What Do We Know About Individual Variability and Its

What Do We Know About Individual Variability and Its Contribution to Disease? Nathaniel Rothman, MD, MPH, MHS Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, NCI, NIH, DHHS

Scope of Genetic Variation Single Nucleotide Changes 10-12 Million Common SNPs > 10% MAF 30-50 Million Uncommon SNPs 1-10% MAF >100 Million Rare SNPs < 1% MAF All 3.1 Billion bp in Human Genome? GWAS: Current chips Interrogate ONLY Common SNPs (> 10%)

Common Variation (MAF > 10%) Represents a Small Part of All Variation 1.0 0.8 SNPs by observations in CEU HapMap 1000 Genomes 6 ) Coun nt (x 10 6 0.6 0.4 02 0.2 0.0 0 10 20 30 40 50 60 Minor Alleles

Published Cancer GWAS Etiology Hits: April 1, 2012 KIF1B 1p36 IRF4 ~230 Disease Loci marked by SNPs 1 Locus marked by a CNV SLC4A7/ 6p22 C2orf43 NEK10 6p21 ITGA9 GPRC6A LSP1 THADA TACC3 FOXP4 DMRT1 TERT/ 10p15.115 1 11p1515 CLPTM1L TERT EHBP1 CLPTM1L GABBR1 JAZF1 BNC2 GSTM1 NAT2 10p14 CDKN2A/ LMO1 deletion REL 3p12.1 HLA-F 5p15 IKZF1 8p21 CDKN2B GATA3 ATM 1p11.2 3p11.2 HLA-A MSMB/ 5p12 EGFR 6p21.33 NCOA4 FAM111A 1q22 2q13 5q11.2 ARIDB5 11q13 3q13 6p21.32 1q21.1 PDLIM5 5q11 HSD17B12 EEFSEC BAK1 LMTK2 PLCE1 EPAS1 ADH1B 5q13 TYR 1q21.3 FOXE1 TET2 DDX4 ECHDC1/ ZNF365 11q23.1 1q32.1 ITGA6 RNF146 EIF3H 9q31.2 FGFR2 PHLDB1 3q26 C6orf97/ 7q32 SPRY4 8q24.21(x5) CCDC26 MYC ABO CTBP2 11q24.1 TARDBP 2q31 ESR1 7q32 TP63 1q41 SLC22A3 PSCA 10q21.2 CASP8 ERG2 1q42 BARD1 LINC00340 10q22.3 1q42.12 2q35 10q26 FARP2 2q37 12p11.23 ATF7IP KRT5 KITLG 12q13.13 ALDH2 KLF5/ KLF12 13q12.125 13q22 CEBPE NKX2-1 BMP4 RAD51L1 15q15 GREM1 15q21.3 TOX3 CDH1 15q23 16q24.1 CHRNA3/ CHRNA5 MC1R 1 Liver 10 CLL HNF1B x 2 COX11/ STXBP4 17q24.3 6 Neuroblastoma 3 Kidney SLC14A1 SMAD7 19p13 RHPN2 CCNE1 19q13.2 PRKD2 KLK2/ KLK3 C20orf54 BMP2 ASIP RTEL1 4 Thyroid 2 Non-Hodgkin 5 Ovary 3 Gastric MX2 21q22 9 Multiple 22q12.25 22q13 4 Pediatric Acute Lymphoblastic Leukemia 7 Esophageal Squamous 48 Prostateate 25* Breast13 Colorectal7 Basal Cell Carcinoma11* Bladde77 Glioma 6 Lung 9 Melanoma 4 Pancreas 7 Nasopharyngeal 6 Testicular BIK NUDT10/ NUDT11 3 Ewing Sarcoma 3 Hodgkins Chung & Chanock 2011

Integrating g Knowledge of Human Variability into Studies of Exposure and Disease Obtain mechanistic insight Clarify dose-response relationships, and more effectively evaluate low levels of risk Identify new environmental health hazards Develop more effective prevention, screening, and treatment strategies

Susceptibility for Bladder Cancer as a model for Gene- Environment Interactions

Spanish Bladder Cancer Study Hospital-based case-control study (1998-2001) DNA from 1,150 cases and 1,149 controls

Data Collection o Data Resources Response Rate CAPI 86% Blood/Buccal Cell 95% Diet Qx. 72% Urination Diary 60% Toenails 77% Hair dye Qx. 85% Specific Areas Demographics Smoking Occupation/Environmental Family history Medical/Drugs Genetic Susceptibility Functional Assays Fluid intake Food Frequency Food Carcinogens Urine ph Urinary freq Arsenic/Selenium Hair Dye

Disinfection Byproducts, GSTT1/GSTZ1, and Bladder Cancer Ubiquitous in all water supplies disinfected with chlorine Precursor levels much higher in surface than ground water sources, creating a major differential in exposure A complex mixture of halogenated organics with mutagenic properties (one class called Trihalomethanes THM)

Exposure Assessment for Disinfection Byproducts Lifetime residential / water source history gathered from each study subject Detailed information on water treatments & sources gathered from all major water utilities in study area average annual THM level at each utility assigned by expert evaluation Personal & utility information merged to create a lifetime yearby-year estimate of THM exposure for each study subject L t l ti k t Long-term average, cumulative, peak, etc. exposures were calculated

GSTT1, GSTZ1 & Disinfection Byproducts GSTT1 Mutagenic activation of trihalomethanes (THM) GSTT1 null deletion eliminates activity GSTZ1 Clears haloacetic and other dihalogenated acids: The CT/TT: M82T (exon7+29c>t, rs 1046428) variant has decreased activity

Trihalomethane Concentration in Drinking Water and Risk of Bladder Cancer OR=1.8 1.8 1.6 1.4 p (trend)=0.03 03 1.2 1 0.8 0.6 0.4 0.2 0 <=8 >8-26 >26-49 >49 Avg THM, μg/l Villanueva et al., Am J Epidemiol 2007

Interaction between THM and GSTT1, p (interaction) = 0.021 OR=2.2 2.5 2 1.5 p (trend)=0.28 p (trend)=0.007 1 0.5 0 <=8 >8-26 >26-49 >49 <=8 >8-26 >26-49 >49 Avg THM, μg/l Avg THM, μg/l GSTT1 Null GSTT1 +, +/+ Cantor et al., EHP, 2010

Interaction between THM and GSTZ1* p (interaction) = 0.018 018 3 OR=2.9 2.5 p (trend)=0.0043 2 1.5 p (trend)=0.28 1 05 0.5 0 <=8 >8-26 >26-49 >49 <=8 >8-26 >26-49 >49 Avg THM, μg/l Avg THM, μg/l *CT/TT: M82T (exon7+29c>t, rs 1046428) Cantor et al., EHP 2010

Trihalomethane Concentration, Combined GSTT1 +/++ and GSTZ1 CT/TT Genotype, and Bladder Cancer; p (interaction) = 0.005 OR=5.9 6 5 p (trend)=0.0012 4 3 2 p (trend)=0.57 1 0 <=8 >8-26 >26-49 >49 <=8 >8-26 >26-49 >49 Avg THM, μg/l Avg THM, μg/l GSTT1 NULL & GSTZ1 CC GSTZ1 +/++ & GSTZ1 CT/TT Cantor et al., Environ Health Perspect 2010

Study Design of Multi-stage NCI GWAS of Bladder Cancer Includes ~12,000 cases and ~53,000 controls from 20 studies Rothman et al Rothman et al., Nat Genet 2010

Bladder Cancer Susceptibility Loci Previous GWAS 8q24.21 3q28 (TP63) 5p11.3 (TERT-CLPTM1L) 8q24.23 (PSCA) 4p16.3 (TMEM129 TACC3-FGFR3) NCI GWAS 1p13.3 (GSTM1) 8p22 (NAT2) 2q37.1 (UGT1A) 19q12 (CCNE1) 22q13.11 (CBX6, APOBEC3A) Rothman et al., Nat Genet 2010 NCI + MD Anderson GWAS NCI MD Anderson GWAS 18q12.3 (SLC14A1) Garcia-Closas et al., Hum Mol Genet, 2011

In the Bladder Cancer GWAS, 7 out 12 SNPs Showed Significant Additive Interactions with Tobacco Odds ratios for joint associations of smoking status (ever vs never smokers) and 12 susceptibility variants with bladder cancer risk Garcia-Closas et al., Submitted

The Level of Risk Discrimination Could be of Public Health Relevance Cumulative 30-year absolute risk for bladder cancer in a 50 year old male in the USA, overall and by quartiles of a polygenetic genetic score. P-additive =1x10-4 RD are risk differences for current vs never smokers Garcia-Closas et al., Submitted

Impact of Eliminating Smoking in 100,000 Current Smokers in Highest vs. Lowest Genetically Susceptible Subgroup of the Population 8,000 vs. 2,000 cases of bladder cancer eliminated

Tobacco smoking Occupational exposures: e.g., aromatic amine dyes, cutting oils Water contaminants: arsenic disinfection by-products pre-carcinogens Liver NAT2 GSTM1 GSTT1 UGT1A6 Urinary Bladder? CCNE1 CBX6, APOBEC3A TMEM129 TACC3-FGFR3 8q24: PSCA, MYC TP63 TERT-CLPTM1L carcinogenic metabolites Urine ph<6 Voiding frequency Kidney SLC14A1 Urinary Bladder UGT1A6

Integrating g Knowledge of Human Variability into Studies of Exposure and Disease Obtain mechanistic insight Clarify dose-response relationships, and more effectively evaluate low levels of risk Identify new environmental health hazards Develop more effective prevention, screening, and treatment strategies

Advances will be accelerated by Collective Intelligence I not only use all of the brains I have, but all I can borrow Woodrow Wilson

Acknowledgements Studies in Europe and US, Genomics, and Analysis NCI New England D. Silverman M. Schwenn M. Garcia-Closas A. Johnson N. Chatterjee M. Karagas K. Cantor L. Prokunina ACS K. Jacobs M. Thun D. Baris E. Jacobs J. Figueroa R. Diver S. Chanock MD Anderson Spain Xifeng Wu N. Malats F. Real Harvard M. Kogevinas I. De Vivo