The effect of interruptions and prolonged treatment time in radiotherapy for nasopharyngeal carcinoma

Purpose: The effect of interruptions and prolonged overall treatment time in radiotherapy for nasopharyngeal carcinoma and the significance of timing of interruption was investigated. Methods and Materials: Treatment records of 229 patients treated with continuous course (CC) and 567 patients treated with split course (SC) radiotherapy for nonmetastatic NPC were reviewed. Overall treatment time without inclusion of time for boost was calculated. Treatment that extended 1 week beyond scheduled time was considered prolonged. Outcome in patients who completed treatment ''per schedule'' were compared with those who had ''prolonged'' treatment. Because of known patient selection bias between CC and SC, patients on the two schedules were analyzed separately. Multivariate analysis was performed for patients on SC. Total number of days of interruption, age, sex, T and N stage, and the use of boost were tested for the whole SC group. Analysis on the effect of timing of interruption was performed in a subgroup of 223 patients on SC who had a single unplanned interruption. Timing of interruption, either before or after the fourth week for the unplanned interruption, was tested in addition to the other variables in multivariate analysis for this subgroup of SC. Results: Twenty-seven (11.8%) patients on CC and 96 (16.9%) patients on SC had prolonged treatment. Patients on SC who had prolonged treatment had significantly poorer loco-regional control rate and disease free survival when compared with those who completed radiotherapy per schedule (p = 0.0063 and 0.001, respectively, with adjustment for stage). For CC, the effect of prolonged treatment on outcome was not significant. The small number of events for patients on CC probably account for the insignificant finding. The number of days of interruption was confirmed as prognostic factor, independent of T and N stages, for loco-regional control and disease-free survival in multivariate analysis for SC. The hazard rate for loco-regional failure increased by 3.3% for each day of interruption. The timing of interruption, at the beginning or towards end of treatment, did not significantly alter outcome. Conclusion: Interruptions and prolonged treatment adversely affect outcome in radiotherapy for NPC and the effect of repopulation was confirmed. Every effort should be made to keep treatment on schedule and interruptions for whatever reasons should be minimized. (C) 1997 Elsevier Science Inc.