Review 117 Medicina (Kaunas) 2012;48(3):117-31 a Meta-Analysis of the Relationship Between NAT2 Polymorphism and Colorectal Cancer Susceptibility

Home , Zhong (surname)

ReviEw 117 Medicina (Kaunas) 2012;48(3):117-31 A Meta-Analysis of the Relationship Between NAT2 Polymorphism and Colorectal Cancer Susceptibility

Hong Liu, Zhong-xue Fu, Chun-yi Wang, Jiang Qian, Lei Xing, Yi-wu Liu Department of General Surgery, the First Affiliated Hospital of Chongqing Medical University, Yuanjiagang, Yuzhong district, China

Key words: N-acetyltransferase 2; colorectal cancer; polymorphism; genetic susceptibility; meta-analysis.

Summary. Background and Objective. Although the association between N-acetyltransferase 2 (NAT2) polymorphism and colorectal cancer (CRC) susceptibility in humans has been extensively investigated, the results are contradictory. The aim of this study was to conduct a meta-analysis of published studies to quantitatively summarize the association between NAT2 polymorphism and risk of CRC. Material and Methods. Relevant studies that had investigated NAT2 polymorphism and CRC susceptibility were identified through a comprehensive search of Pubmed, EMBASE, Medline, Bio- sis, Wiley-Blackwell, ISI Web of Knowledge, CNKI, and Chinese Biomedicine Database until Oc- tober 2011. After selection based on the inclusion and exclusion criteria, the relevant data were extracted from each study, and finally a meta-analysis was performed. Results. Eight phenotype studies (791 cases and 1158 controls) and 45 genotype studies (13 875 cases and 18 879 controls) were included in the present meta-analysis. The pooling of phenotype studies showed no significant association between the NAT2 acetylator status and CRC susceptibility (rapid acetylator, OR, 1.32; 95% CI, 0.92–1.89, P=0.14; slow acetylator, OR, 0.76; 95% CI, 0.53–1.09, P=0.14). The combined ORs for rapid and slow acetylator status and CRC risk in genotype studies were 1.01 (95% CI, 0.94–1.08; P=0.86) and 0.99 (95% CI, 0.93–1.06; P=0.86), respectively. In the subgroup analysis by regions, no increased risks were found in Asians, Europe- ans, Americans, or Australasians. Pooling studies were also conducted on the groups of gender, specific tumor sites, and smoking status, but no significant association in genotype distribution between CRC and control was found as well. Conclusions. These results of our meta-analysis suggest that there is no overall association between NAT2 polymorphism and CRC susceptibility. Introduction gene located on chromosome 8p22 region that con- Colorectal cancer (CRC) is one of the most com- tains an 870-bp open reading frame and encodes a mon cancers in the world. It is generally accepted protein of 290 amino acids (3). Individuals can be that human colorectal carcinogenesis is a complex, divided into 3 different phenotypes based on the al- multistep, and multifactorial process in which many lelic variants of NAT2: fast, intermediate, and slow. factors, such as dietary and lifestyle habits and/or These phenotypes are determined by single nucle- mild genetic predisposition, are implicated (1). Ex- otide polymorphisms in NAT2 (4). As one of the posure to carcinogens, such as heterocyclic amines phase II enzymes, NAT2 plays an essential role in (HCAs), polycyclic aromatic hydrocarbons (PAHs), the detoxification and/or bioactivation of several and other amine compounds, is regarded as a risk carcinogenic compounds, such as HCAs and PAHs, factor for developing CRC (2). Studies have shown found in meat and tobacco smoke (5). It is therefore that individual inherited susceptibility plays an im- conceivable that increased or decreased activities portant role in the pathogenesis of tumor. In the last of this enzyme may be involved in susceptibility to 3 decades, genetic polymorphisms have also been CRC (6). extensively investigated to identify inherited genet- Since Lang et al. (7) first reported an association ic susceptibility for CRC. between the NAT2 acetylator status and CRC risk, a N-acetyltransferase 2 (NAT2) is one of these sus- number of studies have been published to describe ceptibility genes, which has been considered to have the association between NAT2 polymorphisms and an association with CRC risk. It is a polymorphic CRC risk in humans (8–57). However, the results Correspondence to Z. Fu, Department of General Surgery, the are not conclusive. Therefore, in the present study, First Affiliated Hospital of Chongqing Medical University, 1 a meta-analysis of all available published studies was Youyi Road, Yuanjiagang, Yuzhong district, 400016 Chong- carried out to summarize the results of the effect of qing, China. E-mail: [email protected] NAT2 polymorphism on CRC. Medicina (Kaunas) 2012;48(3) 118 Hong Liu, Zhong-xue Fu, Chun-yi Wang, et al.

Material and Methods numbers of cases and controls of different acetyla- Search strategy tion status, phenotyping and genotyping technique, Papers published until October 2011 that had location of tumors, matching, exposure assessment, investigated NAT2 polymorphism and CRC sus- and results of studies. This was checked by a sec- ceptibility were identified through a comprehen- ond independent investigator to avoid data input er- sive search of Pubmed, EMBASE, Medline, Biosis, rors. Any disagreement was resolved by discussion; Wiley-Blackwell, ISI Web of Knowledge, CNKI, a third investigator adjudicated the disagreements if and Chinese Biomedicine Database, using the fol- they could not come to an agreement. lowing search key words: acetyltransferase or N- acetyltransferase 2 or NAT2, genetic polymorphism Statistical Analysis or single nucleotide polymorphism, colon or rec- A meta-analysis was performed separately for tum or colorectal, cancer or carcinoma or tumor. phenotype and genotype studies. The strength of The search was without language restriction and se- the associations between the NAT2 polymorphism lected only those conducted on human subjects. In and CRC susceptibility was estimated by odds ratio addition, the citations in relevant articles were also (OR) with 95% confidence intervals (CI). Hetero- thoroughly examined to further ensure that all ap- geneity was analyzed among the studies using the propriate studies were collected. In situations when Cochran’s Q test and I2 test. Fixed effects model multiple studies were published using the same data was used when I2 was less than 30%. Otherwise, source, only the one that contained the largest data the random effects model was used (I2 >30%).We was taken into account. Unpublished studies were constructed a funnel plot to test the influence of not considered in this literature search. publication bias. Sensitivity analysis was performed by deselecting studies with extreme findings to test Inclusion and Exclusion Criteria the robustness of the results. Statistical analysis was All articles involving studies that investigated performed using the Review Manager 5.1. P<0.05 NAT2 and CRC susceptibility were included. The se- was considered statistically significant. lection criteria were as follows: 1) case-control studies; 2) evaluation of the association between NAT2 Results polymorphism and CRC susceptibility; 3) enough Study Characteristics information about the number of CRC cases and Fig. 1 shows the literature selection process. controls studied with the different NAT2 acetylation Overall, 51 studies of NAT2 acetylator status and status; and 4) clearly description of CRC diagnoses CRC risk were eligible for our final analyses. They and the sources of cases and controls. The major ex- were published between 1986 and 2011. The size of clusion criteria were as follows: 1) no control group; study population ranged from 72 to 3587 individuals 2) overlapping or republished studies; 3) no usable (14 666 cases and 20 037 controls). Of the included information reported; and 4) cases or controls suf- studies, 2 studies evaluated NAT2 acetylator status fered from other caners or other colorectal diseases. by phenotyping and genotyping separately, 6 studies by phenotyping only, and 43 studies by genotyp- Classification of NAT2 Acetylation Status ing only. Characteristics of the included studies in Eligible studies were classified into two types – this meta-analysis are presented in Tables 1 and 2. “phenotype” or “genotype” studies – because the methods for measuring NAT2 acetylation status is Main Results different-measuring phenotypes by using metabolic Phenotype Studies. Of the 8 phenotype studies, response to a particular compound or measuring al- 4 studies identified acetylation phenotype via ad- leles directly, and we did not combine these stud- ministration of sulfamethazine, 1 via administration ies for analysis. In our study, phenotype frequen- of p-aminobenzoic acid (PABA), 4-aminobiphenyl cies were summarized as slow and rapid status. In (ABP), 2-aminofluorene (AF), andβ -naphthylamine genotype studies, rapid acetylators were defined as (BNA), and other 3 used the caffeine test. For the carriers of homozygous or heterozygous for rapid meta-analysis, the test for heterogeneity was statis- acetylator alleles; those individuals who had two tically significant (I2=60%, P=0.01). Figs. 2 and 3 slow acetylation alleles were classified as slow acety- show the summary odds ratio (rapid acetylator, OR, lators, consistent with the definition in most studies. 1.32; 95% CI, 0.92–1.89; P=0.14; slow acetylator, OR, 0.76; 95% CI, 0.53–1.09; P=0.14) for the Data Extraction and Analysis combined rapid or slow acetylator phenotype stud- One of the authors extracted and summarized ies separately, using the random-effect model. It is the following information from each article: first au- noteworthy that a plot of the rapid or slow acetylator thor, publication year, country of origin, study type, status and CRC risk showed a trend toward a less study population, number of cases and controls, and significant association in the studies.

Medicina (Kaunas) 2012;48(3) NAT2 and Colorectal Cancer 119

Records identified through Additional records identified database searching through other sources (n=155) (n=15)

Records after duplicates removed (n=170)

Records screened Records excluded (n=157) (n=85)

Full-text articles assessed for eligibility (n=72) Full-text articles excluded (n=21): Review articles (n=4) Inadequate data reports (n=3) Studies included Uncontrolled case reports (n=4) in qualitative synthesis Reports containing individuals (meta-analysis) with polyps or adenoma (n=51) (n=10)

Fig. 1. Flow diagram

Table 1. Characteristics of Studies Included in the Meta-Analysis for Phenotype

Slow Method Author and Year Country Study type No. of No. of Rapid Rapid Slow Con- for phenotype Matching cases controls cases controls cases trols determination Lang et al. (7), 1986 USA Case-control 43 41 23 13 20 28 Sulfamethazine Age Ilett et al. (8), 1987 Australia Case-control 49 86 28 26 21 60 Sulfamethazine Age, sex, and racial origin Wohlleb et al. (9), USA Case-control 43 41 23 13 20 28 Sulfamethazine – 1990 Kirlin et al. (10), USA Case-control 25 12 12 9 13 3 PABA, ABP, – 1991 AF, and BNA Ladero et al. (11), Spain Case-control 109 96 49 40 60 56 Sulfamethazine Age 1991 Lang et al. (12), USA Case-control 34 205 14 92 20 113 Caffeine test – 1994 Le Marchand et al. USA Case-control 348 466 272 346 76 120 Caffeine test Sex, (13), 2001 ethnicity, age Ishibe et al. (14), USA Case-control 140 211 65 106 75 105 Caffeine test Gender 2002 and age in 5-year intervals PABA, p-aminobenzoic acid; ABP, 4-aminobiphenyl; AF, 2-aminofluorene; BNA, β-naphthylamine.

Genotype Studies For the meta-analysis, ORs were calculated from Of the 45 genotyping studies, 14 studies were the reported frequencies of genotype by NAT2 sta- carried out in Asian countries, 16 in European coun- tus. Overall, the combined results based on all stud- tries, 14 in American countries, and 1 in Australia. ies indicated that no significantly elevated CRC risk In some studies, the authors classified individu- was associated with fast or slow NAT2 genotypes als into 3 categories (slow, intermediate, and rapid between cases and controls (rapid acetylator, OR, acetylators), and in our meta-analysis, intermediate 1.01; 95% CI, 0.94–1.08, P=0.86; slow acetylator, acetylators were reclassified as fast acetylators. OR, 0.99; 95% CI, 0.93–1.06; P=0.86), and there

Medicina (Kaunas) 2012;48(3) 120 Hong Liu, Zhong-xue Fu, Chun-yi Wang, et al. – – – – – – – – – – – 12 Meat Meat Meat Meat history Exposure a ssessment history, and occupational occupational consumption consumption consumption Diet, smoking, Red meat intake exercise, medical exercise, consumption and cigarette smoking – – – – – – – – – 11 Age Age Gender smoking Matching age group Age and sex Sex, ethnicity, Sex and 5-year 5-year intervals and age (2 years) Age (±1 year) and Gender and age in – – – – – – – – – – – – 10 umors colon, rectum t Rectum and rectum side of colon Proximal and Proximal Non-sigmoid Location of the Sigmoid colon, distal colon and of colon, sigmoid Right and left sideRight Colon and rectum Right side and left 9 3 7 7 15 31 62 62 53 203 100 125 120 139 143 157 110 Slow Slow 1156 (M/F) (61/39) ontrols c ontrols (603/553) 8 5 5 24 33 26 71 48 60 60 79 106 175 101 131 930 149 124 Slow Slow c ases (M/F) (61/40) (536/394) 7 3 13 50 34 74 96 81 93 57 98 298 140 134 807 119 115 497 Rapid (M/F) (41/33) ontrols c ontrols (433/374) 6 20 33 96 32 73 81 66 60 98 60 64 208 100 156 694 101 419 c ases Rapid (M/F) (41/32) (376/318) 5 28 36 96 329 112 343 174 221 201 187 100 258 122 200 654 208 1963 No. of No. (M/F) ontrols c ontrols (102/72) (1036/927) Characteristics of Studies Included in the Meta-Analysis for Genotype 4 44 36 234 202 103 275 174 212 114 216 106 120 103 209 543 143 1624 c ases No. of No. (M/F) (102/72) (912/712) Table 2. Table ype 3 Nested Nested Study t Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control 2 UK UK UK UK USA USA USA USA USA Japan Japan Japan Japan Spain Country Portugal Australia Singapore 1 Year Author and elfare et al. (21), 1997 Rodriguez et al. (15), 1993 Oda et al. (16), 1994 Shibuta et al. (17), 1994 Bell et al. (18), 1995 Spurr et al. (19), 1995 Hubbard et al. (20), 1997 Hubbard W Chen et al. (22), 1998 Gil et al. (23), 1998 Lee et al. (24), 1998 Kampman et al. (25), 1999 Yoshioka et al. (26), 1999 Yoshioka Agúndez et al. (27), 2000 Katoh et al. (28), 2000 Butler et al. (29), 2001 Le Marchand et al. (13), Le Marchand 2001 Ishibe et al. (14), 2002

Medicina (Kaunas) 2012;48(3) NAT2 and Colorectal Cancer 121 ------– – – – – – – – – – 12 smoking smoking smoking smoking smoking smoking smoking Cigarette Meat consump Meat consump Meat consump Meat consump Meat consump Meat consump tion and cigarette tion and cigarette tion and cigarette tion and cigarette tion and cigarette tion and cigarette Cigarette smoking - – – – – – – – 11 meat matched intervals) Age- and Caucasian age groups Population Gender, age sex-matched consumption Year of birth, Age- and sex- month/year of ty of residence habits, and red Sex, 5-year age Gender, nation, Gender and age blood collection blood (5-year intervals) place, age (5-year Healthy unrelated Healthy occupation, living living occupation, Sex and by 5-year Sex and by groups, and coun Age, sex, smoking – – – – – – – – – – – – – – – – 10 distal Proximal and Proximal Colon and rectum Continuation) 9 68 50 59 79 69 20 300 349 520 495 318 267 162 152 374 376 126 154 (34/45) (79/75) (306/214) 8 59 18 20 52 47 72 18 68 304 409 146 267 107 192 295 413 143 157 (28/19) (90/67) (247/162) 7 91 48 63 96 237 243 470 362 169 182 176 293 146 227 317 220 135 576 (255/215) (112/108) (327/249) 6 43 65 76 40 83 96 65 54 186 357 112 233 119 168 208 272 197 558 (100/97) (204/153) (313/245) 5 83 537 592 990 857 237 500 443 343 308 243 601 693 107 299 204 222 730 (561/429) (146/153) (406/324) 4 83 92 83 102 490 766 258 500 183 139 360 503 685 226 244 168 122 715 (451/315) (128/116) (403/312) Characteristics of Studies Included in the Meta-Analysis for Genotype ( Table 2. Table 3 Nested Nested Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control 2 UK USA USA USA USA Spain China China China China China China (Taiwan) Hungary (Taiwan) Germany Germany Germany Netherlands Netherlands 1 Tiemersma et al. (30), 2002 Barrett et al. (31), 2003 Slattery et al. (32), 2003 Van der Hel et al. (33), Van 2003* He et al. (34), 2004 Kiss et al. (35), 2004 Chan et al. (36), 2005* Chen et al. (37), 2005 Landi et al. (38), 2005 Borlak and Reamon- Buettner (39), 2006 Lilla et al. (40), 2006 Moslehi et al. (41), 2006 Pistorius et al. (42), 2006 Huang et al. (43), 2007 Jiang (44) 2007 Luo et al. (45), 2007 Mahid et al. (46), 2007 Yeh et al. (47), 2007

Medicina (Kaunas) 2012;48(3) 122 Hong Liu, Zhong-xue Fu, Chun-yi Wang, et al. - - - – – 12 Meat Meat Meat smoking smoking smoking smoking Cigarette consumption consumption consumption and doneness Meat consump Meat consump Meat consump Red meat intake tion and cigarette tion and cigarette tion and cigarette - – – 11 sex age residence Unaffected and area of Gender, age race, and age Race, age, and Sex, ethnicity/ group-matched Gender, nation, Gender and age Gender and age living place, age living (within 3 years), ins in the family (3-year intervals) Sex-matched and siblings and cous siblings – – – – – – – – – – 10 9 9 Continuation) 24 29 425 445 736 497 320 50 (23/27) 110 (52/58) 8 2 12 56 227 211 458 336 281 50 (25/25) 88 (46/42) 7 35 112 405 323 511 201 996 257 71 (33/38) 102 (33/69) 6 64 93 20 273 166 374 656 230 42 (26/16) 59 (25/34) 5 121 830 768 225 286 355 1247 1493 121 (56/65) 212 (85/127) 4 66 92 500 377 832 105 992 286 147 301 (51/41) (90/76) Characteristics of Studies Included in the Meta-Analysis for Genotype ( 3 Table 2. Table Nested Nested Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control Case-control 2 Italy USA USA USA Japan Japan Brazil China Canada Denmark 1 ang et al. (57), 2011 *The individuals are all females. *The individuals Yoshida et al. (48), 2007 Yoshida Butler et al. (49), 2008 Cotterchio et al. (50), 2008 Cotterchio Sørensen et al. (51), 2008 Kobayashi et al. (52), 2009 Kobayashi Nöthlings et al. (53), 2009 Peng et al. (55), 2010 Zupa et al. (54), 2009 Silva et al. (56), 2011 Silva W

Medicina (Kaunas) 2012;48(3) NAT2 and Colorectal Cancer 123

Case Control Weight, Odds Ratio Odds Ratio Study No. of No. of No. of No. of % M-H, Random, Events Total Events Total 95% CI M-H, Random, 95% CI Ilett et al. (8), 1987 28 49 26 86 12.0 3.08 (1.48, 6.38) Ishibe et al. (14), 2002 65 140 106 211 17.6 0.86 (0.56, 1.32) Kirlin et al. (10), 1991 12 25 9 12 4.5 0.31 (0.07, 1.41) Ladero et al. (11), 1991 49 109 40 96 15.1 1.14 (0.66, 1.99) Lang et al. (7), 1986 23 43 13 41 9.7 2.48 (1.02, 6.03) Lang et al. (12), 1994 14 34 92 205 11.9 0.86 (0.41, 1.80) Le Marchand et al. (13), 2001 272 348 346 466 19.6 1.24 (0.89, 1.72) Wohlleb et al. (9), 1990 23 43 13 41 9.7 2.48 (1.02, 6.03) Total (95% CI) 791 1158 100 1.32 (0.92, 1.89) Total events 486 645 Heterogeneity: τ2=0.15; χ2=17.69, df=7 (P=0.01); I2=60% 0.1 0.2 0.5 1 2 5 10 Test for overall effect: Z=1.49 (P=0.14) Favors rapid Favors rapid cases controls

Fig. 2. Meta-analysis of NAT2 rapid acetylator status and CRC susceptibility based on NAT2 phenotypes

Case Control Weight, Odds Ratio Odds Ratio Study No. of No. of No. of No. of % M-H, Random, Events Total Events Total 95% CI M-H, Random, 95% CI Ilett et al. (8), 1987 21 49 60 86 12.0 0.33 (0.16, 0.67) Ishibe et al. (14), 2002 75 140 105 211 17.6 1.16 (0.76, 1.79) Kirlin et al. (10), 1991 13 25 3 12 4.5 3.25 (0.71, 14.92) Ladero et al. (11), 1991 60 109 56 96 15.1 0.87 (0.50, 1.52) Lang et al. (7), 1986 20 43 28 41 9.7 0.40 (0.17, 0.98) Lang et al. (12), 1994 20 34 113 205 11.9 1.16 (0.56, 2.43) Le Marchand et al. (13), 2001 76 348 120 466 19.6 0.81 (0.58, 1.12) Wohlleb et al. (9), 1990 20 43 28 41 9.7 0.40 (0.17, 0.98) Total (95% CI) 791 1158 100 0.76 (0.53,1.09) Total events 305 513 Heterogeneity: τ2=0.15; χ2=17.69, df=7 (P=0.01); I2=60% 0.1 0.2 1 5 Test for overall effect: Z=1.49 (P=0.14) Favors slow Favors slow cases controls

Fig. 3. Meta-analysis of NAT2 slow acetylator status and CRC susceptibility based on NAT2 phenotypes was a statistically significant heterogeneity (I2=39%, descending colon, while the distal colon refers to P=0.005) (Figs. 4 and 5). Therefore, the random- the sigmoid colon and rectum (58). To evaluate the effects model was used to pool the results. effect of NAT2 acetylator status associated with the Subgroup analysis was conducted by regions. No location of the tumors, tests for heterogeneity across association was observed in the subgroups of Asians, the studies showed no evidence of heterogeneity re- Europeans, Americans, or Australasians (Figs. 4 lated to distal and proximal colon. The overall odds and 5). A statistically significant heterogeneity was ratios for the NAT2 status and distal and proximal observed among studies for all analysis. Therefore, colon are presented in Table 4. the random effects model was used to pool the results. It was not possible to perform some subgroup Interactions analyses (ethnicity, stage, histological type, etc.) due Smoking. In terms of the main effect of NAT2 to a lack of adequate data on these subtypes. polymorphisms on CRC susceptibility associated Nine studies provided adequate data for gen- with smoking, 6 of the 45 genotype studies analyzed der (7 studies for both men and women, and 2 for the effect in detail by using a variety of exposure women only). Table 3 summarizes the results of variables. Van der Hel et al. (33) and Silva et al. the meta-analysis of the association between NAT2 (56) showed significant evidence for the modifica- acetylator status and CRC risk by gender. No sig- tion of NAT2 and CRC by smoking. Van der Hel nificantly increased risk was found in different genet al. reported that rapid NAT2 acetylation in com- der groups. bination with smoking significantly increased the Six studies described the location of the tumors; risk of CRC; meanwhile, Silva et al. indicated that in our meta-analysis, the location was reclassified cigarette smoking increased the risk of CRC among into 2 groups: proximal and distal. The proximal slow NAT2 acetylators. Lilla et al. (40) reported colon is defined as the ascending, transverse, and that exposure to environmental tobacco smoke was

Medicina (Kaunas) 2012;48(3) 124 Hong Liu, Zhong-xue Fu, Chun-yi Wang, et al.

Case Control Odds Ratio Odds Ratio Weight, Study or Subgroup No. of No. of No. of No. of % M-H, Random, M-H, Random, Events Total Events Total 95% CI 95% CI Asian Chen et al. (37), 2005 119 139 293 343 1.2 1.02 (0.58, 1.78) He et al. (34), 2005 65 83 169 237 1.1 1.45 (0.80, 2.63) Huang et al. (43), 2007 197 244 220 299 2.0 1.51 (1.00, 2.27) Jiang et al. (44), 2007 96 168 135 204 1.9 0.68 (0.45, 1.04) Katoh et al. (28), 2000 98 103 115 122 0.3 1.19 (0.37, 3.88) Kobayashi et al. (52), 2009 93 105 201 225 0.8 0.93 (0.44, 1.93) Lee et al. (24), 1998 156 216 134 187 1.8 1.03 (0.67, 1.59) Luo et al. (45), 2007 65 83 63 83 0.8 1.15 (0.56, 2.37) Oda et al. (16), 1994 33 36 33 36 0.2 1.00 (0.19, 5.32) Peng et al. (55), 2010 230 286 257 286 1.6 0.46 (0.29, 0.75) Shibuta et al. (17), 1994 208 234 298 329 1.3 0.83 (0.48, 1.44) Yeh et al. (47), 2007 558 715 576 730 3.6 0.95 (0.74, 1.22) Yoshida et al. (48), 2007 64 66 112 121 0.2 2.57 (0.54, 12.27) Yoshioka et al. (26), 1999 101 106 93 100 0.3 1.52 (0.47, 4.96) Subtotal (95% CI) 2584 3302 17.10 0.97 (0.80, 1.18) Total events 2083 2699 Heterogeneity: τ2=0.04; χ2=20.69, df=13 (P=0.08); I2=37% Test for overall effect: Z=0.29 (P=0.78)

European Agúndez et al. (47), 2000 60 120 119 258 1.8 1.17 (0.76, 1.80) Barrett et al. (31), 2003 186 490 243 592 3.7 0.88 (0.69, 1.12) Bell et al. (18), 1995 96 202 50 112 1.7 1.12 (0.71, 1.79) Borlak et al. (39), 2006 40 92 91 243 1.5 1.28 (0.79, 2.09) Gil et al. (23), 1998 66 114 81 201 1.6 2.04 (1.28, 3.25) Hubbard et al. (20), 1997 100 275 140 343 2.7 0.83 (0.60, 1.15) Kiss et al. (35), 2004 233 500 182 500 3.6 1.52 (1.18, 1.96) Landi et al. (38), 2005 168 360 146 308 2.9 0.97 (0.72, 1.32) Lilla et al. (40), 2006 208 503 227 601 3.7 1.16 (0.91, 1.48) Pistorius et al. (42), 2006 83 226 48 107 1.6 0.71 (0.45, 1.14) Sørensen et al. (51), 2008 166 377 323 768 3.6 1.08 (0.85, 1.39) Spurr et al. (19), 1995 32 103 34 96 1.1 0.82 (0.46, 1.48) Tiemersma et al. (30), 2002 43 102 237 537 1.9 0.92 (0.60, 1.42) Van der Hel et al. (33), 2003 112 258 362 857 3.2 1.05 (0.79, 1.39) Welfare et al. (21), 1997 73 174 75 174 1.9 0.95 (0.62, 1.46) Zupa et al. (54), 2009 42 92 71 121 1.3 0.59 (0.34, 1.02) Subtotal (95% CI) 3988 5818 38 1.04 (0.92, 1.19) Total events 1708 2429 Heterogeneity: τ2=0.03; χ2=30.12, df=15 (P=0.01); I2=50% Test for overall effect: Z=0.67 (P=0.50)

American Butler et al. (49), 2008 273 500 405 830 4.0 1.26 (1.01, 1.58) Chan et al. (36), 2005 76 183 176 443 2.5 1.08 (0.76, 1.53) Chen et al. (22), 1998 81 212 96 221 2.2 0.81 (0.55, 1.18) Cotterchio et al. (50), 2008 374 832 511 1247 4.7 1.18 (0.99, 1.40) Ishibe et al. (14), 2002 64 143 98 208 1.9 0.91 (0.59, 1.39) Kampman et al. (25), 1999 694 1624 807 1963 5.5 1.07 (0.94, 1.22) Le Marchand et al. (13), 2001 419 543 497 654 3.4 1.07 (0.82, 1.40) Mahid et al. (46), 2007 54 122 96 222 1.8 1.04 (0.67, 1.63) Moslehi et al. (41), 2006 272 685 317 693 4.1 0.78 (0.63, 0.97) Nöthlings et al. (53), 2009 656 992 996 1493 4.9 0.97 (0.82, 1.15) Rodriguez et al. (15), 1993 20 44 13 28 0.5 0.96 (0.37, 2.49) Silva et al. (56), 2011 59 147 102 212 1.9 0.72 (0.47, 1.11) Slattery et al. (32), 2003 357 766 470 990 4.5 0.97 (0.80, 1.17) Wang et al. (57), 2011 20 301 35 355 1.2 0.65 (0.37, 1.15) Subtotal (95% CI) 7094 9559 43.10 1.00 (0.91, 1.09) Total events 3419 4619 Heterogeneity: τ2=0.01; χ2=19.84, df=13 (P=0.10); I2=34% Test for overall effect: Z=0.09 (P=0.93) Australasian Butler et al. (29), 2001 60 209 57 200 1.9 1.01 (0.66, 1.55) Subtotal (95% CI) 209 200 1.9 1.01 (0.66, 1.55) Total events 60 57 Heterogeneity: Not applicable Test for overall effect: Z=0.05 (P=0.96)

Total (95% CI) 13875 18879 100 1.01 (0.94, 1.08) Total events 7270 9804 Heterogeneity: τ2=0.02; χ2=72.25, df=44 (P=0.005); I2=39% Test for overall effect: Z=0.18(P=0.86) 0.1 0.2 0.5 1 2 5 10 Test for subgroup differences: χ2=0.50, df=3 (P=0.92); I2=0% Favors slow Favors slow cases controls Fig. 4. Meta-analysis of NAT2 rapid acetylator status and CRC susceptibility based on NAT2 genotypes Studies are stratified by regions.

Medicina (Kaunas) 2012;48(3) NAT2 and Colorectal Cancer 125

Case Control Odds Ratio Odds Ratio Weight, Study or Subgroup No. of No. of No. of No. of % M-H, Random, M-H, Random, Events Total Events Total 95% CI 95% CI Asian Chen et al. (37), 2005 20 139 50 343 1.2 0.98 (0.56, 1.73) He et al. (34), 2005 18 83 68 237 1.1 0.69 (0.38, 1.25) Huang et al. (43), 2007 47 244 79 299 2.0 0.66 (0.44, 1.00) Jiang et al. (44), 2007 72 168 69 204 1.9 1.47 (0.96, 2.24) Katoh et al. (28), 2000 5 103 7 122 0.3 0.84 (0.26, 2.72) Kobayashi et al. (52), 2009 12 105 24 225 0.8 1.08 (0.52, 2.25) Lee et al. (24), 1998 60 216 53 187 1.8 0.97 (0.63, 1.50) Luo et al. (45), 2007 18 83 20 83 0.8 0.87 (0.42, 1.80) Oda et al. (16), 1994 3 36 3 36 0.2 1.00 (0.19, 5.32) Peng et al. (55), 2010 56 286 29 286 1.6 2.16 (1.33, 3.50) Shibuta et al. (17), 1994 26 234 31 329 1.3 1.20 (0.69, 2.08) Yeh et al. (47), 2007 157 715 154 730 3.6 1.05 (0.82, 1.35) Yoshida et al. (48), 2007 2 66 9 121 0.2 0.39 (0.08, 1.86) Yoshioka et al. (26), 1999 5 106 7 100 0.3 0.66 (0.20, 2.14) Subtotal (95% CI) 2584 3302 17.10 1.03 (0.85, 1.25) Total events 501 603 Heterogeneity: τ2=0.04; χ2=20.69, df=13 (P=0.08); I2=37% Test for overall effect: Z=0.29 (P=0.78)

European Agúndez et al. (47), 2000 60 120 139 258 1.8 0.86 (0.55, 1.32) Barrett et al. (31), 2003 304 490 349 592 3.7 1.14 (0.89, 1.45) Bell et al. (18), 1995 106 202 62 112 1.7 0.89 (0.56, 1.42) Borlak et al. (39), 2006 52 92 152 243 1.5 0.78 (0.48, 1.27) Gil et al. (23), 1998 48 114 120 201 1.6 0.49 (0.31, 0.78) Hubbard et al. (20), 1997 175 275 203 343 2.7 1.21 (0.87, 1.67) Kiss et al. (35), 2004 267 500 318 500 3.6 0.66 (0.51, 0.84) Landi et al. (38), 2005 192 360 162 308 2.9 1.03 (0.76, 1.40) Lilla et al. (40), 2006 295 503 374 601 3.7 0.86 (0.68, 1.10) Pistorius et al. (42), 2006 143 226 59 107 1.6 1.40 (0.88, 2.24) Sørensen et al. (51), 2008 211 377 445 768 3.6 0.92 (0.72, 1.18) Spurr et al. (19), 1995 71 103 62 96 1.1 1.22 (0.67, 2.20) Tiemersma et al. (30), 2002 59 102 300 537 1.9 1.08 (0.71, 1.66) Van der Hel et al. (33), 2003 146 258 495 857 3.2 0.95 (0.72, 1.26) Welfare et al. (21), 1997 101 174 99 174 1.9 1.05 (0.69, 1.60) Zupa et al. (54), 2009 50 92 50 121 1.3 1.69 (0.98, 2.92) Subtotal (95% CI) 3988 5818 38 0.96 (0.84, 1.09) Total events 2280 3389 Heterogeneity: τ2=0.03; χ2=30.12, df=15 (P=0.01); I2=50% Test for overall effect: Z=0.67 (P=0.50)

American Butler et al. (49), 2008 227 500 425 830 4.0 0.79 (0.63, 0.99) Chan et al. (36), 2005 107 183 267 443 2.5 0.93 (0.65, 1.32) Chen et al. (22), 1998 131 212 125 221 2.2 1.24 (0.85, 1.82) Cotterchio et al. (50), 2008 458 832 736 1247 4.7 0.85 (0.71, 1.02) Ishibe et al. (14), 2002 79 143 110 208 1.9 1.10 (0.72, 1.69) Kampman et al. (25), 1999 930 1624 1156 1963 5.5 0.94 (0.82, 1.07) Le Marchand et al. (13), 2001 124 543 157 654 3.4 0.94 (0.72, 1.23) Mahid et al. (46), 2007 68 122 126 222 1.8 0.96 (0.61, 1.50) Moslehi et al. (41), 2006 413 685 376 693 4.1 1.28 (1.03, 1.59) Nöthlings et al. (53), 2009 336 992 497 1493 4.9 1.03 (0.87, 1.22) Rodriguez et al. (15), 1993 24 44 15 28 0.5 1.04 (0.40, 2.69) Silva et al. (56), 2011 88 147 110 212 1.9 1.38 (0.90, 2.12) Slattery et al. (32), 2003 409 766 520 990 4.5 1.04 (0.86, 1.25) Wang et al. (57), 2011 281 301 320 355 1.2 1.54 (0.87, 2.72) Subtotal (95% CI) 7094 9559 43.10 1.00 (0.92, 1.10) Total events 3675 4940 Heterogeneity: τ2=0.01; χ2=19.84, df=13 (P=0.10); I2=34% Test for overall effect: Z=0.09 (P=0.93)

Australasian Butler et al. (29), 2001 149 209 143 200 1.9 0.99 (0.64, 1.52) Subtotal (95% CI) 209 200 1.9 0.99 (0.64, 1.52) Total events 149 143 Heterogeneity: Not applicable Test for overall effect: Z=0.05 (P=0.96)

Total (95% CI) 13875 18879 100 0.99 (0.93, 1.06) Total events 6605 9075 Heterogeneity: τ2=0.02; χ2=72.25, df=44 (P=0.005); I2=39% Test for overall effect: Z=0.18 (P=0.86) 2 2 0.1 0.2 0.5 1 2 5 10 Test for subgroup differences: χ =0.50, df=3 (P=0.92); I =0% Favors slow Favors slow cases controls

Fig. 5. Meta-analysis of NAT2 slow acetylator status and CRC susceptibility based on NAT2 genotypes Studies are stratified by regions. Medicina (Kaunas) 2012;48(3) 126 Hong Liu, Zhong-xue Fu, Chun-yi Wang, et al.

Table 3. Meta-analysis Results of NAT2 Acetylator Status Associated with Gender

Heterogeneity Gender No. of No. of No. of NAT2 OR 95% CI P Model Studies Cases Controls Status I2 P Fast 0.95 0.84–1.08 0.46 Male 7 2118 2392 Slow 1.05 0.93–1.18 0.46 0% 0.96 Fixed Fast 1.04 0.88–1.24 0.61 Female 9 2085 3397 Slow 0.96 0.81–1.13 0.61 43% 0.08 Random

Table 4. Meta-analysis Results of NAT2 Acetylator Status Associated with the Location of the Tumors

Heterogeneity Tumor No. of No. of No. of NAT2 OR 95% CI P value Model sites studies Cases Controls Status I2 P Fast 1.05 0.86–1.28 0.62 Proximal 6 678 2009 Slow 0.95 0.78–1.16 0.62 0% 0.68 Fixed Fast 0.98 0.77–1.25 0.86 Distal 5 471 1805 Slow 1.02 0.80–1.30 0.86 14% 0.32 Fixed

Table 5. Meta-analysis Results of NAT2 Acetylator Status Associated with Smoking

Heterogeneity Smoking No. of No. of No. of NAT2 OR 95% CI P value Model studies Cases Controls Status I2 P Fast 1.05 0.87–1.28 0.59 Never 5 682 1394 Slow 0.95 0.78–1.15 0.59 0% 0.44 Fixed Fast 1.03 0.80–1.34 0.80 Ever 6 827 1265 Slow 0.97 0.75–1.25 0.80 40% 0.14 Random associated with an increased risk of CRC among be attributable to genetic susceptibility to heterocy- NAT2 fast acetylators, but the elevated CRC risk clic amines, as determined by NAT2 genotype. associated with active smoking was not significantly Of the 45 studies included in our meta-analy- modified by NAT2 genotype. Slattery et al. (32) sis, 15 investigated the interactions between NAT2 found that current smokers who were fast acetyla- acetylator status and meat consumption. Of these tors were at slightly lower risk than current smokers 15 genotype studies that measured meat intake (13, who were slow acetylators; the risk was slightly less 21, 22, 25, 30–32, 36, 40, 49–53, 56), 13 investigat- than would be expected on an additive scale. Chan ed the hypothesized interaction of meat intake and et al. (36) reported that the interactions between NAT2 acetylator genotype. Studies by Welfare et al. genotype and either early or total lifetime smoking (21), Chen et al. (22), Kampman et al. (25), Chan et failed to achieve a statistical significance. Moreover, al. (36), Lilla et al. (40), and Silva et al. (56) showed Yoshida et al. (48) found that the distribution of significant evidence for the modification of NAT2 NAT2 genotypes was not associated with CRC risk and CRC by meat consumption. Welfare et al. re- in ever-smokers. ported that fast acetylators consuming fried meat In order to evaluate the combined ORs of these more than twice a week were at risk of CRC (21). studies, smoking status was reclassified as never or Chen et al. observed a stronger association between ever smoking since the definition of cutoff points for red meat intake and cancer risk among NAT2 rapid pack-years in these studies was a bit different. The acetylators, especially among men aged 60 years and combined results showed no significant evidence for more (22). Kampman et al. reported that the overall the modification of CRC risk associated with NAT2 mutagen index for red and white meat together was status by cumulative smoking exposure (Table 5), significantly positively associated with colon cancer which agreed with the conclusion of a meta-analysis risk among intermediate and rapid acetylators (25). conducted by Raimondi et al. (59). Chan et al. indicated that women with rapid acetylator genotypes experienced a greater risk associated Meat Consumption with intake of ≥0.5 serving of beef, pork, or lamb In an ecological study among 27 countries by as a main dish per day compared to intake of less Ognjanovic et al. (60), the authors concluded that in meat (36). Lilla et al. reported that the frequent con- combination with meat intake, some proportion of sumption of red meat significantly increased CRC the international variability in CRC incidence may risk for NAT2 fast acetylators (40). Silva et al. found

Medicina (Kaunas) 2012;48(3) NAT2 and Colorectal Cancer 127 that among NAT2 fast acetylators, meat consump- Discussion tion more than 3 times a week increased the risk NAT2 is involved in the metabolism of various of CRC (56). In contrast, Le Marchand et al. (13), potential carcinogens, such as HCAs and PAHs, and Tiemersma et al. (30), Barrett et al. (31), Sørensen it has been hypothesized that NAT2 genetic poly- et al. (51), Butler et al. (49), Nöthlings et al. (53), morphism may contribute to risk of CRC. A series and Kobayashi et al. (52) did not report any appar- of studies have been published, but no clear consen- ent effect of interaction between NAT2 genotypes sus has been reached. and meat intake on CRC risk. In 2002, Ye and Parry (61) conducted a meta- Because the categorizations and criteria of meat analysis using the published data from 20 case-con- consumption varied, there were no matching data trol studies. They reported that the pooling of stud- for combinations, and we did not evaluate the com- ies based on phenotyping methods indicated that bined ORs of these studies in this meta-analysis. the overall odds ratio of colon cancer risk associated with rapid acetylator was 1.51 (95% CI, 1.07–2.12). Sensitivity Analysis and Publication Bias However, the calculated overall odds ratio of colon Sensitivity analysis was performed subsequently, cancer risk associated with rapid acetylator from and it showed similar results when deleting a single the studies based on genotyping was 1.03 (95% CI, study involved in the meta-analysis each time, indi- 0.94–1.12), consistent with what was observed in cating that our results were statistically robust. this larger analysis. Pooling studies were also con- A funnel plot was performed to evaluate the ducted on specific tumor sites and ethnic groups. publication bias of the literature. There does not ap- The results showed that the effect of rapid acetyla- pear to be an obvious publication bias among all the tor on colon cancer risk was not obviously different. genotype studies extracted to this meta-analysis, for Therefore, the authors concluded that NAT2 alone the shape of the funnel plot is symmetry, similar to was not an important risk factor for colon cancer, an invert funnel (Fig. 6). However, in the phenotype and NAT2 rapid acetylation status had no specific studies, there may be a publication bias because the effect on the risk of developing colon cancer. plot was asymmetrical (Fig. 7). In the same year, de Jong et al. (62) also performed a meta-analysis to detect low-penetrance genes and their involvement in CRC susceptibility. 0.0 The pooled analysis for phenotype studies revealed a positive association between fast acetylatorship 0.2 and CRC, but genotype studies detected neither an association between CRC and presumed fast acety- 0.4 latorship overall nor in subgroup analyses for ethnicity, gender, and tumor localization. 0.6 Three years later, Chen et al. (63) conducted a SE(log[OR]) meta-analysis to clarify the influence of genetic pol- 0.8 ymorphisms on CRC. They found that NAT2 rapid acetylator phenotype (pooled OR, 1.15), but not 1.0 0.1 0.2 0.5 1 2 5 10 NAT2 rapid acetylator genotype (pooled OR, 1.05), OR had a significantly increased risk of CRC (P<0.05). Why does the pooled analysis of the phenotype- Fig. 6. Funnel plot of genotype studies based studies conducted by Ye and Parry tend to be positive and that by de Jong et al. and Chen et 0.0 al. negative? It is likely that methodological differences, such as different criteria of including a study 0.2 for analysis, may underlie the somewhat different observations across the meta-analysis studies. For 0.4 example, Ye and Parry, and de Jong et al. reported different results, which are likely due to the fact that 0.6 Ye and Parry included additional studies by Wohl- SE(log[OR]) leb et al. (64) and Lang et al. (1994), and de Jong 0.8 et al. included additional studies by Robers-Thom- son et al. (65), which contained the patients with 1.0 colorectal adenomatous polyps, and by Lang et al. 0.05 0.2 1 5 20 (1986), though both studies of Ye and Parry, and OR de Jong et al. included the same 2 studies (8, 11). In Fig. 7. Funnel plot of phenotype studies the meta-analysis of Chen et al., a study (66) inves-

Medicina (Kaunas) 2012;48(3) 128 Hong Liu, Zhong-xue Fu, Chun-yi Wang, et al. tigating NAT (not only NAT2) polymorphism and of which are derived from tobacco smoke (67). The gastrointestinal carcinoma (containing gastric carci- action of NATs on these carcinogens can generate noma) and also 2 studies based on genotype (17, electrophilic ions capable of inducing DNA point 25) were included. Comparing to the study of Ye mutations, so smoking may interact with NAT2 and Parry, our meta-analysis included 4 more stud- polymorphism. However, the results of epidemio- ies (10, 12–14) and still found a less significant as- logic studies were incompatible as described previ- sociation between the rapid or slow acetylator status ously, and it may be partly explained by the com- and CRC risk. plexity of tobacco smoke constituents, variation in Furthermore, phenotype is an expression of the metabolism of smoking, and differences in a study actual ability to metabolize the relevant chemicals. design. In a meta-analysis conducted by Raimondi There is some evidence that the alteration of acety- et al. (59), the authors made a pooling and found lation phenotype was influenced by a number of the a nonsignificant positive interaction between NAT2 following factors: 1) disease status; 2) liver and renal genetic polymorphism and smoking for CRC risk, functions; 3) selection or participation bias of case consistent with our results. and/or controls and analytical method used; and Another factor that has been investigated as a po- 4) overlapping activity of NAT1, and in addition, tential modifier of the NAT2 and CRC association misclassification may also be important because dif- is meat consumption. The consumption of meat, ferent xenobiotics and methods were used to assess especially cooked at high temperature, is associated NAT2 activity. Therefore, it is likely that genotype with exposure to HCAs. It has been shown that after studies more accurately reflect the risk attributable absorption, they need to be bioactivated or detoxi- to NAT2 acetylation status, which was shown by the cated by N-acetyltransferase enzymes or enzymes consistent results of genotype studies in the present of other family, such as cytochrome P450 (CYP), 4 meta-analyses. glutathione S-transferase (GST), and sulfotrans- Our study was built upon these previous meta- ferase (SULT), before they can damage DNA (68). analyses with a more comprehensive and thorough Therefore, the effect modification of the association assessment of NAT2 polymorphism and CRC sus- between meat consumption and CRC risk by NAT2 ceptibility and included some new studies published polymorphisms has been suggested. However, the after 2005. Importantly, we also did a pooled analy- associations were not consistent across studies. The sis of raw data from a large sample size (phenotype role of NAT2 polymorphisms in the effect modifi- studies, 791 cases and 1158 controls; genotype stud- cation of environmental carcinogens should be as- ies, 13 875 cases and 18 879 controls) to corroborate sessed in well-designed, large-scale epidemiological the meta-analysis. The pooling of phenotype stud- studies with comprehensive information on risk fac- ies showed no significant association between the tors for better understanding the etiologic role of NAT2 acetylator status and CRC risk (rapid acety- dietary factors. lator, OR, 1.32; 95% CI, 0.92–1.89, P=0.14; slow Regarding NAT2, the results of our meta-analysis acetylator, OR, 0.76; 95% CI 0.53–1.09; P=0.14). of the genotype studies failed to show a significant The combined ORs for rapid and slow acetylator association between NAT2 polymorphism and CRC status and CRC risk in genotype studies were 1.01 risk. Molecular biologically, because the products of (95% CI, 0.94–1.08, P=0.86) and 0.99 (95% CI, several genes interact, NAT2 polymorphisms may 0.93–1.06, P=0.86), respectively. In the subgroup be not associated with CRC alone, and an associa- analysis by regions, no increased risks were found tion with CRC is still possible in combination with in Asians, Europeans, Americans, or Australa- polymorphisms of other genes. For two of these sians. Pooling studies were also conducted on the combinations, an association with CRC was shown groups of gender and specific tumor sites, but re- with the combined high-risk genotypes of CYP1A2 sults showed no significant association in genotype and NAT2 (12) and of GSTT1 and NAT2 (69). Only distribution between CRC and control as well. It is with specific knowledge of whether other genes are possible that the results would be more confidence involved, it will be possible to recognize epidemio- in the meta-analysis because of the larger numbers logically the exact estimate of risk conferred via one of cases and controls. particular gene. Future studies should also measure It is widely recognized that not only the main the interaction of NAT2 and other genes in studies effect of a gene, but also the influence of gene-en- involving large numbers of patients and controls. vironmental or gene-diet interactions on cancer risk Several limitations of this meta-analysis should are important. be interpreted with caution. Firstly, some included An interaction between NAT2 genetic polymor- studies and stratified analyses were limited by the phism and smoking in cancer risk received great at- relatively small sample size. Secondly, our results tention in some pieces of research. NAT2 is a phase II were based on unadjusted estimates, and a more metabolizing enzyme detoxifying arylamines, some precise analysis would have been performed if all

Medicina (Kaunas) 2012;48(3) NAT2 and Colorectal Cancer 129 individual raw data had been available. Thirdly, of genotype studies may be unbiased. Third, when gene polymorphisms could modify the association comparing with previous meta-analyses (61–63), between smoking and CRC only in specific catego- we considered not only association between NAT2 ries of smokers such as long-term smokers (70), and polymorphism and CRC susceptibility by pheno- other factors such as age at initiation of smoking type and genotype studies separately and stratified and years since smoking cessation for former smok- to subgroup analyses for regions, gender, and tumor ers could play a role. In our analysis, smoking sta- localization, but also paid attention to the impact of tus was reclassified as never or ever smoking, and it NAT2 and environmental factors, such as smoking might be that this classification was not accurate and and meat consumption, on CRC. We could there- did not reflect the real association between NAT2 fore give a more complete picture on the role of polymorphism and smoking in colorectal cancer. NAT2 polymorphisms contributing to CRC risk. Finally, as in most meta-analyses, publication bias must be considered because only published studies Conclusions were included in the meta-analysis. Following the Findings from this meta-analysis and pooled construction of a funnel plot (Fig. 7), we conclude analysis indicate that there is no overall association that there is some degree of publication bias in the between NAT2 polymorphism and CRC suscepti- phenotype studies. Therefore, we cannot exclude bility. It is of great important to conduct large-scale this probability in our meta-analysis, and such a studies using standardized unbiased phenotyping or situation may lead to incorrect conclusions. genotyping methods, homogeneous CRC patients, In spite of these limitations, our meta-analysis and well-matched controls. Moreover, future stud- had some strengths. First, the sufficient number ies evaluating smoking and meat intake should try of cases and controls were pooled from multiple to collect and use standardized exposure measures, studies, which significantly increased the statistical which would greatly help summarize the results of power of our analysis. Second, the symmetry of the related studies. funnel plot among all the genotype studies (Fig. 6) suggests that in these publications, bias is less likely Statement of Conflicts of Interest to have appeared, indicating that the pooled results The authors state no conflict of interest.

References RJ, Yerokun T, et al. Acetylator genotype-dependent ex- 1. de Jong MM, Nolte IM, te Meerman GJ, van der Graaf WT, pression of arylamine N-acetyltransferase in human colon de Vries EG, Sijmons RH, et al. Low-penetrance genes and cytosol from non-cancer and colorectal cancer patients. their involvement in colorectal cancer susceptibility. Cancer Cancer Res 1991;51(2):549-55. Epidemiol Biomarkers Prev 2002;11(11):1332-52. 11. Ladero JM, González JF, Benítez J, Vargas E, Fernández 2. Cross AJ, Sinha R. Meat-related mutagens/carcinogens in MJ, Baki W, et al. Acetylator polymorphism in human the etiology of colorectal cancer. Environ Mol Mutagen colorectal carcinoma. Cancer Res 1991;51(8):2098-100. 2004;44(1):44-55. 12. Lang NP, Butler MA, Massengill J, Lawson M, Stotts RC, 3. Blum M, Grant DM, McBride W, Heim M, Meyer UA. Hu- Hauer-Jensen M, et al. Rapid metabolic phenotypes for man arylamine N-acetyltransferase genes: isolation, chro- acetyltransferase and cytochrome P4501A2 and putative ex- mosomal localization, and functional expression. DNA Cell posure to food-borne heterocyclic amines increase the risk Biol 1990;9(3):193-203. for colorectal cancer or polyps. Cancer Epidemiol Biomark- 4. Walraven JM, Zang Y, Trent JO, Hein DW. Structure/func- ers Prev 1994;3(8):675-82. tion evaluations of single nucleotide polymorphisms in hu- 13. Le Marchand L, Hankin JH, Wilkens LR, Pierce LM, Franke man N-acetyltransferase 2. Curr Drug Metab 2008;9(6): A, Kolonel LN, et al. Combined effects of well-done red 471-86. meat, smoking, and rapid N-acetyltransferase 2 and CY- 5. Hengstler JG, Arand M, Herrero ME, Oesch F. Polymor- P1A2 phenotypes in increasing colorectal cancer risk. Can- phism of N-acetyltransferases, glutathione s-transferases, cer Epidemiol Biomarkers Prev 2001;10(12):1259-66. microsomal epoxide hydrolase and sulfotransferases: influ- 14. Ishibe N, Sinha R, Hein DW, Kulldorff M, Strickland P, ence on cancer susceptibility. Recent Results Cancer Res Fretland AJ, et al. Genetic polymorphisms in heterocyclic 1998;154:47-85. amine metabolism and risk of colorectal adenomas. Phar- 6. Kiyohara C. Genetic polymorphism of enzymes involved in macogenetics 2002;12(2):145-50. xenobiotic metabolism and the risk of colorectal cancer. J 15. Rodriguez JW, Kirlin WG, Ferguson RJ, Doll MA, Gray Epidemiol 2000;10(5):349-60. K, Rustan TD, et al. Human acetylator genotype: rela- 7. Lang NP, Chu DZ, Hunter CF, Kendall DC, Flammang TJ, tionship to colorectal cancer incidence and arylamine N- Kadlubar FF. Role of aromatic amine acetyltransferase in acetyltransferase expression in colon cytosol. Arch Toxicol human colorectal cancer. Arch Surg 1986;121(11):1259-61. 1993;67(7):445-52. 8. Ilett KF, David BM, Detchon P, Castleden WM, Kwa R. 16. Oda Y, Tanaka M, Nakanishi I. Relation between the oc- Acetylation phenotype in colorectal carcinoma. Cancer Res currence of K-ras gene point mutations and genotypes of 1987;47(5):1466-9. polymorphic N-acetyltransferase in human colorectal carci- 9. Wohlleb JC, Hunter CF, Blass B, Kadlubar FF, Chu DZ, nomas. Carcinogenesis 1994;15(7):1365-9. Lang NP. Aromatic amine acetyltransferase as a marker for 17. Shibuta K, Nakashima T, Abe M, Mashimo M, Mori M, colorectal cancer: environmental and demographic associa- Ueo H, et al. Molecular genotyping for N-acetylation poly- tions. Int J Cancer 1990;46(1):22-30. morphism in Japanese patients with colorectal cancer. Can- 10. Kirlin WG, Ogolla F, Andrews AF, Trinidad A, Ferguson cer 1994;74(12):3108-12.

Medicina (Kaunas) 2012;48(3) 130 Hong Liu, Zhong-xue Fu, Chun-yi Wang, et al.

18. Bell DA, Stephens EA, Castranio T, Umbach DM, Watson 35. Kiss I, Németh A, Bogner B, Pajkos G, Orsós Z, Sándor J, M, Deakin M, et al. Polyadenylation polymorphism in the et al. Polymorphisms of glutathione-S-transferase and ar- acetyltransferase 1 gene (NAT1) increases risk of colorectal ylamine N-acetyltransferase enzymes and susceptibility to cancer. Cancer Res 1995;55(16):3537-42. colorectal cancer. Anticancer Res 2004;24(6):3965-70. 19. Spurr NK, Gough AC, Chinegwundoh FI, Smith CA. Poly- 36. Chan AT, Tranah GJ, Giovannucci EL, Willett WC, Hunter morphisms in drug-metabolizing enzymes as modifiers of DJ, Fuchs CS. Prospective study of N-acetyltransferase-2 cancer risk. Clin Chem 1995;41(12):1864-9. genotypes, meat intake, smoking and risk of colorectal can- 20. Hubbard AL, Harrison DJ, Moyes C, Wyllie AH, Cunning- cer. Int J Cancer 2005;115(4):648-52. ham C, Mannion E, et al. N-acetyltransferase 2 genotype 37. Chen K, Jin MJ, Fan CH, Song L, Jiang QT, Yu WP, et al. in colorectal cancer and selective gene retention in cancers [A case-control study on the association between genetic with chromosome 8p deletions. Gut 1997;41(2):229-34. polymorphisms of metabolic enzymes and the risk of colo- 21. Welfare MR, Cooper J, Bassendine MF, Daly AK. Relation- rectal cancer]. Zhonghua Liu Xing Bing Xue Za Zhi 2005; ship between acetylator status, smoking, diet and colorec- 26(9):659-64. tal cancer risk in the northeast of England. Carcinogenesis 38. Landi S, Gemignani F, Moreno V, Gioia-Patricola L, Chab 1997;18(7):1351-4. rier A, Guino E, et al. A comprehensive analysis of phase 22. Chen J, Stampfer MJ, Hough HL, Garcia-Closas M, Willett I and phase II metabolism gene polymorphisms and risk WC, Hennekens CH, et al. A prospective study of N-acetyl- of colorectal cancer. Pharmacogenet Genomics 2005;15(8): transferase genotype, red meat intake, and risk of colorectal 535-46. cancer. Cancer Res 1998;58(15):3307-11. 39. Borlak J, Reamon-Buettner SM. N-acetyltransferase 2 23. Gil JP, Lechner MC. Increased frequency of wild-type ar- (NAT2) gene polymorphisms in colon and lung cancer pa- ylamine-N-acetyltransferase allele NAT2*4 homozygotes tients. BMC Med Genet 2006;7:58. in Portuguese patients with colorectal cancer. Carcinogen- 40. Lilla C, Verla-Tebit E, Risch A, Jäger B, Hoffmeister M, esis 1998;19(1):37-41. Brenner H, et al. Effect of NAT1 and NAT2 genetic poly- 24. Lee ED, Zhao B, Seow-Choen F. Relationship between pol- morphisms on colorectal cancer risk associated with expo- ymorphism of N-acetyltransferase gene and susceptibility sure to tobacco smoke and meat consumption. Cancer Epi- to colorectal carcinoma in a Chinese population. Pharma- demiol Biomarkers Prev 2006;15(1):99-107. cogenetics 1998;8(6):513-7. 41. Moslehi R, Chatterjee N, Church TR, Chen J, Yeager M, 25. Kampman E, Slattery ML, Bigler J, Leppert M, Samow- Weissfeld J, et al. Cigarette smoking, N-acetyltransferase itz W, Caan BJ, et al. Meat consumption, genetic suscep- genes and the risk of advanced colorectal adenoma. Phar- tibility, and colon cancer risk: a United States multicenter macogenomics 2006;7(6):819-29. case-control study. Cancer Epidemiol Biomarkers Prev 42. Pistorius S, Görgens H, Krüger S, Engel C, Mangold E, 1999;8(1):15-24. Pagenstecher C, et al. N-acetyltransferase (NAT) 2 acetyla- 26. Yoshioka M, Katoh T, Nakano M, Takasawa S, Nagata tor status and age of onset in patients with hereditary non- N, Itoh H. Glutathione S-transferase (GST) M1, T1, P1, polyposis colorectal cancer (HNPCC). Cancer Letters 2006; N-acetyltransferase (NAT) 1 and 2 genetic polymorphisms 241(1):150-7. and susceptibility to colorectal cancer. J UOEH 1999;21(2): 43. Huang CC, Chien WP, Wong RH, Cheng YW, Chen MC, 133-47. Chou MC, et al. NAT2 fast acetylator genotype is associated 27. Agúndez JAG, Lozano L, Ladero JM, Sastre J, Cerdán FJ, with an increased risk of colorectal cancer in Taiwan. Dis Diaz-Rubio M, et al. N-acetyltransferase 2 (NAT2) geno- Colon Rectum 2007;50(7):981-9. type and colorectal carcinoma: risk variability according to 44. Yi J, Bing X, Zhan-xiong X, Jun-jie Z, Lai-fang S, Bo Z. tumor site? Scand J Gastroenterol 2000;35(10):1087-91. [Association between N-acetyltransferase 2 gene polymor- 28. Katoh T, Boissy R, Nagata N, Kitagawa K, Kuroda Y, Itoh phisms and genetic susceptibility to sporadic colorectal ad- H, et al. Inherited polymorphism in the N-acetyltransferase enocarcinoma.] Zhonghua Liu Xing Bing Xue Za Zhi 2007; 1 (NAT1) and 2 (NAT2) genes and susceptibility to gastric 27(1):15-8. and colorectal adenocarcinoma. Int J Cancer 2000;85(1): 45. Rui-zhen L, Zhen-ya Z, Yue-ming Y, Lu-jun H, Fang Q. 46-9. The relationship between NAT2 polymorphism and colo- 29. Butler WJ, Ryan P, Roberts-Thomson IC. Metabolic geno- rectal cancer susceptibility. J Pract Med Techn 2007;14(20): types and risk for colorectal cancer. J Gastroenterol Hepatol 2705-7. 2001;16(6):631-5. 46. Mahid SS, Colliver DW, Crawford NP, Martini BD, Doll 30. Tiemersma EW, Kampman E, Bueno de Mesquita HB, MA, Hein DW, et al. Characterization of N-acetyltrans- Bunschoten A, van Schothorst EM, Kok FJ, et al. Meat ferase 1 and 2 polymorphisms and haplotype analysis for consumption, cigarette smoking, and genetic susceptibility inflammatory bowel disease and sporadic colorectal carci- in the etiology of colorectal cancer: results from a Dutch noma. BMC Medical Genet 2007;8:28. prospective study. Cancer Causes Control 2002;13(4):383- 47. Yeh CC, Sung FC, Tang R, Chang-Chieh CR, Hsieh LL. 93. Association between polymorphisms of biotransformation 31. Barrett JH, Smith G, Waxman R, Gooderham N, Light- and DNA-repair genes and risk of colorectal cancer in Tai- foot T, Garner RC, et al. Investigation of interaction be- wan. J Biomed Sci 2007;14(2):183-93. tween N-acetyltransferase 2 and heterocyclic amines as 48. Yoshida K, Osawa K, Kasahara M, Miyaishi A, Nakanishi potential risk factors for colorectal cancer. Carcinogenesis K, Hayamizu S, et al. Association of CYP1A1, CYP1A2, 2003;24(2):275-82. GSTM1 and NAT2 gene polymorphisms with colorectal 32. Slattery ML, Edwards S, Curtin K, Schaffer D, Neuhausen cancer and smoking. Asian Pacific J Cancer Prev 2007;8(3): S. Associations between smoking, passive smoking, GSTM- 438-44. 1, NAT2, and rectal cancer. Cancer Epidemiol Biomarkers 49. Butler LM, Millikan RC, Sinha R, Keku TO, Winkel S, Prev 2003;12(9):882-9. Harlan B, et al. Modification by N-acetyltransferase 1 geno- 33. van der Hel OL, Bueno de Mesquita HB, Sandkuijl L, van type on the association between dietary heterocyclic amines Noord PA, Pearson PL, Grobbee DE, et al. Rapid N-acetyl- and colon cancer in a multiethnic study. Mutat Res 2008; transferase 2 imputed phenotype and smoking may increase 638(1-2):162-74. risk of colorectal cancer in women (Netherlands). Cancer 50. Cotterchio M, Boucher BA, Manno M, Gallinger S, Okey Causes Control 2003;14(3):293-8. AB, Harper PA. Red meat intake, doneness, polymorphisms 34. He LJ, Yu YM, Qiao F, Liu JS, Sun XF, Jiang LL. Genetic in genes that encode carcinogen-metabolizing enzymes, polymorphisms of N-acetyltransferase 2 and colorectal can- and colorectal cancer risk. Cancer Epidemiol Biomarkers cer risk. World J Gastroenterol 2005;11(27):4268-71. Prev 2008;17(11):3098-107.

Medicina (Kaunas) 2012;48(3) NAT2 and Colorectal Cancer 131

51. Sørensen M, Autrup H, Olsen A, Tjønneland A, Overvad search 2009;670(1-2):6-14. K, Raaschou-Nielsen O. Prospective study of NAT1 and 60. Ognjanovic S, Yamamoto J, Maskarinec G, Le Marchand L. NAT2 polymorphisms, tobacco smoking and meat consum NAT2, meat consumption and colorectal cancer incidence: ption and risk of colorectal cancer. Cancer Letters 2008; an ecological study among 27 countries. Cancer Causes 266(2):186-93. Control 2006;17(9):1175-82. 52. Kobayashi M, Otani T, Iwasaki M, Natsukawa S, Shaura K, 61. Ye Z, Parry JM. Meta-analysis of 20 case-control studies on Koizumi Y, et al. Association between dietary heterocyclic the N-acetyltransferase 2 acetylation status and colorectal amine levels, genetic polymorphisms of NAT2, CYP1A1, cancer risk. Med Sci Monit 2002;8(8):CR558-65. and CYP1A2 and risk of colorectal cancer: a hospital-based 62. de Jong MM, Nolte IM, te Meerman GJ, van der Graaf WT, case-control study in Japan. Scand J Gastroenterol 2009; de Vries EG, Sijmons RH, et al. Low-penetrance genes and 44(8):952-9. their involvement in colorectal cancer susceptibility. Cancer 53. Nöthlings U, Yamamoto JF, Wilkens LR, Murphy SP, Park Epidemiol Biomarkers Prev 2002;11(11):1332-52. SY, Henderson BE, et al. Meat and heterocyclic amine in- 63. Chen K, Jiang QT, He HQ. Relationship between metabolic take, smoking, NAT1 and NAT2 polymorphisms, and colo- enzyme polymorphism and colorectal cancer. World J Gas- rectal cancer risk in the multiethnic cohort study. Cancer troenterol 2005;11(3):331-5. Epidemiol Biomarkers Prev 2009;18(7):2098-106. 64. Wohlleb JC, Hunter CF, Blass B, Kadlubar FF, Chu DZ, 54. Zupa A, Sgambato A, Bianchino G, Improta G, Grieco V, Lang NP. Aromatic amine acetyltransferase as a marker for LA Torre G, et al. GSTM1 and NAT2 polymorphisms and colorectal cancer: environmental and demographic associa- colon, lung and bladder cancer risk: a case-control study. tions. Int J Cancer 1990;46(1):22-30. Anticancer Res 2009;29(5):1709-14. 65. Roberts-Thomson IC, Ryan P, Khoo KK, Hart WJ, McMi- 55. Xian-e P, Ying-ying J, Lin L, Zhi-min H, Zhi-jian H, Xi- chael AJ, Butler RN. Diet, acetylator phenotype, and risk of shun S. Relationship between environmental exposures, ge- colorectal neoplasia. Lancet 1996;347(9012):1372-4. netic polymorphism of NAT2 and colorectal cancer. Cancer 66. Hua C, Xiangfu Z, Dongpo X, Yi-fan Z, Chang-sheng S, Res Clin 2010;22(2):89-91. Zhi-hong W. Study on the NAT polymorphisms among 56. da Silva TD, Felipe AV, de Lima JM, Oshima CT, Foro- gastrointestinal carcinoma patients in Fujian. Cancer Res nes NM. N-Acetyltransferase 2 genetic polymorphisms and Prev Treat 1999;26(3):232-3. risk of colorectal cancer. World J Gastroenterol 2011;17(6): 67. Bartsch H, Nair U, Risch A, Rojas M, Wikman H, Alexan- 760-5. drov K. Genetic polymorphism of CYP genes, alone or in 57. Wang J, Joshi AD, Corral R, Siegmund KD, Marchand LL, combination, as a risk modifier of tobacco-related cancers. Martinez ME, et al. Carcinogen metabolism genes, red Cancer Epidemiol Biomarkers Prev 2000;9(1):3-28. meat and poultry intake, and colorectal cancer risk. Int J 68. Nowell SA, Ahn J, Ambrosone CB. Gene-nutrient interac- Cancer 2011; doi: 10.1002/ijc.26199. tions in cancer etiology. Nutr Rev 2004;62(11):427-38. 58. Zhong S, Wyllie AH, Barnes D, Wolf CR, Spurr NK. Re- 69. Welfare M, Monesola Adeokun A, Bassendine MF, Daly lationship between the GSTM1 genetic polymorphism and AK. Polymorphisms in GSTP1, GSTM1, and GSTT1 and susceptibility to bladder, breast and colon cancer. Carcino- susceptibility to colorectal cancer. Cancer Epidemiol Bio- genesis 1993;14(9):1821-4. markers Prev 1999;8(4 Pt 1):289-92. 59. Raimondi S, Botteri E, Iodice S, Lowenfels AB, Maison- 70. Botteri E, Iodice S, Bagnardi V, Raimondi S, Lowenfels neuve P. Gene-smoking interaction on colorectal adenoma AB, Maisonneuve P. Smoking and colorectal cancer: a me- and cancer risk: review and meta-analysis. Mutation Re- ta-analysis. JAMA 2008;300(23):2765-78.

Received 19 January 2012, accepted 13 February 2012

Medicina (Kaunas) 2012;48(3)