Next Article in Journal
Delivering a Quantitative Imaging Agenda
Next Article in Special Issue
The Impact of Neurophysiological Monitoring during Intradural Spinal Tumor Surgery
Previous Article in Journal
Locoregional Treatment in Intrahepatic Cholangiocarcinoma: Which Treatment for Which Patient?
Previous Article in Special Issue
Fat-Containing Soft Tissue Tumors in Children, Adolescents, and Young Adults: Which Require Biopsy?
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Cancers
Volume 15
Issue 17
10.3390/cancers15174218
cancers-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Case-Only Genome-Wide Interaction Study of Smoking and Bladder Cancer Risk: Results from the COBLAnCE Cohort

by
Maryam Karimi
1,2,*,
Sebastian Mendez-Pineda
1,
Hélène Blanché
3,
Anne Boland
4,
Céline Besse
4,
Jean-François Deleuze
3,4,
Xiang-Yu Meng
5,
Nanor Sirab
6,
Karine Groussard
1,
Thierry Lebret
7,
Julia Bonastre
1,2,
Yves Allory
5,8,
François Radvanyi
5,
Simone Benhamou
1,† and
Stefan Michiels
1,2,†
1
Oncostat U1018 Inserm, Équipe Labellisée Ligue Contre le Cancer, Université Paris-Saclay, 94805 Villejuif, France
2
Bureau de Biostatistique et d’Épidémiologie, Gustave Roussy, Université Paris-Saclay, 94805 Villejuif, France
3
CEPH-Biobank, Fondation Jean Dausset-CEPH, 75010 Paris, France
4
Centre National de Recherche en Génomique Humaine (CNRGH), CEA, Université Paris-Saclay, 91057 Evry, France
5
CNRS UMR144, Curie Institute, 75005 Paris, France
6
Curie Institute, CNRS, UMR144, Molecular Oncology Team, PSL Research University, 75005 Paris, France
7
Urology, Foch Hospital, 92150 Suresnes, France
8
UVSQ, Curie Institute, Department of Pathology, Université Paris-Saclay, 92210 Saint-Cloud, France
*
Author to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Cancers 2023, 15(17), 4218; https://doi.org/10.3390/cancers15174218
Submission received: 6 July 2023 / Revised: 2 August 2023 / Accepted: 7 August 2023 / Published: 23 August 2023
(This article belongs to the Special Issue Oncology: State-of-the-Art Research in France)
Browse Figures
Versions Notes

Abstract

:

Simple Summary

Bladder cancer is one of the most common cancers worldwide. The most important known cause of bladder cancer is tobacco smoking. Several genetic variations are also linked to the risk of developing bladder cancer. In this study, we focused on how smoking behavior interacts with genes to affect bladder cancer risk using the French national prospective COBLAnCE cohort (COhort to study BLAdder CancEr). The COBLAnCE cohort comprises approximately 1700 bladder cancer patients with available smoking and genetic information. We identified new chromosomal regions (specifically, genes located at 4q22.1, 12p13.1, and 16p13.3) of which variations may interact with smoking behavior in relation to bladder cancer risk. These findings need to be replicated in other studies.

Abstract

Bladder cancer (BC) is the 6th most common cancer worldwide, with tobacco smoking considered as its main risk factor. Accumulating evidence has found associations between genetic variants and the risk of BC. Candidate gene-environment interaction studies have suggested interactions between cigarette smoking and NAT2/GSTM1 gene variants. Our objective was to perform a genome-wide association case-only study using the French national prospective COBLAnCE cohort (COhort to study BLAdder CancEr), focusing on smoking behavior. The COBLAnCE cohort comprises 1800 BC patients enrolled between 2012 and 2018. Peripheral blood samples collected at enrolment were genotyped using the Illumina Global Screening Array with a Multi-Disease drop-in panel. Genotyping data (9,719,614 single nucleotide polymorphisms (SNP)) of 1674, 1283, and 1342 patients were analyzed for smoking status, average tobacco consumption, and age at smoking initiation, respectively. A genome-wide association study (GWAS) was conducted adjusting for gender, age, and genetic principal components. The results suggest new candidate loci (4q22.1, 12p13.1, 16p13.3) interacting with smoking behavior for the risk of BC. Our results need to be validated in other case-control or cohort studies.

1. Introduction

Bladder cancer (BC) ranked tenth among the most common cancer types worldwide, with an estimated total of 573,278 new bladder cancer cases and 212,536 deaths due to bladder cancer in 2020 [1]. Over three-fourths of all BC cases occur in men [1].
Bladder cancer risk factors can be divided into external exposure and genetic predisposition variants. Studies have suggested an increased risk of bladder cancer associated with occupational exposure to carcinogens [2,3,4] and water contaminants [5]. However, the most important known bladder cancer risk factor is tobacco smoking [6,7,8]. It is estimated that almost 50% of bladder cancer cases are attributable to smoking [9] and can be prevented.
Previous studies suggested a potential role of genetic predisposition in bladder cancer occurrence [10]. Recent genome-wide association studies (GWAS) have been successful in identifying variants associated with bladder cancer [11,12,13,14,15]. These studies identified susceptibility loci at 3q28 (variants with the nearest gene TP63) [14], 4p16.3 (variants with the nearest gene TACC3) [13], and 8q24.21 (variants with the nearest gene MYC) [14]. However, most of these variants are associated with a small bladder cancer risk. To provide more insights into the biological mechanism of bladder cancer, gene-environment (GxE) interactions have been proposed. Interactions between cigarette smoking (as the major risk of bladder cancer) and N-acetyl transferase 2 (NAT2) and Glutathione-s-transferaseM1 (GSTM1) variants, the two best-known candidates, have been consistently associated with BC development [16]. Indeed, studies have shown a higher relative risk of bladder cancer due to smoking for NAT2 slow acetylators compared to NAT2 rapid/intermediate acetylators. This was justified by the fact that NAT2 detoxifies aromatic amines (primary bladder cancer carcinogen in tobacco) [17], and hence the effect of smoking becomes limited to smokers. The GxE interaction studies are important as they help to identify potentially new loci interacting with environmental exposure that would not be revealed by the main GWAS analysis. The results of such studies can help to set up new strategies for bladder cancer prevention and control. To date, most GxE interaction studies have focused on candidate genes, such as carcinogen detoxification genes, and much less attention has been given to ‘agnostic’, free-of-candidate, GxE interactions. Two studies went further and performed these association analyses on a wider range of single nucleotide polymorphisms (SNPs) (using more than 490,000 genotyped data [18] or imputed SNPs [19]), and found some suggestive evidence of interactions (either additive or multiplicative) between smoking and new SNPs and bladder cancer risk [18], but there is still a need for more data on the potential interaction between genetic variants and smoking habits.
However, conducting a GxE interaction study could be challenging due to the size of the study and the limited power for interaction analysis. The standard approach to test the interaction when data comes from case-control studies is to use a regression model by adding an interaction term between the variables of interest (for example, genetic variants and smoking phenotypes). Using this approach to detect an interaction effect of the same magnitude as the main effect (obtained from GWAS analysis), the sample size should be at least four times the sample used for the main effects analysis [20]. In the absence of a control group, the case-only design has been proposed as an alternative approach with more power for interaction tests [21,22].
Here, we aimed to explore the genome-wide gene-smoking interactions in bladder cancer development. While previous studies have mostly focused on smoking status and tobacco consumption [18,23,24,25], we looked at three main smoking phenotypes that characterize individuals’ smoking behavior throughout their life: age at smoking initiation, smoking status (never vs. ever smokers), and average tobacco consumption. We used the COBLAnCE cohort (COhort to study BLAdder CancEr) [26], a large prospective French-based cohort containing bladder cancer patients for whom a wide range of phenotypic and environmental measures and biological samples were prospectively collected. We conducted a genome-wide gene-smoking interaction study in bladder cancer based on ~9,700,000 imputed SNPs data from ~1700 bladder cancer patients of the European ancestry population, using a case-only design framework.

2. Materials and Methods

2.1. Study Population

COBLAnCE is a multicentre cohort of bladder cancer patients (eleven public and three private hospitals). It includes newly diagnosed patients with bladder cancer who were within 1 year of their initial diagnosis and aged 18 years or older. Recruitment was performed from December 2012 to June 2018. A total of 1800 patients were enrolled in the COBLAnCE cohort. Demographic information and detailed smoking data were collected from questionnaires completed by trained nurses who asked questions of patients at baseline.

2.2. Smoking Phenotypes

A complete history of tobacco smoking was ascertained separately for cigarettes, cigars, and pipes. Depending on the availability of data in the COBLAnCE cohort and based on previous genetic associations in the literature, three phenotypes were selected:
Smoking status: whether the patient reported as never smoker (less than 100 cigarettes/lifetime) or ever smoker. This phenotype was considered a binary phenotype coded by 0 (never smoker) or 1 (ever smoker).
Age at smoking initiation: the age at which the patient started smoking (cigarette/pipe or cigar). This phenotype was considered a continuous phenotype and was log-transformed since it is highly skewed.
Average tobacco consumption in grams for ever-smokers: using a complete history of tobacco smoking collected separately for cigarettes, cigars, and pipes. To standardize consumption across products, we calculated product-specific lifetime consumption (in grams of tobacco) based on the International Agency for Research on Cancer estimates of average unit weights [27,28]. The average tobacco consumption was then calculated by dividing the cumulative lifetime tobacco consumption (in grams) by the overall duration of smoking and was log-transformed since this variable is highly skewed. Given its construction, this phenotype combines both tobacco consumption and the duration of smoking across a lifetime.
The analyses of the phenotypes of age at smoking initiation and average tobacco consumption excluded never smokers.

2.3. Genotype Data, Quality Control, and Imputation

Peripheral blood samples were collected from all patients at enrolment (N = 1800). Samples were sent to the CEPH-Biobank, on average, within 48 h after collection for processing. DNA was extracted from buffy-coats using the salting-out method on the Autopure (Qiagen, Hilden, Germany) automated system and quantified using fluorimetry with the Quant-iT DNA Assay kit, Broad Range (Thermo Fisher Scientific, Waltham, MA, USA). Aliquots of DNA from 1765 patients were sent to the National Centre of Research in Human Genomics (CNRGH, CEA, Evry, France) for genotyping using the Illumina Global Screening Array with a Multi-Disease drop-in panel (GSA-MD v1.0, Illumina San Diego, CA, USA), which allows the analysis of ~700,000 polymorphisms. Genotyping was performed at the CNRGH on an Illumina automated high-throughput genotyping platform, according to the manufacturer’s instructions; 2 internal positive controls were included on each plate. All aliquoted samples were genotyped, and genotypes of 1762 patients were transferred for further analysis (failed genotyping for 21 samples, Figure 1).
Quality control (QC) on genotyped data was performed using PLINK version 1.9 [29,30]. We excluded samples with a call rate <95% (N = 2) and with a sex-discordant call (N = 1). Additionally, 2 pairs of related subjects were identified, and for each pair, the patient with the greater number of missing genotype data was excluded. Of the remaining 1757 patients, 1732 were of European ancestry.
Of the 687,572 SNPs that were successfully obtained, we excluded duplicated SNPs (n = 1168) with a call rate <95% (n = 9572), SNPs not in Hardy-Weinberg Equilibrium (HWE) (with the GWAS threshold p < 5 × 10−8) (n = 1214), monomorphic SNPs (n = 104,352) and SNPs with a minor allele frequency (MAF) <1% (n = 63,173). The final set of SNPs retained comprised 508,093 SNPs.
Imputation of missing SNPs was performed only on the European ancestry population (N = 1732 patients and n = 508,093 SNPs), and 1000 Genome Phase 3 was used as the reference panel [31]. Additional filtering was applied to the genotyped data to increase the accuracy of the imputation. We, therefore, excluded the following SNPs from imputation: SNPs with a call rate < 98% (n = 7449) and MAF < 1% (n = 1571), ambiguous SNPs (n = 141,456), SNPs with different frequencies compared to the reference panel (n = 1108), non-autosomic SNPs (n = 8365), and SNPs that could not be linked to the reference panel (n = 348). Therefore, imputation was based on 347,796 genotyped SNPs. A two-stage approach was considered for the imputation: first, we phased GWAS samples with SHAPEIT (v2.r904) [32] using burn, prune, and main parameters (--burn 10 --prune 10 --main 50) and then genotype imputation using IMPUTE2 [33,34]. We considered non-overlapping intervals defined by the centromere locations of the human reference sequence assembly (GRCh37). For the imputation, we set the buffer region to 500 kb and the number of reference haplotypes to 800. The imputation gave us a sample of 15,292,387 SNPs. Once imputed, we excluded imputed variants with imputation info values (r2) < 0.3 (n = 609,891), MAF < 1% (n = 4,760,135), and multi-allelic SNPs (n = 202,747). Our final SNP population was comprised of 9,719,614 SNPs ready for use in our association analyses.
The whole procedure (from blood collection to genotyping and imputation) is outlined in Figure 1.

2.4. Statistical Analysis

GWAS was performed on COBLAnCE cohort participants of European ancestry with available genotyping data after QC and imputation. We employed a case-only study design to estimate gene-environment interactions to test the association between SNPs and smoking phenotypes in bladder cancer patients. The analyses were conducted using logistic regression (for smoking status phenotype) and linear regression models (for age at smoking initiation and average tobacco consumption), with phenotype as the dependent variable against each of the imputed SNPs as the independent variable. Regressions were adjusted for gender, age, and the first ten genetic principal components to control for the study population’s latent genetic structure. The results were presented by Manhattan plots and quantile-quantile (Q-Q) plots, and the significance threshold was set at the established GWAS level (5 × 10−8). Functional annotation was performed for significant SNPs and the top hits using ANNOVAR [35] to identify the overlapping or closest gene to each SNP. For these SNPs, effect size estimates were presented as odds ratios (ORs), 95% confidence interval (CI) or βs, standard error (SE), and p-values. An OR greater than 1.0 and a β greater than 0 suggest a positive interaction for patients with reference alleles compared to those with alternative alleles. The case-only design approach relies on the independence assumption of genetic variants and environmental exposure. If the independence assumption is violated, the number of false positives will increase. A two-step approach was proposed in the literature to overcome the power insufficiency of the case-control approach [36]. In the first step, a case-only design of GxE interaction is performed, and a subset of SNPs is selected. In the second step, the selected SNPs are used in a case-control approach for the GxE interaction. This two-step strategy could help reduce false positive results. Since no control group was available in the COBLAnCE cohort, we opted for a different second step in which we considered a set of 433 SNPs from the literature. The significant association between these SNPs and different smoking phenotypes was shown within a large study of up to 1.2 million individuals [37]. We verified whether one of our top SNPs (10 SNPs with the smallest p-values) was part of the selected SNPs or was in high linkage disequilibrium (LD) with them
We also searched the literature for candidate SNPs that reported a smoking interaction association for bladder cancer risk (selected either from association studies with smoking phenotypes in the case-only population or from gene-smoking interaction studies in the case-control population). We narrowed our research to articles published from 2008 onwards. SNPs were filtered and retained with a reported p < 10−4 within the European population and were not associated with smoking, which led us to a total number of 37 SNPs [16,18]. We investigated the corresponding p-values of these SNPs in our GWAS results.
All statistical analyses were performed using PLINK v1.9 and R version 4.1.1 (https://www.r-project.org/ (accessed on 6 July 2023)).

3. Results

The baseline characteristics of the study population are presented in Table 1. The study population comprised 1732 patients with European ancestry (303 females and 1429 males) with a mean age of 68 years old. The difference in gender ratio was mostly attributable to the higher prevalence of smoking among males than among women in France several years ago. Accordingly, the results for smoking exposure were presented separately for males and females.
Of these 1732 patients, 308 were never smokers, and 1366 were former smokers with missing smoking status for 58 patients (3.35%). The proportion of never smokers was higher among women (37.5%) than men (13.51%).
The mean duration of tobacco consumption was almost the same for men and women (33 years vs. 34 years, respectively). However, men consumed more tobacco than women (19 g on average vs. 14 g, respectively). It seems that men started smoking earlier than women (18 years old vs. 20 years old, respectively).
As shown in Figure 1, analyses for smoking status were carried out on 1674 patients out of 1732 patients because of the missing smoking status of 58 patients. For the age at smoking initiation and average tobacco consumption, analyses were restricted to 1342 patients and 1283 patients, respectively, because of the missing age at smoking initiation for 24 patients, having never smoker status for 308 patients, and missing tobacco consumption for 83 patients. We performed a GWAS using a case-only design. The Manhattan plot of −log10(p) and the Q-Q plots of the observed p-values versus the expected p-values and the genomic inflation factor (λ) of our GWAS analyses for smoking status, age at smoking initiation, and average tobacco consumption phenotypes are shown in Figure 2A,B, Figure 3A,B and Figure 4A,B. No genomic inflation was detected in our GWAS (λ = 1.003, 1.003, and 1.002 for analyses of smoking status, age at smoking initiation, and average tobacco consumption, respectively).
The top ten SNPs for each GWAS are presented in Table 2, for which only four SNPs reached the genome-wide significance threshold (p < 5 × 10−8). None of these SNPs were mentioned previously in the literature in association with smoking phenotypes and were not in high LD with the 433 selected SNPs (squared correlation r L D 2 between pair of SNPs ≤0.90, Supplementary Figure S1). Therefore, we presumed the independence of SNPs and smoking, which is the main hypothesis of a case-only design.
For the smoking status phenotype, none of the SNPs reached the GWAS significance threshold. However, the most significant SNP variants interacting with smoking status for BC risk were rs114073636 and rs116571608, located on chromosome 1p31.3 (OR = 0.31, 95% CI = [0.20–0.47], p = 8.68 × 10−8 and OR = 0.31, 95% CI = [0.20–0.47], p = 8.87 × 10−8, respectively for patients with reference allele compared to those with alternative allele). Both variants were intergenic. Our results also suggest a promising locus on 4q22.1 (not reaching the GWAS threshold but significant at the 10−5 threshold) with an intergenic SNP located between genes TIGD2 and GPRIN3 (Figure 2A). The regional association plot of this locus showed high LD with surrounding SNPs (Figure 2C).
Analyses yielded three SNPs significantly interacting with age at smoking initiation (Figure 3A). The reference allele at rs531756449, an intergenic variant located on chromosome 1q44, interacted positively with age at smoking initiation (β(SE) = 0.3 (0.05), p = 8.26 × 10−9). The reference allele at SNPs located on the 12p13.1 locus showed positive interaction associations with age at smoking initiation and bladder cancer risk (rs77186197 with β(SE) = 0.29 (0.05), p = 3.74 × 10−9, rs78947799 with β(SE) = 0.31 (0.06), p = 3.97 × 10−8 and rs79782126 with β(SE) = 0.18 (0.03), p = 2.76 × 10−7). The regional association plot of the 12p13.1 locus (Figure 3C) mapped between genes GRIN2B and ATF7IP shows elevated LD with two other surrounding variants (>0.4) that did not reach the significance threshold.
The reference allele of an intronic SNP, located on gene GRAMD1B on chromosome 11q24.1, significantly interacted positively with average tobacco consumption (rs2714069 with β(SE) = 0.63 (0.11), p = 1.35 × 10−8). However, the reference alleles of SNPs located on locus 16p13.3 seems to have protective variants for the interaction between average tobacco consumption (rs113683380 with β(SE) = −0.66 (0.12), p = 6.44 × 10−8 and rs113590624 with β(SE) = −0.64 (0.12), p = 2.74 × 10−7). These variants were intronic and located on gene RBFOX1 (Figure 4A). The regional association plot of the 16p13.3 locus showed high LD between these two promising variants (Figure 4C).
We also investigated previously published SNPs in the literature in our GWAS results. Supplementary Table S1 shows a comparison of our results for the 37 selected SNPs. None of the SNPs were close to the GWAS threshold. However, the SNP with the smallest p-value was rs1495741, which showed a positive interaction with smoking initiation (OR = 1.37, 95% CI = [1.11–1.70], p = 0.004 for patients with reference allele compared to alternative allele), located on NAT2.

4. Discussion

The current paper focused on the interaction between SNPs and different smoking phenotypes in relation to bladder cancer risk. To our knowledge, this is the third study that focuses on free-of-candidate GxE interactions.
We performed a genome-wide interaction study of smoking phenotypes and BC risk within a case-only design framework. Our analyses were based on the COBLAnCE study, which to our knowledge, is one of the largest prospective cohorts of bladder cancer patients with numerous data and sequential biological samples collected at baseline and during the follow-up. Such prospective cohorts with large sample sizes are rare because of the major efforts required for patient recruitment and data/biomaterial collection.
Within the COBLAnCE cohort, we identified multiple loci on chromosomes 4q22.1, 12p13.1, and 16p13.3 as new candidates containing SNPs interacting with our smoking phenotypes (smoking status, age at smoking initiation and average tobacco smoking, respectively) for the risk of bladder cancer.
The reference allele at SNPs on locus 4q22.1 (rs542541627 and rs1533294) suggested a positive interaction with smoking status. These variants were intergenic and were mapped between genes TIGD2 and GPRIN3. The closest gene, TIGD2 (Tigger Transposable Element Derived 2), is a protein-coding gene. The other gene, GPRIN3 (G Protein-Regulated Inducer of Neurite Outgrowth 3), a protein-coding gene, is predicted to be involved in neuron projection development and to be active in the plasma membrane. It has also been suggested that GPRIN3 is associated with COPD (severe chronic obstructive pulmonary disease) and emphysema [38], and cocaine addiction [39].
The reference allele at our significant SNPs on locus 12p13.1 (rs77186197 and rs78947799) showed positive interactions with age at smoking initiation and was mapped between GRIN2B and ATF7IP. This suggests that these variants are more relevant in interaction with the early stages of smoking. The closest gene to these variants was GRIN2B (Glutamate Ionotropic Receptor NMDA Type Subunit 2B). A protein-coding gene that encodes a member of the N-methyl-D-aspartate (NMDA) receptor family within the ionotropic glutamate receptor superfamily. Its pathways are Transcriptional Regulation by MECP2 (Methyl-CpG-binding protein 2) (reactome: R-HSA-8986944), which regulates neuronal receptors and channels (reactome: R-HSA-9022699), Disorders of Nervous System Development (reactome: R-HSA-9697154), and Nicotine (KEGG: mpah05033) and cocaine (KEGG: mpah05030) addiction. A new study showed that the overexpression of MECP2 attenuates cigarette smoke extract-induced lung epithelial cell injury [40]. GRIN2B is also associated with nicotine dependence [41] and is one of the genes influencing smoking behaviors [42]. However, the ATF7IP (Activating Transcription Factor 7 Interacting Protein), a protein-coding gene, is a multifunctional nuclear protein associated with heterochromatin. ATF7IP is inactivated in 5% of lung adenocarcinomas among non-smokers [43].
The reference allele of two of the promising SNPs (rs113683380 and rs113590624, with a negative interaction with average tobacco consumption) located on 16p13.3 was intronic and mapped to the RBFOX1 gene. RBFOX1 (RNAS Binding FOX-1 Homolog 1) is a protein-coding gene that has a role in alternative splicing regulation. RBFOX1 (also known as A2BP1) is highly expressed in neurons in brain regions that include the hippocampus, and there is evidence of associations between this gene, smoking cessation, and nicotine dependence [44,45]. Interestingly, one of the pathways of RBFOX1 is transcriptional regulation by MECP2 (reactome: R-HSA-8986944). The RBFOX1 gene was identified as associated with substance dependence or the ability to quit smoking in 13 independent datasets [46], with smoking frequency [47]. In addition, RBFOX1 was presented as a strong candidate for susceptibility to aggressive behavior and to several psychiatric disorders [48,49]. Our results also showed a significant interaction association with average tobacco consumption for the reference allele of an intronic SNP (rs2714069) located on the GRAMD1B (GRAM Domain Containing 1B) gene, a protein-coding gene. A study of breast epithelial cells showed significant upregulation in response to nicotine stress by GRAMD1B [50].
It is hard to compare the results of the current study with previously GxE interaction and bladder cancer risk results since most studies are based on case-control cohorts focusing only on smoking status (never vs. ever smokers). Considering a case-only design framework, our results should be interpreted as the multiplicative interaction effect of smoking, genetic variants, and the risk of bladder cancer. On the other side, most studies in the literature presented the additive interaction effect, besides two studies that also looked at the multiplicative interaction effect by performing stratified analyses [18,19]. For example, looking at the multiplicative interaction effect of NAT2 and smoking status within a meta-analysis study [19], the same direction of the results is obtained within the COBLAnCE cohort, but the p-value of NAT2 in our study did not reach the GWAS significant threshold which could be due to the small sample size compared to a meta-analysis approach.
The strength of our study comes from the COBLAnCE cohort, one of the largest prospective bladder cancer cohorts besides cohorts from Spain [51], the Netherlands [52], the UK [53], and the US [54,55]. Given the structure of the COBLAnCE cohort, it gives an accurate image of practices in bladder cancer patients in France, covering both public and private hospitals. Despite this, one of the limitations of our study is the absence of disease-free individuals. We carried out a case-only design; however, the main assumption of this approach, which is the independence between exposure and genetic variants, could not be validated in this cohort. The second limitation is the absence of replication. However, performing an exact replication study on other bladder cancer cohorts requires the availability of genotyped data and an adequate sample size, which in our case, would be smaller than the COBLAnCE cohort. Additionally, even though our cohort is one of the largest prospective bladder cancer cohorts, some SNPs had a very small MAF, and some important hits could have been lost. It is worth mentioning that we investigated the statistical interaction between genetic variants, smoking behavior, and the risk of bladder cancer. However, it does not directly imply a mechanistic interaction or causality. Our results highlight potential associations that may indicate the presence of complex biological interactions or shared pathways, but additional mechanistic studies are required to establish causality and plausible pathways.
Knowing these limitations, our results need to be investigated further in case-control or cohort studies; a GWAS meta-analysis would also be of utility to identify rare variants.

5. Conclusions

We found new suggestive loci on chromosomes 4q22.1, 12p13.1, and 16p13.3 interacting with different smoking phenotypes. Despite using the COBLAnCE cohort, one of the largest prospective bladder cancer cohorts, some limitations should be mentioned: use of a case-only design approach, absence of replication, and loss of information on some SNPs due to small MAF. Hence, these results need to be investigated in larger cohorts/case-control studies or within a meta-analysis framework.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/cancers15174218/s1, Figure S1: Pair-wise LD between SNPs in association with smoking phenotypes from the literature (shown in black) and the top SNPs from the COBLAnCE cohort (shown in blue). Red indicates a value of 1, and white indicates a value of 0. COBLAnCE, COhort to study BLAdder CancEr. LD, linkage disequilibrium. SNP, single nucleotide polymorphisms; Table S1: Selected SNPs from the literature and their corresponding ORs and βs in the COBLAnCE cohort.

Author Contributions

M.K. and S.M.-P. performed the statistical analyses. M.K., S.B. and S.M. conceived and coordinated the study. K.G. provided overall study management. H.B., A.B., C.B. and J.-F.D. performed blood sample processing and genotyping. M.K. drafted the manuscript. All authors revised the manuscript for important intellectual content and approved the submission of the manuscript. M.K. had full access to the data and takes responsibility for the integrity of the data and the accuracy of the data analysis. The work reported in the paper has been performed by the authors unless clearly specified in the text. All authors have read and agreed to the published version of the manuscript.

Funding

This project is funded by the National Agency of Research (Programme Investissements d’Avenir Cohortes ANR-10-COHO-0007), the National Institute of Health and Medical Research (Soutien aux cohortes pour la Recherche sur le Cancer, AAP 2013), the National League against Cancer (EP 2011) and the National Cancer Institute (Research programme and interventions to reduce and fight against tobacco consumption 2019-1-TABAC). Sponsors were not included in the present study design, collection, analysis, and interpretation of data or in publications that resulted from the study.

Institutional Review Board Statement

The study protocol has been approved by the French regulatory authorities (data protection authority Commission Nationale de l’Informatique et des Libertés (CNIL) N° DR-2012-441, biological collections authority N° DC-2011-1486) and a French ethics committee (Comité de Protection des Personnes Ile de France VII (N° CO-12-001). This project was approved by one French ethics committee (Comité de Protection des Personnes Ile de France VII, reference number CO-12-001) for all participating centers: Hôpital Foch (Suresnes), Centre Hospitalier Universitaire Henri Mondor (Créteil), Centre Hospitalier Universitaire Gabriel-Montpied (Clermont-Ferrand), Centre Hospitalier Régional Universitaire de Lille, Centre Hospitalier Universitaire de Nîmes, Clinique Beau Soleil (Montpellier), Hopital Pasteur (Colmar), Centre Hospitalier René Dubos (Pontoise), Clinique du Parc Rambot (Aix en Provence), Centre Hospitalier Universitaire Grenoble Alpes (Grenoble), Clinique Urologique Atlantis (Nantes), Centre Hospitalier Européen Georges Pompidou (Paris), and Centre Hospitalier de Brabois (Nancy).

Informed Consent Statement

All patients were informed of the objectives of the study and were invited to voluntarily participate at inclusion. Patients who agreed to participate provided written consent.

Data Availability Statement

The GWAS summary statistics will be available in the NHGRI-EBI GWAS Catalog. Pseudoanonymised data and biological samples collected during the study can be obtained for conducting scientific projects. Requests should be made to the corresponding author and will be evaluated by the Scientific Committee of the COBLAnCE cohort. For approved projects, data transfer agreements will be required to be set up.

Acknowledgments

The authors would like to thank the following hospitals for their involvement in patient recruitment: Hôpital Foch (Suresnes), Centre Hospitalier Universitaire Henri Mondor (Créteil), Centre Hospitalier Universitaire Gabriel-Montpied (Clermont-Ferrand), Centre Hospitalier Régional Universitaire de Lille, Centre Hospitalier Universitaire de Nîmes, Clinique Beau Soleil (Montpellier), Hopital Pasteur (Colmar), Centre Hospitalier René Dubos (Pontoise), Clinique du Parc Rambot (Aix en Provence), Centre Hospitalier Universitaire Grenoble Alpes (Grenoble), Clinique Urologique Atlantis (Nantes), Centre Hospitalier Européen Georges Pompidou (Paris), and Centre Hospitalier de Brabois (Nancy), and Caroline Barau, Pascale Maille, Fannie Semprez and Pascale Soyeux at the Platform of Biological Resources and Research Team INSERM U955 of Henri Mondor Hospital, Elodie Guyon and Odette Mariani at the Biological Resources Centre of Curie Institute, Clémentine Krucker and Isabelle Bernard-Pierrot at UMR144 CNRS for the management of biological specimens, Marie-Claude Babron, Pascale Jan, Anthony Mangin and Pierre Guyader at Bureau de Biostatistique et d’Épidémiologie, Gustave Roussy for data management of the COBLAnCE study, Yves Koudou at INSERM U1018 Oncostat for the quality control of genotyped data, and CEPH-Biobank and CNRGH genotyping teams for excellent technical support.

Conflicts of Interest

Julia Bonastre reports fees outside the scope of the submitted work as advisory board for BMS and funding of the MICADO study. Yves Allory reports fees outside the scope of the submitted work as consultant for AstraZeneca, Janssen, MSD, VitaDx, as invited speaker for Astellas, AstraZeneca, Bristol-Myers Squibb, MSD, Sanofi, as COBLAnCE cohort support for Bristol-Myers Squibb, MSD, and public and charity fundings from INCa, ANR, Ligue Contre le Cancer, and Fondation ARC. Simone Benhamou reports fundings outside the scope of the submitted work as COBLAnCE cohort support for Bristol-Myers Squibb, MSD, and public and charity fundings from INCa, ANR, and Ligue Contre le Cancer. Stefan Michiels reports fees outside the scope of the submitted work: Scientific Committee Study member: Roche, Data and safety monitoring member of clinical trials: Sensorion, Biophytis, Servier, IQVIA, Yuhan, Kedrion. The other authors declare no conflict of interest.

References

  1. Sung, H.; Ferlay, J.; Siegel, R.L.; Laversanne, M.; Soerjomataram, I.; Jemal, A.; Bray, F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J. Clin. 2021, 71, 209–249. [Google Scholar] [CrossRef] [PubMed]
  2. Yu, M.C.; Skipper, P.L.; Tannenbaum, S.R.; Chan, K.K.; Ross, R.K. Arylamine exposures and bladder cancer risk. Mutat. Res. Mol. Mech. Mutagen. 2002, 506-507, 21–28. [Google Scholar] [CrossRef] [PubMed]
  3. Chen, H.-I.; Liou, S.-H.; Loh, C.-H.; Uang, S.-N.; Yu, Y.-C.; Shih, T.-S. Bladder cancer screening and monitoring of 4,4′-Methylenebis(2-chloroaniline) exposure among workers in Taiwan. Urology 2005, 66, 305–310. [Google Scholar] [CrossRef] [PubMed]
  4. García-Pérez, J.; Pollán, M.; Boldo, E.; Pérez-Gómez, B.; Aragonés, N.; Lope, V.; Ramis, R.; Vidal, E.; López-Abente, G. Mortality due to lung, laryngeal and bladder cancer in towns lying in the vicinity of combustion installations. Sci. Total. Environ. 2009, 407, 2593–2602. [Google Scholar] [CrossRef] [PubMed]
  5. Fernández, M.I.; López, J.F.; Vivaldi, B.; Coz, F. Long-Term Impact of Arsenic in Drinking Water on Bladder Cancer Health Care and Mortality Rates 20 Years After End of Exposure. J. Urol. 2012, 187, 856–861. [Google Scholar] [CrossRef] [PubMed]
  6. Freedman, N.D.; Silverman, D.T.; Hollenbeck, A.R.; Schatzkin, A.; Abnet, C.C. Association Between Smoking and Risk of Bladder Cancer Among Men and Women. JAMA 2011, 306, 737–745. [Google Scholar] [CrossRef] [PubMed]
  7. Lammers, R.J.; Witjes, W.P.; Hendricksen, K.; Caris, C.T.; Janzing-Pastors, M.H.; Witjes, J.A. Smoking Status Is a Risk Factor for Recurrence After Transurethral Resection of Non–Muscle-Invasive Bladder Cancer. Eur. Urol. 2011, 60, 713–720. [Google Scholar] [CrossRef] [PubMed]
  8. Boström, P.J.; Alkhateeb, S.; Trottier, G.; Athanasopoulos, P.Z.; Mirtti, T.; Kortekangas, H.; Laato, M.; van Rhijn, B.; van der Kwast, T.; Fleshner, N.E.; et al. Sex differences in bladder cancer outcomes among smokers with advanced bladder cancer. BJU Int. 2011, 109, 70–76. [Google Scholar] [CrossRef]
  9. Saginala, K.; Barsouk, A.; Aluru, J.S.; Rawla, P.; Padala, S.A.; Barsouk, A. Epidemiology of Bladder Cancer. Med Sci. 2020, 8, 15. [Google Scholar] [CrossRef]
  10. Burger, M.; Catto, J.W.F.; Dalbagni, G.; Grossman, H.B.; Herr, H.; Karakiewicz, P.; Kassouf, W.; Kiemeney, L.A.; La Vecchia, C.; Shariat, S.; et al. Epidemiology and Risk Factors of Urothelial Bladder Cancer. Eur. Urol. 2013, 63, 234–241. [Google Scholar] [CrossRef]
  11. de Maturana, E.L.; Rava, M.; Anumudu, C.; Sáez, O.; Alonso, D.; Malats, N. Bladder Cancer Genetic Susceptibility. A Systematic Review. Bl. Cancer 2018, 4, 215–226. [Google Scholar] [CrossRef] [PubMed]
  12. Guey, L.T.; García-Closas, M.; Murta-Nascimento, C.; Lloreta, J.; Palencia, L.; Kogevinas, M.; Rothman, N.; Vellalta, G.; Calle, M.L.; Marenne, G.; et al. Genetic Susceptibility to Distinct Bladder Cancer Subphenotypes. Eur. Urol. 2010, 57, 283–292. [Google Scholar] [CrossRef] [PubMed]
  13. A Kiemeney, L.; Sulem, P.; Besenbacher, S.; Vermeulen, S.H.; Sigurdsson, A.; Thorleifsson, G.; Gudbjartsson, D.F.; Stacey, S.N.; Gudmundsson, J.; Zanon, C.; et al. A sequence variant at 4p16.3 confers susceptibility to urinary bladder cancer. Nat. Genet. 2010, 42, 415–419. [Google Scholar] [CrossRef]
  14. A Kiemeney, L.; Thorlacius, S.; Sulem, P.; Geller, F.; Aben, K.K.H.; Stacey, S.N.; Gudmundsson, J.; Jakobsdottir, M.; Bergthorsson, J.T.; Sigurdsson, A.; et al. Sequence variant on 8q24 confers susceptibility to urinary bladder cancer. Nat. Genet. 2008, 40, 1307–1312. [Google Scholar] [CrossRef]
  15. Figueroa, J.D.; Ye, Y.; Siddiq, A.; Garcia-Closas, M.; Chatterjee, N.; Prokunina-Olsson, L.; Cortessis, V.K.; Kooperberg, C.; Cussenot, O.; Benhamou, S.; et al. Genome-wide association study identifies multiple loci associated with bladder cancer risk. Hum. Mol. Genet. 2013, 23, 1387–1398. [Google Scholar] [CrossRef] [PubMed]
  16. Rothman, N.; Garcia-Closas, M.; Chatterjee, N.; Malats, N.; Wu, X.F.; Figueroa, J.D.; Real, F.X.; Van den Berg, D.; Matullo, G.; Baris, D.; et al. A multi-stage genome-wide association study of bladder cancer identifies multiple susceptibility loci. Nat. Genet. 2010, 42, 978–984. [Google Scholar] [CrossRef]
  17. Hein, D.W. Molecular genetics and function of NAT1 and NAT2: Role in aromatic amine metabolism and carcinogenesis. Mutat. Res. Mol. Mech. Mutagen. 2002, 506-507, 65–77. [Google Scholar] [CrossRef]
  18. Figueroa, J.D.; Han, S.S.; Garcia-Closas, M.; Baris, D.; Jacobs, E.J.; Kogevinas, M.; Schwenn, M.; Malats, N.; Johnson, A.; Purdue, M.P.; et al. Genome-wide interaction study of smoking and bladder cancer risk. Carcinog. 2014, 35, 1737–1744. [Google Scholar] [CrossRef]
  19. Koutros, S.; Kiemeney, L.A.; Choudhury, P.P.; Milne, R.L.; de Maturana, E.L.; Ye, Y.; Joseph, V.; Florez-Vargas, O.; Dyrskjøt, L.; Figueroa, J.; et al. Genome-wide Association Study of Bladder Cancer Reveals New Biological and Translational Insights. Eur. Urol. 2023, 84, 127–137. [Google Scholar] [CrossRef]
  20. Smith, P.G.; E Day, N. The Design of Case-Control Studies: The Influence of Confounding and Interaction Effects. Leuk. Res. 1984, 13, 356–365. [Google Scholar] [CrossRef]
  21. Piegorsch, W.W.; Weinberg, C.R.; Taylor, J.A. Non-hierarchical logistic models and case-only designs for assessing susceptibility in population-based case-control studies. Stat. Med. 1994, 13, 153–162. [Google Scholar] [CrossRef] [PubMed]
  22. Gauderman, W.J. Sample Size Requirements for Association Studies of Gene-Gene Interaction. Am. J. Epidemiology 2002, 155, 478–484. [Google Scholar] [CrossRef] [PubMed]
  23. García-Closas, M.; Malats, N.; Silverman, D.; Dosemeci, M.; Kogevinas, M.; Hein, D.W.; Tardón, A.; Serra, C.; Carrato, A.; García-Closas, R.; et al. NAT2 slow acetylation, GSTM1 null genotype, and risk of bladder cancer: Results from the Spanish Bladder Cancer Study and meta-analyses. Lancet 2005, 366, 649–659. [Google Scholar] [CrossRef]
  24. Garcia-Closas, M.; Rothman, N.; Figueroa, J.D.; Prokunina-Olsson, L.; Han, S.S.; Baris, D.; Jacobs, E.J.; Malats, N.; De Vivo, I.; Albanes, D.; et al. Common Genetic Polymorphisms Modify the Effect of Smoking on Absolute Risk of Bladder Cancer. Cancer Res 2013, 73, 2211–2220. [Google Scholar] [CrossRef] [PubMed]
  25. Lubin, J.H.; Kogevinas, M.; Silverman, D.; Malats, N.; Garcia-Closas, M.; Tardón, A.; Hein, D.W.; Garcia-Closas, R.; Serra, C.; Dosemeci, M.; et al. Evidence for an intensity-dependent interaction of NAT2 acetylation genotype and cigarette smoking in the Spanish Bladder Cancer Study. Leuk. Res. 2007, 36, 236–241. [Google Scholar] [CrossRef]
  26. Lebret, T.; Bonastre, J.; Fraslin, A.; Neuzillet, Y.; Droupy, S.; Rebillard, X.; De la Taille, A.; Guy, L.; Villers, A.; Schneider, M.; et al. COBLAnCE: A Prospective Cohort to Study Prognostic and Predictive Factors in Bladder Cancer and to Generate Re-al-World Data on Treatment Patterns, Resource Use and Quality of Life. BMJ Open 2023, in press. [Google Scholar]
  27. IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Tobacco Smoke and Involuntary Smoking; IARC: Lyon, France, 2004. [Google Scholar]
  28. Malhotra, J.; Borron, C.; Freedman, N.D.; Abnet, C.C.; Brandt, P.A.v.D.; White, E.; Milne, R.L.; Giles, G.G.; Boffetta, P. Association between Cigar or Pipe Smoking and Cancer Risk in Men: A Pooled Analysis of Five Cohort Studies. Cancer Prev. Res. 2017, 10, 704–709. [Google Scholar] [CrossRef]
  29. Purcell, S.; Chang, C. PLINK, [v1.9]. Available online: www.cog-genomics.org/plink/1.9/ (accessed on 6 July 2023).
  30. Chang, C.C.; Chow, C.C.; Tellier, L.C.; Vattikuti, S.; Purcell, S.M.; Lee, J.J. Second-generation PLINK: Rising to the challenge of larger and richer datasets. GigaScience 2015, 4, 7. [Google Scholar] [CrossRef]
  31. 1000 Genomes Project Consortium; Auton, A.; Brooks, L.D.; Durbin, R.M.; Garrison, E.P.; Kang, H.M.; Korbel, J.O.; Marchini, J.L.; McCarthy, S.; McVean, G.A.; et al. A global reference for human genetic variation. Nature 2015, 526, 68–74. [Google Scholar]
  32. Delaneau, O.; McVean, G.A.; Donnelly, P.; Lunter, G.; Marchini, J.L.; Myers, S.; Gupta-Hinch, A.; Iqbal, Z.; Mathieson, I.; Rimmer, A.; et al. Integrating sequence and array data to create an improved 1000 Genomes Project haplotype reference panel. Nat. Commun. 2014, 5, 1–9. [Google Scholar] [CrossRef]
  33. Howie, B.N.; Donnelly, P.; Marchini, J. A Flexible and Accurate Genotype Imputation Method for the Next Generation of Genome-Wide Association Studies. PLOS Genet. 2009, 5, e1000529. [Google Scholar] [CrossRef] [PubMed]
  34. Howie, B.; Marchini, J.; Stephens, M. Genotype Imputation with Thousands of Genomes. G3 Genes|Genomes|Genetics 2011, 1, 457–470. [Google Scholar] [CrossRef] [PubMed]
  35. Wang, K.; Li, M.; Hakonarson, H. ANNOVAR: Functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010, 38, e164. [Google Scholar] [CrossRef] [PubMed]
  36. Murcray, C.E.; Lewinger, J.P.; Gauderman, W.J. Gene-Environment Interaction in Genome-Wide Association Studies. Am. J. Epidemiology 2008, 169, 219–226. [Google Scholar] [CrossRef] [PubMed]
  37. Liu, M.; Jiang, Y.; Wedow, R.; Li, Y.; Brazel, D.M.; Chen, F.; Datta, G.; Davila-Velderrain, J.; McGuire, D.; Tian, C.; et al. Association studies of up to 1.2 million individuals yield new insights into the genetic etiology of tobacco and alcohol use. Nat. Genet. 2019, 51, 237–244. [Google Scholar] [CrossRef]
  38. Sakornsakolpat, P.; Morrow, J.D.; Castaldi, P.J.; Hersh, C.P.; Bossé, Y.; Silverman, E.K.; Manichaikul, A.; Cho, M.H. Integrative genomics identifies new genes associated with severe COPD and emphysema. Respir. Res. 2018, 19, 46. [Google Scholar] [CrossRef]
  39. Cheron, J.; D’exaerde, A.d.K. Drug addiction: From bench to bedside. Transl. Psychiatry 2021, 11, 1–14. [Google Scholar] [CrossRef]
  40. Lin, J.; Peng, J.; Liu, G.; Deng, L. Overexpression of MECP2 attenuates cigarette smoke extracts induced lung epithelial cell injury by promoting CYP1B1 methylation. J. Toxicol. Sci. 2020, 45, 177–186. [Google Scholar] [CrossRef]
  41. Korytina, G.F.; Akhmadishina, L.Z.; Kochetova, O.V.; Aznabaeva, Y.G.; Zagidullin, S.Z.; Victorova, T.V. Polymorphic variants of glutamate receptor (GRIK5, GRIN2B) and serotonin receptor (HTR2A) genes are associated with chronic. Mol. Biol. 2017, 51, 533–542. [Google Scholar] [CrossRef]
  42. Vink, J.M.; Smit, A.B.; de Geus, E.J.; Sullivan, P.; Willemsen, G.; Hottenga, J.-J.; Smit, J.H.; Hoogendijk, W.J.; Zitman, F.G.; Peltonen, L.; et al. Genome-wide Association Study of Smoking Initiation and Current Smoking. Am. J. Hum. Genet. 2009, 84, 367–379. [Google Scholar] [CrossRef]
  43. Pros, E.; Saigi, M.; Alameda, D.; Gomez-Mariano, G.; Martinez-Delgado, B.; Alburquerque-Bejar, J.; Carretero, J.; Tonda, R.; Esteve-Codina, A.; Catala, I.; et al. Genome-wide profiling of non-smoking-related lung cancer cells reveals common RB1 rearrangements associated with histopathologic transformation in EGFR-mutant tumors. Ann. Oncol. 2020, 31, 274–282. [Google Scholar] [CrossRef] [PubMed]
  44. Wang, J.; Li, M.D. Common and Unique Biological Pathways Associated with Smoking Initiation/Progression, Nicotine Dependence, and Smoking Cessation. Neuropsychopharmacology 2009, 35, 702–719. [Google Scholar] [CrossRef] [PubMed]
  45. Uhl, G.R.; Liu, Q.-R.; Drgon, T.; Johnson, C.; Walther, D.; Rose, J.E.; David, S.P.; Niaura, R.; Lerman, C. Molecular Genetics of Successful Smoking Cessation. Arch. Gen. Psychiatry 2008, 65, 683–693. [Google Scholar] [CrossRef]
  46. Zhong, X.; Drgonova, J.; Li, C.-Y.; Uhl, G.R. Human cell adhesion molecules: Annotated functional subtypes and overrepresentation of addiction-associated genes. Ann. New York Acad. Sci. 2015, 1349, 83–95. [Google Scholar] [CrossRef] [PubMed]
  47. Thorgeirsson, T.E.; Gudbjartsson, D.F.; Surakka, I.; Vink, J.M.; Amin, N.; Geller, F.; Sulem, P.; Rafnar, T.; Esko, T.; Walter, S.; et al. Sequence variants at CHRNB3–CHRNA6 and CYP2A6 affect smoking behavior. Nat. Genet. 2010, 42, 448–453. [Google Scholar] [CrossRef]
  48. Fernàndez-Castillo, N.; Gan, G.; van Donkelaar, M.M.; Vaht, M.; Weber, H.; Retz, W.; Meyer-Lindenberg, A.; Franke, B.; Harro, J.; Reif, A.; et al. RBFOX1, encoding a splicing regulator, is a candidate gene for aggressive behavior. Eur. Neuropsychopharmacol. 2017, 30, 44–55. [Google Scholar] [CrossRef]
  49. O’leary, A.; Fernàndez-Castillo, N.; Gan, G.; Yang, Y.; Yotova, A.Y.; Kranz, T.M.; Grünewald, L.; Freudenberg, F.; Antón-Galindo, E.; Cabana-Domínguez, J.; et al. Behavioural and functional evidence revealing the role of RBFOX1 variation in multiple psychiatric disorders and traits. Mol. Psychiatry 2022, 27, 4464–4473. [Google Scholar] [CrossRef]
  50. Bavarva, J.H.; Tae, H.; Settlage, R.E.; Garner, H.R. Characterizing the Genetic Basis for Nicotine Induced Cancer Development: A Transcriptome Sequencing Study. PLoS ONE 2013, 8, e67252. [Google Scholar] [CrossRef]
  51. Samanic, C.; Kogevinas, M.; Dosemeci, M.; Malats, N.; Real, F.X.; Garcia-Closas, M.; Serra, C.; Carrato, A.; García-Closas, R.; Sala, M.; et al. Smoking and Bladder Cancer in Spain: Effects of Tobacco Type, Timing, Environmental Tobacco Smoke, and Gender. Cancer Epidemiol. Biomark. Prev. 2006, 15, 1348–1354. [Google Scholar] [CrossRef]
  52. De Goeij, L.; Westhoff, E.; Witjes, J.A.; Aben, K.K.; Kampman, E.; Kiemeney, L.A.; Vrieling, A. The UroLife study: Protocol for a Dutch prospective cohort on lifestyle habits in relation to non-muscle-invasive bladder cancer prognosis and health-related quality of life. BMJ Open 2019, 9, e030396. [Google Scholar] [CrossRef]
  53. Zeegers, M.P.; Bryan, R.T.; Langford, C.; Billingham, L.; Murray, P.; Deshmukh, N.S.; Hussain, S.; James, N.; Wallace, D.M.A.; Cheng, K. The West Midlands Bladder Cancer Prognosis Programme: Rationale and design. BJU Int. 2010, 105, 784–788. [Google Scholar] [CrossRef] [PubMed]
  54. Kwan, M.L.; Kushi, L.H.; Danforth, K.N.; Roh, J.M.; Ergas, I.J.; Lee, V.S.; Cannavale, K.L.; Harrison, T.N.; Contreras, R.; Loo, R.K.; et al. The Be-Well Study: A prospective cohort study of lifestyle and genetic factors to reduce the risk of recurrence and progression of non-muscle-invasive bladder cancer. Cancer Causes Control. 2019, 30, 187–193. [Google Scholar] [CrossRef] [PubMed]
  55. Westhoff, E.; Wu, X.; Kiemeney, L.A.; Lerner, S.P.; Ye, Y.; Huang, M.; Dinney, C.P.; Vrieling, A.; Tu, H. Dietary patterns and risk of recurrence and progression in non-muscle-invasive bladder cancer. Int. J. Cancer 2017, 142, 1797–1804. [Google Scholar] [CrossRef] [PubMed]
Cancers 15 04218 g001
Figure 1. Flowchart of the study population from data collection, genotyping, and imputation to data analyses in the COBLAnCE cohort. GWAS, genome-wide association study. COBLAnCE, COhort to study BLAdder CancEr. HWE, Hardy-Weinberg Equilibrium. MAF, minor allele frequency. SNP, single nucleotide polymorphisms.
Figure 1. Flowchart of the study population from data collection, genotyping, and imputation to data analyses in the COBLAnCE cohort. GWAS, genome-wide association study. COBLAnCE, COhort to study BLAdder CancEr. HWE, Hardy-Weinberg Equilibrium. MAF, minor allele frequency. SNP, single nucleotide polymorphisms.
Cancers 15 04218 g001
Cancers 15 04218 g002
Figure 2. (A) Manhattan and (B) Q-Q plots of GWAS for interactions with smoking status. Blue and red horizontal lines indicate 1 × 10−5 and 5 × 10−8 (the typical GWAS threshold) thresholds, respectively. (C) Regional association plot for the 4q22.1 locus interacting with smoking status. The most significantly associated SNP in the region is presented in a diamond (rs542541627). Other dots correspond to other analyzed SNPs. Dots are colored by their linkage disequilibrium coefficient. GWAS, genome-wide association study. Q-Q plots, quantile–quantile plots. SNP, single nucleotide polymorphisms.
Figure 2. (A) Manhattan and (B) Q-Q plots of GWAS for interactions with smoking status. Blue and red horizontal lines indicate 1 × 10−5 and 5 × 10−8 (the typical GWAS threshold) thresholds, respectively. (C) Regional association plot for the 4q22.1 locus interacting with smoking status. The most significantly associated SNP in the region is presented in a diamond (rs542541627). Other dots correspond to other analyzed SNPs. Dots are colored by their linkage disequilibrium coefficient. GWAS, genome-wide association study. Q-Q plots, quantile–quantile plots. SNP, single nucleotide polymorphisms.
Cancers 15 04218 g002
Cancers 15 04218 g003
Figure 3. (A) Manhattan and (B) Q-Q plots of GWAS for interaction with age at smoking initiation. Blue and red horizontal lines indicate 1 × 10−5 and 5 × 10−8 (the typical GWAS threshold) thresholds, respectively. (C) Regional association plot for the 12p13.1 locus interacting with age at smoking initiation. The most significantly associated SNP in the region is presented in a diamond (rs77186197). Other dots correspond to other analyzed SNPs. Dots are colored by their linkage disequilibrium coefficient. GWAS, genome-wide association study. Q-Q plots, quantile–quantile plots. SNP, single nucleotide polymorphisms.
Figure 3. (A) Manhattan and (B) Q-Q plots of GWAS for interaction with age at smoking initiation. Blue and red horizontal lines indicate 1 × 10−5 and 5 × 10−8 (the typical GWAS threshold) thresholds, respectively. (C) Regional association plot for the 12p13.1 locus interacting with age at smoking initiation. The most significantly associated SNP in the region is presented in a diamond (rs77186197). Other dots correspond to other analyzed SNPs. Dots are colored by their linkage disequilibrium coefficient. GWAS, genome-wide association study. Q-Q plots, quantile–quantile plots. SNP, single nucleotide polymorphisms.
Cancers 15 04218 g003
Cancers 15 04218 g004
Figure 4. (A) Manhattan and (B) Q-Q plots of GWAS for the interaction with average tobacco consumption. Blue and red horizontal lines indicate 1 × 10−5 and 5 × 10−8 thresholds (the typical GWAS threshold), respectively; (C) Regional association plot for the 16p13.3 locus interacting with average tobacco consumption. The most significantly associated SNP in the region is presented in a diamond (rs113683380). Other dots correspond to other analyzed SNPs. Dots are colored by their linkage disequilibrium coefficient. GWAS, genome-wide association study. Q-Q plots, quantile-quantile plots. SNP, single nucleotide polymorphisms.
Figure 4. (A) Manhattan and (B) Q-Q plots of GWAS for the interaction with average tobacco consumption. Blue and red horizontal lines indicate 1 × 10−5 and 5 × 10−8 thresholds (the typical GWAS threshold), respectively; (C) Regional association plot for the 16p13.3 locus interacting with average tobacco consumption. The most significantly associated SNP in the region is presented in a diamond (rs113683380). Other dots correspond to other analyzed SNPs. Dots are colored by their linkage disequilibrium coefficient. GWAS, genome-wide association study. Q-Q plots, quantile-quantile plots. SNP, single nucleotide polymorphisms.
Cancers 15 04218 g004
Table 1. Descriptive characteristics of the study population.
Table 1. Descriptive characteristics of the study population.
Overall (N = 1732)Women (N = 303)Men (N = 1429)
Age (Years)
Mean (SD)68.55 (10.79)68.05 (12.46)68.65 (10.41)
Median (Q1, Q3)68.70 (61.74, 76.34)68.25 (60.27, 77.31)68.73 (62.07, 76.10)
Min; Max22.05; 95.3122.2; 93.3422.05; 95.31
Missing101
Smoking status
Never smoker308 (17.78%)115 (37.95%)193 (13.51%)
Ever smoker1366 (78.87%)176 (58.09%)1190 (83.28%)
Missing58 (3.35%)12 (3.96%)46 (3.22%)
Duration of smoking (Years) (N = 1366)
Mean (SD)34.10 (14.54)33.09 (14.17)34.25 (14.59)
Median (Q1, Q3)35.00 (24.00, 45.00)35.00 (22.00, 44.00)35.00 (24.00, 45.00)
Min; Max1; 761; 621; 76
Missing831766
Average tobacco consumption in grams (N = 1366)
Mean (SD)19.82 (12.65)16.00 (11.45)20.36 (12.72)
Median (Q1, Q3)18.61 (11.43, 23.51)14.41 (9.23, 20.00)19.11 (12.16, 24.67)
Min; Max0.5; 1151; 62.060.5; 115
Missing831766
Age at smoking initiation (Years) (N = 1366)
Mean (SD)18.40 (5.83)20.79 (7.46)18.06 (5.47)
Median (Q1, Q3)18.00 (15.00, 20.00)18 (17.00, 22.00)17 (15.00, 20.00)
Min; Max7; 7310; 567; 73
Missing24519
SD, standard deviation; Q1, first quartile; Q3, third quartile.
Table 2. Top 10 SNPs of GWAS for interaction with smoking phenotypes in the COBLAnCE cohort.
Table 2. Top 10 SNPs of GWAS for interaction with smoking phenotypes in the COBLAnCE cohort.
PhenotypersIDChrPositionLocusREFALTMAFOR (95% CI)β (SE)p-valueAnnotation (Distance)Type
Smoking status (Ever vs. Never)rs1140736361620663401p31.3GA0.040.31 (0.20–0.47) 8.68 × 10−8NFIA (137880), MGC34796 (53574)intergenic
rs1165716081620626941p31.3GA0.040.31 (0.20–0.47) 8.87 × 10−8NFIA (134234), MGC34796 (57220)intergenic
rs211004011878420471q31.1GC0.010.17 (0.08–0.36) 1.88 × 10−6LINC01037 (395693)intergenic
rs5425416274900484664q22.1AAAAAACAAACAAAC0.441.69 (1.39–2.06) 1.81 × 10−7TIGD2 (12414), GPRIN3 (109068)intergenic
rs15332944900481224q22.1CT0.421.63 (1.34–1.98) 1.01 × 10−6TIGD2 (12070), GPRIN3 (109412)intergenic
rs802813694558567074q12TC0.040.36 (0.24–0.55) 1.91 × 10−6KIT (249826), KDR (87941)intergenic
rs1109841941189399534q26TG0.301.62 (1.33–1.98) 2.30 × 10−6LINC02264 (148850), NDST3 (15688)intergenic
rs11531751541292232514q28.2CT0.020.29 (0.17–0.49) 2.64 × 10−6PGRMC2 (14283), LINC02615 (125920)intergenic
rs7626140671104923017q31.1TG0.030.23 (0.13–0.42) 2.39 × 10−6IMMP2Lintronic
rs111121821210513911612q23.3AG0.190.57 (0.45–0.71) 1.16 × 10−6CHST11intronic
Age at smoking initiation (Years)rs53175644912444605771q44CG0.02 0.3 (0.05)8.26 × 10−9ZBTB18 (239797), C1orf100 (55360)intergenic
rs14053857131496544603q25.1CA0.01 −0.24 (0.05)1.04 × 10−6RNF13intronic
rs11542108131496008073q25.1CG0.01 −0.24 (0.05)1.05 × 10−6RNF13intronic
rs148961658941117669p24.2GC0.01 0.26 (0.05)5.03 × 10−7GLIS3intronic
rs117818261116264421411q12.3CT0.01 0.21 (0.04)8.13 × 10−7SLC3A2intronic
rs77186197121431638312p13.1CT0.01 0.29 (0.05)3.74 × 10−9GRIN2B (183093), ATF7IP(202183)intergenic
rs78947799121441581112p13.1AC0.01 0.31 (0.06)3.97 × 10−8GRIN2B (282521), ATF7IP (102755)intergenic
rs79782126121429462112p13.1TC0.02 0.18 (0.03)2.76 × 10−7GRIN2B (161331), ATF7IP (223945)intergenic
rs149790626177271653017q25.1CT0.01 0.35 (0.06)9.76 × 10−8RAB37intronic
rs34177209191847497819p13.11TA0.26 0.06 (0.01)1.57 × 10−6PGPEP1UTR3
Average tobacco consumption (grams)rs11468193011600467121q23.2GT0.01 −0.7 (0.14)2.39 × 10−7KCNJ10 (6762), KCNJ9 (4616)intergenic
rs7975246811600508671q23.2TA0.01 −0.7 (0.14)2.44 × 10−7KCNJ9 (461)upstream
rs14272815112481661731q44GA0.02 −0.58 (0.12)5.94 × 10−7OR2L13intronic
rs1146345072192472302p24.1GA0.02 −0.65 (0.13)3.44 × 10−7LINC01376 (20504), MIR4757 (300960)intergenic
rs1492461422192922232p24.1GA0.02 −0.67 (0.13)4.01 × 10−7LINC01376 (65497), MIR4757 (255967)intergenic
rs27140691112338961411q24.1AG0.03 0.63 (0.11)1.35 × 10−8GRAMD1Bintronic
rs11368338016769253416p13.3GA0.02 −0.66 (0.12)6.44 × 10−8RBFOX1intronic
rs11359062416768608916p13.3GC0.02 −0.64 (0.12)2.74 × 10−7RBFOX1intronic
rs146671367187214612418q22.3CT0.01 −0.92 (0.18)6.89 × 10−7DIPK1C (21621), CNDP2 (17474)intergenic
rs151292117187214547318q22.3GA0.01 −0.94 (0.19)8.84 × 10−7DIPK1C (20970), CNDP2 (18125)intergenic
ALT, alternative allele; CI, confidence interval; COBLAnCE, COhort to study BLAdder CancEr; GWAS, genome-wide association study; MAF, minor allele frequency; REF, reference allele; OR, odds ratio; rsID, SNP identification number; SNP, single nucleotide polymorphism. SNPs that are significant at the GWAS threshold are presented in bold.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Karimi, M.; Mendez-Pineda, S.; Blanché, H.; Boland, A.; Besse, C.; Deleuze, J.-F.; Meng, X.-Y.; Sirab, N.; Groussard, K.; Lebret, T.; et al. A Case-Only Genome-Wide Interaction Study of Smoking and Bladder Cancer Risk: Results from the COBLAnCE Cohort. Cancers 2023, 15, 4218. https://doi.org/10.3390/cancers15174218

AMA Style

Karimi M, Mendez-Pineda S, Blanché H, Boland A, Besse C, Deleuze J-F, Meng X-Y, Sirab N, Groussard K, Lebret T, et al. A Case-Only Genome-Wide Interaction Study of Smoking and Bladder Cancer Risk: Results from the COBLAnCE Cohort. Cancers. 2023; 15(17):4218. https://doi.org/10.3390/cancers15174218

Chicago/Turabian Style

Karimi, Maryam, Sebastian Mendez-Pineda, Hélène Blanché, Anne Boland, Céline Besse, Jean-François Deleuze, Xiang-Yu Meng, Nanor Sirab, Karine Groussard, Thierry Lebret, and et al. 2023. "A Case-Only Genome-Wide Interaction Study of Smoking and Bladder Cancer Risk: Results from the COBLAnCE Cohort" Cancers 15, no. 17: 4218. https://doi.org/10.3390/cancers15174218

APA Style

Karimi, M., Mendez-Pineda, S., Blanché, H., Boland, A., Besse, C., Deleuze, J. -F., Meng, X. -Y., Sirab, N., Groussard, K., Lebret, T., Bonastre, J., Allory, Y., Radvanyi, F., Benhamou, S., & Michiels, S. (2023). A Case-Only Genome-Wide Interaction Study of Smoking and Bladder Cancer Risk: Results from the COBLAnCE Cohort. Cancers, 15(17), 4218. https://doi.org/10.3390/cancers15174218

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop