The review concluded that, compared with hepatic resection or local ablation surgery alone, postoperative treatment with adoptive immunotherapy seemed to increase disease-free survival and reduce recurrence rate in liver cancer (hepatocellular carcinoma) patients. The authors acknowledged that the results should be interpreted with caution due to limitations in the evidence base; the reliability of their conclusions is unclear.

To assess the clinical efficacy and safety of adjuvant adoptive immunotherapy for postoperative hepatocellular carcinoma patients.

MEDLINE, EMBASE and The Cochrane Library were searched up to May 2011 with no date or language restrictions; search terms were reported. Reference lists of retrieved studies were manually searched.

Randomised controlled trials (RCT) that compared the effect of adoptive immunotherapy with no active treatment for postoperative hepatocellular carcinoma were eligible for the review. Trials that included patients with liver metastases or postoperative recurrent hepatocellular carcinoma were excluded. Primary outcomes of interest were overall survival, disease-free survival and recurrence. Secondary outcomes included adverse events, withdrawals and discontinuations.

In the included trials, most participants had a Child-Pugh score of A; the remainder had a score of B (where reported). Included patients mostly underwent hepatic resection surgery, but one trial included patients who had transarterial chemoembolisation sequentially combined with radiofrequency ablation. Adjuvant adoptive immunotherapy included anti-CD3-activated peripheral blood lymphocytes, cytokine-induced killer cells, and lymphokine-activated killer cells. One trial had two treatment arms, with three and six courses of adoptive immunotherapy. Recurrence was measured and assessed the same way in the included trials using at least two imaging methods. The trials were all conducted in Asia.

Two reviewers independently selected trials for the review, with disagreements resolved by consultation with a third reviewer.

Trials were assessed for quality based on the Quality of Reporting of Meta-analyses (QUORUM) statement. Criteria included sequence generation of randomisation, allocation concealment, blinding of outcome assessors, and reporting of intention-to-treat analysis. Trials were considered of low quality if they reported none of the criteria, of moderate quality if they reported on fewer than three criteria and of high quality if they reported on three or four criteria.

Two reviewers independently assessed studies for quality, with discrepancies resolved by consensus.

Data were extracted on the outcomes according to how they were analysed in the individual trials.

Two reviewers independently extracted data, with discrepancies resolved by consensus.

Trials were described in a narrative synthesis, with the results presented in a table. Results from the included trials were compared with five other studies (two RCTs, two clinical trials and one case report).

Four RCTs (423 patients) were included in the review. Overall quality was considered moderate in two trials and low in two trials. None of the trials reported allocation concealment or blinding of assessors. Only one trial had intention-to-treat analyses. Mean follow up ranged from 1.5 to over five years (where reported).

Five-year disease-free survival rate (two trials): Compared with no treatment after surgery, both trials found significantly increased five year disease-free survival with adoptive immunotherapy (p<0.05; results from in Table 3).

One-year recurrence rate (two trial): Compared with no treatment after surgery, both trials found significantly reduced recurrence rate at one year with adoptive immunotherapy (p<0.05; results from in Table 3).

Overall survival (three trial): There was no evidence of statistically different overall survival between treatment and control groups at 1.5 years (one trial) or at five years (two trials) after surgery.

Safety: Twenty-five percent of patients treated with adoptive immunotherapy developed fever (reported by three of four trials), which was well controlled with antipyretic agents (no comparison was made with the no treatment group). All adverse events were grade 1 or 2 and were self limiting.

Compared with hepatic resection or local ablation alone, adjuvant adoptive immunotherapy appeared to increase disease-free survival and reduce recurrence rate in postoperative patients with hepatocellular carcinoma.

The review addressed a clear research question, supported by appropriate inclusion criteria. A range of relevant sources was searched with no language or date restrictions. However, no specific attempts were made to find unpublished studies, so it was possible some trials may have been missed. Appropriate methods were used to select studies, extract data and assess studies for quality, which minimised the chance of reviewer error or bias.

A limited number of trials were identified, all with small sample sizes and none were of high quality, so the reliability of the results was unclear. A narrative synthesis of trials was appropriate as there was marked clinical heterogeneity between trials (operation methods, dosage and types of drugs/cytokines). It was not clear why the authors compared the findings of the included trials with results from other studies; two of these studies were also randomised and some results for overall survival were at variance with the findings from the trials included in the review. Selective outcome reporting seemed likely as no analyses were undertaken of adverse events.

The authors acknowledged that their results should be interpreted with caution due to limitations in the evidence base; the reliability of their conclusions is unclear.

Practice: The authors stated that a multi-modality treatment strategy was needed for patients with hepatocellular cancer that used a combination of interventional and immune-based therapies.

Research: The authors stated that a double blind-multicentre phase III RCT with a large sample size and adequate follow up was needed to assess other important endpoints, such as quality of life, duration of hospital stay and cost effectiveness.

National Natural Science Foundation of China.