The future of endocannabinoid-oriented clinical research after CB₁ antagonists

Abstract

Great interest has been shown by the medical community and the public in the cannabinoid CB₁ receptor antagonists, such as rimonabant, for treatment of obesity, metabolic syndrome, and possibly drug addiction. This novel class of drug has therapeutic potential for other disorders, as the endocannabinoid system is involved in various health conditions. However, rimonabant, the first clinically available member of this class of drugs, has been linked to increased risk of anxiety, depression, and suicidality. Due to those risks, the European Medicines Agency (EMEA) called for its withdrawal from the market in October, 2008. Shortly after this decision, several pharmaceutical companies (Sanofi-aventis, Merck, Pfizer, Solvay) announced they would stop further clinical research on this class of drug. Here, we provide an overview of those events and make several suggestions for continuing such clinical research, while safeguarding the safety of patients and clinical trial subjects.

Obesity and tobacco use are the two highest preventable causes of morbidity and mortality in the developed world (Flegal et al. 2005; Hossain et al. 2007; Ogden et al. 2006; U.S. Department of Health and Human Services 2004). Since blocking the cannabinoid CB₁ receptor was shown to decrease eating and nicotine self-administration in animals (Le Foll et al. 2008; Pacher et al. 2006), there has been a great deal of interest in this novel class of drug. More broadly, the endogenous cannabinoid system is implicated in a substantial number of severe and frequent clinical disorders (Table 1) (Pacher et al. 2006). Therefore, it is possible that blocking endocannabinoid transmission could have benefits in various fields of medicine.

Table 1.

Possible therapeutic indications for cannabinoid CB₁ antagonists/inverse agonists

- Obesity (Christensen et al. 2007; Van Gaal et al. 2008) - Dyslipidemia (Despres et al. 2008; Van Gaal et al. 2008) - Metabolic syndrome (Despres et al. 2008; Van Gaal et al. 2008) - Diabetes (Rosenstock et al. 2008) - Coronary artery disease (Nissen et al. 2008; Sugamura et al. 2009) - Tobacco dependence (Cahill and Ussher 2007; Le Foll et al. 2008) - Other drug dependence (cannabis, alcohol, opiates, psychostimulants) (Le Foll and Goldberg 2005; Soyka et al. 2008; Wiskerke et al. 2008) - Hypotension/shock (Batkai et al. 2004) - Liver disease (Izzo and Camilleri 2008) - Gastro-intestinal disease (Izzo and Camilleri 2008) - Reproductive health (e.g., miscarriage) (Habayeb et al. 2008) - Arthritis (Richardson et al. 2008)

As one example, CB₁ receptor antagonists show promise in treating drug addiction (De Vries and Schoffelmeer 2005; Le Foll et al. 2008; Le Foll and Goldberg 2005; Wiskerke et al. 2008), for which there is a need for more effective treatments (Le Foll and George 2007; Reuter and Pollack 2006). Blocking CB₁ receptors reduces motivation for Δ⁹-tetrahydrocannabinol, the active ingredient of cannabis, in a non-human primate model (Justinova et al. 2008; Tanda et al. 2000) and partially blocks the psychological and cardiovascular effects of smoking a cannabis cigarette in humans (Huestis et al. 2007). Clinically validating these promises would be of value to the field of drug addiction.

Rimonabant (Acomplia,^® Sanofi-aventis) was the first selective CB₁ ligand introduced into clinical practice. This CB₁ receptor antagonist (with inverse agonist action) has been shown efficacious as a treatment for obesity (Despres et al. 2008; Hampp et al. 2008) and for improving dyslipidemias, diabetes, and metabolic syndrome (Despres et al. 2005; Scheen 2008; Van Gaal et al. 2005). It had been approved as an obesity treatment in more than 50 countries worldwide, including the European Union (EU). Rimonabant was also being developed for smoking cessation, with statistically significant, albeit modest, evidence for efficacy (Cahill and Ussher 2007; Le Foll et al. 2008; Rigotti et al. 2009), especially in minimizing weight gain often associated with smoking cessation.

Tempering this promise has been a growing concern about the psychiatric safety of CB₁ receptor antagonists (Christensen et al. 2007; Food and Drug Administration Endocrinologic and Metabolic Advisory June 13, 2007; Rucker et al. 2007), notably increased rates of depression, anxiety and suicidality related to drug use. These psychiatric concerns led to the October, 2008 decision by the European Medicines Agency (EMEA) to suspend marketing of rimonabant in the EU. Following this decision, the drug-maker Sanofi-aventis announced on November 5^th 2008 its decision to withdraw rimonabant from the market worldwide and to discontinue its ongoing rimonabant clinical development program for all indications (Sanofi Aventis 2008). Around the same time, several other CB₁ receptor antagonists/inverse agonists not yet approved for marketing were withdrawn from clinical development by their developers, including taranabant (Merck) and otenabant (Pfizer), both in phase 3, and ibipinabant (Solvay/Bristol-Myers Squibb) and surinabant (Sanofi-aventis) in phase 2 (Jones 2008; Merck 2008)

It should be noted that rimonabant was eventually withdrawn from marketing in the EU because its clinical use was not following the criteria established to maximize its benefit/risk ratio, i.e., community clinicians were prescribing it to patients at high risk for depression and suicidality and many patients were not taking it long enough to achieve the potential benefits. We agree that the public should be protected from medications with unfavorable risk-benefit ratios. However, the unfavorable conditions that existed with rimonabant prescriptions in the community do not exist in the clinical research environment, where subjects are rigorously screened for eligibility and drug administration must adhere to the protocol schedule. We consider it premature to stop further clinical research on cannabinoid antagonists in important human diseases as a result of the termination of clinical development of currently available CB₁ receptor antagonists/inverse agonists. It is possible that several conditions which currently have no treatment may benefit from the use of such a class of drugs (Table 1). Cannabinoid CB₁ receptor antagonists/inverse agonists and other ligands modulating endocannabinoid transmission may have a very positive benefit-risk ratio for other indications. We therefore make several suggestions for continuing such important clinical research, while safeguarding the safety of patients and clinical trial subjects.

First, we suggest that rimonabant and similar compounds remain available for clinical research under strictly controlled circumstances that maximize safety, such as single-dose studies, inpatient studies, or short-term outpatient studies with rigorously screened subjects under close monitoring. This would allow researchers to perform ‘proof of principle’ or ‘proof of concept’ clinical trials. In this regard, full and transparent disclosure of all safety data from clinical trials would advance our understanding of the risks associated with this class of compounds and hopefully improve the design of future clinical trials.

Second, we suggest continued development of more selective CB₁ receptor ligands. It is possible that the observed psychiatric adverse events were due, at least in part, either to inverse agonism properties or to excessive blockade of CB₁ receptors. Both rimonabant and taranabant showed evidence of inverse agonist activity in preclinical studies. It is possible that “cleaner” compounds that are neutral (so-called “pure”) CB₁ antagonists or inverse agonists with little antagonism may have a more favorable pharmacological profile. For example, in animal studies, both CB₁ receptor neutral antagonists and inverse agonist/antagonists reduce food intake and food-reinforced behavior, while only the former cause nausea and vomiting (Salamone et al. 2007). Whether this dissociation between desired effects and side-effects holds for human psychiatric indications remains to be evaluated.

Another approach is development of CB₁ receptor antagonists/inverse agonists that do not act in the central nervous system, and so presumably would be devoid of psychiatric side-effects. Although such novel ligands might be ineffective for CNS-mediated disorders such as drug addiction, there is evidence that a peripheral antagonist can reduce feeding and weight gain in animals (Pavon et al. 2008). Other indications, such as coronary artery disease (Nissen et al. 2008; Sugamura et al. 2009), liver and pancreatic disease (Izzo and Camilleri 2008), inflammatory bowel disorders (Izzo and Camilleri 2008), and arthritis (Richardson et al. 2008) may also benefit from peripheral CB₁ antagonism.

Third, we suggest continued research on pharmacological approaches that bypass the CB₁ receptor and modulate the endocannabinoid system by different means. Targeting the enzyme fatty acid amide hydrolase (FAAH) (Gobbi et al. 2005; Kathuria et al. 2003; Scherma et al. 2008a), which breaks down endocannabinoids, or the endocannabinoid membrane transporter system (Beltramo et al. 1997), may have therapeutic utility, notably for nicotine addiction (Scherma et al. 2008b), while avoiding psychiatric side-effects

In conclusion, there is increasing evidence that the endocannabinoid system plays a significant role in various human illnesses. It is important to continue clinical research in this area and to evaluate promising preclinical findings. The recent termination of clinical development of one class of compounds should not deter continued work in this field, which holds the promise for meaningful clinical benefits.