Outcomes for Abdominoperineal Resections Are Not Worse Than Those of Anterior Resections

Purpose

Local recurrence and cure rates following abdominoperineal resections have been reported to be much worse than sphincter-preserving anterior resections. We compared the oncologic outcomes of patients treated by abdominoperineal resections with those following sphincter-preserving anterior resections.

Methods

The medical records of patients who underwent radical rectal resection for rectal carcinoma at the Colorectal Surgery Department, Singapore General Hospital, during the period from April 1989 to April 2002 were reviewed. A total of 791 cases were studied. Operative procedures were classified as either abdominoperineal resections or anterior resections with either straight or pouch anastomosis. Total mesorectal excision was routinely performed for carcinomas of the lower middle and lower third of the rectum. Sentinel events, including local and systemic recurrences or morbidity and mortality, were tracked prospectively.

Results

There were a total of 93 abdominoperineal resections (12.1 percent), 547 anterior resections with straight anastomoses (71 percent), and 130 anterior resections with pouch anastomoses (16.9 percent). Postoperative mortality was 2.6 percent and postoperative morbidity was 13.6 percent with an overall anastomotic leakage rate of 2.5 percent. The cumulative five-year local recurrence rate was 5.4 percent for abdominoperineal resections, 3.6 percent for anterior resections with straight anastomoses, and 3.8 percent for anterior resections with pouch anastomoses (P = 0.73 by log-rank test). The median time to local recurrence also did not differ significantly between the different procedures (abdominoperineal resections, 17 months, anterior resections with straight anastomoses, 18 months, anterior resections with pouch anastomoses, 13 months). Independent predictors for local recurrence included advanced tumor stage, tumor depth, and poorly differentiated tumors. The five-year cancer-specific survival was 70 percent. The type of anastomosis did not influence disease-free survival with median disease-free survival for patients who underwent abdominoperineal resections being 100 months, survival of anterior resections with straight anastomoses being 135 months, and survival of anterior resections with pouch anastomoses being 121 months (P = 0.33 by log-rank test). The independent factors for poor survival were age greater than 65 years, advanced tumor stage, tumor depth, and poorly differentiated tumors.

Conclusion

Both abdominoperineal resections and sphincter-preserving anterior resections can be performed safely with low morbidity and mortality in a specialized high-volume hospital unit without compromising oncologic outcomes. With appreciation of the anatomic relations in total mesorectal excision and standardized consistent surgical technique, the oncologic outcomes of patients treated by abdominoperineal resections are not worse than those treated by sphincter-preserving anterior resections.