Modifying EPA Radiation Risk Models
Based on BEIR VII
Draft White Paper
Prepared by:
Office of Radiation and Indoor Air
U.S. Environmental Protection Agency
August 1, 2006
-------
Contents
I. Introduction
A. Current EPA Cancer Risk Models 4
B. BEIR VII Models 5
C. Proposed EPA Adjustments and Extensions to BEIR VII Models 5
II. Proposed Methods for Projecting Radiogenic Risk to the U.S. 7
Population
A. Calculating Lifetime Attributable Risk 7
B. Solid Cancer Incidence 7
C. Solid Cancer Mortality 10
D. U.S. Baseline and Census Data 12
E. Combining Results from ERR and EAR Models 13
1. BEIR VII approach 13
2. Proposed EPA approach 14
F. Cancer Sites with Individual Risk Estimates 15
G. Possible Modifications for Radiogenic Lung Cancer Risk 15
Projections
H. Calculating Radiogenic Breast Cancer Mortality Risk 16
I. Leukemia 18
J. Preliminary Risk Calculations 19
K. Uncertainties 19
III. Risks from Higher LET Radiation 21
A. Alpha Particles 21
1. Laboratory studies 21
2. Human data 22
Bone Cancer 22
Leukemia 23
Liver 23
Lung 24
3. Summary and recommendations 26
B. Lower Energy Beta Particles and Photons 27
IV. Risks from Prenatal Exposures 29
V. Skin Cancer Risk 30
VI. Thyroid Cancer Risk 31
VII. References 32
-------
Tables
1. Parameter values for preferred risk models in BEIR VII 9
2. Comparison of the impact of two methods for age-averaging on LAR for 10
solid cancer incidence for selected sites
3. Comparison of impact of two methods for age-averaging on LAR for solid 12
cancer mortality for selected sites
4. Comparison of proposed and BEIR VII method for combining EAR and 14
ERR LAR projections for incidence
5. Baseline lifetime risks and LAR for breast cancer mortality 17
6. Comparison of proposed EPA and BEIR VII LAR calculations 19
7. Lung cancer mortality and RBE 25
8. Estimated parameters for lung cancer mortality risks with 95% confidence 26
intervals for internal lung dose in Mayak Workers and external lung
dose in the Life Span Study Cohort of Japanese Atomic Bomb Survivors
exposed between the ages of 15 and 60
Figure
12-1A Age-time patterns in radiation-associated risks for solid cancer incidence
excluding thyroid and nonmelanoma skin cancer (from BEIR VII)
-------
I. Introduction
In 1994, EPA published a report, referred to as the "Blue Book," which lays out
EPA's current methodology for quantitatively estimating radiogenic cancer risks (EPA
1994). A follow-on report made minor adjustments to the previous estimates and
presented a partial analysis of the uncertainties in the numerical estimates (EPA 1999a).
Finally, the Agency published Federal Guidance Report 13 (FGR-13), which utilized the
previously published cancer risk models, in conjunction with ICRP dosimetric models
and U.S. usage patterns, to obtain cancer risk estimates for over 800 radionuclides, and
for several exposure pathways (EPA 1999b).
The National Research Council (NRC) of the National Academy of Sciences
(NAS) recently released a report on the health risks from exposure to low levels of
ionizing radiation (NRC 2006). Cosponsored by the EPA and several other Federal
agencies, Health Risks from Exposure to Low Levels of Ionizing Radiation BEIR VII
Phase 2 (BEIR VII) primarily addresses cancer and genetic risks from low doses of low-
LET radiation.
In this paper, we outline proposed changes in EPA's methodology for estimating
radiogenic cancers, based on the contents of BEIR VII and some ancillary information.
For the most part, we expect to adopt the models and methodology recommended in
BEIR VII; however, we believe that certain modifications and expansions are desirable or
necessary for EPA's purposes. At this point, EPA is seeking advice from the Agency's
Science Advisory Board's Radiation Advisory Committee (RAC) on the application of
BEIR VII and on issues relating to these modifications and expansions. After receiving
the advisory review, we plan to implement changes in our methodology through the
publication of a revised Blue Book, which we would expect to submit to the RAC for
final review. The revised Blue Book could then serve as a basis for an updated version of
FGR-13.
A. Current EPA Cancer Risk Models
For most cancer sites, radiation risk models are generally derived from
epidemiological data from the life span study (LSS) of the atomic bomb survivors.
EPA's models for esophagus, stomach, colon, lung, ovary, bladder, leukemia, and
"residual" cancers were adapted from the models published by Land and Sinclair based
on a linear, no-threshold fit to the LSS data (Land and Sinclair 1991). For each solid
tumor site, gender, and age-at-exposure interval, there is a model providing a coefficient
for the excess relative risk (ERR) per Gy for cancer mortality, which is assumed to be
constant beginning at the end of a minimum latency period until the end of life. Land
and Sinclair present two sets of models — "multiplicative" and "NIH "— differing in
how one "transports" risk from the Japanese LSS population to another population, e.g.,
to the U.S. population. For the multiplicative model, it is assumed that the ERR/Gy is the
same in all populations, whereas, for the NIH model, it is assumed that the excess
-------
absolute risk is the same in different populations^^ the limited period of epidemiological
follow-up. Given the scarcity of information on how radiogenic cancer risk varies
between populations having differing baseline cancer rates, EPA adopted an intermediate
"GMC" model for each site, where the ERR coefficients were taken to be the geometric
mean of the corresponding ERR coefficients for the multiplicative and NIH models (EPA
1994).
For leukemia, the temporal response in the models was more complex, but the
approach for transporting risk to the U.S. population was analogous. Following the
approach of Land and Sinclair, EPA also developed a GMC model for kidney from the
LSS data. EPA's models for other sites, including breast, liver, thyroid, bone, and skin
were based on various authoritative reports (NCRP 1980; NRC 1988; ICRP 199la, b;
Gilbert 1991). Based primarily on ICRP recommendations at that time, for low doses
and dose rates, each coefficient was reduced by a factor (DDREF) of 2 from what would
be obtained from a linear, no-threshold fit to the LSS data.
B. BEIR VII Models
BEIR VII site-specific models derived from the LSS differ from those of Land
and Sinclair in several significant ways: (1) they are derived primarily from data on
cancer incidence rather than cancer mortality; (2) mathematical fitting is performed to
better reflect the functional dependence of solid cancer risk on age at exposure and
attained age; (3) a weighted average of risk projection models was used to transport risk
from the LSS to the U.S. population; (4) a value for the DDREF of 1.5 was estimated
from the LSS and laboratory data; (5) quantitative uncertainty bounds are provided for
the site-specific risk estimates in BEIR VII.
For breast cancer and thyroid cancer, BEIR VII risk models were based on pooled
analyses of data from the LSS cohort, together with data on medically irradiated cohorts
(Preston et al. 2002, Ron et al. 1995).
C. Proposed EPA Adjustments and Extensions to BEIR VII Models
In implementing its revised methodology for estimating radiogenic cancer risks,
EPA proposes to adopt many of the recommendations in BEIR VII. One significant
extension to be considered is the estimation of risks from exposures to higher LET
radiations, especially to alpha particles, but also to lower energy photons and beta
particles. Particularly important in this regard is the risk from alpha emitters deposited in
the lung and the bone. BEIR VII presents no risk estimates for radiogenic bone cancer.
As in the past, we propose to estimate bone cancer risk from data on radium injected
patients.
BEIR VII also fails to provide quantitative estimates of risk for skin cancer, both
of which might be significant under some exposure conditions. Risks from prenatal
-------
exposures are also not fully addressed by the report. BEIR VII presents a model for
estimating radiogenic thyroid cancer incidence, but not thyroid cancer mortality. We
hope to address these gaps and to consider the findings of an EPA sponsored thyroid
report being drafted by the NCRP, when it becomes available.
As explained in Section II, we intend to employ somewhat different population
statistics than BEIR VII. Consideration is given here to an alternative model for
estimating radiogenic lung cancer, which diminishes what appears to be an anomalously
high lung cancer risk projected in BEIR VII for females. For breast cancer, an alternative
method is introduced for estimating mortality, which takes into account changes in
incidence rates and survival rates over time.
BEIR VII provides quantitative uncertainty bounds for each of its risk
coefficients. Nevertheless, in deriving these bounds, it is clear that some sources of
uncertainty were not included. Most important, no uncertainty was assigned to the form
of the dose-response relationship: it was implicitly assumed that the dose-response
relationship is "linear-quadratic", which allowed the BEIR VII Committee to place
uncertainty bounds on the "DDREF". Mechanisms pertaining to the biological effects of
low level ionizing radiation are being investigated, which could eventually mandate a
different dose-response model, with resulting large changes in estimates of risk at low
doses. Assigning probabilities to alternative models would be highly subjective at this
time. We do not propose to quantify the uncertainty pertaining to low-dose extrapolation,
beyond what was done in BEIR VII, but we would expect to include a brief discussion of
the issue in our revised risk assessment document.
-------
II. Proposed Methods for Projecting Radiogenic Risk to the U.S.
Population
A. Calculating Lifetime Attributable Risk
As in BEIR VII, we propose using lifetime attributable risk (LAR) as our primary
risk measure. For a person exposed to dose (x) at age (e), the LAR is given by:
110
LAR(x,e)= $M(x,e,d)-S(d)/S(e)da. (1)
e+L
where
M(x,e,d) is the excess absolute risk at attained age a from an exposure at age e,
S(a) is the probability of surviving to age a, and L is the latency period (2 years for
leukemia, 5 years for solid cancers). The LAR approximates the probability of a
premature cancer death from radiation exposure, and in BEIR VII (approximate) values
for the LAR are obtained as weighted sums (over attained ages a up to age 100) of the
excess probabilities of radiation-induced cancer incidence or death, M(x,e,d). We
intend instead to calculate the integral (Eq. 1) to age 110 (or perhaps 120) using spline
approximations - not unlike the approach used to calculate EPA's current risk
coefficients (EPA 1999).
The LAR for a population is calculated as a weighted average of the age-at-
exposure specific risks discussed above. The weights are proportional to the number of
people, N(e), who would be exposed at age e. The population-averaged LAR is given by:
110-L
-| 1 1U — l^
LAR(X,pop} = — \N(e)-LAR(x,e)-de. (2)
For the BEIR VII approach, N(e) is the number of people from census data in the U.S.
population at age e for a reference year - BEIR VII used 1999, and N* is the total
summed over all ages. In contrast, for our primary projection, we propose to use a
hypothetical stationary population for which the N(e) are proportional to S(e) based on
observed mortality rates for the year 2000. Under the assumption that there would be no
appreciable change in future mortality rates, this would approximate the radiogenic risk
from a lifetime (chronic) exposure at constant dose rate. A stationary population is being
used for our current risk assessment (EPA 1999).
B. Solid Cancer Incidence
For most cancer sites, separate evaluations of LAR were made using both an
excess absolute risk (EAR) model and an excess relative risk (ERR) model. For most
solid cancers (all but thyroid, and breast cancer), the ERR and EAR models were based
exclusively on analyses of the atomic bomb survivor incidence data. This differs from
-------
the risk models that had been used in previous risk assessments, and had been derived
from mortality data.
Except for breast and thyroid cancers, the preferred BEIR VII EAR and ERR
models are functions of sex, age at exposure, and attained age, and were of the form:
,e,a) or ERR(x,e,a) = J3SD exp(^*)(a 760)' ,
, „ min(e,30)-30
where e* = - - - - - .
10
As seen in Table 1, the values for the parameters J3s,y, and 77 depend on the type of
model - EAR or ERR. For ERR models for most sites:
(3 , the ERR at age-at-exposure 30 and attained age 60, tends to be larger for
females than males;
y = -0.3 implies the radiogenic risk of cancer at age a falls by about 25% for
every decade increase in age-at-exposure up to age 30; and
77 = -1.4 implies the ERR is almost 20% smaller at attained age 70 than at age 60.
Thus, ERR decreases with age-at-exposure (up to age 30) and attained age. In
contrast, for EAR models for most sites, y = -0.41 and 77 = 2.8. EAR decreases with
age-at-exposure but 12-1 A (NAS 2006, increases with attained age. These patterns are
illustrated in BEIR VII (NAS 2006, Figure 12.1 A, p. 270).
AJ» at 10
A§« II 20
Ag* II JO*
TO H
#t KJ •
> ,
a M I
S i'
fi » h
*»^
*3 53 to ?0
Jitolrwd
TO
at m
FIGURE 12-1A Age-time patterns in radiation-associated risks for solid cancer incidence excluding thyroid and
nonmelanoma skin cancer. Curves are sex-averaged estimates of the risk at 1 Sv for people exposed at age 10 (solid
lines), age 20 (dashed lines), and age 30 or more (dotted lines). Estimates were computed using the parameter estimates
shown in Table 12-1 of BEIR VII. FROM BEIR VII
For either type of model, calculating the LAR is relatively straightforward. For
the EAR models, note thatM(x,e,a) = EAR(x,e,a). For ERR models,
M(x,e,a) = ERR(x, e, a) • A7 (a),
8
-------
where A7 (a) is the baseline cancer incidence rate at age a. Values for LAR are then
obtained using equations 1 and 2.
Results of LAR calculations are given in Table 2. Separate calculations were
made using both census data - weights proportional to the number of people of each age
in the year 2000, and a stationary population - based on mortality data for the year 2000.
For most sites, the LAR is about 5-10% larger when based on weights from census data.
Table 1: Parameter values for preferred risk models in BEIR VII1
Cancer
Stomach
Colon
Liver
Lung
Breast
Prostate
Uterus
Ovary
Bladder
Other solid
Thyroid2
ERR model
PM
0.21
0.63
0.32
0.32
PF
0.48
0.43
0.32
1.4
Y
-0.3
-0.3
-0.3
-0.3
11
-1.4
-1.4
-1.4
-1.4
Not used
0.12
0.5
0.27
0.53
0.055
0.38
1.65
0.45
1.05
-0.3
-0.3
-0.3
-0.3
-0.3
-0.83
-1.4
-1.4
-1.4
-1.4
-2.8
0
EAR model
PM
4.9
3.2
2.2
2.3
PF
4.9
1.6
1
3.4
Y
-0.41
-0.41
-0.41
-0.41
11
2.8
2.8
4.1
5.2
See text
0.11
1.2
6.2
1.2
0.7
0.75
4.8
-0.41
-0.41
-0.41
-0.41
-0.41
2.8
2.8
2.8
6
2.8
Not used
1 From Table 12-2 (NAS 2006)
2 Unlike other sites, the dependence on ERR on age-at-exposure is not limited to ages < 30y.
-------
Table 2: Comparison of the impact of two methods for age-averaging on LAR for
solid cancer incidence for selected sites. Projections are made using the BEIR VII
EAR and ERR models. Age-averaging is based on either 2000 census data (NCHS
2004) or a stationary population constructed from 2000 life tables (Arias 2002).
Site
Stomach
Colon
Liver
Lung
Breast
Prostate
Uterus
Ovary
Bladder
Sex
Male
Female
Male
Female
Male
Female
Male
Female
Female
Male
Female
Female
Male
Female
Risk Model
Population Weighting
E^
Census
278
328
182
107
150
84
189
361
463
6
80
47
121
101
LR
Stationary
259
308
169
100
141
80
179
344
423
6
75
44
115
96
El
Census
23
30
256
164
25
9
246
767
507
202
16
75
170
165
IR
Stationary
22
29
240
155
23
9
230
714
465
187
15
69
160
155
C. Solid Cancer Mortality
The ERR and modified versions of the EAR models just discussed were used in
BEIR VII to calculate LAR for radiation-induced cancer death. For ERR, the same
models were used for both incidence and mortality,
M(x,e,d) = ERR(x, e, a) • hM (a) .
For EAR, BEIR VII used essentially the same approach by assuming
.
O) • (3)
, ,, . EAR(x,e,a)
M(x,e,d) = - '
Note that the ratio of age-specific EAR to incidence rate is the ERR for incidence - based
on the EAR model. Equations (1) and (2) are then applied to obtain the LAR. This BEIR
VII approach, equating the incidence and mortality ERR, ignores the "lag" between
incidence and mortality, which could lead to bias in the estimate of mortality risk in at
least two different ways.
10
-------
First, there would be a corresponding lag between the ERR for incidence and
mortality, which might result in an underestimate of mortality risk. For purposes of
illustration, suppose that a particular cancer is either cured without any potential life-
shortening effects or results in death exactly 10 years after diagnosis, and that survival
does not depend on whether it was radiation-induced. Then, with subscripts M and /
denoting mortality and incidence:
The same relationship would hold for EAR, if the baseline cancer rate has the same age-
dependence for A-bomb survivors as for the U.S. population.
Second, since current cancer deaths often occur because of cancers that develop
years ago, application of the EAR-based ERR for incidence can result in a substantial
bias due to birth cohort effects. If age-specific incidence rates increase (decrease) over
time, the denominator in Equation 3 would be too large (small). This could result in an
underestimate (overestimate) of the LAR.
The BEIR VII approach is reasonable for most cancers, because the time between
diagnosis and a resulting cancer death is typically short. An exception is breast cancer,
and an alternative approach is outlined in Section H.
Results of LAR calculations using the BEIR VII approach are given in Table 3.
Similar to what was observed for incidence, LAR for mortality tends to be about 5%
larger using census-based weights than for weights based on a stationary population.
Mortality and incidence data used for the calculations are described in the next section.
11
-------
Table 3: Comparison of impact of two methods for age-averaging on LAR for solid
cancer mortality for selected sites. Projections are made using the BEIR VII
EAR and ERR models. Age-averaging is based on either 2000 census data
(NCHS 2004) or a stationary population constructed from 2000 life tables
(Arias 2002).
Site Sex
Stomach
Colon
Liver
Lung
Breast
Prostate
Uterus
Ovary
Bladder
Male
Female
Male
Female
Male
Female
Male
Female
Female
Male
Female
Female
Male
Female
Risk Model
Population Weighting
EAR
Census
146
183
84
47
114
70
179
312
103
1
19
27
28
33
Stationary
136
173
79
44
108
66
169
298
94
1
18
26
27
32
ERR
Census
12
17
119
72
17
7
229
618
101
32
3
36
34
43
Stationary
11
16
113
69
16
7
214
577
94
30
2
34
32
41
D. U.S. Baseline and Census Data
Cancer specific incidence and mortality rates are based on the Surveillance,
Epidemiology, and End Results (SEER) program of the National Cancer Institute (NCI).
Begun in the early 1970's, SEER collects data from several mostly statewide and
metropolitan cancer registries within the U.S. Rates for this report are calculated using
SEER-Stat and the 1975-2003 SEER public-use data (SEER 2006a,b) available from the
SEER website (http://seer.cancer.gov). The dataset is structured to represent two notable
expansions in the SEER program: from 9 registries to 13 registries (SEER 13) in the early
1990's and most recently to 17 registries (SEER 17). For this report, incidence rates for
each 5 year age-interval are the average of data for the years 1998-2000 from SEER 13,
and SEER 17 data for the years 2000-2002. This contrasts with BEIR VII, which used (a
previous version) of public-use SEER 13 data for the years 1995-99.
SEER regularly revises its statistics on baseline rates, and the baseline rates used
for our final risk assessment will likely be based on SEER statistics for the year 2000 that
12
-------
are not yet available. For example, it is anticipated that the denominator (person years at
risk) for future versions of the SEER cancer data will be derived using 2000 decennial
census results.
SEER areas currently comprise about 26% of the U.S. population, and are not a
random sample of areas within the U.S. Nevertheless the cancer rates observed in the
combined SEER areas are thought to be reasonably similar to rates for the U.S.
population. For the final risk assessment, datasets that would reflect a much larger area
than SEER represents will also be considered, such as data that incorporates the CDC
National Program of Cancer registries.
Finally, we used life tables for the year 2000 (compared to 1999 for BEIR VII),
(Arias 2002) available at http://www.cdc.gov/nchs/data/lt2000.pdf. Life tables for our
final risk assessment will likely be based on NCHS data that incorporates data from the
decennial census.
Changes in any of these data sets for the final risk assessment will likely have
only minor changes in risk projections.
E. Combining Results from ERR and EAR Models
1. BEIR VII approach
BEIR VII calculates LAR values separately based on preferred EAR and ERR
models, and then combines results using a weighted geometric mean. More specifically:
LARB1 = (LARRY (LARA)l~w,
with weight (w) - on results from the ERR model - depending on cancer site. If a weight
(w) equals 0.5, a simple geometric mean would be calculated. Instead for most cancer
sites, BEIR VII recommended weights (w) equal to 0.7 - placing somewhat more
emphasis on results from ERR models. (A notable exception is lung cancer where the
EAR model is given more weight, reflecting near additivity between smoking and gamma
radiation in the A-bomb survivor data.)
A problem with the BEIR VII method for averaging the EAR and ERR
projections is that the geometric mean is not additive in the sense that the geometric mean
of two risk projections for the combined effect of separate exposures is generally not
equal to the sum of the geometric mean projections for the exposures. We circumvent
this problem by first calculating the weighted geometric mean of the excess absolute risk
for the two projection models for each at each age at exposure and attained age. Then,
results can be integrated to calculate risk from chronic lifetime exposure.
13
-------
2. Proposed EPA approach
Thus, we propose calculating the combined age-specific risk (at high dose rates)
according to:
with the LAR at exposure age e calculated as before:
110
LARC (x, e) = MC (x, e, a) • S(d) / S(e)da
e+L
The difference with the BEIR VII approach is that the risk models would be combined
before integrating results according to (1) to obtain the LAR.
The two methods of combining results from EAR and ERR models, BEIR VII and
the proposed EPA approach, are compared in Table 4. The method of combination has
only a minimal impact on age-averaged LAR calculations.
Table 4: Comparison of proposed and BEIR VII method for combining
EAR and ERR LAR projections for incidence.1
Site Sex
Stomach
Colon
Liver
Lung
Breast
Prostate
Uterus
Ovary
Bladder
Male
Female
Male
Female
Male
Female
Male
Female
Female
Male
Female
Female
Male
Female
Method of Combination
Proposed
45
58
214
135
38
17
186
401
423
63
115
58
141
129
BEIR VII
46
58
216
136
39
17
193
428
423
66
120
60
145
134
Results are shown for stationary populations and SEER incidence data
for the years 1998-2002.
14
-------
F. Cancer Sites with Individual Risk Estimates
With few exceptions, we agree with BEIR VII's choice of sites for individual risk
projections. BEIR VII cited two main criteria for deciding on the sites for which
individual risk projections should be provided. First, the cancer should be linked clearly
with radiation exposure, and second, there should be adequate data to develop reliable
risk estimates. In our view, these criteria are generally reasonable. Leukemia and
cancers of the stomach, colon, liver, lung, breast, bladder, ovary and thyroid all satisfy
the first and arguably both criteria. In addition, BEIR VII also provides individual
projections for cancers of the uterus and prostate. In our view, the choice of uterine
cancer is appropriate. The A-bomb survivor data provides sufficient information for
radiogenic uterine cancer to formulate a risk projection of reasonable precision.
Regarding prostate cancer, BEIR VII cites the vastly differing baseline rates for the U.S.
compared to Japan for providing a separate prostate estimate. The BEIR VII estimate for
prostate cancer - based on data that includes less than 300 cancers - is unreliable, being
essentially dominated by the effects of sampling error. Nevertheless, at this point, we
anticipate adopting the BEIR VII model for prostate cancer. As discussed elsewhere, we
plan to provide individual estimates for bone cancer and possibly non-melanoma skin
cancer. A discussion of possible changes to the application of the BEIR VII models for
projecting radiogenic lung cancer risks follows.
G. Possible Modifications for Radiogenic Lung Cancer Risk Projections
In our view, the BEIR VII lung cancer ERR model might not be appropriate for
projecting radiogenic risk to the U.S. population. The BEIR VII ERR projection is based
upon a fit to the A-bomb survivor data without any adjustment for smoking, which
distorts the ERR for radiogenic lung cancer. This is due to substantial birth-cohort
related changes in smoking habits among cohort members, and the likelihood that the
effects of smoking and radiation are approximately additive (Pierce et al. 2003). Thus,
without adjustment for smoking, observed ERR values in the A-bomb cohort are largest
for early birth cohorts who smoked less and had lower baseline lung cancer rates, but
may have similar absolute risks from radiation as other survivors. This would distort the
age-at-exposure effect on radiogenic risk because the observed ERR would tend to be an
overestimate for older ages-at-exposure (early birth dates) as compared to younger ages-
at-exposure (later birth dates). Pierce et al. also discuss how - without adjustment for
smoking - attained-age effects might also be distorted.
Overall, the modeled ERR for radiogenic lung cancer in the A-bomb cohort
would be higher than might be appropriate for the U.S. population. Based on evidence
that radiation and smoking are roughly additive in the A-bomb survivors, the absolute
risk from radiation would be similar in the U.S. and Japan. However, the U.S. smoking-
related lung cancer rates would almost surely be substantially higher than had been
observed in the A-bomb cohort.
15
-------
Not accounting for smoking also affects the estimate of the female:male sex ratio
(fiF/fly), about 4.4 for the BEIR VII ERR models. This is why BEIR VIFs ERR-based
LAR is much larger for females (0.074 per Gy) than for males (0.026 per Gy). Using a
subset of the LSS for which smoking information was available, Pierce et al. found that
"after adjusting for smoking, the radiation-related risks relative to background rates for
nonsmokers are similar to those for other solid cancers: a sex-averaged ERR/Sv of about
0.9 with a female:male sex ratio of about 1.6. Adjusting for smoking removes a
spuriously large female:male ratio in radiation relative risk due to confounding between
sex and smoking level."
The BEIR VII EAR model would not seem to suffer from the same birth-cohort
related effects that distort the ERR observed in the A-bomb survivors. Given that lung
cancer rates are driven predominantly by smoking patterns, and the evidence for additive
effects of smoking and radiation, an EPA projection based exclusively on the BEIR VII
EAR model should be considered.
Another approach for estimating the LAR, suggested by Pierce et al. (p. 519) is as
follows: "Lifetime risk computations involve estimating age-specific absolute lung
cancer rate increases and then summing these over age with weights corresponding to the
chance of survival to each age. The absolute rate increases may best be estimated by
multiplying lung cancer rates for nonsmokers and the ERR/Sv relative to nonsmokers."
Although there are some compelling arguments for this approach, reasons not to use it
include: 1) the apparent presumption that factors (such as radon) - other than smoking -
are multiplicative with low-LET radiation; 2) ambiguity as to whether non-smokers
includes (any) former smokers; 3) difficulty in estimating baseline never- or non-smoker
lung cancer rates.
H. Calculating Radiogenic Breast Cancer Mortality Risk
This section outlines an alternative method for calculating radiogenic breast
cancer risk, compares results with calculations based on the BEIR VII method, and
indicates potential additional modifications.
As before, M(x,e,ac) denotes the excess absolute risk at attained age ac from
an exposure at age e. The density function for radiogenic cancer at ac would be:
For the cancer to result in a death at age a > ac , the patient would have to survive
the interval (ac , a), and then die from the cancer at age a. This and the concept of the
relative survival rate form the basis for the method. The relative survival rate for a breast
cancer patient would be the ratio of the survival rate for the patient divided by the
expected survival rate (without breast cancer). Assuming that the relative survival
depends only on the length of the time interval and is independent of age of diagnosis,
16
-------
and letting R(t) be the relative survival function, the probability of survival with breast
cancer for the interval (ac,d) would be:
S(a)IS(ac)R(a-ac).
Furthermore, assuming the breast cancer mortality rate among those with breast
cancer (h) does not depend on age, the density function for breast cancer death can be
shown to equal:
a
\C(x,e,ac)S(a)IS(ac)R(a-ac)hdac.
Baseline mortality rates can be derived in a similar manner. Baseline lifetime breast
cancer mortality risk and lifetime LAR estimates are compared in Table 5. The rate h
was set to 0.0233, corresponding to a 5 year relative survival rate of about 0.89 (Ries et
al. 2006).
Table 5: Baseline lifetime risks and LAR for breast cancer mortality.
Method
BEIR VII
EPA Alternative
Baseline lifetime risk
2990
4630
Lifetime LAR
92
126
Much of the discrepancy between the two sets of results seems to be a
consequence of observed increases in breast cancer incidence rates and declines in
mortality rates. From 1980 to 2000, age-averaged breast cancer rates increased by about
35% (102.1 to 135.7), whereas the mortality rates declined by about 15% from 31.7 to
26.6 (http://seer.cancer.gov/csr/1975_2003/results_merged/sect_04_breast.pdf).
To understand the possible implications of the trends in incidence and mortality
on the alternative LAR calculations, consider the assumptions on which the projections
are based:
The BEIR VII projection is based on the following formula:
M(x,e,a) = EAR(x,e,a) — .
^(d)
Essentially, the absolute risk of radiogenic cancer death from an exposure at age e is
assumed to be equal to the absolute risk of a radiation-induced cancer multiplied by a
lethality ratio (that depends on attained age). In BEIR VII, the lethality ratio is estimated
as the ratio of current mortality rates divided by current incidence rates. However, since
the time between breast cancer diagnosis and death is relatively long, lethality rates
17
-------
would seem to be best determined by comparing current mortality rates to incidence rates
observed for (much) earlier time periods. If as data indicate, current incidence rates are
much larger than past incidence rates, the BEIR VII denominator is too large, and the
estimated lethality ratio is too small. In contrast, the BEIR VII calculation of lifetime
baseline mortality risk is a valid estimate of risk with respect to current mortality rates,
since in particular it does not depend on an estimate of lethality. Of course, it is only a
snapshot reflecting the overall rates of breast cancer deaths currently observed for all
U.S. women, and would not be a reliable indicator of future breast cancer mortality (that
might result from currently diagnosed cases).
The proposed EPA alternative projection also has limitations. The validity of the
projections would depend on the extent to which estimates of relative survival functions
can be used to approximate mortality rates from breast cancer for people with breast
cancer. There are two potential problems with this approach which merit further
investigation:
One relates to the improvement in survival rates for women with breast cancer.
Long-term survival rates for breast cancer patients are needed to construct valid estimates
for this approach, but since these survival rates can change rapidly, extrapolation of rates
for periods beyond 5 to 10 years may be unreliable. The relative survival rates upon
which the estimates in Table 5 were based reflect "current" rates, and were constructed
from SEER data for 1996-2002. Although a simple exponential function seems to be a
very good approximation to the relative survival up to 5 years, there is no guarantee as to
how appropriate such an approximation would be for (extrapolation to) longer time
periods. It should be noted that the alternative "EPA" estimate of baseline lifetime risk
for breast cancer mortality reflects current incidence rates and "current" (short-term)
breast cancer survival rates. In contrast to the BEIR VII estimate of baseline risk, such an
approach may result in a valid estimate of mortality from currently occurring cases to the
extent that long-term rates can be approximated - based on extrapolations - from recent
data.
The other problem is that the reduced expected survival among breast cancer
patients may be partly attributable to causes other than breast cancer. For example, if
some breast cancers are smoking-related, breast cancer patients as a group may be at
greater risk of dying from lung cancer.
I. Leukemia
We plan to adopt the approach used in BEIR VII with only minor modifications.
Calculations would be based on the BEIR VII ERR and EAR models:
EAR(x,e,a,t] or ERR(x,e,a,t) = j3sx(\ + 6x)exp[re*+S\og(t/25) + 0e*\og(t/25)],
where t is time since exposure. We would use the parameter values recommended in
BEIR VII (Table 12-3). LAR would be calculated using the same methodology proposed
18
-------
for most solid cancer sites. Following the BEIR VII recommendation, a DDREF of 1
will be assumed. Departures from the BEIR VII methods include: (1) calculating a
weighted geometric mean of the excess absolute risk for the two projection models for
each at each age at exposure and attained age, and (2) using weights based on a stationary
population for our primary projection.
J. Preliminary Risk Calculations
Table 6 compares LAR calculations for selected solid cancers based on the
proposed EPA methods with calculations in BEIR VII. With the exception of lung
cancer, the proposed EPA methods are based on the same EAR and ERR models as in
BEIR VII. For lung cancer, the BEIR VII ERR model would not be used.
Table 6: Comparison of proposed EPA and BEIR VII LAR calculations.
Site
Stomach
Colon
Liver
Lung1
Breast
Prostate
Uterus
Ovary
Bladder
Sex
Male
Female
Male
Female
Male
Female
Male
Female
Female
Male
Female
Female
Male
Female
Incidence
EPA
30
39
143
90
25
11
119
229
282
42
15
39
94
86
BEIR VII
34
43
160
96
27
12
140
300
310
44
20
40
98
94
Mortality
EPA
16
22
67
40
18
9
113
199
842
8
3
20
20
25
BEIR VII
19
25
76
46
20
11
140
270
73
9
5
24
22
28
:EPA projection of lung cancer based on BEIR VII EAR model.
2EPA projection based on alternative method as described in the previous section.
K. Uncertainties
The BEIR VII Report includes a quantitative uncertainty analysis with 95%
subjective CIs for each site-specific risk estimate. The analysis focused on three sources
of uncertainty thought to be most important: (1) sampling variability in the LSS data, (2)
the uncertainty in transporting risk from the LSS to the U.S. population, and (3) the
19
-------
uncertainty in the appropriate value of a DDREF for projecting risk at low doses and dose
rates from the LSS data.
The BEIR VII analysis neglected other sources of uncertainty, including: (1)
errors in dosimetry; (2) errors in disease detection and diagnosis; (3) uncertainty in the
age and temporal pattern of risk, especially for individual sites, which was usually taken
to be the same as that derived for all solid tumors; (4) uncertainty in the relative
effectiveness of medical x rays in inducing cancer for those sites where data on medically
irradiated cohorts were used in deriving the risk models.
It should also be noted that the treatment of uncertainty in projecting risk at low
doses and dose rates of low-LET radiation, basically assumes the "linear-quadratic" dose-
response model in which: (1) the risk from an acute dose, D, is of the form oD + flD2 and
(2) the risk from low dose rate radiation is then simply aD. The assumption here that one
can extrapolate risk linearly from moderate acute doses (-100 mGy) to very low dose
fractions remains contentious.
EPA proposes adopting the BEIR VII quantitative uncertainty bounds for most
purposes. It is anticipated, however, that the revised Blue Book would contain an
examination of where these uncertainty bounds might fail to adequately capture the
overall uncertainty. In addition, we would include a brief discussion of the low dose
extrapolation problem, which would acknowledge continuing disagreement on this issue.
Ultimately, the estimates of uncertainties in risk per unit dose can be combined
with estimates of uncertainties in tissue doses for internally deposited radionuclides in
order to obtain uncertainty estimates for inhaled or ingested radionuclides. For alpha
emitting radionuclides this will require additional assessment of risk and uncertainty
beyond what is contained in BEIR VII. Estimation of risk from internally deposited
alpha emitters is addressed in the next section.
20
-------
III. Risks from Higher LET Radiation
A. Alpha Particles
Assessing the risks from ingested or inhaled alpha-emitting radionuclides is
problematic from two standpoints. First, it is often difficult to accurately estimate the
dose to target cells, given the short range of alpha-particles in aqueous media (typically
< 100 um) and the often non-uniform distribution of a deposited radionuclide within an
organ or tissue. Second, there is very little direct human data on cancer induction by
alpha particles. For most tissues, the risk to a given tissue from a given dose of alpha
radiation must be calculated based on the estimated risk from an equal absorbed dose of y
rays multiplied by an "RBE" factor that accounts for different carcinogenic potencies of
the two types of radiation, derived from what are thought to be relevant comparisons in
experimental systems
The high density of ionizations associated with tracks of alpha radiation produces
DNA damage which is less likely to be faithfully repaired than damage produced by low-
LET tracks. Consequently, for a given absorbed dose, the probability of inducing a
mutation is higher for alpha radiation, but so is the probability of cell killing. The
effectiveness of alpha radiation relative to some reference low-LET radiation (e.g., 250
kVp x rays or 60Co y rays) in producing some particular biological end-point is referred to
as the alpha-particle relative biological effectiveness (RBE). The RBE may depend not
only on the observed end-point (induction of chromosome aberrations, cancer, etc.), but
also on the species and type of tissue or cell being irradiated, as well as dose and dose
rate.
In most experimental systems, the RBE increases with decreasing dose and dose
rate, apparently approaching a limiting value. This mainly reflects reduced effectiveness
of low-LET radiation as dose and dose rate are decreased-presumably because of more
effective repair. In contrast, the effectiveness of high-LET radiation in producing
residual DNA damage, transformations, cancer, etc. may actually decrease at high doses
and dose rates, at least in part due to the competing effects of cell killing. For both low-
and high-LET radiations, it is posited that at low enough doses, the probability of a
stochastic effect is proportional to dose, and independent of dose rate. Under these
conditions, the RBE is maximal and equal to a constant RBEM. In order to estimate site-
specific cancer risks for low dose alpha radiation, we need a low-dose, low-LET risk
estimate for that site and an estimate of the RBEM.
1. Laboratory studies
The experimental data on the RBE for alpha particles and other types of high-LET
radiation have been reviewed by the NCRP (NCRP 1990) and the ICRP (ICRP 2003).
From laboratory studies, the NCRP concluded that: "The effectiveness of alpha emitters
is high, relative to beta emitters, being in the range of 15 to 50 times as effective for the
induction of bone sarcomas, liver chromosome aberrations, and lung cancers." The
21
-------
NCRP made no specific recommendations on a radiation weighting factor for alpha
radiation.
The ICRP has reiterated its general recommendation of a radiation weighting
factor of 20 for alpha-particles (ICRP 2003, 2005). However, ICRP Publication 92
further states (ICRP 2003):
Internal emitters must be treated as a separate case because their RBE depends not merely
on radiation quality, but also, and particularly for a-rays with their short ranges, on their
distribution within the tissues or organs. It is, accordingly, unlikely that a single WR
should adequately represent the RBEM for different a emitters and for different
organs.. .The current wRof 20 for a-rays can thus serve as a guideline, while for specific
situations, such as the exposure to radon and its progeny, or the incorporation of 224Ra,
226Ra, thorium, and uranium, more meaningful weighting factors need to be derived.
Another set of recommendations for a-particle RBE is contained in the NIOSH-
Interactive RadioEpidemiological Program (NIOSH-IREP) Technical Documentation
meant for use in adjudicating claims for compensation of radiogenic cancers (NIOSH
2002, Kocher et al. 2005). IREP posits a lognormal uncertainty distribution for its
radiation effectiveness factor (REF, equivalent to RB£M) with a median of 18 and a 95%
CI [3.4, 101]. For leukemia, IREP employs a hybrid distribution: REF=1.0 (25%);
=LN(1,15) (50%); =LN(2,60) (25%) where LN(a,b) represents a lognormal distribution
witha95%CIof[a,b].
Studies comparing groups of animals inhaling insoluble particles to which are
attached alpha or beta emitters have been performed to assess RBE for lung cancer. In a
large long-term study of beagle dogs, Hahn et al. (1999) reported that the RBE was at
least 20. In contrast, from an analogous study of lung cancer induction in CBA/Ca mice,
Kellington et al. (1997) estimated the RBE to be only 1.9.
2. Human data
Results from epidemiological studies of groups exposed to alpha radiation can be
used, directly, to develop risk estimates for alpha radiation; they can also be used in
conjunction with low-LET studies to estimate RBE; finally, it is possible to use results
from these studies in combination with estimates of RBE to derive low-LET risk
estimates where none can be obtained from low-LET studies.
There are four cancer sites for which we have direct epidemiological evidence on
the risks from alpha irradiation: bone, bone marrow, liver, and lung. In each of these
cases except for bone, we also have risk estimates for low-LET radiation derived from
the LSS.
Bone Cancer. Although there is some new information coming from research on
Mayak plutonium workers, the most definitive information on bone cancer risk continues
to be radium dial painters exposed to 226Ra and 228Ra and patients injected with the
shorter-lived isotope 224Ra. The usefulness of the dial painter data for low dose risk
estimation suffers from lack of complete epidemiological follow-up and from the
22
-------
possible complicating effects of extensive tissue damage associated with very high doses
of radiation in the bone. For this reason, EPA has taken its estimates of risk of alpha-
particle-induced bone sarcoma from the BEIRIV analysis of the 224Ra data, which is
consistent with a linear, no-threshold dose response (NRC 1988, EPA 1994). The
corresponding low-LET risk estimate (per Gy) was assumed to be a factor of 20 lower
based on the assumed alpha-particle RBEM of 20. We believe that no major change in
bone cancer incidence risk estimates is warranted; some downward adjustment of
mortality estimates may be warranted to reflect improved survival.
Leukemia. Excess leukemia cases have not been observed in studies of radium
dial painters or patients injected with 224Ra, although in some cases there was evidence of
some blood disorders that may have been undiagnosed leukemias. It is clear from these
studies, however, that bone sarcoma is a more common result of internally deposited
radium, and that the radium leukemia risk is much lower than that calculated using ICRP
dosimetry models together with a leukemia risk coefficient derived from the LSS
weighted by an RBE of 20. As with humans, it appears that Ra isotopes induce bone
cancers but few if any leukemias. In part, the low incidence of leukemia might be
attributed to microdosimetry: i.e., target cells may be non-uniformly distributed in the
bone marrow in such a way that the dose to these cells is considerably lower than the
average marrow dose.
Evidence suggests, however, that microdosimetry is not the full story, and that
high-LET radiation is only weakly leukemogenic. Thorotrast patients, who are expected
to have a more even distribution of alpha-particle energy, do show an excess of leukemia,
but only about twice the risk per Gy as seen in the LSS (ICRP 2003). Moreover, an
RBE of only about 2.5 has been found for neutron induced leukemia in mice (Ullrich and
Preston 1987). BEIR VII low-LET risk estimate for leukemia incidence is roughly 50%
higher than that of UNSCEAR or EPA. Using a Bayesian approach, Grogan et al. (2001)
estimated the alpha-particle leukemia risk to be 2.3xlO"2 Gy"1. Based on the BEIR VII
low-LET leukemia (incidence) risk estimate, this would correspond to an RBE of
approximately 2.5. Through a comparison of Thorotrast and A-bomb survivor data,
Harrison and Muirhead (2003) also estimated the RBE to be 2-3. However, the authors
noted that the Thorotrast doses were likely to be overestimated by a factor of 2-3
(Ishikawa et al. 1999), and that the RBE was perhaps very close to 1. EPA has been
employing an RBE of 1 for leukemia; it would appear that a value of about 1-3 is still
reasonable.
Liver. The LSS study shows a statistically significant excess of liver cancer. The
uncertainty bounds derived by BEIR VII are wide, both because of the large sampling
error and the uncertainty in the population transport (liver cancer rates are about an order
of magnitude lower here than in the LSS cohort). The BEIR VII central estimate is
~ 2xlO"3/Gy. For comparison, an update on Danish Thorotrast patients (Andersson et
al. 1994) yielded an estimate of 7xlO"2 per Gy. Thus, given the large uncertainties and
difference in age and temporal distribution, these findings are reasonably consistent with
the commonly assumed default value of 20 for alpha-particle RBE. Grogan et al. (2001)
also concluded that a value of 20 (GM) with 1.6 GSD was the best estimate of an RBE
23
-------
relative to y rays based on the follow-up of the Thorotrast patients and the atomic bomb
survivors.
We conclude that the use of the liver cancer risks in BEIR VII, modified by an
RBE of 20, provides a suitable estimate of the risk alpha-particle induced liver cancer.
Uncertainty estimates will require consideration of both uncertainties in the BEIR VII
estimates and the uncertainty in RBE.
Lung. Excess lung cancers have been associated with the inhalation of alpha-
emitting radionuclides in numerous epidemiological studies. Cohort studies of
underground miners and residential case-control studies have shown a strong association
with exposure to airborne radon progeny. In addition, a cohort study of workers at the
Mayak nuclear plant, has also shown an association with inhaled plutonium (Gilbert et al.
2004). The miner studies serve as the primary basis for BEIR VI and EPA estimates of
risk from radon exposure (NRC 1999, EPA 2003), and results from the residential studies
are in reasonable agreement with those risk estimates (Darby et al. 2005, Krewski et al.
2005). The Agency has no plans at this time to reassess its estimates of risk from
exposure to radon progeny, but it is our intent to reassess estimates of risk from inhaled
plutonium and other alpha-emitters, along with the lung cancer risk associated with low-
LET exposures.
Table 7 compares summary measures of risk per unit dose for the U.S. population
derived from the LSS in BEIR VII and from the pooled underground miner studies in
BEIR VI. The RBE inferred from this comparison is much lower than what one might
project based on most animal studies. It should be recognized, however, that the risk
model used to derive risk estimates for radon are in certain ways incompatible with the
models for low-LET lung cancer risk in BEIR VII. They differ not only with respect to
age, gender, and temporal factors, but also with respect to the interaction with smoking.
In contrast to the BEIR VII models, the radon risk models do not incorporate a higher
risk coefficient for females or for children. The miner cohorts from the radon models
were derived consisted essentially entirely of adult males, and it is possible that radon
risks are being underestimated for children and females. On the other hand, results from
residential case-control studies on females are generally consistent with projections based
on the miners. The radon risk appears to be almost multiplicative with smoking risk (or
the baseline lung cancer rate), whereas the LSS data suggests an additive interaction. It is
unclear whether these apparent differences with respect to gender and smoking reflect a
real mechanistic difference in carcinogenesis by the two types of radiation exposure
(chronic alpha versus acute gamma) or unexplained errors inherent in the various studies.
24
-------
Table 7: Lung cancer mortality and RBE.
Source of Data
A-bomb mortality
EPA radon risk model
Gender
Male
Female
Combined
Male
Female
Combined
Risk per 106
Person- WLM
640
440
540 (220, 1300)3
Risk per 104
Person-Gy
140 (52, 380)2
270(110, 660)2
210
8001
3501
4301
RBE
1.0
1.0
1.0
5.7
1.3
2.0
1 Risk per Gy to the whole lung, assuming: (1) 12 mGy/WLM, on average, to sensitive cells in the
bronchial epithelium (James, Birchall & Akabani) and (2) lung risk partitioned 1/3 (bronchi): 1/3
(bronchioles): 1/3 (alveoli).
2 95% C.I. (NRC 2006)
3 90% C.I. (EPA 2003)
Table 8 compares lung cancer results from the LSS with those on Mayak workers,
whose lungs were irradiated by alpha particles emitted by inhaled plutonium. Some
issues also arise here in trying to compare ERR/Gy and EAR/Gy estimates from the two
studies. One is that the populations are quite different with respect to gender and age
profile. Males account for about 75% of the PY and over 90% of the lung cancers among
the internally exposed Mayak workers, but for only about 30% and 55% of the PY and
lung cancers, respectively, among the LSS cohort. Another is that the dependence of the
risk on attained age appears to be quite different in the two studies—a monotonically
increasing EAR for the LSS but a sharp decrease in the EAR above age 75 for the Mayak
workers. There are, however, very few data on these older Mayak workers. Focusing
just on lung cancers appearing between ages 55 and 75, one finds that the central
estimates of risk per Sv in the two studies are comparable, consistent with an RBE for
alpha particles of 10 or more.
The risk per unit dose estimate from the plutonium exposed Mayak workers is
considerably higher than that from the radon studies. This is somewhat surprising since
the lung dose from radon progeny is projected to be almost entirely to the epithelial lining
of the airways, whereas the inhaled plutonium dose would be expected to be concentrated
in the alveoli, which is generally thought to be a much less sensitive region for cancer
induction.
25
-------
Table 8
Estimated Paraawteis foi Lung COUCH Mortality Rbto witk 15s?-! Confidence IirtiTah (Cli for Inreuuil Lin|
<;e ii Mayak Woifcen and External Lung Dwe n the Life Span Study flSSi Cohort of Jap;iii«e Atomic
on, Exposed between tk Age? of 15 and SO
Total
By ax
Male,
Feole:
Ferule-mle r?,nc
Mayijt ?"e:keis LI j ccior jje 15-«' it expos-re
JsMB-yMri* Linr E1R '.si ysvffi' Persou-yisn lm| ERR j» never'
(rarreo^p caaca at cnsjiei ?p (SO- ;reic€S;,p caa:e: ar ateaei sp
cfM^'i dea:jiK (?5D:C!) of total datSu §3 (>"5L; C?i
"""052i^4i •] C.S5 ".-*: 1 ?; 1.233.2" ,6S' :03 !-»} ('."6 12i
»'j l.?:SJ,i ]d il,2 11;
E:;cess death- p«r
3y anaus.i aa
Ucdw 55 T-SS".
^"™t*l4 \g^T^
J5~"4 veir;
"5+ ">'E3:;.
Mesi s|6
:0? 332 ••:'£)
o'IPCC i]li
IP 553 :.? "i
-595 [].:'(
•8
65
15-
Ui
2~
6S
1.4 i1.'. i 2 :J
5.1 ;if . ' 11
5.2 {2.1 E Ui
1.4 (".0 12;
-«3.w.;4:; 4s : (--o c.-i»
•^2.-2- ;25. If" 1.5 i' 0 i.3,1
i"/,""S {21; 4^.0 *.2 f3,~, 11. r)
2:i.] 7-: {12: 5-17 R3.;f.5-2:-|
•:•" i
"fc;i'.idn| ooiy fffsot-^ai- sad IILLS ciacs: dssii wbae ictiriul i'icg dj-« cocld'M e-.tmssd: L0wa«r. ill wo:k«:> psd pei-Ki-yesr-
xclr.dei ;a tie aalysei
* litvsm wae calctlasd by dr.'iiiag ±6 d;.a it zrsy> >, s quakr.' factor iQ: j ef 20.
' Sjjfd :i s 21 c3*1 in Hhcii tbs c-oefcistt of "te Jc|3ri*Jn of iittjxei ap w.">>;.««q'jiL :c -2.2.
"EA?. fci fecislei wouS H a fecrsr :l ?.-" ^mslbr 'bin. :be:,? Eiiaiws
There seems to be no fully satisfactory way to reconcile all the results from the
LSS, miner and Mayak worker studies with what one would expect from the dosimetry
and experimental determinations of alpha-particle RBE, even taking into account the
sampling errors in the various epidemiological studies. The Mayak study is ongoing,
with significant improvements in the dosimetry still to be made; the LSS risk estimates
are also somewhat suspect, especially their dependence on gender and age at exposure
(see Section II.G). In particular, it appears strange that the risk is higher in females than
males among the A-bomb survivors, despite the much lower lung cancer incidence
among Japanese women than men. Also, the BEIR VII lung cancer model reflects the
negative trend with age at exposure obtained from the analysis of all solid tumors, but
there are still very little data to directly support a higher lung cancer risk for childhood
exposures.
3. Summary and recommendations
Information on alpha-particle KB EM (relative to y-irradiation) for induction of
cancer is sketchy, especially in humans. Laboratory studies are mostly indicative of a
2Q
-------
value of about 20, but with likely variability depending on cancer site and animal species
or strain. There is also evidence in both animals and humans that the RBEM is much
lower for induction of leukemia. Comparisons of data on lung cancer induction by
inhaled radon progeny or plutonium with data on the A-bomb survivors yields somewhat
conflicting results, suggesting possible errors in the data or in underlying assumptions
regarding the form of the models, internal dosimetry, or the sensitivity of different parts
of the lung. At this point, comparisons between the radon data and the LSS data suggest
an RBE much lower than 20 for lung cancer induction, but the Mayak results so far fail to
substantiate this. Further follow-up of the LSS cohort and additional information on the
Mayak workers may help to resolve this issue.
At this time, we would propose multiplying site-specific gamma ray risk estimates
by an RBE of 20 in order to derive corresponding estimates of risk from alpha radiation,
with two exceptions: (1) an RBE =1-3 for leukemia induced by alpha-emitters deposited
in bone (EPA 1992) and (2) continued use of models derived from BEIR VI to estimate
lung cancer risk from inhaled radon progeny (NAS 1999, EPA 2002). The low dose, y-
ray risk estimate for bone cancer would be obtained by dividing by 20 the risk per Gy
estimated from patients injected with 224Ra.
This approach is consistent with current EPA practice except in the case of breast
cancer, where previously an RBE of 10 was employed rather than 20. The justification
for the lower RBE was that the estimated (y ray) DDREF was 1 for breast cancer but 2
for other solid tumors. The evidence for such a difference in DDREF appears weaker
now, and, for simplicity we would propose using the same nominal DDREF (1.5) and
RBE (20) for all solid tumors, including breast.
For all cancers except leukemia, EPA previously assigned a lognormal
uncertainty distribution to the alpha-particle RBE, with a 90% CI from 5 to 40. The
median value is thus about 14 (EPA 1999a). This distribution still seems reasonable to
us. The uncertainty distribution for leukemia induced by alpha-emitters deposited in the
bone was previously taken to be uniform over the interval [0,1] (EPA 1999a). Recent
analyses of Thorotrast patient would suggest that this distribution be extended upward.
B. Lower Energy Beta Particles and Photons
As energetic electrons lose energy in passing through matter, they become more
densely ionizing: i.e., the average distance between ionization events shrinks, and more
energy is deposited in ionization clusters. As discussed earlier, such clusters can lead to
DSBs and complex DNA damage that is more difficult for the cell to repair. Indeed it has
been suggested that a large fraction of the residual damage caused by low-LET radiation
may stem from these clusters produced at the ends of electron tracks. For this reason, it
might also be expected that lower energy beta particles would be more biologically
damaging than higher energy betas. Furthermore, since the energy distribution of
secondary (Compton) electrons is shifted downward as incident photon energy is
reduced, the biological effectiveness of photons might also be expected to rise with
decreasing energy, implying that x rays might be more damaging than y rays.
27
-------
Results from many studies tend to confirm these predictions for low-LET
radiations, including measurements of chromosome aberrations, mutations, cell
transformation and cancer induction. The most direct source of data on the subject
consists of comparative studies of x and y ray induction of dicentrics in human
lymphocytes. In these studies, 220-250 kVp x rays generally produced 2-3 times as many
dicentrics as 60Co gamma rays (NCRP 1990). The relevance of such measurements for
cancer induction is unclear, however, since a dicentric will render a cell incapable of cell
division. Other laboratory studies directed at ascertaining the RBE for various types of
radiation relative to x rays or y rays provide additional indirect information, suggesting
again that x rays may be a factor of 2-3 times more hazardous than y rays (Kocher et al.
2005, NCRP 1990, NRC 2006). Kocher et al. also conclude that x rays below 30 keV
may have a slightly higher RBE than those in the range 30-250 keV.
Kocher et al. also consider what REFs should be applied to beta particles. They
note that the average energy of a Compton electron produced by an incident 250 keV
photon is 60 keV. It follows that beta particles above 60 keV should have the same RBE
as photons above 250 keV: i.e., 1.0. The only radionuclide that emits a substantial
fraction of its decay energy in the form of lower energy beta is 3H, for which the mean
beta energy is 5.7 keV. For 3H betas, the authors recommend a lognormal uncertainty
distribution with GM=2.4 and a GSD=1.4, corresponding to a 95% CI of (1.2, 5.0). The
reference radiation is again chronic y rays. This range is comparable to what was
recommended for <30 keV photons and is generally consistent with experiments in which
investigators compared 3H with y rays in the induction of various end-points.
No firm conclusions can be drawn from human epidemiological data. Risk
coefficients derived from studies of cohorts medically irradiated with x rays are generally
lower than what has been observed for the A-bomb survivors. Nevertheless, given the
various uncertainties, such as those relating to dosimetry, sampling error, and possible
confounders, it is still possible that medical x rays are significantly more carcinogenic,
per unit dose, than y rays. Based on the available evidence, it would appear reasonable to
assign an RBE of 2 for most medical x rays and an RBE of 2.5, both for x rays below 30
keV and for 3H beta particles.
28
-------
IV. Risks from Prenatal Exposures
Currently, EPA does not include radiogenic cancers induced in utero as part of its
population risk estimate. Although this contribution is expected to represent a small
fraction of the total population risk, it is of particular concern because of the potentially
higher radiation sensitivity in the embryo/fetus, and the evidence for an appreciably
higher risk of radiogenic childhood cancers than that obtained with postnatal childhood
exposures.
First carried out by Stewart and coworkers (1958), case-control studies of
childhood cancer have shown a statistically significant association with diagnostic x rays
in utero. The risk estimate derived from the Stewarts' "Oxford survey" is about 0.06 per
Gy (95% C.I. 0.01-0.126) for all cancers and about 0.025 per Gy for leukemia (Mole
1990, Doll and Wakeford 1997). The LSS cohort results show no statistically significant
excess of childhood cancers, on the other hand, and the upper bound estimate of the
leukemia risk (Doll and Wakeford 1997).
This discrepancy has led to considerable controversy. The review article by Doll
and Wakeford (1997) carefully considered many of the issues involved and concluded
that the elevated risk in the fetus is probably real. They recommend the numerical risk
estimates from the Oxford survey cited in the paragraph above. ICRP has recommended
the value 6xlO"2 Gy"1 for risk of a radiogenic cancer before the age of 14 from in utero x
rays (ICRP 2000), and NCRP cites a value of 5xlO"2 Sv"1 for fetal exposure to internally
deposited radionuclides (NCRP 1998).
One possible contributing factor to the discrepancy between the Oxford survey
and LSS data is the difference in type of radiation: x rays and gamma rays, respectively.
As discussed in another section, for a fixed absorbed dose, the former may be about twice
as effective in causing cancer. Based on this consideration, EPA proposes here a
numerical risk estimate of 6xlO"2 Gy"1 for x rays, but 3xlO"2 Gy"1 for gamma rays and beta
particles (other than from 3H). Survival rates for childhood cancer in the U.S. are
approximately 70-80% for both leukemia and solid tumors (SEER 2006c, Tables
XXVIII-10 and XXIX-6), but this does not include delayed mortality resulting from
second cancers arising from the treatment. In the Oxford study, roughly the same
ERR/Gy was observed for solid tumor incidence as for leukemia. It is then reasonable to
expect that approximately 20-30% of the radiation-induced cases would also be fatal.
Epidemiological follow-up of the atomic bomb survivors has indicated that
individuals irradiated in utero have a risk similar to that of irradiated young children
(Delongchamp et al. 1997). Based on this finding, we propose to adopt the same set of
models employed for calculating risk for exposure to young children to assess the risk of
adult cancers caused by in utero exposure.
29
-------
V. Skin Cancer Risk
Current EPA risk estimates for radiation-induced skin cancer mortality (EPA
1994) are taken from ICRP Publication 59 (ICRP 1991). The one modification made by
EPA was to apply a DDREF of 2 at low doses and dose rates. Recognizing that the great
majority of nonmelanoma skin cancers are not life threatening or seriously disfiguring,
EPA included only the fatal skin cancer cases in its estimates of cancer incidence risk.
The contribution of skin cancers to the radiogenic risk from whole-body irradiation is
then minor: about 0.2% and 0.13% of the total mortality and incidence, respectively.
ICRP's calculation of skin cancer incidence risk employed an ERR of 55% per
Sv, along with U.S. baseline skin cancer incidence rates from the 1970's. The ICRP
mortality estimate was also based on conservative assumptions that: (1) 1/6 of radiogenic
skin cancers would be squamous cell carcinomas (SCC), the remainder basal cell
carcinomas (BCC); (2) essentially all of the BCC would be curable, whereas about 1% of
SCC would be fatal. Predicated on these considerations, ICRP Publication 59 estimated
that 0.2% of the cases would be fatal.
The ICRP risk estimates closely mirror those previously published by Dr. Roy
Shore (1990), who also served as a member of the committee that drafted ICRP
Publication 59. Shore (2001) reviewed the subject again in light of additional
information and concluded that essentially all of the radiation-induced skin cancers at low
to moderate doses would be BCC. He maintains that the fatality rate for BCC is
"virtually nil" but cites a study indicating a rate of 0.05% Weinstock (1994). Shore also
notes that there is no persuasive evidence that radiation-induced BCC would be more
fatal than sporadic cases.
At the same time, there is evidence that the baseline rates for BCC have increased
dramatically since the 1970s, which might also result in a higher (absolute) risk per unit
dose of inducing a radiogenic skin cancer.
It should be possible for EPA to estimate radiogenic skin cancer incidence and
mortality using the ICRP risk model, but with a revised estimate of lethality for such
cancers. Assuming that radiogenic skin cancers induced by low level ionizing radiation
are all BCC, the fraction of cases expected to be fatal should be reduced, perhaps from
0.2% to 0.05%. An updated calculation with the ICRP relative risk model would further
require recent information on baseline BCC incidence rates, which have increased
dramatically over the last 3 decades (Karagas et al. 1999). SEER does not include
nonmelanoma skin cancers in its data base, but statistics exist on the number of new cases
each year. Together with assumptions on the variation of BCC incidence by age, it is
anticipated that an approximate table of age-specific incidence rates could be derived.
30
-------
VI. Thyroid Cancer Risk
The BEIR VII Report presents a model for estimating radiogenic thyroid cancer
incidence based on a combined analysis of seven studies of individuals receiving external
radiation (Ron et al. 1995). All seven studies were used to assess: the shape of the dose-
response relationship; the effect of gender; the influence of age at exposure; temporal
patterns of risk at times after exposure; the effects of fractionation; and the influence of
screening and clinical surveillance on risk estimates (Ron et al. 1995). A pooled analysis
of five of these studies was conducted to arrive at an excess ERR coefficient of 7.7%/Gy
for exposures under the age of 15.
BEIR VII adopts a simplified model in which the ERR depends only on whether
the exposure occurred before age 30, but not on attained age or time since exposure.
Based on the findings of Ron et al., BEIR VII also assigns twice the ERR to females as
males. Given the higher baseline thyroid cancer rate in females (and their longer life
expectancy) this translates into a radiogenic thyroid cancer risk that is five times higher
for females than males. BEIR VII provides no estimate of the fraction of radiogenic
cancers that would be fatal. It also does not deal with the issue of thyroid irradiation by
internal emitters, including the important case of 131I.
We also have available a preliminary draft of an EPA sponsored report from
NCRP on thyroid radiation risk (T. Tenforde, private communication). This report
presents several alternative risk models based primarily on the pooled analysis of Ron et
al. The recommended model in the draft report differs in some ways from that in BEIR
VII: (1) the dependence of ERR on age at exposure is more complex, falling off to zero
by age 30; (2) the ERR peaks after 15 y and then falls off continuously with time since
exposure; (3) the ERR/Gy is assumed to be equal for males and females. The net effect
of these differences is that the population risk projection in the NCRP report is about
40% lower than that in BEIR VII.
Based on SEER data, the NCRP estimates that 5-7% of the radiogenic cancers
would be fatal, and recommends adopting 7%. Our preliminary examination of the most
recent SEER data suggests that the lethality is probably lower, perhaps 3-5%.
NCRP assigns an "RBE" of 0.6 to 131I radiation relative to external x ray or
gamma ray exposure. This is meant to include both the effects of dose fractionation
(DDREF) and non-uniform dose distribution within the thyroid for 131I.
The higher ERR for females cited in BEIR VII was derived from the pooled
analysis, but the results from the individual studies were highly variable in this regard,
and the gender difference was not statistically significant. NCRP's treatment of age and
temporal dependence is more detailed. Overall, we now favor adoption of the NCRP
thyroid cancer model, assuming that we would have a proper reference that can be cited.
The mortality risk can be projected from either the NCRP or BEIR VII incidence model,
using SEER data to estimate the fraction of radiogenic thyroid cancers that are fatal.
31
-------
VII. REFERENCES
Andersson M, M Vyberg, J Visfeldt, B Carstensen and HH Storm. 1994. Primary liver
tumours among Danish patients exposed to Thorotrast. Radiat Res 137: 262-273.
Arias E. 2002. United States Life Tables, 2000. National Vital Statistics Reports, VolSl
No 3. Hyattsville, Maryland: National Center for Health Statistics.
CDC (Centers for Disease Control) and NCI (National Cancer Institute). U.S. Cancer
Statistics Working Group. 2005. United States Cancer Statistics: 2002 Incidence and
Mortality. Atlanta: U.S. Department of Health and Human Services, Centers for Disease
Control and Prevention and National Cancer Institute. Accessed 7/24/06 at CDC web-
site: http://www.cdc.gov/cancer/npcr/uscs/pdf/2002_uscs.pdf
Darby S, D Hill, A Auvinen, et al. 2005. Radon in homes and risk of lung cancer:
collaborative analysis of individual data from 13 European case-control studies. BMJ
330: 223.
Delongchamp RR, K Mabuchi, Y Yoshimoto, DL Preston. 1997. Cancer mortality among
atomic bomb survivors exposed to in utero or as young children, October 1950-May
1992. Radiat Res 147: 385-395.
Doll R and R Wakeford. 1997. Risk of childhood cancer from fetal irradiation. Brit J
Radiol 70: 130-139.
EPA (Environmental Protection Agency). 1994. Estimating Radiogenic Cancer Risks.
EPA Report 402-R -93-076, Washington, DC: U.S. EPA.
EPA (Environmental Protection Agency). 1999a. Estimating Radiogenic Cancer Risks.
Addendum: Uncertainty Analysis. EPA Report 402-R-99-003, Washington, DC: U.S.
EPA.
EPA (Environmental Protection Agency). 1999b. Cancer Risk Coefficients for
Environmental Exposure to Radionuclides. EPA Report 402-R-99-001, Washington, DC:
U.S. EPA.
EPA (Environmental Protection Agency). 2003. EPA Assessment of Risks from Radon in
Homes. EPA Report 402-R-03-003, Washington, DC: U.S. EPA.
Gilbert ES. 1991. Chapter 3: Late somatic effects. In: S Abrahamson, MA Bender, BB
Boecker et al. Health Effects Models for Nuclear Power Plant Accident Consequence
Analysis. Modifications of Models Resulting from Recent Reports on Health Effects of
Ionizing Radiation, Low LET Radiation, Part II: Scientific Bases for Health Effects
Models. NUREG/CR-4214, Rev. 1, Part II, Addendum 1, LMF-132, U.S. Nuclear
Regulatory Commission, Washington, DC.
32
-------
Gilbert ES, NA Koshurnikova, ME Sokolnikov et al. 2004. Lung cancer in Mayak
workers. RadiatRes 162: 505-516.
Grogan HA, WK Sinclair and PG Voilleque. 2001. Risks of fatal cancer from inhalation
of 239'240plutonium by humans: a combined four-method approach with uncertainty
evaluation. Health Phys 80: 447-461.
Hahn FF, BA Muggenburg, RA Guilmette and BB Boecker. 1999. Comparative
stochastic effects of inhaled alpha- and beta-particle-emitting radionuclides in beagle
dogs. RadiatRes 152: S19-S22.
Harrison JD, CR Muirhead. 2003. Quantitative comparisons of induction in humans by
internally deposited radionuclides and external radiation. Int J Radiat Biol 79:1-13.
ICRP (International Commission on Radiological Protection). 1991. 1990
Recommendations of the International Commission on Radiological Protection. ICRP
Publication 60. Ann ICRP 21(1-3).
ICRP (International Commission on Radiological Protection). 1991. The Biological Basis
for Dose Limitation in the Skin. ICRP Publication 59. Ann ICRP 22(2).
ICRP (International Commission on Radiological Protection). 2000. Pregnancy and
Medical Radiation. ICRP Publication 84. Ann ICRP 30(1).
ICRP (International Commission on Radiological Protection). 2003. Relative Biological
Effectiveness (RBE), Quality Factor (Q), and Radiation Weighting Factor (WR). ICRP
Publication 92. Ann ICRP 33(4).
Ishikawa Y, JAH Humphreys, CG Collier, ND Priest, Y Kato, T Mori and R Machinami.
1999. RadiatRes 152: S102-S106.
James AC, A Birchall and G Akabani. 2004. Comparative dosimetry of BEIR VI
revisited. Radiat Prot Dosimetry 108: 3-26.
Karagas MR, ER Greenberg, SK Spencer, TA Stukel and LA Mott. 1999. The New
Hampshire Skin Cancer Study Group. Increase in incidence rates of basal cell and
squamous cell skin cancer in New Hampshire, USA. Int J Cancer 81: 555-559.
Kellington, JP, TM Eldred, K Ambrose, PN Brooks and ND Priest. 1997. Effects of
radiation quality on lung tumour induction in CBA/Ca mice. In: Health Effects of Low
Dose Radiation: Challenges of the 21st Century. Proceedings of BNES conference held in
Stratford-upon-Avon, UK, on 11-14 May 1997. London, UK: British Nuclear Energy
Society.
Kocher, DC, AI Apostoaei and FO Hoffman. 2005. Radiation effectiveness factors for
use in calculating probability of causation of radiogenic cancers. Health Phys 89:3-32.
33
-------
Krewski D, JH Lubin, JM Zielinski et al. 2005. Residential radon and the risk of lung
cancer: A combined analysis of 7 North American case-control studies. Epidemiol 16:
1-9.
Land CE and WK Sinclair. 1991. The relative contributions of different organ sites to the
total cancer mortality associated with low-dose radiation exposure. In: Risks Associated
with Ionising Radiations, Annals of the ICRP 22 (1).
Mole RH. 1990. Childhood cancer after prenatal exposure to diagnostic X-ray
examinations in Britain. Br J Cancer 62: 152-168.
(NCHS) National Center for Health Statistics. 2004. Bridged-Race Population Estimates,
United States, 1990 - 2004, By Age Groups. Compiled from the April 1, 2000 resident
population developed by the Bureau of the Census in collaboration with the NCHS -
accessed from the CDC WONDER On-line Database on June 3, 2006.
NCRP (National Council on Radiation Protection and Measurements). 1980. Induction of
Thyroid Cancer by Ionizing Radiation. NCRP Report No. 64. Bethesda, MD: National
Council on Radiation Protection and Measurements.
NCRP (National Council on Radiation Protection and Measurements). 1990. The Relative
Effectiveness of Radiations of Different Quality. NCRP Report No. 104. Bethesda, MD:
National Council on Radiation Protection and Measurements.
NCRP (National Council on Radiation Protection and Measurements). 1998.
Radionuclide Exposure of the Embryo/Fetus. NCRP Report No. 128. Bethesda, MD:
National Council on Radiation Protection and Measurements.
NIOSH (National Institute for Occupational Safety and Health). 2002. NIOSH-
Interactive RadioEpidemiological Program (NIOSH-IREP), Technical Documentation.
Cincinnaiti, OH: National Institute for Occupational Safety and Health. Available at:
http://www.cdc.gov/niosh/ocas/pfs/irep/irepfnl.pdf. Accessed 6/06.
NRC (National Research Council). 1988. Health Effects of Radon and Other Internally
Deposited Alpha-Emitters (BEIRIV). Washington, DC: National Academy Press.
NRC (National Research Council). 1990. Health Effects of Exposure to Low Levels of
Ionizing Radiation. (BEIR V). Washington, DC: National Academy Press.
NRC (National Research Council). 1999. Health Effects of Exposure to Radon. BEIR VI.
Washington, DC: National Academy Press.
NRC (National Research Council). 2006. Health Risks from Exposure to Low Levels of
Ionizing Radiation. BEIR VIIPhase 2. Washington, DC: National Academy Press.
34
-------
Pierce DA, GB Sharp and K Mabuchi. 2003. Joint effects of radiation and smoking on
lung cancer risk among atomic bomb survivors. Radiat Res 159: 511-520.
Preston DL, A Mattsson, E Holmberg, R Shore, NG Hildreth and JD Boice Jr. 2002.
Radiation effects on breast cancer risk: a pooled analysis of eight cohorts. Radiat Res
158: 220-235.
Ries LAG, D Harkins, M Krapcho, A Mariotto, et al. (eds). 2006. SEER Cancer Statistics
Review, 1975-2003, National Cancer Institute. Bethesda, MD,
http://seer.cancer.gov/csr/1975_2003/, based on November 2005 SEER data submission,
posted to the SEER web site, 2006.
Ron E, JH Lubin, RE Shore, K Mabuchi, B Modan, L Pottern, AB Schneider, MA Tucker
and JD Boice. 1995. Thyroid cancer after exposure to external radiation: a pooled
analysis of seven studies. Radiat Res 141: 259-277.
(SEER 2006a). Surveillance, Epidemiology, and End Results (SEER) Program
(www.seer.cancer.gov) SEER*Stat Database: Incidence and Mortality - SEER 13 Regs
Public-Use, Nov 2005 Sub (1992-2003), National Cancer Institute, DCCPS, Surveillance
Research Program, Cancer Statistics Branch, released April 2006, based on the
November 2005 submission.
(SEER 2006b). Surveillance, Epidemiology, and End Results (SEER) Program
(www.seer.cancer.gov) SEER*Stat Database: Incidence - SEER 17 Regs Public-Use,
Nov 2005 Sub (2000-2003), National Cancer Institute, DCCPS, Surveillance Research
Program, Cancer Statistics Branch, released April 2006, based on the November 2005
submission.
(SEER 2006c). Surveillance, Epidemiology, and End Results (SEER) Program
(www.seer.cancer.gov). SEER Cancer Statistics Review, 1975-2003, National Cancer
Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2003/, based on November 2005
SEER data submission, posted to the SEER web site, 2006.
Shore RE. 1990. Overview of radiation-induced skin cancer in humans. Int J Radiat Biol
57:809-827.
Shore RE. 2001. Radiation-induced skin cancer in humans. Med Pediat Oncol 36:549-
554.
Stewart, A, J Webb and D Hewitt. A survey of childhood malignancies. 1958. Br Med J
1: 1495-1508.
Ullrich, RL and RJ Preston. 1987. Myeloid leukemia in male RFM mice following
irradiation with fission spectrum neutrons or gamma-rays. Radiat Res 109:165-
35
-------
Weinstock MA. 1994. Epidemiologic investigation of nonmelanoma skin cancer
mortality: the Rhode Island follow-back study. J Invest Dermatol 102:65-95.
36
------- |