INTRODUCTION

Necrotizing enterocolitis (NEC) is one of the most common causes of death among preterm infants who survive the first week of life.¹ Up to 12% of very-low-birth-weight (VLBW; <1500 g) infants are affected with this condition.² Neonates who develop NEC are generally made NPO for a period of up to 1 to 2 weeks. Following this period, where total parenteral nutrition is utilized, feedings are generally instituted gradually, because feeding intolerance is a common complication. Feeding intolerance can be manifest as emesis, large gastric residuals, abdominal distention, abdominal tenderness, or occult or frank blood in the stool.³ One mechanism predisposing patients recovering from NEC to feeding intolerance is the intestinal villous atrophy that occurs during the NPO period. In fact, studies in experimental animals and adult humans indicate that being NPO for periods as short as 24 to 72 hours results in substantial loss of villous structure and function.^4,5

We reported that granulocyte colony-stimulating factor (G-CSF) and erythropoietin (Epo) act physiologically in the fetus and neonate as intestinal villous growth factors.^{6,7,8,9,10,11} These two factors are present in fluids swallowed by the fetus and neonate, namely amniotic fluid, colostrum, and human milk, and these factors resist digestion using simulated gastric and small intestinal conditions.^6,7,8 We also observed that receptors for G-CSF and Epo are expressed on the luminal surface of fetal and neonatal intestinal villi,^9,10 and that rodents fed formulas containing rEpo have a dose-dependent preservation of their villous structure and function, compared with those fed isocaloric and isovolemic formulas devoid of recombinant Epo (rEpo).¹¹ Moreover, we reported the stability of a sterile isotonic electrolyte solution for enteral administration containing recombinant G-CSF (rG-CSF) and rEpo¹², and we observed no adverse effects of administering this solution to 30 VLBW infants who were NPO during their first 3 days of life.¹³

Relative to the present study, we speculated that stimulating intestinal villous growth and function would benefit neonates who were recovering from NEC, if this could be accomplished prior to the reinstitution of enteral feedings. As a first step, we postulated that following their NPO period, but before nutrient feedings were resumed, the enteral administration of this sterile isotonic growth factor-containing solution would be well tolerated. Thus, the present study was undertaken as a phase I trial aimed at determining whether subsequent safety and efficacy studies would be reasonable.

METHODS

Patients

Patients were eligible if they had the diagnosis of NEC (Table 1)¹⁴ and a decision had been made by the attending Neonatologist and Surgeon to keep the patient NPO for at least 1 week. Patients were considered ineligible if they had any congenital or acquired anomaly of the gastrointestinal tract (i.e., omphalocele, gastroschisis, tracheoesophageal fistula, intestinal perforation), or had other major congenital anomalies (congenital heart disease, neural tube defect, congenital diaphragmatic hernia, Trisomy, etc.). Also excluded from study were neonates judged as being too ill to be acceptable study candidates; this criterion was set as receiving mechanical ventilation with an FIO₂ >0.60. The study was approved by the Institutional Review Board of the Hospital General Dr. Ignacio Morones Prieto, San Luis Potosi City, San Luis Potosi, Mexico, and informed consent was obtained from parents of all enrolled subjects.

Table 1 - Modified Bell's Staging Criteria for Necrotizing Entercolitis.

Full table

Study Procedures

The hospital pharmacist made the test solution using sterile technique in the manner previously reported.¹² The solution contained 115 meq/l sodium chloride, 17 meq/L sodium acetate, and 4 meq/l potassium chloride, 225 ng/ml Neupogen (Filgrastim, Amgen Inc., Thousand Oaks, CA), 4400 mU/ml Epoetin alfa (Epogen, Amgen). Human serum albumin (5%) (Baxter Healthcare Corp., Hyland Division, Glendale, CA) was added to the infusion bag prior to the addition of the cytokines (final concentration of albumin=0.05%). The fluid contained a 10-fold higher concentration of G-CSF and Epo than natural amniotic fluid (so that 20 ml/kg will supply the amount in 200 ml/kg normally swallowed in utero).

Aliquots were frozen until use. Separate aliquots (10 ml) were frozen and labeled "Priming". Solution in the "Priming" syringe was pushed through the orogastric tube prior to insertion into the patient in order to reduce binding of rEpo to the plastic tubing.¹² Upon enrollment of a subject a full day's dose and a priming syringe were thawed. The full day's dose was then divided into eight equal amounts, to be administered by the bedside nurse every 3 hours. A total of 24 doses were to be given to each enrolled infant. Each aliquot was allowed to warm to room temperature before it was administered, and it was recorded in the medical record as both a medication and an enteral fluid.

A total of three groups consisting of 10 neonates each received the fluid in a dose-escalation fashion; 5, 10, and 20 ml/kg/day. Specifically, the first 10 neonates received 5 ml/kg/day of fluid. The study was then resumed with the second group of 10 neonates receiving 10 ml/kg/day of test solution in a similar fashion. Once enrollment of all of the infants in the second group was completed and the last infant completed the final dose, the final group, consisting of 10 neonates, received 20 ml/kg/day of solution.

Demographics and Measurements of Intolerance

Demographic data collected included gestational age, birth weight, gender, day of life feedings were initiated, the type of feedings, the number of days infants were fed before developing NEC, and the volume of those feedings, and medications given prior to developing NEC. Any antenatal steroid exposure, parenteral use of epoetin alfa (rEpo) or neupogen (rG-CSF) administration were recorded. Once NEC was diagnosed, daily documentation of the following measures was kept; Stage of NEC, presence and character of emesis, character of stools, presence and character of gastric residuals, change in abdominal girth, presence of a skin rash, and blood pressure instability. These measures were also recorded daily during the 3 days of study fluid administration and for each of the next 7 days after the study administration concluded.

Statistical Analysis

Within-group comparisons of time-related observations were made using paired t-tests. Variables were tested for normality by the Kolmogorov–Smirnov and Shapiro–Wilk tests, while the homogeneity of variance in observations was confirmed by the Levine statistic. Only gestational age, birth weight, and volume of feeds prior to NEC satisfied criteria to be parametric, and these data were compared by a one-way ANOVA. Other variables were analyzed by the Kruskal–Wallis test. A p-value of <0.05 was considered statistically significant. Mean and standard deviations were reported for specific characteristics.

RESULTS

In total, 30 neonates were enrolled on study between June 1, 2000 and October 31, 2001. The subjects were grouped by dose of study solution, that is, patients 1 to 10 (5 ml/kg/day); patients 11 to 20 (10 ml/kg/day); and patients 21 to 30 (20 ml/kg/day). A total of 18 males and 12 females were enrolled. The mean gestational age of the subjects was 33.83.1 weeks (meanstandard deviation; range 25.3 to 41.1 weeks), the average birth weight was 1498687 g (range 700 to 3350 g), and feedings were initiated at 5.76.4 days of life (range 0 to 34 days). There were no differences between any of the dosing groups regarding gestational age, birth weight, or day of life when feedings were initiated (p=0.723, 0.826, and 0.618, respectively). Similarly, there were no differences between the groups in the volumes of feedings administered prior to the diagnosis of NEC.

Some of the subjects received other medications in addition to the sterile isotonic electrolyte solution containing select recombinant growth factors. Antenatal steroids (n=8) were the most frequently administered medication. Only one neonate received rEpo; none received intravenously administered rG-CSF. Once neonates developed NEC, antibiotics and dopamine (2.5 mcg/kg/min) were the most frequently administered medications (n=30 and n=21, respectively). Fewer neonates in the 10 ml/kg/day group received inotropes than in either of the other dosing groups (p=0.034).

Most infants were fed prior to developing NEC (n=22). Of those who were fed, 14 received formula exclusively, two received human milk exclusively, and six received a combination. All but one of the eight neonates who developed NEC without being fed was less than 1500 g. One term infant (patient #20) developed NEC without being fed.

Characteristics of each study patient and measures of tolerance during the study and during the week following administration of the study solution were examined. The outcome and tolerance data included the occurrence of emesis, gastric residuals, diarrhea, bloody stools, abdominal distention (measurement of circumference), rashes and death. In all, 16 infants developed stage IA NEC, six developed stage IB NEC and eight developed stage IIA NEC. There were no differences between the groups in the Stage of NEC (p=0.509) or in the number of NPO days (p=0.163). The mean day of life that NEC was diagnosed was 15.726.5 days and the average time to receiving the solution was 19.19.5 days. None of the neonates who received the study solution had the administration of the study fluid prematurely discontinued, and none required surgery for NEC.

The occurrence of gastric residuals during the administration of the study solution was similar between the dosage groups (n=5, n=4, and n=6 and in the 5, 10, and 20 ml/kg/day groups; p=0.694). Similarly, the number of neonates who had gastric residuals in the week after the administration of the study solution was similar between the groups (p=0.873). Likewise, when the occurrence of gastric residuals was examined in each infant during and after the administration of the study solution there were no differences (p=0.96). None of the infants experienced diarrhea, guaiac positive or bloody stools, rashes, or abdominal distention during or after the administration.

When feedings were resumed in the neonates after NEC, none developed diarrhea or guaiac positive or bloody stools. There were no differences between the dosage groups regarding abdominal distention (p=0.339). Death occurred in two study patients (#6 and #20). Both were because of fulminate sepsis 14 (patient #6) and 21 (patient #20) days after completing the study.

DISCUSSION

We performed a multicentered, phase I, dose-escalation trial to assess whether neonates who were NPO because of NEC would tolerate feedings of a sterile isotonic electrolyte solution containing select recombinant growth factors. Although this study enrolled a small number of infants, the characteristics of the patients are similar to those reported for other neonates who develop NEC, in that most; (1) were fed prior to development of NEC, (2) were premature, (3) received formula feedings, and (4) developed NEC during their second to third week of life.^1,2 A significant number of neonates in this study received dopamine once NEC was diagnosed, reflecting the practice at the institutions performing the study to enhance gastrointestinal tract perfusion.¹⁵ While significantly fewer infants in the 5 ml/kg/day group received dopamine when compared to each of the other dosage groups, no other differences were noted between the groups, suggesting that this finding reflects random variability in the initiation of dopamine by the Neonatologists. Each of the neonates tolerated the study solution without any observed adverse effects during the study or in the week following.

The amount of recombinant growth factors in the study solution was set such that 20 ml/kg/day would approximate the upper limit of the normal amount of G-CSF and Epo swallowed per day by a term fetus in utero.^12,13 Our measurements suggested that this preparation was well tolerated in these patients in doses up to and including 20 ml/kg/day for three consecutive days in neonates recovering from NEC.

Most infants in this study had stage I NEC. None of the study patients had more severe NEC than stage IIA, and none of the neonates in this study required surgery for NEC. While the patients in this study represent the majority of neonates who develop NEC, in that they have less severe disease and can be medically managed, it remains to be shown how the study solution is tolerated in neonates with more severe disease. In addition, stage I NEC is considered "suspected NEC", where systemic signs are nonspecific and intestinal findings include gastric residuals and guaiac-positive stools with nonspecific radiographic findings.¹⁴ It is difficult to separate these findings from those of neonates with feeding intolerance.

NEC remains a major problem in neonatal intensive care.^1,2,3 Indeed, many questions remain to be answered regarding prevention strategies¹⁶ and optimal medical and surgical management.¹⁷ One problematic area is the feeding intolerance that is almost universal in these patients once feedings are reinstated following a prolonged period of total parenteral nutrition. A historic cohort study¹⁸ and several animal experiments^11,19,20,21 suggest that rEpo or rG-CSF might be useful in preventing or treating treat villous atrophy in neonates with NEC. Although G-CSF and Epo are included in the formulation we tested, we recognize that other growth factors are also involved in the development of the intestine. In fact, Koldovsky²² reported that as many as 28 peptide and nonpeptide hormones have been identified in human milk, including epidermal growth factor (EGF), insulin-like growth factor (IGF), and transforming growth factor beta (TGF-).

EGF is present in amniotic fluid²³ and in human milk and EGF receptors are expressed on fetal small intestine and colon.²⁴ Binding of EGF with its receptor induces maturation of brush border hydrolytic enzymes and stimulates the proliferation of these cells.²⁴ The parenteral administration of EGF increases intestinal proliferation in human infants with congenital microvillous atrophy or NEC.^25,26 IGF is also contained in amniotic fluid and human milk.^27,28 Like Epo and G-CSF, IGF in milk is relatively protected from digestion in vitro. Similar to other growth factors, enterally administered IGF induces intestinal DNA synthesis and cell growth, resulting in increased villous height, and proliferation of intestinal epithelial cells.²⁹ TGF- is also present in amniotic fluid and human milk.³⁰ TGF- is expressed along the crypt-villous axis, and it is proposed that TGF- supports differentiation of villous enterocytes. TGF- is also expressed by villous enterocytes and may have a role in mucosal healing.³⁰

For simplicity, in the present study we limited the number of growth factors in the study solution to two, and we focused only on the issue of tolerance during the 3 days of treatment and in the week following administration. We observed no intolerance among 30 patients, but we recognize that this relatively small sample size could miss infrequent adverse effects. However, on the basis of the present study, we conclude that phases II and III studies can now commence to examine safety, effectiveness, and benefit to risk ratio of this approach to improving feeding tolerance of neonates recovering from NEC.

References

1. Foglia RP. Necrotizing enterocolitis. Curr Probl Surg 1995;32:757–823.
2. Uauy RD, Fanaroff AA, Korones SB, Phillips EA, Phillips JB, Wright LL. Necrotizing enterocolitis in very low birth weight infants: biodemographic and clinical correlates. J Pediatr 1991;119:630–638. | PubMed |
3. Neu J. Necrotizing enterocolitis: The search for a unifying pathogenic theory leading to prevention. Pediatr Clin North Am 1996;43:409–432.
4. Buchman AL, Moukarzel AA, Bhuta S, et al. Parenteral nutrition is associated with intestinal morphologic and functional changes in humans. J Parenter Enteral Nutr 1995;19:453–460. | ISI | ChemPort |
5. Condino A, Barleycorn A, Lu W, Maheshwari A, Christensen RD, Calhoun DA. Abnormal intestinal histology in human fetuses with congenital defects of the gastrointestinal tract (GIT). Pediatr Res 2002;51:381A.
6. Calhoun DA, Lunoe M, Du Y, Staba SL, Christensen RD. Concentrations of granulocyte colony-stimulating factor in human milk after in vitro simulations of digestion. Pediatr Res 1999;46:767–771.
7. Juul SE, Joyce AE, Zhao Y, Ledbetter DJ. Why is erythropoietin present in human milk? Studies of erythropoietin receptors on enterocytes of human and rat neonates. Pediatr Res 1999;46:263–268. | PubMed | ChemPort |
8. Calhoun DA, Lunoe M, Du Y, et al. Concentrations of granulocyte colony-stimulating factor in human milk after in vitro simulations of digestion. Pediatr Res 1999;46:767–771.
9. Calhoun DA, Donnelly Jr., WH, Du Y, et al. Distribution of granulocyte colony-stimulating factor (G-CSF) and G-CSF- receptor mRNA and protein in the human fetus. Pediatr Res 1999;46:333–338.
10. Juul SE, Yachnis AT, Christensen RD. Tissue distribution of erythropoietin and erythropoietin receptor in the developing human fetus. Early Hum Dev 1998;52:235–249. | Article | PubMed | ISI | ChemPort |
11. Juul SE, Ledbetter DJ, Joyce AE, et al. Erythropoietin acts as a trophic factor in neonatal rat intestine. Gut 2001;49:182–189.
12. Calhoun DA, Juul SE, McBryde EV, Sullivan SE, Christensen RD. Stability of filgrastim and epoetin alfa in a system designed for enteral administration in neonates. Ann Pharmacother 2001;34:1257–1261.
13. Sullivan SE, Calhoun DA, Maheshwari A et al. Tolerance of simulated amniotic fluid in premature neonates. Ann Pharmacother 2002;36:1518–1524. | Article | PubMed |
14. Walsh MC, Kliegman RM, Fanaroff AA. Necrotizing enterocolitis: A practitioner's perspective. Pediatr Rev 1988;9:225.
15. Hentschel R, Hensel D, Brune T, Rabe T, Jorch G. Impact on blood pressure and intestinal perfusion of dobutamine and dopamine in hypotension in preterm neonates. Biol Neonate 1995;68:318–324.
16. Foster J, Cole M. Oral immunoglobulin for preventing necrotizing enterocolitis in preterm and low birth-weight neonates. Cochrane Database Syst 2001; Rev 3:CD001816.
17. Moss RL, Dimmitt RA, Henry MC, Geraghty N, Efron B. A meta-analysis of peritoneal drainage versus laparotomy for perforated necrotizing enterocolitis. J Pediatr Surg 2001;36:1210–1213. | Article | PubMed | ChemPort |
18. Ledbetter DJ, Juul SE. Erythropoietin and the incidence of necrotizing enterocolitis in infants with very low birth weight. J Pediatr Surg 2000;35:178–182. | Article | PubMed | ISI | ChemPort |
19. Aklise M, Kullahcioglu Girgin F, Baka M, Husseyinov A, Kultursay N. The role of recombinant human erythropoietin in lipid peroxidation and platelet-activating factor generation in a rat model of necrotizing enterocolitis. Eur J Pediat Surg 2001;11:167–172.
20. Kling PJ, Sullivan TM, Roberts RA, Philipps AF, Koldovsky O. Human milk as a potential enteral source of erythropoietin. Pediatr Res 1998;43:216–221. | PubMed |
21. Gersting JA, Lunoe M, Staba SL, Christensen RD, Calhoun DA. Evaluation of the enteral administration of recombinant human granulocyte colony-stimulating factor (rhG-CSF) to mouse pups with an intact G-CSF-receptor (G-CSF-R) gene (WT) and pups deficient for the G-CSF-R gene (KO). Pediatr Res 2000;47:287A.
22. Koldovsky O. Hormones and growth factors in milk. Ann Nestle 1996;54:105–112.
23. Romero R, Wu YK, Oyarzun E, Hobbins JC, Mitchell MD. A potential role for epidermal growth factor/alpha-transforming growth factor in human parturition. Eur J Obstet Gynecol Reprod Biol 1989;33:55–60.
24. Pothier P, Menard D. Presence and characteristics of epidermal growth factor receptors in human fetal small intestine and colon. FEBS Lett 1988;228:113–117.
25. Marti U, Burwen SJ, Jones AL. Biological effects of epidermal growth factor, with emphasis on the gastrointestinal tract and liver: An update. Hepatology 1989;9:126–138. | PubMed | ISI | ChemPort |
26. Walker-Smith JA, Phillips AD, Walford N, et al. Intravenous epidermal growth factor/urogastrone increases small-intestinal cell proliferation in congenital microvillous atrophy. Lancet 1985;2:1239–1240.
27. Verhaeghe J, Coopmans W, van Herck E, van Schoubroeck D, Deprest JA, Witters I. IGF-I, IGF-II, IGF binding protein 1, and C-peptide in second trimester amniotic fluid are dependent on gestational age but do not predict weight at birth. Pediatr Res 1999;46:101–108.
28. Nagashima K, Itoh K, Kuroume T. Levels of insulin-like growth factor I in full- and preterm human milk in comparison to levels in cow's milk and in milk formulas. Biol Neonate 1990;58:343–346.
29. Steeb CB, Trahair JF, Read LC. Administration of insulin-like growth factor-I (IGF-I) peptides for three days stimulates proliferation of the small intestinal epithelium in rats. Gut 1995;37:630–638.
30. Saito S, Yoshida M, Ichijo M, Ishizaka S, Tsujii T. Transforming growth factor-beta (TGF-beta) in human milk. Clin Exp Immunol 1993;94:220–224. | PubMed | ISI | ChemPort |