Nancy G. Hildreth

Radiation Effects on Breast Cancer Risk: A Pooled Analysis of Eight Cohorts

Abstract Preston, D. L., Mattsson, A., Holmberg, E., Shore, R., Hildreth, N. G. and Boice, J. D., Jr. Radiation Effects on Breast Cancer Risk: A Pooled Analysis of Eight Cohorts. Radiat. Res. 158, 220–235 (2002). Breast cancer incidence rates after radiation exposure in eight large cohorts are described and compared. The nature of the exposures varies appreciably, ranging from a single or a small number of high-dose-rate exposures (Japanese atomic bomb survivors, U.S. acute post-partum mastitis patients, Swedish benign breast disease patients, and U.S. infants with thymic enlargement) to highly fractionated high-dose-rate exposures (two U.S. tuberculosis cohorts) and protracted low-dose-rate exposure (two Swedish skin hemangioma cohorts). There were 1,502 breast cancers among 77,527 women (about 35,000 of whom were exposed) with 1.8 million woman-years of follow-up. The excess risk depends linearly on dose with a downturn at high doses. No simple unified summary model adequately describes the excess risks in all groups. Excess risks for the thymus, tuberculosis, and atomic bomb survivor cohorts have similar temporal patterns, depending on attained age for relative risk models and on both attained age and age at exposure for excess rate models. Excess rates were similar in these cohorts, whereas, related in part to the low breast cancer background rates for Japanese women, the excess relative risk per unit dose in the bomb survivors was four times that in the tuberculosis or thymus cohorts. Excess rates were higher for the mastitis and benign breast disease cohorts. The hemangioma cohorts showed lower excess risks suggesting ameliorating dose-rate effects for protracted low-dose-rate exposures. For comparable ages at exposure (∼0.5 years), the excess risk in the hemangioma cohorts was about one-seventh that in the thymus cohort, whose members received acute high-dose-rate exposures. The results support the linearity of the radiation dose response for breast cancer, highlight the importance of age and age at exposure on the risks, and suggest a similarity in risks for acute and fractionated high-dose-rate exposures with much smaller effects from low-dose-rate protracted exposures. There is also a suggestion that women with some benign breast conditions may be at elevated risk of radiation-associated breast cancer.

The Risk of Breast Cancer after Irradiation of the Thymus in Infancy

It is well established that exposure to ionizing radiation during or after puberty increases a womans risk for breast cancer, but it is less clear whether exposure to ionizing radiation very early in life is also carcinogenic. We studied the incidence of breast cancer prospectively in a cohort of 1201 women who received x-ray treatment in infancy for an enlarged thymus gland and in their 2469 nonirradiated sisters. After an average of 36 years of follow-up, there were 22 breast cancers in the irradiated group and 12 among their sisters, yielding an adjusted rate ratio of 3.6 (95 percent confidence interval, 1.8 to 7.3). The estimated mean absorbed dose of radiation to the breast was 0.69 Gy. The first breast cancer was diagnosed 28 years after irradiation. The dose-response relation was linear (P less than 0.0001), with a relative risk of 3.48 for 1 Gy of radiation (95 percent confidence interval, 2.1 to 6.2) and an additive excess risk of 5.7 per 10(4) person-years per gray (95 percent confidence interval, 2.9 to 9.5). We conclude that exposure of the female breast to ionizing radiation in infancy increases the risk of breast cancer later in life.

Risk of extrathyroid tumors following radiation treatment in infancy for thymic enlargement.

Two thousand eight hundred and fifty-six individuals who received X-ray treatments in infancy for an enlarged thymus gland and their 5053 nonirradiated siblings have been followed prospectively since 1953 to evaluate the risk of radiation-induced neoplastic disease. The health status of the entire cohort has been ascertained periodically by mail questionnaire survey. Based on the cumulative experience of five surveys of this cohort, the irradiated group has a statistically significant increased risk for both benign and malignant extrathyroid tumors, the age-adjusted relative risks being 2.0 and 2.2, respectively. Benign tumors of the bone, nervous system, salivary gland, skin, and breast (females only) and malignant tumors of the skin and breast (females only) account for the excess incidence of extrathyroid tumors among the thymic-irradiated individuals. Although a radiation-induced excess of extrathyroid tumors was suggested in an earlier survey of this cohort, small numbers restricted attribution of this excess to specific sites. The implications of these findings are discussed. Thyroid tumors are addressed in a separate paper.

Time Variations in the Risk of Cancer Following Irradiation in Childhood

The Japanese atomic bomb survivors and three other cohorts of children exposed to radiation are analyzed, and evidence is found for a reduction in the radiation-induced relative risk of cancers other than leukemia with time following exposure. Multiplicative adjustments to the excess risk either of the form exp[-delta.(time since exposure)] or of the form [time since exposure] gamma give equivalent goodness of fit. Using the former type of adjustment an annual overall reduction of 6.9-8.6% in excess relative risk is indicated (depending on the year after which this reduction might take effect). Using the second type of multiplier an adjustment to the excess relative risk varying between [time after exposure]-2.0 and [time after exposure]-3.2 fits best overall. All these reductions are statistically significant at the 5% level. There is no significant variation by cohort, by sex, by cancer type, or by age at exposure group in the degree of annual reduction in excess relative risk. Although time-adjusted relative and absolute risk models give equivalently good fits within each cohort, there is significant variation between cohorts in the degree of increase of risk with time in the absolute risk formulation, in contrast to the lack of such heterogeneity for the relative risk formulation. It is shown that if the range of observed reductions in relative risk is assumed to operate 40 or more years after exposure in the youngest age groups, the calculated UK population risks would be reduced by 30-45% compared to those based on a constant relative risk model.

Methylene chloride exposure and birthweight in Monroe County, New York

This study examined the relationship between birthweight and exposure to emissions of methylene chloride (DCM) from manufacturing processes of the Eastman Kodak Company at Kodak Park in Rochester, Monroe County, New York. County census tracts were categorized as exposed to high, moderate, low or no DCM based on the Kodak Air Monitoring Program (KAMP) model, a theoretical dispersion model of DCM developed by Eastman Kodak Company. Birthweight and information on variables known to influence birthweight were obtained from 91,302 birth certificates of white singleton births to Monroe County residents from 1976 to 1987. No significant adverse effects of exposure to DCM on birthweight were found. Adjusted birthweight in high exposure census tracts was 18.7 g less than in areas with no exposure (95% confidence interval for the difference between high and no exposure - 51.6, 14.2 g). Problems inherent in the method of estimation of exposure, which may decrease power or bias the results, are discussed. Better methods to estimate exposure to emissions from multiple industrial point sources are needed.

The influence of potential biases on the risk of breast tumors among women who received radiotherapy for acute postpartum mastitis

Data from a 30-year follow-up study of 606 women given radiotherapy for acute postpartum mastitis and two non-irradiated control groups were reanalyzed to determine whether bias could account for the reported increased risk of benign and malignant breast tumors in the irradiated population. The biases examined were comparison group selection (i.e. selection bias); differential medical verification of breast tumors between exposed and unexposed groups; differential detection of breast tumors; and confounding. Overall age-adjusted relative risk (RR) estimates for breast cancer (RR = 2.0; 95% CL:1.3-3.1) and for benign neoplastic (RR = 3.5; 95% CL:1.8-6.8) and non-neoplastic breast tumors (RR = 1.5; 95% CL: 1.2-2.1) were compared to those obtained after adjustment for the above biases. The increased risk for breast cancer among the irradiated women could not be accounted for by any of the biases examined, supporting previous reports based on this cohort. However, the suggested association of radiation with benign breast tumors may represent a selection bias since the RR estimate for non-neoplastic breast tumors was no longer statistically significant (RR = 1.2; 95% CL:0.8-1.6) while that for neoplastic breast lesions declined to 2.0 (95% CL:1.1-1-4.0) when non-irradiated women with acute postpartum mastitis were used as the comparison group instead of the non-irradiated female siblings of the exposed. This finding suggests that acute postpartum mastitis may predispose a woman to later developing a benign breast tumor irrespective of whether or not radiotherapy was used to treat this condition.(ABSTRACT TRUNCATED AT 250 WORDS)

Corrections to the paper "Fitting the Armitage-Doll model to radiation-exposed cohorts and implications for population cancer risks.

A recent paper analyzed patterns of cancer in the Japanese atomic bomb survivors and three other groups exposed to radiation by fitting the so-called multistage model of Armitage and Doll. The paper concluded that the incidence of solid cancer could be described adequately by a model in which up to two stages affected by radiation were assumed but that the data for leukemia within the bomb survivors might not be so well fitted. This was in part because of a failure to account for the observed linear-quadratic dose response that has been observed in the Japanese cohort. It has recently come to our attention that there was a mistake in the fits of the model with two adjacent radiation-affected stages, whereby the quadratic coefficient in dose was being set to zero in all the fits. This paper provides corrections in the calculations for the model and discusses the results.

Lymphocyte subclasses and immunoglobulins in adults receiving radiation treatment in infancy for thymic enlargement.

One-hundred and fifty-three subjects who were irradiated in infancy for an alleged enlarged thymus and 51 controls were studied to quantify the phenotypically different subpopulations of peripheral blood lymphocytes and immunoglobulins. These subjects were selected from a larger cohort which has been followed prospectively since the early 1950s. Previous studies of this cohort have shown a radiation dose-related increased risk of thyroid carcinomas and adenomas in the irradiated group as well as a recent excess of various extrathyroid neoplasms. In this substudy, no significant differences in lymphocyte subpopulations or immunoglobulin levels were observed between the irradiated and control groups. These findings support a hypothesis that the radiation-related excess of neoplasms in this cohort may have resulted from direct cellular damage with subsequent mutations rather than impaired immune function.