Background: Results from cohort 1 of KEYNOTE-224, an open-label, single-arm, multi-country phase II trial, demonstrated that pembro monotherapy was efficacious and tolerable in patients (pts) with advanced HCC previously treated with sorafenib. Here, we report results from KEYNOTE-224 cohort 2, which enrolled pts with advanced HCC and no prior systemic therapy. Methods: Eligible pts in cohort 2 had radiologically, histologically, or cytologically confirmed, incurable HCC not amenable or refractory to locoregional therapy, Child Pugh A liver disease, measurable disease based on RECIST 1.1 by blinded independent central review (BICR), ECOG PS 0-1, and BCLC stage C or B. Pts received pembro 200 mg IV Q3W for ~2 years or until disease progression, unacceptable toxicity, consent withdrawal, or investigator decision. Primary endpoint was ORR (RECIST 1.1 by BICR). Secondary endpoints included DOR, DCR, TTP, PFS, OS, and safety/tolerability. Response was assessed every 9 weeks. Efficacy and safety were assessed in pts who received ≥1 dose of study treatment. DOR was assessed in responders. The estimate and 95% CI of the ORR and DCR were based on the Clopper-Pearson method. Kaplan-Meier method was used to estimate OS, PFS, and DOR. A sample size of ~50 pts was chosen to provide acceptable precision for the assessment of ORR. Results: Cohort 2 enrolled 51 pts. The median time from the first dose to data cutoff (July 31, 2020) was 21 (range, 17-23) mo. The median age of pts was 68 (range, 41-91) years, one pt was HBV+, 80% had alcohol use, 8% were HCV+, 18% had vascular invasion, 35% had extrahepatic disease, 33% had BCLC Stage B disease, and 67% had BCLC Stage C HCC. ORR was 16% (95% CI, 7-29) and was similar across most subgroups. Median DOR was not reached (range, 3-20+ mo); 70% were estimated to have response duration ≥12 mo. Best overall responses were 0 CR, 8 (16%) PRs, 21 (41%) SDs, and 17 (33%) PDs; response was not evaluable or not assessed for 5 (10%) pts. DCR was 57%. The median TTP was 4 (95% CI, 3-8) mo. The median PFS was 4 (95% CI, 2-6) mo, and median OS was 17 (95% CI, 8-NA) mo. PFS rate at 18 mo was 16%, and OS rate at 18 mo was 46%. Treatment-related AEs (TRAEs) occurred in 27 (53%) pts; the most common TRAEs were diarrhea, fatigue, hypothyroidism, and myalgia. Grade ≥3 TRAEs occurred in 7 (14%) pts. TRAEs led to treatment discontinuation in 6% of pts. Immune-mediated AEs and infusion reactions occurred in 11 (22%) pts. One treatment-related death occurred due to myocarditis, with associated immune-related hepatitis. Conclusions: In pts with advanced HCC and no prior systemic therapy, pembro monotherapy provided durable anti-tumor activity, promising overall survival, and demonstrated a safety profile consistent with that previously observed for pembro in advanced HCC. These findings support further evaluation of pembro-based regimens for the treatment of HCC in the frontline setting. Clinical trial information: NCT02702414