ROCKET AF Investigators: Problems With Warfarin Monitoring Device Did Not Skew Results
Potential problems with the point-of-care device used to monitor warfarin in ROCKET AF did not appear to have a meaningful impact on the trial results, an analysis by study investigators shows. The findings come in response to questions about the validity of the trial’s main results, an issue that has also been connected to Robert Califf, MD, the nominee for FDA commissioner.
Califf, then at Duke University, was co-chair of the executive steering committee for ROCKET AF, which showed that rivaroxaban (Xarelto; Bayer/Janssen Pharmaceuticals) was noninferior to warfarin for preventing stroke or systemic embolism in patients with nonvalvular A-fib.
In December 2014, roughly 4 years after the main trial results were reported and 3 years after rivaroxaban was approved for stroke prevention, testing device manufacturer Alere issued an alert about its INRatio and INRatio2 PT/INR Monitor System, which was used to monitor anticoagulation levels in warfarin-treated patients in ROCKET AF.
The company informed users in its medical device correction that patients with certain conditions—including anemia with a hematocrit less than 30%, unusual bleeding or bruising, and conditions associated with elevated fibrinogen levels—should not be tested with the system because of the possibility of obtaining INR results lower than would be expected with an automated, plasma-based laboratory INR test. The announcement, which came in the form of an FDA recall notice, also noted that the instructions for optimal use recommend using the test in patients with a hematocrit of 30% to 55%.
The concern is that a low INR reading could prompt an inappropriately high warfarin dose and, in turn, induce excess bleeding. This would inadvertently lead to rivaroxaban appearing safer than warfarin in the ROCKET AF study.
The possibility of skewed INR findings has focused attention on whether the trial results can still be considered valid and has put heat on the ROCKET AF leaders, who say they were made aware of the issue in October 2015. Califf, in particular, has been affected by the nagging questions as he moves through the nomination process. Both the FDA and the European Medicines Agency have said that they are reviewing data related to the potential impact of the device issues on the ROCKET AF results.
After members of the trial’s executive committee learned about the potential problem, they initiated a series of post hoc analyses, the results of which were published online this week in a research letter in the New England Journal of Medicine.
Results Match Those of Overall Trial
Manesh Patel, MD, of the Duke Clinical Research Institute (Durham, NC), and colleagues performed efficacy and safety analyses in subgroups of patients with and without any of the conditions listed in the device recall notice. Also included in the recall group were patients with hematocrit values outside of the recommended range. Of the 14,236 patients in ROCKET AF, 63% did not have any recall condition and 37% had at least 1.
In the subgroup without recall conditions, results were consistent with the overall trial findings. Rivaroxaban was noninferior to warfarin for preventing stroke or systemic embolism, with similar rates of overall bleeding in both groups. Rivaroxaban carried lower risks of fatal and intracranial bleeding but a higher risk of GI bleeding.
Results also were generally consistent in the patients with recall conditions, although there was a nonsignificant trend toward a higher major bleeding risk with rivaroxaban vs warfarin; this trend was not seen in patients without recall conditions (P = .04 for interaction).
Patel told TCTMD that the finding is likely due to chance considering the number of tests that were performed. “But nevertheless,” he added, “that finding goes against the hypothesis that the device was leading to a bleeding risk.”
Other analyses are ongoing, but these initial ones most directly address the concerns that have been raised, Patel said.
“We take the integrity of the trial and the findings very seriously and as soon as we learned about [the issue] we were able to at least do this first wave of analyses, which have comforted us and give us at least some support to tell our investigators, our patients, and our colleagues that we think the trial results remain consistent,” he said.
In a separate prepared statement, the members of the executive committee concluded that “the re-analysis found that any possible malfunction of the device that could have led to lower INR measures than laboratory testing did not have any significant clinical effect on the primary efficacy and safety outcome measures recorded in the ROCKET AF trial.”
Patel MR, Hellkamp AS, Fox KAA. Point-of-care warfarin monitoring in the ROCKET AF trial [correspondence]. N Engl J Med. 2016;Epub ahead of print.
- The analysis was supported by the Duke Clinical Research Institute.
- Patel reports receiving grant support from Janssen and Johnson & Johnson and personal fees from Bayer and Janssen during the conduct of the study. He also reports receiving grant support from AstraZeneca, CSI, Heartflow, Maquet, Medtronic, and the National Heart, Lung, and Blood Institute and personal fees from AstraZeneca, CSI, Genzyme, Medtronic, and Merck unrelated to this study.