Hypoxic modification of radiotherapy in squamous cell carcinoma of the head and neck – A systematic review and meta-analysis

Background: The importance of tumour hypoxia for the outcome of radiotherapy has been under investi gation for decades. Numerous clinical trials modifying the hypoxic radioresistance in squamous cell car cinoma of the head and neck (HNSCC) have been conducted, but most have been inconclusive, partly dueto a small number of patients in the individual trial. The present meta-analysis was, therefore, performedutilising the results from all clinical trials addressing the specific question of hypoxic modification inHNSCC undergoing curative intended primary radiotherapy alone. Methods: A systematic review of pub lished and unpublished data identified 4805 patients with HNSCC treated in 32 randomized clinical trials,applying, normobaric oxygen or carbogen breathing (5 trials); hyperbaric oxygen (HBO) (9 trials); hyp oxic radiosensitizers (17 trials) and HBO and radiosensitizer (1 trial). The trials were analysed with regardto the following endpoints: loco-regional control (32 trials), disease specific survival (30 trials), overallsurvival (29 trials), distant metastases (12 trials) and complications to radiotherapy (23 trials). Results:Overall hypoxic modification of radiotherapy in head and neck cancer did result in a significant improvedtherapeutic benefit. This was most dominantly observed when using the direct endpoint of loco-regionalcontrol with an odds ratio (OR) of 0.71, 95% cf.l. 0.63–0.80; p < 0.001), but this was almost mirrored in thedisease specific survival (OR: 0.73, 95% cf.l. 0.64–0.82; p < 0.001), and to a lesser extent in the overall sur vival (OR: 0.87, 95% cf.l. 0.77–0.98; p = 0.03). The risk of distant metastases was not significantly influ enced although it appears to be less in the tumours treated with hypoxic modification (OR: 0.87, 95%cf.l. 0.69–1.09; p = 0.22), whereas the radiation related late complications were not influenced by theoverall use of hypoxic modifications (OR: 1.00, 95% cf.l. 0.82–1.23; p = 0.96). The improvement in loco regional control was found to be independent of the type of hypoxic modification. The trials have used different fractionation schedules, including large doses per fraction, which may result in relatively morehypoxia and greater benefit. However, analysis of HNSCC trials using conventional fractionation only, showed that the significant effect of hypoxic modification was maintained. Conclusion: The meta-analysis thus demonstrates that there is level 1a evidence in favour of adding hypoxic modification to radiother apy of squamous cell carcinomas of the head and neck. 2011 Elsevier Ireland Ltd. All rights reserved