2
EPIDEMIOLOGIC INVESTIGATIONS
This chapter considers epidemiologic evidence that has accumulated since the BEIR V report (NRC 1990). The first section identifies epidemiologic studies of low-LET ionizing radiation that have appeared since the BEIR V report and summarizes their results. The second section discusses recent developments in methodology, and the third the relevance of new data to the possible utility and function of a BEIR VII phase-2 committee. The final section discusses the radioepidemiologic tables.
NEW EPIDEMIOLOGIC RESULTS
Table 1 summarizes some of the more important epidemiologic data that have been published since 1990, when the BEIR V report appeared, or that were not available to the BEIR V committee. The table includes studies that the current committee expects to produce new and useful epidemiologic data during the term of a BEIR VII phase-2 committee. Table 1 is intended not to be exhaustive, but rather to be a guide to the new epidemiologic data that have become available since 1990, and are expected to become available in the next few years.
Table 1. Summary of epidemiologic studies of low LET ionizing radiation and cancer since 19901
STUDY |
REFERENCE |
TYPE OF STUDY |
SERIES |
SEX |
NO. IN STUDY |
FOLLOW-UP PERIOD |
CANCER SITES REPORTED |
Ankylosing spondylitis patients |
Weiss and others, 1994 Weiss and others, 1995 |
Cohort Cohort |
Mortality Mortality |
Male and Female Male and Female |
15,577 14,767 |
1935-1992 1935-1992 |
All cancer and multiple cancer sites Leukemia |
Atomic-bomb survivors |
Preston and others, 1994 Thompson and others, 1994 Ron and others, 1995a |
Cohort Cohort Cohort |
Incidence Incidence Incidence |
Male and Female Male and Female Male and Female |
93,696 79,972 80,311 |
1950-1987 1958-1987 1958-1989 |
Leukemia, lymphoma, multiple myeloma Multiple cancer sites (solid tumors) Benign tumors of stomach, colon, and rectum |
Atomic-bomb survivors |
Pierce and others, 1996 |
Cohort |
Mortality |
Male and Female |
86,572 |
1950-1990 |
Non leukemias, leukemia, and multiple cancer sites |
Atomic-bomb survivors |
Land and others, 1994a Land and others, 1994b |
Case control |
|
Female |
Cases: 196 Controls: 566 |
1955-1981 |
Breast cancer |
Atomic-bomb survivors (in utero cohorts) Canadian fluoroscopy |
Delongchamp and others, 1997 Howe, 1995 Howe and McLaughlin, 1996 |
Cohort Cohort Cohort |
Mortality Mortality Mortality |
Male and Female Male and Female Female |
17,601 64,172 31,917 |
1950-1992 1950-1987 1950-1987 |
Non leukemias, leukemia, and multiple cancer sites Lung cancer Breast cancer |
Cervical cancer patients |
Kleinerman and others, 1995 |
Cohort |
Incidence |
Female |
86,193 |
1935-1990 |
Multiple cancer sites |
Contralateral breast (Denmark) |
Storm and others, 1992 |
Case control in cohort |
|
Female |
Cohort: 56,540 Cases: 691 Controls: 691 |
1943-1986 |
Breast cancer |
STUDY |
REFERENCE |
TYPE OF STUDY |
SERIES |
SEX |
NO. IN STUDY |
FOLLOW -UP PERIOD |
CANCER SITES REPORTED |
Contralateral breast (US) |
Boice and others, 1992 |
Case control in cohort |
|
Female |
Cohort: 4,109 Cases: 655 Controls: 1,189 |
1935-1987 |
Breast cancer |
Fallout from Nevada Test Site |
Kerber and others, 1993 Simon and others, 1995 |
Cohort Case control |
Incidence |
Male and Female Male and Female |
2,473 Cases: 1,177 Controls: 5,330 |
1965-1986 1952-1981 |
Thyroid cancer and other thyroid disease Leukemia |
Massachusetts fluoroscopy |
Davis and others, 1989 Boice and others, 1991 |
Cohort Cohort |
Mortality Incidence |
Male and Female Female |
13,385 4,940 |
1929-1986 1925-1986 |
Multiple cancer sites Breast cancer |
Multiple diagnostic xrays of scoliosis patients |
Hoffman and others, 1989 |
Cohort |
Incidence |
Female |
1,030 |
1935-1986 |
Breast cancer |
Nuclear industry workers (combined analysis) |
Cardis and others 1994 Cardis and others, 1995 |
Cohort Cohort |
Mortality Mortality |
Male and Female Male and Female |
95,673 95,673 |
1943-1988 1943-1988 |
Multiple cancer sites Solid tumors and leukemia |
Nuclear workers at Mayak Production Association |
Koshurnikova and Shilnikova, 1996 |
Cohort |
Mortality |
Male and Female |
18,879 |
1948-1993 |
Lung cancer and leukemia |
Pelvic radiotherapy for benign gynecologic disease |
Inskip and others, 1993 |
Cohort |
Mortality |
Female |
12,955 |
1929-1985 |
Multiple hematopoietic cancers |
Pooled analysis of external radiation and thyroid cancer |
Ron and others, 1995b |
Cohort Case control |
Incidence |
Male and Female |
120,000 |
1926-1990 |
Thyroid cancer |
Radiation treatment for benign head and neck conditions (benign thyroid tumors) |
Wong and others, 1996 |
Cohort |
Incidence |
Male and Female |
544 |
1939-1991 |
Benign thyroid nodules |
Radiation treatment for |
Schneider and others, 1993 |
Cohort |
Incidence |
Male and |
4,296 |
1939-1990 |
Thyroid cancer and |
STUDY |
REFERENCE |
TYPE OF STUDY |
SERIES |
SEX |
NO. IN STUDY |
FOLLOW-UP PERIOD |
CANCER SITES REPORTED |
benign head and neck conditions (thyroid cancer and thyroid nodules) |
|
|
|
Female |
|
|
nodules |
Radiation treatment for breast cancer |
Curtis and others, 1992 |
Case control in cohort |
|
Female |
Cohort: 82,700 Cases: 90 Controls: 264 |
1973-1985 |
Leukemia |
Radiation treatment for peptic ulcer |
Griem and others, 1994 |
Cohort |
Mortality |
Male and Female |
3,609 |
1937-1985 |
Multiple cancer sites |
Radiotherapy for Hodgkin disease (breast cancer) |
Hancock and others, 1993 |
Cohort |
Incidence and Mortality |
Female |
885 |
1961-1990 |
Breast cancer |
Radiotherapy for Hodgkin Disease (gastrointestinal cancer) |
Birdwell and others, 1997 |
Cohort |
Incidence and Mortality |
Male and Female |
2,441 |
1961-1993 |
Multiple cancer sites (gastrointestinal only) |
Radiotherapy for metropathia hemorrhagic anemia |
Darby and others, 1994 |
Cohort |
Mortality |
Female |
2,067 |
1940-1991 |
Multiple cancer sites |
Radiotherapy for pituitary adenoma |
Brada and others, 1992 |
Cohort |
Incidence |
Male and Female |
334 |
1962-1986 |
Multiple cancer sites (solid tumors only) |
Radiotherapy for skin, hemangioma in childhood |
Furst and others, 1990 |
Case control in cohort |
|
Male and Female |
Cohort: 14,647 Cases: 94 Controls: 359 |
1920-1986 |
Multiple cancer sites (solid tumors) |
Radiotherapy for thymus enlargement |
Shore, 1990 |
Cohort |
Incidence |
Male and Female |
7,450 |
1953-1989 |
Skin cancer |
Radiotherapy for uterine bleeding |
Inskip and others, 1990 |
Cohort |
Mortality |
Female |
4,153 |
1925-1984 |
Multiple cancer sites |
Tinea capitis (Israel) |
Ron and others, 1989 Ron and others, 1991 |
Cohort Cohort |
Incidence Incidence |
Male and Female Male and Female |
10,834 27,060 |
1950-1986 1950-1980 |
Thyroid cancer and other thyroid disease Melanoma, other skin cancer and benign skin tumors |
STUDY |
REFERENCE |
TYPE OF STUDY |
SERIES |
SEX |
NO. IN STUDY |
FOLLOW-UP PERIOD |
CANCER SITES REPORTED |
Women treated for infertility |
Ron and others, 1994 |
Cohort |
Mortality |
Female |
816 |
1925-1991 |
Multiple cancer sites |
STUDY |
REFERENCE |
DESCRIPTION |
In utero exposure |
Doll and Wakeford 1997 |
A review of case-control and cohort studies of childhood cancers. |
1 Table 1 is a summary of the more important epidemiologic data that have been published since the 1990 publication of the BEIR V report or that are expected to provide new and useful data during the 3-year term of the proposed BEIR VII phase-2 study. Although not exhaustive, the list should serve as a guide to some of the pertinent new and upcoming epidemiologic data on the subject. |
The following list presents categories where additional data have become available since the BEIR V report.
-
Nonleukemia cancer mortality. In a recent mortality update from the Japanese atomic-bomb survivor Life Span Study cohort, Pierce and others (1996) modeled mortality to the end of 1990. This extension of the existing data added 10,500 persons to the cohort with DS86 doses and 1,227 nonleukemia cancers to the mortality data. The increase in the number of cancer deaths for analysis was particularly noticeable among the members of the cohort who were under the age of 20 years at the time of the atomic bombings; in this category the number of deaths increased from 545 to 889 in the most recent 5-yr period of follow-up. The primary focus of this analysis was on modeling the risk of the nonleukemia cancers as a single entity, in that the authors concluded that the apparent variation in site-specific cancer risks could often not be distinguished from random variation. With that approach, the preferred risk model was a linear excess-relative-risk model; the excess relative risk (ERR) per sievert was lower for men (0.375) than for women (0.774) and was reduced with age at exposure by the same exponential factor for men and women.
-
Mortality in the British series of patients treated with x-rays for ankylosing spondylitis. This data has been updated (Weiss and others 1994, 1995).
-
Mortality among radiation workers. A combined analysis of risk estimates can be compared with those obtained at higher doses from other series (Cardis and others 1994, 1995).
-
Site-specific analyses:
-
Leukemia: includes a downwind study (Preston and others 1994; Weiss and others 1994; Simon and others 1995).
-
Breast cancer: possibility that sensitive groups might show up in form of high excess relative risk for early-onset cancer, new evidence on risk of exposure to radiation for various reproductive histories, and new data on transfer of risk between populations with different baseline risks (Land and others 1994a,b; Tokunaga and others 1994; UNSCEAR 1994; Land 1995a).
-
Lung: Cancer evidence relating to dose and dose rate effectiveness factor (DDREF) (Howe 1995).
-
Gastrointestinal cancers: longer follow-up periods by studies such as Birdwell and others (1997).
-
Lymphatic and hematopoietic cancers other than leukemia (Thompson and others 1994).
-
Lung, salivary gland, skin, and central nervous system cancers: evidence of specificity of radiation-related risk in terms of histologic subtype (Land and others 1993, 1996; Land 1995b; and data from the Radiation Effects Research Foundation).
-
Thyroid cancer (including that caused by Iodine-131): combined analysis of childhood exposure to x rays and gamma rays (Kerber and others 1993; Ron and others 1995b); given the recent National Cancer Institute report (National Cancer Institute 1997) estimating thyroid doses to the US population from Iodine-131 in fallout from the Nevada Test Site, BEIR VII phase-2 will be expected to address the issue of thyroid cancers induced by Iodine-131.
-
Other cancers (including atomic-bomb incidence series not addressed above).
-
Noncancer outcomes.
-
-
New data on radiation-related risk in patients known to be genetically susceptible to cancer:
-
Retinoblastoma patients (Tucker and others 1987; Eng and others 1993; Hawkins and others 1996; Wong and others 1997): evidence that ERR for bone sarcoma (Tucker and others 1987) and bone sarcoma and soft-tissue sarcoma (Wong and others 1997) increases with increasing therapeutic radiation dose to the tumor site with dose-specific relative risks comparable with those in survivors of other childhood cancers treated with radiation. Given evidence that baseline rates of bone and soft-tissue sarcoma are orders of magnitude higher among survivors of heritable retinoblastoma, this suggests that the excess rate (or absolute risk) of radiation-related cancer is also orders of magnitude higher among heritable retinoblastoma patients, and that, therefore, these patients constitute a genetic subpopulation highly susceptible to radiation-related bone and soft-tissue sarcoma.
-
-
Swift and others (1991) hypothesis regarding the protein mutated in ataxia telangiectasia (ATM) and breast cancer and increased susceptibility to radiation-related breast cancer.
-
International Commission on Radiological Protection study group on genetic susceptibility to radiation-related cancer (Cox and others in press).
ADVANCES IN METHODOLOGY
In addition to new data, there have been advancements reported for analytical methods including:
-
Adjustment for bias due to random errors in dosimetry (Pierce and others 1990; Gilbert 1998).
-
Systematic presentation of sources of uncertainty in various components of risk estimates and their combined influence (NCRP 1997).
CONSIDERATIONS FOR A BEIR VII PHASE-2 COMMITTEE
From the epidemiologic point of view, the prime motivation for a BEIR VII phase-2 study is the substantial increase in the mount of epidemiologic data that have been published since the BEIR V report. That applies particularly to some subjects on which data have previously been sparse, for example, cancer mortality in those exposed as children to whole-body irradiation. The new data permit the development of richer risk models and alternatives to models presented in the BEIR V report. Furthermore, there have been methodologic developments, such as the incorporation of dose measurement errors in fitting risk models.
The primary purpose of a BEIR VII phase-2 study would be to present a balanced overview of the new epidemiologic evidence and in particular to synthesize results from all the relevant studies, giving appropriate weight to the value of each study.
The committee could develop a generalized strategy for risk modeling and illustrate it with specific examples. Ideally, the strategy would be applied to all relevant exposure circumstances and outcomes; if this task were too onerous, the committee could at least develop a generalized approach that could be applied by others to other relevant situations.
In a general strategy for modeling, models should provide a good fit to the empirical epidemiologic data, be biologically plausible, be readily understood by the scientific community in general (which argues in favor of simple, rather than complex, models), and take into account all the relevant epidemiologic and biologic data.
Obviously some specific issues would have to be considered in such models, including the dose and dose rate effectiveness factor (DDREF) and the shape of the dose-response curve, the temporal distribution of risk after exposure, and the interaction of radiation with other risk factors and with other possible modifying factors, such as sex, age at exposure, attained age, and population differences.
Approaches to the modeling process could include:
-
Fitting of purely empirical models to original data from studies or combined studies.
-
Fitting of purely empirical models with meta-analysis; this is relatively underdeveloped and might be particularly useful when there are a number of studies of a particular outcome such as esophageal cancer.
-
Fitting semiempirical biologically based models to epidemiologic data to improve understanding of the biologic basis of some of the empirical effects observed.
-
Fitting (and testing) of simple models now being used in radiation protection, such as linear nonthreshold models in which the estimated relative risk at 1 sievert might depend upon age at exposure but remains invarient over time after exposure (with a minimal latent period) or an otherwise similar quadratic (linear-quadratic) model with an appropriate DDREF and particular attention given to the principal contending alternatives. Such alternatives include hormesis, threshold models, the Kellerer-Barclay model, and supralinear models.
Committee members will be selected who will be able to access original data from completed or ongoing studies, or who will be able to directly contact the original investigators.
RADIOEPIDEMIOLOGIC TABLES
The NIH radioepidemiologic tables, mandated by Congress, were developed to meet a perceived need for an objective way to present and evaluate compensation claims for adverse health outcomes, such as cancer, that might be related to radiation exposure. The concept is simple: given a documented history of exposure to radiation d1, . . ., dk at ages a1, . . ., ak and a cancer diagnosis at age A, compute the ERR of a cancer at that age. The ''probability of causation" (NIH 1985), or "assigned share" (NRC 1984), computed as ERR/(1 + ERR), is an informed quantitative estimate of the proportion of similar cancers at that age, in a large population of similar people with similar exposure histories that, would not have occurred in the absence of exposure, that is, the proportion of such cancers attributable to radiation. The ERR might depend on exposure history and age at diagnosis, but also on sex, time from each exposure until diagnosis, history of exposure to other carcinogens (such as tobacco), and other risk modifiers (such as reproductive history). Thus, all relevant factors known to influence radiation-related risk can be incorporated, as can various sources of uncertainty.