Abstract
Lung cancer remains a leading cause of disease globally, with smoking being the largest single cause. Phase I enzymes, including cytochrome P450, family 1, subfamily A, polypeptide 1 (CYP1A1), are involved in the activation of carcinogens, such as polycyclic aromatic hydrocarbons, to reactive intermediates that are capable of binding covalently to DNA to form DNA adducts, potentially initiating the carcinogenic process. The aim of the present study was to investigate the association of CYP1A1 gene polymorphisms and haplotypes with lung cancer risk.
A case–control study was carried out on 1,040 nonsmall cell lung cancer (NSCLC) cases and 784 controls to investigate three CYP1A1 variants, CYP1A1*2A (rs4646903; thymidine to cytosine substitution at nucleotide 3801 (3801T>C)), CYP1A1*2C (rs1048943; 2455A>G; substitution of isoleucine 462 with valine (exon 7)) and CYP1A1*4 (rs1799814; 2453C>A; substitution of threonine 461 with asparagine (exon 7)) using PCR restriction fragment length polymorphism methods.
The CYP1A1*2A and CYP1A1*2C variants were significantly over-represented in NSCLC cases compared with controls, whereas the CYP1A1*4 variant was under-represented. CYP1A1 haplotypes (in allele order CYP1A1*4, CYP1A1*2C, CYP1A1*2A) CGC and CGT were associated with an increased risk of lung cancer, whereas AAT was associated with decreased lung cancer risk in this population.
The present study has identified risk haplotypes for CYP1A1 in NSCLC and confirmed that CYP1A1 polymorphisms are a minor risk factor for NSCLC.
Lung cancer is a leading cause of cancer death internationally, with smoking being the largest single cause. Smoking is responsible for 85–90% of lung cancers 1, yet <20% of lifelong smokers develop lung cancer, suggesting that other factors, including genetics, may play a role 1. Phase I enzymes (mainly cytochrome P450) metabolically activate carcinogens, such as polycyclic aromatic hydrocarbons and N-nitrosamines, to reactive intermediates 2. These intermediates are capable of binding covalently to DNA to form DNA adducts, potentially initiating the carcinogenic process. Two functionally important nonsynonymous polymorphisms have been described for the cytochrome P450, family 1, subfamily A, polypeptide 1 (CYP1A1) gene, a base substitution at codon 462 in exon 7, resulting in substitution of isoleucine with valine (Ile462Val (exon 7)) (CYP1A1*2C; National Center for Biotechnology Information single nucleotide polymorphism (SNP) identifier rs1048943; adenine (A) to guanine (G) substitution at nucleotide 2455 (2455A>G)) and a point mutation (thymine (T) to cytosine (C)) at the MspI site in the 3’-untranslated region (CYP1A1*2A; rs4646903; 3801T>C) 3, 4. A third polymorphism, substituting threonine (Thr) for asparagine (Asn) (CYP1A1*4; rs1799814; 2453C>A; Thr461Asn (exon 7)), has also been reported two bases upstream of CYP1A1*2C; however, its functional effects remain to be fully elucidated.
Ethnic differences in the distribution of the CYP1A1*2C and CYPA1*2A genotypes have been reported in lung cancer subjects, with few reports available for CYP1A1*4. An over-representation of the Val allele (CYP1A1*2C) among lung cancer cases has been reported in Asian and Caucasian populations 5–7. Although relatively frequent in Asian populations (0.18–0.25) 8, 9, the Val allele is quite rare in Caucasian control populations, occurring in ∼7–13% of people 7, 10. In a previous study, no interactive effects were shown between CYP1A1*2C and polymorphisms of the phase II enzyme genes encoding glutathione S-transferase pi 1 (GSTP1), theta 1 (GSTT1) or mu 1 (GSTM1). However, risk genotypes of the myeloperoxidase gene (MPO) and CYP1A1*2C interacted to increase the overall risk of nonsmall cell lung cancer (NSCLC) (odds ratio (OR) 2.88; 95% confidence interval (CI) 1.70–5.00; p = 0.001) 11. Similarly, the CYP1A1*2A variant has been strongly associated with increased lung cancer risk in Asian populations, especially in relation to tobacco smoking and in combination with polymorphisms of the phase II enzyme genes GSTM1 and GSTT1 12, 13. In contrast, studies in Caucasian populations have not clearly established an association between CYP1A1 polymorphisms and increased lung cancer risk 14–19. Relatively frequent incidences of CYP1A1*2A variants have been reported in Asian (39%) 20 and Caucasian control populations (18–23%) 14, 16, 21. Conversely, few studies have reported population frequencies for CYP1A1*4 in lung cancer subjects. Cascorbi et al. 6 reported variant allele frequencies of 3% in Caucasian lung cancer cases and controls. In contrast, Song et al. 13 detected no polymorphic sites in an Asian population, supporting similar polymorphism studies of Japanese liver subjects 10. Similar frequencies (2–3%) have been observed in healthy Caucasian populations 22, 23. A recent pooled analysis of Hung et al. 24 reported an increase in lung cancer susceptibility by more than two-fold for CYP1A1*2A and CYP1A1*2C in nonsmokers. The authors suggested these polymorphisms may be implicated in lung carcinogenesis at low levels of tobacco exposure, possibly in combination with phase II enzymes, including GSTM1 24, although this has not been replicated by other groups 21.
Many studies investigating the relationships between CYP1A1 polymorphisms and lung cancer have been limited by small sample numbers, leading to a lack of statistical power. Pooled analyses to increase sample size have tried to address this issue. Conflicting results between groups may be due to population differences (i.e. ethnicity) or failure to control for other potential confounders, including age and sex. CYP1A1 haplotype studies have potential in determining whether combinations of CYP1A1 (CYP1A1*2A, CYP1A1*2C, CYP1A1*4) together confer a greater risk of lung cancer than single polymorphisms. One recent study of 200 case-matched controls from an Indian population, showed that only one haplotype, CGC (3801C, 2455G, 2453C), was significantly associated with increased lung cancer risk, although the study was limited by its small sample numbers 3. Few CYP1A1 haplotype association studies have been conducted in a Caucasian population. A study of Han et al. 25 investigated the genotype frequencies of 13 SNPs found in the promoter region of CYP1A1 in 21 Caucasian individuals. Subsequent functional studies identified two CYP1A1 haplotypes (2923C>T, 2875G>A, 3777T>G; and 2923C>T, 3777T>G, 4553G>A) demonstrating moderate increases in basal activity compared with the wild-type CYP1A1 constructs (1.38 and 1.50, respectively; p<0.05). These were considered unlikely to be of functional significance considering the magnitude of difference in CYP1A1 expression in response to benzo[a]pyrene and cigarette smoke extract 25.
In order to further investigate the role of CYP1A1 polymorphism variation (CYP1A1*2C, CYP1A1*2A and CYP1A1*4) in lung cancer risk, CYP1A1 haplotype analyses were performed in a large sample of Australian lung cancer cases with the aim of identifying risk-modifying CYP1A1 haplotypes.
MATERIALS AND METHODS
Study population
The present study population has been described previously 11, 26. Cases were subjects with confirmed primary lung cancer treated at The Prince Charles Hospital (Brisbane, Australia) during 1980–2007 (n = 1,040). Controls consisted of subjects with chronic obstructive pulmonary disease (COPD) treated at the same hospital (n = 506) or healthy smokers (n = 278) who, at the time of recruitment to the study, did not have doctor-diagnosed lung cancer. The ethnicity of the study population was >99% Caucasian. The study was approved by the Ethics Committee at The Prince Charles Hospital. Cases and controls gave informed written consent for use of resected lung tissue or peripheral blood. The demographics of cases and controls were checked by a research nurse or the treating physician against patient medical charts or the hospital lung cancer database (table 1⇓).
Demographics of cases and controls using the PCR restriction fragment length polymorphism method
Sample preparation and genotyping
DNA from cases diagnosed with NSCLC was extracted from peripheral blood or resected cryopreserved normal lung tissue as described previously 26. Cases and controls from a previous study of CYP1A1*2C were included 11, 26. DNA from control subjects was extracted from peripheral blood. In 592 cases, DNA was extracted from more than one source (blood lymphocytes and fresh frozen normal lung tissue). In these cases, both sources were genotyped, with identical results in all cases, reinforcing the reproducibility of the methods. PCR-based restriction fragment length polymorphism (RFLP) methods were used to analyse CYP1A1*2A 27 and CYP1A1*4 6 polymorphisms. Approximately 10% of the samples were randomly selected for repeat genotyping by PCR-RFLP in order to test reproducibility.
DNA sequencing and single-base extension genotyping
In order to confirm the accuracy of the PCR-RFLP methods, two representative samples per genotype were confirmed by DNA sequencing for each polymorphism. Samples were purified using the Wizard PCR Cleanup and Gel extraction kit (Promega, Madison, WI, USA) and sequenced at the Australian Genome Research Facility (AGRF, Brisbane, Australia) for DNA sequencing using BigDye Terminator V3.1 chemistry (Applied Biosystems, Foster City, CA, USA). The sequencing primers were identical to those used in PCR amplification. Sequences were visualised using Chromas V1.4 (C. McCarthy, School of Biomolecular and Physical Sciences, Griffith University, Brisbane, Australia).
In addition, an independent genotyping method, iPLEX single-base extension (Sequenom, San Diego, CA, USA), was also used to genotype a subset (89%) of subjects genotyped by PCR-RFLP methods. All experiments were performed by the AGRF. Briefly, genomic DNA was amplified using primers with 10-mer tags, designed to amplify a 75–150-bp amplicon. Following this, shrimp alkaline phosphatase was added to each reaction in order to remove any unincorporated deoxyribonucleoside triphosphates (dNTPs) by cleavage of phosphates from dNTP groups. An iPLEX reaction master mix consisting of primer, enzyme, buffer and mass-modified nucleotides was then added and the samples placed in a thermocycler in order to permit addition of nucleotides to the polymorphic site, producing allele-specific base extension products of differing sizes. Products were then run on a matrix-assisted laser desorption/ionisation time-of-flight mass spectrometer in order to determine product size.
Statistical analysis
Distributions of genotypes and demographic variables were compared between cases and controls using Chi-squared tests for categorical outcome variables and two-sided unpaired t-tests for continuous outcome variables. ORs and 95% CIs were estimated in order to measure the association between lung cancer and genotype/haplotype frequency. Standard Chi-squared statistics were used to determine whether or not the three CYP1A1 variants were in Hardy–Weinberg equilibrium (HWE). All statistical analyses for genotypes were performed using the SPSS software package (Version 13.0 for Windows; SPSS, Inc., Chicago, IL, USA). Haplotype analyses (haplotype frequency estimates and linkage disequilibrium) were carried out using Haploview linkage software (Version 4.0) 28. A p-value of <0.05 (two-tailed) was considered significant. Power calculations to evaluate ability to detect associations between NSCLC and CYP1A1 variants among the present study population were determined using an α of 0.05 and 80% power. The present study had the power to detect ORs of 1.39, 1.48 and 1.69 for the variant alleles of CYP1A1*2A, CYP1A1*4 and CYP1A1*2C, respectively, with 80% confidence at an α of 0.05.
RESULTS
Participant characteristics
In order to ensure that observed effects between cases and controls were due to genotype frequency and not other potential confounding factors, such as age, smoking history or sex, mean and frequency distributions were compared between cases and controls (table 1⇑). Mean age differed significantly between cases and controls (p<0.001), with controls being slightly younger. In contrast, there were no significant differences in sex distribution (p = 0.81) or smoking history in pack-yrs (p = 0.47), excluding these factors as potential study confounders. The majority of NSCLC cases were adenocarcinoma or squamous cell carcinoma histological subtypes (39 and 36%, respectively).
In order to confirm the validity of combining COPD subjects and healthy smokers to form one control group, as done previously 26, Chi-squared tests were used to investigate whether or not there were significant differences in genotype frequency between the two groups. No significant difference in variant frequency was observed between COPD and healthy smokers (p>0.05; data not shown), validating the decision to combine these two subgroups.
PCR restriction fragment length polymorphism
Genotype frequencies for all three variants, determined by PCR-RFLP methods, are detailed in table 2⇓. Variant allele frequencies were low in the control population for all of the SNPs. The minor allele frequencies for cases and controls combined were 4.6, 4.5 and 11.4% for CYP1A1*4, CYP1A1*2C and CYP1A1*2A, respectively. Since less than five samples were classified as homozygous for the variant allele in either cases or controls for CYP1A1*2C or CYP1A1*4, homozygous variants were combined with heterozygous genotypes for statistical analysis (table 3⇓). In the present study, genotypes containing the variant allele of CYP1A1*2A occurred in 17.6% of the control population, CYP1A1*4 in 11% and CYP1A1*2C in 5.5%. These values were in relative agreement with previous studies of Caucasian populations (CYP1A1*2A 18–23%; CYP1A1*4 2–3% and CYP1A1*2C 9–10%) 14, 16, 21–23. In the present study population, the genotype frequencies of CYP1A1*2A and CYP1A1*4 were in HWE for cases and controls; however, CYP1A1*2C was not (table 4⇓). Sanger DNA sequencing of two samples per genotype was performed for each polymorphism in order to confirm the accuracy of the PCR-RFLP assays.
Distribution of cytochrome P450 1A1 (CYP1A1) genotypes amongst nonsmall cell lung cancer subjects and controls#
Distribution of cytochrome P450 1A1 (CYP1A1) genotypes and haplotypes amongst nonsmall cell lung cancer (NSCLC) subjects and controls#
Hardy–Weinberg Chi-squared tests for nonsmall cell lung cancer (NSCLC) subjects and controls
From the PCR-RFLP data, it was observed that subjects carrying the CYP1A1*2A variant were over-represented in NSCLC cases compared with controls (TT versus CT/CC; OR 1.43 (95% CI 1.35–1.51); p = 0.003). A decreased risk of NSCLC was also observed for those with the CYP1A1*4 variant (CC versus CA/AA; OR 0.64 (95% CI 0.62–0.64); p = 0.005). The wild-type genotype (CC) was slightly over-represented in cases versus controls (92.8 versus 89%). Data from a previous study on CYP1A1*2C were included in the present study. The variant allele was associated with a greater than two-fold increase in the risk of lung cancer.
In order to investigate associations between the three polymorphisms, haplotype analyses were performed to determine whether combinations of polymorphisms conferred a greater lung cancer risk. Five possible haplotypes were identified amongst cases and the larger control group, where the CAT haplotype was considered to be the wild-type (in allele order CYP1A1*4, CYP1A1*2C, CYP1A1*2A). The frequencies of CAT, CAC, AAT, CGC and CGT are presented in table 3⇑. Two haplotypes were significantly over-represented in NSCLC subjects versus controls, CGC (4.1 versus 2.4%; p = 0.0038; OR 1.76 (95% CI 1.19–2.60)) and CGT (1.4 versus 0.4%; p = 0.0026; OR 3.26 (95% CI 1.46–7.28)). Conversely, AAT was significantly under-represented in cases versus controls (3.4 versus 5.4%; p = 0.0014; OR 0.62 (95% CI 0.45–0.86)). CAC showed a general increase in lung cancer risk, although this did not reach significance (8.5 versus 6.9%; p = 0.0747; OR 1.25 (95% CI 0.98–1.60)). It was found that CYP1A1*2A and CYP1A1*2C were in linkage disequilibrium (normalised disequilibrium constant (D’) 0.749; logarithm of odds (logit; LOD) 66.5; r2 = 0.207; 95% CI 0.67–0.81) but that CYP1A1*4 and CYP1A1*2C (D’ 0.01; LOD 0; r2 = 0; 95% CI 0.01–0.696) and CYP1A1*4 and CYP1A1*2A were not (D’ 0.582; LOD 1.21; r2 = 0.0020; 95% CI 0.17–0.82).
Single-base extension genotyping
As a secondary method of confirming PCR accuracy and to address the issue of Hardy–Weinberg disequilibrium, single-base extension genotyping methods (iPLEX) were used to genotype cases and controls. It was possible to successfully re-genotype 89% of cases and controls with iPLEX, with failure rates of 3–5% for the three genotypes. Tables 2⇑ and 3⇑ illustrate the sample numbers and genotype and haplotype frequencies observed using iPLEX. All variants were in HWE for the control population (table 4⇑), with two of three variants in HWE for cases (table 4⇑).
Similar genotype frequencies were observed for CYP1A1*2C, CYP1A1*4 and CYP1A1*2A, although relatively fewer variant homozygotes were observed for CYP1A1*4 (five for PCR-RFLP versus none for iPLEX in controls). Observed associations between CYP1A1 polymorphisms and lung cancer risk remained consistent, with subjects carrying the CYP1A1*2A variant over-represented among NSCLC cases compared to controls (TT versus CT/CC; OR 1.43 (95% CI 1.35–1.52); p = 0.005), and carriers of the CYP1A1*4 variant under-represented in NSCLC cases (CC versus CA/AA; OR 0.66 (95% CI 0.66–0.67); p = 0.026). Carriers of the CYP1A1*2C variant were also associated with elevated lung cancer risk (AA versus AG/GG; OR 1.83 (95% CI 1.57–2.12); p = 0.002), consistent with previous findings. Although the findings from PCR-RFLP remained consistent, a decrease in significance was observed, possibly due to the small decrease in sample number genotyped by iPLEX.
In order to ensure that the haplotype findings from PCR-RFLP remained consistent, new haplotype analyses were also performed on the iPLEX data. Only samples with iPLEX genotype data for all three polymorphisms were included in these analyses. Four possible haplotypes (CAT, CAC, CGC and AAT) were observed amongst cases and controls (in allele order CYP1A1*4, CYP1A1*2C, CYP1A1*2A; table 3⇑), where CAT was considered to be the wild-type. No CGT haplotypes were observed in this group. Two of these haplotypes were significantly associated with either increased (CGC; 5.3 versus 3.0%; p = 0.0023) or decreased lung cancer risk (AAT; 3.4 versus 4.8%; p = 0.037), with CAC showing a trend towards increased lung cancer risk (8.5 versus 6.7%; p = 0.063). These risk haplotypes were also identified in the present PCR-RFLP haplotype analyses. CYP1A1*2A and CYP1A1*2C remained in linkage disequilibrium (D’ 1.0; LOD 98.17; r2 = 0.324; 95% CI 0.97–1.00), with CYP1A1*4 and CYP1A1*2C (D’ 0.991; LOD 0.67; r2 = 0.002; 95% CI 0.11–0.98) and CYP1A1*4 and CYP1A1*2A not exhibiting disequilibrium (D’ 1.0; LOD 2.76; r2 = 0.006; 95% CI 0.48–1.00), as observed in the PCR-RFLP findings.
DISCUSSION
It has previously been shown that carriers with the CYP1A1*2C Val allele (Ile/Val or Val/Val genotype) are significantly over-represented in NSCLC subjects compared to controls (OR 1.9 (95% CI 1.20–2.90); p = 0.005), especially in females, those aged <64 yrs and those with <46 pack-yrs of tobacco exposure 26. It has also previously been shown that the CYP1A1*2C variant, in combination with the MPO risk allele, confers a significantly increased risk of NSCLC (OR 2.88 (95% CI 1.70–5.00); p<0.0001) 11. In the present report, we present the genotype and haplotype frequencies of two additional CYP1A1 polymorphisms in relation to lung cancer risk. These data indicate an association between CYP1A1*2A and lung cancer risk, and support a recent pooled analysis of 2,451 lung cancer subjects and 3,358 controls that showed a clear association between the homozygous CYP1A1*2A CC allele and lung cancer risk in Caucasians (age- and sex-adjusted OR 2.36 (95% CI 1.16–4.81)) 15. Although a meta-analysis of Houlston 29 provided little support for involvement in lung cancer risk of variations in CYP1A1, a recent review investigating the role of polymorphisms in candidate genes for 18 different cancer sites reported an increased risk of lung cancer for carriers of the CYP1A1*2A variant in Caucasian populations (OR 2.36 (95% 1.16–4.81); p = 0.018) in addition to variants of CYP1A1*2C in Asian populations (OR 1.61 (95% CI 1.24–2.08); p = 0.0003) 30. Associations between CYP1A1 variations and lung cancer risk have also been observed in never-smokers with lung cancer 24, 31.
CYP1A1*2A and CYP1A1*2C have been reported to be in linkage disequilibrium in Caucasian 18 and Asian populations 27, although not in people of African descent 32. Linkage disequilibrium can be influenced by a variety of factors, including genetic linkage, recombination, mutation rates, random drift, nonrandom mating and population structure. In the present study, CYP1A1*2A and CYP1A1*2C were in linkage disequilibrium, confirming reports from previous studies 18. Conversely, as in Cascorbi et al. 6, no evidence of linkage disequilibrium between CYP1A1*2C and CYP1A1*4 was observed. This may be due to the close proximity of the two variants (only one base pair separates the two polymorphic sites) decreasing the rate of recombination.
The function of CYP1A1*4 has not been clearly established, although it has been suggested that it shows the greatest enzymatic efficiency of all of the CYP1A1 polymorphisms 33. In the present study, very few homozygous variants of CYP1A1*4 were identified among cases or controls, limiting any ability to draw strong conclusions for a Caucasian population. However, a general decrease in the frequency of heterozygotes was observed in this study population compared to controls. Others have reported no clear association between lung cancer risk and CYP1A1*4 3.
Very few haplotype analyses studying the interactions between CYP1A1*2A, CYP1A1*2C and CYP1A1*4 have been performed in populations of Caucasian lung cancer subjects. A recent pooled analysis studying interactions between CYP1A1 polymorphisms and GSTM1/GSTT1 in an Asian population identified the GT and AC haplotypes as being associated with lung cancer risk compared to AT (OR 3.41 (95% CI 1.78–6.53) and 1.39 (1.12–1.71), respectively) 34. Among the present population, the haplotypes CGC and CGT (in allele order CYP1A1*4, CYP1A1*2C, CYP1A1*2A) were associated with an increased risk of lung cancer, whereas AAT was associated with a decreased risk. In an Indian population, a four-fold increased lung cancer risk (hazard ratio 3.90 (95% CI 1.00–5.10); p = 0.025) was shown for the CGC haplotype 3, suggesting that this haplotype may be an important risk factor for NSCLC. Although the CAC haplotype has been associated with higher enzymatic activity and decreased cancer risk in prostate cancer 35, Shah et al. 3 showed a trend for increased lung cancer risk, although this did not reach significance. The differences observed may be due to tumour type, gene expression or functional role of CYP1A1. For example, in addition to detoxification of carcinogens, CYP1A1 also has a role in the oxidative metabolism of oestrogens 36, which have been implicated in the aetiology of prostate cancer 37, 38. Chang et al. 35 suggested that, in prostate cancer, CYP1A1’s role of oestrogen metabolism may be more important than that of carcinogen detoxification. Therefore, it is possible that the role of CYP1A1 differs between cancer types, resulting in haplotype associations that vary according to tumour origin.
Interpretation of the data from the present study are curtailed by the low frequency of some variant genotypes (for instance very few CYP1A1*4 AA genotypes were observed) limiting any ability to draw statistically valid conclusions. Subjects in the control group also have the potential to develop lung cancer in the future, a source of misclassification bias. One of the polymorphisms, CYP1A1*2C was not in HWE in either controls or cases using PCR-RFLP, but was in HWE using iPLEX. In addition, although CYP1A1*2A was in HWE using PCR-RFLP, it was not in HWE using iPLEX in NSCLC subjects. This may be due to chance (from the selection of a large subset of the original samples), decreased sensitivity of iPLEX due to differences in sample concentration (10 ng·μL−1 for iPLEX versus 10–40 ng·μL−1 for PCR-RFLP, as determined by gel estimation techniques), a nonrandom population (highly unlikely) or a possible association with NSCLC risk. In addition, CYP1A1*4 and CYP1A1*2C are separated by only one nucleotide on exon 7, raising the possibility that misclassification could occur. In the present study, an attempt was made to overcome this by employing published assays using enzymes that selectively digest each polymorphism irrespective of the neighbouring polymorphism, thereby decreasing the risk of misclassification bias. Finally, the present study evaluated three commonly studied CYP1A1 variants; however, it is possible that other SNPs may be in linkage disequilibrium with these variants, and may also be instrumental in determining lung cancer risk.
Technical validation of the present PCR-RFLP method by iPLEX confirmed the observed associations for all of the CYP1A1 polymorphisms. Similar genotype frequencies were observed for CYP1A1*2C, CYP1A1*4 and CYP1A1*2A in both case and control populations. Haplotype analyses also confirmed association of haplotypes AAT and CGC with decreased and increased lung cancer risk, respectively. No CGT haplotypes were observed in these analyses, and association with lung cancer risk could not be confirmed. Despite the significance of these results decreasing in the iPLEX cohort, it remained possible to confirm the observations obtained using PCR-RFLP, increasing our confidence in the reproducibility of genotyping techniques. Owing to the high reproducibility rate, we are confident that these associations are valid and support a role for CYP1A1 polymorphisms in altering lung cancer risk.
In conclusion, the results of the present study confirm that CYP1A1 polymorphisms are a minor risk factor for NSCLC. Although several studies have confirmed associations between these polymorphisms and lung cancer risk in the past, few have explored linkage effects between the three different CYP1A1 polymorphisms in a Caucasian population. To the present authors’ knowledge, this is the first study to show that haplotypes CGC, CGT and AAT are strongly associated with lung cancer risk in Caucasians. Identification of these haplotypes may assist in risk stratification, early detection and improvement of current treatment options for subjects with lung cancer. Larger studies are required in order to explore these risk CYP1A1 haplotypes further.
Support statement
This study was supported by The Prince Charles Hospital Foundation (Brisbane), a Queensland Smart State (Brisbane) Fellowship (K.M. Fong), a National Health and Medical Research Council (NHMRC; Canberra) Practitioner Fellowship (K.M. Fong), a Queensland Cancer Fund (Brisbane) Clinical Research Fellowship (K.M. Fong) and an NHMRC Career Development Award (I.A. Yang) (all Australia).
Statement of interest
None declared.
Acknowledgments
The authors would like to thank A. Semmler (Thoracic Research Laboratory, Prince Charles Hospital, Chermside, Australia). The authors also wish to thank the physicians, surgeons and pathologists at The Prince Charles Hospital (Brisbane, Australia) who assisted with the project; the subjects and donors who participated in the study; L.H. Passmore for recruiting subjects and confirming clinical data; and technical staff for their assistance with DNA extractions.
- Received August 6, 2008.
- Accepted July 5, 2009.
- © ERS Journals Ltd
References
