Correlation of E-cadherin gene polymorphisms and epidermal growth factor receptor mutation in lung adenocarcinoma.

Epithelial-mesenchymal transition (EMT) was recently discovered related to the efficacy of epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) in NSCLC patients and cell lines. In this study, we aimed to explore the association among the E-cadherin gene (CDH1) genetic variants, TK-domain mutations of EGFR, and clinicopathologic characteristics in patients with lung adenocarcinoma. A total of 280 patients with lung adenocarcinoma were recruited between years 2012 and 2015. All subjects underwent the analysis of CDH1 genetic variants (rs16260 and rs9929218) by real-time polymerase chain reaction (PCR) genotyping. The results showed that CA and CA + AA genotypes of CDH1 single nucleotide polymorphism (SNP) rs16260 were significantly reverse associated with EGFR mutation type (Adjusted odds ratio (AOR) = 0.43, 95% CI = 0.20-0.92 and AOR = 0.46, 95% CI = 0.22-0.96, respectively) in female lung adenocarcinoma patients. Moreover, the significantly reverse associations between CA and CA + AA genotypes of CDH1 rs16260 and EGFR hotspot mutations, namely L858R mutation and exon 19 in-frame deletion, were also demonstrated among female patients. Besides, CA + AA genotype of CDH1 rs16260 was noted significantly reverse associated with the tumor sizes (OR = 0.31, 95% CI = 0.12-0.80; p = 0.012). In conclusion, our results suggested that CDH1 variants are significantly reverse associated with mutation of EGFR tyrosine kinase, especially among the female patients with lung adenocarcinoma. The CDH1 variants might contribute to pathological development in lung adenocarcinoma.

Introduction

Lung cancer is the most common cancer worldwide and also represented the most common death in Taiwan. Based on the National Health Insurance Research Database published in Taiwan, the 5-year survival rate of lung cancer patients was 15.9% 1. The 2015 WHO classification relies on a greater extent of immunohistochemical characterizeation, which allows for precise subtyping, conduction of appropriate treatment strategy, and predicting clinical course. Moreover, the molecular characterization of patients with non-small cell lung cell (NSCLC) is resulting in the use of agents with high levels of antitumor activity, particularly for those with driver mutations 2. The most notorious ones are the mutations in the epidermal growth factor receptor (EGFR) and rearrangement of the anaplastic lymphoma kinase (ALK) gene or ROS1 gene, and mutation in BRAF V600E. Mutations in the EGFR tyrosine kinase are noticed in lung adenocarcinoma and occur more frequently in females and non-smokers 3. In Asia, the incidence of EGFR mutation goes even up to 62% 4.

E-cadherin (CDH1 gene) is a Ca2+ dependent homotypic cell adhesion molecule (CAM) that is important in the formation of adherens junctions to bind cells altogether and functions as a binding partner for β-catenin 5. The E-cadherin plays a key role in cellular adhesion, and its down-regulation is potentially important and highly associated with greater tumor metastasis 6-10. E-cadherin is regarded as an important factor for Epithelial- mesenchymal transition (EMT), which is the critical step for cancer metastasis 11, 12.

Of recent knowledge, EMT is associated with the mutant status of EGFR and the efficacy of EGFR-tyrosine kinase inhibitors (TKIs) in lung cancer cell lines as well as NSCLC patients. Also, the reduction of E-cadherin expression predicts worse overall survival (OS) and disease-free survival/ progression-free survival (DFS/PFS) in patients with NSCLC 13, 14. Furthermore, decreased expression of E-cadherin was associated with decreased sensitivity to EGFR-TKIs, whereas high E-cadherin expression improved tumor cells' sensitivity to EGFR-TKIs 15. However, the correlation between CDH1 gene polymorphisms and EGFR mutations of lung adenocarcinoma has not been well-studied. In this study, we hypothesized that the genetic polymorphisms of CDH1 (rs16260 and rs9929218) may have an effect on the TK-domain mutations of EGFR in patients with lung adenocarcinoma.

Material and Methods

Study Population

Between years 2012 and 2015, a total of 280 patients with lung adenocarcinoma at Cheng-Ching General Hospital in Taichung, Taiwan were recruited. The study was approved by the Institutional Review Board of Cheng-Ching General Hospital (No. HP120009; 22 September 2012). All enrolled patients provided signed informed consent to participate in this study.

Study Variables

The main endpoint of the present study was the prevalence of EGFR mutation among these lung adenocarcinoma patients, and its association with E-cadherin (CDH1) genotypes. We selected two CDH1 SNPs, including rs16260 (-160, C/A) in the promoter region and the intron 2 variant rs9929218 based on their potential involvement in the various cancer types 16-20. Different associating factors relating to the mutations in EGFR were considered and analyzed as previously described 21, 22. Data obtained from medical record of each patient included demographics (age, gender), tobacco smoking status, and tumor staging and differentiation.

Patients Specimens for Genomic DNA Extraction and E-cadherin (CDH1) Genotyping

Venipuncture was performed and withdrawn blood from each participant into Vacutainer blood collection tubes containing EDTA and stored at 4℃. Genomic DNA was extracted from QIAamp DNA blood mini kits according to the manufacturer's instructions as previously described 23. Allelic discrimination of CDH1 rs16260 (C_11934298_10) and rs9929218 (C_11509221_10) gene polymorphisms was assessed with the ABI StepOne™ Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) and analyzed using SDS version 3.0 software (Applied Biosystems) with the TaqMan assay.

Statistical Analysis

Categorical variables, including demographics, smoking status, tumor characteristics, and genotypes polymorphisms, were summarized as number and percentage by EGFR mutation status; continuous variables were expressed as mean and standard deviation. The distributions of demographics, clinical characteristics and genotype frequencies among lung adenocarcinoma patients, as well as clinicopathological features in different genotypes, were analyzed with a χ2-test. A p-value of <0.05 indicated statistically significant.

Results

Characteristics of Study Population

This study included a total of 280 patients, 127 males and 153 females with a mean age of 65 years, for analysis. The baseline demographics and clinical characteristic of enrolled patients were shown in Table 1. There were 111 (39.6%) and 169 (60.4%) patients in the EGFR wild-type and mutation type groups, respectively. These two groups differed with respect to gender, tobacco smoking status, and tumor differentiation (p<0.001). The EGFR mutation type group, compared with subjects in the EGFR wild-type group, were predominantly female (64.5% vs. 39.6%, respectively), non-smoker status (77.5% vs. 45.0%), and mostly well-differentiated (12.4% vs. 7.2%) and moderately-differentiated tumor (81.7% vs. 72.1%).

Table 1

Baseline demographic and clinical characteristics of patients with lung adenocarcinoma by EGFR mutation status (N=280)

Variable	Wild-type(N=111)	EGFR mutation (N=169)	p-value
Age, n (%)
<30	1 (0.9%)	1 (0.6%)	p=0.657
30-39	3 (2.7%)	2 (1.2%)
40-49	11 (9.9%)	16 (9.5%)
50-59	21 (18.9%)	44 (26.0%)
60-69	26 (23.4%)	31 (18.3%)
≥70	49 (44.1%)	75 (44.4%)
Mean ± SD	65.36 ± 13.42	65.76 ± 13.57	p=0.810
Gender, n (%)
Male	67 (60.4%)	60 (35.5%)	p<0.001
Female	44 (39.6%)	109 (64.5%)
Tobacco smoking, n (%)
Non-smoker	50 (45.0%)	131 (77.5%)	p<0.001
Smoker	61 (55.0%)	38 (22.5%)
Pack-years ± SD	46.32 ± 28.21	19.94 ± 23.83	p<0.001
Tumor AJCC staging, n (%)
IA	11 (9.9%)	17 (10.1%)	p=0.570
IB	9 (8.1%)	23 (13.6%)
IIA	5 (4.5%)	7 (4.1%)
IIB	1 (0.9%)	0 (0%)
IIIA	10 (9.0%)	11 (6.5%)
IIIB	17 (15.3%)	19 (11.2%)
IV	58 (52.3%)	92 (54.4%)
Tumor differentiation, n (%)
Well	8 (7.2%)	21 (12.4%)	p=0.001
Moderate	80 (72.1%)	138 (81.7%)
Poor	23 (20.7%)	10 (5.9%)

Distribution of CDH1 Genotypes of Study Population and Its Association with EGFR Mutation by Gender Difference

The distribution frequency of CDH1 genotypes (rs16260 and rs9929218) of patients with lung adenocarcinoma was shown in Table 2. The alleles with the highest distribution frequency for rs16260 and rs9929218 in the enrolled patients were homozygous C/C and homozygous G/G for both EGFR wild-type and mutation type groups, respectively. After adjusting for variance, there was no significant difference between the wild-type and mutation type of EGFR and polymorphisms of the CDH1 genotypes in rs16260 and rs9929218, when compared with wild-type individuals.

Table 2

Distribution frequency of CDH1 genotypes of patients with lung adenocarcinoma and multiple logistic regression analysis of EGFR mutation association

	All cases (N=280)			Male (N=127)			Female (N=153)
GenotypesSNP	Wild type(N=111)n (%)	Mutation type(N=169)n (%)	AOR(95% CI)	Wild type(N=67)n (%)	Mutation type(N=60)n (%)	AOR(95% CI)	Wild type(N=44)n (%)	Mutation type (N=109)n (%)	AOR(95% CI)
CDH1 rs16260
CC	55 (49.5%)	90 (53.3%)	1.00	38 (56.7%)	26 (43.3%)	1.00	17 (38.6%)	64 (58.7%)	1.00
CA	49 (44.1%)	70 (41.4%)	0.90 (0.53-1.53)	24 (35.8%)	31 (51.7%)	1.95 (0.92-4.14)	25 (56.8%)	39 (35.8%)	0.43 (0.20-0.92)
AA	7 (6.3%)	9 (5.3%)	0.81 (0.27-2.45)	5 (7.5%)	3 (5.0%)	0.90 (0.19-4.31)	2 (4.6%)	6 (5.5%)	0.72 (0.13-3.90)
CA+AA	56 (50.5%)	79 (46.7%)	0.89 (0.54-1.48)	29 (43.3%)	34 (56.7%)	1.77(0.85-3.65)	27 (61.4%)	45(41.3%)	0.46 (0.22-0.96)
CDH1 rs9929218
GG	68 (61.3%)	108 (63.9%)	1.00	43 (64.2%)	34 (56.7%)	1.00	25 (56.8%)	74 (67.9%)	1.00
GA	41 (36.9%)	56 (33.1%)	0.87 (0.51-1.49)	23 (34.3%)	25 (41.7%)	1.37 (0.65-2.89)	18 (40.9%)	31 (28.4%)	0.54 (0.25-1.15)
AA	2 (1.8%)	5 (3.0%)	1.17 (0.21-6.57)	1 (1.5%)	1 (1.6%)	0.99 (0.05-18.34)	1 (2.3%)	4 (3.7%)	1.14 (0.12-10.79)
GA+AA	43 (38.7%)	61 (36.1%)	0.89 (0.52-1.50)	24 (35.8%)	26 (43.3%)	1.36 (0.65-2.83)	19 (56.8%)	35 (32.1%)	0.57 (0.27-1.20)

The AORs with 95% CIs were estimated by multiple logistic regression models after controlling for age and smoking.

Note: Bold text indicated a significant association with p value <0.05.

Abbreviations: SNP, single nucleotide polymorphism; AOR, adjusted odds ratio; CI, confidence interval.

To elucidate the association between the polymorphism of CDH1 gene and EGFR mutations in different gender, the distribution frequency of CDH1 single nucleotide polymorphism (SNP) (rs16260 and rs9929218) of EGFR wild-type and mutation type in lung adenocarcinoma patients was estimated in Table 2. In females, CA and CA + AA genotypes of CDH1 rs16260 were significantly reverse associated with EGFR mutation type (AOR = 0.43, 95% CI = 0.20-0.92 and AOR = 0.46, 95% CI = 0.22-0.96, respectively). Hence, further analyses were focused on the association between the polymorphism of CDH1 gene and EGFR hotspot mutations in female patients with lung adenocarcinoma.

Association between Polymorphisms of CDH1 genotypes and EGFR Hotspot Mutations among the Female Lung Adenocarcinoma Patients

Table 3 showed the association between the polymorphisms of CDH1 and the EGFR hotspot mutation in the female patients. Also, the significantly reverse associations between CA and CA + AA genotypes of CDH1 rs16260 and EGFR hotspot mutations, namely L858R mutation (OR = 0.40, 95% CI = 0.18-0.90 and OR = 0.44, 95% CI = 0.20-0.97, respectively) and exon 19 in-frame deletion (OR = 0.39, 95% CI = 0.16-0.95 and OR = 0.41, 95% CI = 0.17-0.97, respectively), were demonstrated.

Table 3

The associations between the polymorphisms of CDH1 and the EGFR hotspot mutations in female patients with lung adenocarcinoma.

	Wild-type	L858R			Exon 19 in-frame deletion
Genotypes	(N=44) n (%)	(N=61) n (%)	OR (95% CI)		(N=43) n (%)	OR (95% CI)
CDH1rs16260
CC	17 (38.6%)	36 (59.0%)	1.00		26 (60.5%)	1.00
CA	25 (56.8%)	21 (34.4%)	0.40 (0.18-0.90)		15 (34.9%)	0.39 (0.16-0.95)
AA	2 (4.6%)	4 (6.6%)	0.94 (0.16-5.67)		2 (4.6%)	0.65 (0.08-5.10)
CA+AA	27 (61.4%)	25 (41.0%)	0.44 (0.20-0.97)		17 (39.5%)	0.41 (0.17-0.97)
CDH1rs9929218
GG	25 (56.8%)	41 (67.2%)	1.00		29 (67.4%)	1.00
GA	18 (40.9%)	17 (27.9%)	0.58 (0.25-1.32)		13 (30.2%)	0.62 (0.26-1.52)
AA	1 (2.3%)	3 (4.9%)	1.83 (0.18-18.56)		1 (2.3%)	0.86 (0.05-14.51)
GA+AA	19 (43.2%)	20 (32.8%)	0.64 (0.29-1.43)		14 (32.6%)	0.64 (0.27-1.52)

Note: bold text indicated a significant association with p value <0.05.

Abbreviations: SNP, single nucleotide polymorphism; OR, odds ratio; CI, confidence interval.

Association between CDH1 SNP rs16260 and Tumor Classification among Male Lung Adenocarcinoma Patients

The AJCC Tumor, Node, Metastasis (TNM) staging system for lung cancer (NSCLC) is an internationally accepted system to describe the extent of disease 24. It combines features of the tumor into disease stage groups that correlate with survival and are linked to recommendations for treatment, as well as an indicator of prognosis. In this study, we further investigated the association between polymorphisms of CDH1 gene and clinicopathologic characteristics among male patients with lung adenocarcinoma. As shown in Table 4, CA + AA genotype of CDH1 rs16260 was noted significantly reverse associated with the “T” classification (primary tumor size and extension) based on eighth edition of AJCC TNM staging system (OR = 0.31, 95% CI = 0.12-0.80; p = 0.012). Furthermore, while stratified male patients based on EGFR mutation status, consistent reverse association was seen in the mutation type male patients (OR = 0.20, 95% CI = 0.04-0.99; p = 0.037). These findings indicated that the polymorphisms of CDH1 rs16260 may be associated with the tumor size and extension of lung cancer.

Table 4

The associations between polymorphic genotypes of CDH1 rs16260 and clinicopathologic characteristics of male patients with lung adenocarcinoma.

	Tumor AJCC “T” Classification
Variable genotypic frequencies	T1 & belown (%)	T2 & aboven (%)	OR(95% CI)	p value
All cases (N=127)	(N=25)	(N=102)
CDH1 rs16260
CC	7 (28.0%)	57 (55.9%)	1.00
CA+AA	18 (72.0%)	45 (44.1%)	0.31 (0.12-0.80)	p=0.012
EGFR Wild type (N=67)	(N=13)	(N=54)
CDH1 rs16260
CC	5 (38.5%)	33 (61.1%)	1.00
CA+AA	8 (61.5%)	21 (38.9%)	0.40 (0.12-1.38)	p=0.139
EGFR Mutation type (N=60)	(N=12)	(N=48)
CDH1 rs16260
CC	2 (16.7%)	24 (50.0%)	1.00
CA+AA	10 (83.3%)	24 (50.0%)	0.20 (0.04-0.99)	p=0.037

Note: bold text indicated a significant association with p value <0.05.

Abbreviations: SNP, single nucleotide polymorphism; OR, odds ratio; CI, confidence interval.

Discussion

Our study investigated 280 patients with lung adenocarcinoma, and the role of E-cadherin (CDH1) gene polymorphism in regards to the EGFR mutation status, as well as a possible prognostic marker for tumor invasiveness and metastasis and predictive marker for the resistance of EGFR-TKI therapy. The significance of E-cadherin (CDH1) polymorphism, as an important epithelial marker/factor of EMT, have been well-studied and established in several human cancers, such as breast cancer 25, 26, gastric cancer 27, pancreatic cancer, ovarian cancer 28, 29, prostate cancer 26, esophageal squamous cell carcinoma 30, head and neck cancer 31 and skin cancer 32-34. Regard to lung cancers, though E-cadherin is highly associated with risk of NSCLC carcinogenesis 14, 35, 36, particular adenocarcinoma, the mutant status of EGFR, the efficacy of EGFR-TKIs therapy 37, and survival prognosis, no study exploring the relationship between CDH1 gene polymorphisms and the mutations of EGFR of lung cancers was conducted. To our best knowledge, this is the first time discovering the statistically significant association between CDH1 SNP rs16260 variant types (CA and CA + AA genotypes) and hotspot mutations (in-frame deletion mutation in exon 19 and L858R mutation) of EGFR, especially in the female lung adenocarcinoma patients.

Advanced NSCLCs that contains characteristic mutations in EGFR are highly sensitive to EGFR-TKIs. The previous study has reported that EGFR mutation was more prevalent in adenocarcinoma than other types of NSCLCs, as well as female and non-smokers 3. The Asian population has the highest incidence of EGFR mutation in the world 4. In Taiwan, Hsu et al. performed a study based on the National Taiwan Lung Cancer Registry and detected the EGFR mutation rate higher than 50% 38. Indeed, as shown in our results, higher frequency of EGFR mutation type was observed in the female patients (female vs. male = 64.5% vs. 35.5%) and in non-smokers (non-smoker vs. smoker = 77.5% vs. 22.5%). These results were consistent with previous studies indicating that the mutation of EGFR was associated with gender, adenocarcinoma, and smoking status 3, 4, 38.

There were very limited studies investigating polymorphisms of different EMT-related genes expression and their relationship with carcinogenesis. Xie et al. 39 have reported lately that EMT-related gene variants, namely ADGRF1, NOTCH3, and CDH1, may be involved in susceptibility to NSCLC. Cai et al. 40 earlier revealed that genetic effect of a nonsynonymous HRH4 polymorphism (rs11662595) is a loss-of-function polymorphism that results in dysfunction of HRH4 gene and attenuates the anti-EMT function of HRH4 in NSCLC. This investigation provides a promising biomarker for prognosis and therapy of NSCLC. Kim S et al. 41 also demonstrates that EMT phenotype was related to PD-L1 overexpression in pulmonary adenocarcinoma cells and patients with EMT-phenotype pulmonary adenocarcinoma may benefit from PD-1/PD-L1-blocking immunotherapy. But to the detail of CDH1 gene polymorphisms and their association with EGFR mutation status, no specific research was found prior to ours.

Our findings relative to association between CDH1 polymorphism and EGFR mutation status suggested CA and CA + AA genotypes of CDH1 SNP rs16260 were significantly reverse associated with EGFR mutation type in female patients with lung adenocarcinoma (AOR = 0.43, 95% CI = 0.20-0.92 and AOR = 0.46, 95% CI = 0.22-0.96, respectively), as compared to the control (CC genotype). AA genotype of the same SNP also showed the same direction of association (AOR = 0.72, 95% CI = 0.13-3.90), but did not reach statistical significance. We further verified this association with the use of two EGFR hotspot mutations, namely L858R mutation and in-frame deletion mutation in exon 19, and obtained the consistent statistically significant relationship. These results may explain that CDH1 variant type is less associated with EGFR mutation type especially among the female population; in the other words, CDH1 wild type (CC genotype) is more related to mutate EGFR tyrosine kinase, which is a more favorable prognosis and strongly predicts for sensitivity to EGFR-TKIs. However, the mechanism by which this SNP (rs16260) modulates the roles of female lung adenocarcinoma patients should be further investigated.

Among male lung adenocarcinoma patients, we found no statistical significance between CDH1 polymorphism and EGFR mutation status. But when we further stratified them based on tumor characteristic of AJCC “T” classification (primary tumor size and extension), CDH1 rs16260 variant type (CA + AA genotype) showed significantly reverse association as T classification progressed, as compared with the control type (CC genotype). This association was still seen in male patients with mutated EGFR gene. In the previous studies, downregulation of E-cadherin expression is highly associated with cancer progression and metastasis 13, 42-45. The inclusion of other clinicopathologic parameters at molecular level should be further investigated.

In conclusion, our results suggested that E-cadherin gene (CDH1) variants are significantly reverse associated with mutation of EGFR tyrosine kinase, especially among the female patients with lung adenocarcinoma. This may be utilized as a prognostic factor for tumor size in Taiwanese patients with lung adenocarcinoma.