Renal damage in the mouse: the response to very small doses per fraction.

Experiments were undertaken to study the effect on the mouse kidney of repeated X-ray doses in the range 0.2 to 1.6 Gy per fraction and neutron doses in the range 0.05 to 0.25 Gy per fraction. A top-up design of experiment was used, so that additional graded doses of d(4)-Be neutrons (EN = 2.3 MeV) were given to bring the subthreshold damage produced by these treatments into the measurable range. This approach avoided the necessity to use a large number of fractions to study low doses per fraction. Renal damage was assessed using three methods: 51Cr-EDTA clearance, urine output, and hematocrit at 16-50 weeks postirradiation. The dose-response curves obtained were resolved best at 29 weeks. However, the results were also examined by fitting second-order polynomials to the data for response versus time postirradiation and using interpolated values from these functions at 29 weeks to construct dose-response curves. This method reduced slightly the variation in the dose-response data, but the interrelationship between the dose-response curves remained the same. The data were used to test the linear-quadratic (LQ) description of the underlying X-ray dose-fractionation relationship. The model fits well down to X-ray doses per fraction of approximately 1 Gy, but lower X-ray doses were more effective per gray than predicted by LQ, as seen previously in skin [M. C. Joiner et al., Int. J. Radiat. Biol. 49, 565-580 (1986)]. This increased X-ray effectiveness and deviation from LQ are reflected directly in a decrease in the RBE of d(4)-Be neutrons relative to X-rays at low doses, since the underlying response to these neutrons is linear in this low-dose region. The RBE decreases from 9.9 to 4.7 as the X-ray dose per fraction is reduced below 0.8 Gy to 0.2 Gy, reflecting an increase in X-ray effectiveness by a factor of 2.1. A model is discussed which attempts to explain this behavior at low doses per fraction.