OC 66.3 - A model-based meta-analysis leveraging total knee replacement (TKR) data for selection of the milvexian dose for a phase 3 study in participants with atrial fibrillation.

Tuesday, June 25, 2024

15:15 – 15:30 ICT

Room: 111 A-C

Co-Author(s)

GA

Gregory E. Alexander

Sr. Scientist
Janssen Research & Development, LLC
Raritan, New Jersey, United States
MZ

Matthew Zierhut

Vice President, Integrated Drug Development
CERTARA
San Diego, California, United States
FC

Fei Chen

Sr. Director, Biostatistics
Janssen Research & Development, LLC
Raritan, New Jersey, United States
ES

EunYoung Suh

Scientific Director
Janssen Research & Development, LLC
Raritan, New Jersey, United States
PZ

Peter Zannikos

Scientific Director
Clinical Pharmacology and Pharmacometrics, Johnson and Johnson Innovative Medicine
Spring House, Pennsylvania, United States
AP

Alexei Plotnikov

Sr. Clinical Program Head
Global Clinical Development of CVM, Johnson and Johnson Innovative Medicine
Raritan, New Jersey, United States
NG

Navin Goyal

Sr. Director CVM PA
Clinical Pharmacology and Pharmacometrics, Johnson and Johnson Innovative Medicine
Raritan, New Jersey, United States

Presenting Author(s)

Chandni Valiathan, PhD

Director Modeling and Decision Sciences
Janssen Research & Development, LLC
San Diego, California, United States

Background: Milvexian is a FXIa-inhibitor being evaluated for prevention of thromboembolic events. Unmet medical need and results from Phase 2 studies (total knee replacement [TKR] - NCT03891524 and secondary stroke prevention [SSP] - NCT03766581) warranted evaluation of milvexian efficacy and safety in participants with atrial fibrillation (AF).

Aims: Dose-selection for the milvexian Phase 3 Librexia-AF study (NCT05757869) was supported by model-based meta-analysis (MBMA), a quantitative method that integrates aggregate-level clinical endpoint data across treatments and studies for comparative assessment of clinical endpoints.

Methods: Available data for Phase 2/3 total knee/hip replacement clinical studies published between 1986 and 2021 were identified. A proprietary joint model was developed for efficacy (venous thromboembolism [VTE]) and safety/bleeding (major, minor, and clinically relevant non-major [CRNM] bleeding) using a non-linear mixed-effects dose-response model. For each endpoint, treatments were assumed to share a common E-max differing only in their ED50, except for milvexian safety endpoints, which were captured by a step-function. Odds ratio for VTE were estimated for several doses of milvexian compared to apixaban 5 mg BID (typical AF dose) for non-inferiority assessment.

Results: Data for 9 active treatments and placebo from 45 clinical studies were included (Figure 1). The median VTE odds ratio showed a dose response favoring milvexian over apixaban for doses ≥75 mg BID (Figure 2). The Phase 2 milvexian TKR study showed no dose-related increase in bleeding, with low incidence of major or CRNM bleeding (0.8% all doses, 25 mg QD to 200 mg BID, combined) limiting the estimation of a dose-response for milvexian bleeding events, and subsequently limiting comparison of bleeding events with other treatments.

Conclusion(s): Considering the totality of available data along with MBMA-based non-inferiority assessment versus apixaban as an active comparator, 100 mg BID of milvexian was selected to be tested in the Phase 3 Librexia-AF study.