Abstract 7260: Predicting the starting dose and step-up regimen for first-in-human trial of T-cell engagers using mechanistic mathematical modeling

Abstract The minimal anticipated biological effect level (MABEL) approach is recommended for selecting a safe clinical starting dose for T-cell engagers (TCEs). This approach estimates the minimal recommended starting dose (MRSD) by setting predicted drug exposure in humans below the concentration that results in a desirable percentage of maximal pharmacological activity (PA) observed in vitro (in the range of 20%-50%, e.g., EC20). However, this method may lead to a low MRSD and multiple dose escalations, resulting in sub-therapeutic doses (Saber et al. Regul Toxicol Pharmacol. 2017 Nov;90:144-152). The aim of our study was to apply quantitative systems pharmacology (QSP) modeling to predict the MRSD and step-up dosing regimen for HPN536, a T-cell engager targeting mesothelin (MSLN). The translational QSP model consists of two parts: an in vitro model and a pharmacokinetic (PK), receptor occupancy (RO), and PA model. The in vitro model was developed to describe published data on T-cell dependent cytotoxicity, cytokine secretion, and T-cell activation in the presence of various HPN536 concentrations, and to identify EC50 expressed in terms of the number of trimers—complexes of HPN536 with CD3 and MSLN in the immunological synapse (IS) between T-cells and cancer cells. PK parameters were fitted using data on cynomolgus monkeys and translated to humans using a standard allometric scaling approach. The PK/RO/PA model for cancer patients includes the distribution of TCE into the tumor (ovarian cancer), binding of HPN536 with CD3 and MSLN, trimer formation in IS, and PA based on EC50 values identified in the in vitro model. PA depends on the number of trimers in the tumor. Variability was implemented in the model, including MSLN expression, PK, T-cell density in the tumor, and tumor size. Starting dose estimates were calculated as a dose range corresponding to 20%-50% PA. The MRSD range predicted by the model was higher than the range calculated by the conventional MABEL PK-driven approach. The starting dose predicted by the model is 40 ng/kg based on cytokine assays, whereas the starting dose in the phase 1 trial was 6 ng/kg (established by the MABEL PK-driven approach), and CRS grade 3 was observed at 54 ng/kg without dexamethasone pretreatment. The optimal step-up regimen is the one that minimizes peak trimer levels and results in the highest trough trimer level at the target dose. Considering once-weekly administration of the HPN536 target dose, the optimal regimen with two step-ups was as follows: 0.001 mg/kg given on day 1, 0.003 mg/kg given on day 4, and 0.01 mg/kg QW started on day 8 (target dose). Applying QSP modeling to predict MRSD in addition to the MABEL PK-driven approach might help reduce the number of dose escalation steps, exposing patients to sub-therapeutic doses in first-in-human trials, and optimize step-up dosing regimens. Citation Format: Oleg Demin Jr, Galina Kolesova, Dmitry Shchelokov. Predicting the starting dose and step-up regimen for first-in-human trial of T-cell engagers using mechanistic mathematical modeling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 7260.