Document zz3Zjg3bmNYmmOG8LaKpkorK7

AXCRICANJ O ~ R S A LOF EPIDEMIOLOGT Copyright Q 1971 by The Johns Hopkins Cniversit? V d . Y3, so. 3 Printed in I-.S.A. ORIGINAL CONTRIBUTIONS OCC~-PATIOSAI,F-iCTORS IS THE E P I D E l I I ( )LO(iY ( ) F LEI-I<EAIIA I S HIKOSHIlIA ASD X-4(;-lS-iK11 TOH.\SOSt-I<II I S H I M ~ I < I ~H. I2l < O l l I - OK.\D.I." T.4Ii.ISOHI TOMI\-.\SI~.'~ T.tIS0 TSt-CHI.\IOTO.' T.\J\;\dHI HOSHISO" ASI) l I I ( ' H I T 0 ICHIM.4 Rt-" ( Rc~.ri\-rtifor piibliixt ion July 20. l!)iO) Ishimaru, T. (Dept. Statisiics, Aromic b m b Casualty Commission, U.S. Marine Corps Air Station, FPO Seattle, Washington 987641, H. Okada, T. Tomiyasu, T. Tsuchimoto, T. Hoshino and M.Ichimaru. Occupational factors in the epidemiology of leukemia in Hiroshima and Nagasaki. Amer I Epidem 93: 157-165, 1971.- This epidemiologic study examined retrospectively occupational and other en- vironmental factors in relation to the incidence of leukemia. A total of 4 9 2 cases of leukemia with onset in 1945-1967 in Hiroshima or Nagasaki City, Japan, and controls, were selected from the Atomic Bomb Casualty Commission Leukemia Registry and interviewed during the period 1966-1969. In 3 0 3 pairs o f adult leukemia cases and controls the risk was approximately 2.5 times higher among those with history o f probable occupational exposure to benzene or its derivatives and medical x-ray. The change in relative risk o f leukemia due to the coexistence of the two factors (radiation and occupation) was also examined. benzene; carcinoma; epidemiology; leukemia; occupations; radiation effects ISTHODI'CTlOS It is well k1lon-nthat exposure to ionizing radiation froill t l l r .l-homh has induced leu- -kemia among tIir IieaviIy exposed survivors ' From the At.omic Bomb Casualty Commission. Hiroshima and Sagasaki, Japan--A Cooperative Research Agency of U.S..4. Xational .irademY of Sciences-National Research Council and Japanese National Institute of Health of the Ministry of Health and Welfare. with funds p r o d e d by USA. atomic Energy Commission. Japanese ?;ah a 1 Institute of Health and rnited States Pnblip Health Service. 'Hiroshima Branch Laboratory. Japanese Na- Ilon:il Institute of Health. and Department of Sta' I s h Atomic Bomb Casualty Commission. a Department of Medicine. Atomic Bomb CasI ' a l t ~Commiwion. ' D(,partment of Medicinr. Hiroshima Vniver- ''I! Rrse:irc+ Institnte for Xiidear hlrdirine and ISiO!o~y.nnd D e p r t n 1 ~ n tof Mrdicinr. .Itomic. I ' ~ l ! l l t ~ C:rsualty Comniission. "8f'c'ond Intrrnal Mrdicinr Departnirnt. Far''''?.of Mcdivinr. Kyoto Vnivrrsity. and Departl'lf*nt of Mrtlirinr. .4toniic. RomIi Ca.w:ilty Com- ~lihrioll, a D('k)artnicnt of Hrm:cto!ogy. Atomic. Disease of the A-bomb in Hiroshirna and Kagasaki 1 ) . Court Brown and Doll (2, 3) have re- ported a high inc~idcnrcof leukemia among patients who received x-ray therapy for ankylosing spondylitis. It is clear that radiation can induce leukemia in man. On the other hand, animal experiments have shown that various chemical agents and viruaey ran also cause leukemia (4 1 . The Leukemia Registry has been in operation at Atomic Bomb Casualty Commi3sion (ABCC) since 1950 for the study of the leukemogenic effect of A-bomb exposure ( 5 ) .In this program, all possible cases of leukemia are searched for from a widr Institute. X:igasaki Vnirersity School of Medicine. and De1)artnifmt of Medicine. .\toniic Bomb Ca,.ualty Commiwion. Tlir authors eqires- their d c ~ pappre.c.iation to nienihrrs of the Field 0prr:rtions Section. Departmpnt of Statistie. ABCC. who c.onducted the intrrviews for this study. and to individuals at tiif. \-nrioirsmrtiical institutions in H i r o s l h i : ~and Sagi\s:tki who pro\idrd i h p nirdir:d inforiixltion for. t h r . I,rri!irini:i Hrgisiry. 158 ISHIMARC ET AL. range of sources. Screening of death certifi- decided to seek only simple information for cates and confirmation of diagnosis have which the chance of error due to the forget- been made for alniost all leukemia deaths fulness of the informant or type of informa- within the citiea of Hiroshima and Saga- tion would be small. Furthermore, inasmucll w k i since 1948. as most of the leukemia patients were dead The present study was designed in the a t the time of the interview. whereas the hope of obtaining leads to the elucidation of controls were often alive, different inter- what, if any, environmental factors other viewing approaches were required to obtain than A-bomb exposure, especially occupa- an environmental history. tion, have contributed to the development Information was obtained for both pa- of leukemia in Hiroshima and Nagasaki tients and controls using the same question- 1 since 1945, and how such factors in combi- naire by interview of the subject or nation with A-bomb exposure have affected closely related person. The choice of in- the development of leukemia. formant in order of preference was as fol- AIETHODS A S D MATERIAL4 lows: the subject, parent. brother or sister, child, a relative who had been living with Subjects the subject, and others. From among the patient> included in the ABCC Leukemia Registry as of December 1968. those satisfying the following condition,. were >elected ancl tlcsignnted a- intlcs cases : Patients diagiiosecl by the liemntologist to be leukemia and whose certainty of diagnosis was "definite" or "probable". Resident a t time of onaet in either Hiroshima or Nagasaki City. O i w t of diseaze betwen 1943 nncl th(5 entl of Deceniher 1967. For e:tch incles (`:I-e. a, inatclic~lc`otitrol nn- >elected from on(' of two -\BC`(` fisotl -nniple- ( 6 . 7 1 . Controls for patient; horn hefore Aiugu+t1945 were d r a w froiii thr > Inster Paniplti ( 6 1 , rvhile controls for pa- tients horn inter were >elected from th(, .:imple I 7 1 for the 1Iortality Study of Off-/)ring of .~-I)omt)Survivors. The controls n - c ~ t inintclicvl n-itli tlic paticlit. on tiit. follon-ing fivo rlial.:irt~,i.irtic>:city: -ex: cl:ittx of hirth z 3 0 iiionthz: esposurc rli-tance I less th:in 1400 111. 1400-1999 in. 2000-9999 111. not in city :it the time of the bombs 1 .iTB 1, born after bomb) ; alive and resident in either Hiroshima or Sagasaki City at time of onset of disease in patient. For adult leukemia cases and controls, the qucbstioning identified persons 1vho had ever worked as a production process worker or R nietlic.:il n-orkcr. For children with onset of leukcmia at less than 15 years of ~ i g eand their controls, information was obtninetl on birth order, mother`s age at time of birth. any abnormality of mother or subject during pregnancy or at childbirth, and history of medical s-ray csl)osure of the mother during pregnancy. 1nforni:ition col ltxctthcl on nI I s ut),i ects in- clutlctl : cnvironnicnt of rc..:iclcnce at the time of o i w t oi r l i > c . n ~ t ~l:ii.-tnry of contact n-itli animals: t l w t l i - c l u t ~t o c:incer among parvntz. sihliiip or c~liiltlrc~:tt nn(l consanguitI ity of parents. Bccauw of the sniall nunil)cr of cases, Hiroshimn m(1S a p s a k i w r e combined in the. rom~):irizonof t h r intlc's t>:1ses and control-. ll7Iien any factor w:is i o u d t o be as- >oci:itc+cI \vi t 11 t IN' (lcw*lopnicntof leukemia, t h c ~(lata w w c.saniitit~t1separately by city to clctrniiinc n.21cthcr ttw m i i a~ssociation coitlil tw swn in each city. Iicfilntivcrisks for i t i t l r ~ s twos :in(l controlz ior c.nch factor ncrta coniparcd using the matched sample tnt~tliotldescrilml by 3I:intcl ancl Haenszel (8,. 111v i m of the difficulties of retrospective d u d y over a span of 10 or 20 year.. it was A total of 492 patictits i n hot11 cities were LEUKEMU IN HIBOSHIMA AND NAGASAKI 159 Controls Kind of infomlnt Subject Parent spwse Siblings Children others Total Proportion Index e m Kind of informant Subject Parent Spouse Siblings Children Others I 9 155 130 37 21 61 2.18 37..53 31.48 8.96 5.08 14.77 Tots1 159 130 63 25 12 24 413 100.0 Proportion 38.50 31.48 15.25 6.05 2.91 5.81 loo. 0 onset of leukemia by the end of December 1967. For 15 of the leukemia patients it was not possible to obtain a control matched on all five factors. For 64 pairs of leukemia caseg and controls, interviews could not be Completed for one or the other because all m i b l e informants had moved out of the Oitiea. Thus, during the period from September 1966 to March 1969, an interview W8a completed only for 413 matched pairs of leukemia patients and controls. Table 1 compares the kind of informants for the index cases and controls in the 413 matched pairs for whom an interview was completed. .The type of informant was the ssPhe in only 156 pairs, or 38 per cent. There was a large difference in the type of informant between cases and controls msinly because the subject was himself the ~ o r r n a nitn 39 per cent of the controls, but in only 2 per cent of the index cases. This W(Ls because most of the leukemia patients had died by the time of the interview, while most of the controls were still living. Although such obvious biases were to be expected in this interview survey, no restriction was made because it was felt to be impractical to exclude the subjects from the interview. In order to reduce problems aris- ing from such differences it was decided to seek only simple items of information which, it was thought, the experienced ABCC investigators could obtain without bias as between patients and controls. The responses to the various questions were compared for the 413 matched pairs of index cases and controls who had been interviewed. History of possible occupational exposure to benzene or medical x-ray The risk of leukemia has been reported to be high among workers who handle benzene or are exposed to medical x-ray. Therefore, the following 15 occupations (9)in which benzene or x-ray exposure may be frequent were selected for a comparison of the index cases and controls. In the study of these occupations, 303 adult cases with onset of leukemia at age 15 years or over and their controls were compared. The occupations were: (a) carbon dioxide gas furnace worker; (b) welder, plater, tinsmith or sheet metal worker; ( c ) battery manufac- turing worker; (d) cabinet maker, furniture maker or finishing worker; (e) worker en- gaged in printing, repairing or cleaning printing machines; (f) platemaker, lithog- 160 ISHIhABU ET AL. TABLE 2 Comparison of occupational history related to potential ezposure to benzene or medical x-ray, frequency of occupational history for 11 occupations Controls Yes No Index cases Yes No _.- 2 12 30 259 Relative risk 2.5 (30/12) x' test x' = 6.881 df = 1 p < .01 TABLE 3 Comparison of occupational history related to potential exposure to benzene or medical x-ray; and duration of occupational history for 11 occupations Controls. Duration (years1 Index cases* 1Yes No -25 -<5 -0 Yes 25 11 8 <5 00 4 No 0 24 6 258 Relative risk: 2 5 years 3.0 (24/8) x' = 7.031, p < .01 <5 years 1.5 (6/4) P > .05 df = 1 1 pair was rejected because of unknown in- formation regarding duration. risk of leukemia was found to be 2.5 times higher among those with a history of any of these 11 occupations in comparison with those without, and the risk increased with longer duration of occupation. Comparisons were made within each city and the relative risk in each city was about the same, 2.6-2.7. Table 4 shows the frequency of each of the 11 occupations in the index cases and controls. The frequency of each individual occupation is very low. However, the frequency of the following six occupations was found to be higher in the index cases: 1) welder, plater, tinsmith or sheet metal worker; 2) cabinet maker, furniture maker or finishing worker; 3) rubber products worker; 4) glass craftsman or pottery painter ;5 ) barber, beautician or dry cleaning worker; 6) radiologist or x-ray technician. No investigation was made of the specific chemical agents handled by the individuals engaged in these occupations. These are, TABLE4 Frequency of history for each of 11 occupations for index cases and conlrols Occupation i1 A Leutemia i cases (303 pairs) rapher; ( 9 ) rubber products worker; (h) leather products worker; (i) glass craftsman or pottery painter; (j) painter; (k) chemical worker; (1) soft drink manufacturing worker; (m) barber, beautician or dry cleaning worker; (n) radiologist or xray technician; (0)physician, or laboratory technician. In fact, however, no leukemia case or control had a history of the occupa- tions (c), ( f ) , ( j ) , ( o ) , so further discussion is limited t o the 11 occupations that were actually reported. The risk of leukemia is shown in table 2 in relation to a history of any of these occupations, and in table 3 by the duration of such work subdivided into five or more years, less than five years, and never. The 1. Carbon dioxide gas furnace worker 11 4 (1.3). 2. Welder, plater, tinsmith or sheet metal worker j6 (2.0) 3. Cabinet maker, furniture maker, 7 or finishing worker ' (2.3) 4. Worker engaged in printing, re- 1 pairing or cleaning printing (0.3) machines 6. Rubber products worker 2 6. Leather producta worker (0.7) 1 7. Glass craftsman or pottery (0.3) 2 8. Chemical worker (0.7) 1 9. Soft drink manufacturing (0.3) 1 worker (0.3) 10. Barber,beauticianor dry clean- 6 ing worker (2.0) 11. Radiologkt or x-ray technician 2 (0.7) Parent6ews show percentages. 161 however, occupations in which organic sol- TABLE 5 vents or x-rays are generally used. Thus, this suggests that these occupations and in particular the occupational use of benzene for a long duration increases the risk of leukemia. The leukemia risk was examined in Relationship between occupational ezposure to benzene or medical z-ray and atomic bomb ezposure status; and frequency of occupational history for 11 selected occupations by three groups of ezposure status relation to a history of any of the 11 OCCU- pations and chronicity of leukemia. The relative risk was 2.9 for acute and 1.8 for chronic leukemia. No relationship was evident between a history of one of the 11 listed occupations 1-Y-es -No .-I ) Inner proximal ex- Yes o 2 Posed (100 rnd or more) 1.5 (3/2) and year of onset of leukemia: For cases with onset between 1945 and 1957 the relative risk was 2.6, and for the years 1958 t o posed (1-99 rad) 4.0 (411) S o 4 36 1967 the risk n-as 2.4. 3) Distal expmed and Yes 1 9 2.4 (22/9) To guard against the possibility that the data might be affected by the fact that the none) kind of informant was not always the same for the leukemia cases and the controls, the pairs were subclassified into those in which the kinds of informant were identical, and those in which they were different. The rel- Hiroshii Xagssaki T65 Dose (md) ative risk was 11.0 (11/1)for the pairs that had similar informants, but only 1.7 (19/ 11) for the pairs where they were different. 1) <1200 rn 2) 12061999 m 3) 2000 m 6 over Non-esposed <I500 m 1500-2499 m 2500 m & over Xon-expmed B 100 1-99 Negligible h'one Clearly, the effect of differences in kind of informant was to reduce markedly the ap- risk of leukemia, particularly chronic leu- parent relative risk. kemia, only among survivors in Hiroshima Effect of A-bomb exposure and history of occupational ezposure to benzene or medical x-ray upon the development of leukemia (2) The cases were therefore classified into the three groups described below, and in table 5 the relative risk of leukemia in these three exposure groups is compared with re- Recent estimates of the radiation dose re- spect to a history of any of the selected ceived by A-bomb survivors (tentative 1965 occupations. The heavily exposed group, in dose-T65D) (10) indicate that doses of which the dose is estimated to have been 100 rad or more had been received within 100 rad or more, had a relative risk from 1200 m in Hiroshima and within 1500 m in the selected occupations of 1.5,whereas the Nagasaki. The dose was lower, generally in low dose group, consisting of outer proxi- the range of 1 to 99 rad between 1200-1999 mally exposed individuals, had a risk of 4.0, m in Hiroshima and between 1500-2499 m and the unirradiated other group had a rel- in Nagasaki, and was negligible beyond ative risk of 2.4.Of course, among the prox- 20o0 m in Hiroshima and 2500 m in Naga- imally exposed persons, a history of any of saki. Exposure to doses of 100 rad or more the selected occupations was noted in only has been found to result in a marked in- eight of the leukemia patients and four of crease in leukemia in both Hiroshima and the controls. These 12individuals are shown Nagasaki, while exposure to lower doses, by T65D, disease type, occupation, etc. in less than 100 rad, has resulted in increased table 6. Examination of the relative risk 162 ISEIIMABU ET AL. TABLE 6 Relationship between occupational ezposure to benzene or medical z-ray and atomic bomb ezposure status; eight prozimal ezposed cases and four controls with a history of any of 11 selected occupations Index ca8~1h: istory of selected occupation Controls: hirtory of aelected occupation IYa. No' 22;MF No. Typet T6SD rad MF --No. Trpe %: D Occupation: f 4: Yes* 249-266 32 CGL 31 2 121-894 Control 361 1 295-993 - Control Not esti- 1 441-238 Control 226 1 & 8 mated 207-042 Control 32 3 No* 400-538 31 AL 295-870 51 AL 220344 51 CGL 437-743 70 AL 400-257 23 AL 233-136 64 AL 138-148 50 AL 573 11 212 10 44 8 38 2 19 3 17 3 62 I Answers to history of 11 selected occupations. t AL = Acute leukemia; CGL = Chronic granulocytic leukemia. $ Number corresponds to occupation in table 4. TABLE 7 Relationship between occupational ezposure to benzene or medical x-ray occupations and atomic bomb ezposure status; frequency of history of 6 selected occupations* I I IYes No 1 1 1 4.0 (24/5) U = 1 No 24 27: P < .01 after classifying the cases grossly into high and low dose groups revealed an occupa' tional risk of 1.0 in the high dose group. The relative risk in the low dose group, however, was 5.0, approximately twice as high as the risk of 2.4 in the distal exposed and non-exposed groups, but the number of cases is small. The six occupations noted at a higher frequency among the index cases than in the controls in table 4 were selected for an examination in more detail of the relation of occupational exposure and A-bomb exposure (table 7). In general, the risk was approximately 5 times higher among those with a history of any of these six occupations in comparison with those without. The relative risks in the proximal group and in the distal and non-exposed group were 6.0 and 4.5, respectively. Age of mother, birth order and abnormality during pregnancy or at childbirth The 106 matched pairs of children with onset of leukemia at less than 15 years of age and their controls are compared with respect to mother's age a t time of birth of subject, and birth order. However, no &atistically significant relation could be found between mother's age or birth order and the LEUKEWA IN EIIBOSESMA AND NAGASAKI 163 risk of leukemia. Information was also obtained as to whether the mother had any abnormality during pregnancy or a t birth of the subject. Abnormalities were reported in 11of the 106 index cases but in only two of the 106 controls. In the leukemia group, nine reported severe "morning sickness" or other complications during pregnancy, but since it is possible that the mother of index cases had greater concern over such abnormalities during pregnancy as compared with the c c ~ t r d s i,t cannot be concluded that abnormalities during pregnancy increase the risk of leukemia. Information on exposure to medical x-ray was sought, but the number of cases was small and no relation could be found. Consanguinity of parents and environmental factors Consanguinity of parents and such factors as socioeconomic status and history of contact with animals, including pets were examined, but none appeared to be factors related to the incidence of leukemia. DISCUSSION Numerous epidemiologic studies have been made on the etiology of leukemia. Of the various causes of leukemia in man, most evident is the effect of radiation. It is well known that radiation can induce leukemia, and that there is a marked increase of leukemia among A-bomb survivors and among Patients receiving radiation therapy for anblosing spondylitis (1-3). On the other hand, the results of animal experiments suggest that viruses may be one of the causes of leukemia (4), but there are few reports of human experience which Provide epidemiologic evidence of a transmission theory beyond the report (11) of the development of eight cases of leukemia between 1957 to 1960 among children at the Primary school in Niles, Illinois. Other causes of leukemia reported from experimental studies include various chemical agents which disturb the hematopoietic function of the bone marrow. A typical ex- ample is the induction of leukemia by benzene or its derivatives. For this reason, 11 classification of occupations in which benzene or its derivatives, or x-ray is frequently handled were selected for a study of the risk of leukemia by whether there had been a history of such occupations. The risk of leukemia was found to be significantly higher (about 2.5 times higher) among those with a history of such occupations as compared with those without, especially among those who had engaged in these occupations for a t least five years. The risk of leukemia was also examined by whether there had been a history of having worked as a production process worker in any occupation other than the above 11,but no significant relationship could be found. The increase of risk was noted specifically only in those occupations in which there is frequent contact with benzene or x-rays. Examination in more detail revealed that the risk was higher in those occupations in which various volatile organic solvents are used. Approximately one-third of all cases of leukemia in Hiroshima and Nagasaki are among the proximally exposed persons and presumably are due to exposure to significant amounts of radiation from the Abomb. As shown in tables 5 and 6, the distal and non-exposed group had a risk of leukemia about 2.5 times higher among those with a history of occupational exposure to benzene or medical x-rays as compared to those without. I n the so-called heavily exposed group, however, the effect of A-bomb radiation is so large that there seemed no change in the risk by whether or not there was a history of any of the selected occupations. In the low dose group, the risk was approximately 5 times higher among those with a history of any of the selected occupations but the numbers are so small that the relative risk is not measured very accurately. However, of the five leukemia patients in the low dose group with a history of such occupations, four were acute leukemia so that this does not explain the high 164 ISHIM?LRU ET AL. risk of chronic leukemia that has been noted in the so-called lightly exposed group of Hiroshima (1). Manning (12), Stewart (131, Macblahon (14),and others have reported that the risk of leukemia is high in children of older mothers and in the first child. Therefore, the risk of leukemia was examined in relation to age of mother and birth order, but no statistically significant evidence was found that these factors affect the risk of leukemia. Awa, Bloom et al. have reported that the incidence of stable chromosome aberrations increase with age (15,16). When considered in association with the report of the possible relation between congenital or acquired chromosome breakage and the development of leukemia (17) and the report of the higher frequency of Down's syndrome among children of older mothers (18), the suggestion that there is a tendency for the risk of leukemia to increase with the age of mother may be epidemiologically significant. However, the mother's age was 40 or over in only six cases of leukemia so that it cannot be concluded from the present investigation that the incidence of leukemia increased ~ignificantlywith age of mother. -4number of reports have been published on the relation between Down's syndrome and leukemia in children (19-21), but m-e did not attempt to study this condition since accurate information concerning such medical matters could not be obtained from interview oi persons closely related to the patients. Questions were asked about congenital abnormalities, but none were reported so that retrospective study of such conditions by interview does not appear to be practical. With regard t o the relation between leukemia and socioeconomic factors, it has been reported in the United States that the risk of leukemia is increased among persons of higher socioeconomic status and among Jews (22). Therefore, in an attempt to study the effect of social factors in the development of leukemia, the social environ- ment of the residence at time of onset and the history of contact with animals were reviewed for possible differences between the leukemia patients and the controls. If it is assumed that leukemia is transmissible and that animals carry the leukemia vim, the proportion of households in which there is greater contact with animals should be higher in the leukemia group as compared with the controls, although Miller s t a h that even if viruses contribute to the development of leukemia, it would be extremely difficultto demonstrate horizontal transm-8- sion of leukemia epidemiologically (8). Moreover, we also wanted to examine any difference in socioeconomic factors between leukemia cases and controls. The present study, however, did not show any significant difference by occupation of head of household, environment of residence, con- sanguinity of parents, whether ever bitten by a dog or whether any animal had been kept during the five years prior to onset, 90 that there was no finding suggesting transmission via man or animal or the importance of socioeconomic factors. The present study was done using the ABCC Leukemia Registry and the data that have been accumulated over the ap- proximately 20 years since 1950, in order to determine whether any factors other than A-bomb exposure are related to leukemia. It was thought that a case history study would be the only possible approach to the investigation of factors involved in a disease with so low an inciderice as leukemia. Even for such a study, however, no effective method of investigating patients who had developed leukemia over the past 10 or more years could be found other than to inteniew closely related individuals. On the other hand, most of the members of the ABCC fixed sample who were chosen for use as controls were alive, whereas the index cases were dead so that there wae some doubt concerning the homogeneity of the two groups as to the accuracy of d&@ collection because of the difference in dd type of informant for each. Nevertheless, it is hoped that the present results will be suggestive for future studies. FLEFEREXCES 1. Ishimaru T, Hoshino T, Ichimaru M, et al: Leukemia in atomic bomb survivors, Hiroshima-Nagasaki, 1 Oct. 1950-30 Sept. 1966. ABCC Technical Report 2549,1969 2. Court Brown WM, Doll R: Leukemia and Aplastic Anemia in Patients Irradiated for An- kylosing Spondylitis. British Medical Research Council Special Report Series No. 295 1-50, London, Her Majesty's Stationery O5ce, 1957 3. Court Brown WM, Doll R : Mortality from cancer and other causes after radiotherapy for ankylosing spondylitis. Brit Med J 2: 1327-1332, 1965 4. Rich MA: Experimental Leukemia. New York, Appleton Century Crofts, 1968 5. Finch SC, Hrubec Z, Nefzger MD, et al: Detection of leukemia and related disorders, Hiroshima-Nagasaki. Research plan. ABCC Technical Report 5-65,1965 6. Beebe GW, Uaegawa M: The major ABCC aamplea. ABCC Technical Report 1268,1968 7. Hato H, Schull WJ: JKIH-ABCC Life Span Study of children born to atomic bomb survivors. Research plan. ABCC Technical Report 4-60,1960 8. Mantel N, Haenszel W: Statistical aspects of the analysia of data from retrospective studies 11. Heath CW, Hwterlik RJ: Leukemia among children in a suburban community. Amer J Med 34: 796-812.1963 12. Manning MD, Carroll BE: Some epidemiological aspects of leukemia in children. J Nat Cancer Inst 19: 1087-1094,1967 13. Stewart A, Webb J, Hewitt D: A survey of childhood malignancies. Brit Med J 1: 1495-1508, 1958 14. MacMahon B, Newill VA: Birth characteriatic of children dying of malignant neoplasms. J Nat Cancer Inst 28: 231-244,1962 15. Awa AA, Bloom AD: A review of cytogenetics at ABCC, including a report of a symposium. Jap J Hum Genet 12: 69-75,1967 16. Bloom AD, Neriishi S, Awa AA, et al: Chromosome aberratione in older atomic bomb mrvivors, Hiroshima-Nagasaki. Lancet 2 : 802805, 1967 17. Sawitsky A, Bloom D, German J: Chromosomal breakage and acute leukemia in congeni- t d telangiectatic erythema and stunted growth, Ann Intern Med 65: 487495,1966 18. Slavin RE, Karnada N, Hamilton HE:A cyto- genetic study of 92 c a m of Down's syndrome, Hiroshima-Nagasaki. Jap J Hum Genet 12: 17-28,1967 19. Stewart A: Aetiology of childhood malignancies congenitally determined leukemia. Brit Med J 1: 452460,1961 20. Wald N, Borges WH, Li CC, et al: Leukemia associated with mongolism. Lancet 1: 1228, ernment Printing Office,1964 10. Milton RC, Shohoji T; Tentative 1965 Dose &mation for A-bomb survivors, HiroshimaNagasaki. ABCC Technical Report 188,1968 1957 23. Miller RW: Persons with exceptionally high risk of leukemia. Cancer Research 9 : 2420- 2423,1967