Factors affecting circumferential resection margin involvement after rectal cancer excision

PURPOSE: This study was designed to assess factors affecting rates of circumferential resection margin involvement after rectal cancer excision, the association between circumferential resection margin involvement rates for patients undergoing anterior resection and abdominoperineal excision within the same unit, and trends in outcomes between units. METHODS: Data about patients undergoing rectal cancer excision between 2000 and 2003 were extracted from the Association of Coloproctology of Great Britain and Ireland database. Multivariate logistic regression analysis was used to identify independent predictors of circumferential resection margin involvement. Pearson correlation coefficient was used to evaluate the association between circumferential resection margin involvement for anterior resection and abdominoperineal excision. RESULTS: A total of 1,430 patients satisfied the inclusion criteria. The circumferential resection margin involvement rate for anterior resection (n = 794) was 6.7 percent, between hospital variability was 0 to 40 percent, and for abdominoperineal excision (n=521) was 17.6 percent, between hospital variability 0 to 100 percent. Independent predictors of circumferential resection margin involvement were T stage (P<0.001), nodal involvement (P = 0.007), and operative procedure (P<0.001). Units with a high circumferential resection margin involvement rate for anterior resection also had a high circumferential resection margin involvement rate for abdominoperineal excision (Pearson correlation = 0.349; P=0.01). CONCLUSIONS: Circumferential resection margin involvement is more common in lymph-node-positive tumors and is more common after abdominoperineal excision compared with anterior resection. This relationship was consistent across units irrespective of their individual circumferential resection margin involvement rates.