An Analysis of EGFR Mutations among 1506 Cases of Non-Small Cell Lung Cancer Patients in Guangxi, China

An association between epidermal growth factor receptor (EGFR) and clinical characteristics of non-small cell lung cancer (NSCLC) was reported ten years ago. In addition, a different type of relationship was seen in different ethic races. However, the relationship between these factors is not well understood in the Guangxi province. Up to now, there are only very limited data on the association of TTF1/EGFR protein positivity and EGFR mutation status in NSCLC. This study aims to investigate the role of EGFR gene mutation status on the clinical characteristics and the relationship with TTF-1/EGFR protein positivity of patients with NSCLC in Guangxi, China. 1506 samples from different patients with NSCLC were detected by amplification refractory mutation system for 29 hotspot mutations. Analysis of the relationship between clinical characteristics and EGFR mutation status was performed by using the crosstabs Chi-square and SPSS 21.0 software. Of 1506 samples, 537 (35.7%) revealed tyrosine kinase inhibitor (TKI) sensitive EGFR mutations with 27 (1.8%) cases harboring TKI resistant EGFR mutations or union co-existing EGFR-TKIs sensitive mutations. EGFR-TKIs sensitive mutations were not significantly associated with age and TNM-M stage (P = 0.863; P = 0.572, respectively). However, they were significantly associated with p-stage, TNM-T stage and TNM-N stage (P = 0.011, P < 0.001, P = 0.036, respectively). Immunohistochemical studies revealed that TTF-1 and EGFR protein expression level were all associated with EGFR mutation status (P < 0.001, P = 0.002, respectively). Of the 537 EGFR-TKIs sensitive mutation cases, the rates of exon 19-del, 18 G719X point, exon 21 L858R and L861Q points were 54.6, 0.9, 42.3 and 0.9%, respectively. EGFR TKI-sensitive mutations commonly occur in female, non-smoking and adenocarcinoma patients. The p-stage, TNM-T stage, TNM-N stage, EGFR and TTF-1 protein expression levels have close relationships with EGFR mutation status.


Introduction
According to the latest global cancer statistics, about 1.8 million new lung cancer cases occurred in 2012, accounting for nearly 13% of the total cancer diagnoses [1]. Lung cancer was still the leading cause of cancer death among males and the leading cause of cancer among females in more developed countries. It was also the second leading cause of cancer death in less developed countries among females [1]. Non-small cell lung cancers (NSCLCs) account for approximately 85% of primary lung cancer cases [2,3].
With the emergence of targeted therapy, there are new treatment choices for patients with NSCLCs. Patients with advanced NSCLCs can benefit from the molecular targeted therapeutic drugs such as tyrosine kinase inhibitors (TKIs) [4][5][6]. The majority of patients who have a good response to TKIs are within a group that possesses the epidermal growth factor receptor (EGFR) active mutation [7][8][9]. The NSCLC patients with EGFR mutation showed a long progression-free survival (11.5 months) [10] and median survival time (15.4 months) [7] compared to the EGFR mutation negative patients. EGFR mutation was one of the most common mutations in NSCLC patients [11,12]. The EGFR mutations showed a close relationship with clinical characteristics of patients with lung cancer and mainly appeared in non-smoking female patients with adenocarcinoma lung cancers [12][13][14]. The frequency of EGFR mutations have a wide variation worldwide, ranging from 31% to 66% in Asian patients [15][16][17] and 7.5% to 40% in Caucasians [18,19].
Thyroid transcription factor-1 (TTF-1) is potentially an important marker to distinguish NSCLCs histological subtype. Ninety-five percent of NSCLCs were positive for expression to this marker. In the subtypes of NSCLCs, the expression of TTF-1 in adenocarcinomas is significantly higher than that for squamous cell carcinomas. Previous studies have shown that high TTF-1 expression predicted better prognosis and it can be an independent predictor of favorable prognosis in adenocarcinoma patients [20,21]. Furthermore, we learnt in previous studies that TTF1-positive tumours are found more frequently in female patients and in non-smokers [22,23]. It would be interesting to discover whether there is a correlation between TTF1 status and EGFR mutation status in the patients from the Guangxi province of China. Therefore, detection of expression of TTF-1 protein may contribute to the decision whether to use TKIs drugs for subsequent treatment of patients. However, the association between TTF-1 protein expression and EGFR mutation in other ethnicities is not well understood. Therefore, we performed a cross-sectional study with NSCLCs patients and explored the associations between their clinical characteristics and EGFR mutational status.

Ethics Statements
The local ethics committee of the Affiliated Tumor Hospital of Guangxi Medical University approved this study, and it was conducted in accordance with the Declaration of Helsinki and current ethical guidelines. The patients or the caretakers of patients for their information to be stored in hospital databases and used for research gave written consent. All data underlying the findings described in this manuscript are fully available without restriction.

Clinical Samples
In this single center study, a total of 1506 NSCLCs lung cancer patients that have taken EGFR mutation tests in exon 18-21 were collected consecutively from August 2010 to June 2016. Those with secondary lung cancer and those patients with a doubtful diagnosis were excluded from the study. All the 1506 samples were collected as fresh resection specimens or paraffin embedding tissue section. Before undergoing EGFR detection, all the patients had not undergone TKIs therapy. All these patients were diagnosed by histodiagnosis. Two board-certified pathologists who examined and verified all the diagnoses crosschecked the results. The criteria used for the pathological diagnosis were set according to the standards of the WHO Classification of Lung Tumors in 2016 [24].
In this study, the TNM stage analysis was according to the new lung cancer staging system [25]. At the same time, a p-stage analysis was also performed. Smoking status was classified as non-smokers (<100 lifetime cigarettes), former smokers (quit>1 year prior to diagnosis), or current smokers (still smoking, or quit<1 year prior to diagnosis). EGFR and TTF-1 protein expression were classified as negative and positive.

EGFR Mutation Detection
EGFR mutations were assessed in all samples within the same laboratory. In this study, all cases were analyzed by amplification refractory mutation system (ARMS) using the human EGFR gene mutations fluorescence polymerase chain reaction diagnostic kit (Amoy Dx, Xiamen, China). This kit is well suited for analyzing specimens such as fresh resection tissue, paraffin embedding tissue sections and peripheral blood. This EGFR kit detects 29 mutations in exons 18 to 21, including T790M, L858R, L861Q, S768I, G719S, G719A, and G719C; three insertions in exon 20; and 19 deletions in exon 19. The kit has its own internal and external controls to monitor the reagents, the quality of DNA obtained and the experimental procedure. In addition, the positive and negative controls would be included along with the samples in each batch of PCR reactions to monitor the cleanliness of environment and correctness of the procedures used. This kit was designed to target the EGFR gene in a relatively conservative section consisting about 100bp, and it even allows for DNA degradation in paraffin embedding tissue sections, such that the internal and external controls can give an estimate of the DNA concentration in the tissue. As usual, the PCR amplification procedure was performed over three stages. The first stage was a 5-minute initial denaturation at 95˚C followed by the second stage which consisted of 15 cycles of 25 seconds at 95˚C, 20 seconds at 64˚C, and 20 seconds at 72˚C and the third stage consisted of 31 cycles of 25 seconds at 93˚C, 35 seconds at 60˚C, and 20 seconds at 72˚C.The data were collected at 60˚C during stage 3.
DNA from all specimens used were extracted using a TIANamp genomic DNA kit (Tiangen Biotech Co., Ltd, Beijing, China) and all subsequent assays followed the manufacturer's protocol, and were detected using the Roche LightCycler 480 real-time PCR system. The LightCycle Adapt software (LightCycler 480 Software, v1.5) was use for analysis of the realtime PCR data. The manufacturer's instructions were used to judge for positive or negative results.

TTF-1 and EGFR Protein Expression Assessment
The mouse anti-TTF-1 monoclonal antibody kit (ZSGB-BIO, ZM-0270) assessed TTF-1 expression. The mouse anti-EGFR monoclonal antibody kit (ZSGB-BIO, ZM-0093) assessed EGFR protein expression.by. All the operations were in accordance with the manufacturer's instructions. The TTF-1 protein was expressed positively in neoplastic cell nuclei and this was shown as tan or brown. The EGFR protein was expressed positively in cell membranes and this was shown as pale yellow or brown as a granular uniform distribution. The assessment of TTF-1 and EGFR protein expression levels fell into two categories: TTF-1/ EGFR positive expression and TTF-1/ EGFR negative expression.

Statistical Analysis
In this study, all the EGFR gene mutation results tested were sorted into two categories: wild type and EGFR mutation positive. In addition, the positive mutations were divided into two sections: EGFR-TKIs sensitive mutations (including exons 18, 19 and 21) and EGFR-TKIs resistant or co-existing mutations (exon 20 or resistant co-existing sensitive mutations). The assessment of TTF-1 and EGFR protein expression levels fell into two categories: TTF-1/ EGFR positive expression and TTF-1/ EGFR negative expression. Analysis of the relationship between clinical characteristics and EGFR mutation status was performed by using the Crosstabs Chi-square and SPSS 21.0 software.

Results
A total of 1506 patients with NSCLCs were analyzed. The mid-age was 58 years (range from 23-85 years). In all the NSCLCs cases, the patients' characteristics associated with the mutation status are shown in Table 1.

Frequency of EGFR Mutations and Clinical Characteristics
The frequency of EGFR mutations associated with clinical characteristics is shown in Table 1. There is no significant difference of EGFR-TKIs sensitive mutations in the age of patients (p = 0.863). However, a significant difference was observed with gender (p < 0.001). There is no significant difference of EGFR-TKIs resistance or co-exiting mutations in the two genders (p = 0.220). In male patients, 28.2% (268/951) harbored EGFR-TKIs sensitive mutations and 1.5% (14/951) harbored resistant or co-existing (sensitive and resistant) mutations. Of 555 female patients, 48.5% (269/555) harbored EGFR-TKIs sensitive mutations and 2.3% (13/555) harbored resistant or co-existing mutations.

The Relationship between Mutation Status and Immunohistochemistry Analysis
TTF-1 and EGFR protein expression were assessed by immunohistochemistry (Figs 1 and 2

Characteristics of EGFR Mutations
The frequency of EGFR-TKIs sensitive mutations is shown in  (Table 3). The frequency of EGFR-TKIs resistant or co-existing (sensitive and resistant) mutations is shown in Table 4. Approximately 1.8% of samples analyzed harbored EGFR-TKIs resistant or co-existing mutations, and the details are shown in Table 5.

Discussion
This study shows that the EGFR mutation is common in females, non-smokers and those patients with lung adenocarcinoma. However, no association between EGFR mutation and age was observed. This is in accordance with other studies in this field [15,16,19,26].
In the cohort of this study, we demonstrate the rate of EGFR-TKIs sensitive mutation in NSCLCs is 35.7% (537/1506), a finding which is in accordance with a previous study in China [17]. However, the EGFR-TKIs sensitive mutation rate of 41.8% in lung adenocarcinoma is lower than the rate of 66.3% in a study with a cohort in Southern China [16], but much higher than the mutation rate of 7.5% in the study with a cohort of Northern European populations [27,28]. The reasons for the difference in different regions may be due to different detection techniques or different populations having different gene backgrounds. Gender and smoking status both show significant differences with mutation status, respectively. According to other studies [18,19,27], it is clear that the high rate of EGFR mutation in females is similar to the high rate of EGFR mutation in the non-smoking population. The EGFR mutation rate in males resembles the rate in the smoking population. Therefore, we have reasons to believe that the different genders have different distributions depending on their smoking status (smokers are mainly distributed in males) and this is one reason that accounts for the difference in mutation rates among the two genders. Other reasons maybe the variations of the physiological makeup of males and females. However, the mechanisms by which gender and smoking status contribute to EGFR mutation is still largely unknown. In addition, the interactions between gender and smoking status still remain unknown.
With regards to the EGFR mutations in the different histological subtypes, this study showed that the highest EGFR-TKIs sensitive mutation frequency occurred in adenocarcinoma (41.8% (493/1178)) and this was followed by adenosquamous carcinoma, (30.9% (21/68)), squamous cell carcinoma (6.6% (12/181)) and others (6.1% (11/79)). There is a significant difference with respect to EGFR-TKIs sensitivity in the various subtypes. However, there is no significant difference regarding EGFR-TKIs resistant or co-existing mutations in the various subtypes. It is clear that the occurrence of EGFR-TKIs sensitive mutation in non-adenocarcinoma was relatively rare. In addition, a previous study reported that the patients who expressed sensitive mutations in non-adenocarcinoma responded poorly and non-effectively to the EGFR-TKIs [29], and hence whether it is even a necessity to undergo a routine mutation test in the absence of adenocarcinomas. However, in the study of Chiu and coworkers, they pointed out that it is useful to have an EGFR mutation test for those patients who have a poor performance status or other comorbidities [30]. In this study, although the rate of EGFR-TKIs sensitive mutation in squamous cell carcinoma and adenosquamous carcinoma was obviously lower than with adenocarcinoma, but it is clear that the rate is higher than other geographical regions [27]. Therefore, we would advise that there is a necessity to have an EGFR mutation test no matter what subtypes are presented, especially in those patients who failed to get benefit from other therapies but where EGFR-TKIs target therapy is being administered.
Based on our analysis, we can hypothesis that the patients with a smaller in situ tumor diameter can be more suited for surgery without any other medical treatment and therefore can avoid the effect of medicines on gene detection. In addition, in the patients with a larger in situ tumor diameter and hence are at an advance stage of lung cancer, are most likely to be inoperable. Those patients with a large tumor diameter may well have had chemo-or radiotherapy to shrink the tumor diameter to suitable size for surgery. Therefore, we can see that the patients with smaller diameter tumors show a high rate of EGFR-TKIs sensitive mutation and the patients with larger diameter tumors show a high rate of EGFR-TKIs resistant or coexisting mutations. Therefore, it is important to detect an EGFR mutation as early as possible, and this will determine the formulation of treatment for the individual patient.
There is some controversy regarding the relationship between EGFR protein expression and EGFR gene mutation. A study in a cohort of 119 Norwegian patients showed that there is no significant relationship between EGFR protein expression and EGFR gene mutation [27]. However, the study in a cohort of Korean showed a close relationship between EGFR protein expression and EGFR gene mutation [31]. In the present study, 1506 cases were examined which is relatively larger than the above studies, and EGFR protein expression level showed a close relationship with EGFR-TKIs sensitive mutation with EGFR protein positive-expression being more frequent in the patients who harbored EGFR mutation. The size of sample cases studied maybe one reason for the different results seen. Other reasons that may account for the differences are ethnic groups or differences in experimental methods and evaluation systems employed. It is probable that gene mutations can lead to high gene expression levels and this would in turn cause EGFR protein positive-expression. Therefore, it is probable that the EGFR protein expression is closely related to EGFR-TKIs sensitive mutation.
TTF-1 is a nucleoprotein with a molecular weight of 40kD and is one of the members of the gene transcription NKx2 family. It was discovered as a transcriptional regulator of thyroglobulin in 1989, and the first studies in relation to lung cancer were conducted in 2007 [32].
Previous studies have shown that the patients with positive TTF-1 protein expression can show improvement in lung adenocarcinoma survival [33][34][35]. In this study, we find that TTF-1 protein expression show a close relationship with EGFR mutation. Positive TTF-1 expression was more frequent in those patients who harbored EGFR mutation. This result is in accordance with studies performed in Chinese, Korean and Caucasian populations [18,31,36]. It is not always possible to detect EGFR gene mutation when the tissue samples obtained are in a poor condition and in those cases it is useful to evaluate TTF-1 or EGFR protein expression as this may give a valid basis for the development of strategies for clinical care for patients with NSCLC.
With respect to the EGFR mutation subgroup analysis, we can demonstrate that 19-del and L858R are the two main subtypes of EGFR-TKIs sensitive mutations in Chinese patients with NSCLCs, including 54.6% and 42.3% of all the EGFR-TKIs sensitive mutations, respectively. There is no significant difference between 19-del and L858R (p = 0.296).This result is in accordance with the study in a cohort of the Chinese population by Lai and coworkers [17]. However, this result is inconsistent with other studies that investigated a cohort of Caucasian [18,37,38]. In the mutation cases, exon 21 and 19 mutation rates show no significant difference with gender, smoking status, pathology, TTF-1 and EGFR protein expression. This result is not consistent with the Japanese study where it was reported that exon 19 deletion mainly occurs in male and exon 21 mainly occurs in females [37].
Although this manuscript has analyzed the status of EGFR gene mutations and EGFR/TTF-1 protein expression levels on 1506 cases of NSCLC patients, the lack of prognoses data and biochemical mechanism research limited the scope of this study.

Conclusions
In summary, in this study based on a large cohort of patients, there were three interesting findings. Firstly, the p-stage has a close relationship with EGFR-TKIs sensitive mutations with the patients who in I stage have the most frequency EGFR-TKIs sensitive mutations. Secondly, the TNM-T stage has a close relationship with EGFR mutations, and that the early TNM-T stage shows more common EGFR-TKIs sensitive mutations whereas the late TNM-T stage shows more common EGFR-TKIs resistant or co-existing mutations. Thirdly, the TNM-N stage has a close relationship with EGFR-TKIs sensitive mutations even though the rate among the N0-N3 show no remarkable changing trend. With respect to the result of the IHC, the TTF-1 and EGFR protein levels showed a close relationship with EGFR-TKIs sensitive mutations. In addition, TTF-1 and EGFR protein positive-expression was associated with an increased frequency of mutations.
Supporting Information S1 File. The original data of this study. (XLSX)