Purpose: This study was designed to assess the local recurrence rate and prognostic factors for local recurrence in patients undergoing curative anterior or abdominoperineal resections without radiotherapy.
Methods: From January 1980 to December 1996, 514 consecutive patients underwent curative resections for rectal cancer. We excluded those with preoperative radiotherapy (n = 23), postoperative radiotherapy (n = 27), local resection (n = 36), and 11 (2.1 percent) patients who died postoperatively. The remaining 417 patients (249 males) with a median age of 64 (range, 21-90) years were analyzed. For upper third lesions, mesorectal tissue was excised down to at least 5 cm below the tumor. Total mesorectal excision was performed for lower and middle tumors. Postoperative chemotherapy was limited to patients with Stage III lesions. Median follow-up (and 95 percent confidence interval) was (5.2 4.3-5.9) years, with 87.7 percent of patients followed up longer than 24 months. Local recurrence was defined as any recurrence within the field of resection, regardless of the presence or absence of distant metastasis.
Results: Five-year local recurrence rate(and 95 percent confidence interval) was 9.7 (6.4-13) percent, with a median time to diagnosis of 15 (10-23) months. Local recurrence rates in Stages I, II, and III were: 3.1, 4.1, and 24.1 percent, respectively (P < 0.0001). In relation to node status, local recurrence rates were N0, 4.1 (1.7-6.5) percent; N1, 12.6 (4.6-20.6) percent; N2, 32.1 (12.1-52.1) percent; and N3, 59.3 (22.5-96.1) percent; (P < 0.00001). Lower third tumors had a higher local recurrence rate than middle and upper third tumors: 17.9, 7.1, and 5.1 percent, respectively (P = 0.002). Adjusted by stage, this difference was maintained only in Stage III tumors. Among lower tumors, those at 6 and 7 cm from the anal verge had a lower local recurrence rate than those below 6 cm (6.7 vs. 26.2 percent, respectively; P = 0.02). Accidental rectal perforation at or near the tumor site occurred in 12 cases (2.9 percent), showing a strong correlation with local recurrence (P < 0.0001). Multivariate analysis showed significant higher risk for lower third tumors (hazard ratio, 2.98) and positive nodes (hazard ratio, 4.78).
Conclusions: Appropriate surgery without irradiation achieves excellent local control in N0 rectal cancers. Node metastasis, lower third localization (especially below 6 cm), and accidental rectal perforation at or near the tumor site are significantly associated with a higher local recurrence rate.