Abstract
The total dose in radiotherapy has been adjusted in the past for different fractionation schedules by the use of empirical formulae such as NSD, TDF and CRE. It is now appropriate to consider fractionation factors which include more biological insight in their formulation than was possible earlier. It has become clear, from both clinical and experimental animal data, that the total dose in multi-fraction irradiations depends more critically on size of dose-per-fraction for late than for early damage to normal tissues. This difference has been interpreted as due to different shapes of the underlying dose-response curves. The late reactions respond with more curvature in the dose-response curve, i.e. with more repair capability at very low doses per fraction, than the early tissue reactions. A linear-quadratic relationship for the dose-response curves has been found to fit experimental data well, with few exceptions. This paper reviews this interpretation and explores some of its implications for radiotherapy and for radiobiology applied to therapy. Of many repair factors that have been suggested, the ratio alpha/beta (of the linear to the quadratic coefficients) is one that should be independent of the level of damage assayed. Values of alpha/beta of about 10 Gy have been reported for a number of early tissue responses but a range of values from about 1 to 5 or 6 Gy for late responses. It is a current challenge to radiobiology to explain why this difference occurs. Once such values are known for different tissues--and the dangers of premature assumptions are emphasized--calculations are possible which might be useful in radiotherapy as an alternative to NSD, TDF, CRE etc. Some data are presented on the magnitude of differences from these previously used empirical formulae, with a discussion about how easily detected the discrepancies might be in clinical practice. Applications to hypofractionation, hyperfractionation and accelerated fractionation are illustrated.