8 responses to “Icosapent Ethyl to Reduce Atherosclerotic Events in Patients with Hypertriglyceridemia”

  1. Superb review, thank you! I do however have a problem with using the NNT taken directly from clinical trials, for the following reason two reasons:

    Firstly, the NNT taken form the trial results assumes that my patient will be treated for exactly as long as the patients in the clinical trial (5 years in REDUCE-IT). Assuming that the drug effect is maintained (as you imply in your cumulative exposure model), the NNT would in fact decline as treatment is continued.

    Secondly, one of the dogmas in interpretations of clinical trials is that one should not generally pay too much attention to sub-group analyses, except as hypothesis-generating observations. In other words, it should be assumed until proven otherwise that the RELATIVE risk reduction (RRR) is a constant. In any event, there was consistent response (RRR) among groups in REDUCE-IT.

    NNT is the inverse of absolute risk reduction (ARR), which in turn is equal to RRR multiplied by baseline (placebo group) risk. In other words, ARR = RRR * baseline risk.

    In other words, the NNT is as much influenced by the patient as it is by the treatment.

    The NNT derived directly from a clinical trial assumes that the patient sitting in front of me has the same risk as the placebo group, and that is, I believe, a faulty assumption.

    Therefore, when faced with an individual patient, I try to calculate his personal risk (eg Framingham or whatever), and then multiply that by the RRR from the clinical trial. That would be his personal ARR, and the inverse of that number is what I call his NITNT (“number of identical twins needed to treat”).

  2. 2 further pieces of information are needed to put the study results and treatment suggestions in perspective. First, we need to know changes in all cause mortality, not just the change in disease specific mortality. We need all cause mortality as treatments may increase other causes of death even as the cause of interest declines.

    Secondly, we need to know how much longer a treated population lives. A gain of a month or two of life expectancy may not be considered of value to some patients considering the costs of a treatment or its side-effects.

  3. very interesting

  4. Its an excellent article which will change mypractice.

  5. Thanks for the concise review of REDUCE-IT. I will certainly give more consideration to the use of icosapent in the future.

    It’s worth pointing out that 70% of the subjects had established atherosclerotic disease (coronary, carotid, or peripheral) and so the results may not be truly generalizable to those without any documented plaques.

  6. No change in mortality.

    Each of the first two authors has more than 10 lines of disclosures with dozens of entities.
    Five authors disclose being “employed by and being a stock shareholder of Amarin Pharma”
    I don’t know but I’d be surprised if Amarin Pharma does not benefit from this publication.

    Not yet replicated that I know of. I(‘m no expert.)

    Are we sure that’s enough to change practice enthusiastically?

  7. this is a very thought provoking article

  8. I thank you all for the thoughtful comments and discussion. I will attempt to answer these questions.

    Dr. Auersperg provides excellent comments regarding measures of treatment effect. My preference is to report both relative risk and absolute risk. Relative risk has the advantage of comparing risk between two groups (i.e. the treatment and control group). Absolute risk provides an estimate of the clinical relevance of the effect. The number needed to treat (NNT) was introduced as a more intuitive measure of a treatment effect, often when communicating with a patient. After all, a patient’s outcome is binary. It is calculated as the inverse of the ARR and therefore has the same limitations, namely that it depends on baseline risk and treatment duration. In this article, the NNT is reported as a result of an editorial request; however, the ARR (4.8%) is available in the publication(1). It is incumbent on the reader to know the strengths and limitations of each measure.

    Dr. Etches discusses the durability of the treatment effect and the presence of adverse effects. I respectfully disagree with these comments. 1) There is no evidence that the treatment effect is not durable. Figure 1 from the REDUCE-IT publication shows Kaplan-Meier curves that are consistent with countless trials of therapies to reduce atherogenic lipoproteins (i.e. LDL), in that the curves separate at 1-year and continue to separate over time(1). This is consistent with our current model of ASCVD. If anything, the (absolute) effect of the therapy only increases with time. 2) The hazard ratio (95% CI) for all-cause mortality is 0.87 (0.74–1.02) with treatment. I would also caution against using all-cause mortality to exclude important adverse effects(1). It is difficult to power a trial for all-cause mortality, even with an effective therapy. The absence of an increased risk of death should not reassure us that there are not important adverse effects. In the case of EPA, the adverse effect profile is favourable reported in Supplementary Table 5(1).

    Dr. Weiss emphasized that, as described in the text of this article, 71% of patients had established cardiovascular disease (secondary prevention) and 29% had diabetes plus an additional risk factor(1). This study population is generalizable to many practices. Furthermore, subgroup analyses showed no interaction for risk stratum (primary prevention versus secondary prevention).

    Dr. Finucane comments on the financial relationships of this trial and investigators. As with most large randomized trial of new medical therapies, this trial is funded by a pharmaceutical company. The funding source of trials should neither negate or reassure us on the merits of the science – we have to judge the data for what it is. For example, let’s review the major breakthrough life-saving therapies for heart failure with reduced ejection fraction over the past decade. The trials for empagliflozin(2) (EMPEROR-Reduced, NEJM 2020, funded by Boehringer Ingelheim), dapagliflozin(3) (DAPA-HF, NEJM 2019, funded by AstraZeneca), sacubitril/valsartan(4) (PARADIGM-HF, NEJM 2014, funded by Novartis), and eplerenone(5) (EMPHASIS-HF, NEJM 2011, funded by Pfizer) were all funded by pharmaceutical companies. When compared with treatment with beta blocker and ACE inhibitor alone, these therapies will increase the survival of a 55-year-old with HFrEF by a remarkable 6.3 years(6). It is unconscionable to deny a patient the benefits of these therapies based on who funded the trial. Conversely, we must not be falsely reassured by investigator-led, grant-funded studies. Just look at the recent falsification of COVID-19 patient data that resulted in the Surgisphere scandal(7),(8). Clearly, personal and academic conflicts of interest may prevail in the absence of pharmaceutical funding. I implore us all to assess trials based on the merits of their science through a systematic critical appraisal.

    Dr. Finucane also comments on the reproducibility of these results. The findings of REDUCE-IT build on the JELIS trial, which compared EPA with statin versus statin alone (there was no placebo group)(9). They are supported by the findings of the subsequent EVAPORATE trial, which showed that EPA results in a 17% reduction in low-attenuation coronary artery plaque volume at 18-months of therapy(10). We have no reason to believe that the findings of REDUCE-IT will be further replicated but, as always, remain vigilant. For the record, I have no disclosures and conflicts of interest.

    I thank Dr. Madoonik, Dr. Bhuiyan, and Dr. Puts for their interest and kind works.


    1. Patel PN, Patel SM, Bhatt DL: Cardiovascular risk reduction with icosapent ethyl. Curr Opin Cardiol 2019; 34:721–727.
    2. Packer M, Anker SD, Butler J, et al.: Cardiovascular and Renal Outcomes with Empagliflozin in Heart Failure. N Engl J Med [Internet] 2020; :1–12. Available from: http://www.ncbi.nlm.nih.gov/pubmed/32865377
    3. McMurray JJV, Solomon SD, Inzucchi SE, et al.: Dapagliflozin in Patients with Heart Failure and Reduced Ejection Fraction. N Engl J Med Massachusetts Medical Society, 2019; .
    4. McMurray JJV, Packer M, Desai AS, et al.: Angiotensin–Neprilysin Inhibition versus Enalapril in Heart Failure. N Engl J Med 2014; 371:993–1004.
    5. Zannad F, McMurray JJV, Krum H, et al.: Eplerenone in Patients with Systolic Heart Failure and Mild Symptoms. N Engl J Med 2011; 364:11–21.
    6. Vaduganathan M, Claggett BL, Jhund PS, et al.: Estimating lifetime benefits of comprehensive disease-modifying pharmacological therapies in patients with heart failure with reduced ejection fraction: a comparative analysis of three randomised controlled trials. Lancet 2020; 396:121–128.
    7. Mehra MR, Ruschitzka F, Patel AN: Retraction—Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis. Lancet [Internet] Elsevier, 2020; 395:1820. Available from: https://doi.org/10.1016/S0140-6736(20)31324-6
    8. Mehra MR, Desai SS, Kuy S, Henry TD, Patel AN: Retraction – Cardiovascular Disease, Drug Therapy, and Mortality in Covid-19. N Engl J Med. DOI: 10.1056/NEJMoa2007621. N Engl J Med [Internet] Massachusetts Medical Society, 2020; 382:2582. Available from:
    9. Yokoyama M, Origasa H, Matsuzaki M, et al.: Effects of eicosapentaenoic acid on major coronary events in hypercholesterolaemic patients (JELIS): a randomised open-label, blinded endpoint analysis. Lancet 2007; 369:1090–1098.
    10. Budoff MJ, Bhatt DL, Kinninger A, et al.: Effect of icosapent ethyl on progression of coronary atherosclerosis in patients with elevated triglycerides on statin therapy: final results of the EVAPORATE trial. Eur Heart J [Internet] Oxford University Press (OUP), 2020; . Available from:

Leave a Reply