Longitudinal Tumor Size and Survival Modeling for Exposure-Response Analysis of Drugs with Frequent Dose Reductions:...

Abemaciclib is an oral anticancer drug indicated for treatment of HR+ HER2- breast cancer. Dose modifications due to side effects are frequent, thus drug exposures change over time as a result of altering the dose or temporarily withholding abemaciclib treatment. This limits the utility of simple exposure-response assessments (such as Kaplan-Meier quartile analyses) using static summary pharmacokinetic measures of drug exposure. The objective of the current work was to characterize the exposure-response relationship of abemaciclib using nonlinear mixed-effects modeling in 663 patients with metastatic or advanced breast cancer. A time-course joint simultaneous pharmacokinetic-tumor size-progression-free survival model accounted for changes in drug exposure due to dose reductions/omissions and described the change in tumor size over time. The change in tumor size was a highly significant predictor of the hazard for progression-free survival in a joint tumor size-time to event model (P < 0.001). Interval censoring was used in the time to event model to account for wide patient visit intervals. Simulations using the combined pharmacokinetic-tumor size-time to event model demonstrated the adequacy of a 150 mg twice daily dose in combination with fulvestrant. There was a negligible impact of dose reductions on efficacy, likely due to the shallow exposure-response relationship. When analyzing survival data where the drug exposure changes significantly within an individual over time, it is important to maximize the use of available longitudinal data through simultaneous modeling of time-course tumor size and survival data.