Polymorphisms in GSTT1, GSTM1, NAT1 and NAT2 genes and bladder cancer risk in men and women.

BACKGROUND
Cigarette smoking is an established risk factor for bladder cancer. Epidemiological and biological data suggest that genetic polymorphisms in activating and detoxifying enzymes may play a role in determining an individual's susceptibility to bladder cancer in particular when in combination with specific environmental exposures such as cigarette smoking. N-acetyltransferase (NAT) enzymes, NAT1 and NAT2, are involved in the activation and detoxification of tobacco smoke constituents. Polymorphisms in these genes alter the ability of these enzymes to metabolize carcinogens, as certain allelic combinations result in a slow or rapid acetylation phenotype. Glutathione S-transferases (GSTs) also detoxify tobacco smoke constituents, and polymorphisms within the GSTM1 and GSTT1 genes can result in a complete lack of enzyme activity.


METHODS
We assessed the association between common polymorphisms identified in the GSTM1, GSTT1, NAT1, and NAT2 genes and the risk of bladder cancer in two nested case-control studies within the Nurses' Health Study (n = 78 female cases, 234 female controls) and the Health Professionals' Follow-up Study (n = 139 male cases, 293 male controls). We also evaluated whether cigarette smoking modified the associations of the genotypes and bladder cancer risk in men and women.


RESULTS
Overall, we observed no statistically significant associations between the polymorphisms and bladder cancer risk among men and women, although given our sample size, we had limited power to detect small to moderate effects. There was however the suggestion of an increased risk among female ever smokers with the NAT2 slow genotype and an increased risk in male never smokers with the GSTM1 null genotype.


CONCLUSION
In summary, these prospective results are consistent with previous literature supporting associations between bladder cancer and the NAT2 slow acetylation and the GSTM1 null genotypes.

Glutathione S-transferases (GSTs) are involved in the metabolism of environmental carcinogens, reactive oxygen species, and chemotherapeutic agents by catalyzing reactions between glutathione and electrophilic compounds [21,22]. Glutathione S-transferase M1 (GSTM1) is involved in the detoxification of polycyclic aromatic hydrocarbons (PAHs) and benzo(a)pyrene found in tobacco smoke [23,24]. Glutathione S-transferase T1 (GSTT1) is involved in activation and detoxification reactions and catalyzes the conjugation of industrial chemicals, e.g. ethylene oxides, with glutathione [25]. Homozygous deletions of the GSTM1 and GSTT1 genes are common and result in a complete loss of enzyme activity [22,26,27]. Because the metabolism of tobaccorelated carcinogens may be influenced by the activity of GSTs, polymorphisms in GSTM1 and GSTT1 may modify the risk of bladder cancer associated with these carcinogens. The GSTM1 null polymorphism has been consistently associated with an increased risk of bladder cancer in pooled and meta-analyses [11,23,28]. Studies investigating the importance of GSTT1 in bladder carcinogenesis are more limited and inconsistent [19,24,[29][30][31][32][33][34][35].
Since exposure to tobacco smoke carcinogens is a risk factor for bladder cancer, genetic modulation of carcinogen metabolism may explain some inter-individual bladder cancer susceptibility. Given their biological role, polymorphisms within the GSTT1, GSTM1, NAT1, and NAT2 genes may be important in determining an individual's susceptibility to bladder cancer. Recent studies have also suggested that although the risk associated with each variant may be small, in combination with other genetic and/ or environmental factors, the effects of the polymorphisms may be increased. We examined the associations between polymorphisms within these key genes and bladder cancer risk in a nested case-control study of women participating in the Nurses' Health Study (NHS) and a nested case-control study of men participating in the Health Professionals' Follow-Up Study (HPFS). Furthermore, we assessed the potential for effect modification by cigarette smoking for the associations between the genotypes and bladder cancer risk.

Nurses' Health Study
The NHS began in 1976 when 121,700 female United States registered nurses aged 30-55 years completed a selfadministered questionnaire. Detailed information on individual characteristics and behaviors was obtained from the questionnaires at baseline and biennially thereafter. Between 1989 and 1990, blood samples were collected from 32,826 women; follow-up for this subcohort exceeds 96%. The methods used for blood collection have been previously detailed [36].

Case-control study
In our study, we included both incident and prevalent bladder cancer cases from the NHS blood cohort. Trained physicians who were blinded to exposure information reviewed hospital records and pathology reports. We were able to confirm approximately 95% of self-reported bladder cancer cases. Eligible incident cases consisted of women with confirmed bladder cancer that had been diagnosed anytime after blood collection (1989-90) and up to June 1, 2000, with no previously diagnosed cancer except for nonmelanoma skin cancer. Prevalent cases were defined as having confirmed bladder cancer diagnosed after questionnaire return in 1976 and before the date of blood collection, with no previously diagnosed cancer except for nonmelanoma skin cancer. Controls were randomly selected participants who had given a blood sample and were free of diagnosed cancer (except nonmelanoma skin cancer) up to and including the interval in which the case was diagnosed. Controls were matched to cases 3:1 according to year of birth, smoking status at blood draw (never versus ever), month of blood collection, and fasting status at blood draw (for possible plasma hormone analyses). This case-control study con-sists of 78 bladder cancer cases (47 incident cases and 31 prevalent cases), and 234 matched controls. In addition, 2646 women who were controls in either a nested casecontrol study of breast cancer (n = 1686) [37,38], colon cancer (n = 461) [39], and colon polyps (n = 499), were free of cancer other than nonmelanoma skin cancer, and were genotyped for the three NAT2 polymorphisms were included in unconditional logistic regression analyses to stabilize confidence intervals. A subset of the extra controls (n = 466) from the breast cancer nested case-control study [40] were genotyped for the GSTM1 and GSTT1 polymorphisms and included in the unconditional logistic regression analyses. The study protocol was approved by the Committee on Use of Human Subjects of the Brigham and Women's Hospital, Boston, MA. Completion of the self-administered questionnaire and receipt of the blood sample were considered to imply informed consent.

Health professionals' follow-up study
The HPFS is an ongoing prospective study of the causes of chronic diseases in men. The cohort began in 1986 when 51,529 U.S. male health professionals aged 40 to 75 years responded to a mailed questionnaire [41]. These men provided detailed information on medical histories and health-related exposures at baseline and biennial questionnaires thereafter. The follow-up for these participants is approximately 93%. Blood samples were collected between 1993 and 1995 from 18,025 participants. Approximately 95% of the samples arrived within 24 hours of blood draw.

Case-control study
In our study, we included both incident and prevalent bladder cancer cases from the HPFS blood cohort. All cases in this analysis were confirmed through histopathologic reports reviewed by a study investigator. Eligible incident cases consisted of men with confirmed bladder cancer that had been diagnosed anytime after blood collection (1993-95) and up to June 1, 2000, with no previously diagnosed cancer except for nonmelanoma skin cancer. Prevalent cases were defined as having confirmed bladder cancer diagnosed after questionnaire return in 1986 and before the date of blood collection, with no previously diagnosed cancer except for nonmelanoma skin cancer. All controls were randomly selected participants who had given a blood sample and were free of diagnosed cancer (except nonmelanoma skin cancer) up to and including the interval in which the case was diagnosed. Controls were matched to cases 2:1 according to year of birth, smoking status at blood draw (current, past, or never), fasting status (for possible plasma hormone analyses), and month of sample collection. For cases accrued from 1998 to 2000, controls were matched to cases 3:1 with the previously described matching factors to increase our statistical power. This case-control study consists of 139 bladder cancer cases (91 incident cases and 48 prevalent cases), and 293 matched controls. In addition, 1028 men who were controls in either a nested case-control study of colon cancer (n = 321) or colon polyps (n = 707), were free of cancer other than nonmelanoma skin cancer, and were also genotyped for the three NAT2 polymorphisms were included in unconditional logistic regression analyses. Completion of the self-administered questionnaire and receipt of the blood sample were considered to imply informed consent.

Sample collection
Venous blood samples were separated into plasma, buffy coat, and red blood cells and stored in liquid nitrogen. Genomic DNA was extracted from 50 μL buffy coat diluted with 150 μL of PBS using the QIAGEN QIAmp (Qiagen, Inc., Chatsworth, CA) 96-spin blood protocol according to the manufacturer's instructions. Genomic DNA concentrations were calculated in 96-well format using PicoGreen technology (Molecular Probes, Eugene, OR).
Genotypes for the GSTT1 and GSTM1 deletions were determined by polymerase chain reaction (PCR) using a previously published protocol [40]. In both assays, the absence of the PCR product was indicative of the null genotype; these assays do not distinguish between heterozygous or homozygous wildtype genotypes.
All genotyping was performed by laboratory personnel blinded to case-control status, and blinded quality control samples were inserted to validate genotyping identification procedures; concordance for blinded samples was 100%.

Statistical analysis
We used a χ 2 test to assess whether the genotypes were in Hardy-Weinberg equilibrium and to determine P values for differences in genotype frequencies between cases and controls. The associations between the genotypes and bladder cancer risk were examined by using conditional and unconditional logistic regression to calculate odds ratios (ORs) and 95% confidence intervals (CIs). Unconditional logistic regression models assessing the main effect of the genotype on bladder cancer risk included the additional control groups to maximize the sample size, stabilize CIs, and to increase the power to detect meaningful associations. In addition to the matching factors, analyses were also adjusted for packyears of smoking (i.e. average reported number of cigarette packs smoked per day multiplied by the number of years of smoking). Inclusion of additional factors, e.g. geographic region, alcohol intake, cruciferous vegetable intake, did not substantially change the point estimates, and therefore these factors were not included in the final models. To test statistical interactions between the genotypes and cigarette smoking in unconditional models, we first used a likelihood-ratio test (LRT) to compare nested models that included terms for all combinations of the genotypes and cigarette smoking status to the models with indicator variables for the main effects only (nominal LRT). To test the statistical interaction between the various genotypes, a similar approach was employed. All p-values are two-sided. We used the SAS Version 8.2 software (SAS Institute, Cary, NC). Because the allele frequencies of NAT1, NAT2, GSTM1, and GSTT1 differ by ethnicity [7,23,24,45,46], all analyses were restricted to Caucasians. Numbers may vary for the different analyses due to missing genotype data. We had 80% power to detect the following ORs for the men: OR = 1.8 for the GSTM1 null genotype assuming a genotype prevalence of 50%; OR = 2.0 for the GSTT1 null genotype assuming a genotype prevalence of 15%; OR = 1.9 for the NAT2 slow acetylation genotype assuming a genotype prevalence of 62%; OR = 1.8 for the NAT1*10 carrier assuming a genotype prevalence of 35%. We had 80% power to detect the following ORs for the women: OR = 2.1 for the GSTM1 null genotype assuming a geno-type prevalence of 50%; OR = 2.4 for the GSTT1 null genotype assuming a genotype prevalence of 15%; OR = 2.3 for the NAT2 slow acetylation genotype assuming a genotype prevalence of 62%; OR = 2.1 for the NAT1*10 carrier assuming a genotype prevalence of 35%.

Population characteristics
Our study population included 78 bladder cancer cases and 234 matched controls from the NHS and 139 bladder cancer cases and 293 matched controls from the HPFS for a total of 217 bladder cancer cases and 526 matched controls. The risk factor profile for female bladder cancer cases and controls were similar with no statistically significant differences between cases and controls (Table 1). Similarly, no significant differences in risk factor distributions were observed in cases and controls in the HPFS (Table 1). Risk factor distributions between incident and prevalent cases in the NHS and incident and prevalent cases in the HPFS were also similar. A comparison of the population characteristics of the second control group with the first control group revealed no material differences in risk factor profiles in the NHS and HPFS.
We compared the distributions of the GSTM1, GSTT1, NAT1, and NAT2 genotypes between incident and prevalent bladder cancer cases in men and women separately to determine whether the variant alleles were associated with survival. All of the genotype frequencies were similar among incident and prevalent cases in the NHS and in the HPFS (p > 0.2). Therefore, all incident and prevalent cases in each cohort were combined to form one case group for all statistical analyses. The genotype frequencies were in Hardy-Weinberg equilibrium for cases and controls. The allele frequencies among the controls for the GSTM1 null allele, GSTT1 null allele, NAT2 rapid genotype, and the NAT1 rapid genotype were similar to previously published reports in Caucasians [10,18,23,26,38,46,47]. Results from the conditional logistic regression analyses using only the matched dataset were similar to the results obtained from the unconditional analyses that included the additional controls.

Nurses' Health Study
In the NHS, the prevalence of GSTM1 null genotype was similar among cases and controls (48.4% versus 52.5%, respectively, p = 0.54), in contrast to the GSTT1 null genotype, which was more prevalent in cases than controls (21.9% vs. 15.2%, respectively, p = 0.16). After adjusting for the matching factors and packyears of smoking, we observed no association between the GSTM1 and GSTT1 polymorphisms and bladder cancer risk in women ( Table  2). We did not observe any statistically significant effect modification by cigarette smoking status for the GSTM1 and the GSTT1 null genotypes and bladder cancer risk (p for interaction > 0.30). Our results also do not support a gene-gene interaction between the GSTT1 and GSTM1 polymorphisms (p for interaction = 0.88).
The frequency of the NAT2 slow acetylation genotype was similar among cases and controls (68.3% versus 62.1%, respectively, p = 0.32) ( Table 3). The adjusted OR for the slow acetylation genotype compared to the rapid acetylation genotype was 1.33 (95% CI, 0.77-2.31). After stratification by cigarette smoking status, the association between NAT2 slow acetylation genotype and bladder cancer risk was stronger but not statistically significant among ever smokers (OR = 1.76; 95% CI, 0.88-3.52) (p for interaction = 0.15). We observed no significant association between bladder cancer risk and individuals carrying the NAT1*10 allele (Table 3), and no effect modification by cigarette smoking status for the NAT1 genotypes and bladder cancer risk (p for interaction > 0.10).

Health professionals' follow-up study
There were no significant differences in GSTM1 and GSTT1 genotype frequencies among cases and controls (p = 0.09 and 0.63, respectively), and no statistically significant associations between the GSTM1 and GSTT1 polymorphisms and bladder cancer risk in males (Table 2). We did observe a statistically significant interaction between the GSTM1 null genotype, cigarette smoking, and bladder cancer risk in males (p for interaction = 0.02). Among never smokers, the GSTM1 null genotype was associated with a statistically significant increased risk (OR = 3.23 (95% CI, 1.38, 7.58)), and among ever smokers, the OR was 0.96 (95% CI, 0.55-1.66). The association between GSTT1 polymorphism and bladder cancer risk remained similar after stratification by smoking status (p for interaction = 0.69), and no significant gene-gene interactions with GSTM1 and GSTT1 were detected (p for interaction = 0.22).
The frequency of the NAT2 slow acetylation genotype was similar among cases and controls (57.3% versus 61.8%, p = 0.32) ( Table 3). The adjusted OR for the slow acetylation genotype compared to the rapid acetylation genotype was 0.78 (95% CI, 0.53-1.15). We observed no significant association between bladder cancer risk and carrying the NAT1*10 allele (Table 3). Furthermore, we did not observe any statistically significant effect modification by cigarette smoking status for the NAT2 or NAT1 genotypes and bladder cancer risk (p for interaction > 0.4).

Discussion
Evidence suggests that polymorphisms in activating and detoxifying enzymes may interact to effect the level of DNA damage sustained by a specific tissue and ultimately influence disease risk [47]. Therefore, imbalances between activation and detoxification processes may result in an increase in bladder cancer risk due to the accumulation of carcinogen metabolites, e.g. from cigarette smoking. Overall, in our population-based case-control study, we observed no statistically significant associations between the GSTT1, GSTM1, NAT1, and NAT2 polymorphisms and bladder cancer risk.
Aromatic amines within tobacco smoke are the most important class of bladder carcinogens [48]. These compounds may be partially responsible for the increased bladder cancer risk observed among slow NAT2 acetylators who have a decreased capacity to detoxify aromatic amines. Our data suggest an association between the NAT2 slow acetylation genotype and increased bladder cancer risk in female smokers. Our findings are biological plausible and are consistent with a recent study in Spain [11], which found that NAT2 slow acetylators have an increased risk of bladder cancer that was stronger for cigarette smokers compared to never smokers. NAT1 polymorphisms may also affect individual bladder cancer risk by interacting with cigarette smoking. Our findings do not support an association between the NAT1 polymorphisms, cigarette smoking, and bladder cancer; however we had little power to detect such an association. Additional studies that have assessed the relationship between NAT1*10 allele and bladder cancer risk have been inconsistent with some studies observing an increased risk [17,20], a decreased risk [15], or no association [11,16,18,19]. Larger epidemiological studies are required to clarify the relationship between bladder cancer risk and the NAT1*10 allele.
We observed no significant associations with the GSTM1 polymorphism and bladder cancer risk, which is in contrast with previous findings [11,23]. The potential role of the GSTT1 polymorphism on bladder cancer susceptibility is less certain. Studies that have addressed the role of GSTT1 polymorphism and bladder cancer have been inconsistent. Of the published studies, some suggest an increased risk [19,31,34,35] with the GSTT1 null genotype, others no association [11,29,30], and only one suggests a decreased risk [24]. We observed no significant association with the GSTT1 polymorphism and bladder cancer risk.
The strengths of our study include the prospective design and the detailed collection of exposure information (i.e. cigarette smoking) prior to bladder cancer diagnosis. We were also able to formally assess and test gene-environment interactions. However, our study has a small sample size resulting in low power to detect minor to modest genotype-disease associations and gene-environment interactions, therefore such associations cannot be ruled out.

Conclusion
In summary, our data support the prior evidence that the GSTM1 null genotype and the NAT2 slow acetylation genotype may affect an individual's bladder cancer risk.

Competing interests
The author(s) declare that they have no competing interests.

Authors' contributions
MM carried out the data analysis and wrote the manuscript. DM and ID conceived of the study, participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.