ML20133D652
| ML20133D652 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 11/30/1984 |
| From: | Standefer F, Standerfer F GENERAL PUBLIC UTILITIES CORP. |
| To: | Snyder B Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20132B167 | List: |
| References | |
| FOIA-85-285, FOIA-850285 0133A, 133A, 4410-84-L-0216, 4410-84-L-216, NUDOCS 8507220264 | |
| Download: ML20133D652 (150) | |
Text
--9 f
I '. e s
p ght77 "c
67 s 2,,M h, DO
^
b\\
A GPU Nuclear Corporation J
- g gf Post Office Box 480 Route 441 South Middletown, Pennsylvania 17057 0191 h
717 944 7621 TELEX 84 2386 Writer's Direct Dial Number:
}$.
(717) 948-8461 4410-84-L-0216 Docm ent ID 0133A November 30, 1984 THI Program Office Attn: Dr. B. J. Snyder Program Director US Nuclear Regulatory Commission Washington, DC 20555
Dear Dr. Snyder:
Three Mile Island Nuclear Station, Unit 2 (TMI-2)
Operating License No. DPR-73 Docket No.53-320 Aamodt Study / Affidavits Attached for your information are GPU Nuclear internal memoranda concerning the "Aamodt Motions for Investigation of Licensee's Reports of Radioactive Releases During the Initial Days of the TMI-2 Accident...".
These memoranda suggest a possible causal relationship between the agent lithi m and the symptoms experienced by those residents interviewed by the Aamodts. As noted by GPU Nuclear Memorandum R&EC 84-314, an assessment of the possible relationship between the residents' symptoms and lithi m toxicity has been conducted. Based on the results of that assessment, GPU Nuclear has concluded that the amount of lithis which
'could have been dispersed from THI-2 was insufficient to cause the symptoms cited in the Aamodt's affidavits. We are continuing to inquire into the matter of the Aamodt Motions and will provide any further information on this subject as it develops.
If you have any questions concerning this information, please call Mr. J. J. Byrne of my staff.
Sincerely, 8507220264 850524 f.$GE O 85 PDR R. Standerfer Vice President / Director, TMI-2 FRS/RDW/jep Attachments cc: Deputy Program Director - TNI Program Office, Dr. W. D. Travers Z6 GPU Nuclear Corporatlor,is a subsidiary of the General Public Utilities Corporation s.
Nuslear memorandum
Subject:
Response to Your 11/2/84 Memorandum Date:
November 9,1984 on Lithium R&EC 84-314 (0073U)
From:
Vice President, R&EC Division Location:
HQ R. W. Heward, Jr.
I To:
Manager, Radiological Health - U. H. Behling pd O *'
C Responding to your memoranda of 11/2/84 and 11/6/84 on the possibility that releases of Lithium from TMI-2 caused the symptoms described in the Aamodts' petition I want So provide some results of the Company's findings in this area. The excerpt you provided from the Physicians' Desk Reference (PDR) describing the symptoms associated with Lithium carbonate (Li CO ) oral
~
2 3 ingestion, clearly match some of the symptoms described in Affidavits 1, 2, 4, 5, 6 and 7.
As we discussed on 11/5/84, we must examine the possible plant and non plant sources, the meteorology and dispersion mechanism, and other data sources to corroborate your concern that Lithium may be the causative agent and that THI-2 may be the source of the Lithium.
As a start on this, we have determin'ed that the TMI-2 Reactor Coolant System (RCS) has Lithiuto Hydroxide (LiOH) added at plant start-up to neutralize boric acid. *When the plant is operating, pH control is achieved via addition of Sodium Hydroxide (NaOH). The, RCS had 0.85 ppa Littiium Hydroxide (LiOH) on 3/27/79, and 4.64 ppm on 4/11/79. The latter number was due to Lithium being removed from the demineralizers by the addition of NaOH.
Given the volume of RCS, and the normat range of LiOH in the RCS of 0.2 to 2.0 ppe, we have calculated the RCS generally contains approximately 1 to 2 pounds of LiOH.
i.
The PDR inaicates that an oral dose of 1800 mg of Lithium carbonate (Li CO ) yields blood serum levels of 1.0 to 1.5 mEq/1.
Based on this 2 3 fact, a.04 to.06 conversion factor can be calculated for Li CO. Assum-2 3 ing this conversion factor applies to LiOH and using your number of 210 mg of LiOH for the toxic body burden, one would have to drink a substantial portion l
of the RCS to reach toxic levels. Thus it appears the RCS could not be a source for LiOH in the affected individuals since transport and dispersion would significantly reduce the amount of LiOH available for uptake.
l l
The possibility of elemental Lithium having been added to the TM1-2 l
systems to consume the hydrogen bubble during the accident has been checked.
We have learned the hydrogen bubble was eliminated by letdown and pres-surizer venting -- no elemental Lithium was used.
l Based on our discussions with Jack DeVine and Ed Fuhrer and your 11/6/64 l
memo, we believe an upper bound for the amount of LiOH in the TMI-2 demineralizers is approximately 40 to 42 pounds. Work is underway to determine whether this material could have been released from the TMI-2 i
domineralizers during the accident -- the release would have gone to the waste l
decay tank and then through roughing, HEPA and charcoal filters before going out the station vent.
i A00o064s s-s3 m
's.,, _,
1 T):U. M. Behling Noveabar 9,1984 I
While it is unlikely that all 40 pounds could have been discharged, we examined the possible source terms which would have been necessary to achieve toxic levels in the environment. Based on the 2mg/m3 LiOH concentration in air being the threshold toxic concentration in humans (The Hazards of 1
Dangerous Industrial Chemicals, 1. R. Sax, 1975) and assuming.the dispersion characteristics in existence during the, days in questi,on in 1979 Werner Heck has calculated (see Attachment 1) that a source term of between 250 pounds / hour and 79,000 pounds / hour would be required. This calculation also assumed the LiOH was a gas' and that the individual was at the center line of the plume.
A second dispersion calenistion was done which assumed 46 ag/l of LiOH in i
blood serum for the toxic body burden (most sensitive individual),3conversion factor of 0.04, 3 liters of blood serum, respiration rate of 0.9 m / hour and exposure time of I hour for the woman in Affidavit No. 6 and 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> for the men in Affidavit No. fl.', This calculation yields source terms of between l
440,000 pounds / hour and,140,000,000 pounds / hour.
Thedispersioncalcu5ationstreatedLiOHasa. gas. The fact that during some of the days in question rain was falling, makes direct application of the gaseous dispersion difficult.
If rain was falling throughout the area when a possible release of LiOH occurred, the deposition would be greater near to the plant and less at distance. Because of the different days involved in the affidavits, and the complexity of the meteorology, terrain, etc., more refined i
i dispersion calculations are not practical. ' Based on the amount of material available at THI-2 for" dispersion, the wind speed and direction data extant during the days in ques' tion, ana the ambunt necessary to cause toxicity, it i
does not seen LiOH from TMI-2 is a likely candidate to have caused the symptoms of the residents cited by the Aamodes.
Efforts continue to determine if possible other causative agents in plant processes as well as alternative sources of Lithium or other compounds could produce the same symptoms. We have artracted a list of area industries which may use Lithium and are pursuing the locations of these versus the meteorological conditions. We are also searching a toxicology data base for a match on symptoms from other chemical and pharmaceutical compounds. Your continued effort in this search is appreciated.
1 RWH/MBR/brh R. W. Heward, Jr. "
CC: Manager, Env Con - TMI - C. C. Baker (w/ attach.)
2 Recovery Engineering Director - J. C. DeVine, Jr. (w/ attach.)
Rad Con Director - THI J. E. Hildebrand (w/ attach.)
Executive Vice President - E. E. Kintner (w/ attach.)
Safety and Env Con Director - M. B. Roche (w/o attach.)
- o.
- e ATTACHMENT 1 1
l Stability Class Distance X/D X
Source Term (km)
(s/m3)
(g/m3)
(g/s)
(1bs/hr)
A 5
4.BE-7 2E-3 4.2E3 3.3E4 2.BE-7 2E-3 1.BE4 7.9E4 15 B
5 6.7E-7 2E-3 3.SE3 2.4E4 15 -
2.6E-7 2E-3 7.7E3 6.1E4 C
5 3.3E-6 2E-3 ~
6.1E2 4.BE3 15 6.9E-7 2E-3 3.3E3 2.6E4 D
5 1.4E-5 2E-3 1.4E2 1.1E3 3.3E-6.'
. 2E--3 6.1E2 4.BE3 15
~
E 5
3.4E-5 2E-3 5.9E1 4.7E2 15 1.SE-5 2E-3 2.BE2 1.6E3
-V F
5 9.3E-5 2E-3 2.2E1 1.7E2 15 2.7E-5 2E-3 7.4E1 5.BE2 6
5 1.7E-4 2E-3 1.2E1 9.5El 15 6.2E-5 2E-3 3.2E1 2.5E2 e
e
^-
K..
Nuslear uomorandum l
l
Subject:
Aamodt Study / Affidavits Date:
November 2, 1984, 1984 1
i From:
Manager,* Radiological Health Location: TMI-2 U. H. Behling To:
Vice President Radiological & Environmental Controls - R. W. Heward I
G In pursuit of explaining the frequently reported observation of metallic taste and associated symptoms with the first 3-4 days of the TMI accident. I have stumbled across something which may have significant implications.
In a previous confidential communication I had referenced Iodine / Iodide as a potential source for metallic taste and a limited nunber of associated effects which, however, are restrictive to iodine hypersensitive individuals.
As was obvious from the start, the concentrations necessary to elicit any reaction, even among the few hypersensitive individuals, would have required astronomical quantities / radioactivities of iodine. Therefore, this explanation was never a viable theory worth considering other than to i
recognize that others might (i.e., Seo Takeshi, who claims that at least 64,000C1ofIodinewerereleased).
However, in trying to identify an alternate agent /cause for not only the metallic taste, but also the complex clinical signs and symptoms described by individuals (submitted as affidavits in the Aamodt Report). I have identified an agent which in definitive terms parallels the complex clinical picture contained in the Aamodt affidavits. Moreover, this material is used in potentially sufficient quantities at TMI and at the time of the accident may have been released in quantities to support sqy suspicion. The conditions of release may have involved the purging and/or venting to the atmosphere of select systems / portions of the primary loop (not necessarily by Containment / Reactor Building air). The agent in question is Lithium. Lithium hydroxide (L10H) is added in stoichiometric q)uantities to the primary co in order to neutralize the boron (boric acid. Throughout the first few days of the accident, loss of coolant and replacement of coolant required additions of boric acid and lithium hydroxide.
Furthermore, under the elevated conditions of core / coolant temperatures, LION in the presence of strong annnnaag a 33
\\
- , =:
l 3
- November 2.1984 reductants (Zircaloy cladding and in the' presence of the H2 that was fomed) can be reduced to lithium hydride (L1H). Without having concrete knowledge, there might also have been elemental lithium added for the purpose of consuming the hydrogen bubble (i.e., elemental L1 reacts with H2 gas to form L1H which is a solid powder which in the presence of H 0 is oxidized to 2
L10H).
In sumary, there exists a potential for lithium (L1+DH';LiH) to have been released by a yet undetemined pathway into the atmosphere. A cursory inspection of venting times coincides with reported cases of metallic taste.
I an enclosing a description of the adverse clinical signs and symptoms associated with the oral and controlled administration of lithium (LiCO )
3 for the treatment of manic depression (PDR).
Keep in mind that exposure for the proposed conditions differs from the oral administration of LiC0 -
3 Additionally if L10H was released its chemically dissociation into L1+ and OH-in H O has the concurrent effect of a strong base (hence the " burning" 2
sensation in lungs, etc). In reviewing the highlighted portions, it will be evident that the clinical signs and symptoms do, in fact. coincide in detail with those identified in the affidavits. They are:
o nausea o vomiting o dehydration o diarrhea o anorexia o muscular weakness o drying of skin (xerosis cutic) o thinning / loss of hair o blurred vision o skin itching (prunitis) o cardiac arrythmia o kidney (renal) failure o circulatory collapse (of organs) o hypothyroidism o teratogenesis o metallic taste A review of the enclosed affidavits 1-8 will identify each of the above described symptoms in either/or both human and animal subjects.
Furthermore, the levels of human toxicity which elicit these reactions occur at relatively low levels (approximately 3 meq/L). These concentrations might readily be obtained under conditions of discrete venting and primary plume dispersion for select areas of plume touchdown.
I believe that there is sufficient cause for concern that lithium may have been the agent responsible for the cited conditions. There is agrwement in both_ time and meteorological conditions to suspect that an airborne agent was involved.
I feel reasonably certain that the symptoms described in the Aamodt affidavits were genuine. I an even more certain that they have nothing to do
- - ~ - -
-~-
N
. s.:>,
3-November 2, 1984 with radiation exposure (inasmuch as they do not fit any patterns of the acute radiation syndrome in dose / time / symptoms, etc.).
It is also sy feeling that the "Beyes Committee" feels that the Aamodt affidavits have merit but, at the same time, concur that radiation exposure was not the cause (see enclosure /
excerpt frun the Beyea Report, pages 38-39).
Because of the sensitive and confidential nature of this possibility. I will await your comments and further instruction in exploring this matter.
Ulrich H. Behling PhD Extension 8582 UNB/ pat enclorures t
e 9
.--n-
- .y
~.
$ Nuclear memorandum
Subject:
Source Terms for L10H for TMI Date:
November 5. 1984 From:
Manager. Rad Health Location: 1MI-2 U. H. Behling 9210-84-1720 T2:
Radiological Controls Director. TMI-2 J. E. Hildebrand Vice Presidcnt. Radiological & Environmental Controls R. W. Howard BETA. Inc. - M. E. Miles 1.
Primary Coo' ant:
For L10H. the concentration of the primary coolant was limited to H ppm which is equivalent to 2 mg/ liter.
does not represent a likely source term due to low con (centration.)This, the 2.
Make-uo purification Resins: These resins were subject to temperatures of up to 3500 C.
of approximately 108 rads.Thest resins were additionally exposed to radiation dose resin for a total volume of 40 cubic feet of main.The content of lithfue was 1 m The density of resin is.7 g/ml. The internal source term can be calculated as follows:
1 1.
Iotal Itasin Volume: 40 cubic ft. = 1132 liters i
1132 liters = 1.132.000 al 2.
Total Weicht = 1.132 x lo a1 x.7 g/m1 s
= 7.92 x 105 g
(
i 3.
Total Quantity of Lf0N/40 cubic ft. of resin l
= 7.92 a 105 g x 1 meg LION /g l
- 7.92 x 105 ang 7.92 x 105 meg. 792 equivalents 1 equivalent g = 5.939 g t
i i
akev.us a.es g.at d118 Illi UdlWil WI
s
'i Therefore:
792 equivalentsti = 5495 g of Li IL1 + 2H 0 -+ 2L10H + H i
~
2 2
l l
Therefore: 792 equivalents or 5495 g of Li will produce 18.216 g of L10H Total source term for L10Hf 18.216 Ke Human toxicity levels (serum) for L10H are 3 3 ang/ liter.
Total serum quantity / standard man + 3 If ters.
Total body burden of L10H needed to induce toxic syndrome:
(3) x (23 ag) W h 11ter.
70 mg/L x 3L = 210 as total body burden.
h.
U. H. Behling Extension 8582 UH8/ pat O
.. 7l, r=
v
=r-y-
m-ete.
s-m--yg
- w-e
-e-s---tvr'-s-
+--=-e9-
>-==e-r-+erv-----r-
w--e-
-.ee
--h-=m err--we--w-ww---
--w-- - - - - -
nw---*
--s--+
+ - -
--w-w--w-*
THREE MILE ISLAND J E9' i,
Pusuc HEALTH FUND o
/
/
'888 ' mat sw g
(
- _,..a mis)s7s sooo mm
/
M 1 9 8WP5. M March 25, 1985 HAND DELIVERY John G. Harkins, Jr., Esquire Fidelity Building 20th Floor Philadelphia, PA 19109
Dear Mr. Harkins:
Please find enclosed a draft copy of a report entitled, "A Review of the Carcinogenic Effects of Low Level Ionizing Radia-tion," authored by Daniel A. Hoffman of the National Cancer Institute and Edward P. Radford, chairman of the'BEIR III committee of the National Academy of Sciences.
This report is a literature review C(2.'
commissioned by the Public Health Fund of the human health effects of low level radiation and was referred to in.the Fund's letter of invitation to the upcoming conference.
The report is still in draft form and is not for quotation or dissemination.
The report, along with two discussion papers on the subject of potential study populations to be circulated shortly, is being made available to stimulate discussion at the conference.
The Public Health Fund is hopeful that the conference will result in a constructive informal discussion to assist the Fund in developing a research program to recommend to the court supervising the Fund.
Invitees are not requested to give formal papers or presen-tations, but to contribute their expertise to the discussion.
- However, should any invitee wish to submit written suggestions for research, the Fund will be glad to receive and circulate them to the conference participants.
I look forward to the conference and a stimulating exchange of ideas.
Please refer any logistical questions to Linda Cooper, the conference coordinator.
Sincerely yours,
\\
/4
~
s s
N nathan Berger, P
~23 xecutive Secretary Public Bealth Fund v=:r3v Enclesures
l LIST OF PARTICIPANTS Dr. Philip Archer Dr. Victor Archer /
Dr. Olav Axelson Dr. Baruch Blumberg 3 Dr. Victor Bond Mr. Stephen Chinn 1 Dr. Sidney Cobb Dr. Ted Colton
/
Dr. Nancy Dreyer Dr. John Gofman Dr. Michael Gould Dr. Maureen Hatch Dr. Daniel Hoffman Mr. Seymour Jablon Dr. Carl Johnson Dr. George Kneale Dr. Charles Land Dr. Joseph L. Lyon Dr. Thomas Mancuso Dr. Baruch Modan y.
Dr. Bruch Molholt, 3
/;is Dr. Thomas Najarian
" !?
Dr. Mary Osbakken Dr. Alice M. Stewart Ms. Lynda S. Taylor Dr. George K. Tokuhata Dr. Kai-guo Wu l
e S
e
'/
m 9gy
,,_,y
._____-__.,,--.-._,.-.___-w__,,_,.-,,,%__.._,_%,
,,.y-
_w.,
...y._
A REVIEW OF INE CARCIXOGENIC EFFECTS OF LOM-LEVEL IONIEIXG RADIATIDM Daniel A.
Moffman, Ph.D.
Edward F. Radford, M.D.
O
- T E..,9.0, I:i. T. T.
Ei p.e u
s Vh'.s W. ? 2 r1 D h 25 L;hi h[:IY;k i.'
d 2
0 tic 2 TCA QC07ATN1:
FRIOR 20 FURLICATICM e
,A%
's
p.
RFT fi A.4 r-
$ %? f,kb5 Y
?.
~
TABLE OF COMTEXTS g
NOT FOR QU"TATIO3 PRIOR 20 P'JBLICATION I.
Introduction
- 1. Radiation induction of cancer II. Current Experimental Fasaarch on Mechanisms of Cancer Induction III. EPidemiologic Studies of Populations Exposed to Ionizing Radiation
- 1. Introduction D'.
u.y l-B.
Japanese Atomic-Bomb Survivors
- 1. Effects of new dosimetry 2.
Lympho-hematologic Malignancies 3.
Breast Cancer 4.
Lung Cancer
- 5. Other Sites l
- 6. Mortality from causes Other than Cancer
~
l l
C.
Studies of Medically Irradiated Populations
- 1. Persons Irradiated for Treatment of Ankylosing Spondylitis
.. (,.
l
e e
- 3. Children Trooted for Tinco CcFitis with X Rcy's 3.
Cancer Following Thysic Irradiation During Childhood 4.
Patients Exposed to Thorium Dioxide (Thorotrast)
- 5. Persons Treated by 1891a
- 6. Persons Receiving stF Treatment for Polycythenia Vera
- 7. Persons Treated by is1 Iodine S.
Nomen Receiving Felvic Irradiation for Treatment of Benign and Malignant Disease
- a. Cervical Cancer Therapy
- h. Benign Gynecological Diseases
- 9. Nomen Exposed to Chest FluoroscoFy
- 10. Nomen Treated by X Rays for Benign Breast Diseases
- 11. Persons Treated with Radium Therapy of the Xasopharynx
- 12. FoPulations Exposed to Diagnostic X Irradiation
- a. Persons Exposed In-Utero to Diagnostic Radiation gjj. }
\\.n,,
- b. Adult Leukemia and Diagnostic X 1ays D. Cancer Among Military Personnel Exposed to Xuclear Neapons Testing Fallout E. Studies of Civilian Populations Exposed to Fallout from Xuclear Weapons Testing
~
- 1. Marshallese Islanders
~
2.
Utah, Xavada, and Arizona Residents
/
D-r V
/
- k 13 I I
\\
y so: FOR C'.'C A: ION FT.ICR 20 FW.L :A:!CI
T. Studico of Populationc ExPosod to Occupctional Ionizing Rad,iation
- 1. Cancer among' Underground Miners
- 2. Muclear Workers (Ranford)
- 3. Studies of Muclear Shipyard Workers
- 4. Cancer Mortality Among Radiologists and X Ray Technicians
- 5. Radium Dial Painters 6.
Plutonium Workers
- 7. Other occupationally Exposed Groups G.
Studies of Cancer Morbidity and Mortality in FoPulations Exposed to Background Radiation.
IV.
SUMMARY
AMD CONCLUSIONS: DIRECTIONS FOR FUTURE STUDIES
($
V.
BIBLIOGRAPHY i
i l
l W
i EFT
~'
yo; yon que:ATION rxIo1 to rus:.Ici:Io2 l
I!l r
1 I. INTRODUCTION This report is meant to be a relatively concisa review of the current knowledge on the biological affects of ionizing radiation as it pertains to cancer induction in aan.
The review will, among other things, include information on current theorias of carcinogenesis, the role of call mutations ead transformations in the induction of cancer, current concepts of initiation and Promotion, various aspects of dose-responsa, both at the cellular and epidemiologic laval, the role of host factors in modifying cancer induction rates, and a current update on the major ePidemiologic studies and what we have learned from these studies as regard's the effects of ionizing radiation exposure on human' h
cancer.
.,p This review is not meant as an exhaustive and complete summary of the total literature on radiation biology or on the k
epidemiology of human cancer induced by radiation as this has pp already been done elsewhere (BEIR III, 1980 UNSCEAR, 1977).
Its Purpose rather, is to provide a current and concisa update on issues in radiation affects, thus the emphasis is on recent evidence obtained within the past two years.
RADIATION IMDUCTIDM OF CAMCER l
The biological effects of ionizing radiation have been I
studied more extensively than those of any other physical or
/ k R
5 a
/
30 FOR QUC ATION FEIGR 20 F M ICATICI
l l
chomical ogont, ospecially the cenoor-inducing offect.
Ionizing radiation has been observed to increase the incidence of cancers in virtuaily every type of tissue in many species including man.
Novaver, the degree of cancer induction is dependent on a variety of factors including organ, cell type, sex, age at irradiation, Physiological condition of the tissues, exposure to promotional agents, and other variables including species and strain of experimental animal.
Cancer induction is also a function of radiation f actors such as type of radiation (X rays, gamma rays, 1
alpha and beta particles and neutrons), dose, dose-rate, linear energy transfer (LET), energy distribution, and seguance of i
exposure (acute, fractionated and protracted).
The effects of th'ese factors as regards experimental studies will be one of the focuses of this review, in addition to a summary of recent epidemiologic studies of populations exposed'to ionizing 6;;g radiation, which will be Presented in the latter Part of this CD report.
II.
CURRIXT ZXPERIMEXTAL RESEARCH OX MICH1XISMS FOR CAMCIR IXDUCTION cancer is now considered to be a disease with a complex, multistage etiology.
While the molecular mechanisms of carcinogenesis remain to be precisely defined, two general models of cancer causation at the cellular level have developed.
The DRAD 50: POR CUD ATION PRIOR *Q P E I Z ICK
first nodal statos that ocztoin citorations of tho genotic i
material are essential to the development of cancer.
This can e
\\
involve adding new genetic material (such as viral genes),
deleting genes, or changing them (transformation) by point mutations or chromosomal rearrangements.
The second model 1
concerns activation or repression of normal genas that does not i
result from DNA alterations (the epigenetic theory).
'Both of i
l these events (transformation or epiganesis) must lead to a l
breakdown in the growth control in the calls so that normal growth regulation is lost.
In support of the first model, Yunis (1983) cites improved methods in tumor call culturing and chromosomal banding techniques have revealed that chromosomal defects are present in most neoplasia.
The most common of the recurring defects is
' ('.
either a band deletion or a reciprocal translocation between the Q':j-two chromosomes.
In most leukemias and in some solid tumors, the malignant cells show a reciprocal translocation.
Most solid tumor calls which have been defined chromosomally show a loss of a specific band or sagaent.
This list includes small call lung cancer and Wilm's tumor.
Yunis (1983) postulated that in neoplasia with a recurring chromosomal translocation, such as I
leukemia or lymphomas, the rearrangements may activate specific oncogenas.
The finding that certain leukemias and lymphomas may share specific chromosomal defects may be interpreted that these-defects affect stem cells and that the number of seguances initiating those human neoplasias may be small.
In vitro studies of radiation-induced cellular b
20: poR tuo:ATION PRIca to FunLIcA:Ioz
tronoformation hevo provided cvidenco es to tho encogenicity of transformed cells.
Radiation-induced transformation involves at g
least two and probably several stays.
The first stay is called initiation, which requires more than one stage.
Although the l
precise biological changes responsible for radiation-induced call initiation have not been completely identified, they most likely involve first, autational events, such as chromosomal rearrangements, or other types of nuclear or cellular changes.
The mutation must than be fixed as a heredity property of the calls, a process requiring call division soon after exposure, yinally, progression and expression of the altered genome requires several call replications and results in a colony of cells distinct in morphology from the normal (BoreX and Iall, 1982).
At present, there is no evidence which precisely locates the
,y) cell targets responsible for radiation-induced call transformation.
Possible targets apart from nuclear DNA include cytoplasmic DX1 and the plasma membrane (Goodhead, 1934).
Gould (1984) has proposed that the call and cell nucleus has multiple targets that interact with the carcinogen (e.g.,
radiation) leading to initiation.
The more targets that are hit, the greater the probability of initiation.
The call becomes maximally initiated before all the available targets are hit and the striking of additional targets does not increase the level 4f initiation, that is, a saturation point is reached.
This theory, if correct, could explain the initial rise in the dose-response curve (for transformation), the eventual plateau in the curve, N
30; FOR SUCTATION F113R 20 yunkICATICE
l cnd tho downturn in tho curvo (the saturation point) at high doses.
c Experiments in which irradiated cultures have been subcultivated at various dilutions after their growth to confluence have been interpreted as implying that transformation is a late-occurring event, the probability of which increases with the number of call divisions fol. lowing irradiation, and that every irradiated call is initiated and' potentially transformable.
Iowever, Xennedy at al. (1984) have demonstrated in 10T1/2 cells 1
that although 100% of the call progeny treated with either methylcholanthrone-or X rays vers' potentially transformable, only a.very small proportion of the treated and initiated cells produced transformed colonies.
This finding indicates that there must be a second and very rare stage in call transformation, h'
which may occur either soon after initiation or later after growth to confluence.
The second stage could be the result of DNA lesions produced by X irradiation.
Tumor promotion, the second stage in the two-stage model of carcinogenesis, has effectively been demonstrated experimentally in animals.
51aga et al. (1978) observed that in nouse skin exposed to a low dose of the tumor initiator, 7-12 dimethylbenzanthracene, no tumors appeared during the period of observation.
However, after an interval of one to two weeks, repetitive treatments with a tumor promoter resulted in the i
development of both benign and anlignant tumors.
Exposure to the tumor promoter alone did not result in neoplasia in uninitiated control mica.
9 DRAR 30: POR QUOTATION PRIOR 20 FfE ICATION
Tho two-stago modol for corcinogonosio is dopondant on tho dose of both the initiator Lnd promoter (Boutvall, 1982).
The shaya of the dose-response curve for skin carcinomas in SKI hairless nice following ultraviolet radiation'(UY1) exposure is modified by the addition of a promoter (Try, 1984).
,Tollowing UV1 initiation, the plot of skin cancer as a function of UY1 dose is curvilinear, with the suggestion of a threshold.
Treatment with the tumor promoter, 12 te t r a d e c an o yl-p ho r b o l-13 -a c e ta t a (TPA) changes the shape of the dose-response curve to a linen *r, no-threshold relationship, consistant with the theory that at low doses, latant initiated cells are made manifest by the promoter.
The results of this study suggest the presence of unknown systemic factors that suppress the growth of UVE-transformed cells, and that exposura to TPA overcomes this suppression (Try, 1984).
These observations suggest that the differences in tumor v).~.a l C9p) response in different tissues in the same species or the same tissue in different species may be due to factors Cyromoters) that influence axyresion.
)
Indocrina stimulation may promote the pathogenesis of other radiation-induced tumors such as those of the murine Harderian gland.
Normones promote carcinogenesis by enhancing call l
replication, which can result in increasing the chance of armor of DNA copying or unmasking latent DXA changes caused earlier by
)
initiating events (autagens).
Other exaarles of tumor promotion l
following radiation-induced initiation include the induction of murina thysic lymphomas (with urethane as the promoting agent) and the enhanced induction of renal tumors in rats (with DRAPI
~
~
c 302 yCR QUCTATION FF.ICR 20 P M ICATION
contraintorcl nophroctony cs tho promotor).
Rorsk and Ea11 -(1982) observed enhancement (promotion) of the transformation frequency of Ramster 10T1/2 cells by treating in vitro cultures with TPA following irradiation.
The promotional effect became apparent during the expressi'on period rather than at initiation.
They also observed that exposure of 1071/2 cells to X rays or gamma rays in the presence of an analog of retinyl, trans-ratinoic acid, significantly inhibited the frequency of neoplastic transformation.
Ratinoids also inhibited the promotional effect of X ray transformation exerted by TFA.
Superoxide disautase (50D) was aliso observed to inhibit X ray induced transformation on 10TS/1 cells, suggesting that the oncogenic effects of X rays are in part mediated by free radical action, as would be expectea on theoretical grounds.
In vitro transformation assays have also shed light on the roles of dose, dose-rate, LIT and. dose schedule on the rate of es11 transformations.
For low-LET radiation, the dose-response curve for inducing chromosomal deletions is approximately linear while for double-strand break aberrations, the desa-effect curves observed were guadratic (proportional.to the square of the dose) or linear-quadratic.
For the linear-quadratic case, with decreasing dose-rate for low-LET radiation, the transformation induction rate decreases because the dose-squared term becomes less important.
For very low dose-rates, only the alpha (linear) term is left and the dosa-effect curve is a simple proportional or linear one (Bander, 1984).
Rossi and Nall (1984) have proposed a dual radiation action 302 pcR SUOTA~ ICE ygIcx 20 FUE:.ICAZICE
~
for low-LET rcdiction at highor desco, whora the shcro of tho dose-response curve would be linear guadratic according to their
(
theory that pairs of sublesions are necessary for a anlignant neoplasm to be expressed.
Therefore, from this theory, the 15E for transformation would be equal to XCDh)-1/2, where Dh is the.
absorbed dose for high-LET radiation.
Iowever, Goodhead (1984) has shown that the 13E values predicted by the Rossi theory varied considerably from those he observed in his transformation assays and that a linear, no threshold dose-response relationship for doses up to 200 rad was a more appropriate indicator for transformation analyses.
If there is a linear component for low-LET dose-response, than the 15E should approach a fixed upper limit at low doses, i.e.,
the ratio of two linear slopes.
In experiments with mouse 10T1/2 cells, Borek and Nall (1982) observed a complex X ray dose response t'alationship for
- - h )
cell transformation.
At doses above 100 rad the transformation
[
data were consistent with a quadratic dependence on dose.
Below 30 rad, the data were consistant with a slope of 1, implying direct proportionality to dose.
Between 30 and 100 rad, the dose-response curve was very shallow.
In fact, within the confidence limits of the data points,.the transformation frequency did not appear to vary with dose over this range.
Based on these observations, Borek and Hall (1984) concluded that a linear or linear-quadratic extrapolation of high dose data would not predict the effects of low doses.
In fact, the line~ar projection might underestimate the effects at low doses.
Variations in radiation dose scheduling and dose rate have 30; 70R qu :ATION yytICR 20 FUILICATIO3
orror-Prono roPQir Procono whilo the coll rayoir process for 1
(
low-LET exposure appears to be error-free.
However, cell repair N
following low dose rate, low-LET irradiation may not be completely error-free.
Consequently, call transformation can occur although at a reduced frequency.
This observation challenges the concept that the risk of cancar following low-1ET, low dosa-rate exposura is minimal or non-existant because of
- armor-free rayair of call damage.
While in vitro transformation assays with animal calls have proven valuable for quantitative studies of oncogenesis, future studies of human cell systems are necessary to assess radiation-induced cancer risk.
Borek and Mall (1982) studied cell transformation in human diploid skin fibroblasts following 400 rad of X rays.
Radiation transformation was enhanced by
,j[
beta-estradiol.
The transformation rate in human cells was lover, than that observed in hamster cells at a dose of 400 rad of x 1
rays.
In human cells, marked chromosome changes did not reflect the neoplastic state.
yurther analysis of X ray-induced human j
call transformation should alucidate the vagaries between animal and human call systems in their Progression from a normal to a neoplastic state.
These recent experimental studies emphasize a number of
~
important points:
(a).
Empirical findings in simple experimental systems support a superlinear, linear, or even a sublinear dose-responsa DRAFT N07 70R CUDIATIIMI MUCR TO PUEICATICI
III.
ZyIDEMIOLOGIC STUDIES OF popULATIQXS EXp0 SED TO
(
IONIZING 1ADIATION Epidemiologic studies of populations exposed to various sources of ionizing radiation have played a central role in defining cancer risks following radiation exposure, in identifying the roles of host factors which modify risks, in defining differential target organ sensitivity, and in describing and guantitating dose-response relationships.
- Iowever, epidemiologic studies, especially those of low-level exposure, have been controversial.
Part of this continuing controversy is
~
a product of the observational nature o,f epidemiologic studies and such methodologic concerns as the sample sizes required to detect effects of a given magnitude, and various design or
., g execution biases.
The statistical problems in conducting I!
epidemiologic studies of pcpulations exposed to low-level radiation exposure have been discussed in detail by Land (1980).
These criticisms are relevant to studies which have demonstrated both " negative" and " positive" results.
However, what is perhaps most remarkable about auch of the human epidemiologic data is the relative consistency of it despite the aforementioned problems.
There are sufficiently sound epidemiologic data with relatively precise dosimetry, at least compared to that for other environmental carcinogens, from which risk ooefficients have been developed for a variety of organs and host factors such as age and sex.
Historical perspectives and controversies surrounding l
10 so; poa qu::ATIos PF10E 20 FUILICAIION
tho devolopcont and uso of those modolo to oxtropointo risk estimates will be discussed later on in this review.
Important sources of the epidemiologic evidence -has been from studies of persons exposed to the atomic bombs in Japan, from persons'madically irradiated for both benign and malignant conditions, and from occupational exposures.
This report, in most instances, will review the most current data from these ongoing studies and will not present much in the way of historical background.
MOTE:
Throughout the discussion of the results from epidemiologic studies, the term, " person-years" will frequently be used.
This is, simply, the integrated measure of the follow-up time accumulated by individuals in a particular group.
It can represent, for example, the time in years from which
.J, follow-up of an individual began, up to the time that person either died, had a cancer diagnosed, or was last examined clinically.
The use of the person-years concept becomes necessary when Persons are followed for differing periods or lengths of tima.
It is a critical concept, as cancer induction l
following radiation exposura usually shows a characteristic time-dependant pattern.
One limitation of the application of this concept is that the collective measurement of person-years includes a mixture of individuals with differing lengths of follow-up.
Consequently, while this is onta advantage to using this method, it is also a vaakness in that the risk associated with a person year for 100 persons followed for 5 years (500 person-years) would not be the same as that for 25 individuals
- DRAFT sc: 7CP. quc 3 :ICI PEIca 20 FUELICATION
fo11cwod for 20.yoors (500 person-yonrs.
Sinco cancor is a "4
tine-doPondant disonso usually cenifonted by incroccing incidonco
(
)
as a population age's, one would probably saa more cases among 25
\\
persons followed for 20 years than among 100 persons followed for 5 years.
Due to this feroblem, the patterns of cancer risks following radiation eMyosure, have baan dafined when possible, by n.
time since exposure, to more precisely characterize the temporal distribution of radiogenic cancer.
Estimates of absoluta risks will be yrasanted for many studies, specific for cancer sita.
~
Absoluta risk is a measure of the observed numbers of cancer cases, above normal ~axpectation, paz million persons paz year of follow-up par rad of exposure.
This measura, which usually allows for a minimal induction parlod, defines the excess risk of developing a particular. cancer following radiation exposure, assuming that $he radiation induced cancers are simply added to
"., )
the normally Exp2cted number of cancer's of the same type.
The relative risk model assumes that radiogenic risk is expressed as a multiple of the natural age-specific cancer rata.
Consequently, absoluta risk could be described as an additive model while relativa risk is a multiplicative model.
Further
. discussion of these models and the impact of their use for risk pro:laction can be found-in Section IV.
s.
I JAPAXESE ATOMIC BOMB 50F. VIVO 15
~
Over the last'35 years, studies of the survivors of the atomic, bombs dropped in Siroshima and Xagasaki have establish'ad l
unequivocally an iperaased risk of maligancias of several sites l1 s
_p _
FK10120 P'XICATICI
.,_,__c.
l including loukonin, and concors of tho bronst, thyroid gland, salivary glands, lung, stomach, asophagus and urinary tract.
t More recent evidence indicates that both colon cancer and multiple myeloma can be added to this list (Schull, 1984).
Effects of New Dosimetry since 1980, there has been an intensive reavaluation of the radiation doses that vara experienced by survivors in Elroshima and Magasaki.
This reevaluation was based initially on the realization that the neutron energy spectrum from the U-235 Kiroshima bomb was in error.
With subsequent work it has been found that for the previous dosimetry developed in the early 1960's and used by RZ1y since about 1970 (called the T65 dosimetry), the water vapor concentrations in air assumed in the two cities, and the extant of shielding of gamma rays by Japanese
(* ',:
houses were also incorrect.
There has also been an intensification of efforts to check the results of dose calculations by measurements made in the two cities and by comparison with field measurements from plutonium bomb tests in i
Xavada (for Magasaki), and by criticality experiments at Los 11amos using a raylica of the Hiroshima bomb casing.
Measurements in the two cities have included thermoluminescence detection in guartz grains from bricks and roof tiles (to determine gamma ray exposure), which avidently can detect gamma -
doses down to 10 or 20 rad with reasonable accuracy, and activation of cobolt in reinforced concrete as well as
]
europium-151 activation in rocks (to determine slow neutron 307 poR (UCTATICX PRIOR 20 FJELICA:ICK
.-,.._,_.--,_..__.,p._,_,.,ryy,,
---,f- - -, - - - -. _. +, _ _,, -,,., _. _,_,,
,.__,--y.
oxposuro).
A few consurczonto nodo in Hiroshino shortly ofter the bombing, of phosphorus-32 resulting izon sulfur exposure to
{
fast neutrons have baan reavaluated to correct for slight calibration armors of the electroscope readings nada in 1945 (Woolson at al.,
1982 Loewa, 1982).
Thus when completed the new dosimetry will have the banafit of both theoretical and anPirical checks.
At the Frasant time, thera myyaars to be ganaral agresuant among the physicists about most aspects of the dosimetry.
The principal outstanding guastien concerns the secondary gamma rays from the ascending fireball in both cities share predictions based on current calculations are consistently i
higher than values observed in the Nevada test sita data.
The following conclusions about the dosimetry seem now to be well-founded First, the uncertainties int a) bomb yield b) anergy sPactrum of prompt radiation c)' fireball emission d) jj k?I transport of ra'diation through air and a) affect of ground reflection and ground characteristics, indicate that the fras-in-air dosas received at any distanca cannot be determined to batter than 25%.
For the large fraction of survivors who vara inside or other structuras or who vara shielded by buildings or terrain, there is additional uncertainity in the absorbed dosa.
Thus for this group, about two-thirds of the exposed Persons, the dosas have a total uncertainty of 35-40%.
These uncertainties apply primarily to the absoluta dosas, less to the relationship with distance.
as a function of distance from the hyPocantar,'the
- Second, neutron dosas in the cities vara nearly the same.
The aqpa
(
)
302 70R QUOTATICI FRIQE 20 FUEJCATICE I
centribution of prompt noutrons to tho totni deso is oignifictat only within a ground distance of about 1200 meters, and was
('
essentially zero by 1500 meters.
Because the numbers of survivors within 1200 meters were relatively few and their shielding characteristics and gamma doses are difficult to J
assess, the data from the 1-bomb survivors cannot yield any valid information about the comparative radiobiological effects of l
neutrons compared to samma radiation. -
Third, depending on resolution of the extent of the delayed gamma radiation, free-in-air gamma doses in Magasaki were about as given for the T65 estimates, yor Hiroshima, because there were many more slow neutrons from the bomb than had been thought (and hence ground level neutron doses vere only about 10% of the T65 dose estimates), the neutron-gamma reactions increased the gamma flux at the ground.
Thus Biroshima free-in-air gamma doses 6
d-may have been about 2 to 3 times higher than the T65 estimates.
Despite this increase in the delayed gamma radiation, the total physical dose (gamma plus tautrons) as a function of aistance is not greatly changed from the T65 estimates.
Fourth, computer simulations of attenuation of gamma rays by traditional Japanese houses indicates that for the 1-bomb spectral range, gamma transmission. coefficients vary depending on location in the house, whether the house is in a cluster, typically the situation in both cities, and the ground angle to" l
the epicenter.
On average, however, a mean transmission coefficient of about 0.40 is appropriate for the many configurations tested (Woolson, 1984).
This is less than half E.
J 30; FOR QUCIATI M
_ l..
FRIOR 20 FUBI*ICATICI
~
the valus unod in the T65 ostincteo, and thus the curfaco kerns 1
dose from gamma radiation for a large fraction of the survivors 4
)
(
was subst'antially less than estimated for T65, especially in Magasaki.
yifth, estimates of body shielding for specific organs as a function of angular orientations to the epicanter shows that for most internal organs the degree of body shielding is relatively unaffected by this angle.
Surface tissues such as female breast and thyroid show more affect of facing toward compared with facing away from the source.
In ganaral, however, the 265 assumption of a random orientation is unlikely to introduce much armor, particularly because the actual orientation of the
~
survivors to the epicenter at the time of the datonation is uncertain in many cases (while the prompt radiation exposure took 4
place almost instantaneously, the delayed samma radiation was
...)
received over a period of about 10 seconds).
II Finally, despite the above factors, which have the effect of reducing the doses to a large number of the survivors, it is evident that if those exposed to more than 0.5 rad (surface karma)'are taken as the exposed group, this group includes people out to a ground range of 2300-2600 matars in E1roshima, and 2500-2800 meters in Magasaki, the range depending on shielding.
That is, well beyond the point at which zero dosas wawra thought to apply in the earlier dosimetry.
A substantial number of those survivors previously in the zero dosa category will now be included in those exposed at low doses.
This applies als'o to the 11,000 new persons added to the Magasaki study to increase the 1
1 ErgE g
4 f
I L
4 so: 70R quanTION palon to F m ICANCA
control population, obout half will octuilly bo oddod to thoso exposed at low doses (1-10 rads karma).
The conclusions that can be derived fron the above dosimetrio changes are that in both cities the radiation exposures were essentially to gamma rays alone except for a small number of persons exposed near the hypocenter who could have received substantial neutron doses.
Therefore, it will be appropriate to combine the results for the two cities for some of the analyses, particularly those concerned with low dose effects.
In addition, the tissue gamma doses for most survivors were substantially less than estimated from the T65 dosimetry, thus the previous estimates of risk.of radiation effects will likely be revised upward, possibly by a factor of about two.
- Moreover, when the 800-odd people who were in the Mitsubishi plants in Magasaki Cand for whom no shielding was previously assumed) are assigned reasonable shielding estimates or are omitted from the i
analysis, then it may be that some of the anomalias in Magasaki dose-response data will be reduced.
In other words, a significant group of people in the middle dose range had their
~
doses consistently overestimated, which could have had the effect of making the dose-response relationship appear to sag in the middle.
Finally, the low-dose discrimination of affects will be improved by proper distribution of the control Czaro-dose) group.
Enormous effort has been expanded to improve the dosimetry" from the two atomic bombs, and the new results are likely to be as definitive as we will ever get.
The fact that the calculated and measured free-in-air doses agree as well as they do suggests
~
MAFT 70 FNL QUQ 3:10I PRIOR 20 FUBLICATION
that no now burprinoc cro likely to altor the ostinnton in a significant'way, though the dosimetric uncertainities identified 3
(
vill remain.
lympho-Iamatologic Malignancias in the Japanese Survivors laukania was the first malignancy to appear in excess in the atomic bomb survivors, but in the last two decades, the numbers of solid tumors hava surpassed those of leukemia in importance.
The leukemia acrtality excess peaked about 5-9 years after exposure, with an absoluke risk coef41cient of 4.1 excess deaths per million PY1 and has since declined.
The most recent estinata for the, 1975-1978 period was 0.4 excess deaths par million PYR (Xato at al.,
1982), and the total risk for the period 1950-1978 was 1,72 excess laukakia deaths par million PYR (Finch, 1984).
A g,)
linear dose-responsa relationship down to 30 rad karm was
.pc4 observed in Hiroshima, while in Magasaki, a dose proportionate relationship was sean only in the high dosa cohort (>100 rad).
This is likely due to the smaller sample siza in Xagasaki and correspondingly greater sampling variation at doses below 100 rads kerma.
The new dose estimates should modify the previous risk coefficients and the between-city differences in them.
Age at exposure CATB) and radiation dose both affect the magnitude of the excesses seen for the various call types of leukemia, with the 0-9 and over 50 ATB age groups having the highest absolute risks.
The risk was lowest in persons 10-19 years of age ATB and thereafter increased with increasing age IE 2
E so: pu quen :Ic3 FRIGE 20 PUBLICATIC3
ATB, n ynttorn consistent with persono treated by X rays for
/
ankylosing. spondylitis.
The peak risk for acute leukemia
(
occurred at a later time and lasted longer in older individuais ATB compared with younger age groups (yinch, 1984).
The risk of developing chronic granulocytic leukemia was greater in younger than in older individuals ATB, with the excess lasting for a
, shorter period among the age group < 20 1T3.
No excess in any dose group was observed for chronic lymphocytic leukemia, an observation consistant with other studies.
The high dose acuta leukemias ayyeared much earlier in time than the low-dose acute leukemias 'for survivors who were 0-9 and 10-19 years ATB, but not for the older cohorts.
The relationship, was reversed for survivors who were 50 years and older 1T3 (Land and Norman, 1978).
The overall absolute risk of dying from h}k leukemia in survivors exposed to 10 or more rads karma was 1.7
- y -l
_p excess cases per million PYR betwaan 1950-1978.
One area where the Japanssa laukasia data are discrepant from other data is for persons who were exposed in utero.
Jablon and kato (1970) observed no increased risk of leukemia or any other malignancy in 1292 person's who were exposed in utero and were less than 1500 maters from the hypocenter.
This finding was inconsistant with the findings from several studies of persons exposed in utero to medical irradiation (Stewart at al.,
1958 MacMahon, 1962), where increased risks of leukemia on the order of 50% were observed.
Excessas for othar solid cancers were also observed in one study (Bithell and Stewart, 1975) but not confirmed in others.
Mole (1974) hypothesired that the l -
ser FOR (UOTATION rxIca ;o FUE:.ICATICI
l discropency botwoon tho Joycnoso data and that of the other studies was a conseguance of the neutron doses received by the
)I
(
fatal bone marrow which resulted in call sterilization and a i
\\
resultant reduction in the risk of leukemia in the exposed i
population.
Eau the new dosimetric calculations, with their lowered neutron component will affect this theory is not known, although it could lead to a new perspective on the call-killing capacity of gamma irradiation.
1 slightly increased risk of lymphoma was observed only in persons exposed to 100+ rads. 'These data are based on only 9 lymphomas in E1roshima and 5 in Magasaki (yinch, 1984).
No significantly greater incidence with increasing dose was observed.
The average time between exposure and diagnosis was 20 years in the 200+ rad group, with no apparent evidence of an age ATB affect on latency.
The absolute risk for the 1950-1978
,.)
period was 0.04 excess deaths per million PY1.
in excess of multiple myeloma has now been observed for the t
first time in the heavily irradiated cohorts (Ichimaru et al.,
1982).
The minimum latency period was 20 years and no excess has yet been observed in individuals less than 20 years of age 1T3.
This is likely due to the fact that these persons are just now entering the peak years of nyelona incidence and further excesses should be observed with continued follow-up.
The estimated absolute risk was 0.2 excess cases per million PYR (Ichimaru et ~
al.,
1982).
Breast Cancer Tokunaga et al (1982) recently presented data that for the EET' I
' I I
302 70R QUO 3T105 yRIon :0 yuaLIcA::cs
-~w,-.
l first timo showod n Indintion related risk in porsono loss then l f
- years of age ATB.
The data, based on 24 cases of breast s.
1 cancer diagnosed between 1950-1980, suggest a. breast cancer risk i
among the women under.10 years ATB exposed to over 50 rads kerna l
to be over 7 times as high as that among woman with less than 10 rads kerna.
This finding has important implications for theories of carcinogenesis in that initiating events can apparently occur
)
i l
years before the budding of breast tissue during the
~
perimenarchal period, a time previously thought to be the aost sensitive to the effsets of radiation.
In another report Tokunaga and Land (1984) show that the previously observed deficit of breast cancer in women 40-49 years ATB has now disappeared with additional follow-up.
The failure to observe a i
significantly increased excess of breast cancer for exposures
('[,h after 40 years ATB is consistent with Japanese statistics for breast cancer incidence.
These patterns show a leveling off of breast cancer incidence after the. age of 40 in contrast with the pattern in Western women, where a sharp increase appears.
The risk of radiation-induced breast cancer increases in a
. linear fashion with increasing dose (Tokunaga et al.,
1980) and the absolute risk per rad is highest in the group 10-19 years ATB with approximately similar lower estimates for ages 20-29 and 30-39.
The minimal latent period for persons less than 20 years i
ATB ranges from 15 to 20 years and 10 years for persons 30 or more years of age ATB.
Breast cancer excesses are now being seen 4
in women exposed to breast tissue doses of 10-20 rads, which may
)
be supportive of the no threshold hypothesis (Tokunaga et al.,
x h y.y 'h O
z.
2/
!j BU, jF
@k t!
~
NOT POR 0;c;ATION m.
'2 -
e n -n wu CIC1:13
,.... _ - -. _ ~,, -. -,
1984).
No rolctionship botwoon radiation dos'o end intoney wcs observed in a recent analysis (Tokunaga and Land, 1984).
An
)
(
ongoing case-control study of breast cancer among the Japanese
\\,,
survivors hopes to identify possible relationships between known I
breast cancer risk factors (reproductive history, family history, etc.) and radiation dose to determine whether these host factors modify radiation-induced breast smacer.
sung Cancer Radiation related lung cancer was first observed among the Japanese survivors about 10-13 years after exposure.
Incidence data from 1959-1978 indicate that the greatest excess occurred in persons over 50 years AT3.
The size of the excess decreases with younger age at exposure.
This pattern is not unexpected since f
the peak years of lung cancer incidence occur after 50 years of age.
It appears that the size of the radiation-related excess is s$$v a function of attained age.
That is, when individuals enter the age range normally associated with the incidence of lung cancer (Beebe, 1978).
Recent analyses of data from a case-control study of lung cancer among the A-bomb survivors indicate that the relationship between radiation exposure and smoking is additive in nature (3. Blot, personal communication).
Other Organs I
l Excess stomach cancer mortality among the Hiroshima survivors first appeared in the 1959-1962 time period (14-17 years after exposure) and the absolute risk steadily increased, with the most recent time period.( 1975-1978) showing an absolute L..*
- f-'.=t.
?
f
'5
' S*?
3 k r'b " t
=
u r
n su/B e a
sc: rca quou rzon prion 20 FUBLICHIcN l
i risk of 1.9 oxcoss ccsos por million PYR.
The pattern of stomach cancer mortality in Magasaki was less pronounced although in the 1975-1978 time period, the excess mortality was markedly increased.
The time lag in stomach cancer mortality excess between the two cities was probably due to the ol' der age distribution of Hiroshima.
Analysis of stomach cancer mortality by age ATB indicated that persons under 35 years of age had a
~
greater absolute risk than persons >35 years.
A slight sex difference for stomach cancer mortality was observed with females having a higher risk (100 rads versus o rads) than males."
Evidence for a radiation-ind'uced pancreas cancer excess among the cases identified through the L55 mortality data (Xato and Schull, 1980) or through the Magasaki Tumor Registry (XTR)
(Wakabayashi et al.,
1983) is not clear.
In the mortality analysis, there was no increased risk of pancreas cancer in any
,g of the ATB age groups through 1978.
Novaver, in the MTR data, a relative risk of 3 was observed among the autopsy cases in the
>100 rad group compared with similarly identified cases in the o dose group.
The relative risk decreased to 2.2 when only the histologically confirmed cases were analyzed, which was not significantly greater than one.
This observation points out the potential problems of misclassification of pancreas cancer, since it is a difficult tumor to correctly diagnosis.
Death
~
certificate diagnosis of pancreas cancer is especially poor
~
(Miyoshi et al.,
1977).
i A significant excess of urinary tract (UT) cancer (kidney l
and bladder) was observed in the incidence data of the MTR b
i 'u r. n e. n N a
OjjY; 4 [ t.;U 1 0 L
23 Roi TOR Q'J0:Atton PRIen -n winer.ree-w.
(Wokobayashi ot cl.,
1983).
Tho obsoluto risk was ostimated to be 1.2 excess cases per million persons per year per rad.
C.
Mortality data show that the first excess of urinary tract cancer appeared about 20 years after exposure (Xato and Schull, 1982).
In the most recent time period (1975-1978), a significant linear trend for UT mortality by dose was observed.
No excess mortality was observed in persons exposed to less than 100 rads.
The influence of age at exposure was not evaluated due to the small numbers of cases.
The absolute risk for UT cancer mortality i
between 1950-1978 was 0.2 excess cases per million pYR.
Recent data from the Japanese survivors now indicate that i
colon cancer is emerging as a radiogenic tumor (Xat_o and Schull, 1982 Wakabayashi et al.,
1983).
The mortality data between 1950-1978 show that colon cancer first began to appear in excess about 20 years after exposure (1963-1966) and has steadily
-)
increased as a function of dose with increasing time since dM' exposure (Xato and Schull, 1982).
In the 1975-1978 period, the absolute risk for colon cancer was 1.8 excess cases per million PYR, although this risk was based primarily on the Hiroshima data as the Magasaki mortality data showed no significant trend.
The excess in Hiroshima was limited to persons with 100+ zads exposure with an absolute risk for the 1950-1978 period estimated I
to be 0.3 excess deaths per million FYR.
In the analysis of the incident cases from the MTR, a significant linear trend with increasing dose was observed.
The absolute risk based on the Magasaki incidence data was 1 excess case per million FYR'which is greater than the absolute risk based on mortality.
f S k!e lin f.al Sy w$ Dh.g y 5.a.
e'j 30T 70K GUCIA!IG3 FRICR 70 PUR:.ICA: ION i2
4 In tho Magnanki incidonco data, the risk of Prestoto concor,
{/
both clinical and occult (autopsy cases), increased as a function of dose.
It is likely that, in part, this relationship is biased due to the differential dose distribution of autopsied cases.
A higher proportion of the high-dose cases were autopsied which increased the likelihood of diagnosing occult prostatic cancer among them compared with the low-dose cases.
Based on the recent mortality and^1ncidence data from Eiroshima and Magasaki, it is now clear that the absolute risks based on mortality underestimate the true risks, possibly substantially so.
Leukemia is the only malignancy where the absolute risks based on both moztality and incidence are similar.
For example, the absolute risk for stomach cancer based on incidence data is 1.5 excess' cases per million PYE compared with 0.2 excess cases per million PYR for the absolute risk derived p
from mortality data.
However, the relative risks (100+ versus 0 dosa) for different cancar sites are relatively consistant when comparing the mortality and incidence data.
The difference in magnitude of the absoluta and relative risks for incidenca I
compared with mortality data is due to the l'arger number of incident cases available for analysis.
Furthermore, there is a shift in time from exposure to onset (induction time) if date of cancer diagnosis instead of date of death is used as the data of onset.
The size of the time shift varies by cancer site, with a l
small shift, f or example, for rapidly fatal malignancias such as cancer of the pancreas.
However, the variation is larger for such sites as breast cancer and thyroid cancer, which have longer 0
$ i;iM.No b lE.fif {
2r d ; YL i 0
c NOT FOR QUOTATION wer e ae
.a w...._
,_-,,-,-n-,
ev - - -
,<,,--~-----------,---v
~ ~ - -
ourvival ticos.
As the yoyulation of Joycncoo curvivors ages Cand 70% are still alive), and the spontaneous cancer incidence increases, the absolute risk will increase accordingly while the relative risk should remain constant.
Conseguently, the Magasaki l
incidence data for all cancers except leukemia are more compatible with the relative risk model.
Therefore, for projection of future cancer excesses, the relative risk model appears to be the more ayyropriate as it myyears to be more I
stable in an aging yoyulation.
Use of the relative risk model l
will increase the projected excess in exposed yoyulations of cancers of all dites except leukemia.
This observation also emphasizes the importance of 1idatine follow-up of exposed cohorts, especially those exposed at a young age and who are just now entering the decades of increased cancer incidence rates.
i Ib Mortality fros'Causes other than Cancer Xato et al. (1982) recently reported on mortality from causas other than cancer for deaths up through 1978.
Except for diseases of the blood and blood-forming organs, no 1'ncreased mortality from non-malignancies wa.s observed in any of the age ATB or sex groups.
No increase in non-cancer mortality with increasing radiation dose was observed except for deaths due to
- diseases of the blood and blood-forming organs, where a significant trend with increasing dose was observed.
Evaluation of possible biases introduced into the data by exclusion of deaths occurring before 1950 was conducted and the authors concluded that.this was unlikely to affect the interpretation of i
h,h' hs }.'; e
~
[
5 e7 L.
?
307 POR CL*C7ATION Ei.V.4
'.'O :
~7
their findings.
Nowavor, Stowart (1982) hypothosized that the
" healthy survivor effect", as she measured by the 30% reduction in mortality due to cerebrovascular accidents CCVA) among the Japanese survivors, a cause she selected as being non-dose related, is an indication of the level of adjustment n'ecessary to accurately estimate the excess cancer risk due to radiation exposure.
She calculated that the current risk estimates for cancer mortality among the' Japanese survivors are low by a f actor i
of 10 as a consequence of this adjustaant.
Ginavan and Fuskin (1983) dispute this interpretation primarily on the basis of the observation that the Standardized Mortality Ratios (SMR) for cVA in the not-in-city (unexposed). Japanese cohort was 75, or a 25%
reduction in nortality, which was similar to that in the exposed groups.
There' fore, the " silent forces" hypothesis was acting in
. -3, a similar fashion in both the exposed and non-exposed groups,
-c5 I'
thereby negating Stewart's argument.
STUDIES OT MIDICALLY IRRADIATID popVLATIONS l
Much of the information on radiation risks comes from
(
studies of popul'ations exposed to radiation for diagnosis or i
treatment of both malignant and nonaalignant diseases.
These
~
l include studies of persons exposed to radioistopes and external
(
beam radiation.
l h;ev'v>aW[i2k e
g pdi g
M ggid %!
t.i E
30; PCR quc;ATICH PRIog to FUBLI::ATICN w--
CAMCIA IX PERSONS TREATED TOR AMXYLOSING SPONDYLITIS cancer mortality has been evaluated in 14,558 persons treated with X rays for ankylosing spondylitis between 1935-1954 in 81 British radiotherapy clinics.
Several reports have been published on this series of patients, the most recent one being that by Smith and Doll (1982).
In the latest analyses, follow-up I
was restricted to those 'atients with only one course of p
treatment, in order to minimize the complications that subsequent irradiation would introduce into evaluating the excess cancer 5:isk.
Thirty-one cases of leukemia were identified through death certificates compared with 6.5 cases expected, based on general population rataa.
The relative risk (observed / expected) for leukemia mortality was greatest in the period 3-5 years after the 1
first treatment and declined thereafter.
The excess death rate (relative risk / person-years) also exhibited a similar temporal f
pattern.
It.is likely that the leukemia effect has now been i
exhausted, a pattern similar to that of the A-bomb survivors.
Leukemia mortality was elevated in all age groups but the greatest relative risk occurred in persons 45-54 years of age at first treatment.
The excess death rate increased with increasing age at treatment.
No clear dose-response relationship was apparent when excess leukemia mortality was plotted as a function i
of mean bone marrow dose.
The greatest excess of leukemia occurred between 100-200 rads.
At higher doses, the risk was reduced except for doses over 500 rads.
A variety of dose response functions were tested and the best fit to the data was l.l..-.;..b;,'d g
I n
302 70R quo:ATICS FRICE TO FUBLIC1.TIO2
{
,,,e-------
- - -*, -. _ y_.. _., _ ~ _, _...
j obtainod by a torn linoor in deso cultipliod by a negativo a
exponential term (Smith, 1984).
One of the issues addressed by
~
this study was the propriety of utilizing a mean bone marrow dose when in reality, parts of the marrow received much higher doses than sections of marrow not in the primary X ray beam.
- Thus, i
l some of the marrow stem cells may be sterilized, leading to the downturn in the dose-response curve over 200 rads.
An excess of deaths from cancers of the " heavily irradiated" sites was also observed in this study (259 observed, 167.5 expected).
This category included a significant excess for deaths due to cancer of the esophagus, stomach, lung, CMS, and non-Modgkin's lymphomas.
Colo-rectal cancer was excluded from this group as spondylitic patients have a high risk of ulcerative colitis which is associated with an increased risk of colon cancer.
In order to separate the effects of radiation from those 4
e i
of spondylitis, a study of 1000 patients who were diagnosed with spondylitis but not treated with X rays was undertaken (Radford et al.,
1977)
No excess of leukemia was found in this group l
and the number of deaths due to other neoplasms was close to
)
normal expectation.
The temporal pattern of cancer risk in the heavily irradiated sites differed from that of leukemia with the highest relative risk occurring to years after the first treatment.
This is consistent with the induction period for i
solid tumors observed in the atomic bomb survivors (Land and Norman, 1978).
There was no apparent increased risk more than 20 years aft'er exposure although this is likely due to the relatively few person-years of follow-up after this time.
The h.
s
- 3..
.ie
?~ 7 k
i.-
se: yoR (UO;A: ION PRIca 20 PunLICATION
rointivo risk for cancero of tho honvily irradinted sitos was similar in all age at first treatment groups although the excess
[
);
risk increased appreciably with increasing age at first
\\.
treatment, a function of the increased baseline nortality rates for cancer of these sites in the older age groups.
This
]
demonstrates that if the relative risk remains constant over the various age at treatment groups, then one can expect to see an
~
increased excess or absolute risk as the population ~ ages.
Smith (1984) observed that the pattern of increased risk with increasing age at exposure indicates possible Ankaraction between radiation and other factors, including tobacco and alcohol use I
Novaver, this study did not collect personal data in order to evaluate this hypothesis.
Darby (1984) conducted a parallel analysis of data from the ankylosing spondylitis series and the atomic bomb survivors.
For
!K'e;$
leukemia mortality, the estimates of both relative and absolute risk in the high dose (>100 rad) 155 group were approximately double the corrasponding values for the spondylitis series.
yor cancers of the heavily irradiated sites, the pattern was reversed with estimates of both relative and absolute risk being slightly higher in the spondylitis patients.
However, the magnitude of the overall radiation-related risk was similar in the two l
studies, whether measured on an absolute or relative scale.
l The patterns of leukemia mortality by both age at exposure.
time since exposure were similar in both studies.
The excess
, leukemia mortality rates increased with increasing age at l
exposure Cafter standardizing for time since exposure).
For both Y; h,) !,'.. aL f.
D.. t.Mf I.:
i g
O' f
- u a sB tr,.o W
h NOT FOR CUOTATION FRIOR 20 FU3LICA !C3 I
l series, there was a clear decrease in the excess leukemia l (
mortality rate with increasing time since exposure, with the excesses in both studies plateauing about 15-20 years after i
exposure, and then returning to normal expectation after 20 years, although in the spondylitis series, few patients had over 20 years of follow-up.
1 Tor solid tumors of the heavily irradiated sites, the i
relative risks showed no a gnificant trends with age at exposure, time since exposure, or age at observation in the spondylitis series.
In the atomic bomb survivors, there was a significant decrease in relative risk with increasing age at exposure.
There was no trend in relative risk with either time since exposure or age at observation, which is similar to the spondylitis series.
Barby (1984) concluded that while the analysis was not precisely dose-specific, there were remarkable similarities between the two Populations as to the risk of specific malignancies and the role of age and time in patterns of radiation risk.
The results of this analysis tends to reaffirm the generalizability of the Japanese atomic bomb data to other populations exposed to radiation.
Cancer Following Thysic It' radiation During Childhood
~
In the last 30 years, several studies have been conducted of persons who were treated for an enlarged thymus gland with x irradiation.
This practica, which was widespread during the 1930's and 1940's, resulted in large numbers of persons receiving s ?' 0 fW
,. x u a t YE U! '.?!
[.'.*
k S
h b
NOT FOR GCC7A" ION PRIOR 20 FUE:,Tt'a-int
fairly largo X ray desos to tho hond end nock topion, ospecially to the thyroid gland.
Nampelmann at al. (1975) reported on tumor incidence among 2,872 persons exposed at less than one year of age and who had been followed for an average of 30 years.
Thirty cases of thyroid cancer were observed in the exposed group compared with only one case among 5,053 unexposed controls (11=53) (shora et al.,
1964).
The average thyroid gland dose was estimated to be 142 rads.
The overall risk estimates were 5.2 excess cancers per million persons-years per rad for females and i
1.8 excess cases per million person-years per rad for males.
The appearance of thyroid malignancies in the exposed group began 6 years after exposure and the excess was still evident more than 40 years after irradiation.
The excess of thyroid cancer appears to be increasing with time.
For thyroid adenomas, the excess began to appear about 12 years post-irradiation and has shown no decrease over t'ine.
Xo relationship between thyroid dose and length of the latent period was demonstrated in this study.
Furthermore, the mean latent period was longer for adenomas than for carcinomas.
Thus, this study doesn't previde support for the theory that adenomas progress into invasive malignanicas.
Analysis of dose-response patterns for thyroid cancer indicate a
(
linear relationship best fit the data and that the guadratic component was not significant C5 hora et al.,,1980).
Evidence for-a dose fractionation effect was not observed as there was no l
apparent decrease in thyroid cancer excess at the highest dose i
levels.
To the contrary, however, dose fractionation appeared to increase the yield of thyroid adenomas (shore et al.,
1984).
1 bns I
l i
3 :.j rj.. Y b 5Gb a
.s y y
i
$bl l?
U go!.FOR QUC ATION l
FEICE 20 FUELICATION
{:
~
~
Nosploonnn at cl.,
(1975) first noticod on cxcosa of thyroid i
cancer among the irradiated Jewish subjects relative to that in I
irradiated non-Jewish subjects, even after controlling for factors such as thyroid dose and age at irradiation.
This excess has yet to be adeguately explained although part of this excess may have been due to more careful thyroid surveillance among the Jewish subjects (shore et al.,
1980),
e In the 1970's, several reports appeared based on the results of recall programs for thyroid screening programs in persons who were irradiated in childhood (Dagroot and paloyan, 1973 Refetoff et al.,
1975 yavus et al.,
1976.Schneider et ~l.,
1978 Royce a
W j
at al.,
1979).
Although these studies provide valuable clinical information on various aspects of radiation-induced thyroid cancer, they are of limited use for evaluating low-dose
,,A carcinogenesis because most of the persons received high
\\ tit radiation doses
(> 500 rads), the population was self-selected, the lack of an unirradiated, comparison group subjected to the
'same self-screening the incompleta follow-up of the exposed cohort, and the questionable importance of microscopic, suhbclinical thyroid tumors, which accounted for 25% of the identified tumors in one study (Tavus et al.,
1976).
patients Exposed to Thorium Dioxide (Thorotrast) l Thorium dioxide (1 1Th) was used as a contrast agent between 1930-1955 during radiographic procedures, primarily intravasculat injection for cerebral angiography and angiography of the upper and lower limbs.
The trade name of this agent was thorotrast.
Thorium, which is an alpha-emitter, was deposited in various body gN gEsq M l
.s p
l l
so FOR quo:ATION FIICR 20 FUBLICATION
~
tissuco including beno narrow and tho livor, cnd rosultod in continuous alpha particle exposure at a low dose rate throughout.
)
A typical 25 al in36ction of thorotrast resulted in an average liver dose of 25 rad / year, a dose to th.e spleen of 70 rad / year, and 16 rad / year to the endosteal 1myer in bone (Kaul and Moffs, 1978).
Three large surveys of persons exposed to thorotrast in Germany (van Kaick at al.,
1978, 1984), Danmark Cyaber, 1979),
and portugal (da Motta et -a1., 1979) have been of particular value due to the large numbers of exposed persons and long follow-up.
Excesses of liver anlignancias and leukemia have been observed in these groups.
yor liver cancars, the three types most commonly observ'ed vara angiosarcomas (which are very rare in the genazal population), bile-duct carcinomas, and hepatic-cell carcinomas.
None of the liver tumors appeared until over'16 years after the first thorotrast injection and were still
,)
N appearing in excess over 40 years after first injection.
In the 1
three European surveys, 3,046 exposed persons survived 10 or more years after exposure for a total of 53,371 person-years at risk 10 or more years after first injection.
Among these patients, 301 liver malignancies have developed compared with about 6 cases normally expected based on incidence rates in each of these i
countries, resulting in an excess of 295 cases of liver cancer l
l associated with exposure to thorotrast.
Estimates of radiation l
l risk are affected by factors such as deposition patterns of
~
l threratrast in the liver, the use of an average organ dose when in fact this may be incorrect, the incomplete follow-up in some i
of these studies leading to an underascertainment of malignancies i
)
Q.g
~
n so: FCA GUo A!!03 FRIoa to Fum:,Ica: Ion
~
=.
(vnn Koick, 1984) cnd the concopt of awasted radiation *, which is
(
a dose expanded in tissue already necrotic and therefore not 1
essential for carcinogenesis (Mole, 1979).
However, the BEIR sommittee estimated that the ultimate lifetime risk of liver I
malignancies associated with thorotrast injection was 300 liver l
cancers per million person-rads of alpha radiation (BEIR, 1980) for persons surviving at least 10 years after infection.
Rn increased incidence of leukemia has also been oberved in the thorotrast exposed patients (van Xaisk at al.,
1984
- Faber, 1979 da Motta et al.,
1979).
The estimated bone marrow dose was 9 rads per year or 270 rads over s' 30 year period following injection (van Kaick at al.,
1984).
In the German study, 27 cases of nyeloproliferative malignancies were observed in the thorotrast group compared with 2 cases in the une$ posed controls (van Xaick at al, 1984).
The majority of cases were acute myeloid leukemia.
The first case in the exposed group appeared 5 years post-injection, with a cumulative incidence of about 1.8%
45 years after injection.
There was no increased incidence of chronic lymphatic leukemia in the thorotrast group, which is r
~
}
consistent with other studies of radiation-induced leukemia.
patients Exposed to Radium (***1a)
In a. study of 897 German patients treated with 1**Ra injection for bone tuberculosis or ankylosing spondylitis following World War II, the incidence of osteosarcoma was greatly increased among 218 persons under the age of 18 years at l
treatment (36 cases) and in 680 adults (18 cases).
Oniy 0.2 cases would be expected in this group so all of the osteosarcosas 5
3(
, a ' j\\ 'h e
Wo.
n.;
l ser poa q'JoTAT108 Muon to nisLIcAnos w.--me+---em eve we,--c-,_
--,et-sr
+---w--w-u e mme
e
=
e
waro attributod to cipha irradiction of tho ondestoni curfaco of the bone (Mays and Speiss, 1984).
The first case appeared 3.5
(
years after injection with a mean induction time of about 11 years.
The temporal distribution of the ostaosarcosas following
- Ra injection was remarkedly similar to the pattern,of leukemia among the atomic bomb survivors (Mays and Speiss, 1984).
Although the risk of osteosarcosas per rad appeared to decrease with increasing age at injection, this could be due to the fai[ure to adjust for duration of follow-up or to the influence of dose-protraction (Mays and spiess, 1984).
In this study, dose-protraction (the time span between injections) resulted in a higher yield of osteosarcomas (27 cases per million person-rads of alpha particles).
Mays postulated that the protraction-enhancement effect was due either to an increase in the stimulus for cell division or prevention of repair of local i '[
~,
damage (Mays and spiess, 1984).
L It is interesting to note that compared with the radium dial painters exposed to azsRa (to be discussed later), the excess risk of osteosarcoma per bone dose was over 10 times greater in the patients injected with 22*Ra.
This apparently reflects the
' differences in dose distribution on the bone as
- Ra with its shorter half-life, is deposited on the endosteal surface of the 5
bone, which is the area where osteosarcosas develop.
Conversel.y, 228Aa is a long-lived bone seeker and its dose is distributed I
more evenly through the bone.
Consequently, in terms of osteosarcoma induction, the dose from ***Ra is more effective in that the sensitive targets (endosteal cells) receive the critical
? W(
D.,yE1 3
3, h
so: FOR quo:ATION nuon to MacICATIos
~
radiction.
- (
patients Receiving 827 for Treatment of Polycythenia Vera Modan and lilienfeld (1965) conducted a follow-up study of i
1,222 patients treated for polycythenia vera, secondar~y polycythenia or guestionable polycythenia at seven clinics between 1937 through 1953.
The modes of treatment were 81F and/or X-irradiation and phlebotomy in conjunction with several chemotherapeutic agents.
Approximately 98% of the total population was traced through 1961 and*57 confirmed cases of acute leukemia were identified from death certificates or hospital records.
The distribution of leukemia by treatment group was 17% in patients treated with both X rays and 81F, 11%
in the 81F group, 90% in the X ray treated group, and 1.0% in persons who received no radiation therapy.
The leukemia excess in the radiation-treated patients was not attributed to increased survival of these patients or to selection biases, although Tubiana et al. (1968) observed that the number of mil 11 curies of' 82F administered to polycythenia patients was greater in those patients with high initial white counts or splenomegaly, suggesting some selection factors in patients treated with 817.
Xo apparent interaction between X rays and 82F was observed in Modan and lilienfeld's study (1965).
The average time from treatment to death from leukemia was 9 years with the greatest incidence occurring 9-13 years after treatment.
In pationts treated solely with 82F, an increased risk of leukemia with increasing number of mil 11 curies of 827 administered was R.P.P'.Id.fa [ N f.*. "
77 L.!
so: pon quo:A:IoM
'e MucR 20 TJnIu Lu
obsorvod.
Although doso octinntos woro not ccdo in thio otudy, spiers et al.,
(1976) estimated the dose from 82p therapy to the
.I marrow in trabecular bone to be about 24 rad per a1111 curie injected, or 142 rada per treatment.
Studies of persons Treated with 181 Iodine several studies have"heen conducted of persons treated with 181I for hyperthyroidism,. cardiac disease or thyroid cancer.
Saenger et al. (1968) observed no increased risk of leukemia in 18,400 patients treated with 1812 for hyperthyroidisa compared with 10,700 patients treated surgically.
While the short follow-up period (mean=6.5 years) has been suggested as o.no reason the study failed to demonstrate an increased leukemia risk in older persons, the whole-body dose was relatively low, ranging from 7 to 13 rads, so that the predicted risk would likely have g
,,r c:
not been detected even in a study population this larga.
It is interesting to note that an excess of leukemia was observed in both the 181I and surgically treated patients compared with population rates (Tompkins, 1970).
Lewis (1971) suggested that there was an excess of acute leukemia in persons 50 years and older in the 181I treated patients, an observation not confirmed by Tompkins (1970).
Hoffman et al. (1982a), in a second follow-up of one clinic in'the original study population, observed no increased cancer mortality or mortality from all causes in the 181I treated group compared with patients treated surgically.
Although this study added nine additional years of follow-up, the study size was small and only a large increase in Q. R I}
(W s ;.. U.5\\
i m )p,.3
,.: isis! $ s!
a
)sd$b Jy 30: TOR GUCIATICI FRIOR TO FUALICATION j
concor cortolity would have boon datocted.
No increase in total l
cancer incidence was observed in the 181I treated population although a slight increased relative risk (1R=1.6) for cancer of 1
4 the heavily irradiated sites (salivary glands, thyroid c\\and, colon, kidney and urinary bladder) was observed in the exposed women (Ioffman at al.,
1982b) compared with patients treated surgically.
Novaver, no increased risks for any sites vara observed in the exposed women when compared with general pop'tlation cancer incidence rates (Ioffman, 1984).
No increased risk of b: east cancer was observed in this same study group although the breast tissue in women treated with 181I for hyperthyroidism was exposed to a mean breast dose of 40 rads (Hoffman and McConahey, 1983).
Again, the study size was too small to detect the level of risk expected in an exposed group of
\\
this age and ra$diation dose.
Larger study sizes will have to be observed'before this guestion can be more thoroughly addressed.
1 Thyroid cancer has not been found following therapeutic i
doses of 181I (Dobyns et al.,
1974 Mola et al.,
1980a
- Nola, 1984), possibly because of the cellular destruction rather than transformation following thyroid gland doses on the o,rder of 5000 to 10,000 rads.
However, the mean length of follow-up was less j
than 10 years in both of these studies.
Consequently, because of j
the long latent period for radiation-induced thyroid cancer, the i
exposed groups were likely not followed long enough to adequately avalua'te the ultimate risk of thyroid carcinoma following 181I therapy.
In a small' study of 215 women treated for thyroid cancer C P @s A; r. : l' 4
30.*;,d/!:'l't!s 3
7 qFes r 4r$ $'
39
> b %E"U NOT POR quCTATION PRIOR 20 FU11.1017:05
with high doses of 181I, pochin 11969) observed four cuses of leukemia compared with 0.08 cases expected based on ganaral
[
}
.s s
population ratas.
A sam 11' breast cancer excess was also observed s
although two of the area.st cancer cases appeared within two years of treatment, which Is-inconsistent with past studies of radiation-3nduced breast can er.
Brinker et al. (1973) also q
observed a, slight increased !1sk of leukemia in 194 Danish woman treated with 181I for thyroid cancer.
Iowever, because of the small numbe'rs of patients in both of these studies, whether or not 1812 therapy for thyroid cancer is laukanogenic has not sufficiently resolved.
Continuing studies of larger numbers of patients with more precise dosimetry is necessary to address this issue.
s s
Rola et al. (1980b) studied oVar 10,000 Swedish patients expos'ad to diagnestic' levels of J11I for_ evaluation of thyroid
)f I function.
No excess of thyroid cancer wa,s observed in this population (14 cases observed, 14.2 case,s expected, Obs/Exp=1.0).
However, the mean age of the study" group at first exposura was 44 years, Sitd only SX of the population was less than 20 years of age at exposura.
Consequently, the issue of whether or not younger persons"are at an increased risk of thyroid cancer from diagnostic levels of 1 1I has not been'adeguately studied.
This is an important health guastion because of the large numbers of persons under the age of.20 years ex1.osed to diagnostic 131I over the last 30 years and-in light of the results from a racent experimental study which for the first tima demonstrated similar thyroid cancer induction i: stas for both X rays and 181I (Lee et
- m. m 6..'P e
[d
'l,..*M[=,Ws
.Ml
~
FC: TOR GUO:ATION FRIOR TO FU3LICATIC3
cl.,
1982).
Tho Matienc1 Contor for Dovicos and Radiologicci Mealth is completing a larger study of children exposed to diagnostic 181I and, hopefully, the results,will soon be s
available for evaluation (p.
Immilton, personal correspondence).
Children Treated for Tinea Capitis several reports of health efacts in persons treated during childhood for tinea capitis by scalp irradiation have appeared within tha last 15 years.
Nodan at al. (1974, 1977) and Ron and Modan (198'0, 1984) observed an excess of thyroid tumors, both banign and malignant, brain tumors, and salivary gland tumors in 11,000 Israeli children treated by scalp irradiation for tinea capitis Fetween 1950 through 1974.
Twenty-three cases of thyroid cancer were observed in the exposed group compared with 5 cases in the unexposed control group (RR=4.6) and a cases in the unexposed siblings of the exposed group (1,R=2.7).
An excess of SN benign thyroid tumors also appeared in the exposed group (RR=9).
The mean radiation dose to the thyroid gland was estimated to be nine rads although considerable individual variation could have occurred due to movement of the child during the scaly epilation procedure (Modan et al.,
(1977).
The excess risk of thyroid cancer in the exposed group given one radiation treatment was estimated to be 8.3 excess cases per million FYR which is higher than studies with larger exposures (Ron and Modan, 1980).
The excess risk was much more pronounced in females than males and in Persons under the age of six at time of irradiation.
A linear trend for increased risk with younger age at'Arradiation was observed even after adjustment for numbers of treatments and b*
b u':.
'. ;l, h. s G
i
$N p
Its E
u 7
30T POR QUQ A!!ON FRIOR 20 FUBLICATION b
durotten of follow-up (Ron end Modtp, 1984).
Marth Africons, especially natives of Morocco and Tunisia, exhibited the highest
)
relative risks for thyroid cancer (RR=25) of any of the ethnic groups studied.
This finding led to a hypothesis that the high rate of heterozygosity for ataxia talangiactasia (AT) amon's North Africans may have accounted for the apparently increased radiosensitivity in this group.
This arguement is somewhat weakened by the f ailu::e to observe an excess of lymphoma in this population, a malignancy known to be strongly associated with AT.
In another study of 2,215 persons treated for tinea capitis (shore et al.,
1976), no cases of thyroid cancer were observed in a group of patients exposed to a mean thyroid dose of 9 rads.
Mowever, an excess of thyroid adenomas (RR=5), salivary gland tumors (RR=4), brain tumors (RR=8) and skin cancers, primarily of the head and neck (RR=5) were observed in the exposed group.
The
)
thyroid and brain tumors begun to appear in excess 5-10 years after exposure while skin and pai:otid tumors had a longer induction period, 15-20 years after exposure.
The discrepancy' between this study and that of Modan et al (1974) regarding thyroid cancer incidence may be due to the smaller size of the exposed group in the former study.
patients Receiving pelvic Radiation for Treatment of Benign and Malignant Diseases
~
several studies have avaulated cancer incidence and mortality in women treated for co'rvical cancer with external beam therapy or radium implants (Mutchison, 1968 zippin at al.,
1971 hoice and Mutchison, 1980 X1einerman et al.,
1982 Boice et al.,
1 M 3 h.
o E
u: pc. w e m ic, PRIOR To FU3LIcr Ics
t 1984).
The cost strikingly consistent finding in theso studios I
~
1 l
l has been the apparent lack of a leukanogenic effect from bone narrow doses ranging from 300 to 1,500 rads (Boice and i
i i
Mutchinson, 1$80).
The authors postulated that the very high I
j doses of radiation to active bone marrow may have been sufficient i
]
to sterilize or kill a high proportion of leukocyte stem calls, j
i l
i thereby reducir.? the risk of leuxemogenesis.
Iowever, in a j
recent report on the expanded Internation Study of Cervical cancer (Boice et al.,
1984), a slight excess of leukemia was l
j observed (77 cases observed, 66 cases expected), although the excess was not statistically significant.
For nonlymphatic and i
j acute leukemias, a significant. excess was observed (58 observed, l
41 expected, 11=1.4 952 CI= 1.1-1.8).
The excess was most prominent 1-4 years after irradiation and the pattern of risk 1
i l
over time was consistent with the latent period distribution l
$?
i normally seen for radiogenic leukemia (Boice et al.,
1984).
I other irradiated sites that showed excess cancers vara the l
rectum, bladder, and pancreas.
The temporal distribution of 1
j these cancers following radiotherapy suggest a, radiation
~
etiology.
An increased risk of lung cancer was also observed in i
this study, although it was attributed to smoking since this is I
also a risk factor for cervical cancer.
There have been several studies of the occurence of leukemia
~
and other malignancies following radiation therapy for benign menstrual conditions (Doll and Smith, 1968 Brinkley and l
Maybittle, 1969 Alderson and Jackson, 1971 Smith and Doll, 1
1976 Wagoner, 1984).
Doll and Smith (1968) and Smith and Doll i
1 g
4 Yb
.J
(
NOT PCR QLTTATICI PRIOR 20 FUBLICATION
(1976) followod 2,068 wooon tractod for metropathin hemorrhagica in Scotland between 1940-1960.
Over 971 of the women had one x
(
)
N ray treatment with a mean bone marrow dose of 135 rads.
patient follow-up continued through 1972, at which time the mean-follow-up was 19 years.
Seven deaths from laukaria were observed in the treated group compared with 2.7 expected deaths (11=2.6).
The greatest excess occurred 5-9 years after treatment.
An excess of cancers of heavily irradiated sites was also observed with increased risks being reported for cancer of the large and small intestines CER=1.7), rectum (RR=1.5), uterus (RR=1.6) and urinary bladder (RR=1.4).
The excess risk for these sites first became apparent 5-9 years after treatment and stayed constant thereafter.
It was limited to women over the age of 46 years at time of first treatment.
For leukemia, an increased risk with increasing bone marrow dose was' observed with increased risks g.I being apparent at marrow doses below 125 reds.
A similar
~
relationship was not seen for cancers of the heavily. irradiated sites.
The excess risk for leukemia was estimated to be 1.2 cases per million FYR (Smith and Doll, 1976).
l In a study of 1,893 Connecticut women treated between 1935 through 1964 with X rays or radium implants for a variety of benign gynecological condition, Wagoner (1984) observed an excess s
of leukemia (1R=2.3) and cancers of heavily irradiated organs, I
~
primarily the kidney (1R=2) and bladder (11=2).
For leukemia, the excess first appeared 1-4 years after treatment with the greatest excess appearing 10-19 years post-treatment (RR=3.5).
1 large excess of uterine sarcomas was also obse:ved in the F
Ai P,
1:
h FW%
t a
po: ron quo:ATIC3 l
FRIoR 20 yu1LICA:ICH l
radiction-trooted group (12 donths obosorved, 1.5 oxpooted).
The
(
greatest risk was observed in the radium treated group which is N.
not unexpected as the radium needles were implanted around the endometrial wall.
A common feature in all of the studies of women treated for benign or malignant gynecological conditions is the deficit of br, east cancer presumably due to radiation ablation of ovarian function.
It is interesting to note that the protective effect was observed even in post-menopausal women, which suggests non-estrogen mediation of breast cancer development in these l
women.
Another feature is the apparent resistance of the ovaries to radiation carcinogenesis as none of these stud.ies showed an excess at this site, even though the organ received substantial radiation doses.
Patients Exposed to Chest Fluoroscopy Several studies of women exposed to repeated chest fluoroscopies during the course of monitoring pneumothorax therapy for tuberculosis have observed an increased risk of
, breast cancer (MacKenzie, 1965 Myrden and Niltz, 1969 Boice and Monson, 1977 Bowe, 1984).
In the largest incidence study (Boice i
and Monson, 1977), a significant excess of breast cancer (41 s
cases observed. 23.3 expected) was observed in 1,047 women treated in two Massachusetts sanatoria between 1930 and 1954.
ko excess of breast cancer was seen in 717 women who were not treated with pneumothorax therapy (15 observed, 14.1 expected).
The women treated by pneumothorax received an average of 102
?MR D.?""
hki U
~
1:
NOT FOR QUOTATION PRIOR To FUELICATICI
fluoroneopios during tho ocurso of trcotcont, resulting in en average breast dose of 150 rads.and an average dose per exposure
('
)
of 1.5 rads (Bolce at al.,
1978).
The excess risk was proportional to dose with an absolute risk estimata of 6.2 cases per million PYR.
Exposure around the time of menarche resulted in the greatest risk.
The minimal tumor appearance tima ranged from 10 to 15 years after exposure, depending on age at expo'sure and the breast cancer excess was observed in the exposed group more than 40 years after the first exposura.
Evaluation of the modifying effect of radiation exposure and breast cancer host factors suggested a slight moderation of risk by parity status although the numbers v. era too.small to meaningfully ava1unta interaction (Boica and Stone, 1978).
i In a recent Canadian study (Howe, 1984) of 23,318 women treated for tuberculosis in the 1930's and 1940's and who were still alive as of 1950, an excess of breast cancer deaths identified though record linkage with Canadian mortality statistics was observed in the 11,284 woman who received fluoroscopy (RR=1.5).
No excess breast cancer mortality was observed in the 12,034 wcaen who never received fluoroscopy i
(RR=1.0).
In contrast with other studies of radiation-induced l
l breast cancer, the Canadian data showed upward curvature of the dose-response model, and the best fit to the data was obtained by
~
a pure quadratic model with the departure from linearity being most noticable for the Nova Scotia patients, who were the sole occupants of the 500+ dose range.
This was due to these patients facing the X ray tube in contrast to the lateral or posterior R
, f*Y
~
,.. ba 307 FOR QL'UATICE FRIOR 20 FUBLICATION v
Positioning of pationts in 'the other provinoos.
Bolow 500 rods,
(
the standardized nortality ratios for Nova Scotia patients were uniformly two to three times that for other provinces in the four dose categories presented, although at 0 dose, the SMR's were comparable.
The influence of age at exposure on breas't cancer risk was consistant with other studies.
No excess mortality was seen in women 40 years and over at exposure and the greatest risk was observed in women 10-19 years of age at exposure.
The minimal in' duction time for breast cancer ranged from 10-15 years after exposure, depending on age at exposure, and the exposed groups continued to show excess risks up to 40 years post-exposure.
Women Treated with X irradiation for Benign Breast Diseases sEh Mettler et al. (1968) followed a group of 606 women treated
~/
for acute post-partum mastitis with X rays and observed an excess of breast cancer (13 cases observed, 5.9 expected).
A more recent follow-up added three comparison groups in addition to seven more years of risk (shora et.al., 1977).
A total of 36 cases of breast cancer was observed in the exposed group (6.3%)
compared with 32 cases in the various comparison 5:oups (3.2%).
.The average cumulative breast dose per patient was 247 rad and the breast cancer excess was not apparent until 10 or more yeasr after the first exposure.
The dose response curve was consistant with linearity up to about 400 rads with a downward curvature after this dose point.
In contrast to other studies of
~
radiation-induced breast cancer, irradiation after the age of 30 l g' g) 5: V,k )..l l
hb5I g i b i' j
bd NOT FOR QUO 3 TION war ?M en IP'TRT.Te A ?TMIIT
yoars conforrod an absoluto risk of concor induction c2 high as that from breast irradiation earlier.in life (7.9 excess breast
(
)
cancer cases for exposure at 15-29 years of age versus 9.2 excess cases for exposure 30-44 years of age) although the numbers of women over 40 years of age at exposure were small.
An evaluation of 1,115 Swedish women treated by X rays for benign breast diseases, primarily fibroadanomatosis, indicated a four-fold excess of breast cancer compared with general population rates.
The mean latent period, which was 24 years, decreased with incre,asing dose to the breast (26 years in the
<500 rad group 19 years in the >1500 group).
This reduction in latent period as a function of breast dose was likely confounded by a higher radiation dose with increasing age at exposure.
A i
significant excess of breast cancer was observed for women over 40 years of age at exposure.
As there was no comparison group fr3y J, (K.t
'2' free of breast disease in this study, the magnitude of risk is l
likely distor+ed since benign breast disease probably predisposes toward breast cancer.
Parallel analyses of the data from the three major studies of radiation-induced breast cancer (Japanese A-bomb survivors, Massachusetts fluoroscopy patients, and post-partum mastitis patients)'showed consistant patterns of risk in all studies by age at exposure, dose-response and latent period (Land at al.,
1980).
The absolute risk estimates by age at exposure were remarkably similar in all three studies.
What is notable about the similarities in risk patterns is that the mode of exposure in the three populations was very different ranging from an acute Q ', Mh.i 6 R
f I
I f.\\ Y. bbs
.s e
e,..
.C 30; PCR qL*0;ATICE PRICR 20 PUBI,1CA 103 9
wholo-body oxposuro with a high doso-roto of tho Japanoso A-bomb survivors to a fractionated, low-LET dose of the tuberculosis g
\\
patients.
yurthermore, the natural breast cancer rates of among Japanese and western women are substantially different, yet the absolute risks were similar.
patients Exposed to Radium Therapy of the Masopharynx Several reports of head and neck tumors following radium treatment of nasopharyngeal hypertrophy have appeared in the last 2,0 years (Wilson et al.,
1958 Walach et al.,
1975).
Three cohort studies of this association have also been conducted.
Mazen et al. (1967) observed no increased risk of benign or l
malignant head and neck tumors in 417 persons treated during childhood.
However, the size of the exposed group was small and 4r the mean follow-up period was only'15 years which may be too short a time for excess tumors to appear, especially in a group l
whose mean age at exposure was 11 years and therefore, would not have been of an age where the spontaneous cancer incidence was high.
Manning and D' Angio (1967) found no increased cancer incidence in 890 persons treated with radium for lymphoid hyperplasia.
However, 38% of the study group was not followed and no comparison group was identified, which limit the usefullness of these results.
In perhaps the most thorough
~
study, Sandler at al. (1982) found an increased risk of benign and malignant tumors of the head and neck in 904 persons treated with radium for lymphoid hypertrophy of the nasopharynx.
The 2Trgest observed risk was for malignant brain tumors where three m$ }.e n..
} g, '...f* l!
.s-Q
% 0 k0:. Qk h-
- f1 ?.i LJ g
30: TOR CUC ATION FRICR TO FUSLICATION v
~
1 ocson vero obsorvod in the expecod group cnd nono among the controls.
The mean dose to the lower brain was estimated to be
(
)
78 rads.
No increased risk of thyroid tumors, either benign or malignant, were observed.
The mean thyroid dose ranged between 5-20 rad, depending on the age of the child at treatment.
The apparent lack of an association between thyroid tumors and radium may have been, in part, due to the small sample size, as a risk of less than 8-fold would not have been detected in this study with adequate statistical power.
In a postscript to this report, the authors stated that one malignant thyroid tumor'was observed in an exposed patient-following completion of this study,.
While the authors suggested that an. increased risk of thyroid tumors might appear in the exposed group with increased follow-up, it would be difficult to account for the increased surveillance bias in the exposed group following notification of radiation.
1)
Persons Exposed In-utero to Diagnostic X Rays' A case-control study of childhood leukemia and other cancers was initiated by Stewart and her colleagues at Oxford University in 1956.
In the first report from,this study (Stewart et al.,
1958), 1,299 children who died of leukemia or other childhood cancers prior to 10. years of age were matched with live controls on age, sex and place of birth.
The mothers of both groups were
~
interviewed and information was obtained on the frequency of maternal diagnostic radiation exposure during the gestational period of each child.
A total of 178 cases had been exposed to diagnostic radiography during the relevant pregnancy compared deniI so: pon quo:ATICE FRICE 20 PUBLICA Ios
with 93 controls (RR=1.9).
This led to the specuintion that abdominal X ray examinations during Pregnancy which involved exposure of the fetus, contributed to the atiology,of childhood malignancias.
A later report from the same group (Bithall and Stavart, analyzed over 8,000 deaths and incident casas of childhood 1975)
The malignancias and included children up to 15 years of age.
association of prenatal irradiation with. childhood cancer was still present although at a somewhat decreased level than that in l
1 the original study (11=1.4).
In analyzing the data by trinaster an increased risk was observed for 1st trinaster i
of exPosura, which was greater than the risk for exposura in exposure CER=9) the 3rd trimastar, a time when most obstatric radiography The tyya of diagnostic X ray procedura differed by occuzzad.
O 1st trinaster with more multiple films being mada during the k.
?f
';$.i The zaasons for taking the radiograph also differed trimaster.
3 d
by trinaster, as expected, with 30 out of the 38 casas exposed i
i 1st trinastar listing non-obstetric conditions as during the A crude
[ustificationfortheradiographbeingtaken.
dose-responsa relationship based on the number of films taken was 4
observed with the shape of the curve being proportional to dosa.
However, this intarPratation was criticized because the width of the confidence limits didn't rule out other models (Newcomba and McGregor, 1971).
The findings of this study were criticized on the basis of the observed association not being due to radiation exposure but to salaction of the " cancer.Prona" for radiography (Burch, 1970 T
gstaat; u pra:
[j9sb~g r/
307 roR QUC~u!ICH FRIOR 20 FUBLICA2IOS
j Tottor end Macphorson, 1981).
Conseguently, Melo (1974) remnalyzed the twin data fr'on the Oxford survey.
The rationale
)
was that twin pregnancies are selected for pre-natal X ray examinations for conditions directly related to twinning and the effects of selection for adverse medical conditions should be minimi=ed.
In Mole's analysis, the relative risk of leukemia in irradiated twins was 2.2 compared with 1.5 for irradiated singletons.
For solid tumors, the relative risk in twins was 1.6 and 1.5'for singletons.
A recent study of cancer risk in 32,000 twins in connecticut followed to age 15 (Marvey at al.,
1983) observed a relative risk of 1.6 'for leukemia and 3.1 for solid tumors in twins exposed to prenatal X rays.
This study, while based on small numbers of cases, provides fur'Eher support for the relationship between prenatal X rays and childhood cancer, at least for leukemia, and minimizes the " medical selection"
/Wh)
Q:'Q",;
arguments.
Xneale and Stewart (1976a and 1976b) further analyzed the Oxford data for a variety of factors including number of films j
taken, trimester of exposure, reasons for X ray exam, and X ray 4 findings and suggested that even after controlling for these, factors, the association between in-utero X ray exposure and childhood malignancies remained intact.
MacMahon (1962) conducted a study of children born in and discharged alive from 37 hospitals in the Northeast United States-in the years 1947-1954.
A one-percent systematic sample of births from each of the hospitals was selected and the frequency of pre-natal X ray examinations was determined in this sample.
$. 6 ;.4 o. -
- u 57 v* i u u.*
n a
.}
- d p,,
a 30; FOR QUOTATION PEIQR 20 Pu nic121c3 i
locortninnant of nortality waa conducted by sontching for donth certificates for deaths occurring between 1947-1960.
A total of 548 cancer deaths was identified in the sample.
The cancer death rate in the X rayed population was 11.0% compared with 7.3% in the non-X rayed group, resulting in a relative risk of 1.4, which is consistant with Bithall and Stewart's findings (1975).
When age at ienth was analyzed, the greatest excess occurred among the 5-7 year olds.
The association with X rays disappeared after 8 years of age, in contrast to the Oxford survey where the association for leukemia was apparent through age 11, and for solid tumors through age 14 (Bithall and stewart, 1975).
In a extension of,MacMahon's original study, Monson and MacMahon (1984) reported on cancer deaths up through 1967 in 1,429,400 children born in 42 maternity hospitals between 1947 through 1960.
The relative risk for laukania was similar to that found N*
in the original study (1.5).
However, the initial findings of an association between in-utero X rays and solid tumors was not confirmed in the extended study.
The relative risks for tumors of the central nervous system and other cancers were 1.2 and 1.1 respectively, which differ from the relative risk for solid tumors in the original study of 1.4.
The reduction in risk for solid tumors may have been due to the lower X ray dose during the latter period of study (1955-1960).
A similar decline in risk by increasing birth cohort year was seen in the Oxford data (31thell-and Stewart, 1975), where the number of films per examination significantly declined between 1946-1967.
Graham et al. (1966) reported on a case-control study of 313 k
, u. :M
. n LM o
30: FOR CUCIATIDE FRIOR 20 FU3:JCATIO3
childhood loukonio casos dingnosed botuaca 1959-1962 in throo states (Maryland, Meu York and Minnesota).
In-utero X ray
)
exposure was reported 402 more frequently by nothers of cases than of controls.
An association with preconception irradiation was'also reported, a finding not observed in the Oxford data (Xneale and Stewart, 1980).
Sibson at al. (1968), observed a possible synergistic effect between in-utero X irradiation, leukemia and several host. factors i
including a reported histery of previous miscariage and childhood viral diseases.
The relative risks ranged from 1.1 for persons 4
with only one host factor (childhood viral disease) to 4.6 for cases with multiple factors (pr.e-conception and in-utero radiation exposure, history *of miscarriage, and childhood viral diseases).
One possible explanation for these findings is that there is a suscept'ible subset of the population that in the presence of radiation exposure, either pre-or post-conception, has an increased risk of leukemia.
Radiation exposure acts as a triggering mechanism for a seguance of events that lead to leukemia induction.
The host factors (viral diseases and history of miscarriage) are markers of susceptibility and do not increase the risk of leukemia independently of radiation.
Novaver, this theory is suspect as several reports have observed an increased risk of leukemia in persons exposed in-utero to certain virusos and not exposed to radiation (Traderic and Alberman, 1972 Curnan et al.i 1974).
3ross et al. (1972, 1974), reporting on the same Tri-state grouped the cases into two subgroups:. cases which
- cases, DRAFT e
33: POR RUO3TICE F1201 To FU3:.ICATIO5 L
s indicated markors of cuscortibility (n11orgios, viral indootions and bacterial infection histories) and those that didn't.
They indicated a relative risk of 8.4 for allergic children exposed in-utero compared with non-exposed, non-susceptible children.
However, Smith (1972) and Rollockar (1972) both refuta this interpretation.
They concluded that children with preclinical leukemia are more susceptible to viral and bacterial infections because of their impaired immune sys tem and, therefore, are more likely to report these infections than non-leukemic children.
Shiono at al. (1980) observed an increased risk for all childhood malignancias following either pra-conception or in-utero radiation exposure.
An association with both pre-conception (11=2.1) and in-utaro (11=2.2) exposure was observed for what the authors categorized as " medium" dose radiographic examinations (a dose.to the fetus of betwaan 70 to 300 mrads).
No interaction between suspected host factors and radiation exposure was observed.
For y' arsons exposed to both in-utaro radiation and viral infections, the relative risk was 0.4, which is not consistant with Bross's theory.
In a large Prospective study, Diamond at al. (1973) observed both increased total mortality and leukemia mortality (11=3) in white children exposed pre-nataly but not in blacks.
The authors were vnable to explain this. discrepancy.
Even after adjusting for differences in socio-economic factors, the excess mortality remained, although somewhat reduced over the crude risk.
OPPanheim at al. (1974) observed no increased mort 11ty or E
(?h g?)
-iR$)
s
.cr 1 rn h4 W k i
k 3:: FOR QUOTATION s
FRIOR 20 FU3LICA: ION
~
incrocsod ocncor cortolity in 1,000 childron routinoly irradicted in-utero at one chicago hospital in 1948.
The authors suggested
(
I' that because they found no association between in-utero radiation and malignancias in a population which was rouItinely irradiated and, therefore, not subject to the purported selection biases of Previous studies, the reported associations in studies such as the Oxford survey were due to indications for selection for a medical X ray and not to radiation exposure.
Novaver, this study was limited by several deficiencies including that of too samil of sample size.
In order to detect a 50X inormasa in leukemia
~
incidence in the exposed group, which is consistent with the findings from previous studies, the authors would need an exposed group 100 times larger than the one they studied.
A lack of an association between in-utaro exposure and an increased risk of childhood cancers has also been reported by court-Brown at al. (1960) and by Jablon and Kato (1970).
In the lattar study, no increased cancer mortality was reported in 1,292 children exposed in-utaro to the atomic bomb.
The authors estimated that based on the risk astimates from the Oxford data and the fatal doses received by the Japanese, that 5.2 excess cancer deaths should have been observed whereas none were found.
l l
Xnaale and Stewart (1978b) speculated that the apparent deficit of cancers among the Japanese children exposed in-utaro was a j
l l
l l
consequence of the increased incidence of spontaneous abortions and early inf ant mortality in Japan following the atomic bombings.
These children likely had a compromised immune system and consequently, died from an infection prior to the time when 3
k"We\\1.
i 5
t SL G
5 E
FEIOR 20 FLEICATICI
j j
they would norcally dovolop cancor.
This would have tha offoot of removing " pre-cancerous" children from the population at risk I
because of early mortality and would result in the observed 1
I cancer deficit in the exposed population.
However, the f
I l
appropriateness of this interpretation has been challenged by j
i Jablon (1973).
4 Mole (1974) has suggested that the discrepancy between the Japanese and Oxford data is a result of the neutron irradiation received by the Japanese in-utero, which may have resulted in cell killing, thus having the effect of reducing the number of cells at risk of malignant transformation and reducing the cancer rate in the exposed population.
Adult Leukemia and Diagnostic X rays i
In 1962, Stewart et al. -reported an association between i
diagnostic X rays and myelocytic leukenia.
However, the author
_.. *G 1
1ater retracted these findings (Stewart, 1973) as being artifactual because the X ray expo.ure was related to diagnosis i
of the disease process Cleukemia) and, therefore, was not causal.
Gibson at al. (1972) observed an association in males between l
diagnostic X rays and leukemia, both acute and chronic myeloid.
The effect was most pronour.ced for radiation to the trunk, an area where most of the active bone marrow is located.
- However, the results of this study are likely biased as the interviewers were not blinded as to case or control status and discrepant
~
results.were obtained when next of kin interviews were used in the analysis (Lines et al.,
1980).
In a recent analysis of this data, Bross at al. (1979) claimed to demonstrate a dose-response
(
3[hi i
- ..en((
S Sh i
\\
O hhk L
30; pcF. q;C;ATION PRIOR :o FUSLICATION
curvo for cdult loukocin end hocrt disocco in tho ono-red rango.
However, both 3ross's method of analysis and interpretation of
)
~
\\
the data have been critisized (Boice and Land, 1979,).
Ginavan (1980) also reviewed the analysis conducted by Berta11 (1977) on the same data set that Bross at al. (1979) analyzed, and concluded that while relatively high doses of diagnostic X rays
(>20 rads) may be associated with a slight laukania risk, this association is confounded by the fact that the reason for taking the X ray may have baan due to the pre-laukemic stage and, therefora, the association is likely indirect.
More recently, Lines et al. (1980) observed no association between diagnostie x ray exposure and laukasia'in a study where the authors were able to accurately determine the number of X ray films taken, a problem in previous studies that relied upon interview data.
However, the inclusion of chronic lymphocytic laukania, which has
.)
- g-yet to be linked with radiation exposure and which accountad for approximately 33% of the total number of lymphocytic casas, greatly decreased the statistical power of this study.
Therefore, while the ascertainment of exposure was very completa, the study was too small to be definitive.
CANCER AM0XG MILITARY PE150XXIL EXPOSED TO MUCLEAR WZ170M5 TESTING Caldwell at al. (1980) reported on leukemia incidence among soldiers exposed to a single nuclear weapons test (Smokey) in
~
1952, a 44-Riloton, tower detonated weapon.
Troops witnessed the blast from a distance of 29 kilometers.
Some soldiers vara ssnt into ground zero soon after'the blast to either angage in field l
@ 4. -
[
59 fI$f -
N 8
HGlia h
s-.1
- yon quonnes l
.pKICR 20 FUELICATION
nanouvors or to rotricvo contamincted oguipnont.
The conn 1957 cumulative external gamma ray dose, based on film badge readings,
. (
was approximately 0.5 rad.
A list of 3224 man believed to have participated in the smokey test was identified, although the completeness of this list was not determined.
Follow-up of the l
Persons on this list was conducted through a variety of resources.
In addition, 447 individuals contacted the investigators in response to videspread publicity about the study.
A total of 9 cases of leu amia were identified among the 76% of individuals located.
The expected number based on D. 3.
population rates was 3.5 deaths resulting in a relative risk of 2.3.
The interval between the test and date of leukemim diagnosis ranged from 2 to 19 years (mean = 14.2 years) which is not consistant with the short latent period for radiation-induced leukemia in the atomic bomb survivors or the spondylitic g
3' patients.
The results of this study indicated that either the soldiers received higher levels of radiation than indicated on their film badges'or that low doses of radiation were more carcinogenic than previously believed.
One ma$or weakness of this study is that half of the leukemia cases came from the "self-referral" individuals, introducing a large selection bias into the analysis.
Elimination of the self-referred persons reduced the relative risk to a non-significant excess.
yurthermore, the dosimetry estimates were uncertain, since film badges would not have recorded the dose from any ingested or inhaled radionuclides'.
In a second report from this study (Caldwell et al.,
1983),
t e@ c Yh
~
NOI 70R CUOTATION FKICR 20 FUELICATICE h
=
the previously roported loukonin oxcons still persisted with additional and more complete follow-up although there was no
)
excess mortality for solid tumors (56 deaths observed, 61.2 expected, 0/Z=1.0).
Compared with the inital zaport, over 95% of the former soldiers were contacted.
Robinette and Jablon (1983) studied cancer mortality in participants of the plumbob series of weapons testing.
This series consisted of 24' nuclear detonations which occurred in 1957, and includes the Smokey shot' reported on by Caldell at al. (1983).
Over 15,000 participants were identified and followed through the records syste's of the Yaterans administration.
Except for the leukemia excess'present among participants of the Smokey test, no significant excesses of leukemia were observed among participants of the other tests in
-i the plumbob series (18 leukemia deaths observed, 15 deaths expected, 0/E=1.2), indicating that the radiation doses received by the Smokey participants were higher than than actually recorded by film badges.'
Xnox et al. (1983) reported on the mortality experience of over 8,000 former British. servicemen and civilians who were involved with nuclear weapons testing in the South Pacific between 19'52 through 1958.
A 60% excess of reticuleendothelial
~
cancer mortality was estimated.
Iowever, large deficits for cancer deaths of other sites were observed.
These findings are very preliminary as only 300 (<4%) of the respondents were.
included in the analysis.
\\
p.c
'1F'
- Fj;;ar
(
l w.s hN g
g 30 y0R QgonM05 MCCR 20 M2LIUlIQ1
I STUDY OT MARSHALLE5E ISLAXDERS EXPOSED TO TALLOUT l
In 1954, an unexpected wind shift during a nuclear weapons test on the Bikini Atoll in the south Pacific resulted in the accidental exposure of 250 Marshallese PacPla to relatively high levels of whole-body gamma and bata radiation.
The estimated whole body ganna dosas ranged from 14 rads to persons on Utirik to 175 rads to the exposed longelay population.
The thyroid gland doses due to a mixtura primarily of shnzt-lived isotopes of radiciodina ranged up to 2000 rads to persons under the age of 10 years at exposura (Conard at al.,
1975).
A recent za-avaluation of dosimetry estimated that the thyroid gland doses to persons
<10 years of age on longelay Is1.and are now approximately 4000 rads (1.
Conard, Farsonal Communication).
A series of
, examinations conducted over the next 27 years identified a high
, incidence of thyroid nodulas and both clinical and subclinical
~
On longelay, thyroid hypofunction 1n the exposed persons.
subclinical evidence of hypothyroidism was observed at doses as low as 350 rads (Larson at al.,
1978).
Savanty-savan Parcant of the individuals on Kongelap who were less than 10 years of age at exposura developed thyroid nodules.
Four cases of thygoid carcinoma vara observed at all ages in the Kongelay people (6%)
compared with less a one Parcant incidence in the unexposed contro'Is (Conard, 1984).
One case of leukemia was observed in a 19 year old longelaP male exposed to 175 rads of whole body gamma -
radiation at one year of age.
Xo other hematopoietic malignancias vara observed in the exposed Population.
Risk astimates based on the new dosimetry and most recent
~
MAFT c
~
m;; TOR 4403:10E FRIOR 20 F E.ICH ICI n-w
clinicci cxamination cyclo dozenstrato that the cbsoluto risk for thyroid nodulas increases.with youngar age at exposure, while for
(
)
thyroid carcinoma, the reversa pattern was observed (1.
- Conard, Farsonal Communication).
Overall, the greatest risk of thyroid noduits occurred among children exposed under the age of tan.
Only one case of thyroid carcinoma was observed in this age group.
It is likely that the ultimata incidence of thyroid carcinoma was not realized in the axyosed groups due to aggressive treatment of thyroid nodulas with surgery and thyroxina replacement therapy, which has baan suggested to result 1n a reduced thyroid tuacr induction rata (Doniach, 1977).
The high doses of the short-lived isotopes of radiolodine (182I, 188I, and 185I) resulted in relatively high dosas to the thyroid follicular calls which may have produced call-killing.
This would hava reduced the risk of thyroid cancer
)'l
~
in this group.
The dose from the longer-lived and lower energy 46 isotoya, 181I, was only about one-third of the total thyroid dose.
The latant period for the thyroid nodules ranged from 9 to 22 years.
The absoluta risk astimate for thyroid cancer of 3.2 excess casas per million FY1.CBEII, 1980) was similar to that for populations exposed to external beam X rays.
52VDIZ5 0F UTAE EZ5IDENTS ZXFOSID TO TALLOUT T10M i
XUCLEAR NEAPONS TESTING IX MZY1DA Epidemiologic studies of populations in Utah exposed to downwind fallout from nuclear weapons testing in Xavada from 1951
~
.. s.J.
)
,N [.y.t f,wS..
M 7-
$Lh I
E2/ u E0; 70? qJOTATION ygIOR TO PU EICAZ10%
^
through 1960 have baan conducted to invostignto the risks of thyroid disease and leukemia associated with this exposura, t
Thyroid Disease Associated with yallout Exposure J
Rallison at al. (1974, 1975) studied 2495 children be. tween the ages of 12-18 in two counties in Utah and Xavada and 2271 children of the same age in one county in Arizona believed to be free of fallout exposure.
The dose to the thyroid gland from 121I ingested by milk-drinking children was estimated between 30 and 240 rad although the cartainty of these dose estimates varias widely.
Physical examinations of the thyroid gland vara conducted between 1965 though 1971.
Six benign thyroid nodules
- were observed in the 1378 exposed Utah-Xavada children (those who were either born or moved into these two counties before 1961) for a prevalence rate of 4.4 cases par 1000 children.
This compared,with 4 benign nodules in 1313 unexposed Utah-Xavada children (moved into the two counties after 1960), a prevalence rate of 3 casas per 1000 children.
The prevalence rata in the unexposed Arizona children was 2.8 cases par 1000 children.
A non-significant prevalence ratio of 1.5 for benign thyroid nodules was observed for the exposed Utah-Xavada children.
Two cases of thy'roid cancer, one in the Utah-Xavada unexposed and one in the Arizona unexposed groups vara observed.
Xo thyroid unlignancias vara observed in the exposed children.
Based on BIIR III (1981) risk estimates for thyroid nodularity, an excess i
of 73 to 100 radioganic thyroid nodules should have been q.
qis<ih c-43 3
=
ds r
NOT 70R QUCTATION
"~
FRIOR 30 FUBLICATICS
O expected.
One explanation for the failure to observe a significant
).
excess of thyroid nodules in the exposed children is that 131I is not as effective as external X rays in inducing thyroid tumors due, possibly, to the lower dose-rata and non-homogeneous dose distribution in thyroid tissue, as much of'the bata-ray dose is wasted in the colloid and doesn't reach the follicular calls.
Iowever, the results of a recent experiaantal study comparing the relativa effectiveness of X rays and 131I irradiation in inducing thyroid tumors in rats Clae at al.,
1982) indicate that there was no difference in induction rates o'i thyroid cancer betwaan 181I and X rays, implying that the RBE for 181I for thyroid cancer induction is 1 and not 1/10th or 1/70th as some reports have suggested CMaxon et s1.,
1977).
Another possible explanation for the failure to observa an
- $A) increased risk of thyroid nodularity in the exposed children is that the follow-up period (mean = 14 years) may have not baan sufficiently long enough to detect an excess of thyroid nodules.
A second follow-up of the original study groups is ongoing in order to address this issue.
This study will add 10-12 more years of follow-up and should more thoroughly investigata any affects on the thyroil gland from fallout radiciodines.
A re-evaluation of the thyroid gland dosimetry is also planned to address past weaknesses in this area.
leukemia in Utah Residents In a mortality study of childhood cancer in Utah between 1944 through 1975, lyon et al. (1979) detected a two-fold DRAR 302 70R GUOTMION 7EICE TO FU LIOC IC3
incrooso in tho'ogo and 00x-cd$detod nortclity rato in a "high exposure cohort" compared with two " low exposure cohorts".
The investigators defined the high exposure group as persons born between 1951 through 1958, the time of the heaviest concentration of atmospheric nuclear weapons testing in Nevada.
The t'wo low exposure groups were comprised of individuals born either between 1944 though 1950 or after 1958.
For individuals born before 1950, their person years of risk were shifted into the high exposure group as they entered the period of active weapons testing.
The investigators also divided the state into 17 high exposure counties, representing 10% of the state's population and the remainder of the state, including the Salt Lake City area, which contains about 90% of the state's population was designated as low exposure counties.
The exposure status by geographical region was based on isodosa contour maps developed after each weapons test.
No individual desa estimates were available.
It is interesting to note that, both before and after the high exposure period, the childhood leukemia mortality rate in the high exposure counties was substantially less than the low exposure counties and also that of the U.
5.
population.
An opposite trend was observed for mortality from other childhood cancers, an observation that one investigator suggested as being inconsistent with a radiation etiology Cland, 1979).
Lyon et al.
(1980) estimated the eff' acts of misdiagnosis of leukemia on the death certificate and migration of the population on their l
observed increased leukemia mortality.
They determined that l
these factors exerted only a minimal affect and would not, by 4
DRAFT DCT FOR q'JOTATION FRIOR 20 FUBLICATION
..-y.-y--._--_.,
_e.__.,__,.-,.,-,_-y
themselves, account for the observed two-fold incronso in leukemia mortality in the high exposure group.
}
This implication that this study provides avidence for a laukanogenic effect of very low levels of ionizing radiation has been criticized by several investigators (Land, 1979
- Enstrom, 1980).
yirst, the findings in the high exposure cohort were based on only 32 cases of leukemia or less than one leuksmia death per year of study.
Therefora, the study was subject to large sampling variations because of small numbers.
Second, the high fallout counties were largely zuzal and the low fallout counties were urban with a growing population.
It is possible that the results were confounded by urban-rural differences, with the observed association with fallout being an indirect one.
Several investigators have observed' increased leukemia in urbanized areas, especially among younger aga groups (Blair et
,.g)
- sg'!
al, 1980 Kessler and Lilienfeld, 1969).
yinally, individual dosimetry was not available at the time of the initial zaport, so the radiation doses reenived by the leukemia casas were not known.
This study is currently being expanded to include all
~
leukemia deaths for all age groups in Utah since 1944.
In extensive program on radiation dosimetry has been developed in conjunction with this expanded study so hopefully, more precise data on this. issue vill be forthcoming.
Land at al. (1984) re-analyzed a portion of Lyon's data
~
(deaths occurring from 1950 through 1978) and could not confirm the findings of an increased childhood leukemia mortality.
Evidence for,this conclusion restad on two general assertions y,tigK 20 yV LIOC IC3
(1) Book cnd Kray (1983) roPorted that the overogo follout exposures from the 1951 through 1958 test series were somewhat higher in northern than southern Utah, contrary to Lyon's characterization of northern Utah as a " low fallout area" (2) the standardized rate ratios (511's) for childhood laukania mortality were elevated in all 3 geographic areas evaluated in Utah (southern interior, southern border, and northern).
yurthermora, elevated 511's were also observed in eastern Oregon and Iowa, two areas believed to be fraa of fallout, for the same tima period.
Land at al. (1984) concluded that Lyon's findings depended on the extraordinarily low leukemia mortality rate observed in southern Utah betwaan 1944-1949, a period before weapons testing bogan.
Land at al. (1984), however, were unable to ray 11cate these rates as the data.wara not available.
It was suggested that this low rate was due to underraPorting of lauksmia in'the largely rural and undavaloped' southern Part of the state.
This was a consequence of misdiagnosis of leukemia as a result of few board-cartified Physicians practicing in this area at this time (Enstrom, 1980).
Johnson (1984) found an increased risk of cancer mortality among 4,100 residents of southwest Utah and northern Arizona Presumably exposed to weapons test fallout between 1951-1962.
288 deaths due to cancer were observed among this group compared with 179 deaths expected (0/Z=1.6).
Johnson's findings ara
~
suspect for several reasons.
yirst, the cancer casas were all self or surrogate raPorted and were not independently validated, even using available resources such as the Utah Cancer Registry, DRAFT
~
S
. 2.T
- rea q.c:Jt:Ios y
n uon to puzLIcAIIcs u
~5 -
rosulting in an unknown dogroo of nisolassification.
Thic deficiency is further substantiated by the lower than expected
),
overall mortality rate from causes other than cancer, which suggests that some of the deaths reported as cancar vara actually due to some other causes.
Second, only about 60% of the exposed population was located leading to an unknown amount of bias in the study findings.
As a consequence of these major methodologic deficiencias, the results from this study are difficult to interpret.
However, although the study suffers from several methodologic deficiencias, tha. types of cancers found to be in excess vara those normally expected to be so following radiation exposure, indicating that the radiation dosas received by some persons in this population vara greater than previously estimated.
STUDIES OT POPULATIONS OCCUPATIONALLY ZXPO5ZD TO IOMIZING 1ADIATIDX Cancers in Underground Miners l
Savaral studies within the last 140 years have identified alavated cancer risks, primarily of the lung and bronchus and possibly stomach, in underground miners.'
The excess of lung cancer observed in uranium minars in Czechoslavakia, Ontario, anB l
Colorado, in Xawfoundland fluorospar minars, in British hematita and tin miners, and in Swedish iron era minars was attributed to exposure to radon gas released in the mining process and the c:
ymIon 20 pus:ac12IO4 I.'
doccy to its short-livod dcughtor producto, 11470, 11870, 11431, and tityb.
Decay products of radon are alpha-emitters, become C.-
adsorbed to dust particles in the ambient m'ir, are inhaled and deposited on the tracheobronchial epithelium, which is then exposed to continuous alpha-irradiation from the embedded particle.
Among U.
- 3. miners in Colorado who worked at least one month underground and were followed through 1974, 159 lung cancer deaths vara observed compared with 25.2 cases expected based on U.
5.
Population rates JBZII, 1980).
The cumulative exposure was 1,180 working level months (NLM), uh,1ch resultad in an averaga lung deze of between 472-944 rads, depending on a variety of factors including inhalation patterns and thickness of the bronchial epithelium.
The lower exposure groups had risk estimates 2-3 times greater than the higher dose groups, as the dose-response curva was relatively steep up to 500 WLM, and then leveled out.
This pattern implies that alpha irradiation at low dosas and dose-ratas is more afficient in inducing lung cancer than it is at higher doses, possibly as a consequence of call-sparing or a call repair effect.
Although lung cancer risk l
was also increased among the non-smoking minars, the risk I
estimates were not very precise as a very small proportion of the minars were non-smokers.
The length of the induction period (time between the beginning of underground air.ing to data of death from lung cancer) was at least 10 years and was affected by both smoking status and age when mining began.
Smokers had a j
considerably shorter induction period than non-smokers, which
?
DRAFT o
n ; FOR QUOTATION MCOR 20 PUEICCIMI
likely reflects the promoting effect of smoking on induction of lung cancer.
Gottlieb and Eusen (1982), in studying a small
)
group of Navajo uranium miners, SSE of whom were non-smokers, observed a shorter induction period'among the smokers (8 years) compared with non-smokers (15 years).
Age when under3round mining began also affected the length of induction period, which was inversely related to age.
For example, the induction period for lung cancer in men who began underground mining at 25 years of age was 26 years compared with 15 years in men who began underground mining at age 45 or more.
This relationship was consistant among both smokers and non-smokers (Archer, 1981).
The pattern of induction period by age when mining began is likely a function of the normal age-specific distribution of lung cancer, which shows an exponential increase in age-specific mortality after the age of 40 years in males (Burbank, 1971).
.b The difference in induction periods between smokers and non-smokers has not been observed in some studies of Swedish iron era miners (Radford, 1983 Dahlgren, 1979), but has been observed in others (Larsson and Damber, 1982).
This could be a function of the higher dose rates experienced by the V. 5. miners which resulted in an increased potential for initiation of bronchial cancer in the presence of a promoting agent such a cigarette smoking (Radford, 1984).
It is also of i'nterest to note that the absolute risk for the U. S. miners is less than that~for both the czechos'lovakian and Swedish miners.
This could be a dose-rate l
i effect as an increased bone cancer rate from 11*Ra, which has a lower dose-rate alpha irradiation, has been observed.
\\2 s 1-dik ILW k*
p g yon quo:A:10H ygIc1 To yU5LI m-ICE w
w
-vw m -
m er-w M
In o recont study of Swodish iron cinors oxposed to low doses of radon daughters, Radford and St. Clair Renard observed a four-fold excess in lung cancer mortality among miners compared with Swedish lung cancar mortality rates.
A unique feature of this study was that the expected values for lung cancer deaths derived from Swedish mortality rates vara corrected for the effects _of smoWing, resulting in a measure of risk unconfounded by smoking history.
No significant excess of lung cancer appeared among the miners until at least 20 years after the first underground exposure.
Age at initial ex?osure had no affect on j
subsequent lung cancer risk.
A strong dose-response relationship was observed with a risk of 2.4 for exposure to less than 50 NLM and increasing to over a 7-fold risk for exposure to greater than 150 NLM.
The relative risk coefficient among nonsmokers q,.
suggestad that the doubling dose for radiogenic inng cancer may
~ y.fi
- be as low as 10 WLM.
The absolute risk coefficient for lung cancer was 19.2 excess deaths par million person-years, par NLM, which is relatively consistent with the data from the Ontario and cracheslovakian uranium minars, but more than twice the risk for U.
5.
uranium miners.
This discrepancy with the U.
5.
data could be the conseguance of an overestimation of dosas for the U.
5.
minars which would lead to lower values of the estimated risk l
coefficient.
The U.
5.
uranium miners study also had a,much shorter follow-up time than the Swedish study, which could lead to an underastimate of affects among the non-smokers.
Analysis of the joint effects of smoking and raliation exposure revealed that the relationhip was additive, with no difference in lung DRAFT 3
I 30; 70R quo;ATION 4;_
FRIOR 20 ru PICA 21Q4 7-
r cancar induction period between the smokers and non-onokors.
Archer.(1981) combined lung cancer data from several mining
)
yoyulations and observed that the shaya of the dosa-response relationship was curva11naar, with the linear part of the surve predominate up to approximately 250 WLN, with a downturn of the curve after this point.
The absolute risk estimates for lung cancer in the combined mining yoyulations ranged from 22-45 excess cases par million yY1 of alpha exposure compared with 3 exoass cases per million FY1 for the atomic bomb survivors and parsons treated with X rays for ankylosing spondylitis'CREII, 1980).
This ratio indicates that the 1BE for alpha exposure to the bronchial arithelium is betwaan 8 and 15, which is consistant with experimental evidence.
The joint effects of uranium mining and cigaretta smoking have been the subject of debate during the past 20 years.
lundin
)
- =-
et al. (1971) Proposed an additive model as best describing the U.S. Uranium miner's study data, while a recent analysis of the same yoyulation suggest that the relationship is multiplicative (Whittamore and McMillan, 1983).
This finding is also' supported by a recent case-control study of lung" cancer in Swedish underground iron ore miners (larsson and Damber, 1982), although the additive model is supported by the large cohort study of Swedish iron miners (Radford and St. Clair Renard, 1984) and data
~
from a recent case-control study of lung cancer among the A-bomb survivors (B. Blot, personal Communication).
e STUDY 0F XUCLEAR N01XE15 P.,y..%
3 2
yA101 0 IUIL10ATIOi Z
i In 1977, Mcacuno at al. (1977) reported on cousos of donth by cumulative mean, radiation dosas among 25,000 employees of the Manford Works, an atomic plant in Richland, Washington.
Analysis i
of 3,520 cartified deaths occurring prior to 1973 ravaaled that i
workers dying from all malignant neoplasms and, specifically, l
from " bone mezzow" cancers (primarily multiple ayaloma) and cancers of the pancreas, brain, kidney, lung and large intestinas had higher mean cumulative radiation dosas (CMD) than employaas dying from other causes.
The authors estimated the doubling dosa (radiation dose required to double the normal mortality rate for a particular cancer) to be 12.2 rads for all cancer mortality, 6.1-7.4 rads for cancers of the lung or pancreas, and 0.8-2 rads for bone or retf.culoendothelial (RIS) neoplasms.
These doubling dose estimates are much lower than previously reported.
In fact,
. ;r,t the estimated doubling dose for RZ5 neoplasms is agual to the 10%'
laval of background radiation.
Conseguently, the report generated a good deal of controversey.
1 In subsequent reports (Xnaale at al.,
1978 Stewart at al.,
1980 Knaala et al.,
1981), the authors analyzad for the possible effects of several confounding variables including sex, age and year of death, years worked and level of monitoring Caxternal and
' internal exposures).
The ravised estimates of doubling dose for all cancers slightly increased (15' rads) although the 95%
confidence limits vara vida, 2-15C rads, if a more complex model incorporating is: tors for latency and age sensitivity were used.
Further analysis of the lung cancer risk demonstrated that the cumulative maan doses for daaths due to non-malignant respiratory M
DRAFT m
NO 70.e. QUCIATICE yRIOR 20 FURLICATION
discasoc, on indirect nonsuro of tho cifoots of besking, presented no similar patterns to lung cancer indicating that
)
~
cigarette smoking was probably not confounding,this observation.
This particular set of data has undergone several re-analyses as the initial findings of Mancuso et al. (1977) were met with considerable criticism because the statistical analysis of the data was non-conventional and the risk estimates were divergent from accepted estimates of radiation risk (BEII, 1980).
Anderson (1978) was critical of the original Manouso at al.
(1977) report beacuse it failed to aga-standardize the mortality rates and because different revisions of the International Classification of Diseases were used to classify observed and expected causes of deaths which led to errors in the risk estimates for EZ5 malignancies.
These criticisms, however, were addressed by the authors in subsequent analyses of the data.
)
Mutchison et al. (1979) re-analyzed the data from the 1977 report, controlling for the effects of certain confounding variables, using the cumulative dose method.
Of the sites originally identified by Mancuso et al. (1977) as having a higher CMD, only pancreatic cancer and multiple myeloma had a l
significant tast for trend with increasing CMD over several dose intervals.
For both these sites, the significant trend test was based on an excess of cases a=ong workers with 10+ rad cumulative
~
dose.
The authors of this review interpreted the excess proportional mortality above 10 ran for these two sites was possibly explainable in terms of a correlate of dose rather than i
in terms of radiation.
They suggested a cohort study of the j
DRAFFL pc; ycR QUc AT105 PRIoK 20 FUP11cA ICE
Hanford workers, both alive and doconced, would provida a coro thorough avaluation of the data than a proportional mortality l \\
analysis as was performed by Mancuso at al. (1977) and which is subject to several limitations of interpretation (MacMahon and i
i Pugh, 1970).
A cohort analysis of the Ranford workers was conducted by scientists at the Battella Pacific Northwest laboratory (Marks at al.,
1978 Gilbert and Marks, 1979 Gilbert and Marks, 1980).
They reported on a retrospective co.hort study of 28,811 employaas at Manford hired prior to 1965.
Their exposure analysis was confined to 12,522 white males anployed at least 2 years, with monitoring data, who survived until 1955.
Vital status was i
determined through a search of the Social Security Administration i
filas.
Consequently, approximately 6-8% of all deaths in the I
cohort vara.not idsntified.
An additional source of information loss included persons known to be deceased but with no death i
i cartificates, approximately 3% of all white male deaths.
In a later report (Gilbert and Marksi 1980), the authors analyzed 2,216 deaths occurring among monitored males employed at least 2 or more years.
Two sites, multiple myeloma and pancreas cancar had significant tests for trend although these were both based on excessas occurring among workers with 15+ zam cumulativa exposure.
Mone of the other sitas vara in excess, as expected due to the reduction of mortality as a consequence of the healthy worker affect.
The most recent analysis of this dhta (Tolley at al.,
1983), based on 15,992 white male workers employed at Ranford for at least 2 years, observed a significant trend for S
DRAFT NOT FOR CUOTAIION ECOR 20 pmc.ICATICS
nultiplo nyolcan, en in provioup analysoc.
Iownvoz, o trond for pancreatic cancer, which was also observed previously, has now
)
disayyeared, Primarily due to the addition of 4, cases of pancreas cancer in the two lowest exposure categories.
One major problem in all the analyses of mortality among the Nanford workers is that the cause of death as listed on the death certificate was not validated.
Misclassification of death certificates is a non-random event and certain sites such as pancreatic cancer and hematologic malignancias have a higher Probability of errors in certification than other sites (Engel at al.,
1980).
Conseguently, until these causes are validated and death certificate acquisition is completed, the results from the Ianford study must still be regarded as preliminary.
Plans have been formed to validata cause of death in deceased Ianford workers and to mscertain deaths not identified p
i through the Social Security system (Petersen and Breitenstein, 1982).
Follow-up studies of the offspring of the Ianford workers have been proposed to identify iosssible genetic effects associated with occupational radiation exposure.
STUDIES OF MUCLEAR MORMERS IX XAVAL SEIFYARDS A study of mortality among f,ormer employees of the Portsmouth Xaval Shipyard CFX5) was reported by Ma$arian and *
~
Colton (1978).
A search of death certificates in New Hampshire, Maine and Massachusetts for deaths occurring from 1959 through 1977 yielded 1722 certificates that listed employment at FXS.
Next-of-kin were contacted for 592 cases (34.4%) to determine
... = =.
yRIDE2p.,FUELIC47,I%
whothor tho docodont woro n film bcdgo or worked with rediction at PMS.
Expected numbers of deaths were calculated from U.
5.
White mala mortality rates and a proportional mortality analysis revealed an excess of cancer deaths among PM5 nuclear workers (pMR=1.8) but not among the non-nuclear workers (pME=1.1,).
The excess was greatest among workers 60-69 years of age at death.
An excess of leukemia was also observed among the nuclear workars (pM1=5.63.
The call types, hovaver, vara not specified.
The risks observed in this study were unexpected as the cumulative exposures to the workers were thought to be guita low (<
10 ram).
The authors vara unable to evaluate monitored expos'ures for the workers.
The study findings vara limitad by incomplate follow-up, subjective biases from next-of-kin reporting exposura status, and possible confounding of exposures to other CShb
- carcinogens in the workplace such as industrial solvents which
-L may have accounted for the excess cancer mortality.
The findings of this study pron'Ptad a larger retrosrective-cohort mortality study of all civilian PMS workers amployed from 1952-1977 (11nsky at al.,
1981).
This study assessed cause of death among 7,615 workers with external radiation exposure (from film badges or thermoluminescent dosimeters, primarily to Co-60) ranging from 0.001 to 91.414 ran, 15,585 non-radiation workers, and 1,345 workers selected for radiation work who received no measurable exposures.
Vital status of 96% of the workers was ascertained.
The mortality rates in the 3 cohorts were indirectly standardized using'U.
5.
mortality rates.
An internal comparison group was used to DRA_R o
. _ _. = _
McQR :0 PUBLICATIZi
i minimize the influence of the healthy worker effect.
This was done by applying cause-specific mortality rater from the
)
non-radiation workers to the appropriate person-years at risk of the exposed cohort to calculate expected numbers of deaths in this group under the null assumption that their risk of death was similar to the non-radiation workers.
Compared with V. S.
population rates, the.3 study groups demonstrated no excess mortality from all causes, all malignant neoplasms, leukemia or all neoplasms of lymphatic and hematopoietic tissue'.
In.
comparing the risk of leukemia between the exposed and non-radiation workers, 7 deaths du'a to leukemia were observed among the exposed workers compared with 10 deaths expected on the basis of mortality rates among the non-radiation workers.
Finally, there was no trend toward increasing risks with incremental increases in cunulative lifetime radiation doses for l.
anlignancias t.f the lymphatic and hematopoietic system.
Majarian (1983), however, disputes the NIOSH interpretation by asserting thats (1) the XIO5E analysis incorrectly compared nuclear workers for differing work periods, resulting in biased SME'st l
(2) a re-ordering of the dose categories resulted in a j
l statistically significant test for trend for hematologic malignancies and (3) the MIO5E analysis included feaths up.to August 15, 1977, even though 3 additional deaths from hematolo,gic l
malignancies occurred during the remainder of 1977.
Adding these 3 deaths along with the corresponding expected fraction to'the different dose categories, increased the relative risk in the >1 rem group from 3 to 4 which is similar to that reported in the DRAFI NOT PCR QUOTATION yEIOR TO Fu1LICA.TICE
i e
originci Ma$arian roport (Ma$nrinn end colton, 1978).
One ondor i
limitation of the PMS study is that not enough time has elapsed j
to allow the development of malignancies with long induction
\\
periods such as multiple myeloma (as reported by the Ianford Study) or solid tumors.
Further follow-up of these cohorts is desirable in order to address this issue.
A claim that the PMS data support a intge, increase in radiogenic lung cancer,20-200 times that predicted by the BEIR (1980) estimates, has been reported by Bross and Driscoll (1981).
In their analysis, they measured the change in lung cancer SMR,'s as a function of cumulative lifetime exposure and estimated a doubling dose for lung cancer of 12.3 ram, which is consistent with that of Mancuso et al. (1977) for the Ianford workers (10 ren).
Novever, Bross's analysis of the changes in the SMR's as a function of lifetime dose is also consistent with a flat response or no increase.
Furthermore, the lung cancer risk is also likely to be confounded by cigarette smoking or exposure to other occupational bronchial carcinogens.
A large retrospective coho'rt study of workers in 8 shipyards that overhauled nuclear powered submarines or surface vessels is i
now underway (Matanoski et al.,
1982).
To date, a population of l
ove'z 700,000 employees of the 8 shipyards has been identified of which approximately 113,000 were nuclear workers.
A mortality follow-up of this group should eventually address important issves in low-dose radiation carcinogenesis.
s DRAFT V
50: POR QCO:A 103 C'~~
pKIOR 00 FUBLICATICE q
CAMCER MORTALITY AMDMG 1ADIOLOGI535 AMD X 11Y TECEMICIAMS
)I Studies of mortality among radioloists and other physician
(
specialities have been conducted in the United, States and in i
Great Britain.
There are several advantages in studying these groups.
Since the members belong to an active professional
)
society, follow-up and ascertainment of vital status is usually guite complate.
Certification of desth among physicians is usually more accurate than that for the ganaral population, which minimizes the problems with misclassification of cause of death characteristic of other cohort studies.
Matanoski et al."(1975a, 1975b) reported on causes of death among 6500 radiologists and compared.their mortality experience with members of three other' physician specialties that i
represented intermediata and minimal radiation exposure levels.
All cohorts were followed throug'h 1974 (Matanoski at al.,
1984).
.)
Comparad with members of the American College of physicians (primarily internists) and the American Academy.of Ophthalmology and Otolaryngology, radiologists experienced increased SMR's for malignant neoplasms of the oral cavity and pharynx, skin, and lym"phatic and hematopolatic system, particularly acute and i
l myeloid leukemia and multiple myeloma.
The excesses were l
primarily limited to radiologists who entered the society prior to 1940, when, y,rasumably, the radiation dosas were much higher.
~
Radiologists who antered the society after 1940 experienced increased risks only for non-melanoma skin cancer.
One finding from this study which has not been replicated in other studies, vas an excess mortality due to non-malignant conditions, 70 TOR QUCIATION nuCR 20 FWEJLCIDI
indienting o possiblo non-spooifio "Oging" offect of chronic Iow-dose exposura.
This effect, present in both the early and more recent cohorts, was primarily manifested as a slight increase in mortality due to arteriosclerotic heart disea'se CASED).
The risk of ASED was approximately 15% greater among radiologists.
The risk of aplastic anemia was also increased among the radiologists, being observed in both the early and lata cohorts
- (Matanoski et al., 19884).
This pattern of mortality may result
)
from selective factors among radiologists, but these factors have never been investigated."
A study of British radiologists (Smith and Doll, 1981) analyzed causes of death among 1,338 members of several British radiologiesi societies followed through 1976.
Excess mortality
?
from malignant neoplasms was reported for members who antered the societies prior to 1921 but not for the more recent entrants (post-1921).
The excess in the early cohort represented"a 75%
increased risk compared with the rates derived from other medical specialties.
Those sites that appeared in excess included deaths from cancers of the lung, pancreas, skin and laukania.
In this study, there were no excess deaths due to non-malignant conditions in the early (5MR=0.9) or late (5MR=0.8) cohorts.
Consequently, the findings do not support the concept of a non-specific aging effect of chronic radiation exposure at suggested by the study of American radiologists (Matanoski at al.,
1984).
However, the specific causas of non-cancer deaths were not evaluated in the British study.
Since the American radiologists experienced.only a 152 excess of ASND, an effect of DRAFT o
M FOR quCIATION yF. ICE 20 yVII.101 ICK
b this size might have been diluted in the British mortality analysis which was limited to total non-cancer mortality.
)
Conseguently, the findings from both studies might be compatible if specific causes were analyzed.
There are several limitations to both of these studies that restrict their interpretation.
First, no individual radiation dosimetry was available so the doses are unknown.
One can only roughly approximate the aggregata doses, especially to the earlier schorts who presumably were exposed to higher doses because of poor technique.
In these cohorts, dose estimates from 200-2000 rads have been reported (Lewis, 1963).
Another aspect that limits the interpretation.of these findings is the lack of information on host factors such as smoking, that may affect the level of risks observed in these studies.
Another study, which mailed guastionnaires to members. of the American College of 4)'
99 Radiology and of the College of American pathologists, has yet to report its findings on mortality and morbidity (Jessup and Silverman, 1981).
Jablon and Miller (1978) reported the findings of a mortality follow-up of 6,500 U.S. Army X ray technicians trained during World War II and compared their experience with that of
~
6,800 Army medical or pharmacy technologists trained at the same i
time.
The overall mortality rate among the technicians was 12.1%
compared with 11.62 among the unexposed technologists.
No excess mortality from malignant neoplasia was observed among the s
technicians with the exception of leukemia, where the X ray technologists had an 802 increase in risk compared to other 31 F
NC7 FOR QUCIATION FF.IOR 20 FU3L,ICA (0,{
Inboratory technologiota, although tho difforonco una not
(
statistically significant.
The radiation doses to the X ray 4
! (
technicians were unknown but they were presumed to be small as occupational exposure lasted for less than 3 years on the j
aver' age.
Only 19% of the group continued,in this occupation
{
after their discharge from the azay.
1ADIUM DIAL PAINTERS Several studies have reported on cancer incidence andl l
mortality among workers exposed to azola and taats, primari'ly female radium dimi painters who ingested large amounts of paint l
containing radium by licking their paint brushes to maintain a fine tip.
21sta is a bone-seeking alpha emitter that deposits evenly throughout the bone matrix, irradiating the endosteal cells.
As of 1980, about 50% of the workers had measurements of radium body burdens.
Part of the exposure received by this group A
was from 1181a, which has a substantially shorter half-life than
, 228Ea but delivers 3 times the alpha-energy per radium aton decay.
C o ns e que nt '
'ha dose response model for osteosarcoma developed by Rowlan_
1 Lucas, (1984) allowed for the systemic intake and decay of both isotopes of radium.
The function, I= (c + #D1)e-7D (where I is the number of bone sarcomas / person-years at risk, C is the expected,naturmi incidence of bone sarcomas / person-year in the population under study, and the radium insult, D, is expressed in teras of systemic intake in units of radium i
DRAFT F
3CT FCR IUCTATICN pKICR 20 yV3LICATICE g,
e.
cctivity), host fit the obsorvod bono strocco diotribution.
Tho dose-squared portion was predominant up to 1200 Ci of
- Ra
}
systerio intake with a downturn in the curve af ter this dosa.
Tor radium-induced malignancies of the mastoid and paranasal sinuses, a simple linear relationship I= C + ap Provided the best fit to the data (Rowland at al.,
1978).
In addition to osteosarcomas and malignancias of the mastoid and paranasal sinuses, tumors of soft tissues have been reported among the radium dial workers.
polednak at al. (1978a) observed a significant excess of colon cancer among 634 radium dial painters first employed prior to 1930 (when the heaviest ingestion occurred).
Adams and Bruas (1980) reported an excess
- ]).J Q'1 of breast cancer among 799 workers first employed prior to 1930 and with measurad body burdens.
This excess was limited to woman with >50 microcuria systemic intaka dose, and it was aspecially notable for nulliparous women in this exposure category.
The average time of aPrearance for the breast cancers among the dial painters was more than 40 years after first exposura.
The external gamma-ray doses, which were suggested as the primary reason for the breast cancer excess, were unknown.
A breast
~
cancer excess was also observed among British radium workers (Baverstock at al.,
1983).
The excess was limited to women less than 30 years of age at the start of employment (and presumably, exposura) who received over 20 rads of gamma-ray exposure.
DNTE sc; Tcn (UCTATICE i
3310R ;0 PWCJC1:103 j
Follow-up of so'it tissus naligncnoios among the Accrienn workors is still ongoing.
(
Multiple myelona has been reported in excess among'the U.S.
radium dial painters (Stabbings et al., 1982). 'This excess (SMR=2.5) is consistent with the review of radiation-induced myelomatosis reported by Cuziek (1981).
Polednak (1978b) analyzed patterns of latency as a function of dose and age at exposure on the risk of osteosarcoma among the radium dial painters.
He found little evidence for such relationships.
Although the mean latent period was shorter for women with body burdens over 1000 C1, this was possibly a function of the force of competing risks, which resulted in early deaths among the more heavily exposed from causes other than osteosarcoma.
other possible effects that are being currently pursued include those on the offspring of the dial painters who were exposed in utero to both gamma-rays from radium daughters fixed in the maternal skeleton and alpha-particles from maternal radium 1
that crossed the placenta and decayed in the fatal soft tissue or skeleton (Rowland and Lucas, 1984).
This represents more than 3,000 children exposed to a skalatal dose ranging from 1 to 10 rem and a whole body dose estimated to be from 10-100 ram.
STUDIES OF P0yULATIONS EXy05ED TO yLUTONIUM
~
5tudies of workers exposed to plutonium (pu) during the procass of nuclear weapons construction have attempted to measure
~
possible health effects as a correlate of body burdens of pu-238 R
L' g,Y
.R 1
l{
I"O i
m ro.u===
MUOR 20 M!BLICATION
-+%- -,_-._-.,. _,,-_..,,, -
y._.
--,._--,_____m_.
and pu-239.
Studios of boogio dogo (park ot al.,
1972) hevo shown a'high incidence of lung cancer following inhalation 0.2 ci
)
of puos.
Bone cancers have also been reported.in beagle dogs following either inhalation or injection with pu (Huggenburg at al.,
1983).
In dogs exposed to pu skeletal doses ranging from 5-350 rads, 18 percent developed bone tumors within 8 years of exposure.
No difference in bone cancer risk was observed between the inhalation and injection mode of exposure.
The results from this study indicate the the beagle dog is highly sensitive to the skeletal effects of irradiation from either pu or 118Ra, as the 4
ratio of dog to human data Chased.on the radium dial painters experience) indicated that dog'had 8 times the number of bone cancers per rad of exposure.
These data also indicate that beagle dogs injected with 239-pu had almost 30 times more bone (jjj )
cancers than those injected with 1**Ra (Muggenburg, et al.,
au 1983).
Based on these data, the risk estimates for bone cancers from inhaled pu in man were calculated to be 1200 bone cancers 1
i per million person rads.
In man, the primary routes of entry of j
pu into the body are through inhalation of particulates and j
contaminated wounds, ingestion accounting for only minimal exposure.
Autopsy tissue measurements of pu in workers have I
demonstrated that the largest concentrations are present usually in the tra'cheobronchial lymph nodes followed in decreasing order by the lungs, liver and bone.
Consequently, the organs of greatest interest in terms of plutonium carcinogenesis include l
the lymphatic and hematopoietic system, respiratory tract, liver and biliary tract, bone, and, to a lesser extent, the b
3c7 70R QUOTATION FEICE TC FU5:,ICATIc5
gastrointactinni tract.
j A study of mortality among 224 Los Alamos white male workers t
\\
with 10 or more nanocuries of pu exposure has shown no excess deaths due to any cause compared with U.S. White male mortality l
rates (Voelz et al.,
1982).
We excess lung cancer was o,bserved in this cohort although some risk models have predicted over 202 of the workers should have developed lung cancer.
Among 26 males
~
l who originally worked on the Manhat, tan project in 1944 and 1945 and who were exposed to relative high levels of pu, no excess mortality has been observed (Voels et al.,
1979).
The body burdens for these workers ranged from 7 to 240 act of Pu.
However, it is laprobable that risk estimates can be derived from such a small study.
Gofman (19813 has been highly critical of 4
the efforts made to study pu workers.
Ne has estimated that over 100,0C0 lung cancer deaths have resulted in the United States j,
since the 1950's as a conseguance of Fu exposure from atmospheric nuclear weapons testing fallout.
Novaver, empirical proof of i
this claim will be nearly impossible to obtain as his estimates are confounded by the effects of cigarette smoking.
A study of malignant melanoma among Los llamos workers failed to confirm an earlier reported excess among employees of the Lawrence Livermore Xational Laboratory Cactuavalla et al.,
1982 Austin at al.,
1981).
Among the 11,300 Los Alamos employees, 6 cases were observed compared with 5.7 expected based ~
on incidence rates for the state of Xew Mexico.
In the Livermore study (Austin et al.,
1981), 19 cases of melanoma were observed among 5,100 employees compared with 6.4 cases expected, a 3-fold Y
Q dr
,sc7 POR quCTATION yRICR 20 PUBLICATICX 2fu
1 I
oxcons.
No explanation uns Opparont for the discroyant findings between the two studies, other than that they might result from i
)
\\
personal factors rather than occupational exposures.
In the Livermore study, no association between gamma exposure and melanoma incidence was observed.
In fact, the inverse relationship was found with the sentrols having higher gamma exposures than the cases.
Both beta and~ gamma exposures ware l
investigated-but no correlation with the excess melanoma rates was noted.
Only one $ob category, chemists, had a greater than expected melanoma incidence (11=7.0).
Analyses of mortality among employees of the Rocky y1ats 1
plutonium plant in Colorado revealed an excess of deaths due to intracranial s (5ME=2.1) (Yoelz et al.,
1982).
1 case-control stsdy of 16 brain tumor deaths occurring among Rocky i
Tlats male employees was conducted to futher investigate the
...,77 3 relationship between these cases and radiation exposure Clayas, et al.,
1983).
No significant association between pu exposure and risk of brain tumors was observed in this study.
The relative risks were all less th'an 1, regardless of which control group was analyzed or whether time limitations (pu body burdens before death) were imposed in the analyses.
A study of cancer ' incidence and mortality among male residents of Los Alamos County, New Mexico, demonstrated a Possible excess of reticuloendothelial system neoplasms compared-with several control counties (Stebbings and Toelz, 1981).
The incidence data suggest that the excess,.if real, is no longer occurring.
Mortality from colo-rectal cancer was also greater ja
.=p..=.
FRIoa :0 PUBLICACC3
w--------.
'8
among residento of Loc Alamos county, although socioecononic
(
factors may explain this excess., Lung cancer incidence or mortality rates were not significantly elevated.,
In summarizing these various studies, no organ site has shown consistently elevated risks among humans occupationally exposed to plutonium.
Iowever, the numbers studied to date are small and further follou-up is necessary to more precisely define the risks from chronic or acute plutonium exposure.
These studies are important because Fu exposure has significant public health implications, not only for the occupationly exposed, but also for the general population, since Fu may one day play a more direct role as a nuclear fuel.
It already is of potential importance as a possible source of Izzga population exposures,
because of the nation's increasing supply of high-level nuclear (f[.},
wasta materials for which a satisfactory storage scheme has yet b-to be implemented.
STUDIES OT OTHER OCCUPATIONALLY EXPOSED GROUPS Various studies of workers exposed to radiation, primarily persons exposed to uranina dust during processing of nuclear fuels fabric'ation, have been reported.
One major limitation of these studies, at least in the reports presented to date, is that the risks are not presented by radiation dose, but by job classification, which is inadequate for precisely estimating radiation risks.
Polednak and'Frome (1981) reported on the mortality Wo! FOR QUOTATIUS l
FF.IOR 20 FUBLICATION
oxporionco of 18,869 whito colos coployod botucon 1943 nnd 1947 in a uranium processing plant.
Mortality follow-up though Social
.I security Administration records revealed an exce'ss of lung cancer (RR=1.5) among a subgroup of workers exposed to alpha and beta radiation and electrical workers who were 45 years of age or older at time of hire.
There was no excess of bone cancer, solid tumors, or leukemia in any of the job classification groups.,
continued follow-up of this group is ylmaned to eval'uate further Possible effects of exposure to uranium dust.
In another study of workers exposed to uranium during the process of nuclear fuels fabrication, no excess mortality or cancer incidence was observed in those job categories where uranium exposure was identified (Radjimichael at al.,
1983).
4
}
Nowever, only a small proportion of the workers was exposed either to whole-body X-or gamma radiation or to alpha radiation
[ $. )
No' as determined by film badge monitoring or urine bioassays.
Consequently, only a very large risk would have been detected in this study.
I i
schofield (1982a), in a preliminary report on nortality among 41,000 current and ex-employees of of a nucinar fuel j
I6 Processing plant in Great Britain (BMTL), observed no excess deaths from any cause in either category of' worker (active or retired).
The findings were consistent with the healthy j
worker effect appearing in previous studies, as manifested by a decreased SMR among current employees.
The force of this effect tended to disappear among the retired employees for whom SM1's on the order of 1 were observed.
No excess was observed among t
00 NOT FOR QUOTATION l
FKICE 20 FUBLICATION I
onployoos of tho Sol'1cfiold pitnt whoro the highost oxposuro readings were record [d.
Further follow-up of these workers is planned and additional reports should appear,in the future (Schoefield, 1982b).
Bross and Driscoll (1982) interpret the BMTL study as showing an excess of lung cancar if the SMRs of the retired amployees (1.0) are compared with that for the current employees (5M1=0.7).
They claim that this 45X excess lung cancer mortality is due to the longer period of. latency experienced by the retired workers, although this discrepancy is observed betwaan current and ex-employaas for all causas of death, both cancer and non-cancer, suggesting that it is the result of a reduction of the healthy worker affect among the retired employees (Schofield, 1982b).
STUDIES OT popVLATIONS EXPOSED TO BACXGROUND RADIATIDX t*?
During the last 20 years, several studies have been conducted of populations residing in areas with high levels of natural background radiation.
Their results have baan inconclusiva, a functio,n of the different study methods and various design weaknesses, which are common to most correlational studies.
Segall (1964) observed no patterns of increased leukemia mortality associated with background radiation, Primarily terrestrial gamma radiation from grantic deposits in three Maw England states.
However, there vara few differences in background radiation levels in these three states.
Tasting'the hypothesis of increased cancer incidence as a function of b
.=p
.==--
EFl0E TO PVELIMIICS l,
bockground radiation in the Unitod Staton io limited by numsrous practical problems.
These include the large amount of
}
- background noise" such as other environmental carcinogens, which complicate the intaryzatation of any observed associations, and the relatively small variation in levels of background radiation.
In certain othat areas of the world, however, there are larga differences in background radiation levels due to unusual
~
geologic formations.
Such a place is the Kerala coast of India, where the presence of thorium-containing monarite sands results in a population exposura up to 1300-3000 mrams per year, a level over 10 times that found in Colorado, which has the highest levels in'the United States.
S'tudies of yoyulations residing in the Kerala region have shown marked cytogenetic effects, particularly Down's Syndrome.
While no marked differences in the frequency of chromatid exchange aberrations were observed between
- .j'lh) p the exposed and comparison groups, there was a higher rate of chromosomal aberrations, primarily deletions, among the exposed population (Kochupillai at al.,
1976).
An increased risk of i
l stomach cancer was observed among the residents of 6 villages in Brittany, France (Fincet and Massa, 1975).
The Primary source of exposure there was from uranium-containing granitic deposits.
However, no excess cancer nortality associated with backgr.ound radiation was reported in studies of the residents of rural Ireland (111 wright at al.,
1983) or of Guangdong. Province, China (Luxin, 1980).
Recently, Ladford at al. (1983) observed an increased prevalence of thyroid nodules in a community survey of long-term residents of two small Pennsylvania towns.
One Wa r
)
Os k-
- 9L n
q l
\\Ll~tt s
3: yog gg :ATION yr.:cF. :o Fuc.I W ION
Population zoco'ivod exposuro to gamma Indiction'frco o nontby uranium vasta site, with radiation levels ranging from 15-40 microRens per hour.
At highest risk were individuals who lived in the area prior to the age 15, which is consistent with other studies of radiation-induced thyroid disease.
While the data are 1
suggestive, they avait further confirmation by larger, carefully i
conducted investigations in areas where there are clear variations in background radiation levels and a relatively stable j
population.
The thyroid gland is a suitable marker to measure the effect of background radiation because of its increased sensitivity to ionizing radiation and because of the relationship between age at exposure and risk.
One source of environmental radiation that has generated 3
recent concern regarding the risk of lung cancer is the effect of l
indoor radon concentrations.
In one U.S.
study (Radford, 1984)
?
the measurements of redon levels in the homes of lung cancer i
ca,ses were found to be higher than in the homes of controls.- A correlation study in Maine demonstrated a significant association between lung cancer and radon levels in well water (Hess et al.,
1980).
Axelson et al. (1979) found an increased frequency of reported residence in stone house dwellings among lung cancer cases than in controls in a study in Sweden.
The measured levels of radon daughter products was higher in stone houses than in wooden ones.
Further evidence for a raden-lung cancer association is perhaps demonstrated by the 4-fold urban-rural gradient among men and the 2.5-fold gradient among women in Sweden for lung cancer rates (National 3, card of Health and OhmI h
n 3C: FQF.qu0 A:10E y7.10p. 20 Fu CIC1:;3 1
Welfare. 1971) which is too large to be explained by differences in smoking' habits.
(
-I Although, at this present juncture, the data on lung cancer and indoor radon exposure are equivocal, this is an inyottant j
public health concern that will demand further attention, j
especially in a society that has become more energy-conscious and j
is restricting air-exchange in homes through more efficient i
insulation techniques, thus raising the indoor radon concentrations.
Recent calculations, derived from lung cancer I
risks in uranium miners, estimate that as much as one third to one half of all lung cancer cases among non-smokers and a small proportion among smokers can be attributed to indoor radon exposures (Narley and Fasternack, 1981).
.}
I IV.
SUMMARY
AMD CONCLUSIQX5:
DIRECTIONS FOR FUTURE STUDIES l
I The large body of both experimental and epidemiologic data on the carcinogenic effects of ionizing radiation allow certain generalizations to be made about the current state of knowledge I
l t
of radiation risk.
1 RADIATIQX INDUCED CANCERS ARE INDISIIMGUI5NAALE FROM.
Sp0XTAMEOU5 MALIGNANCIZ5'.
In studies of yoyulations exposed to 3
ionizing radiation, the pathology or histology of malignancies 4
30; FOR QUcn ;;05 yKics. To Fu1L101:14 l
occuring in the exposed groups is sicilor to those occurring in
/
the non-exposed population.
It is only on the basis of a
(
statistically increased frequency of malignancias among the exposed in contrast with suitable comparison groups coupled with she observations of the tima-and age-dayandant distributions of malignancias that radiation itself can be indicted as a carcinogan.
MOST TYpIS OF CANCER 5 1RE INDUCED BY 20MIZING RADIATION.
With the possible exceptions of chronic lymphocytic laukania, carvical cancer, Modgkins Disaasa, and a few others, ionizing radiation increases the frequency of all other types of malignancias.
The case of chronic lymphocytic leukemia is unique, as an excess cf all other call types of leukemia has haan observed in the Atomic Bomb Survivors and in persons irradiated for treatment of ankylosing spondylitis.
This could be due to the increased sensitivity of the transformed lymphocyte to the call killing effects of ionizing radiation so that killing of the transformed call occurs in persons who have early chronic lymphocytic laukan'ia.
RADIATIOM-IMDUCED MALIGX1MCIE5 MAYE A CHARACTERISTIC INDUCTION PERIOD.
The temporal distribution of tumor development following exposure to ionizing radiation has baan observed to conform to I
cartain characteristic patterns.
Laukasia is the most notable of -
the radiogenic cancers with the inital excess becoming evidcnt 2-4 years following exposure, a peak excess occurring about 6-8 uTu I
i' n 5
k 93 4
f a NN
{
NOT TOR ("0 ATIO;i FRIOR 20 FUBLICATION
-. =
. = - - -.
)
yoors citor cupssure, a pintocu paried extending fren S-25 yaora,'
followed by a decline to nearly normal levels 25 years after exposure.
These observations are derived from the Atomic Bomb survivors and patients treated for ankylosing spondylitis.
For solid tumors, e6ch as the breast and thyroid gland, the minimal I
induction period appears to be at least 10 years after exposure, and is dependent on age at exposure.
For human data, if dose affects the induction paziod, the effect must be relatively small.
DITTE1ENT ORGAX5 kAVE VARYING DEGREES OT SEMSITIVITY WITX REGAR) TO RADIATIOX-IMDUCED CA,1CIXOGEXI5IS.
Studies of populations exposed to ionizing radiation, either partial-or whole-body, have demonstrated that the breast, thyroid gland, and bone marrow have the highest degree of sensitivity to radiation
,,7
)
carcinogenesis.
The relative sensitivity of the various tissues and organs can be measured in terms of either a proportion of the natural risk (relative risk) or as an increment above it l
Cabsolute risk).
The seguance of sensitivity depends on whether the relative or absolute measurement is used, with the thyroid
' tissue having the highest sensitivity under the relative risk measurement and the brecst being highest under the absolute risk measurement.
This reversal in order reflects the much higher incidence rate of breast cancer.
Tissues that have a relativelf Iow sensitivity include the kidney, bladder, prostate gland, and ovary.
I N C; FC R Q'w'C IAT ICI FKICF. To FUBI101:IC5
.---e~
-m ee.-w.<<
--w-
_ _ _ _ ~ _ _ _ _,,.
O t
~
VARIOUS MOST TAC 3015 MAY IMFLUZMCE THE PATTERM5 0F RADIOGEMIC CAMCER RISK.
Although ou'r knowledge of the effects of various host factors on modification of radiation response is far from complete, various studies have identified both basic biologic factors such as hormonal status, immune competence, and BMA repair mechanisms, demographic factors such as age, sex, and ethnic background, and personal factors such as cigarette smoking ss being possible modifiers of radiogenic carcinogenesis.
Age has been identified as an important factor that influences both the induction period (as previously discussed) and level of risk, Radiogenic tumors such as breast 'and, more recently, thyroid cancer, have an inverse relationship between age at exposure and level of risk.
The risk coefficent for breast cancer for exposure at 10-19 years of age was estimated to be 7.3 excess cases per PY.R compared with 6.6 excess cases per PYR for exposure u
over age 20.
For thyroid cancer, the relationship is less clear, as most of the data are derived from individuals who were exposed either in infancy or adolescence.
Only the Japanese A Bomb survivors have a wide enough age distribution that would allow an evaluation of the age effect.
In the most recent analysis of those data, both the absoute and relative risks ~were higher in the <20 years ATB group compared with the >20 ATB group.
One of the major problems in identifying the ultimate role
~
of age at exposure, as it influences cancer risk, is that most of the studies have not yet continued for a long enough time to characterize the lifetime risk of cancer.
As the younger cohorts enter the ages'of high natural cancer incidence, additional 7
eg.
a bsFbIi 13 ECT FOR QUCTATION
==mv w. witar.;c:TIcx
i fo11cu-up is roguirod to nero procisoly gucatify radiogonic cancer risk for solid tumors, as the peak for leukemia appears to
{
have now passed.
j Sex does not seem to be an important modifer of risk, at 1 east in relative taras of measurement.
Iowever, the ab' solute risks are higher in fanales for thyroid and breast cancer, reflecting the auch higher natural incidence of cancers of these sites among females.
Smoking is an important host factor for radiogenic lung cancer, although the nature of its modifying effects on radiation-induced cancer is not clear.
Data from various studies have shown either an additive or multiplicative (interactive) effect of smoking and radiation.
It has recently been suggested that the apparent multiplicative effect of smoking among miners i
any be an artifact of insufficient follow-up time, and that when N
l E'
the follow-up period is long, such as in the Swedish iron ore ainers, the interaction largely disappears and the :loint effect appears to be additive (Radford and 5t. Clair Renard, 1984).
Other factors that have been suggested to modify radiation risk includa genetic fac' tors, such as ataxia talangiactasia, a DNA-repair deficient condition, which confers increased radiation sensitivity, and such conditions as childhood allergies and infectious diseases, which may increase host suspectibility to the affects of fatal irradiation.
These factors all point to th~e existance of possible radiation-sensitive subgroups of populations that may account for a disproportionate fraction of radiogenic cancers.
m sI t'
LNihME
-s
- pawnm, t-Fr.IoK 20 F E CA 105
- -m-wwo,na--m
.w.,,
- - -,-n-,--
- - - - - - ~ - - - ~ ~ -, - ~ - - - - - ~ ~ = ~ ~ - - ~ - - * = ~ ~ " ~ ' - ' " " '
~ ~ " ' ' ~
MODELS OT RI5K pKOJECTION N
The models used to project radiation risk beyond known observation periods or to apply from one population to another are the absolute risk and relative risk models.
The absolute risk model assumes that radiogenic excess of. cancer above that normally expected is independent of the natural incidence rate of However, the absolute risk appears to increase with age cancer.
as demonstrated by the Japanese A Bomb survivors, where the absolute risks for all cancers except leukemia have become auch larger as the exposed population ages, and, therefore, with the concommitant increase in the natural cancer incidence rate.
The re'lative risk model, which assumes that the radiogenic risk is expressed as a multiple of the natural age-specific cancer rate.
is more stable as the population ages.
Use of the relative risk
((!;.}f model, which is now coming into favor based on recent analyses of the A Bomb and ankylosing spondylitis data, results in higher risk projections than the absolute risk model.
None of the radiation studies have long enough follow-up yet to determine which model is more appropriate for risk projection.
It'is likely that neither of the models is entirely applicable for risk projection for all tumors.
This is an area that avaits further refinement as more data are gathered from studies with long-tera follow-up.
~
Dose-Response Relationships The relationship between radiation dose and a biological c
M, t
.s PdT i
nL k
f[
- w. u ;u!,
g(.
$[$.
4 e
u n: rox t.c:A:Ios 7RICR To PUBLICA:ICS
)
i
~
- ' ' ~ ~
- ~ ~ ~~ ~^ ~' ~
~~
~~
e response such as cancer is subject to several sources of
('
. variation.
These include the type of response observed, the conditions of exposure, and variables affecting suspectibility such as the host factors previously discussed.
The wealth of l
experimental literature indicates that dose-response relationships for mutation freguancy'or call transformation are linear or'superlinear (larger proportional effect at low dose) and non-threshold in nature.
The differences in dose-response curves for low-LET (X rays for example) and high-LET (alpha rays) i radiation possibly reflect differences in the repair process occurring following low-LET radiation.
The mode of exposure (fractionated, acute, or protracted ) also affects the slope of the curve with dose fractionation yielding a higher frequency of l
mutation per unit dose than chronic or acute exposure of low-LET
['ll; }
M[e:
radiation (gamma rays).
The dose-response curves for most radiation-induced neoplasia in experimental animals are consistent with the dose response curves for induction of mutations and chromosomal aberrations.
They may be characterized as follows:
(1) the slopes of the curves for high-LET radiation are steeper, more near1y linear, and less dependent on dose rate than are the curves for low-LET radiation and (2) the curves for low-LET radiation tend to increase in slope with increasing dose and dose rate.
One problem with experimental data on I
dose-response is that in most experiments, the shape of the curves are based primarily on the initiating events and do not involve tumor promoters, which can alter the nature of the dose-response r'elationships (Try,'1984).
The situation regarding 7 A. f Y;
W' FC* FCR CUc;ATION n.IOR ;o PUBLICATIC5
I dose-response models for human carcinogenesis based on l
epidemiologic data is much less certain dua to the vagaries and incompleteness of the data.
Munan dose-response data are complex and, ideally, should consider many factors.
These include, the quality of radiation exposure (high versus low-1ET), mode of exposure, susceptibility of the host, including such variables as age, sex, and others, and the duration of follow-u.p.
The complexity of the appropriate models, the many and sometimes unknown assumptions upon which they rest and the uncertainties of i
the human data for model development were highlighted by the level of discussion and dissension of the various members of the BEIR (1980) committee over what was the single most appropriate model to use for risk projection.
That this was not a moot excercise is reflected in the magnitude of difference in the d,.k, projection of excess lifetime mortality following exposure to ionizing radiation, demonstrated by the data in Table is Table 1.
Estimated Excess Mortality Trom all Cancers Per Million persons, Single Exposure to 10 Rads Low-LIT Radiation (BEIR, 1980)
Dose-Response Model Absolute Risk Relative Risk Linear 1671 5014 Linear-Quadratic 766 2255 Quadratic 95 276 5@
3 Er'E l ' L
/
.. l,..
I Es a,[*t $;i'g-. l.<i0
(
ft i f
b
/C/
m rw c=xnc FRIOR 20 PUELICA~IC3 4
An 18-fold difforonco in tho nunbor of oxcons concor donths i
exists between the linear and the guadratic model projections.
I The history of risk projection has been one marked by uncertainties and controversies over two general areas, the i
appropriate dose-response model to use for risk projection (linear, linear-guadratic, or pure guadratio) and the use of the absolute or relative risk model.
The evo,1ution of risk projection over the last 2 decades is summed up in Table 2.
l Table 2.
Comparative Estimates of Lifetime Excess cancer Mortality Induced by Low-LET Radiation-Excess Deaths per Million Persons per Rad source Model single Exposure, continuous Lifetime
)
10 Rads Exposure, 1 Rad /Yr ABS REL ABS REL BEIR (1972)
Linear 117 621 115 568 UX5cEAR (1977)
- 75 - 175 167 501 158 403 3EIR (1980)
=
Lin-Quad 77 226 67 169 Quad 10 28 GOTTMAX (1981) Linear Relative 21sk--3,731 The obvious discrepancias between the different risk, projections are due to the use of different assumptions concerning the nature of radiation carcinogenesis.
These include
- 9. r.. @. VO7, I
b
.s d
d [i ;
I l0k y
Ej"ka.d j '.l Jd U
i NOT F03 QUDI!*IIUM
..,,a.-,av
- ~ ~ - - - ~
the shape of the dose-response curve, the absolute or relative model of risk projection, the age range at risk, and the duration of effect.
yor example, BEIR 1 (1972) used a linear, dose-response curve and the absolute risk model for risk projection.
The 1980 BEIR Report presented both relative and absolute risk projections but the majority report concluded that the absolute risk model and the linear-guadratic dose response projection model was most appropriate for extrapolation for future excess cancer mortality.
However, for individual site and age specific risk coefficients, the linear dose response model was used for all cancers except leukemia and bone cancer, where the linear-quadratic model was used.
Gofman (1981) was highly critical of the BEIR (1980) risk projections because they neglected the data for persons exposed to the Aton Bomb before
.O N 7]?
age 10 on the basis that they were unreliable, while in fact, Gofman claims the increased radiosensitivity of this age group would greatly increase the risk coefficients.
Gofman (1981) also claims that the BEIR (1980) Committee also made several errors of commission in that they used the incozzect percentage of increased excess in cancer mortality per rad of exposure for their calculations of risk, thereby resulting in an unnaturally low risk coeffecient for total cancer mortality, excess per rad.
Gofman also summarily dismisses any use of the absolute risk model as being inappropriate, therefore negating the UN5CEAR (1977) risk projections.
However, Gofmany's estimates have been
~
criticised as being incorrect and unrealistic as they were based on guastionable biological assumptions and. net supported by
]. " '- { I T -
bi.
4 * * ".'s l. i y r,
. '3 f
/d/
- !j
" JV
$ Y $; r.'W.b Y
i WO 703 (UCTA'dC:!
n er., -n u nit te. - - r
appropriate statistical analysis (Orient, 1983).
The entire field of risk projection is continually s
un.dergoing revision as more data become avaliable from populations with long-term follow-up such as the Japanese survivors of the atomic boah.
The importance of long-tera follow-up is especially great here as the natural risk of cancer increases as the population ages.
Cons'equently, except for leukemia, the plateau period for solid tumors has not likely yet been achieved.
This is why lifetime follow-up of exposed cohorts is important.
Lifetime risks are the most relevant sensures of radiation effects.
The current risk projections are likely to be obsolete for the following reas.ons:
(1) the revisions in the dosimetry for the atomic bomb studies will probably increase.the risk coefficients by a yet unknown factor.
They will'also allow pooling of the Hiroshima and Nagasaki data to provide more precise risk estimates, and greatly diminish the importance of neutron radiation as accounting for the between-city differences; (2) the relative risk model is now coming into favor based on new data from the tumor incidence registry in Nagasaki.
The use of the relative risk model will likely increase the radiogenic cancer r.isk estimates as discussed ~ earlier, even though it is probable that the relative risk model is not appr'opriate for alk solid cancers; (3) the linear dose-response rel'acionship is coming back into favor as an appropriate, simple model for risk projection and. radiation neandards setting; and (4) the continued development of cancer incidence d'ats instead of only cancer I
i R. nL,D N.
z.
My a v.;- s j. 3 :,-
,= ra t=u zu
mortality data vill result in changes in risk projection 4
estimates because the nortality data considerably underestimate the risks for most cancer sites, especially breast an'd thyroid.
cancer.
The next 10-15 years should see considerable changes in the risk projection models es all of these factors are studied and are incorporated.
FUTURE RESEARCH:
DIRECTIONS FOR FURTHER STUDY Continued study of populations exposed to ionizing radiation should be a high priority, but only if they are targeted to those populations which have the highest potential for new knowledge or confirmation of existing knowledge.
Although ionizing radiation MS
~
'NF has been the most thoroughly studied human careinogen, there are still gaps in our knowledge which are crucial to more complete understanding of radiation carcinogenesis, and, in turn, mechanisms of carcinogenesis in general.
Areas that need further exploration includes (1) the magnitude of effects at low doses; (2) continued study of the effects of dose fractionation or protraction as regards tumor yield; (3) more precise estimates of differing organ sensitivity, especially those organs that are currently considered to have " low" or "no" sensitivity to exposure to ionizing radiation; (4) studies of interactions between radiation and various known cancer risk factors to determine if and how radiation exposure modifies the effects of these factors; (5) the duration of the effect for excess cancer rates, especially for solid tumors in populations that were
. g 7" F
C..,. ;
l,../- il J. ; : Q
/Cl i I.
1g
--,n
.._--.n-,
--~n
.,,,-_,,r-
l (I
relatively young at exposure; (6) temporal patterns of risk and how they are modified by factors such as age at exposure; and (7) the "best" models for risk projection.
Candidate populations for study should have a balance of i
exposures, including some moderate doses, so that risk estimates are applicable to a wide range of dose conditions.
- Although,
~
from a radiation protection perspective, there is a great need for data from populations exposed to very low doses (on the order of 1-10 rads), these studies are are often subject to small sample size limitations which can (and often do) result in chance associations that arise from sample variation.
These limitations have been discussed in detail by Land (1980a).,one possible exception to this are studies of susceptible populations, such as
. 4::, )
ra#
i persons exposed to low doses in utero or persons with an underlying susceptibility to radiation exposure such as persons with ataxia teleangeactasia, where excess cancer risks have been shown at low doses.
However, the usefulness of these groups is l
limited by the extent to which risk projections derived from them i
can be generalized to the entire population, and raises the j
ethical question of whether radiation protection standards should j
l be determined on the basis of the most susceptible subgroup.
~
Another important factor for study is the 4xistance of precise radiation dosimetry.
We already know that radiation causes cancer so that further ' descriptive studies which lack good dosimetry are superfluous.
More data which will allow relatively f
e precise estimates of cancer risk as a function of dose are e
('
l needed.
Access to complete cancer incidence data instead of r'
~ ~.7'3 r~-
t~ *~
i.
}
'i
(!
$b lDN
. l! '. $ll )
I
' ci q,
l..: :
death certificate data is important as the certification of death k
is subject to misclassification that is not randon, and nortality data usually underestimates the risks derived from incidence data for tumors that are not rapidly fatal such as c a n c e r,' o f the thyroid gland.
Long-ters and complete follow-up of exposed populations is also important in order to minimize potential biases as a consequence of persons lost to follow-up.
Long-term follow-up is also important in investigating the temporal distributions of cancer developme6e in the exposed population.
Dreyer et al. (1981) identified 100 exposed cohorts and ranked them on the feasibility and desireability as candidates for further study.
Six of the cohorts emerged as the most promising radiologic technologists, nuclear power plant N'
workers, all Department of Energy (DOE) workers, DOE workers in the >5 rem group, nuclear shipyard workers, persons exposed in utero to pelvimetry, and persons routinely exposed to diagnostic X rays.
In addition, three other exposed groups were discussed, not for their feasibility of study, but because of the great amount of public attention they had received.
These included the persons, both military and civilian, exposed to fallout from nuclear weapons testing at the Nevada Test Site, persons residing around the Rocky Tlats plutonium processing plant in Colorado. -
and persons residing in 4 counties in Florida where soil levels of phosphate and radon daughters are high..
j We would add to this group studies of populations exposed to frictionated, low dose partial body exposure, such as persons 7f b.b h, 0
l i
/t 7 dc/ l b. L G L!
i
o with scoliosis eho receivo cultiple diagnostic full-spine X roy i
examinations at a time of life when the breast and thyroid gland, (
l both of which are in the ' primary X ray beau.. are highly sensitive i
to the effects of radiation.
l i
Currently, several of these populations are being studied by i
various groups.
Within the next 10 years, there should be much more data with which to evalute, in a precise fashion, the effects of radiation on human populations.
~
4
f 0.
~
- i. l h'D i '.:l
.l}*,
.f.*
C.:I(,sL'.U C;3 (
l
\\
FOT TOR CUCIAi!ON PRIOR 20 FU3LICATICU i 6'i 8
0
V.
BIBLIOGRAPHY Acquavella JT, Iietjen GL, Wilkinson G5, et al.: Malignant
/ melanoma incidence at the Los Alamos National Laboratory.
Lancet 1 1982 483-884.
Adams EE, Brues AM: Breast cancer in female radium dial workers first employed before 1950. J occup Med 1980 22:583-587.
Alderson MR, Jackson SM Long term follow-up of patients with menorrhagia treated by irrad.iation. Er J Radiol 1971 44:295-298.
Alluright spa, Colgan PA, McAulay IR, et al.: Matural background g
radiation and cancer mortality in the Republic of Ireland.
Int J Epidemiol 1983 12:414-418.
Anderson TW: Radiation exposures of Hanford workers: A critique
[
of the Mancuso, Stewart and Xneale report. Health phys 1978 35:743-750.
i Archer VE: Health concerns in Uranium mining and milling. J
~
Occup Med 1981 23:502-506.
Austin DT, Snyder MA, Reynolds PJ et als Malignant melanoma among employees of the Laurence Livermore National Laboratory.
Lancet 1981 2 712-716.
10 1 e
- __y___,
,,__y
f I
s Axelsen 0 Edling C.
Kling H Lung cancer and' residency-A case-referent study on th'a possible impact of exposure to radon and its daughters in duellings. Scand J Work Environ Health 1979 5:10-15.
Baral E. Larsson L. Nattsson 3: Breast cancer following irradiation of the breast. cancer 1977 40:2905-2910.
Baverstock XT, Vennart 48 A note on radium body content and breast
/
cancer in U.X.
radium luminisers. Health phys 1983 44:575-577.
Back HL, Krey PW Radiation exposure in Utah from Nevada nuclear tests. Science 1983 120:18-24.
BEIR Advisory Committee on the Biological Effects of Ionizing
/
Radiation (BEIR Report), National Academy of Sciences-Mational
/
Research Council, US Gov't printing Office, Washington, DC, 1980.
Eender MA: significance of chromosome abnormalities. IM: Radiation
~
carcinogenesist Epidemiology and Biological significance., JD Bo' ice Jr., JT Traumani Jr.,
Eds., Meu York, Raven press, pp 281-290.
1984.
Bertall R X ray exposure and premature aging. J Surg Oncol 1977 9:379-391.
f
\\
\\.
l10
l e
Bithall JF, Stewart *AM: pre-natal irradiation and childhood malignancy: A review of British data from the Oxford survey.
Er J Cancer 1975 31:271-247.
I Blair A, yraumani JT Jr., Mason TJ8 Geographic patterns of leukemia in the United States. J Chron Dis 1980 338251-260.
Boice JD Jr., Monson R18 3reast cancer in women after repeated
/ fluoroscopic examinations of the chest. J Matl Cancer Inst 1977 59:423-832.
Boice JD Jr., Stone BJ8 Interaction between radiation and other breast cancer risk factors. IN Late Biological Effects
. 's
,/ of Ionizing Radiation, Vol 1.' Vienna, International Atomic ys Energy Agency, 1974, pp 231-149.
Boice JD Jr., Rosenstein M. Trout ED: Estimation of breast doses
/ and breast cancer risk associated with fluoroscopic chest examinations of vonen with tuberculosis. Radiat Res 1978 738 373-390.'
i Boise JD Jr.,
Land CEI Adult leukemia follouing diagnostic X rays.
A review of the report by 3ross, Ball and yallen on the Tri-state leukemia survey. Am J pub Health 1979 69:137-145.
i Boice JD Jr., Hutchison G3:
Leukemia in women following radiotherapy for cervical cancers
(*. '
Ten-year follow-up of an international study. J Mati cancer Inst 1980 658115-129.
///
soice JD Jr., Day ME, Anderson A, et al.: Cancer risk following radiotherapy of cervical cancers A preliminary report. IM:
Radiation carcinogenesis Epidemiological and Biological Significance. JD toice Jr., Jr Traumeni Jr., Eds., Meu York, Raven Press, 1943, pp.161-179.
Routuell RK Overview of tumor promotion. IX: prostaglandins and cancer. TJ Poules',' RS Bockman, XV Monn, P Rouve11. Eds., Meu York Alan Liss, Inc., 1942, pp. 143-144.
Borak C.
Hall EJ Effect of split doses of x rays on neoplastic transformation of single cells. Nature. 1974 252:499-501.
Borek C, Hall EJ Oncogenic transformation produced by agents (gyg{i
...yu and modalities used in cancer therapy and its modulations.
IX: cell P'roliferation Cancer, and cancer Therapy. 1 Basarga.
Ed. Ann New York Acad Shi 1942 397 193-110.
Borek C.
Hall EJ Induction and modulation of radiogenic transformation in naamalian cells. IX: Radiation Carcinogenesis
/ Ipidemiology and Biologica1 significance., JD soice Jr., Jr Traumani Jr., Eds., Meu York, Raven Press, 191-302, 1944.
Brincker H.
Hansen H3, Anderson AP Introduction of leukemia by treatment of thyroid carcinoma. 3r J Cancer 1973 28:232-137.
f
\\,
' ll i-
O O
(
Brinkley D.
Maybittle JLs The late effects of artificial menopause by X iradiation. Er J' Radiol 1969.42:519-521.
Bross IDJ, Nataradan Ms Leukemia from lou-level radiation.
j/ Identification of susceptible children. New Engl J Med 1972 247s107-110.
Bross IDJ, Natarajan N: Risk of leukemia in susceptible children exposed to preconception, in utero, and postnatal radiation.
Prev Med 1974 3:361-369.
Bross IDJ, Ball M.
Talen St A dosage response curve for the one rad range Adult risks from diagnostic radiation. Am J
/
(j pub Health 1979 69:130-136.
Bross IDJ, Driscoll DL Direct estimates of lou-level radiation risks of lung cancer at two nuclear installations: Why are the
! new risk estimates 20 to 200 times the old official estimates.
Yale J Biol Med 1941 54:317-324.
Bross IDJ, Driscoll DLs Data on lung cancer in radiation workers. J Royal Stat soc 1942 75(A):424-829.
Burba'nk TI patterns in cancer nortality in the United Statess' 1950-1967. J Matl Cancer Inst Monograph 33 1971.
(
l I.]
Burch PRJ8 prenatal radiation exposure and childhood cancer.
(
i Lancet 1970 2:1189-1190.
Caldwell GG, Kelley DB, Heath CW Leukemia among participants in military maneuvers at a nuclear bomb test. J An Med Assoc 1980 244:1575-1574.
Caldwell GG, Kelley D, Eack M, at al.
Mortality and cancer fretuency among military nuclear test (smokey) participants, t
1957 through 1979. J Am Med Assoc 1943 2508620-624.
Canard RA: summary of thyroid findings in Marshallese 22 years after exposure to radioactive fallout. IN: Radiation-Associated Thyroid Carcinoma, LJ Degroot Ed., Meu York, Grune and Stratton, fy7,q (m.3 1977, pp. 241-257.
Conard rat Late radiation effects in Marshall Islanders exposed to fallout to years ago. IN: Radiation carcinogenesis Epidemiology and Biological Significance., JD Soice Jr.,
Jr rzaumeni Jr.,
Eds.,
New York. Raven press, pp. 57-72, 1984.
Court-Brown WM, Doll R.
Hill A38 Incidence of leukemia after exposure to diagnostic radiation in utero. Et Med J 1960 28 1539-1545.
Curnen MGM, Varna AA0e Christine BW, et al.
Childhood leukemia and maternal infectious diseases during pregnancy. J Hati cancer
/
6 Inst 1974 538943-947.
IIU.
e i
Cuzick Ja Radiation-induced myelomatosis. New Eng J Med 1981 s
/ 304:204-210.
Dahlgren En Lung cancer, cardiovascular disease and smoking in a group of miners. Lakartidningen 1979 76:4811-4814.
Darby 3C Modelling age-and time-dependent changes in the rates of radiation-induced cancers. IM Atomic Bomb survivor Data Utilization and Analysis., RL prentice. DJ Thompson, Eds.,
philadelphia, SIAM, 1984.
DeGroot LJ, paloyan El Thyroid carcinoma and radiations A
/ Chicago endemic. J Am Med Assoc 1973 22,5 487-491.
{
Piamond EL, schmerler H. Lilienfeld AM: The relationsh1P of intrauterine radiation to subsequent mortality.and development of leukemia in children. Am J Epidemiol 1973 97 283-313.
Dobyns BM, sheline GE, Workman JB, et al.: Malignant and benign neoplesas of the thyroid in patients treated for i
hyperthyroidism: A report of the cooperative thyrotoxicosis therapy follou-up study. J Clin Indocrinol Metab 1974 38:
976-998.
4 Doll R, saith 70: The long-tern effects of X irradiation in
! patients treated for metropathis hemorrhagica.
Er J Radiol 1968 41:362-368.
ll 3"
o Doniach It pathology of irradiation thyroid damage. IX:
Radiation I.'
)
Associated Thyroid Carcinoma., LJ DeGroot, Ed.,Meu York, Grune and Stratton, pp 199-211, 1977.
Dreyer MA, Loughlin JE, Friedlander ER, et al.8 Choosing populations to study the health effects of low-dose ionizing radiation. Am J pub Nealth 1981 71:1247-1252.
Elkind MM, Man A.
Mill CK Error-free and error-prone repair in radiation-induced neoplastic call transformation. IM: Radiation
' Carcinogenesist Epidemiology and Biological Significance.,
JD Boice Jr.,
JT Fraumeni Jr., Eds., Meu York, Raven press, pp 303-318, 1944.
)
Engel LW, Strauchen JA, Chiarre L Jr., et al.: Accuracy of death certification with specific attention to salignant neoplasms and vascular diseases. Am J Epidemiol 1940 111899-112.
Erstron JE The nonassociation of fallout radiation with childhood leukemia in Utah. IX: Cancer Incidence in Defined populations 8 Banbury Report 4 J Cairns, JL Lyon, M Skolnick, Eds., cold Spring Harbor Laboratory, 1940, pp.163-177.
Taber Ms Twenty-eight years of continuous follow-up of patients injected with Thorotrast for cerebral angiography. Environ Res 1979 14:37-43.
s OL g
d
o Tavus MJ, Schneider AB. Stachura ME, at al.: Thyroid cancer occurring as a late conseguance of head-and-neck irradiation:
Evaluation of 1056 patients. New Engl J Med 1976 19481019-1025.
Tedrick J. Alberman ED8 Reported influenza in pregnancy and subsequent cancer in the child. Er Med J 1972 2:485-448.
Tinch SC Leukemia and lymphoma in atomic bomb survivors. IMs Radiation carcinofenesiss Epidemiology and Biological Significance.,
JD Soice Jr., JT Fraumeni Jr., Eds., New York, Raven Presse PF g
37-44, 1944.
Try RJM8 Relevance of animal studies to the human experience.
IX: Radiation carcinogenesis Epidemiology and aiological Significance. JD Boice Jr.,a JT Traumeni Jr., Eds., Heu York, Raven Press, pp. 337-346, 1944.
Fujita 5, Awa AA, Pierce DA, et al.: Reevaluation of the biological effects of atomic bomb radiation by the changes of estimation dose.
/ 2H Biological Effects of Low-Level andiation (Proceeding Series),
Vienna, International Atomic Energy Agency, pp 55-69, 1943.
l
~
Gibson R.
Bross ZDJ, Graham 5, et al.: Leukenia in children
/ exposed to multiple risk factors. Heu Eng1 J Med 1964 179:
406-909.
s
/I7
Gibson Re. Graham S.
Lilienfeld AM, et al.: Irradiation in the ll epidemiology of leukemia among adults. J Mat,1, Cancer Inst 1972 44:301-311.
Gilbert ES, Marks St An analysis of the mortality of workers in a
~
nuclear facility. Radiat tes 1979 79:121-144.
l
~
Gilbert E3, Marks 58 An updated analysis of mortality of workers
!,inanuclearfacility. Radiat Res 1940 438740-741.
Ginevan ME: Monlymphatic leukemias and adult exposure to diagnostic X rays: The evidence reconsidered. Health phys 1940 34s124-134.
Ginevan ME, puskin J58 Delayed effects of A-bomb radiation (letter I
I to the editor). J Epidemiol Community Health 1943 34845-44.
Gofman JW Radiation and Human Health. San Francisco, 5
Sierra Club Books, 904 pp. 1941.
Goodhead DT Deductions from sellular studies of inactivation, mutagenes'is, and transformation. ZMs Radiation carciniegenesist
/ Epidemiology and Biological signifiennea. JD seice Jr., Jr rzausani Jr., Eds. Meu York, Raven Press, pp 349-346, 1944.
Gotlieb Ls, Husen 1Al Lung cancer among Navajo uranium miners.
Chest 1942 418449-452.
(,
19 z
e.
e s e
Gould MN Radiation initiation of carcinogenesis in viv.o a rare or common cellular event. IM: Radiation Carcinogenesis Epidemiology y/ and Biological Significance. JD Bolce Jr.,
JT Fraumeni Jr.,
Eds.,
New York, Raven press, pp 347-354, 1944.
Graham 5.
Levin ML, Lilienfeld AM, et al.: preconception, intrauterine, and postnatal irradiation as related to leukemia.
Matl Cancer Inst Monogr.1966 19:347-371.
Hadjimichael CC, Ostfield AM, D'Atti DA, et al.
Mortality and cancer incidencu experience of employees in a nuclear fuels fabrication plant. J Occup Med 1943 25,844-61.
(fj Marley'MH, pasternack 35: A model for predicting lung cancer risks induced by environmental levels of radon daughters. Health phys 1941 40:307-316.
Harvey E.
Boice J.
Honeymon M. et al.: prenatal X ray and childhood cancer in twins. Am J tyidemiol 1943 114:424-425.
Hazen AW, pifer JW, Toyooka ET, et al.1 Meoplasms following irradiation of the heads part 1. Cancer Res 1966 26:305-311.
Hempelmann 1H, Hall WJe phillips M, et al.: Neoplasms in persons treated with X rays in infancys Tourth survey in 30 years.
J Hatl Cancer Inst 1975 55:519-530.
(
)I 9
.u e
van Kaick G. Lorenz D. Muth H. et al.: Malignancies in German
'l
/ thorotrast patients and estimated tissue dose. Health phys 1974 35:127-136.
van Kaick G. Muth N, Xaul A, et al.: Results of the German Thorotrast study. IN: Radiation Carsinogenesis Epidemiology and Biological significance. JD Boise Jr., JF Frauseni JJr., Eds.,
New York, Raven press, pp 253-162, 1944.
Nato He schull WJe studies of th's mortality of the A-bomb survivors. 7. Mortality, 1950s part 1. Cancer mortality.
Radiat Res 1942 90:395-432.
Nato H, Brown CC, Noel DG, et al.: studies of the nortality of A-bomb survivors. Report 7. Mortality, 1950-1974: part II.
?:ortality from causes other than cancer and mortality in early entrants. Radiat Ras 1942 91:243-264.
,' Xaul A, Hofdz W Tissue dose in Thorotrast patients. Health phys 1974 35:113-121.
i Xennedy AR, Cairns J.
Little JB Timing of the steps in transformation of C3H 1071/2 cells by x irradiation. Nature 1944 307:45-46.
Lb
Kerr GD Tindings of recent oak Ridge Naional Laboratory review of dosimetry for the Japanese Atom-bomb s'urvivors.
p/
IX: Reevaluations of Dosimetric Tactors, Hiroshima and Magasaki., Vp Bond, JW Ihiessen, Eds. DOE Symposium Series, No. 55, CONT-8109287 National Technical Information service, Springfield, VA, pp 52-97, 1982.
Kerr GD, pace JV III, Scott WH Tissue kerna vs distance relationships for initial nuclear radiation from the atomic bombs, Hiroshima and Nagasaki. IX: Reassessment of Atomic Bomb Radiation Dosimetry in Hiroshima and Nagasaki., DJ Thompson, Ed.,
Hiroshima, Radiation Effects Research Foundation, pp 57-103, 1983.
f1 Kessler II. Lilienfeld AM: Perspectives in the epidemiology of leukemia. Advances in Cancer Research 1969 12:225-302.
Kleinarman RA, Curtis RE, Boice JD Jr.,
et al.: Second cancers following radiothera;y for cervical cancer. J Hati cancer Inst 1982 69:1027-1033.
Xneale GW, Steu. art AM: Mantel-Haenssel analysis of Oxford 5
/ data.
I.
Independent effects of several birth factors including detal irradiation. J Natl Cancer Inst 1976a 56:1009-1014 Kneale GW, Stewart AM: Mantel-Haenszel analysis of Oxford
(
data. II. Independent effects of fatal irradiation subfactors.
J Hati cancer Inst 1976b 57:1009-1014
/ 2. /.
(
\\
'A Kneale GW, Stewart AM, Mancuso TT Re-analysis of data relating to the Hanford study of the cancer risks of radiation workers.
v!
Late Biokogical Effects of Ionizing Radiation, Vol.
I.,
IX:
Vienna, International Atomic Energy Agency, pp 386-412, 1978a.
Kneale GW, Stewart AM Pre-cancers and liability to other diseases, Br J Cancer 1978b 37:448-457.
Kneale GW, Stewart AM Pre-conception X rays and childhood cancer.
Er J Cancer 1980 41:221-226.
1 Kneale GW, Mancuso TT, Stewart AM: Nanford radiation study. III.
a cohort study of the cancer risks from radiation to workers at b)
.c3 Hanford (1949-1977 deaths) by the method of regression models in life-tables. Er J Indust Med 1981 34 156-166.
Knox ZG. Sozahan T, Stewart AMt Cancer following nuclear weapons l
tests. Lancet 1983 18815.
Kochupillai M. Vazaa IC, Grewal MS, et al.: Down's syndrome
/
and related abnormalities in an area of high background
~
radiation in coastal Kerala. Nature 1976 262 60-61.
Land CE, Horaan JE Latent periods of radiogenic cancers occurring among Japanese A-bomb survivors. IX8 Late Biological Effects of Iorizing Radiation, Vol.
1.,
Vienna, International Atomic Energy A2*ncy. PP 19-47e 1978.
8 7
0 e
i
.\\
Land CE: The bazards of fallout or epidemiologic research. Hau g/ Eng J Med 1979 300:431-432.
Land CE: Estimating cancez risks from lou doses of ionizing v radiation. Fcience 1980a 209:1197-1202.
i
~
Land CE, Boice JD Jr., sho:e RE, et al.: Breast cancer risk from lou-dose exposures to ionizing radiation: Results of parallel analysis of three exposad populations of women. J Natl Cancer Inst 1980b 65:353-376.
Land CE. Tokunaga M Induction period. IN Radiation Carcinogenesis d, !
Epidemiology and Biological Significance., JD Boice Jr., JT Traumeni l
Jr.,
Eds., New York, Raven press, pp 421-436, 1984a.
Land CE, McKay TW, Machado SG Childhood leukemia and fallout from i
the Nevada nuclear tests. science 1984b 223:139-144.
s Larson PR, Conard RA, Knudsen XD, et al.: Thyroid hypofunction after
., ' exposure to fallout from a hydrogen bomb explosion. J Am Med Assoc 1982 247:1571-1575.
Larsson L-G, Damber L Interaction between underground, mining and smoking in the causation of lung cancer A study of nonuranium miners in Northern Sueden. Cancer Detection and prev 1982 5 385-389.
113
Lee W.
Chiacchierini 17, Schleien 3, at al.: Thyroid tumors
-[
l
(
following 181I or localized X irradiation.to the thyroid and the pituitary glands in rats. Radiat Res 1982 928.307-319.
Louis E3: Leukemia, multiple myeloma, and aplastic anaemia in American radiologists. Science 1963 142:1492-1494.
Louis EE: Leukemia, radiation and hyperthyroidism. Science 1971 174:454.
Linos A, Gray JE, Orvis AL,' et al.: Lou-dose radiation and leukemia.
Meu Engl J Med 1980 302:1101-1105.
Loeve WE Revised estimates of neutron and gamma-ray doses at
.' Hiroshima and Nagasaki. IM Reevaluations of Dosimetric Tactors.
Hiroshima and Magasaki. Vp Bond, JW Thiessen, Eds., DOE Symposium series, No. 55, c0XF-810928, National Technical Information Services, Springfield. VA, pp 25-51, 1982.
Lundin TE Jr., Wagoner JK, Archer VE, et al.: Radon daughter exposure and respiratory cancers Quantitative and temporal aspects. Report from the Epidemiological Study of United Staes Uranium Miners.
NIOSH-MIENS Joint Monograph No.
1, Springfield, VA, Natl Tech Information service, 176 pp.,
1971.
, Luxin W: Health survey in high background radiation areas in
(
China. Science 1980 209:477-880.
Lyon JL. Klauber MR, Gardner JW, et al.: Childhood leukemias y associated with fallout froa nuclear testing. New Engl J Med" 1979 300:397-402.
Lyon JL, Gardner JW, West DW, at al.: yurther information on the association of childhood leukemias with atomic fallout.
Banbury Report No. 4: Cancer Incidence in Defined Populations, J Cairns, JL Lyon, M Skolnick, Eds., Cold Spring Harbor Laboratory, pp 145-162, 1980.
MacKenzie Is Breast cancer.following multiple fluoroscopies.
Br J Cancer 1965 1981-8.
' 3'g,
.fy MacMahon 58 prenatal X ray exposure and childhood cancer. J Natl Cancer Inst 1962 28:1173-1191.
MacMahon B, pugh TT Epidemiology: principles and Methods., Boston, Little, Erown and Company., pp 59-60, 1970.
Mancuso TT, Stewart AM. ICneale GW: Radiation exposures of
/
Hanford workars dying of, cancer and causes; Health phys
~
1577 33:359-385.
Manning MD, D' Angio GJL. Irradiation of the nasopharynx in childrens-A survey of patients 5-to years after treatment for lymphoid hyperplasia. Clin Radiol 1967 18:173-175.
~
J e----
--m-e-
a Marks 5, Gilbert Es, Breitenstein 3D: Cancer, mortality in
)
Hanford workers. IX: late Biological Effects of Ionizing
/
Radiation, Vol.
1, Vienna, International Atomic Energy Agency, pp369-346, 1978.
Matanoski GM, seltser 1, Sartuell PE, at al.: The current mortality j/ rates of radiologists and other physician specialists: Deaths from l
all causes and from cancer. Am J Epidemiol 1975a 101:188-198.
s Matanoski GM, seltzer 1, Sartwell PE, rt al.'s The current mortality
(
/ rates of radiologists and other physician specialists: specific l
causes of death. Am J Epidem,iol 1975h 101:199-210.
ffQ)
Matanoski GM, Tonascia J.
Billings C, et al.: Health effects of low-level radiation in shipyard workers. IM: Proceedings of the DOE Radiation Epidemiology contractors Workshop. DE82014065, National Technical Information Service, Springfield, VA, 1982.
l l
Matanoski GM, sattuell P.
Elliott E, et al.: Cancer risks in radiologists and radiation workers. IX: Radiation Carcinogenesis:
Epidemiology and Biological Signifance. JD Boice Jr.,
JT Traumeni Jr.,
Eds., New York, Raven Press, pp 43-96, 1984.
Maxon HR, Thomas $1, Saenger EL, et al.: Ionizing radiation j/ and the induction of clinically significant disease in the human thyroid gland. Am J Med 1977 638967-978.
\\26
(
~
Mays CW, Spiess H Bone sarcomas in patients given
- Ra.
IN:
Radiation Carcinogenesis Epidemiology and Biological Significance.
V JD Boice Jr., JT Traumani Jr., Eds., New York, Raven press, pp 241-252, 1944.
Mettler FA Jr.,
Hempelmann LN, Dutton AM, et al.: Breast neoplasms in uomen treated with X' rays for acute postpartum mastitis. A pilot study. J Matl Cancer Inst 1969 43:403-411.
Modan B.
Lilienfeld AM: polyc'ythenia vera and leukemia-the role of radiation treatment. A study of 1222 patients.
Medicine 1965 44:305-344.
Modan B.
Baidatz D, Mart H.
et al.: Radiation-induced head and neck tumors. Lancet 1974 1:277-279.
Modan B, Ron E.
Werner A.: Thyroid cancer following scalp irradiation. Radiology 1977 1238741-744.
Mole RH Antenatal irradiation and childhood cancer Causation or coincidence. Er J Cancer 1974 30:199-204.
5 Mole RH Carcinogenesis by thorotrast and other sources of irradiatien, especially other alpha-emitters. Environ Res 1979 14:192-215.
/27 7
--~
a Monson RR, MacMahon 38 prenatal X ray exposure and cancer in k'
/ children. IM Radiation Carcinogenesis Epidemiology and Biologi' cal Significance. JD Boice Jr.,
JT Fraumani Jr., Eds., Meu York, Raven press, pp 97-106, 1984.
~
da Motta LC, da Silva Horta J. Tavares MN prospective epidemiological study of Thorotrast-exposed patients in portugal.
Environ Res 1979 18:152-172.
Muggenburg BA, Meuhinney JA, Griffith WC, et al.
Dose-response relationships for bone cancers from plutonium in dogs and people. Health phys 1983 44:529-535.
Myrden JA, Milta JE Breast cancer following multiple fluoroscop.e-A.1 during artificial pneumothorax treatment of pulmonary tuberculos s.
Can Med Assoc J 1969 100:1032-1034.
N:darian T, Colton T Mortality from leukemia and cancer in
~
i shipyard workers. Lancet 1978 1:1018-1020.
Majaria'n T Comment on alternative interpretations of statistics on health effects of lou-level radiation. American Stat 1983 37:
~
455-457.
National Board of Health and Welfare. The cancer Registry: Cancer Incidence in Sweden, 1959-1965. Stockholm, 197 pp.,
1971.
119 9 --
Newcombe HB, McGreg'or JF Childhood cancer following obstetric radiography. Lancet 1971 2:1151-1152.
J Oppenheim BE, Grien ML. Meier P.: The effects of lou-dose
. Prenatal irradiation in humans: Analysis of Chicago lying-in
,/
data and comparison with other studies. Radiat Res 1974 57:
508-544.
crient JM A critique of the statistical methodology in Radiation
'/
~
and Human Health by John W. Gofman. Health Phys 1983 45 823-827.
i Paterson MC, Bech-Hansen NT, Smith PJ, et al.: Radiogenic neoplasia, i
, cellular radiosensitivity, and faulty DNA repair. IX: Radiation f
Carcinogenesis Epidemiology and Biological Significance., JD Boice i
j Jr.,
JF Fraumeni Jr.,
Eds., Meu York, Raven Presse,pp 319-338, 1984.
Petersen GR, Breitenstein BD Jr.: The Hanford DOE health and mortality study. IN: Proceedings of the DOE Radiation Epidemiology Contractors Workshop. DE82014065. Mtaional Technical Informatica service, springfield, VA, 1982.
d i
l 1
l P!ncet J. Masse La Natural radiation and cancer mortality in j
l
~
l' several areas of North Brittany. Int J Epidemiol 1975 4:311-316.
Pochin EE Long-term hazards of radiciodine treatment of thyroid 2
carcinoma. IN: Thyroid Cancer., C Hedinger, Ed.,
Berlin, Springer-
~
Verlag, pp 293-304, 1969.
/27
m Polednak AP, Stahney Ay, Rowland RE: Mortality among women first employed before 1930 in the U S radium dial painting industry. A group ascertained from employment lists. Am J Epidemiol 1978a 107:179-195.
~
Polednak AP Bone cancer among female radium dial workers.
latency periods and incidence rates by time after exposure.
J Matl Cancer Inst 1978b 60:77-82.
Polednak AP, Trome ED: Mortality among men employed between 1943 and 1947 at a uranium processing plant. J Occup Med 1981 23 1
169-178.
..A. )
Yl Radford EP, Doll 1, Smith PG: Mortality among patients with ankylosing spondylitis not given X ray therapy. Meu Engl J Med 1977 297:572-576.
Radford EP Radon daughters in the induction of lung cancer in underground miners. IN: Quantification of Occupational Cancer:
Banbury Report No.
9.,
1 Peto, M Schneiderman, Eds. Cold Spring Harbor Laboratory, pp 151-163, 1981.
O e
(.
13e u
k
~
?. '
Radford EP, Talbott E, Schmaltz R.: Effects of elevated gamma
(
zay exposures on the prevalence of thyroid abnormatlities in a community. IX: Somatic and Genetic Effects., JJ Broerse, GW Barendsen, HB Kal, AJ van der Kogel, Eds. Proceedings of the Seventh Internation Congress of Radiation Research, Amsterdam, Marttinus Nijhoff Publishers, C8-12 (abstract), 1943.
Radford EP Epidemiology of radiation-induced cancer. Environ I
Health Perspect 1983 52:45-50.
Radford EP Radiogenic cancer in underground miners. IM Radiation e Carcinogenesis Epidemiology and Biological Significance. JD Boice
/
Jr.,
JT Fraumeni Jr.,
E d's., New York, Raven Press, pp 225-230, g.: p, 1984.
i c
,.9
^
\\
e Radford EP, St. Clair Renard KG LungcancerinSwahishiron minersexposedtolowdesesofradondaughters.Mf:nglJMed 1984 310:1485-1494.
Rallison ML, Dobyns BM, Keating TR, et al.: Thyroid disease in children A survey of subjects potentially exposed to fallout
,/ radiation.
Am J Med 1974 56:457-a63.
t I
Rallison ML, Dobyns BM, Keating TR Jr., et dl.: Thyroid l
nodularity in children. JJ Am Med 3ssoc 1975 133:1069-1072.
(I IJl
- - + - -
em
-_2___,.,..,_,
4 Refetoff 5.
Marison J. Karanfilski BT, et al.: Continuing i
occurrence of thyroid carcinoma after irradiation to the neck in infancy and childhood. New Engl J Med 1975 292*171-175.
Reyes M. Wilkinson G. Tietjen G, et al.
Case-control study of brain tumors among uhite males employed at the Rocky Flats l
plant. LA-9404-MS, Los Alamos National 4aboratory Report, UC-41, 13 pp. 1943.
Rinsky RA, Zumwalde RD, Waxuailer RJ et al.
Cancer mortality at a naval nuclear shipyard. lancet 1981 1:231-235.
l l
Robinette CD, Jablen St Studies of participants at tests of af I
- lf;h '
nuclear weapons: I. The plumbob series. IX: Somatic and s
Genetic Effects., JJ Broerse, GW Barendsen, HB Kal, AJ van der Kogel, Eds., proceedings of the Seventh International Congress of Radiation Research., Amsterdam, Martinus Nijhoff publishers, C4-13 (abstract), 1943.
Ron E, Modan 38 Benign and malignant thyroid neoplasms after
,/ childhood irradiation for tinea capitis. J Matl Cancer Inst.
1940 65:7-11.
Ron E, Modan 3: Thyroid and other neoplasms following childhood scalp irradiation. IM Radiation carcinogenesis Epidemiology
/
and Biological Significance., JD Boice Jr.,
Jr Fraumeni Jr.,
Eds.,
~Xeu York,. Raven press, pp 139-151, 1984.
is>
l l
I
-,m.
- ____m, r
-,y.
Rossi HN, Hall EJ: The multicellular nature of radiation carcinogenesis. IM Radiation Carcinogenesis 1 Epidemiology and v'
Biological Significance., JD Boice Jr., JT Fraumeni Jr., Eds.,
Iteu York, Raven press, pp 359-364, 1984.
Roula.1d RE, Stahney AT, Lucas NT Jr.: Dose-response relationships for female radium dial workers. Radiat Res 1978 76:368-343.
Rouland RE, Lucas NT Jr.: Radium-dial workers. IX: Radiation Carcinogenesis Epidemiology'and Biological Significance. JD soice Jr., JT Fraumani Jr.,
E d s'., Meu York, Raven press, pp 231-240, 1984.
1 Jh'u'}
Royce PC, MacKay BR, Disabella PM: Value of postizzadiation screening for thyroid nodules. A controlled study of recalled patients. J Am Med Assoc 1979 242:2675-2674.
Saenger EL, Thoma GE, Tompkins EA: Incidence of leukemia following treatment fo hyperthyroidism. preliminary report of the Co' operative Thyrotoxicosis Therapy rollow-up Study. J Am Med Assoc 1968 205 855-862.
Sandler Dy, Comstock GW, Matanoski GM Neoplasms following c'hildhood irradiation of the nasopharynx. J Matl Cancer Inst 1982 68:3-8.
'l s
I?)
l
f Schneider AB. Tavus MJ, Stachura ME, et al.: Incidence, prevalence, and characteristics of radiation-induced thyroid tumors. Am J Med 1978 64:243-252.
l I
Schofield GE: Epidemiological studies of radiation workers t
preliminary communication. J Royal Stat So'c 1982a 75A 450-456.
Schofield GB: Letter to the editor in response to " Data on 5
Oung cancer risk" by Bross and Driscoll. J Royal Stat Soc 1982b 75A 429.
Schull WJa Atomic bomb survivors patterns of cancer risk. IX:
Radiation Carcinogenesis Epidemiology and Biological Significance.a Boice JD Jr., JT Traumeni Jr., Eds., Meu York, Raven press, pp
)(
1984.
Segall As Leukemia and background radiation in Northern New England.
Blood 1964 23:250-261.
S l
Shiono pH, Chung CS, Myrianthopoulos NC: preconception radiation, intrauterine diagnostic radiation, and childhood neoplasia.
J Xatl Cancer Inst 1980 63:681-686.
l shore RE, Albert RE, pasternack 358 'To11ou-up study of patients
/ treated by X ray epilation for tinea capitis Resurvey of post-treatment illness and mortality experience. Arch Environ Health 1976 31:17-24.
lh
c t
Shore RE, Hempelmann LM, Koualuk E, et al.8 Breast neoplasms in women treated with X rays for acute post-partum mastitis.
J Matl cancer Inst 1977 59:413-822.
l Shore RE, Woodward ED, Pasternack 35, et al.: Radiation and host
, factors in human thyroid tumors following thymus irradiation.
Health Phys 1980 348451-465.
shore RE, Woodward ED, Hempelmann LH Radiation-induced thyroid cancer. IM: Radiation carcinogenesis Epidemiology and Biological l'
/
Significance., JD Boice Jr.,
JT Fraumeni Jr.,
Eds., Meu York, Raven Press, pp 131-138, 1984.
A.1%
haith PG, Pike MC, Hamilton LD: Multiple factors in leukamogenesis.
Br Med J 1973 1:442-483.
Smith PG, Doll 1: Late effects of X irradiation in patients treated for metropathia haemorrhagica. Br J Radiol 1976 49:
224-232.
Smith PG, Doll 18 Mortality from cancer and all causes among
~
British radiologists. Er J Radiol 1981 54: 187-194.
l
~
(
Smith PG, Doll 1: Mortality among patients with ankylosing l
spondylitis after a single treatment course with X rays.
Ex Med J 1982 284:449-460.
l l
13 C x
Smith pG: Late effects of X ray treatment of ankylosing spondylitd-IN: Radiation Carcinogenesis Epidemiology and Biological
/
Significance. JD Boice Jr.,
Jy yraumeni Jr.,
Eds., Xeu York, Raven press, pp 107-114, 1984.
Spiers yW. Beddos AH, King SD, et al.
The absorbed dose to the bone marrou in the treatment of polycythenia by asp. Er J Radiol 1976 49:133-140.
Stebbings JH Jr.,
Voels GL Morbidity and mortality in Los Alamos County, Meu Mexico.
- 1. Methodological issues and preliminary results. Environ Res 1981 15:46-105.
Stebbings JH Human health effects of radium: An epidemiologic
,p
)
perspective of research at Argonne National Laboratory. IX:
proceedings of the DOE Radiation Epidemiology Contractors Workshop. DE82014065, springfield, VA, Matl Tech Information Service, 1982.
stewart AM, Webb J, Neuitt Da A survey of childhood malignancies.
^
3r Med J 1958 181495-1508.
Stewart AM, pennybacker W.
Barber 1 Adult leukemias and diagnostic X rays. Er Med J 1962 2:482-490.
Stewart AM: An epidemiologist takes a look at radiation risks.
' Washington, DC.,
US Govt print Off. DHEW publication Ho. (TDA/P'"'
73-4024, 37 pp. 1973.
13 r
Stewart AM, Kneale G, Mancuso TT The Hanford data-a reply to recent criticisms. Ambio 1980 9866-73.
Stewart AM Delayed effects of A-bomb radiations a revieu of recent mortality rates and risk estimates for five-year survivors.
J Epidemiol Community Health 1942 36:40-46.
t Tokunaga M, Norman JE Jr., Asano M, at al.: Malignant breast
[ tumors among atomic bomb survivors, Hiroshima and Magasaki, 1950-1974. J Mati cancer, Inst 1979 62:1347-1359.
Tokunaga M, Land CE, Yamamoto T, et al.: Breast cancer in
/ Japanese A-bomb survivors.
Lancet 1982 2 924.
[)$:
W Tokunaga M, Land CE, Yamamoto T, et al.: Breast cancer among i
atomic bomb survivors. IM8 Radiation Carcinogenesis Epidemiology
/ and Biological Significance. JD Boice, Jr., JT Traumani Jr.,
Eds.
l New York, Raven press, pp 45-56, 1984a.
l Tokunaga M, Land CE, Yamamoto T, et al.: Incidence of female breast cancer among atomic bom'b survivors, Hiroshima and Magasaki, 1950-1980. (unpublished manuscript, 1984b).
~
Tolley HD, Marks S. Suchanon J.
et al.
A further update of the
/ analysis of mortality of workarr in a nuclear facility. Radiat
- Res 1983 95:211-213.
13>
r---
,_,..-,,---------.----_--_.-,-,e-
Tompkins.Es Late effects of radiciodine therapy. IM8 Medical Radionuclides
- Radiation Dose and Effects...RJ Cloutier, CL Eduards, WS Snyder, Eds. AEC Symposium Series 20, CONT-691212, V
Mational Technical Information Service, Springfield, VA, pp 431-440, 1970.
Totter JR. McPherson MGs Do childhood' cancers result from prenatal X rays. Health phys 1941 40:511-524.
Tubiana M',
Flamant 1, Attie E, et al.: A study of hematological complications occuring in patients with polycythenia vera treated with 8tp. Blood 1964 32:536-542.
United Nations Scientific Committee on the Effects of Atomic
' It }
Radiation (UNSCEAR)* Sources and Biological Effects. 1982 Report to the General Assembly, Meu York, United Nations, 1982.
Voelz GL, Hempelmann LH, Laurence JMP, et al.: A 32-year medical f follow-up of plutonium of Manhattan project plutonium workers.
Health phys.1979 37:445-453.
vo.els GL, Wilkinson G3, Acquavella JT, et al.8 A review of epidemiologic studies at Los Alamos Mational Laboratory. IX8 proceedings of the DOE Radiation Epidemiology Contractors Workshop. DE82014065, Springfield, VA, Matl Tech Information services, 1982.
('
19 s
~
Voelz GL, Wilkinson GS, Acquavella Jr. et al.
An update of i
,,e epidemiologic studies of plutonium workers. Health phys 1983 44:49'3-503.
Wagoner JX: Leukemia and other malignancies following radiation therapy for gynecological disorders. IM Radiation Carcinogenesist
/
/
Epidemiology and Biological Significance., JD Boice Jr., JT Fraumeni Jr., Eds., Xeu York, Raven press, pp 153-160, 1984.
Wakabayashi T.
Kato H.
Theda T, et al.
Studies of the mortality of A-bomb survivors. Report 7. III: Incidence of cancer in v/
1959-1978, based on tumor registry. Magasaki. Radiat les 1983 93 112-146.
(Q. ?
Walach N, Hochman A.
Loeventhal Et cancer of the thyroid following radium application to the neck. Oncology 1975 31:153-156.
Whittemore AS, McMillan At Lung cancer mortality among U.
5.
uranium miners: A reappraisal. J Matl Cancer Inst 1983 71 449-499.
Wilson GM, Milpatrick R. Eckert H. et al.: Thyroid neoplasms following itradiation. Er Med J 1958 2 929-934.
/ Woolson WA, Marcus J, Scott WH, et al.8 Building-transmission factors. IX: Reevaluations of Dosimetric Tactors. Hiroshima and Magasaki., VP Sond, JW Thiessen. Eds., DOE Symposium Series, No. 55, CoHr-810928, National Technical Information service.
Springfield. VA, pp 179-200, 1982.
/31
lib.
tunis Je The chromosomal basis of human neoplasia. science,
' /
1943 221:227-236.
i Eippin C, Bailar JC III, Kohn HI, et al.: Radiation therapy for cervical cancer Late effects on life span and on leukemia incidence, cancer 1971 24 937-943.
t e
...: s )
.tli$
i i
i t
f O
O e
g l
l I
1
(~.
I Iu,c
- -.