Sorafenib versus Hepatic Arterial Infusion Chemotherapy as Initial Treatment for Hepatocellular Carcinoma with Advanced Portal Vein Tumor Thrombosis

Objective: To investigate the validity of hepatic arterial infusion chemotherapy with low-dose 5-fluorouracil and cisplatin (LFP) versus sorafenib as first-line treatment for hepatocellular carcinoma (HCC) with portal vein tumor thrombosis (Vp3, Vp4). Patients and Methods: We retrospectively reviewed the cases of Child-Pugh A advanced HCC with Vp3 or Vp4 treated with LFP or sorafenib between October 2002 and December 2013. Results: There were 32 patients in the LFP group and 14 patients in the sorafenib group. The objective response rate/disease control rate was 31.3/56.3% in the LFP group and 0/28.6% in the sorafenib group. The median survival time (MST) (309 vs. 120 days; p = 0.009) and the median time to treatment failure (109 vs. 37 days; p = 0.022) were significantly longer in the LFP group than in the sorafenib group. In the LFP group, a relatively favorable outcome (MST, 622 days) was obtained among the response cases. Among the nonresponse cases in the LFP group, at the time of cessation of LFP, 70.4% of cases were Child-Pugh A and 88.9% of cases maintained a score of ≤7 points; of the cases in whom Child-Pugh A was maintained, the survival period from the time of LFP discontinuation was significantly longer in the cases in whom sorafenib was introduced as a secondary treatment after LFP than in the cases treated with best supportive care (220 vs. 89 days; p = 0.002). The main adverse event with LFP was grade 3 or higher cytopenia, which was manageable, and adverse event-induced discontinuation was significantly lower as compared with sorafenib (p = 0.002). Conclusion: For the treatment of HCC with Vp3/Vp4, it is desirable to initially use LFP and then immediately change to sorafenib if no response is obtained.