1. for The STRADIVARIUS Investigators
STRADIVARIUS Effect of Rimonabant on Progression of Atherosclerosis in Patients with Abdominal Obesity and Coronary Artery Disease
Steven E. Nissen MD
Stephen J. Nicholls MBBS PhD, Kathy Wolski MPH, Josep Rodés-Cabau MD, Christopher P. Cannon MD, John E. Deanfield MD, Jean-Pierre Després PhD, John JP Kastelein MD PhD, Steven R. Steinhubl MD, Samir Kapadia MD, Muhammad Yasin MD, Witold Ruzyllo MD, Christophe Gaudin MD, Bernard Job MD, Bo Hu PhD, Deepak L. Bhatt MD, A. Michael Lincoff MD, and E. Murat Tuzcu MD
for The STRADIVARIUS Investigators
Chairmen, ladies and gentlemen, on behalf of the investigators, I am pleased to present to you the results of the STRADIVARIUS Trial examining the effect of rimonabant on progression of atherosclerosis in patients with abdominal obesity and coronary disease.

2. Background and Objectives
Obesity is increasing at an alarming rate in developed countries; 34% in the US population have BMI >30.
Abdominal obesity is associated with specific metabolic abnormalities that increase the risk of CAD.
Rimonabant, a cannabinoid receptor (CB1) antagonist, enhances weight loss and improves obesity-related metabolic abnormalities.
We sought to determine if rimonabant could reduce progression of coronary atherosclerosis measured by IVUS in abdominally-obese CAD patients.
Obesity is increasing at an alarming rate in developed countries - 34% of the US population have a BMI greater than 30.
Abdominal obesity is associated with specific metabolic abnormalities that increase the risk of coronary disease.
Rimonabant, a cannabinoid receptor or CB1 antagonist, enhances weight loss and improves obesity-related metabolic abnormalities.
We sought to determine if rimonabant could reduce progression of coronary atherosclerosis measured by IVUS in abdominally-obese CAD patients.

3. Methods
Patients selected with abdominal obesity (defined as waist >102 cm for men or >88 cm for women) undergoing angiography for clinical indications.
Inclusion criteria required two additional risk factors of the metabolic syndrome or current smoking.
Intravascular ultrasound (IVUS) was performed to assess atheroma volume in 839 patients randomized to placebo or rimonabant 20 mg.
After 18 months, repeat IVUS was performed in the 676 patients who remained in the trial, regardless of whether they were still taking study drug.
Patients were selected with abdominal obesity, defined as a waist circumference greater than 102 cm for men or 88 cm for women, who were undergoing angiography for clinical indications.
Inclusion criteria required 2 additional components of the metabolic syndrome or current smoking.
IVUS was performed to assess atheroma volume in 839 patients randomized to placebo or rimonabant 20 mg.
After 18 months, repeat IVUS was performed in the 676 patients remaining in the trial, regardless of whether they were still taking study drug.

4. Intravascular Ultrasound Endpoint Calculations
Lumen Area
Atheroma Area
EEM Area
Change in Percent Atheroma Volume
AtheromaCSA
EEMCSA =  n –
(Month 18)
(baseline) n 
Total Atheroma
Volume
EEMCSA 
Median number of slices in all pullbacks –
LumenCSA
The primary IVUS endpoint is shown in the top panel, percent atheroma volume, which represents the percent of the external elastic membrane area occupied by atheroma.
The secondary IVUS endpoint was total atheroma volume, which is the absolute change in the sum of atheroma areas adjusted for the length of each patient’s pullback. n = x
Number of slices in patient’s pullback
Change in
Atheroma Volume
(Month 18)
(baseline) = –

5. Baseline Patient Characteristics (n=839)
Placebo(n=417)
Rimonabant(n=422)
Age (years)
57.5
57.9
Male gender
65.0%
64.9%
Weight (kg)
103.5
Waist Circum. (cm) [in]
117.5 [46.3]
117.3 [46.2]
BMI
35.3
Diabetes
37.4%
38.4%
Metabolic Syndrome
91.6%
94.1%
Current Smoker
26.6%
29.9%
Psychiatric Disease
24.5%
25.6%
At baseline, patients in both groups were approximately 58 years of age and about two-thirds were male.
Weight was 105 kg and waist circumference 117 cm or 46 inches!
The mean BMI was 35 - about one-third of patients were diabetic.
Approximately 25% of participants had a history of prior psychiatric disease.

6. Medications at Randomization (n=839)
Placebo (n=417)
Rimonabant(n=422)
Aspirin
91.1%
91.7%
ß-blocker
70.5%
69.4%
ACEi or ARB
68.6%
Statin
81.8%
82.5%
Oral hypoglycemia agent
29.7%
30.6%
Insulin
11.8%
11.1%
Antidepressant
19.2%
18.2%
Baseline medications included evidence-based therapies directed at cardiovascular risk in a high percentage of patients, including aspirin in more than 90% and statins in more than 80% of patients.
Approximately 20% were taking an anti-depressant.

7. Baseline Lab Values & Blood Pressure (n=676)
Placebo (n=341)
Rimonabant(n=335)
LDL-cholesterol (mg/dL)
89.5
91.9
HDL-cholesterol (mg/dL)
37.6
38.5
Triglycerides* (mg/dL)
140.0
hsCRP* (mg/L)
3.8
3.4
HbA1c (%)
5.8
Systolic BP (mmHg)
129.3
129.4
Diastolic BP (mmHg)
76.7
76.9
Baseline laboratory values and blood pressure indicated good risk factor control with an LDL of approximately 90. HDL levels were low, triglycerides were borderline high, and CRP levels elevated, findings typical of abdominal obesity. Blood pressure averaged approximately 129/77.
*median values

8. Weight and Waist Circumference Changes
Body Weight (kg)
Waist Circumference (cm)
-0.5 kg
-1.0 cm
P < 0.001
P < 0.001
Change in Body Weight (kg)
Change in Waist Circumference (cm)
-4.5 cm
During the course of the trial, rimonabant-treated patients, shown in orange, lost 4.3 kg and 4.5 cm in waist circumference, P<0.001 compared with placebo, shown in blue.
-4.3 kg
Months after Randomization

9. Percent Changes in Biochemical Parameters
HDL-cholesterol
HbA1c in Diabetics (n=248)
22.4%
0.42%
P=0.001
P<0.001
6.9%
-0.13%
hs C-reactive Protein
Triglycerides
Substantial differences in biochemical effects were observed. Rimonabant patients are shown in orange and placebo patients in blue. HDL increased 22% compared with 7% (upper left). C-reactive protein fell 50% compared with 31% (lower left) and triglycerides fell 20% compared with 6% (lower right). In the 248 diabetics, shown in the upper right, hemoglobin A1c decreased in the rimonabant group, -0.13%, and increased 0.42% in the placebo- treated patients. All 4 comparisons were significant at P<0.001.
-6.2%
P<0.001
-30.9%
P<0.001
-20.5%
-50.3%

10. Primary and Secondary IVUS Endpoints
PAV: Primary Endpoint
TAV: Secondary Endpoint
P < 0.001
P = 0.37
0.51
0.88
P=0.22
P=0.03
P =0.09
0.25
In the left panel, PAV, the primary endpoint, showed a lower rate of progression in the rimonabant- treated patients, shown in orange, that did not reach statistical significance, P= However, TAV, the secondary endpoint, in the right panel, showed a significant decrease in atheroma volume in the rimonabant group, p = Therefore, although the study failed to show a significant effect for the primary endpoint, a statistically significant effect favoring rimonabant was observed for the secondary IVUS endpoint.
-2.2
P=0.03
Placebo
Rimonabant

11. Additional Exploratory IVUS Endpoints
Maximum Atheroma Thickness
TAV in Most Diseased Segment
P < 0.001
0.01
P = 0.37
P =0.01
-0.89
P = 0.05
We also computed rates of disease progression for two additional exploratory IVUS endpoints used in other clinical trials. In the left panel, the mean maximal atheroma thickness, showed a statistically significant difference between placebo, in blue, and rimonabant, in orange, with a P value of In the most-diseased 10 mm sub-segment, a numerically lower rate of progression was observed in the rimonabant group, which did not reach statistical significance, P=0.37.
-1.47
P = 0.88
P = 0.001
Placebo
Rimonabant

12. Primary Endpoint in Selected Subgroups
≥ 58 years
< 58 years
Age
Male
Female
Gender
≥ median
< median
BMI
(34.3)
Baseline HDL-C
≥ median
< median
(38.6) }
Statin use
Yes (560) no (116)
P = 0.03
We examined the primary endpoint in non-pre- specified subgroups.
Statistical heterogeneity was observed for two groups, patients not treated with statins and those with triglyceride levels greater than the median value.
These hypothesis-generating findings suggest that there may be subgroups of patients in whom rimonabant produces a greater extent of benefit. }
Triglyc- erides
≥ median
< median
P = 0.03
(140.0)
≥ median
< median
hsCRP
(3.5) 2 1 1 2
Favors Rimonabant
Favors Placebo

13. Major Adverse Cardiovascular Events
Placebo (n=417)
Rimonabant(n=422)
Composite of CV death, nonfatal MI, nonfatal stroke, or hospitalization for revascularization, unstable angina, TIA
11.0%
10.4%
Composite of CV death, nonfatal MI, or nonfatal stroke
1.7%
3.1%
Cardiovascular death
0.5% 0%
All cause mortality*
1.9%
Nonfatal MI
1.0%
2.1%
Fatal or nonfatal stroke
0.2%
0.9%
Hospitalization for revascularization, unstable angina, or TIA
9.6%
8.5%
Major adverse cardiovascular events were similar in both treatment groups, although the study was not powered for clinical outcomes.
A single endpoint approached statistical significance, all cause mortality, which occurred in 1.9% of placebo-treated patients and 0.5% of rimonabant treated patients, p=0.06.
*P = 0.06

14. Psychiatric Adverse Effects
Placebo (n=417)
Rimonabant(n=422) P
value
Psychiatric Disorders
28.4%
43.4%
<0.001
Anxiety
11.8%
18.0%
0.01
Depression
11.3%
16.8%
0.02
Insomnia
9.1%
12.3%
0.14
Depressed Mood
4.8%
6.9%
0.20
Major Depression
2.2%
3.1%
0.41
Suicidal ideation
2.4%
1.7%
0.44
Suicide Attempt – n (%)
1 (0.2%)
0 (0%)
0.50
Completed suicide – n (%)
Severe psychiatric disorders*
3.8%
4.7%
0.52
Psychiatric adverse effects were collected prospectively using a questionnaire and occurred in 43.4% of rimonabant-treated patients compared with 28.4% of placebo treated patients, P< These differences were driven primarily by anxiety and depression.
A single patient in the placebo group attempted suicide and a single patient in the rimonabant group completed suicide.
* Major depression, suicidal ideation, attempted or successful suicide

15. Other Treatment-Emergent Adverse Effects
Placebo (n=417)
Rimonabant(n=422)
P value
GI Disorders
17.8%
33.6%
<0.001
Nausea
5.5%
14.9%
Diarrhea
3.4%
7.8%
0.005
Vomiting
1.9%
0.01
Erectile Dysfunction*
0.7%
3.3%
0.04
Dizziness
12.7%
14.5%
0.47
Fatigue
6.0%
10.9%
Other treatment-emergent adverse effects included GI disorders such as nausea, diarrhea, and vomiting, erectile dysfunction and fatigue, all of which were more common in the rimonabant treatment group.
*n = 545 male patients

16. Conclusions
Treatment of abdominally-obese coronary disease patients for 18 months with rimonabant:
Reduced body weight 4.3 kg and waist circumference 4.5 cm, increased HDL-C 22.4%, reduced triglycerides 20.5%, hsCRP 50.3%, and favorably affected HbA1c.
The study did not demonstrate an effect for rimonabant on the primary endpoint, PAV (P = 0.22), but a favorable effect for the secondary endpoint, TAV (P = 0.03).
Psychiatric and GI adverse effects were more common with rimonabant, which resulted in a higher rate of drug discontinuation.
We conclude that treatment of abdominally-obese coronary disease patients for 18 months with rimonabant reduced body weight 4.3 kg and waist circumference 4.5 centimeters, increased HDL 22%, reduced triglycerides 20%, high sensitivity CRP 50%. and favorably affected hemoglobin A1c.
Although this study did not demonstrate an effect for rimonabant on the primary endpoint, percent atheroma volume, P=0.22, we did observe a favorable effect for the secondary endpoint, total atheroma volume, P=0.03.
Psychiatric and GI adverse effects were more common with rimonabant, which resulted in a higher rate of drug discontinuation.

17. The authors appreciate the opportunity to publish these results simultaneously in the Journal of the American Medical Association, which is now available on line.

18. Some Final Thoughts
Development of effective and durable treatment strategies for management obesity has proven a daunting challenge.
New approaches are greatly needed to reduce the burdens of this global epidemic and its metabolic consequences.
We believe CB1 inhibition shows promise for treatment of obesity-related atherosclerotic disease, but its benefits will need to be confirmed in additional trials, which are currently underway.
Let me then offer some final thoughts. Development of effective and durable treatment strategies for management of obesity has proven a daunting challenge.
New approaches are greatly needed to reduce the burdens of this global epidemic and its metabolic consequences.
We believe that CB1 inhibition shows promise for treatment of obesity-related atherosclerotic disease, but its benefits will need to be confirmed in additional trials, which are currently underway.
Thank you for your attention.

20. Back Up Slides

21. Changes: Lab Values and Obesity Measures
Placebo(n=341)
Rimonabant(n=335)
P value
Body Weight
-0.5 kg
-4.3 kg
<0.001
Waist Circumference
-1.0 cm
-4.5 cm
LDL-cholesterol
1.7%
0.4%
0.78
HDL-cholesterol
6.9%
22.4%
Triglycerides
-6.2%
-20.5%
hsCRP
-30.9%
-50.3%
HbA1c
0.40%
0.11%

22. Changes in HbA1c: All Patients and Diabetics
All Patients (N = 676)
Diabetic Patients (N= 248)
0.40%
0.42%
P <0.001
P <0.001
Change in HbA1c (%)
Change in HbA1c (%)
0.11%
-0.13%
Months after Randomization

23. Time to Permanent Drug Discontinuation
HR = 2.75
Log rank P < 0.001
Placebo
Rimonabant