| Abstract: | In July 1982 the Gastrointestinal Section of The Clinical Oncological Society of Australia began a multicentre randomized trial to assess the value of post-operative pelvic radiotherapy in the local control of stage B and C carcinoma of the rectum. Patients who had undergone a potentially curative resection were randomized either to no further treatment (NO RT) or to pelvic radiotherapy (RT), 45 Gy in 25 fractions (plus a perineal boost of 5 Gy in 2 fractions after abdomino-perineal resection). The trial was prematurely terminated in December 1985 owing to slow accrual after 70 patients had been randomized: 36 patients to RT and 34 patients to NO RT. Two patients in each group were found to be ineligible after randomization and, for a variety of reasons, thirteen patients who were randomized to RT did not receive that treatment. The incidences of post-operative complications were comparable in both groups, indicating that radiotherapy produced no additional problems. Radiotherapy-related morbidity comprised mainly diarrhoea (grade 1 = 2/21; grade 2 = 5/21; grade 3 = 0/21; grade 4 = 1/21). There were no deaths due to radiotherapy and only one patient, with grade 4 diarrhoea, failed to complete the prescribed course of radiotherapy. After an average follow-up period of 52 months there were 15 local recurrences, 8 in the RT group and 7 in the NO RT group, giving an actuarial 2 year local recurrence rate of 16%. There was no difference in the time to local recurrence, either by randomization group (p=0.50) or by actual treatment received (p=0.91). Distant metastases occurred in 25 patients, 12 in the RT group and 13 in the NO RT group. Thirty one patients have died, 26 from cancer, 3 from intercurrent illness without evidence of cancer and two from surgical complications. There was no difference in survival time between the randomized groups: 16 deaths in the RT group and 15 deaths in the NO RT group (p=0.46). The only important factor predictive of local recurrence was age 65 years and over (p=0.002). |
