Impact of ambient air pollution on colorectal cancer risk and survival: insights from a prospective cohort and epigenetic Mendelian randomization study.

BACKGROUND: This study investigates the associations between air pollution and colorectal cancer (CRC) risk and survival from an epigenomic perspective. METHODS: Using a newly developed Air Pollutants Exposure Score (APES), we utilized a prospective cohort study (UK Biobank) to investigate the associations of individual and combined air pollution exposures with CRC incidence and survival, followed by an up-to-date systematic review with meta-analysis to verify the associations. In epigenetic two-sample Mendelian randomization analyses, we examine the associations between genetically predicted DNA methylation related to air pollution and CRC risk. Further genetic colocalization and gene-environment interaction analyses provided different insights to disentangle pathogenic effects of air pollution via epigenetic modification. FINDINGS: During a median 12.97-year follow-up, 5767 incident CRC cases among 428,632 participants free of baseline CRC and 533 deaths in 2401 patients with CRC were documented in the UK Biobank. A higher APES score was associated with an increased CRC risk (HR, 1.03, 95% CI = 1.01-1.06; P = 0.016) and poorer survival (HR, 1.13, 95% CI = 1.03-1.23; P = 0.010), particularly among participants with insufficient physical activity and ever smokers (Pinteraction > 0.05). A subsequent meta-analysis of seven observational studies, including UK Biobank data, corroborated the association between PM2.5 exposure (per 10 µg/m3 increment) and elevated CRC risk (RR,1.42, 95% CI = 1.12-1.79; P = 0.004; I2 = 90.8%). Genetically predicted methylation at PM2.5-related CpG site cg13835894 near TMBIM1/PNKD and cg16235962 near CXCR5, and NO2-related cg16947394 near TMEM110 were associated with an increased CRC risk. Gene-environment interaction analysis confirmed the epigenetic modification of aforementioned CpG sites with CRC risk and survival. INTERPRETATION: Our study suggests the association between air pollution and CRC incidence and survival, underscoring the possible modifying roles of epigenomic factors. Methylation may partly mediate pathogenic effects of air pollution on CRC, with annotation to epigenetic alterations in protein-coding genes TMBIM1/PNKD, CXCR5 and TMEM110. FUNDING: Xue Li is supported by the Natural Science Fund for Distinguished Young Scholars of Zhejiang Province (LR22H260001), the National Nature Science Foundation of China (No. 82204019) and Healthy Zhejiang One Million People Cohort (K-20230085). ET is supported by a Cancer Research UK Career Development Fellowship (C31250/A22804). MGD is supported by the MRC Human Genetics Unit Centre Grant (U127527198).

Systematic investigation of genetically determined plasma and urinary metabolites to discover potential interventional targets for colorectal cancer.

BACKGROUND: We aimed to identify plasma and urinary metabolites related to colorectal cancer (CRC) risk and elucidate their mediator role in the associations between modifiable risk factors and CRC. METHODS: Metabolite quantitative trait loci were derived from two published metabolomics genome-wide association studies (GWASs), and summary-level data were extracted for 651 plasma metabolites and 208 urinary metabolites. Genetic associations with CRC were obtained from a large-scale GWAS meta-analysis (100,204 cases; 154,587 controls) and the FinnGen cohort (4,957 cases; 304,197 controls). Mendelian randomization (MR) and colocalization analyses were performed to evaluate the causal roles of metabolites in CRC. Druggability evaluation was employed to prioritize potential therapeutic targets. Multivariable MR and mediation estimation were conducted to elucidate the mediating effects of metabolites on the associations between modifiable risk factors and CRC. RESULTS: The study identified 30 plasma metabolites and four urinary metabolites for CRC. Plasma sphingomyelin and urinary lactose, which were positively associated with CRC risk, could be modulated by drug interventions (ie, Olipudase alfa, Tilactase). Thirteen modifiable risk factors were associated with nine metabolites and eight of these modifiable risk factors were associated with CRC risk. These nine metabolites mediated the effect of modifiable risk factors (Actinobacteria, BMI, waist-hip ratio, fasting insulin, smoking initiation) on CRC. CONCLUSION: This study identified key metabolite biomarkers associated with CRC and elucidated their mediator roles in the associations between modifiable risk factors and CRC. These findings provide new insights into the etiology and potential therapeutic targets for CRC and the etiological pathways of modifiable environmental factors with CRC.

Fine-mapping analysis including over 254,000 East Asian and European descendants identifies 136 putative colorectal cancer susceptibility genes.

Genome-wide association studies (GWAS) have identified more than 200 common genetic variants independently associated with colorectal cancer (CRC) risk, but the causal variants and target genes are mostly unknown. We sought to fine-map all known CRC risk loci using GWAS data from 100,204 cases and 154,587 controls of East Asian and European ancestry. Our stepwise conditional analyses revealed 238 independent association signals of CRC risk, each with a set of credible causal variants (CCVs), of which 28 signals had a single CCV. Our cis-eQTL/mQTL and colocalization analyses using colorectal tissue-specific transcriptome and methylome data separately from 1299 and 321 individuals, along with functional genomic investigation, uncovered 136 putative CRC susceptibility genes, including 56 genes not previously reported. Analyses of single-cell RNA-seq data from colorectal tissues revealed 17 putative CRC susceptibility genes with distinct expression patterns in specific cell types. Analyses of whole exome sequencing data provided additional support for several target genes identified in this study as CRC susceptibility genes. Enrichment analyses of the 136 genes uncover pathways not previously linked to CRC risk. Our study substantially expanded association signals for CRC and provided additional insight into the biological mechanisms underlying CRC development.

Combining Asian and European genome-wide association studies of colorectal cancer improves risk prediction across racial and ethnic populations.

Polygenic risk scores (PRS) have great potential to guide precision colorectal cancer (CRC) prevention by identifying those at higher risk to undertake targeted screening. However, current PRS using European ancestry data have sub-optimal performance in non-European ancestry populations, limiting their utility among these populations. Towards addressing this deficiency, we expand PRS development for CRC by incorporating Asian ancestry data (21,731 cases; 47,444 controls) into European ancestry training datasets (78,473 cases; 107,143 controls). The AUC estimates (95% CI) of PRS are 0.63(0.62-0.64), 0.59(0.57-0.61), 0.62(0.60-0.63), and 0.65(0.63-0.66) in independent datasets including 1681-3651 cases and 8696-115,105 controls of Asian, Black/African American, Latinx/Hispanic, and non-Hispanic White, respectively. They are significantly better than the European-centric PRS in all four major US racial and ethnic groups (p-values < 0.05). Further inclusion of non-European ancestry populations, especially Black/African American and Latinx/Hispanic, is needed to improve the risk prediction and enhance equity in applying PRS in clinical practice.

Dissecting the pathogenic effects of smoking and its hallmarks in blood DNA methylation on colorectal cancer risk.

BACKGROUND: Tobacco smoking is suggested as a risk factor for colorectal cancer (CRC), but the complex relationship and the potential pathway are not fully understood. METHODS: We performed two-sample Mendelian randomisation (MR) analyses with genetic instruments for smoking behaviours and related DNA methylation in blood and summary-level GWAS data of colorectal cancer to disentangle the relationship. Colocalization analyses and prospective gene-environment interaction analyses were also conducted as replication. RESULTS: Convincing evidence was identified for the pathogenic effect of smoking initiation on CRC risk and suggestive evidence was observed for the protective effect of smoking cessation in the univariable MR analyses. Multivariable MR analysis revealed that these associations were independent of other smoking phenotypes and alcohol drinking. Genetically predicted methylation at CpG site cg17823346 [ZMIZ1] were identified to decrease CRC risk; while genetically predicted methylation at cg02149899 would increase CRC risk. Colocalization and gene-environment interaction analyses added further evidence to the relationship between epigenetic modification at cg17823346 [ZMIZ1] as well as cg02149899 and CRC risk. DISCUSSION: Our study confirms the significant association between tobacco smoking, DNA methylation and CRC risk and yields a novel insight into the pathogenic effect of tobacco smoking on CRC risk.

Combining Asian-European Genome-Wide Association Studies of Colorectal Cancer Improves Risk Prediction Across Race and Ethnicity.

Polygenic risk scores (PRS) have great potential to guide precision colorectal cancer (CRC) prevention by identifying those at higher risk to undertake targeted screening. However, current PRS using European ancestry data have sub-optimal performance in non-European ancestry populations, limiting their utility among these populations. Towards addressing this deficiency, we expanded PRS development for CRC by incorporating Asian ancestry data (21,731 cases; 47,444 controls) into European ancestry training datasets (78,473 cases; 107,143 controls). The AUC estimates (95% CI) of PRS were 0.63(0.62-0.64), 0.59(0.57-0.61), 0.62(0.60-0.63), and 0.65(0.63-0.66) in independent datasets including 1,681-3,651 cases and 8,696-115,105 controls of Asian, Black/African American, Latinx/Hispanic, and non-Hispanic White, respectively. They were significantly better than the European-centric PRS in all four major US racial and ethnic groups (p-values<0.05). Further inclusion of non-European ancestry populations, especially Black/African American and Latinx/Hispanic, is needed to improve the risk prediction and enhance equity in applying PRS in clinical practice.

Exploiting gene dependency to inform drug development for multiple myeloma.

Despite recent advances in therapy, multiple myeloma essentially remains an incurable malignancy. Targeting tumour-specific essential genes, which constitute a druggable dependency, potentially offers a strategy for developing new therapeutic agents to treat MM and overcome drug resistance. To explore this possibility, we analysed DepMap project data identifying 23 MM essential genes and examined the relationship between their expression and patient outcome in three independent series totalling 1503 cases. The expression of TCF3 and FLVCR1 were both significantly associated with progression-free survival. IKBKB is already a drug target in other diseases, offering the prospect of repurposing to treat MM, while PIM2 is currently being investigated as a treatment for the disease. Our analysis supports the rationale of using large-scale genetic perturbation screens to guide the development of new therapeutic agents for MM.

Alcohol consumption, DNA methylation and colorectal cancer risk: Results from pooled cohort studies and Mendelian randomization analysis.

Alcohol consumption is thought to be one of the modifiable risk factors for colorectal cancer (CRC). However, the causality and mechanisms by which alcohol exerts its carcinogenic effect are unclear. We evaluated the association between alcohol consumption and CRC risk by analyzing data from 32 cohort studies and conducted two-sample Mendelian randomization (MR) analysis to examine for casual relationship. To explore the effect of alcohol related DNA methylation on CRC risk, we performed an epigenetic MR analysis with data from an epigenome-wide association study (EWAS). We additionally performed gene-alcohol interaction analysis nested in the UK Biobank to assess effect modification between alcohol consumption and susceptibility genes. We discovered distinct effects of alcohol on CRC incidence and mortality from the meta-analyses, and genetic predisposition to alcohol drinking was causally associated with an increased CRC risk (OR = 1.79, 95% CI: 1.23-2.61) using two-sample MR approaches. In epigenetic MR analysis, two alcohol-related CpG sites (cg05593667 and cg10045354 mapped to COLCA1/COLCA2 gene) were identified causally associated with an increased CRC risk (P < 8.20 × 10-4 ). Gene-alcohol interaction analysis revealed that carriage of the risk allele of the eQTL (rs3087967) and mQTL (rs11213823) polymorphism of COLCA1/COLCA2 would interact with alcohol consumption to increase CRC risk (PInteraction  = .027 and PInteraction  = .016). Our study provides comprehensive evidence to elucidate the role of alcohol in CRC and highlights that the pathogenic effect of alcohol on CRC could be partly attributed to DNA methylation by regulating the expression of COLCA1/COLCA2 gene.

Relationship between genetically determined telomere length and glioma risk.

BACKGROUND: Telomere maintenance is increasingly recognized as being fundamental to glioma oncogenesis with longer leukocyte telomere length (LTL) reported to increase risk of glioma. To gain further insight into the relationship between telomere genetics and risk of glioma, we conducted several complementary analyses, using genome-wide association studies data on LTL (78 592 individuals) and glioma (12 488 cases and 18 169 controls). METHODS: We performed both classical and summary Mendelian randomization (SMR), coupled with heterogeneity in dependent instruments tests, at genome-wide significant LTL loci to examine if an association was mediated by the same causal variant in glioma. To prioritize genes underscoring glioma-LTL associations, we analyzed gene expression and DNA methylation data. RESULTS: Genetically increased LTL was significantly associated with increased glioma risk, random-effects inverse variance weighted ORs per 1 SD unit increase in the putative risk factor (odds ratio [OR]SD) 4.79 (95% confidence interval: 2.11-10.85; P = 1.76 × 10-4). SMR confirmed the previously reported LTL associations at 3q26.2 (TERC; PSMR = 1.33 × 10-5), 5p15.33 (TERT; PSMR = 9.80 × 10-27), 10q24.33 (STN1 alias OBFC1; PSMR = 4.31 × 10-5), and 20q13.3 (STMN3/RTEL1; PSMR = 2.47 × 10-4) glioma risk loci. Our analysis implicates variation at 1q42.12 (PSMR = 1.55 × 10-2), 6p21.3 (PSMR = 9.76 × 10-3), 6p22.2 (PSMR = 5.45 × 10-3), 7q31.33 (PSMR = 6.52 × 10-3), and 11q22.3 (PSMR = 8.89 × 10-4) as risk factors for glioma risk. While complicated by patterns of linkage disequilibrium, genetic variation involving PARP1, PRRC2A, CARMIL1, POT1, and ATM-NPAT1 was implicated in the etiology of glioma. CONCLUSIONS: These observations extend the role of telomere-related genes in the development of glioma.

Phenome-wide association study (PheWAS) of colorectal cancer risk SNP effects on health outcomes in UK Biobank.

BACKGROUND: Associations between colorectal cancer (CRC) and other health outcomes have been reported, but these may be subject to biases, or due to limitations of observational studies. METHODS: We set out to determine whether genetic predisposition to CRC is also associated with the risk of other phenotypes. Under the phenome-wide association study (PheWAS) and tree-structured phenotypic model (TreeWAS), we studied 334,385 unrelated White British individuals (excluding CRC patients) from the UK Biobank cohort. We generated a polygenic risk score (PRS) from CRC genome-wide association studies as a measure of CRC risk. We performed sensitivity analyses to test the robustness of the results and searched the Danish Disease Trajectory Browser (DTB) to replicate the observed associations. RESULTS: Eight PheWAS phenotypes and 21 TreeWAS nodes were associated with CRC genetic predisposition by PheWAS and TreeWAS, respectively. The PheWAS detected associations were from neoplasms and digestive system disease group (e.g. benign neoplasm of colon, anal and rectal polyp and diverticular disease). The results from the TreeWAS corroborated the results from the PheWAS. These results were replicated in the observational data within the DTB. CONCLUSIONS: We show that benign colorectal neoplasms share genetic aetiology with CRC using PheWAS and TreeWAS methods. Additionally, CRC genetic predisposition is associated with diverticular disease.

A genome-wide search for determinants of survival in 1926 patients with advanced colorectal cancer with follow-up in over 22,000 patients.

BACKGROUND: While genome-wide association studies (GWAS) have identified germline variants influencing the risk of developing colorectal cancer (CRC), there has been limited examination of the possible role of inherited variation as a determinant of patient outcome. PATIENTS AND METHODS: We performed a GWAS for overall survival (OS) in 1926 patients with advanced CRC from the COIN and COIN-B clinical trials. For single nucleotide polymorphisms (SNPs) showing an association with OS (P < 1.0 × 10-5), we conducted sensitivity analyses based on the time from diagnosis to death and sought independent replications in 5675 patients from the Study of Colorectal Cancer in Scotland (SOCCS) and 16,964 patients from the International Survival Analysis in Colorectal cancer Consortium (ISACC). We analysed the Human Protein Atlas to determine if ERBB4 expression was associated with survival in 438 patients with colon adenocarcinomas. RESULTS: The most significant SNP associated with OS was rs79612564 in ERBB4 (hazard ratio [HR] = 1.24, 95% confidence interval [CI] = 1.16-1.32, P = 1.9 × 10-7). SNPs at 17 loci had suggestive associations for OS and all had similar effects on the time from diagnosis to death. No lead SNPs were independently replicated in the meta-analysis of all patients from SOCCS and ISACC. However, rs79612564 was significant in stage-IV patients from SOCCS (P = 2.1 × 10-2) but not ISACC (P = 0.89) and SOCCS combined with COIN and COIN-B attained genome-wide significance (P = 1.7 × 10-8). Patients with high ERBB4 expression in their colon adenocarcinomas had worse survival (HR = 1.50, 95% CI = 1.1-1.9, P = 4.6 × 10-2). CONCLUSIONS: Genetic and expression data support a potential role for rs79612564 in the receptor tyrosine kinase ERBB4 as a predictive biomarker of survival.

Genetically predicted physical activity levels are associated with lower colorectal cancer risk: a Mendelian randomisation study.

BACKGROUND: We conducted a Mendelian randomisation (MR) study to investigate whether physical activity (PA) causes a reduction of colorectal cancer risk and to understand the contributions of effects mediated through changes in body fat. METHODS: Common genetic variants associated with self-reported moderate-to-vigorous PA (MVPA), acceleration vector magnitude PA (AMPA) and sedentary time were used as instrumental variables. To control for confounding effects of obesity, we included instrumental variables for body mass index (BMI), body fat percentage, waist circumference and arm, trunk and leg fat ratios. We analysed the effect of these instrumental variables in a colorectal cancer genome-wide association study comprising 31,197 cases and 61,770 controls of European ancestry by applying two-sample and multivariable MR study designs. RESULTS: We found decreased colorectal cancer risk for genetically represented measures of MVPA and AMPA that were additional to effects mediated through genetic measures of obesity. Odds ratio and 95% confidence interval (CI) per standard deviation increase in MVPA and AMPA was 0.56 (0.31, 1.01) and 0.60 (0.41, 0.88), respectively. No association has been found between sedentary time and colorectal cancer risk. The proportion of effect mediated through BMI was 2% (95% CI: 0, 14) and 32% (95% CI: 12, 46) for MVPA and AMPA, respectively. CONCLUSION: These findings provide strong evidence to reinforce public health measures on preventing colorectal cancer that promote PA at a population level regardless of body fatness.

Lack of an association between gallstone disease and bilirubin levels with risk of colorectal cancer: a Mendelian randomisation analysis.

BACKGROUND: Epidemiological studies of the relationship between gallstone disease and circulating levels of bilirubin with risk of developing colorectal cancer (CRC) have been inconsistent. To address possible confounding and reverse causation, we examine the relationship between these potential risk factors and CRC using Mendelian randomisation (MR). METHODS: We used two-sample MR to examine the relationship between genetic liability to gallstone disease and circulating levels of bilirubin with CRC in 26,397 patients and 41,481 controls. We calculated the odds ratio per genetically predicted SD unit increase in log bilirubin levels (ORSD) for CRC and tested for a non-zero causal effect of gallstones on CRC. Sensitivity analysis was applied to identify violations of estimator assumptions. RESULTS: No association between either gallstone disease (P value = 0.60) or circulating levels of bilirubin (ORSD = 1.00, 95% confidence interval (CI) = 0.96-1.03, P value = 0.90) with CRC was shown. CONCLUSIONS: Despite the large scale of this study, we found no evidence for a causal relationship between either circulating levels of bilirubin or gallstone disease with risk of developing CRC. While the magnitude of effect suggested by some observational studies can confidently be excluded, we cannot exclude the possibility of smaller effect sizes and non-linear relationships.

Searching for causal relationships of glioma: a phenome-wide Mendelian randomisation study.

BACKGROUND: The aetiology of glioma is poorly understood. Summary data from genome-wide association studies (GWAS) can be used in a Mendelian randomisation (MR) phenome-wide association study (PheWAS) to search for glioma risk factors. METHODS: We performed an MR-PheWAS analysing 316 phenotypes, proxied by 8387 genetic variants, and summary genetic data from a GWAS of 12,488 glioma cases and 18,169 controls. Causal effects were estimated under a random-effects inverse-variance-weighted (IVW-RE) model, with robust adjusted profile score (MR-RAPS), weighted median and mode-based estimates computed to assess the robustness of findings. Odds ratios per one standard deviation increase in each phenotype were calculated for all glioma, glioblastoma (GBM) and non-GBM tumours. RESULTS: No significant associations (P < 1.58 × 10-4) were observed between phenotypes and glioma under the IVW-RE model. Suggestive associations (1.58 × 10-4 < P < 0.05) were observed between leukocyte telomere length (LTL) with all glioma (ORSD = 3.91, P = 9.24 × 10-3) and GBM (ORSD = 4.86, P = 3.23 × 10-2), but the association was primarily driven by the TERT variant rs2736100. Serum low-density lipoprotein cholesterol and plasma HbA1C showed suggestive associations with glioma (ORSD = 1.11, P = 1.39 × 10-2 and ORSD = 1.28, P = 1.73 × 10-2, respectively), both associations being reliant on single genetic variants. CONCLUSIONS: Our study provides further insight into the aetiological basis of glioma for which published data have been mixed.

Search for multiple myeloma risk factors using Mendelian randomization.

The etiology of multiple myeloma (MM) is poorly understood. Summary data from genome-wide association studies (GWASs) of multiple phenotypes can be exploited in a Mendelian randomization (MR) phenome-wide association study (PheWAS) to search for factors influencing MM risk. We performed an MR-PheWAS analyzing 249 phenotypes, proxied by 10 225 genetic variants, and summary genetic data from a GWAS of 7717 MM cases and 29 304 controls. Odds ratios (ORs) per 1 standard deviation increase in each phenotype were estimated under an inverse variance weighted random effects model. A Bonferroni-corrected threshold of P = 2 × 10-4 was considered significant, whereas P < .05 was considered suggestive of an association. Although no significant associations with MM risk were observed among the 249 phenotypes, 28 phenotypes showed evidence suggestive of association, including increased levels of serum vitamin B6 and blood carnitine (P = 1.1 × 10-3) with greater MM risk and ω-3 fatty acids (P = 5.4 × 10-4) with reduced MM risk. A suggestive association between increased telomere length and reduced MM risk was also noted; however, this association was primarily driven by the previously identified risk variant rs10936599 at 3q26 (TERC). Although not statistically significant, increased body mass index was associated with increased risk (OR, 1.10; 95% confidence interval, 0.99-1.22), supporting findings from a previous meta-analysis of prospective observational studies. Our study did not provide evidence supporting any modifiable factors examined as having a major influence on MM risk; however, it provides insight into factors for which the evidence has previously been mixed.

Lack of association between modifiable exposures and glioma risk: a Mendelian randomization analysis.

BACKGROUND: The etiological basis of glioma is poorly understood. We have used genetic markers in a Mendelian randomization (MR) framework to examine if lifestyle, cardiometabolic, and inflammatory factors influence the risk of glioma. This methodology reduces bias from confounding and is not affected by reverse causation. METHODS: We identified genetic instruments for 37 potentially modifiable risk factors and evaluated their association with glioma risk using data from a genome-wide association study of 12 488 glioma patients and 18 169 controls. We used the estimated odds ratio of glioma associated with each of the genetically defined traits to infer evidence for a causal relationship with the following exposures:Lifestyle and dietary factors-height, plasma insulin-like growth factor 1, blood carnitine, blood methionine, blood selenium, blood zinc, circulating adiponectin, circulating carotenoids, iron status, serum calcium, vitamins (A1, B12, B6, E, and 25-hydroxyvitamin D), fatty acid levels (monounsaturated, omega-3, and omega-6) and circulating fetuin-A;Cardiometabolic factors-birth weight, high density lipoprotein cholesterol, low density lipoprotein cholesterol, total cholesterol, total triglycerides, basal metabolic rate, body fat percentage, body mass index, fasting glucose, fasting proinsulin, glycated hemoglobin levels, diastolic and systolic blood pressure, waist circumference, waist-to-hip ratio; andInflammatory factors- C-reactive protein, plasma interleukin-6 receptor subunit alpha and serum immunoglobulin E. RESULTS: After correction for the testing of multiple potential risk factors and excluding associations driven by one single nucleotide polymorphism, no significant association with glioma risk was observed (ie, PCorrected > 0.05). CONCLUSIONS: This study did not provide evidence supporting any of the 37 factors examined as having a significant influence on glioma risk.

Modifiable pathways for colorectal cancer: a mendelian randomisation analysis.

BACKGROUND: Epidemiological studies have linked lifestyle, cardiometabolic, reproductive, developmental, and inflammatory factors to the risk of colorectal cancer. However, which specific factors affect risk and the strength of these effects are unknown. We aimed to examine the relationship between potentially modifiable risk factors and colorectal cancer. METHODS: We used a random-effects model to examine the relationship between 39 potentially modifiable risk factors and colorectal cancer in 26 397 patients with colorectal cancer and 41 481 controls (ie, people without colorectal cancer). These population data came from a genome-wide association study of people of European ancestry, which was amended to exclude UK BioBank data. In the model, we used genetic variants as instruments via two-sample mendelian randomisation to limit bias from confounding and reverse causation. We calculated odds ratios per genetically predicted SD unit increase in each putative risk factor (ORSD) for colorectal cancer risk. We did mendelian randomisation Egger regressions to identify evidence of potential violations of mendelian randomisation assumptions. A Bonferroni-corrected threshold of p=1·3 × 10-3 was considered significant, and p values less than 0·05 were considered to be suggestive of an association. FINDINGS: No putative risk factors were significantly associated with colorectal cancer risk after correction for multiple testing. However, suggestive associations with increased risk were noted for genetically predicted body fat percentage (ORSD 1·14 [95% CI 1·03-1·25]; p=0·0086), body-mass index (1·09 [1·01-1·17]; p=0·023), waist circumference (1·13 [1·02-1·26]; p=0·018), basal metabolic rate (1·10 [1·03-1·18]; p=0·0079), and concentrations of LDL cholesterol (1·14 [1·04-1·25]; p=0·0056), total cholesterol (1·09 [1·01-1·18]; p=0·025), circulating serum iron (1·17 [1·00-1·36]; p=0·049), and serum vitamin B12 (1·21 [1·04-1·42]; p=0·016), although potential pleiotropy among genetic variants used as instruments for vitamin B12 constrains the finding. A suggestive association was also noted between adult height and increased risk of colorectal cancer (ORSD 1·04 [95% CI 1·00-1·08]; p=0·032). Low blood selenium concentration had a suggestive association with decreased risk of colorectal cancer (ORSD 0·85 [95% CI 0·75-0·96]; p=0·0078) based on a single variant, as did plasma concentrations of interleukin-6 receptor subunit α (also based on a single variant; 0·98 [0·96-1·00]; p=0·035). Risk of colorectal cancer was not associated with any sex hormone or reproductive factor, serum calcium, or circulating 25-hydroxyvitamin D concentrations. INTERPRETATION: This analysis identified several modifiable targets for primary prevention of colorectal cancer, including lifestyle, obesity, and cardiometabolic factors, that should inform public health policy. FUNDING: Cancer Research UK, UK Medical Research Council Human Genetics Unit Centre, DJ Fielding Medical Research Trust, EU COST Action, and the US National Cancer Institute.

Comprehensive analysis of colorectal cancer-risk loci and survival outcome: A prognostic role for CDH1 variants.

PURPOSE: Genome-wide association studies have identified common single nucleotide polymorphisms (SNPs) at 83 loci associated with colorectal cancer (CRC) risk in European populations. Because germline variation can also influence patient outcome, we studied the relationship between these SNPs and CRC survivorship. EXPERIMENTAL DESIGN: For the 83 risk loci, 10 lead SNPs were directly genotyped, 72 were imputed and 1 was not genotyped nor imputed, in 1948 unrelated patients with advanced CRC from the clinical trials COIN and COIN-B (oxaliplatin and fluoropyrimidine chemotherapy ± cetuximab). A Cox survival model was used for each variant, and variants classified by pathway, adjusting for known prognostic factors. We imposed a Bonferroni threshold of P = 6.6 × 10-4 for multiple testing. We carried out meta-analyses of published risk SNPs associated with survival. RESULTS: Univariate analysis identified six SNPs associated with overall survival (OS) (P < 0.05); however, only rs9939049 in CDH1 remained significant beyond the Bonferroni threshold (Hazard Ratio [HR] 1.44, 95% Confidence Intervals [CI]: 1.21-1.71, P = 5.0 × 10-5). Fine mapping showed that rs12597188 was the most significant SNP at this locus and remained significant after adjustment for known prognostic factors beyond multiple testing thresholds (HR 1.23, 95% CI: 1.13-1.34, P = 1.9 × 10-6). rs12597188 was also associated with poor response to therapy (OR 0.61, 95% CI: 0.42-0.87, P = 6.6 × 10-3). No combinations of SNPs within pathways were more significantly associated with survival compared with single variants alone, and no other risk SNPs were associated with survival in meta-analyses. CONCLUSIONS: The CRC susceptibility SNP rs9939049 in CDH1 influences patient survival and warrants further evaluation as a prognostic biomarker.