Second-line tyrosine kinase inhibitors (TKIs) versus immunotherapy (IO) for advanced hepatocellular carcinoma (HCC): Real-world efficacy and safety analysis in patients with varying liver dysfunction.

Abstract

271 Background: Within the last 4 years, the FDA approved five drugs as second-line therapies for advanced HCC. At our cancer center, the majority of our patients are Hispanic with varying liver dysfunction, and this real world population is not represented in registration trials of TKIs and IOs. Therefore, we analyzed survival and toxicities among second-line therapies for HCC at our Hispanic-majority NCI-designated cancer center. Methods: Retrospective analysis of patients with advanced HCC diagnosed at Mays Cancer Center from 1/2015-3/2019 who received second-line therapies, including IO (i.e. nivolumab), TKIs (i.e. cabozantinib, regorafenib), or hospice/best supportive care (BSC). Progression-free survival (PFS) was determined using Kaplan-Meier method, and hazard ratios estimated with Cox proportional hazards model. AEs according to Common Terminology Criteria for AEs v5.0 were analyzed with Fisher’s exact test. Results: Of the patients receiving first-line therapy, the median age was 60 years (n=65), and patients were 75% (n=49) Hispanics. 58 (89%) patients went onto receive second-line therapy. Child-Pugh (CP) score: A 17%, B 55%, C 28%. Median PFS was 3.1 months with TKI (n=6), 3.3 months with IO (n=27), and 1.3 months with BSC (n=25) (Table). There was improved survival with IO when compared to BSC (HR=3.26; 95% CI: 1.58-6.72; p=0.00136). There was no significant difference when comparing IO to TKI (HR=0.94; 95% CI: 0.31-2.86; p=0.92), but a trend to improved PFS with TKI when compared to BSC (HR=3.08; 95% CI: 0.96-9.84; p=0.06). TKI group had significantly more rash (p=0.01) and hand-foot syndrome (HFS) (p<0.001) compared to IO and BSC. All other AEs demonstrated no significant difference between groups (Table). Conclusions: In our Hispanic-majority cohort, patients with varying liver dysfunction, including CP B & C cirrhosis, were more likely to receive IO or BSC. Both second-line treatment groups (IO or TKI) had increased mPFS compared to BSC. Both IO and TKI groups were tolerable compared to BSC, with expected toxicity per class of drug. More prospective studies comparing second-line agents should be done in patients with varying liver dysfunction to understand survival, tolerability, and quality of life. [Table: see text]