Pharmacokinetic and Pharmacodynamic Modeling of the Relationship Between TAK-442 Exposure and Clinical Outcome In Patients Undergoing Elective Knee Arthroplasty

Fran Stringer, Bart Ploeger, Nolan Wood, Patrick Moneuse, Masaki Kawamura, Stuart Kupfer, Jeff Spaeder and Graham Scott


Abstract 3326

Introduction: TAK-442 is an oral direct factor Xa (fXa) inhibitor currently in clinical development for the prevention of venous and arterial thromboembolism. Pharmacokinetic (PK) and pharmacodynamic (PD) modeling is a powerful tool for predicting efficacy and safety in drug development programs. The derivation of PK-PD exposure metrics in patients allows exploration of the key descriptors of the clinical response and can lead to more informed dose selection. The development of PK-PD models using response biomarkers relevant to anticoagulation (e.g., prothrombin time [PT] and fXa inhibition) may predict the antithrombotic effects of the drug in this patient population.

Methods: Population PK-PD methods (non-linear mixed effects modeling program) were used to derive individual exposure parameters, such as area under the plasma concentration time curve (AUC), maximum concentration (Cmax) and minimum concentration at the end of a dose interval (Cmin) and the corresponding PD parameters (PT and fXa inhibition) for all patients receiving TAK-442 in a Phase II dose ranging trial. Patients received 40 or 80 mg p.o. once daily or 10, 20, 40, or 80 mg p.o. twice daily (n=150 per group) for thromboprophylaxis after elective total knee arthroplasty. Logistical regression analysis was used to explore the relationships between the derived PK-PD exposure metrics and the primary efficacy (composite of all-cause mortality, symptomatic and asymptomatic deep vein thrombosis [DVT], and symptomatic pulmonary embolism [PE]) and safety endpoints (combined major, minor, and clinically significant non-major bleeding events).

Results: The PK of TAK-442 were found to be linear over the dose range studied, with no significant time dependencies. The logistical regression analysis identified significant relationships between increasing exposure and a decrease in the probability of the primary efficacy endpoint, with the strongest relationships observed for Cmax and AUC (P-value of ≤0.001 and 0.03, respectively) and fXa inhibition Emax (P-value of ≤0.001). For every 100 ng/mL increase in Cmax, the odds of having a composite efficacy event were reduced by 10.1% (odds ratio=0.899, 95% confidence interval=0.843, 0.958). In contrast, exploration of the safety endpoint for bleeding did not identify any significant relationships between any of the exposure metrics (Cmax, Cmin, and AUC) or the corresponding PD metrics for fXa inhibition or PT.

Conclusions: The PK and PD of TAK-442 in patients undergoing total knee arthroplasty were found to be dose-dependent and predictable. Significant relationships were identified between the PK and PD exposure metrics and the reduction in the primary efficacy endpoint. At the dose ranges studied, no relationship was found between any of the PK-PD exposure metrics and the bleeding endpoint. In conclusion, key predictors of clinical efficacy were identified, whereas no exposure metrics correlated with increased bleeding events.

Disclosures: Stringer: Takeda Global Research & Development Centre, Ltd.: Employment. Ploeger: Takeda Global Research & Development Center, Inc.: Consultancy. Wood: Takeda Global Research & Development Centre, Ltd.: Employment. Moneuse: Takeda Global Research & Development Centre, Ltd.: Employment. Kawamura: Takeda Global Research & Development Center, Inc.: Employment. Kupfer: Takeda Global Research & Development Center, Inc.: Employment. Spaeder: Takeda Pharmaceuticals International: Employment. Scott: Takeda Global Research & Development Centre, Ltd.: Employment.