The review concluded that adjuvant interferon therapy after curative therapy for hepatocellular carcinoma improved recurrence-free survival. The most consistent predictors for improved recurrence-free survival were low percentage of multiple tumours and use of ablation therapy. Caution should be exercised in interpreting the authors’ conclusion due to issues regarding analyses and the possibility of studies being missed.

To determine the efficacy, and predictors of efficacy, of adjuvant antiviral therapy after curative therapy for hepatocellular carcinoma.

PubMed and The Cochrane Library were searched up to December 2008; search terms were reported. Conference abstracts from American Association for the Study of Liver Disease were searched from 2005 to 2008. Reference lists of relevant identified papers were checked. Only full papers available in English were included.

Randomised controlled trials (RCTs) and cohort studies that compared interferon-based therapy with a control (no antiviral therapy) in patients with underlying chronic viral hepatitis B (HVB) or hepatitis C (HVC) who underwent curative surgery or ablation therapy were eligible for inclusion. Eligible studies were required to report recurrence-free survival of hepatocellular carcinoma. Studies of unapproved antiviral therapy, vaccination studies, studies of other medical conditions (such as haemodialysis or thalassaemia), studies of post liver transplantation and studies of hepatitis D were excluded.

All studies used conventional interferon (natural or recombinant) as antiviral therapy. Curative therapy included surgery (nine studies) and ablation (eight studies). Most studies enrolled patients with HCV-related hepatocellular carcinoma; four studies enrolled patients with HBV-related hepatocellular carcinoma. Median tumour size and percentage of patients with cirrhosis did not significantly differ across the groups. Median age was 61.3 years (range 50.0 to 68.5 years) for patients in the interferon group and 63.0 years (range 49.0 to 67.5 years) in the control group. Ten studies reported sustained virological response rate (defined as absence of HCV ribonucleic acid for at least 24 weeks after completion of interferon therapy).

Two reviewers independently selected studies for inclusion in the review; any disagreement was resolved through discussion.

Two reviewers independently assessed study quality based on eight criteria taken from Cochrane guidelines: study size, control treatment, allocation generation, allocation concealment, blinding, follow-up, intention-to-treat and sample size estimation. A summed quality score was provided for each study, with a maximum possible score of 12 for RCTs and 8 for cohort studies (higher scores indicated poorer study quality).

Two reviewers independently extracted data on cumulative hepatocellular carcinoma incidence (or data to enable calculation of cumulative hepatocellular carcinoma incidence). Data on potential prognostic factors (patient age, gender, aetiology of underlying liver disease, type of curative therapy, Child-Pugh classification, stage and virological response to antiviral therapy) were extracted.

Studies were combined in a meta-analysis using fixed-effect and random effects models. Summary estimates for recurrence-free survival were reported as percentages with associated 95% confidence intervals (CIs). Meta-regression was also performed to identify the most significant predictors of overall treatment efficacy; basic model fitting techniques for variable selection, assessment of goodness of fit, and regression diagnostics were used. Separate analyses were carried out for RCTs only and combined RCTs and cohort studies. Subgroup analyses were performed for RCTs that enrolled HCV-related hepatocellular carcinoma. Regression diagnostics were used to identify influential or poorly fitted cases in the regression models.

Thirteen studies (n=1,200) were included in the review: nine RCTs (n=917) and four cohort studies (n=283). Quality scores ranged from 3 to 8.

Pooled results from the RCTs found that interferon improved one-year recurrence-free survival by 9.5% (95% CI 3.7% to 15.3%), two-year by 40.3% (95% CI 35.1% to 45.4%) and three-year by 13.4% (95% CI 7.3% to 19.6%). When data from the cohort studies were added, interferon was found to improve one-year recurrence-free survival by 7.8% (95% CI 3.7% to 11.8%), two-year by 35.4% (95% CI 30.7% to 40.0%) three-year by 14.0% (95% CI 8.6% to 19.4%).

When only RCTs were analysed, meta-regression found that the most consistent predictive factor for improved recurrence-free survival was low percentage of multiple tumours and use of ablation therapy. When data from the cohort studies was incorporated a significant interaction between study types and underlying HCV infection was found at year two and at year three. Further results from subgroup and prognostic factors were reported.

Cumulative evidence indicated a beneficial effect of adjuvant interferon therapy after curative therapy for hepatocellular carcinoma in terms of improved recurrence-free survival. A lower percentage of patients with multiple tumours and use of ablation therapy were the most important predictors for better treatment efficacy.

The review addressed a focused question. Inclusion criteria were clearly defined. Two electronic databases were searched. Unpublished studies were sought through a search of conference abstracts. The restriction to English-language studies may have led to papers being missed. Appropriate steps were taken minimise error and bias in study selection, data extraction and assessment of study validity. A formal assessment of study quality was performed using appropriate criteria, but only a summary score was presented. The authors considered the different study designs separately; pooling RCTs and cohort studies may not have been appropriate. The authors acknowledged that interferon regimes varied widely between studies and it was unclear whether heterogeneity was formally assessed. It was unclear whether the reported results were taken from fixed-effect or random-effects models. Meta-regression was not a good way to explore differences in treatment effect; comparison of treatment effects between different individuals should be performed using subgroup analyses. Meta-regression does not prove causality and should be regarded as hypothesis generating. Given these considerations, the authors' conclusions should be interpreted with caution.

Practice: The authors did not state any implications for practice.

Research: The authors stated that further clinical trials should focus on antiviral regimens with better compliance and antiviral activity.

Ministry of Education, Taiwan; National Science Council, Taiwan; Liver Disease Prevention and Treatment Research Foundation, Taiwan; New Century Health Care Promotion Foundation, Taiwan.

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.