CardioExchange Archive

Thirteen years after first being presented, the results of the PRAISE-2 trial finally have been published in JACC: Heart Failure. The trial itself is now largely irrelevant to current clinical practice, as the hypothesis it tested has long been abandoned, but the long delay in publication may serve to bring even more awareness to the issue of the delay or complete absence of publication of many clinical trials.

An accompanying editorial, by Marc Pfeffer and Hicham Skali, is highly critical of the delay:

Although standards for conduct and reporting of clinical trials have improved since 2000, the failure to fully vet the results of a clinical trial of human volunteers in a peer-reviewed journal was and remains unacceptable.”

PRAISE-2 had its origins in the first PRAISE trial, which was first presented in 1995 and subsequently published in the New England Journal of Medicine in 1996. In that trial there was no difference between amlodipine (Norvasc, Pfizer) and placebo in the rate of mortality or cardiovascular hospitalization in patients with heart failure. However, a prespecified subgroup analysis turned up the highly surprising result that heart failure patients with a nonischemic etiology who received amlodipine had a highly significant 46% reduction in the risk of death compared with placebo patients.

This subgroup analysis was the basis for PRAISE-2, which had a similar design to PRAISE but only included patients with nonischemic heart failure. As first reported in 2000, and now in JACC Heart Failure, there were no significant differences between the groups in PRAISE-2. The trial therefore failed to confirm the promising observation from PRAISE. Citing additional examples of this phenomenon, Pfeffer and Skali write that PRAISE-2 is a

…prominent example in cardiovascular medicine in which a significant p value for an important irrefutable endpoint such as all-cause mortality from a secondary analysis was not confirmed by a more definitive randomized controlled trial directly testing the hypothesis generated from the initial study…”

The editorialists praise the PRAISE investigators for “promptly conducting the definitive trial to test the hypothesis generated from the subgroup analysis from the first PRAISE trial” but go on to write that “equally deserved is the criticism regarding the 13-year incubation between the completion of the trial with neutral results” in 2000 and the “long overdue first publication of the results.”

It should be noted that Milton Packer, the principal investigator of both PRAISE studies, was also the PI for the REVIVE trials, the results of which were published in April, also in  JACC Heart Failure, after a 7-year delay. REVIVE found that although intravenous levosimendan provided symptomatic relief to patients with acute decompensated heart failure, it was associated with an increased risk of adverse cardiovascular events.

In an accompanying editorial about REVIVE, Robert Califf writes that the cardiovascular community “should rejoice in the fact that the REVIVE investigators have finally decided to come out of the data cellar.” He discusses an earlier published subgroup analysis from REVIVE that “advanced the claim that levosimendan is cost effective compared with standard care.” He goes on to comment:

The publication of a non–pre-specified subgroup analysis without first making available the full trial results in a peer-reviewed venue constitutes an example of a publication fostered by commercial interests without appropriate academic participation.”

I asked Packer to respond to the two editorials. He pointed out that he “could have avoided the ‘criticism’ by not publishing at all. But,” he continued:

in reality, it really isn’t criticism; both are simply trying to make a good point. Regarding the reason for the delay, I have no excuse. The usual stuff. There is always something else to do.”