At present only PDF and raw text is avaiable for this file at: http://ratical.org/radiation/CNR/ACCCAo4LoRC1986.pdf http://ratical.org/radiation/CNR/ACCCAo4LoRC1986.txt NOT FOR RELEASE BEFORE SEPTEMBER 9, 1986 R (Pre-publication preprint.) ASSESSING CHERNOBYL'S CANCER CONSEQUENCES:* APPLICATION OF FOUR "LAWS" OF RADIATION CARCINOGENESIS John W. Gofman, M.D., Ph.D. Department of Biophysics and Medical Physics 102 Donner Laboratory University of California at Berkeley Berkeley, California 94720 Presentation at the 192nd National Meeting The American Chemical Society Anaheim, California September 9, 1986 Symposium On Low-Level Radiation Division of Chemical Health and Safety * Country-by-Country Summary Page 39 For contacting author P .0. Box 11207 San Francisco, California 94101 Telephone: 415-664-1933 or 415-776-8299 September 7 -10: Hotel, 714-774-7817, or via Press Office of the American Chemical Society Tel: (714) 740-4529 at the Anaheim Hilton, Monterey Room, Concourse Level Table Of Contents Section (1) THE LAWS OF HUMAN CARCINOGENESISFROM IONIZING RADIATION, p2 Status and Implication s Of the First Law, p2 All Forms Of Cancer (p2); Lin ea r Dose-Risk Relationship (p3); No Protection By Dividing Doses (p4); No Harmless Dose (p4) Implications and Status Of the Second and Third Laws, p4 Implicatio ns For Assessing Chernobyl, p6 Section (2) DISPROOF OF A SAFE "THRESHOLD DOSE" FOR CARCINOGENESIS, p6 Repair-Time and Repair-Domains, p6 Table 1: Millirads Causing Only One Primary Tra ck Per Nucleus, p8 Table 2 : Rads Causing Only One Primary Track Per Repair-Domain, p lO Table 3 : Primary Ionization Tracks Per Domain In 5 Human Studies, pl3 Conclusi on On the "Saf e Threshold", pl3 Implications For the Linear Dose-Risk Model, pl4 Sect io n (3) RECONCILIATIONOF DISPARATE RISK-ESTIMATES, pl5 Corrections Of th e UNSCEARRisk-Estimate, pl5 Corrections Of th e BEIR-3 Risk -Estimate, pl7 Table 4: Observed Rates Of Excess Cancer, A-Bomb Survivors, p20 Correction Of the Radford Ris~-Estimate, p21 Comparison Of N.I.H. and Gofman Est imates, p22 Section (4) THE CANCER-DOSE FOR MIXED AGES, p23 Section (5) THE CHERNOBYL ACCIDENT: CANCER-LEUKEMIA CONSEQUENCES p24 Elements Of the Calculation, p24 Radio-Cesium As the Dominant Menace, p25 Section (6) THE SOURCES OF FALLOUT DATA, p25 Table 5: Percent Of Gamma-Ray Dose Assigned To Cesium-137, p27 Section (7) : METHOD: ILLUSTRATIVE USE OF THE DATA, p28 Me thod 1: Best Type Of Dat a (Denmark), p28 Method 2 : Next Best Type Of Dat a (Poland), p30 Method 3: Last Type Of Data (Italy), p33 Uniform Reduction Of "First-Step" Values , p34 á Cesium-137: Amount Released and Initial Inventory, p35 Section (8) RESULTING ASSESSMENT OF CHERNOBYL'S CANCER CONSEQUENCES, p37 Reality-Check On the Assessment, p37 The Distribution Of Doses Over Time, p38 Table 6: Cancer and Leukemia Tolls From the Chernobyl Accident, p39 The Distribution Of Impact By Age, p40 Section (9) DISCUSSION AND CONCLUSIONS, p40 The Single Most Serious Accident Ever, p40 What We Need and Do Not Need To Assess Chernobyl Accidents, p41 A Prediction, p41 Implications For Medical, Dental, and Occupational Irradiation, p42 Technical Appendix 1: The Basis For Table 1, p43 Technical Appendix 2: The Basis For Table 6, p47 References, p54 1 - Gofman Introduction There exists as large a body of evidence for the human concerning the quantitative aspects of induction of cancer and leukemia by ionizing radiation as exists for any other carcinogen, probably an even larger body of evidence. In 1969, using the human evidence available to that time, Gofman and Tamplin[2] presented three generalizations or "laws" which permit quantitative assessment of the cancer toll which will follow human exposure to ionizing radiation under virtually all circumstances of exposure. It i s gratifying --- as a scientist, but not as a physician --- that all of the human epidemiological evidence which has accumulated since 1969 has provided support for the correctness and usefulness of the three laws. In other words, the laws have been correctly predicting what is being found. The laws apply as well for the incidence of cancer induction by radiation as for the mortality. The validity of the laws made it possible to arrive in 1981 at a single ratio, called the Whole-Body Cancer-Dose for a population of mixed ages, with which ratio the cancer consequences of population exposures could be correctly assessed. The evidence supporting the laws and the step-by-step method of transforming the human evidence into the Whole-Body Cancer-Dose are presented in detail in Chapters 5-10 of Gofman[3]. In the first part of the present paper, the laws, the Whole-Body Cancer-Dose, and the status of challenges to their validity are discussed. Then the idea that cellular repair of DNA and chromosomes may provide some safe threshold dose of ionizing radiation, with respect to human carcinogenesis, is shown to be ruled out by human evidence which already exists. And lastly, specific errors in the reports of the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR)[4,5] and of the N.A.S. Committee on the Biological Effects of Ionizing Radiation (BEIR-3)[6] are identified because those errors would make assessments of Chernobyl's cancer consequences at least 16 to 25 times too low; the disparate risk-estimates are reconciled. In the paper's second part, the Whole-Body Cancer-Dose is applied to the available dose-data from the Chernobyl accident, and the cancer-leukemia consequences are assessed in Table 6 for each of 30 countries; additionally, the methods for handling such data and making such calculations are shown step-by-step so that anyone anywhere can use them when and if a country provides more detailed dose-data. (1) THE LAWS OF HUMAN CARCINOGENESISFROM IONIZING RADIATION Generalization 1: "All forms of cancer, in all probability, can be increased by ionizing radiation, and the correct way to describe the phenomenon is either in terms of the dose required to double the spontaneous mortality rate for each cancer, or alternatively, of the (percent) increase in mortality rate of such cancers per rad of exposure." Generalization 2: "All forms of cancer show closely similar doubling doses and closely similar percentage increases in cancer mortality rate per rad" (at a given age). Generalization 3: "Youthful subjects require less radiation to increase the (cancer) mortality rate by a specified fraction than do adults." Generalization 4, added in 1983 (Gofman, 1983)[7]: "The peak percent increase in cancer rate per rad is reached grossly earlier for such high Linear-Energy-Transfer radiations as alpha-particle irradiation in contrast to the time to reach peak percents for low LET radiation." The fourth law is not discussed or used in this analysis, whose particular application is Chernobyl. While it is clear that the accident caused measurable quantities of plutonium, americium, curium, and other alpha-emitters to fall in Scandinavia and elsewhere, daily data on air concentrations are not available. Since inhalation causes the predominant hazard from such nuclides, we are able to say nothing áexcept that any cancers induced by the transuranics would be additional to those listed in Table 6. STATUS AND IMPLICATIONS OF THE FIRST LAW All Forms of Cancer By now, the human evidence clearly shows that radiation exposure increases the frequency of virtually every major kind of cancer. By major kind, we mean those kinds which, combined, account for about 90% of cancer deaths. Prostate cancer[8] and uterine cancer[9] have recently joined the proven group. By 1982, Dr. Edward Radford, Chairman of the BEIR-3 Committee, said "The point that I feel is important is the consistency with which radiation 3 - Gofman has proved to be carcinogenic in man. It is far and away the most consistent agent that we know of to cause cancer of any type"[10]. In a 1985 report, the National Institutes of Health[11] acknowledged that "it is generally accepted that ionizing radiation may increase the risk of virtually any form of cancer." A Linear Dose-Response Relationship When Generalization 1 speaks of the (percent) increase in mortality rate, it implies that there is only a single percent increase in cancer mortality rate, per rad of exposure, throughout the large dose-range of relevance for humans. A constant risk-per-rad at all dose-levels corresponds to direct proportionality, or the linear dose-response model. As more human evidence accumulates, it keeps confirming that the dose-effect relationship is either linear or quite possibly supralinear, but none of it suggests a quadratic curve with falling risk-per-rad as dose falls. The most recent follow-up data from Hiroshima and Nagasaki again support linearity[12]. Dose-groups in that study range from about 13 rads to 300 rads. Among some 150 types of cancer, only the commonest types like breast cancer and thyroid cancer are providing enough cases to analyze separately. In 1985, the N.I.H. Group[11] acknowledged that the evidence on breast and thyroid cancers "strongly favors linearity", and in 1983, Wakabayashi and co-workers reported that the data on all cancer-types combined among the atomic bomb survivors suggest a linear model, and in a linear-quadratic model, "the linear term is significant whereas the quadratic term is not"[12]. Indeed, the quadratic term was negative. Since the data from Hiroshima-Nagasaki show linearity for all cancer-types considered together, and since the data show linearity for breast and thyroid cancers each analyzed separately, it is obvious that the combined remaining types of cancer must also show direct proportionality between increased cancer risk and radiation dose. In 1970, Stewart[37] had already demonstrated that linearity holds, for diagnostic X-rays, in the region of 0.25 to 1.25 rads. No Protection By Dividing Doses Generalization 1 implies also that the rate of excess cancer per rad of dose is unaffected by the rate of dose-delivery; a single percent increase applies to a rad, whether it is delivered in one instant or over a year. Boice[13] in a careful study of breast-cancer induction in women, concluded: "The observation that multiple low-dose exposures did not produce significantly fewer cancers per unit dose than less highly fractionated 4 - Gofman larger exposures suggests that radiation damage is cumulative and that highly fractionated X-irradiation may be as effective in inducing breast cancer as single or less fractionated exposures." In 1985, the N.I.H. Working Group[11] (p.26) acknowledged this study and others showing that, where human evidence exists on the question of divided doses, it shows no reduced risk per rad when a given dose is divided into smaller doses . No Harmless Dose , For Carcinogenesis Generalization 1 implies that it holds at every dose , all the way down to zero dose. No exception for the lowest dose-range is suggested, and no completely harmless dose is implied. Rather, cancer-risk is proportional to dose, right down to the lowest conceivable dose. In Section 2 of this paper, we shall offer scientific disproof of the notion that a safe "threshold dose" exists with respect to human carcinogenesis. IMPLICATIONS AND STATUS OF THE SECOND LAW One of the important consequences of the second law is that quantitative evidence about radiation-induction of common cancers can be extended to less common cancers for which statistically significant proof is likely to be absent. It is likely to be absent because radiation induces excess cancer in proportion to a cancer's spontaneous rate, as the first law says. An early assault on Generalization 2 was made in 1972 by the BEIR-1 Committee[14], which reached retroactively into a study published in 1965 by Court-Brown and Doll[15]. BEIR-1 raised one of Court-Brown and Doll's crucial dose-estimates by four-fold. In 1980, the BEIR-3 Committee[6] quietly undid this challenge to Generalization 2 by restoring the dose in question approximately to its 1965 value (discussion in Gofman[3], 1981, pp. 327-32). As of today, inspection of the combined evidence from Smith and D0ll[16] , Smith[17], Kato and Schull[18], Schull[19], and Wakabayashi and co-workers[12] provides no reason whatever to think that one type of cancer increases, over its spontaneous rate, by a different percentage per rad than any other type. With respect to the irradiated ankylosing spondylitic patients, Smith[17] reports that "with the exception of tumors of the spinal cord and nerves, the excess mortality from cancers of heavily irradiated sites was increased roughly in proportion to the expected number of cancers of each site based on general population rates." Reporting on the Hiroshima-Nagasaki series, Kato and Schull[18] find the same phenomenon ( p.404): 5 - Gofman "As shown in Figure 1, the 90% confidence limits of the relative risks of breast, stomach, lung, and colon cancers overlap each other, and the relative risk of cancers of all sites (except leukemia) is within the confidence limit, so, statistically it cannot be said that the relative risk differs according to target organ." In Schull's 1984 paper[19] , he confirms that "there is as yet no statistically persuasive evidence that the relative mortality risk differs according to target organs." These statements of Kato and Schull, based upon the single largest body of evidence in existence for human radiation carcinogenesis, provide powerful confirmation of Generalization 2. The reason for some of the early doubt was that for certain cancers, the number of cases available for analysis was very low, and so there were wide confidence limits for the percent increase per rad. With increased follow-up time, the number of cases also increased:, the confirming evidence for Generalization grew steadily, and now very strong statements are coming from Kato and Schull. STATUS AND IMPLICATIONS OF THE THIRD LAW There exists virtually unanimous assent now to the third law. In 1985, the N.I.H. Working Group[11] (p.18) joined the mainstream by saying, "One of the most interesting observations to come out of the Japanese A-bomb survivor studies, which are based on a large population of all ages in 1945, is that the risk of radiation-induced cancer depends strongly on age at exposure." Indeed, we have shown (Gofman[3], 1981, Chapter 9) that, when a population of normally mixed ages is irradiated, about 73% of the radiation-induced cancers will develop in people at or below 20 years of age at exposure. IMPLICATIONS FOR ASSESSING CHERNOBYL Although all three of these laws were regarded in 1969 as outside the mainstream of wisdom, the real-world human evidence accumulated by 1986 has made them the mainstream and has cast the contradictory propositions on the sandy fringes of the riverside. The out-dated propositions must be avoided, if society wants any realism in assessments of Chernobyl's cancer-consequences. 6 - Gofman (2) DISPROOF OF A SAFE "THRESHOLD DOSE" FOR CARCINOGENESIS ¥ THE EXISTENCE OF REPAIR MECHANISMS There has never existed any valid evidence at all for a safe "thre shold dose" of radiation with respect to human carcinogenesis; major flaws in "threshold" studies using doses from natural background and using the very lowest dose-group of 3 A-bomb survivors have been analysed in extenso in Gofman(pp. 227-31, 386-7, 566-72, 672-8 3). But after the Chernobyl accident, the idea has been predictably voiced here and abroad that populations which received its fallout may be protected from any cance r conseque nces by the existence of some safe thr eshold dose. Indeed, some Soviet officials are apparently attempting to deny that the health of anyone outside . U . i . d20 d. h A . d P t he Soviet nion was nJure , accor ing tote ssociate ress. Once upon a time, the safe-dose idea had some plausibility, since r epair mechanisms do exist for at least certain types of damage to chromosomes and genetic molecules, but in the light of accumulatin g human evidence ---both from epidemiology and from cell -studies in vitro ---the safe-dose idea with resp ec t t o human carcinogenesis has become indefensible, as we shall show. ¥ REPAIR-TI ME AND REPAIR-DOMAINS There are cer t ain types of injury to DNA and chromosomes whose repair (and misr epai r) ca n be observed and measured in cell studies. It is observed by workers in the fie ld that it is probably failure to repair DNA, or its misrepair, 21 22 which often causes the chromosome aberrations which endure ' ¥ It is well known by now th at chromosome deletions and tr anslocations are hi ghly associated . h W¡l I 23 . bl 24 k 25 wit i m s tumor , retina astoma , idney cancer ,. and numerous types of 26 blood cancer . It may turn out that certain disturbances in chromosomal 3 material are key to cancers of all types . With r es pect to the question of repair providing a safe threshold dose of radiation, s tudies of human cells in vitro, following X-irradiation, provide some important parameters. (1) Such studies indicate that whatever repair is achieved is compl ete within 21 27 28 6 hours or less af ter irradiation, even at doses of 100 and 200 rads ¥¥ Indeed, almost all the repair occurs within the first 2 hours after irradiation, and by 3 hours, the repair-curve is flat. In Table 3, we have assumed that 8 hours are required for full repair, to be extremely cautious. (2) Such studies confirm what we already know from chromosome studies of 7 -Gofman living humans: not all injuries get repaired, even when cells have infinite rep air -time, and some attempts to repair end in misrepairs. (3) The cell studies suggest that each cell -nucleu s has more than one set or te am of repair enzyme s. The speed at which the early repair is occurring even at doses of 100 and 200 rads per exposure makes it unreasonable to suppose that a single "crash-cart" of repair enzymes is servi cing all the genes and all the chromosomes (see Table 1, Note 2). It is far more likely that each chromosome has at least one repair set of its own, and possibly many. In other words, it seems very unlikely that , with respect to repair , the entire nucleus is a si ngle domain serviced by only one repair team or "crash-cart." When we come to the epidemiological evidence, below, it will be cl ea r how repair-speed and repair-domains are relevant to the indefensibility of a safe threshold dos e. ¥ THE MINIMAL CHALLENGE TO REPAIR MECHANISMS For r epair to provide a safe threshold dose of radiation, below which no radiation-induced cancer occurs, rep ai r of t he relevant injuries must work perfectly bel ow some dose or dose-rate. If human evidenc e shows that r epai r mechanism s are failing to prevent radiation-induced cancer when such mechanisms are faced with the l eas t possi ble challenge, this amounts to pr oof that no safe dose exists with respect to human carcinogenesis . So it is iraportant to explore in detail what dose or dose-rate con s titute s the le as t possible challenge to the repair mechanisms in th e cell nucleu s. ¥ One Primary Ionization Track Per Nucleus By def inition, . a radiation dose of 100 millir ads is simply the dep os ition of 10 ergs of energy per gram of tissue. The deposition is not uniform. The bi ologically import ant characteristics of low LET radiation are that its energy is carried through tissue by high-speed electrons and that the transfers of this energy occur in extremely localized or concentrated fashion, unlike the even diffusion of heat energ y . A dose of 400 rads is equivalent in heat to only 0.001 calorie per gram of tissue enough to provoke the tiny fever of 0.001¡c ---yet 400 rads of ionizing radiation to the whole body will kill about half the humans acutely exposed to it. 8 -Gofman Unlike toxic substances, which can be diluted indefinitely to lower and lower concentration in solution, ionizing radiation cannot be evenly diffused into cells. For low LET radiations such as X-rays, gamma rays, or beta particles, the minimal event is one primary ionization track. Either a cell nucleus experiences such a 29 3 track, or it does not (discussion in Gofmanápp.275 -6 and Gofman , pp.405-7). The minimal possible challenge to the repair mechanism from ionizing radiation is therefore the traversal of the cell nucleus by just one primary ionization track. ¥ DOSE-RATES CORRESPONDING TO ONE PRIMARY TRACK PER NUCLEUS At what dose or dose-rate does every cell-nucleus of an irradiated tissue experience, on the average, only one primary ionization track? Table 1 provides answers, which vary with the energy of the radiation, its linear energy transfer (LET), and the size of irradiated cells. Technical Appendix 1 shows how the answers were obtained, and the basis for 11.4 microns as the appropriate cell-size for most human tissues. Table 1: Millirads Causing Only One Primary Track Per Nucleus In All Cells Of Irradiated Tissue Human Average Cells Cells Cells 10 microns 11.4 microns 15 microns Cesium-137 165 mrads 125 mrads 70 mrads 75 mrads Radium-226 170 mrads 130 mrads 50 KEV X-rays 490 rnrads 375 mrads 215 mrads 40 KEV X-rays 575 mrads 445 mrads 255 mrads 30 KEV X-rays 705 mrads 540 mrads 310 mrads Note 1: At half the given doses, half the cel l nuclei completely escape primary ionization tracks, and for those pa r ticular cells, the dose would be zero. Example: when tissue consisting of 11.4-micron cells receives a dose of 270 millirads from 30 KEV X-rays, half the cellnuclei experience no primary track at all . Note 2: 30 KEV represents the average energy of most medical diagnostic X-rays when peak ki l ovoltage is abou t 90 KEV, and 50 KEV represents the average energy used in cell irradiation studies when the peak kilovoltage is about 150 KEV. A dose of 100 rads from 50 KEV X-rays corresponds (from Table 1) to 100 rads I 0.375 rad per track a 267 primary tracks per nucle us. 9 -Gofman As the dose to an irradi ated tissue falls below the corresp0nding value given in Table 1, the number of injured nuclei falls, but the challenge to the repair system inside those nuclei which are hit does not fall because no cell nucleus ---and no repair mechanism can experience a chal leng e less than one primary ionization track. For 11.4-micron cells irradiated by 30 KEV X-rays, many nuclei receive no hits at all at a dose of 1 millirad, but for the particular nuclei which are hit, the injury experienced at 1 millirad is equal to the injury experienced at 540 millirads. ¥ DOSETO ONE PRIMARY -RATES CORRESPONDING TRACK PER REPAIR-DOMAIN For a given cell -size and type of radiation, Table 1 provides the doses at which the repair mechanism faces the minimal challenge if there is only one repair team or 11crash..cart 11 available to the entire domain of the nucleus . But we find no evidence for only one team, and no evidence that experts in the field consider a single single !'crash-cart" per nucleus to be plausible. If there is more than one "crash-cart" for repair per nucleus, then the minimal challenge for the repair squad is one primary ionization track or hit per repair-domain, not per nucleus. For instance, if there were 10 repair -domain s per nucleus, all the dose-entries in Table 1 would hav e to be rais ed by a factor of 10 in order for each repair -d omain to experience the challenge of 1 primary ionization track. In Table 2, the doses in Table 1 are adjusted to correspond with 23 repair-domains per nucleus (one domain per pair of chromosomes), which seems like a very conservative approach .,when 46 suggests itself as a more likely number, and when the true number may be hundreds or more. Both Table 1 and Table 2 are used for the analysis made in Table 3 . - 10 Gotman Table 2: Rads Causing Only One Primary Track Per Repair-Domain In All Cells Of Irradiated Tissue Basis: 23 repair-domains per cell-nucleus. Human Average Cells Cells Cells 10 microns 11. 4 microns 15 microns Cesium-137 3.8 rads 2.9 rads 1.6 rads Radium-226 3.9 rads 3.0 rads 1. 7 rads 50 KEV X-rays 11. 3 rad s 8.6 rads 4.9 rad s 40 KEV X-rays 13.2 rads 10 . 2 rads 5.9 rad s 30 KEV X-rays 16.2 rads 12.4 rads 7.1 rads ¥ DESCRIPTION OF LOW-DOSE EPIDEMIOLOGICAL EVIDENCE Some values from Tables 1 and 2 can be 'applied to the existing epidemiological evidence of radiation carcinogenesis áat very low doses in humans in order to answer the central question: is "repair" failing to prevent radiati on-in duced cancer even at doses and dose-rates which present the minimal challenge to repair systems ---namely, 1 primary ionization track or hit per repair-domain. There are at least 5 human studies, widely recognized to be well done and "mainstream," which show radiation-induction of cancer at very low dose -rate s. 3 h k d 30, 31 Four of t he f ive studies are b in o man ; t e Baverstoc stu y descried G f of British luminisers appeared later. In three o.f the five studies, the total accumulated dose was hi gh, but the time-interval between small exposures was far more than ample for repairmechanisms to achieve completely whatever protection they can against radiationinduced cancer. Before testing the studies for their dose-rates, we shall briefly describe the nature of each. ¥ The Nova Scotia Fluoroscopy Study 32 Myrden and Hiltz studied 243 women who (in the course of tuberculosis treatment) had chest fluoroscopies with the beam from front _.to back, and with an estimated dose to the breasts of 7.5 rads per fluoroscopy. Time between fluoroscopies was days or weeks, and the total dose accumulated per woman was about 850 rads. Breast cancer was observed at more than six times the expected rate during a limited follow-up period. 11 -Gofman ¥ The Isr~eli Scalp-Irradiation Study 33 Modan and co-workers reported on the excess of thyroid cancer observed among almost 11,000 Isra eli children who received X-radiation for treat ment of tinea ca pitis (ringworm of the scalp). The estimated thyroid dose per child was 7.5 rads total. Thyroid cancer was observed at five times th e expected rate during a limited follow-up period. ¥ The Massachusetts Fluoroscopy Study 34 Boice and Monson also studied women who had received repeated chest fluoroscopy during tuberculosis tr ea t ment; in their series, the beam was usually from back to front , and the estimated breast-dose was 1.5 rads per exam . The accumulated breast dose was about 150 r ads . Among the women whose average age was 20 years at the time of irradiation, breast cancer was observed at more than twic e the expected rate during a limit ed follow-up period. ¥ In -Ut ero Irr adiat i on 35 37 Stewart , 36¥ compared the X-ray history of children who died of cancer or leukemia in childhood with th e X-ray history of matched cont rols who had no malignant disease . She demonstrated that dia gnostic pelvimetry, irradiating the fetus in utero, provoked about a 50% increase in th e frequency of childh ood cancer and leukemia. She was able to demonstrate an excess even from single-film exams, with an estimated dose to th e fetus of 0.25 rad (250 millirads). Her work has been confirmed by several additional analysts and . 38, 39, 40, 41, 42, 43 series ¥ ¥ The British Luminisers 30, 31 In 1981 and 1983, Baverstock and co-workers report ed highly si gnif icant proof of breast-cancer induction by radiation in female workers who applied radium-226 to luminous instruments in Great Britain. Ba,~erstock was able to rule out internal radiation by alpha particles as the cause, and identified external gamma radiation as the source of the radiation injury. The dose-rate of external gamma radiation to the breasts was, by measurement, 500 millirads per week or less. For a 40-hour week, this represents a dose of 12.5 millirads per hour, or 100 millirads per 8 hours (the maximum time consumed by human cells in repairing X-ray injuries). The average dose-rate must have been even lower. Among the women whose average age was 20 years at the time of irradiation, 12 -Gofman breast cancer was observed at twice the expected rate during a limited follow-up period. The total breast dose accumulated by the young women over the work-years was 40 rads. ¥ ANALYSIS OF THE EPIDEMIOLOGICAL EVIDENCE In Table 3, the five studies described above have been tested to ascertain whether or not the excesses of cancer were observed at dose -rates representing the minimal possible challenge to repair-mechanisms in the nucleus. Most of the entries in Table 3 are "1 track," the minimal possible challenge. In other words, existing human evidence already is showing that repair is not able to prevent excess malignancy even at minimal dose-rates. One can quibble about cell-size, uncertain dosimetry in one study or another, and number of repair-domains per nucleus, but even if one invokes the probably preposterous notion of a single "crash-cart," the evidence in Table 3 combines to tell realists that repair provides no safe threshold dose. Repair must work perfectly in order to provide a harmless dose. If it were working perfectly at the minimal dose-rates considered in Table 3, then there would be no cumulative carcinogenic effect to observe from repeated exposures. Repair most probably does a lot of good for human health, but what epidemiolo gists observe here is the residual and unrepaired injury. In the scalp irradiation series, the single "crash -cart" had infinite time to cope with only 14 primary tracks, but it could not. In the Nova Scotia series, the "crash-cart" had far more repair-time between exposures than the few hours in which repair is observed to be completed even in cell studies performed at doses of about 200 rads (about 533 primary tracks per nucleus); yet repair could not cope successfully with only 14 tracks. In the Massachusetts series, the "crash-cart" failed to cope successfully with just 3 primary tracks. When one also considers the evidence from the in-utero and British luminiser series, the story is clear ---if wishful thinking is banished from radiation science. 13 -Gofman Table 3: Primary Ionization Tracks Per Repair-Domain in Five Human Studies Single Repair-Domain 23 Repair -Domains Per Nucleus; Per Nucleus; Divisor from Table 1 Divisor from Table 2 Nova Scotia Flu orosco py 7.5 I 0.540 = 13.88 7. 5 / 12.4 = 0.60 Dose = 7.5 rads per exposure 14 tracks 1 track Energy= 30 KEV X-rays (Some domains having 0) Isra e li Scalp Irradiation 7.5 I 0.540 = 13.88 7.5 I 12.4 = 0.60 Dose= 7.5 rads, one time 14 tracks 1 track Energy= 30 KEV X-rays (Some domains having O) Massachusetts Fluoroscopy 1.5 I 0.540 = 2. 77 1.5 I 12.4 0.12 Dose = 1.5 r ads per exposure 3 tracks 1 track Ener gy = 30 KEV X-rays (Most domains havin g 0) In -Utero Series 0.250 I 0 . 540 = 0.46 0.250 I 12.4 = 0.02 Dose= 0.250 rad per single -fil m exam 1 track 1 track Energy= 30 KEV X-r ays (Many nuclei having O) (Most domains having 0) British Lumini se r s 0 . 100 I o.130 = o. 77 0.100 I 3.0 = 0.03 Dose= 100 millirads per 8 hours 1 track 1 tr ack Energy= Radium-226 (Some nuclei havi ng 0) (Most domains having O) ¥ CONCLUSION LD" ON THE "SAFE THRESHO The expression of radiation-indu ced cancer at minimal dose -rates shows that it is scientifically unr easonab l e --even ludicrous --to entertain hope of finding a "safe threshold" based upon "repair." Any repair-process which fails to prevent radiation -induced cancer at minimal delivery-rates of ionizing radiation i s inherently unable to provide a safe or harmless dose. The combination of physics, epidemiology, and cell studies in vitro, provides overwhelming evidence that a safe "threshold dose" of ionizing radiation simply does not exist for human carcinogenesis. In 1980, the BEIR-3 Committee 6 declared that, from epidemio l ogic science alone, "the existence or non-existence of a threshold dose is practically impossible to determine" because the size of the human series 14 -Gofman required to achieve statistical significance would be huge if the threshold were very low. However, the determination is not impossible if the already existing human evidence from epidemiology is combined with the physics of ionizing radiation. The non-existence of a safe "threshold dose" for human carcinogenesis has now been determ ined. ¥ IMPLICATIONS CURVE FOR THE LINEAR DOSE-RESPONSE Human epidemiological evidence supports the linear dose-response relationship (see Section 1 of this paper). Table 2 shows why the relation shi p must remain linear below a dose of 12.4 rads (for 30 KEV X-r ays). At doses l ower than 12.4 rads, the fraction of repair-domains which experience any traversal decreases in direct proporti on to dose, while th os e domain s which do experie nce traversal continue to experien ce the same minimal dose -rate they experienced at 12.4 rads, namel y 1 primary i oni zation tra ck per r epa ir-domain. The combination of the existing human ev idenc e and the information in Table 2 demonstrates th at linearity cannot possibly exagg e rate the carcinogenic effect of radiati on at very low doses . 15 -Gofman (3) RECONCILIATION OF DISPARATE RISK-ESTIMATES Among the ways to express cancer -risk from radiation exposure are percent increase over the spontaneous rate per rad, doubling dose ( the dose in person-rads which doubles the spontaneous rate), ca ncer-dose ( the dose in person-rads which corresponds to 100 % chance of 1 fatal radiation-induced cancer in the population), and number of radiation-induced cancer deaths per million person-rads. Person-rads are simply th e product of a dose and the number of people who receive it. One hundred person-rads could be the sum of 60 people each receiving 1.5 rads ( 90 person-rads) plus 2 people each receiving 5 rads ( 10 person-rads). ¥ CORRECTIONS OF THE UNSCEAR RISK-ESTIMATE In 1977 and agai n in 19 82, the United Nations Scientific Conunitte e on th e Effects of Atomic Radiation (UNSCEAR)4 ¥ 5 asserted that the cancer-dose for a population of mixed ages is 10,000 person -rads per fatal radiation-induced cancer. ¥ Er ror In the Leukemia -Dose UNSCEAR's value for cancer-dose starts with its mistaken value for leukemia of 1 case per 50,000 person -rads. UNSCEAR, usin g data from the Japanese A-bomb 4 survivors as followed up through 1972 (UNSCEAR, Appendix G, Table 4, p.372), ar ri ved at the wrong value by inv oking an especially large carcinoge n ic effect for neutrons, allegedly present in the Hiroshima dose. By 1982, it was revealed that neutrons were virtually absent 63 ¥ In the meantime, a much larger set of leukemia data fr om the A-bomb survivors was published in 1978 by Beebe, Kato, and 44Land . It was clear from that data, dan 1a so f rom ht e b reast-cancer data .int he Ab b-om .series 64 , htat neutrons --i áf present ---were not having an especially large carcinogenic effect. Therefore, we 44 considered the leukemia data (the Life Span Study through 1974), and analysed it without invoking any "neutron effect . " These data yield a Leukemia-Dose not of 50,000 3 but rather of 10,000 person-rads per case (Gofman). If UNSCEAR had simply analysed its more limited set of leukemia data (through 1972) for both Hiroshima and Nagasaki without invoking any "neutron effect" (since the effect was imaginary), it would have arrived at a Leukemia-Dose of 11,500 person-rads per case, a value remarkably close to the value of 10,000 person-rads derived from the longer follow-up period. There is no doubt that UNSCEAR's leukemia-dose is wrong by a factor of 5. 16 -Gofman ¥ Error In the Cancer-Leukemia Ratio UNSCEAR's second error occurred when it asserted that the ultimate ratio of fatal cancers to leukemia in an exposed population would be 5. In other words, for every radiation-induced leukemia, the population would show 5 fatal radiationinduced cancers. This assertion is the basis for UNSCEAR's setting its cancer -d ose at 10,000 person-rads per fatal case, compared with its leukemia-dose of 50,000 person-rads per case . But UNSCEAR's ratio of S failed to take proper account of what everyone now recognizes: the r ate of radiation-induced leukemia passes its peak 7.5 years after radiation exposure, but at 7.5 years post-irradiation the rate of radiation-induced cancer is just beginning its climb. The delay in the appearance of the solid tumors has been observed in numerous human studies of low LET exposure, not only in the A-bomb study . By 1977, the radiation -indu ced leukemia cases among the A-bomb survivors had peaked, declined, and finished, whereas the rate of r adiation -indu ced cancers was . 12 18 44 45 still rising with no end in sig ht. Follow-up studies published later ' ' ' show that new radiation-induced cancer cases are continuing to appear, and very markedly among those who were young at the time of exposure. The tru e ratio of cancer to leukemia keeps growing. The leukemia-dose in the A-bo mb series is fixed and verifiable because the leukemia "story" there is over. A cancer-dose based exclusively on that one series would be less firm because the cancer " story " there won't be over for another 30 years or so . It would be a serious mistake to base the human cancer-dose on a single series when there are many additiona l studies of exposed humans besides the 3 A-bomb survivors. When th e worldwide human evidence is considered (Gofman), the data show th at the leukemia dose is likely to be 6,500 person-rads per case compared with 268 person-rads per fatal cancer , which reflects a ratio of 24.3, not 5. Because no lifespan follow-up study has been completed anywhere, the ratio of 24.3 necessarily reflects some projections based upon the three "laws" (Section 1), which have already proven repeatedly to predict correctly what the next human evidence will show. By contrast, there is no scientific basis for UNSCEAR's assertion that 5 should be regard ed as the ultimate and universal value for the cancer-to-leukemia ratio. Therefore it is fair to say that UNSCEAR's cancer-to-leukemia ratio is wrong by a factor of (24.3 I 5), or 4.9. 17 -Gofman ¥ Reconciliation Of the Disparit y If we take b oth of UNSCEAR's e rr ors into account, the UNSCEAR value for cancer -d ose is off by the product of 5 and 4 . 9, or 24 . 5. The co rre cted UNSCEAR valu e for a mixed-age population would be (10,000 person-rads pe r fata l cancer) divided by (24.5) 408 person-rads per fatal cancer . This is very close to 268 person-rads per fatal cancer, the value from Gofman used in this paper to analyse th e consequences fr om the Chernobyl accident. It is natural if -scientis t s disagre e about the exact size of UNSCEAR's errors, but it would be unthinkable for responsible member s of the scientific community to use UNSCEAR's obviously mistaken cance r-dose as it is, simply bec ause it carries UNSCEAR's aut hority. ¥ CORRECTIONS OF THE BEIR-3 RISK-ESTIMATE It is well known that members of the Committee on the Biological Effects of 6 Ionizing Radiation (BEIR-3) were split over the risk -values in their final report, which included a vigorous dissenting appendix by th e BEIR-3 Chairman, Dr . Edward P. Radford. After the recall of th e Connnittee ' s 1979 report by the National Academy of Scie nces and its appoi ntment of a special committee to arra nge a compromise, final values emerged in 1980 as if from a black box; no way was provided for the read er to check , follow, or replicate the analysis. However, what we can do is to show, quantitatively, how the BEIR-3 values of percent i ncrease (over the spontaneouás rate of cance r) per rad of exposure ar e about 16 tim es lower than the human evidence req uir es . 18 -Gofman ¥ Erroneous Incorporation of Zero -Values Because the BEIR-3 Committee did not use Generalization 2, it reviewed the data for each typ e of cancer separately. It relied almost exclusively on the Hiroshima-Nagasaki data. Obviously, if one subdivides a set of evidence enough times, eventually the data in many of its se ctions will become statisti cally inconclusive. After the Committee subdivided the existing data by age at irradi ation, by sex, and by site of cancer, it had categories with excess cancer -rates which were not provably significant, and it treated all those categories as zero excess-cancers-per -rad. These zeroes were especially important among A-bomb survi vo rs who were between ages O -9 years in 1945 . Although many studies of childhood exposure have shown that young children are far more sensitive than adults to radiation carcinogenesis 3 (d at a in Gofman, 1981),the first law correctly states that the sensitivity is most fully manifest as the y reach the adult years when their spontaneous cancer rate rises. Therefore, when a population of mixed ages is irradiated, the ~ungest children are the slowest in showing the full consequen ces. Ignorin g this, the BEIR-3 Committee put three shocking entries in its Table V-14, "Estim a ted Excess Cancer Incidence ( Excluding Leuk emia and Bone Cancer) per Million Persons per Year per Rad, 11-30 Years after Exposure, by Site, Sex, and Age at 6 Exposure 11 For males O -9 yea rs old at exposure: Lung cancer 0.00 For females O -9 years old at exposure: Lung cancer 0.00 Breast cancer 0.00 It is nothing short of fantastic to use risk values of zero for two of the most important cancers when the most sensitive age-group is irradiated. Such entries mistakenly and serious l y lower BEIR' s risk-value for a population of mixed ages. Such a blunder is particular l y surprising when the O -9 19 -Gofman year age-group of A-bomb survivors was already showing, with other cancer sites, a risk-rate 5 to 10 times higher than adults (see Table 4). Two other cancers of major importance were prematurely assigned zero values at every age by the BEIR-3 Committee: prostate and uterine cancers. 12 The mistakes were soon evident to all. In 1983, Wakabayashi reported that the excess of prostate cancer in the Nagasaki bomb-survivors was consistent with the excess observed for most of the common cancers. In 1984, 9 Wagoner , reported on the significant excess of uterine cancer observed among women irradiated in the pelvis. It is hardly surprising, after BEIR-3 made so many major. -omissions, that it s risk-values are seriously at variance with reality. If it had used Generalization 2, it would have avoided those errors. Because of our confidence in the three laws, we never corrupted our riskestimates with obviously false zeroes. Instead, we presumed that the "missing" cancers would ultimately display the same behavior as all the other types. And ever since 1969, additional evidence has repeatedly validated the presumption. More recently, we predicted in 1981 (Gofman 3 , 1981, p.260) that female A-bomb survivors who were O -9 years old at irradiation "will demonstrate a startling number of breast cancers induced by radiation .¥ " 45 The very next year, Tokunaga and co-workers demonstrated that additional follow-up of the A-bomb survivors showed, among the females O -9 years old at the time of bombin g , that the relative rate of radiation-induced breast cancer already exc ee ds the rate for any other age-group. Meanwhile, the BEIR-3 Committee had built into its risk-analysis an underestimate of gross proportions for the most important single form of cancer among women. ¥ Consequent Percent-Increase-per-Rad Values As a result of BEIR's false zero values and its erroneous use of a quadratic 6 term, it arrived at the following values for risk (Table V-19, BEIR-3 : Report ) ¥ For males: 0. 11 percent increase per rad. For females: 0. 15 percent increase per rad. Average: 0. 13 percent increase per rad. Since the BEIR-3 Committee relied almost exclusively on the HiroshimaNagasaki study, we will compare its value of 0. 13 % with the actual evidence 20 -Gofman 44 for the period 1950 through 1974, calculated from Beebe, Kato, and Land in Gofman 3 , Chapter 6, and presented in Table 4 below. Table 4: Observed Rates of Excess Cancer Among A-Bomb Survivorsi Through 1974 (1) (2) (3) (4) Age Group Person-Years Crude Percent Increase Product (2) X (3) per Rad 0 to 9 362,100 3.26 % 1,180,446 10 to 19 398 ,500 1.42 % 565 ,870 20 to 34 363,100 0.63 % 228,753 35 to 49 386,300 0.33 % 127,479 so+ 177, 700 0.0 % 0 Sum 1, 510,200 Sum 2,102,548 Table 4 leads to the following value: Weighted Percent In crease per Rad= 2,102,548 I 1,"510,200 = 1.39 %. When thi s value is compared with BEIR's, BEIR is clearly too low by a factor of L39 "/ 0 . 13 = 10. 7 times. But th at is not all. ¥ Erroneous Values From Inadequate Follow-Up Time The cr ud e percent increase per rad of 1.39 % is itself too low by at least 50 %, because it i s ba sed on the fifth throu gh thirtieth years following exposure; the percent is seriously lower than one bas ed on the tenth through fiftieth years following exposure be ca use (a) it is decreased by values f rom the 5th-10th years after exposure when the ratio of irradiated cases to unirr ad iated control cases (ratio of observed/expected) is 1.0 or just be gi nnin g to diver ge grad uall y from unity, and (b) the irradiated children have not even r eached ages where their special sensitivity adds hugely to a population's can ce r toll. If 1.39 % is increased by a reali st ic 50 %,the value becomes 2.085 % per rad. When 2.085 % is compared with the BEIR's O. 13 %, BEIR is clearly too low by a factor of 2.085 I 0. 13 = 16 times. It is regrettable that some analysts insist on using inadequate follow-up periods as a basis for pub l ishing inordinately low cancer rates per rad. And there is no reason for taking such analyses seriously. BEIR's error of 16-fold accounts for the disparity between its cancer-dose of 4,425 person-rads (Table V-4)6 and Gofman's 268 person-rads3. 21 -Gofman ¥ Reconciliation of the 16 -Fold Disparity If the BEIR Connnittee had done its work correctly, its percent increase per rad would have been in the neighborhood of 2.085 %. And how would that have compared with the value presented by Gofman in 1981 for the Whole-Body Cancer-Dose? It would have been in complete harmony. Our value of 268 person-rads per cancer fatality is derived from the male Cancer-Dose for a mixed age population of 235.0 person-rads, and the female Cancer-Dose of 300.2 person-rads (Gofman , 1981, p.294). These, in turn correspond with a male doubling dose of 43.5 person -rads and a f emal e doubling dose of 49.5 person-rads, for populations of mixed ages. And sin ce these values produce a 100 % in crease in cancer rate over its spontaneous rate, i t f ollows that: For males, per cent increas e per person-rad 100 % I 43.5 person-rads, 2.3 % per person-rad. For female s , percent increase per person-rad= 100 % / 49.5 person-rads, 2.02 % per person-rad. Avera ge for males and females 2.16 % per person -r ad. The differ ence between 2.16 % and 2.085 % is utterly negligible. The way to reconcile the 16 -fold disp arit y is no myster y. ¥ CORRECTIO N OF THE RADFORD RISK-ESTIMATE The Chairman of BEIR-3, Edward P. Radford, dis avowed the introduction of a lin ear -qu adr ati c model for risk, and stat ed in his 1980 diss ent that the human evi dence support s the lin ear dose-effect model . Radford' s own analysis of percent 6 in crea se in cancer rat e per rad was ( from hi s Table 1, BEIR-3 Report ) For male s : 0.31 % per rad. For females 0.67 % per rad. Avera ge 0.465 % per rad. The r a tio of the Gofman and Radford estimates was therefore 2 . 16 % I 0.465 %, I whi ch is 4.64 5, in 1980. The ori gin of Radford ' s underestimate is clear (p.303, BEIR-3 Report ): he, .too, mistakenly assigned risk-values of zero to several important cancers. In addition, he failed to correct evidence from a limited-follow-up period so that it could properly represent what it purported to represent: total excess cancer mortality in a population following its irradiation. 47 . k 1 Subsequently, Radford has raised his ris -va ue somewhat. It is presently only 3.7-fold lower th an ours. ¥ COMPARISONWITH THE N. I.H. ESTIMATE The 1985 report of the N.I.H. Working Group virtually adopted BEIR-3's risk values without challenge. However, the N.I.H. Report made three not ab le 11 exceptions First, it explicitly objected (p.242) to BEIR's use of zero to describe breast cance r risk in females exposed between O -9 years of age. In a classic understatement, it said, "The two studies make it plain that the BEIR coeffi cient of zero for women exposed under 10 years of age is inappropriate and sho uld be replaced." Secondly, on breast cancer, it conceded (p.55) that the human evidence demands the use of th e lin ear model rather than th e linear-quadratic model. Thirdly, it made th e same concession about thyroid cancer ( p.55). ¥ Small Remaining Disparity on Breast and Thyroid Cancers How, then, does N.I.H. ' s evaluation of breast-cancer risk compare with our own? To compare risk-values for exposure of 1 rad to the breasts at age 20, we can work from the respective PC (probability of causation) values for a breast cancer appearing at age 44, as an illustrative case. PC= RI (S + R), where R = the share of causation con tributed by th e radiation exposure and S represents the "spontaneous" share contributed by other causes. The N.I .H. Breast Value: N.I.H . 11 (p.244-5) gives R = 0.00606 for such an illustration, where the spontaneous share (normalized)= 1.0 Therefore, PC= 0.00606 I l.00606, or a PC value of 0.602 %. The Gofman Breast Value: (1981,Table 56b)3gives R = 2.01 , where the spontaneous share= 93.35 for the same illustrative case. Therefore, PC= 2 .01/(9 3.35 + 2.01), or a PC value of 2.1 %. The Breast Ratio: The ratio of Gofman to N.I.H. for breast cancer risk per rad is 2.1 I 0.602 = 3.5. As we shall see for th e oth er cancer, it i s the N.I .H. committee which has the higher va lu e. A coMparable illustration can be exp1ored for thyroid cancer. Consider a female, exposed at age ten to 5 thyroid-rads, who shows a thyroid cancer at age 25. The N.I.H. Thyroid Value: N.I.H.(p.257) shows R = 0.297 where spontaneous is 1.0. Therefore PC= 0.297 I 1.297 = 0.23, or a PC of 23.0 %. 23 -Gofman The Gofman Thyroid Value : The Gofman tables for thi s illustration show R 0.22 for 1 thyroid-rad where spontaneous= 8.00. When R is multiplied by 5 for a dose of 5 thyroid-rads, PC= 1.1 /(8.00 + 1.1) = 0.121, or PC= 12.1 %. The Thyroid Ratio: For this cancer, the N.I.H. risk va lue is higher than our own by a factor of 23.0 / 12.1 = 1.9-fold. ¥ SUMMARY ON RECONCILIATION For the two types of cancer which provide enough data for separate analysis, we and the N.I.H. Conunittee arrived at a "wash", with virtually the same risk values. In one case, the N. I.H. is 3.5-fold lower and in the other case, 1.9-fold higher. It is clear where the mainstream is flowing. (4) THE CANCER -ilOSE FOR MIXED AGES The Whole-Bod y Cancer-Dose is a ratio: a whole-body dose in person-rads per 3 one fatal radiation-induced cancer. Gofman has demonstrated extensively how the Whole-Body Cancer-Dose is derived for each sex and at various ages from the existing human evidence. As Generalization 3 indicates, the Cancer-Dose is far lower for children tha n for adults, and is very high for adults age 50 and 3 older. The range (Gofman, 1981, Tables 21 and 22) is from about 65 whole-body per son -r ads for newborn s to about 20,000 whol e-b ody person -r ads at age 55. Ri sk Possibly Unde r es timated: Up to th e 40th yea r following expos ur e to low LET radiation, there is evi dence that the observed I expected ratio for solid cancers is in cre as in g3¥ 9¥ 3o, 61 There are as ye t no studie s with follow-up for lon ger peri ods. But we have da t a for leukemia, whe re th e 0/E ratio peak s about 7.5 years followin g a single exposure and th en dec lin es . Usin g thi s dec lin e as a model, we have included in the Whole-Body Cancer-Dose th e ass umption that the 0/E ratio for solid cancer s from low LET expos ure s will also begin declinin g after it peaks. This assumption may underestimate the risk from such exposures. ¥ THE CANCER-DOSE FOR A MIXTURE OF AGES Following a situation like the Chernobyl accident, the world is interested in assessing the overall excess number of cancer fatalities and non-fatal cancer cases for populatio ns in which all ages are present. A Cancer-Dose for mixed ages is available (Gofman 3 ),We simply took account of the distribution of persons by age and sex in a population ( the U.S. population 24 -Gofman was used) and then weighted the Whole-Body Cancer-Dose for each age by the fraction of the population at that age. In a population of constant size, the distribution also remains virtually constant. The error introduced by thi s approximation is trivial except in a population with age distributions 3 grossly and permanently different from the U.S. ( details in Gofman ,19 81). The result obtained f rom such weighting gave the following Whole-Body Cancer3 Dose for a population of mixed ages (both sexes included) (Gofman,1981,p.294): 268 whole-body person-rads per fatal cancer, or 268,000 whole-bod y person-millirads per fatal cancer. ¥ LEUKEMIA-DOSEFOR A POPULATION There is unanimous agreement on treatin g leukemia separately from the solid cancers, because leukemia behaves differently with respect to speed of appearance, duration of radiation effect, loss of life-expectancy, and the absence of a á definitive age-trend with risk-per-rad. The Whole-Body Leukemia-Dose for a mixed aged population is estimated at 6,500 person-rads, or 6,500,000 person-millirads 46 per leukemia case (Gofman and O'Connor, 1985) . (5) THE CHERNOBYL CONSEQUENCES ACCIDENT: CANCER-LEUKEMIA ¥ ELEMENTSOF THE CALCULATION If in each co untr y, we knew the whole-bod y radiation dose received by each person and if we added up those doses for the entire population , we would know th e total person-rads or person-millirads received i n that country. For illus~ 8 tration, suppose we had a country with 10person-millirads of whole-body dose. Then, 8 Fatal cancers 10person-millirads I 268,000 person-millirads per cancer 373 fatal cancers from the Chernobyl accident. Since there will be approximately one non-fatal cancer induced by radiati on for each fatal cancer, there would be an additional 373 non-fatal cancers. Leukemia cases 108 person-millirads I 6.5 x 106 person-millirads per leukemia 15 leukemia cases. ¥ SOURCES OF DOSE FROM A NUCLEAR POWER ACCIDENT There are several sources of exposure in areas where fallout occurs: (1) Direct radiation of the whole body from gannna rays in the cloud of radionuclides passing over a population; 25 -Gofman (2) Inhalation of radioactive substances from the passing cloud and from radionuclides re-suspended after deposition on the gro und ; (3) Direct external radiation of the whole body by gamma rays emi tt ed from r adionuclides deposited on the ground; (4) Int ernal exposure by ingestion of radionuclides with milk, water, meat, fruits, and vegetables. One or another source of exposure will dominate, according to the type of accident and where a population is located with respect to the event. ¥ RADIOCESIUM AS THE DOMINANT MENACE In most countries receiving fallout from the Chernobyl accident, it is clear that the maj or doses will come from ganuna rays emitted from radionuclides deposited on the ground, and from internal radiation via food and water. Fallout measurements show that a large quantity of radioactive cesium did come out of the Chernobyl reactor. The dose received from cesium-137 (T1 =30.2 years) ~ and cesium -1 34 (T1 = 2.3 years) will be the most important part of the whole-body ~ exposures. Of course, we do not deny additional doses from other nuclides. Even without the incremental dose th ey inflict, we can rea ch a good appreciation of Chernobyl's cance r and leukemia conseq uences if we are able to calculate the doses delivered by the cesium-137 and ces ium-13 4 , both from direct gamma radiation from the ground and from these nuclid es in the food chain. (6) THE SOURCES OF FALLOUT DATA There are two major sources of multi-nation information available. The first 48 i s a series of reports from the World Health Organization (WH0)in Copenhagen, and the second is a series of reports from the United States Environmental Protection Agency (EPA) 49 in Washington, D.C. The organizations issued their last reports, respectively, on June 12 and June 30, 1986. These reports rely upon measurements provided by the various reporting countries . Some cot.mtries reported data rather professionally. Others, such as East Germany, reported none at all. The Soviet Union, in spite of its assurances of being forthcoming with data, provided no cesium-137 or cesium-134 measurements 65 at all, until the Soviet report in late August ¥ In addition to the WHO and EPA reports, there are reports for some single 50 51 52 countries, most particularly Finland ' and the United Kingdom ¥ 26 -Gofman We will de sc ribe below th e kinds of measurem ents avai lable, and their use in estimating doses from cesiurn-137 and cesiurn-134. In this paper's Technical Appendix, dose data are described country-by-country. ¥ Opportunity fo r Future Measurements We can state at the outset that all the requisite measurements for a pe r fect assessmen t of Chernobyl's cancer consequen ces are far fr om ava ilabl e. To match existing measurements with exact population distribution would require a grid over each country with measurements of both the populat ion within a particu lar grid -lo cation and the cesiurn-137 and cesium-134 deposition in that same grid-location. It is regrettable that socie t y is not se t up to provide such informa ti on on a tim ely basis . However, i f th e will '.exists t o obtain such data, the opport unity has not been lost (see Section 8). ¥ THREE KINDS OF AVAILABLE DATA AND THEIR HANDLING ¥ (1) The Best Type of Data Here th e co untr y repo rts the int eg rated dep os iti on of cesium-137 and ces ium-1 34 up thro ugh the entir e period of si gnificant fallout. Among the reports, this occurs relatively r arel y . Some countries provide the deposition val ues fo r a very li mited period of th e fallout, so that the true total deposition must be higher than th e values reported. ¥ (2) The Next Best Type Of Data Here t he country reports the values for gannna-r ay exp os ure f rom the depositi on of al l radio nucl ides on the ground for a specified date following th e accident . These data can be used effe ctivel y to obtain indirectly what the cesium -1 37 and cesium -13 4 depositio ns were at the same l oca t ions . The basis for su ch conversion fr om external gamma-r ay dos e to ces iu m values resides in the provision by th e Finnish Centre f or Rad iation and Nuclear Safety of valu es for the perce nt of the t otal gamma-ray dose whi ch is t o be assigned to 5á cesium -1 37. In the Fi nns' fir st report O, they provide the datum that 1.8 % (1.7% -1.9%) of th e total gamma-ray dose for April 29, 1986 is to be assigned to gamma-rays from cesium-137's decay (vi a barium-137m). In their second report the y provide the datum that 11% of the total gamma-ray dose for May 6-7, 1986 is to be ass igned to gamma-rays from cesium-137. By using the daily decay curve of of gamma-ray dose for Uusikaupunki for the first two weeks, it is possible to interp ola te and extrapolate the percent of the gamma-dose to be assigned t o 27 -Gofman cesium -1 37 for dates others than those for which data a re provided directly. These assignments are listed in Table 5. Table 5: Per ce nt Of Gamma-Ray Dose Assigned To Cesium-137 Gannna-Rays Date Of Gamma-Ray Percent Of Measured Dose Measurement Assigned To Cesium-137 April 29,19 86 1.8% April 30,1986 4.1% May 1,1986 4.9% May 2,1986 5.8% May 3,1986 6 .8 % May 4, 1986 8.0% May 5,1986 9 .1 % May 6,1986 10 .2 % May 7,1986 11. 7% May 8,1986 12.9% May 9,1986 14.4% May 10,1986 16.0% May 11, 1986 17.3 % May 12,1986 18 . 8% The rea son fo r the rising percent of the gamma-ray dose assigned to cesium-137 is that the cesium-137 hardly changes its output of gamma-rays durin g this brief period of about two weeks, whereas many of the short -li ved nu c lid es are decreasing their output due to s ubstantial decay during the same time period . ¥ (3) The Last Type Of Data Here we ar e not provided either with gamma-r ay dose measurements or with radiocesium deposition measurements, but we are provided with iodine-131 deposition measurements. From ana l ysing other data where both I-131 and Cs-137 deposition data are available, we are able t o estimate Cs-137 deposition indirectly from I-131 deposition data. 28 -Gofman (7) METHOD: ILLUSTRATIVE USE OF THE DATA ¥ METHOD1: BEST TYPE OF DATA (DENMARK) Cesium-137 integrated deposition is ava i lab l e. Denmark did provide such data. 4 Denmark pr ovided (WHO June 5, 1986 Report) ~he fol l owing data ( from 10 stations for May 15 through May 27) for countrywide contamination in Becquerels/m 2 of surface soil: Mean S.D. Max Cs-137 1075 758 2943 Cs-134 602 424 3477 These reports note that th e above values are corrected with respect t o cesium -137 still present from weapons-test fallout . 22 For Cs-137: (1075 Bq/m) x (27 Picocuries/Bq) 29025 pCi /m. 22 For Cs-134: (602 Bq/m) x (27 Picocuries/Bq) 16254 pCi /m. ¥ Cesium-137 Dose, Method 1 4 From data for worldwide fallout from weapons testing, described by UNSCEAR and sunnnarized in Gofmaná3(1981, p.548), we calculate that the total absorbed dose 2 commitment is 0.66 millirads for each 1000 pCi/m . This includes th e dose commitment both from external radiation from Cs-137 gamma rays coming fr om the gro und, adjusted by UNSCEAR for weather i ng to an average depth of 3 cm, for body-shielding, and for time spent indoors, and from internal radiation from Cs-137 ingested via th e food chain. Internal doses vary by soil typ e , and here we are using aver age values observed f r om weapons fa ll out. Unfortunately for people in the Ukraine, 4 UNSCEARestimates that a much larger internal dose will be received from cesium there than in most areas, due to special soil characteristics there. But fo r average conditions, of the 0.66 millirads total dose commitment from each 1000 pCi/m 2 , UNSCEAR's estimate is that 70 % is from external dose áand 30 % from internal dose. For Denmark, therefore, we can make the following calculation of dose. External Cs-137 Dose Connnitment External dose= (total depositio n) x (dose per unit of deposition) x 22 (external share)= (29025 pCi/m ) x (0 .66 mrads per 1000 pCi/m ) x (0.70) =13 . 4 millirads. 29 -Gofman Internal Cs-1 37 Dose Commitment The inte rn al sha re changes to 30 %, and therefore internal dose 2 (29025 pCi/m ) x (0.66 mrads per 1000 pCi/m 2) x (0.30) = 5.7 millirads. ¥ Cesium-134 Dose, Method 1 Exte rn al Cs-134 Dose Commitment : There are two factors t o consider in 2 evaluating the dose from Cs-134 for the same number of picocuries/m as for Cs-1 37. (a) The total average ganuna-r ay energy per disintegration for each nuclid e. Ratio of ganuna-ray energy Cs-134 I Cs-137 = 2.52 {b) The mean li fe of Cs-134 atoms versus Cs-1 37 a t oms (mean li fe half l ife I 0.693). Ratio of mean life Cs-1 34 / Cs-137 = 0.076 Rela tive Dose -Effectiveness : The relative dose-effectiveness of Cs-134 2 versus Cs-137 per pCi/m depo s ition is the product of factors (a) and (b). Dose-effectiveness Cs-134 I Cs-1 37 = (2.52) x (0.076) = 0.19. Calcula tion of Dose: The external dose connnitment from Cs-134 is : (Relative Deposi tion Cs-13 4/Cs -137) x (Dose -Effectiveness Factor) x (Exte rn al Cs-1 37 Dose) . For Denmar k, we have therefore Cs-134 external dose= (16254 pCi per square meter I 29025 pCi per square meter) x (0.19) x (13.4 millirads) = 1.4 millirads . Interna l Cs-1 34 Dose Commitment : There are two factors to consider. (a) The averag e peak beta energy of Cs-13 4 versus that of Cs-137; the predomi nant source of internal dosag e is from disintegrations via beta particles . Ratio of beta -particle energy Cs-134 I Cs-137 = 0.65 I 0.51 = 1.275. (Note that some handbook s give 1.17 MEV as th e Cs-137 beta ene r gy . This is true for only 8 % of the disintegrations; 92 % go via the 0.51 MEV disintegration pathway.) (b) The ratio of mean-life, calculated above to be 0.076. Relative Dose -Effectiveness: This factor is th e product of (a) and (b). Dose-effect i veness Cs-134 / Cs-137 = (1.275) x (0.076) = 0.097 Calculation of Dose: The i nt er nal dos e connnitment from Cs-134 is (Relative Deposition of Cs-13 4 I Cs-137) x (Dose-Effectiveness Factor) x (Internal Cs-137 Dose). For Denmark we have, therefore, Cs-134 internal dos e= (16254 pCi per 2 2 m I 29025 pCi per m ) x (0.097) x (5.7 millirads) = 0.3 millirads. 30 -Gofman ¥ Combined Cs-137 and Cs-134 Doses, Method 1 The combined external and internal doses from both nuclides, in Denmark= (1 3 .4 + 5.7 + 1.4 + 0.3) millirads = 20.8 millirads. ¥ Cancer and Leukemia Consequen ces, Denmark (Population Size) x (Dose in millirads) Fata l Cancers=_,__..~~~~~~--'-~---''--~ ~~ ~ ~~~ ~-'--~ ( 268,000 person -millirads per fatal cancer) 6 (5 .1 x 10person s ) x (20.8 millira ds) (268,000 person-millirads per fatal cancer) = 396 fa tal cancers, which we round off to 400, in Denmark . Non-fatal Cance rs, additional= 400 cases. ~~~~~~~~~ (5 .1 x 106 persons) x (20.8 millirads) Leukemias (6,500,000 person-millirads per l eukemia) 16.3 leukemias, rounded off to 16 cases . Thi s completes the analysis for Denmark, bas ed upon what we are ca ll ing the best type of data , namely, integrated Cs-137 and Cs-134 deposition on the ground , averaged ove r the countr y . ¥ METHOD2: NEXT BEST TYPE OF DATA (POLAND) Gamma-ray exposure from deposition of all radi onuclides on the ground is 48 provided. Poland reports data usable for this illustr ative example. The WHoreport of May 30, 1986, provides gamma-ra y exposure from the ground for "all Poland" for th e very early period, April 29, 1986, which is ideal since the 1.8 % factor for the contributi on by cesium-137 applies correctly. Although we would much prefer to have separate gamma-ray measurements and populati on distributi ons for each part of Poland, such data are not supplied. The measure ment supplied is, f or "all Poland", a range of 20-1000 micro-roentgens per hour, or 20-1000 fR/hr. Subt racting 12 pR/hr for background, we have 8-988 fR/hr as the range, outdoors, free-in-air. So we use Criterion II (see Technical Appendix 2) to derive a gamma dose for Poland of 249 pR/hr. 31 -Gofman ¥ Cesium-137 Dose, Method 2 Since cesium-137's contribution to the gannna dose is 1.8%, the Cs-137 gamma dose= (0.018) x (249 pR/hr) = 4.48 fR/hr. We are interested in calculating the Cs-137 dose for the whole first year, and thereafter for the entire mean-life of the Cs-137 atoms. That mean-life is T1 / 0.693, or about 43.5 years. The first ~ year dose is only 2.3 % of the total dose commitment. The total dose is 43.5 times the first year dose. If all the deposited Cs-137 were to remain right on the surface for the first year ( and thereafter), the calculation would simply involve multiplication of the early deposition dose by the number of hours in a year, 8760 hours per year. But the cesium-137 has been found to work its way into the soil during the first year, with the result that the average external dose is appreciably lower than it would be if the Cs-137 had all remained on the surface. How much lower? 53 Devel! and co-workers have given a value for external dose one meter above the ground of 0.0811 mR/yr per 1000 pCi/m 2 ---provided the Cs-137 remains on the surface of the earth for the entire period of one year. 4 . Beck's work is cited by UNSCEAR as leading to the conclusion that the cesium137 works its way into soil, with the establishment of an exponential profile for the Cs-137, with a mean depth of 3 cm. For the average dose in the first year, UNSCEAR gives a value of 0.033 niR/yr per 1000 pCi/m 2. The cesium apparently stabilizes at this distribution in soil, and the average value for the first year can be used for all the subsequent years in estimation of dose commitment over the mean-life of the Cs-137. Therefore, the .value we would get for external dose one meter above the grou nd for deposited Cs-137 is too high if we use the very early dose. The correction factor is 0.0811 I 0.033, or 2.46. In our analysis, we have, for Poland, an external dose of 4.48 pR/hr. This must be divided by 2.46, yielding 1.821 pR/hr as the appropriate external dose per hour for the hours in the first year. Therefore, for the first year the total dose will be (8760 hrs) x (1.821 pR/hr), or 15,952 pR in the first year. And for the total dose commitment over the -mean-life, we have (.43. 5 yrs) x (15,952 pR/yr), or 6. 939 x 10 5 pR. ¥ Correction of the External Cs-137 Dose Commitment The UNSCEAR 4 recommendation is that the external dose in pR should be reduced by a factor of 0.32 prads/pR to take into account back scattering, shielding by the body itself, and time spent indoors, on the average. Therefore, the whole-body 32 -Gofman 5 absorbed dose from external Cs-137 = (0. 32 11rads/uR) .x (6. 939 .x 10 fR) , or 5 5 2.22 x 10 prads. Thus, external Cs-137 dose= (2.22 x 10 frads) x (1 mrad/1000 frads), or 222 millirads. ¥ Internal Cs-137 Dose Connnitment Given the usual distribution of Cs-137 dose ( 70% external and 30 % internal), we must multiply the external dose by (0.3 / 0.7) or 0.43 t o obtain th e internal Cs-137 dose. Therefore, internal dose from Cs-137 = (222 mrads) x (0.43) = 95.5 millirads. ¥ Cesium -1 34 Dose, Method 2 Poland provides no dat a for th e Cs-134 I Cs-137 ratio of deposition. We use the 48 49 average value calculate d from many other data in the WHO and EPA reports. Since the ratio i s fixed in the reactor, use of the average ratio from such measurements is fully j ustified in the absence of actual meas urem ents in a particular country. The average deposition ratio, Cs-134 / Cs-137 = 0.76. ¥ Ext erna l Cs-134 Dose Connnitment The external Cs-134 dose is th e (deposition ratio) .x (dose -effectiveness áfactor) x (external Cs-137 dose). Borrowing th e dose -effectiveness factor of 0.19 from Method 1, we calc ul ate th e exte rn al Cs -134 dose = (0. 76) x (0.19) x (222 mrads) 32.1 mrads. ¥ Int ern al Cs-134 Dose Commitment This is the (deposition ratio) x ( dose -effectiveness facto r) x (internal Cs-137 dose). Borrowing the appropria t e dose-effectiveness factor of 0.097 f rom Method 1, we calculate the internal Cs-134 dose = (0. 76) x (0 . 097) x (95.5 millirads) = 7 .0 mrads. ¥ Combined Cs-137 and Cs-134 Doses, Method 2 Total cesium dose, from above,= 222 + 95.5 + 32.1 + 7.0=356.61Dillirads. ¥ Cancer and Leukemia Consequences, Poland Fatal = (Population Size)x(Dose in mrads) . = (36.9 .x 10 6 persons) :x 056.6 mrads) Cance rs 268,000 person-mrads per fatal cancer 268,000 person-mrads per fatal cancer = 49,099 fatal cancers. This is rounded off to 49 2000 fatal cancers. Non-fatal cancers, additional, are 49,000 cases. 6 = (36.9 x 10 persons) x (356.6 mrads) Leukemia s ¥ 2025 leukemias ( rounded off.) 6,500,000 person-mrads per leukemia 33 -Gofman ¥ METHOD 3: LAST TYPE OF DATA (ITALY) This type of analysis is based upon Iodine-131 deposition on the ground, with conversion of such data to Cesium-137 deposition on the ground. Fortunately, there were few instances whe re this method had to be used. Italy was such a case. The EPA report of May 12, 1986: 9provides values for I-131 deposition in five separate regions. The average is 269 nanocuries/m 2 , or 269,000 pCi/m 2 . From excellent Swedish data on the ratio of I-131 to Cs-137 depositions, daily, in the early days of th e accident, we obtain a factor of 0.202 for converting from Iodine-131 deposition to Cesium-137 deposition. Therefore, average Cs-137 deposition in Italy= (0.202) x (269,000 pCi/m 2) = 54338 pCi/m 2. This value is used as if it were Type (1) data, (see especially EPA Report, May 12,1986). ¥ Cs-137 External and Internal Dose s Total Cs-137 dos e = (0.66 mrads per 1000 pCi/m 2) x (54338 pCi/m 2) = 35 .9 millirads. External share is 70%, or 25.1 millirads. Internal share is 30 %, or 10.8 millirads. ¥ Cs-134 External and Internal Doses Using the deposition ratio from Method 2 and the dose-effectiveness ratio from Method 1, we obtain: External Cs-134 dose (0.76) x (0.19) x (25.1 mrads) 3.6 millirads. Internal Cs-134 dos e (0.76) x (0.097) x (10.8 mrads) 0.8 millirads. ¥ Combined Cs-137 and Cs-134 Doses, Method 3 Total cesium dose, from above,= (25.1 + 10.8 + 3.6 + 0.8) millirads = 40.3 millirads. ¥ Cancer and Leukemia Consequences, Italy (Population size) x (Dose in Millirads) Fatal Cancers (268,000 person-m.illirads per fatal cancer) (5.624 x 107 persons) x (40.3 millirads) (268,000 person-millirads per fatal cancer) 8457 fatal cancers, rounded off to 8450 fatal cancers. Non-fatal Cancers, additional= 8450 cases. (5.624 x 107 persons) x (40.3 millirads) Leukemias (6,500,000 person-millirads per leukemia) 350 leukemias, rounded off. 34 -Gofman ¥ UNIFORMREDUCTIONOF "FIRST-STEP" VALUES The dose commitments from cesium derived above are not the fina l values used to assess the cancer consequences; they are "first-step" values. The final values are the entries in Table 6, which are lower . We are confident that Methods 1, 2, and 3 provide very reas onable dose connnitments from ces iums in the localities where some measure ments were report ed. But we could not know how representative those loca lities were for the whole country. For instance, the localities measured may sometimes have been the arbitr ary lo cat ions of permanent monitoring equipment, or may often have been localiti es where rainfall produced much greater concern and much more fallout. The variability of fallout within some coun trie s was illustrated by Poland, where gamma doses ranged from 8 pR /hour t o á9_88 JJR/hour on the same date. Therefore, before calculating "first-st ep" dose commitments, we tried to correct for such variability by using the two criteria stated at the beginning of Technical Appendix 2. After obtaining "fir st -step" dose commitments for each country by Methods 1, 2 and 3, we obtained reasonable factors by which all "first-step" values cou ld be reduc ed uniformly. We shall call these the "lowerin g f acto rs ." ¥ BASIS OF THE LOWERING FACTORS The "fir st -step" dose commitments from cesiums correspond to "first-step" deposition -values for cesiums. These wer e easily obtained in picocuries per 2 mete r for cesium -1 37 with a single equation. Because in Methods 2 and 3 the ratio is constant for the deposition of Cs-1 34 to Cs-137, the ratio of the dose commitment from each nuclide is likewise constant. The share of th e nuclides' combined dose commitment which is contributed by the Cs-137 is always 0.89. 2 And bec ause a Cs-137 deposition of 1,000 pCi/m gives an absorbed dose commitment of 0.66 millirads (see Method 1), the following equation can be applied for all countries where Methods 2 and 3 were used. Cesium -137 2 deposition in units of 1,000 pCi/m = (0.89) x (dose commitment Cs-137 + Cs-134 2 in mrads) I 0.66 mrads per 1,000 pCi/m ¥ With this equation, we obtained average "first-step" values for cesium-137 2 deposition in pCi/m for every country in Table 6. We multiplied by each 2 country's area in meters to get "first-step" values for total cesium-137 deposition in each country. 35 -Gofman 6 The sum of those "first-step" values was 2.73 x 10 curies of cesium -13 7 deposited in all th e countries combined. By comparing this value with some conservative est imat e s of tot al ce sium-137 relea se d by the accident, we obt ai ned two appropriate lowering factors which we applied to the "first-step" dose commitments. ¥ Cesium : Amount Released and Initi al Inventory Several gro ups have attempted to estimate th e t ota l quantit y of cesium -1 37 re leased from the Cherno byl reactor. Knox 54 suggested a value of 3.0 x 106 5 cur ies. The Imperial College Group ~in England suggested a much lower value 6 of 1. 4 x 10curi es released . It is hard to know whether one of th ese value s i s better than th e oth er . The initial invento r y of Cs-137 at the time o f th e accident depend s on the length of operation and refuelin g schedule. Estimates for the Chernobyl 54 56 . reactor have bee n offered ' , based on approxim ately two yea r s of 6 6 operation, whi ch plac e its cesium-137 inventor y at 5.8 x 10and 6.0 x 10 c-ur ies . However , because our objective is t o determ ine a credible lower-limit on the ca nce r-con seq uences from the accide nt , we hav e used th e much l ower value of 6 3.53 x 10 curies as the cesium -137 invent ory, which corresponds with one yea r' s full -power operatio n. From this mini mal value , we are goin g to derive and apply (separat ely) 65 . f Th . lt . d á h h S á see foot of Table 6 . t wo 1ower1ng actors. e1r resu s are 1n goo agreemen t wit t e ov1et report ¥ FACTOR FOR CESIUM-137 DEPOSITION OF 1,9 90,000 CURIES For one factor, we have assumed that 75% of the minimal cesium-137 inve nt ory was rel eased at the temperatur es and disruption whi ch occurred at th e 6 6 rea cto r: (3.53 x 10curies ) x (0.75) = 2. 65 x 10 curies r elease d. Aft er we assume d th at 25 % of this amount was deposite d on lands and waters not considered in the areas of Table 6, the cesium-137 deposi tion was reduced to 6 6 (2.65 x 10curies) x (0.75) = 1.99 x 10curies. This compares with our 6 "fir s t-s tep" value of 2.73 x 10curies deposited . Therefore th is low e rin g fa ctor for all the "first-step" dose commitments is (1.99 I 2. 73) = 0. 729. ¥ FACTOR FOR CESIUM-137 DEPOSITION OF 1,330,000 CURIES For th e other factor, we have assumed that 50 % of the minimal cesium -1 37 6 6 inventory was released: (3.53 x 10 curies) x (0.50): 1.77 x 10 curies 36 -Gofman released. Then this value was reduced for the 25% "loss" in areas not 6 6 considered: (1.77 x 10curies) x (0.75) = 1.33 x 10curies deposited. 6 Comparison with our "first -step" value of 2.73 x 10 curies leads to the lowering factor of (1.33 I 2.73) = 0.487. ¥ FINAL ENTRIES IN TABLE 6 After a dose commitment is lowered by one of the factors, it is multiplied by the country's population to obtain person -millirads, and then person-millirads are divided by 268,000 person-millirads per fatal cancer and 6,500,000 person-millirads per leukemia to obtain the entries for Table 6, as explained in Section 5 of this paper. The two sets of entries for malignancies in Table 6 correspond to cesium-137 depositions of 1,990,000 curies and 1,330,000 curies respectively (Technical Appendix 2-B illustrates the country-by-country calculation). The lower value of 1.3 million curies is very close to the estimate by the Imperial College Group55 It may be much too low, especially if the initial Cs-137 inventory was about 6 million curies instead of the 3.53 million curies used in this paper. Unfortunately, scientists must be skeptical about the validity of any Soviet statements concerni ng cesium-137 inventory or percentage released. Indeed, one must wonder how much the Soviets can know about the percentage released when the condition of their reactor is hidden under tons of sand, lead, and boron, and when the explosion rendered worthless any measurements at normal vents. Moreover, the Soviets have an obvious interest in underestimating the amount of cesium released, and this interes t is powerfully shared by many nuclear experts in other countries which have nu cl ea r power plants, or plan to have them. 37 -Gofman (8) RESULTINGASSESSMENTOF CHERNOBYL'SCANCERCONSEQUENCES Table 6 shows that the Chernobyl accident will cause between 634,200 and 951, 000 total cases of radiation-induced cancer, and between 13,100 and 19,500 cases of rad iatio n-induc ed leukemia. (Table 6 is on page 39.) ~ ~~nic h end of the ran ge is the more credible? ¥ REALITY-CHECKON TABLE 6's ASSESSMENT For re asons of compassion, we would much pref er that the lower values from Table 6 be t he true ones. On the other hand, we must recognize that it is the higher estimate which corresponds more closely with the "first -step" values derived from ac tual measurements (see Section 7 of this paper). And although we did not tabulate the r esu lts if the cesium-137 inventory was 6,000,000 curies instead of 3,530,000 curies, anyone can see by simple proportion that the total cancers would rise to a ran ge of 1,000,000 to 1,600,000 from t he same analysis. In th e absence of additional measurements, we will use the l ower range b ased on the lower inventory. A way does exist for the scientific community to make a r eality-check on Table 6's assessment. The cesium -1 37 and cesium -1 34 are going to remain as fallout in the various regions for a lon g period of time. Even though cesium -137 measurements, made retroa ctively without ".trays" t o collec t only fresh fallout, are complicated somewhat by residual Cs-1 37 from weapons-testing, the solution is still easy. An independent team of scientists could go to all the affected countries and measure the cesi um-134 contamination, making samples which are coded and split before analysis whenever possibl e. There is no significant cesium-134 left from weapons -testing. From such measurements, reliable values of the Cs-137 fallout from Chernobyl could be obtained . The Soviet Union would neces sar il y have to agree to such testin g by independent scientists. Whether that will ever come to pass in not known . But there can be no doubt that a correct final assessment of the cancer consequences from the Chernobyl accident can be validated if the will for such assessment exists. Meanwhile, Table 6 reveals that a credib l e lower-limit on the cancer consequences from the Chernobyl accident is: 317, 100 317,1 00 13,100 fatal cancers additional non-fatal leukemias. cancers 647,300 malignancies. It must be noted that the number 647,300 excludes cancers from the following 38 -Gofman additional sources of exposure: ¥ (a) from external gamma-dose delivered from the ground by deposited radi onuclides other tha n the radio-ce si ums. Thi s dose will add approximately 3% to each of the t otal s for maligna ncies in Table 6. For the lower es ti mate , th e sum would become 647,300 + 19,400 = 666,700 malignancies. For the hig her es ti mate, th e sum would become 970,500 + 29,100 = 999,600 malignanc ies. ¥ (b) from inhalation and ingestion of the radi o-i odines, which concentra te in the thyroid gland and can cause thyroid cancers and abnormalities; ¥ (c) from interna l dose (via f ood, water, and inhalation) delivered from radionuclides oth er than radi o-cesi ums and radio-i odin es ; ¥ (d) from the passing r adioactive clo ud, which irradiated people directl y with gamma r ays . ¥ THE DISTRIBUTION OF DOSES OVER TIME Exposure from Chernobyl 's radioactive cloud occurr ed only once , but exposu re from Chernobyl ' s cesium fal lout ex t ends through tim e , beca use of th e 2.3 yea r half -li fe of cesium -1 34 and th e 30. 2 yea r half-life of cesium-137. Cal cul at i on shows (Technical Appendix 2-C) that approximatel y 50% of all the dose ever to be received from th e cesiums from th e acc id ent will have been re ceived in a little over ten years. About 2/3 of the dose ever to be receiv ed will have been receiv ed by about the 25th year after th e acc ident. About 75% of the dose ever to be received will have been receiv ed by th e 40th year. The deliver y of about 50% of the dose commitment during the first ten years after the accident means that about 50% of the cancers in Tabl e 6 will result from th at part of th e exposure. However , the malignancies will definitely not appear simultaneously. Even if the dose had occurred in an instant instead of gradually over ten years, the leukemias woul d be spread over 25 years (with the peak excess about 7.5 years after the exposure), and the cancers would be spread over the 39 -Gofman Table 6: á Cancer and Leukemia Tolls From the Chernobyl Nuclear Power Plant Accident (Based Upon Dose Commitments In Millirads From Cesium -137 Plus Cesium -134) Corresponding To Deposition Corresponding To Deposition Of 1 1 990 1 000 Curies Of Cesium-137 Of 1,330 1 000 Curies Of Cesium-137 Country Population Method Dose Fatal Add it' 1 Leuke-Dose Fatal Addit'l Leukeor (see Commit. Cancers Non-fatal mias Comnit . Cancers Non-fatal mias Region text) mrads Cancers mrads Cancers Albania 2 , 500,000 (2) 12 110 110 5 8 73 73 3 Austria 7,600,000 (2) 174 4,900 4,900 200 116 3,300 3,300 135 Belgium 10,000,000 (1) 2 75 75 3 l. 3 50 50 2 Bulgaria 8,600,00 0 (2) 172 5,500 5,500 225 ll5 3,700 3, 700 150 Canada 22,125,000 (3) 0 .4 33 33 1 0.3 22 22 l *Czechosl. 15, 500,000 (2) 52 3,000 3 , 000 125 35 2,000 2,000 83 Denmark 5,100,000 (1) 15 280 280 12 10 190 190 8 **Finland 4,800,000 (2) 249 4,450 4,450 180 166 3,000 3,000 120 France 54,540 ,0 00 (2) 58 11,800 11,80 0 480 39 7,900 7,900 320 Germany,W 61,400,000 (2) 172 39,400 39,400 1,600 ll5 26 , 300 26,300 1,100 Germany,E Greece 17,100,000 9,700,0 00 (2) (1) 201 3 12,800 110 12,800 110 530 5 134 2 8,600 72 8,600 72 350 3 Hungary 10,600,000 (2) 41 1 , 620 1, 620 65 27 1,08 0 1,08 0 43 Irelan d 3, 100,000 (2) 1. 3 15 15 1 0.9 10 10 0 * Italy 56,200,000 (3) 29 6,100 6,100 250 17 4,000 4,000 165 *Japan 119,500,000 (3) 0.8 360 360 15 0.5 240 240 10 S.Korea 33,900,000 (3) 0.6 75 75 3 0.4 áso 50 2 Luxemb'rg 350,000 (2) 12 16 16 l 8 11 11 0 Nether'ds 14,400,00 0 (2) 12 640 640 26 8 430 430 17 Non,ay 4,130,000 (1) 86 1,300 1, 300 55 57 880 880 37 ***Poland 36,900,000 (2) 259 35,700 35 , 700 1,470 173 23,800 23,800 980 Romania 22 , 900,000 (2) 770 66,000 66,000 2,700 513 44,000 44 ,0 00 1,800 Spain 38,200,000 (2) 2 . 6 370 370 15 l. 7 250 250 10 *** Sweden 8,300,000 (1) 496 15,400 15,400 630 331 10,200 10,200 420 Switzer'd 6,500,000 (2) 236 5,700 5,700 240 157 3,800 3 ,80 0 160 Turkey 48,000,000 United K. 56,000,000 U. S.A . 235 ,0 00,000 ¥u.s.s.,. JUkraine 50 , 700,000 (2} (2) (3) (2) 100 65 0.05 936 18,0 00 13,600 44 177,000 18,00 0 13,600 44 177,00 0 740 560 2 7 ,300 67 43 0.03 624 12,000 9,100 29 118, 000 12,000 9,100 29 118,000 490 370 4 ,9 00 Byelor 'a 9,9 00 ,00 0 Moldavia 4,080,000 Baltic R. 7,660,000 Mosco w 8,400,000 Lening'd 4 , 700 , 000 (2) (2) (2) (2) (2) 714 125 104 40 148 26,400 1,900 3,000 1,250 2,600 26,400 1,900 3,000 1,250 2,600 1,100 80 120 50 110 476 83 69 27 100 17 ,6 00 1,300 2,000 830 1,700 17,60 0 1,30 0 2 , 000 830 l, 700 730 55 80 35 75 Yu~oslav. 23,000,000 (.2) 185 15,900 15,900 650 123 10,600 10.600 430 -áá-----á---------------------- -----------------------------~------------------------------------------------------- Sum (all countries) 475,500 475,500 19,500 317,100 317 ,100 13,100(Rounded off) ------------------~---------------------------------------------------------------- Total Malignancies¥ 970,500 Total Malignancies ~ 647 , 300 -------------------------------------~----------------------------------------------------------------------------- *Czechoslovakia, It aly , Japan , USSR: The values in Table 6 are pr obably too low; details in Technical Appendix 2-A. We have no data for the ar ea close t o Chernobyl, and none for the Russian SSR except for Mos cow and Leningrad . **Finland: There have been ser iou s inconsistencies in the Finnish data; detail s in Technical Appendix 2-A. ***Poland and Sweden: Polan d reported extremely high gamma-dose rates in Warsaw during the early days of the accident, but these val ues were later deleted from EPA reports as "too hig h" without any expla nati on (compare EPA reports of May 12 and 14 with the EPA report of June 4 , 1986). Sweden reported extremely high gamma measurements in Uppsala for April 29 , but these high values simply disappeared from later reports witho ut explanation (compare EPA reports of May 8 and 9 with EPA reports of May 12 and there after) . In epidemiological science , authorities cannot select only high measurements for checking; unless low measurements are checked for error with exactly the same amount of diligence, the ne e result i s to create a bias toward lowering a whole set of measurements. Such practice is not acceptable in science. August 22, 1986: The Soviets are estimating 1,000,000 curies of cesium-137 deposition within their own european regions 65 ¥ Table 6 matches extremely well with the Soviet val ue. The higher estimate of dose and malignancies corresponds with cesium-137 deposition of 991,874 curies in european r egions of the Soviet Union; see Technical Appendix 2-B. The lower estimate in Table 6 correspon ds with 2/3 of that value, or 661,458 curies. 40 -Gofman remaining lifespans of the irradiated population (with the peak excess occurring between 30-40 years after exposure). ¥ THE DISTRIBUTION OF IMPACT BY AGE The third "law" of radiation carcinogenesis (Section 1 of this paper) means that children will be the most affected by the cesium fallout. Not only will they experience more fatal cases per 100,000 exposed individuals than will adults, but each cancer fatality means a far greater loss of lifespan for those irradiated young than for those irradiated at older ages. This point i s demonstrated by considering three ages: newborn, age 25, and age 45 at irradiation. When newborn males are irradiated, among those who do develop fatal radiation-induced cancer, the average loss of life expectancy is about 22.3 years. Half of those cases die before reaching age 54.5 years, and half die later. By comparison, if irradiation occurs at age 25, among those who do develop fatal radiation-induced cancer, the average loss of life expectancy (for males) is 12.8 years. Half of such cases die before reaching age 67.5 years, and half die later. And if irradiation occurs at age 45, among those who do develop fat al radiation -induced cancer, the average loss of life expectancy is about 8.7 yea rs. Half of such cases die before reaching age 75.2 years, and half die later. The calculations leading to the statements about loss of life expectancy are based upon Tables 21 and 56 in Gofman3 . (9) DISCUSSION AND CONCLUSIONS ¥ THE SINGLE MOST SERIOUS INDUSTRIAL ACCIDENT EVER It is correct to say that a single event ---the Chernobyl accident ---has caused between 600,000 and a million cases of cancer and leukemia. The radio-cesiums are on the ground, and humans are committed to receive the doses from them. To the extent that a share of the dose has already been received, a share of the malignancies is already underway, even though they will not become manifest, .. clinicall y, for years. The Chernobyl accident obviously represents the most serious industrial tragedy in the history of mankind, and by a very large factor. 41 -Gofman ¥ THE QUALITY OF EVIDENCE With respect t o the proven human ca rcin og ens, the existing quantitative evidence 4 6 of human carcinogenesis by ionizin g r ad i ation is second to non e (UNSCEAR , BEIR, 3 11 GOFMAN ,N.I.H. ).The data on ionizing radi ation may be the strongest of all, and they cover vir tually every site of human cancer. Moreover, several studies examine 25 very low doses ---a total of 250 millirads in one series ; even the A-bomb su rviv ors pr ov id e a lar ge subset of people who re ce i ved less than 20 rads of ex18 posure In addition, studies of occupat ionally and medical ly exposed populations have con trib uted much evidence at low doses. Coupled with the quantitative human eviden ce hard-w on over the past half cen tur y, the three gene r ali zations described in this paper provide a very good assessment of the cance r conseque nces of th e Chernoby l accident . The real pr obl em we have in making s uch an assessment is s i mply th e acq ui sition of do se dat a . The proble m does not have to do with any mystery about consequences, once the doses are known. ¥ WHAT WE NEED, AND DO NOT NEED, TO ASSESS CHERNOBYL ACCIDENTS On Ju ne 6, 1986, Mr. Stuart Loory, br oadcasting from Moscow to many nations on the Cable News Network , reported ~hat an agreement had been reached between Dr. Robert Gal e of the U.S .A. and the Soviet Govern ment t o arrange for a li fetime study of the approximately 100,000 persons who received high doses from Chernobyl and were finally evacuated from the nearby area . Mr. Loory added that such a study might dete rmine for radiat ion and cancer what we already know for cigarett e smoking and cancer. We can imagi ne nothing further from the truth than the suggestion th at science ha s not yet firmly es tablished a causal relati onship betwee n ra di at ion exposure and human cance r . If the fo llow-up study of th e Soviet high-dose group is promoted as necessa ry to.establish thi s relati onshi p, it will represent a cruel decepti on of mankind demonstrates in quantvery l ow doses . concerning itative detail the massive the prodbod y of existin uction of cancer g evidence by radiation, which already and at ¥ A PREDICTION We ca n predict with high confidence that an honest study of the proposed population sample will simply confirm ---but decades from now ---the magnitude of radiation production of cancer, a magnitude we know quite well prior to such 42 -Gofman a s tud y. The existing human evidence provides a solid basis for assessing the Chernobyl toll. The credible lower -l imit of malignancies from the cesium fallout is approximately 640,000 cases, and a credible upper -limit is probably 1,600,000 malignancies. Only additional and reliable measurements of cesium fallout, made by independent scientists, can narrow th e ran ge. ¥ IMPLICATIONS FOR MEDICAL, DENTAL, AND OCCUPATIONAL IRRADIATION The findings in Section 2 of this paper that there cannot be a safe threshold dose of ionizing radiation with respect to human carcinogenesis, and that linearity cannot exaggerate the carcinogenic effect at very low doses, disprove th e "h orme tic" notion that exposure at low doses may protect humans . 1 á . 62 against ma ignancies Also the findings of Section 2 have daily applicability for medical, dental, and occupational exposures. Although lip -service is generall y paid to the absence of any safe dose, in reali t y t he hazard at low doses is often dismissed as "purel y th eore ti ca l." The finding s presented here show why the hazard is not imaginary ---it is real . The agg re gate dose each year from dia gnos ti c radiology is sufficient to cause about 78,000 radiation-induced cancers per year in the United States 46 alone (Gofman-O'Connor , pp.365-70). Occupational exposures, in their aggregate, add another l arge number. The findings in Sections 2 and 3 of this paper provide ample evidence that measure s to redu ce individual doses would constitute a scientif icall y sound method of achieving large reductions in the human cancer-rate. 43 -Gofman TECHNICALAPPENDIX 1 The Basis For Table 1 Part A of this Technical Appendix shows the basis for estimating th e approp ri at e size of most human cells, and Part B demonstrates th e series of ca l cul ati ons which produced Table 1 in th e t ext. ¥ (A) DETERMINATION CELL N OF HUMA-SIZE Because th e size of human cells and their nucl ei has an impact on th e re sults in Table 1, the choice of appropriate size was not made cas ually. A search was made for ele ctr on micrographs of human tissue fixed with glutaraldehyde, for the y sh ould provide th e most r eliable dimensions in -situ for cell nuclei . Several hi s tology atlases 57 ¥ 58 ¥ 59 and the twelve-volume work of 60 Johannessen were consulted for such micrographs. Specif ically sought were micrographs fr om normal human tissu e in which nuclei were prominent . Espec iall y good were th ose micrographs where several nuclei were present , so th at the one with the largest diameter could be chosen from the group. The largest was chosen because, in sectioning the tissue, some nuclei will not have been sectioned thr ough their maximum nu cl ea r dimen sio n. The following nuclea r di mensions were ascertai ned: Mean Nucl ea r Diamet e r Fr om 29 sui t abl e From 6 suitable From 1 suitable From 1 suitable From 1 suitable mi cro gr aphs mic r og r aphs micrograph micrograph micrograph of of of of of non -f etal human cells f e tal human cells non-fetal human thyroid f etal human th y roid non-fetal human breast 5 . 9 micrometers 6 . 1 micrometers 5 .7 micrometers 6 . 9 mic romete r 5.5 mi cr ometer s s For th e epidemiological studies evaluated i n this paper, it is approp ri ate to start with the mean nuclear value , which is 5.9 micrometers . Two cor r ectio ns of this dimension wer e made. First, becau se it is impossible to know that the nuclei pictured were cut exactly through the maximum dimension, a factor of 1.1 increase was applied. Second, because it is possible that fixation may have caused some shrinkage of cells, another factor of 1.1 increase was applied. The total correction applied= 1,1 x 1.1 = 1.21. So, di amete r of nuclei for this paper is (5.9 J.1I11)x (1.2 1) = 7. 1 micromet ers . A very reasonable estimate, based on examining numerous cells in histol ogy texts, i s that cell -diame t er is tw i ce the nuclear diameter. Therefore, for thi s paper, cel l-diam e ter is 14.2 micro meters . A spherical nucle us with a diameter of 7.1 J.1I11has the same volume as a cuboidal 44 -Gofman nucleus of 5.7 pm per edge. A spherical cell with a diamet er of 14.2 pm has the same volume as a cuboidal cell of 11.4 pm per edge . ¥ (B) CALCULATIONSPRODUCINGTABLE 1 We shall present the ca l culations s upport i ng the ent ries in Table 1 for the cell-size of 11.4 microns (cuboid a l). The six steps are the same for cells of any other size. However, once a value for Step 5 has been obtained fo r one cell-s iz e, th e cor resp onding value for ce lls of any si ze can be obtained with the general formula presented here af t e r Step 6. ¥ Expression Of 100 Millirads In MEV 7 1 rad means 100 ergs or 6.25 x 10 MEV de livered per gr am of tissue. 6 100 millirads means the deliver y of 6 .2 5 x 10MEV per gram of ti ss ue. ¥ Step 1: Number Of Cell -Nuclei Per Gram Of Tissue 3 Density of ti ss ue = 1.1 grams/cm . Cell-size (cuboidal) = 11.4 mic r ometers per side . 3 Volume = 1481.5 pm . 3 12 3 -9 Mass of one cell= (1481.5 pm )( 1.1 gms/10 .m) 1. 630 x 10 grams. Number of cells per gram of ti ss ue= 9 8 1.0 gr am I (1.630 x 10-grams per cell) = 6.1 35 x 10cells . 8 Number of nuclei per gram of tis sue = 6.1 35 x 10nucle i. 6 ¥ Step 2: Number Of Primar y El ectrons Required To Deliver 6. 25 x 10MEV Per Gram Of Tis sue By using the energy per photo-electr on below, we exaggera te the energy per pr i mary elec tron since ioni zin g radiation converts also t o Compton el ec tron s of l ower initi a l energy. Beca use th e average energy is really so mewhat lower than s t a t ed , all the va lu es in Step 5 should be a bit lower and all th e doses in Step 6 should be a little higher. We are usin g the hi gher energies in order to sta y conservative in demonstrat ing the case against a safe "threshold dose." For Cesium-137, the gamma ray is actually from Barium-137m decayi ng to Barium-137. Radium's value of 0.596 MEV per photo-ele ctr on repre se nts th e weighted average gamma-ray energ y from Radium-226 and it s daughters. 50 KEV repres~nt s the average energ y of ph oto -electron s when the peak ki lov oltage of X-rays is about 150 KEV. 30 KEV r epresents the average energy of photo -electrons when the peak kilovolt age of X-rays is about 90 KEV (typica l for medical dia gnost ic X-rays). 45 -Gofman (number of primary electrons for dose (MEV per 100 mrads) I (MEV per electron) of 100 mrads) Cesium-137 Radium-226 50 KEV X-rays 40 KEV X-ray s 30 KEV X-rays 6.25 6.25 6.25 6.25 6.25 x x x x x 6106MEV 106MEV 106MEV 106MEV 10 MEV I I I I I 0.662 0.596 0.050 0.040 0.030 MEV per MEV per MEV per MEV per MEV per e e e e e = = = 9.44 1.05 1.25 1.56 2.08 x x x x x 10~ electrons. 108 electrons. 108 electrons. 108 electrons. 10 electrons. ¥ Step 3: Number Of Cells Traversed By Each Primary Electron The di sta nce traveled by each type of photo~electron is its initial energy divided by the energy it l oses per micrometer of tissue traversed (its linear energy transfer, or LET). The number of cells trav ersed is th e distance divided by the cell-size of 11.4 micrometers. (initial energy in KEV) I (LET) = distance per electron Cs-137 (0 . 662 MEV x 1000 KEV/MEV) I (O. 28 KEV /1Jm) 2,364 micrometers. Ra-226 (0.596 MEV x 1000 KEV/MEV) I (O. 2 9 KEV IJ.1D1) 2,055 micrometers. 50 KEV X-rays (50 KEV) I (O. 84 KEV /.m) 59.5 micrometers. 40 KEV X-rays (40 KEV) I (1. 00 KEV /J.lDl) = 40.0 micrometers. 30 KEV X-ra ys (30 KEV) I (1. 20 KEV /.m) 25.0 micrometers. (distan ce) I (11.4) = (cel l s traversed by each primary electron) Cesi um-137 2,364 pm I 11.4 .m per cell 207 . 4 cells Radium-226 2,055 pm I 11.4 pm per cell 180.3 cells 50 KEV X-ray s 59.5 pm I 11. 4 J.lDl per cell 5. 22 cells 40 KEV X-ra ys 40.0 .m I 11.4 pm per cell 3. 51 cells 30 KEV X-rays 25.0 .m I 11. 4 J.lm per cell = 2.1 9 cells ¥ Step 4: Numbe r Of Cell -Nuclei Traversed By Primary Electrons Delivering 100 Milli r ads To a Gram Of Tissue-Cells An electron approachin g normally to a cell "sees" a nuclear area which is~ the area of the total cell 's area. Therefore, the number of nuclei traversed by primary electrons will be about~ of the number of cells traversed. In the calculation bel ow, the number of primary electrons (from Step 2) times the number of cells traversed by each (from Step 3) provides th e number of cells traversed, and this is reduced by a factor of 0 . 25 to obtain the number of nuclei traversed. (number of nuclei For 100 mrads: (electrons) x (cells per e) x (0.25) = traversed) 6 8 Cesium-137 (9.44 X 10) (207.4) (0. 25) = 4.89 X 10 7 8 Radium-226 (1.05 X 10) (180. 3) (0.25) = 4.73 X 10 8 8 50 KEV X-rays (1.25 X 10) (5.22) (0.25) 1.63 X 108 40 KEV X-rays (1.56 X 108) 8(3. 51) (0.25) = 1.37 X 108 30 KEV X-rays (2. 08 X 10) (2.19) (0.25) = 1.14 X 10 46 -Gofman ¥ Step 5: Number Of Primary Tracks Per Cell -Nucleus At 100 Millirads The number of primary ionization tracks which pass through a cell-nucleus is obtained by div idin g the number of cell-nuclei traver sed (from Step 4) by the number of nuclei present in a gram of tissue (from Step 1). Traversal by aáá pr i mary ionization tr ack is commonly called a "hit." For 100 millirads: 8 Cesium-137 (4. 89 X nuclei hit) / (6.135 x 10nucl ei)= 0.7971 hit 8 Radium-226 (4. 73 X nu clei hit) / (6.135 x 108 nuclei)= 0.7710 hit 50 KEV X-r ays (1. 63 X nuclei hit) / (6.135 x 108 nuclei) 0 . 2657 hit 40 KEV X-rays (1. 37 nuc l ei hit) / (6.135 x 108 nuclei) 0.2233 hit 30 KEV X-rays (1. 14 nuclei hit) I (6.135 x 10 nuclei) 0.1858 hit Since a nucle us is either hi t or it is not, th e use of average values with fractional hits is just a preparatory device for Step 6. ¥ Step 6: Millirads Required To Cause an Aver age Of One Primary Ionization Track In All Cell -Nuclei Of an Irradiated Tissue This value is th e dos e of 100 millirads divided by the average number of primary tracks pe r nucleus occurring at 100 millirad s. Cesium -137 When 100 mrads cause 0. 7971 tra ck, then 125 mrads cause 1 track . Radium-226 When 100 mrad s cause 0.7710 track, then 130 mrads cause 1 track. 50 KEV X-rays When 100 mrads cause 0.2657 tra ck, then 376 mrads cause 1 tr ack. 40 KEV X-ray s When 100 mrads cause 0.2233 tra ck , then 448 mrads cause 1 track . 30 KEV X-r ays When 100 mrads cause 0. 1858 t rack, then 538 mrads cause 1 track . ¥ Formula To Convert Values For Other Cell-S i zes The general relationship of values from Step 5, according to cell size, is New Value (Number of nuclei hit, old size) x (old/new size) 3 =(Number of nuclei present, old size) x (old / new size) New Value = (Old result from Step 5) x 1 _(_o_l_d-/n_e_w_s_i_z_e_) 2 ¥ Example: To convert the value in Step 5 from 11.4 micron cells to 10 micron cel l s, for 30 KEV X-ray s, we writ e New Value= (0.1858 hit) x 1 = 0.14296 hit. 2 (11. 4/10) 47 -Gofman TECHNICAL APPENDIX 2 The Basis Of Table 6 Part A of this Technical Appendix shows the handling of fallout data, country by count ry . Part B provides the area of each country and demonstrates how the 1,990,000 curies of ces ium-1 37 are distributed country-by-country. Part C shows the calculations supporting the statement that 50 % of the dose commitment from th e cesiums will occur during the first ten years after the Chernobyl accident. ¥ (A) TYPES, DATES, AND HANDLING OF FALLOUT MEASUREMENTS, BY COUNTRY ¥ General Criteria Examinatio n of all the fallout data from the various reporting countries shows that there is a high degree of variability of results within a single country . This is wholly expected, lar ge ly because rainfall can grossly increase deposition of r adionuclides, and also because cloud plumes seldom cover a count r y uniforml y . In ou r endeavor to obtain the be st representation of the average dose received by residents of any region, we have established some cri teria for handling the limited quantity of fallout data provided from the various coun tri es . ¥ Criterion I: Any country can be divided into four quadrants. When data are presented for each of the quadrants, we shall use the data as presented . When data are provided for three quadrants, we shall assign a zero value for th e fourth quadrant, and then shall average all four val ue s. When data are provided for two quadrants, we shall assign a zero valu e for th e two remaining quadrants be fore averagin g th e four values . When data are provided for a single quadrant, we shall assign a zero value to each of the remaining quadrants before averag in g . Thi s set of pro cedures provides a cautious method of estimation. ¥ Criterion II: Some data are rep or t ed as a range of values. If values are provided within the range and if all four quadrants are represented, we shall average the values given. Where only th e outer l imits of the range are provided, we shall take th ese to represent two of the four quadrants of the country, and shall assign zero values to two other quadrants, and then average al l four values. ¥ CONVERSION OF UNITS 9 There ar e 10 picocuries per millicurie. Ther e are 27 picocuries per Becquerel. There are 100 rads per Gray. There are 100 rems per Sievert. Ther e is 1 rem per rad, for gannna and for X-rays. 48 -Gofman COUNTRY SOURCE COMMENTS Albania WHO Report June J2, 1986 Data are given for only one site. Therefore, t hree zeroeswere assigned to other sites. Tbe final result is 1/4 the value given for the one site. Date for the direct gamma dose was not given. Therefore, Cesium-137 % is taken as 1.8%, th e lowest possible value, This effective l y ----------------------~-minimizes_the_fallout_estimate._~-~~--------~~-~-------~-------------Austria EPA Report Excellent data are provided for the direct gamma dose. The average dose is based upon reports June 11, 1986 from 322 stations, The peak value for direct gamma dose was almost always for May 2,1986 . ------~--------~ ---------ThereforeL_the_a22r£J?ria te _Cesium-137_% is 5.8% of the tota.!_samma doseá-----------~---Belgium WHO Report Data are given as a ran ge for Cs-137 deposition, for May 9 ,1986. Therefore , two zeroeswere June 5, 1986 assigned, and the average of these plus the r ange limits were used to obtain average Cs-137 ---~------------------~--dee osition. --~~------~~------~--~ -----~------~------------------Bulgaria WHO Report Data for the direct gamma dose are given for five separated sites, The average of these is June 5, 1986 taken. Values are for May8, 1986. The appr opriate Cs-137 % is 12.9% of the total gamma dose. --------------~-~---~-----------------------~-----~---~--~-~ ----~------------------ Canada EPA Report Data are given for Iodine-131 deposition on the ground for nine widely separated locations -June 11, 1986 a r easonable representation of Canada. Most deposition i>alues given were for May 12 or May 13. The conversion factor ( in Method 3) for conversion from Iodine -131 to Cs-137 -------------~-~---------takes_these_dates_into_account_a22!£J?riatel~ --~~-------------~ Czecho-WHO Report A single peak value of 200 uR/hr is given for the dire ct gamma dose. Three additi onal slovakia June 5, 1986 values of zero were assigned, giving an average v.alue of 50JJR/hr -for use in Method (2) calculation . Since the only indication for the date of this one readin g was that it was before May 6, 1986, caution requires using the l owest Cs-137 %, the value of 1.8% of the total gamma dose. The effect is to make the cancer estimate given here too low, if the -------------------------~true_date were later than_Aeril_22,_1986._~---~-------~------------------ Denmark WHO Report Excellent data are provided. The mean va lue for the integrated Cs-137 and Cs-134 depositions June 5, 1986 on the ground are provided for the period between May 7 and May 27, 1986. The data were obtained as a mean for 10 separate locations, label ed as "countrywide". It is not clear whether there may have been add iti onal depositions before the May 7, 1986 date. If there were additional depositi ons , the Cs-137 and Cs-134 deposition totals here are too low, and _______ ______ ___________ the_cancer_estimates are also too low. ------------------------- - In the May 30, 1986 WH0report, the statement is ..;.de that "the deposition of Cs-137 varied Finland WHO Report 2 May 30, 1986 between 100 and 1300 kJlq/m ." These values would lea d to an extremely high cancer r ate and compared with the ones in Table 6 of this paper. WHO Report In the June 5, 1986 WHO report, these data hsv e just disappeared and the following data, 2 June 5, 1986 bearing no resemblance, are presented for the cesiums: "Contamination of surface soil in 2kBq/m plus (in-situ measurements) 2-7May in Southern Finland was as follows: Cs-137 3 to 40 kBq/m ; Communication Cs-134 0. 9 to 24 kJlq/m ¥ " with Inquiry produced from the Finnish Cent re For Radiation and Nuclear Safety the reply that Finnish Auth-''WHO made an obvious err or in their first figures fr om Finland. We straightened out that orit ies. mistake, but why WHO did not inform in their next report about their misprint , I do not know." Also The letter, dat ed July 16, 1986, was signed by Olli Paakkola, Acting Director of th e SurveilFinnish lance Dept., Finnish Center for Radiation and Nuclear Safety. Reports: In the same l et ter , it is stated that "only half the country was affected by Chernobyl STUK-B-VALO44 fallout," which is the basis for usin g half the area in Technical Appendix 2-B. STUK-B-VALO 45 Finni sh authorities are designating one-third of the 5,amma-dose measured for Uusikaupunki as representative of Finnish exposures (STUK-11-VALO ¥ That value is the basis for 45) ------------------------Method 2 calculations for Finland,_ and for the entries in Table 6. ---------------- Fran c e EPAReport Only a single value is given for the direct gamma dose rate, It is for Paris for May 4, 1986. June 11, 1984 Three additional values were assigned as zero, and hence the average is 1/4 of the value for the Paris datum . The appropriate value for Cs-137 % of gamma dose rate is 8.0%. It is remarkable that France, a sophisticated nation in the field of nuclear power , provides ---------------------------so little data to WHO and the EPAá--~-~--~--~--~-----------~----~--~ East Germany provided no data at all to the World Health Organization. Since it lies between East provided Germany, No data Poland and West Germany, it is rea sonable to assign i t a dos e intermediate to that of Poland and West Germany. Since West Poland most probably had a lesser dose than East Poland, we have weighted the West German dose twice as heavily as the Poland dose, to arrive a reasonab le es timate for East Germany. It is certainly regrettable that the East German authorities ------~--~-------------saw fit_to_refuse to~ videan~ measurements. --~-~ Germany, WHO Report The WHO Report provides an "avera ge " value for the direct gamma dose rate for Southern West May 30, 1986 Germany for May 4, 1986. A comparison of air values for many stations in Northern Germany showed that the fallout was heavier in the Southern regio n than the north ern region, By using such comparisons, a value was estimated for Northern Germany, It appears that most of the data review ed are for the eastern region of Southern Germany. Therefore, two additional values of zero were assigned for the western quadrants, north and south,in arriving at an appropriate valu e for the gamma dose rate. Since the gamma dose rate {e reported for Kay 4, --------------~--------1986L_it is ae2ro2Eiate to take 8.0% as the Cs-137 % of the total ~aerate. ----Greece WHO Report A single value for Ca-137 depoaition ia given as follows: 2 May 30, 1986 " Hay 9-11 0.8 kllq/111" Thie is difficult to interpret, since the data aa reported suggest that the value reflects and WHO Report only deposition for the period between Kay 9-May 11, rather than the entire surface conJun e 5, 1986 taaination with Cs-137 on the ground, Nevertheless we have used this value here. Since only a aiagle value ia given . ve have aaeigned a ~ero value to three other quadrants. giving a final value 1/4 that of the single value given. ___ _;;Bo.th WHO Reports show the_aame inadequate statement concernin~a-137 deposition. Hungary WHO Report Direct g&1111adose rates are presented ass range for May 1, 1986. Therefore, the outer limits June 5 , 1986 of the range plus two asaigned valuea'of zero for two other quadrants are all uaed. The final average ia 1/2 the aid-point of the given range. Thi¥ vaa uaed in Method (2) calculation of Ca-137 deposition, For May 1, 1986 aeaauraoenta, the appropriate Ca-137 %of toc.l gmaaa ---------------doae rate 1a 4.9 %._____ _ 49 -Gofman COUNTRY SOURCE COMMENTS Irelan d WHO Repor t A sin gle measurement i s given for di r ect gamna dose rate for May 7, 1986 . Therefore, three June 12, 1986 additional values of zero were assigned for other quadrant¥ , giving a final value to be used in Hethod (2) of 1/4 the measurement given. For Hay 7, 1986 .eaaureaenta, the appro_____________________ __J!riate Cs-137 _% of tota!Jamna do se rate la 11. 7 %. -------------------- Italy EPA Repor t Data are given for Iodine-131 depositions for five separate locations in Italy, for dates Hay 12, 1986 ranging fr om May 1 through May 3, 1986. To be cautious, we a re treating these values as cumulative depositions, but if they are values for ain gle days, we are underestimating Cs-1 37 -------------- -------------de~sition and dose bt Hethod_{llá---------------------------------------------- Jap an \lHO Report Deposi ti on of Iodine-131 is given for four separate locations in Japan. The re sults are June 5, 1986 giv en "by day" , so that they may not reflect th e cumulative deposition of Iodine -1 31. I f and thi s is tr ue the Cs-137 deposition es t imated by Method (3) is to o low, and the ca ncer EPA Report estimates pres ent ed here are a l so to o lo w. __________ _ June _ll ~_l986 ________________ _____ __ _ Korea WHO Report No direct data are given for South Korea. However, measurement of Iodine-131 in air in (South) June 5,1986 Seoul, Korea is available for comparison with Iodine-131 in air in Kanagawa , Japa n for and the same day. So, an indireci: calc ul atio n can be made bas ed upon the Japanese deposit ion EPA Report data plus the Korea -Japa n comparison for air data. June 11, 1986 While thi s i s not an i deal ba s is for calcula ti on, it cer t a inly gives t he order-of-mag -----------------------nitude _l evel_fo r_ canc er s in_Koreaá -------------------------------- Luxembour g WHO Repor t A single direc t gaam,a dose rat e measu r ement i s given for May 2, 1986. Therefore, a zero June 5, 1986 value was ass igned for three additio nal quadrants, and the final average value used in Method (2) is 1/4 the given valu e. For May2, 1986 mea surements, the appr opriate Cs-137 % ____ __________ ________ __ of total~amma dose_rat e is_5 .8%. _____ _____ ___ _ Norway WHO Report Excelle nt data are given for Cs-137 depositi on on the ground. The data are presented as June 5, 1986 the cumulative surface soil contamination by Cs-137 f or th e peri od between Mayland -----------------Mal 22, 1986. _The results are_b as ed~ on 70 se f"~ampl es bav~ been measured. _ Netherlands WHO Report A direct gamma dose rat e i s gi ven for Hay 4, 1986 and thereaft er. However, since it is not June 5, 1986 clear whether this dose is for a singl e locati on or is an average, we have, for caution, assigned three additional zero values to other quadrants. The final value used in Method(2) is, therefore, 1/4 of th e given value. For May 4 , 1986 measurements , the appropriate Cs-137 % _______ __________ __ of total 11amma dose rat e is 8.0%á------------------------------- Pol and WHO Report Multip le direct ganm,a-d ose rat e meas urements are provided as a ran ge for "all Poland". J une 5, 1986 The two extremes of the range are taken and an additi onal two zero values are assigned to two quadrants. Therefore, the final value used in Method (2) calculations is 1/4 of the midpoint of the range giv en. For measurmen ts made on April 29, 1986 , the appropriate Cs-137 % of tot al gamma dose rate is 1.8%. Early EPA reports showed extremely high values for gamma-dose rates in Warsaw, Poland in the early period of fall out. These values were deleted in later EPA reports, a s noted in ____________ ___ ___ _____ Table_6. _ Ingui!)'._revealed tha t EPA did not_Jcnow the reaso!!_i_ "must too_hi11~as sted. be _s u1111e Romania WHO Report Multi ple dire ct gamma dose rates are given as a range for the period April 29, 1986 through June 5, 1986 May 8 ,1986. The two extreflles of the range for Hay 1 are used and zero val ues ar e ass igned for two addit i onal quadr ants. The final valu e for Hay 1 used in Method (2) ca lculati ons is 1/4 of the midpoint of the range for that date. For May 1, 1986 measurements, the appro --------------- -----------f!:1 ate value_for Cs-13 7 % of total _!!!_mma dose rate i s 4.9%. --------------------- Spain WHO Report Di rect gamma dos e rates are given as a range for the period April 29 to May 8 , 1986. June 12, 1986 The extremes of the range fo r Apr il 29 are taken and zero values ar e a ssigned for two additional quadrants. Ther efore, the final value used in Method (2) calculations i s 1/4 of the midpoint of the range for April 29. For April 29 measurements, the appropriate __ ___ _____________ ____ value for Cs-137 % of total 11amma dose rate_is l._8_ %_._ _ _ Sweden WHO Report Detailed data are provided for Cs-1 37 deposition on the groun d f or e ight separate stat ion s . May 30, 1986 Four statio ns report deposition for May 15, 1986 and four other stations report deposit ion data for April 30, 1986. While th e early data may be very much too l ow for measuring the cumulative deposition of Cs-137, th ose data were avera ged 1n with th e data for May 15. It i s puzzling that Sweden did not con tinue reporting measurements af ter April JO a t four of the stat ions. Also it is puzzling that very hi gh gamma-doses reported from Uppsala on Apr il 29, in the EPA repor ts of May 8 and 9, simply disappeared as not ed in Table 6. EPA i s le ft in its May 12 report and there.aft er with a single value for Uppsala (1,000 uR/hr on May 4) and no other data at all for that city. Since the eight stations reporting on cesium deposi tion were mainly in ~tern Sweden, we ele cted t o assign zero value s for western Sweden. Therefore, the final value used is half th e average for the eight reporting stations. á This approach may undere stimate radiation-induced can cers in Sweden. The basi s !gr uaillg half th e area of Sweden 1n Technical Appendix 2-B is the map on page 32 _____________ ______ of_Hoheoemser ¥ Switzerland EPA Report Di rec t gamma dose rates are given for four parts of the country, central, east, vest, and June 11, 1986 south . These values are for Kay 4, 1986. The average of these four gamma dose rates 1¥ used for indirect estimation of Ca-137 by Method (2) . For Hay 4, 1986 measurements, the ---------------___!fF~P!iate value for Ce-137 %of total 1¥--¥_d~o~se.c..;r~a~t~e;_;i~¥;....;8~á~0_;.;%~á------á----Turkey WHO Report A range of valuea for the direct g..,... dos e rate is given for the period Kay 4-Kay 7, 1986. June 12, 1986 The two extremes of the range are taken for Hay 4 and a zero value is asaigned to two additional quadrants. The final value used in Method (2) calculations 1¥, therefore, 1/4 of the aid-l)Oint of tha range for Kay 4, 1986. For Kay 4, 1986 .easuraaenta, the appropriate _________ _ _________ ce-137 % of total a-dose rate is 8 . ... %_. _ _ 0 ___ ___________________ 50 -Gofman COUNTRY SOURCE COMMENTS United Report in Fry, F.A., Clarke, R.H., and O'Riordan, H.C. published a paper entitled "Early Estimates of Kingdom NATURE UI: Radiation Doses from the Chernobyl Reactor". Thia useful paper provide¥ representative Volume 321 data for ganna dose rates, vsighted by population diatribution for tvo sajor regions of the 52 15 Hay, 1986 United Kingdom. "South" 1a the description of the region with 82.1% of the UK population, and ''Nort h" 1e the region with 17. 9 % of the U1( population (including the northwest of England, North Wales, Scotland, and Northern Ireland). Theae doae rates for Hay 2 ,1986 were used here for estimation o! Cs-137 de position by the indirect ¥ethod (Method (2)). ----------------------For H!I _~ _ l986 measureaientaL the ¥eeroeriate_Ca-137 % of total_6a....,_dose_ratea is 5 . 8 %. __ United EPA Repor t Deposition of Iodine-131 on the ground is reported for fifteen widel y separated stations States Hay 11, 1986 in the United States. These data are aatisfactory for indirect estimation of Ca-137 depo a ition by Method (3). The iodine-131 dep osition data are for Hay 5-Hay 8, 1986, and --------------------------I-131 to_Cs-137 _converaion __factors_for_thoae dates were used. __________________ _ Yugoslavia WHO Report Direct gamma dose rates are provided for three separate regi ons. Peak gamma dose rates were Jun e 11, 1986 reached Hay 2 -May 3, 1986. Two of the regions were close together, so the average of these two was used as a single value. Zer o values were assigned to two additional quadrants. Then an average was taken of the four values so derived. This average was us ed in Method (2) calculation of Cs-137 deposition. For measurements of Hay 2, 1986, the appropriate value of _____ ____ ________ __ Cs-137 % of total gamma dose rate h_5.8%. ____________________ _ Note: The values in Table 6 are rounded off. Some may have preferred that we do not round off so as to facilitate cross-checking between column entries. Others complain that the goodness of the data do not justify keeping the number of significant figures which would be present with out rounding off. Thia dilemma ia ever-present. The reader simply needs to keep in mind that rounding has been done, when the reader makes use of Table 6. u.s.s.R . WHO Report Dire ct gamma dose rates are report ed for Oster, just north of Kiev, startin g with May 9, Ukraine Jun e 12, 1986 1986. The data for Hay 10 are used as a first step in the indirect estimation for Cs-137 deposition. A second usabl e value is that for Kishinev, Moldavia, vhich borders the Ukraine in the southwestern region. Theref ore , we have assigned two zero values to cover the other quadrants of the Ukraine. The final avera ge for gamma dose rate for Hay 10 is that obtained by avera ging the values for Oster , for Kishine v, and the two assigned zero values. For Hay 10 measurements, the appropriate value for Cs-137 % of total gamma dose rate is 16.0 %. We should note that the Ukraine is one of the regions where Cs-137 remai.ns available to plants through th e root -so il pathway for longer periods than is the case elsewhere. As a result, our estimat e of the internal dose from Cs-137 to residents of the Ukraine _______________ __________ mar_be too low. ~-----~----- ---- u.s . s.R . WHO Report Direct gaDDDadose ra tes are reported for Bialystok, Pol and on the vest border of Bye loru ss ia. Byelo-June 12, 1986 And, as mentioned above, direct gaDDDadose rates are available for Oster (north of Kiev , russia and 100 km south of the southern border of Byelorusaia) . It appears reas onable that the average of these two result s can be used as representative of the southern 1/3 of Byelorussia . Therefore, we have assigned a zero value for each of the other 1/3 segments of Byelorussia. The final average is 1/3 of the value midway between the values for Bialystok and Oater. For measurements in Bialystok (data for April 29) , the appropriate value for Cs-137 % of total gamma dose rate is 1.8 %. For Oster , as stated above (for Hay 10 measurements), the appropriate value for Cs-13 7 % of total ga11111a rate is 16.0 %. These adjustments were dose -'-e"'m"e..;.ncct..cs.c ____ _ _____________ made before_ combini!!& the_ Bial~k and Oster meas-"ur-'.. ____ _ U.S.S .R. WHO Report Direct gamma dose rate data are provi ded by the Soviet Union for Kishinev, the capitol of Moldavian June 12, 1986 th e Moldavian Republic, start in g with Hay 10, 1986. The data used here are for Hay II, 1986. Republic Three additional values of zero ver e ass i gned for other quadrants of Moldavia where we have no measurements. There fore, the average value used in Method (2) calculations is 1/4 of the value for Kishinev. For measurements of Hay II, 1986, the appropriate value for Cs-137 % ------------------of total__aa111118dose rate is 17.3_%. ----- -------------------- U.S.S.R. WHO Report No really uaeful data for Cs-137 or gamma dose rates are provided for the Baltic Republics. Baltic June 12, 1986 But, data are available for direct gamma dose rates for si tes in Poland (Bialystok, Olsztyn) Republics bordering these Republi cs, for Southern Finland not far from the northern part of these Latvia republics, and from Sweden to the nortlwest of these Republics. Lithuania From all these data, a minimal esti,i;,te of 100 JJR/hr as the peak direct galDB dose rate Estonia has been here assigned to the Baltic Republics . This appears cautious and reaeonable. Further, to err on the aide of underestimation of cesium dose, ve ahall assign this value for April 29, -----------------198~ for vhi ch the aeeroeriate value_for Cs-137 % of total~-dose rate ta l.~ %~-- U.S . S.R. EPA Report Soiae values for dir ect gamaa dose rate are pro vided, atarting with data for Hay 5, 1986." Moscow June 11, 1986 We ahall used the Hay 5 data for indire ct estimate of Cs-137 deposition by Method (2). For and Hay 5 ¥easurementa, the appropriate value for Cs-137 % of total ga11111& rate is 9.1%. dose Suburbs --~----------------- 51 -Gofman COUNTRY SOURCE COMMENTS U.S.S.R. EPA Report Direct gamna dose rates are provided for May 2-May 7, 1986. The peak gamma dose rates Leningrad June 11, 1986 are reported for May 7 ,1986, and the se data are used in the indirect estimate of Cs-137 and by Method (2). For May 7 , 1986 measurements, the appropriate Cs-13 7 % of total gamma dose __ Suburbs ____ ______ ___ rate is 11. 7 _ % ¥ ----------------------------------------- U.S. S.R. No really satisfactory data are available which enable us to provide any estimates for Russian Cs-137 deposition in this largest of the Soviet Republics, aside from the data for Moscow Soviet and Leningrad, which are described abo v e. This is regrettable, since this Russian Republic ___ Reeublic _____ _______ __ __ is_not_onli_the_laE_Se&t g~raehicallrL but is_also_the most_eoeulous_of_the Soviet Reeublics. U.S.S.R. No data This are a very near the Chernobyl nuclear power plant had some very high dose s , since Chernobyl radiati on sickne ss and deaths have occurred there . Since no data have been made available _Reg!on _______ _________ for_this_seecial_regionL..no cancer calculation s_have been made. _______ _______ ______ _ U.S.S.R. No data have been provided for all these other Soviet Republics, nor are there any data All No data for regions cl os e by from whi c h sny reason a ble estimates of Cs-137 dep o sition can be made. __ Other_ Soviet Republi c s We_ther e fore refrain from making any cancer calculations for these Republics. __________ _ 52 -Gofman ¥ (B) DISTRIBUTION OF 1,990,000 CURIES OF CESIUM~l37, BY COUNTRY The tabulation below corresponds with the left-hand side of Table 6. 22 Countr:i DeEosition ~ECi[m} Area in meters Deeosition Total, in Curies Albania 1. 618 X 104 2.886 X 10lO 467 A;;stria 2. 346 X 105 8.417 X lOlO 19,746 Belgium 2.697 X 103 3.063 X lOlO 83 1011 X X Bulgaria 2.319 105 1. 113 25,810 Canada 0 .539 X 103 1 x 101) 539 4 11 Czechoslovakia 7.012 X 101.284 X 109,003 4 Denmark 2.022 X 104.324 X 10lO 874 11 Finland(~ area)* 3.358 X 105 1.692 X 10á56,817 4 11 France 7.821 X 105.491 X }0 42,945 105 11 Germany, West 2.319 X 2.495 X 1057,859 105 1011 Germany, East 2. 710 X 1.086 X 29,431 10 11 Greec e 4.045 X 103 1.325 X 536 Hungary 5. 529 X 104 9.340 X lOlO 5,164 Ireland 1. 753 X 103 7.055 X }0lO 124 4 11 Italy 3.911 X 103.024 X 10 11,8 27 3 11 Japan 1.079 X 103. 738 X }0 403 9.887 X 10lO 80 4 9 Korea, South 0.809 X 10) Luxembour g 1.618 X 102.590 X 1042 4 4.100 X 10lO 663 Net herland s 1. 618 X 10 Norway 1.160-x 105 1.627 X 10 11 18,87 3 5 10 11 Poland 3.493 X 103.139 X 109,6 45 }0 11 Romania 1. 038 -x 106 2.383 X 247 ,3 55 11 Spain 3. 506 X 103 5.067 X 10 1, 776 Sweden(~ area)* 6.68 8 X 105 2.258 X 101} 151,015 }05 1010 Switzerland 3.182 X 4.145 X 13,189 5 11 Turkey J.348 X 107.836 X 10105,629 4 1011 X United Kingdom 8 . 765 X 102.45 21,4 74 3 X 1012 UnJ.ted Stat es 0. 067 X 107.60 469 Ukraine 1. 262 :x 106 6.032 :x 10 11 761, 238 11 Byelorussia 9.628 X 105¥ 2.083 X 10200,551 Moldavia 1. 686 -x 105 3.370 X 1010 5,662 Baltic Republic s J.4 02 X }0 5 1. 742 :x 101I 24 , 423 1011 Yugoslavia 2.495 X 105 2.568 X 64,0 72 Moscow and Leningrad not computed, because area is so small Sum of All Depositions, in Curies 1,987,784 (0.JS)(.0.75)(.).53 x 106) m ---------- ----1,985,625 .. .. See Technical Appendix 2-A Comparison with cesium -137 deposition from weapons fallout: 4 According to UNSCEAR(p.146), the deposition of ces i um-137 in the temperate latitudes of the northern hemisphere from all the atomspheric nuclear bomb-tests 5 of the United States, Soviet Union, and Britain combined was 136,000 or 1.36 x 10 picocuries per square meter. 53 -Gofman ¥ (C) FROM CESIUMS TIME-DISTRIBUTION FOR DOSE COMMITMENT Cesium-134 with its half-life of 2.3 years will deliver its committed dose to exposed populations very much earlier th an is the case for cesium-137, with its half-life of 30.2 years. Calculations below show what fraction of the total dose commitment (over all time, from the cesiums combined) is delivered by the end of each decade following the accident . To calculate, we used the observations (from Section 7 of this paper) that Cesium-134 (internal+ external) acc ounts for 11% of the total dose from cesiums; Cesium-137 (internal+ external) accounts for 89% of the total dose from cesiums; and of the 89%, the internal share is 30% and the external share is 70%. ¥ First Decade CESIUM-1 34 will deliver 94.6 % of both its internal and external doses; this amounts to (0.946) x (11 %) = 10.4 % of the total dose from cesiums. CESIUM-137 will deliver approximately 95% of its internal dose in the first decade; this amounts to (0.95) x (0.30) x (89%) = 25.4% of the total dose from cesiums. CESIUM-137 will deliver 20.0 % of its external dose in the first decade; this amounts to (0.20) x (0.70) x (89 %) = 12.5 % of the total dose from cesiums. COMBINED DELIVERY (%) BY THE END OF THE FIRST DECADE= 10.4 + 25.4 + 12.5 = 48.3% of total. ¥ Second Decade CESIUM-134 will deliver 5.4 % of (11%) = 0.59 % of the total dose. CESIUM-137 (internal) will deliver 5% of (0.30)(89 %) = 1.34 % of the total dose. CESIUM-137 (extern a l) will de liver 16% of (0.70)(89 %) 10.0 % of the total dose. COMBINEDDELIVERY(%) BY THE END OF THE SECOND DECADE= 48.3 + 0.59 + 1.34 + 10.0 = 60.2 % of the total dose committed. ¥ Third Decade The only new contribution will be from external cesium-137 because int e rnal contributions from the cesiums are essentially over . CESIUM-1 37 will deliver 14% of (0.70)(89 %) = 8.7 %. COMBINEDDELIVERY (%) BY THE END OF THE THIRD DECADE= 60.2 + 8 .7 = 69% of the total dose committed from the cesiums . ¥ Fourth Decade Additional contribution from external CESIUM-137 is 10% of (0.70)(89%) = 6.2% of the total dose. COMBINEDDELIVERY(%) BY THE END OF THE FOURTH DECADE 69 + 6.2 = 75.2% of the total dose committed from the cesiums. 54 - Gofman References (J) University of California at Berkeley (2) Gofman, J.W.; Tamplin, A.R. "Low Dose Radiation and Cancer" I.E.E.E. Transactions on Nuclear Science, Part 1, 1970, NS-17, 1-9. (3) Gofman, J.W. Radiation and Human Health; Sierra Club Books: San Francisco, 1981 .. (4) UNSCEAR (United Nations Scientific Committee on the Effects of Atomic Radiation) 1977; Sources and Effects of Ionizing Radiation Report to the General Assembly, with annexes. United Nations, New York. (5) UNSCEAR 1982; Ionizing Radiation: Sources and Biological Effects Report to the General Assembly, with annexes . United Nations, New York. (6) BEIR-III (The Advisory Committee on the Biological Effects of Ionizing Radiation) 1980; (Final Report); The Effects on Populations of Exposure to Low Levels of Ionizing Radiation; National Academy of Sciences (Typescript Edition), Washington . (7) Gofman, J.W. The Cancer -Leukemia Risk from Ionizing Radiation: Let's Have a Closer Look A.A.A.S. Symposium Presentation, Detr oit, Michigan May 30, 1983. (Copies available from author, P.O.B 11207, San Francisco, Ca. 94101). (8) Radford, E.P. "Radiogenic Cancer in Underg round Miners" In Radiation Carcinogenesis: Epidemiology and Biological Significance; Boice, J.D. Jr., J.F. Fraumeni, Eds.; Raven Press, New York, 1984, 225-230. (9) Wagoner, J .K. "Leukemia and Other Malignancies Following Radiation Therapy for Gynecological Disorders" In Radiation Carcinogenesis: Epidemi ology and Biolog ical Significance; Boice, J.D. Jr., J.F. Fraumeni, Eds. ; Raven Press, New York, 1984, 153-159. (10) Radford, E.P . In A Roundtable: with radiation, how little is t oo much? New York Times, Week in Review, Section 4, p. Ey 19. (11) Report of the National Institutes of Health Ad Hoc Working Group to Develop Radioepidemiological Tables; National Institutes of Health, Publication No. 85-2748, 1985, 15. (12) Wakabayashi, T.; Kato, H.; Ikeda,T.; Schull, W.J. "Studies of the Mortality of A-Bomb Survivors, Report 7, Part III. Incidence of Cancer in 1959-1978, Based on the Tumor Registry, Nagasaki" Radiation Research 1983, 93, 112-146. (13) Boice,J.D. Jr.; Land, C.E . ; Shore, R.E.; Norman, J.E.; Tokunaga, M. Radiology 1979, 131, 589-597. (14) BEIR-I; (See above re BEIR-III) . This report is for 1972. (15) Court-Brown, W.M.; Doll, R. "Mortality from Cancer and Other Causes a fter Radiotherapy for Ankylosing Spondylitis" British Medical Journal 1965, l, 1327-1332. (16) Smith, P .G.; Doll, R.; "Age-and Time-dependent Changes in the Rates of Radiationinduced Cancers in Patients with Ankylosing Spondylitis Following a Single Course of X-ray Treatment" In Late Biological Effects of Ionizing Radiation, Vol 1; International Atomic Energy Agency : Vienna, 1979 . (17) Smith, P.G. "Late Effects of X-ray Treatment of Ankylosing Spondylitis" In Radiation Carcinogenesis: Epidemiology and Biological Significance; Boice, J.D. Jr., J.F. Fraumeni, Eds. ; Raven t'ress, New York, 1984, 107-118. 55 -Gofman (18) Kato, H.; Schull , W. J. "Studi es of the Mortality of A-Bomb Survivors . 7. Mortality, 1950-1978: Part 1. Cancer Mortality" Radiation Research 1982, 90, 395-432. (19) Schull, W.J. "Atomic Bomb Survivors: Patterns of Cancer Risk" In Radiation Carcinogenesis: Epidemiology and Biological Significance; Boice, J.D. Jr., J.F. Fraumeni, Eds. ; Raven Press, New York, 1984, 21-36. (20) Associated Press. "Pravda Sneers at Oregon's Bill" as carried in San Francisco Examiner, August 10, 1986, A-16. (21) Natarajan,A.T.; Csukas,I.; Degrassi, F.; van Zeeland, A.A.; Palitti, F.; Tanzarella, C.; de Salvia, R.; Fiore, M. "Influence of Inhibition of Repair Enzymes on the Induction of Chromosomal Aberrations by Physical and Chemical Agents" In Progress in Mutation Research, Vol. 4 A.T. Natarajan et al, Eds.; Elvesier Biomedical Press, Amsterdam, 1982, 47-59. (22) Kihlman, B.A.; Natarajan, A.T. "Potentiation of Chromosomal Alterations by Inhibitors of DNA Repair" In DNA Repair and Its Inhibition Collins,A., Downes, C.S., R.T. Johnson, Eds.; IRL Press, Oxford, 1984, 319-33 9. (23) Riccardi, V.M.; Sujansky,E.; Smith, A.C.; Francke, U. "Chromosomal Imbalance in the Aniridia -Wilm's Tumor Association: 11 p Interstitial Deletion" Pediatrics, 1978, g, 604-610. (24) Zellweger, H.; Simpson,J. "Chromosomes of Man" (Clinics in Developmental Medicine Nos. 65/6?),Spastic International Medical Publications, J.B. Lippincott, Philadelphia, 1977, 153-156. (25) Cohen, A.J.; Li, F.P.; Berg, S.; Marchetto, D.J.; Tsai, S.; Jacobs, S.C.; Brown, R.S.: "Hereditary Renal-cell Carcin~a Associated with a Chromosomal Translocation" New England Journal of Medicine 1979, 301, 592-595. (26) Croce, C.M. "Chromosomal Translocations, Oncogenes, and B-cell Tumors" Hospital Practice January 15,1985, 20, 41-48, (27) Bender, M.A.; Preston, R.J. "Role of Base Damage in Aberration Formation: Interaction of Aphidicolin and X-rays" In Progress in Mutation áResearch, Vol. 4 A.T. Natarajan et al, Eds.; Elvesier Biomedical Press, Amsterdam, 1982, 37-46. (28) Preston, R.J. "The Effect of Cytosine Arabinoside on the Frequency of X-rayInduced Chromosome Aberrations in Normal Human Leukocytes" Mutation Research, 1980, .¤2., 71-9. (29) Gofman, J.W.; Tamplin, A.R. "Epidemiol ogical Studies of Carcinogenesis by Ionizing Radiation" In Proceedings of the Sixth Berkeley Symposium on Mathematical Statistics and Probability: Volume VI: Effects áof Pollution on Health Lecam, L.M., Neyman, J., E. Scott, Eds.; University of California Press, Berkeley, 1970, 235-277. (30) Baver stock, K.F.; Papworth, D.; Vennart, J. "Risk of Radiation at Low Dose Rates" Lancet 1981, _! (Feb. 21), 430-433. (31) Baverstock, K.F.; Vennart, J. "A Note on Radium Body Content and Breast Cancers in U.K. Radium Luminisers" Health Physics , 1983, 44, Suppl. No.1, 575-577. 56 -Gofman (32) Myrden, J.A.; Hiltz, J.E. "Breast Cancer Following Multiple Fluoroscopies During Artificial Pnemnothorax Treatment of Pul monary Tuberculosis" Canadian Medical Association Journal, 1969, 100, 1032-1034. (33) Modan, B.; Ron,E.; Werner, A. "Thyroid Cancer Following Scalp Irradiation" Radiology 1977, 123, 741-744. (34) Boice, J.D. Jr.; Monson, R.R. "Breast Cancer in Women After Repeated Fluoroscopic Examinations of the Chest" Journal of the National Cancer Institute, 1977, 59, 823-832. (35) Stewart, A.M.; Webb, J .W.; Giles, B.D.; Hewitt, D. "Preliminary Communication: Malignant Disease in Childhood and Dia gnostic Irradiation in-Utero" Lancet, 1956, ±_, 447. (36) Stewart, A.M.; Webb,J.W.; Hewitt, D. "A Survey of Childhood Malignancies" British Medical Journal 1958, 2, 1495-1508. (37) Stewart, A.M.; Kneale, G.W. "Radiation Dose Effects in Relation to Obstetric X-rays and Childhood Cancers" Lancet, 1970, _!_, 1185-1188. (38) MacMahon, B. "Pre-natal X-ray Exposure and Childhood Cancer" Journal of the National Cancer Institute ,1962, ~. 1173-1191. (39) Newcombe, H.B. ; McGregor, J .F . "Childhood Cancer Following Obstetric Radio Graphy" Lancet, 1971, 1, 1151-1152. (40) Holford, R.M. "The Relation Between Juvenile Cancer and Obstetric Radiography" Health Physics, 1975, 28, 153-156. (41) Mole, R.H. "Antenatal Irradiation and Childhood Cancer; Causation or Coincidence?" British Journal of Cancer 1974, lQ., 199-208. (42) Mole, R.H. "Radiation Effe c ts on Pre-natal Development and Their Radiological Significance" British Journal of Radiology, .1979, 52, 89-101. (43) Monson, R.R.; MacMahon, B. "Prenatal X-ray Exposures and Cancer in Children" In Radiation Carcinogenesis: Epidemiology and Biological Significance; Boice, J.D. Jr., J.F. Fraumeni, Eds .; Raven Press, New York, 1984, 97-106. (44) Beebe, G.W.; Kato, H.; Land, C.E. "Studies of the Mortality of A-bomb Survivors: 6. Mortality and Radiation Dose, 1950-1974" Radiation Research 1978, ..?2_, 138201. (45) Tokunaga, M.; Land, C.E.; Yamamoto, T.; Asano, M.; Tokuoka,S.; Ezaki, H.; Nishimori,!. "Breast Cancer in Japanese A-Bomb Survivors" Lancet 1982, ii, (Oct.23), 924. (46) Gofman, J.W.; O'Connor, E. X-rays: Health Effects of Common Exams; Sierra Club Books, 1985, 282-297. (47) Hoffman, D.A.; Radford, E.P. A Review of the Carcinogenic Effects of Low Doses of Ionizing Radiation; Three-Mile Island Public Health Fund, Philadelphia, 1985(April), Table 2, p.124. (48) World Health Organization" Updated Backg round Information on the Nuclear Reactor Accident, (Chernobyl) USSR" 1986, (1) May 30, 1986, (2) June 5, 1986, (3) June 12, 1986. Copenhagen, Denmark. Contact Person: J.I. Waddington. (49) United States Environmental Protection Agency "Chernobyl Radiation Data Summary" 1986, A series issued between May 8, 1986 and June 30, 1986. Contact Person: H. Michael Mardis, Environmental Studies and Statistics Branch, ASD (ANR-461), Office of Radiation Programs, U.S.E.P.A., Washington,D.C. 20460. 57 -Gofman (50) Finnish Centre For Radiation and Nuclear Safety "Interim Report on Fallout Situation in Finland from April 26 to May 4, 1986" STUK-B-VALO 44, May, 1986. PL-P.O. Box 268, SF-00101 Helsinki 10, Finland. (51) Finnish Centre For Radiation and Nuclear Safety "Second Interim Report on Radiation Situation in Finland from 5 to 16 May 1986" STUK-B-VALO 45, May, 1986. Address as in Reference (50). (52) Fry, F.A.; Clarke, R.H.; O;Riordan, M.C. "Early Estimates of UK Radiation Doses" Nature, 1986, 321, 193-195. (53) Devell, L.; Tovedal, H.; Bergstrom, u. ; Appelgren, A.; Chyssler,J.; Anderson,L. "Initial Observations of Fallout frOll) the Reactor Accident at Chernobyl" Nature, 1986, 321, 192-193. (54) Knox, J.B. "Description of ARAC 1-131 and Cs-137 Deposition and Dose Calculations: Preli1llinary" May 7, 1986. From Livermore National Laboratory, California. (55) Apsimon,H.M.; Wilson, H.N. "Preliminar y Analysis of Dispersion of th e Chernobyl Release", Paper given at the (U. K.) Nuc lear Insp ec torate on 20th May, 1986 and Helen ApSirnon, letters to Frank von Hippel, June 19, 22 , 1986. (Cited in v0n Hippel, F. ; Cochran , T.B in an article in pre ss , September, 1986, Bulletin of th e Atomic Scientists . ) (56) Hohenemser, C.; Deicher, M. ; Ernst,A.; Hofsass, H.;, Lind er , G.; Recknagel,E. "Chernobyl: An Early Report" Environment .1986, ~. (June) 6-43. (57) Hammersen, F. Sobotta/Hammersen Histology; Color Atlas of Microscopic Anatomy; Urban and Schwarzenberg, Baltimore, 1985. (58) Elias, H.; Paul y , J.E.; Burns,E.R. Histolo gy and Human Microanatomy; John Wile y & Sons ; New York, 1978. (59 ) Gardner, D. L. ; Dodds, T.C. Human Hist ology : An Introdu c ti on to the Study of Histopatholog y ; Churchill Livingstone, Edinburgh, 1976 . (60) Johannessen, J.V. (Editor); Electron Microscopy in Human Medicine; In 12 Volumes, through 1985; McGraw-Hill Internation al Book Co., New York. (61) Tokunaga, M.; Land, C.E.; Yamamoto, T.; Asano, M.; Tokuoka , S.; Ezaki, H. ; Ni shimora, I.; Fujikura, T. "Br e ast Cancer Among Atomic Bomb Survivors" In Radiation Carcinogenesis: Epidemiology and Biological Signi f icance; Boice, áá J.D. Jr., J.F. Fraumeni, Eds.; Raven Press, New York, 1984, 45-56. (62) Luckey, T.D. "Physiological Benefits from Low Level s of Ionizing Radiation" Health Physics, 1982, 43, 771-789. (63) Loewe, W.E.; Mendelsohn, E. "Neutron and Gamma Ray Doses at Hiroshima and Nagasaki" Nuclear Science and Engineering,1982, .!!..!_,325-350. (64) McGregor, D.H.; Land, C.E.; Choy, K.; Tokuoka, S.; Liu, P.I.; Wakabayashi, T.; Beebe , G.W. "Breast Cancer Incidence Among Atomic Bomb Survivors, Hir os hima and Nagasaki, 1950-1969" Journal of the National Cancer Institute, 1977, 59, 799-811. ~ (65) State Committee For Using the Atomic Energy Of the U.S.S.R. The Accident At the Chernobyl AES and Its Consequences; Data Prepared For the International Atomic Energy Agency Expert Conference 25-29 August 1986 in Vienna. ¥ BIOGRAPHICAL INFORMATION ABOUT THE AUTHOR¥ John Gofman is Professor Emeritus of Medical Physics at the University of California at Berkeley, and lecturer at the Department of Medicine, University of California School of Medicine at San Francisco. Author of Radiation and Human Health, 1981. He is the author of approximately 150 scientific papers in peer-review journals in the fields of nuclear/physical chemistry, coronary heart disease, ultracentrifugal analysis of the serum lipoproteins, the relationship of human chromosomes to cancer, and the biological effects of ionizing radiation with particular reference to cancer and leukemia induction. Since 1980, he has been serving as a peer-reviewer for Health Physics journal. While a graduate student at Berkeley, he co-discovered Protactinium-232 and Uranium -232, Protactinium-233 and Uranium-233, and proved the slow and fast neutron fissionability of U-233. He worked during the war on the Manhattan Project atomic bombs, and developed some of the first methods for chemically extracting plutonium from irradiated uranium. After graduating from medical school, he began his r esearch on coronary heart disease and, by developing special flotation ultracentrifugal techniques, demonstrated the existence of high-density and low-density lip oproteins . His work on their chemistry and health consequences has been widely honored, in cluding the St ouffer Prize (shared) in 1972 for outstanding contributions to research in arteriosclerosis, and selecti on in 1974 by the American College of Cardiology as one of the twenty-five leadin g resear chers in cardiology of the past quarter -century . Meanwhile, the Atomic Energy Commission asked him to establish the Biomedical Research Division at the Lawrence Livermore National Laboratory. From 1963 through 1969, he serv e d as Associate Director of the Laboratory, and as the first Director of its Biomedical Division, where he also carried out his own laboratory research in cancer, chromosome s , and radiation, as well as his analyses of the data on the Japanese atomicbomb survivors and other irradiated human populations. In 1969, he published his finding (with Dr. Arthur Tamplin) that exposure to ionizin g radiation is far more serious than previousl y recognized, and stated the first three "laws" of radiation carcinogenesis . In 1973, he returned to full-time teaching at the University of California at Berkeley, and continued independently to analyze the accumulatin g human evidence of health effects -from radiation in the low-dose region. His book RADIATION AND HUMAN HEALTH (1981; still i n print) integrated the existing worldwide evidence for the first time. It has been described as a "remarkable and important book" by the Journal of the American Medical Association (19 March 1982), and recommended by the journal of the American Nuclear Society Nuclear News (January 1982) as "n ot a tome to be treated casually, since there is a lot of material here which should be read carefully and given thought and evaluation." His most recent book (1985) X-RAYS: EXAMS (with O'Connor) has HEALTH EFFECTS OF COMMON been selected by the Library Journal (15 April 1986) as one of the most useful and important reference books published during al l of 1985. The New England Journal of Medicine (6 February 1986) said, "This book is practical and important. It is destined to represent a watershed in the controversial field of low-dosage radiobiology and will be of inestimable value to radiologists, other physicians, dentists, and patients." Addresses: 102 Donner Lab, University of California, Berkeley, Calif. 94720, USA, or via Main Post Office Box 11207, San Francisco, Calif. 94101, USA~ Tel: 415-664-1933.