Investigation of leptin receptor rs1137101 G>A polymorphism with cancer risk: evidence from 35936 subjects

Leptin receptor (LEPR) signaling may be involved in promoting angiogenesis and proliferation, inhibiting apoptosis and playing a vital role in the progression of carcinogenesis. A number of studies have focused on the association of LEPR rs1137101 variants with susceptibility of cancer, however, the observed results were controversial. We searched literature on the relationship of LEPR rs1137101 G>A polymorphism with cancer risk by using PubMed and Embase databases, covering all publications up to 14 October 2018. In total, 44 case–control studies with 35,936 subjects were included. After combining all eligible studies, we identified null relationship between LEPR gene rs1137101 G>A polymorphism and overall cancer risk [A vs. G: odds ratio (OR ) =  0.97, 95% confidence interval (CI ) =  0.89–1.06, P = 0.547; AA vs. GG: OR  =  0.93, 95% CI  =  0.78–1.13, P = 0.476; AA/GA vs. GG: OR  =  0.99, 95% CI  =  0.91–1.09, P= 0.890 and AA vs. GA/GG: OR  = 0.92, 95% CI  =  0.82–1.04, P= 0.198]. However, in a subgroup analysis, there was an increased susceptibility of oral and oropharyngeal cancer in AA vs. GA/GG genetic model (OR, 1.83; 95% CI, 1.01–3.33; P=0.048). Considering the limited participants were included, the findings might be underpowered. Sensitivity analysis identified that any independent study omitted did not materially influence the pooled ORs and CIs. The results of publication bias detection showed that there was no evidence of bias. In summary, this analysis indicates that no significant association of cancer risk was identified to be correlated with rs1137101 G>A variants, even in stratified analyses.


Introduction
Cancer is one of the common health burden worldwide. Due to the prevalence of smoking and drinking, environmental pollution, as well as population aging, the incidence of malignancy is rising. According to the estimation of Global Cancer Statistics 2018, approximately 18.1 million new cancer cases and 9.6 million cancer-related deaths may have occurred in 2018 [1]. In the developing countries, the prognosis of cancer could be poorer than that in the developed countries. The reason of this phenomenon may be due to the diagnosis at an advanced stage combined with limited treatment. Some effective measures can potentially contribute to relieve the global cancer burden, including the application of precise early detection and treatment, tobacco and alcohol control, vaccine injection, sufficient fruits and vegetables intake and appropriate physical exercise. Overweight/obesity influences the health of more than 200 million people [2]. Overweight/obesity may play a vital role in growing morbidity of malignancy [3,4,5]. Thus, it is believed that overweight/obesity-related genes could affect the development of cancer. heterogeneity. Then a random-effects model (DerSimonian and Laird method) was used to assess the association of LEPR rs1137101 polymorphism with cancer susceptibility [28,29]; otherwise, a fixed-effects model (Mantel-Haenszel method) was harnessed to pool the data [30]. Stratified analysis was conducted to explore the source of heterogeneity. We performed one-way sensitivity analysis to evaluate the effect of an individual study on pooled ORs and CIs. Begg's funnel plot and the Egger's test were used to detect the potential bias in included publications, and a P<0. 10 was considered significant [31,32]. Statistical analysis of the present study was calculated with STATA 12.0 software (StataCorp, College Station, Texas, U.S.A.).

Heterogeneity analysis
Significant heterogeneity among the eligible studies was found in this pooled analysis (Table 3). In the present study, we carried out subgroup analysis to explore the sources of heterogeneity. We found that Caucasians, breast cancer, hepatocellular carcinoma, other cancers, small sample sizes (<1000) and hospital-based case-control studies might contribute the major source to heterogeneity.

Sensitivity analysis
Sensitivity analysis of one-way method identified that any individual study deleted did not materially influence the pooled ORs and CIs under all genetic comparisons. These findings indicated that our observations were stable and reliable (Supplementary Figure S1).  Figure S2).

Discussion
Recently, variants in LEPR gene and their potential associations with cancer risk have been explored. Rs1137101 G>A polymorphism is one of the important variants in LEPR gene. LEPR rs1137101 G>A polymorphism is located on the exon region of LEPR gene, and it has been thought to be involved in the development of cancer by a number of studies. Several case-control studies reported that LEPR rs1137101 G>A polymorphism might be associated with the decreased risk of cancer [16,22,26,41,42]. However, several primary studies also suggested that LEPR rs1137101 locus could promote the progression of cancers [15,18]. Meanwhile, two meta-analyses were carried out to clarify the correlation of this SNP with susceptibility of overall cancer [9,10]. The results of these meta-analyses indicated that LEPR rs1137101 G>A polymorphism might not be associated with the risk of cancer. Moreover, more epidemiologic data were reported [11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26]. Therefore, an updated meta-analysis is necessary to calrify this issue precisely. In this meta-analysis, data of 44 case-control studies involving 13,711 cases and 22,225 controls, which is higher compared with these previous pooled-analyses mentioned above, were included and analyzed. Therefore, the obtained results may be more convincing.
LEP/LEPR signaling may promote cell proliferation and inhibit epithelial cell apoptosis [6]. In addition, Ben et al. [49] reported that LEPR rs1137101 G>A SNP may affect plasma LEP levels and BMI. A previous study suggested that leptin level was associated with the development of breast cancer [50]. However, null associations of LEPR rs1137101 locus with cancer susceptibility was identified, which was analogous to the results reported in previous meta-analyses [9,10,51,52], but unlike the other four meta-analyses [48,[53][54][55]. Due to lack of sufficient data, the findings of previous systematic reviews and meta-analyses might be conflicting. Ethnicity also may be a vital factor for the difference. The minor allele frequency of LEPR rs1137101 G>A polymorphism was difference among different populations, but in stratified analysis by race, null relationship was found. Additionally, results of stratified analyses by sample size and source of control both found no relationship between LEPR rs1137101 G>A polymorphism with overall cancer susceptibility, highlighting that these variables could not influence the negative findings either.
According to the findings of stratified analysis, we found that LEPR rs1137101 locus might be associated with the susceptibility of oral and oropharyngeal cancer. However, the increased risk of cancer was dubious and hard to explain. Sample size was an important factor to determine the relationship between LEPR rs1137101 G>A polymorphism and cancer risk. In this subgroup, only 460 oral and oropharyngeal cancer cases and 503 controls were included for analysis, the findings might be underpowered.
Significant heterogeneity was found among the eligible studies in multiple genetic models. Thus, we carried out stratified analysis by ethnicity, cancer type, sample size, and source of control. It was obvious that Caucasians, hepatocellular carcinoma, breast cancer, other cancers, small sample sizes (<1000 subjects) and hospital-based case-control studies might contribute to heterogeneity.
In this meta-analysis, as mentioned in results, no publication bias was detected. In addition, findings of sensitivity analysis also indicated that our observation was convincing. Overall, results of this meta-analysis were stable and credible for the studied populations.
However, there were several limitations in the present pooled-analysis. First, some small sample size studies were recruited in this meta-analysis, which could promote the power of study. On the other hand, they led to potential publication bias and significant heterogeneity as well. Second, our findings were based on the crude pooled-results of the eligible case-control studies, the adjustment on characteristics and risk factors (e.g. BMI, physical exercise, age, gender, smoking, alcohol consumption, vegetable and fruit intake, and so on) was not performed. In the future, a more detailed assessment is needed, in which the characteristics and potential risk factors should be considered to adjust the findings. Third, although the cases and controls in eligible studies were fully matched, there was heterogeneity in different cancer types which may influence our results. Fourth, only PubMed and Embase databases were searched to retrieve the eligible. Finally, significant heterogeneity was found in all genetic models, which might influence the findings of our study. Thus, these results should be interpreted with caution.
In summary, this meta-analysis may be the largest sample size so far to assess the potential association of cancer risk with rs1137101 G>A polymorphism in LEPR gene. There is no significant association of cancer risk was identified to be correlated with rs1137101 G>A variants in the overall comparison, and the similar findings was also found in stratified analysis by ethnicity, cancer type, sample size, and source of control. In the future, more large-scale studies are needed to confirm or refute our findings.