Validity of Targeted Next-Generation Sequencing in Routine Care for Identifying Clinically Relevant Molecular Profiles in Non–Small-Cell Lung Cancer
2018; Elsevier BV; Volume: 20; Issue: 4 Linguagem: Inglês
10.1016/j.jmoldx.2018.04.002
ISSN1943-7811
AutoresAntoine Legras, Marc Barritault, Anne Tallet, Elizabeth Fabre, Alice Guyard, Bastien Rance, William Digan, Nicolas Pécuchet, Étienne Giroux-Leprieur, Catherine Julié, Stéphane Jouveshomme, Véronique Duchatelle, Véronique Giraudet, Laure Gibault, Alain Cazier, Jean Pastré, Françoise Le Pimpec‐Barthes, Pierre Laurent‐Puig, Hélène Blons,
Tópico(s)RNA modifications and cancer
ResumoTheranostic assays are based on single-gene testing, but the ability of next-generation sequencing (NGS) to interrogate numerous genetic alterations will progressively replace single-gene assays. Although NGS was evaluated to screen for theranostic mutations, its usefulness in clinical practice on large series of samples remains to be demonstrated. NGS performance was assessed following guidelines. TaqMan probes and NGS were compared for their ability to detect EGFR and KRAS mutations, and NGS mutation profiles were analyzed on a large series of non–small-cell lung cancers (n = 1343). The R2 correlation between expected and measured allelic ratio, using commercial samples, was >0.96. Mutation detection threshold was 2% for 10 ng of DNA input. κ Scores for TaqMan versus NGS were 0.99 (95% CI, 0.97–1.00) for EGFR and 0.98 (95% CI, 0.97–1.00) for KRAS after exclusion of rare EGFR (n = 40) and KRAS (n = 60) mutations. NGS identified 693 and 292 mutations in validated and potential oncogenic drivers, respectively. Significant associations were found between EGFR and PI3KCA or CTNNB1 and between KRAS and STK11. Potential oncogenic driver mutations or gene amplifications were more frequent in validated oncogenic driver nonmutated samples. This work is a proof of concept that targeted NGS is accessible in routine screening, including large screening, at reasonable cost. Clinical data should be collected and implemented in specific databases to make molecular data meaningful for direct patients' benefit. Theranostic assays are based on single-gene testing, but the ability of next-generation sequencing (NGS) to interrogate numerous genetic alterations will progressively replace single-gene assays. Although NGS was evaluated to screen for theranostic mutations, its usefulness in clinical practice on large series of samples remains to be demonstrated. NGS performance was assessed following guidelines. TaqMan probes and NGS were compared for their ability to detect EGFR and KRAS mutations, and NGS mutation profiles were analyzed on a large series of non–small-cell lung cancers (n = 1343). The R2 correlation between expected and measured allelic ratio, using commercial samples, was >0.96. Mutation detection threshold was 2% for 10 ng of DNA input. κ Scores for TaqMan versus NGS were 0.99 (95% CI, 0.97–1.00) for EGFR and 0.98 (95% CI, 0.97–1.00) for KRAS after exclusion of rare EGFR (n = 40) and KRAS (n = 60) mutations. NGS identified 693 and 292 mutations in validated and potential oncogenic drivers, respectively. Significant associations were found between EGFR and PI3KCA or CTNNB1 and between KRAS and STK11. Potential oncogenic driver mutations or gene amplifications were more frequent in validated oncogenic driver nonmutated samples. This work is a proof of concept that targeted NGS is accessible in routine screening, including large screening, at reasonable cost. Clinical data should be collected and implemented in specific databases to make molecular data meaningful for direct patients' benefit. Non–small-cell lung cancer (NSCLC) remains an important cause of cancer-related death worldwide. Current guidelines for treatment decisions1Socinski M.A. Evans T. Gettinger S. Hensing T.A. VanDam Sequist L. Ireland B. Stinchcombe T.E. Treatment of stage IV non-small cell lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines.Chest. 2013; 143: e341S-e368SAbstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar rely on tumor stage, patient's performance status, histology, and, for advanced stage patients, the presence of tumor genetic alterations. Several molecular events driving lung carcinogenesis have been identified and can be successfully targeted by different tyrosine kinase inhibitors (TKIs).2Lynch T.J. Bell D.W. Sordella R. Gurubhagavatula S. Okimoto R.A. Brannigan B.W. Harris P.L. Haserlat S.M. Supko J.G. Haluska F.G. Louis D.N. Christiani D.C. Settleman J. Haber D.A. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib.N Engl J Med. 2004; 350: 2129-2139Crossref PubMed Scopus (10020) Google Scholar, 3Paez J.G. Jänne P.A. Lee J.C. Tracy S. Greulich H. Gabriel S. Herman P. Kaye F.J. Lindeman N. Boggon T.J. Naoki K. Sasaki H. Fujii Y. Eck M.J. Sellers W.R. Johnson B.E. Meyerson M. EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy.Science. 2004; 304: 1497-1500Crossref PubMed Scopus (8482) Google Scholar In that context, tumor genetic testing has become standard of care to select efficient treatments in patients with NSCLC. In many cases, targeted drugs received approval in molecularly defined subgroups of patients,4Mazza V. Cappuzzo F. Treating EGFR mutation resistance in non-small cell lung cancer: role of osimertinib.Appl Clin Genet. 2017; 10: 49-56Crossref PubMed Scopus (22) Google Scholar and targeted therapies have either become a standard for patients with epidermal growth factor receptor (EGFR) mutated or ALK and ROS1 rearranged tumors5Solomon B. Varella-Garcia M. Camidge D.R. ALK gene rearrangements: a new therapeutic target in a molecularly defined subset of non-small cell lung cancer.J Thorac Oncol. 2009; 4: 1450-1454Abstract Full Text Full Text PDF PubMed Scopus (263) Google Scholar, 6Scheffler M. Schultheis A. Teixido C. Michels S. Morales-Espinosa D. Viteri S. Hartmann W. Merkelbach-Bruse S. Fischer R. Schildhaus H.-U. Fassunke J. Sebastian M. Serke M. Kaminsky B. Randerath W. Gerigk U. Ko Y.-D. Krüger S. Schnell R. Rothe A. Kropf-Sanchen C. Heukamp L. Rosell R. Büttner R. Wolf J. ROS1 rearrangements in lung adenocarcinoma: prognostic impact, therapeutic options and genetic variability.Oncotarget. 2015; 6: 10577-10585Crossref PubMed Scopus (72) Google Scholar, 7Mazières J. Zalcman G. Crinò L. Biondani P. Barlesi F. Filleron T. Dingemans A.-M.C. Léna H. Monnet I. Rothschild S.I. Cappuzzo F. Besse B. Thiberville L. Rouvière D. Dziadziuszko R. Smit E.F. Wolf J. Spirig C. Pecuchet N. Leenders F. Heuckmann J.M. Diebold J. Milia J.D. Thomas R.K. Gautschi O. Crizotinib therapy for advanced lung adenocarcinoma and a ROS1 rearrangement: results from the EUROS1 cohort.J Clin Oncol. 2015; 33: 992-999Crossref PubMed Scopus (291) Google Scholar or a therapeutic option for patients with BRAF, MET, NTRK, or RET alterations.8Lee G.D. Lee S.E. Oh D.-Y. Yu D.-B. Jeong H.M. Kim J. Hong S. Jung H.S. Oh E. Song J.-Y. Lee M.-S. Kim M. Jung K. Kim J. Shin Y.K. Choi Y.-L. Kim H.R. MET exon 14 skipping mutations in lung adenocarcinoma: clinicopathologic implications and prognostic values.J Thorac Oncol. 2017; 12: 1233-1246Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar, 9Baik C.S. Myall N.J. Wakelee H.A. Targeting BRAF-mutant non-small cell lung cancer: from molecular profiling to rationally designed therapy.Oncologist. 2017; 22: 786-796Crossref PubMed Scopus (78) Google Scholar, 10Song Z. Yu X. Zhang Y. Clinicopathologic characteristics, genetic variability and therapeutic options of RET rearrangements patients in lung adenocarcinoma.Lung Cancer. 2016; 101: 16-21Abstract Full Text Full Text PDF PubMed Scopus (29) Google Scholar, 11Vaishnavi A. Capelletti M. Le A.T. Kako S. Butaney M. Ercan D. Mahale S. Davies K.D. Aisner D.L. Pilling A.B. Berge E.M. Kim J. Sasaki H. Park S. Kryukov G. Garraway L.A. Hammerman P.S. Haas J. Andrews S.W. Lipson D. Stephens P.J. Miller V.A. Varella-Garcia M. Jänne P.A. Doebele R.C. Oncogenic and drug-sensitive NTRK1 rearrangements in lung cancer.Nat Med. 2013; 19: 1469-1472Crossref PubMed Scopus (464) Google Scholar Because the number of predictive markers entering clinical practice is rapidly growing and because molecular screening has to be done using small biopsy specimens in a limited time, the implementation of next-generation sequencing (NGS) is necessary. Companion tests linked to one drug will soon be replaced by NGS-based companion tests that offer a more comprehensive analysis of molecular alterations. Indeed, NGS is quickly entering clinical practice because of technical improvements. It is suitable for formalin-fixed, paraffin-embedded (FFPE) samples and low quantities of DNA within the nanogram range. In the context of somatic mutation screening in cancer, NGS provides advantages over single-gene approaches, such as tissue saving, which is important for small biopsy or cytology samples. Moreover, NGS is a quantitative and sensitive technology. With mutant allele detection cutoff around 2%, it bypasses detection problems attributable to low percentage of tumor cells and allows the identification of tumor subclones. Finally, bioinformatics treatment of NGS coverage data can point out gene amplifications.12McCourt C.M. McArt D.G. Mills K. Catherwood M.A. Maxwell P. Waugh D.J. Hamilton P. O'Sullivan J.M. Salto-Tellez M. Validation of next generation sequencing technologies in comparison to current diagnostic gold standards for BRAF, EGFR and KRAS mutational analysis.PLoS One. 2013; 8: e69604Crossref PubMed Scopus (97) Google Scholar, 13Frampton G.M. Fichtenholtz A. Otto G.A. Wang K. Downing S.R. He J. et al.Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing.Nat Biotechnol. 2013; 31: 1023-1031Crossref PubMed Scopus (1450) Google Scholar, 14Pant S. Weiner R. Marton M.J. Navigating the rapids: the development of regulated next-generation sequencing-based clinical trial assays and companion diagnostics.Front Oncol. 2014; 4: 78Crossref PubMed Scopus (63) Google Scholar, 15Deeb K.K. Hohman C.M. Risch N.F. Metzger D.J. Starostik P. Routine clinical mutation profiling of non-small cell lung cancer using next-generation sequencing.Arch Pathol Lab Med. 2015; 139: 913-921Crossref PubMed Scopus (24) Google Scholar Altogether, NGS allows for a more comprehensive screening of tumors; however, its implementation in a routine workflow remains challenging in many ways. The heterogeneity of routine samples in terms of quality, percentage of tumor cells, and size and type of samples may hamper proper identification of mutations, and the clinical implication of rare molecular alterations detected by high-throughput sequencing may be difficult to evaluate. In France, the National Cancer Institute (INCa) developed and supported, since 2009, a large national program to set up molecular screening for patients with advanced tumors. This was done to ensure equal access to innovative treatment to all patients in the country. Therefore, all lung cancer patients are being offered free molecular testing with a rapid turnaround testing time for validated biomarkers. The implementation of additional tests was discussed with the development of INCa guidelines on emerging biomarkers and trials, such as the AcSé program (INCa), that give access to potentially useful drugs outside their approved indications.16Buzyn A. Blay J.-Y. Hoog-Labouret N. Jimenez M. Nowak F. Deley M.-C.L. Pérol D. Cailliot C. Raynaud J. Vassal G. Equal access to innovative therapies and precision cancer care.Nat Rev Clin Oncol. 2016; 13: 385-393Crossref PubMed Scopus (37) Google Scholar Efforts are underway in all of the INCa laboratories to overcome many of the challenges to multibiomarker, including the development of high-throughput molecular screenings as routine tools. In the context of a growing number of markers, the INCa supported the implementation of NGS with the same objectives as for EGFR or ALK testing: all patients, equal access, and quality control assessment. Herein, we report our experience of validation and application to the first 1343 tests of targeted NGS testing for the diagnostic of druggable alterations in NSCLC. This study is the first proof of concept that equal access to precision cancer care is feasible and cost effective compared with any companion test. Targeted NGS validation was performed in three steps. First, method performance was tested using reference DNA samples (Horizon Discovery, Waterbeach, UK) to investigate repeatability, reproducibility, robustness, and detection threshold, as recommended.17Bennett N.C. Farah C.S. Next-generation sequencing in clinical oncology: next steps towards clinical validation.Cancers (Basel). 2014; 6: 2296-2312Crossref PubMed Scopus (43) Google Scholar Second, NGS was compared with competitive allele-specific TaqMan probes for common EGFR mutations and with TaqMan assays for KRAS mutations. Finally, we report the application of NGS in a series of 1343 NSCLC samples corresponding to 2 years of experience in routine molecular testing. We showed that targeted NGS demonstrated its robustness and efficiency to assess molecular alterations in NSCLC in a clinical setting and discussed strength and drawbacks focusing on clinical implications. Reproducibility and repeatability tests were performed on the basis of the INCa guidelines.18Jennings L.J. Arcila M.E. Corless C. Kamel-Reid S. Lubin I.M. Pfeifer J. Temple-Smolkin R.L. Voelkerding K.V. Nikiforova M.N. Guidelines for validation of next-generation sequencing-based oncology panels: a joint consensus recommendation of the Association for Molecular Pathology and College of American Pathologists.J Mol Diagn. 2017; 19: 341-365Abstract Full Text Full Text PDF PubMed Scopus (378) Google Scholar Commercial FFPE samples (HD200 and HD301) and genomic DNAs (HD701 and HD729) with validated allelic ratio for various mutations (Horizon Discovery) were used. Triplicates within the same run and triplicates from three different experiments were used to assess repeatability and reproducibility. Allelic ratios were defined as the end points. To assess robustness, FFPE results were compared with genomic DNA, and the potential impact of low DNA inputs was analyzed on results using range of DNA templates. Finally, detection cutoff was estimated after serial dilutions (½, ¼, ⅛, and 1/16) of HD301 and HD802 nontumor genomic DNAs. Beyond initial performance testing, HD samples were used as positive internal controls in all sequencing runs. A total of 1343 consecutive samples of NSCLC were addressed to our laboratory for molecular diagnosis (following current guidelines,1Socinski M.A. Evans T. Gettinger S. Hensing T.A. VanDam Sequist L. Ireland B. Stinchcombe T.E. Treatment of stage IV non-small cell lung cancer: diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines.Chest. 2013; 143: e341S-e368SAbstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar cIIIB or cIV NSCLC with treatment intent) from July 2015 to July 2017. Molecular data were prospectively registered in the laboratory record informatics system that provides result reporting to clinicians [French National Commission for Informatics and Liberties (CNIL) declaration number 1922081 v 0]. A pathologist (C.J., V.D., L.G., or A.C.) qualified tumor sampling and quantified the tumor cell content. For samples with G: EGFR_6224_mu Hs00000102_mu; 19 deletions in exon 19: EGFR_ex19dels_mu Hs00000228_mu; EGFR_6240mu; and p.Thr790Met c.2369C>T Hs00000102_mu and reference assay: EGFR_rf Hs00000173_rf) and TaqMan probes for KRAS (p.Gly12Ser/Cys/Asp/Ala/Val and p.Gly13Asp) (Thermo Fisher Scientific) (Table 1).Table 1In-House TaqMan Probe Design for KRAS (p.Gly12Ser/Cys/Asp/Ala/Val, p.Gly13Asp)AssayForward primer sequenceReverse primer sequenceWild-type probe (FAM)Mutant probe (VIC)KRAS p.Gly12Asp c.35G>A5′-AGGCCTGCTGAAAAT-GACTGAATAT-3′5′-GCTGTATCGTCAA-GGCACTCTT-3′TTGGAGCTGGTGGCGTTGGAGCTGATGGCGTKRAS p.Gly12Cys c.34G>T5′-AGGCCTGCTGAAAAT-GACTGAATAT-3′5′-GCTGTATCGTCAA-GGCACTCTT-3′CCTACGCCACCAGCTCTACGCCACAAGCTKRAS p.Gly12Ala c.35G>C5′-AGGCCTGCTGAAAAT-GACTGAATAT-3′5′-GCTGTATCGTCAA-GGCACTCTT-3′CCTACGCCACCAGCTCTACGCCAGCAGCTKRAS p.Gly12Ser c.34G>A5′-AGGCCTGCTGAAAAT-GACTGAATAT-3′5′-GCTGTATCGTCAA-GGCACTCTT-3′CTACGCCACCAGCTCCTACGCCACTAGCTCKRAS p.Gly12Val c.35G>T5′-AGGCCTGCTGAAAAT-GACTGAATAT-3′5′-GCTGTATCGTCAA-GGCACTCTT-3′CTACGCCACCAGCTCACGCCAACAGCTCKRAS p.Gly13Asp c.38G>A5′-AGGCCTGCTGAAAAT-GACTGAATAT-3′5′-GCTGTATCGTCAA-GGCACTCT-3′TGGTGGCGTAGGCACTGGTGACGTAGGCAThe probe was from Thermo Fisher Scientific. Open table in a new tab The probe was from Thermo Fisher Scientific. Each real-time quantitative PCR, one per probe, was run in a final volume of 5 μL in 384-well plates, including 2.5 μL of 2× TaqMan genotyping master mix (Applied Biosystems, Foster City, CA), 0.5 μL of 10× Assay Mix, 1 μL of deionized water, and 10 ng of DNA template. Samples were run in duplicate on an ABI Prism 7900 HT sequence detection system (Applied Biosystems) using standard thermocycling conditions and analyzed with the SDS software version 2.4 (Applied Biosystems). NGS was performed with a dedicated panel of 92 amplicons (Ion AmpliSeq Colon-Lung Cancer Research Panel version 2; Life Technologies, Carlsbad, CA), covering >500 hotspot mutations in KRAS, EGFR, BRAF, PIK3CA, AKT1, ERBB2, PTEN, NRAS, STK11, MAP2K1, ALK, DDR2, CTNNB1, MET, TP53, SMAD4, FBXW7, fibroblast growth factor receptor 3 (FGFR3), NOTCH1, ERBB4, FGFR1, and FGFR2 (Supplemental Table S1). Multiplex PCR libraries were prepared using 30 ng of DNA whenever possible and 3 μL of DNA for samples with DNA concentration <10 ng/μL by AmpliSeq technology (Ion AmpliSeq library kit version 2, Ion library equalizer kit; Life Technologies). Clonal amplification and sequencing were done on the Ion Chef System (Ion PI Hi-Q Chef, Ion PI Chip Kit v3) and Ion Torrent Proton sequencer (Life Technologies). Data were analyzed by the Torrent Suite 4.4.3 and 5.0.4 (Life Technologies) using optimized parameters: detection threshold of 2% for single-nucleotide variants, 5% for insertions/deletions, and 1% for hotspots (Supplemental Table S2). Variant call files from the variant caller were loaded on a galaxy platform19Digan W. Countouris H. Barritault M. Baudoin D. Laurent-Puig P. Blons H. Burgun A. Rance B. An architecture for genomics analysis in a clinical setting using Galaxy and Docker.Gigascience. 2017; 6: 1-9Crossref PubMed Scopus (6) Google Scholar and annotated using the Safir2report tool (https://github.com/OvoiDs/IonTorrentReport). Quality criteria used as end points were a detection threshold of 2% and a minimum coverage depth of 300×. The panel used herein is the first in-line routine panel used in our laboratory. In this panel, MET exon 14 is only partially analyzed. In patients with EGFR and KRAS wild tumors, subsequent screenings include MET (exon 14 and TK domain), ALK (TK domain), and ROS1 (TK domain) analysis (AcSé panel) and fluorescence in situ hybridization for ROS1. Moreover, ALK immunohistochemistry was performed systematically for all samples coming from our institution (n = 624), using ALK p80 primary antibody (Mob416; Diagnostic BioSystems, Pleasanton, CA) used at 1:40 dilution on the Ventana BenchMark ULTRA IHC/ISH Staining Module (Ventana Medical Systems, Tucson, AZ). Concerning samples from other centers, ALK immunohistochemistry was not performed in our laboratory. Mutated genes were classified in different groups to facilitate NGS interpretation: validated oncogenic drivers (VODs; EGFR, KRAS, BRAF p.Val600 or p.Lys601, NRAS, and MET exon 14) and potential oncogenic drivers (PODs; STK11, PIK3CA, non-p.Val600 or non-p.Lys601 BRAF, PTEN, ERBB2, MAP2K1, ERBB4, DDR2, FGFR1, FGFR2, FGFR3, non–exon 14 MET, ALK, and AKT1). Other mutated genes (TP53, CTNNB1, SMAD4, FBXW7, and NOTCH1) were labeled nontargetable driver. In samples with variants of unknown significance in VOD genes, it might have been more accurate to classify alterations on the basis of the type of variant. However, this situation was rare, and we found it clearer to always report genes the same way on molecular reports with an adapted conclusion for variants of unknown significance. NGS coverage depth data were used to identify gene amplifications using an algorithm developed in our laboratory on the basis of the identification of outliers from the expected coverage means + 3 SDs, calculated using the entire run data (Supplemental Code S1 and Supplemental Table S3). This method was validated for EGFR, ERBB2, MET, and KRAS amplification using real-time quantitative PCR technology (Hs04960197_cn, Hs00817646_cn, Hs00774321_cn, and Hs06962121_cn, respectively; Applied Biosystems) (n = 23, n = 10, n = 14, and n = 16, respectively). Single-gene assays' costs were assessed in a previous publication.20Blons H. Rouleau E. Charrier N. Chatellier G. Côté J.-F. Pages J.-C. de Fraipont F. Boyer J.-C. Merlio J.P. Morel A. Gorisse M.-C. de Cremoux P. Leroy K. Milano G. Ouafik L. Merlin J.-L. Le Corre D. Aucouturier P. Sabourin J.-C. Nowak F. Frebourg T. Emile J.-F. Durand-Zaleski I. Laurent-Puig P. MOKAECM Collaborative GroupPerformance and cost efficiency of KRAS mutation testing for metastatic colorectal cancer in routine diagnosis: the MOKAECM study, a nationwide experience.PLoS One. 2013; 8: e68945Crossref PubMed Scopus (21) Google Scholar Concerning NGS, reagent costs were US$89 per sample when using a single primer-pool library and a full loading of 96 samples on Ion PI Proton chips. Turnover time was 2.5 days of technical work; data interpretation and reports take another 3 days, including 1 day of biological validation. Pearson's correlation coefficients were calculated to compare means. Comparisons of categorical variables were calculated using the Wilcoxon test. A two-tailed P < 0.05 was taken to indicate statistical significance. Statistical analyses were performed using the JMP software version 10.0 (SAS Institute JMP, Brie Comte Robert, France). The analytic workflow is depicted in Figure 1. Concerning repeatability, the R2 assessed correlations between expected and measured allelic ratio (Supplemental Table S4 and Supplemental Figure S1) and was 0.96, 0.991, 0.97, and 1.00 for HD701, HD729, HD200, and HD301, respectively. Concerning reproducibility, the R2 values were 0.99, 0.99, 0.98, and 1.00 for HD701, HD729, HD200, and HD301, respectively (Supplemental Table S5 and Supplemental Figure S2). The detection threshold was 2% down to 10 ng of DNA input. Robustness analyses showed that mutations with allelic ratio of 5% were constantly called with 10, 5, or 1 ng of DNA input. The detection threshold after dilution in genomic DNA was affected by low DNA inputs. Indeed, different mutations with an expected 2% ratio were not called (Supplemental Table S6). Coverage histograms for HD701, HD729, HD200, and HD301 are shown in Supplemental Figure S3. Amplicon coverage for EGFR, KRAS, ERBB2, and BRAF was constantly >300×. Four amplicons (ERBB4 exon 3, PTEN exon 3, STK11 exon 6, and FBXW7 exon 7) recurrently had a lower coverage. A total of 1343 consecutive NSCLC samples were analyzed. Pathology characteristics are reported in Table 2.Table 2Description of the Population AnalyzedParametern%Histologic type Adenocarcinomas79459 Squamous cell564 Sarcomatoid100.7 Large cell60.5 Adenosquamous20.1 Small cell40.3 Undifferentiated16913 Unknown30222% of tumor cells 070.5 5048636 Unknown27020Samples with 0% of tumor cells were noncontributory. Open table in a new tab Samples with 0% of tumor cells were noncontributory. The tumor cell percentage per sample is reported in Table 2. Mean DNA concentration was 51.7 ng/μL (95% CI, 45.7–57.7) and was significantly associated with tumor cell percentage (P < 0.001). Small samples may be difficult to macrodissect, leading to an association between low DNA concentration and low tumor cell content. Low DNA concentration was related to NGS failure (P < 0.001). FFPE samples are prone to sequencing artifacts attributable to formalin fixation DNA damage, and recurrent artifacts attributable to primer mispriming may occur in low-quality samples. Therefore, the global number of variants detected by the pipeline (all germline and somatic calls) was used as a quality criterion. Mean number of variants per sample was 10.6 (95% CI, 9.1–12). More than 30 variants were found in 6% of samples. These showed either abnormal levels of C>T/G>A changes in the 1% to 10% allele frequency range or mispriming alterations with a mean allelic ratio of 8.4% (range, 3 to 100) (Supplemental Table S7) that were reported as artifacts. Thus, NGS results were totally (low coverage depth and/or excess of C>T/G>A and/or mispriming alterations) or partially (excess of C>T/G>A and/or mispriming alterations) not interpretable in 75 (5.5%) and 26 (2.0%) of cases, respectively. Among the 1343 consecutive samples, 64 were only assessed by NGS. Therefore, data comparisons are based on 1279 samples (Figure 2). TaqMan results were not contributory for 67 cases (5.3%). Among them, 22 (33%) were correctly sequenced using NGS, 4 (6%) led to partial NGS results, and 41 (61%) remained not contributory. NGS results were not contributory for 75 cases (5.5%). Among them, 41 (55%) were not contributory using TaqMan assays. Considering global results, TaqMan assays identified 190 EGFR mutations versus 230 using NGS (κ score, 0.77; 95% CI, 0.73–0.81) and 302 KRAS mutations versus 364 using NGS (κ score, 0.77; 95% CI, 0.73–0.80) (Table 3). Among the EGFR mutations exclusively detected by NGS, there were 40 rare EGFR mutations that were not screened by TaqMan probes and 4 real false-negative mutations: 1 EGFR p.Leu858Arg and 3 EGFR Del19, including 2 complex deletions. The corrected κ score was 0.85 (95% CI, 0.81–0.88) after exclusion of the 40 rare EGFR mutations and 0.99 (95% CI, 0.97–1.00) after exclusion of noncontributory results. In two cases, a false-positive Del19 was identified using the EGFR_ex19dels_mu TaqMan assay. NGS reported that the genomic alteration was a multiple nucleotide variant (c.2239_2240delinsCC) leading to a p.Leu747Pro mutation. Among the KRAS mutations missed by TaqMan probes, 58 were rare KRAS mutations and 9 were real false-negative mutations (Supplemental Table S8). All KRAS mutations detected with TaqMan probes were also detected by NGS. The corrected κ score was 0.86 (95% CI, 0.83–0.89) after exclusion of the 58 undetectable rare KRAS mutations and 0.98 (95% CI, 0.97–1.00) after exclusion of noncontributory results.Table 3Concordance Tables between TaqMan Probes and NGS Results, for EGFR p.Leu858Arg and Del19, p.Thr790Met, and KRAS MutationsProbeNGSTotalWild typeMutatedNIEGFR TaqManWild type, n94944301023Mutated, n01828190NI, n2144166Total, N970230791279EGFR T790M TaqManWild type, n4228Mutated, n311014Total, N713222KRAS TaqManWild type, n8066729902Mutated, n02939302NI, n3044175Total, N836364791279NGS, next-generation sequencing; NI, not interpretable. Open table in a new tab NGS, next-generation sequencing; NI, not interpretable. When relapse after first line of TKI was mentioned, samples were screened for the EGFR p.Thr790Met secondary resistance mutation using the corresponding TaqMan assay. A p.Thr790Met mutation was found in 14 (64%) of the 22 cases. In this subset of samples, NGS identified the p.Thr790Met mutation in 13 cases. However, the κ score was 0.43 (95% CI, 0.00–0.86) because of five mismatches (Table 3). A total of 1268 NSCLC samples were included in the analysis (1343 consecutive samples, 75 do not have contributory results). A total of 1649 mutations were identified in 1009 cases (Figure 3). Distribution and association of mutations between VOD and POD are reported in Tables 4 and 5.Table 4Description and Association of Mutations in Validated Oncogenic Driver GenesValidated oncogenic driver genesn%EGFR23617.6 Second423.1 Third30.2KRAS36827.4 EGFR associated∗Del19, p.Val765Met, and p.Trp477Cys.30.1 KRAS (second)30.2 BRAF associated10.4 NRAS associated10.1BRAF (p.Val600 or p.Lys601)141.0NRAS161.2MET exon 1410.1Not interpretable756.0No mutation63346.8Total Cases with one mutation58043.2 Cases with two mutations523.9 Cases with three mutations30.2 Cases with at least one mutation63547.3∗ Del19, p.Val765Met, and p.Trp477Cys. Open table in a new tab Table 5Description and Association of Mutations in Potential Oncogenic Driver GenesPotential oncogenic driver genesn%STK111027.6 FGFR3 associated20.1 ERBB4 associated10.1 DDR2 associated10.1 ALK associated10.1PIK3CA523.9 PTEN associated20.1 FGFR3 associated10.1 ERBB4 associated10.1 DDR2 associated20.1 ALK associated10.1BRAF (except p.Val600 and p.Lys601)352.2 STK11 associated30.2 ERBB4 associated10.1PTEN201.5 MAP2K1 associated20.1 PTEN (second)10.1ERBB2181.3 ERBB4 associated10.1 DDR2 associated10.1MAP2K1110.8 Second10.1ERBB480.6 Second10.1DDR280.6FGFR380.6FGFR240.3MET (except exon 14)30.2ALK20.1 ERBB4 associated10.1 DDR2 associated10.1AKT110.1FGFR110.1Total Cases with one mutation24518.2 Cases with two mutations191.4 Cases with three mutations30.2 Cases with at least one mutation26719.9 Open table in a new tab A total of 692 VOD mutations were found in 635 cases: one VOD mutation in 580 cases, two VOD mutations in 52 cases, and three VOD mutations in three cases. The most frequent co-occurring VOD mutations were complex EGFR events. All samples with three VOD mutations were EGFR mutated cases with a combination of r
Referência(s)