Abstract
Background: In non-small cell lung cancer (NSCLC) the phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) gene mutations have been reported in fewer than 5% of primary tumors. Materials and Methods: We assessed PIK3CA gene mutations in 145 tissue samples from central nervous system (CNS) metastases of NSCLC using three polymerase chain reaction (PCR) techniques: high resolution melting-PCR (HRM-PCR), allele-specific-quantitative PCR (ASP-qPCR) and TaqMan PCR. Results: HRM analysis allowed us to select three PIK3CA-positive specimens (2.1% of the studied group) and ASP-qPCR techniques identified them as one E542K and two H1047R substitutions, which were confirmed by TaqMan probes. The PIK3CA mutations were indicated only in males (3% of all males). One of the patients was reported to be a non-smoker with adenocarcinoma (AC; 2.5% of the AC group), however, the other two patients were smokers with squamous cell carcinoma (SCC; 3.4% of SCC group). Conclusion: This is the first report of the presence of PIK3CA gene mutation in CNS-metastatic lesions of NSCLC worldwide that could broaden therapeutic choices in such patients.
Phosphatidylinositol 3-kinases (the PI3K protein family) are lipid kinases encoded by the phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) gene. PI3Ks play a major role in important cellular functions, including growth, adhesion, survival, motility and cell proliferation. The function of PI3Ks is regeneration of phosphorylated forms of phosphatidylinositol, which is a key mediator between growth factor receptors and intracellular downstream signaling pathways (1-4).
PIK3CA gene mutations are somatic missense substitutions that are generally clustered in exon 9 (E542K, E545K, E545Q), which encodes the PI3K helical domain, and in exon 20 (H1047R, H1047L), which encodes the catalytic domain of PI3K (3-6). It has been shown that mutations of the PIK3CA gene can induce oncogenic cellular transformation, which is correlated with enhancing of PI3K enzymatic activity that leads to constitutive activation of murine thymoma viral oncogene (AKT) signaling pathway in cancer cells. It has been associated with more aggressive diseases, which correlate with a poor prognosis for patients (3-7).
PIK3CA gene mutations are most frequently reported in colorectal cancer, gliobloastoma, and gastric and breast cancer (30%) (1, 2, 5). In non-small-cell lung cancer (NSCLC), PIK3CA gene mutations have been reported in fewer than 5% of primary tumors, especially in adenocarcinoma (AC) and squamous cell carcinoma (SCC) subtypes. Moreover, the coexistence of PIK3CA gene mutations and other oncogene mutations (mainly epidermal growth factor receptor (EGFR) and vital rapidly accelerated fibrosarcoma homolog B1 (BRAF) in lung adenocarcinoma have also been described (6-9).
Reliable and sensitive molecular detection assays are in demand for predicting response or survival benefit during standard or targeted therapy in patients with NSCLC (10). However, only exact reporting of methodology will allow for detectability of mutations by different molecular methods to be evaluated (11). In spite of low sensitivity (especially in poorly cellular biopsies and degraded formalin-fixed paraffin-embedded (FFPE) tissue samples), direct DNA sequencing was considered a gold standard for the detection of driver mutations (11-13). Currently, techniques certified for in vitro diagnosis followed by melting-curve analysis or hybridization probes in real-time polymerase chain reaction (PCR) technique are described as an efficient approach to screening of well-known mutations (14). The high detectability of driver mutations is also provided by next-generation sequencing (NGS). In the future, the NGS method may play a major role both in the qualification of patients for personalized therapy and in the detection of residual disease (15).
To date analysis of E542K, E545K, E545Q,H1047R, and H1047L substitutions for the PIK3CA gene was performed mainly in primary tumors of NSCLC. However, in our study we estimated their incidence in central nervous system (CNS) metastases of NSCLC. Moreover, we compared the results of PIK3CA mutation detection using high-resolution melting PCR (HRM-PCR), allele-specific quantitative real-time PCR (ASP-qPCR) and real-time PCR with TaqMan hydrolysis probes. With these approaches, we aimed to determine the screening process for PIK3CA gene status by determining the concordance rate between the results obtained for the analyzed samples and control DNA with the presence or absence of PIK3CA mutation.
Materials and Methods
Patients and Materials. A total of 145 FFPE tissue samples were obtained from surgically resected CNS metastases of Caucasian patients with advanced NSCLC. Thirty corresponding primary NSCLC tumors were simultaneously available. The patients underwent routine neurosurgical procedures with a palliative intent. The median overall survival (OS) time from cancer diagnosis to death was 13.5 months (range=0.1-78.2 months; information available from 119 patients). All of the studied patients were untreated by chemotherapy, radiotherapy or other molecularly targeted therapies. According to the number of smoked cigarettes, patients were qualified as heavy smokers (≥15 pack-years), light smokers (<15 pack-years) and non-smokers. Detailed characteristics of the studied group are presented in Table I.
The study was approved by the Ethics Committee of the Medical University of Lublin, Poland (no. KE-0254/86/2013).
Mutation analysis. DNA was isolated from FFPE samples of metastatic tissue using QIAamp a DNA FFPE Tissue Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Estimation of PIK3CA gene mutations (substitutions E542K, E545K, H1047R) was conducted using three methods based on a real-time PCR reaction. The HRM-PCR technique was used as a screening method. ASP-qPCR was performed for the identification of PIK3CA mutation type. TaqMan (Applied Biosystems, Foster City, CA, USA) hydrolysis probes, as well as real-time PCR technique, were used to confirm the results obtained with the two previously mentioned methods. DNA isolated from PIK3CA gene mutation-positive cell lines (SW48 cell-line: substitution E542K; MCF10A cell-line: substitution E545K and H1047R) acted as positive control for the analysis. The control DNA was provided by Horizon Discovery (Horizon Discovery, Cambridge, UK) at a concentration of 20 ng/μl. DNA isolated from peripheral blood leukocytes of healthy individuals acted as the negative control. In addition, in our previous studies, we published the incidence of EGFR, human epidermal growth factor receptor 2 (HER2), discoidin death receptor 2 (DDR2) and Kirsten rat sarcoma virus (KRAS) gene mutations in the same analyzed material (16).
Study group characteristics.
HRM-PCR method. PIK3CA gene substitutions were analyzed using real-time PCR method with DNA intercalating dye and two pairs of primers flanking the mutated site of the gene. One pair of primers flanked the mutations located in exon 9 (substitutions E542K and E545K) and a second pair of primers flanked a mutation located in exon 20 (substitution H1047R). Before HRM-PCR, all DNA samples were diluted to a concentration of 20 ng/μl. The total reaction mixture (15 μl) was prepared with 8 μl of HRM Primer Design Master Mix (PrimerDesign Ltd, Southampton, UK) with DNA-intercalating dye (Chromofy™), 1 μl of purified genomic DNA (20 ng/μl), 1 μl of each forward and reverse primer and 4 μl of nuclease-free water. The amplification of the examined region was performed in 48-well plates using an Eco real-time PCR device (Illumina, San Diego, CA, USA) in the following steps: pre-denaturation at 95°C for 10 min and 45 cycles of: 95°C for 15 s and 62°C for 60 s. In the next step, the HRM procedure was performed for PCR products as follows: 92°C for 5 s, 72°C for 20 s, and 92°C for 5 s. Different genotypes of PIK3CA were distinguished according to the raw data plots and difference graph derived from the normalization data. A comparison of amplification and melting curves in the positive and negative controls allowed us to distinguish the mutant and wild-type samples.
ASP-qPCR method. The PIK3CA gene mutations were analyzed using the ASP-qPCR method with DNA-intercalating dye and allele-specific primers for mutant and wild-type PIK3CA gene. Some of the reagents and DNA, as well as total volume of PCR mixture, were the same as in the HRM procedure. The amplification of the examined region was performed in 48-well plates using an Eco real-time PCR device (Illumina) in the following steps: pre-denaturation at 95°C for 10 min and 35 cycles of: 95°C for 15 s and 63°C for 60 s. The mutant and wild-type PIK3CA gene were tested in separate reactions with specific forward primers for these variants of the gene. Samples were assessed as positive if amplification in the ASP-qPCR was observed both for mutant and wild-type PIK3CA gene. Samples with late amplification (Ct>32-cycle) of wild-type PIK3CA gene were excluded from analysis and samples with late amplification (Ct>32-cycle) of mutant region of the PIK3CA gene were assessed as negative.
Characteristics of patients with non-small-cell lung cancer with phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) gene mutation.
Real-time PCR with TaqMan probes. The TaqManFAM™-labeled hydrolysis probes are able to bind only to mutant regions of DNA (i.e. emission of fluorescence is observed only in positive samples). The PCR was performed in a final volume of 15 μl: 10 μl of 2× TaqMan® genotyping master mix (Applied Biosystems), 1 μl of 10× TaqMan hydrolysis probe, 3 μl of nuclease-free water and 1 μl of DNA template (20 ng/μl) in 48-well plates using an Eco real-time PCR device (Illumina). Thermo cycling conditions were recommended by Applied Biosystems: 95°C for 10 min, five cycles of 92°C for 15 s and 58°C for 1 min, then 40 cycles of 92°C for 15 s and 60°C for 60 s. Samples were assessed as mutant for PIK3CA if we observed amplification between 25-30 cycles. The results were compared with amplification plots for positive and negative controls. Samples with (Ct>40-cycle) or without late amplification were assessed as having wild-type PIK3CA gene.
Analysis of results. Results obtained in originally designed HRM- and ASP-qPCR methods were compared to the results produced by TaqMan probes that were dedicated to detection of E542K, E545K and H1047R substitutions in PIK3CA gene. In all three analyses, the threshold was established according to amplification plots for positive and negative controls.
Coexistence of different driver mutations with PIK3CA mutations has also been described. We published the incidence of EGFR, HER2, DDR and KRAS gene mutations in NSCLC metastases to the CNS in our previous studies (16).
Results
In the first part of the analysis, the HRM-PCR technique was used for screening of mutations in exons 9 and 20 of the PIK3CA gene. Flanking HRM primers allowed amplification of selected PIK3CA gene regions in all analyzed samples and controls (mean Ct=25 cycles). Based on the differences in melting curves of positive and negative controls, we were able to distinguish mutant and wild-type PIK3CA gene (Figure 1). We selected three samples with probable PIK3CA gene mutations (one in exon 9 and two in exon 20). In the second part of the analysis, we performed ASP-qPCR on all examined samples to define the type of screened mutations and to compare the sensitivity of the methods (Figure 2). Using the ASP-qPCR technique we were able to detect mutations in positive controls and in all three samples previously selected by HRM-PCR. We found one E542K and two H1047R substitutions of the PIK3CA gene. Amplification of mutant PIK3CA gene was observed on average in the 25th cycle. In these cases, the amplification of wild-type PIK3CA gene was observed on average in the 20th cycle. The presence of amplification of the mutant PIK3CA gene in the negative control was not shown. In the third part of the analysis, we used TaqMan assays (dedicated to the distinction of E542K, E545K and H1047R substitutions in the PIK3CA gene) as a reference technique (Figure 3). This analysis confirmed our results obtained both in HRM-PCR and ASP-qPCR. We observed amplification curves (mean Ct=35th cycle) for three samples positive by HRM-PCR and ASP-qPCR and in the positive control. In wild-type samples, no amplification was observed.
We detected one E542K and two H1047R substitutions in the PIK3CA gene in three CNS metastases of male patients with NSCLC. Incidence of PIK3CA gene mutations was 2.1%. In all cases, the PIK3CA gene was mutated exclusively (without EGFR, HER2, DDR and KRAS gene mutations). Two patients with PIK3CA gene mutations had lung adenocarcinoma (2.5% of patients with AC) and one patient squamous cell lung cancer (3.4% of patients with SCC). One patients with AC was described as a never smoker (77 years old); however, the other PIK3CA gene mutation-positive patients had reported a smoking history (20 pack-year 54-year-old patient with AC; and 174 pack-year 67-year-old patient with SCC). The median OS for patients without PIK3CA gene mutation with NSCLC treated with resection of CNS metastases was 13.5 months. A never-smoking patient with AC with PIK3CA mutation died in the postoperative period. However, OS for a mutation-positive patient with SCC was 5.8 months and in a second mutation-positive patient with AC was 18.7 months (Table II). Analysis of corresponding primary tumors indicated only wild-type PIK3CA gene. Corresponding primary tumors were not available for PIK3CA mutation-positive patients with CNS metastases.
Melting curves of phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) gene amplification product in high-resolution melting-polymerase chain reaction analysis. Curves represent: 1: Amplification of wild-type PIK3CA region of negative control; 2: amplification of mutant PIK3CA region in a central nervous system metastasis from non-small-cell lung cancer; 3: amplification of mutant PIK3CA region in positive control DNA.
Discussion
We detected three PIK3CA gene mutations in CNS metastases of NSCLC (3/145; 2.1%). This result is in accordance with results of the majority of studies reporting a frequency of PIK3CA gene mutations of lower than 5% in primary NSCLC tumors (1, 14, 17, 18). However, unlike in previous studies, PIK3CA mutations were detected here in heavily smoking patients and in CNS metastases of NSCLC. Moreover, in our studied group, PIK3CA mutations were mutually exclusive. In the primary tumor, coexistence of PIK3CA and EGFR mutations was previously described. An et al. reported 20 cases with PIK3CA mutation in 452 patients with NSCLC (4.4%). Mutation presence was insignificantly related with histological type of NSCLC and smoking history (mutations were more frequently detected in adenocarcinoma and in non-smokers). Moreover, they observed overlapping of PIK3CA mutations and cellular-mesenchymal to epithelial transition factor (c-MET) amplification in smokers, both with SCC and AC subtypes (17). Kawano et al. indicated eight cases with PIK3CA gene mutation (8/235; 3.4%) in NSCLC samples. The PIK3CA mutation incidence was lower in those with AC (2/135; 1.5%) than in SCC (5/77; 6.5%). Moreover, in three cases PIK3CA mutations co-existed with EGFR gene mutations. Kawano et al. also mentioned that the most frequent mutations were E545K (26.7%), H1047R (12.9%) and E542K (8.9%) substitutions, whereas the incidence of E545Q and H1047L substitutions was estimated to be in the range of 0.5-2.5% (14). Schildgen et al. detected two cases with PIK3CA mutation (4.9%) in 41 EGFR-negative NSCLC samples (18). However, there are also data of an extremely low (<1%) or high (7%) frequency of PIK3CA gene mutations in patients with NSCLC (7, 19). Kris et al. identified only six cases with PIK3CA gene mutations in 733 NSCLC primary specimens (0.82%) (7). On the other hand, Bar et al. described an exceedingly high incidence of PIK3CA gene mutations (7/100; 7%). Moreover, Bar et al. suggested that the variability in PIK3CA gene mutations may be associated with genetic predisposition to lung cancer (19).
The pre-clinical data have shown that the activation of PIK3CA mutations in EGFR-mutated lung cancer cell lines provides acquired resistance to EGFR tyrosine kinase inhibitors (TKIs). Therefore, treatment based on combination of EGFR TKIs with orally bioavailable inhibitors of class I PI3K isoforms such as NPV-BKM120 or inhibitor of mammalian target of rapamycin (mTOR) (everolimus) have shown promising activity in such cases (3, 4, 8, 9, 20, 21). Moreover, the combination of dual PI3K/mTOR inhibitors (PI-103 or NVP-BEZ235) with selumetinib (selective mitogen-activated protein kinase 1/2 (MEK1/2) inhibitor) can enhance anti-angiogenic and antitumor effects in gefitinib-resistant NSCLC cell line models (20-23). The combination of pictilisib, a novel, oral, highly specific, small-molecule, ATP-competitive class I PI3K inhibitor (GDC-0941) with standard chemotherapy (paclitaxel or carboplatin) revealed better antitumor activity against lung cancer cells than did PI3K inhibition alone (3, 20-24). It was also mentioned that the dual PI3K/mTOR inhibitor dactolisib (NVP-BEZ235) has also shown efficacy against both PIK3CA- and KRAS-mutated NSCLC tumors (25). Therefore, the development of effective methods for PIK3CA mutation testing may be important for the future qualification of molecularly targeted therapies in patients with NSCLC.
Amplification curves for the phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) gene in allele-specific quantitative polymerase chain reaction. Curves represent: 1: Amplification of wild-type PIK3CA region in a negative control; 2: amplification of mutant PIK3CA region in a central nervous system metastasis from non-small-cell lung cancer; 3: amplification of mutant PIK3CA region in positive control DNA; 4: lack of amplification of mutant PIK3CA region in negative control DNA (pair of primers specific to the mutant PIK3CA region).
Amplification curves for the phosphatidylinositol-4,5-bisphosphate 3-kinase, catalytic subunit alpha (PIK3CA) gene in real-time polymerase chain reaction with TaqMan probes. Curves represent: 1: Amplification of mutant PIK3CA region in positive control DNA; 2: amplification of mutant PIK3CA region in a central nervous system metastasis from non-small-cell lung cancer.
In the subsequent analysis, we used originally designed techniques (HRM-PCR and ASP-qPCR methods) that allowed us to detect PIK3CA gene mutations in CNS metastases of NSCLC and in DNA from mutation-positive cell lines. ASP-qPCR was more useful in the identification of PIK3CA mutation types. Moreover, a real-time PCR technique with TaqMan hydrolysis probes confirmed HRM-PCR and ASP-qPCR results. Lindeman et al. suggested that all assays with detectability of mutation in specimens with 10% cancer cells can be used in the screening of routine driver mutations (26). Previously, we described that our ASP-qPCR method is able to detect mutations in material with >1% content of mutant DNA (27). However, the melting-curve analysis may be useful as a prescreening method allowing the selection of specimens for sequence analysis and genotype confirmation (10, 28, 29). Heideman et al. (10) and Kramer et al. (29) described sequencing of HRM-PCR products as a sensitive technique in the identification of mutation types in FFPE tissue samples. However, in the future, the NGS method may play a major role in the screening of mutations (15, 30). Tuononen et al. (30) reported a high (96.3-100%) concordance of the results between targeted NGS and certified real-time PCR tests (for EGFR, KRAS and BRAF screening). Surprisingly, the NGS method showed lower sensitivity than real-time PCR in which four additional mutations were observed. This may be due to FFPE sample purity, DNA degradation and low content of mutant DNA (<20%). A low total number of reads and hybridization-based enrichment NGS method can also favor the wild-type allele. However, NGS allows new and unknown variants of driver mutations to be identified. The implementation of the NGS technique into routine diagnostics will be possible when data for the impact of novel gene variants on cancer pathogenesis, treatment possibilities and prognosis become available (15, 30).
Based on the overall data, we would like to conclude that the designed PCR techniques were sensitive enough to detect PIK3CA gene mutations in the CNS metastases of NSCLC. The HRM-PCR technique can be useful in the selection of positive samples, however, ASP-qPCR or real-time PCR with TaqMan probes allow identification of the types of mutation. The incidence of PIK3CA gene mutations in CNS metastases of NSCLC does not exceed 5% and mutation frequency is probably similar to that in primary tumors.
Footnotes
Conflicts of Interest
The Authors declare no conflicts of interest.
- Received February 26, 2016.
- Revision received April 7, 2016.
- Accepted April 11, 2016.
- Copyright© 2016 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.