Modifiers of radiation effects on breast cancer incidence revealed by a reanalysis of archival data of rat experiments. (7th January 2023)

• Record Type: Journal Article
• Authors: Imaoka, Tatsuhiko; Nishimura, Mayumi; Daino, Kazuhiro; Kakinuma, Shizuko
• Is Part Of: Journal of radiation research. Volume 64:Number 2(2023)
• Page Start: 273
• Page End: 283
• Publication Date: 2023-01-07
• Subjects: animal experiment -- effect modifiers -- individual response -- interaction -- risk model -- data archive; Radiology, Medical -- Periodicals; Radiobiology -- Periodicals; Radiation -- Periodicals; 616.0757
• DOI: 10.1093/jrr/rrac090
• Languages: English
• ISSNs: 0449-3060

Abstract

Abstract: Cancer risk after exposure to ionizing radiation can vary between individuals and populations, but the impact of factors governing those variations is not well understood. We previously conducted a series of carcinogenesis experiments using a rat model of breast cancer, in which 1654 rats born in 2002–2012 were exposed to γ rays at various doses and ages with or without non-radiation factors including high-fat diet, parity and chemical carcinogens. We herein reanalyze the incidence data from these archival experiments to clarify the effect of age at exposure, attained age, radiation dose and non-radiation factors (i.e. fat, parity, chemicals and birth cohorts) on radiation-related mammary cancer incidence. The analysis used excess relative risk (ERR) and excess absolute risk (EAR) models as well as generalized interaction models. Age-at-exposure dependence displayed a peak of susceptibility at puberty in both the ERR and EAR models. Attained age decreased ERR and increased EAR per unit radiation dose. The dose response was concordant with a linear model. Dietary fat exhibited a supra-multiplicative interaction, chemicals represented a multiplicative interaction, and parity and birth cohorts displayed interactions that did not significantly depart from additivity or multiplicativity. Treated as one entity, the four non-radiation factors gave a multiplicative interaction, but separation of the four factors significantly improved the fit of the model. Thus, the present study supports age and dose dependence observed in epidemiology, indicates heterogenous interactions between radiation and various non-radiation factors, and suggests the potential use of more flexible interaction modeling in radiological protection.