• Rectal cancer;
  • circumferential resection margin;
  • extralevator abdominoperineal resection;
  • low anterior resection;
  • conventional abdominoperineal resection


Aim  Abdominoperineal excision (APR) has been associated with higher circumferential resection margin (CRM) involvement and local recurrence rates than extralevator APR for low rectal cancer. This study aimed to evaluate the CRMs in APR and low anterior resection (LAR) specimens and to identify factors influencing CRM involvement.

Method  All pathological specimens from consecutive patients with rectal cancer who underwent curative resection at the Cleveland Clinic Florida, from January 2000 to July 2010, were reviewed by two pathologists. Demographics, tumour characteristics, operative data, postoperative pathology and Dworak’s tumour regression grade were compared between specimens with positive and negative CRMs.

Results  One-hundred and fifty-four patients underwent curative APR (n = 65) or LAR (n = 69). Mean tumour size was 3.6 cm, and mean distance from the dentate line was 5.4 cm. Nine (6.8%) patients had a positive CRM (n = 6 APR, n = 3 LAR), which was associated with tumour size > 5.9 cm (= 0.002), a distance of ≤ 2.6 cm from the dentate line (= 0.013), microvascular invasion (= 0.009), perineural invasion (< 0.001), number of positive lymph nodes (= 0.046) and incomplete total mesorectal excision (TME) (< 0.001). APR specimens were three times more likely than LAR specimens to have an incomplete mesorectum (9.8%vs 2.9%, = 0.322).

Conclusions  Factors associated with a positive CRM were tumour size > 5.9 cm, a distance of ≤ 2.6 cm from the dentate line, incomplete TME, number of positive nodes and microvascular and perineural invasion. The incidence of a positive CRM was not significantly different between LAR and APR (= 3 LAR and n = 6 APR).