Rio de Janeiro, Brazil - A new meta-analysis of rimonabant clinical trials has concluded that the drug provides "modest" yet meaningful weight loss in overweight and obese subjects but points to some "deficiencies" in the methodologies used in the studies [1]. The review comes as market-watchers are increasingly questioning the drug's longer-than-expected delay in obtaining FDA approval in the US.

In a paper in the Cochrane Database of Systematic Reviews, Drs Cintia Curioni and Charles André (State University of Rio de Janeiro, Brazil) conclude from their analysis that the use of rimonabant over one year could produce a modest weight loss of approximately 5% of initial body weight.

To heartwire, Curioni emphasized that even modest weight loss is "potentially beneficial," but there are caveats.

"Important details should be mentioned," Curioni said. "The quality of the studies is generally poor, and health-related quality and cost analysis were not evaluated by authors. More high-quality research is needed. . . . If people who lose weight could maintain this loss for a long time, it would be very good. But this question was not evaluated in this review. Studies of this new drug with follow-up [beyond] the end of treatment and more rigorous in the quality of methodology should be done before definitive recommendations can be made."

The investigators conducted a literature review of studies that included overweight or obese people and compared rimonabant with placebo or another active intervention. In all, only the four Sanofi-Aventis-sponsored studies met the inclusion criteria: RIO-North America, RIO-Diabetes, RIO-Lipids, and RIO-Europe; for the purposes of the pooled analysis, only the first 12 months of study results were included in the analysis (two of the studies followed patients for two years), yielding results for a total of 6625 patients.

The authors report that weight and waist circumference were significantly reduced in the 20-mg rimonabant group, as compared with placebo. Statistically significant but minimal reductions in weight and waist circumference were also seen in the 5-mg group compared with placebo. General adverse effects and serious adverse effects were significantly greater in the 20-mg group as compared with placebo and with the 5-mg dose. Blood-pressure changes associated with the 20-mg dose were not "reliable," due to the heterogeneity of the studies, the authors concluded; however, triglycerides were significantly reduced, and HDL cholesterol was significantly increased in the 20-mg group compared with the placebo-treated patients.

Curioni and André also evaluated the methodologies used in the studies, identifying several "deficiencies." Only two studies reported an "appropriate method" of randomization, they note, and most did not describe the procedure used to conceal treatment allocation. No methods were described as to how double blinding was done, at what stage blinding was performed, or whether outcome assessors (who took samples or carried out lab tests) were blinded. Only one study reported the power calculation used, and only one study actually described the numbers of patients who discontinued treatment because of adverse events. No study described the intention-to-treat population in detail, Curioni and André write.

To heartwire, Curioni explained that the purpose of a Cochrane collaboration is to conduct systematic reviews looking carefully at methodological quality of studies and to consider the effectiveness of the intervention in light of the studies' quality. "We carried out a review following this principle," Curioni said. "The low quality of the included studies could bias results in different ways. It is difficult to interpret the obtained results if studies do not have adequate randomization, allocation concealment, blinding, and power-calculations analysis. All these factors could interfere with results."

The high attrition rate in the study could also influence outcomes, Curioni added, noting that the authors called the adverse effects "modest and transient" and stated in the papers that the main reasons for study withdrawal were no different from the side effects in people who remained in the study. But "more details are necessary," Curioni said. "Did they lose weight? How long did they participate in the study? The problem is to know whether they discontinued the treatment only due to adverse effects or also due to lack of effectiveness of the studied drug."

While the modest reduction in weight loss seen with rimonabant might potentially be beneficial, the drug should be considered only as an additive strategy, the authors emphasized.

"Always people should be alerted that drugs for obesity are part of a treatment program," Curioni told heartwire. "There is not a 'magic pill.' Efforts should be done in terms of lifestyle changes, including diet and exercise. This review warns that it is difficult to treat obesity, and prevention remains the best strategy."

Such advice may fall on deaf ears in the US, where soaring obesity rates and flagging enthusiasm for low-carb and other fad diets are only fueling anticipation of the drug's approval. As reported by heartwire, the FDA issued rimonabant's manufacturer, Sanofi-Aventis, an "approvable" letter for the drug in February 2006; it was approved with a weight-loss indication in Europe in June and is already available in several countries.

Now, analyst reports are highlighting the fact that a full seven months have passed, yet there has been no hint of a pending FDA approval, nor has a meeting of the FDA's Endocrinologic and Metabolic Drugs Advisory Committee appeared on the tentative schedule for the remainder of 2006. Analysts predict that the FDA is unlikely to approve rimonabant without obtaining the committee's advice, because of concerns over the drug's neurological effects and past blunders in approving eagerly anticipated weight-loss drugs that subsequently had to be yanked off the market due to adverse effects not picked up in clinical trials.

After the fanfare that greeted the RIO program results when they were first unveiled in 2004, Sanofi-Aventis had hinted that the company expected rimonabant to obtain US market in early, then late, 2006—a prediction that now seems increasingly unlikely to materialize.