Impact of prior immune checkpoint inhibitor (ICI) and interval from ICI exposure on outcomes with enfortumab vedotin and pembrolizumab (EVP) in advanced urothelial carcinoma (aUC).

704 Background: EVP is the preferred frontline regimen for patients (pts) with aUC. However, pivotal trials excluded pts with prior ICI exposure, while data on its impact on EVP outcomes are limited. We hypothesized that prior ICI exposure, but not the interval between prior ICI and EVP, would be associated with outcomes. Methods: Pts in the retrospective UNITE study treated with EVP were separated based on any prior ICI exposure vs no ICI exposure. Among ICI-exposed pts, additional assessed variables included interval from ICI completion to EVP start (ICI–EVP interval; analyzed as categorical variable at ≥3, ≥6, ≥12 mos), prior ICI therapy setting, prior pembrolizumab, and best response with prior ICI treatment in the metastatic setting. Potential associations between these factors and progression-free (PFS) and overall survival (OS) were evaluated using Kaplan–Meier and Cox regression, while associations with observed response rate (ORR) were analyzed using logistic regression in pts with scans after ≥1 EVP cycle. Results: Among 493 pts treated with EVP, 68 (21 adjuvant, 46 metastatic, 1 both) had prior ICI. Adjuvant ICI included nivolumab (n = 19), pembrolizumab (n = 1), and 1 unspecified regimen. ICI in metastatic setting included pembrolizumab (n = 26, 1 in combination without EV), switch maintenance avelumab (n = 12), other ICI regimens (n = 7), and 1 unknown. In this cohort, median age was 68 yrs, 68% male, 67% pure urothelial histology, 83% ECOG PS 0/1, 43% had visceral metastases (excluding bone). With a median follow-up of 16 mos from EVP start, median PFS and OS were 7 mos (95% CI 5.3–8.5) and 13 mos (95% CI 10.8–NR), respectively; ORR was 47% (28/59; 95% CI 34 – 61). Median ICI to EVP interval was 6 months (range 0.2 – 79.9). Pts with prior ICI had inferior survival compared to ICI-naive pts, and although the difference in ORR did not reach statistical significance, a trend toward lower ORR was observed in the prior-ICI cohort (Table). Among ICI-exposed pts, outcomes did not differ by ICI–EVP interval, prior pembrolizumab exposure, prior ICI setting, or progression on prior ICI. These findings were also consistent across adjuvant and metastatic ICI settings. Conclusions: In this multisite retrospective analysis, prior ICI exposure was associated with inferior survival with EVP in pts with aUC. The ICI–EVP interval was not associated with outcomes, suggesting that timing of EVP after prior ICI might not influence EVP efficacy. Prospective validation of our hypothesis-generating findings is warranted. ORR Median PFS Median OS Prior ICI (N=68) 47% (28/59)(95% CI 34 – 61) 6.7 mos(95% CI 5.3 – 8.5) 13.3 mos(95% CI 10.8 – NR) ICI-naive (N=425) 57% (211/369)(95% CI 52 – 62) 11.7 mos(95% CI 8.7 – 16.1) 21.9 mos(95% CI 18.6 – 47.8) Prior ICI vs ICI-naive OR: 0.68(95% CI 0.39 –1.18), p = 0.2 HR: 1.60(95% CI 1.16 –2.19), p = 0.02 HR: 1.53(95% CI 1.06 – 2.21), p = 0.003