From the Journals

Findings of most heart failure trials reported late or not at all


 

FROM THE JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY

A large proportion of results from heart failure trials registered with clinicaltrials.gov are published a year or more after completion or not at all, which violates the U.S. FDA Amendments Act (FDAAA), according to a detailed analysis of the interventional and observational trials in this database.

Dr. Christopher M. O'Connor, president, Inova Heart and Vascular Institute, Falls Church, Va.

Dr. Christopher M. O'Connor

Of the 1,429 heart failure trials identified, 75% of which were randomized interventional studies and the remainder of which were observational, fewer than 20% met the FDAAA 1-year reporting requirement, and 44% have yet to be published at all, reported a team of collaborative investigators led by cardiologists from the Inova Heart and Vascular Institute (IHVI), Falls Church, Va.

“I believe the critical issue is that the FDAAA has thus far never been enforced,” reported Christopher M. O’Connor, MD, a cardiologist and president of IHVI. He was the senior author of the study, reported in the Journal of the American College of Cardiology.

To improve systematic reporting of clinical trials, including negative results, clinicaltrials.gov was created in 2000. In 2007, the FDAAA enacted rules to broaden the requirements for reporting and to make timely reporting of results mandatory.

Ten years later, the FDA was finally authorized to issue a penalty of $10,000 for failure to release results in a timely fashion, a provision of the 2007 amendment but not confirmed at that time, the investigators reported. In the majority of cases, timely reporting was defined as within 12 months of completion of the trial.

The new study shows that reporting of completed trials, timely or otherwise, remains low. Of the 1,243 trials completed after 2007, the proportion meeting the 1-year reporting requirement was just 20%. Although a significant improvement over the 13% reporting in this time frame before 2007, more than 80% of findings are not being released in a timely manner more than 10 years after this was made mandatory.

There are a number of reasons to consider this to be a serious issue, according to Mandeep R. Mehra, MD, of Brigham and Women’s Hospital, Boston. One of the authors of an accompanying editorial regarding this analysis, Dr. Mehra called underreporting “a public health matter because it is an impediment to medical discovery and poses plausible threats to patient safety.”

Among studies registered after 2007, publication rates were higher for trials funded by the National Institutes of Health (71%) relative to industry (49%) or the U.S. Veterans Affairs (45%).

Publication rates were also higher among interventional relative to observational trials (59% vs. 46%) and trials that enrolled more than 1,000 patients relative to those enrolling fewer than 150 (77% vs. 51%), although trial size was not a significant predictor of publication on multivariate analysis. Clinical endpoints, such as death or hospitalization, were also associated with a greater likelihood of publication relative to nonclinical endpoints.

Of the 251 trials terminated before completion, findings were published within 1 year in only 6%. Two years after completion, only 20% were published at all.

Results consistent with the primary hypothesis did not predict timely publication, but only 39% of the studies listed a primary hypothesis. Since 2017, this is another violation of the FDAAA, according to Dr. O’Connor.

The problem is not unique to heart failure trials, according to the authors who cited numerous studies showing low rates of timely publication in other therapeutic areas. Heart failure was selected for evaluation in this study mainly to keep the analysis feasible, although the authors contend this is an area with an urgent need for better treatments.

The problem needs to be fixed, according to Dr. Mehra. In his editorial, he called for rules to be “transitioned to regulations and action taken for underreporting.” Dr. O’Connor agreed.

“A combination of carrots and sticks might be needed to achieve sufficient result sharing,” Dr. O’Connor said. He suggested that stakeholders, such as investigators, sponsors, regulators, and journal editors, should collaborate to address the problem.

So far, the FDA has never levied a fine for lack of reporting or for failure to report in a timely manner. Routine imposition of large fines might not be viable, given the complex reasons that delay or inhibit publication of trial findings, but it would be a large source of revenue.

“According to the FDAAA TrialsTracker, a live tool that tracks FDAAA compliance and promotes trial transparency, the U.S. government could already have imposed more than $2.8 billion in fines for trials due after January 2018,” Dr. O’Connor reported.

The first and senior authors are among those who report financial relationships with pharmaceutical companies.

SOURCE: Psotka MA et al. J Am Coll Cardiol. 2020;75:3151-61.

Next Article: