Abstract
During the period 1970–1989, 256 patients with squamous cell carcinoma of the head and neck were treated with radical surgery and postoperative split-course radiotherapy. The 3-week rest period was compensated by a 10% increase in the total radiation dose to 66 Gy. The 5-year actuarial local control rate was 66% in tonsillar, 60% in mobile tongue, 64% in floor of the mouth, 51% in lower gingiva, 63% in laryngeal, and 35% in hypopharyngeal cancer. Failures were observed in 102 patients (40%) and 51 (20%) died of intercurrent diseases. Except in mobile tongue cancer, the results in the logistic regression analyses showed that T- and N-category and pretreatment Karnofsky index had the strongest association with local control, whereas in Cox's proportional hazard's regression analyses T-category and pretreatment Karnofsky had the strongest association with survival. In mobile tongue cancer, the histologic grade and the time interval between surgery and radiotherapy had the strongest association with local control and survival respectively. However, the lengthening of the time interval was often caused by factors or events which could have influenced the prognosis. On the basis of both univariate and multivariate analyses, when the tumour resection margins were free, a time interval of less than 3 months, the overall treatment time and the duration of the split of radiotherapy, became less important for local control and survival when doses of 65 Gy were given by fraction of 2 Gy or greater.