Background: Based on interim results from the phase 3 KEYNOTE-045 (NCT02256436) study comparing pembro and investigator?s choice of chemotherapy (chemo), pembro was approved for the treatment of locally advanced or metastatic UC that has progressed during or after a platinum-containing regimen. Updated results after 2-year follow-up are presented. Methods: Eligible patients (histologically or cytologically confirmed UC, progression after platinum, ECOG PS 0-2, measurable disease per RECIST v1.1, ≤2 lines of systemic therapy) were randomly assigned 1:1 to receive pembro 200 mg Q3W or investigator?s choice of paclitaxel 175 mg/m² Q3W, docetaxel 75 mg/m² Q3W, or vinflunine 320 mg/m² Q3W. Primary efficacy end points were OS and PFS (RECIST v1.1, blinded central review). ORR (RECIST v1.1, blinded central review) was a secondary end point. Results: As of Oct 26, 2017, among 542 enrolled patients (pembro, 270; chemo, 272), median follow-up was 27.7 mo. Median OS was significantly longer with pembro vs chemo (10.3 vs 7.3 mo; HR, 0.70; P< 0.0002). OS benefit with pembro vs chemo was seen in all PD-L1 expression subgroups (HR; combined positive score [CPS] < 1, 0.82; CPS ≥1, 0.58; CPS < 10, 0.75; CPS ≥10, 0.56) and was maintained regardless of age, ECOG PS, prior therapy, liver metastases, baseline hemoglobin, time from last chemo, histology, risk factor group, and choice of chemo. PFS was not different between arms (2.1 vs 3.3 mo; HR, 0.96; P= 0.32). ORR was higher with pembro vs chemo (21.1% vs 11.0%). Median duration of response was longer with pembro (not reached [1.6+ to 30.3+ mo] vs 4.4 mo [1.4+ to 29.9+ mo]), and a greater proportion of responses lasted ≥12 mo (68% vs 35%) per Kaplan-Meier method. Fewer patients with pembro vs chemo experienced a treatment-related adverse event of any grade (62.0% vs 90.6%) and a grade ≥3 adverse event (16.5% vs 50.2%). Conclusions: Results observed over 2-year follow-up, including OS benefit and superior safety with pembro vs chemo, were consistent with the interim analyses that led to the approval of pembro in locally advanced or metastatic UC that progressed during or after platinum-based chemotherapy. Clinical trial information: NCT02256436