STRENGTH CV outcomes trial of omega-3 fatty acid stopped for futility

Carl E. Orringer, MD, FNLA

There were a number of differences between STRENGTH and REDUCE-IT. STRENGTH had HDL enrollment criteria and REDUCE-IT did not. There were slightly different inclusion criteria for triglycerides: In REDUCE-IT, it was 150 mg/dL to 499 mg/dL, in STRENGTH it was 180 mg/dL to 499 mg/dL. REDUCE-IT had a mineral oil placebo, STRENGTH had a corn oil placebo.

Both studies had patients on what appeared to be good statin regimens. In REDUCE-IT, baseline LDL was 40 mg/dL to 100 mg/dL. In STRENGTH, baseline LDL was less than 100 mg/dL but could be greater than or equal to 100 mg/dL on maximally tolerated statin therapy.

Putting those differences aside, a couple of things are of interest. In REDUCE-IT, the beneficial effect of icosapent ethyl was statistically greater in the subgroup with low HDL (≤35 mg/dL) and higher triglycerides (≥ 200 mg/dL), but the baseline HDL level was not part of the inclusion criteria, as it was in STRENGTH. One might have predicted a greater chance for beneficial effects in STRENGTH, as the investigators selectively employed a low HDL/high triglyceride population.

However, there are differences in DHA vs. EPA. Both lower triglycerides and non-HDL, but DHA plus EPA raises LDL levels, especially in those with very high triglycerides and low HDL, whereas EPA doesn’t raise and may lower LDL. In fact, in patients with triglycerides 200 mg/dL to 500 mg/dL treated with statins, EPA lowers LDL by about 6%, which is similar to doubling the dose of a statin. Also, DHA plus EPA has varying effects on apolipoprotein B levels, but EPA has been shown to lower ApoB. And REDUCE-IT showed EPA lowered the inflammatory marker high-sensitivity C-reactive protein, whereas DHA plus EPA does not consistently affect CRP levels. There have also been mechanistic studies suggesting EPA may have membrane-stabilizing effects, whereas that may not be the case with compounds containing DHA plus EPA.

The bottom line is that we did not know whether therapy with DHA plus EPA would provide similar, greater or lesser benefit on CV outcomes compared with those demonstrated with icosapent ethyl, a highly purified EPA product. The premature termination of STRENGTH for futility tells us that this formulation of marine omega-3 fatty acids was not effective in preventing these events. We do not know whether any of the above differences in the patient characteristics, study design or the biological characteristics of the two marine omega-3 fatty acid preparations explain the divergent outcomes.

What we know is that EPA, when given in the form of icosapent ethyl to patients with diabetes older than 50 years with additional risk factors or to patients with established atherosclerotic CVD 45 years or older, reduces risk in patients with triglycerides 150 mg/dL (perhaps 135 mg/dL) to 499 mg/dL. In addition, the JELIS study from Japan used a different form of EPA (Epadel, Mochida) and showed benefit in patients who were not significantly triglyceridemic and had relatively high cholesterol.

It seems that EPA might be where the action is in terms of benefit. We all look forward to future studies that will explore the mechanisms for these differences.

How much of this is an anti-inflammatory effect? How much of this is a lipid-related effect? Are there other mechanisms at play? Does icosapent ethyl work in other populations?

A big question is, if you have a patient on maximally tolerated statins plus ezetimibe with mildly elevated triglyceride levels and an LDL level that is persistently mildly elevated, do you go to a PCSK9 inhibitor or EPA first? Both drugs are associated with favorable ASCVD outcomes, but comparative efficacy studies have not been performed.

The role of diet and exercise in this context should also be explored further. There was no intense dietary intervention in REDUCE-IT. Perhaps if more careful attention is given to counseling on heart-healthy dietary and exercise habits, fewer patients might need icosapent ethyl.

At this time, we need to be cautious about using DHA-containing compounds and we should be moving toward using EPA in certain hypertriglyceridemic patients who are maximally treated with statins but still maintain high risk.

This was a very interesting and somewhat surprising result. This is why we do the studies.

JoAnn E. Manson, MD, DrPH, FAHA

It seems that the formulation of the omega-3s tested may be extremely important. At this time, it appears that high-dose EPA, especially in the formulation of icosapent ethyl, is particularly protective against CVD. We need to have a better understanding of whether there are other formulations of marine omega-3s, when given in high doses, that can have similar benefits. More research is needed in this area, as omega-3s hold great promise for reducing residual risk and have relatively low rates of adverse events.

We also need to explore further some of the mechanisms that may explain why the addition of DHA may be blunting the CV benefit. This may go beyond LDL changes. And we need to understand whether icosapent ethyl or other omega-3s could benefit high-risk populations beyond those meeting criteria for inclusion in these recent trials.

The divergent findings of the STRENGTH and REDUCE-IT trials highlight the need for additional research to understand the role of formulation and dose of omega-3 fatty acids and the effects on other important clinical outcomes.

In the VITAL trial (Manson JE, et al. N Engl J Med. 2018;doi:10.1056/NEJMoa1811403), we saw reductions in coronary events but not stroke with Omacor (BASF/Pronova BioPharma), which contained 840 mg of marine omega-3 fatty acids (465 mg EPA and 375 mg DHA).  For MI specifically, the reduction was 28%, with greater reductions in African Americans. The benefits for MI in African Americans were dramatic but require replication and further exploration. Also, because DHA appears important for cognition, it will be important to look at cognitive function over time in these trials.

In a recent meta-analysis we performed (Hu Y, et al. J Am Heart Assoc. 2019;doi:10.1161/JAHA.119.013543), there appeared to be a dose-response gradient for omega-3s and CVD; for every 1 g increase in omega-3s tested in the trials, an 8% reduction in CVD was observed. These findings were consistent with the CVD findings in the VITAL trial. But now, while we have not seen the detailed results of STRENGTH, that trial raises questions about whether some of the formulations may not be protective, even at higher doses. There is now a bit of a disconnect because it looked like the dose-response gradient was tremendously important. We will need to look carefully at the final results of STRENGTH.

In our recent meta-analysis, trials such as VITAL, GISSI-P and others suggested that the combination of EPA and DHA had some benefits for coronary outcomes but not stroke. Uniquely, icosapent ethyl reduced risk for stroke in REDUCE-IT, which contributed to a significant reduction in the composite CVD endpoint. We need to better understand why icosapent ethyl was particularly protective, why some of these other formulations have shown limited benefits, and why this latest trial suggests minimal if any benefit.

We now have even more reason to believe that the formulation and composition of the omega-3s are important determinants of their efficacy for CVD prevention.