Bridging sunitinib exposure to time-to-tumor progression in hepatocellular carcinoma patients with mathematical modeling of an angiogenic biomarker.

Abstract

e14690 Background: Hepatocellular carcinoma (HCC) is the most common primary liver malignancy with 250,000 new annual cases in the US. Sunitinib (SU), an oral tyrosine kinase inhibitor, inhibits tumor cell proliferation, selectively binds to the angiogenesis biomarker (VEGFR2), and is metabolized to an equipotent metabolite (SU12662). The objective of this study is to utilize mathematical modeling to bridge drug exposure and VEGFR2 dynamics to tumor growth inhibition (TGI) and time-to-tumor progression (TTP). Methods: Plasma concentrations of SU, SU12662, and VEGFR2, and tumor growth and TTP measurements were obtained from a phase-II study in 16 patients with unresectable HCC. SU was administered at 37.5 mg daily 7-days prior to chemoembolization, then on days 15-35. A MAP Bayesian approach was utilized to model SU and SU12662 pharmacokinetics (PK), both captured with a two-compartment model including linear clearances. Inhibition of VEGFR2 production was mediated by predicted active unbound concentrations (ACub) of SU and SU12662. VEGFR2 concentrations were the driver for TGI. The TTP probability was modeled with exponential hazard function dependent on a time varying covariate -the difference in VEGFR2 concentration from its baseline (ΔVEGFR2). Results: Drug and metabolite clearances (CLd, CLm) were estimated at 30.3 (19, %RSE) and 19.7L/h (14). Their volumes of distribution (Vd, Vm) were 1,780 (39) and 1840L (25). SU exposure was within target range for VEGFR2 inhibition. VEGFR2 half-life in plasma was calculated at 4h. The slope (α) for ACub effect on VEGFR2 production was 0.77(µg/mL)-1 (14). The ΔVEGFR2 effect on TGI was assumed maximal, whereas its potency (ΔIC50) was estimated at 1.83x10-2µg/mL (41). The between-subjects variabilities were 37.4, 51, 44.7, 64.8, 21, and 36% for CLd, CLm, Vd, Vm, α, and ΔIC50. The median observed tumor size baseline was 91 mm3. The simulated mean TTP was 7 months. Conclusions: The time-course of SU, SU12662, and VEGFR2 were captured well with final PK/pharmacodynamic models. VEGFR2 profiles successfully linked drug exposure to TGI and TTP. This model may serve as a general platform for the dynamics of anti-angiogenic drugs.