Accelerated fractionation in the radiation treatment of head and neck cancer. A critical comparison of different strategies.

There is strong clinical and radiobiological evidence that protraction of overall treatment time has an adverse influence on the radiocurability of certain human tumors. Overall treatment time can be reduced without recourse to large dose fractions by the use of accelerated fractionation, but in patients with head and neck cancer acute mucosal reactions limit the extent to which treatment can be accelerated. Three different prototypical schedules for accelerated fractionation have been devised to avoid exceeding acute mucosal tolerance. Type A consists of an intensive short course in which the overall duration of treatment is markedly decreased with a corresponding substantial reduction of total dose; type B achieves a modest decrease in overall time without reduction of total dose by using a split-course technique; type C also achieves a modest decrease in overall time without reduction of total dose by means of the concomitant boost technique. A hybrid schedule combining features of types B and C allows additional shortening of overall treatment time without reduction of total dose. Available radiobiological and clinical data suggest that schedules of types B or C which do not compromise total dose are generally preferable to those of type A in which there is a trade-off between total dose and overall time. For a given total dose and overall time, a continuous treatment of type C is likely to produce more cell kill than a split-course of type B, although the latter will be better tolerated. Because of the increased acute toxicity associated with all schedules of accelerated fractionation, rational selection of patients for such treatment is important. New techniques to measure the potential doubling time of human tumors in vivo offer this prospect.