35]. The
or constant) [7]. Both positions could have led to a better relatively small variability of the nebulizer output (5  %)
aerosol delivery with both ventilation modes and should [6] and images quantification performed by two blind cli-
be assessed in further in vivo studies. nicians ensured the accuracy and the reproducibility of
Many factors may explain the lower lung deposition the analysis.
during pressure support. The lower respiratory rate The aerosol distribution between both lungs was also
(while keeping a similar inspiratory time and tidal vol- variable among patients although a majority presented a
ume) during pressure support led to a decrease in lung higher aerosol deposition in the right lung. Two studies
deposition as compared to volume control. Reducing the also found a higher aerosol delivery in the right lung of
respiratory rate without increasing neither the inspira- intubated patients following open-heart surgery [14, 16].
tory time nor the tidal volume led to an increase in the Our patients had healthy lungs unaffected by the surgical
expiratory time and hence reduced aerosol delivery distal procedure. Repartition of the ventilation in the right/left
to the endotracheal tube during continuous nebulization lung depicted by the 81mKr ventilation scan was close to
[28]. Aerosol lung deposition normalized to the respira- 55/45 % indicating a normally distributed ventilation vol-
tory rate becomes similar to the two modes (p = 0.378). ume [36]. The increased right lung deposition could be
The low respiratory rate in pressure support was due to explained by this physiological increased right lung ven-
the minimal sedation mandatory to the safe postopera- tilation [37]. The fact that the patient’s head was turned
tive management of neurosurgery patients. Less patient right for all patients may also explain the predominant
sedation may lead to a higher respiratory rate and hence right lung deposition because of an uncontrolled twist
better lung deposition during pressure support providing of the endotracheal tube. This had no effect on a volume
that there is no increase of the inspiratory flow rate. tracer such as 81mKr, but lead to more proximal bronchus
The higher inspiratory flow rate during pressure sup- deposition on the right side. As shown in Fig. 3b, c, this
port ventilation could also explain the lower lung deposi- effect is rather versatile.
tion with this mode. It is well known that the higher the High aerosol impaction on the endotracheal tube and
inspiratory flow rate, the lower the lung deposition of the trachea (20–27 % of the nominal dose) was measured
aerosolized particles [29, 30]. However, lung deposition with both modes. Larger particles which were generated
was independent of the inspiratory flow rate variation at the inlet of the endotracheal tube directly impacted the
between the two modes tested in this study. lumen of the endotracheal tube. Higher impaction during
The aerosol penetration from the inner to the outer pressure support is explained by the higher inspiratory
part of the lungs was similar to the two modes. One flow rate inherent to this ventilation mode. Previous scin-
major factor that influences aerosol penetration is the tigraphic studies reported a comparable aerosol deposi-
particle size [31]. A comparable MMAD around 2  µm tion within the ventilator circuit around 30 % [13, 14, 16].
distal to the endotracheal tube was obtained whatever However, we found a majority of aerosolized particles
the inspiratory flow pattern. deposited on the proximal pieces of the ventilator circuit
As reported by Thomas et  al. [16], we found a pen- (proximal flow sensor, the catheter mount and the nebu-
etration index around 0.5 for both lungs. A penetration lizer T-piece).The expiratory filter captured around 20 %
index below 1 indicates a more central deposition of of aerosolized particles. Our nebulizer generated parti-
aerosol [32, 33]. Poor aerosol penetration during invasive cles proximally during the entire breathing cycle. Parti-
mechanical ventilation could be explained by particles cles were captured on proximal pieces of the ventilator
impaction and trickling from the endotracheal tube to circuit or directly lost on the expiratory limb during the
the trachea and main bronchi [14, 16, 34]. long expiratory time measured in this study [7, 38].
The high intersubject variability of the aerosol lung Our study has several limitations. First, extrapolation of
dose, lung penetration and the heterogeneous regional our results is limited due to the nebulization performed
distribution whatever the ventilation mode in patients in patients with healthy lungs, using a radiotracer instead
with healthy lungs is the major finding of this study. We of a relevant drug, using a minimal mandatory sedation
expected that the uncontrolled ventilatory pattern dur- during pressure support ventilation and a ventilator cir-
ing pressure support ventilation would induce a higher cuit without heated and humidifier system. However, the
variability of aerosol lung deposition. Surprisingly, a aim of the study was to focus on the effect of mechanical
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 9 of 10

ventilation on nebulization efficiency and not on its clini- Vaud, Filière physiothérapie, University of Applied Sciences and Arts Western
Switzerland, Avenue de Beaumont 21, 1011 Lausanne, Switzerland. 8 Louvain
cal benefit. Second, the continuous nebulizer was con- Drug Research Institute (LDRI), Université catholique de Louvain, Avenue
nected to the endotracheal tube. At the beginning of the Hippocrate 10, 1200 Brussels, Belgium. 9 Soins Intensifs, Clinique Saint-Pierre,
study, the distal position was not known as the optimal Avenue Reine Fabiola 9, 1340 Ottignies, Belgium. 10 Radiologie Intervention‑
nelle, Cliniques universitaires Saint-Luc, Avenue Hippocrate 10, 1200 Brussels,
position to reduce aerosol loss considering the presence Belgium. 11 Anesthésiologie, Cliniques universitaires Saint-Luc, Avenue Hip‑
of the proximal flow sensor on the ventilator circuit [7]. pocrate 10, 1200 Brussels, Belgium. 12 Neurochirurgie, Cliniques universitaires
Further clinical study with nebulizer positioned at 30, Saint-Luc, Avenue Hippocrate 10, 1200 Brussels, Belgium.
45  cm above the Y-piece or at the inspiratory outlet of Competing interests
the ventilator is required. Third, quantification was per- The authors declare that they have no competing interests.
formed based on counts derived from a single anterior
Funding
planar image instead of a geometric mean of anterior The author(s) received no specific funding for this work.
and posterior counts that could decrease the accuracy of
the quantification [23]. Safety concern imposed to avoid Received: 20 April 2016 Accepted: 28 June 2016
patient’s transfer to the examination table and posterior
images of the lungs could not be acquired due to bed
structure; this bias, if of any relevance, was, however,
constant for all patients in both groups. References
In conclusion, volume-controlled ventilation was 1. Ehrmann S, Roche-Campo F, Sferrazza Papa GF, Isabey D, Brochard L,
associated with higher aerosol delivery to the lungs as Apiou-Sbirlea G, et al. Aerosol therapy during mechanical ventilation: an
international survey. Intensive Care Med. 2013;39(6):1048–56.
compared to pressure support ventilation with a vibrat- 2. Dhand R, Guntur VP. How best to deliver aerosol medications to mechani‑
ing-mesh nebulizer connected to the endotracheal tube cally ventilated patients. Clin Chest Med. 2008;29(2):277–96.
in the specific conditions of the study. Lung deposition 3. Ferrari F, Liu ZH, Lu Q, Becquemin MH, Louchahi K, Aymard G, et al.
Comparison of lung tissue concentrations of nebulized ceftazidime in
during pressure support ventilation was nonetheless ventilated piglets: ultrasonic versus vibrating plate nebulizers. Intensive
substantial and should be further assessed using lighter Care Med. 2008;34(9):1718–23.
sedation. Aerosol therapy is totally unpredictable in term 4. Ari A, Fink JB. Factors affecting bronchodilator delivery in mechanically
ventilated adults. Nurs Crit Care. 2010;15(4):192–203.
to the location of aerosol deposition through the lungs. 5. Ari A, Areabi H, Fink JB. Evaluation of aerosol generator devices at 3 loca‑
A high intersubject variability of lung doses, particles tions in humidified and non-humidified circuits during adult mechanical
penetration and distribution between both lungs was ventilation. Respir Care. 2010;55(7):837–44.
6. Ari A, Atalay OT, Harwood R, Sheard MM, Aljamhan EA, Fink JB. Influence
demonstrated with the two ventilation modes. Before of nebulizer type, position, and bias flow on aerosol drug delivery in
promoting volume control mode to administer an aerosol simulated pediatric and adult lung models during mechanical ventila‑
during invasive mechanical ventilation, we should assess tion. Respir Care. 2010;55(7):845–51.
7. Dugernier J, Wittebole X, Roeseler J, Michotte JB, Sottiaux T, Dugernier
the clinical benefit of the average 50 % differences in lung T, et al. Influence of inspiratory flow pattern and nebulizer position
deposition in patients with a need for aerosol therapy on aerosol delivery with a vibrating-mesh nebulizer during invasive
such as inhaled antibiotics. mechanical ventilation: an in vitro analysis. J Aerosol Med Pulm Drug
Deliv. 2015;28(3):229–36.
8. Rouby JJ, Bouhemad B, Monsel A, Brisson H, Arbelot C, Lu Q, et al.
Additional file Aerosolized antibiotics for ventilator-associated pneumonia: lessons from
experimental studies. Anesthesiology. 2012;117(6):1364–80.
Additional file 1. Detailed description of particle size analysis and image 9. Sassoon CS. Ventilator-associated diaphragmatic dysfunction. Am J Respir
correction for background, decay and attenuation for deposition analysis. Crit Care Med. 2002;166(8):1017–8.
10. Sassoon CS, Zhu E, Caiozzo VJ. Assist-control mechanical ventilation
attenuates ventilator-induced diaphragmatic dysfunction. Am J Respir
Crit Care Med. 2004;170(6):626–32.
Authors’ contributions 11. Conway J. Lung imaging—two dimensional gamma scintigraphy, SPECT,
JD, GR, VD, JR, MAD, CR, FJ, and PFL were involved in data collection. JD, FJ, CT and PET. Adv Drug Deliv Rev. 2012;64(4):357–68.
and GR were involved in study design. JD, VD, FJ, and GR were involved in data 12. Newman S, Fleming J. Challenges in assessing regional distribu‑
analysis. JD, GR, XW, and TD prepared the manuscript. JD, GR, XW, JR, VD, TS, tion of inhaled drug in the human lungs. Expert Opin Drug Deliv.
JBM, RV, TD, PG, MAD, CR, PH, FJ, and PFL reviewed the manuscript. All authors 2011;8(7):841–55.
read and approved the final manuscript. 13. Harvey CJ, O’Doherty MJ, Page CJ, Thomas SH, Nunan TO, Treacher DF.
Effect of a spacer on pulmonary aerosol deposition from a jet nebuliser
Author details during mechanical ventilation. Thorax. 1995;50(1):50–3.
1
 Soins Intensifs, Médecine Physique, Cliniques universitaires Saint-Luc, Avenue 14. Harvey CJ, O’Doherty MJ, Page CJ, Thomas SH, Nunan TO, Treacher DF.
Hippocrate 10, 1200 Brussels, Belgium. 2 Médecine Physique, Cliniques univer‑ Comparison of jet and ultrasonic nebulizer pulmonary aerosol deposition
sitaires Saint-Luc, Avenue Hippocrate 10, 1200 Brussels, Belgium. 3 Institut de during mechanical ventilation. Eur Respir J. 1997;10(4):905–9.
Recherche Expérimentale et Clinique (IREC), Pneumologie, ORL & Dermatolo‑ 15. O’Riordan TG, Palmer LB, Smaldone GC. Aerosol deposition in mechani‑
gie, Université catholique de Louvain, 1200 Brussels, Belgium. 4 Soins Intensifs, cally ventilated patients. Optimizing nebulizer delivery. Am J Respir Crit
Cliniques universitaires Saint-Luc, Avenue Hippocrate 10, 1200 Brussels, Care Med. 1994;149(1):214–9.
Belgium. 5 Médecine Nucléaire, Cliniques universitaires Saint-Luc, Avenue 16. Thomas SH, O’Doherty MJ, Fidler HM, Page CJ, Treacher DF, Nunan TO.
Hippocrate 10, 1200 Brussels, Belgium. 6 Soins Intensifs, Clinique Notre-Dame Pulmonary deposition of a nebulised aerosol during mechanical ventila‑
de Grâce, Chaussée de Nivelles 212, Gosselies, Belgium. 7 Haute Ecole de Santé tion. Thorax. 1993;48(2):154–9.
Dugernier et al. Ann. Intensive Care (2016) 6:73 Page 10 of 10

17. Goldstein I, Wallet F, Nicolas-Robin A, Ferrari F, Marquette CH, Rouby JJ. 29. Fink JB, Dhand R, Grychowski J, Fahey PJ, Tobin MJ. Reconciling in vitro
Lung deposition and efficiency of nebulized amikacin during Escheri‑ and in vivo measurements of aerosol delivery from a metered-dose
chia coli pneumonia in ventilated piglets. Am J Respir Crit Care Med. inhaler during mechanical ventilation and defining efficiency-enhancing
2002;166(10):1375–81. factors. Am J Respir Crit Care Med. 1999;159(1):63–8.
18. Niederman MS, Chastre J, Corkery K, Fink JB, Luyt CE, Garcia MS. BAY41- 30. O’Riordan TG, Greco MJ, Perry RJ, Smaldone GC. Nebulizer function dur‑
6551 achieves bactericidal tracheal aspirate amikacin concentrations ing mechanical ventilation. Am Rev Respir Dis. 1992;145(5):1117–22.
in mechanically ventilated patients with Gram-negative pneumonia. 31. Heyder J. Deposition of inhaled particles in the human respiratory tract
Intensive Care Med. 2012;38(2):263–71. and consequences for regional targeting in respiratory drug delivery.
19. Luyt CE, Eldon MA, Stass H, Gribben D, Corkery K, Chastre J. Pharmacoki‑ Proc Am Thorac Soc. 2004;1(4):315–20.
netics and tolerability of amikacin administered as BAY41-6551 aerosol 32. Fleming J, Conway J, Majoral C, Katz I, Caillibotte G, Pichelin M, et al. Con‑
in mechanically ventilated patients with gram-negative pneumonia and trolled, parametric, individualized, 2-D and 3-D imaging measurements of
acute renal failure. J Aerosol Med Pulm Drug Deliv. 2011;24(4):183–90. aerosol deposition in the respiratory tract of asthmatic human subjects
20. Luyt CE, Clavel M, Guntupalli K, Johannigman J, Kennedy JI, Wood C, for model validation. J Aerosol Med Pulm Drug Deliv. 2015;28:432–51.
et al. Pharmacokinetics and lung delivery of PDDS-aerosolized amikacin 33. Majoral C, Fleming J, Conway J, Katz I, Tossici-Bolt L, Pichelin M, et al. Con‑
(NKTR-061) in intubated and mechanically ventilated patients with noso‑ trolled, parametric, individualized, 2D and 3D imaging measurements of
comial pneumonia. Crit Care. 2009;13(6):R200. aerosol deposition in the respiratory tract of healthy human volunteers:
21. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, Coates A, et al. in vivo data analysis. J Aerosol Med Pulm Drug Deliv. 2014;27(5):349–62.
Standardisation of spirometry. Eur Respir J. 2005;26(2):319–38. 34. MacIntyre NR, Silver RM, Miller CW, Schuler F, Coleman RE. Aerosol
22. Miller DD, Amin MM, Palmer LB, Shah AR, Smaldone GC. Aerosol delivery delivery in intubated, mechanically ventilated patients. Crit Care Med.
and modern mechanical ventilation: in vitro/in vivo evaluation. Am J 1985;13(2):81–4.
Respir Crit Care Med. 2003;168(10):1205–9. 35. Thomas SH, O’Doherty MJ, Page CJ, Nunan TO. Variability in the
23. Newman S, Bennett WD, Biddiscombe M, Devadason SG, Dolovich MB, measurement of nebulized aerosol deposition in man. Clin Sci (Lond).
Fleming J, et al. Standardization of techniques for using planar (2D) 1991;81(6):767–75.
imaging for aerosol deposition assessment of orally inhaled products. J 36. Biddiscombe MF, Meah SN, Underwood SR, Usmani OS. Comparing lung
Aerosol Med Pulm Drug Deliv. 2012;25(Suppl 1):S10–28. regions of interest in gamma scintigraphy for assessing inhaled therapeu‑
24. Pitcairn GRNS. Tissue attenuation corrections in gamma scintigraphy. J tic aerosol deposition. J Aerosol Med Pulm Drug Deliv. 2011;24(3):165–73.
Aerosol Med. 1997;10(3):187–98. 37. Moller W, Meyer G, Scheuch G, Kreyling WG, Bennett WD. Left-to-right
25. De Backer W, Devolder A, Poli G, Acerbi D, Monno R, Herpich C, et al. Lung asymmetry of aerosol deposition after shallow bolus inhalation depends
deposition of BDP/formoterol HFA pMDI in healthy volunteers, asthmatic, on lung ventilation. J Aerosol Med Pulm Drug Deliv. 2009;22(4):333–9.
and COPD patients. J Aerosol Med Pulm Drug Deliv. 2010;23(3):137–48. 38. Diblasi RM. Clearing the mist from our eyes: bronchodilators, mechanical
26. Ferrari F, Lu Q, Girardi C, Petitjean O, Marquette CH, Wallet F, et al. Nebu‑ ventilation, new devices, locations, and what you should know about
lized ceftazidime in experimental pneumonia caused by partially resist‑ bias flow. Respir Care. 2010;55(7):942–6.
ant Pseudomonas aeruginosa. Intensive Care Med. 2009;35(10):1792–800.
27. Lu Q, Girardi C, Zhang M, Bouhemad B, Louchahi K, Petitjean O, et al.
Nebulized and intravenous colistin in experimental pneumonia caused
by Pseudomonas aeruginosa. Intensive Care Med. 2010;36(7):1147–55.
28. Thomas SH, O’Doherty MJ, Page CJ, Treacher DF, Nunan TO. Delivery of
ultrasonic nebulized aerosols to a lung model during mechanical ventila‑
tion. Am Rev Respir Dis. 1993;148(4 Pt 1):872–7.