Breast Cancer Among Women Given X-Ray Therapy for Acute Postpartum Mastitis2

Abstract

Acute postpartum mastitis (APM) is an inflammatory-infectious condition of the breast, occurring commonly at childbirth or during lactation. A series of 601 women who received x-ray therapy for APM during the 1940's or 1950's have been followed up by mail questionnaire, with medical verification of pertinent conditions, to ascertain their incidences of breast cancer. Control subjects consisted of a series with APM who did not receive irradiation, plus the female siblings of both the APM groups, for a total of 1,239 controls. The groups have been followed up to 45 years; the average was 29 years. The relative risk (RR) for breast cancer, adjusted for age and interval since irradiation (or an equivalent entry definition for controls), was 3.2 for the irradiated breasts; the 90% confidence interval (CI) was 2.3-4.3. For a linear multiplicative model, the risk increased by 0.4% per rad (90% Cl of 0.2-0.7). The dose-response curve appeared to be essentially linear, except for a diminution of risk at high doses (greater than or equal to 700 rad). The fact that there were no treated breasts with doses between 0 and 60 rad, however, means that it was not possible to evaluate the curvature with the maximum contrast between low and high doses. The dose fractionation analyses showed that neither the number of dose fractions, the number of days between fractions, nor the dose per fraction had any apparent effect on breast cancer risk when the variables were analyzed separately. Similarly, when the fractionation variables were considered jointly in a Cox regression analysis, none was significant once total breast dose was controlled for. Analyses of age at irradiation did not show appreciable differences between age groups, although the numbers were too small to be clear-cut (only 64 women greater than 34 yr old at irradiation). Other studies have shown diminished risk associated with an older age at irradiation. The lack of diminished risk in this study may occur because during pregnancy and lactation the breasts are under increased proliferative stimulation by hormones, by comparison with the normal condition of breasts at older ages. An analysis of the temporal relationship of radiation to breast cancer showed that the RR did not vary systematically with interval since irradiation, but the absolute risk increased over time. This finding agrees with other studies that have also suggested a better fit for the multiplicative model.

Dose- and time-response for breast cancer risk after radiation therapy for benign breast disease

“…The working hypothesis of the mechanisms for carcinogenesis is that it is of single-cell origin (UNSCEAR, 1993). Also, the conclusion of a linear effect in the low-dose region accords with most other breast cancer studies in this field (Land et al, 1980;Shore et al, 1986;Miller et al, 1989;Boice et al, 1991;Tokunaga et al, 1994). At dose levels above 5 Gy a cell-killing effect became obvious (Figure 1).…”

“…This effect was best described by a loglinear dose term. Cell-killing effects have been documented in some other studies, as for example in the New York Mastitis Study (Shore et al, 1986), in which a significant downward curvature was observed at doses > 3 Gy. Other studies have not found this effect, but they have generally little information at high dose levels (Miller et al, 1989;Boice et al, 1991;Tokunaga et al, 1994).…”

“…Consequently, this report does not support the BEIR-V report (1990), which proposed a curvature with time for the ERR. Instead it accords with the latest RERF report (Thompson et al, 1994) and the latest published breast cancer studies (Shore et al, 1986;Boice et al, 1991;Tokunaga et al, 1994) except for the Canadian Fluoroscopy Study (Miller et al, 1989). However, the Canadian study was based on mortality, which might be a less sensitive indicator of radiation effects than incidence.…”

“…The analysis, which was based on 278 breast cancer cases, showed a linear dose-response relationship at low to medium dose levels with a cell-killing effect of 5% Gy-' (95% confidence interval 2-9%). For a given absorbed dose and age at first exposure the time-related excess was proportional to the background rates with a suggestion that the excess remains throughout life.Keywords: ionising radiation; breast cancer; cohort study; dose-response; time-response It is well known that exposure of the breast to ionising radiation increases the risk of subsequent breast cancer, especially among young women (Baral et al, 1977; Land et al., 1980;Shore et al, 1986;Hildreth et al, 1989;Miller et al, 1989;Boice et al, 1991;Tokunaga et al, 1994). According to radiobiological target theories, the cancer risk is expected to increase approximately linearly at low doses with an upward curvature at medium dose levels (UNSCEAR, 1993).…”

High prevalence of activating ret proto-oncogene rearrangements, in thyroid tumors from patients who had received external radiation

“…Since then, an increased incidence of thyroid tumors has been observed in populations including atomic bomb survivors (Wood et al, 1969), inhabitants of regions a ected by a thermonuclear test (Conard et al, 1970), patients with a history of external radiation for benign or malignant conditions (Shore et al, 1985), and most recently, in children contaminated in Ukraine and Belarus as a consequence of the Chernobyl accident (Kazakov et al, 1992). In the latter case, most of the tumors appeared after a short latent period, and were papillary carcinomas which were assumed to be the consequence of the contamination of the exposed children by radioactive iodine isotopes.…”