Missed Opportunities in the Trial on Proton-Pump Inhibitor Therapy in Bleeding Peptic Ulcers

IN RESPONSE:

We thank all the correspondents for their considered comments. Ms. Yih and Dr. Tejani, Drs. Lin and Hsu, and Dr. Piscoya raised concerns about the use of placebo as a comparator in our study of patients with peptic ulcer bleeding. When this study was planned, the efficacy of intravenous PPI therapy for treatment of peptic ulcer bleeding was still in question, and no approved comparator treatment existed. After full discussion of the design with the U.S. Food and Drug Administration and the European Medicines Agency, it was agreed that the study should be placebo-controlled, which resulted in the first approval of intravenous PPI therapy for this indication in several countries. Now that the efficacy of intravenous PPI therapy has been accepted by health authorities, we agree that it is logical to explore different doses of intravenous therapy and the possible utility of oral PPI therapy. A comparison of oral with intravenous therapy may require many patients, however, even if a noninferiority design is chosen to demonstrate similar efficacy.

Drs. Lin and Hsu also raised concerns about the low mortality and rebleeding rates observed in our study. Although the study by Lau and colleagues (1) focused on preemptive therapy with omeprazole and thus had a different design, it did have a placebo group, permitting comparison of mortality rates with our study, although we do accept that patients in Lau and colleagues' study were probably in a lower risk group. However, other recent studies (2, 3) in patients with peptic ulcer bleeding have reported similarly low mortality rates. We also attempted to provide other explanations for the low mortality rates, such as exclusion of patients with the worst prognosis (that is, highest American Society of Anesthesiologists scores). Some earlier studies also included more patients who first bled as “inpatients” rather than “outpatients,” making them clinically high-risk patients.

Concerning the low rebleeding rates that we observed, we emphasized efficient endoscopic treatment, which we consider to be one of its strengths because even with such state-of-the-art therapy, profound acid suppression still provided additional benefit. The issue of whether to standardize endoscopic therapy in our study was thoroughly discussed within the steering committee, with the view that various endoscopic approaches were used in different centers. We made a deliberate decision to allow “real world” endoscopic treatment, to evaluate the true value of adjuvant esomeprazole therapy in current clinical practice. If one controls too many variables in a study, they may be criticized for making the design remote from reality.

Finally, Ms. Yih and Dr. Tejani raised an issue concerning multiethnicity in our study. Although our study population was largely white, a sufficient number of Asians were included to permit separate analyses of their data, allowing them to be examined as an “internal control” group.