Virchows Arch DOI 10.1007/s00428-017-2226-8
ORIGINAL ARTICLE
EGFR T790M mutation testing of non-small cell lung cancer tissue and blood samples artificially spiked with circulating cell-free tumor DNA: results of a round robin trial Jana Fassunke 1 & Michaela Angelika Ihle 1 & Dido Lenze 2 & Annika Lehmann 2 & Michael Hummel 2 & Claudia Vollbrecht 2 & Roland Penzel 3 & Anna-Lena Volckmar 3 & Albrecht Stenzinger 3 & Volker Endris 3 & Andreas Jung 4 & Ulrich Lehmann 5 & Silke Zeugner 6 & Gustavo Baretton 6 & Hans Kreipe 5 & Peter Schirmacher 3 & Thomas Kirchner 4 & Manfred Dietel 2 & Reinhard Büttner 1 & Sabine Merkelbach-Bruse 1
Received: 7 March 2017 / Revised: 22 July 2017 / Accepted: 22 August 2017 # Springer-Verlag GmbH Deutschland 2017
Abstract The European Commision (EC) recently approved osimertinib for the treatment of adult patients with locally Virchows Archiv conforms to the ICMJE recommendation for qualification of authorship. The ICMJE recommends that authorship be based on the following 4 criteria: • Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND • Drafting the work or revising it critically for important intellectual content; AND • Final approval of the version to be published; AND • Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This article is part of the Topical Collection on Quality in Pathology Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00428-017-2226-8) contains supplementary material, which is available to authorized users. * Jana Fassunke
[email protected]
1
Institue of Pathology, University Hospital Cologne, Kerpener Strasse 62, 50924 Cologne, Germany
2
Institute of Pathology, Charité—University Hospital Berlin, Campus Mitte, Virchowweg 15, 10117 Berlin, Germany
3
Institute of Pathology, University Hospital Heidelberg, Im Neuenheimer Feld 224, Heidelberg, Germany
4
Pathologisches Institut of the Ludwig-Maximilinas-Universität München, Thalkirchnerstrasse 36, 80336 Munich, Germany
5
Institute of Pathology, Hannover Medical School, Carl-Neuberg-Strasse 1, 30625 Hannover, Germany
6
Institute of Pathology, University Hospital Dresden Carl Gustav Carus, Fetscherstrasse 74, 01307 Dresden, Germany
advanced or metastatic non-small-cell lung cancer (NSCLC) harboring EGFR T790M mutations. Besides tissue-based testing, blood samples containing cell-free circulating tumor DNA (ctDNA) can be used to interrogate T790M status. Herein, we describe the conditions and results of a round robin trial (RRT) for T790M mutation testing in NSCLC tissue specimens and peripheral blood samples spiked with cell line DNA mimicking tumor-derived ctDNA. The underlying objectives of this twostaged external quality assessment (EQA) approach were (a) to evaluate the accuracy of T790M mutations testing across multiple centers and (b) to investigate if a liquid biopsy-based testing for T790M mutations in spiked blood samples is feasible in routine diagnostic. Based on a successfully completed internal phase I RRT, an open RRT for EGFR T790M mutation testing in tumor tissue and blood samples was initiated. In total, 48 pathology centers participated in the EQA. Of these, 47 (97.9%) centers submitted their analyses within the predefined time frame and 44 (tissue), respectively, 40 (plasma) successfully passed the test. The overall success rates in the RRT phase II were 91.7% (tissue) and 83.3% (blood), respectively. Thirty-eight out of 48 participants (79.2%) successfully passed both parts of the RRT. The RRT for blood-based EGFR testing initiated in Germany is, to the best of our knowledge, the first of his kind in Europe. In summary, our results demonstrate that blood-based genotyping for EGFR resistance mutations can be successfully integrated in routine molecular diagnostics complementing the array of molecular methods already available at pathology centers in Germany. Keywords T790M mutation . Osimertinib . EGFR . NSCLC . FFPE . Tumor tissue . Blood and plasma samples . Liquid biopsy . DNA sequencing
Virchows Arch
Introduction Until recently, non-small-cell lung cancer (NSCLC) was a neoplasm with limited therapeutic options and dismal prognosis [1]. However, over the past decade, we have learned that the historically known and histologically defined subtypes of NSCLC that relied primarily on morphologic features can be further characterized at the molecular level by oncogenic driver mutations that can be exploited therapeutically. EGFR mutations in NSCLC patients The epidermal growth factor receptor (EGFR)-rat sarcoma (RAS)/RAS-associated factor (RAF)/mitogen-activated protein kinase (MAPK) signaling cascade is an important pathway in cancer development. In 2004, two independent research groups identified activating somatic mutations in the tyrosine kinase (TK) domain of the EGFR in NSCLC patients [2, 3]. These mutations, ranging from 10 to 15% of the specimens in the USA [4] and Europe [5] to 30–50% in East Asian and 23% in Indian populations [6, 7], occur mostly within the TK encoding exons 18–21 of the EGFR gene and are highly co-existent with an increased sensitivity to first- and secondgeneration EGFR tyrosine kinase inhibitors (TKI) gefitinib, erlotinib, and afatinib [2, 3, 6, 8, 9]. Oncogenic mutations of EGFR occur with a significantly higher frequency in NSCLC with adenocarcinoma histology, patients of younger age, females, and light (< 10 pack years) or non-smokers [10]. Deletions within exon 19 and the single-point substitution mutation L858R within exon 21 account for almost 90% of all EGFR mutations [11]. Followed by this, we witnessed a crucial and rapid development in both molecular genotyping and individualized therapy in the management of NSCLC harboring driver mutations. With the growing number of oncogenic drivers being identified and the emergence of targeted therapeutic options as new standard of care in these settings, molecular pathology is increasingly called upon to provide comprehensive and reliable genotyping of lung cancer subsets not only by standard tissue biopsy, but also by low-invasive procedures such as liquid biopsy. Liquid biopsy In a substantial number of patients suffering from NSCLC open surgery, biopsy samples cannot be obtained [12]. Closed biopsies such as fine-needle aspirations (FNAs) or core needle biopsies (CNB) are less invasive procedures, but still cause discomfort to patients, require local anesthesia, and bear the risk of undue complications, the commonest being pneumothorax. Needle biopsies may not always be feasible because of the location of the tumor and/or bear the risk of providing insufficient material [12, 13]. Moreover, taking
blood samples is less expensive than taking a biopsy, and the turnaround time of processing the blood sample is shorter. The genomic profile of a blood sample can be generated more quickly [14]. Genotyping of surrogate sources of tumor DNA including biofluids such as blood, which contains circulating cell-free tumor DNA (ctDNA) or circulating tumor cells (CTC), is an evolving field and a new strategy for tumor genotyping, which has vast clinical implications [12]. In the upcoming era of personalized medicine, the molecular screening of a patient’s tumor, monitoring of resistance mechanisms, and treatment response are only some examples where liquid biopsy might find its place. To what extend liquid biopsies will change clinical practice and influence patients’ outcome is currently under debate [15, 16]. EGFR T790M mutation in advanced NSCLC The development of acquired resistance is a major clinical problem for the targeted therapy of cancer patients. NSCLC patients with sensitizing EGFR mutations treated with gefitinib or erlotinib have an overall response rate (ORR) of 60–80% [8, 17] despite promising initial responses. However, the majority of the patients develop a resistance against firstgeneration tyrosine kinase inhibitors and experience progressive disease after a median of 9 to 13 months [11, 18]. The most frequently reported mechanism of secondary resistance is the EGFR point mutation T790M in exon 20 [19–21], which accounts for more than 50% of progressing tumors [18, 21, 22]. After the manifestation of T790M mutation, the median survival time is less than 2 years [21]. The unmet need of a targeted therapeutic strategy for patients with acquired EGFR TKI resistance has led to the development of highly selective third-generation EGFR TKIs. Osimertinib (AZD9291, Tagrisso®) is a US Food and Drug Administration (FDA)-[23] and European Medicine Agency (EMA)-approved [24] irreversible covalent-bound inhibitor of EGFR sensitizing (EGFRm) and resistance (T790M) mutations, with a reduced potency against the wild-type EGFR [25]. This EGFR TKI received an accelerated approval as the first indicated treatment for patients with EGFR T790M mutation-positive metastatic NSCLC [26]. Meanwhile, a phase III trial (AURA3 ClinicalTrials.gov: NCT02151981) has confirmed the efficacy of osimertinib after failure of first-generation EGFR TKI [27]: Osimertinib significantly improved progression-free survival (PFS) compared with platinum-pemetrexed: hazard ratio 0.30; 95% confidence interval: 0.23, 0.41; p < 0.001 (median 10.1 vs 4.4 months). The result was consistent with PFS analysis by blinded independent central review (BICR): HR 0.28; 95% CI: 0.20, 0.38; p < 0.001 (11.0 vs 4.2 months). ORR was significantly improved with osimertinib (71%) vs platinum-pemetrexed (31%); odds ratio 5.39 (95% CI: 3.47, 8.48; p < 0.001). Median duration of
Virchows Arch
response (DoR) was 9.7 months (95% CI 8.3, 11.6) with osimertinib and 4.1 months (95% CI 3.0, 5.6) with platinum-pemetrexed. After an initial conditional approval, osimertinib has been now granted a full approval by the EU Commission. To make appropriate use of this new therapeutic option, accurate testing for the presence of T790M mutation is an essential and required step in the subsequent treatment decision. However, if tissue availability limits the genotyping of EGFR T790M mutation, blood-based EGFR mutation testing might be an alternative or complementary to genotyping of tumor tissue. The first objective of this study was to evaluate the accuracy (clinical validity) of T790M mutations testing in a round robin trial (RRT) based on external quality assessment (EQA) employing tissue specimens. The second objective was to evaluate if liquid biopsy-based testing of T790M mutations using artificially spiked blood samples is feasible. This also includes the preparation, detection, and shipment of such blood samples in order to allow a molecular testing at remote sites. This RRT for blood-based EGFR testing initiated in Germany is currently to our knowledge the first one in Europe.
Material and methods Patient material Tissue specimens and processing For the preparation of the RRT 22 of formalin-fixed and paraffin-embedded (FFPE) non-small-cell lung cancer tissue, specimens were collected from the archives of the Institutes of Pathology in Cologne, Berlin, Heidelberg and Hannover (Supplementary Table 1). This cohort included samples that were positive for the EGFR T790M mutation as well as tissue specimens positive for EGFR mutations other than T790M and wild type (wt) for EGFR. All tissue blocks had at least a tumor cell content > 20%. Five blocks had enough material to split the tissue. The cases were histologically re-evaluated by experienced pathologists from the Institute of Pathology in Cologne according to the combined current criteria of the World Health Organization Classification and of the Association for the Study of Lung Cancer/American Thoracic Society/European Respiratory Society classification [28, 29]. FFPE tissue samples were obtained as part of routine clinical care under approved ethical protocols in compliance with the local ethics committees. Blood collection and processing Peripheral blood samples were obtained from multiple healthy volunteers from the institute of pathology and the lung cancer group (Cologne). The samples were spiked with fragmented genomic DNA from the EGFR T790M and L858R-mutated cell line H1975 and the
EGFR wt cell line MCF-7 (both purchased from the American Type Culture Collection [ATCC]) which served as a surrogate for ctDNA. Therefore, DNA was crushed into fragments of about 165 bp size using a Covaris system (Covaris, Woburn, MA, USA) using appropriate settings (volume 50 μl; peak incident power 175 W; duty factor 10%; cycles per burst 200; treatment time 200 s). DNA quality was checked with the DNF-474 High Sensitivity NGS Fragment Analysis Kit on a Fragment Analyzer™ (Advanced Analytical). Only probes with a clear peak at 165 bp heights were spiked into blood samples of healthy donors. Blood specimens were spiked with different amounts of DNA immediately and shipped the same day. Validation of the RRT test cases for phase I and II Pre-testing of all cases prior to the delivery for the open RRT was done as a collective effort by the institute of pathology in Berlin, Dresden, Cologne, Hannover, Heidelberg, and Munich. For the phase I RRT, 10 test cases with suitable amounts of tissue blocks and DNA quality were selected and validated in Cologne. The mutation status of the EGFR gene of each case was assessed. For the phase II RRT, additional tissue blocks had to be validated, and some of the blocks from phase I was split due to the high need of tissue material (Suppl. Table 1). Additional blocks from Cologne were cross-checked successfully in Hannover and Heidelberg and three blocks containing T790M-mutated NSCLC tissue from Berlin, Hannover, and Heidelberg were successfully validated in Cologne. All FFPE blocks had tissue with a tumor cell content of ≥ 20% which was macro-dissected from two 10-μm thick sections. After overnight hydrolysis in a proteinase K containing buffer, DNA was extracted using the Maxwell 16 LEV DNA kit (Promega, Madison, WI, USA) as described previously [30]. Mutational verification was performed by massive parallel sequencing using a custom QIAGEN Lung Cancer Panel (QIAGEN, Hilden, Germany) and the Illumina MiSeq platform (Illuminia, San Diego, CA, USA) at the lead center in Cologne. Sequencing was carried out with 2 × 150 bp and the V2 chemistry according to the manufacturer’s instructions. For all samples, the results matched the previous findings. To assess the amplification performance of the short DNA fragments which were used for spiking of the blood samples, the artificially generated ctDNA was analyzed also by parallel sequencing as described above (data not shown). Different amounts (twice 10, 22, 45, and 96 ng of the H1975 cell line, twice 100 ng of the MCF-7 cell line) of the artificial ctDNA were pipetted into 10-ml Cell-Free DNA BCT blood collection tubes (Streck, La Vista, NE, USA), each containing 9 ml of fresh whole blood samples from healthy donors to simulate the presence of ctDNA. The lowest spiking amount is 10 ng which are 1515 genome copies of the heterozygous cell line H1975. That corresponds to a minor allele frequency (MAF)
Virchows Arch
of 5, 9% when we assume a total circulating cell-free DNA (ccfDNA) amount of 75 ng from a healthy blood donor. Twenty-four hours after the blood draw, plasma was isolated by centrifugation at 3000 rpm for 10 min at 4 °C in a 50-ml collection tube. The supernatant, thus plasma, was subsequently centrifuged at high speed (at 16,000 g, 10 min, 4 °C) in low-binding DNA tubes (DNA LoBind Tubes, Eppendorf AG, Hamburg, Germany) and stored at − 80 °C. Before ctDNA extraction, the samples were equilibrated to room temperature (15–25 °C). An additional blood sample was collected from each healthy donor to determine the ccfDNA concentration. CcfDNA concentration from healthy donors ranged from 0, 12–0, 82 ng/μl. The ctDNA extraction was carried out using the QiaSymphony device (QIAGEN, Hilden, Germany) using the circulating DNA kit according to the manufacturer’s protocol. Mutational analysis was performed by massively parallel sequencing using an Ion AmpliSeq Custom DNA Panel (Thermo Fisher Scientific, Waltham, MA, USA) and the Ion AmpliSeq Library Kit 2.0 (Thermo Fisher Scientific) according to the Ion AmpliSeq Library Preparation User Guide (Thermo Fisher Scientific). The Ion AmpliSeq Library Kit 2.0 is well-suited for a low DNA input amount. After multiplex PCR, libraries were generated by adapter ligation and target enrichment using the Gene Read DNA Library I Core Kit, the Gene Read DNA I Amp Kit (QIAGEN, Hilden, Germany), and the NEXTflex DNA Barcodes (Bioo Scientific, Austin, TX, USA) [31]. Twelve pM of the constructed libraries were sequenced on the MiSeq (Illumina, San Diego, CA, USA) with a MiSeq reagent kit V2 (300 cycles) (Illumina) following the manufacturer’s recommendations. For all samples, the results matched the previous findings.
Table 1 Reporting framework and scoring scheme of round robin trial (RRT) phase I and II
Setup of RRT phase I For the internal RRT (phase I), 10 FFPE samples (2 × 10 μm slides and one slide for H&E staining) were sent to the five additional participating reference centers (Berlin, Dresden, Hannover, Heidelberg, Munich) on December 9th, 2015, and arrived either the next day or the day after. The test cohort consisted of two cases with wild-type EGFR status, five cases of EGFR T790M-mutated samples (including one with an allele frequency of 5–10%) and three samples with EGFR mutations (EGFR Ex 19 or 21 mut.), but wild type for T790M. All participants in the phase I RRT were blinded to the sequencing results of the samples. The results had to be returned by December 23, 2015. The blood samples were taken on January 4, 2016 and spiked with the artificial ctDNA for EGFR T790M mutation in different amounts (10, 22, 45, and 96 ng) of the human NSCLC cell line H1975, which harbors both the EGFR exon 20 mutation T790M and the exon 21 mutation p.L858R. For the EGFR wild-type samples, the human breast adenocarcinoma cell line MCF-7 was used for spiking (100 ng). The blood samples were shipped on the day of blood draw and arrived the following day. The results had to be returned by January 20, 2016. In parallel, a blinded blood sample test set was analyzed in Cologne. All six participants were requested to provide the following information upon completion of the analysis: method of ctDNA extraction, analysis method, T790M status, and allele frequency of the mutation (Table 1). A total of 20 points could be reached for each test, and the tests could be passed separately (Table 1). The FFPE test was passed with 19, and the blood test with 18 out of 20 points. As the blood test had to be prepared freshly on the day of the
Reporting results for 10 FFPE and 10 blood samples – Turnaround time: 14 days – T790M mutated, yes or no – ctDNA extraction method – Analysis method for FFPE and blood tests – cDNA change, amino acid change, and allele frequency Scoring scheme RRT phase I and II Material Maximum score Successful pass Correct call (T790M wild-type or T790M-mutated) Exclusion of 1 blood samplea Technical problem with 1 sample
10 FFPE samples 20 points
10 blood samples 20 points
19 of 20 points 2 points
18 of 20 points 2 points 20–2×–2 1-point penalty
1-point penalty
Essential criteria and requirements of the RRT: Each participating test center had to report the specific analysis results stated above. Performance of each participant was assessed by a pre-defined scoring system a
Phase I: Sample drops out if more than three participants failed to analyze the sample. Phase II: dropout in a confirmatory testing at the centers in Berlin and Munich
Virchows Arch
shipping, it was not possible to evaluate and analyze the samples beforehand. In the case of a sample dropout of more than three participants, one sample would have been excluded from the calculation of the total points. The test would then have been passed with 20–2×–2 points. Setup of RRT phase II For the open RRT (phase II) comprising 42 registered participants, a total of four splits of the FFPE tissue blocks were generated. Tissue blocks without sufficient amounts of tissue or tumor cells were replaced by a comparable block in the next split. From each individual FFPE block, 12 test sets were prepared containing two slides for genotyping (10 μm) and one slide for H&E staining (4 μm). The last set was re-tested by the lead institute in Cologne for the EGFR mutation status. Due to the time consuming fresh preparation of the blood samples, we chose 2 days for shipment of the samples in March 2016 and 2 days in April. In total, blood samples (14 Streck collection tubes at a time) from 40 healthy donors were collected to prepare the blood sets. Each blood set comprised eight mutated and two wt samples. To identify possible dropouts, one blood set from each shipping day was sent to the Department of Pathology either in Berlin or Munich for confirmatory testing. All samples were shipped on March 15 and 16 or April 19 and 20, 2016, and arrived the following day. The results had to be returned by April 2 or May 2, 2016, respectively. With the exception of the dropout criteria, the criteria and the requirements of the RRT used in phase I (Table 1) were also used to determine the performance of the participant centers in RRT phase II.
Results On December 2015, shortly after Tagrisso™ (osimertinib, AZD9291) was approved by the US Food and Drug Administration (FDA) and shortly before it was granted a conditional approval by the European Commision as treatment for patients with EGFR T790M mutation-positive metastatic non-small cell lung cancer on February 2016, the QuIP (Quality Initiative in Pathology, the official German organization for quality control in pathology) launched a two-staged RRT to ensure quality-approved T790M mutation testing for EGFR. Pre-screening and RRT phase I (reference centers) Tissue samples From the sample pool described in Material and methods, five test sets each consisting of two cases of NSCLC with wt EGFR, five with EGFR-T790M mutation, and three with
EGFR mutations other than T790M were produced. For each case, three tissue sections were prepared with one being H&Estained and two for DNA extraction. These sets were sent to the five reference centers (Berlin, Dresden, Hannover, Heidelberg, and Munich) for a phase I RRT. Additionally three T790M-mutated blocks from Berlin, Hannover, and Heidelberg were sent to Cologne anticipating a higher number of participants for this EQA. These three blocks were successfully validated in Cologne (Supplementary Table 1). All reference centers completed the analysis within the pre-defined time frame of 14 days which was one of the parameters of the EQA to warrant fast turnaround times. Of the six reference centers in phase I, different technologies were used to analyze the samples where massively parallel sequencing was the most frequent (five of six; Table 2). In summary, all sites completed the phase I RRT successfully (Fig. 1).
Blood samples Peripheral blood samples from healthy donors were spiked with artificial ctDNA containing EGFR T790M mutation (NSCLC cell line H1975) in different concentrations (10, 22, 45, and 96 ng) or EGFR wt samples (human breast adenocarcinoma cells line MCF-7) resulting in a concentration of 100 ng (Fig. 2). Again, all reference centers completed their analysis within the specified time frame of 14 days. In this setting, the number of different techniques was smaller and all but one centre employed massively parallel sequencing (Table 3); Sanger sequencing was not applied. Sample no. 11 containing the smallest amount of fragmented gDNA (10 ng) dropped out by more than three participants and was therefore excluded from the evaluation so that the EQA was passed with 16 (20– 2×–2) points (Table 1). In total, five out of six sites successfully completed the phase I RRT liquid biopsy (Fig. 2). Taken together, the successful internal phase I RRT provided the following results. ➊ The analysis of FFPE tissue samples was successfully performed at all five reference centers in that the T790M mutation or the wild-type status was correctly identified in all tissue samples by all sites. ➋ Blood sample collection and shipping in 10 mL Streck BCTs were feasible. ➌ Fragmented genomic cell line DNA spiked-in in peripheral blood samples is a feasible setting to produce analyzable samples simulating liquid biopsies and ctDNA. ➍ The analysis of blood samples spiked with only 10 ng fragmented gDNA proved to be difficult since T790M detection was not possible by three of the reference centers. Therefore, we removed one of the two samples from the calculation. ➎ The RRT phase I was passed with 16 points. Five out of six reference centers passed the internal RRT for T790M testing form blood samples.
Virchows Arch
a
LDT laboratory-developed test
6
5a
Only the results of massively parallel sequencing were included in the evaluation. All used techniques revealed basically the same results
Tumor tissue
Custom Panel, QIAGEN
Colon Lung Panel, Thermo Fisher Scientific LCPv2 Custom Panel, Thermo Fisher Scientific LCPv1 Custom Panel, Biomers.net LDT, Pyromark Gold, QIAGEN Therascreen EGFR RGQ PCR Kit, V2, QIAGEN, LDT, BigDye Terminator v.3.1 ThermoFisher
Based on the results of phase I, an open RRT trial for EGFR T790M mutation testing in tumor tissue and blood samples was initiated. In total, 40 pathology centers from Germany and two centers from Switzerland registered for this RRT.
Massively parallel sequencing PGM/Ion Torrent, Thermo Fisher Scientific Massively parallel sequencing PGM/Ion Torrent, Thermo Fisher Scientific Massively parallel sequencing, Pyrosequencing, Real-time PCR, PGM/Ion Torrent, Thermo Fisher Scientific, Sanger sequencing Pyromark Q24 Rotorgene Q, QIAGEN ABI3130, Applied Biosystems Massively parallel sequencing MiSeq, Illumina 3 4
Colon Lung Panel v2, Thermo Fisher Scientifc EGFR Pyro Kit 24, V1, QIAGEN Therascreen EGFR RGQ PCR Kit, V2, QIAGEN PGM/Ion Torrent, Thermo Fisher Scientifc Pyromark Q24 Rotorgene Q, both QIAGEN Massively parallel sequencing Pyrosequencing Real-time PCR 1 2
Platform DNA sequencing method Reference center
Table 2
Overview of the detection technology used in RRT phase I (tissue samples)
Reagents
RRT phase II (open EQA)
Reflecting the large number of applying participants, nine additional FFPE tissue blocks and splits from tissue blocks that already had been tested were included (Supplementary Table 1). In addition, the phase II RRT test set consisted of six instead of five wild-type cases and of only four instead of five EGFR T790M-mutated cases. Of the 42 participants, the most often used method was pyrosequencing [13] followed by allele-specific PCR (n = 12), massive parallel sequencing (n = 9), and Sanger sequencing (n = 7). Participants used often more than one technique. Table 4 gives an overview of the methods of DNA sequencing, technology platforms, and gene panels that were applied. As none of the cases in this EQA failed to pass the internal phase I RRT, at least 19 points are necessary for successful participation. Thirty eight of 42 test centers (90.5%) received at least 19 points and thus completed the RRT successfully (Fig. 3). All of the participants using pyrosequencing [13] or massively parallel sequencing [9], respectively, passed the RRT successfully as did 11 of 12 participants with allelespecific PCR (92%) as well as 5 of 7 centers with Sanger sequencing (71%). Case number 10 of participant number 20 did not contain enough tumor cells and was therefore excluded from the evaluation. Participant number 20 detected two T790M-mutated samples using MALDI-TOF analysis with allele frequencies of 0.5 and 0.1%, respectively. Those two cases were wild type in the RRT phase I and II with all other methods and were thus kept classified as being wild type. Sequencing system: 0, not stated; 1, allele-specific PCR; 2, Sanger sequencing; 3, massive parallel sequencing; 4, pyrosequencing; 5, PCR and reverse hybridization; 6, matrixassisted laser desorption/ionization (MALDI)-TOF analysis; 7, melting point determination; 8, digital droplet PCR (ddPCR). Asterisk indicates no result notification, double asterisks indicate case number 10 did not contain enough tumor cells and was excluded from the result’s calculation, EGFRm = EGFR-mutated case. Blood samples Of the 42 test centers, allele-specific PCR (n = 19) was the most often used method for DNA sequencing, followed by massively parallel sequencing (n = 13), pyrosequencing
Virchows Arch Fig. 1 Results of tissue test of the RRT phase I. Numbers indicate achieved points; two score points indicate the correct identification of the T790M mutation or the wild-type status, respectively. In cases with only one point, technical problems with the sample emerged. wt wild type
Status of T790M mutation
Other EGFR mutations
1
Wt
2
Case
Reference center 1
2
3
4
5
6
Wt
2
2
2
2
2
2
p.T790M
p.E746_A750del
2
2
2
2
2
2
3
Wt
p.E746_A750del
2
2
2
2
2
2
4
p.T790M
p.L858R
2
2
2
2
2
2
5
p.T790M
p.L858R
2
2
2
2
2
2
6
Wt
p.E746_A750del
2
2
2
2
2
2
7
p.T790M
p.E746_A750del
2
2
2
2
2
2
8
Wt
Wt
2
2
2
2
2
2
9
p.T790M
p.L858R
2
2
2
2
2
2
10
Wt
p.E746_A750del
2
2
2
2
2
2
20
20
20
20
20
20
Score achieved (max. achievable test score: 20): Abbr.: wt, wild type.
(n = 10), and Sanger sequencing (n = 2). Table 5 gives an overview of the methods of DNA sequencing, technology platforms, and gene panels that were used for the detection of T790M mutation in blood samples. Fig. 2 Results of the RRT phase I (blood samples). Two (two points) stand for the correct identification of the T790M mutation or the wild-type status, respectively. In cases with only one point, technical problems with the sample emerged. wt stand for the incorrect classification as wild type. Number sign indicates sample could not be analyzed; Double number signs indicate sample 11 dropped out by more than three participants and was excluded from the calculation. The test was then passed with 20–2×–2 points
None of the tested samples showed a massive dropout so that the RRT was passed with at least 18 points. Thirty five of 42 test centers (83.3%) successfully completed this RRT (Fig. 4). Sixteen of 19 centers passed successfully with allele-
Sam ple
Spiked DNA in 9 mL blood [ng]
11##
10
p.T790M
12
96
13
Status of T790M mutation
Other mutations
Reference center 1
2
3
4
5
6
EGFR : p.L858R
2
wt
wt
wt
2
1#
p.T790M
EGFR: p.L858R
2
2
2
Wt
2
2
100
wt
PIK3CA: E545K
2
2
2
2
2
2
14
22
p.T790M
EGFR: p.L858R
2
2
wt
Wt
2
1#
15
96
p.T790M
EGFR: p.L858R
2
2
2
Wt
2
2
16
45
p.T790M
EGFR: p.L858R
2
2
2
Wt
2
2
17
22
p.T790M
EGFR: p.L858R
2
2
2
Wt
2
2
18
100
wt
PIK3CA: E545K
2
2
2
2
2
2
19
10
p.T790M
c.2573T>G p.L858R
2
2
2
Wt
2
2
20
45
p.T790M
c.2573T>G p.L858R
2
wt
2
Wt
2
2
Score achieved (max. achievable test score: 20##): 18
16
16
4
18
17
#
Sample could not be analyzed; ## Sample 11 dropped out by more than three participants
and was excluded from the calculation. The test was then passed with 20–2x–2 points.
Maxwell RSC ccfDNA Plasma kit, Promega Colon Lung Panel v2, Thermo Fisher Scientific QIAamp circulating nucleic acid kit, QIAGEN EGFR Pyro Kit 24, V1, QIAGEN Therascreen EGFR RGQ PCR Kit, V2, QIAGEN QIAamp circulating nucleic acid kit, QIAGEN Colon Lung Panel v2, Thermo Fisher Scientific QIAamp circulating nucleic acid kit, QIAGEN LCPv2 Custom Panel, Thermo Fisher Scientific QIAamp circulating nucleic acid kit, QIAGEN Custom Panel, QIAGEN LDT, PyroMark Gold, QIAGEN Therascreen EGFR RGQ PCR Kit, V2, QIAGEN AmoyDx, EGFR 29 Mutations Detection Kit, Amoy QIASymphony circulating DNA kit, QIAGEN Custom Panel, QIAGEN
specific PCR (84%), 11 of 13 centers with massively parallel sequencing (85%), and 9 of 10 centers with pyrosequencing (90%). Both centers applying Sanger sequencing (0%) did not pass the RRT. Blood samples that were sent for the confirmatory testing to Berlin arrived partly hemolytic but were analyzed correctly.
Only the results of massively parallel sequencing were included in the evaluation. All used techniques revealed basically the same results a
Magnetic bead-based DNA extraction Massively parallel sequencing
QIAsymphony, QIAGEN MiSeq, Illumina
Discussion
LDT laboratory-developed test
5a
4
3
2
6
Maxwell RSC, Promega PGM/Ion Torrent, Thermo Fisher Scientific Manually, QIAGEN Pyromark Q24 Rotorgene Q, QIAGEN Manually, QIAGEN PGM/Ion Torrent, Thermo Fisher Scientific Manually, QIAGEN PGM/Ion Torrent, Thermo Fisher Scientific QIAcube, QIAGEN PGM/Ion Torrent, Thermo Fisher Scientific, Pyromark Q24, Rotorgene Q, QIAGEN, Real-Time PCR Magnetic bead-based DNA extraction Massively parallel sequencing Column-based DNA extraction Pyrosequencing Real-time PCR Column-based DNA extraction Massively parallel sequencing Column-based DNA extraction Massively parallel sequencing Column-based DNA extraction Massively parallel sequencing Pyrosequencing Real-time PCR 1
Platform Extraction and DNA sequencing method Reference center
Table 3
Overview of DNA isolation systems and detection technology of the RRT phase I (blood samples)
Reagents
Virchows Arch
Tyrosine kinase inhibitors for the blockage of phosphorylation activity of epidermal growth factor receptor are the new standard of care for advanced EGFR mutation-positive NSCLC patients [32]. Despite the remarkable initial response to this targeted therapy [8, 10, 11, 17], most patients develop progressive disease during the first 12 to 24 months due to secondary resistance [11, 17, 18, 33]. T790M mutations in exon 20 of the EGFR gene were identified as the most frequent cause of this secondary resistance [11, 18, 34]. Thus, reliable and sensitive detection of various types of EGFR mutations is of utmost importance for adequate treatment decisions. To ensure the correct EGFR mutation testing at various places, external quality assessment by round robin trials is the method of choice. Herein, we describe the conditions and results of the first German EQA for T790M mutation testing in NSCLC tissue specimens and peripheral blood samples spiked with artificially generated ctDNA. A two-staged RRT (phase I pre-testing performed by 6 reference centers) and phase II (open RRT) testing scheme was initiated in September 2015 and finished in May 2016 with the successful completion of an open RRT trial. Thus, RRT was performed under the umbrella of QuIP, a national quality assurance program of the DGP (German Society of Pathology). In a first step, an internal RRT was performed at all six reference centers. Based on the correct and uniform identification T790M mutations in the selected tissue specimens and based on the correct detection of this mutation in the blood samples in five of the 6 reference centers, the open RRT was initiated. In total, 48 pathology centers from Germany (n = 46) and Switzerland (n = 2) participated in RTT. Of these, 47 submitted their analyses within the pre-defined time frame, and 44 (tissue-based) and 40 (blood-based) participants successfully performed the test. The overall success rates in the RRT phase II were 91.7% (tissue) and 83.3% (blood), respectively. Thirty-eight out of 48 participants (79.2%) successfully passed both parts of the RRT. This EQA was initiated prior to the conditional approval of the third-generation EGFR TKI osimertinib (AZD9291, Tagrisso®) for the treatment of adult patients with locally advanced or metastatic EGFR T790M mutation-positive non-small cell lung cancer by the EMA in February 2016
Virchows Arch Table 4
Overview of sequencing methods and platforms in the RRT phase II (tissue samples)
DNA sequencing method
Platform
Number
Pyrosequencing
13
Sanger sequencing PCR + reverse hybridization
– Pyromark/Therascreen Pyro Kit (n = 9) – Pyromark/in-house assay (n = 4) – Home-brew (n = 1) – Rotorgene/Therascreen (QIAGEN) (n = 3) – ABI7300/CAST PCR (Thermo Fisher Scientific) (n = 1) – Cobas/Cobas EGFR mutation test (Roche) (n = 4) – LightCycler/AmoyDx T790M (Roche/Zytomed) (n = 2) – Step-one/AmoyDx T790M (Thermo Fisher Scientific/Zytomed) (n = 1) – MiSeq/GeneRead (Illumina/QIAGEN) (n = 4) – MiSeq/TST15 or. custom (Illumina) (n = 1) – NextSeq/Neoplus (Illumina/Agilent) (n = 1) – Ion Torrent/Fragment Library or. Oncomine Kit (Thermo Fisher Scientific) (n = 3) ABI Genetic Analyzer/BigDye (Thermo Fisher Scientific) LDT
Melting point determination
N.a.
1
Droplet digital PCR (ddPCR)
BioRad
1
MALDI-TOF
Agena MASSarray/Agena Bioscience
1
Allele-specific PCR
Massively parallel sequencing
12
9
7 1
Not indicated
2
LDT laboratory-designed test, MALDI-TOF matrix-assisted laser desorption/ionization-time of flight, n.a. not available
The statements and recommendations for EGFR T790M mutation testing that can be given on the basis of our RRT results are:
[24]. With this approval, the reliable identification of NSCLC patients with EGFR T790M mutations became increasingly relevant. Since, as discussed above, tissue availability in NSCLC limits genotyping, particularly for patients with recurrence and metastasis; only 20–50% of the NSCLC patients could provide sufficient tissue material for molecular testing even in large and well-designed clinical trials [35]. Taking in consideration that tumors almost always release DNA into patient’s blood-circulation [36, 37], blood-based EGFR mutation testing (liquid biopsy) might become a reasonable alternative, or at least can complement genotyping done from biopsy tissue. It may thus expand the number of patients in whom valuable genetic information can be obtained, which may lead to optimized treatment stratification [12, 38–40].
&
&
The molecular testing for EGFR T790M mutations in biopsy tissue as well as in blood samples can be done, as recommended in the guidelines from the College of American Pathologists and the International Association for the Study of Lung Cancer [41], within an acceptable time frame of a 14-day turnaround time. As expected and already demonstrated in different studies [42], molecular testing for EGFR T790M mutation can be performed reliably on FFPE tissue. If tissue is available, it remains the gold standard.
Test center No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36* 37 38 39 40 41 42
Case No. /Sequencing system
1
2, 3
0
4
3
1, 3
2
4
4
4
2
1
4
4
3
1
5
4
3
6
4
4 3, 1 1
3
2
1
4
4
1
2
4
3
4
1
0
1 (T790M)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
2
2 (wild-type)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
0
2
3 (EGFRm)
2
1
1
2
2
2
2
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
4 (T790M)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
5 (EGFRm)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
6 (EGFRm)
2
2
2
2
2
2
2
2
2
2
2
1
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
7 (wild-type)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
8 (T790M)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
9 (T790M)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
10 (wild-type)
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
**
2
2
2
2
2
2
2
2
2
2
20
19
19
20
20
20
20
19
20
20
20
19
20
20
20
20
20
20
20
2, 1 1
3
1 7, 2 8
2
2
2
0
2
2
2
2
2
2
2
2
2
2
0
2
0
2
2
2
2
0
2
2
0
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
18/18 20 20 20 20 20 20 20 20 20 16 16 20 20 20 20
0
20 20 20 20 14 20
Sequencing system: 0, not stated; 1, allele-specific PCR; 2, Sanger-sequencing; 3, massive parallel sequencing; 4, pyrosequencing; 5, PCR and reverse hybridization; 6, matrix-assisted laser desorption/ ionization (MALDI)-TOF analysis; 7, melting point determination; 8, digital droplet PCR (ddPCR). * No result notification, ** Case number 10 did not contain enough tumor cells and was excluded from the result’s calculation, EGFRm = EGFR mutated case.
Fig. 3 Results of the RRT phase II (tissue testing). In the second row, the detection system used is encoded. In the following rows, numbers indicate the achieved points; two score points stand for the correct
identification of the T790M mutation or the wild-type status, respectively. In cases with one point technical problems emerged, no point was given in case of a wrong result
Virchows Arch Table 5
Overview of detection methods and platforms in the RRT phase II (blood samples)
DNA sequencing method
Platform
Number
Allele-specific PCR
19
Sanger sequencing PCR + reverse hybridization
- LDT (n = 3) - Rotorgene/Therascreen (QIAGEN) (n = 7) - ABI7300/CAST PCR (ThermoFisher) (n = 1), - COBAS/COBAS EGFR mutation test (Roche) (n = 3) - LightCycler/Therascreen (Roche/QIAGEN) (n = 1) - LightCycler/AmoyDx T790M (Roche/Zytomed) (n = 2) - Step-one/AmoyDx T790M (ThermoFisher/Zytomed) (n = 2) - MiSeq/GeneRead (Illumina/QIAGEN) (n = 3) - MiSeq/TST15 or custom (Illumina) (n = 4) - NextSeq/Neoplus (Illumina/Agilent) (n = 1) - Ion Torrent/Fragment Library or Oncomine Kit (ThermoFisher) (n = 5) - Pyromark/Therascreen Pyro Kit (n = 5) - Pyromark/in-house assay (n = 5) ABI Genetic Analyzer/BigDye (ThermoFisher) (n = 2) LDT
Melting point determination
N.a.
1
Droplet Digital PCR (ddPCR) MALDI-TOF
BioRad Agena MASSarray/Ageno Bioscience
1 1
Massively parallel sequencing
Pyrosequencing
13
10 2 1
Not indicated
2
LDT laboratory-developed test, MALDI-TOF matrix-assisted laser desorption/ionization-time of flight, n.a. not available
& &
&
manufacturer’s instructions of the blood stabilization tubes. Shipping of blood samples needs to be improved.
Since all methodologies used for tissue analysis yielded comparable results, all these methods can be equally recommended. With an overall success rate of 83.3%, the RRT supports the use of plasma testing to identify EGFR mutations and/ or monitor TKI treatment [35], particularly when tumor tissue is not available or inappropriate for molecular testing. Blood samples that were sent for the confirmatory testing arrived partly hemolytic although we followed the Test center No.
1
ctDNA extraction
2
3
4
5
0
A
B
A
Samples /Sequencing system
1
1, 3
0
4
6
7
8
9
10
11
12
13
A
C
D
A
B
A
A
B
3
1, 3
2
4
4
1,4
3
1
As with all studies, there are limitations of the present RRT that should be considered. The analysis of ctDNA is technically challenging, and the ctDNA extraction and sequencing methods described in the literature are—just as with our study (see Tables 2, 3, Figs. 3, 4)—numerous. Test assays actually used to detect treatment-altering NSCLC mutations vary widely in their sensitivity and specificity. Reck et al. showed
14
15
16
17
18
19
20
21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36* 37 38 39 40 41 42
A
A
A
B
D
D
A
B
B
4
3, 4
3
1
5
3
3
6
4 4, 1 1
A
B
A
A
D
4 3, 1 3
C
B
D
B
A
A
A
A
D
0
A
D
A
B
B
A
1 4, 1 3
1
1
1
3
4
1
0
1
1
3
1 7, 2 8
T790M/ 96 ng
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
T790M/ 10 ng
2
0
2
1
1
2
0
2
0
0
0
0
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
1
2
2
2
2
2
2
2
0
2
2
2
2
2
2 2
T790M/ 22 ng
2
0
2
2
2
2
0
2
2
2
0
0
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
2
T790M/ 45 ng
2
1
2
2
2
2
2
2
2
0
2
0
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
2
wild type
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
T790M/ 45 ng
2
0
2
2
2
2
2
2
2
2
0
0
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
2
T790M/ 10 ng
2
0
2
1
2
2
2
2
2
2
2
0
0
2
2
2
2
2
2
2
2
0
2
2
2
2
2
1
2
2
2
2
2
2
2
0
2
2
2
2
0
2 2
T790M/ 96 ng
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
T790M/ 22 ng
2
1
2
2
2
2
0
2
2
2
2
0
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
2
wild type
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
0
2
2
2
2
0
2
20
10
20
18
19
20
14
20
18
16
14
4
18
20
20
20
18
20
20
20
20 18 20 20 20 20 20 18 20 20 20 20 20 20 20
0
20 20 20 20
4
20
Platform: A, manual, QIAGEN; B, manual, others; C, automated, QIAGEN; D, automated, Maxwell. Sequencing system: 0, not stated; 1, allele-specific PCR; 2, Sanger sequencing; 3, Parallel sequencing; 4, pyrosequencing; 5 PCR and reverse hybridization; 6, matrix-assisted laser desorption/ ionization (MALDI)-TOF; 7, melting point determination; 8, digital droplet PCR (ddPCR). * No result notification.
Fig. 4 Results of the RRT phase II (blood samples). In the second and third rows, platforms and sequencing system used are encoded respectively. In the following rows, numbers indicate the achieved points. Two points stand for the correct identification of the T790M mutation or the wild-type status, respectively. In cases with one point technical problems emerge, no point was given in case of a wrong result.
Platform: A, manual, QIAGEN; B, manual, others; C, automated, QIAGEN; D, automated, Maxwell. Sequencing system: 0, not stated; 1, allele-specific PCR; 2, Sanger sequencing; 3, Parallel sequencing; 4, pyrosequencing; 5 PCR and reverse hybridization; 6, matrix-assisted laser desorption/ionization (MALDI)-TOF; 7, melting point determination; 8, digital droplet PCR (ddPCR). Asterisk indicates no result notification.
Virchows Arch
in a large European-Japanese diagnostic study for EGFR testing (ASSESS), which enrolled 1311 patients with advanced lung cancer that the positive and the negative predictive value of genotyping differs depending on which DNA tests were used [43]. But even in this study, in which the physicians were free to choose whichever genotyping method they wanted, the concordance of mutation status between matched biopsy and blood samples (n = 1162) was high (89%). However, the experience with this RRT taught us that the sensitivity of the Sanger sequencing might be insufficient for T790M mutation blood testing, especially when dealing with low allele frequency samples. Moreover, since there is no lower threshold for circulating tumor DNA and given that our sample collection only contained blood samples with ≥ 10 ng of the artificial ctDNA, we can neither confirm nor refute the possibility of false negative diagnosis with one of the methods used in the ring trial. Currently, there is no scientific consensus about the clinical importance of extremely low frequencies of the T790M mutation in tumor tissue. One participant found the T790M mutation in two supposed wild-type samples with allele frequencies of 0.1 and 0.5%, respectively. Crosschecking in the pre-phase of the RRT, the 9 participants from the same split in the open RRT, and a digital droplet PCR did not detect the T790M mutation in those samples. In this context, our findings indicate that there is a need for further studies, and standardization might be desirable in future T790M mutation detection methodologies. One possibility might be found in the parallel detection and evaluation of primary EGFR gene mutations as reference mutations as these are necessarily associated with T790M mutation. Against this background, a lower limit detection level for artificially designed RRT needs to be discussed. As described above, the rarity of clinically derived mutation-positive NSCLC tissue samples is not only a significant problem for clinical diagnosis but also for laboratory proficiency testing, certification, and quality control procedures. In order to enable all interested pathological institutes to participate in the RRT, some of the pre-selected FFPE tissue blocks had to be replaced after the second or third split due to insufficient tumor cell content. For the planning of future RRT, the collection of sufficient patient tissue specimens for T790M mutation testing is a challenge. For the future and to overcome this paucity of cancer tissue, it could be considered to generate artificial samples containing T790M mutations that would appropriately reflect human biopsy samples. Another conceivable alternative is the use of artificial FFPE cell blocks made from mutated cell lines after processing cultivated cell lines derived from NSCLC-like tissue samples to generate H&E slides. Alternatively, these cell lines might also be used to generate patient-derived xenografts (PDX) by transplanting standard NSCLC cell lines with EGFR mutations or wild type within an appropriate immunodeficient mouse model [44].
Conclusions In this EQA, we investigated and proved the multicenter reliability of an EGFR T790M mutation testing by RRT. Furthermore, the study shows that blood-based genotyping can successfully be implemented in the given and preexisting routines of a pathology center. Funding Phase I (internal round robin trial) of this study was funded by AstraZeneca. Compliance with ethical standards Conflict of interest Author JF has received speaker honoraria from AstraZeneca and Roche. Author AS has received speaker honoraria from AstraZeneca and is a member of the Diagnostic Content advisory board of AstraZeneca. VE received speaker honoraria from AstraZeneca and is a advisory board members of Astra Zeneca. Author AJ has received speaker honoraria from Amgen, AstraZeneca, Merck-Serono, Roche Pharmaceuticals and is a member of committee (advisory board): Amgen, AstraZeneca, Biocartis, Merck-Serono, Novartis. Author UL has received speaker honoraria from AstraZeneca and Roche. Author HK has received honorarium from Astra Zeneca as an advisory board member. Author PS received speaker honoraria from Astra Zeneca and is an advisory board member of AstraZeneca. Author TK has received speaker honoraria from Merck and AstraZeneca, is a member of AdBoard (Consulting): Amgen, AstraZeneca, Merck KGaA, MSD, Novartis, Pfizer, Roche and has received research grants from Merck and Roche. Author RB has received speaker honoraria from AstraZeneca, Boehringer Ingelheim, BMS, MSD, Pfizer, Qiagen and RocheVentana, has served in Scientific Advisory Boards from AstraZeneca, Boehringer Ingelheim, BMS, MSD, Pfizer, Qiagen and RocheVentana, and is a Cofounder and Chief Scientific Officer of Targos Molecular Pathology, Kassel, Germany. Author SMB has received speaker honoraria from AstraZeneca, Bristol Myers Squibb, Roche and Novartis and has received research grants from Astra Zeneca and Novartis. MI, DL, MH, CV, AL, RP, AV, SZ, GB, and MD declare that they have no conflict of interest. Informed consent The samples were collected with informed consent and under approved local ethical protocols from each patient.
References 1. 2.
3.
4.
5.
6.
Siegel R, Naishadham D, Jemal A (2013) Cancer statistics, 2013. CA Cancer J Clin 63:11–30 Lynch TJ, Bell DW, Sordella R et al (2004) Activating mutations in the epidermal growth factor receptor underlying responsiveness of nonsmall-cell lung cancer to gefitinib. N Engl J Med 350:2129–2139 Paez JG, Janne PA, Lee JC et al (2004) EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science (New York, NY) 304:1497–1500 Ellison G, Zhu G, Moulis A, Dearden S, Speake G, McCormack R (2013) EGFR mutation testing in lung cancer: a review of available methods and their use for analysis of tumour tissue and cytology samples. J Clin Pathol 66:79–89 Szumera-Cieckiewicz A, Olszewski WT, Tysarowski A et al (2013) EGFR mutation testing on cytological and histological samples in nonsmall cell lung cancer: a Polish, single institution study and systematic review of European incidence. Int J Clin Exp Pathol 6:2800–2812 Pao W, Miller V, Zakowski M et al (2004) EGF receptor gene mutations are common in lung cancers from Bnever smokers^ and
Virchows Arch
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23. 24.
are associated with sensitivity of tumors to gefitinib and erlotinib. Proc Natl Acad Sci U S A 101:13306–13311 Choughule A, Noronha V, Joshi A et al (2013) Epidermal growth factor receptor mutation subtypes and geographical distribution among Indian non-small cell lung cancer patients. Indian J Cancer 50:107–111 Mitsudomi T, Morita S, Yatabe Y et al (2010) 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 11:121–128 Park K, Tan EH, O'Byrne K et al (2016) Afatinib versus gefitinib as first-line treatment of patients with EGFR mutation-positive nonsmall-cell lung cancer (LUX-Lung 7): a phase 2B, open-label, randomised controlled trial. Lancet Oncol 17:577–589 D'Angelo SP, Pietanza MC, Johnson ML et al (2011) Incidence of EGFR exon 19 deletions and L858R in tumor specimens from men and cigarette smokers with lung adenocarcinomas. J Clin Oncol 29:2066–2070 Mok TS, Wu YL, Thongprasert S et al (2009) Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med 361:947–957 Fenizia F, De Luca A, Pasquale R et al (2015) EGFR mutations in lung cancer: from tissue testing to liquid biopsy. Future Oncol (London, England) 11:1611–1623 Chouaid C, Dujon C, Do P et al (2014) Feasibility and clinical impact of re-biopsy in advanced non small-cell lung cancer: a prospective multicenter study in a real-world setting (GFPC study 12– 01). Lung Cancer (Amsterdam, Netherlands) 86:170–173 Mellert H, Foreman T, Jackson L et al (2017) Development and clinical utility of a blood-based test service for the rapid identification of actionable mutations in non-small cell lung carcinoma. J Mol Diagn 19:404–416 Ionescu DN (2013) Impact of the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology clinical practice guidelines for EGFR and ALK testing in lung cancer in Canada. Curr Oncol (Toronto, Ont) 20:220–226 Oxnard GR, Thress KS, Alden RS et al (2016) Association between plasma genotyping and outcomes of treatment with osimertinib (AZD9291) in advanced non-small-cell lung cancer. J Clin Oncol 34:3375–3382 Zhou C, Wu YL, Chen G et al (2011) 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 12:735–742 Sequist LV, Waltman BA, Dias-Santagata D et al (2011) Genotypic and histological evolution of lung cancers acquiring resistance to EGFR inhibitors. Sci Transl Med 3:75ra26 Kobayashi S, Boggon TJ, Dayaram T et al (2005) EGFR mutation and resistance of non-small-cell lung cancer to gefitinib. N Engl J Med 352:786–792 Pao W, Miller VA, Politi KA et al (2005) Acquired resistance of lung adenocarcinomas to gefitinib or erlotinib is associated with a second mutation in the EGFR kinase domain. PLoS Med 2:e73 Yu HA, Arcila ME, Rekhtman N et al (2013) Analysis of tumor specimens at the time of acquired resistance to EGFR-TKI therapy in 155 patients with EGFR-mutant lung cancers. Clin Cancer Res 19:2240–2247 Balak MN, Gong Y, Riely GJ et al (2006) Novel D761Y and common secondary T790M mutations in epidermal growth factor receptor-mutant lung adenocarcinomas with acquired resistance to kinase inhibitors. Clin Cancer Res 12:6494–6501 FDA (2016) Osimertinib Journal. https://www.fda.gov/drugs/ informationondrugs/approveddrugs/ucm549683.htm EMA (2016) Osimertinib. Journal. http://www.ema.europa.eu/ema/ index.jsp?curl=pages/medicines/human/medicines/004124/ human_med_001961.jsp&mid=WC0b01ac058001d124
25.
26. 27.
28. 29.
30.
31.
32. 33.
34.
35.
36.
37. 38.
39.
40.
41.
42.
43.
44.
Cross DA, Ashton SE, Ghiorghiu S et al (2014) AZD9291, an irreversible EGFR TKI, overcomes T790M-mediated resistance to EGFR inhibitors in lung cancer. Cancer Discov 4:1046–1061 Janne PA, Yang JC, Kim DWet al (2015) AZD9291 in EGFR inhibitorresistant non-small-cell lung cancer. N Engl J Med 372:1689–1699 Ramalingam S, Yang JC, Lee CK et al (2016) LBA1_PR: osimertinib as first-line treatment for EGFR mutation-positive advanced NSCLC: updated efficacy and safety results from two phase I expansion cohorts. J Thorac Oncol 11:S152 Kurman RJCM, Herrington CS, Young RH (2014) WHO classification of tumours, Volume 6, WHO Press ISBN-13 9789283224358 Travis WD, Brambilla E, Nicholson AG et al (2015) The 2015 World Health Organization classification of lung tumors: impact of genetic, clinical and radiologic advances since the 2004 classification. J Thorac Oncol 10:1243–1260 Heydt C, Fassunke J, Kunstlinger H et al (2014) Comparison of preanalytical FFPE sample preparation methods and their impact on massively parallel sequencing in routine diagnostics. PLoS One 9:e104566 Konig K, Peifer M, Fassunke J et al (2015) Implementation of amplicon parallel sequencing leads to improvement of diagnosis and therapy of lung cancer patients. J Thorac Oncol 10:1049–1057 Landi L, Cappuzzo F (2014) Pharmacotherapy targeting the EGFR oncogene in NSCLC. Expert Opin Pharmacother 15:2293–2305 Gil-Bazo I, Rolfo C (2016) AZD9291 in TKI EGFR resistance in non-small cell lung cancer and the new concept of phase I trials. Transl Lung Cancer Res 5:85–88 Stewart EL, Tan SZ, Liu G, Tsao MS (2015) Known and putative mechanisms of resistance to EGFR targeted therapies in NSCLC patients with EGFR mutations—a review. Transl Lung Cancer Res 4:67–81 Sun W, Yuan X, Tian Y et al (2015) Non-invasive approaches to monitor EGFR-TKI treatment in non-small-cell lung cancer. J Hematol Oncol 8:95 Mack PC, Holland WS, Burich RA et al (2009) EGFR mutations detected in plasma are associated with patient outcomes in erlotinib plus docetaxeltreated non-small cell lung cancer. J Thorac Oncol 4:1466–1472 Tsang JC, Lo YM (2007) Circulating nucleic acids in plasma/serum. Pathology 39:197–207 Murtaza M, Dawson SJ, Tsui DW et al (2013) Non-invasive analysis of acquired resistance to cancer therapy by sequencing of plasma DNA. Nature 497:108–112 Sholl LM, Aisner DL, Allen TC et al (2016) Liquid biopsy in lung cancer: a perspective from members of the pulmonary pathology society. Arch Pathol Lab Med 140:825–829 Yanagita M, Redig AJ, Paweletz CP et al (2016) A prospective evaluation of circulating tumor cells and cell-free DNA in EGFRmutant non-small cell lung cancer patients treated with erlotinib on a phase II trial. Clin Cancer Res 22:6010–6020 Lindeman NI, Cagle PT, Beasley MB et al (2013) Molecular testing guideline for selection of lung cancer patients for EGFR and ALK tyrosine kinase inhibitors: guideline from the College of American Pathologists, International Association for the Study of Lung Cancer, and Association for Molecular Pathology. Arch Pathol Lab Med 137:828–860 Naidoo J M, Drilon A (2014) Molecular diagnostic testing in non-small cell lung cancer. Am J Hematol Oncol. Available on http://www.gotoper.com/publications/ajho/2014/2014sep/ molecular-diagnostic-testing-in-non-small-cell-lung-cancer. Last accessed October 20, 2016 Reck M, Hagiwara K, Han B et al (2016) ctDNA determination of EGFR mutation status in European and Japanese patients with advanced NSCLC: the ASSESS study. J Thorac Oncol 11:1682–1689 Zhang XC, Zhang J, Li M et al (2013) Establishment of patientderived non-small cell lung cancer xenograft models with genetic aberrations within EGFR, KRAS and FGFR1: useful tools for preclinical studies of targeted therapies. J Transl Med 11:168