INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pulmonary surfactant deficiency is the main cause of infant respiratory distress syndrome (IRDS) in preterm infants (1). Prenatal corticosteroids and exogenous surfactant have greatly reduced morbidity and mortality caused by IRDS. (2) However, some infants respond poorly or only temporarily to surfactant therapy. Information on the surfactant system in infants is very limited, mainly because radioactive isotopes cannot be used in humans. The available methods for assessing the surfactant system are serial measurements of the amount of phosphatidylcholine (PC) in tracheal aspirates (3), or data on surfactant kinetics can be obtained from the exponential disappearance of phosphatidylglycerol (PG) and sphingomyelin (Sph) from repetitive airway samples after the endotracheal administration of exogenous surfactant (4, 5). The latter method relies on the assumption that only exogenous and not endogenous surfactant contains PG or Sph, and that their metabolism is similar to that of PC, the most prominent lipid of pulmonary surfactant. These techniques do not give direct information on the metabolism of surfactant PC. Obtaining data on the pharmacokinetics of PC of exogenous surfactant in humans would be highly desirable for optimizing surfactant therapy (6).

Surfactant kinetics have been extensively studied with radioactive isotopes in cell cultures and in animals (7, 8). These studies showed that the clearance and catabolism of pulmonary surfactant depend both on age and on the species investigated (8). In rabbits and sheep, the half-lives of surfactant phospholipids are longer in newborn than in adult animals (8). In addition, PC half-lives in newborn rabbits and newborn sheep differ. The half-life of PC in the newborn lamb is about 6 d (9), in the newborn rabbit 3.5 d, and in the adult rabbit only 8 h (12, 13). Surfactant half-life also depends on the surfactant preparation used (14), the mode of ventilation (15), the use of certain drugs, such as -receptor agonists (16), and possibly on the underlying lung disease.

We have recently described a new method based on stable isotope technology that is suitable for the study of endogenous surfactant metabolism in humans. The method is based on the intravenous infusion of different surfactant precursors labeled with the stable, nonradioactive isotope ¹³C and measurement of the ¹³C enrichment of PC obtained from sequential tracheal aspirates (17, 18).

We here describe a novel and safe method for measuring the pharmacokinetics of exogenous surfactant in human infants with this new methodology. To accomplish this, we labeled exogenous surfactant with a PC tracer labeled with ¹³C (19).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients and Study Design

We studied eight newborn infants with IRDS, who were admitted to the Neonatal Intensive Care Unit of the Department of Pediatrics of the University of Padua and who required mechanical ventilation and received exogenous surfactant during the first 24 h of life. The study protocol was approved by the Ethics Committee of the Department of Pediatrics, University of Padova and written informed consent was obtained from at least one parent.

The diagnosis of IRDS was based on clinical data, chest radiograms (20), and the exclusion of biochemical and microbiologic signs of infection. Newborns with congenital malformations, sepsis, or renal or liver failure were excluded from the study. All infants whose clinical characteristics are reported in Table 1 received two doses of 100 mg/kg/dose of porcine surfactant (Curosurf; Chiesi Farmaceutici S.p.a., Parma, Italy) as rescue treatment for IRDS. A tracer dose of dipalmitoylphosphatidylcholine (DPPC), with both palmitic acids (PA) uniformly labeled with the stable isotope ¹³C (U-¹³C-PA), was added to the intratracheal doses of exogenous surfactant given to the subject infants for the treatment of IRDS. We used 5 mg/kg/dose of (U-¹³C-PA)-DPPC for labeling the exogenous surfactant, mixing this with exogenous surfactant as described by Ikegami and colleagues (21). The ¹³C used for labeling the PAs of DPPC came from Martek Biosciences, Columbia, MD.

The indication for surfactant treatment was a fraction of inspired oxygen (FI_O2) higher than 0.40 or a mean airway pressure () higher than 7.5 cm H₂O. Surfactant was administered as a bolus via a small catheter inserted through the endotracheal tube. After the procedure, the neonates were hand ventilated for 1 min and then reconnected to the mechanical ventilator at pretreatment settings. The mode of ventilation during the study period was standardized, with an inspiratory time of 0.3 to 0.5 s, initial respiratory rate of 50 to 65 breaths/min, and positive end-expiratory pressure (PEEP) of 3 to 4 cm H₂O. Peak inspiratory pressure was adjusted so that the Pa_O2 ranged from 50 to 70 mm Hg, the oxygen saturation (Sa_O2) exceeded 88% but was less than 96%, and the PCO₂ ranged from 40 to 50 mm Hg. Ventilatory parameters were recorded before the start of the study and subsequently every 12 h. The ventilatory index was calculated as follows: VI = (f ·  · FI_O2)/1,000, where f represents the ventilatory rate per minute and is the mean airway pressure. Tracheal aspirates of the infants enrolled in the study were collected before administration of the first dose of surfactant (time = 0), and at 3 h, 6 h, and then every 6 h up to 72 h, and thereafter every 12 h until extubation.

The airways were routinely suctioned with 0.5 ml of 0.9% saline injected through the endotracheal tube. To accomplish this, the neonate was ventilated by hand bagging for 30 s, after which tracheal secretions were collected through a Lukens trap. The tracheal aspirate was kept at 4° C for no longer than 3 h, and was brought to a final volume of 3 ml with 0.9% saline. This sample was gently vortexed for 1 min and centrifuged at 400 × g for 10 min. The supernatant was recovered and kept at 20° C until analysis. Fifty microliters of each dose of exogenous surfactant labeled with (U-¹³C-PA)-DPPC was stored at 20° C in order to determine the ¹³C enrichment of disaturated phosphatidylcholine (DSPC).

Analytical Methods

Lipids from tracheal aspirates and from exogenous surfactant were extracted according to the method of Bligh and Dyer (22), after addition of the internal standard heptadecanoylphosphatidylcholine. One third of the extract was oxidized with osmium tetroxide (23). Saturated PC (DSPC) was isolated from the lipid extract by thin layer chromatography (TLC) (24). The DSPC fatty acids were derivatized (25) as pentafluorobenzyl derivatives, extracted with hexane, and stored at 20° C. Half the TLC spot of exogenous surfactant DSPC was derivatized as methyl ester, and the amount of DSPC was measured by gas chromatography as previously described (26). Tracheal aspirates with visible blood were discarded.

The enrichment of (U-¹³C-PA)-DSPC in the tracheal aspirates was measured by gas chromatography-mass spectrometry in the negative ionization mode, and the results were expressed in atom percent excess (APE), which is equivalent to the specific activity used for radioactive tracers. The APE represents the increase in percentage of atoms of ¹³C above the level in the naturally occurring molecules.

Calculations

The half-life of (U-¹³C-PA)-DSPC was calculated by exponential curve-fitting at the final, monoexponential part of the downslope of the curve of enrichment values over time after administration of the second surfactant dose. The time interval between the first and the second dose did not allow calculation of the monoexponential curve-fit after the first dose (Figure 1).

View larger version (25K): [in this window] [in a new window]   Figure 1.   Isotopic enrichment of DSPC obtained from tracheal aspirates of eight preterm infants with IRDS and treated with exogenous surfactant (panels 1 to 8). On the y-axis the 13C-enrichment of (U-13C)-palmitate in surfactant DSPC is given in APE after normalization for the isotopic enrichment of DSPC palmitate in the dose of exogenous surfactant. The APE shows an exponential decay after the first (closed squares) and the second (closed circles) surfactant doses. Panel 9 depicts the mean values of the DSPC palmitate enrichment for the eight infants, together with the 95% confidence interval and the exponential regression line. In order to draw the data for all eight patients on the same graph, the isotopic enrichment of DSPC palmitate was normalized to the 13C enrichment of the administered surfactant. All of the first dose curves were assumed to start at time = 0, corresponding to the time of the first surfactant administration, with the second doses starting at the mean time of the second surfactant administrations.

The size of the surfactant pool into which the exogenous surfactant phospholipids were distributed after the first and the second doses of surfactant (the "apparent surfactant pool size") was calculated on the basis of the Fick principle (4). Calculations were made by the least-squares method from the linear regression line representing the decay of the log-transformed (U-¹³C-PA)-DPPC APE over time. This was applied for both the first and the second surfactant doses. The APE at zero time was defined as the y-axis intercept of the linear regression curve of the log-transformed (U-¹³C-PA)-DPPC enrichment values (4, 5). The "apparent DSPC pool size" was equal to the amount of exogenous surfactant DSPC · exogenous surfactant (U- ¹³C-PA)-DSPC APE/calculated (U-¹³C-PA)-DSPC in lung effluent at the time of the first and the second surfactant doses, respectively. The "endogenous surfactant pool size" was calculated as the difference between the apparent surfactant pool size and the amount of exogenous surfactant administered.

The turnover rate of surfactant DSPC was calculated as 1/(1.44 × DSPC half-life), assuming that the surfactant pool size was constant after the second surfactant dose (mean postnatal age at the second dose: 35.5 ± 11.5 h; range: 20 to 55.5 h; Table 1)

Data are expressed as group mean ± SD. Statistical analyses were performed with Excel 97 software (Microsoft, Seattle, WA).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Clinical data are shown in Table 1. The birth weight of the infants was 1.4 ± 0.2 kg (mean ± SD), and all infants presented with at least Grade 2 IRDS and received the first surfactant dose within 8 h of birth. Surfactant replacement was given twice to all infants, with an interval of 32.7 ± 11.8 h between the first and the second surfactant doses. No infant showed any sign of infection during the study period.

Figure 1 (panels 1 to 8) shows the (U-¹³C-PA)-DPPC isotopic enrichment of DSPC obtained from tracheal aspirates of the eight preterm infants in the study. The mean values of the DSPC enrichments for the eight infants are shown collectively in panel 9. At 3 h after the first and second doses, all but one of the tracheal aspirates showed greater than expected (U-¹³C-PA)-DPPC enrichment on the basis of the linear regression analysis (data not shown). This suggests either a nonuniform distribution of labeled DPPC in the alveolar pool or inability of the immature lung to contribute a significant amount of surfactant to the alveolar pool during the first 3 h after dosing. Because of this, the 3-h time point was excluded from the calculation of surfactant pool size for all infants. All curves show a rapid decrease in enrichment after administration of the first and second doses of labeled surfactant. This rapid decrease in isotopic enrichment after surfactant administrations probably represents a distribution phenomenon. After this rapid decrease, clear monoexponential curve fitting was observed  36 h after the second dose of surfactant in all patients (Figure 1). Main results are given in Table 2. DSPC half-life, as calculated from the monoexponential part of the decay curve of the second surfactant doses for the eight preterm infants, was 34.2 ± 9.4 h (mean ± SD). Surfactant DSPC turnover was 0.5 ± 0.2 d¹. The DSPC pool sizes were 5.8 ± 6.1 mg/kg and 17.3 ± 13.6 mg/kg before the first and second surfactant doses, respectively (Table 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We describe a novel and safe method that is applicable to the study of surfactant kinetics in human infants who require endotracheal intubation. We used a DPPC tracer with the PA moiety uniformly labeled with the stable isotope ¹³C to trace exogenous surfactant, and we report for the first time on the direct measurement of the pharmacokinetics of exogenous surfactant DSPC in preterm infants with IRDS.

Unlike previous studies, in which indirect indicators for surfactant pharmacokinetics were used (4, 5), we were able to trace PC directly. We were able to follow the metabolism of this most abundant surfactant component by labeling DSPC surfactant molecules. The validity of the model of endotracheal injection of a tracer for the study of surfactant was demonstrated by Jacobs and coworkers with radioactive isotopes (13). Jacobs and coworkers injected a trace dose of [³H]choline-labeled surfactant mixed with synthetic [¹⁴C]DPPC intratracheally into 3-d-old rabbits, and demonstrated that the synthetic DPPC functioned metabolically in the same way as that administered in the form of natural surfactant. On the basis of these studies, we used (U-¹³C-PA)-DPPC instilled into the trachea in a mixture with exogenous surfactant to study exogenous surfactant metabolism in vivo in humans. The most obvious advantage of stable isotopes is that they are nonradioactive and thus not hazardous to human subjects. In fact, they can even be used during lactation or pregnancy (27).

As mentioned earlier, surfactant metabolism has been studied in preterm human infants through indirect methods involving the use other surfactant metabolic markers, such as PG or Sph, under the assumptions that there is no PG or Sph in preterm human lungs with RDS, and that a known and measurable amount of each of these phospholipids is present in the exogenous surfactant administered to infants (4, 5). A third important assumption is that the metabolic fate of PG and Sph is representative of that of surfactant PC. To the best of our knowledge, this has not yet been demonstrated. Surfactant half-life, apparent surfactant pool size, and surfactant turnover can be estimated on the basis of the assumptions that: (1) the distribution of exogenous surfactant is similar to the distribution of endogenous surfactant; (2) the phospholipid composition of surfactant in various pulmonary compartments is uniform; (3) the surfactant system is pulse labeled; (4) there is no endogenous synthesis of the marker; and (5) the pool size at a specified time after administration of exogenous surfactant is constant. These points were discussed by Hallman and coworkers (4), and are discussed individually as follows:

1. 1. The distribution of exogenous surfactant is similar to the distribution of endogenous surfactant. The distribution of different phospholipid classes after administration of exogenous surfactant has been measured in vivo in the tracheal aspirates of preterm infants who had received exogenous surfactant (4, 5). During the first few hours after surfactant administration, there were no detectable differences in phospholipid contents from those of the exogenous surfactant, suggesting an homogeneous distribution within the lungs.
2. 2. The phospholipid composition of surfactant in various compartments is uniform. According to animal studies done with radioactive tracers, there is a fast bidirectional flux of intact surfactant PC or DPPC between the alveolar space and the lamellar bodies (28, 29).
3. 3. The surfactant system is pulse labeled. The approach used in our study directly pulse labeled the surfactant surface-active phospholipids in the lungs.
4. 4. There is no endogenous synthesis of the marker. We feel that the (U-¹³C-PA)-DPPC used in our study comes close to being an ideal tracer. Unlike PG or Sph (U-¹³C-PA)- DPPC cannot be endogenously synthesized, and unlike tracers with low isotopic enrichments (6), (U-¹³C-PA)-DPPC does not overlap with any of the naturally occurring molecular species of PA during measurements.
5. 5. The pulmonary surfactant pool size at a specified time after exogenous surfactant remains constant. According to animal studies, the pool size of lavageable surfactant increases severalfold within a few hours after preterm delivery, after which the increase is more gradual (11, 30). No data are available on postnatal changes in the pool size of human surfactant, although it is likely to increase after birth in IRDS patients (3, 4). Data from preterm baboons receiving 100 mg/kg of exogenous surfactant at birth showed that the total lung DSPC pool size after 6 d of mechanical ventilation was 170 µmol/kg, and that a similar value (around 140 µmol/kg) could be calculated at birth by adding the baboons' endogenous DSPC pool size to the amount of DSPC administered with the exogenous surfactant (33).

From the decay curve of the enrichment of (U-¹³C-PA)- DPPC in serial tracheal aspirates, we calculated the endogenous DSPC pool size to be 5.8 ± 6.1 mg/kg before the first surfactant dose. Although we are aware that such estimates must be interpreted with caution, data obtained from lung lavage of infants who died from RDS and who did not receive exogenous surfactant (34) support our findings. Pool size estimates of 7.5 mg and 9 mg as measured with indirect methods were reported by Hallman and coworkers (4) and Griese and colleagues (5).

We also estimated the DSPC pool size before the second surfactant dose as being 17.3 ± 13.6 mg/kg, which was about twice as great as the value found before the first surfactant dose. Similar mean values have been found in baboons after exogenous surfactant administration. We intend to study the apparent DSPC pool size in a larger number of infants before each surfactant dose, in order to see whether or not the need for retreatment is associated with a low pool size. We believe that the large difference in pool size before the second surfactant dose given to the eight infants in our study reflects an important and clinically relevant difference among patients. This could be due to differences in endogenous surfactant synthesis among patients (V. P. Carnielli, unpublished observation) and/or to the different capacity of the immature and sick lung to retain exogenous surfactant. Further work, possibly combining the measurement of endogenous surfactant synthesis and the metabolism of exogenous surfactant, is necessary to answer this question.

The half-life of DSPC-PA in this study was 34.2 ± 9.4 h, which is in agreement with data obtained through indirect methods in two groups of preterm infants with RDS whose clinical characteristics were similar to those of the infants in our study (4, 5). These half-life values are longer than those for the adult rabbit (8 h), but shorter than those found in the newborn rabbit (3.5 d) (13).

In a previous study of preterm infants with RDS (17), we found the half-life of endogenous surfactant to be 113 ± 25 h. In that study we infused U-¹³C glucose intravenously as a tracer, and therefore labeled the entire surfactant pool (both the alveolar and lung surfactant pools). Among factors that could have caused longer surfactant half-lives in that particular study in comparison with the present one were that: (1) the continuous infusion of the tracer for 24 h could have caused a delay in the curve of enrichment versus time in comparison with a bolus administration; (2) when a label is infused into the systemic circulation, it is likely to be incorporated in several body organs and to be released at a later time; and (3) the half-life of the total surfactant pool may be longer than that of the alveolar pool. In the present study the label was given with the exogenous surfactant, with the result that the alveolar pool was labeled first and the total surfactant pool became enriched only after equilibration with the alveolar surfactant. Other factors, however, may be responsible for the difference between the findings in our first study of endogenous surfactant and those in the present study: It is plausible that the metabolism of exogenous animal surfactant given for the treatment of IRDS differs from that of the more physiologic, endogenously synthesized human surfactant. Other differences may be determined by the commercial preparation of surfactant (5), and possibly by the type of ventilation used for a particular patient. Despite these variables, which may explain the difference in surfactant half-life found in our first study and in the present one, we believe that intratracheal instillation of the label is a better approach for studying the pharmacokinetics of exogenous surfactant.

In conclusion, we describe here a new method for measurement of the kinetics of exogenous surfactant in vivo in humans. We measured the time course of disappearance of the stable isotope (U-¹³C-PA)-DPPC after the pulselike application of a labeled, exogenous (¹³C)-DPPC surfactant, using gas chromatography mass-spectrometry. In future studies, this method could lead to: (1) better understanding of the efficacy of therapies that enhance lung/surfactant maturation; (2) a stronger rationale for the dosing and timing of exogenous surfactant in different lung diseases; and (3) the ability to compare different surfactant preparations.