Purpose: Endocavitary radiation is a treatment option for selected patients with rectal cancer, but concern exists for the effectiveness of salvage abdominoperineal resection. This study was designed to examine outcomes after salvage abdominoperineal resection for recurrence after endocavitary radiation.
Methods: A prospective database was used to identify patients undergoing abdominoperineal resection after endocavitary radiation from 1985 to 2001. Office records and a tumor registry were used for disease status and survival data. Survival was calculated using the Kaplan-Meier method and groups compared using the Mantel-Haenszel test.
Results: Thirty-eight patients underwent salvage abdominoperineal resection. The mean time to recurrence after completion of endocavitary radiation was 21 +/- 27 months, with 29 percent having persistent disease, 63 percent recurrent disease, and 8 percent a second primary. At abdominoperineal resection, 47 percent had tumor transection, specimen perforation, or injury to the genitourinary or gynecologic tract. Nine patients (24 percent) had positive radial margins. The mean time to perineal wound healing was 56 +/- 74.1 days postoperatively, with 36.8 percent taking more than 60 days. Seventeen patients (45 percent) re-recurred at a mean of 21 +/- 25 months after salvage, with a local control rate of 26 percent at 45 +/- 37 months of follow-up. Median disease-specific survival from completion of endocavitary radiation was 115.5 months, with a five-year, disease-specific survival rate of 66 percent. Patients with recurrent disease after endocavitary radiation had significantly (P = 0.025) better disease-specific survival than those with persistent disease (median survival 115 vs. 25 months).
Conclusions: Although technically difficult and associated with a high morbidity, abdominoperineal resection can salvage a significant fraction (55 percent) of patients failing endocavitary radiation. A high index of suspicion for recurrence and a tenacious approach to its diagnosis are essential for optimal outcomes.