Breast cancer in women with x-ray exposure: models of dose, time, and host susceptibility

Acute postpartum mastitis (APM) is an inflammatory/infectious condition of the breast, occurring commonly at childbirth or during lactation. A series of 600 women who received x-ray therapy for APM during the 1940s or 1950s have been followed up by mail questionnaire, with medical verification of pertinent conditions, to ascertain their incidence of breast cancer. The groups have been followed for up to 45 years, with an average of 29 years. The relative risk of breast cancer, adjusted for age and interval since irradiation (or an equivalent entry definition for controls) was 3.2 for the irradiated breasts. The dose-response curve appeared to be essentially linear, except for a diminution of risk at high doses (≥ 700 rad). The fact that there were no breasts with doses between zero and 50 rad, however, means it was not possible to evaluate the curvature with the maximum contrast between low and high doses. The dose fractionation analyses showed that the number of dose fractions or the number of days between fractions had no apparent effect upon breast cancer risk, but there was a suggestion that lower doses per fraction led to a higher risk, which runs counter to what one would expect based on radiobiological theory. However, a Cox regression analysis, controlling for total breast dose, did not yield a significant effect for any of the fractionation variables