Abstract
Introduction
Three genome-wide association studies identified a region on chromosome 15q25.1 associated with lung cancer and measures of nicotine addiction. This region includes nicotinic acetylcholine receptor subunit genes CHRNA3 and CHRNA5. These studies were conducted in European or European American populations and do not provide risk estimates for African Americans. The goal of this study was to determine whether recently identified genetic variation in 3 SNPs (rs1051730, rs931794, rs8034191) on chromosome 15q25.1 contributes to risk of lung cancer in African Americans.
Methods
Data were derived from three case-control studies. Participants included 1058 population-based non-small cell lung cancer cases selected from the Detroit area SEER registry and 1314 controls matched within study by age, race, and sex. Thirty-nine percent of participants were African American.
Results
Risk associated with rs1051730 (odds ratio 1.59; 95% confidence interval 1.16–2.19) and rs931794 (odds ratio 1.39; 95% confidence interval 1.09–1.78) increased in ever smoking African Americans adjusting for cigarettes smoked per day. Among white cases, the number of cigarettes smoked varied by genotype at all three SNPs, and when smoking quantity was included in the models, risk was not significantly associated with any of the three SNPs.
Conclusions
These findings suggest that SNPs in the CHRNA3 and CHRNA5 region contribute to lung cancer risk, and while variant alleles are less frequent in African Americans, risk in this group may be greater than in whites and less likely to reflect an indirect effect on lung cancer risk through nicotine dependence.
Keywords: Non-small cell lung cancer, Smoking, SNPs
Lung cancer continues to be the leading cause of cancer death in the United States.1 Incidence rates are higher in African Americans than in whites, with average annual age-adjusted rates, 2001 to 2005, at 50.3 and 41.3 per 100,000, respectively.1 With more than 85% of lung cancer risk attributable to cigarette smoking, this racial variation in incidence occurs even though African Americans smoke fewer cigarettes per day than whites.2 In addition, an understanding of the influence of genetic susceptibility is not fully appreciated. A genetic contribution to lung cancer risk has been suggested by studies demonstrating familial aggregation of lung cancer. 3 In addition, a study in high-risk lung-cancer families reported linkage on chromosome 6q23–25 (146–164 cM) to familial lung cancer;4 however, not all high-risk families have been linked to this region and only about 1% of lung cancer patients have such extensive family histories.
Running parallel to studies of rare genes, studies of candidate genes have produced mixed results5 leaving open the question of the contribution to risk from common, low-penetrant genes. Alternative strategies to gene identification, such as genome-wide association studies (GWAS), have recently been completed to address this question. Three GWAS, conducted in European and European American populations, reported an association between non-small cell lung cancer (NSCLC) risk and SNPs on 15q25.1.6–8 This region has also been associated with familial lung cancer risk in an additional genome-wide association study.9 The region of interest contains nicotinic acetylcholine receptor subunit genes CHRNA3 and CHRNA5, and one of the GWAS supports a stronger association with nicotine addiction than lung cancer.8 Genetic variation in this region has also been linked to smoking behavior,10 age-dependent nicotine addiction,11 and variability in internal dose of tobacco-specific nitrosamine per cigarette.12
To further evaluate risk of lung cancer and smoking behavior associated with this region, we genotyped three SNPs on 15q25 in a pooled sample from three case-control studies conducted in metropolitan Detroit. Each study identified population-based cases from the Metropolitan Detroit Cancer Surveillance System, a participant in NCI’s SEER program and included a relatively large sample of African Americans.13–15
MATERIALS AND METHODS
Institutional review board approval was obtained for all studies, and informed consent was obtained from all participants. Table 1 describes the three studies included in the analysis. All studies were conducted by the same study staff using identical procedures, with cases ascertained through the population-based Metropolitan Detroit Cancer Surveillance System, an NCI-funded SEER registry. Studies differed only in the eligibility of cases, with the family history study focused on never smokers and cases diagnosed before the age of 50 years. The women’s epidemiology of lung diseases study focused on women, and the exploring health, ancestry, and lung epidemiology (EXHALE) study focuses on African Americans. The EXHALE study is an ongoing study. With the exception of the EXHALE study, population-based controls were chosen using random digit dialing methods. The EXHALE study relies on volunteer controls identified through research registries, advertisement, and word of mouth. All study controls were frequency matched to cases by 5-year age group, sex, and race. Questions about family history and smoking history were obtained using the same questionnaires in each study. The family history study case interview response rate was 67.7%, and biospecimens, including blood, buccal cells through buccal brush or mouthwash collection, and normal tissue from paraffin-embedded tissue, were collected for 53.5% of these cases. The control participation rate was 93.1% of those completing the eligibility screeners, with 50.3% providing biospecimens. The women’s epidemiology of lung diseases study had an interview participation rate for cases of 54.2%, for controls of 69.6%, with biospecimens collected for 98.6% of subjects. The EXHALE study is still recruiting and, to date, 87.3% of subjects have provided a biospecimen. Volunteer controls were selected for the EXHALE study that focuses on using admixture mapping to identify novel genes for lung cancer. It is unlikely that unknown genotype influences volunteer participation or participation in any of the studies. Because 92% of all cases had NSCLC, the results are restricted to 1058 NSCLC cases and 1314 controls providing a biospecimen. Thirty-nine percent of the cases and 34% of the controls included were African Americans.
TABLE 1.
Study Characteristics
| Study Population | Eligibility Criteria | Years of Diagnosis | Ages of Diagnosis (yr) |
Control Selection |
Number of NSCLC Cases Genotyped |
Number of Controls Genotyped |
|---|---|---|---|---|---|---|
| Family history study13,15,22 | Never smokers + | From November 1, 1984, to June 30, 1987 + From September 15, 1990, to September 30, 2003 | 40–84 + <50 | Random digit dialing | 299 | 570 |
| Early onset with any smoking history | ||||||
| Men and women | ||||||
| All races | ||||||
| All histologies | ||||||
| Women’s epidemiology of lung disease14 | All smoking histories | From November 1, 2001 to October 31, 2005 | <74 | Random digit dialing | 567 | 562 |
| Only women | ||||||
| All races | ||||||
| Only NSCLC | ||||||
| Exploring health, ancestry and lung epidemiology | All smoking histories | From November 1, 2005, to October 31, 2009 | 18–74 | Volunteers | 192 | 182 |
| Men and women | ||||||
| Only African Americans | ||||||
| All histologies | ||||||
| Total NSCLC | 1058 | 1314 |
NSCLC, non-small cell lung cancer.
DNA was isolated from blood using a Gentra AutoPure Kit (Qiagen, Valencia, CA), buccal swabs with the BuccalAmp DNA Extraction Kit (Epicenter Technologies, Madison, WI), mouthwash samples with the Gentra Puregene Kit (Qiagen), and paraffin-embedded tissue with the QIAamp DNA Mini Kit (Qiagen) using standard protocols. TaqMan Genotyping Assays (Applied Biosystems, Foster City, CA) were used to detect polymorphisms in rs8034191, rs1051730, and rs931794. DNA isolated from buccal cells or from paraffin blocks was preamplified in an outer polymerase chain reaction for added sensitivity. Either 25 ng DNA or 1 µL of the outer nest was amplified, with primers designed using Primer Express software (Applied Biosystems), and detected using an AB 7900 Sequence Detection System (Applied Biosystems). For quality control, 5% of the products were sequenced and 10% were directly repeated.
Cases were compared with controls on demographic factors using χ2 tests for categorical variables and t tests for continuous variables. Hardy-Weinberg equilibrium was assessed using a χ2 goodness-of-fit test for each SNP in African American and white controls separately. Unconditional logistic regression was used to examine the effect of these SNPs on lung cancer risk, controlling for age, sex, and one of the three measures of smoking behavior: ever smoking, pack-years of smoking, or cigarettes smoked per day. Analyses were conducted separately by race for whites and African Americans. Results for models including pack-years and cigarettes smoked per day were very similar; hence, the unadjusted models and models including cigarettes per day are presented. Participants of other races (26 cases and 26 controls) were excluded from the analysis. Analyses were conducted assuming a log additive model. Analyses were repeated stratified by smoking history, family history of lung cancer in a first-degree relative, and age. A dominant model was also tested for each SNP, with carriers of one or two copies of the minor allele compared with carriers of two major alleles, and results were very similar to those under the log-additive model so these results are not presented. For each SNP, a model including race and a race × genotype interaction term was also created to evaluate the significance of the difference between SNP associations by race. The association between mean cigarettes smoked per day and genotype was evaluated using one-way analysis of variance within racial group. Linkage disequilibrium (LD) between SNPs was determined using r2 values within each racial group. All statistics were derived using SAS version 9.1 (SAS Institute, Cary, NC), using the genetics module for analyses using SNPs.
RESULTS
All SNPs were in Hardy-Weinbnerg equilibrium (Table 2). Minor allele frequencies varied by race, with African Americans less likely than whites to carry the minor alleles at any of the three SNPs (Table 2). In a log-additive model, risk of NSCLC associated with any of the SNPs among whites was increased by 20–25%, when adjusting for age, sex, and ever smoking (Table 3). However, after adjusting for cigarettes smoked per day or pack-years, risk estimates in whites were reduced and were no longer statistically significant. Among African Americans, risk was not significantly associated with rs8034191. In African Americans, significantly increased risk was associated with rs931794 (odds ratio [OR] 1.41; 95% confidence interval [CI] 1.10 –1.81) and rs1051730 (OR 1.66; 95% CI 1.20–2.30) after adjusting for age, sex, and ever smoking. Risk remained unchanged and significantly elevated among African Americans when adjustment was made for cigarettes smoked per day for rs931794 and rs1051730. None of the SNPs were associated with lung cancer among never smokers (Table 4). In the subset of ever smokers, ORs associated with risk of lung cancer were 1.18–1.36 in whites and 1.18–1.71 in African Americans before adjusting for smoking amount. Risk estimates decreased and were no longer statistically significant in whites after adjustment for cigarettes smoked per day, but similar decreases in risk were not observed among African Americans for rs931794 and rs1051730. In models including race, race was a significant predictor of risk independent of genotype, and there was a significant interaction between race and genotype for rs1051730 (p = 0.04).
TABLE 2.
Study Participant Characteristics and Genotype Frequencies by Race
| White |
African American |
|||
|---|---|---|---|---|
| Cases | Controls | Cases | Controls | |
| (N = 630) | (N = 853) | (N = 402) | (N = 435) | |
| Smoking status | ||||
| Never smoked | 111 (17.6) | 451 (52.9) | 41 (10.2) | 196 (45.1) |
| Ever smoked | 519 (82.4) | 402 (47.1) | 361 (89.8) | 239 (54.9) |
| Family history of lung cancer | ||||
| No | 491 (77.9) | 763 (89.5) | 322 (80.1) | 380 (87.4) |
| Yes | 139 (22.1) | 90 (10.5) | 79 (19.7) | 55 (12.6) |
| Unknown | 1 (0.2) | |||
| Age at diagnosis/interview (yr) | ||||
| <50 | 211 (33.5) | 411 (48.2) | 121 (30.1) | 149 (34.7) |
| 50+ | 419 (66.5) | 442 (51.8) | 281 (69.9) | 286 (65.8) |
| rs8034191a | ||||
| A/A | 185 (34.3) | 326 (40.3) | 231 (64.2) | 300 (71.3) |
| A/G | 264 (49.0) | 367 (45.4) | 119 (33.1) | 106 (25.2) |
| G/G | 90 (16.7) | 116 (14.3) | 10 (2.8) | 15 (3.6) |
| **p = 0.43 | **p = 0.15 | |||
| rs931794a | ||||
| A/A | 214 (34.6) | 336 (40.3) | 215 (54.4) | 270 (62.9) |
| A/G | 287 (46.4) | 368 (44.2) | 154 (39.0) | 137 (31.9) |
| G/G | 117 (18.9) | 129 (15.5) | 26 (6.6) | 22 (5.1) |
| **p = 0.09 | **p = 0.40 | |||
| rs1051730a | ||||
| G/G | 207 (35.6) | 344 (40.8) | 279 (72.5) | 353 (81.7) |
| A/G | 280 (48.1) | 379 (44.9) | 96 (24.9) | 74 (17.1) |
| A/A | 95 (16.3) | 121 (14.3) | 10 (2.6) | 5 (1.2) |
| **p = 0.32 | **p = 0.62 | |||
Genotype data were missing for 191 subjects for rs8034191, 45 subjects for rs931794, and 77 subjects for rs1051730.
p value for χ2 test of Hardy-Weinberg Equilibrium in controls.
TABLE 3.
Risk Estimates for NSCLC for Chromosome 15q25 SNPs by Race
| White |
African American |
|||
|---|---|---|---|---|
| SNP | ORa (95% CI) | ORb (95% CI) | ORa (95% CI) | ORb (95% CI) |
| rs8034191 | 1.25 (1.04–1.49) | 1.10 (0.91–1.33) | 1.17 (0.88–1.57) | 1.17 (0.88–1.56) |
| rs931794 | 1.25 (1.06–1.46) | 1.13 (0.96–1.34) | 1.41 (1.10–1.81) | 1.39 (1.09–1.78) |
| rs1051730 | 1.20 (1.01–1.42) | 1.07 (0.90–1.28) | 1.66 (1.20–2.30) | 1.59 (1.16–2.19) |
Log additive model adjusted for age at diagnosis/interview, sex, and smoking status (ever/never).
Log additive model adjusted for age at diagnosis/interview, sex, and cigarettes smoked per day.
OR, odds ratio; CI, confidence interval.
TABLE 4.
Estimates of NSCLC Risk Associated With SNPs on 15q25 by Race, Smoking History, Family History, and Age at Diagnosis/Interview
| rs8034191 |
rs931794 |
rs1051730 |
||||
|---|---|---|---|---|---|---|
| Subject Characteristics | ORa (95% CI) | ORb (95% CI) | ORa (95% CI) | ORb (95% CI) | ORa (95% CI) | ORb (95% CI) |
| Smoking history | ||||||
| Whites | ||||||
| Never smokers | 0.94 (0.61–1.43) | 0.94 (0.61–1.43) | 1.01 (0.75–1.37) | 1.01 (0.75–1.37) | 0.90 (0.63–1.28) | 0.90 (0.63–1.28) |
| Ever smokers | 1.32 (1.08–1.61) | 1.18 (0.95–1.46) | 1.36 (1.12–1.65) | 1.21 (0.98–1.49) | 1.31 (1.07–1.59) | 1.21 (0.98–1.49) |
| African Americans | ||||||
| Never smokers | 1.03 (0.44–2.43) | 1.03 (0.44–2.43) | 1.06 (0.61–1.84) | 1.06 (0.61–1.84) | 1.47 (0.70–3.06) | 1.47 (0.70–3.06) |
| Ever smokers | 1.20 (0.88–1.63) | 1.18 (0.86–1.61) | 1.54 (1.15–2.05) | 1.54 (1.15–2.06) | 1.71 (1.18–2.48) | 1.67 (1.15–2.43) |
| Family history of lung cancer | ||||||
| Whites | ||||||
| No | 1.19 (0.99–1.44) | 1.05 (0.86–1.29) | 1.20 (1.01–1.42) | 1.10 (0.92–1.32) | 1.11 (0.93–1.33) | 1.00 (0.82–1.21) |
| Yes | 1.43 (1.07–1.89) | 1.30 (0.96–1.78) | 1.43 (1.09–1.86) | 1.31 (0.98–1.75) | 1.55 (1.17–2.06) | 1.41 (1.03–1.91) |
| African Americans | ||||||
| No | 1.17 (0.87–1.59) | 1.15 (0.85–1.56) | 1.42 (1.09–1.85) | 1.40 (1.07–1.82) | 1.59 (1.13–2.24) | 1.54 (1.09–2.16) |
| Yes | 1.16 (0.73–1.85) | 1.29 (0.81–2.05) | 1.35 (0.89–2.05) | 1.39 (0.93–2.11) | 1.80 (1.07–3.05) | 1.83 (1.10–3.06) |
| Age at diagnosis/interview (yr) | ||||||
| Whites | ||||||
| <50 | 1.37 (1.04–1.80) | 1.27 (0.95–1.69) | 1.35 (1.04–1.75) | 1.23 (0.94–1.60) | 1.36 (1.04–1.77) | 1.25 (0.95–1.65) |
| 50+ | 1.14 (0.90–1.45) | 0.97 (0.75–1.25) | 1.17 (0.95–1.44) | 1.06 (0.86–1.33) | 1.06 (0.85–1.33) | 0.93 (0.73–1.19) |
| African Americans | ||||||
| <50 | 1.53 (0.91–2.57) | 1.36 (0.81–2.28) | 1.32 (0.85–2.05) | 1.21 (0.78–1.90) | 1.83 (1.04–3.22) | 1.68 (0.95–2.96) |
| 50+ | 1.05 (0.74–1.50) | 1.12 (0.79–1.59) | 1.46 (1.07–1.99) | 1.48 (1.09–1.99) | 1.56 (1.04–2.33) | 1.52 (1.03–2.25) |
Log additive model adjusted for age at diagnosis/interview, sex and smoking status (ever/never).
Log additive model adjusted for age at diagnosis/interview, sex and cigarettes smoked per day.
OR, odds ratio; CI, confidence interval.
In a subset of cases with a family history of lung cancer, all SNPs were associated with increased risk in whites when cigarettes smoked per day was not included in the model, and rs1051730 remained significant in whites with a family history of lung cancer after adjusting for sex and cigarettes smoked per day (OR 1.41; 95% CI 1.03–1.91; Table 4). This SNP was also associated with lung cancer risk among African Americans with a family history of lung cancer (OR 1.83; 95% CI 1.10–3.06). In general, among African Americans, family history did not greatly change estimates of risk for any of the three SNPs. Risk varied by age at lung cancer diagnosis in whites, with young onset disease associated more strongly with these SNPs than in later onset lung cancer, but again after adjustment for cigarettes per day no significant increases in lung cancer risk were evident. In African Americans, risk patterns were similar across age groups, but statistically significant risks were only seen in those 50 years of age or older. There were no significant differences by sex (data not shown).
Because of possible associations between these SNPs and nicotine dependence, we also evaluated whether mean cigarettes smoked per day varied by genotype. White cases carrying the variant allele at any of the three SNPs reported significantly higher numbers of cigarettes smoked per day than white cases carrying major alleles (Table 5). A similar association was not seen in white controls. Cigarettes smoked per day did not vary significantly by genotype in African American cases or controls.
TABLE 5.
Mean Number of Cigarettes Smoked Per Day by Genotype and Race Among Ever Smoking Cases and Controls
| Mean Cigarettes/day |
||||
|---|---|---|---|---|
| White |
African American |
|||
| Controls | Cases | Controls | Cases | |
| SNPs | (N = 397) | (N = 517) | (N = 238) | (N = 359) |
| rs8034191 | ||||
| A/A | 17.4 | 25.0 | 15.2 | 18.1 |
| A/G | 19.0 | 28.1 | 15.1 | 20.6 |
| G/G | 18.0 | 29.2 | 11.9 | 18.9 |
| pa | 0.42 | 0.02 | 0.67 | 0.15 |
| rs931794 | ||||
| A/A | 17.1 | 24.9 | 15.1 | 19.4 |
| A/G | 18.8 | 28.3 | 15.4 | 18.7 |
| G/G | 17.7 | 29.3 | 10.1 | 22.0 |
| pa | 0.33 | 0.009 | 0.33 | 0.43 |
| rs1051730 | ||||
| G/G | 17.1 | 25.4 | 15.0 | 18.5 |
| A/G | 18.9 | 28.3 | 15.4 | 20.2 |
| A/A | 18.5 | 29.4 | 5.7 | 22.5 |
| pa | 0.27 | 0.03 | 0.35 | 0.30 |
One-way ANOVA.
There is strong LD between all three SNPs in the population of European ancestry in HapMap (Fig. 1A), and, in our study population, r2 values in whites were more than 0.88 between any combination of the three SNPs. In a stepwise logistic regression analysis in whites that did not include cigarettes smoked per day, only rs931794 remained in the model as significantly associated with lung cancer risk. However, once smoking quantity was included, this SNP was not significantly associated with lung cancer risk in whites. In the Yoruban population in HapMap, there is little LD in this region (Fig. 1B). In our population of African Americans, r2 values were 0.24 or less between SNPs. When all three SNPs were included in a logistic regression model in a stepwise fashion in African Americans, only rs1051730 was predictive of risk.
FIGURE 1.
A, LD structure at 15q25 in individuals of European ancestry. Boxes are shaded according to r2 values derived in Haploview (v3.2). Triangular border areas denote haplotype blocks assigned according to the method of Gabriel et al.23 B, LD structure at 15q25 in Yorubans. Boxes are shaded according to r2 values derived in Haploview (v3.2). Triangular border areas denote haplotype blocks assigned according to the method of Gabriel et al.23
DISCUSSION
The frequency of genotypes in 15q25 SNPs, rs8034191, rs931794, and rs1051730, in the white population was consistent with reported frequencies in the GWAS studies.6–8 This is the first report on the frequency of these genotypes in African Americans. Among African Americans, the homozygous rare variant genotypes for each SNP were relatively uncommon and occurred in fewer than 6% of the controls. A recent study in Asians evaluated seven SNPs in this region including rs8034191 and rs1051730.16 Allele frequencies in Asians for these two SNPs were lower than what we observed in whites and African Americans in our study; for rs8034191 the minor allele frequencies were reported to be 0.37, 0.16, and 0.02 in whites, African Americans, and Asians, respectively. Likewise, for rs1051730 the minor allele frequencies were reported to be 0.37, 0.10, and 0.01 in whites, African Americans and Asians, respectively.
Estimates of risk associated with these three SNPs are consistent with the GWAS results6–8 and with the existence of a gene associated with lung cancer, directly or indirectly through nicotine dependence, on chromosome 15q25. Among white cases in our studies, all three SNPs were significantly related to smoking amount measured as cigarettes smoked per day and may contribute to lung cancer risk through increased exposure to tobacco smoke. Risk associated with these SNPs is highest in ever smoking whites after adjustment for cigarettes per day for rs931794, but these findings did not reach statistical significance. Amos et al.6 also report weak evidence of a relationship between genotype and smoking behavior and there is extensive literature indicating that SNPs in this region associate with smoking behavior8,17. It is possible that this study was underpowered to detect a small increase in lung cancer risk, on the order of 10–15%, independent of smoking quantity in whites.
The effect of smoking amount is not evident among African Americans, where the increased lung cancer risk associated with two of the three SNPs was independent of smoking dose. Results were most significant in African Americans for rs1051730, which is in a LD block encompassing CHRNA3/CHRNA5. Although variant alleles were less frequent in African Americans, risk in this group was similar if not greater than that in whites. Reasons underlying the racial difference in the genotype/smoking amount associations are unknown. It is possible that cigarettes per day do not adequately measure nicotine dependence in African Americans. It is also possible that a marker linked to the SNPs studied is the causal variant, and the frequency of that marker and its association with number of cigarettes smoked varies by race.
In this study population, no increased risk of lung cancer was associated with these SNPs among never smokers. Similar results were shown in one of the replication studies presented by Amos et al.6 and an additional replication population from Texas,18 but in the study of Hung et al.7 lung cancer in never smokers was associated with rs8034191 genotype. In the study conducted in Japan, a haplotype including rs8034191 and rs1051730 in combination with rs16969968 was associated with lung cancer risk both in smokers and non-smokers.16 The numbers of never smokers with lung cancer included in any of these studies are relatively small and this is an area that deserves more attention in future studies.
Liu et al.9 report higher risks associated with 15q polymorphisms for familial lung cancer. In that small study, risks were increased more than threefold among homozygous carriers of the rare alleles. In this study, the only significant finding of risk among whites, after adjusting for cigarettes per day, was for rs1051730 in those with a family history of lung cancer. A stronger association at this SNP was seen for African Americans with a family history, but risk associated with these SNPs, in general, did not vary greatly between African Americans with and without a family history of lung cancer. Liu et al. included individuals with three or more family members affected, whereas in our study, most subjects with a family history had only one affected relative.
In addition to the myriad of carcinogens present in tobacco smoke, nicotine and the nicotinic acetylcholine receptor pathway may be involved in the pathogenesis of lung cancer through inhibition of proapoptotic pathways and promotion of cell proliferation, invasion, and angiogenesis.19–21 Candidate genes in this 15q region include CHRNA3 and CHRNA5. The strong link between smoking amount and lung cancer makes it difficult to determine whether lung cancer risk is more directly associated with variation in this genetic region or indirectly through increased exposure because of nicotine addiction. Three of the four GWAS studies and this study did not include measures of nicotine dependence other than smoking dose and duration making nicotine addiction difficult to classify. SNPs in this region have been associated with nicotine dependence severity among subjects initiating daily smoking at or before age 16 years11 and heavy smoking. 10 These studies were conducted in healthy populations and included only whites. Most recently, Le Marchand et al.12 demonstrated that variants in CHRNA3 (rs1051730, and included in this study) and CHRNA5 (rs16969968, which is in near perfect LD with rs1051730 in whites) alter internal dose of a carcinogenic tobacco-specific nitrosamine per cigarette smoked, as measured in urine. Although that study included small numbers of African Americans, results were not presented by race. Taken together, reported findings suggest that more sensitive markers of smoking dose are needed to fully understand the genetics underlying nicotine dependence and lung cancer.
The study presented here has a number of strengths including population-based case selection. It is the first report of the role of SNPs on 15q24–25 in lung cancer and smoking behavior in African Americans. There are also some limitations to this study. Response rates for population-based lung cancer case-control studies, in general and in our studies, tend to be low because of the poor survival and great morbidity associated with this disease. However, genotype at any of the three SNPs was not associated with survival, so survival bias is not likely to account for the results presented. In addition, the only measures of nicotine dependence available were smoking dose and duration.
As the first of the GWAS results are being published, it is reassuring that four GWAS in lung cancer identified the same region and that there are biologically plausible genes in this region. What is more daunting is how best to use the data coming from these studies. GWAS data are analyzed as if SNPs are independent, and there is little evaluation of gene-gene or gene-environment interactions. Because of genetic variation, GWAS are typically conducted in European or European American populations. Replication of findings in other populations with different LD patterns and in this case, different smoking patterns, may provide important clues in attempts to fine map regions of interest. Further identification of the specific genetic changes associated with NSCLC and nicotine dependence should include the study of African Americans who have higher rates of lung cancer than whites, but have different patterns of cigarette use.
ACKNOWLEDGMENTS
This research was funded in part by the NIH grants R01-CA87895, R01-CA60691, R01CA-134791, and R01- CA121197 and contracts N01-PC35145 and P30CA22453.
Footnotes
Disclosure: The authors declare no conflicts of interest.
REFERENCES
- 1.Ries LAG, Krapcho M, Stinchcomb DG, et al., editors. SEER Cancer Statistics Review, 1975–2005. Bethesda, MD: National Cancer Institute; [Accessed February 12 2009]. Available at: http://seer.cancer.gov/csr/1975_2005/. [Google Scholar]
- 2.Haiman CA, Stram DO, Wilkens LR, et al. Ethnic and racial differences in the smoking-related risk of lung cancer. N Engl J Med. 2006;354:333–342. doi: 10.1056/NEJMoa033250. [DOI] [PubMed] [Google Scholar]
- 3.Matakidou A, Eisen T, Houlston RS. Systematic review of the relationship between family history and lung cancer risk. Br J Cancer. 2005;93:825–833. doi: 10.1038/sj.bjc.6602769. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bailey-Wilson JE, Amos CI, Pinney SM, et al. A major lung cancer susceptibility locus maps to chromosome 6q23-25. Am J Hum Genet. 2004;75:460–474. doi: 10.1086/423857. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schwartz AG, Prysak GM, Bock CH, Cote ML. The molecular epidemiology of lung cancer. Carcinogenesis. 2007;28:507–518. doi: 10.1093/carcin/bgl253. [DOI] [PubMed] [Google Scholar]
- 6.Amos CI, Wu X, Broderick P, et al. Genome-wide association scan of tag SNPs identifies a susceptibility locus for lung cancer at 15q25.1. Nat Genet. 2008;40:616–622. doi: 10.1038/ng.109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Hung RJ, McKay JD, Gaborieau V, et al. A susceptibility locus for lung cancer maps to nicotinic acetylcholine receptor subunit genes on 15q25. Nature. 2008;452:633–637. doi: 10.1038/nature06885. [DOI] [PubMed] [Google Scholar]
- 8.Thorgeirsson TE, Geller F, Sulem P, et al. A variant associated with nicotine dependence, lung cancer and peripheral arterial disease. Nature. 2008;452:638–642. doi: 10.1038/nature06846. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Liu P, Vikis HG, Wang D, et al. Familial aggregation of common sequence variants on 15q24-25.1 in lung cancer. J Natl Cancer Inst. 2008;100:1326–1330. doi: 10.1093/jnci/djn268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Stevens VL, Bierut LJ, Talbot JT, et al. Nicotinic receptor gene variants influence susceptibility to heavy smoking. Cancer Epidemiol Biomarkers Prev. 2008;17:3517–3525. doi: 10.1158/1055-9965.EPI-08-0585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Weiss RB, Baker TB, Cannon DS, et al. A candidate gene approach identifies the CHRNA5-A3-B4 region as a risk factor for age-dependent nicotine addiction. PLoS Genet. 2008;4:e1000125. doi: 10.1371/journal.pgen.1000125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Le Marchand L, Derby KS, Murphy SE, et al. Smokers with the CHRNA lung cancer-associated variants are exposed to higher levels of nicotine equivalents and a carcinogenic tobacco-specific nitrosamine. Cancer Res. 2008;68:9137–9140. doi: 10.1158/0008-5472.CAN-08-2271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Cote ML, Kardia SL, Wenzlaff AS, Ruckdeschel JC, Schwartz AG. Risk of lung cancer among white and black relatives of individuals with early-onset lung cancer. JAMA. 2005;293:3036–3042. doi: 10.1001/jama.293.24.3036. [DOI] [PubMed] [Google Scholar]
- 14.Schwartz AG, Wenzlaff AS, Prysak GM, et al. Reproductive factors, hormone, use estrogen receptor expression and risk of non small-cell lung cancer in women. J Clin Oncol. 2007;25:5785–5792. doi: 10.1200/JCO.2007.13.3975. [DOI] [PubMed] [Google Scholar]
- 15.Wenzlaff AS, Cote ML, Bock CH, Land SJ, Schwartz AG. GSTM1, GSTT1 and GSTP1 polymorphisms, environmental tobacco smoke exposure and risk of lung cancer among never smokers: a population-based study. Carcinogenesis. 2005;26:395–401. doi: 10.1093/carcin/bgh326. [DOI] [PubMed] [Google Scholar]
- 16.Shiraishi K, Kohno T, Kunitoh H, et al. Contribution of nicotine acetylcholine receptor polymorphisms to lung cancer risk in a smoking-independent manner in the Japanese. Carcinogenesis. 2009;30:65–70. doi: 10.1093/carcin/bgn257. [DOI] [PubMed] [Google Scholar]
- 17.Bierut LJ, Stitzel JA, Wang JC, et al. Variants in nicotinic receptors and risk for nicotine dependence. Am J Psychiatry. 2008;165:1163–1171. doi: 10.1176/appi.ajp.2008.07111711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Spitz MR, Amos CI, Dong Q, Lin J, Wu X. The CHRNA5-A3 region on chromosome 15q24-25.1 is a risk factor both for nicotine dependence and for lung cancer. J Natl Cancer Inst. 2008;100:1552–1556. doi: 10.1093/jnci/djn363. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Tsurutani J, Castillo SS, Brognard J, et al. Tobacco components stimulate Akt-dependent proliferation and NFkappaB-dependent survival in lung cancer cells. Carcinogenesis. 2005;26:1182–1195. doi: 10.1093/carcin/bgi072. [DOI] [PubMed] [Google Scholar]
- 20.Maneckjee R, Minna JD. Opioids induce while nicotine suppresses apoptosis in human lung cancer cells. Cell Growth Differ. 1994;5:1033–1040. [PubMed] [Google Scholar]
- 21.Egleton RD, Brown KC, Dasgupta P. Nicotinic acetylcholine receptors in cancer: multiple roles in proliferation and inhibition of apoptosis. Trends Pharmacol Sci. 2008;29:151–158. doi: 10.1016/j.tips.2007.12.006. [DOI] [PubMed] [Google Scholar]
- 22.Cote ML, Kardia SL, Wenzlaff AS, Land SJ, Schwartz AG. Combinations of glutathione S-transferase genotypes and risk of early-onset lung cancer in Caucasians and African Americans: a population-based study. Carcinogenesis. 2005;26:811–819. doi: 10.1093/carcin/bgi023. [DOI] [PubMed] [Google Scholar]
- 23.Gabriel SB, Schaffner SF, Nguyen H, et al. The structure of haplotype blocks in the human genome. Science. 2002;296:2225–2229. doi: 10.1126/science.1069424. [DOI] [PubMed] [Google Scholar]