Correlation of adverse events (AEs) and clinical outcomes in first-line enfortumab vedotin–pembrolizumab (EVP) treatment for locally advanced or metastatic urothelial cancer (La/mUC).

717 Background: Neuropathy and rash are the most common AEs with EVP. Their association with clinical outcomes in standard-of-care settings remains unclear. We hypothesized that these AEs may be associated with improved survival. Methods: We reviewed our institutional database of patients with La/mUC treated with first-line EVP (March 2022–July 2025). AEs and grades were extracted from provider notes; missing grades were inferred per CTCAE v5. Kaplan–Meier and time-dependent Cox regression analyses were performed, modeling AEs as time-dependent variables to account for immortal-time bias and variable onset. OS was measured from EVP initiation to death or last follow-up. Results: Among 185 patients, 141 (76%) were male, and 95 (48%) had de novo metastatic disease. Urinary bladder primary and pure urothelial histology were seen in 129 (70%) and 123 (66%), respectively. Baseline details are listed in the Table. At a median follow up of 9.7 months (95%CI: 8.2-11.9), any grade of rash and neuropathy was reported in 71 (38%) and 51 (28%) patients, at median onset of 0.4 (IQR 0.2-0.9) months and 3.2 (IQR: 2.3-4.4) months respectively. Both were reported in 26 (14%) patients. Grade 3/4 rash and neuropathy were reported in 8 (4.3%) and 1 (0.5%) patient respectively. One Grade 5 rash was reported. All but one Grade ≥3 rash developed within 2 weeks of initiation. EV dose reduction occurred in 83 (45%), most commonly for neuropathy (25, 30.1%) and rash (17, 20.4%). Discontinuation of EV alone, P alone, and both drugs was seen in 30 (16%), 5 (2.7%), and 21 (11%), respectively. Median OS for the cohort was 36.5 months (17.5–not reached). OS was longer in patients with neuropathy vs no neuropathy (36.5 vs 17.6 months; HR 0.46, 95%CI 0.21–1.04; p = 0.06) and rash vs no rash (not reached vs 17.5 months; HR 0.34, 95%CI 0.17–0.68; p = 0.002). After adjusting for prior neuropathy (only for neuropathy analysis), liver metastases, ECOG, starting dose, and time on EV, neuropathy and rash were associated with OS HRs of 0.54 (95%CI 0.24–1.24; p = 0.14) and 0.42 (95%CI 0.20–0.86; p = 0.02), respectively. Conclusions: The prevalence and timing of neuropathy and rash observed in our cohort are consistent with major trials of EVP in La/mUC, underscoring the reproducibility of these safety findings. In first-line EVP-treated La/mUC, these hypothesis-generating findings suggest susceptibility to AEs may correlate with survival benefit. Alternative explanations include germline variation in payload metabolism. Limitations include small numbers, short follow-up, retrospective collection, and potential residual immortal-time bias despite modelling. n (%) (Total N=185) Baseline ECOG > 0 113 (61) Baseline neuropathy 38 (21) Liver Metastases 37 (20) Starting EV Dose- 1.25/1/0.75/0.5 mg/kg 145(78)/ 26(14)/ 13(7)/ 1(0.5) Median EV doses received 7 (IQR: 3-12)