Background: Oral multikinase inhibitors that have shown improvements in overall survival (OS) in HCC are associated with clinically important toxicities that commonly require dose adjustment or discontinuation (D/C) due to intolerance. REACH and REACH-2 studied RAM in patients (pts) with HCC who progressed on or were intolerant to sorafenib (SOR), and REACH-2 only enrolled pts with baseline AFP ≥400 ng/mL. In REACH-2 RAM treatment (trt) improved OS compared to placebo (P), supporting findings in REACH pts with baseline AFP ≥400 ng/mL. An exploratory analysis of outcomes by reason for D/C of SOR was performed. Methods: Pts had advanced HCC, Child-Pugh A, ECOG PS 0-1, and prior SOR. Pts were randomized to RAM 8 mg/kg or P Q2W. A pooled independent pt data analysis (stratified by study) of REACH-2 and REACH pts (AFP ≥400 mg/mL) was performed. Results are reported by reason for SOR D/C (intolerance or disease progression). OS and PFS were evaluated using Kaplan-Meier method and Cox proportional hazard model. Objective response rate (ORR), disease control rate (DCR) and safety are reported. Results: Baseline characteristics in the pooled population were generally balanced between trt arms in each subgroup. Median durations of prior SOR were 2.5 mo for SOR intolerant (n = 70) and 4.0 mo for SOR progressors (n = 472). Median OS (RAM v P) was 10.2 v 6.7 mo for SOR intolerant and 8.0 v 4.7 mo for SOR progressors (Table). Rates of D/C due to trt-related adverse events (AEs) (Table) (7% in each subgroup), and Grade ≥3 AEs (most frequently hypertension) were consistent with those observed in each study. Conclusions: Acknowledging limitations of sample size, the RAM trt benefit in SOR intolerant pts was consistent with that in the ITT population. RAM was well tolerated in SOR intolerant pts with low rates of D/C due to related-AEs. Clinical trial information: NCT01140347, NCT02435433