Erythromycin as a Prokinetic Agent in Preterm Infants

*Christos Costalos, *Anthony Gounaris, Evangelia Varhalama, Fedra Kokori, *N. Alexiou, and*Efi Kolovou

Departments of *Neonatal Medicine and Ultrasonography, Piraeus General Hospital, Nikea, Piraeus, Greece

ABSTRACTBackground: The macrolide antibiotic erythromycin is a pro-kinetic agent that stimulates gastrointestinal motility. The aimof the study was to determine the effect of erythromycin on thegastrointestinal motility of preterm infants.Methods: Erythromycin 10 mg/kg, 8 hourly or a placebo, wasgiven orally for 7 days in a double-blind randomized, crossoverstudy of 20 preterm infants with a median gestational age of 32weeks (range, 2634 weeks). Antral contractility was deter-mined by using ultrasonography to measure the decrease in thegastric antral cross-sectional area after a feed. The whole guttransit time was assessed by timing the transit of carmine redthrough the gut.

Results: Antral contractility lasted for a shorter period of timeduring erythromycin treatment than during placebo treatment(mean [standard deviation], 31 minutes [9.9 minutes] vs. 70minutes [13 minutes]; P < 0.01). Whole gut transit time wasalso shorter during erythromycin treatment (mean, 23.1 hours[12.9 hours] vs. 49.3 hours [29 hours]; P < 0.01). All infantstolerated the drug well.Conclusions: Oral erythromycin in food-intolerant preterm in-fants enhances both antral contractility and whole gut transittime. JPGN 34:2325, 2002. Key Words: ErythromycinNeonatesGut motility. 2002 Lippincott Williams &Wilkins, Inc.

The macrolide antibiotic erythromycin competitivelyinhibits the binding of motilin to its receptors (1). Eryth-romycin stimulates phase 3 of the migrating motor com-plex in dogs, like motilin (2). It increases antral motility,enhancing gastric emptying in patients with diabetic gas-troparesis (3), postvagotomy gastroparesis (4), andchronic intestinal pseudoobstruction (5). Erythromycinhas been used successfully as a prokinetic agent in younginfants with intestinal dysmotility (6), but its role forpreterm infants is less clear.

The aim of the present study was to determine theeffect of erythromycin on gastrointestinal motility of pre-term infants.

Radioisotope studies are the gold standard for estimat-ing gastric emptying, and although noninvasive, they in-volve irradiation and access to a gamma camera. Studiesusing either marker dilution and a single aspiration tech-nique (7) or a modified double sampling marker dilutiontechnique (8) are not so accurate and may not be welltolerated by smaller infants. The same applies to manom-

etry (9). Nonisotopic methods involving minimal distur-bance to the infant are those using either a 13C-octanoicacid breath test (10) or an ultrasound technique.

Ultrasonography has been used successfully in pre-term infants and has been validated against radioisotopemeasurements in adults (11). This is the method we usedin the present study (12).

METHODSThe study was of a double-blind, randomized, crossover de-

sign. Preterm infants consecutively admitted to the NeonatalDepartment at Piraeus General Hospital Nikea, Greece, wereeligible for the study. Patients were selected on the basis offeeding intolerance defined as a total gastric aspirate volume ofat least half of enteral intake over 24 hours. Infants were ex-cluded from the study if they were receiving drugs known toaffect gastrointestinal motility (e.g., cisapride).

Twenty healthy infants were recruited with a median birthweight of 1,315 g (range, 9801,450 g), a median gestationalage of 32 weeks (range, 2634 weeks), a median postnatal ageat first examination of 17 days (range, 841 days), and receiv-ing an average of 150 mL/kg daily (range, 130165 mL/kgdaily) of a modified formula (Prenan Nestle; Nestle, Vevey,Switzerland) by bolus administration via a nasogastric tubewith an interval of 2 hours between feeds. Fifteen infants weresmall for gestational age (birth weight less than the tenth per-centile for gestation). The study was approved by the hospitalethics committee, and informed parental consent was obtainedfrom all participants.

On recruitment, infants were allocated according to a prede-termined list of random numbers by the hospital pharmacy to

Received March 17, 2000; accepted May 14, 2001.Address correspondence and reprint requests to Dr. Christos

Costalos, 130 Artemidos Str., Paleo Faliro 17562, Greece.

This article is accompanied by an editorial. Please seeKaul A. Erythromycin as a prokinetic agent [editorial].J Pediatr Gastroenterol Nutr 2002;34:1315.

Journal of Pediatric Gastroenterology and Nutrition34:2325 January 2002 Lippincott Williams & Wilkins, Inc., Philadelphia

23

receive on the first occasion either 7 days of oral erythromycin10 mg/kg every 8 hours or a placebo (hydroxypropyl methylcellulose) of equivalent volume. The alternate drug was usedfirst on the second occasion. Each infant thus served as his owncontrol. The hospital pharmacy supplied erythromycin and aplacebo preparation of identical appearance, and the investiga-tors were blind to infant allocation until after the study wascomplete.

Antral contractility and whole gut transit time were mea-sured on the seventh day of each treatment. Antral contractilitywas assessed ultrasonically by serially measuring the change inthe gastric antral cross-sectional area (ACSA) after a feed usinga previously described method (12). The ACSA was measuredby an experienced radiologist by obtaining real time ultrasoundcross-sectional images of the gastric antrum with the infantlying in the right lateral position. Vertebral bodies, the aorta,and the superior mesenteric artery were used as constant land-marks to ensure that the same cross-sectional plane of the gas-tric antrum was measured on each occasion. Images were ob-tained using a Contron Ultrasound Imaging System (RocheCo., Switzerland) with a 7.5 MHz probe. The same operator didall ACSA measurements. Before each feed, the stomach wasaspirated using a nasogastric feeding tube. The ACSA wasmeasured before the test feed. A bolus feed of the pretermformula (10 mL/kg) was given over 10 minutes. On completionof the feed, the ACSA was remeasured. The ACSA was mea-sured subsequently at 10-minute intervals until 30 minutes hadelapsed and then continued at 15-minute intervals until 2 hoursafter the feed or the prefeed value was reached. Treatmentswere compared by calculating the 50% ACSA, defined as thetime taken for ACSA to decrease by half of the maximumchange seen during gastric filling. This value provides a simple,single measurement of antral contractility.

Whole gut transit time was measured right after the ACSAby timing the transit of the marker carmine red through thegastrointestinal tract. Carmine red (125 mg) was given via anasogastric feeding tube over 1 minute and the time was re-corded. The dye was always given immediately after a feed.Median feed volume was 21 mL (range, 1532 mL). At eachsubsequent diaper change, any feces were inspected for evi-dence of the dye. Whole gut transit time was defined as theelapsed time between carmine red administration and its firstappearance in the feces (13). At the second measurement, eachbaby received the same volume and type of milk feed as he hadreceived during the first median feed volume.

Statistical analysis was performed using the paired t test. AP value of less than 0.05 was regarded as significant.

RESULTS

Mean (standard deviation) 50% ACSA was signifi-cantly shorter during erythromycin treatment as com-pared with placebo: 31 minutes (9.9 minutes) versus 70minutes (13 minutes); P < 0.01 (Fig. 1). Mean (standarddeviation) whole gut transit time was also significantlyshorter during erythromycin treatment: 23.1 hours(12.9 hours) versus 49.3 hours (29 hours); P < 0.01(Fig. 2).

All infants tolerated erythromycin well. There were noadverse effects associated with its use (e.g., diarrhea,cardiac arrhythmias).

DISCUSSION

Prokinetic agents have been used in infants to treatgastrointestinal dysmotility syndromes. Bethanechol, amuscarinic agonist, was one of the first agents used.However, it has a high frequency of undesirable sideeffects (14). Metoclopramide, a dopamine receptor an-tagonist, also has been used, but its major drawback isthe distressing extrapyramidal side effects (15). Of thenewer drugs, domperidone is a benzimidazole derivativewith dopamine antagonistic properties. Unlike metoclo-pramide, it does not readily enter the central nervoussystem and normally is devoid of extrapyramidal reac-

FIG. 2. Whole gut transit time (hours) during erythromycin andplacebo treatment. The horizontal bar represents mean value.

FIG. 1. Time (in minutes) taken for the antral cross-sectionalarea to decrease by half of the maximum change seen duringgastric filling (50% antral cross-sectional area) during erythromy-cin and placebo treatment. The horizontal bar represents mean50% antral cross-sectional area.

C. COSTALOS ET AL.24

J Pediatr Gastroenterol Nutr, Vol. 34, No. 1, January 2002

tions, but in young babies in whom the bloodbrain bar-rier function is not yet fully developed, such a reactionmay occur (16).

Cisapride is the first of a subgroup of substitutedbenzamides that have an effect on gastrointestinal motorfunction via indirect cholinergic mechanisms without in-terfering with dopamine receptors. Concern, however,has been raised over its safety because it prolongs theQTc interval and can lead to serious arrhythmias andeven death (17,18).

Erythromycin is a motilin receptor agonist. It inducesbouts of intense motor activity in the stomach and smallintestine that resemble phase 3 migrating motor complexactivity (19). Despite its use in adults and older childrenas a prokinetic agent (5), experience in neonates remainsscarce. A beneficial effect was first shown by Simkiss etal. (6) on four infants with postoperative intestinal dys-motility, two of whom received erythromycin enterallyand two intravenously at a dose of 12 mg/kg given every6 hours. In another study (20), the antral motility indexwas increased fourfold in six healthy infants of fewerthan 30 weeks gestation given a single erythromycindose of 0.75 mg/kg.

Stenson et al. (21), using erythromycin in ventilatedinfants of fewer than 31 weeks gestation in a dose of 15mg/kg given three times daily, showed that the drug didnot lead to earlier establishment of enteral feeding. How-ever, infants treated with erythromycin were prescribedfewer glycerine suppositories for constipation.

Our results confirm that in food-intolerant infants,erythromycin promotes both antral contractility andwhole gut transit time when administered at the moderateenteral dose of 10 mg/kg given three times daily, withoutcausing any side effects. Because erythromycin toxicityis higher when using the parenteral rather than the enteralroute (22), it is to the infants benefit to use the enteralroute in as low a dose as is clinically effective. In fact, asshown in dogs, a high erythromycin dose may be lesseffective than a lower dose because it prolongs the mi-gratory motor complex cycle period of the intestine (23).A recent study by Jadcheria and Berseth (24) on healthypreterm infants confirms this by showing that a dose ofthe drug as small as 1.5 mg/kg has significant prokineticeffects.

In conclusion, oral erythromycin given at a dose of10 mg/kg every 6 hours promotes both antral contractil-ity and whole gut transit time in food-intolerant preterminfants. Larger studies will be required to determine theleast effective prokinetic dose of the drug as well as itsrole in neonatology.

REFERENCES1. Peeters T, Mathijs G, Depoortere I, et al. Erythromycin is a motilin

receptor agonist. Am J Physiol 1989;257:G4704.

2. Itoh Z, Nakaya M, Sukuzi T, et al. Erythromycin mimics exog-enous motilin in gastrointestinal contractile activity in the dog. AmJ Physiol 1984;247:G68894.

3. Kerlin P, Zinsmeister A, Philips S. Relationship of motility to flowof contents in the human small intestine. Gastroenterology 1982;82:7016.

4. Morzwecz H, Pavel D, Pitrak D, et al. Erythromycin stearate asprokinetic agent in postvagotomy gastroparesis. Dig Dis Sci 1990;35:9025.

5. Di Lorenzo C, Lachman R, Hyman PE. Intravenous erythromycinfor postpyloric intubation. J Pediatr Gastroenterol Nutr 1990;11:457.

6. Simkiss DE, Adams IP, Myrdal U, et al. Erythromycin in neonatalpostoperative intestinal dysmotility. Arch Dis Child 1994;71:F1289.

7. Husband J, Husband P. Gastric emptying of water and glucosesolutions in the newborn. Lancet 1969;ii:40911.

8. George JD. New clinical method for measuring the rate of gastricemptying: The double sampling test meal. Gut 1968;9:23742.

9. Arndorfer RC, Steff JJ, Dodds WJ, et al. Improved infusion systemfor intraluminal oesophageal manometry. Gastroenterology1977;73:237.

10. Van Den Driessche M, Peeters K, Marien P, et al. Gastric emptyingin formula-fed and breast-fed infants measured with the 13C-octanoic acid breath test. J Pediatr Gastroenterol Nutr 1999;29:4651.

11. Marzio L, Giacobbe A, Conoscitore P, et al. Evaluation of the useof ultrasonography in the study of liquid gastric emptying. Am JGastroenterol 1989;84:496500.

12. Newell SJ, Chapman S, Booth IW. Ultrasound assessment of gas-tric emptying in the preterm infant. Arch Dis Child 1993;69:326.

13. Rubaltelli FF, Largajolli G. Effect of light exposure on gut transittime in jaundiced newborns. Acta Paediatr Scand 1973;62:1468.

14. Levi P, Marmo F. Betanechol versus antacids in the treatment ofgastroesophageal reflux. Helvetica Paediatric Acta 1985;40:34959.

15. Reid M. Dystonic reactions to metoclopramide. Ulster Med J 1977;46:3840.

16. Verlinden M, Welburn P. The use of prokinetic agents in thetreatment of gastrointestinal motility disorders in childhood. In:Milla PJ, ed. Disorders of gastrointestinal motility in childhoodChichester, UK: Wiley and Sons, 1988;12540.

17. Bernardini S, Semama DS, Huet F, et al. Effects of cisapride onQTc interval in neonates. Arch Dis Child 1997;77:F2413.

18. Ahmad SR, Wolfe SM. Cisapride and torsades de pointes [letter].Lancet 1995;345:508.

19. Tomomasa T, Kuroume T, Arai H, et al. Erythromycin inducesmigrating motor complex in human gastrointestinal tract. Dig DisSci 1986;31:15761.

20. Tomomasa T, Miyazaki M, Koizumi T, et al. Erythromycin in-creases gastric antral motility in human premature infants. BiolNeonate 1993;63:34952.

21. Stenson BJ, Middlemist L, Lyon AJ. Influence of erythromycin onestablishment of feeding in preterm infants: Observations from arandomized controlled trial. Arch Dis Child Fetal Neonatal Ed1998;79:F2124.

22. Farrar HC, Walsh Sukys MC, Kyllonen K, et al. Cardiac toxicityassociated with intravenous erythromycin lactobionate: Two casereports and a review of the literature. Pediatr Infect Dis 1993;12:68890.

23. Otterson MF, Sarna SK. Gastrointestinal motor effects of erythro-mycin. Am J Physiol 1990;259:G35563.

24. Jadcheria SR, Berseth CL. Prokinetic effects of erythromycin inpreterm and term infants are related to gestational age [abstract].Pediatr Res 1995;37:124A.

ERYTHROMYCIN IN PRETERM INFANTS 25

J Pediatr Gastroenterol Nutr, Vol. 34, No. 1, January 2002