Adjuvant hepatic arterial infusion chemotherapy is beneficial for selective patients with Hepatocellular carcinoma undergoing surgical treatment

Abstract Background Recurrence rate after curative surgical resection of Hepatocellular carcinoma (HCC) remains high. Postoperative hepatic arterial infusion chemotherapy (HAIC) has been suggested to improve survival. This study is to investigate the efficacy of HAIC in the patients with poor tumor factors such as vascular invasion or multiplicity. Methods From 2006 to 2014, 221 patients with HCC undergoing hepatectomy and pathologically staged as ≧ T2 (American Joint Committee on Cancer TNM staging system, 7th edition) were included. 61 patients received adjuvant HAIC with 5-fluorouracil, cisplatin, and epirubicin. 160 patients received surgery alone. The overall survival time (OST) and disease free survival time (DFST) were compared between the two groups. Results In all patients, the multivariate analysis of survival data showed that resection margin less than 10 mm was the independent poor prognostic factors. The median OST and DFST between the HAIC and surgery alone groups were 56.4 vs. 56.9 months (p = 0.76), and 50.6 vs. 54.5 months (p = 0.905), respectively. There was no significant difference. For patients with multiple tumors and concomitantly microvascular invasion, the OST was better in the HAIC group (69.7 vs. 54.6 months, p  Based on the image and operative finding, we classified multiple HCC's into two types. Type A: multiple small nodules were close to each other or a huge tumor with several satellite nodules. Type B: two or more tumors scattering in separate segments. Our study showed that type A group benefits from adjuvant HAIC much more than type B. (the median OST in type A versus type B were 85.06 vs. 41.53 months, p = 0.0036). Conclusion The surgical outcome for the patients with multiple HCC's and vascular invasion was poor. Our study showed adjuvant HAIC was beneficial in these patients and formed the basis for further randomized controlled trials.