Outcomes of patients with advanced urothelial cancer who develop infection while on treatment with pembrolizumab.

4573 Background: Over the past decade, studies have shown the benefit of immune checkpoint inhibitors (IO) in patients with advanced urothelial cancer. These agents work by reconditioning the adaptive anti-cancer immune response within the tumor microenvironment. Immune-related adverse events have been well documented, but there is limited data evaluating infections in patients treated with IO. We performed a retrospective analysis to assess the incidence of infection and its effect on morbidity and mortality in patients treated with pembrolizumab for advanced urothelial cancer. Methods: Data was collected from a network of 7 hospitals for patients who received pembrolizumab for advanced urothelial cancer from 1/1/2017-8/1/2021. Date of last follow up was 12/1/2021. Covariates compared among infected and non-infected cohorts included age, gender, race, comorbidities, ECOG, anti-infective therapy at IO initiation, and line of therapy (1L, 2L, > 2L). Univariable analysis with reported odds ratio (OR) and 95% confidence interval (CI) was used to assess risk factors for infection. Outcome measures included all-cause emergency department (ED) visits, inpatient and intensive care unit (ICU) admissions, median number of IO cycles, and overall survival (OS). OS was evaluated using the Kaplan-Meier model. All analyses were deemed statistically significant if the p-value was < 0.05. Results: A total of 51 patients were identified. Of these, 34 (66.7%) had at least one documented infection and 17 (33.3%) had no reported infections. Baseline characteristics were similar across cohorts. Compared to non-infected patients, infected patients received fewer cycles of IO (median 4 vs 8, p = 0.016). At last follow-up, 20 (58.8%) patients in the infected cohort and 4 (23.5%) in the non-infected cohort died (p = 0.017). Median OS was 7.4 months (95% CI: 3.4-24.9) among patients with infection while not reached in those without infection (p = 0.014). ED visits (p = 1.00), inpatient admissions (p = 0.21), and ICU admissions (p = 0.17) did not significantly differ between cohorts. Univariable analysis did not identify significant risks among covariates. Conclusions: The incidence of infection in patients treated with pembrolizumab for advanced urothelial cancer is high and associated with fewer cycles of IO therapy and shorter OS. Further study of infectious process prevention is of value to maximize immunotherapy benefit.