Investigation of IL-4, IL-10, and HVEM polymorphisms with esophageal squamous cell carcinoma: a case–control study involving 1929 participants

Abstract It is believed that an individual’s hereditary factors may be involved in the development of esophageal cancer (EC). The present study recruited 721 esophageal squamous cell carcinoma (ESCC) cases and 1208 controls and explored the roles of single nucleotide polymorphisms (SNPs) in the interleukin-4 (IL-4), IL-10, and herpesvirus entry mediator (HVEM) genes in contributing to ESCC risk. IL-4, IL-10, and HVEM SNPs were analyzed by employing an SNPscan method. After adjustment for body mass index (BMI), smoking, drinking, age and gender, we identified that the rs2070874 T>C locus in IL-4 gene decreased the risk of ESCC (CC vs. TT: P=0.008; CC vs. TT/TC: P=0.010). After a stratified analysis, we suggested that the IL-4 rs2070874 T>C variants might be a protective factor for ESCC in male, ≥63 years old, never smoking, drinking and BMI < 24 kg/m2 subgroups. In addition, we identified that the rs2243263 G>C polymorphism in IL-4 gene was a risk factor for ESCC development in the BMI ≥ 24 kg/m2 subgroup (GC vs. GG: P=0.030 and GC/CC vs. GG: P=0.018). We identified an association of the IL-4 rs2070874 T>C SNP with the decreased susceptibility of ESCC in stage I/II subgroup. Finally, we found an association of the IL-10 rs1800872 T>G SNP with a worse differentiation (TG vs. TT: P=0.048 and GG/TG vs. TT: P=0.032). In conclusion, the findings indicate a potential importance of IL-4 rs2070874 T>C, IL-4 rs2243263 G>C and IL-10 rs1800872 T>G SNPs in the development of ESCC.


Introduction
In China, esophageal cancer (EC) is the fourth most frequently diagnosed form of malignant tumor in males and the fifth most commonly diagnosed form in females, approximately 320800 and 157200 cases occurred in 2015, respectively [1]. The incidence of EC in Eastern Asia is in the top five worldwide, including China. Esophageal squamous cell carcinoma (ESCC) is a major histological subtype, accounting for 90% of all EC cases. The complex interaction of economical and environmental conditions with individual's hereditary factors may lead to EC development [2,3]. The etiology and development of EC is not fully understood, despite many investigations have payed close attention to the importance of immunity [4,5]. Recently, it was hypothesized that some important variants in immune-related genes may influence the susceptibility of ESCC. 87 years old. The controls were not related to any ESCC case. Using a pre-structured questionnaire, we collected epidemiological data from participants. The ESCC patients and normal controls signed consent forms.

Statistical analysis
The difference in alcohol consumption, body mass index (BMI), gender, cigarette use, and age were tested by using χ 2 test. Mean age was calculated by using a Student's t test. We used a Chi-square test (χ 2 ) or Fisher's exact test to determine whether the frequencies of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 variants in ESCC cases and controls were different. A multivariate logistic regression analysis method was used to calculate the crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) (SAS 9.4 software package; SAS Institute Inc., Cary, NC, U.S.A.). The relationship of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 polymorphisms with ESCC development was determined by ORs and 95% CIs. The statistical significance of all analyses was P<0.05 (two-sided). An internet-based Hardy-Weinberg equilibrium (HWE) test (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl) was also harnessed to assess whether the distribution of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 genotypes could represent the included population.

Baseline characteristics
In total, 721 ESCC cases and 1208 controls were recruited (Table 1). Of these ESCC cases, 170 were females and 551 were males, average age was 62.59 + − 8.18 years. In the control group, there were 309 females and 899 males with an average age of 62.92 + − 8.94 years. There was no difference in terms of mean age (P=0.413). The categorical variables, age and gender, were well-matched (P>0.05). However, the distribution of other categorical variables (e.g. tobacco use, BMI, and drinking status) were significantly different (all P<0.001). Among ESCC cases, there were 405 (56.17%) with lymphatic metastasis. The AJCC version 8.0 criteria (2018) was used to determine the ESCC stage; and 328 ESCC cases were stage I/II and 393 were stage III/IV. After genotyping the 1929 participants, the association of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 genotypes with ESCC risk was assessed.
The minor allele frequencies (MAFs) of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 loci are shown in Table 2. They are similar to the data of Chinese population. As presented in Table  2, the HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 genotypes in controls accorded with HWE.
HVEM rs2234167, IL-4 rs2243263 and IL-10 rs1800896 and rs1800872 genotypes are shown in Table 3. Both crude and adjusted comparisons indicated that HVEM rs2234167, IL-4 rs2243263, and IL-10 rs1800896 and rs1800872 loci were not associated with the risk of ESCC (Table 4).  Additionally, a subgroup analysis was conducted by ESCC stage. We identified an association between IL-4 rs2070874 T>C SNP and the decreased susceptibility of ESCC in stage I/II subgroup (CC vs. TT: P=0.022; CC vs. TT/TC: P=0.025, Table 4). However, this association could not been identified for other SNPs.

Relationship of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 loci with ESCC in stratified analyses
In a stratified analysis, the IL-4 rs2070874 genotypes are listed in Table 5. After an adjustment, we suggested that IL-4 rs2070874 C allele was a protective factor for ESCC in five subgroups (male subgroup: In other subgroups, no association of L-4 rs2070874 with ESCC risk was found ( Table 5).
The IL-4 rs2243263 G>C genotypes in the stratified analysis are listed in Table 6. After adjustment, we identified that IL-4 rs2243263 G>C polymorphism was a risk factor for ESCC development in the BMI ≥ 24 kg/m 2 subgroup (GC vs. GG: P=0.030 and GC/CC vs. GG: P=0.018, Table 6).
In other stratified analyses, adjustment comparisons suggested that HVEM rs2234167, and IL-10 rs1800872 and rs1800896 loci did not confer a risk of ESCC (data not shown).

Association of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 loci with lymphatic metastasis in ESCC cases
Among the 721 ESCC patients, 405 patients had lymphatic metastasis. As presented in Table 7, we found a null association of HVEM rs2234167, IL-4 rs2070874, rs2243263 and IL-10 rs1800896 and rs1800872 SNPs with different lymph node status.

Association of HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872 loci with tumor grade of ESCC cases
As presented in Table 1, 142 patients had well-differentiated tumors, 405 had moderately differentiated, tumors and 174 has poorly differentiated tumors. We found an association of the IL-10 rs1800872 T>G SNP with a worse differentiation (TG vs. TT: P=0.048 and GG/TG vs. TT: P=0.032, Table 8).

Discussion
Immunotherapy is altering how we comprehend malignancies and offers new methods to treat them. EC is a representative model of immune and inflammation-related cancer [39]. Recently, some studies indicated that the SNPs in inflammation and immune-related genes might influence the risk of EC [40,41]. In this study, we explored the role of immune-related gene SNPs (HVEM rs2234167, IL-4 rs2070874, and rs2243263, and IL-10 rs1800896 and rs1800872) to ESCC development. We observed that IL-4 rs2070874 T>C could decrease a risk to ESCC, even in the stage I/II subgroup. However, in BMI ≥ 24 kg/m 2 subgroup, IL-4 rs2243263 G>C might increase the risk of ESCC. We also found an association of the IL-10 rs1800872 T>G SNP with a worse differentiation. IL-4 is an important regulator of immune and inflammation pathways. Some reports have suggested that IL-4 levels are higher in untreated ESCC patients than in controls [42][43][44]. It is considered that IL-4 levels may be implicated in the development of ESCC. The IL-4 rs2070874 T>C polymorphism is a 5 -UTR SNP. In a high-risk gastric cancer (GC) region, a previous study suggested that rs2070874 C allele in the IL-4 gene might decrease the susceptibility to GC in a Chinese population [45]. Lu et al. reported that the rs2070874 C allele increased the risk of HCC in a male subgroup [46]. However, Chang et al. and Wang et al. found that the IL-4 rs2070874 polymorphism might not influence the susceptibility of cancer in Chinese population [47,48]. In this study, we included 1929 subjects and investigated the correlation of this SNP to ESCC susceptibility. We found that IL-4 rs2070874 T>C polymorphism seemed to be a protective factor for ESCC development. Our findings were similar to a previous meta-analysis that suggested that the IL-4 rs2070874 C allele could be associated with a decreased susceptibility of gastrointestinal cancer [14]. A functional study indicated that the IL-4 rs2070874 allele C could promote a higher level of IL-4 in plasma [49]. IL-4 has an anti-inflammatory effect and may decrease the risk of ESCC by inhibiting the inflammation. FitzGerald et al. reported that the IL-4 rs2070874 allele C could decrease the risk of prostate cancer specific mortality [50]. Consistent with that report, we identified an association between the IL-4 rs2070874 T>C SNP and a decreased susceptibility to ESCC in the stage I/II subgroup. However, we did not find an association of IL-4 rs2070874 T>C polymorphism with lymphatic metastasis. This might be due to the limited sample sizes. In the future, the relationship of the rs2070874 SNP in IL-4 gene with progress and prognosis should be further explored. Rs2243263 G>C, an intron SNP in the IL-4 gene, was studied for the relationship of this SNP to some diseases. This SNP might decrease the risk of asthma in the African American children, while this relationship was not identified in Caucasians [51]. Hsiao et al. reported that the IL-4 rs2243263 C allele was a protective factor for HBV surface antigen reverse seroconversion in non-Hodgkin lymphoma cases undergoing rituximab treatment. A previous study investigated the relationship of IL-4 rs2243263 G>C with colon and rectal cancer risk [18], but no association was found. However, in a large simple size study, Lan et al. found that the IL-4 rs2243263 G>C SNP increased the susceptibility to non-Hodgkin lymphoma [19]. In this study, we found that the IL-4 rs2243263 G>C might increase the risk of ESCC in obese and overweight subjects (Table 6). It was reported that the IL-4 level in mothers was inversely linked to overweight in early childhood and might influence the metabolic profile of childhood [52]. In addition, the level of IL-4 decreased with antipsychotic-induced weight gain [53]. It is suggested that the level of IL-4 could influence obesity and overweight. Introns are regulatory sequences that can affect the expression of genes. Here, we found that the rs2243263 G>C polymorphism, a SNP in IL-4 intron region, might alter the risk of ESCC. It is presumed that the rs2243263 G>C polymorphism influences the level of IL-4 by regulating gene transcription. In the future, a functional study should be considered to explore the potential mechanism.
The IL-10 rs1800872 T>G is a promotor SNP. Torres-Poveda et al. reported that the expression of IL-10 mRNA and the level of serum IL-10 were significantly higher in subjects with the IL-10 rs1800872 T allele [54]. A recent study found that IL-10 rs1800872 T>G SNP promoted the risk of EC [25]. A meta-analysis also confirmed this association [55]. In our case-control study, we did not find the association of IL-10 rs1800872 T>G SNP with the development of EC, even in stratified analyses and reviewing different lymph node status. Additionally, Liu et al. reported that IL-10 rs1800872 GG genotypes predicted the worse survival of diffuse large B-cell lymphoma patients treated with rituximab-CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) [56]. In this study, we found that the IL-10 rs1800872 G allele was associated with poorly differentiated tumor. Thus, in the future, the association of the IL-10 rs1800872 T>G SNP and the survival of ESCC cases should be further studied.
Limitations in the present study should be acknowledged. First, in the present study, we only included five functional SNPs and explored the association of the risk to ESCC. Second, there were other environmental risk factors (e.g. vegetable and fruit intake, aspirin and NSAIDs use, and physical exercise), which we did not consider for their influence to the development of ESCC. Third, the number of ESCC patients was limited and our study may be