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
-------
|