MUTAGENIC AND CARCINOGENIC AGENTS 



(i) Patients heavily irradiated for treatment of ankylosing 

 spondylitis showed a significant incidence of myeloid leu- 

 kaemia. Analysis of the data by Court-Brown and Doll 

 (1957) indicated that there was an approximately linear 

 relation between dose and the likelihood of leukaemia and 

 a latent period with a mode about three to four years after 

 a single heavy dose of radiation. 



(ii) Retrospective study of leukaemia in childhood shows 

 that children under 10 with leukaemia include a higher 

 proportion (15 %) who were irradiated in utero than is found 

 in normal control children (8%) (Stewart et al, 1956). The 

 implications of this finding are {a) that 85 % of leukaemia in 

 childhood is certainly, and another 8 % probably, not due 

 to irradiation in utero. (b) The normal chance of death from 

 leukaemia in childhood (about i in 2000) is doubled (to i in 

 1000) by the fact of foetal X-radiation. 



(iii) Persons exposed to atomic-bomb radiation at Hiro- 

 shima and Nagasaki in 1 945 have shown an increased incidence 

 of leukaemia, again with an approximate linear relationship 

 of incidence to dose. 



It can be calculated (Lewis, 1957) from what has been 

 published that the effect of i roentgen of whole-body radia- 

 tion is to give a probability of dying of leukaemia of about 

 I to 2 X io~^ per person per annum. 



This does not, however, indicate that the whole of the 

 increase in leukaemia since 1920 has been due to the in- 

 creasing use of X-rays. Unless there is some quite unsuspected 

 flaw in data or logic, the very striking increase in leukaemia 

 in this century involving all ages is due to some aetiological 

 factor other than ionizing radiation, which is unlikely to be 

 responsible for more than 5 to 10% of the increase. The 

 nature of the other factor or factors provides one of the im- 

 portant challenges to present-day preventive medicine and it 

 may be of interest, although not perhaps very closely relevant 

 to my main theme, to enumerate the epidemiological indica- 

 tions which may be relevant to the nature of these factors : 



197 



