Epidermal growth factor receptor (EGFR), a transmembrane receptor tyrosine kinase, is frequently overexpressed or dysregulated in a variety of malignancies.1 EGFR tyrosine kinase inhibitor (TKI) was developed as an anticancer therapy by blocking the overactive EGFR signaling in cancer cells. Although the clinical efficacy was not impressive in the early clinical trials, some “super-responders” were observed in lung adenocarcinomas. Because of two landmark studies,2,3 it is clear that specific somatic mutations at the EGFR tyrosine kinase domain are the key determinants for the effectiveness of EGFR-TKI. Today, it is well accepted that EGFR mutation is the driving force of some lung adenocarcinomas, and EGFR-TKI can very effectively kill these tumors by hitting their Achilles’ heel.4 In fact, several large randomized trials have consistently shown that, in terms of progression-free survival (PFS), EGFR-TKI is superior to conventional chemotherapy in treatment-naïve EGFR-mutant lung adenocarcinoma patients.5–8
Sanger direct sequencing (DS) is the classic method to detect gene mutations and is considered the current standard method. DS can detect both known and novel mutations; however, its detection sensitivity is low and at least 25% of the sample needs to be mutant DNA.9 Its low detection sensitivity is not an issue in basic research but may cause significant deficiency in clinical practice. On one hand, there are not only EGFR-mutant cancer cells but also a lot of EGFR-wild-type (EGFR-WT) stromal cells in tumors. On the other hand, emerging evidence suggested that not all tumor cells harbor (the same) EGFR mutation in an EGFR-mutant tumor.10,11 Consequently, some clinical samples may give rise to a false-negative testing result by DS and this phenomenon may partly explain the high response rate of EGFR-TKI in EGFR-WT patients in previously reported clinical series including ours.12
In the past few years, several methods have been developed to improve the detection sensitivity of EGFR mutations.9 One strategy is to use the mutant allele-specific polymerase chain reaction (PCR) combined with the real-time detection technology, such as Scorpion/Amplification Refractory Mutation System (ARMS), which can lower the detection limit down to 1% mutant DNA present in the sample. Therefore, certain abundance of EGFR-mutant DNA may result in discrepant results determined by low- and high- sensitivity mutation detection methods. Whether the discrepant EGFR mutation results will contribute to different EGFR-TKI treatment outcomes in clinical patients is still controversial.13,14 In this study, by using high-sensitivity Scorpion/ARMS method, we reanalyzed the EGFR mutation status in serial lung adenocarcinoma samples which were found to have no EGFR mutations by DS method. Both the patient and tumor characteristics were evaluated and the patients’ clinical outcome to EGFR-TKI treatment was compared with those of a control group whose tumors were found to have EGFR mutations by DS method.
PATIENTS AND METHODS
The database at the Department of Pathology and Laboratory Medicine in Taipei Veterans General Hospital was used to retrieve tumors that were diagnosed with lung adenocarcinoma and had been tested for EGFR mutation by DS method from a period of 2008 to 2010. Among tumors diagnosed as EGFR-WT, we further identified cases with adequate remaining DNA (of the same batch for DS method) to submit for EGFR mutation testing by Scorpion/ARMS method. Patients whose tumors were found to have the sensitive EGFR mutations (exon 19 deletion, L858R, G719X, and L861Q) by DS method in the same time period were selected as the positive control group. Patients with other EGFR mutations were excluded because the function of those mutations is either EGFR-TKI-resisting or not clear. Patients’ clinical characteristics and EGFR-TKI treatment outcomes were reviewed through the electronic chart record and serial chest computed tomography imaging. This study was approved by the Committee of Pathology Specimens and the Institution Review Board of Taipei Veterans General Hospital.
Formalin-fixed, paraffin-embedded tissue sections were used for EGFR mutation testing. One of the consecutive sections was stained with hematoxylin and eosin and reviewed by pathologists to select tumor region(s) for genomic DNA extraction. The percentage of cancer cells in the selected region(s) was assessed in a 10% increment by two independent pathologists. The selected tumor region(s) was marked on deparaffined tissue sections, manually microdissected, and followed by genomic DNA extraction using PicoPure DNA extraction kit (Arcturus/Applied Biosystems, Foster City, CA). The extracted DNA was evaluated and quantified by Nanodrop 2000 (Thermo Fisher Scientific, Waltham, MA).
EGFR Mutation Detection by DS Method
Exons 18, 19, 20, and 21 of the EGFR gene were amplified as previously described12 with minor modification, namely, nested PCR was only performed on specimens when their first PCR products could not be visualized on 2% agarose gel electrophoresis. The first PCR was carried out in a volume of 25 μl containing 2 μl of DNA, ×1 Taq Master Mix Red (Ampliqon III, Odense, Denmark), and 0.5 μM of each primer. The PCR reaction was carried out for 35 cycles at 95°C for 40 seconds, at 56°C for 40 seconds, and at 72°C for 40 seconds, followed by 5 minutes extension at 72°C. For the nested PCR, the DNA amplification was performed with the same PCR reaction program by using 2 μl of first PCR products as template, ×1 Taq Master Mix Red, and 0.5 μM of each of primers. Sanger’s sequencing was performed with forward or reverse primers, and the sequence analysis was carried out by Mutation Surveyor software (SoftGenetics, State College, PA).
EGFR Mutation Detection by High-Sensitivity Scorpion/ARMS Method
The remaining DNA after DS analysis was subjected to high-sensitivity EGFR mutation analysis using EGFR RGQ PCR kit (Qiagen, Manchester, United Kingdom) on a Rotor-gene platform. The kit contained primers designed for detection of 29 common EGFR mutations (L858R, L861Q, S768I, T790M, 3 G719 missense mutations, 19 deletions in exon 19, and 3 insertions in exon 20) based on ARMS PCR and Scorpion detection technology. Analysis was carried out as described in the operation instruction of the kit.
The association between patients and tumor characteristics was analyzed by the χ2 test and the Fisher’s exact test. EGFR-TKI treatment response among different groups was compared by nonparametric test. Survival curves were plotted by the Kaplan–Meier method and compared by log-rank test. When multiple comparisons were performed, the cutoff level of α error was reduced using Bonferroni correction. Analyses and figures were carried out with PASW Statistics 18.0 (SPSS Inc., Chicago, IL).
Patient Characteristics and Tumor EGFR Mutation Status
From December 2008 to September 2011, a total of 851 samples were received for EGFR mutation testing by DS method and 471 (55.3%) of them were positive for mutation. Among those without EGFR mutations, 137 consecutive samples that fulfilled the inclusion criteria of this study were identified. Seven samples were excluded for further analysis: five were not primary lung cancer and two were cell block materials. Among these 130 primary lung adenocarcinoma tumors that were regarded as EGFR-WT by DS method, 28 were found to have EGFR mutations by Scorpion/ARMS method. Taking the high-sensitivity method as the standard, the false-negative rate of DS method to detect sensitive EGFR mutations was 21.5% (95% confidence interval, 14.4%–28.7%). Patient characteristics are summarized in Table 1. Discrepant EGFR mutation results were more common in samples from nonsmokers than in samples from smokers (30.7% versus 9.1%; p = 0.003). There was no difference in age and sex.
The Roles of Nested PCR and Tumor Cell Percentage on the Discrepant EGFR Mutation Results
Due to the limited amount of available tumor tissue, nested PCR was performed in most specimens (108 of 130, 83.1%) to make enough PCR products for DS. Most discrepant EGFR mutation results occurred in samples which needed nested PCR (27 of 108) and only one happened in those which made enough products by simple PCR (1 of 22; p = 0.044).
The distribution of the tumor cell percentage in each sample is shown in Figure 1. There was no significant association between the tumor cell percentage and discrepant EGFR mutation results (p = 0.102). In a further analysis, the false-negative rates of EGFR mutation testing by DS method were not significantly different in samples with tumor cell percentage either 50% or less or more than 50% (26.2% versus 16.9%; p = 0.201).
The Roles of Tissue Types on the Discrepant EGFR Mutation Results
The most common type of specimens sent for EGFR mutation analysis was lung tissue (n = 104), followed by lymph nodes (n = 14), pleura (n = 8), and others (n = 4). Interestingly, the false-negative rates of EGFR mutation testing by DS method were significantly higher in pleural than nonpleural samples (62.5% versus 18.9%; p = 0.012). Among the eight pleural samples, seven were obtained by percutaneous biopsy and one was through video-assisted thoracoscopy. In seven percutaneous biopsied samples, four were indeed EGFR-mutant tumors but missed by DS method. The tumor cell percentages in these samples were 60%, 50%, 40%, and 10%, respectively. DS method also failed to identify the EGFR mutation in the thoracoscopic biopsied sample which contained 10% tumor cells.
Survival in Patients with Discrepant EGFR Mutation Testing Results
Eighty-five of the 130 patients in our cohort had received single-agent EGFR-TKI treatment; 64 (group A) had EGFR-WT tumors defined by both mutation detection methods, and 21 (group B) had tumors with EGFR mutations which were detected by Scorpion/ARMS but not DS method. A positive control group (group C) from the same time period was established, consisting of 58 lung adenocarcinoma patients who had sensitive EGFR mutations detected by DS method and had been treated with EGFR-TKI. Patient and tumor characteristics are summarized in Table 2.
During EGFR-TKI treatment, patients with concordant testing results of EGFR-WT tumors had a shorter PFS than those with EGFR-mutant tumors defined by either Scorpion/ARMS or DS method (median PFS = 1.8, 13.4, and 10.9 months, respectively; p < 0.001; Fig. 2). The overall survival (OS) was also significantly different among these patients (median OS = 18.2, 37.6 and 27.5 months, respectively; p = 0.001). Interestingly, there was no difference in both PFS and OS among the patients with EGFR-mutant tumors no matter they were diagnosed by low-sensitivity DS method or high-sensitivity Scorpion/ARMS method alone (p = 0.225 and 0.594, respectively).
EGFR-TKI Treatment Response in Patients Categorized by Different EGFR Mutation Testing Methods
According to the definition used in Response Evaluation Criteria in Solid Tumors criteria, 41, 19, and 55 patients had measurable disease in groups A, B, and C, respectively. The objective response rate was significantly higher in patients with EGFR-mutant tumors (groups B and C combined) than those with EGFR-WT tumors (group A) (70.3% versus 9.8%; p < 0.001). However, there was no difference between patients whose tumor EGFR mutation status was analyzed using Scorpion/ARMS (group B) or DS (group C) method (73.7% versus 69.1%; p = 0.706).
In this study, in a high EGFR mutation prevalence area, we showed that Scorpion/ARMS method could identify sensitive EGFR mutations in up to 22% of lung adenocarcinomas which were considered EGFR-WT by DS method. We also demonstrated that patients with EGFR mutations missed by DS had similar response rate and survival to those with mutations detected by DS. These findings highlight the major limitation of DS for EGFR mutation detection in current clinical practice. It has been clear that EGFR-TKI can provide better treatment outcome and quality of life than cytotoxic chemotherapy as the front-line therapy in patients with EGFR-mutant non–small-cell lung cancer. Therefore, using DS as the primary strategy for EGFR mutation detection could significantly compromise patients’ prognosis because it may misplace one in every five DS-defined “EGFR-WT” patients into inappropriate treatment.
The failure rate of EGFR mutation detection by DS method in our study was similar to a few previous reports15,16 but higher than the others.17,18 We consider it is reasonable because the negative predictive value of a test should be closely related to the prevalence of the disease in a given population. Therefore, it was not surprising that EGFR mutation detection failure rates by a low-sensitivity method were higher in East-Asian studies,15,16 including ours, than those with Caucasian-predominant population.17,18 It also explained our observation that discrepant EGFR mutation results by low- and high-sensitivity methods were more common in nonsmokers than smokers because EGFR mutations were more frequent in nonsmoker lung adenocarcinomas.19
Low percentage of tumor cells in clinical samples was generally considered the reason for the false-negative result of EGFR mutation by DS, and most recommendations suggested that the tumor cell percentage should be at least 50%.20–22 However, when we vigorously examined the tumor cell percentage of each sample in this study, we could not find a correlation. We therefore speculated that the percentage of mutant DNA, but not tumor cell per se, was the decisive factor for the sensitivity of mutation detection.9 It is well documented that EGFR gene amplification is a frequent event in EGFR-mutant tumors. In addition, more complicatedly, an EGFR-mutant tumor may actually consist of not only EGFR-mutant but also EGFR-WT tumor cells.10,11 Therefore, the summation of the percentage of tumor cells in a given sample, the proportion of EGFR-mutant versus EGFR-WT tumor cells, and EGFR gene amplification status in EGFR-mutant tumor cells will contribute to the final ratio of mutant-to-WT DNA and then the result of EGFR mutation detection by DS method.
In addition to the mutant DNA percentage in raw samples, types of PCR reaction (simple or nested PCR) will also play a role in the final result of Sanger sequencing. Theoretically, in a sample with a fixed mutant-to-WT DNA ratio, the final mutant-to-WT signals detected by DS will be amplified with the increasing cycles of PCR reaction. Because of limited amount of DNA obtained from clinical samples, nested PCR was routinely performed before DS in our earlier study.12 Along with the improvement of the personnel and technique, nested PCR has been carried out only when the first PCR products could not be visualized on agarose gels. As expected, although nested PCR still had to be executed for most specimens (83.1%), avoidance of the second PCR significantly reduced the false-negative rate of EGFR mutation detection by DS method.
Interestingly, we found that tissue type may also be a factor contributing to the EGFR mutation testing results by DS method. The false-negative rate was extraordinarily high in pleural specimens (62.5%). The exact mechanism was not clear. Both the tumor cell percentage and the necessity of nested PCR did not differ significantly in pleural versus nonpleural tissues. One of the possible reasons was that tumor cell percentage in pleural samples was indeed significantly low but was overestimated because of the difficulty in differentiating between adenocarcinoma cells and reactive mesothelial cells by hematoxylin and eosin staining.23 Another plausible explanation was that reactive mesothelial cells might have EGFR gene amplification, which has not yet been explored. Finally, and maybe most likely, it could simply be a biased conclusion caused by a small sample size (n = 8).
Predictably, patients with EGFR-mutant tumors have better response and longer survival than those with EGFR-WT tumors when they receive EGFR-TKI treatment. However, in contrast to a previous study by Zhou et al.,13 we did not find significant differences in objective response rate and PFS in patients with tumors bearing EGFR mutations detected by DS method or only by high-sensitivity Scorpion/ARMS method. There were several differences in terms of methodology between Zhou’s study and ours. First, they included patients with not only adenocarcinoma but also other non–small-cell types. Second, all their patients received gefitinib, but our patients received gefitinib and erlotinib at about equal frequency. Third, they only included patients with tumor samples which consisted of more than 50% tumor cells. In our study, about 60% of the samples had tumor cells less than 50%. Although they had excluded samples with lower tumor cell percentage, the false-negative rate of EGFR mutation by DS method in their series was similar to ours (26.1% and 21.5%, respectively). This finding concurs with our conclusion that tumor cell percentage itself does not have significant impact on EGFR mutation testing by DS method.
There are several limitations in this study. First of all, it was a retrospective study; therefore, the clinical data were not perfectly collected. Second, we were not able to evaluate the EGFR gene copy number which in combination with tumor cell percentage could give a better estimation of the mutant DNA amount. However, because EGFR-mutant tumors may also have WT EGFR allele, we have to develop a method which can differentially detect WT and mutant EGFR gene copy numbers. Third, although Scorpion/ARMS method is highly sensitive, it could detect only 29 variants of EGFR mutations. Tumors with some rare mutations may still have response to EGFR-TKI.24,25 It could partly explain why some patients whose tumors were EGFR-WT by Scorpion/ARMS method still achieved partial response in this study. Fourth, we did not perform Scorpion/ARMS assay in our positive control group (EGFR-mutant by DS method).
In conclusion, EGFR mutation testing by DS method may misdiagnose up to one-fifth of “EGFR-WT” lung adenocarcinoma patients in a high EGFR mutation prevalence area. This can have significant adverse impact on the treatment decision for both physicians and patients. It will also severely jeopardize the clinical outcomes of EGFR-mutant patients if they are excluded from the use of EGFR-TKI due to a false-negative testing result. Our study suggested a similar EGFR-TKI response and survival in patients with EGFR mutation no matter it was determined by DS or Scorpion/ARMS alone. Future prospective studies are warranted to recapitulate our current retrospective observation.
The authors thank Zih-Ying Wu for providing the technical assistance. Parts of this study were supported financially by grant V102C-195 from Taipei Veterans General Hospital, grant NSC99-2320-B-010-023-MY3 from the National Science Council, grant “Center of Excellence for Cancer Research at Taipei Veterans General Hospital” (DOH101-TD-C-111-007) from the Department of Health, and grant “Aim for the Top University Plan” from the Ministry of Education, Taiwan.
1. Nicholson RI, Gee JM, Harper ME. EGFR and cancer prognosis. Eur J Cancer. 2001;37(Suppl 4):S9–S15
2. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med. 2004;350:2129–2139
3. Paez JG, Jänne PA, Lee JC, et al. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science. 2004;304:1497–1500
4. Gazdar AF, Shigematsu H, Herz J, Minna JD. Mutations and addiction to EGFR: the Achilles ‘heal’ of lung cancers? Trends Mol Med. 2004;10:481–486
5. Mitsudomi T, Morita S, Yatabe Y, et al.West Japan Oncology Group. Gefitinib versus cisplatin plus docetaxel in patients with non-small-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): an open label, randomised phase 3 trial. Lancet Oncol. 2010;11:121–128
6. Maemondo M, Inoue A, Kobayashi K, et al.North-East Japan Study Group. Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR. N Engl J Med. 2010;362:2380–2388
7. Zhou C, Wu YL, Chen G, et al. Erlotinib versus chemotherapy as first-line treatment for patients with advanced EGFR mutation-positive non-small-cell lung cancer (OPTIMAL, CTONG-0802): a multicentre, open-label, randomised, phase 3 study. Lancet Oncol. 2011;12:735–742
8. Rosell R, Carcereny E, Gervais R, et al.Spanish Lung Cancer Group in collaboration with Groupe Français de Pneumo-Cancérologie and Associazione Italiana Oncologia Toracica. Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol. 2012;13:239–246
9. Pao W, Ladanyi M. Epidermal growth factor receptor mutation testing in lung cancer: searching for the ideal method. Clin Cancer Res. 2007;13:4954–4955
10. Taniguchi K, Okami J, Kodama K, Higashiyama M, Kato K. Intratumor heterogeneity of epidermal growth factor receptor mutations in lung cancer and its correlation to the response to gefitinib. Cancer Sci. 2008;99:929–935
11. Jiang SX, Yamashita K, Yamamoto M, et al. EGFR genetic heterogeneity of nonsmall cell lung cancers contributing to acquired gefitinib resistance. Int J Cancer. 2008;123:2480–2486
12. Chou TY, Chiu CH, Li LH, et al. Mutation in the tyrosine kinase domain of epidermal growth factor receptor is a predictive and prognostic factor for gefitinib treatment in patients with non-small cell lung cancer. Clin Cancer Res. 2005;11:3750–3757
13. Zhou Q, Zhang XC, Chen ZH, et al. Relative abundance of EGFR mutations predicts benefit from gefitinib treatment for advanced non-small-cell lung cancer. J Clin Oncol. 2011;29:3316–3321
14. Kim HJ, Lee KY, Kim YC, et al. Detection and comparison of peptide nucleic acid-mediated real-time polymerase chain reaction clamping and direct gene sequencing for epidermal growth factor receptor mutations in patients with non-small cell lung cancer. Lung Cancer. 2012;75:321–325
15. Kimura H, Fujiwara Y, Sone T, et al. High sensitivity detection of epidermal growth factor receptor mutations in the pleural effusion of non-small cell lung cancer patients. Cancer Sci. 2006;97:642–648
16. Liu Y, Liu B, Li XY, et al. A comparison of ARMS and direct sequencing for EGFR mutation analysis and tyrosine kinase inhibitors treatment prediction in body fluid samples of non-small-cell lung cancer patients. J Exp Clin Cancer Res. 2011;30:111
17. Ellison G, Donald E, McWalter G, et al. A comparison of ARMS and DNA sequencing for mutation analysis in clinical biopsy samples. J Exp Clin Cancer Res. 2010;29:132
18. Angulo B, Conde E, Suárez-Gauthier A, et al. A comparison of EGFR mutation testing methods in lung carcinoma: direct sequencing, real-time PCR and immunohistochemistry. PLoS One. 2012;7:e43842
19. Shigematsu H, Gazdar AF. Somatic mutations of epidermal growth factor receptor signaling pathway in lung cancers. Int J Cancer. 2006;118:257–262
20. Sequist LV, Joshi VA, Jänne PA, et al. Epidermal growth factor receptor mutation testing in the care of lung cancer patients. Clin Cancer Res. 2006;12(14 Pt 2):4403s–4408s
21. Eberhard DA, Giaccone G, Johnson BENon-Small-Cell Lung Cancer Working Group. . Biomarkers of response to epidermal growth factor receptor inhibitors in Non-Small-Cell Lung Cancer Working Group: standardization for use in the clinical trial setting. J Clin Oncol. 2008;26:983–994
22. Pirker R, Herth FJ, Kerr KM, et al.European EGFR Workshop Group. Consensus for EGFR mutation testing in non-small cell lung cancer: results from a European workshop. J Thorac Oncol. 2010;5:1706–1713
23. Pereira TC, Saad RS, Liu Y, Silverman JF. The diagnosis of malignancy in effusion cytology: a pattern recognition approach. Adv Anat Pathol. 2006;13:174–184
24. He M, Capelletti M, Nafa K, et al. EGFR exon 19 insertions: a new family of sensitizing EGFR mutations in lung adenocarcinoma. Clin Cancer Res. 2012;18:1790–1797
25. Wu JY, Yu CJ, Chang YC, Yang CH, Shih JY, Yang PC. Effectiveness of tyrosine kinase inhibitors on “uncommon” epidermal growth factor receptor mutations of unknown clinical significance in non-small cell lung cancer. Clin Cancer Res. 2011;17:3812–3821
Lung adenocarcinoma; Epidermal growth factor receptor; Mutation; Tyrosine kinase inhibitor
Copyright © 2014 by the European Lung Cancer Conference and the International Association for the Study of Lung Cancer.