Background and purpose: Theoretical calculations suggest that pulsed dose-rate irradiation (PDR) should have approximately the same effectiveness as continuous low dose-rate (CLDR) when the same total dose is given in the same overall time, unless large doses per pulse (> 2 Gy) are used and/or non-exponential or very short half-times of repair (< 0.5 h) are present in the irradiated tissues. However, few animal experiments have been reported to test this theory, and some of them gave contradictory results. We have carried out experiments to determine whether PDR irradiation of 18 mm of cervical spinal cord in the rat was more or less effective than CLDR at 0.5-1 Gy/h, when the overall average dose rate during each day of PDR was close to the overall CLDR average dose rate.
Materials and methods: PDR was simulated at a within-pulse dose rate of 4 Gy/h by filtered 18 MV X-rays from a linear accelerator. Two PDR schedules were used, 0.69 Gy at 1 h repetition (9 pulses per day) and 2 Gy at 3 h repetition (4 pulses per day), with overnight intervals of 16 and 15 h, respectively. The CLDR was delivered from iridium-192 wires in two concentric rings around a collar designed to fit the necks of rats so that they could eat and drink during the 72 h that was always the duration of the CLDR. Dose rate was then proportional to total CLDR dose. A range of doses was used to obtain dose response-curves, with a 15 Gy top-up dose (at 2 Gy/min, HDR) given on the day after the end of the PDR or CLDR irradiations. Animals were observed for at least 9 months to see whether fore-limb myelopathy developed. A total of 6-8 rats was irradiated per dose point, in two sets of experiments at an interval of 12 months.
Results: A set of 2 Gy fractions (at HDR) given daily, followed by the same top-up dose of 15 Gy at HDR, was available from a previous experiment for planning. Its ED50 was 61.2 Gy. The ED50 values found for the PDR schedules with 2 Gy at 3 h and 0.69 Gy at 1 h were 59.9 and 60.2 Gy, respectively. These were just 2% more effective than the daily HDR fractions, similar to expectations from theory if two components of repair are present. However, the CLDR irradiations resulted in no myelopathy even after doses up to 68 Gy at 0.94 Gy/h.. Thus PDR over 7 days (not at nights) appears to be more effective than CLDR over 3 days, with an effective dose-modifying factor of at least 1.1 to 1.17.
Discussion and conclusions: Reasons for this absence of effect with CLDR in these experiments are discussed, the most likely explanation being that a substantial component of repair with very short T1/2 (< 0.5 h) was present in spinal cord of these rats. There is evidence from other experiments elsewhere and in our laboratory for such a fast component of repair.