Genetic Susceptibility to Cancer the GWAS era David Hunter - PowerPoint PPT Presentation

Genetic Susceptibility to Cancer – the GWAS era David Hunter Program in Genetic Epidemiology and Statistical P i G ti E id i l d St ti ti l Genetics Harvard School of Public Health Harvard School of Public Health Channing Laboratory, Brigham and Women’s g y, g Hospital Broad Institute of MIT and Harvard

Common variants and breast cancer – April 2007 CASP8 CASP8

Genome-wide CGEMS Scan – Breast Cancer FGFR2 P<0.00001 P<0.01 Hunter et al, Nat Gen 2007

Common variants and breast cancer – May 2007 FGFR2 2q35 TOX3 TOX3 MAP3K1 8q24 LSP1 S nGWAS ~ 1600

The case of the missing heritability The case of the missing heritability NATURE|Vol 456|6 November 2008

Holtzman and Marteau, Will genetics revolutionize medicine? NEJM 2000

Common variants and breast cancer – Sept 2010 FGFR2 2q35 TOX3 TOX3 MAP3K1 8q24 LSP1 S 5p12 16q12 1p11.2 RAD51L1 RAD51L1 3p24 17q23.2 CASP8 9p21 10p14 10q21 11q13 11q13 ESR1 nGWAS ~ 5000 5p12

Common variants and breast cancer – April 2012 FGFR2 12p11 2q35 12q24 TOX3 TOX3 21 21 21q21 MAP3K1 9q31 8q24 LSP1 S 5p12 16q12 1p11.2 RAD51L1 RAD51L1 3p24 17q23.2 CASP8 9p21 10p14 10q21 11q13 11q13 ESR1 nGWAS ~ 7000 5p12

Common variants and breast cancer – April 2013 FGFR2 12p11 C19orf61- 2q35 12q24 MLLT10-DNAJC1 TOX3 TOX3 21q21 21 21 DNAJC1 DNAJC1 MAP3K1 9q31 TCF7L2 8q24 PEX14 EMID1- LSP1 S MKL1 PTPN22- 5p12 HNF4G METAP1D- 16q12 CDCA7 ARHGEF5-NOBOX 1p11.2 DIRC3 DKFZp761E198- RAD51L1 RAD51L1 ITPR1 EGOT ITPR1-EGOT NTN4 NTN4 3p24 TGFBR2 BRCA2- 17q23.2 TET2 PAX9-SLC25A21 CASP8 ADAM29 RAD51L1 9p21 RAB3C CCDC88C 10p14 PDE4D MIR1972-2-FTO 10q21 EBF1 CDYL2 11q13 11q13 FOXQ1 FOXQ1 CHST9 CHST9 ESR1 RANBP9 SSBP4- nGWAS ~ 10000 5p12 MIR1208 Chr 2, 8, 8, 9, 10, 11, 12, 18

17 further hits (>100 in total) from 1000 1000 genomes imputation - Manhattan plots genomes imputation+meta-analysis g p y (62,890 cases and 61,872 controls, GWAS+iCOGS) Combine GWAS+iCOGS Combine GWAS+iCOGS Excluding known regions Excluding known regions Kyriaki Michailidou et a

Genetic Variants and Breast Cancer Risk Jan 2014 Easton, D in press p Unexplained: 54%* ~28 new hits from fine-mapping (7%) BRCA1 BRCA1 ~19 new SNPs post iCOGS BRCA2 (2%) CHEK2 45 iCOGS SNPs TP53 TP53 ATM ATM 27 pre-iCOGS SNPs (5%) PTEN PALB2 (9%) LKB1 * For overall breast cancer in Europeans Lower for ER- disease, early onset disease, and breast cancer in non-Europeans

GAME-ON OncoArray

Published Genome-Wide Associations through 12/2012 Published GWA at p ≤ 5X10 -8 for 17 trait categories NHGRI GWA Catalog www.genome.gov/GWAStudies

73 variants: OR in EAs vs AAs 3 a a s O s s s 1.3 1.25 EAs 1 2 1.2 OR in E 1.15 O 1.1 1.05 1.05 1 0 8 0.8 1 1 1 2 1.2 1 4 1.4 OR in AAs Haiman et al.

Value of tumor-subset GWAS Value of tumor subset GWAS Kraft and Haiman, Nat Gen, Oct 2010

Lessons from the GWAS o Small Relative Risks (RR 1.05-1.1) can be Small Relative Risks (RR 1 05 1 1) can be discovered and reproduced o Very large sample sizes are necessary to maximize power and reproducibility o False positives can be minimized with very large sample sizes pooled and analyzed de novo l i l d d l d d o Beware results reported from small studies o Beware results reported from small studies Hunter DJ. Epidemiology. 2012 23(3):363-7.

GENE-ENVIRONMENT INTERACTIONS DO CLASSIC BREAST CANCER RISK FACTORS SYNERGIZE WITH GWAS SNPS? 16,285 BC cases and 19,376 controls 16,285 BC cases and 19,376 controls 39 GWAS-assoc SNPS x 8 “Env” Risk Factors AAM AAM Parity AAMeno Height g BMI FH Smoking Alcohol Alcohol “After correction for multiple testing, no significant [multiplicative] interaction between SNPs and established risk factors...was found.” Barrdahl et al, BPC3, in preparation

GENE-ENVIRONMENT INTERACTIONS Good examples of supramultiplicative (synergistic) interactions between strong exogenous environmental risk factors (e.g. smoking, alcohol) and genetic exogenous environmental risk factors (e.g. smoking, alcohol) and genetic variants known to be on the pharmacogenetic pathway (e.g. NAT2, ALDH2). Few examples of synergistic interactions between lifestyle and environmental risk factors and GWAS associated SNPs factors and GWAS-associated SNPs.

GENE-ENVIRONMENT INTERACTIONS Good examples of supramultiplicative (synergistic) interactions between strong exogenous environmental risk factors (e.g. smoking, alcohol) and genetic exogenous environmental risk factors (e.g. smoking, alcohol) and genetic variants known to be on the pharmacogenetic pathway (e.g. NAT2, ALDH2). Few examples of synergistic interactions between lifestyle and environmental risk factors and GWAS associated SNPs factors and GWAS-associated SNPs.

GENE-ENVIRONMENT INTERACTIONS Good examples of supramultiplicative (synergistic) interactions between strong exogenous environmental risk factors (e.g. smoking, alcohol) and genetic exogenous environmental risk factors (e.g. smoking, alcohol) and genetic variants known to be on the pharmacogenetic pathway (e.g. NAT2, ALDH2). Few examples of synergistic interactions between lifestyle and environmental risk factors and GWAS associated SNPs factors and GWAS-associated SNPs. Actually, this is good news – multiplicativity makes risk modelling much more robust and predictable An indication that the GWAS variants represent biological processes independent of what we know from established risk factors Absence of G-E interaction simplifies our public health messages on E

RISK PREDICTION RISK PREDICTION

Clinical Utility of breast ca risk scores? • To select women at higher risk for prevention trials • To stratify screening? • As modifiers of high penetrance alleles

Loci of proven relevance to etiology of cancers May lead to new understanding of gene-specific mechanisms May lead to new understanding of gene-specific mechanisms May lead to new biologic understanding e.g. role of intergenic regions

Loci of proven relevance to etiology of cancers May lead to new understanding of gene-specific mechanism May lead to new understanding of gene-specific mechanism May lead to new biologic understanding e.g. role of intergenic regions

SUMMARY • Hundreds of new cancer “risk factors”, many more to come • Extremes of risk prediction approaching clinical utility • Sample size rules • Whole genome sequence data – much harder to g q interpret due to the very large number of potentially functional variants found per genome • Absence of G-E synergy on the multiplicative scale make our life easier • Insights into biology and mechanism likely to be the major contribution of genetic epidemiology

Acknowledgements CGEMS & DCEG HSPH-BWH ACS Stephen Chanock Peter Kraft Gilles Thomas Gilles Thomas David Cox David Cox Michael Thun Michael Thun Robert Hoover Sue Hankinson Heather Feigelson Kevin Jacobs Sara Lindstrom Ryan Diver Meredith Yeager Rulla Tamimi Vickie Stevens Richard Hayes Connie Chen J Joseph Fraumeni h F i Carolyn Guo EPIC Daniela Gerhard Julie Buring Elio Riboli Patricia Hartge Dan Chasman Afshan Siddiq Demetrius Albanes Rudolf Kaaks Sholom Wacholder Federico Canzian Federico Canzian Nilanjan Chatterjee Daniela Campa MEC Zhaoming Wang C Haiman Kai Yu B Henderson Margaret Tucker L Kolonel L Kolonel Jesus Gonzalez Bosquet Jesus Gonzalez Bosquet F Schumacher Montse Garcia-Closas Charles Chung Julia Ciampi

Acknowledgements DRIVE Discovery, Biology, and Risk of Inherited Variants in Breast Cancer C Cancer Research UK R h UK HSPH HSPH U i University of Utah it f Ut h Jack Cuzick Peter Kraft David Goldgar Connie Chen Dana Farber Cancer Institute Sara Lindstroem Vanderbilt University John Quackenbush Amit Joshi Wei Zheng Matthew Freedman Jirong Long Andrew Beck Alejandro Quiroz-Zarate Mayo Clinic University Hawaii Judy Garber Fergus Couch Loic Le Marchand Alexander Miron Alexander Miron National Institutes of Health NCI Program Office Fred Hutchinson CRI Stephen Chanock Daniela Seminara Paul Auer Ross Prentice Stanford University Alice Whittemore German Cancer Research Center (DKFZ) Federico Canzian University of Cambridge Rudolf Kaaks Douglas Easton Daniele Campa Daniele Campa University of Chicago BWH Brandon Pierce Rulla Tamimi Habibul Ahsan Sue Hankinson Aditi H Aditi Hazra University of Southern California Imperial College Brian Henderson Elio Riboli Chris Haiman Fred Schumacher