Apixaban Has Edge Over Other NOACs in Large, Real-World Analysis

ORLANDO, FL—Apixaban beat out two of its non-vitamin K antagonist oral anticoagulant (NOAC) competitors in a large, manufacturer-funded analysis of US claims data.

In patients with nonvalvular A-fib, “apixaban use was associated with a lower risk of stroke/systemic embolism, major bleeding, and net clinical outcome in comparison to dabigatran and rivaroxaban,” said Steven Deitelzweig, MD (Ochsner Health System, New Orleans, LA).

Moreover, dabigatran was associated with lower risks of major bleeding and net clinical outcome (combining stroke/systemic embolism and major bleeding) when compared with rivaroxaban, he reported at the American College of Cardiology 2018 Scientific Session here.

Although the study, funded by the makers of apixaban, is believed to be the largest comparison of NOACs to date, Deitelzweig said, “randomized controlled trials are needed to confirm the results.”

Indeed, no head-to-head trials have compared the NOACs, which have been shown to be at least as safe and effective as warfarin for preventing stroke/systemic embolism in patients with nonvalvular A-fib. Several prior observational analyses have explored differences between the agents, but they have often relied on single data sources, Deitelzweig said.

In the current study, dubbed ARISTOPHANES, investigators pooled five sources of claims data covering more than half of the US population, tapping into one Medicare database and four US commercial claims databases to examine the real-world performance of three NOACs: dabigatran (Pradaxa; Boehringer Ingelheim), rivaroxaban (Xarelto; Bayer/Janssen), and apixaban (Eliquis; Bristol-Myers Squibb). The most recently approved agent, edoxaban (Savaysa; Daiichi Sankyo), was not included because of the low number of patients who received a prescription for it.

The analysis included 194,532 adult patients with A-fib who had at least one pharmacy claim for rivaroxaban (53.2%), apixaban (32.6%), or dabigatran (14.2%) between January 2013 and September 2015 and had not had a claim for oral anticoagulation in the preceding 12 months.

Before adjustment, apixaban-treated patients were older on average and had the highest CHA2DS2-VASc and HAS-BLED scores compared with those receiving either of the other NOACs. Despite that, they had the lowest rates of stroke/systemic embolism, major bleeding, and net clinical outcome.

The investigators performed propensity-score matching, leaving 27,096 patient pairs for comparisons of apixaban and dabigatran, 62,219 for comparisons of apixaban and rivaroxaban, and 27,538 for comparisons of dabigatran and rivaroxaban.

For stroke/systemic embolism, apixaban was associated with a lower risk compared with both dabigatran (1.03% vs 1.42%; P = 0.0003) and rivaroxaban (1.21% vs 1.42%; P = 0.0038), whereas there was no difference between dabigatran and rivaroxaban (1.40% vs 1.23%; P = 0.0792).

Major bleeding also favored apixaban versus either dabigatran (2.68% vs 3.30%; P = 0.0001) or rivaroxaban (3.06% vs 5.33%; P < 0.0001), with dabigatran carrying a lower risk compared with rivaroxaban (3.28% vs 4.76%; P < 0.0001). The same pattern of relationships was seen for the net clinical outcome.

Deitelzweig pointed out that the analysis was subject to the usual limitations of observational studies, including the inability to establish causality, the potential for residual confounding, and the lack of more detailed information on laboratory results and biomarker levels. That latter issue precludes the ability to take a deeper look at how NOAC doses were being chosen for patients.

Jeanne Poole, MD (University of Washington, Seattle), who co-chaired the session at which Deitelzweig presented the results, highlighted another potential limitation. Large databases do not include information on whether patients are taking rivaroxaban with meals as recommended. If patients are not doing so, absorption of the drug can be decreased by 40%, which would affect comparisons of safety and efficacy, she said. Moreover, some patients in the analysis were receiving a dose of rivaroxaban typically used for deep vein thrombosis prophylaxis rather than the standard dose used for stroke prevention in A-fib, representing another potential confounder.

In comments made following Deitelzweig’s presentation, James Reiffel, MD (Columbia University Medical Center, New York, NY), also suggested additional limitations inherent to all real-world analyses: the inability to fully account for how physicians select drugs for certain patients and the inability to account for differences in the severity of conditions that might be included in statistical adjustments. “All the real-world analyses we have to take with a little grain of salt,” he said.

Findings Not Surprising

Commenting for TCTMD, Sana Al-Khatib, MD (Duke University Medical Center, Durham, NC), said the findings are largely consistent with prior observational studies comparing the NOACs and also with findings from the agents’ pivotal trials.

“When I see that alignment I’m more reassured that the results are credible and believable, but you can’t really confirm anything with observational studies,” she said.

That said, the results indicating that apixaban might have some advantages over dabigatran and rivaroxaban jibe with what Al-Khatib has taken away from looking at the individual trials.

“I don’t find these results surprising,” she said. “From my interpretation and careful review of all the trials, I feel that apixaban has the strongest data that support its use, but you have to take other things into account.”

Cost, for example, is a consideration because not all insurance plans cover all of the NOACs, Al-Khatib said. On the other hand, she pointed out, some patients feel very strongly about taking a once-a-day drug.

“If you have a patient [in whom] you’re concerned about compliance, if they tell you, ‘I’m not going to take medication more than once a day,’ then I think you should think of rivaroxaban, which is dosed once a day.”

All of the approved NOACs should be considered good options depending on these other factors, however.

“In all of the trials we see a significant safety advantage that these medications have over warfarin, especially when it comes to the most dreaded complication of intracranial bleeding,” Al-Khatib said. She agreed with Deitelzweig that head-to-head trials would be needed detect differences between the NOACs, but said, “I doubt that these trials will ever be done.”