Abstract

Background

Colorectal cancer is a major global public health problem, with approximately 950 000 patients newly diagnosed each year. We report the first comprehensive field synopsis and creation of a parallel publicly available and regularly updated database (CRCgene) that catalogs all genetic association studies on colorectal cancer ( http://www.chs.med.ed.ac.uk/CRCgene/ ).

Methods

We performed two independent systematic reviews, reviewing 10 145 titles, then collated and extracted data from 635 publications reporting on 445 polymorphisms in 110 different genes. We carried out meta-analyses to derive summary effect estimates for 92 polymorphisms in 64 different genes. For assessing the credibility of associations, we applied the Venice criteria and the Bayesian False Discovery Probability (BFDP) test.

Results

We consider 16 independent variants at 13 loci ( MUTYH, MTHFR, SMAD7, and common variants tagging the loci 8q24, 8q23.3, 11q23.1, 14q22.2, 1q41, 20p12.3, 20q13.33, 3q26.2, 16q22.1, and 19q13.1) to have the most highly credible associations with colorectal cancer, with all variants except those in MUTYH and 19q13.1 reaching genome-wide statistical significance in at least one meta-analysis model. We identified less-credible (higher heterogeneity, lower statistical power, BFDP >0.2) associations with 23 more variants at 22 loci. The meta-analyses of a further 20 variants for which associations have previously been reported found no evidence to support these as true associations.

Conclusion

The CRCgene database provides the context for genetic association data to be interpreted appropriately and helps inform future research direction.

Colorectal cancer is a major global public health problem, with approximately 950 000 patients newly diagnosed each year ( www.who.int ; International Agency for Research on Cancer GLOBOCAN). The risk of developing colorectal cancer increases steeply with age and its incidence is rising in many industrialized countries as life expectancy and the number of elderly people increase. Incidence is also rising in many developing countries as diet and lifestyle change and increasingly resemble those in industrialized countries. Overall, 5-year survival of colorectal cancer remains at best only approximately 50%. Because of the rising disease burden, health care costs, including costs for diagnosis and treatment, are substantial.

Approximately one-third of variance in colorectal cancer is attributed to inherited genetic factors ( 1 ) and disease risk is increased by two- to fourfold for first-degree relatives of patients. Excess risk of familial cancer can be accounted for by a combination of rare high-penetrance mutations and large numbers of common genetic variants, each of which confers a small risk. These variants combine to confer a range of susceptibilities in the population ( 2 ).

The associations between genetic variants and human diseases described so far were highly dependent on the study designs used to identify them. Highly penetrant mutations with large effects have been identified in a number of genes responsible for heritable colorectal cancer susceptibility syndromes (eg, DNA mismatch repair genes, APC , SMAD4 , LKB1/STK11 , MUTYH, and a linkage region on 15q) and these mutations account for about 3% of incident colorectal cancers in industrialized countries. Identification of further high-penetrance loci has proven intractable to genetic linkage approaches, probably because of locus heterogeneity, and confounding, because of the segregation of multiple alleles in collected families. Association studies conducted in general population samples using common genetic markers have typically identified variants with very small effects. To date, genome-wide association studies (GWASs) have reported 14 common genetic variants that influence the risk of colorectal cancer and account for approximately 6% of the excess familial risk ( 2–10 ). Future resequencing studies are expected to identify rarer variants (eg, with 0.05%–5% prevalence) with intermediate, or perhaps even large, effects ( 2 ). GWASs of structural variations will likely identify deletions, amplifications, and other copy-number variations that may also influence risk of colorectal cancer.

For this report, we have undertaken a comprehensive review of genetic factors that appear to be associated with colorectal cancer by using previously published guidelines for the assessment of cumulative evidence on genetic association studies ( 11 , 12 ) and following a format similar to previous overview meta-analyses ( 13–15 ). We have cataloged all genetic association studies published in this field and conducted meta-analyses of variants with genotypes available in four or more independent case-control studies; any variants that had been typed in two large GWASs were also included in the meta-analyses. The results of the search strategy and meta-analyses are publicly available on a regularly updated Internet database (CRCgene). This represents the first attempt to systematically capture all published genetic association data for colorectal cancer and conduct a meta-analysis.

The core aim is to provide an up-to-date systematic review of the state of the art across the field of colorectal cancer genetics for the research community. We conducted a critical review of all published “candidate gene” study data, incorporating relevant candidate gene data from the GWASs available to us, and then performed meta-analyses of these data. This approach enabled us to summarize available evidence from larger sample sizes, thus gaining greater precision in odds ratio estimates. We have presented these data within a defined statistical and causal inference framework to aid correct interpretation of data ( 16 ).

The broader medium-term aim of this work is to identify genetic variants for which there is robust evidence of influence on risk of colorectal cancer. This will help to inform future research efforts and to identify variants that can serve as a basis for providing risk estimates for population groups. It will also provide new insights into the fundamental biological mechanisms involved in colorectal carcinogenesis.

Methods

Literature Search and Data Collection

The first step was to undertake a comprehensive systematic literature review of all current published data on genetics and colorectal cancer. To identify gene association studies aimed at risk of colorectal cancer, we used the Medline database via the Ovid gateway and the search terms comprising medical subject headings (MeSH) and keywords relating to colorectal neoplasms, the MeSH heading “genetic predisposition to disease,” and the keywords “gene$” and “associate$” were applied to terms in the entire article. We cross-checked these 10 125 findings against those listed in the HuGENet phenopedia ( 17 ). Of the 10 145 articles thus identified, first we screened the abstracts for eligibility, and then, if necessary, the full texts. We used the following inclusion and exclusion criteria. The article must have evaluated the association between a polymorphic genetic variant [one with a minor allele frequency (MAF) ≥0.01 in the general population based on the data on the reference panel of the 1000 GenomesTable 1; ( 18 )] and sporadic colorectal cancer. Studies that examined associations with only premalignant conditions such as adenomas, polyps, or dysplastic tissue were not included. In addition, studies of hereditary colorectal cancer syndromes, such as familial adenomatous polyposis, hereditary nonpolyposis colorectal cancer, juvenile polyposis syndrome, and Gardner’s syndrome were excluded because our focus was on sporadic colorectal cancer. All studies needed to relate to human participants; any study that was concerned solely with investigating the progression or histological phenotype of colorectal cancer was excluded. Case–control studies and appropriate cohort and GWASs were included. The study had to be published in English in a peer review journal before June 30, 2010. For variants that were identified through the GWAS, the search was repeated and extended until March 31, 2011. Any research that had only been reported in abstracts (eg, presented in scientific conferences yet to be fully published) was excluded. Nine family-based studies were also excluded. We generated a list with all variants to be summarized using meta-analysis and compared it with a list of variants that were included in two GWASs, from Scotland and Canada, for which we had access to individual-level data. The Scottish study comprised patients aged 16–79 years who were diagnosed with colorectal cancer throughout Scotland during the period 1999–2006 ( 6 , 19 ). Potentially eligible population-based control subjects were selected through the Community Health Index, a national register of all individuals registered with a general practitioner in Scotland, and individually matched with patients on age (± 2 years), sex, and area of residence. The Canadian study was based on patients aged 20–74 years who were diagnosed during the period July 1, 1997, to June 30, 2000, who were identified and recruited from the population-based Ontario Cancer Registry ( 3 , 20 ). Random digit dialing was used to select population-based control subjects who were frequency-matched with patients on sex and 5-year age group. In both of these studies, DNA from blood was genotyped using multiple platforms; here, we included data obtained in both studies from the Illumina HumanHap300 and HumanHap240S arrays on the Infinium platform. If a variant was found to be included in either of these GWASs, then genotype counts were included in the meta-analysis of this variant. For the other studies, we contacted the authors of studies with missing data, with an approximate successful response rate of 0.2.

Data Entry, Management, and Abstraction

Once the search was completed, the references of the articles in the search were entered into a web-based database, “RefWorks” ( http://www.refworks.com/ ), thus freezing the search at that point in time. Their abstracts and/or full texts were screened to assess their eligibility for inclusion in the field synopsis. Review articles and meta-analyses on genetic associations of colorectal cancer were also kept alongside so that the references they used could be screened for eligibility if they had been missed in the Medline search.

Data from all studies that met the final inclusion and exclusion criteria were abstracted into two standardized tables. The first of these listed study characteristics and the second table listed allele and genotype frequencies. We abstracted key variables with regard to the study identifiers and context, study design and limitations, intervention specifics, and outcome effects.

Statistical Analysis

Statistical analysis was conducted using Intercooled STATA, version 11.0 ( 21 ). Meta-analysis was performed for all variants with case-control data available from four or more independent samples. We obtained summary crude odds ratios (ORs) and 95% confidence intervals (95% CI) for two additive models (variant [var]/wild-type [wt] vs wt/wt; and var/var vs wt/wt), one recessive model (var/var vs var/wt and wt/wt), and one dominant model (var/var and var/wt vs wt/wt). We applied either the fixed-effect model (Mantel–Haenszel method) or, in case of heterogeneity, the random-effect model (DerSimonian–Laird method). Between-study heterogeneity was quantified by calculating the Q statistic, with a P value less than .05 being the threshold. We also calculated the I2 heterogeneity metric and its 95% confidence interval (CI). Although sometimes we summarized studies that were very heterogeneous, we recognized that because of the variation in study methods and outcome definitions, the meta-estimates should be interpreted cautiously. To assess any small-study effects, we performed funnel plot analyses and tested for statistical significance using the Harbord modification of the Egger test, implemented in STATA ( http://ideas.repec.org/a/tsj/stataj/v9y2009i2p197-210.html ). A negative result from small study–effects testing does not entirely exclude publication bias. In addition, the test for small-study effects may be underpowered if there are approximately 10 or fewer studies and it may be inappropriate in the presence of large heterogeneity ( 22 ). We also estimated the power that each meta-analysis had to detect a statistically significant effect by using the Power and Sample Size Program ( 23 ) and specifying α = .05 as the level of statistical significance and the effect sizes and allele frequencies estimated from the meta-analyses (an integral component of the BFDP analysis).

The relative risk of colorectal cancer in a sibling that was attributable to a given single-nucleotide polymorphism (SNP) was calculated using the following formula ( 24 , 25 ):

 
(1)
formula

where p is the population frequency of the referent allele, q = 1– p , and r1 and r2 are the relative risks (estimated as ORs from the meta- analyses) for heterozygotes and variant homozygotes, relative to wild-type homozygotes. Assuming a multiplicative interaction, we calculated the proportion of the familial risk attributable to an SNP as log(λ*)/log(λ 0 ), where λ 0 is the overall familial relative risk estimated from epidemiological studies, assumed to be 2.2. Although family studies are normally used to estimate the relative risk for siblings, this formula can be used in population-based studies ( 24 , 25 ).

Finally, we repeated the analysis by excluding any studies that were conducted in a nonwhite population (with white populations defined as Europeans, North Americans, and Australians) for the SNPs that were found to be statistically significantly associated with risk of colorectal cancer in any of the genetic models at a threshold level of α = .05.

Credibility of Genetic Association

To assess the credibility of genetic associations, we considered the BFDP ( 26 ) and the Venice criteria ( 11 , 12 ). The BFDP assesses the noteworthiness of an observed association. For variants that were found to be statistically significantly associated with risk of colorectal cancer in any of the genetic models (at P < .05), the BFDP was estimated using the Excel Calculation Spreadsheet ( http://faculty.washington.edu/jonno/cv.html ). The BFDP threshold for noteworthiness was set up to be equal to 0.20, based on the assumption that a false discovery would be four times more costly than a false nondiscovery. We chose to calculate BFDP values for two levels of prior probabilities: at a medium or low prior level (0.05 to 10 –3 ) that would be close to what would be expected for a candidate gene; and at a very low prior level (10 –4 to 10 –6 ) that would be close to what would be expected for a random SNP.

With regard to the Venice criteria, we operationalized the criterion of volume of evidence on the basis of statistical power to detect an association of the desired magnitude: A, 80% or more; B, 50%–79%; or C, less than 50%. For replication, we used the I ² criterion proposed by Ioannidis et al. ( 11 ). For protection against bias, we considered that complete reporting was problematic. The phenotype definition was addressed by our inclusion criterion ––namely, that case subjects would have newly incident colorectal cancer. In general, genotyping error rates are low ( 27 ), and the criterion of replication across studies in part addressed potential concern about variation in genotyping quality between studies; some genotyping issues are discussed in relation to specific findings below. Whereas population stratification may impact gene discovery ( 28 , 29 ), the effect on the magnitude of association in general appears to be small ( 30 , 31 ).

We classified the genetic associations in three categories according to the findings after the BFDP analysis and the application of the Venice criteria. Associations were classified as positive if they fulfilled the following criteria: 1) they were statistically significant (at P < .05) in at least two of the genetic models; 2) they had a BFDP less than 0.20 (at least when P < .05); 3) they had a statistical power greater than 80%; and 4) they had an I2 less than 50%. Associations were classified as less-credible positives if they were statistically significant ( P < .05) in at least one of the genetic models, their BFDP was greater than 0.20, and their statistical power was between 50% and 79% ( I2 ranged from 0% to 89% , but this criterion was not taken into account for this category). All other associations were classified as negative.

Model-free Approach

For those SNPs that were identified as true positives after applying the BFDP and Venice criteria, we applied the model-free meta-analysis approach described by Thompson et al. ( 32 ). This model uses a maximum likelihood estimator, assumes a fixed-effect meta-analysis and is similar to a bivariate meta-analysis of the effects for the heterozygotes and variant homozygotes. It gives an estimate of the lambda (λ), which represents the underlying genetic model (and it represents the heterozygote effect as a proportion of the homozygote effect), with its confidence interval limits. If the confidence interval of λ is too wide, there is not enough power to estimate λ. The value of λ is not restricted. Values equal to 0, 0.5, and 1.0 correspond to the recessive, additive, and dominant genetic models, respectively. Values greater than 1 or smaller than 0 suggest heterosis, which is when the risk of the heterozygotes is higher or lower, respectively, than the risk of either of the homozygote genotypes. However, heterosis is relatively uncommon and λ values greater than 1 or smaller than 0 could represent chance fluctuation, reflecting low statistical power to estimate the magnitude of effect for heterozygotes.

Results

Literature Search and Data Collection

After screening approximately 10 145 titles and abstracts ( Supplementary Table 1 , available online) dated until June 30, 2010 (or March 31, 2011 for GWAS-identified hits), we identified and extracted data from 635 publications reporting on 445 polymorphisms in 110 different genes. More than 9231 (88%) of these studies were published after 1995, and about half of those were published during the past 3 years.

Table 1.

List of genes and variants that were selected for meta-analysis (sorted by gene pathway)*

Genes, by category Variant rs number Case vs control subjects (number of samples) Ref allele Ref allele frequency of case subjects Ref allele frequency of control subjects Attributable familial risk Result of most recent meta-analysis; case subjects/ control subjects (samples) (reference) Other meta- analyses Source of minor allele frequency 
Adhesion molecules 
CDH1 C-160A rs16260 7493 vs 7329 (5†) 0.74 0.72 0.19%  No assoc ( 63 )  — 1000 Genomes 
MMP1 G-1607GG rs1799750 1007 vs 1032 (5) 0.44 0.47 0.62%  Positive assoc; 1343/1590 (7) ( 65 )   ( 64 )  Current study 
MMP3 AAAAA-1612AAAAAA rs3025058 857 vs 932 (4) 5A 0.38 0.41 0.03%  No assoc; 1296/1543 (7) ( 65 )   ( 64 )  Current study 
MMP9 1562C/T ‡ rs3918242 575 vs 836 (4) — — —  No assoc; 575/836 (4) ( 64 )   ( 64 )  1000 Genomes 
Alcohol metabolism 
ADH1B Arg47His rs1229984 1931 vs 2898 (5) His 0.75 0.76 0.32% n/a — 1000 Genomes 
ADH1C Ile349Val (1045A>G) rs698 3168 vs 6229 (7) Ile (A) 0.69 0.64 0.14% n/a — 1000 Genomes 
ALDH2 Glu487Lys rs671 2209 vs 3383 (8) Glu (G) 0.76 0.74 0.02%  Inverse assoc; 1960/3163 (6) ( 66 )  — 1000 Genomes 
Angiogenesis 
VEGF 936 C>T rs3025039 1317 vs 1192 (4) 0.84 0.83 0.60% n/a — 1000 Genomes 
VEGF G634C § rs2010963 1508 vs 1308 (4) — — 0.00% ||  No assoc; 1508/1308 (4) ( 131 )  — 1000 Genomes 
Base-excision repair 
MGMT Leu84Phe ¶ rs12917 1524 vs 4646 (5) 0.89 0.88 0.14%  No assoc; 1524/4646 (5) ( 137 )  — 1000 Genomes 
MGMT Il3143Val ¶ rs2308321 1326 vs 3520 (4) 0.89 0.87 0.19%  No assoc; 1326/3520 (4) ( 137 )  — 1000 Genomes 
MUTYH G396D rs36053993 26 592 vs 19 207 (15) 0.99 0.99 0.87% Positive assoc; 25 616/18 521 (15) (48)  ( 138 , 139 , ( 140 , 141 )  Current study 
MUTYH Y179C rs34612342 26 370 vs 19 042 (15) 1.00 1.00 0.00% Positive assoc; 25 392/18 362 (15) (48)  ( 138 , 139 , ( 140 , 141 )  Current study 
OGG1 Ser326Cys rs1052133 4713 vs 6165 (9) Ser (C) 0.71 0.65 0.02% n/a — 1000 Genomes 
XRCC1 Arg194Trp rs1799782 6635 vs 8488 (11†) 0.91 0.88 0.16% No assoc; 1709/3233 (9) (70)  ( 69 )  1000 Genomes 
XRCC1 Arg280His rs25489 3114 vs 3679 (5) 0.94 0.95 0.04% No assoc; 931/1547 (4) (70) — 1000 Genomes 
XRCC1 Arg399Gln rs25487 7247 vs 8786 (12†) 0.67 0.67 0.19% No assoc; 2776/4402 (14) (70)  ( 69 )  1000 Genomes 
XRCC3 Thr241Met rs861539 4484 vs 5235 (10†) 0.72 0.71 0.00% No assoc; 3183/3926 (7) (69) — 1000 Genomes 
Inflammation or immune response 
IL6 174G>C rs1800795 6676 vs 7942 (10†,#) 0.61 0.60 0.05% n/a — 1000 Genomes 
IL8 251T/A rs4073 3228 vs 3772 (7†) 0.54 0.54 0.02% n/a — 1000 Genomes 
IL10 1082G/A rs1800896 2964 vs 3621 (5†) 0.52 0.51 0.04% n/a — 1000 Genomes 
PPAR-ɣ C1431T rs3856806 5574 vs 7035 (7†) 0.87 0.87 0.08% No assoc; 486/941 (3) (106)  ( 107 )  1000 Genomes 
PPAR-ɣ Pro12Ala rs1801282 15 091 vs 18 690 (17†,#) 0.88 0.88 0.06%  Inverse assoc; 6878/9391 (10) ( 106 )  — 1000 Genomes 
PTGS2/COX2 A1195G rs689466 4756 vs 6030 (7†) 0.73 0.74 0.03%  Positive assoc; 1196/ 1691 (2) ( 142 )  — 1000 Genomes 
PTGS2/COX2 A1803G rs4648298 4229 vs 4279 (5†) 0.97 0.97 0.00%**  No assoc; 480/657 (2) ( 142 )  — 1000 Genomes 
PTGS2/COX2 C427T rs5275 4745 vs 5756 (7†) 0.66 0.66 0.01% n/a — 1000 Genomes 
PTGS2/COX2 G306C rs5277 4269 vs 4735 (5†) 0.85 0.84 0.24% n/a — 1000 Genomes 
PTGS2/COX2 G765C rs20417 5459 vs 7272 (11†) 0.86 0.88 0.26%  Positive assoc; 3322/5166 (10) ( 123 )   ( 142 )  1000 Genomes 
PTGS2/COX2 T1532C rs5273 2843 vs 3216 (5†) val/val 1.00 1.00 —  No assoc; 670/1113 (4) ( 142 )  — 1000 Genomes 
TNF-α 308G>A rs1800629 3843 vs 4098 (9†) 0.80 0.81 0.02%  No assoc; 1372/1458 (7) ( 132 )  — Current study 
NOD2 3020incC rs5743293 4222 vs 2988 (8) 0.96 0.97 0.95%  Positive assoc; 2571/1856 (7) ( 72 )  — Current study 
NOD2 G908R rs2066845 4541 vs 3820 (6†) 0.99 0.99 —  Positive assoc; 1442/1109 (5) ( 72 )  — Current study 
NOD2 R702W rs2066844 3445 vs 2731 (6†) 0.96 0.97 0.00%  Positive assoc; 1436/1109 (5) ( 72 )  — 1000 Genomes 
Inhibition of cell growth 
CCND1 870A rs17852153 4747 vs 6783 (13) 0.46 0.46 0.18%  Positive assoc; 2289/3232 (13) ( 73 )  — Current study 
TGFB1 C509T †† rs1800469 994 vs 2335 (5) 0.42 0.48 1.99%  No assoc; 994/2335 (5) ( 76 )  — 1000 Genomes 
TGFBR1 TGFBR1*6A rs11466445 3217 vs 4539 (8) 0.90 0.92 0.52%  Positive assoc; 5666/8450 (7) ( 79 )   ( 77 , 78 )  Current study 
Insulin related 
IGF1 CA-repeat n/a 7900 vs 9161 (6) 19/19 0.62 0.61 0.02%  No assoc; 3672/4125 (4) ( 117 )   ( 116 )  — 
IGFBP3 202A>C rs2854744 7296 vs 10 452 (6) 0.48 0.50 0.00%  No assoc; 2834/3520 (3) ( 117 )  — 1000 Genomes 
Iron metabolism 
HFE C282T rs1800562 5177 vs 6150 (6†#) 0.92 0.93 0.25% n/a — 1000 Genomes 
Lipid metabolism 
ApoE e2 rs7412 5821 vs 6754 (5†) e3 0.81 0.81 0.80% n/a — 1000 Genomes 
ApoE e4 rs429358 3808 vs 4684 (5†) e3 0.79 0.79 0.09% n/a — 1000 Genomes 
Mitotic control 
STK15 F31 ‡ rs2273535 4860 vs 4629 (4) 0.75 0.76 0.31%  Positive assoc; 2302/1769 (3) ( 143 )  — 1000 Genomes 
One-carbon metabolism 
MTHFR C677T rs1801133 27 372 vs 39 867 (52§) 0.67 0.67 0.24%  Inverse assoc; 9345/18 887 (37) ( 56 )   ( 51 , 52 , 53 , 54 , 55 , 88 )  1000 Genomes 
MTHFR A1298C rs1801131 17 178 vs 24 792 (34#) 0.70 0.70 0.06%  Inverse assoc; 4764/6592 (9) ( 53 )  — 1000 Genomes 
MTR A2756G rs1805087 11 829 vs 15 975 (14#) 0.81 0.80 0.01%  No assoc; 7804/8184 (9) ( 110 )   ( 88 )  1000 Genomes 
MTRR A66G rs1801394 6170 vs 8732 (9) 0.59 0.61 0.02% n/a — 1000 Genomes 
TS TSER rs34743033 3519 vs 5289 (5) 3R/3R 0.57 0.57 0.24% n/a — Current study 
TS Ts1494del6 rs34489327 3262 vs 4518 (4) ins/ins 0.67 0.67 0.03% n/a — Current study 
Rare, high penetrance 
APC E1317Q rs1801166 6898 vs 6668 (6) 0.99 0.99 —  No assoc; 3794/4484 (8) ( 122 )  — Current study 
APC D1822V rs459552 6282 vs 7038 (6) Asp 0.78 0.77 0.38% n/a — 1000 Genomes 
MLH1 I219V rs1799977 2956 vs 5071 (7†) 0.71 0.71 0.02% n/a — 1000 Genomes 
MLH1 -93 G>A rs1800734 4524 vs 5544 (6†) 0.77 0.78 0.12% n/a — 1000 Genomes 
Substrate metabolism 
CYP1A1 2454A>G rs1048943 10 274 vs 11 978 (13†,#) 0.91 0.92 0.18%  Positive assoc; 5336/6226 (13) ( 87 )   ( 88 )  1000 Genomes 
CYP1A1 3698T>C rs4646903 4897 vs 6559 (7) 0.84 0.83 0.20%  No assoc; 234/250 (2) ( 88 )  — 1000 Genomes 
CYP1A2 163C>A rs762551 3051 vs 5326 (9) 0.68 0.68 0.00% n/a — 1000 Genomes 
CYP1B1 4326C>G rs1056836 8514 vs 9721 (6†) 0.53 0.53 0.00% n/a — 1000 Genomes 
CYP2C9 430C>T rs1799853 5134 vs 6164 (6†) 0.86 0.86 0.95% n/a — HapMap 
CYP2C9 1057A>C rs1057910 5379 vs 6531 (6†) 0.93 0.94 1.85% n/a — 1000 Genomes 
CYP2E1 1053C>T rs2031920 4456 vs 5077 (8#) 0.90 0.88 0.64% n/a — 1000 Genomes 
CYP2E1 1293G>C rs3813867 3424 vs 4686 (7) 0.94 0.93 1.06%  No assoc; 4979/6012 (10) ( 133 )  — 1000 Genomes 
GSTA1 GSTA1*B allele ‡‡ — 1648 vs 2039 (4) — — —  No assoc; 1648/2039 (4) ( 92 )  — — 
GSTM1 Null variant n/a 18 845 vs 26 662 (43) present 0.48 0.49 —  Positive assoc; 11 998/17 552 (44) ( 92 )   ( 93 , 91 , 89 , 90 , 88 )  — 
GSTP1 IIe105Val rs1695 9267 vs 12 902 (22†) IA 0.71 0.72 0.09%  No assoc; 5421/7671 (19) ( 92 )   ( 111 , 88 )  1000 Genomes 
GSTP1 Ala114Val rs1138272 5183 vs 5457 (6†,#) 0.92 0.92 0.22% n/a — 1000 Genomes 
GSTT1 Null variant n/a 13 410 vs 20 455 (35) present 0.65 0.68 —  Positive assoc; 8596/13 589 (34) ( 92 )   ( 88 , 95 , 96 )  — 
NAT1 slow/rapid n/a 4791 vs 6628 (15) slow 0.69 0.68 0.12%  No assoc; 520/433 (3) ( 88 )  — — 
NAT2 slow/rapid n/a 12 908 vs 16 483 (26) slow 0.67 0.66 0.04%  No assoc; 6741/8015 (18) ( 88 )   ( 108 , 51 , 109 )  — 
NQO1 Pro187Ser (C609T) rs1800566 5084 vs 5932 (8) 0.81 0.79 0.00%  Positive assoc; 1783/2494 (6) ( 130 )  — 1000 Genomes 
Tumor suppressor genes 
TP53 Arg72Pro §§ rs1042522 7414 vs 9872 (27) — — 0.01%  No assoc; 7414/9872 (27) ( 119 )   ( 118 , 120 , 121 )  1000 Genomes 
TP53 intron 3 16bp |||| rs17878362 1637 vs 1874 (5) Del — — 0.00%  No assoc; 1637/1874 (5) ( 144 )  — — 
MDM2 309 T/G ¶¶ rs2279744 2543 vs 2115 (7) 0.46 0.43 0.42%  No assoc; 2543/2115 (7) ( 145 )   ( 146 )  1000 Genomes 
Vitamin D and calcium metabolism 
VDR BsmI (60890GA) rs1544410 5607 vs 6202 (7) 0.63 0.60 0.12%##  Inverse assoc; 3285/1497 (4) ( 147 )  — 1000 Genomes 
VDR FokI rs10735810 7646 vs 8968 (9#) 0.61 0.60 0.00%  No assoc; 1331/2943 (5) ( 147 )  — 1000 Genomes 
VDR TaqI rs731236 946 vs 1184 (4) 0.68 0.70 0.01% n/a — 1000 Genomes 
Common low penetrance 
SMAD7 rs4939827 rs4939827 37 650 vs 36 154 (13#) 0.55 0.51 0.64% n/a — 1000 Genomes 
SMAD7 rs12953717 rs12953717 33 771 vs 32 364 (11#) 0.63 0.65 0.30% n/a — 1000 Genomes 
SMAD7 rs4464148 rs4464148 15 999 vs 15 216 (7†) 0.62 0.66 0.47% n/a — 1000 Genomes 
8q24 rs6983267 rs6983267 40 604 vs 42 672 (19) 0.48 0.51 1.06%  Positive assoc; (17) ( 148 )  — 1000 Genomes 
8q24 rs10505477 rs10505477 18 580 vs 20 147 (14) 0.46 0.49 0.59% n/a — 1000 Genomes 
9p24 rs719725 rs719725 13 290 vs 14 774 (13) 0.37 0.39 0.16%  Positive assoc; 14 064/15 933 (17) ( 102 )  — 1000 Genomes 
19q13.1 rs10411210 rs10411210 25 607 vs 26 477 (17) 0.89 0.88 0.09% n/a — 1000 Genomes 
16q22.1 rs9929218 rs9929218 26 191 vs 27 409 (18) 0.74 0.72 0.23% n/a — 1000 Genomes 
15q14 rs4779584 rs4779584 13 656 vs 12 635 (9) 0.64 0.65 0.84% n/a — 1000 Genomes 
1q41 rs6691170 rs6691170 17 740 vs 19 776 (11) 0.62 0.64 0.18% n/a — 1000 Genomes 
3q26.2 rs10936599 rs10936599 17 802 vs 19 795 (11) 0.77 0.75 0.12% n/a — 1000 Genomes 
12q13.13 rs11169552 rs11169552 17 148 vs 19 739 (11) 0.74 0.72 0.76% n/a — 1000 Genomes 
20q13.33 rs4925386 rs4925386 17 847 vs 19 832 (11) 0.71 0.68 0.39% n/a — 1000 Genomes 
14q22.2 rs4444235 rs4444235 18 607 vs 19 576 (13) 0.53 0.55 0.21% n/a — 1000 Genomes 
20p12.3 rs961253 rs961253 18 118 vs 19 006 (13) 0.66 0.68 0.22% n/a — 1000 Genomes 
8q23.3 rs16892766 rs16892766 17 180 vs 17 840 (4†) 0.88 0.90 0.10% n/a — 1000 Genomes 
10p14 rs10795668 rs10795668 20 026 vs 20 682 (6†) 0.72 0.69 0.52% n/a — 1000 Genomes 
11q23.1 rs3802842 rs3802842 33 004 vs 31 654 (14) 0.67 0.70 0.37% n/a — 1000 Genomes 
Genes, by category Variant rs number Case vs control subjects (number of samples) Ref allele Ref allele frequency of case subjects Ref allele frequency of control subjects Attributable familial risk Result of most recent meta-analysis; case subjects/ control subjects (samples) (reference) Other meta- analyses Source of minor allele frequency 
Adhesion molecules 
CDH1 C-160A rs16260 7493 vs 7329 (5†) 0.74 0.72 0.19%  No assoc ( 63 )  — 1000 Genomes 
MMP1 G-1607GG rs1799750 1007 vs 1032 (5) 0.44 0.47 0.62%  Positive assoc; 1343/1590 (7) ( 65 )   ( 64 )  Current study 
MMP3 AAAAA-1612AAAAAA rs3025058 857 vs 932 (4) 5A 0.38 0.41 0.03%  No assoc; 1296/1543 (7) ( 65 )   ( 64 )  Current study 
MMP9 1562C/T ‡ rs3918242 575 vs 836 (4) — — —  No assoc; 575/836 (4) ( 64 )   ( 64 )  1000 Genomes 
Alcohol metabolism 
ADH1B Arg47His rs1229984 1931 vs 2898 (5) His 0.75 0.76 0.32% n/a — 1000 Genomes 
ADH1C Ile349Val (1045A>G) rs698 3168 vs 6229 (7) Ile (A) 0.69 0.64 0.14% n/a — 1000 Genomes 
ALDH2 Glu487Lys rs671 2209 vs 3383 (8) Glu (G) 0.76 0.74 0.02%  Inverse assoc; 1960/3163 (6) ( 66 )  — 1000 Genomes 
Angiogenesis 
VEGF 936 C>T rs3025039 1317 vs 1192 (4) 0.84 0.83 0.60% n/a — 1000 Genomes 
VEGF G634C § rs2010963 1508 vs 1308 (4) — — 0.00% ||  No assoc; 1508/1308 (4) ( 131 )  — 1000 Genomes 
Base-excision repair 
MGMT Leu84Phe ¶ rs12917 1524 vs 4646 (5) 0.89 0.88 0.14%  No assoc; 1524/4646 (5) ( 137 )  — 1000 Genomes 
MGMT Il3143Val ¶ rs2308321 1326 vs 3520 (4) 0.89 0.87 0.19%  No assoc; 1326/3520 (4) ( 137 )  — 1000 Genomes 
MUTYH G396D rs36053993 26 592 vs 19 207 (15) 0.99 0.99 0.87% Positive assoc; 25 616/18 521 (15) (48)  ( 138 , 139 , ( 140 , 141 )  Current study 
MUTYH Y179C rs34612342 26 370 vs 19 042 (15) 1.00 1.00 0.00% Positive assoc; 25 392/18 362 (15) (48)  ( 138 , 139 , ( 140 , 141 )  Current study 
OGG1 Ser326Cys rs1052133 4713 vs 6165 (9) Ser (C) 0.71 0.65 0.02% n/a — 1000 Genomes 
XRCC1 Arg194Trp rs1799782 6635 vs 8488 (11†) 0.91 0.88 0.16% No assoc; 1709/3233 (9) (70)  ( 69 )  1000 Genomes 
XRCC1 Arg280His rs25489 3114 vs 3679 (5) 0.94 0.95 0.04% No assoc; 931/1547 (4) (70) — 1000 Genomes 
XRCC1 Arg399Gln rs25487 7247 vs 8786 (12†) 0.67 0.67 0.19% No assoc; 2776/4402 (14) (70)  ( 69 )  1000 Genomes 
XRCC3 Thr241Met rs861539 4484 vs 5235 (10†) 0.72 0.71 0.00% No assoc; 3183/3926 (7) (69) — 1000 Genomes 
Inflammation or immune response 
IL6 174G>C rs1800795 6676 vs 7942 (10†,#) 0.61 0.60 0.05% n/a — 1000 Genomes 
IL8 251T/A rs4073 3228 vs 3772 (7†) 0.54 0.54 0.02% n/a — 1000 Genomes 
IL10 1082G/A rs1800896 2964 vs 3621 (5†) 0.52 0.51 0.04% n/a — 1000 Genomes 
PPAR-ɣ C1431T rs3856806 5574 vs 7035 (7†) 0.87 0.87 0.08% No assoc; 486/941 (3) (106)  ( 107 )  1000 Genomes 
PPAR-ɣ Pro12Ala rs1801282 15 091 vs 18 690 (17†,#) 0.88 0.88 0.06%  Inverse assoc; 6878/9391 (10) ( 106 )  — 1000 Genomes 
PTGS2/COX2 A1195G rs689466 4756 vs 6030 (7†) 0.73 0.74 0.03%  Positive assoc; 1196/ 1691 (2) ( 142 )  — 1000 Genomes 
PTGS2/COX2 A1803G rs4648298 4229 vs 4279 (5†) 0.97 0.97 0.00%**  No assoc; 480/657 (2) ( 142 )  — 1000 Genomes 
PTGS2/COX2 C427T rs5275 4745 vs 5756 (7†) 0.66 0.66 0.01% n/a — 1000 Genomes 
PTGS2/COX2 G306C rs5277 4269 vs 4735 (5†) 0.85 0.84 0.24% n/a — 1000 Genomes 
PTGS2/COX2 G765C rs20417 5459 vs 7272 (11†) 0.86 0.88 0.26%  Positive assoc; 3322/5166 (10) ( 123 )   ( 142 )  1000 Genomes 
PTGS2/COX2 T1532C rs5273 2843 vs 3216 (5†) val/val 1.00 1.00 —  No assoc; 670/1113 (4) ( 142 )  — 1000 Genomes 
TNF-α 308G>A rs1800629 3843 vs 4098 (9†) 0.80 0.81 0.02%  No assoc; 1372/1458 (7) ( 132 )  — Current study 
NOD2 3020incC rs5743293 4222 vs 2988 (8) 0.96 0.97 0.95%  Positive assoc; 2571/1856 (7) ( 72 )  — Current study 
NOD2 G908R rs2066845 4541 vs 3820 (6†) 0.99 0.99 —  Positive assoc; 1442/1109 (5) ( 72 )  — Current study 
NOD2 R702W rs2066844 3445 vs 2731 (6†) 0.96 0.97 0.00%  Positive assoc; 1436/1109 (5) ( 72 )  — 1000 Genomes 
Inhibition of cell growth 
CCND1 870A rs17852153 4747 vs 6783 (13) 0.46 0.46 0.18%  Positive assoc; 2289/3232 (13) ( 73 )  — Current study 
TGFB1 C509T †† rs1800469 994 vs 2335 (5) 0.42 0.48 1.99%  No assoc; 994/2335 (5) ( 76 )  — 1000 Genomes 
TGFBR1 TGFBR1*6A rs11466445 3217 vs 4539 (8) 0.90 0.92 0.52%  Positive assoc; 5666/8450 (7) ( 79 )   ( 77 , 78 )  Current study 
Insulin related 
IGF1 CA-repeat n/a 7900 vs 9161 (6) 19/19 0.62 0.61 0.02%  No assoc; 3672/4125 (4) ( 117 )   ( 116 )  — 
IGFBP3 202A>C rs2854744 7296 vs 10 452 (6) 0.48 0.50 0.00%  No assoc; 2834/3520 (3) ( 117 )  — 1000 Genomes 
Iron metabolism 
HFE C282T rs1800562 5177 vs 6150 (6†#) 0.92 0.93 0.25% n/a — 1000 Genomes 
Lipid metabolism 
ApoE e2 rs7412 5821 vs 6754 (5†) e3 0.81 0.81 0.80% n/a — 1000 Genomes 
ApoE e4 rs429358 3808 vs 4684 (5†) e3 0.79 0.79 0.09% n/a — 1000 Genomes 
Mitotic control 
STK15 F31 ‡ rs2273535 4860 vs 4629 (4) 0.75 0.76 0.31%  Positive assoc; 2302/1769 (3) ( 143 )  — 1000 Genomes 
One-carbon metabolism 
MTHFR C677T rs1801133 27 372 vs 39 867 (52§) 0.67 0.67 0.24%  Inverse assoc; 9345/18 887 (37) ( 56 )   ( 51 , 52 , 53 , 54 , 55 , 88 )  1000 Genomes 
MTHFR A1298C rs1801131 17 178 vs 24 792 (34#) 0.70 0.70 0.06%  Inverse assoc; 4764/6592 (9) ( 53 )  — 1000 Genomes 
MTR A2756G rs1805087 11 829 vs 15 975 (14#) 0.81 0.80 0.01%  No assoc; 7804/8184 (9) ( 110 )   ( 88 )  1000 Genomes 
MTRR A66G rs1801394 6170 vs 8732 (9) 0.59 0.61 0.02% n/a — 1000 Genomes 
TS TSER rs34743033 3519 vs 5289 (5) 3R/3R 0.57 0.57 0.24% n/a — Current study 
TS Ts1494del6 rs34489327 3262 vs 4518 (4) ins/ins 0.67 0.67 0.03% n/a — Current study 
Rare, high penetrance 
APC E1317Q rs1801166 6898 vs 6668 (6) 0.99 0.99 —  No assoc; 3794/4484 (8) ( 122 )  — Current study 
APC D1822V rs459552 6282 vs 7038 (6) Asp 0.78 0.77 0.38% n/a — 1000 Genomes 
MLH1 I219V rs1799977 2956 vs 5071 (7†) 0.71 0.71 0.02% n/a — 1000 Genomes 
MLH1 -93 G>A rs1800734 4524 vs 5544 (6†) 0.77 0.78 0.12% n/a — 1000 Genomes 
Substrate metabolism 
CYP1A1 2454A>G rs1048943 10 274 vs 11 978 (13†,#) 0.91 0.92 0.18%  Positive assoc; 5336/6226 (13) ( 87 )   ( 88 )  1000 Genomes 
CYP1A1 3698T>C rs4646903 4897 vs 6559 (7) 0.84 0.83 0.20%  No assoc; 234/250 (2) ( 88 )  — 1000 Genomes 
CYP1A2 163C>A rs762551 3051 vs 5326 (9) 0.68 0.68 0.00% n/a — 1000 Genomes 
CYP1B1 4326C>G rs1056836 8514 vs 9721 (6†) 0.53 0.53 0.00% n/a — 1000 Genomes 
CYP2C9 430C>T rs1799853 5134 vs 6164 (6†) 0.86 0.86 0.95% n/a — HapMap 
CYP2C9 1057A>C rs1057910 5379 vs 6531 (6†) 0.93 0.94 1.85% n/a — 1000 Genomes 
CYP2E1 1053C>T rs2031920 4456 vs 5077 (8#) 0.90 0.88 0.64% n/a — 1000 Genomes 
CYP2E1 1293G>C rs3813867 3424 vs 4686 (7) 0.94 0.93 1.06%  No assoc; 4979/6012 (10) ( 133 )  — 1000 Genomes 
GSTA1 GSTA1*B allele ‡‡ — 1648 vs 2039 (4) — — —  No assoc; 1648/2039 (4) ( 92 )  — — 
GSTM1 Null variant n/a 18 845 vs 26 662 (43) present 0.48 0.49 —  Positive assoc; 11 998/17 552 (44) ( 92 )   ( 93 , 91 , 89 , 90 , 88 )  — 
GSTP1 IIe105Val rs1695 9267 vs 12 902 (22†) IA 0.71 0.72 0.09%  No assoc; 5421/7671 (19) ( 92 )   ( 111 , 88 )  1000 Genomes 
GSTP1 Ala114Val rs1138272 5183 vs 5457 (6†,#) 0.92 0.92 0.22% n/a — 1000 Genomes 
GSTT1 Null variant n/a 13 410 vs 20 455 (35) present 0.65 0.68 —  Positive assoc; 8596/13 589 (34) ( 92 )   ( 88 , 95 , 96 )  — 
NAT1 slow/rapid n/a 4791 vs 6628 (15) slow 0.69 0.68 0.12%  No assoc; 520/433 (3) ( 88 )  — — 
NAT2 slow/rapid n/a 12 908 vs 16 483 (26) slow 0.67 0.66 0.04%  No assoc; 6741/8015 (18) ( 88 )   ( 108 , 51 , 109 )  — 
NQO1 Pro187Ser (C609T) rs1800566 5084 vs 5932 (8) 0.81 0.79 0.00%  Positive assoc; 1783/2494 (6) ( 130 )  — 1000 Genomes 
Tumor suppressor genes 
TP53 Arg72Pro §§ rs1042522 7414 vs 9872 (27) — — 0.01%  No assoc; 7414/9872 (27) ( 119 )   ( 118 , 120 , 121 )  1000 Genomes 
TP53 intron 3 16bp |||| rs17878362 1637 vs 1874 (5) Del — — 0.00%  No assoc; 1637/1874 (5) ( 144 )  — — 
MDM2 309 T/G ¶¶ rs2279744 2543 vs 2115 (7) 0.46 0.43 0.42%  No assoc; 2543/2115 (7) ( 145 )   ( 146 )  1000 Genomes 
Vitamin D and calcium metabolism 
VDR BsmI (60890GA) rs1544410 5607 vs 6202 (7) 0.63 0.60 0.12%##  Inverse assoc; 3285/1497 (4) ( 147 )  — 1000 Genomes 
VDR FokI rs10735810 7646 vs 8968 (9#) 0.61 0.60 0.00%  No assoc; 1331/2943 (5) ( 147 )  — 1000 Genomes 
VDR TaqI rs731236 946 vs 1184 (4) 0.68 0.70 0.01% n/a — 1000 Genomes 
Common low penetrance 
SMAD7 rs4939827 rs4939827 37 650 vs 36 154 (13#) 0.55 0.51 0.64% n/a — 1000 Genomes 
SMAD7 rs12953717 rs12953717 33 771 vs 32 364 (11#) 0.63 0.65 0.30% n/a — 1000 Genomes 
SMAD7 rs4464148 rs4464148 15 999 vs 15 216 (7†) 0.62 0.66 0.47% n/a — 1000 Genomes 
8q24 rs6983267 rs6983267 40 604 vs 42 672 (19) 0.48 0.51 1.06%  Positive assoc; (17) ( 148 )  — 1000 Genomes 
8q24 rs10505477 rs10505477 18 580 vs 20 147 (14) 0.46 0.49 0.59% n/a — 1000 Genomes 
9p24 rs719725 rs719725 13 290 vs 14 774 (13) 0.37 0.39 0.16%  Positive assoc; 14 064/15 933 (17) ( 102 )  — 1000 Genomes 
19q13.1 rs10411210 rs10411210 25 607 vs 26 477 (17) 0.89 0.88 0.09% n/a — 1000 Genomes 
16q22.1 rs9929218 rs9929218 26 191 vs 27 409 (18) 0.74 0.72 0.23% n/a — 1000 Genomes 
15q14 rs4779584 rs4779584 13 656 vs 12 635 (9) 0.64 0.65 0.84% n/a — 1000 Genomes 
1q41 rs6691170 rs6691170 17 740 vs 19 776 (11) 0.62 0.64 0.18% n/a — 1000 Genomes 
3q26.2 rs10936599 rs10936599 17 802 vs 19 795 (11) 0.77 0.75 0.12% n/a — 1000 Genomes 
12q13.13 rs11169552 rs11169552 17 148 vs 19 739 (11) 0.74 0.72 0.76% n/a — 1000 Genomes 
20q13.33 rs4925386 rs4925386 17 847 vs 19 832 (11) 0.71 0.68 0.39% n/a — 1000 Genomes 
14q22.2 rs4444235 rs4444235 18 607 vs 19 576 (13) 0.53 0.55 0.21% n/a — 1000 Genomes 
20p12.3 rs961253 rs961253 18 118 vs 19 006 (13) 0.66 0.68 0.22% n/a — 1000 Genomes 
8q23.3 rs16892766 rs16892766 17 180 vs 17 840 (4†) 0.88 0.90 0.10% n/a — 1000 Genomes 
10p14 rs10795668 rs10795668 20 026 vs 20 682 (6†) 0.72 0.69 0.52% n/a — 1000 Genomes 
11q23.1 rs3802842 rs3802842 33 004 vs 31 654 (14) 0.67 0.70 0.37% n/a — 1000 Genomes 

* n/a = not applicable. See Supplementary Table 1 (available online) for gene names.

Includes unpublished data from SOCCS.

McColgan and Sharma 2009 ( 64 ).

§ Liu et al. 2011 ( 131 ).

|| ref allele frequency taken from 1000 Genomes data.

Zhong et al. 2010 ( 137 ).

# Includes unpublished data from Ontario.

** OR for homozygote estimated as square of OR for heterozygotes.

†† Fang et al. 2010 ( 76 ).

‡‡ Economopoulos and Sergentanis 2010 ( 92 ).

§§ Economopoulos et al. 2010 ( 119 ).

|||| Hu et al. 2010 ( 144 ).

¶¶ Tomlinson 2008 was based on 10 samples.

## Based on the white-only analysis.

Statistical Analysis

Meta-analyses were conducted for 92 polymorphisms in 64 different loci ( Table 1 ) with genotypes available in four or more candidate or GWA studies. On average, these meta-analyses were based on data from 5281 case patients (median interquartile range = 13 472 – 3384) and 6594 control subjects (median interquartile range = 16 102 – 4534) originating from eight (median interquartile range = 13 – 6) case–control samples. Unpublished data from a Scottish and/or a Canadian GWAS were included in the analyses of 37 SNPs. Summary crude odds ratios (ORs) and 95% confidence intervals for two additive models (var/wt vs wt/wt and var/var vs wt/wt) for variants that were identified from candidate studies are presented in Table 2 and for variants that were identified from GWAS, in Table 3 . Summary crude odds ratios and 95% confidence intervals for a recessive (var/var vs var/wt and wt/wt) and a dominant model (var/var and var/wt vs wt/wt) are presented in Supplementary Table 2 (available online), for variants that were identified from candidate studies, and Supplementary Table 3 (available online), for variants that were identified from GWAS. For 37 associations in which the 95% confidence intervals excluded unity, we repeated the analysis excluding studies in nonwhite populations ( Supplementary Tables 4 and 5 , available online). We checked the linkage disequilibrium (LD) for all the polymorphisms that were included in our study and we found that the following variants were in LD ( r2 threshold ≥ 0.60): 1) rs4939827 (in SMAD7 ) and rs12953717 (in SMAD7, r2 = 0.60); 2) rs3813867 (in CYP2E1 ) and rs2031920 (in CYP2E1, r2 = 0.87); 3) rs10505477 (in 8q24 ) and rs6983267 (in 8q24, r2 = 0.88); 4) rs731236 (in VDR ) and rs1544410 (in VDR, r2 = 1.00).

Table 2.

Summary crude odds ratios (ORs) and 95% confidence intervals (95% CIs) for two additive models for variants that were identified from candidate studies

Gene Variant rs number Case vs control subjects (no. of samples) Additive Model: var/wt vs wt/wt Additive Model: var/var vs wt/wt Credibility  
OR (95% CI) P I2 (95% CI)  Power OR (95% CI) P I2 (95% CI)  Power BFDP* Venice criteria grade† 
Adhesion molecules 
CDH1 C-160A‡ rs16260 7493 vs 7329 (5§) 0.91 (0.85 to 0.97) .005 49 (0 to 81) 0.78 0.84 (0.74 to 0.96) .01 34 (0 to 75) 0.76 0.67 BBB 
MMP1 G-1607GG‡ rs1799750 1007 vs 1032 (5) 1.05 (0.84 to 1.33) .66 15 (0 to 82) 0.07 1.54 (1.00 to 2.37) .05 53 (0 to 83) 0.61 0.88 BBB 
MMP3 AAAAA-612AAAAAA‡ rs3025058 857 vs 932 (4) 0.79 (0.60 to 1.03) .08 0 (0 to 85) 0.43 1.16 (0.86 to 1.56) .33 0 (0 to 85) 0.20 0.92 CAB 
MMP9 1562C/T|| rs3918242 575 vs 836 (4) — — — — — — — — — — — — 
Alcohol metabolism 
ADH1B Arg47His rs1229984 1931 vs 2898 (5)  1.06 (0.83 to 1.36)  .63  70 (24 to 88)  0.15  1.22 (0.94 to 1.58)  .13  26 (0 to 70)  0.35 0.97 CCB 
ADH1C Ile349Val(1045A>G) rs698 3168 vs 6229 (7)  0.96 (0.87 to 1.07)  .49  0 (0 to 78)  0.12  0.88 (0.66 to 1.16)  .35  65 (21 to 84)  0.4 0.98 CBB 
ALDH2 Glu487Lys rs671‡ 2209 vs 3383 (8) 0.88 (0.79 to 0.99) .04 0 (0 to 65) 0.59 0.89 (0.58 to 1.36) .59 58 (12 to 80) 0.17 0.89 CCB 
Angiogenesis 
VEGF 936 C>T rs3025039 1317 vs 1192 (4)  0.94 (0.71 to 1.24)  .69  55 (0 to 85)  0.13  1.19 (0.72 to 1.99)  .50  15 (0 to 87)  0.10 0.97 CBB 
VEGF G634C¶ rs2010963 1508 vs 1308 (4)  0.89 (0.72 to 1.10)  — — —  1.17 (0.93 to 1.47)  — — — 0.96 --C 
Base-excision repair 
MGMT Leu84Phe** rs12917 1524 vs 4646 (5)  0.84 (0.70 to 1.00)  .05 — —  0.97 (0.56 to 1.66)  .97 — — 0.91 --C 
MGMT Il3143Val** rs2308321 1326 vs 3520 (4)  0.86 (0.66 to 1.12)  .26 — —  1.01 (0.56 to 1.81)  .69 — — 0.91 --C 
MUTYH G396D rs36053993†† 26 592 vs 19 207 (15) 15 1.07 (0.91 to 1.27) .41 0 (0 to 54) 0.13 6.15 (2.34 to 16.15) .00 0 (0 to 65) 0.50 0.17 CAB 
MUTYH Y179C rs34612342‡ 26 370 vs 19 042 (15) 14 1.34 (1.02 to 1.77) .04 0 (0 to 55) 0.55 3.35 (1.14 to 9.89) .03 0 (0 to 75) 0.37 0.89 BAB 
OGG1 Ser326Cys rs1052133 4713 vs 6165 (9)  1.02 (0.94 to 1.12)  .60  48 (0 to 76)  0.08  1.05 (0.89 to 1.23)  .58  42 (0 to 73)  0.14 0.99 CBB 
XRCC1 Arg194Trp rs1799782 6635 vs 8488 (11§) 10  0.96 (0.87 to 1.07)  .50  0 (0 to 62)  0.13 10  1.10 (0.82 to 1.48)  .52  16 (0 to 57)  0.13 0.98 CAB 
XRCC1 Arg280His rs25489 3114 vs 3679 (5)  1.06 (0.83 to 1.34)  .65  41 (0 to 80)  0.08  1.16 (0.37 to 3.62)  .31  14 (0 to 91)  0.06 0.97 CBB 
XRCC1 Arg399Gln rs25487‡ 7247 vs 8786 (12§) 12 0.99 (0.92 to 1.06) .72 11 (0 to 51) 0.06 12 0.88 (0.79 to 0.97) .02 0 (0 to 58) 0.67 0.99 BAB 
XRCC3 Thr241Met rs861539 4484 vs 5235 (10§) 10  0.92 (0.84 to 1.01)  .09  36 (0 to 69)  0.48 10  0.95 (0.72 to 1.24)  .68  57 (12 to 79)  0.12 0.95 CBB 
Inflammation or immune response 
IL6 174G>C rs1800795‡ 6676 vs 7942 (10§,||) 10 1.03 (0.91 to 1.17) .65 56 (11 to 78) 0.13 10 0.94 (0.79 to 1.12) .48 56 (11 to 78) 0.24 0.98 CCB 
IL8 251T/A rs4073 3228 vs 3772 (7§) 1.03 (0.92 to 1.15) .61 44 (0 to 77) 0.08 1.05 (0.91 to 1.20) .53 0 (0 to 71) 0.11 0.99 CBB 
IL10 1082G/A rs1800896 2964 vs 3621 (5§) 0.96 (0.85 to 1.08) .49 0 (0 to 79) 0.11 0.93 (0.81 to 1.07) .30 35 (0 to 75) 0.18 0.98 CAB 
PPARγ C1431T rs3856806 5574 vs 7035 (7§) 1.04 (0.95 to 1.13) .44 0 (0 to 79) 0.14 0.95 (0.50 to 1.78) .86 69 (22 to 88) 0.06 0.98 CAB 
PPARγ Pro12Ala rs1801282 15 091 vs 18 690 (17§,||) 13 0.98 (0.86 to 1.11) .72 65 (36 to 80) 0.09 12 0.91 (0.73 to 1.12) .37 0 (0 to 58) 0.14 0.98 CCB 
PTGS2 A1195G rs689466 4756 vs 6030 (7§) 1.04 (0.95 to 1.13) .42 41 (0 to 75) 0.16 1.08 (0.77 to 1.51) .66 64 (19 to 84) 0.19 0.98 CBB 
PTGS2 A1803G rs4648298 4229 vs 4279 (5§) 1.00 (0.83 to 1.22) .96 49 (0 to 85) 0.05 n/a n/a n/a n/a 0.98 CBB 
PTGS2 C427T rs5275 4745 vs 5756 (7§) 1.01 (0.93 to 1.09) .87 0 (0 to 71) 0.06 1.03 (0.91 to 1.17) .65 0 (0 to 71) 0.08 0.99 CAB 
PTGS2 G306C rs5277 4269 vs 4735 (5§) 0.97 (0.88 to 1.06) .45 23 (0 to 68) 0.10 0.85 (0.65 to 1.11) .24 5 (0 to 80) 0.23 0.99 CAB 
PTGS2 G765C rs20417 5459 vs 7272 (11§) 10 1.03 (0.95 to 1.13) .45 45 (0 to 74) 0.11 1.21 (0.93 to 1.57) .15 0 (0 to 68) 0.3 0.99 CBB 
PTGS2 T1532C rs5273 2843 vs 3216 (5§) n/a n/a n/a n/a n/a n/a n/a n/a — — 
TNF-α 308G>A rs1800629 3843 vs 4098 (9§) 1.11 (0.88 to 1.40) .37 72 (45 to 86) 0.65 1.13 (0.92 to 1.39) .24 47 (0 to 75) 0.21 0.97 BCB 
NOD2 3020incC‡ rs5743293 4222 vs 2988 (8) 1.39 (1.15 to 1.69) .001 0 (0 to 68) 0.94 2.81 (0.87 to 9.05) .08 0 (0 to 79) 0.26 0.34 AAB 
NOD2 G908R rs2066845 4541 vs 3820 (6§) 1.41 (1.04 to 1.91) .03 0 (0 to 75) 0.63 n/a n/a n/a n/a 0.87 BBB 
NOD2 R702W‡ rs2066844 3445 vs 2731 (6§) 1.22 (1.00 to 1.50) .06 12 (0 to 78) 0.49 1.23 (0.41 to 3.70) .71 0 (0 to 85) 0.06 0.91 BAB 
Inhibition of cell growth 
CCND1 870A‡ rs17852153 4747 vs 6783 (13) 13 1.13 (1.03 to 1.25) .02 0 (0 to 57) 0.70 13 1.16 (0.98 to 1.38) .09 51 (8 to 74) 0.78 0.85 BAB 
TGFB1 C509T‡‡ rs1800469 994 vs 2335 (5)  1.12 (0.91 to 1.37)  — — —  1.62 (1.30 to 2.02)  — — — 0.96 --C 
TGFBR1 TGFBR1*6A‡ rs11466445 3217 vs 4539 (8) 1.15 (1.01 to 1.31) .03 0 (0 to 68) — 1.51 (0.69 to 3.31) .31 52 (0 to 87) — 0.89 -AB 
Insulin-related 
IGF1 CA-repeat n/a 7900 vs 9161 (6)  1.06 (0.99 to 1.14)  .08  0 (0 to 75)  0.4  1.07 (0.97 to 1.18)  .15  0 (0 to 75)  0.31 0.97 CAB 
IGFBP3 202A>C rs2854744 7296 vs 10 452 (6)  1.02 (0.94 to 1.10)  .72  0 (0 to 75)  0.08  1.00 (0.91 to 1.10)  .96  0 (0 to 75)  0.05 0.99 CAB 
Iron metabolism 
HFE C282T rs1800562 5177 vs 6150 (6§,||)  1.08 (0.97 to 1.21)  .15  0 (0 to 75)  0.29  0.90 (0.55 to 1.48)  .69  0 (0 to 79)  0.07 0.97 CAB 
Lipid metabolism 
ApoE e2 rs7412 5821 vs 6754 (5§)  0.94 (0.85 to 1.04)  .23  66 (11 to 87)  0.21  0.75 (0.46 to 1.22)  .25  0 (0 to 85)  0.21 0.98 CCB 
ApoE e4 rs429358 3808 vs 4684 (5§)  1.04 (0.94 to 1.15)  .49  14 (0 to 82)  0.12  1.12 (0.83 to 1.52)  .45  0 (0 to 79)  0.12 0.98 CAB 
Mitotic control 
STK15 F31I rs2273535 4860 vs 4629 (4)  1.03 (0.94 to 1.12)  .58  0 (0 to 85)  0.1  1.30 (0.96 to 1.76)  .09  52 (0 to 84)  0.56 0.99 CAB 
One-carbon metabolism 
MTHFR C677T rs1801133†† 27 372 vs 39 867 (52||) 52 1.00 (0.94 to 1.06) .92 53 (36 to 65) 0.05 52 0.87 (0.82 to 0.91) .00 35 (10 to 53) 1.00 0.99 CBB 
MTHFR A1298C rs1801131 17 178 vs 24 792 (34||) 34  1.01 (0.97 to 1.06)  .51  0 (0 to 37)  0.08 34  0.94 (0.87 to 1.01)  .09  22 (0 to 49)  0.40 0.99 CAB 
MTR A2756G rs1805087 11 829 vs 15 975 (14||) 14  0.97 (0.92 to 1.02)  .27  12 (0 to 50)  0.21 14  0.96 (0.84 to 1.09)  .50  48 (4 to 72)  0.05 0.99 CAB 
MTRR A66G rs1801394 6170 vs 8732 (9)  0.98 (0.90 to 1.07)  .66  15 (0 to 57)  0.10  1.04 (0.94 to 1.14)  .47  23 (0 to 64)  0.17 0.99 CAB 
TS TSER‡ rs34743033 3519 vs 5289 (5) 0.86 (0.78 to 0.95) .003 18 (0 to 83) 0.87 0.83 (0.73 to 0.94) .004 17 (0 to 83) 0.85 0.55 AAB 
TS Ts1494del6 rs34489327 3262 vs 4518 (4)  0.96 (0.88 to 1.06)  .45  0 (0 to 85)  0.13  1.03 (0.88 to 1.19)  .73  0 (0 to 85)  0.07 0.98 CAB 
Rare, high penetrance 
APC E1317Q rs1801166 6898 vs 6668 (6)  1.13 (0.88 to 1.47)  .34  0 (0 to 75)  0.15 n/a n/a n/a n/a 0.96 CAB 
APC D1822V‡ rs459552 6282 vs 7038 (6) 0.99 (0.92 to 1.07) .83 24 (0 to 68) 0.06 0.84 (0.71 to 0.98) .03 0 (0 to 75) 0.55 0.99 CAB 
MLH1 I219V rs1799977 2956 vs 5071 (7§)  1.09 (0.90 to 1.32)  .40  55 (0 to 81)  0.42  1.01 (0.85 to 1.21)  .88  43 (0 to 76)  0.05 0.97 CCB 
MLH1 -93 G>A rs1800734 4524 vs 5544 (6§)  1.06 (0.97 to 1.15)  .23  0 (0 to 79)  0.27  1.15 (0.95 to 1.39)  .15  26 (0 to 71)  0.33 0.97 CAB 
Substrate metabolism 
CYP1A1 2454A>G‡ rs1048943 10 274 vs 11 978 (13§,||) 13 1.28 (1.01 to 1.63) .05 85 (7 to 91) 1.00 12 1.47 (1.17 to 1.85) .001 3 (0 to 60) 0.93 0.89 ACB 
CYP1A1 3698T>C rs4646903 4897 vs 6559 (7)  0.94 (0.86 to 1.04)  .23  0 (0 to 71)  0.27  0.84 (0.56 to 1.27)  .42  54 (0 to 80)  0.38 0.98 CAB 
CYP1A2 163C>A rs762551 3051 vs 5326 (9)  1.13 (0.95 to 1.34)  .18  63 (25 to 82)  0.75  1.07 (0.92 to 1.26)  .40  42 (0 to 73)  0.16 0.96 ACB 
CYP1B1 4326C>G rs1056836 8514 vs 9721 (6§)  0.99 (0.92 to 1.06)  .69  0 (0 to 75)  0.06  0.98 (0.90 to 1.07)  .70  0 (0 to 77)  0.08 0.99 CAB 
CYP2C9 430C>T rs1799853 5134 vs 6164 (6§)  0.93 (0.85 to 1.02)  .13  21 (0 to 65)  0.34  1.29 (0.99 to 1.70)  .06  0 (0 to 75)  0.46 0.97 CAB 
CYP2C9 1057A>C rs1057910 5379 vs 6531 (6§)  1.08 (0.83 to 1.40)  .57  73 (33 to 88)  0.26  0.68 (0.37 to 1.26)  .22  11 (0 to 86)  0.23 0.97 CCB 
CYP2E1 1053C>T rs2031920 4456 vs 5077 (8||)  0.93 (0.83 to 1.05)  .23  0 (0 to 68)  0.27  1.23 (0.92 to 1.63)  .16  35 (0 to 71)  0.34 0.97 CAB 
CYP2E1 1293G>C rs3813867 3424 vs 4686 (7)  1.17 (0.92 to 1.48)  .21  53 (0 to 80)  0.61  1.83 (0.94 to 3.57)  .08  0 (0 to 75)  0.55 0.95 BCB 
GSTA1 GSTA1*B§§  1648 vs 2039 (4)  1.03 (0.89 to 1.19)  — — —  1.09 (0.90 to 1.32)  — — — 0.98 — 
GSTM1 Null variant n/a 18 845 vs 26 662 (43) n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a — 
GSTP1 IIe105Val rs1695 9267 vs 12 902 (22§) 22  1.05 (0.99 to 1.12)  .11  0 (0 to 46)  0.38 22  0.95 (0.86 to 1.05)  .32  36 (0 to 62)  0.17 0.98 CAB 
GSTP1 Ala114Val rs1138272 5183 vs 5457 (6§,||)  1.02 (0.91 to 1.13)  .77  0 (0 to 75)  0.07  0.87 (0.55 to 1.37)  .55  12 (0 to 78)  0.09 0.99 CAB 
GSTT1 Null variant n/a 13 410 vs 20 455 (35) n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a — 
ΝΑΤ1 slow/rapid‡ n/a 4791 vs 6628 (15) 0.80 (0.68 to 0.93) .003 38 (0 to 74) 0.86 0.98 0.79 to 1.22) .97 0 (0 to 71) 0.06 0.57 ABB 
ΝΑΤ2 slow/rapid n/a 12 908 vs 16 483 (26) 15  1.01 (0.83 to 1.22)  .94  81 (70 to 88)  0.06 15  0.95 (0.76 to 1.20)  .68  64 (37 to 79)  0.17 0.98 CCB 
NQO1 Pro187Ser(C609T) rs1800566 5084 vs 5932 (8)  1.14 (0.96 to 1.35)  .12  64 (22 to 83)  0.89  1.10 (0.76 to 1.59)  .63  56 (2 to 80)  0.19 0.95 ACB 
Tumor suppressor genes 
TP53 Arg72Pro|||| rs1042522 7414 vs 9872 (27) 27  1.01 (0.89 to 1.14)  .90 — — 27  1.04 (0.82 to 1.31)  .77 — — 0.99 --C 
TP53 intron3 16bp|||| rs17878362 1637 vs 1874 (5)  1.58 (0.98 to 2.56)  — — —  1.14 (0.84 to 1.55)  — — — 0.92 --C 
MDM2 309 T/G|||| rs2279744 2543 vs 2115 (7)  0.73 (0.62 to 0.86)  — — —  0.86 (0.57 to 1.30)  — — — 0.10 --C 
Vitamin D and calcium metabolism 
VDR BsmI(60890GA) rs1544410‡ 5607 vs 6202 (7) 0.77 (0.58 to 1.02) .07 89 (81 to 94) 1.00 0.51 (0.28 to 0.90) .02 95 (92 to 97) 1.00 0.92 ACB 
VDR FokI rs10735810 7646 vs 8968 (9||)  0.97 (0.85 to 1.11)  .66  67 (32 to 83)  0.16  0.99 (0.82 to 1.20)  .94  70 (41 to 85)  0.06 0.98 CCB 
VDR TaqI rs731236 946 vs 1184 (4)  1.06 (0.87 to 1.30)  .55  0 (0 to 85)  0.09  1.00 (0.57 to 1.75)  .99  73 (8 to 92)  0.05 0.97 CAB 
Gene Variant rs number Case vs control subjects (no. of samples) Additive Model: var/wt vs wt/wt Additive Model: var/var vs wt/wt Credibility  
OR (95% CI) P I2 (95% CI)  Power OR (95% CI) P I2 (95% CI)  Power BFDP* Venice criteria grade† 
Adhesion molecules 
CDH1 C-160A‡ rs16260 7493 vs 7329 (5§) 0.91 (0.85 to 0.97) .005 49 (0 to 81) 0.78 0.84 (0.74 to 0.96) .01 34 (0 to 75) 0.76 0.67 BBB 
MMP1 G-1607GG‡ rs1799750 1007 vs 1032 (5) 1.05 (0.84 to 1.33) .66 15 (0 to 82) 0.07 1.54 (1.00 to 2.37) .05 53 (0 to 83) 0.61 0.88 BBB 
MMP3 AAAAA-612AAAAAA‡ rs3025058 857 vs 932 (4) 0.79 (0.60 to 1.03) .08 0 (0 to 85) 0.43 1.16 (0.86 to 1.56) .33 0 (0 to 85) 0.20 0.92 CAB 
MMP9 1562C/T|| rs3918242 575 vs 836 (4) — — — — — — — — — — — — 
Alcohol metabolism 
ADH1B Arg47His rs1229984 1931 vs 2898 (5)  1.06 (0.83 to 1.36)  .63  70 (24 to 88)  0.15  1.22 (0.94 to 1.58)  .13  26 (0 to 70)  0.35 0.97 CCB 
ADH1C Ile349Val(1045A>G) rs698 3168 vs 6229 (7)  0.96 (0.87 to 1.07)  .49  0 (0 to 78)  0.12  0.88 (0.66 to 1.16)  .35  65 (21 to 84)  0.4 0.98 CBB 
ALDH2 Glu487Lys rs671‡ 2209 vs 3383 (8) 0.88 (0.79 to 0.99) .04 0 (0 to 65) 0.59 0.89 (0.58 to 1.36) .59 58 (12 to 80) 0.17 0.89 CCB 
Angiogenesis 
VEGF 936 C>T rs3025039 1317 vs 1192 (4)  0.94 (0.71 to 1.24)  .69  55 (0 to 85)  0.13  1.19 (0.72 to 1.99)  .50  15 (0 to 87)  0.10 0.97 CBB 
VEGF G634C¶ rs2010963 1508 vs 1308 (4)  0.89 (0.72 to 1.10)  — — —  1.17 (0.93 to 1.47)  — — — 0.96 --C 
Base-excision repair 
MGMT Leu84Phe** rs12917 1524 vs 4646 (5)  0.84 (0.70 to 1.00)  .05 — —  0.97 (0.56 to 1.66)  .97 — — 0.91 --C 
MGMT Il3143Val** rs2308321 1326 vs 3520 (4)  0.86 (0.66 to 1.12)  .26 — —  1.01 (0.56 to 1.81)  .69 — — 0.91 --C 
MUTYH G396D rs36053993†† 26 592 vs 19 207 (15) 15 1.07 (0.91 to 1.27) .41 0 (0 to 54) 0.13 6.15 (2.34 to 16.15) .00 0 (0 to 65) 0.50 0.17 CAB 
MUTYH Y179C rs34612342‡ 26 370 vs 19 042 (15) 14 1.34 (1.02 to 1.77) .04 0 (0 to 55) 0.55 3.35 (1.14 to 9.89) .03 0 (0 to 75) 0.37 0.89 BAB 
OGG1 Ser326Cys rs1052133 4713 vs 6165 (9)  1.02 (0.94 to 1.12)  .60  48 (0 to 76)  0.08  1.05 (0.89 to 1.23)  .58  42 (0 to 73)  0.14 0.99 CBB 
XRCC1 Arg194Trp rs1799782 6635 vs 8488 (11§) 10  0.96 (0.87 to 1.07)  .50  0 (0 to 62)  0.13 10  1.10 (0.82 to 1.48)  .52  16 (0 to 57)  0.13 0.98 CAB 
XRCC1 Arg280His rs25489 3114 vs 3679 (5)  1.06 (0.83 to 1.34)  .65  41 (0 to 80)  0.08  1.16 (0.37 to 3.62)  .31  14 (0 to 91)  0.06 0.97 CBB 
XRCC1 Arg399Gln rs25487‡ 7247 vs 8786 (12§) 12 0.99 (0.92 to 1.06) .72 11 (0 to 51) 0.06 12 0.88 (0.79 to 0.97) .02 0 (0 to 58) 0.67 0.99 BAB 
XRCC3 Thr241Met rs861539 4484 vs 5235 (10§) 10  0.92 (0.84 to 1.01)  .09  36 (0 to 69)  0.48 10  0.95 (0.72 to 1.24)  .68  57 (12 to 79)  0.12 0.95 CBB 
Inflammation or immune response 
IL6 174G>C rs1800795‡ 6676 vs 7942 (10§,||) 10 1.03 (0.91 to 1.17) .65 56 (11 to 78) 0.13 10 0.94 (0.79 to 1.12) .48 56 (11 to 78) 0.24 0.98 CCB 
IL8 251T/A rs4073 3228 vs 3772 (7§) 1.03 (0.92 to 1.15) .61 44 (0 to 77) 0.08 1.05 (0.91 to 1.20) .53 0 (0 to 71) 0.11 0.99 CBB 
IL10 1082G/A rs1800896 2964 vs 3621 (5§) 0.96 (0.85 to 1.08) .49 0 (0 to 79) 0.11 0.93 (0.81 to 1.07) .30 35 (0 to 75) 0.18 0.98 CAB 
PPARγ C1431T rs3856806 5574 vs 7035 (7§) 1.04 (0.95 to 1.13) .44 0 (0 to 79) 0.14 0.95 (0.50 to 1.78) .86 69 (22 to 88) 0.06 0.98 CAB 
PPARγ Pro12Ala rs1801282 15 091 vs 18 690 (17§,||) 13 0.98 (0.86 to 1.11) .72 65 (36 to 80) 0.09 12 0.91 (0.73 to 1.12) .37 0 (0 to 58) 0.14 0.98 CCB 
PTGS2 A1195G rs689466 4756 vs 6030 (7§) 1.04 (0.95 to 1.13) .42 41 (0 to 75) 0.16 1.08 (0.77 to 1.51) .66 64 (19 to 84) 0.19 0.98 CBB 
PTGS2 A1803G rs4648298 4229 vs 4279 (5§) 1.00 (0.83 to 1.22) .96 49 (0 to 85) 0.05 n/a n/a n/a n/a 0.98 CBB 
PTGS2 C427T rs5275 4745 vs 5756 (7§) 1.01 (0.93 to 1.09) .87 0 (0 to 71) 0.06 1.03 (0.91 to 1.17) .65 0 (0 to 71) 0.08 0.99 CAB 
PTGS2 G306C rs5277 4269 vs 4735 (5§) 0.97 (0.88 to 1.06) .45 23 (0 to 68) 0.10 0.85 (0.65 to 1.11) .24 5 (0 to 80) 0.23 0.99 CAB 
PTGS2 G765C rs20417 5459 vs 7272 (11§) 10 1.03 (0.95 to 1.13) .45 45 (0 to 74) 0.11 1.21 (0.93 to 1.57) .15 0 (0 to 68) 0.3 0.99 CBB 
PTGS2 T1532C rs5273 2843 vs 3216 (5§) n/a n/a n/a n/a n/a n/a n/a n/a — — 
TNF-α 308G>A rs1800629 3843 vs 4098 (9§) 1.11 (0.88 to 1.40) .37 72 (45 to 86) 0.65 1.13 (0.92 to 1.39) .24 47 (0 to 75) 0.21 0.97 BCB 
NOD2 3020incC‡ rs5743293 4222 vs 2988 (8) 1.39 (1.15 to 1.69) .001 0 (0 to 68) 0.94 2.81 (0.87 to 9.05) .08 0 (0 to 79) 0.26 0.34 AAB 
NOD2 G908R rs2066845 4541 vs 3820 (6§) 1.41 (1.04 to 1.91) .03 0 (0 to 75) 0.63 n/a n/a n/a n/a 0.87 BBB 
NOD2 R702W‡ rs2066844 3445 vs 2731 (6§) 1.22 (1.00 to 1.50) .06 12 (0 to 78) 0.49 1.23 (0.41 to 3.70) .71 0 (0 to 85) 0.06 0.91 BAB 
Inhibition of cell growth 
CCND1 870A‡ rs17852153 4747 vs 6783 (13) 13 1.13 (1.03 to 1.25) .02 0 (0 to 57) 0.70 13 1.16 (0.98 to 1.38) .09 51 (8 to 74) 0.78 0.85 BAB 
TGFB1 C509T‡‡ rs1800469 994 vs 2335 (5)  1.12 (0.91 to 1.37)  — — —  1.62 (1.30 to 2.02)  — — — 0.96 --C 
TGFBR1 TGFBR1*6A‡ rs11466445 3217 vs 4539 (8) 1.15 (1.01 to 1.31) .03 0 (0 to 68) — 1.51 (0.69 to 3.31) .31 52 (0 to 87) — 0.89 -AB 
Insulin-related 
IGF1 CA-repeat n/a 7900 vs 9161 (6)  1.06 (0.99 to 1.14)  .08  0 (0 to 75)  0.4  1.07 (0.97 to 1.18)  .15  0 (0 to 75)  0.31 0.97 CAB 
IGFBP3 202A>C rs2854744 7296 vs 10 452 (6)  1.02 (0.94 to 1.10)  .72  0 (0 to 75)  0.08  1.00 (0.91 to 1.10)  .96  0 (0 to 75)  0.05 0.99 CAB 
Iron metabolism 
HFE C282T rs1800562 5177 vs 6150 (6§,||)  1.08 (0.97 to 1.21)  .15  0 (0 to 75)  0.29  0.90 (0.55 to 1.48)  .69  0 (0 to 79)  0.07 0.97 CAB 
Lipid metabolism 
ApoE e2 rs7412 5821 vs 6754 (5§)  0.94 (0.85 to 1.04)  .23  66 (11 to 87)  0.21  0.75 (0.46 to 1.22)  .25  0 (0 to 85)  0.21 0.98 CCB 
ApoE e4 rs429358 3808 vs 4684 (5§)  1.04 (0.94 to 1.15)  .49  14 (0 to 82)  0.12  1.12 (0.83 to 1.52)  .45  0 (0 to 79)  0.12 0.98 CAB 
Mitotic control 
STK15 F31I rs2273535 4860 vs 4629 (4)  1.03 (0.94 to 1.12)  .58  0 (0 to 85)  0.1  1.30 (0.96 to 1.76)  .09  52 (0 to 84)  0.56 0.99 CAB 
One-carbon metabolism 
MTHFR C677T rs1801133†† 27 372 vs 39 867 (52||) 52 1.00 (0.94 to 1.06) .92 53 (36 to 65) 0.05 52 0.87 (0.82 to 0.91) .00 35 (10 to 53) 1.00 0.99 CBB 
MTHFR A1298C rs1801131 17 178 vs 24 792 (34||) 34  1.01 (0.97 to 1.06)  .51  0 (0 to 37)  0.08 34  0.94 (0.87 to 1.01)  .09  22 (0 to 49)  0.40 0.99 CAB 
MTR A2756G rs1805087 11 829 vs 15 975 (14||) 14  0.97 (0.92 to 1.02)  .27  12 (0 to 50)  0.21 14  0.96 (0.84 to 1.09)  .50  48 (4 to 72)  0.05 0.99 CAB 
MTRR A66G rs1801394 6170 vs 8732 (9)  0.98 (0.90 to 1.07)  .66  15 (0 to 57)  0.10  1.04 (0.94 to 1.14)  .47  23 (0 to 64)  0.17 0.99 CAB 
TS TSER‡ rs34743033 3519 vs 5289 (5) 0.86 (0.78 to 0.95) .003 18 (0 to 83) 0.87 0.83 (0.73 to 0.94) .004 17 (0 to 83) 0.85 0.55 AAB 
TS Ts1494del6 rs34489327 3262 vs 4518 (4)  0.96 (0.88 to 1.06)  .45  0 (0 to 85)  0.13  1.03 (0.88 to 1.19)  .73  0 (0 to 85)  0.07 0.98 CAB 
Rare, high penetrance 
APC E1317Q rs1801166 6898 vs 6668 (6)  1.13 (0.88 to 1.47)  .34  0 (0 to 75)  0.15 n/a n/a n/a n/a 0.96 CAB 
APC D1822V‡ rs459552 6282 vs 7038 (6) 0.99 (0.92 to 1.07) .83 24 (0 to 68) 0.06 0.84 (0.71 to 0.98) .03 0 (0 to 75) 0.55 0.99 CAB 
MLH1 I219V rs1799977 2956 vs 5071 (7§)  1.09 (0.90 to 1.32)  .40  55 (0 to 81)  0.42  1.01 (0.85 to 1.21)  .88  43 (0 to 76)  0.05 0.97 CCB 
MLH1 -93 G>A rs1800734 4524 vs 5544 (6§)  1.06 (0.97 to 1.15)  .23  0 (0 to 79)  0.27  1.15 (0.95 to 1.39)  .15  26 (0 to 71)  0.33 0.97 CAB 
Substrate metabolism 
CYP1A1 2454A>G‡ rs1048943 10 274 vs 11 978 (13§,||) 13 1.28 (1.01 to 1.63) .05 85 (7 to 91) 1.00 12 1.47 (1.17 to 1.85) .001 3 (0 to 60) 0.93 0.89 ACB 
CYP1A1 3698T>C rs4646903 4897 vs 6559 (7)  0.94 (0.86 to 1.04)  .23  0 (0 to 71)  0.27  0.84 (0.56 to 1.27)  .42  54 (0 to 80)  0.38 0.98 CAB 
CYP1A2 163C>A rs762551 3051 vs 5326 (9)  1.13 (0.95 to 1.34)  .18  63 (25 to 82)  0.75  1.07 (0.92 to 1.26)  .40  42 (0 to 73)  0.16 0.96 ACB 
CYP1B1 4326C>G rs1056836 8514 vs 9721 (6§)  0.99 (0.92 to 1.06)  .69  0 (0 to 75)  0.06  0.98 (0.90 to 1.07)  .70  0 (0 to 77)  0.08 0.99 CAB 
CYP2C9 430C>T rs1799853 5134 vs 6164 (6§)  0.93 (0.85 to 1.02)  .13  21 (0 to 65)  0.34  1.29 (0.99 to 1.70)  .06  0 (0 to 75)  0.46 0.97 CAB 
CYP2C9 1057A>C rs1057910 5379 vs 6531 (6§)  1.08 (0.83 to 1.40)  .57  73 (33 to 88)  0.26  0.68 (0.37 to 1.26)  .22  11 (0 to 86)  0.23 0.97 CCB 
CYP2E1 1053C>T rs2031920 4456 vs 5077 (8||)  0.93 (0.83 to 1.05)  .23  0 (0 to 68)  0.27  1.23 (0.92 to 1.63)  .16  35 (0 to 71)  0.34 0.97 CAB 
CYP2E1 1293G>C rs3813867 3424 vs 4686 (7)  1.17 (0.92 to 1.48)  .21  53 (0 to 80)  0.61  1.83 (0.94 to 3.57)  .08  0 (0 to 75)  0.55 0.95 BCB 
GSTA1 GSTA1*B§§  1648 vs 2039 (4)  1.03 (0.89 to 1.19)  — — —  1.09 (0.90 to 1.32)  — — — 0.98 — 
GSTM1 Null variant n/a 18 845 vs 26 662 (43) n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a — 
GSTP1 IIe105Val rs1695 9267 vs 12 902 (22§) 22  1.05 (0.99 to 1.12)  .11  0 (0 to 46)  0.38 22  0.95 (0.86 to 1.05)  .32  36 (0 to 62)  0.17 0.98 CAB 
GSTP1 Ala114Val rs1138272 5183 vs 5457 (6§,||)  1.02 (0.91 to 1.13)  .77  0 (0 to 75)  0.07  0.87 (0.55 to 1.37)  .55  12 (0 to 78)  0.09 0.99 CAB 
GSTT1 Null variant n/a 13 410 vs 20 455 (35) n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a — 
ΝΑΤ1 slow/rapid‡ n/a 4791 vs 6628 (15) 0.80 (0.68 to 0.93) .003 38 (0 to 74) 0.86 0.98 0.79 to 1.22) .97 0 (0 to 71) 0.06 0.57 ABB 
ΝΑΤ2 slow/rapid n/a 12 908 vs 16 483 (26) 15  1.01 (0.83 to 1.22)  .94  81 (70 to 88)  0.06 15  0.95 (0.76 to 1.20)  .68  64 (37 to 79)  0.17 0.98 CCB 
NQO1 Pro187Ser(C609T) rs1800566 5084 vs 5932 (8)  1.14 (0.96 to 1.35)  .12  64 (22 to 83)  0.89  1.10 (0.76 to 1.59)  .63  56 (2 to 80)  0.19 0.95 ACB 
Tumor suppressor genes 
TP53 Arg72Pro|||| rs1042522 7414 vs 9872 (27) 27  1.01 (0.89 to 1.14)  .90 — — 27  1.04 (0.82 to 1.31)  .77 — — 0.99 --C 
TP53 intron3 16bp|||| rs17878362 1637 vs 1874 (5)  1.58 (0.98 to 2.56)  — — —  1.14 (0.84 to 1.55)  — — — 0.92 --C 
MDM2 309 T/G|||| rs2279744 2543 vs 2115 (7)  0.73 (0.62 to 0.86)  — — —  0.86 (0.57 to 1.30)  — — — 0.10 --C 
Vitamin D and calcium metabolism 
VDR BsmI(60890GA) rs1544410‡ 5607 vs 6202 (7) 0.77 (0.58 to 1.02) .07 89 (81 to 94) 1.00 0.51 (0.28 to 0.90) .02 95 (92 to 97) 1.00 0.92 ACB 
VDR FokI rs10735810 7646 vs 8968 (9||)  0.97 (0.85 to 1.11)  .66  67 (32 to 83)  0.16  0.99 (0.82 to 1.20)  .94  70 (41 to 85)  0.06 0.98 CCB 
VDR TaqI rs731236 946 vs 1184 (4)  1.06 (0.87 to 1.30)  .55  0 (0 to 85)  0.09  1.00 (0.57 to 1.75)  .99  73 (8 to 92)  0.05 0.97 CAB 

* BFDP value for the heterozygous additive model (var/wt VS wt/wt) at prior probability of 0.05. BFDP level of noteworthiness 0.2. For more information please see Supplementary Table 6

Venice criteria grade for the the heterozygous additive model (var/wt VS wt/wt). For the third criterion (protection from bias), all meta-analyses of candidate gene association studies were scored with B. There was no obvious bias in the studies included, but there was considerable missing information on the generation of evidence. For the published meta-analyses, from which neither power nor I2 could be determined, we considered that there is considerable potential for bias (there were few small studies), so that third category would be “C.”

The associations considered to be “less-credible positives”.

§ Includes unpublished data from Scottish GWAS. || Includes unpublished data from Canadian GWAS. (IV) Liu et al. 2011 ( 131 ).

# McColgan and Sharma 2009 (64).

** Zhong et al. 2010 ( 137 ). †† The associations considered to be “positives”.

‡‡ Fang et al. 2010 ( 76 ).

§§ Economopoulos and Sergentanis 2010 ( 92 ).

|||| Economopoulos et al. 2010 ( 119 ).

¶¶ Hu et al. 2010 ( 144 ).

## Fang et al. 2011 ( 145 ).

Table 3.

Summary crude odds ratios (ORs) and 95% confidence intervals (95% CIs) for two additive models for variants that were identified from genome-wide association studies

Gene Variant Case vs control subjects (no. of samples) Additive Model: var/wt vs wt/wt Additive Model: var/var vs wt/wt Credibility 
OR (95% CI) P I2 (95% CI)  Power OR (95% CI) P I2 (95% CI)  Power BFDP* Venice criteria grade† 
Common, low penetrance             
SMAD7 ‡  rs4939827 37 650 vs 36 154 (13§) 13 0.89 (0.86 to 0.92)  3.7 × 10 -12 0 (0 to 57) 1.00 13 0.75 (0.71 to 0.79)  3.5 × 10 –41 39 (0 to 69) 1.00  <10 –3 AAA 
SMAD7 ‡  rs12953717 33 771 vs 32 364 (11‡) 11 1.11 (1.07 to 1.15)  1.4 × 10 –7 46 (0 to 73) 1.00 11 1.23 (1.16 to 1.29)  6.0 × 10 -13 36 (0 to 68) 1.00 0.35 ABA 
SMAD7 ‡  rs4464148 15 999 vs 15 216 (7||) 1.14 (1.08 to 1.19)  4.9 × 10 –7 6 (0 to 72) 1.00 1.30 (1.21 to 1.40)  7.5 × 10 -13 0 (0 to 71) 1.00  <10 –3 AAA 
8q24‡ rs6983267 40 604 vs 42 672 (19) 19 1.23 (1.18 to 1.27)  7.4 × 10 –30 38 (0 to 64) 1.00 19 1.45 (1.39 to 1.51)  6.9 × 10 –68 0 (0 to 49) 1.00  <10 –3 ABA 
8q24¶¶ rs10505477 18 580 vs 20 147(14) 14 1.21 (1.15 to 1.28)  1.0 × 10 -13 25 (0 to 60) 1.00 14 1.33 (1.26 to 1.41)  6.2 × 10 -23 44 (0 to 70) 1.00  <10 –3 ABA 
9p24¶¶ rs719725 13 290 vs 14 774 (13) 13 1.08 (1.00 to 1.16) .04 0 (0 to 57) 0.57 13 1.15 (1.07 to 1.24) .0002 0 (0 to 57) 0.96 0.93 BAA 
19q13.1‡ rs10411210 25 607 vs 26 477 (17) 17 0.87 (0.81 to 0.93)  7.8 × 10 -5 55 (23 to 74) 1.00 17 0.81 (0.70 to 0.93) .003 0 (0 to 51) 0.89 0.03 ABA 
16q22.1‡ rs9929218 26 191 vs 27 409 (18) 18 0.93 (0.90 to 0.97) .0001 13 (0 to 49) 0.98 18 0.84 (0.78 to 0.90)  1.9 × 10 –7 0 (0 to 50) 1.00 0.39 AAA 
15q14¶¶ rs4779584 13 656 vs 12 635 (9) 1.13 (1.02 to 1.24) .02 51 (0 to 77) 0.99 1.38 (1.09 to 1.73) .006 71 (47 to 85) 1.00 0.83 ABA 
1q41‡ rs6691170 17 740 vs 19 776 (11) 11 1.12 (1.07 to 1.17)  2.9 × 10 –7 19 (0 to 59) 1.00 11 1.19 (1.12 to 1.27)  1.2 × 10 –7 0 (0 to 60) 1.00  <10 –3 AAA 
3q26.2‡ rs10936599 17 802 vs 19 795 (11) 11 0.90 (0.86 to 0.94)  7.1 × 10 –7 33 (0 to 67) 1.00 11 0.85 (0.78 to 0.93) .0004 0 (0 to 60) 0.95 0.003 ABA 
12q13.13¶¶ rs11169552 17 148 vs 19 739 (11) 11 0.92 (0.88 to 0.96) .0003 0 (0 to 60) 0.97 11 0.75 (0.66 to 0.86)  1.2 × 10 -5 53 (7 to 76) 1.00 0.11 AAA 
20q13.33‡ rs4925386 17 847 vs 19 832 (11) 11 0.91 (0.87 to 0.95)  2.0 × 10 -5 0 (0 to 60) 0.99 11 0.80 (0.75 to 0.87)  6.2 × 10 -9 20 (0 to 59) 1.00 0.02 AAA 
14q22.2‡ rs4444235 18 607 vs 19 576 (13) 13 1.09 (1.04 to 1.14) .0004 12 (0 to 52) 0.95 13 1.18 (1.12 to 1.25)  1.3 × 10 –8 21 (0 to 59) 1.00 0.14 AAA 
20p12.3‡ rs961253 18 118 vs 19 006 (13) 13 1.13 (1.08 to 1.18)  1.6 × 10 –7 4 (0 to 58) 1.00 13 1.22 (1.14 to 1.30)  2.3 × 10 –8 23 (0 to 60) 1.00  <10 –3 AAA 
8q23.3‡ rs16892766 17 180 vs 17 840 (4)# 1.27 (1.20 to 1.33)  9.0 × 10 -20 0 (0 to 85) 1.00 1.38 (1.12 to 1.71) .003 0 (0 to 85) 0.90  <10 –3 AAA 
10p14‡ rs10795668 20 026 vs 20 682 (6)# 0.89 (0.80 to 0.99) .04 68 (23 to 86) 0.57 0.77 (0.72 to 0.82)  6.2 × 10 -15 40 (0 to 76) 1.00 0.90 BCA 
11q23.1‡ rs3802842 33 004 vs 31 654 (14) 14 1.15 (1.12 to 1.19)  1.2 × 10 -17 33 (0 to 64) 1.00 14 1.29 (1.23 to 1.36)  1.310 –20 33 (0 to 64) 1.00  <10 –3 ABA 
Gene Variant Case vs control subjects (no. of samples) Additive Model: var/wt vs wt/wt Additive Model: var/var vs wt/wt Credibility 
OR (95% CI) P I2 (95% CI)  Power OR (95% CI) P I2 (95% CI)  Power BFDP* Venice criteria grade† 
Common, low penetrance             
SMAD7 ‡  rs4939827 37 650 vs 36 154 (13§) 13 0.89 (0.86 to 0.92)  3.7 × 10 -12 0 (0 to 57) 1.00 13 0.75 (0.71 to 0.79)  3.5 × 10 –41 39 (0 to 69) 1.00  <10 –3 AAA 
SMAD7 ‡  rs12953717 33 771 vs 32 364 (11‡) 11 1.11 (1.07 to 1.15)  1.4 × 10 –7 46 (0 to 73) 1.00 11 1.23 (1.16 to 1.29)  6.0 × 10 -13 36 (0 to 68) 1.00 0.35 ABA 
SMAD7 ‡  rs4464148 15 999 vs 15 216 (7||) 1.14 (1.08 to 1.19)  4.9 × 10 –7 6 (0 to 72) 1.00 1.30 (1.21 to 1.40)  7.5 × 10 -13 0 (0 to 71) 1.00  <10 –3 AAA 
8q24‡ rs6983267 40 604 vs 42 672 (19) 19 1.23 (1.18 to 1.27)  7.4 × 10 –30 38 (0 to 64) 1.00 19 1.45 (1.39 to 1.51)  6.9 × 10 –68 0 (0 to 49) 1.00  <10 –3 ABA 
8q24¶¶ rs10505477 18 580 vs 20 147(14) 14 1.21 (1.15 to 1.28)  1.0 × 10 -13 25 (0 to 60) 1.00 14 1.33 (1.26 to 1.41)  6.2 × 10 -23 44 (0 to 70) 1.00  <10 –3 ABA 
9p24¶¶ rs719725 13 290 vs 14 774 (13) 13 1.08 (1.00 to 1.16) .04 0 (0 to 57) 0.57 13 1.15 (1.07 to 1.24) .0002 0 (0 to 57) 0.96 0.93 BAA 
19q13.1‡ rs10411210 25 607 vs 26 477 (17) 17 0.87 (0.81 to 0.93)  7.8 × 10 -5 55 (23 to 74) 1.00 17 0.81 (0.70 to 0.93) .003 0 (0 to 51) 0.89 0.03 ABA 
16q22.1‡ rs9929218 26 191 vs 27 409 (18) 18 0.93 (0.90 to 0.97) .0001 13 (0 to 49) 0.98 18 0.84 (0.78 to 0.90)  1.9 × 10 –7 0 (0 to 50) 1.00 0.39 AAA 
15q14¶¶ rs4779584 13 656 vs 12 635 (9) 1.13 (1.02 to 1.24) .02 51 (0 to 77) 0.99 1.38 (1.09 to 1.73) .006 71 (47 to 85) 1.00 0.83 ABA 
1q41‡ rs6691170 17 740 vs 19 776 (11) 11 1.12 (1.07 to 1.17)  2.9 × 10 –7 19 (0 to 59) 1.00 11 1.19 (1.12 to 1.27)  1.2 × 10 –7 0 (0 to 60) 1.00  <10 –3 AAA 
3q26.2‡ rs10936599 17 802 vs 19 795 (11) 11 0.90 (0.86 to 0.94)  7.1 × 10 –7 33 (0 to 67) 1.00 11 0.85 (0.78 to 0.93) .0004 0 (0 to 60) 0.95 0.003 ABA 
12q13.13¶¶ rs11169552 17 148 vs 19 739 (11) 11 0.92 (0.88 to 0.96) .0003 0 (0 to 60) 0.97 11 0.75 (0.66 to 0.86)  1.2 × 10 -5 53 (7 to 76) 1.00 0.11 AAA 
20q13.33‡ rs4925386 17 847 vs 19 832 (11) 11 0.91 (0.87 to 0.95)  2.0 × 10 -5 0 (0 to 60) 0.99 11 0.80 (0.75 to 0.87)  6.2 × 10 -9 20 (0 to 59) 1.00 0.02 AAA 
14q22.2‡ rs4444235 18 607 vs 19 576 (13) 13 1.09 (1.04 to 1.14) .0004 12 (0 to 52) 0.95 13 1.18 (1.12 to 1.25)  1.3 × 10 –8 21 (0 to 59) 1.00 0.14 AAA 
20p12.3‡ rs961253 18 118 vs 19 006 (13) 13 1.13 (1.08 to 1.18)  1.6 × 10 –7 4 (0 to 58) 1.00 13 1.22 (1.14 to 1.30)  2.3 × 10 –8 23 (0 to 60) 1.00  <10 –3 AAA 
8q23.3‡ rs16892766 17 180 vs 17 840 (4)# 1.27 (1.20 to 1.33)  9.0 × 10 -20 0 (0 to 85) 1.00 1.38 (1.12 to 1.71) .003 0 (0 to 85) 0.90  <10 –3 AAA 
10p14‡ rs10795668 20 026 vs 20 682 (6)# 0.89 (0.80 to 0.99) .04 68 (23 to 86) 0.57 0.77 (0.72 to 0.82)  6.2 × 10 -15 40 (0 to 76) 1.00 0.90 BCA 
11q23.1‡ rs3802842 33 004 vs 31 654 (14) 14 1.15 (1.12 to 1.19)  1.2 × 10 -17 33 (0 to 64) 1.00 14 1.29 (1.23 to 1.36)  1.310 –20 33 (0 to 64) 1.00  <10 –3 ABA 

|| Includes unpublished data from Scottish GWAS.

* Bayesian False Discovery Probability (BFDP) value for the heterozygous additive model (var/wt vs wt/wt) at prior probability of 0.05. BFDP level of noteworthiness 0.2. For more information, please see Supplementary Table 6 .

Venice criteria grade for the heterozygous additive model (var/wt vs wt/wt). For the third criterion (protection from bias), all meta-analyses of variants identified by GWAS were scored with A. Reporting was generally more transparent (at least for the discovery datasets, but somewhat variable for the replication datasets). All of the studies included the discovery data with the replication data in estimating magnitude of effect and arguably this may bias the magnitude of the association. However, it is unlikely to affect direction of association. Therefore, we rated the third category for GWASs as “A.”

# Tomlinson 2008 was based on 10 samples.

§ Includes unpublished data from Canadian GWAS.

‡ The associations considered to be “positives”.

¶¶ The associations considered to be “less-credible positives”.

To assess the credibility of genetic associations, we considered the Venice criteria ( 11 , 12 ) and the BFDP ( 26 ) ( Supplementary Table 6 , available online). Applying these filters indicate that associations with 16 variants (17% of the meta-analyzed SNPs) tagging 13 loci ( SMAD7, MUTYH, MTHFR, and the 8q24, 8q23.3, 11q23.1, 14q22.2, 1q41, 20p12.3, 20q13.33, 3q26.2, 16q22.1, and 19q13.1) represent the most credible findings and these will be referred as “positive” SNP associations ( Tables 2 and 3 ; Supplementary Tables 2 and 3 and Supplementary Figures 1–64 , available on the CRCgene website). These findings are based on accrued data on 16 000–40 604 case patients and on 15 216–42 672 control subjects, with a median of 1616 case patients per study. All variants that were included in the “positive” SNP associations reached genome-wide statistical significance ( P ≤ 1.7 × 10 –7 ) in at least one meta-analysis model, apart from the ones in MUTYH and in 19q13.1.

We identified “less-credible positive” associations (higher heterogeneity, lower statistical power, BFDP > 0.2) with 23 variants (25% of the meta-analyzed SNPs) of 22 loci ( Tables 2 and 3 ; Supplementary Tables 2 and 3 and Supplementary Figures 65–146 , available on the CRCgene website. These findings were based on accrued data on 857–26 370 case patients and on 932–26 662 control subjects, with a median of at least 600 case patients per study. Variants in the 10p14 and 15q14 loci reached genome-wide statistical significance ( P ≤ 1.7 × 10 –7 ) in at least one meta-analysis model.

For those SNPs that were identified as “positives” or “less-credible positives” after applying the BFDP and Venice criteria, we applied the model-free meta-analysis approach as described by Thompson et al. ( 32 ) ( Supplementary Table 7 , available online), which gives an estimate, λ, of the underlying genetic model. Funnel plots were produced for all the positive and less-credible associations. There was no evidence of small-study effects, apart from the associations with rs1799750 ( MMP1 ), rs34743033 ( TS ), GSTM1 deletion, rs36053993 ( MUTYH ), rs10411210 (19q13.1), rs10936599 (3q26.2). However, only the GSTM1 deletion and rs10411210 (19q13.1) and rs10936599 (3q26.2) tests were based on more than 10 studies and satisfied the conditions of no heterogeneity. Therefore, the results of the other SNPs, for which there was some evidence of a small-study effect, should be interpreted with caution. The remaining 53 (58%) meta-analyzed SNPs of 37 loci are designated as “negatives,” based on accrued data from 575 to 17 178 case patients and 836 to 24 792 control subjects, with an average of at least 600 case patients per study.

Discussion

This is the first systematic, comprehensive field synopsis of genetic association studies on colorectal cancer to our knowledge. It differs from the recently published database of cancer genetic association ( 33 ) in that it has collated and extracted data from more than 600 publications on more than 400 polymorphisms in 110 different genes and presents the results of more than 90 new meta-analyses, rather than acting as a portal to already published meta-analyses and GWASs. Furthermore, we have classified the results of our analysis as “positive associations,” “less-credible positive associations,” and “negative associations” within a defined statistical and causal inference framework, including the Venice criteria ( 11 , 12 ) the BDFP, and the model-free meta-analysis approach.

Two SNPs at the 8q24 locus (rs6983267 and rs10505477) identified in the GWASs had a large volume of evidence and showed positive associations irrespective of the genetic model considered, with moderate heterogeneity, and the model-free approach suggested an underlying additive model ( Table 3 , Supplementary Tables 3 and 7 ). The SNP rs6983267 may be a somatic target in colorectal cancer ( 34 ) and may be associated with enhanced responsiveness to Wnt signaling ( 35 ). Further, rs6983267 has also been found to be associated with other types of cancer ( 36 , 37 ), including prostate cancer ( 38–40 ). Interaction with the MYC proto-oncogene has been controversial ( 41–45 ), but recently in functional studies in cell lines, interaction between enhancer elements in the 8q24 locus and the MYC promoter, via transcription factor Tcf-4 binding and allele-specific regulation of MYC expression, has been demonstrated ( 46 ).

One variant close to SMAD7 (rs4939827), which was inversely associated with colorectal cancer, and two variants (rs12953717 and rs4464148), which were associated with increased risk for the disease, were again identified through GWASs with large numbers of participants. The associations were apparent irrespective of the genetic model, with moderate heterogeneity, and the model-free approach suggested an underlying additive model ( Table 3 , Supplementary Tables 3 and 7 ). The SMAD7 protein is an inhibitor for the TGFβ signaling pathway ( 47 ). The otherwise-tentative associations with the TGFβ1 rs1800469 (C5097) and TGFβR1 rs11466445 (*6A) variants may be consistent with this finding.

Dysfunction of base-excision repair, the major pathway for repairing oxidative damage, has been implicated in the development of multiple colorectal adenomas and colorectal cancer (MUTYH-associated polyposis [MAP] syndrome). Two variants of MUTYH were associated with increased risk of colorectal cancer, with little heterogeneity apparent ( Table 2 , Supplementary Table 2 ), providing further evidence that the single variants can act as lower-penetrance risk alleles of smaller effect, validating the recent results for Y179C (rs34612342) ( 48 ) and also highlighting the possible role of G396D (rs36053993). The BFDP was substantially less for rs36053993 than for rs34612342, possibly because of the very low MAF of the rs34612342 variant, and therefore, rs36053993 was classified as having a positive association, whereas rs34612342 was classified as having a less-credible positive association with colorectal cancer ( Supplementary Table 6 ). Although there was formally statistically significant evidence of small-study effects for rs36053993 ( Supplementary Figures 189–190 , available on the CRCgene website), the magnitude of this effect was small and has to be interpreted in the context of multiple tests for small-study effects. The model-free approach suggested a recessive model for rs36053993 and an additive (but with wide CI) for rs34612342 ( Supplementary Table 7 ).

An inverse association with homozygosity for the MTHFR rs1801133 (commonly reported as C677T) variant was observed in 52 studies comparing more than 27 000 case patients and almost 40 000 control subjects, with moderate heterogeneity and an underlying recessive model as suggested by the model-free approach ( Table 2 , Supplementary Tables 2 and 7 ). A similar association was observed in a smaller body of evidence with the rs1801131 (commonly known as A1298C) variant, with little heterogeneity between 34 studies ( Table 2 , Supplementary Table 4 ). In most populations, strong LD between the variants has been observed ( 49 , 50 ). These findings extend and concur with previous systematic reviews and meta-analyses ( 51–56 ). All of the studies considered these MTHFR variants as candidates for association with colorectal cancer on the basis of knowledge about their effects on enzyme function and associations with blood levels of folate and related metabolites. However, the evidence is further reinforced by GWASs of plasma homocysteine confirming some of the earlier work on the effects of the MTHFR C677T variant ( 57–60 ).

The other variants with which associations were observed were all based on accumulated data on a total of at least 17 000 case patients and 18 000 control subjects, with a median of 1506 case patients per study. Positive associations with colorectal cancer were identified with variants at the 8q23.3 (rs16892766), 11q23.1 (rs3802842), 14q22.2 (rs4444235), 1q41 (rs6691170), and 20p12.3 (rs961253) loci, and inverse associations were identified with variants at the 20q13.33 (rs4925386), 3q26.2 (rs10936599), 16q22.1 (rs9929218), and 19q13.1 (rs10411210) loci. Although there was formally statistically significant evidence of small-study effects for 3q26.2 (rs10936599) ( Supplementary Figures 209–210 , available on the CRCgene website) and 19q13.1 (rs10411210; Supplementary Figures 203–204 ), the magnitude of these effects was not large, and the findings of GWASs are less prone to the problem of selective reporting than studies that target only a few risk factors at a time with only selected findings being published ( 61 ). In a study of somatic genetic changes, no allelic imbalance targeting 11q23.1 (rs3802842), 14q22.2 (rs4444235), 16q22.1 (rs9929218), 19q13.1 (rs10411210), or 20p12.3 (rs961253) was observed ( Table 3 , Supplementary Table 3 ) ( 62 ).

The less-credible positive associations with 23 variants of 22 genes involved the following pathways—adhesion ( CDH1, MMP1, and MMP3 ); alcohol metabolism ( ALDH2 ); base-excision repair ( XRCC1 and MUTYH ); inflammation and immune response ( IL6 , two variants of NOD2 ); inhibition of cell growth ( CCND1, TGFB1, and TGFBR1 ); one-carbon metabolism ( TS ); substrate metabolism ( CYP1A1, GSTM1, GSTT1, and NAT1 ); vitamin D metabolism ( VDR )—common low-penetrance variants at 9p24 (rs719725), 10p14 (rs10795668), 15q14 (rs4779584), and 12q13.13 (rs11169552); and the common rs459552 (D1822V) variant of APC , for which rare variants confer a high risk of colorectal cancer. Some of these have been considered in previous meta-analyses.

The association with the CDH1 rs16260 (C160A) variant remained apparent when the analysis was restricted to populations of European origin ( Supplementary Tables 4 and 5 ). No association was apparent in an earlier meta-analysis by Wang et al. ( 63 ), but this meta-analysis included less than one-tenth of the number of case and control subjects included in the present study. Previous meta-analyses have found positive associations with the MMP1 rs1799750 (G-1607GG) but not the MMP3 rs3025058 (AAAAA 1612 AAAAAA) variant ( 64 , 65 ). We found evidence of small-study effects for MMP1 rs1799750 ( Supplementary Figures 149–150 , available on the CRCgene website.

With regard to ALDH2 , our findings are similar to those of Wang et al. ( 66 ). All of these studies were carried out in Asian populations, in which the rs671 (Glu487Lys) variant is more common than in the other populations (MAF is 0 in white, 0.16 in Chinese, and 0.23 in Japanese populations). The association between the ALDH2 variant and colorectal adenomas has been investigated, but it has been interpreted as null ( 67 , 68 ).

Inverse associations were observed with the XRCC1 rs25487 (Arg399Gln) variant in the additive and recessive models with heterogeneity, which persisted when the analysis was restricted to populations of European origin ( Table 2 , Supplementary Tables 2, 4, and 5 ). An earlier meta-analysis with about half the number of participants also detected an inverse association for the recessive model ( 69 ), but another meta-analysis that included a smaller number of participants (but more studies) was interpreted as null ( 70 ). It has been suggested that the association between this XRCC1 variant and colorectal adenomas is stronger than that apparent with colorectal cancer ( 71 ), but we note that the volume of evidence relating to adenomas is substantially smaller. Three variants of NOD1 (rs5743293, rs2066845, and rs2066844) are associated with increased risk when a dominant model is assumed ( Supplementary Table 2 ); this finding is consistent with that of Tian et al. ( 72 ).

Assuming a dominant model, variants at three gene loci that are involved in the inhibition of cell growth are associated with an increased risk of colorectal cancer. First, the analysis of the CCND1 rs17852153 (G870A) variant includes approximately five times as many participants as the earlier study of Tan et al. ( 73 ) ( Supplementary Table 4 ). In the study of Tan et al., there was a concern about small-study effects, but we found no evidence of this problem ( Supplementary Figures 165–166 ). Two studies of colorectal adenomas give inconsistent results ( 74 , 75 ). Second, the TGFB1 rs1800469 (C5097) variant has been investigated solely in Asian populations ( Supplementary Table 2 ) ( 76 ). Third, previous meta-analyses have drawn different conclusions about the association with the TGFBR1 rs11466445 (*6A) variant ( 77–79 ). The magnitude of association in the present meta-analysis is similar to that determined in previous reports ( 78 , 79 ), but the inclusion of data on an additional 2000 participants from two studies has increased statistical power but not heterogeneity ( Supplementary Table 2 ).

Our finding of an inverse association with the TS rs34743033 (TSER) variant assuming either an additive or dominant model remained apparent when the analysis was restricted to populations of European origin ( Table 2 , Supplementary Tables 2, 4, and 5 ), although there was evidence of small-study effects for the comparison between homozygotes under the additive model. Two studies of colorectal adenomas do not suggest an overall association with this variant ( 80 , 81 ), and one study did not detect an association with adenoma recurrence ( 82 ).

There has been a longstanding interest in the possible effects of genetic variants influencing the metabolism of diverse substrates, including potential carcinogens, on cancer risk. Cytochrome P-450 CYP1A1 is involved in the metabolism of polycyclic aromatic hydrocarbons and estrogens, and perhaps cruciferous vegetables and heterocyclic amines ( 83 ). Further, CYP1A1 mRNA has been detected in the large bowel ( 84 , 85 ), and human colon cell lines also express CYP1A1 at the protein and mRNA levels ( 86 ). These observations stimulated many candidate gene studies of the association between colorectal cancer and CYP1A1 variants. The positive association with presence of the rs1048943 (A2454G, Ile462Val) variant apparent in the present analysis is in line with the results of a smaller recent meta-analysis ( 87 ), whereas an earlier meta-analysis ( 88 ) did not detect this association ( Table 2 , Supplementary Table 2 ). No association was apparent with the rs4646903 variant, but there was low statistical power to detect this association ( Table 2 , Supplementary Table 2 ).

Another intensively investigated gene polymorphism that is thought to affect the metabolism of multiple substrates is the GSTM1 null variant, with which we observed a weak positive association in data from more than 18 000 case patients and more than 26 000 control subjects from 43 studies ( Supplementary Table 2 ), with some evidence of small-study effects. The weak positive association was also apparent when analysis was restricted to 25 studies in populations of European origin ( Supplementary Table 5 ). There have been several previous meta-analyses; the earliest of these studies indicated no overall association ( 88–90 ), whereas the more recent studies showed a positive association ( 91–93 ). No clear association between GSTM1 and adenomas has been observed ( 94 ). In data from more than 13 000 case patients and 20 000 control subjects from 35 studies, there was also a positive association with the GSTT1 null deletion, in line with previous meta-analyses ( Supplementary Table 2 ) ( 88 , 92 , 95 , 96 ). The positive association remained apparent when analysis was restricted to approximately half the number of participants from 20 studies in populations of European origin ( Supplementary Table 5 ). Studies of GSTT1 and adenomas have not detected a clear association ( 94 ).

In data from nearly 5000 case patients and more than 6000 control subjects from 15 studies, there was a positive association with heterogeneity for NAT1 alleles classified as being associated with rapid acetylation ( Table 2 , Supplementary Table 2 ). No association was apparent with homozygosity for alleles associated with rapid acetylation or assuming other models of inheritance. Furthermore, the association was not apparent when analysis was restricted to populations of European origin ( Supplementary Tables 4 and 5 ). A previous meta-analysis found no association with carriage of the NAT1*10 variant in less than 1000 participants from three studies ( 88 ).

There was an inverse association between colorectal cancer and the VDR rs1544410 (Bsml 6089GA) variant, but with considerable heterogeneity between studies ( Table 2 , Supplementary Table 2 ). This heterogeneity was reduced when analysis was limited to populations of European origin ( Supplementary Tables 4 and 5 ). No association between this variant and colorectal adenomas ( 97–99 ) or recurrence ( 100 , 101 ) has been observed.

On the basis of GWASs, a positive association with the rs719725 variant at 9p24 was observed with moderate heterogeneity in data from more than 11 000 case patients and 13 000 control subjects from 13 studies. This finding is consistent with the recent meta-analysis of Kocarnick et al. ( 102 ).

Assuming a recessive mode of inheritance, an inverse association was found with the rs10795668 at 10p14, based on data from more than 20 000 case patients and more than 20 000 control subjects from six studies ( Table 3 ). The magnitude of association was attenuated when the analysis was restricted to populations of European origin ( Supplementary Table 4 ). In a study of somatic genetic changes, no allelic imbalance targeting this SNP was observed ( 62 ).

We considered the associations with one variant at 15q14 (rs4779584) and one at 12q13.13 (rs11169552) to lie in the less-credible category of associations because of heterogeneity between studies. Based on data from more than 13 500 case patients and 12 500 control subjects from nine studies, there was a positive association with the rs4779584 variant ( Table 3 , Supplementary Table 3 ), and based on data from nearly 18 000 case patients and nearly 20 000 control subjects from 11 studies, an inverse association with the rs11169552 variant ( Table 3 , Supplementary Table 3 ). In a study of somatic genetic changes, no allelic imbalance targeting the former variant was observed ( 62 ).

On the basis of studies carried out solely in populations of European origin, homozygosity for the common rs459552 (D1822V) variant of APC, rare truncating germline variants of which cause familial adenomatous polyposis (FAP), was inversely associated with colorectal cancer ( Supplementary Table 4 ). No marginal association between this variant and colorectal adenomas ( 103 , 104 ) or advanced adenomas (ie, ≥1cm or containing villous elements or high-grade dysplasia) of the distal colon ( 105 ) has been observed.

For variants for which we did not find associations in our meta-analyses, the accumulated number of case patients ranged between nearly 600 and slightly more than 15 000, with 20 variants in which the accumulated number of case patients exceeded 5000 ( Table 2 , Supplementary Table 2 ).

The lack of association with the PPARγ rs1801282 variant does not support an earlier meta-analysis ( 106 ), whereas the lack of association with the PPARγ rs3856606 variant is consistent with the results of Lu et al. ( 107 ). Earlier work on NAT2 suggested a positive association with the rapid acetylator phenotype, but no association with acetylator status could be inferred from genotype ( 51 , 88 , 108 , 109 ). No overall association was apparent in an earlier meta-analysis with homozygosity for the MTR rs1805087 (A2576G) variant, based on nine studies ( 110 ). The current analysis approximately doubles the data on colorectal cancer and the GSTP1 rs1695 (Ile105Val) variant. As in the meta-analysis of Economoupoulos and Sergentanis ( 92 ), it does not confirm an earlier analysis that found an inverse association when a recessive model was assumed ( 111 ). Studies of colorectal adenomas do not suggest any association with this GSTP1 variant, although subgroup effects have been reported ( 112–115 ).

The results presented here are in accordance with previous meta-analyses that did not show any association with the IGF1 CA-repeat polymorphism ( 116 , 117 ), although one of these showed an association with a higher serum concentration of IGF1 ( 117 ). Our finding of no association with the IGFBP3 rs2854744 (A202C) variant is consistent with that reported by Chen ( 117 ).

There have been numerous studies assessing the possible relationship between common variants of the TP53 tumor suppressor gene and several cancer types. The relationship between colorectal cancer and the rs1042522 (codon 72) variant has been investigated in previous meta-analyses, none of which suggest an association across multiple small studies, although marked heterogeneity was observed ( 118–121 ). Of note, one of these highlights the differences in results according to genotyping method and source of tissue for genotyping case patients ( 118 ). With regard to the APC rs1801166 (E1317Q) variant, one previous analysis of colorectal cancer was interpreted as indicating no association ( 122 ).

The current analysis examines more than double the amount of genetic association data for the XRCC1 variant (Arg194Trp) and is consistent with previous meta-analyses indicating no association ( 69 , 70 ).

Because of the inverse association between aspirin and colorectal neoplasia, many investigations have been carried out on genetic variants that might influence cyclooxygenase (COX) metabolism. We did not find any association with PTGS2/COX2 variants: neither with the rs20417 (G765C) variant (for which there is most evidence), in accordance with Cao ( 123 ), nor with the other variants. Individually, adenoma studies did not suggest associations for newly detected adenomas with the rs20417 ( 124–127 ), the rs5275 ( 124 , 125 ), the rs5277 ( 124 , 125 ), or the rs689466 variants ( 125 , 126 ). With regard to increased risk of adenoma recurrence, individual studies found no associations with the rs20417 ( 128 , 129 ) or the rs5275 variants ( 129 ), but one study reported an increased risk associated with homozygosity for the uncommon variant of rs5277 ( 129 ).

The lack of association with the NQ01 rs1800566 variant was also found in an earlier meta-analysis based on about one-third the number of subjects ( 130 ). With regard to the other variants for which we conducted meta-analyses, our finding of no association is consistent with earlier meta-analyses for MMP9 ( 64 ), VEGF rs2010963 (G634C) ( 131 ), XRCC1 rs25489 (Arg280His) ( 70 ), XRCC3 rs861359 (Thr241Met) ( 69 ), TNFα rs1800629 (308G>A) ( 132 ), and CYP2E1 rs3813867 (Pstl/Rsal, 1293G>C) ( 133 ).

We were not able to identify previous meta-analyses of ADH1B rs1229984, ADH1C rs698, VEGF rs3025039, OGG1 rs1052133, IL8 rs4073, IL10 rs1800896, APOE rs429358, TS rs34489327, CYP1A1 rs4646903, CYP1A2 rs762551, CYP2E1 rs2031920, GSTP1 rs1138272, VDR rs731236 or the MLH1 rs1799977, and rs1800734 variants.

The substantial majority of studies are from Western Europe, North America, and parts of Asia (more than one study from China, Japan, Korea, Singapore, and Taiwan). Investigation of more diverse populations is important, as it will enable variants to be considered in populations with more diverse patterns of LD and gene–environment interaction.

The number of common, low-penetrance variants that appear to be associated with colorectal cancer is very much less than anticipated, therefore decreasing the feasibility of combining variants as a profile in a prediction tool for stratifying screening modalities or primary prevention approaches ( 134 ). In addition, the variants so far identified account for only a small proportion of the familial risk (6.1% for “positive” and 14.1% when adding the “less-credible positive” variants; Table 1 ). As for other common diseases, there is increasing interest in the effects of lower-frequency SNPs and rare variants, types of genetic variation other than SNPs, and the effects of possible gene–environment and gene–gene interactions. Further, advances in technology enable the capture of large amounts of data on types of biological variations, in addition to genomic, including epigenetic, proteomic, transcriptomic, and metabolomic variations. These developments make this field synopsis the beginning of a work in progress, and we plan regular searches of the literature using the methodology described here and will continuously update the website when new data are available ( http://www.chs.med.ed.ac.uk/CRCgene/ ).

There is an overlap of some of these loci and variants with the loci of genetic risk factors for other common complex diseases and disease traits ( 135 ). In particular, several of these genes show genome-wide associations ( P < 5 × 10 –8 ; replicated in at least one other study) with other cancers: breast cancer ( CCND1 ), prostate cancer (8q24), and bladder cancer ( GSTM1 ); or other diseases or disease traits: mean corpuscular hemoglobin concentration and alanine aminotransferase levels ( ALDH2 ), ulcerative colitis ( CDH1 ), diastolic blood pressure ( CYP1A1 ), CRP ( IL6 ), serum matrix metalloproteinase ( MMP1 and MMP3 ), systolic blood pressure, and plasma homocysteine ( MTHFR ), Crohn’s disease and inflammatory bowel disease ( NOD2 ). The pleiotropic associations with 8q24, MTHFR, and ALDH2 are with the same SNP that has been reported to have an association with colorectal cancer. These genome-wide associations are listed at http://www.genome.gov/gwastudies/ , but because the causal genetic variant is not known in most reports, there may be some instances of wrong assignment of genes to these observed associations. Nevertheless, reports of robust associations with other diseases in 11 genes (35% of genes showing association with colorectal cancer and 17% of all meta-analyzed genes) highlight the importance of pleiotropy. This presumably reflects a common underlying pathological process underlying these conditions, for example, between (i) colorectal, breast, esophageal, bladder, and prostate cancers ( CND1 , MYC , GSTM1 ) and (ii) colorectal cancer and inflammatory bowel disease ( CDH1 ). Other pleiotropic links are less readily explained but may yield clues to novel underlying pathological processes.

We have conducted a comprehensive exercise to capture and meta-analyze all SNP data for variants with MAFs in the range 0.01–0.49. The analysis clearly identifies 16 variants for which there is robust evidence of impact on risk of colorectal cancer; 23 variants for which further evidence through international collaboration should be generated; and 20 variants for which the overall evidence does not support association and on which further research is not warranted. With increasing availability of data from multiple SNPs, it is clear that studies to test associations must achieve very high levels of statistical stringency. It has been suggested that even for candidate SNPs, that statistical support should reach genome-wide thresholds. Nonetheless, the analysis here provides a resource for mining available data and puts into context the sample sizes required for the identification of true associations. This study highlights a number of SNP associations that could be incorporated into genetic risk–prediction algorithms as further risk factors are identified and highlights the loci at which further research effort should be targeted. The data are all lodged on the CRCgene database ( http://www.chs.med.ed.ac.uk/CRCgene/ ).

Funding

Cancer Research, UK (C348/A3758 and A8896, C48/A6361); Medical Research Council (G0000657-53203); Scottish Executive Chief Scientist’s Office (K/OPR/2/2/D333, CZB/4/449); and a Centre Grant from CORE as part of the Digestive Cancer Campaign; Cancer Research UK Fellowship (C31250/A10107 to ET); Canadian Institutes of Health Research (CIHR) Team in Interdisciplinary Research on Colorectal Cancer (CTP-79845); a CIHR pilot project grant in colorectal cancer screening (200509CCS-152119-CCS-CECA-102806); the Cancer Risk Evaluation (CaRE) Program Grant from the Canadian Cancer Society Research Institute (18001).

References

1.
Lichtenstein
P
Holm
NV
Verkasalo
PK
et al
Environmental and heritable factors in the causation of cancer–analyses of cohorts of twins from Sweden, Denmark, and Finland.
N Engl J Med
 
2000
343
(
2
):
78
85
2.
Tenesa
A
Dunlop
MG
New insights into the aetiology of colorectal cancer from genome-wide association studies.
Nat Rev Genet
 
2009
10
(
6
):
353
358
3.
Zanke
BW
Greenwood
CM
Rangrej
J
et al
Genome-wide association scan identifies a colorectal cancer susceptibility locus on chromosome 8q24.
Nat Genet
 
2007
39
(
8
):
989
994
4.
Tomlinson
I
Webb
E
Carvajal-Carmona
L
et al
;
CORGI Consortium
A genome-wide association scan of tag SNPs identifies a susceptibility variant for colorectal cancer at 8q24.21.
Nat Genet
 
2007
39
(
8
):
984
988
5.
Tomlinson
IP
Webb
E
Carvajal-Carmona
L
et al
;
CORGI Consortium
EPICOLON Consortium
A genome-wide association study identifies colorectal cancer susceptibility loci on chromosomes 10p14 and 8q23.3.
Nat Genet
 
2008
40
(
5
):
623
630
6.
Tenesa
A
Farrington
SM
Prendergast
JG
et al
Genome-wide association scan identifies a colorectal cancer susceptibility locus on 11q23 and replicates risk loci at 8q24 and 18q21.
Nat Genet
 
2008
40
(
5
):
631
637
7.
Broderick
P
Carvajal-Carmona
L
Pittman
AM
et al
;
CORGI Consortium
A genome-wide association study shows that common alleles of SMAD7 influence colorectal cancer risk.
Nat Genet
 
2007
39
(
11
):
1315
1317
8.
Jaeger
E
Webb
E
Howarth
K
et al
;
CORGI Consortium
Common genetic variants at the CRAC1 (HMPS) locus on chromosome 15q13.3 influence colorectal cancer risk.
Nat Genet
 
2008
40
(
1
):
26
28
9.
Houlston
RS
Webb
E
Broderick
P
et al
;
Colorectal Cancer Association Study Consortium
CoRGI Consortium
International Colorectal Cancer Genetic Association Consortium
Meta-analysis of genome-wide association data identifies four new susceptibility loci for colorectal cancer.
Nat Genet
 
2008
40
(
12
):
1426
1435
10.
Houlston
RS
Cheadle
J
Dobbins
SE
et al
;
COGENT Consortium
CORGI Consortium
COIN Collaborative Group
COINB Collaborative Group
Meta-analysis of three genome-wide association studies identifies susceptibility loci for colorectal cancer at 1q41, 3q26.2, 12q13.13 and 20q13.33.
Nat Genet
 
2010
42
(
11
):
973
977
11.
Ioannidis
JP
Boffetta
P
Little
J
et al
Assessment of cumulative evidence on genetic associations: interim guidelines.
Int J Epidemiol
 
2008
37
(
1
):
120
132
12.
Khoury
MJ
Bertram
L
Boffetta
P
et al
Genome-wide association studies, field synopses, and the development of the knowledge base on genetic variation and human diseases.
Am J Epidemiol
 
2009
170
(
3
):
269
279
13.
Allen
NC
Bagade
S
McQueen
MB
et al
Systematic meta-analyses and field synopsis of genetic association studies in schizophrenia: the SzGene database.
Nat Genet
 
2008
40
(
7
):
827
834
14.
Bertram
L
McQueen
MB
Mullin
K
Blacker
D
Tanzi
RE
Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database.
Nat Genet
 
2007
39
(
1
):
17
23
15.
Chatzinasiou
F
Lill
CM
Kypreou
K
et al
Comprehensive field synopsis and systematic meta-analyses of genetic association studies in cutaneous melanoma.
J Natl Cancer Inst
 
2011
103
(
16
):
1227
1235
16.
Campbell
H
Rudan
I
Interpretation of genetic association studies in complex disease.
Pharmacogenomics J
 
2002
2
(
6
):
349
360
17.
Yu
W
Clyne
M
Khoury
MJ
Gwinn
M
Phenopedia and Genopedia: disease-centered and gene-centered views of the evolving knowledge of human genetic associations.
Bioinformatics
 
2010
26
(
1
):
145
146
18.
1000 Genomes Project Consortium
A map of human genome variation from population-scale sequencing
Nature.
 
2010
467
1061
1073
19.
Theodoratou
E
Kyle
J
Cetnarskyj
R
et al
Dietary flavonoids and the risk of colorectal cancer.
Cancer Epidemiol Biomarkers Prev
 
2007
16
(
4
):
684
693
20.
Cotterchio
M
Manno
M
Klar
N
McLaughlin
J
Gallinger
S
Colorectal screening is associated with reduced colorectal cancer risk: a case-control study within the population-based Ontario Familial Colorectal Cancer Registry.
Cancer Causes Control
 
2005
16
(
7
):
865
875
21.
StataCorp.
Stata Statistical Software: Release 11 .
 
College Station, TX
StataCorp LP;
2009
22.
Ioannidis
JP
Trikalinos
TA
The appropriateness of asymmetry tests for publication bias in meta-analyses: a large survey.
CMAJ
 
2007
176
(
8
):
1091
1096
23.
Dupont
WD
Plummer
WD
Jr.
Power and sample size calculations. A review and computer program.
Control Clin Trials
 
1990
11
(
2
):
116
128
24.
Houlston
RS
Ford
D
Genetics of coeliac disease.
QJM
 
1996
89
(
10
):
737
743
25.
Cox
A
Dunning
AM
Garcia-Closas
M
et al
A common coding variant in CASP8 is associated with breast cancer risk.
Nat Genet
 
2007
39
(
3
):
352
358
26.
Wakefield
J
A Bayesian measure of the probability of false discovery in genetic epidemiology studies.
Am J Hum Genet
 
2007
81
(
2
):
208
227
27.
Pompanon
F
Bonin
A
Bellemain
E
Taberlet
P
Genotyping errors: causes, consequences and solutions.
Nat Rev Genet
 
2005
6
(
11
):
847
859
28.
Marchini
J
Cardon
LR
Phillips
MS
Donnelly
P
The effects of human population structure on large genetic association studies.
Nat Genet
 
2004
36
(
5
):
512
517
29.
Balding
DJ
A tutorial on statistical methods for population association studies.
Nat Rev Genet
 
2006
7
(
10
):
781
791
30.
Wacholder
S
Rothman
N
Caporaso
N
Population stratification in epidemiologic studies of common genetic variants and cancer: quantification of bias.
J Natl Cancer Inst
 
2000
92
(
14
):
1151
1158
31.
Ioannidis
JP
Ntzani
EE
Trikalinos
TA
‘Racial’ differences in genetic effects for complex diseases.
Nat Genet
 
2004
36
(
12
):
1312
1318
32.
Thompson
JR
Minelli
C
Abrams
KR,
et al
Combining information from related meta-analyses of genetic association studies.
Appl Statist.
 
2008
57
103
115
33.
Schully
SD
Yu
W
McCallum
V
et al
Cancer GAMAdb: database of cancer genetic associations from meta-analyses and genome-wide association studies.
Eur J Hum Genet
 
2011
19
(
8
):
928
930
34.
Tuupanen
S
Niittymäki
I
Nousiainen
K
et al
Allelic imbalance at rs6983267 suggests selection of the risk allele in somatic colorectal tumor evolution.
Cancer Res
 
2008
68
(
1
):
14
17
35.
Tuupanen
S
Turunen
M
Lehtonen
R
et al
The common colorectal cancer predisposition SNP rs6983267 at chromosome 8q24 confers potential to enhanced Wnt signaling.
Nat Genet
 
2009
41
(
8
):
885
890
36.
Park
SL
Chang
SC
Cai
L
et al
Associations between variants of the 8q24 chromosome and nine smoking-related cancer sites.
Cancer Epidemiol Biomarkers Prev
 
2008
17
(
11
):
3193
3202
37.
Wokolorczyk
D
Gliniewicz
B
Sikorski
A
et al
A range of cancers is associated with the rs6983267 marker on chromosome 8.
Cancer Res
 
2008
68
(
23
):
9982
9986
38.
Yeager
M
Orr
N
Hayes
RB
et al
Genome-wide association study of prostate cancer identifies a second risk locus at 8q24.
Nat Genet
 
2007
39
(
5
):
645
649
39.
Thomas
G
Jacobs
KB
Yeager
M
et al
Multiple loci identified in a genome-wide association study of prostate cancer.
Nat Genet
 
2008
40
(
3
):
310
315
40.
Eeles
RA
Kote-Jarai
Z
Giles
GG
et al
Multiple newly identified loci associated with prostate cancer susceptibility.
Nat Genet
 
2008
40
(
3
):
316
321
41.
Pomerantz
MM
Ahmadiyeh
N
Jia
L
et al
The 8q24 cancer risk variant rs6983267 shows long-range interaction with MYC in colorectal cancer.
Nat Genet
 
2009
41
(
8
):
882
884
42.
Cicek
MS
Slager
SL
Achenbach
SJ
et al
Functional and clinical significance of variants localized to 8q24 in colon cancer.
Cancer Epidemiol Biomarkers Prev
 
2009
18
(
9
):
2492
2500
43.
Prokunina-Olsson
L
Hall
JL
No effect of cancer-associated SNP rs6983267 in the 8q24 region on co-expression of MYC and TCF7L2 in normal colon tissue.
Mol Cancer
 
2009
8
96
44.
Wright
JB
Brown
SJ
Cole
MD
Upregulation of c-MYC in cis through a large chromatin loop linked to a cancer risk-associated single-nucleotide polymorphism in colorectal cancer cells.
Mol Cell Biol
 
2010
30
(
6
):
1411
1420
45.
Verzi
MP
Hatzis
P
Sulahian
R
et al
TCF4 and CDX2, major transcription factors for intestinal function, converge on the same cis-regulatory regions.
Proc Natl Acad Sci U S A
 
2010
107
(
34
):
15157
15162
46.
Sotelo
J
Esposito
D
Duhagon
MA
et al
Long-range enhancers on 8q24 regulate c-Myc.
Proc Natl Acad Sci U S A
 
2010
107
(
7
):
3001
3005
47.
Loh
YH
Mitrou
PN
Wood
A
et al
SMAD7 and MGMT genotype variants and cancer incidence in the European Prospective Investigation into Cancer and Nutrition (EPIC)-Norfolk Study.
Cancer Epidemiol
 
2011
35
(
4
):
369
374
48.
Theodoratou
E
Campbell
H
Tenesa
A
et al
A large-scale meta-analysis to refine colorectal cancer risk estimates associated with MUTYH variants.
Br J Cancer
 
2010
103
(
12
):
1875
1884
49.
Shi
M
Caprau
D
Romitti
P
Christensen
K
Murray
JC
Genotype frequencies and linkage disequilibrium in the CEPH human diversity panel for variants in folate pathway genes MTHFR, MTHFD, MTRR, RFC1, and GCP2.
Birth Defects Res Part A Clin Mol Teratol
 
2003
67
(
8
):
545
549
50.
Mao
R
Fan
Y
Chen
F
Sun
D
Bai
J
Fu
S
Methylenetetrahydrofolate reductase gene polymorphisms in 13 Chinese ethnic populations.
Cell Biochem Funct
 
2008
26
(
3
):
352
358
51.
Houlston
RS
Tomlinson
IP
Polymorphisms and colorectal tumor risk.
Gastroenterology
 
2001
121
(
2
):
282
301
52.
Sharp
L
Little
J
Polymorphisms in genes involved in folate metabolism and colorectal neoplasia: a HuGE review.
Am J Epidemiol
 
2004
159
(
5
):
423
443
53.
Huang
Y
Han
S
Li
Y
Mao
Y
Xie
Y
Different roles of MTHFR C677T and A1298C polymorphisms in colorectal adenoma and colorectal cancer: a meta-analysis.
J Hum Genet
 
2007
52
(
1
):
73
85
54.
Hubner
RA
Houlston
RS
MTHFR C677T and colorectal cancer risk: A meta-analysis of 25 populations.
Int J Cancer
 
2007
120
(
5
):
1027
1035
55.
Taioli
E
Garza
MA
Ahn
YO
et al
Meta- and pooled analyses of the methylenetetrahydrofolate reductase (MTHFR) C677T polymorphism and colorectal cancer: a HuGE-GSEC review.
Am J Epidemiol
 
2009
170
(
10
):
1207
1221
56.
Zacho
J
Yazdanyar
S
Bojesen
SE
Tybjærg-Hansen
A
Nordestgaard
BG
Hyperhomocysteinemia, methylenetetrahydrofolate reductase c.677C>T polymorphism and risk of cancer: cross-sectional and prospective studies and meta-analyses of 75,000 cases and 93,000 controls.
Int J Cancer
 
2011
128
(
3
):
644
652
57.
Tanaka
T
Scheet
P
Giusti
B
et al
Genome-wide association study of vitamin B6, vitamin B12, folate, and homocysteine blood concentrations.
Am J Hum Genet
 
2009
84
(
4
):
477
482
58.
Hazra
A
Kraft
P
Lazarus
R
et al
Genome-wide significant predictors of metabolites in the one-carbon metabolism pathway.
Hum Mol Genet
 
2009
18
(
23
):
4677
4687
59.
Paré
G
Chasman
DI
Parker
AN
et al
Novel associations of CPS1, MUT, NOX4, and DPEP1 with plasma homocysteine in a healthy population: a genome-wide evaluation of 13 974 participants in the Women’s Genome Health Study.
Circ Cardiovasc Genet
 
2009
2
(
2
):
142
150
60.
Lange
LA
Croteau-Chonka
DC
Marvelle
AF
et al
Genome-wide association study of homocysteine levels in Filipinos provides evidence for CPS1 in women and a stronger MTHFR effect in young adults.
Hum Mol Genet
 
2010
19
(
10
):
2050
2058
61.
Khoury
MJ
Little
J
Gwinn
M
Ioannidis
JP
On the synthesis and interpretation of consistent but weak gene-disease associations in the era of genome-wide association studies.
Int J Epidemiol
 
2007
36
(
2
):
439
445
62.
Niittymäki
I
Tuupanen
S
Li
Y
et al
Systematic search for enhancer elements and somatic allelic imbalance at seven low-penetrance colorectal cancer predisposition loci.
BMC Med Genet
 
2011
12
23
63.
Wang
GY
Lu
CQ
Zhang
RM
Hu
XH
Luo
ZW
The E-cadherin gene polymorphism 160C->A and cancer risk: A HuGE review and meta-analysis of 26 case-control studies.
Am J Epidemiol
 
2008
167
(
1
):
7
14
64.
McColgan
P
Sharma
P
Polymorphisms of matrix metalloproteinases 1, 2, 3 and 9 and susceptibility to lung, breast and colorectal cancer in over 30,000 subjects.
Int J Cancer
 
2009
125
(
6
):
1473
1478
65.
Peng
B
Cao
L
Wang
W
et al
Polymorphisms in the promoter regions of matrix metalloproteinases 1 and 3 and cancer risk: a meta-analysis of 50 case-control studies.
Mutagenesis
 
2010
25
(
1
):
41
48
66.
Wang
J
Wang
H
Chen
Y
Hao
P
Zhang
Y
Alcohol ingestion and colorectal neoplasia: a meta-analysis based on a Mendelian randomization approach.
Colorectal Dis
 
2011
13
(
5
):
e71
e78
67.
Takeshita
T
Morimoto
K
Yamaguchi
N
et al
Relationships between cigarette smoking, alcohol drinking, the ALDH2 genotype and adenomatous types of colorectal polyps in male self-defense force officials.
J Epidemiol
 
2000
10
(
6
):
366
371
68.
Hirose
M
Kono
S
Tabata
S
et al
Genetic polymorphisms of methylenetetrahydrofolate reductase and aldehyde dehydrogenase 2, alcohol use and risk of colorectal adenomas: Self-Defense Forces Health Study.
Cancer Sci
 
2005
96
(
8
):
513
518
69.
Jiang
Z
Li
C
Xu
Y
Cai
S
A meta-analysis on XRCC1 and XRCC3 polymorphisms and colorectal cancer risk.
Int J Colorectal Dis
 
2010
25
(
2
):
169
180
70.
Wang
B
Wang
D
Huang
G
Zhang
C
Xu
DH
Zhou
W
XRCC1 polymorphisms and risk of colorectal cancer: a meta-analysis.
Int J Colorectal Dis
 
2010
25
(
3
):
313
321
71.
Naccarati
A
Pardini
B
Hemminki
K
Vodicka
P
Sporadic colorectal cancer and individual susceptibility: a review of the association studies investigating the role of DNA repair genetic polymorphisms.
Mutat Res
 
2007
635
(
2-3
):
118
145
72.
Tian
Y
Li
Y
Hu
Z
Wang
D
Sun
X
Ren
C
Differential effects of NOD2 polymorphisms on colorectal cancer risk: a meta-analysis.
Int J Colorectal Dis
 
2010
25
(
2
):
161
168
73.
Tan
XL
Nieters
A
Kropp
S
Hoffmeister
M
Brenner
H
Chang-Claude
J
The association of cyclin D1 G870A and E-cadherin C-160A polymorphisms with the risk of colorectal cancer in a case control study and meta-analysis.
Int J Cancer
 
2008
122
(
11
):
2573
2580
74.
Lewis
RC
Bostick
RM
Xie
D
et al
Polymorphism of the cyclin D1 gene, CCND1, and risk for incident sporadic colorectal adenomas.
Cancer Res
 
2003
63
(
23
):
8549
8553
75.
Schernhammer
ES
Tranah
GJ
Giovannucci
E
et al
Cyclin D1 A870G polymorphism and the risk of colorectal cancer and adenoma.
Br J Cancer
 
2006
94
(
6
):
928
934
76.
Fang
F
Yu
L
Zhong
Y
Yao
L
TGFB1 509 C/T polymorphism and colorectal cancer risk: a meta-analysis.
Med Oncol
 
2010
27
(
4
):
1324
1328
77.
Kaklamani
VG
Hou
N
Bian
Y
et al
TGFBR1*6A and cancer risk: a meta-analysis of seven case-control studies.
J Clin Oncol
 
2003
21
(
17
):
3236
3243
78.
Skoglund
J
Song
B
Dalén
J
et al
Lack of an association between the TGFBR1*6A variant and colorectal cancer risk.
Clin Cancer Res
 
2007
13
(
12
):
3748
3752
79.
Liao
RY
Mao
C
Qiu
LX
Ding
H
Chen
Q
Pan
HF
TGFBR1*6A/9A polymorphism and cancer risk: a meta-analysis of 13,662 cases and 14,147 controls.
Mol Biol Rep
 
2010
37
(
7
):
3227
3232
80.
Ulrich
CM
Bigler
J
Bostick
R
Fosdick
L
Potter
JD
Thymidylate synthase promoter polymorphism, interaction with folate intake, and risk of colorectal adenomas.
Cancer Res
 
2002
62
(
12
):
3361
3364
81.
Hubner
RA
Liu
JF
Sellick
GS
Logan
RF
Houlston
RS
Muir
KR
Thymidylate synthase polymorphisms, folate and B-vitamin intake, and risk of colorectal adenoma.
Br J Cancer
 
2007
97
(
10
):
1449
1456
82.
Hubner
RA
Muir
KR
Liu
JF
et al
Folate metabolism polymorphisms influence risk of colorectal adenoma recurrence.
Cancer Epidemiol Biomarkers Prev
 
2006
15
(
9
):
1607
1613
83.
Little
J
Sharp
L
Masson
LF
et al
Colorectal cancer and genetic polymorphisms of CYP1A1, GSTM1 and GSTT1: a case-control study in the Grampian region of Scotland.
Int J Cancer
 
2006
119
(
9
):
2155
2164
84.
McKay
JA
Murray
GI
Weaver
RJ
Ewen
SW
Melvin
WT
Burke
MD
Xenobiotic metabolising enzyme expression in colonic neoplasia.
Gut
 
1993
34
(
9
):
1234
1239
85.
Mercurio
MG
Shiff
SJ
Galbraith
RA
Sassa
S
Expression of cytochrome P450 mRNAs in the colon and the rectum in normal human subjects.
Biochem Biophys Res Commun
 
1995
210
(
2
):
350
355
86.
Lampen
A
Bader
A
Bestmann
T
Winkler
M
Witte
L
Borlak
JT
Catalytic activities, protein- and mRNA-expression of cytochrome P450 isoenzymes in intestinal cell lines.
Xenobiotica
 
1998
28
(
5
):
429
441
87.
Jin
JQ
Hu
YY
Niu
YM
et al
CYP1A1 Ile462Val polymorphism contributes to colorectal cancer risk: a meta-analysis.
World J Gastroenterol
 
2011
17
(
2
):
260
266
88.
Chen
K
Jiang
QT
He
HQ
Relationship between metabolic enzyme polymorphism and colorectal cancer.
World J Gastroenterol
 
2005
11
(
3
):
331
335
89.
Smits
KM
Gaspari
L
Weijenberg
MP
et al
Interaction between smoking, GSTM1 deletion and colorectal cancer: results from the GSEC study.
Biomarkers
 
2003
8
(
3-4
):
299
310
90.
Ye
Z
Parry
JM
A meta-analysis of 20 case-control studies of the glutathione S-transferase M1 (GSTM1) status and colorectal cancer risk.
Med Sci Monit
 
2003
9
(
10
):
SR83
SR91
91.
He
HQ
Chen
K
Zhang
Y
Tong
F
Fan
CH
Song
L
Glutathione S-transferase M1 polymorphism and the risk on colorectal cancer: a multilevel meta regression.
Zhonghua Liu Xing Bing Xue Za Zhi
 
2005
26
(
12
):
992
994
92.
Economopoulos
KP
Sergentanis
TN
GSTM1, GSTT1, GSTP1, GSTA1 and colorectal cancer risk: a comprehensive meta-analysis.
Eur J Cancer
 
2010
46
(
9
):
1617
1631
93.
Gao
Y
Cao
Y
Tan
A
Liao
C
Mo
Z
Gao
F
Glutathione S-transferase M1 polymorphism and sporadic colorectal cancer risk: An updating meta-analysis and HuGE review of 36 case-control studies.
Ann Epidemiol
 
2010
20
(
2
):
108
121
94.
Raimondi
S
Botteri
E
Iodice
S
Lowenfels
AB
Maisonneuve
P
Gene-smoking interaction on colorectal adenoma and cancer risk: review and meta-analysis.
Mutat Res
 
2009
670
(
1-2
):
6
14
95.
Liao
C
Cao
Y
Wu
L
Huang
J
Gao
F
An updating meta-analysis of the glutathione S-transferase T1 polymorphisms and colorectal cancer risk: a HuGE review.
Int J Colorectal Dis
 
2010
25
(
1
):
25
37
96.
Wan
H
Zhou
Y
Yang
P
Chen
B
Jia
G
Wu
X
Genetic polymorphism of glutathione S-transferase T1 and the risk of colorectal cancer: a meta-analysis.
Cancer Epidemiol
 
2010
34
(
1
):
66
72
97.
Ingles
SA
Wang
J
Coetzee
GA
Lee
ER
Frankl
HD
Haile
RW
Vitamin D receptor polymorphisms and risk of colorectal adenomas (United States).
Cancer Causes Control
 
2001
12
(
7
):
607
614
98.
Kim
HS
Newcomb
PA
Ulrich
CM
et al
Vitamin D receptor polymorphism and the risk of colorectal adenomas: evidence of interaction with dietary vitamin D and calcium.
Cancer Epidemiol Biomarkers Prev
 
2001
10
(
8
):
869
874
99.
Boyapati
SM
Bostick
RM
McGlynn
KA
et al
Calcium, vitamin D, and risk for colorectal adenoma: dependency on vitamin D receptor BsmI polymorphism and nonsteroidal anti-inflammatory drug use?
Cancer Epidemiol Biomarkers Prev
 
2003
12
(
7
):
631
637
100.
Hubner
RA
Muir
KR
Liu
JF
Logan
RF
Grainge
MJ
Houlston
RS
Members of UKCAP Consortium
Dairy products, polymorphisms in the vitamin D receptor gene and colorectal adenoma recurrence.
Int J Cancer
 
2008
123
(
3
):
586
593
101.
Egan
JB
Thompson
PA
Ashbeck
EL
et al
Genetic polymorphisms in vitamin D receptor VDR/RXRA influence the likelihood of colon adenoma recurrence.
Cancer Res
 
2010
70
(
4
):
1496
1504
102.
Kocarnik
JD
Hutter
CM
Slattery
ML
et al
Characterization of 9p24 risk locus and colorectal adenoma and cancer: gene-environment interaction and meta-analysis.
Cancer Epidemiol Biomarkers Prev
 
2010
19
(
12
):
3131
3139
103.
Slattery
ML
Samowitz
W
Ballard
L
Schaffer
D
Leppert
M
Potter
JD
A molecular variant of the APC gene at codon 1822: its association with diet, lifestyle, and risk of colon cancer.
Cancer Res
 
2001
61
(
3
):
1000
1004
104.
Tranah
GJ
Giovannucci
E
Ma
J
Fuchs
C
Hunter
DJ
APC Asp1822Val and Gly2502Ser polymorphisms and risk of colorectal cancer and adenoma.
Cancer Epidemiol Biomarkers Prev
 
2005
14
(
4
):
863
870
105.
Wong
HL
Peters
U
Hayes
RB
et al
Polymorphisms in the adenomatous polyposis coli (APC) gene and advanced colorectal adenoma risk.
Eur J Cancer
 
2010
46
(
13
):
2457
2466
106.
Xu
W
Li
Y
Wang
X
et al
PPARgamma polymorphisms and cancer risk: a meta-analysis involving 32,138 subjects.
Oncol Rep
 
2010
24
(
2
):
579
585
107.
Lu
YL
Li
GL
Huang
HL
Zhong
J
Dai
LC
Peroxisome proliferator-activated receptor-gamma 34C>G polymorphism and colorectal cancer risk: a meta-analysis.
World J Gastroenterol
 
2010
16
(
17
):
2170
2175
108.
Brockton
N
Little
J
Sharp
L
Cotton
SC
N-acetyltransferase polymorphisms and colorectal cancer: a HuGE review.
Am J Epidemiol
 
2000
151
(
9
):
846
861
109.
Ye
Z
Parry
JM
Meta-analysis of 20 case-control studies on the N-acetyltransferase 2 acetylation status and colorectal cancer risk.
Med Sci Monit
 
2002
8
(
8
):
CR558
CR565
110.
Yu
K
Zhang
J
Zhang
J
et al
Methionine synthase A2756G polymorphism and cancer risk: a meta-analysis.
Eur J Hum Genet
 
2010
18
(
3
):
370
378
111.
Gao
Y
Pan
X
Su
T
Mo
Z
Cao
Y
Gao
F
Glutathione S-transferase P1 Ile105Val polymorphism and colorectal cancer risk: a meta-analysis and HuGE review.
Eur J Cancer
 
2009
45
(
18
):
3303
3314
112.
Moore
LE
Huang
WY
Chatterjee
N
et al
GSTM1, GSTT1, and GSTP1 polymorphisms and risk of advanced colorectal adenoma.
Cancer Epidemiol Biomarkers Prev
 
2005
14
(
7
):
1823
1827
113.
Tijhuis
MJ
Wark
PA
Aarts
JM
et al
GSTP1 and GSTA1 polymorphisms interact with cruciferous vegetable intake in colorectal adenoma risk.
Cancer Epidemiol Biomarkers Prev
 
2005
14
(
12
):
2943
2951
114.
Skjelbred
CF
Saebø
M
Hjartåker
A
et al
Meat, vegetables and genetic polymorphisms and the risk of colorectal carcinomas and adenomas.
BMC Cancer
 
2007
7
228
115.
Northwood
EL
Elliott
F
Forman
D
et al
Polymorphisms in xenobiotic metabolizing enzymes and diet influence colorectal adenoma risk.
Pharmacogenet Genomics
 
2010
20
(
5
):
315
326
116.
Chen
X
Guan
J
Song
Y
et al
IGF-I (CA) repeat polymorphisms and risk of cancer: a meta-analysis.
J Hum Genet
 
2008
53
(
3
):
227
238
117.
Chen
W
Wang
S
Tian
T
et al
Phenotypes and genotypes of insulin-like growth factor 1, IGF-binding protein-3 and cancer risk: evidence from 96 studies.
Eur J Hum Genet
 
2009
17
(
12
):
1668
1675
118.
Dahabreh
IJ
Linardou
H
Bouzika
P
Varvarigou
V
Murray
S
TP53 Arg72Pro polymorphism and colorectal cancer risk: a systematic review and meta-analysis.
Cancer Epidemiol Biomarkers Prev
 
2010
19
(
7
):
1840
1847
119.
Economopoulos
KP
Sergentanis
TN
Zagouri
F
Zografos
GC
Association between p53 Arg72Pro polymorphism and colorectal cancer risk: a meta-analysis.
Onkologie
 
2010
33
(
12
):
666
674
120.
Tang
NP
Wu
YM
Wang
B
Ma
J
Systematic review and meta-analysis of the association between P53 codon 72 polymorphism and colorectal cancer.
Eur J Surg Oncol
 
2010
36
(
5
):
431
438
121.
Wang
JJ
Zheng
Y
Sun
L
et al
TP53 codon 72 polymorphism and colorectal cancer susceptibility: a meta-analysis.
Mol Biol Rep
 
2011
38
(
8
):
4847
4853
122.
Rozek
LS
Rennert
G
Gruber
SB
APC E1317Q is not associated with Colorectal Cancer in a population-based case-control study in Northern Israel.
Cancer Epidemiol Biomarkers Prev
 
2006
15
(
11
):
2325
2327
123.
Cao
H
Xu
Z
Long
H
Li
XQ
Li
SL
The -765C allele of the cyclooxygenase-2 gene as a potential risk factor of colorectal cancer: a meta-analysis.
Tohoku J Exp Med
 
2010
222
(
1
):
15
21
124.
Gunter
MJ
Canzian
F
Landi
S
Chanock
SJ
Sinha
R
Rothman
N
Inflammation-related gene polymorphisms and colorectal adenoma.
Cancer Epidemiol Biomarkers Prev
 
2006
15
(
6
):
1126
1131
125.
Siezen
CL
Bueno-de-Mesquita
HB
Peeters
PH
Kram
NR
van Doeselaar
M
van Kranen
HJ
Polymorphisms in the genes involved in the arachidonic acid-pathway, fish consumption and the risk of colorectal cancer.
Int J Cancer
 
2006
119
(
2
):
297
303
126.
Ueda
N
Maehara
Y
Tajima
O
Tabata
S
Wakabayashi
K
Kono
S
Genetic polymorphisms of cyclooxygenase-2 and colorectal adenoma risk: the Self Defense Forces Health Study.
Cancer Sci
 
2008
99
(
3
):
576
581
127.
Gong
Z
Bostick
RM
Xie
D
et al
Genetic polymorphisms in the cyclooxygenase-1 and cyclooxygenase-2 genes and risk of colorectal adenoma.
Int J Colorectal Dis
 
2009
24
(
6
):
647
654
128.
Hubner
RA
Muir
KR
Liu
JF
Logan
RF
Grainge
MJ
Houlston
RS
Members of the UKCAP Consortium
Polymorphisms in PTGS1, PTGS2 and IL-10 do not influence colorectal adenoma recurrence in the context of a randomized aspirin intervention trial.
Int J Cancer
 
2007
121
(
9
):
2001
2004
129.
Barry
EL
Sansbury
LB
Grau
MV
et al
Cyclooxygenase-2 polymorphisms, aspirin treatment, and risk for colorectal adenoma recurrence–data from a randomized clinical trial.
Cancer Epidemiol Biomarkers Prev
 
2009
18
(
10
):
2726
2733
130.
Chao
C
Zhang
ZF
Berthiller
J
Boffetta
P
Hashibe
M
NAD(P)H:quinone oxidoreductase 1 (NQO1) Pro187Ser polymorphism and the risk of lung, bladder, and colorectal cancers: a meta-analysis.
Cancer Epidemiol Biomarkers Prev
 
2006
15
(
5
):
979
987
131.
Liu
L
Liu
L
Zeng
F
et al
Meta-analysis of the association between VEGF-634 G>C and risk of malignancy based on 23 case-control studies.
J Cancer Res Clin Oncol
 
2011
137
(
6
):
1027
1036
132.
Fan
W
Maoqing
W
Wangyang
C
et al
Relationship between the polymorphism of tumor necrosis factor-a-308 G>A and susceptibility to inflammatory bowel diseases and colorectal cancer: a meta-analysis.
Eur J Hum Genet
 
2011
19
(
4
):
432
437
133.
Zhou
GW
Hu
J
Li
Q
CYP2E1 PstI/RsaI polymorphism and colorectal cancer risk: a meta-analysis.
World J Gastroenterol
 
2010
16
(
23
):
2949
2953
134.
Hawken
SJ
Greenwood
CM
Hudson
TJ
et al
The utility and predictive value of combinations of low penetrance genes for screening and risk prediction of colorectal cancer.
Hum Genet
 
2010
128
(
1
):
89
101
135.
Frazer
KA
Murray
SS
Schork
NJ
Topol
EJ
Human genetic variation and its contribution to complex traits.
Nat Rev Genet
 
2009
10
(
4
):
241
251
136.
Butler
WJ
Ryan
P
Roberts-Thomson
IC
Metabolic genotypes and risk for colorectal cancer increased risk in individuals with glutathione transferase theta 1 (GSTT1) gene defect.
Gastroenterology.
 
1997
112
A542
137.
Zhong
Y
Huang
Y
Huang
Y
et al
Effects of O6-methylguanine-DNA methyltransferase (MGMT) polymorphisms on cancer: a meta-analysis.
Mutagenesis
 
2010
25
(
1
):
83
95
138.
Tenesa
A
Campbell
H
Barnetson
R
Porteous
M
Dunlop
M
Farrington
SM
Association of MUTYH and colorectal cancer.
Br J Cancer
 
2006
95
(
2
):
239
242
139.
Lubbe
SJ
Di Bernardo
MC
Chandler
IP
Houlston
RS
Clinical implications of the colorectal cancer risk associated with MUTYH mutation.
J Clin Oncol
 
2009
27
(
24
):
3975
3980
140.
Avezzù
A
Agostini
M
Pucciarelli
S
et al
The role of MYH gene in genetic predisposition to colorectal cancer: another piece of the puzzle.
Cancer Lett
 
2008
268
(
2
):
308
313
141.
Balaguer
F
Castellví-Bel
S
Castells
A
et al
Identification of MYH mutation carriers in colorectal cancer: a multicenter, case-control, population-based study.
Clin Gastroenterol Hepatol
 
2007
5
(
3
):
379
387
142.
Pereira
C
Medeiros
RM
Dinis-Ribeiro
MJ
Cyclooxygenase polymorphisms in gastric and colorectal carcinogenesis: are conclusive results available?
Eur J Gastroenterol Hepatol
 
2009
21
(
1
):
76
91
143.
Ewart-Toland
A
Dai
Q
Gao
YT
et al
Aurora-A/STK15 T+91A is a general low penetrance cancer susceptibility gene: a meta-analysis of multiple cancer types.
Carcinogenesis
 
2005
26
(
8
):
1368
1373
144.
Hu
Z
Li
X
Qu
X
et al
Intron 3 16bp duplication polymorphism of TP53 contributes to cancer susceptibility: a meta-analysis.
Carcinogenesis
 
2010
31
(
4
):
643
647
145.
Fang
F
Yu
XJ
Yu
L
Yao
L
MDM2 309 T/G polymorphism is associated with colorectal cancer risk especially in Asians: a meta-analysis.
Med Oncol
 
2011
28
(
4
):
981
985
146.
Wilkening
S
Bermejo
JL
Hemminki
K
MDM2 SNP309 and cancer risk: a combined analysis.
Carcinogenesis
 
2007
28
(
11
):
2262
2267
147.
Theodoratou
E
Farrington
SM
Tenesa
A
et al
Modification of the inverse association between dietary vitamin D intake and colorectal cancer risk by a FokI variant supports a chemoprotective action of Vitamin D intake mediated through VDR binding.
Int J Cancer
 
2008
123
(
9
):
2170
2179
148.
Hutter
CM
Slattery
ML
Duggan
DJ
et al
Characterization of the association between 8q24 and colon cancer: gene-environment exploration and meta-analysis.
BMC Cancer
 
2010
10
670

Notes

JL holds a Tier 1 Canada Research Chair in Human Genome Epidemiology.

The study sponsors had no role in the design of the study; the collection, analysis, or interpretation of the data; the writing of the manuscript; or the decision to submit the manuscript for publication.

JL and HC conceived the study; JL, HC, ET, ZM, and SH designed it; ET, ZM, HC, and JL wrote the article with input from other authors; ET and ZM undertook data manipulations and statistical analysis; GCdLA and JG undertook the literature review; ET, ZM, SH, GCdLA, JG, VT, IK, MT, AD, LZ, DL, HEB, and SHR coordinated and/or undertook related abstraction, handling, and curation of the data; SMF, AT, and ZM provided the GWAS data for Scotland and Canada; SMF, IR, AT, and MGD provided consultation in their areas of expertise. JL, HC, MD, and ET obtained funding for the study.

The team in Edinburgh would like to thank Mrs Gisela Johnstone, Stephanie Scott, and Rosa Bisset for their administrative support and Mr Colin Pride for his IT support. The team in Ottawa would like to thank Dr Philip Ryan (University of Adelaide) for providing information that clarified the data published by Butler et al. ( 136 ).