‘Massive loss’ of patients
Critiquing the study, Dr. Nissen highlighted the large dropout and protocol violation rate.
“There was a massive loss of patients over the 6- to 8-year follow-up, and the Kaplan-Meier curves didn’t start to diverge until after 4 years, by which time many patients had dropped out. It would have been a very selective population that lasted 6 years in the study. Patients that drop out are different to those that stay in, so they are cherry-picking the patients that persist in the trial. There is enormous bias here,” he commented.
“Another weakness is the open-label design. Everyone knew who is getting what. Blinding is important in a study. And there was no control treatment in this trial,” he noted.
The researchers also selected patients with low EPA levels at baseline, Dr. Nissen added. “That is completely different hypothesis to what was tested in the REDUCE-IT and STRENGTH trials. And even with all these problems, the results are still statistically insignificant.”
On the post hoc subgroup analysis showing a significant benefit, Dr. Nissen said, “they compared a subgroup in the active treatment arm who had large increases in EPA to a subgroup of control patients who had the smallest increase in EPA. That would be like comparing patients who had the largest reductions in LDL in a statin trial to those in the control arm who had no reductions or increases in LDL. That’s scientifically totally inappropriate.”
Supportive data
But Dr. Bhatt argues that the RESPECT-EPA trial supports the two previous trials showing benefits of EPA.
“Some may quibble with the P value, but to me this study has shown clear results, with obvious separation of the Kaplan-Meier curves,” he said.
“It is an investigator-initiated study, which is good in principle but has some of the usual caveats of such a study in that – probably as a consequence of budget constraints – it has an open-label design and is underpowered. But as they did not use a placebo and still showed a benefit of EPA, that helps resolve the issue of the placebo used in REDUCE-IT for those who were concerned about it,” Dr. Bhatt noted.
He pointed out that the 1,800-mg dose of EPA is the same dose used in the JELIS trial and is the dose used in Japan. The REDUCE-IT trial used a higher dose (4 g), but in general, Japanese people have higher levels of EPA than Western populations, he explained.
“While this trial included patients with lower levels of EPA, what is considered low in Japan is much higher than average American levels,” he added.
Magnitude of benefit uncertain?
Discussant of the study at the Late Breaking Clinical Trials session, Pam R. Taub, MD, professor of medicine at the University of California, San Diego School of Medicine, said, “Despite being underpowered with a sample size of 2,460, RESPECT-EPA shows benefit in decreasing composite coronary events.”
“There is benefit with EPA, but the magnitude of benefit is uncertain,” she stated.
Dr. Taub pointed out that there is a signal across studies for new-onset atrial fibrillation, but the absolute increase is “rather small.”
She noted that more mechanistic and clinical data are needed to hone in on which patients will derive the most benefit, such as those with elevated high-sensitivity C-reactive protein or highest change in EPA levels. But she concluded that in clinical practice, physicians could consider addition of EPA for reduction of residual risk in secondary prevention patients.
The RESPECT-EPA study was supported by the Japan Heart Foundation. Dr. Daida reports peakers’ bureau/honorarium fees from Novartis Pharma, Bayer Yakuhin, Sanofi, Kowa Company, Taisho Pharmaceutical, Abbott Medical Japan, Otsuka Pharmaceutical, Amgen, MSD, Daiichi Sankyo, Pfizer Japan, FUKUDA DENSHI, Tsumura, and TOA EIYO and research funding from Philips Japan, FUJIFILM Holdings, Asahi Kasei, Inter Reha, TOHO HOLDINGS, GLORY, BMS, Abbott Japan, and Boehringer Ingelheim Japan.
A version of this article first appeared on Medscape.com.