Nature Genetics: doi: /ng.450

Transcription

1 Supplementary Figure 1(a) UK data. Quantile-quantile plot for the test statistics (Cochran-Armitage 1df chi-squared trend tests) for stage 2. Shaded area indicates the 95% confidence region under the global null hypothesis. Inflation was assessed using the lowest 90% of the test statistics (expected values less than 2.71).

2 Supplementary Figure 1(b) Australian data, unstratified. Quantile-quantile plot for the test statistics (Cochran-Armitage 1df chi-squared trend tests) for stage 2. Shaded area indicates the 95% confidence region under the global null hypothesis. Inflation was assessed using the lowest 90% of the test statistics (expected values less than 2.71).

3 Supp Figure 1(c) Australian data, stratified into MCCS and EOPCS/RFPCS. Quantile-quantile plot for the test statistics (Cochran-Armitage 1df chi-squared trend tests) for stage 2. Shaded area indicates the 95% confidence region under the global null hypothesis. Inflation was assessed using the lowest 90% of the test statistics (expected values less than 2.71).

4 Supplementary Figure 1(d) stratified UK & Australian data combined. Quantile-quantile plot for the test statistics (Cochran-Armitage 1df chi-squared trend tests) for stage 2. Shaded area indicates the 95% confidence region under the global null hypothesis. Inflation was assessed using the lowest 90% of the test statistics (expected values less than 2.71).

5 Supplementary Figure 2. Scatter plot of the first two principal components for individuals in stage 2, based on uncorrelated SNPs (r 2 <0.1). Blue: UK; green: Australia RFPCS and EOPCFS; red: Australia MCCS

6 Supplementary Table 1. Summary of distribution of 1df P-trend test statistics for SNPs in stage 2 (n=43,671). Significance Level Number of SNPs Expected number

7 Supplementary Table 2. Summary of 116 SNPs significant at P<10-6 in stages 1 and 2 combined. SNP Stage Chr Position Allele version case-11 case-12 case-22 control-11 control-12 control-22 Ptrend Ptrend combined rs AG E E E-07 rs CT E E E-08 rs GA E E E-07 rs AG E E E-09 rs AG E E E-09 rs AG E E E-08 rs AG E E E-07 rs AG E E E-08 rs AG E E E-08 rs CT E E E-07 rs TC E E E-07 rs AG E E E-08 rs CT E E E-08 rs CT E E E-07 rs AC E E E-07 rs TC E E E-08 rs GA E E E-07 rs TC E E E-07 rs AG E E E-07 rs CT E E E-09 rs AG E E E-07 rs AG E E E-07 rs CT E E E-07 rs CA E E E-08 rs GA E E E-09 rs CT E E E-07 rs AG E E E-07 rs CT E E E-09 rs GT E E E-07 rs AG E E E-07 rs GT E E E-08 rs GA E E E-08 rs CT E E E-08 rs TC E E E-08 rs CA E E E-08 rs GT E E E-07 rs AG E E E-07 rs TG E E E-07 rs GA E E E-08 rs TC E E E-08

8 Supplementary Table 2: Summary of 116 SNPs significant at P<10-6 (cont.) SNP Stage Chr Position Allele version case-11 case-12 case-22 control-11 control-12 control-22 Ptrend Ptrend combined rs AG E E E-07 rs TC E E E-08 rs GA E E E-07 rs CT E E E-12 rs AG E E E-10 rs AG E E E-12 rs TC E E E-07 rs GA E E E-12 rs AG E E E-12 rs TC E E E-07 rs TC E E E-08 rs AG E E E-08 rs CT E E E-12 rs GT E E E-23 rs AG E E E-13 rs AG E E E-12 rs AC E E E-08 rs GA E E E-07 rs AC E E E-23 rs GA E E E-24 rs CA E E E-24 rs CA E E E-18 rs GA E E E-20 rs GA E E E-18 rs TG E E E-24 rs CT E E E-16 rs CT E E E-19 rs GA E E E-18 rs TC E E E-18 rs TC E E E-28 rs CT E E E-07 rs AG E E E-12 rs TC E E E-07 rs AG E E E-11 rs TG E E E-08 rs AG E E E-07 rs TG E E E-07 rs AG E E E-07 rs AG E E E-07

9 Supplementary Table 2: Summary of 116 SNPs significant at P<10-6 (cont.). SNP Stage Chr Position Allele version case-11 case-12 case-22 control-11 control-12 control-22 Ptrend Ptrend combined rs TG E E E-10 rs AG E E E-10 rs AG E E E-13 rs AG E E E-10 rs TG E E E-10 rs TC E E E-07 rs GA E E E-07 rs CT E E E-07 rs TC E E E-11 rs TC E E E-18 rs CA E E E-15 rs GA E E E-07 rs GT E E E-16 rs GA E E E-08 rs CT E E E-12 rs TC E E E-07 rs TC E E E-10 rs AG E E E-07 rs AG E E E-08 rs AG E E E-07 rs CT E E E-08 rs GA E E E-09 rs TC E E E-17 rs AG E E E-17 rs GT E E E-09 rs X GA NA NA E NA NA E E-09 rs X CT NA NA E NA NA E E-10 rs X AG NA NA E NA NA E E-10 rs X CT NA NA E NA NA E E-10 rs X CA NA NA E NA NA E E-07 rs X TC NA NA E NA NA E E-07 rs X GA NA NA E NA NA E E-08 rs X TC NA NA E NA NA E E-07 rs X AG NA NA E NA NA E E-07 rs X CT NA NA E NA NA E E-08 rs X GA NA NA E NA NA E E-08 rs X CT NA NA E NA NA E E-08

10 Supplementary Table 3. Estimated per-allele odds ratios and significance tests (1df P- trend) for association in stage 2, after adjustment for the first and second principal components (PC), for the 12 SNPs selected for genotyping in stage 3. Marker rs rs rs rs rs rs rs rs rs rs rs rs Unadjusted Adjusted for first Adjusted for first & second PCs OR (95%CI) OR (95%CI) PC OR (95%CI) P-value P-value 1.13 ( ) 1.13 ( ) P-value 1.13 ( ) ( ).67 ( ).67 ( ) 1.5x x x ( ).85 ( ).85 ( ) 4.7x x x ( ) 1.14 ( ) 1.14 ( ) 9.6x x ( ).88 ( ).89 ( ) ( ) 1.14 ( ) 1.14 ( ) ( ) 1.13 ( ) 1.12 ( ) ( ) 1.14 ( ) 1.15 ( ) x ( ) 1.27 ( ) 1.27 ( ) 2.4x x x ( ) 1.17 ( ) 1.17 ( ) 3.5x x x ( ).85 ( ).85 ( ) 1.1x x x ( ).83 ( ).83 ( ) 6.3x x x10-8

11 Supplementary Table 4. Summary of studies contributing to stage 3. Study Name and Site Site Total number Number included after QC Cases Controls Cases Control Type of Study Age range of cases (years) Case identification Control Ascertainment/ matching BiPAS Birmingham, UK Hospital Based Hospital Based Non PrCa patients in PSA and Lower Urinary Tract Symptom clinics CHSH China Hospital based Hospital based Ethnic and Age FHCRC Seattle, US 1,409 1,335 1,401 1,301 Population based Cancer registry based Age Telephone dialing FMHS Michigan, US Community based Hopital based Race Community based probability sample HaPCS Hannover, Germany Hospital based Hospital based Ethnic origin Male blood donors MAYO Minnesota, US Hospital based cases Hospital-based Geographically Population via epidemiology study MCCS- RFPCS- EOPCFS MEC Melbourne, Australia Multiethnic Cohort Study: US, California, Hawaii 1,532 1,546 1,410 1,499 prospective cohort study-populationbased case-control study-family study of cases diagnosed at age <56 yrs 2,216 2,024 2,216 2,024 Nested case-control within population based cohort Cancer registry based (at entry to cohort) Cancer registry based MOFFIT Florida, US Hospital based Hospital based Age NC_CCPC California, US Population based Cancer registry Age Controls for stage 3 are a random sample of the MCCS cohort Age at entry and race/ethnicity

12 based Race/ethnicity Random digit dialing PCMUS Bulgaria Hospital based Hospital based Age and ethnicity PENN (also Philadelphia, Health System-Based Health System Age and Sex known as US Study Based SCORE) ProtecT UK 1,801 1,799 1,706 1,709 Population based PSA screen detected disease Ethnic origin Age QUEENSLA ND TAMPERE TASPRAC UKGPCS Australia (incorporating the ProsCan Study and the Retrospective Queensland Study) Tampere, Finland Tasmania, Australia UK (incorporating The Prostate Cancer Research Foundation Study 956 1, ,354 Population based Identified through clinicians in public hospitals and private clinics Two groups: population based matched for sex, age and post-code; blood donor controls matched for sex only 1,769 1,774 1,769 1,774 Population based Hospital based Sex only Blood donors& PSA screening participants Population based case control / Familial PC Study 1,316 1,179 1,289 1,130 Population and hospital based Cancer registry based Age +/- 5 years Sex Electoral Roll Hospital based Ethnic origin & age GP practices ULM Ulm, Germany Population and hospital based USC California, US 1, , Population based Cancer registry based Hospital based Geographically Age +/- 5 years Age, geographically, Neighbourhood

13 UTAH Utah, US Population based Utah Population Database Valais Valais, Switzerland Population based Population based TOTAL 16,939 15,366 16,229 14,821 block walking Age Geographically, Blood donors, non PrCa patients in private clinics

14 Supplementary Table 5. Joint analysis SNPs rs and rs on chromosome 4. Model rs Stage 3 Stage 3 OR 1 (95% CI) P-value All Stages P-value rs Stage 3 Stage 3 OR 2 (95% CI) P-value All Stages P-value Unadjusted.90 ( ) 1.8x x ( ) 1.8x x10-11 Adjusted.92 ( ) x ( ) Per-allele OR 2. Adjusted ORs based on a multiple logistic regression including both SNPs.

15 Supplementary Table 6. Estimated per-allele ORs and significance levels in the CGEMS GWAS, for SNPs typed in stage 3. Marker CGEMS Current study Current study+cgems combined P-value OR (95%CI) P-value rs ( ) x x10-8 rs ( ) x x10-23 rs x x10-15 rs ( ) x x10-12 rs ( ) x x10-15 rs ( ) x x10-6 rs ( ) x x10-8 rs ( ) x x10-30 rs ( ) x x10-32 rs ( ) rs ( ) rs ( ) x x10-31

16 Supplementary Table 7. Estimated per-allele ORs and Minor Allele Frequencies for each SNP by population in stage 3. Marker rs rs rs rs rs rs rs rs rs rs rs S SNP16 rs European Asian African-American OR (95%CI) OR (95%CI) OR (95%CI) P-value 1 MAF P-value 1 MAF P-value 1 MAF 1.08 ( ) 1.24 ( ) 1.05 ( ) ( ).77 ( ).55 ( ) ( ) 1.03 ( ).91 ( ) ( ).93 ( ).98 ( ) ( ).90 ( ) 1.05 ( ) ( ) 1.12 ( ) 1.16 ( ) One study (N/A) ( ) 1.09 ( ) 1.04 ( ) ( ) 1.27 ( ) 1.27 ( ) ( ) 1.20 ( ) 1.09 ( ) ( ) 1.07 ( ) 1.01 ( ) ( ) No observation No observation One study (N/A).41 N/A N/A N/A N/A.98 ( ).97 ( ).97 ( ) One study (N/A) ( ).97 ( ).97 ( ) One study (N/A) ( ).92 ( ).92 ( ) Test for heterogeneity of studies

17 Supplementary Table 8. Screen/Clinical detected-specific per-allele ORs for PrCa subdivided by studies in populations where screening is more vs. less common in stage 3. Marker Screening more common 1 Screening less common 2 rs ( ) 1.04 ( ).59 rs ( ).74 ( ) 1.00 rs ( ).89 ( ).18 rs ( ) 1.08 ( ).80 rs ( ).87 ( ).75 rs ( ) 1.02 ( ).37 rs ( ) 1.01 ( ).53 rs ( ) 1.21 ( ).24 rs ( ) 1.26 ( ).33 rs ( ).99 ( ).31 rs * No observation 1.01 ( ) N/A rs ( ).95 ( ).43 rs ( ).87 ( ).07 * Only one study 1. US Studies plus the UK ProtecT Study 2. Non US Studies minus ProtecT 3. Test of difference in per-allele ORs P 3

18 Supplementary Table 9. PSA levels by genotype in controls in stage2. Marker Geometric mean PSA (95%CI) p-value 1 rs GG GA AA 1.40 ( ) 1.43 ( ) 1.50 ( ).42 rs AA AG GG 1.42 ( ) 1.44 ( ).93 ( ).97 rs CC CT TT 1.51 ( ) 1.32 ( ) 1.43 ( ).04 rs CC CA AA 1.38 ( ) 1.42 ( ) 1.46 ( ).53 rs CC CA AA 1.41 ( ) 1.37 ( ) 1.60 ( ).41 rs GG GA AA 1.41 ( ) 1.41 ( ) 1.47 ( ).58 rs CC CT TT 1.42 ( ) 1.42 ( ) 1.38 ( ).70 rs GG GA AA 1.33 ( ) 1.46 ( ) 1.46 ( ).04 rs GG GA AA 1.43 ( ) 1.36 ( ) 1.65 ( ).80 rs GG GA AA 1.44 ( ) 1.42 ( ) 1.35 ( ).64 rs TT TC CC 1.43 ( ) 1.43 ( ) 1.36 ( ).73 rs GG GT TT 1.42 ( ) 1.36 ( ) 1.53 ( ).29 1 Test for trend in log(psa) by allele dose.

19 SupplementaryTable 10. Per-allele ORs for each SNP in stage 3, for cases with high and low Gleason score (compared with controls). Marker GS < 7 GS 7 + P 1 P 2 rs ( ) 1.05 ( ) rs ( ).79 ( ) rs ( ).91 ( ) rs ( ) 1.07 ( ) rs ( ).85 ( ) rs ( ) 1.06 ( ) rs ( ) 1.03 ( ) rs ( ) 1.16 ( ) rs ( ) 1.19 ( ) rs ( ) 1.02 ( ) rs No observation No observation N/A N/A rs ( ).94 ( ) rs ( ).85 ( ) P-value for the difference in OR between cases with high and low Gleason score (see methods). 2 P-value for the trend in OR by Gleason score, treated as an ordinal trait (see methods).

20 Supplementary Table 11. Per-allele ORs for the association between disease and each SNP in stage 3 by age at diagnosis. Marker Age at diagnosis (years) P-trend 1 < rs ( ) 1.09 ( ) 1.07 ( ) 1.10 ( ) 1.03 ( ).23 rs ( ).75 ( ).69 ( ).78 ( ).85 ( ) 1.1x10-5 rs ( ).87 ( ).90 ( ).90 ( ).94 ( ).18 rs ( ) 1.13 ( ) 1.06 ( ) 1.06 ( ) 1.06 ( ).10 rs ( ).85 ( ).95 ( ).89 ( ).93 ( ).02 rs ( ) 1.06 ( ) 1.03 ( ) 1.11 ( ).95 ( ).09 rs ( ) 1.07 ( ) 1.03 ( ) 1.03 ( ) 1.12 ( ).19 rs ( ) 1.22 ( ) 1.18 ( ) 1.18 ( ) 1.12 ( ).13 rs ( ) 1.24 ( ) 1.22 ( ) 1.24 ( ) 1.12 ( ).006 rs ( ).95 ( ).98 ( ) 1.01 ( ) 1.02 ( ).22 rs ( ).98 ( ) 1.09 ( ).99 ( ).98 ( ).66 rs ( ) 1.03 ( ).98 ( ) 1.02 ( ).96 ( ).25 rs ( ).84 ( ).84 ( ).85 ( ).88 ( ) df trend test for trend in OR by age, using case only analysis.

21 Supplementary Table 12. Per-allele ORs for each SNP in stage 3, for cases with and without a family history of PrCa. Marker Family history No Yes rs ( ) 1.09 ( ).25 rs ( ).76 ( ).82 rs ( ).91 ( ).69 rs ( ) 1.05 ( ).27 rs ( ).86 ( ).002 rs ( ) 1.07 ( ).63 rs ( ) 1.08 ( ).35 rs ( ) 1.22 ( ).08 rs ( ) 1.27 ( ).10 rs ( ).99 ( ).85 rs ( ).90 ( ).19 rs ( ) 1.00 ( ).99 rs ( ).85 ( ).82 P 1 1 P-value for the difference in OR between cases with and without a family history (see methods).

22 Supplementary Table 13. Estimated power to detect each association at P<10-7, based on the estimate OR in stage 3 1. SNP Power rs % rs % rs % rs % rs % rs <0.1% rs % rs % rs % rs % 1 Based on cut-off of P<.059 after stage 1 and P<10-6 after stage 2, and effective sample sizes of 1,874; 1,536; and 15,109 cases and controls in stages 1-3.

23 DETAILED DISCUSSION OF CANDIDATE GENES IN SEVEN NEW LOCI (see also Figure 1) rs on chromosome 2p21 is in intron 30 of THADA, a gene disrupted by rearrangement in thyroid adenomas and part of the death receptor pathway 1, 2. Another SNP, rs is in exon 24 of THADA and has been associated with type 2 diabetes 3. Alleles in TCF2 and JAZF1 have also been shown to be susceptibility loci for both diabetes and PrCa. This locus may provide another example of this phenomenon. rs on 2q31 is in intron 1 of ITGA6, the gene encoding integrin alpha 6. Integrins control cell attachment to the extracellular matrix and mediate cell proliferation, migration and survival. ITGA6 may also be considered as a potential therapeutic target, as its up-regulation is associated with a metastatic phenotype and an increase in cancer cell motility, including for PrCa 4. rs and rs on 4q22 are in introns 7 and 9, respectively, of PDLIM5 5. LIM domain-containing proteins are scaffolds involved in cytoskeleton organization, cell lineage specification, organ development, and oncogenesis. rs on 4q24 lies 90kb upstream of TET2, coding for a DNA binding zinc finger protein. TET2 is widely expressed, with highest expression levels found in prostate and bone marrow, and mutations of TET2 have been reported in myelodysplasia cases 6. rs and rs are 90kb apart on 8p21. Thus, they may be markers for the same causal variant, but given that they are in separate LD blocks, they could also indicate the presence of two distinct causal alleles. An interesting candidate gene in this general region is NKX3.1 (rs is 10kb downstream of NKX3.1), which codes for the androgenregulated homeobox protein NKX3.1. This is important for maintenance of normal prostate tissue and is expressed at all stages of prostate development and in adult prostate 7. Loss of heterozygosity at 8p21 is frequently observed in early PrCa. NKX3.1 is up-regulated by androgens and simultaneous loss of NKX3.1 and PTEN is common in PrCa initiation 8. NKX3.1 stabilises TP53 via Akt independent mechanisms 8 ; it associates with HDAC1, preventing deacetylation and destabilisation of TP53 by HDAC1/MDM2, thereby promoting apoptosis. Of note, HDAC inhibitors have been developed as targeted therapies and are currently in trial in metastatic PrCa 9. NKX3.1 also downregulates expression of PSA by PDEF 10 (an epithelium-specific Ets transcription factor, which plays a role in PrCa), therefore loss of NKX3.1 results in increased PSA expression. rs is in a region on 11p15 that includes several plausible candidate genes: IGF2, IGF2AS, INS and TH. The first three are members of the insulin family of polypeptide growth factors 11. A SNP in H19, a regulator of IGF2 that lies telomeric to this region, has been associated with breast cancer risk 12. rs lies on chromosome 22q13, for which evidence of linkage to PrCa has been found 13,14, 15.

24 This SNP is ~7Mb downstream from the previously reported linkage peak. There are several genes of interest in the LD block (see figure 1). SUPPLEMENTARY TEXT: DESCRIPTION OF THE PRACTICAL CONSORTIUM GROUPS Subjects were included from 21 studies comprising PrCa cases and controls: eight from Europe, nine from North America, one from China, and three from Australia. These comprised 16, 332 PrCa cases and 16, 344 male controls. Details are shown in supplementary Table 4. Some have previously been described as part of work in the PRACTICAL Consortium 16. The Mayo Clinic and Utah studies oversampled cases from multiple case families and only one case per family was genotyped. Three of the studies contained men of Asian ancestry (China and Japan); six of the studies contained men of African American ancestry and two studies contained men of Latino (Hispanic) origin; one study was from Hawaii (see Figure 2). The remaining studies were predominantly of men of European ancestry. All studies have the relevant IRB approval in each country in accordance with the principles embodied in the Declaration of Helsinki. Details of each study set are given below, and a summary of the studies is given in Supplementary Table 4. BiPAS: The Birmingham Prostatic Neoplasms Association Study (BiPAS) A Genetic and Environmental Case Control Study The Birmingham Prostatic Neoplasms Association Study base consists of men living in the south Birmingham area, United Kingdom aged >50 years. The study recruited men with lower urinary tract symptoms (LUTS) and/or high serum prostate specific antigen (PSA) levels referred for prostate biopsies between March 2007 until October PrCa cases were recruited from the Queen Elizabeth Medical Centre, Birmingham. Cases are defined as men with histologically confirmed adenocarcinoma of the prostate. Controls were also recruited from the Queen Elizabeth Medical Centre and Selly Oak Hospital, Birmingham. Men with a normal repeat PSA and a negative biopsy were categorized as benign controls. A blood sample from every hospital based subject was obtained using standard venepuncture methods, and collected in a 5ml tube containing EDTA. Samples were transported to the laboratory immediately in a cool bag with cool packs and stored at 4 C. DNA was extracted using the QIAGEN maxi blood kit. CHSH: China

25 All samples were from men of Chinese Han origin from Shanghai and its surrounding city. Patients with PrCa enrolled in the study were diagnosed by transrectal ultrasonographic prostate biopsy, and confirmation of the pathologic diagnoses for those who underwent radical prostatectomy. They were from two hospitals; Changhai hospital and Changzheng Hospital in Shanghai, China. Controls are hospital based and are from other clinical divisions which are treating non cancer patients; these have a PSA level of < 3.0ng/ml, and are age matched (+/-3 years). FHCRC: Fred Hutchinson Cancer Research Center, Seattle USA The study population consists of participants from two population-based case-control studies in Caucasian and African American residents of King County, Washington (Study I and Study II), which have been previously described. Incident cases with histologically confirmed PrCa were ascertained from the Seattle-Puget Sound Surveillance, Epidemiology and End Results cancer registry. In Study I, cases were diagnosed between January 1, 1993, and December 31, 1996 and were years of age at diagnosis. In Study II, cases were diagnosed between January 1, 2002, and December 31, 2005 and were years of age at diagnosis. Overall, 2,244 eligible PrCa patients were identified and 1,754 (78%) were interviewed. Blood samples yielding sufficient DNA for genotyping were drawn from 1,457 (83%) cases who completed the study interview. A comparison group of controls without a history of PrCa, residing in King County, Washington, was identified for each study using random digit telephone dialing. Controls were frequencymatched to cases by five-year age groups and recruited evenly throughout each ascertainment period for cases. A total of 2,448 men were identified who met the eligibility criteria and 1,645 (67%) completed a study interview. Blood samples were drawn and DNA prepared from 1,352 (82%) interviewed controls. FMHS: Michigan, US The Flint Men s Health Study (FMHS) is a community-based, case-control study of PrCa in African American men living in Genesee County, MI, US conducted from 1996 to Controls were recruited from a probability sample of African American men aged years with intentional over-sampling from older age groups. Cases were identified from the Genesee County Community-Wide Hospital Oncology Program registry. Participants provided blood samples from which DNA and serum were isolated and completed detailed interviews which addressed potential risk factors for PrCa, urinary symptoms, PrCa screening history and general medical history, socio-economic factors, and access to and use of health care. Additionally, controls underwent urological examinations and PSA screening and cases provided access to

26 medical records pertaining to their PrCa diagnoses. A total of 383 controls participated in all portions of the study. Nineteen of those controls were diagnosed with PrCa during the time of the study and subsequently recruited as cases. DNA is currently available for 356 of the remaining controls. A total of 136 cases participated in all portions of the study and DNA is currently available for 133 cases. HaPCS: Hannover, Germany A hospital-based series of 499 unselected Caucasian patients with PrCa who were treated with brachytherapy between October 2000 and September 2007 at Hannover Medical School, were enrolled for this study 17. All patients had biopsy-proven adenocarcinoma of the prostate. Indication for permanent brachytherapy was clinically localized low risk early PrCa (ct2a or less with a PSA serum level < 10 ng/ml and a Gleason score < 7) following the European Society for Therapeutic Radiology and Oncology/European Assocation of Urology/European Organization for Research and Treatment of Cancer recommendations. The median age at diagnosis was 67 years in this patient series (range years). For comparison, a series of 504 genomic DNA samples was established from ethnically matched adult male blood donors at Hannover Medical School in the period from MAYO: The MAYO clinic: Rochester, Minnesota, USA The Mayo Clinic study consisted of hospital-based cases, including 476 affected men from 185 families with PrCa, 445 men with sporadic PrCa, 199 with aggressive (Gleason score > 7) PrCa, and 500 population-based controls. The controls (all males) were randomly selected from a sampling frame of Olmsted County, Minnesota, provided by the Rochester Epidemiology Project. The methods used to ascertain familial and sporadic PrCa patients, as well as controls, have been described previously 18. All individuals from the Mayo Clinic study included in this report were of self-reported European descent. MCCS: Melbourne Collaborative Cohort Study, Melbourne, Australia MCCS/RFPCS/EOPCFS: Cancer Council Victoria, Melbourne, Australia The Melbourne Collaborative Cohort Study (MCCS) is a prospective cohort of 17,154 men aged 40 to 69 years at recruitment in MCCS participants were diagnosed with PrCa during follow-up to mid 2006 and were ascertained through linkage with the Victorian Cancer Registry and the National Cancer Statistics Clearing House that includes diagnoses from other States in Australia. The Risk Factors for Prostate Cancer Study (RFPCS) is a population based casecontrol study that in the period recruited through State Cancer Registries men resident in Perth and Melbourne diagnosed with prostate cancer at age less than 70 years. The

27 Early-Onset Prostate Cancer Family Study (EOPCFS) is a population-based study of prostate cancer in men diagnosed at age less than 56 years ascertained through the Victorian Cancer Registry. Controls for stage 3 were a random sample of the MCCS participants that were not diagnosed with PrCa during follow-up. All study subjects were of Caucasian origin. Participants in the three studies whose samples were used for stage 2 were excluded from stage 3. MEC: The Multiethnic Cohort Study The Multiethnic Cohort Study 19 is a population-based prospective cohort study that was initiated between 1993 and 1996 and includes subjects from various ethnic groups -African-Americans and Latinos primarily from California (mainly Los Angeles) and Native Hawaiians, Japanese- Americans, and European Americans primarily from Hawaii. State drivers license files were the primary sources used to identify study subjects in Hawaii and California. Additionally, in Hawaii, state voter's registration files were used, and, in California, Health Care Financing Administration (HCFA) files were used to identify additional African American men. All participants (n=215,251) returned a 26-page self-administered baseline questionnaire that obtained general demographic, medical and risk factor information. In the cohort, incident cancer cases are identified annually through cohort linkage to population-based cancer Surveillance, Epidemiology, and End Results (SEER) registries in Hawaii and Los Angeles County as well as to the California State cancer registry. Information on stage and grade of disease are also obtained through the SEER registries. Blood sample collection in the MEC began in 1994 and targeted incident PrCa cases and a random sample of study participants to serve as controls for genetic analyses. This nested PrCa case-control study in the MEC consists of 2,792 invasive PrCa cases and 2,377 controls. This study was approved by the Institutional Review Boards at the University of Southern California and at the University of Hawaii and informed consent was obtained from all study participants. MOFFITT: Moffitt Study, Tampa, Florida, US This is a hospital-based incident study of 646 patients with primary adenocarcinoma of the prostate (560 whites, 55 African Americans, 28 white Hispanics and 3 others). They were recruited from 2002 to 2007 at the H. Lee Moffitt Cancer Center (Tampa, FL, US) and James A. Haley Veterans Affairs Hospital (Tampa, FL, US). Ninety-five percent of the case subjects who were asked to participate in the study agreed. All cancer cases were histologically confirmed by the Department of Pathology at each institution.

28 The controls consisted of 320 subjects (302 whites, 10 African Americans, 6 white Hispanics and 2 others) who were visiting the Lifetime Cancer Screening Center, which is affiliated with the H. Lee Moffitt Cancer Center. At this center, routine screenings are offered to men for cancers of the prostate, colorectum, and skin. Men could have been self-referred or directed for screening by their primary healthcare provider. All control subjects were male and had had no previous diagnosis of cancer. The control subjects were frequency matched to the patients by age at diagnosis (± 5 years). Eighty-three percent of the control subjects who were asked to participate in the study consented. Non-genetic risk factor data for the present study were obtained through in-person interviews with the patients and controls at enrollment. The questionnaire covered demographic information, family history of cancer (ie, whether they have one or more first-degree family member with PrCa), medical history, and detailed tobacco consumption. For the patients, data on cancer stage, Gleason score, and prostate specific antigen level were abstracted from the medical records. The subjects were asked to provide a blood or buccal sample after the interview as a source of genomic DNA 20. NC_CCPC: San Francisco, California USA This population-based case-control study of advanced PrCa in non-hispanic white and African- American men was conducted in the San Francisco Bay Area. Newly diagnosed cases aged years were identified through the regional cancer registry, which is part of the Surveillance, Epidemiology and End Results (SEER) Cancer Registry Program. Non-Hispanic white cases were diagnosed between July, and February 28, 2000, and African-American cases were diagnosed between July 1, 1997 and December 31, Overall, 1,015 patients with a first primary advanced PrCa were identified. Of these, 106 were deceased at the time of contact, 33 were enrolled in another study and thus not available, 12 were declined contact by their physician, and 76 no longer lived in the San Francisco Bay area or did not meet other eligibility criteria. Of 788 eligible cases contacted, 568 (72%) completed the interview and 533 (68%) provided a biospecimen sample. DNA from blood samples was available for 389 cases. Non-Hispanic white and African-American population controls aged years were identified through random-digit dialing. In addition, controls aged years were randomly selected from the rosters of beneficiaries of the Health Care Financing Administration (HCFA). Controls were frequency-matched to cases by five-year age group and race. Of 1,081 controls selected into the study 16 were deceased at the time of contact, 41 had a history of PrCa, and 156 did not meet other eligibility criteria. Of 868 eligible controls contacted, 545 (63%) completed the interview and 525 (60%) provided a biospecimen sample. DNA from blood samples was available for 256 controls [John et al., 2005].

29 PCMUS: Bulgaria The Bulgarian sample of PrCa patients consist mainly of newly diagnosed cases, which are histopathologically confirmed. The patients (N=114, age range 51-91) are of Bulgarian origin. Transrectal biopsy was performed at the Urology Clinic, Alexandrovska University Hospital, mainly because of an elevated PSA. Some of the patients were referred from other centers to the tertiary university hospital after being previously diagnosed with PrCa. A small subset of patients had previously had definitive treatment (mainly radical prostatectomy), and they were called retrospectively for invitation to join the study. The control group is matched to the patients by sex, age, and ethnicity. It consists of two groups: (i) 96 healthy males, age range 51-86, presenting to our institution with lower urinary tract symptoms caused by benign prostatic hypertrophy (BPH) who had a PSA <3.5. The majority of them subsequently underwent surgical treatment with histological verification of the BPH; (ii) an additional healthy control group of 110 anonymous males matched to the PrCa patients by age and ethnicity, but with no PSA data. All participants gave written informed consent and anonymous controls have been selected from the available DNA bank at Molecular Medicine Center, Medical University Sofia. PENN, (SCORE), University of Pennsylvania, Philadelphia, US Incident PrCa cases were identified through Urologic Clinics between 1995 and Case status was confirmed by medical records review using a standardized abstraction form. Cases were excluded from this study if they reported having exposure to finasteride (Proscar) at the time of their PrCa diagnosis. Patients who were non-incident cases (i.e., those diagnosed more than twelve months prior to the date of study ascertainment), or had a prior diagnosis of cancer at any site except non-melanoma skin cancer, were also excluded. Risk factor, medical history, PrCa screening history, and PrCa diagnostic information was obtained by using a standardized questionnaire and review of medical records. Information collected included personal history of benign prostatic hyperplasia (BPH) and vasectomy, previous cancer diagnoses, and demographic information, and PrCa screening history. Existence of BPH was confirmed by medical records review. Genomic DNA for the present study was self-collected by each study participant using sterile cheek swabs (Cyto-Pak Cytosoft Brush, Medical Packaging Corporation, Camarillo, CA), and processed using either a protocol modified from Richards et al. (1993) 21 as described previously 22 or using a modified protocol on the Qiagen 9604B robot with the QIAamp 96 DNA Buccal Swab Biorobot Kit (Valencia, CA).

30 ProtecT, UK ProtecT [ Donovan et al, 2003] is a national study of community-based PSA testing and a randomised trial of subsequent PrCa treatment. Approximately 200,000 men between the ages of 50 and 69 years, ascertained through general practices in nine regions in the UK, were approached and 100, 000 were recruited. Men known to be non-white were excluded. For this study, 1800 cases identified by PSA screening within the ProtecT study were analysed. Controls (1800) with normal PSA levels (<3ng/ml) were selected from the same GP register and 5 year age band as the cases. QUEENSLAND: Australia The Queensland cases were ascertained from two studies: (i) a series of men recruited through QUT/QIMR within two years of their diagnosis of prostate cancer (Retrospective Queensland Study), and identified through physician referrals from three hospitals in Brisbane, Queensland (N=176, age range years) (ii) a longitudinal cohort study (Prostate Cancer Supportive Care and Patient Outcomes Project: ProsCan) being conducted through CCQ in Queensland, through which men newly diagnosed with prostate cancer from 26 private practices and 8 public hospitals were directly referred to ProsCan at the time of diagnosis by their treating clinician (N=780, age range years) 23. All cases had histopathologically confirmed prostate cancer, following presentation with an abnormal serum PSA and/or lower urinary tract symptoms. Controls, recruited through QUT and QIMR, comprised healthy male blood donors with no personal history of prostate cancer, from the (i) the Australian Red Cross Blood Services in Brisbane (N=834, age range years) 24 and (ii) the Australian Electoral Commission (N=559, age and post-code/ area matched to ProsCan, age range years). TAMPERE: Finland Samples (3543 total) were collected in Tampere and Helsinki and are all of Finnish origin. The mean age of diagnosis for the 1485, unselected consecutive PrCa patients from Tampere was 69 years (range 43-95). The patients were diagnosed with PrCa in in the Tampere University Hospital, Department of Urology. Tampere University Hospital is a regional referral center in the area for all patients with PrCa, which results in an unselected, population-based collection of patients. The rest of the cases, 284 men, were diagnosed in in the Finnish PSA-screening Trial from the Hospital district of Helsinki and Uusimaa. Their mean age at diagnosis was 66 years (range 62-74) For controls, two groups of samples were used: (i)the 835 Finnish population controls consisted of DNA samples from volunteer, anonymous, healthy male blood donors obtained from the Blood Center of the Finnish Red Cross in Tampere, and (ii) the 938 men with a PSA of <1ng/ml were selected from the PSA-screening trial.

31 TASPRAC: Tasmania, Australia The thirty-one familial PrCa cases genotyped for this study were drawn from the Tasmanian Familial Prostate Cancer Study, which has recruited families with multiple individuals affected with PrCa. These families were identified using the records of the Tasmanian Cancer Registry (a register of all cancer diagnoses in Tasmania, Australia since 1978) and the genealogical records from the Menzies Research Institute Genealogical Database; they comprise at least 2 affected first-degree relatives. One case per family was selected for inclusion. Blood samples, pathology specimens and pathology reports are available for these familial cases. The remaining 492 sporadic PrCa cases were drawn from the Tasmanian Prostate Cancer Case Control Study. Cases were again identified from the Tasmanian Cancer Registry, and eligible cases were men <70 years diagnosed with histologically confirmed PrCa diagnosed between 1996 and Controls were randomly selected from the Tasmanian electoral roll (registration on the electoral roll is compulsory in Australia for individuals > 18 years). Eligible controls were sex and age-matched within 5-year age groups to the sporadic cases and self-reported as unaffected with PrCa. Blood samples, physical measures, dietary history, environmental exposure data and family history have been collected from participating individuals. Of note, controls diagnosed with PrCa subsequent to their recruitment have been removed from the control dataset. Notification of their subsequent diagnosis with PrCa was determined by their registration as a confirmed case by the Tasmanian Cancer Registry as of the end of , 26 UKGPCS, UK Samples were ascertained through the UKGPCS (UK Genetic Prostate Cancer Study) and through a systematically collected series from PrCa clinics in the Urology Unit at The Royal Marsden NHS Foundation Trust over a 14 year period. Controls were collected as part of the Gene-Environment Interactions in Prostate Cancer or The Prostate Cancer Research Foundation studies run from the University of Nottingham from GP practices participating in the UKGPCS. They had no personal or family history of PrCa. ULM: Germany Cases were recruited in two different ways. Familial PrCa probands (index cases) were ascertained from all over Germany. They were advised by their attending physicians to contact the Clinic of Urology of Ulm. The positive family history was then verified by reviewing medical records or death certificates of family members. In each case, only one member of each family (e.g. the proband) was enrolled in the present study. Sporadic cases, who reported no relatives affected with PrCa, were almost exclusively collected at Ulm during their course of treatment