Background: In the DAD study (NCT04724018) the combination of SG+EV demonstrated favorable safety determining a maximum tolerated dose of 10 mg/kg and 1.25 mg/kg IV D1,8 every 21 days along with encouraging efficacy in treatment resistant mUC as evidenced by an objective response rate (ORR) of 70% in 23 evaluable patients. Given cumulative toxicities and efficacy across dose levels, the recommended phase 2 dose was SG 7.5 mg/kg and EV 1.25 mg/kg D1,8 of 21-day cycle. As EV+P is now a standard of care for front line urothelial carcinoma, NCT04724018 was amended to include 2 cohorts of distinct patient populations; one expanding SG+EV (DAD) and a second exploring the SG+EV+P (DAD-IO) combination. Methods: In DAD cohort, patients with mUC progressing on platinum and immunotherapy or on one line of therapy receive SG 7.5 mg/kg + EV 1.25 mg/kg intravenously (IV) on day(D)s 1 and 8 every 3 weeks (1 cycle) until progression or intolerable toxicities. In DAD-IO cohort, patients with treatment naïve mUC are given SG 7.5 mg/kg+EV 1.25 mg/kg IV D1, 8 of a 21-day cycle with P 200 mg every 3 weeks or 400 mg every 6 weeks per investigator discretion. In both cohorts, prophylactic granulocyte colony stimulating factor (GCSF) support is given per supportive care guidelines with cycle 1. Following first dose on trial, SG and EV can be reduced and held independently for toxicity with reductions of SG to 5 mg/kg, and EV to 1, 0.75 and 0.5 mg/kg. P dose reductions are not allowed but can be held as needed for immune related adverse events. Laboratory evaluation is performed on days 1 and 8 every cycle. Imaging is performed at baseline and every 6 weeks for the first 4 cycles followed by every 9 weeks thereafter. In both cohorts, the primary endpoint is RECIST 1.1 ORR with secondary endpoints being progression free survival, overall survival, and safety. DAD aims to enroll 41 eligible patients, assuming an ORR of 65% for EV+SG, to detect a 25% improvement with over 90% power at a one-sided type I error rate of 0.05 using a single-stage exact one-sample binomial design; ³ 23 responses would rule out a null of 40% ORR. DAD-IO plans to enroll 41 eligible patients, assuming an ORR of 75% for SG+EV+P to detect a 20% improvement with 90% power at a one-sided type I error rate 0.09 based on a Simon optimal two-stage design. If ³ 11 responses are observed in the first 18 patients, an additional 23 patients will be enrolled. At least 27 responses among a total of 41 will be needed to confirm its efficacy. The study includes a 12-patient safety lead in using a Bayesian toxicity monitoring. If ≤2 dose limiting toxicities (DLTs) occur in the first 6 patients, the next 6 are enrolled. The study halts if ≥5 DLTs occur among 12 evaluable patients. Clinical trial information: NCT04724018.