Main

Surfactant treatment for neonatal RDS has become routine in most developed countries. The introduction of this therapy resulted in a reduction in mortality and complications caused by RDS (1,2). However, the surfactant instillation procedure has raised concerns about safety and efficacy (3–5). Concerns about safety were initiated by reports that described rapid fluctuation in blood pressure, cerebral perfusion, and cerebro-electrical activity after surfactant instillation (6–11). It is suggested that these fluctuations in CBF may enhance the development of intraventricular hemorrhage in the premature newborn (4). This hypothesis is supported by the fact that the incidence of intraventricular hemorrhage has not decreased markedly since the introduction of surfactant therapy (12,13). Concerns about the efficacy of the instillation procedure involve the experience that not all infants respond to surfactant treatment in terms of a sustained improvement in lung function. In addition, many infants develop bronchopulmonary dysplasia after RDS (2). Nonuniform distribution of surfactant throughout the lungs may be one of the explanations for the incomplete success of this therapy, because animal studies have shown that a more uniform distribution is correlated with a superior clinical response (14).

Modifications of surfactant treatment procedures (e.g. dosage, volume of doses, number of doses, duration of instillation, body position) have been studied extensively in animal models. From these studies it can be concluded that a high volume of natural surfactant, quickly instilled in one or two doses, is the most efficacious on lung function and distribution throughout the lungs, but hemodynamic side effects may occur (15–18). Slow infusion resulted in fewer hemodynamic side effects, but it was far less efficacious on lung function and surfactant distribution (18,19). Nebulization of surfactant might be an attractive alternative but must be technically improved before it can be clinically applicable (20–27). It is remarkable that to date, investigations on variations in ventilation regimens during surfactant instillation are scarce, although many variations are clinically applied.

Therefore, we performed an experimental study to evaluate one of those variations. We investigated in rabbits with severe respiratory failure whether maintenance of airway pressures, by avoiding disconnection of the endotracheal tube during surfactant instillation, improves the effect on lung function and promotes uniformity of surfactant distribution. In addition, we studied whether avoidance of tube disconnection prevents the adverse effects of surfactant instillation on blood pressure and CBF.

METHODS

Animals. The experiments were performed in the central animal laboratory of the University of Groningen, under approved institutional animal care protocols with concern for animal welfare. Twelve young adult Chinchilla rabbits (2.8 ± 0.1 kg BW) were anesthetized with sodium pentobarbital 30 mg/kg BW (Nembutal, Abbott, the Netherlands), tracheotomized, intubated with an endotracheal tube with side lumen, and paralyzed with pancuronium bromide 0.1 mg/kg BW (Pavulon, Organon, the Netherlands). Additional sodium pentobarbital was given hourly by 15-min intravenous infusion. Additional pancuronium bromide was given when necessary.

The rabbits were artificially ventilated with a pressure-limited, time-cycled, neonatal ventilator (Babylog 8000, Drägerwerk AG, Lübeck, Germany). Initial ventilator settings were as follows: frequency of 60 per min, inspiratory time equal to expiratory time (0.5 sec), and gas flow 1.0 L/min per kg BW, which yielded a tidal volume varying from 7 to 8 mL/kg BW. Peak inspiratory pressure was limited at 20 cm H2O; FiO2 was set at 1.0.

After 15-min stabilization, the lung lavage procedure was performed described earlier (28,29). Saline (35 mL/kg BW at 38°C) was endotracheally instilled and removed again. This procedure was repeated five times. After each lavage, the rabbit was allowed to recover for 5 min. The PEEP was increased stepwise to 8 cm H2O. Peak inspiratory pressure did not exceed 20 cm H2O. After 15 min, PEEP was decreased to 5 cm H2. After another 15 min, surfactant replacement was started, t = 0 (see "Surfactant treatments"). Two hours after surfactant instillation, by withholding pancuronium bromide, the animals were allowed to breathe spontaneously. In the subsequent hour, the ventilator was supported by synchronized intermittent mandatory ventilation and when more than 75% of the minute volume was achieved by spontaneous breathing, the ventilator was set on CPAP mode. During the last hour, CPAP was decreased from 5 cm H2O to 2.5 cm H2O. During CPAP, FiO2 was lowered stepwise every 15 min, as long as PaO2 > 8.0 kPa. The experiment ended when FiO2 was 0.4, when PaO2 was <8.0 kPa, or when the animal died.

Surfactant treatments. Alveofact (45 mg/mL; Dr. Karl Thomae GmbH, Biberach an der Riss, Germany), a bovine pulmonary surfactant, was labeled with 99mTc-Nanocoll (5 MBq in 0.5 mL saline) and stirred gently. Nanocoll (Solco, Biomedica S.p.A. Vercelli, Italy) contains human albumin as a nanocolloid with 95% of the particles smaller than 80 nm. In a previous study, we demonstrated that the pulmonary distribution of 99mTc-Nanocoll labeled surfactant correlated well with a labeled surfactant component (14C-DPPC) and could be used as a valid label to assess surfactant distribution (26,27).

In six rabbits, one bolus of 100 mg/kg BW labeled surfactant was instilled after disconnecting the rabbit from the ventilator tubing within 10 s through a cannula that was positioned with the tip at the end of the endotracheal tube. Immediately after instillation, the rabbit was reconnected to the ventilator tubing again without changing the settings and without changing the position of the animal. In six other rabbits, 100 mg/kg BW labeled surfactant was endotracheally instilled in 10 s through the side lumen of the endotracheal tube without disconnecting the tube from the ventilator and, thus, without interrupting ventilation. This tube had a side lumen with a hole at the tip. The ventilator settings were not changed during and after instillation.

Measurements. Every 15 to 30 min, sequential blood samples were taken to assess PaO2, PaCO2, and pH, using an ABL 330 blood gas analyzer (Radiometer Co, Copenhagen, Denmark). Dynamic compliance was assessed using computerized flow- and pressure measurements of the Babylog 8000 by a flow sensor in the Y-piece before and after surfactant administration and after the weaning. Just before killing, the animals were paralyzed again and briefly ventilated on intermittent positive-pressure ventilation with 5 cm H2O of PEEP to allow the measurement of dynamic compliance at the same ventilator settings as before.

Intravascular arterial blood pressure was continuously measured using a pressure transducer (Uniflow, Baxter Healthcare Corporation, Santa Ana, CA). The CBF was continuously measured using an ultrasonic transit time flow probe (Transonic Systems Inc, Ithaca, NY) around the left carotid artery.

Immediately after the animals were killed, the lungs were removed and degassed in a vacuum jar. Quasi static pressure-volume relations were obtained by inflating the lungs manually to 30 cm H2O. Then the lungs were deflated in six steps of 30 sec each. In each step, the pressure in the lung was lowered 5 cm H2O. The volume of the lungs at the end of each step was noted. Quasi static compliance was calculated by dividing the volume at the maximal pressure of 30 cm H2O by that pressure (30 cm H2O) and by the weight of the animal. The stability index according to Gruenwald was calculated by two times the volume at a pressure of 5 cm H2O, added to the volume at a pressure of 10 cm H2O, and then divided by two times the maximal volume at 30 cm H2O (30). The expansion index according to Clements was calculated by dividing the volume at 5 cm H2O by the maximal volume at 30 cm H2O and multiplying this ratio by 100% (31).

Processing of the lungs. After the rabbits were killed, the lungs were removed, separated in lobes, which were further separated in central and peripheral parts, frozen by immersion in liquid nitrogen, and cut in 200 pieces. Each piece was weighed, and the radioactivity of 99mTc was counted in a gamma counter (Packard 5000, Packard Instrument Co., Inc., Meriden, CT).

Data analysis. All values are presented as means ± SEM unless stated otherwise. MABP and CBF are measured in mm Hg and mL/min, respectively. To overcome subject variation at t = 0 min these data are expressed as percentage change from their values at t = 0 min. Differences in means between groups were tested for significance by unpaired t test. Multiple comparison analysis of pretreatment and sequential posttreatment values within each group were assessed using analysis of variance with Dunnett's post hoc test. p < 0.05 was considered statistically significant. Statistical tests were carried out using INSTAT software (Graphpad Software, San Diego, CA).

The distribution of surfactant was expressed as normalized value for each lung piece. The normalized value was calculated by dividing the radioactivity per milligram of tissue for each lung piece (cpm/mg) by the average value of radioactivity per milligram of tissue of all the lung pieces for each rabbit (mean cpm/mg). These normalized values are grouped intervals of 0.1, which is 10% of the mean value (1.0). Lung pieces having normalized values of <0.10 or >2.0 were grouped at the extremes of the distribution intervals.

RESULTS

Physiologic measurements. All animals survived the total duration of the study, including the weaning. After surfactant instillation with interruption of ventilation, PaO2 increased quickly in 15 min from 8.7 ± 1.3 to 24.9 ± 6.4 kPa and maintained this level for 120 min (Fig. 1). After surfactant instillation without interruption, PaO2 rose quickly in 15 min from 8.4 ± 0.8 to 32.4 ± 4.3 kPa. During the weaning, initially, PaO2 rose further in both groups, and decreased when FiO2 was lowered. There were no statistically significant differences between the two groups.

Figure 1
figure 1

Sequential measurements of PaO2 divided by FiO2 and PaCO2 after surfactant instillation with or without interruption of ventilation in lung lavaged rabbits. All values are means and SEM for n = 6. *different from pretreatment values (t = 0), with p < 0.05. N.S., no significant differences between the two groups.

Full size image

After surfactant instillation with interruption of ventilation, PaCO2 decreased in 15 min from 4.69 ± 0.51 to 3.61 ± 0.26 kPa (Fig. 1). After instillation without interruption, PaCO2 increased from 5.06 ± 0.41 to 4.13 ± 0.23 kPa. During the weaning, initially, PaCO2 values increased and then decreased when FiO2 was lowered. There were no statistically significant differences between the two groups.

After instillation with interruption of ventilation, the dynamic compliance rose from 0.40 ± 0.05 to 0.51 ± 0.03 mL/cm H2O kg BW. After instillation without interruption, dynamic compliance increased from 0.46 ± 0.05 to 0.51 ± 0.03 mL/cm H2O kg BW. Two hours after surfactant instillation, dynamic compliance was 0.45 ± 0.02 mL/cm H2 kg BW after interruption and 0.45 ± 0.03 mL/cm H2 kg BW without interrupting ventilation. After the weaning, the dynamic compliances were 0.40 ± 0.03 and 0.39 ± 0.03 mL/cm H2O kg BW. There were no statistically significant differences in time or between the two groups.

After surfactant instillation with interruption of ventilation, the MABP first decreased slightly, followed by a sharp rise of 28 ± 6.6% in the first 30 s (Fig. 2). This sharp rise was followed by a decrease of 22 ± 5.3% in 8 min after instillation. After surfactant instillation without interruption, blood pressure decreased slightly, then increased quickly 16 ± 4.2%, and then decreased 40 ± 10% in 8 min. There were no statistical differences between the two modes of surfactant administration. The effects on the CBF were more pronounced. After instillation with interruption, CBF slightly decreased, followed by a pronounced increase of 45 ± 14% in 30 s. This increase was followed by a drop of max 64 ± 11% in 8 min. When interruption was avoided, the initial rise of 15 ± 4.9% tended to be less pronounced than with interruption. However, this difference between the two groups just didn't reach statistical significance (p = 0.06). After this initial rise, the CBF decreased to max 61 ± 11% 4 min after instillation. During the following hours, there were no statistically significant differences between the two groups.

Figure 2
figure 2

Continuous measurements of mean arterial blood pressure (MABP) and cerebral blood flow (CBF) in left carotid artery after surfactant instillation, with or without interruption of ventilation in lung lavaged rabbits. All values are expressed as mean percentage change from pretreatment values and 95% confidence intervals (gray area) for n = 6. No statistical differences between groups were found.

Full size image

The quasi static compliance after instillation with interruption of ventilation was 0.92 ± 0.09 mL/cm H2O kg BW and after instillation without disruption was 1.05 ± 0.04 mL/cm H2O kg BW. The stability index after instillation with interrupting ventilation was 0.91 ± 0.05 mL/mL and after instillation without interruption was 0.99 ± 0.02 mL/mL. The expansion index after instillation with interruption of ventilation was 56.8 ± 5.9% and after instillation without interruption was 60 ± 1.9%. There were no statistically significant differences between the two modes of surfactant administration.

Surfactant distribution. Both modes of surfactant administration resulted in nonuniform distribution patterns (Fig. 3), with many lung pieces in which large amounts of surfactant were deposited; 28 ± 1.3 lung pieces contained more than two times the mean amount of surfactant after instillation without interruption and 31 ± 4.2 lung pieces after instillation with interruption (not significantly different). A small number of lung pieces contained the mean amount of surfactant; 38 ± 6.1 lung pieces contained the mean ± 25% amount of surfactant after surfactant instillation without interrupting ventilation and 36 ± 2.2 lung pieces after instillation with interruption (not significantly different). However, after surfactant instillation with interruption of ventilation, more lung pieces contained less than 10% of the mean amount of surfactant (24 ± 5.8 lung pieces) than when ventilation was not interrupted (10 ± 1.1 lung pieces).

Figure 3
figure 3

Distribution histograms of surfactant after instillation with or without interruption of ventilation in lung lavaged rabbits. All bars represent the number of lung pieces within each distribution interval of 10%. All values are means and SEM for n = 6.

Full size image

In contrast, the lobar surfactant distribution was more uniform after instillation with interruption than after instillation without interruption (Fig. 4). After instillation with interruption, lobar distribution rates were not significantly different from the mean amount of surfactant deposition with a normalized value of 1.0. Whereas, after instillation without interrupting ventilation, lobar distribution rates were significantly different from 1.0 for the right upper lobe (1.68 ± 0.25), right middle lobe (0.37 ± 0.10), and left upper lobe (0.49 ± 0.04). In both groups, a significantly larger deposition rate was found in the central lung pieces than in the peripheral lung pieces. Moreover, significantly larger amounts of surfactant were deposited in the central lung pieces after instillation without interruption than after instillation with interruption of ventilation.

Figure 4
figure 4

Lobar distribution of surfactant after surfactant instillation with or without interruption of ventilation in lung lavaged rabbits. Lobes are right upper lobe (RUL), right middle lobe (RML), right lower lobe (RLL), left upper lobe (LUL), and left lower lobe (LLL), and right versus left lung and peripheral versus central lung pieces. All values are normalized values of surfactant and expressed as means and SEM for n = 6. *different from mean value 1.0 with p < 0.05. #significantly difference between two groups with p < 0.05.

Full size image

DISCUSSION

We have shown that avoiding interruption of ventilation tends to prevent the potential adverse effect of a rapid rise in CBF during instillation, whereas having no detrimental effect on respiratory function. Surfactant distribution tends to be slightly more uniform with less lung pieces containing very small amounts of surfactant, but, on the lobar level, distribution is less uniform.

We found that avoiding interruption of airway pressure during surfactant instillation did not affect gas exchange and elastic recoil characteristics of the lungs. PaO2, PaCO2, dynamic compliances, and isolated lung compliances were not significantly different.

Furthermore, we found that avoidance of interruption tended to improve uniformity of surfactant distribution with less lung pieces containing very small quantities of surfactant. This could imply that maintenance of PEEP enhances deposition of, at least, some surfactant in those areas that otherwise would not be reached by instilled surfactant. We expected that the peripheral lung pieces would especially benefit from this mechanism. However, this could not be confirmed by our data. In contrast, peripheral deposition rates even declined when interruption was avoided.

Our findings are in line with those of Merritt et al. (32). They reported an improvement in uniformity of distribution of KL4 when this surfactant was administered while maintaining PEEP in premature monkeys. Especially the right middle lobe and left lower lobe benefited with a higher deposition rate by this method. However, no distinction was made between central and peripheral surfactant deposition.

In apparent contrast, we found that lobar surfactant distribution was less uniform when PEEP was maintained during instillation. In the right middle lobes and the left upper lobes, less than the mean amount of surfactant was deposited, whereas in the right upper lobe, more than the mean amount was deposited. The same tendency for lobar distribution was found by other investigators after slow infusion of surfactant and after nebulization of surfactant (18,20). During these procedures, airway pressures were maintained also.

Several hypotheses have been put forward to explain distribution patterns of surfactant. First, regional differences in lung maturation of premature animals may lead to differences in lung injury and consequently in surfactant spreading. Second, gravity may lead to a preferential spreading of surfactant into the right upper lobe. Third, physical properties of the instilled liquid may influence its spreading over the airways. Davis et al. (33) showed in lung lavaged piglets that surfactant spreads more rapidly and more uniformly over the lungs than saline, which implies the importance of the physical properties of the instilled drug. Espinosa et al. (34) confirmed this in a model of surfactant spreading on a one-dimensional thin liquid layer. They concluded that surfactant spreading is driven primarily by surface tension gradients within the lungs and that endogenous surfactant enhances the spreading of exogenous surfactant.

We speculate that these interfacial forces promote the spreading of instilled surfactant to those alveoli that remain expanded when PEEP is maintained during instillation. In contrast, when ventilation is interrupted, the airway pressure drops and alveoli collapse. In collapsed lung parts, there are no interfacial forces like surface tension gradients and therefore, surfactant will not be deposited there. Furthermore, we speculate that these mechanisms especially influence the spreading of surfactant peripherally in the branching structure of the airways. More proximal, other variables like the position of the tube, speed of instillation, gravity, and airway pressures may influence the spreading in the larger airways. Nevertheless, the knowledge about variables that influence the spreading of surfactant is still far from complete.

Interestingly, we furthermore found a tendency that the initial quick rise in CBF during surfactant instillation is reduced by avoiding interruption of ventilation. It is suggested that the hemodynamic consequences of surfactant instillation are related to transient obstruction of the airways by the instilled liquid, followed by rapid improvement in ventilation and blood gasses, when surfactant spreads throughout the lungs (3–5). Consequently, changes in intrathoracic pressures influence the venous return of blood to the heart and cardiac output. Additionally, improvements in ventilation might be accompanied by hyperoxemia and hypocarbia, which has shown to correlate significantly with CBF (25,35,36). After the first 30 s, we found no differences in CBF between the two modes of surfactant instillation. Both groups showed a significant drop in CBF during 8 min that recovered during the 3 h thereafter. The latter is in concordance with the fact that, in this study, blood gasses and compliance indices were equally affected by both procedures.

Although many animal studies have been evaluating different modes of surfactant administration, it has not been fully elucidated yet which variables influence the spreading of surfactant. Although it can be concluded that a high volume of a natural surfactant instilled quickly in one or two doses, preferentially before starting mechanical ventilation, is most efficacious on lung function and uniformity of distribution but could be accompanied by detrimental hemodynamic side effects (15–19,37).

Nebulization of surfactant can be an attractive alternative in improving uniformity of surfactant distribution and preventing hemodynamic consequences, but technical innovations are warranted before it can be applied clinically (20–27). Slow infusion of surfactant may prevent the hemodynamic side effects but is less efficient on lung function and surfactant distribution (18,19). However, Saliba et al. (11) found that slow infusion of Exosurf in 14 neonates with severe RDS had no detrimental effects on gas exchange when compared with rapid instillation. This apparent discrepancy may be caused by differences in standardization, surfactant preparation, ventilator settings, and subject characteristics.

In conclusion, we found that avoiding interruption of ventilation during surfactant instillation tends to prevent the potential adverse effects of a rapid rise in CBF and tends to improve uniformity of surfactant distribution although peripheral deposition tended to be less, while, at the same time, having no detrimental effect on respiratory function. This implies that surfactant instillation without interruption of ventilation is equally efficient and possibly safer than instillation of surfactant with interruption of ventilation. Prospective clinical studies are warranted to evaluate this hypothesis.