Commentary: Reflections on G. M. Lower and colleagues’ 1979 study associating slow acetylator phenotype with urinary bladder cancer: meta-analysis, historical refinements of the hypothesis, and lessons learned
1Division of Cancer Epidemiology and Genetics, NCI, NIH, DHHS, Bethesda, MD20892, USA.
2Department of Pharmacology & Toxicology and James Graham Brown Cancer Center, University of Louisville School of Medicine, Louisville, KY, USA.
*Corresponding author. Occupational and Environmental Epidemiology Branch, Division of Cancer Epidemiology and Genetics, NCI, NIH EPS 8116, Bethesda, MD 20892, USA. E-mail: rothmann{at}mail.nih.gov
Keywords Bladder neoplasms, genetic susceptibility, NAT2, N-acetylation, tobacco
Accepted 5 February 2007
In 1979, Gerald Lower and colleagues published the first epidemiologic study to directly test the hypothesis that individuals exposed to aromatic amines would be at higher risk of urinary bladder cancer if they had the slow acetylator phenotype.1 Not only was the paper a landmark in the history of bladder cancer carcinogenesis, it was one of the very first efforts to study the effect of a metabolic polymorphism on risk for any cancer.
Lower began his paper1 by summarizing and integrating the empirical and conceptual basis for his hypothesis. After noting the initial observation of the bladder cancer excess among dye workers by Rehn in 1895,2 he described the key findings by investigators who had made seminal observations in the epidemiology of bladder cancer; performed the fundamental experimental studies of aromatic amine carcinogenesis; made the key clinical observations that the slow N-acetylation phenotype was associated with isoniazid neurotoxicity; carried out genetic studies of acetylator phenotype; and linked acetylator phenotype with rapid and slow N-acetylation of aromatic amines (citing work by Case, Cole, Hoover, Fraumeni, and Wynder; Hueper, James and Elizabeth Miller, Poirier, King, Radomski, John and Elizabeth Weisburger, and Kadlubar; and Hughes, Evans, and Weber, among others). Lower had also carried out his own studies of aromatic amine carcinogenesis while at the University of Wisconsin-Madison, well before his epidemiologic studies, reporting in 1973 on the relative ability of liver cytosols from several species including the dog to N-acetylate 4-aminobiphenyl and 2-aminonapthelene. He then went on to integrate his findings with previous observations, particularly reports that several arylamines were carcinogenic in the dog, which cannot N-acetylate, to lend support to the hypothesis that N-acetylation was not necessary for aromatic amine-induced bladder carcinogenesis, and was, in fact, a likely detoxification mechanism.3
Lower was compelled to directly test this hypothesis in human populations, and carried out two small case-control studies of slow acetylator phenotype and bladder cancer. He studied 71 cases and 74 controls from Copenhagen, and 115 cases and 118 controls from a rural region in Sweden. He reported in his 1979 paper a suggestive increased effect for the slow acetylator phenotype [odds ratio (OR) = 1.74, 95% confidence interval (CI) = 0.85–3.59] in the former, and no association in the latter (OR = 1.13, 95% CI = 0.63–2.04). Lower speculated that the positive association might reflect higher exposures to aromatic amines in the urban setting of the Danish study, and this hypothesis received further support in a subsequent review in which the effect of acetylator phenotype on bladder cancer risk was compared among studies with and without documented exposure to aromatic amines.4 However, the CIS suggest that the effects from his Danish and Swedish studies were consistent with each other, and with our overall meta-analysis estimate of the main effect of slow acetylation and bladder cancer risk (OR = 1.37, 95% CI = 1.22–1.54, P = 2 x 10–7), among Caucasians only, and among Caucasians in European studies (Figure 1). In fact, the central pooled estimate of his two studies (OR = 1.35, 95% CI = 0.86–2.10) is very similar to that from the meta-analysis of the 22 studies in Caucasian populations carried out subsequently!
|
Lower finished his article with a discussion of some of the basic and now well-established principles of molecular epidemiology, which was probably the very first description of a broad framework for considering the application of biomarkers in cancer epidemiology.5 He emphasized the multi-disciplinary nature of the evolving field of molecular epidemiology, and noted the key importance of collecting detailed information on external exposure, including age of initial exposure, intensity and duration; the use of biologic markers to show that external agents are absorbable; the importance of broadly assessing individual susceptibility including both activation and detoxification capacity; and the need to study how these factors relate to each other. It was also clear that he was interested in the broader societal, philosophical and organizational context of scientific inquiry, which he developed further in an article in 1982,6 and which became and still is an important focus of his life.
Much has ensued in the study of the acetylator phenotype and genotype, aromatic amine exposure and bladder cancer since the publication of Lower's paper in 1979, both in the laboratory and in population studies. A key advancement is our understanding that the slow acetylator phenotype reflects the metabolic capacity of the N-acetyltransferase 2 (NAT2) isozyme. The molecular biology and biochemistry of NAT2 and its many functional variants has been recently reviewed.7 Here, we comment on the epidemiologic studies of NAT2 and bladder cancer that have been carried out since 1979.
There are two lines of research that ensued from Lower's observations—one in the general population setting, where bladder cancer is caused overwhelmingly by tobacco, and the other in the occupational setting, among workers who could have been exposed to very high levels of carcinogenic aromatic amines in their workplace. To summarize the effect of NAT2 slow acetylation phenotype and genotype, we have updated our previous meta-analysis.8 We have identified a total of 36 studies that have analysed NAT2 phenotype or genotype and bladder cancer, and were carried out in a general population setting. We excluded studies that were formally, or informally, nested within cohorts of workers exposed to aromatic amines, as well as studies carried out where infectious agents are likely to be a key aetiological risk factor (i.e. shistosomiasis endemic regions). The overall pooled estimate of NAT2 slow acetylation and bladder cancer risk is highly significant (Figure 1). As the test for heterogeneity of OR across studies is significant (Figure 1), we stratified studies by ethnicity and geographic region i.e. Caucasian (and within this group, studies in Europe and the USA), and Asian. The estimate for the 24 studies that have been conducted among Caucasians was homogenous and without evidence of publication bias (Figure 1). The estimated effect is weaker in the US than in Europe, however, most studies have been relatively small, and a final conclusion about the association may change, as results from a number of larger studies should be available in the not-too-distant future. We note that a study testing this hypothesis in Caucasians would need at least 670 cases and 670 controls to be able to detect the current estimate of the effect of NAT2 slow acetylation with 80% power, which is larger than any single published study to date in the US (Figure 1). We found a borderline, statistically significant effect among Asians, of similar magnitude to that observed in Caucasians. However, power to detect a risk of 1.37 is substantially lower among Asians compared to Caucasians, given the markedly lower prevalence of the NAT2 slow acetylator phenotype in the former. Indeed, one would need to study at least 1200 cases and 1200 controls to detect an OR of 1.37 in a study of Asians (assuming a genotype prevalence of 14% and power of 80%), larger than the total number of cases from all studies published to date (Figure 1).
Can we conclude that the association between NAT2 slow acetylation and bladder cancer in the general population is a true positive finding? There is a large body of work from experimental in vitro and in vivo studies, and from studies of intermediate endpoints in humans,9,10 that support the biological plausibility of the association and justify a relatively high prior probability that the association is likely.11 Given the observed relative risk estimate, 95% CI, and associated P-value from our meta-analysis (Figure 1), we can reasonably conclude that there is a high likelihood that the association is a true one, using criteria that have been proposed by several investigators,11,12 and that Lower's basic hypothesis has been confirmed.
We have used a case-only approach to test Lower's prediction that NAT2 slow acetylation is likely to impact on bladder cancer in the general population more substantially among those who are exposed to aromatic amines, i.e. smokers. We found evidence of a significant, albeit modest multiplicative interaction that satisfies tests of heterogeneity (Figure 2) and is without evidence of publication bias. However, the magnitude of the effect, the 95% CI and the associated P-value of the association (Figure 2) are such that the interaction cannot be accepted with certainty, and will require confirmation by additional data. Further, stratification of the analysis by ethnicity and geographic location suggests that the interaction is present only among European studies, with little evidence in US or Asian studies. However, the meta-analyses of studies in the US and Asia are both underpowered (i.e. the minimum sample size required to attain 80% power to detect an interaction of 1.39 was estimated to be about 1800 and 3100 cases, respectively—Figure 2). Nevertheless, this observation does lend some support to the hypothesis that the higher aromatic amine content present in black tobacco,10 used much more commonly in Europe, provides conditions where the influence of NAT2 phenotype on bladder cancer risk may be greatest, and that at lower aromatic amine concentrations NAT2 phenotype could be less relevant. Indeed, in the largest study of NAT2 genotype, tobacco use, and bladder cancer reported to date, which was carried out in Spain, the data suggested that the interaction between NAT2 genotype and tobacco is stronger with increasing level of tobacco intensity.8,13 At the same time, others have suggested that such interactions might be expected to take place primarily among those exposed to lower levels of tobacco.14 We note that a recent Bayesian analysis of NAT2 phenotype/genotype and bladder cancer studies found convincing evidence that NAT2 slow acetylation was associated with bladder cancer, but only among smokers.15
|
The second line of research that was inspired and explicitly recommended by Lower was the study of populations with high occupational exposure to aromatic amines.1 In this instance, findings have turned out to be more complex than Lower anticipated. The NAT2 slow acetylator phenotype/genotype has been associated for the most part with increased risk of bladder cancer among workers exposed to aromatic amines4,16 (though usually not specified as to type, and if so, without the provision of supporting data), beginning with Cartwright's notable paper in the Lancet,17 although the magnitude of the effect he reported for slow acetylators (OR = 16.7, 95% CI = 2.2–129.1) has not been replicated. In contrast, two studies of bladder cancer among workers with documented exposure to benzidine,18,19 from a cohort of workers in China with a 25-fold increased risk of bladder cancer overall and a 158-fold excess among the highest exposed workers,20 found that NAT2 slow acetylation was protective, with a pooled estimate of OR = 0.3 (95% CI = 0.1–1.0).19 In addition, a study of workers exposed to benzidine or benzidine-based dyes found that the predominant DNA adduct in exfoliated urothelial cells was N-acetylated.21 Further, experimental studies have shown that the N-acetylation of benzidine, an aryldiamine, produces a far better substrate for subsequent oxidation to its hydroxylamine.22 This body of work suggests that in contrast to arylmonoamines, such as 4-aminobiphenyl and 2-naphthylamine, N-acetylation of aryldiamines such as benzidine is an activation rather than a detoxification step. Further complicating the situation is that benzidine is a better substrate for NAT1 than NAT2.23 Taken together, it would seem that the NAT2 slow acetylation does not increase risk for bladder cancer among workers exposed to benzidine and may in fact be protective.
Since Lower's 1979 paper, other potential modifiers of the aromatic amine and bladder cancer relationship are coming to light, as a broad approach to identifying sources of susceptibility for bladder cancer is being applied in current studies. These include evidence that water intake,24 urination frequency,25,26 and acidic urine pH (which can substantially shorten the time needed for aromatic amine glucuronides to deconjugate and become bioavailable),22,25–28 may alter risk of bladder cancer. There are no consistent findings for studies of variants in genes that activate aromatic amines (e.g. cytochrome P450 enzymes), an area of research that Lower also recommended, which is likely due to the fact that genetic variation in several key genes that oxidize arylamines may be masked by environmental factors that affect enzymatic expression and activity. In contrast, there is evidence that genetic variants in other processes relevant for aromatic amine carcinogenesis, particularly nucleotide excision repair,29–31 may impact bladder cancer risk. In addition, functional studies using assays in cultured lymphocytes, such as DNA repair capacity of damage induced by 4-aminobiphenyl,32 and the mutagen sensitivity assay,31,33 also are showing that DNA repair and other factors reflected in these assays are likely to be important in bladder cancer carcinogenesis. Finally, the ability to achieve the large sample sizes needed to explore the subtleties of interactions between variants in NAT2 and other genes and intensity and duration of exposure to tobacco and occupational and environmental carcinogens has been facilitated through the efforts of the Genetic Susceptibility to Environmental Carcinogens (GSEC) database (http://www.upci.upmc.edu/research/ccps/ccontrol/g_intro.html),8,16,34,35 and the recent formation of the International Consortium of Case-Control Studies of Bladder Cancer, which includes almost every active case-control study of bladder cancer in the world.
The foundation of the aromatic amine and NAT2 acetylation hypothesis was based on an extraordinary series of observations by epidemiologists, toxicologists, clinicians, pharmacologists and geneticists; integrated and first tested by Lower; and followed up by molecular biologists who cloned NAT2 and demonstrated the unequivocal functionality of its most important variants,7 and by molecular epidemiologists who tested Lower's hypothesis in population studies in many parts of the world. Lower's body of work, those who inspired it and those who came afterwards serve to remind us of the importance of incorporating biologically-based hypothesis-driven basic research into epidemiological studies. Furthermore, as we venture into the era of agnostic genome-wide association studies using tag markers, this history teaches us that future findings of association will need to be replicated in diverse settings, followed up by identification of biologically meaningful and relevant variants, and tested for interactions with well characterized and measured exposures. All students and practitioners of molecular epidemiology would be enriched by studying this historical saga.
 |
Supplementary Data |
|---|
 Top Supplementary Data AcknowledgementsReferences |
|---|
Supplementary data are available at IJE Online.
|
Acknowledgements |
|---|
|
TopSupplementary DataAcknowledgementsReferences |
|---|
This research was supported in part by the Intramural Research Program of the National Cancer Institute, National Institutes of Health.
Conflict of interest: None declared.
|
References |
|---|
|
TopSupplementary DataAcknowledgementsReferences |
|---|
1 Lower GM Jr, Nilsson T, Nelson CE, Wolf H, Gamsky TE, Bryan GT. N-acetyltransferase phenotype and risk in urinary bladder cancer: approaches in molecular epidemiology. Preliminary results in Sweden and Denmark. Environ Health Perspect (1979) 29:71–79. (Reprinted Int J Epidemiol 2007;36:11–18.)[Web of Science][Medline]
2 Rehn L. Blasengeschwulste bei Fuchsin-Arbeinern. Arch Klin Chirurgie (1895) 50:588–600.
3 Lower GM Jr, Bryan GT. Enzymatic N-acetylation of carcinogenic aromatic amines by liver cytosol of species displaying different organ susceptibilities. Biochem Pharmacol (1973) 22:1581–88.[CrossRef][Web of Science][Medline]
4 Hein DW. Acetylator genotype and arylamine-induced carcinogenesis. Biochim Biophys Acta (1988) 948:37–66.[Medline]
5 Schulte PA. A conceptual and historical framework for molecular epidemiology. In: Molecular epidemiology: principles and practices—Schulte PA, Perera FP, eds. (1993) New York: Academic Press. 4–44.
6 Lower GM Jr. Concepts in causality: chemically induced human urinary bladder cancer. Cancer (1982) 49:1056–66.[CrossRef][Web of Science][Medline]
7 Hein DW. N-acetyltransferase 2 genetic polymorphism: effects of carcinogen and haplotype on urinary bladder cancer risk. Oncogene (2006 Mar 13) 25:1649–58.[CrossRef][Web of Science]
8 Garcia-Closas M, Malats N, Silverman D, et al. NAT2 slow acetylation, GSTM1 null genotype, and risk of bladder cancer: results from the Spanish Bladder Cancer Study and meta-analyses. Lancet (2005) 366:649–59.[CrossRef][Web of Science][Medline]
9 Vineis P, Caporaso N, Tannenbaum SR, et al. Acetylation phenotype, carcinogen-hemoglobin adducts, and cigarette smoking. Cancer Res (1990) 50:3002–4.
10 Bartsch H, Malaveille C, Friesen M, Kadlubar FF, Vineis P. Black (air-cured) and blond (flue-cured) tobacco cancer risk. IV: molecular dosimetry studies implicate aromatic amines as bladder carcinogens. Eur J Cancer (1993) 29A:1199–207.[CrossRef]
11 Wacholder S, Chanock S, Garcia-Closas M, El Ghormli L, Rothman N. Assessing the probability that a positive report is false: an approach for molecular epidemiology studies. J Natl Cancer Inst (2004) 96:434–42.
12 Colhoun HM, McKeigue PM, Davey Smith G. Problems of reporting genetic associations with complex outcomes. Lancet (2003) 361:865–72.[CrossRef][Web of Science][Medline]
13 Lubin JH, Kogevinas M, Silverman D, et al. Evidence for an intensity-dependent interaction of NAT2 acetylation genotype and cigarette smoking in the Spanish Bladder Cancer Study. Int J Epidemiol (2007) 36:236–41.
13 Vineis P, Bartsch H, Caporaso N, et al. Genetically based N-acetyltransferase metabolic polymorphism and low-level environmental exposure to carcinogens. Nature (1994) 369:154–56.[CrossRef][Medline]
14 Salanti G, Higgins JP, White IR. Bayesian synthesis of epidemiological evidence with different combinations of exposure groups: application to a gene-gene-environment interaction. Stat Med (2006) [Epub ahead of print].
15 Vineis P, Marinelli D, Autrup H, et al. Current smoking, occupation, N-acetyltransferase-2 and bladder cancer: a pooled analysis of genotype-based studies. Cancer Epidemiol Biomarkers Prev (2001) 10:1249–52.
16 Cartwright RA, Glashan RW, Rogers HJ, et al. Role of N-acetyltransferase phenotypes in bladder carcinogenesis: a pharmacogenetic epidemiological approach to bladder cancer. Lancet (1982) 2:842–45.[Web of Science][Medline]
17 Hayes RB, Bi W, Rothman N, et al. N-acetylation phenotype and genotype and risk of bladder cancer in benzidine-exposed workers. Carcinogenesis (1993) 14:675–78.
18 Carreon T, Ruder AM, Schulte PA, et al. NAT2 slow acetylation and bladder cancer in workers exposed to benzidine. Int J Cancer (2006) 118:161–68.[CrossRef][Web of Science][Medline]
19 Bi W, Hayes RB, Feng P, et al. Mortality and incidence of bladder cancer in benzidine-exposed workers in China. Am J Ind Med (1992) 21:481–89.[Web of Science][Medline]
20 Rothman N, Bhatnagar VK, Hayes RB, et al. The impact of interindividual variation in NAT2 activity on benzidine urinary metabolites and urothelial DNA adducts in exposed workers. Proc Natl Acad Sci USA (1996) 93:5084–89.
21 Zenser TV, Lakshmi VM, Hsu FF, Davis BB. Metabolism of N-acetylbenzidine and initiation of bladder cancer. Mutat Res (2002) 506–507:29–40.
22 Zenser TV, Lakshmi VM, Rustan TD, et al. Human N-acetylation of benzidine: role of NAT1 and NAT2. Cancer Res (1996) 56:3941–47.
23 Michaud DS, Spiegelman D, Clinton SK, et al. Fluid intake and the risk of bladder cancer in men. N Engl J Med (1999) 340:1390–97.
24 Kadlubar FF, Dooley KL, Teitel CH, et al. Frequency of urination and its effects on metabolism, pharmacokinetics, blood hemoglobin adduct formation, and liver and urinary bladder DNA adduct levels in beagle dogs given the carcinogen 4-aminobiphenyl. Cancer Res (1991) 51:4371–77.
25 Silverman DT, Devesa SS, Moore LE, Rothman N. Bladder Cancer. In: Cancer Epidemiology and Prevention—Schottenfeld D, Fraumeni JF Jr, eds. (2006) 3rd edn. NY: Oxford University Press. 1101–27.
26 Kadlubar FF, Miller JA, Miller EC. Hepatic microsomal N-glucuronidation and nucleic acid binding of N-hydroxy arylamines in relation to urinary bladder carcinogenesis. Cancer Res (1977) 37:805–14.
27 Rothman N, Talaska G, Hayes RB, et al. Acidic urine pH is associated with elevated levels of free urinary benzidine and N-acetylbenzidine and urothelial cell DNA adducts in exposed workers. Cancer Epidemiol Biomarkers Prev (1997) 6:1039–42.[Abstract]
28 Garcia-Closas M, Malats N, Real FX, et al. Genetic variation in the nucleotide excision repair pathway and bladder cancer risk. Cancer Epidemiol Biomarkers Prev (2006) 15:536–42.
29 Andrew AS, Nelson HH, Kelsey KT, et al. Concordance of multiple analytical approaches demonstrates a complex relationship between DNA repair gene SNPs, smoking and bladder cancer susceptibility. Carcinogenesis (2006) 27:1030–37.
30 Wu X, Gu J, Grossman HB, et al. Bladder cancer predisposition: a multigenic approach to DNA-repair and cell-cycle-control genes. Am J Hum Genet (2006) 78:464–79.[CrossRef][Web of Science][Medline]
31 Lin J, Kadlubar FF, Spitz MR, Zhao H, Wu X. A modified host cell reactivation assay to measure DNA repair capacity for removing 4-aminobiphenyl adducts: a pilot study of bladder cancer. Cancer Epidemiol Biomarkers Prev (2005) 14:1832–36.
32 Matthew BS, Margaret RS, Barton GH, et al. Genetic instability in bladder cancer assessed by the comet assay. J Natl Cancer Inst (2003) 95:540–47.
33 Engel LS, Taioli E, Pfeiffer R, et al. Pooled analysis and meta-analysis of glutathione S-transferase M1 and bladder cancer: a HuGE review. Am J Epidemiol (2002) 156:95–109.
34 Taioli E, Raimondi S. Genetic susceptibility to bladder cancer. Lancet (2005) 366:610–12.[CrossRef][Web of Science][Medline]
35 Laird NM, Mosteller F. Some statistical methods for combining experimental results. Int J Technol Assess Health Care (1990) 6:5–30.[Medline]
36 Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics (1994) 50:1088–101.[CrossRef][Web of Science][Medline]
37 Egger M, Davey SG, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ (1997) 315:629–34.
38 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) 9:461–67.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. J. Khoury, L. Bertram, P. Boffetta, A. S. Butterworth, S. J. Chanock, S. M. Dolan, I. Fortier, M. Garcia-Closas, M. Gwinn, J. P. T. Higgins, et al. Genome-Wide Association Studies, Field Synopses, and the Development of the Knowledge Base on Genetic Variation and Human Diseases Am. J. Epidemiol., August 1, 2009; 170(3): 269 - 279. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Figueroa, N. Malats, M. Garcia-Closas, F. X. Real, D. Silverman, M. Kogevinas, S. Chanock, R. Welch, M. Dosemeci, Q. Lan, et al. Bladder cancer risk and genetic variation in AKR1C3 and other metabolizing genes Carcinogenesis, October 1, 2008; 29(10): 1955 - 1962. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||