Abstract 6006: Modeling antibody drug conjugate sensitivity using urothelial carcinoma patient-derived models

Abstract Introduction: Enfortumab Vedotin (EV), an antibody-drug conjugate (ADC) targeting Nectin-4, has emerged as a new standard-of-care for patients with metastatic urothelial cancer (UC). Promising results have also been observed in bladder cancer (BLCA) patients with ADCs targeting HER2. To identify predictive biomarkers of ADC sensitivity, we characterized a large and expanding biobank of patient-derived organoid (PDO) and xenograft (PDX) models for Nectin-4 and HER2 expression, and Nectin-4 and HER2-targeted ADC sensitivity. Methods: We generated 19 PDOs and 34 PDXs that reflect the genomic and biologic heterogeneity of human UC. A subset of these models was further characterized using a multiplatform approach, including bulk and single-cell DNA and RNA sequencing and immunoblot for Nectin-4 and HER2 expression. We leveraged these models to study the biologic determinants of sensitivity to EV and the HER2-targeted ADC trastuzumab deruxtecan (T-DXd). PDX models were treated with EV (5 mg/kg IV on day 1, 3, 6) or T-DXd (10 mg/kg IV on day 1, 21, 42) or vehicle only as control. We also studied sensitivity to T-DXd and EV, as well as to their cytotoxic payloads (exatecan and vedotin) in vitro using paired PDO models. Results: We observed high degree of variability of HER2 and Nectin-4 expression across UC PDOs and PDXs. In PDO models, sensitivity to EV correlated strongly with Nectin-4 expression, whereas HER2 expression was not strongly correlated with sensitivity to T-DXd. We also noted a poor correlation between Nectin-4 expression in PDOs and PDXs derived from the same patient. A subset of models, for example, SMBO-104 (ERBB2 wildtype with low HER2 expression), exhibited significantly greater sensitivity to EV than to T-DXd. Conversely, T-DXd was significantly more active than EV in SMBO-170, a model with ERBB2 amplification (~40 copies) and HER2 3+ expression. Sensitivity to exatecan, an analogue of the T-DXd payload deruxtecan, as measured by inhibition of cell proliferation and induction of apoptosis, was more predictive of T-DXd sensitivity than HER2 expression level. Conclusion: In this study, we generated and biologically characterized UC patient-derived models with the goal of identifying predictive biomarkers of sensitivity to Nectin-4 and HER2-targeted ADCs. We observed significant heterogeneity of HER2 and Nectin-4 expression across models. Models were identified in which T-DXd was superior to EV, and vice versa. Sensitivity to the T-DXd cytotoxic payload was more predictive of PDX response than HER2 expression levels, with significant tumor regression observed with T-DXd in BLCA models with low HER2 expression. Our data suggest that clinical trials of HER2-targeted ADCs such as T-DXd should not be restricted to patients with HER2 overexpression and that novel biomarkers will be needed to identify those UC patients most likely to benefit from FDA-approved and investigational ADCs. Citation Format: Xinran Tang, Ziyu Chen, Jasmine Thomas, Karan Nagar, John Christin, Naryan Rustgi, Sizhi Gao, Elisa de Stanchina, Jonathan Coleman, Michael Shen, Hikmat Al-Ahmadie, Gopa Iyer, Kwanghee Kim, David Solit. Modeling antibody drug conjugate sensitivity using urothelial carcinoma patient-derived models abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (6Suppl): Abstract nr 6006.