Genome-wide association study of colorectal adenoma in the Nurses Health Study and the Health Professionals Follow-up Study. A. D. Joshi1, A. Hazra1,2, C. Chen1, R. B. Hayes3, P. Kraft1, U. Peters4, A. T. Chan2,5 1) Program in Molecular and Genetic Epidemiology, Department of Epidemiology, Harvard School of Public Health, Boston, MA; 2) Channing Division of Network Medicine, Department of Medicine, Brigham and Womens Hospital and Harvard Medical School, Boston, MA; 3) Division of Epidemiology, New York University School of Medicine, New York, NY; 4) Cancer Prevention Program, Fred Hutchinson Cancer Research Center, Seattle, WA; 5) Division of Gastroenterology and Hepatology, Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, MA.

   The vast majority of colorectal cancers (CRC) arise from adenomatous polyp (adenoma) precursors. Although several genome-wide association studies (GWAS) have identified germline susceptibility loci for CRC, there are relatively little data characterizing genetic risk of adenomas. We therefore conducted a GWAS of advanced adenomas among participants enrolled in the Nurses Health Study (NHS) and the Health Professionals Follow-up Study (HPFS) cohorts. Advanced adenoma cases (N = 513 in NHS, 314 in HPFS) had adenomatous polyps diagnosed on endoscopy with 1cm size and/or tubulovillous, villous, or high-grade dysplasia histological type. Controls (N = 578 in NHS, 351 in HPFS) had no adenoma diagnosis on colonoscopy, and were matched to cases on age, reason for endoscopy and time of most recent endoscopy. Genotyping was done using Illumina OmniExpress array and imputed to CEU reference panel (HapMap II release 24). Study-specific effect estimates were adjusted for age at blood draw and the top 3 eigenvectors and combined using an inverse-variance weighted meta-analysis. We also conducted an in silico meta-analysis GWAS of a total of 2,298 adenoma cases (all histological types and sizes) and 10,392 controls of European ancestry, from GWAS studies nested within NHS and HPFS, that were originally designed to study other outcomes, including type 2 diabetes, coronary heart disease, kidney stones, breast cancer, endometrial cancer, mammographic density, as well as controls from GWAS studies of CRC. The 1p31.1 locus, rs17520771 was associated with an odds ratio of 0.56 (0.46, 0.69; p= 3.2x10-8) for advanced adenoma. In addition, we found that CRC-susceptibility SNPs previously identified in GWAS were also associated with risk of advanced adenomas (rs4939827, SMAD7, p= 0.00014) and all adenomas (rs6983267, 8q24, p= 0.00036). However, after corrections for multiple comparisons, we did not identify significant associations between genotyped/ imputed SNPs and overall adenoma risk. In this case-control study nested within two large, population-based cohorts, we observed a possible association between the 1q31.1 locus and risk of advanced adenoma. Moreover, in an in silico analysis of GWAS data collected within other studies nested in these cohorts, we observed associations between established CRC susceptibility SNPs and adenoma risk. These data support the need for further studies examining genetic predisposition to different stages of colorectal tumorigenesis.

You may contact the first author (during and after the meeting) at