Epidermoid carcinoma of the anal canal is an uncommon disease, and most institutions have only a small series of patients. The current study of a large series of patients treated with radiation therapy in a single institution evaluates the outcome, prognostic factors, and the late complications for these patients.

METHODS. From 1972 to 1991, 270 patients with anal canal epidermoid carcinoma without evident distant metastasis were irradiated with curative intent in the Radiotherapy Department of Tenon Hospital. The sex ratio was 1 man/5.7 women, with a mean age of 67.5 years. The histology included 59.6% well-differentiated epidermoid carcinoma, 32.2% moderately or poorly differentiated epidermoid carcinoma, and 8.2% cloacogenic. The T-classification was: T1: 8.5%; T2: 51.1%; T3: 30.4%; T4: 10%.

Abnormal inguinal lymph nodes were present in 12.5% of the patients. Patients were irradiated by external beam. They received a first course of photon irradiation consisting of (mostly 18 mV or 25 mVl; some Co60 or 6 mV) 40-45 Gy (box technique) in the pelvis for 4-5 weeks.

After a rest of 4-6 weeks, a second course of 15-20 gy in 2 weeks was given through a perineal field by an electron beam of suitable energy. When rectal involvement was important, a four-field, small box technique was used. Fourteen patients were given a booster irradiation of 30 Gy by interstitial brachytherapy (Iridium 192 sources), and four patients were treated with interstitial brachytherapy alone, to a mean dose of 62.5 Gy.

RESULTS. At 5 and 10 years, determinate survival rates were: T1: 86% and 86%; T2: 86.2% and 82.5%; T3: 60.1% and 56.8%; T4: 45% and 45%, respectively. The overall local control rate was 80%. The overall anal conservation rate was 67%. In 154 patients (57%), the anus had maintained its normal function. At 5 and 10 years, determinate survival was 76% and 73.7%, respectively, for N0 and 53.5 and 53.5% for clinically involved inguinal lymph nodes. According to the log-rank test, survival comparisons between T2 and T3 classifications and of tumor sizes less than or equal to 4 cm in length and greater than or equal to 5 cm in length were significant (P = 0.0001 and P &lt: 0.0001, respectively). The presence of clinical abnormal inguinal lymph nodes had a significant negative influence on survival rates (P = 0.047).

Multivariate analysis indicated that T-classification and tumor size in centimeters were the only predictive variables. Nonpredictive variables included nodal status, histology, age, total dose, overall treatment time, and irradiation technique. The grade 3 complication rate requiring surgical treatment was 27/270 (10%), considering all patients (27/190 represents a 14% rate for patients who had local tumor control after radiation therapy alone without secondary salvage amputation). There was no significant relationship between complication rate and the aforementioned variables. Because of the homogeneity of the irradiation doses, no significant relationship was found between dose, local control rate, or complication rate.

CONCLUSIONS. After radiation therapy, recognizing the distinction between tumor sizes of less than or equal to 4 cm in length and more than 4 cm in length (which is not considered in TNM Classification criteria [International Union Against Cancer, 1987]) could help to improve treatment strategies. For tumors more than 4 cm in length and/or with clinically involved lymph nodes, the treatment should be more extensive with combined chemotherapy and radiation therapy, but the increased local control with the addition of cytotoxic chemotherapy to irradiation has not been proven.

Touboul E, Schlienger M, Buffat L, Lefkopoulos D, Pène F, Parc R, Tiret E, Gallot D, Malafosse M, Laugier A.

Service de Radiothérapie A et B, Hôpital Tenon, Paris, France.
Cancer. 1994 Mar 15;73(6):1569-79.