3:00 PM Abstract No. 32 Ablation versus surgical resection for small hepatocellular carcinoma: a risk-adjusted SEER-Medicare analysis

Abstract

To compare the all-cause survival, cancer-specific survival and progression-free survival of surgical resection (SR) versus ablation (AB) for small hepatocellular carcinoma. Our sample consisted of treatment-naïve patients who underwent AB or SR for small HCC (≤5 cm) from 2009-2016 in the SEER-Medicare database. Patients who had a non-HCC malignancy diagnosis within 5 years, lost entitlement 1 year prior to treatment or underwent liver transplant were excluded. Unadjusted and Cox regression model analysis of all-cause survival, cancer-specific survival and progression-free survival were compared. Cox regression model included basic demographic factors, socioeconomic factors, medical comorbidities, and liver disease prognostic indicators. A total of 767 patients who received AB and 214 who received SR met our study inclusion criteria. There were no significant differences between groups for age, gender, socioeconomic factors or the number of medical comorbidities. However, there were significant differences in prognostic indicators for and etiology of liver disease (all P values < 0.05). Unadjusted and adjusted all-cause survival, cancer-specific survival and progression-free survival were all superior in the SR group with all P values <0.001. With SR as a reference group, the unadjusted HRs (95% CI) for all cause survival, cancer-specific survival and disease-specific survival following AB were 2.32 (1.82-2.94), 2.45 (1.70-3.53), and 2.48 (2.01-3.06), respectively. The fully adjusted HRs (95% CI) for all cause survival, cancer-specific survival and disease-specific survival following AB were 1.83 (1.41-2.36), 2.28 (1.54-3.38) and 2.08 (1.65-2.62), respectively. In this national sample of patients with small hepatocellular carcinoma, compared to AB, SR is associated with better all cause survival, cancer-specific survival, and disease-specific survival. After adjusting for known socioeconomic factors, medical comorbidities and liver disease prognostic indicators, the differences between groups decreased. Unmeasured and unaccounted for confounders may account for at least part of the remaining differences between these groups.