A question that won't go away.

Abstract

Maternal and Infant Use of Erythromycin and Other Macrolide Antibiotics as Risk Factors for Infantile Hypertrophic Pyloric Stenosis Mahon BE, Rosenman MB, Kleiman MB. J Pediatr 2001;139:380–4. Summary: The charts were reviewed of infants born at a large urban hospital between 1993 and 1999 for evidence of macrolide antibiotic use either by their mothers during pregnancy or by the infants themselves during the first 3 months of life and for the incidence of infantile hypertrophic pyloric stenosis (IHPS). Prescriptions from the hospital pharmacy records for erythromycin, azithromycin, or clarithromycin were viewed as proof of medication use. Single-dose prophylaxis and topical use, including erythromycin ophthalmic ointment used at delivery, were recorded. Hospital records were then searched electronically for the diagnosis of IHPS using several key words. The study eventually included 14,876 infants (and their mothers) of whom 469 received prescriptions for oral or intravenous erythromycin. Pyloric stenosis was found in 43 of these infants (0.29%). The risk of pyloric stenosis was increased in only the infants who had prescriptions for oral or intravenous erythromycin (not topical). Pyloric stenosis developed in none of the infants who received erythromycin after the first 2 weeks of life. Pyloric stenosis developed in 3% of the 201 infants with prescriptions for more than 2 weeks of use, whereas IHPS developed in none of 215 with prescriptions for fewer than 14 days. Only six infants in the study received azithromycin, and none received clarithromycin; therefore, no conclusions about the risk of these macrolides could be reached. Erythromycin ophthalmic ointment use did not increase the incidence of IHPS. Findings suggest that from the 3,346 mothers who received macrolide antibiotics during pregnancy, erythromycin use during the last 10 weeks of pregnancy was a possible risk factor for IHPS, but no increase in risk was associated with use earlier in pregnancy. Comment: During the years, studies have suggested that early infant exposure to erythromycin is associated with increased risk of IHPS. Most of the studies have been either brief case reports or have been from institutions in which a modest cohort of newborn infants exposed to pertussis before immunization received the medication as prophylaxis. The oversimplified explanation for the association between IHPS and erythromycin is that erythromycin, a potent motilin agonist, causes muscular hypertrophy in the antrum or pylorus by prolonged overstimulation of the antral musculature. The current study does not speak to cause. Its strength is the large number of patients included and the obvious care expended on the record review. The risk of IHPS in the general population of infants is only 0.25%. Thus, a large number of study infants are needed to statistically support what prior studies have suggested. The risk estimated by this large study is that 2.65% of infants treated with erythromycin in the first 2 weeks of life will develop IHPS, a 10-fold increase over the general population. This study also clarifies the timing of erythromycin administration (before 2 weeks of age seems to be the period of risk), the risk of topical erythromycin ointment (none), and the duration of use (higher risk with use for more than 14 days). The dosages of erythromycin prescribed were appropriate, and no dosage-related effect was detected in the affected infants. Unfortunately, the numbers were not large enough to determine whether maternal use was a risk factor, although use late in pregnancy may prove, in bigger future studies, to be a factor. The authors did not extend their study to look at the incidence of IHPS in the infants of breast-feeding mothers who received erythromycin postnatally. The weakness of this study is obvious. Using prescription records to define actual antibiotic use may be unreliable; however, the large number of neonates involved in this study decreases that concern somewhat. The increase in incidence of IHPS was the same in infants receiving drug therapy in the hospital as for outpatients. This finding supported the use of pharmacy records as being synonymous with drug administration in the outpatients. Ninety-seven percent of the infants treated with erythromycin did not develop IHPS. This suggests that other factors relating to IHPS remain to be identified. It would have been interesting to see whether there were any differences in erythromycin dosage or duration of use in the 97% of treated infants who did not develop IHPS. We congratulate the authors for their effort in extracting so much information in this retrospective study. Their findings will have an impact on several aspects of general pediatric care of the newborn, and will also have an impact on gastrointestinal care. Specifically, in the young infant with blood-streaked stool, we should wait for Campylobacter culture results before starting erythromycin therapy.