The impact of treatment time and smoking on local control and complications in T1 glottic cancer.

PURPOSE: To define the optimal treatment regimen, patients with T1N0M0 glottic larynx carcinoma were treated with six different radiotherapy (RT) schedules. To assess the influence of patient characteristics, complication rates, and to evaluate the overall larynx preservation. METHODS AND MATERIALS: Out of a consecutive series of 383 patients treated for T1N0M0 glottic larynx carcinoma between 1965 and 1992, 352 evaluable patients were treated with six different "standard" fractionation schedules: 65 Gy (20 x 3.25 Gy), 62 Gy (20 x 3.1 Gy), 61.6 Gy (22 x 2.8 Gy), 60 Gy (25 x 2.4 Gy), 66 Gy (33 x 2 Gy) and 60 Gy (30 x 2 Gy). The median follow-up of all patients was 89 months. Patient factors analyzed included: age, sex, concurrent illness, smoking habits, tumor localization and extension, tumor differentiation, the effect of tumor biopsy or stripping of the vocal cord, and the presence of visible tumor at the start of radiotherapy. Treatment parameters evaluated were: year of treatment, beam energy, treatment planning, field size, fractionation schedule, fraction size, number of fractions, total dose, treatment time and treatment gap, the use of wedges, and neck diameter.
RESULTS: The overall 5-year actuarial locoregional control was 89%, varying between 83 and 93% for the different schedules. Univariately, local control decreased with increasing treatment time. This could not be explained by the confounding variables sex, tumor extension, and field length (p = 0.0065). Adjusted for these variables, 5-year local control percentage decreased from 95% (SE 2%) for 22-29 days to 79% (SE 6%) for treatment time > or = 40 days. The overall complication rate (grade I-IV) at 5 years was 15.3%, and varied between the different schedules, from 7 to 17%. No relation was found between complications and treatment factors. Patients who continued smoking had a higher complication rate than those who never smoked or stopped smoking, univariately as well as adjusted for tumor extension, macroscopic tumor, and neck diameter (p = 0.0038). Twenty-eight percent (SE 6%) of the patients who continued smoking had complications at 10 years, compared to about 13% (SE 4%) of those who stopped before or after RT. No evidence was found for any other relation between complications and patient or tumor factors. Severe edema and necrosis (grade III and IV) were not observed in the 2 Gy fraction schedules. A laryngectomy was performed in 36 patients: 30 for recurrence, 3 for complications (at 40, 161, and 272 months), and 3 for a second primary. The overall larynx preservation was 90% at 10 years, and for the different schedules it was 20 x 3.25 Gy: 97%; 20 x 3.1 Gy: 96%; 22 x 2.8 Gy: 92%; 25 x 2.4 Gy: 89%; 33 x 2 Gy: 78%; and 30 x 2 Gy: 80%.
CONCLUSION: Overall treatment time is the most significant factor for locoregional control of T1 glottic cancer. A schedule of 25 x 2.4 Gy appeared to be the optimal treatment schedule considering both tumor control and long term toxicity. The complication rate was increased in patients who continued smoking.