PII-146 - POPULATION PHARMACOKINETIC MODELING OF CAMRELIZUMAB IN PATIENTS WITH VARIOUS CANCER TYPES, INCLUDING ADVANCED HEPATOCELLULAR CARCINOMA.

N. DJEBLI¹, B. Wang², C. Yang³, S. Adriaens⁴, K. Shen⁵, S. Jang⁶, M. Zhou⁷; ¹Luzsana Biotechnonology (subsidiary of Jiangsu Hengrui Pharmaceuticals Co., Ltd), Luzsana Biotechnonology (subsidiary of Jiangsu Hengrui Pharmaceuticals Co., Ltd), , ²Amador Bioscience, Amador Bioscience - Pleasanton, USA, ³Jiangsu Hengrui Pharmaceuticals, 3Jiangsu Hengrui Pharmaceuticals - Shanghai, China, ⁴Amador Bioscience, Amador Bioscience - Ann Arbor, MI, USA, ⁵Jiangsu Hengrui Pharmaceuticals Co., Ltd, Lianyungang - Jiangsu, China, ⁶Elevar Therapeutics, Salt Lake City, USA, ⁷Luzsana Biotechnonology (subsidiary of Jiangsu Hengrui Pharmaceuticals Co., Ltd), Princeton - NJ, USA.

Background: Camrelizumab is a humanized monoclonal antibody (mAb) that specifically binds to programmed cell death receptor 1 (PD-1). Population analysis was conducted to characterize the pharmacokinetic (PK) properties and identify potential covariates (demographic and clinical covariates) of camrelizumab PK in patients with various cancer types including advanced hepatocellular carcinoma (HCC).

Methods: Camrelizumab PK data (6347 PK samples from 672 cancer patients) were pooled from 10 clinical studies including a pivotal Phase III study in HCC patients in combination with rivoceranib, a small molecule VEGFR-2 inhibitor, for population analysis. The final model was evaluated using visual predictive checks and bootstrap to determine if the predictive ability was adequate. Forest plots were generated to further assess the impact of covariates on camrelizumab exposure in cancer patients.

Results: Upon intravenous infusion administration, the PK of camrelizumab were adequately described by a 2-compartment model with parallel time-varying first-order and time-independent nonlinear (Michaelis-Menten type) elimination pathways. Estimated systemic clearance of the first-order elimination pathway (CL: 0.302 L/day [28.7 %CV]) and central and peripheral volume of distribution (Vc: 3.51 L [14.7% CV]; Vp: 2.75 L) of camrelizumab were similar to those of other therapeutic immunoglobulins. CL and Vc were affected by body weight and modeled using allometric scaling with estimated exponents. Post-hoc analyses demonstrated that time-dependent CL was associated with disease status change (time-varying albumin and time-varying tumor size). Baseline albumin level, renal function, sex, time-varying anti-drug antibody formation, and concomitant administration of rivoceranib were identified as significant PK covariates. However, the predicted effects of these covariates on camrelizumab exposure were generally within 20% of the reference and were not considered clinically relevant.

Conclusion: PK of camrelizumab was similar to that of therapeutic antibodies. As with other anti-PD-1 mAbs, CL was reduced over time. Age, race, ECOG score, tumor size and cancer type had no significant effect on camrelizumab PK.