Examination of outcome following abdominoperineal resection for adenocarcinoma in Oxford

Abstract

Abdominoperineal resection (APR) has been shown to have poor outcomes compared with anterior resection (AR) in the treatment of rectal cancer. We compared APR outcomes with those for low AR. Lower third rectal cancers treated at the John Radcliffe Hospital with APR and low AR were examined using a prospectively collected database augmented with review of patient records. For all cases (APR and low AR), a range of patient, cancer and outcome data were collected. A selected group was created on the basis of exclusions. Outcomes for the global and selected APR and low AR groups were compared using the Kaplan-Meier method. CRM+ve and CRM-ve APR cases were compared. Between 1994 and 2003, 70 APR and 93 low AR were performed. After exclusions, 42 APR and 81 low AR remained. Median follow-up was 4.8 years. Five year survival for the APR group was significantly worse than for the low AR group. The APR group showed significantly fewer T0 cancers and significantly more T3 cancers. CRM R1 involvement was significantly higher for the APR group. The CRM+ve APR group contained significantly more later stage cancers, more defective resection specimens, more abscesses and fistulas and was associated with more local recurrence. These data showed that APR led to worse results than low AR in terms of overall survival and circumferential margin involvement, but that the cancers treated with APR tended to be more locally advanced.