Detection of EGFR and BRAF mutations by competitive allele‑specific TaqMan polymerase chain reaction in lung adenocarcinoma

Abstract. Epithelial growth factor receptor (EGFR)-tyrosine kinase inhibitors are the standard first‑line treatment for patients with metastatic non‑small cell lung cancer (NSCLC) expressing sensitive EGFR‑mutants. Other drugs target different driver mutants, including the serine/threonine-protein kinase B-raf (BRAF) inhibitor dabrafenib, which has exhibited promising efficacy for treating patients with metastatic BRAF‑mutated NSCLC. Therefore, identifying patients carrying mutations that may be treated using targeted therapies is important. However, the methods of molecular detection presently applied in clinical practice, particularly detection of BRAF in NSCLC patients, require further investigation. Therefore, more sensi- tive and economic methods are required. The present study applied the competitive allele‑specific TaqMan polymerase chain reaction (CastPCR) technology to the molecular detec- tion of EGFR (del2235‑2249, del2236‑2250, T790M, L858R) and BRAF (V600E, G469A, D594G) mutations in 144 treat- ment‑naive patients with lung adenocarcinoma, and analyzed the association between the mutation rates and patients’ clini- copathological features. 51.4% (74/144) cases were identified harboring EGFR mutations. A total of 40.3% (58/144) patients carried sensitizing mutations (exon 19 deletion or L858R) and 14.6% (21/144) carried T790M mutations. 6.9% (10/144) mutation‑positive patients were double‑mutated. Total EGFR mutation rate was significantly increased in female compared with that of males (60.9 vs. 43.8%, P<0.05), in non‑smokers compared with that of smokers (62.8 vs. 34.5%, P<0.05). In total, 8.3% (12/144) patients were identified with BRAF mutations. 16.7% were V600E (2/12) and 83.3% (10/12) were non‑V600E mutants. Among the 10 non‑V600E mutations, D594G accounted for 90.0% (9/10) and G469A accounted for 10.0% (1/10). Statistical analysis demonstrated that the BRAF mutation rate was not associated with any of the following clinicopathological features: Sex, age, smoking history, clin- ical stages, distant metastasis, differentiation degree, tumor size and regional lymph node metastasis (P≥0.05). CastPCR technology is a robust method with high sensitivity for the molecular detection of EGFR and BRAF mutations in clinical formalin‑fixed paraffin‑embedded samples.

Introduction
Lung cancer is one of the most common types of cancer and the leading cause of cancer‑associated mortality worldwide. According to data from the World Health Organization GLOBOCAN study (1), the estimated number of new cases was1.8 million in 2012 (12.9% of the total diagnosed cancer cases), and lung cancer was estimated to be responsible for nearly one in five (1.59 million deaths, 19.4% of the total) incidences of mortality from cancer. The traditional platinum‑based chemo- therapeutic modalities have, for the past few decades, reached a therapeutic plateau (2); fortunately, the identification of driver mutations, including those to the epithelial growth factor receptor (EGFR) and anaplastic lymphoma kinase, and the use of targeted therapy have improved progression‑free survival, overall response rate and quality of life of patients (3-9).At present, driver mutations are continuing to be identified; there are a number of clinical trials regarding targeted thera- pies in progress (10,11). Serine/threonine‑protein kinase B‑raf (BRAF) is a member of the RAF kinase family, which serves an important function in the MAPK signaling pathway (12,13). Once abnormally activated, the kinases transmit extracellular signals that result in the promotion of cell proliferation, survival and invasion (14,15). In non‑small cell lung cancer (NSCLC), BRAF inhibitors are showing promise according to the results of a phase II clinical trial (no. NCT01336634) assessing theclinical activity of dabrafenib in metastatic NSCLC patients (of which 96% were adenocarcinomas) with BRAF V600E mutations (16). However, the incidence of BRAF mutations in NSCLC has not been conclusively determined. The reported incidence differs greatly and ranges between 0.5 and 9% (17). BRAF mutations are observed almost exclusively in adenocar- cinomas (18); however, other details remain to be elucidated.To identify as many mutation‑positive patients who may benefit from targeted therapies as possible, a rapid and robust molecular detection assay is required.

In clinical practice in China, the detection of genetic mutations is usually performed by Sanger sequencing and the amplification refractory muta- tion system (ARMS) (19). Sanger sequencing is considered the ‘gold standard’ genotyping technique, owing to its low false‑positive rate and high specificity (20). However, it has a number of disadvantages, including low sensitivity, the time taken to perform the assay, the requirement of high‑quality tissue samples and the necessity of manual interpretation (21). Although the sensitivity (1%) (22) of ARMS is increased compared with that of Sanger sequencing (10‑25%) (20,23-26), and certain other disadvantages of Sanger sequencing may be overcome by ARMS, it remains inferior to certain novel testing methods (27). Recently, gene detection using next generation sequencing (NGS) technology has attracted extensive atten- tion; however, these methods are time‑consuming and too expensive for the detection of a specific gene allele. Besides, the data produced by NGS is not easy for routine clinical analysis. Competitive Allele‑Specific TaqMan® polymerase chain reaction (CastPCR) has a high sensibility and specificity; its allele‑specific primer and locus‑specific primer guarantee the amplification of the mutant allele, while an oligonucleotide blocker suppresses amplification of the wild type allele (28). Bao et al (29) demonstrated that CastPCR technology may robustly detect mutated alleles in a wild type background as low as 0.1% and has >99% concordance with other technologies, including PCR‑based technology and sequencing. Although small sample sizes of~30 were used, Didelot et al (30) and Roma et al (22) concluded that CastPCR is highly sensitive for the specific detection of EGFR mutations in NSCLC clinical samples. Li et al (31) demonstrated that CastPCR technology is a valuable validation tool for NGS detection of multiple gene mutations, including those to EGFR.To the best of our knowledge, the present study is the first to evaluate the validity of CastPCR in EGFR and BRAF mutation detection in >100 formalin‑fixed paraffin‑embedded (FFPE) samples of lung adenocarcinoma.

The current study also investigated the association between EGFR/BRAF mutation incidence and the clinicopathological features of patients.Materials. A total of 144 FFPE samples of lung adenocar- cinoma patients diagnosed between November 2010 and November 2015 were collected from Nanjing Drum‑Tower Hospital (Nanjing, China). The sex ratio of the 144 patients enrolled was 1.25 (male:female) and the age range was between 37 and 75 years (mean, 60.8 years). Clinicopathological features are provided in Table I. Each FFPE sample was cut into 4‑µm‑thick slices. For every FFPE sample, a randomslice underwent hematoxylin and eosin staining for 60 min at room temperature (CoverStainer; Agilent Technologies, Inc., Santa Clara, CA, USA). A senior pathologist from the Pathology Department of Drum Tower Hospital identified the area of tumor tissue of the stained sample using a light microscope (magnification, x40), and a sample of tumor tissue (0.6‑1.0 mm2) was removed and placed into an Eppendorf tube for later DNA extraction. The clinicopathological features, including sex, age, smoking history, distant metastasis, clinical stages of patients (according to the 7th edition of tumor‑node‑metastasis staging for lung tumors outlined by the American Joint Committee on Cancer) (32), differentiation of tumor, and, for patients that underwent surgery, information of tumor size and regional lymph node metastasis were collected. Ethical approval for the present study was provided by The Medical Ethics Committee of Drum Tower Hospital. All patients provided written informed consent for the publication of the present study.DNA extraction. DNA was extracted using the TIANamp FFPE DNA kit (Tiangen Biotech Co., Ltd., Beijing, China) according to manufacturer’s instructions from the 0.6‑1.0 mm2 tumor tissue on the tissue section.CastPCR. Extracted DNA samples from FFPE tissues were analyzed by the following Taqman® Mutation Detection Assay kits(Life Technologies; Thermo Fisher Scientific, Inc., Waltham, MA, USA): EGFR_reference (Catalog Number: 4465807, Assay ID: Hs00000173_rf) and BRAF_reference (Catalog Number: 4465807, Assay ID: Hs00000172_rf); EGFR_6223_ mu (Catalog Number: 4465804, Assay ID: Hs00000156_mu), EGFR_6225_mu (Catalog Number: 4465804, Assay ID: Hs 00000157_mu), EGF R_ 6224 _mu (Catalog Number: 4465804, Assay ID: Hs00000102 _mu) and EGFR_6240_mu (Catalog Number: 4465804, Assay ID: Hs00000106_mu) for detecting the 2235‑2249 del, 2236‑2250 del, L858R and T790M, respectively; BRAF_460_mu (Catalog Number: 4465804, Assay ID: Hs00001351_mu), BRAF_467_ mu (Catalog Number: 4465804, Assay ID: Hs00000996_mu) and BRAF_475_mu (Catalog Number: 4465804, Assay ID: Hs00001384_mu) for detecting G469A, D594G and V600E, respectively. The PCR reaction was performed using a Stratagene MX3000P real‑time PCR system (Stratagene; Agilent Technologies, Inc.).

The PCR conditions were as follows: 95˚C for 10 min, followed by 5 cycles at 92˚C for 15 sec and 58˚C for 1 min, then 40 cycles at 92˚C for 15 sec and 60˚C for 1 min. Prior to the detection of clinical specimens, the limit of detection (LOD) of CastPCR for detecting EGFR and BRAF mutations was assessed and a dilution of mutant alleles in a wild‑type background was prepared using DNA extracted from the corresponding cell lines. The wild‑type background was set as DNA extracted from EGFR and BRAF mutation‑nega- tive cell line A549. Cell line PC9 was mutation‑positive for deletion (2235‑2249) in exon 19 of EGFR, cell line H1975 was mutation‑positive for T790M in exon 20 and L858R in exon 21 of EGFR and cell line A375 was positive for the V600E muta- tion in exon 15 of BRAF. All cell lines were obtained from our own laboratory, and were cultured in RPMI‑1640 (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetalbovine serum (Minhai Bio Engineering, Lanzhou, China). The cells were incubated at 37˚C in a humidified atmosphere of 5% CO2. In addition, the cell lines were used within ten passages after thawing of frozen cells. All the mutations of the cell lines were confirmed using pyrosequencing by Shanghai HuaGene Biotech Company (Shanghai, China). DNA extracted from cell lines (as aforementioned) was initially diluted at 20 ng/µl and then serial dilutions of 1:100, 1:500, 1:1,000 and 1:2,000 were performed. The prepared DNA was put into the CastPCR reaction wells. CastPCR was run on a 96‑well plate, and each well had a final volume of 10 µl. This included 5 µl Taqman® Genotyping Master Mix (2X) (Life Technologies; Thermo Fisher Scientific, Inc.), 2 µl prepared gDNA sample, 2 µl nuclease‑free water (Beyotime Institute of Biotechnology, Haimen, China) and 1 µl TaqMan® Mutation Detection Assay (10X) (Life Technologies; Thermo Fisher Scientific, Inc.) according to the manufacturer’s protocol.

The detection of each DNA sample was repeated three times. The concentration of extracted DNA was determined by a NanoDrop spectro- photometer (NanoDrop Technologies, Inc., Thermo FisherScientific, Inc.), and adjusted to a concentration of 10 ng/µl for further analysis. For G469A and D594G, no available mutation positive cell lines could be obtained; therefore, the assessment was not performed. However, according to the manufacturer, the sensitivity of detecting the two mutations was 0.1%.DNA is qualified for detection when the Cq value (gene reference assay) is between 17‑33, or re‑extraction would be performed for the sample. The mutation status was determined by calculating the ΔCq value between amplification reactions for a mutant allele assay and a corresponding gene r1eference assay, namely ΔCq=Cq (mutant allele assay)‑Cq (gene refer- ence assay) as outlined by the TaqMan® Mutation Detection Assay handbook provided by CastPCR (33). The cut‑off ΔCq values provided by Life Technologies were used. The mutation was detected if the ΔCq≤ΔCq cut‑off and mutation was not detected if the ΔCq>ΔCq cut‑off. According to the cut‑off ΔCq values provided by Life Technologies, the cut‑off ΔCq values of EGFR 2235‑2249 del, 2236‑2250 del, L858R and BRAF V600E, G469A, D594G are 9.96; That of EGFR T790M is 9.61.Statistical analysis. Data were presented as mean ± standard deviation. Statistical analyses were performed using SPSS software (version 18; SPSS, Inc., Chicago, IL, USA). Variables measured in the study were assessed for association using the χ2 test or Fisher’s exact test, as appropriate. P<0.05 was considered to indicate a statistically significant difference. All P‑values were two‑sided.

Results
L858R and 1:100 T790M detection, respectively. The ΔCq value was 9.90±0.04 for 1:1,000 V600E detection. Therefore, in accordance with the sensitivity value provided by the manu- facturer, the LOD for EGFR exon 19 deletion and L858R were 0.1%, that of T790M was 1% and that of BRAF V600E was 0.1%.EGFR mutations. In total, 51.4% (74/144) patient samples were identified as harboring EGFR mutations: 40.3% (58/144) patients carried EGFR sensitizing mutations (exon 19 deletion or/and L858R) and 14.6% (21/144) carried EGFR T790M mutations. In addition, 6.9% (10/144) mutation‑positive patients were EGFR double‑mutated: 3 patients had exon 19 deletions and T790M, 5 had exon 19 deletion and L858R, and 2 had T790M and L858R.Statistical analysis demonstrated that the EGFR mutation rate was significantly increased in female patients compared with that in males (60.9 vs. 43.8%, P<0.05), in non‑smokers compared with that in smokers (62.8 vs. 34.5%, P<0.05), as expected (Table II). Age, clinical stage (stratified as stage I‑III and stage IV), distant metastasis, degree of differ- entiation, tumor size and metastasis of regional lymph nodes (the last two features applying to surgical patients only) were not associated with EGFR mutation status (P≥0.05; Table II).BRAF mutations. In total, 8.3% (12/144) of patients were identified as possessing BRAF mutations: 16.7% (2/12) were V600E mutations and 83.3% (10/12) were non‑V600E muta- tions. Among the 10 non‑V600E mutations, D594G accounted for 90.0% (9/10) of mutations and G469A accounted for 10.0% (1/10) (Table III).Statistical analysis demonstrated that the BRAF mutation rate was not associated with any of the clinicopathological features assessed in the present study, including sex, age, smoking history (including the stratified classification of smokers, former smokers and non‑smokers; data not shown), distant metastasis, degree of differentiation, tumor size and metastasis of regional lymph nodes (the last two features applying to surgical patients only; P≥0.05; Table IV).

Discussion
Identifying EGFR as an oncogene and the successful appli- cation of EGFR‑tyrosine kinase inhibitors (TKIs) have revolutionized the treatment of NSCLC. Personalized medicine that targets driver mutations using precision medicine is key to eliciting an improved therapeutic effect. Molecular detection of driver mutations within the tumor tissue of patients may enable precision treatment of lung cancer.CastPCR technology is highly specific and sensitive and may detect small quantities of mutated DNA in a sample that contains large amounts of normal, wild‑type genomic DNA. Tan et al (34,35) demonstrated that CastPCR exhibits TaqMan® assay‑like sensitivity, linearity and dynamic range and may detect a single mutant molecule in the presence of 1 million wild‑type molecules. CastPCR may be performed to detect certain mutant alleles and is cost‑effective (30,34-37). Compared with ARMS, the most commonly used method in clinical practice in China, CastPCR has improved sensi- tivity and specificity owing to its oligonucleotide blocker that suppresses the wild‑type allele, which does not exist in ARMS (28). Previous studies have demonstrated this inmultiple types of malignant tumor (22,38,39). Two previous studies used CastPCR technology to validate detected mutation results obtained by NGS (31,40). For economic consideration in clinical practice, the present study evaluated the molecular detection of EGFR and BRAF mutations using CastPCR in 144 lung adenocarcinoma samples.Although previous studies have performed EGFR muta- tion detection in NSCLC using CastPCR, the number of patients was relatively small, with no more than 30 samples included (22,30,41). Therefore, the present study collected samples from 144 patients with lung adenocarcinoma in order to verify its sensitivity and feasibility in larger clinical samples. The frequency of EGFR mutations in lung adenocar- cinoma was 51.4% (74/144) in the present study, which was concordant with that in previous studies conducted in Asia (3,42,43). In the present study, 40.3% (58/144) of patients exhibited EGFR mutations, values close to those obtained by Shi et al and Wu et al (43,44).

In the present study, EGFR mutations appeared more frequently in women than in men(60.9 vs. 43.8%, P=0.04), more frequently in non‑smokers than in smokers (62.8 vs. 34.5%, P=0.001), as expected (43,45-47).To identify the LOD of CastPCR for detecting EGFR mutations, EGFR‑mutation‑positive and ‑negative cells were mixed. The LOD for exon 19 deletion and L858R was 0.1% and that of T790M was 1%, which matched the LOD provided by the manufacturer. Considering the LOD of CastPCR, EGFR mutation detection in T790M (1%) may not be satis- fying compared with that of exon 19 deletion and L858R (0.1%). In EGFR‑mutated NSCLC, the reported incidence of de novo T790M substitution varied, ranging between 0 and 78.9% worldwide (48). This range of figures was primarily associated with different detection methods (Table V). The LOD of different methods for detecting T790M mutations is provided (Table VI).In the present study, 6.9% (10/144) of patients were found to possess double mutations. EGFR double mutations have been reported in previous studies (49-55). For example, Zhang et al (54) found that 6.5% (5/77) of lung adenocarci- noma patients possessed double mutations. Masago et al (51) found that this figure was 4.4% (4/90). In EGFR‑mutated lung adenocarcinoma, the double mutation rate was 13.5% (10/74) in the present study. Keam et al (50) found a double muta- tion incidence 1.5% (4/274) in Japanese EGFR‑mutated lung adenocarcinoma patients. The incidence of double mutation in the present study may have been increased due to the increased sensitivity of the CastPCR technology. Among double EGFR‑mutated patients in the present study, 3 were exon 19 deletions and T790M, 5 were exon 19 deletions and L858R and 2 were T790M and L858R. Exon 19 deletions and L858R are considered sensitizing mutations since the majority of patients harboring this type of mutation respond to EGFR‑TKIs (55). The clinical significance of exon 19 deletions/L858R and T790M is, to the best of our knowledge, unknown.

Blakely et al (56) found that EGFR‑TKI treatment may lead to the expansion of BRAF V600E‑expressing tumor cells, resulting in acquired EGFR‑TKI resistance that may be reversed by treatment with a BRAF inhibitor, which implies that the BRAF V600E mutation could be one of the mechanisms of EGFR‑TKI resistance. Recently, a phase II clinical trial (no. NCT01336634) has made its results public (16). Dabrafenib exhibited clinical activity in BRAF V600E‑positive meta- static NSCLC; these results indicated that dabrafenib could represent a treatment option for a population of patients with limited therapeutic options (16). Certain studies (57-62) have evaluated BRAF mutations in NSCLC samples; however, the use of CastPCR in assessing these mutations has not, to the best of our knowledge, been reported. Therefore, the present study performed BRAF mutation detection using CastPCR in 144 lung adenocarcinoma samples.The BRAF gene is mutated in 1‑5% of NSCLC cases, and the majority of these are mutated in adenocarcinomas (17,18,61,62). When NGS was used, the incidence of BRAF mutation appeared to be 6.5% (33/510) (63). In the present study, this inci- dence was 8.3%, closer to the result obtained by NGS (17,63). The incidence of BRAF mutation in the present study may have been increased since BRAF mutations are more common in Asian populations (59,60). Compared with that of Sanger sequencing (64), CastPCR exhibited superior sensitivity in detecting BRAF mutations in clinical FFPE samples in ourstudy, which is comparable with that of NGS. Previous studies demonstrated that V600E mutation accounted for at least 50% of BRAF mutations in NSCLC (58,59,65). In recent years, a number of studies applying the NGS demonstrated thatnon‑V600E mutations represent the majority of BRAF muta- tions: Kinno et al (61), Zheng et al (63) and Carter et al (17) revealed that the proportion of non‑V600E in BRAF‑mutated NSCLC was 70.0, 78.8 and 86.0%, respectively. The presentstudy revealed an incidence of non‑V600E BRAF muta- tions of 83.3% (10/12), similar to that demonstrated by NGS studies (17,61,63).

Of the 10 non‑V600E‑mutated patients in the present study, 75.0% (9/12) carried D594G and 8.3% (1/12) carried G469A, so the mutation rates were may be represented as D594G>V600E>G469A, which is concordant with the results obtained by Carter et al (17). Therefore, the results of the present study indicated that CastPCR is a valuable tool for the detection of BRAF mutations in clinical FFPE samples.The present study selected certain clinicopathological features, including sex, age, smoking history, clinical stages, distant metastasis, differentiation degree, tumor size and lymph node metastasis, to identify the association between BRAF mutations and these features (Table IV). None of these clinicopathological features was associated with BRAF muta- tions. The association between BRAF mutation status and sex, age and smoking history was consistent with previous research (18,66). In addition, the incidence among former, current and never smokers was assessed and no statistical significance was found (data not shown). At present, few studies focus on the pathological features of BRAF‑mutated patients. According to Marchetti et al (65), no association was found between BRAF mutation rate and tumor size or lymph node metastasis. No association was identified between BRAF mutation and differentiation degree. EGFR mutations tend to appear in individuals that are Asian, female, non‑smokers and exhibit adenocarcinoma, therefore this may aid physicians in recognizing patients who may benefit from EGFR‑TKIs. Therefore it is recommended that patients should be tested for EGFR mutations in order to personalize the treatment regime. Since BRAF inhibitors have already demonstrated promise in patients with advanced stage non‑small‑cell lung cancer and who are BRAF-positive (16), it is necessary to obtain the information concerning the type of population that typically exhibits BRAF mutations. However, no consensus exists for this and further studies are required. BRAF‑mutated patients are provided (Table III). No signif- icant difference was found between V600E and non‑V600E mutation rates according to sex, smoking history and distant metastasis (P>0.05; data not shown).

A total of 7/12 patients exhibited concurrent EGFR mutations. Prior studies revealed that BRAF and EGFR mutations are exclusive in NSCLC (66-68) This phenomenon seems plausible since BRAF mutation was demonstrated as one of the mechanisms of resistance to EGFR‑TKIs (69). However, concurrent BRAF and EGFR mutations were observed in further studies (61,70). Among the 7 BRAF+EGFR mutation‑positive patients in the present study, all of their BRAF mutations were D594G; the BRAF mutations in the BRAF+EGFR mutation‑positive patients in other studies were also non-V600E (61,70). Of the aforementioned BRAF muta- tions, V600E is the only kinase‑activating one and could result in the activation of downstream mitogen‑activated protein kinase kinase‑extracellular signal‑regulated kinase signaling pathway resistance to EGFR‑TKIs (71). However, to the best of our knowledge, no evidence has demonstrated that non‑V600E mutations of BRAF are resistant to EGFR‑TKIs. Therefore, just as the EGFR mutations are divided into ‘sensitizing’ and ‘resistant’ categories to more accurately select treatment regimens for patients, BRAF‑mutated patients should also be distinguished for V600E and non‑V600E status to aid the design of more pertinent experiments and clinical trials. CastPCR is a robust method with high sensitivity for molecular detection of EGFR and BRAF mutations in FFPE samples. Therefore, EGFR Brimarafenib and BRAF mutation detection could represent a feasible strategy in clinical practice.