Sharing insights after study results with icosapent ethyl

Hello, my name's Kausik Ray. I'm a cardiologist from Imperial College London, and with me today is my friend and colleague, Professor Gabriel Steg from Paris. Gabriel, lovely to be with you. Very nice seeing you.

Gabriel, we have focused on lipids for many years, and before we've just had LDL lowering therapies, and a few years ago, icosapent ethyl was really a game changer in that we saw this huge effect in the REDUCE-IT trial, 25% relative risk reduction. There was a little bit of controversy potentially about the placebo arm, but a lot of that seems to have now been resolved with emerging data. Maybe you could actually just summarize some of the developments that we have now and where it stands.

First of all, I think it's fair to say that the investigators themselves from REDUCE-IT were as surprised as anybody by the magnitude of the benefit we saw because we were in true equipoise. We were really pleasantly surprised. In fact, it's the most positive trial I've ever contributed to in my career. It was a robust effect and very consistent across subgroups, but then questions were raised because we noticed, and we published in the primary paper that there was an increase in some biomarkers in the placebo arm, particularly CRP and LDL cholesterol were modest, but statistically significant increased. When you have 8,000 patients in a trial easily becomes statistically significant. Then it raised questions about whether what we were seeing was benefit from icosapent ethyl or potential harm from the placebo control of mineral oil. Now, we have a number of pieces of evidence that have emerged that I think are reassuring in that respect. First of all, we did a responder analysis of the trial patients in the placebo arm looking at those who increased CRP or LDL cholesterol or both compared to those who did not, and it's very striking that their events are the same. That does not seem to account for direct harm in the placebo patients from mineral oil.

Second, we had pieces of evidence from outside trials. There was previously the JELIS trial done in Japan using icosapent ethyl at the dose of 1.8 grams per day versus control. There was no placebo that had shown benefit, both in primary and secondary prevention. Then more recently a year ago at the American Heart Association meeting, our Japanese colleagues reported another trial called RESPECT-EPA testing, again, 1.8 grams per day of EPA, purified EPA in patients with elevated triglycerides and a favorable EPA to arachidonic acid ratio. Again, they reported a signal of benefit, a composite cardiovascular outcomes trial. Again, it was not against placebo, it was against usual care as a control. We now have three trials with relatively high-dose EPA that are consistently showing benefit, one of which uses a placebo, and two of which use no placebo. I think that really now becomes a very clear picture that high-dose EPA is beneficial.

The question for me now is really how do we implement this rather than should we use it because with the magnitude of the benefit we're seeing, it'd be a shame not to have our patients benefit from that. I think that's really important because this is a patient group that we talk a lot about residual risk. We know that we've got an increase in obesity, diabetes, we see that with triglycerides elevated, we've seen with, for example, other studies now that the fibrates don't really reduce cardiovascular events on a background of statin therapy. This group of individuals with a TG of 125, 135 up to 500, let's say, that's a high residual risk group that, at the moment, we don't have anything specific for. We now have that external validity, the total events analysis, which really show you how cost-effective this is.

re there any potential signals because we've seen with other fish oils or marine derivatives, that where there was an excess risk of atrial fibrillation and potentially I think in REDUCE-IT, there was a small signal for increased bleeding. What should the clinician do? What are the clinical implications, if any? That's a very important question. Of course, if we have a benefit, we have to see is there a risk and what's the balance of risk and benefit. First of all, there's clearly an increase in the risk of atrial fibrillation or flutter with icosapent ethyl in the trial. Since then it's been confirmed in other trials, it's actually quite consistent across trials and it appears to be dose-related. Our Scandinavian colleagues have published an elegant analysis where they show that there's a direct relationship between the EPA content and on one side the cardiovascular benefit. On the other side, the potential signal for AFib. Interesting. However, a couple of things have to be said. First of all, there was a dramatic reduction in stroke both overall and in that subgroup of patients who got AFib, either at the onset of the trial or who developed AFib as an incident event. There's no reason to deprive patients from the benefit of the drug because of fear of AFib that would cause stroke because actually icosapent ethyl reduces stroke substantially.

The second observation is, there's an increase in bleeding, borderline increase, and fortunately, it's not an increase in intracranial bleeding or fatal bleeding, but there's an increase in minimal or minor bleeding, borderline significant. That again, is consistent with prior evidence of some antiplatelet efficacy of omega-3 fatty acid that's been reported 20 years ago. It raises the question of should we restrict this drug to patients who are not on DAPT. Because of the concern with increased bleeding. We did an analysis where we looked at patients with recent ACS because this is a group that's frequently on DAPT, and this is a group where we might be particularly concerned with the risk of an incident AFib. What we found when we looked at the recent ACS group within REDUCE-IT is twofold. First, this is the group that has the highest magnitude of benefit from icosapent ethyl. Because they are at the highest risk. Second, there is a modest increase in the incidence of AFib consistent with what is seen in the overall trial. There's a reduction in stroke, there are lower rates of stroke in that group, and there's no increase in bleeding. Even in the subset of that subset who is on DAPT. That intrigued us and we believe the explanation is the following, that the modest antiplatelet efficacy of omega-3 fatty acid can be measured and perceived if patients are not on potent antiplatelet therapy, but if they're on DAPT, that is overwhelmed by platelet inhibition related to the DAPT itself. If you're on DAPT, it makes no difference whether you're taking icosapent ethyl or not with respect to bleeding.

I think it's an important thing because it tells us that we don't have to deprive the highest-risk patients, those who've had a recent ACS, from the benefit of the drug, we can start the drug in hospital even if they're on anticoagulants or on DAPT. That's really been reflected by the recent ESC post-ACS guidelines and the management where essentially once you are obviously starting people on statins as part of routine clinical care, antiplatelet agents, and there's a recommendation that if you're triglycerides are elevated in line with the inclusion criteria of the REDUCE-IT trial, that basically these are people that we should at an early stage consider for high dose icosapent ethyl, 2 grams twice a day. Essentially I think the challenge now is for how we put that into clinical practice.

One of the things, for example, is that yes, we are using triglycerides as a way of identifying the people that will benefit, but also we know that we don't actually have to, unlike LDL cholesterol where you're talking about targets, et cetera, in this scenario. It's a bit more like an antiplatelet drug where you just give the person the treatment and you are using triglycerides if you like, to enrich and identify who is going to benefit. One of the things I always now try to look at is, okay, I've given all the first-line treatments, who are the patients with high residual risk? The patients with diabetes, obesity are usually going to be people with a high TG. For example, if you find a low HDL, there's usually a high TG as well.

Who do you try and identify for this potential treatment and how? I think it's generally a good idea to apply the evidence to exactly the patients like those in the trial. I'm trying to identify a second prevention patients who are on statin with an LDL cholesterol between 40 and 100 and who have TGs between 135 and 500 exactly like the patients in the trial. The second thing is LDL cholesterol management is critical. There's absolutely no question about that, but I think that managing LDL cholesterol and addressing the risk of TGs and elevated TGs is not an either/or proposition. It's both, it's all of the above. You want to address the LDL cholesterol and you want to give icosapent ethyl to these patients because it is so beneficial and the side effects are modest and the cost is modest. I think it's a good deal for the patient. It's particularly important because in the trial, all the patients were on statins. Again, it's not either/or. With well-controlled LDL. You've got the LDL controlled and they've got high event rates based on the triglycerides.

So I think we've probably covered everything that we needed to in the time that we've got available. I want to thank you for your insights and I think the key messages are, there's a clear consistent benefit now that we've got three different trials in different settings that have clearly shown the same large treatment effect which is great. Also, you provided assurance about not having to worry about excess of bleeding or atrial fibrillation increasing the risk of strokes. I mean, that clearly is, goes in the opposite direction. I think that's really, again, a good thing. We've now got guidelines that talk about implementing these, particularly in the highest-risk patients. The real challenge for all of us now is to actually try and put all of this into clinical practice and reduce some of that residual risk that otherwise exists, so I want to thank you for your time and sharing your insights. It was a pleasure. Thank you.