Professional Documents
Culture Documents
the Netherlands). Preterm infants (, 37 wk) were included if they TABLE 1. PATIENT CHARACTERISTICS
failed nasal continuous positive airway pressure (nCPAP) and needed
high-frequency oscillatory ventilation (HFOV) for a suspected diagno- 5-min
sis of RDS within 72 hours after birth. All patients were ventilated in Subject No. GA (wk) Weight (g) Apgar Score Age (h) mOI
the supine position and were not sedated or paralyzed. The study was 1 27.6 1,000 8 4 9.76
approved by the institutional board and written informed consent was 2 28.3 1,070 8 2 7.17
obtained from both parents. 3 29.4 1,200 9 8 4.44
HFOV was used as a primary mode and combined with an open lung 4 25.3 750 6 3 4.49
ventilation strategy using oxygenation as an indirect marker for lung vol- 5 28.1 760 9 4 4.53
ume. Briefly, the continuous distending pressure (CDP) was increased 6 26.9 960 9 16 9.52
stepwise (1–2 cm H2O) until oxygenation no longer improved or frac- 7 28.0 720 8 3 8.70
tional inspired oxygen (FIO2) was less than or equal to 0.25 (CDPo). 8 26.4 750 7 4 8.99
Next, CDP was decreased with 1- to 2-cm H2O steps until oxygenation 9 28.6 820 7 33 3.48
deteriorated (CDPc). Finally, the lung was once more recruited (CDPo) 10 29.3 1,095 8 2 3.08
and then stabilized with a CDP 2 cm H2O above CDPc (CDPopt). 11 28.7 1,220 8 14 3.45
Previous studies have shown that this open lung approach is feasible 12 28.7 880 8 24 6.67
13 31.9 1,080 9 4 10.63
in the majority of preterm infants with RDS and does not lead to
14 29.6 1,000 7 10 5.27
hemodynamic instability (17, 18).
15 26.3 1,020 9 2 11.36
Surfactant was then administered as a bolus via a closed system cath-
Median 28.3 1,000 8 4 6.67
eter. After a stabilization period of 10 minutes, the CDPc, CDPo, and
CDPopt were once more determined. If CDPc could be successfully Definition of abbreviations: CDP ¼ continuous distending pressure; GA ¼ ges-
reduced to 8 cm H2O without oxygenation compromise, the procedure tational age; mOI ¼ modified oxygenation index at the start of recruitment using
was stopped and this pressure was designated as CDPopt. the following equation: CDP 3 FIO2/SpO2; SpO2 ¼ oxygen saturation as measured
Changes in lung impedance (DZ) were continuously recorded by pulse oximetry.
during the above-described procedures, using a scan rate of 44 Hz.
Relative DZ was calculated for each decremental pressure step before This volume increase was reached after a median time of 241
and after surfactant treatment using a 30-second reference period at (interquartile range [IQR], 118–300) seconds. The median ven-
the start of recruitment and just before surfactant administration, re-
tral to dorsal ratio in lung volume changed significantly from 1.16
spectively. All DZ values were normalized setting the DZ at presurfac-
tant CDPo at 100%. Using the pressure/impedance pairs, individual (IQR, 0.79–1.89) before to 0.81 (IQR, 0.40–0.96) after surfactant
pre- and postsurfactant deflation limbs were plotted according to administration, whereas the right to left ratio remained un-
Venegas and colleagues, and the upper inflection points (UIPs), max- changed (Table 2).
imum compliance of the respiratory system (Crsmax), and pressure at The fitting curves of the deflation limbs before and after sur-
Crsmax were calculated (19). The immediate effect of surfactant on lung factant could be constructed in all patients with a mean goodness
volume was assessed by calculating the increase in DZ after adminis- of fit (R2) of 0.98 6 0.03 and 0.97 6 0.05, respectively. The UIP
tration and the stabilization time. The above-described analyses were after surfactant (10.4 6 2.4 cm H2O) was significantly lower com-
performed for the whole chest cross-section as well as for the ventral, pared with the UIP before surfactant (16.4 6 3.1 cm H2O) (P ,
dorsal, right, and left lung regions. The regional effect of surfactant on
0.01) (Table 2). Crsmax increased significantly from a median of
lung volume was assessed by calculating the ventral/dorsal and right/
left ratios at presurfactant CDPopt, directly after surfactant but before
6.9 (IQR, 4.9–8.1) %/cm H2O before to 9.0 (IQR, 8.3–11.8) after
the first decremental pressure step and at postsurfactant CDPopt. surfactant administration. The pressure at which Crsmax was
At these same time points the effect of surfactant treatment on oscillation reached also changed from 12.0 (IQR, 8.8–13.2) before to 7.2
volume was calculated using band-pass filtering (600 6 15/min). In (IQR, 4.7–7.8) cm H2O after surfactant treatment (Table 3).
addition, the ventral to dorsal and left to right distribution of the These changes were also observed in the different analyzed lung
normalized oscillatory DZ was calculated using functional images (20, regions.
21). The area under the curve and geometrical centers (area under the
curve ventral ¼ dorsal) were calculated for the ventral, dorsal, left, and
Surfactant Effect on Ventilation
right regions.
Nonlinear regression was used to plot the deflation limbs. For com- Compared with the presurfactant time point (¼ 100%), the me-
parative analyses the Mann-Whitney or Wilcoxon rank test was used for dian oscillation volume decreased significantly to 73.6% (IQR,
skewed data and the Student t test for a normal distributed data. A 62.1–80.6) immediately after surfactant treatment followed by
P value less than 0.05 was considered statistically significant. an increase to 85.2% (IQR, 75.2–91.4) at optimal postsurfactant
inflation (Table 2). Transcutaneous carbon dioxide pressure ini-
RESULTS tially increased after surfactant administration but reached sig-
nificantly lower values at optimal inflation, despite the fact that
Fifteen newborn infants were included in the study and com- the mean oscillation pressure amplitude decreased from 20.4 6
pleted the recruitment procedure and surfactant administration 2.6 cm H2O before to 18.5 6 3.5 cm H2O after surfactant treat-
without complications (Table 1). The mean CDP at the start of ment (Table 3).
lung recruitment (CDPst) was 8.3 6 1.7 cm H2O with a mean In general, oscillatory ventilation showed a slight asymmetri-
FIO2 of 0.73 6 0.27. The mean CDPo, CDPc, and CDPopt before cal distribution favoring the ventral lung regions along the ver-
surfactant treatment were, respectively, 19.3 6 1.7, 11.0 6 1.6, tical axis (Table 2 and Figure 2). Furthermore, oscillatory
and 13.0 6 1.6 cm H2O, with a mean FIO2 of 0.26 6 0.04. Sur- ventilation was more ventrally located in the right lung com-
factant was administered in a mean dose of 136 6 27 mg/kg, after pared with the left lung. Surfactant administration did not
which CDP could be decreased in all patients to 8.0 cm H2O, with change this regional distribution (Table 2, Figures 2 and 3).
a mean FIO2 of 0.21.
DISCUSSION
Surfactant Effect on Lung Volume To our knowledge, this is the first study that continuously mea-
Lung volume increased in all patients after surfactant adminis- sured regional lung volume changes during and after surfactant
tration with a mean increase of 61 6 39% expressed as per- treatment in preterm infants with RDS. Up to now, the effect of
centage of the presurfactant lung volume at CDPo (Figure 1). exogenous surfactant on lung volume has only been determined
102 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 184 2011
in conventionally ventilated preterm infants, using different tech- completed in the majority of infants within 5 minutes after sur-
niques to measure lung volume changes and different time points factant administration. To our knowledge, this observation has
after administration (5–10, 13). These studies showed that surfac- not been previously reported.
tant increased FRC by 20 to 100% (4, 12, 13). In the present study, In addition to measuring the changes in lung volume in the
surfactant was administered during HFOV and after a lung re- whole chest cross-section after surfactant treatment, EIT also pro-
cruitment procedure. The main reason for recruiting the lung vided regional information on these changes, which ideally should
before surfactant treatment was to minimize ventilation- be homogenously distributed. Our study shows that the increase in
induced lung injury as much as possible during the presurfactant lung volume was most prominent in the dorsal lung region (i.e., in
ventilation period. Surfactant administration under these condi- the dependent lung parts). This finding seems to support previous
tions is different from the studies using conventional ventilation animal experimental data suggesting that gravity plays an impor-
in which surfactant also plays an important role in lung recruit- tant role in the distribution of exogenous surfactant (22).
ment. Despite these differences, we found a comparable increase The observed increase in lung volume can be caused by either
in lung volume (60%) in preterm infants on HFOV. But more alveolar/saccular distension or recruitment. Considering the fact
importantly, we were able to show that this volume increase was that the lungs of these preterm infants were already recruited
Miedema, de Jongh, Frerichs, et al.: Effect of Surfactant on Lung Function in Infants 103
TABLE 2. CHANGES IN LUNG VOLUME, OSCILLATION VOLUME DISTRIBUTION, AND FRACTIONAL VENTILATION DISTRIBUTION DIRECTLY
AFTER SURFACTANT INSTILLATION AND AT THE OPTIMAL PRESSURE POSTSURFACTANT
Optimal Pressure Postsurfactant before the First Optimal Pressure
Presurfactant Decremental Pressure Step Postsurfactant
Definition of abbreviations: AUCven ¼ area under the curve of the ventral lung region; DZosc ¼ relative oscillation impedance change; TcPCO2 ¼ transcutaneous carbon
dioxide pressure.
Values are shown as mean 6 SD or median (IQR, 25–75).
* Indicates a significant change in contrast to optimal pressure before surfactant, P , 0.05.
y
Indicates a significant change in contrast to directly after surfactant treatment, P , 0.01.
z
Indicates a significant change in contrast to optimal pressure before surfactant, P , 0.01.
x
Indicates a significant change in contrast to directly after surfactant treatment, P , 0.05.
k
Indicates a significant difference in contrast to the right half of the chest, P , 0.01.
before surfactant administration, it is most likely that the observed technique (7). Bjorklund and colleagues also found a surfac-
volume increase after surfactant administration was, for the larger tant-induced increase in Crsmax at lower airway pressures (7).
part, caused by distension. The fact that we observed only a small In addition, the presented deflation limbs suggested a stabiliz-
improvement in oxygenation after surfactant treatment (FIO2 from ing effect of surfactant on lung volume, although more objec-
0.26 to 0.21) supports this assumption, as only alveolar recruit- tive variables, such as UIPs, were not provided. In contrast to
ment and not distension will improve oxygenation. our study, FRC and TLC showed no increase 20 to 40 minutes
To assess the effect of surfactant on lung mechanics, we after surfactant administration. As noted by the authors, this
mapped the deflation limb of the pressure/impedance curve be- discrepancy with our study may be, in part, explained by the fact
fore and after treatment. As indicated by the significant de- that presurfactant FRC was overestimated by applying repeated
crease in the UIP of the deflation limb, surfactant stabilized deep inflations before each measurement.
lung volume at much lower pressures compared with the By filtering the impedance data in the 10-Hz domain, we were
surfactant-deficient lung. Surfactant treatment also resulted able to establish the effect of surfactant on oscillatory ventila-
in a significant increase in Crsmax and a reduction of the air- tion. Immediately after surfactant administration, the oscillation
way pressure at which the maximal compliance was reached. volume decreased, probably as a result of airway obstruction.
Both the stabilization of lung volume and the increase in com- Over time the oscillation volume improved as airway pressures
pliance are consistent with previous animal studies (4, 13). were reduced and surfactant moved into the periphery of the lung.
Studies in preterm infants have mainly focused on the changes The fact that the oscillation volumes did not return to presurfac-
in compliance and showed conflicting results (5, 11, 13, 23). tant values after completing the postsurfactant recruitment proce-
Besides differences in the timing and method (static versus dure is probably best explained by the reduction in the pressure
dynamic) of compliance measurement, the present study amplitude. Despite this decrease in pressure amplitude and oscil-
seems to suggest that part of this inconsistency can be lation volume, transcutaneous carbon dioxide pressure levels were
explained by the fact that some studies failed to reduce airway lower after surfactant, indicating improved alveolar ventilation as
pressures after surfactant treatment and therefore did not de- also shown in previous studies in premature infants (5, 11–13).
tect the surfactant-induced increase in compliance (3, 4, 13). Analysis of the regional distribution of the oscillatory volume
Only one study mapped the deflation limb of the pressure/ showed that ventilation was more prominent in the right and ven-
volume curve in paralyzed preterm infants before and 30 tral lung region. This (novel) finding is probably best explained
minutes after surfactant treatment, using a tracer gas washout by the anatomical position of the heart, resulting in less lung
TABLE 3. CHANGES IN THE DEFLATION LIMB CHARACTERISTICS BEFORE AND AFTER SURFACTANT TREATMENT
Deflation Limb before Surfactant Deflation Limb after Surfactant
UIP, cm H2O 16.4 6 3.1 15.7 6 2.8 16.2 6 3.8 10.4 6 2.4* 10.5 6 2.8* 10.5 6 3.2*
Crsmax, %/cm H2O 6.9 (4.9–8.1) 3.6 (2.7–4.4) 3.7 (2.8–5.9) 9.0 (8.3–11.8)† 4.9 (4.1–6.0) 5.3 (3.6–8.3)
Pressure at Crsmax, cm H2O 12.0 (9.3–12.9) 10.0 (8.6–12.2) 11.4 (8.3–12.6) 7.2 (4.7–7.7)* 6.7 (5.0–7.5)* 6.7 (4.7–7.6)†
Definition of abbreviations: Crsmax ¼ maximal compliance of the respiratory system; UIP ¼ upper inflection point.
Values are shown as mean 6 SD or median (IQR, 25–75).
* Indicates a significant change in contrast to the deflation limb before surfactant, P , 0.01.
y
Indicates a significant change in contrast to the deflation limb before surfactant, P , 0.05.
104 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 184 2011
tissue and thus oscillatory ventilation in the left ventral region. have adopted the use of early nCPAP makes our study results
This finding seems to support the validity of the presented EIT clinically relevant.
data. It was interesting to observe that surfactant treatment did Despite these limitations, our study has important implica-
not change the distribution of oscillatory ventilation. Although tions for clinical practice. First, this study shows that surfactant
this finding might suggest a homogenous surfactant distribution, treatment during high-frequency ventilation increases lung vol-
other explanations should also be considered. First, surfactant ume comparable to treatment during conventional (tidal) venti-
treatment in an already recruited, and thus homogeneously aer- lation. Second, based on the stabilizing effect of surfactant on
ated lung might not impact ventilation distribution even if surfac- lung volume, clinicians should reduce the airway pressures after
tant distribution is heterogeneous. Second, the oscillatory volumes treatment to avoid possible overdistension and to profit from the
may be too small to pick up more subtle changes in ventilation increased compliance. Third, this reduction in pressure should be
distribution. This might be different during conventional mechan- started within 5 minutes after surfactant therapy. Finally, rescue
ical ventilation, although a recent animal study does not seem to surfactant treatment in a recruited lung seems to result in a rel-
support this explanation (2). atively homogenous distribution, although there is a possible
This study has several limitations that need to be addressed. gravity-dependent tendency for surfactant to move to the more
First, EIT only monitored lung volume changes in one transver- dependent lung parts. This latter finding suggests that body po-
sal “slice” of the lung. RDS is suggested to be a homogenous sitioning might be an effective way to augment regional surfac-
lung disease, making it highly likely that the presented findings tant deposition in heterogeneous lung disease.
are representative for the entire lung. Second, although not In conclusion, this study shows that rescue surfactant treat-
essential, this study did not provide information on the absolute ment in open lung high-frequency ventilated preterm infants
changes in lung volume as the calibration of the EIT signal has, with RDS causes a rapid increase (minutes) and subsequent sta-
so far, not been feasible. Third, all the patients in this study bilization of lung volume, which is most prominent in depen-
were treated with open lung HFOV. The results may be differ- dent lung regions. In addition, surfactant treatment increases
ent during conventional (tidal) ventilation, during which surfac- maximal compliance, but at lower airway pressures. Ventilation
tant also plays an important role in lung recruitment. However, distribution is relatively homogeneous and not affected by
the similar findings on FRC and compliance in animal and hu- surfactant.
man studies using conventional ventilation seem to be reassur- Author Disclosure: M.M. does not have a financial relationship with a commercial
ing (2, 7). Finally, surfactant was administered relatively late entity that has an interest in the subject of this manuscript. F.H.d.J. was an expert
(. 2 h) after birth, which may have reduced its efficacy. This witness for Teva Pharmaceutical. I.F. received travel accommodations from Car-
dinal Health. M.B.v.V. does not have a financial relationship with a commercial
delay is probably best explained by the fact that all patients in entity that has an interest in the subject of this manuscript. A.H.v.K. received
our unit are started on nCPAP (1). The fact that many clinicians grant support from Chiesi Pharmaceuticals.