Randomised controlled trials (RCTs) that compared hepatectomy plus adjuvant transarterial chemoembolisation versus hepatectomy alone in patients with primary hepatocellular carcinoma were eligible for inclusion. Chemotherapeutic drugs could be used in both treatment groups. Patients with any stage of hepatocellular carcinoma were eligible, but they had to have no previous management, no distant metastasis, good or moderate liver function, and no contraindication for laparotomy. Trials had to report on at least one primary outcome. Trials were excluded if they did not use lipiodol or embolic agent, or if they involved intravenous, portal vein or oral chemotherapy, or immune treatment.
Primary outcomes included mortality, disease free survival rate, tumour recurrence, or median survival. Secondary outcomes included adverse events.
The included trials studied transarterial chemoembolisation with lipiodol dose ranging from 2 to 20 mL, in combination with one to five courses of chemotherapy (less than two weeks to 12 weeks postoperative) in patients with hepatocellular carcinoma. The type of chemotherapy included adriamycin, carboplatin, cisplatin, doxorubicin, epirubicin, 5-fluorouracil, or mitomycin. Most of the included patients underwent curative or radical resection, but palliative resection was also included. Participants primarily had stage IIIa disease; the proportion of patients with cirrhosis ranged from 72 to 85% (where reported). In most patients, the diameter of the tumour was over 5cm; some patients had portal vein or hepatic vein invasion.
Two reviewers independently performed study selection and disagreements were resolved by discussion.