PII-216 - EXPANDING THE SCOPE OF AN EXISTING QSP MODEL OF CD3 BISPECIFICS TO CREATE A PLATFORM MODEL FOR HEMATOLOGICAL MALIGNANCIES FOR CLINICAL APPLICATIONS

S. Pallikonda Chakravarthy¹, A. Saini¹, M. Roy², K. Aditya Raman¹, A. Shaliban¹, M. Rengaswamy¹; ¹Vantage Research, ²Vantage Research, Lewes, Delaware, USA.

Background: CD3 bispecific antibodies (CD3 BsAbs) function by engaging T cells and redirecting them towards cancer cells, resulting in targeted and potent immune responses against cancer cells.

Recently, anti-BCMA and anti GPRC5D bispecifics have been approved for Multiple Myeloma (MM). Despite promising efficacy, combination bispecifics are of interest to address non-responders. We developed and calibrated the QSP model for these two CD3 BsAbs mono therapies and predicted the combination therapy results [1,2,3].

Ours and several groups have published mPBPK/PD models of bispecifics to capture PK across preclinical and clinical studies [4,5,6]. Hosseini et al [7] has published a comprehensive QSP model for CD3 BsAbs for Non-Hodgkin Lymphoma. In this work we have recalibrated the QSP model from [7] for Multiple Myeloma.

Objectives
- Modify existing model [7] to capture the clinical efficacy and CRS of anti-BCMA and anti-GPRC5D monotherapies in Multiple Myeloma patients.
- Predict the combo efficacy and CRS of anti-BCMA + anti-GPRC5D, and validate the results against the clinical data.
- Leverage this mechanistic platform approach to develop combination dosing strategy for bispecifics in liquid cancers.

Methods: We used the Hosseini et al model [7] as the base model and then changed model equations/parameters to modify the physiology of interest from B Cell lymphoma to Multiple Myeloma. While we retained other aspects of model structure to capture complex PBPK characteristics, T Cell activation and exhaustion, margination and trafficking, tumor growth inhibition and CRS [7]. The model was calibrated to anti-BCMA and anti-GPRC5D, obtained from the clinical trials NCT04557098 and NCT03399799. Model captures the Overall Response Rate (ORR) and cytokine levels for both therapeutic agents as mono and in combinations.

Results: The model predicts the ORR for the combination of two bispecifics within < 10% of the reported clinical data. This indicates that the model has captured bispecific PK/PD in multiple myeloma and can be used to determine combination strategy.

Conclusion: CD3 Bispecific QSP platform model for hematological malignancies can be used to optimize dose and dosing regimens, inform clinical trial designs, for mono and combination therapies. Alternate dosing strategies including dose priming can be tested for their impact on efficacy and CRS.