Epidermal Growth Factor Receptor Exon 20 Mutation Increased in Post-Chemotherapy Patients with Non-Small Cell Lung Cancer Detected with Patients' Blood Samples

Abstract

PURPOSE: Patients with non-small cell lung cancer (NSCLC) and epidermal growth factor receptor (EGFR)-mutations have excellent response to EGFR tyrosine kinase inhibitors (TKIs), and exon 20 mutation accounts for most of TKI drug resistance. Nested polymerase chain reaction (PCR) was used to detect EGFR exon 20 mutations of patients with NSCLC after chemotherapy. The same is being analyzed with patients' characteristics. METHODS: Peripheral blood samples were collected from 273 patients with NSCLC, including 143 with adenocarcinoma (ADC) and 130 with squamous cell carcinoma (SCC), after chemotherapy. DNA was extracted from whole blood for nested PCR amplification and purification. Sequencing was carried out in an automated 3730 sequencer, followed by analysis of EGFR exon 20 mutations from nested PCR products. RESULTS: The mutations of EGFR exon 20 were mainly point mutations in rs1050171 (c.2361A>G) and rs56183713 (c.2457G>A). The point mutation was 28.21%, 28.46%, and 27.97% in patients with NSCLC, ADC and SCC, respectively. Men had an equivalent mutation (27.18%) to women (30.77%). The mutation in smokers and nonsmokers was 27.68% and 29.17%, respectively. In unselected patients, there was no correlation between EGFR exon 20 mutations and patients' characteristics of age, gender, smoking history, histologic type, or tumor-node-metastasis (TNM) staging system. In subgroup analyses, the EGFR mutation of patients with SCC was correlated with TNM stage [P = .013; odds ratio = 1.758; 95% confidence interval (CI) = 1.125–2.747]. CONCLUSIONS: The data indicate that the chemotherapy may induce EGFR-TKI-resistant mutation in NSCLC cells and EGFR-TKI should be used in the early stage of NSCLC but not after chemotherapy.

Introduction

Lung cancer is the leading cause of cancer death in the United States [1] and China [2]. Non-small cell lung cancer (NSCLC) is the most common histologic type, which affects approximately 80% to 85% of all patients with lung cancer. Although much progress has been made in the last 10 years for lung cancer (e.g., screening, minimally invasive techniques for diagnosis and treatment, and targeted therapy) [3], only 15.9% of all lung cancer patients can be alive for 5 years or longer after diagnosis [4]. Because more than 70% of patients with lung cancer are diagnosed with advanced-stage disease, systemic treatment plays main role in clinical management. However, routine platinum-based chemotherapy prolongs median survival time by only a few months in these patients compared with supportive care. Currently, median survival time is about 8 to 18 months for patients with metastatic and locally advanced diseases, respectively [5].

Epidermal growth factor receptor (EGFR) is a transmembrane receptor that is detectable in approximately 80% to 85% of patients with NSCLC, and the level of expression varies widely on a continual scale. The activation of EGFR pathways results in the initiation of cancer proliferation, increased metastasis potential, and neoangiogenesis.

In recent years, selective EGFR tyrosine kinase inhibitor (TKI) has emerged as an alternative treatment option for advanced NSCLC. Both erlotinib and gefitinib have demonstrated clinical efficacy in the second- or third-line treatment of NSCLC, especially among never smokers, females, East Asians, and adenocarcinoma (ADC) cell type [6–9]. However, despite the fact that the majority of patients with EGFR mutations benefited from these drugs, in excess of 20% of patients experienced resistance and all tumors ultimately developed resistance following initial response [10]. This is because EGFR mutations are more common in patients who are Asian, female, nonsmokers, and have ADC [11–15], all mentioned characteristics being linked to the known clinical predictors of gefitinib sensitivity [16–19].

Inhibition of EGFR kinase activities by EGFR-TKI, such as gefitinib and erlotinib, results in effective treatment for patients with NSCLC [6,20,21]. The IRESSA Pan-Asia Study clinical trial in Asia and randomized controlled studies on advanced NSCLC have confirmed EGFR-activating mutations as the main predictor of clinical outcome with TKI therapy [7,22,23]. The most commonly EGFR-activating mutations were in exon 19 (Del19) and exon 21 (L858R) that were found in approximately 10% of Caucasian patients with NSCLC and up to 50% of Asian patients [24].

Other drug-sensitive mutations include point mutations at exon 21 and exon 18 [25]. The pretreatment EGFR T790M mutation was associated with shorter progression-free survival (PFS) compared to EGFR-TKI therapy in patients with NSCLC [26]. The mutation play an important role in predicting PFS in patients with NSCLC and EGFR-activating mutation treated with TKI [27]. The prevalence of reported de novo T790M varied from 1% to 38% [27–30]. Han et al.

[31] found that EGFR mutations were detected in 39.4% (13/33) and 54.5% (18/33) of the serum samples of pre-chemotherapy and post-chemotherapy, respectively. Data show that PFS is improved with use of EGFR-TKI in patients with EGFR mutations when compared with standard chemotherapy, although overall survival is not statistically different [32].

Primary resistance to TKI therapy is associated with mutation at exon 20 [33–35]. However, the role of EGFR exon 20 mutations in lung cancer progression, especially when it is present before TKI treatment, is still being debated and lacks systematic exploration.

In mainland China, “adenocarcinoma” and “nonsmoker” were independent predictors for EGFR mutations in exons 18, 19, and 21 [36], while the correlation between EGFR exon 20 mutations and patients' characteristics remains unclear. The purpose of this study is to detect the relationship between EGFR exon 20 mutations and patients' characteristics in mainland China.

Patient and Methods

Patient Population

To be eligible for the study, patients were required to have pathologically confirmed stage NSCLC [either ADC or squamous cell carcinoma (SCC)], Eastern Cooperative Oncology Group performance status of 0 to 2, and available plasma. Only patients treated at the Zhejiang Cancer Hospital from May 2012 to December 2012 were enrolled. Exclusion criteria included history of previous primary cancer other than lung cancer and independent of other lung-related disease, justifying this to avoid any probable interference from overlapping genes. All subjects provided their informed consent, and the study was approved by the ethics committee of Zhejiang Cancer Hospital.

DNA Preparation and Sequencing

DNA was isolated from whole blood by using the AxyPrep Blood Genomic DNA Miniprep Kit (Axygen Biosciences, Union City, CA). Atwo-step “boost/nest” polymerase chain reaction (PCR) strategy was employed where the boost reaction generated a larger fragment used as a template for the nest reaction. The nest products were bidirectionally sequenced on ABI 3730 XL sequencers (ABI Company, Oyster Bay, NY) with ABI BigDye Terminator v3.1 chemistry. Base calling was performed by the Agent system (Paracel, Alameda, CA). Sequence tracings were visually inspected to confirm accurate variant detection by the base-calling software. Coding variants were confirmed on repeat PCRs to exclude PCR artifacts. GenBank NM_201283 was used as the reference cDNA for nucleotide positions.

Nested PCR Amplification and Purification

Primer sequences for EGFR exon 20-nested PCR and extension were designed by using the Assay Designers Software version 3.0 (Sequenom, San Diego, CA) and were processed following standard protocols for iPLEX chemistry. Primers were synthesized by Sangon Biotech (Shanghai, China). Primer sequences for boost PCR (502 bp) were forward 5′-TGACTCCGACTCCTCCTTTA-3′ and reverse 5′ATCTCCCTTCCCTGATTAC-3′. Primer sequences for nest PCR (365 bp) were forward 5′-GTCCCTGTGCTAGGTCTTTT-3′ and reverse 5′-ATCTCCCTTCCCTGAT-TAC-3′. The first PCR assays were carried out in a 50-µl volume that contained 1 µl of extracted DNA, 1 µl of primers (R1 and F1), and 0.5 µl of Taq DNA polymerase. DNA was amplified for 35 cycles at 95°C for 3 minutes, 94°C for 30 seconds, 55 to 60°C for 35 seconds, 72°C for 40 to 50 seconds, followed by a 5- to 8-minute repaired extension at 72°C. The second PCRs were carried out in a 25-µl volume, including 2 µl of DNA from the first PCR products, 0.3 µl of primers (R1 and F1), and 0.3 µl of Taq DNA polymerase. DNA was amplified at 95°C for 5 minutes, the next 35 cycles at 95°C for 30 seconds, 55°C for 35 seconds, 72°C for 30 seconds, followed by a 10-minute extension at 72°C. Nested PCR products were electrophoresed in 1% agarose gel. Only PCR products with a positive band were purified using PCR Product Purification Kit (SanPrep, SK1141).

Sequencing Analysis

Before sequencing, PCR sequencing reaction for PCR products was used by BigDye Terminator v1.1 Kit (Applied Biosystems, Foster City, CA) and performed in the light of the kit's request. All PCR assays were carried out in a 20-µl volume, including 1 µl of purified PCR products, 8 µl of BigDye (2.5x), and 1 µl of primers (3.2 pmol/µl). PCR sequencing reaction was performed at 96°C for 1 minute, then followed by 25 cycles at 96°C for 10 seconds, 50°C for 5 seconds, and 60°C for 4 minutes. Sequencing was carried out in an ABI 3730 Genetic Analyzer (Applied Biosystems). All sequence variants were confirmed by sequencing the products of independent PCR amplifications.

Statistical Analysis

We used the χ² to assess the relationship between EGFR exon 20 mutations and each of the clinical and pathologic parameters. The relationship between EGFR exon 20 mutations and demographic characteristics was analyzed by logistic regression. The statistical analysis was performed using SPSS v13.0. All P values were two-sided, and P < .05 was considered significant.

Results

Patients' Characteristics

A total of 300 patients met the enrollment criteria and were enrolled in the study. The patients consisted of 217 men and 83 women, all of which accepted one to four cycles of platinum-based combined chemotherapy. There were 155 patients with lung ADC and 145 with SCC. A total of 197 patients were smokers and 103 were nonsmokers. The patients' clinical and disease characteristics are listed in Table 1. DNA was not available for 27 patients (20 males and 7 females), and 273 patients remained in the analysis list.

Table 1.

Patients' Clinical and Disease Characteristics.

Variables	No. of Patients (N = 300)	Percentage of Patients
Age, years
Mean	58.20	-
SD	9.52	-
Sex
Male	217	72.33
Female	83	27.67
Smoking history
Smokers	199	66.33
Nonsmoker	101	33.67
Histologic type
ADC	155	51.67
SCC	145	48.33
Disease stage
I	15	5.00
II	23	7.67
III	100	33.33
IV	162	54.00

Relationship between EGFR Mutations and Patients' Characteristics

The EGFR exon 20 wild-type DNA sequence was GAAGCC-TACGTGATGGC-CAGCGTGGACAACCCCCACG-TGTGCCGCCTGCTGGGCATCTGCCTCACCT-CCACCGTGCAGCTCATCACGCAGCTCATGC-CCTTCGGCTGCCTCCTGGAC-TATGTCCGGGAACACAA-AGACAATATTGGCTCCCAGTACCTGCTCAACTG-GTGTGTGCAGATCGCAAAG (189 bp; Figure 1).

Figure 1.

EGFR exon 20 wild-type DNA sequence.

The mutations of EGFR exon 20 were mainly point mutations in rs1050171 (c.2361A>G) and rs56183713 (c.2457G>A; Figure 2). The results showed that the EGFR mutation was 28.21%, 28.46%, and 27.97% in patients with NSCLC, ADC, and SCC, respectively. Men had an equivalent mutation (27.18%) to women (30.77%). The mutation in smokers and nonsmokers was 27.68% and 29.17%, respectively. Differences between ADC and SCC, men and women, and smokers and nonsmokers did not reach statistical significance (Table 2).

Figure 2.

Mutations of EGFR exon 20 rs1050171 and rs56183713.

Table 2.

Association between EGFR Exon 20 Mutations and Clinicopathologic Parameters.

Item	No. of Patients with Mutation	No. of Patients without Mutation
Age, years
±60	44	119
>60	33	77
P	.58
Sex
Male	53	142
Female	24	54
P	.55
Smoking history
Smoker	49	128
Nonsmoker	28	68
P	.79
Histologic type
ADC	40	103
SCC	37	93
P	.93
Disease stage
I	5	9
P	.52
II	7	9
P	.15
III	22	70
P	.26
IV	43	108
P	.91

Multivariate analysis was performed to analyze correlation between EGFR mutation and patients' variables. The dependent variable was EGFR status (mutant and wild type). The covariates included age (years), gender (male and female), histology type (ADC and SCC), smoking history (smoker and nonsmoker), and tumor-node-metastasis (TNM) stage (I, II, III, and IV). There was no correlation between the five variables and EGFR mutation (Table 3).

Table 3.

Association between EGFR Exon 20 Mutations and Clinical Features in Patients with NSCLC.

Parameter	P	Multivariate Analysis
		OR	95% CI
Age	.77	0.92	0.54–1.58
Sex	.29	1.76	0.61–5.13
Smoking history	.46	0.67	0.24–1.92
Histologic type	.69	1.13	0.61–2.12
Disease stage	.45	1.14	0.81–1.62

For its part, in subgroup analysis, multivariate analysis was used to identify the correlation. In patients with SCC, the EGFR mutation was correlated with TNM stage [P = .013; odds ratio (OR) = 1.758; 95% confidence interval (CI) = 1.125–2.747], but the three variables (age, gender, and smoking history) were not correlated with EGFR mutation (Table 4). In an almost equal way, in patients with ADC the EGFR mutations were not correlated with age, gender, smoking history, or TNM stage (Table 5).

Table 4.

Association between EGFR Exon 20 Mutations and Clinical Features in Patients with SCC.

Parameter	P	Multivariate Analysis
		OR	95% CI
Age	.82	0.92	0.43–1.96
Sex	1	-	-
Smoking history	1	-	-
Disease stage	.01	1.76	1.12–2.75

Table 5.

Association between EGFR Exon 20 Mutations and Clinical Features in Patients with ADC.

Parameter	P	Multivariate Analysis
		OR	95% CI
Age	.73	1.15	0.53–2.47
Sex	.22	0.49	0.16–1.53
Smoking history	.99	0.99	0.32–3.07
Disease stage	.40	0.81	0.49–1.34

Discussion

It has been confirmed that EGFR-activating mutations represents the main predictor of clinical outcome with TKI therapy [7,22,23]. According to Hirsch and Bunn [24], the most commonly EGFR-activating mutations are in exon 19 (Del19) and exon 21 (L858R). Additionally, Riely et al. [25] indicated that other drug-sensitive mutations include point mutations at exon 21 and exon 18. Primary resistance to TKI therapy is associated with mutation at exon 20 [33–35]. However, the role of EGFR exon 20 mutations in lung cancer progression, especially its presence before TKI treatment, is still being debated and lacks systematic exploration.

In this study, we found, in unselected patients with NSCLC treated with platinum-based combined chemotherapy, that the EGFR exon 20 mutation was 28.21%, the same being 28.46% and 27.97% in ADC and SCC, respectively. We failed to find the correlation between EGFR exon 20 mutations and patients' characteristics of age, gender, smoking history, histologic type, or TNM stage. This may be ascribed to the greatly increased EGFR-TKI-resistant mutation post-chemotherapy, thus leading to the reduction of therapeutic effect of EGFR-TKI. Therefore, if EGFR-TKI is used in NSCLC early stage, the benefits that patients get from the drug will be maximized. Recent data also suggest that erlotinib should be used as first-line systemic therapy in patients with EGFR mutations documented before the mentioned therapy [32,37–39]. Strikingly, we also found that the patients with SCC had a higher TNM stage and more mutation (P = .013) in subgroup analysis. These data indicated that the chemotherapy may induce NSCLC cells to restrain EGFR-TKI. Thus, the higher the TNM stage, the greater the resistance. These results suggest that the EGFR-TKI should be used in the early stage of NSCLC but not after chemotherapy.

It is also possible that mutation can be present before treatment and mutation tumor cells may undergo a relatively indolent progression in tumorigenesis [40–42]. De novo mutation might contribute to poor survival besides the drug resistance [27].

Initially, EGFR exon 20 mutations were found in specimens of NSCLC patients with EGFR-TKI treatment failure and currently in untreated specimens. It is likely that primary tumor samples of patients with NSCLC have significant de novo EGFR exon 20 mutation and de novo TKI-resistant cells before TKI therapy. These TKI-resistant clones grow in the selective environment of TKI therapy because of their survival advantage [26]. Similarly, chemotherapy can kill TKI-sensitive NSCLC cells, while TKI-resistant cells can still grow.

It was reported that the specimens to detect EGFR mutation can be derived from tissue, blood, and pleural fluid. Nevertheless, sequence analysis of tumor DNA from tissue is still the most widely applied method for detecting EGFR mutation and is considered the most direct and reliable approach. In this study, we used peripheral blood to detect EGFR exon 20 mutations. In the future, it is necessary to detect EGFR mutations in both tissue and blood.

In conclusion, our data suggested that the platinum-based chemotherapy in this study can induce EGFR-TKI-resistant mutation in NSCLC cells at EGFR exon 20 and that EGFR-TKI should be used in the early stage of NSCLC but not after chemotherapy.