Time-dose-response relationships in postoperatively irradiated patients with head and neck squamous cell carcinomas.

BACKGROUND AND PURPOSE: To define the influence of the dose and time on the response to treatment in postoperatively irradiated head and neck cancer patients and to establish a good prediction of failure. METHODS AND MATERIALS: From January 1985 to December 1995, 214 patients with histologically proven head and neck squamous cell carcinomas were irradiated after radical surgery or single tumour resection according to surgical and histopathological findings. The total doses given ranged between 50 and 75 Gy to the primary bed tumour and between 42 and 56 Gy to the neck with fraction sizes of 1.7-2 Gy/day. The median length of the time interval between surgery and radiotherapy, time of irradiation and total treatment time were 81, 59 and 139 days, respectively. The end-point analyzed was the local-regional tumour control rate at the primary tumour bed and neck for 5 years from the beginning of radiotherapy. Univariate and multivariate analyses were used to determine predictors of failure from among the following studied variables: (i), clinical stage (T/N) of the patients; (ii), tumour grade; (iii), neck surgery; (iv), tumour margins; (v), histological tumour nodal extension; (vi), chemotherapy; (vii), normalized total dose; (viii), time interval between surgery and radiotherapy; (ix), time of irradiation; and (x), total treatment time. RESULTS: The actuarial 5-year tumour control rate for the entire group was 72%, and 92% of the patients who achieved local control are currently alive without disease. Tumour control was inversely related to T stage (83% for T2 vs. 57% for T4) and the probability of local control within each stage was dependent on the N status (> or =71% for T3-T4/N0 vs. 31-44% for T3-T4/N1-N3). Histological N status and tumour margins, but not tumour grade, impacted significantly on tumour control. When local control was analyzed as a function of the dose to the primary, a non-significant negative dose-response relationship was found. The total treatment time was a significant prognostic factor, and the time interval between surgery and irradiation proved to be an independent predictor of failure. CONCLUSIONS: Despite the absence of a statistically significant dose-response relationship, the present results suggest that postoperative irradiation treatment given to patients with head and neck squamous cell carcinomas should not be unduly prolonged, in order to minimize the amount of tumour cell proliferation. In these patients, nodal involvement, positive margins of the resected specimens and time interval between surgery and irradiation were the most important prognostic factors.