Hepatocellular carcinoma etiology drives survival outcomes: A population-based analysis.

Abstract

e16142 Background: Previous studies on survival differences according to etiology of hepatocellular carcinoma (HCC) (5-year observed survival of 23%) are limited to hospital-based series and restricted cohorts. We performed a population-based survival analysis by major HCC etiologies, inclusive of hepatitis-C virus (HCV), hepatitis-B virus (HBV), alcohol-related liver disease (ALD), and non-alcoholic fatty liver disease (NAFLD) as singular causes, in addition to unexplored dual etiological groups, (HCV&HBV, HCV&ALD, and HBV&ALD), and accounting for age at diagnosis, tumor characteristics, and socio-demographics. Methods: All 19,956 cases of HCC diagnosed in Florida during 2005–2018 from the Florida cancer registry were linked for HCC etiology using Agency for Healthcare Administration and Florida Department of Health viral hepatitis data. Survival analysis was performed for the 15,759 linked HCC cases, comparing HCC etiologic groups (4 single and 3 dual). Results included age-adjusted 5-year survival, Kaplan-Meier curves, and adjusted hazard ratios from Cox regression models. Results: The leading single etiology was ‘HCV only’ (n=4,983; 31.6%); the leading dual etiology was ‘HCV&ALD’ (n=2,694; 17.1%). The age-adjusted 5-year survival was low (<22%) across all HCC etiologies; however, non-ALD causes showed higher survival [‘HCV only’ (21.5%; 95%CI:20.2%–22.8%), ‘HBV only’ (20.2%; 95%CI:16.6%–23.9%), ‘NAFLD only’ (19.9%; 95%CI:18.7%–21.2%)] than ALD-related etiologies [‘ALD only’ (14.4%; 95%CI:12.7%–16.0%), ‘HCV&ALD’ (10.4%; 95%CI:9.0%–11.9%), ‘HBV&ALD’ (8.1%; 95%CI:2.2%–14.1%)]. Compared to the reference category ‘HCV only,’ age-adjusted hazard ratios were 1.69 (95%CI: 1.41–2.03; p-value<0.0001) for ‘HBV&ALD,’ 1.33 (95%CI:1.26–1.40; p-value<0.0001) for ‘HCV&ALD,’ and 1.35 (95%CI:1.28–1.43; p-value<0.0001) for ‘ALD only.’ Multivariable analysis including additional factors showed minimal variation from these results. Conclusions: Significant differences in survival based on HCC etiology are important for prevention, surveillance, and targeted treatments based on etiologic-specific biomarkers. Specific pathophysiological mechanisms for ALD in addition to viral hepatitis could be responsible for poorer outcomes for dual etiologies involving alcohol use, compared to viral etiology alone. To increase overall survival, improved screening is needed for patients with multiple HCC risk factors. [Table: see text]