Eskimos, Fish Oils and EPA Deniers


Eskimos: More than half a century ago, Danish researchers Bang and Dyerberg made the astute discovery that Eskimos residing in the Danish territory of Greenland experienced a much lower rate of coronary heart disease than Europeans living in Denmark.  As Eskimos consumed high amounts of marine derived omega-3 fatty acids from whale blubber (see photo above), salmon and other fish, they reasoned that the low incidence of myocardial infarction was due to a diet enriched in EPA. 

As it turns out, Bang and Dyerberg were right! 

With the results of RESPECT-EPA presented at last week’s American Heart Association Scientific Sessions, we now have 4 clinical outcome trials and 2 arteriographic studies pointing in the same direction; simply stated, purified EPA reduces coronary (CAD) events, arteriographic CAD progression and overall cardiovascular risk.  Whether studies included patients in primary prevention, secondary prevention or post-ACS, they have generated reproducible results in favor of EPA therapy. In effect, icosapent ethyl the purified form of EPA (marketed in the U.S. as Vascepa, the abbreviated form of Vascular EPA) is currently indicated for patients with high triglycerides (HTG) and CAD or for Type 2 diabetes mellitus (T2DM) with additional CAD risk factors.    Now admittedly, RESPECT-EPA had shortcomings that included an open-label design and was underpowered (low event rates/high early withdrawal rates) that would not have garnered significant attention despite the 22% reduction in the primary endpoint (cardiovascular [CV] death, nonfatal myocardial infarction, nonfatal stroke, revascularization and unstable angina requiring hospitalization) (p=0.054) and 27% reduction in the secondary endpoint (CAD events, stroke and mortality) (P=0.03) in Japanese patients with stable CAD assigned to 1.8 grams of EPA daily. However, when RESPECT-EPA is placed within the context of other EPA based studies, there is a clear and consistent trend of improved cardiovascular risk.  These results are directionally aligned with the much larger REDUCE-IT trial that used 4 grams daily of highly purified icosapent ethyl (IPE) in men and women with elevated triglycerides and CAD or T2DM that resulted in 25% and 26% reductions in the primary and secondary endpoints, respectively (P< 0.0001; P< 0.0001).  In the earlier JELIS study of nearly 20,000 Japanese patients with elevated cholesterol there was an overall 19% reduction in risk of coronary events (p=0.011).  Finally in the study by Nosaka and colleagues, assignment to 1.8 grams EPA within 24 hours following percutaneous coronary intervention in the setting of an acute coronary syndrome was associated with a 58% reduction in cardiac events during the 1-year follow-up period (P=0.02).  Two angiographic studies CHERRY and EVAPORATE further demonstrated reduced atherosclerotic progression when EPA was added to statin therapy.   

Taken together, ALL 6 EPA trials demonstrate improvement in cardiovascular risk.

Fish Oils: In contrast to EPA only studies, the use of fish oils, most commonly a combination of EPA/DHA has not been successful in lowering CV risk. Some have suggested that the daily dose of fish oils used was too low (e.g., 1 gram daily), thereby limiting the rise of circulating EPA levels (that many believe to be the primary basis for the CV benefits)  and/or did not test a sufficiently high-risk study population.  Nonetheless, the STRENGTH study that used 4 grams of fish oil daily in a high-risk population (HTG and CAD/CAD risk factors) also failed to demonstrate benefit.  So why have EPA-only but not EPA/DHA studies improve cardiovascular risk? In this regard, basic science investigations led by my colleague, Dr. Preston Mason has repeatedly shown that DHA offsets favorable actions of EPA (e.g., cellular inflammation, oxidative stress, etc).  Does this mean that we should recommend avoiding DHA containing foods?  That would not be my recommendation because DHA is 1) vital for brain growth and development (most pronounced during the first 5 years of life) and 2) cardioprotective marine-derived fish contain both EPA/DHA. Thus, while we don’t recommend discontinuing fish consumption, we do  recommend purified EPA for our high-risk patients based upon REDUCE-IT inclusion criteria.

Supplements: The most common dietary supplements in the U.S. are fish oil derived. Unfortunately, while the FDA regulates over-the-counter (OTC) drugs, they do not regulate dietary supplements because they are considered food derived products.  Not only are fish oil supplements ineffective in reducing cardiovascular risk but they may also contain contaminants such as atherogenic saturated fatty acids.  Hence, we do not recommend fish oil supplements and would be very wary of any physician/health care professional who does…even if/especially if they run for Congress!

EPA Deniers: EPA deniers have found a variety of reasons to minimize the fact that purified EPA is highly effective.  For one, they point to the open label design (as is customary in Japan) even though both REDUCE-IT and EVAPORATE drew similar/positive conclusions using placebo-controlled groups.  Then they will complain that the CV benefit was not due to EPA per se, but rather the mineral oil placebo group that inflicted damage and accelerated CVD events due to LDL-C and hs-CRP elevations.  Surely if this were the case, placebo-treated subjects with the highest on-treatment LDL-C or hs-CRP levels would have markedly greater CVD event rates than placebo treated patients without such elevation.  However, that line of reasoning was faulty because there were no differences in CVD events irrespective of on-treatment LDL-C or hsCRP in the mineral oil group. In fact, IPE reduced CVD events similarly in these groups.  Finally, they will extol that mineral oil caused such severe increases in inflammatory biomarkers that that must have been the reason for the success of REDUCE-IT.  Yes, it is true that the mineral oil group evidenced increases in biomarkers that included a 28.9% increase in interleukin-1β (0.06 to 0.08 pg/ml) and 16.2% increase in interleukin-6 (3.27 to 3.76 pg/ml) after 1 year.  On the surface this appears to be problematic until the normal range is revealed: interleukin-1β (0.5-12 pg/ml), interleukin-6 (0-43.5 pg/ml). In other words, while there were statistically significant increases in these levels, they were still within the normal range! One is certainly hard-pressed in this case to equate statistical significance with clinical relevance.  Hence, the modest effects of mineral oil can in no way account for the sizeable benefits observed in REDUCE-IT, a fact corroborated by the FDA upon issuing their approval indications for IPE in 2019 following unanimous agreement of the scientific advisory team.   

Moving forward, let’s direct our focus towards optimizing lipid lowering treatment for our patients at highest CVD risk.  The proof is in the pudding when it comes to LDL-C lowering (statins/ezetimibe/PCSK9 inhibitors) and TG lowering (icosapent ethyl) and our patients should receive these therapies as indicated.  We await novel therapies currently in clinical trial testing that may provide incremental risk reduction to benefit our patients.

Michael Miller, MD is a cardiologist and Chief of Medicine at the Corporal Michael J Crescenz VA Medical Center and Vice Chair of Medicine at the Hospital of the University of Pennsylvania. Dr. Miller is Scientific Advisor to Amarin, Inc and 89bio. Prior to his arrival in Philadelphia, he was Professor of Cardiovascular Medicine, Epidemiology and Public Health at the University of Maryland School of Medicine. He is the author of numerous scientific publications and several books, including Heal Your Heart published by Penguin Random House. Check him out on twitter: @mmillermd1