|Home | About | Journals | Submit | Contact Us | Français|
Colorectal cancer literature regarding the interaction between polymorphisms in carcinogen-metabolizing enzymes and red meat intake/doneness is inconsistent. A case-control study was conducted to evaluate the interaction between red meat consumption, doneness and polymorphisms in carcinogen-metabolizing enzymes. Colorectal cancer cases diagnosed 1997-2000, aged 20-74 years, were identified through the population-based Ontario Cancer Registry and recruited by the Ontario Family Colorectal Cancer Registry (OFCCR). Controls were sex- and age-group matched random sample of Ontario population. Epidemiologic and food questionnaires were completed by 1095 cases and 1890 controls; blood was provided by 842 and 1251, respectively. Multivariate logistic regression was used to obtain adjusted odds ratio (OR) estimates. Increased red meat intake was associated with increased colorectal cancer risk [OR (>5 servings/week vs. ≤2 servings/week) =1.67 (1.36, 2.05)]. Colorectal cancer risk also increased significantly with well-done meat intake [OR (>2 servings/week well-done vs. ≤ 2 servings/week rare-regular) = 1.57 (1.27, 1.93)]. We evaluated interactions between genetic variants in 15 enzymes involved in the metabolism of carcinogens in overcooked meat (CYPs, GSTs, UGTs, SULT, NATs, mEH, AHR). CYP2C9 and NAT2 variants were associated with colorectal cancer risk. Red meat intake was associated with increased colorectal cancer risk, regardless of genotypes; however, CYP1B1 combined variant and SULT1A1-638G>A variant significantly modified the association between red meat doneness intake and colorectal cancer risk. In conclusion, well-done red meat intake was associated with an increased risk of colorectal cancer regardless of carcinogen-metabolizing genotype, although our data suggests persons with CYP1B1 and SULT1A1 variants had the highest colorectal cancer risk.
Colorectal cancer is the third most common cancer in Canada and has a poor 5-year survival rate of 60% (1). Both genetic and environmental factors are involved, with 20% of colorectal cancer patients exhibiting a familial component in which relatives have a doubling of risk (2). Since less than 5% of colorectal cancer is explained by identified genetic syndromes (3-4) common inherited polymorphisms of low penetrance are likely important.
Many case-control and prospective epidemiologic studies, as well as recent meta-analyses of cohort studies, report that red meat consumption is associated with an increased risk of colorectal cancer and it has been suggested that this may be due to carcinogenic polycyclic aromatic hydrocarbons (PAH) and heterocyclic amines (HCA) produced when meat is cooked at high temperatures (e.g., 5-16). Furthermore, among the few studies to include assessment of cooking methods, the association with colorectal cancer was found to be strongest for well-done red meat (7,13,17-19).
It is well documented that genes that encode enzymes involved in metabolism/activation of carcinogens such as PAH and HCA found in overcooked meat (e.g., CYP1A1/1A2/1B1,2E1, GSTs, UGTs, mEH) exhibit variation (e.g., 20-26). In general, bioactivation of pre-carcinogens is carried out by Phase I enzymes such as cytochromes P450 (CYP) (62-63) whereas Phase II enzymes such as glutathione S-transferases (GST) and UDP-glucuronosyltransferases (UGT) usually “detoxify” reactive metabolites by conjugation and thus prevent metabolites from binding to DNA (23-24). Genetic variants may alter enzyme function and carcinogen metabolism and consequently modify the association between well-done meat intake and increased colorectal cancer risk (27-31).
Colorectal cancer literature regarding the interaction between carcinogen-metabolizing genes and red meat intake/doneness is limited and inconsistent, although suggests certain genetic variants may modify the association between red meat and colorectal cancer risk (18-19,29,32-38). To summarize the main studies to date, CYP2E1, GSTT1 and SULT1A1 were shown to significantly modify the association between red meat intake and colorectal cancer, while interaction findings for NAT2 vary between studies and CYP1A1, GSTM1, UGT1A7 and mEH did not modify risk (18,29,32-37,39). Two studies observed multiple-way interactions with red meat intake, such as the combination of CYP1A2 and NAT2 (19) and the combination of several variants in CYP genes (1A2, 2E1, 1B1, 2C9) (38). There is a paucity of data on the interaction between many of the genetic variants in carcinogen-metabolizing enzymes and red meat intake and doneness as regards colorectal cancer risk, and many previous studies had very limited power to detect statistically significant gene-environment interactions.
We evaluated the association between red meat intake, doneness and genetic variants in key carcinogen-metabolizing enzymes and colorectal cancer risk among over 2,000 cases and controls participating in the population-based Ontario Family Colorectal Cancer Registry (OFCCR). Understanding the interaction between modifiable risk factors and genetic susceptibility may aid development of more tailored colorectal cancer primary prevention strategies.
The OFCCR is one of six international sites participating in the Cooperative Family Registry for Colorectal Studies (CFR-Colon) established by the US National Cancer Institute (40). The methods of the OFCCR have been described previously (41-43) and are reiterated below. Colorectal cancer cases and population controls participating in the OFCCR were used to conduct this study.
The population-based Ontario Cancer Registry (OCR) was used to identify and recruit into the OFCCR, living, incident colorectal cancer cases (pathology confirmed; International Classification of Diseases – 9th revision codes 153.0-153.9, 154.1-154.3, 154.8) (44) aged 20-74 and diagnosed between July 1, 1997 and June 30, 2000. The OCR registers all cases of cancer diagnosed among residents of Ontario using computerized probabilistic record linkage to resolve the four main sources of cancer information (pathology reports, hospital discharge summaries, reports from Ontario's regional cancer centres, and death certificates).
Controls recruited into the OFCCR were a random sample of Ontario residents identified using two methods. Population-based controls were randomly selected and frequency-matched, within sex and 5-year age groups, to the colorectal cancer cases. In 1999-2000, persons were identified using a list of residential telephone numbers in Ontario provided by Infodirect (Bell Canada). Households were randomly selected from this list, and telephoned to obtain a census of household members (age, sex). One eligible person within each household was randomly selected and invited to participate. To increase the sample size and approach a 1:2 case:control ratio, additional population-based controls were recruited in 2001. An age- and sex-stratified random sample of persons was selected from a listing of all Ontario residents (homeowners and occupants) based on assessment rolls maintained by the provincial government. A re-abstraction study was able to link more than 95% of persons in the OCR to this population database, suggesting that its' accuracy and completeness are high (45).
Physicians, identified from pathology reports, were asked to permit contact with their patient(s), and to provide the patient address, telephone number and vital status. Over 90% of physicians consented. Once a physician provided consent, patients were mailed a package containing a letter, a family history questionnaire (FHQ), and a brochure describing the various phases of the OFCCR (epidemiologic and food questionnaires, provide blood sample and enrol kin). A reminder post-card was sent several weeks after this mailing and non-responders were followed up with a telephone call several months after the initial mailing. Of the 6695 colorectal cancer cases mailed an FHQ, 3781 (57%) returned this questionnaire (participated). The median lag between diagnosis date (pathology report) and mailing of case invitation package was approximately six months (for all cases including non-responders and refusals).
Following the completion and return of the FHQ (phase one), pedigrees were constructed. Each colorectal cancer case was then classified as 1) high familial risk (satisfying hereditary nonpolyposis colorectal cancer (HNPCC) Amsterdam criteria) (46), 2) intermediate familial risk, or 3) low (sporadic) risk. Intermediate familial risk has a very broad definition and consists of cases satisfying at least one of the following: a) two relatives with HNPCC cancers (this includes 14 cancer sites), and two (of three) are first degree kin, b) case and relative both with colorectal cancer < 50 years of age, and c) any relative with colorectal cancer < 35 years of age. All other cases not classified as high or intermediate familial risk were classified as sporadic (with the exception of a few cases categorized as intermediate due to selected “pathology criteria” such as multiple polyps (described in Cotterchio et al., (41)). All high and intermediate risk cases and a 25% random sample of the low risk cases were selected to participate in phase two of the OFCCR. These participants were then asked to i) complete the self-administered epidemiologic questionnaire designed by the CFR-Colon and food frequency questionnaire (FFQ), ii) provide a blood sample, and iii) provide permission to contact their relatives.
Controls were mailed a cover letter along with the family history, epidemiologic, and food frequency questionnaires, and were also asked to provide a blood sample.
Dietary intake was determined using both the CFR-Colon epidemiologic questionnaire and an FFQ. The CFR-Colon epidemiology questionnaire contained several questions regarding dietary intake two years ago including questions about red meat intake and cooking methods. For example, “how often a fruit/vegetable serving was eaten”, “how often a red meat serving (2-3 ounces) was eaten”, “how often a red meat serving cooked by broiling/grilling/barbequing/pan-frying was eaten” and “how well-done was red meat (inside and outside appearance)”. The FFQ adopted by the CFR-Colon asked about foods “eaten about two years ago” and was analyzed using food composition databases that include values for macro- and micro-nutrients (47); this information was used to derive several potential confounding dietary variables.
Two variables were derived to describe red meat consumption: red meat servings per week and red meat doneness intake (reported in the epidemiologic questionnaire). “Red meat servings per week” was defined as the reported number of servings of red meat consumed per week and included beef, pork, veal, lamb and venison. “Red meat doneness consumption” was defined by combining the number of servings of red meat per week that were cooked by broiling, grilling, barbecuing or pan-frying and the degree to which the meat was cooked. Rare was defined as red/pink inside and light/medium brown outside; well-done was defined as brown inside or heavily browned/blackened outside. Four mutually exclusive “red meat doneness consumption” categories were derived: ≤ 2 servings/week of rare-regular red meat, ≤ 2 servings/week of well-done red meat, >2 servings/week of rare-regular red meat, >2 servings/week of well-done red meat.
The 32-page CFR-Colon epidemiologic questionnaire also included many close-ended questions about colorectal cancer screening, medical conditions, medication use, reproductive factors, physical activity, alcohol consumption, smoking history, sociodemographic, and anthropometric measures.
The 1536 incident colorectal cancer cases selected to participate in phase two of the OFCCR were mailed epidemiologic and food questionnaires and asked to provide a blood sample. 1095 (72%) cases completed the questionnaires (epidemiology questionnaire queried meat intake and cooking methods) and are included in the meat intake data analysis. Eighty-three subjects were excluded due to extreme caloric intakes (females: < 700 or > 4200 kcal and males: < 800 or > 4900 kcal). The OFCCR classified cases based on their familial cancer history: 42 (4%) were high (HNPCC) risk, 483 (44%) were intermediate familial risk (defined above), and 570 (52%) were low risk. Of these 1095 colorectal cancer cases, 842 (77%) provided blood (DNA) and thus were available for genotype analyses and evaluation of the possible interaction between meat intake and genetic variants.
Of the 4876 eligible controls identified and invited to participate, 2131 refused (43%), and of 2745 mailed the questionnaire package, 1928 (70%) completed the FFQ, and 1944 (71%) completed the epidemiology questionnaire. 1890 controls are included in the red meat intake data analysis (i.e., completed both questionnaires and had reasonable caloric intake). Of the 1890 controls who completed both questionnaires, 1251 (66%) provided a blood (DNA) sample, and these persons comprised the control dataset used for genotype analyses and evaluation of the possible interaction between meat intake and genetic variants.
Reasons for non-participation included language barrier, illness, too busy, and questionnaire too long; however, the majority of cases and controls did not provide a reason. 95% of participants in the OFCCR are Caucasian.
The OFCCR obtained 40 ml of blood from participating cases and controls. DNA was extracted from lymphocytes using organic solvents or spin columns (Qiagen Inc.) and banked at 4°C.
A priori, genetic variants were chosen for investigation based on an estimated minor allele frequency of ≥ 5% with preference given to polymorphisms with a potential impact on function. Genotyping assays are standard assays from the literature. Assays to query the single nucleotide polymorphism (SNP) of interest were performed using the TaqMan 5′ nuclease allele discrimination assay (Applied Biosystems). In general an allele-specific oligonucleotide probe, labeled with a fluorescent reporter and quencher dye, is cleaved during the amplification process generating an increased intensity of fluorescence directly related to the accumulation of polymerase chain reaction (PCR) product. The reaction mix consisted of 5μl Taqman Universal Master Mix – no UNG (Applied Biosystems), combined primer and probe mix (per manufacturer's instructions), 20-50 ng of DNA template and water for a total reaction volume of 10μl. Cycling conditions for the reaction were 95 C for 10 minutes, followed by 40-45 cycles of 94 C for 15 seconds and 60 C for 1 minute. Following PCR amplification, end-point fluorescence was read using an ABI 7900HT Sequence Detection System and genotypes were assigned using Allelic Discrimination Software (Applied Biosystems SDS Software v2.1). Appropriate controls representative of each genotype and multiple template controls were included in each analysis. Microsatellite fragment analysis was used to genotype UGT1A1*28. Briefly, PCR was performed on 50 ng of DNA in buffer [100 mM Tris-HCl (pH 8.0), 500 mM KCl, 1.5 mM MgCl2, 0.2 mM dNTP, 0.2 μM of each primer and 1 unit of Taq Polymerase (Applied Biosystems)]. Cycling conditions were: initial denaturation at 95°C for 2 minutes followed by 30 cycles of denaturation at 94°C for 30 seconds, annealing at 55°C for 30 seconds and extension at 72°C for 45 seconds, with a 15 minute final extension at 72°C. Microsatellite fragment analysis was performed using the ABI 3730 DNA Analyzer and Genemapper 3.5 software (Applied Biosystems). A multiplex PCR for the simultaneous analysis of GSTM1-null and GSTT1-null was done using albumin as an internal positive control.
Approximately 5% of the samples were randomly selected for blinded duplicate analysis (verification), with an estimated genotyping error rate of 0.25%.
Associations between red meat intake and red meat doneness variables and colorectal cancer risk were examined by computing age-adjusted odds ratio (AOR) estimates and approximate 95% confidence intervals (CI) (48). Multivariate logistic regression analyses were performed to obtain multivariate odds ratio (OR) estimates while simultaneously controlling for potential confounders (49). For both the red meat and meat doneness analyses, potential confounding variables were evaluated based on the 10% change-in-OR estimate methods (50). Potential confounders included: family history of colorectal cancer, diagnosis of inflammatory bowel disease, body mass index (BMI), colonic screening, non-steroidal anti-inflammatory drug use, smoking history, reproductive factors, lifetime physical activity, vegetable intake, fruit intake, folate, calcium, dietary fibre, saturated fat, total energy (calories), and alcohol consumption. No variables were identified as confounders in our dataset, and thus the final multivariate models were the most parsimonious and contained only sex and age group. All statistical analysis was done using SAS v8.2 (SAS Institute, Cary, North Carolina, USA).
The possibility of interactions between selected genetic variants (CYP1A2 163C>A, rs762551; CYP2E1 1293G>C, rs3813867; CYP2E1 7632T>A, rs6413432; CYP2C9 430C>T, rs1799853; CYP2C9 1075A>C, rs1057910; CYP1A1 2455A>G, rs1048943; CYP1A1 3801T>C, rs4646903; CYP1B1 142C>G, rs10012; CYP1B1 4326C>G, rs1056836; CYP1B1 4390A>G, rs1800440; GSTM1-locus deletion; GSTT1-locus deletion; GSTM3delAGG, rs1799735; UGT1A7 W208R, rs11692021; UGT1A7 N129K, rs17868323; UGT1A1*28 A(TA)6TAA > A(TA)7TAA; mEH 17673T>C, rs1051740; SULT1A1 638G>A, rs9282861; NAT1 459G>A, rs4986990; NAT1 1088T>A, rs8190861; NAT2 341T>C, rs1801280; NAT2 590G>A, rs1799930; NAT2 857G>A, rs1799931; ARH 1661G>A, rs2066853) and red meat intake/doneness was initially evaluated by stratified analyses (50). In addition, several composite genotype variables were derived based on the known phenotype of a variety of SNP combinations within the same gene (e.g., CYP1B1 – wild type and increased activity) – these definitions are footnoted in Table 2. Interaction was formally assessed by the statistical significance of the Likelihood Ratio Statistic (LRS P<0.05) after the addition of the product term to the model (51).
Ethics approval was granted from the Office of Research Services, University of Toronto, the Mount Sinai Hospital Research Ethics Board (REB), and the Hospital for Sick Children REB.
Table 1 shows the distribution and AOR estimates among colorectal cancer cases and controls for red meat intake and red meat doneness intake. Increased consumption of red meat was associated with increased colorectal cancer risk [OR (>5 servings/week vs. < 2 servings/week) =1.67 (1.36, 2.05]. Increased consumption of well-done red meat was associated with an increased colorectal cancer risk [OR (>2 servings/week of well-done red meat vs. < 2 servings/week of rare-regular red meat) = 1.57 (1.27, 1.93)].
Table 2 shows the distribution of 29 genetic polymorphisms in 15 selected genes involved in carcinogen metabolism. For the majority of variants there was no association with colorectal cancer risk although a statistically significant association was observed for variants in CYP2C9 430C>T, and NAT2 combined variants (fast acetylator). Also, CYP1A1 3801T>C was associated with colorectal cancer risk, though this was of borderline statistical significance (possibly due to very small number of carriers) and CYP1B1 4390A>G was also of borderline significance.
The same 29 genetic variants encoding carcinogen-metabolizing enzymes were assessed for interaction with red meat intake and red meat doneness (genotypes are listed in the Methods section and Table 2). Increasing red meat intake was associated with increasing colorectal cancer risk, regardless of genotype stratification – thus, no effect modification was observed (data not shown). However, as regards red meat doneness, several genetic variants modified the association between intake and colorectal cancer risk, and these data are presented in Table 3. Specifically, CYP1B1-combined variant and SULT1A1-638G>A were found to be statistically significant effect modifiers of the association between red meat doneness intake and colorectal cancer risk, and the interaction between CYP1B1-4326C>G and red meat doneness was of borderline significance. Colorectal cancer risk increased to four-fold among persons in the highest red meat doneness category (> 2 servings of well-done red meat per week) who also carried the combined CYP1B1 wildtype variants (i.e., not increased activity) (g*e interaction p value: 0.04). Carriers of the SULT1A1-638 GG genotype who consumed > 2 well done meat servings/week had a greater than doubling of colorectal cancer risk and in comparison, a borderline statistically significant association of lower magnitude was observed with this same meat intake level among persons with AA/GA genotypes (g*e interaction p value: 0.03). In addition, AHR-1661 was a borderline significant effect modifier (LRS interaction P=0.07, data not shown). The majority of genetic variants are not presented in Table 3 because no effect modification was observed by eyeballing the data and furthermore no statistically significant interaction was apparent (i.e., increasing red meat doneness intake was associated with increased colorectal cancer risk regardless of genetic variant, and p-value > 0.10 for interaction). As the sample size was limited, some gene-meat interactions may not have been detected in our dataset and it is also possible that spurious associations were found due to the many comparisons made.
The association between red meat intake/doneness and cancer risk was evaluated separately for microsatellite instability MSI-high and MSI-low/stable colorectal cancer cases. We did not observe a statistically significant difference; red meat intake was associated with a statistically significant increased risk of colorectal cancer for both these types of tumours (as compared with controls) – data not shown.
As a disproportionate number of cases in the OFCCR were from families classified as high/intermediate risk (vs. low risk), effect modification by familial risk was also assessed. No differences across familial risk strata were identified with respect to the association between red meat intake/doneness and colorectal cancer risk (data not shown).
We report that increased consumption of both red meat and well-done red meat were significantly associated with increased colorectal cancer risk. Our findings are consistent with several meta-analyses (primarily of cohort studies) which concluded that red meat consumption is associated with an increased risk of colorectal cancer (5-6,11), although to our knowledge we are the first to evaluate this comprehensively among Canadians. We also evaluated interactions between red meat intake/doneness and polymorphic genes that encode enzymes involved in the metabolism of carcinogens found in well-done meat (CYPs, GSTs, UGTs, SULT, NATs, mEH, AHR). Our study assessed many genetic variants in 15 enzymes central to the metabolism of PAHs and HCAs produced by overcooking red meat. Polymorphisms that lead to a known change in function of these enzymes or that were previously shown to be associated with risk of colon or other cancers were studied; however, rare variants were excluded a priori because statistical power was not sufficient. Of the many genetic polymorphisms assessed, two were found to be significantly associated with colorectal cancer risk - CYP2C9-430C>T and NAT2 fast/slow variant. Upon evaluation of possible effect modification, red meat intake was found to be associated with colorectal cancer risk, regardless of genotype (no effect modification observed); however, two genetic variants (CYP1B1-combined variant and SULT1A1-638G>A) significantly modified the association between red meat doneness intake and colorectal cancer risk. Further investigation of these possible interactions is warranted.
Our findings suggest that CYP1B1 and SULT1A1 variants may modify the association between well-done red meat intake and colorectal cancer risk. The positive association with colorectal cancer increased to four-fold among persons in the highest red meat doneness category (> 2 servings of well-done red meat per week) who also carried the combined CYP1B1 wildtype variants (i.e., not increased activity). We are the first study to specifically assess several CYP1B1 variants, well-done red meat intake and colorectal cancer risk. Our finding is plausible since CYP1B1 bioactivates carcinogens such as PAHs found in burnt meat, and polymorphisms in the essential exon-3 heme-binding region alters this activity (52-54). Furthermore, CYP1B1 is highly expressed in the colon and in colon cancers, and thus is available to interact with carcinogens within the colon itself (55-56). We found that carriers of the SULT1A1-638 GG genotype who consumed > 2 well done meat servings/week had a greater than doubling of colorectal cancer risk while no statistically significant association was observed with this same meat intake level among persons with AA/GA genotypes. SULT1A1 is involved in Phase II metabolism and is also expressed in numerous tissues including the colon (57). Somewhat supporting our finding, a recent German study reported modification of the red meat - colorectal cancer risk association by SULT1A1 genotype although meat doneness was not assessed (37). Our findings could also be due to chance or bias, especially since many genetic variants were assessed and the study response rate was less than optimal. Thus, replication by future studies is essential to further investigate the hypothesis that genetic variants in carcinogen-metabolizing enzymes may modify colorectal cancer risk associated with well-done red meat intake.
Consistent with our findings, many epidemiologic studies and several meta-analyses report that red meat consumption is associated with an increased risk of colorectal cancer (e.g., 5-16). Few studies have evaluated red meat doneness and colorectal cancer risk, although, consistent with our findings most studies found the association with colorectal cancer was strongest for well-done red meat (7,13,17-18). To our knowledge only one Canadian study has assessed red meat intake and colon cancer risk; however, the doneness of meat was not considered (16). Consistent with our findings, they observed an association between red meat and colon cancer, in particular proximal colon cancer risk (16). It is important to conduct Canadian studies since nutrient values (such as fat and protein involved in PAH /HCA production) for Canadian and American beef differ because production methods are not the same [Beef Information Centre: http://www.beefinfo.org/nutrient_data.cfm#faq11] (69).
Several studies have evaluated some genetic variants, red meat intake and colorectal cancer risk, though data are sparse or non-existent for certain carcinogen-metabolizing genetic variants. To summarize, CYP2E1, GSTT1 and SULT1A1 significantly modified the association between red meat intake and colorectal cancer risk, while CYP1A1, GSTM1, UGT1A7 and mEH did not modify risk and NAT2 findings varied between studies (18,29,32-37,39). A case-control colorectal cancer study conducted in Utah and California found no interaction between red meat intake, well-done red meat consumption and CYP1A1 genotype, nor was the association between colorectal cancer risk and red meat consumption modified by the combination of CYP1A1 and GSTM1 genotypes (32). A separate publication by these authors reported that the association between rectal cancer risk and red meat consumption/doneness was not significantly modified by NAT2 phenotype or GSTM1 genotype (18). Another American case-control study reported little to no association between many red meat intake variables and colon cancer risk, though the NAT2 variant slightly modified these associations while the GSTM1 variant had no impact (33). A British colorectal cancer case-control study reported some evidence of an interaction between GSTT1 and red meat intake; however, no interaction was observed for mEH, CYP1A1 or GSTM1 (35). A case-control study in Hawaii assessed well-done red meat intake, genetic variants and colorectal cancer risk and observed that the largest statistically significant association was seen for the three-way interaction between well-done red meat, rapid CYP1A2 phenotype and rapid NAT2 genotype (19). A subsequent Hawaiian study reported that CYP2E1 (increased activity) may modify the rectal cancer risk associated with red meat intake (29). The Nurses' Cohort Study reported an interaction between NAT2 (fast/slow acetylator) and red meat intake as regards colorectal cancer risk (of borderline statistical significance) (34). A recent German colorectal cancer case-control study reported a moderate (though not statistically significant) interaction between NAT1/NAT2 combined genotype and red meat intake (36). A recent case-control study conducted in France reported that the combination of several variants in CYP genes (1A2, 2E1, 1B1, 2C9) modified (exacerbated) the association between red meat intake and colorectal cancer risk (38). Two small case-control studies recently evaluated SULT variants, meat intake and colorectal cancer (37,58). Lilla et al (37) reported modification of the red meat-colorectal cancer risk by SULT1A1 genotype while the other study reported no effect modification (58), however the latter study was underpowered to detect an interaction with less than 300 cases participating.
While most previous studies that assessed effect modification of the red meat colorectal cancer risk association included only a limited number of genetic variants and were limited by small sample sizes, we evaluated 29 genetic polymorphisms in 15 selected genes known to be central to the metabolism/bioactivation of carcinogens among nearly 900 cases and 1200 controls. We are the first large study to investigate whether mEH modifies the association between red meat and colorectal cancer, with only one small cancer study previously published on this topic (35). Similar to our findings, this study reported that mEH does not modify the red meat colorectal cancer association. We are the first to investigate red meat doneness, SULT variants and colorectal cancer risk.
HCAs are formed during the pyrolysis of proteins in meat, and the quantity depends on cooking temperature and duration, while PAHs are produced from the pyrolysis of fat (15,59). Most chemical carcinogens require metabolic bioactivation in order to bind to DNA and form DNA adducts that exert a carcinogenic effect (60-61). Bioactivation of pre-carcinogens is usually carried out by Phase I enzymes such as CYPs (62-63) whereas Phase II enzymes such as GST and UGT usually “detoxify” reactive metabolites by conjugation and thus prevent metabolites from binding to DNA (23-24,64). Enzymes such as CYP1A2 and CYP1A1 are important in bioactivation of PAHs and HCAs involved in carcinogenesis (63,65-66). Factors that alter the level or activity of these enzymes may influence the body's response to carcinogens (30-31). For example, individuals with a rapid CYP1A2 phenotype who also excreted high levels of PhIP (an HCA), had the lowest levels of PhIP DNA adducts in their colon (67).
Survival bias is a possible limitation of our study since fatal cases were excluded and thus cases with better survival are over-represented. In addition, the lag between diagnosis and recruitment into the OFCCR may have created a possible survivor bias since the survival rate for colorectal cancer is moderate, though varies greatly by stage at diagnosis. However, it is reassuring that participation in the OFCCR was not statistically different for early versus late (metastatic) stage colorectal cancer cases (41). Although it has been reported that most colon cancer risk factors do not differ by stage of disease (68), survival bias may be a concern if red meat intake affects survival. Although our response rate was not optimal, both cases and controls were selected from population-based sampling frames and many established risk factors were found to be associated with colorectal cancer risk in our dataset suggesting the cases and controls are representative (42). Response bias is always a possible limitation when response rates are not optimal; however, it is unlikely that non-response would be associated with inherited carcinogen-metabolizing genotypes. In an attempt to assess possible response bias as regards sociodemographic factors, we previously published that the age and sex distribution of colorectal cancer cases participating in the OFCCR did not differ from non-participating cases; however, colorectal cancer cases in rural areas were slightly more likely to participate (41). Possible confounding by many colorectal cancer risk factors was evaluated, and adjusted for, in our analyses. Although potential confounders were evaluated, residual or unknown confounding always remains a possibility. Case-control studies are susceptible to recall bias because cases may report exposures differently than controls. Although we could not directly measure HCA and PAH, the CFR-Colon epidemiologic questionnaire asked not only about red meat consumption but also the doneness of red meat eaten. As our sample size was moderate, it is possible that some gene-environment interactions were not detected. It is also plausible that the observed interactions are spurious since many comparisons were made. Although multiple comparisons were made, this study was conducted with specific a priori hypotheses based on a candidate gene pathway approach that focused on enzymes involved in carcinogen metabolism/activation and certain genetic variants likely to be functional. Lastly, the incomprehensive gene coverage due to the small number of variants selected per gene is a limitation of this study which used the candidate gene approach to investigate genetic interactions.
This study adds to the growing body of evidence that suggests consumption of red meat (especially well-done meat) increases the risk of colorectal cancer, with this being the first Canadian study to evaluate well-done red meat intake and colorectal cancer risk. In general, the increased colorectal cancer risk among consumers of red meat was observed regardless of carcinogen-metabolizing genotype, although our data suggests consumers of well-done red meat who carry CYP1B1 and SULT1A1 variants may exhibit higher colorectal cancer risk. Future studies are needed with greater power to simultaneously examine combinations of relevant genetic polymorphisms, red meat intake, doneness and colorectal cancer risk.
We would like to thank the Ontario Family Colorectal Cancer Registry (OFCCR) staff, in particular Darshana Daftary (Cancer Care Ontario), the study co-ordinator, and Teresa Selander (Mount Sinai Hospital), the biospecimen manager, for their dedication to this study. We would also like to thank Zhanquin Liu (The Centre for Applied Genomics, Hospital for Sick Children) and Hui Zhang (Hospital for Sick Children) for their invaluable assistance in the laboratory. We thank Nancy Deming for her assistance with manuscript preparation. The content of this manuscript does not necessarily reflect the views or policies of the NCI, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government or the Consortium of Familial Registries.
Financial Support – Supported by the National Cancer Institute of Canada (NCIC) with funds from the Canadian Cancer Society (grant no. 013208) and the National Cancer Institute, National Institutes of Health, USA (RFA # CA-95-011, grant no. U01-CA74783).