Document zzwj6GLZavJprvro5M09Jpye3

//' Evidence for Radiation and Chemicals as Leukemogenic Agents ) E U G E N E P. CRONKITE, M.D. L UPTON, N.Y. Introduction Modern society is replete with hazards of life on all hands. The hazards in the home, daily living, industry, and recreation can only be considered sensibly when divorced from politics, diplomacy, and emotions, and compared with the benefits to society at large or the personal satisfactions gained in particular. It is human for an individual to take a risk. It is reprehensible for individuals to be forced to submit to risks of which they have neither knowledge nor the opportunity to choose to accept the risk. For the public to accept a risk they must know what it is and the magnitude of this risk. Since the average citizen can neither understand nor equate the hazards, these must be performed by appropriate scientific groups and duly considered by legislative bodies for appropriate action. The heart of the problem in the induction of leukemia by industrial hazards rests upon quantitation of the agent to the yield of leukemia. Two agents used in industry have been correlated with an increased incidence of leukemia in human beings. The first, ionizing radiation, is unquestionably able to increase the incidence of leukemia. The second agent, benzol, probably can produce an increased incidence of leukemia, but the data are not as good as for the former. First, the finger of suspicion must be pointed at any agent which is able to produce an aplasia of the bone marrow, assuming it will probably be able to produce leukemia also. Second, there is no reason to doubt that any agent which will produce a cancer elsewhere in the body will not also be able to produce a leukemia if the offending agent is transported to the hemopoietic tissues. The types of radiation exposure of human beings that have resulted . : ~ Commission. Head, Division of Experimental Pathology, Medical Research Center, Brookhaven National Laboratory. Presented at the 13th Annual Meeting of the American Academy of Occupational Medicine, Detroit, Feb. 8 to 10, 1961. in an increased incidence of leukemia, the dose-eff ect relationships, and evidence for chemical leukemogens will be reviewed. Induction of Leukewia by Ionizing Radiation in Man.-Historically, leukemia was correlated to overexposure to x-ray in the professional use of x-rays over 40 years ago? Since the initial correlation there has been a continual procession of cases. Most of the earlier observations correlating radiation exposure in the practice of medicine or industrial use of radiation with the later development of leukemia only serves to demonstrate the fact that radiation can be leukemogenic, and does not answer the important problem of whether there is a threshold dose or the type of the dose-effect relationship. I n discussing the leukemogenic action of radiation, it is necessary to recognize that this constitutes only a small fraction of the total problem of the radiation hazard, and that one is also concerned with carcinogenesis in other organs, life shortening, decreased fertility, cataracts, and genetic damage. There are 2 sets of data; one involves a single, relatively large exposure of human beings, and the other involves chronic exposure. The largest single amount of data is that which has been gleaned from the survivors of the Japanese atomic bomb e~plosion,2w.~hich has been reviewed critically.* After the single large exposure as occurred in Japan, many of the individuals exposed had a severe temporary aplasia of the bone marrow. In those that recovered from this, there was a definite increase in the incidence of leukemia that seemed to commence about 11/2 years after the bombings. However, in reality the minimum latent period may actually be shorter, since adequate clinical studies were not practical due to the chaotic conditions in the earlier period after the bombing. The largest number of cases of leukemia appeared in 1951 and 1952. Thereafter, the number apparently decreased, but cases of leukemia in excess of the expected incidence were still being found in 1960. Many observers have analyzed the incidence of leukemia in Japan. The modal incubation time is approximat 5 years, and it has been deduced by et al.: on the basis of both the Japandata and the incidence of leukemia in s ~ - ' dylitics given therapeutic spinal irradiatiqa that probably !30% of the cases will occur within approximately 10 years after expo- sure. In view of the various analyses and & data published, it appears that an assumpti& that the incidenke of leukemia will continut at a constant rate throughout the duration of life as has been assumed by Lewis? is not justified. The most valuable p u b l i s u data for studying the dose-effect relationship of leukemia is that of Heyssel et al.8 M q ; investigators have plotted the incidence of leukemia against the computed dose of ation in Japan. Burch has shown that data can be treated as a curvilinear relati ship. Heyssel et al.g have chosen to it as best fitting a linear no-thresh tionship. When one treats the data incidence versus dose were linear, ognizing that the slope of the line is not known with precision, one can era1 estimate that the incidence o per rad, measured outside Japanese per million people at risk per a single exposure, is about 1 case per. population at risk for at least the first 10 15 years after the exposure. S the dose within the homes nor the dose is known adequately, it is not ye sible to estimate the incidence absorbed dose, the entity in whi interested. However, the incid be greater for absorbed dose th above. Another set of data is the of leukemia in persons with spond were exposed to therapeutic spi These data have been extensi by Court-Brown and gators indicate that the relationship of dose to leukemia may be linear. They the analysis of the data that there may not be a threshold dose, and that the dose which would double the incidence of of the order of 40 rads. They deduce the probability of developing leukemia per year UKEMOGENIC AGENTS system to be approximately that of children have been irradiated shortly was deduced for the Japanese. In after birth for a presumed enlarged thymus. g, it is of interest to note that the nce limits of their threshold dose therapy resulted in an increase in the incibetween 0 and approximately 400 dence of leukemia. However, because of uncertainties in dose, the port size, the pre- very critical study from the standpoint cise area radiated, and the absence of an determining the relative magnitude of a adequate control population, it is not feasible nimal threshold dose for leukemogenesis to deduce any realistic yield of leukemia per the asserted correlation of in utero expo- rad of exposure to this area. from pelvic diagnostic x-rays to the A series of observations on x-ray therapy, development of leukemia in the off- repetitive diagnostic x-rays, radiothorium, Stewart et a19 have reported a and 1131 therapy for cancer indicates that incidence of leukemia in children cx- all of these situations can be leukemogenic, in utero to diagnostic radiation than but it is not practical to deduce any realistic control population. I t is imperative dose-effect re1ati0nship.l~ Since IlS1may be blish whether diagnostic x-rays of the used under various industrial conditions a in pregnancy will induce leukemia. further discussion of the hazard of this if this small dose (1-4 rads) is isotope appears in order. Thousands of peo- nic it indicates a very low thresh- ple have been treated with multimillicurie that the fetus is extremely sensitive amounts of for hyperthyroidism. How- development of leukemia. However, ever, there is no well-documented case of a very difficult problem to analyze leukemia in these individuals that could be ally. Court-Brown and Doll lo have reasonably associated with the prior treat- ly published the 10-year incidence of ment with for hyperthyroidism.14 ia in 40,OOO children who received in However, following a case of the treatment o irradiation. The incidence of leukemia of thyroid cancer with doses of P1totaling se 40,OOO children who had received in excess of 1 curie the incidence of leuke- in utero was essentially the same as mia is greater than would be expected.l8 ence in the control population of Lastly, one comes to the consideration . Whereas this is reassuring, it does of occupational exposure in radiologists, answer satisfactorily the question either nurses, technicians, and physicists, and what itivity of the fetus or of a threshold. might be expected to occur in the chemical assumes the same sensitivity of the processing of radioisotopes and the utiliza- and the adult and computes the in- tion of radioactive sources and isotopes in incidence of leukemia in the children industry. The incidence of leukemia in e first 10 years after birth, one would American physicians is about 1.7 times what difference of only one case,* assum- it is in the population at 1arge.l5 The inci- rad exposure and using the doubling dence of leukemia is higher in radiologists radiation suggested by Court-Brown than in nonradiologists.l8 In Great Britain Thus, to answer this question, the incidence of leukemia is not higher in ,OOO children would be needed, but, radiologists than in nonradiologists. This is nearly, 4,000,000 births would have to probably ascribable to the fact that almost all x-rays in Great Britain were taken in diation of the thymic area alone has hospitals by radiological technicians in concorrelated with a later development of trast to the United States, where a high iaI2 and reviewed.' A large number fraction are taken in private offices by the thod of analysis suggested by Dr. Britain. The existence of an. increased in- ' any realistic dose-effect relationship. Another important medical legaLc Radium was administered to many indi- sation and industrial problem is the viduals for therapeutic purposes before its hazard was well appreciated, and a larger relationship when a case of leukemia number of individuals were exposed in industry utilizing radium and mesothorium individual who has had prior for watch dials. The incidence of osteosar- , radiation. If the leukemia de coma, cancer of the nasal sinuses, and aplas- 15 years after exposure, and i tic anemia was high in these individuals. unproved linear n e t In the watch dial painters, Martland's de- scriptions" of an intense replacement of the normal fatty marrow by a red actively regenerating bone marrow with numerous primitive red and white cell precursors is classic. However, the conclusion of Martland, -"it would seem very likely that a myeloid leukemia could easily develop" has than 100 rads and with chroni these estimates will have little It is well known that develo not been realized except in one case report. The reason for the relatively small yield of leukemia with radium poisoning is not clear. The industrial problem becomes then one process, and it is impossible to is going to happen However, after the of ability to predict the hazard and make one can compute the probabili calculated risk decisions. The problem of relationship to prior exposure, the dose rate dependence is therefore .crit- has done: ical. For the single dose, the Japanese data As an example, one can take a case are reasonably good and g'ive one a reason- able estimate of the excess leukemia to be expected in a population receiving doses in excess of 100 rad in a single exposure. For doses less than 100 rad it is not as reliable because of the few observations between no dose and the lowest, well-documented dose. None of the human sets of data provide a satisfactory dose-effect relationship below 100-rad exposure or for chronic radiation. However, one can lean in part upon the animal observations of Mole,l* who has shown that animal leukemogenesis is of leukemia with age, the chances being responsible are approximately if he develops leukemia at age 40, at age 60,and 20 out of 2% at age dependent upon dose rate. However, in his experiments, the dose rate was also relative- ly high compared with background, diag- nostic, fall-out, and industrial exposure. I t is hazardous to make quantitative extrapola- Before leaving radiati tions from animals to man; however, it is ment on the maximum equally hazardous to set up assumptions for man that are in a variance with the wealth these have been reduced of animal experimentation available. If one from 100 r e m per year LEUKEMOGENIC AGENTS -per y- ear. I t is mistakenly believed that the that 4 ounces of Brazil nuts contains 3 times reductions have been refommended by the as much radium and thorium as the entire National Commission of Radiation Protec- bone and soft tissues of a normal adult. In tion and the International Commission of typical British irony, "If the level of SrW Radiation Protection b e c a u s e evidence in food in this country represents a radia- shows somatic injuries at the earlier levels tion hazard, and apparently does to the arm- of 100 rem. This is not the case. There is chair calculator, then the whole process of not a single case of proved personal injury eating food must be regarded as highly to an individual exposed in the so-called dangerous." maximum permissible dose range. The max- Chemicals as Leukemogenic Agents in imum permissible dose ranges have been Man.-The case for radiation being leuke- reduced purely because of genetic considera- mogenic can be considered air tight. HOW- tions-possible but not proved harm to gen- ever, the case for chemical leukemogenic erations yet unborn. The atomic energy agents in man is not nearly SO good. AS industry has a remarkable record of safety stated in the introduction, any agent capa- in respect to radiation injury, and the plan- ble of producing an aplasia of the bone mar- ners are to be commended. row should be considered as potentially Philosophically, in respect to somatic and leukemogenic. The chemical substances that c hazards, Bruesm has pointed out have been accused of being responsible for `the doctor or the health physicist is the later development of leukemia in animals practical questions and, since progress are legion. However, to date only benzol does not or cannot wait for the scientific has been investigated sufficiently well to wers, is expected to give an answer be- make it appear a serious contender for the adequate data are present. In the case part of a chemical leukemogen in man2l on he is forced to employ such A much better sound statistical examination as, "safe," "permissible," or "tol- of the danger to industrial porkers of ben- in order to satisfy the human, social, zol is needed comparable to the extensive ethical requirements. This problem be- retrospective and prospective studies with omes extremely unscientific when one be- sources of radiation. mes involved in the area of low dose Apparently the first case of leukemia in effects. In this area the theoretical scientist a benzol worker*l was published in 1928, may take a low incidence and multiply it by and the second21 in 1932. The typical his- tion of the world and come out tory is that of prolonged exposure to benzol at appears to be an intolerable situ- vapors in rubber or other industry for many on, ignoring the calculated risks in all years, with ultimate development of fatigue, an endeavors. When death, disease, or anemia, leukocytosis, immature cells in the ury is being produced by a factor in so- peripheral blood, and death. Cause of death ty it can only be considered in relation in some instances was ascribed to aplastic the benefit obtained. An interesting com- anemia, and in others, to leukemia. The which helps to place the problem in autopsy studies of Mallory et a1.22 and rots if not correct perspective is an others implicate leukemia in some cases and icle that appeared in the London Times, aplastic anemia in others. When the correla- .9, 1959. Professor Mayneord ( a dis- tion between an exposure to a toxic sub- `shed x-ray physicist) and associates stance and the later development of disease zed carefully elements of the British is questionable in man, one naturally turns nd concluded that a typical high carbo- to animal experimentation to implicate the ate British diet contains amounts of agent as being potentially leukemogenic. and thorium equivalent in effect to1 I n this case the numerous animal studies ?me 300 times the present intake of SrgO remain equivocal. The published human . - benzol to leukemia and other blood dyscrasias is akin to the situation with benzol vapor, and whether the exposur radium. Martland,17 in his classical series continuous or intermittent. Again, as in of studies on the late pathologic ef- case of radiation, individual susceptibi fects of radium poisoning, described the cannot be measured experimentally. cases which developed osteosarcoma, carci- length of exposure has nearly always noma of the nasal sinuses, and aplastic ane- very considerable, varying from sev mia. No mention was made of the hundreds years to several decades, and the disease or thousands of individuals who did not appeared many months after the ex& develop an obvious late effect. It was only has ceased. A few exceptions h in 1959 that a study was commenced in noted (the shortest duration of New Jersey of the total population of those being 8 mopths). The degree of exposed to radium in 1914-1919 and still also needs to be m?rked, and must alive, to try to obtain a better evaluation of high concentrations of benzol in the the hazard in a statistical sense. I n the in- and hematopoietic organs. dustries using benzol, extensive studies have been made on obvious hemopoietic diseases Summary and Conclusions that developed in these workers in Massa- Both single and prolonged ex chusetts by Bowditch and Elkin,22 x-ray and y-rays has been de and Mallory et al?` Eighty-nine individuals produce an increase in the inc were found with hematologic abnormalities kemia in man. who had been exposed to benzol. Ten died, There is no reason to question t one from acute leukemia. Additional cases mogenic capacity of all types and were studied by Mallory, and from 19 cases, of radiation irrespective of the SOU autopsies or marrow punctures, cases with viding that the hemojmietic tissues aplasia and normal cellularity, hyperplasia diated. of the marrow, and acute leukemia were ob- After single doses of penetrati served. The spleen in many instances or gamma rays, reasonable estim showed more or less widespread foci of increased incidence of leukemia hematopoiesis which, in advance cases, be- pected can be made. Comparable came extremely large and undifferentiated. after chronic exposure, intermitt The transition to leukemia thus appeared to sure, and deposition of radioisoto be gradual in this material. In addition, in bone cannot be made. some cases, fibrosis of the marrow accom- Many chemicals and drugs h panied extramedullary hematopoiesis, and shown to be capable of inducing leu this fact has been interpreted as signifying animals. Of all the toxic agents that benzol can cause a syndrome of myelo- man is exposed in industry, a r fibrosis and myeloid metaplasia as well as cause and effect relationship has been leukemia with aplasia of the marrow as a lished only fur benzol. third possibility. The relation between leu- The absence of a positive correlat kemia and myelofibrosis is one which has been discussed in great detail by Dameshek and There can be no doubt that there have been "epidemics" of blood dyscrasias in industries using benzol, which certainly incriminates benzol as being a hematopoietic toxin. In these epidemics the cases of leukemia and myelofibrosis have been few indeed. As with radiation, one leukemia induction with other agents d m not give these other toxic agents a "el& *-bill of health," since all carcinogenic 1esses are statistical, and the unequiva 1 demonstration of a small effect may "54- : 'tate a sample size in excess of 1Oa. Si@ ian exposed group is a rarity, even in h u g industries. When a positive effect is d&- i onstrated in a small group the haza i must consider individual susceptibility, relatively great. The leukemia burdent LC U K C M U b C N 1L AIcrCIV 1 3 ..- .L---:- :-a..----- ...A.=+.... ram,,;n~ r,n- 9. Stewart. A.: W&b. T.: Giles. D.. and Hewitt- d a t e and weigh the risk demands quantitation often lacking or not evaluated because nf a n assiimerl exorhitant rnst in dollars. In ___- -_.- - - - merit of Thorax in Infancy, Can& 10:42, 1957. 13. Maloney, w. C.: Induction of Leukemia in Man bv Radiation. svmoosium on Radiation Biolop -------.- --*-----,, aven National Laboratory, Upton, L.I., 17. Martland, H. S.: The Occurrence of Malig- nancy in Radioactive Persons: Review of Data REFERENCES Gathered in Study of the Radium Dial Painters, Amer. J. Cancer 15:2435, 1931. Jag%, N. ; Schwarz, G., and von Sieben- 18. Mole, R. H. : The Development of Leukemia Bfutbefunde bei Rontgenofogen, KIin in Irradiated Animals, Brit. Med. Bull. 14:174, 1958. 2.`M010ne~W,. C.,and Lange, R. D.: Leukemia Atomic B o m b Survivors, Blood, 9663, 1954. 3. Heyssel, R.; A. B. ; Woodbury, L. A. ; W-ra, E. T,; Ghme, T.;Hahino, T., and ymaki, M.:Leukemia in Atomic Bomb Sur- E. P.; Maloney, w., and ~ ~ on Leukemogenesis, Amer. J. Med. 19. Bond, V. P.: Medical Effects of Radiation, Proceedings of the 13th Annual Convention of the National Association of claimants' Compensation Attorneys, 1959, PP. 117-129. 20. Brues, A. M.: Somatic Effsts, Low Level Irradiation, Washington, D.C., American ASSOci~ationdfor t,he Advancement of science, 1960, p. 72. 21. Dameshek, W., and Gunz, F.: Leukemia, New York, Grune & Stratton, Inc., 1959, Chap. V, S.;Miller, M., and Wald, N.: On the PP. 44-52- On of the Incubation Period in Malignant 22. Bawditch, M., and Elkin, H. B.: Chronic \ J. Chron Dis. 9:385, 1959. Exposure to Benzene (Benzol): I. The Industrial %Mrt-Brown, W. M., and Doll, R. : Nuclear Aspects, J. Industr. Hyg. Toxicol. 21 :321, 1939. f!hkd Radiations and the Incidence of Leu- 23. Hunter, F. T.: Chronic Exposure to Benzene: 11. The Chemical Effects, J. Industr. s E. P. : Leukemia and Ionizing Radia- Hyg. Toxicol. 21 :331, 1939. 24. Mallory, T. B.; Gall, E. A,, and Brickley, P. R. J. : Radiation Carcinogenesis: W. J. : Chronic Exposure to Benzene (Benzol) : pothesis, Nature (Lond.) 185:135, 111. The Pathologic Results, J. Industr. Hyg. Tnu;cnl 71 - 7 5 5 1979 F