- Email: [email protected]
Efficacy of oral erythromycin for treatment of feeding intolerance in preterm infants
EFFICACY OF ORAL ERYTHROMYCIN FOR TREATMENT OF FEEDING INTOLERANCE IN PRETERM INFANTS PRACHA NUNTNARUMIT, MD, MSC, PAKAPHAN KIATCHOOSAKUN, MD, WACHAREE TANTIPRAPA, MD, SUPPAWAT BOONKASIDECHA, MD
AND
Objective To determine the efficacy and safety of oral erythromycin (EM) for feeding intolerance in preterm infants < 35 weeks gestation.
Study design In this randomized, double-blinded, placebo-controlled trial, preterm infants with feeding intolerance were randomly allocated to a treatment group given EM ethyl succinate 10 mg/kg every 6 hours for 2 days, followed by 4 mg/kg every 6 hours for another 5 days, or to a control group given placebo. The primary outcome was time to full feeding (150 mL/kg/day) after the start of treatment. Results Each group comprised 23 preterm infants, almost all of whom were < 32 weeks gestation. Baseline characteristics were similar between the 2 groups. Times to full feeding were significantly shorter and the number of withheld feeds were significantly less in the EM group than the control group; the respective medians (interquartile ranges) were 7 days (6 to 9 days) versus 13 days (9 to 15 days) (P < .001) and 1 episode (0 to 2 episodes) versus 9 episodes (2 to 13 episodes) (P < .001). No significant differences in episodes of sepsis, necrotizing enterocolitis, and cholestasis were observed. Conclusions Oral EM was effective and safe for treatment of feeding intolerance in preterm infants. (J Pediatr 2006;148:600-5)
dvancing feedings to full enteral feeding is frequently difficult, especially in very low birth weight infants, due to feeding intolerance, and may take up to 2 to 3 weeks. This problem necessitates prolonged parenteral nutrition with its attendant complications. Feeding intolerance is often attributable to absence of the migratory motor complex (MMC) in the small intestine and the lack of mature response to feeding, resulting in gastrointestinal dysmotility, characterized by delayed gastric emptying and prolonged gastrointestinal transit time.1,2 From the Department of Pediatrics, Faculty Erythromycin (EM), a macrolide antibiotic, has potent prokinetic activity through of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand; Department its agonist action on motilin receptors in the stomach and proximal small intestine in both of Pediatrics, Faculty of Medicine, Srinagar3-5 humans and animals. EM at low doses (3 to 12 mg/kg/day) induces phase III MMC ind Hospital, Khon Kaen University, Khon Kaen, Thailand; Department of Pediatrics, of the small intestine through activation of motilin neuronal receptors and at high doses Faculty of Medicine, Chiang Mai University, (antimicrobial doses, ⱖ 40 mg/kg/day) initiates antral contractility and causes sustained Chiang Mai, Thailand; and Department of 6-11 contractility of the small intestine through stimulation of motilin muscular receptors. Pediatrics, Faculty of Medicine, Srinakarinwirot University, Bangkok, Thailand. However, a recent systematic analysis of trials of EM as a prokinetic agent in preterm Supported by the Ramathibodi Fund (grant infants revealed conflicting results due to differences among the trials including doses, 46018). routes and modes of administration, durations of treatment, feeding protocols, and The medication (erythromycin ethyl succinate) and placebo used in this study were definitions of feeding intolerance.12 Most of the trials failed to show any benefit for the provided by Siam Pharmaceutical Ltd, prophylactic use of EM on the day that feedings were started.13-15 In terms of the Bangkok, Thailand, which had no involvetherapeutic use of EM for feeding intolerance, although several clinical trials have ment in any aspect of the study, including study design; data collection, analysis, and reported favorable outcomes, these trials were not randomized and controlled and ininterpretation; manuscript preparation; and 10,11,16 –18 To date there have been only 3 randomized volved only small sample sizes. the decision to submit the manuscript for publication. The first author wrote the first controlled trials,19-21 1 with high doses (48 mg/kg/day) reporting a shorter time to full draft of the manuscript; no grant or other feeding and reduced cholestasis and the other 2 with low doses (6 to 15 mg/kg/day) form of payment was received by anyone reporting no beneficial effect. involved in manuscript preparation. Submitted for publication Jun 18, 2005; last Currently the role of EM in the management of feeding intolerance in premature revision received Nov 3, 2005; accepted infants remains equivocal and warrants further study. Low doses appear to be ineffective, Dec 7, 2005. whereas high (antimicrobial) doses may cause such adverse effects as hypertrophic pyloric Reprint requests: Pracha Nuntnarumit, MD, Department of Pediatrics, Faculty of Medistenosis, cardiac arrhythmias, and elevated liver enzymes.22-24 An intermediate dose may cine, Ramathibodi Hospital, Mahidol Uniprove to be effective while minimizing adverse effects. Consequently, we conducted a versity, Rama VI, Bangkok, 10400, Thailand.
A
EM MMC
600
Erythromycin Migratory motor complex
NEC PDA
Necrotizing enterocolitis Patent ductus arteriosus
E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2006 Elsevier Inc. All rights reserved. 10.1016/j.jpeds.2005.12.026
randomized, placebo-controlled trial to determine whether or not an intermediate dose of oral EM would benefit preterm infants with feeding intolerance and to assess the adverse effects of this treatment modality.
METHODS Patients Preterm infants admitted to neonatal care nurseries of 3 university hospitals (Ramathibodi Hospital, Srinagarind Hospital, and Chiang Mai University Hospital) between December 2002 and February 2005 were consecutively enrolled in the trial. Inclusion criteria included gestational age ⬍ 35 weeks, birth weight ⬍ 1800 g, postnatal age at least 5 days, clinically stable (defined as normal blood pressure and no recurrent severe episodes of hypoxemia or bradycardia), and a gastric residual ⬎ 50% of the feed volume given over the previous 3 hours on at least 2 occasions during a 24-hour period. Exclusion criteria included major congenital anomalies, anatomic gastrointestinal abnormalities, suspected or proven necrotizing enterocolitis (NEC) within 7 days before the onset of feeding intolerance, cyanotic heart disease, major gastrointestinal surgery within 2 weeks, suspected clinical or proven sepsis, metabolic or electrolyte disturbances, and therapy with any of the following medications at the onset of feeding intolerance: fentanyl, indomethacin, pancuronium, and vercuronium. The study design was approved by the Ethics Committee at each study site, and informed consent was obtained from the parents of all participants. Sample size was calculated based on the primary outcome: time to full feeding. The number of events (full feeding) were estimated using the formula of George and Desu.25 Assuming that the median time to full feeding was 7 days in the treatment group (based on our pilot study)18 and 15 days in the control group, the estimated hazard ratio was 2.2. With an ␣ error of 0.05 (2-tailed) and a power of detection of 0.8, the number of infants in each group was 23. No censoring was assumed, meaning that infants entering the study remained in the trial until they had reached full enteral feeding. Randomization Enrolled infants were randomly assigned to receive either EM (the EM group) or placebo (the control group) by staff not involved in care of the infants. For treatment allocation purposes, the infants were stratified by gestational age (⬍ 32 or ⱖ 32 weeks), because this was the major influencing factor. A block of 4 randomizations was used to ensure a balance of infants in each allocation. The allocation concealment was kept in an opaque sealed envelope, and the investigators, the patient care team, and the assessors were blinded to the treatment allocation. All of the subjects followed the same feeding protocol. Drug Administration The infants were given 10 mg/kg of EM ethyl succinate every 6 hours for the first 2 days, followed by 4 mg/kg every Efficacy Of Oral Erythromycin For Treatment Of Feeding Intolerance In Preterm Infants
6 hours for another 5 days. The oral route of administration was used because it is simpler and more practical than the intravenous route. The dosing schedule was derived from the results of several studies reported in the literature. Nogami et al 17 reported in 2001 that 2 mg/kg of EM lactobionate given intravenously enhanced gastrointestinal motility even in infants who did not respond to 1 mg/kg of intravenous EM; presumably, 2 mg/kg of intravenous EM produced an effective serum concentration of EM. Other pharmacokinetic studies in preterm infants have shown that at doses of 10 mg/kg, the peak serum concentration of EM at 2 hours was approximately twice as high after intravenous EM than after oral EM ethyl succinate.26,27 It was inferred that an oral EM ethyl succinate dose of 4 mg/kg could produce a similarly effective serum level of EM as an intravenous EM dose of 2 mg/kg; thus, an oral EM ethyl succinate dose of 4 mg/kg was selected for the present study. Incidentally, most studies of low-dose intravenous EM used doses of 1 to 3 mg/kg, which are presumably equivalent to oral EM ethyl succinate doses of 2 to 6 mg/kg.8,16,17 Finally, in studies using low-dose intravenous EM, an EM loading dose of 15 to 30 mg/kg/day was given for the initial few days to ensure adequate blood level of EM.8,16 Thus, we chose an arbitrary loading dose of 40 mg/kg/day for the first 2 days. Our regimen of oral EM ethyl succinate was tested in a pilot study of 10 preterm infants and determined to be safe.18 The EM ethyl succinate (Erysil; Siam Pharmaceutical Ltd, Bangkok, Thailand) was diluted to 40 mg/mL with sterile water and given 30 minutes before feeding. The infants in the control group received a placebo with a similar appearance and ingredients except for the active drug (EM). Both the drug and the placebo were code-numbered and prepared by a staff members not involved in patient care. During the study period, no other prokinetic agents were allowed. Electrocardiography was performed before the start of treatment and again on the fourth day of treatment. Liver function tests were carried out before and after treatment.
Protocol for Enteral and Parenteral Nutrition Infants were started on parenteral nutrition at 2 days of life with 0.5 g/kg of amino acid and 0.5 g/kg of lipid. The doses of both nutrients were subsequently increased in increments of 0.5 mg/kg/day up to a maximum of 3 g/kg/day. The initial glucose infusion rate was 4 to 8 mg/kg/min in increments of 1 to 2 mg/kg/min to a maximum of 10 to 12 mg/kg/min to maintain blood sugar concentrations within the normal range and not higher than 160 mg/dL. Enteral feeding was initiated at day 3 or 4 of life when the infants were clinically stable. Infants were fed their own mothers’ milk whenever possible, but preterm infant formulas were also used in accordance with parents’ preference. Enteral feeding was given through a nasogastric or an orogastric tube as an intermittent bolus in 30 to 60 minutes, beginning with 10 to 20 mL/kg/day and increasing in increments of 10 to 15 mL/kg/ day for infants ⬍ 32 weeks of gestation and 15 to 20 mL/ kg/day for infants ⱖ 32 weeks of gestation. Enteral feeding 601
Table I. Adjustment of enteral feeding Volume of gastric content (% of preceding feed volume)
Gastric content
Amount of subsequent feeding
ⱕ15 16 to 30
Put back Put back
31 to 50 ⱖ51
Put back Discarded
Given as planned Deducted from gastric content Withheld until next feeding Withheld and given next feeding with the volume of previous day feed
was adjusted as a function of the volume of gastric residual capacity, as summarized in Table I. All infants were examined at least twice a day and closely monitored for vomiting, diarrhea, abdominal distention, and volume of gastric residual. Gastric aspirate was measured every 3 hours before each feeding and examined for any bile-stained content. Abdominal circumference was measured before feeding at 12-hour intervals; an increase in abdominal circumference of ⬎ 1.5 cm over a 12-hour interval was considered abnormal. Any vomiting, regurgitation, and bile-stained aspirate were recorded. The enteral feeding was stopped if vomiting occurred more than twice in 24 hours or there were clinical signs and symptoms suggesting NEC or any intra-abdominal pathology. An isolated incidence of bile-stained or blood-stained gastric aspirate with normal physical examination was not an indication for stopping feeding. The duration of withholding feeding was at the discretion of the attending physician. Enteral feeding and EM administration were resumed as soon as the aforementioned signs and symptoms subsided. The resumed feed was started at half the volume given before the feeding was withheld.
Data Analysis The primary outcome was the time to full enteral feeding (150 ml/kg/day) for at least 3 consecutive days after the start of treatment. Secondary outcomes were incidence of NEC and septicemia, length of hospital stay, suspected adverse effects of EM (eg, elevated liver enzymes, hypertrophic pyloric stenosis), and complications related to parenteral nutrition (eg, cholestatic jaundice, catheter-related sepsis). Data were expressed as median, interquartile range, and percent. For comparison, the Mann-Whitney U-test was used for continuous data, and the 2 test or Fisher’s exact test was used for categorical data. Because the primary outcome was time to full enteral feeding, survival analysis was used, and survival curves were constructed for each group. Survival curves between groups were compared using a log-rank test, with the results displayed as Kaplan-Meier survival curves. Cox regression was performed for multivariate analysis using type of feeding (breast milk vs formula or combination), presence of patent ductus arteriosus (PDA), and indomethacin treatment as covariates. SPSS software version 11.0 (SPSS Inc, Chicago, Ill) was used for all statistical tests. Analysis was performed on an intent-to-treat basis. A P value ⬍ .05 was considered statistically significant.
RESULTS Characteristics of the Study Population During the study period, 133 infants were born at ⬍ 32 weeks gestation and 152 were born between 32 and 34 weeks gestation. The respective number of infants from these 2 groups who met the study inclusion criteria were 42 (32%) and 4 (2.6%); thus, a total of 46 preterm infants were enrolled in the study. Each group (EM and control) comprised 23 infants, 21 at ⬍ 32 weeks gestation and 2 at 32 to 34 weeks gestation. Baseline characteristics of the 2 groups are similar, as shown in Table II.
Table II. Baseline characteristics of the study infants Characteristics
Erythromycin (n ⴝ 23), median (IQR)
Placebo (n ⴝ 23), median (IQR)
P
Birth weight (g) Gestational age (weeks) Small for gestational age [n, (%)]* Male gender [n, (%)]* Apgar score at 1 minute Apgar score at 5 minutes FiO2 at enrollment Age at enrollment (days) Prenatal steroid [n, (%)]* Presence of PDA [n, (%)]* Indomethacin use [n, (%)]* Presence of umbilical catheter [n, (%)]* Duration of umbilical catheter (days) Breast milk [n, (%)]* Volume of feeding at enrollment (mL/kg/day)
1100 (870, 1500) 30 (29, 32) 7 (30) 9 (39) 6 (3, 8) 8 (6, 9) 0.21 (0.21, 0.30) 7 (6, 8) 15 (65) 12 (52) 11 (48) 15 (65) 7 (6, 8) 9 (39) 25 (9, 40)
1065 (940, 1270) 29 (28, 31) 3 (13) 13 (56) 7 (4, 9) 9 (8, 10) 0.21 (0.21, 0.30) 6 (6, 8) 17 (74) 17 (73) 14 (61) 19 (82) 8 (6, 10) 4 (17) 15 (7, 31)
0.26 0.65 0.28 0.38 0.55 0.08 0.97 0.41 0.75 0.22 0.55 0.31 0.14 0.19 0.17
*Expressed as number (percentage).
602
Nuntnarumit et al
The Journal of Pediatrics • May 2006
Table III. Clinical outcomes of the two groups Characteristics
Erythromycin (n ⴝ 23), median (IQR)
Placebo (n ⴝ 23), median (IQR)
P
Time to full feeding (days) Numbers of withheld feeds or gastric residuals ⬎ 50% Duration of parenteral nutrition (days) Day to regain birth weight (days) Length of stay (days) Discharge weight (g)
7 (6, 9) 1 (0, 2) 13 (11, 15) 11 (10, 14) 46 (24, 74) 2170 (1987, 2587)
13 (9, 15) 9 (2, 13) 17 (13, 25) 12 (11, 15) 60 (43, 89) 2560 (2130, 3600)
⬍0.001* ⬍0.001* 0.03* 0.49 0.07 0.06
*Mann-Whitney U test, P ⬍ .05.
Time to Full Feeding After Enrollment Time to full feeding after the beginning of treatment was significantly shorter in the EM group than in the control group (median, 7 days vs 13 days; P ⬍ .001) (Table III). In the EM group, 50% of the infants reached full feeding in 1 week during the course of treatment, and all achieved full feeding within 2 weeks. In comparison, in the control group, only half of the infants were able to achieve the equivalent feedings during the same period (Figure). The effect of EM was also seen when the data were analyzed separately according to age strata (⬍ 32 weeks vs ⱖ 32 weeks). Nine infants in the EM group and 4 infants in the control group were breast-fed. When the type of feeding was entered into the Cox regression analysis model, treatment with EM was the only significant independent variable, with a hazard ratio of 4.55 (95% confidence interval ⫽ 2.22 to 9.34; P ⬍ .001). Similarly, when PDA or indomethacin use was selected as a covariate in the regression analysis, treatment with EM remained the only significant independent variable. Secondary Outcomes Duration of parenteral nutrition, number of withheld feeds, and significant gastric residual (⬎ 50% of feeding volume) were significantly lower in the EM group compared with the control group (Table III). There were trends toward shorter lengths of hospital stay and lower discharge weights in the EM group. Complications of Parenteral Nutrition and Adverse Effects Related to EM There were no significant differences in the incidence of sepsis, NEC, and cholestatic jaundice and in mortality rate between the 2 groups (Table IV). The organism causing sepsis in the EM group was coagulase-negative Staphylococcus in all 3 infants. In the control group, the causative organism was coagulase-negative Staphylococcus in 2 infants and Streptococcus group D in the other 2 infants. Two infants in EM group died, 1 from severe bronchopulmonary dysplasia on the 92nd day of life and the other from NEC stage III occurring after 4 days of full feeding and 11 days after discontinuation of EM. Three infants in the control group developed NEC stage II. No significant adverse effects related to EM (eg, Efficacy Of Oral Erythromycin For Treatment Of Feeding Intolerance In Preterm Infants
Figure. Probability of achieving full feeding in the EM group and in the control group.
elevated liver enzymes, abnormal electrocardiogram, pyloric stenosis) were observed.
DISCUSSION The present study was a randomized controlled trial of EM involving 2 comparable groups of preterm infants with feeding intolerance. EM was efficacious in improving feeding tolerance in quite immature infants, almost all of whom were ⬍ 32 weeks gestation. The results are encouraging, as evidenced by the differences in the time to full feeding and the number of withheld feeds with secondary reduction in the duration of parenteral nutrition between the EM group and the control group. Although more infants in the control group were formula-fed, developed PDA, or received indomethacin treatment, any of which could increase feeding intolerance, Cox regression analysis still showed that EM treatment was efficacious in improving feeding tolerance. Studies of EM for feeding intolerance in preterm infants published in the English literature are of limited quantity and quality.8,10,11,16 –21 The results are inconsistent and contradictory, making comparison with our findings difficult. The discrepancies are due to many factors, particularly the 603
Table IV. Morbidity and mortality of the 2 groups
Characteristics
EM group, n (%)
Control group, n (%)
Sepsis NEC ⱖ stage II Cholestatic jaundice Prolonged QTc interval Hypertrophic pyloric stenosis Dead
3 (13) 1 (4) 3 (13) 0 0 2 (9)
4 (17) 3 (13) 2 (9) 0 0 0
P 1.0 .61 1.0
.49
severity of feeding intolerance and the dose of EM, both of which are not clearly defined. A recent systematic review conducted in 2005 found a total of 7 studies eligible for analysis.12 Two of the 3 randomized controlled trials involving the prophylactic approach using an antimicrobial dose (45 mg/kg/day given intravenously in 1 study 13 and 48 mg/kg/day given enterally in the other study14) showed no beneficial effect on feeding in preterm infants; however, the third study 15 using low-dose EM (10 mg/kg/day given orally) found EM to be useful. In contrast, 1 of the 4 randomized placebo control studies used a rescue approach with an antimicrobial dose of EM of 50 mg/kg/day given enterally and found EM to be useful.19 Low-dose EM (6 to 15 mg/kg/day given intravenously in 1 study and given enterally in the other 2 studies) showed no beneficial effect.20,21,28 In all likelihood, the low EM dose used in the rescue treatment is not sufficient to affect the gastrointestinal tract. Ng et al 21 reported the efficacy of an oral antimicrobial dose of EM for feeding intolerance in preterm infants. Our study involved rescue treatment with oral EM ethyl succinate at a dose between the low dose and the antimicrobial dose (40 mg/kg/day for the first 2 days, followed by 16 mg/kg/day for another 5 days). The results demonstrate that this intermediate dose is adequate to ameliorate feeding intolerance in infants ⬍ 32 weeks gestational age and confirm the positive findings reported by Ng et al. Our pilot study in 10 preterm neonates also found EM to be useful in promoting enteral feeding.18 These favorable results substantiate the usefulness of oral EM in this application. The present study used an intermediate dose of EM for only 7 days, compared with an antimicrobial dose for 14 days in the study of Ng et al.21 The feeding intolerance in our study could be considered mild to moderate, based on the fact that our subjects were enrolled at an early age and almost all of them tolerated full feeding within 3 weeks after enrollment. In contrast, the study of Ng et al 21 involved a more advanced age at study entry and a mean of 31 days to achieve full feeding, indicating a more severe degree of feeding intolerance in their subjects. A milder degree of feeding intolerance might indicate a more mature gastrointestinal motility network and thus a better response to exogenous motilin agonist. It seems logical that mild feeding intolerance requires a smaller EM dose and a shorter duration of treatment, whereas 604
Nuntnarumit et al
severe feeding intolerance needs a larger dose and a longer duration of therapy. In this regard, the results of these 2 studies are mutually supportive. It remains to be seen whether our EM regimen will also be efficacious for severe feeding intolerance. The mechanisms whereby EM may help advance feeding in preterm neonates are still not well understood. The motilin-like activity of EM on the MMC has been the most commonly cited mechanism to explain the efficacy of EM. However, in preterm infants, MMC is absent before 32 weeks gestation and does not appear in its mature form until 35 weeks gestation;29 therefore, EM may not have the same effects on the MMC as in term neonates, children, or adults. Other mechanisms have been proposed, but none appears to satisfactorily explain this prokinetic action of EM in preterm infants.30-33 The intermediate dose of EM and the duration used in the present study seem not only adequate, but also devoid of any adverse effects. There were no differences between the EM group and the control group in terms of the incidence of cholestatic jaundice or elevated liver enzyme levels. There were no cases of cardiac arrhythmia or hypertrophic pyloric stenosis. Exposure to an antimicrobial dose of EM for more than 14 days was associated with a 10-fold increase in the incidence of hypertrophic pyloric stenosis.24 The incidences of sepsis and NEC were similar in the 2 groups. These complications also were not reported in previous studies using higher EM doses and longer treatment durations for either prophylactic or rescue proposes.12 However, it should be emphasized that the sample size in the present study was calculated to have sufficient power to enable evaluation of the efficacy of EM, but not of the incidence of adverse effects. Our results are promising and lend support to considering oral EM for preterm infants with feeding intolerance; however, a recommendation for routine use of the present regimen awaits further confirmation from larger clinical trials. We express our sincere thanks and gratitude to Phienvit Tantibhedhyangkul, Amnuay Thithapandha, Sumitr Sutra, Sungkom Jongpiputavanich, and Mahippathorn Chinnapha for their valuable comments and suggestions, and to Marion Haynes for her editorial assistance.
REFERENCES 1. Berseth CL. Neonatal small intestinal motility: motor responses to feeding in term and preterm infants. J Pediatr 1990;117:777-82. 2. Berseth CL. Gastrointestinal motility in the neonate. Clin Perinatol 1996;23:179-90. 3. Asai T. Erythromycin and motilin as gastrointestinal prokinetic agents. Eur J Anaesthesiol 1997;14:3-6. 4. Curry JI, Lander TD, Stringer MD. Review article: erythromycin as a prokinetic agent in infants and children. Aliment Pharmacol Ther 2001;15:595-603. 5. Jadcherla SR, Berseth CL. Effect of erythromycin on gastroduodenal contractile activity in developing neonates. J Pediatr Gastroenterol Nutr 2002;34:16-22. 6. Tomomasa T, Miyazaki M, Koizumi T, Kuroume T. Erythromycin increases gastric antral motility in human premature infants. Biol Neonate 1993;63:349-52.
The Journal of Pediatrics • May 2006
7. Itoh Z, Suzuki T, Nakaya M. Gastrointestinal motor-stimulating activity of macrolide antibiotics and analysis of their effects on the canine gut. Antimicrob Agents Chemother 1984;26:863-9. 8. Kubota M, Nakamura T, Motokura T, Mori S, Nishida A. Erythromycin improves gastrointestinal motility in extremely low birth weight infants. Acta Paediatr Jpn 1994;36:198-201. 9. Coulie B, Tack J, Peeters T, Janssens J. Involvement of two different pathways in the motor effects of erythromycin on the gastric antrum in humans. Gut 1998;43:395-400. 10. Costalos C, Gavrili V, Skouteri V, Gounaris A. The effect of low-dose erythromycin on whole gastrointestinal transit time of preterm infants. Early Hum Dev 2001;65:91-6. 11. Dellagrammaticas HD, Iacovidou N, Megaloyanni E, Papadimitriou M, Kapetanakis J. Effect of low-dose oral erythromycin on gastric aspirates in ventilated neonates less than 32 weeks of gestation: preliminary results. Biol Neonate 2002;81:213-6. 12. Patole S, Rao S, Doherty D. Erythromycin as a prokinetic agent in preterm neonates: a systematic review. Arch Dis Child Fetal Neonatal Ed 2005;90(4):F301-6. 13. Stenson BJ, Middlemist L, Lyon AJ. Influence of erythromycin on establishment of feeding in preterm infants: observations from a randomised controlled trial. Arch Dis Child Fetal Neonatal Ed 1998;79:F212-4. 14. Patole SK, Almonte R, Kadalraja R, Tuladhar R, Muller R, Whitehall JS. Can prophylactic oral erythromycin reduce time to full enteral feeds in preterm neonates? Int J Clin Pract 2000;54:504-8. 15. Oei J, Lui K. A placebo-controlled trial of low-dose erythromycin to promote feed tolerance in preterm infants. Acta Paediatr 2001;90:904-8. 16. Su BH, Lin HC, Peng CT, Tsai CH. Effect of erythromycin on feeding intolerance in very low birth weight infants: a preliminary observation. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 1998;39:324-6. 17. Nogami K, Nishikubo T, Minowa H, Uchida Y, Kamitsuji H, Takahashi Y. Intravenous low-dose erythromycin administration for infants with feeding intolerance. Pediatr Int 2001;43:605-10. 18. Sekteera W, Nuntnarumit P, Supapannachart S. Oral erythromycin for treatment of feeding intolerance in preterm infants: a preliminary report. J Med Assoc Thai 2002;85(suppl 4):S1177-82. 19. Ng PC, So KW, Fung KS, Lee CH, Fok TF, Wong E, et al. Randomised controlled study of oral erythromycin for treatment of gastrointestinal dysmotility in preterm infants. Arch Dis Child Fetal Neonatal Ed 2001;84:F177-82. 20. El Hennawy AA, Sparks JW, Armentrout D, Huseby V, Berseth CL.
Efficacy Of Oral Erythromycin For Treatment Of Feeding Intolerance In Preterm Infants
Erythromycin fails to improve feeding outcome in feeding-intolerant preterm infants. J Pediatr Gastroenterol Nutr 2003;37:281-6. 21. Ng SC, Gomez JM, Rajadurai VS, Saw SM, Quak SH. Establishing enteral feeding in preterm infants with feeding intolerance: a randomized controlled study of low-dose erythromycin. J Pediatr Gastroenterol Nutr 2003;37:554-8. 22. Sims PJ, Waites KB, Crouse DT. Erythromycin lactobionate toxicity in preterm neonates. Pediatr Infect Dis J 1994;13:164-7. 23. Gouyon JB, Benoit A, Betremieux P, Sandre D, Sgro C, Bavoux F, et al. Cardiac toxicity of intravenous erythromycin lactobionate in preterm infants. Pediatr Infect Dis J 1994;13:840-1. 24. Mahon BE, Rosenman MB, Kleiman MB. Maternal and infant use of erythromycin and other macrolide antibiotics as risk factors for infantile hypertrophic pyloric stenosis. J Pediatr 2001;139:380-4. 25. George SL, Desu MM. Planning the size and duration of a clinical trial studying the time to some critical event. J Chronic Dis 1974;27:15-24. 26. Patamasucon P, Kaojarern S, Kusmiesz H, Nelson JD. Pharmacokinetics of erythromycin ethylsuccinate and estolate in infants under 4 months of age. Antimicrob Agents Chemother 1981;19:736-9. 27. Waites KB, Sims PJ, Crouse DT, Geerts MH, Shoup RE, Hamrick WB, et al. Serum concentrations of erythromycin after intravenous infusion in preterm neonates treated for Ureaplasma urealyticum infection. Pediatr Infect Dis J 1994;13:287-93. 28. Cairns P, Craig S, Tubman R. Randomised controlled trial of low-dose erythromycin in preterm infants with feed intolerance [abstract]. Arch Dis Child 2002;86(suppl I):G24. 29. Jadcherla SR, Klee G, Berseth CL. Regulation of migrating motor complexes by motilin and pancreatic polypeptide in human infants. Pediatr Res 1997;42:365-9. 30. Tomomasa T, Kuroume T, Arai H, Wakabayashi K, Itoh Z. Erythromycin induces migrating motor complex in human gastrointestinal tract. Dig Dis Sci 1986;31:157-61. 31. Peeters T. Erythromycin and other macrolides as prokinetic agents. Gastroenterology 1993;105:1886-99. 32. Tack J, Janssens J, Vantrappen G, Peeters T, Annese V, Depoortere I, et al. Effect of erythromycin on gastric motility in controls and in diabetic gastroparesis. Gastroenterology 1992;103:72-9. 33. Tomomasa T, Tabata M, Nako Y. The effect of oral erythromycin on gastric emptying in neonates with gastric retention [abstract]. Gastroenterology 1997;112(suppl):A839.
605
AND
Objective To determine the efficacy and safety of oral erythromycin (EM) for feeding intolerance in preterm infants < 35 weeks gestation.
Study design In this randomized, double-blinded, placebo-controlled trial, preterm infants with feeding intolerance were randomly allocated to a treatment group given EM ethyl succinate 10 mg/kg every 6 hours for 2 days, followed by 4 mg/kg every 6 hours for another 5 days, or to a control group given placebo. The primary outcome was time to full feeding (150 mL/kg/day) after the start of treatment. Results Each group comprised 23 preterm infants, almost all of whom were < 32 weeks gestation. Baseline characteristics were similar between the 2 groups. Times to full feeding were significantly shorter and the number of withheld feeds were significantly less in the EM group than the control group; the respective medians (interquartile ranges) were 7 days (6 to 9 days) versus 13 days (9 to 15 days) (P < .001) and 1 episode (0 to 2 episodes) versus 9 episodes (2 to 13 episodes) (P < .001). No significant differences in episodes of sepsis, necrotizing enterocolitis, and cholestasis were observed. Conclusions Oral EM was effective and safe for treatment of feeding intolerance in preterm infants. (J Pediatr 2006;148:600-5)
dvancing feedings to full enteral feeding is frequently difficult, especially in very low birth weight infants, due to feeding intolerance, and may take up to 2 to 3 weeks. This problem necessitates prolonged parenteral nutrition with its attendant complications. Feeding intolerance is often attributable to absence of the migratory motor complex (MMC) in the small intestine and the lack of mature response to feeding, resulting in gastrointestinal dysmotility, characterized by delayed gastric emptying and prolonged gastrointestinal transit time.1,2 From the Department of Pediatrics, Faculty Erythromycin (EM), a macrolide antibiotic, has potent prokinetic activity through of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand; Department its agonist action on motilin receptors in the stomach and proximal small intestine in both of Pediatrics, Faculty of Medicine, Srinagar3-5 humans and animals. EM at low doses (3 to 12 mg/kg/day) induces phase III MMC ind Hospital, Khon Kaen University, Khon Kaen, Thailand; Department of Pediatrics, of the small intestine through activation of motilin neuronal receptors and at high doses Faculty of Medicine, Chiang Mai University, (antimicrobial doses, ⱖ 40 mg/kg/day) initiates antral contractility and causes sustained Chiang Mai, Thailand; and Department of 6-11 contractility of the small intestine through stimulation of motilin muscular receptors. Pediatrics, Faculty of Medicine, Srinakarinwirot University, Bangkok, Thailand. However, a recent systematic analysis of trials of EM as a prokinetic agent in preterm Supported by the Ramathibodi Fund (grant infants revealed conflicting results due to differences among the trials including doses, 46018). routes and modes of administration, durations of treatment, feeding protocols, and The medication (erythromycin ethyl succinate) and placebo used in this study were definitions of feeding intolerance.12 Most of the trials failed to show any benefit for the provided by Siam Pharmaceutical Ltd, prophylactic use of EM on the day that feedings were started.13-15 In terms of the Bangkok, Thailand, which had no involvetherapeutic use of EM for feeding intolerance, although several clinical trials have ment in any aspect of the study, including study design; data collection, analysis, and reported favorable outcomes, these trials were not randomized and controlled and ininterpretation; manuscript preparation; and 10,11,16 –18 To date there have been only 3 randomized volved only small sample sizes. the decision to submit the manuscript for publication. The first author wrote the first controlled trials,19-21 1 with high doses (48 mg/kg/day) reporting a shorter time to full draft of the manuscript; no grant or other feeding and reduced cholestasis and the other 2 with low doses (6 to 15 mg/kg/day) form of payment was received by anyone reporting no beneficial effect. involved in manuscript preparation. Submitted for publication Jun 18, 2005; last Currently the role of EM in the management of feeding intolerance in premature revision received Nov 3, 2005; accepted infants remains equivocal and warrants further study. Low doses appear to be ineffective, Dec 7, 2005. whereas high (antimicrobial) doses may cause such adverse effects as hypertrophic pyloric Reprint requests: Pracha Nuntnarumit, MD, Department of Pediatrics, Faculty of Medistenosis, cardiac arrhythmias, and elevated liver enzymes.22-24 An intermediate dose may cine, Ramathibodi Hospital, Mahidol Uniprove to be effective while minimizing adverse effects. Consequently, we conducted a versity, Rama VI, Bangkok, 10400, Thailand.
A
EM MMC
600
Erythromycin Migratory motor complex
NEC PDA
Necrotizing enterocolitis Patent ductus arteriosus
E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2006 Elsevier Inc. All rights reserved. 10.1016/j.jpeds.2005.12.026
randomized, placebo-controlled trial to determine whether or not an intermediate dose of oral EM would benefit preterm infants with feeding intolerance and to assess the adverse effects of this treatment modality.
METHODS Patients Preterm infants admitted to neonatal care nurseries of 3 university hospitals (Ramathibodi Hospital, Srinagarind Hospital, and Chiang Mai University Hospital) between December 2002 and February 2005 were consecutively enrolled in the trial. Inclusion criteria included gestational age ⬍ 35 weeks, birth weight ⬍ 1800 g, postnatal age at least 5 days, clinically stable (defined as normal blood pressure and no recurrent severe episodes of hypoxemia or bradycardia), and a gastric residual ⬎ 50% of the feed volume given over the previous 3 hours on at least 2 occasions during a 24-hour period. Exclusion criteria included major congenital anomalies, anatomic gastrointestinal abnormalities, suspected or proven necrotizing enterocolitis (NEC) within 7 days before the onset of feeding intolerance, cyanotic heart disease, major gastrointestinal surgery within 2 weeks, suspected clinical or proven sepsis, metabolic or electrolyte disturbances, and therapy with any of the following medications at the onset of feeding intolerance: fentanyl, indomethacin, pancuronium, and vercuronium. The study design was approved by the Ethics Committee at each study site, and informed consent was obtained from the parents of all participants. Sample size was calculated based on the primary outcome: time to full feeding. The number of events (full feeding) were estimated using the formula of George and Desu.25 Assuming that the median time to full feeding was 7 days in the treatment group (based on our pilot study)18 and 15 days in the control group, the estimated hazard ratio was 2.2. With an ␣ error of 0.05 (2-tailed) and a power of detection of 0.8, the number of infants in each group was 23. No censoring was assumed, meaning that infants entering the study remained in the trial until they had reached full enteral feeding. Randomization Enrolled infants were randomly assigned to receive either EM (the EM group) or placebo (the control group) by staff not involved in care of the infants. For treatment allocation purposes, the infants were stratified by gestational age (⬍ 32 or ⱖ 32 weeks), because this was the major influencing factor. A block of 4 randomizations was used to ensure a balance of infants in each allocation. The allocation concealment was kept in an opaque sealed envelope, and the investigators, the patient care team, and the assessors were blinded to the treatment allocation. All of the subjects followed the same feeding protocol. Drug Administration The infants were given 10 mg/kg of EM ethyl succinate every 6 hours for the first 2 days, followed by 4 mg/kg every Efficacy Of Oral Erythromycin For Treatment Of Feeding Intolerance In Preterm Infants
6 hours for another 5 days. The oral route of administration was used because it is simpler and more practical than the intravenous route. The dosing schedule was derived from the results of several studies reported in the literature. Nogami et al 17 reported in 2001 that 2 mg/kg of EM lactobionate given intravenously enhanced gastrointestinal motility even in infants who did not respond to 1 mg/kg of intravenous EM; presumably, 2 mg/kg of intravenous EM produced an effective serum concentration of EM. Other pharmacokinetic studies in preterm infants have shown that at doses of 10 mg/kg, the peak serum concentration of EM at 2 hours was approximately twice as high after intravenous EM than after oral EM ethyl succinate.26,27 It was inferred that an oral EM ethyl succinate dose of 4 mg/kg could produce a similarly effective serum level of EM as an intravenous EM dose of 2 mg/kg; thus, an oral EM ethyl succinate dose of 4 mg/kg was selected for the present study. Incidentally, most studies of low-dose intravenous EM used doses of 1 to 3 mg/kg, which are presumably equivalent to oral EM ethyl succinate doses of 2 to 6 mg/kg.8,16,17 Finally, in studies using low-dose intravenous EM, an EM loading dose of 15 to 30 mg/kg/day was given for the initial few days to ensure adequate blood level of EM.8,16 Thus, we chose an arbitrary loading dose of 40 mg/kg/day for the first 2 days. Our regimen of oral EM ethyl succinate was tested in a pilot study of 10 preterm infants and determined to be safe.18 The EM ethyl succinate (Erysil; Siam Pharmaceutical Ltd, Bangkok, Thailand) was diluted to 40 mg/mL with sterile water and given 30 minutes before feeding. The infants in the control group received a placebo with a similar appearance and ingredients except for the active drug (EM). Both the drug and the placebo were code-numbered and prepared by a staff members not involved in patient care. During the study period, no other prokinetic agents were allowed. Electrocardiography was performed before the start of treatment and again on the fourth day of treatment. Liver function tests were carried out before and after treatment.
Protocol for Enteral and Parenteral Nutrition Infants were started on parenteral nutrition at 2 days of life with 0.5 g/kg of amino acid and 0.5 g/kg of lipid. The doses of both nutrients were subsequently increased in increments of 0.5 mg/kg/day up to a maximum of 3 g/kg/day. The initial glucose infusion rate was 4 to 8 mg/kg/min in increments of 1 to 2 mg/kg/min to a maximum of 10 to 12 mg/kg/min to maintain blood sugar concentrations within the normal range and not higher than 160 mg/dL. Enteral feeding was initiated at day 3 or 4 of life when the infants were clinically stable. Infants were fed their own mothers’ milk whenever possible, but preterm infant formulas were also used in accordance with parents’ preference. Enteral feeding was given through a nasogastric or an orogastric tube as an intermittent bolus in 30 to 60 minutes, beginning with 10 to 20 mL/kg/day and increasing in increments of 10 to 15 mL/kg/ day for infants ⬍ 32 weeks of gestation and 15 to 20 mL/ kg/day for infants ⱖ 32 weeks of gestation. Enteral feeding 601
Table I. Adjustment of enteral feeding Volume of gastric content (% of preceding feed volume)
Gastric content
Amount of subsequent feeding
ⱕ15 16 to 30
Put back Put back
31 to 50 ⱖ51
Put back Discarded
Given as planned Deducted from gastric content Withheld until next feeding Withheld and given next feeding with the volume of previous day feed
was adjusted as a function of the volume of gastric residual capacity, as summarized in Table I. All infants were examined at least twice a day and closely monitored for vomiting, diarrhea, abdominal distention, and volume of gastric residual. Gastric aspirate was measured every 3 hours before each feeding and examined for any bile-stained content. Abdominal circumference was measured before feeding at 12-hour intervals; an increase in abdominal circumference of ⬎ 1.5 cm over a 12-hour interval was considered abnormal. Any vomiting, regurgitation, and bile-stained aspirate were recorded. The enteral feeding was stopped if vomiting occurred more than twice in 24 hours or there were clinical signs and symptoms suggesting NEC or any intra-abdominal pathology. An isolated incidence of bile-stained or blood-stained gastric aspirate with normal physical examination was not an indication for stopping feeding. The duration of withholding feeding was at the discretion of the attending physician. Enteral feeding and EM administration were resumed as soon as the aforementioned signs and symptoms subsided. The resumed feed was started at half the volume given before the feeding was withheld.
Data Analysis The primary outcome was the time to full enteral feeding (150 ml/kg/day) for at least 3 consecutive days after the start of treatment. Secondary outcomes were incidence of NEC and septicemia, length of hospital stay, suspected adverse effects of EM (eg, elevated liver enzymes, hypertrophic pyloric stenosis), and complications related to parenteral nutrition (eg, cholestatic jaundice, catheter-related sepsis). Data were expressed as median, interquartile range, and percent. For comparison, the Mann-Whitney U-test was used for continuous data, and the 2 test or Fisher’s exact test was used for categorical data. Because the primary outcome was time to full enteral feeding, survival analysis was used, and survival curves were constructed for each group. Survival curves between groups were compared using a log-rank test, with the results displayed as Kaplan-Meier survival curves. Cox regression was performed for multivariate analysis using type of feeding (breast milk vs formula or combination), presence of patent ductus arteriosus (PDA), and indomethacin treatment as covariates. SPSS software version 11.0 (SPSS Inc, Chicago, Ill) was used for all statistical tests. Analysis was performed on an intent-to-treat basis. A P value ⬍ .05 was considered statistically significant.
RESULTS Characteristics of the Study Population During the study period, 133 infants were born at ⬍ 32 weeks gestation and 152 were born between 32 and 34 weeks gestation. The respective number of infants from these 2 groups who met the study inclusion criteria were 42 (32%) and 4 (2.6%); thus, a total of 46 preterm infants were enrolled in the study. Each group (EM and control) comprised 23 infants, 21 at ⬍ 32 weeks gestation and 2 at 32 to 34 weeks gestation. Baseline characteristics of the 2 groups are similar, as shown in Table II.
Table II. Baseline characteristics of the study infants Characteristics
Erythromycin (n ⴝ 23), median (IQR)
Placebo (n ⴝ 23), median (IQR)
P
Birth weight (g) Gestational age (weeks) Small for gestational age [n, (%)]* Male gender [n, (%)]* Apgar score at 1 minute Apgar score at 5 minutes FiO2 at enrollment Age at enrollment (days) Prenatal steroid [n, (%)]* Presence of PDA [n, (%)]* Indomethacin use [n, (%)]* Presence of umbilical catheter [n, (%)]* Duration of umbilical catheter (days) Breast milk [n, (%)]* Volume of feeding at enrollment (mL/kg/day)
1100 (870, 1500) 30 (29, 32) 7 (30) 9 (39) 6 (3, 8) 8 (6, 9) 0.21 (0.21, 0.30) 7 (6, 8) 15 (65) 12 (52) 11 (48) 15 (65) 7 (6, 8) 9 (39) 25 (9, 40)
1065 (940, 1270) 29 (28, 31) 3 (13) 13 (56) 7 (4, 9) 9 (8, 10) 0.21 (0.21, 0.30) 6 (6, 8) 17 (74) 17 (73) 14 (61) 19 (82) 8 (6, 10) 4 (17) 15 (7, 31)
0.26 0.65 0.28 0.38 0.55 0.08 0.97 0.41 0.75 0.22 0.55 0.31 0.14 0.19 0.17
*Expressed as number (percentage).
602
Nuntnarumit et al
The Journal of Pediatrics • May 2006
Table III. Clinical outcomes of the two groups Characteristics
Erythromycin (n ⴝ 23), median (IQR)
Placebo (n ⴝ 23), median (IQR)
P
Time to full feeding (days) Numbers of withheld feeds or gastric residuals ⬎ 50% Duration of parenteral nutrition (days) Day to regain birth weight (days) Length of stay (days) Discharge weight (g)
7 (6, 9) 1 (0, 2) 13 (11, 15) 11 (10, 14) 46 (24, 74) 2170 (1987, 2587)
13 (9, 15) 9 (2, 13) 17 (13, 25) 12 (11, 15) 60 (43, 89) 2560 (2130, 3600)
⬍0.001* ⬍0.001* 0.03* 0.49 0.07 0.06
*Mann-Whitney U test, P ⬍ .05.
Time to Full Feeding After Enrollment Time to full feeding after the beginning of treatment was significantly shorter in the EM group than in the control group (median, 7 days vs 13 days; P ⬍ .001) (Table III). In the EM group, 50% of the infants reached full feeding in 1 week during the course of treatment, and all achieved full feeding within 2 weeks. In comparison, in the control group, only half of the infants were able to achieve the equivalent feedings during the same period (Figure). The effect of EM was also seen when the data were analyzed separately according to age strata (⬍ 32 weeks vs ⱖ 32 weeks). Nine infants in the EM group and 4 infants in the control group were breast-fed. When the type of feeding was entered into the Cox regression analysis model, treatment with EM was the only significant independent variable, with a hazard ratio of 4.55 (95% confidence interval ⫽ 2.22 to 9.34; P ⬍ .001). Similarly, when PDA or indomethacin use was selected as a covariate in the regression analysis, treatment with EM remained the only significant independent variable. Secondary Outcomes Duration of parenteral nutrition, number of withheld feeds, and significant gastric residual (⬎ 50% of feeding volume) were significantly lower in the EM group compared with the control group (Table III). There were trends toward shorter lengths of hospital stay and lower discharge weights in the EM group. Complications of Parenteral Nutrition and Adverse Effects Related to EM There were no significant differences in the incidence of sepsis, NEC, and cholestatic jaundice and in mortality rate between the 2 groups (Table IV). The organism causing sepsis in the EM group was coagulase-negative Staphylococcus in all 3 infants. In the control group, the causative organism was coagulase-negative Staphylococcus in 2 infants and Streptococcus group D in the other 2 infants. Two infants in EM group died, 1 from severe bronchopulmonary dysplasia on the 92nd day of life and the other from NEC stage III occurring after 4 days of full feeding and 11 days after discontinuation of EM. Three infants in the control group developed NEC stage II. No significant adverse effects related to EM (eg, Efficacy Of Oral Erythromycin For Treatment Of Feeding Intolerance In Preterm Infants
Figure. Probability of achieving full feeding in the EM group and in the control group.
elevated liver enzymes, abnormal electrocardiogram, pyloric stenosis) were observed.
DISCUSSION The present study was a randomized controlled trial of EM involving 2 comparable groups of preterm infants with feeding intolerance. EM was efficacious in improving feeding tolerance in quite immature infants, almost all of whom were ⬍ 32 weeks gestation. The results are encouraging, as evidenced by the differences in the time to full feeding and the number of withheld feeds with secondary reduction in the duration of parenteral nutrition between the EM group and the control group. Although more infants in the control group were formula-fed, developed PDA, or received indomethacin treatment, any of which could increase feeding intolerance, Cox regression analysis still showed that EM treatment was efficacious in improving feeding tolerance. Studies of EM for feeding intolerance in preterm infants published in the English literature are of limited quantity and quality.8,10,11,16 –21 The results are inconsistent and contradictory, making comparison with our findings difficult. The discrepancies are due to many factors, particularly the 603
Table IV. Morbidity and mortality of the 2 groups
Characteristics
EM group, n (%)
Control group, n (%)
Sepsis NEC ⱖ stage II Cholestatic jaundice Prolonged QTc interval Hypertrophic pyloric stenosis Dead
3 (13) 1 (4) 3 (13) 0 0 2 (9)
4 (17) 3 (13) 2 (9) 0 0 0
P 1.0 .61 1.0
.49
severity of feeding intolerance and the dose of EM, both of which are not clearly defined. A recent systematic review conducted in 2005 found a total of 7 studies eligible for analysis.12 Two of the 3 randomized controlled trials involving the prophylactic approach using an antimicrobial dose (45 mg/kg/day given intravenously in 1 study 13 and 48 mg/kg/day given enterally in the other study14) showed no beneficial effect on feeding in preterm infants; however, the third study 15 using low-dose EM (10 mg/kg/day given orally) found EM to be useful. In contrast, 1 of the 4 randomized placebo control studies used a rescue approach with an antimicrobial dose of EM of 50 mg/kg/day given enterally and found EM to be useful.19 Low-dose EM (6 to 15 mg/kg/day given intravenously in 1 study and given enterally in the other 2 studies) showed no beneficial effect.20,21,28 In all likelihood, the low EM dose used in the rescue treatment is not sufficient to affect the gastrointestinal tract. Ng et al 21 reported the efficacy of an oral antimicrobial dose of EM for feeding intolerance in preterm infants. Our study involved rescue treatment with oral EM ethyl succinate at a dose between the low dose and the antimicrobial dose (40 mg/kg/day for the first 2 days, followed by 16 mg/kg/day for another 5 days). The results demonstrate that this intermediate dose is adequate to ameliorate feeding intolerance in infants ⬍ 32 weeks gestational age and confirm the positive findings reported by Ng et al. Our pilot study in 10 preterm neonates also found EM to be useful in promoting enteral feeding.18 These favorable results substantiate the usefulness of oral EM in this application. The present study used an intermediate dose of EM for only 7 days, compared with an antimicrobial dose for 14 days in the study of Ng et al.21 The feeding intolerance in our study could be considered mild to moderate, based on the fact that our subjects were enrolled at an early age and almost all of them tolerated full feeding within 3 weeks after enrollment. In contrast, the study of Ng et al 21 involved a more advanced age at study entry and a mean of 31 days to achieve full feeding, indicating a more severe degree of feeding intolerance in their subjects. A milder degree of feeding intolerance might indicate a more mature gastrointestinal motility network and thus a better response to exogenous motilin agonist. It seems logical that mild feeding intolerance requires a smaller EM dose and a shorter duration of treatment, whereas 604
Nuntnarumit et al
severe feeding intolerance needs a larger dose and a longer duration of therapy. In this regard, the results of these 2 studies are mutually supportive. It remains to be seen whether our EM regimen will also be efficacious for severe feeding intolerance. The mechanisms whereby EM may help advance feeding in preterm neonates are still not well understood. The motilin-like activity of EM on the MMC has been the most commonly cited mechanism to explain the efficacy of EM. However, in preterm infants, MMC is absent before 32 weeks gestation and does not appear in its mature form until 35 weeks gestation;29 therefore, EM may not have the same effects on the MMC as in term neonates, children, or adults. Other mechanisms have been proposed, but none appears to satisfactorily explain this prokinetic action of EM in preterm infants.30-33 The intermediate dose of EM and the duration used in the present study seem not only adequate, but also devoid of any adverse effects. There were no differences between the EM group and the control group in terms of the incidence of cholestatic jaundice or elevated liver enzyme levels. There were no cases of cardiac arrhythmia or hypertrophic pyloric stenosis. Exposure to an antimicrobial dose of EM for more than 14 days was associated with a 10-fold increase in the incidence of hypertrophic pyloric stenosis.24 The incidences of sepsis and NEC were similar in the 2 groups. These complications also were not reported in previous studies using higher EM doses and longer treatment durations for either prophylactic or rescue proposes.12 However, it should be emphasized that the sample size in the present study was calculated to have sufficient power to enable evaluation of the efficacy of EM, but not of the incidence of adverse effects. Our results are promising and lend support to considering oral EM for preterm infants with feeding intolerance; however, a recommendation for routine use of the present regimen awaits further confirmation from larger clinical trials. We express our sincere thanks and gratitude to Phienvit Tantibhedhyangkul, Amnuay Thithapandha, Sumitr Sutra, Sungkom Jongpiputavanich, and Mahippathorn Chinnapha for their valuable comments and suggestions, and to Marion Haynes for her editorial assistance.
REFERENCES 1. Berseth CL. Neonatal small intestinal motility: motor responses to feeding in term and preterm infants. J Pediatr 1990;117:777-82. 2. Berseth CL. Gastrointestinal motility in the neonate. Clin Perinatol 1996;23:179-90. 3. Asai T. Erythromycin and motilin as gastrointestinal prokinetic agents. Eur J Anaesthesiol 1997;14:3-6. 4. Curry JI, Lander TD, Stringer MD. Review article: erythromycin as a prokinetic agent in infants and children. Aliment Pharmacol Ther 2001;15:595-603. 5. Jadcherla SR, Berseth CL. Effect of erythromycin on gastroduodenal contractile activity in developing neonates. J Pediatr Gastroenterol Nutr 2002;34:16-22. 6. Tomomasa T, Miyazaki M, Koizumi T, Kuroume T. Erythromycin increases gastric antral motility in human premature infants. Biol Neonate 1993;63:349-52.
The Journal of Pediatrics • May 2006
7. Itoh Z, Suzuki T, Nakaya M. Gastrointestinal motor-stimulating activity of macrolide antibiotics and analysis of their effects on the canine gut. Antimicrob Agents Chemother 1984;26:863-9. 8. Kubota M, Nakamura T, Motokura T, Mori S, Nishida A. Erythromycin improves gastrointestinal motility in extremely low birth weight infants. Acta Paediatr Jpn 1994;36:198-201. 9. Coulie B, Tack J, Peeters T, Janssens J. Involvement of two different pathways in the motor effects of erythromycin on the gastric antrum in humans. Gut 1998;43:395-400. 10. Costalos C, Gavrili V, Skouteri V, Gounaris A. The effect of low-dose erythromycin on whole gastrointestinal transit time of preterm infants. Early Hum Dev 2001;65:91-6. 11. Dellagrammaticas HD, Iacovidou N, Megaloyanni E, Papadimitriou M, Kapetanakis J. Effect of low-dose oral erythromycin on gastric aspirates in ventilated neonates less than 32 weeks of gestation: preliminary results. Biol Neonate 2002;81:213-6. 12. Patole S, Rao S, Doherty D. Erythromycin as a prokinetic agent in preterm neonates: a systematic review. Arch Dis Child Fetal Neonatal Ed 2005;90(4):F301-6. 13. Stenson BJ, Middlemist L, Lyon AJ. Influence of erythromycin on establishment of feeding in preterm infants: observations from a randomised controlled trial. Arch Dis Child Fetal Neonatal Ed 1998;79:F212-4. 14. Patole SK, Almonte R, Kadalraja R, Tuladhar R, Muller R, Whitehall JS. Can prophylactic oral erythromycin reduce time to full enteral feeds in preterm neonates? Int J Clin Pract 2000;54:504-8. 15. Oei J, Lui K. A placebo-controlled trial of low-dose erythromycin to promote feed tolerance in preterm infants. Acta Paediatr 2001;90:904-8. 16. Su BH, Lin HC, Peng CT, Tsai CH. Effect of erythromycin on feeding intolerance in very low birth weight infants: a preliminary observation. Zhonghua Min Guo Xiao Er Ke Yi Xue Hui Za Zhi 1998;39:324-6. 17. Nogami K, Nishikubo T, Minowa H, Uchida Y, Kamitsuji H, Takahashi Y. Intravenous low-dose erythromycin administration for infants with feeding intolerance. Pediatr Int 2001;43:605-10. 18. Sekteera W, Nuntnarumit P, Supapannachart S. Oral erythromycin for treatment of feeding intolerance in preterm infants: a preliminary report. J Med Assoc Thai 2002;85(suppl 4):S1177-82. 19. Ng PC, So KW, Fung KS, Lee CH, Fok TF, Wong E, et al. Randomised controlled study of oral erythromycin for treatment of gastrointestinal dysmotility in preterm infants. Arch Dis Child Fetal Neonatal Ed 2001;84:F177-82. 20. El Hennawy AA, Sparks JW, Armentrout D, Huseby V, Berseth CL.
Efficacy Of Oral Erythromycin For Treatment Of Feeding Intolerance In Preterm Infants
Erythromycin fails to improve feeding outcome in feeding-intolerant preterm infants. J Pediatr Gastroenterol Nutr 2003;37:281-6. 21. Ng SC, Gomez JM, Rajadurai VS, Saw SM, Quak SH. Establishing enteral feeding in preterm infants with feeding intolerance: a randomized controlled study of low-dose erythromycin. J Pediatr Gastroenterol Nutr 2003;37:554-8. 22. Sims PJ, Waites KB, Crouse DT. Erythromycin lactobionate toxicity in preterm neonates. Pediatr Infect Dis J 1994;13:164-7. 23. Gouyon JB, Benoit A, Betremieux P, Sandre D, Sgro C, Bavoux F, et al. Cardiac toxicity of intravenous erythromycin lactobionate in preterm infants. Pediatr Infect Dis J 1994;13:840-1. 24. Mahon BE, Rosenman MB, Kleiman MB. Maternal and infant use of erythromycin and other macrolide antibiotics as risk factors for infantile hypertrophic pyloric stenosis. J Pediatr 2001;139:380-4. 25. George SL, Desu MM. Planning the size and duration of a clinical trial studying the time to some critical event. J Chronic Dis 1974;27:15-24. 26. Patamasucon P, Kaojarern S, Kusmiesz H, Nelson JD. Pharmacokinetics of erythromycin ethylsuccinate and estolate in infants under 4 months of age. Antimicrob Agents Chemother 1981;19:736-9. 27. Waites KB, Sims PJ, Crouse DT, Geerts MH, Shoup RE, Hamrick WB, et al. Serum concentrations of erythromycin after intravenous infusion in preterm neonates treated for Ureaplasma urealyticum infection. Pediatr Infect Dis J 1994;13:287-93. 28. Cairns P, Craig S, Tubman R. Randomised controlled trial of low-dose erythromycin in preterm infants with feed intolerance [abstract]. Arch Dis Child 2002;86(suppl I):G24. 29. Jadcherla SR, Klee G, Berseth CL. Regulation of migrating motor complexes by motilin and pancreatic polypeptide in human infants. Pediatr Res 1997;42:365-9. 30. Tomomasa T, Kuroume T, Arai H, Wakabayashi K, Itoh Z. Erythromycin induces migrating motor complex in human gastrointestinal tract. Dig Dis Sci 1986;31:157-61. 31. Peeters T. Erythromycin and other macrolides as prokinetic agents. Gastroenterology 1993;105:1886-99. 32. Tack J, Janssens J, Vantrappen G, Peeters T, Annese V, Depoortere I, et al. Effect of erythromycin on gastric motility in controls and in diabetic gastroparesis. Gastroenterology 1992;103:72-9. 33. Tomomasa T, Tabata M, Nako Y. The effect of oral erythromycin on gastric emptying in neonates with gastric retention [abstract]. Gastroenterology 1997;112(suppl):A839.
605