Background
TAK-102 is an investigational GPC3 targeted CAR-T therapy armoured with IL-7 and CCL19 to enhance proliferation/persistence and induce host-immune cell infiltration respectively. An open-label, non-randomized phase-1 study (NCT04405778) was conducted in patients with GPC3+ solid tumors to evaluate safety and tolerability of TAK-102 as well as identifying recommended Phase-2 dose (RP2D) level.
Methods
As of March 31st, 2024, 11 patients have been infused with TAK-102 at dose-levels of 10 million (DL1), 100 million (DL2) and 500 million (DL3) CAR+ cells/patient after standard lymphodepletion chemotherapy regimen of cyclophosphamide and fludarabine. Cellular kinetics (CK) of TAK-102 was measured by PCR and flow cytometry-based assays and non-compartmental analysis (NCA) was performed to extract cellular kinetic parameters. Humoral immunogenicity was evaluated by measuring anti-drug antibody (ADA) titers against CAR extracellular domain (ECD), IL7 and CCL19. Pharmacokinetics (PK) of IL-7 and CCL19 along with a large panel of cytokines/chemokines were measured pre- and post- TAK-102 infusion to assess their associations with TAK-102 exposure. Integrated CK-pharmacodynamic (PD) relationships were developed using longitudinal tumor volume measurements along with other indication-specific biomarkers (e.g., alpha fetoprotein (AFP) levels). A physiologically based CK-PD model was developed which mechanistically describes the multiphasic CK profile of TAK-102 and its distribution in relevant tissues and solid tumors. The model also incorporated the impact of lymphodepletion on circulating homeostatic cytokine levels and tumor heterogeneity in GPC3 expression at the site of CAR-Target engagement.
Results
Among the 11 patients treated so far, no DLT or neurotoxicity was observed, whereas 6/11 patients experienced mild cytokine release syndrome (CRS, mostly grade 1) managed by appropriate medications (e.g., tocilizumab), suggesting acceptable safety profile of TAK-102. Stable disease was achieved in five patients, and in one patient with hepatocellular carcinoma (HCC) antitumor activity lasting 6 months was also confirmed by monitoring of peripheral blood AFP levels. An increase in TAK-102 exposure (AUC, Cmax) was observed when escalating from DL1 to DL2, whereas further dose-escalation to DL3 only led to decrease in Tmax values. A dose-dependent kinetics of IFN-gamma, IL-6 and CCL19 suggested higher signal of activity at higher dose-levels. A mechanistic PBCK-PD model was able to adequately integrate dataset(s) pertaining to extent of lymphodepletion, homeostatic cytokines, TAK-102 CK, tumor volume and key biomarker measurements by goodness of fits.
Conclusions
The integrated CK-PD analysis presented here could facilitate identification of relative contribution of factors impacting exposure/response of TAK-102 and help understand the mechanism of action of this therapy to further enable future dose-optimization.
Trial Registration
NCT04405778.