Comparison of circumferential margin involvement between restorative and nonrestorative resections for rectal cancer

To study circumferential margin involvement (CMI) in patients undergoing restorative, compared with nonrestorative, surgery for rectal cancer.

Descriptive multicentre study, using routinely collected clinical data from the Association of Coloproctology of Great Britain and Ireland (ACPGBI) Bowel Cancer Audit database. The study encompassed 1403 newly diagnosed patients with rectal cancer undergoing either restorative (anterior resection (AR)), or nonrestorative (abdominoperineal excision of rectum (APER) or Hartmann's procedure), surgery. Operations were carried out in 39 hospitals during a variable period between April 1999 to March 2002. A logistic regression analysis was used to control for variables associated with circumferential margin involvement.

One thousand and thirty-six patients satisfied the inclusion criteria. The average CMI was 12.5% (range 0-33.3% between hospitals). CMI for anterior resection was 7.5% (n = 629) compared with a CMI of 16.7% for APER (n = 306) and a CMI of 31.7% for Hartmann's procedure (n = 101); P < or = 0.001. CMI for patients undergoing curative surgery was 7.1% (423 anterior resections, CMI 3.8% (n = 16); 181 APER, CMI 13.3% (n = 24); 29 Hartmann's procedure, CMI 17.2%). On multivariate analysis, having controlled for Dukes' stage and operative intent, the CMI was significantly different between APER and AR (odds ratio 3.3, 95%CI 2.0-5.4), but less so between Hartmann's procedure and AR (odds ratio 2.2, 95%CI 1.1-4.2).

APER is associated with a significantly higher CMI than anterior resection. Attention to surgical technique, with a wide perineal dissection and the use of pre-operative adjuvant therapy, may reduce CMI in patients undergoing APER.