Abstract
The mechanisms underlying the association of menopausal hormone therapy (MHT) with reduced colorectal cancer (CRC) risk are unknown and the identification of genetic modifiers may yield further insight. We explored the effect modification of MHT-associated CRC risk in postmenopausal women by 47 polymorphisms with known or putative functional relevance in 16 candidate genes related to hormone metabolism (COMT, CYP1A1, CYP1A2, CYP1B1, CYP2C9, CYP2C19, CYP3A4, CYP17A1, GSTP, and HSD17B1), transport (ABCB1), and signaling (ESR1, ESR2, SHBG, PGR, and NR1I2). A total of 685 CRC patients and 684 healthy controls from a German population-based case–control study (DACHS) were genotyped. Multiplicative statistical interaction between polymorphisms and ever MHT use as well as duration of use was assessed using multivariate logistic regression. CRC risk associated with ever MHT use as well as with duration was significantly modified by rs1202168 in the transporter gene ABCB1 (P interaction=0.04). The MHT-associated risk reduction was not significant in homozygous non-carriers (odds ratio (OR) ever use=0.84, 95% confidence interval (CI) 0.53–1.34; OR per 5 year duration=0.94, 95% CI 0.83–1.08), while homozygous carriers of the minor T allele had a 57% lower risk with ever use of MHT (95% CI 0.21–0.88) and a 22% lower risk per 5 years of MHT use (95% CI 0.62–0.97). Significant effect modification was also observed for the ESR1_rs910416 polymorphism (P interaction=0.03 for ever use and 0.07 for duration of use), whereby the decreased risk was attenuated in homozygous carriers of the minor C allele (OR ever use=0.87, 95% CI 0.48–1.60, OR per 5 year duration=0.99, 95% CI 0.83–1.18). Results of this exploratory study provide first evidence that polymorphisms in genes related to estrogen transport and signaling may modify MHT-associated CRC risk but warrant replication in an independent population.
Introduction
An increasing body of evidence points to the involvement of sex steroids in development and progression of colorectal cancer (CRC; Kennelly et al. 2008, Koo & Leong 2010). Inverse associations between CRC risk and menopausal hormone therapy (MHT) observed in both, clinical and observational studies support this hypothesis (Chlebowski et al. 2004, La Vecchia et al. 2005). A combined analysis of two randomized clinical trials showed a relative risk for CRC of 0.64 (95% confidence interval (CI) 0.45–0.92) for women assigned to combined estrogen–progestagen therapy (Beral et al. 2002). The reduction of risk for developing CRC associated with MHT use has been found in case–control and cohort studies to be in the range of 20–40% (La Vecchia et al. 2005, Hoffmeister et al. 2009). Nevertheless, inconsistent findings also exist. Some studies reported a reduced risk only with a certain type of therapy, e.g. with estrogen monotherapy (Hildebrand et al. 2009), or no association with MHT at all (Kabat et al. 2008, Tsilidis et al. 2010).
How MHT exerts its protective effect on CRC risk is largely unknown. The effect of estrogen on tumor development may be mediated through its receptors (Slattery et al. 2005, Kennelly et al. 2008). Both estrogen receptor α (ERα) and ERβ can be detected in colon tissue, while ERβ is found to be more prevalent than ERα (Foley et al. 2000, Konstantinopoulos et al. 2003, Jassam et al. 2005). Loss of ERβ has been related to advanced tumor progression (Kennelly et al. 2008) and recent cell line studies support the assumption that estrogen acts via ERβ as an inhibitor of malignant cell growth (Hartman et al. 2009, Wilkins et al. 2010).
Action of sex steroids on tissues and cells depends on concentration levels of hormones. Xenobiotic compounds as well as steroid hormones are metabolized through oxidative, reductive, and hydrolytic reactions catalyzed by phase-I enzymes (Nebert & Russell 2002). This process is often followed by a conjugation to charged species such as glutathione, which involves phase-II enzymes and leads to more soluble metabolites (Roediger & Babidge 1997). Therefore the function of enzymes of the phase-I and phase-II biotransformation system has an impact on concentrations of sex steroids. Binding and transporter proteins further influence the availability of steroid hormones on the cellular level. The bioavailable fraction of estrogens and androgens is regulated by the sex hormone-binding globulin (Xita & Tsatsoulis 2010). On the cellular level, P-glycoprotein might be involved, an efflux transporter protein expressed in colon tissue (Berggren et al. 2007).
To give further insight into the biological mechanisms underlying the role of steroid hormones in colon carcinogenesis, we investigated the interaction between MHT and predominantly functional single nucleotide polymorphisms (SNPs) in 16 candidate genes. The genes under investigation are involved in sex steroid binding and signaling (ESR1, ESR2, SHBG, PGR, and NR1I2), metabolism (COMT, CYP1A1, CYP1A2, CYP1B1, CYP2C9, CYP2C19, CYP3A4, CYP17A1, GSTP, and HSD17B1), and transport (ABCB1). A total of 47 SNPs were evaluated in postmenopausal female participants of a German case–control study on CRC.
Materials and methods
Study subjects
The DACHS study is a German population-based case–control study on CRC and has been described in detail elsewhere (Brenner et al. 2006, Lilla et al. 2006). In brief, cases were patients with a first diagnosis of CRC receiving in-patient treatment and were recruited in hospitals of the Rhein-Neckar-Odenwald region. Controls were randomly selected from population registry lists and frequency matched according to sex, 5-year age groups, and county of residence. To be eligible, participants had to be at least 30 years old, resident in the study region, and able to complete the interview. From January 01, 2003 until December 31, 2007, 1945 (811 women and 1134 men) cases and 2399 (1013 women and 1386 men) controls were recruited (Brenner et al. 2011). Two hundred thirty-nine female controls completed only a short questionnaire and were thus not considered in the current study.
The 1585 (811 cases and 774 controls) fully participating women were asked to provide a blood and a mouthwash sample. Overall, 746 (92.0%) female cases and 732 (94.6%) female controls donated a biological sample and were genotyped. Genomic DNA was extracted from peripheral blood mononuclear cells using FlexiGene Kit 250 (Qiagen). For 19 study subjects, blood samples were not available and DNA was obtained from mouthwash samples using Qiagen Mini Kit (Qiagen). In order to have a homogenous study population, premenopausal women and women aged <55 years with unclear menopausal status (N=109) were not included in the study population. The final study population consisted of 685 postmenopausal female cases and 684 corresponding postmenopausal controls. Information was collected through in-person interviews by trained interviewers using a standardized questionnaire including demographic information as well as information on former colorectal endoscopies, reproductive and life style factors, anthropometry, and medical and family history.
Women recruited between January 2003 and October 2004 provided basic information about MHT, including start, end, and duration of use (Hoffmeister et al. 2007). Between November 2004 and December 2007, detailed information for each period of MHT was collected during the interview. For women entering the study until December 31, 2006, we sought to validate self-reports on MHT by sending a special questionnaire to the physicians of those women. The process of MHT validation in the DACHS study is described in detail elsewhere (Hoffmeister et al. 2009). Complete validation of all self-reported periods of MHT use was attained for 60% of women who reported MHT use, and incomplete validation was attained for additional 10%. For 49 women who entered the study after December 31, 2006, self-reported MHT use has not been validated. The median date of recruitment was January 11, 2005 for cases and August 23, 2005 for controls. Therefore, it is unlikely that changes in MHT prescription over time affected cases and controls differently.
The present study was approved by the Ethics Committees of the State Medical Boards of Baden-Württemberg and Rhineland-Palatinate (Germany) and the University of Heidelberg (Germany). Every participant gave written informed consent.
Variable definitions
Variables were defined according to the reported history at the reference date, which was the date of the interview for controls and the date of CRC diagnosis for cases. Women were defined as postmenopausal if they reported that menstrual bleedings had stopped naturally, after bilateral oophorectomy, radiation therapy, or chemotherapy, or if they were older than 55 years. Menopausal status can be masked if a hysterectomy was the cause for the end of the menstrual cycle or if MHT was started before natural menopause. Therefore, we also classified women aged ≤55 years as postmenopausal if their menstrual bleedings had stopped and they additionally used MHT for more than four years (the duration of perimenopause for most women (Dudley et al. 1998)).
We considered women to be ‘ever users’ if they reported ≥3 months of total MHT duration. Participants who used MHT <3 months were considered as ‘never users’, which applied to 18 controls and 6 cases. Duration of use was calculated by adding phases of any MHT use. Duration of phases of herbal or other non-hormonal therapies was excluded from total duration, if phase-specific information was available.
SNP selection and genotyping
We conducted an extensive literature search concerning the hypothetical mechanisms of action of MHT through sex steroid receptors, transporter, and metabolizing enzyme genes to select candidates that might be associated with CRC development and progression. SNPs were selected primarily according to their known or putative functional consequences, i.e. their modifying influence on protein structures, transcription levels or alternative splicing mechanisms. Polymorphisms were also assessed on the basis of NCBI data and with respect to linkage disequilibrium (LD).
Forty-seven SNPs in 10 hormone metabolism genes (COMT, CYP1A1, CYP1A2, CYP1B1, CYP2C9, CYP2C19, CYP3A4, CYP17A1, GSTP, and HSD17B1), one transport gene (ABCB1) and five signaling and binding genes (ESR1, ESR2, SHBG, PGR, and NR1I2) were selected for this study. Cases and controls were allocated randomly on 384-well plates (Applied Biosystems, Foster City, CA, USA). Genotyping was performed at the German Cancer Research Center (DKFZ) using KASPar assays (KBiosciences, Hoddesdon, Hertfordshire, UK). PCR products were analyzed with the ABI Prism 7900HT detection system applying the SDS 2.2 Software (Applied Biosystems).
We excluded 11 cases and 10 controls with a genotyping rate below 80%, assuming poor DNA quality and consequently unreliable measurements. Call rates of individual SNPs were >94.6% in both cases and controls. Samples (6.5%) were randomly selected and included as duplicate for internal quality control. A concordance rate of ≥99.2% was observed between the original and the duplicate samples.
Statistical analysis
All statistical analyses were performed using SAS, version 9.2 (SAS Institute, Cary, NC, USA). We used a χ2 test (1 degree of freedom (df)) to check whether genotype distributions in controls are in Hardy–Weinberg equilibrium (HWE) with P values <0.01 indicating a significant HWE deviation. A χ2 test (α=5%) was also used to test for differences in proportions of potential risk factors and protective factors between the case and control group.
Unconditional logistic regression was used to assess marginal genetic and environmental associations, CRC risk associated with MHT use in genotype subgroups, and gene–environment interaction. All models were adjusted for the matching factors age and county of residence, and additionally for putative confounders of MHT-associated CRC risk: former colorectal endoscopy (excluding endoscopies that led to the diagnosis of CRC), diagnosis of diabetes, body mass index (BMI) ≥5 years before diagnosis/date of interview, former general health check-ups, ever use of oral contraceptives, and physical activity in the last 12 months (metabolic equivalent of task, MET (hours/week)). When assessing marginal genetic associations with CRC risk, ever use of MHT was additionally included as covariate. A log-additive mode of inheritance (1 df) was assumed when evaluating marginal genetic effects and interactions with ever use or duration of MHT. Effect modification was assessed with a likelihood ratio test, comparing models with and without a multiplicative interaction term. Analyses stratified by genotype were performed if the genetic variant significantly modified the association of MHT with CRC risk. Odds ratios (ORs) and 95% CIs for ever use of MHT and duration of use per 5-year increment are presented.
Additionally, haplotype associations were assessed for genes incorporating variants that significantly modified MHT-associated CRC risk (ABCB1 and ESR1). Genotypes of controls were analyzed to calculate LD between genetic variants and define haplotype block structures using the Software Haploview (Version 4.2, Cambridge, MA, USA; Barrett et al. 2005). Block structures were determined according to the method of Gabriel et al. (2002). Haplotypes were estimated using SAS (Proc Haplotype). All haplotypes with frequencies below 1% were grouped together into one category. To assess interactions with MHT use, all estimated haplotypes and respective interaction terms were simultaneously included in multivariate regression models, using the most frequent haplotype as reference. Significance of effect modification was evaluated by the means of Wald statistics derived from the corresponding estimates of the logistic regression model.
All tests were two-sided and considered to be statistically significant with a P value of <0.05. Formal adjustment for multiple comparisons was not performed because of the exploratory character of these analyses.
Results
Population characteristics are shown in Table 1. Compared with controls, cases were more likely to have a higher BMI (P=0.03) and were more often diagnosed with diabetes mellitus (P<0.01). On the other hand, they less often reported a former colorectal endoscopy (P<0.01) and ever use of oral contraceptives (P<0.01). The prevalence of ever MHT use was 49.0% in the control group and 31.2% in the case group. The median duration of use was 10 years in controls (ranging from 3 months to 41.3 years) and 8 years in cases (ranging from 3 months to 45.5 years). Ever use of MHT was associated with CRC risk with an OR of 0.65 (95% CI 0.50–0.85). The OR for CRC risk associated with MHT duration was 0.88 (95% CI 0.81–0.95) per 5-year increment of duration.
Distribution of selected epidemiological variables in the postmenopausal female DACHS study population
| Variable | Controls, (N=684 N (%)) | Cases, (N=685 N (%)) | P value |
|---|---|---|---|
| Age (years) | |||
| 40 to <50 | 1 (0.1) | 6 (0.9) | 0.12 |
| 50 to <60 | 66 (9.6) | 73 (10.7) | |
| 60 to <70 | 260 (38.0) | 229 (33.4) | |
| 70 to <80 | 231 (33.8) | 232 (33.9) | |
| 80+ | 126 (18.4) | 145 (21.2) | |
| Body mass index (kg/m2) ≥5 years before diagnosis/date of interview | |||
| <18.5 | 13 (1.9) | 10 (1.5) | 0.03 |
| 18.5 to <25 | 316 (46.2) | 277 (40.4) | |
| 25 to <30 | 250 (36.5) | 234 (34.2) | |
| 30 to <35 | 79 (11.5) | 112 (16.4) | |
| 35 to <40 | 15 (2.2) | 26 (3.8) | |
| >40 | 7 (1.0) | 6 (0.9) | |
| Unknown | 4 (0.6) | 20 (2.9) | |
| Average physical activity in METs in last 12 months (hours/week) | |||
| <84.6 | 175 (25.6) | 211 (30.8) | 0.06 |
| 84.6 to <122.5 | 170 (24.9) | 154 (22.5) | |
| 122.5 to <183.0 | 168 (24.6) | 153 (22.3) | |
| ≥183.0 | 166 (24.3) | 135 (19.7) | |
| Unknown | 5 (0.7) | 32 (4.7) | |
| Any first-degree family history of colorectal cancer? | |||
| No | 570 (83.3) | 556 (81.2) | 0.31 |
| Yes | 94 (13.8) | 107 (15.6) | |
| Unknown | 20 (2.9) | 22 (3.2) | |
| Any diagnosis of diabetes (through a physician)? | |||
| No | 602 (88.1) | 540 (78.8) | <0.001 |
| Yes | 81 (11.8) | 136 (19.9) | |
| Unknown | 1 (0.01) | 9 (1.3) | |
| Ever had colorectal endoscopy? | |||
| No/unknown | 311 (45.5) | 549 (80.1) | <0.001 |
| Yes | 373 (54.5) | 135 (19.8) | |
| Unknown | 0 (0.0) | 1 (0.1) | |
| Ever use hormone replacement therapy? | |||
| No | 343 (50.1) | 463 (67.6) | <0.001 |
| Yes | 335 (49.0) | 214 (31.2) | |
| Unknown | 6 (0.9) | 8 (1.2) | |
| Duration of hormone replacement therapy | |||
| Never | 343 (50.1) | 463 (67.6) | <0.001 |
| ≥0.25 to <3 years | 59 (8.6) | 51 (7.4) | |
| ≥3 to <5 years | 42 (6.1) | 23 (3.4) | |
| ≥5 to <10 years | 59 (8.6) | 45 (6.6) | |
| ≥10 years | 175 (25.6) | 95 (13.9) | |
| Unknown | 6 (0.9) | 8 (1.2) | |
| Ever use oral contraceptives? | |||
| No | 374 (54.7) | 447 (65.3) | <0.001 |
| Yes | 309 (45.2) | 236 (34.5) | |
| Unknown | 1 (0.1) | 2 (0.3) | |
| CRC localization | |||
| Colon | N/A | 467 (68.2) | N/A |
| Rectum | N/A | 218 (31.8) |
N/A, characteristic only available for cases.
Information on each of the 47 genetic variants, genotype distributions, and associations of polymorphisms with CRC risk are shown in Table 2. The genotype distributions in controls of two polymorphisms showed a significant deviation from HWE (P<0.01), i.e. rs4646903 in the CYP1A1 gene and rs2740574 in the CYP3A4 gene. We excluded these from further investigation.
Genotype frequencies and colorectal cancer risk estimates of polymorphisms in genes related to sex steroid signaling, transport, and metabolism in the female postmenopausal DACHS study population
| Gene | Variant | Genotype | Effect estimate | |||
|---|---|---|---|---|---|---|
| dbSNP rs# | Amino acid substitution and functional consequence | MAF | Genotype | N Co/N Ca | OR (95% CI)a | P trend |
| ABCB1 rs1045642 | I1145I, T allele associated with lower protein expression (Hoffmeyer et al. 2000) | 49% | T/T | 183/190 | 1.00 | |
| T/C | 329/336 | 1.04 (0.77–1.40) | ||||
| C/C | 167/156 | 0.98 (0.69–1.39) | 0.914 | |||
| ABCB1 rs2229109 | S400N, functional consequences unknown | 6% | G/G | 596/614 | 1.00 | |
| G/A | 86/64 | 0.69 (0.47–1.03) | ||||
| A/A | 1/1 | 1.00 (0.03–33.9) | 0.078 | |||
| ABCB1 rs1202168 | Intronic, functional consequences unknown, in high LD interval (Soranzo et al. 2004) | 40% | C/C | 249/218 | 1.00 | |
| C/T | 312/337 | 1.08 (0.82–1.42) | ||||
| T/T | 119/127 | 1.02 (0.71–1.46) | 0.829 | |||
| ABCB1 rs9282564 | N21D, results in a net charge change of the protein (Brinkmann and Eichelbaum 2001) | 10% | A/A | 545/544 | 1.00 | |
| A/G | 117/124 | 1.05 (0.76–1.45) | ||||
| G/G | 12/9 | 0.73 (0.23–2.27) | 0.997 | |||
| ABCB1 rs2214102 | 5′-UTR, located at a translation initiation site (Hoffmeyer et al. 2000) | 7% | G/G | 593/580 | 1.00 | |
| G/A | 81/97 | 1.33 (0.91–1.92) | ||||
| A/A | 6/6 | 0.96 (0.26–3.51) | 0.208 | |||
| SHBG rs6259 | D356N, A allele is associated with reduced clearance of SHBG (Cousin et al. 1998) | 11% | G/G | 525/549 | 1.00 | |
| G/A | 141/115 | 0.69 (0.50–0.95) | ||||
| A/A | 6/10 | 1.13 (0.36–3.60) | 0.061 | |||
| ESR2 rs1255998 | 3′-UTR, functional consequences unknown, G allele associated with endometrial cancer risk (Ashton et al. 2009) | 12% | C/C | 519/547 | 1.00 | |
| C/G | 138/114 | 0.76 (0.55–1.04) | ||||
| G/G | 14/5 | 0.33 (0.11–1.01) | 0.015 | |||
| ESR2 rs928554 | 3′-UTR, G allele might create new acceptor splice site (MARIE-GENICA-Consortium 2010a) | 40% | A/A | 246/222 | 1.00 | |
| A/G | 316/317 | 1.19 (0.90–1.56) | ||||
| G/G | 111/133 | 1.48 (1.03–2.13) | 0.032 | |||
| ESR2 rs4986938 | 3′-UTR, potentially affects pre-mRNA splicing (Zheng et al. 2003) | 38% | G/G | 260/264 | 1.00 | |
| G/A | 310/306 | 0.97 (0.74–1.27) | ||||
| A/A | 99/106 | 1.02 (0.70–1.48) | 0.997 | |||
| ESR2 rs1271572 | 5′-UTR, might modulate binding of transcription factors (MARIE-GENICA-Consortium 2010a) | 42% | G/G | 231/208 | 1.00 | |
| G/T | 329/330 | 1.29 (0.97–1.70) | ||||
| T/T | 120/138 | 1.40 (0.98–2.00) | 0.046 | |||
| ESR1 rs851984 | 5′-UTR, functional consequences unknown, T allele associated with breast cancer risk (MARIE-GENICA-Consortium 2010a) | 41% | C/C | 221/231 | 1.00 | |
| C/T | 341/314 | 0.95 (0.72–1.25) | ||||
| T/T | 103/122 | 1.07 (0.74–1.55) | 0.833 | |||
| ESR1 rs2881766 | 5′-UTR, functional consequences unknown, G allele associated with breast cancer risk (MARIE-GENICA-Consortium 2010a) | 18% | T/T | 461/453 | 1.00 | |
| T/G | 190/195 | 1.02 (0.77–1.35) | ||||
| G/G | 25/22 | 0.95 (0.48–1.87) | 0.989 | |||
| ESR1 rs2071454 | 5′-UTR, functional consequences unknown | 11% | T/T | 536/551 | 1.00 | |
| T/G | 132/120 | 0.81 (0.59–1.12) | ||||
| G/G | 5/8 | 1.75 (0.51–6.03) | 0.467 | |||
| ESR1 rs2077647 | S10S, potentially affects mRNA structure (Tanaka et al. 2003) | 47% | A/A | 183/187 | 1.00 | |
| A/G | 347/349 | 0.97 (0.73–1.30) | ||||
| G/G | 140/135 | 0.93 (0.65–1.33) | 0.697 | |||
| ESR1 rs827421 | Intronic, functional consequences unknown | 48% | T/T | 172/178 | 1.00 | |
| T/C | 350/355 | 1.02 (0.76–1.37) | ||||
| C/C | 151/143 | 0.91 (0.64–1.30) | 0.620 | |||
| ESR1 rs2234693 | Intronic, C allele introduces a binding site for transcription factor B-myb, leading to altered transcription (Herrington et al. 2002) | 46% | T/T | 188/209 | 1.00 | |
| T/C | 348/341 | 0.93 (0.70–1.24) | ||||
| C/C | 136/120 | 0.87 (0.61–1.24) | 0.429 | |||
| ESR1 rs9340799 | Intronic, functional consequences unknown, may lead to altered transcription or splicing (Schuit et al. 2004) | 36% | A/A | 274/287 | 1.00 | |
| A/G | 319/317 | 1.01 (0.78–1.32) | ||||
| G/G | 83/72 | 0.94 (0.63–1.42) | 0.864 | |||
| ESR1 rs3798577 | 3′-UTR, functional consequences unknown | 45% | T/T | 203/189 | 1.00 | |
| T/C | 328/330 | 1.14 (0.86–1.52) | ||||
| C/C | 139/154 | 1.27 (0.89–1.80) | 0.181 | |||
| ESR1 rs910416 | 3′ near gene, functional consequences unknown, modified estrogen monotherapy-associated breast cancer risk (MARIE-GENICA-Consortium 2010a) | 48% | T/T | 178/188 | 1.00 | |
| T/C | 338/314 | 0.81 (0.61–1.10) | ||||
| C/C | 148/159 | 0.92 (0.65–1.31) | 0.599 | |||
| PGR rs1042838 | V660L, SNP possibly affects receptor dimerization, nuclear localization, ligand binding, cofactor interaction (Agoulnik et al. 2004) | 15% | C/C | 495/478 | 1.00 | |
| C/A | 154/181 | 1.32 (0.99–1.76) | ||||
| A/A | 24/16 | 0.91 (0.44–1.88) | 0.202 | |||
| PGR rs1379130 | G393G, functional consequences unknown | 36% | G/G | 272/298 | 1.00 | |
| G/A | 311/268 | 0.74 (0.56–0.96) | ||||
| A/A | 88/108 | 1.01 (0.69–1.47) | 0.412 | |||
| PGR rs10895068 | 5′-UTR, creates new transcription start site, increasing expression of PR-B isoform (De Vivo et al. 2002) | 4% | G/G | 615/619 | 1.00 | |
| G/A | 58/59 | 0.89 (0.58–1.38) | ||||
| A/A | 1/1 | 1.94 (0.05–80.6) | 0.670 | |||
| PGR rs518162 | 5′-UTR, SNP is located between PR-B and PR-A transcription start sites, potentially affecting PR-A/B expression (De Vivo et al. 2002) | 7% | G/G | 580/586 | 1.00 | |
| G/A | 93/88 | 1.06 (0.74–1.53) | ||||
| A/A | 3/4 | 1.23 (0.21–7.06) | 0.710 | |||
| NR1I2 rs1523127 | 5′-UTR, belongs to a haplotype incl. SNPs that introduce new transcription factor binding sites (Zhang et al. 2001) | 39% | A/A | 245/258 | 1.00 | |
| A/C | 326/317 | 0.89 (0.68–1.17) | ||||
| C/C | 98/88 | 0.89 (0.61–1.32) | 0.450 | |||
| NR1I2 rs2276706 | Intronic, belongs to a haplotype incl. SNPs that introduce new transcription factor binding sites (Zhang et al. 2001) | 38% | G/G | 251/267 | 1.00 | |
| G/A | 329/324 | 0.92 (0.70–1.20) | ||||
| A/A | 95/83 | 0.89 (0.60–1.31) | 0.474 | |||
| NR1I2 rs1464603 | Intronic, functional consequences unknown | 32% | T/T | 303/307 | 1.00 | |
| T/C | 310/291 | 1.05 (0.80–1.36) | ||||
| C/C | 65/78 | 1.11 (0.73–1.69) | 0.608 | |||
| NR1I2 rs6785049 | Intronic, G allele associated with increased induction of CYP3A (Zhang et al. 2001) | 38% | A/A | 260/264 | 1.00 | |
| A/G | 323/313 | 1.00 (0.77–1.31) | ||||
| G/G | 94/101 | 1.10 (0.75–1.60) | 0.698 | |||
| NR1I2 rs2276707 | Intronic, T allele associated with increased induction of CYP3A (Zhang et al. 2001) | 17% | C/C | 446/439 | 1.00 | |
| C/T | 180/190 | 1.00 (0.75–1.32) | ||||
| T/T | 21/24 | 1.19 (0.58–2.44) | 0.794 | |||
| NR1I2 rs1054191 | 3′-UTR, A allele associated with decreased induction of CYP3A (Zhang et al. 2001) | 14% | G/G | 499/518 | 1.00 | |
| G/A | 161/143 | 0.95 (0.70–1.28) | ||||
| A/A | 17/12 | 0.80 (0.35–1.87) | 0.581 | |||
| NR1I2 rs3814057 | 3′-UTR, C allele associated with decreased induction of P-glycoprotein (Zhang et al. 2001) | 17% | A/A | 458/440 | 1.00 | |
| A/C | 177/201 | 1.15 (0.87–1.52) | ||||
| C/C | 22/24 | 1.20 (0.60–2.41) | 0.307 | |||
| COMT rs4680 | V158M, A allele leads to thermo-labile protein and 2–3 fold lower catalytic activity (Dawling et al. 2001) | 49% | G/G | 171/177 | 1.00 | |
| A/G | 343/327 | 0.88 (0.66–1.19) | ||||
| A/A | 162/175 | 1.05 (0.74–1.48) | 0.802 | |||
| HSD17B1 rs605059 | G313S, functional consequences unknown, C allele has been associated with lower estradiol levels (Setiawan et al. 2004) | 46% | T/T | 199/167 | 1.00 | |
| T/C | 330/340 | 1.30 (0.97–1.75) | ||||
| C/C | 144/170 | 1.41 (0.99–2.01) | 0.051 | |||
| CYP1B1 rs1800440 | N453S, G allele leads to enzyme which catalyzes hydroxylation of estradiol more efficiently (Hanna et al. 2000) | 19% | A/A | 452/467 | 1.00 | |
| A/G | 202/187 | 0.89 (0.68–1.18) | ||||
| G/G | 26/22 | 0.81 (0.40–1.62) | 0.342 | |||
| CYP1B1 rs1056836 | L432V, G allele leads to enzyme with increased activity (Shimada et al. 1999) | 41% | C/C | 224/220 | 1.00 | |
| C/G | 339/320 | 1.00 (0.75–1.32) | ||||
| G/G | 106/128 | 1.31 (0.91–1.88) | 0.209 | |||
| CYP1B1 rs1056827 | A119S, T allele leads to enzyme which catalyzes hydroxylation of estradiol more efficiently (Hanna et al. 2000) | 31% | G/G | 317/323 | 1.00 | |
| G/T | 304/297 | 0.90 (0.70–1.17) | ||||
| T/T | 57/55 | 0.85 (0.54–1.35) | 0.362 | |||
| CYP1B1 rs10012 | R48G, G allele may increase enzyme activity (Hanna et al. 2000) | 30% | C/C | 322/329 | 1.00 | |
| C/G | 299/295 | 0.91 (0.70–1.18) | ||||
| G/G | 54/55 | 0.91 (0.57–1.44) | 0.495 | |||
| CYP1A1 rs4646903 | 3′ near gene, C allele has been associated with increased enzyme activity (Bartsch et al. 2000) | 9% | T/T | 564/539 | N/A | |
| T/C | 102/134 | N/A | ||||
| C/C | 13/6 | N/A | N/A | |||
| CYP1A2 rs762551 | Intronic, C allele associated with reduced CYP1A1 activity in smokers (Sachse et al. 1999) | 28% | A/A | 353/354 | 1.00 | |
| A/C | 280/261 | 0.93 (0.72–1.21) | ||||
| C/C | 47/63 | 1.08 (0.68–1.72) | 0.937 | |||
| CYP2C19 rs12248560 | 5′-UTR, T allele associated with enhanced enzyme activity (Sim et al. 2006) | 22% | C/C | 417/403 | 1.00 | |
| C/T | 216/229 | 1.03 (0.79–1.35) | ||||
| T/T | 42/39 | 0.74 (0.43–1.28) | 0.584 | |||
| CYP2C19 rs4244285 | P227P, leads to splicing defect in exon 5 and loss of enzyme activity (Wedlund 2000) | 16% | G/G | 470/494 | 1.00 | |
| G/A | 186/175 | 0.84 (0.64–1.11) | ||||
| A/A | 18/11 | 0.43 (0.18–0.98) | 0.046 | |||
| CYP2C9 rs1799853 | R144C, T allele associated with decreased enzyme activity (King et al. 2004) | 12% | C/C | 524/525 | 1.00 | |
| C/T | 137/141 | 0.97 (0.71–1.31) | ||||
| T/T | 12/9 | 0.80 (0.28–2.25) | 0.695 | |||
| CYP2C9 rs1057910 | I359L, C allele associated with decreased enzyme activity (King et al. 2004) | 6% | A/A | 599/588 | 1.00 | |
| A/C | 74/89 | 1.31 (0.89–1.92) | ||||
| C/C | 2/4 | 2.22 (0.35–13.8) | 0.113 | |||
| CYP3A4 rs2740574 | 5′-UTR, ambiguous findings regarding functional consequences, association with ovarian cancer (Pearce et al. 2009) | 5% | A/A | 616/624 | N/A | |
| A/G | 50/50 | N/A | ||||
| G/G | 6/0 | N/A | N/A | |||
| CYP3A4 rs11773597 | 5′-UTR, functional consequences unknown | 6% | G/G | 592/598 | 1.00 | |
| G/C | 84/80 | 0.90 (0.62–1.32) | ||||
| C/C | 1/2 | 0.85 (0.04–18.1) | 0.586 | |||
| CYP17A1 rs743572 | 5′-UTR, potentially enhances expression, modified breast cancer risk associated with estrogen monotherapy (MARIE-GENICA-Consortium 2010b) | 39% | A/A | 252/213 | 1.00 | |
| A/G | 326/355 | 1.31 (1.00–1.72) | ||||
| G/G | 103/115 | 1.24 (0.86–1.81) | 0.128 | |||
| GSTP1 rs1695 | I105V, G allele associated with increased enzyme activity (Hu et al. 1997) | 33% | A/A | 298/302 | 1.00 | |
| A/G | 308/300 | 1.04 (0.80–1.35) | ||||
| G/G | 70/72 | 1.14 (0.74–1.75) | 0.556 | |||
| GSTP1 rs1138272 | A114V, T allele associated with decreased enzyme activity (Srivastava et al. 1999) | 9% | C/C | 553/570 | 1.00 | |
| C/T | 118/103 | 0.95 (0.68–1.33) | ||||
| T/T | 5/4 | 1.00 (0.22–4.45) | 0.803 | |||
N Co/N Ca, N controls/N cases; N/A, no estimate due to violation of Hardy–Weinberg equilibrium.
Models adjusted for age, county of residence, former colorectal endoscopy (yes/no), ever use of hormone replacement therapy (yes/no), ever diagnosis of diabetes (yes/no), body mass index (in categories of <23 kg/m2, 23 to <25 kg/m2, 25 to <27 kg/m2, 27 to <30 kg/m2, and >30 kg/m2), former general health checks (yes/no), ever use of oral contraceptives (yes/no), and physical activity in METs (in categories of <84.6 h/week, 84.6 to <122.5 h/week, 122.5 to <183.0 h/week, and ≥183.0 h/week).
Significant allele–dosage associations with CRC risk were found with three polymorphisms in the ESR2 gene, rs1255998, rs928554, and rs1271572 (P trend=0.02, 0.03, and 0.05, respectively). The polymorphism rs4244285 in the phase-I metabolism gene CYP2C19 showed a marginally significant allele–dosage effect per minor allele (P trend=0.05).
We assessed possible interactions of each of the 47 SNPs with ever use of MHT as well as with duration of therapy. For all 47 investigated genetic variants, interaction ORs and P values are shown in Supplementary Table 1, see section on supplementary data given at the end of this article. For polymorphisms that showed significant modification of the MHT effect, ORs associated with MHT according to genotype are shown in Table 3.
Effect modification of menopausal hormone therapy-associated colorectal cancer risk by genotype
| Genetic variant | Genotype | |||
|---|---|---|---|---|
| MHT measure | Common homozygous (N controls/N cases) OR (95% CI)a | Heterozygous (N controls/N cases) OR (95% CI)a | Rare homozygous (N controls/N cases) OR (95% CI)a | P interaction* |
| ABCB1_rs1202168_C>T | C/C (237/198) | C/T (303/305) | T/T (116/111) | |
| MHT ever versus never | 0.84 (0.53–1.34) | 0.62 (0.42–0.91) | 0.43 (0.21–0.88) | 0.039 |
| MHT duration (per 5-year increment) | 0.94 (0.83–1.08) | 0.85 (0.75–0.95) | 0.78 (0.62–0.97) | 0.038 |
| ESR1_rs910416_T>C | T/T (172/171) | T/C (327/285) | C/C (148/144) | |
| MHT ever versus never | 0.41 (0.23–0.72) | 0.71 (0.48–1.04) | 0.87 (0.48–1.60) | 0.030 |
| MHT duration (per 5-year increment) | 0.83 (0.71–0.97) | 0.86 (0.76–0.96) | 0.99 (0.83–1.18) | 0.071 |
*P value for testing of effect modification by genotype utilizing an interaction term of MHT measure and genetic polymorphism assuming a log-additive mode of inheritance (1df).
Models adjusted for age, county of residence, former colorectal endoscopy (yes/no), ever diagnosis of diabetes (yes/no), body mass index (in categories of <23 kg/m2, 23 to <25 kg/m2, 25 to <27 kg/m2, 27 to <30 kg/m2, and >30 kg/m2), former general health checks (yes/no), ever use of oral contraceptives (yes/no), and physical activity in METs (in categories of <84.6 h/week, 84.6 to <122.5 h/week, 122.5 to <183.0 h/week, and ≥183.0 h/week).
We observed that CRC risk associated with ever use of MHT was modified by the polymorphism rs1202168 in ABCB1, a gene related to estrogen transport (interaction OR=0.69, 95% CI 0.48–0.98, P interaction=0.04). The reduced risk for CRC was only present in carriers of the minor T allele. While risk in heterozygous MHT users was observed to be similar to the risk in the whole study population (OR=0.62, 95% CI 0.42–0.91; Table 3), CRC risk associated with ever use of MHT in women homozygous for the minor T allele was somewhat lower (OR=0.43, 95% CI 0.21–0.88; Table 3). Similarly, rs1202168 modified risk association with duration of MHT (interaction OR=0.89, 95% CI 0.80–1.00, P interaction=0.04). In carriers of the minor T allele, MHT duration was significantly associated with decreased CRC risk (OR=0.85, 95% CI 0.75–0.95 in women heterozygous for the T allele and OR=0.78, 95% CI 0.62–0.97 in homozygous women, respectively), but not in women homozygous for the common C allele (OR=0.94, 95% CI 0.83–1.08).
Applying of the method of Gabriel et al. (2002) to our data revealed the presence of one haplotype block in ABCB1 (Supplementary Figure 1, see section on supplementary data given at the end of this article). Interaction ORs for haplotypes in ABCB1 and MHT use are shown in Table 4. The results support the risk-modifying effect of rs1202168. Carriers of the GT haplotype are at lower risk for CRC associated with MHT use compared with carriers of the referent GC haplotype. Significant interaction was observed for both, ever use of MHT (interaction OR=0.67, 95% CI 0.46–0.98, P interaction=0.04) and duration of MHT (interaction OR=0.89, 95% CI 0.79–0.99, P interaction=0.04).
Interaction effects between haplotypes in ABCB1 and menopausal hormone therapy use and duration on colorectal cancer risk
| Interaction effects | |||||
|---|---|---|---|---|---|
| Haplotype | MHT ever versus never | MHT duration (per 5-year increment) | |||
| Genetic variantsb | Haplotype frequency | Interaction OR (95% CI)a | P interactionc | Interaction OR (95% CI)a | P interactionc |
| ABCB1_rs2229109 ABCB1_rs1202168 | |||||
| GC | 0.525 | 1.00 | 1.00 | ||
| GT | 0.418 | 0.67 (0.46–0.98) | 0.039 | 0.89 (0.79–0.99) | 0.040 |
| AC | 0.056 | 0.75 (0.29–1.91) | 0.543 | 0.94 (0.70–1.25) | 0.653 |
| HRare | 0.001 | N/A | N/A | N/A | N/A |
N/A, no estimates due to low haplotype frequency. HRare, category comprising haplotypes with frequencies below 1%.
Model adjusted for age, county of residence, former colorectal endoscopy (yes/no), ever diagnosis of diabetes (yes/no), body mass index (in categories of <23 kg/m2, 23 to <25 kg/m2, 25 to <27 kg/m2, 27 to <30 kg/m2, and >30 kg/m2), former general health checks (yes/no), ever use of oral contraceptives (yes/no), and physical activity in METs (in categories of <84.6 h/week, 84.6 to <122.5 h/week, 122.5 to <183.0 h/week, and ≥183.0 h/week). All haplotypes and respective interaction terms were included simultaneously in the model.
Minor alleles are underlined.
Wald statistic of the interaction term in the multivariate model.
Furthermore, we observed a significant effect modification of MHT ever use by rs910416 in the ER gene ESR1 (interaction OR=1.49, 95% CI 1.04–2.13, P interaction=0.03). Strong reduction in CRC risk associated with MHT use was found in women homozygous for the common T allele (OR=0.41, 95% CI 0.23–0.72), while less pronounced or no reduction in risk was observed in women carrying the minor C allele (OR=0.71, 95% CI 0.48–1.04 in heterozygous women and OR=0.87, 95% CI 0.48–1.60 in homozygous women, respectively). ESR1_rs910416 showed a similar modifying effect on MHT duration (interaction OR=1.10, 95% CI 0.99–1.23, P interaction=0.07).
Using LD estimates for polymorphisms in ESR1, two haplotype blocks were defined (Supplementary Figure 2, see section on supplementary data given at the end of this article). None of them included the variant rs910416 and significant statistical interaction of haplotypes was observed neither with MHT use, nor with duration (Supplementary Table 2, see section on supplementary data given at the end of this article).
Our data also provided some indication that rs4986938 in ESR2 altered CRC risk associated with MHT use (interaction OR=1.38, 95% CI 0.96–1.99, P interaction=0.08) as well as MHT duration (interaction OR=1.10, 95% CI 0.99–1.22, P interaction=0.07; Supplementary Table 1, see section on supplementary data given at the end of this article).
Discussion
This is the first investigation that identified possible genetic modifiers of MHT-associated CRC risk. We investigated 47 SNPs in 16 candidate genes related to sex steroid transport, metabolism and signaling. Significant modification was observed with polymorphisms in the P-glycoprotein gene ABCB1 and in the ER alpha gene ESR1.
The gene product of ABCB1, P-glycoprotein, is located at the plasma membrane of cells and is expressed at high concentrations in the large intestine (Ho et al. 2003). P-glycoprotein is a transporter protein known to be involved in the cellular efflux of a wide variation of substrates and it has been implicated in the development of drug resistance of tumor cells (Gutmann et al. 2010). Results of in vitro studies suggest that estrogens are substrates of P-glycoprotein and that it decreases the accumulation of a wide range of steroids (Barnes et al. 1996, Kim & Benet 2004). Additionally, sex steroids seem to influence P-glycoprotein expression (Mutoh et al. 2006, Bebawy & Chetty 2009). Progesterone, its natural metabolic products, as well as synthetic progestins, have also been found to effectively inhibit transport by P-glycoprotein (Barnes et al. 1996, Hamilton et al. 2001, Frohlich et al. 2004). Based on that evidence, complex hormone–P-glycoprotein interactions are conceivable and it seems likely that MHT influences P-glycoprotein function and that vice versa P-glycoprotein modifies actions of sex hormones administered via MHT.
In the present study, the inverse association of MHT ever use with CRC was only seen in women carrying at least one copy of the minor T allele of ABCB1_rs1202168. The finding was supported by a significant statistical interaction observed with duration of MHT use. Additionally, a haplotype including rs1202168 showed a modifying effect on MHT-associated CRC risk. The polymorphism rs1202168 is located in an intronic region of the ABCB1 gene and functional consequences are not yet known. The variant might affect gene transcription or splicing and consequently affect the transporting function of P-glycoprotein and cellular levels of hormones. We cannot rule out that the observed effect modification may be induced by a causal genetic variant in LD with rs1202168. The variant rs1202168 seems to be located in a high LD interval surrounding rs1045642, where a causal variant for altered P-glycoprotein activity is anticipated (Soranzo et al. 2004).
ESR1 encodes ERα, a member of the steroid receptor superfamily. Estrogen receptors influence cellular growth by regulating gene transcription (Ascenzi et al. 2006). In the colon, ERα is expressed at low levels, and ERβ is thought to be the primary mediator of sex steroid effects on colonic tissue (Kennelly et al. 2008). ERα and ERβ differ in their transcriptional functions, although they are highly homologous in their structural and functional domains and show similar DNA-binding properties and estrogen-binding affinities (McInerney et al. 1998, Delaunay et al. 2000, Ascenzi et al. 2006). Estrogen signaling via ERα has been found to promote cell proliferation, while ERβ mediates pro-apoptotic effects and seems to attenuate ERα activity (Ascenzi et al. 2006, Chen et al. 2008). However, an in vitro study in colon cancer cells observed activation of the p53 pathway via ERα, leading to apoptosis (Hsu et al. 2006). Furthermore, presence of functioning ERβ as well as ERα inhibited Apc-dependent tumor development in the colon of Min/+ mice (Cho et al. 2007). These findings suggest similar effects of both ERs in colonic tissue.
We found a significant effect heterogeneity of MHT-associated CRC risk with rs910416 in ESR1. Only women homozygous for the major T allele were at significantly decreased risk of developing CRC after taking menopausal hormone replacement therapy. A similar, but non-significant, effect modification was seen for duration of MHT with rs910416. ESR1_rs910416 is located in an untranscribed downstream region of the gene and a functional effect is not known. Therefore, it is likely that rs910416 is in LD with the functionally relevant genetic variant. Interestingly, rs910416 has been found to modify the effect of estrogen monotherapy in a study on breast cancer risk (MARIE-GENICA-Consortium 2010a). In line with our results, a significantly increased risk for breast cancer associated with estrogen monotherapy was only seen in women homozygous for the common T allele.
Regarding ERβ, we observed a non-significant modification of CRC risk associated with MHT use by ESR2_rs4986938. Similar to ESR1_rs910416, women homozygous for the common G allele had a reduced risk for CRC when under MHT, while this association was attenuated in carriers of the variant A allele. The A allele of the polymorphism rs4986938 might affect pre-mRNA splicing, as it degenerates the exonic splicer enhancer motif for the splice factor SC35 (Zheng et al. 2003). The altered splicing might then consequently limit functioning of ERβ.
So far, gene–environment interaction regarding MHT-associated CRC risk has rarely been studied. One study, which assessed the association of genetic variants in five sex steroid-related genes (ESR1, ESR2, PGR, CYP19A1, and HSD17B2) with CRC risk in 158 CRC cases and 563 controls, mentioned that current MHT use did not appear to modify the reported non-significant genetic associations without providing results (Lin et al. 2010). Another study of 1021 female cases and 1798 controls did not assess effect modification directly related to MHT use but compared the risk of CRC associated with polymorphisms in the ERα, ERβ, and androgen receptor gene in ‘estrogen negative’ (postmenopausal women not using MHT) and in ‘estrogen positive’ women (who were either premenopausal or used MHT). They reported significantly stronger associations of two ERβ gene polymorphisms, the SNP rs1256049 and a CA repeat in intron 5, with the ‘estrogen negative’ group (Slattery et al. 2005). The variant rs1256049 was not included in this study.
We investigated gene–MHT interactions among 685 female postmenopausal CRC patients and 684 controls participating in the DACHS study. Strengths of the present study comprise its population-based design, the candidate gene approach and the validation of the information collected on MHT use. Although recall bias might be potentially relevant, a former study in the DACHS sample did not observe differential deviations of the self-reported compared with the record-based MHT duration among cases and controls with complete validation (Hoffmeister et al. 2009). Misclassification as well as differential selection in cases and controls regarding MHT use and genotypes cannot be ruled out, but these common concerns in epidemiologic studies are unlikely to strongly affect our results. When a multiplicative interaction term is used to assess gene–environment interactions, misclassification may lead to a decline in power, but inflated type I error rates were not observed (Lindstrom et al. 2009). Furthermore, gene–environment interactions are not prone to bias by differential selection, assuming that the genotype does not influence participation conditional on exposure and disease status (Morimoto et al. 2003).
This study had 60% power to detect the interaction OR of 0.69 seen with ABCB1_rs1202168, given a prevalence of MHT use of 40%. Accordingly, we had 68% power to detect the interaction OR of 1.49 observed with ESR1_rs910416. It is therefore possible that some existent genetic modifying effects have not been detected due to insufficient power. We also did not investigate the genetic modification of CRC risk associated with type-specific MHT use (e.g. estrogen monotherapy and combined estrogen–progestagen therapy) or different administration routes of MHT due to limited sample sizes in the specific subgroups. Additionally, the investigated SNPs were selected based on their potential functional consequences and represent only a small subset of the actual genetic variation in the selected candidate genes. Therefore, important effect modification of MHT-associated CRC risk by genetic variants that are not in high LD with the investigated variants might have been missed.
Although we observed some significant interactions at the 0.05 level, none of these findings remain statistically significant after correction for multiple testing. Consequently, our results have to be interpreted with caution and require replication in independent study populations.
In summary, the results of the present study suggest that the transporter gene ABCB1 and the ERα gene ESR1 modulate the effect of exogenous hormones on colorectal tumorigenesis. Given that little is known about the biological mechanisms underlying the role of exogenous sex steroids regarding CRC development, we believe our exploratory findings are valuable. Furthermore, gaining insight into biological pathways is crucial to direct further clinical research, potentially leading to an improved prevention of CRC. These findings warrant replication in sufficiently large studies. Regarding the complexity of factors influencing CRC development, it is likely that additional genes and genetic variants are involved in modulating the effect of MHT on colorectal carcinogenesis.
Supplementary data
This is linked to the online version of the paper at http://dx.doi.org/10.1530/ERC-11-0057.
Declaration of interest
The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research reported.
Funding
The DACHS study was supported by grants from the German Research Council (Deutsche Forschungsgemeinschaft, grant numbers BR 1704/6-1, BR 1704/6-3, BR 1704/6-4 and CH 390 117/1-1), and the German Federal Ministry of Education and Research (grant numbers 01KH0404 and 01ER0814). This work was funded by the NGFN+ (Nationales Genomforschungsnetz), grant number 01GS08181.
Acknowledgements
We thank all participants of the DACHS study and the interviewers, physicians and recruiting hospitals who have been involved in realizing the study. We also acknowledge the excellent technical assistance by Belinda-Su Kaspereit and Ute Handte-Daub.
References
Agoulnik IU, Tong XW, Fischer DC, Korner K, Atkinson NE, Edwards DP, Headon DR, Weigel NL & Kieback DG 2004 A germline variation in the progesterone receptor gene increases transcriptional activity and may modify ovarian cancer risk. Journal of Clinical Endocrinology and Metabolism 89 6340–6347 doi:10.1210/jc.2004-0114.
Ascenzi P, Bocedi A & Marino M 2006 Structure–function relationship of estrogen receptor alpha and beta: impact on human health. Molecular Aspects of Medicine 27 299–402 doi:10.1016/j.mam.2006.07.001.
Ashton KA, Proietto A, Otton G, Symonds I, McEvoy M, Attia J, Gilbert M, Hamann U & Scott RJ 2009 Estrogen receptor polymorphisms and the risk of endometrial cancer. BJOG : an International Journal of Obstetrics and Gynaecology 116 1053–1061 doi:10.1111/j.1471-0528.2009.02185.x.
Barnes KM, Dickstein B, Cutler GB Jr, Fojo T & Bates SE 1996 Steroid treatment, accumulation, and antagonism of P-glycoprotein in multidrug-resistant cells. Biochemistry 35 4820–4827 doi:10.1021/bi952380k.
Barrett JC, Fry B, Maller J & Daly MJ 2005 Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21 263–265 doi:10.1093/bioinformatics/bth457.
Bartsch H, Nair U, Risch A, Rojas M, Wikman H & Alexandrov K 2000 Genetic polymorphism of CYP genes, alone or in combination, as a risk modifier of tobacco-related cancers. Cancer Epidemiology, Biomarkers & Prevention 9 3–28.
Bebawy M & Chetty M 2009 Gender differences in P-glycoprotein expression and function: effects on drug disposition and outcome. Current Drug Metabolism 10 322–328 doi:10.2174/138920009788498996.
Beral V, Banks E & Reeves G 2002 Evidence from randomised trials on the long-term effects of hormone replacement therapy. Lancet 360 942–944 doi:10.1016/S0140-6736(02)11032-4.
Berggren S, Gall C, Wollnitz N, Ekelund M, Karlbom U, Hoogstraate J, Schrenk D & Lennernas H 2007 Gene and protein expression of P-glycoprotein, MRP1, MRP2, and CYP3A4 in the small and large human intestine. Molecular Pharmacology 4 252–257 doi:10.1021/mp0600687.
Brenner H, Chang-Claude J, Seiler CM, Sturmer T & Hoffmeister M 2006 Does a negative screening colonoscopy ever need to be repeated? Gut 55 1145–1150 doi:10.1136/gut.2005.087130.
Brenner H, Chang-Claude J, Seiler CM, Rickert A & Hoffmeister M 2011 Protection from colorectal cancer after colonoscopy: a population-based, case–control study. Annals of Internal Medicine 154 22–30.
Brinkmann U & Eichelbaum M 2001 Polymorphisms in the ABC drug transporter gene MDR1. Pharmacogenomics Journal 1 59–64 doi:10.1038/sj.tpj.6500001.
Chen GG, Zeng Q & Tse GM 2008 Estrogen and its receptors in cancer. Medicinal Research Reviews 28 954–974 doi:10.1002/med.20131.
Chlebowski RT, Wactawski-Wende J, Ritenbaugh C, Hubbell FA, Ascensao J, Rodabough RJ, Rosenberg CA, Taylor VM, Harris R & Chen C et al. 2004 Estrogen plus progestin and colorectal cancer in postmenopausal women. New England Journal of Medicine 350 991–1004 doi:10.1056/NEJMoa032071.
Cho NL, Javid SH, Carothers AM, Redston M & Bertagnolli MM 2007 Estrogen receptors alpha and beta are inhibitory modifiers of Apc-dependent tumorigenesis in the proximal colon of Min/+ mice. Cancer Research 67 2366–2372 doi:10.1158/0008-5472.CAN-06-3026.
Cousin P, Dechaud H, Grenot C, Lejeune H & Pugeat M 1998 Human variant sex hormone-binding globulin (SHBG) with an additional carbohydrate chain has a reduced clearance rate in rabbit. Journal of Clinical Endocrinology and Metabolism 83 235–240 doi:10.1210/jc.83.1.235.
Dawling S, Roodi N, Mernaugh RL, Wang X & Parl FF 2001 Catechol-O-methyltransferase (COMT)-mediated metabolism of catechol estrogens: comparison of wild-type and variant COMT isoforms. Cancer Research 61 6716–6722.
Delaunay F, Pettersson K, Tujague M & Gustafsson JA 2000 Functional differences between the amino-terminal domains of estrogen receptors alpha and beta. Molecular Pharmacology 58 584–590.
De Vivo I, Huggins GS, Hankinson SE, Lescault PJ, Boezen M, Colditz GA & Hunter DJ 2002 A functional polymorphism in the promoter of the progesterone receptor gene associated with endometrial cancer risk. PNAS 99 12263–12268 doi:10.1073/pnas.192172299.
Dudley EC, Hopper JL, Taffe J, Guthrie JR, Burger HG & Dennerstein L 1998 Using longitudinal data to define the perimenopause by menstrual cycle characteristics. Climacteric 1 18–25 doi:10.3109/13697139809080677.
Foley EF, Jazaeri AA, Shupnik MA, Jazaeri O & Rice LW 2000 Selective loss of estrogen receptor beta in malignant human colon. Cancer Research 60 245–248.
Frohlich M, Albermann N, Sauer A, Walter-Sack I, Haefeli WE & Weiss J 2004 In vitro and ex vivo evidence for modulation of P-glycoprotein activity by progestins. Biochemical Pharmacology 68 2409–2416 doi:10.1016/j.bcp.2004.08.026.
Gabriel SB, Schaffner SF, Nguyen H, Moore JM, Roy J, Blumenstiel B, Higgins J, DeFelice M, Lochner A & Faggart M et al. 2002 The structure of haplotype blocks in the human genome. Science 296 2225–2229 doi:10.1126/science.1069424.
Gutmann DA, Ward A, Urbatsch IL, Chang G & van Veen HW 2010 Understanding polyspecificity of multidrug ABC transporters: closing in on the gaps in ABCB1. Trends in Biochemical Science 35 36–42 doi:10.1016/j.tibs.2009.07.009.
Hamilton KO, Yazdanian MA & Audus KL 2001 Modulation of P-glycoprotein activity in Calu-3 cells using steroids and beta-ligands. International Journal of Pharmaceutics 228 171–179 doi:10.1016/S0378-5173(01)00836-5.
Hanna IH, Dawling S, Roodi N, Guengerich FP & Parl FF 2000 Cytochrome P450 1B1 (CYP1B1) pharmacogenetics: association of polymorphisms with functional differences in estrogen hydroxylation activity. Cancer Research 60 3440–3444.
Hartman J, Edvardsson K, Lindberg K, Zhao C, Williams C, Strom A & Gustafsson JA 2009 Tumor repressive functions of estrogen receptor beta in SW480 colon cancer cells. Cancer Research 69 6100–6106 doi:10.1158/0008-5472.CAN-09-0506.
Herrington DM, Howard TD, Brosnihan KB, McDonnell DP, Li X, Hawkins GA, Reboussin DM, Xu J, Zheng SL & Meyers DA et al. 2002 Common estrogen receptor polymorphism augments effects of hormone replacement therapy on E-selectin but not C-reactive protein. Circulation 105 1879–1882 doi:10.1161/01.CIR.0000016173.98826.88.
Hildebrand JS, Jacobs EJ, Campbell PT, McCullough ML, Teras LR, Thun MJ & Gapstur SM 2009 Colorectal cancer incidence and postmenopausal hormone use by type, recency, and duration in cancer prevention study II. Cancer Epidemiology, Biomarkers & Prevention 18 2835–2841 doi:10.1158/1055-9965.EPI-09-0596.
Ho GT, Moodie FM & Satsangi J 2003 Multidrug resistance 1 gene (P-glycoprotein 170): an important determinant in gastrointestinal disease? Gut 52 759–766 doi:10.1136/gut.52.5.759.
Hoffmeister M, Raum E, Winter J, Chang-Claude J & Brenner H 2007 Hormone replacement therapy, body mass, and the risk of colorectal cancer among postmenopausal women from Germany. British Journal of Cancer 97 1486–1492 doi:10.1038/sj.bjc.6604066.
Hoffmeister M, Raum E, Krtschil A, Chang-Claude J & Brenner H 2009 No evidence for variation in colorectal cancer risk associated with different types of postmenopausal hormone therapy. Clinical Pharmacology and Therapeutics 86 416–424 doi:10.1038/clpt.2009.134.
Hoffmeyer S, Burk O, von Richter O, Arnold HP, Brockmoller J, Johne A, Cascorbi I, Gerloff T, Roots I & Eichelbaum M et al. 2000 Functional polymorphisms of the human multidrug-resistance gene: multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. PNAS 97 3473–3478 doi:10.1073/pnas.050585397.
Hsu HH, Cheng SF, Chen LM, Liu JY, Chu CH, Weng YJ, Li ZY, Lin CS, Lee SD & Kuo WW et al. 2006 Over-expressed estrogen receptor-alpha up-regulates hTNF-alpha gene expression and down-regulates beta-catenin signaling activity to induce the apoptosis and inhibit proliferation of LoVo colon cancer cells. Molecular and Cellular Biochemistry 289 101–109 doi:10.1007/s11010-006-9153-3.
Hu X, O'Donnell R, Srivastava SK, Xia H, Zimniak P, Nanduri B, Bleicher RJ, Awasthi S, Awasthi YC & Ji X et al. 1997 Active site architecture of polymorphic forms of human glutathione S-transferase P1-1 accounts for their enantioselectivity and disparate activity in the glutathione conjugation of 7beta,8alpha-dihydroxy-9alpha,10alpha-ox y-7,8,9,10-tetrahydrobenzo(a)pyrene. Biochemical and Biophysical Research Communications 235 424–428 doi:10.1006/bbrc.1997.6777.
Jassam N, Bell SM, Speirs V & Quirke P 2005 Loss of expression of oestrogen receptor beta in colon cancer and its association with Dukes' staging. Oncology Reports 14 17–21.
Kabat GC, Miller AB & Rohan TE 2008 Oral contraceptive use, hormone replacement therapy, reproductive history and risk of colorectal cancer in women. International Journal of Cancer 122 643–646 doi:10.1002/ijc.23079.
Kennelly R, Kavanagh DO, Hogan AM & Winter DC 2008 Oestrogen and the colon: potential mechanisms for cancer prevention. Lancet Oncology 9 385–391 doi:10.1016/S1470-2045(08)70100-1.
Kim WY & Benet LZ 2004 P-glycoprotein (P-gp/MDR1)-mediated efflux of sex-steroid hormones and modulation of P-gp expression in vitro. Pharmaceutical Research 21 1284–1293 doi:10.1023/B:PHAM.0000033017.52484.81.
King BP, Khan TI, Aithal GP, Kamali F & Daly AK 2004 Upstream and coding region CYP2C9 polymorphisms: correlation with warfarin dose and metabolism. Pharmacogenetics 14 813–822 doi:10.1097/00008571-200412000-00004.
Konstantinopoulos PA, Kominea A, Vandoros G, Sykiotis GP, Andricopoulos P, Varakis I, Sotiropoulou-Bonikou G & Papavassiliou AG 2003 Oestrogen receptor beta (ERbeta) is abundantly expressed in normal colonic mucosa, but declines in colon adenocarcinoma paralleling the tumour's dedifferentiation. European Journal of Cancer 39 1251–1258 doi:10.1016/S0959-8049(03)00239-9.
Koo JH & Leong RW 2010 Sex differences in epidemiological, clinical and pathological characteristics of colorectal cancer. Journal of Gastroenterology and Hepatology 25 33–42 doi:10.1111/j.1440-1746.2009.05992.x.
La Vecchia C, Gallus S & Fernandez E 2005 Hormone replacement therapy and colorectal cancer: an update. Journal of the British Menopause Society 11 166–172 doi:10.1258/136218005775544264.
Lilla C, Verla-Tebit E, Risch A, Jager B, Hoffmeister M, Brenner H & Chang-Claude J 2006 Effect of NAT1 and NAT2 genetic polymorphisms on colorectal cancer risk associated with exposure to tobacco smoke and meat consumption. Cancer Epidemiology, Biomarkers & Prevention 15 99–107 doi:10.1158/1055-9965.EPI-05-0618.
Lin J, Zee RY, Liu KY, Zhang SM, Lee IM, Manson JE, Giovannucci E, Buring JE & Cook NR 2010 Genetic variation in sex-steroid receptors and synthesizing enzymes and colorectal cancer risk in women. Cancer Causes & Control 21 897–908 doi:10.1007/s10552-010-9518-5.
Lindstrom S, Yen YC, Spiegelman D & Kraft P 2009 The impact of gene-environment dependence and misclassification in genetic association studies incorporating gene–environment interactions. Human Heredity 68 171–181 doi:10.1159/000224637.
MARIE-GENICA-Consortium on Genetic Susceptibility for Menopausal Hormone Therapy Related Breast Cancer Risk Polymorphisms in genes of the steroid receptor superfamily modify postmenopausal breast cancer risk associated with menopausal hormone therapy International Journal of Cancer 126 2010a 2935–2946 doi:10.1002/ijc.24892.
MARIE-GENICA-Consortium on Genetic Susceptibility for Menopausal Hormone Therapy Related Breast Cancer Risk Postmenopausal estrogen monotherapy-associated breast cancer risk is modified by CYP17A1_−34_T>C polymorphism Breast Cancer Research and Treatment 120 2010b 737–744 doi:10.1007/s10549-009-0490-2.
McInerney EM, Weis KE, Sun J, Mosselman S & Katzenellenbogen BS 1998 Transcription activation by the human estrogen receptor subtype beta (ER beta) studied with ER beta and ER alpha receptor chimeras. Endocrinology 139 4513–4522 doi:10.1210/en.139.11.4513.
Morimoto LM, White E & Newcomb PA 2003 Selection bias in the assessment of gene–environment interaction in case–control studies. American Journal of Epidemiology 158 259–263 doi:10.1093/aje/kwg147.
Mutoh K, Tsukahara S, Mitsuhashi J, Katayama K & Sugimoto Y 2006 Estrogen-mediated post transcriptional down-regulation of P-glycoprotein in MDR1-transduced human breast cancer cells. Cancer Science 97 1198–1204 doi:10.1111/j.1349-7006.2006.00300.x.
Nebert DW & Russell DW 2002 Clinical importance of the cytochromes P450. Lancet 360 1155–1162 doi:10.1016/S0140-6736(02)11203-7.
Pearce CL, Near AM, Van Den Berg DJ, Ramus SJ, Gentry-Maharaj A, Menon U, Gayther SA, Anderson AR, Edlund CK & Wu AH et al. 2009 Validating genetic risk associations for ovarian cancer through the international Ovarian Cancer Association Consortium. British Journal of Cancer 100 412–420 doi:10.1038/sj.bjc.6604820.
Roediger WE & Babidge W 1997 Human colonocyte detoxification. Gut 41 731–734 doi:10.1136/gut.41.6.731.
Sachse C, Brockmoller J, Bauer S & Roots I 1999 Functional significance of a C→A polymorphism in intron 1 of the cytochrome P450 CYP1A2 gene tested with caffeine. British Journal of Clinical Pharmacology 47 445–449 doi:10.1046/j.1365-2125.1999.00898.x.
Schuit SC, Oei HH, Witteman JC, Geurts van Kessel CH, van Meurs JB, Nijhuis RL, van Leeuwen JP, de Jong FH, Zillikens MC & Hofman A et al. 2004 Estrogen receptor alpha gene polymorphisms and risk of myocardial infarction. Journal of the American Medical Association 291 2969–2977 doi:10.1001/jama.291.24.2969.
Setiawan VW, Hankinson SE, Colditz GA, Hunter DJ & De Vivo I 2004 HSD17B1 gene polymorphisms and risk of endometrial and breast cancer. Cancer Epidemiology, Biomarkers & Prevention 13 213–219 doi:10.1158/1055-9965.EPI-03-0241.
Shimada T, Watanabe J, Kawajiri K, Sutter TR, Guengerich FP, Gillam EM & Inoue K 1999 Catalytic properties of polymorphic human cytochrome P450 1B1 variants. Carcinogenesis 20 1607–1613 doi:10.1093/carcin/20.8.1607.
Sim SC, Risinger C, Dahl ML, Aklillu E, Christensen M, Bertilsson L & Ingelman-Sundberg M 2006 A common novel CYP2C19 gene variant causes ultrarapid drug metabolism relevant for the drug response to proton pump inhibitors and antidepressants. Clinical Pharmacology and Therapeutics 79 103–113 doi:10.1016/j.clpt.2005.10.002.
Slattery ML, Sweeney C, Murtaugh M, Ma KN, Wolff RK, Potter JD, Caan BJ & Samowitz W 2005 Associations between ERalpha, ERbeta, and AR genotypes and colon and rectal cancer. Cancer Epidemiology, Biomarkers & Prevention 14 2936–2942 doi:10.1158/1055-9965.EPI-05-0514.
Soranzo N, Cavalleri GL, Weale ME, Wood NW, Depondt C, Marguerie R, Sisodiya SM & Goldstein DB 2004 Identifying candidate causal variants responsible for altered activity of the ABCB1 multidrug resistance gene. Genome Research 14 1333–1344 doi:10.1101/gr.1965304.
Srivastava SK, Singhal SS, Hu X, Awasthi YC, Zimniak P & Singh SV 1999 Differential catalytic efficiency of allelic variants of human glutathione S-transferase Pi in catalyzing the glutathione conjugation of thiotepa. Archives of Biochemistry and Biophysics 366 89–94 doi:10.1006/abbi.1999.1217.
Tanaka Y, Sasaki M, Kaneuchi M, Shiina H, Igawa M & Dahiya R 2003 Polymorphisms of estrogen receptor alpha in prostate cancer. Molecular Carcinogenesis 37 202–208 doi:10.1002/mc.10138.
Tsilidis KK, Allen NE, Key TJ, Sanjoaquin MA, Bakken K, Berrino F, Fournier A, Lund E, Overvad K & Olsen A et al. 2010 Menopausal hormone therapy and risk of colorectal cancer in the European Prospective Investigation into Cancer and Nutrition. International Journal of Cancer 128 1881–1889 doi:10.1002/ijc.25504.
Wedlund PJ 2000 The CYP2C19 enzyme polymorphism. Pharmacology 61 174–183 doi:10.1159/000028398.
Wilkins HR, Doucet K, Duke V, Morra A & Johnson N 2010 Estrogen prevents sustained COLO-205 human colon cancer cell growth by inducing apoptosis, decreasing c-myb protein, and decreasing transcription of the anti-apoptotic protein bcl-2. Tumour Biology 31 16–22 doi:10.1007/s13277-009-0003-2.
Xita N & Tsatsoulis A 2010 Genetic variants of sex hormone-binding globulin and their biological consequences. Molecular and Cellular Endocrinology 316 60–65 doi:10.1016/j.mce.2009.08.025.
Zhang J, Kuehl P, Green ED, Touchman JW, Watkins PB, Daly A, Hall SD, Maurel P, Relling M & Brimer C et al. 2001 The human pregnane X receptor: genomic structure and identification and functional characterization of natural allelic variants. Pharmacogenetics 11 555–572 doi:10.1097/00008571-200110000-00003.
Zheng SL, Zheng W, Chang BL, Shu XO, Cai Q, Yu H, Dai Q, Xu J & Gao YT 2003 Joint effect of estrogen receptor beta sequence variants and endogenous estrogen exposure on breast cancer risk in Chinese women. Cancer Research 63 7624–7629.
