Barrett’s Esophagus Progression and Proton Pump Inhibitors

Abstract & Commentary

Synopsis: Examining data from a 20-year time period, correlations were sought between antisecretory drug therapy and cumulative incidence of dysplasia. Incidence of dysplasia was significantly lower in patients who had received proton pump inhibitor (PPI) therapy vs no therapy or H2-receptor antagonists (H2RAs).

Source: Alter MJ, et al. Ann Intern Med. 2004;141:715-717.

Chronic acid reflux is thought to underlie the pathogenesis of Barrett’s esophagus (BE; (change from normal squamous mucosa to specialized columnar metaplasia). BE underlies most or all cases of esophageal adenocarcinoma, and it is believed by many that continued acid exposure increases adenocarcinoma risk. It has been speculated that acid suppression might alter cancer risk in this setting, but results of studies have not been conclusive one way or the other. This retrospective analysis of patients followed at the Tucson, Arizona, VA hospital between 1981 and 2000 attempts to compare patients with BE who received acid suppressive therapy with those who did not; 236 patients who initially presented without dysplasia were analyzed. Also, 155 patients (66%) received a PPI, and 149 patients got H2RAs, and 21 received neither. Fifty-six (56) patients developed dysplasia during a total follow-up of 1170 patient years creating an incidence rate of 4.7% per year. Dysplasia developed in 9 of 19 patients on neither medication, and in 25 of 64 patients taking

H2RAs. PPI therapy given after diagnosis was associated with less dysplasia. At 5 years, dysplasia with PPI therapy was 11% vs 37%; at 10 years, results were 21% vs 58%. Time of diagnosis and length of Barrett’s mucosa and grade of dysplasia did not alter these findings. The authors believe that these data strongly suggest that potent acid suppression can alter the progression from metaplasia to dysplasia.

Comment by Malcolm Robinson, MD, FACP, FACG

Alter and colleagues discuss some of the potential defects of the study. For example, prescriptions obtained outside of the VA system would not have been tracked. There were 2 cases of adenocarcinoma in patients who didn’t demonstrate previous dysplasia. Both had received PPIs. Alter et al provide lots of useful information on the probable contributions of acid reflux and chronic inflammation to abnormal cellular proliferation. They mention the failure of prior studies of medical and surgical antireflux therapy to alter evolution to esophageal adenocarcinoma. Finally, they admit that prospective randomized data collection is necessary to confirm these findings.

An accompanying editorial by Dr. Thomas Schnell outlines other potential flaws to the conclusions of the Tucson VA study. He points out that acid suppression itself has been accused of promoting cancer, and he reiterates the considerable existing data that negate any protective effect of acid suppression on cancer progression in BE. Alter et al and the editorial both mention the fact that there is poor inter-observer agreement regarding the presence or absence of low-grade or indeterminate dysplasia (vs no dysplasia). There is of course no data on what happened to these patients prior to study entry, and many factors could impact later cancer risk (toxin exposure, diet, etc). The relatively brief exposure to PPIs in this study (only 1-2 years generally) was described by Dr. Schnell as terribly brief to have an effect on years of carcinogenesis. Nevertheless, on balance, this article is certainly ammunition for those who urge that all patients with BE be treated with PPIs (whether they have symptoms or not). I am not so sure about this conclusion, but I must admit that my doubt regarding PPI use in this setting has been shaken. We must all hope that well designed future prospective studies will provide an answer that all of us can accept in terms of the continued chronic and indefinitely prolonged use of expensive PPIs in any major subset of GERD patients.

Dr. Robinson, Medical Director, Oklahoma Emeritus Clinical Professor of Medicine, University of Oklahoma College of Medicine Oklahoma City, OK, is Associate Editor of Internal Medicine Alert.