Response of rat spinal cord to very small doses per fraction: lack of enhanced radiosensitivity

Our previous work with rat spinal cord demonstrated that the linear quadratic (LQ) model based on data for large fraction sizes ((α(β)) of 2.4 Gy) failed to predict isoeffective doses between 1 and 2 Gy per fraction, and under-estimated the sparing effect of small doses per fraction given once daily. In contrast, data from mouse skin and kidney, and recent in vitro results revealed a paradoxical increase in radiosensitivity at below 1 Gy per fraction. To assess whether enhanced radiosensitivity is present in the spinal cord below 1 Gy per fraction, the rat spinal cord (C2-T2) was irradiated initially with three daily doses of 10.25 Gy (top-up doses representing 90% of tolerance), followed by graded single doses or fractionated doses in 1.5, 1.0, 0.8, 0.6 or 0.4 Gy fractions given once daily. To limit the overall treatment time to ≤ 8 weeks, a small number of the 0.6- and 0.4-Gy fractions were given twice daily with an interfraction interval of 16 h. The end-point was forelimb paralysis secondary to white matter necrosis, confirmed histologically. The ED50 values, excluding the top-up doses, were 5.8, 10.6, 14.8, 15.2, 15.9 and 19.1 Gy for a single dose and doses in 1.5-, 1.0-, 0.8-, 0.6- and 0.4-Gy fractions, respectively. The data gave an (α(β)) of 2.1 Gy (95% CI, 1.4, 2.7 Gy). Pooling the data separately, the (α(β)) value was 2.3 Gy (95% CI, 0.82, 3.7 Gy) for fraction sizes ≥ 1 Gy, and 1.2 Gy (95% CI, 0.16, 2.3 Gy) for the 0.8-, 0.6- and 0.4-Gy experiments. These results in which top-up doses were given initially are consistent with a large sparing effect of very small fraction sizes in rat spinal cord provided sufficient time is allowed for repair of sublethal damage between fractions, and provide no evidence for a paradoxical increase in radiosensitivity in the rat spinal cord below 1 Gy down to 0.4 Gy per fraction