Designing the Next Generation of Diabetes Drug Trials

Lawrence A. Leiter, MD; Mark Harmel, MPH


September 29, 2017

During the past few years, a large number of so-called diabetes cardiovascular outcomes trials or safety trials have been completed.[1] But these trials also have given rise to several questions. Have they been worth the cost and effort? Could we do things better?

What have we learned from these trials? We now know about the cardiovascular safety of several newer antihyperglycemic agents, and in some cases, specifically with empagliflozin, liraglutide, and semaglutide, we know that there is some cardiovascular benefit. We have also discovered a few unexpected adverse events: perhaps an elevated risk for heart failure with saxagliptin and alogliptin, perhaps a very small increase in risk for pancreatitis with the dipeptidyl peptidase-4 (DPP-4) inhibitors.

Without a doubt this has been informative.

At the same time, these drug trials are expensive. Each has cost several hundred million dollars. People have questioned whether the money could have been better spent [elsewhere].

I believe that the information we have learned has certainly been clinically relevant. It has changed our clinical practice. Therefore, I would say that the money has been well spent. Obviously, however, the cost of these trials eventually is incorporated into the cost of the medications.

Can we do things better going forward? Absolutely. The completed trials to date have focused on high-risk patients, primarily patients with known cardiovascular disease, and the trials have lasted only about 1.5 to 3 years.

One of my colleagues has said that we have "crash tested" these drugs. We have shown that they are safe in our sickest patients, and therefore we are quite comfortable extrapolating the safety to our less sick patients. At the same time, some people have argued that these trials may be limited because they have only focused on patients with advanced disease. What about the 80% of patients with diabetes who do not have known cardiovascular disease?

Moving forward, we need longer trials that incorporate a greater proportion of patients without known cardiovascular disease. In other words, we need study populations that more accurately reflect the patients we see in our daily practice.

These trials to date have also been placebo controlled, and as we develop more and more efficacy data, it is perhaps inappropriate to compare a new drug against placebo. I would like to see comparisons of new drugs versus another drug with a known cardiovascular safety profile.

People have also questioned the typical major adverse cardiovascular event (MACE) primary endpoints—cardiovascular mortality, nonfatal myocardial infarction, or nonfatal stroke—because in the completed trials, the components of the primary endpoint have not necessarily all been affected positively or negatively. Sometimes they diverge. There is also the suggestion that we should be including other relevant endpoints, such as heart failure and renal outcomes.

Last, we certainly must incorporate to a greater degree patient-reported outcomes, as well as outcomes that are important not only to the regulators and clinicians, but to the ultimate consumer of the medications: our patients.

Yes, we have learned a lot over the past few years. We will learn a lot more over the next few years. But in our next wave of study designs, we must be more creative and think of innovative ways of conducting longer, less expensive trials with lower-risk patients.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: