Skip Navigation


Carcinogenesis Advance Access originally published online on January 3, 2008
Carcinogenesis 2008 29(3):573-578; doi:10.1093/carcin/bgm277
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
29/3/573    most recent
bgm277v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Zabaleta, J.
Right arrow Articles by Ochoa, A. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zabaleta, J.
Right arrow Articles by Ochoa, A. C.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2008. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Interactions of cytokine gene polymorphisms in prostate cancer risk

Jovanny Zabaleta1,*, Hui-Yi Lin2, Rosa A. Sierra3, M.Craig Hall4,8, Peter E. Clark5, Oliver A. Sartor6, Jennifer J. Hu7 and Augusto C. Ochoa3

1 Department of Genetics, Louisiana State University Health Sciences Center, 533 Bolivar Street, CSRB 455, New Orleans, LA 70112, USA
2 Medical Statistics Section, University of Alabama at Birmingham, Birmingham, AL 35294, USA
3 Department of Pediatrics, Louisiana State University Health Sciences Center, New Orleans, LA 70112, USA
4 Department of Urology and Comprehensive Cancer Center, Wake Forest University School of Medicine, Winston-Salem, NC 27157, USA
5 Department of Urologic Surgery and Vanderbilt-Ingram Comprehensive Cancer Center, Vanderbilt University Medical Center, Nashville, TN 37232, USA
6 Lank Center for Genitourinary Oncology, Dana Farber Cancer Institute, Harvard Medical School, Harvard University, Boston, MA 02115, USA and
7 Sylvester Comprehensive Cancer Center and Department of Epidemiology and Public Health, University of Miami School of Medicine, Miami, FL 33136, USA
8 Present address: Piedmont Urological Associates, High Point, NC 27262, USA

* To whom correspondence should be addressed. Tel: +(504) 599 0920; Fax: +(504) 599 0911; Email: jzabal{at}lsuhsc.edu


   Abstract
Top
 Abstract
Introduction
 Materials and methods
 Results
 Discussion
 Funding
 References
 
Prostate cancer (CaP) is the second leading cause of cancer death in American men. Chronic inflammation has been one of several factors associated with the development of CaP. Single-nucleotide polymorphisms (SNPs) in cytokine genes have been associated with increased inflammation, increased cytokine production and possibly increased CaP risk. However, the effects of cytokine SNPs on CaP susceptibility have not been consistent. Using the genomic DNA collected in a CaP case–control study (557 cases and 547 controls), we pilot tested the interactions of nine functionally characterized SNPs of three cytokine genes in CaP risk using the multivariate adaptive regression splines (MARS)–logit models. African-Americans with the IL10–819TT genotype had a lower CaP risk [odds ratio (OR) = 0.27, 95% confidence interval (CI) = 0.07–1.01], but subjects with the genotype combination of IL1B–511CT/TT and IL10–592CC had a higher CaP risk (OR = 2.56, 95% CI = 1.09–6.02). In Caucasians, higher CaP risk was associated with the IL10–1082AG/GG genotype (OR = 3.62, 95% CI = 1.42–9.28), the genotype combination of IL10–1082AA plus IL1B–31TT/TC (OR = 2.92, 95% CI = 1.13–7.55) and the genotype combination of TNF–238GG plus IL10–592AA (OR = 2.14, 95% CI = 1.05–4.38). Our results highlight the importance of cytokine SNPs and their interactions in CaP risk.

Abbreviations: BPH, benign prostatic hyperplasia; CaP, prostate cancer; CI, confidence interval; HWE, Hardy–Weinberg equilibrium; IL, interleukin; MARS, multivariate adaptive regression spline; OR, odds ratio; SNP, single-nucleotide polymorphism; TNF, tumor necrosis factor


   Introduction
Top
 Abstract
 Introduction
Materials and methods
 Results
 Discussion
 Funding
 References
 
Prostate cancer (CaP) is the second leading cause of cancer death in American men with >27 500 deaths estimated in 2007 (1). The incidence of CaP increases with age (1), family history and race/ethnicity (1–3). A recent study suggests that ~42% of the risk for CaP may be explained by heritable factors (4). Several works based on epidemiological and genetic studies have proposed genes such as HPC1 (5), CAPB (6), BRCA1 and BRCA2 (7), as susceptibility genes for CaP. Chronic inflammation has been associated with increased risk in CaP (8). The initiation, maintenance and pathology of the inflammatory response depend upon pro- and anti-inflammatory signals. Interleukin (IL) 1β, tumor necrosis factor (TNF)-{alpha} and IL10 are critical in the regulation of inflammation (9–12). Differential production of these cytokines has been associated with single-nucleotide polymorphisms (SNPs) (13–19). IL1β is essential in promoting inflammation. Several IL1B SNPs have been associated with different types of cancer (20–23). TNF-{alpha} is a proinflammatory molecule involved in the initiation and maintaining of the inflammatory response. A G to A transition at position –308 of the TNF-A gene (TNF) has been associated with increased protein production in vitro (15,18) in several diseases including cancer and infectious diseases (24–27). IL10 is an antiinflammatory cytokine that has been shown to inhibit vascular epithelial growth factor (14), which may contribute to angiogenesis and survival of the tumor cells (28,29). Reduced levels of IL10 have been associated with the presence of the IL10–1082A/–819T/–592A haplotype on the IL10 gene promoter (13,17) as well as with increased risk of cancer in several populations (25,30,31). The objective of this study was to evaluate the associations of nine functionally characterized SNPs (IL1B–511C>T, IL1B–31T>C, IL1B+3954C>T, IL10–1082A>G, IL10–819C>T, IL10–592C>A, TNF–857C>T, TNF–308G>A and TNF–238G>A) with CaP risk.


   Materials and methods
Top
 Abstract
 Introduction
 Materials and methods
Results
 Discussion
 Funding
 References
 
Study populations
DNA samples used for this clinic-based study were obtained from CaP cases and controls from the Departments of Urology and Internal Medicine of the Wake Forest University School of Medicine with sequential patient population as described previously (32). We anticipated that the sample size for African-American cases would be about 10–15%. Therefore, we oversampled African-American controls (case:control = 1:2 ratio) to increase statistical power. For Caucasian, we used case:control ratio of 1:1. All subjects received a description of the study and signed their informed consent according to the Medical Center's Institutional Review Board. The inclusion and exclusion criteria are described elsewhere (32). A total of 200 African-Americans and 889 Caucasians were included in this study.

SNP determination
In this study, we evaluated the association of the following nine SNPs with CaP risk: three in the IL1B gene (IL1B–511C>T, IL1B–31T>C and IL1B+3954C>T), three in the IL10 gene (IL10–1082A>G, IL10–819C>T and IL10–592C>A) and three in the TNF gene (TNF–857C>T, TNF–308G>A and TNF–238G>A). All SNPs were determined by TaqMan genotyping assays (Applied Biosystems, Foster City, CA) using probes labeled with either FAM or VIC dyes. Briefly, 4 ng DNA was mixed with 2x TaqMan Universal Master Mix (Applied Biosystems), water and the respective SNP mix, heated at 95°C for 10 min and subjected to 40 cycles of 15 s at 92°C and 1 min at 60°C. The polymerase chain reaction product was analyzed using a 7900 HT instrument (Applied Biosystems) for the presence of VIC or FAM fluorescent, or both, using the Sequence Detection System (Applied Biosystems) that determines the genotype. Controls of known genotype for each polymorphic locus were always run in parallel with each experiment per SNP analysis. In addition, four internal controls were included in each 96-well plate. The concordance of the genotypes for both types of controls was >98.5%.

Statistical analysis
Chi-square tests were used to compare the demographic and clinical characteristics between cases and controls. With differential distributions of case–control in the two racial groups, we performed all the statistical analyses stratified by race. The Hardy–Weinberg equilibrium (HWE) was examined for both Caucasians and African-American control groups by using the exact test (33). Genotypes deviated from HWE were excluded before performing association analysis. Logistic regression models were used for testing the one-to-one association between CaP risk and each of the nine SNPs. The potential confounding/effect modifiers include age, smoking history, family history and benign prostatic hyperplasia (BPH). The smoking history was evaluated by whether the individual has ever smoked at least 100 cigarettes in lifetime. The existence of family history is determined by his first-degree relatives (father or brothers) with CaP. The crude and adjusted odds ratios (ORs) and 95% confidence interval (CI) were presented.

In order to evaluate SNP–SNP interactions, we pilot tested the multivariate adaptive regression splines (MARS)–logit models. The modes of inheritance (dominant, recessive or additive) and interaction patterns were automatically selected by MARS (Salford Systems, San Diego, CA). In this study, we allowed a maximum of 70 basis functions and 10-fold cross-validation for MARS model selection and tested up to three-way interactions (34). In the MARS–logit hybrid model, MARS was applied as a variable screening tool and the selected terms with their extended terms were plugged in a logistic regression model for the second-step variable selection. The parent terms of the MARS-selected terms are their main and lower order interaction terms. The extended terms include all such terms selected from a MARS model with a lower order interaction (35). For example, the parent terms of A x B x C were the main effects of A, B and C and the two-way interactions of A x B, A x C and B x C. In a three-way interaction MARS model, all terms and their parent terms selected from the one-, two- and three-way MARS model were included. Then, we applied the second-step variable selection in a logistic model by using the stepwise automatic selection with P values of 5% as entry and removal criteria. The final model selection was based on the Bayesian information criterion (36). Lower Bayesian information criterion value represents a better fit model. The best models for each of ethnic group were displayed as trees. SAS 9.1 (SAS Institute, Cary, NC) and MARS 2.0 were used for data management and analyses. All frequencies were model frequencies from complete data set used in the final MARS–logit models.


   Results
Top
 Abstract
 Introduction
 Materials and methods
 Results
Discussion
 Funding
 References
 
The demographic and clinical characteristics of patients and controls are summarized in Table I. Except for smoking history, all other variables were significantly different between CaP patients and controls. About half of the CaP patients were in a range of 60–69 years old. Most controls (54.4%) reported a positive history of BPH, whereas most of CaP cases (56.6%) had a negative history of BPH. About 17% of controls reported having at least one first-degree relative with CaP, whereas almost 30% of the CaP cases reported positive family history.


View this table:
[in this window]
[in a new window]

  		Table I. Demographic and clinical characteristics of study subjects

 
The distribution of genotypes in our study was compared with that reported previously in African-American and Caucasians (Table II). Our results are within the range of published studies (37–41). Distributions of genotypes were different comparing that in African-Americans and Caucasians. Based on this information, subsequent analyses were stratified by race.


View this table:
[in this window]
[in a new window]

  		Table II. Distribution of genotypes in healthy African-American and Caucasian controls

 
The associations between each SNP and CaP risk with or without controlling for age, family history, smoking history and BPH are summarized in Table III. After adjusting for the demographic and clinical factors, African-American individuals with genotype IL10–819TT had a lower risk of CaP compared with those with the genotype IL10–819CC. Three SNPs (IL10–1082A>G, IL1B–31T>C and TNF–238G>A) did not follow HWE in African-Americans and were not included in the final analyses. In Caucasians, a reduced risk of CaP was associated with the IL10–1082AA (OR = 0.67, 95% CI = 0.45–1.0), IL10–592CA (OR = 0.7, 95% CI = 0.52–0.93) and IL10–819CT (OR = 0.72, 95% CI = 0.53–0.98) genotypes after adjusting for age, family history, smoking history and BPH (Table II). There was one SNP (TNF–857C>T) that did not follow HWE in Caucasians.


View this table:
[in this window]
[in a new window]

  		Table III. Distribution and ORs of CaP of cytokine gene polymorphisms in Caucasians and African-Americans

 
One main effect and one two-way gene–gene interaction were associated with the risk of CaP in African-Americans using the MARS–logit model (Figure 1). The association of the IL10–819TT genotype and lower CaP risk was marginally significant (OR = 0.27, 95% CI = 0.07–1.01, P = 0.052). The genotype combination of IL1B–511CT/TT and IL10–592CC was significantly associated with a higher risk of CaP (OR = 2.56, 95% CI = 1.09–6.02) after adjusting for age, family history, smoking history and BPH. As shown in Figure 2, in Caucasians, one main and two two-way gene–gene interactions were significantly associated with CaP risk. An increased risk of CaP was associated with IL10–1082AG/GG (OR = 3.62, 95% CI = 1.42–9.28), the genotype combination of IL10–1082AA and IL1B–31TT/TC (OR = 2.92, 95% CI = 1.13–7.55) and the genotype combination of TNF–238GG and IL10–592AA (OR = 2.14, 95% CI = 1.05–4.38). We also analyzed haplotype frequencies and their association with CaP risk. However, none of the haplotypes was associated with CaP risk (results not shown).


Figure 1
View larger version (10K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1. A two-way gene–gene interaction is associated with the risk of CaP in African-Americans. Interaction of the different SNPs was determined by using the MARS–logit model, as described in Materials and Methods. A marginally significant protective role of IL10-819TT (OR = 0.27, 95% CI = 0.07–101, P = 0.052) was observed. Individuals carrying a combination of IL1B-511CT/TT and IL10-592CC genotypes had increased risk of CaP (OR = 2.56, 95% CI = 1.09–6.02). *Frequency of case/control for complete data in the model. **All OR adjusted for age, family history, smoking history and BPH.

 


Figure 2
View larger version (12K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2. Two two-way gene–gene interactions are associated with CaP risk in Caucasians. Interaction of the different SNPs was determined by using the MARS–logit model, as described in Materials and Methods. Individuals carrying the IL10-1082AG/GG genotype had increased risk of CaP (OR = 3.62, 95% CI = 1.42–9.28). Similarly, individuals carrying the combination of IL10-1082AA and IL1B-31TT/TC genotypes or the genotype combination of TNF-238GG and IL10-592AA were found to have increased risk of CaP (OR = 2.92, 95% CI = 1.13–7.55 and OR = 2.14, 95% CI = 1.05–4.38, respectively). *Frequency of case/control for complete data in the model. **All OR adjusted for age, family history, smoking history and BPH.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
Funding
 References
 
Defining the genetic effect on CaP has proven to be a complex issue despite the fact that several studies have shown that family history is a risk factor strongly associated with the disease (3). The results from twin studies suggested that, for most of the cancers, environmental factors seem to play a major role in determining the degree of risk of developing the malignancy (4). However, it has been shown that heritable factors may play a particularly important role in determining the risk of CaP (4,42). Several genes have been linked to CaP development (4–7); however, confirmatory results have been inconsistent may be due to poorly defined genetic differences between populations and by the effect of gene–gene or gene–environment interactions.

Chronic inflammation has been directly associated with the risk of developing CaP. By following a cohort of patients for 5 years, MacLennan et al. (8) have shown that a higher percentage of those patients initially diagnosed with the presence of chronic inflammation developed newly diagnosed prostate adenocarcinoma (20 versus 6% in those with no chronic inflammation involved). Since inflammation seems to be pivotal for the development of several malignancies including CaP (8,43–47) and because no major gene has been consistently linked to CaP risk, it would be possible to think of SNPs on cytokine genes as modifiers of gene activity and modifiers of CaP risk. Even though SNPs have been associated with increased risk of several types of cancer (20–22,25,26), studies about their association with CaP have been contradictory. Using 247 CaP patients and 263 controls (all CaP patients and controls were Caucasian) from London, McCarron et al. (48) found that CaP patients had a significant increased frequency of IL10-1082AA genotype compared with that in controls (31.6 versus 20.6%, respectively; P = 0.01, OR = 1.78, 95% CI = 1.14–2.77). However, a more recent study by Michaud et al. (49) did not confirm their findings and reported two other IL10 SNPs associated with risk of CaP.

In this study, we evaluated the frequency of nine functionally characterized SNPs in three cytokine genes and their association with CaP. Our results show that Caucasians carrying the IL10–1082AA SNP had a reduced risk of CaP (OR = 0.67, 95% CI = 0.45–1.0). Similarly, Caucasian individuals carrying the genotypes IL10–592CA and IL10–819CT also had reduced risks of CaP (OR = 0.7, 95% CI = 0.52–0.93 and OR = 0.72, 95% CI = 0.53–0.98, respectively). These three SNPs are in linkage disequilibrium (17) forming an haplotype (IL10–1082A/–819T/–592A) that has been associated with reduced levels of IL10 (13,17) and increased risk of cancer in several populations (25,30,31). However, in our study this haplotype did not modify the risk of CaP. Other studies have suggested that higher levels of IL10 may be deleterious and in fact may promote the development of solid tumors. Fortis et al. (50) has shown that patients with malignant melanoma, pancreatic carcinoma and stomach cancer had increased levels of serum IL10 when compared with healthy controls opening the possibility that increased levels of IL10 in CaP patients are in fact involved in the pathogenesis of the disease. It is not clear how IL10 may play a dual role in the development of malignancy. By suppressing the Th1 response and by inhibiting phagocytic functions, IL10 may promote the tumor cells to evade the immune system and promote uncontrolled metastasis. In contrast, higher levels of IL10 have been associated with reduced angiogenesis via reduction of vascular epithelial growth factor expression (14), controlling the progression of the tumor by limiting the access to blood supply. The reported differences on the role of IL10 in carcinogenesis may be in part due to the different origins of the tumor cells as well as the advanced stage of the disease.

The lack of consistency of single SNP analysis in CaP susceptibility may be due to the relatively minor effect that a single SNP may have in the expression or function of the gene. It is more probable that combinations of SNPs in haplotypes or SNP–SNP interactions may modify the risk of developing a malignancy. It was recently shown that when analyzed separately, single SNPs did not modify the risk of breast cancer (51); however, it was noted that the risk was greatly modified by SNP–SNP interactions. Using multivariate logistic models, these authors showed that SNPs on genes involved in different pathways interact and increase the risk of developing breast cancer. For example, genes like XPD (Lys751Gln), involved in DNA repair, were consistently associated with IL10–1082G>A increasing the risk depending on the alleles in both genes (46). Other SNP–SNP interactions suggested the interaction of pathways like cell cycle and metabolism in the regulation of the overall breast cancer risk.

Because traditional logistic regression has limitations in testing high-order SNP–SNP interactions (52), we used the MARS–logit hybrid model to pilot test the high-order SNP–SNP interactions. The MARS–logit model (35) combines MARS (53) and a logistic model. MARS is considered the most flexible method to determine gene interactions, compared with classification and regression trees, and traditional logistic regression (35) and it has performed better than artificial neural networks (54). The MARS–logit tended to be more powerful than MARS alone in detecting SNP–SNP interactions (55). Interestingly, we found different risk patterns in African-Americans and Caucasians. CaP risk was associated with one main effect and one two-way interaction in African-Americans whereas two two-way gene–gene interactions were observed in Caucasians. In African-Americans, individuals with IL10–819TT had lower CaP risk and those with IL1B–511CT/TT plus IL10–592CC combination had higher CaP risk. Caucasians with the IL10–1082AG/GG genotype and the genotype combination of IL10–1082AA plus IL1B–31TT/TC had higher CaP risk. In addition, Caucasians with the combination of TNF–238GG plus IL10–592AA presented increased risk of CaP. It is noteworthy that single SNPs may not have significant effect on CaP risk but their interactions may impact CaP risk. In African-Americans, the main effects of IL1B–511 and IL10–592 were not significantly associated with CaP risk but their interactions were. In Caucasians, the main effects of IL1B–31 and TNF–238 were not significantly associated with CaP risk but there were two-way SNP–SNP interactions in CaP risk. In general, the traditional modeling in detecting interaction follows the hierarchical rule, in which main effects and lower order interactions contained in a significant interaction must remain in the model even if they are not significant. Without this restriction, MARS can detect the SNP–SNP interactions with no or weak main effects.

The mechanisms for differential SNP–CaP risk interactions between African-Americans and Caucasians remain unclear. First, it may be due to the frequency of individual SNPs in each population. A second possibility would be the presence of other factors (SNPs and transcription factors) that may be modulating the interaction of the different genetic markers in each population and then modifying the risk of CaP. In addition, the effect of the cytokine SNP interactions described here would be further modulated by environmental risk factors (inflammation, viral infections, smoking, environment exposures and others). Taken together, all these factors would, at the end, impact the definitive risk to the disease.

In summary, our results highlight the importance of studying the distribution of genetic determinants in different racial/ethnic groups to investigate their role in malignancy. The outcome of this study shows that multiple immune response pathways may interact with each other and modify the risk of CaP. More extensive work is needed to understand the effect that single SNP and SNP–SNP interactions may play in predicting the risk of CaP as well as in modifying the severity and the final outcome of the disease.


   Funding
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Funding
References
 
American Cancer Society (CNE-101119) to J.J.H.; National Research Foundation to the Wake Forest University's General Clinical Research Center (M01-RR07122); National Cancer Institute, National Institutes of Health (CA82689, CA107974 [GenBank] ) to A.C.O.


   Acknowledgments
 
The authors are grateful to study participants. We also want to acknowledge the contributions of Frank M.Torti, Robert Lee, Charles J.Rosser, Dean G.Assimos, Elizabeth Albertson, Dominck J.Carbone, William Rice, Francis O'Brien, Ray Morrow, Franklyn Millman, Nadine Shelton, Joel Anderson, Shirley Cothren, Eunkyung Chang, the General Clinical Research Center, the Urology Clinic and the Internal Medicine Clinic at Wake Forest University School of Medicine.

Conflict of Interest Statement: None declared.


   References
Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Funding
 References
 

  1. Jemal A, et al. Cancer statistics, 2007. CA Cancer J. Clin. (2007) 57:43–66.[Abstract/Free Full Text]
  2. Carter BS, et al. Hereditary prostate cancer: epidemiologic and clinical features. J. Urol. (1993) 150:797–802.[Web of Science][Medline]
  3. Lesko SM, et al. Family history and prostate cancer risk. Am. J. Epidemiol. (1996) 144:1041–1047.[Abstract/Free Full Text]
  4. Lichtenstein P, et al. Environmental and heritable factors in the causation of cancer—analyses of cohorts of twins from Sweden, Denmark, and Finland. N. Engl. J. Med. (2000) 343:78–85.[Abstract/Free Full Text]
  5. Smith JR, et al. Major susceptibility locus for prostate cancer on chromosome 1 suggested by a genome-wide search. Science (1996) 274:1371–1374.[Abstract/Free Full Text]
  6. Gibbs M, et al. Evidence for a rare prostate cancer-susceptibility locus at chromosome 1p36. Am. J. Hum. Genet. (1999) 64:776–787.[CrossRef][Web of Science][Medline]
  7. Ford D, et al. Risks of cancer in BRCA1-mutation carriers. Breast Cancer Linkage Consortium. Lancet (1994) 343:692–695.[CrossRef][Web of Science][Medline]
  8. MacLennan GT, et al. The influence of chronic inflammation in prostatic carcinogenesis: a 5-year followup study. J. Urol. (2006) 176:1012–1016.[CrossRef][Web of Science][Medline]
  9. Dinarello CA. Biologic basis for interleukin-1 in disease. Blood (1996) 87:2095–2147.[Abstract/Free Full Text]
  10. Hart PH, et al. Comparison of the suppressive effects of interleukin-10 and interleukin-4 on synovial fluid macrophages and blood monocytes from patients with inflammatory arthritis. Immunology (1995) 84:536–542.[Web of Science][Medline]
  11. Moore KW, et al. Interleukin-10 and the interleukin-10 receptor. Annu. Rev. Immunol. (2001) 19:683–765.[CrossRef][Web of Science][Medline]
  12. Vassalli P. The pathophysiology of tumor necrosis factors. Annu. Rev. Immunol. (1992) 10:411–452.[CrossRef][Web of Science][Medline]
  13. Crawley E, et al. Polymorphic haplotypes of the interleukin-10 5' flanking region determine variable interleukin-10 transcription and are associated with particular phenotypes of juvenile rheumatoid arthritis. Arthritis Rheum. (1999) 42:1101–1108.[CrossRef][Web of Science][Medline]
  14. Huang S, et al. Regulation of tumor growth and metastasis by interleukin-10: the melanoma experience. J. Interferon Cytokine Res. (1999) 19:697–703.[CrossRef][Web of Science][Medline]
  15. Kroeger KM, et al. The –308 tumor necrosis factor-alpha promoter polymorphism effects transcription. Mol. Immunol. (1997) 34:391–399.[CrossRef][Web of Science][Medline]
  16. Rad R, et al. Cytokine gene polymorphisms influence mucosal cytokine expression, gastric inflammation, and host specific colonisation during Helicobacter pylori infection. Gut (2004) 53:1082–1089.[Abstract/Free Full Text]
  17. Turner DM, et al. An investigation of polymorphism in the interleukin-10 gene promoter. Eur. J. Immunogenet. (1997) 24:1–8.[Web of Science][Medline]
  18. Wilson AG, et al. Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc. Natl Acad. Sci. USA (1997) 94:3195–3199.[Abstract/Free Full Text]
  19. Wu WS, et al. DNA polymorphisms and mutations of the tumor necrosis factor-alpha (TNF-alpha) promoter in Langerhans cell histiocytosis (LCH). J. Interferon Cytokine Res. (1997) 17:631–635.[Web of Science][Medline]
  20. Barber MD, et al. A polymorphism of the interleukin-1 beta gene influences survival in pancreatic cancer. Br. J. Cancer (2000) 83:1443–1447.[CrossRef][Web of Science][Medline]
  21. El-Omar EM, et al. Interleukin-1 polymorphisms associated with increased risk of gastric cancer. Nature (2000) 404:398–402.[CrossRef][Medline]
  22. Ito LS, et al. Significant reduction in breast cancer risk for Japanese women with interleukin 1B -31 CT/TT relative to CC genotype. Jpn. J. Clin. Oncol. (2002) 32:398–402.[Abstract/Free Full Text]
  23. Machado JC, et al. Interleukin 1B and interleukin 1RN polymorphisms are associated with increased risk of gastric carcinoma. Gastroenterology (2001) 121:823–829.[CrossRef][Web of Science][Medline]
  24. Cabrera M, et al. Polymorphism in tumor necrosis factor genes associated with mucocutaneous leishmaniasis. J. Exp. Med. (1995) 182:1259–1264.[Abstract/Free Full Text]
  25. El-Omar EM, et al. Increased risk of noncardia gastric cancer associated with proinflammatory cytokine gene polymorphisms. Gastroenterology (2003) 124:1193–1201.[CrossRef][Web of Science][Medline]
  26. Machado JC, et al. A proinflammatory genetic profile increases the risk for chronic atrophic gastritis and gastric carcinoma. Gastroenterology (2003) 125:364–371.[CrossRef][Web of Science][Medline]
  27. McGuire W, et al. Severe malarial anemia and cerebral malaria are associated with different tumor necrosis factor promoter alleles. J. Infect. Dis. (1999) 179:287–290.[CrossRef][Web of Science][Medline]
  28. Mattern J, et al. Association of vascular endothelial growth factor expression with intratumoral microvessel density and tumour cell proliferation in human epidermoid lung carcinoma. Br. J. Cancer (1996) 73:931–934.[Web of Science][Medline]
  29. Takahashi Y, et al. Expression of vascular endothelial growth factor and its receptor, KDR, correlates with vascularity, metastasis, and proliferation of human colon cancer. Cancer Res. (1995) 55:3964–3968.[Abstract/Free Full Text]
  30. Havranek E, et al. An interleukin-10 promoter polymorphism may influence tumor development in renal cell carcinoma. J. Urol. (2005) 173:709–712.[CrossRef][Web of Science][Medline]
  31. Nikolova PN, et al. Association of cytokine gene polymorphisms with malignant melanoma in Caucasian population. Cancer Immunol. Immunother. (2007) 56:371–379.[CrossRef][Web of Science][Medline]
  32. Lockett KL, et al. DNA damage levels in prostate cancer cases and controls. Carcinogenesis (2006) 27:1187–1193.[Abstract/Free Full Text]
  33. Guo SW, et al. Performing the exact test of Hardy-Weinberg proportion for multiple alleles. Biometrics (1992) 48:361–372.[CrossRef][Web of Science][Medline]
  34. Steinberg DMG. MARS User's Guide (2001) San Diego, CA: Salford Systems.
  35. Cook NR, et al. Tree and spline based association analysis of gene-gene interaction models for ischemic stroke. Stat. Med. (2004) 23:1439–1453.[CrossRef][Web of Science][Medline]
  36. Schwart G. Estimating the dimension of a model. Ann. Stat. (1978) 6:461–464.[CrossRef]
  37. Hassan MI, et al. Racial differences in selected cytokine allelic and genotypic frequencies among healthy, pregnant women in North Carolina. Cytokine (2003) 21:10–16.[CrossRef][Web of Science][Medline]
  38. Cancer Genome Anatomy Project. SNP500Cancer Database (2007) http://snp500cancer.nci.nih.gov (November 2007, date last accessed).
  39. Rady PL, et al. Comprehensive analysis of genetic polymorphisms in the interleukin-10 promoter: implications for immune regulation in specific ethnic populations. Genet. Test. (2004) 8:194–203.[CrossRef][Web of Science][Medline]
  40. Zabaleta J, et al. Association of interleukin-1 beta gene polymorphisms with precancerous gastric lesions in African Americans and Caucasians. Am. J. Gastroenterol. (2006) 101:163–171.[CrossRef][Web of Science][Medline]
  41. Zabaleta J, et al. Ethnic differences in cytokine gene polymorphisms: potential implications for cancer development. Cancer Immunol. Immunother. (2008) 57:107–114.[CrossRef][Web of Science][Medline]
  42. Ahlbom A, et al. Cancer in twins: genetic and nongenetic familial risk factors. J. Natl Cancer Inst. (1997) 89:287–293.[Abstract/Free Full Text]
  43. Clevers H. At the crossroads of inflammation and cancer. Cell (2004) 118:671–674.[CrossRef][Web of Science][Medline]
  44. Coussens LM, et al. Inflammatory cells and cancer: think different! J. Exp. Med. (2001) 193:F23–F26.[Free Full Text]
  45. Coussens LM, et al. Inflammation and cancer. Nature (2002) 420:860–867.[CrossRef][Medline]
  46. Fitzpatrick FA. Inflammation, carcinogenesis and cancer. Int. Immunopharmacol. (2001) 1:1651–1667.[CrossRef][Web of Science][Medline]
  47. Itzkowitz SH, et al. Inflammation and cancer IV. Colorectal cancer in inflammatory bowel disease: the role of inflammation. Am. J. Physiol. Gastrointest. Liver Physiol. (2004) 287:G7–G17.[Abstract/Free Full Text]
  48. McCarron SL, et al. Influence of cytokine gene polymorphisms on the development of prostate cancer. Cancer Res. (2002) 62:3369–3372.[Abstract/Free Full Text]
  49. Michaud DS, et al. Genetic polymorphisms of interleukin-1B (IL-1B), IL-6, IL-8, and IL-10 and risk of prostate cancer. Cancer Res. (2006) 66:4525–4530.[Abstract/Free Full Text]
  50. Fortis C, et al. Increased interleukin-10 serum levels in patients with solid tumours. Cancer Lett. (1996) 104:1–5.[CrossRef][Web of Science][Medline]
  51. Onay VU, et al. SNP-SNP interactions in breast cancer susceptibility. BMC Cancer (2006) 6:114.[CrossRef][Medline]
  52. Ritchie MD, et al. Multifactor-dimensionality reduction reveals high-order interactions among estrogen-metabolism genes in sporadic breast cancer. Am. J. Hum. Genet. (2001) 69:138–147.[CrossRef][Web of Science][Medline]
  53. Friedman J. Multivariate adaptive regression splines. Ann. Stat. (1991) 19:1–66.[CrossRef]
  54. Veaux RDD, et al. A comparison of two nonparametric estimation schemes: MARS and neural networks. Comput. Chem. Eng. (1993) 17:819–837.[CrossRef]
  55. Lin HY, et al. Variable selection in the multivariate adaptive regression splines (MARS)-logit models for detecting gene-gene interactions. In: Proceedings of the American Statistical Associaton, Biometrics Section [CD-ROM] (2006) American Statistical Association, Alexandria, VA, pp. 256–257.
Received August 1, 2007; revised November 26, 2007; accepted November 27, 2007.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
29/3/573    most recent
bgm277v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Zabaleta, J.
Right arrow Articles by Ochoa, A. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zabaleta, J.
Right arrow Articles by Ochoa, A. C.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?