Reflux med use in infants with GERD symptoms

This opinion appeared on Medscape.

Proton pump inhibitors (PPIs) are not US Food and Drug Administration-approved for use in infants, but their use has increased dramatically since 2000. The authors point out that to date, the studies evaluating treatment effects of PPIs for gastroesophageal reflux disease (GERD) suffered from methodological issues such as low patient numbers, nonrandomized, nonplacebo-controlled, etc. This study sought to remedy many of the methodological issues of the previous studies.

This Journal Pediatrics multicenter study, conducted in the United States and Poland, enrolled infants during 2006 and 2007. The 162 subjects were randomized 1:1 to intervention drug or identical placebo, and all subjects had been diagnosed with GERD by at least clinical evaluation. Not all subjects received diagnostic testing such as impedance probes or pH probes or endoscopy.

Eligible infants were subjected to 7 to14 days of nonpharmacological intervention (limiting tobacco smoke exposure, changes to feeding strategies and positioning strategies) before randomizing to drug or placebo. All subjects had sufficiently abnormal scores on parental questionnaires (validated) that assessed infant GERD and symptoms associated with GERD. All infants included in the study also had to exhibit daily crying within the hour after feeding for at least 25% of feeding episodes in the 4 days prior to randomization; these data were from a daily diary completed by the parents.

The intervention drug was lansoprazole, given at doses of 0.2 to 0.3 mg/kg/day for infants younger than 11 weeks old and 1.0 to 1.5 mg/kg/day for infants 11 weeks or older. Subjects were to continue their assigned treatment for at least 1 week. They could then go on open-label (not blinded) lansoprazole if the site investigator deemed the treatment ineffective. Either way, data collection on symptoms continued for up to 4 weeks.

The primary outcome variable was the number and frequency of crying episodes during the first hour after feeds, obtained from daily symptom diaries completed by the parents. Identical data obtained from the 7 days before treatment served as the baseline for each subject. The authors were able to calculate responder rate at 4 weeks after randomization. Subjects were labeled "responders" at week 4 if they experienced at least a 50% reduction in percentage of feedings associated with crying or a 50% decrease in overall crying duration in the first hour after feeds.

The subjects were 50% men, 80% white. Approximately 25% of the sample were born prematurely. In regard to primary outcome, both groups had a 54% response rate to treatment (intervention or placebo). In a similar manner, the rate of discontinuing assigned therapy due to perceived non-response was 35% in lansoprazole group and 36% in placebo (nonsignificant). Reduction in crying, reduction in regurgitating, and reductions in feed refusal and back arching were virtually identical between the 2 groups. For example, the percentage of feeding episodes associated with crying and arching of back after feeds decreased about 20% in both groups.

The parents' global assessment of whether the subject improved was 56% in treatment group and 51% in the placebo group. The investigators identified higher rates of lower respiratory tract infections (5% vs 1%) in those on the treatment drug, but the difference was not significant. The authors conclude that their study did not identify a difference reduction of symptoms attributed to GERD between lansoprazole and placebo.

Viewpoint

This study excels by more than tripling the number of subjects involved in any previous randomized, placebo-controlled, blinded trial, but the fact that the subjects did not have objective evidence of GERD (via pH or impedance probe, for example) will remain an opening for criticism by skeptics. However, I think it is worth taking several issues into account. First, many children are treated with medications, including PPIs, for GERD without test documentation of GERD. In that regard, this investigative approach in this study more closely mirrors daily clinical practice. Given that, the study shows at a minimum that placing infants on PPIs based on only history and physical examination findings is no better than placebo. Also, it is worth noting that the previous studies, with their methodological limitations, did not demonstrate an improvement with PPI use. This consistency of findings between different studies is for me the most convincing aspect of the series of studies. The authors are correct to note that one can't extend these findings to children with documented erosions or older children, but the data do support a more stringent standard for which infants should receive PPIs.

Bonnie - how sad. If you read this carefully, you'll see that this is a microcosm of what is wrong with our health care system. As I have said for a while now, it is downright criminal to use PPIs in infants, especially long-term. This is so frustrating to me because I have seen the damage done by these drugs in adult clients. I cannot imagine the host of long-term effects they will create in infants.