Efficacy and Safety of Rimonabant for Improvement of Multiple Cardiometabolic Risk Factors in Overweight/Obese Patients

Abstract

OBJECTIVE—To better define the efficacy and safety of rimonabant, the first selective cannabinoid type 1 (CB1) receptor antagonist, in a large population of overweight and obese patients using pooled efficacy data from three Phase III nondiabetes Rimonabant in Obesity and Related Metabolic Disorders (RIO) studies, selected efficacy data from the RIO-Diabetes study, and pooled safety data for all four RIO studies. RESEARCH DESIGN AND METHODS—The RIO studies enrolled patients who were either overweight (BMI >27 kg/m2) with at least one comorbidity (i.e., hypertension, dyslipidemia, or, for RIO-Diabetes, type 2 diabetes) or obese. All patients received daily treatment with rimonabant (5 or 20 mg) or placebo for 1 year plus a hypocaloric diet (600 kcal/day deficit) and advice on increased physical activity. RIO-Europe (n = 1,508), RIO-North America (n = 3,045), and RIO-Lipids (n = 1,036) excluded patients with type 2 diabetes; untreated dyslipidemia was an entry requirement for RIO-Lipids. RIO-Diabetes (n = 1,047) required the presence of type 2 diabetes inadequately controlled by sulfonylurea or metformin monotherapy. RESULTS—The pooled intention-to-treat population comprised 5,580 patients without diabetes (3,165 completed treatment) and 1,047 patients with diabetes (692 completed treatment). Most efficacy measures improved during the 4-week placebo run-in period, except that HDL cholesterol decreased as expected in the early phase of a hypocaloric diet. After 1 year of randomized treatment, changes from baseline with 20 mg rimonabant in the nondiabetic population were as follows: body weight −6.5 kg, waist circumference −6.4 cm, HDL cholesterol +16.4%, triglycerides −6.9%, fasting insulin −0.6 μU/ml, and homeostasis model assessment for insulin resistance (HOMA-IR) −0.2 (all P < 0.001 vs. placebo). In the diabetic population, 20 mg rimonabant reduced A1C levels by 0.6% (P < 0.001 vs. placebo). Regression analysis of change in HDL cholesterol, triglycerides, adiponectin (in RIO-Lipids), and A1C (in RIO-Diabetes) versus body weight at 1 year by ANCOVA suggested that 45–57% of the effect of rimonabant could not be explained by the observed weight loss. At 1 year, adverse events more frequently reported with rimonabant were gastrointestinal, neurological, and psychiatric in nature. Serious adverse events were infrequent and almost equivalent to placebo. Overall discontinuation rates were similar across treatment groups, except discontinuation from adverse events, which occurred more frequently with 20 mg rimonabant versus placebo (most commonly, depressive disorders [1.9 vs. 0.8%], nausea [1.4 vs. 0.1%], mood alterations with depressive symptoms [1.0 vs. 0.6%], and anxiety [1.0 vs. 0.3%]). A thorough review of psychiatric and neurological adverse events was performed. CONCLUSIONS—In overweight/obese patients, 20 mg/day rimonabant produced weight loss and significant improvements in multiple cardiometabolic risk factors such as waist circumference, A1C, HDL cholesterol, and triglycerides. Rimonabant was generally well tolerated, with more frequently reported adverse events being gastrointestinal, neurological, and psychiatric in nature.