AlperJCCC Molecular Epidemiology Symposium Molecular Epidemiology of Lung

  • Slides: 45
Download presentation
Alper-JCCC Molecular Epidemiology Symposium Molecular Epidemiology of Lung Cancer in Chinese Population Hongbing Shen,

Alper-JCCC Molecular Epidemiology Symposium Molecular Epidemiology of Lung Cancer in Chinese Population Hongbing Shen, M. D. , Ph. D Professor of Epidemiology Department of Epidemiology & Biostatistics Nanjing Medical University School of Public Health Tel: 86 -25 -86862756; Email: hbshen@njmu. edu. cn

Age-standardized Incidence Rates for Lung Cancer 2002 (per 100, 000) Parkin et al. CA

Age-standardized Incidence Rates for Lung Cancer 2002 (per 100, 000) Parkin et al. CA Cancer J Clin. 2005

Rural Men Woman Time trends for lung cancer mortality rates during 1987– 1999 in

Rural Men Woman Time trends for lung cancer mortality rates during 1987– 1999 in China, using agestandardized rate (ASR) by sex and area Urban Yang et al. Int J Cancer 2003

Sales Volume of Cigarettes in China, 1981 -1995 +100% Yang et al. JAMA 1999

Sales Volume of Cigarettes in China, 1981 -1995 +100% Yang et al. JAMA 1999

Age-Specific Prevalence Rates of Current and Former Smoking in Men and Women in China

Age-Specific Prevalence Rates of Current and Former Smoking in Men and Women in China 2001 35 -74 10. 6 % Yang et al. JAMA 1999; 60. 2 % Gu et al. Am J Public Health 2004 6. 9 %

Prevalence Rates of Smoking in a Rural Area of South Jiangsu Province, China (2004

Prevalence Rates of Smoking in a Rural Area of South Jiangsu Province, China (2004 -2005) Male Age No Smk Rate*(%) Female No Smk Rate*( %) Total No Smk Rate*( %) < 30 606 45. 7 856 0. 5 1462 19. 4 30~ 1530 70. 8 2388 0. 4 3918 28. 4 40~ 2176 75. 4 2865 0. 7 5041 33. 4 50~ 2405 73. 6 3050 1. 0 5455 33. 5 60~ 1302 64. 9 1611 0. 7 2913 29. 8 70~ 762 51. 8 1183 2. 5 1945 22. 7 Total 8781 68. 4 11953 1. 7 20734 30. 0 * Including former smokers Unpublished data

Men Woman Prevalence of environmental tobacco smoke (ETS) exposure among nonsmokers by gender and

Men Woman Prevalence of environmental tobacco smoke (ETS) exposure among nonsmokers by gender and age group in China, 2000 -2001 Gu et al. Am J Public Health. 2004

Background • Lung cancer incidence rate has been increasing significantly in the last two

Background • Lung cancer incidence rate has been increasing significantly in the last two decades in China. • More than 80% lung cancer can be attributed to smoking, yet only <10% smokers develop lung cancer. • Lung cancer is an excellent model for the research of gene-environment interactions.

Multi-step Carcinogenic Process of Lung Cancer Nicotine Dependent Metabolic activation Exposure Carcinogen uptake Apoptosis

Multi-step Carcinogenic Process of Lung Cancer Nicotine Dependent Metabolic activation Exposure Carcinogen uptake Apoptosis DNA damages Metabolic detoxification Excretion Persistency miscoding Mutation accumulation DNA Repair Normal DNA Cancer

Smoking Caused DNA Damage and Related Repair Pathways

Smoking Caused DNA Damage and Related Repair Pathways

Hypothesis: DNA repair gene SNPs (genotypes) DRC (Phenotype) Lung Cancer Risk Association? Gene-environment interaction?

Hypothesis: DNA repair gene SNPs (genotypes) DRC (Phenotype) Lung Cancer Risk Association? Gene-environment interaction? The genetic polymorphisms of the DNA repair genes are associated, independently or coordinately, with increased risk of lung cancer

Subjects & Methods Ø Hospital-based case-control studies of lung cancer Inclusion Criteria Cases: Incident

Subjects & Methods Ø Hospital-based case-control studies of lung cancer Inclusion Criteria Cases: Incident lung cancer patients Newly diagnosed, histopathologically confirmed No previous radiotherapy or chemotherapy Ø Controls: Cancer-free subjects Recruited at the same time period as cases and frequented matched to the cases on age, sex All the subjects are Han Chinese (Nanjing).

‘Sequence-based’ candidate SNPs approach Potentially functional SNPs of DNA repair gene XRCC 1 and

‘Sequence-based’ candidate SNPs approach Potentially functional SNPs of DNA repair gene XRCC 1 and risk of lung cancer A case-control study of 710 lung cancer cases and 710 cancer-free controls Hu Z, … Shen H. Pharmacogenetics & Genomics 2005 15(7): 457 -64.

Functional polymorphisms of DNA repair gene XRCC 1: Promoter: T-77 C (2004) Coding region:

Functional polymorphisms of DNA repair gene XRCC 1: Promoter: T-77 C (2004) Coding region: Arg 194 Trp (C>T) (DNA pol , PARP domain ) Arg 399 Gln (G>A) (PARP domain) Case-control study of 710 lung cancer cases and 710 cancer-free controls

XRCC 1 EXON Chromosome: 19 q 13. 2 1 2 5’UTR 3 4 5

XRCC 1 EXON Chromosome: 19 q 13. 2 1 2 5’UTR 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 3’UTR ATG -77 T>C Exon 6 Arg 194 Trp Exon 10 Arg 399 Gln 1246 -bp -77 T p 77 T XRCC 1 promoter Luciferase -1152 +93 -77 C Kpn. I p 77 C +1 Nhe. I XRCC 1 promoter -1152 p. GL 3 Luciferase +93 57– 64% p. GL 3

Distribution of Select Variables and XRCC 1 Variant Alleles in Lung Cancer Cases and

Distribution of Select Variables and XRCC 1 Variant Alleles in Lung Cancer Cases and Cancer-free Controls Variable Cases (n=710) No. % Controls (n=710) No. % Age (years) P value 0. 7100 ≤ 60 > 60 374 336 52. 7 47. 3 366 344 51. 5 48. 5 Sex Male Female Smoking status Non-smokers Ever-Smokers XRCC 1 T-77 C T allele C allele XRCC 1 C 194 T C allele T allele XRCC 1 G 399 A G allele A allele 0. 7207 520 190 73. 2 26. 8 513 197 72. 3 27. 7 <0. 0001 231 479 32. 5 67. 5 370 340 52. 1 47. 9 0. 0003 1198 222 84. 4 15. 6 1264 156 89. 0 11. 0 0. 8389 981 439 69. 1 30. 9 986 434 69. 4 30. 6 0. 4488 1040 380 73. 2 26. 8 1022 398 72. 0 28. 0

XRCC 1 T-77 C Polymorphism and Lung Cancer Risk XRCC 1 Genotype Cases(710) No

XRCC 1 T-77 C Polymorphism and Lung Cancer Risk XRCC 1 Genotype Cases(710) No % Controls(710) No % Adjusted OR (95%CI) XRCC 1 T-77 C TT (ref. ) 500 70. 4 558 78. 6 1. 00 CT 198 27. 9 148 20. 9 1. 51 (1. 17 -1. 94) CC 12 1. 7 4 0. 6 2. 98 (0. 93 -9. 59) CT/CC 210 29. 6 152 21. 4 1. 55 (1. 21 -1. 98) C allele 222 15. 6 156 11. 0 P = 0. 0003 Hu Z, … Shen H. Pharmacogenetics & Genomics 2005 15(7): 457 -64.

XRCC 1 T-77 C and Cumulative Smoking -77 TT Adjusted OR -77 CT/CC Pack-years

XRCC 1 T-77 C and Cumulative Smoking -77 TT Adjusted OR -77 CT/CC Pack-years of smoking P value for the test of additive gene-smoking interaction: 0. 027 Hu Z, … Shen H. Pharmacogenetics & genomics 2005 15(7): 457 -64.

This significant association with lung cancer was validated in another independent case-control study in

This significant association with lung cancer was validated in another independent case-control study in a Chinese population Hao et al. 2006 Oncogene

‘block-free’ linkage disequilibrium-based SNPs approach Polymorphisms of NER pathway core genes (XP) and susceptibility

‘block-free’ linkage disequilibrium-based SNPs approach Polymorphisms of NER pathway core genes (XP) and susceptibility of lung cancer A case-control study of 1010 lung cancer cases and 1011 cancer-free controls Hu Z, … Shen H. Carcinogenesis 2006; Jul; 27(7): 1475 -1480 Hu Z, … Shen H. CEBP 2006; 15(7): 1336 -40.

NER Core Genes 1 From NIEHS SNPs database Modified from James E. Cleaver, Nature

NER Core Genes 1 From NIEHS SNPs database Modified from James E. Cleaver, Nature Review Cancer 2005

Linkage disequilibrium-based tag. SNPs selection 8 candidate core genes in NER pathway 41 Tagging

Linkage disequilibrium-based tag. SNPs selection 8 candidate core genes in NER pathway 41 Tagging SNPs were selected / 32 were used in the analyses Based on the resequencing data of 90 individuals in the Environmental Genome Project (EGP) database (designed before the release of Hap. Map data) Based on the calculation of pairwise linkage disequilibrium (LD). LD parameter r 2 threshold 0. 5 (Carlson et al. ).

Laboratory Assays 5’-nuclease (Taq. Man) assay ABI PRISM 7900 HT Sequence Detection System 384

Laboratory Assays 5’-nuclease (Taq. Man) assay ABI PRISM 7900 HT Sequence Detection System 384 -well format Chinese National Human Genome Center at Shanghai, Quality control: Two blank control (water) and two duplicated samples in each 384 -well format The intensity of each SNP should meet the criteria of three clear clusters in two scales generated by SDS software (ABI).

Taq. Man assay results output: 5’-nuclease (Taq. Man) assay in 384 -well format

Taq. Man assay results output: 5’-nuclease (Taq. Man) assay in 384 -well format

Taq. Man Assay Results Output:

Taq. Man Assay Results Output:

Statistical analyses Chi-square test, Logistic regression Single locus analysis (ORs and 95% CIs) Multivariate

Statistical analyses Chi-square test, Logistic regression Single locus analysis (ORs and 95% CIs) Multivariate logistic regression model Haplotype inference: We used the PHASE 2. 0 program to infer haplotype frequencies based on the observed genotypes of each gene.

Single SNPs Analyses in the Dominant (Blue Line) and Recessive Model (Green Line)

Single SNPs Analyses in the Dominant (Blue Line) and Recessive Model (Green Line)

Analysis of association between the DDB 2(XPE) polymorphisms and risk of lung cancer DDB

Analysis of association between the DDB 2(XPE) polymorphisms and risk of lung cancer DDB 2 Genotype rs 830083 CC CG GG CG/GG rs 37816203 CC CG GG CG/GG Cases No. (%) 961 372 (38. 7) 483 (50. 3) 106 (11. 0) 589 (61. 3) Controls No. (%) 960 429 (44. 7) 426 (44. 4) 105 (10. 9) 531 (55. 3) Crude OR (95%CI) Adjusted OR (95%CI) a 1. 00 1. 31 (1. 08 -1. 58) 1. 16 (0. 86 -1. 58) 1. 28 (1. 07 -1. 53) 1. 00 1. 31 (1. 08 -1. 60) 1. 22 (0. 89 -1. 67) 1. 30 (1. 08 -1. 56) 967 978 429 (43. 9) 437 (44. 7) 112 (11. 5) 549 (56. 2) 1. 00 1. 29 (1. 07 -1. 56) 1. 04 (0. 77 -1. 41) 1. 24 (1. 03 -1. 48) 1. 00 1. 28 (1. 06 -1. 56) 1. 11 (0. 81 -1. 51) 1. 25 (1. 04 -1. 50) 374 491 102 593 (38. 7) (50. 8) (10. 6) (61. 4) Hu … Shen et al, Carcinogenesis, 2006 Jul; 27(7): 1475 -1480

Stratified analyses between the combined DDB 2 rs 830083 genotypes and lung cancer risk

Stratified analyses between the combined DDB 2 rs 830083 genotypes and lung cancer risk DDB 2 rs 830083 Cases (n=961) CC N (%) Controls (n=960) CG/GG N (%) CC N (%) CG/GG N (%) Adjusted OR (95% CI) CC CG/GG Age (years) ≤ 60 175 (37. 2) 296 (62. 9) 206 (44. 4) 258 (55. 6) 1. 00 1. 40 (1. 07 -1. 83) > 60 197 (40. 2) 293 (59. 8) 223 (45. 0) 273 (55. 0) 1. 00 1. 20 (0. 92 -1. 56) 112 (38. 0) 183 (62. 0) 207 (45. 1) 252 (54. 9) 1. 00 1. 29 (0. 96 -1. 75) 1 -29 95 (39. 8) 144 (60. 3) 105 (40. 4) 155 (59. 6) 1. 00 1. 01 (0. 70 -1. 45) >29 165 (38. 6) 262 (61. 4) 117 (48. 5) 124 (51. 5) 1. 00 1. 48 (1. 07 -2. 04) Pack-years of smoking 0 Family history of cancer

ERCC 2 and ERCC 3 polymorphisms and lung cancer risk Genotype ERCC 2 rs

ERCC 2 and ERCC 3 polymorphisms and lung cancer risk Genotype ERCC 2 rs 1618536 AA AG GG AG/AA ERCC 2 rs 1799786 CC CT TT CT/TT ERCC 2 rs 1799793 GG AG AA AG/AA ERCC 2 rs 3916823 AAAA/AAAA/---/-AAAA/-- + --/-- Cases No. (%) 975 177 492 306 798 18. 2 50. 5 31. 4 81. 9 965 844 117 4 121 87. 5 12. 1 0. 4 12. 5 970 850 116 4 120 87. 6 12. 0 0. 4 12. 4 986 744 225 17 242 75. 5 22. 8 1. 7 24. 5 Controls No. (%) 985 199 20. 2 488 49. 5 298 30. 3 786 79. 8 986 877 89. 0 108 11. 0 1 0. 1 109 11. 1 986 874 88. 6 111 11. 3 1 0. 1 112 11. 4 987 750 76. 0 227 23. 0 10 1. 0 237 24. 0 Adjusted OR (95%CI) 1. 00 1. 10 (0. 86 -1. 40) 1. 15 (0. 89 -1. 51) 1. 12 (0. 89 -1. 41) 1. 00 1. 12 (0. 84 -1. 49) 6. 18 (0. 68 -56. 52) 1. 16 (0. 87 -1. 54) 1. 00 1. 06 (0. 80 -1. 41) 6. 13 (0. 67 -56. 11) 1. 10 (0. 83 -1. 46) 1. 00 1. 01 (0. 81 -1. 25) 1. 88 (0. 84 -4. 21) 1. 04 (0. 84 -1. 29) Hu … Shen et al, Cancer Epidemiology Biomarkers & Prevention 2006, 15(7): 1336 -40.

ERCC 2 rs 13181 AA AC CC AC/CC ERCC 3 rs 2271026 TT CT

ERCC 2 rs 13181 AA AC CC AC/CC ERCC 3 rs 2271026 TT CT CC CT/CC ERCC 3 rs 4150441 AA AG GG AG/GG Combined genotypes 0 -1 ‘at risk’ locus 2 ‘at risk’ loci 3 ‘at risk’ loci >=4 ‘at risk’ loci 975 827 141 7 148 997 84. 8 14. 5 0. 7 15. 2 865 127 5 132 1002 774 214 14 228 787 200 13 213 952 78. 7 20. 0 1. 3 21. 3 1. 00 1. 07 (0. 86 -1. 34) 1. 02 (0. 46 -2. 24) 1. 07 (0. 86 -1. 33) 18. 2 48. 8 33. 1 81. 9 1. 00 1. 20 (0. 93 -1. 56) 1. 28 (0. 97 -1. 69) 1. 24 (0. 97 -1. 58) 20. 7 38. 8 25. 6 14. 9 1. 00 1. 29 (0. 98 -1. 70) 1. 38 (1. 02 -1. 85) 1. 51 (1. 09 -2. 10) P for trend: 0. 02 975 15. 7 49. 8 34. 6 84. 4 178 475 322 797 847 136 331 231 149 1. 00 1. 16 (0. 89 -1. 51) 1. 64 (0. 50 -5. 36) 1. 18 (0. 91 -1. 53) 1000 77. 3 21. 4 22. 8 149 474 329 803 86. 8 12. 7 0. 5 13. 2 899 16. 1 39. 1 27. 3 17. 6 186 349 230 134

ERCC 1 genotypes and lung cancer risk Genotyope Cases (n=1010) No. rs 3212948 %

ERCC 1 genotypes and lung cancer risk Genotyope Cases (n=1010) No. rs 3212948 % Controls (n =1011) No. 992 Adjusted OR (95%CI) % 986 GG 594 59. 9 521 52. 8 CG 338 34. 1 403 40. 9 CC 60 6. 1 62 6. 3 CG/CC 398 40. 2 465 47. 2 rs 1007616 P 835 1. 00 0. 005 0. 73 (0. 60 -0. 88) 0. 96 (0. 65 -1. 41) 0. 003 0. 76 (0. 63 -0. 91) 908 CC 513 61. 4 493 54. 3 CT 270 32. 3 354 39. 0 TT 52 6. 2 61 6. 7 CT/TT 322 38. 5 415 45. 7 Ma et al, Pharmacogenetics and Genomics, 2007. in press 1. 00 0. 009 0. 72(0. 59 -0. 89) 0. 90(0. 61 -1. 35) 0. 004 0. 75(0. 62 -0. 91)

Combined Effects of NER Core Genes and Lung Cancer * Six gene in dominant

Combined Effects of NER Core Genes and Lung Cancer * Six gene in dominant model and two in recessive model Unpublished data

Adjusted OR Stratified Analysis of Dichotomized NER Combined Diplotypes by Cumulative Smoking Pack-years of

Adjusted OR Stratified Analysis of Dichotomized NER Combined Diplotypes by Cumulative Smoking Pack-years of smoking P value for the test of gene-smoking interaction (multiplicative): 0. 004 Unpublished data

Two Factor Gene-gene (Diplotype) Interaction and Gene-environment Interaction Unpublished data

Two Factor Gene-gene (Diplotype) Interaction and Gene-environment Interaction Unpublished data

‘block-based’ linkage disequilibrium mapping Haplotyope-Tagging SNPs in MGMT (O 6 -alkylguanine. DNA alkyltransferase) and

‘block-based’ linkage disequilibrium mapping Haplotyope-Tagging SNPs in MGMT (O 6 -alkylguanine. DNA alkyltransferase) and lung cancer susceptibility Case-control study of 500 lung cancer cases and 517 cancer-free controls Hu Z, … Shen H. Human Mutation 2007 published online.

MGMT Gene Structure and Haplotype Block for Beijing Han Chinese From Hap. Map No.

MGMT Gene Structure and Haplotype Block for Beijing Han Chinese From Hap. Map No. of SNPs: 4 14 7 1 1 3 1 11

Illumina High Throughout Genotyping Platform (CHGC-Shanghai)

Illumina High Throughout Genotyping Platform (CHGC-Shanghai)

MGMT Tagging SNPs Based on the Gene Blocks (by Haploview program ) ht. SNPs

MGMT Tagging SNPs Based on the Gene Blocks (by Haploview program ) ht. SNPs 1 1 2 2 Gabriel et al. 2 2 10 ht. SNPs from the 6 blocks Stram et al. Rh 2 > 0. 80 25 informative SNPs (MAF>0. 05) in our controls

Summary of the findings: • Single variant analysis: no significant main effects was observed

Summary of the findings: • Single variant analysis: no significant main effects was observed for each single genetic variant in MGMT on lung cancer risk. • Haplotype analysis: no significant difference was observed for frequencies of haplotypes in different blocks between cases and controls. • Diplotype analysis: only the diplotype carrying 1 variant copy of the block 3 was associated with a significantly decreased lung cancer risk. • Stratification: The frequencies of haplotypes in blocks 1 and 2, block 4, 5 were significantly different between cases and controls in different stratum of smoking dose. Hu Z, … Shen H. Human Mutation 2007

MDR Models of Selected Gene Regions and Co-variables Best models CVC * Avg. Testing

MDR Models of Selected Gene Regions and Co-variables Best models CVC * Avg. Testing Accuracy Sign Test P Value One Factor: Pack-years of smoking 100/100 0. 5827 0. 0009 0. 0330 Two Factors: Pack-years of smoking; block 3 99/100 0. 5992 0. 0018 0. 0344 Three Factors: Pack-years of smoking; block 5; rs 1625649: C>A 100/100 0. 6145 0. 0000 0. 0453 Four Factors: Pack-years of smoking; block 3; block 5; rs 1625649: C>A (pre-block SNP) 100/100 0. 6374 0. 0000 0. 0146 Five Factors: Pack-years of smoking; block 3; block 5; block 6; rs 1625649: C>A 68/100 0. 5349 0. 5398 0. 3194 * CVC: cross-validation consistency

Conclusions • DNA repair gene polymorphisms do contribute to lung cancer risk and large

Conclusions • DNA repair gene polymorphisms do contribute to lung cancer risk and large prospective studies are needed to validate the findings. • The candidate gene and sequence-based SNPs selection approach have a high priority to find a biological relevant association but our knowledge on the function of different kind of SNPs are still limited. • The candidate gene and Mapping-based SNPs selection approach may have an increased power to detect a haplotype-based disease risk, however, multiple testing is a big issue when analyzing many SNPs and interactions using different combinations.

The resources for molecular epidemiological studies in Nanjing, China Ø Lung Cancer Case-control Study

The resources for molecular epidemiological studies in Nanjing, China Ø Lung Cancer Case-control Study >1200 lung cancer cases and matched controls >500 lung cancer cases with survival information Ø Breast Cancer Case-control Study >1000 breast cancer cases and matched controls Ø Gastric Cancer Case-control Study >800 gastric cancer cases and matched controls Ø Changzhou population-based cohort study Established baseline information and blood samples for 21, 000 participants (30 -75 yrs) in the suburb of Changzhou city, Jiangsu province. Expect to recruit more than 40, 000 in the region.

Acknowledgement Dr. Qinyi Wei Dr. Dongxin Lin Drs. Daru Lu; Li Jin Dr. Wei

Acknowledgement Dr. Qinyi Wei Dr. Dongxin Lin Drs. Daru Lu; Li Jin Dr. Wei Huang Dr. Tangchun Wu U T M. D. Anderson Cancer Center Chinese Academy of Medical Science Fudan University CHGC, Shanghai Huazhong University of Science & Technology Nanjing Medical University Jiangsu Cancer Hospital First Affiliated Hospital of Nanjing Medical University Shanghai Cancer Hospital Wuhan Zhongnan Hospital The Molecular Epi Lab in NJMU

Thank you!

Thank you!