266O - A population pharmacokinetic model: Assessment of pharmacokinetic similarity of HLX01 and rituximab in diffuse large B-cell lymphoma

Background

HLX01, the first China (CN)-manufactured rituximab (RTX) biosimilar, was approved by the National Medicinal Products Administration (NMPA) for the treatment of non-Hodgkin's lymphoma on 22-February-2019, and was concurrently developed as a novel drug for rheumatoid arthritis (RA) since the indication has not been approved in CN. Here, we describe the development of a population pharmacokinetic (PopPK) model derived from the PK data of HLX01 and rituximab in patients with RA, and external validation in patients with CD20⁺ diffuse large B-cell lymphoma (DLBCL).

Methods

A PopPK model of HLX01 and the European-sourced RTX (EU-RTX) from HLX01-RA01 study in 196 RA pts (serum sample number=4289) was developed using non-linear mixed-effect modelling (NONMEM^®) with the first-order estimation with interaction (FOCEI) method. PK and PK-pharmacodynamic (PK-PD) relationship were characterised with various covariates (i.e., demographics, pathophysiologic/disease conditions and medical history, etc) which were examined by using forward addition (p < 0.01) / backward elimination (p < 0.001). The model was then examined using Bayesian bootstrapping and visual predictive check (VPC) followed by 1000 simulations were tested using the observed covariates. The final model was validated using PK samples of HLX01 and CN-RTX from HLX01-NHL03 study in ∼110 patients with CD20⁺ DLBCL and bridged the Chinese data to other races by simulation of Caucasian RA data from publication.

Results

A two-compartment model with first-order elimination provided the best model fit. The estimated clearance (CL), central volume (V_c), peripheral compartment volume (V_p) and clearance-of-distribution from the central to the peripheral compartment (Q) were 27.32%, 16.56%, 21.61%, and 40.79%, respectively. The correlation between CL and V_c was 0.02239. The observed concentrations and simulations-based 95% confidence interval for the corresponding model successfully predicted all subjects in the dataset, demonstrating PK similarity of HLX01 and RTX in patients with RA and CD20⁺ DLBCL. No significant difference in area under the curve from zero to infinity (AUC_0-inf) was observed in two different sourced RTXs. This model was similar to the existing model in Caucasian population.Table: 266O

Estimated parameter of HLX01 and EU-rituximab from the final PopPK model

Typical values for baseline creatinine clearance (BCR), baseline albumin (BALB), body surface area (BSA)are 125 mL/min, 41.2 g/L, 1.62 m² respectively.

Conclusions

This PopPK model derived from RA patients adequately predicts HLX01 and CN-RTX in patients with CD20⁺ DLBCL. HLX01 and EU-/CN-RTXs had similar PK parameters and influential PK covariates. These results provided further evidence for PK similarity between HLX01 and RTXs in patients with RA or DLBCL regardless of ethnicity.

Clinical trial identification

HLX01-RA01 Phase 1/2 study (NCT03355872) in moderately-to-severely active rheumatoid arthritis; HLX01-NHL03 Phase 3 study (NCT02787239) in CD20⁺ diffuse large B-cell lymphoma.

Editorial acknowledgement

Legal entity responsible for the study

Shanghai Henlius Biotech, Inc.

Funding

Shanghai Henlius Biotech, Inc.

Disclosure

Y. Shi: Advisory / Consultancy: Shanghai Henlius Biotech, Inc. X. Zhang: Full / Part-time employment: Shanghai Henlius Biotech,Inc. W. Jiang: Full / Part-time employment: Shanghai Henlius Biotech, Inc. S. Liu: Full / Part-time employment: Shanghai Henlius Biotech,Inc. E. Liu: Full / Part-time employment: Shanghai Henlius Biotech,Inc. K. Chai: Full / Part-time employment: Shanghai Henlius Biotech,Inc. A. LUK: Full / Part-time employment: Shanghai Henlius Biotech,Inc. D. Yao: Full / Part-time employment: Shanghai Henlius Biotech,Inc. All other authors have declared no conflicts of interest.