August 10th, 2011

Rethinking Trilipix — And the Process for Approving Lipid-Modifying Drugs

Editor’s Note: In an article published in the New England Journal of Medicine, three members of the FDA’s Endocrinologic and Metabolic Drugs Advisory Committee (Allison Goldfine, Sanjay Kaul, and William Hiatt) offer their perspective on the May 19 committee meeting to review the controversial ACCORD-Lipid Study. Here, one of those authors, Sanjay Kaul, provides his own account of the session for CardioExchange members.

After reviewing data from the ACCORD Lipid Study, the 13-person Endocrinologic and Metabolic Drugs Advisory Committee (EMDAC) of the FDA has voted unanimously to request that Abbott conduct a new trial of its lipid-modifying drug Trilipix (fenofibric acid). Specifically, EMDAC advises that the trial identify whether men and women who are at high risk for coronary heart disease (CHD) and who have already reached their LDL-cholesterol goals on a statin — but who have persistently high triglycerides and low HDL levels — benefit from Trilipix as add-on therapy.

The original, NHLBI-sponsored ACCORD Lipid Study involved 5518 high-risk patients with type 2 diabetes and LDL levels of 60 to 180 mg/dL (mean, 101 mg/dL). Combination therapy with daily Trilipix (160 mg) plus simvastatin showed no advantage over placebo plus simvastatin in preventing adverse cardiovascular events during a mean follow-up of 4.7 years (annual rate, 2.2% vs. 2.4%). Among women treated with Trilipix, there was a suggestion of harm from the drug; among the subgroup of patients with baseline triglyceride levels ≥204 mg/dL plus HDL levels ≤34 mg/dL (the so-called “dyslipidemic profile”), there was a suggestion of benefit.

Penetrating the nuances of the new EMDAC recommendation requires some background beyond these facts.

A Brief History of Fibrates

After statins, fibrates are the most widely used lipid-modifying agents. Annual sales in the U.S. for the three fibrates currently approved by the FDA to treat hyperlipidemia — gemfibrozil, fenofibrate, and fenofibric acid (Trilipix) — are in the billions of dollars. Trilipix is the only fibrate approved as combination therapy with a statin to reduce triglycerides and increase HDL levels in patients who have mixed dyslipidemia and CHD or a CHD risk equivalent (and who are on optimal statin therapy to achieve LDL goals). The Trilipix/statin combination-therapy approval was based on three 12-week studies of the drug’s impact on lipid biomarkers. No clinical-outcome data currently support this indication.

Indeed, unlike statins, fibrates have had mixed results in placebo-controlled trials when it comes to reducing CHD risk and actually saving lives. Studies of gemfibrozil (conducted before the statin era) have shown primary- and secondary-prevention benefits, whereas trials of fenofibrates have not (see this recap of major fibrate trials). And as summarized above, the recent ACCORD Lipid Study did not show a benefit of fenofibric acid as an add-on to simvastatin in high-risk patients with type 2 diabetes and mixed dyslipidemia.

Nevertheless, fibrate use in the U.S. doubled from 2002 to 2009. Furthermore, the two agents with no clear demonstrated benefit — fenofibrate and fenofibric acid — are gradually displacing gemfibrozil in terms of market share: Fenofibrates and fenofibric acid now command about three-quarters of the fibrate market, compared with about one-quarter for gemfibrozil. Fenofibric acid specifically is used (as add-on to a statin) primarily in dyslipidemic patients, in line with the indication on the prescribing label.

The FDA Committee’s Discussion and Vote

From its review of the ACCORD Lipid Study, EMDAC determined that the trial was not specifically designed to assess the clinical benefit of treating patients with residual low HDL and high triglyceride levels who were on a full-dose statin. Thus, the trial could not adequately evaluate the benefits of Trilipix as add-on therapy.

With regard to the two prespecified subgroup analyses (gender and “dyslipidemic profile”), the committee expressed general caution against acting on the basis of these data. They cited, among other concerns, that the trial as a whole was negative, that the analyses were limited by relatively small numbers of patients and events, and that other relevant adjustments were not performed.

The unanimous vote to call for a new trial came despite the fact that the trial’s aim would be to prove whether the indications already listed on the current medication label are supported by evidence of clinical benefit. With regard to how to handle that conundrum, the committee’s vote was split: 3 in favor of maintaining the current label, 6 in favor of revising the label, and 4 in favor of withdrawing the approved indication.

The committee also discussed the regulatory standard for approving lipid-modifying agents (currently based on proof of improvements in surrogate lipid biomarkers). Some members said that the FDA should instead consider requiring a preapproval cardiovascular-outcomes trial. Others felt that the current surrogate-biomarker–based approval process allows potentially beneficial therapeutic options to become available before results from a definitive outcome study can be obtained.

The Regulatory Fallout

In my opinion, the EMDAC decision has major regulatory implications. Taken together, the data from the ACCORD Lipid Study, ILLUMINATE, and AIM-HIGH strengthen the argument for changing the standard for approving lipid-modifying therapies: from an emphasis on surrogate lipid biomarkers to one on clinical benefits and cardiovascular outcomes. Although such outcome trials are likely to be time- and resource-intensive, if the duration of drug patents were extended, financial risk to the manufacturers would be mitigated. Pharmaceutical companies could fulfill the preapproval outcome-trial requirement yet still have sufficient time to market the drug and earn a return on their investments.

Specifically with regard to Trilipix, if the FDA were to ask the sponsor to conduct the clinical trial recommended by EMDAC, the following conditions would need to be considered:

  • Start the trial ASAP, preferably within 6 to 9 months of the FDA’s decision.
  • Enroll high-risk cohorts of interest (at least 30% to 40% women and minority groups; at least 50% patients with a “dyslipidemic profile” of elevated triglycerides plus low HDL).
  • Ensure proper trial design and conduct, and minimize rates of crossover and dropout.
  • Enroll subjects quickly and complete the study within 3 to 5 years.
  • Allow the FDA to withdraw the indication if the sponsor does not complete the trial in a timely and efficient manner. Only then would the FDA be able to discourage the sponsor from taking an approach like Merck’s with ezetimibe or GlaxoSmithKline’s in RECORD.

What are your thoughts on the EMDAC recommendations with respect to Trilipix? Does this development affect how you view the general process for approving lipid-modifying drugs?

5 Responses to “Rethinking Trilipix — And the Process for Approving Lipid-Modifying Drugs”

  1. Dan Hackam, MD PhD says:

    Fibrates have shown consistent benefits only in patients with atherogenic dyslipidemia defined variably as the dyslipidemia commonly accompanying metabolic syndrome, and in five large fibrate trials on the basis of such parameters as body mass index, low HDL and high Trigs. This subgroup produced benefits for three species of fibrates in five trials and meta-analysis of the 5 trials suggests about a 35% reduction in risk (even on top of statins – ie in ACCORD and added on in FIELD in some patients). Yes it’s a subgroup analysis, but it’s extremely consistent from trial to trial and entirely consistent with the pharmacophysiology of these drugs. Do we need a trial to prove this prospectively? It might be nice, but I’m not waiting for one to prescribe fibrate therapy in this specific setting.

    Competing interests pertaining specifically to this post, comment, or both:

  2. Consistency and pharmacophysiology are “slippery slope” arguments! Remember Huxley’s often quoted caveat “The tragedy of science, the slaying of a beautiful hypothesis by ugly facts”.

    Results of metaanalyses, especially when based on pooling disparate trials (primary plus secondary prevention, fibrates with or without statin, etc.) that use nonuniform TG and HDL cutoffs to define dyslipidemia, is not what I would consider strong evidence.

    Medicine works best when the ‘art’ and ‘science’ merge. If one were going to prescribe fibrates in this specific setting (atherogenic dyslipidemia in high-risk patients), then one should first optimize LDL-lowering with a statin before adding on fibrates.


  3. Dan Hackam, MD PhD says:

    Subgroup analyses, which are fraught with positives (specifically false positive and negative results), are still the only solution within large randomized trials for the problem of “personalized medicine”. Large trials give population-based averages (or sample-based averages, if you will); whereas the patient in front of us in clinic is unique and may or may not respond to therapy. How do we know if they will respond or not, or whether 20-30 years of fibrate therapy will essentially expose them to risk with no benefit accrued?

    They key to my mind is that the atherogenic dyslipidemia subgroup has been replicated across five separate mega-trials, with three distinct fibrates, in various populations – on and off statin, primary and secondary prevention, etc. I feel that patients with a large plaque area burden (a debate perhaps for another time!) with atherogenic dyslipidemia and LDL already to target (either with or without statin) deserve a fibrate. This isn’t an ISIS-2 consellation of Virgo issue. This is a repeatedly replicated subgroup analysis (admitted with varying definitional cutpoints, but always getting at the underlying issue of metabolic syndrome, present in 1 in 4 Americans).

    I would love to know whether Abbott plans to replicate, prospectively, the above findings in a brand new trial as per the panel recommendations, or whether they feel it is not worth it (due, for example, to loss of patent issues).

    The other issue is that AIM-HIGH has locked us into a corner. If we take evidence-based medicine at face value, we are really limited to four approaches:
    -higher dose statins are better than lower dose statins
    -omega-3 fatty acids
    -fibrates in metabolic syndrome

    Forget about ezetimibe, cholestyramine, probucol and niacin.

    As you can see, how one rates the evidence in any particular area has to be based, in part, on the evidence underlying competing approaches (such as niacin or ezetimibe). I am simply trying to do my best for my patients.

  4. We agree that fibrates might be considered in appropriate patients whose LDL levels are at target (off or on statins).

    The patent for Trilipix is set to expire in January 2025 (Tricor patent expires this year). So, if the FDA requires the sponsor to conduct the trial as per panel recommendations, it is in their best interest to do it in a timely and efficient manner.

    It is interesting to note that the hypothesis-generating observation that fibrates require a permissive environment of dyslipidemia for them to be effective was first made in a post hoc analysis of the Helsinki Heart Study. Nearly 2 decades and 4 major trials later, we still don’t have any validation of this hypothesis!

    Subgroup analysis are best utilized to evaluate consistency of treatment effects across subsets (quantitative interactions), not inconsistency (qualitative interactions).

  5. Dan Hackam, MD PhD says:

    I like your perspective but I know you are a scholar of Bayesian statistics. Rather than dismissing pharmacophysiology of fibrates as a “slippery slope” (which, as with you, I am inclined to do with any surrogate marker), I would suggest the pharmacophysiology plus the subgroup analyses form a very strong pre-test probability of benefit for fibrates. Absent the clinical trial you propose (which I believe will never occur), we are left with a bit of a black box. As a clinician who cares for patients at high risk for events, I must weigh these uncertainties and coupled with the scant risks of fibrates and the probably benefits (again that very strong pre-test probability), I will often reach for a fibrate for a patient with so-called “atherogenic dyslipidemia”.

    Here is a news piece on an article by Ovbiagele, again confirming this strong “pre-test” probability.

    By Nikki Withers

    07 June 2011

    Atherosclerosis 2011; Advance online publication

    MedWire News: Fibrate treatment is effective for reducing cardiovascular risk in patients with atherogenic dyslipidemia, a meta-analysis of published studies suggests.

    Specifically, the team found that fibrate therapy was associated with a 16% to 29% reduction in vascular risk in these patients.

    “The effect was most pronounced in persons with presence of both elevated serum triglycerides and reduced high-density lipoprotein (HDL) cholesterol levels,” write Bruce Ovbiagele (University of California, San Diego, USA) and colleagues in the journal Atherosclerosis.

    The researchers’ systematic review and meta-analysis included six studies involving patients with atherogenic dyslipidemia (serum HDL cholesterol 200 mg/dl [2.26 mmol/l]).

    Ovbiagele et al report a significant 29% reduction in risk for vascular events (defined as the composite of nonfatal myocardial infarct, nonfatal stroke, and vascular death) among patients with hypertriglyceridemia and low HDL-cholesterol (n=5068) who were using fibrates, compared with placebo.

    Among patients with either hypertriglyceridemia (n=7389) or low HDL cholesterol (n=15,303), fibrate use was associated with a 25% and 16% reduction in vascular event risk, respectively, compared with placebo.

    Fibrates had no significant effect in reducing vascular risk in patients who did not have atherogenic dyslipidemia, notes the team.

    Ovbiagele and co-authors say: “Our results further reinforce the [National Cholesterol Education Program] Adult Treatment Panel III guidelines which stress the importance of incorporating targeted modification of elevated serum triglycerides and low HDL cholesterol levels as part of an optimal vascular risk reduction strategy.”

    They suggest that, to substantially reduce future vascular events, fibrate treatment should be directed at patients who present with markers of atherogenic dyslipidemia.

    MedWire ( is an independent clinical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2011

    Free abstract
    Efficacy of fibrates for cardiovascular risk reduction in persons with atherogenic dyslipidemia: A meta-analysis

    Meng Lee

    , Jeffrey L. Saver

    , Amytis Towfighi

    , Jessica Chow

    , Bruce Ovbiagele

    Received 5 November 2010; received in revised form 23 March 2011; accepted 15 April 2011. published online 18 May 2011.
    Full Text
    Supplemental Materials



    Recent data suggest that non-targeted treatment with fibrates modestly reduces the risk of incident cardiovascular events. However the effect of fibrate treatment may be particularly beneficial in patients with guideline-endorsed indications for therapy due to evidence of atherogenic dyslipidemia. We conducted a systematic review and meta-analysis to investigate the influence of fibrates on vascular risk reduction in persons with atherogenic dyslipidemia.


    Systematic search of Pubmed, CENTRAL and recent reviews was conducted to identify atherogenic dyslipidemia (serum high density lipoprotein cholesterol [HDL-C]200mg/dl) cohorts from randomized controlled trials. RR with 95% CI was used as a measure of the association between fibrate therapy and risk of cardiovascular diseases, after pooling data across trials in a random-effects model.


    Six trials met selection criteria. Compared to placebo, the greatest benefit with fibrate treatment was seen in 7389 subjects with high triglycerides, fibrate therapy reduced risk of vascular events (RR 0.75, 95% CI 0.65 to 0.86, P<0.001); and in 5068 subjects with both high triglycerides and low HDL-C (RR 0.71, 95% CI 0.62 to 0.82, P<0.001). Less benefit was noted in 15,303 subjects selected for low HDL-C (RR 0.84, 95% CI 0.77 to 0.91, P<0.001). Among 9872 subjects with neither high triglycerides nor low HDL-C, fibrate therapy did not reduce subsequent vascular events (RR 0.96, 95% CI 0.85 to 1.09, P=0.53).


    Fibrate treatment directed at markers of atherogenic dyslipidemia substantially reduce subsequent vascular event risk.