Cross-cancer genome-wide pleiotropy analysis based on GAME-ON and GECCO across five common cancers: lung, ovary, breast, prostate and colon cancer

Rayjean HUNG, Mount Sinai Hospital, Canada
FEHRINGER G. 1 , KRAFT P. 2 , SEMINARA D. 6 , CHATTERJEE N. 6 , SCHUMACHER F. 3 , EELES R. 5 , PHAROAH P. 4a , HAIMAN C. 3a , HUNTER D. 2 , RAFNAR T. 7a , SHEN H. 7b , SHIRAISHI K. 7c , BOSSÉ Y. 7d , OBEIDAT M. 7e , FREEDMAN M. 7f , PETERS P. 8 , GRUBER S. 3b , AMOS C. 9 , SELLERS T. 10 , EASTON D. 4b

1 Lunenfeld-Tanenbaum Research Institute of Mount Sinai Hospital, Toronto, Canada.
2 Harvard T.H. Chan School of Public Health, Boston USA, on behalf of Discovery, Biology, and Risk of Inherited Variants in Breast Cancer (DRIVE)
3 University of South California, Los Angeles, USA, on behalf of (a) The Elucidating Loci in Prostate Cancer Susceptibility (ELLIPSE), African American Breast Cancer Consortium (AABC), African Ancestry Prostate Cancer Consortium (AAPC), Japanese American Prostate Cancer Consortium (JAPC), Latino American Breast Cancer Consortium (LABC), Latino American Prostate Cancer Consortium (LAPC); (b)on behalf of Colorectal Transdisciplinary (CORECT) Study
4 University of Cambridge, Cambridge, UK, on behalf of (a) Ovarian Cancer Association Consortium (OCAC); (b) Breast Cancer Association Consortium (BCAC)
5 Institute of Cancer Research, London, UK, on behalf of The PRACTICAL Consortium
6 National Cancer Institute, Bethesda, USA.
7 Replication and eQTL studies: (a) deCODE genetics, Amgen, Reykjavik, Iceland. (b) Nanjing Medical University School of Public Health, Nanjing, China. (c) Division of Genome Biology, National Cancer Center Research Institute. Tokyo, Japan. (d) Institut universitaire de cardiologie et de pneumologie de Québec, Department of Molecular Medicine, Laval University, Québec, Canada. (e) University of British Columbia Centre for Heart Lung, Innovation, St. Paul's Hospital. Vancouver, Canada. (f) Dana-Farber Cancer Institute, Boston, USA.
8 Fred Hutchinson Cancer Research Center, Seattle, USA, on behalf of Genetic Epidemiology of Colorectal Cancer Consortium (GECCO)
9 Geisel School of Medicine, Dartmouth College, Lebanon, USA, on behalf of the Transdisciplinary Research for Cancer of Lung (TRICL)
10 Moffitt Cancer Center, Tampa, USA, on behalf of Transdisciplinary Cancer Genetic Association and Interacting Studies (FOCI)

Background: Identifying genetic variants with pleiotropic associations can uncover common pathways influencing multiple cancers and further understanding of cancer susceptibility. Method: We conducted a genome-wide cross-cancer pleiotropy analysis across five common cancers:  lung, ovary, breast, prostate and colon cancer from the GAME-ON/GECCO Network with a total of 61,851 cases, 61,820 controls of European ancestry using the association analysis based on subsets (ASSET) method, and validated the results in additional independent studies from Harvard, deCODE and Collaborative Oncological Gene-Environment Study (iCOGS) with a total of 55,789 cases and 330,490 controls of European ancestry. We have also evaluated the generalizability in Chinese, Japanese, Latinos and African Americans. Results: We identified a novel pleiotropic association at 1q22 with a variant associated with breast and lung squamous cell carcinoma (overall P-value for both cancers combined=8.9 x 10-8) in European descendants and the results were validated in the replication set.  The eQTL analysis of this region showed a consistent association with ADAM15/THBS3 gene expression in lung tissues in three independent studies. New pleiotropic associations were also found at previously known cancer loci: variants at a known BRCA2 locus for lung and breast cancer were associated with serous ovarian cancer (overall p-value=4.0 x 10-8); a known breast cancer locus, CASP8/ALS2CR12, with a variant associated with prostate cancer (overall P-value=1.9 x 10-8), and a known breast cancer locus, CDKN2B-AS1, where one variant was associated with lung adenocarcinoma (overall P-value=1.0 x 10-5) and a second was associated with prostate cancer (overall P-value=9.5 x 10-7). Conclusions: Our results provide important insights into common carcinogenesis across multiple major cancers and highlight the value of pleiotropy analysis.