Cutaneous toxicity and clinical outcomes with enfortumab vedotin and pembrolizumab in patients with locally advanced or metastatic urothelial carcinoma.

4574 Background: Enfortumab vedotin and pembrolizumab (EV/P) recently became the standard of care frontline treatment for patients with locally advanced or metastatic urothelial carcinoma (la/mUC). In patients treated with EV monotherapy, previous studies reported a potential association between cutaneous toxicities and improved clinical outcomes. While cutaneous toxicities are common with EV/P, the correlation with response has not been described. Methods: This is a retrospective cohort of patients treated with first line EV/P for la/mUC at Memorial Sloan Kettering Cancer Center. Clinical data were collected by chart review. Response to EV/P was defined by investigator assessment adhering to RECIST 1.1. Cutaneous toxicity events were recorded and classified as mild (managed with topical agents and/or oral antihistamines) or moderate/severe (requiring further pharmacological interventions and/or dose modifications). Association between clinical characteristics and response were analyzed using univariable and multivariable logistic regression models. Results: 186 patients who started EV/P between October 2018 and September 2024 were identified, 166 with distant metastases (89%) and 20 (11%) with locally advanced disease. Median age was 72 years, 68% were male, 31% had upper tract primary, and 44% had subtype/divergent histology component. Rates of bone, lung, and liver metastases were 29%, 24%, and 19%, respectively, and 27% had lymph node only disease. Observed response rate was 64% (119/186; 95% CI 57%, 71%), including a complete response rate of 18% (33/186; 95% CI 13%, 24%). Cutaneous toxicities occurred in 106 patients (57%), predominantly within the first 12 weeks (94/106), 70% of events were classified as mild and 30% as moderate/severe. On univariable analysis, cutaneous toxicity at any time, and before 12 weeks were significantly associated with response to EV/P, with odds ratio (OR) 2.6 (95% CI 1.44 - 4.94, p = 0.002) and 1.91 (95% CI 1.04 - 3.53, p = 0.037). Greater effect was seen based on severity: OR 2.24 (95% CI 1.17 - 4.42, p = 0.017) for mild events and OR 4.12 (95% CI 1.62 - 12.0, p = 0.005) for moderate/severe events compared to patients without cutaneous toxicity. Response to EV/P was less likely with distant metastatic disease (vs. locally advanced; OR 0.28, p = 0.05), and with bone metastases (OR 0.44, p = 0.015), and more likely with lymph node only disease (OR 2.89, p = 0.007). When adjusted for treatment setting (metastatic vs locally advanced) and time on treatment in multivariable analysis, the effect of cutaneous toxicity ≤12 weeks on response was no longer significant (OR 1.7, 95% CI 0.9 - 3.25, p = 0.11). Conclusions: In our large, real-world cohort there was a non-significant trend for response to EV/P in patients with cutaneous toxicities. Further studies are needed to define the potential correlation.