We analyzed in a retrospective series of patients treated by conservative irradiation for an epidermoid cancer of the anal canal (ECAC) the prognostic factors of locoregional control (LRC), survival, late severe complications (LSC), and sphincter conservation (SC).
Methods and materials: From 1976 until 1994, 118 patients presenting with an ECAC were conservatively treated (mean age, 65 years). According to the 1987 International Union Against Cancer (TNM) classification, they were: 19 T1, 70 T2, 22 T3, 7 T4, 94 N0, and 24 N1-3. The treatment started with external beam irradiation (EBI) (36 Gy in 3 weeks or 45 Gy in 5 weeks). Concomitant chemotherapy (5-fluorouracil and mitomycin C) was delivered to 31 patients. Two months later, a boost of 20 Gy was delivered by interstitial 192Ir brachytherapy to 101 patients and EBI in 5. Twelve other patients had an abdominoperineal resection (APR). The mean follow-up was 6 years.
Results: At 5 years the overall survival was 60%, and specific survival (SS) was 75%; it was 94% for T1, 79% for T2, 53% for T3, and 19% for T4. In multivariate analysis, tumor size (> or = 4 cm), node involvement, and no response to the EBI were factors of poor prognosis for SS. Thirty-two locoregional recurrences occurred of which 21 were local recurrences in the 106 patients treated by a conservative schedule. Only tumor size and response to the EBI were prognostic factors on multivariate analysis for local and LRC. A total of 17 patients presented with LSC (Grade 3, 16 patients; and Grade 4, 1 patient), which was treated by APR in 4 patients and colostomy in 11 (of which 7 were definitive). The only significant prognostic factor for LSC in the multivariate analysis was the total extrapolated response dose of irradiation. The definitive rate of SC after conservative treatment in cured patients was 100% for T1, 82% for T2, 58% for T3, and 100% for T4. Since 1989, improvements of the technique have allowed reduction of the LSC in maintaining the same local control.
Conclusion: The results of this series are similar to those of the literature. The confirmation of pretherapeutic prognostic factors related to response to the treatment should allow us to adapt the therapeutic intensity for each case to obtain better tumor control, with as few sequelae as possible, to yield a better rate of SC.