Polymorphisms in Inflammation Genes and Bladder Cancer: From Initiation to Recurrence, Progression, and Survival
http://www.100md.com
《临床肿瘤学》
the Departments of Epidemiology, Urology, Genitourinary-Medical Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, TX, and the Scott Department of Urology, Baylor College of Medicine, Houston TX
ABSTRACT
PURPOSE: Since chronic inflammation contributes to tumorigenesis, we hypothesized that the risk and clinical outcome of bladder cancer (BC) might be modulated by genetic variations in inflammation genes.
METHODS: Using the TaqMan method, we genotyped single nucleotide polymorphisms in interleukin (IL) -6 (–174 GC), IL-8 (–251 TA), tumor necrosis factor-alpha (TNF-; –308 GA), and peroxisome proliferator-activated receptor (PPARG; Pro12Ala), and determined their associations with BC initiation and clinical outcome.
RESULTS: We found that the IL-6 variant genotype (C/C) was associated with an increased BC risk (OR, 1.77; 95% CI, 1.25 to 2.51). There were joint effects between the variant IL-6 genotypes and smoking status, and between the variant genotypes of IL-6 and other genes. To assess effect on recurrence, we grouped non-muscle-invasive BC patients according to intravesical Bacillus Calmette-Guerin (BCG) treatment status: no BCG, induction BCG (iBCG), and maintenance BCG (mBCG). In the Cox proportional hazards model, the variant IL-6 genotype was associated with an increased recurrence risk (hazard ratio [HR], 4.60; 95% CI, 1.24 to 17.09) in patients receiving mBCG. The variant PPARG genotype was associated with a reduced recurrence risk (HR, 0.41; 95% CI, 0.20 to 0.86) among untreated patients. In patients with non-muscle-invasive BC, the variant IL-6 genotype was associated with an increased progression risk (HR, 1.88; 95% CI, 0.80 to 4.11). In patients with invasive BC, variant IL-6 was associated with improved 5-year overall and disease-specific survival (HR, 0.43; 95% CI, 0.19 to 0.94 and HR, 0.39; 95% CI, 0.15 to 1.00, respectively).
CONCLUSION: Inflammation gene polymorphisms are associated with modified BC risk, treatment response, and survival.
INTRODUCTION
Inflammation, a crucial, complex host defense against biologic, chemical, physical, and endogenous irritants, is usually self-limiting. However, persistent inflammation can cause cellular damage resulting in many diseases, including cancer. The functional association between inflammation and cancer dates back to Virchow,1 who in 1863 hypothesized that cancer arises in sites of inflammation because prolonged irritation, tissue injury, and activated local host response ultimately favored cell proliferation. Although it is currently understood that mere cell proliferation does not result in cancer, it is also clear that sustained cell proliferation in an environment abundant in inflammatory cells, growth factors, activated stroma, enhanced angiogenesis, and DNA-damaging agents may provide the conditions needed for tumor formation and progression. Indeed, it has been estimated that cancer is preceded by chronic inflammation in up to a third of all cases.2 Examples include inflammatory bowel disease and colorectal cancer, chronic bronchitis and lung cancer, papillomavirus infection and cervical cancer, Barrett's metaplasia and esophageal cancer, H pylori-induced gastritis and gastric cancer, chronic pancreatitis and pancreatic cancer, hepatitis B and C and liver cancer, asbestosis and mesothelioma, and persistent inflammation of the bladder, as with schistosomiasis or chronic indwelling catheters, and bladder cancer.3-8
The precise molecular events leading to malignant transformation in an environment of chronic inflammation have not been fully elucidated. Cancer cells, in order to facilitate a survival advantage, may usurp some physiologic functions of the immune system. For instance, tumor associated macrophages frequently infiltrate neoplastic tissues and provide the tumor with angiogenic and lymphangiogenic growth factors, cytokines, and proteases—all of which enhance cancer progression.9 In addition, proinflammatory signals that target the elimination of infection in the acute phase of inflammation subsequently switch their function from the killing of the intruder to tissue healing, thereby providing further growth opportunities for incipient tumors.10
Cytokines, expressed by various cancer and immune cells, bind to specific receptors and activate distinct signal pathways to transcriptionally activate a plethora of downstream factors. Interleukin (IL) -6 activates mainly the JAK1/STAT3 signal pathway and functions as a proinflammatory factor. Furthermore, high levels of IL-6 may favor a T-helper-2 (Th2) pattern of humoral immune response, which does not contribute to combating cancer.11 IL-8 activates G proteins that initiate multiple signaling cascades, including mitogen-activated protein kinase (MAPK) pathways. Under its influence, cancer cells acquire the motility that enables them to migrate, invade, and metastasize.12, 13 Members of the tumor necrosis factor (TNF) protein family function in intercellular signaling in a number of inflammatory pathways. For instance, TNF--initiated signaling pathways activate proinflammatory gene expression via the transcription factor nuclear factor-kappa B (NF-B). The NF-B family comprises a group of cytoplasmic receptors that act downstream of TNF and on activation by numerous stimuli translocate into the nucleus, where they bind to specific DNA binding sites to regulate the expression of antimicrobial peptides, cytokines, chemokines, stress-related proteins, and antiapoptotic factors.14 The peroxisome proliferator activated receptor- (PPARG), a ligand-activated transcription factor belonging to the nuclear receptor superfamily, counteracts inflammation by inhibiting the expression of TNF-, IL-6, IL-8, and other proinflammatory factors. This inhibition occurs at the transcription level, where PPARG antagonizes the AP-1, STAT, and NF-B pathways.15
Bladder cancer (BC) represents 2% of all human malignancies. It is estimated that more than 63,210 new patients will be diagnosed with BC in the United States in 2005 and 13,180 patients will die of this disease.16 Although smoking is a well-recognized risk factor for BC, direct and indirect evidence points to inflammation as another predisposing factor. The use of chronic indwelling catheters and chronic bladder inflammation are two factors associated with an increased BC risk.7 Furthermore, inflammatory infiltrates are common in the stroma of patients with BC but not in healthy control subjects.17 Variations in individual inflammatory responses could explain the considerable variability in the clinical course of disease and therapy response among patients with tumors of similar grades and stages.
Single nucleotide polymorphisms (SNP) in inflammation genes have been shown to alter their expression and functions. A G/C SNP in the promoter region (–174) of IL-6 was shown to affect transcription and alter plasma IL-6 levels.18, 19 The A-allele of an IL-8 SNP in the promoter region (T-251A) has been associated with increased IL-8 production by lipopolysaccharide-stimulated whole blood.20 A G-to-A transition in the promoter region (–308) of the TNF- gene results in higher expression of TNF- in vitro and in vivo.21 A nonsynonymous SNP of PPARG (Pro12Ala) is associated with decreased receptor activity, lower body mass index, and improved insulin sensitivity.22 Recently, the aforementioned SNPs in PPARG, IL-8, and TNF- were associated with a decreased risk for colorectal cancer, whereas the IL-6 SNP was associated with an increased risk.23
In this study, we evaluated whether polymorphisms in genes that regulate inflammatory processes alter risk for developing BC and clinical outcomes. We used a case-control study to examine SNPs in IL-6, IL-8, TNF-, and PPARG, and their associations with cancer risk. Then we followed the cases to investigate the role of these polymorphisms in recurrence risk, progression, and survival.
PATIENTS AND METHODS
Study Subjects
Between 1995 and 2003, 519 histologically confirmed incident BC patients were recruited from the Department of Urology at The University of Texas M.D. Anderson Cancer Center (Houston, TX) and from the Scott Department of Urology at Baylor College of Medicine (Houston, TX). No patient received chemotherapy or radiation before inclusion. Five hundred five control subjects without a history of cancer (except nonmelanoma skin cancer) were recruited from the Kelsey-Seybold clinics, a large multispecialty managed-care organization in the Houston, TX, metropolitan area. Cases and controls were matched by age, sex, and ethnicity. Ethnicity was defined by self-report in response to a specific question in our questionnaire. The categories used to classify ethnicity included: white, Anglo, Caucasian; Spanish Origin (Hispanic); black, African American; Asian; American Indian; and Other. We restricted our analysis to only white participants.
Epidemiology and Clinical Data Collection
In a 45-minute interview, trained M.D. Anderson staff interviewers collected data on demographics, smoking history, and family history of cancer. At the end of the interview, a 40 mL blood sample was drawn into coded heparinized tubes. Subjects who had smoked at least 100 cigarettes in their lifetimes were defined as ever smokers. Participants who had quit smoking at least 1 year before the study began were categorized as former smokers. Clinical data on tumor size, grade, stage, presence of carcinoma-in-situ, number of tumor foci, intravesical therapy, dates of recurrence and progression events, systemic chemotherapy, radical cystectomy, pathologic findings at cystectomy, and mortality were summarized from patients' charts.
According to the American Joint Committee on Cancer's 1997 TNM staging system,24 non-muscle-invasive (superficial) BC was defined as a tumor confined to the mucosa or lamina propria, and invasive BC was defined as a tumor that infiltrated the muscle layer. Patients with non-muscle-invasive BC were followed with periodical cystoscopic examinations and intravesical treatment as indicated. Patients with first tumors were diagnosed within the year before study induction. Some of the patients with non-muscle-invasive disease had a long history before referral to M.D. Anderson and they were excluded from the study. Intravesical treatment consisted of either Bacillus Calmette-Guerin (BCG) for six weekly instillations (induction course [iBCG], n = 85); BCG induction + maintenance [mBCG] in accordance with the SWOG protocol (n = 38)25; or intravesical cytotoxic agents (n = 12). Ninety-eight patients received no intravesical treatment. In patients with non-muscle-invasive BC, the study end points included tumor recurrence, defined as a newly found bladder tumor following a previous negative follow-up cystoscopy, and progression, defined as the transition from non-muscle-invasive to invasive or metastatic disease. Patients with invasive BC were offered radical cystectomy with or without systemic chemotherapy, which included one or more of the following drug combinations: methotrexate, vinblastine, doxorubicin, and cisplatin; cisplatin, gemcitabine, and ifosfamide; gemcitabine and cisplatin; ifosfamide, gemcitabine, and doxorubicin; paclitaxel, ifosfamide, and cisplatin; or paclitaxel, methotrexate, and cisplatin. To increase the power of analysis, we combined all cases treated with systemic chemotherapy. The study end points were overall and disease-specific survival rates. Because neither superficial nor invasive tumors received delayed treatment, such treatments were not treated as time-dependent covariates in the analyses.
All participants signed informed consents, and the Institutional Review Boards of the M.D. Anderson Cancer Center, Baylor College of Medicine, and Kelsey-Seybold Clinic, in accordance with the US Department of Health and Human Services, approved the study.
Molecular Analysis
Genomic DNA was extracted from peripheral blood lymphocytes by proteinase K digestion, followed by isopropanol extraction and ethanol precipitation. DNA samples were stored at –80°C. We based our SNP selection on how frequently these SNPs appeared in the population and on established effect on gene expression, generally following the methods of Landi et al,23 who have determined the impact of these SNPs in colorectal cancer. The following SNPs were selected: IL-6 (–174G>C), IL-8 (–251T>A), TNF- (–308T>A), and PPARG (pro12ala). Genotyping was performed using a 5' nuclease assay with dual fluorescent reporter probes VIC and FAM. The primer and probe sequences for each SNP can be found in the report by Landi et al.23 Reactions were completed and read in a 7900 HT TaqMan sequence detector system (Applied Biosystems). Amplification mixes (5 μL) contained sample DNA (5 ng), 1X TaqMan buffer A, deoxynucleotide triphosphates (200 μmol/L), MgCl2 (5 mmol/L), AmpliTaq Gold (0.65 units), each primer (900 nmol/L), and 200 nmol/L of each probe. The thermal cycling conditions consisted of 1 cycle for 10 minutes at 95°C, 40 cycles for 15 seconds at 95°C, and 40 cycles for 1 minute at 60°C. Sequence Detector System (SDS) version 2.1 software (Applied Biosystems, Foster City, CA) was used to analyze end point fluorescence according to the allelic discrimination technique. Water control, ample internal controls, and previously genotyped samples were included in each plate to ensure the accuracy of genotyping.
Statistical Analysis
The 2 and Fisher's exact tests were used to assess patient characteristics by genotype and to compare subjects with controls. Student's t test was used to test for differences between the subjects and controls for continuous variables. Odds ratios (ORs) were calculated as an estimate of relative risk. Unconditional multivariable logistic regression was performed where appropriate to control for possible confounding by age, sex, ethnicity, and smoking status. Kaplan-Meier plots and the log-rank test were applied using the time for events-free survival analysis, which was calculated from date of BC diagnosis to the date of end point event or the end of patient follow-up. The end points for patients with non-muscle-invasive BC included number of recurrences, time to first recurrence, and disease progression defined as recurrence with invasive tumor or need for cystectomy. The end points in patients with invasive BC included overall and disease-specific mortality. Multivariate analysis using the Cox proportional hazards model was used to assess the effect of individual SNPs on the risk for end point events. The reference hazard ratio of 1 was set for the wild type genotype. STATA 8.0 software (STATA Corp, College Station, TX) was used to perform statistical analyses. Because of the exploratory nature of our study, no attempt was made to correct for multiplicity of analyses and nominal P values were reported for all analyses. Any factors reaching nominal significance will be investigated further in future validation studies using independently collected data.
RESULTS
Subject Characteristics
Of the 519 case subjects and 505 controls enrolled onto this study, 89% were white, and therefore analysis was restricted to this subset. There were a total of 450 white control subjects and 465 white case subjects. Cigarette smoking was more common in subjects as compared with controls (74.2% ever smokers among the case subjects v 53.1% among the controls; P = .000). In addition, the average number of pack-years was greater among case subjects than among controls (42.9 pack-years v 29.3 pack-years; P = .000).
Inflammation and Cancer Risk
Overall, the variant IL-6 C/C genotype was associated with an increased BC risk (OR = 1.77; 95% CI, 1.25 to 2.51; Table 1). This association was found in ever smokers (OR = 2.07; 95% CI, 1.34 to 3.18), but not in never smokers (OR = 1.16; 95% CI, 0.63 to 2.15). We observed a joint effect between smoking status and variant IL-6 genotype, with the highest BC risk occurring in current smokers with the variant IL-6 genotype (Fig 1). Furthermore, heavy smokers (pack-years, > 29) with the variant IL-6 genotype were at greater risk for BC (OR = 2.22; 95% CI, 1.19 to 4.17) compared with light smokers (pack-years 29) with the variant genotype (OR = 1.65; 95% CI, 0.87 to 3.12). In the logistic model, the interaction (on the log-odds scale) between smoking and IL-6 variant alleles was not statistically significant (OR = 1.50; 95% CI, 0.83 to 2.69). We did not find significant associations between variant genotypes for IL-8, TNF-, and PPARG and BC risk after adjusting for age, sex, and smoking status (Table 1). However, there was evidence of gene-gene joint effects between the variant IL-6 genotypes and the variant genotypes for PPARG, TNF-, and IL-8 (Fig 2). For example, the BC risk was not increased in individuals with IL-6 CC or IL-8 (GA + AA) variant alleles (OR = 0.88, 95% CI, 0.42 to 1.87; and OR = 0.97, 95% CI, 0.66 to 1.42, respectively); however, for individuals with both IL-6 CC and IL-8 (GA + AA) variant alleles, the risk was significantly increased (OR = 2.06; 95% CI, 1.24 to 3.44). There was a significant interaction between IL-6 and IL-8 variant genotypes (OR = 2.31; 95% CI, 1.01 to 5.34 for the interaction). Although we observed joint effects between IL-6 and TNF- variant genotypes and IL-6 and PPARG variant genotypes (Fig 2), there were no statistically significant interactions for IL-6 and PPARG variant genotypes (OR = 1.68; 95% CI, 0.74 to 3.84), or for IL-6 and TNF- variant genotypes (OR = 1.18; 95% CI, 0.56 to 2.50).
Genotype Distribution and BC Stage
Of 465 patients with BC, 233 presented with non-muscle-invasive disease, and 232 presented with invasive disease. Invasive BC was associated with older patient age and longer smoking history (mean age, 65.2 years v 62.3 years, P = .007; and 32.7 years v 30.0 years of smoking, P = .057). Sex distribution, smoking status, and the number of pack-years were not statistically different between invasive and non-muscle-invasive BC patients. In the multiple logistic regression analysis, invasive BC was associated with age (OR = 1.02; 95% CI, 1.01 to 1.04) and years of smoking (OR = 1.01; 95% CI, 1.00 to 1.02). The frequency of variant genotypes was significantly higher in invasive tumors than in non-muscle-invasive tumors for PPARG (30.5% v 21.1%) and TNF- (39.9% v 26.5%), which conferred a 1.91-fold increased BC risk for TNF- and a 1.61-fold increased BC risk for PPARG (Table 2).
Genotype and Outcome in Non-Muscle-Invasive BC: Recurrence and Progression
During a median follow-up period of 20.8 months (range, 1 to 74.5 months), tumor recurrence occurred in 121 patients (62%) with non-muscle-invasive BC. When patients with non-muscle-invasive BC were not stratified according to type of intravesical BCG therapy, there was no significant difference in recurrence risk between variant and wildtype genotypes. However, when analyzed according to intravesical treatment method using the Cox proportional hazards model adjusted for age, sex, tumor grade, and smoking status, when compared with wildtype genotype (GG), the variant IL-6 genotypes were associated with an increased recurrence risk (hazard ratio [HR] = 4.31; 95% CI, 1.09 to 17.09 for GC, and HR = 5.47; 95% CI, 1.05 to 28.44 for CC) in patients receiving mBCG. When we combined the GC and CC together, the variant IL-6 genotypes were associated with an increased recurrence risk (HR = 4.60; 95% CI, 1.24 to 17.09) in patients receiving mBCG. When we put patients receiving no BCG, iBCG, or mBCG in the same Cox proportional hazards model, a borderline significant positive interaction was observed between IL-6 genotypes (GG versus GC + CC) and BCG treatment (no BCG or iBCG v mBCG), with an HR of 1.88 (95% CI, 0.95 to 3.74; P = .07). Among untreated patients, the variant PPARG genotype was associated with a reduced recurrence risk (HR = 0.41; 95% CI, 0.20 to 0.86), but this finding was not evident for treated patients (Table 3). A significantly positive interaction was observed between PPARG genotypes (CC v CG + GG) and BCG treatment (no BCG v mBCG; OR = 2.29; 95% CI, 1.21 to 4.31; P = .011).
Disease progression occurred in 37 patients (18%). A total of 15 patients (6.3%) died; seven patients as a result of BC, three as a result of treatment complications, and five from unrelated causes. Clinical factors associated with an increased risk for progression included multiple tumor foci, aggressive histologic subtypes (sarcomatoid, micropapillary, signet ring carcinoma, and small cell carcinoma), the presence of carcinoma-in-situ, and tumor size (Table 4). There was a significant over-representation of the variant IL-6 (C/C) genotype in patients with progression (41.67%) compared with patients without progression (21.43%) adjusted for age, sex, tumor grade, and smoking status (OR = 2.64; 95% CI, 1.18 to 5.95). In the Cox proportional hazards model, the variant IL-6 (C/C) genotype was associated with an HR of 1.88 for progression (95% CI, 0.80 to 4.41; Table 5).
Genotype and Outcome in Invasive BC
Of 232 patients with invasive BC, 200 patients (86.2%) underwent radical cystectomy; 164 patients had the surgery performed at the M.D. Anderson Cancer Center, and 36 patients were operated on elsewhere. Of the 32 patients who did not undergo radical cystectomy, 10 patients were diagnosed with metastatic disease at presentation, and 14 others received upfront chemotherapy and sustained disease progression or severe toxicity, which precluded proceeding with surgery. Eight other patients did not undergo radical cystectomy due to coexisting significant medical problems.
A total of 141 patients (51.3%) received chemotherapy; 67 patients and 35 patients received chemotherapy in the neoadjuvant and adjuvant settings, respectively. There were two patients who received both neoadjuvant and adjuvant chemotherapy, and 37 others to whom chemotherapy was given without surgery because of bladder preservation, metastatic disease, or disease progression and performance status decrease following chemotherapy.
During a median follow-up duration of 14.2 months (range, 1 to 151 months), 68 patients (24.7%) with invasive BC died; 38 patients died as a result of BC, four patients as a result of treatment complications (two from surgical complications and two due to chemotherapy toxicity), and 26 patients died as a result of unrelated causes. Of the 215 patients currently alive, 140 patients (65%) have no evidence of disease, 46 patients (21%) continue to live with cancer, and 29 patients (13%) have been lost to follow-up and their disease status could not be assessed. The variant IL-6 genotype was associated with improved 5-year overall survival (HR = 0.43; 95% CI, 0.19 to 0.94) and disease-specific survival (HR = 0.39; 95% CI, 0.15 to 1.00) in patients with invasive BC (Table 6; Fig 3).
DISCUSSION
This study provides evidence for the association between the variant IL-6 genotype and an overall increased BC risk. We also observed gene-environment and gene-gene interactions. Regarding the effect of genotype on disease progression, an association was found between the variant IL-6 genotypes with increased recurrence risk in patients treated with maintenance BCG treatment. The variant IL-6 genotypes, as well as the variant PPARG genotypes, were associated with an increased progression risk in non-muscle-invasive BC patients.
Inflammation and Tumor Initiation
It is now evident that a substantial proportion of cancer cases worldwide arise from infection and inflammation. Both inflammatory and tumor cells produce an assorted array of cytokines and chemokines, which mediate all aspects of inflammation and profoundly affect the development of chronic diseases, including cancer. In this study, consistent with a recent study of colorectal cancer,23 we found an association between the variant C allele of the –174G>C SNP in the IL-6 promoter region with an increased BC risk. The IL-6 variant genotype (C/C) frequency varied dramatically in literature from different ethnic groups, for example, it is very low in African Americans, Native Americans, and Asian Americans. However, in white subjects, the frequency has been fairly consistent in different studies in Europe and the United States. Most studies with relatively large sample sizes have listed the frequency between 15% and 20%. The largest study of white control subjects in the United Kingdom (2,751 subjects) gave a frequency of "CC variant" of 18%.26 Our sample size was relatively large, our population consisted of white case subjects, and our control frequency is comparable with other studies of white case subjects from different countries. Several studies have shown that the variant C allele is associated with lower IL-6 levels.18, 19, 27 It is possible that attenuated IL-6 expression following an injury may modify the host immune response in a manner that creates a survival advantage for transformed or malignant cells. In our study, a joint effect between the variant IL-6 allele and smoking with BC indicates that inflammation may play a more visible role in smoking-related BC.
Inflammation and BC Recurrence, Progression, and Survival
Carcinogenesis is a multistage process. Initiation mainly involves DNA alterations and can persist in otherwise normal tissue indefinitely until promotion occurs. Inflammatory cells produce favorable microenvironments for latent DNA damage to proceed to carcinogenesis. The protumor actions of inflammatory cells include releasing growth and survival factors, promoting angiogenesis, evading apoptosis, subverting the host immune response, and remodeling the microenvironment to facilitate tumor migration and metastasis. Therefore, alterations in inflammatory genes caused by functional SNPs may also modulate the prognosis of BC.
Intravesical Bacillus Calmette-Guerin is the prevailing choice of immunotherapy for non-muscle-invasive BC, with a 60% to 70% response rate. Induction of T-helper-type 1 (Th1) response (cellular immunity) is essential for successful BCG treatment of non-muscle-invasive BC. Th1 response typically leads to the activation of cytotoxic T lymphocytes, natural killer cells, macrophages, and monocytes, all of which fight intracellular pathogens and attack cancerous cells. Th2 response (humoral immunity), in contrast, primarily targets extracellular organisms via the upregulation of antibody production. Th1 and Th2 cells produce characteristic cytokines and can cross-inhibit each other. The dramatic increase of urinary interferon- and other Th1 cytokines (IL-2, IL-12, and so on) is associated with BCG success in BC treatment, whereas higher levels of Th2 cytokines, such as IL-6 and IL-10, are associated with BCG failure.28 Therefore, factors affecting the direction of the generated immune response to BCG (Th1 v Th2) may determine the efficacy of this therapy.29
There is evidence suggesting an intricate interplay between IL-6 and IFN- production.30 Specifically, there appears to be a bell-shaped IL-6 expression/response curve in which both low and high IL-6 concentrations hamper INF- production. Endogenously induced relatively low levels of IL-6 support a Th1 response, whereas a too high concentration favors a Th2 response. Therefore, to achieve maximal therapeutic benefit optimal IL-6 expression should follow BCG administration. The –174 G> C SNP in the IL-6 promoter region leads to insufficient IL-6, resulting in lower IFN- production and a suboptimal Th1 immune response on BCG induction—perhaps explaining the association between the variant IL-6 genotype and an increased recurrence risk in patients receiving mBCG. This association was not observed in patients with non-muscle-invasive BC who have not received BCG therapy. Considering that urinary concentration of cytokines is low before BCG exposure, the impact of decreased IL-6 expression due to the SNP would not be expected to have clinical significance.
In patients with invasive BC, the variant IL-6 allele was associated with improved overall survival and improved disease-free survival. Patients with invasive BC have higher baseline plasma IL-6 levels compared with healthy controls.31 Increased IL-6 serum level in patients with advanced cancer has been associated with poor prognosis.32-34 Decreased IL-6 expression caused by the variant allele of IL-6 may favor Th1 response, resulting in a better outcome. In fact, the variant C allele of this IL-6 SNP has been associated with improved outcome in patients with other types of advanced cancer, such as breast and ovarian cancers.35, 36
PPARG, a nuclear receptor and anti-inflammatory factor that functions by inhibiting transcription of proinflammatory factors, also affects BC outcome. The PPARG SNP (Pro12Ala) is associated with decreased receptor activity, lower body mass index, and improved insulin sensitivity.22 Decreased PPARG receptor activity due to variant alleles leads to reduced anti-inflammatory and antiproliferative activity and thus may provide favorable conditions for tumor growth. We found that the variant PPARG allele was associated with an increased risk for BC recurrence and progression in patients who received BCG therapy. In contrast, a protective effect of the variant PPARG genotype was observed in patients not treated by BCG. This opposite effect of PPARG in the environment of BCG treatment could have profound clinical significance. Patients with PPARG variant genotype could be used to predict a failure to respond to BCG. Therefore, an alternative treatment plan should be considered. The predictive value of the PPARG genotype alone is not sensitive and specific enough for the clinical setting; however, we could envision that an individualized treatment plan might be achieved if combined with other predictive markers of BCG failure in a recurrence model. Additional functional studies are necessary to explain the opposite effects of the PPARG SNP on BC progression in patients with versus without intravesical BCG treatment.
We observed that variant TNF- was more frequent among patients with invasive BC than in patients with non-muscle-invasive BC. TNF- plays a vital role in inflammation and many studies have shown that it is a tumor promoter and an active contributor to carcinogenesis.37 Expression studies have consistently shown aberrant high concentrations of TNF- in advanced tumors. The TNF- –308G >A SNP is associated with an increased expression of TNF-, which may account for the over-representation of the variant allele in invasive BC patients.
We did not find an association between the variant IL-8 allele and BC occurrence and outcome. Although there was circumstantial evidence suggesting that this IL-8 SNP (–251T>A) affects IL-8 level in lipopolysaccharide-stimulated whole blood,20 the functional significance of this SNP under physiologic and pathologic conditions remains to be determined.
In conclusion, we have shown that IL-6 plays an important role in the pathogenesis of BC and that polymorphisms in other inflammation genes may modify the outcome of this cancer. Because BCG treatment is the predominant choice for non-muscle-invasive BC, and currently there have been no predictors for the response and toxicity of this treatment, using a profile of genetic polymorphisms in inflammatory genes, either alone or in combination with other genetic markers, in order to predict clinical outcomes of BCG therapy will have great clinical significance.
Study Limitations
There are a few limitations to this study. We only studied a few selected candidate polymorphisms; pathway-based genotyping of more SNPs in inflammation genes and haplotype analysis are warranted to confirm and extend our findings. Functional studies are needed to measure phenotypes and evaluate genotype and phenotype correlation in the context of BC development. In addition, further studies are needed to better define the risk associated with IL-6 level for each disease stage and to determine if pharmacologic intervention to modulate the IL-6 level will have a therapeutic role in BC treatment. Finally, despite strong biologic plausibility and strong, independent associations, we cannot rule out that some of the associations might be false-positives as a result of inherent limitation of candidate gene approach in association studies and the effect of population heterogeneity. We attempted to limit the confounding effect of ethnicity by restricting our analysis to white subjects; however, the country of origin of ancestors among these white participants could possibly affect the frequencies of certain genotypes and cause spurious associations. Detailed information in this regard is warranted in future studies to avoid spurious findings resulting from heterogeneous ethnicity.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported by grant Nos. CA74880 and CA91846 from the National Cancer Institute, and a fellowship from the American Physician Fellowship Organization (D.L.).
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Balkwill F, Mantovani A: Inflammation and cancer: Back to Virchow? Lancet 357:539-545, 2001
Ames BN, Gold LS, Willett WC: The causes and prevention of cancer. Proc Natl Acad Sci U S A 92:5258-5265, 1995
Bargen JA, Sauer WG, Sloan WP, et al: The development of cancer in chronic ulcerative colitis. Gastroenterology 26:32-37, 1954
Campbell AH, Lee EJ: The relationship between lung cancer and chronic bronchitis. Br J Dis Chest 57:113-119, 1963
Imai T, Kubo T, Watanabe H: Chronic gastritis in Japanese with reference to high incidence of gastric carcinoma. J Natl Cancer Inst 47:179-195, 1971
Boyd DP, Norton RA: Esophagitis, hiatus hernia, and cancer. Geriatrics 23:118-121, 1968
Groah SL, Weitzenkamp DA, Lammertse DP, et al: Excess risk of bladder cancer in spinal cord injury: Evidence for an association between indwelling catheter use and bladder cancer. Arch Phys Med Rehabil 83:346-351, 2002
Rosin MP, Anwar WA, Ward AJ: Inflammation, chromosomal instability, and cancer: The schistosomiasis model. Cancer Res 54:1929S-1933S, 1994 (suppl 7)
Schoppmann SF, Birner P, Stockl J, et al: Tumor-associated macrophages express lymphatic endothelial growth factors and are related to peritumoral lymphangiogenesis. Am J Pathol 161:947-956, 2002
Nathan C: Points of control in inflammation. Nature 420:846-852, 2002
Kidd P: Th1/Th2 balance: The hypothesis, its limitations, and implications for health and disease. Altern Med Rev 8:223-246, 2003
Inoue K, Slaton JW, Eve BY, et al: Interleukin 8 expression regulates tumorigenicity and metastases in androgen-independent prostate cancer. Clin Cancer Res 6:2104-2119, 2000
Karashima T, Sweeney P, Kamat A, et al: Nuclear factor-kappaB mediates angiogenesis and metastasis of human bladder cancer through the regulation of interleukin-8. Clin Cancer Res 9:2786-2797, 2003
Hanada T, Yoshimura A: Regulation of cytokine signaling and inflammation. Cytokine Growth Factor Rev 13:413-421, 2002
Delerive P, Fruchart JC, Staels B: Peroxisome proliferator-activated receptors in inflammation control. J Endocrinol 169:453-459, 2001
Jemal A, Tiwari RC, Murray T, et al: Cancer statistics, 2005. CA Cancer J Clin 55:10-30, 2005
Cresswell J, Robertson H, Neal DE, et al: Distribution of lymphocytes of the alpha(E)beta(7) phenotype and E-cadherin in normal human urothelium and bladder carcinomas. Clin Exp Immunol 126:397-402, 2001
Fishman D, Faulds G, Jeffery R, et al: The effect of novel polymorphisms in the interleukin-6 (IL-6) gene on IL-6 transcription and plasma IL-6 levels, and an association with systemic-onset juvenile chronic arthritis. J Clin Invest 102:1369-1376, 1998
Terry CF, Loukaci V, Green FR: Cooperative influence of genetic polymorphisms on interleukin 6 transcriptional regulation. J Biol Chem 275:18138-18144, 2000
Hull J, Thomson A, Kwiatkowski D: Association of respiratory syncytial virus bronchiolitis with the interleukin 8 gene region in UK families. Thorax 55:1023-1027, 2000
Wilson AG, Symons JA, McDowell TL, et al: Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci U S A 94:3195-3199, 1997
Deeb SS, Fajas L, Nemoto M, et al: A Pro12Ala substitution in PPARgamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity. Nat Genet 20:284-287, 1998
Landi S, Moreno V, Gioia-Patricola L, et al: Association of commom polymorphisms in inflammatory genes interleukin (IL)-6, IL8, tumor necrosis factor-, NFB1 and peroxisome proliferator activated receptor- with colorectal cancer. Cancer Res 63:3560-3566, 2003
Fleming ID, Cooper JS, Hensen DE, et al: AJCC Cancer Staging Manual, 5th ed. Philadelphia, PA, Lippincott-Raven, 1997, pp 241-243
Lamm DL, Blumenstein BA, Crissman JD, et al: Maintenance bacillus Calmette-Guerin immunotherapy for recurrent TA, T1 and carcinoma in situ transitional cell carcinoma of the bladder: A randomized Southwest Oncology Group Study. J Urol 163:1124-1129, 2000
Humphries SE, Luong LA, Ogg MS, et al: The interleukin-6-174 G/C promoter polymorphism is associated with risk of coronary heart disease and systolic blood pressure in healthy men. Eur Heart J 22:2243-2252, 2001
Belluco C, Olivieri F, Bonafe M, et al: -174 G>C polymorphism of interleukin 6 gene promoter affects interleukin 6 serum level in patients with colorectal cancer. Clin Cancer Res 9:2173-2176, 2003
O'Donnell MA, Chen X, Dewolf WC: Maturation of the cytokine immune response to BCG in the bladder: Implications for treatment schedules. J Urol 155:568A, 1996 (suppl, abstr 1030)
Arnold J, de Boer EC, O'Donnell MA, et al: Immunotherapy of experimental bladder cancer with recombinant BCG expressing interferon-gamma. J Immunother 27:116-123, 2004
Luo Y, Chen X, O'Donnell MA: Role of Th1 and Th2 cytokines in BCG-induced IFN-gamma production: Cytokine promotion and simulation of BCG effect. Cytokine 21:17-26, 2003
Andrews B, Shariat SF, Kim JH, et al: Preoperative plasma levels of interleukin-6 and its soluble receptor predict disease recurrence and survival of patients with bladder cancer. J Urol 167:1475-1481, 2002
Mouawad R, Rixe O, Meric JB, et al: Serum interleukin-6 concentrations as predictive factor of time to progression in metastatic malignant melanoma patients treated by biochemotherapy: A retrospective study. Cytokines Cell Mol Ther 7:151-156, 2002
Chow KC, Chiou SH, Ho SP, et al: The elevated serum interleukin-6 correlates with the increased serum butyrate level in patients with nasopharyngeal carcinoma. Oncol Rep 10:813-819, 2003
Pfitzenmaier J, Vessella R, Higano CS, et al: Elevation of cytokine levels in cachectic patients with prostate carcinoma. Cancer 97:1211-1216, 2003
DeMichele A, Martin AM, Mick R, et al: Interleukin-6-174G->C polymorphism is associated with improved outcome in high-risk breast cancer. Cancer Res 63:8051-8056, 2003
Hefler LA, Grimm C, Ackermann S, et al: An interleukin-6 gene promoter polymorphism influences the biological phenotype of ovarian cancer. Cancer Res 63:3066-3068, 2003
Szlosarek PW, Balkwill FR: Tumour necrosis factor alpha: A potential target for the therapy of solid tumours. Lancet Oncol 4:565-573, 2003(Dan Leibovici, H. Barton )
ABSTRACT
PURPOSE: Since chronic inflammation contributes to tumorigenesis, we hypothesized that the risk and clinical outcome of bladder cancer (BC) might be modulated by genetic variations in inflammation genes.
METHODS: Using the TaqMan method, we genotyped single nucleotide polymorphisms in interleukin (IL) -6 (–174 GC), IL-8 (–251 TA), tumor necrosis factor-alpha (TNF-; –308 GA), and peroxisome proliferator-activated receptor (PPARG; Pro12Ala), and determined their associations with BC initiation and clinical outcome.
RESULTS: We found that the IL-6 variant genotype (C/C) was associated with an increased BC risk (OR, 1.77; 95% CI, 1.25 to 2.51). There were joint effects between the variant IL-6 genotypes and smoking status, and between the variant genotypes of IL-6 and other genes. To assess effect on recurrence, we grouped non-muscle-invasive BC patients according to intravesical Bacillus Calmette-Guerin (BCG) treatment status: no BCG, induction BCG (iBCG), and maintenance BCG (mBCG). In the Cox proportional hazards model, the variant IL-6 genotype was associated with an increased recurrence risk (hazard ratio [HR], 4.60; 95% CI, 1.24 to 17.09) in patients receiving mBCG. The variant PPARG genotype was associated with a reduced recurrence risk (HR, 0.41; 95% CI, 0.20 to 0.86) among untreated patients. In patients with non-muscle-invasive BC, the variant IL-6 genotype was associated with an increased progression risk (HR, 1.88; 95% CI, 0.80 to 4.11). In patients with invasive BC, variant IL-6 was associated with improved 5-year overall and disease-specific survival (HR, 0.43; 95% CI, 0.19 to 0.94 and HR, 0.39; 95% CI, 0.15 to 1.00, respectively).
CONCLUSION: Inflammation gene polymorphisms are associated with modified BC risk, treatment response, and survival.
INTRODUCTION
Inflammation, a crucial, complex host defense against biologic, chemical, physical, and endogenous irritants, is usually self-limiting. However, persistent inflammation can cause cellular damage resulting in many diseases, including cancer. The functional association between inflammation and cancer dates back to Virchow,1 who in 1863 hypothesized that cancer arises in sites of inflammation because prolonged irritation, tissue injury, and activated local host response ultimately favored cell proliferation. Although it is currently understood that mere cell proliferation does not result in cancer, it is also clear that sustained cell proliferation in an environment abundant in inflammatory cells, growth factors, activated stroma, enhanced angiogenesis, and DNA-damaging agents may provide the conditions needed for tumor formation and progression. Indeed, it has been estimated that cancer is preceded by chronic inflammation in up to a third of all cases.2 Examples include inflammatory bowel disease and colorectal cancer, chronic bronchitis and lung cancer, papillomavirus infection and cervical cancer, Barrett's metaplasia and esophageal cancer, H pylori-induced gastritis and gastric cancer, chronic pancreatitis and pancreatic cancer, hepatitis B and C and liver cancer, asbestosis and mesothelioma, and persistent inflammation of the bladder, as with schistosomiasis or chronic indwelling catheters, and bladder cancer.3-8
The precise molecular events leading to malignant transformation in an environment of chronic inflammation have not been fully elucidated. Cancer cells, in order to facilitate a survival advantage, may usurp some physiologic functions of the immune system. For instance, tumor associated macrophages frequently infiltrate neoplastic tissues and provide the tumor with angiogenic and lymphangiogenic growth factors, cytokines, and proteases—all of which enhance cancer progression.9 In addition, proinflammatory signals that target the elimination of infection in the acute phase of inflammation subsequently switch their function from the killing of the intruder to tissue healing, thereby providing further growth opportunities for incipient tumors.10
Cytokines, expressed by various cancer and immune cells, bind to specific receptors and activate distinct signal pathways to transcriptionally activate a plethora of downstream factors. Interleukin (IL) -6 activates mainly the JAK1/STAT3 signal pathway and functions as a proinflammatory factor. Furthermore, high levels of IL-6 may favor a T-helper-2 (Th2) pattern of humoral immune response, which does not contribute to combating cancer.11 IL-8 activates G proteins that initiate multiple signaling cascades, including mitogen-activated protein kinase (MAPK) pathways. Under its influence, cancer cells acquire the motility that enables them to migrate, invade, and metastasize.12, 13 Members of the tumor necrosis factor (TNF) protein family function in intercellular signaling in a number of inflammatory pathways. For instance, TNF--initiated signaling pathways activate proinflammatory gene expression via the transcription factor nuclear factor-kappa B (NF-B). The NF-B family comprises a group of cytoplasmic receptors that act downstream of TNF and on activation by numerous stimuli translocate into the nucleus, where they bind to specific DNA binding sites to regulate the expression of antimicrobial peptides, cytokines, chemokines, stress-related proteins, and antiapoptotic factors.14 The peroxisome proliferator activated receptor- (PPARG), a ligand-activated transcription factor belonging to the nuclear receptor superfamily, counteracts inflammation by inhibiting the expression of TNF-, IL-6, IL-8, and other proinflammatory factors. This inhibition occurs at the transcription level, where PPARG antagonizes the AP-1, STAT, and NF-B pathways.15
Bladder cancer (BC) represents 2% of all human malignancies. It is estimated that more than 63,210 new patients will be diagnosed with BC in the United States in 2005 and 13,180 patients will die of this disease.16 Although smoking is a well-recognized risk factor for BC, direct and indirect evidence points to inflammation as another predisposing factor. The use of chronic indwelling catheters and chronic bladder inflammation are two factors associated with an increased BC risk.7 Furthermore, inflammatory infiltrates are common in the stroma of patients with BC but not in healthy control subjects.17 Variations in individual inflammatory responses could explain the considerable variability in the clinical course of disease and therapy response among patients with tumors of similar grades and stages.
Single nucleotide polymorphisms (SNP) in inflammation genes have been shown to alter their expression and functions. A G/C SNP in the promoter region (–174) of IL-6 was shown to affect transcription and alter plasma IL-6 levels.18, 19 The A-allele of an IL-8 SNP in the promoter region (T-251A) has been associated with increased IL-8 production by lipopolysaccharide-stimulated whole blood.20 A G-to-A transition in the promoter region (–308) of the TNF- gene results in higher expression of TNF- in vitro and in vivo.21 A nonsynonymous SNP of PPARG (Pro12Ala) is associated with decreased receptor activity, lower body mass index, and improved insulin sensitivity.22 Recently, the aforementioned SNPs in PPARG, IL-8, and TNF- were associated with a decreased risk for colorectal cancer, whereas the IL-6 SNP was associated with an increased risk.23
In this study, we evaluated whether polymorphisms in genes that regulate inflammatory processes alter risk for developing BC and clinical outcomes. We used a case-control study to examine SNPs in IL-6, IL-8, TNF-, and PPARG, and their associations with cancer risk. Then we followed the cases to investigate the role of these polymorphisms in recurrence risk, progression, and survival.
PATIENTS AND METHODS
Study Subjects
Between 1995 and 2003, 519 histologically confirmed incident BC patients were recruited from the Department of Urology at The University of Texas M.D. Anderson Cancer Center (Houston, TX) and from the Scott Department of Urology at Baylor College of Medicine (Houston, TX). No patient received chemotherapy or radiation before inclusion. Five hundred five control subjects without a history of cancer (except nonmelanoma skin cancer) were recruited from the Kelsey-Seybold clinics, a large multispecialty managed-care organization in the Houston, TX, metropolitan area. Cases and controls were matched by age, sex, and ethnicity. Ethnicity was defined by self-report in response to a specific question in our questionnaire. The categories used to classify ethnicity included: white, Anglo, Caucasian; Spanish Origin (Hispanic); black, African American; Asian; American Indian; and Other. We restricted our analysis to only white participants.
Epidemiology and Clinical Data Collection
In a 45-minute interview, trained M.D. Anderson staff interviewers collected data on demographics, smoking history, and family history of cancer. At the end of the interview, a 40 mL blood sample was drawn into coded heparinized tubes. Subjects who had smoked at least 100 cigarettes in their lifetimes were defined as ever smokers. Participants who had quit smoking at least 1 year before the study began were categorized as former smokers. Clinical data on tumor size, grade, stage, presence of carcinoma-in-situ, number of tumor foci, intravesical therapy, dates of recurrence and progression events, systemic chemotherapy, radical cystectomy, pathologic findings at cystectomy, and mortality were summarized from patients' charts.
According to the American Joint Committee on Cancer's 1997 TNM staging system,24 non-muscle-invasive (superficial) BC was defined as a tumor confined to the mucosa or lamina propria, and invasive BC was defined as a tumor that infiltrated the muscle layer. Patients with non-muscle-invasive BC were followed with periodical cystoscopic examinations and intravesical treatment as indicated. Patients with first tumors were diagnosed within the year before study induction. Some of the patients with non-muscle-invasive disease had a long history before referral to M.D. Anderson and they were excluded from the study. Intravesical treatment consisted of either Bacillus Calmette-Guerin (BCG) for six weekly instillations (induction course [iBCG], n = 85); BCG induction + maintenance [mBCG] in accordance with the SWOG protocol (n = 38)25; or intravesical cytotoxic agents (n = 12). Ninety-eight patients received no intravesical treatment. In patients with non-muscle-invasive BC, the study end points included tumor recurrence, defined as a newly found bladder tumor following a previous negative follow-up cystoscopy, and progression, defined as the transition from non-muscle-invasive to invasive or metastatic disease. Patients with invasive BC were offered radical cystectomy with or without systemic chemotherapy, which included one or more of the following drug combinations: methotrexate, vinblastine, doxorubicin, and cisplatin; cisplatin, gemcitabine, and ifosfamide; gemcitabine and cisplatin; ifosfamide, gemcitabine, and doxorubicin; paclitaxel, ifosfamide, and cisplatin; or paclitaxel, methotrexate, and cisplatin. To increase the power of analysis, we combined all cases treated with systemic chemotherapy. The study end points were overall and disease-specific survival rates. Because neither superficial nor invasive tumors received delayed treatment, such treatments were not treated as time-dependent covariates in the analyses.
All participants signed informed consents, and the Institutional Review Boards of the M.D. Anderson Cancer Center, Baylor College of Medicine, and Kelsey-Seybold Clinic, in accordance with the US Department of Health and Human Services, approved the study.
Molecular Analysis
Genomic DNA was extracted from peripheral blood lymphocytes by proteinase K digestion, followed by isopropanol extraction and ethanol precipitation. DNA samples were stored at –80°C. We based our SNP selection on how frequently these SNPs appeared in the population and on established effect on gene expression, generally following the methods of Landi et al,23 who have determined the impact of these SNPs in colorectal cancer. The following SNPs were selected: IL-6 (–174G>C), IL-8 (–251T>A), TNF- (–308T>A), and PPARG (pro12ala). Genotyping was performed using a 5' nuclease assay with dual fluorescent reporter probes VIC and FAM. The primer and probe sequences for each SNP can be found in the report by Landi et al.23 Reactions were completed and read in a 7900 HT TaqMan sequence detector system (Applied Biosystems). Amplification mixes (5 μL) contained sample DNA (5 ng), 1X TaqMan buffer A, deoxynucleotide triphosphates (200 μmol/L), MgCl2 (5 mmol/L), AmpliTaq Gold (0.65 units), each primer (900 nmol/L), and 200 nmol/L of each probe. The thermal cycling conditions consisted of 1 cycle for 10 minutes at 95°C, 40 cycles for 15 seconds at 95°C, and 40 cycles for 1 minute at 60°C. Sequence Detector System (SDS) version 2.1 software (Applied Biosystems, Foster City, CA) was used to analyze end point fluorescence according to the allelic discrimination technique. Water control, ample internal controls, and previously genotyped samples were included in each plate to ensure the accuracy of genotyping.
Statistical Analysis
The 2 and Fisher's exact tests were used to assess patient characteristics by genotype and to compare subjects with controls. Student's t test was used to test for differences between the subjects and controls for continuous variables. Odds ratios (ORs) were calculated as an estimate of relative risk. Unconditional multivariable logistic regression was performed where appropriate to control for possible confounding by age, sex, ethnicity, and smoking status. Kaplan-Meier plots and the log-rank test were applied using the time for events-free survival analysis, which was calculated from date of BC diagnosis to the date of end point event or the end of patient follow-up. The end points for patients with non-muscle-invasive BC included number of recurrences, time to first recurrence, and disease progression defined as recurrence with invasive tumor or need for cystectomy. The end points in patients with invasive BC included overall and disease-specific mortality. Multivariate analysis using the Cox proportional hazards model was used to assess the effect of individual SNPs on the risk for end point events. The reference hazard ratio of 1 was set for the wild type genotype. STATA 8.0 software (STATA Corp, College Station, TX) was used to perform statistical analyses. Because of the exploratory nature of our study, no attempt was made to correct for multiplicity of analyses and nominal P values were reported for all analyses. Any factors reaching nominal significance will be investigated further in future validation studies using independently collected data.
RESULTS
Subject Characteristics
Of the 519 case subjects and 505 controls enrolled onto this study, 89% were white, and therefore analysis was restricted to this subset. There were a total of 450 white control subjects and 465 white case subjects. Cigarette smoking was more common in subjects as compared with controls (74.2% ever smokers among the case subjects v 53.1% among the controls; P = .000). In addition, the average number of pack-years was greater among case subjects than among controls (42.9 pack-years v 29.3 pack-years; P = .000).
Inflammation and Cancer Risk
Overall, the variant IL-6 C/C genotype was associated with an increased BC risk (OR = 1.77; 95% CI, 1.25 to 2.51; Table 1). This association was found in ever smokers (OR = 2.07; 95% CI, 1.34 to 3.18), but not in never smokers (OR = 1.16; 95% CI, 0.63 to 2.15). We observed a joint effect between smoking status and variant IL-6 genotype, with the highest BC risk occurring in current smokers with the variant IL-6 genotype (Fig 1). Furthermore, heavy smokers (pack-years, > 29) with the variant IL-6 genotype were at greater risk for BC (OR = 2.22; 95% CI, 1.19 to 4.17) compared with light smokers (pack-years 29) with the variant genotype (OR = 1.65; 95% CI, 0.87 to 3.12). In the logistic model, the interaction (on the log-odds scale) between smoking and IL-6 variant alleles was not statistically significant (OR = 1.50; 95% CI, 0.83 to 2.69). We did not find significant associations between variant genotypes for IL-8, TNF-, and PPARG and BC risk after adjusting for age, sex, and smoking status (Table 1). However, there was evidence of gene-gene joint effects between the variant IL-6 genotypes and the variant genotypes for PPARG, TNF-, and IL-8 (Fig 2). For example, the BC risk was not increased in individuals with IL-6 CC or IL-8 (GA + AA) variant alleles (OR = 0.88, 95% CI, 0.42 to 1.87; and OR = 0.97, 95% CI, 0.66 to 1.42, respectively); however, for individuals with both IL-6 CC and IL-8 (GA + AA) variant alleles, the risk was significantly increased (OR = 2.06; 95% CI, 1.24 to 3.44). There was a significant interaction between IL-6 and IL-8 variant genotypes (OR = 2.31; 95% CI, 1.01 to 5.34 for the interaction). Although we observed joint effects between IL-6 and TNF- variant genotypes and IL-6 and PPARG variant genotypes (Fig 2), there were no statistically significant interactions for IL-6 and PPARG variant genotypes (OR = 1.68; 95% CI, 0.74 to 3.84), or for IL-6 and TNF- variant genotypes (OR = 1.18; 95% CI, 0.56 to 2.50).
Genotype Distribution and BC Stage
Of 465 patients with BC, 233 presented with non-muscle-invasive disease, and 232 presented with invasive disease. Invasive BC was associated with older patient age and longer smoking history (mean age, 65.2 years v 62.3 years, P = .007; and 32.7 years v 30.0 years of smoking, P = .057). Sex distribution, smoking status, and the number of pack-years were not statistically different between invasive and non-muscle-invasive BC patients. In the multiple logistic regression analysis, invasive BC was associated with age (OR = 1.02; 95% CI, 1.01 to 1.04) and years of smoking (OR = 1.01; 95% CI, 1.00 to 1.02). The frequency of variant genotypes was significantly higher in invasive tumors than in non-muscle-invasive tumors for PPARG (30.5% v 21.1%) and TNF- (39.9% v 26.5%), which conferred a 1.91-fold increased BC risk for TNF- and a 1.61-fold increased BC risk for PPARG (Table 2).
Genotype and Outcome in Non-Muscle-Invasive BC: Recurrence and Progression
During a median follow-up period of 20.8 months (range, 1 to 74.5 months), tumor recurrence occurred in 121 patients (62%) with non-muscle-invasive BC. When patients with non-muscle-invasive BC were not stratified according to type of intravesical BCG therapy, there was no significant difference in recurrence risk between variant and wildtype genotypes. However, when analyzed according to intravesical treatment method using the Cox proportional hazards model adjusted for age, sex, tumor grade, and smoking status, when compared with wildtype genotype (GG), the variant IL-6 genotypes were associated with an increased recurrence risk (hazard ratio [HR] = 4.31; 95% CI, 1.09 to 17.09 for GC, and HR = 5.47; 95% CI, 1.05 to 28.44 for CC) in patients receiving mBCG. When we combined the GC and CC together, the variant IL-6 genotypes were associated with an increased recurrence risk (HR = 4.60; 95% CI, 1.24 to 17.09) in patients receiving mBCG. When we put patients receiving no BCG, iBCG, or mBCG in the same Cox proportional hazards model, a borderline significant positive interaction was observed between IL-6 genotypes (GG versus GC + CC) and BCG treatment (no BCG or iBCG v mBCG), with an HR of 1.88 (95% CI, 0.95 to 3.74; P = .07). Among untreated patients, the variant PPARG genotype was associated with a reduced recurrence risk (HR = 0.41; 95% CI, 0.20 to 0.86), but this finding was not evident for treated patients (Table 3). A significantly positive interaction was observed between PPARG genotypes (CC v CG + GG) and BCG treatment (no BCG v mBCG; OR = 2.29; 95% CI, 1.21 to 4.31; P = .011).
Disease progression occurred in 37 patients (18%). A total of 15 patients (6.3%) died; seven patients as a result of BC, three as a result of treatment complications, and five from unrelated causes. Clinical factors associated with an increased risk for progression included multiple tumor foci, aggressive histologic subtypes (sarcomatoid, micropapillary, signet ring carcinoma, and small cell carcinoma), the presence of carcinoma-in-situ, and tumor size (Table 4). There was a significant over-representation of the variant IL-6 (C/C) genotype in patients with progression (41.67%) compared with patients without progression (21.43%) adjusted for age, sex, tumor grade, and smoking status (OR = 2.64; 95% CI, 1.18 to 5.95). In the Cox proportional hazards model, the variant IL-6 (C/C) genotype was associated with an HR of 1.88 for progression (95% CI, 0.80 to 4.41; Table 5).
Genotype and Outcome in Invasive BC
Of 232 patients with invasive BC, 200 patients (86.2%) underwent radical cystectomy; 164 patients had the surgery performed at the M.D. Anderson Cancer Center, and 36 patients were operated on elsewhere. Of the 32 patients who did not undergo radical cystectomy, 10 patients were diagnosed with metastatic disease at presentation, and 14 others received upfront chemotherapy and sustained disease progression or severe toxicity, which precluded proceeding with surgery. Eight other patients did not undergo radical cystectomy due to coexisting significant medical problems.
A total of 141 patients (51.3%) received chemotherapy; 67 patients and 35 patients received chemotherapy in the neoadjuvant and adjuvant settings, respectively. There were two patients who received both neoadjuvant and adjuvant chemotherapy, and 37 others to whom chemotherapy was given without surgery because of bladder preservation, metastatic disease, or disease progression and performance status decrease following chemotherapy.
During a median follow-up duration of 14.2 months (range, 1 to 151 months), 68 patients (24.7%) with invasive BC died; 38 patients died as a result of BC, four patients as a result of treatment complications (two from surgical complications and two due to chemotherapy toxicity), and 26 patients died as a result of unrelated causes. Of the 215 patients currently alive, 140 patients (65%) have no evidence of disease, 46 patients (21%) continue to live with cancer, and 29 patients (13%) have been lost to follow-up and their disease status could not be assessed. The variant IL-6 genotype was associated with improved 5-year overall survival (HR = 0.43; 95% CI, 0.19 to 0.94) and disease-specific survival (HR = 0.39; 95% CI, 0.15 to 1.00) in patients with invasive BC (Table 6; Fig 3).
DISCUSSION
This study provides evidence for the association between the variant IL-6 genotype and an overall increased BC risk. We also observed gene-environment and gene-gene interactions. Regarding the effect of genotype on disease progression, an association was found between the variant IL-6 genotypes with increased recurrence risk in patients treated with maintenance BCG treatment. The variant IL-6 genotypes, as well as the variant PPARG genotypes, were associated with an increased progression risk in non-muscle-invasive BC patients.
Inflammation and Tumor Initiation
It is now evident that a substantial proportion of cancer cases worldwide arise from infection and inflammation. Both inflammatory and tumor cells produce an assorted array of cytokines and chemokines, which mediate all aspects of inflammation and profoundly affect the development of chronic diseases, including cancer. In this study, consistent with a recent study of colorectal cancer,23 we found an association between the variant C allele of the –174G>C SNP in the IL-6 promoter region with an increased BC risk. The IL-6 variant genotype (C/C) frequency varied dramatically in literature from different ethnic groups, for example, it is very low in African Americans, Native Americans, and Asian Americans. However, in white subjects, the frequency has been fairly consistent in different studies in Europe and the United States. Most studies with relatively large sample sizes have listed the frequency between 15% and 20%. The largest study of white control subjects in the United Kingdom (2,751 subjects) gave a frequency of "CC variant" of 18%.26 Our sample size was relatively large, our population consisted of white case subjects, and our control frequency is comparable with other studies of white case subjects from different countries. Several studies have shown that the variant C allele is associated with lower IL-6 levels.18, 19, 27 It is possible that attenuated IL-6 expression following an injury may modify the host immune response in a manner that creates a survival advantage for transformed or malignant cells. In our study, a joint effect between the variant IL-6 allele and smoking with BC indicates that inflammation may play a more visible role in smoking-related BC.
Inflammation and BC Recurrence, Progression, and Survival
Carcinogenesis is a multistage process. Initiation mainly involves DNA alterations and can persist in otherwise normal tissue indefinitely until promotion occurs. Inflammatory cells produce favorable microenvironments for latent DNA damage to proceed to carcinogenesis. The protumor actions of inflammatory cells include releasing growth and survival factors, promoting angiogenesis, evading apoptosis, subverting the host immune response, and remodeling the microenvironment to facilitate tumor migration and metastasis. Therefore, alterations in inflammatory genes caused by functional SNPs may also modulate the prognosis of BC.
Intravesical Bacillus Calmette-Guerin is the prevailing choice of immunotherapy for non-muscle-invasive BC, with a 60% to 70% response rate. Induction of T-helper-type 1 (Th1) response (cellular immunity) is essential for successful BCG treatment of non-muscle-invasive BC. Th1 response typically leads to the activation of cytotoxic T lymphocytes, natural killer cells, macrophages, and monocytes, all of which fight intracellular pathogens and attack cancerous cells. Th2 response (humoral immunity), in contrast, primarily targets extracellular organisms via the upregulation of antibody production. Th1 and Th2 cells produce characteristic cytokines and can cross-inhibit each other. The dramatic increase of urinary interferon- and other Th1 cytokines (IL-2, IL-12, and so on) is associated with BCG success in BC treatment, whereas higher levels of Th2 cytokines, such as IL-6 and IL-10, are associated with BCG failure.28 Therefore, factors affecting the direction of the generated immune response to BCG (Th1 v Th2) may determine the efficacy of this therapy.29
There is evidence suggesting an intricate interplay between IL-6 and IFN- production.30 Specifically, there appears to be a bell-shaped IL-6 expression/response curve in which both low and high IL-6 concentrations hamper INF- production. Endogenously induced relatively low levels of IL-6 support a Th1 response, whereas a too high concentration favors a Th2 response. Therefore, to achieve maximal therapeutic benefit optimal IL-6 expression should follow BCG administration. The –174 G> C SNP in the IL-6 promoter region leads to insufficient IL-6, resulting in lower IFN- production and a suboptimal Th1 immune response on BCG induction—perhaps explaining the association between the variant IL-6 genotype and an increased recurrence risk in patients receiving mBCG. This association was not observed in patients with non-muscle-invasive BC who have not received BCG therapy. Considering that urinary concentration of cytokines is low before BCG exposure, the impact of decreased IL-6 expression due to the SNP would not be expected to have clinical significance.
In patients with invasive BC, the variant IL-6 allele was associated with improved overall survival and improved disease-free survival. Patients with invasive BC have higher baseline plasma IL-6 levels compared with healthy controls.31 Increased IL-6 serum level in patients with advanced cancer has been associated with poor prognosis.32-34 Decreased IL-6 expression caused by the variant allele of IL-6 may favor Th1 response, resulting in a better outcome. In fact, the variant C allele of this IL-6 SNP has been associated with improved outcome in patients with other types of advanced cancer, such as breast and ovarian cancers.35, 36
PPARG, a nuclear receptor and anti-inflammatory factor that functions by inhibiting transcription of proinflammatory factors, also affects BC outcome. The PPARG SNP (Pro12Ala) is associated with decreased receptor activity, lower body mass index, and improved insulin sensitivity.22 Decreased PPARG receptor activity due to variant alleles leads to reduced anti-inflammatory and antiproliferative activity and thus may provide favorable conditions for tumor growth. We found that the variant PPARG allele was associated with an increased risk for BC recurrence and progression in patients who received BCG therapy. In contrast, a protective effect of the variant PPARG genotype was observed in patients not treated by BCG. This opposite effect of PPARG in the environment of BCG treatment could have profound clinical significance. Patients with PPARG variant genotype could be used to predict a failure to respond to BCG. Therefore, an alternative treatment plan should be considered. The predictive value of the PPARG genotype alone is not sensitive and specific enough for the clinical setting; however, we could envision that an individualized treatment plan might be achieved if combined with other predictive markers of BCG failure in a recurrence model. Additional functional studies are necessary to explain the opposite effects of the PPARG SNP on BC progression in patients with versus without intravesical BCG treatment.
We observed that variant TNF- was more frequent among patients with invasive BC than in patients with non-muscle-invasive BC. TNF- plays a vital role in inflammation and many studies have shown that it is a tumor promoter and an active contributor to carcinogenesis.37 Expression studies have consistently shown aberrant high concentrations of TNF- in advanced tumors. The TNF- –308G >A SNP is associated with an increased expression of TNF-, which may account for the over-representation of the variant allele in invasive BC patients.
We did not find an association between the variant IL-8 allele and BC occurrence and outcome. Although there was circumstantial evidence suggesting that this IL-8 SNP (–251T>A) affects IL-8 level in lipopolysaccharide-stimulated whole blood,20 the functional significance of this SNP under physiologic and pathologic conditions remains to be determined.
In conclusion, we have shown that IL-6 plays an important role in the pathogenesis of BC and that polymorphisms in other inflammation genes may modify the outcome of this cancer. Because BCG treatment is the predominant choice for non-muscle-invasive BC, and currently there have been no predictors for the response and toxicity of this treatment, using a profile of genetic polymorphisms in inflammatory genes, either alone or in combination with other genetic markers, in order to predict clinical outcomes of BCG therapy will have great clinical significance.
Study Limitations
There are a few limitations to this study. We only studied a few selected candidate polymorphisms; pathway-based genotyping of more SNPs in inflammation genes and haplotype analysis are warranted to confirm and extend our findings. Functional studies are needed to measure phenotypes and evaluate genotype and phenotype correlation in the context of BC development. In addition, further studies are needed to better define the risk associated with IL-6 level for each disease stage and to determine if pharmacologic intervention to modulate the IL-6 level will have a therapeutic role in BC treatment. Finally, despite strong biologic plausibility and strong, independent associations, we cannot rule out that some of the associations might be false-positives as a result of inherent limitation of candidate gene approach in association studies and the effect of population heterogeneity. We attempted to limit the confounding effect of ethnicity by restricting our analysis to white subjects; however, the country of origin of ancestors among these white participants could possibly affect the frequencies of certain genotypes and cause spurious associations. Detailed information in this regard is warranted in future studies to avoid spurious findings resulting from heterogeneous ethnicity.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported by grant Nos. CA74880 and CA91846 from the National Cancer Institute, and a fellowship from the American Physician Fellowship Organization (D.L.).
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Balkwill F, Mantovani A: Inflammation and cancer: Back to Virchow? Lancet 357:539-545, 2001
Ames BN, Gold LS, Willett WC: The causes and prevention of cancer. Proc Natl Acad Sci U S A 92:5258-5265, 1995
Bargen JA, Sauer WG, Sloan WP, et al: The development of cancer in chronic ulcerative colitis. Gastroenterology 26:32-37, 1954
Campbell AH, Lee EJ: The relationship between lung cancer and chronic bronchitis. Br J Dis Chest 57:113-119, 1963
Imai T, Kubo T, Watanabe H: Chronic gastritis in Japanese with reference to high incidence of gastric carcinoma. J Natl Cancer Inst 47:179-195, 1971
Boyd DP, Norton RA: Esophagitis, hiatus hernia, and cancer. Geriatrics 23:118-121, 1968
Groah SL, Weitzenkamp DA, Lammertse DP, et al: Excess risk of bladder cancer in spinal cord injury: Evidence for an association between indwelling catheter use and bladder cancer. Arch Phys Med Rehabil 83:346-351, 2002
Rosin MP, Anwar WA, Ward AJ: Inflammation, chromosomal instability, and cancer: The schistosomiasis model. Cancer Res 54:1929S-1933S, 1994 (suppl 7)
Schoppmann SF, Birner P, Stockl J, et al: Tumor-associated macrophages express lymphatic endothelial growth factors and are related to peritumoral lymphangiogenesis. Am J Pathol 161:947-956, 2002
Nathan C: Points of control in inflammation. Nature 420:846-852, 2002
Kidd P: Th1/Th2 balance: The hypothesis, its limitations, and implications for health and disease. Altern Med Rev 8:223-246, 2003
Inoue K, Slaton JW, Eve BY, et al: Interleukin 8 expression regulates tumorigenicity and metastases in androgen-independent prostate cancer. Clin Cancer Res 6:2104-2119, 2000
Karashima T, Sweeney P, Kamat A, et al: Nuclear factor-kappaB mediates angiogenesis and metastasis of human bladder cancer through the regulation of interleukin-8. Clin Cancer Res 9:2786-2797, 2003
Hanada T, Yoshimura A: Regulation of cytokine signaling and inflammation. Cytokine Growth Factor Rev 13:413-421, 2002
Delerive P, Fruchart JC, Staels B: Peroxisome proliferator-activated receptors in inflammation control. J Endocrinol 169:453-459, 2001
Jemal A, Tiwari RC, Murray T, et al: Cancer statistics, 2005. CA Cancer J Clin 55:10-30, 2005
Cresswell J, Robertson H, Neal DE, et al: Distribution of lymphocytes of the alpha(E)beta(7) phenotype and E-cadherin in normal human urothelium and bladder carcinomas. Clin Exp Immunol 126:397-402, 2001
Fishman D, Faulds G, Jeffery R, et al: The effect of novel polymorphisms in the interleukin-6 (IL-6) gene on IL-6 transcription and plasma IL-6 levels, and an association with systemic-onset juvenile chronic arthritis. J Clin Invest 102:1369-1376, 1998
Terry CF, Loukaci V, Green FR: Cooperative influence of genetic polymorphisms on interleukin 6 transcriptional regulation. J Biol Chem 275:18138-18144, 2000
Hull J, Thomson A, Kwiatkowski D: Association of respiratory syncytial virus bronchiolitis with the interleukin 8 gene region in UK families. Thorax 55:1023-1027, 2000
Wilson AG, Symons JA, McDowell TL, et al: Effects of a polymorphism in the human tumor necrosis factor alpha promoter on transcriptional activation. Proc Natl Acad Sci U S A 94:3195-3199, 1997
Deeb SS, Fajas L, Nemoto M, et al: A Pro12Ala substitution in PPARgamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity. Nat Genet 20:284-287, 1998
Landi S, Moreno V, Gioia-Patricola L, et al: Association of commom polymorphisms in inflammatory genes interleukin (IL)-6, IL8, tumor necrosis factor-, NFB1 and peroxisome proliferator activated receptor- with colorectal cancer. Cancer Res 63:3560-3566, 2003
Fleming ID, Cooper JS, Hensen DE, et al: AJCC Cancer Staging Manual, 5th ed. Philadelphia, PA, Lippincott-Raven, 1997, pp 241-243
Lamm DL, Blumenstein BA, Crissman JD, et al: Maintenance bacillus Calmette-Guerin immunotherapy for recurrent TA, T1 and carcinoma in situ transitional cell carcinoma of the bladder: A randomized Southwest Oncology Group Study. J Urol 163:1124-1129, 2000
Humphries SE, Luong LA, Ogg MS, et al: The interleukin-6-174 G/C promoter polymorphism is associated with risk of coronary heart disease and systolic blood pressure in healthy men. Eur Heart J 22:2243-2252, 2001
Belluco C, Olivieri F, Bonafe M, et al: -174 G>C polymorphism of interleukin 6 gene promoter affects interleukin 6 serum level in patients with colorectal cancer. Clin Cancer Res 9:2173-2176, 2003
O'Donnell MA, Chen X, Dewolf WC: Maturation of the cytokine immune response to BCG in the bladder: Implications for treatment schedules. J Urol 155:568A, 1996 (suppl, abstr 1030)
Arnold J, de Boer EC, O'Donnell MA, et al: Immunotherapy of experimental bladder cancer with recombinant BCG expressing interferon-gamma. J Immunother 27:116-123, 2004
Luo Y, Chen X, O'Donnell MA: Role of Th1 and Th2 cytokines in BCG-induced IFN-gamma production: Cytokine promotion and simulation of BCG effect. Cytokine 21:17-26, 2003
Andrews B, Shariat SF, Kim JH, et al: Preoperative plasma levels of interleukin-6 and its soluble receptor predict disease recurrence and survival of patients with bladder cancer. J Urol 167:1475-1481, 2002
Mouawad R, Rixe O, Meric JB, et al: Serum interleukin-6 concentrations as predictive factor of time to progression in metastatic malignant melanoma patients treated by biochemotherapy: A retrospective study. Cytokines Cell Mol Ther 7:151-156, 2002
Chow KC, Chiou SH, Ho SP, et al: The elevated serum interleukin-6 correlates with the increased serum butyrate level in patients with nasopharyngeal carcinoma. Oncol Rep 10:813-819, 2003
Pfitzenmaier J, Vessella R, Higano CS, et al: Elevation of cytokine levels in cachectic patients with prostate carcinoma. Cancer 97:1211-1216, 2003
DeMichele A, Martin AM, Mick R, et al: Interleukin-6-174G->C polymorphism is associated with improved outcome in high-risk breast cancer. Cancer Res 63:8051-8056, 2003
Hefler LA, Grimm C, Ackermann S, et al: An interleukin-6 gene promoter polymorphism influences the biological phenotype of ovarian cancer. Cancer Res 63:3066-3068, 2003
Szlosarek PW, Balkwill FR: Tumour necrosis factor alpha: A potential target for the therapy of solid tumours. Lancet Oncol 4:565-573, 2003(Dan Leibovici, H. Barton )