On ionising radiation and breast cancer risk

Abstract

Background: It is well known that exposure to ionising radiation increases the risk for breast cancer occurrence. However, better understanding of breast cancer risks in terms of dose-response relationships formulated according to radiobiological target theories, age-at-exposure patterns and temporal dependence are important for the understanding of the underlying biological mechanism of radiation carcinogenesis as well as for the concern about possible risks associated with the very low doses received from routine mammographic screening.

Methods: Paper I, II, and IV regard a cohort of 3,090 women with clinical diagnosis of benign breast disease (BBD). Of these, 1,216 were treated with radiation therapy during 1925-54 (median age 40 years). The mean dose to the breasts was 5.8 Gy (range 0-50 Gy). Among other organs the lung received the highest scattered dose (0.75 Gy; range 0.004-8.98 Gy) and the rectum the lowest (0.008 Gy; range 0-0.06 Gy). In paper III a pooled analysis of eight breast cancer incidence cohorts was done. Included were tumour registry data on breast cancer incidence among women in the Life Span Study cohort of atomic bomb survivors; women in Massachusetts who received repeated chest fluoroscopies during lung collapse treatment for tuberculosis; women who received x-ray therapy for acute post-partum mastitis; women who were irradiated in infancy for enlarged thymus glands; two Swedish cohorts of women who received radiation treatments during infancy for skin hemangioma; and the BBD cohort. Together the cohorts included almost 78,000 women (~35,000 were exposed), around 1.8 million woman-years and 1500 cases. The breast cancer incidence rate as a function of breast dose was analysed using linear-quadratic Poisson regression models. Cell-killing effects and other modifying effects were incorporated through additional log-linear terms. Additive (EAR) and multiplicative (ERR) models were compared in estimating the age-at-exposure patterns and time related excess. Paper V evaluates the carcinogenic risks associated with radiation in marnmographic mass screening. Assessment was in terms of breast cancer mortality and years of life. Effects were related to rates not influenced by a marnmographic mass screening program and based on a hypothetical cohort of 100,000 40-year old women with no history of breast cancer being followed to 100 years of age. Two radiation risk assumptions were compared.

Results and conclusions: Breast cancer risk the dose-response relationship is linear with little support in data for an upward curvature at low to medium doses. The competing effect of cell killing is obvious at higher doses (5-10 Gy) leading to a flattening or downward curvature. Fractionation of dose does not seem to modify risk. Protracted exposures may carry lower risks than fractionated or acute exposures. Women with benign breast diseases is more sensitive compared to women with normal breasts. Age-at-exposure is a strong modifier of excess additive risk. For excess relative risk with adjustment for attained age, only the BBD-cohort shows strong dependence on age-at-exposure. Even if risk decreases with age-at-exposure, ages over 40 years at exposure carry increased risk. EAR models provides simpler description of excess risks over populations with different background rates. In EAR-models modification-terms are needed to describe the increasing excess risk by attained age. Radiogenic breast cancers occur at same ages as non-radiogenic breast cancers occur. Time from exposure to occurrence is inversely related to age-at-exposure. The excess absolute risk pattern by attained age is similar to background rates pattern. Breast cancer risks are raised throughout life. Other cancers: In the BBD) study cancer risks in other organs than the breast was studied as well. Results indicated that scattered doses from breast irradiation may increase the risk of cancer from other sites but the small number of cases in different locations precludes strong interpretation. Mammographic mass screening: radiation risk is not, under careful consideration of dose, a crucial factor for the endorsement of a mammographic screening program of women from 40 years of age.