350 SPECULATIONS ON CELLULAR ACTIONS 



Another important physical factor is the time pattern of irradiation. 

 I use this general term to include variation in dose rate and also various 

 more or less complicated schemes of fractionated irradiation. Although, 

 among the hundreds of papers on the subject, complete dose-effect 

 curves are all too rarely found, the following general statements may be 

 made (2). (a) A large but not overwhelming fraction of the known 

 radiobiological actions depends more or less on the time pattern of ir- 

 radiation. (6) As the time consumed in delivering the dose is lengthened 

 — by decreasing the dose rate, by fractionation, or by combinations of 

 the two — a small fraction of the time-pattern-dependent actions is en- 

 hanced; that is, the effectiveness of a given total dose increases, (c) On 

 the other hand, for the overwhelming majority of time-pattern-depend- 

 ent actions, the reverse is true; that is, the effectiveness of a given total 

 dose decreases. 



In terms of our model (Fig. 2), it appears that the time pattern might 

 exert its influence in two general ways. First, the biological system 

 might change during irradiation — either because of natural cellular 

 activities or because of some alteration due to irradiation (possibly pro- 

 duced by some of the processes designated as "irrelevant" in Fig. 2) — 

 and such a change in the biological system might develop to greater or 

 less degree as the irradiation is prolonged. In this event, as pointed out 

 in the discussion of influence of other biological factors, such as mitosis, 

 it is possible for any of the parameters m, n, r, s, or h either to increase 

 or to decrease, and accordingly either an increase or decrease of 

 effectiveness might result from prolongation of the total irradiation 

 period. 



The second general way in which the time pattern might exert its in- 

 fluence is by changing h through differential alteration of the probabilities 

 of the relevant and of the various competing processes. As in the dis- 

 cussion of RET, let us consider this possible mode of influence with re- 

 spect to the time at which the decisive processes occur. If the decisive 

 processes occur early, that is, when the participating entities are local- 

 ized in or close to the ionization track, then each wave of relevant proc- 

 esses proceeding from a single track acts independently, and the time 

 pattern can exert no influence on any of the parameters of the dose- 

 effect relation, unless the dose rate is varied to such extremely high 

 values that there is appreciable probability that ionization tracks may 

 spatially coincide during their very brief lifetimes. On the other hand, 

 if the decisive processes are late, the rates of the relevant and of the com- 

 peting processes depend on the concentrations of the participating en- 

 tities when they are distributed singly at random in the cell, and these 

 concentrations obviously can be affected by the time pattern. 



