Abstract

The N-acetyltransferase 2 gene (NAT2) product is an enzyme important in carcinogen metabolism via activation and detoxification pathways. Therefore, NAT2 variants may represent underlying susceptibility to breast cancer. Because a number of studies of the association of NAT2 with breast cancer have been published, the authors performed a meta-analysis. They extracted all relevant data to examine evidence for a main effect (i.e., the effect in a model that does not include any interactions) of NAT2 phenotype and genotype on breast cancer risk. They summarized the evidence for modification by smoking and meat intake, sources of exposure to aromatic and heterocyclic amines, respectively, which are metabolized by NAT2. The authors identified seven studies that measured NAT2 phenotype and 20 studies that deduced phenotype via genotyping. They found no evidence for heterogeneity (Cochran's Q statistic p = 0.74) and no statistically significant increased risk from NAT2 acetylation (slow/rapid) for breast cancer (summary odds ratio = 1.02, 95% confidence interval: 0.95, 1.08). These results suggest that there is no overall association between the NAT2 slow- or rapid-acetylation phenotype and breast cancer risk. However, some evidence suggests that smoking may modify this association.

GENE

The N-acetyltransferase 2 gene (NAT2) is located on chromosome 8p21.3-23.1 and codes for a phase II xenobiotic metabolizing enzyme (1). The NAT2 enzyme plays an important role in the metabolism of both aromatic and heterocyclic amines, via N- and O-acetylation, which are pathways responsible for deactivation and activation, respectively (2).

GENE VARIANTS AND FREQUENCY

A previous NAT2 human genome epidemiology (HuGE) review detailed gene variants and their population frequencies (3). Briefly, NAT2 phenotypes are characterized as being slow, intermediate, or rapid acetylators, which refers to their ability to metabolize or activate xenobiotics.

DISEASE(S) OR OTHER OUTCOMES

Previous epidemiologic work implicates NAT2 variants in carcinogenesis, especially for bladder, colon, and breast cancer (3–5). This finding is biologically plausible owing to the ability of the NAT2 enzyme to N-acetylate or O-acetylate carcinogens to inert or noninert compounds capable of forming DNA adducts. Cigarette smoking and intake of well-done meat involve exposure to aryl and heterocyclic amines, substrates of NAT2; therefore, it is plausible that underlying NAT2 genotype may modify risk of cancer based on the ability to activate or detoxify heterocyclic and aromatic amines (6, 7).

OBJECTIVES OF THE CURRENT REVIEW

Since there have been a number of published reports on NAT2 and breast cancer, we conducted a meta-analysis to summarize the results of the effect on breast cancer of the NAT2 acetylation phenotype and NAT2 genotype and also to discuss evidence for an interaction of NAT2 with exposure to smoking, meat intake, and meat cooking method.

METHODS

Study selection and inclusion criteria

We searched for relevant papers published before August 2006 by using the Human Genome Epidemiology Network (HuGENet) and the MEDLINE database (National Library of Medicine, Bethesda, Maryland). We searched MEDLINE by using the following terms: “arylamine N-acetyltransferase,” “acetyltransferases,” “NAT2,” “genetic polymorphism,” “restriction fragment polymorphism,” “single nucleotide polymorphism,” “breast cancer,” and “breast neoplasms.” We included all articles involving case-control or nested case-control studies that investigated NAT2 (determined by using phenotyping or genotyping methods) and breast cancer risk. We also searched the reference lists of the published studies found this way.

Data extraction

We extracted the following information from each manuscript: year of publication, country, number of cases and controls, matching, phenotyping technique, genotyping technique, alleles measured, allele and genotype frequencies, method for phenotype classification, phenotype frequencies, mean age of case and controls, menopausal status, covariates, and results by smoking or meat intake. We also extracted risk estimates for gene-environment interactions.

Classification of phenotype

Because of the inherent differences in methods for measuring NAT2 status—measuring phenotypes by using metabolic response to a particular compound or measuring alleles directly—we did not combine these studies for analysis. Respectively, these two types of studies are herein referred to as “phenotype” or “genotype” studies. Subsequently, we summarized phenotype frequencies (slow and rapid status). In the published genotype studies, individuals were classified as having a slow phenotype if homozygous for any of the slow-activity alleles and as having a rapid phenotype if homozygous or heterozygous for wild-type alleles. In some cases, we reclassified intermediate acetylators as rapid when the authors provided frequencies for an intermediate category (having one slow and one rapid allele) (6).

Statistical analysis

We performed a meta-analysis separately for phenotype and genotype studies. For each analysis, we investigated among-study heterogeneity by using Cochran's Q statistic, and we examined fixed- and random-effect models based on the method of DerSimonian and Laird (8) in Number Cruncher Statistical software (9). We constructed a funnel plot to examine the influence of publication bias.

RESULTS

Phenotype studies

Seven studies were identified that analyzed differences in the proportion of cases and controls having a slow-acetylator phenotype (10–16). Table 1 summarizes these studies. Six were conducted in Europe, and the samples were primarily Caucasian. The sample size ranged from 79 to 515, totaling 1,330 women. Six studies measured acetylation phenotype via administration of arylamine drugs (four sulfamethazine, two dapsone); the other used isoniazid. We pooled the numbers across these studies and calculated the prevalence of the slow-acetylator phenotype to be 56 percent overall: 52 percent in cases and 58 percent in controls.

TABLE 1.

Description of phenotype-only studies included in a meta-analysis of the association of N-acetyltransferase 2 variants with breast cancer

First author, year (reference no.) Geographic region Source/type of cases Source/type of controls No. of cases No. of controls Method for phenotype determination % with the slow phenotype 
Cases Controls 
Bulovskaya et al., 1978 (10Russia Advanced breast cancer Healthy, but some with age-associated cardiovascular disturbances 41 38 Sulfamethazine (not given) 31.7 63.2 
Cartwright et al., 1984 (11United Kingdom Invasive breast cancer Not given 93 112 Dapsone (slow: MADDS/DDS* <0.33) 45.2 58.0 
Ladero et al., 1987 (13Spain Histologically confirmed Healthy or with disease not related to acetylator phenotype 81 75 Sulfamethazine (slow: ACMZ/SMZ*<0.45 in plasma or <0.77 in urine) 60.5 60.0 
Philip et al., 1987 (14United Kingdom Malignant breast disease Cardiovascular disease and normal 181 337 Dapsone (slow: MADDS/DDS <0.30) 54.7 55.2 
Webster et al., 1989 (16United Kingdom Histologically confirmed 68 healthy volunteers, 32 with breast abnormalities 100 100 Isoniazid (slow: ACINH/INH* <1.5) 57.0 59.0 
Sardas et al., 1990 (15Turkey Advanced breast cancer Healthy volunteers 28 51 Sulfamethazine (slow: not given) 39.3 64.7 
Ilett et al., 1990 (12Australia Surgical resection of primary breast carcinoma No breast disease 45 48 Sulfamethazine (slow: ACMZ/SMZ <0.6) 55.6 64.6 
First author, year (reference no.) Geographic region Source/type of cases Source/type of controls No. of cases No. of controls Method for phenotype determination % with the slow phenotype 
Cases Controls 
Bulovskaya et al., 1978 (10Russia Advanced breast cancer Healthy, but some with age-associated cardiovascular disturbances 41 38 Sulfamethazine (not given) 31.7 63.2 
Cartwright et al., 1984 (11United Kingdom Invasive breast cancer Not given 93 112 Dapsone (slow: MADDS/DDS* <0.33) 45.2 58.0 
Ladero et al., 1987 (13Spain Histologically confirmed Healthy or with disease not related to acetylator phenotype 81 75 Sulfamethazine (slow: ACMZ/SMZ*<0.45 in plasma or <0.77 in urine) 60.5 60.0 
Philip et al., 1987 (14United Kingdom Malignant breast disease Cardiovascular disease and normal 181 337 Dapsone (slow: MADDS/DDS <0.30) 54.7 55.2 
Webster et al., 1989 (16United Kingdom Histologically confirmed 68 healthy volunteers, 32 with breast abnormalities 100 100 Isoniazid (slow: ACINH/INH* <1.5) 57.0 59.0 
Sardas et al., 1990 (15Turkey Advanced breast cancer Healthy volunteers 28 51 Sulfamethazine (slow: not given) 39.3 64.7 
Ilett et al., 1990 (12Australia Surgical resection of primary breast carcinoma No breast disease 45 48 Sulfamethazine (slow: ACMZ/SMZ <0.6) 55.6 64.6 
*

MADDS/DDS, monoacetyldapsone/dapsone ratio; ACMZ/SMZ, N-acetylmethazine/methazine ratio; ACIHN/INH, acetylisoniazid/isoniazid ratio.

Three of the seven studies reported a higher prevalence of the rapid-acetylator phenotype in the breast cancer cases (10, 11, 15), but only two of the studies were statistically significant (10, 15). For the meta-analysis, the test for heterogeneity was not statistically significant at p < 0.05 (Q statistic = 10.9, p = 0.09). Figure 1 shows the individual study odds as well as the summary odds ratio (0.77, 95 percent confidence interval: 0.61, 0.96) for the combined phenotype studies (where the reference group is rapid acetylators) using the fixed-effect model.

FIGURE 1.

Summary plot for phenotype studies (first author, date of publication (reference no.)) and combined estimate of odds ratio and 95% confidence interval. Reference group is rapid acetylators. Studies are sorted according to odds ratio. The sizes of the symbols are proportional to the study sizes.

FIGURE 1.

Summary plot for phenotype studies (first author, date of publication (reference no.)) and combined estimate of odds ratio and 95% confidence interval. Reference group is rapid acetylators. Studies are sorted according to odds ratio. The sizes of the symbols are proportional to the study sizes.

Genotype studies

Two duplication studies and two abstracts were identified that were later published as full manuscripts (17–20). These two particular full manuscripts were included in the meta-analysis. Another study, a case-cohort study (21), was brought to our attention by the reviewers of our paper. Excluding overlapping papers and including the case-cohort study, we identified 15 case-control studies, four nested case-control studies, and one case-cohort study, 20 in total (21–40). Table 2 summarizes these studies. Ten studies were population based, and eight were hospital based. Most studies were conducted among Caucasian populations. The only study that included a larger sample of African-American women was published by Millikan et al. (35). Three studies were conducted among Asian populations (two in China and one in Korea) (29, 33, 39). Two recruited women based on menopausal status; the Iowa Women's Health Study recruited postmenopausal women only (26), while Matheson et al. (34) included only premenopausal women. Of the 16,091 women included in the 20 studies, there were 7,479 cases and 8,612 controls.

TABLE 2.

Description of genotype studies included in a meta-analysis of the association of N-acetyltransferase 2 variants with breast cancer

First author, year (reference no.) Geographic region Study type Source of cases Source of controls No. of cases No. of controls Matching 
Agundez et al., 1995 (22Spain Case-control Not given Healthy women from the same region 160 132 None 
Ambrosone et al., 1996 (24New York Case-control Erie and Niagara Counties, New York Randomly selected using the DMV* and HCFA* 304 327 Age and county 
Hunter et al., 1997 (30United States Nested case-control (Nurses' Health Study) From cohort Cancer free from cohort 465 466 Year of birth, menopausal status, month and time of blood draw, fasting status at blood draw, hormone use 
Millikan et al., 1998 (35North Carolina Case-control Primary, invasive, North Carolina Central Cancer Registry Randomly selected using the DMV and HCFA 488 472 None 
Huang et al., 1999 (29Taiwan, Republic of China Case-control One hospital in Taiwan Hospital based, randomly selected 139 133 None 
Delfino et al., 2000 (27California Case-control Breast centers in Orange County, California Subjects with benign masses recruited from the same breast centers 113 107 None 
Morabia et al., 2000 (36Geneva, Switzerland Case-control Surviving cases who remained residents of Geneva (from an earlier study) Surviving controls who remained residents of Geneva (from an earlier study) 160 162 Age (within 10 years) 
Deitz et al., 2000 (26Iowa Nested case-control (Iowa Women's Health Study) Postmenopausal, from cohort Postmenopausal and cancer free from cohort 174 387 None 
Krajinovic et al., 2001 (32Montreal, Canada Case-control Three hospitals in Montreal Randomly selected from Montreal and a hospital DNA bank 149 203 Neighborhood 
Wu et al., 2002 (39Taiwan, Republic of China Case-control One hospital in Taiwan Cancer free from the same area 60 60 Age, smoking, family history (first-degree relative) 
Matheson et al., 2002 (34Melbourne, Australia Case-control Premenopausal, recruited from Victoria and New South Wales cancer registries Premenopausal, randomly selected from electoral rolls 157 157 Age range 
Chang-Claude et al., 2002 (25Southern Germany Case-control All hospitals in two regions Randomly selected from the same regions 422 887 Age (within 5 years), study region 
Egan et al., 2003 (28Massachusetts, New Hampshire, and Wisconsin Case-control State registries Randomly selected using the DMV and HCFA 791 797 None 
Lee et al., 2003 (33Seoul, Korea Case-control Three hospitals in Seoul Hospital based, cancer free 251 288 None 
van der Hel et al., 2003 (38The Netherlands Nested case-control Netherlands Cancer Registry and regional registries Randomly selected from cohort 229 264 Age (within 5 years), menopausal status, residence 
Kocabas et al., 2004 (31Turkey (central and western) Case-control Three hospitals in Ankara University staff, students, and others from hospital, no family history of cancer 84 103 Age (within 10 years), time of blood draw, menopausal status 
Alberg et al., 2004 (23Maryland Nested case-control (Clue II cohort) Washington County cancer registry Cancer free from cohort 110 113 Age (within 1 year), race, menopausal status, day of menstrual cycle (premenopausal only), date of blood collection 
Sillanpaa et al., 2005 (37Finland Case-control One hospital in Finland Recruited from the National Population Register 478 479 None 
van der Hel et al., 2005 (21The Netherlands Case-cohort Regional and the Netherlands cancer registry Population-based screening cohort 845 875 None 
Lissowska et al., 2006 (40Warsaw and Łódź, Poland Case-control Five hospitals and the Cancer Registry Randomly selected residents of the same cities 1,900 2,200 Age (within 5 years), city 
First author, year (reference no.) Geographic region Study type Source of cases Source of controls No. of cases No. of controls Matching 
Agundez et al., 1995 (22Spain Case-control Not given Healthy women from the same region 160 132 None 
Ambrosone et al., 1996 (24New York Case-control Erie and Niagara Counties, New York Randomly selected using the DMV* and HCFA* 304 327 Age and county 
Hunter et al., 1997 (30United States Nested case-control (Nurses' Health Study) From cohort Cancer free from cohort 465 466 Year of birth, menopausal status, month and time of blood draw, fasting status at blood draw, hormone use 
Millikan et al., 1998 (35North Carolina Case-control Primary, invasive, North Carolina Central Cancer Registry Randomly selected using the DMV and HCFA 488 472 None 
Huang et al., 1999 (29Taiwan, Republic of China Case-control One hospital in Taiwan Hospital based, randomly selected 139 133 None 
Delfino et al., 2000 (27California Case-control Breast centers in Orange County, California Subjects with benign masses recruited from the same breast centers 113 107 None 
Morabia et al., 2000 (36Geneva, Switzerland Case-control Surviving cases who remained residents of Geneva (from an earlier study) Surviving controls who remained residents of Geneva (from an earlier study) 160 162 Age (within 10 years) 
Deitz et al., 2000 (26Iowa Nested case-control (Iowa Women's Health Study) Postmenopausal, from cohort Postmenopausal and cancer free from cohort 174 387 None 
Krajinovic et al., 2001 (32Montreal, Canada Case-control Three hospitals in Montreal Randomly selected from Montreal and a hospital DNA bank 149 203 Neighborhood 
Wu et al., 2002 (39Taiwan, Republic of China Case-control One hospital in Taiwan Cancer free from the same area 60 60 Age, smoking, family history (first-degree relative) 
Matheson et al., 2002 (34Melbourne, Australia Case-control Premenopausal, recruited from Victoria and New South Wales cancer registries Premenopausal, randomly selected from electoral rolls 157 157 Age range 
Chang-Claude et al., 2002 (25Southern Germany Case-control All hospitals in two regions Randomly selected from the same regions 422 887 Age (within 5 years), study region 
Egan et al., 2003 (28Massachusetts, New Hampshire, and Wisconsin Case-control State registries Randomly selected using the DMV and HCFA 791 797 None 
Lee et al., 2003 (33Seoul, Korea Case-control Three hospitals in Seoul Hospital based, cancer free 251 288 None 
van der Hel et al., 2003 (38The Netherlands Nested case-control Netherlands Cancer Registry and regional registries Randomly selected from cohort 229 264 Age (within 5 years), menopausal status, residence 
Kocabas et al., 2004 (31Turkey (central and western) Case-control Three hospitals in Ankara University staff, students, and others from hospital, no family history of cancer 84 103 Age (within 10 years), time of blood draw, menopausal status 
Alberg et al., 2004 (23Maryland Nested case-control (Clue II cohort) Washington County cancer registry Cancer free from cohort 110 113 Age (within 1 year), race, menopausal status, day of menstrual cycle (premenopausal only), date of blood collection 
Sillanpaa et al., 2005 (37Finland Case-control One hospital in Finland Recruited from the National Population Register 478 479 None 
van der Hel et al., 2005 (21The Netherlands Case-cohort Regional and the Netherlands cancer registry Population-based screening cohort 845 875 None 
Lissowska et al., 2006 (40Warsaw and Łódź, Poland Case-control Five hospitals and the Cancer Registry Randomly selected residents of the same cities 1,900 2,200 Age (within 5 years), city 
*

DMV, Department of Motor Vehicles; HCFA, Health Care Financing Administration.

Web table 1 shows the NAT2 alleles typed for each study as well as the frequency of slow- and rapid-acetylation phenotypes deduced from genotyping results by each author. (This information is described in the first of two supplementary tables; both are referred to as “Web table” in the text and are posted on the website of the Human Genome Epidemiology Network (http://www.cdc.gov/genomics/hugenet/reviews.htm) as well as on the Journal's website (http://aje.oupjournals.org/).) In spite of the differences in nomenclature across the 20 genotype studies included in the meta-analysis, there were few actual differences in the particular alleles measured. In Caucasians, the majority of variation in slow-acetylator phenotype is due to three polymorphic sites (41, 42). Each study similarly measured the three alleles responsible for the majority of the slow-acetylator phenotype and classified women as having slow acetylators if they had two slow alleles. Therefore, carrying one copy of the *4 allele resulted in classification as a rapid acetylator. Seven studies measured additional rapid alleles (*12 or *13) (21, 26, 31–33, 38, 40), while six measured the *14 allele as an additional slow allele (23, 26, 27, 29, 35, 37). Intermediate acetylators were reclassified as rapid for the purpose of this review and, in each case the definition of intermediate was the same, as having only one rapid allele. There is a confusion regarding differences in nomenclature for NAT2 alleles. An excellent review (6) as well as a website (www.louisville.edu/medschool/pharmacology/NAT.html) help to clarify the allele name with the specific nucleotide changes.

We reclassified intermediate acetylators as fast acetylators for the analysis in those studies in which the authors classified women into three categories (slow, intermediate, and rapid). Following this reclassification of intermediate phenotypes, we found that, overall, 54.4 percent of cases and 54.3 percent of controls were slow acetylators. From the 16 studies that included primarily Caucasian women and the Caucasian sample provided by Millikan et al. (35), we calculated prevalence of the slow genotype in Caucasians to be 56.9 percent: 57.0 percent in cases and 56.9 percent in controls. For the sample of African-American women, prevalence of the slow phenotype was 40 percent overall: 41 percent for cases and 40 percent for controls (35). In the three Asian studies, the pooled prevalence of the slow phenotype was 20 percent: 21 percent in cases and 18 percent in controls.

Associations

Web table 2 shows that three of the 20 studies reported a significant association between NAT2 acetylation and breast cancer. Alberg et al. (23) showed that rapid-acetylation status increased risk both in the whole sample and among postmenopausal women. Conversely, Sillanpaa et al. (37) in the Finland sample and Huang et al. (29) in the Taiwanese population reported an increased risk for slow acetylators, which may also be highest for postmenopausal women.

For the meta-analysis of the genotype studies, we found no evidence for among-study heterogeneity (Q statistic = 14.58, p = 0.75). Figure 2 shows the individual odds ratios, as well as the summary odds ratio (1.02) and 95 percent confidence interval (0.95, 1.08) for the main effect of NAT2 acetylator genotype obtained by using a fixed-effect model (8).

FIGURE 2.

Forest plot of genotype studies (first author, date of publication (reference no.)) and combined odds ratio and 95% confidence interval. Reference group is rapid acetylators: those with at least one rapid N-acetyltransferase 2 allele. Studies are sorted according to odds ratio. The sizes of the symbols are proportional to the study sizes.

FIGURE 2.

Forest plot of genotype studies (first author, date of publication (reference no.)) and combined odds ratio and 95% confidence interval. Reference group is rapid acetylators: those with at least one rapid N-acetyltransferase 2 allele. Studies are sorted according to odds ratio. The sizes of the symbols are proportional to the study sizes.

Figure 3 is a funnel plot showing the association of NAT2 with breast cancer by study type (phenotype or genotype) and study size. There does not appear to be an obvious publication bias toward positive or negative findings among the genotype studies. However, in the phenotype studies, there may be a publication bias toward manuscripts showing risk associated with rapid phenotypes because the plot is asymmetrical.

FIGURE 3.

Funnel plot of the association of the N-acetyltransferase 2 (NAT2) genotype with breast cancer, by study size. Odds ratios for the main effect of NAT2 are shown. Reference group is rapid acetylators.

FIGURE 3.

Funnel plot of the association of the N-acetyltransferase 2 (NAT2) genotype with breast cancer, by study size. Odds ratios for the main effect of NAT2 are shown. Reference group is rapid acetylators.

Smoking

In terms of a main effect of smoking, 15 of the 20 genotype studies analyzed smoking and breast cancer risk by using a variety of exposure variables: smoking status (active, former, passive smoking), cumulative exposure, duration, and intensity via number of cigarettes smoked per day (which may have been measured as many as 20 years prior to diagnosis). Of these 15 studies, nine found no significant association between smoking and breast cancer risk. Hunter et al. (30) reported increased risk for women who smoked 15 or more cigarettes per day (10 years prior to diagnosis) and women with 20–30 pack-years of exposure, a finding replicated by Millikan et al. (35), who reported increased risk of breast cancer with more than a 20-year duration of active smoking. Morabia et al. (36) reported increased risk for current and former smokers. Egan et al. (28) found increased risk for ever smokers and increasing pack-years; however, the association with increasing pack-years was significant for postmenopausal women only. Lissowska et al. (40) reported increased risk for women less than age 45 years in terms of smoking status, increasing number of cigarettes per day and duration, and age at which smoking started. None of the studies showed a decreased risk of breast cancer with active or passive smoking.

Interactions

Smoking.

Of the 15 genotype studies that measured smoking, 13 investigated the hypothesized interaction of smoking and NAT2 acetylator genotype. Some studies further stratified by menopausal status.

Among the studies that examined the NAT2 × smoking interaction according to menopausal status, Ambrosone et al. (24) was the first to show increased risk of breast cancer for slow-acetylating, postmenopausal women who were current or former smokers (using cigarettes/day 2 years ago, cigarettes/day 20 years ago, and lifetime pack-years). These findings were later replicated by Egan et al. (28) (pack-years) and Alberg et al. (23) (active smoking), who reported the strongest association among postmenopausal women. Sillanpaa et al. (37), van der Hel et al. (38), and Matheson et al. (34) reported significant increased risk for smokers who were slow acetylators.

Krajinovic et al. (32), Delfino et al. (27), Hunter et al. (30), and Millikan et al. (35) showed no significant evidence for modification of NAT2 and breast cancer by smoking. Krajinovic et al. reported a nonsignificant increase in risk among rapid acetylators who were current or former smokers, which became significant in the case-only analysis. Delfino et al. reported no trend in risk for rapid or slow acetylators, while Hunter et al.'s weak, nonsignificant trend was among slow acetylators. Millikan et al. reported an increased, yet nonsignificant trend in risk for duration of smoking among postmenopausal rapid acetylators. Interestingly, Chang-Claude et al. (25) reported increased risk of breast cancer in rapid acetylators exposed to passive smoke, a finding later replicated by Morabia et al. (36).

Two of the studies conducted a case-only analysis. Results from Ambrosone et al. (24) suggest that, among cases, the odds of being a slow acetylator were greatest for women who smoked more than 20 cigarettes per day (20 years ago), smoked for more than 20 pack-years, or initiated smoking at an earlier age (≤16 years). Conversely, Krajinovic et al. (32) reported increased odds for being a rapid acetylator for ever smokers.

Meat intake.

There is much published literature on the association of breast cancer with red meat intake and meat doneness (43). Five of the studies included in this review investigated potential modification of the association of NAT2 with breast cancer according to charred or well-done meat intake. The only study that found evidence for interaction between meat doneness and NAT2 status was that of Deitz et al. (26), who reported a significant increasing trend in risk among rapid acetylators with increasing meat doneness score.

Ambrosone et al. (44) reported no significant association between meat consumption and breast cancer, nor did they find evidence of a significant interaction between meat intake and NAT2 acetylation status. Similarly, Gertig et al. (45) found no association of red meat intake or meat cooking method with breast cancer risk in the Nurses' Health Study, and no significant interaction with NAT2. Delfino et al. (46) also found no evidence of increased risk for intake of well-cooked meat nor an interaction with NAT2. The last of the studies, by Krajinovic et al. (32), did not report a significant elevated risk for rapid acetylators who consumed well-done meat. In each of these studies, the trends in risk according to increasing meat intake or charred meat intake were not strong enough to suggest whether NAT2 slow or rapid acetylators are at increased or decreased risk.

DISCUSSION

Main findings

The results of this meta-analysis show no evidence for an overall effect of NAT2 acetylation capacity on the risk of breast cancer. There is suggestive evidence that acetylation may underlie a susceptibility to breast cancer if there is also exposure to tobacco smoke, but there is little evidence for effect modification by intake of well-done meat. Future studies evaluating gene variants in entire pathways may enable better evaluation of susceptibility to breast cancer in the presence of other factors.

There are a number of challenges in summarizing studies of NAT2 variants; there are not only differences in single nucleotide polymorphism selection and technology available at the time of the study but also a reliance on phenotypic characterization of NAT2 variants. Any misclassification of individuals as slow, intermediate, or rapid acetylators would more likely result in bias toward the null and therefore may partly explain the lack of an overall significant association between NAT2 acetylation status and breast cancer.

Following construction of a funnel plot (figure 3) of both phenotype and genotype studies, we conclude that there is some degree of publication bias in the phenotype studies. These studies were more likely to be published if they found a protective effect for slow acetylators. The distribution of odds ratio estimates in the genotype studies suggests that, in these publications, bias is less likely to have occurred.

Our original goal was to also conduct a quantitative review of the potential factors implicated in modifying the association of NAT2 with breast cancer, specifically smoking and meat intake. However, once we examined the studies and discovered the many different exposure assessments used to classify individuals, we concluded that the difficulty associated with summarizing the effect of different exposure variables would be better overcome by using a pooled design in which primary data are obtained from authors of published manuscripts. At the same time, we found that a pooled analysis, in which the exposure measurements can be better summarized and a more comprehensive analysis performed, is already under way.

Biology

Approximately 26 NAT2 variants are known to exist in humans, some silent and some functional variants responsible for changes in enzymatic activity. An excellent review and explanation of genotype-phenotype correlation was published by Hein et al. (6). This paper helps to clarify the confusion regarding NAT2 allelic nomenclature as well as the functional significance and classification of the different alleles.

For the phenotype studies, our meta-analysis suggests evidence of decreased breast cancer risk for women with a slow phenotype (odds ratio = 0.77, 95 percent confidence interval: 0.62, 0.96) or, conversely, increased breast cancer risk for women with a rapid phenotype (odds ratio = 1.29, 95 percent confidence interval: 1.04, 1.62) when slow is the reference group. The published reports also suggest increased risk for rapid acetylators; however, the majority of these studies were unable to report a significant association. Evans (47) combined the results of Bulovskaya et al. (10) and Cartwright (11) and found an odds ratio of 2.05, suggesting an increased risk for carriers of the rapid allele. These studies were of limited size, however, which may have increased sampling error. In addition, misclassification may also be important because different xenobiotics and methods were used to assess NAT2 activity. The NAT2 phenotype may be influenced by disease status; therefore, the results of the genotype studies may be more reliable because genotype is not affected by disease status.

The genotype studies purport a modifying effect of smoking in the association of NAT2 with breast cancer, especially in postmenopausal women. Postmenopausal women are more likely to have a greater duration of smoking simply because of their age, which may explain these findings. These studies therefore suggest a role for NAT2 in breast cancer carcinogenesis, especially for women who also smoke. This hypothesis can be supported by evidence from Pfau et al. (48), who showed that slow acetylators have a higher quantity of DNA adducts than rapid acetylators. A recent review and meta-analysis of the potential modifying effect of smoking suggests that, among slow acetylators, smoking may increase breast cancer risk, an effect that is strongest among postmenopausal women (odds ratio = 2.4, 95 percent confidence interval: 1.7, 3.3) (49), which agrees with our assessment of the studies included in our review.

The increased risk of breast cancer for women exposed to passive smoke who are also rapid acetylators is very interesting (25, 36) in that passive smoke contains 2-amino-α-carboline, a heterocyclic amine (50). If exposed to passive or sidestream smoke, rapid acetylators are at increased risk since they are more likely to O-acetylate the N-hydroxy heterocyclic amines that have been processed in the liver by CYP1A2 and subsequently form compounds capable of creating DNA adducts (50). Further studies are needed to confirm this association.

Biologically, slow acetylators who have lower concentrations of active NAT2 enzyme and are exposed to aromatic amines are more likely to undergo metabolism via the hydroxylation pathway. This process may then lead to electrophilic intermediates capable of binding DNA and initiating carcinogenesis, as compared with the process in rapid acetylators, whose aromatic amine exposure is metabolized via N-acetylation to become an inactive metabolite (51). Firozi et al. (52) reported a higher frequency of smoking-related adducts in breast cancer cases with the slow-acetylation phenotype, which suggests that the hypothesized increased risk of breast cancer and smoking is especially important for slow acetylators.

The varying action of NAT2 under different exposure scenarios may partially explain the null findings for the main effect of NAT2 and breast cancer. While smoking involves exposure to both aromatic amines and heterocyclic amines, N- and O-acetylation may be taking place. Thus, the competing or heterogeneous nature of the pathway depending on exposure may attenuate the association. In spite of this heterogeneity, many of the studies found evidence of increased risk for slow-acetylating women who also smoke.

Another exposure that has been investigated as a potential modifier of the NAT2 and breast cancer association is meat intake. Meat intake, especially intake of charred meat, is associated with exposure to heterocyclic amines because high meat intake results in increased exposure to 2-amino-1-methyl-6-phenylimidazo[4,5-b]-pyridine (PhIP) (53). PhIP is the most abundant heterocyclic amine in cooked meat (53). Women with the rapid-acetylation phenotype and high intake of well-done or charred meat may especially be at increased breast cancer risk, since O-acetylation by NAT2 predominates, resulting in reactive metabolites capable of forming DNA adducts and mutations. Zhu et al. (54) detected PhIP-DNA adducts in breast tissue and reported that rapid acetylators, compared with slow acetylators, had a higher level of PhIP-DNA adducts, conferring increased risk of breast cancer in rapid acetylators who have high exposure to PhIP.

Regarding NAT2, the molecular explanation for why particular exposures more readily result in N- or O-acetylation may be an important cause of the lack of significant main effect of the NAT2 genotype and breast cancer risk. The results of the meta-analysis of the genotype studies are therefore not unexpected. Aryl and heterocyclic amines differ in how they are metabolized, namely, the propensity toward N- or O-acetylation pathways for these respective exposures. Without specific knowledge of which pathway may predominate and whether other genes are involved, it is difficult to discern epidemiologically the exact estimate of risk conferred via one particular gene. Future studies should also measure and analyze the interaction of NAT2 and other genes, such as CYP1A2, in the carcinogen metabolism pathway since knowledge of underlying variation in both genes may more fully characterize an individual's carcinogen metabolism phenotype (55).

Future studies that investigate smoking and meat intake should attempt to collect and use standardized exposure measures. Doing so would greatly help summarize the results of published studies.

Abbreviations

    Abbreviations
  • NAT2

    N-acetyltransferase 2; PhIP, 2-amino-1-methyl-6-phenylimidazo[4,5-b]-pyridine

This work was supported in part by US Public Health Service National Institutes of Health training grant R25 CA094186 (H. O.-B.), research grants RR03655 and GM28356 (R. E.), and Cancer Center support grant P30CAD43703 (R. E.).

The authors thank Dr. Christine Ambrosone from the Roswell Park Cancer Institute for her constructive comments regarding the manuscript.

Conflict of interest: none declared.

References

1.
Smith
CA
Smith
G
Wolf
CR
Genetic polymorphisms in xenobiotic metabolism
Eur J Cancer
 , 
1994
, vol. 
30A
 (pg. 
1921
-
35
)
2.
Hein
DW
Acetylator genotype and arylamine-induced carcinogenesis
Biochim Biophys Acta
 , 
1988
, vol. 
948
 (pg. 
37
-
66
)
3.
Brockton
N
Little
J
Sharp
L
, et al.  . 
N-acetyltransferase polymorphisms and colorectal cancer: a HuGE review
Am J Epidemiol
 , 
2000
, vol. 
151
 (pg. 
846
-
61
)
4.
Marcus
PM
Hayes
RB
Vineis
P
, et al.  . 
Cigarette smoking, N-acetyltransferase 2 acetylation status, and bladder cancer risk: a case-series meta-analysis of a gene-environment interaction
Cancer Epidemiol Biomarkers Prev
 , 
2000
, vol. 
9
 (pg. 
461
-
7
)
5.
Marcus
PM
Vineis
P
Rothman
N
NAT2 slow acetylation and bladder cancer risk: a meta-analysis of 22 case-control studies conducted in the general population
Pharmacogenetics
 , 
2000
, vol. 
10
 (pg. 
115
-
22
)
6.
Hein
DW
Doll
MA
Fretland
AJ
, et al.  . 
Molecular genetics and epidemiology of the NAT1 and NAT2 acetylation polymorphisms
Cancer Epidemiol Biomarkers Prev
 , 
2000
, vol. 
9
 (pg. 
29
-
42
)
7.
Weber
W
The acetylator genes and drug response
1987
New York, NY
Oxford University Press
8.
DerSimonian
R
Laird
N
Meta-analysis in clinical trials
Control Clin Trials
 , 
1986
, vol. 
7
 (pg. 
177
-
88
)
9.
Number Cruncher Statistical Systems
 
Kaysville, UT. (www.ncss.com)
10.
Bulovskaya
LN
Krupkin
RG
Bochina
TA
, et al.  . 
Acetylator phenotype in patients with breast cancer
Oncology
 , 
1978
, vol. 
35
 (pg. 
185
-
8
)
11.
Cartwright
RA
Epidemiological studies on N-acetylation and C-center ring oxidation in neoplasia. Banbury report 16: Genetic variability in responses to chemical exposure
1984
New York, NY
Cold Spring Harbor Laboratory
12.
Ilett
KF
Detchon
P
Ingram
DM
, et al.  . 
Acetylation phenotype is not associated with breast cancer
Cancer Res
 , 
1990
, vol. 
50
 (pg. 
6649
-
51
)
13.
Ladero
JM
Fernandez
MJ
Palmeiro
R
, et al.  . 
Hepatic acetylator polymorphism in breast cancer patients
Oncology
 , 
1987
, vol. 
44
 (pg. 
341
-
4
)
14.
Philip
PA
Rogers
HJ
Millis
RR
, et al.  . 
Acetylator status and its relationship to breast cancer and other diseases of the breast
Eur J Cancer Clin Oncol
 , 
1987
, vol. 
23
 (pg. 
1701
-
6
)
15.
Sardas
S
Cok
I
Sardas
OS
, et al.  . 
Polymorphic N-acetylation capacity in breast cancer patients
Int J Cancer
 , 
1990
, vol. 
46
 (pg. 
1138
-
9
)
16.
Webster
DJ
Flook
D
Jenkins
J
, et al.  . 
Drug acetylation in breast cancer
Br J Cancer
 , 
1989
, vol. 
60
 (pg. 
236
-
7
)
17.
Deitz
AC
Zheng
W
Leff
MA
, et al.  . 
N-Acetyltransferase-2 (NAT2) acetylation polymorphism, well-done meat intake and breast cancer risk among post-menopausal women. (Abstract)
Proc Am Assoc Cancer Res
 , 
1999
, vol. 
40
 pg. 
148
 
18.
Ambrosone
CB
Freudenheim
JL
Marshall
JR
, et al.  . 
The association of polymorphic N-acetyltransferase (NAT2) with breast cancer risk
Ann N Y Acad Sci
 , 
1995
, vol. 
768
 (pg. 
250
-
2
)
19.
Chern
HD
Huang
CS
Wang
CY
, et al.  . 
Association of N-acetyltransferase polymorphism with the susceptibility of breast cancer in Taiwan. (Abstract)
Proc Am Assoc Cancer Res
 , 
1997
, vol. 
38
 pg. 
214
 
20.
Lilla
C
Risch
A
Kropp
S
, et al.  . 
SULT1A1 genotype, active and passive smoking, and breast cancer risk by age 50 years in a German case-control study
Breast Cancer Res
 , 
2005
, vol. 
7
 (pg. 
R229
-
37
)
21.
van der Hel
OL
Bueno-de-Mesquita
HB
van Gils
CH
, et al.  . 
Cumulative genetic defects in carcinogen metabolism may increase breast cancer risk (The Netherlands)
Cancer Causes Control
 , 
2005
, vol. 
16
 (pg. 
675
-
81
)
22.
Agundez
JA
Ladero
JM
Olivera
M
, et al.  . 
Genetic analysis of the arylamine N-acetyltransferase polymorphism in breast cancer patients
Oncology
 , 
1995
, vol. 
52
 (pg. 
7
-
11
)
23.
Alberg
AJ
Daudt
A
Huang
HY
, et al.  . 
N-acetyltransferase 2 (NAT2) genotypes, cigarette smoking, and the risk of breast cancer
Cancer Detect Prev
 , 
2004
, vol. 
28
 (pg. 
187
-
93
)
24.
Ambrosone
CB
Freudenheim
JL
Graham
S
, et al.  . 
Cigarette smoking, N-acetyltransferase 2 genetic polymorphisms, and breast cancer risk
JAMA
 , 
1996
, vol. 
276
 (pg. 
1494
-
501
)
25.
Chang-Claude
J
Kropp
S
Jager
B
, et al.  . 
Differential effect of NAT2 on the association between active and passive smoke exposure and breast cancer risk
Cancer Epidemiol Biomarkers Prev
 , 
2002
, vol. 
11
 (pg. 
698
-
704
)
26.
Deitz
AC
Zheng
W
Leff
MA
, et al.  . 
N-Acetyltransferase-2 genetic polymorphism, well-done meat intake, and breast cancer risk among postmenopausal women
Cancer Epidemiol Biomarkers Prev
 , 
2000
, vol. 
9
 (pg. 
905
-
10
)
27.
Delfino
RJ
Smith
C
West
JG
, et al.  . 
Breast cancer, passive and active cigarette smoking and N-acetyltransferase 2 genotype
Pharmacogenetics
 , 
2000
, vol. 
10
 (pg. 
461
-
9
)
28.
Egan
KM
Newcomb
PA
Titus-Ernstoff
L
, et al.  . 
Association of NAT2 and smoking in relation to breast cancer incidence in a population-based case-control study (United States)
Cancer Causes Control
 , 
2003
, vol. 
14
 (pg. 
43
-
51
)
29.
Huang
CS
Chern
HD
Shen
CY
, et al.  . 
Association between N-acetyltransferase 2 (NAT2) genetic polymorphism and development of breast cancer in post-menopausal Chinese women in Taiwan, an area of great increase in breast cancer incidence
Int J Cancer
 , 
1999
, vol. 
82
 (pg. 
175
-
9
)
30.
Hunter
DJ
Hankinson
SE
Hough
H
, et al.  . 
A prospective study of NAT2 acetylation genotype, cigarette smoking, and risk of breast cancer
Carcinogenesis
 , 
1997
, vol. 
18
 (pg. 
2127
-
32
)
31.
Kocabas
NA
Sardas
S
Cholerton
S
, et al.  . 
N-acetyltransferase (NAT2) polymorphism and breast cancer susceptibility: a lack of association in a case-control study of Turkish population
Int J Toxicol
 , 
2004
, vol. 
23
 (pg. 
25
-
31
)
32.
Krajinovic
M
Ghadirian
P
Richer
C
, et al.  . 
Genetic susceptibility to breast cancer in French-Canadians: role of carcinogen-metabolizing enzymes and gene-environment interactions
Int J Cancer
 , 
2001
, vol. 
92
 (pg. 
220
-
5
)
33.
Lee
KM
Park
SK
Kim
SU
, et al.  . 
N-acetyltransferase (NAT1, NAT2) and glutathione S-transferase (GSTM1, GSTT1) polymorphisms in breast cancer
Cancer Lett
 , 
2003
, vol. 
196
 (pg. 
179
-
86
)
34.
Matheson
MC
Stevenson
T
Akbarzadeh
S
, et al.  . 
GSTT1 null genotype increases risk of premenopausal breast cancer
Cancer Lett
 , 
2002
, vol. 
181
 (pg. 
73
-
9
)
35.
Millikan
RC
Pittman
GS
Newman
B
, et al.  . 
Cigarette smoking, N-acetyltransferases 1 and 2, and breast cancer risk
Cancer Epidemiol Biomarkers Prev
 , 
1998
, vol. 
7
 (pg. 
371
-
8
)
36.
Morabia
A
Bernstein
MS
Bouchardy
I
, et al.  . 
Breast cancer and active and passive smoking: the role of the N-acetyltransferase 2 genotype
Am J Epidemiol
 , 
2000
, vol. 
152
 (pg. 
226
-
32
)
37.
Sillanpaa
P
Hirvonen
A
Kataja
V
, et al.  . 
NAT2 slow acetylator genotype as an important modifier of breast cancer risk
Int J Cancer
 , 
2005
, vol. 
114
 (pg. 
579
-
84
)
38.
van der Hel
OL
Peeters
PH
Hein
DW
, et al.  . 
NAT2 slow acetylation and GSTM1 null genotypes may increase postmenopausal breast cancer risk in long-term smoking women
Pharmacogenetics
 , 
2003
, vol. 
13
 (pg. 
399
-
407
)
39.
Wu
FY
Lee
YJ
Chen
DR
, et al.  . 
Association of DNA-protein crosslinks and breast cancer
Mutat Res
 , 
2002
, vol. 
501
 (pg. 
69
-
78
)
40.
Lissowska
J
Brinton
LA
Zatonski
W
, et al.  . 
Tobacco smoking, NAT2 acetylation genotype and breast cancer risk
Int J Cancer
 , 
2006
, vol. 
119
 (pg. 
1961
-
9
)
41.
Bell
DA
Taylor
JA
Butler
MA
, et al.  . 
Genotype/phenotype discordance for human arylamine N-acetyltransferase (NAT2) reveals a new slow-acetylator allele common in African-Americans
Carcinogenesis
 , 
1993
, vol. 
14
 (pg. 
1689
-
92
)
42.
Lin
HJ
Han
CY
Lin
BK
, et al.  . 
Slow acetylator mutations in the human polymorphic N-acetyltransferase gene in 786 Asians, blacks, Hispanics, and whites: application to metabolic epidemiology
Am J Hum Genet
 , 
1993
, vol. 
52
 (pg. 
827
-
34
)
43.
Boyd
NF
Stone
J
Vogt
KN
, et al.  . 
Dietary fat and breast cancer risk revisited: a meta-analysis of the published literature
Br J Cancer
 , 
2003
, vol. 
89
 (pg. 
1672
-
85
)
44.
Ambrosone
CB
Freudenheim
JL
Sinha
R
, et al.  . 
Breast cancer risk, meat consumption and N-acetyltransferase (NAT2) genetic polymorphisms
Int J Cancer
 , 
1998
, vol. 
75
 (pg. 
825
-
30
)
45.
Gertig
DM
Hankinson
SE
Hough
H
, et al.  . 
N-acetyl transferase 2 genotypes, meat intake and breast cancer risk
Int J Cancer
 , 
1999
, vol. 
80
 (pg. 
13
-
17
)
46.
Delfino
RJ
Sinha
R
Smith
C
, et al.  . 
Breast cancer, heterocyclic aromatic amines from meat and N-acetyltransferase 2 genotype
Carcinogenesis
 , 
2000
, vol. 
21
 (pg. 
607
-
15
)
47.
Evans
D
Kalow
W
Goedde
H
Agarwal
D
Acetylation
Ethnic differences in reactions to drugs and xenobiotics
 , 
1986
New York, NY
Alan R Liss, Inc
48.
Pfau
W
Stone
EM
Brockstedt
U
, et al.  . 
DNA adducts in human breast tissue: association with N-acetyltransferase-2 (NAT2) and NAT1 genotypes
Cancer Epidemiol Biomarkers Prev
 , 
1998
, vol. 
7
 (pg. 
1019
-
25
)
49.
Terry
PD
Goodman
M
Is the association between cigarette smoking and breast cancer modified by genotype? A review of epidemiologic studies and meta-analysis
Cancer Epidemiol Biomarkers Prev
 , 
2006
, vol. 
15
 (pg. 
602
-
11
)
50.
King
RS
Teitel
CH
Kadlubar
FF
In vitro bioactivation of N-hydroxy-2-amino-alpha-carboline
Carcinogenesis
 , 
2000
, vol. 
21
 (pg. 
1347
-
54
)
51.
Ambrosone
CB
Impact of genetics on the relationship between smoking and breast cancer risk
J Women's Cancer
 , 
2001
, vol. 
3
 (pg. 
17
-
22
)
52.
Firozi
PF
Bondy
ML
Sahin
AA
, et al.  . 
Aromatic DNA adducts and polymorphisms of CYP1A1, NAT2, and GSTM1 in breast cancer
Carcinogenesis
 , 
2002
, vol. 
23
 (pg. 
301
-
6
)
53.
Sinha
R
Gustafson
DR
Kulldorff
M
, et al.  . 
2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine, a carcinogen in high-temperature-cooked meat, and breast cancer risk
J Natl Cancer Inst
 , 
2000
, vol. 
92
 (pg. 
1352
-
4
)
54.
Zhu
J
Chang
P
Bondy
ML
, et al.  . 
Detection of 2-amino-1-methyl-6-phenylimidazo[4,5-b]-pyridine-DNA adducts in normal breast tissues and risk of breast cancer
Cancer Epidemiol Biomarkers Prev
 , 
2003
, vol. 
12
 (pg. 
830
-
7
)
55.
Murtaugh
MA
Sweeney
C
Ma
KN
, et al.  . 
The CYP1A1 genotype may alter the association of meat consumption patterns and preparation with the risk of colorectal cancer in men and women
J Nutr
 , 
2005
, vol. 
135
 (pg. 
179
-
86
)

Author notes

Editor's note:This paper is also available on the website of the Human Genome Epidemiology Network (http://www.cdc.gov/genomics/hugenet/).