Table of Contents - RERF · 2017. 8. 22. · RERF Update Volume 13, Issue 1, Spring 2002 RERF News...

Table of Contents

RERF News

The 36th Board of Directors Meeting ........................................................................... 1

The 28th Scientific Council Meeting ............................................................................ 2

New Chairman and Chief of Research ........................................................................ 4

Staff News .......................................................................................................................... 4

Multinational Peer Review of Clinical Studies ......................................................... 5

Current Status of Dosimetry Revision ......................................................................... 7

Articles

Ten Thousand Days Atop Hijiyama, by Akio Awa .................................................. 9

Radiation Risk, Longevity, and the Impact of the Comparison Group ............. 17 on Low-Dose Risk Estimates, by John B. Cologne and Dale L. Preston

Multifactorial Diseases in the Post-Genomic-Sequencing Era and ................... 23 Our Current Position, by Norio Takahashi

Facts and Figures ........................................................................................................ 28

Historical Vignette 1955, by Robert W. Miller ...................................................... 29

In Memoriam, R.M. Heyssel, ABCC Chief of Medicine 1956–58 ...................... 30

Research Protocols and Publications

Research Protocols ........................................................................................................... 31

Recent Publications ......................................................................................................... 32

This newsletter is published by the Radiation Effects Research Foundation (formerly the AtomicBomb Casualty Commission), established in April 1975 as a private, nonprofit Japanese foundation.It is supported equally by the government of Japan through its Ministry of Health, Labour andWelfare and that of the United States through the National Academy of Sciences under contractwith the Department of Energy.

RERF conducts research and studies—for peaceful purposes—on the medical effects of radiationon humans with a view toward contributing to the maintenance of the health and welfare of atomic-bomb survivors and to the enhancement of the health of all mankind.

Editor: Donald Pierce, Department of StatisticsTechnical Editor: Yuko Ikawa, Editorial & Publications Section

Editorial PolicyContributions to Update receive editorial review only and do not receive scientific peer review.

The opinions expressed herein are those of the authors only and do not reflect RERF policies orpositions.

Contact: Mailing address: RERF Editorial & Publications Section, 5-2 Hijiyama Park,Minami-ku, Hiroshima 732-0815 JapanTelephone: 81-82-261-3131; Facsimile: 81-82-263-7279Internet: [email protected]

Radiation Effects Research Foundation News and ViewsHiroshima and Nagasaki, Japan

��

mailto:[email protected]

�

�� RERF Update Volume 13, Issue 1, Spring 2002

RERF News

The 36th meeting of the Board of Directors washeld on May 26–27, 2001 in the Auditorium of theHiroshima Laboratory. Seventeen people were in at-tendance, including directors, supervisors, andobservers from the Ministry of Health, Labour andWelfare (Japan), Department of Energy (U.S.), andNational Academy of Sciences (U.S.). The Boardnormally meets once per year to consider importantoperational issues of RERF.

Following approval of the minutes of the 34th and35th Board meetings, Dr. Shigenobu Nagataki, RERFChairman, presented three reports: “Summary of thepast four years,” “Present problems,” and “Future re-search plans.” With regard to the Health Effects Studyon the Children of A-bomb Survivors, it was reportedthat the full-scale mail survey had been initiated basedon the agreement with the All Japan Second Genera-tion A-bomb Victims Liaison Council.

Dr. Nagataki further commented on three addi-tional issues: “Role of the Operating Committee,”“Role of the Chief Scientist,” and “Future plans.” Interms of the future plans, he emphasized the need topromote collaboration with other research organiza-tions.

Dr. Seymour Abrahamson, Vice Chairman andChief of Research, explained in detail proposals forestablishing liaisons with statistics programs at uni-versities and promoting the recruitment of researchscientists, based on the recommendations of theMultinational Peer Review of the Statistics Program.That was followed by deliberation on the past year’sresearch activities and audit reports, current year’sworking budget, and coming year’s provisional bud-get plan, all of which were approved.

Finally, the Board elected Dr. Burton G. Bennett(former Secretary of the United Nations ScientificCommittee on the Effects of Atomic Radiation[UNSCEAR]) as Chairman, Senjun Taira (RERFPermanent Director) as Vice Chairman, and Dr. EiichiTahara (professor emeritus of Hiroshima UniversitySchool of Medicine) as Permanent Director. Theirterms of office begin on 1 July 2001 and extend for aperiod of four years.

List of Participants

Permanent Directors:Shigenobu Nagataki, ChairmanSeymour Abrahamson, Vice Chairman and Chief of

Research

Senjun Taira, Permanent Director

Visiting Directors:Patricia A. Buffler, Dean Emeritus and Professor of

Epidemiology, School of Public Health, Univer-sity of California, Berkeley

Jonathan M. Samet, Professor and Chairman, Depart-ment of Epidemiology, The School of Hygieneand Public Health, The Johns Hopkins Univer-sity

Richard B. Setlow, Senior Biophysicist, BrookhavenNational Laboratory, and Adjunct Professor ofBiochemistry and Cell Biology, State Universityof New York at Stony Brook (submitted a letterof attorney)

Kazuaki Arichi, Councilor, The Japan Institute ofInternational Affairs

Toshiyuki Kumatori, Consultant, Radiation EffectsAssociation (submitted a letter of attorney)

Masumi Ohike, Former Chairman, Board of Direc-tors, Japan Anti-Tuberculosis Association

Supervisors:David Williams, Senior Financial Advisor, National

Academy of SciencesShudo Yamazaki, Former Director-General, National

Institute of Infectious Diseases

Scientific Councilor:Tomio Hirohata, Professor Emeritus, Kyushu Uni-

versity Faculty of Medicine

Representatives of Allied Agencies:Kazuhiro Kanayama, Chief, Medical Care Activities

Unit, General Affairs Division, Health ServiceBureau, Ministry of Health, Labour and Welfare

James H. Hall, Minister-Counselor (Science), Em-bassy of the United States of America

Giulia R. Bisconti, Energy Attaché, Director, U.S.Department of Energy Asia Office, Embassy ofthe United States of America

Evan Douple, Director, Board on Radiation EffectsResearch, National Research Council, NationalAcademy of Sciences

Catherine S. Berkley, Administrative Associate,Board on Radiation Effects Research, NationalResearch Council, National Academy of Sciences

Secretariat:Masaharu Yoshikawa, Chief of SecretariatRichard D. Sperry, Administrative Advisor, Secre-

tariat

��

�


RERF News

The 28th meeting of the Scientific Council was heldin Hiroshima on April 9–11, 2001. The meeting wasco-chaired by Dr. J. Martin Brown of Stanford Uni-versity and Dr. Tomio Hirohata of Kyushu University.After introductory remarks by RERF Chairman, Dr.Shigenobu Nagataki, and Vice Chairman/Chief of Re-search, Dr. Seymour Abrahamson, presentations weremade by each of the RERF departments: Clinical Stud-ies, Epidemiology, Statistics, Radiobiology, andGenetics.

In addition to overview presentations, more spe-cific reports of RERF projects were presented asfollows:

Breast cancer molecular analysis (Yuko Hirai, Ra-diobiology)Immunological homeostasis (Yoichiro Kusunoki,Radiobiology)In utero cytogenetics (Yoshiaki Kodama, Genet-ics)Molecular analysis of induced mutations in mice(Junichi Asakawa, Genetics)Progress in DNA microarray technology (NorioTakahashi, Genetics)Plans for the F1 clinical study (Saeko Fujiwara,Clinical Studies)Microbial infection in the Adult Health Study(Masayuki Hakoda, Clinical Studies)F1 mail survey and mortality in the F1 population(Akihiko Suyama, Epidemiology, Nagasaki)Radiation interactions in lung, breast, and livercancer (Gerald B. Sharp, Epidemiology)DS86 dosimetry revision (Shoichiro Fujita andHarry M. Cullings, Statistics)Effect of radiation on menopause (MichikoYamada, Clinical Studies)Natural menopause in Nagasaki women (ShizueIzumi, Statistics)Role of body mass index, serum cholesterol, sys-tolic blood pressure, and menopause (MasazumiAkahoshi, Clinical Studies, Nagasaki)

The Scientific Council affirmed the emphasisgiven at RERF to the main core projects on thehealth effects of radiation. In addition to the epi-demiology and clinical studies, the Councilrecognized that research to elucidate the etiologyof diseases is important to keep RERF in the fore-front of scientific research and to attract youngscientists. The primary general recommendationswere as follows, along with brief indication of RERFfollow-up on them:• Reiteration of the previous Scientific Council reco-mmendation that there be small-group “brain-

storming” sessions where scientists discuss researchgoals. [Such sessions have been initiated or resumedat the department level and are contributing to de-velopment of a Long Range Plan for RERF.]• The Scientific Council might henceforth conducttheir formal review only on alternate years, with someless formal interaction in the non-review years. [TheBoard of Directors requested that a decision on thisbe postponed. Discussions are continuing in regardto a possible restructuring of the review process.]• Careful consideration should be given to the pos-sibility of individual RERF scientists obtaining theirown research grants. [The Ministry of Education,Culture, Sports, Science and Technology (MEXT) hasrecently recognized RERF as eligible for submittinggrant applications, and proposals for Grants-in-Aidfrom MEXT have been made for 13 projects. Sevenof these have now been awarded.]• Consideration should be given to longitudinalcontinuation of the new F1 clinical study beyond theinitial investigation, with biennial mail contact andrepeat clinical visits each 10 years. [Very serious con-sideration is being given to that, including clinicalvisits more frequently than each 10 years.]

Extensive recommendations were then made foreach research department, which are summarized be-low.

Radiobiology – Appointment of 2–3 new staff at thedoctoral level should be made. The animal facilitiesshould be upgraded. Additional tumor tissue, espe-cially breast cancer tissue, should be obtained fromthe Adult Health Study (AHS) participants for use instudies utilizing modern genomic techniques such asprofiling of genetic expression in tumors, e.g., in earlyand late onset breast cancers. The department shouldcontinue to develop expertise in this area and insti-tute the necessary collaborations to perform thesestudies.

Genetics – The current direction of the departmentto develop expertise in new molecular-based tech-nologies to identify genetic alterations is supported.The staff should be expanded with young, short-terminvestigators. Collaborations should be establishedwith other scientists both within and outside RERF.Joint departmental retreats organized around a re-search theme may be helpful for problem solving,technology sharing, and generation of new researchideas. The department should be involved in theRERF F1 study during the planning and implementa-tion stages to be sure that appropriate data andspecimen collection plans are included.

��

�


RERF News

Clinical Studies – The F1 study should be made intoa longitudinal study using mail contact with partici-pants every 2 years, and a repeat full clinical studyshould be performed every 10 years. Current studiesthat are inactive or making poor progress should beclosed and a mechanism should be set up to limitapproval of clinical research proposals to five yearswith resubmission with a progress report if it is nec-essary for the study to continue for a longer period.A clinical research review committee should be cre-ated, made up of clinicians and statisticians to reviewclinical research proposals for scientific validity. Itshould include outside experts as well as RERF per-sonnel.

Epidemiology –The continued surveillance of theLife Span Study (LSS) sample, in utero cohort, andF1 cohort is essential for the mission of RERF. Be-cause of improved survival of cancer patients,analysis based on cancer incidence is becoming moreimportant than mortality analysis. RERF should con-tinue to be involved in the tumor registries inHiroshima and Nagasaki cities in order to maintainthe current high standards of the tumor registries,which enables RERF to make nearly complete as-sessment of incidence cancer cases among survivors.The maintenance of the current system to abstractcancer records of survivors at hospitals in bothHiroshima and Nagasaki cities is desirable. The in-formation obtained during the surveillance of manyyears of the LSS cohort, such as several mail surveysshould be explored. Cooperation between the De-partments of Clinical Studies and Epidemiology isdesirable to conduct “nested” case-control studiesusing stored serum. Careful consideration should begiven to develop first-rate scientific hypotheses tobe tested in such studies.

Statistics – The Council endorsed the recommenda-tions of the recently completed peer review of theDepartment of Statistics. In response to these, thedepartment will make further efforts to increase itsvisibility to the Japanese statistical community andto recruit Japanese as well as foreign statisticians toRERF. At least one statistician should be added tothe Scientific Council in the future. Efforts to developnew statistical techniques applicable to RERF data

will be increased. These will include: improved doseresponse models for predicting health effects at lowradiation exposures and continued exploration ofmechanistic models for cancer induction. Analysesof data on the primary RERF cohorts will continueto be a major focus, and attention will be given to thecommittee’s specific suggestions in this regard. Theseinclude use of adjusted survivor doses taking intoaccount random errors. More discussion within thedepartment and throughout RERF will be necessaryon confounding, adjustment, and interpretation issues.Also, further discussion will be needed to considerwhether there should be an oversight group to con-sider design, analysis, access, and documentation ofdatabases for current and future studies.

Members of the Scientific Council

Tomio Hirohata, Professor Emeritus, Kyushu Uni-versity Faculty of Medicine/Professor, NakamuraGakuen University

Yusuke Nakamura, Director of Human Genome Cen-ter, Laboratory of Molecular Medicine, Instituteof Medical Science, The University of Tokyo (ab-sent)

Masao Sasaki, Professor Emeritus, Kyoto Univer-sity (absent)

Yasuhito Sasaki, Chairman, Board of Directors, Na-tional Institute of Radiological Sciences

Shinichiro Ushigome, Visiting Professor, Jikei Uni-versity School of Medicine

J. Martin Brown, Professor and Division Chairman,Division of Radiation Biology, Department ofRadiation Oncology, Stanford University Schoolof Medicine

Joe W. Gray, Professor of Laboratory Medicine, Ra-diation Oncology, University of California, SanFrancisco (absent)

Gloria M. Petersen, Professor of Clinical Epidemi-ology, Mayo Medical School

Theodore L. Phillips, Professor and Chairman, Ra-diation Oncology, Cancer Center, School ofMedicine, University of California, San Francisco

Susan Preston-Martin, Professor, Department of Pre-ventive Medicine, Keck School of Medicine,University of Southern California

�


RERF News

!�"��

At the Board of Directors meeting held inHiroshima on 26–28 May 2001, Dr. Burton G. Bennettwas appointed to succeed Dr. Shigenobu Nagataki asChairman and Dr. Eiichi Tahara was appointed the newChief of Research. The Chief of Research’s positionhad been held by Dr. Seymour Abrahamson who alsoconcurrently held the position of Vice Chairman. Dr.Senjun Taira was continued as a Director for a secondterm and was now appointed Vice Chairman.

This series of appointments was a complete breakwith past practice since the founding of RERF withDr. Bennett being the first American Chairman, Dr.Taira being the first Japanese Vice Chairman, andDr. Tahara becoming the first Japanese Chief of Re-search. The terms of all the Directors began 1 July2001 and are for four years.

Dr. Bennett comes to RERF after having been formany years Secretary of the United Nations Scien-tific Committee on the Effects of Atomic Radiation(UNSCEAR). Since UNSCEAR serves as a primaryorganization utilizing the radiation and health re-search results from RERF, he is in a good position toprovide leadership in maintaining and improvingRERF contributions to the radiation protection com-munity. Dr. Bennett had his basic education in physicswith a M.S. degree from University of Washingtonand later a Ph.D. degree in environmental health sci-ences from New York University. Dr. & Mrs. Bennettlived in London for ten years where he worked onthe United Nations Environment Programs and theylater moved to Vienna for twelve years when he be-came Secretary of UNSCEAR. Most recently he hadreturned to the DOE Environmental MeasurementsLaboratory in New York City before coming to RERF.

When Dr. & Mrs. Bennett arrived in Hiroshima totake up the position of Chairman they were surprisedas they got off the train to be unexpectedly met by a

television crew and reporter with questions. This washis introduction to the interest that has been shown bythe media in the first American Chairman of RERF.

Dr. Tahara comes to RERF following a distin-guished career as a pathologist at Hiroshima University,and is the first RERF Director to be appointed fromthe local community. In his work he has advanced thearea of molecular methods in pathology, and is thuswell suited to provide leadership to RERF in the newresearch areas stemming from the revolution ingenomics. Dr. Tahara has most recently been a visit-ing professor at the University of California, San Diegoafter retiring from Hiroshima University in March2000. He had a long career at Hiroshima University,beginning in 1968, with positions including Chief ofthe Pathology Department and Director of the Facultyof Medicine. He spent two years in Germany at BonnUniversity in the early part of his career. Dr. Taharaplayed a central role in establishing the HiroshimaCancer Seminar Foundation in 1992.

Dr. Senjun Taira is the veteran of this group, nowstarting his second term as an RERF Director after along career in the Ministry of Health, Labour andWelfare. He had assignments in many posts includ-ing the Japan International Cooperation Agency(JICA) before coming to RERF in 1997.

It is appropriate to mention here that the Chief Sci-entist, Dr. Charles Waldren, very ably complementsthese three directors in functions of the ExecutiveCommittee and other leadership of RERF research.Dr. Waldren spent most of his career at the Universityof Colorado Health Sciences Center, working in mu-tagenesis and DNA repair and their relationship togenetic disease. Before coming to RERF he was Pro-fessor of Radiological Sciences at the Colorado StateUniversity School of Veterinary Medicine.

��

!�"�

Drs. Hidetaka Eguchi and Kazue Imai joined theDepartment of Epidemiology in July 2001 as Re-search Scientists. Both came from the Saitama CancerCenter Research Institute, to join Dr. Nakachi (Chief,Epidemiology), also from there, in the molecularepidemiology program at RERF. Dr. Eguchi receiveda Ph.D. in Life Chemistry at Tokyo Institute of Tech-nology and Dr. Imai received a Ph.D. in Psychologyat Tokyo Metropolitan University.

Dr. Masahiro Ito, Department of Genetics, retired

in December 2001. Dr. Ito graduated from TokyoUniversity of Agriculture in 1967, taking a researchfellow position at University of Hokkaido. In the earlydays of the ABCC genetics program, Dr. Ito followedDr. Awa from Hokkaido to ABCC in 1971, taking aposition in Nagasaki (see Dr. Awa’s article in thisissue of Update). Dr. Ito played an important rolefor 20 years in Nagasaki, continuing in Hiroshimawhen the Nagasaki Radiobiology Department wasmerged with Hiroshima departments.

�


RERF News

The fifth Multinational Peer Review of RERF de-partments was held on November 28–30 at HiroshimaLaboratory Auditorium. Annual peer reviews havebeen held since 1997 in response to the Blue RibbonPanel’s recommendation of this practice. This reviewof Clinical Studies for both Hiroshima and Nagasakicompletes the cycle of all departments, including also:Radiobiology, Epidemiology, Genetics, Statistics.

The meeting opened with an address by Dr. BurtonG. Bennett, Chairman of RERF, and an RERF over-view given by Chief of Research Dr. Eiichi Tahara.Following that, Dr. Gen Suzuki, Hiroshima ClinicalStudies Department Chief reported on the major ac-tivities of his department and the Adult Health Study(AHS) program, and Dr. Masazumi Akahoshi,Nagasaki Clinical Studies Department Chief, summa-rized the activities of his department. Subsequently,the following presentations were made and active dis-cussions held: F1 Health Study (Saeko Fujiwara); AHSlongitudinal data analysis methodology (MichikoYamada); Thyroid diseases (Misa Imaizumi); Inflam-matory response and immune response to microbialinfection (Masayuki Hakoda); Radiation exposure anddiabetes mellitus (Saeko Fujiwara); Radiation expo-sure and senile cataract (Kazuo Neriishi);Cardiovascular disease (Masazumi Akahoshi); Inter-national Collaboration Studies (Ni-Hon-Sancomparative study on cardiovascular diseases, Ni-Hon-Sea comparative study on dementia) (MichikoYamada); International and domestic collaborationstudy (Ni-Hon comparative study on osteoporosis)(Saeko Fujiwara); Future research plan—Sicca syn-drome (Ayumi Hida; read by Masazumi Akahoshi);Future RERF cohort study (Gen Suzuki).

As the RERF departments having direct contactwith the A-bomb survivors and their children, Clini-cal Studies in Hiroshima and Nagasaki havecontributed in many ways to the health and welfareof the A-bomb survivors and their children, includ-ing through the various research programs mentionedabove. The departments also play a crucial rolethrough the systematic collection of clinical data andbiological samples that are important to research workof other RERF departments. The departments haveplaced special emphasis on research of the radiationeffects on non-lethal diseases. The results are reportedin twenty to thirty-plus scientific papers publishedeach year in the major scientific journals in Englishand in Japanese.

The departments reported on future plans to con-duct, in addition to the ongoing F1 health examinations,research using stored biological samples and to per-form analyses of newly found modifying factors andgenetic traits related to the onset of diseases.

On the final day of the meeting, Dr. Theodore L.Phillips, panel chairman, provided a preliminary over-view of the recommendations, and some other panelmembers added more specific comments. The de-tailed recommendations, received later, aresummarized below. The review panel highly evalu-ated the research achievements of the clinical groupand the papers they have published. It also recom-mended that the departments further strengthen theanalytical program of modifying factors (confound-ers) related to the onset of diseases. As a conclusion,the panel encouraged the staff, saying that RERF willcontinue to be acknowledged as a globally impor-tant research institute if its future direction isestablished as utilizing RERF’s long-accumulateddata and samples for the research in elucidating theetiologies of diseases.

Summary of recommendations received in writ-ing later is as follows:• Initial review of new studies should be expanded

to include outside reviewers when the RERF pro-tocol committee does not have expertise in thearea of study proposed.

• There should be a systematic strategy for priori-tizing hypotheses and the use of blood products.Very strict policies on the use of preserved bio-logic specimens and international review of anyuse of limited specimens should be instituted. Di-vision of specimens into multiple tubes shouldonly take place after such approval for their use,not on a routine basis.

• An annual review of open studies should be con-ducted by an RERF committee, to determine thatprogress is adequate to justify the investment ofresources.

• The Departments of Clinical Studies should be-come more involved in studies of multifactorialinfluences on cancer incidence and outcome.

• The F1 study should conduct the initial baselineexaminations and contacts with the cohort asplanned. During the initiation of this process theexpected number of events in common diseasesshould be calculated in order to refine and re-duce the number of clinical endpoints and allow

� � � � � � � #�� $�"% � ��

Performance highly evaluated, with recommended strengthening of theanalysis of modifying factors related to onset of disease

�


RERF News

focus on high probability events in future re-ex-aminations. High subject participation rates mustbe encouraged. Every effort should be made tomaximize the power of the research to identifythe hypothesized effects.

• In the AHS longitudinal study more attentionshould be paid to validation of endpoints such asdeath certificate data versus case-control studies.

• The classification systems for disease should bestandardized.

• The thyroid disease study is a very nice piece ofresearch. Consideration should be given to ex-tending the study to Hiroshima and to initiatingfollow-up studies in subclinical hypothyroidism.

• The studies of infectious agents have shown thatindividual agents are not important but that theinitiation of a chronic inflammatory process maybe very important in cardiovascular disease andaging in general. Future studies should focus onmechanisms that trigger chronic inflammatoryresponses.

• The diabetes mellitus studies are very interestingbut the discrepancy between results in Hiroshimaand Nagasaki needs to be resolved by means ofHLA typing.

• The cataract study may be very important in show-ing that radiation accelerates aging. It should becompleted and consideration given to adding reti-nal photographs and fluorescein angiograms tothe studies.

• The cardiac disease studies should concentrateon specific biochemical risk factors such as cho-lesterol levels etc. rather than on surrogatemarkers.

• The osteoporosis work is outstanding. Futurestudies may want to focus on environmental and

lifestyle factors influencing osteoporosis and oncomparison of Japanese and Caucasians.

• The international studies are of high quality, ben-efit RERF, and should be continued and expandedif opportunities arise.

• Broader interaction of the clinical departmentswith the epidemiology departments is encouraged.

Peer Review Panel Members

Theodore L. Phillips, Professor and Chairman, Ra-diation Oncology, Cancer Center, School ofMedicine, University of California, San Francisco(Panel Chair and RERF Scientific Councilor)

Yasuhito Sasaki, Chairman, Board of Directors, Na-tional Institute of Radiological Sciences (RERFScientific Councilor)

John Danesh, Professor, Epidemiolgy and Medicine,Head of the Department of Public Health andPrimary Care, University of Cambridge, Instituteof Public Health

Donald R. Harkness, Professor Emeritus, Universityof Wisconsin Medical School

Yoshitomo Oka, Professor, Molecular Metabolismand Diabetes, Internal Medicine, Tohoku Univer-sity School of Medicine

Hajime Orimo, Director, Metropolitan GeriatricHospital

Kazuo Ueda, Professor, Kyushu University Schoolof Health Sciences

Lon R. White, Principal Investigator, Honolulu-AsiaAging Study, Senior Neuroepidemiologist, PacificHealth Research Institute, Professor with jointappointments in the Schools of Nursing and Medi-cine, University of Hawaii at Manoa

�


RERF News

In the previous issue of RERF Update some in-formation was given about reasons for revision ofthe DS86 radiation dose estimates for A-bomb sur-vivors, and the progress at that time. Since then therehas been substantial progress, including a meeting inHiroshima in April 2002 of the Joint U.S.-JapaneseWorking Groups on the Reassessment of A-bomb Do-simetry. Below we will provide verbatim the pressrelease following that meeting.

It is now expected that the fundamentals of a newdosimetry system may be completed by Fall 2002,with implementation of that system and preparationof a final report continuing for several months afterthat. Although there apparently will be smaller over-all changes in dose estimates than anticipated, therewill be some more specific improvements in accu-racy. For example, there should be substantiallyimproved shielding calculations for a large numberof Nagasaki factory workers who received high doses,and for low-dose survivors in both cities who wereat some distance from the bombs but shielded by ter-rain. Generally, the process of having very carefullyconsidered every aspect of revision will lead to moreconfidence in the dosimetry system.

The April 4, 2002 press release is as follows:

Since the detonation of the A-bombs, the survivorsof the bombings and the international community re-sponsible for radiation protection standards havedepended on the A-bomb radiation dosimetry systemsfor the determination of accurate radiation doses. Asadvances in technology for dose calculation from thebombs have been made, and the ability to check thosecalculations using activation measurements has im-proved, changes have been made in the dosimetry fromT57D to T65D to the current DS86 system.

It is widely recognized that the DS86 system ac-curately calculates gamma rays, which constitute themajority of the radiation exposure to survivors. Sincethe implementation of DS86, thermal neutron acti-vation measurements have been made that appearto differ from those calculated by the dosimetry sys-tem. Although neutrons are a small fraction of thetotal dose, it was thought to be important that thisapparent discrepancy be addressed in order to givethe survivors and the radiation risk assessment com-munity confidence that the dosimetry system wasaccurate in all significant details.

In the 16 years since DS86 was implemented,remarkable advances have been made in computa-tional capabilities and measurement technology.

These improvements now make it possible for scien-tists to measure and calculate trace amounts ofactivation from the bombs in ways and details notpossible in the 1980’s. Over the past 18 months, allof this new technical capability has been brought tobear in a further re-evaluation of the dosimetry sys-tem conducted by the Working Groups on theReassessment of A-bomb Dosimetry in the U.S. andJapan.

The joint Japanese and U.S. Working Groups onthe Reassessment of A-bomb Dosimetry met on April3rd and 4th at the Radiation Effects Research Foun-dation in Hiroshima to review their work and discussfinal preparations for a new dosimetry system, DS02,to replace DS86. The most detailed recalculation todate of the Hiroshima and Nagasaki bombs and anexhaustive evaluation of the data indicate the neu-tron discrepancy that gave rise to this reassessmentwill be resolved.

Changes in the nuclear data and calculation tech-niques of DS02, and extensive verification usingnuclear test data, have refined the Nagasaki dosim-etry. These refinements confirm the DS86 conclusionthat 21 kilotons and height of 503 meters above thecity are the most accurate parameters for the deto-nation. The degree to which the calculations for theNagasaki detonation have been verified againstnuclear tests now makes the Nagasaki calculation abenchmark for dose reconstruction. The refinementsin the DS02 calculation for Nagasaki produce anincrease of less than 10% in the dose from gammarays while decreasing the neutron dose by 15 to 45%at distances of 1,000 to 2,000 meters from the bomb.Significant refinements in calculation of the shield-ing for workers in the torpedo factories reduceoverall doses for these workers by 20 to 40%.

Similar refinements in the new DS02 calculationof the Hiroshima bomb produce better agreementbetween neutron calculations and measurements.New measurements of high-energy neutron activa-tion at Hiroshima were particularly important inconfirming the calculations. A measurementintercomparison is being carried out to confirm thisagreement. The reassessment for DS02 confirms theDS86 conclusion that the bomb produced an explo-sion that was equivalent to 15 kilotons. A carefulreview of the data for the height of the detonationindicates that best agreement is achieved by increas-ing the burst height by approximately three percentto 600 meters. While this change produces betteragreement between the calculation and the measure-ments, it produces almost no change in survivor

��&��$��

�


RERF News

doses. Gamma-ray doses are unchanged within thefirst 2,000 meters of the bomb. Neutron doses areessentially unchanged for survivors out to approxi-mately 1,500 meters, beyond which they are slightlydecreased.

Thus, within the uncertainties of such difficultmeasurements and complex calculations, acceptableagreement has been achieved for both the neutronand gamma radiations. As a practical matter, while

the DS02 system will be more accurate than DS86,it will produce little change in the overall radiationdoses for survivors. However, DS02 should producea much more confident basis for the entire radiationdose to the survivors than was previously available.

Measurement results of exposed samples collectedhave been essential for dose reassessment. We wouldlike to express our sincere appreciation to the citi-zens who have helped us collect the samples.

�Ten Thousand Days Atop Hijiyama


Ten Thousand Days Atop Hijiyama

Akio Awa

PreambleAs of June 30, 1995, I retired from the Radiation

Effects Research Foundation (RERF). I joined theAtomic Bomb Casualty Commission (ABCC) onJanuary 16, 1967 and have spent 28 and half yearswith this organization. Alternatively put, I have spent10,393 days at Hijiyama—8 years and 3 months withthe ABCC and the remaining 20 years and 3 monthswith the RERF. During this period, I was under theguidance of two ABCC Directors (Drs. George B.Darling and LeRoy R. Allen) and three RERF Chair-men (Drs. Hisao Yamashita, Masao Tamaki, andItsuzo Shigematsu).

It was my pleasure to participate in research atthe Cytogenetics Laboratory. A number of col-leagues have gathered here and departed. At present,there is virtually no one who is familiar with thehistory of the laboratory. Before everything fadesaway and is forgotten, as a former permanent mem-ber of the laboratory, I feel obliged to prepare amessage regarding its history. This document is in-tended to take the form of a personal account of theevents I encountered over the past decades in theCytogenetics Laboratory. Fortunately, Dr. SeymourAbrahamson, former Chief of Research and editor-in-chief of RERF Update kindly offered me anopportunity to publish this here. I am aware thatthis article will inevitably contain some errors andam solely responsible for them.

The document will consist of two parts, publishedseparately. This part one details the genesis of Cyto-genetics Laboratory, and a series of chromosomestudies based on the Adult Health Study (AHS) popu-lation and a summary of this work. Part two willdescribe an outline of another important research mis-sion, i.e., cytogenetic study of the children ofatomic-bomb survivors, the so-called F1 study. In-cluded here also are my recollections of many of thoseunforgettable people to whom I am indebted in manyways, and who have—even now—guided me spiri-tually and scientifically.

Part 1. Cytogenetics Laboratory at ABCC/RERF — Past and Present

1. PrologueIn 1952, I entered the Hokkaido University un-

dergraduate course in science, and completed theBiology course in the Faculty of Science in 1956. Bycoincidence, the period between 1952 and 1956 wasan epoch-making era in the twentieth century de-

velopment of biology. In April of 1953, Watson andCrick first uncovered a famous double-helix structureof deoxyribose nucleic acid (DNA), and published ashort but path-breaking paper in Nature.1 A few yearslater, in early 1956, Tjio and Levan demonstratedunequivocally that the true number of human chromo-somes is 46.2 These important findings had a strongimpact on me throughout my scientific career.

In the postgraduate course at Hokkaido Univer-sity, my thesis work was to study human andmammalian chromosomes under the guidance of Pro-fessor Sajiro Makino (1906–1989). As will be fullydescribed later, Dr. Makino has been internationallyrecognized as a human and mammalian cytogeneti-cist. He was also well known as an excellent leaderfor the education of his youngster colleagues.

I was a lazy postgraduate student with a short andcapricious temper unsuitable for scientific research.Surprisingly, I was fond of microscopic work, andwas able to accurately draw and analyze chromo-somes directly under the microscope. This may berelated to my hobby since childhood of drawing car-toons. This acquired habit became an indispensabletool for performing time-consuming work on detec-tion of chromosome aberrations.

2. Human Cytogenetics — çà et làA. All about Chromosomes in Man*

Chromosomes are present in the cell nucleus. Theirbasic molecular component is DNA. As mentionedbefore, the exact number of human chromosomes is46. They are small and slender, only visible by the useof a microscope; even the longest of the chromosomesis less than 1/100 mm. Each chromosome carries aconstriction called “centromere,” which is an impor-tant landmark to characterize individual chromosomes.

Chromosomes can be observed only when cellsare at a stage of cell division (mitosis) called“metaphase.” The period of metaphase is extremelyshort, so that in order to examine chromosomes, it isnecessary to obtain tissue samples whose cells areactively proliferating. Until tissue culture techniquewas introduced to biological research, it was diffi-cult to get the required fresh materials from healthypeople. This was the reason why progress in humanchromosome study had been greatly delayed.

In early 1950’s, there occurred a remarkable waveof technical progress in biology. Tissue culture meth-ods, as mentioned above, became in routine use for

��Ten Thousand Days Atop Hijiyama


medical and biological research. Particularly note-worthy was a new development of peripheral bloodculture by Moorhead and his colleagues in 1960.3

Fresh human specimens for subsequent culture wereeasily obtained aseptically by vein puncture. For chro-mosome analysis, it requires only a few milliliters ofperipheral blood. Moreover, it takes only two daysof cultivation in vitro to yield sufficient number ofmitotic cells.

Besides tissue culture techniques, other techno-logical improvements have helped expand theknowledge of human chromosomes. These include(1) use of colchicines that arrest the mitotic cells atmetaphase, and (2) an “air-dry” method4 that canspread metaphase chromosomes two-dimensionallyon microscopic slides, and more importantly, over-lapping of chromosomes is reduced considerably. Useof enlarged photographic prints of metaphases alsohelped facilitate the efficiency of chromosome analy-sis.

Such technical advances have made detailed chro-mosome analyses far easier, and have prompted thedevelopment of human clinical genetics. Certain ge-netic diseases are now known to be associated withchromosome anomalies due either to changes in num-ber (chromosome aneuploidy) or in structure (structuralrearrangement). For instance, the chromosome num-ber of those with Down syndrome is 47, with anaddition of an extra chromosome 21.

The introduction of lymphocyte culture method alsohas expanded the research area in radiation cytoge-netics. A research group in Edinburgh Universitystudied lymphocyte chromosomes in patients given X-ray treatment for ankylosing spondylitis.5 Theyreported that a variety of X-ray-induced chromosomedamage was observed. Furthermore, they later con-firmed not only their previous findings but alsodemonstrated that cells carrying aberrations couldpersist for years in patients’ circulating lymphocytes.6

These findings aroused the interest of scientists in ra-diation research, and need for investigation of radiationeffects onto atomic-bomb survivors of Hiroshima andNagasaki.

B. All about Radiation-induced ChromosomeDamagea. Formation of chromosome aberrations

Chromosomes are packed in a mass-like structure,called a nucleus, in the interphase cell—during an in-terval between two mitoses. When the nucleus isexposed either to ionizing radiation, or some otheragents such as toxic chemicals and certain kind of vi-ruses, chromosome threads are broken, and the numberof induced breaks is proportional to the amount of doseadministered. With the aid of repair enzyme, brokenends of affected chromosomes are largely restored toan original state. However, as the radiation dose in-creases, breaks are (1) left unrestored as chromosome

fragments, or (2) joined between wrong broken part-ners , thus newly producing chromosomeaberrations—more precisely, exchange aberrations,or structural rearrangements of chromosomes. Theevidence from experiments to date has shown thatchromosome aberration frequency increases with in-creasing radiation dose, and that the frequency isinfluenced by radiation quality such as neutron, α-,β-, γ-, and X-rays.

b. Types of chromosome aberrationsAs shown in Figure 1, an exchange between two

breaks within a chromosome results in the formationof either a ring chromosome (plus a fragment with-out centromere, called acentric fragment) or aninversion of a chromosome. These categories of ab-errations are termed “intra-chromosomal exchange.”Theoretically, both types of aberrations occur withan equal probability. Similarly, exchanges of breaksbetween different chromosomes result in a dicentricchromosome plus an acentric fragment, on the onehand, and on the other hand, two-translocated chro-mosomes, called “inter-chromosomal exchange.”“Dicentric” means a chromosome having two cen-tromeres. Although there are many other types ofchromosome aberrations, for convenience I will dealexclusively with the following four main types ofaberrations, i.e., rings, dicentrics, inversions, andtranslocations.

Chromosome aberrations are classified in differ-ent ways, as seen in Figure 1. One group consists ofrings and dicentrics, while the other with inversionsand translocations. The former is called “asymmet-ric exchange,” and the latter “symmetric exchange.”Due to their morphological peculiarity, the formertype of aberration is unequivocally detectable. Onthe contrary, symmetric exchange is difficult to iden-tify with certainty, because the changes in shape andlength of aberrant chromosomes often are so subtlethat they are not discriminated as aberrant chromo-somes. To score inversions and translocations withfull efficiency requires long experience and exper-tise.

However, dicentrics and rings are known to causemitotic disturbance because of their structural pe-culiarity. The cells carrying such aberrations are lostin the subsequent cell generations. Thus, their fre-quency decreases sharply with time, being adrawback in use of dicentrics and rings as a bio-logical marker for those who were irradiated manyyears prior to chromosome examination. In contrast,there is no such disadvantage for translocations andinversions at mitosis, and the level of such frequencyis maintained constantly for decades after exposureto radiation. For this reason, asymmetric exchangesare also called unstable aberrations (Cu type), andsymmetric ones as stable aberrations (Cs type). Ishall use this terminology exclusively throughoutthis article.



The types of chromosome aberrations and the wayto detect them have been fully described in the RERFHome Page (www.rerf.jp/Gene/eng/giemsa.htm).

3. Birth of Cytogenetics LaboratoryThe first ABCC genetics program was the exten-

sive search for untoward pregnancy outcomes(1948–1954) initiated by Dr. James V. Neel and col-leagues. In that era chromosome analysis was ratherprimitive, e.g. even the number of chromosomes wasnot established until the work of Tjio and Levan in1956.2 It was only in the 1960’s that the ABCC cy-togenetic studies began to flourish. See Dr. WilliamJ. Schull’s book7 for further historical perspective.

During the period between 1963 and 1964, bothDrs. Michael A. Bender and Ernest H.Y. Chu, cyto-genetic experts in Oak Ridge National Laboratory,came to ABCC for a site visit to seek the possibilityof initiating cytogenetic programs there. Based ontheir recommendation, Dr. Richard E. Slavin andcolleagues in ABCC Pathology Department startedthe cytogenetic project. One of their results on thecytogenetic screening of Down syndrome cases inHiroshima was published in 1967.8

Dr. Arthur D. Bloom began a full study in 1965.Prior to his assignment with the ABCC, he receivedtraining for human chromosomes under Dr. J.H. Tjio.The Cytogenetics Laboratory was transferred fromthe Department of Pathology to the Clinical Labora-tory headed by Dr. Howard B. Hamilton.

In collaboration with Dr. Nanao Kamada(Hiroshima University, former director of ResearchInstitute of Radiation Biology and Medicine) and Dr.Tetsuya Iseki (Nagasaki University, current presidentof Nagasaki Prefectural Medical Association), bothDrs. Bloom and Shotaro Neriishi (Nagasaki BranchLaboratory) started the cytogenetic survey on the

Adult Health Study (AHS) population. They alsostudied some survivors who experienced prenatal ex-posure to A-bomb radiation (in utero survivors).

Their initial AHS study consisted of 174 proxi-mally exposed survivors, whose estimated radiationdose (T65D system) was more than 200 rad,** andthe 181 controls (0 rad group). They reported a sig-nificant elevation in the frequency of grosschromosome damage (such as dicentrics, rings, acen-tric fragments, and abnormal marker chromosomesof exchange type).9,10 In this study, however, they didnot analyze the data for demonstration of a dose-re-sponse relationship.

In the 1960’s Japanese scientists outside ABCCpublished reports on chromosome aberrations inatomic-bomb survivors. In 1968 Sasaki and Miyataexamined 52 Hiroshima atomic-bomb survivorsand published an excellent paper in Nature.11 Theyscored mainly dicentrics, rings, and acentric frag-ments out of an average of more than 1,000

Figure 1. Diagrammatic representation of four major types of chromosome aberrations followingexposure to ionizing radiation

Front row (from left to right): Dr. Shotaro Neriishi, Dr.Sajiro Makino, and Dr. Howard B. Hamilton. Backrow: Dr. Michihiro C. Yoshida, Dr. Takeo Honda, andDr. Toshio Sofuni [1967]

http://www.rerf.jp/Gene/eng/giemsa.htm



metaphases per person, and made radiation doseestimations for individual survivors on the basisof their bio-dosimetric formula, called the Qdrmethod. Their results showed a close relationshipbetween biological and physical dose estimates,the latter of which was derived from both the dis-tance from the hypocenter and the types ofshielding materials between A-bomb burst pointand individual survivors.

4. Expanded Cytogenetic Study: 1967–1975In early December of 1966, Professor Makino in

Hokkaido University was invited to ABCC to explorecollaboration between Hokkaido University andABCC. I accompanied Dr. Makino to ABCC, andnegotiated with the representatives of ABCC, i.e.,Drs. Hamilton, Bloom, and Hiroshi Maki (ABCC As-sociate Director). Dr. Darling, ABCC Director, didnot attend this meeting since he was out town. Aftera lengthy discussion, we reached to the agreementthat Hokkaido University would support ABCC cy-togenetic projects by participation of Dr. Makino’sresearch staff.

Only a month later, in January of 1967, Dr. TakeoHonda and I were assigned to ABCC as permanentresearch associates. In the same year, two more mem-bers also came to ABCC, Dr. Michihiro Yoshida(1967–1969) to Nagasaki and Dr. Toshio Sofuni(1967–1979) to Hiroshima. In the subsequent years,others from Hokkaido joined the ABCC: Messrs.Hachiro Shimba and Kazuo Ohtaki in Hiroshima, andMr. Masahiro Itoh in Nagasaki.

The addition of research staff continued. Dr.Mimako Nakano from Hiroshima University joinedus in 1977. Mr. Yoshiaki Kodama came to RERF in1980 as a replacement for Dr. Sofuni, who was as-signed to National Institute of Hygienic Science inTokyo. Finally in 1994, Dr. Nori Nakamura, currentlyChief of Genetics Department, was transferred fromthe Radiobiology Department. Dr. Sadayuki Ban wasalso part of the Cytogenetics Laboratory from 1985to 1994.

Dr. Bloom returned to the United States after three

years with the ABCC and joined to Dr. Neel’s groupat University of Michigan Medical School in AnnArbor. Indeed, he did a good job and his name shouldlong be remembered as a founder of ABCC Cytoge-netics Laboratory. Mr. Shozo Iida, then chieftechnician of the laboratory, also left for the U.S. tohelp Dr. Bloom set up a new cytogenetic laboratoryin Michigan.

After Dr. Bloom’s departure to the U.S., I wasasked to take over the management of the laboratory.I carefully looked at the ABCC organization chartand tried to remember the names of key staff mem-bers in other departments. I also went around inABCC to confirm what I had remembered. All ofthis helped me a great deal when we had to establishin inter-departmental management system for Dr.Neel’s new project on biochemical screening of theF1 children of survivors.

Prior to the onset of laboratory management, Ifaced with two urgent issues. Firstly, we all felt thestudy should be oriented to the determination of thedose-response relationship in A-bomb survivors. Todo this, however, we needed to expand the samplesize to cover survivors in all dose ranges. Sampleselection from among the AHS participants was madein collaboration with a statistician in charge of cyto-genetic project, Mr. Takashi Matsui, now professorin Statistics at Dokkyo University. Indeed, he wasour best partner. In those days, only statisticians coulddeal with radiation dose data (T65D). Any membersother than statisticians were not allowed to obtainindividual dose of survivors, to avoid any observers’bias prior to or in the course of survey.

Secondly, the level of techniques both for lym-phocyte culture and preparation of microscopic slideswas rather poor in our laboratory. At that time, it wasdifficult for us to routinely observe more than 100cells per person. In addition, it was already more than20 years since A-bomb explosions in 1945. We hadthus anticipated that most of the cells carrying un-stable aberrations might be eliminated fromcirculating blood of survivors. Consequently, evenfor heavily exposed survivors, we might fail to de-tect any unstable aberrations within the ranges of 100analyzable metaphases per person. We empiricallyknew that we could find on an average about 5% ofcells bearing stable aberrations at 100 rad.

We thus decided as a routine procedure to scoreall types of stable and unstable aberrations per 100metaphases per survivors. Our decision was quitecontrary to the general strategy of scoring exclusivelyunstable aberrations, an easy and well established bio-dosimetry employed in most of cytogeneticlaboratories in the world. We knew that those re-searchers who analyze stable aberrations were in theminority group. We were convinced that stable aber-ration scoring could equally be efficient for the

Dr. Howard B. Hamilton[1981]

Dr. Arthur D. Bloom[1983]



analysis of dose-response relationship, provided thatthe pooled data were analyzed by breakdown of sur-vivors into 100 rad-interval dose groups.

In the course of the routine microscopic work,what worried us was finding too many metaphaseswith false negative aberrations, i.e., metaphasesjudged as normal even when true aberrations werepresent. Our data would have lost credibility if aber-rations were overlooked by misjudgment of theobservers, though it is impossible to eliminate allmistakes. In order to minimize the number of falsenegative aberrations, all of the cells in which defi-nite and suspected structural aberrations weredetected by direct microscopy were then photo-graphed and subjected to karyotype analysis. Wherepossible, we vigorously pursued the origin of chro-mosomes involved in the exchanges of aberrationsby karyotyping. Our obligatory scoring procedurewas (1) to record on the laboratory score sheets suchitems as case ID number, slide number, cell number,and the X-Y axis location of the cell on the micro-scopic stage, and (2) to take photographs for allmetaphases suspected by examiners of having anyabnormality. Considering the cost of a frame of 35mm negative films, this procedure seemed to be themost reliable and the cheapest. By this system, abil-ity of observers to detect aberrations became verymuch improved and consistent. It is worth noting thatnearly 400,000 frames of metaphase pictures in 35mm negative films have been kept to date in the largefilm cabinets of the laboratory.

The results of expanded AHS surveys have beenintermittently reported elsewhere.12,13,14 Major find-ings of the study are as follows: (1) The frequency ofaberrant cells increased with increasing radiationdose, and was generally higher in Hiroshima than inNagasaki. (2) The mode of dose-response relation-ship was linear in Hiroshima, while it wasdose-squared fashion in Nagasaki. The observed dif-ference might reflect the difference in a mixed ratioof neutron and gamma rays between the two cities.This view, however, was re-evaluated upon adoptinga new dosimetry system of DS86, in which neutrondose was drastically reduced from Hiroshima radia-tion spectrum. The observed inter-city difference inthe neutron-gamma rays thus dimished. Even now,however, inter-city difference in the frequency ofchromosome aberrations still exists, though lesserin degree, when data are reanalyzed using DS86system. (3) Frequency of stable aberrations (trans-locations and inversions) is predominantly higherthan that of unstable aberrations (dicentrics and rings).The former type of aberration is the major contribu-tor for the dose-response relationship. Thedose-response for unstable aberrations was also dem-onstrated. (4) There was evidence of in vivo clonesof cells with cytogenetically identical aberrations inhigh dose survivors. (5) There was a small fractionof survivors either with high dose and low aberra-

tion frequency or with low dose and high aberrationfrequency. These survivors were regarded as over-dispersion cases outlying the normal range of the doseresponse relationship. We termed them “cytogeneticoutliers.” Our interpretation was that the outlyingvalues stemmed from errors in dose estimate ratherthan due to biological individual difference. Obvi-ously, we needed further validation of cytogeneticallyoutlying cases.

In February 1975, before the re-organization ofABCC to RERF, a special scientific committee meet-ing was held at ABCC, chaired by Dr. James F. Crow(professor in Genetics, University of Wisconsin),called the “Crow Committee.” The charge to the com-mittee was to carefully review each item of the ABCCscientific programs as to which of the items shouldbe taken over by the new foundation. When the rec-ommendation of the committee15 was reported, wecytogenetic members were very happy to know thatthe report included the issue of “cytogenetic outli-ers,” and recommended pursuing this even morevigorously. Part of the Crow Committee recommen-dation is directly cited as follows.

“It would be especially valuable to study personswhose cytogenetic findings are grossly discrepantwith regard to estimated dose. On the one hand, sucha study may lead to improved dosimetry; on the other,it might reveal possible human phenotypes with ex-treme radiation resistance or susceptibility. Muchconcern has been expressed for persons who may beespecially susceptible to chemicals; this might wellbe a concern also with radiation.”

After the re-organization of ABCC to RERF, ex-cept for the F1 cytogenetic work our AHS study wasconcentrated to conduct repeat examinations of sur-vivors who were cytogenetically categorized indifferent groups. In one of the categories, we ran-domly chose an appropriate number of survivors, forwhom cytogenetic survey had already conducted, inall dose ranges, to see if aberration frequency notedpreviously would be consistently maintained. Alter-natively put, had our ability to detect aberration beenmaintained constantly over time? Important inclu-sions of survivors into repeat examination were the

Dr. James F. Crow (left) and the author [1997]



outliers, both high-dose with low aberration fre-quency and low-dose with high aberration. Alsoincluded were survivors carrying a clone of cells withcytogenetically identical aberration.

The results of repeated examinations reconfirmedour previous findings. Aberration frequency for eachof survivors had been maintained consistently for longperiod of time. The same held true for both outliersand clone-carriers. As for controversial cases of out-liers, it was proved that the discrepancy of aberrationfrequency with regard to the dose was found to stemfrom errors in dosimetry rather than from differencein radiosensitivity or radioresistance of survivors.Interestingly, clones of aberrant cells with a constantfrequency still continued to persist for many years inthe body of clone carriers.

In 1982, the T65D A-bomb radiation dose esti-mate system was criticized as having many problems.Thus a team of scientists from the U.S. and Japanwas formed, and started a re-evaluation work thattook them several years to complete. In 1986, a newsystem “Dosimetry System 1986,” called DS86, wasfinally established. One of the major changes fromthe previous system was a substantial reduction ofneutron dose and increase in gamma dose inHiroshima. Thus the ratio of neutron to gamma esti-mates differed less between Hiroshima and Nagasakithan before.

When a new system was available, our cytoge-netic data were reanalyzed using DS86 values. Theresults indicated that the inter-city difference in ab-erration frequency per unit dose became smaller thanthat of the previous result. The degree of differencein the shape of dose-response curves also becamesmaller, but there still existed some inter-city differ-ence; Hiroshima frequency is still higher than inNagasaki.16

We had a difficult time between 1982 and 1986when dosimetric work was still underway. All of ourstudies were directly related to A-bomb radiationdose. Doing studies without dose was actually im-possible. Thus we spent most of the time carryingout a detailed analysis of the types and frequenciesof stable aberrations, data being derived from bothG-band and conventional staining analyses. The datarevealed that translocations were found to be abouttwo-thirds of the total stable aberrations, and theywere a primary contributor to the dose-response re-lationship.17 It was also demonstrated that, providedthat G-band method can detect all types of aberra-tions with a full efficiency, conventional analysiscould detect aberrations of about 70% of the G-bandefficiency.

The year 1988 was a dawn of a new molecularstudy at the Cytogenetics Laboratory. In close col-laboration with Lawrence Livermore National

Laboratory (LLNL) in California, a new techniquecalled a chromosome painting method, or more pre-cisely the FISH technique, an abbreviated form offluorescence in situ hybridization, was introduced toour laboratory. By this technique translocations canbe detected easily, rapidly, and accurately, so that thetechnique seemed to be promising for identificationof stable aberrations in A-bomb survivors, and itwould be eventually employed as our routine cyto-genetic procedure. Since the procedure of techniqueis too complicated and highly technical, I do not de-scribe it here in detail. In principle, composite probesfrom chromosomes 1, 2, and 4 are hybridized to thesame chromosomes in the microscopic slide. Thusthe chromosome of interest is selectively painted,while the rest of non-target chromosomes are leftunstained (or stained with other dye). Any transloca-tions involving chromosomes 1, 2, and 4 arevisualized as a bicolor chromosome(s) so that, evenfor lay people, they are easily distinguishable fromother elements in the metaphase. In 1989, a new jointresearch program between LLNL and RERF wasapproved.18 This research protocol was to exam-ine the feasibility of the FISH technique as aroutine screening procedure for A-bomb survivors(www.rerf.jp/Gene/eng/fish.htm).

When I visited LLNL in the fall of 1988, I wasstrongly convinced that the technique seemed highlyfeasible, as far as I could see from observing all pro-cedures with my own eyes. With a classicalmorphological analysis, I used to engage in analyz-ing stable aberrations in A-bomb survivors in atime-consuming process, but now it was possible todetect translocations more objectively and accuratelythan with the classic conventional technique I hademployed. When those irradiated people who wereexposed to ionizing radiation decades before exami-nation are studied, it is my belief that stableaberrations are the most reliable cytogenetic marker.I thought that the dreams could come true, and wemay be able to get rid of the minority.

A preliminary FISH study was conducted in com-bination with G-band analysis on 22 cases of

Microscopy Room in Cytogenetics Laboratory(Hiroshima) [1988]

http://www.rerf.jp/Gene/eng/fish.htm



Hiroshima A-bomb survivors. The results satisfac-torily showed that, as had been anticipated, there wasgood agreement of the data between FISH and G-band measurements. The FISH method was thusvalidated, and further indicated the utility of translo-cation frequency analysis for assessment of the levelof acute exposure to radiation.19 Based on the resultof the preliminary FISH study, a new screeningproject was proposed to examine AHS participantsto further analyze the relationship between genomictranslocation frequency and DS86 radiation dose forHiroshima and Nagasaki survivors.20

Now this document regarding the AHS cytoge-netic study is close to the end. Before closing thischapter, I have some questions addressed to myself.Where do we stand now with a large body of cytoge-netic data? How have the results of our data beenevaluated now by others? Finally, have we done thingsin the right direction? It is still premature and thusdifficult to evaluate our work by ourselves. Part ofthe answers may be in the evidence described below.

Our data suggests that a better-fitted dose-responserelationship can be obtained when restricting to thosesurvivors who were in individual (Japanese-style)houses at the time of explosion. It is these survivorswhose estimated doses seem to be the most reliable.Furthermore, the difference in the shape of dose-re-sponse curves between Hiroshima and Nagasaki isthereby reduced to a certain extent.

Recently, Nakamura et al collected teeth obtainedfrom Hiroshima survivors for whom cytogeneticanalysis had already been completed, and estimatedgamma-ray doses received by 69 Hiroshima survi-vors using electron spin resonance (ESR) of toothenamel. The resulting measurement was comparedwith the corresponding stable aberration data. For40 donors examined by both methods, there was thesame pattern of the dose-response between ESR-es-timated gamma dose and aberration data.21

Evidence accumulated to date has shown a pos-sible future study for survivors who were in thefactories at the time of bombing in Nagasaki. Theywere relatively proximally exposed, but owing tocomplex shielding situations their physically esti-mated dose is either unavailable or inaccurate evenif an estimate was attempted. Since they were ex-posed to appreciable radiation, and cancer riskestimation is limited by the numbers exposed at suchlevels, further study of the Nagasaki factory workersis urgently needed.

At present, no one knows the role and function ofcytogenetically abnormal lymphocytes in the humanbody. It seems to me that the presence of lympho-cytes with chromosome aberrations may have nodirect bearing on the health of survivors. One of thepotential future objectives would be to elucidate theclonal development of impaired lymphocytes at thesite of the lymphoid stem cells, and determine theirgenetic and immunological implications.

– End of Part One –

Notes to the readers: * The following is the definition of the terms that

are used frequently in this article. The terms“chromosome study” and “cytogenetic study”have been used synonymously and indiscrimi-nately among the cytogeneticists. The study onthe structure, function, and role of the chromo-somes is called “cytogenetics.” The term“karyotype” is applied to “a systematized arrayof the chromosomes of a single cell prepared ei-ther by drawing or by photography, with extensionin meaning that the chromosomes of a single cellcan typify the chromosomes of an individual oreven a species.”22

** The term “rad” is an old radiation unit. A newunit is “gray (Gy)” (1 Gy = 100 rad).

References1. Watson JD, Crick FHC: Molecular structure of deoxypentose nucleic acids. Nature 171:738–9, 1953

2. Tjio JH, Levan A: The chromosome number in man. Hereditas 42:1–6, 1956

3. Moorhead PS, et al.: Chromosome preparations of leukocytes cultured from human peripheral blood. Exp CellRes 20:613–6, 1960

4. Rothfels KH, Siminovitch L: An air-drying technique for flattening chromosomes in mammalian cells grown invitro. Stain Technol 33:73–7, 1962

5. Tough IM, et al.: X-ray-induced chromosome damage in man. Lancet ii:849–51, 1960

6. Buckton KE, et al.: A study of the chromosome damage persisting after X-ray therapy for ankylosing spondylitis.Lancet ii:676–82, 1962

7. Schull WJ: Effects of Atomic Radiation. John Wiley & Sons, Inc., New York, 1995

8. Slavin RE, et al.: A cytogenetic study of Down’s syndrome in Hiroshima and Nagasaki. Jpn J Hum Genet 12:17–28, 1967

9. Bloom AD, et al.: Cytogenetic investigation of survivors of the atomic bombings of Hiroshima and Nagasaki.Lancet ii:672–4, 1966



10. Bloom AD, et al.: Chromosome aberrations in leucocytes of older survivors of the atomic bombings of Hiroshimaand Nagasaki. Lancet ii:802–5, 1967

11. Sasaki MS, Miyata H: Biological dosimetry in atomic bomb survivors. Nature 220:1189–92, 1968

12. Awa AA, et al.: Chromosome-aberration frequency in cultured blood-cells in relation to radiation dose of A-bomb survivors. Lancet ii:903–5, 1971

13. Awa AA: Cytogenetic and oncogenic effects of the ionizing radiations of the atomic bombs. In: German JL (ed).Chromosomes and Cancer. John Wiley & Sons, Inc., New York, 1974, pp. 637–74

14. Awa AA, et al.: Relationship between the radiation dose and chromosome aberrations in atomic bomb survivorsof Hiroshima and Nagasaki. J Radiat Res (Tokyo) 19:126–40, 1978

15. Crow JF, et al.: Report of the committee for scientific review of ABCC, February 1975. ABCC Technical Report21-75, 1975

16. Preston DL, et al.: Comparison of the dose-response relationships for chromosome aberration frequencies betweenthe T65D and DS86 dosimetries. RERF Technical Report 7-88, 1988

17. Ohtaki K: G-banding analysis of radiation-induced chromosome damage in lymphocytes of Hiroshima A-bombsurvivors. Jpn J Hum Genet 37:245–62, 1992

18. Awa AA, et al.: Pilot study to examine a new fluorescence in situ hybridization-based method for chromosomeaberration frequency analysis and a new method for glycophorin A variant erythrocyte frequency analysis todetermine the utilities of these methods in assessing genetic damage in A-bomb survivors. RERF ResearchProtocol 10-89, 1989 [Editor’s note: Superseded by Research Protocol 8-93.]

19. Lucas JN, et al.: Rapid translocation frequency analysis in humans decades after exposure to ionizing radiation.Int J Radiat Biol 62:53–63, 1992

20. Kodama Y, et al.: Cytogenetic study in the Adult Health Study population by fluorescence in situ hybridization(FISH). RERF Research Protocol 8-93, 1993 [Editor’s note: Supersedes Research Protocol 10-89.]

21. Nakamura N, et al.: A close correlation between electron spin resonance (ESR) dosimetry from tooth enamel andcytogenetic dosimetry from lymphocytes of Hiroshima atomic-bomb survivors. Int J Radiat Biol 73:619–27,1998

22. International System for Human Cytogenetic Nomenclature (ISCN): Report of the Standing Committee on HumanCytogenetic Nomenclature. Cytogenet Cell Genet 21:309–404, 1978

��


Longevity of A-bomb Survivors

Radiation Risk, Longevity, and the Impact of theComparison Group on Low-Dose Risk Estimates

John B. Cologne and Dale L. PrestonDepartment of Statistics

Note: This article is based on several recent RERF publications, including: “Longevity of atomic-bomb survivors” (Cologne and Preston [2000]1), a modified version of this paper published in Japanese(Cologne et al [2001]2), and “Impact of comparison group on cohort dose response regression: An ex-ample using risk estimation in atomic-bomb survivors” (Cologne and Preston [2001]3).

In recent years articles appearing in the popular newspapers and magazines, such as The Washington Post,The New York Times, and Time magazine, have contained claims that atomic-bomb survivors are outliving theirunexposed peers. The statements in these articles were based on some reports such as those of Mine et al (1990)4

that report increased life expectancy for male survivors in Nagasaki who received doses between 0.5 and 1.5 Gy,and Hayakawa et al (1989)5 who report that all-cause death mortality rates for survivors who were more than 1km from the hypocenter were lower for than those for non-exposed residents of Hiroshima Prefecture. Kondo(1993)6 presents a summary of these and other findings (including data from RERF reports) to support hisargument for reduced mortality (increased life expectancy) at low doses. On the other hand, RERF’s analyses ofmortality in the Life Span Study provide clear evidence for dose-related increases in mortality rates from bothcancer, Pierce et al (1996),7 and non-cancer diseases, Shimizu et al (1999).8 In view of these contrasting re-sults—apparent radiation-associated increases in both cancer and non-cancer mortality and reports of greaterlongevity in some dose groups—we undertook to analyze recent Life Span Study mortality data (follow-upthrough 1994) in order to estimate longevity and to investigate the impact of the choice of a zero-dose compari-son group on estimates of risk and longevity.

Increased mortality and life shortening caused byradiation

Table 1 presents our estimates of the risk for totalmortality and longevity for the Life Span Study (LSS)cohort members in various dose groups. Risks fortotal mortality are measured relative to that for co-hort members with a total shielded kerma estimateof 0 Gy. Longevity, or median life expectancy, is de-

Table 1. Life expectancy by radiation dosea

Dose range (Gy)

Mean dose (Gy)

No. of people

No. of deaths

Relative risk

Median age at death

0 (

��



40 60 80 1000.0

0.2

0.4

0.6

0.8

1.0

0 Gy (

��



Table 2. Standardized mortality ratio (SMR) for geographically distinct zero-dose groups in the Life Span Study

Group People SMR

Estimate 95% confidence interval

Survivors (within 10 km of the hypocenter ATBa )

within 3 km 8,532 0.95 0.92, 0.98

3−5 km 18,352 1.01 0.99, 1.03

5−7 km 5,188 1.03 0.995, 1.07

7−10 km 1,992 1.07 1.00, 1.13

Local residents absent from city ATB

Not in city 26,531 0.95 0.93, 0.97

aAt the time of the bombs

mortality dose response, solid cancers are likely tocontribute a greater proportion of the total life lost atlower doses. Thus, cancer is the major cause, but notthe sole cause, of radiation related life shortening inthe LSS. While the LSS results in Table 1 clearlyindicate radiation-associated increases in risk (anddecreases in lifespan) at doses in excess of 0.25 Gythey also indicate that mortality risk and longevitychanges associated with lower doses are small. Theseresults do not support claims that radiation-exposedatomic-bomb survivors are living longer than theirpeers.

Risk estimates and the choice of comparison groupBecause the changes in risk associated with low-

dose exposures are small and estimation of themrequires the use of a comparison group, it is difficultto develop accurate estimates of these risks. The idealcomparison group should be comparable to the ex-posed cohort in every way except for the radiationexposure. However, detection of small changes inrisks is complicated by the fact that, for reasons otherthan the radiation exposure, rates in seemingly simi-lar populations can easily differ by as much or morethan the likely low-dose radiation effects. It may bepossible to measure and make statistical adjustmentsfor some of the factors that cause these differences,but even when some such adjustments are feasible(as is the case, for example, with age and sex) theycannot completely eliminate differences between theexposed and unexposed populations. If the unexposedgroup is not comparable to the exposed group, riskestimates, particularly those at low doses, will be dis-torted (biased).

The LSS includes almost 94,000 survivors whowere within 10 km of the hypocenter at the time ofthe bomb and a group of a little more than 26,000people who were temporarily away from the cities atthe time of the bombs. This latter group is referred toas the not-in-city (NIC) group. In order to reduce thelikelihood of bias, the cohort was constructed to in-sure that the age and sex distribution of distalsurvivors and the NIC group is similar to that for

survivors in the cohort who received significant ra-diation exposures (i.e., more than 5 mGy). As hasbeen noted in many RERF reports, despite this match-ing, cancer and non-cancer death rates for the NICportion of the cohort are lower than those for survi-vors whose estimated dose is less than 5 mGy.Because of this little-understood difference in rates,the NIC group has not been used in most recent analy-ses of LSS mortality and cancer incidence. It is alsouseful to consider the degree of variability in deathrates seen for survivors in the LSS cohort who re-ceived little or no radiation exposure from the bomb.This “zero” dose group includes the 25,532 LSS co-hort members who were between 3 and 10 km fromthe hypocenter at the time of the bombs and 8,532 ofthe 68,137 cohort members who were within 3 km ofthe hypocenters.

In order to investigate the degree of heterogene-ity in the risks for the zero dose group we dividedthis group into four parts based on distance from thehypocenter and estimated the standardized mortalityratios (SMR) for these subgroups relative to mortal-ity rates of for all zero-dose in-city survivors withadjustment for city, gender, age, and birth cohort. Theresults are shown in Table 2; an SMR of 1 meansthat the rates do not differ from the average rate forall zero-dose cohort members who were within 10km of the hypocenter at the time of the bombs.

These estimates suggest that background deathrates vary by about 10% within the zero-dose survi-vor group used for most RERF analyses. They alsosuggest that the SMR tends to increase with distancefrom the hypocenter. (A test for this trend indicatesthat it is statistically significant with P < 0.001.) Varia-tion in mortality rates with distance in the zero-dosesurvivor group could be due to geographic differ-ences in lifestyle, socioeconomic status, regionaldifferences in health care, and/or occupation. In 1945,areas that were more than 3 km from the hypocenterwere more rural than the urban central city areas onwhich the bombs were dropped and residents of thoseareas were generally poorer than city residents. The

��



Table 3. Effect of comparison group on estimated excess relative risk (ERR) of mortality

Comparison group

ERR per Gy

Deviation from positive-dose-only

ERR (%)

P-value for non-linearity

Positive dose only a 0.212 — 0.11

0−3 km 0.224 5.7 0.22

3−10 km 0.196 −7.5 0.025

0−10 km 0.206 −2.8 0.059

a The intercept was estimated by the dose-response regression fit to the non-zero-dose persons; the zero-dose groups were fit by separate parameters.

low rates for proximal zero-dose survivors (whichwe will take as survivors who were within 3 km ofthe hypocenter at the time of the bombs) may alsopartially reflect a healthy-survivor selection effect.Although more work on selection is necessary, re-sults from analyses of LSS non-cancer mortality byShimizu et al (1999),8 suggest that there is evidencefor such selection for noncancer deaths, but that im-pact of this selection had largely disappeared by thelate 1960’s.

It is also noteworthy that the SMR for the NICgroup (for whom no selection effects would be ex-pected) is significantly less than 1 and quite similarto that seen for the proximal zero-dose survivors. Thereasons for this are unclear. However, it is useful toconsider how the NIC group was selected. The pri-mary selection was based on data obtained duringspecial daytime censuses of people in Hiroshima orNagasaki cities on October 1, 1950 and, forHiroshima, supplementary censuses carried out onOctober 1 of 1951 and 1952. The unexposed groupwas defined to include people identified in these sur-veys whose family registry (honseki) was inHiroshima or Nagasaki but who were not in or nearthe cities at the time of the bombs. These surveyswere more likely to identify people who lived in orfairly near the city centers than people who lived inthe more rural outlying areas. Thus, one might ex-pect members of the NIC group to be more like theproximal-exposed survivors than some of the distally-exposed survivors. The LSS mail survey data pro-vide some evidence to support such a view.

It is of interest to consider the impact of variabilityin the death rates among the zero-dose survivors onradiation risk estimates obtained from the LSS. To dothis we examined how the choice of the zero-dosecomparison group affects both the linear dose-responseslope estimate and the evidence for dose-response non-linearity. Table 3 summarizes our findings (presentedin more detail in Cologne and Preston [2001]3). Forthe analyses summarized in the first row of this table,all information about baseline risks is derived fromsurvivors with non-zero (positive) dose estimates, us-ing a dose-response model in which the baseline riskis estimated by the model parameters with dose equal

to zero. In the other analyses the 0–3 km zero-dosegroup, the 3–10 km zero-dose group, or all zero-dosesurvivors were used to determine the level of thebaseline risks.

These results indicate that the choice of the com-parison group has a relatively small effect on theestimate of, or inference about, the slope in a lineardose response model. Neither did we find that thechoice of comparison group affected inferences ongender effects or age-time patterns in the excess risks.However, the choice of the comparison group has amarked impact on inference about the shape of thedose response and, hence, low-dose risk estimatesfor total mortality. The results presented in Pierceand Preston (2000)9 also suggest that choice of thecomparison group has some impact on inferenceabout low dose risks and the shape of the dose re-sponse. Analyses based solely on proximal survivors(positive dose only or all proximal survivors) pro-vide no suggestion of significant non-linearity in thedose response, while in analyses in which the moredistal survivors were allowed to contribute to the de-termination of the baseline rate level, there wereindications of a lack of linearity, specifically signifi-cant upward curvature in the dose response as can beseen in Figure 3.

The present work reveals that small biases in therisk estimate can result from the choice of zero-dosecomparison groups in analyzing atomic-bomb survi-vor data, but demonstrates that primary results fromthe LSS do not depend to a large extent on the tradi-tional comparison group. However, because of thevariability in death rates seen for various groups ofzero-dose survivors in the LSS, it is questionablewhether it is appropriate to include persons beyond3 km in the analyses, when one is concerned with thenature of the dose response relationship over the lowdose range (e.g., 5 to 200 mGy). Because of the largenumber of cohort members with very low doses (e.g.,5 to 10 mGy), background mortality rates can be pre-cisely estimated from the data for proximal survivorsor even from survivors with positive dose estimates.Our results suggest that detailed analyses of low-doseeffects should focus on the proximally-exposed(within 3 km) individuals only.

��



0 1 2 3 4

1.0

1.2

1.4

1.6

Proximal zero-dose group Distal zero-dose group Combined zero-dose groups

Rel

ativ

e r

isk

Dose (Gy)

ConclusionsThe LSS cohort is well-suited for addressing issues

of relative risk, loss of life, and low-dose radiationeffects for several reasons. It is a well-defined cohortwith virtually complete follow-up, doses are wellcharacterized, and there are substantial numbers ofpeople exposed to low doses. Because of thesefeatures, internal standardization of the risk regressionestimates is possible, so that an external estimate ofbackground rates is not required. Furthermore,follow-up began five years after exposure, whichwould eliminate mortality due to acute radiationeffects and other bomb-related trauma but not mostdelayed radiation effects on mortality (except for asmall number of early leukemia deaths).

Variation with dose in the relative risk of deathfrom all causes and changes in lifespan are two waysof describing and quantifying the same phenom-enon—the effect of exposure on the populationage-specific rate of death. An increase in one mea-sure implies a decrease in the other; when the relativerisk increases, the age-specific mortality rate becomeshigher than the background rate, and so some peoplewill die at a younger age as a result of their exposure.Reasons for favoring the use of relative risks or ex-cess rates to summarize radiation risk include theirdirect relationship to impact of radiation exposureon health and the relative ease with which they canbe computed and adjusted for confounding or modi-fying factors. However, relative risks and excess ratescannot really be understood without rather extensiveknowledge and understanding of background mor-

tality rates. Loss-of-life is a seemingly simple, rela-tively intuitive, general summary that has somemeaning to people outside of the fields of statisticsand epidemiology. Despite its intuitive appeal, lossof life is not an adequate summary of radiation ef-fects in a population since, among other things, itdoes not provide any information on the proportionof the population that are actually affected by a givenexposure, it fails to indicate when these exposure-associated cases might be expected to occur, and it isnot very useful when one needs to focus on the im-pact of an exposure on specific outcomes (e.g., solidcancer).

Using total mortality and avoiding the use of anyparticular dose-response model by using dose groups,we found compelling evidence that radiation expo-sure has increased total mortality rates (and henceshortened lives) for atomic-bomb survivors withdoses in excess of 0.25 Gy and some indications, al-beit not statistically significant, that there is slightlife shortening at lower doses. There is no evidenceof increased longevity in any dose range in the LSS.We suspect that others’ claims of greater longevityin certain dose groups may partly reflect biases re-sulting from the choice of comparison group. At leastfor total mortality in the L

Date post:	31-Jan-2021
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times

Table of Contents - RERF · 2017. 8. 22. · RERF Update Volume 13, Issue 1, Spring 2002 RERF News...

Documents