OTC CIMETIDINE HEARTBURN RELIEF STUDIES ARE "WITHOUT PRECEDENT"

OTC CIMETIDINE HEARTBURN RELIEF STUDIES ARE "WITHOUT PRECEDENT," SmithKline Beecham Clinical R&D Group Director Carl Friedman, MD, declared in justifying the company's clinical trial approach to establishing the efficacy of low-dose cimetidine. In a Sept. 9 presentation before FDA's Nonprescription Drugs and Gastrointestinal Drugs Advisory Committees, Friedman reported that, "a review of the literature shows several small studies [but] there are no placebo-controlled trials demonstrating relief in duration or intensity of established heartburn." Friedman highlighted the challenge facing the company in designing studies that would demonstrate the effectiveness of cimetidine in an OTC setting for an already established and self- limiting condition such as episodic heartburn. He noted that because the condition is self-limiting and subjective, studies were showing "a very high placebo rate." In addition, both the test treatments and the controls have to be taken with water, which "has some palliative effect," Friedman pointed out. In his presentation to the committees, Friedman defended the methods used by the company to analyze the clinical trial data. "In analyzing these data and developing our methods it became obvious to us that our episodic heartburn data contained many antithetical values that were not normally distributed," he said. Consequently, he said the company decided to use "median" values in analyzing the data rather than averages. Acknowledging that the committee members might not be "accustomed" to this approach, Friedman asserted that "medians are the best way to ethically represent individual data when they are not normally distributed." Defending the one-sided statistical comparison used by SmithKline to interpret the data, Friedman explained that this decision was based on the "concept that it was virtually impossible for placebo to give rise to better acid suppression than cimetidine." The company arrived at this assumption based on cimetidine's proven effect on gastric acid secretion and the lack of such an effect with placebo. Friedman also reported that FDA asked SmithKline Beecham to conduct a reanalysis of the clinical data based on three parameters: a first episode analysis, a last episode analysis and an analysis limited only to responsive episodes. Friedman said that FDA's rationale for requesting the first episode analysis was to eliminate confounding factors that may have distorted the clinical results when subsequent heartburn episodes are analyzed. Also, FDA felt that this approach was necessary, Friedman explained, because of the varying number of episodes experienced by patients in the treatment groups. The SmithKline researcher pointed out that much of the statistical power of the study is lost when the data analysis is confined to the drug's effect on each test subject's first heartburn episode. "You are going from approximately 15,000 episodes when all are analyzed to approximately 1,200 episodes when only the first episode is analyzed," he said. The study was not designed to assess efficacy in one dose, he said. "It was designed to analyze efficacy with repeated doses over the course" of one to two weeks. Also, the first episode analysis "does not really assess the true clinical value of cimetidine which when administered for intermittent, repeated [heartburn episodes] over one to two weeks is clearly effective," Friedman asserted. Nonetheless, Friedman noted, "despite this loss of power, we still observe differences in the first episode analysis." However, Friedman acknowledged that "we lose significance" when all centers are analyzed with a two-sided test. The company's difficulties in proving cimetidine's effectiveness notwithstanding, SmithKline Beecham was applauded for its efforts by Gastrointestinal Drugs Advisory Committee member Vay Liang Wong Go, MD, UCLA School of Medicine. "The sponsors have done really a pioneering kind of job at reaching the issue," Go commented. "The design may not be perfect," he said, "but they really have done something that I'm pleased to see for the first time in OTC drugs." In addition, Gastrointestinal Drugs Advisory Committee Chairperson Rosemarie Fisher, MD, Yale University, also "congratulated" the company "on this difficult study." SmithKline Beecham also received accolades from FDA Office of OTC Drug Evaluation Director Michael Weintraub, MD, for designing OTC clinical trials that focused on consistency of response. Although the company was raised for its efforts, the committee was clearly dissatisfied with SmithKline Beecham's analyses and conclusions. In reference to Friedman's assertion that "200 mg is clearly effective [and] 100 mg is also effective but less consistently so," G-I committee member Ralph D'Agostino, PhD, Boston University, said that SmithKIine's "inference [is] not at all in the analysis." D'Agostino added: "I don't think there's anything in the data that I can see that is able to clarify that comparison." In response, Friedman said: "I think you get a gut feeling that 200 mg is very effective. Then go a little bit extra to show that 100 mg is effective." Peter Banks, MD, Brigham and Women's Hospital, asked whether "there was anything about whether, during or after the study, the patients felt that they were better off than they had been." Describing the results of a questionnaire filled out by still- blinded study subjects, Friedman answered: "The good news is that cimetidine-treated patients all thought they got at least the relief they were getting, if not better than, with their ordinary antacid," he said. "The bad news is: so did those who were treated with placebo and antacid."