The PROactive Study—The Glass Is Half Full

Abstract

Type 2 diabetes is associated with an increased risk of cardiovascular events that carry a poor prognosis. People with diabetes have not benefited to the same degree as those without diabetes in terms of recent reductions in cardiovascular mortality. Obesity-related insulin resistance, which is accompanied by a cluster of cardiovascular risk factors, is clearly associated with cardiovascular disease and accounts for much of this increased risk of macrovascular disease in people with type 2 diabetes (1). Furthermore, recent data suggest that insulin resistance is associated with a several novel factors that have been implicated in the pathogenesis of cardiovascular disease, including inflammation and endothelial dysfunction (2). A variety of therapeutic interventions have been developed to treat insulin resistance. The most successful of these has been the peroxisome proliferator-activated receptor (PPAR) agonists such as the thiazolidinediones, also known as glitazones. Although these drugs are only approved for the treatment of diabetes, they have attracted much attention because of their potential for reducing cardiovascular events. They improve insulin sensitivity and directly or indirectly improve a wide array of cardiovascular risk factors, many of which may be pathogenic in inducing cardiovascular events (2). Although the glitazones improve a number of cardiovascular risk factors, much of these data come from animal studies and small human studies. Ultimately, therefore, the promise of any drug to alleviate human disease must be tested in large outcome-based clinical trials. Several such trials have been initiated, using PPAR agonists in a variety of settings from prediabetes to late-stage diabetes, to examine their role in primary as well as secondary prevention of cardiovascular events (3). The first of these was the Prospective Pioglitazone Clinical Trial in Macrovascular Events Study (PROactive), which recently reported its results and is therefore being viewed with considerable interest (4). PROactive was a prospective, randomized, controlled trial studying 5238 patients with type 2 diabetes who had evidence of cardiovascular disease. These patients were randomized to pioglitazone, titrated to 45 mg daily, or matching placebo. Importantly, the study drug was taken in addition to the patient’s usual glucose-lowering medications, and therefore this study was designed to assess the pure effect of pioglitazone, independent from any of its effects on lowering blood glucose. The results of the study show that pioglitazone had only a modest, and not statistically significant, 10% reduction in the risk of the primary composite endpoint, which consisted of all-cause mortality, nonfatal myocardial infarction, stroke, acute coronary syndrome, and revascularization or amputation. However, the “main secondary endpoint,” consisting only of certain of the primary outcome measures, namely all-cause mortality, myocardial infarction, and stroke, was significantly reduced by 16%. Despite the negative result on the primary endpoint, the primary publication states in its conclusion and the paper subtitle that “pioglitazone reduces mortality, myocardial infarction, and stroke.” Such a conclusion is of course noteworthy for a population at high risk for these complications. However, the publication and presentation of the data have also been received with considerable skepticism of the interpretation, as well as condemnation of the methodology used, particularly the statistical analysis (5, 6). It is important for the practicing clinician to put these into perspective so that appropriate, clinically relevant conclusions can be drawn. The study had major limitations that are likely to impede rapid acceptance in clinical practice. First, it was carried out in a population of mainly Caucasian subjects who were possibly less insulin resistant than the population of the United States and other countries, and who had a relatively high rate of cigarette smoking and a relatively low rate of statin use for patients who had already established macrovascular disease. It is not surprising, therefore, that the event rates were higher than expected, which, coupled with a remarkably higher than expected rate of recruitment, led to the study being concluded somewhat earlier than previously anticipated. A glance at the figures in the primary publication suggest that the Kaplan-Meier curves of the time to primary endpoint only start separating after 18 months, and the 36 months of the trial might not have been adequate to detect a true difference. This is also understandable, given the well-known slow onset of action of this class of drugs, which may be related to their primary mechanism of action. The design of the study was previously published (7), but the “main secondary endpoint” was not clearly stated at that First Published Online November 22, 2005 Abbreviations: CHF, Congestive heart failure; PPAR, peroxisome proliferator-activated receptor.