Squamous-cell Carcinoma of the Anal Canal: Predictors of Treatment Outcome

Purpose The incidence of anal canal squamous-cell carcinoma is increasing. Limited data exist on predictors of treatment failure. This study was designed to identify predictors for relapse/persistence after first-line therapy. Methods Using one database, we identified 131 Stages I-III patients treated for primary anal canal squamous-cell carcinoma at our institution from December 1986 to August 2006, with minimum six-month follow-up. Demographic, pathologic, treatment, and outcome data were extracted. Treatment failure was defined as biopsy-proven persistence or relapse (local and/or distant). Univariate, bivariate, and multivariate survival analyses were performed. Results Of 131 patients (median age, 58.3 years; median follow-up, 2.9 (range, 0.6–11.2) years), 66 percent were females, 43.5 percent were Stage II, and 11 (8 percent) were HIV-positive. Surgery only (local excision) was uncommon (6.9 percent, n = 9). One hundred twenty-two patients (93.1 percent) received radiotherapy; two required preradiotherapy diversion. Although 114 (93.4 percent) completed radiotherapy, most required treatment breaks, making total duration of radiotherapy longer than planned. Almost all patients undergoing radiotherapy (96.7 percent, 118/122) also had chemotherapy: 118 (100 percent, Stages I-III) had concurrent chemotherapy: (98 (83.8 percent) mitomycin/5-fluorouracil, 12 (10.2 percent) cisplatin/5-fluorouracil, 8 (6.8 percent) 5-fluorouracil alone); 35 of 46 (76 percent) Stage III patients received induction chemotherapy (34 (97.1 percent) cisplatin/5-fluorouracil, 1 (2.8 percent) 5-fluorouracil alone). Many (44 percent Stages I/II, 48.9 percent Stage III) required dose adjustments. Thirty-seven patients (28.2 percent) failed first-line therapy. There were no differences between patients with relapse (n = 22) or persistence (n = 15) of disease. Bivariate analyses demonstrated that T stage (P = 0.0019), completion of radiotherapy, and total radiotherapy dose (P = 0.03) were all significantly associated with treatment failure. On multivariate analyses, disease stage (P = 0.05) and completion of radiotherapy (P = 0.01) remained significant predictors of relapse-free survival. Conclusions Tolerance of chemoradiation seems to be an important predictor of treatment success. Effective therapies with less acute toxicity must be identified. Dr. Temple is funded by the Society of University Surgeons and by the American Society of Clinical Oncology. Read at the meeting of The American Society of Colon and Rectal Surgeons, June 2 to 6, 2007. No reprints available. An erratum to this article can be found at