Lipid Luminations: Whither Fibrates?

The role of fibrates in treating patients with hyperlipidemia remains controversial.1-3 Recent randomized clinical trials of fibrates, alone or in combination with statins, have been inconclusive.4,5

Fibric acid derivatives exert favorable lipid effects through the ligand-dependent transcription factor PPARα, which can regulate multiple target genes.6 Mechanistically, fibrates have been shown to induce lipoprotein lipase transcription7,8 and reduce hepatic apo-CIII production.9-11 Fibrates also increase synthesis of the HDL apolipoproteins A-1 and A-II via stabilization of MRNA transcription,12-14 which likely increases HDL “functionality.” This class of agents has also been shown to increase LDL particle size and decrease LDL particle triglyceride content, thereby increasing affinity of LDL particles for the apo B/E hepatic LDL receptor.15-18

Despite what appears to be very favorable lipid effects, randomized clinical trial results have been disappointing. Trials comparing gemfibrozil to placebo did show favorable results.19,20 However, these trials did not employ standard-of- care “background” statin therapy, and the interference of gemfibrozil with statin glucucoridation and subsequent increased myopathy risk makes combining these agents untenable.21-24

Thus there was great promise with fenofibrate, which does not interact unfavorably with statins.21 In the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) trial — designed as a fibrate monotherapy trial — more than 9,000 subjects were randomized to placebo or fenofibrate, 200mg/day. Median baseline triglyceride level in each group was 154mg/dl, with median HDL level of 43mg/dl. The trial failed to meet its primary endpoint, with no reduction in total CHD events with fenofibrate. However, in the analysis of the subgroup with lipid characteristics of the metabolic syndrome, a statistically significant 14-percent reduction in the primary endpoint was achieved with fenofibrate. Additionally, statin “drop-ins” were greater in the placebo arm, with 40 percent of placebo group patients taking a statin by year five of the study.5

In the Accord Lipid Study, more than 5,000 patients with type II diabetes mellitus were treated with simvastatin therapy, and were randomized to placebo or fenofibrate. Again, the primary endpoint of a reduction in myocardial infarction, stroke, or death was not achieved with fenofibrate in the trial. However, in the prespecified analysis of subjects in the highest tertile of baseline triglyceride level (>204mg/dl) and lowest tertile of baseline HDL (<34mg/dl), a borderline statistically significant reduction in primary endpoint was achieved with the addition of fenofibrate. (12.4 percent fenofibrate group; 17.3 percent placebo group; p=.06).4

Similar “lipid-dependent” results have been seen in trials of other fibrates. Bezafibrate, available in Europe, was evaluated compared to placebo in the randomized Bezafibrate Infarction Prevention (BIP) Study. In this study, more than 3,000 patients with coronary artery disease were randomized to 400mg bezafibrate per day or placebo. Subjects with triglyceride levels >300mg/dl were excluded, but all had HDL cholesterol levels <45mg/dl. Insulin-dependent diabetes was also excluded. Open-label use of additional lipid-modifying therapy was almost twice as common in the placebo group (237 subjects) as in the bezafibrate group (136 subjects). Again, the primary endpoint of MI or sudden death failed to be reduced in the bezafibrate group. Post-hoc analysis of events, based upon triglyceride levels, revealed no benefit of bezafibrate in subjects with baseline triglyceride <200mg/dl, but a nearly 40 percent reduction in primary endpoint was achieved in subjects with triglyceride levels > 200mg/dl (p=.02).25

The gemfibrozil monotherapy trials, referred to earlier, yielded similar results. Though the overall trial results were positive with gemfibrozil in the Helsinki Heart Study, the benefit was most striking in subjects with LDL/HDL ratio >5.0, and triglyceride levels > 200mg/ dl. In this group, gemfibrozil reduced ischemic coronary events by 71 percent.19 Similarly, a benefit of gemfibrozil was seen overall in the VA-HIT Study, but was most pronounced in those subjects with the most severe insulin resistance. In subjects with diabetes (25 percent of the entire study population) the combined ischemic endpoint of nonfatal MI, CHD death, or stroke, was reduced by 32 percent with gemfibrozil (p=.004), whereas in the larger nondiabetic study population, a statistically significant reduction was not achieved (18 percent; p=.07). Furthermore, the benefits of gemfibrozil increased with each quartile of fasting plasma insulin levels. In the lowest quartile, no reduction in events was noted, while in the highest quartile of fasting insulin level (> 39uU/ml) a 35-percent reduction in events was seen with gemfibrozil.

So, where does this leave us? Clearly, fibrates exert many beneficial effects in atherogenic dyslipidemias,6  and many trials of fibrates have clearly shown differential “lipid-dependent” beneficial effects of fibrate therapy.4,5,19,25,26 Yet, without a recent positive trial with contemporary fibrate therapy, we are left without an evidence-basis for our lipid-subgroup- driven beliefs. Of course, no large fibrate trial has ever studied a population with baseline median triglyceride levels >200mg/dl, despite strong evidence that it is only in this group that fibrates have benefit. Let us hope that the planned VA FIT trial, a randomized trial of fenofibrate with triglyceride levels <200mg/dl an exclusion criterion, in fact commences and proceeds to completion. Without such much-needed evidence, very relevant clinical questions may remain unanswered, and the role for fibrates in hyperlipidemia treatment and event reduction may remain undefined.

Disclosure statement: Dr. Shurmur has no disclosures to report.

References are listed on page 34.