Effects of IL-4-590C/T (rs2243250) Polymorphism on the Susceptibility of Smoking-Related Cancer: A Meta-Analysis Involving 11,407 Subjects

Background Several previous studies have assessed the relationship between IL-4-590C/T gene polymorphism and smoking-related cancer in recent years; however, the results remain controversial. Based on it, the study intends to clarify whether IL-4-590C/T variant increases the risk of smoking-related cancer through meta-analysis. Methods We searched PubMed, EMBASE, Web of Science, Cochrane Library database, China National Knowledge Infrastructure, and Wanfang data information service platform to collect qualified case-control studies in strict accordance with the inclusion and exclusion standards. The 95% confidence interval (95% CI) and its odds ratio (OR) were adopted to access the relation between IL-4-590C/T gene polymorphism and smoking-related cancer; sensitivity analysis and publication bias assessment were carried out after the studies' quality evaluation. Results 17 studies were included in total, with 5,061 patients and 6,346 control cases. A significant association between IL-4-590C/T variant and smoking-related cancer in total population was revealed in our meta-analysis results, and IL-4-590C/T variant might have a relatively protective effect on smoking-related cancer (CT vs. TT: P=0.026, OR = 0.900, 95% CI: 0.820–0.987). Subgroup analysis by ethnicity showed that the IL-4-590C/T polymorphism was associated with a decreased risk of smoking-related cancer in the Asian population (CT vs. TT: P=0.008, OR = 0.878, 95% CI: 0.798–0.967; CC + CT vs. TT: P=0.030, OR = 0.903, 95% CI: 0.824–0.990). Subgroup analysis based on types of cancer demonstrated the IL-4-590C/T variant achieved a lower risk in renal cell cancer (CC vs. TT: P=0.046, OR = 0.640, 95% CI: 0.412–0.993). Conclusion There is a conspicuous association between IL-4-590C/T polymorphism and decreased risk of smoking-related cancer, particularly in Asians. And IL-4-590C/T polymorphism may have a protective effect on renal cell cancer.


Introduction
Smoking is currently an important risk factor for a variety of cancers. Almost half of male urinary tumors and 1/3 of female urinary tumors are caused by smoking [1]. Compared with nonsmokers, smokers can increase the prevalence of bladder cancer by two to three times [2]. Cigarette smoke itself is a rich free radical that induces DNA damage in cells by inducing it as a triggering factor for tumors. In addition, more than 60 carcinogens have been found in cigarette smoke, like aromatic amines, polycyclic aromatic hydrocarbons, and specific nitrosamines in tobacco; all of them are recognized as carcinogens in humans [3][4][5]. e genotoxic patterns of these chemicals have been delineated and related to the DNA damage in cells. DNA adducts can be caused by smoking [6], as a result of the formation of covalently bound DNA damage by the production of electrophilic substances. e formation of DNA adducts is a cancerous potential result, and DNA adducts can mislead DNA replication, resulting in mutations [7]. ese specific genetic mutations affect critical areas that control cell function, which can lead to tumorigenesis [8]. Compared with individuals who smoke slightly, individuals who smoke heavily have a higher risk of cancer because of the higher concentration of carcinogens in their body [9,10]. e International Agency for Research on Cancer (IARC) has defined smoking-related cancer as the cause of cancers of the lung, oral cavity, gastric, bladder, and so on [11]. According to the latest cancer report of 2018, smoking-related cancers became a major public health problem with extremely high morbidity and mortality [12].
Interleukin-4 (IL-4) gene is oriented in the cytokine gene cluster on chromosome 5q31-33. It is a cytokine that can promote the proliferation of 2 cells while inhibit the proliferation of 1 cells, and finally reduce the immune response mediated by 1 [13]. Many studies have shown that interleukin-4 can promote tumor progression and metastasis by affecting apoptosis of tumor cells [14][15][16][17]. At present, there have been several case-control researches to explore the relationship between IL-4-590C/T (rs2243250) gene polymorphism and smoking-related cancer, but the results are still controversial. e polymorphism of IL-4-590C/T (rs2243250) is a C to T base mutation, and the T-allele gene means mutational allele gene and C-allele gene means the wild allele gene. Our aim is to study whether the mutational T-allele and TT genotype increase the susceptibility of smoking-related cancer compared with CT and CC containing the wild C allele through systematic review and meta-analysis. e inclusion criteria were as follows: (1) case-control articles about the polymorphism of IL-4 referring to smoking-related cancers; (2) the studies including IL-4-590C/T (rs2243250) variant; (3) studies having sufficient data for examining the odds ratio (OR) with 95% confidence intervals (CIs); (4) the articles reporting the risk of smoking-related cancers which were defined according to the IARC monograph; (5) the genotypes accorded with Hardy-Weinberg equilibrium (HWE).

Exclusion Criteria.
e exclusion criteria were as follows: (1) the control group gene distribution contained in these studies was not subject to HWE; (2) literature with incomplete data or data not available.

Literature Quality Evaluation and Data Extraction.
e quality of the included literature was accessed using the evaluation criteria of the case-control study in the Newcastle-Ottawa scale (NOS). e evaluation included (1) the selection of study subjects, (2) the inter-group comparability, and (3) the exposure evaluation. e NOS divides the quality of the literature into 0-9 points. e higher the score, the better the quality of the literature. e score ≥5 is considered to be a literature of high quality. e data extraction includes the studies' names, the publication year, the target country, the race, the cancer type, the amount of case group and control group, and its genotype distribution and genotyping methods.

Statistical
Methods. Meta-analysis was performed using Stata12.0 software. Q test and heterogeneity coefficient I 2 were used to judge the heterogeneity between studies. If there is statistical heterogeneity (I 2 > 50%, or P < 0.1), the random effect model is used for meta-analysis; otherwise, the fixed-effect model is used. Using odds ratio (OR) as an effector to reflect the strength of the correlation between IL-4 rs2243250 and smoking-related cancer. And we included five models: (1) allele model, (2) dominant model, (3) codominant model, (4) homozygote model, and (5) recessive model, to analyze the association, respectively. e T allele is a mutational allele, and the C allele is a wild allele. In addition, the heterogeneity source was explored by subgroup analysis. And the P value of Egger's test, Begg's test, and funnel chart was obtained to evaluate the bias induced from publication.

Results
A total of 1,324 potentially relevant researches were identified in accordance with the search strategy ( Figure 1). Ultimately, 17 case-control articles were selected into our meta-analysis.

Traits of the Selected Studies.
e traits of the selected studies are presented in Table 1. For IL-4 rs2243250 polymorphism, 24 articles were investigated. However, deviations from HWE were found in 7 articles [18][19][20][21][22][23][24], so we have to remove these 7 articles (Figure 1), while the other 17 articles were in accordance with HWE. Of these 17 articles [25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41] (including 5,061 cases and 6,346 controls), seven studied the relation between IL-4-590C/T variant and the susceptibility of gastric cancer, three between IL-4-590C/T variant and the susceptibility of lung cancer, and three between IL-4-590C/T variant and the susceptibility of oral carcinoma. e others studied the association between studied IL-4-590C/T variant and the susceptibility of bladder cancer, hepatocellular cancer, and renal cell carcinoma. For genotyping methods, seven were using PCR-RFLP, six were using TaqMan, three were using PCR, and one was using MASSARRAY. e NOS scores of the 17 documents were all more than 5, meaning that all of them were high-quality studies ( Table 1). Table 2 listed the main results of the meta-analysis of IL-4-590C/T (rs2243250) gene polymorphism and the risk of smoking-related cancer in overall population. Table 3 shows the main results of subgroup analysis by ethnicity.

Association between IL-4-590C/T Polymorphism and the Risk of Smoking-Related Cancer.
Seventeen articles including 5,061 cancer cases and 6,346 normal controls were investigated. As illustrated in Table 2, a significant relation was found for IL-4-590C/T gene polymorphism and the risk of smoking-related cancer in overall population (CT vs. TT: P � 0.026, OR � 0.900, 95% CI: 0.820-0.987). And the CT genotype carriers have a slightly lower incidence of smoking-related cancer compared to that of TT carriers and the TT variants might be a potential risk factor for smokingrelated cancer susceptibility in overall populations (Table 2 and Figure 2).

Association between IL-4-590C/T Polymorphism and Smoking-Related Cancer Risk in Subgroup Analysis by
Ethnicity.
e heterogeneity of IL-4-590C/T variant and smoking-related cancer was complicated by multiple indexes, so subgroup analysis of different ethnicities was carried out. A significant association between IL-4-590C/T variant and the susceptibility of the smoking-related cancer in Asian population (CT vs. TT: P � 0.008, OR � 0.878, 95% CI: 0.798-0.967; CC + CT vs. TT: P � 0.030, OR � 0.903, 95% CI: 0.824-0.990) was found. Our results indicated that CC/ CT genotype carriers had a lower risk compared with TT carriers and the CC/CT variant might be a protective factor for smoking-related cancer susceptibility in Asian population. However, any association between smoking-related cancer risk and IL-4-590C/T variant was found in Caucasians (Table 3).

Association between Smoking-Related Cancer Susceptibility and IL-4-590C/T Polymorphism in Subgroup Analysis by Cancer Type.
Our results demonstrated that IL-4-590C/T polymorphism had a lower risk for renal cell cancer (CC vs. TT: P � 0.046, OR � 0.640, 95% CI: 0.412-0.993). And the CC genotype might be a protective factor in renal cell cancer.
However, for bladder cancer, gastric cancer, oral carcinoma, lung cancer, and hepatocellular cancer, we did not find any obvious association in different genotype carriers ( Figure 3).

Sensitivity
Analysis. Sensitivity analysis was performed by deleting a study one by one, and the combined results showed no significant changes, denoting that the results of this study were relatively steady ( Figure 4).

Publication Bias. Publication bias was evaluated by
Begg's funnel plots and Egger's test (the allelic contrast of   Figure 5).

Discussion
In terms of originality, our meta-analysis is the first paper to study the relationship between IL-4-590C/T (rs2243250) polymorphism and smoking-related cancer. We use five genetic models ((1) allele model; (2) dominant model; (3) codominant model; (4) homozygote model; (5) recessive model) to determine the association between IL-4-590C/T (rs2243250) polymorphism and smoking-related cancer. As a result, we found that IL-4-590C/T (rs2243250) polymorphism was associated with the decreased risk of smoking-related cancer in overall population. A slightly lower incidence of smoking-related cancer was observed in CT carriers compared to TT carriers and the TT variant might be a risk factor resulting in smoking-related cancer susceptibility in different races. In subgroup analysis by ethnicity, our results indicated that the polymorphism of IL-4-590C/T was associated with the decreased risk of smoking-related tumors in Asian population. e CT/CC genotype was a protective factor for the susceptibility of smoking-related cancer while the TT genotype was a risk factor for smoking-related cancer susceptibility. No obvious association was observed in the population of Caucasians. It might account for that genetic polymorphisms are greatly different in various racial groups, which means that we can pay more attention to IL-4-590C/T gene polymorphism when screening for smoking-related cancer, especially among Asians.
For cancer-type subgroup analysis, the susceptibility of renal cell carcinoma was observed to be associated with the polymorphism of IL-4-590C/T. And the CC genotype might be a protective factor in renal cell cancer.
Our results were consistent with Chu et al. [38]; they found that the polymorphism of IL-4 rs2243250 was also associated with the lower risk of renal cell carcinoma. But our result was different from Cozar et al. [25]; they did not find that the polymorphism of IL-4 rs2243250 was also associated with the susceptibility of renal cell carcinoma. In addition, Pan et al. [32] and Tan et al. [37] have found that IL-4-590C/T variant was relevant with the increased risk of lung cancer and gastric cancer. e possible reasons for the different results of the above studies have been summarized as follows: (1) the genetic background of the population was different; (2) the sample volume of Asians we included in this meta-analysis was much larger than that of Caucasians; (3) the methods of genotyping were different in these included studies; (4) the different types of smoking-related cancers had different relationships with IL-4 rs2243250 polymorphism. In addition, IL-4 rs2243250 polymorphism was involved in the metabolism of a variety of carcinogens associated with tobacco smoke, which may be affected by complex factors like multiple genes, environmental factors, individual genetic background, and dietary habits.
Some advantages feature our findings. First, we performed the NOS quality test on all the included literature, ensuring that the literature works included were of high quality. Second, HWE tests were conducted on included documents, and we excluded those that did not meet the Hardy-Weinberg Equilibrium.
ird, more than 11,407 subjects consisting of 5,061 patients with smoking-related cancer and 6,346 controls were contained in the study. erefore, the sample size was large enough to confirm the results of our analysis. Moreover, sensitivity analysis and Egger's test showed the results were stable.
One of the important things about meta-analysis is heterogeneity, and ignoring the heterogeneity may lead to scientific errors. We found the heterogeneity was significant in many models. erefore, we conducted a subgroup analysis and finally found that the type of cancer may be the source of this heterogeneity.

Study ID
Subtotal (I-squared = 64.4%, p = 0.094) Lu [31] Chu [38] Renal cell carcinoma  only included English and Chinese studies, so it might cause language bias. erefore, there is a need for more multicentric researches with large samples to be carried out in future to gain more insights into the association between IL-4 rs2243250 polymorphism and smoking-related cancer.

Conclusion
Our study indicates that smoking-related cancer susceptibility is associated with IL-4-590C/T polymorphism in the total population. e CT genotype carriers have a slightly lower incidence of smoking-related cancer compared to TT carriers in the overall populations. And the IL-4-590C/T polymorphism has different susceptibility to smoking-related tumors in the different populations, especially in Asians. However, in order to determine a comprehensive conclusion on the correlation between smoking-related cancer susceptibility and IL-4-590C/T polymorphism, more prospective cohort studies are still needed in the future.

Disclosure
Guangyuan Chen, Cong Hu, and Yuxuan Song contributed equally and should be considered as the co-first authors.

Conflicts of Interest
e authors declare that they have no conflicts of interest.

Authors' Contributions
Peng Huang was responsible for study design with methodology checking and data analysis guidance. Guangyuan Chen and Cong Hu were involved in data collection. Guangyuan Chen, Cong Hu, Penghui Lai, and Haifei Zhang analysed the data; Guangyuan Chen, Cong Hu, and Yuxuan Song wrote the manuscript. Yuxuan Song, Yiling Zhang, and Song Li carried out grammar checking.