E-003 - A QUANTITATIVE SYSTEMS PHARMACOLOGY QSP) MODEL FRAMEWORK FOR COMBINED CLINICAL EFFICACY AND HEMATOLOGICAL TOXICITY PRECITIONS FOR ANTIBODY DRUG CONJUGATES.

B. PALEJA¹, R. DUTTA², S. SAHOO², R. KUMAR³; ¹VANTAGE RESEARCH, TAMILNADU, INDIA, ²VANTAGE RESEARCH, TAMIL NADU, INDIA, ³VANTAGE RESEARCH, LEWES, DE, USA.

Background: Antibody-drug conjugates (ADCs) signify a major leap forward in cancer treatment, providing a precise method to deliver potent cytotoxic drugs directly to cancer cells. Although ADCs are generally well tolerated by patients, certain treatment-related adverse events are commonly reported and lead to patient discontinuation. Dose-limiting toxicities (DLTs) are frequently observed across various ADCs that administer the same cytotoxic payload, regardless of the targeted antigen or the specific type of cancer being treated. Here we outline a quantitative systems pharmacology (QSP) model framework developed to predict both clinical efficacy and hematological toxicity for ADCs.

Methods: In the current work we have utilized a previously reported (1), QSP ADC model as the base model. This model comprehensively represents both the cellular and systemic dynamics of drug disposition, taking into account both drug and target antigen related properties such as expression of the target antigen, formation of drug-antigen complexes and their internalization. Additionally, the model integrates modules to simulate drug distribution within a tumor, inhibition of tumor growth and the phenomenon of bystander effect induced by the drug. For capturing ADC payload induced thrombocytopenia, we have repurposed chemotherapy induced myelosuppression (2,3). The model was informed and calibrated with in-vivo preclinical and clinical data for Trastuzumab emtansine (T-DM1) and used for clinical therapeutic index and alternate dosing strategy predictions.

Results: The developed model effectively captured in-vivo preclinical PK and efficacy data for T-DM1 (4). The model output for translation to humans accurately described clinical response in terms of tumor growth inhibition (5) and the model predicted platelet dynamics in the published data (6).

Conclusion: The QSP model framework described here can be utilized for effective prediction of efficacy and toxicity for ADCs. The model can be used to determine dosing regimen and therapeutic window. Given availability of data, the model can be used for informing translation and early clinical development of novel ADCs.