In Reply The letter from Dr Olshansky and colleagues proposes several hypotheses to explain the different outcomes in the REDUCE-IT and STRENGTH trials. They suggest that the greater content of EPA (4 g/d) administered in REDUCE-IT produced superior results but also suggest that the results were replicated in JELIS, an open-label study that used only 1.8 g/d. If EPA dosage is the critical issue, then the 4 g/d administered in STRENGTH should have produced some benefit. These authors suggest a threshold effect, but for the top tertile of change in EPA in STRENGTH (>435% increase), there was no benefit. With respect to mineral oil, the best evidence for toxicity comes from REDUCE-IT, in which mineral oil treatment increased high-sensitivity C-reactive protein (hs-CRP) by 32%, low-density lipoprotein cholesterol (LDL-C) by 10.9%, and non–HDL-C by 12% after 12 months. In STRENGTH, corn oil decreased hs-CRP by 6.3% and both LDL-C and non–HDL-C by 1.1%. Both the CHERRY and EVAPORATE trials were very small, with only 97 and 31 treated patients. EVAPORATE was widely criticized for showing large differences between treatment groups that were not biologically plausible. This trial further supported the conclusion that mineral oil is toxic, showing a 109% increase in low-attenuation plaque and a 32% increase in fibrofatty plaque in the mineral oil treatment group. It would be tragic if patients receive an ineffective drug based on a false-positive trial due to a toxic placebo.
