Background: Lenvatinib (L, E7080) is an oral multi-targeted tyrosine kinase inhibitor of VEGFR1-3, FGFR1-4, PDGFRβ, RET and KIT. Given the role of angiogenesis in hepatocellular carcinoma (HCC), a phase 1/2 open-label study evaluated the safety and efficacy of L in 46 patients with advanced disease and Childs-Pugh (CP) A liver function status (Kudo ILCA 2013). Patients were treated with a starting dose of L 12 mg qd (28-d cycles) until disease progression or development of unmanageable toxicities. Median time to progression (TTP) was 12.8 months (mo; 95% confidence interval [CI] 7.23–14.7) and median overall survival (OS) was 18.7 mo (95% CI 12.8–25.1). The most common adverse events were hypertension 76% (Gr3 54%), palmar-plantar erythrodysesthesia syndrome 61% (Gr3 7%), proteinuria 59% (Gr3 20%), anorexia 57% (Gr3 2%), thrombocytopenia 50% (Gr3 33%), and fatigue 48% (Gr3 0%). Overall response rate (ORR) was 37%; 45.7% had stable disease. Methods: Based on these phase 2 data, a global, randomized, open-label phase 3 trial was designed to determine if L is non-inferior or superior compared to sorafenib (S) in advanced HCC. Eligible patients (N=940) with Barcelona Clinic Liver Cancer Stage B or C HCC, CP A status, and ECOG 0-1 will be randomized 1:1 to either L 12 mg or 8 mg orally qd (based on body weight [BW]) or S 400 mg orally bid. Patients will be stratified by region; macroscopic portal vein invasion, extrahepatic spread, or both; ECOG-PS; and BW (<60 vs ≥60 kg). The primary endpoint is OS. Secondary endpoints include progression-free survival, TTP, ORR (modified RECIST criteria), safety and PK/PD. Given an estimated median OS for S of approximately 10 mo, a 2.5 mo improvement was derived to achieve a hazard ratio (HR) of 0.8. Statistical study power (using a non-inferiority test by the 95% CI lower-limit method on log HR for OS) was determined based on this HR and a non-inferiority margin of 1.08, corresponding to 60% retention of S effect vs placebo. Based on these assumptions, the study power to declare non-inferiority or superiority is approximately 97% and 82%, respectively. The overall false positive rate is 0.05 (2-sided). Clinical trial information: NCT01761266.