Abstract
Prostate cancer is the second most frequent cancer in men. The frequency of the prostate cancer is greatly varies in different populations of the world. Three common polymorphisms in promoter region of interleukin-10 (IL-10) gene viz. -1082 A>G, -819 C>T and -592 C>A are extensively studied in prostate cancer with inconclusive results. So, a meta-analysis was performed to assess the association between these three IL-10 gene polymorphisms and risk of prostate cancer susceptibility. Suitable studies were retrieved by electronic databases search and odds ratios (ORs) with 95% confidence intervals (CIs) were used as association measure. All the statistical analyses were conducted in the Open Meta-Analyst program.
In our meta-analysis we included 17 studies (10,718 samples), 11 studies (8,391 samples) and 13 studies (7,801 samples) for -1082 A>G, -819 C>T and -592 C>A polymorphisms respectively. The result of the -592 C>A polymorphism revealed low heterogeneity with no association in the overall analysis (ORAvs.C= 1.05, 95% CI= 0.99-1.12, p= 0.09, I2= 35.89%). In ethnicity based stratified analyses, significant association was found in Caucasian population with prostate cancer using allele contrast model (ORAvs.C= 1.08, 95% CI= 1.01-1.16, p= 0.02, I2= 19.37%), homozygote model (ORAAvs.CC= 1.24, 95% CI= 1.00-1.52, p= 0.04, I2= 0%), and dominant model (ORAA+CAvs.CC= 1.10, 95% CI= 1.00-1.22, p= 0.05, I2= 32.57%). No such results were found in the Asian population. In the other two polymorphisms i.e. -1082 A>G and -819 C>T, no significant association with prostate cancer was observed.
In conclusion, results of present meta-analysis suggested that IL-10 -592 C>A polymorphism plays a role in the progression of the prostate cancer in the Caucasian population.
Introduction
Prostate cancer is the second most common cancer and fifth leading cause of deaths in males worldwide. The worldwide incidence and mortality rate of prostate cancer is 13.5% and 6.7% respectively. The incidence of prostate cancer greatly varies in different regions of the world (highest in Australia/New Zealand and lowest in South Central Asia). While the mortality is highest in Southern Africa and lowest in South Central Asia [1]. The progression of prostate cancer is often slow and it remains localized firstly and later when the cells of the prostatic tissue abnormally proliferate it may spread to nearby tissues and organs and then metastasizes. The etiology of the prostate is least understood despite being a common cancer. A few genetic factors were identified which are consider as the risk for the prostate cancer including age, family history of prostate cancer and ethnicity. Smoking, diet, androgen, and obesity are the secondary factors that may contribute to the risk of prostate cancer [2].
Recently inflammation has been added to the list of secondary factors as it contributes to proliferation, malignancy, angiogenesis, metastasis, modulation of adaptive immunity, and unresponsiveness to hormones and chemotherapeutic agents [3]. Interleukin-10 (IL-10) is an anti-inflammatory cytokine primarily produced by monocytes and to a lesser amount by lymphocytes like-TH2, CD4+ and CD8+ T cells, macrophages, and dendritic cells [4]. IL-10 down regulates the cytokine production of the TH1 cells [5]. It enhances the proliferation of B cells, thymocytes, and mast cells and also stimulates the antibody production [6, 7].
The IL-10 cytokine in humans is coded by interlukin-10 (IL-10) gene, which is located on the chromosome 1 (1q31-32) and contains 5 exons [8]. Numerous polymorphisms are reported on the IL-10 gene but clinically important three polymorphisms on the promoter region viz. - 1082A>G (rs1800896), -819C>T (rs1800871) and -592C>A (rs1800872) have been extensively studied in different types of cancers like-breast [9], lung [10], gastric [11], cervical [12, 13], and bladder cancer [14].
IL-10 may possibly increase tumor development, as it suppresses the anti-tumor immune response. Involvement of IL-10 polymorphisms in relation to cancer development is controversial as it has both immune-suppressive as well as anti-angiogenic properties [15]. The role of IL-10 in prostate cancer susceptibility and tumor development is still not clear [16, 17]. A number of association studies were published which evaluated the role of IL-10 gene polymorphisms in prostate cancer susceptibility [3, 18-23]. These studies investigated the role of IL-10 gene polymorphisms in the prostate cancer risk with contradicted results. The reason might be due to the limited number of samples or due to the other factors like ethnicities. Therefore, authors performed a meta-analysis to check the association of these three IL-10 gene polymorphisms with the prostate cancer susceptibility.
Materials and methods
Study selection
For the present study, we searched PubMed, Google scholar, and Springer Link databases up to July, 2020. The following keyword were used for the literature search “prostate cancer” along with the combination of “IL-10”, “Interleukin 10”, “IL-10 -1082”, “IL-10 -819”, “IL-10 - 592”. Retrieved articles were searched manually for the additional citations.
Inclusion and exclusion criteria
The inclusion criteria are-i) case-control study; ii) provided number of alleles and/or genotypes in both case and controls. We excluded the studies which are i) only case reports; ii) not providing enough data to calculate OR and 95%CI; iii) conducted on the animal model; iv) containing overlapping data; and v) review article, editorial, letter to editor etc.
Date extraction
From all the included articles the following information were retrieved-i) first author’s name; ii) year of publication; iii) ethnicity; iv) country of study; v) number of cases and controls; and vi) number of alleles and genotypes in both cases and controls. All this information were extracted by the authors individually and if any discrepancy were found it was resolved by the discussion.
Meta-analysis
The odds ratio (OR) with 95% confidence interval (CI) were calculated for the five genetic models viz. allele contrast model, dominant model, homozygote model, heterozygote model and recessive model to check the association between prostate cancer risk and IL-10 gene polymorphisms. The between study heterogeneity was checked by the Q-statistics (p< 0.05 was considered as statistically significant heterogeneity) and I2 statistic was used to quantify the inconsistency between study estimates. When a significant heterogeneity was found random effect model [24] was used otherwise fixed effect model [25] was applied. Further analyses were performed on the basis of sub-group analysis based on ethnicity. The evaluation of the symmetry of the funnel plot was used for the identification of the publication bias. Egger’s linear regression method was used for further evaluation of the publication bias [26]. All the statistical analyses were performed in the Open Meta-Analyst program [27]. All p-values were two-tailed with a significance level of 0.05.
Results
For the present meta-analysis, we followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guideline [28]. Flow chart of article selection was shown in Fig. 1 with specific reasons. Eighteen studies were found to be eligible for inclusion in the present meta-analysis after applying the inclusion and exclusion criteria [3, 18-23, 29-39].
Eligible studies
For IL-10 -1082
A total of 16 studies were included in IL-10 -1082 meta-analysis [3, 18, 20, 22, 23, 29-39]. The study conducted by Zabaleta et al. (2008)20 analyzed two different populations, so we considered them as separate studies. Hence, total 17 studies with 4,937 cases and 5,781 controls were included in this meta-analysis. In the study of Xu et al. [18] only the number of alleles were given. Out of 17 studies, subjects of four studies were Asian, subjects of 11 studies were Caucasian, subjects of one study was Afro-American and another one study subjects was of mixed ethnicity.
For IL-10 -819
Total 10 studies were retrieved and included in IL-10 -819 meta-analysis (20-23, 29, 30, 32, 34-36). The study conducted by Zabaleta et al. [20] contained two different populations so, we considered them as separate studies. Hence, we have total 11 studies with 3,743 cases and 4,648 controls. Out of selected 11 selected studies, four studies were Asian, five from Caucasian, one Afro-American and one of mixed ethnicity.
For IL-10 -592
Twelve studies were found eligible for IL-10 -592 meta-analysis [18-23, 30, 33-37]. The study conducted by Zabaleta et al. [20] contained two different populations so we considered them as separate studies. Total 13 studies with 3,771 cases and 4,030 controls were included in the meta-analysis. In the study of Xu et al. [18] only the number of alleles were provided. Out of selected 13 studies, three were Asian, nine from Caucasian, and one belongs to Afro-American ethnicity.
Meta-analysis
For IL-10 -1082
High heterogeneity was found in allele contrast model with no statistically significant association with prostate cancer (ORGvs.A= 0.98, 95% CI= 0.87-1.10, p= 0.76, I2= 75.48%) (Fig. 2, Table 1). Insignificant association were found in all other four genetic models- for dominant model (GG+AG vs. AA) OR= 0.99, 95% CI= 0.81-1.20, p= 0.94, I2= 73.54%; for homozygote model (GG vs. AA) OR= 0.93, 95%CI= 0.72-1.22, p= 0.63, I2= 73.49; for co-dominant model (AG vs. AA) OR= 1.01, 95%CI= 0.84-1.21, p= 0.89, I2= 65.15%; and for recessive model (AA+AG vs. GG) OR= 1.05, 95%CI= 0.86-1.28, p= 0.62, I2= 66.71. High heterogeneity was found in all the studies so random effect model was applied (Table 1).
In subgroup analysis, no significant association was found in any ethnic population using five genetic models and heterogeneity was high in all genetic models (Table 1).
For IL-10 -819
Insignificant association with higher heterogeneity was found in the allele contrast model (ORTvs.C= 1.00, 95%CI= 0.89-1.12, p= 0.98; I2= 53.51%) (Fig. 3, Table 2). Similarly, no significant association was observed in any genetic model- for dominant model (TT+CT vs. CC: OR= 1.00, 95%CI= 0.83-1.20, p= 0.96, I2= 65.37%), for homozygote model (TT vs. CC: OR= 0.92, 95%CI= 0.79-1.08, p= 0.36, I2= 10.71), for co-dominant model (CT vs. CC: OR= 1.01, 95%CI= 0.82-1.23, p= 0.91, I2= 66.8%), and for recessive model (CC+CT vs. TT: OR= 1.03, 95%CI= 0.90-1.19, p= 0.60, I2= 0%). In the sub-group meta-analyses, no significant association was found in any genetic model in any sub-group (Table 2).
For IL-10 -592
Non-significant association was found in the allele contrast model with low heterogeneity (ORAvs.C= 1.05, 95%CI= 0.99-1.12, p= 0.09, I2= 35.89) (Fig. 4, Table 3). Similarly no significant association with low heterogeneity was observed in all genetic models- for dominant model (AA+CA vs. CC: OR= 1.07, 95%CI= 0.97-1.17, p= 0.14, I2= 23.57%), for homozygote model (AA vs. CC: OR= 1.10, 95%CI= 0.93-1.30, p= 0.26, I2= 30.19%), for co-dominant model (CA vs. CC: OR= 1.05, 95%CI= 0.96-1.16, p= 0.24, I2= 22.36%) and for recessive model (CC+CA vs. AA: OR= 0.93, 95%CI= 0.79-1.09, p= 0.37, I2= 37.04%) (Table 3).
Ethnicity based sub-group analyses were performed and no statistically significant association was found in any genetic model in the Asian sub-group (Table 3). However, in the Caucasian sub-group meta-analysis, marginal association was observed in the allele contrast model (ORAvs.C= 1.08, 95%CI= 1.01-1.16, p= 0.02, I2= 19.37%). Similar association was also found in the dominant model (AA+CA vs. CC) OR= 1.10, 95%CI= 1.00-1.22, p= 0.05, I2= 32.57%; and homozygote model (AA vs. CC) OR= 1.24, 95%CI= 1.00-1.52, p= 0.04, I2= 0% (Fig. 5, Table 3). But no association was observed in the other two models i.e. co-dominant (CA vs. CC) and recessive (CC+CA vs. AA) models (Table 3).
Sensitivity analysis
The sensitivity analysis was performed by the removal of the studies in which the control population were not in the Hardy-Weinberg equilibrium (HWE). For IL-10 -1082 four studies were deviated from the HWE [31, 20 (Afro-American), 32, 35]. Removal of these studies has no effect on the overall meta-analysis or subgroup analyses. For IL-10 -819 two studies were not in HWE [22, 36]. When we performed meta-analysis after removal of these studies no effect was observed in overall meta-analysis or sub-group meta-analyses. For IL-10 -592 only one study [22] was deviated from HWE. When this study was removed from the analysis, we found that the heterogeneity was decreased and a marginally significant association was observed in the allele contrast model (ORAvs.C= 1.07, 95%CI= 1.00-1.14, p= 0.03, I2= 19.37%). While no such effect was found in any other genetic model in overall as well as in the sub-group meta-analyses.
Publication bias
Publication bias was assessed by the evaluation of the funnel plot and a symmetry in the funnel plot demonstrates lack of publication bias. Symmetrical funnel plots were found in all genetic models (Fig. 6). To confirm this statistically, we performed the Egger’s test and the p-value of Egger’s test is greater than 0.05 in all the genetic models in all the selected polymorphisms both in overall as well as in sub-group analyses (Table 1-3).
Discussion
In the present study, we examined the role of three common promoter polymorphisms (−1082A>G, -819C>T and -592C>A) of the IL-10 gene in the susceptibility of the prostate cancer. The result of the present study indicates that the -592 polymorphism of the IL-10 gene shows significant association with the prostate cancer risk in the Caucasian population in allele contrast (ORAvs.C= 1.08, 95%CI= 1.01-1.16, p= 0.02, I2= 19.37%), dominant model (ORAA+CAvs.CC= 1.10, 95%CI= 1.00-1.22, p= 0.05, I2= 32.57%) and homozygote model (ORAAvs.CC= 1.24, 95%CI= 1.00-1.52, p= 0.04, I2= 0%). No significant association was observed in overall meta-analysis or in the sub-group analyses in the other two polymorphisms.
IL-10 cytokine demonstrates both anti-inflammatory and anti-angiogenic properties. The anti-inflammatory mechanisms of IL-10 are suspected to have pro-tumorigenic potential by allowing cancerous cells to avoid immune surveillance. The IL-10 also have anti-angiogenic properties and it decreases the tumor growth and angiogenesis both in animals and in in-vitro models [30]. The expression of the IL-10 cytokine is influenced by the promoter polymorphisms (−1082A>G, -819C>T and -592C>A). Reduced level of IL-10 is reported in the presence of IL-10 -1082G/-819T/-592A haplotype [19, 40, 41]. Several studies were published examining the role of IL-10 gene polymorphisms and susceptibility to prostate cancer but the results were inconclusive. The reasons for this is not clear however, several possibilities may exist. The first reason might be due to the differences in the distribution of the IL-10 gene polymorphisms amongst the different ethnic populations. In addition, the prostate cancer is a multi-factorial disease and the exposure of various environmental factors may influence the expression of some genes e.g. the interaction of promoter genetic variants of IL-10 gene with exposure to tobacco may affect the severity as well as susceptibility of prostate cancer [42].
During literature search we found five meta-analyses, which examined the role of IL-10 gene polymorphisms in the susceptibility of prostate cancer [17, 43-46]. Shao et al. [43] have found modest association of -819 C>T (ORTvs.C= 1.16, 95% CI= 1.03-1.30, p= 0.01) and -592 C>A (ORAvs.C= 1.13, 95% CI= 1.01-1.26, p= 0.03) polymorphisms with prostate cancer. Three meta-analyses [17, 44, 45], did not report any association between IL-10 gene polymorphisms and prostate cancer risk. Men et al. [46] included 17 studies in their meta-analysis and reported marginally significant association of -1082A>G (OR=1.10, 95% CI= 1.00-1.21, p= 0.05, I2= 64.3%). In the present meta-analysis, we included three polymorphisms (−1082A>G, -819C>T and -592C>A) with 18 studies and found no association between these polymorphisms and prostate cancer. Further analysis were performed on the basis of the ethnicity and we found that the -592C>A polymorphism statistically associated with the risk of prostate cancer in the Caucasian population.
Meta-analysis is an statistical tool for combining the studies with low sample size and low power. In recent years meta-analysis is frequently used to check the association between alleles/genotypes or some environmental factors with some diseases/disorders. In past few years a number of meta-analyses were published like – recurrent pregnancy loss [47], Down syndrome [48, 49], neural tube defects [50], Glucose 6-phosphate dehydrogenase deficiency [51], osteoporosis [52], epilepsy [53], bipolar disorder [54], schizophrenia [55, 56], Alzheimer [57], breast cancer [58-60], colorectal cancer [61], digestive tract cancer [62], esophageal cancer [63], and prostate cancer [64].
Our meta-analysis has few advantages over the previously published meta-analyses such as – i) this is the largest study on the basis of number of included studies (18 studies) and sample size; ii) included three polymorphisms (−1082A>G, -819C>T and -592C>A); iii) absence of publication bias; and iv) included studies from all the ethnicities. At the same time we want to narrate few limitations associated with our meta-analysis like – i) included only studies published in English; ii) only genetic factors were considered; and iii) crude odds ratio was used.
Conclusions
The result of our meta-analysis showed that the -592C>A polymorphism is significantly associated with the risk of prostate cancer in the Caucasian population, while no such effect is found in other ethnicities or in the other gene polymorphisms. For the future studies we suggest that one should also measure the level of IL-10 expression because these polymorphisms significantly influences the level of IL-10 cytokine. Furthermore, if gene-environmental interaction is considered in the future studies it would be an added advantage to elucidate the clear picture of the IL-10 gene polymorphisms in the development of the prostate cancer.
Data Availability
All the data is given in the manuscript.
Acknowledgements
Upendra Yadav is highly grateful to the VBS Purvanchal University, Jaunpur, for providing financial assistance to him in the form of PDF.