RADIATION-THERAPY IN EARLY GLOTTIC CARCINOMA - UNIVARIATE-ANALYSIS AND MULTIVARIATE-ANALYSIS OF PROGNOSTIC FACTORS AFFECTING LOCAL-CONTROL

Purpose: The purpose of this report is to clarify prognostic factors affecting local control of T1 and T2 glottic tumors and to define an optimal regimen for radiation therapy. Methods and Materials: Two hundred and ten patients (199 males, 11 females, age range 30 to 86 years with an average of 62 years) with previously untreated invasive squamous cell carcinoma of the glottis were treated with radiation therapy at the University of Tokyo between January 1972 and December 1989. Endoscopic microsurgery was introduced as an integral part of treatment in 1974. From 1974 to 1979 the radiation dose was gradually reduced, reaching a mean of 20 Gy in 2 weeks in 1979. From 1980 to 1983, the total dose increased to 50.4 Gy, with a fraction size of 1.8 Gy, over a mean of 5.6 weeks. From 1984 onward, the mean total radiation dose increased to 60 Gy with a fraction of 2 Gy. Results: Recurrence-free 5 year survival rates for T1a, T1b, and T2 were 79%, 73%, and 67%, respectively. When the relationship between radiation dose and local control rates was analyzed for each year from 1974 to 1989, total doses were strongly associated with local control for patients with Tla disease. Age, sex, daily dose, total dose, radiation machine (Co-60 or 10 MV Lineac), treatment technique (anterior wedged pair or parallel opposed fields), treatment volume, use of endoscopic microsurgery, and involvement of the anterior commissure were examined for effects upon relapse-free survival in T1a disease by uni- and multivariate analysis. Total dose was the only significant factor for T1a disease (p < 0.02). The effect of these variables upon relapse-free survival in T2 disease as well as the effect of cord mobility, and number of involved sites was examined by multivariate analysis. Total dose (p < 0.03), cord mobility (p < 0.05), and number of involved sites (p < 0.04) significantly affected relapse-free survival in T2 disease. Conclusion: At least 50 Gy is required for treatment of T1 disease when 2 Gy is used as a daily dose, even if endoscopic microsurgery is performed. Better local control of T2 disease in patients with impaired cord mobility or more than three involved sites leads to an improved prognosis; we recommend doses of at least 70 Gy or use of hyperfractionation in such patients with these factors. Although the daily dose did not significantly affect prognosis in multivariate analyses, 1.8 Gy is not recommended for treatment of T2 tumors instead of 2 Gy.