Repair between dose fractions: a simpler method of analyzing and reporting apparently biexponential repair

Increasing numbers of animal experiments in situ are reporting that repair of sublethal radiation damage in vivo slows down with time, usually described as two components of (monoexponential) repair. For repair of DNA strand breaks, plotting the reciprocal of proportion unrepaired as a function of time yielded straight lines. Two processes have been suggested as causing this: (1) a second-order process (bimolecular) instead of first-order (exponential) and (2) a skewed distribution of monoexponential rates. The present paper investigates whether such plots of hyperbolic or reciprocal repair are relevant for laboratory animal tissue results. Published repair data were reanalyzed from laboratory animal experiments that employed split doses or two fractions per day. Graphs are presented of the reciprocal proportion of damage remaining as a function of the interval between the two doses. If the reciprocal model applies, the graphs would be straight lines. Different animal data showed no inconsistency with straight reciprocal plots. These reciprocal plots describe well with one parameter tau, the first half-time, repair curves previously thought to be "biexponential", and to require three parameters. Straight reciprocal plots mean that in a constant time interval tau the unrepaired damage falls from 1 to (1/2), then from (1/2) to (1/3), then (1/3) to (1/4), etc. A much larger proportion of damage would therefore remain unrepaired at several half-times than is estimated by current mono- or biexponential models. The practical implications for clinical radiotherapy are important.