STRENGTH Rough estimate of likely upper bound on performance
Looking for a quick rough estimate of the maximum likely RRR we should expect from the STRENGTH trial of Epanova. Epanova capsules are a gram each, at least 850 mg of which is PUFA. They are 50-60% EPA and 15-25% DHA. We’ll look at the 4 gram dose (I believe they study 2 and 4). The trial specifies a low cholesterol diet, and subjects are on statin, so presumably LDL is well controlled.
To give a best case estimate (for Epanova, worst case from Amarin’s perspective), ignore the likely LDL-C increase from the DHA and any negative effects of extra material and assume DHA has 1/3 the benefit of EPA (due to good trig reduction benefit but lacking many other benefits of EPA). So, 4 Epanova capsules have 4x0.60=2.4g of EPA and 4x0.25=1g of DHA which we’ll call equivalent to a total of 2.4+0.33=2.73g of EPA. Vascepa is 96% EPA, so a dose is 3.84g of EPA, and the Epanova dose has 0.711 as much EPA as the V dose.
Next, we’ll consider the difference in trial duration. From the RI and JELIS trials it is reasonable to assume a startup delay of between 1 and 2 years before full benefits, let’s conservatively call it 1.1 years. If I recall correctly STRENGTH duration was expected to be 3.5 years but they may be running a bit long, so let’s call it 3.7 years. RI was 4.9 years. So, STRENGTH had a startup degradation factor of (3.7 – 1.1)/3.7 = 0.703. V’s was 0.776, so the ratio is 0.703/0.776 = 0.906. Assuming local linearity of RRR vs EPA dose (which is reasonable based on JELIS and RI), with RI reporting 38% RRR for the high trig, low HDL subgroup, our estimate of maximum likely STRENGTH RRR = 0.711x0.906x38% = 24%. Allowing another 2% for statistical aberration (with possible placebo effect), I would say that STRENGTH seems fairly unlikely to exceed 26%, and very likely to be under 30%, although no one ever knows for sure with clinical trials.
If we do a best guess estimate instead of a likely upper bound (using 20% DHA, 55% EPA, and duration 3.6 years for Epanova, 1.25 years for startup delay, and DHA benefit 0.3 as much as EPA) that gives 21% (not adjusted for active placebo).
BOTTOM LINE: A rough guess at STRENGTH MACE RRR is about 21% after adjusting for EPA dose and trial duration. It seems fairly unlikely to exceed 26% and very unlikely to exceed 30%, which is well under Vascepa’s 38% MACE RRR for the high trig/low HDL subgroup.
The corn oil placebo is active and will likely exaggerate performance at least a little. I don't know how to predict the impact, this analysis focuses on the more predictable dose and duration effects. Placebo impact will be estimable after the trial from the K-M curves.
Market Data 
Markets 
AI
