Nothing Fishy: Triglyceride Lowering With High-Dose EPA


Interviewer: John M. Mandrola, MD; Interviewee: Jane M. Armitage, MBBS, FRCP, FFPH


November 13, 2018

John M. Mandrola, MD. Hi, everyone. This is John Mandrola from the | Medscape Cardiology, and I'm here at the American Heart Association (AHA) meeting in Chicago. I'm really pleased to be joined by Professor Jane Armitage from Oxford University. We're going to discuss the big trial from AHA, and that is the REDUCE-IT trial.[1] My main question is, what do you think? Do you believe that the results are this strong.

Jane M. Armitage, MBBS, FRCP, FFPH: I thought it was a really interesting result and I think people were surprised by it. But if you think about it logically, perhaps it's not so surprising. Let me just go back and explain. We've known for many years that triglycerides are related to the risk for coronary disease, but typically in epidemiology we've adjusted the association for HDL cholesterol (HDL-C) levels, and then the association with triglycerides goes away. And people have believed for many years that actually it's the HDL-C that's important rather than the triglycerides, but there's been emerging evidence—both from genetics and other sources—that perhaps we were looking at the wrong target.

That's one bit of the prior evidence. The other is that we lipidologists have known for many years that high doses of fish oils will lower triglyceride levels, and they're in our therapeutic toolbox for people with high triglycerides—particularly people who cannot take fibrate drugs for some reason.

This was obviously a highly concentrated preparation of EPA (eicosapentaenoic acid) that has lowered triglycerides in the REDUCE-IT trial, by about 0.4 mmol/L (35 mg/dL), which is a good reduction. The important thing about REDUCE-IT was that they took a high-triglyceride population; they took a high-risk population who all have vascular disease, and they were all on statins so their LDL-C was well controlled, but they still had high triglycerides. That is a particularly high-risk group of patients, and we haven't really known how to manage them. In lipid clinics, we've tended to use high-dose statins to get LDL-C as low as possible.

Mandrola: In REDUCE-IT, there was a 25% relative risk reduction, a massive 4.5% of absolute risk reduction. You were a discussant for the VITAL trial,[2] in which lower-dose omega-3 oils showed no difference in outcomes. How do you square the differences between these two trials?

Armitage: The VITAL trial used a 1-g mixture of EPA and DHA [docosahexaenoic acid]; that may not matter, but the dose definitely does matter. If we look just at the triglyceride effect of 1 g per day, this wasn't reported in VITAL, but we know from other trials that it is pretty modest—about 0.1-0.16 mmol (9-14 mg/dL)—so it's perhaps not surprising that you're not getting the cardiovascular benefit on 1 g/day.

Mandrola: You think the dose is more important than the purity?

Armitage: I do. But I don't know that. My hunch is that it's more important.

Mandrola: Do you think that the amount of the triglyceride reduction in REDUCE-IT is enough to account for the benefit or do you think that there are other potential mechanisms that the high-dose fish oil could have worked upon?

Armitage: There was a small reduction in non-HDL cholesterol in REDUCE-IT alongside the reduction in triglycerides, and that could have accounted for some of the benefit. Whether there are other benefits on other pathways is difficult to know. I think that we now need to go back to the epidemiologic relationship and try to see. I don't think it's very out of line with what you might expect.

It may be that they were a little lucky to get 25% rather than 20% [reduction], but I don't think it's that implausible.

Mandrola: One of the other points of discussion about REDUCE-IT, at least in the social media space, has been that the placebo (mineral oil) group had an increase in LDL-C. And there's some speculation that the mineral oil could have been less than an inert placebo.

Armitage: I don't know what was in the mineral oil; it's a rather odd term to use. I would like to know whether it was an entirely neutral oil, because I guess it is potentially feasible that it could be toxic in some way and that you're not comparing with a neutral placebo. But I don't know what "mineral oil" means in this context.

Mandrola: From a Bayesian perspective, you don't agree that all of the previous trials really showed no signal of benefit. You believe that there is a prior signal, and REDUCE-IT isn't that much of an outlier—is that correct?

Armitage: We don't have other trials of high-dose fish oils on clinical outcomes, so from that point of view we have no priors. But we do have trials of fibrate drugs, which also lower triglycerides by a fairly comparable amount to what we saw in REDUCE-IT.

Now, if you look at meta-analyses of the older fibrate trials—none of them done in the current era—the overall benefit is around 10%-12% reduction in risk.[3] And those tended to be in populations of all-comers.

In those older trials, if you look at the subgroups of patients who have high triglycerides and low HDL-C, typically the risk reduction was around 25%. There are published meta-analyses of those subgroups.[4,5] So from that point of view, I think there are some data from the past, which is supportive of what we saw replicated in REDUCE-IT.

Mandrola What percentage of people do you think are like the REDUCE-IT trial population, just in an all-comers cardiology clinic?

Armitage: I don't know exactly. But my estimation is that if you have a secondary-prevention population, well treated with high-intensity statins, it may be about 30% of the population but it will depend on the obesity levels, the diabetes levels and the other drivers of high triglycerides.

Mandrola: Where I practice and in some other US states with high obesity rates, it may be a significant number of people.

Armitage: It could well be a significant number of people, and one of the things one can do is look back at the trial populations to see what proportion of patients it is.

Mandrola: If we come across a patient who's like the REDUCE-IT population, what percentage of these patients can be modified with diet lifestyle, if they'll do that?

Armitage: I think probably the vast majority. But, of course, we know that we give these people that advice and it's very hard for them to take it. If people can get down to a normal weight, then quite possibly many of the lipid profiles would return to normal. We know that people find that very difficult.

But in practice, we all have seen patients who come in with grossly abnormal lipids, and if they really can get from, say, a BMI of 28 or 30 down to a BMI of 23, then often their lipid profiles will entirely normalize.

Mandrola: You really think that this is a practice-changing piece of evidence?

Armitage: I do. Yes. I think it's going to need replicating, and one of the questions will be: Does it need to be this purified preparation, or if we just give 4 g of a standard fish oil preparation, would it have the same impact? It's a fairly simple experiment to take a relatively small group of hypertriglyceridemic patients and actually compare the effects of 4 g of the preparations that have been tested in the other trials against this. That would be an initial interesting first step.

Mandrola: Excellent. Thank you for your expert commentary. I appreciate it.


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.