Systematic review and meta-analysis of the influence of circumferential resection margin involvement on survival in patients with operable oesophageal cancer

Background: The prognostic role and definition of circumferential resection margin (CRM) involvement in operable oesophageal cancer remain controversial. The College of American Pathologists (CAP) and Royal College of Pathologists (RCP) define CRM involvement as tumour found at the cut resection margin and within 1 mm of the cut margin respectively. This systematic review and meta-analysis was performed to determine the influence of CRM involvement on survival in operable oesophageal cancer. Methods: PubMed, MEDLINE and the Cochrane Library (January 1990 to June 2012) were searched for studies correlating CRM involvement with 5-year mortality. Statistical analysis of dichotomous variables was performed using the odds ratio (OR) as the summary statistic. Results: Fourteen studies involving 2433 patients with oesophageal cancer who had undergone potentially curative oesophagectomy were analysed. Rates of CRM involvement were 15·3 per cent (173 of 1133) and 36·5 per cent (889 of 2433) according to the CAP and RCP criteria respectively. Overall 5-year mortality rates were significantly higher in patients with CRM involvement compared with CRM-negative patients according to both CAP (OR 4·02, 95 per cent confidence interval (c.i.) 2·25 to 7·20; P < 0·001) and RCP (OR 2·52, 1·96 to 3·25; P < 0·001) criteria. CRM involvement between 0·1 and 1 mm was associated with a significantly higher 5-year mortality rate than CRM-negative status (involvement more than 1 mm from CRM) (OR 2·05, 95 per cent c.i. 1·41 to 2·99; P < 0·001). Conclusion: CRM involvement is an important predictor of poor prognosis. CAP criteria differentiate a higher-risk group than RCP criteria, but overlook a patient group with similar poor outcomes.