Early growth failure of extremely preterm infants is a risk factor for diseases including chronic lung disease (1) and retinopathy of prematurity (2), and has long-term negative effects on childhood growth and neurodevelopment (3). Providing adequate nutrition to such infants, so as to achieve a body composition and growth pattern similar to those of a normal fetus of the same postmenstrual age (4), remains challenging. Especially, certain nutritional strategies seek to restrict fluid intake (5) in the presence of a patent ductus arteriosus or a respiratory compromise that simultaneously increases metabolic demand (1). Extended intravenous parenteral nutrition is associated with the risks of catheter-related sepsis, thrombosis, and liver damage; however, such nutrition is essential in case of feeding difficulties. Pancreatic exocrine functions are not fully developed in preterm infants (6) and the maturation of such functions is not well understood.
The fecal concentration of human pancreatic-specific elastase 1 (PSE1) has been used to estimate the level of exocrine pancreatic function. A low fecal PSE1 level is considered to be a highly sensitive and specific marker of exocrine pancreatic deficiency (7). Fecal PSE1 levels are low in meconium and feces excreted during the first days of life (8,9) but increase steadily thereafter (8–11), attaining the concentrations encountered in children and adults (>200 μg/g) within the first week of life. Fecal PSE1 levels have consistently been reported to be lower in preterm, compared with term infants (9,11).
Fecal PSE1 levels have been used to diagnose the exocrine pancreatic insufficiency that causes malabsorption and nutrient deficiencies. PSE1 concentrations of >200 μg/g are considered to reflect normal pancreatic function. Pancreatic insufficiency is assumed if PSE1 levels are <100 μg/g, whereas concentrations of 100 to 200 μg/g are thought to indicate mild-to-moderate impairment of pancreatic function (12). Impairments in exocrine pancreatic function can be countered by administration of porcine pancreatic enzyme preparations protected within acid-resistant pH-sensitive microspheres; such exogenous pancreatic enzymes are routinely given to patients diagnosed as having pancreatic exocrine deficiencies including chronic pancreatitis and cystic fibrosis (13).
In the present study, we explored the longitudinal development of PSE1 levels in a cohort of infants of very low and extremely low birth weight. We correlated these data with the ability of the infants to gain weight, and any potential need for enzyme supplementation. We hypothesized that very-low-birth weight infants would exhibit low levels of PSE1 because pancreatic function was immature; that infants with low PSE1 levels would show only poor weight gain even when fed a diet that was calorically adequate; and that pancreatic enzyme replacement would restore growth to the intrauterine trajectory in very-low-birth weight infants with low fecal PSE1 concentrations receiving full enteral nutrition.
The present retrospective analysis considered data on all surviving preterm newborn infants of gestational age <32 weeks and birth weight <1250 g who were sequentially admitted to our neonatal intensive care units for a 1-year period (April 16, 2009–April 15, 2010). Data on 25 infants who died because of extreme prematurity or the existence of perinatal complications were not included in the analysis. One included infant was later diagnosed as having trisomy 21. No infant was diagnosed as having an inborn metabolic disorder.
Infants were treated using our local standard operating procedures; the feeding regime aimed to elevate the growth rate to recognized fetal growth percentile values (14). Fluids were commenced at 70 mL · kg−1 · day−1 and were increased stepwise to 150 to 180 mL · kg−1 · day−1 during 7 days. Parenteral nutrition, including amino acids and lipids, commenced on the first day of life and was maintained until the level of enteral feeding attained 130 to 150 mL/kg. Enteral nutrition was commenced on the first day of life, at approximately 10 mL · kg−1 · day−1, and was increased daily by 10 mL · kg−1 · day−1 whenever possible. Human milk, preferably the mother's own milk, was used. Mother's milk was pasteurized if the mother tested positive for immunoglobulin G antibodies to cytomegalovirus, until infants attained 32 weeks of gestational age. Human donor milk was always pasteurized. The protein content of milk was increased by addition of bovine milk protein–based fortifiers when the volume of milk given daily exceeded 90 mL. The daily target fluid volume (combined enteral and parenteral intake) was 150 to 180 mL in infants ages 7 days or more. The target weight of combined parenteral amino acid and enteral protein intake were 4.0 to 4.5 g · kg−1 · day−1 by day 5 of life in infants <1000 g birth weight and 3.5 to 4.0 g · kg−1 · day−1 in infants >1000 g birth weight. The target volume was adjusted upward when weight gain was inadequate. In infants with significant pulmonary morbidity, and in those experiencing hemodynamically significant patent ductus arteriosus, the daily fluid intake was restricted to a maximum of 150 mL/kg. The target energy intake of 130 kcal · kg−1 · day−1 was increased to 150 kcal · kg−1 · day−1 in infants with persistent pulmonary morbidity and in those exhibiting growth failure.
Determination of PSE1 Levels
Fecal PSE1 concentrations were determined, at the discretion of the attending physician, in serial stool samples obtained from very-low-birth weight infants. PSE1 levels were assayed every 2 weeks using an enzyme-linked immunosorbent assay (ScheBo; Biotech AG, Gießen, Germany). Results were expressed as microgram PSE1 per gram of stool.
Enteral Pancreatic Enzyme Supplementation
Infants with low PSE1 levels (<200 μg/g), in whom weight gain did not attain appropriate fetal growth percentiles (14) despite provision of adequate energy intake, received enteric-coated purified porcine pancreatic enzymes (Kreon; Solvay Arzneimittel, Hannover, Germany) at the discretion of the attending physician. The levels given were 15,000 to 20,000 U · kg−1 · day−1, divided into 8 to 12 doses, depending on the feeding regime used. Treatment commenced when laboratory PSE1 results became available and was usually discontinued when PSE1 levels exceeded >200 μg/g and/or when infants showed satisfactory weight gain. Coated drug pellets were powdered with a mortar and pestle before administration to allow the preparation to pass, with milk, through the gastroenteric tube.
Medical charts of all infants were reviewed; the charts contained data on weight gain and the type(s) of nutrition given (parenteral nutrition only, mixed parenteral and enteral nutrition, or enteral nutrition only) (Table 1). Energy intake (kcal · kg−1 · day−1) was calculated every 7 days, and relative weight gain was calculated as the weekly weight gain per average level of daily energy intake.
The present study was approved by our local institutional review board (Ethikkommission der Charité, approval no. EA2/139/09) and our data-handling safety supervisor.
Data were analyzed with the aid of appropriate nonparametric tests (the Friedman, Kruskal-Wallis, and Spearman tests.) A P value <0.05 was considered to indicate statistical significance. All analyses were performed employing SPSS 19.0 software (SPSS Inc, Chicago, IL).
Between April 16, 2009 and April 15, 2010, a total of 96 preterm infants of birth weight <1250 g and of gestational age <32 weeks were born in our tertiary care center; 81 of these infants (84%) were admitted to our neonatal intensive care unit and 71 infants survived long term. Of these 71 infants, PSE1 levels were measured in the stools of 69 at least once, and these data were included in the analysis. Thirty infants were girls (44%) and 31 (45%) were nonsingleton births. The median (range) birth weight was 857 g (382–1240 g) and the median (range) gestational age was 273/7 weeks (233/7–311/7 weeks). Forty-six infants (67%) were <28 weeks of gestational age and 23 (33%) were ≥28 weeks of gestational age. Gestational age and birth weight data are shown in Figure 1.
A total of 168 stool samples were collected from the 69 neonates, commencing on day 7 of life and continuing to the time of discharge from our neonatal care unit. Because of the number of assay data collected, we limited our statistical analysis to results obtained at 3 timepoints. When PSE1 levels were measured at 2 (N = 56), 4 (N = 46), and 6 (N = 23) weeks of age, the PSE1 concentration was found to increase from week 2 (median [interquartile range] 84 [48–187] μg/g) to week 4 (164 [87–251] μg/g; P < 0.001), but not thereafter (169 [82–298] μg/g at week 6). Such a maturational increase in PSE1 levels was observed only in infants of gestational age <28 weeks (P < 0.001). At 2 weeks, the PSE1 level was lower in infants <28 weeks of gestational age, compared with that in infants of gestational age ≥28 weeks (77 [43–110], N = 37 vs 165 [56–300] μg/g, N = 19; P = 0.019). This difference was less pronounced at 4 weeks (153 [77–226], N = 34 vs 230 [108–503] μg/g, N = 12; P = 0.070) and had disappeared by 6 weeks (163 [76–258], N = 18 vs 175 [85–418] μg/g, N = 5; P = 0.576) (Fig. 2).
Weight Gain and Effect of Addition of Pancreatic Enzymes
In infants receiving full enteral nutrition at 4 weeks of age (N = 57), PSE1 levels measured at this time were available for 40 and were correlated with the extent of weight gain per unit of energy intake (Rs = 0.431, P = 0.005). Weight gain per unit of energy intake was lower (P = 0.040) in infants with PSE1 levels <200 μg/g (0.110 [0.081–0.139] g/kcal, N = 25) than in those with PSE1 levels ≥200 μg/g (0.139 [0.117–0.157] g/kcal, N = 15). Exogenous pancreatic enzyme supplementation was commenced at a median age of 23 days (range 11–53 days) and extended for a further 29 days (6–54 days) in 30 infants with low PSE1 concentrations (<200 μg/g). We next examined data on infants on full enteral nutrition, in whom PSE1 concentrations were <200 μg/g at day 28 of life. Infants who did (N = 12) or did not (N = 13) receive exogenous pancreatic enzyme supplementation were of a similar gestational age at birth (25 [24–27] vs 26 [24–31] weeks; P > 0.1) and were also similar in terms of birth weight (683 [382–1010] vs 895 [618–1200] g; P > 0.1); however, such infants were of lower weight at day 28 of life (942 [600–1400] vs 1200 [930–1640] g; P = 0.019) than were infants in whom PSE1 levels were ≥200 μg/g. Notably, weight gain per unit of energy intake did not differ significantly between these 2 groups (0.105 [0.073–0.116] vs 0.124 [0.087–0.149] g/kcal; P > 0.1). More information on these patients is shown in Table 2.
In the present study, we have shown that preterm infants of extremely low gestational age have only limited exocrine pancreatic function during the first weeks of life, a period when vigorous growth is desirable. Our study extends previous findings (8–10) showing that PSE1 concentrations were low (<200 μg/g) in meconium of all infants studied, regardless of gestational age (9), and that fecal PSE1 concentrations increased more rapidly in term, as opposed to preterm, newborn infants (8,10). We found that in infants <28 weeks of gestational age, average fecal PSE1 concentrations remained <100 μg/g at 2 weeks of life but increased steadily thereafter. This may be a result of a positive effect of early enteric feeding; exposure of the duodenal mucosa to nutrients is important for stimulation of pancreatic enzyme secretion. Few data on fecal PSE1 concentrations have been reported previously, and in those data the absolute standard deviations were close to the means (9). This may explain discrepancies between the data of the cited report and those presented herein. We did not observe complete normalization of PSE1 levels during the first month of life in preterm infants of birth weight <1250 g and of gestational age <32 weeks, and average fecal PSE1 levels remained below the reported normal range of adults (>200 μg/g) in such infants at 4 and 6 weeks of life.
The association between weight gain in preterm infants on full enteric feeding and fecal PSE1 levels has not been investigated previously. We found that 4-week-old preterm infants on full enteral feeding, and who had low PSE1 levels (<200 μg/g), exhibited less weight gain per unit of energy intake, compared with infants in whom the PSE1 concentration was >200 μg/g. No correlation between the extent of fecal nitrogen loss and fecal PSE1 concentration has been observed in preterm infants with average PSE1 levels >200 μg/g (11). Although all newborn infants show transient low levels of PSE1, the duration of relative pancreatic insufficiency may extend for several weeks in extremely preterm infants.
In healthy adults, postprandial cumulative pancreatic enzyme output far exceeds the quantity required for digestion; 5% to 10% of normal enzyme output is thought to be sufficient to maintain the routine digestive process (13). Low-level exocrine pancreatic function may be more critical in newborn infants who secrete only low levels of gastric pepsin (15) and who have low pancreatic triglyceride lipase activity. These deficiencies are partially overcome by activity of the bile salt–stimulated lipase present in fresh human milk (16); however, this lipase is not present in formula, and the activity is lost upon pasteurization of banked human donor milk or pumped breast milk from cytomegalovirus-positive mothers (17,18).
Optimization of weight gain via exogenous enzyme replacement therapy in infants with low-level exocrine pancreatic function seems to be a logical and straightforward procedure, and no lower age limit has been imposed for use of the commercially available enzyme preparations licensed for treatment of exocrine pancreatic insufficiency; however, we found no significant beneficial effects of exogenous pancreatic enzymes when these enzymes were given to infants with PSE1 levels <200 μg/g and who showed poor weight gain. Several possible explanations may be advanced. Because the decision to commence enzyme replacement was at the discretion of the attending physician, enzyme replacement was more likely prescribed for infants exhibiting growth failure. This means that potential beneficial effects may be concealed by selection bias. Furthermore, inactivation of the enzymes may have rendered supplementation ineffective. The enzymes were administered via a feeding tube; the enzyme pellets were ground in a mortar and pestle because the pellet diameter was too great to allow passage through the tube. Such treatment damages the enteric coating, rendering the enzymes susceptible to inactivation by gastric acids. It appears likely that the enzymes may have been inactivated in this manner. There are 2 approaches recommended to circumvent this problem. First, the enzyme dose may be augmented to compensate for the loss. In clinical practice, this approach is limited by blockage of the feeding tube. Second, the gastric pH may be increased by treatment with proton pump inhibitors or histamine blockers, or gavage with an alkaline solution (19). It may be worthwhile to test these approaches in term infants; however, such methods may not be safe in extremely preterm infants; any increase in gastric pH may elevate the incidences of nosocomial infections and necrotizing enterocolitis (20,21).
The analysis presented here, especially our observations on pancreatin substitution, has several limitations. First, our work was entirely retrospective in nature and descriptive in character. We did not follow a treatment protocol characteristic of a randomized controlled trial. Especially, administration of pancreatic enzymes was initiated at the discretion of attending physicians. Therefore, both the time of initiation and the duration of treatment varied within wide ranges. Second, because of the limited dataset, we conducted statistical analysis on information obtained at only selected timepoints. Thus, we cannot draw any firm conclusions on an association between the growth rates of infants with low PSE1 levels and administration of exogenous pancreatic enzymes.
In summary, the data suggest that exocrine pancreatic function matures more slowly in preterm infants of extremely low gestational age, compared with infants that are not as preterm, or term infants. The resulting exocrine pancreatic insufficiency is transient but appears to significantly compromise weight gain in infants that receive only enteric food. Attempts to address this issue using exogenous pancreatic enzyme replacement therapy have not yet been successful. Future work on enzyme replacement will need to address the problems (described above) associated with enzyme administration, including passage of coated pellets through the gavage tube.
We thank Boris Metze for statistical advice and help with graphics.
1. Biniwale MA, Ehrenkranz RA. The role of nutrition in the prevention and management of bronchopulmonary dysplasia. Semin Perinatol
2. Hellstrom A, Hard AL, Engström E, et al. Early weight gain predicts retinopathy in preterm infants: new, simple, efficient approach to screening. Pediatrics
3. Casey PH. Growth of low birth weight preterm children. Semin Perinatol
4. Agostoni C, Buonocore G, Carnielli VP, et al. Enteral nutrient supply for preterm infants: commentary from the European Society of Paediatric Gastroenterology, Hepatology and Nutrition Committee on Nutrition. J Pediatr Gastroenterol Nutr
5. Bell EF, Acarregui MJ. Restricted versus liberal water intake for preventing morbidity and mortality in preterm infants. Cochrane Database Syst Rev
6. Manson WG, Weaver LT. Fat digestion in the neonate. Arch Dis Child Fetal Neonatal Ed
7. Soldan W, Henker J, Sprössig C. Sensitivity and specificity of quantitative determination of pancreatic elastase 1 in feces of children. J Pediatr Gastroenterol Nutr
8. Nissler K, von Katte I, Huebner A, et al. Pancreatic elastase 1 in feces of preterm and term infants. J Pediatr Gastroenterol Nutr
9. Kori M, Maayan-Metzger A, Shamir R, et al. Faecal elastase 1 levels in premature and full term infants. Arch Dis Child Fetal Neonatal Ed
10. Campeotto F, Kapel N, Kalach N, et al. Low levels of pancreatic elastase 1 in stools of preterm infants. Arch Dis Child Fetal Neonatal Ed
11. Corvaglia L, Paoletti V, Battistini B, et al. Lack of correlation between fecal elastase-1 levels and fecal nitrogen excretion in preterm infants. J Pediatr Gastroenterol Nutr
12. Lüth S, Teyssen S, Forssmann K, et al. Fecal elastase-1 determination: ‘gold standard’ of indirect pancreatic function tests? Scand J Gastroenterol
13. Ferrone M, Raimondo M, Scolapio JS. Pancreatic enzyme pharmacotherapy. Pharmacotherapy
14. Fenton TR. A new growth chart for preterm babies: Babson and Benda's chart updated with recent data and a new format. BMC Pediatr
15. Armand M, Hamosh M, Mehta NR, et al. Effect of human milk or formula on gastric function and fat digestion in the premature infant. Pediatr Res
16. Lindquist S, Hernell O. Lipid digestion and absorption in early life: an update. Curr Opin Clin Nutr Metab Care
17. Hamprecht K, Maschmann J, Jahn G, et al. Cytomegalovirus transmission to preterm infants during lactation. J Clin Virol
18. Hamele M, Flanagan R, Loomis CA, et al. Severe morbidity and mortality with breast milk associated cytomegalovirus infection. Pediatr Infect Dis J
19. Ferrie S, Graham C, Hoyle M. Pancreatic enzyme supplementation for patients receiving enteral feeds. Nutr Clin Pract
20. Guillet R, Stoll BJ, Cotten CM, et al. Association of H2-blocker therapy and higher incidence of necrotizing enterocolitis in very low birth weight infants. Pediatrics
21. Terrin G, Passariello A, De Curtis M, et al. Ranitidine is associated with infections, necrotizing enterocolitis, and fatal outcome in newborns. Pediatrics