Population Pharmacokinetic and Exposure-Response Modeling for the EZH1/2 Dual Inhibitor DS-3201b in Patients with Non-Hodgkin Lymphomas

Introduction: Epigenetic regulators are an emerging new target class for cancer therapy. Enhancer of zeste homolog 1 (EZH1) and EZH2 are alternative subunits of polycomb repressive complex 2 (PRC2), and catalyze tri-methylation of the 27th lysine residue of histone H3 (H3K27). The tri-methylation of H3K27 (H3K27me3) plays an important role in the repression of genes associated with tumor suppression and cell differentiation. DS-3201b is a dual inhibitor of EZH1/2 with activity against hematological cancer cell lines that include acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), diffuse large B-cell lymphoma (DLBCL), and adult T-cell leukemia-lymphoma (ATL). The first-in-human (FIH) study (NCT02732275) is on-going to assess safety, tolerability, pharmacokinetics (PK), pharmacodynamics (PD) and potential anti-tumor activity in patients with relapsed or refractory non-Hodgkin lymphoma (NHL), including ATL. This preliminary analysis was aimed to develop a population PK model, to explore the exposure-response (E-R) relationships for key safety and PD endpoints, and to find the optimal dose and schedule for DS-3201b.

Methods: As of August 2, 2017, clinical data that include plasma levels of DS-3201a (a free form of DS-3201b) and platelet counts were collected from 15 patients with NHL who were participated in the FIH study. They received DS-3201b oral once-daily dosing ranging from 150 mg to 300 mg in 28-day cycles. Population PK analysis for DS-3201a in plasma was carried out and the potential influence of patient characteristics on the PK of DS-3201a was investigated in a covariate analysis. The PK/PD model for platelet counts was developed using a semi-mechanistic model which consists of compartments that imitate platelet proliferation, maturation, circulation, and a feedback on proliferation. Preliminary E-R analyses were also performed for PD biomarker.

Results: A 2-compartment PK model with first order elimination and absorption lag-time best characterized the plasma concentration-time profile of DS-3201a. Serum albumin was a significant covariate for total clearance. In the FIH study, platelet counts decreased to a nadir during the first treatment cycle and it returned to baseline without drug interruption in most patients who received 150 mg or 200 mg doses, whereas all patients who received 300 mg required dose interruption. The PK/PD model with drug effect on platelet proliferation and feedback compartments aptly described the time-course of platelet counts. Model-based simulations showed that up-titration schedule based on each patient's platelet count after cycle 1 treatment could allow administration of higher DS-3201b doses to a subset of patients with low risk of grade ≥3 thrombocytopenia. The results of E-R relationship for the inhibition of H3K27me3 in surrogate tissues will also be presented.

Conclusion: We developed preliminary population PK and E-R models for DS-3201b in patients with relapsed or refractory NHL. Model-based simulations suggested that up-titration approach can be effective for DS-3201b to ensure continuity of treatment benefit while minimizing dose-dependent thrombocytopenia.