Icosapent Ethyl for Residual Risk: The Comparator Arm That Wasn't Inert
Two years after a stented NSTEMI, his LDL is controlled and his triglycerides are not. The drug that addresses them was tested against a placebo that turned out not to be neutral, and the debate is about how much of that 25% still belongs to the drug.
J.B., a 58-year-old man, has taught high-school shop class for over twenty years, and for most of that time his wife handled the cooking. Since their divorce finalized this spring he has been teaching himself to cook, which is how a follow-up lipid panel ordered mostly out of habit turned into a longer conversation than either he or his cardiologist expected. Two years ago he had an NSTEMI, was found to have single-vessel disease in the right coronary artery, and received a drug-eluting stent. He has taken rosuvastatin 40 mg since, walks most evenings, and his LDL-C has held at 62 mg/dL for over a year — by the numbers his lipid management looks finished.
It isn't. His triglycerides have sat in the 220–260 mg/dL range across three separate draws despite the statin, the walking, and a diet that has actually improved since he started cooking for himself. That combination — established ASCVD, LDL-C already at goal, triglycerides persistently elevated on a maximally tolerated statin — is exactly the population REDUCE-IT enrolled, and exactly the population in which icosapent ethyl showed a 25% reduction in major cardiovascular events. What complicates recommending it isn't his numbers; it's what the trial's own placebo turned out to be. REDUCE-IT used mineral oil as the comparator, and mineral oil modestly raised LDL-C and inflammatory markers in the placebo arm over the trial's several years — raising a real question of how much of the observed 25% reflects icosapent ethyl helping versus the comparator arm quietly performing worse than a truly inert placebo would have. Whether that confound explains a meaningful fraction of the result, or only a marginal one, is not something his numbers alone can answer.
At the lipid follow-up
This is the REDUCE-IT population exactly — established ASCVD, LDL-C at goal on a maximally tolerated statin, triglycerides persistently between 135 and 499. A 25% relative risk reduction in major cardiovascular events is not a marginal signal, and the 2021 ACC Expert Consensus Decision Pathway on persistent hypertriglyceridemia states that adding icosapent ethyl may be reasonable in this exact profile. I'd start 2 g twice daily.
I want to be honest about what the placebo arm actually was before we treat 25% as a clean number. Mineral oil isn't biologically inert — it modestly raised LDL-C and hs-CRP in that arm over the trial, which means some of the separation between arms could reflect the comparator doing worse rather than icosapent ethyl doing better.
The comparison I keep coming back to is STRENGTH, which used a corn-oil placebo — genuinely closer to inert — and tested a related omega-3 formulation. STRENGTH was neutral. That doesn't prove REDUCE-IT's result is an artifact, but it means we can't treat the 25% as settled the way we'd treat a result from a trial with an uncontroversial comparator.
I don't think the placebo question resolves cleanly either way, and I don't think it has to for us to act. The post hoc analysis by baseline LDL-C showed benefit regardless of how low his LDL already was, which is at least consistent with a triglyceride- or inflammation-mediated mechanism distinct from anything statins are doing. Given his profile and the absence of a real alternative with comparable outcome data, I'd start icosapent ethyl 2 g twice daily and continue rosuvastatin unchanged — while telling him plainly that the size of the benefit, not its existence, is the part still genuinely argued about.
Started icosapent ethyl 2 g twice daily in addition to unchanged rosuvastatin 40 mg. Repeat lipid panel and tolerability check in 3 months.
Not agreed, and stated to him directly rather than smoothed over:
A meaningful, drug-attributable reduction in his recurrent-event risk, independent of further LDL-lowering.
A real but smaller benefit than 25% — still probably worth the modest pill burden, but not the number usually quoted to patients.
He was told both readings are defensible from the same trial, and that his own decision to start the drug didn't require the argument to be settled first.