ABSTRACT Aim: Tumors use the PD-1 receptor-ligand pathway to evade immune surveillance. The anti-PD-1 monoclonal antibody pembro has demonstrated antitumor activity in advanced solid tumors. We assessed the safety, tolerability, and antitumor activity of pembro in pts with recurrent or metastatic urothelial cancer in the KEYNOTE-012 study (Clinicaltrials.gov: NCT01848834). Methods: Archival or newly obtained tumor samples from pts with advanced carcinoma of the renal pelvis, ureter, bladder, or urethra were screened for PD-L1 expression using a prototype immunohistochemistry assay. PD-L1 expression in stroma or ≥1% of tumor cells was required for study entry. Pts received pembro 10 mg/kg every 2 wk until complete response, progression, or unacceptable toxicity. Pts deriving benefit could remain on pembro beyond initial progression. Response was assessed every 8 wk per RECIST v1.1 by independent central review (primary efficacy end point). Results: 33 pts enrolled, including 30 with transitional cell histology and 3 with nontransitional cell or mixed histology. Median age was 70 y (range 44-85), 70% had ECOG PS 1, 52% received ≥2 prior therapies for advanced disease, and 21% had liver metastases. 22 pts (67%) received ≥3 pembro doses. Median follow-up duration was 11 mo (range 10-13), and 7 pts (21%) remain on therapy. 61% of pts reported ≥1 drug-related AE, most commonly fatigue (n = 6), periphereal edema (n = 4), and nausea (n = 3); 4 pts (12%) reported grade 3-4 drug-related AEs, with only rash seen in >1 pt (n = 2). 29 pts received ≥1 dose of pembro and had a baseline scan with measurable disease and were evaluable for response. ORR by central review was 24% (95% CI 10%-44%), with 3 (10%) complete responses. Response duration is 16 to 40+ wk (median not reached), with 6 of 7 responses ongoing. In the pts evaluable for response, median PFS is 8.6 wk. In all pts, median OS is 9.3 mo (6-mo OS rate, 58%). Analysis of the relationship between PD-L1 expression and pembro efficacy is ongoing. Conclusions: Pembro shows acceptable safety and tolerability and provides promising antitumor activity in pts with advanced urothelial cancer. These data support the continued development of pembro in advanced urothelial cancer. Disclosure: E.R. Plimack: has received research funding from Merck & Co., Inc., Whitehouse Station, NJ; J. Bellmunt: Has served on advisory boards, without compensation, for Merck & Co., Inc., Genentech, and Bristol-Myers Squibb; B. Montgomery: Received research funding from Janssen, Medivation, and Tokai; E.J. Gonzalez, R. Perini, J. Cheng and C. Gause: Is employee of and holds stock in Merck & Co., Inc.; J. Pulini: Former employee of Merck & Co., Inc.; M. Dolled-Filhart, K. Emancipator and K. Pathiraja: Is employee of Merck & Co., Inc. All other authors have declared no conflicts of interest.