Background: A retrospective study of 1,493 head and neck cancer patients was designed to test current radiobiological thinking, postulating the detrimental effect of protracted overall treatment times (OTT) and/or split course (SC) regimes in radiation therapy on local tumor control.
Methods: Primary squamous cell carcinomas of the oral cavity (OC), oropharynx (OP), hypopharynx (HP), nasopharynx (NP), and larynx radiated with a dose of at least 50 Gy were analyzed. Those patients treated by brachytherapy and/or primary surgery were excluded. A detailed analysis of the 997 cancers of the larynx was recently published. This paper focuses on the relationship between local tumor control and treatment characteristics for the 496 tumors originating from the OC, OP, HP, and NP. Total doses of radiation ranged from 50 to 79 Gy, with a mean of 64 Gy.
Results: A local failure (LF) was observed for 278 patients. Using Cox regression analysis, T stage and site were strongly related to LF. Corrected for T stage and with reference to OP, tumors in the NP, HP, and OC had a relative LF rate of 0.5, 1.6, and 1.8, respectively. Patients treated with continuous course (CC) and higher doses of radiation therapy fared best. No association was found with OTT and the use of chemotherapy.
Conclusions: The results observed for the OC, OP, HP, and NP are in line with the findings for the larynx. Analyzing all 1,493 patients, for SC regimes lower local control rates were observed as opposed to the CC treatment series. Moreover, for the normalized total doses, a dose-effect relationship could be established. This study corroborates that disruption of the treatment per se and/or the use of suboptimal total doses of RT are detrimental; it is argued that these observations could be of relevance when designing combined modality protocols.