Negative Fish Oil Studies Shouldn't Diminish REDUCE-IT

Results of the STRENGTH and OMEMI trials, reported at the virtual scientific sessions of the American Heart Association, showed that two formulations of omega-3 fatty acids failed to reduce cardiovascular events in patients with elevated triglycerides and low HDL cholesterol. Steven Nissen, MD, chair of STRENGTH's executive committee, told MedPage Today that the findings suggest that the REDUCE-IT study (with a different omega-3 product) was falsely positive.

In this exclusive MedPage Today video, Deepak Bhatt, MD, of the Brigham and Women's Hospital Heart and Vascular Center in Boston and principal investigator of REDUCE-IT, pushes back on that assertion.

Following is a transcript of his remarks:

There was a lot of discussion after the various trials of omega-3 fatty acids were presented. In fact, there were two different trials that were presented. One was the STRENGTH trial, a large cardiovascular outcome trial looking at a specific mixture of omega-3 fatty acids, four grams a day of a mixture of DHA [docosahexaenoic acid] and EPA [eicosapentaenoic acid], as well as some other fatty acids in there, and that trial was negative -- that is, there was no benefit of the drug.

It was a large, well-done study, and I've got to congratulate the investigators. It was definitive in its findings.

It also shows, I thought, something really interesting that may not have fully come out, that the drug significantly reduced triglyceride levels, but it didn't reduce ischemic events, and I think that's an important disconnect.

Despite being adequately powered for events, there was no reduction in ischemic events, so I think there's an important message there. Not everything that reduces triglycerides necessarily is going to reduce cardiovascular risk. It's sort of an old story where we look at surrogate markers and sometimes can be misled, so it's an important message. The triglyceride story still needs to be resolved.

There are other trials that are starting or ongoing looking at other ways of lowering triglycerides -- in some cases within the same range as this drug, and in some cases several-fold greater reduction, some with the new RNA-based therapeutics, for example.

Those trials will answer definitively whether triglyceride reduction per se is sufficient to reduce cardiovascular risk when it's a large triglyceride reduction in some of those compounds produced, or whether we have been misled by thinking about triglyceride as a modifiable risk factor, whereas it might just be a risk factor -- and that we know for sure -- but perhaps one that isn't modifiable.

Also presented at this AHA, another well-done trial, the OMEMI trial, looking at 1.8 grams a day of a mixture of DHA, and so another omega-3 fatty acid trial, and this one was negative as well, specifically in STEMI [ST-elevation myocardial infarction] patients. Once again, I thought reasonably well-powered, but no signal of benefit.

In fact, both trials did have a suggestion of harm -- that is, a significant increase in atrial fibrillation strength, and I'd say statistics notwithstanding, an increase in the OMEMI trial as well. I can argue about the P value, but I think the investigators appropriately interpreted it as a real signal, so no benefit and some detriment.

So two trials that were negative, but each highly informative.

I think, when viewed in a larger context, this goes together with many trials of mixtures of EPA and DHA that haven't been positive. That's been a pretty consistent story. Now, there have been some exceptions if you look at subgroups of some of the trials, or individual endpoints -- the ASCEND trial for example. Vascular death was nominally significantly reduced, but the overall trial was neutral.

The VITAL trial as well, the overall trial was neutral. But if you look at some subgroups or specific secondary endpoints, there were some hints of positivity there, but overall the trials have not met their primary endpoint for the most part.

There are, however, two very notable exceptions to that. The REDUCE-IT trial, which I presented as a late-breaker at AHA 2018, and that trial published in the New England Journal of Medicine, detailed a large and significant reduction in cardiovascular events, including death from cardiovascular causes, with one specific prescription type of highly-purified EPA, so not a mixture of omega-3 fatty acids, but EPA or eicosapentaenoic acid. There, that was four grams a day, so a high dose of pure EPA and a very positive trial. We did see an increase in atrial fibrillation there as well.

Now, the other trial that preceded that -- it was also positive -- also used EPA, the JELIS trial, published in The Lancet in 2007. Another well-done, important trial, and there, with 1.8 grams a day in a largely Japanese population, or an entirely Japanese population, there was a significant reduction in ischemic events. That was a population that had a higher baseline EPA level that was given a lower dose of EPA than we used in REDUCE-IT and still a positive trial.

There were many differences between REDUCE-IT and JELIS, but one is particularly important in the context of a lot of discussions that came out of the STRENGTH trial, and that had to do with placebo choice.

Whereas in REDUCE-IT we used a mineral oil placebo -- at smaller quantities, of course, a substance generally believed to be inert when it is used in that quantity -- in the JELIS trial, they used no placebo. It was a so-called open-label trial, so randomized, but no actual placebo. Patients either got the EPA or they didn't get it, and the endpoints were adjudicated and blindly assessed in that trial.

That's a so-called probe design to minimize any sort of bias. We weren't really expecting bias in a long-term study of the sort JELIS was. So, two different trials of EPA that were positive, each using icosapent ethyl as the active ingredient. A lot of people don't realize that about JELIS.

But in that trial, no placebo, in REDUCE-IT, a placebo, but both essentially are coming to the same place with significant reductions in ischemic events, a 19% relative risk reduction with the 1.8 grams a day of icosapent ethyl used in JELIS, and a 25% relative risk reduction with the four grams a day used in REDUCE-IT, so actually a very nice cohesive story.

It does seem that, as is often the case in medicine, what we're studying matters in terms of the actual drug. Everything in this case can't be lumped together as a class effect because omega-3 fatty acids and their biology are very complex, and omega-3 fatty acids can be very different. The doses, the exact omega-3 fatty acid, and also the formulation.

Of course, the formulation matters a lot for any drug. That may be even more true with omega-3 fatty acids. Because if they're prepared or processed or packaged in a way that leads to oxidation of the omega-3 fatty acid, then it could lose any potential biological benefit it might have.

Really fascinating science coming out of the AHA about omega-3 fatty acids, but when viewed in a larger context, really quite in line with what we've seen with prior studies.

Comment