Advisory on Agency Draft White Paper Entitled Modifying EPA Radiation Risk Models Based on BEIR VII

5. UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
? WASHINGTON D.C. 20460
OFFICE OF THE ADMINISTRATOR
SCIENCE ADVISORY BOARD

January 31,2008

EPA-SAB-08-006

The Honorable Stephen L. Johnson
Administrator
U.S. Environmental Protection Agency
1200 Pennsylvania Avenue, N.W.
Washington, DC 20460

Subject: Advisory on Agency Draft White Paper entitled "ModifyingEPA Radiation
Risk Models Based on BEIR VII"

Dear Administrator Johnson:

The Radiation Advisory Committee (RAC) of the Science Advisory Board has completed
its review of the Agency's draft White Paper entitled "Modifying EPA Radiation Risk Models
Based on BEIR VII, " dated August 1, 2006. In this draft White Paper, the Agency' s Office of
Radiation and Indoor Air (ORIA) outlined proposed changes in the EPA's methodology for
estimating cancers from exposure to low levels of ionizing radiation. The EPA sought the
RAC's advice on the application of the National Research Council of the National Academies'
Committee to Assess Health Risks from Exposure to Low Levels of Ionizing Radiation (BEIR
VII, 2006) cancer risk estimates and on issues relating to the proposed modifications and
expansions desirable or necessary for EPA's purposes.

The RAC endorses EPA's proposal to base its approach to low dose risk estimation on
BEIR VII. Specifically, for purposes of establishing radiation protection policy, the RAC
endorses the EPA's use of a Linear Non-Threshold (LNT) model combined with the Dose and
Dose Rate Effectiveness Factor (DDREF). That is, the slope of the dose-response relationship in
the high dose region is modified by the DDREF which corrects for the decreased biological
effectiveness of low dose and dose-rate exposures. The resulting lower slope is then linearly
extrapolated into the very low dose and dose-rate region in which epidemiological data usable in
analyses have not and may not be obtained. By low dose, the RAC follows BEIR VII's
definition; that is, doses below 100 mSv (0.1 Sv), in the context of low Linear Energy Transfer
(LET) radiation.

The RAC agrees with the EPA that the BEIR VII methodologies using incidence models
and data should be used wherever possible. The RAC accepts the EPA's use of BEIR VII
methodologies for deriving risk estimates for cancers of the stomach, colon, liver, prostate,
uterus, ovary, bladder, other solid cancers, and leukemia. The RAC did not find compelling
evidence to suggest the use of the alternative lung cancer model discussed by EPA and

-------
recommends that the EPA use the BEIR VII methodologies for deriving risk estimates for
radiogenic lung cancer risk.

There were several areas not addressed by BEIR VII, for which EPA requires a cancer
risk estimate. They include:

• in utero - The RAC concludes that it would be reasonable for the EPA to use the
referenced estimates of cancer risk from in utero exposure to external radiation sources,
and the dose coefficients provided by the International Commission on Radiological
Protection (ICRP) as a basis for developing its risk estimates for in utero radiation
exposure from internally-deposited radionuclides.
• bone - The EPA proposes to divide the bone cancer risk observed in humans exposed to
alpha particles from 224Ra by an Relative Biological Effectiveness (RBE) to estimate the
bone cancer risk from 90Sr. The RAC concurs with this practical, operational approach to
radiation protection.
• non-melanoma skin cancer (NMSC) - The RAC supports EPA's proposed use of the
1991 ICRP model to estimate the incidence and mortality risks of radiogenic NMSC.
The RAC concurs with EPA that because of the high baseline incidence rates and low
mortality due to NMSC, it is inappropriate to include risk estimates for radiogenic NMSC
in the estimate of the incidence and mortality risk for radiogenic cancer.
• higher LET radiation
o alpha particles - The RAC is supportive of the use of a generally accepted
Maximum Relative Biological Effectiveness (RBEm) value, such as 20 which is
currently being used. The RAC recommends using data specific to particular
radionuclides where such human cancer risk data are available (e.g., lung, liver,
bone, or bone marrow). For other organs and tissues, the RAC is supportive of
the general approach of using the low-LET cancer risk from BEIR VII multiplied
by RBEM.
o lower energy photons - The RAC concurs that an RBE in the range of 2 to 2.5
seems reasonable for low-energy photons and electrons for purposes of setting
radiation protection standards.
o beta particles - The RAC concurs that an RBE in the range of 2 to 2.5 is
reasonable for estimating the cancer risk from exposure to tritium.
• additional uncertainty -An additional source of uncertainty in risk estimates is
associated with the mechanistic biophysical model that is used in BEIR VII to support the
LNT model in the low dose region. In Appendix A, the RAC provides a brief review of
current research and recommends that the EPA remain aware of the research
continuously updating the biophysical model used to support the estimates of radiation
risk following low dose radiation exposure.

The RAC finds that the EPA is warranted in modifying the BEIR VII methodologies in
several specific areas where the EPA's particular application requires some adaptation of the
BEIR VII approach. The RAC agrees that the proposed estimation of radiogenic cancer risks for
the U.S. population using a standard stationary population, that is for a 'fixed cohort' based on
death rates for the year 2000, is a reasonable adaptation of the BEIR VII approach. The RAC

-------
agrees that the EPA's proposed use of the most current cancer-specific incidence and mortality
rates available is an appropriate and scientifically valid adaptation of the BEIR VII approach.

The RAC concurs with EPA's exploration of alternative methods for estimating the
relative risk for radiogenic breast cancer. In particular, the RAC concurs with the EPA's
proposal to relate current breast cancer mortality rates to retrospective incidence rates rather than
current incidence rates to better reflect the influence of life style changes, earlier breast cancer
detection and treatment that could influence survival and hence mortality rates over an extended
period.

The RAC strongly endorses the EPA's desire to estimate uncertainty bounds for its
radiogenic cancer risk estimates. The uncertainty bound estimates should incorporate, to the
extent possible, all sources of error and/or uncertainty, including the three main sources
identified in BEIR VII. Other sources of error and/or uncertainty identified by the EPA which
should be considered include dosimetry, disease detection, disease classification, temporal
patterns, and appropriate RBE values.

The RAC considered several additional complications that could influence uncertainty.
One such complication arises in the extrapolation, to lower dose ranges, of radiation effects seen
at dose levels for which statistically significant increases in cancer mortality or incidence have
been observed in the Life Span Study (LSS) and other epidemiological studies of exposed
populations. At such lower dose ranges, extrapolation may result in the risk estimates associated
with doses in the low-dose range having larger relative uncertainties than those in the higher
dose range.

BEIR VII specifically considered adaptive response, genomic instability, and bystander
effects, and concluded that currently there are insufficient quantitative data to include these
effects in the dose-response model. The RAC recommends that EPA discuss potential problems
associated with the use of LNT dose response model risk estimates in these very low dose
settings.

It is important to note that there is an opportunity to implicitly include (qualitative)
uncertainties in the choice of risk model per se in the overall (quantitative) uncertainty analysis.
That is, a major issue with the choice of the LNT model is whether it is appropriately applied at
low doses. In the quantitative uncertainty analysis, this qualitative uncertainty in model choice
can be included as a quantitative uncertainty in the DDREF value. The RAC also strongly
endorses the EPA's intention to include uncertainty in DDREF in the overall uncertainty
analysis.

Uncertainties in risk estimates also change as a function of time into the future, being
smallest in the near time frame. This is due to several factors, including changes in future

-------
populations (actual as opposed to a 'stationary population'), future background cancer incidence,
and future medical advances (since the case fatality rate may decrease as a result of better
treatment interventions in the future). Uncertainties thus become greater as the risk estimates are
applied further into the future. The RAC recommends that EPA include a discussion of these
concepts in its final report.

The RAC considers it premature to offer any advice to the EPA on estimating the risk of
radiogenic thyroid cancer. A major review of radiogenic thyroid cancer is being completed by
the National Council on Radiation Protection and Measurements (NCRP). This information
should be considered by the EPA as it will reflect more recent data that could improve the
thyroid cancer risk estimates provided by BEIR VII.

In summary, the SAB finds that the draft dated August 1, 2006 and entitled "Modifying
EPA Radiation Risk Models Based on BEIR VII, " is an important document to provide the basis
for EPA's update of radiogenic cancer risk estimates. The RAC appreciates the opportunity to
review this draft document and hopes that the recommendations contained herein will enable
EPA to implement changes in the methodology for estimating radiogenic cancers and revise the
"Blue Book". We look forward to your response to the recommendations contained in this
Advisory.

Sincerely,
/Signed/ /Signed/

Dr. M. Granger Morgan Dr. Jill Lipoti
Chair Chair, Radiation Advisory Committee
Science Advisory Board Science Advisory Board

-------
                                       NOTICE

        This advisory has been written as part of the activities of the EPA Science Advisory
Board (SAB), a public advisory group providing extramural scientific information and advice to
the Administrator and other officials of the Environmental Protection Agency. The SAB is
structured to provide balanced, expert assessment of scientific matters related to problems facing
the Agency. This advisory has not been reviewed for approval by the Agency and, hence, the
contents of this advisory do not necessarily represent the views and policies of the
Environmental Protection Agency, nor of other agencies in the Executive Branch of the Federal
government, nor does mention of trade names of commercial products constitute a
recommendation for use.  Reports and advisories of the SAB are posted on the EPA website at
http://www.epa.gov/sab.

-------
                  U.S. Environmental Protection Agency (EPA)
                          Science Advisory Board (SAB)
                      Radiation Advisory Committee (RAC)

CHAIR
Dr. Jill Lipoti, Director, Division of Environmental Safety and Health, New Jersey Department
of Environmental Protection, Trenton, NJ

MEMBERS
Dr. Bruce Boecker1, Scientist Emeritus, Lovelace Respiratory Research Institute, Albuquerque,
NM

Dr. Thomas B. Borak, Professor, Department of Environmental and Radiological Health
Sciences, Colorado State University, Fort Collins, CA

Dr. Antone L. Brooks, Professor, Radiation Toxicology, Washington State University Tri-
Cities, Richland, WA

Dr. Brian Dodd, Consultant, Las Vegas, NV

Dr. Shirley A. Fry, Consultant, Indianapolis, IN

Dr. William C. Griffith, Associate Director, Institute for Risk Analysis and Risk
Communication, Department of Environmental and Occupational Health Sciences,  University of
Washington, Seattle, WA

Dr. Helen A. Grogan1, President, Cascade Scientific, Inc., Bend, OR

Dr. Richard W. Hornung1, Director of Biostatistics and Data Management, Cincinnati
Children's Hospital Medical Center, Division of General and Community Pediatrics, Cincinnati,
OH

Dr. Jonathan M. Links, Professor, Department of Environmental Health Sciences, Bloomberg
School of Public Health, Johns Hopkins University, Baltimore, MD

Mr. Bruce A. Napier, Staff Scientist, Radiological Science & Engineering Group, Pacific
Northwest National Laboratory, Richland, WA

Dr. Daniel O. Stram, Professor, Department of Preventive Medicine, Division of Biostatistics
and Genetic Epidemiology, Keck School of Medicine, University of Southern California, Los
Angeles,  CA

Dr. Richard J. Vetter, Radiation Safety Officer, Professor of Biophysics, Mayo Clinic,
Rochester, MN
       Former RAC member
                                          11

-------
SCIENCE ADVISORY BOARD STAFF
Dr. K. Jack Kooyoomjian, Designated Federal Officer, US EPA, Science Advisory Board
(1400F), 1200 Pennsylvania Avenue, NW, Washington, DC, 20460
                                       in

-------
                     U.S. Environmental Protection Agency
                             Science Advisory Board

CHAIR
Dr. M. Granger Morgan, Lord Chair Profesor in Engineering, Professor and Department Head,
Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, PA

SAB MEMBERS

Dr. Gregory Biddinger, Coordinator, Natural Land Management Programs, Toxicology &
Environmental Sciences, ExxonMobil Biomedical Sciences, Inc, Houston, TX

Dr. James Bus,  Director of Externar Technology, Toxicology and Environmental Research and
Consulting, The Dow Chemical Company, Midland, MI

Dr. Deborah Cory-Slechta, J. Lowell Orbison Distinguished Alumni Professor of
Environmental Medicine, Department of Environmental Medicine, School of Medicine and
Dentistry, University of Rochester, Rochester, NY

Dr. Maureen L. Cropper, Professor, Department of Economics, University of Maryland,
College Park, MD, and Lead Economist, The World Bank, Washington, DC

Dr. Virginia Dale, Corporate Fellow, Environmental Sciences Division, Oak Ridge National
Laboratory, Oak Ridge, TN

Dr. Kenneth Dickson, Professor, Institute of Applied Sciences, University of North Texas,
Denton, TX

Dr. Baruch Fischhoff, Howard Heinz University Professor, Department of Social and Decision
Sciences, Department of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh,
PA

Dr. James Galloway, Professor, Department of Environmental Sciences, University of Virginia,
Charlottesville, VA

Dr. Lawrence Goulder, Shuzo Nishihara Professor of Environmental and Resource Economics,
Department of Economics, Stanford University, Stanford, CA

Dr. James K. Hammitt, Professor of Economics and Decision Sciences, Harvard Center for
Risk Analysis, Harvard University, Boston, MA

Dr. Rogene Henderson, Scientist Emeritus, Lovelace Respiratory Research Institute,
Albuquerque, NM

Dr. James H. Johnson, Professor and Dean, College of Engineering, Architecture & Computer
Sciences, Howard University, Washington, DC
                                         IV

-------
Dr. Agnes Kane, Professor and Chair, Department of Pathology and Laboratory Medicine,
Brown University, Providence, RI

Dr. Meryl Karol, Professor Emerita, Graduate School of Public Health, University of
Pittsburgh, Pittsburgh, PA

Dr. Catherine Kling, Professor, Department of Economics, Iowa State University, Ames, IA

Dr. George Lambert, Associate Professor of Pediatrics, Director, Center for Childhood
Neurotoxicology, Robert Wood Johnson Medical School-UMDNJ, Belle Mead, NJ

Dr. Jill Lipoti, Director, Division of Environmental Safety and Health, New Jersey Department
of Environmental Protection, Trenton, NJ

Dr. Michael J. McFarland, Associate Professor, Department of Civil and Environmental
Engineering, Utah State University, Logan, UT

Dr. Judith L. Meyer, Distinguished Research Professor Emeritus, Institute of Ecology,
University of Georgia, Athens, GA

Dr. Jana Milford, Associate Professor, Department of Mechanical Enginweering, University of
Colorado, Boulder, CO

Dr. Rebecca Parkin, Professor and Associate Dean, Environmental and Occupational Health,
School of Public Health and Health Services, The George Washington University Medical
Center, Washington, DC

Mr. David Rejeski,  Director, Foresight and Governance Project, Woodrow Wilson International
Center for Scholars,  Washington, DC

Dr. Stephen M. Roberts, Professor, Department of Physiological Sciences, Director, Center for
Environmental and Human Toxicology, University of Florida, Gainesville, FL

Dr. Joan B. Rose, Professor and Homer Nowlin Chair for Water Research, Department of
Fisheries and Wildlife, Michigan State University, East Lansing, MI

Dr. Jerald Schnoor, Allen S. Henry Chair Professor, Department of Civil and Environmental
Engineering, Co-Director, Center for Global and Regional Environmental Research, University
of Iowa, Iowa City, IA

Dr. Kathleen Segerson, Professor, Department of Economics, University of Connecticut, Storrs,
CT

Dr. Kristin Shrader-Frechette, O'Neil Professor of Philosophy, Department of Biological
Sciences and Philosophy Department, University of Notre Dame, Notre Dame, IN

-------
Dr. Philip Singer, Professor, Department of Environmental Sciences and Engineering, School of
Public Health, University of North Carolina, Chapel Hill, NC

Dr. Robert Stavins, Albert Pratt Professor of Business and Government, Environment and
Natural Resources Program, John F. Kennedy School of Government, Harvard University,
Cambridge, MA

Dr. Deborah Swackhamer, Interim Director and Professor, Institute on the Environment,
University of Minnesota, St. Paul, MN

Dr. Thomas L. Theis, Director, Institute for Environmental Science and Policy, University of
Illinois at Chicago, Chicago, IL

Dr. Valerie Thomas, Anderson Interface Associate Professor, School of Industrial and Systems
Engineering, Georgia Institute of Technology, Atlanta, GA

Dr. Barton H. (Buzz) Thompson, Jr., Robert E. Paradise Professor of Natural Resources Law
at the Stanford Law School and Director, Woods Institute for the Environment, Stanford
University, Stanford, CA

Dr. Robert Twiss, Professor Emeritus, University of California-Berkeley, Ross, C A

Dr. Terry F. Young, Consultant, Environmental Defense, Oakland, CA

Dr. Lauren Zeise, Chief, Reproductive and Cancer Hazard Assessment Branch, Office of
Environmental Health Hazard Assessment, California Environmental Protection Agency,
Oakland, CA
SCIENCE ADVISORY BOARD STAFF
Mr. Thomas Miller, Designated Federal Officer, US EPA, Science Advisory Board (1400F),
1200 Pennsylvania Avenue, NW, Washington, DC, 20460
                                          VI

-------
                            TABLE OF CONTENTS
1.  EXECUTIVE SUMMARY	1

2.  INTRODUCTION	6

   2.1  Background	6
   2.2  Review Process and Acknowledgement	6
   2.3  Current EPA Cancer Risk Models	7
   2.4  BEIR VII Models	8
   2.5  EPA's Proposed Adjustments and Extensions to BEIR VII Models	8
   2.6  EPA Charge to the Committee	9

3.  RAC's PHILOSOPHY OF APPROACH FOR RESPONSE TO THE CHARGE	11

4.  RESPONSE TO CHARGE QUESTION 1: APPLICATION OF THE OVERALL APPROACH AS
             DESCRIBED IN THE DRAFT WHITE PAPER	12

5.  RESPONSE TO CHARGE QUESTION 2: WHITE PAPER MODIFICATIONS & EXTENSIONS	14

   5.1  Charge Question # 2	14
   5.2  Response to Charge Question* 2a	15
   5.3  Response to Charge Question* 2b	 15
   5.4  Response to Charge Question #2c	15
   5.5  Response to Charge Question #2d	16
   5.6  Response to Charge Question #2e	18
   5.7  Response to Charge Question #2f.	18
   5.8  Response to Charge Question #2g	21
   5.9  Response to Charge Question #2h	22


6.  RESPONSE TO CHARGE QUESTION 3: UNCERTAINTIES NOT QUANTIFIED IN BEIR VII	23

7.  RESPONSE TO CHARGE QUESTION 4: ISSUES RELATING TO RADIOGENIC THYROID
             CANCER NOT QUANTIFIED IN BEIR VII AND ISSUES BEYOND THE CHARGE ....26

8.  ISSUES BEYOND THE CHARGE	26

TABLE 1 - COMPARISON OF THE EPA WHITE PAPER (WP) and BEIR VII METHOD FOR
             COMBINING EAR and ERR LAR PROJECTIONS FOR LUNG CANCER INCIDENCE
             	17

REFERENCES CITED	27

APPENDIX A - ON-GOING RESEARCH AND PARADIGMS ASSOCIATED WITH BIOLOGICAL
             RESPONSES TO LOW DOSES OF RADIATION	35

APPENDIX B - ACRONYMS, SYMBOLS AND ABBREVIATIONS	39
                                         vn

-------
1. EXECUTIVE SUMMARY

The Radiation Advisory Committee (RAC) of the Science Advisory Board (SAB) has
completed its review of the Agency's draft White Paper entitled "Modifying EPA Radiation Risk
Models Based on Biological Effects of Ionizing Radiation (BEIR) VII, " dated August 1, 2006
(U.S. EPA. ORIA. 2006a). In this draft White Paper, the EPA's Office of Radiation and Indoor
Air (ORIA) outlined proposed changes in the Agency's methodology for estimating radiogenic
cancers. The EPA sought the RAC's advice on the application of BEIR VII's (U.S. National
Academy of Science /National Research Council 2006) cancer risk estimates and on issues
relating to proposed modifications and expansions desirable or necessary for EPA's purposes.

In providing advice to the Agency, the RAC had to consider the important distinction
between the current state of scientific knowledge and the need for a practical, operational public
health approach to radiation protection and standards setting. The RAC endorses EPA's
proposal to base its approach to low dose risk estimation on BEIR VII. Specifically, for the
purposes of establishing radiation protection policy, the RAC endorses the EPA's use of a Linear
Non-Threshold (LNT) model combined with the Dose and Dose Rate Effectivenesss Factor
(DDREF) for estimating cancer risks following low dose exposures. That is, the slope of the
dose-response relationship in the high dose region is modified by the DDREF which corrects for
the decreased biological effectiveness of low dose and dose-rate exposures. The resulting lower
slope is then linearly extrapolated into the very low dose and dose-rate region in which
epidemiological data usable in analyses have not and may not be obtained. By low dose, the
RAC follows BEIR VII's definition; that is, doses below 100 mSv (0.1 Sv), in the context of low
Linear Energy Transfer (LET) radiation.

With respect to recent advances in the scientific knowledge of radiation biology and
carcinogenesis, the RAC wishes to emphasize that considerable uncertainties remain in the risk
estimates for radiation-induced cancers, especially at low doses and low dose rates. The
epidemiological data below 100 mSv are not sufficient by themselves for risk estimation and
considerable cellular and animal data suggest complexities beyond the application of a simplified
deoxyribonucleic acid (DNA) damage model which historically has been used as support for an
LNT dose-response model. The RAC also emphasizes the additional complexities introduced
with varying Relative Biological Effectiveness (RBE) and dose-rate. Thus, while the RAC
endorses EPA's use of the LNT model, the Agency is advised to continue to monitor the science
of the biological mechanisms underlying cancer induction at low doses of ionizing radiation and
of their influence on the biophysical models used to estimate the cancer risk in this dose range.
Additional discussion of the biophysical models of radiation effects in the low-dose region is in
Appendix A.

-------
radiogenic lung cancer risk. However, the RAC finds that the EPA is warranted in modifying
the BEIR VII methodologies in several specific areas as discussed below.

The RAC agrees that the proposed estimation of radiogenic cancer risks for the U.S.
population using a standard stationary population, that is for a "fixed cohort," based on death
rates for the year 2000, is a reasonable adaptation of the BEIR VII approach. It is consistent
with the EPA's established approach to cancer risk estimation from exposures to chemicals.

The RAC agrees that the EPA's proposed use of the most current cancer-specific
incidence and mortality rates available is an appropriate and scientifically valid adaptation of the
BEIR VII approach.

The RAC agrees with the EPA's proposed approach for projecting risk estimates from
the Japanese A-bomb survivors to the U.S.A. population by combining the age-specific results
from the Excess Absolute Risk (EAR) and Excess Relative Risk (ERR) models using the
weighted geometric mean before calculating the lifetime attributable risk. This approach is a
modification of that used in BEIR VII, but it has the advantage of allowing the estimates of risk
from multiple exposures to be integrated, enabling the risk from chronic lifetime exposure to be
calculated. Additionally, this method was previously used by the EPA in Federal Guidance
Report 13 (FOR-13).

The RAC understands that EPA requires a rationale to estimate risks from exposures to
higher LET radiation, especially alpha particles, lower energy photons and beta particles, but this
subject was beyond the scope of BEIR VII. For alpha particles, the RAC is supportive of the use
of a generally accepted Maximum Relative Biological Effectiveness (RBEM) value, such as 20
which is currently being used. For those radionuclides for which human cancer risk data are
available (lung, liver, bone, or bone marrow), the RAC recommends that this information be
used directly whenever possible. For other organs and tissues, the RAC is supportive of the
general approach of using the low-LET cancer risk from BEIR VII multiplied by RBEM.

For low-energy photons and electrons, the EPA draft White Paper suggests that the RBE
for medical x-rays is about 2 to 2.5. X-rays are not uniquely different from gamma rays with
respect to their biological effects, so the RAC recommends that any risk estimate association
with exposure to photons should be correlated with energy rather than the method of production.
The RAC concurs that an RBE factor in the range of 2 to 2.5 is reasonable for low-energy
photons and electrons for purposes of setting radiation protection standards. The RAC concurs
that an RBE factor in the range of 2 to 2.5 is reasonable for tritium.

-------
The RAC recognizes that although the BEIR VII committee chose not to provide risk
estimates for non-melanoma skin cancer (NMSC) induced by ionizing radiation, EPA has an
operational need for such estimates. The RAC supports EPA's proposed use of the 1991
International Commission on Radiological Protection (ICRP) model to estimate the incidence
and mortality risks of radiogenic NMSC taking into account more recent findings that most of
the NMSCs attributable to low to moderate doses of low-LET ionizing radiation are of the basal
cell carcinoma (BCC) type (Shore 2001), and that the incidence rates of BCC have been
increasing substantially in recent decades among the general population based on a study of New
Hampshire cancer rates (Karagas et al. 1999). However, the RAC concurs with EPA that
because of the high baseline incidence rates and low mortality due to NMSC, it is inappropriate
to include risk estimates for radiogenic NMSC in the estimate of the incidence or mortality risk
for radiogenic cancer.

The risk of bone cancer from low-LET radiation is not specified in the BEIR VII report
but such information is required to consider the cancer risk from a bone-seeking beta-emitting
radionuclide such as 90Sr. The EPA proposes to divide the bone cancer risk observed in humans
exposed to alpha particles from 224Ra by an RBE to estimate the bone cancer risk from 90Sr. The
RAC concurs with this practical, operational approach to radiation protection.

BEIR VII does not provide risk estimates for in utero exposure to radiation, but the EPA
requires an estimate for its guidance documents. The RAC concludes that it would be reasonable
for the EPA to use the referenced estimates of cancer risk from in utero exposure to external
radiation sources, and the dose coefficients provided by the ICRP as a basis for developing its
risk estimates for in utero radiation exposure from internally-deposited radionuclides.

The RAC considers that it is premature for RAC to offer any advice to ORIA on
estimating the risk of radiogenic thyroid cancer. A major review of radiogenic thyroid cancer is
being completed by the National Council on Radiation Protection and Measurements (NCRP).
This information should be considered by the EPA as it will reflect more recent or more relevant
data that could improve the thyroid cancer risk estimates provided by BEIR VII.

The RAC strongly endorses the EPA's desire to estimate uncertainty bounds for its
radiogenic cancer risk estimates. The uncertainty bound estimates should incorporate, to the
extent possible, all sources of error and/or uncertainty, including the three main sources
identified in BEIR VII (sampling variability in the Life Span Study (LSS) data, transport of risk
from LSS to the U.S.A. population, and the appropriate value for DDREF at both high and low
doses of low-LET radiation). Other sources of error and/or uncertainty identified by the EPA
which should be considered include dosimetry, disease detection, disease classification, temporal
patterns, and appropriate RBE values.

The RAC considered several additional complications that could influence uncertainty.
One such complication arises in the extrapolation, to lower dose ranges, of radiation effects seen
at dose levels for which statistically significant increases in cancer mortality or incidence have
been observed in the LSS and other epidemiological studies of exposed populations. At such
lower dose ranges, risk estimates are based on an assumed LNT dose-response model and
method of extrapolation from higher-dose/higher-response data. This extrapolation may result in

-------
the risk estimates associated with doses in the low-dose range having larger relative uncertainties
than those in the higher dose range.

BEIR VII specifically considered adaptive response, genomic instability, and bystander
effects, and concluded that currently there are insufficient quantitative data to include these
effects in the dose-response model. The EPA proposes at the present time to follow BEIR VII
and use the LNT model combined with a DDREF for calculation of radiation risk. In the
absence of compelling scientific evidence to do otherwise, the RAC endorses the EPA's plan in
this regard.

When estimating radiation-induced cancer risks in any human population it is important
to recognize that typically the baseline overall cancer incidence and mortality rates are high and
variable, representing >40% of the 15 leading causes of illness and about 23% of the 15 leading
causes of death in the U.S.A. in 2003 (CDC/NCHS National Vital Statistics System, 2003).
Baseline cancer rates have been found to be influenced by various environmental factors such as
chronic infections, life style, diet and human factors such as genetic background (WHO, Stewart
and Kleihues 2003). The dose of interest to any "radiation exposed" population is in addition to
a highly variable natural background radiation dose (lower limit lifespan dose 60 mSv) that
changes as a function of elevation, geographical location and human activities. Depending on
the study design, epidemiological studies typically match the "exposed" or "diseased" study
population to a "non-exposed" or "non-diseased" comparison population with respect to the
variables known to influence baseline cancer rates so as to statistically relate the effect of the
exposure to the health outcome of interest as precisely as possible. At radiation exposures in the
range of natural background, it is difficult to distinguish radiation-induced changes in risk from
the baseline. Thus, as a cautionary note, the RAC recommends that the EPA discuss potential
problems associated with the use of LNT dose response model risk estimates in very low dose
settings. Currently at these low doses, statistically significant differences between the cancer
rates among "exposed" (defined study populations) and "non-exposed" (defined comparison
populations) are not observed. These near background doses are only a fraction of those that
have been found to be associated with statistically significant differences in cancer frequency
between "exposed" and "non-exposed" populations.

It is important to note that there is indeed opportunity to include uncertainties in the
model - that is, uncertainties in high-dose versus low dose behavior - in the overall uncertainty
analysis. In BEIR VII and the EPA's proposed approach to uncertainty estimation, this
"additional" uncertainty is contained within the uncertainty in the value for DDREF, since
DDREF is only invoked at lower doses. The RAC thus strongly endorses the EPA's intention to
include uncertainty in DDREF in the overall uncertainty analysis.

Uncertainties in risk estimates also change as a function of time into the future, being
smallest in the near time frame. This is due to several factors, including changes in future
(actual) populations (as opposed to a 'stationary population'), future background cancer
incidence, and future medical advances (since the case fatality rate may decrease as a result of
better treatment interventions in the future). Uncertainties thus become greater as the risk
estimates are applied further into the future. The RAC recommends that EPA include a
qualitative discussion of these concepts in its final report.

-------
       An additional source of uncertainty in risk estimates is associated with the mechanistic
biophysical model that is used in BEIR VII to support the LNT model in the low dose region. In
Appendix A, the RAC provides a brief review of current research and recommends that the EPA
remain aware of the research continuously updating the biophysical model used to support the
estimates of radiation risk following low dose radiation exposure.

       These recent advances provide a scientific basis for the observed non-linear dose-
response relationships seen in many biological systems (BEIR VII, Ko et al. 2006, Mitchel et al.
2004). They suggest that the mechanism of action of radiation-induced damage is different
following exposure to high doses than it is after low radiation doses.  It becomes important to
consider new paradigms associated with the biological responses to low doses of radiation and to
modify and further develop the models used to support the extrapolation of dose-response
relationships into dose regions where it is not possible  to measure changes in radiation-induced
cancer incidence/mortality in human populations.

-------
2. INTRODUCTION
2.1 Background

In 1994, the EPA published a report, referred to as the "Blue Book," which lays out the
EPA's methodology for quantitatively estimating radiogenic cancer risks (U.S. EPA. 1994)
http://epa.gov/radiation/docs/assessment/402-r-93-076.pdf A follow-on report made minor
adjustments to the previous estimates and presented a partial analysis of the uncertainties in the
numerical estimates (U.S. EPA. 1999a) http://epa.gov/radiation/docs/assessment/402-r-99-
003.pdf. Finally, the Agency published Federal Guidance Report 13 (U.S. EPA. 1999)
http://epa.gov/radiation/docs/federal/402-r-99-001.pdf which utilized the previously published
cancer risk models, in conjunction with International Commission on Radiological Protection
(ICRP) dosimetric models and the U.S.A. usage patterns, to obtain cancer risk estimates for over
800 radionuclides, and for several exposure pathways. These were later updated (U.S. EPA.
1999b) http://epa.gov/radiation/federal/techdocs.htmtfreportl3.

In 2006, the National Academy of Sciences/ National Research Council (NAS/NRC)
released Health Risks from Exposure to Low levels of Ionizing Radiation BEIR VII Phase 2
which primarily addresses cancer and genetic risks from low doses of low-LET radiation (BEIR
VII) (U.S. NAS/NRC. 2006) http://newton.nap.edu/catalog/11340.htmltftoc). In the EPA draft
White Paper: Modifying EPA Radiation Risk Models Based on BEIR VII, the Agency proposes
changes to the EPA's methodology for estimating radiogenic cancers, based on the contents of
BEIR VII (U.S. EPA. 2006a). The Agency expects to adopt the models and methodology
recommended in BEIR VII, but believes that certain modifications and expansions are desirable
or necessary for the EPA's purposes. EPA's ORIA requested the SAB to review the Agency's
draft White Paper and provide advice regarding the proposed approach to dose-response
assessment of radionuclides.
2.2 Review Process and Acknowledgement

In response to EPA's request, the SAB Radiation Advisory Committee (RAC) was
initially briefed on the draft White Paper topic at its public planning meeting of December 21,
2005 which was held at the National Air and Environmental Radiation Laboratory (NAERL) in
Montgomery, Alabama (see 70 Fed. Reg. 69550, November 16, 2005). The EPA issued its
external draft White Paper entitled "Modifying EPA Radiation Risk Models Based on BEIR VII, "
on August 1, 2006 (U.S. EPA. 2006a). The charge questions to the SAB were formally
submitted on August 31, 2006 (U.S. EPA. 2006b).

There are various levels of reviews which EPA can request from the SAB. These include
reviews, advisories, and commentaries. The request from EPA was for an "advisory" review of
the draft White Paper. The EPA requested consensus advice from the RAC on how to
incorporate the changes in cancer risk models described by BEIR VII and to extend the BEIR
VII models to areas not specifically addressed by the BEIR VII committee. The EPA's request
was described as a "mid-course correction" for the RAC to provide advice on a series of

-------
questions posed by the Agency to guide the EPA in incorporating the latest scientific thinking
into the Agency's risk estimates. The RAC was not asked to provide policy direction, therefore
the RAC did not consider the implications to EPA standards which may be an outcome of the
changes to the risk estimates. The RAC's only departure from the EPA's request was in the
Issue Beyond the Charge.

       The SAB RAC met in a public teleconference meeting on September 6, 2006 and
conducted a face-to-face public meeting on September 26, 27 and 28, 2006 for this advisory (see
71 Fed. Reg., 45545, August 9, 2006).  Additional public conference calls took place on
November 28, 2006, December  18, 2006, and March 9, 2007 (see 71 Fed. Reg., 62590, October
26, 2006. These notices, the charge to the RAC and other supplemental information may be
found at the SAB's Web site (http://www.sab.gov/sab). The quality review draft advisory dated
July  18, 2007 was forwarded to the Chartered SAB for their September 5, 2007 public
teleconference meeting (see  72 Fed. Reg., 46057, August 16, 2007). This advisory reflects the
suggested editorial changes from the Charter SAB.

       The draft document "Modifying EPA Radiation Risk Models Based on BEIR VII, "
August 1, 2006 was well written and provided much needed background. Similarly, with the
BEIR VII report, presentations by the EPA staff and other information provided to the RAC in
the course of the public meetings were found to be helpful. During the meetings, the EPA staff
worked diligently to augment their draft White Paper with additional pieces of information that
the RAC felt were necessary to assist with the advisory. The staff took care to honor all the
RAC's requests and demonstrated their patience.

       2.3    Current EPA Cancer Risk Models

       For most cancer sites, radiation risk models are derived primarily from epidemiologic
data  from the Life Span Study (LSS) of the atomic bomb survivors.  The EPA's models for
esophageal, stomach, colon,  lung, ovarian, bladder and "residual" cancers and leukemia were
adapted from the models published by Land and  Sinclair based on a linear non-threshold (LNT)
fit to the LSS data (Land and Sinclair. 1991).

       For each solid tumor site, gender, and age-at-exposure interval, there is a model
providing a coefficient for the excess relative risk (ERR) per gray (Gy) for cancer mortality,
which is assumed to be constant beginning at the end of a minimum latency period until the end
of life. Land and Sinclair present two sets of models known as the "multiplicative" and the
"National Institutes of Health (NIH)" models that differ in how one "transports" risk from the
Japanese LSS population to the United States population.  In the multiplicative model, it is
assumed that the ERR/Gy is the same in all populations, whereas, in the NIH model, it is
assumed that the excess absolute risk (EAR) is the same in different populations for the limited
period of epidemiological follow-up. Given the scarcity of information on how radiogenic
cancer risk varies between populations having differing baseline cancer rates, the EPA
previously adopted an intermediate geometric mean coefficient (GMC) model for each site,
where the risk coefficients were taken to be the weighted geometric mean of the corresponding
ERR and EAR coefficients for both the multiplicative and the NIH models (U.S. EPA.  1994).

-------
For leukemia, the treatment of the temporal response in the models was more complex,
but the approach for transporting risk to the U.S.A. population was analogous. Following the
approach of Land and Sinclair, the EPA also developed a GMC model for kidney cancer from
the LSS data. The EPA's models for other site- or type-specific cancers, including breast, liver,
thyroid, bone, and skin were based on various authoritative reports (NCRP 1980; NRC 1988;
ICRP. 199la, b; Gilbert. 1991.). Based primarily on ICRP recommendations at that time (ICRP
199la), for low doses and dose rates, for low LET, each coefficient was reduced by a factor of
two, dose and dose-rate effectiveness factor (DDREF), from that which would be obtained from
a LNT fit to the LSS data.

2.4 BEIR VII Models

BEIR VII cancer site-specific models (which focused mainly on low-LET radiation)
derived from the LSS differ from those of Land and Sinclair in several notable ways: (1) they are
derived primarily from cancer incidence rather than cancer mortality data; (2) mathematical
fitting is performed to better reflect the functional dependence of solid cancer risk on age at
exposure and attained age, (i.e., age at diagnosis of a cancer or age at death due to cancer
depending on the end-point of interest); (3) a weighted average of risk projection models is used
to transport risk from the LSS to the U.S.A. population; (4) a value for the DDREF of 1.5 is
estimated from the LSS and laboratory data; (5) quantitative uncertainty bounds are provided for
the site-specific risk estimates in BEIR VII.

For breast cancer and thyroid cancer, BEIR VII risk models are based on pooled analyses
of data from the LSS cohort, together with data from epidemiologic studies of medically
irradiated cohorts (Preston et al. 2002; Ron et al. 1995).

2.5 EPA's Proposed Adjustments and Extensions to BEIR VII Models

In the draft White Paper: Modifying EPA Radiation Risk Models Based on BEIR VII
(U.S. EPA. ORIA 2006a), the EPA staff outlined proposed changes in the Agency's
methodology for estimating radiogenic cancers, based on the contents of BEIR VII and some
ancillary information. For the most part, the Agency proposed to adopt the models and
methodology recommended in BEIR VII; however, the Agency outlined certain modifications
and expansions which it deemed desirable or necessary for the EPA's purposes. The objective of
BEIR VII was to derive/update cancer risk estimates for radiation exposures of 100 mSv or less,
primarily from external photon radiation based on the most current valid epidemiological and
experimental data available. In order to satisfy EPA's broader mission, the EPA established a
written basis for estimation of cancer risks outside BEIR VII's scope.

One significant extension proposed by EPA for consideration was the estimation of
cancer risks from exposures to higher Linear Energy Transfer (LET) radiations, especially to
alpha particles, and also to lower energy photons and beta particles. An important expansion
proposed by EPA to be considered was the estimation of risks from exposures to alpha particles,
and also to alpha emitters deposited in the lung and the bone. BEIR VII does not present any
risk estimates for radiogenic bone cancer. The EPA proposes to estimate bone cancer risk from

-------
data on radium injected patients and to multiply that risk by a quality factor to estimate the risk
from internally deposited beta-gamma emitting radioactive materials.

BEIR VII does not provide quantitative estimates of risk for skin cancer. It does not fully
address prenatal exposures. BEIR VII presents a model for estimating the risk of the radiogenic
thyroid cancer incidence, but not of mortality due to radiogenic thyroid cancer.

In the White Paper, the EPA proposed to use somewhat different population statistics
from BEIR VII. Consideration was given to an alternative model for estimating radiogenic lung
cancer. For breast cancer, the EPA proposed an alternative method for estimating mortality,
which takes into account changes in incidence rates and survival rates over time.

BEIR VII provides quantitative uncertainty bounds for each of its risk coefficients. The
uncertainty analysis focused on the three sources which the BEIR VII committee considered to
be most important, including sampling variability in the LSS data, extrapolation of the risk from
the LSS population to the US population, and the uncertainty in the DDREF at low doses. In the
White Paper, EPA noted a number of additional sources of uncertainty including the uncertainty
assigned to the form of the dose-response relationship. It was implicitly assumed that the dose-
response relationship followed the hypothetical dose-response curve depicted in Figure 10-1 in
BEIR VII. This figure implies a progression of linear approximations (as the tangent to the
curve at different doses), with different slopes (i.e., potentially different risk coefficients) at
different doses. The ratio of the slope at high doses and that of the tangent at zero dose provides
the definition of the DDREF. This definition allowed the BEIR VII Committee to place
uncertainty bounds on the DDREF. Of importance, however, mechanisms pertaining to the
biological effects of low-level ionizing radiation are still being investigated, which could
eventually lead to adoption of a different dose-response model, potentially resulting in changes
in estimates of risk at low doses (and, as a result, to the DDREF). EPA proposed to adopt the
BEIR VII quantitative uncertainty bounds for most purposes, but to include a brief discussion of
low dose extrapolation issues.

After receiving the advisory review of the Agency's draft White Paper , the Agency plans
to implement changes in their methodology through the publication of a revised Blue Book,
which it would expect to submit to the SAB for final review. The revised Blue Book could then
serve as a basis for an updated version of Federal Guidance Report 13 (FGR-13).

2.6 EPA Charge to the Committee

1) BEIR VII provides incidence models for many cancer sites as a basis for calculating the
risk from low-dose, low-LET radiation. Please comment on EPA 's application of this overall
approach as described in the draft White Paper.

2) In addition to the overall approach described in BEIR VII, the draft White Paper
presents specific modifications and extensions. Please comment on the soundness of the
follow ing proposals:

-------
a. Calculation of the risk to the life table (stationary) population instead of the actual
U.S. population (see Sections II.A.-C.); this is consistent with our current approach.

b. Use of more recent incidence and mortality data from SEER and/or other sources
(see Section II.D.); BEIR VII used a previous version of SEER data for the years
1995-1999.

c. Method for combining BEIR VII 's models for projecting risk from Japanese A-bomb
survivors to U.S. population (see Section II.E.). In contrast to BEIR VII, we propose
to combine the two risk models before integration to calculate the lifetime
attributable risk.

d. Adoption of an alternative model for radiogenic lung cancer risk which may better
account for the effects of smoking than the BEIR VII approach (see Section II. G.).

e. Method for calculating breast cancer mortality risk, accounting for the relatively
long time from detection until death (see Section II.H.).

f. Proposed approaches for extending risk estimates to radiations of different LET's - in
particular, deriving site-specific risk estimates for alpha or x radiations based on
models derived from the A-bomb survivors, who were primarily exposed to gamma
rays (see Section III).

g. Estimation of risks for sites not specified in BEIR VII, specifically bone and skin, for
which we propose to update our current approaches (see Sections III. A. and V,
respectively).

h. Estimation of risk due to prenatal exposure. EPA 's current lifetime risk estimates do
not include risk from prenatal exposure, and BEIR VII does not provide them. The
draft White Paper uses ICRP recommendations to project its risks of childhood
cancers induced by in utero exposure. Please comment on the soundness of the
approach described in the draft White Paper to apply ICRP as described in Section
IV.

3) BEIR VII provides quantitative uncertainty bounds for each of its risk coefficients. EPA
proposes to adopt this methodology with some additional discussion of the uncertainties not
quantified in BEIR VII. Please comment on the adequacy of this approach (see Section U.K.).

4) In Section VI, the draft White Paper discusses some issues relating to radiogenic thyroid
cancer. Does the RAC have any specific suggestions for dealing with this risk; e.g., does the
RAC have any advice on gender specificity, effectiveness of iodine -131 compared to gamma
rays, or estimation of thyroid cancer mortality?
10

-------
3. RAC's PHILOSOPHY OF APPROACH FOR RESPONSE TO
THE CHARGE

In providing advice in response to the Agency's specific request, the RAC had to
consider the important distinction between the current state of scientific knowledge and the need
for a practical, operational public health approach to radiation protection and standards setting.
In this Advisory, the RAC wishes to comment on both issues.

For the purposes of providing estimates of the risks of radiation-induced cancers as a
basis for setting radiation protection standards, the RAC endorses EPA's proposal to base its
approach to low dose risk estimation on BEIR VII. Specifically, for purposes of establishing
radiation protection policy, the RAC endorses the use of an LNT model combined with the
DDREF for estimating risks following low dose exposures. To elaborate, the slope of the dose-
response relationship in the high dose region is modified in the low-dose region by the DDREF
which corrects for the decreased biological effectiveness of low dose and dose-rate exposures.
The resulting lower slope is then linearly extrapolated into the very low dose and dose-rate
region, in which epidemiological data usable in analyses have not and may not be obtained. By
"low dose," the RAC follows BEIR VIFs definition; that is, doses below 100 mSv (0.1 Sv), in
the context of low-LET radiation. The RAC endorses the concept of using DDREF factors for
estimating the risk in the low dose region.

With respect to recent advances in the scientific knowledge of radiation biology and
carcinogenesis, the RAC wishes to emphasize that considerable uncertainties remain in the risk
estimates for radiation-induced cancers, especially at low doses and low dose rates. As BEIR
VII acknowledges, the epidemiological data below 100 mSv (0.1 Sv) are not sufficient by
themselves for risk estimation, and considerable cellular and animal data suggest complexities
beyond the application of a simplified DNA damage model which historically has been used as
support for an LNT dose-response model. The RAC also wishes to emphasize the additional
complexities introduced with varying RBE and dose rate. Thus, while the RAC endorses EPA's
use of the LNT model, the Agency is advised to continue to monitor the science of the biological
mechanisms underlying cancer induction at low doses of ionizing radiation and of their influence
on the biophysical models used to estimate the cancer risk in this dose range. Additional
discussion of the biophysical models of radiation effects in the low-dose region is in Appendix
A.
11

-------
4. RESPONSE TO CHARGE QUESTION 1: APPLICATION OF THE
OVERALL APPROACH AS DESCRIBED IN THE DRAFT WHITE PAPER
Charge Question 1: BEIR VII provides incidence models for many cancer sites as a basis for
calculating the risk from low-dose, low-LET radiation. Please comment on EPA 's application of
this overall approach as described in the draft White Paper.

The Radiation Advisory Committee (RAC) agrees with the EPA that the BEIR VII
methodologies using incidence models and data should be used wherever possible. The RAC
accepts the EPA's use of BEIR VII methodologies for deriving risk estimates for cancers of the
stomach, colon, liver, prostate, uterus, ovary, bladder, and other solid tumors. Furthermore, if
one of the four following conditions applies, then the RAC agrees that the EPA is warranted in
modifying the BEIR VII methodologies. The four possible conditions are:

1) Information and data are needed about subject matter not addressed in BEIR VII;
2) More recent or more relevant data exist which could improve or otherwise
influence the risk estimates;
3) Compelling evidence suggests the use of a more appropriate scientific method; or
4) The EPA's implementation requirements for practicality or applicability
necessitate an adaptation or other alternative to BEIR VII methodologies.

The RAC grouped all of the charge issues according to these conditions. For example,
under condition one, the RAC considered prenatal exposures, bone and skin cancers, x- and
alpha-particle radiations and tritium as areas not addressed by BEIR VII, and for which the EPA
has a need to derive a basis for risk estimates. An example of applying condition two is that the
use of the most recent Surveillance, Epidemiology, and End Results (SEER) data would improve
the risk estimate. Examples of condition three are issues where a more appropriate scientific
method was considered, i.e. in development of breast cancer risk estimates and the estimation of
uncertainty. An example of condition four is the use of a stationary or a standard population to
remove the variability in risk estimates associated with differences in cancer rates in age and
race distributions across locations and calendar years in the U.S.A. population.

The RAC concludes that the EPA's use of the gray (Gy) as the unit of radiation absorbed
dose is appropriate and agrees that modifying factors should be applied to the risk rather than
dose.

The RAC's approach to giving advice to the EPA is predicated on the basic premise that
risk estimates are for use in assessing population (or specific subgroup) risk, rather than risk to a
specific individual in that population or subgroup. In general, the EPA's risk estimates are
calculated separately for each age group, gender, and cancer site. At present, the EPA has not
produced risk estimates for those with increased susceptibility, although the RAC notes that after
reviewing human and animal data, the 1998 International Commission on Radiological
Protection (ICRP) report Genetic Susceptibility to Cancer selected a single best estimate of a 10-
fold increase in risk as appropriate for the purposes of modeling radiological impact. However,
12

-------
ICRP went on to state, "The presence of familial (high penetrance) genetic disorders in the
population is too low (<1%) for there to be a significant impact on risk in typical human
populations	The current estimates of radiation cancer risk already include an unknown
contribution from genetically radiosensitive subpopulations	Because of the high risk of
spontaneous cancer in familial disorders, low doses of radiation... are most unlikely to impact
significantly on life-time cancer risk in an affected individual." (ICRP Publication 79, 1998).
The EPA's policy is based on limiting the risk to a general population (of all possible
susceptibilities).
                                           13

-------
5. RESPONSE TO CHARGE QUESTION 2: WHITE PAPER
MODIFICATIONS AND EXTENSIONS
5.1 Charge Question # 2
In addition to the overall approach described in BEIR VII, the draft White Paper presents
specific modifications and extensions. Please comment on the soundness of the following
proposals:

a. Calculation of the risk to the life table (stationary) population instead of the actual U.S.
population (see Sections II.A.-C.); this is consistent with our current approach.

b. Use of more recent incidence and mortality data from SEER and/or other sources (see
Section II.D.); BEIR VII used a previous version of SEER data for the years 1995-1999.

c. Method for combining BEIR VII 's models for projecting risk from Japanese A-bomb
survivors to U.S. population (see Section II.E.). In contrast to BEIR VII, we propose to
combine the two risk models before integration to calculate the lifetime attributable risk.

d. Adoption of an alternative model for radiogenic lung cancer risk which may better
account for the effects of smoking than the BEIR VII approach (see Section II. G.).

e. Method for calculating breast cancer mortality risk, accounting for the relatively long
time from detection until death (see Section II.H.).

f. Proposed approaches for extending risk estimates to radiations of different LET's - in
particular, deriving site-specific risk estimates for alpha or x radiations based on models
derived from the A-bomb survivors, who were primarily exposed to gamma rays (see
Section III).

g. Estimation of risks for sites not specified in BEIR VII, specifically bone and skin, for
which we propose to update our current approaches (see Sections III. A. and V,
respectively).

h. Estimation of risk due to prenatal exposure. EPA 's current lifetime risk estimates do not
include risk from prenatal exposure, and BEIR VII does not provide them. The draft
White Paper uses ICRP recommendations to project its risks of childhood cancers
induced by in utero exposure. Please comment on the soundness of the approach
described in the draft White Paper to apply ICRP as described in Section IV.
14

-------
5.2 Response to Charge Question # 2a

Calculation of the risk to the life table (stationary) population instead of the actual U.S.
population (see Sections II.A.-C.); this is consistent with our current approach.

The RAC agrees that the proposed estimation of radiogenic cancer risks for the U.S.A.
population using a standard stationary population based on the year 2000 death rate, or fixed
cohort is a reasonable adaptation of the BEIR VII approach. Specifically, the use of a stationary
population produces risk estimates standardized to a population with fixed age and race
distributions. This approach removes the variability in risk estimates associated with differences
in cancer rates in age and race distributions across locations and calendar years in the U.S.A.
population. The RAC notes that estimates of the risk of radiogenic cancer per unit dose always
will be subject to modification due to changes in population demographics. Depending on
circumstances, it may be appropriate to use the actual population. The RAC notes that the
proposed approach is also consistent with the EPA's established approach to cancer risk
estimation from exposures to chemicals which may be useful for harmonization of EPA
approaches to contaminants (U.S. EPA. 2005a, U.S. EPA. 2005b, Also FR Vol 70, No. 66, pp
17765, April 7, 2005).

5.3 Response to Charge Question #2b

Use of more recent incidence and mortality data from SEER and/or other sources (see
Section II.D.); BEIR VII used a previous version of SEER data for the years 1995-1999.

The RAC agrees that the EPA's proposed use of the most current cancer-specific
incidence and mortality rates available is an appropriate and scientifically valid adaptation of the
BEIR VII approach.

It is anticipated that incidence or mortality data for the years 1998-2002 will be available
for the final calculations of radiogenic cancer incidence risk estimates from NCI's SEER
program. In contrast, only data from this program for 1995-1999 were available to BEIR VII.

Although other potential sources of valid, nationally representative data will be
considered by the EPA, the RAC considers that the most current SEER data are adequate and
preferred for consistency with the BEIR VII approach. The EPA may want to consider the latest
vital statistics report produced from the 2000 census for mortality rates if they become available
before the final report is produced.

5.4 Response to Charge Question #2c

Method for combining BEIR VII's models for projecting risk from Japanese A-bomb
survivors to U.S. population (see Section II.E.). In contrast to BEIR VII, we propose to
combine the two risk models before integration to calculate the lifetime attributable risk.

The RAC notes that there is considerable uncertainty in the application of risk estimates
developed from the Japanese atomic bomb survivors to the U.S.A. population. This uncertainty
15

-------
results from different genetic and lifestyle characteristics of the two populations and differences
in the baseline cancer risks. The RAC agrees with the EPA's proposed approach for projecting
risk estimates from the Japanese A-bomb survivors to the U.S.A. population by combining the
age-specific results from the Excess Absolute Risk (EAR) and Excess Relative Risk (ERR)
models using the weighted geometric mean before calculating the lifetime attributable risk. This
approach is a modification of that used in BEIR VII but is consistent with the method used
previously by the EPA in FGR-13. The Committee agrees with the general approach to deal
with the uncertainty in transport for cancers with background rates that differ between the USA
and Japan. The general approach is to perform some kind of averaging between relative risk and
absolute risk model results and using the weighted geometric mean is one reasonable choice.
EPA's proposed change to the BEIR methodology solves a technical problem: i.e. that if you
compute the geometric average of a risk projection for one exposure and the geometric average
for another exposure then the sum of these two is not equal to the risk projection for the sum of
the two exposures (also computed as a geometric average between EAR and ERR models). This
contradicts the fact that using either the ERR or the EAR models separately, the risks due to the
two exposures do sum up. Solving this technical problem is the rationale for the specific change
proposed by EPA, and the committee agrees that this is a reasonable modification of the method.

5.5 Response to Charge Question #2d

Adoption of an alternative model for radiogenic lung cancer risk which may better account
for the effects of smoking than the BEIR VII approach (see Section II.G.).

The RAC recommends that the EPA use the BEIR VII methodologies for deriving risk
estimates for radiogenic lung cancer risk. The RAC does not find compelling evidence to
suggest the use of the alternative model discussed by EPA.

The lung cancer risk estimates reported by BEIR VII are primarily based on analyses of
the LSS data. These estimates were not adjusted for cigarette smoking which is potentially an
important confounder and/or effect modifier. This problem of lack of adjustment for cigarette
smoking is further compounded by the fact that lung cancer incidence rates are lower in Japan
than the U.S.A. and the lung cancer incidence rate ratio of males to females is considerably
higher in Japan than in the U.SA. The BEIR VII Committee was aware of this problem and
chose to deal with it by using a risk transport model that more heavily weighted the EAR
estimates relative to ERR estimates, i.e., assigning the weight of 0.7 for EAR and 0.3 for ERR.
The BEIR VII Committee justified this approach based on mechanistic arguments and the
finding reported by Pierce (Pierce el al. 2003), that in the LSS population of Japanese atomic
bomb survivors the interaction between low LET radiation and smoking was consistent with an
additive effect. This weighting scheme results in a Lifetime Attributable Risk (LAR) that is
roughly twice as great among females as among males.

The EPA draft White Paper provided an alternative model to the BEIR VII lung cancer
risk estimates. EPA was concerned that the lack of adjustment for cigarette smoking and birth
cohort effects would result in an overestimate of risk in the U.S.A. population as well as female
to male incidence rate ratio that was too high. EPA proposed to use a pure EAR model for lung
cancer, equivalent to a weighting of 1.0 for EAR and 0.0 for ERR risk models.
16

-------
The RAC requested additional work on this problem from the EPA consisting of the
following tasks:

• Compare results of the calculation of LAR using BEIR VII weighting to 100% EAR model
and to alternative weighting schemes and/or the use of arithmetic, AM, or geometric, GM,
means.
Table 1 : Comparison of the EPA White Paper (WP) and BEIR VII Method for
Combining EAR and ERR LAR Projections for Lung Cancer Incidence.1

Sex
Male
Female
Combination
Method
RR weight2 = 0.0
WP
179
344
BEIR VII
179
344
Combination
Method
RR weight3 = 0.3
WP
186
401
BEIR VII
193
428
Combination
Method
RR weight = 0.5
WP
195
460
BEIR VII
203
495
Combination
Method
RR weight = 0.7
WP
206
541
BEIR VII
213
573
Combination
Method
RR= 1.0
WP BEIR VII
230 230
714 714
NOTE: Number of cases per 100,000 persons exposed to 0. 1 Gy . Because of the uncertainty the results do not
incorporate DDREF adjustment.
Results are shown for stationary populations and SEER incidence data for the years 1998-2002.
2Weight for projection based on EPA proposal
3 Weight for projection using BEIR VII.
• Consider how the additive ERR model for smoking and radiation provides evidence for the
appropriate weighting scheme.

• Consider papers additional to Pierce (2003) on the nature of the smoking /radiation
interaction.

Based upon EPA's response to these requests, Table 1 illustrates the effect upon LAR
estimates for lung cancer incidence of several different weighting schemes for the EAR and ERR
risk models. The columns labeled White Paper (WP) and BEIR VII reflect differences in how
the weighting was applied. BEIR VII used a weighted average of the final age-adjusted ERR
and EAR estimates on a log scale, while EPA first weighted each age-specific stratum and then
combined the weighted age-specific risk estimates. Inspection of the table reveals that the
difference in application of the weights produced very small changes in the WP and BEIR VII
LAR estimates. The weighting of 0.0 for ERR proposed by EPA produces LAR estimates that
are somewhat smaller than the weight of 0.3 for ERR chosen by BEIR VII, most notably for
17

-------
females. The RAC also notes that the evidence for a purely additive model is not compelling
based upon the literature review performed by EPA. There is some support for an interaction
between radiation exposure and cigarette smoking that is intermediate between additive and
multiplicative, similar to the weighting scheme selected by BEIR VII.

Accordingly, due to a lack of compelling evidence to depart from the weighting approach
used by BEIR VII, the RAC recommends that EPA should not employ alternative weighting
schemes.

5.6 Response to Charge Question #2e

Method for calculating breast cancer mortality risk, accounting for the relatively long time
from detection until death (see Section II.H.).

The RAC notes that the EPA adopts the approach used by BEIR VII to estimate the risk
of breast cancer in females, and that this approach differs from that used by BEIR VII to estimate
the risks for the majority of other solid cancers. However, the EPA questions some aspects of
BEIR VIFs breast cancer risk estimation method, in particular the changing clinical course of
breast cancer in conjunction with the relatively long survival period. Thus, the EPA has
identified several alternative methods for estimating the relative risk for radiogenic breast cancer
in an effort to take into account some of the temporal features that can influence the cancer's
clinical course and hence the risk estimates. The RAC concurs with the EPA's decision to
explore these alternative methods.

Specifically, the RAC concurs with the EPA's proposal to relate current breast cancer
mortality rates to retrospective incidence rates rather than current incidence rates to better reflect
the influence of life style changes, earlier breast cancer detection and treatment that could
influence survival and hence mortality rates over an extended period.

The RAC notes the potential for development of second cancers during the cancer
survival period. Such an event could be spontaneous or related to treatment of the initial cancer.
In the case of breast cancer, it could impact mortality reporting and loss of deaths attributed to
breast cancer.

The RAC suggests that the EPA explore the feasibility of using the BEIR VII approach
with the proposed method (above) with retrospective lagging incidence rates relative to current
mortality rates.

5.7 Response to Charge Question #2f

Proposed approaches for extending risk estimates to radiations of different LET's - in
particular, deriving site-specific risk estimates for alpha or x radiations based on models
derived from the A-bomb survivors, who were primarily exposed to gamma rays (see Section
III).

A significant extension requiring subject matter not addressed in BEIR VII is guidance
on how to deal with the estimation of risks from exposures to different LET radiation, especially
18

-------
alpha particles and lower energy photons and beta particles. Knowledge of these risks is
required particularly for dealing with the possible health risks from chronic irradiation from
alpha, beta, or gamma emissions from internally deposited radionuclides. A key feature of the
low-LET radiation exposures used in the analyses available in the BEIR VII report, especially
those based on the Japanese atomic bomb survivors, is that they involved a very brief, whole-
body exposure to radiation from an external source. In such a situation, all of the organs and
tissues of the body were irradiated and the long-term risks to these organs and tissues have been
studied directly. When dealing with internally deposited radionuclides, the situation is different
because the radionuclide is likely to be distributed non-uniformly in the body, with only a few
organs and tissues receiving most of the dose. This can change the spectrum of cancers
produced. Also, because of the possible long-term retention of some long-lived radionuclides,
the dose can continue to accumulate at a low dose rate over months or years. Dealing with these
differences is important but not necessarily straightforward as discussed below.

Higher LET Radiation

The RAC noted that the draft White Paper only considered alpha particles for
radionuclides inhaled or ingested.

Alpha Particles

The EPA draft White Paper discusses three possible approaches to estimating the lifetime
health risks from internally deposited alpha-emitting radionuclides. These three approaches are
discussed below:

a) Data from human populations exposed to alpha-emitting radionuclides.

Reliable risk data are available for the following organs and tissues (U.S. NAS/NRC.
1988: U.S. NAS/NRC. 1999; Koshurnikova et al. 2000; Gilbert et al. 2004):
- Bone cancer from radium dial painters and radium chemists exposed to 226'228Ra;
- Bone Cancer from ankylosing spondylitis patients exposed to 224Ra;
- Liver cancer from patients given Thorotrast (232Th) as an imaging agent;
- Leukemia from patients given Thorotrast (232Th) as an imaging agent;
- Lung cancer from uranium miners who inhaled 222Rn and progeny; and
- Lung cancer from Mayak Russian workers who inhaled 239Pu.

Since the lung, liver, bone and bone marrow are the major organs at risk for internally
deposited, alpha-emitting radionuclides, these populations provide important information on
carcinogenic risk for alpha-emitting radionuclides. The RAC notes that this information is based
on site-specific cancer mortality among groups whose total doses are generally well above the
low-dose region.

b) Data from life-span studies of laboratory animals exposed via various routes to graded
activity levels of alpha-emitting radionuclides.
19

-------
       Sizeable data bases are available for different species of laboratory animals exposed via
various routes to graded activity levels of different beta-, gamma- or alpha-emitting
radionuclides and studied for their lifetimes. These studies provide much information on the
life-span health effects but the number of variables involved including species, route of
exposure, animal husbandry and other factors make it difficult to extrapolate the risk results
directly to human populations in a consistent manner. However, they do provide useful
information on radionuclides for which no human data are available.  Such studies also help
define the influence of dose distribution and the relative effectiveness of high- and low-LET
radiations in those cases where studies of the high and low-LET emissions were examined in a
parallel manner under similar conditions.

   c) The most recent cancer risk data from the RERF studies of atomic bomb survivors
   exposed to low-LET radiation multiplied by a general RBEiu factor for alpha particles.

       This third, more general, approach assumes that an appropriate value for RBEM is known
and that it is appropriate to use this value with the cancer risk seen after a brief, high dose-rate
exposure received by the atomic bomb survivors to estimate cancers risks in a broad range of
organs and tissues for which no data are available for alpha-particle exposure.

       As discussed in Section III.A.3, Summary and Recommendations of the White Paper, the
EPA proposes to multiply site-specific gamma-ray cancer risk estimates by an RBE of 20 to
derive corresponding estimates of cancer risk from alpha radiation, with two exceptions:

       a) An RBE of 1 to 3 for leukemia induced by alpha emitters deposited in bone; and
       b) Continued use of models derived  from BEIR VI to estimate lung cancer risk from
   inhaled radon  progeny.

       The RAC recognizes the problems that the EPA has to deal with in adding consideration
of alpha-emitting radionuclides to the information already provided for low-LET radiation in the
BEIR VII report.  This particular issue is one example of the need for a practical, operational
public health approach to radiation protection and standards setting mentioned earlier in this
Advisory. On this basis, the RAC is supportive of the use of a generally accepted RBEM value
such as the 20 that they are using currently.  For those radionuclides for which human cancer risk
data are available  for the lung, liver, bone, or bone marrow, the RAC recommends that this
information be used directly whenever possible. For other organs and tissues, the RAC is
supportive of the general approach (except for bone cancer as discussed in Section 5.8) of using
the low-LET cancer risk from BEIR VII multiplied by RBEM.

       Low-energy Photons and Electrons

       The EPA draft White Paper suggests that the relative biological effectiveness (RBE) for
medical x rays is about 2-2.5. However, x-rays are not uniquely different from gamma-rays
except for their production. Any risk estimate associated with exposure to photons needs to be
correlated with the energy of the photon rather than the method of production.
                                          20

-------
Reviews by ICRU (1986) and Kocher et al. (2005) show that RBEs for low energy
photons, < 30 keV, and low energy electrons, <15 keV, are higher than one when compared to
higher energy x-rays and ^Co gamma-rays. A probability distribution by Kocher et al. (2005)
showed a median radiation effectiveness factor of approximately 2.4 for photons less than 30
keV and for ^H beta particles. Thus, an effectiveness factor for these low energy radiations in
the range of 2 to 2.5 seems reasonable.

Beta particles

The RAC concurs that an RBE in the range of 2-2.5 is reasonable for estimating the
cancer risk from exposure to tritium.

5.8 Response to Charge Question #2g

Estimation of risks for sites not specified in BEIR VII, specifically bone and skin, for which we
propose to update our current approaches (see Sections III.A. and V, respectively).

The risk of bone cancer from low-LET radiation is not specified in the BEIR VII report
but such information is required to consider the cancer risk from a bone-seeking beta-emitting
radionuclide such as 90Sr. In this case, the EPA proposes to do the reverse of what is discussed
above in Section 2f. Instead of multiplying a low-LET cancer risk by an RBE to estimate a high-
LET cancer risk, it proposes to divide the bone cancer risk observed in humans exposed to alpha
particles from 224Ra by an RBE to estimate the bone cancer risk from 90Sr (NCRP 1991). Once
again, this practical, operational approach to radiation protection and standards setting seems
appropriate and conservative for the task at hand.

The RAC recognizes that although the BEIR VII committee chose not to provide risk
estimates for non-melanoma skin cancer (NMSC) induced by ionizing radiation, EPA has an
operational need for such estimates. This presents the EPA with certain methodological
challenges given the high incidence and low mortality rates of NMSC among the US general
population and the limitations of available data. The RAC recognizes that squamous cell
carcinoma (SCC) is not without individual or social cost - removal can cause significant
cosmetic deformity and requires short term and continuing follow-up because of the potential for
metastases.

The RAC supports EPA's proposed use of the 1991ICRP model to estimate the incidence
and mortality risks of radiogenic NMSC taking into account more recent findings that most of
the NMSCs attributable to low to moderate doses of low-LET ionizing radiation are of the basal
cell carcinoma (BCC) type (Shore. 2001.). Also, based on a study of New Hampshire cancer
rates, the incidence rates of BCC have been increasing substantially in recent decades among the
general population (Karagas et al. 1999).

However, the RAC concurs with EPA that because of the high baseline incidence rates
and low mortality due to NMSC (both BCC and SCC), it is inappropriate to include risk
estimates for radiogenic NMSC in the estimate of the total risk for radiogenic cancer. The RAC
also notes that as ionizing radiation is not considered to be a risk factor for melanoma skin
cancer there is no rationale for risk estimation in this instance.
21

-------
5.9 Response to Charge Question #2h

Estimation of risk due to prenatal exposure. EPA 's current lifetime risk estimates do not
include risk from prenatal exposure, andBEIR VII does not provide them. The draft White
Paper uses ICRP recommendations to project its risks of childhood cancers induced by in
utero exposure. Please comment on the soundness of the approach described in the draft
White Paper to apply ICRP as described in Section IV.

BEIR VII does not provide risk estimates for in utero exposure to radiation. Even though
the risk from in utero exposure is a minor component of the overall radiogenic cancer risk, the
EPA requires an estimate for radiation protection and standard setting purposes.

Few human data exist on which to base an estimate of radiogenic cancer risk for in utero
exposure to radiation from either external sources or internally deposited radioactive materials.

The primary sources of data for external exposures are the Oxford Survey of Childhood
Cancer (Stewart et al., 1958; Mole, 1990) and as reviewed by Mettler and Upton, (1995) and by
Doll and Wakefield, (1997) and the studies of Japanese atomic bomb survivors exposed in utero
(Delongchamp et al., 1997). When all sources of uncertainty are taken into account, the risk
estimates from these studies are not incompatible with each other (Wakeford & Little, 2003).

The dose to the embryo/fetus from internally-deposited radionuclides has been reviewed
(NCRP, 1998; ICRP 2000) and ICRP (2001) provides organ/tissue dose coefficients (Sv/Bq) to
the embryo/fetus from chronic intake of individual radionuclides by the mother. These data can
be used to develop cancer risk estimates for the embryo/fetus exposed coincidentally to radiation
delivered at low dose rates from the same sources.

Given the paucity of the epidemiological data available for estimating cancer risks of in
utero exposure to low or high LET, the RAC advises that the EPA continue to monitor advances
in the subject area, as well as the science of the biological mechanisms underlying cancer
induction in such situations and of their influence on the biophysical models used to estimate the
cancer risk of in utero exposure.

The RAC concludes therefore that it would be reasonable for the EPA to use the cancer
risk estimates from the published studies of populations exposed to photons in utero as a basis
for developing its estimates of cancer risk for such exposures. The RAC similarly advises EPA
to use the dose coefficients provided by ICRP as a basis for developing its estimates for in utero
radiation exposure from internally-deposited radionuclides.
22

-------
6. RESPONSE TO CHARGE QUESTION 3: UNCERTAINTIES NOT
QUANTIFIED IN BEIR VII

Charge Question 3: BEIR VII provides quantitative uncertainty bounds for each of its risk
coefficients. EPA proposes to adopt this methodology with some additional discussion of the
uncertainties not quantified in BEIR VII. Please comment on the adequacy of this approach (see
Section ILK.).

The RAC strongly endorses the EPA's desire to estimate uncertainty bounds for its
radiogenic cancer risk estimates. Indeed, given the range of possible operational uses of the risk
estimates, as much effort should go into estimating the uncertainty bounds as into producing the
central or point risk estimates themselves.

Ideally, the uncertainty analysis would involve the development of a probability density
function for (site-specific) estimated risk, rather than bounds around a central or point risk
estimate. Such an approach, which has previously been considered by other national and
international committees, would facilitate risk estimation based on other than the average risk.
For example, such an approach might facilitate the identification of a minimum cost-of-errors (or
'loss') risk estimate for operational use (e.g., in risk-informed regulation). Under-estimation and
over-estimation of risk potentially lead to under- and over-regulation, respectively, each of
which have costs associated with them. However, the RAC believes that such an approach is not
likely to be practically achievable, and endorses the EPA's approach (central risk estimate with
uncertainty bounds, following BEIR VII).

The uncertainty bound estimates should incorporate, to the extent possible, all major
sources of error and/or uncertainty, including the three main sources identified in BEIR VII
(sampling variability in the LSS data, transport of risk from LSS to the U.S.A. population, and
the appropriate value for DDREF at both high and low doses of low-LET radiation (or,
equivalently, the appropriate use of the LNT dose-response model used for low dose
extrapolation)). Other sources of error and/or uncertainty identified by the EPA (including
dosimetry (of which neutron RBE is a factor), disease detection, disease classification, temporal
patterns, and appropriate RBE values) should also be considered.

By this the RAC suggests that the EPA should consider performing a quantitative
analysis of uncertainty in the components of the risk assessment equations to establish
uncertainty in the final estimate of risk. This process should be expanded to include a sensitivity
analysis that establishes a ranking of the input parameters. This ranking can provide a valuable
tool for determining which components merit further consideration, with the possible acquisition
of additional data, and those that do not merit further consideration because the influence of
these uncertainties on the final result is small.

There is some value to producing two sets of uncertainty bounds, one representing the
bounds on the (site-specific) central or point risk estimate for the method of combining the RR
and AR that the EPA finally chooses to use, the other representing combinations ranging from
100% RR through 100% AR. The former gives a measure of the uncertainty of the central risk
23

-------
estimate derived from the method specifically used, and the latter gives an indication of the
range in which the true value (independent of method) likely resides.

In coming to these recommendations, the RAC considered several additional
complications that could influence uncertainty. One such complication arises because the
uncertainties associated with the current risk estimates for radiogenic cancers are smallest for the
doses at which statistically significant increases in cancer mortality or incidence have been
observed in the LSS and other epidemiological studies of exposed populations (Cardis, et.al,
2005). However, such increases have been observed over a limited range of individual doses.
At doses below this range, risk estimates are based on an assumed LNT dose-response model
and method of extrapolation from higher-dose/higher-response data. This extrapolation may
result in the risk estimates associated with doses in the low-dose range having larger relative
uncertainties than those in the higher dose range.

When estimating radiation-induced cancer risks in any human population it is important
to recognize that typically the baseline overall cancer incidence and mortality rates are high and
variable, representing >40% of the 15 leading causes of illness and about 23% of the 15 leading
causes of death in the U.S.A. in 2003 (Heron et al. CDC/NCHS, National Vital Statistics
Reports, 2007). Baseline cancer rates have been found to be influenced by various
environmental factors such as chronic infections, life style, diet and human factors such as
genetic background (WHO, Stewart and Kleihues 2003). The dose of interest to any "radiation
exposed" population is in addition to a highly variable natural background radiation dose (lower
limit lifespan dose 60 mSv) that changes as a function of elevation, geographical location and
human activities. Depending on the study design, epidemiological studies typically match the
"exposed" or "diseased" study population to a "non-exposed" or "non-diseased" comparison
population with respect to the variables known to influence baseline cancer rates so as to
statistically relate the effect of the exposure to the health outcome of interest as precisely as
possible. At radiation exposures in the range of natural background, it is difficult to distinguish
radiation-induced changes in risk from the baseline. Thus, as a cautionary note, the RAC
recommends that the EPA discuss potential problems associated with the use of LNT dose
response model risk estimates in very low dose settings. Currently at these low doses,
statistically significant differences between the cancer rates among "exposed" (defined study
populations) and "non-exposed" (defined comparison populations) are not observed. These near
background doses are only a fraction of those that have been found to be associated with
statistically significant differences in cancer frequency between "exposed" and "non-exposed"
populations.

It is important to note that there is an opportunity to implicitly include (qualitative)
uncertainties in the choice of risk model per se in the overall (quantitative) uncertainty analysis.
That is, a major issue with the choice of the LNT model is whether it is appropriately applied at
low doses. The RAC recommends that EPA discuss potential problems associated with the use
of LNT dose response model risk estimates in these very low dose settings.

In the quantitative uncertainty analysis, this qualitative uncertainty in model choice can
be included as a quantitative uncertainty in the DDREF value. The RAC strongly endorses the
EPA's intention to include uncertainty in DDREF in the overall uncertainty analysis.
24

-------
BEIR VII specifically considered adaptive response, genomic instability, and bystander
effects, and concluded that currently there are insufficient quantitative data to include these
effects in the dose-response model. The RAC recommends that the EPA include a (qualitative)
discussion of modern cellular and molecular biological concepts in its final report.

There is also a need to evaluate uncertainty following exposure to high doses delivered at
low dose-rates. In addition to the DDREF, it may be necessary to have a dose rate effectiveness
factor (DREF). The major data sets for these types of exposure come from internally deposited
radioactive materials both in experimental animals and in humans where the dose rates can be
low, but the total lifetime dose can be very high. The prime examples of such exposures in
humans are the doses to the lungs of uranium miners from inhaled radon/radon daughters inhaled
by uranium miners (U.S. NAS/NRC. 1999. BEIR VI, page 67) and the dose to bone from
internally deposited radium in the radium dial painters (Roland 1994).

In the uranium miners, an "inverse dose rate effect" was observed among miners exposed
to high total doses (WLM Working Level Months) of alpha-particle radiation over a relatively
short time period (months to a few years). These miners show a lower risk than that seen in
miners with the same total dose or WLM accumulated over many years. However, it was
determined that the mechanism of action for high and low dose rate exposures were different and
that ".. .the inverse exposure-rate effect found in the miner data should not modify the risks for
typical (radon) indoor exposures." (BEIR VI, page 9). Supralinearity associated with dose
fractionation is seen in studies of animals exposed to alpha-particle radiation.

The bone cancer frequency in the radium dial painters remained essentially at zero until
the total bone dose from the internally deposited alpha emitting radionuclide reached about 10
Gy, after which it increased rather markedly. This has been used to suggest a threshold dose
below which bone sarcomas are not induced by radiation exposure. Similar data were seen in
dogs that were exposed to beta emitting 90Sr-90Y by either inhalation (Gillett et al. 1992) or
ingestion (White et al. 1993). These low dose-rate exposures caused non-detectable changes in
cancer risk or life shortening until the total dose became very high (Raabe et al. 1981).

This discussion illustrates that the cancer risk estimates derived for acute exposure, even
with a DDREF of 1.5-2.0, do not result in accurate prediction of cancer risk to populations
exposed to high doses delivered at low dose rates. Such information needs to be considered
when predicting long term risk from low dose-rate exposures.

Uncertainties in the estimates are also a function of time into the future, being smallest in
the near time frame. This is due to several factors, including changes in future (actual)
populations (as opposed to a 'stationary population'), future background cancer incidence, and
future medical advances (since the case fatality rate may decrease as a result of better treatment
interventions in the future). Uncertainties thus become greater as the risk estimates are applied
further into the future. The RAC recommends that the EPA include a (qualitative) discussion of
these concepts in its final report.
25

-------
7. RESPONSE TO CHARGE QUESTION 4: ISSUES RELATING TO
RADIOGENIC THRYOID CANCER NOT QUANTIFIED IN BEIR VII
AND ISSUES BEYOND THE CHARGE

Charge Question 4: In Section VI, the draft White Paper discusses some issues relating to
radiogenic thyroid cancer. Does the RAC have any specific suggestions for dealing with this
risk; e.g., does the RAC have any advice on gender specificity, effectiveness of iodine -131
compared to gamma rays, or estimation of thyroid cancer mortality?

The RAC believes that it is premature to offer any advice to the Agency on this issue. A
major review of radiogenic thyroid cancer is being completed by the National Council on
Radiation Protection and Measurements. This information should be considered by the EPA as
more recent or more relevant data which could improve the risk estimates provided by BEIR VII.
8. ISSUES BEYOND THE CHARGE

The RAC received written and oral comments from members of the public which raised
concern about the need to set radiation protection standards for the most sensitive population for
specific cancer end points, instead of the use of "Reference Man." Some commenters expressed
strong opinions that the basis for risk estimates were too focused on "young Caucasian males"
and suggested that EPA expand the basis of risk estimates to include women, children, and non-
Caucasians. The RAC has addressed the issue in this report.

The RAC's approach to providing advice to the Agency, as outlined in Section 3 of this
report, is predicated on the basic premise that risk estimates are for use in assessing population
(or specific subgroup) risk, rather than risk to a specific individual in that population or
subgroup. In general, the EPA's risk estimates are calculated separately for each age group,
gender, and cancer site. At present, the EPA has not proposed risk estimates for subgroups
known to be genetically more susceptible to radiation than the general population as a whole.
This is an area of on-going research at the mechanistic level.

The EPA's present policy is based on limiting the risk to a general population (which
includes individuals of all possible susceptibilities). The RAC notes, however, that in the
existing Federal Guidance Report 13, the EPA has already used the current ICRP age groups
(infant, 5-10, 15-20 year olds) in calculating both the cancer risk coefficients and the underlying
radiation dose coefficient. However, the EPA has not explicitly accounted for gender or other
susceptibilities. The RAC recommends the EPA consider the concept described in ICRP
Publication 89 (ICRP, 2002) as a Reference Family, because it contains reference information on
persons at ages from newborns to adults and both genders; it also considers the results of studies
of Asian reference populations.
26

-------
                              REFERENCES CITED

Azzam, El and Little, JB. 2004. The radiation-induced bystander effect: Evidence and
significance. Human and Experimental Toxicology 23(2): 61-65, 2004.

Barcellos-Hoff, MH and Brooks, AL. 2001. Extracellular signaling through the
microenvironment: A hypothesis relating carcinogenesis, bystander effects and genomic
instability. Radiation Research 156(5 Pt 2): 618-627, 2001.

Barcellos-Hoff, MH. 2005.1ntegrative radiation carcinogenesis: Interactions between cell and
tissue responses to DNA damage. Seminars in Cancer Biology 15(2): 138-148, 2005.

Breckow, J. 2006. Linear-no-threshold is a radiation protection standard rather than a
mechanistic effect model. Radial. Environ. Biophys, 44:257-260, 2006.

Brooks, AL.  2004. Evidence for "bystander effects" in vivo. Human and Experimental
Toxicology 23(2): 67-70, 2004.

Brooks, AL.  2005. Paradigm shifts in radiation biology: Their impact on intervention for
radiation-induced disease. Radiation Research 164(4 Pt 2): 454-461, 2005.

Burma S, Chen BP, Murphy M, Kurimasa A, and Chen DJ. 2001. ATM phosphorylation histone
H2AX in response to DNA double-strand breaks. Journal of Biological Chemistry 276(45):
42462-467,2001.

Cardis E., Vrijheid M., Blettner M., Gilbert E., Hakama M, Hill C., Howe G, Kaldor J, Muirhead
CR, Schubauer-Berigan M, Yoshimura T, Bermann F, Cowper G, Fix J, Hacker C, Heinmiller B,
Marshall M,  Thierry-Chef I, Utterback D, Ahn Y-O, Amoros E, Ashmore P, Auvinen A, Bae J-
M, Bernar Solano J, Biau A, Combalot E, Deboodt P, Diez Sacristan A, Eklof M, Engels H,
Engholm G,  Gulis G, Habib R, Holan K,  Hyvonen H, Kerekes A, Kurtinaitis J, Malker H,
Martuzzi  M,  Mastauskas A, Monnet A, Moser M, Murata M, Pearce MS,  Richardson DB,
Rodriguez-Artalejo F, Rogel A, Tardy H, Telle-Lamberton M, Turai I, Usel M, Veress K. 2005.
Risk of cancer after low doses of ionising radiation - retrospective cohort study in 15 countries.
British MedicalJournal 331(7508):77.

Cardis E, Kesminiene A, Ivanov V, Malakhova I, Shibata Y, Khrouch V,  Drozdovitch V,
Maceika E, Zvonova I, Vlasov O, Bouville A, Goulko G, Hoshi M, Abrosimov A, Anoshko YA,
Astakhova L, Chekin S, Demidchik E, Galanti R, Ito M, Korobova E, Lushnikov E, Maksiutov
M, Masyakin V, Nerovnia A, Parshin V, Piliptsevich N, Pinchera A, Polyakov S,  Shabeka N,
Suonio E, Tenet V, Tsyb A, Yamashita S, Williams D. 2005. Risk of thyroid cancer following
Iodine-131 exposure in childhood. Journal of the National Cancer Institute, 97(10): 724-732,
2005.
                                          27

-------
Coleman MA and Wyrobek AJ. 2006. Differential transcript modulation of genes after low vs.
high doses of ionizing radiation, Chapter 10.6, In: Advances in Medical Physics, Edt: A.B.
Wolbarst, R.G. Zamenhof, and W.R. Hendee, Medical Editors: M.E. Clouse, A. Dritschilo, and
G. Cook, Medical Physics Publishing, Madison, Wisconsin, 2006.

Coleman MA, Yin E, Peterson LE, Nelson D, Sorensen K, Tucker JD and Wyrobek, AJ. 2005.
Low-dose irradiation alters the transcript profiles of human lymphoblasoid cells including genes
associated with cytogenetic radioadaptive response. Radiation Research 164(4 Pt 1): 369-382,
2005.

Delongchamp RR, Mabuchi K, Yoshimoto Y,  Preston DL. 1997. Cancer mortality among
atomic bomb survivors exposed in utero or as young children. Radiation Research 147: 385-395,
1997.

Di Masi A, Antoccia A, Dimauro I, Argentino-Storino A, Mosiello A, Mango R, Novell! G, and
Tanzarella C. 2006. Gene expression and apoptosis induction in p53-heterozygous irradiated
mice. Mutation Research 594(1-2): 49-62, 2006.

Ding L-H, Shingyoji M, Chen F, Hwang J-J, Burma S, Lee C, Cheng J-F, and Chen DJ. 2005.
Gene expression profiles of normal human fibroblasts after exposure to ionizing radiation: A
comparative study of low and high doses. Radiation Research 164(1): 17-26, 2005.

Dodd, B. 1990. The Validity of Population Dose and Cancer Risk Coefficients in the
Determination of Latent Cancer Fatalities. HP S Newsletter., April, 1990.

Doll R and Wakeford R.  1997. Risk of childhood cancer from fetal irradiation. Brit J Radiol 70:
130-139, 1997.

Federal Register Notice Citations:
FR, Vol. 70, No. 66, April 7, 2005, pp. 17765-17817 (U.S. EPA 2005 Guidelines for
Carcinogen Risk Assessment and Supplemental Guidance for Assessing Susceptability from Earl-
Life Exposure to Carcinogens).

FR, Vol. 70, No. 220, November 16, 2005, pp. 69550-69551;
FR, Vol. 71, No. 153, August 9, 2006, pp. 45545-45546;
FR, Vol. 71, No. 207, October 26, 2006, pp.  62590-62591;
FR, Vol. 72, No. 37, February 26, 2007, pp.  8379-8380; and
FR, Vol. 72, No. 158, August 16, 2007, pp. 46057-46058.

Gilbert ES.  1991. Chapter 3: Late somatic effects. In:  S Abrahamson, MA Bender, BB Boecker
et al. Health Effects Models for Nuclear Power Plant Accident Consequence Analysis.
Modifications of Models Resulting from Recent Reports on Health Effects of Ionizing Radiation,
Low LET Radiation, Part II: Scientific Bases for Health Effects Models. NUREG/CR-4214, Rev
1, Part II, Addendum 1, LMF-132, U.S. Nuclear Regulatory Commission, Washington, DC.
                                          28

-------
Gilbert ES, Koshurnikova NA, Sokolnikov ME, Shilnikova NS, Preston DL, Ron E, Okatenko
PV, Khokhryakov VF, Vasilenko EK, Miller S, Eckerman K, Romanov SA. .2004. Lung cancer
in Mayak workers. Radiation Res. 2004 Nov;162(5):505-16.

Gillett NA, Pool RR, Taylor GN, Muggenburg BA, Boecker BB. 1992. Strontium-90 induced
bone tumors in beagle dogs: effects of route of exposure and dose rate. InternationalJournal of
Radiation Biology. 61, 821-831.

Heron MP, Smith BL. 2007. U.S. Centers for Disease Control and Prevention, National Vital
Statistics Reports, Volume 55, Number 10, "Death: Leading Causes for 2003," March 15, 2007.
http://www.cdc.gov/nchs/data/nvsr/nvsr55_10.pdf

International Commission on Radiological Protection. 199la. Recommendations of the
International Commission on Radiological Protection. ICRP Publication 60. Ann ICRP 21 (1-3).

International Commission on Radiological Protection. 1991b. The Biological basis for Dose
Limitation in the Skin. ICRP Publication 59. Ann ICRP 22(2).

International Commission on Radiological Protection. 1998. Genetic Susceptibility to Cancer.
ICRP Publication 79. Ann ICRP Volume 28/1-2.

International Commission on Radiological Protection. 2000. Pregnancy and Medical Radiation.
ICRP Publication 84, Volume 30.1, Elsevier Science Ltd. New York. 2000.

International Commission on Radiological Protection. 2001. Doses to the Embryo and Fetus
from Intakes of Radionuclides by the Mother. ICRP Publication 88. Volume 31. 1-3. Elsevier
Science ltd. New York. 2001.

International Commission on Radiological Protection. 2002. Basic Anatomical and
Physiological Data for Use in Radiological Protection Reference Values, ICRP Publication 89,
Ann ICRP 32/3-4 (2002).

International Commission on Radiation Units and Measurements. The Quality Factor in
Radiation Protection. ICRU Report No. 40. Bethesda, MD. 1986.

Ishizaki K, Hayashi Y, Nakamura H, Yasui Y, Komatsu K, and Tachibana A. 2004. No
induction of p53 phosphorylation and few focus formation of phosphorylated H2AX suggest
efficient repair of DNA damage during chronic low-dose-rate irradiation in human cells. Journal
of Radiation Research 45: 521-525, 2004.

Kadhim MA, Moore SR, and Goodwin EH. 2004. Interrelationships amongst radiation-induced
genomic instability, bystander effects, and the adaptive response, Mutation Research 568(1): 21-
32, 2004.
29

-------
Karagas MR, Greenberg ER, Spencer SK, Stukel TA, and LA Mott. 1999. The New Hampshire
Skin Cancer Study Group: Increase in incidence rates of basal cell and squamous cell skin cancer
in New Hampshire, USA. Int J Cancer 81: 555-559, 1999.

Kennedy AR, Zhou Z, Donahue JJ and Ware JH. 2006. Protection against adverse biological
effects induced by space radiation by the Bowman-Birk inhibitor and antioxidants. Radiation
Res. 166 (2): 327-332, 2006.

Ko M, Lao X-Y, Kapadia R, Elmore E and Redpath JL. 2006. Neoplastic transformation in vitro
by low doss of ionizing radiation: Role of adaptive response and bystander effects. Mutation
Research 597: 11-17,2006.

Kocher DC, Apostoaei AI, Hoffman FO. 2005. Radiation effectiveness factors for use in
calculating probability of causation of radiogenic cancers. Health Phys. Jul;89(l):3-32, 2005.

Koshurnikova NA, Gilbert ES, Sokolnikov M, Khokhryakov VF, Miller S, Preston DL,
Romanov SA, Shilnikova NS, Suslova KG, Vostrotin VV. 2000. Bone cancers in Mayak
workers. RadiatRes. 2000 Sep;154(3):237-45.

Land CE and Sinclair WK. 1991. The relative contributions of different organ sites to the total
cancer mortality associated with low-dose radiation exposure. In: Risks Associated with Ionizing
Radiations. Annals of the ICRP 22(1), 1991.

Little JB. 2006. Cellular radiation effects and the bystander response. Mutation Research 597:
113-118,2006.

Lubin JH, Boice JD, edling C, Hornung RW, et al. 1995. Radon exposed underground miners
and inverse exposure-rate (protraction enhancement) effects. Health Physics 69:494-500, 1995.

Marchetti F, Coleman MA, Jones IM, and Wyrobek AJ. 2006. Candidate protein biodosimeters
of human exposure to ionizing radiation. InternationalJournal of Radiation Biology 82(9): 605-
639, 2006.

Mettler FA and Upton AC, 1995. Medical Effects of Radiation, pp 331-334. W.B. Saunders,
Philadelphia, 1995.

Mitchel REJ, Jackson JS, and Carlisle SM. 2004. Upper dose thresholds for radiation-induced
adaptive response against cancer in high-dose-exposed, cancer-prone, radiation-sensitive Trp53
heterozygous mice. Radiation Research 162: 20-30, 2004.

Mole R. 1990. Childhood cancer after prenatal exposure to diagnostic x-ray examinations in
Britain. BrJ Cancer 62: 152-168, 1990.

Morgan WF. 2003. Is there a common mechanism underlying genomic instability, bystander
effects and other nontargeted effects of exposure to ionizing radiation? Oncogene 22(45): 7094-
7099,2003.
30

-------
NCRP 1980. Induction of Thyroid Cancer by Ionizing Radiation. NCRP Report No 64. Bethesda,
MD: National Council on Radiation Protection and Measurements.

NCRP 1991. Some Aspects of Strontium Radiobiology; NCRP Report No. 110. Bethesda, MD:
National Council on Radiation Protection and Measurements.

NCRP 1998, Radionuclide exposure of the embryo /fetus, NCRP Report No 128, Bethesda MD:
National Council on Radiation Protection and Measurements.

Olivieri G, Bodycote J, and Wolff S. 1984. Adaptive response of human lymphocytes to low
concentrations of radioactive thymidine. Science 223(4636): 594-597, 1984.

Oxford Survey of Childhood Cancer, see Mettler and Upton. No date.

Pierce DA, Sharp GB, Mabuchi K.2003. Joint effects of radiation and smoking on lung cancer
risk among atomic bomb survivors. Radiation Research 2003 Apr; 159(4):511-20.

Ponnaiya B, Cornforth MN, and Ullrich RL. 1997. Radiation-induced chromosomal instability
in BALB/c and C57BL/6 mice: The difference is as clear as black and white. Radiation Research
147: 121-125, 1997.

Preston DL, Mattsson A, Holmberg E, Shore RE, Hildreth NG, and Boice Jr. JD. 2002.
Radiation effects on breast cancer risk: a pooled analysis of eight cohorts. Radiation Research
158:220-235,2002.

Raabe OG, Parks NJ, Book SA. 1981. Dose-Response Relationships for Bone Tumors in
Beagles Exposed to 226Ra and 90Sr. Health Physics 40, 863-880.

Roland RE. 1994. Radium in Humans, A Review of U.S. Studies, ANL/ER-3, UC-408.

Ron E, Lubin, JH, Shore, RE, Mabuchi,  K, Modam, B, Pottern, L,  Schneider, AB, Tucker, MA,
and Boice, JK. 1995.  Thyroid cancer after exposure to external radiation; a pooled analysis of
seven studies. Radiation Research 141: 259-277, 1995.

Shore RE. 1990.  Overview of radiation-induced skin cancer in humans. Int JRadiat Biol. Apr;
57(4):809-27, 1990.

Shore RE. 2001.  Radiation-induced skin cancer in humans. MedPediatr Oncol. 36(5):549-54,
May, 2001.

Spitz, DR, Azzam, El, Li, JJ, and Gius, D. 2004. Metabolic oxidation/reduction reactions and
cellular responses to ionizing radiation: A unifying concept in stress response biology. Cancer
and Metastasis Reviews 23(3-4): 311-322, 2004.
                                          31

-------
Stewart A, Webb J, Hewitt D. 1958. A survey of childhood malignancies. Br MedJ 1: 1495-
1508, 1958.

Thierry-Chef I, Marshall M, Fix JJ, Hermann F, Gilbert ES, Hacker C, Heinmiller B, Murray W,
Pearce MS, Utterback D, Bernar J, Deboodt P, Eklof M, Griciene B, Holan K, Hyvonen H,
Kerekes A, Lee M-C, Moser M, Pernicka F and E. Cardis. 2007. The 15-Country Collaborative
Study of Cancer Risk Among Radiation Workers in the Nuclear Industry: Study of Errors in
Dosimetry. Radiation Research. 167, 380-395, 2007.

Tubiana, M. 2005.  Dose-effect relationship and estimation of the carcinogenic effect of low
doses of ionizing radiation: The joint report of the Academie des Sciences (Paris) and of the
Academie Nationale de Medicine. Inter nationalJ. of Radiation: Oncology -Biology -Physics
63(2): 317-319,2005.

U. S. EPA (Environmental Protection Agency). 1994. Estimating Radiogenic Cancer Risks
("Blue Book"), Washington, DC (EPA 402-R-93-076), June, 1994:
http://epa.gov/radiation/docs/assessment/402-r-93-076.pdf

U.S. EPA (Environmental Protection Agency) / OAR (Office of Air and Radiation).  1999.
Federal Guidance Report (FGR)-13. Federal Guidance Report 13:  Cancer Risk Coefficients for
Environmental Exposure to Radionuclides, Washington, DC (EPA 402-R-99-001), September,
1999  http://epa.gov/radiation/docs/federal/402-r-99-001.pdf

U. S. EPA (Environmental Protection Agency) 1999a. Estimating Radiogenic Cancer Risks
Addendum: Uncertainty Analysis, Washington, DC (EPA 402-R-99-003), May, 1999:
http://epa.gov/radiation/docs/assessment/402-r-99-003.pdf

U.S. EPA. (Environmental Protection Agency) 1999b.  Update to the Federal Guidance Report
No. 13 and CD Supplement: http://epa.gov/radiation/federal/techdocs.htmtfreportl3

U.S. EPA SAB. 2002.  "PanelFormation Process: Immediate Steps to Improve Policies and
Procedures: An SAB Commentary, " EPA-SAB-EC-COM-02-003, May 17, 2002.

U.S. EPA. 2005a. Guidelines for Carcinogen Risk Assessment, EPA/630/P-03/001F, March 29,
2005

U. S. EPA. 2005b. Supplemental Guidance for Assessing Susceptability from Early-Life
Exposure to  Carcinogens. EPA/630/R-03/003F, March 29, 2005

U.S. Environmental Protection Agency, Office of Radiation and Indoor Air (ORIA). 2006a
"Modifying EPA  Radiation Risk Models based on BEIR VII, " Draft White Paper, Prepared by:
ORIA, U.S. Environmental Protection Agency, August 1, 2006
http://epa.gov/radiation/news/recentadditions.htm
                                          32

-------
U.S. EPA, Office of Radiation and Indoor Air. 2006b. Memorandum from Elizabeth A.
Cotsworth, Director, ORIA to Vanessa Vu, Director, Science Advisory Board Staff Office,
entitled "Advisory Review of the Draft 'White Paper: Modifying EPA Radiation Risk Models
Based on BEIR VII, "' August 31, 2006

U.S. NAS/NRC. 2006. BEIR VII. Health Risks from Exposure to Low levels of Ionizing
Radiation BEIR VIIPhase 2, National Academies of Sciences (NAS), National Research
Council, Committee to Assess Health Risks from Exposure to Low levels of Ionizing Radiation,
http://newton.nap.edu/catalog/11340.htmltftoc

U.S. NAS/NRC 1988. BEIR IV. Health Risks of Radon and Other Internally Deposited Alpha-
Emitters. National Research Council, Committee on the Biological Effects of Ionizing Radiation.
Washington, DC: National Academy Press.

U.S. NAS/NRC 1999. BEIR VI. Health Effects of Exposure to Radon. National Research
Council, Committee on Health Risks of Exposure to Radon.

Vrijheid M, Cardis E, Blettner M, Gilbert E, Hakama M, Hill C, Howe G, Kaldor J, Muirhead
CR, Schubauer-Berigan M, Yoshimura T, Ahn YO, Ashmore P, Auvinen A, Bae JM, Engels H,
Gulis G, Habib RR, Hosoda Y, Kurtinaitis J, Malker H, Moser M, Rodriguez-Artalejo F, Rogel
A, Tardy H, Telle-Lamberton M, Turai I, Usel M, Veress K. 2007. "The 15-Country
Collaborative Study of Cancer Risk Among Radiation Workers in the Nuclear Industry: design,
epidemiological methods and descriptive results." Radiation Research 167(4):361-79, April,
2007.

Wakeford R, Little MP. 2003. Risk coefficients for childhood cancer after intrauterine
irradiation: a review. Int JRadiat Biol. 79(5):293-309, May, 2003.
White RG, Raabe OG, Culbertson MR, Barks NJ, Samuels SA, Rosenblatt LS. 1993. Bone
Sarco
1993.
Sarcoma Characteristics and Distribution in Beagles Fed 90Sr. Radiation Research 136, 178-189,
WHO/OMS. 2003. World Health Organization, International Agency for Research on Cancer,
World Cancer Report, Edited by Stewart BS, Kleihues P. IARC Press, Lyon, 2003.

Zablotska LB, Ashmore JP, Howe GR. 2004. "Analysis of mortality experience amongst
Canadian nuclear power industry workers following chronic low-dose exposure to ionizing
radiation," Radiation Research 161(6) 633-41, 2004.
33

-------
                        Web-based Citations and Hotlinks
U. S. EPA (Environmental Protection Agency). 1994. Estimating Radiogenic Cancer Risks
("Blue Book"), Washington, DC (EPA 402-R-93-076), June, 1994:
http://epa.gov/radiation/docs/assessment/402-r-93-076.pdf

U.S. EPA (Environmental Protection Agency) / OAR (Office of Air and Radiation).  1999.
Federal Guidance Report (FGR)-13. Federal Guidance Report 13: Cancer Risk Coefficients for
Environmental Exposure to Radionuclides, Washington, DC (EPA 402-R-99-001), September,
1999  http://epa.gov/radiation/docs/federal/402-r-99-001.pdf

U. S. EPA (Environmental Protection Agency) 1999a. Estimating Radiogenic Cancer Risks
Addendum: Uncertainty Analysis, Washington, DC (EPA 402-R-99-003), May, 1999:
http://epa.gov/radiation/docs/assessment/402-r-99-003.pdf

U.S. EPA. (Environmental Protection Agency) 1999b. Update to the Federal Guidance Report
No. 13 and CD Supplement: http://epa.gov/radiation/federal/techdocs.htmtfreportl3

U.S. EPA. 2006.  Office of Radiation and Indoor Air (ORIA), Draft White Paper: Modifying
EPA Radiation Risk Models Based on BEIR VII, August 1, 2006
http://epa.gov/radiation/news/recentadditions.htm

U.S. NAS/NRC. 2006. BEIR VII. Health Risks from Exposure to Low levels of Ionizing
Radiation BEIR VIIPhase 2, National Academies of Sciences (NAS), National Research
Council, Committee to Assess Health Risks from Exposure to Low levels of Ionizing Radiation,
http://newton.nap.edu/catalog/11340.htmltftoc
                                          34

-------
APPENDIX A -ON-GOING RESEARCH AND PARADIGMS
ASSOCIATED WITH BIOLOGICAL RESPONSES TO LOW DOSES OF
RADIATION

According to the BEIR VII report, "Atomic bomb data for solid tumors combined
provide statistical evidence of a radiation-associated excess at doses down to around 100 mSv;
these combined data are well described by a linear no-threshold dose-response, although some
low dose nonlinearity is not excluded (US NAS/NRC. 2006. BEIR VII, p. 245)." "It is
abundantly clear that direct epidemiological and animal approaches to low dose cancer risk are
intrinsically limited in their capacity to define possible curvilinearity or dose thresholds for risk
in the range of 0-100 mSv. For this reason the present report has placed much emphasis on the
mechanistic data that can underpin such judgments (US NAS/NRC. 2006. BEIR VII, p.245)."

The uncertainty associated with the use of the epidemiological data to estimate risk in the
low dose range has been covered in detail in Charge Question 3: Uncertainties not Quantified in
BEIR VII. Recent articles examining the risk of cancer in cohorts of workers in the nuclear
industry in 15 countries showed that there is a small excess risk of cancer, even at the low doses
and dose rates typically received by nuclear workers (Cardis, et.al, 2005; Vrijheid, et al, 2007;
Thierry-Chef, et.al, 2007). However, these findings, including the risk estimates, are preliminary
and appear to be influenced by the as yet unexplained large cancer risk estimate for the
subcohort of Canadian nuclear workers (Zablotska et al, 2004) included in the study.

An additional source of uncertainty in risk estimates is associated with the DDREF and
the mechanistic biophysical model that is used in BEIR VII to support the LNT in the low dose
region. It is well established that it is not possible to use a linear extrapolation from health
effects produced by high radiation doses to predict those induced in the low dose and dose-rate
region. To make this low dose estimate, the slope of the dose-response relationship in the high
dose region is modified by the (DDREF) which corrects for the decreased biological
effectiveness of low dose and dose-rate exposures. The resulting lower slope is then linearly
extrapolated into the very low dose and dose-rate region below where useful epidemiological
data is obtainable. The major question discussed in this appendix is the applicability of the LNT
model in this very low dose region.

Although the BEIR VII committee conducted an extensive review of the cell and
molecular literature relative to biological responses at low doses and discussed the recent
advances, they concluded that the mechanistic cell and molecular biological research supported
the current biophysical model that they use (US NAS/NRC. 2006. BEIR VII, pp. 63-64).
However, the rapid increase in information on the biological responses to low doses of radiation
suggest new paradigms in radiation biology (Brooks 2005) that may modify the biophysical
model used in the BEIR VII report.

BEIR VII uses a biophysical model that suggests that each and every ionization increases
the probability of a DNA breakage (Burma et al. 2001) and that this results in a linear increase in
the risk for mutations and therefore in the risk for cancer (US NAS/NRC. 2006. BEIR VII, pp.
10-11). This model assumes independent action of cells and a lack of cell communication. The
model suggests that there is no change in response as a function of previous radiation exposure
35

-------
and that there is a linear link between unrepaired DNA damage, rare mutational events and the
development of cancer. Recent research has been conducted to provide a solid data base on the
response of molecules, cells, tissues and organisms to very low doses and dose rates of radiation
(Ko et al. 2004.; Azzam and Little 2004.; Little 2006.; Brooks 2005.; Mitchel et al. 2004.). This
research has suggested that several of the assumptions used in the BEIR VII biophysical model
may no longer be valid (Tubiana 2005). The data base that questions the assumptions used by
BEIR VII include information on dose dependent changes in gene expression, radiation induced
changes in redox status of the cells, apoptosis, bystander effects, adaptive responses, and
genomic instability (Spitz et al. 2004; Di Masi et al. 2006.; Coleman et al. 2005.; Azzam and
Little 2004.; Little 2006.; Brooks 2004.). The BEIR VII report has discussed each of these
effects and concluded that until molecular mechanisms of action involved in the induction of low
dose biological effects are elucidated, they cannot be utilized in modification of dose-response
relationships. This appendix provides a brief review on the mechanistic research being
conducted and to suggest the need for continuously updating the biophysical model used to
support the estimates of radiation risk following low dose radiation exposure.

It is well known that cells communicate by a variety of direct and indirect mechanisms
(Kadhim et al, 2004; Azzam and Little, 2004). Many new radio-biological observations indicate
that cells do not respond to radiation independently. This communication results in modification
of responses to low dose and dose-rate radiation.

Using recently developed microbeams and other technology to expose individual cells
and study the response of the "hit" cells and the response of neighboring cells demonstrated the
presence of "bystander effects." These effects demonstrate that a cell traversed by an alpha
particle or "hit" by a focused low LET beam communicate with neighboring cells and can
produce changes in "non-hit" cells. These changes have been shown to be both "harmful" and
"protective" and are most marked following exposure to high-LET radiation (Little 2006.).
Bystander effects impact the current use of "hit-theory" in defining radiation risk since the
radiation target is much larger than the individual cell. The research demonstrates that cells
communicate within each tissue making the assumption of independence of action of individual
cells used in the BEIR VII biophysical model inappropriate. Since non-hit cells show biological
responses, it may not be appropriate to calculate radiation dose to individual cells or cell types in
tissues (US NAS/NRC 2006. BEIR VII page 54). Bystander effects also make it more difficult
to define the biological target for the interaction of radiation with cells and the induction of
cancer. The data suggest that tissues and organs respond as a whole and that the biological
response is related to the dose to the whole organ/tissue, which is the metric used by BEIR VII in
all the human studies, rather than to the dose to individual cells (Barcellos-Hoff and Brooks
.2001.).

It has been demonstrated that following exposures to low doses of radiation there are
unique dose-dependent changes in gene and protein expression which were not recognized or
identified when the BEIR VII biophysical models were developed (Ding et al. 2005.; Coleman
and Wyrobek 2006.; Marchetti et al. 2006.). Low dose activation of such mechanisms supports
the existence of non-linear dose-response relationships for low-LET radiation. For bystander
effects following exposure to high LET radiation, the mechanisms can support either protective
mechanisms or mechanisms that increase risk. For gene activation from low LET radiation the
36

-------
mechanisms are related to adaptive and protective mechanisms. Identification of these genes is
providing a scientific basis for defining metabolic pathways activated by radiation and
determining mechanisms of action.

Previous radiation exposure can alter the response producing diminished biological
effects. This is called the "adaptive response." Two different types of adaptive responses have
been identified (Azzam and Little 2004.). The first is where low doses of radiation decrease the
amount of damage observed relative to background levels (Ko et al. 2006.). The second is where
a small "priming dose" of radiation given before a high acute "challenge dose" results in a
decreased response relative to the high dose alone (Olivieri et al .1984.). The ability to produce
an adaptive response is dependent on genetic background of the cells. Different sets of genes are
up and down regulated in cells capable of adaptation compared to cells that cannot adapt to
radiation exposure. Cells and tissues that demonstrate an adaptive response following low dose
exposures have repair and stress genes up regulated (Coleman et al. 2005.).

Research has been conducted to understand cell/cell and cell/tissue interactions and how
they modify cancer frequency (Barcellos-Hoff 2005.). Tissue interactions have been shown to
modify the expression of cellular and molecular damage and to be critical in the expression of
cancer. There is evidence that under certain experimental conditions, radiation damage can be
modified in vitro (Kennedy et al. 2006). Also administration of stable iodine considerably later
than the period normally prescribed to block exposure to radioactive iodine was unexpectedly
associated with a decreased risk of thyroid cancer incidence among a population at risk of
exposure as a result of the Chernobyl accident. The authors suggested that this finding may be
related to a modification of radiation-induced cellular or molecular damage in the presence of
stable iodine (Cardis et al. 2005). Data from this research verified that the initial DNA damage
increases linearly with radiation dose, that DNA damage triggers many molecular responses and
that even the initial DNA damage and repair is modified by radiation type, dose and dose-rate
(Ishizaki et al. 2004.). Importantly, it has been shown that biological repair of this damage as
well as the other cellular and organ responses are very non-linear over the low dose region.
These new findings may have significance in quantifying the safety margins associated with
regulatory standards.

Genomic instability suggests that, in addition to rare mutational events, frequent
radiation-induced changes following exposure may play an important role in cancer induction.
Radiation-induced genomic instability is seen at a high frequency in cells many cell divisions
after the radiation exposure (Morgan 2003; Ponnaiya et al. 1997.). The instability results in
increased frequency of mutations, chromosome aberrations, and cell killing. Radiation-induced
genomic instability seems to be one of the early stages in the carcinogenesis process and has
been seen both in vitro and in vivo. These observations challenge the relative importance that
initial mutations play in radiation-induced cancer (Kadhim et al. 2004.). The BEIR VII
biophysical model suggests that since DNA damage increases as a linear function of acute
radiation dose that there must be a linear increase in cancer risk (BEIR VII pp. 245). Genomic
instability and the ability to modify responses after the radiation exposure both challenge the
linear relationship between initial DNA damage and cancer frequency.
37

-------
       The magnitude of the response for all of these phenomena has been shown to be
dependent on the genetic background of the cells, tissues and organisms in which they are being
measured (Coleman et al. 2005; Ponnaiya et al. 1997; Azzam and Little 2004.; Little 2006.). A
better definition of the range of inter-individual variability and the development of analytical
methods and tools may make it possible to identify individuals that are either sensitive or
resistant to either the early or late effects of radiation or both.  However, currently it is not
possible to identify either radiation resistant or radiation sensitive individuals, or to use this
information in a regulatory framework.

       These recent advances provide a scientific basis for the observed non-linear dose-
response relationships seen in many biological systems (US NAS/NRC. 2006. BEIR VII; Ko et
al. 2006; Mitchel et al. 2004). They suggest that the mechanism of action of radiation-induced
damage is different following exposure to high doses than it is after low radiation doses. It
becomes important to consider new paradigms associated with the biological responses to low
doses of radiation and to modify and further develop the models used to support the
extrapolation of dose-response relationships into dose regions where it is not possible to measure
changes in radiation-induced cancer incidence/mortality in human populations.
                                           38

-------
          APPENDIX B -ACRONYMS, Symbols and Abbreviations
A-Bomb
AM
AR
BCC
BEIR
BEIR VII
CDC
CFR
Co
DREF
DDREF
DFO
DNA
EAR
EPA
ERR
FR
FOR-13
GM
GMC
GSD
Gy

H
I
IARC
ICRP
ICRU
IREP
keV
LAR
LET
LNT
LSS
mSv
NAS

NCHS
Atomic Bomb
Arithmetic Mean
Absolute Risk
Basal Cell Carcinoma
Pertains to committees of the Board of Radiation Effects, National Research
Council of the National Academy (now the National Academies'), charged with
assessing the Biological Effects of Ionizing Radiation
The report entitled "Health Risks from Exposure to Low Levels of Ionizing
Radiation BEIR VII - Phase 2 " published (2006) by the Committee to Assess
Health Risks from Exposure to Low Levels of Ionizing Radiation of the Board on
Radiation Effects Research, National Research Council of the National
Academies
Centers for Disease Control and Prevention
Code of Federal Regulations
Chemical symbol for Cobalt (60Co isotope)
Dose Rate Effectiveness Factor
Dose and Dose-Rate Effectiveness Factor
Designated Federal Officer
Deoxyribonucleic Acid
Excess Absolute Risk
Environmental Protection Agency (U.S. EPA)
Excess Relative Risk
Federal Register
Federal Guidance Report 13
Geometric Mean
Geometric Mean Coefficient
Geometric .Standard Deviation
Gray., SI unit of radiation absorbed dose (IGy is equivalent to 100 rad in
traditional units)
Chemical symbol for Hydrogen (3H isotope)
Chemical symbol for Iodine (131I isotope)
International Agency for Research on Cancer
International Commission on Radiological Protection
International Commission on Radiation Units and Measurements, Inc.
Interactive RadioEpidemiological Program
kiloelectron Volts
Lifetime Attributible Risk
Linear Energy Transfer
Linear Non Threshold
Life .Span Study
milli-Stievert
formerly the National Academy of .Sciences (U.S. NAS), now known as the
National Academies'
National Center for Health Statistics
                                         39

-------
NCI
NCRP
NIH
NIOSH
NMSC
NRC
OAR
ORIA
PAG
Pu
QA
QC
QA/QC
R
Ra
RAC
rad

RBE
RBEm
REF
rem

RERF
R/h
Rn
RR
SAB
SCC
SEER
SI

Sr
Sv

Th
UNSCEAR
US
WLM
WP
National Cancer Institute
National Council on Radiation Protection and Measurements
National Institutes of Health
National Institute for Occupational Safety and Health
Non-Melanoma Skin Cancer
National Research Council
Office of Air and Radiation (U.S. EPA/OAR)
Office of Radiation and Indoor Air (U.S. EPA/OAR/ORIA)
Protective Action Guide
Chemical symbol for Plutonium (239Pu Isotope)
Quality Assurance
Quality Control
Quality Assurance/Quality Control
Roentgen
Chemical symbol for Radium (Isotopes include 224Ra, 226Ra,
228Ra,and236Ra)
Radiation Advisory Committee (U.S. EPA/SAB/RAC)
Traditional unit of radiation absorbed dose in tissue (a dose of 100 rad is
equivalent to 1 gray (Gy) in SI units)
Relative Biological Effectiveness
Maximum Relative Biological Effectiveness
Radiation Effectiveness Factor
Radiation  equivalent in man; traditional unit of effective dose equivalent (equals
rad x tissue weighting factor) (100 rem is equivalent to 1 Sievert (Sv))
Radiation Effects Research Foundation
Roentgen per hour; traditional measure of exposure rate
Chemical symbol for Radon  (222Rn Isotope)
Relative Risk
Science Advisory Board (U.S. EPA/SAB)
Squamous Cell Carcinoma
Surveillance, Epidemiology,  and End Results
International System of Units (from NIST, as defined by the General Conference
of Weights & Measures in 1960)
Chemical Symbol for Strontium (90Sr Isotope)
Sievert, SI unit of effective dose equivalent in man (1 Sv is equivalent to 100 rem
in traditional units)
Thorotrast (232Th Isotope)
United Nations Scientific Committee on the Effects of Atomic Radiation
United States of America - used interchangeably with USA
Working Level Months
White Paper
End of Document
                                         40

-------