Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
AN ABSTRACT OF THE THESIS OF
Kevin A. Makinson for the degree of Master of Science in Radiation Health Physics presented
on July 2, 2009.
Title: Tissue Weighting Factors for Radiation Protection: Derivation and Parametric Analysis.
Abstract approved:
___________________________________________________________________________
David M. Hamby
Tissue weighting factors, wT, used to convert equivalent dose to effective dose,
account for differences in radiosensitivity of various organs in the human body and allow
users to compare individual organ detriment to whole body risk. They are explicitly
calculated in ICRP 26, 60, and 103, although the methods of calculation, as well as the
weighting factors themselves, vary greatly between reports. The ICRP 26 report bases its
weighting factors solely on fatal cancer risk for eleven organs, whereas ICRP 60 and 103 use
detriment (with unique definitions in each report) to calculate the tissue weighting factors
for twenty‐two and twenty‐eight organs, respectively. Each new report introduces levels of
uncertainty into the calculation of wT, which ultimately may reduce the significance of
individual weighting factors. A review of the three calculational methods used in ICRP 26,
60, and 103 is covered, as well as an examination of the uncertainties introduced by each
weighting factor parameter utilized in ICRP 103 and the total uncertainties of the various wT
values. Finally, the effective dose is calculated for several exposure scenarios using the ICRP
26 and 60 methodologies for comparison to ICRP 103 effective dose calculation and
uncertainties thereof.
©Copyright by Kevin A. Makinson
July 2, 2009
All Rights Reserved
Tissue Weighting Factors for Radiation Protection: Derivation and Parametric Analysis
by Kevin A. Makinson
A THESIS
submitted to
Oregon State University
in partial fulfillment of the requirements for the
degree of
Master of Science
Presented July 2, 2009 Commencement June 2010
Master of Science thesis of Kevin A. Makinson presented on July 2, 2009.
APPROVED:
___________________________________________________________________________
Major Professor, representing Radiation Health Physics
___________________________________________________________________________
Head of the Department of Nuclear Engineering and Radiation Health Physics
___________________________________________________________________________
Dean of the Graduate School
I understand that my thesis will become part of the permanent collection of Oregon State University libraries. My signature below authorizes release of my thesis to any reader upon request.
___________________________________________________________________________
Kevin A. Makinson, Author
ACKNOWLEDGEMENTS
I would like to express my sincerest appreciation to my family for their continued
support of my academic, athletic, and musical dreams. I would not be writing this without
their love and support. My advisor, Dr. David Hamby, deserves respect for his guidance on
this project, as well as in life. I would also like to thank the rest of my committee members
for agreeing to support me in this endeavor: Dr. Kathy Higley, my first contact at Oregon
State; Dr. John Edwards, my “psychology advisor,” and fellow musician; and Dr. Paul
Vincent, my Graduate Council Representative.
TABLE OF CONTENTS
Page
1 Chapter 1 – Introduction ................................................................................................... 1
1.1 The International Commission on Radiological Protection ....................................... 1
1.2 Effective Dose Scenario ............................................................................................. 3
1.3 10.CFR.20 Proposed Revisions .................................................................................. 4
2 Chapter 2 – Literature Review ........................................................................................... 6
2.1 Pre ICRP 2 .................................................................................................................. 6
2.2 ICRP 2 ......................................................................................................................... 7
2.2.1 Relevant Terminology Used in ICRP 2 ............................................................... 7
2.2.2 Maximum Permissible Body Burden ................................................................. 8
2.2.3 Maximum Permissible Concentration ............................................................. 10
2.3 Linear Energy Transfer ............................................................................................. 12
2.4 Absorbed Dose ........................................................................................................ 13
2.5 ICRP 26 ..................................................................................................................... 14
2.5.1 Quality Factor .................................................................................................. 14
2.5.2 Tissue Weighting Factor and Detriment .......................................................... 16
2.6 ICRP 60 ..................................................................................................................... 19
2.6.1 Radiation Weighting Factor ............................................................................. 19
2.6.2 Tissue Weighting Factor and Detriment .......................................................... 21
2.7 ICRP 103 ................................................................................................................... 27
2.7.1 Radiation Weighting Factor ............................................................................. 27
2.7.2 Tissue Weighting Factor and Detriment .......................................................... 29
3 Chapter 3 – Methods ....................................................................................................... 35
3.1 ICRP 103 ................................................................................................................... 35
3.2 Uncertainty Analysis ................................................................................................ 36
3.2.1 Cancer Incidence Risk ...................................................................................... 40
3.2.2 Heritable Effects (Gonads)............................................................................... 46
TABLE OF CONTENTS (Continued)
Page
3.2.3 Lethality Fraction (k) .......................................................................................... 49
3.2.4 Minimum Weight for Quality of Life (qmin) ........................................................ 51
3.2.5 Relative Length of Life Lost................................................................................ 52
4 Chapter 4 –Results ............................................................................................................. 56
4.1 Uncertainty Analysis Output ‐ Parameters ................................................................ 56
4.2 Uncertainty Analysis Output ‐ Figures ....................................................................... 56
5 Chapter 5 – Discussion ...................................................................................................... 59
5.1 Skin ............................................................................................................................ 59
5.2 Normalization ............................................................................................................ 62
5.3 ICRP 103 Tissue Weighting Factor Comparison ......................................................... 63
5.4 ICRP 60 Tissue Weighting Factor Comparison ........................................................... 65
5.5 ICRP 26 Tissue Weighting Factor Comparison ........................................................... 66
5.6 Tissue Weighting Factor Comparison Summary........................................................ 68
6 Chapter 6 – Conclusion ...................................................................................................... 70
Bibliography ............................................................................................................................... 73
APPENDICIES .............................................................................................................................. 77
Appendix A – Steps in the development of the tissue weighting system, taken from (ICRP, 2007) ........................................................................................................................... 78
Appendix B – A summary of all input parameters in the uncertainty analysis ..................... 80
Appendix C – Relative detriment plots generated by Crystal Ball ......................................... 84
LIST OF FIGURES Figure Page
2.1 A comparison of Equation (1.17) and the step function described in Table 2.1 .... 21
2.2 Equation (1.24) expressed graphically .................................................................... 29
3.1 An example of a uniform distribution generated by Crystal Ball ............................ 37
3.2 An example of a triangular distribution generated by Crystal Ball ......................... 37
3.3 An example of a normal distribution generated by Crystal Ball ............................. 38
3.4 Normal‐probit plot line fit ....................................................................................... 39
3.5 Log‐probit plot line fit ............................................................................................. 39
3.6 The combined distribution from several uniform distributions used in the calculation of heritable risk ..................................................................................... 49
3.7 Number of years of life lost due to cancer in the esophagus ................................. 53
4.1 Crystal Ball graphical output for the results of the uncertainty analysis on the bone relative detriment .......................................................................................... 57
4.2 Crystal Ball graphical output for the results of the uncertainty analysis on the skin relative detriment ............................................................................................ 58
5.1 Input log‐normal distribution for skin incidence risk with GM of 1000 and GSD of 3 .......................................................................................................................... 60
5.2 Input triangular distribution for skin lethality fraction with bounds of 0 and 1 and a most likely value of 0.002.............................................................................. 61
5.3 Input uniform distribution for skin years of life lost with bounds of 0 and 30 ....... 62
LIST OF FIGURES (Continued) Figure Page
5.4 Box‐plot of the ICRP 103 uncertainty analysis data, with the deterministic tissue weighting factors from the various documents highlighted. The box represents the middle two quartiles of the data.. .................................................. 68
LIST OF TABLES
Table Page
1.1 The inputs of Equation (1.1), the tissue weighting factors, dose, and radiation weighting factors ...................................................................................................... 3
1.2 The outputs of Equation (1.1), the effective dose using weighting factors from ICRP 26, 60 and 103 .................................................................................................. 4
2.1 ICRP 2 RBE values ...................................................................................................... 8
2.2 Unrestricted LET, quality factor relationship .......................................................... 15
2.3 ICRP 26 quality factors ............................................................................................ 16
2.4 The ICRP 26 calculation of radiation weighting factors .......................................... 18
2.5 ICRP 60 radiation weighting factors ........................................................................ 20
2.6 ICRP 60 calculation of lethality fraction .................................................................. 22
2.7 The ICRP 60 calculation of relative life lost ............................................................. 24
2.8 ICRP 60 parameter values used in the calculation of detriment ............................ 25
2.9 ICRP 60 calculation of tissue weighting factors for various organs ........................ 26
2.10 ICRP 103 radiation weighting factors ...................................................................... 28
2.11 A summary of the terms used to define detriment in ICRP 60 and 103 ................. 29
2.12 ICRP 103 detriment parameters ............................................................................. 31
LIST OF TABLES (Continued)
Table Page
2.13 The ICRP 103 calculation of tissue weighting factors ............................................. 32
2.14 A summary of tissue weighting factors in ICRP 26, 60, and 103 ............................. 33
3.1 Relative and absolute LAR from BEIR VII (National Research Council, 2006) ......... 44
3.2 ICRP 103 nominal risks taken as the GM; GSD estimates obtained from BEIR VII confidence intervals ........................................................................................... 45
3.3 Current estimates of genetic risk from continuing exposure to low‐LET, low dose or chronic irradiation (UNSCEAR, 2001) with an assumed doubling dose of 1 Gy. .................................................................................................................... 48
3.4 Risk coefficients (per 10,000) for the reproductive and the total population obtained up to two generations when the population sustains radiation exposure generation after generation. ................................................................... 48
3.5 Lethality fractions and confidence intervals from USCS 2001‐2005 ...................... 50
3.6 Uniform distributions fit to the ICRP 103 reported lethality fraction using the confidence interval percentages from Table 3.5 .................................................... 51
3.7 Triangular distribution parameters defined for two sites ...................................... 51
3.8 Triangular distribution parameters defined for qmin ............................................... 52
3.9 Log‐normal distribution parameters for years of life lost calculation .................... 54
3.10 Non‐log‐normal distributions for years of life lost parameter ............................... 55
LIST OF TABLES (Continued)
Table Page
4.1 Log‐normal distribution parameters for the results of the uncertainty analysis ... 56
5.1 The normalization of the geometric means from the uncertainty analysis ........... 63
5.2 A comparison of the ICRP 103 tissue weighting factors and the uncertainty analysis .................................................................................................................... 64
5.3 A comparison of the ICRP 60 tissue weighting factors and the uncertainty analysis .................................................................................................................... 65
5.4 A comparison of the ICRP 26 tissue weighting factors and the uncertainty analysis .................................................................................................................... 66
5.5 A comparison of the renormalized ICRP 26 tissue weighting factors and the uncertainty analysis ................................................................................................ 67
LIST OF APPENDIX FIGURES
Figure Page
C.1 Crystal Ball graphical output for the results of the uncertainty analysis on the esophagus relative detriment ................................................................................. 84
C.2 Crystal Ball graphical output for the results of the uncertainty analysis on the stomach relative detriment .................................................................................... 84
C.3 Crystal Ball graphical output for the results of the uncertainty analysis on the colon relative detriment ......................................................................................... 85
C.4 Crystal Ball graphical output for the results of the uncertainty analysis on the liver relative detriment ........................................................................................... 85
C.5 Crystal Ball graphical output for the results of the uncertainty analysis on the lung relative detriment ........................................................................................... 86
C.6 Crystal Ball graphical output for the results of the uncertainty analysis on the bone relative detriment .......................................................................................... 86
C.7 Crystal Ball graphical output for the results of the uncertainty analysis on the skin relative detriment ............................................................................................ 87
C.8 Crystal Ball graphical output for the results of the uncertainty analysis on the breast relative detriment ........................................................................................ 87
C.9 Crystal Ball graphical output for the results of the uncertainty analysis on the ovary relative detriment ......................................................................................... 88
C.10 Crystal Ball graphical output for the results of the uncertainty analysis on the bladder relative detriment ...................................................................................... 88
LIST OF APPENDIX FIGURES (Continued)
Figure Page
C.11 Crystal Ball graphical output for the results of the uncertainty analysis on the thyroid relative detriment ...................................................................................... 89
C.12 Crystal Ball graphical output for the results of the uncertainty analysis on the gonads relative detriment ....................................................................................... 90
C.13 Crystal Ball graphical output for the results of the uncertainty analysis on the remainder relative detriment ................................................................................. 90
LIST OF APPENDIX TABLES
Table Page
B.1 Nominal risk input parameters used in the uncertainty analysis ........................... 80
B.2 Lethality fraction input parameters used in the uncertainty analysis .................... 81
B.3 Minimum lethality weight input parameters used in the uncertainty analysis ...... 82
B.4 Relative length of life lost input parameters used in the uncertainty analysis ...... 83
LIST OF ACRONYMS
Acronym Definition
ALARA As Low As Reasonably Achievable
ALI
Annual Limit of Intake
BEIR Biological Effects of Ionizing Radiation
CFR Code of Federal Regulations
DAC Derived Air Concentration
DDREF Dose and Dose Rate Effectiveness Factor
EC Commission of the European Communities (‘European Commission’)
EPA
Environmental Protection Agency
ESU Electrostatic Unit
IAEA International Atomic Energy Agency
ICRP International Commission on Radiological Protection
ICRU International Commission on Radiation Units and Measurements
IEC International Electrotechnical Commission
ILO International Labor Organization
IRPA International Radiation Protection Association
ISO International Standards Organization
IXRPC International X‐Ray and Radium Protection Committee
LET Linear Energy Transfer
LSS Life Span Study
MPBB Maximum Permissible Body Burden
LIST OF ACRONYMS (Continued)
MPCa Maximum Permissible Concentration in air
MPCw Maximum Permissible Concentration in water
MPD Maximum Permissible Dose
NRC Nuclear Regulatory Committee
OECD/NEA
Nuclear Agency of the Commissions of Economic Co‐operation and Development
STP Standard Temperature and Pressure
RBE Relative Biological Effectiveness
REB Roentgen Equivalent Biological
REM Roentgen Equivalent Man
REP Roentgen Equivalent Physical
UNEP United Nations Environment Program
UNSCEAR United Nations Scientific Committee on the Effects of Atomic Radiation
WHO World Health Organization
Tissue Weighting Factors for Radiation Protection: Derivation and Parametric Analysis
1 Chapter 1 – Introduction
Tissue weighting factors are mandated for use in the field of radiation protection to
convert equivalent dose to effective dose. The International Commission on Radiological
Protection (ICRP) is the international body that provides the method of calculation for these
weighting factors and recommends their use. With the recent proposal in the United States
to update 10.CFR.20 to incorporate the newest weighting factor calculations from the ICRP,
it is necessary to address the two objectives of this work:
• to understand the meaning of the tissue weighting factor; and
• given their inherent uncertainties, determine how the weighting factors have
changed over the past 30 years.
1.1 The International Commission on Radiological Protection
The International Commission on Radiological Protection (ICRP) was established in
1928 by the Second International Congress of Radiology. At the time, it bore the name,
“The International X‐Ray and Radium Protection Committee” (IXRPC). It wasn’t until 1950
that the committee reorganized into its present form to more effectively cover the
expanding field of radiation protection (ICRP, 2007).
Although the Commission still has a special relationship with the Congress of
Radiology, it now works closely with its sister organization, the International Commission on
Radiation Units and Measurements (ICRU), and has official relationships with the United
Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), the World
Health Organization (WHO) and the International Atomic Energy Agency (IAEA). It also
Page: 2 works closely with the International Labor Organization (ILO), the United Nations
Environment Program (UNEP), and other United Nation’s bodies. The group collaborates
with the Commission of the European Communities (‘European Commission’, EC), the
Nuclear Agency of the Commissions of Economic Co‐operation and Development
(OECD/NEA), the International Standards Organization (ISO), the International
Electrotechnical Commission (IEC) and the International Radiation Protection Association
(IRPA) (ICRP, 2007).
The primary goal of the recommendations of the ICRP, as stated in ICRP Publication
103 (2007) is to:
“Contribute to an appropriate level of protection for people and the environment against the detrimental effects of radiation exposure without unduly limiting the desirable human actions that may be associated with such exposure” (ICRP, 2007).
The ICRP does not, however, have regulatory authority in the United States, nor any
other country; rather, it provides recommendations agreed upon by the international
community with the intent of influencing national laws. In the United States, these laws can
be found in the Code of Federal Regulations (CFR) and are enforced by governmental
organizations such as the Nuclear Regulatory Commission (NRC) and the Environmental
Protection Agency (EPA).
The ICRP released several documents which are especially pertinent to this work.
ICRP 2 lists permissible radionuclide intake limits. ICRP 26 simplifies this method and is the
first to define tissue weighting factors, determined from fatal cancer risk. ICRP 60 and 103
update the weighting factors by incorporating new parameters into the calculation.
Page: 3 1.2 Effective Dose Scenario
According to US regulation, the effective dose (dose of record) is calculated using the
ICRP 26 tissue weighting factors. If ICRP 60 or 103 weighting factors were used instead, the
reported effective dose could change appreciably. For example, assuming a radiation
worker received an internal dose (see Table 1.1) of 5 rad to the thyroid from electrons, 1 rad
to the bone from alpha particles, 2 rad to the breast from photons, and a whole body dose
of 0.5 rad from photons, we would obtain the effective dose using:
( )( )( ) ( )( )( ), ,,E T thyroid thyroid T bone bone RR e
H w D w w D w α−⎡ ⎤
( )( )( ) ( )( )( ),
, , , , T breast breast R T WB WB Rw D w w D wγ γ
⎡ ⎤= + +⎣ ⎦⎣ ⎦⎡ ⎤ ⎡ ⎤+⎣ ⎦ ⎣ ⎦
(1.1)
where:
• wT is the tissue weighting factor [unitless];
• D is the equivalent dose to a particular organ [rem] and;
• wR is the radiation weighting factor [unitless].
Table 1.1 The inputs of Equation (1.1), the tissue weighting factors, dose, and radiation weighting factors
wT
Site ICRP 26 ICRP 60 ICRP 103 D[rad] wR Thyroid 0.03 0.05 0.04 5 1 Bone 0.03 0.01 0.01 1 20 Breast 0.15 0.05 0.12 2 1 Whole body 1.00 1.00 1.00 0.5 1
The effective dose (Table 1.2) recorded for this individual, depending on whether
ICRP 26, 60, or 103 tissue weighting factors were used, would vary from 1.55 rem, to 1.05
rem, to 1.14 rem, respectively.
Page: 4
Table 1.2 The outputs of Equation (1.1), the effective dose using weighting factors from
ICRP 26, 60 and 103
Effective dose (rem)
Organ ICRP 26 ICRP 60 ICRP 103 Thyroid 0.15 0.25 0.20 Bone 0.60 0.20 0.20 Breast 0.30 0.10 0.24 Whole Body 0.50 0.50 0.50 Total: 1.55 1.05 1.14
Between ICRP 26 and 103, there is about a 35% difference in the reported effective
dose and between ICRP 26 and 60 there is about a 50% difference. As observed from this
simulation, depending on the methodology used, there can be large differences in the
reported effective dose. This could be of particular concern if, for example, a regulation was
exceeded using the ICRP 26 methodology, but not the ICRP 103 methodology. It is therefore
necessary to investigate the uncertainties involved in this calculation.
1.3 10.CFR.20 Proposed Revisions
For several years, there has been discussion in the US regarding the update of
10.CRF.20 to align with the international community. As of April 9, 2009, the NRC had
approved plans to begin revising 10.CFR.20 (Vietti‐Cook, 2009) to be in alignment with ICRP
103. The regulation that will require the most attention is the annual occupational dose to a
radiation worker. The ICRP 103 Committee recommends 2 rem, while the US currently has a
limit of 5 rem. The 5 rem limit was derived based on placing radiation workers at the same
level of risk as the average US industrial worker in the 1950’s. Since then, however, industry
worldwide has attained higher levels of safety, while the dose limit remains the same for
Page: 5 nuclear workers. The 2 rem limit would simply bring the estimated risk from radiation
exposure (using the linear no‐threshold hypothesis) in line with other US industries of today.
Among the many other quantities suggested for update are the tissue weighting
factors. The tissue weighting factors currently used in 10.CFR.20 are specified in the
“definitions” section and are based on the ICRP 26 values. The NRC is currently proposing to
keep the definitions of the weighting factors in the “definitions” section, but to move the
numeric values from that section to an appendix of 10.CFR.20 (Borchardt, 2008). The
possibility of adding an appendix for radiation and tissue weighting factors would allow for
greater ease in updating the document in the future.
As explained in Section 1.2, the effective dose can change appreciably depending on
the methodology used. However, the weighting factors themselves may not be statistically
different from one another. This work will explore the differences in tissue weighting
factors over the years in comparison to an uncertainty analysis of the ICRP 103 data.
Page: 6 2 Chapter 2 – Literature Review
For the past 50 years, the ICRP has released publications addressing a multitude of
topics in the field of radiation protection. The current series began in 1959 with ICRP
Publication 1. Since then, the ICRP has released over 100 publications in the series. For the
sake of simplicity, the titles have been shortened; for example, ICRP Publication 1 is
commonly referred to as ICRP 1. The reports summarized below (chronologically) are ICRP
2, 26, 60, and 103, which all pertain to radiation protection methodology.
2.1 Pre ICRP 2
In 1928, the first report published by the IXRPC focused on the protection of medical
professionals by limiting their work with radioactive sources (IXRPC, 1928). These
recommendations were qualitative in nature and primarily concerned with the avoidance of
threshold doses. At the time, there was no official system of dose quantification; hence no
dose limits could be defined. In 1934, the IXRPC defined a threshold limit of what can be
estimated today as about ten times the current annual occupational dose limit (ICRP, 2007;
IXRPC, 1934).
In the years following, due to epidemiological evidence of excess disease among
American radiologists, and the first reports of excess leukemia in the Japanese atomic bomb
survivors, support for a radiation risk threshold diminished and the ICRP released
supplemental material to their earlier 1951 report (ICRP, 1955). The 1955 report was the
first by the ICRP to officially recognize the possibility of stochastic effects following radiation
exposure (ICRP, 2007). A stochastic (or probabilistic) effect is one where the risk (e.g., of
fatal cancer) appears to have a random outcome, and the probability of which increases
Page: 7 proportionally with dose. Conversely, a non‐stochastic, or deterministic effect, is one in
which there are no discernable effects below a certain dose threshold.
At the time, the ICRP advised that “every effort [should] be made to reduce
exposures to all types of ionizing radiation to the lowest possible level” (ICRP, 1955). This
statement has become the cornerstone of radiation protection and, as the years have
passed, gradually transformed: from “as low as practicable” in the first report (ICRP
Publication 1, (ICRP, 1959), to “as low as readily achievable” (ICRP, 1966), and finally, as it is
known today, “as low as reasonably achievable” (ALARA) (ICRP, 1973).
2.2 ICRP 2
The second publication in the current ICRP series, commonly referred to as ICRP 2,
was also released as the third issue of the Health Physics journal in 1960.
2.2.1 Relevant Terminology Used in ICRP 2
The rem was defined in ICRP 2 as the dose in tissue which results in biological
damage equivalent to that produced per rad of X‐radiation (of about 200 kV) having a linear
energy transfer (LET) (see Section 2.3) in water of 3.5 keV/µm.
Before the creation of quality factors and radiation weighting factors (see sections
2.5.1, 2.5.2, 2.6.1, and 2.6.2), the relative biological effectiveness (RBE) factor was used. To
convert from rad to rem, one would simply multiply the absorbed dose by the RBE value
(Table 2.1). The resulting value would be called the RBE dose, or simply the dose expressed
in rem. To determine the RBE of a particular radiation, one would take the inverse ratio of
the absorbed dose producing the same degree of a defined biological end point as 250 kVp
(peak voltage) x‐rays or cobalt‐60 photons. For example, given that the RBE for cobalt‐60
Page: 8 photons is defined as 1, and 100 rads of radiation from cobalt‐60 photons produced an
increase in cancer mortality equivalent to 10 rads of 5 MeV alpha radiation, the RBE for the
alpha radiation would therefore be 10.
Table 2.1 ICRP 2 RBE values
Radiation RBE
γ, β‐, β+, e‐ 1
Low energy β‐, β+, e‐ (if Emax≤ 0.03 MeV) 1.7
Alpha 10
Recoil Atoms 20
At the time, internal dose limitations were expressed by two main quantities:
“maximum permissible body burden” (MPBB or q), and “maximum permissible
concentration” specified for air and water (MPCa and MPCw, respectively).
2.2.2 Maximum Permissible Body Burden
The maximum permissible body burden (MPBB) was defined by the ICRP 2 Task
Group as the internal activity (in μCi) of a particular radionuclide that results in a maximum
permissible dose (MPD) to the whole body, or to one or more organs in the body (ICRP,
1960). For radionuclides that do not localize in the bone, MPDs of 0.1 rem per week were
used for gonads and total body, 0.6 rem per week for the skin and thyroid, and 0.3 rem for
all other soft tissues. The method of calculating the non‐bone localizing MPBB was as
follows:
Page: 9
[ ][ ]
4 6 52
ergs rad100 g Rg rad week
μCint ergs secs MeV3.7 10 1.6 10 6.05 10 ntμCi MeV week
mq
x x x f ε−
⎡ ⎤ ⎡ ⎤⋅ ⋅⎢ ⎥ ⎢ ⎥⋅ ⎣ ⎦⎣ ⎦=⎡ ⎤ ⎡ ⎤ ⎡ ⎤ ⎡ ⎤⎢ ⎥ ⋅ ⋅ ⋅ ⋅⎢ ⎥ ⎢ ⎥ ⎣ ⎦⎢ ⎥ ⎣ ⎦ ⎣ ⎦⎣ ⎦
(1.2)
or
[ ][ ]3
2
week MeV μCi rad2.8 10 g Rnt rad g week
μCiMeV
nt
x mq
f ε
− ⎡ ⎤⋅ ⋅ ⎡ ⎤⋅⎢ ⎥ ⎢ ⎥⋅ ⋅ ⎣ ⎦⎣ ⎦=⎡ ⎤⋅ ⎣ ⎦
(1.3)
where:
• m is the reference organ mass [g];
• R is the MPD per week limitation [rad/week];
• 2f is the reference organ body burden fraction [unitless]; and
• ε is the effective absorbed energy per disintegration of a given radionuclide
[MeV/nt].
In the case of radionuclides that do localize in the bone, such as 90Sr and 239Pu, the
MPBB was determined from a direct comparison with 0.1 μCi of 226Ra (ICRP, 1960). These
special cases for MPBB were calculated as follows:
[ ] [ ] ( )2
2 2
MeVμCi 0.1 0.99 110ntμCiMeV
nt
RaRa Raq fq
f f
ε
εε
⎡ ⎤⎣ ⎦= ⋅ =⎡ ⎤⎣ ⎦
⋅ (1.4)
or
Page: 10
[ ]2
MeV11 μCi ntμCiMeV
ntq
f ε
⎡ ⎤⎣ ⎦=
⎡ ⎤⎣ ⎦
(1.5)
where:
• = 0.1 [µCi] the MPBB for 226Ra, used for comparison to other radionuclides; Raq
• 2Raf =0.99, the fraction of radium in the body that resides in the skeleton;
• 2f is the fraction of the radionuclide in the body that resides in the skeleton;
• MeV110 ntRaε ⎡= ⎣
⎤⎦ , the effective absorbed energy per disintegration of radium;
• ε is the effective absorbed energy per disintegration of a radionuclide, or
( )MeVntEF RBE n⎡ ⎤ ⋅⎣ ⎦ ⋅∑ ; and
• n is the relative damage factor, which was 1 if the radionuclide was a daughter of
radium, and 5 in all other cases.
The MPBB was the predecessor to the annual limit of intake (ALI), described in ICRP 30
(1982a). The MPBBs for 240 different radionuclides were calculated in ICRP 2 and listed in
over 200 pages of tables. The MPBB provided the health physicist with a look‐up value
utilized for determining internal dose limitations. To date, the US still requires several of
these calculations to be performed by health physicists in nuclear power plants as specified
in 10.CFR.50 Appendix I (Borchardt, 2008).
2.2.3 Maximum Permissible Concentration
The maximum permissible concentration (MPC) was simply the concentration in air
or water that would result in the MPBB, given a certain breathing rate or water
Page: 11 consumption rate. The calculation of the MPC was somewhat more complicated than the
calculation of MPBB.
For example, if the radionuclide considered was the parent of a chain of k daughters
and the stomach was the critical tissue, the corresponding formulas for MPCa and MPCw
were:
( )[ ]
( )
103
3a /24
0 01
0
μCi MeV2.5 10cm g ntμCiMPC cm MeV 1
nt
hik ir
a i j ii hj
h p hpp h
x m g
efλ
ε λλ λ λ
−
−
= ==
=≠
⎡ ⎤⋅ ⋅⎢ ⎥⋅ ⋅⎣ ⎦⎡ ⎤ =⎢ ⎥⎣ ⎦ ⎛ ⎞ −⎡ ⎤ ⋅ ⋅⎜ ⎟⎢ ⎥⎣ ⎦ ⎝ ⎠ −∑ ∑∏
∏
(1.6)
and:
( )[ ]
( )
63
3 /24
0 01
0
μCi MeV2.2 10cm g ntμCiMPC cm MeV 1
nt
hw ik ir
w i j ii hj
h p hpp h
x m g
efλ
ε λλ λ λ
−
−
= ==
=≠
⎡ ⎤⋅⋅⎢ ⎥⋅ ⋅⎣ ⎦⎡ ⎤ =⎢ ⎥⎣ ⎦ ⎛ ⎞ −⎡ ⎤ ⋅ ⋅⎜ ⎟⎢ ⎥⎣ ⎦ ⎝ ⎠ −
∑ ∑∏∏
(1.7)
where:
• m is the reference organ mass [g];
• λj, p, h is the effective decay constant for various radionuclides j, p, h in the decay
chain [s‐1];
• k is the number of daughters in the decay chain [unitless];
• ε i is the effective absorbed energy per disintegration for radionuclide, i [MeV/nt];
• af is the inhalation fraction that is retained in the critical organ [unitless]; and
• wf is the ingestion fraction that is retained in the critical organ [unitless].
Page: 12 A list of all MPC equations used in ICRP 2 was considered beyond the scope of this work, and
thus was omitted. A list of all MPC equations used and complete derivations can be found
in ICRP 2 on pages 16‐39 (ICRP, 1960).
Just as the MPBB was the predecessor to the ALI, MPC was the predecessor to the
derived air concentration (DAC), also defined in ICRP 30 (1982a).
The MPC’s for 240 different radionuclides were calculated in ICRP 2 and listed,
together with MPBB’s, in over 200 pages of tables. The MPC’s provided health physicists
with a look‐up value utilized for the determination of air and water concentrations that
correspond to the MPBB for a given radionuclide.
2.3 Linear Energy Transfer
The restricted LET was defined in ICRU Report 19 (1971) as:
dELdΔ
Δ
⎛ ⎞= ⎜ ⎟⎝ ⎠l
(1.8)
where Δ is a specified energy cutoff in electronvolts and dE is the energy lost by a
charged particle traversing a distance dl in a medium.
The unrestricted linear energy transfer (LET) was defined in ICRU Report 16 (1970)
as:
dELd∞ = l
(1.9)
Δ = ∞ ). or the restricted LET with no energy cutoff (i.e.,
Page: 13
Photons and electrons are considered to be of “low LET” due to their energy
deposition over a relatively long path length. Alpha particles and other heavy charged
particles deposit their energy in a very short distance, and thus are considered to be of “high
LET.”
2.4 Absorbed Dose
Absorbed dose, D, was defined in ICRU Report 19 (1971) as,
dDdmε
= (1.10)
where dε is the mean energy imparted by ionizing radiation to the matter in a
volume element of mass . The special unit, as the ICRP refers to non‐SI units, for
absorbed dose is the rad, where 1 rad was defined as 100 ergs g‐1 or 0.01 J kg‐1. The SI unit,
introduced later, is the Gray (Gy), where 1 Gy was defined as 1 J kg‐1 or 100 rad.
dm
The unit for absorbed dose is the rad, which corresponds to the energy absorption of
ionizing radiation of 100 ergs per gram in any medium. Its predecessor was the rep or
“roentgen equivalent physical.” There is widespread belief that the rad is an acronym for
“radiation absorbed dose.” This seems reasonable because several other radiation
protection units at the time were acronyms (e.g., rem, rep, reb). This is not the case,
however, as was specifically addressed by Dr. Lauriston Taylor, Chairman Emeritus of the
ICRU in a 1990 issue of ICRU News, “The term rad was simply suggested as a word by itself.
Since then it has frequently been improperly referred to as an abbreviation for ‘radiation
absorbed dose.’ This is simply incorrect” (Taylor, 1990).
Page: 14 2.5 ICRP 26
ICRP 26 was published in 1977 as an update to the ICRP 2 radiation dosimetry
methodology. It contained several very useful tools that are still used today in radiation
protection, as specified in 10.CFR.20. Among them were the concepts of detriment, the
quality factor, and the tissue weighting factor.
As is the case in ICRP 26, an informative way to describe radiation dose is through
the use of “detriment.” For effects on health, if pi, the probability of suffering the effect i, is
small and the severity of the effect is expressed by a weighting factor gi, then the detriment
to health, G, in a group of P persons is given by:
i ii
G P p g= ∑ (1.11)
Generally speaking, the detriment to a population is defined as the mathematical
“expectation” of the harm incurred from radiation exposure (ICRP, 1977). The ICRP uses
detriment to quantify the risk of radiation exposure.
2.5.1 Quality Factor
The first measure of detriment introduced in ICRP 26 was the dose equivalent. The
dose equivalent (H) was defined (ICRU, 1970) as:
H DQN= (1.12)
or the absorbed dose weighted by the modifying factors of Q, the quality factor (see Table
2.3), and N, the product of all other modifying factors specified by the Commission. The
Page: 15 value of N could, for example, account for the absorbed dose rate and fractionation effects
on detriment. However, in most cases, N is assigned a value of 1.
The special unit for dose equivalent is the rem, “roentgen equivalent man.” The rem
replaced the reb, “roentgen equivalent biological” resulting from a confusion of speech from
its creator, Herbert M. Parker, in the 1950s (Kathren & Parker, 1986). It was reported that
“during one of his early presentations of the new unit, Parker was suffering from a cold,
which led to difficulty in differentiating it from the rep. Accordingly, the name of the unit
was changed to rem” to avoid future confusion. The rem has since been replaced by the SI
unit, the Sievert (Sv), where 1 Sv is equivalent to 100 rem. According to NIST (2008), the use
of the rem today is “strongly discouraged,” in an attempt to standardize dose units.
Quality factors, or Q‐values, replaced the RBE factors of ICRP 2. Quality factors were
defined as a function of the unrestricted LET, L∞, in water (Table 2.2).
Table 2.2 Unrestricted LET, quality factor relationship
L∞ in water (keV/μm) Q 3.5 (and less) 1 7 2 23 5 53 10 175 (and above) 20
Because the distribution of LET is not always known, an approximation of average
LET per radiation type was recommended by the ICRP (Table 2.3).
Page: 16
Table 2.3 ICRP 26 quality factors
Radiation Quality Factor (Q) X rays, γ rays, and electrons 1 Neutrons, protons, and singly‐charged particles of rest mass greater than one atomic mass unit of unknown energy
10
α particles and multiply‐charged particles (and particles of unknown charge) of unknown energy
20
The quality factor values in Table 2.3 were based on an extrapolation from higher
absorbed doses at which deleterious effects could be directly assessed. Therefore, the Q‐
values were also not necessarily representative of RBE values for other observed effects,
such as stochastic effects at low absorbed doses, and non‐stochastic effects at larger
absorbed doses.
Oftentimes, to estimate the dose equivalent for the entire residence time of a
radionuclide in the human body, the “committed” dose equivalent, H50, is used:
0
0
50
50 ( )t y
tH H t dt
+= ∫ (1.13)
where ( )H t is the relevant dose‐equivalent rate, t0 is the time of intake, and the integration
time is 50 years.
2.5.2 Tissue Weighting Factor and Detriment
An interesting quantity recommended for the regulation of dose‐equivalent limits
was the weighted committed dose equivalent, or the doubly weighted committed absorbed
dose. This limit was defined in ICRP 26 as the product of the dose equivalent and a tissue
weighting factor, wT, summed over all tissues, T, being below the annual dose‐equivalent
limit, Hwb,L, where:
Page: 17
,T T wb LT
w H H≤∑ (1.14)
It is a common misconception that ICRP 26 introduced the quantity of “effective
dose equivalent.” This is not the case. It was not until several years later (1982) that this
quantity made its first appearance in ICRP 33 (ICRP, 1982b). The effective dose equivalent,
HE, was defined as the dose equivalent multiplied by a tissue weighting factor, wT, summed
over all tissues, T, or:
E TT
H w TH= ⋅∑ (1.15)
The unit for effective dose equivalent was the same as dose equivalent, i.e., the Sievert or
rem.
The ICRP 26 methodology for quantifying detriment to a given organ or tissue relies
on the risk of fatal cancer, only. The risk is normalized and rounded to yield a tissue
weighting factor (Table 2.4).
Page: 18
Table 2.4 The ICRP 26 calculation of radiation weighting factors
Organ Risk (fatal cancer per Sv)
% Total wt
Gonads 4.00E‐03 0.242 0.25
Red Bone Marrow 2.00E‐03 0.121 0.12
Bone Surface 5.00E‐04 0.030 0.03
Lung 2.00E‐03 0.121 0.12
Thyroid 5.00E‐04 0.030 0.03
Breast 2.50E‐03 0.152 0.15
Other 5.00E‐03 0.303 0.30
Total: 1.65E‐02 1.000 1.00
The ICRP 26 report was vague about the origin of the values of fatal cancer risk in Table 2.4,
although it is suspected that the values are from BEIR I (National Research Council &
National Academy of Sciences, 1972). The ICRP 30 (ICRP, 1982a) Committee used ICRP 26
methodology to calculate limits for the intake of radionuclides by workers.
The ICRP clearly prescribes the correct (and incorrect) uses of effective dose
equivalent, in that:
• HE is calculated using reference values for a reference person or group (not for an
individual);
• HE is to be used for the planning of prospective situations;
• HE is not to be used for retrospective dose and risk assessments of exposure of
individuals; and
• HE is not to be used for epidemiological studies.
Page: 19 The ICRP states that when it is inappropriate to use effective dose equivalent, an RBE‐
weighted dose, RBExD [Sv or rem] should be used in its place.
2.6 ICRP 60
The ICRP 60 report was published in 1991 as an update to ICRP 26. It contained
several revisions to ICRP 26, including enhancements to the radiation weighting factor and
tissue weighting factor methods.
2.6.1 Radiation Weighting Factor
The ICRP 60 method updated the definition of dose equivalent in two ways. First,
the Commission changed dose equivalent to “equivalent dose”; second, at this time, they
call the quality factor the “radiation weighting factor.” Equivalent dose is defined in ICRP 60
as the absorbed dose multiplied by a radiation weighting factor (Table 2.5), or:
, ,T R R T RH w D= ⋅ (1.16)
where:
• wR is the radiation weighting factor [unitless];
• DT,R is the absorbed dose averaged over a tissue, T, from radiation, R [Gy or rad]; and
• HT,R is the equivalent dose, to tissue T, from radiation, R [Sv or rem].
Page: 20
Table 2.5 ICRP 60 radiation weighting factors
Radiation Energies Radiation Weighting Factors
X‐Rays, γ rays, and electrons All energies 1 Neutrons <10 keV 5 10‐100 keV 10 100 keV‐2 MeV 20 2‐20 MeV 10 >20 MeV 5 Protons <20 MeV 5
Alpha particles, fission fragments, heavy nuclei
20
Due to the significant variation of secondary radiation with neutron energy, the
biological effectiveness of neutrons incident on the human body is strongly dependent on
their energy. For energies below 100 keV, significant fractions of the absorbed dose are
deposited by secondary photons via the H(n, γ) reaction, thereby reducing the biological
effectiveness. The maximum weighting factor occurs between 100 keV and 2 MeV and has a
value of 20. This weighting factor is not an experimentally determined value; rather, it is a
representative value accounting for the broad range of RBE values from animal testing data
using reactor generated neutrons of around 1 MeV. It was thus decided (ICRP, 1991) that a
step function, dependent on energy, provided a more accurate measure of neutron
detriment than the single value used in ICRP 26.
To provide consistency in calculations, a smooth curve was fit (ICRP, 1991) to the
step function for neutrons (Figure 2.1). The mathematical relationship for neutron radiation
weighting factors (as a function of energy in MeV) is:
(1.17) ( )( )2ln 2 /6( ) 5 17 E
Rw E e−= +
Page: 21
0
5
10
15
20
25
1.E‐08 1.E‐06 1.E‐04 1.E‐02 1.E+00 1.E+02 1.E+04
Radiation Weighting
Factor
Neutron Energy (MeV)
Neutron Radiation Weighting Factors: ICRP 60
Smoot Fit
Step Function
Figure 2.1 A comparison of Equation (1.17) and the step function described in Table 2.1
2.6.2 Tissue Weighting Factor and Detriment
As in ICRP 26 (1977), the determination of the tissue weighting factors in ICRP 60 is
based on organ detriment. However, unlike the 1977 definition, detriment to organs or
tissues was defined in ICRP 60 (1991) as the product of fatal cancer risk, severe genetic
effects (for gonads only), the relative non‐fatal contribution, and the relative length of life
lost.
“Fatal cancer risk” is simply defined as the number of fatal cancers per Sievert, per
10,000 persons. These data were obtained primarily from the Japanese atomic bomb
survivor Life Span Study (LSS) cohort, summarized in BEIR V (National Research Council,
1990).
Page: 22
“Severe genetic effects” were defined in ICRP 60 as the incidence probability of
genetic effects due to radiation exposure over all generations per person in the total
population. This was found by the ICRP (1991) to be 0.014 Gy‐1. The ICRP points out that
“because some of the mulifactorial diseases are less detrimental… this probability should
not be added as such without some weighting for the severity of the effects.” The ICRP
admits that “this weighting is necessarily somewhat arbitrary.” And, it was decided
therefore to reduce the risk of genetic effects by a factor of about 1/3, to 0.01 Gy‐1 (ICRP,
1991).
The “relative non‐fatal contribution” to detriment is a function of the probability of
fatal cancers and the lethality fraction for cancer in a given organ. The lethality fraction, k,
was determined from the available lethality data at the time (Table 2.6).
Table 2.6 ICRP 60 calculation of lethality fraction
Organ 5 year Lethality
20 year Lethality
Proposed Lethality, k
Bladder 0.22 0.58 0.50 Breast 0.24 0.62 0.50 Colon 0.45 0.62 0.55 Liver 0.95 0.98 0.95 Lung 0.87 0.96 0.95 Esophagus 0.92 0.97 0.95
Ovary 0.62 0.74 0.70 Stomach 0.85 0.90 0.90 Thyroid 0.06 0.15 0.10
Using the lethality fraction, the non‐fatal contribution was calculated as follows. For
a given tissue, there are F fatal cancers in a given population (i.e., the probability of fatal
cancer). The total probability of cancers, T, in the population, is therefore defined as:
Page: 23
FTk
= (1.18)
The probability of non‐fatal cancers, N, is therefore:
( ) F1N T F kk
= − = − (1.19)
The weighted detriment, D, is then defined (so far) as the sum of the fatal cancer probability
and the lethality fraction multiplied by the probability of non‐fatal cancers:
(1 ) FD F kN F k kk
⎛ ⎞= + = + − ⋅⎜ ⎟⎝ ⎠
(1.20)
or
( )2D F k= − (1.21)
where (2‐k) is now defined as the relative non‐fatal contribution.
The relative length of life lost is a dimensionless ratio of two values, l , defined as
the organ‐specific length of life lost per fatal cancer, and l , the average length of life lost
due to any fatal cancer. The ICRP (1991) defined l as 15 years; this value was obtained by
dividing the expected years of life lost summed over all cancers, by the total number of fatal
cancers. Values of (see l Table 2.7) for bladder, bone marrow, breast, colon, lung,
esophagus, ovary, stomach and remainder were obtained from Land and Sinclair (1991).
Values of for bone surface, liver, skin and thyroid could not be determined in 1991 and l
Page: 24
were arbitrarily set to 15, making the ratio ll equal to unity. The value of l for gonads
was also uncertain and was arbitrarily set to 20 years.
Table 2.7 The ICRP 60 calculation of relative life lost
Organ Life Lost (years): l Relative Life lost: ll
Bladder 9.8 0.65
Bone Marrow 30.9 2.06
Bone Surface 15.0* 1.00
Breast 18.2 1.21
Colon 12.5 0.83
Liver 15.0* 1.00
Lung 13.5 0.90
Esophagus 11.5 0.77
Ovary 16.8 1.12
Skin 15.0* 1.00
Stomach 12.4 0.83
Thyroid 15.0* 1.00
Remainder 13.7 0.91
Gonads 20.0* 1.33
* values were arbitrarily chosen l
( )
After accounting for the relative life lost, the total detriment, D, to a tissue is calculated
using:
2 lD F k l= − (1.22)
Using the data in Tables 2.6 and 2.7, the relative contribution of detriment can be calculated
(Table 2.8).
Page: 25
Table 2.8 ICRP 60 parameter values used in the calculation of detriment
Probability of fatal cancer (10,000
people/Sv)
Severe genetic effects (10,000
people/Sv)
Relative length of life lost
( )ll
Relative non‐fatal
contribution (2‐k)
Detriment (D)
Bladder 30 0.65 1.5 29.25
Bone Marrow 50 2.06 1.01 104.0
Bone Surface 5 1.00 1.30 6.5
Breast 20 1.21 1.50 36.3
Colon 85 0.83 1.45 102.3
Liver 15 1.00 1.05 15.8
Lung 85 0.90 1.05 80.3
Esophagus 30 0.77 1.05 24.3
Ovary 10 1.12 1.30 14.6
Skin 2 1.00 2.00 4.0
Stomach 110 0.83 1.10 100.4
Thyroid 8 1.00 1.90 15.2
Remainder 50 0.91 1.29 58.7
Gonads 100 1.33 133.3
Total Detriment: 724.9
It was determined that there was insufficient data in 1991 to use probability of
cancer incidence for determining detriment. Instead, probability of fatal cancer was used.
The relative organ contribution to total detriment (see Table 2.9) was generated by
determining the fractional contribution of organ detriment to total detriment. Taking into
account the uncertainties involved in the generation of detriment factors, the ICRP decided
that the relative organ contributions could be rounded and grouped into a “simple system of
weights of adequate accuracy” that “would use no more than four groups of weights and
require no more than about a factor of 2 rounding between the relative contributions and
the assigned weight” (ICRP, 1991). The possible values for weights were selected in such a
way as to limit the amount of manipulation involved in forcing the weights to sum to unity.
Page: 26 The possible weighting factors selected were: 0.01, 0.05, 0.12, and 0.20. The tissue
weighting factors (wT) decided upon by ICRP 60 (ICRP, 1991) are summarized in Table 2.9.
Table 2.9 ICRP 60 calculation of tissue weighting factors for various organs
Organ Relative Contribution
wT
(from ICRP 60) Bladder 0.040 0.05
Bone Marrow 0.144 0.12
Bone Surface 0.009 0.01
Breast 0.050 0.05
Colon 0.141 0.12
Liver 0.022 0.05
Lung 0.111 0.12
Esophagus 0.033 0.05
Ovary** 0.020 ‐‐
Skin 0.006 0.01
Stomach 0.139 0.12
Thyroid 0.021 0.05
Remainder* 0.081 0.05
Gonads** 0.184 0.2
Total 1.000 1
*Remainder is composed of 10 tissues: adrenals, brain, upper large intestine, small intestine, kidney, muscle, pancreas, spleen, thymus, and uterus. ** Detriment for gonads and ovaries were summed to yield one wt
The weighting factor for the remainder tissues was divided equally between 10
tissues yielding a wT of 0.005 for each; which is purposefully lower than the lowest wT for
other tissues.
For simplicity sake, effective dose equivalent (HE) was shortened to effective dose (E)
in ICRP 60. According to the Commission, “[effective dose equivalent] is unnecessarily
Page: 27
T TT
E w H
complicated, especially in more complex combinations such as collective committed
effective dose equivalent (ICRP, 1991).” Effective dose was defined as:
= ⋅∑ (1.23)
where wT is the tissue weighting factor and HT is the equivalent dose.
2.7 ICRP 103
The ICRP 103 report was released in 2007 as an update to ICRP 60. The overall
estimates of deterministic and stochastic effects remain fundamentally the same. Even the
ICRP admits, “The overall estimates of cancer risk attributable to radiation exposure have
not changed appreciably in the past 16 years” (ICRP, 2007).
2.7.1 Radiation Weighting Factor
The ICRP 103 radiation weighting factors (Table 2.10) are fairly similar to the ICRP 60
factors (Table 2.5), in that photons and electrons are given a weighting factor of 1. Protons’
and other charged particles’ (not including electrons or muons) weights were reduced from
5 to 2. For neutrons, aside from the use of a continuous function (as opposed to a step
function as was utilized in ICRP 60), the two most significant changes are the decrease of wR
in the low‐energy range (which accounts for the large contribution of secondary photons to
the absorbed dose in the human body), and the decrease of wR at energies above 100 MeV
(ICRP, 2007). It was noted by the ICRP that most neutron doses involve a range of energies
and the use of a continuous function was based on practical considerations and does not
imply greater precision.
Page: 28
Table 2.10 ICRP 103 radiation weighting factors
Radiation Type Radiation weighting factor, wR Photons 1 Electrons and muons 1 Protons and charged particles 2 Alpha particles, fission fragments, heavy ions 20 Neutrons A continuous function of neutron
energy (see Equation (1.24) and Figure 2.2) ranging from 2 to 20
The energy dependent radiation weighting factor for neutrons (see Figure 2.2) can
be found by using the empirically derived, continuous Equation (1.24) as a function of
neutron energy, En:
(1.24)
[ ]
[ ]
[ ]
2
2
2
ln( ) /6
ln(2 ) /6
ln(0.04 ) /6
2.5 18.2 , 1 MeV
( ) 5.0 17.0 , 1 MeV 50 MeV
2.5 3.25 , 50 MeV
n
n
n
En
ER n n
En
e E
w E e E
e E
−
−
−
⎧ + <⎪⎪= + ≤ ≤⎨⎪
+ >⎪⎩
For energies above 1 MeV, very little animal data was available to determine the
biological effectiveness of neutrons; however, the scant data that were available show a
clear decrease in RBE with increasing neutron energy. At energies above 50 MeV, it was
decided (ICRP, 2007) that the weighting factor should asymptotically approach a value close
to that of protons due to their similar behavior at higher energies.
Page: 29
0
5
10
15
20
25
1E‐06 1E‐04 1E‐02 1E+00 1E+02 1E+04Radiation Weighting
Factor
Neutron Energy (MeV‐1)
Neutron Radiation Weighting Factors: ICRP 103
Figure 2.2 Equation (1.24) expressed graphically
2.7.2 Tissue Weighting Factor and Detriment
Although most of the calculation of the ICRP 103 tissue weighting factors will be
covered in this section, Appendix A presents an enlightening step‐by‐step procedure,
written by the ICRP 103 committee, for calculating tissue weighting factors. For the
following calculations, it is useful to reference Table 2.11 for a comparison of terminology
used in ICRP 60 and 103.
Table 2.11 A summary of the terms used to define detriment in ICRP 60 and 103
ICRP 60 Terminology ICRP 103 Terminology Definition F
,F TR Fatal risk
(1 ) FN kk
= − , , ( (1 ))T F TR R k q k= + −NF Nonfatal risk
ll Tl Relative length of life lost
FTk
= IR Nominal risk for fatal and non fatal cancers
Page: 30
, ,( )T F T T NF T TD R q R l
As with ICRP 60, tissue weighting factors are based on detriment to a tissue, T.
Detriment is defined by ICRP 103 as:
= +
min min(1 )Tq q q k
(1.25)
where qT is based on lethality and subjective quality of life accounting for pain and suffering
due to cancer in a tissue, T, and is defined as:
= + − ⋅ (1.26)
The value of qmin is defined as the minimum weight for non‐lethal cancers and attempts to
adjust for pain and suffering for different cancer types. The values assigned to qmin were 0
for skin, 0.2 for thyroid and 0.1 for all other organs. In most cases the weighted detriment
was not terribly sensitive to the value chosen (ICRP, 2007). The qmin adjustment primarily
has an impact on cancers in tissues that have a low lethality rate, such as the skin.
The ICRP assigned a qmin value of 0 to skin cancer because, “radiogenic skin cancer is
almost exclusively of basal cell type which is usually associated with very little pain, suffering
or treatment sequelae” (ICRP, 2007). They assigned a qmin value of 0.2 to thyroid cancer
because they believed that cancer of the thyroid was significantly more painful than other
cancers, although they never specified the origin of this belief. There was little conclusive
evidence to support different qmin values for other organs, therefore the rest of the organs
received a qmin value of 0.1.
Detriment estimates were based on cancer incidence in ICRP 103, whereas cancer
mortality was the basis for tissue weighting factors in ICRP 60. Since that time, however,
Page: 31
T
there has been more progress made on the data analysis of the Life Span Study (ICRP, 2007),
such that the ICRP felt comfortable using incidence data to determine detriment.
Table 2.12 ICRP 103 detriment parameters
Site Nominal risk based on
incidence(RI)(per 10,000)
Lethality (k)
Non‐fatal weight (qT)
Relative life lost ( l )
Absolute Detriment
(DT)
Esophagus 15.1 0.93 0.935 0.87 13.1 Stomach 79.1 0.83 0.846 0.88 67.8 Colon 65.4 0.48 0.530 0.97 48.0 Liver 30.3 0.95 0.959 0.88 26.6 Lung 114.2 0.89 0.901 0.80 90.4 Bone 7 0.45 0.505 1.00 5.1 Skin 1000 0.002 0.002 1.00 4.0 Breast 112.1 0.29 0.365 1.29 79.0 Ovary 10.6 0.57 0.609 1.12 9.9 Bladder 43.4 0.29 0.357 0.71 16.8 Thyroid 32.5 0.07 0.253 1.29 12.9 Bone Marrow 41.9 0.67 0.702 1.63 61.6 Other Solid* 143.8 0.49 0.541 1.03 113.4 Gonads 20 0.80 0.820 1.32 25.4
Total: 574.1
* Other Solid is a grouping of 14 tissues: Adrenals, Extrathoracic (ET) region, Gall bladder, Heart, Kidneys, Lymphatic nodes, Muscle, Oral mucosa, Pancreas, Prostate, Small intestine, Spleen, Thymus, and Uterus/cervix.
Utilizing the values presented in Table 2.12 and Equation (1.25), the relative
detriment was calculated (Table 2.13). Similar to ICRP 26 and 60, the possible values for
weights were selected in such a way as to limit the amount of manipulation involved in
forcing the weights to sum to unity. The possible weighting factors selected were: 0.01,
0.04, 0.08, and 0.12.
Page: 32
Table 2.13 The ICRP 103 calculation of tissue weighting factors
Site Absolute Detriment
(DT)
Relative Detriment
wT
Esophagus 13.1 0.023 0.04 Stomach 67.8 0.118 0.12 Colon 48.0 0.084 0.12 Liver 26.6 0.046 0.04 Lung 90.4 0.157 0.12 Bone 5.1 0.009 0.01 Skin 4.0 0.007 0.01 Breast 79.0 0.138 0.12 Ovary* 9.9 0.017 ‐‐ Bladder 16.8 0.029 0.04 Thyroid 12.9 0.022 0.04 Brain ‐‐ ‐‐ 0.01 Salivary glands ‐‐ ‐‐ 0.01 Bone Marrow 61.6 0.107 0.12 Remainder 113.4 0.198 0.12 Gonads* 25.4 0.044 0.08 Total: 574.1 1.000 1.00
*Gonad and ovary data were summed to yield one weighing factor for the combined organs
The wT for the reminder tissues was divided equally between 14 tissues yielding a wT
of 0.0086 for each; which is purposefully lower than the lowest wT for other tissues. While
not specifically quantifiable, cancer risk to the salivary glands and the brain are thought to
be greater than that of the other tissues in the remainder category, and thus were ascribed
a value of 0.1 (ICRP, 2007).
In summary, Table 2.14 is provided to track the evolution of tissue weighting factors
over 30 years from 1977 (ICRP 26) to 2007 (ICRP 103).
Page: 33
Table 2.14 A summary of tissue weighting factors in ICRP 26, 60, and 103
Organ ICRP 26 ICRP 60 ICRP 103
Gonads 0.25 0.2 0.08 Bone marrow 0.12 0.12 0.12 Lung 0.12 0.12 0.12 Breast 0.15 0.05 0.12 Thyroid 0.03 0.05 0.04 Bone Surfaces 0.03 0.01 0.01 Remainder 0.3 0.05 0.12 Colon 0.12 0.12 Stomach 0.12 0.12 Bladder 0.05 0.04 Liver 0.05 0.04 Esophagus 0.05 0.04 Skin 0.01 0.01 Salivary glands 0.01 Brain 0.01
The weighting factors stayed fairly consistent between the publications with a few
exceptions. The weighting factor for gonads changed significantly from 0.20 in ICRP 60 to
0.08 in ICRP 103. This primarily is due to genetic risk being estimated for only two
generations into the future. The ICRP (2007) points out that, “the equilibrium value used in
Publication 60 is judged to be of limited scientific validity because of the unsupported
assumptions necessary on selection coefficients, mutation component, and population
changes over hundreds of years“ (ICRP, 2007).
The weighting factor for breast dropped by a factor of three (0.15 to 0.05) between
ICRP 26 and 60. This is most likely due to advances in cancer therapy resulting in a drop in
the lethality of breast cancer between 1977 and 1991. The weighting factor, however,
Page: 34 jumped up to 0.12 in ICRP 103. This is due to ICRP 103’s method of using cancer incidence
risk instead of cancer mortality risk in the detriment calculation.
The bone surface weighting factor also dropped by a factor of three (0.03 to 0.01)
between ICRP 26 and 60. This is due to the publication of BEIR IV (1988) on the effects of
radium and other internally deposited alpha emitters. The ICRP 60 and 103 reports both use
the data from BEIR IV in their calculation of detriment to bone surface; ICRP 26 does not.
The deterministic values of the tissue weighting factors are difficult to compare
unless an estimate of uncertainty is determined. Chapter 3 explains the process of
determining the uncertainty of the ICRP 103 relative detriment calculation; Chapter 4
presents the results of that analysis, while Chapter 5 compares the tissue weighting factors
from each publication to the uncertainty analysis.
Page: 35
3 Chapter 3 – Methods
In order to compare the tissue weighting factors of ICRP 103 to the weighting
factors of ICRP 26 and 60, and to determine the significance of any changes, one first needs
to calculate the uncertainties in the newly developed ICRP 103 relative detriment
calculation. The method of assessing uncertainty is similar to that used by Hamby (1993), in
an estimation of atmospheric tritium dose. The goal of this uncertainty analysis is to
quantify the uncertainties in the ICRP 103 relative detriment calculations. Crystal Ball
uncertainty and sensitivity analysis software* was utilized (as a plug‐in to Microsoft Excel)
for this task.
3.1 ICRP 103
The report of ICRP 103 estimates relative detriment due to cancer in a specific organ
from three primary factors: nominal risk based on incidence, a non‐fatal weighting factor
(including the probability of lethality), and relative life lost due to radiogenic cancer death.
Each factor carries an associated uncertainty, distributions of which have been captured
herein. These probability distributions can be deduced, given the availability of data
appropriate for the estimate. If the original data set is not available, or if parameter values
were chosen arbitrarily by the Commission, one can use similar data to determine trends,
functional forms, and variances of the known distribution. For example, if the ICRP used US
cancer incidence data from 2005, but one only has access to US cancer incidence data from
2002, an appropriate assumption would be that the distributions of the 2002 and 2005 data
are sufficiently similar to use one distribution in estimating the other. It therefore can also
be assumed that the variance of the 2002 data is reasonably similar to the variance of the
* By Oracle; online at: http://www.oracle.com/crystalball/
Page: 36 2005 data such that the same geometric standard deviation (GSD) (or some other
multiplicative expression of variance, depending on the distribution) can be applied to the
2005 data.
To carry the example further, if the 2002 cancer incidence of a particular organ was
fit to a log‐normal distribution with a GSD of 1.5, a geometric mean (GM) of 20 (per 10,000),
and the ICRP reported a cancer incidence of 15 (without sufficient information to locate the
original data), one would take 15 as the new GM of a lognormal distribution and apply a
GSD of 1.5. Applying this method and using original data sets where available, probability
distributions for each parameter were generated for the three primary factors in the ICRP
103 detriment calculation to determine the final uncertainties in the relative organ
detriment.
3.2 Uncertainty Analysis
Each parameter in an uncertainty analysis is defined by a probability distribution of
possible values. The choice of distribution is based on one’s knowledge of the parameter
and the values appropriate for the calculation. For example, if the variable of interest is a
fraction, but nothing else is known, a uniform distribution (Figure 3.1) would be fit with
bounds of 0 and 1.
Page: 37
Figure 3.1 An example of a uniform distribution generated by Crystal Ball
If the most likely value of the variable on interest is 0.5, a more informative,
triangular distribution (Figure 3.2), could be assigned.
Figure 3.2 An example of a triangular distribution generated by Crystal Ball
If additional information is available to indicate that the data are normally
distributed, with a mean of 0.5 and a standard deviation of 0.2, a normal distribution (Figure
3.3) would be assigned.
Page: 38
Figure 3.3 An example of a normal distribution generated by Crystal Ball
Unfortunately, not all of the ICRP 103 original data were available to determine
information of distribution shape. If it were, standard statistical tests could be implemented
to determine an appropriate distribution and a measure of goodness‐of‐fit.
Oftentimes, confidence intervals are reported alongside deterministic parameter
estimates. This information can be utilized to determine how data are distributed. For
example, in order to determine if data are normally distributed, one can plot the mean and
confidence intervals on a normal‐probability (often referred to as normal‐probit) plot
(Figure 3.4). If a straight line can be fit to the points, then the data are normally distributed.
One could then use the fit line to determine the standard deviation as the average distance
from the mean at the 16th and 84th percent.
Page: 39
Figure 3.4 Normal‐probit plot line fit
If the plot does not form a straight line on a normal‐probit plot, another functional
fit must be determined. To determine if data are log‐normally distributed, the median and
confidence intervals can be plotted on a log‐probit plot (Figure 3.5); if the points form a
straight line, the data are log‐normally distributed. If the data do not fit a normal or a log‐
normal distribution, other methods must be employed to determine an appropriate
probability distribution to describe the data.
Figure 3.5 Log‐probit plot line fit
Page: 40
To determine a geometric standard deviation (GSD) from a log‐probit plot, the
multiplicative difference between the 84th and 16th percentiles (represented by F(84%) and
F(16%), respectively) and the geometric mean are averaged, expressed below:
(84%)
2 2 (16%)F GM GSD
GM F+ =
⋅ ⋅ (2.1)
Oftentimes, 95% confidence intervals are given. In that case, the GSD can be determined as
follows. Given:
(2.2) 1.96 (97.5%)GM GSD F⋅ =
1.96 (2.5%)GM FGSD
= (2.3)
Rearrangement of the above equations provides the GSD from knowledge of the median
and confidence interval, expressed as:
1.96 1.961 (97.5%)2 (2.5%)
F GGSDGM F
M⎡ ⎤⎛ ⎞⎛ ⎞= +⎢ ⎥⎜ ⎟⎜ ⎟⎝ ⎠⎢ ⎥⎝ ⎠⎣ ⎦
(2.4)
3.2.1 Cancer Incidence Risk
The nominal organ risk coefficients utilized in ICRP 103 were taken from BEIR VII
(2006). BEIR VII is a compilation of several studies (including the Life Span Study), and
represents the most up to date, detailed analysis of the biological effects of radiation.
Instead of quantifying risk by cancer mortality, ICRP 103 uses a measure of incidence
rate, defined by BEIR VII (2006) as “new cases of cancer occurring among previously
Page: 41 unaffected individuals.” The ICRP 103 report uses BEIR VII’s measure of the risk from the
incidence, or the “probability that an individual develops cancer over a specified interval of
time, given that the individual is alive and disease free at the start of the time period
(National Research Council, 2006).”
To accurately measure the increase in cancer risk from radiation exposure, one must
account for the base rate of a particular cancer. If we allow λE(t) and λU(t) to describe the
incidence rates of the exposed (E) and unexposed (U) groups, respectively, and if disease
occurrence is unrelated to exposure, it is expected that λE(t)=λU(t). A lack of equality
indicates an association between cancer occurrence and radiation exposure.
BEIR VII uses two measures of this discrepancy between incidence rates, excess
absolute risk (EAR) and excess relative risk (ERR). The first measure, EAR, is defined as the
difference between the incidence rates of the exposed and unexposed groups, or:
( ) ( ) ( )E UEAR t t tλ λ= − (2.5)
Excess absolute risk describes the additive increase in incidence rate associated with
exposure. For example, if the EAR is constant (EAR=a), then the effect of exposure is to
increase the incidence rate by the constant amount, a, for all time periods.
The second common measure of discrepancy is the relative risk (RR), defined as the
quotient of incidence rates between exposed and unexposed groups, or:
( )( )( )
E
U
tRR tt
λλ
= (2.6)
Page: 42 The relative risk describes the multiplicative increase in incidence rate associated with
exposure. When RR is constant (RR=r) the effect of exposure is to alter incidence rate by a
factor of r.
The ERR is often a more useful quantity than the RR and is defined as the percent
increase of cancer risk from exposed populations, or:
( ) ( ) 1ERR t RR t= − (2.7)
For example, if the relative risk from a particular cancer was 1.3, the ERR(t) would indicate a
30% increase in risk from exposure.
To model risk from radiation exposure, BEIR VII uses lifetime attributable risk (LAR),
which is an approximation of the risk of exposure‐induced death (REID), the measure used
by UNSCEAR (2000) to estimate the probability that an individual will die from (or develop)
cancer associated with the exposure. The LAR for a person receiving a dose D at age e is
expressed as:
100 ( )( , ) ( , , )
( )a
S aLAR D e M D e aS e
=∑ (2.8)
where:
• M(D, e, a) is the EAR or ERR depending on whether the LAR is expressed as absolute
or relative risk;
• a denotes the attained age in years, or e+L;
• L is the risk‐free latent period (L=5 for solid cancers and L=2 for leukemia); and
Page: 43
• ( )( )
S aS e
is the probability of surviving to age a conditional on survival to age e.
Unfortunately, the LAR cannot be observed directly from the LSS data set. Because
the atomic bomb survivors were exposed to high doses, and BEIR VII presents the risk from
low‐dose exposures, the risk is extrapolated using the linear no threshold (LNT) model of
radiation risk (except for leukemia, which uses a linear‐quadratic model). The LNT model
does not account for the effects of dose rate (although the linear‐quadratic model does). It
is known that higher dose rates are more harmful than low dose rates due to the body’s
ability to repair itself over time. It was calculated by BEIR VII that a dose and dose rate
effectiveness factor (DDREF) of 1.5 should be used to account for this discrepancy (see
Annex 10B, BEIR VII for the complete derivation). The ICRP (2007) used this model, but
rounded their DDREF to 2 in an attempt to communicate that the DDREF is a “board
judgment which embodies elements of both subjectivity and probabilistic uncertainty.”
Table 3.1 appears in BEIR VII as a summary of the LAR based on relative and
absolute risk transport.
Page: 44
Table 3.1 Relative and absolute LAR from BEIR VII (National Research Council, 2006)
Males Females Cancer Site LAR based
on Relative Risk
Transport
LAR based on
Absolute risk
Transport
Combined and
Adjusted by DDREF (subjective 95% CI)
LAR based on
Relative Risk
Transport
LAR based on
Absolute risk
Transport
Combined and
Adjusted by DDREF (subjective 95% CI)
Incidence Stomach 25 280 34
(3, 350) 32 330 43
(5, 390) Colon 260 180 160
(66, 360) 160 110 96
(34, 270) Liver 23 150 27
(4, 180) 9 85 12
(1, 130) Lung 250 190 140
(50, 380) 740 370 300
(120, 780) Breast ‐‐ ‐‐ 510
(not used) 460 310
(160, 610) Prostate 190 6 44
(0, 1860) ‐‐ ‐‐
Ovary 66 47 40 (9, 170)
Bladder 160 120 98 (29, 330)
160 100 94 (30, 290)
Other 470 350 290 (120, 680)
490 320 290 (120, 680)
Thyroid 32 No model 21 (5, 90)
160 No model 100 (25, 440)
Leukemia ‐‐ ‐‐ 800* (400, 1590)
‐‐ ‐‐ 1310* (690, 2490)
*No DDREF adjustment was made for leukemia
The absolute and relative LAR were log‐averaged and weighted by EAR and ERR,
respectively, to yield the combined and adjusted (by DDREF) risk values, which reflects the
excess cancer risk across all populations. The BEIR VII committee weighted EAR:ERR as
0.3:0.7 for all organs except for thyroid and lung, which used 0.0:1.0 and 0.7:0.3,
respectively. The ICRP 103 committee reasoned that “only for lung, breast, and thyroid was
it considered that there was sufficient information to justify a representative [EAR:ERR ratio]
Page: 45 other than 0.5:0.5.” The ICRP used a 1.0:0.0 EAR:ERR ratio for breast and bone marrow, a
0.0:1.0 ratio for thyroid and skin, and a 0.7:0.3 ratio for lung.
By plotting the confidence intervals on log‐probit plots, it was observed that all
averages (except bone) corresponded to a log‐normal distribution. The GSD’s were
obtained from these distributions (Table 3.2) and applied to the ICRP 103 estimates with the
same GSD and a GM of the reported ICRP 103 nominal risk value.
Table 3.2 ICRP 103 nominal risks taken as the GM; GSD estimates obtained from BEIR VII confidence intervals
Organ Distribution Type GM GSD Stomach Log‐Normal 79.1 2.40 Colon Log‐Normal 65.4 1.50 Liver Log‐Normal 30.3 2.20 Lung Log‐Normal 114.2 1.40 Breast Log‐Normal 112.1 1.41 Ovary Log‐Normal 10.6 1.41 Bladder Log‐Normal 43.4 2.35 Thyroid Log‐Normal 32.5 2.89 Bone Marrow* Log‐Normal 41.9 2.35 Bone** Normal 7.0 1.80 Skin Log‐normal 1000 3 Esophagus Log‐normal 15.1 3 Remainder Log‐Normal 143.8 1.85
*Leukemia risk from BEIR VII was used ** Normal mean and standard deviation are reported
The ICRP 103 report points out that “nominal risk estimates for bone and skin are
those used in Publication 60 (1991).” In accordance with ICRP 60, bone risk was taken from
BEIR IV (National Research Council, 1988), reported as 133 ± 36 (bone sarcomas/106 person‐
rad). The uncertainty of the bone risk was assumed to be normally distributed based on
BEIR IV’s presentation of an additive standard error. This assumed standard deviation was
then changed to a percentage of the mean (133 ± 27%) and applied to ICRP 103’s value of
Page: 46 the mean (7 per 10,000), and a normal distribution was assumed with a standard deviation
of 1.8 (i.e., 27% of 7).
Skin distribution estimates proved to be more difficult. The ICRP (2007) references
ICRP 60, which references ICRP 59 (1992), which references a compilation of several studies
on skin cancer from the 1970’s but reports no uncertainty. To fit a distribution by extension,
it was assumed the skin was most likely log‐normally distributed based on all other organ
incidence risks (except bone) being log‐normally distributed. The largest GSD complied for
any organ risk was just under 3, so a conservative estimate of a GM of 1000 and a GSD of 3
was assigned to skin cancer incidence risk.
Esophagus risk was also not reported in BEIR VII. The same method was applied
from above and a GM of 15.1 and a GSD of 3 was applied to the esophagus incidence risk.
3.2.2 Heritable Effects (Gonads)
The three primary disease classes that the ICRP uses to determine the heritable
effects of radiation are Mendelian diseases, chronic diseases and congenital abnormalities.
The risk per unit dose is calculated for each disease class and summed to obtain the total
risk of heritable effects per unit radiation dose.
To calculate the risk per unit dose for a particular disease class, Equation (2.9) is
utilized (ICRP, 2007):
1Risk per unit dose = P MC PCRF
DD⋅ ⋅ ⋅ (2.9)
where:
• P is the baseline frequency of the genetic disease class;
Page: 47
• DD is the doubling dose, or radiation dose required to produce as many mutations
as those that arise spontaneously in one generation;
• MC is the mutation component, or a relative measure of the relationship between
change in mutation rate and increase in disease frequency, or (ΔP/P)/(Δm/m) where
m is the spontaneous mutation rate; and
• PCRF is the potential recoverability correction factor for different classes of
mutation. PRCF allows for differing degrees of recoverability of mutations in live
births, i.e., the fraction of mutations that is compatible with embryonic/fetal
development.
It was decided that an in‐depth analysis of the uncertainties involved in the genetics of
the above variables is beyond the scope of this work. For more information see Appendix
A.6 of ICRP 103 (2007), pages 217‐241. For the purpose of this paper, the uncertainties
reported in ICRP 103 were deemed sufficient. Genetic risk (see Table 3.3) from continuing
exposure to low‐LET, low dose radiation is reported in the 2001 report by the United
Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR). The ICRP uses
the genetic risk per Gy to the second generation in their calculation of gonadal risk (Table
3.3).
Page: 48 Table 3.3 Current estimates of genetic risk from continuing exposure to low‐LET, low dose or
chronic irradiation (UNSCEAR, 2001) with an assumed doubling dose of 1 Gy. Disease Class: Baseline Frequency
(per 10,000 live births):Risk per Gy
(per 10,000 progeny): 1st Generation 2nd Generation Mendelian Autosomal dominant and
X‐linked 165 ~7.5 to ~15 ~13 to ~25
Autosomal Recessive 75 0 0 Multifactorial Chronic 6,500 ~2.5 to ~12 ~2.5 to ~12 Congenital abnormalities 600 ~20 ~24 to ~30
The above estimates apply only to the reproductive population. The estimate of
genetic risk to an entire population will be lower because the “genetic damage sustained by
germ cells of individuals who are beyond the reproductive period, or who are not
procreating for any reason, poses no genetic risk (ICRP, 2007).” Given that the average life
expectancy is about 75 years, and the mean reproductive age is 30 years, the dose received
by a person at 30 years of age should be 40% (30/75) of the total dose (ICRP, 2007).
Table 3.4 Risk coefficients (per 10,000) for the reproductive and the total population obtained up to two generations when the population sustains radiation exposure generation
after generation.
Disease Class Reproductive Population Total Population Range Average Average
Mendelian Diseases 13 to 25 19 8 Chronic Disease 3 to 12 8 3 Congenital Abnormalities
24 to 30 27 11
Total 54 22
The ranges given in Table 3.4 were multiplied by 40%, to account for the differences
in total population and reproductive population, and used as the bounds for uniform
Page: 49 distributions. These distributions were then averaged to yield the combined distribution
(Figure 3.6).
Figure 3.6 The combined distribution from several uniform distributions used in the calculation of heritable risk
Although the distribution appears normal, the beta distribution was the best fit with
a mean of 21.4 and variance of 3.46. The beta distribution was not used in the final
uncertainty calculation, as it is only an approximation of the average of three uniform
distributions, which were utilized instead.
3.2.3 Lethality Fraction (k)
The ICRP 103 Committee states that it used “national cancer statistics” to determine
the lethality fraction of various cancers. Because the exact data set that ICRP used for its
calculation was not available, the United States Cancer Statistics document (National
Page: 50 Program of Cancer Registries (NPCR), 2001) was used in its place. Utilizing the USCS, the
follow data were obtained:
Table 3.5 Lethality fractions and confidence intervals from USCS 2001‐2005
Site Lethality Fraction
LCI UCI 2*LCI % 2*UCI%
Esophagus 0.88 0.88 0.90 100.0% 102.3% Stomach 0.57 0.58 0.58 101.4% 102.4% Colon excluding Rectum 0.42 0.42 0.42 99.6% 100.4% Liver and I.H. Bile Duct 0.89 0.89 0.89 99.8% 100.0% Lung and Bronchus 0.78 0.78 0.78 100.1% 100.2% Bones and Joints 0.44 0.44 0.50 100.0% 112.5% Male and Female Breast 0.21 0.21 0.21 100.1% 100.4% Ovary 0.69 0.69 0.69 99.7% 100.3% Testis 0.06 0.04 0.05 67.9% 98.2% Urinary Bladder 0.20 0.20 0.20 98.2% 99.5% Thyroid 0.06 0.06 0.06 101.1% 100.0% Leukemias 0.60 0.60 0.60 100.8% 100.5%
The conservative assumption was made that the confidence intervals represented
only one standard deviation. Therefore, twice the distance from the mean (reported
lethality fraction) was used to obtain the upper and lower CI columns. The lower and upper
confidence interval percentages were applied to the lethality fractions from ICRP 103 to fit
uniform distributions to the data set, with occasional numeric manipulation to account for
inconsistencies in the data (Table 3.6).
Page: 51
Table 3.6 Uniform distributions fit to the ICRP 103 reported lethality fraction using the confidence interval percentages from Table 3.5
Site Lethality Fraction Lower Bound Upper Bound
Esophagus 0.93 0.92 0.95 Stomach 0.83 0.82 0.86 Colon 0.48 0.47 0.49 Liver 0.95 0.94 0.96 Lung 0.89 0.88 0.89 Bone 0.45 0.44 0.47 Breast 0.29 0.28 0.30 Ovary 0.57 0.56 0.58 Bladder 0.29 0.28 0.30 Thyroid 0.07 0.06 0.08 Bone Marrow 0.67 0.66 0.68 Gonads 0.80 0.56 0.81
For the statistics that were not available in the USCS archives, a much broader
uncertainty measure was used; triangular distributions were fit with bounds of 0 and 1, and
a most likely value equal to the ICRP 103 reported value (Table 3.7).
Table 3.7 Triangular distribution parameters defined for two sites
Site Lethality Fraction (most likely value)
Lower Bound Upper Bound
Skin 0.002 0 1 Remainder 0.49 0 1
3.2.4 Minimum Weight for Quality of Life (qmin)
Because qmin is a purely subjective quantity, measuring the pain and suffering
brought on by a particular cancer, not much information can be concluded about its
parametric uncertainty. The ICRP selected only three possible values for qmin: 0, 0.1, and
0.2. It appears most reasonable to select triangular distributions with a most likely value
Page: 52 taken as the ICRP 103 reported value (Table 3.8), and lower bound of 0. The upper bound
was set at 0.3 in to give qmin=0.2 a non‐zero likelihood.
Table 3.8 Triangular distribution parameters defined for qmin
Site qmin (likeliest value)
Lower Bound
Upper Bound
Esophagus 0.1 0 0.3 Stomach 0.1 0 0.3 Colon 0.1 0 0.3 Liver 0.1 0 0.3 Lung 0.1 0 0.3 Bone 0.1 0 0.3 Skin 0.0 0 0.3 Breast 0.1 0 0.3 Ovary 0.1 0 0.3 Bladder 0.1 0 0.3 Thyroid 0.2 0 0.3 Bone Marrow 0.1 0 0.3 Gonads 0.1 0 0.3 Remainder 0.1 0 0.3
3.2.5 Relative Length of Life Lost
To determine the relative uncertainty related to the length of life lost due to the
development of a cancer, it was assumed that all death occurs at the average age of death
in the US, 77.5 years of age. Therefore, if death from cancer occurs at age 77.5 or later, 0
years of life is lost as a result of the cancer. Conversely, if death from cancer occurs at age
0, 77.5 years of life is lost. Given this assumption, cancer mortality data by age, presented
in ICRP 103 Tables A.4.10‐A4.17, were used to determine the number of years of life lost
due to a given cancer. Thyroid data were not available in ICRP 103; the USCS 2001‐2005
data were used as a surrogate. These years of life lost were normalized to sum to unity,
yielding the probability of losing a certain number of years of life given that cancer has
already developed. The probabilities were then multiplied by the years of life lost for that
Page: 53 particular probability and plotted versus years of life lost to determine the distribution
shape (see Figure 3.7 as an example).
0.000.200.400.600.801.001.201.401.601.80
0 10 20 30 40 50 60 70 80 90
Years Lost*P
roba
bility (years)
Esophagus Life Lost
Years Lost (years)
Figure 3.7 Number of years of life lost due to cancer in the esophagus
These data were graphed on a log‐probit plot and it was concluded that they were
log‐normally distributed. From this analysis, the data in Table 3.9 were obtained.
Page: 54
Table 3.9 Log‐normal distribution parameters for years of life lost calculation
Site Years of Life lost GSD
Esophagus 13.05 1.71 Stomach 13.20 1.82 Colon 14.55 1.77 Liver 13.20 1.81 Lung 12.00 1.69 Breast 19.35 1.66 Ovary 16.80 1.72 Bladder 10.65 1.76 Thyroid* 19.35 1.76 Bone Marrow 24.45 2.31 Gonads** 19.80 1.72 Average 15.00 1.86
*Thyroid data from USCS 2001‐2005 **Ovary GSD was used because gonads and ovaries are eventually combined in the detriment calculation
The ICRP points out that data were not available for the calculation of years of life
lost for bone and skin. The ICRP (2007) uses the average life lost for both organs as a result.
For this reason, a uniform distribution was fit with bounds of 0 and 30 years to reflect this
assumption (Table 3.10). It is unclear where the “remainder” life lost calculation originated.
However, since the ICRP reports a value of 15.45, it appears that the number is more certain
than the bone and skin life lost values. For this reason, a triangular distribution was fit for
the remainder category with bounds of 0 and 30 years.
Page: 55
Table 3.10 Non‐log‐normal distributions for years of life lost parameter
Site Distribution Likeliest Lower Bound Upper Bound Remainder Triangular 15.45 0 30
Bone Uniform ‐‐ 0 30
Skin Uniform ‐‐ 0 30
A summary of all input parameters used in the uncertainty analysis can be found in
Appendix B.
Page: 56 4 Chapter 4 –Results
Crystal Ball, along with the input parameters in Chapter 3 were used to perform the
ICRP 103 relative detriment calculation. The results of the uncertainty analysis follow.
4.1 Uncertainty Analysis Output ‐ Parameters
The following table is a summary of the deterministic relative uncertainty calculation
by ICRP 103 and the uncertainty analysis described in Chapter 3.
Table 4.1 Log‐normal distribution parameters for the results of the uncertainty analysis
Organ Distribution Mean* Standard Deviation** 2.50% 97.50% Esophagus Log‐normal 0.008 4.38 0.000 0.145 Stomach Log‐normal 0.045 3.37 0.004 0.487 Colon Log‐normal 0.032 3.07 0.004 0.288 Liver Log‐Normal 0.017 3.58 0.001 0.207 Lung Log‐normal 0.06 2.86 0.008 0.471 Bone Log‐normal 0.003 2.83 0.000 0.023 Skin*** ‐‐ 0.349 0.272 0.000 0.893 Breast Log‐normal 0.053 2.83 0.007 0.407 Ovary Log‐normal 0.006 3.08 0.001 0.054 Bladder Log‐normal 0.011 3.7 0.001 0.143 Thyroid Log‐normal 0.007 3.89 0.000 0.100
Bone Marrow Log‐normal 0.04 4.25 0.002 0.682 Gonads Log‐normal 0.017 2.8 0.002 0.128
Remainder Log‐normal 0.062 3.35 0.006 0.663 * For log‐normal distributions, GM is displayed, otherwise arithmetic mean is displayed ** For log‐normal distributions, GSD is displayed, otherwise standard deviation is displayed *** No fit was available for skin data; mean and standard deviation reflect only the spread of data
4.2 Uncertainty Analysis Output ‐ Figures
Appendix C displays uncertainty plots of the relative detriment generated by Crystal
Ball. Each figure (except skin) was fit to a log‐normal distribution with parameters given in
Table 4.1.
Page: 57
Figure 4.1 Crystal Ball graphical output for the results of the uncertainty analysis on the bone relative detriment
The bone relative detriment uncertainty (Figure 4.1) was fit to a log‐normal
distribution with a GM of 0.003 and a GSD of 2.83. The above figure is similar in appearance
and fit to all other (except skin) relative detriment uncertainty plots (see Appendix C).
Page: 58
Figure 4.2 Crystal Ball graphical output for the results of the uncertainty analysis on the skin relative detriment
The skin relative detriment uncertainty (Figure 4.2) could not be successfully fit to any
of the continuous distributions† available in Crystal Ball. The statistical mean and standard
deviation were determined as 0.349 and 0.272, respectfully. These values are suspect due to
the predictability of all other distributions, and the significantly lower deterministic relative
uncertainty of 0.007. See Section 5.1 for more details.
The results presented in Chapter 4 are discussed in Chapter 5 and compared with the
tissue weighting factors in ICRP 26, 60, and 103.
† Available continuous distributions in Crystal Ball: beta, betaPERT, exponential, gamma, logistic, lognormal, max extreme, normal, Pareto, student’s t, triangular, uniform, Weibull
Page: 59 5 Chapter 5 – Discussion
Although it is enlightening to understand the origin of the relative organ detriment
calculation, this quantity is not used for radiation protection purposes; tissue weighting
factors are, however. To transfer from relative detriment to tissue weighting factors, a
significant amount of subjective manipulation is involved. Historically, four, fairly arbitrary
values are chosen and each relative detriment is grouped into these categories. The relative
detriments aren’t necessarily assigned the closest weighting factor, they are decided based
on what is necessary to sum to one, as long as the weighting factor is within “about a factor
of two,” of the relative detriment. This seems reasonable in that the uncertainties (GSD)
determined herein are greater than 2 for each of the 14 organs analyzed.
5.1 Skin
The results of the uncertainty analysis give a skin relative detriment distribution mean
as 0.35 with a standard deviation of 0.27. No distribution fit the shape of the data, which is
suspect, especially when all other organ relative detriments fit well into log‐normal
distributions.
The inputs are responsible for the behavior of the output. The ICRP (2007) reported
an incidence risk of 1000 Sv‐1 (per 10,000) for skin cancers. No uncertainty was located for
these data; a log‐normal distribution was assumed, based on the prevalence of the
distribution for other cancer risks. A conservative estimate for the GSD of 3 was ascribed
due to 3 being larger than all other GSDs. These two assumptions left a relatively large
distribution (Figure 5.1).
Page: 60
Figure 5.1 Input log‐normal distribution for skin incidence risk with GM of 1000 and GSD of 3
In an uncertainty analysis, it may be acceptable if there is significant uncertainty in
one parameter, as long as it is not the most sensitive parameter of the model. The risk of
incidence for skin cancer (Figure 5.1), however, is a very sensitive parameter, spans two
orders of magnitudes, and is considerably larger than incidence risks from other cancers.
Uncertainties for the skin lethality fraction were difficult to determine. Because the
lethality fraction is bounded by 0 and 1, and a likely value was reported, 0.002, a triangular
distribution (Figure 5.2) was fit for the lethality fraction.
Page: 61
Figure 5.2 Input triangular distribution for skin lethality fraction with bounds of 0 and 1 and a most likely value of 0.002
All other lethality fraction uncertainties were input as uniform distributions with
bounds relatively close to the expected value. In this case, the triangular distribution
communicates much more uncertainty through to the relative detriment estimate.
For years of life lost, not even a deterministic value was available for ICRP 103 to
display. The ICRP 103 Committee admitted that there was insufficient data to report
anything other than the average years of life lost for this parameter. Due to this lack of
data, a uniform distribution was assumed with bounds of 0 and 30 years (Figure 5.3).
Page: 62
Figure 5.3 Input uniform distribution for skin years of life lost with bounds of 0 and 30
5.2 Normalization
One of the most interesting aspects of tissue weighting factors is their necessity to
sum to unity, and therefore their dependence on one another. Before the relative
detriments from the uncertainty analysis could be normalized, the results from the skin
needed to be manipulate to preserve the integrity of the data. The results show detriment
to the skin as 0.35, whereas the largest detriment to any other organ is for the stomach
(0.047), nearly an order of magnitude smaller than the detriment calculated for the skin. If
the relative detriments were normalized without addressing this issue, the skin detriment
would dominate the normalization, making all other organs appear less sensitive, and the
skin appear even more sensitive.
One can observe that the skin and bone have about the same deterministic relative
detriments, 0.007 and 0.009, respectively (Table 2.13). Skin and bone were also assigned
Page: 63 the same weighting factor (0.01) in both ICRP 60 and 103, suggesting that they have
relatively similar sensitivities to radiation. The GM for bone from the uncertainty analysis
was 0.003; to follow the trend of the detriments, skin was assigned an unnormalized relative
detriment of 0.002. This prevents skin from dominating the normalization (Table 5.1).
Table 5.1 The normalization of the geometric means from the uncertainty analysis
Organ GM of Relative Detriment Normalized GM
Skin 0.002 0.006 Bone 0.003 0.008 Ovary 0.006 0.017 Thyroid 0.007 0.019 Esophagus 0.008 0.022 Bladder 0.011 0.030 Liver 0.017 0.047 Gonads 0.017 0.047 Colon 0.032 0.088 Bone Marrow 0.04 0.110 Stomach 0.045 0.124 Breast 0.053 0.146 Lung 0.06 0.165 Remainder 0.062 0.171 Sum 0.363 1.000
5.3 ICRP 103 Tissue Weighting Factor Comparison
A comparison of the tissue weighting factors utilized in ICRP 103 and the GM and GSD
results from the uncertainty analysis (Table 5.2) are provided.
Page: 64 Table 5.2 A comparison of the ICRP 103 tissue weighting factors and the uncertainty analysis
Organ ICRP 103 wT
Normalized GM
(Uncertainty Analysis)
GSD (Uncertainty Analysis)
wT Within 1 GSD of GM?
Esophagus 0.04 0.022 4.38 TRUE Stomach 0.12 0.124 3.73 TRUE Colon 0.12 0.088 3.07 TRUE Liver 0.04 0.047 3.58 TRUE Lung 0.12 0.165 2.86 TRUE Bone 0.01 0.008 2.83 TRUE Skin 0.01 ‐‐ ‐‐ Breast 0.12 0.146 2.83 TRUE Bladder 0.04 0.030 3.70 TRUE Thyroid 0.04 0.019 3.89 TRUE Bone Marrow 0.12 0.110 4.25 TRUE Remainder 0.12 0.171 3.35 TRUE Gonads* 0.08 0.063 2.80 TRUE Salivary Glands 0.01 ‐‐ ‐‐ ‐‐ Brain 0.01 ‐‐ ‐‐ ‐‐
*Gonads includes risk to ovaries
It is useful to observe how many geometric standard deviations the weighting
factors lie away from the geometric mean. As observed above, all tissue weighting factors
from ICRP 103 are within one GSD of the normalized median.
The weighting factors for the salivary glands and the brain could not be compared
with the uncertainty analysis because the weighting factors were chosen (ICRP, 2007) based
on the subjective belief that there was slightly more risk to those organs than the ones in
the “remainder” category.
As explained earlier, there was not enough data available to accurately quantify the
uncertainty in the relative detriment to the skin; hence, no comparison could be made to
the tissue weighting factor presented in ICRP 103.
Page: 65 5.4 ICRP 60 Tissue Weighting Factor Comparison
One fundamental question this paper sought to answer was, “Given their inherent
uncertainties, have the tissue weighting factors changed in the last 30 years?” To answer
this question involves comparing the results of the uncertainty analysis with the tissue
weighting factors of ICRP 60 and 26.
Table 5.3, below, compares tissue weighting factors derived in ICRP 60 and the
normalized GM and GSD from the ICRP 103 uncertainty analysis.
Table 5.3 A comparison of the ICRP 60 tissue weighting factors and the uncertainty analysis
Organ ICRP 60 wT
Normalized GM
(Uncertainty Analysis)
GSD (Uncertainty Analysis)
wT Within 1 GSD of GM?
wT Within 2 GSD of GM?
Esophagus 0.05 0.022 4.38 TRUE Stomach 0.12 0.124 3.73 TRUE Colon 0.12 0.088 3.07 TRUE Liver 0.05 0.047 3.58 TRUE Lung 0.12 0.165 2.86 TRUE Bone 0.01 0.008 2.83 TRUE Skin 0.01 ‐‐ ‐‐ Breast 0.05 0.146 2.83 FALSE TRUE Bladder 0.05 0.030 3.70 TRUE Thyroid 0.05 0.019 3.89 TRUE Bone Marrow 0.12 0.110 4.25 TRUE Remainder 0.05 0.171 3.35 FALSE TRUE Gonads* 0.20 0.063 2.80 FALSE TRUE *Gonads reflects gonads and ovaries
All organs are still within two GSDs of the GM. The ICRP 60 weighting factor for
gonads is farther than one GSD from the GM, even though the ICRP 103 weighting factor
was relatively close to the GM. This is reasonable since the ICRP changed the way they
calculated the heritable risk to gonads from the risk to all generations in ICRP 60, to the risk
to only the first two generations in 103.
Page: 66 It is also not surprising that the remainder category is beyond 1 GSD from the
normalized GM. The ICRP 60 remainder category includes fewer tissues than the ICRP 103’s
remainder category. To maintain approximately the same level of sensitivity, the remainder
category had to be given more weight in ICRP 103 to include more organs.
Although skin uncertainty data was not available, ICRP 103 did not update the
method of calculating skin cancer risk. The ICRP 59 methodology was used in ICRP 103 as
well as in ICRP 60; hence there was no change from ICRP 60 to 103.
Brain and salivary were not singled out in ICRP 60. They were expressed as part of
the remainder category. No comparison could be made between the salivary glands and
brain weighting factors due to the purely subjective nature of those particular weighting
factors in ICRP 103.
5.5 ICRP 26 Tissue Weighting Factor Comparison
Table 5.4, below, compares tissue weighting factors derived in ICRP 26 and the GM
and GSD from the ICRP 103 uncertainty analysis.
Table 5.4 A comparison of the ICRP 26 tissue weighting factors and the uncertainty analysis
Organ ICRP 26 wT
Normalized GM
(Uncertainty Analysis)
GSD (Uncertainty Analysis)
wT Within 1 GSD of GM?
wT Within 2 GSD of GM?
Lung 0.12 0.165 2.86 TRUE Bone 0.03 0.008 2.83 FALSE TRUE Breast 0.15 0.146 2.83 TRUE Thyroid 0.03 0.019 3.89 TRUE Bone Marrow 0.12 0.11 4.25 TRUE Remainder 0.30 0.171 3.35 TRUE Gonads* 0.25 0.063 2.80 FALSE TRUE Total: 1.00 0.626 *Gonads reflects gonads and ovaries
Page: 67 As expected, the gonads weighting factor did lie outside of one GSD of the GM, due
to the change in the heritable risk calculation in ICRP 103. It is however, surprising that the
bone surface weighting factor changed significantly (by a factor of 3) from ICRP 26 to 60.
This could perhaps be a reflection of the progress made in internal dosimetry methods for
alpha emitters since 1977. Specifically, BEIR IV (1988) was not available to the ICRP 26
Committee, but was utilized by the 60 and 103 Committees.
It is interesting that the ICRP 26 breast weighting factor is within 1 GSD of the ICRP
103 normalized GM, while the ICRP 60 breast weighting factor lies outside this range. This is
due to the decrease in lethality over the years, but the approximately constant incidence
rate.
Because ICRP calculates risk to only six organs plus a remainder category, the
normalized GM from the ICRP 103 uncertainty analysis can be renormalized to account for
the seven sites defined ICRP 26.
Table 5.5 A comparison of the renormalized ICRP 26 tissue weighting factors and the uncertainty analysis
Organ ICRP 26 wT
Renormalized GM
(Uncertainty Analysis)
GSD (Uncertainty Analysis)
wT Within 1 GSD of GM?
Lung 0.12 0.242 2.86 TRUE Bone 0.03 0.012 2.83 TRUE Breast 0.15 0.214 2.83 TRUE Thyroid 0.03 0.027 3.89 TRUE Bone Marrow 0.12 0.161 4.25 TRUE Remainder 0.30 0.251 3.35 TRUE Gonads* 0.25 0.093 2.80 TRUE Total: 1.00 1.000
*Gonads reflects gonads and ovaries
Page: 68 Once the renormalization is performed (Table 5.5), all statistically significant
differences between ICRP 103 and 26 weighting factors disappear.
5.6 Tissue Weighting Factor Comparison Summary
As concluded from the previous sections, the tissue weighting factors have not
changed beyond two geometric standard deviations of the ICRP 103 uncertainties. There is,
however, noticeable change between the publications (Figure 5.4), as discussed below.
Figure 5.4 Box‐plot of the ICRP 103 uncertainty analysis data, with the deterministic tissue weighting factors from the various documents highlighted. The box represents the middle
two quartiles of the uncertainty analysis data.
Due to the publication of BEIR IV (National Research Council, 1988) the ICRP 26
weighting factor for bone (0.03) is a factor of three different than the ICRP 60 and 103
weighting factors, and appears outside of the inner two quartiles. Because of the different
method of calculation, the ICRP 26 and 60 gonads weighting factor also appears outside of
the inner two quartiles. The remainder category is the least consistent of any site. This is
Page: 69 not surprising due to the different number of organs represented in the remainder category
from each document.
Page: 70 6 Chapter 6 – Conclusion
In the health physics community, it is commonly thought that tissue weighting factors
are intended to be exact measures of radiation risk. This is not the case, as can be
concluded from the results of the previous chapters. The proper way to interpret tissue
weighting factors is that they are an indication of organ risk, relative to other organs. For
example, if organ A had a tissue weighting factor of 0.20 and organ B had a weighting factor
of 0.05, organ A is not four times as radiosensitive as organ B. The proper interpretation
would be to say that organ A is significantly more radiosensitive than organ B, but by an
uncertain amount.
Quantification of radiation risk has come a long way since the first report published
by the IXRPC in 1928, proposing dose limitations to medical workers exposed to radiation.
Since then, the ICRP has published several incredibly influential documents, including ICRP 2
(1960), 26 (1977), 60 (1991), and 103 (2007), which have guided, and continue to guide
radiation protection methodology. It is, however, important that the recommendations be
fully understood before implementing, to avoid using a quantity in a way which was not
intended. The intent of this work, therefore was:
• to understand the meaning of the tissue weighting factor; and
• given their inherent uncertainties, determine how the weighting factors have
changed over the past 30 years.
The history of the tissue weighting factors was explored through the years, tracking
changes to the weighting factors based on the discovery of new information about radiation
risk. The ICRP 103 report is certainly not the last publication by the ICRP to recommend
Page: 71 adjustments to the weighting factors. When new data is obtained suggesting a need to
update the weighting factors, the ICRP will release an update to ICRP 103. Judging by the
relative frequency of the previous documents, this update won’t be at least until 2020. Until
then, the ICRP 103 weighting factors are the most up to date calculation of organ risk
available to the international community.
The ICRP 26 tissue weighting factors were based only on the risk of fatal cancer to six
organs plus a remainder category (five organs). The ICRP 60 Committee used risk of fatal
cancer in its calculation, but also incorporated the lethality (to determine risk from non‐fatal
cancers), genetic risk (only to gonads), and years of life lost from a particular cancer for
twelve organs plus a remainder (ten organs). The ICRP 103 report uses a similar calculation,
but uses cancer incidence instead of mortality, and a factor to account for the pain and
suffering from the development and treatment of a particular cancer for fourteen organs
plus a remainder (fourteen organs).
The ICRP 103 tissue weighting factors are a complex set of numbers that say a lot more
than one might think at first glance. They not only attempt to measure the likelihood of
cancer in a particular organ, but also the lethality, pain and suffering, and the years of life
lost. Each weighting factor also is not necessarily representative of relative detriment. Four
fairly arbitrary factors are selected and organs are assigned weighting factors based the
necessity to sum to unity, which implies that the weighting factors are not independent of
one another. Because the weighting factors measure risk relative to other organs, one
organ’s radiosensitivy influences all other organ’s weighting factors. For example, if organ A
Page: 72 is particularly radiosensitive, it will make organ B appear less radiosensitive relative to organ
A, even though organ B’s absolute radiosensitivity is not dependant on other organs.
In light of the results from the uncertainty analysis above, it is suggested that 10.CFR.20
be updated to reflect a mixture of the ICRP 26 and 103 methodologies. Specifically, the
tissue weighting factors could be calculated using only cancer mortality data (as in ICRP 26)
for the fourteen organs (plus remainder) defined in ICRP 103. This method would
significantly simplify the definition of the tissue weighting factors without changing them
beyond the bounds of uncertainties introduced by the additional complexities of ICRP 103.
It is not surprising that there is dialog regarding the usefulness of updating the radiation
weighting factors. According to this work, other than including more organs, there has been
no appreciable change of the tissue weighting factors in the last 30 years; however, given
the right conditions, there could be a significant difference in the recorded effective dose.
Page: 73 Bibliography
Boetticher, H. V., Lachmund, J., Looe, H. K., Hoffmann, W., & Poppe, B. (2008). 2007 recommendations of the ICRP change basis for estimation of the effective dose; what is the impact on radiation dose assessment of patient and personnel? (In English). Publication: Ro Fo; Fortschritte auf dem Gebiete der Ro ntgenstrahlen und der Nuklearmedizin, 180(5), 391‐395.
Borchardt, R. W. (2008). Options to revise radiation protection regulations and guidance with respect to the 2007 recommendations of the International Commission on Radiological Protection.
Hamby, D. M. (1993). A probabilistic estimation of atmospheric tritium dose. Health Physics, 65(1), 33‐40.
ICRP. (1951). International Recommendations on Radiological Protection. Revised by the International Commission on Radiological Protection and the 6th International Congress of Radiology. Br. J. Radiol. (pp. 46‐53). London.
ICRP. (1955). Recommendations of the International Commission on Radiological Protection.
ICRP. (1959). Recommendations of the International Commission on Radiological Protection. ICRP Publication 1. Oxford, UK: Pergamon Press.
ICRP. (1960). Report of Committee II on Permissible Dose for Internal Radiation (1959). ICRP Publication 2, 3.
ICRP. (1966). Recommendations of the International Commission on Radiological Protection. ICRP Publication 9. Oxford, UK: Pergamon Press.
ICRP. (1973). Recommendations of the International Commission on Radiological Protection. ICRP Publication 22. Oxford, UK: Pergamon Press.
ICRP. (1977). Recommendations of the International Commission on Radiological Protection. Annals of the ICRP Publication 26, 1(3).
ICRP. (1982). ICRP Publication 30: Limits for Intakes of Radionuclides by Workers. Annals of the ICRP Publication 30, 6(1‐3).
ICRP. (1982). Protection Against Ionizing Radiation from External Sources Used in Medicine. Annals of the ICRP Publication 33, 9(1).
ICRP. (1991). 1990 Recommendations of the International Commission on Radiological Protection. ICRP Publication 60, Annals of the ICRP, 21(1‐3).
ICRP. (1992). The biological basis for the dose limitation in the skin. ICRP Publication 59.
ICRP. (2007). The 2007 Recommendations of the International Commission of Radiological Protection. ICRP Publication 103, Annals of the ICRP, 37(2‐4).
Page: 74 ICRP. (2008). Book Review: ICRP Publication 103: Recommendations of the ICRP. Radiation
Protection Dosimetry, 129(4), 500‐507.
ICRU. (1970). Linear Energy Transfer. International Commission on Radiation Units and Measurements, ICRU Report 16.
ICRU. (1971). Radiation Quantities and Units. International Commission on Radiation Units and Measurements, ICRU Report 19.
IXRPC. (1928). X ray and Radium Protection. Recommendations of the 2nd International Congress of Radiology. Br. J. Radiol., 12, 359‐363.
IXRPC. (1934). International Recommendations for X ray and Radium Protection. Revised by the International X ray and Radium Protection Commission and adopted by the 4th International Congress of Radiology. Br. J. Radiol., 7, 1‐5.
Kathren, R., & Parker, H. M. (1986). Publications and Other Contributions to Radiological and Health Physics. Battelle Press. Retrieved April 8, 2009, from http://www.orau.org/PTP/articlesstories/names.htm.
Land, C. E., & Sinclair, W. K. (1991). The relative contributions of different organ sites to the total cancer mortality associated with low‐dose radiation exposure. Annals of the ICRP, 22, 31‐57.
Land, C. E., Hayakawa, N., Machado, S. G., Yamada, Y., Pike, M. C., Akiba, S., et al. (1994). A Case‐Control Interview Study of Breast Cancer among Japanese A‐Bomb Survivors. I. Main Effects. Cancer Causes & Control, 5(2), 157‐165.
Leggett, R. W. (2001). Reliability of the ICRP's dose coefficients for members of the public. 1. Sources of uncertainty in the biokinetic models. Radiation Protection Dosimetry, 93(3), 199‐213.
Lindell, B., Dunster, H. J., & Valentin, J. (1998). International Commission on Radiological Protection: History, Policies, Procedures. Swedish Radiation Protection Institute, SE‐171(16).
Lubin, J. H., Jr., J. D. B., Edling, C., Hornung, R. W., Howe, G. R., Kunz, E., et al. (1995). Lung Cancer in Radon‐Exposed Miners and Estimation of Risk From Indoor Exposure. Journal of the National Cancer Institute, 87, 817‐827.
National Academy of Sciences. (1980). The Effects on Populations of Exposure to Low Levels of Ionizing Radiation. The Biological Effects of Ionizing Radiations. Washington, D.C.: National Research Council.
National Program of Cancer Registries (NPCR). (2001). United States Cancer Statistics (USCS). Retrieved June 10, 2009, from http://apps.nccd.cdc.gov/uscs/.
Page: 75 National Research Council. (1988). Health Risks of Radon and Other Internally Deposited
Alpha‐Emitters. BEIR IV. Biological Effects of Ionizing Radiation. Washington, D.C.: National Research Council of the National Academies.
National Research Council. (1990). Health Risks from Exposure to Low Levels of Ionizing Radiation: BEIR V. Biological Effects of Ionizing Radiation. Washington D.C.: National Academy Press.
National Research Council. (2006). Health Risks from Exposure to Low Levels of Ionizing Radiation: BEIR VII Phase 2. Biological Effects of Ionizing Radiation. Washington D.C.: National Research Council of the National Academies.
National Research Council, & National Academy of Sciences. (1972). The Effects on Populations of Exposure to Low Levels of Ionizing Radiation. Report of the Advisory Committee on the Biological Effects of Ionizing Radiations. BEIR I. Washington, D.C.
Pierce, D. A., Sharp, G. B., & Mabuchi, K. (2003). Joint Effects of Radiation and Smoking on Lung Cancer Risk among Atomic Bomb Survivors. Radiation Research, 159, 511‐520.
Prasad, K. N. (1995). Radiobiology: Handbook of ‐ (2nd ed.). Boca Raton, Florida: CRC.
Preston, D. L., Ron, E., Tokuoka, S., Funamoto, S., Nishi, N., Soda, M., et al. (2007). Solid Cancer Incidence in Atomic Bomb Survivors: 1958–1998. Radiation Research, 168, 1‐64.
Preston, D. L., Mattsson, A., Holmberg, E., Shore, R., Hildreth, N. G., & Jr., J. D. B. (2002). Radiation Effects on Breast Cancer Risk: A Pooled Analysis of Eight Cohorts. Radiation Research, 158, 220‐235.
Preston, D. L., Pierce, D. A., Shimizu, Y., Cullings, H. M., Fujita, S., Funamoto, S., et al. (2004). Effect of Recent Changes in Atomic Bomb Survivor Dosimetry on Mortality Risk Estimates. Radiation Research, 162, 377‐389.
Preston, D. L., Shimizu, Y., Pierce, D. A., Suyama, A., & Mabuchi, K. (2003). Studies of Mortality of Atomic Bomb Survivors. Report 13: Solid Cancer and Noncancer Disease Mortality: 1950–1997. Radiation Research, 160, 381‐407.
Russ, S. (1918). A suggestion for a new X‐ray unit in radiotherapy. Archives of Radiology and Electrotherapy, (23), 226‐32.
Taylor, L. (1990). 80 Years of Quantities and Unites ‐ Personal Reminiscences. ICRU News.
Travis, L. B., Gospodarowicz, M., Curtis, R. E., Clarke, E. A., Anderson, M., Glimelius, B., et al. (2002). Lung Cancer Following Chemotherapy and Radiotherapy for Hodgkin's Disease. Journal of the National Cancer Institute, 94, 182‐192.
UNSCEAR. (2000). Sources and Effects of Ionizing Radiation. UNSCEAR Report to the General Assembly, Volume II: Effects. New York: United Nations.
Page: 76 UNSCEAR. (2001). United Nations Scientific Committee on the Effects of Atomic Radiation.
Hereditary Effects of Radiation. The 2002 Report to the General Assembly with Scientific Annex. New York: United Nations. Retrieved June 12, 2009, from http://www.unscear.org/docs/reports/2001/2001Annex_pages%208‐160.pdf.
Vietti‐Cook, A. L. (2009). Staff Requirements ‐ SECY‐08‐0197 ‐ Options to Revise Radiation Protection Regulations and Guidance With Respect to the 2007 Recommendations of the International Commission on Radiological Protection. NRC.
Page: 77
APPENDICIES
Page: 78 Appendix A – Steps in the development of the tissue weighting system, taken from (ICRP,
2007)
ICRP 103 listed a very detailed procedure that was heavily utilized in this work. It is
presented in ICRP 103 (2007) as Box A.1, page 191.
The development of the tissue weighting system was based upon relative radiation
detriment primarily for cancer. The sequential steps used were as follows:
• Determine lifetime cancer incidence risk estimates for radiation‐associated cancers:
for 14 organs or tissues, male and female lifetime excess cancer risk were estimated
using both the excess relative risk (ERR) and excess absolute risk (EAR) models and
were then averaged across sexes.
• Apply a dose and dose‐rate effectiveness factor (DDREF): the lifetime risk estimates
were adjusted downward by a factor of two to account for a DDREF (except for
leukemia, where the linear‐quadratic model for risk already accounts for the
DDREF).
• Transfer risk estimates across populations: To estimate radiation risk for each
cancer site, a weighting of ERR and EAR lifetime risk estimates was established that
provided a reasonable basis for generalizing across populations with different
baseline risks (ERR:EAR weights of 0:100% were assigned for breast and bone
marrow, 100:0% for thyroid and skin, 30:70% for lunch and 50:50% for all others).
• Nominal risk coefficients: these weighted risk estimates, when applied to and
averaged across seven western and Asian populations, provided the nominal risk
coefficients given in tables A.4.1 and A.4.2 [in ICRP 103].
Page: 79
• Adjustment for lethality: The lifetime risks for respective cancer sites, which were
based on excess incident cancers, were converted to fatal cancer risks by
multiplying by their lethality fractions, as derived from representative national
cancer survival data.
• Adjustment for quality of life: A further adjustment was applied to account for the
morbidity and suffering associated with non‐fatal cancers.
• Adjustment for years of life lost: since the age distributions of types of cancers
differ, the average ages of several types of cancer were estimated from national
cancer data and converted to average years of life lost when a cancer occurs. An
adjustment for years of life lost was then applied to the result of the previous steps.
• Radiation detriment: The results of the calculations above yielded and estimate of
radiation detriment associated with each type of cancer. These, when normalized
to sum to unity, constitute the relative radiation detriments in table A.4.1 [in ICRP
103]
• Tissue weight factors: since the detailed relative radiation determents in Table A.4.1
[in ICRP 103] are imprecise because of uncertainties associated with their
estimation, they were grouped into four categories broadly reflecting the relative
detriments. A group of residual ‘remainder tissues’ was also added to account for
radiation risks to organs or tissues for which detailed radiation risk calculations were
uninformative.
Page: 80 Appendix B – A summary of all input parameters in the uncertainty analysis
Table B.1 Nominal risk input parameters used in the uncertainty analysis
Parameter Distribution Type Mean* Standard Deviation**
Nominal Risk
Esophagus Log‐Normal 15.1 3
Stomach Log‐Normal 79.1 2.4
Colon Log‐Normal 65.4 1.50
Liver Log‐Normal 30.3 2.20
Lung Log‐Normal 114.2 1.40
Bone Normal 7 1.8
Skin Log‐Normal 1000 3
Breast Log‐Normal 112.1 1.41
Ovary Log‐Normal 10.6 1.41
Bladder Log‐Normal 43.4 2.35
Thyroid Log‐Normal 32.5 2.89
Bone Marrow Log‐Normal 41.9 2.35
Gonads Beta 21.4 1.86
Remainder Log‐Normal 143.8 1.85
* For log‐normal distributions, GM is displayed; for normal distributions, regular mean; for triangular distributions, most likely value
** For log‐normal distributions, GSD is displayed, for normal distributions, regular SD, for triangular and uniform distributions, bounds
Page: 81
Table B.2 Lethality fraction input parameters used in the uncertainty analysis
Parameter Distribution Type Mean* Standard Deviation**
Lethality Fraction
Esophagus Uniform ‐‐ 0.92, 0.95
Stomach Uniform ‐‐ 0.82, 0.86
Colon Uniform ‐‐ 0.47, 0.49
Liver Uniform ‐‐ 0.94, 0.96
Lung Uniform ‐‐ 0.88, 0.89
Bone Uniform ‐‐ 0.44, 0.47
Skin Triangular 0.002 0, 1
Breast Uniform ‐‐ 0.28, 0.30
Ovary Uniform ‐‐ 0.56, 0.58
Bladder Uniform ‐‐ 0.28, 0.30
Thyroid Uniform ‐‐ 0.06, 0.08
Bone Marrow Uniform ‐‐ 0.66, 0.68
Gonads Uniform ‐‐ 0.56, 0.81
Remainder Triangular 0.49 0, 1 * For log‐normal distributions, GM is displayed; for normal distributions, regular mean; for triangular distributions, most likely value
** For log‐normal distributions, GSD is displayed, for normal distributions, regular SD, for triangular and uniform distributions, bounds
Page: 82
Table B.3 Minimum lethality weight input parameters used in the uncertainty analysis
Parameter Distribution Type Mean* Standard Deviation** Minimum Lethality Weight (qmin)
Esophagus Triangular 0.1 0, 0.3
Stomach Triangular 0.1 0, 0.3
Colon Triangular 0.1 0, 0.3
Liver Triangular 0.1 0, 0.3
Lung Triangular 0.1 0, 0.3
Bone Triangular 0.1 0, 0.3
Skin Triangular 0.0 0, 0.3
Breast Triangular 0.1 0, 0.3
Ovary Triangular 0.1 0, 0.3
Bladder Triangular 0.1 0, 0.3
Thyroid Triangular 0.2 0, 0.3
Bone Marrow Triangular 0.1 0, 0.3
Gonads Triangular 0.1 0, 0.3
Remainder Triangular 0.1 0, 0.3
* For log‐normal distributions, GM is displayed; for normal distributions, regular mean; for triangular distributions, most likely value
** For log‐normal distributions, GSD is displayed, for normal distributions, regular SD, for triangular and uniform distributions, bounds
Page: 83
Table B.4 Relative length of life lost input parameters used in the uncertainty analysis
Parameter Distribution Type Mean* Standard Deviation**
Relative length of life lost
Esophagus Log‐Normal 13.05 1.71
Stomach Log‐Normal 13.20 1.82
Colon Log‐Normal 14.55 1.77
Liver Log‐Normal 13.20 1.81
Lung Log‐Normal 12.00 1.69
Bone Uniform ‐‐ 0, 30
Skin Uniform ‐‐ 0, 30
Breast Log‐Normal 19.35 1.66
Ovary Log‐Normal 16.80 1.72
Bladder Log‐Normal 10.65 1.76
Thyroid Log‐Normal 19.35 1.76
Bone Marrow Log‐Normal 24.45 2.31
Gonads Log‐Normal 19.80 1.72
Remainder Triangular 15.45 0, 30
* For log‐normal distributions, GM is displayed; for normal distributions, regular mean; for triangular distributions, most likely value
** For log‐normal distributions, GSD is displayed, for normal distributions, regular SD, for triangular and uniform distributions, bounds
Page: 84 Appendix C – Relative detriment plots generated by Crystal Ball
Figure C.1 Crystal Ball graphical output for the results of the uncertainty analysis on the esophagus relative detriment
Figure C. 2 Crystal Ball graphical output for the results of the uncertainty analysis on the stomach relative detriment
Page: 85
Figure C.3 Crystal Ball graphical output for the results of the uncertainty analysis on the colon relative detriment
Figure C.4 Crystal Ball graphical output for the results of the uncertainty analysis on the liver relative detriment
Page: 86
Figure C.5 Crystal Ball graphical output for the results of the uncertainty analysis on the lung relative detriment
Figure C.6 Crystal Ball graphical output for the results of the uncertainty analysis on the bone relative detriment
Page: 87
Figure C.7 Crystal Ball graphical output for the results of the uncertainty analysis on the skin relative detriment
Figure C.8 Crystal Ball graphical output for the results of the uncertainty analysis on the breast relative detriment
Page: 88
Figure C.9 Crystal Ball graphical output for the results of the uncertainty analysis on the ovary relative detriment
Figure C.10 Crystal Ball graphical output for the results of the uncertainty analysis on the bladder relative detriment
Page: 89
Figure C.11 Crystal Ball graphical output for the results of the uncertainty analysis on the thyroid relative detriment
Figure C.12 Crystal Ball graphical output for the results of the uncertainty analysis on the bone marrow relative detriment
Page: 90
Figure C.13 Crystal Ball graphical output for the results of the uncertainty analysis on the gonads relative detriment
Figure C.14 Crystal Ball graphical output for the results of the uncertainty analysis on the remainder relative detriment