1. EXECUTIVE SUMMARY
1.1 Background
This briefing package is being provided to the Endocrinologic and Metabolic Drugs Advisory Committee of the United States (US) Food and Drug Administration (FDA) for the public meeting scheduled on 13 June 2007. During this meeting, the committee will discuss the safety and efficacy of sanofi-aventis’ new drug application (NDA) 21-888, ZIMULTI® (rimonabant), 20 mg.
Sanofi-aventis completed a global development program with rimonabant and submitted a NDA to the FDA in April 2005. Following the filing of the NDA, the FDA administratively divided the application to facilitate the review of the obesity/diabetes indications within the Division of Metabolism and Endocrinology Products and the indication for smoking cessation within the Division of Anesthetic, Analgesia, and Rheumatology Products. Following the review of the NDA, the FDA issued an Approvable letter to sanofi-aventis in February 2006 to support the efficacy of rimonabant 20-mg for obesity and a Not Approvable action letter for the indication of smoking cessation due to lack of efficacy.
Representatives from sanofi-aventis and the FDA met to review the action letters and discussed the actions needed to facilitate review of the NDA for obesity with cardiovascular risks. Sanofi-aventis submitted a complete response to the Approvable Action Letter in October 2006, this submission provided revised labeling with the following proposed indications:
- ZIMULTI® is indicated as an adjunct to diet and exercise for the treatment of overweight patients with body mass index (BMI) >27 kg/m2 and at least 1 other cardiovascular risk factor, or for the treatment of obese patients with a BMI ≥30 kg/m2;
- ZIMULTI® is also indicated in combination with metformin or a sulfonylurea to improve glycemic control and reduce weight in patients with type 2 diabetes and a BMI >27 kg/m2 when diet and exercise plus a single agent do not result in adequate control.
During the advisory committee meeting, sanofi-aventis will focus on the safety and efficacy characteristics of rimonabant, the medical need and benefit-risk of the drug in obese patients and patients with type 2 diabetes, and the proposed risk management plan for ZIMULTI®.
12/120
1.2 Development plan/global regulatory status
Obesity is a chronic and highly prevalent illness frequently associated with numerous and sometimes fatal diseases. Contrary to just being a medical condition or risk factor for other diseases, obesity is a complex disease of multifaceted etiology (including environmental factors and genetic predisposition), with its own disabling capacities, pathophysiology, and comorbidities. Some of the comorbidities of obesity include type 2 diabetes mellitus, dyslipidemia, metabolic syndrome, and cardiovascular disease. Despite the severity of this condition, there are very few anti-obesity agents approved for marketing and none targeting the complex metabolic disorders of this disease. An effective treatment of the causes of the metabolic complications of obesity is still needed, and current guidelines for the treatment of obesity support pharmacotherapy as an adjunct to diet and exercise.
There have been increasing numbers of government initiatives from the Food and Drug Administration (FDA), the National Institute of Health (NIH), and others to address the epidemic of obesity and associated comorbidities including hypertension, type 2 diabetes, and dyslipidemia. In 1996, the FDA published the draft guidance for the Clinical Evaluation of Weight-Control Drugs. This guidance outlined the criteria for regulatory approval for drugs to treat obesity, which included a minimum of one year of placebo-controlled exposure in 1500 patients and a second year of exposure in up to 500 patients. The efficacy criteria established by the guidance stipulated that the mean weight loss be 5% greater in drug versus placebo-treated patients or, that the proportion of patients losing 5% of their weight be greater in the drug versus placebo-treated group. Additionally, the populations to be studied in these trials were recommended to be patients with a body mass index (BMI) greater than 30 or greater than or equal to 27 with comorbidities such as hypertension, type 2 diabetes, and dyslipidemia. The pivotal clinical studies with rimonabant performed by sanofi-aventis met these criteria established by the FDA guidance document.
In 1998-2000, the NIH established clinical guidelines for the treatment of overweight and obesity. In addition to establishing a BMI classification based on epidemiology data that showed increases in mortality with BMIs greater than 25, the guidelines also stated that “weight loss medications should be used only by patients who are at increased medical risk because of their weight and should not be used for cosmetic weight loss.” Furthermore, weight loss medications should never be used without concomitant lifestyle modifications. Importantly, the guidelines also recognized that obesity is a chronic disorder (disease) and that short-term treatment with anti-obesity pharmacotherapy is not helpful. Rimonabant has been developed for those patients at increased medical risk due to their overweight or obesity. Rimonabant, in combination with lifestyle modifications, provides a medical therapy for a chronic disease.
13/120
In September 2004, the Endocrinologic and Metabolic Drugs Advisory Committee met to consider revisions to the draft 1996 FDA guideline. The committee agreed that the size of trials for the development of obesity drugs should be driven by the safety of the drug and that one-year placebo-controlled exposure in these trials should be supported. The efficacy criteria established by the 1996 FDA guideline that the mean weight loss be 5% greater in drug versus placebo-treated patients or, that the proportion of patients losing 5% of their weight be greater in the drug versus placebo-treated group was also supported by the committee. Importantly, the committee strongly recommended the continued evaluation of safety and efficacy in patient populations with BMIs greater than 30 or greater than or equal to 27 with comorbidities.
Earlier this year, the FDA published for public review and comment, a revised draft guideline entitled: Developing Products for Weight Management. The rimonabant development program has met the safety and efficacy criteria outlined in the guidance document.
ZIMULTI® (rimonabant) is a new molecular entity and is the first cannabinoid type 1 (CB1) receptor antagonist to be developed for indications in obesity and for the treatment of patients with type 2 diabetes mellitus. The recommended therapeutic dose of rimonabant 20-mg was determined following the completion of a Phase 2 dose-ranging study in obese subjects. A total of 12 836 patients were exposed to rimonabant in Phase 3 studies (7447 to rimonabant 20-mg; 3478 patient-years), 1008 patients were exposed in 6 Phase 2 studies and 1190 healthy subjects were exposed to rimonabant in 40 Phase 1 studies [Table (1.7) 1]. The presentation in this briefing package of the safety and efficacy data from these studies is focused on supporting the proposed indications in obesity and patients with type 2 diabetes mellitus. Where appropriate (SAEs, seizures, and suicidality-related events), other populations will be considered.
In addition to the completed studies, there were 11 ongoing blinded clinical studies as of 01 March 2007, including 3 Phase 1 studies and 8 pivotal clinical trials, involving an additional 14 280 patients, that had randomization ratios of 1:1 (rimonabant: placebo). These studies are evaluating the safety and efficacy of rimonabant in obese patients with comorbidities including type 2 diabetes mellitus, dyslipidemia, and cardiovascular risk factors. Included in these trials are a cardiovascular outcomes study, CRESCENDO, and 2 morphometric imaging studies. Analysis of rare safety events of interest is also presented for these ongoing studies.
1.3 Mechanism of action
CB1 receptors are part of the endocannabinoid system or ECS. The ECS role encompasses regulation, coordination, and integration of the central behavioral aspects of nutrient intake with the peripheral modulation of nutrient transport, metabolism, and storage in gut, liver, adipose tissue, muscle, and the pancreas. The over activity of the ECS translates into centrally mediated increases in energy intake, while favoring peripheral energy storage, to the detriment of its metabolic use. Moreover, ECS dysregulation is implicated in the underlying etiology of obesity and related metabolic disorders.
14/120
Rimonabant is the first potent and selective antagonist of CB1 receptors. As such rimonabant affects CB1 receptor blockade that contributes to: 1) the normalization of metabolic functions via central effects on energy intake leading to a decrease of body weight; and 2) peripheral effects in muscle, liver, and adipose tissue favoring insulin sensitivity and glucose homeostasis. These effects in turn translate into beneficial modulation of glycemia, glycosylated hemoglobin (hemoglobin A1c) (HbA1c), and triglyceride (TG) flux [TG and high-density lipoprotein cholesterol (HDL-C) levels], above and beyond the body weight decrease, as reported in clinical studies with rimonabant in obese/overweight patients.
1.4 Nonclinical studies
In a comprehensive program of nonclinical studies, rimonabant was shown to have very limited potential to induce toxicity. No specific target organ pathology was identified in the completed animal studies. The nonclinical studies suggest a weak, proconvulsant potential for rimonabant when combined with pentylenetetrazole that alone induces convulsions under experimental procedures in some species. These data are consistent with the view that rimonabant is without proconvulsant potential in the absence of other stressors. Results from the literature and from in-house studies showed that rimonabant had no effect on neuronal excitability in animals when administered alone. Based on the animal findings, seizures were carefully monitored during controlled clinical studies.
1.5 Clinical pharmacology
The pharmacokinetic profile of rimonabant was characterized by a rapid absorption; a large distribution with high plasma protein binding, metabolism, and elimination mainly by the liver; and a long terminal half-life. The terminal half-life of rimonabant is 6 to 9 days in healthy (normal weight) subjects and the time to reach steady state after a once-daily dose is 13 days, with a 3.3-fold accumulation. Higher body weight results in a higher peripheral volume of distribution of rimonabant, which leads to a longer terminal half-life (16 days) and time to reach steady state (25 days) observed in obese subjects. Since clearance after oral administration is not affected by body weight, there are fewer fluctuations in peak-to-trough plasma concentrations, but no difference in total plasma exposure (AUC0-24) in obese subjects compared with normal weight subjects was observed.
The global pharmacokinetic profile justifies a once-daily dose regimen without dose adjustment in age, gender, or body weight. Black patients appeared to be less exposed to rimonabant than Caucasians. An interaction study with ketoconazole, a potent cytochrome P450 (CYP) 3A inhibitor, caused an approximate 2-fold increase in rimonabant exposure; therefore, significant drug-drug interactions due to CYP3A inhibition are unlikely. This observation was further supported by the lack of signal detected in the analysis of the extensive clinical database of rimonabant in smokers and obese patients. Rimonabant dose modifications are not recommended in coadministration with substrates, inhibitors, or inducers of other CYP isozymes or substrates of P-gp, or based on smoking status or hepatic (mild and moderate) or renal impairment.
15/120
Studies were performed in healthy volunteers with rimonabant up to 300 mg as a single dose and up to 80 mg as repeated doses. In addition, an electrocardiogram (ECG) study of 4 weeks duration was performed with the proposed therapeutic dose of rimonabant 20-mg and a supratherapeutic dose of 60 mg (maximum tolerated dose) up to 28 days in duration. This study was conducted in accordance to the International Conference on Harmonisation (ICH) guidance. In this study, rimonabant did not prolong cardiac repolarization at either the therapeutic or supratherapeutic dose.
1.6 Clinical efficacy
1.6.1 Obesity
The 2 doses used in the Phase 3 clinical development program of rimonabant for the treatment of obesity were selected from the dose-ranging study conducted in 287 patients (BMI ≥27 kg/m2 with 3 doses of rimonabant (5-, 10-, and 20-mg, once daily) and a placebo. Compared with the placebo, rimonabant significantly reduced body weight in a dose-related manner in obese patients over a 4-month period of treatment [difference versus placebo: for 5-mg -1.6 kg (-3.5 lbs) (p=0.009), for 10-mg -1.8 kg (-4.0 lbs) (p=0.03), and for 20-mg -2.9 kg (-6.4 lbs) (p=0.0001)]. Additionally, a study conducted at 40-mg for 1 month in 44 patients showed a significant body weight loss (-2.9 kg) (-6.4 lbs) but the rate of nausea was 39.1%. Based on theses results, the Phase 3 studies evaluated 20-mg and 5-mg doses versus placebo.
The efficacy of rimonabant in the treatment of obesity has been demonstrated in 4 adequate and well-controlled studies (the RIO studies), for up to 1 year and for 2 years in 2 of the studies.
Rimonabant 20-mg used in conjunction with a modest calorie diet and physical exercise significantly decreases body weight and waist circumference in overweight or obese patients [Table (1.6.1) 1].
Table (1.6.1) 1 - Mean weight change (lbs) from baseline in obese and overweight patients treated with rimonabant 5-mg and rimonabant 20-mg versus placebo at 1 year (ITT analysis)
| Study | Placebo | Rimonabant 5 mg | Rimonabant 20 mg |
|---|---|---|---|
| RIO-North America | N=590 -3.4 | N=1191 -6.4 | N=1189 -13.8 |
| RIO-Europe | N=302 -4.0 | N=597 -7.4 | N=595 -14.5 |
| RIO-Lipids | N=334 -3.3 | N=340 -6.8 | N=344 -15.3 |
| RIO-Diabetes | N=345 -3.2 | N=355 -5.1 | N=336 -11.7 |
p<0.001 for the difference versus placebo in all comparisons ITT: Intent-to-treat; RIO: Rimonabant in obesity
16/120 Specific details regarding the designs of the 4 Phase 3 studies are presented in Figure (6.1) 1.
In addition, in all 4 studies, rimonabant improved several metabolic parameters associated with obesity, including HDL-C, TG [Table (1.6.1) 2], and insulin, above and beyond the effect of body weight loss.
Table (1.6.1) 2 - HDL-C and TG (mmol/L) at 1 year (*) - RIO studies
| RIO-North America | RIO-Europe | RIO-Lipids | RIO-Diabetes | |||||
|---|---|---|---|---|---|---|---|---|
| Placebo (N=607) | 20 mg (N=1219) | Placebo (N=305) | 20 mg (N=599) | Placebo (N=342) | 20 mg (N=346) | Placebo (N=348) | 20 mg (N=339) | |
| % Change - HDL-cholesterol (HDL-C) | ||||||||
| Mean LS Mean | 5.4 | 12.6 7.2 | 13.4 | 22.3 8.9 | 11.0 | 19.1 8.1 | 7.1 | 15.4 8.4 |
| % Change -Triglycerides (TG) | ||||||||
| Mean LS Mean | 7.9 | -5.3 -13.2 | 8.3 | -6.8 -15.1 | -0.2 | -12.6 -12.4 | 7.3 | -9.1 -16.4 |
p<0.001 for the difference versus placebo in all comparisons (*) Intent-To-Treat (ITT) analysis LS Mean: least squares mean
In the 3 studies that included nondiabetic patients, the proportion of patients who lost at least 5% of baseline body weight was 50.8% in the rimonabant 20-mg group and 19.7% in patients in the placebo group (p<0.001). Twenty-seven percent of patients in the rimonabant 20-mg group lost at least 10.0% of baseline body weight (7.8% in placebo, p<0.001). The efficacy criteria established by the FDA guidance document stipulated that the mean weight loss be 5% greater in drug versus placebo treated patients or, that the proportion of patients losing 5% of their weight be greater in the drug versus placebo-treated groups.
In severely obese patients (BMI ≥40 kg/m2), rimonabant 20-mg decreased body weight and improved metabolic parameters. Rimonabant 20-mg almost halved the percentage of patients with severe obesity after 1 year of treatment. One patient out of 4 lost more than 10% of their baseline body weight, tripling the effect of the diet alone (placebo).
As expected for this chronic disease, when treatment is discontinued after 1 year, body weight regain occurred. In contrast, the body weight loss and metabolic improvements are maintained when rimonabant was continued for up to 2 years as demonstrated in 2 studies [Figure (6.1) 3].
The 5-mg dose had an effect of limited clinical interest on metabolic parameters in spite of a clinically significant effect on body weight compared to placebo.
17/120
1.6.2 Type 2 Diabetes
Two studies were conducted in type 2 diabetic patients:
• RIO-Diabetes was conducted in patients who did not achieve adequate glycemic control on a diet plus a single oral antidiabetic agent (metformin or sulfonylureas). Rimonabant at the dose of 20-mg once daily significantly decreased HbA1c, body weight and TG, and significantly increased HDL-C [Figure (1.6.2) 1]. The effects were demonstrated equally in patients taking metformin or a sulfonylurea. Approximately half of these effects on glucose control and lipids are independent of body weight loss (Section 6.4).
Figure (1.6.2) 1 - RIO-Diabetes study results
• SERENADE was a study comparing rimonabant 20-mg and placebo in oral antidiabetic drug-naive patients. Rimonabant significantly decreased HbA1c and body weight [Figure (1.6.2) 2], and also decreased TG over placebo and increased HDL-C over placebo.
Figure (1.6.2) 2 - SERENADE study results
18/120
In the treatment of obesity, the efficacy of rimonabant 20-mg used in conjunction with a reduced calorie diet and physical exercise was demonstrated in 4 adequate and well-controlled studies with reproducible results, and sustained over time, up to 1 year in all 4 studies, and up to 2 years in 2 studies. Rimonabant 20-mg also demonstrated favorable metabolic effects including increased HDL-C and decreased TG, above and beyond its significant effect on body weight loss. In the treatment of diabetes, 2 studies demonstrated that rimonabant 20-mg significantly improved glucose control and decreased body weight, with the same improvements in dyslipidemia associated with type 2 diabetes as in nondiabetic patients.
1.7 Clinical safety
Overall rimonabant clinical program
The rimonabant clinical program included 59 clinical studies completed as of 01 March 2007, involving over 16 000 subjects or patients, of whom 15 034 were exposed to at least 1 dose of rimonabant, as follows:
- 1190 healthy subjects were enrolled in 40 Phase 1 studies during which they received single (1 mg up to 300 mg) or multiple (1 mg up to 80 mg daily for 6 days up to 4 weeks) doses of rimonabant;
- 1008 patients were enrolled in 6 Phase 2 studies conducted in various indications (obesity, smoking cessation, alcoholism, and schizophrenia), where they received multiple doses of rimonabant 5, 10, 20, or 40 mg daily from 6 weeks up to 24 weeks;
- 12 836 patients were enrolled in 13 Phase 3 studies conducted in 3 different programs [Table (1.7) 1] where they received 5-mg or 20-mg for up to 2 years.
19/120
Table (1.7) 1 - Number of patients exposed to any treatment in completed Phase 3 studies
| Study Name | Placebo | Rimonabant | ||
|---|---|---|---|---|
| 5 mg | 20 mg All doses | |||
| Obese patients | ||||
| RIO-Europe RIO-North America RIO-Lipids RIO-Diabetes* Binge Eaters REBA (Eating behavior) EFC5745 Sub-total obesity Sub total pt-year | 305 1233 342 348 146 80 20 2474 2177 | 603 1214 345 358 ---2520 2421 | 599 1219 346 339 143 76 20 2742 2529 | 1202 2433 691 697 143 76 20 5262 |
| Smokers patients | ||||
| STRATUS-US STRATUS-EU STRATUS-META STRATUS-WW CIRRUS Sub-total smoking cessation Sub total pt-year | 261 260 268 -1453 494 | 262 256 2351 -2869 766 | 261 267 262 3023 754 4567 889 | 523 523 262 5374 754 7436 |
| TOTAL obesity and smoking cessation Total pt-year | 3927 2671 | 5389 3187 | 7309 3419 | |
| Type 2 diabetes patients | ||||
| RIO-Diabetes* SERENADE Sub-total Sub total pt-year | 348 140 488 341 | 358 -358 283 | 339 138 477 328 | 697 138 835 |
| TOTAL obesity, type 2 diabetes and smoking cessation Total pt-year | 4067 2734 | 5389 3187 | 7447 3478 | 12836 |
* Patients from RIO-diabetes study are presented both in obesity and type 2 diabetes and are counted in the totals.
In addition to this large development program, with 3 478 patient-years exposed to rimonabant 20-mg, there were 11 ongoing studies as of 01 March 2007 (3 Phase 1 studies and 8 Phase 3 studies) involving a large population (14 280 patients, blinded treatment,
1:1 randomization ratio of placebo to rimonabant 20-mg, representing 7855 patient-years of which half [3927] apply to rimonabant 20-mg).
20/120
Overall safety profile
The primary focus of this briefing document is to present the data from the obesity and diabetes programs. Global pooling is presented for rare events, and the smoking cessation program is discussed where the profile of safety is different from that observed in the obesity program.
Extensive clinical safety data were collected from this clinical development program and allowed a comprehensive review of the safety profile of rimonabant. In the Phase 1 or Phase 2 studies, the adverse events (AEs) more frequently reported with rimonabant than in placebo treated-individuals were gastrointestinal (GI) (nausea, vomiting), nervous system (dizziness, paresthesia, tremor), psychiatric (anxiety, insomnia), and general (asthenia/fatigue, disturbance in attention) disorders. In the Phase 3 obesity program, the most commonly reported AEs with rimonabant when compared with placebo were consistent with those observed in the Phase 1 and Phase 2 studies [Table (1.7) 2]. In the diabetic population, the events were of the same nature when compared to the obese population as a whole, with the exception of some specific events, possibly condition-related events (eg, hypoglycemia) [Table (7.4.1.2.1) 2].
Table (1.7) 2 - Obesity program - AEs reported in ≥2% in the rimonabant 20-mg group and ≥1% over placebo
| SOC PT | Placebo (N=2474) % | Rimonabant 5 mg (N=2520) % | Rimonabant 20 mg (N=2742) % |
|---|---|---|---|
| Gastrointestinal disorders | |||
| Nausea Diarrhea Vomiting | 4.7 5.8 2.3 | 6.9 7.5 3.0 | 13.6 7.7 4.7 |
| Nervous system disorders | |||
| Dizziness | 4.1 | 5.9 | 7.3 |
| Psychiatric disorders | |||
| AnxietyInsomniaMood alterations with depressive symptoms Depressive disorders | 2.1 3.4 2.8 1.7 | 2.9 3.2 3.6 2.8 | 5.9 5.8 4.7 3.9 |
| Miscellaneous | |||
| Influenza Asthenia/fatigue Contusion Hot flush | 9.1 4.4 1.1 0.8 | 9.2 5.0 1.9 1.3 | 10.3 6.1 3.1 2.0 |
N= patients exposed to treatment at any time.
21/120
The incidence of SAEs displayed by SOC reported in ≥0.5% of patients in the rimonabant 20-mg group and reported more often than placebo is presented in Table (1.7) 3. In the Phase 3 obesity program, the rates of serious AEs (SAEs) were similar across treatment groups [Table (1.7) 3]. Consistent rates were observed in the diabetic population. When considering the global pooling of the obesity and smoking cessation studies, the rates of SAEs were 3.6% in the rimonabant 20-mg group, 3.8% in the rimonabant 5-mg group, and 3.9% in the placebo group.
Table (1.7) 3 - Obesity program - incidence of SAEs displayed by SOC reported in ≥0.5% of the rimonabant 20-mg group and reported more often than placebo
| n (%) | Placebo (N=2474) | Rimonabant 5 mg (N=2520) | Rimonabant 20 mg (N=2742) |
|---|---|---|---|
| Any SAEs Including Fatal SAEs | 106 (4.3) 3 ( 0.12) | 153 (6.1) 3 ( 0.12) | 175 (6.4) 4 (0.15) |
| Neoplasms benign, malignant Infections & Infestations Injury & procedural complications Gastrointestinal disorders Psychiatric disorders | 15 (0.6) 13 (0.5) 8 (0.3) 13 (0.5) 2 (<0.1) | 22 (0.9) 15 (0.6) 17 (0.7) 14 (0.6) 6 (0.2) | 27 (1.0) 21 (0.8) 20 (0.7) 18 (0.7) 15 (0.5) |
N= patients exposed to treatment at any time
SOC = system organ class
Overall, 18 fatal outcomes were reported in the rimonabant clinical program completed as of 01 March 2007; 4 cases were off drug and 14 cases occurred during the study treatment period. Of the 14 cases that occurred during study treatment, 10 were observed in the obesity studies with similar rates across treatment groups (0.15% in the rimonabant 20-mg group, 0.12% in the rimonabant 5-mg group, and 0.12% in the placebo group) [Table (1.7) 4]. Concerning the 4 other cases, 1 case of subdural hemorrhage was observed in the SERENADE study in the placebo group, and the 3 others occurred during the smoking cessation studies (2 in the rimonabant 20-mg group: coronary artery sclerosis and road traffic accident, and 1 cardio-respiratory arrest in the placebo group). Thus, deaths were equally distributed across treatment groups and did not show any specific pattern.
22/120
Table (1.7) 4 - Obesity program - fatal SAEs
| Reason for death | Placebo (N=2474) | Rimonabant 5 mg (N=2520) | Rimonabant 20 mg (N=2742) |
|---|---|---|---|
| Fatal SAEs | 3 (0.12) | 3 (0.12) | 4 (0.15) |
| Cardiac arrest Cardiac failure Coronary artery disease Cerebral hematoma/CVA Cerebral hemorrhage Pulmonary embolism Septic shock Completed suicide Road traffic accident (as passenger) Uterine cancer (end-stage) | ---1 1 1 ---- | 1 -----1 1 -- | -1 1 -----1 1 |
N= patients exposed to treatment at any time.
In the Phase 3 program, the primary reasons for discontinuing rimonabant reflected the common safety profile of the drug and the requirement to discontinue for antidepressant treatment since the use of antidepressants often cause weight gain [Table (1.7) 5].
Table (1.7) 5 - Obesity program - discontinuation due to AEs reported in ≥0.5% in the rimonabant 20-mg group
| Placebo (N=2474) % | Rimonabant 5 mg (N=2520) % | Rimonabant 20 mg (N=2742) % | |
|---|---|---|---|
| Patient discontinuations due to AEs | 7.6 | 11.0 | 15.5 |
| Depressive disorders Nausea AnxietyMood alterations with depressive symptoms Dizziness Pregnancy*Headache Insomnia | 1.0 <0.1 0.2 0.7 0.1 0.4 0.3 0.2 | 1.5 0.2 0.3 1.0 0.2 1.3 0.4 0.1 | 2.3 1.5 1.2 1.2 0.7 0.7 0.5 0.4 |
N= patients exposed to treatment at any time. * pre-specified withdrawal criterion as per protocols
23/120
Two categories of AEs of interest have been identified and are being closely monitored. These are psychiatric events (anxiety and depressive events, including suicidality) and neurological events (sensory changes, motor impairment, and cognitive difficulties) including seizures.
Psychiatric events
Anxiety disorders and symptoms included various manifestations of anxiety (such as anxiety, nervousness, panic attack, or panic reaction), with higher incidences in the rimonabant groups when compared with placebo [Table (1.7) 6]. Most of these cases were mild and transient.
These events were nonserious in a large majority of cases, but 2 cases were serious under rimonabant. Corrective treatment was reported (mostly anxiolytics in 48% and 41%of cases for placebo and, rimonabant respectively). Eighty percent of patients experiencing these events continued rimonabant treatment.
Table (1.7) 6 - Obesity programs - anxiety disorders and symptoms
| ANXIETY DISORDERS AND SYMPTOMS | Obesity studies | |||
|---|---|---|---|---|
| Placebo (N=2474) n (%) | 5 mg (N=2520) n (%) | 20 mg (N=2742) n (%) | ||
| Anxiety disorders & symptoms | 100 (4.0) | 141 (5.6) | 278 (10.1) | |
| Anxiety symptoms Panic disorders Specific and social phobic disorders Stress disorders Anxiety disorders nec Fear symptoms | 95 (3.8) 1 (<0.1) 2 (<0.1) 2 (<0.1) 1 (<0.1) 0 (0.0) | 130 ( 5.2) 4 ( 0.1) 0 ( 0.0) 7 ( 0.3) 1 (<0.1) 0 ( 0.0) | 246 ( 9.0) 23 ( 0.8) 5 ( 0.2) 6 ( 0.2) 3 ( 0.1) 1 (<0.1) | |
N= patients exposed to any treatment throughout entire study
24/120
Depressive events were more frequent under rimonabant [Table (1.7) 7]. In the clinical Phase 3 program, patients with a past history of severe depression or patients with current severe psychiatric illness were excluded. Antidepressant treatment was not permitted and warranted mandatory treatment discontinuation, as prespecified by the protocols.
Table (1.7) 7 - Number (%) of patients with depressed mood disorders and disturbances (≥0.1%) in obesity
| Placebo (N=2474) n (%) | 5 mg (N=2520) n (%) | 20 mg (N=2742) n(%) | |
|---|---|---|---|
| Depressed mood disorders and disturbances | 112 ( 4.5) | 158 ( 6.3) | 231 ( 8.4) |
| Depressive disorders Depression Dysthymic disorder Major depression | 43 ( 1.7) 34 ( 1.4) 0 ( 0) 9 ( 0.4) | 70 ( 2.8) 56 ( 2.2) 7 ( 0.3) 7 ( 0.3) | 106 ( 3.9)87 ( 3.2)4 ( 0.1)15 ( 0.5) |
| Mood alterations with depressive symptoms Anhedonia Decreased interest Depressed mood Depressive symptom Feeling of despair Tearfulness | 70 ( 2.8) 1 (<0.1) 1 (<0.1) 57 ( 2.3) 14 ( 0.6) 0 ( 0) 0 ( 0) | 91 ( 3.6) 0 ( 0) 0 ( 0) 76 ( 3.0) 15 ( 0.6) 1 (<0.1) 0 ( 0) | 129 ( 4.7)3 ( 0.1)0 ( 0)96 ( 3.5)29 ( 1.1) 0 ( 0)4 ( 0.1) |
N= patients exposed to any treatment throughout entire study
Two different categories of events were reported: depressive disorders and mood alterations with depressive symptoms. In the obesity program, depressive disorders (depression, major depression, and dysthymic disorders) were reported more frequently in the rimonabant 20-mg group (3.9%) when compared with the placebo group (1.7%). Around 70% of patients required corrective treatment (placebo: 31/43; rimonabant 5-mg: 48/70; rimonabant 20-mg: 76/106) and the rate of discontinuation due to depressive disorders was around 60% (placebo: 25/43; rimonabant 5-mg: 39/70; rimonabant 20-mg: 64/106). Four patients under rimonabant 20-mg were hospitalized. These events occurred early in the course of the treatment, with one-half of the events starting within 3 months after rimonabant was given. Past history of depressive disorders is the main risk factor that has been identified (8-fold, regardless of the group) for depressive disorders.
Mood alterations were reported more frequently in the rimonabant 20-mg group when compared with the placebo group in obese patients (rimonabant 20-mg: 4.7%, placebo: 2.8%). Contrary to what was observed for depressive disorders, most of the patients with mood alteration events were not treated (corrective therapy 28.7% for rimonabant compared to 34.8% for placebo), and the study treatment was continued for 75% of the patients with this type of event, whatever the group of treatment. No mood alteration event led to hospitalization.
25/120
Although the literature is unclear with respect to the relationship between weight loss and depression, increased rates of psychiatric events, including depression, have been reported with weight loss agents in their package inserts: orlistat (Xenical) (depression: 3.4% versus 2.5% at Year 2) and sibutramine (Meridia) (depression: 4.3% versus 2.5%, and emotional liability: 1.3% versus 0.6%) for active drug versus placebo, respectively.
In the completed studies, to evaluate the risk of suicidality with the use of rimonabant, a specific analysis was performed by the Sponsor on the cases of suicide, suicide attempt, or suicide ideation reported as AEs or on associated symptoms of a psychiatric AE. In addition, an independent, blinded assessment was also performed by Columbia-Classification of Adult Suicidality Assessment (C-CASA) to identify definite cases (completed suicide, suicide attempt, preparatory acts toward imminent suicide behavior, suicidal ideation) and possible cases (not enough information fatal and non fatal).
From this analysis, no difference was observed when pooling all indications (obesity, smoking, alcohol, and schizophrenia). An imbalance was seen, however, in suicidal ideation in the obesity studies (0.63% in rimonabant versus 0.38% in placebo). All cases of suicidal ideation were associated with depressive events or adjustment disorders. Only one case was fatal (rimonabant 5-mg) assessed as a possible case (not enough information). This case had been reported as a completed suicide by the Investigator.
Section 7.5.1.2 details the C-CASA method of classification of cases according to the 9 categories of interest.
In the ongoing studies, the C-CASA was not yet applied. As of 01 March 2007 all cases of suicide and suicide attempts (as reported by the Investigator) were unblinded, 1 suicide was reported in a patient treated with rimonabant 20-mg and 2 suicide attempts in patients receiving placebo. The rate of suicidal ideation (0.3%) is similar to what was observed in completed studies. Around 40% of them were serious and unblinded. Among them an imbalance was observed (0.25% in rimonabant versus 0.11% in placebo). This trend is similar to that observed in the complete Phase 3 obesity studies.
Sanofi-aventis is proposing that rimonabant should not be initiated in patients with uncontrolled psychiatric illness such as a major depression. There is limited data in patients taking antidepressant medication in combination with rimonabant; therefore, the use of rimonabant is not recommended in these patients.
In addition to proposed labeling, sanofi-aventis is proposing a Risk Minimization Action Plan (Risk MAP) to reduce the possibility of rimonabant use inconsistent with the labeling in patients with diseases, conditions, or concomitant therapy that raise identified or potential safety concerns. This will be done by increasing the knowledge of key stakeholders (including prescribers/health care professionals, pharmacists, and patients) on the safety and efficacy profile of rimonabant through a Targeted Education and Outreach Program and a Reminder System. The overall Risk MAP effectiveness will be assessed periodically and further ongoing and iterative development is expected with
26/120
follow-up actions implemented, as needed. Details regarding the proposed Risk MAP are presented in Section 9 of this briefing package.
Neurological events
In general, neurological events occurred more frequently in the rimonabant groups when compared with the placebo groups. Events were classified in 3 main categories: sensory changes, motor impairments, and cognitive difficulties [Table (1.7) 8].
Table (1.7) 8 - Neurological AEs reported with an incidence of ≥1% in obesity studies
| Neurological Symptoms | Placebo (N=2474) N (%) | 5 mg (N=2520) N (%) | 20 mg (N=2742) N (%) |
|---|---|---|---|
| Any Class - any event | 310 (12.5) | 403 (16.0) | 554 (20.2) |
| Sensory changes | |||
| Any event Dizziness Paresthesia Dysgeusia Hypoesthesia Sciatica | 230 ( 9.3) 101 ( 4.1) 22 ( 0.9) 6 ( 0.2) 21 ( 0.8) 16 ( 0.6) | 299 (11.9) 148 ( 5.9) 25 ( 1.0) 6 ( 0.2) 35 ( 1.4) 23 ( 0.9) | 406 (14.8) 200 ( 7.3) 41 ( 1.5) 8 ( 0.3) 39 ( 1.4) 34 ( 1.2) |
| Motor impairment | |||
| Any event Tremor | 57 ( 2.3) 2 (<0.1) | 68 ( 2.7) 6 ( 0.2) | 93 ( 3.4) 24 ( 0.9) |
| Cognitive difficulties | |||
| Any event Somnolence | 51 ( 2.1) 5 ( 0.2) | 66 ( 2.6) 12 ( 0.5) | 113 ( 4.1) 14 ( 0.5) |
Among sensory changes, dizziness was the most frequent event. Episodes of dizziness occurred in 7.3% of patients in the rimonabant 20-mg group compared to 4.1% of patients in the placebo group, generally during the first month following the first study drug intake. Reports of dizziness were more frequent in the rimonabant groups compared with the placebo groups for patients with type 2 diabetes mellitus (9.6% versus 4.3%, respectively) and in elderly patients (15.3% versus 2.7%, respectively). One case of dizziness was serious in the rimonabant 20-mg group and dizziness led to study drug discontinuation in 9% of patients in that group. These events were not associated with hypotension, hypoglycemia, or falls.
Paresthesia was a common AE among patients with type 2 diabetes mellitus (2.9% with rimonabant versus 0.6% for placebo).
Tremor was the most frequently reported AE within the motor impairment category. Tremor was reported with a higher incidence in the rimonabant 20-mg groups (0.9%) compared with the placebo group (<0.1%), occurred within the first 2 months of introduction of rimonabant, and resulted in treatment discontinuation in 20% of patients in the rimonabant 20-mg groups and none in the placebo groups.
27/120
Cognitive difficulties were more commonly reported with rimonabant compared to placebo (4.1% versus 2.1%, respectively, in obesity studies) and included somnolence/sedation/lethargy and memory loss (1.5% versus 0.7% in obesity studies). These events were more frequently described in the smoking cessation studies and could be linked to nicotine withdrawal. None of the cases were serious and they rarely led to treatment discontinuation for any of the treatment groups.
For patients receiving rimonabant in the Phase 3 obesity and smoking studies, 0.2% reported confusion or disorientation. Other neurological events were of various types and included ischemic or hemorrhagic cerebral stroke, localized motor deficits, or other atypical neurological signs. These reports were either isolated cases or reported in comparable proportions of patients across treatment groups, including the placebo group.
Seizures
Nonclinical studies suggest a weak proconvulsant potential for rimonabant when combined with pentylenetetrazole, which alone induces convulsions, or under stressful conditions (experimental procedures) in repeated-dose toxicity studies in some species. Due to its potential impact on the central nervous system (CNS), intensive electroencephalogram (EEG) monitoring was systematically performed throughout the Phase 1 clinical program. Results showed a comparable frequency of EEG changes in the rimonabant group compared with the placebo group.
In the Phase 3 clinical studies, patients with treated epilepsy were excluded. However, patients with a prior history of seizures were permitted. Rare cases of seizures were reported during the clinical development program (completed studies), regardless of the indication: placebo: 0.232% patient-years versus all doses of rimonabant: 0.158% patient-years [relative risk 0.68; CI (0.31, 1.51)].
Sanofi-aventis is recommending that rimonabant be used with precaution in patients with epilepsy. Results from non-clinical studies suggest that rimonabant has a weak proconvulsant potential when administered to animals, which is not confirmed in the clinical situation. A comprehensive evaluation of the available clinical data has shown that cases of seizures have been rarely reported in patients treated with rimonabant with no evidence of an increase in the seizure rate in patients treated with rimonabant 20-mg compared to placebo. Rimonabant has not been studied in patients being treated for epilepsy. Rimonabant should be used with caution in patients being treated for epilepsy.
28/120
In addition to proposed labeling, sanofi-aventis is proposing a Risk Minimization Action Plan (RiskMAP) to reduce the possibility of rimonabant use inconsistent with the labeling in patients with diseases, conditions, or concomitant therapy that raise identified or potential safety concerns. This will be done by increasing the knowledge of key stakeholders (including prescribers/health care professionals, pharmacists, and patients) on the safety and efficacy profile of rimonabant through a Targeted Education and Outreach Program and a Reminder System. The overall Risk MAP effectiveness will be assessed periodically and further ongoing and iterative development is expected with follow-up actions implemented, as needed. Details regarding the proposed Risk MAP are presented in Section 9 of this briefing package.
Other safety parameters
Standard laboratory tests did not provide evidence of safety concerns with respect to the main biological functions (liver, renal, hematology, electrolytes, and metabolism). Similarly, the review of vital signs [supine blood pressure (BP) and heart rate (HR)] did not raise any safety concerns. In the specific ECG Phase 1 study designed to evaluate the electrocardiographic safety of rimonabant, there was no evidence of a potential for rimonabant to prolong ventricular repolarization. Extensive ECG records from the Phase 3 studies confirmed these observations.
In conclusion, the safety of rimonabant has been characterized by the completion of the development program in obesity and patients with type 2 diabetes mellitus. Additional safety information was obtained from the studies conducted in smoking cessation. Based on these trials, the most frequently reported adverse reactions were of GI, nervous system, or psychiatric origin. Adverse events of interest have been identified and are being closely monitored and assessed in both the ongoing studies with rimonabant, as well as from the post-marketing surveillance reports from countries where the drug is currently marketed. A risk management plan has been instituted in the European Union and proposed in the US to facilitate both minimization and management practices in order to provide the greatest benefit to patients prescribed rimonabant.
1.8 Post-marketing experience
In June 2006, the European Medicines Evaluation Agency (EMEA) approved rimonabant for marketing authorization under the trade name ACOMPLIA®. As of April 2007, ACOMPLIA has been launched in the following countries: Argentina, Austria, Chile, Columbia, Cyprus, Denmark, Finland, France, Germany, Greece, Iceland, Ireland, Mexico, Norway, Sweden, and the United Kingdom (UK). A total of 108 730 people have been prescribed rimonabant in these countries as of 01 March 2007.
In addition, rimonabant has received marketing authorization but has not yet been launched in Belgium, Brazil, Bulgaria, the Czech Republic, Estonia, Guatemala, Hong Kong, Hungary, Italy, Kuwait, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Switzerland, the United Arab Emirates, and the Ukraine.
29/120
Assessment of drug risks and benefits starts during drug development and continues as part of risk management/minimization as the drug product moves out of the development stage. At product approval, the company works with the relevant regulatory authorities on product labeling that describes the conditions for safe and effective use of the drug. Product labeling is the basis for further risk management activities as the drug is marketed.
Based on a thorough review of all available preclinical and clinical safety data from the premarketing studies, sanofi-aventis developed a risk management plan (RMP) in Europe to monitor and manage identified and potential risks. Agreed upon by the EMEA, the first European RMP (EU-RMP) was dated 22 June 2006.
Preliminary results available on the EU-RMP from the European countries that have launched rimonabant indicate that the drug is used consistent with product labeling in more than 95% of the cases.
1.8.1 US proposed Risk Management Plan
In the US, sanofi-aventis is proposing a risk management plan to minimize identified and potential important risks, with particular emphasis on exposure of the drug to patient populations or conditions where patient safety risks may exist, or where safety information is not complete. In addition, since obesity is a chronic disorder, short term use is not recommended and the company does not intend to promote the product for short term cosmetic use. The risk minimization activities that have been launched successfully in the European Union will serve as the foundation to the RiskMAP proposed in the US. Sanofi-aventis is committed to providing information to key stakeholders (eg, healthcare prescribers, pharmacists, and patient) on the safety and efficacy profile of ZIMULTI® (rimonabant) and use of the drug consistent with product labeling and risk minimization.
The overall strategy proposed in the US RiskMAP is to accomplish risk minimization through a Targeted Education and Outreach Program and a Reminder System using some of the tools described later in the briefing package. The effectiveness of the proposed RiskMAP will be assessed periodically and the RiskMAP will be subject to ongoing and iterative development. Details regarding the proposed US RiskMAP are provided in Section 9.0 of this briefing package.
1.9 Benefit-risk ratio assessment
The efficacy and the safety of rimonabant in obese and overweight patients with cardiovascular risks were demonstrated in 4 adequate and well-controlled studies with reproducible results, and sustained over time, for up to 1 year (in all 4 studies) and 2 years (in 2 studies). In type 2 diabetic patients, 2 studies were conducted: RIO-Diabetes in patients not adequately controlled by single oral antidiabetic agent and SERENADE in drug naive patients. The 2 studies demonstrated that rimonabant 20-mg significantly decreases HbA1c values compared to single oral antidiabetic agents or placebo. The 20-mg dose of rimonabant had favorable metabolic effects including improved insulin sensitivity,
30/120
increased HDL-C and decreased TG, and improved glucose control in type 2 diabetes, above and beyond its effect on body weight loss. Based on this body of data, rimonabant used in conjunction with a reduced calorie diet and physical exercise will be an important option in the management of obese and overweight patients with associated cardiovascular risk, including patients with type 2 diabetes.
Despite all efforts from the medical community and patients to fight the obesity epidemic using diet, counseling, and approved and non approved drugs, there is still a medical need for effective new pharmacotherapies that can help to achieve significant body weight loss and prevent short term relapse. Meaningful mean body weight loss, confirmed by the 2-to 3-fold higher rate of 5% and 10% responders compared with diet alone, observed after 1 year, demonstrates that rimonabant 20-mg results in sustained body weight loss and waist circumference reduction. Severe obesity (BMI of 40.0 kg/m² or higher) is increasing much faster than obesity, leading to more and more bariatric surgery and its related morbidity/mortality. Rimonabant 20-mg almost halved the percentage of patients with severe obesity after 1 year of treatment. One patient out of 4 lost more than 10% of their baseline body weight, tripling the effect of the diet. In severely obese patients, rimonabant demonstrated improvements in lipids and glucose/insulin homeostasis similar to the improvements seen in non-severely obese patients.
In 1996, the FDA published the draft guidance for the Clinical Evaluation of Weight-Control Drugs. This guidance outlined the criteria for regulatory approval for drugs to treat obesity, which included a minimum of one year of placebo-controlled exposure in 1500 patients and a second year of exposure in up to 500 patients. The efficacy criteria established by the guidance stipulated that the mean weight loss be 5% greater in drug versus placebo-treated patients or, that the proportion of patients losing 5% of their weight be greater in the drug versus placebo-treated group. Additionally, the populations to be studied in these trials were recommended to be patients with a body mass index (BMI) greater than 30 or greater than or equal to 27 with comorbidities such as hypertension, type 2 diabetes, and dyslipidemia. Earlier this year, the FDA published for public review and comment, a revised draft guideline entitled: Developing Products for Weight Management. The rimonabant development program has met the safety and efficacy criteria outlined in both the 1996 and the 2007 draft guidance documents.
Additionally, in 1998-2000 the NIH established clinical guidelines for the treatment of overweight and obesity. In addition to establishing a BMI classification based on epidemiology data that showed increases in mortality with BMIs greater than 25, the guidelines also stated that “weight loss medications should be used only by patients who are at increased medical risk because of their weight and should not be used for cosmetic weight loss.” Furthermore, weight loss medications should never be used without concomitant lifestyle modifications. Importantly, the guidelines also recognized that obesity is a chronic disorder (disease) and that short-term treatment with anti-obesity pharmacotherapy is not helpful. Rimonabant has been developed for those patients at increased medical risk due to their overweight or obesity. Rimonabant, in combination with lifestyle modifications, provides a medical therapy for a chronic disease.
31/120
Both the American Heart Association (AHA) and American Diabetes Association (ADA) emphasize the importance of weight management in cardiovascular risk including diabetes. Thus, there is a need for treatment of type 2 diabetic patients that will not only lower HbA1c, but will also decrease body weight and the metabolic disorders associated with excess weight. The RIO-Diabetes study in treated type 2 diabetic patients showed that rimonabant, at the dose of 20-mg once daily, could accomplish these goals, when diet plus the single agent (metformin or sulfonylureas) did not result in adequate glycemic control. The 0.7% decrease in HbA1c over the placebo effect, with 67.9% of patients reaching an HbA1c level <7% demonstrated a significant improvement in glucose control. The 6-month SERENADE study in treatment-naive type 2 diabetic patients confirmed the benefits seen with rimonabant in the 12-month RIO-Diabetes study. The SERENADE study showed a placebo-adjusted decrease in HbA1c of 0.5% in the overall population and a 1.2% decrease in patients with a baseline HbA1c above 8.5%. About one-half of the benefit was beyond what was expected from weight loss alone. Moreover, the 4.0 kg
(8.8 lbs) placebo-adjusted weight loss and the 3.7 cm decrease in waist circumference point toward a reduction in cardiovascular risk. Thus, both the RIO-Diabetes and SERENADE studies showed that rimonabant significantly improves glycemic control in type 2 diabetic patients, with significant body weight loss and a positive impact on the lipid profile, a triple benefit that should be considered a very important addition to the currently available treatments, and in direct contrast to the majority of anti-diabetic treatments that increase body weight. These results suggest that the addition of 20-mg rimonabant to metformin or sulfonylurea appears as an effective therapeutic option for the management of patients with type 2 diabetes.
The safety profile of rimonabant was documented by a database comprising the obesity and diabetes programs, including more than 13 000 individuals treated, contributing more than 3000 patient-years of experience at the 20-mg dose. A similar safety profile was observed between the obesity and diabetes populations with GI events (such as nausea, vomiting, and or diarrhea) psychiatric disorders (such as insomnia, anxiety, depressive mood disorders, and disturbances), and nervous system disorders (such as dizziness) were being reported in rimonabant-treated patients.
Anxiety, depressed mood, depressive disorders, and dizziness, were observed in the 20-mg rimonabant group, and were usually mild and transient. The discontinuation rate due to depressive disorders, in part, reflects the protocol mandate for discontinuation when antidepressive therapy was required due to the confounding effect of antidepressants on body weight. Based on this safety experience, therapy with rimonabant should not be initiated in patients with uncontrolled serious psychiatric illness such as a major depression. Appropriate treatment of this condition should be initiated first and therapy with rimonabant considered once this psychiatric condition is controlled and the patient is no longer taking antidepressant medication. As there is limited data in patients with antidepressant medication in combination with rimonabant, use of rimonabant is not recommended in these patients. In type 2 diabetes, rimonabant was well tolerated with a somewhat greater incidence of paresthesia. Importantly, there was a low risk of hypoglycemia, a finding of particular interest for use in patients already on medications that have such adverse reactions.
32/120
Dizziness, paresthesia, hypoesthesia, and with a less frequency tremor, memory impairment, confusion, and disorientation all occurred in a small number of patients more frequently in the rimonabant 20-mg group than in the placebo group. No serious neurological AEs occurred that were related to rimonabant treatment. A total of 3 (2 during rimonabant 5-mg treatment; 1 during placebo) cases of multiple sclerosis were reported in the clinical development program. No cases were reported in the rimonabant 20-mg group or during the second year of exposure. In addition, 1 pati