T4 rectal cancer treated with preoperative chemoradiation to the posteriorpelvis followed by multivisceral resection: Patterns of failure and limitations of treatment

Purpose: To analyze the overall pattern of treatment failure and sites of p elvic disease recurrence relative to the radiation fields used in treating patients with clinically staged T4 rectal cancer with preoperative chemorad iation followed by multivisceral resection. Methods and Materials: Between 1990 and 1998, 45 patients with T4 rectal ca ncer were treated with preoperative chemoradiation. Clinical staging was ac cording to the system of the American Joint Cancer Committee and was based on endoscopic ultrasonography, chemotherapy (CT), and physical examination. A diagnosis of T4 disease required evidence of invasion of a contiguous st ructure on CT (n = 31) or endorectal ultrasonography (n = 6), vaginal mucos al involvement on pelvic examination (n = 6), or a combination of these fin dings (n = 2). Chemoradiation was delivered with 18 MV photons using a 3-fi eld belly-board technique. The median total dose was 45 Gy in all patients (range 45-63). Nine patients received a boost with external beam radiothera py (EBRT) (n = 5, 1.8-18 Gy), intraoperative RT (n = 3, 10-20 Gy), or inter stitial brachytherapy (n = 1, 20 Gy). All patients received concurrent chem otherapy consisting of protracted venous infusion 5-fluorouracil (300 mg/m( 2), 5 d/wk). Resection was not performed in 13 (29%) of the 45 patients bec ause of metastases detected before resection or patient refusal. Multivisce ral resection and pelvic exenteration was required in 21 (66%) and 11 (34%) of 32 patients, respectively. We compared the location of pelvic disease r ecurrence with the RT simulation films. The Kaplan-Meier method was used to calculate the 4-year actuarial pelvic and distant recurrent rates and the overall survival rate. Results: The median length of follow-up was 31.0 months for all patients an d 40.0 months for patients alive at last-follow-up. When only the resected cases were considered, the local recurrence rate was 20%. Distant metastase s occurred in 44% of cases; the overall survival rate was 69%. When all pat ients were considered, the local recurrence rate was similar (24%), but the rate of distant recurrence (51%) was higher and the overall survival rate lower (50%). Pelvic disease was controlled in all 8 patients whose disease responded well to chemoradiation (either a histologically complete response or microscopic residual disease). Three of 4 patients with close or positi ve margins had pelvic recurrences despite intraoperative RT and brachythera py. Nine of the 10 pelvic recurrences occurred in the radiation field. Elec tive external iliac nodal irradiation was not used, and nodal metastases we re not seen in that region. In 1 case, marginal recurrence occurred in a co mmon iliac node at the superior edge of the treatment field. Conclusions: Despite aggressive multimodality therapy including multiviscer al resection, a high rate of pelvic and distant disease recurrence occurred in patients with clinically staged T4 disease. Regional disease recurred a lmost exclusively in the radiation field. The intraoperative RT and interst itial brachytherapy doses used did not prevent pelvic disease recurrence in patients with close or positive margins. Novel strategies such as higher p reoperative doses of RT with or without altered fractionation or more effec tive radiosensitizers are needed to improve locoregional control in patient s with T4 disease. Future strategies must also include more effective syste mic therapy. (C) 2001 Elsevier Science Inc.