Study designs of evaluations included in the review
Randomised controlled trials (RCTs), specifically phase III trials, were eligible for inclusion. Randomised phase IIb trials, quasi-randomised studies, non-randomised trials, case-series, case-control and other observational studies were excluded.
Specific interventions included in the review
Studies of medical therapies used as primary treatment were eligible for inclusion. Studies that compared two or more antitumoral treatments, or treatment versus conservative management or suboptimal therapies, were included. Locoregional and systemic treatments were used in the included studies.
Resection was compared with percutaneous ethanol injection (PEI). Radiofrequency ablation (RF) was compared with PEI or percutaneous acetic acid injection (PAI) or microwave coagulation, and microwave coagulation plus Chinese herbal medicine was compared with microwave coagulation alone. Embolisation plus PEI or RF was compared with PEI or RF alone, transarterial chemoembolisation (TACE) plus PEI was compared with TACE alone. Lipiodol-adriamycin was compared with lipiodol-carboplatin. Tamoxifen and octreotide were compared with placebo. Tamoxifen with and without octreotide was compared with anti-androgen therapies. Doxorubicin was compared with combination chemotherapy.
Participants included in the review
Studies in patients with HCC were eligible for inclusion. Trials including patients with liver metastases were excluded. Patients with hepatitis C and B virus and alcohol aetiologies were included in the trials reviewed. In the one trial of surgery, 78% of the participants had cirrhosis, and approximately 40% had a tumour size greater than 2 cm. Where reported, all the participants in the other trials of percutaneous treatments had cirrhosis and 40 to 72% had tumours greater than 2 cm. The proportion with cirrhosis was not reported in all studies of systemic treatments, but where reported it varied from 44 to 93%. Only three of the studies of systemic treatments reported performance status: either all or most of the participants had scores of 0 or 0.5.
Outcomes assessed in the review
Outcomes were not defined in the selection criteria. The main outcome of interest appeared to be survival, which was the primary outcome in most of the included studies. Studies of locoregional treatments variously reported 1-, 2- and 4-year survival, as well as local recurrence and complete necrosis. Most of the studies of systemic treatments reported 1-year survival. Adverse effects were not reported systematically in the review.
How were decisions on the relevance of primary studies made?
The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.