INTRODUCTION Radiotherapy as a single modality or combined with synchronous chemotherapy is capable of high rates of tumour control. A local control rate of 90% at 5 years would be expected both when radiotherapy is administered as the sole treatment for a T1-glottic carcinoma and when HPV-associated oropharyngeal cancer is treated with synchronous chemoradiation in patients without a signicant smoking history.1-6
However, these high rates of local control come at considerable cost to the patient. Acute side effects can include mucositis which often requires the use of strong opioid analgesia, taste disturbance, hair loss, acute xerostomia, dysphagia requiring the use of temporary tube feeding and skin reactions. Long-term side effects can include the risk of osteoradionecrosis, xerostomia and the risk of stricture formation and brosis necessitating long-term tube feeding. There is no strong evidence that prophylactic treatment of these side effects is effective other than the use of intensity modulated radiotherapy to avoid late xerostomia.7-10 On the contrary an adverse effect on local control was observed in one study when patients were randomized to receive prophylactic granulocyte-colony stimulating factor (G-CSF) during irradiation with no improvement in the rate of acute mucositis, a nding with possible implications for prophylactic G-CSF use during chemoradiation.11 Death during or within 8 weeks
of completion of treatment, when actually reported in randomized trials, occurs in 0-7% of patients undergoing radiotherapy alone and 0-14% with chemoradiation. Deaths denitely ascribable to treatment are reported in 0-3% and 0-5% of patients respectively.11-23
It is therefore essential that this therapeutic ratio i.e. the benet to the patient (altered chance of cure) vs the detriment to the patient (altered chance of mortality/ morbidity) is considered in detail when any modication to the radiotherapy or chemoradiotherapy schedule is made.24-28 In this chapter the physical, chemical and biological processes involved in radiotherapy delivery will be briey considered to enable an appreciation of the way in which these processes have already been or may be modied in the future to affect this ratio. The radiobiological modelling of such modications will then be considered.