Play an ADAGIO with a STRADIVARIUS: the right patient for CB1 receptor antagonists?

Vincenzo Di Marzo  About the author

Accumulation of large quantities of intra-abdominal visceral fat usually accompanies overt obesity (BMI 30 kg/m²), but is also found in overweight individuals (BMI 25–29.9 kg/m²). Visceral or intra-abdominal adiposity is increasingly recognized as one of the major risk factors for type 2 diabetes, atherosclerotic plaque formation, and associated cardiovascular events.¹ A strong phenotypic association exists between abdominal obesity—which can be easily assessed in clinical practice by measurements of waist circumference or waist-to-hip ratio—and two other major cardiovascular risk factors: high triacylglycerol and low HDL-cholesterol levels.¹ Accumulation of visceral fat is accompanied by chronic inflammation, which can trigger atherosclerosis and insulin resistance. In abdominally obese individuals, these events are partly caused by low levels of adipocyte-derived adiponectin, which has a key role in insulin sensitization of the skeletal muscle.^{2, 3} Adiponectin also counteracts the release of inflammatory cytokines from adipose tissue and thus decreases their inhibitory effects on insulin sensitivity and free fatty acid oxidation.^{2, 3} These cytokines, in turn, contribute to atherosclerosis via enhanced endothelial cell activation, transendothelial migration of monocytes and endothelial adhesion of monocytes and neutrophils. Inflammatory cytokines also decrease the capability of adipocytes to store triacylglycerols,⁴ which leads to hypertriglyceridemia and ectopic fat formation in the liver, as well as to dyslipoproteinemia.

Evidence is accumulating that dysregulation of the endocannabinoid system has a role in the development of intra-abdominal obesity and subsequent lipotoxicity. In healthy adipocytes, endocannabinoids act at cannabinoid CB1 receptors to induce fatty-acid synthesis and triacylglycerol accumulation in several different ways, and are under negative-feedback control of insulin, leptin, and peroxisome proliferator-activated receptor .⁵ In obese rodents and humans, however, the endocannabinoid system becomes more active in visceral adipose depots and less active in peripheral subcutaneous adipose depots, which possibly contributes to the eventual accumulation of intra-abdominal fat at the expense of 'healthier' subcutaneous fat.^{5, 6, 7} Furthermore, activation of CB1 receptors in mature adipocytes decreases adiponectin expression.⁶ High plasma levels of endocannabinoid 2-arachidonoylglycerol in obese patients are less strongly associated with high BMI than with increased intra-abdominal obesity, high triacylglycerol levels, low HDL cholesterol, and insulin resistance.^{8, 9} The pathologic role of increased CB1 receptor activation in the visceral adipose tissue of patients with intra-abdominal obesity is suggested by the observation that chronic administration of a CB1 receptor antagonist or inverse agonist to rodents and obese individuals reduces body weight independently of these agents' anorectic effects, and significantly ameliorates the altered metabolic and lipoprotein profiles that accompany obesity.^{5, 10} One such compound, rimonabant, was developed by Sanofi-Aventis. After having completed four phase III studies—the Rimonabant In Obesity (RIO) trials¹¹—this agent is currently marketed in more than 50 countries, as an adjunct to increased physical activity and exercise, to aid weight reduction in obese or overweight individuals with metabolic complications such as type 2 diabetes and dyslipidemia.

The efficacy of this pharmacological approach toward simultaneous targeting of several cardiometabolic risk factors raised initial excitement. Enthusiasm for CB1 receptor antagonism has, however, lately been attenuated by concerns about its psychiatric side effects (primarily anxiety and depression), which led an FDA board of experts to advise against marketing of rimonabant in the US.^{10, 11} Furthermore, until March 2008, little evidence existed that chronic attenuation of CB1 activity could preferentially reduce the accumulation of intra-abdominal fat and counteract atherosclerosis. Two new clinical trials of rimonabant now seem to provide answers to these controversial issues, and have fuelled renewed hope for the clinical use of CB1 receptor antagonists or inverse agonists.

Data from STRADIVARIUS¹² were communicated at the ACC meeting in March 2008 and published concurrently in the Journal of the American Medical Association. Nissen and colleagues carried out a randomized, double-blind, placebo-controlled, parallel-group trial that compared rimonabant with placebo in 839 patients at 112 centers in North America, Europe, and Australia. Unlike previous studies with rimonabant, patients receiving treatment for depression were not excluded from the trial. Patients were randomly assigned to receive rimonabant (20 mg daily) or placebo, and underwent coronary intravascular ultrasonography at baseline (n = 839) and at study completion (n = 676). The primary and secondary efficacy parameters were change in percent atheroma volume (PAV) and change in normalized total atheroma volume, respectively. Rimonabant exerted its established¹¹ beneficial effects on body weight, waist circumference, and levels of triacylglycerols, HDL cholesterol, glycated hemoglobin, and C-reactive protein. In addition, rimonabant slowed the progression of atherosclerosis (PAV increased by 0.25% in rimonabant-treated patients versus 0.57% in placebo-treated patients, and total atheroma volume fell by 1.95 mm³ in the rimonabant group and increased by 1.19 mm³ in the placebo group). Only rimonabant's effect on total atheroma volume was statistically significant. Importantly, in the subgroup analyses, rimonabant also significantly reduced PAV in patients who were not taking statins (by 1.31% compared with placebo) and in those with high baseline triglyceride levels (140 mg/ml), among whom a decrease in PAV of 0.77% versus placebo was observed. Also notable was the observation that psychiatric adverse effects that led to discontinuation of therapy were more frequent than in previous studies, and more common in rimonabant-treated than placebo-treated patients (9.5% versus 3.1%).¹²

At a late-breaking presentation during the 77^th meeting of the European Atherosclerosis Society, the investigators of the ADAGIO-Lipids study presented key findings of a 1-year trial that examined the effects of rimonabant on lipoprotein size and other cardiometabolic risk factors.¹³ The study was conducted in 799 patients with severe abdominal obesity, the typical high triacylglycerol–low HDL-cholesterol lipid profile, and no history of depression. Patients were randomly allocated to receive either rimonabant 20 mg daily or placebo, and followed a diet that reduced their daily calorie intake by 600 kcal. Participants were followed up for 12 months. Importantly, a CT substudy was conducted to test, for the first time, the hypothesis that rimonabant could induce a preferential loss of visceral fat over subcutaneous fat, and also decrease liver fat. The results again confirmed the consistent effects of rimonabant on several markers of cardiometabolic risk. For example, in rimonabant-treated patients, HDL-cholesterol levels increased by 7.4% (P <0.0001) and triglyceride levels were reduced by 18.0% (P <0.0001) compared with placebo-treated patients. The drug did not modify LDL-cholesterol levels, but induced a major shift in the distribution of the size of LDL particles, with a substantial reduction in the proportion of small, atherogenic LDL particles (-6.5% versus placebo, P <0.0001) and a concomitant increase in the concentration of large LDL particles (+4.8% versus placebo). Markers of HDL concentration and quality also improved with rimonabant, including an increase in apolipoprotein AI (+3.2% versus placebo, P = 0.02) and HDL particle size (+0.9% versus placebo, P <0.001). These effects could have been predicted from preclinical studies in obese mice, in which overactivity of CB1 receptors was described in the liver, leading to increased hepatic synthesis of free fatty acids and triacylglycerols, and hence to potential modifications of lipoprotein size. Inflammation was also ameliorated, as revealed by a 17% reduction in C-reactive protein levels compared with placebo (P <0.01), and a highly significant increase in the blood concentration of adiponectin (+18.9% versus placebo, P <0.0001). Rimonabant reduced participants' blood pressure by 3.3 mmHg more than placebo, an effect that has not always been observed in previous studies of this drug.¹¹ In the CT analysis, rimonabant reduced visceral adipose tissue (-10.1% versus placebo, P <0.0005) to a greater extent than it did subcutaneous fat (-5.1% versus placebo, P <0.005). Shown for the first time in humans, rimonabant significantly decreased fat accumulation in the liver, which could have been expected from the observed changes in lipoprotein size distribution and from preclinical studies.¹⁴ Finally, the reduction in body weight in rimonabant-treated patients was less than has been observed in previous RIO trials,¹¹ which might account for the authors' estimate that 60–70% of rimonabant's effects on triacylglycerol, adiponectin, and HDL levels were not attributable to weight loss. On the other hand, in ADAGIO-Lipids, the incidence of psychiatric side-effects that led to discontinuation was essentially similar to (if not less than) that previously observed in the RIO studies.¹¹

What are the key messages of these new studies? First, obese patients with marked intra-abdominal obesity, characterized by high triacylglycerol levels, low HDL-cholesterol levels and low insulin sensitivity, seem to benefit most from chronic treatment with rimonabant—these are the individuals in whom the greatest reduction in cardiometabolic risk factors, and eventually atherosclerosis, is likely to be observed. Second, patients with no history of depression can be treated with CB1 receptor antagonists or inverse agonists more safely than those who have experienced this disorder prior to treatment. These compounds are more likely to worsen pre-existing psychiatric disorders than to cause them de novo.¹⁵ This finding is consistent with the 2-year results of the RIO-Europe study,¹⁶ which showed that no increase in psychiatric adverse events over placebo treatment was observed with 20 mg daily rimonabant in the second year of the trial, and that a further dissociation between psychiatric adverse effects and the beneficial effects of rimonabant on waist circumference, HDL cholesterol and triacylglycerol levels might be obtained by using daily doses lower than 20 mg. Indeed, given the association between depression and extreme obesity,¹⁰ my opinion is that treatment with this new class of drugs might be safer and more valuable in patients with a BMI of 27–33 kg/m², dyslipidemia, prediabetes, and a large waist circumference, than in very obese individuals. Ongoing trials aim to assess the efficacy of prolonged (2 years) treatment with rimonabant at reducing carotid intima thickness (the AUDITOR trial), the likelihood of developing type 2 diabetes (the RAPSODI trial), and the long-term risk of cardiovascular events (the CRESCENDO trial). Together with studies of other potential CB1 antagonists or inverse agonists,^{5, 15} these studies will elucidate whether selection of the right patient groups will improve the risk:benefit ratio afforded by this new drug class, as STRADIVARIUS and ADAGIO seem to suggest.

Competing interests

V Di Marzo has received research support and Speakers' Bureau honoraria from Sanofi-Aventis.

ARTICLE NAVIGATION -ARTICLE TOOLS

Full text
• Download PDF
• Send to a friend
• Export citation
• CrossRef lists 5 articles citing this article
• Rights and permissions
• Order commercial reprints

ISSUE NAVIGATION - THIS ISSUE

• Table of Contents
• Previous article
• Next article