Association of three promoter polymorphisms in interleukin-10 gene with cancer susceptibility in the Chinese population: a meta-analysis

Numerous studies have examined the associations of three promoter polymorphisms (-1082A/G, -819T/C and -592A/C) in IL-10 gene with cancer susceptibility in the Chinese population, but the results remain inconclusive. To gain a more precise estimation of this potential association, we conducted the current meta-analysis based on 53 articles, including 26 studies with 4,901 cases and 6,426 controls for the -1082A/G polymorphism, 33 studies with 6,717 cases and 8,550 controls for the -819T/C polymorphism, and 42 studies with 9,934 cases and 13,169 controls for the -592A/C polymorphism. Pooled results indicated that the three promoter polymorphisms in IL-10 gene were significantly associated with an increased overall cancer risk in the Chinese population. Stratification analysis showed that the association was more pronounced for hepatocellular carcinoma and low quality studies for the -1082A/G polymorphism, lung cancer and oral cancer for the -819T/C polymorphism. However, the -592A/C polymorphism was associated with a statistically significant increased risk for lung cancer, oral cancer, hospital-based studies and low quality studies, but a decreased risk for colorectal cancer. We further investigated the significant results using the false-positive report probability (FPRP) test. Interestingly, FPRP test results revealed that only IL-10 -1082A/G polymorphism was truly associated with an increased overall cancer risk. In the subgroup analysis, only the low quality studies, lung cancer and colorectal cancer remained significant at the prior level of 0.1. Although this association needs further confirmation by considering large studies, this meta-analysis suggested an association between IL-10 gene polymorphisms and cancer risk in the Chinese population.


INTRODUCTION
Cancer is still a global public health problem. According to the GLOBOCAN estimates, about 14.1 million new cancer cases and 8.2 million deaths occurred in 2012 worldwide [1]. In China, cancer has become the leading cause of death since 2010, with an estimate of 4292,000 new cancer cases and 2814,000 cancer deaths in 2015 [2]. As a multifactorial disease, it involves both genetic and environmental factors [3]. Accumulating evidence has indicated that inflammation plays a vital role in cancer development [4][5][6], and approximately 20% of all cancers are associated with chronic inflammation [7].
The IL-10 gene is located on chromosome 1q31-32, and is composed of five exons and four introns. IL-10 gene promoter region is highly polymorphic, and three promoter single nucleotide polymorphisms (SNPs) such as -1082A/G (rs1800896), -819T/C (rs1800871) and -592A/C (rs1800872) have been reported to regulate IL-10 expression [11,12] and alter the susceptibility to various types of cancers [13][14][15][16]. In the Chinese population, numerous case-control studies were performed to investigate the role of IL-10 -1082A/G, -819T/ C and -592A/C polymorphisms in cancer risk. However, the results remain inconclusive. Hence, we performed the present meta-analysis to investigate the association between three polymorphisms in IL-10 gene and cancer susceptibility in the Chinese population.

Meta-analysis results
The main results regarding the association between IL-10 -1082A/G polymorphism and cancer risk are shown in Table 2  The overall results regarding the association between IL-10 -819T/C polymorphism and cancer risk are shown in Table 2. A significant association was found between IL-10 -819T/C polymorphism and overall cancer risk (homozygous: OR = 1.19, 95% CI = 1.00-1.41, P < 0.001; recessive: OR = 1.17, 95% CI = 1.00-1.36, P < 0.001 and allele comparison: OR = 1.08, 95% CI = 1.00-1.18, P < 0.001). In the subgroup analysis, a statistically significant association was found for lung cancer The detailed results regarding the association between IL-10 -592A/C polymorphism and cancer risk are shown in Table 2. A significant association was found between IL-10 -592A/C polymorphism and increased overall cancer risk (homozygous: OR = 1.13, 95% CI = 1.00-1.28, P = 0.001). In the subgroup analysis, a statistically significant increased risk was found for lung cancer (homozygous: OR = 1.

Heterogeneity and sensitivity analysis
Substantial heterogeneities were found among all studies regarding IL-10 -1082A/G polymorphism and overall cancer risk (homozygous: P = 0.025; heterozygous: P < 0.001; dominant: P < 0.001 and allele comparison: P < 0.001), but not under the recessive model (P = 0.242) ( Table 2). Considerable heterogeneities were also observed for the -819T/C (all P < 0.001) and -592A/C (homozygous: P = 0.001; heterozygous: P = 0.001; recessive: P = 0.035; dominant: P < 0.001 and allele comparison: P < 0.001) polymorphisms. Therefore, the random-effect model was used to generate wider CIs. Sensitivity analysis was conducted and the results indicated that each individual study did not influence the pooled ORs obviously (data not shown).

Publication bias
The funnel plot was symmetric for the -1082A/G (Figure 3

False-positive report probability (FPRP) test analysis
The significant findings were assessed using the FPRP test and the results are shown in Table 3. With a prior probability of 0.1, assuming that the OR for a specific genotype was 0.67/1.50 (protection/risk), with statistical power of 0.857, the FPRP value was 0.179 for the -1082A/G polymorphism and cancer risk under the dominant model, and a positive association was also found for low quality studies (dominant: FPRP = 0.053 and allele comparison: FPRP = 0.129). As regards the -819T/C polymorphism, a positive association was found

DISCUSSION
In this meta-analysis, we comprehensively investigated the associations between three promoter variants (-1082A/G, -819T/C and -592A/C) in IL-10 gene and cancer risk in the Chinese population through 53 articles. The results revealed that all the three IL-10 gene polymorphisms we considered were associated with an increased overall cancer risk. Stratification analysis showed that the association between the -1082A/G polymorphism and cancer risk was more evident for hepatocellular carcinoma and low quality studies, the association between the -819T/C polymorphism and cancer risk was more obvious for lung cancer and oral cancer. However, the -592A/C polymorphism showed a statistically significant increased risk for lung cancer, oral cancer, hospital-based studies and low quality studies, but a decreased risk for colorectal cancer. To our knowledge, this is so far the first meta-analysis that has assessed multiple promoter polymorphisms in IL-10 gene with cancer risk in the Chinese population.
Three meta-analyses including international studies have investigated the association of IL-10 -1082A/G, -819T/C and -592A/C polymorphisms with overall cancer susceptibility. The study carried out by Wang et al. [92] analyzed IL-10 -1082A/G polymorphism, consisting 61 international studies with a total of 14,499 cases and For each study, the estimated OR and its 95% CI are plotted with a box and a horizontal line. ◊, pooled ORs and its 95% CIs. www.impactjournals.com/oncotarget 16,967 controls, in which no significant association was found between this polymorphism and overall cancer risk. Another meta-analysis [93] including 15,942 cases and 22,336 controls investigated IL-10 -819C/T polymorphism and cancer risk, without finding any significant association between this polymorphism and overall cancer risk. The study carried out by Ding et al. [94] considered IL-10 -592C/A polymorphism, in which a decreased risk of overall cancer was found with the AA genotype. Other meta-analyses with international studies have assessed the association between polymorphisms in IL-10 gene and susceptibility to some types of cancer. For example, two studies [95,96] revealed no significant association between IL-10 -1082A/G, -819T/C and -592A/C polymorphisms with non-Hodgkin lymphoma susceptibility. Some of the significant associations found in the previous studies were not validated in our metaanalysis, for example, IL-10 -1082A/G polymorphism was associated with an increased lung cancer risk [92]. We also found some significant associations that were not observed in previous analyses. For instance, we found that IL-10 -592A/C polymorphism was associated with a decreased colorectal cancer risk. The discrepancy occurred because our analysis was carried out only in the Chinese population, suggesting that the polymorphisms on cancer risk might vary among different study subjects' ethnicity or lifestyle factors.
To make our significant findings more noteworthy, FPRP analysis was performed. Interestingly, FPRP test results revealed that only the association between IL-10 -1082A/G polymorphism and overall cancer risk remained significant at the prior probability level of 0.1. In the subgroup analysis, only the low quality studies, lung cancer and colorectal cancer remained significant. Other findings were false-positive, which might be due to the limited sample size.
Our present meta-analysis has some highlights. First, it identified the significant association between IL-10 -1082A/G, -819T/C and -592A/C polymorphisms and an increased overall cancer risk in the Chinese population. Second, the quality of each included study was evaluated by the quality score criteria. Third, no publication bias was detected in the study, indicating the robustness of the results. Finally, the significant findings were further validated using the FPRP test, making the results more authentic. However, some possible limitations should be considered. First, the total sample size in each individual study was less than 1000 in all but four studies [69,82,85,86], which might reflect a difficulty to evaluate the real association. Second, our results were based on unadjusted estimates, which might cause confounding bias. Third, in the subgroup analysis by cancer type, only two studies were included for some types of cancer, which might affect the detection of the real association. Finally, the potential gene-gene, and gene-environment interactions were not assessed due to the lack of information in the original studies.  In conclusion, this meta-analysis suggested an association between IL-10 gene polymorphisms and cancer risk in the Chinese population, especially for lung cancer, colorectal cancer and low quality studies. Welldesigned studies with large sample size are required to verify our findings.

Search strategy
A systematic literature search was conducted in PubMed, Embase, CNKI and Wanfang database using the following MeSH terms and their synonyms: ("interleukin-10" or "interleukin 10" or "IL-10" or "IL 10") AND ("polymorphism, single nucleotide" [MeSH] or "SNP" or "single nucleotide polymorphism" or "polymorphism" or "variant" or "variation") AND ("neoplasms" [MeSH] or "neoplasia" or "neoplasm" or "tumor" or "malignancy" or "cancer"), up to 19 January, 2017. In addition, review articles and references of the selected articles were manually searched to identify additional relevant articles. Only the most recent publications or the ones with most participants were included in the final meta-analysis in cases of overlapping data.

Inclusion and exclusion criteria
The inclusion criteria were as follows: (1) studies investigating the association between IL-10 -1082A/G, -819T/C and -592A/C polymorphisms with cancer risk in Chinese populations; (2) case-control studies; (3) studies providing sufficient data for calculation of ORs and 95% CIs. Studies were excluded if any of the following aspects existed: (1) not a case-control study; (2) duplicate publications; (3) studies without available genotype data; (4) review articles, meta-analyses, conference abstracts or editorial articles; and (5) genotype frequencies in the controls departure from HWE.

Data extraction
Two investigators independently extracted the relevant data from all included studies based on the inclusion criteria listed above. Disagreement was resolved by discussion with a third investigator. The following information was extracted from each included study: first author's surname, publication year, cancer type, control source (hospital-based or population-based), genotyping methods, and number of cases and controls with different genotypes.

Quality assessment
Two independent investigators assessed the qualities of all included studies according to the criteria from a previous meta-analysis [97]. Quality scores of studies ranged from 0 (lowest) to 15 (highest), and the studies with scores > 9 were considered of high quality.

Statistical analysis
The strength of association between IL-10 -1082A/ G, -819T/C and -592A/C polymorphisms and cancer risk was assessed by calculating the ORs and the corresponding 95% CIs. The pooled ORs were calculated for the homozygous model, heterozygous model, recessive model, dominant model and an allele comparison. The betweenstudy heterogeneity was quantified by chi-square based Q test and the fixed-effects model (the Mantel-Haenszel method) [98] was used when no significant heterogeneity was observed (P > 0.1); otherwise, the random-effects model (the DerSimonian and Laird method) [99] was adopted. Subgroup analysis was performed by cancer type (if one cancer type contained less than two studies, it was merged into the "other cancers" group), control source (hospital-based studies and population-based studies), and quality scores (≤ 9 and > 9). Sensitivity analysis was performed to assess results stability. Publication bias was examined using Begg's funnel plot and Egger's linear regression test.
The FPRP was calculated to examine the significant associations found in the present meta-analysis. FPRP was calculated with 0.2 as a FPRP threshold and a prior probability of 0.1 was assigned to detect an OR of 0.67/1.50 (protective/risk effects) for an association with the genotypes under investigation [100]. FPRP values below threshold 0.2 were considered as noteworthy associations. All the statistical tests were performed using STATA version 12.0 (Stata Corporation, College Station, TX). All the P values were two-sided, and P < 0.05 were considered statistically significant.

Author contributions
Ping Wang, Junling An and Yanfeng Zhu performed the research study and collected the data; Ping Wang, Xuedong Wan and Hongzhen Zhang analyzed the data; Shoumin Xi and Sanqiang Li designed the research study; Ping Wang wrote the paper. All authors read and approve the final manucript.

ACKNOWLEDGMENTS
This work was supported by the Key Scientific and Technological Project of Henan province (Grant No. 162102310413).