LEUKAEMOGENESIS: QUANTITATIVE ASPECTS AND CO-FACTORS 15 



LiNDOP, P. J., and Rotblat, J. (19G1). Proc. roy. Soc. 154B, 332. 



MacMaiion, B. (19G2). J. nat. Cancer Inst. 28, 1173. 



Mathe, G. (19G2). Bull. World Filth Org. 26, 585. 



Mery, Anne-Marie (1961). Contribution a I'etude des facteurs etiologiques des 



leucemies de Fenfant. These, Paris. 

 Mole, R. H. (1958). Brit. med. Bull. 14, 184. 

 Mole, R. H. (1959a). United Nations Peaceful uses of Atomic energy. Proceedings of the 



Second International Conference, Geneva, September 1958, 22, 145. 

 Mole, R. H. (1959b). Brit. J. Radiol. 32, 497. 

 Mole, R. H. (1962). In "Some Aspects of Internal Irradiation" (T. F. Dougherty, ed.). 



Pergamon Press, Oxford. 

 Murray, R. W., and Hempelmann, L. H. (1961). In "Radioactivity in Man" (G. R. 



Meneely, ed.). Charles C. Thomas, Springfield, Illinois. 

 Rous, P. (1959). Nature, Lond. 183, 1357. 

 SoRiEUL, S., and Ephrussi, B. (1961). Nature, Lond. 190, 653. 

 TOMONAGA, M. (1962). Bull. World Hlth Org. (in press). 

 Upton, A. C, Kimball, A. W., Furth, J., Christenberry, K. W., and Benedict, 



W. H. (1960). Cancer Res. 20, No. 8 (part 2) 1. 

 Wise, M. E. (1961). Hlth Phys. 4, 2.50. 



DISCUSSION 



trPTON: What would be the expectation if radiation could produce both kinds of events? 

 MOLE: It depends entirely on the relative probabilities. If you accejit the basic idea of 

 Armitage and Doll which they propounded theoretically just to account for the distribu- 

 tion of age-specific mortality rates in man for every different kind of cancer they could 

 lay their hands on, then cells which have undergone the first kind of change gradually 

 accumulate during life and further have some selective advantage so that they multiply 

 progressively. However, no ostensible tumour appears until a second change occurs and 

 then there is an inevitable progression towards what we recognize as a real cancer. If the 

 probability of producing the first kind of change with radiation is much higher than pro- 

 ducing the second, then all you are doing with radiation is just ageing the individual by 

 producing more of these cells which he would have got anyway if he had lived another 10 

 years or 5 years, whatever it might be. If, on the other hand, the probability is much 

 higher of producing the second type of change, then you are not increasing the risk of 

 getting leukaemia right through life, you are just increasing the risk temporarily. If ceUs 

 don't undergo the second kind of change, then nothing is going to happen; if they do 

 suffer the change presumably leukaemia shows itself fairly soon. It is interesting that in 

 Japan the actual incidence of leukaemia in the survivors at both Hiroshima and 

 Nagasaki is now in one city three times, in the other six times, what you would expect 

 from the natural life-span incidence of leukaemia in unirradiated population, so that 

 radiation has produced a real excess of leukaemia, it hasn't just altered the distribution 

 of leukaemia in time or with age. That seems to suggest that radiation, if you accept this 

 very mathematical idea, must produce the second kind of change more easily than the 

 first, but that doesn't mean to say it doesn't also produce the first change. If you assume 

 that this first change leads to clones of cells which proliferate and that when radiation 

 does something to cells in such clones, from that moment on the thing inevitably becomes 

 an ostensible leukaemia or cancer, then you can only assume that there is going to be a 

 persistent increased risk for the rest of life if these doubly altered cells can persist for the 

 rest of life without undergoing progressive multiplication. Now I think the basic idea of 

 both Fisher, and Armitage and Doll is that this doesn't happen, but that when the second 



