Apixaban for treatment of venous thromboembolism (VTEtx): Use of Model-Based Meta-Analysis (MBMA) to support Phase 3 dose selection and beyond
Rebecca Boyd,1 Wonkyung Byon,1 John Thompson,1 Margot Johnson,1 Jaap Mandema2
1Pfizer, Groton, CT, 2Certara, Quantitative Solutions Inc., Menlo Park, CA.
Objectives:
Apixaban is an orally available, direct, selective inhibitor of the coagulation factor Xa that reversibly binds to the active site of FXa, and exerts anticoagulant and antithrombotic effects by diminishing the conversion of prothrombin to thrombin. Apixaban is approved in multiple regions for VTEtx, based on the results of a Phase 3 study (AMPLIFY [1]). Model-Based Meta-Analyses (MBMA) can inform drug development decisions and increase the probability of technical success. Here we describe how the results of MBMA were used to support dose selection for the Phase 3 VTEtx study of apixaban (apixaban 10 mg twice daily [BID] for 7 days, followed by 5 mg BID in AMPLIFY) and how the models were updated and used during the conduct of the study to address questions and inform decisions.
Methods: Efficacy and safety dose-response data from 20 treatments evaluated in the prevention of VTE (VTEp) following orthopedic surgery (representing > 39,000 patients in 63 trials) were fit with logistic regression models that specified a similar shape of the dose-response (D-R) curve across compounds [2].D-R MBMA of total VTE and major bleeding (MB) for anticoagulants in VTEp were linked to efficacy (symptomatic VTE) and safety (MB) in VTEtx to generate D-R relationships across compounds, with doses expressed as enoxaparin equivalents. Separate relationships were determined for acute (5-14 days) and chronic (>14 days up to 6 months) effects of the initial treatment. Using these models, VTE and MB event frequencies for various dose regimens of apixaban were estimated. As additional clinical trial data became available for both VTEp and VTEtx during the conduct of the AMPLIFY trial, the models and expected outcomes were updated and used, along with blinded event rates, to assess the impact of an increase in sample size using clinical trial simulations, as well as any impact on hierarchical statistical testing.
Results: An apixaban regimen of 10 mg BID for 7 days followed by 5 mg BID was predicted to result in similar or better efficacy and safety vs. standard of care (subcutaneous enoxaparin followed by vitamin K antagonist) for the total treatment period as the model supported the benefit of a higher dose during the acute treatment period. The updated original MBMA predicted a high probability of meeting the primary objective of non-inferior efficacy of apixaban relative to enoxaparin/warfarin, but only ~50% probability of achieving superiority on efficacy with the pre-specified maximum sample size of 5400. A new network MBMA using apixaban data from the AMPLIFY-Extension study [3] was used to refine the probabilities of efficacy and bleeding outcomes in the AMPLIFY study. The probability of achieving superiority for MB was estimated to be greater than that for recurrent VTE. Based on this result, the order of hierarchical testing in the AMPLIFY study was pre-specified in the statistical analysis plan prior to the database lock, to be 1) non-inferiority for VTE, 2) superiority for MB, and 3) superiority for VTE.
Conclusions: The MBMA supported selection of a dose of 10 mg BID for 7 days followed by 5 mg BID for the 6-month treatment of VTE. In AMPLIFY, consistent with the model-based predictions, this regimen of apixaban was non-inferior for VTE and superior for MB, with observed relative risk (95% CI) of 0.84 (0.60-1.18) and 0.31 (0.17-0.55), respectively [1].
This work was presented at ASCPT 2015 [Clinical Pharmacology & Therapeutics, Volume 97, Issue S1, February 2015, Pages: S99–S100]
References:
[1] Agnelli G et al., N Engl J Med 2013; 369:799-808 [2] Mandema J, Boyd RA, DiCarlo L. Clin Pharmacol Ther 2011; 90:820-827 [3] Agnelli G et al., N Engl J Med. 2013; 368:699-708