The combination of irradiation and total mesorectal excision for rectal carcinoma has significantly lowered the incidence of local recurrence. However, a new problem is represented by the patient with locally recurrent cancer who has received previous irradiation to the pelvis. In these patients, local recurrence is very often not easily resectable and reirradiation is expected to be associated with a high risk of late toxicity. The aim of this multicenter phase II study is to evaluate the response rate, resectability rate, local control, and treatment-related toxicity of preoperative hyperfractionated chemoradiation for locally recurrent rectal cancer in patients previously irradiated to the pelvis.

Patients with histologically proven pelvic recurrence of rectal carcinoma, with the absence of extrapelvic disease or bony involvement and previous pelvic irradiation with doses < or =55 Gy; age > or =18 years; performance status (PS) (Karnofsky) > or =60, and who gave institutional review board-approved written informed consent were treated by preoperative chemoradiation. Radiotherapy was delivered to a planning target volume (PTV2) including the gross tumor volume (GTV) plus a 4-cm margin, with a dose of 30 Gy (1.2 Gy twice daily with a minimum 6-h interval). A boost was delivered, with the same fractionation schedule, to a PTV1 including the GTV plus a 2-cm margin (10.8 Gy). During the radiation treatment, concurrent chemotherapy was delivered (5-fluorouracil, protracted intravenous infusion, 225 mg/m(2)/day, 7 days per week). Four to 6 weeks after the end of chemoradiation, patients were evaluated for tumor resectability, and, when feasible, surgical resection of recurrence was performed between 6-8 weeks from the end of chemoradiation. Adjuvant chemotherapy was prescribed to all patients, using Raltitrexed, 3 mg/square meter (sm), every 3 weeks, for a total of 5 cycles. Patients were staged using the computed tomography (CT)-based F-classification (F0: no side-wall involvement; F1, F2, F3: 1, 2, and 3-4 side-walls involved, respectively). Toxicity was evaluated on the basis of the Radiation Therapy Oncology Group (RTOG) criteria.

Fifty-nine patients (38 male, 21 female; median age, 62 years; range, 43-77 years) were enrolled in the study, by 12 different Italian radiotherapy departments. Previous surgery was anterior resection in 45 patients (76.3%) and abdominal-perineal resection in 14 patients (23.7%); previous radiotherapy dosage ranged between 30 and 55 Gy (median, 50.4 Gy); the median interval between prior radiation therapy to the onset of reirradiation was 27 months (range, 9-106 months); 44 patients (74.6%) had received some form of previous chemotherapy (concurrent and/or adjuvant). Fifty-one of 59 patients (86.4%) completed chemoradiation without treatment interruptions: 6 patients (10.2%) had temporary treatment interruption due to toxicity or patient compliance, and 2 patients (3.4%) had definitive treatment interruption. The incidence of Grade 3 lower gastrointestinal acute toxicity was only 5.1%. No patient developed Grade 4 acute toxicity. After chemoradiation, 5 patients (8.5%) had complete response (CR), 21 patients (35.6%) had partial response (PR), 31 patients (52.6%) had no change (NC) and 2 patients (3.4%) showed progressive disease (PD). Overall, the response rate (PR + CR) was 44.1% (95% confidence interval, 29.0-58.9%). Twenty of 24 patients (83.3%) with pelvic pain before treatment had symptomatic response. Tumor resection was performed in 30 of 59 patients (50.8%) including 2 local excisions, 4 anterior resections, 18 abdominoperineal resections, and 6 other. Surgical resection resulted as R0 and R1 in 21 patients (35.6%) and 3 patients (5.1%), respectively. The possibility of radical resection was influenced by tumor response to chemoradiation (PD/NC: 7/33; PR/CR: 14/26; p = 0.009). Thirty-three patients received adjuvant chemotherapy, which was completed in 30 (50.8%). At a median follow-up of 36 months (range, 9-69 months), 28 patients (47.5%) developed local recurrence or tumor progression in the unresected pelvic disease and 18 patients (30.5%) developed distant metastasis. Seven patients showed late toxicity, including 2 skin fibrosis, 2 impotence, 2 urinary complications requiring nephrostomy, and 1 small bowel fistula requiring surgical diversion. Overall median survival was 42 months. Five-year actuarial survival was 39.3%; 66.8% in R0 resected patients and 22.3% in patients treated without surgery or undergoing subtotal tumor removal. Local control and disease-free survival were significantly correlated with the interval between surgical treatment for primary tumor and local recurrence (p = 0.028 and p = 0.003, respectively). Radical resection significantly influenced local control, disease-free survival, and overall survival (p = 0.010, p = 0.010, and p = 0.050 respectively). The multivariate analysis confirmed the impact of surgery-relapse interval on local control (p = 0.016) and disease-free survival (p = 0.002), and confirmed the correlation between R0 surgery with local control and disease-free survival (p = 0.016).

Use of hyperfractionated chemoradiation was associated with a low rate of acute toxicity and an acceptable incidence of late complications. Pain control was excellent. The overall 5-year survival was 39%. Despite 87.4% of patients having F1-3 stage disease, approximately one-third (35%) achieved R0 resection, and two-thirds of patients in this cohort of patients were alive at the 5-year mark. However, further studies using innovative treatment algorithms are warranted to, hopefully, improve the local tumor response and control.