March 12th, 2014
Clinical Trials For Diabetes Drugs Need To Consider Heart Failure
Is heart failure the missing 800 pound gorilla in diabetes trials? That’s the argument proposed by a group of prominent cardiovascular and diabetes researchers.
It was long believed that by virtue of their glucose-lowering properties diabetes drugs would confer substantial cardiovascular benefits. That belief is no longer widely held, however, and the FDA now requires cardiovascular outcome trials for new diabetes drugs. But, write the researchers in an article published in the Lancet Diabetes & Endocrinology, these trials are failing to track and analyze one key cardiovascular endpoint, thereby diminishing the value of these trials in assessing the cardiovascular effects of diabetes drugs.
John McMurray, Hertzel Gerstein, Rury Holman, and Marc Pfeffer write that heart failure is “a cardiovascular outcome in diabetes that can no longer be ignored.” In response to the firestorm over rosiglitazone (Avandia) the FDA and the EMA have insisted that cardiovascular outcomes trials be performed with new diabetes drugs. But for the most part these trials, which now include about 150,000 patients, have been designed to track a traditional MACE (major adverse cardiovascular events) endpoint, which includes cardiovascular death, MI, and stroke, but which does not include heart failure.
This omission could have serious repercussions, they write, because the available evidence from epidemiological studies and clinical trials indicates that heart failure occurs quite frequently in patients with diabetes (though they acknowledge that it is unclear whether there is a precise causal relationship). Further, the occurrence of heart failure is clearly related to a poor prognosis. It was once believed that the heart failure symptoms associated with one class of diabetes drugs– the thiazolidinediones (which includes rosiglitazone)– was an artifact that did not have the same prognostic significance as heart failure in other patients. But, McMurray and colleagues write, evidence from the RECORD trial “suggests otherwise.” Equally ominous is that recent trials of a newer class of diabetes drugs, the DPP-4 inhibitors, appeared to have resulted in a clear heart failure signal.
Summarizing their position, they write that “hospital admission for heart failure is one of the most common and prognostically important cardiovascular complications of diabetes, and the one cardiovascular outcome for which the risk has been shown unequivocally to be increased by certain glucose-lowering therapies.” They propose that heart failure should be “systematically evaluated” in these outcome trials.
Sanjay Kaul provided the following comment about the Lancet paper:
I agree with the authors that the evidence linking some, but not all, antidiabetic drugs to increased risk of heart failure is stronger than the evidence linking to increased risk of atherothrombotic MACE events. However, the underlying mechanisms remain unclear. Nonetheless, the authors make a defensible case for including heart failure as an outcome of special interest in future diabetes trials. Rather than combining it with MACE as a primary composite outcome, I prefer heart failure to be included as a separate coprimary endpoint, or a key secondary endpoint as it might not lie on the same pathophysiologic pathway as MACE (primarily an atherothrombotic event), and the treatment might potentially yield divergent effects on these endpoints, thereby challenging the interpretation of the composite endpoint.