March 14th, 2010

Does Fenofibrate ACCORD with the Treatment of Diabetes?

and

CardioExchange welcomes Dr. Henry N. Ginsberg to discuss the findings of the ACCORD Lipid Study Group. In this paper, researchers randomized 5,500 diabetic patients to fenofibrate therapy versus placebo. All patients received simvastatin. Although triglyceride levels improved markedly with fenofibrate, the incidence of cardiovascular events was not affected. 

CardioExchange Editors: Should we abandon fenofibrate for treating patients with diabetes?

I believe that fenofibrate should be used when a patient has significant dyslipidemia (high triglycerides and low HDL-C) after statin therapy. My cutpoints for using fenofibrate in combination with a statin are based on our study (which were admittedly arbitrary, given they are tertile cutpoints), the FIELD study (similar cutpoints), and the BIP study (a cutpoint greater than 200 mg/dL). Therefore, my cutpoints are TG greater than 200 mg/dL — this is always associated with an HDL less than 40 mg/dL in a man and less than 50 mg/dL in a woman. If a patient has triglyceride levels lower than 200 mg/dL, and a very low HDL, then I am less sure about the approach — fenofibrate vs. niacin.  But in a diabetic, fenofibrate would be my first try.

CardioExchange Editors: Should we extrapolate these findings to all fibrates?

Based on published evidence, one might say that gemfibrozil was a better fibrate — however, its effects on lipids, particularly in diabetics in the VA-HIT study were not much different than the fenofibrate effects in the present study and the VA-HIT population had a much lower HDL-C at baseline. I am not sure the lipid changes actually explain any benefit seen with the fibrates. But remember, the combination of gemfibrozil and a statin has a much greater risk (albeit still small) for myositis than the combination of fenofibrate and and a statin (no greater than statin alone).

CardioExchange Editors: In your study, Mean HDL levels increased in the fenofibrate group from 38 mg/dL to 41.2 mg/dL. Why do you think that the drug did not have a greater effect on HDL? 

I don’t have a good answer – the literature on fibrates and lipids is skewed by recruitment of severely dyslipidemic subjects and relatively short studies. In VA-HIT, the diabetics had no more of an HDL increase than in ACCORD. In our dyslipidemic subgroup, the HDL increase was larger than in the overall group.

CardioExchange Editors: In a prespecified subgroup analysis, women in your study had higher rates of adverse events with fenofibrate than with placebo. What do you make of this gender interaction?

On further analyses (not done yet), I think that is will be related to a greater proportion of women without prior CVD and with lower triglyceride levels and higher HDL levels compared to the men. In our dyslipidemic subgroup, there was no gender difference. I would suspect (but have not looked yet) that more women in our dyslipidemic group had prior CVD. Notably, in the FIELD study, there was no gender difference.

CardioExchange Editors: What questions do our readers have about the study? Do you think that we should abandon fenofibrate fro the treatment of diabetes?

2 Responses to “Does Fenofibrate ACCORD with the Treatment of Diabetes?”

  1. Triglycerides >200 mg/dl. Is this the right cu

    Henry comments that his threshold for fenofibrate treatment is “TG greater than 200 mg/dL — this is always associated with an HDL less than 40 mg/dL in a man and less than 50 mg/dL in a woman.”

    Yet only individuals with elevated TG and HDL <34 mg/dl (which was half the patients with high TG) were reported in the study. Were the date analyzed by men with HDL <40 mg/dland women with HDL <50 mg/dl?

  2. Why would you reach for fenofibrate first in the diabetic subgroup with TG > 200?