Results of an aggressive approach to resection of locally recurrent rectal cancer.

ABSTRACT

BACKGROUND: The value of resection for locally recurrent rectal cancer (LRRC) remains controversial. We analyzed outcomes of an aggressive approach to resection of LRRC. METHODS: We conducted a retrospective chart review of 52 consecutive patients who underwent resection of LRRC from September 1997 through August 2005. Overall and disease-free survival (OS, DFS) curves were constructed by the Kaplan-Meier method, and compared by log-rank analysis. Median follow-up time was 29 months (range 3-72). RESULTS: Thirty-one patients (60%) were male. Median age was 60 years (range 36-88). Forty-six of the 52 patients were resected with curative intent, while 6 had known distant metastases at the time of resection. All 52 patients underwent grossly complete resection of local disease, and 41 (79%) had microscopically clear resection margins. An en bloc sacrectomy was performed in 28 (54%) patients. Postoperative mortality was nil; significant complications developed in 42% of patients. The complication rate was higher in patients with sacrectomy than without (50 vs. 33%, P = 0.017, Chi square). For the entire cohort of 52 patients, median OS and DFS were 40 and 24 months, respectively. Survival was equivalent in patients with and without sacrectomy. In the 46 patients who had resection with curative intent, 4-year OS was 48%. Median OS in the six patients with distant metastases at the time of resection was 21 months. OS was predicted by the presence of metastases (P = 0.01), and margin status (P < 0.0001). DFS was predicted by margin status (P = 0.0001). CONCLUSIONS: In this series of patients who underwent resection of LRRC, microscopic margin status was the most significant predictor of OS and DFS. Requirement for en bloc sacrectomy was not associated with inferior survival. Carefully selected patients with distant metastases may benefit from resection of LRRC.