907R03090
 RADIATION
 RISK  ASSESSMENT
              WORKSHOP PROCEEDINGS
                           November 5 - 7, 2001
                           Las Vegas, Nevada
CO-SPONSORED BITS
 U.S. Environmental Protection Agency
           xvEPA

 Japan Atomic Energy Research Institute

-------
THIS PABE INTENTIONALLY LEFT BLANK

-------
     TABLE CDF  CONTENTS
     AGENDA
     PARTICIPANTS                                                 5

     INTRODUCTION                                                7

     RADIOBIOLOGY SESSION                                       9
     Background                                                      9
     Papers from Radiobiology Session                                     9
       Antone L Brooks                                               10
       Shin Saigusa                                                  18
       Charles R. Geard                                               23
       Rick Jostes                                                    28
       Miroslav Pinak                                                 30
       Mary Helen Barcellos-Hoff                                        41
       Ritsuko Watanabe and Kimiaki Saito (presented by Miroslav Pinak)         48
       Lowell Ralston                                                 56
     CURRENT ISSUES IN DOSIMETRY SESSION                     65
     Background                                                     65
     Papers from Dosimetry Session                                      65
       Evan B. Douple                                                66
       Fumiaki Takahashi and Yasuhiro Yamaguchi                          71
       Yukio Sakamoto, Shuichi Tsuda, Osamu Sato, Nobuaki Yoshizawa and
       Yasuhiro Yamaguchi                                            79
       Keith F. Eckerman and Akira Endo                                  88
       Yukio Sakamoto and Yasuhiro Yamaguchi                            94
       Kaoru Sato, Hiroshi Noguchi, Kimiaki Saito, Y. Emoto and S. Koga         102
       Keith F. Eckerman                                             111
       Sakae Kinase, Maria Zankl, Jun Kuwabara, Kaoru Sato, Hiroshi
       Noguchi, Jun Funabiki and Kimiaki Saito                             118
                 U.S. Environmental Protection Agency
                 Region 5, Library (PL-12J)
                 77 West Jackson Boulevard, 12th Floor
                 Chicago. IL  60604-3590                                           &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
DEVELOPMENTS IN RADIATION RISK ASSESSMENT SESSION   129
Background                                                      129
Papers from Radiation Risk Assessment Session                         129
  David J. Pawel, R. W. Leggett, K. F. Eckerman and C. B. Nelson           130
  Teruyuki Nakayama and Shohei Kato                                140
  Akira Endo, Yasuhiro Yamaguchi and Fumiaki Takahashi                 151
  Michael Boyd and Keith Eckerman                                   157
CURRENT ISSUES IN RISK MANAGEMENT & RADIATION
PROTECTION POLICY SESSION                                161
Background                                                      161
Papers from Risk Management & Radiation Protection Policy Session         161
  Michael Boyd and Shohei Kato                                     162
  Neal Nelson                                                    165
  Akihiro Sakai and M. Okoshi                                        175
  Robert Meek                                                    187
  Scott Monroe                                                   190
  Hideo Kimura, Seiji Takeda, Mitsuhiro Kanno, and Naofumi Minase         194
                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     AGENDA

EPA/JAERI  WORKSHOP AGENDA
     NOVEMBER 5

     8:15          Registration/Coffee and Pastries Available in Skyview I Room
     9:00          Welcoming Remarks - Richard Hopper, Deputy Director, EPA's Radiation
                   and Indoor Environments Laboratory, Las Vegas and Shohei Kato, Deputy
                   Director, Department of Health Physics, JAERI

     9:20          Radiobiology Session (Part 1)
                   Keynote Address: Recent Findings from DOE-funded Research into the
                   Biological Effects of Exposure to Low Level Radiation - Antone Brooks,
                   University of Washington
                   JAERI Funded  Research  on Molecular and Cellular Mechanisms  of
                   Radiation  Induced  Cancer - Shin  Saigusa, Radiation  Risk Analysis
                   Laboratory, JAERI

     10:30 -10:50  Break
     10:50         The Application of Site-Specific Microbeam Irradiation  in Defining a
                   Bystander Effect - Charles Geard,  Center for Radiological Research,
                   Columbia University
                   BEIR VII Committee Update - Rick Jostes, Board on Radiation Effects
                   Research, National Academy of Sciences
                   Open Discussion
     12:00 - 1:30   Lunch (on your own)
     1:30          Radiobiology Session (Part 2)
                   Molecular Dynamics Simulation of Damaged DNA's and Repair Enzymes
                   - Miroslav Pinak, Radiation Risk Analysis Laboratory, JAERI
                   How Tissues Respond to Damage at the  Cellular Level: Radiation Effects
                   on  Cell-Cell Communication - Mary Helen Barcellos-Hoff, Lawrence
                   Berkeley National Laboratory
                   Monte Carlo Simulation of Initial Process of Radiation-Induced DNA
                   Damage - Presented  by  Miroslav Pinak,  JAERI, on behalf of Ritsuko
                   Wanatabe and Kimiaki Saito

     3:00 - 3:30    Break (refreshments provided)
     3:30          A Regulator's Perspective on Mechanistic  Approaches to the Study of
                   Radiation Oncogenesis and Risk Assessment - Lowell Ralston, Radiation
                   Protection Division, US EPA
     4:00 - 4:30    Session Wrap-Up: Open Discussion with All Presenters


                                                                                    &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
             NOVEMBER 6
             8:00          Coffee and Pastries, Skyview I
             8:30          Current Issues in Dosimetry Session
                           Review of the NAS Report, A Status of the Dosimetry for the Radiation
                           Effects Research Foundation  (DS86) (2001)  - Evan Douple, Director,
                           Board on Radiation Effects Research, National Academy of Sciences
                           Conversion from Tooth Enamel Dose to Organ Doses for ESR Dosimetry -
                           Fumiaki Takahashi, External Dosimetry Laboratory, JAERI
                           Dose  Conversion Coefficients for High-Energy Radiations - Yukio
                           Sakamoto, External Dosimetry Laboratory, JAERI
                           Review of Work Related to ORNL's Collaboration with JAERI - Keith
                           Eckerman, Oak Ridge National Laboratory
             10:00 -10:30  Break
             10:30         Shielding Calculation Parameters for Effective Dose Evaluation - Yukio
                           Sakamoto, JAERI
                           Development  of CT Voxel Phantoms for Japanese  - Hiroshi Noguchi,
                           Head, Internal Dosimetry Laboratory, JAERI
                           Current  ICRP Committee  2 Issues (Weighting Factors, New GI model,
                           etc.) - Keith Eckerman, ORNL
                           Evaluation of Specific  Absorbed Fractions in Voxel  Phantoms Using
                           Monte Carlo Simulation - Hiroshi Noguchi (JAERI)
             12:00 -1:30   Lunch (on your own)
             1:30          Developments in Radiation Risk Assessment Session
                           An Uncertainty Analysis of EPA's Current Cancer Risk Coefficients -
                           David Pawel, Radiation Protection Division, US EPA
                           Effects of Baseline on Uncertainty of Radiation Risk Models - Shohei
                           Kato, JAERI
                           Detailed Dose Assessment for the Two Heavily Exposed Workers in the
                           Tokai-mura Criticality Accident - Fumiaki Takahashi, JAERI
                           Open Discussion
             3:00 - 3:30    Break (refreshments provided)
             3:30          An Overview of the  Methodology  Used  to  Develop  Cancer  Risk
                           Coefficients in Federal Guidance Report No.  13 - Michael Boyd (EPA)
                           and Keith Eckerman (ORNL)
             4:30 - 5:00    Wrap-up of Day 2
SEPA
        2                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      NOVEMBER 7

      8:00          Coffee and Pastries, Skyview I
      8:30 -10:00   Special Session on Current Issues in Risk Management and Radiation
                   Protection Policy
                   Update on  the  ICRP's Proposed Changes to the System of Radiation
                   Protection - Shohei Kato (JAERI) and Michael Boyd (EPA)

                   Unfinished  Business: Assessing Genetic and Fetal Risks - Neal Nelson,
                   Radiation Protection Division, EPA
                   Derivation  of Clearance Levels for Solid Materials in Japan - Akihiro
                   Sakai, Department of Decommissioning and Waste Management, JAERI
                   Developing a Technical Basis for Release of Solid Materials - Robert
                   Meek,  U.S.  NRC

      10:00 -10:30  Break
      10:30         A Status Report on Recent Activities Related to the  WIPP and Yucca
                   Mountain Projects - Scott Monroe (EPA)
                   Safety  Analyses for Shallow-Land Disposal of Alpha-Bearing Wastes -
                   Hideo Kimura, Department of Fuel Cycle Safety Research, JAERI

      11:30 -12:30  Wrap-up Discussion and Adjourn Formal Meeting
      12:30 - 2:00   Lunch

      2:00          Optional Tours  of EPA's Radiation  and Indoor Environments National
                   Laboratory and DOE's Yucca Mountain Visitors Center (Las Vegas)
                                                                                    oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 3

-------
           THIS PABE INTENTIONALLY LEFT BLANK
ww~    4                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     PARTICIPANTS
     LIST OF WORKSHOP PARTICIPANTS

     The following people attended the Workshop.  They are listed below along with contact
     information:
     ANTONEBROOKS
     Washington State University Tri-Cities
     tbrooks@tricity. wsn. edu

     CHARLES BEARD
     Center for Radiological Research,
     Columbia University, New York, N. Y.
     crg4@columbia. edu

     MlROSLAV PlNAK
     Japan Atomic Energy Research Institute
     pinak@ism\vs001.tokai.jaeri.go.jpe

     MARY HELEN BARCELLOS-HOFF
     Lawrence Berkley National Laboratory
     (LBNL)
     mhbarcellos-hoff@lbl. gov

     SHOHEI KATO
     Japan Atomic Energy Research Institute
     shkato@popsvr. tokai.go.jp

     FUMIAKI TAKAHASHI
     Japan Atomic Energy Research Institute
     taka@frs. tokai.jaeri.go.jp

     YUKJO SAKAMOTO
     Japan Atomic Energy Research Institute
     Mitsubishi Research Institute, Inc.

     HlROSHI NOBUCHI
     Japan Atomic Energy Research Institute

     LOWELL RALSTON
     Environmental Protection Agency -
     Radiation Protection Division
     ralston. lowell@epa. gov

     EVAN DOUPLE
     National Academy of Sciences
     edouple@nas. edu
KEITH ECKERMAN
Oak Ridge National Laboratory (ORNL)
eckermankf@ornl.gov

DAVID PAWEL
Environmental Protection Agency -
Radiation Protection Division
pawel. david@epa. gov

MICHAEL BOYD
Environmental Protection Agency -
Radiation Protection Division
boyd.mike@epa.gov

NEAL NELSON
Environmental Protection Agency -
Radiation Protection Division
nelson, neal@epa.gov

AKIHIRO SAKAI
Japan Atomic Energy Research Institute

ROBERT MECK
US Nuclear Regulatory Commission
ram2@nrc.gov

SCOTT MONROE
Environmental Protection Agency -
Radiation Protection Division
monroe.scott@epa.gov

HIDEO KIM LIRA
Japan Atomic Energy Research Institute
Department of Fuel Cycle Safety

RICK JOSTES
National Academy of Sciences
rjostes@nas. edu

SHIN SAIGUSA
Japan Atomic Energy Research Institute
Radiation Risk Analysis Laboratory
                                                                                 &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
             JIM BENETTI
             Environmental Protection Agency -
             Radiation and Indoor Environments
             National Laboratory
             benetti.james@epa.gov

             MARSHA SMITH, III
             Environmental Protection Agency -
             Radiation and Indoor Environments
             National Laboratory
             sm ithiii. martha@epa. gov
DICK HOPPER
Environmental Protection Agency -
Radiation and Indoor Environments
National Laboratory
hopper. r@epa. gov

PHIL NEWKIRK
Environmental Protection Agency -
Radiation Protection Division
ne\vkirk.philip@epa. gov
vvEPA
                                         RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      INTRDDUCTIDN
      For many years, the U.S. Environmental Protection Agency (EPA) and the Japan Atomic
      Energy Research Institute (JAERI)  have been exchanging radiation science information
      under the terms of a Memorandum of Understanding supported by both organizations.  In
      1989 and 1994, the agencies held jointly sponsored workshops on residual radioactivity and
      recycling.  The first workshop was held in the United States and the second was in Japan.
      The title of this third joint workshop, which was held in Las Vegas, Nevada, in November
      2001, is "Radiation Risk Assessment in the 21st Century."  The workshop explored recent
      scientific advances  that contribute  to  improved  human  health  risk  assessments for
      exposures to radionuclides at environmental levels.

      The three-day workshop was designed to increase understanding of the state of the science
      in both radiobiology and internal radiation dosimetry.  There was also a session devoted to
      exploring the challenging,  and sometimes contentious,  issues that policymakers are
      wrestling with in assessing  and managing radiation  risk and in developing criteria for
      protecting human health. Presenters included  radiation  experts from JAERI,  EPA, the
      Nuclear Regulatory Commission (NRC), the U.S Department of Energy, the University of
      Washington, Columbia University, the National Academy of Sciences, and Oak Ridge and
      Lawrence Berkeley National Laboratories.

      Risk  assessment and the science  behind it were the focus of the first  two days of the
      workshop. There were presentations on current cellular-level research into low dose effects
      being funded by the U.S. Department of Energy, computer simulation of radiation-induced
      DNA damage being conducted by JAERI, and current developments in biokinetics and the
      internal dosimetry models of the ICRP.  As highlighted through these presentations, key
      questions that will need to be addressed in the,new century are:
         >  What is the dose-response relationship at low doses of radiation exposure?
         >•  How do laboratory observations in vitro compare to what actually is happening in a
             complex organism in vivo?
         >•  How can we improve internal dosimetry models and better account for radiation
             dose distribution as a function of age, gender, and body type?
         >  How do we account for radiosensitive subpopulations?

      The last day of the workshop was devoted to current events in the areas of risk management
      and radiation protection, including the follow-up to the licensing and certification efforts at
      the Waste Isolation Pilot Plant (WIPP) and recent activities regarding the Yucca Mountain
      regulations.  The  workshop concluded with  a tour of EPA's  Radiation and Indoor
      Environments Laboratory in Las  Vegas.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                  7

-------
            THIS fABE INTENTIONALLY LEFT BLANK
4>EPA
        8                            RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS

-------
     RADIOBIDLQGY  SESSION

BACKGROUND
     The Radiobiology Session was the largest session of the workshop with eight presentations.
     Distinguished radiobiologists and other radiation researchers shared the latest research in
     low-level radiation does effects, molecular and cellular mechanisms of radiation induced
     cancers, and microbeam irradiation / bystander effects.  The research included topics in
     molecular dynamics of damaged DNA and repair enzymes, tissue response  to cellular
     damage, and Monte Carlo simulation of DNA damage. Additional presentations included a
     regulator's perspective to the mechanistic approach to risk assessment and an update on
     BEIR VII activities.

PAPERS FROM RADioBiaLOBY SESSION
     To follow are the papers written by the following conference presenters:
        >  Antone Brooks
        >  Shin Saigusa
        >  Charles Geard
        >  Rick Jostes
        >  Miroslav Pinak
        >-  Mary Helen Barcellos-Hoff
        >  Ritsuko Wanatabe and Kimiaki Saito (presented by Miroslav Pinak)
        >  Lowell Ralston
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     RECENT FINDINGS FROM DOE - FUNDED RESEARCH ON THE
 BIOLOGICAL EFFECTS OF EXPOSURE TO Low LEVELS OF RADIATION

 ANTON E L. BROOKS
     Washington State University Tri-Cities

 ABSTRACT

     This paper provides a brief review of the U.S. Department of Energy's (DOE) Low Dose
     Radiation Research Program and highlights some of the scientific advances made in the
     program. It discusses the problems associated with estimating the cancer risk following
     exposure to low doses of ionizing radiation and indicates that the high background rate for
     both radiation dose and cancer incidence makes it impossible to estimate risk at levels of
     radiation that are of concern in radiation protection.  The DOE research program is then
     discussed as a new approach  to helping with risk estimates.   This paper  reviews new
     paradigm shifts that are the result of the research and may  have an impact on standards.
     These included: "Adaptive Response" versus "Additive or Synergistic Effects", the "Hit
     theory" versus  "Bystander Effects", the "Role  of "Mutations" versus "Gene Induction" in
     Cancer and the "Single Cell" versus  "Tissue"  responses.  The research on these areas is
     providing a strong scientific base for the setting of radiation standards that are  adequate and
     appropriate.

 INTRODUCTION

     The  DOE  Low  Dose  Radiation Research  Program addresses  the  old  problem  of
     determining health effects following exposure to low doses of ionizing radiation. The
     research in  the program is founded on extensive past scientific investigations triggered in
     part by the concern from fallout associated with nuclear weapons testing. As we know, the
     fallout was in and on everything,  and resulted  in low-level doses to both people and the
     environment. However, the question regarding fallout radiation exposure remains: Did the
     low doses from the fallout actually do anything that results in measurable health effects? If
     there were health effects from these low radiation doses, then it is very important for us to
     characterize them, since the methods of estimating the number of radiation-induced cancer
     at  low doses can be  rather large based on linear-no-threshold extrapolations.  Current
     standards are set using  this linear-no-threshold model rather than on real scientific data.
     Such  extrapolations suggest  that  one  particle or  ionization results  in  one  mutation,
     producing  one  cancer. Is the dose-response truly linear at  these low doses, or are there
     biological processes that result  in  sub-linear or even super-linear responses to these very
     low doses?  The DOE Low Dose Radiation Research Program addresses these  questions.

     PROBLEMS ASSOCIATED WITH Law DOSE RADIATION RISK ESTIMATES

     Two major problems make it hard to estimate cancer risks associated  with  low doses of
     ionizing radiation: the variable background exposure to ionizing radiation; and the high and
     variable background rate of cancer.

     It is a well-known fact that we get about 370 mrem per year from different environmental
     radiation sources. Radon is calculated  to be responsible  for about half of this exposure.
     Our background dose can change according to where we live.  The background level of
     radiation from  cosmic-ray exposures can double from 24 mrem/year  at sea level to  50
i o                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     mrem/year in high elevation cities like Denver, Albuquerque and Salt Lake City (NCRP
     1987). In other words, we can alter the dose or number of mrem we get by choosing where
     we live.

     The level of radiation exposure that can be produced by any single exposure site is subject
     to regulation.  For example, there is a discussion on whether to use 15 or 25 mrem as a
     clean-up standard for  waste disposal sites.  The difference between these two levels is
     trivial relative to increased health risk or the background radiation that we receive each
     year, but there is a very large cost associated with cleaning-up to the lower standard.  It is
     important to realize that environmental exposures are usually a small fraction of the natural
     background and, as illustrated above, you can move from one location to another with a
     higher elevation and change your radiation dose.  Many other  factors can also influence
     background radiation such as the level of radon in our homes.  Most of us aren't concerned
     about these low exposures relative to the location of our homes, but we are concerned about
     other changes that may impact background radiation. Multiple studies have tried to link
     cancer incidence  to  background radiation and have not demonstrated  an  association
     between them.

     The other variable that makes it difficult to detect  changes in risk following low doses of
     radiation is the high  and variable background rate of cancer.  Cancer frequency  in any
     population  is  related to  a large  number of variables  such  as genetic  background,
     environmental exposures,  cigarette smoke, diet,  and life  style. All of these variables
     influence cancer risk.  It is  of interest to evaluate the cancer frequency as a function of the
     geographical distribution of the population in the United States.  The National Institute of
     Health has evaluated cancer risk  as a function of the county  (Devesa et al.  1998).  They
     have broken  down in the  cancer rate into percentiles.  The cancer rate in the top 10
     percentile of the population is as high as 800/100,000 per year. The rate in the bottom 10%
     is only 90/100,000. Such data illustrate that there  is a huge variability in the cancer rates
     throughout the United States.  For example, there is a strip of high cancer rates that runs up
     the  lower  Mississippi River.  In  addition, several big cities  have  high cancer rates, in
     comparison to other places. It's  clear that this variability  in cancer rate is  not directly
     related to radiation because there are so many other factors that  go  into the equation.  The
     NIH has published a map for every county  site and for every cancer type. The cancer types
     are  further  broken down according to age,  sex,  and race.  In fact, there is an entire
     book/series of these maps (Devesa et  al. 1998).  These maps are interesting and illustrate
     some of the problems of conducting epidemiology  studies to  relate cancer  to radiation
     exposure.  It is critical that both the background radiation and the background cancer rate
     be considered when trying to detect small changes in cancer rate associated with a small
     change in radiation exposure.   Because of the limited sensitivity of epidemiology studies
     related to the discussion above,  it is important to try to determine if changes in health
     effects  are  present, even  if they  are not detectable  using standard toxicological or
     epidemiological approaches.

     THE OBJECTIVES or THE DOE Law DOSE RESEARCH PROGRAM

     The Low Dose Research Program, started almost four  years ago  by the  Department of
     Energy, was projected to last at least ten years with a funding level reaching $21 million
     per year.  Currently, 54 projects are funded annually in hopes of better understanding the
     basic biological  mechanisms that  occur  at  low  doses.   That way, standards  can be
     developed based on the best possible science.
                                                                                         &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                  1 i

-------
              Why was this research program  started?   In reality, research  funding in  the  field of
              radiation biology has been decreasing rather rapidly in the  United States and in other
              countries around the world.  At the same time that funding for radiation research was
              decreasing, there were major breakthroughs in the fields of genetics, molecular biology and
              cancer biology. In the past, every kind of exposure and response was measured in a wide
              range of biological systems.  Most of this work was conducted  following high levels of
              radiation exposure and the response predicted or extrapolated to the low dose region.  This
              was dictated because of the  lack of technology and biological techniques to detect changes
              after low-level exposures. Now, with these recent scientific breakthroughs, many of which
              were associated with the genome program, it is now possible  to measure the  responses at
              low doses delivered at minimum dose-rates.  This is something that was not possible in the
              past. It was thus important to apply new biology to the old radiation risk problem.

              Currently, changes in gene expression in thousands of genes can be measured at once. This
              makes it possible to develop  fingerprints for  radiation exposure and to predict radiation
              dose.  The ability to  rapidly  sequence genes  opens many doors  for research on  genetic
              susceptibility that was only a dream a few years ago.

              Newly developed technology  is being merged with new biological techniques  in  this
              program. These  new  developments  require shifts in radiation related paradigms.  Four
              specific paradigm  shifts  resulted from the research conducted in this program. These
              paradigm shifts require a reevaluation of how radiation interacts with cells, which means
              starting from a different base to develop the  standards.  This paper reviews those shifts
              which may have an impact on standards.
                  >   "Adaptive Response" versus  "Additive or Synergistic Effects"
                  >   "Hit theory" versus  "Bystander Effects"
                  >   Role of "Mutations" versus "Gene Induction" in Cancer
                  >   "Single Cell" versus "Tissue" responses.

               'ADAPTIVE RESPONSE* VERSUS "ADDITIVE OR BYNERBIBTIC EFFECTS'

              The original results on the adaptive responses were seen for chromosome damage in human
              blood lymphocytes.   If blood lymphocytes were given a small priming dose of radiation
              (1.0-2.0 cGy) before a big challenge dose (200 cGy), the frequency of chromosome damage
              was less than would be  observed in response to the large  challenge dose given alone
              (Olivieri et al. 1984). The adaptive response has been observed in many systems over the
              past several years, making one think that it is an important biological process.  However,
              the adaptive response was not produced in all  systems and there was individual variability
              in the response.  Some individuals responded and some did not. The adaptive response has
              been further demonstrated in whole animals systems where the adaptive dose decreases the
              risk for cancer induced by the large challenge dose (Mitchell et al. 1999).  Additional
              research is being conducted to determine if this response is present at low doses and dose-
              rates.

-------
      Another form of adaptive response,  which deserves evaluation,  is the change in the
      background  cancer  or cell transformation  frequency following exposures to very low
      radiation doses.  Recent studies reported the influence of small  doses of radiation on the
      background  frequency of cell transformation.  This research has provided ample data
      following exposures below 10 cGy. By using a delayed plating protocol, it was possible to
      demonstrate that the frequency of cell transformation following exposure of cells to low
      doses of radiation is  actually lower than the background cell  transformation  frequency
      (Redpath et al.  2001).  These  observations  create important questions that must be
      scientifically addressed. These include:
         >•  What genes  are being activated during the adaptive response?
         >  What processors are being initiated that may alter post-translational processing of
             proteins and alter protein function?
         >  What are the mechanisms involved in low dose and adaptive responses?

      Research in the Low Dose Program is currently underway to address these questions.

      "HIT THEORY* VERBUB 'BYSTANDER EFFEBTB*

      Another paradigm shift is  associated with "hit" theory and the bystander effects.  In the
      past, cells were thought to become damaged by the direct interaction or hitting of the cells
      by the radiation and  the deposition of energy in  the damaged cell.   It has now been
      demonstrated that cells do not have to be "hit" by the radiation to express radiation-related
      changes.  The development of several microbeams makes it possible to target  individual
      cells (Nelson et al. 1996). These targeted cells can be "hit" with very well defined numbers
      of alpha particles, protons, or even with a focused x-ray.  This research has demonstrated
      that a variety of different repair, cell cycle control or apoptosis genes can be activated in
      "non-hit" bystander cells.   It  is also possible to  produce chromosome  aberration and
      mutation in cells that are not targeted.  The big question now is to determine the impact of
      bystander cells and whole  organ or tissue  response on the  risk  or impact  of radiation
      exposure.

      The  bystander paradigm has been modeled and used to compare the  response of cells
      directly hit and cells that respond but are not hit.  This model was  termed the Bystander and
      Direct Hit (BaD) model (Brenner et al. 2001).  The BaD model assumes that all bystander
      effects are detrimental, that the effects  seen in cells can be added and that these cellular
      effects reflect cancer risk.  The emphasis of the model was on the induction of mutations
      and chromosome aberrations that are produced  in non-hit cells.  The total response was
      modeled as detrimental and promotes the formation of cancer and other diseases  especially
      following exposure to high LET radiation.  However, there are other effects  produced in
      bystander cells that  must also be considered.  The change in  gene expression in non-hit
      cells  is one of the main observations that must be considered in risk assessment (Azzam et
      al. 1998). Many of these changes may be protective and eliminate damaged DNA or cells
      from the population.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                  i 3

-------
              ROLE OF "MUTATIONS* VERSUS "BENE INDUCTION* IN CANCER

              The relative role of gene  mutation versus gene expression in the development of late
              effects, especially cancer, leads to the next paradigm shift.  In the past, everything in the
              environment was considered a mutagen and thus  a  carcinogen.   There  was  a serious
              concern about all the compounds and mixtures in the environment that were found to result
              in mutations in bacterial tester strains.  It was predicted that all exposures that produced
              mutations would also be responsible for the production of cancer.  Research that helped
              understand the role of mutagens and anti-mutagens in the environment helped put the role
              of environmentally  induced mutations and cancer in better prospective. It is of interest to
              evaluate the characteristics of mutations and determine if these characteristics are essential
              parts of the cancer process. Radiation-induced mutations result in a wide spectrum of rare
              events that can be passed from one cell generation to the next.  These events seem to be so
              rare  that it is  difficult  to  assign them the  primary  role  in either  cancer or  cell
              transformation, both of which are  rather frequent events.  However, radiation-induced
              gene-induction is a very frequent event induced in many cells in the population by very low
              radiation  doses.  In  most cases, the  activation or change in gene  expression is very
              transient.   Occasionally, the changes in gene expression result in an alter  phenotype that
              may not be transient.  Such observations make it essential to evaluate the changes in gene
              expression as a function of very low radiation doses and dose-rates.

              Such research has determined that there are many changes in gene expression induced at
              both  high and low radiation doses  (Amundson et al.  1999). The important part of these
              studies is the recognition that there is a different set of genes turned on or off at  high does
              than those genes that are altered by low radiation doses.  Thus, if a completely different
              gene set is turned on as a function of radiation dose, it is very difficult  to make linear
              extrapolations between the risks and response from high and low radiation doses. At very
              low doses,  there are  a large numbers  of genes both up and  down regulated.  The next
              challenge is to repeat such studies and determine the site of action  and  the phenotypic
              changes induced by these gene changes and to determine the significance of these changes
              during  radiation-induced cancer.   With these two types  of genetic  alterations,  mutations
              and changes in gene expression, it is possible to suggest that changes in gene expression
              can result in exactly the same set of phenotypic changes as induced by gene mutation. In
              the final  analysis,  it may become  possible  to determine the role  of changes in gene
              expression and mutation on phenotypic  changes such as altered cell proliferation, program
              cell death or apoptosis, and cancer. Such research may  suggest that cancer is  related to
              general physiological changes and regulatory  changes  as the critical steps  of  cancer
              induction and the mode of action for carcinogens are further characterized.

               "SINGLE CELL" VERSUS "TISSUE* RESPONSES

              In the past, it was traditional to treat each cell as responding dependency to a mutagen or
              carcinogen.  Tissues  were thought of as a bag of cells  with each  individual cell acting
              independently. If a mutation was produced in a cell, then that mutated cell did its thing and
              was  capable of producing cancer. However, current research suggests  that  extensive
              cell/cell and cell/matrix communication exists and that different cell types and the extra-
              cellular matrix all influence the response of the tissues to any environmental stress.  It is
              necessary to start  thinking about the response of cells to radiation in terms  of total cell
              biology with the responses being made at the tissue level rather than at the level of the
              individual cell.  In turn, these tissues interact to influence the response of the whole animal.


&EPA
          14                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      This level of organization is  what is important  when setting our radiation protection
      standards.  Current research illustrates the contrast between the current mutation theories,
      in which cells act independently, and the tissue theory where cells interact and normal cells
      prevent mutated cells from expressing the abnormal  phenotype associated with cancer
      (Barcellos-Hoff and Brooks 2001).  The current single cell theory results in linear kinetics
      with a single radiation hit producing  a single mutation and the cell with that  mutation
      producing  a cancer.  In contrast, using the tissue theory, cells that are irradiated respond
      with a range of changes hi gene expression.  Then the cells interact to help regulate the
      expression  of a  changed phenotype.   At large radiation  doses,  the  inter-cellular
      communication and the communication between the cells and the matrix can be altered to
      allow changes in gene expression to remain for long periods of time.  Because  of these
      changes, some of the cells do not down-regulate their gene expression, this results in a cell
      that acts like a mutated cell. The cells are expressing genes and proteins they shouldn't, can
      change phenotype, and through these changes can progress to form cancer in the absence of
      any mutagenic event.

      What we do know is that the number of cancers that are radiation-related are small and that
      these  are produced by both direct and bystander effects.   The  interaction between the
      negative and positive effects in cells that are not radiated is an important  research area
      where additional work is needed.  With new techniques and tools it is possible to get down
      into the dose range of interest and start looking at the responses to very low doses to help
      understand how radiation interacts with cells to produce disease.

      The gene expression paradigm  instead  of a mutation theory is just beginning to be studied
      and understood. The technology is moving fast, making it possible for scientists to custom
      make the gene chip microarrays  for any sets of genes. The genes  of importance can be
      identified and evaluated using  sequencing technology,  and the frequency and type of the
      polymorphic changes can be studied. Now one of the major challenges is to find methods to
      evaluate the large data sets produced. The scientists are using informatic approaches to sort
      out whether any of the changes in gene expression mean anything in terms  of risk or the
      development of disease. The ability to generate large amounts of data very rapidly leads to
      additional  problems in trying to interpret/analyze the data.  It will be  difficult to  take the
      large volumes of information being generated and move it  from the cell molecular to the
      tissue level onto the whole animal and from the whole animal studies to predict radiation
      related risk in humans. The DOE Low Dose Program is addressing this challenge.

  CONBLUBIONB

      In the past, much of the cellular and molecular data was developed in vitro using model cell
      systems that are of little use for extrapolation to the level of biological organization needed
      for standard setting. The current  research needs to expand on such data and supplement it
      with more  appropriate cell systems that can provide useful information for risk assessment.
      Research needs to  start pushing back toward using  more meaningful cells  and tissue
      systems and insist on using modern techniques in complex biological systems. Once these
      techniques are developed, moved from in vitro tissue culture systems  on to whole tissue
      and experimental animals and finally into humans, then the results can start influencing the
      setting of radiation standards.   With such data, it will become possible to use information
      on individual variability and sensitivity in risk assessment.

      Historically, research was  limited to few cell types  and many  of these were fibroblasts,
      which were  genetically  altered and  well on their way toward   becoming  cancer.

                                                                                          &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 1 5  . .4^>.?~
                                                                                           X^MfjfB^

-------
     Scientifically,  one  can make a  strong argument for selecting  a system that is well
     characterized and from which one can get dose-response relationships. As a result of this
     type of thinking, many inappropriate systems continue to survive  in the research world.
     However, it is well known that normal  epithelial cells are the cell types of concern and
     must be the major focus of future research. These cells live in a complex environment that
     is sending them messages and controlling their growth and differentiation. Future research
     needs to consider all of these factors.

     A need exists to develop transgenic animal models, which make it possible to know exactly
     which genes have  changed.   And  in  fact, DOE has  a whole  program  funding the
     development of useful animal models. The problem is that there is only a limited amount
     of money. The decision was made early on that DOE would not fund epidemiology studies
     or long-term dose-response animal  studies.   However, DOE  is  interested  in funding
     mechanistic studies  hi  whole animals,  which  attempt to understand how the cellular
     changes develop into cancers in the "normal" environment of the intact animal.

     The information generated by the DOE Low Dose Radiation Research Program may or
     may not result in changes  in radiation standards based on cell molecular biology. But, as
     the research develops it will form a strong scientific base for extrapolation from effects at
     high levels of exposure to those predicted to occur after low levels of radiation exposure.
     This will generate some very nice tools to be used to understand the  biological responses to
     very low radiation exposures. Since radiation risk from low dose radiation exposure cannot
     be predicted in epidemiology studies, it is necessary to develop these mechanistic data.

     By combining advanced technologies with advances in cell-molecular biology, it's possible
     to detect changes at very low doses. These observed changes have required basic paradigm
     shifts in the way we think  about radiation. It is not as important if this Low Dose Research
     Program alters  standards  as it is  that the research provides a scientific basis that  will
     influence the way we  think about how radiation interacts with cells, tissues  and  whole
     organisms. As we understand the role of these basic radiation-induced biological changes,
     it will be possible to better understand radiation-induced cancer  risk.  Such understanding
     will help ensure that the standards  are adequate and appropriate. This research can also help
     the public better understand  that the standards being setting are  based on the best science
     available.

     The DOE Low Dose Radiation Research Program focuses on developing a scientific basis
     for  radiation standards.   They have done  this through  development  and  use of new
     technology and genomic biology and applying these methods to the old problem to estimate
     the  health effects of low  doses of radiation.  This approach has  been  successful  at the
     cellular and molecular level and is currently being applied to whole tissues and animals.
     With  further development, it will be possible to use the information developed in this
     program to impact the basic radiation biological paradigms discussed in this paper. Finally,
     the knowledge associated with radiation-induced  cancer and disease will provide increased
     public understanding of the basis on which the standards are being developed.
16                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  REFERENBEB
  Amundson, S.A., K.T. Do, A.J. Fomace, Jr. (1999) Induction of Stress Genes by Low Doses of
      Gamma Rays, Radiation. Res. 152, 225-231.
  Azzam, E.I.; deToledo, S.M.;  Gooding, T.; Little, J.B. (1998) Intercellular communication is
      involved in the bystander regulation of gene expression in human cells exposed to very low
      fluences of alpha particles. Radiation Res. 150, 497-504.
  Barcellos-Hoff,  M.H.,  Brooks,  A.L.  (2001)  "Extra   cellular  Signaling  through   the
      Microenvironment: A Hypothesis Relating Carcinogenesis, Bystander Effects and Genomic
      Instability", Radiation, Res. 156, 618-627.

  Brenner, DJ. Little, J.B.,  Sachs, R.K. (2001) The bystander effect in radiation ontogenesis, II.
      A quantitative model.  Radiation Res. 155, 402-408.

  Devesa, S.S., Grauman, D.J., Blot, W.J., Pennello, G.A., Hoover, R.N., Fraumeni, J.F. (1999)
      "Atlas of Cancer Mortality in the United States", National Institutes of Health, National
      Cancer Institute, NIH Publication No. 99-4564.
  Mitchell, R.E.J., Jackson, J.S., McCann, R.A., and  Boreham, D.R. (1999) The adaptive
      response modifies latency for  radiation-induced myeloid  leukemia in CBA/H   Mice.
      Radiation  Res.  152,  273-279.  NCRP  National  Council on  Radiation  Protection  and
      Measurements, 1987.  Exposure of the population in the United States and  Canada from
      natural background radiation. NCRP Report No 94. Issued December 30, 1987.  Bethesda,
      Maryland. National Council on Radiation Protection and Measurements
  Nelson, J.M., A.L. Brooks, N.F. Metting, M.A. Khan, R.L. Buschbom, A. Duncan,  R. Miick,
      L.A. Brady (1996) Clastogenic effects of defined numbers of 3.2 MeV alpha particles on
      individual CHO-K1 cells, Radiation Res. 145, 568-574.

  Olivieri, G., J.  Bodycote and S. Wolff (1984) Adaptive response of human lymphocytes to low
      concentrations of radioactive thymidine, Science, 223, 594-597.
  Redpath, J.L., Liang, D. Taylor,  T.H., Christie, C., Elmore,  E. (2001) "The shape of the dose-
      response curve for radiation-induced neoplastic transformation in vitro. Evidence for an
      adaptive response against neoplastic transformation at  low doses of low-LET  radiation.
      Radiation. Res. 156, 700-707.
                                                                                       &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS                                17

-------
              JAERI FUNDED RESEARCH  ON MOLECULAR AND CELLULAR
                       MECHANISMS OF RADIATION INDUCED CANCER

           SHIN SAIQUSA
              Radiation Risk Analysis Laboratory
              Japan Atomic Energy Research Institute

           INTRODUCTION

              The radiation effects research in Japan started after the Second World War.  In  1947, the
              U.S. National Academy of Sciences, under the Atomic Energy Commission, established the
              Atomic  Bomb Casualty  Commission (ABCC) at Hiroshima  and Nagasaki to  study the
              biomedical effects of atomic bomb survivors in cooperation with the National Institute of
              Health of the Ministry of Health and Welfare of Japan.  In 1954, fishermen on the  boat
              "Lucky-Dragon," were exposed to radioactive fallout from a hydrogen bomb test in the
              Pacific Ocean.  Around this time, radioactive fallout was detected throughout Japan. These
              incidents motivated the Science Council of Japan to initiate scientific research on atomic
              radiation in Japan.  The  Japanese government decided to explore atomic energy research
              and radiation science and they took the following actions:  It started with the establishment
              of Japan Atomic Energy Research Institute  (JAERI) in 1956 with the aim of exploring
              applications  of atomic energies. In 1957, the National Institute of Radiological Sciences
              (NIRS)  was established  for studying radiological sciences, including radiation, physics,
              radiation biology, radiation medicine, etc.  In 1959. the Japan Radiation Research Society
              (JRRS)  was  organized and played a central role in the promotion of radiation research in
              Japan.  Furthermore, in 1961. the Japan Health Physics Society (JHPS) was initiated and
              contributed to  the development of radiation protection in Japan.   Since  this time, the
              International Congress of Radiation Research (ICRR) has been held continuously nearly
              every four years. The sixth congress was organized in Tokyo in 1969.

              In 1975, the Radiation  Effects Research  Foundation (RERF) succeeded  the ABCC to
              continue the study of the survivors.  The RERF has been jointly operated by the Japanese
              Ministry of Health and Welfare and the U.S. National Academy of Sciences. In order to
              promote education and  research  on radiation science, the Ministry built four research
              institutes and centers and also established 18  departments in national universities by 1976.
              All these institutions were devoted to  the study of various aspects of radiology, health
              physics,  radiation physics, radiation protection,  nuclear medicine, and radio  oncology.
              However, many of these departments are now renamed or reorganized into the different
              departments, which are mainly  associated with cancer research or bio-informatics or other
              similar sciences.

              From the standpoint of the practice of reasonable radiation protection, the past two decades
              of research efforts have contributed to the knowledge and recognition of low dose radiation
              effects.  In 1990, the Institute for Environmental Sciences (IES) was established.  In 2000,
              the Low Dose Radiation Research Center (LDRRC) of the Central Research Institute of
              Electric Power  Industry  was established.  Both institutes are concerned with low  dose
              animal studies and long-term animal studies, but their essence  is completely different.  The
              former is studying dose and dose rate effects on radiation induced cancer, but the latter is
              studying the effects of low dose radiation to the chemically induced cancer.
oEPA
  JXL   ia                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  ONBOINB STUDIES ON LOW DOSE RADIATION EFFECTS

     Table 1 shows the ongoing studies on low dose radiation effects in Japan.

                                       TABLE 1 :
            ONGOING STUDIES ON  Law DOSE RADIATION  EFFECTS IN JAPAN
IES, LDRRC
HRF
REA
RERF
JAERI
Experimental study of mice whole body exposure.
Epidemiological study in high background area in China.
Epidemiological study of nuclear industry workers in Japan.
Epidemiological study of Chernobyl accident survivors.
Funding low dose radiation effects researches in Japan.
Database for radiation exposure and safety assessment.
Research on radiation risk analysis.
     The Institute of Environmental Sciences (IES) is carrying out a low dose and low dose rate
     carcinogenesis study (1), which is planned to terminate in 2003. A total of 4,000 B6C3F1
     mice were irradiated with gamma-rays and the life span study had been proceeded. At the
     end of September 2001, 3,970 mice of 4,000 mice examined had already died and 30 mice
     remained. The pathological diagnosis is also carried out for all organs of each individual
     mouse  and  almost  1,700  mice have been done  at present1.  The Health  Research
     Foundation (HRF) has been cooperating with High Background Radiation Research Group
     of China (HBRRG), promoting the epidemiological study in high background area (HBGA)
     in  China during the past  decade (2).   The  Radiation  Effects Association (REA) is
     performing a series  of the  epidemiological study of nuclear industry workers, which is
     consignment research from Ministry of Science and Technology in Japan. The first analysis
     was carried out from 1990 to 1994 and the second was carried out from 1995 to 1999 (3, 4).
     The RERF is well known with the series of the life span studies of atomic bomb survivors
     (5, 6) but is also associated with epidemiological studies of Chernobyl accident survivors.

     JAERI has been funding the Japanese low dose radiation effects researches  since 1989 to
     obtain useful information  and data as a basis of scientific radiation risk estimates and
     reasonable radiation protection. JAERI also started to construct and maintain the database
     for radiation exposure and safety assessment  from 1994 and the database, "DRESA," was
     opened to  the  public  in  May 2001.   This  database  is  mainly used  for the risk
     communication purpose and for research on radiation risk analysis. Furthermore, in 1999,
     JAERI started its own radiation risk analysis research, which will be further described later.

  •JAERI FUNDED RESEARCH

     The JAERI  research funding  program has been put into operation in three  stages.  The
     research program of the first stage, 1989 to 1993, was mainly focused on the accumulation
     of cancer and hereditary risk  data.   The research subjects in these five years were; a)
     Expression of mechanisms of radiogenic cancer and  its  variation factors, b) Radiation
     induced  and endogenously induced damage to DNA and c) Development  of detection
     technique of radiation induced  gene and chromosome mutation.   Table  2-shows the
     research products derived from the research programs of the first period.  The third and
i
 Experiment was terminated in 2002.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                    l 9

-------
oEPA
                fourth research products shown in Table 2 are dealing with hereditary effects.  Research on
                the radiation induced hereditary effect is one of the differences between the DOE's Low
                Dose Radiation Research Program and JAERI's program.

                                                   TABLE  2:
               RESEARCH PRODUCTS DERIVED FROM THE  FIRST RESEARCH  PROGRAM
                                                (1 9S9-T 993)
                                1)   Experimental data on the biological effectiveness of low-
                                    energy photons and neutrons in human peripheral blood
                                    lymphocytes.
                                2)  Analysis of genes responsible for susceptibility to radiation
                                    lymphomagenesis.
                                3)  Spontaneous and induced chromosome aberrations in human
                                    spermatozoa.
                                4)  Risk evaluation of low dose radiations by the observation of
                                    congenital malformations.
                The research program  of the second  stage, 1994 to 1998, was focused on analysis of
                radiogenic somatic effects and hereditary effects from genome to  individual level.   The
                research subjects  in  these  five years  were; d)  Initial process of radiogenic cancer, e)
                Radiogenic genome instability, f) Radiogenic oncogenesis related genes and g) Radiation
                effect on the germ cell.  Table 3 shows the research products derived from the research
                programs of the second period.
                                                    TABLE 3:
                    RESEARCH PRODUCTS DERIVED  FROM THE FIRST RESEARCH PRDBRAM
                                                 ( 1 99-4- 1 99B)
                                1)   Mechanisms of induction of radiogenic genetic instability in
                                    lymphoblast cell.
                                2)  Detection and isolation of genes responsible for susceptibility
                                    to radiation thymic Lymphoma in mice.
                                3)  Gene alterations in human liver tumor associated with
                                    Thorotrast injection.
                                4)  Gene mutation related with hereditary effects in Medaka fish,
                                    Oryzias latipes.
                The  research  program of the third stage,  the ongoing stage,  is focused on  molecular
                analysis  of radiogenic somatic  effects and hereditary  effects  from the genome to the
                individual level, based on the latest molecular biology techniques. The research subjects of
                these ongoing years are; h)  Radiation response and signal transduction, i) Mechanism of
                radiogenic   somatic mutation,  j)  Gene  regulation  of  radiation  effect  expression,  k)
                Radiogenic hereditary effects and 1) Mathematical modeling of radiation biological effects.
                The fifth subject about modeling is not an experimental study, but is quite significant and
                indispensable  to correlate the fundamental biology  and radiation risk  estimation.  The
                research programs of this period have been still continuing.
                                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  TYPICAL RESULT OF THE PROBRAM
      One  of the  interesting studies  funded was the genetic  analysis of Thorotrast induced
      cholangiocellular carcinoma by Dr. Manabu Fukumoto   (7).  Thorotrast is the colloidal
      suspension  of radioactive  232ThO2 that emits  alpha  particle,  which  was  used  as  a
      radiographic contrast agent in the  1930's - 1950's.  The study has been carried out to
      analyze genetic  alterations  occurring in human  liver  tumors induced  by  Thorotrast
      injections.   Since  Thorotrast was  deposited in the  internal organs (mainly  liver) and
      continuously emitting the alpha particles, Thorotrast exposure have  been considered as a
      carcinogenesis model of human chronic exposure by  alpha particles. The  materials used
      were archival tissue sections of Thorotrast  induced tumors, cholangiocellular carcinoma,
      from 22 patients more  than 30 years  after  injection.   DNA was extracted from paraffin
      embedded tissue and PCR were performed.  Inductions of K-ras in exon 12 and p-53
      mutations, as well as chromosome aberrations and 8-hydroxy-guanine were examined.

      In the tumor samples from 22 patients, a total of 3 mutations in K-ras were detected, only
      from tumor sites. All these mutations were transition type G to A. Furthermore, a total of
      eight mutations in p-53 were detected in exon 6, 7, 8 but not in exon 5  (no hot-spot was
      detected).   These eight mutations  were detected from six patients  and six  of the  eight
      mutations were same type, A to G transition, which seems to be  Thorotrast specific.  Four
      of the six mutations were detected in the tumor site  and two  in the non-tumor site,
      suggesting that the p-53 mutation is not  tumor specific.  Observations of  pathologic
      diagnosis suggest that the p53 mutations mainly occurred in undifferentiated carcinomas
      rather than differentiated carcinomas. Results of chromosome aberrations induction and 8-
      hydroxy guanine detection showed no relationship with 220Rn concentration in breath.

  FUNDINB THE PROBRAM

      Table 4-1 shows the amounts of funding for all stages of research. A total of $3.8 million
      dollars was funded  for  13  years. Table  4-2  shows  the  participating  scientists  and
      laboratories during these periods. A total 35  scientists from 33 laboratories were involved.
                     TABLE 4-1
                                                          TABLE 4-2
               TOTAL AMOUNT OF THE FUNDING
                 1989-1993    $1,500,000
                 1994-1998    $ 1,500,000
                1999-2001
$ 800,000
                7bfa/oft3yrs. $3,800,000
                         PARTICIPATING SCIENTISTS
                             (LABORATORIES)
                           1989-1993    11(10)
                           1994-1998    11(10)
1999-2001
13(13)
                           Total of 13 yrs. 35(33)
           4-1 Seiryo-machi, Aobaku, Sendai, 980-8575 JAPAN
' Dr. Manabu Fukumoto: Institute of Development, Aging and Cancer (IDAC), Tohoku University
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                          oEPA
                                                                                      21

-------
 RESEARCH ACTIVITIES OF RADIATION RISK ANALYSIS LABORATORY, JA.ERI

     The Radiation Risk Analysis Laboratory, Department of Health Physics, JAERI, is mainly
     promoting computer  simulation analysis rather than the "wet experiments" to investigate
     the  radiation-induced initial damages from molecular to cellular  levels,  as a basis of
     fundamental steps of the radiation health risks.

     The research activity  involves  molecular dynamics simulation of the damaged DNA and
     repair enzymes, Monte Carlo  simulation of initial  process of radiation induced DNA's
     damage, and the mathematical model  study of radiation-induced cellular oncogenesis.
     Furthermore, the uncertainty analysis of radiation risk estimation is performed based on the
     vital statistics data of Japanese and U.S. populations.  A final goal of our studies  is to
     connect the  initial biological  damages and the  radiation risk  estimation in terms of
     computer simulation analysis.

 REFERENCES
 1.  S.  Tanaka, I II  B.  Tanaka, K.  Ichinohe, M. Saito, S. Matsushita, S.  Sasagawa, T.
     Matsumoto,  H. Otsu  and  F.  Sato.  Long-term Low-dose-rate  Continuous Gamma-ray
     Irradiation on  Mice  -Interim Report  2—.  P.  20 in: Proceedings of the International
     Symposium on Biological Effects of Low Dose Radiation, 2002.
 2.  High Levels of Natural Radiation 1996: Proceedings of the 4th International Conference on
     High Levels of Natural Radiation. (Eds. L. Wei, T.  Sugahara and Z. Tao) Elsevier, 1996.

 3.  Epidemiological Study Group of Nuclear Workers (Japan). First Analysis of Mortality of
     Nuclear Industry Workers in Japan, 1986-1992. Journal of Health Physics. 32: p. 173-184
     (1997).
 4.  S. Ohshima and M. Murata. Second Analysis of Mortality of Nuclear Industry Workers in
     Japan, 1986-1997. Journal of Health Physics. 36: p. 141-147 (2001), in Japanese.
 5.  D.A. Pierce, Y. Shimizu, D.L. Preston, M. Vaeth and K. Mabuchi. Studies of the Mortality
     of Atomic Bomb Survivors. Report  12, Part I. Cancer. 1950-1990. Radiation Research.
     146: p. 1-27(1996).
 6.  Y. Shimizu, D.A. Pierce, D.L. Preston and K. Mabuchi. Studies of the Mortality of Atomic
     Bomb Survivors. Report 12, Part II. Noncancer Mortality. 1950-1990. Radiation Research.
     152: p. 374-389(1999).
 7.  H. Momoi, H.  Okabe, T. Kamikawa, S. Satoh, I. Ikai, M. Yamamoto, A. Nakagawara, Y.
     Shimahara, Y. Yamaoka and M. Fukumoto. Comprehensive allelotyping of humans.
22                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                 MICRDBEAM MEDIATED CELLULAR EFFECTS

  CHARLES R. GEARD
      Center for Radiological Research, Columbia University, New York, N. Y.

      At the Radiological Research Accelerator Facility (RARAF), a charged particle microbeam
      was developed based on a Van de Graaff Accelerator.

      Interest in microbeams stems from the fact that a microbeam allows the  irradiation of
      individual cells with  an exact number of particles, including, and very importantly, one
      particle. The majority of individual cells in the majority of individuals never receive more
      than one  alpha particle in their lifetime, in contrast to the situation in miners, where a
      significant number of cells may receive more than one particle.

      That was the initial rationale for the development of the microbeam - to provide a tool to
      examine this particular problem.  It also allows for the irradiation of some cells while
      missing others, which relates to the bystander effect - a major area of interest over the last
      few years.

      Many cells can be irradiated in a highly localized spatial region, the nucleus of a cell where
      the DNA is based, the cytoplasm of the cell,  or miss the cell completely and irradiate
      intercellular  space.  The irradiation can be  controlled such  that a particle through the
      nucleus of a cell at time zero, can wait ten seconds, a minute, ten minutes, or an hour before
      a second (or more) particle passes through the cell.  The microbeam is a tool for precisely
      targeting all or a fraction of a population of cells with a defined number of alpha particles,
      and this relates specifically to the bystander effect.

      What is the bystander effect? The bystander effect is seen when non-hit, or non-treated,
      cells show a biological response similar to hit, or treated, cells.

      Cells not  hit by radiation show a response  as if they had been hit by radiation. With the
      microbeam, we can then undertake several types of studies.  As previously  indicated, we
      can irradiate the medium between the cells, the cytoplasm of all cells, the nucleus of all
      cells, the  nucleus of a known fraction of cells,  and the cytoplasm of a known fraction of
      cells.

      Ranges of endpoints have been studied for the  bystander effect at the RARAF. We can
      look at delays in cell  cycle progression,  which in large part is a  consequence  of the
      expression of individual genes that have been switched on by the radiation in the hit cells or
      in the bystander cells.  In terms of chromosomal damage, we  can look at frequencies of
      micronuclei,  we can  look at sister chromatid  exchanges, we can look at mutation in
      particular genes, and we can look at oncogenic  transformation.  This range of end points
      was examined at RARAF in  studies using the microbeam to evaluate the bystander effect in
      a variety of cell types.

      This bystander response could originate in the medium, it could originate in cell cytoplasm
      or cell membranes, or it could originate in the cell nucleus. Given these possibilities, how
      then  do you devise a means of addressing  these questions?  It can  be argued  that  a
      microbeam is the optimal tool to undertake these  types of studies.
                                                                                         oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 23

-------
            One method of studying the bystander effect is the transfer of medium from irradiated cells
            to non-irradiated cells. This assumes the release into the medium of factors, which then
            influence the non-irradiated bystander cells. Another approach uses very low influences of
            alpha particles fhim isotopic sources where it is estimated \  at  only  a  relatively small
            fraction of cells have been hit and yet it is  determined  that many more cells  show a
            response than could possibly have been hit.

            Then there is the charged particle microbeam.  One step in cell studies using a microbeam
            is to record the position of the cells prior to irradiation.  When the cells are placed on the
            microscope and imaged, the position of all  cells is recorded, so that the position of all cells
            is known  - in this respect, it can be determined which cells have been hit with alpha
            particles and which cells have not and how near or far they are from the hit cells.

            Initially, we irradiated 100 percent of cells with precisely one particle,  precisely  two
            particles or more, and then moved to irradiating known fractions of cells within the range
            of two to fifty percent, and compared that with 1 00 percent. This was accomplished using
            an approach where a fraction of the  cells have a  nuclear dye,  which is seen by  the
            fluorescence microscope, and a fraction of cells have a cytoplasmic dye, which is not seen
            by the microscope. Only those cells with the nuclear dye are irradiated.  In other words, we
            irradiate cell populations at known proportions and are able to discriminate between them.

            The  question formulated was: In known  un-irradiated cells, what is the frequency of
            micronuclei?   Our expectation would be  that  the  frequency  of micronuclei  in non-
            irradiated  bystander cells  should be the same as that in the controls.  There  was clear
            evidence of particle number dependent micronuclei  in hit cells.   We  found an elevated
            incidence  of micronuclei in the known non-hit cells, which is considered proof positive of a
            bystander effect.

            Because we know exactly how many cells we started with, if you examine the cells  as a
            function of time post irradiation, you can look at cell growth, in other words, the percentage
            increase in cell number.  Cell  growth declines as the number of particles passing through
            the cells increases, but the same is also true in the non-hit cells, i.e. non-hit bystander cells
            are slowed in progression through the cell cycle.

            We can look at these  cells in situ and ask about gene  expression at the  protein level using
            immunofluorescence.  A question raised is, 'what is the expression of p-53 and of p-2 1 in
            the hit  cells versus the non-hit cells?'  P-21  is  a very important protein regulating the
            movement of cells through the  cell cycle.

            The  expression of p-2 1 is enhanced in the bystander cells, as was  the expression  of p-53,
            but not to the same extent as  in the hit cells.  At the gene expression level,  there is an
            elevation in these particular proteins.

            Further, following irradiation  with  the microbeam,  the dish of cells can be placed on a
            microscope adjacent to the microbeam microscope and individual cells  removed with a
            micromanipulator. RNA for a particular gene can then be quantified from single cells.

            One question to answer is 'What is the expression ofp-21 at the RNA level?'  With  time
            post single cell microbeam irradiation, there is a rapid increase in the p-21 gene product in
            individual cells, with up to a ten-fold increase in the RNA level in the cells irradiated with
            10 alpha particles. There is however, a dramatic difference between individual cells. Also
&!.-    24                                  RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      in the majority of cells, at three hours post-irradiation, the  synthesis of RNA for  this
      particular gene has declined dramatically from that at an hour to three hours.  However, in
      some cells, you have the same level of response as applied at an hour, therefore the hit cells
      do not exhibit a uniform response. When looking at a population of cells, one is looking at
      an overall response.

      In bystander cells there is an equally variable but only 2-3-fold increase in the p-21 gene
      product.  That is, known non-hit bystander cells respond similarly to hit cells at both
      molecular and cellular levels.

      Another subject area is mutation. Using a cell line derived from a Chinese hamster,  which
      has  a human chromosome  4 present in it, mutations on this human chromosome  are
      assessed. Both precise nuclear and cytoplasmic irradiations have been undertaken.

      Cells are imaged and the particles are placed through the cytoplasm deliberately missing
      the nucleus.  The paradigm accepted for many years is based on the notion that you have to
      hit DNA with ionizing radiation in order to initiate a  deleterious consequence, hi  this
      experiment, we specifically ask the question, does  cytoplasmic irradiation alone increase
      the incidence of mutation? Based on our results, the simple answer is yes.

      For  survival, the effect is slight, but there is some  effect, hi cells irradiated with a very
      high number of alpha particles, essentially none will survive. How can a non-surviving cell
      - a dead cell - mutate?  It cannot. If mutations are witnessed in a population of cells  where
      only a fraction of the cells has been irradiated, then those mutations must come from the
      bystander  cells.  Results show  that when twenty percent of cells are  hit with 20 alpha
      particles, there is a dramatic increase in mutation over the control incidence.

      However, by treating the cells with the substance lindane, you reduce this effect.  Lindane
      prevents the  communication of signals from cell to  cell  - also called  intercellular
      communication - via gap junctions,   hi  other words,  the signal that was causing  the
      bystander  cells to mutate was  transferred from  the hit cell to the non-hit  cell through
      intercellular communication.

      In a recent study, 100% or fewer  cells were irradiated  with one  alpha  particle.   The
      incidence of mutation for 20 percent was the same as at 100 percent of cells being hit.

      The last area covered  is oncogenic transformation in  C3H  10  Tl/2 cells.   Initially,  a
      comparison was undertaken between an exact number of alpha particles through the  nuclei
      of cells and an average number of alpha particles delivered with track segment irradiation.
      A mean of one particle per cell nucleus was more effective than exactly one particle per
      nucleus.

      The argument  then  is that the  oncogenic transformation process requires two  or more
      particles, and the response seen is due to that fraction of cells. The number of particles per
      cell  is Poisson distributed, that is for a mean of one, about 1/3 of cells received two or more
      particles.

      When ten percent of cells were irradiated versus 100 percent of cells, looking at the fraction
      of cells killed, one particle kills about 15 percent of cells and 85 percent survive.
                                                                                          oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 25

-------
     The other question then is what happens for oncogenic transformation?  And  surprising
     though it may seem, the incidence of oncogenic transformation is very similar, whether 10
     percent of the cells are irradiated or 100 percent of the cells are irradiated.  This is a very
     similar finding to the results for mutation.

  SUMMARY

     Conclusions  from RARAF microbeam experiments  include the following.  Deliberately
     missing  cells and irradiating  culture medium between  them produces  no  detectable
     response. Irradiating cell cytoplasm produces no detectable increase in the frequency of
     micronuclei.  It does induce cell cycle delay and it does induce mutation.

     Irradiating cell nuclei, which, of course, includes cytoplasm and medium, results in fluence
     dependent  increases in cell  cycle  delays, increases in  gene  expression, mutation  and
     oncogenic transformation. This response specifically includes exactly one alpha particle
     with some caveats for one alpha particle for the end point oncogenic transformations.

     Irradiating a fraction of cells through the  nuclei produces a response in known non-hit
     bystander cells,  which is not dependent on the number of particles through the hit cells.
     Reducing the proportion of hit cell nuclei results in a proportional lessening of response in
     the bystander cells, but it is not a dramatic lessening of response.

     Irradiating hit cells through the cytoplasm does not produce a response in bystander cells.
     The expression of a bystander effect in non-hit cells  originates from damage to  the nuclei
     of hit cells, even if the fraction of hit cells is as low as five percent.

     Enhanced gene expression as determined by single cell RT-PCR is the most sensitive  end
     point in both hit and bystander cells.  There is dramatic intercell variability in  cells with
     known radiation histories for both single cell RT-PCR and  single cell immunofluorescence
     cytometry.

     The final conclusion is that the charged particle microbeam and single cell assays provide
     well controlled systems for evaluating cellular responses to  site specific damage.

  A.DKNO WLEDBEMENTB

     The National Institutes of Health, P41 RRl 1623, CA49062, CA75061 and the Department
     of Energy Low Dose Program, DE-FG02-98ER62667 supports the micro beam studies at
     the Center for Radiological Research.
26                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  PERTINENT COLUMBIA UNIVERSITY MICROBEAM BASED PUBLICATIONS
  Randers-Pehrson G., Geard, C.R., Johnson, G., Elliston, C.D., Brenner, D.J., The Columbia
     University single-ion microbeam. Radiat Res. 2001 156(2): 210-4.

  Zhou, H.,  Suzuki, M., Geard, C.R.,  Hei, T.K.  Effects of irradiated medium with or without
     cells on bystander cell responses.  Mutat Res. 20; 499-135-41, 2002.

  Sawant, S., Randers-Pehrson, G., Geard, C.R., Brenner, D.J., and Hall, E.J.  The bystander
     effect  in radiation oncogenesis: I. Transformation in C3H 10T1/2 cells in vitro can be
     initiated in the unirradiated neighbors of irradiated cells.  Radiat. Res. 155:397-401, 2001.
  Hei, T.K.,  Zhou, H.N., Wu, L.J., Randers-Pehrson, G., Waldren, C. and Geard, C.R.  Radiation
     induced  genotoxic  damage  in mammalian cells: from  cytoplasm  to  nucleus and  the
     bystander phenomenon. In:  Free Radicals in Chemistry, Biology and Medicine, (ed. T.
     Yoshikawa,  S. Toyokuni and Y. Yamamoto, OICA International, London) p.  241-247,
     2000.

  Miller, R.C., Randers-Pehrson, G., Geard,  C.R., Hall, E.J. and Brenner, D.J. The oncogenic
     transforming potential  of the passage  of single alpha particles through mammalian cell
     nuclei. Proc. Natl. Acad. Sci. USA 96(1): 19-22, 1999.
  Wu,  L.J.,  Randers-Pehrson, G., Xu, A.,  Waldren,  C.A.,  Geard, C.R., Yu, Z., Hei, T.K.
     Targeted cytoplasmic irradiation with  alpha particles induces mutations in mammalian
     cells. Proc. Natl. Acad. Sci. USA 96(9):4959-4964, 1999.
                                                                                      &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               2<7   
-------
                                 BEIR VII COMMITTEE UPDATE

           RICK JDSTES
              National Academy of Sciences

              The following is an update on BEIR VII, Phase II. Bier VII Phase I set out to determine
              whether there was enough information, since the last low dose, low LET study, BEIR V, to
              conduct a full-fledged study.  It was determined that, "yes," there is  ample information
              available to conduct a full-fledged study. Numerous publications, well  over a thousand of
              them, are available on this subject. Many of those might not be relevant, yet it is surprising
              the amount of material that is out there since the last low-dose study. The emphasis will be
              on published, peer-reviewed information.

              The BEIR  series of reports have been a major resource on the health  effects of ionizing
              radiation for more than 22 years.  BEIR I, III, V, and now VII are focused on low levels of
              the low LET radiation; the BEIR VII report is currently in progress.   BEIR IV and VI
              focused on high LET radiation.  BEIR II focused on cost benefit analysis.

              The purpose of the BEIR VII, Phase II study is to update the BEIR V study on the health
              effects of low dose, low LET ionizing radiation in light of all the new scientific information
              collected over the past decade.  Sponsors of the BEIR VII study are: The Department of
              Energy, The Department of Defense, the Environmental Protection Agency and the Nuclear
              Regulatory Commission.

              The committee has two subcommittees, Epidemiology and Biology:

              1) Richard Monson, an epidemiologist from Harvard  School of Public Health, is the chair
              of the full  committee  and  the  epidemiology subcommittee.   The  epidemiology
              subcommittee consists of Eula Bingham, oncologist; Pat Buffler, epidemiologist; Elisabeth
              Cardis, epidemiologist; Scott Davis, epidemiologist; Ethel Gilbert,  biostatistician, Albrecht
              Kelleher,   physicist;   Dan   Krewski,  epidemiologist;  Katherine   Rowan,   a  risk
              communications individual who has been very helpful in some of the outreaches and
              interactions that we've had with the public on this highly  visible study;  and Leonard
              Stefanski who was put on as an expert on linear and non-linear models from outside of the
              radiation community to give another perspective to the modeling.

              2) The biology subcommittee is chaired by Jim Cleaver, DNA repair; Roger Cox, who is
              familiar with cell  and  animal  radiation biology; Bill  Dewey, a cell radiation biologist;
              Tomas Lindahl, DNA repair expert; K. Sankaranara;- anan, an expert in  germ cell genetics;
              Bob Ulrich, an animal radiation biologist;  and Herb  Abrams, a radiologist from Stanford
              University.

              The statement of task's primary objective is to develop the best possible risk estimate for
              exposure to low dose, low LET  radiation in human subjects.  In order to do this, the
              committee will  conduct a comprehensive review of all relevant epidemiological data
              related to the risk from exposure to low dose, low LET radiation, and define and establish
              principles on which quantitative analyses  of low dose and low dose rate effects can be
              based, including requirements for epidemiological data and cohort characteristics.
&EPA
         2B                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      With respect to the biologic  data, the  committee will consider relevant biologic factors,
      such as the dose and dose rate effectiveness factor, relative biologic effectiveness, genomic
      instability, and adaptive response. It will assess the  current status and relevance to risk
      models of biologic data in models of carcinogenesis, including critical assessment of all
      data that might  affect the shape  of the dose response curve and consider any recent
      evidence regarding genetic effects not relating to cancer.

      With respect to the modeling,  the committee will:
          >   Develop appropriate risk models for all cancer sites and other outcomes for which
             there is adequate data to support a quantitative  estimate of risk, including benign
             disease and genetic effects.
          >   Provide examples of specific risk calculations based on the models and explain the
             appropriate use of the risk models.
          >   Describe and define the limitations and uncertainties of the risk models and their
             results.
          >•   Consider relevant biologic factors and appropriate methods to develop etiologic
             models favoring simple, as opposed to complex, models.
          >   Assess the current status and relevance to risk models of biologic data and models
             of carcinogen sis, including critical assessment of all data that might affect the
             shape of the response  curve at low doses.
          >   Discuss the role and effects of modifying factors, such as individual susceptibility
             and variability, age and sex, environment and life style factors, and
          >•   Identify critical gaps in knowledge that should be filled by future research.

      Sources of the information fall into some general categories:
          >   The biologic information,
          >   Hiroshima and Nagasaki  studies,
          >   Occupational radiation studies,
          >   Medical radiation studies, and
          >   Environmental radiation studies.

      The committee will also  consider accidents and, in particular, situations that resulted in
      high exposures to populations such as Chernobyl and Myak.

      BEIR VII  is a five-year study.  The study has been extended two years from the original
      three years in order to access modified dosimetry at the RERF.

      There will be a one meeting in 2002 when the first draft of the full report will be presented.
      Two meetings in 2003 will be scheduled to fine tune  and finalize the report. Estimated date
      of final report is October of 2003.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                  29

-------
       MOLECULAR DYNAMICS  SIMULATION OF DAMAGED DNA's
                            AND REPAIR ENZYMES

 M IRQ SLAV PlNAK
     Japan Atomic Energy Research Institute

 ABSTRACT

     The molecular dynamics (MD) studies of the several radiation originated lesions on the
     DNA  molecules are presented with respect to the proper recognition of the lesion by the
     respective repair  enzyme.  The pyrimidine lesions (cytosinyl  radical,  thymine dimer,
     thymine glycol) and purine lesion (8-oxoguanine) were subjected to the MD simulations for
     several hundred picoseconds (ps), (between 200 ps for cytosinyl radical and 2 nanosecond
     for 8-oxoguanine) using MD simulation code AMBER 5.0 (4.0) and its  respective force
     field modified for the lesion. The simulations were performed as all atoms simulations for
     fully solvated solute molecules in water.  The negative charges of DNA phosphates were
     neutralized by sodium counter ions NA+ that are essential for the double helical structure.

     In most cases, the significant structural changes  in the DNA double helical structure are
     observed:

       a)   the breaking of the hydrogen bonds network between complementary  bases and
           resulting opening of the double helix (cytosinyl radical, 8-oxoguanine);
       b)   the sharp bending of the DNA helix centered at the lesion site (thymine dimer,
           thymine glycol); and finally
       c)   the flipping-out adenine on the strand complementary to the lesion (8-oxoguanine).

     These changes are related to the overall collapsing double helical  structure around the
     lesion and may facilitate the docking of the repair enzyme into the DNA and formation of
     DNA-enzyme complex. The stable DNA-enzyme complex is a necessary condition for the
     onset  of the enzymatic repair process. In addition to the structural changes, the specific
     values of electrostatic interaction  energy are calculated at  several lesion sites (thymine
     dimer, thymine glycol and 8-oxoguanine). The specific  electrostatic energy (thymine
     dimer, thymine glycol) is considered a factor that enables the repair enzyme to discriminate
     the lesion from the native site.

     Keywords: Molecular dynamics, DNA lesions, repair enzymes, radiation risk

  INTRODUCTION

     In order for specific DNA transcription to occur, recognition and binding at specific sites
     on DNA by regulatory enzymes is  essential. In addition to specific  DNA transcription, the
     functioning  of repair  enzymes removing the damaged DNA parts  is very important to
     ensure correct cell proliferation  and to  eliminate  potential mutagenic  cells.  Several
     nucleotide sequences of specific DNA binding sites that are involved in gene transcription
     regulation have been described, suggesting that a  code for recognition between DNA
     regulatory and repair enzymes and DNA sites exists [1, 2, 3, 4].  Considerable information
     regarding enzyme/DNA interaction has been  gamed  from  biological  experiments. In
     several of these systems, both prokaryotic and eukaryotic, a DNA recognition alpha helix
     within the enzyme's DNA binding domain, has been observed [5, 6].  It  is known that
     sequence specific DNA binding by repair and regulatory enzymes occurs as  a result of
     multistage hydrogen bonding and van der Waals interactions between the DNA recognition
3O                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      amino acid chains of enzyme and nucleotide base sites of DNA. However, the underlying
      mechanisms  by which  enzymes recognize specific or damaged sites on  DNA are  the
      subject of debate [e.g.  7,  8]. This paper reports  the results  of several lesioned  DNA
      molecules examined with respect to the enzyme/DNA interactions between amino acids
      and nucleotides. The method used in the study was molecular dynamics (MD)  simulation.
      Where available, the simulation data are compared with experimental (X-ray and NMR
      crystallography) results. In the center of interest were radiation  damages, as are  8-
      oxoguanine,  thymine dimer, thymine  glycol  and  cytosinyl radical, and their  potential
      impact on the DNA structure. Particularly the  van der Waals and electrostatic interaction
      energies were  calculated.  These interactions  between  DNA  and  enzyme may induce
      breakage of  Watson-Crick nucleotide base pairing hydrogen bonds, further resulting in
      bending of the DNA, strand elongation and its unwinding. The formation of a stable DNA-
      enzyme complex that results from the onset of the repair process was studied as well.

  METHOD

      In our studies, the molecular dynamics technique is the main tool. Since a DNA molecule is
      not a rigid, static structure, the x-ray diffraction and NMR results usually show average
      structural parameters.   In  reality,  every DNA  molecule  is  under constant thermal
      fluctuations, which result in local twisting, stretching, bending and unwinding of the double
      helix. In this  aspect, the molecular dynamics, a simulation technique  that yields static and
      dynamic properties of a molecular system, may provide useful scientific data showing the
      DNA in its dynamical mode. The classical MD is based on solving Newton's equations of
      motion for each atom in the system. This way it is capable to simulate the behavior of a
      system consisting of N atoms. Solving of these  equations produces new atomic coordinates
      that can be used to calculate a new set of forces. Static and dynamic properties  of the
      system  are then obtained as a time  averages over the trajectory. For the simulations, the
      molecular dynamics program package AMBER 5.0 (AMBER 4.0  in  the case of the
      cytosinyl radical) was used  [9].

      The simulated  molecules are subjected to several hundred picoseconds (ps) up to  1-2
      nanoseconds  (ns) of MD simulation under molecular dynamics protocol consisting of the
      following sequential steps:
         1)  Preparation of solute molecule(s). Solute molecules are usually the non-damaged
             DNA segments having a certain part replaced by lesion (e.g. 8-oxoguanine). The
             structural and chemical parameters of the lesion must be defined prior to the
             insertion of the modified part into the solute molecule. These parameters, such as
             lengths of chemical bonds, angles and charges, may be taken from  existing
             experimental data where available, or for small molecules may be calculated by
             quantum chemical methods. The structure of modified solute molecules is then
             optimized using the program AMBER 5.0 in order to achieve stabile molecular
             configuration with minimal potential energy.
         2)  Locating the solute  molecule into the simulation cell.
         3)  Neutralization the negative charges of DNA phosphates by adding the sodium
             counterions at the initial positions bisecting the O-P-O angle at certain distance (~5
             A) from each phosphorus atom.
         4)  Solvation of the  solute molecules in the water (several thousand water molecules
             are usually used to solvate solute molecules).
         5)  Minimization of the potential energy of the system.
                                                                                        oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                31

-------
         6)   Heating up to a required temperature (e.g. 310K (36.85°C), human body
             temperature) during sequential MD runs.
         7)   Stabilization of the density of the system during constant pressure MD runs.
         8)   Production molecular dynamics with constant volume.

     COMPUTATIONAL. DETAILS

     Solvation of solute  molecule usually requires a large number of water molecules that
     increase the requirements on the capacity of RAM and CPU time. To be able to handle
     such large systems, the original AMBER 5.0 code was partly parallelized and then installed
     on a FUJITSU VPP5000 vector/parallel supercomputer using an auto-vectorizing compiler.
     Its sequential  and parallel flags were also changed in order to compile a program on the
     VPP5000 computer. After introducing  these changes and required resizing,  the  current
     program is capable of dealing with a system consisting of up to 100,000 atoms within a
     reasonable CPU time. Production MD simulations are performed on the Fujitsu VPP5000
     supercomputer or on the Hitachi SR8000  parallel supercomputer.  Preparatory steps as
     formation  of solute molecules,  minimization, heating  and density  stabilization  are
     performed on  the scalar workstations (SUN). Supercomputers that are used in simulations
     are at the Center for Computational Science and Engineering of the Japan Atomic Energy
     Research Institute. The samples of CPU simulation time required to accomplish 1 ps of MD
     are shown in Table 1.

     In  MD  simulation, the  constant  dielectric  functions are  used  and  1-4  electrostatic
     interactions (electrostatic interactions separated by only three bonds), which are scaled by
     factor  1.2 that is recommended value for AMBER 5.0 force field.  Particle Mesh Ewald
     Sum technique is used as in implemented in AMBER 5.0 [10]. In this method, a Gaussian
     charge  distribution of opposite sign is superimposed upon  the  original point charges,
     producing a screened charge distribution. The electrostatic interaction between the screened
     charges  is then short ranged. The original  distribution is recovered by  adding a second
     Gaussian charge distribution identical to the  first, but of opposite sign, hi the  calculation of
     electrostatic interactions no cut-off distance  is applied and thus all water molecules in the
     system were included. The van der Waals interactions are calculated within the defined cut-
     off distance (usually  10-12 A). Periodic Boundary Conditions are applied throughout the
     entire simulation.
                                        TABLE 1 :
    EXECUTION TIME REQUIRED TO  ACCOMPLISH THE  1  PS OF M D  SIMULATION OF THE
               SYSTEM  COMPOSING OF NEARLY 4D,PDD ATOMS IN TOTAL.
EXECUTION TYPE / MACHINE
CPU (750MHz), scalar
1 CPU (VU-9.6Gflops, 333MHz), scalar
1 CPU, vector*
4CPU,vector*-parallel
1 CPU (1.56Gflops, 375 MHz), scalar
1 node (8 CPU), parallel***
4 node (32-CPU), parallel***
EXECUTION TIME**
SUN BLADE 1000,
SCALAR
248 sec.






SR8000
SCALAR-PARALLEL




1914 sec.
316 sec.
179 sec.
VPP5000
VECTOR-PARALLEL

2434 sec.
286 sec.
91 sec.



       * Vector mode means execution by auto-vectorized compilation, vectonzation ratio rs 96%
       ** Execution time ts the elapsed time
       *** Pseudo-vectonzation function, i.e fast supplying of the data from the memory for the CPU processing.
32
                                    RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  RESULTS or MO SIMULATION TECHNIQUE APPLIED FOR SEVERAL DNA LESIONS

     Cytosinyl radical (5-hydroxy-6-cytosinyl) - is a lesion produced  by  indirect  radiation
     action, i.e. interaction of active OH' water radical with cytosine base [11], (See Figure 1).

     This lesion is important in the study of strand break formation through the intermolecular
     process of H-abstraction  form  the sugar (pentose) and it emphasizes the importance of
     initial base damage in connection to strand breaks. The 200 ps of MD simulation of DNA
     dodecamer d(CGCGAATTC*GCG)2 with cytosinyl radical at the position  9; C*(9) reveals
     the strong bending at the A(6) and T(7) DNA segment. Since this large bending is not
     observed at the damaged site C  (9) it suggests intermolecular interactions among C*(9) and
     A(6) and T(7). In addition to the bending,  large distortions and disruptions of hydrogen
     bonding network between bases of neighboring pairs are observed.

                                      FIGURE 1 :
                   STRUCTURE DF 5-HYDRoxY-6-cYTosiNYL RADICAL
                   (AXIAL POSITION OF QH is MARKED BY SHADOW).
     Thymine dimer (5,6 cis.sin cyclobuthane thymine dimer) - is photo lesion produced by
     UV radiation in sunlight and is one major factor causing the skin cancer. It is formed as a
     covalently bonded complex of two adjacent thiamines on the single strand of DNA. This
     damage is very frequent but almost 90% of TDs are repaired within a short time, order of
     minutes and only few are experimentally observable and originate future changes on the
     cell level, [12] (See Figure 2).

     This study was  conducted  with DNA dodecamer d(TCGCGTATGCGCT)2, where TAT
     indicates the thymine dimer. The results of 600 ps of MD simulation shows that this lesion
     doesn't disrupt double helical structure and hydrogen bonds are well preserved throughout
     the simulation. Thymine dimer lesioned DNA, if compared with the native one, has sharp
     bending at the TD site which is originated by the two covalent bonds C(5)-C(5) and C(6)-
     C(6) between the adjacent thymine bases forming the thymine dimer, (See Figure 3). This
     bending discriminates  the  lesion  from  the  native DNA segment  and originates the
     conformation that facilitates the formation of DNA-enzyme complex by complementary
     structural shapes of the repair enzyme and bent DNA, (See Figure 3).
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBB                                33

-------
                                              FIGURE 2:
            THYMINE DIMER AB A COMPOSITION or TWO ADJACENT THYMINE BASES COVALENTL.Y
            JOINED BETWEEN C(5)-C(5) AND C(6)-C(S) ATOMS O F AD J AC ENT THYM I N E BASES.
                                                covalent bonds
                                               FIGURE 3:
                STRUCTURE OF T4 ENDONUCLEASE V AND THYMINE DIMER LESIDNED DNA
         AT 3DD PS OF MD. THE ARROWS SHOW THE POSITION OF THE CATALYTIC CENTER OF THE
                                 ENZYME AND THE DIMER ON THE DNA.
                                                            -Thymine
                                                            Dimer
              Thymine dimer is repaired by the repair enzyme T4 Endonuclease V that slides on non-
              target sequences and progressively incises at all dimers within the DNA molecule. This
              enzyme binds to DNA double strand in a two-step process: at first it scans non-target DNA
              by  electrostatic interactions to search  for damaged sites, and at second it sequentially
              specifically recognizes the dimer sites. The process of binding of T4 Endonuclease V to
              thymine dimer lesioned DNA was simulated with MD method. Considering the limitations
              arising from the simulations of large systems and requirements for  CPU time, the only
              catalytic center of enzyme was  subjected to the simulations. 1 he key amino acid of the
              enzyme - glutamic acid 22 of which carboxyl chain plays  a crucial role in the cleavage of
              N-glycosyl bond in DNA (base excision repair) together with surrounding 9 amino acids (8
              of HI and 2 of H2 helices) were selected to form the simulated part of enzyme. After nearly
              100 ps of the MD simulation,  the catalytic  part of enzyme approached the DNA at the
              thymine  dimer site,  docked into it, and this complex remained stable  afterwards (the
              simulation was performed for 500 ps) (See Figure 4).
&EPA
                                           RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                       FIGURE 4:
   COMPLEX QFTHYMINE DIMER LEBIDNED DNA MOLECULE WITH CATALYTIC CENTER OF
  REPAIR ENZYME T4 ENDONUCLEASE V FORMED DURING 1 DO  PS OF MD SIMULATION.
      When the same simulation was performed with the non-lesioned native DNA molecule, the
      catalytic center didn't fuse into the DNA molecule and the DNA-enzyme complex was not
      formed.  Further consideration of the factors that caused the fusion on the DNA and repair
      enzyme, the electrostatic interaction energy between the dimer lesion and catalytic center
      was calculated. It has been found that while the electrostatic energy of thymine dimer is
      negative around -10  kcal/mol,  the electrostatic energy  of glutamic acid 23 (the closest
      amino acid to the C5' atom of phosphodiester bond of dimer) is around +10 kcal/mol. The
      value of electrostatic energy represents the total  electrostatic interaction in the selected
      molecules  was calculated  by using Particle Mesh  Ewald  Sum technique for infinite
      simulated volume of repeating units through periodic boundary conditions; i.e. no cut-off
      distance was applied.  Since  the electrostatic energy of the native thymine is  nearly 0
      kcal/mol, the value of electrostatic energy represents a factor discriminating the thymine
      dimer lesion from the native thymine [13].

      Thymine glycol  (5,6-dihydroxy-5,6-dihydro-pyrimidine) -  is  observed in DNA after
      irradiation in vitro as well in vivo and after oxidation by chemicals (See Figure 5).

                                       FIGURE 5:
      MOLECULE OF THE THYMINE GLYCOL (5,6-DiHYDROXY-5,6-DiHYDROTHYMiDiNE).
                               H2O
                                                                                        oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                    35

-------
              Thymine glycol is known as causing Cockayne Syndrome - an inherited disorder in which
              people are sensitive to sunlight, have short stature and have the appearance of premature
              aging.  It is repaired with the repair enzyme Endonuclease III, which removes a number of
              damaged pyrimidines  from  DNA  via  its glycosylase  activity  and also  cleaves  the
              phosphodiester backbone at apurinic/apyrimidinic sites via an B-eliminatiori mechanism. To
              study the time  evolution  of the recognition processes of TG lesioned  DNA by repair
              enzyme Endonuclease  III  the 2 ns  of MD simulation of the following  molecules were
              performed: DNA 30-mer d(CCAGCGCACGACGCA'TG'GCACGACGACCGGG)2 where
              'TG' refers to thymine glycol; and repair enzyme Endonuclease III [14, 15].

              Analysis of the results  of 1 ns MD simulation shows that the double helical structure and
              hydrogen bonding are well kept through the simulation (except the base  pair of cytosine
              C5' - guanine C3' end, in which hydrogen bond pairing collapsed after 850 ps). DNA
              began to bend at thymine  glycol site after 500 ps of MD and bending continued until
              simulation was terminated at 1 ns. At the TG site the kink was observed, that relocated TG
              closer to DNA surface. Bending associated with kink at TG site dislocated glycosyl  bond at
              C5' atom closer to DNA surface, enabling it to be eventually approached by repair enzyme,
              (See Figure 6).

                                               FIGURE 6:
                SNAPSHOTS DF DNA MOLECULE DURING THE  COURSE OF MD SIMULATION.
              A DNA molecule is shown from the same side and angle with respect to the simulation
              box. The cytosine C5" end and guanine C3' end of DNA molecule are shown.  A molecule
              at 600, 800 and 1000 ps is bent and kinked at the thymine glycol site (shown as a Connolly
              surface).  Bending is expressed as the value of the angle measured between phosphates of
              the guanine (position 41), thymine glycol (position 16) and guanine (position 13); (numbers
              in degrees).

              8-oxoguanine (7,8-dihydro-8-oxoguanine) - is formed by oxidation of a guanine base in
              DNA,  (See  Figure 7).  It is considered to  be one  of the major endogenous  mutagens
              contributing broadly  to spontaneous  cell transformation.  Its frequent rnispairing with
              adenine during replication  increases the number of G-C  —» T-A  trans version mutations.
              This mutation is among the most common somatic mutations in human cancers.
&EPA
         36                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                      FIGURE 7:
                       MOLECULE OF NUCLEOTIDE WITH B-oxoG
     The 8-oxoguanine is recognized  and subsequently  repaired by  the  DNA  glycosylase
     (hOGGl in humans).  DNA glycosylases acting on single-base lesions use an extrahelical
     repair mechanism during which the enzyme recognizes oxidative damaged guanines and
     excludes normal DNA bases. The study on the 8-oxoguanine lesion was aimed to describe
     structural and energetic changes on the DNA molecule that are caused by this lesion, and to
     discuss how these changes may be significant in the formation of a complex with the repair
     enzyme.  The study method was  MD simulation (2 ns) of the two B-DNA molecules
     (native  DNA 15-mer, d(GCGTCCAGGTCTACC)2 and 8-oxoG  lesioned DNA 15-mer,
     d(GCGTCCA'8-oxoG'GTCTACC)2.

     In the  8-oxoguanine lesioned DNA  molecule, the disruptions  of weak hydrogen bonds
     between respective bases caused locally collapsed B-DNA structure.  While the hydrogen
     bonds between 8-oxoguanine and opposite cytosine 23 are well kept, the neighboring base
     pairs (adenine 7 - thymine 24, and guanine 9 -cytosine 22) are broken.  The hydrogen
     bonding of base pair thymine 10 - adenine 21 cease to exist very early, after around 50 ps
     of MD  simulation (See Figure 8).  hi the case of the native DNA, the B-DNA structure
     around native guanine 8 is well preserved.

     Adenine 21 on the complementary strand (separated from 8-oxoguanine by 1 base  pair) is
     completely flipped-out of DNA double helix (Figure 8).   This  extrahelical position is
     caused by the disrupted hydrogen bonds and by the strong electrostatic repulsion between
     the atoms in the region after 1 ns of MD. The cytosine 22 is also severely dislocated  form
     its  intrahelical position and its hydrogen  bonding to guanine 9 is  not  existing.   The
     extrahelical position of adenine 21 forms a hole in the  double helix that may favor docking
     of the repair enzyme into DNA during the repair process. The flipped-out base may also be
     inserted into the enzyme cavity, further ensuring the stability of DNA-enzyme complex
     [16].
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                37

-------
                                       FIGURE B:
                  FUIPPED-DUT ADENINE 2 1  ON THE COMPLEMENTARY
                           STRAND TD STRAND WITH B-DXDG.
       The figure also indicates non-existence of hydrogen bonds beMeen guanme 9 and cytosme 22. since the cytosme 22
       is severely dislocated from its mtrahelical position.

     CONTRIBUTIONS or THE CURRENT MD STUDIES TO RADIATION RIBK
     RESEARCH

     Ionizing radiation damages DNA and causes mutation and chromosomal changes in cells
     and in organisms.  Certain type of damages to DNA can lead to cell transformation 01 to
     cell death.  Radiation as well other chemical agents may damage DNA molecules several
     ways, directly or indirectly by interaction with DNA itself or with its environment.  Some
     damages caused by  ionizing radiation are chemically similar to damage that occurs
     naturally in the cell: this "spontaneous" damage arises from the thermal instability of DNA
     as well  from the endogenous  and enzymatic processes.   Several metabolic pathways
     generate oxidative radicals within the cells, and these radicals can attack DNA to give both
     DNA damage and breakage.  The significance of effective repair of these damages comes
     from two facts:
         >•   DNA is the repository of hereditary information.
         >   DNA is the blueprint for  operation of individual cells.

     Considering these important features,  it can be concluded that nearly all DNA damage is
     harmful.   Therefore, it is essential to reduce this damage  to a tolerable level.   The
     importance and the complexity of a DNA repair can be seen from the facts that:
         >   DNA is the only biomolecule that is specifically repaired, all the others are
             replaced;
         >•   More that 100 genes participate in various aspects of DNA repair, even in
             organisms with very small genomes;
         >•   Cancer is caused by mutations. In most cases, the "genetic instability" (elevated
             mutation rate) is required to permit accumulation of sufficient mutations to
             generate cancer during a human lifetime.  DNA repair mechanisms promote
             genomic stability and prevent cancer.  Many, perhaps most, cancers are thus at
             least partially attributable to defects in DNA repair.
38                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     Numbers of human genes encoding enzymes involved in the repair process of DNA have
     already been discovered (e.g. hRADSl, XRCC4, hRAD52, hREV3, hNTH, hOGGl, etc.).
     These enzymes are involved in DNA repair via several pathways and are functioning in
     certain phases of the complex repair process.  For the study of qualification of radiation
     risk, it is necessary to determine the specific pathways and features that are typical for
     radiation originated DNA damage.  In addition, the quantification of radiation risk would
     involve the study of relative efficiency of the repair processes in respect to the increased
     incidence of DNA damages above the endogenous level.

     The following are samples of several radiation originated human disorders characterized by
     defects in DNA repair.
         >•   Patients with xeroderma pigmentosum (XP) have clinical sun sensitivity, extensive
             freckle-like lesions in sun-exposed skin, increase in developing of skin cancer
             (basal cell carcinoma, squamos cell carcinoma and melanoma). All XP cells have
             been detected to be deficient in DNA repair.
         >   Patients with another human disorder - Cockayne syndrome (CS) - have increased
             sun sensitivity, short stature and progressive neurological degeneration. Cultured
             cells  from CS patients have defective DNA repair.
         >   Patients with trichodiostrophy have short stature, mental retardation and brittle
             hair.  Their cells have also defective DNA excision repair.

     Complexity of the repair processes doesn't allow a simple  approach to their solutions.  The
     MD simulation of recognition  may provide the stepwise description  of  bimolecular
     reactions at  the  radiation lesion  site through the capability to govern chemical and
     chemical-physical reactions in time intervals  that correspond to real-time formation and
     breakage of chemical bonds (order of femtoseconds).  The specific structural conformation
     and energetic properties of the lesion  may  be factors  that guide  a repair enzyme  to
     discriminate a radiation lesion from an endogenous one as well from a native DNA part.

  CONCLUSIONS

     This paper comprises of the results of MD simulation of several radiation lesions on a DNA
     molecule.  The studied lesions were 3 pyrimidine base lesions - cytosinyl radical, thymine
     dimer and thymine glycol, and 1 purine lesion - 8-oxoguanine. Except thymine glycol, the
     other three lesions are considered to originate neoplasic transformation of the cell and are
     found in human cancers.  The common features observed for all lesions are the specific
     conformations originated at the lesions site, like disruption of hydrogen bonding networks
     (cytosinyl radical, 8-oxoguanine), sharp bending at the lesion site (thymine dimer, thymine
     glycol), flipping-out the base on the strand complementary to the lesion and specific values
     of the electrostatic interaction energy at the lesion (8-oxoguanine).

     Among  these changes,  the most important is considered the  flipping-out base since it
     creates the empty space in the DNA double strand and this space may serve as a template
     for the docking of the enzyme and for the formation of the DNA-enzyme complex.  The
     strong bending that was observed in the thymine dimer lesioned DNA molecule forms  a
     complementary shape in respect to the repak enzyme T4 Endonuclease V and facilitates the
     formation of the complex. The electrostatic  interaction energy at several lesion sites differs
     from its  values at the native DNA site (thymine dimer, thymine glycol, 8-oxoguanine) and
     is considered as contributing to  the proper recognition  of  the respective lesion by
                                                                                         &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS                                 39

-------
     discriminating the lesion  from the native site.  This recognition  is important during
     electrostatic scanning of the DNA surface by the repair enzyme.

     The results of MD simulation, in addition to the existing crystallographic and molecular
     biology techniques, may contribute to the studies of radiation risk and DNA radiation
     damage repair by the dynamical  description of the structural and chemical processes that
     are under way at the lesioned DNA molecule.  It may also contribute to the determination
     of the key factors in the process of recognition of the lesion by the repair enzyme.

 ACKNOWLEDGMENTS

     The author  wishes  thank  to Mr. Toshiyuki Nemoto of The Research  Organization for
     Information Science and Technology for the  installation, maintenance and adjustment of
     the AMBER 5.0  code on supercomputers  VPP5000 and SR8000. The  valuable support
     from the all members of The Radiation Risk Analysis Laboratory, JAERI Tokai Research
     Establishment is also highly acknowledged.

 REFERENCES
 1. Harrison, S. and Aggarwal, A. Annu. Rev. Biochem. 59 (1990) 933.
 2. Gicquel-Sanzey, B. and Cossart, P. EMBO J.  1 (1982) 591.
 3. Ham, J., Thompson, A., Nedham, M., Webb,  P. and Parker, M. Nucleic Acid Res. 16:12
     (1988)5263.
 4. Beato, M. Cell 56 (1989) 335.
 5. Harris, L, Sullivan, M. and Hickok,  D. Computers and Mathematics with Applications 20
     (1990)25.
 6. Marx, J. Science 229 (1985) 846.
 7. Matthews, B. Nature 335 (1988) 294.
 8. Harris, L, Sulliwan, M. and Hickok, D. Proc. Natl. Acad. Sci. USA 90 (1993) 5534.
 9. Case, D.A., Pearlman, D.A., Caldwell, J.W., Cheathman III, T.E., Ross, W.S., Simmerling,
     C.L., Darden, T.A., Merz, K.M., Stanton, R.V., Cheng, A.L., Vincent, J.J., Crowley, M.,
     Ferguson, D.M., Radmer, R.J., Seibel, G.L., Weiner, P.K. and Kollman, P.A., AMBER 5.0,
     (1997) University of California San Francisco.
 10. Smith, P.E. and Petit, B.M. J. Chem. Phys. 105 (1996) 4289.
 11. Pinak, M., Yamaguchi, H. and Osman, R. J.Radiat.Res. 37 (1996) 20.
 12. Pinak, M. J.Mol.Struct.:Theochem 466 (1999) 219.

 13. Pinak, M. J.Mol.Struct.:Theochem 499 (2000) 57.
 14. Pinak, M. Jaeri-Research 2001-038, (2001).

 15. Pinak, M. J.Comput.Chem. Vol. 22, Iss.15 (2001) 1723.
 16. Pinak, M. J.Mol. Struct. :Theochem (2001) Submitted.
4D                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
          TISSUE RESPONSES TO IONIZING RADIATION  IN TERMS
                      OF MULTI-STAGE CARCINOGENESIS

  MARY HELEN BARCEULOS-HOFF
     Lcnvrence Berkeley National Laboratory, Life Sciences Division

     A general paradigm developed over the last several decades indicates that carcinogenesis is
     a multi-step process, which encompasses a series of events that are broadly categorized into
     three classes:
         >   Initiation, whereby a cell acquires the ability to become cancerous,
         >   Promotion by which a cell which expands into a population, and
         >•   Progression where that population of cells then acquires additional features that
             allow them to circumvent the normal constraints on their growth, which can result
             in disorganization and invasion into the surrounding tissue.

     The duration of this process depends on the tissue, but usually requires anywhere from
     years to decades.   For example, epidemiological data  indicate  that the breast requires
     almost 40 years from radiation exposure to actually develop cancer.

     Recently, a number of new cell-based phenomenons have been widely documented that are
     at odds  with a  simple DNA target-based model  of physiological radiation responses.
     Understanding the changing paradigm in radiation carcinogenesis  begin with reviewing
     current data  on  how  cells respond  to ionizing  radiation, which are briefly  summarized
     below.

     Radiation induced genes are studied with microarrays, which permit the study of several
     different genes at the same time, and proteomics, which is technology to study the patterns
     of many different proteins. The questions addressed by these studies include:
         >   "What are the proteins induced by radiation?"  This is a cataloging type of
             experiment. The next question which becomes more difficult to answer is,
         >•   "How do these particular proteins contribute to radiation responses, like cell death
             and cell cycle delay?"

     Another interesting  question/idea, with regard to non-homogeneous exposure,  like charged
     particle ionizing radiation, is:
         >   "How does the induction of proteins in one set of cells affect the other set of cells?"

     There is the additional idea that gene expression induced by radiation can, depending on
     the dose of the cell system, possibly  be a program of protection. These studies of adaptive
     response usually give a small dose of radiation to a population followed by a larger dose,
     and then compare the response to the challenge dose to that of cells primed by the smaller
     dose. The system is considered to have an adaptive response if the response to the second
     dose is reduced compared to cells irradiated with only one dose.  The  altered response is
     thought to occur because the cells have a persistence memory of being irradiated and have
     already entered into a  'yellow' alert state.
                                                                                         &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBB                                 41

-------
               Bystander effect is the term applied to the phenomenon where un-irradiated cells near
               those irradiated exhibit a response characteristically induced by  ionizing radiation. These
               studies suggest that unirradiated cells can exhibit increased:
                  >   Mutagenesis,
                  >   Sister chromatin exchange,
                  >   Apoptosis, and
                  >   Damage response signaling.

               Do the irradiated cells induce damage in un-irradiated cells? Although bystander is widely
               interpreted as deleterious, adding to adverse effects, we think that bystander responses can
               be a protective process, especially when the response  is increased apoptosis and damaged
               response signaling.  Apoptosis in a  multi-cellular organism is not a damage response; it's
               what is done to make tissues, to make fingers, to organize cells.  Apoptosis in itself is not
               damage to an organism because cells are built and recycled at the rate of billions of cells a
               day.

               A fourth phenomenon  is delayed effects -  the  observation that  populations  of cells
               surviving radiation exhibit altered  behavior many generations  after the initial radiation
               response.  In an irradiated population, some cells are killed, some cells persist. The cells
               that persist can give rise to daughter cells, and some of those  daughter cells acquire or
               exhibit non-clonal chromosomal aberrations, increased apoptosis, and delayed cell death,
               many generations after a radiation exposure.

               How do these  observations  of cellular  responses relate  to the idea that multicellular
               responses dictate the response of organisms to radiation?  I pose this question in  order to
               ask the following two questions:
                  >   How do tissues respond to damage at the cellular level?
                  >   Is the tissue response the sum of its parts, or can we consider the whole greater
                       than the sum?

               My hypothesis  is that multi-cellular tissues, organized as multicellular units  of function,
               respond differently to ionizing radiation compared to isolated  cells (Barcellos-Hoff, 1998).
               Cells are organized into tissues in order to perform specific functions.  The production of
               billions of cells a day and the degradation of those cells (in effect recycling them) indicates
               that any individual  cell in  a particular tissue  cannot be critical to  the health of that
               organism. However, it is also known that an  individual cell can give rise to cancer, i.e.
               almost all cancer can be said to be clonal in origin. Most research is concerned with what
               goes wrong in  single cells,  e.g.  after irradiation,  in  a tumor, etc., because one  cell can
               become a cancer.

               But in order to understand carcinogenesis, which is a process, the whole context  must be
               considered. Cancer is a disease of tissues, not cells.  For most solid cancers, the context
               consists of cells in an epithelium, which interacts with supporting cells from the stoma, the
               vascular and the immune system. Even the peripheral nervous system has been shown to
               impact the development of cancer.

               Multicellular responses are important to think about in order to begin to understand the risk
               associated with radiation exposure.   So to understand cancer, one must give consideration

3-EPA
          42                                  RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     to the fact that it takes a tissue to make a rumor (Barcellos-Hoff, 2001).  The following
     questions need to be asked:
         >   What are the important features of the tissue that allows a cell to express its
             potential?
         >   What allows a cell to perform its differentiated function? And,
         >   What allows some cells to express this abnormal capacity to form a tumor?

     The first thing that needs to be recognized is that individual cells such as liver cells, eye
     cells, gut cells, skin cells, all have the same genome, the same gene sequence. Likewise the
     epithelium of a tissue is different than the stoma, which is different from the vasculature.
     Not because of a change in  genetic sequence, but rather a change in the way the genes are
     expressed. Genome expression is mediated by signals by interactions and from signals that
     cells receive from the microenvironment, that include:
         >   Proteins at the cell surface known as adhesion molecules,
         >   Proteins that the cells reside in the extra-cellular matrix and growth factors, and
         >•   Soluble peptide proteins that signal between cell compartments, between individual
             cells, and between tissues.

     Why is understanding how tissues are  organized and how individual cells express their
     potential important for understanding how radiation causes cancer?  Because  in doing so
     we can ask, "does radiation alter these microenvironment signals?"   We  have used the
     mouse mammary gland to illustrate the response of the tissue to different doses of ionizing
     radiation, to different qualities of ionizing radiation as a function of time post irradiation,
     and also as a function of genotype. Here is a summary of the  responses tissues have to
     ionizing radiation:
         >   The response to ionizing radiation is extremely rapid. The activation of the TGF-P
             (transforming growth factor beta) protein occurs within minutes to an hour of
             radiation exposure (Barcellos-Hoff et al., 1994).
         >   These events can be quite sensitive.  The activation of TGF-P has been studied in
             great detail. The activation of TGF-P can be detected after a total body dose of 0.1
             Gy three  days after ionizing radiation exposure (Ehrhart et al., 1997).
         >   The activation of these signaling peptides or signaling proteins  is exquisitely
             sensitive  to small perturbations in the tissue.
         >   Furthermore, microenvironment responses are dynamic. The activation of TGF-P
             leads to the induction of collagen, which is also acted on by proteases in the extra-
             cellular environment (Ehrhart et al., 1997).

     Over the course of tune, some of these events resolve following a single exposure and some
     of them persist for weeks after a single  exposure to radiation.  This is an orchestrated
     response and in many ways has features that are very similar to wound healing.

     The induction of active TGF-P is really one of the primary mediators of wound healing in
     cetaceous tissues.  It can be found in ionizing radiation. And what's been shown in a series
     of other experiments  is that the reason TGF-P  is activated very efficiently under both of
     these scenarios is because TGF-P itself,  latent  TGF-P, is  redox sensitive (Barcellos-Hoff
     and Dix,  1995).  Focusing  on reactive  oxygen  activity causes or allows an orchestrated
     response in all the tissues, in all the cells. These events are radiation quality dependent.

                                                                                          &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 43

-------
            When comparing the effect of gamma radiation to a 1 GeV ion, we find some of the events
            occur at a different time course.  Other effects are specific to charged particle ionizing
            radiation. The basement membrane is specifically disrupted only in tissues exposed to high
            LET  radiation (Ehrhart  et al.  1996; Costes et al., 2000).  It is  also tissue specific, so
            although  these events occurred in  mammary glands,  similar,  but different, events are
            evident in the liver and in the skin.

            TGF-P activation is critical to understanding how individual cells respond to radiation. The
            biology of TGF-P is not only implicated in wounding, it is also implicated in a numerous
            other tissue damage processes.  TGF-P is really one of the truly amazing proteins.  All cells
            in the body produce it and all cells in the body have receptors. But the response to TGF-P
            is very  much cell type  specific, content specific, and concentration specific.  In general,
            though, this signal transduction leads to two major effects in epithelial cells. One is growth
            arrest and the other is apoptosis.

            We have asked the question, "What is TGF-P doing in the irradiated tissue?" To do this, we
            took animals that were depleted of TGF-P because of a lack of one copy of the Tgffil gene.
            Because TGF-P regulates itself, the heterozygote animals actually have a 90% reduction in
            tissue levels.  In these studies, we found that radiation-induced apoptosis  is abrogated hi
            Tgfp heterozygotes, indicating that  signaling from extracellular protein is critical to the
            cellular response to radiation.

            These studies led us to believe that some of these events in the microenvironment are
            crucial to determining the cell fate decisions acted upon by individual cells as a function of
            radiation exposure.  In  the multi-stage  carcinogenesis paradigm  genetic changes  drive
            proliferation and the  acquisition of  new phenotypes. But,  cells do not exist in isolation.
            Cells have neighbors. And based on cell biology, what an individual cell does depends on
            the signals it receives from its neighbors  and from extra cellular sources.  The question
            arises, "what do these tissue radiation responses indicate about  the risk of carcinogens"?

            hi  order to understand  carcinogenesis, we also have to understand the seed and the soil.
            Under what conditions does a cell  actually acquire the proliferate potential to expand and to
            invade the tissue?  Normal tissues are very efficient in suppressing cancer, but abnormal
            microenvironments can actually promote cancer.  We have shown that radiation induces an
            altered microenvironment (Barcellos-Hoff,  1993).

            But do these events actually contribute to the ability of radiation to act as a carcinogen?
            The mammary gland  has a unique feature,  in that it develops after birth.  So at the time of
            birth or up to three-weeks old, the mammary gland consists of a  mammary stoma, which is
            essentially adipose.  How can this  be used to test for radiation-induced carcinogenesis?
            Since the mammary gland develops  after birth, at three weeks  of age the epithelial bud can
            be cut out, leaving just the stoma.  The result is known as a cleared fat pad, but it's actually
            an epithelial-free gland. If epithelial cells are taken from another mammary gland at any
            point in the animal's life, and transplanted  into a cleared fat pad, the entire development of
            the mammary gland is recapitulated,  hi short, normal epithelial cells will grow out into the
            mammary fat pad  into  a normal  ductile tree.  This feature  of the tissue has been  used
            extensively to test for the presence of initiated cells in a donor mammary gland exposed to
            a carcinogen, either chemical or radiation, followed by transplanting those cells to a new
            host to then look for the frequency of tumors.
,:p,,i~   44                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
4*w=«u
£?*£-

-------
     In our experiments, rather than irradiating the donor cells, the host stoma was irradiated
     (Barcellos-Hoff and Ravani, 2000).  We then transplanted a cell line called COMMA-ID,
     which were derived from a pregnant, non-carcinogen exposed animal.  The COMMA ID
     have a stem cell population that allows them, when transplanted into  cleared fat pad, to
     branch out into complete ductile trees (Figure la).  But they have two mutations in p53,
     providing them with limited cancer forming potential. When injected into an adult mouse
     fat pad, small tumors are formed at low frequency. Even so, the majority of tumors in un-
     irradiated hosts actually regress if these animals are allowed to live ten weeks. However, if
     these cells are injected into three-week-old mice, they do not form tumors;  if injected
     subcutaneous, they do not form tumors; if injected into a nude mouse, they do not form
     tumors. In the following studies, ten weeks old mice  have been irradiated with a total body
     dose of 4 Gy. A period of six weeks was allocated before observation for tumor formation.

                                  FIGURE LEGEND

     A: COMMA ID outgrowth in unirradiated fat pad.

     B: COMMA ID  outgrowth in fat pad from mouse irradiated (4Gy) three days prior to
     transplantation.
     If COMMA-ID cells are injected into an animal that's been irradiated three days before,
     large tumors form at high frequency (Figure Ib).  Furthermore, this effect was persistent
     out to 14 days post irradiation. Tumor frequency increases from less than 20% to 80% of
     the transplanted fat pads,  but these tumors  were biologically different in that they're
     smaller in sham host than in the irradiated host.  Because these studies were whole body
     irradiations, one concern  was whether this  is a  local effect or a systemic effect.   By
     performing hemi-body irradiations where the animal was irradiated one side versus being
     injected with Comma D cells into both sides, tumors  only formed in the irradiated side.
     Thus, the neoplastic potential of COMMA ID cells is suppressed in a normal environment,
     but  in  irradiated  environment,  they quickly  form large tumors  that do not regress.
     Essentially, we think that the perturbations that we're inducing with this dose of radiation is
     sufficient to either spare damaged cells that then persist in the environment or to release
     damaged cells from tissue constraints that normally suppress them.
                                                                                         oEPA
RADIATION RIBK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                     45

-------
     Another of the questions we'd like to  address  is whether or not radiation  also causes
     persistent changes in epithelial cells, and in particular, whether it occurs in human cells.
     Does the microenvironment actually mediate cancer in a human epithelial tissue?   Our
     preliminary studies  show that the progeny  of cells  surviving ionizing radiation exhibit
     disrupted extra-cellular signaling, an aberrant cell extra-cellular matrix, and a disorganized
     multi-cellular cluster.

     Our studies underscore two significant concepts regarding how tissues respond to damage
     at the cellular level.  First, tissue response to cell radiation is orchestrated program limits
     damage and is aimed towards restoring the homeostasis.  When cells  signal to each other,
     they are not necessarily signaling only bad things; they are also signaling things that will
     help the tissue recover, similar to the healing of a  wound in a complex tissue.

     Second, this program can be corrupted.  Meaning, the program is corrupted when it begins
     with an abnormal population, e.g. a pre-neoplastic population.  When mat occurs, then
     these same events - the induced gene expression and bystander effect - can take a different
     turn and lead to genomic instability and carcinogenesis.  Indeed, we have proposed that
     rather than thinking of genomic instability as an  induced process, it can be considered the
     absence of the multicellular process that normally suppresses aberrant behavior (Barcellos-
     Hoff and Brooks, 2001).

     One of the characteristics that a neoplastic cell takes on is that it becomes antisocial. It just
     doesn't listen anymore; it doesn't care what other cells say.  And because of that, it has an
     effect  on the  society of cells around it.  Actual tumor  cells may be oblivious to their
     surroundings, but at the stage of pre-neoplasia, potential cancer cells not only increase in
     their own numbers, but also recruit normal  cells.  The stoma associated with a tumor is
     quite different than the stoma of a normal tissue  in that it's an active process by which the
     tumor  cells send out signals that then recruit and alter the stoma, changing  the vascular
     response, changing the supporting cell response, and affecting the immune response.  So at
     the same time that the tumor cells are changing, there can be changes in the stoma. There
     is evidence that  an abnormal stoma can support tumor  genesis or  even  initiate cancer.
     Chronic inflammation wounding is a co-carcinogens and a number of transgenic models in
     which all cells express an ontogeny but tumors are found only at wound sites.

     However, not all tumor cells are impervious to signals from surrounding cells. Chronic
     myelogenous leukemia regresses or undergoes remission when the patient is  treated with
     interferon gamma. When those patients are treated with interferon gamma, it causes the
     tumor cells to re-express  pi  integrin, allowing them to attach back to the stoma (Bhatia,
     1994). When they attach back to the stoma, they normalize. The leukemic cells eventually
     overcome it, but there is the potential that you can actually treat tumors to cause them to re-
     associate, to re-establish their response to the signals  from the society of cells  and actually
     behave better.
46                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  REFERENCES
  Barcellos-Hoff,  M.  H.  Radiation-induced transforming growth  factor p  and subsequent
     extracellular matrix reorganization in murine mammary gland. Cancer Res., 53: 3880-3886,
     1993.
  Barcellos-Hoff, M. H., Derynck, R.,  Tsang, M. L.-S., and Weatherbee, J. A.  Transforming
     growth factor-p activation in irradiated murine mammary gland. J. Clin. Invest., 93: 892-
     899, 1994.
  Barcellos-Hoff, M. H. and Dix, T. A. Redox-mediated activation of latent transforming growth
     factor-pi. Molec. Endocrin., 10: 1077-1083, 1996.
  Barcellos-Hoff, M. H. How do tissues respond to damage at the cellular level?  The role of
     cytokines in irradiated tissues. Radiation Res., 150: S109-S120, 1998.
  Barcellos-Hoff,  M. H. and Ravani, S. A. Irradiated mammary gland stoma  promotes the
     expression of rumorigenic potential by unirradiated epithelial cells. Cancer Res., 60: 1254-
     1260,2000.
  Barcellos-Hoff,  M.  H.  It  takes a tissue to  make a  tumor: Epigenetic,  cancer  and the
     microenvironment. J. Mammary Gland Biol. Neoplasia, 6: 213-221, 2001.
  Barcellos-Hoff, M. H.  and Brooks, A. L. Extracellular signaling via the microenvironment: A
     hypothesis relating carcinogenesis, bystander effects and genomic instability. Radiate. Res.,
     756:618-627,2001.
  Bhatia, R., Wayner, E. A,,  McGlave, P.  B., and Verfaillie, C. M. Interferon-alpha restores
     normal adhesion of chronic myelogenous  leukemia hematopoietic progenitors  to bone
     marrow stoma by correcting impaired beta 1 integrin receptor function [see comments]. J
     Clin Invest, 94: 384-391, 1994.
  Costes, S. V., Streuli, C. H., and Barcellos-Hoff, M. H. Quantitative image analysis of lamina
     immunoreactivity in  1 GeV/amu iron particle irradiated skin basement membrane. Radiate.
     Res., 154: 389-397, 2000.
  Ehrhart, E. J., Gillette, E. L., and Barcellos-Hoff, M. H. Immunohistochemical  evidence of
     rapid extracellular matrix remodeling after iron-particle irradiation of mouse mammary
     gland. Rad. Res., 145: 157-162,  1996.
  Ehrhart, E. J., Carroll, A., Segarini, P., Tsang, M. L.-S., and Barcellos-Hoff,  M. H.  Latent
     transforming growth factor activation in  situ:   Quantitative and functional  evidence
     following low dose irradiation. FASEB J, 11: 991-1002, 1997.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                47

-------
             MONTE CARLO SIMULATION OF INITIAL PROCESS OF RADIATION-
                                     INDUCED  DNA DAMAGE

           RITSUKD WATANABE AND KIMIAKI SAITO (PRESENTED BY Mi ROB LAV PINAK)
              Radiation Risk Analysis Laboratory, Department of Health Physics,
              Japan Atomic Energy Research Institute

           ABSTRACT

              Study by Monte Carlo simulation of induction process of DNA strand breaks in aqueous
              solution is  presented for monoenergetic electrons.   The relationship  of electron  track
              structure, yields of chemical species after water radiolysis, and DNA strand break mediated
              by water radicals as a function  of electron energy was investigated.   Assumption of
              induction mechanism of single- and double-strand breaks (SSB and DSB)  used in the
              simulation are that SSB is induced by indirect action (OH or H reaction with DNA) and
              DSB is induced by two SSB on the opposite strands within 6 bp or 10 bp. The yields of
              SSB and DSB for all examined electron energies lie well within the experimental data
              when the probability of SSB induction per OH or H reaction with DNA is assumed  to be
              around 0.1 to 0.2.  The yield of SSB has a minimum at 1 keV, while the yield of DSB  has a
              maximum at 1 keV in the examined energies. 1 keV electrons form the strand breaks most
              densely.  The yield of SSB has a relation .) the  amount of the OH radical in steady  state.
              The yields  of DSB  dose not  directly correspond to the yield  of OH  radical,  though it
              reflects the amount of the reactions among chemical species.  This result shows  that the
              localization of chemical species enhances the production of DSB.

           INTRODUCTION

              The  quantification of the radiation biological  effect with low dose  or low  dose rate
              exposure is the  main concern  in risk estimation for radiation.  However, it is difficult to
              detect  directly the low dose or dose rate effect.  To estimate the low dose or dose rate
              effect,  it is important to clarify the mechanism of radiation effect started from physical and
              chemical processes after single-track irradiation.  Computer  simulation of DNA damage
              based  on track structure using the Monte Carlo method has been a powerful tool  to
              understand the biological effects of single-track irradiation.

              Low-energy secondary electrons  produced  in  interaction of radiation with matter  are
              essential for radiation effect on biological systems.  Previous studies suggest that these low-
              energy electrons or high LET radiation induce DNA damage difficult to repair because of
              clustering of damage sites by the dense energy deposition [1, 2, 3, 4]. Such DNA damage
              is predicted to be  likely to lead to the serious biological consequences [1]. Direct energy
              deposition (direct action) is obviously important for formation of clustering of  damage.
              However, even in  a cellular environment with high concentration of radical scavenger, still
              about  half of DNA  damage  is  ascribed to the  diffusible  radicals produced  in  water
              radiolysis (indirect action) [5].  Therefore, it is also important to know the behavior of
              indirect action on DNA damage concerning the difference of track structure.

              Present study is that  on the relationship between electron track-structure, yields of water
              radicals and DNA strand break mediated by water radicals as a function of electron energy.
              This study also examines whether the experimental observations concerning the effect of
              photon energy on the yields of SSB and DSB could be explained by common hypothesis of

oEPA
         4-B                               RADIATION RISK ASSESSMENT  WORKSHOP PROCEEDINGS

-------
     induction of strand break by indirect action.  In this study, DSB is treated as the damage
     representing clustering of damages, and SSB as the damage representing simple damages.
     Because clustered DNA damage difficult to be repaired is suggested to be more complex
     than simple DSB [1],  however, there has not been enough experimental data showing the
     exact structure of complex damages.

     First, the energy deposition patterns and reaction of water  radicals will be shown as  a
     function of electron energy from 100 eV to 1  MeV.  Secondary, the examination on the
     mechanism of strand  break induction and property of DNA damage by radicals will be
     presented.

  COMPUTATIONAL METHODS

     The simulation  was performed using  the  Monte Carlo  code system DBREAK [6,7].
     DBREAK allows the estimation of DNA strand break induction through simulation of track
     structures in  liquid water, production,  diffusion and  reactions  of  chemical species, and
     radical attack on plasmid DNA.  The target molecule considered is super coiled pBR322
     plastid DNA (4362 bp) modeled in atomistic level. A  detail of the DBREAK code system
     was described in papers by H. Tomita et al [6,7].  The reliability of the models has been
     proved by  comparison  with  other experimental  and theoretical studies concerning the
     evaluation  of dielectric function, the time dependent yields  of chemical species, and the
     yields of DNA strand breaks.  The explanation of the models and calculation conditions
     specific to present study were given  in detail  elsewhere  [8,9].   Figure  1  shows  the
     simulation flow of the code system.

     Simulation of track  structure,  radical production and radical  diffusion: Briefly,  the
     simulation  code  of electron track follows the primary and all the secondary generated in
     liquid water until they are  thermalized.  The code uses  liquid inelastic,  vapor  vibration
     excitation, and vapor elastic collision cross-sections.  The ionization and excitation events
     are assumed to generate water radicals; H, OH, FT1^,, e' aq and  O. The diffusion and reaction
     process of  these radicals were simulated during the period of during 10"12 - 10"6 s.  The
     dissolved oxygen at  atmospheric pressure  and OH radical scavenger are treated as  a
     continuum  [10],  then  each chemical species is  surrounded by homogeneously distributed
     oxygen and scavenger molecules. The chemical species considered in the diffusion process
     are H, OH, H+aq, e" aq , OH", H2O2, O, O2, O2", HO2, HO2", and O. The reactions among the
     water radicals considered in the simulation were described in the previous work [6,8].  To
     save the  computational time of the chemical stage, the independent reaction time (IRT)
     method [11] is applied.

     DNA model:  The three-dimensional  conformation  of super coiled  plastid  DNA was
     determined by Monte  Carlo simulation  based on the algorithm of Vologodskii et al. [12].
     The coordinate of atom of B-DNA was taken from study of crystallography [13] and fitted
     spirally to the conformation in the order of base sequence of pBR322.

     Strand break scoring: Pathway of DNA strand break induction was assumed as follows.
     Only indirect effect was considered in this study.  SSB was scored when OH or H reaction
     with DNA  occurs.  DSB is scored when two SSB on the opposite strands  are produced
     within  6  bp or 10 bp.  One DSB was also  scored as  two SSB.  The distance of 6 bp is
     shown to be critical in the experiment by Hanai et al. [14], and 10 bp is the most generally
     used value  in modelling studies [e.g.  15, 16, 17].  The strand break induction probability
     was assumed to be the same for OH and H reactions. OH and H tested several probabilities
     for SSB induction.

                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 49

-------
                                       FIGURE 1 :
                       SIMULATION FLOW OF THE CODE SYSTEM.
               Track structure
            Radical production

              ~1      .... IT.
                                              ,.
                                                             DNA
                                                             strand break
                      Modeling of plasmid DNA  Radical diffusion
  BEHAVIOR OF CHEMICAL SPECIEa AFTER WATER RADIOLYBIB
     At  first, effect of electron energy on the yields of chemical species produced by water
     radiolysis will be  shown in the context of the difference of track structure.  Figure 2(a)
     shows the yields at 10'12 s (initial yields) when the species begin to diffuse.  The initial
     products are H, OH, H2, O, Haq+ and eaq".  There is little change in the initial yields of all
     species for the incident electron energy in the energy region higher thanl keV. The initial
     yields of OH, Haq+ and H decreases with the decrease of the electron energy below 1 keV.
     Figure 2 (b) shows the primary yields (yields at  10"6 s) as a function of electron energy
     under oxygenated conditions.  The primary yields show only a little change at energy
     higher than 100 keV. In the energy range from 100 keV to 1 keV, the yields of OH and Haq+
     decreased with  decreasing  electron energy, while the  yields of H2O2 increased.   This
     tendency was reversed at energy lower than 1 keV.  This result agrees to the previous
     studies using similar method [18,19].
  (A) INITIAL ( 1 D"
                   FIGURE 2:
S ) AND (B) PRIMARY (1 Q"S S) YIELDS OF MAJOR CHEMICAL SPECIES
          6
                    ° .4. r, ..o. -a-
                                  >
                     103
       104
                      Energy (eV)
105
106
                                               £
                                               o
                                               o
                           •s  2
                                                     (b)
                                          •-OH
                                          *'  H +
                                          a  OH-
10
  IO"

Energy (eV)
10s
10°
5D
                                    RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
               FIGURE 3:
    TWD-DIMENSIQNAL PLOT OF THE
   EXAMPLE OF THE DISTRIBUTION OF
 ENERGY DEPOSITION  EVENTS IN A CUBE
      WITH  SIDE LENGTH OF  1 [1M.
                                FIGURE 4:
                 AVERAGE NUMBER OF CHEMICAL SPECIES
                   IN SPHERICAL DOMAINS OF DIFFERENT
                  RADII OF O.Z-ZD NM AS A FUNCTION OF
                            ELECTRON ENERGY.
                                                                       02 nm
                                                                         nm
                                                                    -••-20 nm
                                                                    -0-100  nm
                                                    102   103   104
                                              10b
             10 keV,1 Gy
1 MeV-1 Gy
Energy (eV)
     Such energy dependence of the primary yields can be attributed to the variation in the
     spatial distribution of energy deposition events.  Figure 3 shows the example of energy
     deposition position in the cube with the side length of 1 urn for 100 eV, 1 ke V, 10 keV and
     1 MeV. Absorbed dose in each cube is approximately 1.28 Gy. Since chemical species are
     generated along to the electron track, initial distribution of species directly reflects the track
     structure.   To ascribe the variation of spatial distribution of initial (at 10"12 s) species
     qualitatively, the number of species existing within a spherical domain around each species
     was counted for four different radii of 0.2 - 20 nm.  Figure 4 shows the average number of
     species within a spherical domain for different radii plotted as a function of initial electron
     energy. The number of species contained in the domain increases with the domain size.
     The electron energy corresponding to the maximum number becomes higher as the sphere
     becomes larger.  The highest density of the chemical species at 1 keV is given for radius of
     20 nm. Therefore, the enhanced reactions among the chemical species at 1  keV (Fig, 2(b))
     can be related to the initial density of chemical species in the region of 20 nm radius. Thus,
     the average number of chemical species existing in the sphere at 20 nm around a  species is
     found to be an appropriate indicator of the  number of the subsequent reactions and the
     primary yields.  This domain size  is comparable  with  the cube size, which shows the
     highest energy deposition localization at 1 keV.

  DNA STRAND BREAKS MEDIATED BY INDIRECT ACTION

     Next, the effect of water radicals on DNA strand breaks will be shown as a function of
     electron energy.  The yields of SSB and  DSB were calculated for approximately  1 Gy
     irradiation.  The yield of DSB is estimated for 6 bp or 10 bp as a critical induction distance.
     The calculated yields of SSB and DSB were compared  with available experimental
     observations performed for plasmid DNA in  diluted aqueous solution.  The SSB and DSB
     yields for different SSB induction probability  (PSSB) are plotted respectively in Figure 5 (A)
     and (B) as  a function of electron energy.  The yields of SSB and  DSB were estimated for
     three different PSSB, 0.2, 0.13 and 0.08.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                    51


-------
                                                FIGURE 5:
             YIELDS DF (A)  SSB AND (B)  DSB AS A FUNCTION DF ELECTRON/PHOTON ENERGY.
_
c

CO
Q
(?

~
tb
CD
W
o
-o
0)
o
7

6

5
4


3
2

1
n
• (a)

•
*fc
y >- •
:
D :
D •' D 4

: 4 ° .* "*" -
4 •

i i < i i
                         101  102  103  104 1(P 10s  107

                                  Energy (eV)
^
o

0

"i.
O
o

00
CO
o
o
O)
>-
o
1

6

5

4

3

2

1
0
1
- (b)

a



o x
X A
-
a

D +•
l
4

31 102 103
.

•

a

-
1

* c i
D • :
X
+- a
A :
10* lOf5 I03 1C
Energy (eV)
             Experimental data are shown by (0) Tomita et al (M) Shao et al. (a) Calculated data are shown by (O) PSSB = 20
             (O) Psst = 0.13 and (A) PSSB = 0.08 (b) Calculated data for d = 10 bp are shown by (O) PSSB = 2.0 (H) PSsn = 0.13
             and (A) PSSB = 0.08. Calculated data for d = 6 bp are shown by (X) PSSB = 2.0 and (+) PSSB = 0.13 PSSB: strand break
             induction probability, d- maximum distance to induce DSB.

               Some experimental data obtained for  different energy photons are shown in the  same
               graphs for comparison with the calculation.  In the low energy region, almost all of incident
               photons are directly  absorbed by photoelectric effect by water molecules,  and produce
               electrons with energies close to  the  initial photon energies.  Although  the secondary
               electrons of 60Co y-rays have a wide energy spectrum, the effective energy is considered to
               be comparable with 1 MeV electron.   Thus, the experimental data for photons could be
               compared with the calculated  data.   The  difference  of radiation source  between the
               calculated and the experimental values  should be considered especially in the comparison
               of the data below 10 keV. Irradiation with monochromatic soft X-rays principally induces
               photoelectric effect so that nearly all of photon energy is divided into photoelectron and
               Auger electrons. As a result, energy deposition density in the  close  vicinity of a photo
               absorption point is higher than that is along an electron track of the same energy.  Such
               localization of energy deposition by soft X-rays may lead relatively higher DSB yield than
               energy deposition by monoenergetic electrons. As shown in Figure 5(a) and (b), the yields
               of SSB and DSB for all examined electron energies lie well within the experimental data
               when PSSB  is assumed to  be  between 0.08 and 0.2. The SSB  induction probability 0.2
               comes from the knowledge that 10 - 20 % of OH radical reacts with sugar-phosphate group
               [5].   The probability  0.13  is based  on the estimation by  Milligan et al. [20].  The
               comparison  of the recent experimental data for photons with our  calculation supported
               these break induction probabilities for OH radical reaction.

               Figure 5(a) and (b) also show that the  yields of both SSB  and DSB significantly change
               depending on the electron energy.  The SSB yield has a minimum at 1 keV and a maximum
               at 1 MeV. Inversely, the yield of DSB has a minimum at 1 MeV and a maximum at 1 keV.
               The experimental data by Tomita et al.  [21,22] shows that the yield of SSB decreases with
               photon energy and the yield of DSB increases with photon energy higher than around 2
               keV.  Also, the energy dependence of the SSB yield reported by Fulford et al. [23]  has a
               minimum around 1 to 5 keV.  This photon energy dependence for the yields of SSB and
               DSB  is in  agreement with the data obtained in the present calculation.  The reasonable
               agreement of calculated energy effect on the strand break yields as well as their absolute
               values with the comparable experimental data  is considered to support the validity of the
               models and assumptions used in this simulation.
oEPA
         52                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     It is clear that the variation of track structure and initial distribution of radicals relates to the
     strand break yield. The yield of SSB due to indirect action shows a close relationship to the
     amount of OH radical in steady state.  The yield of OH radical decreases as time elapses
     after  it was  produced according  to  the  chemical  reactions. T he reaction rates among
     radicals are subject to their spatial distributions since the radicals in the closer positions are
     more likely  to react  at higher probability. As shown in above, the initial radical density
     within the spherical  domain of 20 nm radial has a close relationship to  the amount of the
     following chemical reactions and the radical yields at 10~6 sec.  Accordingly, the amount of
     OH radicals  becomes minimum for  1 keV electrons. This results in the minimum SSB
     yield for 1 keV electrons. On the other hand, direct relation is not observed between the
     yields of DSB and its main cause,  OH radical. This observation shows that the localization
     of radicals enhances the production of DSB. Dense energy deposition just around DNA
     could induce the plural  reactions  of radicals with DNA  close to each other.  As a result,
     higher DSB yield is  obtained for the electrons around 1 keV. This result indicates that the
     number of radicals within about 20 nm is also essential to the yields of DSB  in the present
     condition. The mean diffusion distance of OH radical in the simulated system (lifetime ~ 6
     x 10"8 s) is around 30 nm, which may relate to the essential volume  for the strand break
     induction in aqueous system.

     To analyze the spatial distribution of strand breaks on DNA, the distance between the two
     closest breaks was calculated for  1.28 Gy irradiation. The frequency distributions  of the
     distance between the two closest  breaks were obtained by measuring the  number of base
     pairs between every two closest break sites in more than 104 plasmid DNA molecules. The
     difference of the localization of strand breaks depending on the electron energy is observed
     from the comparison of the distributions for different  energies.  The average  distance
     estimated from the distribution in the range of 0 - 2000 bp has a minimum of 14 bp for 1
     keV electrons, and a maximum of 24 bp for 1 MeV electron. Those for 100 eV and 10 keV
     electrons are 20 bp and 23 bp, respectively.

     The analysis of spatial  distribution of strand break shows that strand breaks  are formed
     most closely by  1 keV  electrons.  The  localization of strand breaks leads to high rate
     production of DSB.  This result also indicates that DSB observed for 1 keV electrons is
     likely to be complex DSB, such as the ones having more than three strand breaks within a
     few tens base pairs.  This fact suggests that the serious biological effect is expected to be
     induced  by  the low energy electron around 1  keV, though it should  be noted that the
     effective energy might depend on the target  size.  The estimation of the number  of DNA
     molecules having more than two breaks showed that the deletion or short fragment could be
     produced more frequently for 1 keV electrons than other energy electrons.

     SUMMARY AND CONCLUBIONB
     It is shown that the assumption on the strand break induction probability  in the range of 0.1
     - 0.2 for mainly OH reactions with DNA and on critical distance of about 10 bp for DSB
     induction could  explain the  corresponding experimental data.  The  inverse  energy
     dependence  experimentally observed for SSB and DSB are reproduced on the assumption
     of strand break induction mediated by radicals.  These agreements of the calculation with
     the comparable  experimental  data  support  the validity of the simulation models and
     assumption used.  Reduction of the  yield of SSB  around 1 keV reflects the OH radical
     yields, while the enhancement of the yield of DSB around 1 keV could be explained by the
     localization of the radicals generated close proximity of DNA.  Electrons around 1 keV are
     shown to be apt to produce  complex  damages  where more  than two  damages  are
     concentrated in a  small  DNA region.  This indicates that biologically significant DNA
     damage is produced by track end as 1  keV electrons at high rates.

                                                                                         &EPA
RADIATION  RISK ASSESSMENT WORKSHOP  PROCEEDINGS                                 53

-------
             ACKNOWLEDGEMENT

             The authors wish to  thank  Dr.  Hiroyuki Tomita  and co-workers  for providing the
             simulation code. RW would like to say many thanks to Professor Kotaro Hieda for his
             support and advice. The authors wish to thank Dr. Miroslav Pinak and Mr. Kenji Masuko
             for the help with preparation of the manuscript.

         REF-EKENCES

         1.  Goodhead DT, Thacker J, Cox R (1993) Effects of radiations of different qualities on cells:
             Molecular mechanisms of damage and repair. Int J Radiat Res 63: 543-556
         2.  Ward  JF (1988) DNA damage produced  by ionizing  radiation in  mammalian cells:
             Identities, mechanisms of formation and reparability. Prog Nucleic Acid Res 35: 95-125

         3.  Jenner T, DeLara CM, O'Neill P, Stevens DL (1993) The induction and rejoining of DNA
             double  strand breaks in V79-4  mammalian cells by gamma and alpha irradiation. Int J
             Radiat Res 64: 265-273
         4.  Price KM, Folkard M, Newman HC, Michael BD (1994) Effect of radiation quality on
             lesion complexity in cellular DNA. Int J Radiat Res 66: 537-542
         5.  Von Sontag C (1987) The chemical basis of radiation biology, Taylor & Francis, London

         6.  Tomita H, Kai M, Kusama T, Ito A (1997) Monte Carlo simulation  of physicochemical
             processes of liquid water radiolysis. Radiat Environ Biophys 36: 105-116
         7.  Tomita H, Kai M, Kusama T, Ito A (1997) Monte Carlo simulation of DNA strand-break
             induction in super coiled plastid pBR322  DNA from indirect  effects. Radiat Environ
             Biophys 36: 235-241
         8.  Watanabe R,  Saito K  (2001) Monte Carlo simulation of water radiolysis in oxygenated
             condition for monoenergetic electrons from 100 eV to 1 MeV. Radiat Phys Chem 62: 217-
             228
         9.  Watanabe R, Saito K, Monte Carlo simulation of strand-break induction on plasmid DNA
             in aqueous solution by monoenergetic electrons. Radiat Environ Biophys (submitted)
         10. Pimblott SM, Pilling MJ, Green NJB (1991) Stochastic models of spur kinetics in water.
             Radiat Phys Chem 37: 377-388
         11. Green NJB.,  Pilling MJ,  Pimblott  SM, Clifford P  (1990)  Stochastic modeling of fast
             kinetics in a radiation track. J Phys Chem 94: 251-258
         12. Vologodski AV,  Levene SD, Klenin KV, Frank-Kamenetskii M,  Cozzarelli NR (1992)
             Conformational  and thermodynamic properties of super coiled  DNA. J  Mol Biol 227:
             1224-1243
         13. Saenger W (1984) Principals of nucleic acid structure. Springer, Berlin Heidelberg New
             York
         14. Hanai R, Yazu M, Hieda K  (1998) On. the experimental distinction between SSBs and
             DSBs in circular DNA. Int J Radiat Biol 73: 475-479
         15. Nikjoo H, O'Neill P. Goodhead DT and Terrissol, M (1997) Computational modeling of
             low-energy electron-induced DNA damage  by early physical and  chemical events, hit J
             Radiat Biol 71: 467-483
,.^po-    54                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
^fijJSSS
 °~\~'»S"'

-------
  16. Terrissol M (1994) Modeling of radiation damage by  1251 on a nucleoside. Int J Radiat
     Biol, 66:447-451
  17. Paretzke HG, Goodhead DT, Kaplan IG, Terrissol M (1995) Track structure quantities. In:
     Atomic and Molecular Data for Radiotherapy and Radiation Research, IAEA-TECDOC-
     799, IAEA
  18. Hill MA, Smith FA (1994) Calculation of initial and primary yields in the radiolysis of
     water. Radiat Phys Chem 43: 265-280

  19. Pimblott SM, LaVerne JA (1998) Effect of electron  energy on the radiation chemistry of
     liquid water. Radiat Res 150: 159-169
  20. Milligan JR, Agulera JA, Ward JF (1993)  Variation of single-strand break yield with
     scavenger concentration for plasmid DNA irradiated in aqueous solution. Radiat. Res 133:
     158-162

  21. Tomita M, Hieda, M, Watanabe R, Takakura K, Usami N, Kobayashi K, Hieda K (1997)
     Comparison between the yields of DNA strand breaks and ferrous ion oxidation in a Fricke
     solution induced by monochromatic photons (2.147-10 keV). Radiat Res 148: 481-482
  22. Tomita M (1998) On the  mechanism of the  induction of DNA strand breaks by ionizing
     radiations in aqueous solution. [Master thesis] Rikkyo (St. Paul's) University, Tokyo, Japan
  23. Fulford J, Bonner P, Goodhead DT, Hill MA, O'Neill P (1999) Experimental determination
     of the dependence of OH radical yield on photon energy: A comparison with theoretical
                                                                                      &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                55   ,4S.?-

-------
           A  REGULATOR'S PERSPECTIVE ON MECHANISTIC APPROACHES TO
           THE STUDY OF RADIATION ONCOBENESIS AND RISK ASSESSMENT

           LOWELL RALSTON
              Radiation Protection Division, Environmental Protection Agency

              This paper discusses the EPA's view of the  mechanistic approaches to the study  and
              estimation of radiogenic cancer risks, focusing particularly on how these new approaches
              may strengthen the scientific basis of our low dose risk estimates and possibly change the
              way that we do radiation protection in our policies in the future.

              It will include an overview of the EPA's current risk assessment approach and discuss some
              of the uncertainties and questions in our present methods.

              It will  highlight some  of the recent  mechanistic  studies  that reveal important new
              information about radiation effects and mechanisms  observed primarily  in low doses  and
              dose rates.

              And finally, it will include comments on the implications of these new studies with respect
              to future risk assessment and protection approaches.

              As it has for the last three decades, EPA is one  of a handful of federal agencies  in the
              United  States  charged  with  radiation  protection.   Through  executive  order  and
              congressional  legislation, the EPA has the authority  and  responsibility to protect human
              health and  the environment  from uncontrolled releases of radioactivity and unnecessary
              exposures to  ionizing radiation.  Under this authority,  EPA conducts its  mission  by
              developing and  providing guidelines,  standards, and  policies for a wide  range  of
              environmental and occupational exposures.  In support of these efforts, the  agency also
              develops methods for estimating radiation doses and lifetime cancer risks.

              Because it's central to our work, radiation risk assessment is central to our program.  Risk
              assessment is a  complex and uncertain process  involving more than just dose response
              relationships.  In fact, it consists of many steps, including the characterization of all man-
              made and naturally occurring radioactive sources  in the environment and the  transport of
              radionuclides through the environment and the identification and quantifications of internal
              and external  exposures for both acute and chronic  scenarios.  These assessments often
              involve multiple radionuclides and sometimes-hazardous  contaminants  and chemical
              contaminants, as well.

              Models are commonly used to estimate dose risk to individuals, or when run in reverse, to
              calculate activity concentrations for specific radionuclides in various environmental media
              to correspond to target dose or risk limits.

              To assist risk assessors in this  process,  EPA  develops radionuclide specific cancer  risk
              conversion factors, or risk  coefficients, for  ingestion, inhalation, and external exposure.
              For inhalation exposures, we consider  a number of factors.  For example,  we assume
              constant concentrations of radionuclides in the environmental  media, but adjust for age and
              gender  specific ingestion and inhalation rates.  Using element specific biogenetic models,
              we account for the distribution, retention, and  excretion of each nuclide in  the body over
              time in order to calculate  time variant activity  concentrations in specific tissues after

oEPA
         56                                RADIATION RISK ASSESSMENT WORKSHOP  PROCEEDINGS

-------
     uptake. We then use dosimetric models to estimate tissue specific absorbed dose rates and
     then combine these with tissue, age, and gender  specific risk per unit dose estimates to
     obtain final estimates of the total risk to these tissues, as well as to the total body per unit
     intake,  which are the risk  coefficients.  As I mentioned, we correct for age and gender
     dependent usage rates for both inhalation and ingestion.

     And we use ICRP's current physiologically based systemic biogenetic  and  dosimetric
     models, including the new lung model, correcting  for age dependent absorption of activity
     from the gut to blood, as well  as for  age dependent organ masses and specific absorbed
     fractions.

     If you  don't  live long enough, you can't contract or die from cancer.  To correct for all
     competing causes of death using a life table approach, we use recent estimates of total
     mortality rate using the U.S. population, as well as estimates of cancer mortality rates for
     the same population for the same time period.

     For each cancer site, EPA applies age and gender specific cancer models.  Below 20
     centiGray  (cGy), we extrapolate risk linearly  with dose  without a  threshold, and we
     calculate our high and low LET risk separately. For low LET radiation exposures delivered
     chronically at low doses and dose rates, we decrease our risk estimates using a dose and
     dose rate effectiveness factor of two for all types — except breast, we use one — and for
     high LET  exposures, we increase risk estimates using a relative biological effectiveness
     factor of 20 for all sites, except for breast, where we use ten,  and one for leukemia. We now
     implement our models using a computer code called DCAL developed for us by Oakridge
     National  Laboratories,  and with it,  we've  tabulated risk  coefficients  for about  800
     radionuclides in our Federal Guidance Report  No. 13, which is available  electronically
     from our website.

     EPA is aware that there are large uncertainties  attached to our risk coefficients. And to
     address this  issue, we have begun to quantify  the uncertainties for many  of our model
     inputs,  including some of the ones I've already talked  about.  Of these, it seems to be the
     shape of the dose response curve below ten centigrade that has captured most people's
     attention.   And  this  is  understandable,  since  most occupational  and  environmental
     exposures occur in this region, and ten cGy  may be thought of as a nominal detection limit
     for our  current epidemiological data and approaches.

     So the question becomes how are we going to reduce the uncertainties in our models and
     improve confidence in our risk estimates?  Well,  to quote a famous  western philosopher,
     the secret to finding something is knowing where it is.  And as proof of this principle, I ask
     you to consider the treatment of chronic myeloid leukemia (CML) using a new drug called
     Gleevac.

     As many  of you know, CML occurs because  of a reciprocal translocation between
     chromosomes nine  and 22, resulting  in  the so-called Philadelphia  Chromosome.   This
     abnormal chromosome encodes for a fusion protein that conveys the prolific growth
     advantage to  immature white blood cells.

     From my understanding of the  mechanisms involved, researchers developed Gleevac, a
     kinas inhibitor, to block activation of this protein by phosphorylation.  As a result, these
     investigators  were able to  essentially  shut  down  the unregulated growth  of white  blood
     cells and thereby provide almost complete remission, with little or no side effects,  in the

                                                                                         &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 57

-------
              majority of patients taking this drug. In this example, the secret to finding a cure for CML
              was, in a sense, knowing where it was. In other words, in understanding the mechanisms
              involved.  And with this example in mind, we can ask why or can mechanistic approaches
              to the study of radiation carcinogenesis improve risk assessment?

              Lately, many people now believe that  traditional approaches alone may not provide
              complete answers to all our radiation protection questions necessitating the need to look at
              the new mechanistic approaches made possible by recent advances in biotechnology. And
              these approaches make sense, because we know that cancer involves alterations in cellular
              DNA and changes in gene expression in response to endogenous and exogenous damage.
              We now have maps and markers to help guide us through this unchartered territory made
              possible largely by the human genome  project and derivative programs.  These have
              provided us with a wealth of new information, such  as the identities and locations of tumor
              suppressor genes and ontogenesis, as well as a better understanding of the cell cycle, signal
              transudation pathways, and of DNA damage and repair mechanisms.

              As  you've heard from many of the speakers today, we have powerful new tools, such as
              fish for cytogenesis studies, and DNA micro rays for gene expression. And these allow us
              to ask and  answer questions previously unanswerable and offer  improvements  in  the
              sensitivity and specificity of our measurements, thereby improving the signal to noise ratios
              for detection of early events  in carcinogen sis.  In addition,  we have a core group of
              dedicated, talented, and hard working scientists who are working together in creative new
              ways to solve these pressing problems.

              Moreover, in the  United States, where there  is renewed interest in nuclear power  and
              nuclear waste issues,  interest  is high  and funding  is available  for programs, such as  the
              DOE's Low Dose Radiation Program. So the take-home message is it's a very good time to
              be doing this kind of work.

              But what  do we hope to  learn from these types of  studies?  Well, obviously, we'd like to
              learn everything we can,  along the way seeking answers to some important questions in
              radiation  protection.  For example, why do some individuals develop cancers and others
              don't  when given seemingly  equal doses  of radiation?  And for  that  matter, why  do
              different tissues and different cells respond differentially?  Are there limits to radiation bio-
              effects, and if so, where  are those limits? Can we use alterations in DN A and changes in
              gene expression as biomarkers or fingerprints for ionizing radiation?  And can we apply our
              understanding of these alterations in our mechanisms for the early detection and treatment
              of some kinds of cancers, as I alluded to for the example with CML with Gleevac.

              And  certainly, there are many other questions we'd like to answer, but what  have we
              learned so far? Well, we know from recent studies, both theoretically and experimentally,
              especially the recent micro beam studies, that a single track of ionizing radiation of both
              types can cause a wide spectrum of damage in cellular DNA, from simple base damages to
              single strand breaks to double strand breaks to more complex damage. We can show  that
              these lesions occurred in both targeted cells through nuclide cytoplasm irradiation, as well
              as in non-targeted bystander cells.

              And in evaluating the biological significance of these damages, we now know that it's the
              clustered  damages that are very important and possibly unique to ionizing radiation. These
              are particularly complex  lesions.  They are difficult to repair, especially by the process of
              non-homologous enjoining.

&EPA
         58                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     We also know that in those cells surviving this kind of damage, incorrect repair can lead to
     permanent aberrations and mutations.   These types of alterations have been shown to
     increase linearly without a threshold down to at least a few cGy and are biologically
     relevant with respect to carcinogenesis in people.

     Understandably, not everybody agrees with EPA's low dose radiation risk estimates.  Some
     believe that the risks should be higher than what we project them to be, and others believe
     that they should be lower.  And you've heard from a number of speakers today that there
     may be new evidence to support both cases.  For example, bystander and inverse dose rate
     effects  may argue for increased risks, whereas threshold models and adaptive responses
     may suggest lower or no effects  at low  doses.  And things like DDREF and RBE may
     modulate in either direction,  either  increasing or decreasing our  risk estimates, so it's
     extremely important that we learn as much as we can  about these phenomenons and the
     magnitude of the effects they may convey.

     There are a number of investigators looking at the radiation induced bystander effect for a
     number of different end points and for both high and low LET radiation. For this particular
     example, the study of Brenner, Little, and Sachs recently proposed a quantitative model for
     the bystander phenomenon based primarily on the single cell, single particle micro beam
     studies  at  Columbia University,  as  well as the studies  of bystander  effects  by other
     investigators.  The  model  predicts  that  alpha particle  induced  in  vitro  ontogeny
     transformation appears to be non-linear below about 40 cGy  due to direct and indirect, or
     bystander, effects.  Direct effects arise when every nucleus in the cell in the population is
     irradiated with exactly one or exactly more than one alpha particle and take the form of a
     linear dose response curve, whereas bystander effects are evident when a small fraction of
     the cells in the population, say  one in ten or so, are irradiated with exactly one or more than
     one alpha particle.  In fact, the model projects that the bystander effect acts  in a binary
     effect, in all or none fashion, predominating at the lowest doses, one or two alpha particles,
     and saturating  as more  and more cells are hit.   The authors concluded that if their
     postulating  mechanisms  are applicable  in  vivo, the  consequences  for low dose risk
     estimation might be major.

     Our current models assume a linear relationship between risk and dose, but also dose rate.
     However, the Villicheck and  Newton example  used data from several investigators to
     arrive at this wonderful non-linear parabolic curve with dose rate versus response.  And
     specifically, they looked at mutations in both somatic and germ line cells in mice and found
     a region between a tenth  and one  cGy per minute. Which on  the log scale is minus one to
     one and where there is essentially error free DNA repair. This was named the minimum
     mutagenic  dose rate region, or MMDR.  In this range, they discovered that low LET
     radiation produced as much damage as the cell sees in one minute from endogenous
     sources of oxidative damages.   And these are simple damages, such as base damages and
     simple strand breaks.

     And they postulated that irradiation in this dose region increases the normal damage rate by
     about ten to one hundred percent and that the cell repairs itself with few, if any, errors.
     And they concluded  that the  cell must  be  exquisitely tuned to the damage frequency.
     However, it's speculated that below a tenth of a cGy per minute, the cell doesn't detect the
     damage signal  and doesn't repair the excess damage, resulting in an  inverse dose rate
     response curve where the mutation rate actually increases with decreasing dose  rate.  On
     the other hand, above a cGy per minute, the cell can't keep up with the damage, and so you
                                                                                         &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 59

-------
              get the conventional linear response with increasing mutation rate with increasing dose
              rate.

              They further suggested that their observation explained, at least in part, adaptive responses,
              which  many  studies deliver  adapting  doses  in  this dose rate region.  However, they
              questioned the significance of this region with respect to environmental exposures, pointing
              out that natural background dose rate from low LET radiation in about 10 microR per hour
              or about .2 to .4 per minute is a million times smaller than this region.

              As I mentioned, EPA applies a dose and dose rate effectiveness factor below 20 cGy to
              reduce its low dose, low LET risk estimates. In this example, I show the work of Sorenson
              and co-workers who studied in vivo dose rate effects by looking at translocations in mice
              following acute chronic and fractionated gamma irradiation at cumulative doses between
              zero and 350  cGy.  As shown, they observed a linear quadratic dose response curve —
              that's the top curve for acute exposures -- and lesser linear dose response curves, the bottom
              ones, for chronic in fractionated exposures. And based on these curves, they calculated the
              DDREF of 14 at 350 cGy, and a DDREF of 303 at 50 cGy.  Currently, we use a DDREF of
              two at 20 cGy, which appears to agree very well with their findings.   Of course, other
              experimental systems and damage end points may lead to different values for DDREF.

              As an example of threshold models is a study by Rowland of bone cancer  mortality in
              female iridium painters.  Rowland found no excess cases of osteosarcoma below about a
              thousand  cGy, leading him to conclude a  dose threshold. The table  also shows  EPA's
              estimates based  on our  DCAL  modeling  of the  expected number  of  cases for  the
              corresponding dose ranges and sample  sizes.  Comparing the observed and the expected,
              we conclude that the estimates of less than one cancer case  in our columa are consistent
              with the observation of zero cases. Essentially, you can't have less than one person dying
              and see it, especially in a small group like this.

              So this finding may  suggest an alternative explanation to Rowland's postulated threshold.
              Unfortunately, there are just too few cases in this cohort to arrive at  any firm conclusions
              using current epidemiological techniques. In order to address questions  like this, we need a
              better understanding of the cellular and molecular events involved. And if we look again to
              the micro beam studies, we find that the traversal of a nucleus of a single cell by a single
              particle at the  lowest possible dose increases oncogenic transformation, ait least in vitro.
              And similar effects have been seen in non-targeted bystander cells. And these results may
              provide plausible mechanistic arguments against threshold.

              Notwithstanding, if EPA were to set thresholds for certain  cancers, we would still need to
              ask the question where do  we  draw  the  line, given  that people  vary  widely  and
              unpredictably in their response to irradiation?  Adaptive responses, like threshold models,
              fall into a category phenomenon that some people believe support  lower risk at low doses
              and dose rates and these have been studied by  many investigators  for a number  of end
              points, in this particular example by Assam and co-workers, and measured the  rate of
              induced malignant transformation and control in irradiated mice embryo cells preceded or
              not by an adaptive  dose, and they found that the transformation rate and the treatment
              growth through receiving or being exposed to an adaptive  dose of one hundred milliGray
              (mGy) followed by a challenging dose of four Gray(Gy) was approximately two and a half
              times lower than the transformation rate observed in the group exposed only to an acute
              dose of four gray.
&EPA
         6D                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     Their  approach  and findings are consistent  with those  in  other adaptive response
     experiments and serve as a useful example for illustrating a few key points about these
     types of studies.

     First, radiation induced adaptive responses do appear to reduce the magnitude of certain
     biological end points. However, I'd like to point out that in the studies that we're aware of
     anyway, adaptive responses do not entirely eliminate the adverse effects that they cause, as
     skewed by the fact that most controls are lower than even the treatment group receiving the
     adaptive dose, so there is some damage happening.

     Second, most such studies are either too short in duration or not designed  specifically to
     investigate possible late term effects perhaps caused by undetected, complex lesions.

     Third, the  relevant adaptive response with respect to environmental exposures is  highly
     questionable, because as  discussed earlier, the adaptive doses are delivered at a rate, which
     are orders of magnitude higher than natural background exposure rates.

     And finally, in most adaptive response studies, the reduction after all of that is only a factor
     of three or less in those cases.   And so in viewing  all of these phenomena, we now
     appreciate  that several opposing forces may affect the shape of the dose response  curve,
     and hence  our risk estimate, below ten cGy.  And all of them or none of them may be
     operational at any given point in time.

     Today most of these phenomena have been observed only in vitro and we clearly need to
     see if the same effects apply in vivo.  And while all of these phenomena are very intriguing
     and worthy of investigation in their own right, we must remember that our ultimate goal is
     to use the knowledge gained from this research  to improve our understanding  and
     assessment of human health risk from ionizing radiation.

     In particular,  we need to find ways to  extrapolate these results from  cell systems as
     transgenic  animal studies to man, and I'm confident that collectively, we will find a way to
     do this. In the meantime, EPA will continue to review and analyze the data that comes in.
     We don't do this all by ourselves.  We often look to our science advisory board for crucial
     consultations.  We sponsor critical assessments by  nationally recognized  organizations,
     such as our recent co-sponsorship getting the BEIR VI study underway.

     We look to the international community for radiation protection advice, limits, and models.
     And  in addition, we apply our own criteria and systematic approach to weighing  the
     evidence.  When that weight of evidence shifts, we do make changes to our model and  our
     radiation protection policies, albeit slowly at times.

     EPA  is responsible for protecting people  of all ages and both genders from uncontrolled
     releases of radioactivity and unnecessary occupational and environmental exposures. Risk
     assessment is an integral part of radiation protection.  It's a complex and uncertain process
     involving more than dose response relationships. To assist risk assessors, EPA uses state-
     of-the-art models to develop radionuclide specific risk coefficients.

     And to reduce the uncertainties in our models and increase the confidence in our  risk
     estimates,  EPA  is  looking  beyond traditional  approaches to  evolving  mechanistic
     approaches to study of radiation carcinogenesis.  It is believed that these new approaches
     may provide us with a more comprehensive understanding of the  cellular and molecular


                                                                                         &EPA
RADIATION  RISK  ASSESSMENT WORKSHOP PROCEEDINGS                                 61

-------
     events and processes involved in cancer induction. This knowledge may also help us to
     better answer questions important to radiation risk assessment and radiation protection.

     What about doing radiation protection strictly on a basis relative to background level?  In
     other words, you have a background level, and that background has certain variability in it.
     Do you make a scale not based on health effect or on anything else, but based on how much
     it is relative to the normal background?

     By mathematical construct, we can't go to a threshold because we would be projecting
     linearly. But we are making projections in this particular example in dose ranges where we
     should have fairly good confidence in making those types of estimates.  In other words, a
     thousand cGy is a thousand rad.  That's a pretty hefty dose and it depends on how many
     alpha tracks that are hit.  The point is that observation alone may not answer the question
     for us in terms of epidemiological observation. Modeling may not necessarily answer the
     questions either, but if we understand the mechanisms  involved, changes  that we  can
     predict, maybe somehow we can arrive at the truth somewhere in between.

     Not  sure what the  answer is at this point, but there could  be something  other than a
     threshold which we're observing simply because we're limited by seeing whole people die,
     rather than fractions of people for small populations.  We can only see these effects in a
     very large level in terms of excess death.

     We do not have effective dose  equivalents  for our risk modeling.  We do have absorbed
     dose rates. And then we apply the tissue and age and gender specific risk coefficients to
     those tissues, so it's not a fixed  system. It incorporates all of the ICRP models to date for
     dosimetry, for biogenetics, and for the lung model. All of those mechanisms are in place.

     Is there a policy or  a policy in development that goes beyond the consideration of average
     man to the consideration of the most sensitive subgroup?

     We have an executive order signed by President Clinton that  says to all of the federal
     agencies that we should consider children's health risk in our risk assessments and we
     might consider them to be a sensitive subpopulation.  The only problem is that it's unclear
     as to how to do that yet with respect to age specific intake and risk estimates. Although our
     model does incorporate that into it, we don't  do it specifically.

     We don't have any other formal policy on any other groups of people, but we are asked to
     consider in our  risk assessment the maximally  exposed individual and we use our  risk
     coefficients,  which were age-averaged lifetime cancer risk estimates  to  sort of cover
     everybody of every age.  We have the capability of doing it, but we don't have an official
     policy to do it.

     Regarding genetic sensitizes, it  is well recognized that genetics is a significant component
     in the development  of human cancers, both for a wide variety of cancers and most probably
     and certainly in some instances  for the radiation induced cancers. There are low penetrates
     and high penetrates with respect to genetics, so there is a small fraction of the population
     that has  severe genetic predisposition to cancer,  less  than a percent or two, and there is  a
     higher group that has lower penetrates. This is assumed to be automatically taken care of in
     our risk models, because we're using the whole population, vital statistics, and so all of the
     sensitive populations are included in that. They're just not specifically identified. In other
     words, as an aggregate, they are included in our group, but we  don't selectively call them
62                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS

-------
     out. There is the problem with this: when you set a limit, if you set the limit based on a
     specific population, you'll probably have to apply it to the whole U.S. population. We can't
     run around identifying individual people. So all you can do is switch the general limits up
     or down for the population. You can't identify the population and control it that well.

     What  our models will tell us  is that the  children in maybe the  first ten years, they
     accumulate 30 percent of their risk, but because we cannot identify,  although we can tell,
     we should have a limit here for age one to two, here for age four to five, here for age six to
     seven, et cetera.  That's impractical. So, in order to protect children, all we can do is lower
     the national limit. However, the children will still be at the same relative proportional risk
     that they were at to begin with, in that wherever the new limit is, they still accumulate 30
     percent of the risk in the first ten years, so they are the most sensitive population.

     Almost unless you can regulate on an individual basis, it won't do much good to know what
     proportion are more susceptible. If you knew that, given the current doses that are released
     by the current generation of CAT scans  in pediatric conditions, in children, many  of the
     CAT scanners give  relatively high doses,  certainly higher than is reasonable for young
     children, so that one could then recommend that certain radiological procedures  not be
     applied.

     Unfortunately, we don't have the authority to regulate medical for the general population
     but we do have for federal agencies. The challenge isn't so much how we protect him from
     an environmentally  contaminated site, but  how  we  protect  him  from  his genetic
     background. The radiation background  may have nothing to do  with  it.  There are
     numerous diseases that have very high incidence relative to genetic predisposition  but aren't
     necessarily  related to radiation exposure. The real challenge  is protecting him  from his
     genetic background and how we modify whatever we need to do to protect him from that.
     We do the best we can to cover peripheral issues, like genetic susceptibility.

     We should be very careful to introduce the concept of adaptive risk to radiation protection.
     There are- three points.

     Most of the in vitro adaptive response experiment is done in the low dose, the laboratory
     dose, not the low dose area of the practical protection, practical low dose.  I think that it is
     about  three different from the low dose in laboratory compared with the dose, which we
     talk about in the radiation protection. For instance, in an animal study done in the Institute
     of Radiological Science, four groups were irradiated: the control and the total dose 20 mGy
     and 400 mGy and 8,000 mGy. And even the 20 mGy total dose has diminished mortality.
     This kind of data cannot be applied from the in vitro adaptive risk experiment.

     And the second point is that most of the data would be data response data, the positive data
     and easy spotlights. Consideration must be given about the narrative data.

     And the third  point is that the point of adaptive response is what happened in the low dose,
     but sometimes it  is misunderstood with low dose rate, and when we introduce the concept
     of adaptive response or something like that, we have to figure out what happened in those
     low dose area and what happened in those low dose rate areas.

     These three points are very important points to introduce the adaptive response concept to
     the radiation protection field.
                                                                                          &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 63

-------
            THIB fABE INTENTIONALLY LEFT BLANK
oEPA
        64                          RADIATION RISK ASSESSMENT WORKSHOP1 PROCEEDINQB

-------
     CURRENT  ISSUES  IN  DOSIMETRY  SESSION
BACKGROUND

     This session featured eight dosimetry experts sharing their latest research. Updates on the
     National Academy of Sciences (NAS), ORNL and JAERI collaborations, and International
     Commission on Radiation Protection (ICRP) Committee 2 were presented.  The research
     topics included; tooth enamel to organ dose using electron spin resonance dosimetry, high
     energy radiation dose conversion coefficients, shielding calculations for dose evaluation,
     CT Voxel phantoms, ICRP new GI models and specific absorbed fractions in a Voxel
     phantom.

PAPERS FROM DOSIMETRY SESSION

     To follow are the papers written by the following conference presenters:
        > Evan B. Douple
        > Fumiaki Takahashi
        > Yukio Sakamoto (two papers)
        > Keith Eckerman (two papers)
        >• Hiroshi Noguchi
                                                                           oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS                           65   -4'"^-

-------
THE STATUS OF A REVISED DOSIMETRY FOR THE RADIATION EFFECTS
         RESEARCH FOUNDATION'S RISK ASSESSMENT OF THE
                             A-BOMB  SURVIVORS

  EVAN B. DOLJPLE
     National Research Council, Board on Radiation Effects Research

  ABSTRACT

     The studies conducted by the Radiation Effects Research Foundation (RERF) of the health
     effects in the survivors of the A-bombs in Hiroshima and Nagasaki, Japan; provide one of
     the world's most important risk estimates for ionizing radiation. One of the strengths of the
     RERF studies is the use of a relatively sound dosimetry estimate for the survivor radiation
     doses, a Dosimetry System established in 1986 known as "DS86."  Shortly after DS86 was
     instituted, the National Academy of Sciences established a committee of scientists charged
     with a review of all new information that was relevant to the Atomic-bomb dosimetry. In
     1996, that committee recommended in a letter report to the U.S. Department of Energy that
     a number of things should be done to improve the accuracy of DS86.  U.S. and Japanese
     working groups were especially encouraged to address the apparent discrepancy between
     estimates of numbers of neutrons derived from measurements during the last decade and
     those numbers  of neutrons predicted by DS86.  In its most recent assessment (2001) of
     results of ongoing measurements, the National Research Council committee reports on the
     status of the dosimetry.  The report provides a number of recommendations of work that
     should be done to reduce the associated uncertainties related to the dosimetry, incorporate
     new scientific  information obtained, and methodologies developed since 1986, and to
     attempt to resolve the issue of the neutron discrepancies.

  INTRODUCTION

     When the dosimetry for the Radiation Effects Research Foundation (RERF) was updated in
     1986 (to Dosimetry   System  1986, or  DS86"), there was  some  concern  that not all
     measurements leading to estimates of the neutron fluency from the atomic bombs were in
     agreement with those predicted by DS86. DS86 was a relatively  sophisticated dosimetry
     that enabled RERF scientists to reconstruct the doses for individual A-bomb survivors.  The
     recognized importance and strength of the RERF risk assessments were due in part to the
     level of confidence hi, and credibility  of,  the dosimetry — the denominator in the risk-
     assessment calculations and which is often a weakness or limitation in epidemiological
     assessments  of the  world's populations  exposed to radiation.   But  by 1996,  new
     measurements were continuing to report discrepancies between  their  predicted  neutron
     fluencies and DS86 predictions.

     It was clear to a National  Research  Council  committee of scientists  responsible for
     monitoring the science relevant to  DS86 that a series of experiments needed to be funded
     and completed to examine the various issues and input parameters that were challenging
     the  credibility of DS86.  That committee delivered a letter report2' to the U.S. Department
     of Energy (DOE) with recommendations.   The letter report stressed the urgency of
     completing the recommended work because:
         >   The world's radiation protection standards rely on the best RERF risk estimates
            and they are continuously under revision,
         >   Key scientists who have been studying the A-bomb dosimetry are retiring,
66                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
         >  Research teams are disbanding
         >  Facilities are losing their capability to conduct the needed studies due to reduced
            levels of funding, and
         >  Copper wire needed to do critical measurements to resolve the neutron discrepancy
            issue needs to be located or it will be lost forever.

  METHODS

     The National Research Council committee, chaired by Warren Sinclair, was  expanded in
     1998 to include scientists with the expertise needed to perform a comprehensive assessment
     of the  status of DS86.  The committee members  were charged by the DOE  to prepare a
     report  that would examine the recent results of studies by members of the Japanese  and
     U.S. working groups, review the proposed methodologies that might be brought to bear on
     improving DS86,  and recommend additional work that should be done to improve DS86
     and especially  resolve the neutron-discrepancy issue.  There were three major  factors
     leading to the request for a focused effort to assess and revise, if necessary, DS86.  The first
     was the 1986 letter report and the fact that DS86's credibility would continue to erode as
     long as the issues of uncertainty persisted. There was a need to collate and examine  closely
     the results of a variety of measurements that were  made by a number of investigators since
     DS86  was implemented more than 10 years ago.  A series of experiments needed to be
     developed, completed, and analyzed in order to ascertain the  best estimate of the "neutron
     discrepancy" if verified as a significant phenomenon.  Second was the recognition that
     there were things that were not incorporated into DS86, such as terrain shielding, that could
     and should be factored  into the dosimetry  calculations.   Third  was development of
     computing technology, which has been dramatic  since 1986 and which now enabled
     computational analyses that were not realistic in 1986.

     To do its work, the Research Council  committee  held a series  of meetings to  receive
     presentations by some members of the Japanese and U.S. working groups, as well as other
     scientists  including European scientists, in order to review the latest experimental results
     that were relevant to DS86. One of the committee meetings was held in Hiroshima, Japan,
     in conjunction with a joint meeting of the Japan and U.S. working groups in order that the
     committee could  receive valuable  input from  Japanese  scientists.   Two committee
     members, Wayne  Lowder and Harold Beck, worked closely with Takashi Maruyama from
     the Radiation Effects Association in Tokyo and Harry Cullings from RERF in Hiroshima in
     order to collect and collate the measurement results from all of the scientists working on A-
     bomb dosimetry so that the database could be analyzed by the committee members for their
     assessment.  Shiochiro Fujita and Dale Preston at  RERF, and Werner Ruehm  from the
     University of Munich, have also been instrumental  in assisting the  committee  and in
     locating copper samples for analysis by the two working groups.

  RESULTS AND DISCUSSION

     Addition of fast neutrons, either from leakage through a cracked bomb casing, or from an
     alternative plausible source term, could account  for the  increase in thermal-neutron
     activation or still agree  with  the well-known fast-neutron  activation measurements of
     sulfur-32  made in situ  soon  after the bomb  explosion. A method involving fast-neutron
     activation of copper to  nickel—63Cu(n,p)63Ni—has been developed to address  the neutron-
     discrepancy issue. The 63Ni is being measured in Japan on the basis of radioactivity and in
     the United States and Germany with accelerator mass spectrometry of copper samples from


                                                                                        4>EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 67

-------
              known  locations at  the  time of the bombing in Hiroshima  and Nagasaki.   If those
              measurements are completed, they should contribute to a better  understanding of the
              uncertainties in the estimates of neutrons and they have the potential to lead to a resolution
              of the neutron discrepancy.  The uncertainty in the gamma-ray fluencies measured and
              calculated is around +20%, but arbitrarily reducing the  measurements by 20% does not
              improve agreement with DS86 calculations.

              Many revisions  in  parameters  of DS86  have been  proposed since  1986, which  if
              incorporated, would improve the calculations in the revised dosimetry. For example, there
              have been changes in transport cross-sections and transport codes and refinement of the
              calculations by  using  increased numbers  of  gamma ray  and neutron  energy  groups.
              Complete evaluation of uncertainty in all aspects of DS86 is  still needed.   Uncertainty
              analysis has become more feasible because of the availability of new  information on
              possible sources of  uncertainty  and the availability of faster  computers, which permit
              benchmark and sensitivity studies.

              While biologic  dosimetry is  not generally expected to  be as  precise as good physical
              measurements, two methods  of biologic assays — measurements  of stable chromosomal
              aberrations and measuring electron-spin resonance in tooth samples — have been employed
              and have yielded results consistent with DS86 estimates for the same people, except for the
              Nagasaki  factory workers.  Those  data have  provided  motivation for recalculating the
              radiation attenuation and transport through the high-density materials that presumably
              provided shielding for the factory workers.

              With additional  funding provided by  the Japanese  and  U.S.  governments  (Japanese
              Ministry of Health, Labor and Welfare and U.S. Department of Energy), scientists in the
              Japanese and U.S. working groups have been coordinated to complete the analyses and to
              develop and  document  a revised dosimetry.   Robu * Young and  George Kerr are
              coordinating the U.S. working group's contribution to the new dosimetry. It is hoped that a
              new dosimetry will be  subjected to a review in 2002 with a final dosimetry published and
              made available for use by RERF's risk assessment in 2003.
               The committee's report emphasizes that although DS86 is a good system for specifying
               dose to the survivors  and  for  assessing risk,  it needs to  be updated  and revised.
               Uncertainties need to be fully evaluated. While the calculated gamma-ray influences agree
               well with measured values, more work needs to be done to provide a better estimate of the
               neutron component of the radiation exposures. Specifically, the report concludes:
                   >   The present program of 63N measurements should be pui ,-,ued to completion.
                   >   All thermal-neutron activation measurements, particularly those with 36C1 and
                      152Eu, should be evaluated with regard to uncertainties and systematic errors,
                      especially background.
                   >   Critical efforts to understand the full releases from the Hiroshima bomb by Monte
                      Carlo methods should be continued.
                   >   Ad joint methods of calculation (i.e., going back from the field situation to the
                      source term) should be pursued to see whether they help solve the neutron problem.
                   >•   Local shielding and local-terrain problems should be resolved.
&EPA
         68                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
         >   The various parameters of the Hiroshima explosion available for adjustment,
             including the height of burst and yield, should be reconsidered in the light of all
             current evidence in order to make the revised system as complete as possible.
         >   A complete evaluation of uncertainty in all stages of the revised dosimetry system
             should be undertaken and become an integral part of the new system.
         >   The impact of the neutron contribution on gamma-ray risk estimates in the new
             system should be determined.

REFERENCES
  1.  Roesch, W.  US-Japan Joint Reassessment of  Atomic Bomb Radiation Dosimetry  in
     Hiroshima and Nagasaki—Final Report.   Vol.  1  and 2.   Radiation Effects Research
     Foundation, Hiroshima, Japan (1987).
  2.  Letter report to Frank C. Hawkins (U.S. Department of Energy) from  National Research
     Council signed by Warren K. Sinclair (August 26, 1996).
  3.  National Research Council.  Status of the  Dosimetry for the Radiation Effects Research
     Foundation (DS86). National Academy Press, Washington, D.C. (2001).

  ACKNOWLEDGMENTS

     The Academies is indebted to the hard work of  members of the committee, to the many
     scientists who provided their scientific data and advice, and to Drs. Maruyama and Cullings
     for their assistance  in coordinating the collection and sharing of the data by investigators.
     Support for this study was provided by the U.S. Department of Energy through cooperative
     agreement No. DE-FC03-97SF21318 with the Office of International Health Programs.

  RECOMMENDATIONS

     Recommendations of the  Committee on Dosimetry for the  Radiation Effects Research
     Foundation in a letter to Frank C. Hawkins dated August 26, 1996:
         >   That investigators  vigorously pursue experiments that will lead to improved
             confidence in a revised DS86.
         >   That investigations to resolve the neutron uncertainty be pursued, including:
             *   Evaluation (quality assurance) and intercomparison of U.S. and Japanese
                measurements of thermal neutrons in order to assess the handling of background
                problems (including the use  of samples from long distances) and to assess total
                uncertainty in each measurement.
             •   Application of the 63Cu(n, p)63Ni reaction for fast-neutron measurements by both the
                U.S. and Japan (this requires an intensive  search for copper samples, particularly up to
                500 m and beyond, if possible, in both Hiroshima and Nagasaki).

         >   Calculations of weapon leakage and nitrogen cross-section experiments.
         >•   That a revised DS86 include a re-evaluation  of gamma rays at Hiroshima, yield,
             height of burst, the U.S. Army map (survivor locations), and shielding.
         >•   That a strong effort be initiated to quantify uncertainties in all phases of DS86 and
             any later revision with a view to upgrading all estimates of uncertainty that are an
             integral part of the dosimetry system.
                                                                                        oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                69

-------
          MEMBERS or THE COMMITTEE
              Chair of the Committee on Dosimetry for the Radiation Effects Research Foundation is:
                 >  Warren K. Sinclair  Escondido, CA
              Members of the Committee on Dosimetry for the Radiation Effects Research Foundation
              are:
                 >•  Harold Agnew      Solana Beach, CA
                 >  Harold L. Beck      New York, NY
                 >  Robert F. Christy    California Institute of Technology, Pasadena, CA
                 >  Sue B. Clark        Washington State University
                 >•  Naomi H. Harley    NYU School of Medicine
                 >•  Albrecht M. Kelleher University of Munich
                 >  Kenneth J. Kopecky  Fred Hutchinson Cancer Center, Seattle, WA
                 >•  Wayne M. Lowder  Valhalla, NY
                 >•  Alvin M. Weinberg  Oak Ridge Associated Universities, Oak Ridge, TN
                 >  Robert W. Young    INSIGHT, Winter Springs, FL
                 >  Marco Zaider       Memorial Sloan-Kettering Cancer Center, New York, NY
sxEPA
         7O                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      CONVERSION FROM TOOTH ENAMEL DOSE TO ORGAN DOSES
                             FOR ESR DOSIMETRY

  FUMIAKI TAKAHASHI AND YASUHIRO YAMABUCHI
     Japan Atomic Energy Research Institute (JAERI)

  ABSTRACT

     Conversion from tooth enamel dose to organ doses against external photon exposure was
     studied in order to develop a method that can retrospectively estimate organ doses by the
     Electron Spin Resonance (ESR) dosimetry using tooth samples. Monte Carlo calculations
     using EGS4 code were performed to obtain dose to tooth enamel and organ doses by using
     a modified MIRD-type phantom. The calculations gave quantitative relations between tooth
     enamel dose and organ doses against external photon exposure. ESR dosimetry using tooth
     samples was  carried out with a realistic  physical phantom. Dose to teeth was also'
     investigated by  measurements using thermo-luminescence dosimeters (TLDs). A Voxel-
     type phantom was constructed from CT images of the physical phantom. Monte  Carlo
     calculations with the Voxel-type phantom  were performed to verify the  results of the
     experiments and the enamel doses calculated by use of the modified MIRD-type phantom.
     The obtained data are to be useful for retrospective assessments of individual dose in past
     exposure events by ESR dosimetry using tooth enamel.

  7. INTRODUCTION

     The Electron  Spin Resonance (ESR) dosimetry using teeth is considered to be a useful
     method to assume exposure in past radiation events where no available information can be
     taken from personal dosimeters l\ This method is based on measurements of radiation
     induced CO}3' radicals in hydroxyapatite of tooth enamels. Since the hydroxyapatite crystal
     in teeth can easily trap free electrons and the signal in exposed dental enamel remains
     stable for a long time, this method has been applied to retrospective dose assessments2)'3)'
     4). The intensity of the ESR signal has been related to the dose accumulated in teeth5''6)'7).
     Estimation of individual dose, however, ultimately requires doses to organs of interest.

     In the present work, Monte Carlo  calculations  were performed to obtain quantitative
     relations between dose to tooth enamel (enamel dose) and organ doses by using a human
     model with a newly defined teeth part. ESR dosimetry was carried out with tooth samples
     contained in a realistic physical head phantom. The absorbed dose to the teeth region was
     also  measured  with  thermoluminescence  dosimeters  (TLDs)  placed in the physical
     phantom. The results of the calculations and the experiments were  verified by additional
     Monte Carlo calculations with a  computational  model, which was constructed from
     computed topographic (CT) data of the physical phantom.
                                                                                    oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               71

-------
 z. METHODS

     2. 1 CALCULATION WITH A MODIFIED MIRDTYPE PHANTOM

     An adult MIRD-5 type phantom 8) designed by Cristy was used to calculate enamel dose
     and organ doses against external photon exposure. A region for teeth was newly defined in
     the head of the phantom9). Figure 1 shows an overview of the MIRD type phantom and a
     cross section of the head at the level of newly added teeth area. The teeth were grouped into
     five parts to examine the distribution of the enamel dose in the  mouth. Tooth  enamels,
     however, were not specified in the teeth model. The elemental composition for  the teeth
     region was based on the data for a whole  tooth in ICRP Publication 23 10). Two kerma
     factors for a whole tooth and tooth enamel were prepared to calculate the enamel dose9).

     The Electron Gamma Shower Code Version 4 (EGS4) U)  in conjunction with user's code
     UCGEN 12) was used to calculate absorbed dose to organs  and tissues. The data of photon
     cross section used in the radiation transport were taken from the library edited by Turbey et
     al13). Eight energies of incident photons were selected in the region between 30  keV and
     2500 keV.  Photon  parallel beams were  assumed  to  be incident on  the phantom.
     Calculations were performed for 12 incident angles with 30 degrees  interval to study the
     angular characteristics of the enamel dose and organ doses.

     2.2 EXPERIMENT

     Experiments were made with a realistic head phantom, which is made of tissue-equivalent
     plastic  and contains human skull. The trunk  of an Alderson RANDO  phantom was
     connected to the head phantom to take radiations scattered by a human body into account.
     Teeth were inserted in the upper and lower jaws of the phantom. The phantom was exposed
     to gamma rays emitted from  a 60Co source in Anterior-Posterior  (AP) and Posterior-
     Anterior  (PA)  geometries. After  the irradiation,  dental  enamels  were  separated
     mechanically from other parts of the teeth and subjected to ESR measurements.

     In addition  to the tooth samples, thermoluminescence  dosimeters (TLDs) were set in the
     head phantom to measure the absorbed dose  to the teeth  region. The TLD is made of a
     CaSO4 crystal and has a diameter of 4mm. Two gamma ray sources of 60Co and 137Cs were
     used.

     2.3 CALCULATION WITH A VOXELTYPE PHANTOM

     A generally called "Voxel (volume pixel)-type" phantom  14)- 15) was constructed from
     computed topography (CT) images of the physical phantom,  which  had been taken with
     1mm interval. One CT image has 512x512 pixels (picture  elements).  Each pixel in the CT
     images was segmented into soft tissue area, bone area, teeth area and cavity area, according
     to its CT value and location. Tooth enamels, however, could not be distinguished from
     other parts of teeth.

     Absorbed dose to the teeth region was calculated by using the EGS4 code in conjunction
     with user's code UCPIXEL 16). Eight energies of incident  photons  were selected in the
     region between 30 keV and 2500 keV. The AP and PA  geometries were  considered for
     irradiations of photon parallel beam. The material of teeth region was defined as a whole
     tooth or CaSO4to verify results of the experiments. The enamel dose was derived with two
     kerma factors for a whole tooth and dental enamel as described in section 2.1.
72                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  3. RESULTS AND DISCUSSION

     3. 1 DOaEB CALCULATED BYUBINB THE MODIFIED MIRD-TYPE PHANTOM

     Some of calculated enamel doses and organ doses are listed in Table 19). The values in the
     Angl (Avr.) geometry were obtained by averaging the doses all over horizontal incident
     angles. The data are given in the form  of the ratio of the organ or tissue dose to the air
     kerma, in the unit of Gy/Gy. The results show significant dependence of enamel dose on
     energy and direction of incident photons.

     Figure 2 depicts the calculated doses as a function of photon  energy for the Angl (Avr.)
     geometry. The enamel dose indicates different behavior  from  other organ or tissue doses
     for low photon energies. Since tooth enamel contains elements with higher atomic numbers
     such as calcium and phosphorus more than soft tissue and bone tissue, the enamel dose
     increases due to energy absorption through photoelectric effect. On the other hand, the
     enamel dose is near to other organ doses in the energy region above 300 keV, where the
     Compton scattering process is dominant interaction with tissues.

     Figure 3 shows  the dependence of enamel dose and some organ doses on the incident
     direction of 1250keV photons. Since teeth are  located at the  front part in the head, the
     absorbed dose to enamel is smaller than dose to the colon in the PA geometry. On the
     contrary, the enamel dose is larger than the colon dose for the lateral irradiation geometries,
     because colon is well shielded by the human body tissues. The angular dependence of the
     enamel dose is similar to that of dose to the thyroid, which is located just below teeth.

     3.2 E8R DOBIMETRYAND DOSE MEABUREMENTB WITH TLDB

     Table 2 summarizes distributions of the enamel dose in the mouth obtained by the ESR
     dosimetry and the calculations for a 60Co source. Since the relation between intensity of the
     ESR signal and enamel dose has not been determined yet, the ESR signal of teeth at the
     middle- and the back- part are given with  relative values to those at the front part, which
     are normalized to 1.0. The values  in the calculations are based on the result of 1250keV
     photons. The results of the ESR dosimetry agree with those of the calculation for the AP
     geometry. A steep gradient of dose can be seen  in the results of the calculations using the
     MIRD-type phantom for the PA geometry, while the distribution of enamel doses  is not
     clearly observed in  the results of the calculations using  the Voxel-type phantom and the
     ESR dosimetry for the same irradiation geometry. More photons are absorbed to soft tissue
     before reaching teeth in the MIRD-type phantom than the Voxel-type phantom and the
     physical phantom, as the mouth of MIRD-type  is filled  with soft tissue and the physical
     phantom has cavity area in the mouth.

     Table 3 shows the comparison of the measured dose with TLDs and the results calculated
     by using the Voxel-type phantom.  Numerical calculated  enamel doses are also presented.
     The  measured doses agree well with the results  of the calculations, where the material of
     teeth region was defined as CaSO4. The difference between the enamel dose and dose to
     teeth region with the material of CaSO4 does not exceed 7% in the calculations using the
     same computational code and human model. It can be mentioned here that doses given by
     the measurements using TLDs indicate  almost  same values as the enamel doses against
     external exposure of 662keV and 1250 keV photons.
                                                                                       oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                73

-------
     3.3 COMPARISON or CALCULATED RESULTS BETWEEN TWO HUMAN MODELS

     The calculated enamel doses by the two human models were compared in Table 4. The
     difference of the enamel dose between the two human models does not exceed 10% for
     energy region above 662keV, although there is an exception in the PA geometry of 662keV
     photons. The enamel dose by the Voxel-type phantom, however, is about 60% larger than
     that by the MIRD-type phantom in the  case, where 30keV photons were incident to the
     body from the backside. This result suggests that the size and structure of the human head
     can affect the enamel dose against external exposure of low energy photons.

  4. CONCLUSION

     Enamel dose was quantitatively related to  organ doses by the Monte Carlo  calculations
     using EGS4 code and  a modified MIRD-type phantom. The calculated enamel doses by
     using the Voxel-type phantom were valid to the results in the experiments. The model of
     the head did  not significantly affect enamel doses  for most cases.  The enamel dose,
     however, can be influenced by  the size and structure of the  head for photons  below
     lOOkeV.  The conversion coefficients from  enamel doses to organ doses obtained in this
     study 9) are to be useful for retrospective dose assessments  by the ESR dosimetry using
     teeth.

  REFERENCES
  1. Jacob, P., Bailiff, I., Bauchinger, M., Haskell, E. and Wieser, A., Retrospective Assessment
     of Exposures to Ionizing Radiation, ICRU NEWS June 2000, 5-11, (2000).
  2. Ikeya,  M., Miki,  T., Kai, A. and Hoshi,  M. ESR Dosimetry of A-Bomb Radiation Using
     Tooth Enamel and Granite Rock, Radiat. Dosim. Prot., 17, 181-184 (1986).

  3. Serezhenkov,  V.A., Dormacheva, E.V.,  Klevezal,  G.A, Kulikov,  S.M., Kuznetsov, S.A,
     Mordvintcev, P.I., Sukhovskaya, L.I., Schklovsky-Kordi, N.E., Vanin, A.F., Voevodskaya,
     N.V. and Vorobiev, A.I.  Radiation Dosimetry for Residents of the Chernobyl Region: A
     Comparison of Cytogenetic and Electron Spin Resonance  Method, Radiat. Prot. Dosim.  ,
     42, 33-36, (1992).
  4. Romanyukha, A.A., Ignatiev, E.A., Vasilenko, E.K., Drozhko, E.G., Wieser, A., Jacob, P.,
     Keirim-Markus, I.E., Kleschenko, E.D., Nakamura, N. and Miyazawa, C.  EPR Dose
     Reconstruction for Russian Nuclear Workers, Health Phys., 78(1),  15-20, (2000).
  5. Iwasaki, M., Miyazawa, C., Uesawa, T., Suzuki, E., Hoshi, H. and Niwa, K. Exposure Rate
     Dependence of the CO33-  Signal Intensity in ESR Dosimetry of Human Tooth Enamel,
     Radioisotopes, 41, 642-644 (1992).
  6. Iwasaki, M., Miyazawa, C., Uesawa, T., Ito, I. and  Niwa,  K.  Differences  in Radiation
     Sensitivity of Human Tooth Enamel in an Individual and among the Individuals in Dental
     ESR Dosimetry, Radioisotopes, 44, 785-788 (1995).
  7. Iwasaki, M., Miyazawa, C. and Uesawa, T. Effect of Tooth Position in the Oral Cavity for
     Various  Irradiation Geometries in Dental  ESR Dosimetry,  Radioisotopes, 48, 530-534
     (1999).

  8. Cristy, M. Mathematical Phantom Representing Children  of Various  Ages for Use in
     Estimates of Internal Doses, MUREG/CR-1159 (1980).
74                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  9. Takahashi, F., Yamaguchi, Y.,  Iwasaki, M., Miyazawa, C. and Hamada, T.,  Relation
     between Tooth Enamel Dose and Organ Doses for the Electron Spin Resonance Dosimetry
     against External Photon Exposure, Radiat. Prot. Dosim., 95, 101-108, (2001).
  10. International  Commission on Radiological Protection.  Report of the  Task Group  on
     Reference man, ICRP Publication 23 (Oxford: Pergamon Press) (1974).
  11. Nelson, W.R., Hirayama, H. and  Rogers, D.W.O. The EGS4 Code System,  SLAC-265
     (1985).
  12 . Takagi, S., Sato, O., Iwai, S., Uehara, T. and Nojiri, I. Development and Benchmarking of
     General Purpose User  Code of EGS4, Proc. of the 1st International Workshop on EGS4
     (Tsukuba), 86-96, (1997).
  13. Trubey, O.K., Berger, MJ. and Hubbell,  J.H. Photon  Cross-Sections for ENDF/B-VI,
     Advanced in Nuclear Computation and Radiation Shielding,  American  Nuclear Society
     Topical Meeting (1989).
  14. Zankl, M., Panzer, W. and Drexler,  G., Topographic anthropomorphic  models: Part II:
     organ doses from  computed topographic examination in pediatric radiology, GSF-Berict
     No.30/93, Forschungszentrum fur Umwelt und Gesundheit, (1993).
  15. Saito, K., Wittmann, A., Koga, S., Ida, Y., Kamei, T., Funabiki, J. and Zankl, M., The
     construction of a computed topographic phantom for a Japanese male adult and the dose
     calculation system, Radiat. Environ. Biophys, 40, 69 (2000).
  16. Funabiki, J., Terabe, M., Zankl, M., Koga, S. and Saito, K., An EGS4 user code with Voxel
     geometry and a Voxel phantom generation system, Proc. of the 2nd International Workshop
     on EGS4, Tsukuba, Japan, 8-12 August, 2000, KEK Proceedings 2000-20, 56 (2000).
                                                                                     &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               75

-------
                                                    TABLE 1 :
                 ABSORBED DOSE TO ORGAN OR TISSUE PER AIR KERMA IN FREE AIR (GY/BY)
(A) 50 KEV
Red bone marrow
Lung
Stomach
Bone Surface
Enamel
(B) 1250 KEV
Red bone marrow
Lung
Stomach
Bone Surface
Enamel
AP
0.428
0.986
1.34
1.95
7.23
AP
0.859
1.00
1.09
0.927
1.04
PA
0.700
1.11
0.464
2.40
0.889
PA
1.01
1.06
0.815
0.990
0.624
RLAT
0.288
0.437
0.0636
1.38
4.33
RLAT
0.721
0.731
0.450
0.749
0.956
ANGL (AVR.)*
0.447
0.749
0.651
1.86
4.13
ANGL. (AVR.)
0.854
0.896
0.832
0.891
0.889
                            ' Values obtained by averaging doses all over horizontal incident angles

                                                    TABLE 2:
                        ENAMEL DOSE DISTRIBUTION IN A MOUTH FOR A 6DCo SOURCE
(A) AP GEOMETRY
ESR Dosimetry
Calculation (Voxel)
Calculation (MIRD)
(B) PA GEOMETRY
ESR Dosimetry
Calculation (Voxel)
Calculation (MIRD)
RELATIVE VALUE*
FRONT TEETH
1.0
1.00
1.00
MIDDLE TEETH
1.0
1.00
0.95
BACK TEETH
0.9
0.95
0.90
RELATIVE VALUE*
FRONT TEETH
1.0
1.00
1.00
MIDDLE TEETH
1.1
0.95
1.14
BACK TEETH
1.1
1.06
1.32
                            * The signal intensities or the enamel dose at middle and back part are relative
                           values to those at front part, which are normalized to 1.0

                                                    TABLE 3:
                         DOSE TO TEETH REGION PER AIR KERMA IN FREE AtR (GY/C5Y)
SOURCE, GEOMETRY

137Cs, AP irradiation
60Co, AP irradiation
60Co, PA irradiation
MEASUREMENT

1.04
0.929
0.672
CALCULATION*1
CAS04*2
101
0.949
0646
ENAMEL DOSE
1.05
0.976
0.689
                     *1. Human Model, Voxel-type phantom
                     *2. Material of teeth region CaSOj (TLD)
oEPA
          7e
                                                RADIATIQN RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
5
o

H '
5
z
TO
Z
n
o

•g

8
a
ni •
n
o
z
                   Dose per air kerma in free air

                             (Gy/Gy)
                                                                -. 2

                                                                < -
                                                                •0 Z

                                                                '— 1
                                                                PI '''
                                                                D m
                                                                2 >
                                                                           o
                                                                            I
                                                                           Tl

                                                                           so

                                                                           I
                                                                           a.
                                                                           CD
                                                                           Q.

                                                                           C?
                                                                           CO
                                                                           •a
                                                                           a>
                                                                                                                         H
                                                                                                                         I
                                                                                                                         n
                                                                                                                         n
                                                                                                                         a
                                                                                                                         in
                                                                                                                         m in
                                                                                                                         in n
                                                                                                                         n I
                                                                                                                         n n
                                                                                                                         d 2
                                                                                                                         1
                                                                                                                         n
PI
r  —
                                                                                                                              c
                                                                                                                           2 •"
PI  <->

H

II
0  0
"2
                                                                                                                         z
                                                                                                                         PI
                                                                                                                         D

                                                                                                                         H
                                                                                                                         FI
                                                                                                                         PI

                                                                                                                         I
ro — »• as
gto O5
en ro
O O 3^-
7 o>
^ - _
o o — >•
— 1 *>. O




000
CO CO CO
ro co en
— *• oo en




pop
•~J O3 O)
CO CO *»
CO CO O5

o o o
'-*i co en
— >• ro j^.
OO ** CO

-^1 O CO

co — ». oo en co
o en o o o
0 ° 7f *r TT

 ~j ^
j*. bo co ro co
en — * ^j co 4x




p p p p p
OO CO OO CO CD
CO 2 5 3 CD




P r' r1- r' P
CO C> "~J O O
co ro — J CD o^
ro ro
CO
p p ^ p p
gco o> co o
CO CO OO CO
4»- tn CD -~<
CJI

1^. o ^- ro 05
O CO -^ CO O>

•5? -a
™ S
x 3
1
m
z
m
33
O
O
P
g 5
=o >
a "°
o
m
O
< m
O 3J
X ^
m
CL
3
70
o

o

I—
3 S
S 5
a >
o
m
O
^ m
X 5
1
                                                r
                                                n
                                                c
                                                n
                                                o
                                                ni
                                                Z
                                                >
                                                2
                                                n
                                                r

                                                0
                                                D
                                                in
                                                ni
                                                in  _,

                                                n  >
                                                n  u
   m

-------
                                           FIGURE 3:
                             ANGULAR DEPENDENCE DF CDLDN DOSE,
                     THYROID DOSE AND ENAMEL DOSE FOR  1 25DKEV PHOTONS.
                                                Front
                                                          1250keV
                                 Left
                                                                Right
                                                 'I-0   -••- Cobn
                                                        r • Thyroid
                                                        3— Enam el
                                                Back
                                                          UnitG v/G y
&EPA
        VB
                                        RADIATION RISK ABBEBBMENT WORKSHOP PROCEEDINGS

-------
  DOSE CONVERSION COEFFICIENTS FOR HIGH-ENERGY RADIATIONS

  YUKIO SAKAMOTO, SHUICHI TSUDA,  OSAMU SATO, NOBUAKI YOSHIZAWA AND
  YASUHIRO YAMAQUCHI
      Yukio Sakamoto, Shuichi Tsud: Japan Atomic Energy Research Institute
      Osamu Sato, Nobuaki Yoshiza\va and Yasuhiro Yamaguchi: Mitsubishi Research Institute

  ABSTRACT

      The dose conversion  coefficients for  high-energy radiations are  indispensable for the
      shielding design of high-energy accelerator facilities and dose estimation against cosmic
      rays in  high altitude flight.  But there were no data of dose conversion coefficients for
      photons above 10 MeV and for neutrons above 180 MeV in the recent ICRP Publication
      74.  For photons,  neutrons and protons up to 10 GeV and electrons up to 100 GeV, the
      absorbed dose to tissues and organs were calculated with Mote Carlo transport code system
      HERMES in conjunction with a MIRD-5 type anthropomorphic phantom, and the effective
      dose was evaluated by applying radiation weighting factors and tissue weighting factors.
      The effective dose equivalent was also evaluated by conventionally used quality factors.

      At the same time, the Istituto Nazionale di Fisica Nucleare (INFN) group in Italy has been
      evaluating the effective dose and ambient dose equivalent with FLUKA code system. The
      effective dose conversion coefficients  for photons and  electrons  above  10 MeV  were
      almost same between two results. The effective dose  for neutrons below 200 MeV was
      almost same between them, but there was maximum difference in the energy region from 1
      GeV to 10 GeV by a factor of 2.  The effective dose for protons in the energy range from
      50 MeV to 10  GeV was also almost same between two results.   From the comparison
      between effective dose and  effective dose equivalent for neutrons and protons, it was
      proven that the radiation weighting factors proposed for high-energy neutrons and protons
      were overestimated from a viewpoint of effective quality factor.

      New data of dose  conversion coefficients for high-energy radiations are going to be  more
      important for shielding design and dose evaluation in future construction of high-energy
      accelerator facilities.

  INTRODUCTION

      The dose conversion coefficients for high-energy radiations are needed in shielding designs
      of accelerator facilities and in dose estimation of cosmic rays in space missions and high
      altitude flight. In ICRP publication 51(1), there were dose conversion coefficients data for
      high-energy photons, electrons, positrons, neutrons, protons, pions  and muons.  In ICRP
      1990  recommendations (ICRP publication 60(2)),  a new concept of effective  dose was
      introduced by using radiation-weighting factors, and the tissue weighting factors and Q-L
      relationship were changed. There were no data of dose conversion coefficients for high-
      energy radiations based on ICRP publication 60 at the  early 1990s.  So the evaluation of
      dose conversion coefficients was  started for high-energy photons, electrons, neutrons and
      protons based on ICRP publication 60.
                                                                                      &EPA
RADIATION RISK ASSESSMENT WORKSHOP PRDCEEDINBS                                79   f*&'L

-------
              As protection doses for the human body, there are two kinds of dose, effective dose (E) and
              effective dose equivalent (HE), defined by following formulas:

                                           E= 2 wTHT= 2 WT E WR
              Where wR and DT,R are the radiation weighting factor and absorbed dose in tissue T for
              specific radiation, ^T and DT are the averaged quality factor and absorbed dose in tissue T,
                                                                   T_T
              wT is the tissue weighting factors for tissue T,  HT and   Tare equivalent dose and dose
              equivalent of tissue.

              The absorbed doses in each tissue and organ are calculated with a mathematical phantom
              model  and radiation transport code.   For the  lower energy  radiations, the  difference
              between the effective dose and the effective dose equivalent is very small.  The difference
              between the two doses has a great interest for high-energy radiations.

              As the operational quantities for measurement, there is the ambient dose equivalent defined
              in ICRU sphere and slab phantom.  The difference between the effective dose and the
              ambient dose equivalent also has a great interest for high-energy radiations.

           STATUS OF DOSE CONVERSION COEFFICIENTS

              The status of dose conversion coefficients for high-energy radiations is shown in Table 1.
              In ICRP publication 51, the ambient  dose  equivalents in the slab phantom with 30 cm
              thickness, 1 cm depth dose equivalent and maximum dose  equivalent, were  cited.  The
              upper energies were 20 GeV for photons and electrons, and 100 GeV for neutrons and
              protons. These data were based on old Q-L relations.   In ICRP publication 74(3), the
              effective doses based on ICRP publication 60 were  cited, but these data were limited below
              10 MeV for photons and electrons, and 180 MeV for neutrons. There were  no data for
              protons.

              In Japan, the effective doses were evaluated for photons'4', neutrons and protons'5"6' up to
              10 GeV and for electrons'7' up to 100 GeV with HERMES code system'8'.  The effective
              dose equivalents'5'6' were also  evaluated by using  same tissue weighting factors.  INFN,
              Italian group has evaluated the effective doses'9'  for photons and electrons up to 100 GeV,
              and for neutrons and protons up to 10 TeV with FLUKA code system'10'.  They evaluated
              also the ambient  dose equivalents'9'.  IHEP group of Russia evaluated the ambient dose
              equivalents'11'  for neutrons with HADRON code'12'.  Recently, Georgia  Tech. Group of
              USA evaluated the effective dose'13' for photons and neutrons with MCNPX code'14'.

           CALCULATION METHOD

              The component of HERMES code system established by  KFA is shown in Figure 1.  The
              hadrons cascade code, HETC-kFA2 simulates the behaviors of neutrons, protons, ions with
              mass heavier  than  10, pions, muons and residual nuclides.  The  behaviors  of neutrons
              below 15 MeV and secondary photons are simulated with MORSE-CG code. NDEM code
              calculates the  photon spectra from de-excitation  of excited residual nuclei.  The behaviors
              of electrons, positrons  and photons are simulated  with electro-magnetic cascade code
              EGS4(15).  For the evaluation of dose equivalents, the quality  factor database of secondary
              charged particles was developed for a wide  energy range  and kerma factors weighted with


&EPA
         BO                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      quality factors were developed for neutrons below 15 MeV. As the mathematical phantom
      model, the MIRD-5 type phantom(16) was used.

      Averaged quality factors for pions, protons, a and 16O charged particles based on new Q-L
      relationship*2' up to  100 GeV are shown in Figure 2.   Averaged quality factors were
      obtained by the averaging of quality factors from incident energy to stoppage energy. The
      curves for pions  and protons have single peak, and the curves of heavy ions have two
      peaks. The Q-L relationship and stopping powers of charged particles against energies are
      shown in Figure 3. The maximum stopping power of pions and protons is smaller than 100
      keV per micrometer. So these particles give only one peak. In the case of heavy ions with
      maximum  stopping power  over  100 keV per micrometer, one  peak  corresponds to the
      maximum value of Q-L relationship and the other peak corresponds to the maximum of
      linear energy transfer.

      The quality factors are defined by the final charged particles, which deposit the energy into
      the human  tissues and organs.  On the  other hand, the radiation weighting factors are-
      defined by the incident radiation,  itself.

  DOSE CONVERSION COEFFICIENTS F~OR PHOTONS

      In Figure 4, the effective dose per  unit photon fluency at Anterior-Posterior irradiation,
      front irradiation is shown. The filled circles and triangles give the HERMES code results(4)
      and FLUKA code results(9), respectively.  Two results gave almost same behaviors. These
      results included the effect of electron transport.  As the electrons produced by the high-
      energy photons in the human body penetrated tissues and organs, effective dose approached
      the  constant above  1  GeV.   In Figure 4, open  circles give the  results with kerma
      approximation, that was no electron transport and the  energy of electrons and positrons was
      deposited in the  vicinity of collided point.  It was proved that the results with kerma
      approximation overestimated the results including the electron transport above 50 MeV.  In
      the low energy, AP irradiation gave the maximum dose among irradiations. As the energy
      increase, the maximum dose was shifted to Posterior-Anterior and Lateral geometries.

      Figure 5  shows the ambient dose  equivalent per unit photon fluency at each depth of ICRU
      sphere with maximum effective dose.  Square symbols give the  maximum effective doses
      among irradiation geometries.  It was proved that the 1 cm depth dose equivalent was not
      the proper operational quantity for high-energy photons, and the 15 cm or 20 cm depth dose
      equivalents were very similar to the maximum effective dose.

  DOSE CONVERSION COEFFICIENTS FOR ELECTRONS

      Figure 6  shows the effective dose per unit electron fluency  at AP irradiation. Filled circles
      and  triangles give the results calculated with  HERMES code(7) and FLUKA code(9),
      respectively. Two results gave almost same behaviors and effective dose approached the
      constant  above 50 MeV.  As the electrons and positrons in the human body occurred the
      energy deposition, the effective dose per unit electron fluency was greater than that per unit
      photon fluency from the viewpoint of efficiency of electron production.

      For the very high-energy electrons, the contribution of hadrons cascades to dose was large
      with the  contribution of electromagnetic cascades.  Hadrons such as neutrons and protons
      were produced by the photonuclear reaction, for example, (y, n) reaction.  The degree  of
      secondary particle contribution to absorbed dose was estimated  to be about 1  % and that
      contribution to dose equivalent was estimated to be about 5 %.

                                                                                        &EPA
RADIATION  RISK ASSESSMENT WORKSHOP PROCEEDINQB                                a i   .^S L

-------
           DOSE CONVERSION COEFFICIENTS FOR NEUTRONS

              Figure 7 shows effective doses and effective dose equivalents per unit neutron fluency at
              AP irradiation.  Filled circles  and triangles give  effective doses  of HERMES code's
              results'5'6' and FLUKA code's results'9', respectively.  The two results gave almost same
              behaviors below 500 MeV, but there was some difference between the two results above 1
              GeV. This was  caused by  the  difference  of cross section data.  Open circles give the
              effective dose equivalents per unit neutron fluency at AP irradiation.  There was large
              difference between effective dose and effective dose  equivalent. As the absorbed doses and
              tissue weighting factors of each tissue were same ones, the difference was caused by the
              difference between radiation weighting factors and averaged quality factors.

              Quality factors averaged over body'5'6' and radiation weighting factors'2' for neutrons are
              shown in Figure 8.  Symbols of filled circles, open circles and boxes give quality factors
              averaged over body for AP,  PA and ISO irradiations. Lines including the broken lines give
              radiation-weighting factors cited in ICRP publication  60.  From the comparison between
              quality factors averaged over body and radiation weighting factors, the latter was about 30
              % overestimated for neutrons above 100 MeV.

           DOSE CONVERSION COEFFICIENTS FOR PROTONS

              Figure 9 shows effective doses and effective dose equivalents per unit proton fluency at AP
              irradiation.  Filled circles and triangles give effective doses of HERMES code's results'5'6'
              and FLUKA code's  results'9', respectively.  HERMES code's  results were  greater than
              FLUKA code's results below 50 MeV and from 1 GeV to 2 GeV regions. This was caused
              by the difference of cross section data as same as neutron case.  Open circles give the
              effective dose equivalents of HERMES code's results'5'6'.  There was also large difference
              between effective dose and effective dose equivalent.

              Quality factors averaged over body and radiation-weighting factors'2'  for protons are shown
              in  Figure  10.   Symbols of filled 'circles, open circles and boxes give quality factors
              averaged over body'5'6' for AP, PA and ISO irradiations obtained with HERMES code'5'6'.
              Line gives the radiation-weighting factor (wry=5) cited in ICRP publication 60.  Two types
              of triangles give averaged quality factors at 1cm depth and maximum dose positions of
              ICRU sphere with FLUKA code'9'.

              Averaged quality factors gave almost same behaviors between HERMES code and FLUKA
              code  calculations.   The radiation  weighting factors were larger than quality  factors
              averaged over body by a factor of 2.5 above 100 MeV protons.

           SUMMARY

              A new data set of dose conversion coefficients based on ICRP 1990 recommendations for
              high-energy photons, electrons, neutrons and protons was evaluated by using HERMES
              code  system and the MIRD-5 type phantom.  HERMES code's results were  almost same
              results calculated with  FLUKA code systems but there were some differences in neutron
              and proton doses caused by the differences of cross section data.   From the  comparison
              between effective doses  and  effective dose equivalents for neutrons and protons, it  was
              proved that radiation weighting  factors for neutrons and protons cited in ICRP publication
              60 were larger than quality factors averaged over body.

              In the OECD/NEA SATIF group, accelerator shielding task group, cross section data and
              absorbed dose has been compared'17'.


oEPA
 ,.£:sL.   82                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  REFERENCES
  1.   International Commission on Radiological Protection, "Data for Use in Protection Against
      External Radiations: ICRP Publication 51", Ann. ICRP 17, No.2/3 (1987).
  2.   International Commission on Radiological  Protection, "1990 Recommendations of the
      International Commission on Radiological Protection: ICRP Publication 60", Ann. ICRP
      21 (1-3) (Oxford: Pergamon) (1991).
  3.   International Commission on Radiological Protection, "Conversion Coefficients for Use in
      Radiological Protection Against External Radiation: ICRP Publication 74", Ann. ICRP 26
      (3/4) (Oxford: Elsevier Science) (1996).
  4.   Sato, O., Yoshizawa, N., Takagi, S., Iwai, S., Uehara, T., Sakamoto, Y., Yamaguchi, Y.
      and Tanaka, S., "Calculations of Effective Dose and Ambient Dose Equivalent Conversion
      Coefficients for High Energy Photons", J. Nucl. Sci. Technol., 36, 977 (1999).
  5.   Yoshizawa, N.,  Sato, O., Takagi, S., Furihata, S., Iwai,  S., Uehara, T., Tanaka, S. and
      Sakamoto, Y., "External Radiation Conversion  Coefficients using Radiation Weighting
      Factor and Quality Factor for Neutron and Proton from 20 MeV to 10 GeV", J. Nucl. Sci.
      Technol., 35, 928(1998).
  6.   Yoshizawa, N., Sato, O., Takagi, S., Furihata, S., Funabiki, J., Iwai, S., Uehara, T., Tanaka,
      S. and Sakamoto, Y., "Fluence to Dose Conversion Coefficients for High-Energy Neutron,
      Proton and Alpha Particles", J. Nucl. Sci. Technol., Supplement 1, 865 (2000).
  7.   Tsuda, S., Endo, A., Yamaguchi, Y. and Sato, O., "Fluency to Effective Dose Conversion
      Coefficients for Electrons from  1 MeV to 100 GeV", Radiat. Prot. Dosim. 95, 5 (2001).
  8.   Cloth, P., Filges, D., Neef, E.D., Sterzenbach, G., Reul, Ch., Armstrong, T.W., Colborn,
      B.L., Anders, B. and Briickmann,  H., "HERMES: A Monte Carlo Program System for
      Beam-Materials Interaction Studies", Jiil-2203 (1988).
  9.   Pelliccioni, M.,  "Overview of Fluence-to-Effective Dose and  Fluence-to-Ambient Dose
      Equivalent Conversion  Coefficients for High Energy Radiation Calculated  Using the
      FLUKA Code", Radiat. Prot. Dosim. 88, 279  (2000).
  10.  Fasso, A., Ferrari, A., Ranft, J. and Sala, P.R., "New Developments  in FLUKA Modeling of
      Hadronic and EM Inter actions", in Proc. of the Third Workshop on Simulating Accelerator
      Radiation Environments (SARE-3), Tsukuba (Japan), Hirayama, H. ed., KEK Proceedings
      97-5,32(1997).
  11.  Sannikov, A.V. and Savitskaya, E.N., "Ambient Dose  Equivalent Conversion Factors for
      High Energy Neutrons on the ICRP-60 Recommendations", Radiat. Prot. Dosim. 70, 383
      (1997).
  12.  Savitskaya, E.N. and Sannikov, A.V., "High Energy Neutron and  Proton Kerma Factors
     for Different Elements", Radiat. Prot. Dosim. 60, 135 (1995).
  13.  Sutton, M.R., Hertel,  N.E.  and Wtears, L.S,  "Fluence-to-Effective Dose  Conversion
      Coefficients for High-Energy Radiations Calculated -with MCNPX", in Proc. of the  Fifth
      Meeting of the Task Force on Shielding Aspects  of Accelerators,  Tragets and Irradiation
      Facilities (SATIF-5), Paris (France), July 2000, OECD/NEA Nuclear Science Documents,
      297(2001).
  14.  Hughes, H.G., Prael, R.E. and Little, R.C., "MCNPX-  The LAHET/MCNP Code Merger",
      Los Alamos National Laboratory, XTM-RN(u) 92-012 (1997).
  15.  Nelson, W. R., Hirayama, H. and Rogers, D.  W. O., "The EGS4 Code System", SLAC-265
      (1985).
  16.  Yamaguchi, Y.,  "DEEP Code to Calculate Dose Equivalents  in Human  Phantom for
      External Photon Exposure by Monte Carlo", JAERI-M 90-235 (1990).
  17.  Yoshizawa, N., Sakamoto, Y., Iwai, S.  and  Hirayama, H., "Benchmark  Calculation with
      Simple Phantom for Neutron Dosimetry", in  Proc. of the Fifth Meeting of the Task Force
      on Shielding Aspects of Accelerators, Targets  and Irradiation Facilities (SATIF-5), Paris
      (France), July 2000, OECD/NEA Nuclear Science Documents, 253 (2001).

                                                                                       oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                B3  ,4^>-~
                                                                                        S£.MW»>'\

-------
                                      TABLE  1 :
                            STATUS DF DOSE  CONVERSION
                    COEFFICIENTS FOR HIGH-ENERGY RADIATIONS.
RADIATION
PHOTONS
ELECTRONS
NEUTRONS
PROTONS

ICRP 51<1>
H*(10),HMAx
<20GeV
<100GeV
slab phantom
ICRP74P)
E
<0.01GeV
<0.18GeV
-
H'(10)
JAERl/MRK4.5'6'7)
E
<10GeV
<100GeV
<10GeV
HE
(HERMES code)
INFN<9>
E
<100GeV
<10,OOOGeV
H*(10),Hmax
(FLLIKAcode)
       Others: IHEP(HADRON code, n, H"'(10))"", Georgia Tech. (MCNPXcode, r, «, E)"3'
                                     FIGURE 1 :
          ORGANIZATION OF HERMES CODE SYSTEM AND PHANTOM MODEL.
 71, Jl
   e,e ,y
HETC-kFA2 code


\
exited
residual
, nuclei
[NDElVi


photon
            V
                               n(<15MeV)
                               Y
     V
        deexcitationW
EGS4 |  |MORSE-CG|  /m
           QKERMAl
                                                                Lower large intestine
84
                                  RADIATION RISK ABBEBSMENT WORKSHOP' PROCEEDINGS

-------
  0

  »
  i
  5
  s
  n
  z •
  H

  s.
  0


  i
  o
  R
  0
  z
                    E (pSvcm2 per photon cm"2)
                                                                    Charged Particle Energy (MeV)
                                                                                                                       Average Quality Factor
                                            O   -
PI

•n
m
n
m

o 3
o n
w c
m 71
•n m
D J,
                                                     I
                                                     D

                                                     D
                                                     Z
                                                     to
                                                 r
                                                 PI
                                                 n
                                                 I
                                                0
                                                TJ
                                                >  •*]


                                                M

                                                5S
                                                in  10
                                                7.
                                                0

                                                a
                                                 i
                                                r
                                                                                                     ni

                                                                                                     5
                                                                                                     a
                                                                                                     z
                                                                                 NJ
                                                                                 o
CO
o
                                                       51
                                                       5?
                                                       s§
                                                                         Quality Factor (Q)
    n n
    2 c
                                                                                               m
                                                                                               0
                                                                                               n
                                                                                               r
                                                                                               m
  ID
  (n;
     «>
(7 '

-------
o
0>
                  Dose
(pSvcm2 per neutron cm" )
f

0
0
n

n

n

H


D
o o o
_i, N> W -**-



So



Q
w





*
f~\
• >
•**_J ~ l$" 1
i : IliJ
?? S
4*
vx.
*« *»4
™ 9^ ^ —
•. %
* .
* w *
• \ i ^
. • * .
: 6 M :
O *• « liijn
jT" ^"
i ^ .


0
0
in
m
M -n
3 o
/D
Z m
N i
C
H "
TO
a
z
U1


                                                               E (pSvcm2 per electron cm"2)
                                                          So
                                                •>•
                                                        '.
                                                                                             n
n
n


n

D

in
n
•n
D
             Dose (pSvcm per photon cm"


                  o      o       o
                   _i       M       W
                                                                         3   5
                                                                               ^
                                                                                             n
                                                                                             r
                                                                                             n
0
Z
in
                                                                                                      Oa>
R
Z
H

0
D
in
n

R
€
C
>s
r "
ni ui
a
TO

•o
i
D
H
D
Z
in
H

s
n
n
n
a
z
D
a

-------
                                      FIGURE B:
                     AVERAGED QUALITY FACTORS FDR NEUTRONS


o:
5
V-
0
O
73

-------
                ORNL - JAERI COLLABORATION: NUCLEAR DECAY DATA

           KEITH F.  ECKERMAN AND AKIRA ENDO
              Oak Ridge National Laboratory
              Japan Atomic Energy Research Institute

           ABSTRACT

              Oak Ridge National Laboratory (ORNL)  and Japan Atomic  Energy Research Institute
              (JAERI) have begun a collaborative effort to provide an updated compilation of nuclear
              decay data for radionuclicies  of importance in occupational, environmental, and medical
              radiation dosimetry.  Information on the energies and intensities of the emitted radiations is
              the starting point of any evaluation of the radiation dose and health risk from the intakes of
              radionuclides.   This paper briefly reviews  the existing  compilations that  have been
              designed specifically for the dosimetrist and discusses ORNL and JAERI's experience in
              creating and maintaining such compilations.  The collaboration will result in an update of
              the  International  Commission on  Radiological  Protection  (ICRP)  compilation  of
              Publication 38.

           INTRODUCTION

              In 1969, the Medical Internal Radiation Dose (MIRD) Committee of the Society of Nuclear
              Medicine published the first  compilation of decay  data specifically designed for use by
              dosimetrists.  That publication, MIRD  Pamphlet 4 (1), was prepared at ORNL by L.T.
              Dillman.  The  methods documented in Pamphlet  4  were a precursor to the EDISTR
              computer code authored by Dillman (2). Initially, EDISTR's input values were based on a
              review of the basic nuclear literature, but later versions of EDISTR used the Evaluated
              Nuclear Structure Data File (ENSDF) data sets of the US DOE Nuclear Data Project (3).
              EDISTR calculates both the nuclear emissions and those associated with relaxation of the
              atomic electron structure. The EDISTR code was used to prepare Publication 38 of the
              International Commission on Radiological Protection (ICRP) (4).

              For the most part, the dosimetrist's needs have been  served by information on the unique or
              average energy of the emitted radiations.   However, there is  increasing interest in the
              spatial  distribution of the dose to cells at risk; particularly the depth dose into the epithelial
              structures  of the airways of the lung, stomach, colon and urinary bladder.  Calculation of
              the dose to cells at risk in the wall of these organs requires detailed information on the
              emitted radiations, including the beta spectra. In many instances the current compilations
              are not sufficiently detailed to provide the appropriate  level of spatial  resolution of the
              absorbed dose.

              About  3,340 radionuclides are known to exist with about 1,500 having half-lives of greater
              than a  minute.  ICRP Committee 2, in its Publication 30 (5), only considered the intake of
              radionuclides of half-life greater than 10 minutes —  about 1020 radionuclides.
oEPA
         88                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  CURRENT COMPILATIONS

      Two compilations are currently available. Publication 38 of the International Commission
      on Radiological  Protection (ICRP) tabulates data for about  825 radionuclides. That
      publication has been the basis for all nuclide-specific  dose coefficients published  by the
      ICRP during the past twenty-plus years. Nuclides were included in Publication 38  if they
      had a half-life of 10 minutes or greater or were a member of the decay chain of a parent
      nuclide of half-life of 10 minutes or greater. In addition, it was required that the ENSDF
      data set for the nuclide be sufficient to yield a valid tabulation of the emitted radiations.  If
      the ENSDF data set did not pass EDISTR's energy balance test, that radionuclide was not
      included in the publication.  As a result, only 825 of the potential  1020 radionuclides
      survived the process. The tabulations of Publication 38, in some instances, were abridged to
      limit the size of the publication.  Unabridged data were made  available in electronic form,
      including the beta spectra (6).

      Many of the ENSDF data sets used in preparing Publication 38 were based on the literature
      prior to 1970.  ENSDF data sets are generally updated on a  cycle  of about six years, so
      these data sets have been updated to reflect the later literature.

      The second compilation, DECDC (7), was prepared by JAERI in 2000 and addresses 1,027
      radionuclides, including all the nuclides of Publication  38. In preparing that compilation,
      JAERI used a modified version of EDISTR and critically reviewed each ENSDF data set.
      The data sets of the 1997 vintage were modified as necessary to satisfy the criteria of the
      computer code.  The compilation is available electronically from  the  Radiation  Safety
      Information Center (RSIC) at ORNL and from the Nuclear Energy Agency (NEA).

      JAERI's review process began with a critical examination of the ENSDF data set for each
      nuclide.  Where necessary,  corrections were made to the half-life, spin, and parities of the
      levels, and to the Q values  for each decay mode (8). The revised files were processed by
      EDISTR. If an acceptable energy balance was  indicated for a nuclide, the  input was
      considered adequate and EDISTR's  output was used in the  compilation.  If the energy
      balance was unacceptable, then a further review was undertaken including an examination
      of the literature.  Additionally, some utility  codes were used to  derive estimates  of the
      forbiddingness of beta spectra. The general flow of the analysis is shown in Figure 1.

  OBSERVATIONS

      Comparison of the DECDC and  ICRP Publication 38 compilations provides  some insight
      into the  advances made  in the primary experimental  literature.  Somewhat unexpected
      differences were noted in the half-lives assigned to long-lived radionuclides; see Table  1.
      Selenium-79 of Table 1 is an example of a very long-lived radionuclide whose half-life, it
      appears, may be in error in either compilation.  Although errors in the half-life for such
      long- lived radionuclides are of little concern  in the calculation of committed dose
      coefficients, is they are, nevertheless, an issue with respect to waste disposal.

      Table 2 compares the half-lives for some radionuclides of general importance with  values
      reported by the US National Institute of Science and Technology (NIST)(9).  The  values
      assigned in either the ICRP38  or DECDC compilations are in reasonable agreement with
      the reported NIST values.
                                                                                        oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                B9

-------
     The total emitted energy is of great importance in the dosimetry, although its distribution
     among the emitted radiations is also of importance. EDISTR relies on the Q value for the
     decay mode and partitions the available energy among the emitted radiations.  For some
     radionuclides, substantial differences  from source  to  source in the emitted energy  were
     noted. Table 3 compares these data for a few selected radionuclides for whom substantial
     differences were  indicated.  It  is evident that substantial  difference  in emitted energy
     (factors exceeding two) can arise, depending on the quality of the ENSDF data sets.

 COLLABORATIVE EFFORTS

     Before preparing a compilation to replace ICRP Publication 38, the methods contained
     within the EDISTR code will be reviewed.  That review is  currently underway.   The
     various data libraries used by EDISTR will be updated to embody the latest experimental
     and theoretical data; e.g., the yield of fluorescence x-rays.  In addition, EDISTR's default
     assignments, which are used when data are missing in the  ENSDF  data  set, will be
     critically evaluated.  It is expected that the past experience with EDISTR will guide any
     refinement in the default assignments.  It is also expected that EDISTR's treatment of the
     Auger electron cascade can be improved and extended to the outer atomic shells.

     Evaluated  or reference data of importance in the preparation of a new compilation are
     available from a number of organizations.  Data sets reviewed by these organizations will
     be considered in the preparation of the new compilation. Some of the organizations are:
         >  International Union of Pure and Applied Chemistry (IUPAC)
         >  International Commission on Radiological Units (ICRU)
         >  International Atomic Energy Agency (IAEA)
         >  National Institute for Science and Technology (NIST)

     The collaboration between ORNL and JAERI will result in a new compilation of nuclear
     decay data for use by ICRP Committee 2 in calculations following the next revision of the
     ICRP's primary recommendations.
go                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  REFERENCES
  1.  Dillman, L.T. Radionuclide Decay Schemes and Nuclear Parameters for Use in Radiation-
     Dose Estimation. MIRD Pamphlet No. 4, New York: The  Society of Nuclear Medicine,
     1969.
  2.  Dillman, L.T. EDISTR: A Computer Program to Obtain a  Nuclear Decay Data Base for
     Radiation Dosimetry. ORNL/TM-6689 (Oak Ridge National Laboratory, Oak Ridge, TN),
     1980.
  3.  Tuli, J.K. Evaluated Nuclear Structure Data File: A Manual for Preparing of Data Sets.
     BNL-NCS-51655-Rev. 87 (Brookhaven National Laboratory) 1987.
  4.  ICRP, Radionuclide Transformations: Energy and Intensity of Emissions. ICRP Publication
     38, Oxford:Pergamon Press), 1983.
  5.  ICRP. Limits for  Intakes  of Radionuclides by Workers.  ICRP Publication 30, Part  1
     (Oxford:Pergamon Press) 1979.
  6.  Eckerman, K.F., Westfall, R.J., Ryman, J.C., and Cristy, M.  Availability of Nuclear Decay
     Data in  Electronic Form, Including Beta  Spectra not Previously Published. Health Phys.
     67(4), 338-345, 1994.
  7.  Endo, A. and Yamaguchi, Y. Compilation of New Nuclear Decay Data Files Used for Dose
     Calculation, J. Nucl.  Sci. Technol. 38(8), 689-696, 2001.
  8.  Audi, G.,  Bersillon, O., Blachlot,  J, and  Wapstra, A.H.  The NUBASE Evaluation  of
     Nuclear and Decay Properties, Nucl. Phys. A624(l),  1-124, 1997.
  9.  Unterweger,  M.P., Hoppes, D.D., and Schima, FJ. New and  revised half-life measurements
     results, Nucl. Instrum. Meth. Phys. Res. A312, 349-352, 1992.
                                                                                     oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               91

-------
                                         TABLE  1 :
                        RADiniMUDLlDES WITH SIGNIFICANT DIFFERENCES IN
                        HALF-LIFE VALUES IN THE CURRENT COMPILATIONS.

NUCLIDE
Te-123
Fe-60
Se-79
Ag-108m
Sn-126
Hg-194
Tb-157
Si-32
Pb-202
HALF-LIFE
ICRP-38
10 Ty
100 ky
65 ky
127 y
100ky
260 y
150 y
450 y
300 ky
DECDC
600 Ty
1.5 My
650 ky
418 y
207 ky
444 y
71 y
172 y
52.5 ky
                                         TABLE Z:
                    COMPARISON or HALF-LIVES IN  CURRENT COMPILATIONS WITH
                             VALUES RECENTLY MEASURED AT NIST.

NUCLIDE
Na-24
Co-57
Zn-65
1-125
Cs-137
Th-228
ASSIGNED HALF-LIFE
ICRP-38
15h
270.9 d
243.9 d
60.1 d
30 y
1.9131 y
DECDC
1 4.959 h
271. 79 d
244.26 d
59.402 d
30.07 y
1.912 y
NIST
14.9513 + 0.0032 h
272.11+ 0.26 d
244. 164 + 0.099 d
59.49 ± 0.13d
30.157 ± 0.055 y
1.9127 ±0.0019 y
&EPA
        92
                                      RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                    TABLE 3:
      RADIONUCLIDEB WITH SIGNIFICANT DIFFERENCES IN TOTAL. EMITTED ENERGY.

NUCLIDE
Sr-80
Th-231
Fr-223
Pb-202
Ba-126
Tb-156m
TI-194
Zr-97
Os-180
Yb-162
lr-190n
Te-123
Ce-135
EMITTED ENERGY (MEV/NT)
ICRP-38
0.0135
0.191
0.459
0.00865
0.183
0.0495
0.809
0.880
0.0927
0.167
1.68
0.0261
2.02
DECDC
0.480
1.50
1.89
0.0116
0.605
0.0495
1.49
1.60
0.160
0.288
0.0868
0.00307
0.852
                                    FIGURE 1 :
         FLOWCHART OF THE DATA ANALYSIS USED IN THE PREPARATION OF THE
                   DECDC COMPILATION OF NUCLEAR DECAY DATA.
                        ENSDF
C                      Review Rasic     ^
                    Nuclear Properties  /
                      EDISTK   L
                    	Input File	)
                       EDISTR
(  Kvnluat
V^   input \
 te / Revise
t data file
                                     No
J)
                                                                                  oEPA
RADIATION RISK ASSEBSMENT WORKSHOP PROCEEDINGS
                                                                               93

-------
         SHIELDING  CALCULATION PARAMETERS FDR EFFECTIVE
                               Da BE EVALUATION

 YUKIO SAKAMOTO AND YASUHIRO YAM ABU CHI
     Japan Atomic Energy Research Institute

 ABSTRACT

     Dose  quantity in shielding design calculations has been  changed from  ambient dose
     equivalent  to effective dose  on  the  occasion of  the  introduction of International
     Commission on Radiological Protection (ICRP) 1990 Recommendations (ICRP Publication
     60) into domestic laws.  In shielding calculations for the radiation facilities, simple dose
     estimation methods by using shielding calculation parameters are effective and widely used
     instead  of  calculations of radiation energy spectra behind shielding materials.   These
     shielding calculation parameters depend on the dose quantity to be estimated and those for
     the evaluation of ambient dose equivalents  needed  to be replaced by  those for the
     evaluation  of effective dose,   hi  this  work, the shielding calculation parameters were
     evaluated for photons, neutrons and Bremsstrahlung from beta ray. The transmission data
     of photon  dose  have  been  calculated with  the standard  data of photon attenuation
     coefficients and gamma-ray buildup factors evaluated in American Nuclear Society, and
     the effective conversion coefficients from air kerma to effective dose evaluated with direct
     integration  radiation transport code BERMUDA.  The transmission  data of neutron dose
     have been calculated with one-dimensional discrete ordinate code ANISN.

     For mono-energetic photons  with  energies from 0.015 MeV to 10  MeV,  effective dose
     buildup factors, effective conversion coefficients from  air kerma to effective dose and
     transmission data  of effective dose were calculated.  Effective  dose rate constants, which
     represent an  effective dose value  at 1 m apart from  a source without shielding, and
     transmission data  of effective  dose were also calculated for gamma-rays and X-rays from
     33 radioisotopes, Bremsstrahlung from  13  beta-decay radioisotopes and 4 neutron sources.
     These data have been employed in "Calculation Manual for Shielding at Radiation Facility"
     and widely used in Japan.  These data will  also be applicable to  shielding calculations
     outside of Japan.

  INTRODUCTION

     This spring, the ICRP 1990 recommendations (ICRP Publication 60(1)) were adopted into
     the domestic laws related to radiation protection in Japan. Table 1 shows the items before
     and after the adoption of ICRP 1990 recommendations. As the evaluation dose in shielding
     design calculations, the ambient dose equivalent was changed to the effective dose.  For the
     dose conversion coefficient, data for Anterior Posterior irradiation were applied. Figure 1
     shows the dose conversion coefficients for photons and neutrons.  Solid and broken lines
     give the ambient dose equivalent cited in ICRP publication 51(2' and the effective dose for
     AP irradiation cited in ICRP publication 74(3), respectively.  Effective dose for photons is
     smaller than ambient dose equivalent by 10 % or more  in all photon energy. In the case of
     neutrons, the effective dose is  larger or  smaller than ambient dose equivalent depending on
     neutron energy.  The ambient dose equivalents of neutrons based on ICRP  publication 51
     were recommended to multiply the factor 2 by the statement from 1985 Paris meeting of
     ICRP(4>. In the effective doses based on ICRP publication 60, there is no need to multiply
     the factor 2. Dose limit for individuals was changed from 50 mSv per year to 100 mSv per
     5 years or not to exceed 50 mSv in any year. Setting criterion  of controlled area was also
94                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     changed from 0.3 mSv per week to 1.3 mSv per 3 months.  This dose rate is one third of
     that before  dose criterion.  The reduction of dose setting criterion near the boundary of
     controlled area needs recalculation of dose rate in terms of effective dose.

     In shielding calculations of many radioisotope handling facilities and radiation generation
     machines,  the  convenience  methods  were  applied  using  the  shielding calculation
     parameters, which corresponded to the evaluation doses,  hi the calculations of effective
     doses, the  shielding calculation parameters for effective  dose  are  needed.   So,  new
     shielding calculation parameter data set composed of gamma-ray buildup factors, effective
     dose from  bare source  and transmission data of radiations in shielding materials was
     calculated for mono-energetic photons, y-rays  and X-rays from radioisotopes, neutron
     source and Bremsstrahlung from p-rays.

     CALCULATION METHODS ar DOSE

     Calculation methods of doses are classified into sophisticated shielding transport codes and
     simple calculations with parameters. The calculations with radiation transport codes such
     as Monte Carlo codes and discrete ordinate codes, offer the energy spectra at evaluation
     points. The  dose (D) is obtained with energy  spectra (E)dE

     where E is radiation  energy.

     As the simple calculations with shielding calculation parameters for a point source, there
     are  two representative methods.   One is  the point kernel method  with the  usage  of
     attenuation  coefficients (^) and buildup factors (B), which is very popular with gamma-ray
     shielding calculations,

           D = S x RO x exp (-ut) x l/(4nf) x B (ut)

     where S  is source strength, RO is  the  dose conversion coefficient corresponding to the
     source energy, t is the thickness of shielding material, and r is the distance between the
     source  and evaluation  point.   The other   is  transmission  data  method represented by
     following formula:

           D = S x I7r2 x  T (t)

     where T is the dose rate  constant from bare  source  without shield and T  (t) is the
     transmission data of radiations in term of radiation dose.

     CALCULATION PARAMETERS FOR PHOTONS

     Table 2  shows the shielding calculation parameters for photons  and  the source of data.
     Mass attenuation coefficients and exposure buildup factors were based on the standard data
     in American Nuclear Society(5). In the database of buildup factors, the Japanese data were
     included  for  high  atomic  number elements.   For the  effective  dose evaluation,  the
     conversion  factors from air kerma to effective  dose  in the shielding materials were
     introduced.   This  data were calculated with direct integration shielding transport code,
     BERMUDA (6) developed in JAERI.  Photon emission data from radioisotope were based
     on ENSDF, Evaluated Nuclear Structure Data File (7), and DECDC (8'9).  DECDC is the
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                 95

-------
            photon data library from ENSDF for dose evaluation produced in JAERI.  The interpolated
            buildup factors were obtained with fitting function, GP formula<10), which was designed in
            Japan  and was adopted in standard data of ANSI/ANS (5) as fitting formula.  From these
            data, the dose rate constants from bare source (F), transmission data in shielding materials
            (T (t)) and effective dose buildup factors.

            Table 2 also shows the objective of photon sources and shielding materials. Photon sources
            were mono-energetic photons from 0.015 MeV to 10 MeV, gamma rays and X-rays from
            33 radioisotopes picked up from 18F to 241 Am at the first stage.  Transmission data were
            calculated for four shielding materials such as iron, lead, ordinary concrete and water.  In
            former manual (11), the density of ordinary concrete was 2.3 g/cm3.  In the view of costs, it
            is difficult to manufacture the new buildings in general radiation facilities with concrete of
            the density 2.3 g/cm3. From the survey results of building after Hanshin earthquake, the
            densities of concrete in general buildings  were assumed to be about 2.2 g/cm3.  If there is
            no certification of concrete density, the usage of 2.1 g/cm3 is recommended as the density
            for concrete with a margin (12).  In this work, this value was set for the density of ordinary
            concrete.

            Table  3 shows the ambient dose equivalent rate constants and effective dose rate constants
            (F) for five isotopes, 24Na, 60Co, 137Cs, 192Ir and 241Am. These data are corresponding to the
            dose rates  (uSv/h)  of ambient dose  equivalents  and effective doses at 1m apart from a
            source having the activity of 1 MBq. It was proven that effective dose  rate was smaller
            than ambient dose equivalent rate by about 15 % in many isotopes, and smaller than that by
            about 25 % for lower energy photons  emitted from 241Am.

            Figure 2 shows the transmission data of 60Co gamma ray against the thickness of four
            shielding materials from the view of effective dose. The lower scale gives the thickness of
            lead and concrete and the thickness of concrete and the upper scale gives water.  There was
            no difference of transmission data between effective dose and ambient dose equivalent.

            CALCULATION PARAMETERS FOR NEUTRONS

            Table  4 shows the shielding calculation parameters for neutrons and the source of data. The
            transmission data of  neutrons and  secondary  gamma rays  were  calculated  with  one-
            dimensional discrete ordinate code, ANISN-JR(13).  The cross section data of neutrons and
            photons were based on JENDL (Japanese Evaluated Nuclear Data Library) version 3.2(14)
            and PHOTX library (15).  Neutron sources picked up were the four types of spontaneous
            fission source of 252Cf, (a, n) source of 241Am-Be, d-T source and d-D source. As shielding
            materials, polyethylene and heavy concrete were added to 4 materials mentioned before.

            Table 5 shows the dose  rate constants for  ambient dose  equivalents and effective doses.
            These data were corresponded to the dose rates (pave/h) for ambient dose equivalents and
            effective doses at 1m apart  from a source emitting a  neutron per second.  The differences
            between effective dose rate constants and ambient dose equivalent rate constants are within
            about 15%.

            The transmission data of neutrons and secondary gamma rays from 252Cf source against
            thickness of concrete  shield are shown in Figure 3.  The axis of ordinate gives the dose
            rates multiplied by the 4Ttr2 where r is the distance from a point source  to the evaluation
            point.  In this case, neutron doses become smaller than  doses of secondary gamma rays
.f>,;,.    96                                 RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBB

-------
     beyond about 150 cm concrete thickness.  The differences among effective dose, ambient
     dose equivalents cited ICRP publication 51 and ICRP publication 74 were very small.

     CALCULATION PARAMETERS FOR BREMBBTRAHLUNB orfi-RAYS

     Table 6 shows the shielding calculation parameters for Bremsstrahlung of p-rays and the
     source of data. The Bremsstrahlung spectra from P-rays were calculated with BETABREM
     code (16) developed in JAERJ,  where the Bremsstrahlung spectra of  mono-energetic
     electrons were  based on Ward's  approximation formula (17).  The target of P-rays was
     assumed to calcium having  atomic number 20  and acryl.  The  transmission data of
     Bremsstrahlung were calculated with same procedure with  y-ray  and  x-rays  from
     radioisotope.  As p-emitters, 13 radionuclides were picked up from 3H to 147Pm.  Shielding
     materials were same as photons.

  SUMMARY

     New  data  set  of shielding  calculation parameters  was calculated for effective dose
     evaluations on mono-energetic photons, y-ray  and x-rays  from radioisotopes, neutron
     sources and Bremsstrahlung  from P-emitters.   These data were summarized in JAERI-
     Data/Code<18) and cited in new manual(12>, and have been widely used in Japan.
                                                                                     &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               97    .»;:, L

-------
          REFERENCES
          1.  International Commission on Radiological  Protection, "1990  Recommendations of the
              International Commission on Radiological Protection: ICRP Publication 60", Ann. ICRP
              21 (1-3) (Oxford: Pergamon) (1991).
          2.  International Commission on Radiological Protection, "Data for Use in Protection Against
              External Radiations: ICRP Publication 51", Ann. ICRP 17, No.2/3 (1987).
          3.  International Commission on Radiological Protection, "Conversion Coefficients for Use in
              Radiological Protection Against External Radiation: ICRP Publication 74'", Ann. ICRP 26
              (3/4) (Oxford: Elsevier Science) (1996).
          4.  International Commission on Radiological  Protection, "Statement  from the 1985 Paris
              Meeting of the International Commission on Radiological Protection", Radiat. Prot. Dosim.
              11,134(1985).
          5.  ANS Standard Committee Working Group ANS-6.4.3, "American National Standards for
              Gamma-Ray Attenuation Coefficients and Buildup Factors  for Engineering Materials",
              ANS/ANS-6.4.3-1991 (1991).
          6.  Suzuki, T., Hasegawa, A., Tanaka, S. and Nakashima, H., "Development of BERMUDA:
              Radiation Transport Code System, Part II, Gamma-Ray Transport Codes", JAERI-M 93-
              143 (1993).
          7.  Maintained by the National Nuclear Data  Center at Brookhaven  National  Laboratory,
              "Evaluated Nuclear Structure Data File".
          8.  Endo, A., Tamura, T. and Yamaguchi, Y., "Compilation of Nuclear Decay Data Used for
              Dose Calculation:  Revised Data for Radionuclides Not Listed in ICRP Publication 38",
              JAERI-Data/Code 99-035 (1999).
          9.  Endo, A.  and Yamaguchi, Y., "Compilation  of Nuclear Decay  Data  Used  for Dose
              Calculation:  Revised Data for  Radionuclides Listed in ICRP Publication 38", JAERI-
              Data/Code 2001-004 (2001).
          10. Harima, Y., Sakamoto,  Y., Tanaka, S. and Kawai, M., "Validity of the Geometrical
              Progression Formula in Approximating Gamma-Ray Buildup Factors", Nucl. Sci. Eng., 94,
              24 (1986).
          11. "Calculation  Manual  for X-ray Shielding  at  Radiation  Facility",  Nuclear  Safety
              Technology Center, (1989).
          12. "Calculation Manual for Shielding at Radiation Facility, 2000", Nuclear Safety Technology
              Center, (2000).
          13. Koyama, K., Taji, Y. and Minami, K., "ANISN-JR, A One-Dimensional Discrete Ordinates
              Code for Neutrons and Gamma-Ray Transport Calculations", JAERI-M 6954 (1977).
          14. Nakagawa, T., Shibata, K., Chiba, S., and et  al., "Japanese Evaluated Nuclear Data Library
              Version 3 revision-2: JENDL-3.2", J. Nucl. Sci. Technol., 32,  1259 (1995).
          15. Trubey, O.K., Berger, M.J. and Hubbell, J.H.,  "Photon Cross Sections for ENDF/B-VI",
              Advanced in Nuclear Computation and Radiation Shielding, American  Nuclear Society
              Topical Meeting (1989).
          16. Sakamoto,  Y.,  Nakane, Y. and Tanaka, S.,  "Calculations  of Bremsstrahlung Photon
              Spectrum from Radioisotope (II)", Proceedings of the Fifth EGS4 Users' Meeting in Japan,
              1995, KEK Proceedings 95-9, 119 (1995).
          17. Wyard, S.J., Proc. Roy. Soc., A65, 377 (1952).
          18. Sakamoto,  Y.,  Endo, A., Tsuda,  S.,  Takahashi, F. and  Yamaguchi,  Y.,  "Shielding
              Calculation  Constants foe Use in Effective Dose Evaluation for Photons, Neutrons and
              Bremsstrahlung from Beta-ray", JAERI-Data/Code 2000-044 (2001)  (in Japanese).

&EPA
         98                               RADIATION  RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                  TABLE 1 :
                THE CHANGE DF EVALUATION ITEMS AFTER ADOPTION
                  DF ICRP 1 99O RECOMMENDATIONS IN JAPAN.
ITEM
EVALUATION DOSE
DOSE LIMIT FOR INDIVIDUALS
SETTING CRITERION OF CONTROLLED
BEFORE RECOMMENDATIONS
ambient dose equivalent, H*(10)
50 mSv/y
0.3 mSv/week
AFTER RECOMMENDATIONS
effective dose, E(AP)
100 mSv/5y, 50 mSv/y
1.3mSv/3month
                                  TABLE 2:
               SHIELDING CALCULATION PARAMETERS FCJR PHOTONS.
TYPE OF DATA
mass attenuation coefficients (\jJp)
exposure buildup factors (Bj
conversion factors from air Kerma to effective dose
photon emission data from Rl
Fitting function of buildup factors
dose rate from bare source(F)
transmission data T(t), effective dose buildup factors
PHOTON SOURCES
SHIELDING MATERIALS
SOURCE OF DATA
standard data of ANSI/ANS-6.4.3-1991
BERMUDA code (direct integration shielding transport)
ENSDF, DECDC
GP (Geometrical Progression) formula

monoenergetic photon: 0.015 - 10 MeV
radioisotope: 33 nuclides
18F, 2*Na, 51Cr, ^Mn, 59Fe, 56Co, 67Co, 60Co, MCu,
65Zn, 67Ga, «>Ge/68Ga, 75Se, 8W1lW, 85Kr, 86Sr,
"Mo/^Tc*, 99mTc, '03Pd/'03mRh, 110mAg, »1ln, 124Sb,
i23|_ 125(| 131^ i33Xe, 137CS*, 192lr, 198Au, 197Hg, ^Tl,
226Ra/daughter, 241Am
iron, lead, ordinary concrete (p=2.10 g cm-3), water
                                  TABLE 3:
       DOSE RATE CONSTANTS (F)FOR RADI O I SOTO PES. (UNIT!  SV
RADIO ISOTOPE
24Na
60Co
137Cs
i92|r
24tAm
H*(10)
0.492
0.354
0.0927
0.139
0.00529
E(AP)
0.429
0.305
0.0779
0.117
0.00395
E(AP)/H*(10)
0.87
0.86
0.84
0.84
0.75
                                                                             oEPA
RADIATION RISK ASSESSMENT WORKSHOP PRQCEEDINBB
                                                                         99

-------
                                          TABLE 4:
                      SHIELDING CALCULATION PARAMETERS FOR NEUTRONS.
TYPE OF DATA
TRANSMISSION DATA
NEUTRON SOURCES
SHIELDING MATERIALS
SOURCE OF DATA
ANISN code (one dimensional Sn code)
JENDL 3.21 PHOTX
(neutron/photon cross section data)
262Cf, 241Am-Be, d-T source, d-D
iron, lead, ordinary concrete, water, polyethylene, heavy concrete
                                          TABLE 5:
              DOSE RATE CONSTANTS FOR NEUTRON SOURCES (F). (UNIT! PS VH~'
                                                                            S)
NEUTRON SOURCE
252Cf
241Am-Be
d-T
d-D
H*(10)
9.78
10.6
14.9
10.4
E(AP)
10.0
11.9
14.2
11.7
E(AP)/H*(10)
1.02
1.12
0.95
1.13
                                          TABLE £>:
                SHIELDING CALCULATION PARAMETERS FOR BREMSSTRAHLUNG P-RAYS.
TYPE OF DATA
BREMSSTRAHLUNG SPECTRA
TRANSMISSION DATA
P-EMM1TERS
SHIELDING MATERIALS
SOURCE OF DATA
BETABREM code (Wyard's formula)
target: Ca(z=20), acryl
same procedure
as Rl y-rays and X-rays
13nuclides
3H, UC, 32P, 33P, 35S, 45Ca, 63Ni, »5Kr, ®Sr, *>Sr,
90Yi 90Sr/90Y, 147Pm
iron, lead, ordinary concrete, water
3-EPA
        1 QO

                                       RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                        FIGURE 1 :
                            DOSE CONVERSION COEFFICIENTS.
                    100
                 §
                 (D
                 &
                "M
                 £
                 o
                     10
                 0>
                 V)
                 8
  	H(10)|CRP51
                           1000
                        |   100
                        5
                        i   10
                        a.
                        V
                        •fl
                        5       iF       lo2
                         Fhoton Energy (M=V)
                              10"6   10'3    10°    103
                                 Neutron Energy
                                        FIGURE 2:
  TRANSMISSION OF 6DCo GAMMA  RAYS.

0        60        120       180       240
                 -
                 cb
                 E
                 E
                 c
                 o
                    ia3
                     -
                 I  10-
                 to
                                                                       300
                       Concrete
                                         V\tter
                               Lead
0         30         60        90        120
              Thickness of Iron and Lead (err)

                  FIGURE 3:
TRANSMISSION OF  25ZCF SOURCE NEUTRONS
                                                                        150
                   10-
                ,0,
                0)
                V  10
                   10-
                                            |g-252 Source Neutronj
                                                Ordinary Concrete
                                                     (ANISN)
           — H(10)|CRP74


            "H(10)|CRP51
                                50
                   100       150
                     Radius (cm)
200
250
                                                                                            &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                       101

-------
                 DEVELOPMENT OF CT VOXEL. PHANTOMS FOR JAPANESE

           KAORU SATO, HIRDSHI NQDUCHI, KIMIAKI SAITO, Y. EMOTD AND S. KOQA
              Kaoru Sato, Hiroshi Noguchi, Kimiaki Sato:  Japan Atomic Energy Research Institute
              Y. Emoto and S. Koga: Fujita Health University

           ABSTRACT

              For calculating doses due to radioactivity taken in a body,  specific absorbed fractions
              (SAFs) are used. So far MIRD type phantoms, which are described by simple geometry,
              have been used for the calculation of SAFs. In recent years, more realistic phantoms called
              voxel (volume pixel) phantoms  have been developed  on  the basis of computational
              tomography (CT) scans or magnetic resonance imaging (MRI) of actual persons. The voxel
              phantoms have begun to be used for calculating SAFs, since they can accurately describe
              sizes, shapes and locations of organs, which would affect SAFs.

              We are now developing Japanese adult voxel phantoms for internal dosimetry by using CT
              images. Until now, CT scans for three Japanese volunteers were performed under supine
              and upright positions to study the effect  of a body size  and posture  on  SAFs. They are
              healthy male adults and their body sizes range from small to large. The height and weight
              of the middle size man is almost coincident with the averages for Japanese adult male. So
              far the development of voxel phantom has been almost finished for the middle size man
              (voxel-phantom-MM). The voxel size is 0.98x0.98x1 mm3, which is expected to represent
              small or thin tissues precisely. Characteristics of the phantom, such as organ masses and
              shapes, were examined. It was found that even small size organs  such  as thyroid were
              realistically modeled. The  examination  showed that  voxel-phantom-MM had  realistic
              structure, which would enable us to calculate reliable SAFs.

           INTRODUCTION

              Specific absorbed  fractions (SAFs) are used for the assessment of dose coefficients for
              internal dosimetry. SAFs are defined as an absorbed fraction (AF)  divided by mass of a
              target organ. AF is the fraction of energy emitted by radioactivity in a source organ, which
              is  absorbed  to a  target organ. SAFs currently  used by International  Commission on
              Radiological Protection (ICRP) have been calculated on the basis  of a human phantom
              called the MIRD5 type phantom.  Snyder et al. developed the original MIRD5 phantom
              (1969,  1974). The organ masses of the MIRD5 phantom are determined in accordance with
              the ICRP Reference Man (ICRP 1975), which is based on Caucasian. The MIRD5 phantom
              is  mathematically  described by simple  equations such as elliptical  cone, ellipsoid and
              cylinder. Later, Cristy  (1980) modified to the shapes, locations, densities and chemical
              compositions of organs and tissues for the original MIRD5 phantom.  However, there is a
              concern that the MIRD5 type phantoms  might not provide  adequate SAFs, because the
              phantoms do not represent the organ shapes of actual persons.
oEPA
    *~~   1Q2                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     In recent years, more realistic human phantoms have become available on the basis of 3-D
     medical imaging techniques, such as computational tomography (CT) scans or magnetic
     resonance imaging (MRI) of actual persons. The organs and tissues of these phantoms are
     defined by sets of small rectangular block units called voxel (volume pixel). The phantom
     consisting of voxels is called a voxel phantom. The organ shapes of a voxel phantom can
     be modeled with high accuracy using a small voxel size. So far, several voxel phantoms
     have been developed. Characteristics of these voxel phantoms are summarized in Table 1,
     together  with the averages  of height  and weight  of Japanese adult males and females.
     BABY, CHILD (Zankl et al. 1988), Golem (Zankl and Wittmann 2001), Otoko (Saito et al.
     2001) and Onago (Kinase et al. 2001; Saito et al.) were constructed on the basis of CT
     unages.

     The CT images of BABY and CHILD were obtained from an 8-week-old baby and 7-year-
     old child, respectively (Zankl et al.  1988). The CT images of Otoko (Saito et al. 2001) and
     Golem (Zankl and Wittmann 2001) were derived from adult males. The Otoko and Onago
     phantoms are respectively the first Japanese male and female voxel phantoms (Kinase et al.
     2001; Saito  et al. 2001; Saito et al.). The height  and weight of Otoko  are close to the
     averages  of Japanese adult males (Tanaka and Kawamura 1996). The NORMAN phantom
     was derived from MRI of an adult male. The phantom was normalized to be 170 cm tall
     and to weigh 70 kg, the values of ICRP Reference Man  (Dimbylow 1996). The VIP-Man
     phantom, which has the finest resolution on voxel  size (0.33x0.33x1 mm3), was made on
     the basis  of color photographs of 1 mm-thick slices of a cadaver (Xu et al. 2000).

     A voxel  size affects the shapes and  volumes of  small or thin organs  and tissues. For
     example, the thyroid of the Otoko phantom is not realistically modeled compared with the
     anatomical structure for actual persons, because of the long vertical height (10 mm) of the
     voxel. The skin mass of Golem, of which voxel size is relatively large (2.08x2.08x8 mm3),
     is 4703 g (Zankl and Wittmann 2001), is about twice as heavy as that of ICRP reference
     man (ICRP  1975).  Therefore, we  have begun  developing precise human phantoms of
     Japanese with a smaller voxel size to calculate reliable SAFs. In this work, CT images for
     three Japanese  adult males with different sizes were taken hi supine and upright positions to
     study the effects of a body size and posture on SAFs. This report describes a construction
     method of a voxel phantom based on the CT images and characteristics of a voxel phantom
     (the voxel-phantom-MM) of which construction is almost completed.

  CONSTRUCTION OF VOXEL. PHANTOM

     SUBJECTS AND CT SCAN

     CT scans were carried out  for three healthy  Japanese adult male volunteers, who  have
     respectively  small (MS), middle (MM) and large (ML) body sizes, to  study the effect of
     body size. As shown in Table 2, the height and weight of the volunteer, MM, are almost the
     same as the  averages of Japanese adult males (Tanaka and Kawamura 1996). The height
     and weight of the volunteers, MS and ML, are respectively smaller and larger by about 1
     standard  deviation  (SD) than  the  Japanese  averages.  In addition  to  the  scans  hi  a
     conventional supine position, CT images were taken in an upright position, to examine the
     changes of organ shapes and locations from a supine position.

     The Ethics Committee of the University Hospital performed the whole-body CT scans in a
     supine position at Fujita Health University Hospital after the  approval for the plan and


                                                                                      oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              1 Q3

-------
              objectives. The CT images at 512 x 512-pixel resolution were taken using the helical CT
              scanner (Aquiline, Toshiba Medical Systems Co. Ltd.). This resolution gives a voxel size
              of 0.98x0.98x1 mm3. The scan image data were stored on digital audiotapes. On the other
              hand, the CT images in an upright position were taken using a cone-beam CT scanner
              (Hitachi Ltd.) under the  same conditions of the supine position in  terms of  the pixel
              resolution and slice thickness. Unfortunately the CT images in the upright, position have a
              lower contrast compared with conventional CT images. Besides the scan area at a time is
              limited to a  sphere of 25 cm diameter.  Thus the scans were performed at four different
              heights to cover the trunk area of the volunteers. The CT images were checked from the
              viewpoint of medical diagnosis and it was ensured that there were no problems for the
              construction of voxel phantoms.

              CONSTRUCTIVE METHOD OF VOXEL PHANTOM

              All tissues, organs and contents in gastrointestinal tracts related to dose calculation have to
              be segmented from CT images in order to construct a voxel phantom. The segmentation is
              carried out according to a method originally developed by Gesellschaft fur Strahlen und
              Umweltforschung (GSF) (Zankl and Wittmann 2001) and modified by Saito et al. (2001).
              Since image-processing techniques are required for the segmentation,  a commercial
              software called Visilog4 is used with a SGI (Silicon Graphics Inc.) O2 workstation. Firstly,
              soft tissue, lungs, adipose and bone with  different density are segmented on the basis of CT
              values to decide the tissue regions of each pixel on CT images. The "values represent the
              attenuation coefficients and the density of pixels corresponding to each tissue. Soft tissue,
              adipose, lungs and  bone can be easily segmented, because their CT values are much
              different from each other. On the other hand, organs in soft tissue regions are unable to be
              segmented by only the CT value data, because the CT values are very similar for most soft
              tissue organs. Therefore, the image processing techniques such as erosion,  dilation and
              filling holes are  used to segment the soft tissues. Finally, organ specific identification
              numbers are assigned to voxels belonging to each organ region, hi the present study, the
              above procedure has been begun with the middle size man. This phantom is temporarily
              named MM in this paper. So far identification numbers are assigned for about  100 regions
              of  organs and tissues  of the MM phantom. Different  identification numbers will be
              assigned to more than 120 regions the whole body MM phantom. Skin is assumed to be one
              voxel layer at the outer surface of the body. Since a skin thicknes; referred to  the ICRP
              reference man is  1.3 mm (ICRP 1975), the skin thickness of 0.98 mm in the MM phantom
              is close to the ICRP reference value.

              CHARACTERISTICS OF VOXEL PHANTOM IN SUPINE POSITION

              Until now the following 19 regions have been segmented from the original CT images of
              MM: brain, bronchi, colon, esophagus, eyes, eye lenses, gall bladder, heart, kidneys, lower
              large intestine,  lungs,  spleen, stomach,  small  intestine, thyroid, trachea,  upper large
              intestine, urinary bladder and cortical bone in skeleton. The segmentations of testis, thymus
              and skeleton containing trabecular bone and marrow cavity are not finished. Figure 1 shows
              the anterior views of selected organs of voxel-phantom-MM and the MIRD5 phantom. A
              comparison of the MM and MIRD5 organ shapes indicates that the thyroid and pancreas of
              MM are represented more realistically than those of the MIRD5 phantom, hi conclusion,
              the MM phantom realistically reproduces the organs and tissues of a Japanese adult male.

              Table 3  shows  some organ masses of voxel-phantom-MM,  Otoko (Saito et al. 2001),
              MIRD5 (Cristy and Eckerman 1987), VIP-Man (Xu et al. 2000) and Golem (Zankl  and

&EPA
         1 O4                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS

-------
     Wittmann 2001), along with the Japanese averages  (Tanaka 1992). The organ mass of
     voxel-phantom-MM is  calculated by multiplying the volume of each organ composed of
     voxel aggregate  with a soft tissue density (1.05 g/cm3) (Zankl and Wittmann 2001).  The
     masses  of the kidneys, liver, spleen and thyroid of voxel-phantom-MM, except for the
     brain, agree with the Japanese averages (Tanaka 1992) within 1 SD. In particular, even for
     small size organs such as spleen and thyroid, their masses are quite close to the averages of
     Japanese adult males. On the other hand, the masses of spleen and thyroid for the Otoko
     phantom are  about 50  % of the Japanese average. Saito et al. (2001) suggested that the
     large deviations  in spleen and thyroid masses between Otoko and Japanese averages were
     caused by individual variations.

     An advantage of a voxel phantom is that it can easily change the size of each organ by
     adding or deleting voxel layers to or from the outer surface of an organ. An example is
     shown in Figure 2. An adjustment of the liver mass, which was less than that of the
     Japanese average, was  tried. The correction was performed by addition of voxel layers at
     the outer surface of the organ phantom. As a result, it was found that the addition of three
     voxel layers makes the  mass close to that of the Japanese average. This result suggests that
     the mass correction method would be useful to examine the effects of organ sizes on SAFs.

     EFFECTS or POSTURE ON BHAPEB AND LOCATIONS OF ORGANS AND TISSUES

     CT images in an upright position were taken for the same volunteers to examine the effects
     of posture on SAFs, because there may be differences in the shapes  and locations of organs
     between the supine and upright positions.  It was observed that there are some differences in
     shapes or locations of the diaphragm, stomach, spine and lower  abdomen between supine
     and upright positions.  Figure 3 shows the CT  images of the volunteer, MS, at coronal
     sections through the stomach. The middle part of the stomach in  the upright position
     markedly  moves downward compared with that in the supine position. It was also found
     that the upper part of the stomach in the supine position moved toward the back. Figure 4
     shows the CT images  of the  volunteer,  MS, at sagittal sections through the spine.  The
     lumbar vertebra  in the  upright position moves forward compared with that in the supine
     positions. The inflation of the  lower abdomen  and the downward movement of the
    . diaphragm are observed in the  upright position (Figure  4). The gravity is obviously
     responsible for the changes  of the shapes and locations of the spine, stomach, diaphragm
     and lower abdomen.

  CONCLUSIONS

     In order to construct precise Japanese voxel phantoms based on medical images,  CT scans
     were performed  for three healthy Japanese adult male volunteers. The volunteers whose
     body sizes were  respectively small, middle and large were selected to study the effect of
     body sizes on specific absorbed fractions. To examine the influence of posture, CT images
     in an upright position were taken as well as in a conventional supine position. The
     resolution of the pixels of images is  0.98x0.98 mm and the slice thickness is 1  mm. The
     voxel size of 0.98x0.98x1 mm3 enables us to represent even small and complicated organs
     such as thyroid realistically.  It  was found that  the masses of  main  organ phantoms
     constructed from the CT images of the middle size man were within one sigma deviation of
     the averaged organ masses for Japanese adult males, except for the brain. The construction
     of a voxel phantom for a male with the average Japanese adult size is almost completed.
     The calculation of specific  absorbed fractions will be started  soon using the constructed
     voxel phantom


                                                                                       &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBB                               1 05

-------
 ACKNOWLEDGEMENTS

     The authors wish to thank Mr. J. Kuwabara of Japan Atomic Energy Research Institute for
     his helpful advice on the present work.

 REFERENCES

 Cristy, M (1980) Mathematical phantoms  representing children  of various ages for use in
     estimates of internal dose.  Oak  Ridge, TN: Oak Ridge National Laboratory; Annual
     Progress Report 5, ORNL/NUREG/TM-367.

 Cristy, M. and Eckerman, K. F. (1987) Specific absorbed fractions  of energy at various ages
     from internal photon sources. Oak Ridge, TN:  Oak Ridge National Laboratory; Report
     ORNL/TM-8381. Vol. 1-7.

 Dimbylow, P. J. (1996) The development of realistic voxel phantom for electromagnetic field
     dosimetry. Proc.  Int. Workshop on voxel  phantom development; National Radiological
     Protection Board Report pp. 1-7.

 ICRP  (1975) Publication 23.  Reference  man:  anatomical, physiological  and metabolic
     characteristics. International  Commission  on Radiological Protection. Pergamon  Press,
     Oxford.

 Kinase, S., Zankl, M., Kuwabara, J., Sato, K., Noguchi, H., Funabiki, J. and Saito, K. (2001)
     Evaluation of specific absorbed factions in voxel phantoms using Monte Carlo simulation.
     Submitted to Proceedings of Radiation Risk Assessment in the 21 the Century, EPA/JAERI
     Workshop, Las Vegas, November 5-7 (2001).
 Saito, K., Wittmann,  A., Koga,  S., Ida, Y., Kamei, J., Funabiki,  J. and Zankl, M. (2001)
     Construction of a computed topographic phantom for  a Japanese male  adult and dose
     calculation system. Radiat. Res. Biophys. 40, 69-76.
 Saito, K et al. The construction of a voxel  phantom based on CT data for a Japanese female
     adult (in preparation).

 Snyder, W. S., Ford,  M. R., Warner, G. G.  and Fisher, Jr. H. L. (1969) Estimates of absorbed
     fractions  for monoenergetic photon sources uniformly distributed in various organs of a
     heterogeneous phantom. MIRD Pamphlet No.5. J. Nucl. Med. 10, 7-52.
 Snyder, W. S., Ford, M. R. and Warner, G. G. and Watson, S. B. (1974) Estimates of absorbed
     fractions  for monoenergetic photon sources uniformly distributed in various organs of a
     heterogeneous phantom. Revision of MIRD Pamphlet No.5, ORNL-4979.
 Tanaka, G. (1992) NIRS-M-85. Reference Japanese Vol.1. Anatomical data.
 Tanaka, G., Kawamura, H. (1996) NIRS-M-115.  Anatomical and physiological characteristics
     for Asian Reference Man male and female of different ages.
 Xu, X. G., Chao, T. C. and Bozkurt, A. (2000) VIP-Man: An image-based whole-body adult
     male model constructed  from color photographs of the visible  human project for multi-
     particle Monte Carlo calculations. Health Phys. 78, 476-486.

 Zankl, M., Veit, R.,  Williams, G., Schneider, K.,  Fendel,  H., Petoussi, N. aind Drexler, G.
     (1988) The construction of computer topographic phantoms  and their  application in
     radiology and radiation protection. Radiat. Res. Biophys. 27, 153-164.
 Zankl, M. and  Wittmann, A. (2001) The adult male voxel model "Golem" segmented from
     whole-body CT patient data. Radiat. Res. Biophys. 40,  153-162.
i as                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                   TABLE 1 :
           CHARACTERISTICS OF VOXEL PHANTOMS PREVIOUSLY DEVELOPED.
VOXEL PHANTOM
BABY (Germany)
CHILD (Germany)
Golem (Germany)
NORMAN (UK)
VIP-Man (USA)
Otoko (Japan)
Onago (Japan)
Japanese male
average3'
Japanese female
average3'
ORIGINAL
IMAGE
DATA
CT
CT
CT
MRI
Photograph
CT
CT
	
	
VOXEL SIZE
(mmxmmxmm)
0.85x0.85x4
1.54x1.54x8
2.08x2.08x8
1.95x1.95x2.04
0.33x0.33x1
0.98x0.98x10
0.98x0.98x10
—
—
HEIGHT
(cm)
57
115
176
170
186
170
162
170±6"'
155±5b'
WEIGHT
(kg)
4
22
69
70
104
65
57
64±9b'
52±7»'
AGE
8 weeks
7 years
Adult
Adult
Adult
Adult
Adult
Adult
Adult
SEX
Female
Female
Male
Male
Male
Male
Female
Male
Female
    a)Tanaka and Kawamura (1996)
    b)mean±SD

                                   TABLE 2:
        BODY CHARACTERISTICS OF THREE JAPANESE ADULT MALE VOLUNTEERS.
VOLUNTEERS

MS
MM
ML
HEIGHT
(cm)
158
171
178
WEIGHT
(kg)
51
65
78
WIDTH OF
THE BODY
(cm)
43
48
52
FRONT TO BACK
(cm)
20
22
25
                                   TABLE 3:
     ORGAN MASSES OF VOXEL-PHANTOM-MM, DTOKO, MIRD5, VIP-MAN, GOLEM
                         AND THE JAPANESE AVERAGES
ORGANS
BRAIN
KIDNEYS
LIVER
PANCREAS
SPLEEN
THYROID
ORGAN MASS (o)
JAPANESE AVERAGE a>
MEAN
1462
319
1569
128
141
19
SD
115
38
329
35
50
5
VOXEL-
PHANTOM
-MM
1638
303
1460
130
145
22
OTOKO
1472
266
1191
109
76
10
MIRD5
1420
299
1910
94
183
21
VIP-MAN
1574
335
1938
83
244
28
GOLEM
1218
316
1592
72
174
26
    a)lanaka (1992)
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS
                                                                              &EPA
                                                                         1O7

-------
                                        FIGURE 1 :
          THE ANTERIOR VIEWS OF SELECTED ORGAN PHANTOMS OF VOXEL-PH ANTO M-M M (A)
                                 AND MIRD5 PHANTOM (B)
                   (a)
                                                     Thyroid
                                                      Pancreas
                   (b)
                    (Cnsty andEckerman 1987)(b).
&EPA
                                                       Thyroid
                                                       Pancreas
        10B
                                     RADIATION RISK ABBEBBMENT WORKSHOP PROCEEDINGS

-------
                                 FIGURE Z:
     CORRECTION OF LIVER MASS DUE TO THE ADDITION OF VOXEL LAYERS ON THE
                    OUTER SURFACE OF THE LIVER PHANTOM.
2000
1800
1600
g 1400 ^
S3 1200
C3
s 100°
3 800
600
400
200
n
4
)
-
-
o Original liver mass
• .Corrected liver mass
„ : Japanese average
                         1234
                         Additional numbers of voxel layer
                                 FIGURE 3:
    CORONAL SECTIONS OF THE CT IMAGES OF THE VOLUNTEER, MS, THROUGH THE
STOMACH IN SUPINE (A) AND UPRIGHT (B) POSITIONS. ARROWS INDICATES THE STOMACH.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                           &EPA
                                                                      109

-------
                                 FIGURE 4:
 SAGITTAL SECTIONS OF THE CT IMAGES OF THE VOLUNTEER, MS, THROUGH THE SPINE
   IN SUPINE (A) AND UPRIGHT (B) POSITIONS. ARROWS INDICATES THE SP (SPINE),
                   Di (DIAPHRAGM) AND LA (LOWER ABDOMEN).
1 1Q
                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
          ICRP COMMITTEE 2  RADIATION  PROTECTION ISSUES

  KEITH F. ECKERMAN
     Oak Ridge National Laboratory

  ABSTRACT

     The International Commission on Radiological Protection may  release  new  primary
     radiation protection guidance before 2005. Thus, Committee 2 has underway review of all
     aspects of its formulations and data used in the calculation of the dose per unit intake of
     radionuclides and the dose per unit exposure to external radiation fields. This paper briefly
     outlines the work plans of Committee 2 during its current term, 2001-2005, in anticipation
     of the new primary recommendations.

  INTRODUCTION

     During  the  current  term (2001-2005),  the International  Commission on Radiological
     Protection (ICRP) expects to issue new primary radiation protection guidance to supercede
     that of Publication 60 (1).  Each of the Commission's  committees has been directed to
     define and  address  the technical  issues associated with  implementation  of  new
     recommendations.  Committee 2  has the role of translating the Commission's primary
     recommendations into  quantities that can be  used for planning of work practices. The four
     committees of the Commission are:
         >  Committee 1. Radiation Effects
         >  Committee 2. Doses from Radiation Exposures
         >  Committee 3. Protection in Medicine
         >  Committee 4. Application of Recommendations

     The work of the committees is generally carried out within task groups or, in some cases,
     working parties.  Committee 2, chaired by C. Streffer, consists of 17 members with four
     currently active task groups.  The tasks groups are:
         >  DOCAL: Dose Calculational Task Group
         >  HAT: Human Alimentary Tract Task Group
         >  INDOS: Internal Dosimetry Group
         >  REM: Reference Man Task Group

     In  addition,  Committee  2  has  a working  party  on the interpretation of bioassay
     measurements and liaison activities with various  organizations, some formally and other
     carried  out  by individual scientists.   Committee  2 has strong interactions  with  the
     International Commission on Radiological  Units  (ICRU) on various issues, the National
     Council on Radiation Protection (NCRP) on the subject of contaminated puncture wounds,
     and Medical Internal  Radiation Dose  (MIRD)  Committee  of the Society of Nuclear
     Medicine regarding issues in anatomical modeling and bone dosimetry. Those Committee
     members serving in a liaison role help to avoid any duplication  of efforts  among the
     organizations.
                                                                                      &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINSB                               ill

-------
              HUMAN ALIMENTARY TRACT TASK GROUP (HAT)

              HAT, chaired by H. Metivier, is developing a new model of the alimentary tract that will be
              age and gender-specific with regard to the physiology  and anatomy.  In addition,  it is
              intended to yield biologically meaningful doses and hence will address the cells considered
              to be at stochastic risk. This task group will complete its publication during this term.

              REFERENCE MAN TASK GROUP (REM)

              REM,  chaired by B. Boecker, is updating the  anatomical and physiological data within
              Publication 23 (2).  The task group has devoted considerable attention to age and gender
              considerations and will provide extensive comparisons of the reference parameters  with
              values for various population groups,  hi addition, the update includes new information on
              the anatomical and physiological parameters for the in utero period.  The task group will
              complete it publication, as an addendum to Publication 23, during this term.

              INTERNAL DOSIMETRY TASK BROUP (INDO8)

              INDOS, chaired by J. Stather, is responsible for establishing the biokinetic models for the
              behavior of materials within the systemic circulation and assignment of parameters of the
              respiratory and alimentary tract models.   INDOS works  closely  with DOCAL  it its
              activities and has taken the lead in revising the text of the publication that will replace
              Publication 30(3).

              Da BE CALCULATION TASK GROUP (DOCAL)

              DOCAL, chaired by K. Eckerman, is responsible for the calculation of dose coefficients for
              the intake of radionuclides and recently has  been asked to establish a computational
              capability for  external radiation  fields.   DOCAL is  developing a  new  series  of
              computational models of the human anatomy, revising some of the dosimetric models, and
              preparing a replacement for Publication 38(4) on nuclear decay data.

              The efforts of each task group are directed to provide Committee 2 with the technical basis
              for implementation  of the new primary radiation protection guidance expected from the
              Main Commission during this term.  Figure 1 shows the time line for the  work within the
              task groups  such that new dose coefficients for workers might be calculated within this
              term of the Commission, that is, by 2005. Note that the publication replacing Publication
              30  is planned to include both the prospective calculations of dose per unit intake and the
              expected excretion rates following a unit intake. This will be the first time that the ICRP
              has published both calculations in a single report.  Since the models will be applied in both
              directions, this further highlights the importance of having physiologically meaningful
              models with realistically assigned parameter values. The emphasis on realistic models and
              parameters is a very important perspective that has been part of Committee 2's work during
              the past ten years.  The scope of work indicated in Figure 1 is quite extensive, particularly
              at a time when funding by national authorities is limited.
oEPA
         112                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      DamiMETRIC IBBUEB

      The issuance  of new primary recommendations provides an opportunity for an in depth
      review of ICRP's dosimetric formulations.  Although in recent years the formulation has
      been extended to deal with age-specific issues, with minor changes it remains largely that
      of developed by MIRD (5) in the late sixties and carried forward in the mid seventies into
      the ICRP dosimetric system. Specifically, some of the issues of concern are:
         >   Location of cells at risk
         >   Dosimetry of bone seekers
         >   Dosimetry of Auger emitters
         >   Radon dosimetry
         >   Significance of the ETj dose

      All the items  involve concern regarding the local dose  distribution (depth dose).  This is
      well illustrated by the high-calculated doses to the ETi region (anterior portion of the nose)
      using the respiratory  tract model of Publication 66  (6).   The resultant dose values are
      numerically unreasonable.  The dose  in this region is  computed for a 10 um thick layer of
      cells located 40 jam from the surface (a mass of 20 mg  in the adult male). While there is
      little evidence that such cells are at risk for either deterministic or stochastic effects, it does
      appear reasonable to question the validity of the dose averaged over such a small volume.

      Figures  2  and 3  highlight the conservative  nature of the absorbed fraction  data  of
      Publication 30 for calculations of the dose to  the wall  of the stomach and the endosteal
      surfaces of bone. Again, additional information regarding the appropriate target cells and
      consideration  of the energy dependence of the absorbed fraction appear warranted.  The
      current assumption that the dose to  the endosteal  tissue, the dose associated with bone
      cancer, be  averaged over a 10 um layer of soft tissue adjacent to the bone surfaces is also
      subject to review.

      ANATOMICAL. MODELS

      A major effort is underway within DOCAL to develop a  new series of reference anatomical
      models to be used in computation of dose from both internal sources and external radiation
      fields. The current mathematical representation of the anatomy, commonly referred to as
      the MIRD  phantom (7), was developed in the mid sixties  to assess the dose from internal
      photon sources using Monte Carlo methods.  Later efforts (8) extended the modeling to
      children and the female;  however, throughout the years  the geometry remained that of
      simple conic sections intersected by planes.  Digital medical images now provide a basis
      for the establishment of a computational model with a high degree  of realism as  evident in
      Figure 3. The computational model, however, must represent the reference anatomical data
      established by Committee 2 and provide a geometric description of both the source regions
      where radionuclides reside while within the body and the target tissues considered at risk.
      Not all of these  regions can be visualized in the medical  images;  e.g., the airways of the
      lung or active marrow of the skeleton.
                                                                                         &EPA
RADIATION  RIBK ASSESSMENT WORKSHOP PROCEEDINGS                               1 i a

-------
     Each cross-sectional medical image, of course,  is an averaging over a depth of tissue.
     Thus, a picture element of the image, a pixel, is in reality a 3-dimensional volume element,
     which is called a voxel.  And thus, the computational phantom consists of a stack of cubes
     each with identifiers as to the organ or tissue. About a million voxels would be required to
     represent the anatomy of the adult male.

     An important consequence of the desire for a realistic anatomical model is that no longer
     can a single model represent both genders as in the current hermaphrodite models.  Thus,
     for example,  occupational radiation protection  will be  based  on calculations for each
     gender.  The effective dose coefficient is e where  hT denotes the equivalent dose coefficient
     for organ T and WT the assigned tissue weighting factor (1).  This is the approach that was
     adopted in  Publication 74 (9).  This gender-averaged effective dose coefficient would be
     used for planning work practices. The above equation might require further modification
     if, in addition to breast, other gender-specific tissue weighting factors are indicated.

     Considerable  discussion and  speculation has been devoted to  what changes might be made
     in the tissue-weighting factors, WT. Since the weighting factors are normalized to unity, the
     factor for any tissue reflects our understanding of the risks  among all trie tissues of the
     body. It appears that one might  expect little change in the radiogenic cancer risk estimates;
     however, a decrease in the genetic risks might be in order.   The current set of weighting
     factors is  based on a rather subjective measure  of health detriment,  which might be
     replaced by a more transparent  measure such morbidity.  So, putting this all together, we
     might expect  that the weighting factors for some non-gondal tissues might increase.  The
     weights for some organs, e.g., the thyroid, would be quite sensitive to whether mortality or
     morbidity is used as the measure of detriment.

     The manner in which the remainder tissue group is handled in computing the effective dose
     has also been of some concern.   The  remainder group is those  tissues  that are not
     specifically assigned a  tissue-weighting factor.   Currently  this group  of tissues  has a
     weighting factor of only 5%. The "splitting" rule and the mass-averaging procedures in the
     current formulations of the effective  dose is a complication that may warrant revision in the
     future formulations.

  CONCLUSIONS

     Committee 2 has an ambitious  program underway to update a number of  elements of its
     dosimetric  methodology.   The guiding  objective is to provide realistic models  and
     parameters that can be used for both  prospective and retrospective analyses.
114                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  REFERENCES
  1.  International  Commission  on Radiological  Protection,  1990 Recommendations  of the
     International  Commission  on Radiological Protection, ICRP Publication 60  (Pergamon
     Press, Oxford), 1991.
  2.  International  Commission  on Radiological  Protection,  Report  of the  Task Group  on
     Reference Man, ICRP Publication 23 (Pergamon Press, Oxford), 1975.
  3.  International Commission on Radiological Protection, Limits for Intakes by Workers, ICRP
     Publication 30, Part 1 (Pergamon Press, Oxford), 1979.
  4.  International  Commission  on  Radiological Protection,  Radionuclide Transformations:
     Energy and Intensity of Emissions, ICRP Publication 38 (Pergamon Press, Oxford), 1983.

  5.  R. Loevinger, and M. Herman. A Schema for Absorbed-Dose Calculations for Biologically
     Distributed Radionuclides, MIRD  Pamphlet No. 1 (Society  of Nuclear Medicine, New
     York, NY), 1968.
  6.  International Commission on Radiological Protection, Human Respiratory Tract Model for
     Radiological Protection, ICRP Publication 66 (Pergamon Press, Oxford), 1994.
  7.  W.S.  Snyder, M.R. Ford,  G.G. Warner, and  H.L.  Fisher,  Jr.  Estimates  of Absorbed
     Fractions for Monoenergetic Photon Sources Uniformly Distributed in Various Organs of a
     Heterogeneous Phantom, MIRD Pamphlet No. 5 (Society  of Nuclear Medicine, New York,
     NY),  1969.
  8.  M. Cristy and K. F. Eckerman.  Specific Absorbed Fractions of Energy at Various Ages
     from  Internal Photon  Sources, ORNL/TM-8381/V1-7 (Oak Ridge  National Laboratory,
     Oak Ridge, TN), 1987.
  9.  International Commission on Radiological Protection, Conversion Coefficients for use in
     Radiological Protection Against External Radiation, ICRP Publication 74 (Pergamon Press,
     Oxford), 1996.
  10. J.  W. Poston, Jr.,  K.A. Kodimer, W.E. Bolch, and J.W.  Poston, Sr. "Calculation of
     absorbed energy in the gastrointestinal tract, " Health Phys. 71, 300-306, 1996.
  11. K.F. Eckerman and M.G. Stabin. "Electron absorbed fractions and dose conversion factors
     for marrow and bone by skeletal region," Health Phys. 78,  199-214, 2000.
  12. PJ. Dimbylow.  "The Development of Realistic Voxel Phantoms for Electromagnetic Field
     Dosimetry," in  Voxel Phantom Development (ed. PJ.  Dimbylow), Proceedings of an
     International Workshop (National Radiation Protection Board, UK), 1995.
                                                                                      &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              115

-------
                                          FIGURE 1 :
             TIME LINE OF THE ACTIVITIES OF THE VARIOUS TASK GROUPS DF COMMITTEE 2.
                       2001
2002
2003
2004
2005
2006


                                           DOCALTG
                                         Dosimetric Code

                                   INDOS TG
                             Systemic Biokinetic Models
                                          FIGURE 2:
        COMPARISON OF THE ENERGY-DEPENDENT SPECIFIC ABSORBED FRACTION IN THE MUCOUS
         OF LAYER OF THE STOMACH  (REF ID) WITH THE ENERGY-INDEPENDENT VALUES DF 1C RP
                                    SAF(ML <-- St Cont)
                             10
                                             10'1          10"
                                         Electron Energy (MeV)
              Publication 30 (Ref3)
&EPA
                                       RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
        Figure 3:
        Comparison of energy-dependent absorbed fractions in the endosteal layer oftrabecular bone (Ref. 11) it ith the
        ICRP Publication 30 values (Ref 3).
                        10
                        10
                                               10"        10U
                                      Electron Energy (MeV)
101
                                        FIGURE 4:
       THE MIRD ANATOMICAL MDDEL, ON THE LEFT, THE NRPB NPRMAN MODEL
           (REF 1 2) IN THE CENTER, AND THE HIGH-RESOLUTION VISIBLE MALE
                    FROM THE U.S. NATIONAL LIBRARY OF MEDICINE.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                            oEPA
                                                                                      I 1 7

-------
              EVALUATION OF SPECIFIC ABSORBED FRACTIONS IN VOXEL
                     PHANTOMS USING MONTE CARLO SIMULATION

         SAKAE KINASE, MARIA ZANKL, JLJN KUWABARA, KADRU SATO,
         HlROSHI NOGUCHI, JUN FUNABIKI AND KlMIAKI SAITO
            Sakae Kinase, Jim Kuwabara, Kaorn Sato, Hiroshi Noguchi, Kimiaki Saito:
            Japan Atomic Energy Research Institute

            Maria Zankl: GSF- National Research Center for Environment and Health

            Jun Funabiki: Mitsubishi Research Institute

         ABSTRACT

            There exists a need to calculate specific absorbed fractions (SAFs) in voxel phantoms for
            internal dosimetry. For the purpose, an EGS4 user code for calculating SAFs using voxel
            phantoms was developed on the basis of the EGS4 user code (UCPIXEL). in the developed
            code, the transport of photons, electrons and positrons in voxel phantoms can be simulated,
            particularly the transport simulations  of secondary  electrons in voxel phantoms can be
            made. The evaluated SAFs for the GSF "Child" voxel phantom using the developed code
            were found to be in  good agreement with the GSF evaluated data. In addition,  SAFs in
            voxel phantoms developed at JAERJ were evaluated using the developed code and were
            compared with several published data.  It was found that SAFs depend on the organ masses
            and would be affected by differences in the structure of the human body.

         INTRODUCTION

            The fraction of energy emitted as a specified radiation in a source tissue, which is absorbed
            in a unit target tissue, the so-called specific absorbed fractions (SAFs)  are needed for
            internal dosimetry (U-3). The SAFs used by the International Commission on Radiological
            Protection (ICRP) were obtained on the basis of calculations using the Medical Internal
            Radiation Dose (MIRD) Committee of the Society  of Nuclear Medicine  Pamphlet No.5
            type phantoms (hereafter MIRD 5  type phantoms) (2'3).  However, the  MIRD 5 type
            phantoms, which  use  mathematical expressions using plane,  cylindrical,  elliptical  or
            spherical surface, do  not  model real  human  bodies. Hence,  SAFs calculations  for
            sophisticated models are necessary to evaluate internal doses accurately.

            In recent years, voxel phantoms, which  use computed tomography (CT) and magnetic
            resonance imaging (MRI) sections to provide three-dimensional representations of the
            human body, have  received considerable attention from  the  standpoint of internal
            dosimetry. Several studies have been made on calculation of SAFs in voxel phantoms using
            Monte Carlo codes (4-5'6-7-8). However, to our knowledge, few studies have been  carried out
            on  calculating SAFs by code considering correlations between primary  photons and
            secondary electrons in voxel phantoms.

            At the Japan Atomic Energy Research Institute (JAERI), an EGS4 (9) user code (UCPIXEL)
            (10)  with voxel geometry was developed  for the calculation  of organ doses for external
            exposure of photons and electrons and two voxel phantoms were constructed from CT data
            of Japanese male adult (10) and Japanese female adult (11). The UCPIXEL code,  which is
,,*&>L-    iis                             RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     useful for external  dosimetry using the Japanese voxel phantoms, cannot treat internal
     dosimetry since source particles within voxel phantoms are unavailable.

     This work was carried out to extend the UCPIXEL code for internal dosimetry using voxel
     phantoms and to validate the code, hi addition, this work was performed to evaluate SAFs
     in voxel phantoms  developed at JAERI using the developed code and to compare with
     several published data, in order to investigate what has a great influence on SAFs.

  MATERIALS AND METHODS

     MONTE CARLO CODE

     An EGS4 user code for calculating SAFs in voxel phantoms was developed on the basis of
     the UCPIXEL code, was named EGS4-UCSAF code. Although the EGS4 code is very
     popular as a general-purpose package for Monte Carlo simulation of the coupled transport
     of electrons and photons in an arbitrary geometry  for particles with energies above a few
     keV up to several TeV, users have to write complicated geometries such as voxel phantoms
     in an extended FORTRAN language known as Mortran.  It is hard for users to write the
     geometries accurately. Therefore, the UCSAF code has been developed as a package of
     subroutines plus phantom data, and prevents users from writing geometries in Mortran. The
     division between the EGS4 and the UCSAF is shown in Figure 1. As the original user code
     of EGS4 consists of a MAIN program, the subroutines HOWFAR to specify the geometry
     and the subroutines AUSGAB to score and output the results, the UCSAF code uses the
     subroutines in conjunction with voxel phantom data.

     In the EGS4-UCSAF code  (hereafter UCSAF code), the radiation transport  of electrons,
     positrons and photons in voxel phantoms can be simulated, and correlations between
     primary and  secondary  particles are  included. The Parameter Reduced Electron-Step
     Transport Algorithm (PRESTA) to improve the electron transport in the low-energy region
     is used. The cross-section data for photons are taken from PHOTX for EGS4 code (12J3),
     and the data for electrons and positrons are taken from ICRU report 37 (14-15).

     VOXEL PHANTOM

     In this work, four whole-body voxel phantoms were used: the  GSF female "Child"
     phantom, the  JAERI male "Otoko", female "Onago" phantoms and a MIRD  5 type
     hermaphrodite phantom. The Child phantom was constructed by Veit et al.(16). The Otoko
     and Onago phantoms were constructed by Saito et al.(10>11). The MIRD 5 type phantom(17'18>
     was represented as a voxel phantom by Kinase and Takagi<19). The voxel phantoms were
     made by the construction techniques in the previous development of the voxel phantoms at
     GSF. The Child, Otoko and Onago phantoms are constructed from CT data of real persons.
     The CT data are 256 pixel*256 pixel resolution for the Child phantom and 512 pixelx512
     pixel resolution for the Otoko  and Onago phantoms.  The data of MIRD 5 type voxel
     phantom are the almost same resolution as the Otoko and Onago phantoms. The voxel size
     is 1.54x1.54x8.00 mm3 for the  Child phantom, 0.98x0.98x10.0 mm3  for the Otoko and
     Onago phantoms and l.OOxl.OOxl.OO mm3 for the MIRD 5 type voxel phantom. Every
     voxel belongs to tissue, which is assigned a unique identification number, and appropriate
     attenuation properties, which are assumed to be uniform in all voxels within the tissue.

     In the same way with the Child phantom,  the portion of red bone marrow (RBM) of the
     Otoko and Onago phantoms can be assessed in each single skeletal voxel directly from the


                                                                                     &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINSB                              1 1 9

-------
              CT data. For the MIRD type voxel phantom, no voxel is assigned to the RBM.  Voxels with
              grey values below 800 in bone regions are assumed to consist of RBM, voxels with grey
              values of 2,040 and above are considered to consist of a hard bone only. For voxels having
              grey values between 800 and 2,040, the portion of RBM in each voxel is estimated by
              linear interpolation. Bone surface is assumed to be the entire skeleton since bone surface is
              not distinguished as a separate volume. Table 1 compares the tissue  masses  of the  voxel
              phantoms (Child, Otoko, Onago and MIRD type voxel) with the MIRD 5 type phantom and
              three other voxel phantoms (4'5'6'7'8-20-21)- There are discrepancies between the tissues of the
              Otoko,  Onago phantoms and those of the  other phantoms, and between the tissues of the
              MIRD 5 type voxel phantom and those of the original MIRD 5 type phantom.

              VALIDATION OF UCSAF CODE

              To validate the UCSAF code for calculating SAFs, SAFs for the transport of photons  in the
              Child  phantom were evaluated  by  the UCSAF  code and  were  compared with  those
              evaluated at GSF (22). The source of the photons was assumed to be monoenergetic  in the
              energies 30 keV,  100 keV and 1 MeV, and be uniformly distributed in source tissues. The
              source tissues for photons were kidneys, and the target tissues were over 100  according to
              the identifications. The SAFs were evaluated as the fraction  absorbed in a target tissue  of
              that in the kidneys -AF- divided by the mass of the target organ. The number of history of
              the simulations was determined to be twenty million or eight hundred million in order to
              reduce statistical uncertainties below 5 %. No variance reduction technique was used.

              In addition, SAFs by the kerma approximation were evaluated to examine the differences
              of SAFs between those by considering deposited energies due to secondaiy electrons and
              those by considering energy deposited at the point of photon interaction,  i.e. kerma. The
              source tissue was a kidney, and the photon energy was 30 keV, 100 keV and 1 MeV.

              CALCULATION OF SPECIFIC ABSORBED FRACTION

              SAFs for photons in the Otoko, Onago and MIRD 5 type voxel phantoms  were calculated
              using  the  UCSAF code and  were compared  with the  published  data<4>5'6'7) so  as  to
              investigate what has a great influence on SAFs. The source of the photons  was assumed to
              be monoenergetic in the energy range 10 keV to 4 MeV, and be uniformly distributed in the
              source tissue.  The source tissues for photons were adrenals, kidneys, liver, lungs, pancreas
              and spleen. The target tissues were over 100. The number of history of the simulations was
              determined to be a million in order to reduce statistical uncertainties below 50 %.

           RESULTS AND DISCUSSION

              VALIDATION OF UCSAF CODE

              SAFs in the Child phantom were evaluated by the UCSAF code. Table 2 shows the results
              in comparison with the SAFs obtained at GSF. The SAFs by the UCSAF code agree with
              the GSF data. The statistical uncertainty for the SAF at the worst case was 4.0  %. There are
              a few differences of the SAFs for the RBM as the target tissue between  those evaluated
              here and those evaluated at GSF, because in the latter calculations, dose enhancement to the
              RBM due to increased secondary electron production in the hard bone proportion  of the
              skeleton has not been explicitly modeled, and the kerma approximation has been used.
              Further discrepancies may be attributed to different cross section data used in code.  While
              we used PHOTX as the cross section data, FIUGO (23) was used at GSF. Comparison  of the
              SAFs considering energies deposition due to secondary electrons induced by photons with

&EPA
         1 2O                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     those by the kerma approximation is shown in Table 3. The SAFs considering secondary
     electrons  in  the  Child phantom are  apparently identical  to those by the kerma
     approximation in the photon energy 30 keV, 100 keV and 1 MeV.

     Consequently, the above results substantiated that the UCSAF code can be used to calculate
     SAFs.

     SPECIFIC ABSORBED FRACTION FOR PHOTONS

     SAFs for  monoenergetic photons were evaluated for some source tissues in the Otoko,
     Onago and MIRD 5 type voxel phantoms. To investigate what  influences on SAFs, the
     SAFs by the UCSAF code were compared with the published data (5'6-7>8).  Tables 4A-C
     compare SAFs for photon energies 30 keV, 100 keV and 1 MeV in some tissues (source =
     target) of the Otoko, Onago and MIRD 5 type voxel phantoms with those in the MIRD 5
     type phantom and two other voxel phantoms. The statistical uncertainties for the SAFs hi
     the Otoko, Onago and MIRD 5 type voxel phantoms were below 0.5 %. From the tables, it
     can be  stated that each  phantom has their SAP.  The discrepancies are mainly due to
     differences of the organ masses, and may be partly further influenced by different shapes.

     Figure 2 shows SAFs for photons in the Otoko, Onago and MIRD  5 type voxel phantoms in
     the energy range 10 keV  to 4 MeV. The kidneys were the  source/target tissues. The SAFs
     in the MIRD 5 type phantom, the Golem and the Voxelman are also shown in the figure.
     There was found good agreement of the SAFs for all phantoms except for the Voxelman,
     whose kidneys have a higher mass than those of all the other phantoms, and the SAFs are,
     consequently, lower. The smaller discrepancies between the SAFs in the other phantoms
     are also attributed mainly to differences in organs mass, the different anatomy of the
     phantoms, different cross section data used in codes and the different transport calculations
     of secondary electrons in  codes.

  CONCLUSIONS

     An EGS4 user  code -UCSAF - for calculating specific  absorbed fraction (SAF) using
     voxel phantom was developed and SAFs evaluated by the code  were compared with the
     published data. It was found that the EGS4-UCSAF code is validated and that SAFs largely
     depend on the organ masses and would be affected by differences in the structure of the
     human body. It could be also stated that cross section data used in codes influences SAFs.

     We reached  the conclusion  that the  uncertainties  in SAFs are attributed mainly to
     differences of phantoms. We  therefore suggest that many phantoms with high voxel
     resolution should  be made to obtain  a  "representative" phantom or SAF needed for
     radiation protection. Voxel phantoms with voxel resolution (0.98x0.98x1.00 mm3 ), which
     give a sufficiently accurate approximation for Organ Masses, Are Constructed At JAERI.

  ACKNOWLEDGEMENTS

     The authors express their sincere thanks to Dr. Y. Sakamoto and Mr. F. Takahashi of the
     Japan Atomic Energy Research Institute for their valuable advice.
                                                                                     oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              121

-------
 REFERENCES

 1.  Cristy, M. and Eckerman, K. F. Specific Fractions of Energy at Various Ages from Internal
     Photon Sources.   Report ORNL/TM-8381: Vol. 1-7 (Oak Ridge National Laboratory, Oak
     Ridge, Tennessee) (1987).
 2.  Snyder, W. S., Ford, M. R., Warner, G. G., Fisher, H. L. Estimates of Absorbed Fractions
     for Monoenergetic  Photon  Sources Uniformly Distributed  in  Various  Organs of  a
     Heterogeneous Phantom, Medical Internal Radiation Dose Committee (MIRD) Pamphlet
     No.5, J. Nucl. Med. 10, Supplement No. 3 (1969).
 3.  Snyder, W. S., Ford, M. R., Warner, G. G. Estimates of Specific  Absorbed Fractions for
     Monoenergetic  Photon  Sources  Uniformly Distributed  in Various  Organs  of  a
     Heterogeneous Phantom, MIRD Pamphlet No. 5, Revised. Society of Nuclear Medicine,
     New York, NY (1978).
 4.  Petoussi-HenB, N., Zankl, M. Voxel  Anthropomorphic Models  as a Tool for Internal
     Dosimetry, Radiat. Prot. Dosim. 79, 415-418 (1998).
 5.  Yoriyaz, H., Santos, A., Stabin, M. G., Cabezas, R., Absorbed  Fractions in a Voxel-based
     Phantom Calculated with the MCNP-4B Code, Med. Phys. 27, 1555-1562 (2000).
 6.  Smith, T., Petoussi-HenB, N., Zankl, M. Comparison of Internal Radiation Doses Estimated
     by MIRD and Voxel Techniques for a "Family" of Phantoms, Eur.  J. Nucl. Med. 27, 1388-
     1398(2000).
 7.  Smith, T., Phipps, A., Petoussi-HenB, N., Zankl, M. Impact on Internal Doses of Photon
     SAFs Derived with the GSF Adult Male Voxel Phantom, Health Phys. 80, 477-485 (2001).
 8.  Chao,  T C, Xu, X G, Specific Absorbed Fractions  from the Image-based VIP-Man Body
     Model and EGS4-VLSI Monte Carlo Code: Internal Electron  Emitters, Phys. Med. Biol.
     46,901-927(2001).
 9.  Nelson, W. R., Hirayama, H. and Rogers, D. W. O. The EGS4 Code System.  SLAC-265
     (1985).
 10. Saito,  K., Wittmann, A., Koga, S., Ida,  Y., Kamei,  T.,  Funabiki,  J. and  Zankl, M.
     Construction of a Computed Tomographic Phantom for a Japanese Male Adult and Dose
     Calculation System. Radiat. Environ. Biophys. 40, 69-76 (2001).
 11. Saito, K. Unpublished.
 12. RSIC.  DLC-136/PHOTX Photon Interaction  Cross  Section Library (contributed by
     National Institute of Standards and Technology) (1989).
 13. Sakamoto, Y. Photon Cross Section Data PHOTX for PEGS4. In: Proc. Third EGS4 Users'
     Meeting in Japan. Tsukuba, July 1993.  KEK Proceedings 93-15, 77-82 (in  Japanese)
     (1993).

 14. Berger, M. J., Seltzer,  S. M. Stopping Power and Ranges of Electrons  and Positrons
     (Second Edition), U. S. Department of Commerce Report NBSIR 82-2550-A (1983).
 15. ICRU. Stopping Powers for Electrons and Positron, ICRU Report 37 (1984).
 16. Veit R., Zankl M., Petoussi-HenB, N., Mannweiler E., William G., Drexler G. Tomographic
     Anthropomorphic Models. Part .  Construction Technique and Description of Models of an
     8 Week Old baby and a 7 Year Old Child. GSF-Bericht  3/89 GSF National Research
     Center for Environment and Health, Neuherberg, Germany (1989).
1 22                             RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  17. Cristy, M. Mathematical Phantoms Representing Children of Various Ages for Use in
     Estimates  of Internal  Dose,  Report  ORNL/NUREG/TM-367  (Oak  Ridge National
     Laboratory, Oak Ridge, Tennessee) (1980).
  18. Iwai, S., Sato, O. and Tanaka, S. Evaluation of Fluency to Dose Equivalent Conversion
     Coefficients for High Energy Photons. In: Proc. Third EGS4 Users' Meeting in Japan.
     Tsukuba, July 1993. KEK Proceedings 93-15, 10-53 (in Japanese) (1993).

  19. Kinase, S andTakagi, S. Unpublished.
  20. Zankl, M., Wittmann, A. The Adult Male Voxel Model "Golem" Segmented from Whole
     Body CT Patient Data, Radiat. Environ. Biophys. 40, 153-162 (2001).
  21. Zubal, I. G., Harrell, C. R., Smith, E. O., Rattner, Z., Gindi, G., Hoffer, P. B. Computerized
     Three-dimensional Segmented Human Anatomy, Med. Phys. 21, 299-302 (1994).

  22. Petoussi-HenB, N., Zankl M., Henrichs K. Tomographic Anthropomorphic Models. Part D.
     Specific Absorbed Fractions of Energy for a  Baby and a Child from Internal Photon
     Sources. GSF-Bericht 7/97 GSF National Research Center for Environment  and Health,
     Neuherberg, Germany (1997).
  23. RSIC. DLC-099/HUGO Photon Interaction Data in ENDF/B-V Format (contributed by
     National Institute of Standards and Technology)  (1983).
                                                                                    «*EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              1 23

-------
                                          TABLE  1 :
COMPARISON or SELECTED TISSUE MASSES FOR THE CHILD  PHANTOM('
                       GOLEM"1'S''7'201,VOXELMAN15'21' AND
TISSUE
Adrenals
Bladder
Brain
Kidneys
Liver
Lungs
Pancreas
RBM
Skeleton
Skin
Spleen
Thymus
Thyroid
CHILD(16>
0.004
0.025
1.316
0.188
0.733
0.153
0.030
1.228
2.048
1.180
0.151
0.030
0.005
OTOKOl10)
0.021
0.012
1.472
0.266
1.191
1.546
0.109
2.834
7.776
2.195
0.076
0.005
0.010
ONAGO<11)
0.020
0.024
1.148
0.257
1.448
0.996
0.053
2.415
7.133
1.975
0.091
0.002
0.006
MIRD
VOXEL(19>
0.016
0.047
1.400
0.300
1.908
1.000
0.064
—
9.870
2.990
0.181
0.021
0.021
MIRD<1>
0.016
0.048
1.420
0.299
1.910
1.000
0.094
1.120
10.000
3.010
0.183
0.021
0.021
GOLEM
(4,6,7,20)
0.023
0.068
1.218
0.316
1.592
0.729
0.072
1.177
10.450
4.703
0.174
0.011
0.026
VOXEL-
MAN*"1)*
0.004
0.212
1.230
0.512
1.967
1.038
0.053
1.391
7.336
20.480
0.374
—
0.007
VIP-
MANW
0.008
0.041
1.574
0.335
1.938
0.911
0.083
11.245
2.253
0.244
0.011
0.028
    * Voxelman is a model from the vertex down to mid-thigh; the lower part of the legs is not contained This affects the
    masses of red bone marrow, skeleton and skin.
                                        TABLE 2:
COMPARISON OF SAFs IN THE CHILD PHANTOM(ISI
 LJCSAF CODE AND THOSE EVALUATED AT THE G S F"(    | N THE PHOTON  ENERGY OF
           KEV,  1 OD  KEV AND 1 MEV,  SOURCE TISSUES ARE THE KIDNEYS.
                                                       BETWEEN THOSE ESTIMATED BY
                                                        (Z2)
THE
3D
TARGET
ORGAN
Adrenals
Bladder
Brain
Colon
Kidneys
Liver
Lungs
Pancreas
RBM
Skeleton
Skin
Spleen
Stomach
Thymus
Thyroid
30 KEV
THIS WORK
2.4x10-1
24x10-3
1.6x10-6
1 4x10-1
1.6
8.1x10-2
5.0x10-3
1.9x10-1
8.2x10-3
4.4x10-2
5.6x10-3
1.4x10-1
3.7x10-2
7.7x10-"
1.7x10-"
GSFI22)
2.5x1 0-1
2.4x10-3
1.4x10'6
1.4x10-1
1.7
8.3x10-2
4.7x10-3
2.0x10-1
7.2x10-3
3.2x10-2
5.4x10-3
1.4x10-1
3.7x10-2
7.4x10^
1.4x10-"
100 KEV
THIS WORK
9.5X10-2
8.0x10-3
1.6x10-"
6.2x10-2
33x10-1
4.2x10-2
9.4x10-3
8.5x10-2
8.8x10-3
2.5x10-2
5.1x10-3
5.5x10-2
3.2x10-2
4.4x10-3
2.0x10-3
GSFI22)
9.6x1 0-2
8.2x10-3
1.6x10-"
6.4x10-2
3.4x10-1
4.4x10-2
9.5x10-3
87x10-2
1.1x10-2
1.8x10-2
5.1x10-3
5.6x10-2
3.3x10-2
4.6x10^
2.1x10-5
1MEV
THIS WORK
7.9X10-2
7.7x10-3
5.7x10-"
5.0x10-2
3.2x10-1
3.5X10-2
8.6x10-3
6.7x10-2
9.0x10-3
7.8x10-3
5.6x10-3
4.6x10-2
2.5x10-2
4.7x10-3
2.8x10-3
GSFI22)
7.9x10-2
7.6x10-3
5.7x10""
5.0x10-2
3.3x10-1
3.5X10-2
8.7x10-3
6.6x10-2
9.6x10-3
8.1x10-3
5.8x10-3
4.6x10-2
2.5x10-2
4.6x10-3
2.6x10-3
1 24
                                      RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                   TABLE 3:
    COMPARISON OF SAFs IN THE CHILD PHANTOM'IS) BETWEEN THOSE CONSIDERING
ENERGIES DEPOSITION BY SECONDARY ELECTRONS AND THOSE BY KERMA APPROXIMATION
             IN THE PHOTON ENERGY OF 3D KEV, 1 DD KEV AND 1 M E V.
                        SOURCE TISSUES ARE THE  KIDNEYS.
TARGET
ORGAN
Adrenals
Bladder
Brain
Colon
Kidneys
Liver
Lungs
Pancreas
RBM
Skeleton
Skin
Spleen
Stomach
Thymus
Thyroid
30 KEV
ELECTRON
TRANSPORT
2.4x10-1
2.4x10-3
1.6x10-6
1.4x10-1
1.6
8.1x10-2
5.0x10-3
1.9x10-1
8.2x10-3
4.4x10-2
5.6x10-3
1.4x10-1
3.7x10-2
7.7x10-^
1.7x10-4
KERMA
2.4x1 0-1
2.4x10-3
1.5x10-*
1.4x10-1
1.6
8.1x10-2
5.0x10-3
1.9x10-1
8.2x10-3
4.4x10-2
5.6x10-3
1.4x10-1
3.7x10-2
7Jx104
1.7x10-4
100 KEV
ELECTRON
TRANSPORT
9.5x10-2
8.0x10-3
1.6x10-4
6.2x10-2
3.3x10-1
4.2x10-2
9.4x10-3
8.5x10-2
8.8X10-3
2.5x10-2
5.1x10-3
5.5x10-2
3.2x10-2
4.4x10-3
2.0x10-3
KERMA
9.4x10-2
7.8x10-3
1.6x10-4
6.2x10-2
3.3x10-1
4.2x10-2
9.5x10-3
8.5x10-2
8.6x10-3
2.5X10-2
5.1x10-3
5.5x10-2
3.2x10-2
4.4x10-3
1.9x10-3
1MEV
ELECTRON
TRANSPORT
7.9x10-2
7.7x10-3
5.7x10^
5.0x10-2
3.2x10-1
3.5x10-2
8.6x10-3
6.7x10-2
9.0x10-3
7.8x10-3
5.6x10-3
4.6x10-2
2.5x10-2
4.7x10-3
2,8x10-3
KERMA
7.6x10-2
7.6x10-3
5JX10"1
5.0x10-2
3.3x10-1
3.5x10-2
8.6x10-3
6.7x10-2
8.7x10-3
7.9x10-3
5.8x10-3
4.6x10-2
2.5x10-2
4.7x10-3
2.5x10-3
                                   TABLE 4A:
    SPECIFIC ABSORBED FRACTIONS (KG"') IN SOME TISSUES (SO U RC E = TARG ET) FOR
        0' AND DNAGO" " PHANTOMS, MIRD 5 TYPE VOXEL PHANTOM*191,  MIRD 5 TYPE
        PHANTOM11', BOLEM(4'6'V>ZD',VDXELMAN(5'21'. PHOTON ENERGY OF 3D KEV
TISSUE
Adrenals
Kidneys
Liver
Lungs
Pancreas
Spleen
OTOKO'10'
5.7
1.2
0.43
0.18
2.4
3.4
ONAGO<11>
6.3
1.2
0.38
0.24
4.1
3.0
MIRD
VOXEL<19>
7.4
1.1
0.30
0.25
3.3
19
MIRD")
6.8
0.99
0.28
0.24
2.4
1.8
GOLEM
(4,6,7,20)
5.3
1.1
0.36
0.29
3.5
2.0
VOXEL-
MANlWD
20
0.71
0.28
0.22
4.4
1.1
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                oEPA
                                                                           125

-------
                                      TABLE 4B:
    SPECIFIC ABSORBED FRACTIONS (KG"1) IN SOME TISSUES (SO U RC E = TARB ET)  FOR
 DTOKO"0' AND DNAGO" " PHANTOMS, MIRD 5 TYPE VOXEL PHANTO M(' % MIRD 5 TYPE
       PHANTOM1", GOLEM(4>6''7'2a',VaXEUMANC5'Z1).  PHOTON ENERGY OF 1 CUD KEV
TISSUE
Adrenals
Kidneys
Liver
Lungs
Pancreas
Spleen
OTOKO<10>
1.0
0.26
0.12
0.041
0.51
0.68
ONAGO<11>
1.1
0.26
0.11
0.055
0.81
0.61
MIRD
VOXEL<19>
1.4
0.23
0.093
0.053
0.65
0.43
MIRD'"
1.3
0.23
0.092
0.053
0.52
0.42
GOLEM
(4,6,7,20)
0.94
0.24
0.11
0.062
0.69
0.43
VOXEL-
MAN<5^1)
3.4
0.17
0.091
0.051
0.84
0.26
                                      TABLE 4C:
    SPECIFIC ABSORBED FRACTIONS (KG"1) IN SOME TISSUES (SO U RC E —TARG ET)  FOR
                                                                         MIRD 5 TYPE
QTOKo'10' AND DNAGO" 1! PHANTOMS, MIRD  5 TYPE VOXEL PHANTOM1191
         PHANTOM11', GOLEM(4'S>V'ZD1,VOXELMAN(5'Z1'. PHOTON ENERGY OF 1  M EV
TISSUE
Adrenals
Kidneys
Liver
Lungs
Pancreas
Spleen
OTOKO<10>
1.0
0.25
0.11
0.036
0.48
0.68
ONAGO<11>
1.1
0.25
0.098
0.048
0.79
0.59
MIRD
VOXEL*19)
1.3
0.22
0.079
0.046
0.63
0.40
MIRD'1'
1.5
0.23
0.081
0.047
0.53
0.41
GOLEM
(4,6,7,20)
1.0
0.23
0.094
0.055
0.69
0.42
VOXEL-
MAN<5,21>
3.8
0.16
0.077
0.045
0.71
0.24
1 26
                                   RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                  FIGURE 1 :
              FLOW CONTROL WITH USER USING  EGS4-LJCSAF CODE.


f User ]
Control Data



User
Code
EGS f Me
Code f





dia Data

3EGS
'




V






User
Control Data






i



MAIN

/
/
1ATCH

\
*-



INPUT
PHAMAS(O)
SRCORG
SOURCE
OUTDOf

OUTPUT
ECNSV1

NTALLYetc

SHOWER



I

1 Block Data
1 (Default)
Subroutines HATCH



SHOWER
HOWFAR
AUSGAB





• • • to establish media data
• • • to initiate the cascade
• • • to specify the geometry
• • • to score and output the
Tf
» ELECTR
	 '
-*\ MSCAT |

->|BHABHA[-

-»|MOLLER[-
i 	 1
-*\ BREMS |-


N
Information j
Extracted from
SHOWER J
T

AUSGAB



i — i
NX '• y
x\
PHOTON


t I














— | PAIR [«

IpHOjok





1
1 	 »

unui I.
                                  FIGURE 2:
    SPECIFIC ABSORBED FRACTIONS IN THE KIDNEYS TISSUES (SO U RC E =TARB ET) FOR
       DTDKO(la' AND DNAGO" " PHANTOMS, MIRD 5 TYPE VOXEL PHANTOM1'9''
            MIRD 5 TYPE PHANTOM'",  GOLEM1'1'6''7'20' AND VoxELMAN15'21'
               IN THE PHOTON ENERGY RANGE OF 1 D  KEV TO 4 M EV.
                O)

                lo-
                CO
t E Otoko
Kg A Onago
I
10°



1C'1
1C'2
1C
[ i D MIRD 5 type voxe
K * MIRD 5 type
: T, * Golem
n
• R H Voxelman
H i 1 § i gK
H H H *
H I
" H

)-2 10'1 10°
\
,
•
-
'•
•

~
•
10
                                 ENERGY (MeV)
                                                                              &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                         1 2V

-------
             THIS PAGE INTENTIONALLY LEFT BLANK
vvEPA
        1 2B                           RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     DEVELOPMENTS  IN  RADIATION

     RISK ASSESSMENT SESSION


BACKGROUND

    This  session featured three presentations; two were on uncertainties  and one was a
    discussion on the detailed analysis of the dose distributions for the workers involved in the
    Tokai-mura accident. Uncertainties in the current cancer risk coefficients was presented by
    EPA.  The uncertainties in the use of epidemiological data from Hiroshima-Nagasaki
    atomic bomb survivors and how they effect the radiation risk models were presented and
    discussed.

PAPERS FROM RADIATION RISK ASSESSMENT SESSION

    To follow are the papers written by the following conference presenters:
       >•  David Pawel
       >  Shohei Kato
       >  Fumiaki Takahashi
       >  Mike Boyd and Keith Eckerman
                                                                     &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINQB                         129

-------
                      UNCERTAINTIES IN ESTIMATES DF CANCER RISK
                 FROM ENVIRONMENTAL EXPOSURE To  RADIONUCLIDES

          DAVID J. PAWEL, R. W. LEQQETT,  K. F. ECKERMAN AND C. B. NELSON
              U.S. Environmental Protection Agency, Office of Radiation and Indoor Air
              Oak Ridge National Laboratory, Life Sciences Division

          ABSTRACT

              This report outlines an analysis of uncertainties for the risk coefficients tabulated for more
              than  800 radionuclides in Federal  Guidance Report No. 13 (EPA, 1999).   The risk
              coefficient for intake of a radionuclide in air, food, or water is an estimate  of the probability
              of radiogenic  cancer mortality or morbidity per  unit activity taken into the body.  This
              analysis incorporates uncertainties associated with biokinetic, dosimetric and radiation risk
              models, and dose modifying factors such as the relative biological effectiveness (RBE) and
              dose and dose rate  effectiveness factor  (DDREF).  The uncertainty  analysis did not
              consider uncertainties associated with absorbed  dose as a measure of radiogenic cancer
              risk, idealized representations of the population and  exposure, and other  uncertainties that
              may be highly dependent  on the type of application. A summary of the main  results is
              tabulated for ingestion of radionuclides.

          INTRODUCTION

              This report, essentially excerpts from a draft technical report (Legget et al), outlines an
              analysis of uncertainties for the risk coefficients tabulated for more than 800 radionuclides
              in Federal Guidance  Report No. 13  (EPA, 1999).  The risk coefficient for intake of  a
              radionuclide in air, food, or water is an estimate of the probability of radiogenic cancer
              mortality or morbidity per unit activity taken into the body.  A  risk coefficient may be
              interpreted either as the average risk per unit exposure for persons exposed throughout life
              to a constant activity concentration of a radionuclide in an environmental medium, or as the
              average risk per unit exposure for persons exposed for a brief period to the radionuclide in
              an environmental medium.   The risk coefficients in FGR 13 apply to an average member
              of the public, in the  sense that estimates of risk are averaged over the age and gender
              distributions of hypothetical populations with mortality rates and air, food and water
              intakes based on recent data for the U.S.

              FGR   13 provides semi-quantitative estimates of uncertainties  for  a  few  selected
              radionuclides.  Although FGR 13 is not the first document to examine uncertainties in
              radiogenic risk estimates,  prior studies such as NCRP (1997) and EPA  (1999)  evaluated
              uncertainties only for whole-body irradiation (where  all tissues receive equal doses). It has
              been  suggested that the uncertainty analysis in  FGR 13 needed  to be  expanded for the
              purpose of setting priorities for deciding what additional information is  needed for
              improving the confidence in risk assessments.

              For this uncertainty analysis, the complex computational approach used for deriving the
              risk coefficients in FGR 13 was simplified. Computations in FGR 13 incorporated models
              for the biological behavior of elements in the human body, the doses to  radiosensitive
              tissues from radiation originating hi the body or in an external medium, and the age-
              specific excess cancer rates per unit dose to these tissues. Also incorporated was data that
              characterized  the mortality and the usage of air, food and water in  the U.S. population.

4>EPA
         1 3D                              RADIATION RISK ASSESSMENT WORKSHOP  PROCEEDINGS

-------
     Data characterizing  mortality and usage  patterns is  reasonably  reliable.  In contrast,
     biokinetic, dosimetric,  and radiation risk models  generally have been  derived  from
     considerably less complete information and in many cases have substantial uncertainties
     associated with their predictions.  This analysis included only those uncertainties associated
     with the biokinetic, dosimetric and radiation risk models, and dose modifying factors such
     as the RBE and DDREF.

     A summary of the main results is tabulated here for ingestion of radionuclides - in the draft
     technical report (Leggett et al) results are also presented for inhalation. The analysis was
     based on subjective probability distributions for parameters associated with the biokinetic,
     dosimetric,  and  risk  models, and  the RBE and DDREF  dose  modifying factors.
     Distributional assumptions for the RBE and DDREF paralleled those made in NCRP report
     126 (NCRP  1997), and risk model distributional assumptions were based  on an  expert
     elicitation (NRC-CRC  1998).  Choices of models, relevant parameters, and  distributional
     assumptions  for characterizing the biokinetics were derived in part by considering the
     extent to  which assumptions  and parameter values underlying ICRP models  might be
     reasonably altered. The uncertainty analysis did not consider uncertainties associated with
     absorbed  dose  as  a  measure of radiogenic cancer risk, idealized representations  of the
     population and exposure, and other uncertainties that may be highly dependent on the type
     of application.

  METHODS

     THE SIMPLIFIED COMPUTATIONAL MODEL USED IN THE PRESENT ANALYSIS

     It is not feasible to  apply the computational model used in FOR 13 in an uncertainty
     analysis for several thousand risk coefficients, due to  its complex formulation involving
     numerous parameters that  depend on time, age, and/or gender.   A  simpler model  was
     formulated whose predictions are consistently close to those of the more complicated
     model and with easier to assess components. For inhalation or ingestion of a radionuclide,
     the simpler model is

     Cancer Mortality Risk =  (dj/a, + D, b,)Ri.                 (1)

     where d, and D; are, respectively, low- and high-LET absorbed doses for tissue i, integrated
     over a period of 20 y assuming acute intake of the radionuclide by an average adult; R,  is
     the age- and gender-averaged site-specific cancer mortality risk estimate for tissue i for
     low-LET uniform  irradiation of the tissue  at high dose and dose rate; a; is  the low-LET
     effectiveness factor at low dose or dose rate and fy is the biological effectiveness of high-
     LET radiation relative to high dose, high dose-rate low-LET radiation. The variables a and
     b correspond, respectively, to the DDREF  and high-dose RBE.  The dose and dose rate
     effectiveness factor is used to account for an apparent decrease of the risk of cancer per unit
     dose at low doses or low dose rates for most cancer sites compared with observations made
     at high, acutely delivered doses in epidemiologic  studies such as that of the  atomic bomb
     survivor cohort.   RBE typically refers to the relative biological  effectiveness of alpha
     radiation in producing  fatal cancers, compared  with  200 kV x-rays at doses less than 0.2
     Gy. Here,  "high  dose RBE" is the relative biological effectiveness of alpha radiation to
     200kV x rays when both are received at doses greater than 0.2 Gy.  Nominal values for
     DDREF, RBE and high dose RBE are  given in Table 1.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               i 31

-------
                                                TABLE 1 :
                         NOMINAL. VALUES FDR DDREF AND RBE FRDM FGR 1 3.
TISSUE
Breast
Leukemia
All others
DDREF3
1
2
2
RBEA
10
1
20
HIGH DOSE
RBE4
10
0.5
10
              Even with this simpler model, a full-scale parameter uncertainty analysis is prohibitive
              because  of the  large number of cases to be  considered  and difficulties  in  assigning
              uncertainty  distributions to some of the parameter values of  Eq.  1.   For  each  risk
              coefficient,  a limited analysis based on  propagation of uncertainties was performed to
              assess the sensitivity of predictions of Eq. 1 to dominant sources of uncertainty in each of
              the parameter values di, ai, Di,   Ri, and DDREFi,.  The uncertainties were propagated
              through assignment of continuous uncertainty distributions to each of the parameter values
              Ri, ai, di,  and  Di  , and  application  of random simulation  techniques to the model
              represented by Eq. 9 to generate a range of possible values of each risk coefficient.  The
              5% and  95% values from the generated range formed the basis  for assigning a nominal
              uncertainty  interval for each risk coefficient.  This incorporated  evaluation of subjective
              judgements derived  from an expert elicitation  (NRC-CEC 1998), previously  published
              reports on  uncertainties (such as NCRP 1997; EPA  1999),  and additional subjective
              judgments of the authors.

              Assignment  of  uncertainties to the values R, (age- and gender-averaged  risk model
              coefficients for high dose for tissues i=l,2,...) was based on recently published judgments
              of nine  independent  experts on the health  effects  of radiation (NRC-CEC 1998).
              Assignment of uncertainties to the alpha RBEs, a,, and the dose and dose rate effectiveness
              factor, DDREF!, were the same as in an EPA report on uncertainties from whole-body low-
              LET radiation (EPA 1999).  Assigned uncertainties for the RBEs were based on ranges of
              values  determined  from  experimental  and epidemiological  studies  of  the  relative
              carcinogenic effects of low- and high-LET radiation data, as discussed in recent documents
              (NAS, 1988; NCRP, 1990;  ICRP 1991; EPA, 1991; EPA 1999). Conclusions on DDREFs
              were based on subjective evaluations of evidence from animal, laboratory, and to a limited
              extent  on   epidemiological  studies  applied  to  competing  dose-response  models.
              Uncertainties for the parameter  values R,,  a,, and  the  DDREF,  were assumed to be
              independent of the radionuclide and exposure mode.

              Characterization  of  uncertainties  in  the tissue-specific  dose  estimates  dj and Dj
              (respectively, low- and high-LET dose estimates for tissues i=l,2...) was more difficult -
              these uncertainties depend strongly on the radionuclide as well as the exposure  mode and
              this topic has rarely been addressed in the literature. As described  later, uncertainties in the
              values dj and D, were judged from results of a  separate sensitivity analysis in which the
              typically dominant components  of the ICRP's biokinetic  and dosimetric scheme  were
              varied within plausible ranges of values.
         J For doses < 0.2 Gy
         4 For doses >0.2 Gv
oEPA
          1 32
                                            RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  ABBIBNMENT OF UNCERTAINTIES To COMPONENTS OF THE SIMPLIFIED MODEL
     RISK MODEL COEFFICIENTS FOR HIBH DOSE AND Daae RATE

     The U.S. Nuclear Regulatory Commission (NRC) and the  Commission of European
     Communities (CEC)  recently conducted  a  joint  study  aimed  at  characterizing the
     uncertainties in predictions of the consequences of accidental releases of radionuclides into
     the environment (NRC-CEC, 1997, 1998). As part of the exercise, the experts were asked
     to provide 5%, 50%, and 95% quantiles of subjective probability distributions for the total
     number of radiation-induced cancer deaths and for the numbers of tissue-specific cancer
     deaths over  a lifetime in a typical population of 100 million persons, each receiving a
     whole body  dose of 1 Gy  low LET radiation at a uniform  rate over  1 min.  With minor
     exceptions, the tissues considered in the NRC-CEC study are the same as those addressed
     in this report, hi the present analysis, the uncertainty in site-specific cancer mortality risk
     estimates for high-dose, low-LET radiation was based on the judgments of the NRC-CEC
     experts.

     In our analysis, a set of lognormal distributions represented the uncertainties in estimates of
     site-specific  cancer deaths following a high dose of radiation at a high dose rate. For each
     cancer site, a lognormal distribution was  constructed to match the conclusions of a given
     expert. Parameters of the resulting lognormal distributions representing the uncertainty in
     the age- and gender-averaged risk model coefficients, R;, for high  dose and dose rate are
     given in Table 2.

                                      TABLE 2:
        MEAN AND STANDARD  DEVIATION OF DISTRIBUTIONS  REPRESENTING THE
  UNCERTAINTIES IN THE LOB TRANSFORMED CANCER MORTALITY RISK COEFFICIENTS
          (CANCER DEATHS PER PERSON-GY)  FOR HIGH DOSE AND DOSE RATES'.
TISSUE
Bone
Breast
Colon
Leukemia
Liver
Lung
Stomach
Skin
Thyroid
Residual6
MEAN
-7.90
-5.03
-4.90
-4.80
-7.08
-3.90
-5.92
-809
-7.47
-3.78
STANDARD
DEVIATION
1.50
0.85
0.96
0.50
1.49
0.80
1.27
1.34
1.23
0.98
5 Distributions are based on judgments of nine experts on the health effects of radiation (NRC-CEC, 1997).
* As defined in NCR-CEC (1997)
                                                                                      oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
1 33   <*&'>

-------
     TISSUE-SPECIFIC DDREF8

     In Federal Guidance Report No. 13, a DDREF of 2 was applied to all cancer sites except
     breast, for which a value of 1 was applied.  The distributions used here for representing
     uncertainties in the DDREF are described in the EPA report on uncertainties from whole-
     body low-LET radiation (EPA 1999), and are based on the approach developed in NCRP
     Report No. 126. For most sites, we have adopted a distribution that is uniform from 1 to 2,
     and falls off exponentially for values greater than 2. The two parts of the distribution are
     normalized so that: (1) the probability density function is continuous and (2) the integrals
     of the  uniform and exponential portions are each  0.5. Mathematically, this probability
     density for the DDREF, f(x), can then be written:
              f(x) = 0.5                  1   x   2                         (2a)
              f(x) = 0.5 e'(x"2)             x  > 2

     The probability density function for breast given in Eq. 2b is somewhat narrower to reflect
     linear  dose  response  results observed  in  several study populations  and the  apparent
     invariance in risk with dose fractionation  (Hrubec et al.  1989, NAS 1990, Howe 1992,
     Tokunaga et al. 1994).

              f(x) = 2 e2(1-x) (2b)

     TISSUE-SPECIFIC RBEs

     In the derivation of the risk coefficients tabulated in FGR 13, alpha RBEs of 1, 10, and 20
     were applied to red marrow (leukemia), breast, and all other tissues, respectively. For this
     analysis, uncertainty distributions assigned to tissue-specific RBEs were the same as those
     described in the EPA report on uncertainties  from whole-body low LET radiation (EPA
     1 999). For most tissues, a lognormal distribution with geometric mean equal to the square
     root of 50, and a 90% probability assigned to the interval 2.5 to 20 was used. For leukemia,
     the uncertainty in RBE is represented using a uniform distribution between 0 and 1 .

  ESTIMATES Or ABSORBED DOSE

     Assignment of uncertainty distributions to the radionuclide-specific parameter values dt and
     Dj (respectively, low- and high-LET dose estimates for tissues i=l,2...) for internally
     deposited radionuclides is particularly difficult because these values are end products of
     complex calculations involving a collection of uncertain biokinetic and dosimetric models,
     parameters, and assumptions.  Current biokinetic models for elements  generally are  not
     process models, and their parameter values often do not represent measurable quantities.
     Conversion from internally distributed activity to tissue doses involves the application of
     specific energies  (SE values) for numerous  pairs of target and source organs,  and the
     uncertainty in a given SE value depends on the types and energies of emitted radiations.
     Even if the information were available to assign meaningful uncertainty distributions to all
     parameter values of all biokinetic and dosimetric models applied in this report, this would
     not be a feasible task due to the numerous cases considered.
     Assignment of uncertainty distributions to the radionuclide-specific parameter values
     D, (respectively,  low- and hig; -LET dose estimates for tissues i=l,2...)  for internally
     deposited radionuclides is particularly difficult because these values are end products of
     complex calculations involving a collection of uncertain biokinetic and dosimetric models,
     parameters, and assumptions.  Current biokinetic models for elements generally are not
1 34                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      process models, and their parameter values often do not represent measurable quantities.
      Conversion from internally distributed activity to tissue doses involves the application of
      specific  energies  (SE values) for numerous  pairs of target and source organs,  and the
      uncertainty in a given SE value depends on the types  and energies of emitted radiations.
      Even if the information were available to assign meaningful uncertainty distributions to all
      parameter values of all biokinetic and dosimetric models applied in this report, this would
      not be a feasible task due to the numerous cases considered.

      In view of such difficulties,  a  systematic scheme was  devised to produce a Monte Carlo
      simulation of the absorbed doses.  First, we created a data set of dose estimates that were
      calculated for  a  limited number of plausible alternatives of components that typically
      dominate the biokinetic and dosimetric models. The dominant components were identified
      using a relatively detailed sensitivity analyses for selected radionuclides.  Then for each
      radionuclide addressed in this document, we constructed a few substantially different but
      plausible variants for each of those dominant components.  The data set was based on a
      factorial design in which the absorbed dose estimates were calculated for each combination
      of the  selected variants  for the  dominant components, with all other  aspects of the
      biokinetic and dosimetric models left
      unchanged.     Finally   for  each
      radionuclide, the data set was used to
      derive    continuous    distributions
      relating  to each  of the  identified
      components from which doses were
      simulated.
012
      The  following  components  were
      judged to  represent the  dominant
      uncertainties in most situations: the
      rate   of   absorption   from   the
      respiratory   tract  to  blood,  the
      gastrointestinal  absorption fraction
      (fi value),  the  systemic  biokinetic
      model, and SE  values for certain
      combinations of source and target organs and radiation types. For each radionuclide, we
      used 3 different values for/;, 3 different systemic models, and 2 different values for SE.
      Thus for ingestion, at least 18  different sets of dose estimates corresponding to the  18 =
      3x3x2 combinations of variations  of the above  components,  were considered.   For
      inhalation of a radionuclide of a  given absorption type, at  least 54 combinations were
      considered.   Thus, the data set included at least 180 dose estimates for ingestion (for low
      LET radiation there are 18  estimates for each of 10 sites) per radionuclide and at least 540
      dose estimates for inhalation. A portion of this data set is shown in Table 4, which shows
      dose estimates obtained using  the ICRP value for SE for two  of the ten tissue sites for
      ingestion of Ru-106.

      For each radionuclide, there is an important difference between the variants selected for the
      systemic models and the  variants selected for  the other  components.  In general, the
      selected variants for the// value, SE value, and rate of absorption from the respiratory tract
      were chosen with the aim to include a "low" and "high" value that encompass the range of
      most plausible  values.  This is much more difficult to do for systemic models, since for
      most radionuclides the universe of plausible models cannot be coherently defined using a
      single one-dimensional parameter.  For each radionuclide, we assumed that the selected
                                                                                          oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                    135

-------
v>EPA
              systemic models used to construct the data set were randomly selected from the universe of
              all plausible systemic models.  In contrast, we assumed the selected variants for/;, SE, and
              the absorption rate from the respkatory  tract  represent quantile values from continuous
              distributions  that  represent  expert  subjective opinion  on plausible  values for these
              parameters.  To illustrate the difference, suppose there is a radionuclide for which the/} and
              SE values are known, and for which the only uncertainties relate to the proper choice  of a
              systemic model.   The probability that the three selected systemic models (that form the
              basis for the data set) would all yield colon doses less than "true" colon dose would be 0.53
              =  0.125.  In contrast, the selected/;  values  are the 5, 50, and 95% quantiles  for the
              continuous distribution of plausible/ values. According to expert opinion, there would be
              only about a 5% chance that all three/ values are less than the true value.

              For each radionuclide,  doses were simulated separately for each of the systemic biokinetic
              models that were  considered. This was  accomplished by first estimating the functional
              relationship between the dose to each tissue and variables representing the other sources of
              variation (such  as the / value) using the data set  of dose estimates.   Then for each
              radionuclide,  distributions  were  assigned  to the  gastrointestinal  absorption fraction,
              standardized SE values, and in the case of inhalation absorption from the respiratory tract.
              Simulated doses to each tissue were  then calculated by applying the estimated functional
              relationship to simulated values for the sources  of variation (such as the/ value).

                                                TABLE  3:
           COLON AND STOMACH DDSE ESTIMATES USING THE ICRP VALUE FOR SET FOR RU-
SYSTEMIC
MODEL
first
first
first
ICRP
ICRP
ICRP
third
third
third
F1 -VALUE
0.005
0.05 (ICRP)
0.07
0.005
0.05 (ICRP)
0.07
0.005
0.05 (ICRP)
0.07
SE VALUE
ICRP
ICRP
ICRP
ICRP
ICRP
ICRP
ICRP
ICRP
ICRP
COLON DOSE
(GY/BQ)
4.59E-08
4.44E-08
4.37E-08
4.60E-08
4.55E-08
4.54E-08
4.59E-08
4.49E-08
4.45E-08
STOMACH
DOSE (GY/Bo)
1.69E-09
1.71E-09
1.72E-09
1.83E-09
3.13E-09
3.71 E-09
1.76E-09
2.46E-09
2.77E-09
               The final step of the simulation was to fully account for uncertainties in risks associated
               with choice of the systemic absorbed dose model.  We plan to provide details on how this
               was accomplished in an EPA/ORNL technical report.

           RESULTS

               Table 4  summarizes results  from the  Monte Carlo  simulations  in  which  minimal
               uncertainties for the ingestion of risk coefficients were quantified using either 90%, 80%,
               or 50% credible intervals. The 90% credible intervals were the intervals that encompass
               90% of the simulated risk coefficients between Q5 and Q95 (where Q5 is the 5% sample
               quantile of the risk coefficients, and Q95 is the 95% quantile).  The 80% arid 50% credible
               intervals were obtained using Qio,  Q9o, and Q25 and Q75 respectively.   The main results of
               our analysis are summarized in the first three columns.  For about 50% of the radionuclides
               the ratio of Qgs/Q; was less than 23. The ratio Qys/Chs was much smaller; for about 50% of
          136
                                             RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     the radionuclides - not necessarily the same ones - the ratio of the midrange values was
     less than  3.4.  By taking the square roots of the ratios in the first columns, one may
     conclude that subject to the limitations of this analysis, the accuracy of most  of the risk
     coefficients ranges from a factor of about 4 (rounded square root of 12.3) to about 25
     (about the square root of 540).  (All values in the interval from Q5 to Q95 are within a factor
     of roughly  (Qgs/Qs)1'2  of the  risk coefficient, provided the risk coefficient is near the
     geometric mean of Qj and QPJ.)

                                       TABLE 4:
                PUANTILES  FOR THE RATIOS OF UPPER AND LOWER LIMITS
               OF SUBJECTIVE UNCERTAINTY INTERVALS, OBTAINED USING
     MONTE CARLO SIMULATIONS, FOR CANCER RISK COEFFICIENTS  FOR INGESTION.
% OF RADIONUCLIDES
WITH SMALLER RATIOS OF
UPPER TO LOWER LIMITS
5
20
40
50
60
80
95
Q95/Q5
11.5
15.6
20
23
26
49
562
Q90/Q10
6.6
8.2
9.8
11.0
12.3
19.6
104
Q75/Q25
2.6
2.9
3.2
3.4
3.6
4.6
10.4
     As part of the analysis, an assessment was made of the relative contribution of each of the
     different sources of uncertainty. For this analysis, the uncertainties were categorized as to
     whether they relate to 1) models for the calculation of absorbed dose, 2) radiogenic cancer
     risk models for low-LET radiation at high dose and high dose rate, or 3) the dose modifiers
     RBE  and DDREF.  This was accomplished by comparing a) the variance of the log
     transformed risk  values generated when factors associated with all but one  source of
     uncertainty type of model  were varied with b) the variance of the transformed risk values
     when factors associated with each of the sources were varied simultaneously.

     A particular source  of uncertainty was  considered to be dominant if its contribution
     accounted for more than half of the variance of the log transformed risk models.  For 221
     out of 758 radionuclides the  dominant source of uncertainty was  "absorbed dose", and for
     483  radionuclides the dominant source was the  "risk model".   For the  remaining 54
     radionuclides  none of the three sources of uncertainty dominated.   The term "absorbed
     dose" refers to uncertainties relating to the use of both biokinetic and dosimetric models for
     estimating the absorbed doses for each tissue type. The biokinetic models characterize the
     biokinetics of a radionuclide in the lungs and gastrointestinal tract and its absorption to
     blood, as well as  its  systemic biokinetics.  Dosimetric models relate to the conversion of
     activity distributed in the human body to absorbed dose to tissues.  The term "risk model"
     includes only the uncertainties relating to the assessment of risk per unit absorbed dose for
     low LET radiation at high  doses/rates (and therefore does not include uncertainties relating
     to DDREF or RBE).   It should be noted that for ingestion, there was no radionuclide for
     which the dominant source of uncertainty relates to the absorbed dose modifiers RBE and
     DDREF. Uncertainty tends to be smallest for radionuclides for which the dominant source
     of uncertainty is the "risk model" and greatest when the dominant source is associated with
     determination of absorbed  dose.
                                                                                        oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                   137

-------
           DlBCUBBION

              The uncertainties summarized in the previous section were based on a simplified model in
              which the risk per unit activity for each type of radiation (high-LET  or  low-LET) is
              expressed as the product of three components: first, the risk per absorbed dose received by
              target tissues (for low-LET high dose and dose rate radiation); second, modifying factors
              applied to the first component to account for type of radiation,  dose, and dose rate; and
              finally the absorbed dose per unit activity.  This formulation allows a convenient allocation
              of uncertainties  associated with  the models used to derive the risk coefficients, and is a
              logical  extension of formulations  in previous evaluations of uncertainties in risks from
              whole-body irradiation (NCRP 1997; EPA  1999).

              Results from this uncertainty analysis need to placed in perspective, since it is true that
              subjective judgment played a role in almost every step of the process used to generate the
              simulations.   We did not attempt to evaluate uncertainties relating to the validity of the
              linear-no-threshold hypothesis, since this simply is not feasible.  As discussed earlier, we
              based our analysis on a simplified risk model, which did not account for age-dependencies
              in either  absorbed  doses  or risks per absorbed doses.   It follows that uncertainties for
              radionuclides that concentrate in bones (for long periods of time) may be understated in this
              report.

              With these limitations in mind, we nevertheless hope that this report provides a reasonable
              evaluation of the uncertainties for the ingestion of radionuclides in FGR13.
&EPA
  :4>,L    i as                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  REFERENCES
  EPA (1991).  Final Draft for the Drinking Water Criteria Document on Radium, NTIS:
     PB 91225631  (Prepared by Life Systems, Inc., for the Environmental Protection Agency,
     Washington, DC).
  EPA (1999). Cancer Risk Coefficients for Environmental Exposure to Radionuclides, Federal
     Guidance Report No. 13,  EPA 402-R-99-001 (U. S. Environmental Protection Agency,
     Washington, DC).
  Hrubec Z, JD Boice Jr., RR Monson and M Rosenstein (1989). Breast cancer after multiple
     chest fluoroscopies: second follow-up of Massachusetts women with tuberculosis.  Cancer
     Res 49, 229-234.
  ICRP (1991). International Commission on Radiological Protection, "1990 Recommendations
     of the  International  Commission on  Radiological Protection",  ICRP  Publication 60
     (Pergamon Press, Oxford).
  Leggett RW, KF  Eckerman, CB Nelson, and DJ Pawel. Uncertainties in estimates of cancer
     risk from environmental exposure to radionuclides - draft EPA/ORNL technical report.
  NAS (1988). Health Risks of Radon and Other Internally Deposited Alpha-Emitters (BEIR IV)
     (National Academy of Sciences, National Academy Press, Washington, DC).
  NAS (1990). Health Effects  of Exposure  to Low Levels of Ionizing Radiation (BEIR V)
     (National Academy of Sciences Press, Washington , DC).
  NCRP (1980). Influence of Dose and Its  Distribution in Time on Dose-Response Relationships
     for Low-LET Radiations, NCRP Report 64 (National Council on Radiation Protection and
     Measurements, Bethesda, MD).
  NCRP (1990). The Relative Biological Effectiveness of Radiations of Different Quality, NCRP
     Report No.  104 (National Council on Radiation Protection and Measurements, Bethesda,
     MD).
  NCRP (1997).  Uncertainties in Fatal Cancer Risk Estimates Used in Radiation Protection,
     NCRP Report No. 126  (National Council on Radiation Protection and Measurements,
     Bethesda, MD).
  NRC-CEC (1997).  Probabilistic Accident Consequence Uncertainty  Analysis.   Late  Health
     Effects  Uncertainty Assessment,  NUREG/CR-6555; EUR  16774;  SAND97-2322 (U.S.
     Nuclear Regulatory Commission,  Washington, DC; Office for Publications of the European
     Communities, Luxembourg).
  NRC-CEC  (1998).  Probabilistic  Accident  Consequence  Uncertainty Analysis.  Uncertainty
     Assessment for Internal Dosimetry, NUREG/CR-6571; EUR 16773; SAND98-0119 (U.S.
     Nuclear Regulatory Commission,  Washington, DC; Office for Publications of the European
     Communities, Luxembourg).
  Tokunaga M, CE Land, S Tukoka, I Nishimori,  M Soda and S Akiba (1994).  Incidence of
     female breast cancer among atomic bomb survivors, 1950-1985.  Radiat Res 138, 209-223.
                                                                                     &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              1 39

-------
                EFFECTS OF BASELINE ON UNCERTAINTY OF
                           RADIATION RISK MODELS

 TERUYUKI NAKAYAMAAND SHQHEI KAra
     Radiation Risk Analysis Laboratory, Department of Health Physics,
     Japan Atomic Energy Research Institute

 ABSTRACT

     ICRP, BEIR, UNSCEAR, and EPA developed the radiation risk projection models, which
     are  based  on the  epidemiological data  especially of Hiroshima-Nagasaki atomic  bomb
     survivors. To apply the data to the other population, cancer mortality data and survival data
     are used as the baseline. The purpose of this study is to examine the effects of baseline on
     the radiation risk projection models. At first, using the multiplicative risk projection model,
     we consider whether or not the ICRP's risks are statistically significant in the present. For
     Japan, there exist the significant differences in most of cancer sites except for  esophagus
     and leukemia. For the USA, there are a fewer sites where the difference is more  significant
     than Japan. In Japan, the years that the risk on a year is effective in the future are only one
     year in colon and total cancers etc., and a few years in most of the other cancer sites. By
     extrapolating cancer mortality, we predict the risks  in the future. Also, using  the excess
     relative risk based on attained age,  which are included in the  radiation  risk  projection
     model, the effects of baseline are examined.

 INTRODUCTION

     Using the radiation risk projection model, we can estimate the lifetime excess cancer
     mortality risk in a certain population, where the excess  relative risk (ERR)  coefficient
     obtained by the epidemiological study, mainly of Japanese atomic bomb survivors, is used.
     Then, to take a difference between the population into consideration,  spontaneous cancer
     mortality data and survival data in a population are applied as the baseline. NCRP (1997)
     and EPA  (1999)  evaluate the uncertainties  in  the radiation risk projection  model  by
     assuming the statistical distributions to the uncertain  sources, which are dose and dose rate
     effectiveness  factor (DDREF),  population transfer, epidemiology,   error in  the  death
     diagnosis, dosimetry and so on.  In uncertainty analyses of both organizations,  though the
     assumed distributions are different, the DDREF has  the largest contribution. On the other
     hand, the  population transfer is a  source that the  order of the contribution is greatly
     different between two organization's results, that is, it means  that  the contribution of
     uncertainty varies greatly according to the distribution for the baseline.

     ICRP (1991) derives the lifetime excess cancer mortality risks  by averaging  the values
     calculated from each baseline data of five countries including Japan (mortality data in 1978
     and survival data in 1986-1987) and USA (mortality in 1973-1977 and survival in  1985),
     whose details are given by Land and Sinclair  (1991). However, since these mortality data
     are  the older data than twenty years, we wonder whether the risks projected by ICRP are
     available in the present.
14O                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     Let e be the age at exposure and D an exposure dose. For a cancer site i, suppose that M,(x)
     denotes the mortality rate at age x, S(x) denotes the proportion of survival at age  x, and
     ERR,(e,£>) denotes the ERR given by the exposure age and dose. Then, the multiplicative
     risk projection model is expressed as:
                                                           (1)
     where the minimum latency time L is 2 if leukemia, 10 if else and the plateau period P is 40
     if leukemia, positive infinity if else. Risk for total cancer is obtained by summing up u,(e,D)
     for all z. To exclude the uncertain effect by DDREF and simplify the projection, we set the
     acute exposure dose 1 Sv. Then, the model is transformed such as follows:


                            M,(*)	dx.                   (2)
                              1 \  / f~, /  -^                       '•-'
     Based on the  age  at  exposure, the ERR is estimated  by the epidemiological  study.
     However, the ERR based on the attained age are recently proposed by Kellerer and Barclay
     (1992), and Pierce and Mendelsohn (1999), which state that the ERR based on the attained
     age is more fit to the  data of atomic bomb survivors than the  one based on  the  age at
     exposure. Therefore, we apply the ERR based on the attained age to the risk projection
     model. Then, the model is given as the function of the age at exposures and the attained
     age:a, and expressed by:


                            +PM,(x)-dX.             (3)
     In this study,  following to Land and Sinclair (1991), we deal with esophagus, stomach,
     colon, lung, female breast, ovary, bladder, leukemia and residuals as target cancer sites.
     Also, Six kinds of ages (0, 10, 20, 30, 40, 50) are used as the age at exposure.

     This paper is organized as follows. In the following section, using the test of equality, we
     examine whether or not the ICRP's risks are statistically significant in the present. In the
     next section, the years that the baseline  data is effective are studied. In the subsequent
     section, the risks in future are projected by the extrapolation of the baseline. In the next
     section, using the ERR based on the attained age, the same significance as the above  is
     examined.

     TEBTB OF THE EQUALITY FOR RISKS

     In this section, from the statistical viewpoint,  we examine  whether or not Japanese and
     USA's risks given by ICRP (1991), whose baseline data are given by Lang and Sinclair
     (1991, Table 2, 3), are effective in the present, respectively. As the latest baseline data, we
     can get Japanese mortality in 1999<7), Japanese survival in 1999 from the homepage of
     Ministry of Health, Labour and Welfare, and USA's baselines in  1998  from the homepage
     of National Center for Health Statistics.
                                                                                        oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               141

-------
     Let the index c be the country; J (Japan) or U (USA), the index 5 be the sex; M (male) or F
     (female), and the index y be the data year or ICRP (data  used by ICRP). The risk is
     projected using the equation (2), and  is expressed by uf(e;c,s,y). That  is,  for a  site  an
     exposure age e, the risk obtained by using the Japanese male's baseline data in 1999 and I
     is  expressed by u,(e-J,M,1999). Then, for each sex, by  comparing u,(e;J, *, 1999)  with
     u,(e;J,*ICRP\ and  u,(e;U,*,1998) with u,(e;U,*,ICRP\ the  effectiveness of the ICRP's
     risks in the present is statistically examined.

     It is assumed that the number of cancer-site-specified death and the number of survival are
     independent random variables, each of which follows a binomial distribution, respectively.
     By iterating that  we generate the random numbers according to these  assumptions and
     calculate the risk projection model, the distributions of the risks are investigated. In this
     case, the iteration is done 5000 times.  Then, by illustrating the histogram or the Q-Q plot, it
     seems that each risk has normal  distribution. Therefore, we may use the test of the equality.
     Since we can use the same test regardless of the site, the exposure age, the country and the
     sex, we explain the case of Japanese male for a cancer site and an exposure age.

     Let U,(e;J,M,ICRP) and \J,(e^,M,1999) be a random variable independently distributed as
     normal with the means m[CRp, mjggg and the variances S:JCRP, s'w?, respectably. The null
     hypothesis of the testing problem is expressed as mICRP = m!999. Then,  for a site an exposure
     age e, and I the test statistic is given by

      „ , .   U,(e,J,M,ICRP)-U,(e,J,M,]999-)
      Z,00 =	1  2       ,  	,          (4)
                         \SICRP "*"  ^1999

     which is distributed as  a normal with the mean 0 and the variance  1.  Since we may regard
     u,(e'rJ,M,1999) as the risk of population mean in Japanese male of 1999 under a specific
     condition, which  is sufficient for the large  population, this value can be  calculated  by
     substituting u, for U,. Then, by the normality of Z,, the probability that the null hypothesis is
     rejected, which is called as/7-value, is obtained. Here, we assess the testing problem by the
     significance level of 0.05. That  is, when the /7-value is below 0.05, it means that the null
     hypothesis is rejected and that the risk given by ICRP (1991) is significantly different from
     the risk basing on the baseline data in 1999. The /rvalues for  Japan and USA are shown in
     Tables 1 and 2, respectively.

     In case of Japan (Table 1), except for all ages of esophagus and for most ages of leukemia,
     it is seen that there exist the significant differences between two risks  and mat most ICRP's
     risks are  statistically  not  effective  in the  present.  Since  this  result  depends  on the
     differences of the baseline data, we may say that the baseline affects the risk projection
     model very much. In case of USA (Table 2), though there are  more sites that the difference
     between two risks is not significant than Japan, especially  for female, the ICRP's risks in
     some sites are statistically not effective in the present. So, for the risk projection model, we
     examine the years that the Japanese baseline data is applicable in future.
1 "*2                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                          TABLE 1 :
                        THE P-VALUES DFTHE STATISTIC IN JAPAN,
                       CLASSIFIED BY SEX,  SITE AND EXPDSURE ACE.

        When thep-value < 0.05, there exists the significant difference between u,(e;J,*.ICRP) and u,(e;J,*,1999)
SEX
M
F
EXPOSURE
AGE
Esophagus
Stomach
Colon
Lung
Bladder
Leukemia
Residual
Total
Esophagus
Stomach
Colon
Lung
Breast
Ovary
Bladder
Leukemia
Residual
Total
0
0.221
<0.01
<0.01
<0.01
<0.01
0.086
<0.01
<0.01
0.192
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
0.098
<0.01
<0.01
10
0.236
<0.01
<0.01
<0.01
<0.01
0.271
<0.01
<0.01
0.189
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
0.230
<0.01
<0.01
20
0.242
<0.01
<0.01
<0.01
<0.01
0.585
<0.01
<0.01
0.189
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
0.478
<0.01
<0.01
30
0.237
<0.01
<0.01
<0.01
<0.01
0.674
<0.01
<0.01
0.186
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
0.920
<0.01
<0.01
40
0.246
<0.01
<0.01
<0.01
<0.01
0.044
<0.01
<0.01
0.170
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
0.144
<0.01
<0.01
50
0.474
<0.01
<0.01
<0.01
<0.01
<0,01
<0.01
<0.01
0.158
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
RADIATION RISK ABSESSMENT WORKSHOP PROCEEDINQS
                                                                                         143


-------
                                                TABLE 2:
          THE  P-VALUES DFTHE STATISTIC IN USA, CLASSIFIED BY SEX, SITE AND EXPOSURE AGE.

                When the p-value < 0.05, there exists the significant difference behveen u/e,-U,*,ICRPJ and «,fe;U,*,1998>
SEX
M
F
EXPOSURE
AGE
Esophagus
Stomach
Colon
Lung
Bladder
Leukemia
Residual
Total
Esophagus
Stomach
Colon
Lung
Breast
Ovary
Bladder
Leukemia
Residual
Total
0
0.016
<0.01
0.265
<0.01
0.748
0.127
<0.01
0.999
0.686
<0.01
<0.01
<0.01
0.164
0.674
0.386
0.179
0.394
<0.01
10
0.019
<0.01
0.223
<0.01
0.698
0.417
<0.01
<0.01
0.699
<0.01
<0.01
<0.01
0.143
0.659
0.373
0.370
0.433
<0.01
20
0.019
<0.01
0.214
<0.01
0.688
0.775
<0.01
<0.01
0700
<0.01
<0.01
<0.01
0.146
0.678
0.371
0.537
0.423
0.715
30
0.020
<0.01
0.202
<0.01
0.665
0.233
<0.01
<0.01
0.700
<0.01
<0.01
<0.01
0.186
0.736
0.369
0.969
0.398
<0.01
40
0.020
<0.01
0.232
<0.01
0.652
<0.01
<0.01
<0.01
0.642
<0.01
0.011
<0.01
0.421
0.936
0.369
0.193
0.260
<0.01
50
0.014
<0.01
0.302
<0.01
0.740
<0.01
<0.01
<0.01
0.441
<0.01
0.035
<0.01
0.762
0.352
0419
<0.01
0.075
<0.01
               EFFECTIVE YEARB or BASELINE

               In this section, using the Japanese baseline data from 1980 to 1999 every one-year, the
               years that the baseline is trustworthy and available in future is examined in view of the
               lifetime excess cancer mortality risk, which depends  on the baseline.  At first, let 1985,
               1990 and 1995 years be three representative points. For each point, by ordering the values
               obtained by the same simulation as the previous section (the iteration times is 2000), we
               can obtain  the  boundary value  deciding  the  95%  confidence intervals  (CI) on  each
               representative point. Also, we consider the linear regression models for the risks calculated
               on every one-year such as:
&EPA
J32x2 ,
                                                         J32x2
               where x and y denote the data year and the risk, respectively. The parameters for each
               model are estimated by the regression analysis. By the Cp criterion (Mallows (1973), which
               is one of the methods to select statistically the fittest regression model, one model of them
               is selected and the degree of regression model is 2 or 3 for most sites and exposure ages.
               Then, we  can consider the effectiveness of the  risk by the 95% points and the fittest
               regression model. Three examples are shown in Figures 1, 2 and 3. Since Figure 1 shows
               that the risks based on data of 1985, 1990 and 1995 are available for one year or two years,
               we conclude that, in the meaning of "at least", the risk for colon cancer of a Japanese male
               exposed at age 40 is effective for one year. Similarly, Figure 2 shows that the risks based
          1 44                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      on data of 1985, 1990 and 1995 are available for nine years, six years and more than four
      years, respectively. Therefore, we conclude that the risk for esophagus cancer of a Japanese
      male exposed at age 40 is effective for four to six years. By Figure 3, the risk for esophagus
      cancer of a Japanese  female exposed at age 40 is effective for more than four years. The
      conclusion is summarized in Table 3. For example, in the case of stomach cancer, when we
      project the risk using the baseline data in 2000, this means that its risk is effective until
      2002 or 2003. As a whole, the effective years in future are a few in most sites.  Therefore,
      we can say that the present risk projection model is affected by the baseline.

                                          FIGURE 1 :
         THE LIFETIME EXCESS CANCER MORTALITY  RISKS FDR COLON CANCER OF
       JAPANESE MALE EXPOSED AT ABE 4D  AND THE FITTEST REGRESSION MODEL.

                                 Male,  Colon,  Exposure age=40
                      Di  c
                         r.
                       'Si
                       'J}
                       o
                       X  10.
                                     1985
                                                1990
                                               Year
                                                          1493
    The x- andy-ca.es denote the year of the baseline data and the riskper 10,000 persons, respectively The squares are the
    risks obtained by the baseline on every one-year. The line is the fittest model selected by the Cp criterion. The vertical lines
    m 1985, 1990 and 1995 denote the 95% CJ. In case of the risk in 1995, since the 95% CI intersects with the regression
    model by 1998 (dashed line), we can express that the risk in 1995 is effective for two years in future. Similarly, both risks
    in 1985 and 1990 are effective for one year.

                                          FIGURE 2:
           THE LIFETIME EXCESS CANCER MORTALITY RISKS FDR ESOPHAGUS  DF
       JAPANESE MALE EXPOSED AT AGE  4D  AND THE FITTEST REGRESSION MODEL.

                               Male,  Esophagus,  Exposure age=40
                      X  to
                      UJ  f"
                           1980
                                     1985
                                                1990
                                               Year
                                                          1995
    The risks in 1985, 1990 and 1995 are effective for nine, six and more than four years, respectively.
    Therefore, we conclude that this risk is effective for four to six years.
                                                                                                 oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                           145

-------
                                                FIGURE 3:
                   THE LIFETIME EXCESS CANCER MORTALITY RISKS FOR ESOPHAGUS OF
              JAPANESE FEMALE EXPOSED AT AGE 4D AND THE FITTEST REGRESSION MODEL.

                                   Female,  Esophagus,  Exposure  age=40
                                                     <> •
                                1980
                                          1985
                                                    1990
                                                    Year
                                                               1915
                              All risks m 1985, 1990 and 1995 are included m 95% CI.
                              Therefore, we conclude that this risk is effective for more than four years

                                                TABLE 3:
                              EFFECTIVE YEARS OF THE BASELINE IN FUTURE,
                         WHICH IS GIVEN REGARDLESS  DFTHE AGE AT EXPOSURE.
SITE (SEX)
Total(M.F), Colon(M,F),
Lung(M), Residual(M)
Stomach(M,F),
Bladder(M),
Lung(F), Breast(F),
Residual(F)
Esophagus(M), Ovary(F),
Bladder(F)
Esophagus(F),
Leukemia(M,F)
YEARS
0-1
2-3
4-6
4-
               FUTURE RISK BY EXTRAPOLATION OF- BASELINE

               As described in the previous section, in most cancer sites, the years that the baseline data is
               effective in the future is not so long.  So, we predict the risk in the future by extrapolating
               cancer mortality.

               For a site, a sex and an age group, by applying the simple linear regression (Y=A+Bt) to
               cancer mortality data from 1980 to 1999 every one-year, and extrapolating its regression,
               cancer mortality in the future is predicted. Here, the exponential regression  (Y=AeBt) is
               applied when the mortality decreases sharply. Then, the risk in future can be  obtained by
               applying the  extrapolated cancer mortality to the risk projection model (2). Two examples
               that the risks in 2005 and 2010 are predicted are shown.
&EPA
         146
                                            RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     In Figure 4, the mortality in future is extrapolated by the linear regression. It seems that the
     predicted future risks in 2005 and 2010 are the points on a straight line, hi Figure 5, the
     future mortality is extrapolated by the exponential regression. It seems that the predicted
     future risks are the points on straight line. For other exposure ages and other sites, similar
     results are observed. That is, these results indicate that the future risks are predicted as the
     points on straight line regardless of the regression function. Therefore, in this method, we
     can predict the future risk by the linear regression of several existing risks.

     EFFECTS OF ERR MODELS

     There are two kinds of ERR model. One is the age at exposure model, which means that the
     ERR coefficient is estimated by the epidemiological data basing on the  exposure age, and is
     used in the  previous sections  and is  also adopted in almost  all  risk projection models
     (model (2)). The other is the age  attained model, which means that the ERR coefficient is
     estimated by the  epidemiological data based on the  attained age, and has been recently
     discussed by Kelleher and Barclay (1992), and Pierce  and Mendelssohn (1999) etc.  (model
     (3)).
                                        FIGURE  4:
     THE LIFETIME EXCESS CANCER MORTALITY RISKS AND THE PREDICTED RISKS FDR
                LUNG CANCER OF A JAPANESE MALE EXPOSED AT AGE 4D.

                               Male,  Lung,  Exposure age=40
                   w
                   00 c
                   r, o-
                   O "
                   O
                   C(-l c
                              Exact Risk
                              Predicted Risk
                        litSO
                               1985
                                      1490
                                             1995
                                            Year
    The x- and y-axes denote the year of the baseline data and the risk per 10,000 persons, respectively. The squares are the
    risks on every one-year The circles are the predicted risks. The mortality m future is extrapolated by the linear regression.
                                                                                            oEPA
RADIATION Riatc ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                       FIGURE 5:
    THE LIFETIME EXCESS CANCER MORTALITY RISKS AND THE PREDICTED RISKS FOR
              STOMACH CANCER OF JAPANESE MALE EXPOSED AT AGE 4D.
                           Male,  Stomach,  Exposure  age=40
                x
                0,
                1)  °-l
                                                       ExactJUsk
                                                       Predicted_Ri=k
                     1980
                             1985
                                    1990
                                           1995
                                           Year
                                                          2005
                                                                  2010
                   The mortality in future is extrapolated by the exponential regression.

                                       TABLE 4:
        THE /^VALUES  OF TESTING  STATISTIC IN JAPAN, WHICH IS CLASSIFIED BY
                         ERR MODEL, SEX AND EXPOSURE AGE.
TYPE
1
II
III
SEX
M
F
M
F
M
F
0
<0.01
<0.01
<0.01
<0.01
<0.01
0.397
10
<0.01
<0.01
<0.01
<0.01
<0.01
0.365
20
<0.01
<0.01
<0.01
<0.01
<0.01
0203
30
<0.01
<0.01
<0.01
<0.01
<0.01
0.032
40
<0.01
<0.01
<0.01
<001
<0.01
<0.01
50
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
       When the p-value < 0 05, there exists the significant difference b't the 1CRP 's risk and the risk based on 1999.

     In this section, the effects of baseline data for the risk projection model are examined by
     using the above ERR models. We consider three types of ERR models. Type I is the model
     where ERR coefficients are estimated based on the age at exposure, and given as values in
     Land and Sinclair (1991, Table 1).  The Type II model, though ERR coefficients are based
     on the age at exposure, is given as a function by Kellerer and Barclay (1992, Table 1). Type
     III is the model that ERR coefficients are based on the attained age, and given as a function
     by Kellerer and Barclay (1992, Table  1).   Here, the  target site  is "all cancers except
     leukemia". With the same method as the section of "Tests of the Equality  for Risks", we
     examine whether or not Japanese  and USA's risks given  by ICRP are effective in  the
     present.  As the results of statistical test, the p-values  for Japan and USA are shown in
     Tables 4 and 5, respectively.

     Table 4 is the p-values of testing statistic between the ICRP's risks and the risks obtained
     by 1999 baseline data in Japan. If p-value is below 0.05, we can judge that two risks are
     significantly different. As seen in table, though the significant differences are observed in
1 48
                                   RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINBS

-------
      most, they are not observed only in female exposed at  childhood and youth of the age
      attained model (type III).  This means that, in Japanese female exposed at childhood and
      youth, the age attained ERR model is affected by the baseline less than the age at exposure
      ERR model. Table 5 is the similar results for two USA's risks (ICRP vs. USA 1998). In
      young age of the age attained model  (type III), the significant differences are not observed,
      though they  are observed in the age at exposure model (type I and II). Therefore, though
      the significant differences are observed in type III model of Japanese male, we can say as
      the whole that the age at exposure model is dependent on the baseline data, and that the age
      attained ERR model is hardly affected by the baseline for the exposure at young age.

                                        TABLE 5:
         THE P-VALUES OF TESTING STATISTIC IN USA, WHICH IS CLASSIFIED  BY
                         ERR MODEL, SEX AND EXPOSURE AGE.
TYPE
1
II
ill
SEX
M
F
M
F
M
F
0
0.058
<0.01
<0.01
<0.01
0.252
0.659
10
<0.01
<0.01
<0.01
<0.01
0.262
0.671
20
<0.01
0.888
<0.01
<0.01
0.162
0.581
30
<0.01
<0.01
<0.01
<0.01
0.097
0.32
40
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
50
<0.01
<0.01
<0.01
<0.01
<0.01
<0.01
        When thep-value - 0.05, there exists the significant difference between the ICRP's risk and the risk basing on 1998

  CONCLUBIdNB

      In this study, we have examined the effect of baseline data on the multiplicative risk
      projection model by comparing the risks given by ICRP with the latest risks.

      hi Japan, for almost cancer sites, it is seen that there exist the statistical differences between
      the ICRP's risks and the latest  risks, and that the  ICRP's risks  are not effective in the
      present. It seems one of the reasons that the ICRP's risks are based on the baseline data
      about twenty years ago. Also, the years that the ICRP's risks are effective are  not so long.
      When we  predict the future risks, though the risks  for several  years  are needed,  it  is
      sufficient to extrapolate the risks by a linear regression. However, since this can be  applied
      only  in the  case that the baseline  cancer  mortality is  extrapolated by a linear or an
      exponential  regression,  it should be  remarked that there exist several limitations to
      interpretation of the results, and that more suitable predictions in future must be considered.
      These are future subjects. In the USA, though the sites that the differences are significant
      are a fewer than Japan, there are many sites that the ICRP's risks are not effective in the
      present. We will be interested in the future risks or baseline data.

      In case of the age at exposure ERR model, the significant differences between the  ICRP's
      risks  and the latest risks exist. In the age attained ERR model, the significant differences
      between two risks are not seen only for the exposure at young age. That is, the risk using
      the age attained ERR model is hardly affected by the baseline. Therefore, if the ERR based
      on the attained age, which is more fit to data of atomic bomb survivors than the one based
      on the age at exposure, is used in the radiation risk projection model, we can trust the risk
      for a longer time than the present condition.
                                                                                         oEPA
RADIATION RIBK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                    149

-------
          REFERENCES

          EPA (1999). Estimating Radiogenic Cancer Risks, Addendum: Uncertainty Analysis. US EPA
              Report, 402-R-99-003, US Environmental Protection Agency, Washington, DC.
          ICRP (1991).  Recommendations of International Commission on Radiological Protection,
              ICRP Publication 60. Annals of the ICRP 21, Pergamon Press, Oxford.
          Kellerer, A. M. and Barclay,  D. (1992). Age dependences in the modeling  of radiation
              carcinogenesis. Radiat. Prot. Dosim. 41, 273-281.
          Land, C. E. and Sinclair, W. K. (1991). The relative contributions of different organ sites to the
              total cancer mortality associated with low-dose radiation exposure. Annals of the ICRP
              22(1) 31-37. Pergamon Press, Oxford.

          Mallows, C. L. (1953). Some remarks of Cp. Technometrics 15, 661-675.
          NCRP (1997). Uncertainties  in  Fatal Cancer Risk Estimates Used in Radiation Protection.
              NCRP  Report No. 126, National Council on Radiation Protection and  Measurements
              Bethesda, Maryland.
          Pierce, D. A. and Mendelsohn, M. L.  (1999). A model for Radiation-Related Cancer Suggested
              by Atomic Bomb Survivor Data. Radiat. Res. 152, 642-654.
          Statistics and Information Department,  Ministry of Health, Labour and Welfare (2001). Vital
              Statistics of Japan, 1999.  Volume 3. Health and Welfare Statistics Association, Tokyo (in
              Japanese).
&EPA
         1 SO                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
DETAILED DOSE ASSESSMENT FOR THE HEAVILY EXPOSED WORKERS
                IN THE TOKAI-MURA CRITICALITY ACCIDENT

  AKIRA ENDO, YASUHIRO YAMABUCHI AND FUMIAKI TAKAHASHI
     Department of Health Physics, Japan Atomic Energy Research Institute

  ABSTRACT

     The present paper describes a summary of dose distribution analysis using a numerical
     simulation technique for  the  heavily exposed workers  in the Tokai-mura  criticality
     accident.

  INTRODUCTION

     At around 10:35 on 30 September 1999, a criticality accident occurred in a uranium
     processing plant in Tokai-mura, Ibaraki, Japan.1) Three workers on the spot were heavily
     exposed as a result of the accident. Two of them, who were pouring uranium solution into a
     tank, were heterogeneously exposed to neutrons and y rays associated with the nuclear
     fission reaction. The exposure conditions influenced the clinical progress observed in the
     workers.2) It is therefore necessary to clarify dose distributions in the body by neutrons and
     y rays for the understanding of the biological effects caused by heavy exposures to neutrons
     and y rays.

     By request from the National Institute of Radiological Science (NIRS), a detailed analysis
     of the dose distributions for the two workers was carried out using a numerical simulation
     technique as a joint research  program between JAERI  and NIRS.  The  present paper
     describes a summary of the analysis reported in JAERI-Research 2001-03 5.3)

  METHOD

     A  system  was developed using a  numerical simulation  technique for analyzing dose
     distribution in various postures by  neutron, photon and electron exposures. The system
     consists of Monte Carlo codes, MCNP-4B4) and MCNPX,5) and mathematical human
     phantoms with movable arms and legs.6) The MCNP code  systems were  used for
     simulations of the criticality  and radiation transport  phenomena. MCNPX  is based on
     MCNP-4B and includes mesh tally capabilities  that  are useful  for  the dose distribution
     analysis inside the human trunk. A cross section library FSXLIBJ3R2,7) compiled from the
     latest evaluated data library JENDL-3.28) of the Japanese Nuclear Data Committee, was
     used for neutron  transport calculation, and MCPLIB029) was used for photon  transport
     calculation. Neutron  kerma coefficients and energy absorption coefficients of photons of
     ICRU Report 461C) were used to calculate absorbed doses on the skin and in the body.

     The mathematical phantom with movable  arms and legs developed at JAERI6) shown in
     Figure  1 was used to model the postures of the workers. The phantom is based on the IRD
     phantom11) and can set the positions of the arms and legs independently from  the body.
     Elemental compositions of soft tissue, bone tissue and lung tissue  are those  of MIRD
     Pamphlet.12).  The tank including uranyl nitrate  solution was precisely modeled and the
     postures of the workers were established by the procedures shown in Figure 2. Hearing was
     carried from the  worker to estimate  the  positions and postures at  the moment of the
     accident (Fig 2(a)). Based on the hearing, the experiment was carried out using a mock-up


                                                                                    oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINSS                              151

-------
    facility that reproduces the tank and the room of the accident (Figure 2(b)). And probable
    positions  and postures  under which the two workers could readily pour the uranium
    solution into the tank were determined (Figure 2(c)). As shown in Figure 2(c), the phantom
    standing by the side of the tank represents Mr. A, who was supporting a funnel, and the
    other phantom represents Mr. B, who was pouring the uranium solution into the tank.

    The kcode method of MCNP was applied to calculate the eigenvalue keff for the  critical
    system and to simulate radiation transport of neutrons and y rays from the tank and inside
    the body.

RESULTS

    Figure 3 shows computed neutron and j ray spectra emitted from the surface of the tank.

    Remarkable differences were not recognized in the spectral shapes between the side  and the
    top of the tank. In the y ray spectra, a peak of captured y rays of 2.2 MeV was not clearly
    found. It indicates that most y rays emitted from the tank were due to the nuclear fission
    reaction. Table l(a) shows absorbed doses in the whole body adjusted to the measured 24Na
    specific activities of Mr. A and Mr. B. It was found that about 10 % in the total y doses is
    due to the secondary y rays produced from the capture of neutrons in the human body.
    Table l(b) shows the estimates of NIRS2) from the measured specific activity of 24Na in the
    blood using the methods of ORNL13)/IAEA.14) The methods give only the average doses in
    the whole body by neutrons. The y doses were then estimated from the ratio of neutron to y
    ratio from the monitoring data obtained around the  site15) and the ratio of the y to neutron
    kerma in the IAEA report.14) Both for Mr. A and Mr. B, the doses calculated by the  present
    simulation showed a good agreement with those by NIRS using the ORNL/IAEA methods.
    Figure 4 and Figure 5  show dose distributions on the skin and  inside the trunk of Mr. A,
    respectively.

    The state of heterogeneous exposure was clarified. For instance, the maximum dose on the
    skin of the abdomen was calculated to be 27 Gy for neutrons and 35 Gy for y rays. It is five
    times higher than the average neutron dose in the whole body and three times higher in y
    rays. It was also found the dose decreases with the depth inside the trunk and that the
    tendency is remarkable in the neutron dose rather than the y dose.

SUMMARY

    A numerical simulation technique was developed and applied to  the  dose  distribution
    analysis for the  heavily  exposed  workers  in the Tokai-mura criticality accident. The
    average whole body dose, skin dose and depth dose distributions by neutrons and y rays
    were analyzed. The details of the  simulation technique developed and calculated results
    were reported in  JAERI Research 2001-035.3) These were supplied  to medical teams and
    NIRS, who performed intensive medical care and initial dose assessment of the workers.
    These results are very useful for scientific understanding of the biological effect by heavy
    exposures to neutrons and y rays.
52                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  REFERENCES
  1) The Criticality Accident Investigation Committee. The Report of the Criticality Accident
     Investigation Committee. The Japanese Nuclear Safety Commission (1999). (in Japanese)

  2) National Institute of Radiological Sciences. NIRS Report on the Criticality Accident in a
     Uranium Processing Plant in Tokai-mura. NIRS-M-143 (2001). (in Japanese)
  3) Endo, A., Yamaguchi, Y. and Ishigure, N. Analysis of Dose Distributions for the Heavily
     Exposed Patients  in  the Criticality Accident at Tokai-mura:  Joint Research Program
     between JAERI and NIRS. JAERI-Research 2001-035 (2001). (in Japanese)
  4) Briesmeister, J.F., Ed.  MCNP — A General Monte Carlo N-Particle Transport Code. LA-
     12625-M(1997).

  5) Waters, L.S., Ed. MCNPX User's Manual, Version 2.1.5. TPO-E83 G-UG-X-00001, Rev. 0
     (1999).
  6) Yamaguchi, Y. FANTOME-90: A Computer Code to Calculate Photon External Doses for a
     Phantom with Movable Arms and Legs. Hoken Butsuri, 27, 143-148 (1992). (in Japanese)
  7) Kosako, K.,  Maekawa,  F.,  Oyama,  Y., Uno, Y. and Maekawa, H.  FSXLIB-J3R2: A
     Continuous Energy  Cross  Section  Library  for MCNP  based on JENDL-3.2. JAERI-
     Data/Code 94-020  (1994).
  8) Nakagawa T., Shibata S., Chiba S.,  Fukahori T., Nakajima Y., Kikuchi Y., Kawano T.,
     Kanda Y., Ohsawa T., Matsunobu H., Kawai M., Zukeran A., Watanabe  T., Igarasi S.,
     Kosako K. and Asami T. Japanese Evaluated Nuclear Data Library Version 3 Revision-2:
     JENDL-3.2. J. Nucl. Sci. Technol., 32, 1259-1271 (1995).

  9) Hughes, H.G. Information on the MCPLIB02 Photon Library.  LANL Memorandum X-
     6:HGH-93-77 (1996).
  10) International Commission on Radiation Units and Measurements. Photon, Electron, Proton
     and Neutron Interaction Data for Body Tissues. ICRU Report 46 (1992).
  11)  Snyder, W.S.,  Ford, M.R.,  Warner, G.G. and  Fisher, H.L.Jr.  Estimates  of Specific
     Absorbed Fractions for Photon Sources Uniformly Distributed in Various Organs of a
     Heterogeneous Phantom. J. Nucl. Med., 10, Supplement No.3 (1969).
  12) Snyder, W.S., Ford, M.R. and Warner, G.G. Estimation of Specific Absorbed Fractions for
     Photon Sources Uniformly  Distributed  in Various Organs and Heterogeneous  Phantom.
     NM/MIRD Pamphlet No. 5 (Revised), J. Nucl. Med.,  19, Supplement, 5-67 (1987).
  13) Feng, Y., Brown, K.S., Casson, W.H., Mei, G.T., Miller, L.F. and Thein, M. Determination
     of Neutron Dose from Criticality Accidents  with Bioassays for Sodium-24 in Blood and
     Phosphorus-32 in Hair. ORNL/TM-12028 (1993).
  14) International Atomic Energy Agency.  Dosimetry for Criticality Accidents: A Manual.
     IAEA Technical Report Series No. 211 (1983)

  15) Endo, A. Yamaguchi, Y., Sakamoto, Y., Yoshizawa, M.  and Tsuda, S. External Doses in
     the Environment from the Tokai-mura Criticality Accident. Radiat. Prot. Dosimetry, 93,
     207-214 (2001).
                                                                                     oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              1 53   .4%«V

-------
                                         TABLE 1 :
            (A) ABSORBED DOSES ADJUSTED TO THE Z"*NA SPECIFIC ACTIVITY

Mr. A
Mr.B
ABSORBED DOSE (Gv)
NEUTRON
5.01 (4.7)
2.61 (2.4)
yRAY
PRODUCED IN THE BODY
1.0
0.6
PRODUCED IN THE TANK
10.7
4.4
        The absorbed doses calculated using the kerma coe.cients of red marrow for the bone tissue.  The values in
       parentheses are the absorbed doses calculated using the kerma coe.cients of cortical bone for the bone tissue.

                                   (B) NIRS ESTIMATES

Mr. A
Mr.B
ABSORBED DOSE (Gv)
NEUTRON
5.4
2.9
yRAY
8.5-13
4.5-6.9
24NA SPECIFIC ACTIVITY
(104BQG')
8.24
4.33
                                         FIGURE 1 :
       (A)  MIRD-TYPE PHANTOM AND (B)  PHANTOM WITH MOVABLE ARMS AND LEGS
                                   DEVELOPED AT JAERI.
                                                       (bl
1 54
                                      RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                    FIGURE Z:
              FLOW FDR THE ESTABLISHMENT DF SIMULATION GEOMETRY.
 (A) LOCATION TESTI.ED BY THE WORKER,  (B) BEHAVIOR SIMULATION EXPERIMENT USING A
     MOCK-UP FACILITY, AND (c)  MODELING OF POSTURES FOR DOSE CALCULATION.
                                    FIGURE 3:
      (A) NEUTRON SPECTRA AND (B) y RAY SPECTRA AT THE SURFACE OF THE TANK.
                      10
                    S 10"
Side
T.-p
                                                    £
                                                   w-
                                                      ^
                                     1 'J •     111
                                    Neutron enemy t'MeVj
                                                     '"1.
                                     I01       10"
                                      Photon enti 3» iMe'vi
RADIATION RISK ASSEBBMENT WORKSHOP PROCEEDINGS
                                                                                   &EPA
                                                                              1 55

-------
                                      FIGURE 4:
                ABSORBED DDSE DISTRIBUTIONS ON THE SKIN OF MR.A.
            (Qy)
            61.8 |


            46.3


            30.9


            15.4


             0.0
                                Front
                                      FIGURE 5:
             ABSORBED DOSE DISTRIBUTIONS INSIDE THE TRUNK DF MR.A.
             z-axis
                        7u cm
                         0 cm
             Precipitation tar
                 o
                                              20 cm
                                          nf trunk
Fiqht iK
 •"'"ftiunk
                                                   r>f ti unk
                                                                K= 19-kOcm J
                                                                   trunk
                                                                          t.-ink.
          0  13 S 27 5  41 3 55.0  6« 8
                    (Gv)
                                          o\\ei end
                                          if trunk
                                  '^r end
                                 i. f trunk
15S
                                   RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  AN OVERVIEW OF THE METHODOLOGY USED TO  DEVELOP CANCER
     RISK COEFFICIENTS IN FEDERAL GUIDANCE REPORT No. 1 3

  MICHAEL Bo YD AND KEITH ECKERMAN
     U.S. Environmental Protection Agency, Radiation Protection Division
     Oak Ridge National Laboratory

     I am going to be telling you a little about Federal Guidance Report Number 13 (FOR 13).
     You've also heard some of the details of how the risk coefficients were calculated in Lowell
     Ralston's presentation yesterday and again in David Pawel's presentation today.

     The purpose of FGR 13 was to provide Federal and State agencies and other organizations
     with consistent, technically  sound methods for  assessing  cancer  risks from  exposure to
     radionuclides in the environment. Basically, it allows all the Federal agencies to be using
     the same methodology for doing radiation risks assessments.

     The report describes the methods and models used for estimating cancer risk from internal
     or external environmental exposure to radionuclides and  provides tabulations  of  cancer
     mortality and morbidity  risk coefficients for  assessing  exposure to radionuclides hi
     different environmental media.

     We've issued a series of these technical reports that provide information that is used in
     implementing radiation protection programs. You may be familiar with Federal Guidance
     Reports 11 and 12 (FGR 11  and FGR 12). FGR 11, issued in 1988, provides the dose
     coefficients for radionuclide intakes as well as limiting values of radionuclide intakes and
     air concentrations based on ICRP Publication 30. This is still, in the United States,  the
     principal reference  for dose conversion factors. We have not, in the United States, moved
     to the ICRP 60-based system of dosimetry, although I hope  we're beginning to move hi that
     direction.

     FGR  12, issued in  1993, provides dose coefficients for external exposure to radionuclides
     in air, water and soil. FGR 13 continues this series and provides cancer risk coefficients for
     assessing intakes and external exposure for over 800 radionuclides, the same ones included
     in FGR 11.

     In FGR 13, we  are using the age- and organ-specific dose totals from ICRP Publications
     56, 67, 69 and 71.  This method replaces the old Radrisk dose model used previously.  We
     also moved to new life table and baseline cancer rate information for the United  States,
     replacing the old 1979 through 1981 vital statistics with those from the 1989 to 1991
     assessment. As I understand it, these assessments, which are done every decade, take a few
     years  to compile. We probably will  not have the 1999 to 2001 data for a couple of years.
     So this is the most recent data that we have.

     The exposure modes include inhalation, ingestion of food, ingestion of water, and external
     exposure.  External  is divided into  submersion  (exposure  to radionuclides in  the
     surrounding  air),  exposure  to radionuclides on  the ground surface and exposure to
     radionuclides uniformly distributed in soil.  Cancer risk coefficients for mortality  and
     morbidity are provided for inhalation and ingestion in terms of risk per unit intake (i.e., risk
     per becquerel)  and for external exposure from contaminated  soil  as kilograms  per
     becquerel-second,  which  is  essentially  risk/second per becquerel/kilogram.    Using
     traditional units, the  external coefficient for contaminated  soil is  often converted to
     risk/year per picocurie/gram of soil.

                                                                                      &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              1 57

-------
In developing the risk coefficients, we start out with the cancer risk coefficients from
human epidemiologic studies.  This is the kind of information  that is  in the National
Academy of Sciences BEIR V report now and that will eventually be in BEIR VII, which is
in development.  Depending on the type of cancer, we use either a relative risk or absolute
risk model.  With a relative risk model, the excess radiogenic cancer rate is a function of
the baseline cancer rate.  With the absolute risk model, the number  of excess cancers  is
independent of the baseline  rate for that type of cancer. For cancer sites that we  believe
have a relative risk, we use the U.S. vital statistics to establish the current baseline cancer
rates for the  U.S. population.  By applying U.S. health statistics to the  epidemiological
cancer risk data, we are able to calculate iifetime risk per unit of absorbed dose at each age
for males and females in one-year steps from age 0 to 120.  The methods and models we
use are described in the 1994 EPA Report, "Estimating Radiogenic Cancer Risk," which is
EPA's  methodology that was approved by our Science Advisory Board. We use age- and
gender-specific models for  14 cancer sites that have  been updated for the 1989-91 U.S.
cancer mortality  rates and  life table  data.  Using this data,  lifetime  cancer risks are
calculated.  For a uniform dose to all organs and tissues, the average lifetime mortality is
now calculated as 6 x 10"2 Gray.

The next piece of the equation is to use  age-specific biokinetic models as a function of time
following a unit-activity intake. The absorbed dose rates in Grays per day for each target
site (or organ) are calculated as a  function of time following a unit activity intake,  and the
age-specific biokinetic and dosimetric models are taken from the ICRP publications 56, 67,
69 and 71  for age-dependent doses to  members of the general public following intake of
radionuclides.

The  age-specific  inhalation models from ICRP Publication 66, "Human Respiratory Tract
Model for Radiological Protection," have replaced the old ICRP Task Group model.  Dose
rates per unit external exposure came from Federal Guidance Report 12.

If you know the  lifetime risk  per unit dose and the absorbed dose, then you can get the
lifetime risk per unit activity intake at each age. That's the next step.  Gender-specific risk
for each cancer site is calculated  for a unit activity intake for persons from age 0 to 120.
Risks are calculated for each site  by integrating the  product of the absorbed dose rate as a
function of time following the age  of a unit  activity intake, the lifetime risk per unit
absorbed dose as a function of time following the age of intake, and the fraction of  survival
following the age of intake. The last term accounts for competing causes of death in the
average  population,  allowing  us  to avoid predicting  excess cancers in a portion of the
population that has died from other causes before  an excess cancer would be expected.
Risks  from low  LET radiation  and  high LET radiation  are calculated separately and
combined.

We  now look at the  lifetime cancer  risk for a constant activity  concentration  in an
environmental medium and the age and gender-specific usage data for that environmental
medium.   The  gender-specific  cancer  risks  are calculated  for  a constant  activity
concentration in an intake medium. The activity intake rates are proportional to usage rates
which are gender and age-specific. The inhalation rates are from ICRP Publication 66 and
the ingestion intake rates depend on the medium.  FGR  13 provides references  for age-
specific  tap water usage in liters per day, food energy in kilocalories per day, and milk
usage  in liters per day. Note that we vary the dietary and water intake as a function of age.
                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     Finally, this leads us to the risk coefficient itself, which is the average lifetime cancer risk
     per unit activity. Lifetime cancer risk per unit activity intake, given as risk per becquerel,
     is averaged over age and gender. It's calculated as the lifetime risk for intake at a constant
     activity concentration divided by the lifetime intake.  It retains the effect of relative age-
     and gender-specific usage rates.  However, it allows assessments to be made using default
     reference values such as two liters per day for tap water consumption as a per capita value
     and allows food concentrations in per capita usage to be in the customary units of becquerel
     per kilogram and kilogram per day respectively.

     To  summarize, the internal dose biokinetic models used in FOR 13 are those recommended
     by ICRP.  The external  dose values come from FOR 12. The media usage rates were taken
     from recognized sources such as the ICRP, EPA, and other sources.  The information was
     reviewed by EPA, DOE and NRC.  A complete draft of FGR 13 was reviewed externally
     by  five persons  who were chosen on the basis of their having either general or specific
     expertise in radiological risk assessment. Other EPA offices and agencies participated in
     the Interagency  Steering Committee on Radiation Standards (ISCORS) review of this
     document, and the Science Advisory Board also reviewed it extensively.  Copies of FGR
     13 are available for download from EPA's website, www.epa.gov/radiation.
                                                                                       &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               1 59

-------
            THIS RABE INTENTIONALLY LEFT BLANK
£EPA
        1 eo                         RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     CURRENT ISSUES  IN  RISK  MANAGEMENT &

     RADIATION  PROTECTION  POLICY SESSION


BACKGROUND

     Some of the pressing current issues were examined. This session focused on some of the
     proposed changes in the radiation protection policy by the ICRP. Areas needing more
     attention, such as assessing genetic and fetal risks, were discussed. Decommissioning and
     waste management clearance levels for solid materials in Japan were presented.  The NRC
     reviewed how they develop a technical basis for the release of solid materials.

PAPERS  FROM
RISK MANAGEMENT & RADIATION PROTECTION  POLICY SESSION

     To follow are the papers written by the following conference presenters:
       >•  Michael Boyd
       >  Neal Nelson
       >  Akihiro Sakai
       >  Robert Meek
       >  Scott Monroe
       >•  Hideo Kimura
                                                                  SEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                       161

-------
           UPDATE ON THE ICRP's PROPOSED CHANGES TO THE SYSTEM OF

                                     RADIATION PROTECTION

           MICHAEL BCTYD AND SHQHEI KATO
              U.S. Environmental Protection Agency, Radiation Protection Division
              Japan Atomic Energy Research Institute

           MICHAEL. BOYD

              Mr. Kato and I serve on the Expert Group on the Evolution of the System of Radiation
              Protection (EGRP), which has been reviewing the evolving proposals  of the International
              Commission on Radiological Protection (ICRP) and  its chairman, Prof.  Roger Clarke.
              Before I begin, let me tell you where this Expert Group is located.  The United States
              belongs to the Organization for  Economic Cooperation and Development (OECD), along
              with many European countries, Japan, Korea,  Canada, Mexico,  and others.  Within the
              OECD, there are a number of agencies including the Nuclear Energy Agency (NEA).  The
              Committee on Radiation Protection and Public Health  (CRPPH) is part of the NEA.  This
              committee is where radiation protection issues are discussed within the NEA framework
              and this is the committee that formed the EGRP. There are  12 countries represented on the
              EGRP.

              I'd like to start out with a brief review of the existing ICRP system of radiation protection.
              The ICRP's  last general recommendations were released in 1990 as  Publication 60. In
              these recommendations of the commission, they created a distinction between a practice
              and an intervention.  A practice is something ongoing.  You can think of it as something
              that has served to increase the levels of radioactivity either in the environment or to the
              worker.    An  intervention  is  an  activity that  will  serve  to  decrease radioactivity.
              Interventions are things like remediation or cleanup of contaminated sites and that sort of
              thing.

              With this  distinction, ICRP tended  to focus their suggestions for limits  in  the area of
              practices and to be less specific  about what levels you need to  achieve for an intervention.
              Also in ICRP 60, they reinforced the traditional system of justification, optimization and
              limitation, the three legs of the system of radiation protection.

              They  addressed the  categories  of occupational exposure, public  exposure and medical
              exposure.  Under optimization, they reinforced the concept of collective dose practices
              where you look at the total  exposure to a critical group. The idea of a dose constraint was
              encouraged for certain practices, whereby source specific  limits would be set below the
              overall public dose limit of one millisievert per year.  They suggested that constraints
              should be on the order of 0.25 mSv and  that you  could  still  optimize below those
              constraints.

              The suggested worker limit of 100 mSv in five years (10 rem per five years), or an average
              of two rems per year, was a decrease from five rems per year.   For the public again, the
              threshold is 1 mSv or 100 mrem/year. For dosimetry, they established the term, effective
              dose, which took advantage of the new biokinetic models and changed the tissue weighting
              factors from a fixed defined quantity to suggested quantities.
oEPA
         162                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
      Prof. Clarke is now proposing to change the system of radiation protection to simplify it
      and to make it more coherent and explicit.  The main points of what Roger Clarke is now
      talking about is what he calls a shift from a utilitarian to an egalitarian system. A utilitarian
      system would be  where you focus on protecting the population as  a  whole  while
      occasionally not having to worry  about every individual.   In this system, an occasional
      individual might get a larger dose than the recommended limit for the average member of
      the  group.  In the  new system, the  egalitarian system, protection of every individual  is
      given a little more emphasis.

      So in the egalitarian  system, you're  focusing your  efforts on protecting an individual as
      opposed to protecting society as a whole. These are two different philosophies, but both
      have the overall goal  of protecting the  members of a population.   Chairman  Clarke
      proposes to reform the principle of optimization, including abandonment of collective dose,
      and to replace ALARA  (as low as reasonably achievable) with ALARP (as low as
      reasonably practical),  hi this  case, the EGRP cannot see the  benefit of changing the
      terminology.

      Prof. Clarke  is also proposing to replace the public and worker dose limits with a series of
      bands, called Protective Action Levels, which are multiples of natural background. The
      higher the band, the more  serious the threat and  the more immediate the concern for
      protecting exposed individuals.   The  latest  proposals are also  moving  away from the
      distinction between practice  and intervention, which is more in line with what we do in the
      United States.  We  don't make that  distinction as yet,  and because of  the proposed
      protective action  levels, you might see  a move away from a precise distinction between
      worker and public dose limits. You would have just a series of bands of protection.  Also,
      by moving to these Protective Action Levels, there is no longer a single public dose limit
      covering all sources of exposure.

      Prof. Clarke  is also  proposing to simplify  the WR and WT,  the radiation and  tissue
      weighting factors.  As I mentioned earlier, these ideas are continuing to change and evolve.

      In surveying the NEA member countries and discussing these ideas within the EGRP, we
      identified several areas of concern.   The first concern is  centered around the issue of
      whether a change is needed.  The American phrase, "If it ain't broke, don't fix it," was said
      in much more eloquent terms by our European and Asian counterparts.  But there were a lot
      of questions about whether the system really was broken and whether there was a pressing
      need for changing it at this  time.  Some people are asking whether the proposed system
      would be relaxing the level of protection that we've come to expect under the ICRP 60 and
      previous recommendations.

      As you heard yesterday from Keith Eckerman's presentation, there is talk about changing or
      re-evaluating the  tissue weighting  factors.  There was  a question particularly among the
      Europeans who have  just  adopted  ICRP 60 dosimetry as a European Union directive, of
      whether we really need to  change these factors.  Is the difference really worth the cost and
      the  effort of changing? Also, a lot  of people were upset  with  the idea of abandoning
      collective dose.

      In the United States, where we have to do regulatory impact or cost benefit analyses, if we
      don't have  a collective dose approach, we really don't  have a tool  to make  those
      determinations. And then  there was a lot of feeling  among members of the EGRP against
                                                                                         &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                l 63

-------
              changing  familiar terminology, for example  changing ALARA,  which  we are all
              comfortable with, to ALARP.

              At one point, there seemed to be a move towards keeping practice and intervention as
              concepts but changing what they were named. There was a considerable resistance to this
              approach,  particularly among the EU countries.  With many countries having just adopted
              ICRP 60,  they want a period of stability.  They  really don't support major changes in the
              near term. They prefer evolution of the system of radiation protection, not revolution.

              Among the principles of justification, optimization, and limitation, many agree that the idea
              of optimization should be less quantitative, more  of a common-sense approach, than a  strict
              quantitative assessment. Roger Clarke is concerned that people are going to great extremes
              and doing things like multi-attribute analyses and intricate cost benefit analyses to do the
              optimization, when what is needed is a practical, common-sense and far less quantitative
              approach.   While agreeing with this sentiment,  most on the EGRP support retaining the
              concept of optimization and ALARA.

              When we got into our discussions of justification, our group was asking questions about
              who is doing the justification. What is the role of the stakeholder?  What is the role of the
              government regulator?  The buzzwords now at conferences are "decision-making" and
              "decision-aiding."  The regulators and the governments  are the decision-makers. But the
              stakeholders, citizens groups, environmental groups and industry  groups are considered
              decision-aiders.  I think, as a result of this new emphasis on stakeholder involvement, it's
              begun to influence the ICRP's view of how radiation protection should be done.  There is
              now more explicit language about involvement of the public and groups within the public.
              So ICRP's ideas are in flux.

              Quite a while  back, Roger Clarke quit using the  phrase "controllable dose,"  although,
              "controllable" is seeping back in. I would say that the modifications and the writings that
              you see coming from the ICRP Chairman are increasingly less radical. At first, it seemed
              he was proposing a complete departure from ICRP 60. Now I think you're seeing elements
              of ICRP 60 being put back into the proposals. And  I think they're still looking at 2005 for
              the next recommendations, although that may be optimistic. I am sure that these proposals
              will continue to evolve  and I encourage you to stay informed and take advantage of the
              ICRP's invitation for stakeholder involvement in this process.
oEPA
 .-rf^fi-    164                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
  -V^t"

-------
UNFINISHED BUSINESS: ASSESSING GENETIC AND FETAL RISKS

  NEAL NELSON
      U.S. Environmental Protection Agency, Radiation Protection Division

      The Uranium Fuel Cycle documents for 40 CFR 190, written in the mid-1970s, was the
      first time the numerical risk developed in the 1972 BEIRI report was used in assessing the
      potential lifetime risks following exposure to radioactive environmental contamination. The
      original 1972 BEIR report was based on the 1967 US population. However, when looking
      at the population tree, it's clear that a real population tree suggests many types of losses. In
      short, you would have to have a spontaneous generation of people to fill in  the age gaps,
      which were left by these losses, such as  wars and epidemics.  Switching to  a life table, a
      stationary population would allow the estimates of population effects to be stable as long as
      the stationary population was used. The initial aim of these population estimates was to be
      able to make comparisons between alternatives or regulations.

      Use of risk assessment models and research on them has taken on a life of it's own since
      the mid-1970s.  EPA has been collaborating with people at Oak Ridge since the 1970s, and
      the first combined dose/risk pathway risk assessment model, DARTAB, came out in about
      1981. EPA has continued its long history of collaboration with Oak Ridge, most recently,
      with the DCAL program.

      However, throughout the whole process,  a number of things have been left undone.  Most
      of our dose estimates are really based on specific activity, activity per gram of tissue, which
      had been converted to energy deposit per gram of tissue. This is fine for a constant external
      source.  But, it's not good for alpha emitters, for many of the electron emitters, nor is it
      good for some aspects of gamma radiation or X-rays.

      Most unfinished business relates to  problems associated  with dose  and/or risks from
      internal emitters.  Until now, the  estimation of internal emitter dose based  on what was
      essentially  a specific activity  calculation, i.e.  activity per  gram converted to energy
      deposited per gram, was acceptable. However, advances in radiation biology  have allowed
      identification of "sensitive cells" in several organs with promise of further identification of
      target cells in other organs.  To determine the dose to  these more localized sites will be
      difficult.

      Local dose from alpha emitters and beta/gamma emitters deposited in tissue may extend for
      a  few cell diameters around the source  of an emitted radiation. Alpha particles have a
      maximum range in tissue on the order of 50 to 100 microns depending on energy (about 10
      \\.  per MeV of energy).  Beta particles and electrons have  maximum ranges in tissue of
      microns to centimeters depending  on energy.  A 100 keV (or  less) particle has a range of
      about 200 microns (or less). Many emissions may have these energy levels ( conversion
      electrons, capture electrons, Auger electrons).  In addition,  some low energy x-rays (<10
      keV) are produced by beta emitters. These can enter localized  photoelectric interactions in
      cells causing short-range electron radiations.

      Is such localized radiation hitting target cells [which we  don't necessarily know either]? Is
      this why calculated risks for internal emitters don't always match observations? Improved
      organ data [voxels] won't help.  What about Auger electrons?
RADIATION RISK ABSESSMENT WORKSHOP PROCEEDINOB                               165

-------
     The need for better data and models for localization of internal  radiation  emitters has
     already come to the attention of the radiotherapists. The need to accurately target radiation
     dose from .deposited radionuclides is particularly  important for targeted radiation therapy
     using alpha emitters, Auger electron emitters, and some high-energy beta emitters. While
     considering  MIRD  calculations  adequate  for  diagnostic   uses,  they  are  adapting
     microdosimetry with Monte Carlo, dose point-kernel, voxel S factor, and other approaches
     to more targeted dose calculations. (Three references include: 1) H.  M. Thierens, M. A.
     Monsieurs, B. Brans, T. Van Driessche, I. Christiaens and R. A. Dierckx, "Dosimetry from
     organ to cellular dimensions", Comput. Med. Imaging Graph. 25(2): 187-193 (2001); 2)
     M. Bardies  and P. Pihet, "Dosimetry and microdosimetry of targeted  radiotherapy", Curr.
     Pharm.  Des. 6(14): 1469-1502  (2000); and 3) P. B.  Zanzonico,  "Internal  radionuclide
     radiation dosimetry: a review of basic concepts and recent developments", J. Nucl. Med.
     41(2): 297-308 (2000)).  We need at least as accurate a localization  of the dose  from
     deposited radionuclides.  This would also include local doses from K and L shell x-rays.

     In  their discussion of gamma photon absorption by photoelectric processes, Mine and
     Brownell (Radiation Dosimetry, G. J. Hine and  G.  L. Brownell,  Academic Press,  New
     York, 1956) defined the photoelectric absorption coefficient  as having two  components.
     One was the fraction of total photon energy absorbed and turned into  electron motion, the
     second was  the part  of photon energy radiated from  the site of  interaction.  Hine and
     Brownell concluded,  "For low  atomic  number materials in  which we are  interested, a
     different state of affairs exists. Now the binding energy is very small, being of the order of
     500 eV  for tissue. Thus the photoelectron acquires almost all of the energy of the photon,
     and the fluorescent radiation (~ 500eV) is so soft as to be absorbed at its point of origin and
     converted into electronic motion.  Thus in low atomic number material one  can consider
     that all the energy of the photon is truly absorbed when a photoelectric  process occurs."

     Some common radionuclides have gamma emissions <100 keV, some  < 10 keV.  As Table
     1  below shows,  photons  with energies  around  5  keV  are  absorbed and generate
     photoelectrons within microns of the site of emission.  When (not if)  sensitive cells are
     located  within microns of the site of emission, it will be of more  importance to quantify
     these local doses.
                                      TABLE  1 :
           5 TO 1 OO KEV PHOTONS ENERGIES,  COEFFICIENTS AND RANESES
APPROXIMATE
ENERGY (KEV)
5
10
15
20
30
40
50
80
100
PHOTOELECTRIC
ABSORPTION
COEFFICIENT
37.3
4.5
1.3
0.5
0.15
0.065
0.039
0.025
0.025
RANGE IN MM (1/E) (ABSORBING
63% OF PHOTON ENERGY)
0.27
2.22
7.69
20.0
66.7
153
256
400
400
(270 11)
(2.2 mm)
(7.7 mm)
(2.0 cm)
(6.7 cm)
(15.3cm)
(25.6 cm)
(40.0 cm)
(40.0 cm)
166
                                   RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     Examination of radioactive decay schemes tabulated in Nuclear Data (NuDat) Retrieval
     files show that in addition to the alpha and beta emissions [eg. Auger electrons], x-rays
     with energies below 10 keV can be emitted frequently. See examples below.
DECAY RADIATIONS
Mass Number:
Element:
T'A:
Decay Mode:
Sort order:
137 | Radiation:
CS 1 Radiation Energy (keV):
| Radiation Intensity:

Mass number, Proton number, Half-Life, and Radiation





A ELEMENT Z
137 CS 55
137 CS 55
137 CS 55
137
137
137
137
137
137
137
137
137
137
137
CS
CS
CS
CS
CS
CS
CS
CS
CS
CS
CS
55
55
55
55
55
55
55
55
55
55
55
Decay
Mode
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
Half-Llfe
30.07 Y 0.03
30.07 Y 0.03
30.07 Y 0.03
30.
30.
30.
30.
30.
30.
30.
30.
30.
30.
30.
.07
.07
.07
.07
.07
.07
.07
07
.07
.07
.07
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
.03
.03
.03
.03
,03
.03
,03
03
.03
,03
03
Rad.
Type
B-
B- TOT
B-
B-
E
E
E
E
G
G
G
G
G
G

AU
AD
CE
CE
X
X
X
X



L
K
K
L
L
KA2
KA1
KB


Radiation End-point Radiation
Energy Energy Intensity Dose
(keV) (keV) (%) (G-RAD/OC:
174.32 0.07 513.97 0.17 94.40 0.20 0.351
187.87 0.07 100.0 0.3 0.400
300.57 0.07 892.13 0.20 0.00058(8) 0
416.26 0.08 1175.
3.670
26.40
624.216 0.003
655.668 0.003
4.470
31.8171(3)
32.1936(3)
36.40
283.50 0.10
661.657 0.003
63 0.17 5.
7
0.
7.
1.
1.
1.
3.
1.
,60 0
.2
.757 0.
.66 0
.39 0
,0
,96 0
62 0
.32 0
.20
0.5
024
.23
.05
0.3
.06
.11
.05
0.00058 (8)
85.
10 0
.20
0
0
0
0
0
0
0
0
0
0
1
.0497
.0006
.0004
.102
.0194

.0013
.0025
.0010

.20
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                   &EPA
                                                                              167

-------
DECAY RADIATIONS
Mass Number:
Element:
T'/k
Decay Mode:
Sort order:
131
I


Radiation:
Radiation Energy (keV):
Radiation Intensity:

I
r
i
1
i
Mass number, Proton number, Half-Life, and Radiation !


Decay
A ELEMENT 2 Mode
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
53 B-
Rad.
Half-life
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8
8.
8 .
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
02070(11)
02070(11)
02070(11)
02070 (11)
02070(11)
02070(11)
02070 (11)
02070 (11)
02070 (11)
02070(11)
02070 (11)
02070 (11)
02070 (11)
02070 (11)
02070(11)
02070(11)
.02070 (11)
02070 (11)
.02070 (11)
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
. 02070 (11)
.02070(11)
02070 (11)
.02070(11)
.02070 (11)
.02070(11)
.02070(11)
.02070 (11)
. 02070 ( 11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
.02070 (11)
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
.02070 (11)
.02070(11)
.02070(11)
.02070 (11)
Type
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
B-
B-
B-
B-
B-
B-
B-
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E



TOT



AD L
AD K
CE K
CE K
CE L
CE M
CE N+
CE L
CE M
CE N+
CE K
CE 1
CE M
CE K+
CE K
CE L
CE M
CE H+
CE K
CE K
CE K
CE L
CE K
CE M
CE N+
CE L
CE M
CE K
CE N+
CE K
CE L
CE K
CE M
CE L
CE M
CE L
CE M
CE N+
CE 1
Radiation
Energy
(keV)
69.36 0.19
86.94 0.20
96.62 0.20
181.92 0.24
191.58 0.23
200.22 0.23
283.24 0.23
3.430
24.60
45.6236(20)
51.34 0.20
74.7322(21)
79.0430(23)
79.9770(23)
80.45 0.20
84.76 0.20
85.69 0.20
142.6526(20)
171.7612(21)
176.0720(23)
177.0060 (23)
197.62 0.15
226.73 0.15
231.04 0.15
231.97 0.15
237.937 0.017
249.744 0.005
261.24 0.20
267.045 0 . 017
267.84 0.20
271.356 0.017
272.290 0.017
278.852 0.005
283.163 0.005
283.527 0 016
284.097 0.005
290.09 0.03
290 35 0.20
291.228 0.004
294.66 0.20
296.95 0 20
301.26 0.20
312.635 0.016
316.946 0.016
317.880 0.016
319.20 0.03
End-point
Energy
(keV)
247.9 0.6
303.9 0.6
333.8 0.6
Radiation
Intensity
Dose
(%) (G-R&D/ncr
2.10 0.03
0.651 0.023
7.27 0.10
100.5 0.8
606.3 0.6 89.9 0.8
629.7 0.6
806.9 0.6



0.


0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.

0
0.
0
0
o
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0.050 0.023
0.480 0.010
5.1 0.3
0.604 0.014
3.54 0.16
00013(9)
0.464 0.021
0.094 0.005
0239 0.0011
00004(5)
.000011(10)
000003(2)
0507 0.0017
0114 0.0004
00237(9)
000574(20)
.00025(4)
000048 (7)
.000010(1)
.0000025(4)
.00264(10)
0.252 0.008
.000017 (8)
.000358 (22)
. 000042 (6)
.000072(5)
.0000179(9)
.0439 0.0014
.0090 0 0003
.00236(9)
.00221(7)
.00060(8)
000002(1)
.0078 0.0007
.0000004(2)
.0000053(7)
.0000010(1)
.000303(12)
.000061(2)
.0000155(6)
.000085(12)
0.0031
0.0012
0.0150
0.389
0.367
0.0002
0.0029
0.0004
0.0003
0.0034
0
0.0007
0.0002
0
0
0
0
0.0002
0
0
0
0
0
0
0
0
0 . 0013
0
0
0
0
0
0.0003
0
0
0
0
0
0
0
0
0
0
0
0
0
&EPA
        168
                                    RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
131
131
131
131
131
131
131
131
131
I
I
I
I
I
I
I
I
I
8.02070(11)
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
131
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
53
53
53
53
53
53
53
53
53
D
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
53
B-
B-
B-
B-
B-
B-
B-
B-
B-
E CE
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
B-
8.
8
8
8
8
8.
8
8
8
N+
8
8
8
a
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
8
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
358.19
.02070(11)
.02070(11)
.02070 (11)
.02070(11)
.02070(11)
02070 (11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
.02070(11)
.02070(11)
.02070 (11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070(11)
.02070 (11)
.02070(11)
D
D
D
D
D
D
D
D
D
0
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
D
E
E
E
E
E
E
E
E
E
.20
E
E
E
E
E
E
E
E
E
E
E
E
E
E
E
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
G
CE
CE
CE
CE
CE
CE
CE
CE
CE

CE
CE
CE
CE
CE
CE
CE
CE
CE
CE
CE
CE
CE
CE
CE
X
X
X
X


















L
M
K
H+
M
Kt
K
L
M

L
M
H+
K
L
M
m-
K
L
K
K
L
L
K
L
L
KA2
KA1
KB


















320.336 0.004
323.51 0.03
323.84 0.20
324.44 0.03
324.647 0.005
325.581 0.005
329.928 0.005
352.95 0.20
357.26 0.20
0.0000025(9) 0
359.036 0.005
363.347 0.005
364.281 0.005
370.253 0.004
399.361 0.004
403.672 0.005
404.606 0.005
468.443 0.004
497.551 0.004
602.428 0.004
608.158 0.005
631.536 0.004
637.266 0 005
688.350 0.005
717.458 0.005
4.110
29.4580(10)
29.7790(10)
33.60
80.1850(20)
85.90 0.20
177.2140 (20)
232.18 0.15
272.498 0.017
284.305 0.005
295.80 0.20
302.40 0.20
318.088 0.016
324.65 0.03
325.789 0.004
358.40 0.20
364.489 0.005
404.814 0.004
503.004 0.004
636.989 0.004
642.719 0.005
722.911 0.005
0
0
0
0
0
0

0
0


0
0
0
0
0
0
0
0
0
0
0
0
0
0





0

0
0

0
0
0
0



0




.00101(9)
.000017(3)
.00034(13)
. 0000042 (7)
.000203(18)
.000052(5)
1.55 0.07
.000050(18)
.000010(4)

0.246 0.011
.0507 0.0022
.0123 0.0006
.00083(8)
.000115(5)
.000024(1)
.0000060(3]
.00269(12)
.000389(17)
.0288 0.0013
.00085(3)
.00395(18)
.000117(5)
.0070 0.0004
.00087(4)
0.57 0.18
1 38 0.04
2.56 0.06
0.91 0.03
2.62 0.04
.00009(5)
0.270 0.004
.0032 0.0004
.0578 0.0012
6.14 0.07
.0018 0 0009
.0047 0.0006
.0776 0.0017
.0212 0.0025
0.274 0.022
0. 016 O.OOc
81 7 0.8
.0547 0.0017
0.360 0.004
7 . 17 0.10
0.217 0.005
1.77 0.03
0
0
0
0
0
0
0
0






.0109

0131 I

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0

.0019
.0004







.0004



.0001


0009
.0016
.0007
.0045

.0010

.0003
.0372


.0005
.0001
.0019
.0001
.634
.0005
.0039
.0973
.0030
.0273
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                       &EPA
                                                                   169

-------
DECAY RADIATIONS
Mass Number:
Element:
T'/2:
Decay Mode:
Sort order:

210
PB


Mass number,

! Radiation:
| Radiation Energy (keV):
i Radiation Intensity:
|
Proton number, Half-Life, and Radiation





	 __ 	 )
Radiation End-point Radiation
Decay
A ELEMENT Z Mode
(G-RAD/UCI-H)
210 PB 82 ft. 22
210 PB 82 B- 22
210 tB 82 B- 22
210 PB 82 B- 22
210 PB 82 B- 22
210 PB 82 B- 22
210 PB 82 B- 22
210 PB 82 B- 22
210 PB 82 B- 22
Rad.
Half-Life Type

.3 Y 0.2 A
.3 Y 0.2 B-
.3 Y 0.2 B- TOT
.3 Y 0.2 B-
. 3 Y 0 . 2 E An L
.3 Y 0.2 E CE L
.3 Y 0.2 E CE M
.3 Y 0.2 G X L
.3 Y 0.2 6
Energy Energy
(keV) (keV)

3720. 20.
4.16 0.13 16.6 0.5
6.08 0.17
16.16 0.13 63.1 0.5
8.150
30.1515(11)
42.5399(11)
10.80
46.5390(10)
Intensity
(*)

0.0000019(3)
84 3.
100. b
16. 3.
35. 3.
60. 3 1.9
14.3 0.5
25. 3.
4.25 0.04
Dose
(G-RAD
/UCI-H)

0
0.0074
0.0130
0.0055
0.0061
0.0388
0.0129
0.0058
0.0042
DECAY RADIATIONS
Mass Number:
Element:
TA:
228
RA

Radiation:
Radiation Energy (keV):
Radiation Intensity:






Decay Mode:
Sort order:

Decay
A ELEMENT Z Mode
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 68 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
228 RA 88 B-
Mass number,

Rad.
Half -Life Type
5.75 Y 0.03 B-
5.75 Y 0.03 B-
Proton number, Half-Life, and Radiation
Radiation End-point
Energy Energy
(keV) (keV)
3.21 0.23 12.8 0.9
6.48 0.23 25.7 0.9
5.75 Y 0.03 B- TOT 7.2 0.3
5.75 Y 0.03 B-
5.75 Y 0.03 B-
5.75 Y 0.03 E CE
5 75 Y 0.03 E CE
5.75 Y 0.03 E CE
5.75 Y 0.03 E CE
5.75 Y 0.03 E CE
5 75 Y 0.03 E AD
5.75 Y 0.03 E CE
5.75 Y 0.03 G
5 75 Y 0.03 G
5.75 Y 0 03 G X
5.75 Y 0.03 G
5.75 Y 0.03 G
5.75 Y 0 03 G
9.94015 39.2 0.9
10 . 04 0 .25 39 . 6 0.9
M 1.28 0.03
M 1.668 0.021
L 6.56 0.10
M 7.75 0.05
M 8.518 0.021
L 9.280
M 21 40 0.10
6.28 0.03
6.670 0.020
L 12.70
12.75 0.05
13 520 0.020
26.40 0.10
Radiation
Intensity
U )
30. 10.
20. 6.
100 12.
40.00
10.00
7.500
37 .50
2.211
2.250
7.31 0.23
1.08 0.11
0.5910
0.0000014
0.0000311(6)
1.13 0.11
0 . 30 0 . 07
1.600
0.01402
Dose
(G-RAD
/UCI-H)
0.0021
0.0028
0.0154
0.0085
0.0021
0.0002
0.0013
0.0003
0.0004
0.0013
0.0002
0.0003
0
0
0.0003
0
0.0005
0
170
                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
DECAY RADIATIONS
Mass Number:
Element:
TA.
Decay Mode:
Sort order:
1 238
' U
!
i
1 Mass number,
| Radiation:
' Radiation Energy (keV):
! Radiation Intensity:


	 	 	 	


Proton number, Half-Life, and Radiation
A ELEMENT Z
238 U 92
238 0 92
238
238
238
238
238
238
238
238
238
238
238
238
238
238
238
U
U
0
U
U
D
tJ
U
0
U
U
U
D
D
U
92
92
92
92
92
92
92
92
92
92
92
92
92
92
92
Decay
Mode Half-Life
A 4.468E+9I3)
A 4.468E+9(3)
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
4
4.
4
4
4
4
4
4
4
4
4.
4.
4
4
4.
.468E+9I3)
.468E+9I3)
.468E+9(3)
.468E+9(3)
.468E+9I3)
.468E+9I3)
.468E+9I3)
.468E+9I3)
.468E+9I3)
.468E+9(3)
.468E+9I3)
.466E+9(3)
.468E+9I3)
.468E+9(3)
.468E+9I3)
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Rad.
Type
A
A
A
E
E
E
E
E
E
E
E
G
G
G
G
G
G

CE K
AD L
CE L
CE M
AU K
CE L
CE M
CE N+
X L

X KA2
X FA1
X KB

Radiation End-point Radiation
Energy Energy Intensity Dose
(keV) (keV) (%) (G-RAD/DCI-H)
4038. 5. 0.078 0.012 0.0067
4151. 5. 21. 3. 1.85
4198.
3.
9.
29.
44.
69.
93.
108.
112.
13.
49.
3.
.85 0.10
.480
08 0.06
.37 0.06
.20
03 0.10
32 0.10
.17 0.10
00
55 0.06
89.9530(20)
93.3500(20)
105.
113.
0
50 0.10
79. 3.
0.0024 0.0004
7.4 1.3
15.3 2.0
4.2 0.6
0.000058 (9)
0.047 0.008
0.0131 0.0020
0.0048 0.0008
8.0 1.3
0.064 0.008
0.00070(11)
0. 00114(17)
0.00053(8)
0.0102 0.0015
7.06
0
0.0015
0.0095
0.0039
0
0
0
0
0.0022
0
0
0
0
0
      There are eight areas that need further attention:

      1)  Female Risk.   EPA calculations of dose and risk are for the adult hermaphrodite.
      Separate dose and risk calculations should be made for adult male and adult female since
      reference man has mass data for both. The mass files for 15,  10, 5, and 1 year old and for
      the newborn include masses for ovaries, uterus and breast and so could be split into a male
      file and a female file. If files are kept separate, changes can be made as new data becomes
      available  without mixing up the male/female data.  This separation is more important for
      adults [pregnancy, lactation] than it is for younger ages.

      While this discrepance is present in all current risk calculations, it has a major effect only
      for certain radionuclides,  organs, and ages.   The greatest effect can be seen in the bone
      seeking radionuclides, e.g., Ra, U, etc. and to a lesser extent  in radionuclides  deposited in
      liver, (e.g., Th, Fe, etc.).   The organs involved are of course those that have a large
      difference between male  and female organ masses.   The risks at different ages are  a
      reflection of the interaction of the biokinetic/dose models  and  the  organ risk models.
      Results are often quite interesting.

      An example is given In Table 2 below. Comparison is made between females (Using all
      female physiological data) and hermaphrodite females (Female genital organs added to
      adult male physiology).  The female is higher reflecting the decreased mass, particularly for
      the skeleton.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                     1 71

-------
               A COMPARISON
                TABLE 2:
OF RISKS IN COHORTS WITH DIFFERENT STARTING AC3ES FOR
 INGESTION OF 1 BQ IN DRINKING WATER.
MORTALITY
RISKS
RA-226
Female
Hermaphrodite
Female
RA-228
Female
Hermaphrodite
Female
RA-226
Female
Hermaphrodite
Female
RA-228
Female
Hermaphrodite
Female
CANCER
AGE
(YRS)
0-5
5-15
15-25
25-70
0-110

bone
bone
F2
F2
3.526E-09
3.507E-09
6.808E-09
5.831 E-°9
9.136E-09
7.253E-09
1.173E-09
8.799E-10
2.365E-09
1.896E-09

bone
bone
F2
F2
1.820E-08
1.816E-°8
2.085E-08
1.840E-°8
2.193E-08
1.720E-"8
2.494E-09
1875E-OS
6.069E-09
4.967E-"9

Total
Total
F
F
4.61 4E-°8
4.595E-08
3.153E-08
2.905E-08
2.596E-08
2.139E-08
4.275E-09
3403E-"9
9.147E-09
7.895E-09

Total
Total
F
F
1.694E-07
1.689E-07
1.039E-07
9.71 2E-°8
6.581E-08
5.486E-"8
9.301 E-°9
7.543E-09
2.533E-08
2.248E-08
              2) Non-cancer risks.  Literature on the effects of ionizing radiation on animals always
              referenced "non-specific life shortening" as one component of the long-term effects of
              radiation exposure. For a long time this was pointed out as one of the differences between
              effects in animals and effects in man.  In man, there were no "non-specific life shortening
              effects", all life shortening was due to cancer mortality.

              Now, there are reports of "non-specific life shortening effects" in humans.  Because of the
              good dosimetry and comparison groups in the Japanese Atomic Bomb Survivors, Cologne
              and Preston (J. B. Cologne and D. L. Preston, "Longevity of Atomic-Bomb  Survivors",
              The  Lancet, 356: 303-307  (2000), we  have been able to estimate a  "non-specific life
              shortening effect'" of about l%-2% per Gray.

              Earlier, Wong, et al (F. L. Wong, M.. Yamada, H. Sasaki, K. Kodama.  S. Akiba, K.
              Shimaoka and Y. Hosoda, "Noncancer Disease Incidence in the Atomic Bomb Survivors:
              1958-1986", Rad. Research, 135: 418-430 (1993)) had reported that a  significant excess
              risk  of nonmalignant disorders was  becoming  evidenced.  Increased risks of uterine
              myoma,  liver disease and cirrhosis, and nonmalignant thyroid  disease  were noted.  The
              increased risk of nonmalignant thyroid disease was noted particularly in those who were
              age 20 or younger at the time of exposure and a risk of myocardial infarction in those who
              were age 40 or younger at the time of exposure.  The later supports an earlier finding of a
              dose related increase in coronary  heart disease  mortality. (Y.  Shimizu, H. Kato, W. J.
              Schull and D. G. Hoel, "Studies of the mortality of A-bomb survivors. Report 9. Mortality,
              1950-1985: Part  3. Noncancer mortality based on the revised doses (DS86), Radiat. Res.
              130: 249-256 (1992)). [RERF TR 2-91].
oEPA
                                           RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     If these findings continue to be  supported,  and perhaps  other non-cancer effects be
     identified, establishing the organs and/or cells at risk for these effects will not be a trivial
     task. Developing the biokinetic and dosimetric models will also be a difficult task.

     3) Embryo/Fetus Dose and Risk (Cancer, Teratologic, and other Risk).  Dosimetry for
     internal  emitters has not been established in the case of exposure of the embryo/fetus.
     Doses  from  radionuclides  transferred  from  the maternal  blood  stream  and  from
     radionuclides deposited locally in maternal tissues must both be considered.  The major
     difficulties will be the; changes in the placenta, changes in size of organs, developmental
     level of tissues, changes in metabolism; the complex development from conception to birth.
     Perhaps time of exposure should be expressed in trimesters. See Table 3.

     The dose response is not well defined; mental  retardation may be non threshold, whereas
     malformations are usually considered to  have  a threshold.  However, some animal  data
     appears linear down to one rem, the lowest dose tested. It is prudent to consider all non
     threshold unless proven otherwise.  Preimplantation loss appears to be a lethal response and
     is expected to have  a threshold.  If the threshold is low, you could  see no evidence  of it
     anyway.

                                        TABLE 3:
                 POSSIBLE EFFECTS OF IN UTERD RADIATION EXPOSURE
TYPE OF RISK TO CONCEPTUS
Cancer Incidence
Mental Retardation a (exposure at 8-15 weeks)
Mental Retardation b (exposure at 16-25 weeks)
Malformation b (exposure at 2-8 weeks)
Pre-implantation Loss (exposure at 0-2 weeks)
RISK PER RAD
6x10-"
4x10-3
1x10-3
5x10-3
1x10-2
      a A threshold for mental retardation following exposure at 8-15 weeks of gestational age may depend on the
      mechanism of action.

      b A threshold is expected for mental retardation following exposure during the 16-25 week period of gestation
      and for many types of malformations following exposures at early gestational age.

      4) Breast Milk, Transfer of internal emitters from the maternal blood stream to the breast
      and then to milk provides another complex dose scenario.  The extent to which milk  can
      concentrate radionuclides is  unclear. Breast and  glandular tissue sizes change during
      pregnancy and lactation. The size of the reservoir for milk  and its distribution in  the
      mammae is of possible importance.  Localized radiation dose may be the most important
      dose for the mother, but  radionuclide concentration potential most important for neonate
      doses.

      5) Genetic Risk.  Genetic disorders are still expected to be induced by ionizing radiation
      dose to the gonads, but dose response estimates are not certain.  Induction of genomic
      instability leading to later increased susceptibility to carcinogens is a distinct possibility.
      Assessing these risks  may require dose  estimates at the level  of the  cell nucleus and
      cytoplasm.   Use  of RBEs derived  from  cellular  studies and  animal  studies may be
      misleading. Human germ cells may not have the same sensitivity as those from laboratory
      animals. See Table 4.
                                                                                           &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                                TABLE 4:
                   ESTIMATED FREQUENCY OF GENETIC DISORDERS IN A BIRTH COHORT
                    DUE TO EXPOSURE DF EACH OF THE PARENTS TO D.D 1  GY (1 RAD)
                                PER REPRODUCTIVE  GENERATION (3D YR).
RADIATION
Low Dose Rate, Low-LET
High Dose Rate, Low-LET
High-LET
SERIOUS HERITABLE DISORDERS
(CASES PER 106 LlVEBORN)
FIRST GENERATION
20
60
90
ALL GENERATIONS
260
780
690
              6) Lifetime Dose for Constant Intake or Discontinuous Intake.  At this time, there are
              dose conversion factors for 6 ages: newborn, infant, 5y, lOy, 15y, and adult. A consensus
              model for integration of dose across various age intervals up to lifetime would be useful.
              However, the six ages bring up another problem.  Human females become adults younger
              than males. When the female model is incorporated, there will have to be sex specific
              designation of some parameters. How this will be accomplished has not yet been defined.

              Likewise, eventually we must consider geriatric aspects of metabolism, physiology, etc. as
              they affect dosimetry. The adult model is that of a health young adult (age 25). Although
              only two percent  to three percent  of lifetime risk from  life  time exposure is  due to
              exposures from age 70 to age  120 (the end of the lifetable population), it would be nice for
              symmetry to address geriatric ages.  We go down hill after about age 40; and past age 60
              many of our physiological functions are running at 50 percent or less of what the were at
              age 25.  In an aging population, such as we now have, assessing geriatric dosimetry may
              take on greater importance. We need to get started now to address the problem.

              7) Short Half-life Isotopes. ICRP report 38 stopped too soon. Isotopes with half-lives less
              than 10 minutes were not included.  At the present time, some facilities, usually research or
              medical facilities, employ radionuclides with half lives less than 10 minutes.  These short
              half-life nuclides should be included in our dosimetry and risk models. The work of JAERI
              on development of data for the nuclear decay database using ENSDF and EDISTR should
              be continued. This work provides revised data and the decay schema for radionuclides not
              in ICRP 38, but which is required for extending our dose and risk files.

              8) Skin.  While ingestion, inhalation, submersion, and immersion pathways are considered,
              so far everyone has been  avoiding the problem of skin absorption.  We know it occurs; eg.
              for H-3, for Iodines, for noble gasses. Intake has not been quantified for adult males, much
              less  for females and non-adults.  This  may be an important pathway of exposure in some
              situations.  For example, ICRP estimates that 1/3 of the dose in an adult male exposed to an
              atmosphere with  tritiated water vapor  comes  through  skin absorption.  2/3  through
              inhalation. It is past time to start developing models for skin absorption.

              It is easy to see that much work remains to be done.  As new data is added and new models
              employed, we hope to provide a more realistic picture of  radiation doses and risks.  At
              least, we can hope.
3-EPA
          1 -74
                                            RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINQS

-------
                  DERIVATION DF CLEARANCE LEVELS FOR
                          SOLID  MATERIALS IN JAPAN

  AKIHIRO SAKAI AND M. OKOSHI
     Japan Atomic Energy Research Institute

  ABSTRACT

     To establish the clearance levels, the Nuclear  Safety  Commission (NSC) has  been
     discussing the  clearance levels since May  1997. The NSC derived the  unconditional
     clearance levels for the solid  materials, namely  concrete  and metal, arising  from the
     operation and  dismantling of nuclear reactors and  post irradiation examination (PIE)
     facilities. Two  destinations of the  cleared materials, namely disposal and  recycle/reuse,
     were  considered. Deterministic calculation models were established to assess individual
     doses resulting from 73 exposure pathways,  and realistic parameter values  were selected
     considering the Japanese natural  and social  conditions.  The  clearance  levels  for  21
     radionuclides of nuclear reactors and for 49 of PIE facilities were derived as radioactivity
     concentration equivalent to the individual doses of 10 uSv/y. Most of calculated clearance
     levels (e.g., y-ray emitters such as Co-60 and a-ray emitters such as Pu-239) were nearly
     the same as those shown in IAEA-TECDOC-855. Some (e.g. p-ray emitters  such as Tc-99
     and 1-129), however,  were different. It  is considered that the  major reasons depend on
     differences of fixed scenarios and of selected values of parameters.

  INTRODUCTION

     Contaminated solid materials with radioactivity are generated from the operation of nuclear
     facilities, medical and industrial radioisotope uses, and dismantling of nuclear facilities.
     Some of them can be  regarded as non-radioactive materials and may be released from the
     regulatory control,  because they only give  rise  to trivial radiation  hazards.  In IAEA-
     TECDOC 855 of 1996 [1], the release  of such materials from regulatory control is defined
     as "clearance" and the corresponding levels of activity concentrations are called clearance
     levels.

     In Japan, plans  of dismantling nuclear facilities have been going forward in recent years, so
     it has been expected to institutionalize  the clearance as early as possible to manage a  great
     deal  of solid  materials arising from  them, especially  nuclear  reactors,  safely  and
     reasonably.  With these points  as background, the Nuclear  Safety Commission  (NSC),
     which is one of the advisory organizations to the Japanese government, has been discussing
     the unconditional clearance levels, under which solid materials can be handled as  non-
     radioactive ones without any conditions, since May 1997.

     The NSC published  the reports of the clearance levels for solid materials arising from light
     water reactors (LWR) and gas cooled reactors (GCR) in March 1999[2], and heavy water
     reactors  (HWR) and  fast breeder reactors (FBR)  in  July 2001 [3].  The NSC has  been
     continuously discussing the clearance levels for solid materials arising from post irradiation
     examination (PIE) facilities,  hi order to support these NSC's  discussion,  Japan Atomic
     Energy Research Institute (JAERI) has been conducting technical analyses to derive the
     clearance levels.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               1 75

-------
           CLEARANCE LEVELS FOR THE SOLID MATERIALS ARISING FROM NUCLEAR
           REACTORS

              METHODOLOGY

              The applied methodology to derive the clearance levels consists of the following steps:
                  >   Establishment of dose criterion to derive the clearance levels.
                  >   Choice of a source term and characterization of this source from the physical,
                      chemical, and radioactive point of view.
                  >   Description of the pathways that can result in exposure to people.
                  >   Establishment of calculation models based on the pathways.
                  >•   Choice of values for the parameters of these scenarios.
                  >   Calculation of radioactivity concentrations equivalent to the individual dose
                      criteria.

              The outlines of each step mentioned above are described in the following sections.

              DOSE CRITERION

              Individual risk resulted from cleared materials must be sufficiently low not to warrant
              regulatory concern. The Radiation Council, which is another advisory organization to the
              Japanese government, states that the radiation control of the disposal site is not needed if
              the doses to individual of the  critical group due to the near surface disposal are less than 10
              uSv/y. [4] And also, the ICRP [5] and the IAEA [6] have suggested doses to individuals of
              the critical group of the  order of 10 uSv in a year from each exempt practice or source.
              Therefore the NSC applies 10 uSv/y to derive the unconditional clearance levels.

              SOURCE TERM

              Though both operation and dismantling  of  a  nuclear reactor generate  contaminated
              materials with radioactivity,  the quantities  arising  from the dismantling is much greater
              than that from the operation.  Additionally metal and concrete are major materials from the
              dismantling. Therefore, investigation was made on only contaminated metal and concrete
              from the dismantling. Table I shows the estimated amounts of materials arising from major
              types of reactors.

              In this analysis, it was assumed that the cleared materials were disposed of or recycled with
              the non-radioactive  materials. For disposal,  it was assumed that the amount of disposed
              materials was 500 thousands metric  tons and that the ratio of cleared materials  to the
              amount was 0.1.  For recycle, the assumed ratios of cleared materials to the  amounts were
              0.1 for concrete and 0.1 or 0.7 for metal depending on exposure pathways.
&EPA
         1 76                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                               TABLE  1:
          ESTIMATED AMOUNTS OF MATERIALS ARISING FROM THE  DISMANTLING  OF
                     NUCLEAR REACTORS  (UNIT: THOUSAND METRIC TONS)
REACTOR
TYPE
BWR
(110MWE)
PWR
(110MWE)
GCR
(160 MWE)
HWR
(FUGEN)
FBR
(JOYO)
WASTE
CATEGORY
MATERIALS
Metal
Concrete
Total
Metal
Concrete
Total
Metal
Concrete
Total
Metal
Concrete
Total
Metal
Concrete
Total
LLW
(I)
0.1
0
0.1
0.1
0.1
0.2
0.2
3
3
O.2
0
<0.2
LLW (II)
2
<1
2
2
<1
3
2
10
12
1
0
1
<1
0
<1
VLLW
<10
<10
10
2
1
3
3
5
8
2
1
3
<1
1
<2
BELOW
CLEARANCE
LEVEL
21
8
30
3
9
10
1
10
10
4
30
34
1
1
3
NON-
RADIOACTIV
E MATERIALS
8
487
500
34
443
480
6
115
720
10
320
330
2
280
282
TOTAL
40
500
540
40
450
500
10
140
160
20
350
370
<10
280
290
         Note 1; LLW (I) means the radioactive wastes whose radioactivity levels are greater than the upper bound
         concentrations for near surface disposal in Japan. On the other hand, LLW (II) means radioactive waste whose
         radioactivity levels are less than the upper bound concentrations for near surface disposal.
         Note 2: FUGEN is the heavy water moderated and boiling light water cooled reactor and JOYO is the experimental
         FBR operated in JNC.
         Note 3; The quantities of the materials below the clearance levels are estimated by using the representative values
         that are shown in IAEA -TECDOC-855 [1].
         Note 4: These numerous values are shown in this table are rounded So, the total is not equal to the sum of each
         numerous value.
RADIATION  RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                                           oEPA
                                                                                                     177

-------
              EXPOSURE PATHWAYS

              For this analysis, two scenarios were considered:
                  >   Landfill disposal in both terrestrial and marine environments, and
                  >   Recycling of steel and concrete, and reuse of equipment (with surface
                      contamination).

              The reason why the landfills in  both terrestrial and marine environments were considered
              was that both are  common methods for  disposal of municipal and industrial wastes in
              Japan.  These two scenarios were subdivided into various sub-scenarios and exposure
              pathways describing the activities of specific  individuals.  First, all possible exposure
              pathways, 202 total pathways, were considered. Then, the exposure pathways, which might
              result in small doses, were omitted from the consideration. Finally, 41 exposure pathways
              for disposal and 32 exposure pathways for recycle/reuse were chosen, which are shown in
              TABLE 2 and TABLE  3 respectively.  Humans  who are  involved  in these exposure
              pathways may be exposed to radiation in three main ways:
                  >   Exposure to external radiation.
                  >   Inhalation of radioactive gases or small particles.
                  >   Ingestion of foodstuffs containing radionuclides or radioactive material.

              The exposure ways were considered about  each exposure pathways.

              CALCULATION MODEL AND PARAMETERS

              In this analysis, deterministic calculation models  were established to assess individual dose
              from selected 73 exposure pathways. Numerical formulas to express calculation model are
              described in the NSC's report [2], and reference [7].

              The realistic parameter values,  namely  mean or most probable values were selected,
              considering natural and social conditions in Japan. It, however, was very  difficult to select
              an  appropriate value for each parameter,  especially for  ones  depending  on  natural
              conditions such as groundwater velocity and length of saturated zone. Therefore, the JAERI
              also performed a stochastic approach to validate the  calculation results obtained with the
              deterministic one [8]. On the basis of the results, the values of some  parameters were
              modified to more appropriate ones.

              The dose conversion factors  for inhalation  and ingestion were taken from the JAERI's
              reports [9] [10] based on International Commission  on Radiation Protection  Publication
              No. 30 and No.48 [11] [12]. All parameter values and calculation models used in this
              analysis are described in the NSC's report  [1].
oEPA
         178                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                TABLE 2:
       DESCRIPTIONS CDF EXPOSURE PATHWAYS FOR THE DISPOSAL SCENARIO
SUB-SCENARIO
OPERATIONS OF THE
WASTES DISPOSAL
DISTURBANCE OF THE
LANDFILL SITE AFTER
THE CLOSURE
GROUNDWATER
MIGRATION
SEA RECLAMATION
EXPOSURE PATHWAY
Loading wastes to truck
Transportation
Landfill
Inhalation of tritium gas
Construction of a house
Residence
Agriculture
Livestock farming
Ingestion of crops cultivated in the landfill
site
Ingestion of livestock grown with the feeds
cultivated in the site
Ingestion of well water
Irrigation cultivation for food crops with well
water
Irrigation cultivation for feed with well water
Ingestion of crops cultivated with well water
Ingestion of livestock grown with the feeds
cultivated with well water
Ingestion of livestock grown with well water
Ingestion of freshwater products cultivated
with well water
Fishery on the river
Swimming in the river
Ingestion of freshwater products caught in
the river
Activities on the river shore
Handling of the fishery nets
Ingestion of salt
Fishery on the marine
Swimming in the marine
Ingestion of products caught in the marine
Activities on the sea shore
Inhalation of the sprayed sea water
Handling of the fishery nets
INDIVIDUAL
CONSIDERED
Equipment operator
Truck driver
Equipment operator
Equipment operator
Off-site resident
Construction worker
Resident
Farmer
Farmer
Consumer
Consumer
Consumer
Farmer
Farmer
Consumer
Consumer
Consumer
Consumer
Fisherman
Swimmer
Consumer
Worker
Fisherman
Consumer
Fisherman
Swimmer
Consumer
Worker
Resident
Fisherman
CATEGORY OF
EXPOSURE
External Inhalation
External Inhalation
External Inhalation
Inhalation
External Inhalation
External Inhalation
External Inhalation
External Inhalation
Ingestion
Ingestion
Ingestion
External Inhalation
External Inhalation
Ingestion
Ingestion
Ingestion
Ingestion
External
External
Ingestion
External Inhalation
External
Ingestion
External
External
Ingestion
External Inhalation
Inhalation
External
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                          &EPA
                                                                     179

-------
                                       TABLE 3:
              DESCRIPTIONS OF EXPOSURE PATHWAYS FOR RECYCLE/REUSE SCENARIO
SUB-SCENARIO
SCRAP METALS
PROCESSING
CONSUMER USE OF
ITEMS MADE FROM
RECYCLED METAL
CONCRETE
PROCESSING
CONSUMER USE OF
ITEMS MADE FROM
RECYCLED
CONCRETE
REUSE
EXPOSURE PATHWAY
Unloading scrap metals
Transportation
Pretreatment
Smelting and casting
Treatment of slag
Fabrication
Inhalation of dust and ingestion of
vegetables contaminated with downwind
plume from the smelting factory
Use of the refrigerator
Use of the bed
Ingestion of food cooked with the frying
pan
Ingestion of the caned beverage
Residence in the room built with the
reinforcement bars
Ingestion of tap water through the water
pipes
Automobile
Truck
Ship
Desk
Numerical controlled lathe
Slag use in asphalt parking lot
Crushing concrete
Inhalation of dust and ingestion of
vegetables contaminated with downwind
plume from the concrete crushing factory
Residence in the room built with the
aggregates
Concrete use in asphalt parking lot
Reuse of equipment
INDIVIDUAL CONSIDERED
Worker
Driver
Worker
Worker
Worker
Worker
Downwind resident
Consumer
Consumer
Consumer
Consumer
Resident
Resident
Driver
Driver
Sailor
Office worker
Lathe operator
Manager of the parking lot
Worker
Downwind resident
Resident
Manager of the parking lot
Worker
CATEGORY OF
EXPOSURE
External
Inhalation
External
External
Inhalation
External
Inhalation
External
Inhalation
Inhalation
Ingestion
External
External
Ingestion
Ingestion
External
Ingestion
External
External
External
External
External
External
External
Inhalation
Inhalation
Ingestion
External
External
External
Inhalation
Ingestion
&EPA
        1 BO
                                     RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
DERIVATION RESULTS
      Using the refined parameter values, the clearance levels for major 21 radionuclides were
      derived. The Clearance  levels are  expressed  in units of Bq/g or Bq/cm2 equivalent to
      individual dose of 1 OuSv/yr. The minimum value of all exposure pathways was defined as
      unconditional clearance  level  for  each  radionuclide.  Table 4 shows the unconditional
      clearance levels and the limited scenario and exposure pathway caused the value.

                                       TABLE 4:
   DERIVED UNCONDITIONAL CLEARANCE LEVELS FDR  SOLID MATERIALS ARISING FROM
                NUCLEAR REACTORS AND LIMITING EXPOSURE PATHWAYS
RADIONUCLIDE
H-3
C-14
CI-36
Ca-41
Mn-54
Fe-55'1
Co-60
Ni-59
Ni-63
Zn-65
Sr-90
Nb-94
Tc-99
1-129
Cs-134
Cs-137
Eu-152
Eu-154
Pu-239
Am-241
CLEARANCE
LEVELS
(BQ/G)
200
5
2
80
1
3000*1
0.4
600
2000
1
1
0.2
0.3
0.7
0.5
1
0.4
0.4
0.2
0.2
LIMITING SCENARIO AND EXPOSURE PATHWAY
SCENARIO
Disposal
Recycle/
reuse
Disposal
Recycle/
reuse
EXPOSURE PATHWAY
Ingestion of crops cultivated in the landfill
site
Ingestion of freshwater products cultivated
with well water
Ingestion of crops cultivated with well water
External exposure on waste disposal
External exposure from reused equipment
External exposure on waste disposal
Ingestion of crops cultivated with well water
Ingestion of livestock grown with the feeds
cultivated in the site
External exposure on waste disposal
Ingestion of crops cultivated in the landfill
site
External exposure of the resident in the
landfill site
Ingestion of crops cultivated in the landfill
site
Ingestion of well water
External exposure on the asphalt parking
lot built with slag
Inhalation of dust on unloading scrap
metals
TECDOC-855 (Bo/G)
RANGES
1000-
10000
100-
1000
100-
1000
SINGLE
3000
300
300
N.A.*2
0.1-1
100-
1000
0.1-1
0.3
300
0.3
N.A."2
1000-
10000
0.1-1
1-10
0.1-1
100-
1000
10-100
0.1-1
0.1-1
0.1-1
3000
0.3
3
0.3
300
30
0.3
0.3
0.3
N.A.'2
0.1-1
0.1-1
0.3
0.3
*7 • The unit of the clearance level Jor Fe-55 is Bq/cm' because the limiting pathway is reuse of the surface contaminated equipment.
*2. Clearance levels for these radionuclides are not available in IAEA-TECDOC-855.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                    181

-------
            CLEARANCE LEVELS FOR SOLID MATERIALS ALIBINB FROM F*IE FACILITIES
               METHOOaLOBY

               In the derivation of the clearance levels for solid materials arising from PIE facilities, the
               applied methodology was  the  same  as one for nuclear reactors. Additionally the same
               source term, the same exposure pathways and the same deterministic calculation models
               were used, considering that the dismantling of PIE facilities generate mostly concrete and
               metal and that the amounts of them and the ratios of the cleared materials to the amounts
               were smaller than those of nuclear  reactors.  The  same  source term leads  conservative
               values in the derivation of clearance levels for PIE facilities. Table 5  shows the amounts of
               materials arising from the dismantling of major PIE facilities.

               To calculate the clearance levels, major 49 were selected among radionuclides contained in
               irradiated nuclear fuels and materials examined in PIE facilities. These include 13, which
               had already been discussed on the derivation for nuclear reactors.

               The same parameter values used in the discussion on nuclear reactors were adopted, and for
               new parameters peculiar to new 36 radionuclides, the realistic values were selected with the
               same way in the discussion on nuclear reactors

                                                  TABLE 5:
               THE ESTIMATED AMOUNTS OF MATERIALS ARISING FROM THE DISMANTLING OF
                                PIE FACILITIES (UNIT:  THOUSAND METRIC TON)
PIE FACLITIES
HOT
LABORATORY
FMF
WASTE
CATEGORY
MATERIALS
Metal
Concrete
Lead
Total
Metal
Concrete
Lead
Total
LLW(I)
0
0
0
0
0
0
0
0
LLW (II)
0.1
0
0
0.1
0.8
0
0
0.8
VLLW
0.05
0
0
0.05
0.2
0
0
0.2
BELOW
CLEARANCE
LEVEL
0.7
0.2
0.15
1
1
0.4
0.003
1
NON-
RADIOACTIVE
MATERIALS
0.3
21
0.05
21
2
68
0.3
70
TOTAL
1
21
0.2
22
4
68
0.3
72
            Note 1 • LLW (I) means the radioactive wastes whose radioactivity levels are greater than the upper bound concentrations for near
            surface disposal m Japan On the other hand, LLW (II) means radioactive waste whose radioactivity levels are less than the upper
            bound concentrations for near surface disposal
            Note 2: Hot Laboratory and FMF are representative PIE facilities operated in JAERI and JNC, respectively
            Note 3: The quantities of the materials below the clearance levels are estimated by using the representative values that are shown
            m IAEA -TECDOC-855 [1J.
            Note 4: These numerous values are shown m this table are rounded. So, the total is not equal to the sum of each numerous value.
oEPA
          1 82
                                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                  TABLE S:
 DERIVED UNCONDITIONAL CLEARANCE LEVELS FOR SOLID MATERIALS ARISING FROM PIE
                  FACILITIES AND LIMITING EXPOSURE PATHWAYS
RADIONUCLIDE

H-3'3
C-14'3
Sc-46
Mn-54'3
Fe-55*1*3
Co-58
Co-60*3
Zn-65*3
Sr-89
Sr-90*3
Y-91
Zr-95
Nb-94*3
Nb-95
Ru-103
Ru-106
Ag-108m
Ag-110m
In- 114m
Sn-113
Sn-119m*1
Sn-123
Sb-124
Sb-125
Te-125m
Te-127m
Te-129m
Cs-134'3
Cs-137'3
Ce-141
CLEARANCE
LEVELS
(BQ/G)
200
5
2
1
3000*1
0.9
0.4
1
600
1
200
1.1
0.2
1
2
5
0.3
0.4
9
3
800
100
0.5
2
200
60
10
0.5
1
10
LIMITING SCENARIO AND EXPOSURE PATHWAY
SCENARIO
Disposal
Recycle/
reuse
Disposal
Recycle/
reuse
Disposal
Recycle/
reuse
Disposal
Recycle/
reuse
Disposal
EXPOSURE PATHWAY
Ingestion of crops cultivated in the landfill site
Ingestion of freshwater products cultivated
with well water
External exposure on waste disposal
External exposure from reused equipment
Ingestion of crops contaminated with plume
Ingestion of crops cultivated in the landfill site
External exposure on waste disposal
External exposure of the resident in the landfill
site
External exposure on waste disposal
External exposure of transportation of waste
External exposure of the resident in the landfill
site
External exposure on waste disposal
External exposure of transportation of waste
External exposure from reused equipment
External exposure on waste disposal
Ingestion of crops contaminated with plume
External exposure of transportation of waste
External exposure on the asphalt parking lot
built with slag
External exposure of transportation of waste
TECDOC-855 (BQ/G)
RANGES
1000-
10000
100-1000
SINGLE
3000
300
N.A.*2
0.1-1
100-1000
1-10
0.1-1
0.1-1
100 - 1000
1-10
0.3
300
3
0.3
0.3
300
3
N.A.*2
N.A.*2
0.1-1
0.3
N.A.*2
N.A.*2
1-10
3
N.A.'2
0.1-10
0.3
N.A."2
N.A'2
N.A/2
N.A.*2
0.1-10
0.3
N.A.*2
N.A.*2
N.A.'2
N.A.*2
0.1-1
0.1-1
0.3
0.3
N.A.*2
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                              vvEPA
                                                                                ~f

-------
RADIONUCLIDE

Ce-144
Pm-148m
Eu-154*3
Eu-155
Gd-153
Tb-160
Hf-181
Ta-182
Pu-238
Pu-239*3
Pu-240
Pu-241
Am-241*3
Am-242m
Am-243
Cm-242
Cm-243
Cm-244
CLEARANCE
LEVELS
(BQ/G)
20
05
0.4
10
10
0.9
1
0.7
0.2
0.2
0.2
10
0.2
0.2
0.2
5
0.3
0.4
LIMITING SCENARIO AND EXPOSURE PATHWAY
SCENARIO

Recycle/
reuse
Disposal
Recycle/
reuse
EXPOSURE PATHWAY

External exposure on waste disposal
External exposure on the asphalt parking lot
built with slag
External exposure of transportation of waste
External exposure on waste disposal
External exposure of transportation of waste
External exposure on waste disposal
Inhalation of dust on unloading scrap metals
TECDOC-855 (Bo/G)
RANGES
10-100
SINGLE
30
N.A.*2
N.A.*2
N.A.*2
N.A*2
N.A.*2
N.A.*2
N.A.*2
N.A.'2
0.1 - 1
0.1-1
10-100
0.1-1
0.3
0.3
30
0.3
N.A.'2
N.A.*2
N.A.*2
N.A.*2
0.1-1
0.3
           */: The unit of the clearance level for Fe-55 and Sn-119m is Bq/cm2 because the limiting pathway is reuse of the surface
           contaminated equipment.
           *2. Clearance levels for these radionuclides are not available in IAEA-TECDOC-855
           *3: Radionuc/ides drawn to the underline are also derived at the nuclear reactors' derivation.
oEPA
           1B4
                                                    RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
  DISCUSSION

     In Table 4 and  6, the results of the clearance levels for both nuclear reactors and PIE
     facilities are  compared with those described in IAEA- TECDOC-855.  Most of derived
     clearance levels (e.g., y-ray emitters such as Co-60 and a-ray emitters such as Pu-239) were
     nearly the same  as those in the IAEA - TECDOC-855. Some (e.g., p-ray emitters such as
     Tc-99 and 1-129), however, were different. The major differences between this analysis and
     IAEA-TECDOC-855 are as follows:
         >   Derived  value of Fe-55 was higher than that of IAEA by one order of magnitude.
         >   Derived  value of H-3, C-14, Co-58, Fe-59 and 1-129 were lower than that of IAEA
             by one order of magnitude.
         >   Derived  value of Cl-36 and Tc-99 were lower than that of IAEA by two orders of
             magnitude.

     It is difficult to make these differences clear because IAEA's levels were  derived based on
     review of some literatures. Major reasons for these differences might be as follows:
         >•   The mixture with cleared scrap metals and non-radioactive metals  was not
             considered in the literature refereed for Fe-55  in IAEA.
         >   The common limiting pathway for H-3 and Tc-99, which is the ingestion of crops
             cultivated in the landfill site, is finally omitted from the consideration in IAEA.
         >•   In this analysis the limiting exposure pathways for C-14, Cl-36 and 1-129 are the
             related ones to the radionuclides migration via groundwater, but these pathways are
             not considered or finally omitted in the consideration in IAEA.
         >•   For Fe-59 and Co-58, parameter values such as the mixture ratio of cleared waste
             to non-radioactive one and working hours of operator were  different between in
             this analysis and IAEA.
         >•   Calculation model and parameter values were different between these derivations.

     The clearance levels  of PIE facilities will be authorized  after calculation results are
     validated with the stochastic approach. The clearance levels for the solid materials arising
     from other facilities such as radioisotope utilization facilities and accelerators will be
     derived after  establishment of a source term, exposure pathways, value of parameters and
     so on.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINQS                                185

-------
 REFERENCES
 [1]   International Atomic Energy  Agency, "Clearance Levels for Radionuclides in Solid
      Materials; Application of Exemption Principles", Interim Report for Comment-, IAEA-
      TECDOC-855, IAEA, Vienna (1996)
 [2]   Nuclear Safety Commission, "Clearance Levels for Solid Materials Arising from Main
      Reactors", NSC, Tokyo (1999) (in Japanese)
 [3]   Nuclear Safety Commission, "Clearance Levels for Solid Materials Arising from Heavy
      Water Reactors and Fast Breeder Reactors", NSC, Tokyo (1999) (in Japanese)
 [4]   Radiation Council, Exemption Dose  Criteria for the Near Surface Disposal of Solid
      Radioactive Wastes, Tokyo (1987) (in Japanese)

 [5]   International Commission On Radiological Protection,  Radiation Protection Principles
      for the Disposal of Solid Radioactive Waste, Publication 46, Pergamon Press, Oxford
      (1985)
 [6]   International Atomic Energy Agency, Principles for the Exemption of Radiation Sources
      and Practices from Regulatory  Control, Safety Series No.89, IAEA, Vienna (1988)
 [7]   M. Okoshi, et al., "Deterministic Approach towards Establishing of Clearance Levels in
      Japan", Proc. Int. Conference on Radioactive Waste Management and Environmental
      Remediation, Nagoya, Japan (1999)
 [8]   T.  Takahashi, et al.,  "Stochastic Approach to Confirm the Derivation of Clearance
      Levels", Proc. Int. Conference on Radioactive Waste Management and Environmental
      Remediation, Nagoya, Japan (1999)

 [9]   K.  Kawai, H. Tachibana, T. Hattori and S.  Suga, Table of Committed Effective Dose
      Equivalent etc. per Unit Intake based on ICRP Publication 30, JAERI-M 87-172, JAERI,
      Tokyo (1987) (in Japanese)
 [10] K.  Kawai, O. Togawa,  Y. Yamaguchi, S. Suga and T. Numakunai, Table of Committed
      Effective Dose  Equivalent  etc.  per Unit Intake of Actinide Elements Conformable to
      Radiation Protection Regulations (Supplement to JAERI-M 87-172), JAERI-M 90-022,
      JAERI, Tokyo (1990) (in Japanese)
 [11] International Commission On Radiological  Protection,  Limits   for  Intakes  of
      Radionuclides by  Workers, ICRP  Publication 30 Part 1  (and  subsequent parts  and
      supplements), Vol. 2 No. 3-4 through Vol. 8 No. 4, Pergamon Press, Oxford (1977-1982)
 [12] International Commission On Radiological Protection, The Metabolism of Plutonium and
      Related Elements, ICRP Publication 48, Pergamon Press, Oxford (1986)
186                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
            DEVELOPING A TECHNICAL BASIS FOR RELEASE OF
                                SOLID MATERIALS

  ROBERT MECK
      U.S. Nuclear Regulatory Commission

      Seven years ago, I met with this group in Tokai-mura, Japan, to compare the U.S. and
      Japanese plans for clearance of materials and equipment. This was my first trip to Japan.
      The experience was very enjoyable and gave many happy memories.

      After seven years, let's look at the current status of NRC activity on the control of solid
      materials. For this presentation, first I'll set the stage from the overall regulatory viewpoint.
      Then, I'll show activities at the NRC. We will briefly talk about how licensees can handle
      materials and equipment and how these processes could be updated.

      The main part of this presentation will be on the activities that could support a change and
      the next steps. Finally, you will see a schedule and conclusions.

      First, there are some terms that we need to understand in the same way. "Clearance" is a
      process removing radiological controls  that implies that controls are already in place. If
      one is removing  the radiological controls from, say, metal or trash, then that person is said
      to clear the materials.

      To clear materials or equipment, measurements of the concentrations of radioactivity are
      often required.  Most of these concentrations have been  determined to correspond to the
      highest reasonable dose of radiation to an individual or a group. These concentrations are
      called "clearance levels."   This is a generalized  representation of the  regulatory control
      system.

      The box on the left represents all radiation sources.  Some radiation sources such as
      potassium-40 in the body are unamenable to control. For that reason, they never enter into
      regulatory control. That's exclusion.

      Other sources have very small quantities and small concentrations of radioactivity and are
      intrinsically safe.  An example would be smoke  detectors  used in the home.  These are
      exempted from regulatory control by the regulatory authorities. This is exemption.

      Some practices that are under regulatory control release small amounts of radioactivity into
      the environment as a gas or a liquid under normal operations. The benefits of the normal
      operations  outweigh the detriment of the  environmental releases. This is referred to as
      "authorized discharge."

      Higher  quantities  and concentrations  may require  disposal  in  a licensed  repository.
      Internationally, this is called "authorized disposal."  Materials and equipment may leave
      regulatory control directly  through clearance. Authorized release is a middle step between
      regulatory  control and clearance. Conceptually,  there can be a  lessening of regulatory
      control in a transition. This would involve materials and equipment that are controlled so
      that they enter an intermediate step or process before they are cleared.
                                                                                        oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                               1 87

-------
     Usually, it is a process that removes or decays radioactivity or limits exposure of people by
     its use. For example, using metal as bridge girders.

     So today, how do NRC licensees release materials and  equipment?  It depends.  If the
     licensed radioactivity is part of the material itself, such as in neutron-activated metal, then
     the metal must go to an authorized disposal.  That disposal could be in a low-level waste
     repository or on a case-by-case basis into another disposal, on site, a municipal landfill, or
     an industrial landfill.

     If the radioactivity  is on the surface,  then the materials  licensee can release material or
     equipment  at  "Fuel Cycle  83-23," levels,  which  are equivalent to the more familiar
     "Regulatory Guide 1.86," levels.  Generally, these levels for beta gamma emitters is 5,000
     dpm per 100  centimeters squared and for alpha emitters and certain other nuclides, 100,
     1,000 or 5,000 dpm per 100 square centimeters  depending on the list in a group.  These
     averages are taken over a one-square meter area or less.

     There are also criteria for the maximum concentration and the removable concentration,
     which are three times and one-fifth of the average respectively. Reactor licensees may not
     release any detectable radioactivity associated with materials or equipment. In general, they
     must be able to detect radioactivity at the environmental levels.

     Where such measurements are impractical because of the size, shape or characteristic of the
     material  or equipment  and if the radioactivity  is on the surface, then they  may use
     deduction methods capable of detecting Regulatory Guide 1.86 levels. While materials and
     equipment are being  released  daily  from licensees,  improvements  could be  made.
     Improvements that have been presented as possibilities are consistency, for example, there
     are no regulations in place that apply to clearance.

     You just heard the various different kinds of way that  materials can be released from NRC
     licensees today.  We need consistency.  Current policies and practices treat reactor and
     material's licensees differently.

     They could be treated consistently under a regulation.  The levels used now are not dose or
     risk based.  While there is adequate  safety provided by the  levels, it is uneven  among
     different radionuclides.  Generic clearance levels would provide  relief for both regulators
     and licensees for making case-by-case determinations.   Specification of both  land and
     surface clearance levels would fill a regulatory vacuum.  Given this situation, what is the
     NRC doing? The next slide lists these actions.

     Basically, the Commission needs more  information and  they're going to get it in  several
     ways.   The National Academies' study is to investigate  alternatives for the release of
     contaminated materials. They are to consider the issue of recycling as a separate issue from
     the release of slightly contaminated solid materials.

     This direction may be interpreted  as  relating  to  the  general concepts of clearance,
     authorized release, and authorized disposal that we saw in an earlier slide.  We understand
     that the National Academies' committee has prepared a draft of their report but we are not
     allowed to know its contents until immediately before the  release to the public.
1 88                                RADIATION  RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     Next, we're developing a technical basis.  And this is what the program said that I would be
     talking about and I'm certainly willing to answer questions, either in the format that would
     keep us on schedule or at lunchtime.

     The dose assessments for individuals will be finalized in NUREG-1640. A supplement will
     cover  materials other  than  ferrous  metals,  aluminum, copper and  concrete. The  final
     version will respond to technical comments.

     In addition, analysis of the inventory of materials and equipment that are likely to  be
     cleared, collected doses and some costs are ongoing. NRC is also developing methods to
     measure low levels of radioactivity on and in materials and equipment.

     On the next point, international initiatives, NRC continues to support the IAEA efforts to
     establish clearance levels.  In addition, we are keeping informed on the implementation of
     clearance levels in  the EU and other countries.  Staff recommendations on how best to
     proceed will be an analysis of the options from the National Academies' report as well as
     staff input as appropriate.

     Finally, there is work with EPA and DOE.  EPA is in the process of posting an updating
     dose assessments for individuals on their clean metals website. EPA is not actively making
     a regulation on clearance.  DOE is in the process of developing an Environmental Impact
     Statement for the release  of metals with associated radioactivity  on surfaces.  .NRC is
     actively coordinating assessments of DOE processes and inventory with DOE

     So what is the schedule? The National Academies' report is due in February 2002. As for
     the technical bases,  individual dose assessments and the finalization of NUREG-1640, they
     are expected in the summer of 2002. The supplement, which will address individual doses
     from other materials, will be published later. Inventory, collective dose and some costs are
     expected in the fall of 2002. Coordination  with IAEA, EPA, and the DOE are ongoing.
     We will provide input as appropriate.  Three months after the National Academies' report
     is published, the staff recommendations are due. That makes it the summer of 2002.  It is
     difficult to predict when the Commission will respond to those recommendations.

     The conclusions that one  might draw from  the current status are — the Commission is
     actively supporting  the development of information to help them decide if rulemaking on
     clearance and the control of solid material is to proceed.  In view of the establishment of
     clearance criteria in other countries, it seems  likely that the U.S. will need to at  least
     address imports of cleared goods so some regulatory actions may be expected.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                1 89

-------
      A STATUS REPORT ON RECENT ACTIVITIES RELATED TO THE
                 WIPP AND YUCCA MOUNTAIN PROJECTS

 SCOTT MONROE
     U.S. Environmental Protection Agency, Radiation Protection Division

     This paper is about Waste Isolation Pilot Plant (WIPP) and Yucca Mountain. First, a quick
     overview of the current activities on the work we've done to date on WIPP.  I'll review the
     key elements of our radioactive waste disposal standards for the Yucca Mountain facility,
     issued in June of this year.

     We'll begin with WIPP. Just to review quickly the roles and responsibilities of two of the
     major Federal agencies involved in this project,  the Department of Energy (DOE) is the
     owner and operator of the WIPP site.  They are responsible for the disposal operations and
     for complying with applicable Federal and State regulations.

     The Environmental Protection  Agency (EPA), by Congressional decree, was tasked with
     developing radioactive waste disposal regulations for the WIPP site for certifying whether
     or not the site could meet those regulations.  If the answer was yes, then to recertify every
     five years that the site continues to be in compliance. And through each five-year period,
     during the operational phase, we would maintain an  ongoing regulatory  oversight role.
     Operations at the WIPP are expected to last about 35 years.

     Some key  dates. . . In 1985, EPA issued general radioactive waste disposal standards for
     transuranic waste.  That's plutonium-contaminated trash, really, from the defense program
     in the United States, and also  spent nuclear fuel and high-level waste. Those  standards
     were revised in 1993.

     In 1996, we issued disposal regulations that took the general standards and applied them
     specifically to WIPP.  We received an application from the  DOE.   We  reviewed that
     application and decided in 1998 that the site would comply with our disposal regulations.
     That decision allowed the  WIPP to  receive waste. The first shipment was received at the
     WIPP in March 1999. That shipment came from Los Alamos National Laboratories in
     New Mexico. This has to  do with the need for the Department to demonstrate to the State
     of New Mexico that the first shipment did not contain mixed waste.

     In October 1999, the State of New Mexico issued a hazardous waste disposal permit for
     WIPP and that allowed DOE to dispose of mixed radioactive and hazardous waste in this
     disposal facility.  So  the period we're in right now is preparing  for the  first  WIPP
     recertification, which should occur in 2004, five years after waste was first sent to WIPP.

     So to move into our current activity, we are preparing for recertification. I'm speaking as a
     regulator here. These are regulatory functions that we're engaged in, so we issue guidance
     to the department about what they need to do with their recertification application.

     We issued guidance in December of last year.  We may have to issue additional  guidance.
     Mainly what we're doing is engaging in very frequent communication with the Department
     staff about what needs to be done in order for recertification to be successful. And we have
     to communicate with the concerned public, with the State of New Mexico and elsewhere.
i go                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     As I  mentioned,  we also have an ongoing regulatory oversight role in addition to the
     recertification decision every five years.  That takes the form of two main areas of activity.
     First, we complete  inspections at WIPP sites  and at sites operated by DOE around the
     country where transuranic waste is stored and characterized prior to disposal.

     Also, we track changes that are initiated by the Department of Energy that could affect our
     certification.  Their application contains an enormous number of  commitments by the
     Department.  We make sure that they hold  true to those commitments.  And if we
     determine that  a  change proposed by DOE is a significant enough departure  from the
     application, we will modify our certification. That involves proposing to modify and then
     accepting public  comment on that proposal and then making a final decision.  And, of
     course, with these activities, we also have to communicate with the public.

     WIPP has now been operating for a little over two years.  They've disposed of roughly
     11,000 55-gallon  drums, with about 4000 shipments to date from these sites. These are all
     considered major sites with a larger amount of transuranic waste.  The total volume is
     roughly 2,600 cubic meters.  They're just getting started. And EPA has completed over 40
     inspections to  date.  That's  in about a three-year period at both the WIPP site  and
     transuranic waste  sites.

     I'd like to say a little about why and where we do inspections. Inspections  are a powerful
     tool for the regulator  to verify compliance.   When we  go to the  WIPP site  to do an
     inspection, we look at several things.  We look at their quality assurance program. There is
     a stringent standard for quality assurance that we apply to the WIPP program.  We look at
     their  environmental  and disposal  system monitoring.   This would be monitoring for
     releases on the  surface and also geological processes in the mine that could be related to
     performance. Also, we look at waste emplacement.

     With regard to transuranic waste sites, depending how you define them, there are eight to
     ten major sites around the country and 13 to 18 smaller-quantity sites.  These sites presently
     characterize waste prior to disposal, specifically to  identify limited waste  components.
     There are  components such  as metals, organic  materials and radionuclides  that were
     determined by  DOE to be  important  to  the  WIPP's ability to contain radionuclides
     effectively. We also look at the quality assurance programs for those processes. So there is
     a quality aspect and also the technical evaluation that we do.

     Lastly on WIPP, one point I'd like to  make is that we have very broad regulatory authority
     over  this program.   We are  authorized to  suspend the certification temporarily.  We're
     authorized to modify the certification to accommodate changes in activities or conditions.
     As I said, that would involve public comment.  And we can also revoke the certification in
     the event of a failure to comply with our regulations. That  would be basically a release into
     the accessible environment that exceeded our standards.

     We have not modified our certification to date.  We expect to have to.  That could happen
     within the next year or so.  And there was an event this summer where we suspended all
     shipments to the  site until we could complete an investigation of a noncompliance at the
     site in Idaho. But that's the closest we've come to a suspension.

     I'm going to move now to Yucca Mountain and explain the basic elements of the disposal
     standards that we issued this summer. You may already know this,  but I'll run through it
     quickly.
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                191

-------
              The DOE again is the owner and operator of this disposal site. It's a disposal site for spent
              nuclear fuel and high-level waste.  The EPA developed radiation protection standards for
              the  site  specific  to that site.   And the Nuclear  Regulatory Commission (NRC) is the
              licensing authority  for  this site.   They would develop the licensing criteria for Yucca
              Mountain based on EPA's standards and would review the license application from DOE
              and issue the license and oversee compliance from that point.

              Some key dates for the Yucca Mountain project, from EPA's point of view. . .  Earlier, I
              noted that we had issued general  standards for radioactive  waste disposal that included
              high-level waste and spent nuclear fuel.  In 1992,  Congress directed EPA first to
              commission a study by the Academy of Sciences about Yucca Mountain.  That report was
              issued in 1995.  Then, on the basis of that report, issue disposal standards that would be
              specific to Yucca Mountain. In other words, these would be separate and distinct from our
              general or generic disposal standards. So, as I said, we issued those standards.

              The standards can be divided into two components:  one requirement that applies to the
              management and storage of waste during the operational period; and a set of standards that
              apply to the disposal period. The management and storage standard is that, while the site is
              operating, no member of the public will receive more  than  150 microsieverts committed
              effective dose equivalent.

              For disposal of waste, the period of performance that we applied was 10,000 years.  There
              are  three  basic  requirements,    hi each  case, DOE  must demonstrate  a reasonable
              expectation of compliance with the limits, since absolute proof of compliance is not to be
              had.

              All  of the limits are based on  the mean of the  projected  doses and they apply to the
              reasonably maximally exposed  individual, or RMEI, as opposed to  a population.  So the
              first of  the  disposal standards  applies to undisturbed conditions, meaning no  human
              intrusion is factored into the analysis. The standard there is that the RME1 will receive no
              more than 150 microsieverts annually from radionuclide releases in undisturbed conditions.

              Some assumptions built into this are that exposure to the RMEI is from all pathways, that
              the RMEI lives above the point where the contaminated underwater plume has the highest
              concentration of radioactive contamination, and that the diet  and lifestyle of the RMEI are
              consistent with the town of Amargosa Valley, which is about 18 miles south of the Yucca
              Mountain  facility,  has about   1,400 residents,  and  is predominantly  an agricultural
              community.

              We then introduced  a  stylized  human  intrusion  scenario that  the  DOE would have to
              consider. And that standard is that DOE must determine the earliest time after the site is
              closed that a driller could puncture a waste package without detecting the presence of the
              package.

              If that time is less than 10,000 years, then the protection standard that applies is that the
              RMEI will receive no more than 150 microsieverts per year.  If the time of this penetration
              is later than 10,000 years, DOE has to report that  time in their Environmental  Impact
              Statement.   That's  a report that DOE would have  to  prepare prior to  initiating waste
              disposal operations.  And  the public is involved in the review of  that document.  This
              human intrusion scenario assumes that there is a single borehole above the facility that
              punctures a single waste package.
vxEPA
    >      1 92                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
   t-

-------
      We introduced our groundwater protection standard that is  consistent with the agency's
      standards for groundwater protection nationally developed by the EPA Office of Water.
      And this standard applies to undisturbed conditions where natural processes are considered
      and no human intrusion is assumed.

      This standard is that radionuclide concentrations in a representative volume of groundwater
      must be less than 5 picocuries (pCi) per liter combined radium 226 and 228,  15 pCi per liter
      gross alpha and 40 microsieverts (  Sv) per year to the whole  body or organ from beta
      photon radiation.

      hi this section, we review some other elements of the standards.  Peak dose  to the RMEI is
      anticipated well past 10,000 years, at about 100,000 to 200,000 years.  DOE must calculate
      the peak dose to the RMEI and must identify that in the Environmental Impact Statement.

      The point of compliance for the dose calculations can extend up to 18 kilometers south of
      the Yucca Mountain facility, which is consistent with the direction of groundwater flow
      and no more than five kilometers in any other direction, north, west, or east of the facility.
      DOE must preserve knowledge of the site.  However, we did not specify any assurance
      measures in our disposal regulations.  We left that to the licensing authority, NRC, to
      determine.

      And lastly, factors other than hydrologic, geologic, and climate  changes --  that is, natural
      processes ~ are assumed to remain constant. The future states assumption says that drilling
      technologies and social and political structures are assumed to be the same in the future as
      they are today because they cannot be predicted with any degree of reliability.

      So what happens next for Yucca Mountain?  EPA has issued its standards. Our rule has
      been completed. The next step  is that NRC will finalize its licensing criteria now that they
      have our final standards.  They can do that by the end of this year.

      The DOE is expected to recommend that Yucca Mountain be used to dispose of radioactive
      waste. That recommendation will go to the President and could  happen by the end of this
      year or early next year.  If the President  and Congress agree with DOE that  the site is
      suitable, they will authorize disposal and the Department will  submit a license application
      to NRC.

      However, we have been sued on our disposal standards and the Federal Courts will have to
      decide the outcome  of those  lawsuits. We don't know when they will do that.  It could be
      spring of next year.  Obviously, the outcome of those lawsuits could affect some of these
      other actions and their timing.

      So to conclude, I'll just say that the WIPP program  is a very complex project.  Therefore
      we, as the regulators, are very active and there are a number of unprecedented, technical
      and policy issues that we have to confront.  It's really a fascinating project to work on. For
      Yucca Mountain, our standards  are  out. We think that they are, without question, a crucial
      part of evaluating the suitability of the site for waste disposal and we believe that these
      regulations are appropriate and will be protective of human health and the environment.
                                                                                         &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                1 93

-------
                   SAFETY ANALYSES FOR 5HALLOW-LAND DISPOSAL OF
                                    ALPHA-BEARING WASTES

           HIDEO  KIMURA, SEMI TAKEDA, MITSUHIRD KAN NO, AND NAQFUMI MINASE
              Japan Atomic Energy Research Institute, Department of Fuel Cycle Safety Research

           ABSTRACT

              Safety analyses for the  shallow-land disposal of alpha-bearing wastes were performed
              using the deterministic and probabilistic safety  assessment models.   The deterministic
              analyses show that the dose calculation in the residence scenario is of great importance
              owing to the influence of daughters built up  by uranium decay chain.  The parameter
              uncertainties  for the important pathways in residence  scenario  are estimated from the
              probabilistic analyses using the statistical methodology.  The uncertainty analysis indicates
              that the influence of parameter uncertainty is the most remarkable  in the estimation for the
              inhalation of radon gas with residence.

           INTRODUCTION

              "Uranium Wastes" or RI and Research Wastes represent the  alpha-bearing wastes. The
              contaminated materials by uranium are generated through the operation and dismantling of
              facilities for  smelting, converting,  enriching etc. It  is  considered that  there are some
              possibilities of safely implementing a shallow-land disposal for most of the uranium wastes
              because those concentration levels are distributed to be relatively low. The uranium wastes
              are characterized by the existence of long-lived radionuclides, the growth of daughters
              associated with uranium decay chain, the emanation of radon from wastes containing Ra-
              226 etc.  To evaluate the performance of the waste disposal system over  long time scales, a
              deterministic  approach is used for quantitative estimates of peak dose to an individual.
              However, uncertainties with respect to parameters, scenarios etc. are inherent in the long-
              term assessment for uranium waste disposal.  At the  next step of safety assessment, it is
              essential to estimate the uncertainties quantitatively. JAERI has developed the deterministic
              and probabilistic safety assessment system to estimate the long-term radiation effect owing
              to the shallow-land disposal of uranium wastes.

           ASSESSMENT METHODOLOGY

              In this study, uranium wastes are distinguished from solid wastes such as the residues from
              the mining of uranium alpha-bearing ores.  The  major  materials  are concrete, metal and
              incinerated wastes arising from the operation and  dismantling  of the  facilities.  The
              amounts of uranium wastes considered in this analysis are 100 thousands m3. It is assumed
              that  the  quantities of the materials are disposed in shallow-land burial  of trench.  The
              performance  of the shallow-land disposal  system may  be  affected  by  two events in the
              future: subsequent natural process and human intrusion.  The natural process leading to
              human exposure is  represented by radionuclide migration  in groundwater flow with the
              associated process of diffusion, dispersion, sorption and decay. There is  also the possibility
              of inadvertent human intrusion such as house construction on the shallow repository. The
              scenarios considered here are both the site-reuse scenarios associated with  human intrusion
              into the disposal site and the groundwater migration scenarios with radionuclides migration
              in groundwater.  The  events of human intrusion at the  disposal site are considered to be
              exposure  events with house construction (construction scenario)  and with residence

4>EPA
  ,*ftti-   194-                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     (residence  scenario).   The  exposure events of using river  water on the basis of
     contaminated  groundwater migration (groundwater migration  scenario on use of river
     water) were considered to derive the upper bounds of radioactive  concentration in the
     wastes acceptable for the LLW disposal in Japan [1] and are also referred here. Viewed
     from scenario  uncertainty on the estimate of long-term radiation effect, the events of using
     water  extracted from  a  well after the  migration of radionuclides  in  groundwater
     (groundwater  migration scenario on use of well  water) are additionally considered here.
     These  four scenarios  are respectively divided  into exposure pathways describing the
     activities of specific individuals as shown in TABLE I. Humans in the pathways may be
     exposed to external  radiation, inhalation of radioactive particles and ingestion of foodstuffs
     containing radionuclides. Some exposure  pathways due to the  inhalation of radon gas,
     which  are distinctive in the disposal system of uranium wastes, are also included in these
     scenarios.

                                       TABLE 1 :
    DESCRIPTIONS OF SCENARIOS AND EXPOSURE PATHWAYS IN THE SHALLOW-LAND
                            DISPOSAL OF URANIUM WASTES
Scenario
Construction scenario
Residence scenario
Groundwater migration
scenario on use of
river water
Groundwater migration
scenario on use of
well water
Exposure pathways
External exposure with house construction
Inhalation of contaminated particles with house construction
Inhalation of radon gas emanated from the site under construction
Ingestion of crops cultivated in the disposal site
External exposure with residence
Inhalation of radon gas emanated from the site with residence
Ingestion of river water
Ingestion of radon released through the use of river water for living
Ingestion of freshwater products caught in the river
Ingestion of livestock grown with river water
Ingestion of well water
Ingestion of crops cultivated with well water
Ingestion of livestock grown with the feeds cultivated with well water
External exposure with agriculture
Exposed individual
Construction worker
Construction worker
Construction worker
Resident
Resident
Resident
Consumer
Resident
Consumer
Consumer
Consumer
Consumer
Consumer
Farmer
  DETERMINIBTIC ABBEBBMENT METHODOLOGY

     The main  procedure to  evaluate the individual  doses  for  the  site-reuse scenario and
     groundwater migration scenario consists of the estimates for:
         >   Release rates of radionuclides from the disposal facilities and quantities of
             radionuclides remaining in the facilities.
         >   Migration of radionuclide with groundwater leached from the facilities and
             concentration of radionuclide in the river and well water.
         >   Concentration of radionuclide remaining in soil cultivated with the well water.
         >   Movement of radon in porous materials such as the waste and soil, and
             concentration of radon in the outdoor and indoor space.
         >   Individual doses for each exposure pathway.
                                                                                        &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS
                                                                                   195

-------
               Assuming the radionuclides are released on the basis of distribution equilibrium, the
               radionuclides fluxes from the facilities and the quantities of radionuclides remaining in the
               facilities are calculated by dynamic compartment model. This model is described by the
               following simultaneous ordinary differential equations:

               Equation (1):

                 H(~*
                  J/ = -(n, + A) • cw + A,_I • c   .,
                  dt                                                              (l)

                     Q
               where  w'! is the amounts of radionuclide '  in the disposal facilities,  /7( is release rate of
               radionuclide ', and decay constant of radionuclide '. The release rate ^' is given by the
               following Equation (2):
                 n, =•
               where  P  is infiltration rate,   w is thickness of waste layer,  £/ is porosity of waste layer,
               Pw is bulk density of waste layer and    "•'  is distribution coefficient of radionuclide '  in
               waste layer. The released radionuclides migrate through a saturated zone with groundwater
               and flow to the water body such as a river. The migration  of radionuclides is estimated
               from solving the 1-D advection dispersion equation. The concentration in the river and well
               water,  which is used for drinking,  agriculture, and etc., is  calculated talcing  account  of
               mixing with  contaminated and non-contaminated water volume.  The  concentration  of
               radionuclide remaining in cultivated soil owing to using the  well water is calculated from
               the application of the dynamic compartment model to the cultivated soil layer under the
               input condition of estimated concentration in the irrigation water.

               The calculations  of the radon impact use the following models to estimate the rates  of
               radon emanation  from the wastes and the soil mixed with the wastes containing Ra-226.
               Generally, the radon exhalation rate from a soil into open atmosphere depends on many
               environmental factors such as water content and particle  size  of the  soil, wind velocity etc.
               Assuming a homogenous radium distribution in the waste layer, the radon flux density is
               obtained by use of the following Equation (3):
                      = CRa(t)-pw-F- ARn •      • tanh x*-
                                         *              A                          ,~


               where   w  is radon flux density at the upper surface,   Ra^ '  is Ra-226 concentration in the

               waste layer at time  t ,  F is emanation factor,   Rn  is decay constant of 222Rn,   w is
                                                                             y-
               effective diffusion coefficient of Rn-222 in the waste layer,  and   "  is thickness of the
               waste layer.

               If non-contaminated soil is covered with the waste layer,  the radon flux density into open
               air is represented by the following Equation (4):
oEPA
          1 96                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
        Jc(t) =

      where,  c is radon flux density into open air,   c  is thickness of the covered soil,  DC is
      effective diffusion coefficient of Rn-222 in the covered soil. The concentration of Rn-222
      in outdoor air, which is emanated from a finite area such as the disposal area, is described
      by the following Equation (5):
             C* ( f\
      where   °v ' is radon concentration in outdoor air, H  is height of open space, u  is wind
      velocity, and L is length of emanation area toward the velocity direction. The solution of
      this equation with the initial condition  c°'°'    is given by the following Equation (6):
      The concentration of radon in indoor air is used for the estimate on the inhalation of radon
      gas for the resident. It depends on the design and construction of the building structure, on
      meteorological parameters, and on the living habits of the occupants, which can affect the
      air exchange rate in the building [3]. Most of the houses in Japan have a vented crawl
      space, in which the height should be more than 45 cm according to the Building Standards
      Act in Japan. Considering the house construction and living habits in Japan, the estimates
      of the Rn-222 concentration in a house is based on the following:
         >•  Radon concentration in the house is determined by two infiltrations. One is the
             infiltration with the ventilation from windows, and another is the distribution
             infiltrated through the crawl space.
         >  In addition, the concentration in the crawl space is calculated by the summation of
             both the emanated concentration from the underlying soil and the infiltrated
             concentration from a ventilating opening in the crawl space.

      These concentrations are estimated from the equations,  which are on a mass  balance of
      radon gas in indoor space or the crawl space, respectively such as Equation (5).  Radon gas
      may  be released from water to air due to using the contaminated water in a house.  The
      radon concentration is expressed by as follows [3], [4]:
      where  ' \' is radon concentration due to water degassing, °w ^ is radon concentration in
      the water used for living, ^ is the amount of water used per unit time, G is the degassing
      efficiency and V is the volume of the reference house. Finally, individual doses owing to
      external radiation,  inhalation  of radioactive particles, ingestion of foodstuffs containing
      radionuclides, and the inhalation of radon  gas are calculated based on the estimated
      concentration for the radionuclides in the disposal site, the cultivated soil, the river and well
      water, air etc.


                                                                                          oEPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                                i 97

-------
          PROBABILISTIC ABBEBBMENT METHODOLOBY

              In the deterministic analysis, the individual dose for the exposure pathway is calculated
              using a conservative or realistic value for each parameter. However, most of the parameters
              have their uncertainty and variability. To estimate the influence of parameter uncertainty in
              the long-term  assessment  for uranium  wastes, we have  developed  a probabilistic
              assessment methodology. In this study, the probabilistic analyses for the residence scenario,
              which result in a critical scenario from the  deterministic analyses, are performed for the
              quantitative estimate of the parameter uncertainty. Probabilistic assessment code system
              consists of a parameter sampling code, the assessment code for site-reuse scenario,  and a
              statistical  analysis code. The Latin  Hypercube Sampling  (LHS)  method  is  used for
              sampling of parameter  sets on the basis  of Monte Carlo technique [5]. The Probability
              Density Function  (PDF) with  the definition  of variable range and distribution type
              describes the variability of each parameter. The PDFs in the parameter sampling code are
              defined by four kinds of distribution type, uniform, log uniform, normal and lognormal. In
              the PDF of log uniform or lognormal distributions, the minimum and maximum values for
              the parameter are treated as the values of 0.1 percentile and 99.9 percentile  in the PDF
              respectively. At the first step of probabilistic analysis, the parameters, which may bring the
              uncertainty and variability, are picked up, and the PDF is defined for each parameter. The
              next step, the parameter sets  are  generated  by the  LHS method. The probabilistic
              calculations of the individual doses for the sampled data sets are carried out using the same
              models as the deterministic analysis.  The statistical analysis  code is applied  to the peak
              dose associated with the sampled parameter sets. The consideration on the statistical results
              such as scatter plot, Cumulative Distribution Function (CDF), Complementary Cumulative
              Distribution  Function (CCDF) of peak  dose values etc. leads to evaluation of parameter
              uncertainty  and variability, hi  addition, parameter sensitivity  or  importance can  be
              estimated from the consideration of partial rank correlation coefficients (PRCCs) for each
              sampled parameter against peak dose values.

           SAFETY ANALYSES

              ASSUMPTIONS AND PARAMETERS

              In this analysis, initial inventory  in uranium wastes is  determined from the  level  of the
              representative enrichment 4.5% used in Japan. On the basis of specific activity values for
              uranium  at various levels of  enrichment  in the IAEA's report [6], the ratios of specific
              activity for U-238, U-235 and U-234 are estimated to be 0.13, 0.04 and 0.83, respectively.
              The  analyses  are carried out  for  the radionuclides with half-life of more than  10 days,
              including in 4N+2 and 4N+3 chains. For the daughters with half-life of less than 10 days,
              their dose conversion  factors are added to those of their  parents,  assuming to  be  in
              radioactive equilibrium with their parents. Internal dose conversion factors for ingestion
              and inhalation are cited from ICRP publication 68 [7], and external dose conversion factors
              of radionuclides are calculated using QAD-CGGP2  [8]. Parameters  associated  with the
              disposal  facilities, groundwater, geosphere, exposure pathways are chosen based  on the
              parameters used hi the NSC's report [1]. The values of distribution coefficient and transfer
              factors to foodstuffs are basically cited from IAEA  TRS 364 [9]. Parameters on radon
              migration refer to  the values in the UNSCEAR's reports [2], [3], and parameters such  as
              house scale,  air exchange rate, infiltration rate from the floor, etc. are determined  by the
              data  on  Japanese house construction. The safety analyses for site-reuse  scenario are
              performed under two assumptions: one is the conservative assumption taking account of no


oEPA
         1 98                               RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     release of radionuclides from the facilities over long time scales and another is the realistic
     assumption taking account of release rates of radionuclides from the facilities.

     UNCERTAINTY or INPUT DATA

     On the basis of the results of the deterministic analyses, the probabilistic analyses are
     carried out for the important scenario, residence scenario, including the following exposure
     pathways:
         >•   Ingestion of crops cultivated in the disposal site,
         >   External exposure with residence, and
         >   Inhalation of radon gas emanated from the site with residence.

     Major PDFs for variable  parameters, which are used  in  calculations  of those three
     pathways, are shown in Table 2. The PDFs for the parameters are defined from the  review
     of existing reports such as IAEA TRS 364 [9] and the consideration of natural and social
     conditions in Japan. Some parameters, amount of uranium wastes, plugging ratio of waste
     material, thickness  of covered  soil and height of crawl  space,  are treated as the fixed
     parameters  in  the  probabilistic analyses.  In the estimate  of  external  exposure  with
     residence, it is  considered that external dose conversion factor  for each radionuclide may
     depend on the uncertainty for thickness of borrowed soil, which is a non-contaminated soil
     for house construction. This probabilistic analysis system provides data library with respect
     to the external dose conversion factor corresponding to the variability of the thickness. The
     conversion factor associated with the thickness of borrowed soil is calculated using the
     interpolation [10].

     REBULTB or  DETERMINISTIC ANAL.YBIB

     The results of safety analyses for specific  activity  1.0 Bq/g of total uranium  with
     enrichment 4.5% are shown in  Figure 1 (a) and (b). Figure 1 (a) indicates the calculated
     individual dose for  four scenarios (construction scenario, residence scenario, groundwater
     migration scenario on use of river water and well water). The  dose history in each scenario
     is summed up  from the results  for the exposure  pathways including in each scenario, as
     shown in TABLE I. These analyses are extended to times beyond the highest value of the
     dose (maximum dose), hi the case of no-released radionuclide from the disposal facilities,
     the dose in residence scenario is the highest and its maximum dose of around 2.2E-4  Sv/y
     is reached about two hundred thousand years  after the disposal.  The component of total
     dose in residence scenario is shown in Figure 1 (b). The exposure for inhalation of radon
     gas with residence is the most critical in this scenario because of increasing radon
     concentration in the site depending on the accumulated concentration of Ra-226. The doses
     derived from Pb-210  and  Ra-226 are dominant,  respectively in  Ingestion pathway  of
     cultivated crops and  in  external exposure  pathway.  This result indicates that  the  dose
     evaluation in residence scenario is of great  importance in the safety assessment  owing to
     the influence of daughters built up by uranium decay chain. Considering the release  rates
     from the disposal facilities as realistic assumption, the maximum dose in residence scenario
     decreases to about  8.0E-6  Sv/y. The results for residence  scenario are  sensitive  to the
     release condition of radionuclides from the facilities over long-term period.

     Viewed from scenario uncertainty on the estimate  of long-term radiation effect, the  dose
     calculations in groundwater migration scenario were performed  for use of both river water
     and well water in a biosphere.  The calculated dose for use  of well water is about three


                                                                                         &EPA
RADIATION RISK ASSESSMENT  WORKSHOP PROCEEDINGS                                199

-------
&EPA
               orders of magnitude greater than that for use of river water. This is based on the difference
               of flow rates between well and river. The dose history in the case of well water indicates
               two peaks at four hundred years and at thirty thousand years. The former is the peak dose
               derived from U-234, the latter is the one derived from its daughters. The maximum dose for
               use  of well water is about 5.0E-5 Sv/y. Under the realistic assumption taking account of
               release rates of radionuclides from the facilities, the maximum dose for use of well water is
               higher than that in residence scenario. Therefore, the scenario uncertainty on use of the
               contaminated water in a biosphere over long time scales has a great influence on the dose
               calculation.

                                                 TABLE 2:
                    MAJOR PDFS FOR VARIABLE PARAMETERS IN RESIDENCE SCENARIO
Parameter
Unit
Deterministic
value
Distribution
type
Minimum
value
Maximum
value
Parameters on the disposal facilities
Amount of uranium wastes
Width of waste layer
Length of waste layer
Thickness of waste layer
Plugging ratio of waste material
Bulk density of disposal site
Porosity of disposal site
Infiltration rate into waste layer
Start time of radionuclide release
after closure
Thickness of covered soil
Depth of excavation
Thickness of borrowed soil
m3
m
m
m
-
g/cm3
-
m/y
y
m
m
m
100,000
350
350
5
0.163
2
0.2
0.4
0
1.8
3
0.3
Constant
Calculation'1
Uniform
Uniform
Constant
Uniform
Normal
Lognormal
Uniform
Constant
Uniform
Uniform
-
-
70
1
-
1
0.15
0.1
0
-
0.5
0
-
-
700
10
-
2.3
0.3
1
300
-
10
1
Parameters on external exposure pathway
Annual exposure time
Shielding factor in residence
h/y
-
8760
0.2
Normal
Uniform
3000
0
8760
0.66
Parameters on ingestion of crops cultivated in the disposal site
Absorption factor from plant root*2
Ingestion rate of rice
Ingestion rate of green vegetable
Ingestion rate of root crop
Ingestion rate of fruit
Dilution factor of crops in a market
-
kg/y
kg/y
kg/y
kg/y
-
1
71
12
45
22
1
Loguniform
Normal
Normal
Normal
Normal
Uniform
0.002
0
0
0
0
0
1
149
36
139
81
1
Parameters on inhalation of radon gas emanated from the disposal site
Emanating power
Radon diffusion coefficient in waste layer
Radon diffusion coefficient in covered soil
Radon diffusion coefficient in borrowed soil
Height of outdoor space
Length of emanation area
Wind velocity
Air exchange rate in crawl space
Height of crawl space
Air exchange rate in indoor space
Height of indoor space
Radon infiltration rate through the floor
Equilibrium factor in indoor space
Equilibrium factor in outdoor space
Ratio of outdoor living
Annual exposure time
-
m2/s
m2/s
m2/s
m
m
m/s
m
s-1
s-1
m
s-1
-
-
-
h/y
0.2
2.0E-06
2.0E-06
2.0E-06
3
180
3
0.45
9.9E-04
1.1E-04
2.5
1.0E-04
04
0.8
0.2
8760
Lognormal
Lognormal
Lognormal
Lognormal
Uniform
Uniform
Normal
Constant
Lognormal
Lognormal
Loguniform
Lognormal
Normal
Normal
Uniform
Normal
0.01
1.0E-10
1 OE-10
1 OE-10
1
70
1 4
-
5.6E-05
1.4E-05
2
1.4E--06
0.1
0.1
0
3000
0.8
3.0E-06
3.0E-06
3.0E-06
5
700
5.5
-
3.1E-03
1.4E-03
5
1.0E-04
07
1
0.66
8760
            *1) This \alue is calculated from the sampled size of waste layer and constant \ralues of waste wlume and plugging
            *2) This factor accounts for the ratio of root which reaches the waste layer and absorbs radionuclides.
          zoo
                                              RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
                                         FIGURE: 1 :
  RESULTS OF DETERMINISTIC ANALYSES: (A)  INDIVIDUAL DOSES FDR FOUR SCENARIOS,
 (B) INDIVIDUAL DOSES FOR THREE EXPOSURE PATHWAYS IN RESIDENCE SCENARIO UNDER
                 NO RELEASE CONDITION FROM THE DISPOSAL FACILITIES.
   1E-3
   1E-5
   1E-8
   1E-9
  1E-10
             Readencescersrio ResidencescerBno
             (Releaseccncitcn) 
-------
               The parameter importance  in  each exposure pathway can  be estimated from using the
               partial rank correlation coefficients (PRCCs) for each sampled parameter against peak dose
               values.  The PRCC value is used as an indicator for the screening of parameter to the
               calculated peak doses.  The absolute value of PRCC indicates the  extent of parameter
               importance,  and the correlation between the parameter and the dose is represented by plus
               and minus of PRCC value. Figure 3 shows the PRCC values of parameters used in the dose
               calculation for the inhalation of radon gas. The  important parameters  identified by high
               PRCC value are depth of excavation, thickness of borrowed soil, and diffusion coefficient
               of radon in the soils. The PRCC value for the distribution coefficient of uranium in waste
               layer is also high under the release condition from the facilities. The depth of excavation
               determines the mixing rate between the waste layer and the covered soil (non-contaminated
               soil), and this  leads  to its tendency of high PRCC value. The  PRCC values for  two
               parameters, thickness of the borrowed soil and radon diffusion coefficient in the borrowed
               soil, are especially high.  This  indicates that the  parameters with respect to the diffusion
               migration of radon gas in the surface soil are of great importance in the radon estimate.

                                                   FIGURE 3:
                    PRCC  VALUES OF PARAMETERS USED IN THE DOSE CALCULATION FDR
                                          INHALATION OF RADON BAS
                                         Length of emanation area
                                         Thickness of waste layer
                                      Bulk density of disposal site
                                          Porosity of disposal site
                                     Infiltration rate into waste layer
                        Start time of radionuclide release after closure
                                            Depth of excavation
                                        Thickness of borrowed soil
                             Distribution coefficient in waste layer (U)
                            Distribution coefficient in waste layer (Pa)
                            Distribution coefficient in waste layer (Th)
                            Distribution coefficient in waste layer (Ac)
                            Distribution coefficient in waste layer (Ra)
                            Distribution coefficient in waste layer (Po)
                            Distribution coefficient in waste layer (Pb)
                                               Emanating power
                            Radon diffusion coefficient in waste layer
                            Radon diffusion coefficient in covered soil
                           Radon diffusion coefficient in borrowed soil
                                          Height of outdoor space
                                                  Wind velocity
                                    Air exchange rate in crawl space
                                   Air exchange rate in indoor space
                                           Height of indoor space
                              Radon infiltration rate through the floor
                                   Equilibrium factor in indoor space
                                  Equilibrium factor in outdoor space
                                           Ratio of outdoor living
                                            Annual exposure time
oEPA
            CONCLUSION

                The JAERI has developed the deterministic and probabilistic safety assessment system to
                estimate the  long-term radiation effect owing to the shallow-land disposal  of  uranium
                wastes.  The safety and uncertainty analyses for the waste disposal were performed using
                the developed  deterministic and probabilistic  safety  assessment system. The results  are
                summarized as follows:
          202
                                                RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS

-------
     The  safety analysis shows  that the  dose evaluation in residence scenario  is of great
     importance in the  safety assessment owing to  the influence of daughters built up by
     uranium decay chain. The dose in residence scenario is sensitive to the release condition of
     radionuclides from the facilities over a long-term period.

     The  uncertainty  analysis based  on  the probabilistic methodology  indicates that  the
     influence  of parameter uncertainty is  the  most remarkable in the  estimation for  the
     inhalation of radon  gas with  residence. The important parameters identified by high PRCC
     value are  depth of excavation, thickness of borrowed soil,  and diffusion coefficient of
     radon in the soils.

  REFERENCES
  (1) NSC: Nuclear Safety Commission, "Radioactive Concentration Upper bounds for the Safety
     Regulations Governing the Shallow Land Disposal of Low-Level Solid Radioactive Wastes
     (Second Interim Report)", NSC, Tokyo, (in Japanese), 1992.
  (2) UNSCEAR:  United Nations Scientific  Committee on the Effects of Atomic Radiation,
     "Sources, Effects and Risks of Ionizing Radiation", United Nations, New York, 1993.
  (3) UNSCEAR:  United Nations Scientific  Committee on the Effects of Atomic Radiation,
     "Sources, Effects and Risks of Ionizing Radiation", United Nations, New York, 1988.
  (4) E. P. Lawrence, R.  B. Wanty, and P. Nyberg, "Contribution of 222Rn in Domestic Water
     Supplies to 222Rn  Indoor Air in Colorado Homes", Health Physics, Vol.62,  2, 171-177,
     1992.
  (5) R. L. Iman, M. J. Shortencarier and J. D. Johnson, "A FORTRAN 77 Program and Use's
     Guide for the Calculation of Partial Correlation and Standardized Regression Coefficients",
     NUREG/CR-4122, U. S. Nuclear Regulatory Commission, 1985.
  (6) IAEA," Advisory Material for the IAEA Regulations for the Safe Transport of Radioactive
     Material (1985 Edition), Third Edition (As Amended 1990)", Safety Series No.37, IAEA,
     Vienna, 1990.
  (7) ICRP, "Dose Coefficients for Intakes of Radionuclides by Workers, Replacement of ICRP
     Publication 61", ICRP Publication 68, ICRP, 1994.
  (8) Y. Sakamoto and S. Tanaka, "QAD-CGGP2 and G33-GP2: Revised  Versions of QAD-
     CGGP2 and G33-GP (Codes with the Conversion Factors from Exposure to Ambient and
     Maximum Dose Equivalents)", JAERI-M 90-110, JAERI, Tokyo,  1990.
  (9) IAEA, "Handbook  of Parameter Values for the Prediction of Radionuclide  Transfer in
     Temperate Environments", Technical Report Series No.364, IAEA, Vienna, 1994.
  (10) A. Akima, "A New Method of Interpolation and  Smooth Curve Fitting  Based on Local
     Procedures", J. ACM, 17, 4,  1970.
                                                                                     &EPA
RADIATION RISK ASSESSMENT WORKSHOP PROCEEDINGS                              2O3    .4>>.L

-------