Readjudication Supports Original RECORD Results

Miriam E. Tucker

August 26, 2013

Results from a readjudication of the Rosiglitazone Evaluated for Cardiovascular Outcomes and Regulation of Glycemia in Diabetes (RECORD) trial, supporting the original trial conclusion that the diabetes drug rosiglitazone does not increase the risk for adverse cardiovascular outcomes, have now been published in full.

A summary of the analysis played a key role in the deliberations of a Food and Drug Administration (FDA) advisory committee back in June; the panel ultimately voted to recommend that the current restrictions on rosiglitazone prescribing be eased or lifted entirely.

The full analysis is published in the August issue of the American Heart Journal, by Kenneth W. Mahaffey, MD, from Duke University, Durham, North Carolina, and colleagues.

Following a previous review of rosiglitazone in 2010, the FDA had required the drug's manufacturer, GlaxoSmithKline, to commission a reevaluation of all RECORD deaths and major CV end points, including MIs and strokes. The company hired the Duke Clinical Research Institute (DCRI) to perform the task and supplied the team with all raw patient-level trial data.

In the analysis reported by Dr. Mahaffey and colleagues, total follow-up for mortality was 25,833 person-years, including an additional 328 person-years identified during the readjudication. A total of 184 deaths of CV or unknown cause, 137 MIs, and 119 strokes were identified in the readjudication using the same end points as were used in the original RECORD trial.

The primary end point — time to first occurrence of death due to CV or unknown cause, MI, or stroke — did not differ significantly between rosiglitazone and metformin/sulfonylurea. The hazard ratio for the readjudication was 0.95, similar to the 0.93 found in the original RECORD trial analysis.

Comparisons for the individual end points of mortality, MI, and stroke were also similarly nonsignificant in the readjudication as in the original trial results.

Additional analysis using new FDA end-point definitions did not change the results significantly (0.95 vs 0.97 for the old vs new end points). The new FDA definitions resulted in a small number of additional stroke and MI events, primarily because they did not require hospitalization.

A series of sensitivity analyses attempting to account for potential bias also did not show significant difference from the original trial results, Dr. Mahaffey and colleagues note.

The authors caution that the number of additional patient-years available for the readjudication were limited due to logistical problems (eg, some of the participating centers had shut down). On the other hand, deaths in the readjudication were more likely to be categorized as "unknown" rather than "non-CV" when insufficient information was available.

The authors also cite the limitations imposed by RECORD's original open-label design. That and the problem of missing data were concerns that had been discussed at length during the FDA advisory committee hearing.

The study was funded by GlaxoSmithKline. Dr. Mahaffey has received research grants from Amylin and Bristol-Myers Squibb and research grants and consultant fees from GlaxoSmithKline, Johnson & Johnson, Merck, and Eli Lilly. Disclosures for the coauthors are listed in the article.

Am Heart J. 2013;166:240-249.e1. Full text


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: