The effect of fraction time in intensity modulated radiotherapy:: theoretical and experimental evaluation of an optimisation problem

Background and purpose: In intensity modulated radiotherapy (IMRT), the complexity and the number of treatment fields have expanded. This may imply that the delivery time for each fraction becomes prolonged. In a number of IMRT techniques used in the clinic, the delivery time per fraction is usually 10-15 min, sometimes more than 15 min. In studies on human skin, prolonged delivery time is shown to cause significant reduction of radiation effects compared with acute irradiation. In this paper the effect of changes in fraction delivery time was studied by in vitro irradiation of mammalian cells. Material and methods: Chinese hamster fibroblasts (V79-379-A) were used for simulating clinical situations. Most experiments Were performed with 2 Gy/fraction with 4-h intervals in 40-60 replicates. Each fraction was divided into different subfractions, simulating the delivery of a complicated treatment. The effect of changing the delivery time for each fraction was studied. Parameters for the cell survival curve and repair kinetics were determined experimentally. The same methods were also used for large fraction sizes (8 Gy). The validity of the most widely used models in the literature, all derived from linear-quadratic formalism, were tested against the experimental results. Results: The effect of prolonging the fraction time for 2-Gy fractions was underestimated by the biological models. The experiments showed that 10-min prolonged delivery time gave a ratio between surviving fractions at 2 Gy (S-ratio) of 1.054 with a 95% confidence interval (CI) 1.030-1.080, while the models predicted 1.007 and 1.009. Extending the fraction time to 20 min gave an S-ratio of 1.063 with CI of 1.045-1.080, while the models predicted 1.012 and 1.014. For 8-Gy fractions, there was a good agreement between predications and experimental results. The ratio between surviving fractions at 8 Gy is 1.370 with CI of 1.300-1.440, while the models predicated 1.37 and 1.35. Conclusions: The effect of prolonging fraction time at conventional dose/fraction is underestimated by biological models. Prolonging the fraction time will spare tissues with a fast DNA repair. There is a risk for sparing tumours. This Should be considered when IMRT technique is implemented in the clinic. (C) 2003 Elsevier Ireland Ltd. All rights reserved.