Effect of enteral erythropoietin on feeding-related complications in preterm newborns: A pilot randomized controlled study

BACKGROUND AND STUDY AIMS

To evaluate the effects of enteral administration of recombinant human erythropoietin (rhEPO) on feeding-related complications in preterm infants.

PATIENTS AND METHODS

This double-blind, randomized controlled pilot study enrolled 120 preterm infants born ≤ 32 weeks' gestation who were admitted to the neonatal intensive care unit in a tertiary hospital; 60 patients randomly received recombinant human erythropoietin while the other 60 received placebo. Newborns who underwent cardiopulmonary resuscitation, infants with genetic syndromes, infants with inborn errors of metabolism, infants with major congenital or acquired gastrointestinal tract malformations, infants with previous use of parenteral growth factors such as recombinant human erythropoietin and granulocyte-macrophage colony-stimuating factor (GM-CSF) and infants previously treated with intravenous immunoglobulin were excluded. Overall, 48 patients withdrew from the study because of intravenous haematopoietic growth factor intake or death before treatment was completed. A total of 72 preterm infants remained in the study: 36 preterm infants in the erythropoietin (EPO) group, and 36 preterm infants in the placebo group. The day that enteral feeding was successfully started, the time to establishing one-half, two-thirds, and full enteral feedings (reaching at least 150 mL/kg/day), the number of episodes of feeding intolerance, the time to regain birth weight and the incidence of necrotizing enterocolitis (NEC) were recorded.

RESULTS

Both groups showed no significant difference in the time to achieve one-half, two-thirds, or full enteral feeding, no signs of feeding intolerance, and no cases of NEC were recorded.

CONCLUSION

Enteral erythropoietin does not appear to affect feeding intolerance or NEC incidence.

Recent Advances in Neonatal Pharmacotherapy

Objective:

To provide commentary and reviews and brief discussions in controversial or innovative recent advances in neonatal pharmacotherapy. To discuss cutting edge drug delivery systems that may become useful in neonatal drug delivery in the future.

Data Sources:

Articles were identified through searches of MEDLINE (1990–October 2005), key articles in the authors' files, and in some cases, through data generated and/or published by the author of a particular topic.

Data Selection:

Article selection and relevance to the topics under discussion was determined by individual authors.

Data Synthesis:

Therapeutic strategies addressed in this review include the use of hematopoietic growth factors including a simulated amniotic fluid preparation containing these growth factors for neonates with selected gastrointestinal problems, erythropoietin for neuroprotection following perinatal asphyxia, drug therapy advances in treatment of patent ductus arteriosus (PDA), evaluation of advances in transdermal drug delivery, and its potential application to neonates and advances in the treatment of persistent pulmonary hypertension (PPHN) of the newborn.

Conclusions:

Despite being over 30 years old, the practice of neonatology is as much of an art as a science. Advances in the basic science research have improved our understanding of use of pharmacologic agents in the premature and full-term neonate including drug disposition pathways. Expanding our knowledge on issues such as physiology of hematopoietic factors, the pharmacologic responses of conditions such as PDA and PPHN, and newer technologies for drug administration, as well as other pharmacologic reponses in the neonate are vital in the development of safe and efficacious treatments for neonates. Many questions remain unanswered, and every clinician must make an effort to contribute to the knowledge and understanding of pharmacotherapy in this patient population.

In the critically ill, nothing-by-mouth infant, would enteral administration of simulated amniotic fluid improve feeding tolerance compared with the current practice of no therapy? An evidence-based review

Necrotizing enterocolitis (NEC ) can be a devastating disease in the NIC U population. The current standard of practice of not providing enteral nutrition to the critically ill and the premature infant is thought to result in pathophysiologic changes of the gastrointestinal (GI) tract that may result in the development of NEC. Various methods of preventing or reducing the incidence of NEC in the NIC U have been explored. One such method is the enteral administration of simulated amniotic fluid (SAF). This article compares, contrasts, and reviews the available evidence regarding the use of SAF feedings as a means of reducing GI tract changes associated with nothing-by-mouth (NPO) status.

Treating feeding intolerance with an enteral solution patterned after human amniotic fluid: a randomized, controlled, masked trial

OBJECTIVE

Feeding intolerance is a common problem in the neonatal intensive care unit (NICU) and some cases might be causally related to atrophic changes in the small bowel mucosa. We speculated that for such patients, feeding tolerance might improve after oral administration of enterocyte growth factors in a sterile, isotonic solution patterned after amniotic fluid.

STUDY DESIGN

Twenty neonates meeting criteria for feeding intolerance were eligible for this trial. They were randomized to either Group 1 (test solution) or Group 2 (control). Group 1 received 2.5 ml of test solution/kg every 3 h by oral-gastric or nasal-gastric (OG/NG) tube. This was begun when the patient was NPO because of feeding intolerance and continued until 80 ml/k/day of milk feedings were tolerated, or for a maximum of 7 days. Group 2 received a sham OG/NG administration every 3 h, until 80 ml/k/day of milk feedings were tolerated, or for a maximum of 7 days. Only the bedside nurse and the NICU pharmacist were aware which patients received the test solution and which received the sham administrations. The volumes of milk feedings were increased by order of the attending neonatologist and nurse practitioner. The study outcome was enteral calories/kg/day during and for 7 days after the cessation of the treatments.

RESULTS

Eleven patients were randomized to receive the test solution and nine to receive sham administrations. At study entry, the two groups were not different in gestational age, postnatal age, signs of feeding intolerance or cal/k/day taken enterally during the previous 3 days. The study doses were given for an average of just under 6 days (range, 2 to 7 days). During the week following the administrations, the test solution recipients trended toward more enteral calories. Specifically, they had an increase averaging 78+/-20.8 cal/k/day more than before the study, whereas the sham recipients had an increase averaging 55.9+/-33 cal/k/day more than before the study (P=0.05 for a one-sided test and P=0.10 for a two-sided test). The test solution recipients also had a trend toward fewer formula changes than did the sham recipients (P=0.10). In this small, randomized, controlled, masked trial, the administration of a sterile, non-caloric, growth factor containing solution patterned after human amniotic fluid was associated with trends that we interpret as reflecting better tolerance of milk feedings. On this basis, we suggest that a phase III efficacy trial should be accomplished, using the present data for sample size calculations.

A phase 1 trial testing an enteral solution patterned after human amniotic fluid to treat feeding intolerance

OBJECTIVE

Feeding intolerance is a common problem in neonates; in some cases it may be related to atrophic changes in the small bowel mucosa. In these infants, feeding tolerance might improve after oral administration of enterocyte growth factors in a sterile, isotonic, noncaloric solution patterned after human amniotic fluid.

STUDY DESIGN

Ten infants who met the study criteria for feeding intolerance were enrolled in this pilot trial. The test solution was administered, 2.5 mL/kg every 3 hours, until 80 mL/kg/day of milk feedings were tolerated, or for a maximum of 7 days. When milk feedings were prescribed, they were mixed with the test solution. Volumes of milk feedings were increased in amounts determined by the attending neonatologist and neonatal nurse practitioner. The outcome was the net average enteral kilocalories per kilogram of body weight per day (kcal/kg/day) for 3 periods: (1) the 3 days before the test treatment was begun, (2) the days the test treatment was administered, and (3) the 7 days following cessation of the test treatment.

RESULTS

At study entry the infants' chronological ages ranged from 3 to 65 days. All infants tolerated the test solution, which was given for a mean of 5 days (range 3 to 7 days). During that period, the infants received more enteral calories (mean 18 kcal/kg/day, range 1 to 32) than during the preceding 3 day period (mean 3 kcal/kg/day, range 0 to 7, P < 0.05). During the 7 days following the test treatment the infants received 87 enteral kcal/kg/day (range 11 to 109, P < 0.02 compared to the treatment period).

CONCLUSIONS

In this Phase 1 pilot trial, 10 heterogeneous infants in the neonatal intensive care unit with feeding intolerance had a significant increase in milk feedings after treatment with a sterile, isotonic, noncaloric solution patterned after human amniotic fluid.

Enteral administration of a simulated amniotic fluid to very low birth weight neonates

OBJECTIVE

To reduce feeding intolerance among very low birth weight neonates.

STUDY DESIGN

A total of 10 neonates with birth weights of 750 to 1250 g were given oral-gastric boluses (2.5 ml/kg every 3 hours) of a solution patterned after amniotic fluid. When milk feedings were begun the milk was mixed with the test solution. The solution was given at a constant daily dose of 20 ml/kg/day while the volumes of milk feedings were gradually increased. When milk feedings reached 80 ml/kg/day the test solution was discontinued. A comparison group consisted of neonates who met study criteria but were cared for during the period immediately preceding the study. The outcome was the number of calories taken enterally over the first 21 days of life.

RESULTS

The test solution was begun an average of 27 hours after birth (range, 4 to 45). In the test group the first milk feedings were introduced 74 hours after birth (range, 18 to 144), which was similar to the time milk was introduced in the comparison subjects (79 hours; range, 18 to 168). After milk feedings were started, the test patients had a total of four NPO days (0.4 NPO days per patient) during their first 21 days, while the comparison group had 34 NPO days (3.4 NPO days per patient). During the first 14 days of life the test solution recipients had a median of 26.5 enteral cal/kg/day (range, 4.3 to 68.9), while the comparison neonates had 8.5 (range, 0.2 to 25; p < 0.05). During the first 21 days of life the test solution recipients had a median of 56.9 enteral cal/kg/day (range, 11.5 to 89.4), while the comparison neonates had 19.2 (range, 0.9 to 52.8; p < 0.05).

CONCLUSION

In all, 10 VLBW infants tolerated the test solution for periods up to 14 days with no significant adverse effects. A randomized trial to determine whether this solution reduces feeding intolerance among VLBW neonates should be conducted.

Effects of enterally administering granulocyte colony-stimulating factor to suckling mice

Gastrointestinal (GI) tract development is influenced by multiple growth factors, some of which are delivered directly to the GI lumen, as they are swallowed constituents of amniotic fluid, colostrum, and milk. Granulocyte colony-stimulating factor (G-CSF), traditionally known as a granulocytopoietic growth factor, is an example of one such factor. However, it is not clear whether the large amounts of G-CSF that are normally swallowed by the fetus and neonate have systemic effects on circulating neutrophils or local effects in the developing intestine. To assess this, we administered either active or heat-denatured (control) recombinant human G-CSF to 5- to 7-d-old C57BL/6 x 129SvJ mice. Pups received either a low dose (3 ng) that was calculated to approximate the amount of G-CSF swallowed in utero from amniotic fluid or an isovolemic high dose 100 times larger (300 ng). Oral dosing was performed daily for either 3 or 7 d, after which pups were killed and measurements were made on the blood and the GI tract. Absolute blood neutrophil counts and immature to total neutrophil ratios did not differ from controls in any of the test groups. However, intestinal villus area, perimeter, length, crypt depth, and proliferating cell nuclear antigen index increased significantly among those that were treated with active G-CSF. Thus, in suckling mice, enterally administered G-CSF had no effect on the concentration of circulating neutrophils but had trophic effects on the intestine. We speculate that the G-CSF present in amniotic fluid, colostrum, and milk acts as a topical intestinal growth factor and has little or no granulocytopoietic action.

Tolerance of an enterally administered simulated amniotic fluid-like solution by neonates recovering from surgery for congenital bowel abnormalities

OBJECTIVE

We report a single-centered, Phase I pilot trial, testing the enteral administration of an experimental amniotic fluid-like solution to 10 neonates who were otherwise "NPO" following surgery for congenital bowel abnormalities. The overall hypothesis was that the trophic effect of the solution on intestinal villi would facilitate advancement to full enteral feedings. The specific hypothesis tested in this pilot trial was that the solution would be tolerated.

STUDY DESIGN

Ten neonates who were NPO following surgery for congenital bowel abnormalities, were studied before any "trophic" feedings were begun. Each received an experimental, sterile, isotonic, amniotic fluid-like solution at a dose of 20 ml/kg/day enterally. When milk feedings were begun they were mixed with the experimental solution. Increases in the volume of milk feedings occurred at the discretion of the neonatologist and surgeon, and the experimental solution was discontinued any time the neonatologist or surgeon felt it was not tolerated, or when 100 ml of milk feedings/kg/day was achieved. We quantified the amount and character of emesis, stools, and gastric residuals, measured abdominal girth and blood pressure, looked for skin rashes, and sought any signs of intolerance or adverse events. We recorded the days to achieve milk feedings of 20, 50, 100, and 120 ml/kg/day and length of hospital stay.

RESULTS

The experimental solution was begun 4 to 32 days after surgery, invariably prior to the institution of "trophic" milk feedings. All subjects completed the doses with no evidence of intolerance. All achieved 100 ml/kg of milk feedings 14 days, or fewer, following institution of the experimental solution (mean 11.1 days, range, 3 to 14). All lived and were discharged home 20.2 days (range, 8 to 42) after the experimental solution was begun.

CONCLUSIONS

In this pilot trial involving 10 neonates who had surgery for congenital bowel abnormalities, the enteral administration of a sterile, isotonic, amniotic fluid-like solution was tolerated.

Endogenous and recombinant erythropoietin levels decline in human amniotic fluid and fetal plasma in vitro at 37 degrees C

OBJECTIVE

To determine the in vitro stability of endogenous and recombinant erythropoietin (EPO) incubated at 37 degrees C in amniotic fluid (AF) and fetal plasma.

STUDY DESIGN

Endogenous and recombinant EPO in AF, fetal plasma and phosphate buffer were incubated in vitro for 21 days at 37 degrees C. Serial aliquots were analyzed for EPO and the rates of EPO decline were compared within and between groups.

RESULTS

Endogenous and recombinant EPO declined significantly in plasma and AF at 37 degrees C. Endogenous EPO displayed a similar linear rate of decline in AF and plasma, with nearly 70% of the initial hormone concentration remaining at 21 days. Recombinant EPO incubated in buffer did not change.

CONCLUSIONS

Using the rate of decline in endogenous EPO we observed, EPO levels measured in AF or plasma within 21 days of fetal demise can be extrapolated back to the level likely present at fetal death.

Hematopoietic growth factors in neonatal medicine: the use of enterally administered hematopoietic growth factors in the neonatal intensive care unit

The practice of complete bowel rest in prematurely delivered neonates and those who have undergone surgery for congenital anomalies of the gastrointestinal (GI) tract is common in neonatal intensive care units (NICU). However, increased recognition of the critical role of growth factors in GI development suggests that this practice might be modified to include the administration of synthetic amniotic fluid-like solutions designed to bridge the neonate between their intra-uterine environment and that of the NICU. This article reviews advances in administering synthetic amniotic fluid-like solutions in the NICU.

Recombinant granulocyte colony-stimulating factor administered enterally to neonates is not absorbed

Granulocyte colony-stimulating factor (G-CSF) is present in liquids swallowed by the fetus and neonate; specifically, amniotic fluid, colostrum, and human milk. The swallowed G-CSF has local effects on enteric cells, which express the G-CSF receptor. However, some portion of the G-CSF ingested by the fetus and neonate might be absorbed into the circulation and have systemic actions, such as stimulating neutrophil production. To assess this possibility we sought to determine if circulating G-CSF concentrations of neonates increase after enteral administration of recombinant human granulocyte colony-stimulating factor (rhG-CSF). This was a single-center, prospective, blinded, randomized, 2 x 2 crossover study, with each infant receiving 1 dose of rhG-CSF (100 microg/kg) and 1 dose of placebo. Plasma G-CSF concentrations were measured at 2 and 4 hours after administration of the test solution. No significant change in plasma G-CSF concentration was observed after the enteral administration of rhG-CSF. On this basis, we conclude that orally administered rhG-CSF is not absorbed in significant quantities, and we speculate that the G-CSF swallowed by the fetus and neonate has local but not systemic effects.