Association between EGFR Gene Mutation and Antioxidant Gene Polymorphism of Non-Small-Cell Lung Cancer

EGFR mutation status is considered as an important predictor of therapeutic responsiveness in non-small-cell lung carcinoma patients. Recent evidence suggests that antioxidant gene polymorphisms are potential predictors of lung cancer risk. Thus, stratification of EGFR mutation-related phenotypes by antioxidant gene polymorphism status can be an effective approach in terms of improving the prognosis of lung cancer patients. The present study was designed to evaluate the distribution frequency of antioxidant gene polymorphisms in lung adenocarcinoma, as well as its association with hotspot EGFR mutations. The study findings revealed that a statistically significant association exists between EGFR L858R mutation and AG + GG genotypes of SOD rs4880 polymorphism. Furthermore, the subgroup analysis data revealed that compared to AA genotype of SOD rs4880, AG + GG genotypes were significantly associated with advanced cancer stage and distant metastasis. Taken together, these findings can be utilized clinically to predict cancer aggressiveness, metastatic, potential and therapeutic responsiveness of lung cancer patients.


Introduction
Lung cancer is the leading cause of death worldwide. It is a highly invasive and metastatic cancer with very poor prognosis. The death rate associated with lung cancer is higher than that observed in colon, breast and pancreatic carcinomas, and the 5-year survival rate of lung cancer is lower than 18% [1]. Among all lung cancer types, small-cell lung carcinoma and non-small-cell lung EGFR-mutated lung adenocarcinoma. Demographic characteristics and medical information of the patients, including gender, smoking status, AJCC clinical staging, tumor 'T' classification, lymph node status, distant metastasis, and tumor differentiation, were obtained from their medical records. For EGFR gene sequencing, paraffin-embedded cancer tissues were collected; in addition, whole blood samples were collected from the participants and kept in EDTA-containing sterile tubes for genotyping. This study was approved by the Institutional Review Board (IRB) of Changhua Christian Hospital at Changhua (IRB No. 140205, date of approval 9 March 2014) and Cheng-Ching General Hospital (No. HP120009, 22 September 2012).

Statistical Analysis
The distributions of patient demographic features and genotype frequencies between wild-type and EGFR-mutated lung adenocarcinomas, as well as the clinicopathological features of EGFR-mutated lung adenocarcinoma patients in polymorphic genotypes of SOD rs4880 were analyzed by χ 2 -test. After controlling for other covariables, the odds ratio and 95% CIs of the association between the genotype frequencies and EGFR mutation risk and the clinicopathological features were calculated using multiple logistic regression models. A p value of <0.05 was considered statistically significant. All statistical analyses were carried out using SAS statistical software (Version 9.1, 2005; SAS Institute, Cary, NC, USA).

Demographic and Clinical Characteristics of Participants
A total of 314 lung adenocarcinoma patients were enrolled for the study. Table 1 shows the demographic and clinical details of the participant. Of all participants, 117 had lung adenocarcinoma with wild-type EGFR (the WT group) and 197 had EGFR-mutated lung adenocarcinoma (the EGFR-mutated group). The average ages of the participants in the WT and EGFR-mutated groups were 65.7 and 65.1, respectively. There were statistically significant differences between the WT and EGFR-mutated groups in regard to gender, smoking status, tumor 'T' classification and tumor differentiation. Than the WT group, the EGFR-mutated group had higher numbers of female (126 vs. 48) and non-smoker (158 vs. 57) participants. Moreover, the numbers of participants with well-(26 vs. 7) and moderately differentiated (158 vs. 89) tumors were higher in the EGFR-mutated group. SD is the abbreviation of standard Deviation. AJCC is the abbreviation of American Joint Committee on Cancer. The AJCC Cancer Staging Manual remains the gold standard reference for oncologists, surgeons, pathologists, radiologists, cancer registrars and medical professionals world-wide to ensure that all those caring for cancer patients are fully versed in the language of cancer staging. *: p value is less than 0.05. Table 2 shows the distribution frequency of antioxidant gene polymorphisms (SOD rs5746136 and SOD rs4880; CAT rs769218; OGG1 rs1052133; and TXN2 rs4821494) of lung adenocarcinoma patients. The alleles with the highest distribution frequency for SOD rs5746136 and SOD rs4880 in study participants were homozygous C/C and homozygous A/A for both the WT and EGFR-mutated groups, respectively. For CAT rs769218; OGG1 rs1052133; and TXN2 rs4821494, the alleles with the highest distribution frequency among participants were homozygous G/G, homozygous C/C and homozygous G/G, respectively, for both the WT and EGFR-mutated groups. For all antioxidant gene polymorphisms, there was no statistically significant association between different genotypes and EGFR mutation status in lung adenocarcinoma patients.  Table 3 represents the association between antioxidant gene polymorphisms (SOD rs5746136 and SOD rs4880; CAT rs769218; OGG1 rs1052133; and TXN2 rs4821494) and EGFR hotspot mutations in study participants. In the present study, two hotspot EGFR mutations including L858R and exon 19 in-frame deletions were evaluated. A significant association between L858R mutation and AG + GG genotypes of SOD rs4880 polymorphism was observed (AOR = 1.90; 95% CI = 1.01-3.58; p = 0.047). However, there was no statistically significant association between SOD rs4880 genotypes and exon 19 in-frame deletions. For SOD rs5746136, CAT rs769218, OGG1 rs1052133 and TXN2 rs4821494 polymorphisms, no statistically significant association was observed between different genotypes and EGFR hotspot mutations. Adjusted odds ratios (AORs) with 95% confidence intervals (CIs) estimated by logistic regression models after controlling for age and gender. *: p value is less than 0.05.

Subgroup Analysis of EGFR Mutations Based on Polymorphic Genotypes of SOD rs4880
To evaluate the correlation between clinicopathological features of EGFR hotspot mutations and SOD rs4880, a subgroup analysis of all lung adenocarcinoma cases as well as EGFR L858R and Exon 19 deletion mutations were performed ( Table 4). The subgroup analysis data revealed that compared to AA genotype, AG + GG genotypes of SOD rs4880 were associated with more aggressive lung adenocarcinoma phenotypes in terms of cancer staging, tumor size, lymph node status, distant metastasis and tumor differentiation. Moreover, a statistically significant correlation was observed between AG + GG genotypes and higher cancer staging irrespective of the EGFR mutation status. When considering all lung adenocarcinoma patients (WT and EGFR-mutated), AG + GG genotypes were found to correlate significantly with distant metastasis; however, similar statistically significant correlation was not observed for EGFR-mutated tumors. I + II is population of stage I and stage II, we compared early stage and advance stage. T1 + T2 is population of T1 and T2, this is only analysis tumor size and genetic variants association. AOR is the abbreviation of adjusted odds ratios, adjust gender and age. CI is the abbreviation of confidence interval. *: p value is less than 0.05.

Subgroup Analysis of EGFR Mutations in Non-Smoking Patients Based on Polymorphic Genotypes of SOD rs4880
Finally, we evaluated the correlation between clinicopathological features in non-smoking NSCLC patients with EGFR mutations and SOD rs4880 ( Table 5). The subgroup analysis data revealed that compared to AA genotype, AG + GG genotypes of SOD rs4880 were associated with more aggressive lung adenocarcinoma phenotypes in terms of cancer staging (III + IV). Moreover, a statistically significant correlation was observed between AG + GG genotypes and higher cancer staging with non-smoking status (AOR = 3.15; 95% CI = 1.28-7.72; p = 0.012). Especially highly significant correlation was observed in female non-smoking populations (AOR = 3.55; 95% CI = 1.25-10.13; p = 0.047). Table 5. Clinicopathologic characteristics of lung adenocarcinoma patients with EGFR mutation in non-smoking population, stratified by polymorphic genotypes of SOD rs4880.  I + II is population of stage I and stage II, we compared early stage and advance stage. T1 + T2 is population of T1 and T2, this is only analysis tumor size and genetic variants association. AOR is the abbreviation of adjusted odds ratios, adjust gender and age. CI is the abbreviation of confidence interval. *: p value is less than 0.05.

Discussion
The present study was designed to evaluate the association between antioxidant gene polymorphisms (SOD rs5746136 and SOD rs4880; CAT rs769218; OGG1 rs1052133; and TXN2 rs4821494) and EGFR-mutated lung adenocarcinoma. To the best of our knowledge, this is the first study of its kind, and the study findings may help identify more appropriate treatment approaches for better cancer management.
Our demographical and clinical data indicated that the number of female and non-smoker participants as well as participants with well-and moderately differentiated tumors was significantly higher in the EGFR-mutated group compared to that in the WT group (Table 1). While studying the distribution frequency of antioxidant gene polymorphisms in patients with WT-or EGFR-mutated lung adenocarcinoma, we found no statistically significant association between different polymorphic genotypes and EGFR mutation status ( Table 2). The analysis performed to investigate the relationship between antioxidant gene polymorphisms with EGFR hotspot mutations (L858R and Exon 19 deletion mutations) revealed that a statistically significant association exists between L858R mutation and AG + GG genotypes of SOD rs4880 polymorphism (Table 3). Furthermore, our subgroup analysis data revealed that AG + GG genotypes of SOD rs4880 were associated with more aggressive lung adenocarcinoma phenotypes compared to AA genotype of SOD rs4880. In particular, AG + GG genotypes were found to be correlated significantly with higher tumor stage and tumors with distant metastasis (Table 4). Additionally, we analyzed for clinical correlation in non-smoking NSCLC patients with EGFR mutations and SOD rs4880 (Table 5). Increased risk was observed in females; non-smokers; and with EGFR mutation.
The analysis of antioxidant gene polymorphisms in different cancer types has gained interest because of the significant impact of oxidant/antioxidant balance in the onset and progression of cancer [26,27]. There is a growing-pool of evidence suggesting that antioxidant gene polymorphisms are important predictors of cancer risks [18][19][20]. These studies have indicated that the distribution frequency of different polymorphic genotypes varies between different cancer subtypes, indicating the potential value of antioxidant gene polymorphisms as cancer biomarkers [28]. Moreover, genetic polymorphism-driven changes in the activity of cellular antioxidants can alter the therapeutic responsiveness of cancer patients through increased oxidative stress; thus, genetic polymorphism as a biomarker can be effective in identifying patient-specific therapeutic interventions and augmenting the responsiveness of personalized medicines [29]. The current findings showed that genetic variants in oxidative stress related genes may modify prognosis in EGFR TKIs-treated NSCLC patients. Cellular redox state is associated with the efficacy of EGFR TKIs treatment in NSCLC patients with activating EGFR mutations [21]. SOD2 overexpression was related to metastatic phenotype in cancers [30][31][32]. Several researchers have hypothesized that SOD2 overexpression promotes metastasis by increasing the steady-state concentration of H2O2 [33,34]. In our findings, AG + GG genotypes of SOD rs4880 were associated with more aggressive lung adenocarcinoma phenotypes compared to AA genotype of SOD rs4880 in advance NSCLC. Furthermore, we speculate that, AG + GG genotypes of SOD rs4880 increase SOD and promote NSCLC progression, especially in non-smoking female. Unfortunately, in this study, we do not have survival data to further validate the role of clinicopathological parameters in SNP by the Kaplan-Meier method or multivariate analysis.
Of all antioxidant genes, SOD genetic polymorphism is well-documented in the literature. SOD being the major first-line antioxidant neutralizes highly reactive superoxide free radicals into less reactive hydrogen peroxide [20]. In the present study, the in-depth analysis of the relationship between polymorphic genotypes of SOD rs4880 and clinic-pathologic features of EGFR-mutated lung adenocarcinoma is particularly important, because SOD polymorphism-mediated stratification of EGFR mutation-related phenotypes, which are considered to be a major predictor of therapeutic responsiveness, can be more advantageous and effective in predicting lung cancer risk as well as therapeutic responsiveness.

Conclusions
The present study findings revealed that SOD rs4880 polymorphism is significantly associated with a specific EGFR hotspot mutation, L858R. Moreover, AG + GG polymorphic genotypes of SOD rs4880 are significant correlated with advancer cancer stage and distant metastasis in EGFR-mutated lung adenocarcinoma patients. These findings can be utilized clinically to predict cancer aggressiveness, metastatic potential and therapeutic responsiveness. Funding: This study was funded by the Changhua Christian Hospital (104-CCH-IRP-099) of Taiwan.

Conflicts of Interest:
The authors declare no conflict of interest.