Bristol-Myers Squibb and Pfizer Announce Publication of ARISTOTLE Subanalysis in Circulation

Bristol-Myers Squibb Company and Pfizer Inc. announced that results from a prespecified subanalysis of the ARISTOTLE trial were published in Circulation, the peer-reviewed journal of the American Heart Association. Results from this subanalysis showed that the reductions in stroke or systemic embolism, number of major bleeding events and mortality demonstrated with Eliquis® (apixaban) compared to warfarin in the ARISTOTLE trial were consistent across subgroups defined based on levels of International Normalized Ratio (INR) control in patients with nonvalvular atrial fibrillation.

“Concerning the quality of warfarin treatment, there is a large variation in time in therapeutic range among different countries and centers, which affects outcomes. This subanalysis was conducted to determine whether the treatment effects of apixaban were similar in centers and patients with high quality warfarin care,” said study lead author Dr. Lars Wallentin of Uppsala University in Uppsala, Sweden. “These additional analyses supported that the primary results of ARISTOTLE were consistent across a broad range of quality of warfarin management.”

Variations in time in therapeutic range (TTR) can affect outcomes for atrial fibrillation patients being treated with Vitamin K antagonists such as warfarin for stroke prevention, leading to an increased risk of stroke when INR levels are below, or bleeding when INR levels are above, the therapeutic range. For patients in the ARISTOTLE trial, the quality of warfarin management was defined by TTR, with a target INR of 2.0 – 3.0. In the ARISTOTLE trial, patients in the warfarin group had an INR in the therapeutic range (2.0 to 3.0) for a median of 66.0% of the time. For context, the median time for INR in the therapeutic range varies across the globe. In clinical practice settings in the United States, it is approximately 57-59%.

The ARISTOTLE trial randomized 18,201 patients from 1,034 clinical centers in 39 countries. In this subanalysis, for each patient, a center average TTR (cTTR) was estimated using a linear mixed model based on the real TTRs in warfarin treated patients with a fixed effect for country and random effect for center. Study centers were placed into one of four similarly sized quartile groups based on cTTR (<60.5%; 60.6%-66.3%; 66.4%-71.1%; and  >71.2%). The rates of stroke or systemic embolism, major bleeding and mortality were consistently lower with Eliquis than warfarin across the cTTR quartiles. Similar results were seen when an individual TTR (iTTR), predicted using a model including patient characteristics, was examined in a post-hoc analysis.

While demonstrating consistency across a broad range of warfarin control, results of this subanalysis suggest a trend toward reduction of the treatment effects at centers and in patients with predicted excellent INR control. In these centers, interaction tests are less reliable because of low numbers of events, and thereby lack statistical power.

Based on the results of the subanalysis, the benefits of Eliquis compared with warfarin for stroke or systemic embolism, bleeding, and mortality appear similar across the range of centers’ and patients’ quality of INR control.