Radiation fractionation: the search for isoeffect relationships and mechanisms

Review the historical basis for the use of fractionated radiation in radiation oncology.

The history of dose fractionation in radiation oncology is long and tortuous, and the radiobiologist's understanding of why fractionation worked came decades after radiation oncologists had adopted multi-week daily-dose fractionation as 'standard'. Central to the history is the search for 'isoeffective' formulas that would allow different radiation schedules to be compared. Initially, this meant dealing with different lengths of treatment, leading to the 1944 Strandqvist formulation that dominated thinking for decades. Concerns about the number of fractions, not just the total time, led to the 1967 Ellis NSD formulation that held sway through the 1980s. The development of experimental radiotherapy in 1970s (e.g. Fowler's work at the Gray Laboratory, and Fischer's work at Yale) led to biologically-based approaches that culminated with the Biologically Effective Dose (BED) concept. BED is the current dogma for treatment optimization, but it must be used with caution, as there are multiple formulations, and some parameters have debatable values. There is also a controversy about whether BED is biologically-based or a 'curve-fitting' exercise. These latter issues are beyond the scope of this article, but the history of fractionation models suggests that our current concepts are probably wrong, although when used with caution they are clearly useful.