Abstract

Common polymorphisms in the N-acetyltransferase 2 gene (NAT2) modify the association between cigarette smoking and bladder cancer and have been hypothesized to determine whether active cigarette smoking increases breast cancer risk. The authors sought to replicate the latter hypothesis in a prospective analysis of 6,900 breast cancer cases and 9,903 matched controls drawn from 6 cohorts (1989–2006) in the National Cancer Institute’s Breast and Prostate Cancer Cohort Consortium. Standardized methods were used to genotype the 3 most common polymorphisms that define NAT2 acetylation phenotype (rs1799930, rs1799931, and rs1801280). In unconditional logistic regression analyses, breast cancer risk was higher in women with more than 20 pack-years of active cigarette smoking than in never smokers (odds ratio (OR) = 1.28, 95% confidence interval (CI): 1.17, 1.39), after controlling for established risk factors other than alcohol consumption and physical inactivity. However, associations were similar for the slow (OR = 1.25, 95% CI: 1.11, 1.39) and rapid/intermediate (OR = 1.24, 95% CI: 1.08, 1.42) acetylation phenotypes, with no evidence of interaction (P = 0.87). These results provide some support for the hypothesis that long-term cigarette smoking may be causally associated with breast cancer risk but underscore the need for caution when interpreting sparse data on gene-environment interactions.

Tobacco smoke contains over 4,000 chemicals, among which more than 60 are listed as class 1 or class 2 carcinogens by the International Agency for Research on Cancer (IARC) (1). Approximately 20 of the constituents in cigarette smoke designated as definite or probable human carcinogens cause mammary tumors in rodents. However, whether active cigarette smoking increases breast cancer risk in humans has been debated for decades. In its most recent evaluation, the IARC judged the evidence for active smoking as “limited” (2), while both the California Environmental Protection Agency (Cal/EPA) (3) and a panel convened in Canada in May 2009 (4) concluded that “the associations between active smoking and both pre- and postmenopausal breast cancer are consistent with causality” (4, p. 11). Both the IARC and Cal/EPA designations acknowledge that bias, confounding, and chance cannot be excluded with reasonable certainty, even though the Cal/EPA terminology appears to be more definitive.

Part of the controversy surrounding cigarette smoking as a possible cause of breast cancer concerns whether common genetic polymorphisms predispose some female smokers to substantially greater risk. Polymorphisms in the N-acetyltransferase 2 gene (NAT2) have received the most attention. NAT2 is one of 2 human N-acetyltransferases that acetylate and thereby detoxify aromatic amines, an important class of carcinogens in tobacco smoke. The hypothesis that female smokers with the slow acetylation phenotype may have higher risk of breast cancer than those with the intermediate or rapid acetylation phenotype has been examined in case-control studies with retrospectively assessed information on smoking history, as well as in case-control studies nested in cohorts with prospectively assessed smoking history. The results of individual studies have been inconsistent. Findings have been summarized in a meta-analysis that included 6,758 breast cancer cases and a pooled analysis of 4,264 cases, both carried out by Ambrosone et al. (5). In the pooled analysis, Ambrosone et al. reported a statistically significant 1.4- to 1.5-fold increase in breast cancer risk among premenopausal and postmenopausal women who were slow acetylators and had accrued more than 20 pack-years of cigarette smoking, as compared with slow acetylators who had never smoked (5). The increase in risk was confined to persons with the slow acetylation phenotype; no statistically significant increase in risk was observed among rapid/intermediate acetylators who reported more than 20 pack-years of smoking (P for interaction not reported).

We sought to replicate this finding in a consortium of 6 large cohort studies being conducted in the United States and Europe with prospectively collected information on smoking history. Included in our analyses were 6,900 breast cancer cases and 9,903 controls. While the primary aim of our study was to assess interaction between smoking and the relevant NAT2 polymorphisms, we also assessed whether 2 polymorphisms associated with both nicotine addiction and lung cancer risk in genome-wide association studies were associated with breast cancer in our cohorts.

MATERIALS AND METHODS

Study population

The Breast and Prostate Cancer Cohort Consortium, which includes large, well-established cohorts assembled in the United States and Europe, and the participating cohorts have been described in detail elsewhere (6). Further details regarding the numbers of cases and controls obtained from each cohort, as well as the distributions of age and menopausal status at baseline, are shown in Table 1. Dates ranged from 1989 to 2006. Informed consent was obtained from all subjects, and each cohort study was approved by the relevant appropriate institutional review board. A relatively small number of cases (n = 392) from the Nurses’ Health Study cohort were included in the previous meta-analysis by Ambrosone et al. (5), and none were included in the prior pooled analyses. Because our main comparisons were with the prior pooled analyses (5), we did not exclude this small number of subjects from the analyses described here. In order to minimize issues of population stratification, we restricted our analyses to Caucasian women.

Table 1.

Contribution of Each Cohort in the Breast and Prostate Cancer Cohort Consortium to an Analysis of Interaction Between Cigarette Smoking and Relevant N-Acetyltransferase 2 Polymorphisms in Breast Cancer Risk, 1989–2006

Cohort Year Blood Collection Began No. of Cases No. of Controls Mean Age, years (SD)a
 
Postmenopausala
 
Cases Controls Cases
 
Controls
 
No. No. 
Cancer Prevention Study 2 1998 622 862 62.2 (6.2) 62.0 (6.1) 589 94.7 813 94.3 
European Prospective Investigation into Cancer and Nutrition 1992 2,456 3,352 54.5 (7.6) 54.6 (8.0) 1,565 63.7 2,099 62.6 
Multi-Ethnic Cohort Study 1996 522 574 60.6 (8.3) 59.2 (8.7) 463 88.7 472 82.2 
Nurses’ Health Study 1989 1,632 2,629 61.5 (8.1) 62.4 (8.2) 1,151 70.5 1,934 73.6 
Nurses’ Health Study II 1999 469 1,239 45.9 (4.2) 45.6 (4.1) 144 30.7 368 29.7 
Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial 1993 798 1,000 62.3 (5.0) 62.3 (5.1) 789 98.9 992 99.2 
Women’s Health Study 1993 701 696 56.0 (7.3) 56.0 (7.3) 446 63.6 418 60.1 
Cohort Year Blood Collection Began No. of Cases No. of Controls Mean Age, years (SD)a
 
Postmenopausala
 
Cases Controls Cases
 
Controls
 
No. No. 
Cancer Prevention Study 2 1998 622 862 62.2 (6.2) 62.0 (6.1) 589 94.7 813 94.3 
European Prospective Investigation into Cancer and Nutrition 1992 2,456 3,352 54.5 (7.6) 54.6 (8.0) 1,565 63.7 2,099 62.6 
Multi-Ethnic Cohort Study 1996 522 574 60.6 (8.3) 59.2 (8.7) 463 88.7 472 82.2 
Nurses’ Health Study 1989 1,632 2,629 61.5 (8.1) 62.4 (8.2) 1,151 70.5 1,934 73.6 
Nurses’ Health Study II 1999 469 1,239 45.9 (4.2) 45.6 (4.1) 144 30.7 368 29.7 
Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial 1993 798 1,000 62.3 (5.0) 62.3 (5.1) 789 98.9 992 99.2 
Women’s Health Study 1993 701 696 56.0 (7.3) 56.0 (7.3) 446 63.6 418 60.1 

Abbreviation: SD, standard deviation.

a

Age and menopausal status as determined at baseline (blood collection).

Breast cancer cases were confirmed by medical records, pathology reports, and/or linkage with population-based tumor registries. Controls were matched to cases by age at study entry (study baseline), and in some cohorts, additional matching criteria were employed (for example, country of residence in the European Prospective Investigation into Cancer and Nutrition).

Using a standardized approach, we genotyped the 3 most common polymorphisms that define NAT2 acetylation phenotype (rs1799930, rs1799931, and rs1801280), as well as 2 polymorphisms related to nicotine addiction and lung cancer risk (rs12914385 (7) and rs8034191 (8)), in 6,900 breast cancer cases and 9,903 matched controls with baseline information on tobacco smoking history and established breast cancer risk factors. All studies included blinded replicate samples for assessment of genotype reproducibility. No quality control discrepancies were detected in genotyping these replicates; genotyping success rates were high in all studies (>95%); and no deviations from Hardy-Weinberg equilibrium were observed.

NAT2 acetylation phenotype was determined using the NAT2PRED Web server (9). While NAT2PRED allows for the use of 6 polymorphisms in NAT2 to determine acetylation phenotype, linkage disequilibrium is high between the 3 single nucleotide polymorphisms (SNPs) we used and the other 3 SNPs possible (D′ > 0.80), and these same polymorphisms were used in the majority of the studies in the meta-analysis of Ambrosone et al. (5). Therefore, we included only these SNPs in our analyses.

Duration of cigarette smoking and pack-years of smoking were calculated using smoking data collected at study baseline. Both factors were analyzed as continuous and categorical variables. Smoking duration was stratified into ≤15 years and >15 years, and pack-years were stratified into ≤20 and >20, as in the study by Ambrosone et al. (5). Where data were available, we also calculated smoking duration and pack-years during the period between menarche and age at first full-term pregnancy (or menopause for nulliparous women). We combined intermediate acetylators with rapid acetylators in order to make our analyses more comparable with prior studies. Power calculations were carried out using QUANTO (10, 11). Unconditional logistic regression analyses controlling for age at baseline, body mass index at baseline, ever use of menopausal hormone therapy at baseline, parity, and cohort were used to estimate associations and 95% confidence intervals, with product terms included to evaluate interactions. All P values reported here are 2-sided. Tests for heterogeneity across cohorts were carried out using Cochran’s Q test. Statistical testing was carried out using SAS, version 9.1 (SAS Institute Inc., Cary, North Carolina).

RESULTS

Breast cancer risk was weakly associated with both duration of smoking and pack-years of smoking (for each additional year of smoking and for each additional pack-year, odds ratio = 1.01, 95% confidence interval: 1.00, 1.01; Table 2). No main-effect association was seen between NAT2 acetylation phenotype and breast cancer risk (Table 2). The associations observed between breast cancer and the 2 smoking parameters (for more than 20 pack-years of smoking, odds ratio = 1.28, 95% confidence interval: 1.17, 1.39) were not modified by NAT2 status, irrespective of whether the smoking variables were specified as continuous or categorical (all P’s for interaction ≥ 0.03; Table 3). No heterogeneity across the 6 cohorts in the main effects of either NAT2 acetylation phenotype or smoking history was observed (all P’s for heterogeneity > 0.30; Table 2). Assuming a null association between NAT2 acetylation phenotype (rapid/intermediate vs. slow acetylators) and breast cancer risk and a 1% increase in breast cancer risk per pack-year or year of cigarette smoking, our study had greater than 85% power to detect an odds ratio of 1.1 for interaction between NAT2 acetylation phenotype and cigarette smoking.

Table 2.

Association Between N-Acetyltransferase 2 Phenotype and Smoking History in Breast Cancer Risk in the Breast and Prostate Cancer Cohort Consortium, 1989–2006

 Cases
 
Controls
 
Odds Ratioa 95% Confidence Interval P for Heterogeneityb 
 No. Mean (SD) No. Mean (SD) 
NAT2 phenotypec          
    Rapid acetylator 456 6.6  625 6.3  Reference 0.95 
    Intermediate acetylator 2,369 34.1  3,366 33.8  0.97 0.85, 1.11 0.70 
    Slow acetylator 4,128 59.4  5,974 60.0  0.96 0.84, 1.09 0.40 
Smoking history          
    Years (continuous)   24.0 (13.1)   23.2 (13.1) 1.01 1.00, 1.01 0.33 
    Pack-years (continuous)   20.5 (18.7)   18.8 (17.2) 1.01 1.00, 1.01 0.37 
 Cases
 
Controls
 
Odds Ratioa 95% Confidence Interval P for Heterogeneityb 
 No. Mean (SD) No. Mean (SD) 
NAT2 phenotypec          
    Rapid acetylator 456 6.6  625 6.3  Reference 0.95 
    Intermediate acetylator 2,369 34.1  3,366 33.8  0.97 0.85, 1.11 0.70 
    Slow acetylator 4,128 59.4  5,974 60.0  0.96 0.84, 1.09 0.40 
Smoking history          
    Years (continuous)   24.0 (13.1)   23.2 (13.1) 1.01 1.00, 1.01 0.33 
    Pack-years (continuous)   20.5 (18.7)   18.8 (17.2) 1.01 1.00, 1.01 0.37 

Abbreviations: NAT2, N-acetyltransferase 2; SD, standard deviation.

a

Logistic regression was used to estimate odds ratios and 95% confidence intervals, adjusted for age at baseline, body mass index at baseline, ever use of menopausal hormone therapy at baseline, parity, and cohort.

b

P for heterogeneity across cohorts using Cochran’s Q test for all odds ratios (including the trend test).

c

Test of trend from rapid acetylation phenotype to slow acetylation phenotype: P = 0.52.

Table 3.

Interactions Between N-Acetyltransferase 2 Phenotype and Smoking History in Breast Cancer Risk, Breast and Prostate Cancer Cohort Consortium, 1989–2006

Smoking History and NAT2 Phenotype Cases
 
Controls
 
Odds Ratioa 95% Confidence Interval P for Interactionb 
Mean (SD) No. Mean (SD) No. 
Duration of smoking (continuous), years         0.98 
    Rapid + intermediate acetylators 23.9 (13.0)   23 (12.9)   1.01 1.00, 1.01  
    Slow acetylators 24.0 (13.2)   23.2 (13.1)   1.01 1.00, 1.01  
Pack-years of smoking (continuous)         0.87 
    Rapid + intermediate acetylators 20.0 (18.0)   18.7 (17.2)   1.01 1.00, 1.01  
    Slow acetylators 20.7 (19.1)   19.0 (17.3)   1.01 1.00, 1.01  
Duration of smoking between menarche and first full-term pregnancy (continuous), years         0.03 
    Rapid + intermediate acetylators 7.4 (7.7)   7.7 (8.1)   1.00 1.00, 1.00  
    Slow acetylators 7.4 (7.7)   7.9 (8.6)   1.00 1.00, 1.00  
Pack-years of smoking between menarche and first full-term pregnancy (continuous)         0.14 
    Rapid + intermediate acetylators 5.4 (7.5)   5.3 (6.9)   1.01 1.00, 1.02  
    Slow acetylators 5.5 (7.4)   5.6 (7.5)   1.00 1.00, 1.01  
Duration of smoking (categorical), years         0.97 
    Rapid + intermediate acetylators          
        Never smoking  1,431 51.7  2,165 55.4 Reference  
        ≤15  410 14.8  559 14.3 1.14 0.98, 1.32  
        >15  927 33.5  1,182 30.3 1.17 1.04, 1.30  
    Slow acetylators          
        Never smoking  2,090 51.5  3,205 54.8 Reference  
        ≤15  606 14.9  865 14.8 1.07 0.95, 1.21  
        >15  1,362 33.6  1,775 30.4 1.18 1.07, 1.29  
Pack-years of smoking (categorical)         0.87 
    Rapid + intermediate acetylators          
        Never smoking  1,431 52.7  2,163 56.5 Reference  
        ≤20  772 28.4  1,045 27.3 1.13 1.01, 1.27  
        >20  514 18.9  619 16.7 1.24 1.08, 1.42  
    Slow acetylators          
        Never smoking  2,090 52.8  3,205 56.1 Reference  
        ≤20  1,108 28.0  1,571 27.5 1.08 0.92, 1.19  
        >20  757 19.1  936 16.4 1.25 1.11, 1.39  
Smoking History and NAT2 Phenotype Cases
 
Controls
 
Odds Ratioa 95% Confidence Interval P for Interactionb 
Mean (SD) No. Mean (SD) No. 
Duration of smoking (continuous), years         0.98 
    Rapid + intermediate acetylators 23.9 (13.0)   23 (12.9)   1.01 1.00, 1.01  
    Slow acetylators 24.0 (13.2)   23.2 (13.1)   1.01 1.00, 1.01  
Pack-years of smoking (continuous)         0.87 
    Rapid + intermediate acetylators 20.0 (18.0)   18.7 (17.2)   1.01 1.00, 1.01  
    Slow acetylators 20.7 (19.1)   19.0 (17.3)   1.01 1.00, 1.01  
Duration of smoking between menarche and first full-term pregnancy (continuous), years         0.03 
    Rapid + intermediate acetylators 7.4 (7.7)   7.7 (8.1)   1.00 1.00, 1.00  
    Slow acetylators 7.4 (7.7)   7.9 (8.6)   1.00 1.00, 1.00  
Pack-years of smoking between menarche and first full-term pregnancy (continuous)         0.14 
    Rapid + intermediate acetylators 5.4 (7.5)   5.3 (6.9)   1.01 1.00, 1.02  
    Slow acetylators 5.5 (7.4)   5.6 (7.5)   1.00 1.00, 1.01  
Duration of smoking (categorical), years         0.97 
    Rapid + intermediate acetylators          
        Never smoking  1,431 51.7  2,165 55.4 Reference  
        ≤15  410 14.8  559 14.3 1.14 0.98, 1.32  
        >15  927 33.5  1,182 30.3 1.17 1.04, 1.30  
    Slow acetylators          
        Never smoking  2,090 51.5  3,205 54.8 Reference  
        ≤15  606 14.9  865 14.8 1.07 0.95, 1.21  
        >15  1,362 33.6  1,775 30.4 1.18 1.07, 1.29  
Pack-years of smoking (categorical)         0.87 
    Rapid + intermediate acetylators          
        Never smoking  1,431 52.7  2,163 56.5 Reference  
        ≤20  772 28.4  1,045 27.3 1.13 1.01, 1.27  
        >20  514 18.9  619 16.7 1.24 1.08, 1.42  
    Slow acetylators          
        Never smoking  2,090 52.8  3,205 56.1 Reference  
        ≤20  1,108 28.0  1,571 27.5 1.08 0.92, 1.19  
        >20  757 19.1  936 16.4 1.25 1.11, 1.39  

Abbreviations: NAT2, N-acetyltransferase 2; SD, standard deviation.

a

Logistic regression was used to estimate odds ratios and 95% confidence intervals, adjusted for age at baseline, body mass index at baseline, ever use of menopausal hormone therapy at baseline, parity, and cohort. For continuous variables, persons with no exposure comprised the reference category.

b

Interaction P values were tested using a multiplicative interaction term in the models.

Upon examining smoking duration and pack-years during the period between menarche and first full-term pregnancy (or menopause for nulliparous women), only the pack-years variable was marginally associated with breast cancer risk (odds ratio = 1.01, 95% confidence interval: 1.00, 1.01; Table 3). No differences in this association were observed by NAT2 acetylation phenotype status. Stratifying analyses by menopausal status at baseline did not reveal any differences in the association between smoking and breast cancer risk by NAT2 acetylation phenotype status in either menopausal group (Table 4). Similarly, no associations with breast cancer risk were observed for rs12914385 or rs8034191—SNPs reported to be associated with nicotine addiction and lung cancer risk, respectively, in previous genome-wide association studies on lung cancer (7, 8). In addition, these SNPs did not interact statistically with NAT2 genotype (Appendix Table 1) to influence breast cancer risk.

Table 4.

Relation Between Cigarette Smoking and N-Acetyltranserase 2 Phenotype in Breast Cancer Risk, According to Menopausal Status at Baseline, Breast and Prostate Cancer Cohort Consortium, 1989–2006

Smoking History, Menopausal Status, and NAT2 Phenotype Mean (SD)
 
Odds Ratioa 95% Confidence Interval P for Interactionb 
Cases Controls 
Duration of smoking (continuous), years      
    Premenopausal     0.99 
        Rapid + intermediate acetylators 8.2 (11.5) 6.8 (10.1) 1.00 0.99, 1.01  
        Slow acetylators 8.0 (11.3) 7.1 (10.5) 1.00 0.99, 1.01  
    Postmenopausal     0.71 
        Rapid + intermediate acetylators 12.6 (15.8) 11.6 (15.1) 1.00 1.00, 1.01  
        Slow acetylators 10.8 (13.6) 11.7 (15.6) 1.01 1.00, 1.01  
Pack-years of smoking      
    Premenopausal     0.98 
        Rapid + intermediate acetylators 5.6 (10.4) 4.8 (9.1) 1.00 0.99, 1.01  
        Slow acetylators 5.6 (10.3) 5.0 (9.7) 1.00 0.99, 1.01  
    Postmenopausal     0.65 
        Rapid + intermediate acetylators 10.6 (17.3) 9.4 (16.1) 1.01 1.00, 1.01  
        Slow acetylators 11.1 (18.3) 9.6 (16.3) 1.01 1.00, 1.01  
Smoking History, Menopausal Status, and NAT2 Phenotype Mean (SD)
 
Odds Ratioa 95% Confidence Interval P for Interactionb 
Cases Controls 
Duration of smoking (continuous), years      
    Premenopausal     0.99 
        Rapid + intermediate acetylators 8.2 (11.5) 6.8 (10.1) 1.00 0.99, 1.01  
        Slow acetylators 8.0 (11.3) 7.1 (10.5) 1.00 0.99, 1.01  
    Postmenopausal     0.71 
        Rapid + intermediate acetylators 12.6 (15.8) 11.6 (15.1) 1.00 1.00, 1.01  
        Slow acetylators 10.8 (13.6) 11.7 (15.6) 1.01 1.00, 1.01  
Pack-years of smoking      
    Premenopausal     0.98 
        Rapid + intermediate acetylators 5.6 (10.4) 4.8 (9.1) 1.00 0.99, 1.01  
        Slow acetylators 5.6 (10.3) 5.0 (9.7) 1.00 0.99, 1.01  
    Postmenopausal     0.65 
        Rapid + intermediate acetylators 10.6 (17.3) 9.4 (16.1) 1.01 1.00, 1.01  
        Slow acetylators 11.1 (18.3) 9.6 (16.3) 1.01 1.00, 1.01  

Abbreviations: NAT2, N-acetyltransferase 2; SD, standard deviation.

a

Logistic regression was used to estimate odds ratios and 95% confidence intervals, adjusted for age at baseline, body mass index at baseline, ever use of menopausal hormone therapy at baseline, parity, and cohort. Persons with no exposure comprised the reference category.

b

Interaction P values were tested using a multiplicative interaction term in the models.

DISCUSSION

Active smoking is designated as being causally related to cancer at 17 or more sites or subsites. Our study found a modest association between breast cancer and both duration and pack-years of active cigarette smoking but no evidence of modification of the smoking relation by NAT2 genotype.

The modest association that we observed between breast cancer and long-term smoking is inconclusive, partly because our study was not sufficiently large to restrict our analyses to nondrinkers. Smoking and alcohol consumption are closely related in terms of both duration and intensity. An influential collaborative analysis restricted to nondrinkers found no association between ever smoking and breast cancer (12), and on the basis of this finding, the investigators proposed that future analyses of this issue be restricted to nondrinkers to avoid residual confounding by alcohol consumption. However, interpretation of that report’s results has been criticized because of the crude definition of smoking used (ever vs. never) and the failure to consider smoking status, duration of smoking, or initiation during periods of potentially greater susceptibility (4). Our analyses also could not control for physical inactivity, which is associated with both long-term smoking and breast cancer risk and could confound the observed association.

Several genetic polymorphisms have been proposed to modify associations between tobacco exposure and cancer risk. The evidence for effect modification by common NAT2 polymorphisms on the risk of bladder cancer from tobacco exposures is strong (13–17). This association is attributed to incompletely N-acetylated (detoxified) aromatic amines from tobacco smoke being held in the bladder before being eliminated. On the basis of these findings for bladder cancer, it has been hypothesized that NAT2 polymorphisms might also modify the association between tobacco exposure and breast cancer risk; it would be modified similarly by these same polymorphisms. Furthermore, the inconsistencies observed among studies with respect to an association between cigarette smoking and breast cancer risk may be due to the existence of subpopulations that are more susceptible to the effects of cigarette smoking based on genetic or other environmental exposures.

In a recent meta-analysis including 6,758 breast cancer cases, Ambrosone et al. (5) observed significant associations between cigarette smoking and breast cancer risk only among women carrying the slow acetylation phenotype. However, in our analyses we did not see any modification of the effect of smoking on breast cancer risk by this phenotype. Indeed, the odds ratios for long-term smoking, measured by either duration or pack-years, were almost equivalent in women with the two acetylation phenotypes.

It is not entirely clear why our results differed from those of the pooled analysis by Ambrosone et al. (5). Our analyses included 6,900 breast cancer cases and was therefore slightly larger than the pooled analyses of Ambrosone et al. (4,264 cases). The cohort investigators in our study assessed smoking history prospectively, before the diagnosis of breast cancer, whereas most of the studies included in the pooled analyses of Ambrosone et al. were retrospective case-control studies. It is plausible that recall bias might differentially affect the reporting of smoking histories by cases and controls but not that it would differentially affect the genetic subgroups. The percentages of never smokers were similar among the cases in our study (52%) and in the pooled analysis by Ambrosone et al. (47%), as were the percentages of controls (55% and 50%, respectively, as calculated from Ambrosone et al.’s Table 4 (5)). Sampling variation with respect to NAT2 polymorphisms may be greater in small, geographically dispersed case-control studies than in a small number of large cohort studies. The most likely explanation for the divergent results, other than chance, is that the studies included by Ambrosone et al. were all drawn from published articles on this issue and were therefore more susceptible to publication bias than were the studies included in our analysis.

Our results do not support the prior reports of effect modification by NAT2 acetylation phenotypes of the association between tobacco exposure and breast cancer risk. Given our large sample size, we can exclude all but very small interaction risk estimates. This study had greater than 85% power to detect an interaction odds ratio of 1.1 for a continuous environmental exposure at a relatively conservative alpha value of 0.01, under the assumption of modest associations with smoking and no overall association between NAT2 acetylation phenotype and breast cancer risk.

It has also been hypothesized that smoking exposure during the period between menarche and first full-term pregnancy, a period of mammary gland development, may be particularly relevant to breast cancer risk and therefore may be modified by NAT2 phenotype. Similar to our overall findings, we did not see any effect modification by NAT2 acetylation phenotype of the risk of breast cancer conferred by tobacco exposures incurred during this critical life period.

Discordance between studies of cigarette smoking and breast cancer risk may still be explained by environmental, lifestyle, or genetic factors. However, we did not observe any modification by NAT2 acetylation phenotype on the putative risk of breast cancer associated with tobacco exposure. Our results suggest that it is not necessary to assess common variations in NAT2 genotype in order to evaluate whether a smoking-breast cancer association exists. Despite the lack of modification of risk by NAT2 acetylation phenotype, reducing exposure to cigarette smoke remains of great importance to reducing risks of various diseases, including many cancers.

Abbreviations

    Abbreviations
  • Cal/EPA

    California Environmental Protection Agency

  • IARC

    International Agency for Research on Cancer

  • NAT2

    N-acetyltransferase 2

  • SNP

    single nucleotide polymorphism

Author affiliations: Lyon Cancer Research Center, Unité 1052, Institut National de la Santé et la Recherche Médicale (INSERM), Lyon, France (David G. Cox); School of Public Health, Faculty of Medicine, Imperial College London, London, United Kingdom (David G. Cox); Division of Clinical Epidemiology, German Cancer Research Center, Heidelberg, Germany (Lucie Dostal); Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts (David J. Hunter); Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, Maryland (Robert Hoover, Regina G. Ziegler); Epidemiology Program, University of Hawai’i Cancer Center, University of Hawai’i, Honolulu, Hawaii (Loïc Le Marchand); and American Cancer Society, Atlanta, Georgia (Michael J. Thun).

This work was funded by the National Cancer Institute (cooperative agreements U01-CA98233 with Dr. David J. Hunter, U01-CA98710 with Dr. Michael J. Thun, U01-CA98216 with Drs. Elio Riboli and Rudolf Kaaks, and U01-CA98758 with Dr. Brian E. Henderson) and by the Intramural Research Program of the National Institutes of Health/National Cancer Institute, Division of Cancer Epidemiology and Genetics. Dr. David G. Cox is the recipient of a grant from INSERM and the French Institut National du Cancer.

The Breast and Prostate Cancer Cohort Consortium: American Cancer Society—W. Ryan Diver, Victoria L. Stevens (American Cancer Society, Atlanta, Georgia); European Prospective Investigation into Cancer and Nutrition—Pilar Amiano (Subdireccion de Salud Publica de Gipuskoa Gobierno Vasco, San Sebastian, Spain); Marie-Christine Boutron-Rualt (Centre for Research in Epidemiology and Population Health, INSERM U1018, Institut Gustave Roussy, Villejuif, France, and UMRS 1018, Paris South University, Villejuif, France); Daniele Campa (Genomic Epidemiology Group, German Cancer Research Center, Heidelberg, Germany); Fränzel J. B. van Duijnhoven (National Institute for Public Health and the Environment, Bilthoven, the Netherlands); Inger T. Gram (University of Tromsø, Tromsø, Norway); Rudolf Kaaks (Division of Clinical Epidemiology, German Cancer Research Center, Heidelberg, Germany); Kay-Tee Khaw (Clinical Gerontology Unit, Department of Public Health and Primary Care, School of Clinical Medicine, University of Cambridge, Cambridge, United Kingdom); Elio Riboli (School of Public Health, Faculty of Medicine, Imperial College London, London, United Kingdom); Malin Sund (Department of Surgical and Perioperative Sciences, Surgery and Public Health, Nutrition Research, Umea University, Umea, Sweden); Demitrios Trichopoulos (Department of Epidemiology, Harvard School of Public Health (HSPH), Boston, Massachusetts, and Bureau of Epidemiologic Research, Academy of Athens, Athens, Greece); Rosario Tumino (Cancer Registry and Histopathology Unit, Azienda Ospedaliera “Civile M. P. Arezzo,” Ragusa, Italy); Ulla Vogel (National Research Centre for the Working Environment, Copenhagen, Denmark, and National Food Institute, Technical University of Denmark, Søborg, Denmark); Harvard University—Peter Kraft (Program in Molecular and Genetic Epidemiology, Department of Epidemiology, HSPH, and Department of Biostatistics, HSPH, Boston, Massachusetts); Julie E. Buring (Department of Epidemiology, HSPH, and Divisions of Preventive Medicine and Aging, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School (BWH/HMS), Boston, Massachusetts); Susan E. Hankinson (Department of Epidemiology, HSPH, and Channing Laboratory, Department of Medicine, BWH/HMS, Boston, Massachusetts); I-Min Lee (Department of Epidemiology, HSPH, and Divisions of Preventive Medicine and Aging, Department of Medicine, BWH/HMS, Boston, Massachusetts); Shumin M. Zhang (Divisions of Preventive Medicine and Aging, Department of Medicine, BWH/HMS, Boston, Massachusetts); Sara Lindstrom (Program in Molecular and Genetic Epidemiology, Department of Epidemiology, HSPH, Boston, Massachusetts); US National Cancer Institute—Christine D. Berg (Division of Cancer Prevention, National Cancer Institute, Bethesda, Maryland); Stephen Chanock (Division of Cancer Genetics and Epidemiology, National Cancer Institute, Bethesda, Maryland); Claudine Isaacs (Lombardi Comprehensive Cancer Center, School of Medicine, Georgetown University, Washington, DC); Catherine McCarty (Center for Human Genetics, Marshfield Clinic Research Foundation, Marshfield, Wisconsin); University of Southern California—Christopher A. Haiman, Brian E. Henderson (Department of Preventive Medicine, Keck School of Medicine, University of Southern California, Los Angeles, California).

Conflict of interest: none declared.

Appendix Table 1.

Relation Between Single Nucleotide Polymorphisms Previously Found to Be Associated With Lung Cancera and Breast Cancer Risk, Breast and Prostate Cancer Cohort Consortium, 1989–2006

SNP and Genotype Cases
 
Controls
 
Odds Ratiob 95% Confidence Interval P for Interactionc 
No. No. 
rs12914385d       0.37 
    C/C 2,706 39.1 3,766 39.1 Reference  
    C/T 3,185 46.1 4,607 46.5 0.96 0.90, 1.03  
    T/T 1,025 14.8 1,533 15.5 0.93 0.85, 1.02  
rs8034191e       0.11 
    T/T 3,078 44.4 4,311 43.4 Reference  
    T/C 3,045 43.9 4,372 44.1 0.98 0.91, 1.05  
    C/C 809 11.7 1,242 12.1 0.92 0.83, 1.02  
SNP and Genotype Cases
 
Controls
 
Odds Ratiob 95% Confidence Interval P for Interactionc 
No. No. 
rs12914385d       0.37 
    C/C 2,706 39.1 3,766 39.1 Reference  
    C/T 3,185 46.1 4,607 46.5 0.96 0.90, 1.03  
    T/T 1,025 14.8 1,533 15.5 0.93 0.85, 1.02  
rs8034191e       0.11 
    T/T 3,078 44.4 4,311 43.4 Reference  
    T/C 3,045 43.9 4,372 44.1 0.98 0.91, 1.05  
    C/C 809 11.7 1,242 12.1 0.92 0.83, 1.02  

Abbreviations: NAT2, N-acetyltransferase 2; SNP, single nucleotide polymorphism.

a

SNPs reported to be associated with nicotine addiction and lung cancer risk, respectively, in genome-wide association studies (7, 8).

b

Logistic regression was used to estimate odds ratios and 95% confidence intervals, adjusted for age at baseline, body mass index at baseline, ever use of menopausal hormone therapy at baseline, parity, and cohort.

c

P for interaction between SNP and NAT2 phenotype. Interactions were tested using a multiplicative interaction term in the models.

d

P-trend = 0.11.

e

P-trend = 0.15.

References

1.
International Agency for Research on Cancer
Tobacco Smoke and Involuntary Smoking.
  
(IARC monographs on the evaluation of carcinogenic risks to humans, vol 83). Lyon, France: International Agency for Research on Cancer; 2002)
2.
Secretan
B
Straif
K
Baan
R
, et al.  . 
A review of human carcinogens—part E: tobacco, areca nut, alcohol, coal smoke, and salted fish
Lancet Oncol.
 , 
2009
, vol. 
10
 
11
(pg. 
1033
-
1034
)
3.
Office of Environmental Health Hazard Assessment and California Air Resources Board
California Environmental Protection Agency
Proposed Identification of Environmental Tobacco Smoke as Toxic Air Contaminant
 , 
2005
Sacramento, CA
California Environmental Protection Agency
4.
Collishaw
NE
Boyd
NF
Cantor
KP
, et al.  . 
Canadian Expert Panel on Tobacco Smoke and Breast Cancer Risk
 , 
2009
Toronto, Ontario, Canada
Ontario Tobacco Research Unit
5.
Ambrosone
CB
Kropp
S
Yang
J
, et al.  . 
Cigarette smoking, N-acetyltransferase 2 genotypes, and breast cancer risk: pooled analysis and meta-analysis
Cancer Epidemiol Biomarkers Prev.
 , 
2008
, vol. 
17
 
1
(pg. 
15
-
26
)
6.
Hunter
DJ
Riboli
E
Haiman
CA
, et al.  . 
A candidate gene approach to searching for low-penetrance breast and prostate cancer genes
Nat Rev Cancer.
 , 
2005
, vol. 
5
 
12
(pg. 
977
-
985
)
7.
Broderick
P
Wang
Y
Vijayakrishnan
J
, et al.  . 
Deciphering the impact of common genetic variation on lung cancer risk: a genome-wide association study
Cancer Res.
 , 
2009
, vol. 
69
 
16
(pg. 
6633
-
6641
)
8.
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
, vol. 
40
 
5
(pg. 
616
-
622
)
9.
Kuznetsov
IB
McDuffie
M
Moslehi
R
A web server for inferring the human N-acetyltransferase-2 (NAT2) enzymatic phenotype from NAT2 genotype
Bioinformatics.
 , 
2009
, vol. 
25
 
9
(pg. 
1185
-
1186
)
10.
Gauderman
WJ
Sample size requirements for matched case-control studies of gene-environment interaction
Stat Med.
 , 
2002
, vol. 
21
 
1
(pg. 
35
-
50
)
11.
Gauderman
WJ
Morrison
JM
QUANTO 1.1: A Computer Program for Power and Sample Size Calculations for Genetic-Epidemiology Studies
 , 
2006
Los Angeles, CA
University of Southern California
 
12.
Hamajima
N
Hirose
K
Tajima
K
, et al.  . 
Alcohol, tobacco and breast cancer—collaborative reanalysis of individual data from 53 epidemiological studies, including 58,515 women with breast cancer and 95,067 women without the disease
Br J Cancer.
 , 
2002
, vol. 
87
 
11
(pg. 
1234
-
1245
)
13.
Gu
J
Liang
D
Wang
Y
, et al.  . 
Effects of N-acetyl transferase 1 and 2 polymorphisms on bladder cancer risk in Caucasians
Mutat Res.
 , 
2005
, vol. 
581
 
1-2
(pg. 
97
-
104
)
14.
Hirao
Y
Kim
WJ
Fujimoto
K
Environmental factors promoting bladder cancer
Curr Opin Urol.
 , 
2009
, vol. 
19
 
5
(pg. 
494
-
499
)
15.
Hung
RJ
Boffetta
P
Brennan
P
, et al.  . 
GST, NAT, SULT1A1, CYP1B1 genetic polymorphisms, interactions with environmental exposures and bladder cancer risk in a high-risk population
Int J Cancer.
 , 
2004
, vol. 
110
 
4
(pg. 
598
-
604
)
16.
Lubin
JH
Kogevinas
M
Silverman
D
, et al.  . 
Evidence for an intensity-dependent interaction of NAT2 acetylation genotype and cigarette smoking in the Spanish Bladder Cancer Study
Int J Epidemiol.
 , 
2007
, vol. 
36
 
1
(pg. 
236
-
241
)
17.
Ma
QW
Lin
GF
Chen
JG
, et al.  . 
Polymorphism of N-acetyltransferase 2 (NAT2) gene polymorphism in Shanghai population: occupational and non-occupational bladder cancer patient groups
Biomed Environ Sci.
 , 
2004
, vol. 
17
 
3
(pg. 
291
-
298
)