Abstract:
We hypothesized that women carrying particular polymorphisms in DNA double strand break repair (DSB) genes would be more susceptible to breast cancer, particularly when exposed to ionizing radiation, a DNA DSB inducing agent. We examined the association between 12 DNA DSB repair single nucleotide polymorphisms (SNPs) and breast cancer risk among female breast cancer cases (n= 858) and controls (n= 1083) selected from the United States Radiologic Technologist cohort; in multivariate analyses, we found no statistically significant associations between these SNPs and overall breast cancer risk (p < 0.05). We also evaluated potential effect modification of radiation and breast cancer risk by genotype, using exposure information generated through a major dosimetry reconstruction effort for this population that provided individual annual breast dose estimates. In multiplicative models, NBS1 P672P (rs1061302) and XRCC3 Ex2+2A>G (rs1799794) had marginally statistically significant protective effects on the radiation-breast cancer dose response relationship (p=0.1). A larger sample size and refinement of the dosimetery may allow for more conclusive results from the breast cancer effect modification analysis.
To corroborate the dose estimates from the dosimetry reconstruction effort, we analyzed the dose-translocation frequency (as measured by fluorescence in situ hybridization) relationship among a group of radiologic technologist (n=149) whose work history spanned the work history of the USRT cohort; a significantly positive relationship would indicated that the assumptions used in the dose reconstruction effort were defensible. In multivariate analyses, a one cGy increase in bone marrow dose was found to be associated with an increase of 0.1 translocations/100 cell equivalents; this association was not statistically significant (p=0.2). Given that the point estimate was greater than the expected 0.06 translocations/100 cell equivalents and that total number of years worked was statistically significantly associated with translocation frequency, there may be issues of non-differential misclassification and underestimation of doses. The variability in translocation frequencies and uncertainties in the dose estimates precluded arriving at a strong conclusion regarding the validity of the dose reconstruction.