Does tumor control decrease by prolonging overall treatment time or interrupting treatment in laryngeal cancer?

Abstract

To evaluate treatment results of radiotherapy in laryngeal cancer and to retrospectively investigate the influence of treatment interruptions on local control rates. A multivariate analysis was performed to investigate the role of overall treatment time on local control rates. From 1962 to end 1986, 864 patients with cancer of the larynx were treated with primary radiotherapy, with a variation of doses and treatment times. In 352 patients a systematic interruption of 2 weeks was done after 50 Gy (split-course). Local control rates were calculated by the actuarial method. A multivariate analysis was subsequently performed in order to identify independent variables influencing local control rates. The univariate analysis showed that in patients treated with a split-course schedule, local control rates were significantly lower in glottic cancer (p < 0.001), with differences of 7% in T1 (p = 0.05) and 11% in T2 (p = 0.11). In node-negative supraglottic cancer a difference of 10% (p = 0.16) was observed. The multivariate analysis showed the following factors to significantly influence local control: site, total dose, total treatment time, T and N stage. Split-course treatment was not an independently significant factor, probably caused by time being a stronger factor. In subgroup analysis, the effect of total treatment time remained statistically significant in T2 glottic cancer. A significant negative influence of treatment interruption was seen in glottic cancer, and a trend toward significance in node-negative supraglottic cancer in univariate analysis. A multivariate analysis showed a negative influence of prolonging treatment time in the total patient group. The effect remained significant in T2 glottic cancer in subsequent subgroup analysis, possibly due to the fact that there is little variation in tumour volume in this group. In larger tumours, where there is substantial variation in tumour volume and a higher probability of heterogeneity in radiosensitivity, the effect of dose-time variations could be more difficult to discern.