Article Text

Download PDFPDF

Randomised controlled trial of cisapride in preterm infants
  1. Academic Children's Hospital
  2. Free University of Brussels
  3. Laarbeeklaan 101
  4. 1090 Brussels, Belgium

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Editor—I read with great interest the study by McClure et al 1reporting delayed gastric emptying and a non-significant increase in whole gut transit time in premature infants treated with cisapride.1 In contrast, placebo controlled studies evaluating cisapride in paediatrics have almost consistently reported improvement in symptoms of gastro-oesophageal reflux diseases or improvement in oesophageal pH results and/or manometry.2As acknowledged by the authors, other studies in premature infants have reported a reduction in gastric residue and improved feed tolerance. In addition, we have reported a reduction in gastro-oesophageal reflux in premature infants.3

In this latest trial the authors chose hydroxypropylmethyl cellulose as a placebo, given at the “same volume” as cisapride; however, the actual quantity administered is not clear. Cellulose derivatives are commonly employed as laxatives, where it is proposed that they act by absorbing water, so softening the faeces and increasing stool volume. This in turn stimulates faecal propulsion.4 Clearly, depending on the dose of hydroxypropylmethyl cellulose used, the choice of this compound as a placebo is nonsensical as there could be a marked impact on gastrointestinal transit time. From the publication it is also not clear why gastrointestinal motility was measured after three days dosing; it is likely that any cathartic effect from a laxative would be less marked after more prolonged treatment time.

The manufacturer had recommended against the use of cisapride in premature infants because of concern that the metabolic pathway of cisapride may not be fully developed. However, in most European countries and the United States, cisapride is commonly used in premature (gestational age > 34 weeks) infants, and, despite this widespread use, the incidence of clinically important cardiovascular effects is very low. Such events are often associated with high doses (> 0.8 mg/kg/day) or concomitant administration of drugs known to inhibit cytochrome P450 that, like cisapride, are known to prolong the QT interval.2

In my opinion, in view of the clinical experience with cisapride in severe gastro-oesophageal reflux disease and feed intolerance in premature infants, and the few clinically relevant cardiovascular events, treatment with cisapride with appropriate monitoring (electrocardiogram before and after two or three days of treatment) should not be withheld from these infants if clinically indicated.


Dr McClure responds: We agree, and indeed stated in our own paper1-1 that cisapride has been shown in children to improve symptoms of gastro-oesophageal reflux disease and to hasten both gastric emptying and gastrointestinal transit time. We believe that it is dangerous to assume that the immature gas of the preterm infant will react in the same manner. We further believe that our study has the strongest methodology of any published that has directly examined the effect of cisapride on gastrointestinal motility in the preterm infant. As stated in our discussion, we did not measure gastro-oesophageal reflux and so cannot comment on the efficacy of cisapride for this condition. However, in view of our findings, until there is published evidence in a peer reviewed journal of cisapride's efficacy for this condition, we hold to our concluding statement that it should be contraindicated as recommended by the Medicines Control Agencies.1-2

We do not believe that the choice of hydroxypropylmethyl cellulose as a placebo inadvertently affected our study. The formulation of cisapride suspension used in our study contained hydroxypropylmethyl cellulose. This was why this agent was chosen as placebo. The strength of cisapride was 1 mg/ml, and therefore infants typically received 0.2–0.3 ml per dose when receiving placebo. Cellulose derivatives act as laxatives by increasing faecal bulk. We do not believe that an aqueous solution of this volume to be anywhere near sufficient to cause this effect. Lastly, the significant finding of our study was delayed gastric emptying time during cisapride treatment, not whole gastrointestinal transit time.

The pharmacokinetics of cisapride in the preterm infant are unclear. A period of three days dosing before measurement of gastrointestinal motility was considered necessary to allow both time for acquisition of a steady serum cisapride level and adequate elimination of any previous cisapride treatment.


  1. 1-1.
  2. 1-2.