Background: Mutations of COMPASS-related proteins KMT2D, KMT2C, and KDM6A are observed in 66% of advanced urothelial carcinomas (UC). In vivo, the administration of tazemetostat, an enzymatic EZH2 inhibitor (EZH2i), reduced tumor burden and enhanced the immune response in animal models of UC. We hypothesized that tazemetostat might improve response to immunotherapy in advanced UC. Methods: ETCTN 10183 (NCT03854474) is a multicenter pilot study evaluating the safety and efficacy of tazemetostat 800 mg BID+ pembrolizumab 200 mg given every three weeks for up to two years in patients (pts) with platinum-refractory (Arm A) or cisplatin/chemo-ineligible advanced UC (Arm B). The primary objective was to identify the recommended phase two dose (RP2D) of tazemetostat in combination with pembrolizumab. Secondary objectives included safety, response rate, and progression-free survival. Translational objectives included determining tumor mutational burden in COMPASS-related genes, tumor subtyping, TCR clonality, T cell infiltration, and PDL-1 expression. Results: There were no dose-limiting toxicities in the safety lead-in phase (n=6 pts), and the RP2D for tazemetostat was established at 800 mg BID. Baseline characteristics for Arm A N=12, Arm B N=13: 72% males; 96% white, 4% Black; 88% ECOG PS 0-1; 88% had bladder primary, 20% Stage III, 72% stage IV; 80% had visceral or bone metastases. The median number of cycles was 5 in Arm A (range 1-35) and 4 in Arm B (range 2-12). Treatment related G3/4 adverse events (AEs) were observed in 8 patients, most common were anemia and lymphopenia (n=2 each). Treatment related G1/2 AEs included nausea (n=6); skin and subcutaneous tissue disorders (n=3); anemia, lymphopenia, and thrombocytopenia (n=2 each). Among evaluable patients (n=10 in each arm), partial response rate was: Arm A 30%, Arm B 20%; stable disease: Arm A 30%, Arm B 30%; and progressive disease: Arm A 30%, Arm B 20%. 8 pts were on study treatment for ≥ 9 cycles (5 in Arm A, 3 in Arm B). All pts are off protocol treatment. Median progression-free survival was 3.06 (95% CI: 1.38-7.75) months for Arm A and 3.02 (95%CI: 2.4-5.65) for Arm B. Conclusions: Tazemetostat 800 mg BID and pembrolizumab combination is feasible and well-tolerated in patients with advanced UC. In this pilot study, improvement in efficacy relative to historical controls of pembrolizumab alone appears to be modest. Response in tumors with COMPASS-related mutations and EZH2 activity will be evaluated. Clinical trial information: NCT03854474.