Methods: Data from dose-escalation stage (Phase 1) and dose-expansion stage (Phase 2) from clinical study ASTX660-01 were included. Subjects recruited into Phase 1 received tolinapant in either powder (15, 30, 60, 120 and 180 mg; n=16) or capsule formulation (180, 210 and 270 mg; n=27). Subjects in Phase 2 only received capsule formulation (90, 120, 150 or 180 mg). A population PK model was developed with NONMEM v. 7.3 using first-order conditional estimation with eta-epsilon interaction (FOCE-I). Model selection was based on goodness-of-fit plots, objective function values, prediction and variance corrected visual predictive check (pvcVPC), and model plausibility. Confidence intervals (CIs) around the parameters were computed using the sampling importance resampling (SIR) method.
Results: The data comprised 3427 tolinapant concentration measurements from 163 subjects (Phase 1; n=43, Phase 2; n=120) aged 23 to 84 years. tolinapant PK was best described using a two-compartment nonlinear elimination model with absorption described by a transit compartment model. The rate of absorption was dependent on formulation and was described by separate transit rate constants (KTR) for powder and capsules. Population estimates of Michaelis-Menten constant (Km), maximum elimination rate (Vmax), and apparent central volume of distribution (Vc/F) were 918.3 ng/mL, 72.2 ng/L and 488.6 L, respectively. Between subject variability included on Vmax and Vc/F were 21.3% and 40%, respectively. An additive error model on log transformed data was used to account for the unexplained residual variability. All parameters were estimated with acceptable precision. The predictive performance of the model assessed by the pvcVPC indicated that the data were adequately described by the model.
Conclusion: A population PK model was developed for tolinapant in subjects with advanced solid tumors and lymphomas. This work is the first description of the tolinapant PK, and the next steps are to explore the exposure and efficacy relationships and investigate the anti-tumor activity of tolinapant in one or more selected tumor types.
1. Dosne AG, Berstrand M, Harling K, Karlsson MO. Improving the estimation of parameter uncertainty distributions in nonlinear mixed effects models using sampling importance resampling. J Pharmacokinet Pharmacodyn, 43(6):583–596, 2016.