Mini Oral Australian Epidemiology Association ASM 2018

Aging-associated changes in blood DNA methylation and cancer risk and survival (#77)

Pierre-Antoine Dugué 1 2 , Julie K Bassett 1 , JiHoon E Joo 3 , Chol-Hee Jung 4 , Ee M Wong 3 5 , Daniel D Buchanan 6 7 8 , Dallas R English 1 2 , Melissa C Southey 3 5 , Graham G Giles 1 2 , Roger L Milne 1 2
  1. Cancer Epidemiology and Intelligence Division, Cancer Council Victoria, Melbourne
  2. Centre for Epidemiology and Biostatistics, Melbourne School of Population and Global Health, The University of Melbourne, Parkville, VIC, Australia
  3. Genetic Epidemiology Laboratory, Department of Pathology, University of Melbourne, Parkville, VIC, Australia
  4. Melbourne Bioinformatics, The University of Melbourne, Melbourne, VIC, Australia
  5. Precision Medicine, School of Clinical Sciences at Monash Health, Monash University, Clayton, VIC, Australia
  6. Colorectal Oncogenomics Group, Department of Clinical Pathology, Melbourne Medical School, The University of Melbourne, Parkville, VIC, Australia
  7. Genomic Medicine and Family Cancer Clinic, Royal Melbourne Hospital, Parkville, VIC, Australia
  8. University of Melbourne Centre for Cancer Research, Victorian Comprehensive Cancer Centre, Parkville, VIC, Australia

Background: Aging is associated with widespread changes in DNA methylation. We aimed to investigate whether these are associated with cancer risk and survival.

Methods: Blood DNA methylation was measured using the Illumina HM450 assay at 485,512 cytosine-guanine sites (CpGs). We used linear mixed models to assess cross-sectional associations between age and DNA methylation using samples from 2,775 controls in case-control studies (of colorectal, kidney, lung, prostate and urothelial cancers, and mature B-cell lymphomas) nested in the Melbourne Collaborative Cohort Study. We then included data for 3,046 cases and used conditional logistic regression to assess associations of age-associated CpGs with cancer risk. We applied Cox regression on case data only to assess associations with cancer survival. All analyses were adjusted for age and other potential confounders. The Bonferroni correction was used to account for multiple testing.

Results: We identified 45,070 CpGs associated with age (P<10-7). Of these, two were associated (P<1.1x10-6=0.05/45070) with overall cancer risk (one near gene GPR68: OR=-0.55, P=2x10-8; one in MGMT: OR=1.51, P=3x10-7); associations were consistent across cancer sites for MGMT but not GPR68 (P-homogeneity=0.69 and 0.04, respectively). We found 91 CpGs associated with overall cancer survival, with strongest associations near genes ATP10A (HR=1.35, P=3x10-10), CUTA (HR=0.66, P=2x10-9), and FAM107A (gene body: HR=0.65, P=9x10-9).

Conclusion: Our study expands the list of age-associated CpGs and suggests that some age-associated methylation changes are associated with cancer risk and survival, independently of age. We are investigating what biological pathways are enriched in the genes identified and refining findings by cancer site.