Skip to Content

Molecular Epidemiology of Radiation-Associated Cancers

Several nested case-control studies of breast and thyroid cancer have been completed within the U.S. Radiologic Technologist cohort. Because DNA repair, apoptotic, inflammatory, and hormone metabolic pathways are important in radiation carcinogenesis and are involved in breast cancer specifically, we evaluated potential modification of the relationship between ionizing radiation exposure and breast cancer risk by polymorphic variants in related genes. We found that the dose-response relationship between either occupational dose or cumulative personal diagnostic radiation exposure and breast cancer risk was significantly modified by specific variants in WRN, BRCA1, PRKDC, ERCC5, ERCC2, IL1A, PTGS2, and CYP1B1 after adjustment for relevant co-exposures. Evaluation of known breast cancer risk variants discovered in genome wide scans and ionizing radiation exposure on breast cancer risk revealed interaction with variants in the H19 and MRPS30 genes.

For thyroid cancer, we evaluated approximately 29,000 variants in candidate genes and pathways for their association with thyroid cancer risk. These genes or pathways included maintenance of genomic integrity, hormone metabolism, the 8q24 region, detoxification and Phase I and II metabolism, body mass index and obesity, and immunity. We found suggestive associations for variants in several genes including HDAC4, CYP19A1, FTO, and SERPINA5. We found an association with variants in HEMGN, a gene near the region where known thyroid cancer susceptibility loci are located. Several known somatic mutations including BRAF and RAS and rearrangements in RET/PTC were evaluated between the early 1970’s to the present using standardized pathology criteria. The prevalence of BRAF mutations and RET/PTC rearrangements were unchanged but an increase in RAS point mutations suggested a change in an unidentified environmental exposure.

Follow-up screening studies of persons exposed to radioactive fallout from the Chernobyl accident before the age of 18 have established a strong relationship between childhood exposure to iodine-131 and subsequent risk of thyroid cancer. This provides an excellent setting to study the biological basis of radiation carcinogenesis in humans. In our recent study of genome-wide mRNA expression in 63 paired (tumor/normal tissue) RNA specimens from post-Chernobyl papillary thyroid cancers, we identified a promising set of 11 genes with evidence of differential radiation dose-expression relationship. In the same set of tumors we studied the prevalence of common somatic mutations including BRAF/RAS point mutations and RET/PTC rearrangements and their relationship with I-131 dose. Following a feasibility pilot study, we plan to initiate whole genome sequencing of radiation-related thyroid cancers in order to obtain a comprehensive view of the thyroid cancer genomic landscape and the role of radiation in thyroid carcinogenesis.

Gliomas account for approximately 80% of all primary malignant brain tumors and remain associated with considerable morbidity despite recent improvements in clinical care. Early work in our hospital-based case-control study indicated possible associations of glioma with candidate gene single nucleotide polymorphisms in the cell-cycle/apoptosis, DNA repair, oxidative stress, and immune function pathways. More recently, we conducted a cohort-based genome-wide association study of glioma using data from 1,856 cases and 4,955 controls. Our results strongly confirmed three of the seven previously reported associations from case-control studies at 20q13.33 (RTEL), 5p15.33 (TERT), and 9p21.3 (CDKN2BAS), and were consistent for the remaining four at 7p11.2 (EGFR both loci), 8q24.21 (CCDC26) and 11q23.3 (PHLDB1). While the direction and magnitude of the signal were consistent for samples from cohort and case-control studies, the strength of the association was more pronounced for loci rs6010620 (20q,13.33; RTEL) and rs2736100 (5p15.33, TERT) in cohort studies despite the much smaller number of cohort samples, likely due to the relatively higher percentage of high grade tumor in cohort studies. In addition, we performed a meta-analysis of the 85 most promising markers identified in our study in three replication sets (5,015 cases and 11,601 controls), but no new markers reached genome-wide significance. Initial efforts are underway to conduct a large pooled study of existing GWAS data on glioma taking into account differences by subtype.

For more information, contact Martha Linet.