INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Respiratory distress syndrome (RDS) remains one of the leading causes of death in preterm infants, and insufficient surfactant production is one of the hallmarks of this disease (1, 2). Prenatal administration of corticosteroids and postnatal administration of exogenous surfactant have greatly reduced morbidity and mortality caused by RDS (3, 4). Some infants, however, respond poorly or only temporarily to surfactant therapy. The different clinical responses to exogenous surfactant may reflect differences in the disease process that causes the respiratory failure, but individual differences are also to be expected in the pharmacokinetics of exogenous surfactant (5) and in the postnatal surge of endogenous surfactant synthesis. Information on the timing and magnitude of the endogenous surfactant synthesis is needed. Although determination of the amount of surfactant in sequential tracheal aspirates provides indirect data on surfactant synthesis (6, 7), the variability from sample to sample is very large and little information can be obtained on synthesis and catabolism. In animals, these processes have been studied with radioactive isotopes (8), but this approach is not ethically acceptable in humans.

We have reported on the feasibility of measuring synthesis and turnover of pulmonary surfactant in humans using safe stable isotopes (12). We have used two techniques to date: in our first study we used carbon-13-labeled glucose (13), and in a second one we infused ¹³C-labeled fatty acids (FA) (14) as metabolic precursors for the biosynthesis of surfactant phosphatidylcholine (PC). Both studies showed that these techniques were suitable for measuring surfactant synthesis, but the patient populations of the two studies were different, comprising, respectively, preterm infants with RDS treated with exogenous surfactant (13) and 1- to 12-mo-old infants with acute respiratory distress syndrome (ARDS) not treated with exogenous surfactant (14). In the first study we reported that the fractional synthesis rate (FSR) of surfactant PC palmitic acid (PC-PA) from plasma glucose was 2.7 ± 1.3% per day and in the second one that FSR from plasma free fatty acids (FFA) were much higher, at approximately 33% per day for PA and 50% per day for linoleic acid (LLA). Whether these differences were due to the type of precursor molecule used or to the different patient populations is unclear, though Martini and coworkers have recently shown in adult pigs that plasma palmitate is a better precursor for surfactant PC than plasma glucose (15).

In the present study, in preterm infants with RDS we infused carbon-13-labeled PA and LLA to determine (1) if data on surfactant kinetics could be obtained even in extremely low-birth-weight preterm infants by an intravenous infusion of labeled FA, (2) if surfactant synthesis and catabolism calculated using either PA or LLA would yield different results, (3) if the administration of exogenous surfactant affects endogenous surfactant kinetics, and (4) if surfactant kinetics from plasma lipid precursors would differ from that obtained in our previous study using intravenously administered glucose in a similar population.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients and Study Design

Surfactant kinetics was studied in 11 preterm infants whose clinical characteristics are reported in Table 1. All patients were admitted to the Neonatal Intensive Care Unit of the Department of Pediatrics, University of Padua. Inclusion criteria were: (1) gestational age between 25 and 29 wk; (2) respiratory failure requiring endotracheal intubation for an estimated length of time of at least 48 h; (3) arterial and venous lines placed for clinical monitoring; (4) written informed consent obtained from the parents. Exclusion criteria were congenital infections and chromosomal abnormalities. Exogenous surfactant (Curosurf; Chiesi Farmaceutici S.p.a, Parma, Italy) was administered endotracheally at a dose of 100 mg/kg, if the mean airway pressure exceeded 7.5 cm of water or if the fraction of inspiratory oxygen (FI_O2) was higher than 0.40. Infants received a second dose if the same criteria were met after the first dose.

All patients received a 24-h constant intravenous infusion of labeled fatty acids [as previously described (14, 16)] while they were receiving a lipid-free intravenous infusion. The isotope infusion started in all patients within 16 h from birth and was terminated before 40 h of life. The start of the study (t = 0) was defined by the start of the infusion. In those patients who received exogenous surfactant, the infusion was started at the time of the first surfactant administration. [U-¹³C]PA and [U-¹³C]LLA (Martek, Columbia, MD) were bound to human albumin and infused intravenously at a constant rate by a high-precision, calibrated syringe pump (M22; Harvard Apparatus, Inc., Natick, MA). Chemical and isotopic purity was confirmed by gas chromatography-mass spectrometry (GC-MS). We used a central venous line (umbilical) for tracer infusion and an arterial line for blood sampling.

Blood (0.6 ml) was drawn at time 0:00, 5:30, 6:00, 12:00, 18:00, and 24:00 h from the start of the study to determine the isotopic enrichment of PA and LLA in plasma FFA. The blood drawn was placed in tubes containing ethylenediaminetetraacetic acid (EDTA) and immediately centrifuged at 1,300 × g. After separation plasma was stored in tubes containing pyrogallol as antioxidant at 20° C until analysis.

Tracheal aspirates were obtained before the start of the infusion, at 3 and 6 h, and every 6 h thereafter until 72 h. Subsequently, the samples were collected every 12 h until extubation or until study day 10. Tracheal aspirates were performed as follows: 20 s after 0.5 ml normal saline was injected into the endotracheal tube, and after a few manual breaths with a balloon, the suction was done with the tip of the suctioning catheter beyond the tip of the endotracheal tube. All tracheal aspirates were brought with saline solution to a final volume of 3 ml. The sample was then gently vortexed and centrifuged at 150 × g for 10 min. The supernatant was stored at 20° C until analysis.

Analytical Procedure

Plasma lipids were processed as previously described (16, 17). Plasma was delipidated with chloroform and methanol, according to Folch and coworkers (18). Lipid classes were separated by thin-layer chromatography, and their FA derivatized as methyl esters. The separation and identification of FA methyl esters from the plasma lipid classes were performed by capillary gas chromatography (GC) (16, 17).

Lipids were extracted from tracheal aspirates according to Bligh and Dyer (19), and surfactant PC was isolated by thin-layer chromatography (20). The PC was derivatized (21) and the fatty acid methyl esters were extracted with hexane and stored at 20° C. Tracheal aspirates containing visible blood were not analyzed. Concentration as well as isotopic enrichments of PA and LLA of the intravenous albumin solutions were measured for each individual patient by GC and GC-MS, by the same method used for plasma analysis.

Determination of the Isotopic Enrichment

The isotopic enrichment of plasma FFA and surfactant PC, PA, and LLA was carried out with a Fisons MD 800 (Fisons, Rodano, Milan, Italy) gas chromatograph quadrupole mass spectrometer using the negative chemical ionization mode with methane (5, 16). Selective ion monitoring of both FA was carried out at m/z 255-m/z 271 for natural PA and [U¹³C]PA and m/z 279-m/z 297 for natural LLA and [U¹³C]LLA. The plasma concentrations of free PA and LLA were measured by conventional GC. Each sample was determined in duplicate (16).

The enrichment was expressed as atom percent excess (APE), which represents the increase in ¹³C-labeled molecules (tracer molecules) above the baseline enrichment (before isotope infusion).

Calculations

Animal studies have shown that recycling is much faster than the de novo synthesis and clearance of surfactant (8, 11). Accordingly, for our calculation, tissue-bound and alveolar surfactant was regarded as one pool. The following surfactant PC kinetic parameters were calculated by using both PA and LLA: Secretion time (ST) was defined as the time lag between the start of the [U-¹³C]PA and [U-¹³C]LLA infusion and the appearance of the respective labeled FA in surfactant PC. ST was calculated by extrapolating the regression line for the ascending part of the enrichment curve to the baseline enrichment (13, 22). FSR of PC-PA and PC-LLA was defined as the percentage of the total surfactant PC pool synthesized from the respective plasma FA per day. It was calculated by dividing the slope of the linear increase of PC enrichment by the plasma steady-state enrichment of the respective FA (13, 23). Half-life (HL) of PC was calculated from the final, decreasing, and monoexponential portion of the enrichment versus time curve. Peak time (PK) was the time of maximal enrichment of surfactant PC calculated from the start of the infusion.

Data Analysis

Data are presented as individual values and group means (± SD) (range). Comparisons of groups were done by the Mann-Whitney or Wilcoxon tests for independent and related samples respectively. The level of significance accepted was p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The clinical characteristics of the study patients are shown in Table 1. Six of the 11 infants (Patients 1 through 6) had moderate or severe RDS and required exogenous surfactant, and five had mild or no respiratory disease and were not treated because they did not meet treatment criteria. Among the six treated infants all but one (Patient 5) required two doses of exogenous surfactant and all showed good clinical response to it.

In all infants, the isotopic enrichments of plasma FFA reached steady state from time 5:30 to 24 h during the isotope infusion. The slope of the linear regression of the enrichment values versus time was not significantly different from zero. Plasma FFA, free plasma PA, and free plasma LLA concentrations were stable during the study period (not shown).

The mole percent values of PA and LLA in surfactant PC were 62.0 ± 5.5 and 4.2 ± 1.9 mole% (mean ± SD), and they were quite constant (variability of less than 15%) in the individual patients during the study period. No significant changes due to surfactant treatment were measured.

A significant incorporation of the intravenously infused labeled PA and LLA was measurable in the surfactant PC from all patients. Figure 1 shows the time curves of the ¹³C enrichment of PC-PA and PC-LLA. Panel a depicts data from study patient 10 who did not receive surfactant treatment; panel b shows data from Patient 6 who received two doses of exogenous surfactant.

View larger version (17K): [in this window] [in a new window]   Figure 1.   Isotopic enrichments above baseline (atom percent excess, APE) of PA and LLA incorporated in surfactant PC after a 24-h intravenous infusion of [U-13C]PA and [U-13C]LLA in a preterm infant with minimal RDS (Patient 10) who did not receive exogenous surfactant ( panel a) and in a patient with mild RDS (Patient 6) who received two doses of exogenous surfactant ( panel b).

Kinetic data calculated for each individual patient are shown in Table 2. FSR, ST, PK, and HL were calculated using both PA and LLA. Surfactant PC-PA and PC-LLA HL could be reliably calculated for only eight infants (Patients 1-6, 10, 11), as three patients (Patients 7-9) were extubated shortly after 48 h from the start of the study. FSR for PA was 12.1 ± 7.7% per day, approximately half of that of LLA which was 24.6 ± 14.5% per day, p = 0.018. 

ST for PA and LLA were similar: 19.9 ± 15.3 and 16.4 ± 17.9 h, respectively. PK of surfactant PC-LA and of PC-LLA were reached after 71 ± 26 h and 85 ± 27 h from the start of infusion, for PA and LLA respectively. PC-PA had a much longer HL than PC-LLA (94.7 ± 18.8 h versus 46.6 ± 32.6 h, p = 0.028); however, large individual differences were found (ranges: 73.0 to 128.4 h for PC-PA and 18.4 to 100.5 h for PC-LLA).

When comparing the infants who did receive surfactant with those who did not (Table 2, bottom), we found that the administration of exogenous surfactant did not affect the calculation of FSR with either PA or LLA. FSR of PC-PA was 12.2 ± 9.8 versus 12.1 ± 5.4% h, treated versus nontreated infants and that of PC-LLA was 22.1 ± 16.6 versus 23.5 ± 17.1% h, treated versus nontreated infants. ST and PK were, however, significantly prolonged by surfactant treatment (Table 2).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We have recently shown that an intravenous infusion of labeled FA is a suitable method for measuring endogenous surfactant synthesis in young infants who did not receive exogenous surfactant (14). The first question we asked in the present study was if such a method could also be applicable to study small preterm infants with RDS who are often treated with exogenous surfactant. In this respect, we wanted to demonstrate if the preterm immature lung is capable of using lipid precursors for surfactant synthesis as compared with glucose, which is believed to be the preferred substrate during fetal life (24), and if reliable isotopic enrichment curves could be obtained despite the use of exogenous surfactant. We were able to measure a clear increase in the isotopic enrichment of the FA of PC from tracheal aspirates in all study infants, both treated and nontreated, with up to two doses of 100 mg/kg of exogenous surfactant. This is an important finding because we feared that the administration of exogenous surfactant, especially in conjunction with conditions of reduced synthesis, could have caused a too low isotopic enrichment for our detection system. By infusing labeled PA, which is the most abundant FA in surfactant PC, we intended to measure the surfactant PC-PA synthesized and secreted into the alveolar space by the type-II cells using plasma free PA as its metabolic precursor.

A limitation inherent in the use of PA as metabolic precursor for surfactant PC synthesis is that an unknown amount of the surfactant PC-PA is synthesized through de novo lipogenesis from glucose and other metabolic sources, thereby diluting the labeled PA incorporated into surfactant PC (13, 15, 25). A second limitation is that PA is concentrated in surfactant PC by a deacylation and reacylation pathway within the type II cell lamellar bodies (29). Martini and coworkers speculated that this pathway in adult pig is responsible for the much higher percentages of PA in surfactant PC ( 70%) than of plasma FFA ( 30%) (15). The rationale for infusing labeled LLA together with PA was that LLA is an essential FA and cannot be endogenously synthesized, and likely it is not concentrated within the lamellar bodies. Under the assumption that LLA is not discriminated in comparison with other FA during PC synthesis (29, 30) we believe that, because of its essentiality, LLA estimates total surfactant PC synthesis. As expected and as found in our previous work, we found that surfactant FSR and HL calculated with PA and LLA yielded different results (14). FSR of PC-PA was approximately half of that of PC-LLA (12.1 ± 7.7% versus 22.7 ± 15.9% per day). In our study infants we can reasonably assume a steady-state condition because we measured relatively constant amounts of PC-LLA and of PC-PA from tracheal aspirates during the study period in either the surfactant-treated or non-surfactant-treated infants, and because the FA composition of the exogenous surfactant, used in the present study, is rather similar to that of the infants' PCs.

During steady state, the rate of appearance equals the rate of disappearance, and it is plausible that the faster FSR of PC-LLA (synthesis) is associated with a shorter HL (catabolism). We do not have a clear biologic explanation for the difference in turnover between PA and LLA because most of the animal work has focused on the metabolism of disaturated PC and little attention has been paid to the metabolism of polyunsaturated FA (29, 30). Long HL of PC-PA could result from a reduced catabolism of PA containing PC molecules or from "active" PA recycling, which is known to exhibit a strong preference for saturated FA (9).

Considering only the patients who did not receive exogenous surfactant, STs were 7.7 ± 1.0 h for PA and 4.9 ± 3.4 h for LLA; these values are comparable to those obtained in critically ill infants studied with the same method (ST for PA 8.7 ± 8.9 h and for LLA 10 ± 7.2 h) (14). Although the limited number of infants in study does not allow for any statistical evaluation, it is unlikely that ST is markedly different between these two FA. Slightly shorter times (5 h) than those found by us have been reported in 10-d-old preterm ventilated lambs after an intravenous injection of [³H]PA (8). Similar ST for PA and LLA does suggest a similar processing during de novo PC synthesis, in agreement with the finding of Cogo and coworkers (14) and in vitro studies (31).

PK times in preterm infants with RDS were reached at 71 ± 25 h and 85 ± 27 h for PA and LLA, respectively (all patients). We found a comparable value of 70 ± 18 h for PC-PA, when infusing labeled glucose in preterm infants with RDS who were treated with exogenous surfactant (13). Shorter PK times were found in the infants who did not receive exogenous surfactant in comparison with the treated ones (54 versus 83 h and 61 versus 97 h for PA and LLA respectively). Furthermore, PK values of the non-surfactant-treated preterm infants were not markedly different from those obtained in critically ill 1- to 12-mo-old infants (PK-PA 49 ± 9 versus 54 ± 12 and PK-LLA 46 ± 19 versus 61 ± 10 h for critically ill infants and preterm infants respectively) (14). PK values in our preterm infants study were also not different from those measured in newborn lambs and rabbits (35 to 60 h after bolus of radiolabeled PA) (8, 32, 33). Patients treated with exogenous surfactant exhibited longer ST and delayed PK times. ST may be longer in immature lungs (slow synthesis) but is prolonged in the event of a large amount of exogenous surfactant in the alveoli. This will result in dilution of the label and therefore a reliable kinetic calculation cannot be made. Panel b of Figure 1 clearly illustrates that the second dose of exogenous surfactant caused a cessation in the increasing enrichment which was detectable already after 12 h. As ST could not be reliably measured from these points, it was calculated after the second surfactant dose. Longer ST in infants treated with exogenous surfactant were also found in our first study using labeled glucose (13).

FSR of PC-PA in the six infants who were treated with exogenous surfactant was 12.2 ± 9.8%/day, and it was 12.1 ± 5.4%/day in the five infants who had milder or no respiratory distress and did not require exogenous surfactant; likewise FSR of PC-LLA was 22.1 ± 16.6 and 27.6 ± 12.6%/day for the surfactant-treated and the nontreated groups. These remarkably similar mean values suggest that in our study surfactant treatment did not affect the FSR.

FSR represent the percentage of incorporation of single precursor (PA for example) into an unknown amount of product (surfactant PC-PA pool size) and over a given period. If we assume identical surfactant pool sizes between surfactant-treated and non-surfactant-treated infants, then similar absolute synthesis rates are to be expected. If pool sizes were different, absolute synthesis would differ proportionally to the difference in pool size at any given FSR. The amount of PC recovered from tracheal aspirates tended to be higher in the nontreated group than in the treated one (data not shown). Although the amount of PC recovered from tracheal aspirates may not represent a reliable estimate of surfactant pool size, this finding supports our belief that pool sizes were not markedly different in treated infants with mild to severe RDS in comparison with the untreated babies with minimal or no RDS. The somewhat lower amount of PC from tracheal aspirates in infants who were treated with exogenous surfactant should be confirmed in a larger number of subjects. We are currently measuring pulmonary surfactant pool size in vivo in healthy and diseased lungs by stable isotope probes (5).

The FSR from plasma PA and LLA found in this study were higher than those measured from plasma glucose in our previous work in a comparable population of preterm infants with RDS. In the study of Bunt and coworkers (13), the rate of PC-PA synthesis from plasma glucose was 2.7% per day, which is markedly lower than the 12% per day for PA and the 22.7% for LLA of the present study. This finding supports the hypothesis that glucose may not be the main precursor substrate for surfactant PC in the small preterm infant. Animal work in adult pigs also demonstrates that PA is a much better precursor than plasma glucose (15).

We measured surfactant PC-PA HL to be 95 ± 19 h, with no apparent effect by the administration of exogenous surfactant (Table 2). Individual values ranged, however, between 3 and 5 d (72 to 128 h). Because of the large variability and the limited number of infants studied, we did not attempt any correlation with the clinical status of the infants. Yet, our data indicate a slow turnover of endogenous surfactant in preterm infants, in agreement with the results of Bunt and coworkers (13). A long HL of surfactant PC is also compatible with findings in term newborn sheep and rabbits (8, 34).

In a very recent study in preterm infants with RDS (5) we measured HL of exogenous surfactant by labeling the exogenous surfactant with stable isotopes and obtained a mean value of 34 h, which is in agreement with the estimates of Hallman and coworkers (6) and Griese and coworkers (35). Longer HL of endogenous surfactant may be due to a combination of several factors: (1) compared with a bolus administration (used to trace exogenous surfactant), the continuous infusion of the tracer for 24 h, as used by us to trace endogenous surfactant, can cause a delay in the enrichment versus time curve; (2) a tracer infused in the systemic circulation is likely incorporated into body organs and then released at a later time, causing a prolonged incorporation into the lungs; (3) exogenous animal surfactant and endogenous human surfactant may have a different metabolism (35).

In summary, in the present study we demonstrate the following (1) Surfactant kinetics can be studied in preterm infants with RDS by an intravenous infusion of stable isotope labeled FA. (2) Surfactant FSR and HL calculated with PA and LLA yield different results that are in agreement with available animal work. (3) The administration of exogenous surfactant does not affect the calculation of FSR but it determines longer ST and delayed PK times. (4) We found in this study that the FSR of surfactant PC-PA and PC-LLA from plasma PA and LLA are higher than those measured from plasma glucose in our previous work in a comparable population of preterm infants with RDS. The latter finding possibly indicates that plasma FFA are a better precursor for surfactant synthesis than plasma glucose even in the small preterm infant during the first hours of life.