(S-105) POPULATION PHARMACOKINETIC (POPPK) MODELING OF IRINOTECAN AND SN38 IN PARTICIPANTS WITH METASTATIC COLORECTAL CANCER (MCRC) IN PHASE 2 AND 3 STUDIES: A SEQUENTIAL PARENT-METABOLITE MODEL

Steven Sun – UCSD; Erik Hahn – Pfizer Inc

Objectives: Irinotecan and its active metabolite, SN-38, are inhibitors of topoisomerase I-DNA complex, which prevent religation of single-strand DNA breaks and induce cytotoxicity. Currently, irinotecan is approved for use in first- and second-line settings to treat mCRC. The objective of this analysis was to develop a popPK model to characterize irinotecan and SN-38 disposition across multiple clinical trials in mCRC in a sequential parent-metabolite modeling approach.

Methods: The popPK analysis was based on seven Phase 2 and 3 studies, in participants with mCRC who received irinotecan in first- or second-line (5-fluorouracil refractory) setting. PopPK modeling was performed using non-linear mixed effects modeling approach (NONMEM v7.5.0) with first order-conditional estimation with interaction. A sequential parent-metabolite modeling approach was employed, where individual post-hoc estimates of irinotecan clearance (CL), central volume (V1), intercompartmental clearance (Q), and peripheral volume (V2) from the parent model were incorporated as subject-specific parameters in the metabolite model. Stepwise covariate modeling assessed the influence of a range of covariates (e.g., age, bilirubin concentration, sex, and if patients received chemotherapy prior to screening) on key pharmacokinetic parameters. Model validation included bootstrap resampling (n = 1000) and visual predictive checks.

Results: A sequential parent-metabolite model was developed to describe the pharmacokinetics of irinotecan and SN-38 using data from 707 individuals with 6428 observations. A two-compartment model with first-order elimination best described irinotecan and SN-38 disposition. Irinotecan CL, V1, Q, and V2 were estimated to be 26.4 L/h, 147 L, 12.6 L/h, and 188 L, respectively. Statistically significant effects of age and bilirubin on CL and bilirubin and chemotherapy on V1 were identified. Based on the 10th and 90th percentiles of continuous covariate values, no covariate changed CL or V1 more than ±25%. Receiving previous chemotherapy was associated with a 10.3% decrease on irinotecan V1. SN-38 CL, V1, Q, and V2 were estimated to be 95.8 L/h, 15.1 L, 67 L/h, and 536 L, respectively. The final model demonstrated adequate goodness-of-fit, parameter precision, and predictive performance for both parent and metabolite disposition.

Conclusions: A sequential parent-metabolite popPK model was successfully developed to describe the pharmacokinetics of irinotecan and SN38 in patients with mCRC. The final model identified significant covariates influencing irinotecan disposition, which may inform dose optimization strategies. The employed model evaluation supported its predictive performance and potential application in clinical settings.

Citations: NA

Keywords: population pharmacokinetics, oncology