Shelley Wood

September 25, 2013

BARCELONA, SPAIN — A single-center, randomized trial testing metformin on top of statins in nondiabetic subjects in the hopes of slowing or halting the progression of cardiovascular disease has come up empty-handed.

But Dr David Preiss (University of Glasgow, Scotland), presenting the Carotid Atherosclerosis: Metformin For Insulin Resistance (CAMERA) trial results here yesterday, said that "impressive" weight loss seen in the metformin group—one of the trial's secondary end points—is worth exploring further.

CAMERA randomized 173 subjects with established coronary disease and a large waist circumference, but no diabetes, to either metformin 850 mg twice daily or placebo.

Dr David Preiss

As Preiss explained to heartwire , the rationale for the study came from previous observations, including the UKPDS study, that metformin reduced not only diabetes end points, but also MI. Moreover, the CVD benefits appeared to be unrelated to glycemic improvement.

In CAMERA, investigators saw glycemia levels drop over the trial's 18-month follow-up in the metformin-treated group, as expected. But for the primary end point of carotid intima-media thickness (CIMT), Preiss and colleagues saw absolutely no differences between groups. Carotid plaque score, the main secondary outcome, was also no different between groups.

In retrospect, Preiss commented to heartwire , the choice of CIMT for the primary end point, as a surrogate for CVD, was not ideal.

"I'm not a diabetologist, but if any group of physicians should be cautious about a study with surrogate markers, you'd think it would be these guys," Preiss mused.

I'm not a diabetologist, but if any group of physicians should be cautious about study with surrogate markers, you'd think it would be these guys.

Other surrogate measures of CVD also showed no benefit of metformin, including total cholesterol, HDL, and non-HDL. Very "marginal" differences in triglycerides and high-sensitivity C-reactive protein (hs-CRP) favored the metformin group, but these were "not impressive," Preiss commented.

The story for metabolic end points was slightly different, however. CAMERA did show "quite an impressive fall in weight," Preiss said—a difference of 3.2 kg between groups. There were also significantly greater declines in body-mass index (BMI) in the metformin group.

To heartwire , Preiss agreed the CIMT results were disappointing, but given the lack of differences in other CVD markers, a change in CIMT in the absence of cholesterol improvement "would have been a big surprise, and hard to explain."

The question of whether the weight change and potentially the signal of anti-inflammatory benefits with metformin will translate into a reduction in hard cardiovascular disease end points may be answered by forthcoming trials.

The GLINT study, now under way, is a multicenter CV primary-prevention trial with a planned enrollment of 12 000 nondiabetic subjects. Participants will have to have a predicted 10-year risk of CVD of 20% (by Framingham) and a high-normal HbA1c; they are being randomized to either metformin or placebo, with a follow-up of five years.

Two smaller studies may also provide some insights: REMOVAL is a 500-patient, multinational study looking at add-on metformin in type 1 diabetes, with an end point of change in CIMT. Another is GIPS III , a Dutch study of 380 post-STEMI patients, looking at the ability of metformin to reduce progression to heart failure.

So more answers on metformin in CVD are coming, Preiss said. "You wouldn't dismiss its chances [of succeeding] based on our study alone," he said.


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.