The authors concluded that transplantation for patients with early hepatocellular carcinoma, was superior to resection, in some settings, but this was not found in the intention-to-treat analysis, and resection remained a viable option. Despite several review limitations, the authors' conclusions were appropriately cautious and are likely to be reliable.

To compare overall survival following transplantation or resection, in patients with early hepatocellular carcinoma and well-compensated cirrhosis.

PubMed was searched, for articles in English, from 1990 to March 2011; search terms were reported. Bibliographies of relevant articles and reviews were scanned for additional studies.

Eligible studies were those comparing overall survival, in patients with hepatocellular carcinoma and cirrhosis, who were undergoing transplantation or resection. Studies had to report five-year overall survival percentages, and patients had to have tumours that met the Milan Criteria (definition provided). Studies comparing primary versus salvage liver transplantation, and those including patients with non-cirrhotic hepatocellular carcinoma, fibrolamellar hepatocellular carcinoma, or hepatocholangiocarcinoma were excluded.

The patients, in the included studies, were classified as Child-Pugh class A, B or C. Studies were conducted in Europe, the USA, Canada, or Asia, and published between 1999 and 2010. In just over half of the studies, the patients had well-compensated cirrhosis.

The authors did not state how many reviewers selected studies for inclusion.

The authors did not state that they assessed study quality.

Five-year overall survival was extracted and used to calculate odds ratios and 95% confidence intervals. Where possible, data were extracted for the intention-to-treat analysis, in primary studies. The authors did not state how many reviewers extracted the data.

Pooled odds ratios and 95% confidence intervals were calculated, using a random-effects model. Statistical heterogeneity was assessed using Ι².

Meta-regression was conducted to explore the differences between studies, using sample size, year of publication, use of intention-to-treat analysis, and well-compensated cirrhosis, as variables. Publication bias was explored using funnel plots, and the Egger and Begg tests. Subgroup analyses were conducted of studies that included patients with well-compensated cirrhosis, those that used an intention-to-treat analysis, and both of these.

Ten studies were included in the review (1,763 participants). Sample sizes ranged from 37 to 379. All the studies were retrospective. Six studies reported an intention-to-treat analysis.

Overall survival was statistically significantly better for transplant patients than resection patients, with early hepatocellular carcinoma (OR 0.58, 95% CI 0.36 to 0.94; 10 studies). There was significant statistical heterogeneity for this analysis (Ι²=77%). There was no evidence of publication bias, but three studies were outliers on the funnel plot. When these three studies were removed from the analysis, statistical heterogeneity was no longer significant (Ι²=26%). No other variables explained heterogeneity.

The results of the subgroup analyses were similar, except for the analysis that only included studies using an intention-to-treat analysis (six studies), which found no significant difference between treatments.

Transplantation for patients with early hepatocellular carcinoma was superior to resection, in some settings, but this was not found in the intention-to-treat analysis, and resection remained a viable treatment option.

The review question was clear and detailed inclusion criteria were reported. Only one database was searched, and inclusion was limited to studies in English, so some relevant data may have been missed. It was unclear whether appropriate methods to reduce reviewer error and bias were used throughout the review process. Study quality does not appear to have been formally assessed, but the authors evaluated the impact of intention-to-treat analysis. They acknowledged that most of the data were of insufficient quality and confounding factors were not evaluated. The analysis appears to have been appropriate and potential sources of heterogeneity were explored.

Despite several review limitations, the authors' conclusions were appropriately cautious and are likely to be reliable.

Practice: The authors stated that the lack of inferiority for transplantation and resection, in the intention-to-treat analysis, demonstrated that either was viable for patients with early hepatocellular carcinoma and well-compensated cirrhosis.

Research: The authors stated that further well-designed comparative studies were needed, and these should include patients who were matched for the extent of their tumour and underlying liver disease.

Not stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.