Sorafenib vs surgical resection for hepatocellular carcinoma with macrovascular invasion: A propensity score analysis

Background & Aim: Sorafenib is the standard of care for patients with hepatocellular carcinoma (HCC) and macrovascular invasion (MVI), with limited survival. Retrospective surgical studies have reported prolonged survival in this situation. This study aimed to compare the overall survival of patients with HCC and MVI treated with surgical resection or sorafenib. Methods: A total of 143 patients with HCC and MVI but no extra-hepatic spread, treated with surgical resection (SR-patients; n=75) or sorafenib (SOR-patients; n=68) in four French centres between 1990 and 2013 were reviewed retrospectively. A propensity score analysis was performed to reduce bias. Results: SR-patients were significantly younger and had a lower tumour burden than SOR-patients. Median overall survival (OS) rates were 10.1 months [95% CI: 4.1-16.1] in SR-patients and 12.9 months [95% CI: 7.9-17.9] in SOR-patients (P=.959). The 90 day mortality rate was 16% (n=12) in SR-patients and 7.5% (n=5) in SOR-patients (P=.196). SR-patients had a median disease-free survival of 4.60 months [95% CI: 3.35.9]. Under the propensity analysis, median OS was 12.0 months [95% CI: 5.5-18.5] in SR-patients vs 9.7 months [95% CI: 6.1-13.3] in SOR-patients (P=.682). Under multivariate analysis, extensive MVI (HR=1.956, P=.024) and bilirubin >17 mu mol/L (HR=1.738, P=.011) were the two factors significantly associated with mortality. Conclusion: Under a propensity score analysis, the overall survival of patients with HCC and MVI undergoing surgical resection was similar to that achieved with sorafenib. We were not able to identify a patient subgroup experiencing a surgery-related improvement in survival, and quality of life was not evaluable.