Cardiovascular Risk Assessment in the Development of New Drugs for Obesity

Abstract

THE US FOOD AND DRUG ADMINISTRATION (FDA), CENter for Drug Evaluation and Research, convened a meeting of the Division of Metabolism and Endocrinology Products on March 28-29, 2012, to consider cardiovascular safety requirements for approval of new drugs to treat obesity. The broader context at that time involved 4 recent individual drug product reviews, resulting in no new antiobesity drugs approved and 1 drug (sibutramine) withdrawn from the market. For pharmacologic weight loss to be more than cosmetic, it is essential to demonstrate measurable health benefits. Because obesity is associated with excess cardiovascular risk and weight loss treatments could potentially mitigate that risk, it is important to assess the effect of these drugs on cardiovascular morbidity and mortality. Stimulation of sympathetic activity is the principal mechanismofactionofseveralantiobesitydrugsapprovedbytheFDA during the past 50 years. These drugs have the potential to increase blood pressure, heart rate, or both, thereby contributing directly to increased cardiovascular risk, possibly offsetting the clinical benefits of weight reduction. For example, a postmarketingcardiovascularoutcomestrial(CVOT)ofsibutramine,asympatheticagentthatinhibitsserotonin-norepinephrine reuptake,demonstratedanincreasedriskofmajoradversecardiac events (MACEs), leading to market withdrawal in 2010. This along with the previously documented adverse cardiac effects of valvulopathy and pulmonary hypertension associatedwithfenfluramine(anothersympathomimeticagent)have raised concerns about the cardiovascular safety of all new antiobesity drugs. Evaluating the cardiovascular effects of these drugs, particularly in the population with underlying cardiovascular disease, is a current regulatory focus that has broad implications for future antiobesity drug development. Weight loss of approximately 5% achieved through lifestyle modification is clinically meaningful and, accordingly, is the recommended milestone for drug development. Recent pharmacologic candidate agents have relatively modest effects on weight loss at 1 year. Mean placebosubtracted weight loss ranges from 3% for lorcaserin to 9% with phentermine/topiramate (TABLE). The degree of weight loss early in treatment may identify subpopulations likely to sustain clinically meaningful long-term weight reductions. Combination drug products appear to be associated with greater magnitude weight loss than a single agent, but at a potential expense of reduced safety and tolerability. Cardiovascular safety informationaboutnewdrugshaspreviously been limited because the premarketing trials have included relatively healthy individuals (mostly women) at low cardiovascular risk (event rates of ~0.5% annually). Consequently, theactualnumberof cardiovascular events is lowand the attribution uncertain due to lack of prospective adjudication.Forexample,thenaltrexone/bupropionpremarketingtrials accumulatedonly4MACEs(1strokewithplaceboand3myocardial infarctions with drug), phentermine/topiramate providedonly12events,andlorcaserinhadonly11events(Table). At the March 2012 meeting, the FDA advisory committee supported several strategies to assess cardiovascular risk. The principal recommendation was to rule out a prespecified unacceptable level of cardiovascular risk similar to the FDA draft guidance for new oral agents to treat type 2 diabetes. This “exclusion of excess risk” approach requires the upper bound of 95% CI of the risk difference to be no worse than what is clinically acceptable as defined by the overall benefit-risk balance of the drug: tolerating a greater degree of risk for greater weight loss and ancillary benefits in cardiometabolic profile. Excluding risk is dependent on acquiring a sufficient number of events. For example, if the population MACE rate is 0.5 events per 100 patient-years and the upper bound of the 95% CI to be excluded is a risk ratio of 2.0 (a more liberal margin than required for diabetes drugs), 87 total events would be required. Given this event rate, a CVOT would need an estimated 17 400 patient-years of exposure. However, if the study population event rate were increased to 1.5 events per 100 patient-years, the number of total patient-years of exposure would be reduced to an estimated 5800. Several strategiesweresuggested to increase theMACErate. Thesecouldincludepoolingthesafetydatabasesacrossallphase 2 and 3 randomized trials and performing a meta-analysis of all cardiovascular events with prospective adjudication by an independent,blindedclinicaleventscommittee.Enrichingstudy populations with individuals at higher risk would require enrollment of obese patients with diabetes, a prior history of cardiovasculardisease,or thosewithmultiplecardiovascular risk factors. This strategy was used for sibutramine to obtain a placebo MACE annual rate of 2.9%. One potential limitation of thisapproachisthegeneralizabilityofevaluatingsafetyinahighrisk population who may be less tolerant of potential adverse eventswhentheprimary intendeduse is in lower-riskpatients.