Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Biomarker testing for advanced lung cancer by next-generation sequencing; a valid method to achieve a comprehensive glimpse at mutational landscape

Biomarker testing for advanced lung cancer by next-generation sequencing; a valid method to... Background: Next-generation sequencing (NGS) based assay for finding an actionable driver in non-small-cell lung cancer is a less used modality in clinical practice. With a long list of actionable targets, limited tissue, arduous single-gene assays, the alternative of NGS for broad testing in one experiment looks attractive. We report here our experience with NGS for biomarker testing in hundred advanced lung cancer patients. Methods: Predictive biomarker testing was performed using the Ion AmpliSeq™ Cancer Hotspot Panel V2 (30 tumors) and Oncomine™ Solid Tumor DNA and Oncomine™ Solid Tumor Fusion Transcript kit (70 tumors) on Ion- Torrent sequencing platform. Results: One-seventeen distinct aberrations were detected across 29 genes in eighty-six tumors. The most commonly mutated genes were TP53 (43% cases), EGFR (23% cases) and KRAS (17% cases). Thirty-four patients presented an actionable genetic variant for which targeted therapy is presently available, and fifty-two cases harbored non-actionable variants with the possibility of recruitment in clinical trials. NGS results were validated by individual tests for detecting EGFR mutation, ALK1 rearrangement, ROS1 fusion, and c-MET amplification. Compared to single test, NGS exhibited good agreement for detecting EGFR mutations and ALK1 fusion (sensitivity- 88.89%, specificity- 100%, Kappa-score 0.92 and sensitivity- 80%, specificity- 100%, Kappa-score 0.88; respectively). Further, the response of patients harboring tyrosine kinase inhibitor (TKI) sensitizing EGFR mutations was assessed. The progression-free-survival of EGFR positive patients on TKI therapy, harboring a concomitant mutation in PIK3CA- mTOR and/or RAS-RAF-MAPK pathway gene and/or TP53 gene was inferior to those with sole-sensitizing EGFR mutation (2 months vs. 9.5 months, P = 0.015). (Continued on next page) * Correspondence: anumehta11@gmail.com Department of Laboratory, Transfusion and Molecular Diagnostic Services, Rajiv Gandhi Cancer Institute & Research Centre, Sector-V, Rohini, Delhi 110085, India Department of Research, Rajiv Gandhi Cancer Institute & Research Centre, Rohini, Delhi 110085, India Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Mehta et al. Applied Cancer Research (2020) 40:4 Page 2 of 12 (Continued from previous page) Conclusions: This is the first study from South Asia looking into the analytical validity of NGS and describing the mutational landscape of lung cancer patients to study the impact of co-mutations on cancer biology and treatment outcome. Our study demonstrates the clinical utility of NGS testing for identifying actionable variants and making treatment decisions in advanced lung cancer. Keywords: Lung cancer, Driver mutations, Single-gene assay, High-throughput sequencing, Compound mutations Background otherwise specified and a few squamous cell carcinoma Cancer is recognized as a genetic disorder. Genetic alter- patients selected by physician based on clinical features ations in the lung adenocarcinoma (LUAD) have been with a high likelihood of finding a driver mutation. Pre- well documented and are observed in more than 60% of dictive biomarker testing by NGS and single gene tests the cases. Identifying a driver mutation in one of the were performed at the time of diagnosis. several oncogenes like EGFR, ALK, ROS1, BRAF, KRAS, A total of hundred patients were tested on NGS using AKT1, HER2, MEK1, MET, NRAS, PIK3CA and RET can ‘Ion AmpliSeq™ Cancer Hotspot Panel V2’ (4475346, have therapeutic bearings in LUAD [1, 2]. Driver Thermo Fisher Scientific) for the first thirty cases, and mutation-based targeted therapies, wherever possible, ‘Oncomine™ Solid Tumor DNA and Oncomine™ Solid improve the median overall survival of patients with Tumor Fusion Transcript Kit’ (A26761, A26762, metastatic LUAD by at least one year [3]. Single-gene as- Thermo Fisher Scientific) for the later seventy cases. says to identify actionable mutations is the current The genes interrogated by Ion AmpliSeq™ Cancer Hot- standard of care in advanced LUAD. However, for spot Panel v2 (hotspot regions of 50 oncogenes/ tumor identifying actionable driver mutations beyond EGFR suppressor genes) and the Oncomine™ Solid Tumor sensitizing mutations, ALK1 and ROS1 fusion rearrange- DNA & Oncomine™ Solid Tumor Fusion Transcript kits ments, next-generation sequencing (NGS) is the most (DNA somatic variants in 22 key solid tumor genes and practical option, given the limited availability of biopsy RNA fusion transcript in ALK, RET, ROS1 and NTRK1) material and arduous single-gene assays. NGS based has been shown in Additional file 1. genetic profiling of advanced solid tumors is a relatively new technique and allows comprehensive search for pre- Isolation of nucleic acid dictive biomarkers in a resource and tissue proficient Formalin-fixed and paraffin-embedded (FFPE) tumor bi- manner [4–8]. Moreover, the broader molecular profile opsies were reviewed by AM. Slides with ~ 10% or more by NGS allows evaluation of the variants of potential tumor area were selected or obtained after macrodissec- clinical significance (Tier II genetic alterations) leading tion. Genomic DNA was isolated from FFPE sections to many additional patients drawing benefits of targeted (5x, 10 μm) using the Promega ReliaPrep™ FFPE gDNA therapy [9]. Miniprep System (A2352, USA) and RNA was extracted To establish the theoretical advantages of NGS over using the Promega ReliaPrep™ FFPE Total RNA Mini- single-gene assay in the real world, we undertook bio- prep System (Z1002, USA). Nucleic acid was quantitated marker testing for advanced non-small-cell lung cancer by Qubit® 3.0 Fluorometer (Invitrogen Life Technolo- (NSCLC) patients by NGS and compared the results gies). cDNA was generated from 10 ng of RNA using the with single-gene assays to determine its accuracy, re- SuperScript™ VILO™ cDNA synthesis kit (11754050, liability, and benefits in understanding the cancer biology USA). The protocol followed was according to the in relation to the effect of co-mutations on treatment vendor’s insert. results and survival statistics. Library preparation and ion-torrent based NGS Methods Amplicon library was prepared using 10 ng of DNA/ The study was approved by our ‘Institutional Review cDNA using the respective targeted panel. Primers were Board’ (RGCIRC/ IRB/ 277/ 2019) and conducted in ac- partially digested and the amplicons were phosphory- cordance with the Declaration of Helsinki. Clinical char- lated with the FuPa reagent. Sample barcoding was per- acteristics, treatment details, and outcome were curated formed using Ion Dx barcodes and the samples were from the electronic medical record of the institute. adaptor-ligated. The amplified product was purified and The study is single institutional; the time period spans the sequencing library was prepared with Ion AmpliSeq™ from January 2015 to December 2018. The cases in- Library Kit Plus (A35907, Thermo Fisher Scientific). cluded were consecutive advanced lung cancer patients Consequently, emulsion PCR was performed using Ion diagnosed as NSCLC- adenocarcinoma, NSCLC- not PI™ Hi-Q™ One Touch2 (OT2) 200 template kit Mehta et al. Applied Cancer Research (2020) 40:4 Page 3 of 12 (A26434) and the library was enriched on Ion One Immunohistochemistry for ALK1 protein Touch™ ES instrument. Sequencing was performed ALK1 protein was tested by immunohistochemistry employing Ion PGM™ Hi-Q™ Sequencing kit (A25592) (IHC) using anti-ALK (D5F3) rabbit monoclonal primary on Ion 318™ Chip v2 (8–10 samples on a single chip for antibody with other proprietary components of the each sequencing run), on the Ion Personal Genome Ma- VENTANA ALK (D5F3) CDx assay (790–4796, Roche) chine™ (PGM™) System (Thermo Fisher Scientific). The on Ventana BenchMark XT automated slide stainer protocol followed at each step was in line with manufac- (using Ventana’s OptiView DAB IHC Detection kit and turers’ instructions without any modifications. OptiView Amplification kit); performed according to the manufacturer’s recommendations. NGS data analysis Sequencing data were checked for quality metrics using Detection of ROS1 rearrangement and c-MET gene the Torrent Suite version 5.0.2 (Thermo Fisher Scien- amplification by fluorescence in situ hybridization tific). Somatic variants and fusion transcripts were called Fluorescence in situ hybridization (FISH) was performed by Ion Reporter using the specific BED files employing on FFPE lung tissue sections of 4–5 μm, placed on posi- appropriate filters. Integrative Genomic Viewer (IGV) tively charged slides. The specimens used for this study version 2.3 (or higher) was used to verify the variants were hybridized using break-apart probe set (ZytoLight® called and to identify short reads with potential mis- SPEC ROS1 Dual Color Break Apart Probe, ZytoVision, priming events. The Oncomine™ Knowledgebase Re- GmbH, Germany), according to the manufacturer’s porter Software (Thermo Fisher Scientific) was used for instructions. FISH measurements were performed using the final report generation, based on the sequence align- fluorescence microscope Leica DM6000B (Leica, Japan). ment with the reference genome hg 19. The threshold of The hybridized sections were examined under 100x the mutation frequency was 2% at a median coverage magnification for break-apart signals. A distance of depth of >1000X. Variants of unknown significance were more than 1 signal diameter between red and green checked on the VarSome search engine which allows ac- signals was considered positive. Lesser than 5 split cess to publications, ClinVar and all in silico prediction signals were reported negative and > 25 split signals tools on one single site [10, 11]. Further, the OncoPrint were considered positive on a count of 50 cells. In heat map and mutation plots were generated using the the case of 6–24 split signals, a second operator online customized tools OncoPrinter and MutationMap- repeated the count. An average of ≥15% signals was per, respectively, at cBioPortal for Cancer Genomics [12, considered positive. 13]. All the pathogenic mutations were checked in the c-MET in situ hybridization was done as per the NCBI and COSMIC databases [14, 15]. TP53 mutations manufacturer’s protocol (ZytoLight directly labeled were compared and analyzed in the IARC TP53 database LSI MET DNA probe; green and CEN-7 probe; (version R19) [16]. orange). A centromeric 7 probe to MET signal ratio > 2.5 with an average number of ≥5MET signals Single tests was considered positive. For determination of analytical and clinical validity of NGS vis-à-vis reference method of single-gene assay, Targeted therapy response in EGFR mutated tumors four single-gene analyses for sensitizing mutations in Therapeutic decision was made according to the NGS the EGFR gene, rearrangement of ALK1 and ROS1 test results. Based on the NGS profile, the EGFR mu- and amplification of c-MET were performed on all tated patients were separated into two groups as those cases where adequate tissue or cytology material was with (i) isolated EGFR mutations and (ii) compound mu- available. tations with a concurrent mutation in PIK3CA – mTOR pathway/ MAPK pathway and/ or TP53 mutation. The EGFR mutation analysis response of the patients put on small molecule Tyrosine Mutational analysis for EGFR was done using Qiagen Kinase Inhibitor (TKI) was assessed. Those patients who EGFR Therascreen® RGQ PCR Kit (870111). Five sec- showed complete/ partial response were classified as tions of 4 μm each were collected in an Eppendorf tube responders. While those with stable/ progressive with manual macro-dissection to enrich tumor fraction disease were grouped as non-responders to the treat- wherever necessary. DNA was extracted using Qiagen ment. The radiological response was evaluated by the DNeasy blood and tissue kit (69504). Multiplex Real- treating physician according to the Response Evalu- Time PCR was carried out on Rotor-Gene Q thermal ation Criteria In Solid Tumors (RECIST). The interval cycler (Qiagen) in 8 tubes along with positive and no- between the computed tomography scan was once in template controls. The interpretation was done as per 3–4 months/ a smaller interval as directed by the the vendor’s insert. treating physician. Mehta et al. Applied Cancer Research (2020) 40:4 Page 4 of 12 Statistical analysis for ALK1, RET, ROS1 and NTRK1 were tested in 70 tu- Descriptive statistics were used to summarize data. As- mors. The variant data summary as visualized on the sociation between the EGFR mutation status and gen- OncoPrint heat map has been shown in Fig. 1. Most pa- der/ smoking history was analyzed using Pearson’s chi- tients (44 cases) presented single-gene mutation, 27 squared test. The diagnostic test evaluation was per- cases presented mutations in two genes, 10 cases har- formed on MedCalc for Windows, version 15.8 (Med- bored mutations in three genes and a lesser number (5 Calc Software bvba, Ostend, Belgium). The sensitivity, cases) presented more than three mutations. Among specificity, and positive and negative predictive values these, the commonest genetic alteration was in the TP53 (NGS versus single tests) were computed by taking the gene (43% cases), followed by EGFR (23%) and KRAS single-gene assay results as a reference. Further, the (17%). The frequency of mutation in other genes ranged concordance of the two techniques was measured by from 1 to 7% (Fig. 1). The mutational plots for TP53, Cohen’s Kappa statistics. EGFR and KRAS genes have been displayed in Fig. 2 (A- The Progression-Free Survival (PFS) was calculated C). Most of the pathogenic TP53 mutations localized to from the date of the start of TKI till the date of radio- the DNA-binding domain. In total 40 distinct TP53 mu- logical progression/ death. Kaplan-Meier survival curves tations were detected, among them the p.P72R was the for the single and compound EGFR gene mutation most frequent variant (n = 10). Missense TP53 mutations groups were plotted and compared by the log-rank test. were common (35 of the 40 distinct TP53 variants, The limit of statistical significance was set as 0.05 (5% 87.5%), while 2 microindels (p.G293fs and p.P301fs) level). Statistical analyses were performed using the SPSS leading to a frameshift (5%) and three protein-truncating version 23.0 software package (IBM Corp, Armonk, NY). nonsense mutations p.R213*, p.R306* and p.E349* (7.5%) were observed (Fig. 2A). The mutations were further an- Results alyzed on the IARC TP53 database (Additional file 2). Patient characteristics All the TP53 mutations identified in the cohort have The baseline characteristics of the 100 advanced NSCLC been previously reported. One significant observation patients tested on the multigene panel have been pre- that emerged from the IARC TP53 database was that 32 sented in Table 1. Majority tumors (98%) were adeno- of the 35 missense TP53 mutations were pathogenic as carcinoma and two (2%) were squamous cell carcinoma. per the SIFT predictions (Additional file 2E). EGFR gene alterations were detected in 23/100 cases. Distribution of the oncogenic driver mutations in the Among the adenocarcinoma patients, the EGFR muta- cohort tions were more frequent in females than in the males Targeted sequencing identified pathogenic alterations in (32.6% versus 15.4%, P = 0.045) and in never-smokers 28 different genes (Additional file 1). Moreover, copy than the ever-smokers (32.3% versus 9.5%, P = 0.041). number change in the gene CDK4 (not included in the Deletion in exon 19 was the commonest mutation (Fig. panel) was called for in a single case. The gene fusions 2B). A coexistent p.T790M mutation was observed in three tumors. Four uncommon EGFR Exon 20 insertion mutations were detected (p.D770delinsES, p.D770_ Table 1 Baseline characteristics of the cohort (N = 100, lung 771insG, p.A767_S768insSVD and p.P772_H773insHV). cancer subjects) Of interest was a single subject who showed a gamut of Patient characteristics N = 100 (%) EGFR alterations that included inframe exon 19 deletion, Age EvIII fusion and copy number gain. Median (Range) 57 (26–85) Somatic mutation in the KRAS gene occurred in 17/100 Mean ± SE 55.3 ± 13.5 subjects. These mutations were missense substitutions that changed the amino acid glycine in codon 12/13 (Fig. 2C). Gender In total NGS detected 117 distinct pathogenic alter- Male 52 ations in 29 different cancer-driver genes. To correlate Female 48 these aberrations clinically, they were categorized into Stage three groups (i) Tier I: driver mutations that are actionable IIIB 11 by Food and Drug Administration (FDA) approved tar- IV 89 geted therapies (EGFR, ALK1 fusions, ROS1 fusions and BRAF (p.V600E) (ii) Tier II: alterations in well-known can- Smoking history cer oncogenes, actionable by targeted agents not-yet- Ever-smokers 21 approved by FDA (ERBB2, RET and MET amplification, Never-smokers 64 and MET Exon 14 skipping mutation) (iii) Tier III: Clinic- Not assessed 15 ally significant, non-actionable variants (all other genes). Mehta et al. Applied Cancer Research (2020) 40:4 Page 5 of 12 Fig. 1 OncoPrint showing the distribution of genomic alterations in 29 genes and 100 lung cancer cases. Note: For cases 1–30 the IonAmpliSeq™ Cancer Hotspot Panel V2 was used, and in cases 31–100 the Oncomine™ Solid Tumor DNA and Oncomine™ Solid Tumor Fusion Transcript kit was employed. Frequency (%) for the genes APC, ATM, FLT3, GNAQ, IDH2, JAK3, KDR, KIT, SMARCB1 and SMO have been calculated in 30 patients; and for fusion transcripts in ALK, RET and ROS1 has been calculated among the 70 tested cases. The frequency for all other genes, common to both the panels has been calculated in 100 cases. The red, green and blue asterisk symbol (*) indicates patients with ≥3, 2 and 1 pathogenic mutation(s), respectively. No mutation was detected in 14 cases Clinically actionable and non-actionable variants detected III) (Fig. 3). Two novel Tier III variants in the genes by sequencing FGFR3 (p.G90del) and IDH2 (p.T138A) were identi- The alterations observed related to approved/ emerging fied which have not been reported previously (Additional treatments have been enlisted in Additional file 3.In file 3). Twenty-two patients harbored more than one total, fourteen different variants/ mutation subtypes were mutation belonging to Tiers I & II/ I & III/ II & III / I, seen in Tier I genes (12%) and six in Tier II genes (5%). II & III. Ninety-seven alterations (83%) were detected in Tier III genes (Fig. 3, Additional file 3). In terms of patient Comparison of NGS and single testing methods for population showing Tier I and Tier II gene mutation detecting genomic alterations in EGFR, ALK1, ROS1 and type, 29 cases (29%) and 5 cases (5%), respectively, were c-MET identified. There were 14 cases (14%) without any We next validated the NGS results for four genes for somatic mutation in the tested genes. While 52 cases which single assay as mentioned in the methods section (52%) had non-actionable genomic alterations (Tier were performed to study the concordance and analytical Mehta et al. Applied Cancer Research (2020) 40:4 Page 6 of 12 Fig. 2 Lollipop plots depicting the distribution of (A) TP53 (B) EGFR (C) KRAS mutations detected on NGS in the study group (N = 100, carcinoma lung subjects). Please note: The four uncommon EGFR gene mutations (p.D770delinsES, p.D770_771insG, p.A767_S768insSVD, p.P772_H773insHV) have been depicted as ‘U’ in the panel B. (GenBank Reference TP53: NM_000546, EGFR: NM_005228, KRAS: NM_033360) validity of NGS. Due to limited tissue availability, single assay was 88.89 and 100%, respectively. Both methods tests were not done in all the samples. Also, NGS test achieved almost perfect agreement (Kappa-score = 0.92) for identifying rearranged ALK1 and ROS1 were per- (Table 3). Similarly, for the detection of ALK1 rearrange- formed in 70 cases (those tested on the Oncomine™ ment, both NGS and IHC methods positively confirmed 4 Solid Tumor Fusion panel). Therefore, in total 75, 49, 43 cases, whereas the result was discordant for a single patient. and 62 cases each have been compared for EGFR The sensitivity and specificity estimates were 80 and 100%, mutation, ALK1 rearrangement, ROS1 fusion and c-MET respectively (Kappa-score = 0.88). With respect to the de- amplification, respectively, for the single test and NGS tectionof ROS1fusionand c-MET expression, the specifi- outcome (Table 2). city of both diagnostic methods was high, but sensitivity Among the seventy-five EGFR gene mutation tested was not determined. The single ROS1 fusion variant patients, results were concordant in all except for two detected by NGS went undetected on FISH. Also, cases. Discordance was observed in two subjects harbor- results varied for c-MET amplification. NGS missed the ing p.L858R mutation. Compared to multiplexed real- four c-MET amplified cases that were detected on FISH time PCR, the sensitivity and specificity of the NGS (Table 3). Mehta et al. Applied Cancer Research (2020) 40:4 Page 7 of 12 Fig. 3 Dough-nut plot showing the percentage of patients with clinically actionable (Tier I and Tier II) and non-actionable (Tier III) genetic alterations detected by NGS in the study group (N = 100, Lung cancer patients) Table 2 Summary of single test and NGS performed in the cohort Total Total Single test vs. NGS Single test NGS test Single test NGS EGFR mutation n = 75 (%) N = 100 (%) n = 75 (%) Mutant 18 (24) 23 (23) 18 (24) 20 (26.7) Exon 19 deletion 9 10 9 9 p.L858R 6 6 6 4 Exon 19 deletion and p.T790M 2 2 2 2 p.L858R and p.T790M 1 1 1 1 Uncommon mutations _ 4 _ 4 Wild 57 (76) 77 (77) 57 (76) 55 (73.3) Concordant cases 73 (97.3) Discordant cases 2 (2.7) ALK1 rearrangement n = 73 (%) n = 70 (%) n = 49 (%) Mutant 5 (6.8) 5 (7.1) 5 (10.2) 4 (8.2) Wild 68 (93.2) 65 (92.9) 44 (89.8) 45 (91.8) Concordant cases 48 (98) Discordant cases 1 (2) ROS1 fusion n = 58 (%) n = 70 (%) n = 43 (%) Mutant 0 (0) 3 (4.3) 0 (0) 1 (2.3) Wild 58 (58) 67 (95.7) 43 (100) 42 (97.7) Concordant cases 42 (97.7) Discordant cases 1 (2.3) c-MET amplification n = 62 (%) N = 100 (%) n = 62 (%) Mutant 4 (6.5) 2 (2) 4 (6.5) 0 (0) Wild 58 (93.5) 98 (98) 58 (93.5) 62 (100) Concordant cases 58 (93.5) Discordant cases 4 (6.5) Mehta et al. Applied Cancer Research (2020) 40:4 Page 8 of 12 Table 3 Comparison of the clinical performance of NGS and single testing platforms for detecting genomic alterations; taking single test as the reference method NGS Single Sensitivity (%) Specificity (%) PPV NPV (%) Accuracy (%) Cohen’s kappa (κ) test (%) (+) (−) EGFR mutation (n = 75) Multiplex Real-Time PCR *(+) 16 0 88.89 (65.29– 100 (93.73–100) 100 96.61 (88.53– 97.33 (90.70–99.68) 0.92 (0.82–1.00) 98.62) 99.06) (−)2 57 Almost perfect agreement ALK1 rearrangement IHC D5F3 (n = 49) assay (+) 4 0 80 (28.36– 100 (91.96–100) 100 97.78 (88.4– 97.96 (89.15–99.95) 0.88 (0.64–1.00) 99.49) 99.61) (−)1 44 Almost perfect agreement ROS1 fusion (n = 43) FISH (+) 0 1 – 97.67 (87.71–99.94) – 100 97.67 (87.71–99.94) – (−)0 42 c-MET amplification FISH (n = 62) (+) 0 0 – 100 (93.84–100) – 93.55 93.55 (84.30–98.21) – (−)4 58 Please Note: 95% CI values have been bracketed PPV: Positive Predictive Value NPV: Negative Predictive Value *The four uncommon mutations in EGFR have been excluded from the comparison Response to targeted therapy in patients with single and NSCLC. The standard single-gene assays are demanding compound EGFR gene mutations in terms of both tissue and time. Next-generation se- To determine the impact of co-mutations on treatment quencing techniques interrogate several cancer-driver- response, we examined the PFS of the twenty-three cases gene alterations, thereby providing a mutational portrait harboring isolated EGFR mutations, or EGFR mutations even in those tumors which have a low tumor fraction. along with mutations in genes involved in the RAS/RAF/ Short turnaround time can be another advantage if the MEK/ERK/MAPK or PIK3CA/AKT/mTOR pathway volume of tests available is optimal for chip/ flow cell and/ or concomitant TP53 mutation (Additional file 4). usage. Despite these benefits, molecular testing requires The response was not evaluated in three cases (two ex- performance characteristics of the NGS techniques ac- ternal outpatient cases and one terminally ill patient). ceptable in terms of analytical validity, clinical validity Also, the four cases with uncommon EGFR gene and clinical utility vis-a-vis the single-gene assays. This mutations were excluded from the PFS analysis, as these is the first study from South Asia looking into the ana- patients were not treated by TKI inhibitors. lytical validity of NGS and describing the mutational TP53 gene mutations were observed in 8/23 (34%) sub- landscape of lung cancer patients to study the impact of jects. While 6/23 (26%) subjects each showed co-mutations co-mutations on cancer biology and treatment outcome. of EGFR with KRAS-BRAF-MAPK pathway genes or with NGS technique was applied to FFPE tumor blocks of PIK3CA-mTOR pathway genes. In total, 16 patients re- hundred lung cancer subjects. The most commonly mu- ceived TKI therapy. Eight patients (50%) harbored co- tated gene was TP53 (43%). This frequency is similar to mutations, and showed a significantly shorter PFS than a study by Tsoulos N. et al. on 502 NSCLC patients but those with single EGFR gene mutation [median PFS = 2 is lower to ‘The Cancer Genome Atlas’ (TCGA) data months, 95% CI (0.00–5.46) versus 9.5 months, 95% CI (51.8%) [17–20]. About 88% of the TP53 mutations (0.52–18.5), respectively; P =0.015] [Fig. 4 (A-C)]. identified in our study were missense mutations and yet Log Rank were pathogenic as per the IARC database. The high rate Discussion of missense pathogenic mutations is unique to TP53 Predictive biomarker identification with cognate targeted where unlike other genes the frameshift and nonsense therapy has improved the treatment outcomes in mutations are infrequently pathogenic. Also, we Mehta et al. Applied Cancer Research (2020) 40:4 Page 9 of 12 Fig. 4 (a) Kaplan-Meier survival curve showing Progression-Free Survival (PFS) in patients with isolated single and compound EGFR mutations (n = 16). (b) Pathway specific co-mutations (encircled) as observed in the compound EGFR mutant group. (c) OncoPrint presenting the summary of the twenty-three EGFR mutated tumors for the single [1–11] and compound [12–23] mutation groups and the patients’ response to TKI therapy. The response could not evaluated in the cases 9, 10, 11, 20, 21, 22 and 23 as they were outpatients/ not treated by TKI. A: Afatinib, E: Erlotinib, G: Gefitinib, O: Osimertinib observed a high prevalence TP53 alteration ‘p.P72R’ (10 meta-analysis evaluating “thestrengthofevidenceof cases, 19.2%) at an allele frequency of approximately published candidate-genes association studies in lung 50% suggesting germline status of this variant. Shi et al. cancer”, to have a significant association with lung reported a high allele frequency of the p.P72R in Eastern cancer susceptibility [23]. This SNP has been widely Asia [21]. This SNP is known as PEX4 (pleomorphism researched [22, 24, 25] and has been linked with a in exon 4), has been reported at reasonably high rates small and definitive risk of several sporadic cancers worldwide though its prevalence in the Indian popula- due to reduced functional efficiency of p53 protein tion is unknown because of a lack of an available coded on this SNP. database. The residue 72 in the TP53 protein is not Around 23% of the cases presented aberrations in conserved residue to severely hamper protein struc- the EGFR gene. This frequency matches a large ture and function. Besides, proline is physiochemically cohort study from India by Chougule et al. [26]. Also, not too different from arginine [22]; yet this SNP has consistent with previous studies, the EGFR gene been shown by Wang and colleagues in their large mutations were more frequent in women and in non- Mehta et al. Applied Cancer Research (2020) 40:4 Page 10 of 12 smokers in the LUAD patients [27]. Four rare EGFR EGFR active mutations are associated with inferior clin- Exon 20 insertion mutations were detected by NGS, ical outcome in EGFR mutated LUAD patients [20, 33, and it’sratein EGFR mutated patients 17.4% (4/23) 34]. The study by Barnet et al. observed significantly was comparable to a recent study from India that has re- shorter PFS in the NSCLC patients harboring compound ported the rate as 18.07% (15/83) [28]. Also, similar to mutations (dual EGFR/ PIK3CA mutations) than the sin- previous studies all the KRAS mutations detected in our gle EGFR mutated cases [34]. Some studies indicate that study were located on Exon 2 [29]. A comparison of the TP53 effects P13K/AKT and ERK pathways and mutated prevalence of frequent driver gene alterations observed in TP53 fails to induce apoptosis in response to TKI [35, this study with the data from the TCGA database for 36]. A similar observation was made in the current study LUAD and that of a recent study in East-Asian patients also with 5/16 EGFR positive patients with a co- has been presented in Additional file 5, where a significant mutation in the TP53 gene showed an inferior response difference in the prevalence of EGFR and KRAS mutations to TKI therapy. were observed. Patients with sole EGFR sensitizing mutations Among the NGS tested cases, 34% (34/100) patients showed significantly longer PFS to those who showed presented an actionable genomic variant according to additional mutations in any one of the three alluded the NCCN guidelines [30, 31]. Among them, the results pathways. Future studies dedicated to the compound of NGS and single-gene assays for identifying EGFR mu- gene mutation group of the patients including the un- tations and ALK1 fusion demonstrated good agreement, common mutation as a group are required to further while the results were more discordant for MET and establish the usage of current targeted therapy in such ROS1 genes. Copy number gain false negativity is likely patients. Also, the imaging performed early in non- due to failure to get adequate amplicons during library responders could lead to a shorter response interval in preparation compared to normalization genes. FISH the group. A prospective study comparing cohorts with assay for ROS1 is fraught with technical inconsistencies isolated sensitizing EGFR mutations against those with like the inability of probes to hybridize and difficult in- dual or several concurrent mutations with a predeter- terpretations and may have contributed to false-negative mined interval for response evaluation will bring fur- results on the FISH assay. Significantly, patients detected ther clarity on this issue. by NGS to have ROS1 fusions which were negative by Lastly, it must be admitted that though NGS allow FISH responded to Crizotinib and prove the superiority multiplexing and offer broad coverage for oncogenes/ of the NGS platform to detect this biomarker. In a retro- tumor suppressor genes, it is associated with variable spective study by Legras et al., TaqMan probes and NGS error rate (0.1–15%) that is often encountered for short were compared for their ability to detect EGFR and reads obtained from FFPE samples [37–39]. This neces- KRAS mutations, and NGS mutation profiles were stud- sitates a larger prospective study with objective response ied on a large series of NSCLC patients (n = 1343) [7]. rates and survival being the study endpoint apart from The results showed a high concordance of the two tech- analytical validity of the NGS. niques, with a Kappa-score for the detection of EGFR gene mutation as 0.99 (95% CI: 0.97–1.00) [7]. The Kappa-value observed in the present study 0.92 (95% CI: Conclusions 0.82–1.00) agrees with the aforementioned study. Study Our study demonstrates the value of NGS in biomarker by Tsoulos et al. showed 100% concordance between testing for advanced/ metastatic NSCLC. The high con- high resolution melting curve analysis and NGS for de- cordance of NGS and single-gene assay results establish tecting mutations in EGFR gene (exons 18, 19, 20 and the analytical validity of NGS. Further, NGS allows the 21), KRAS (exon 2, 3 and 4), and NRAS (exons 11 and identification of additional patients with clinically action- 15), wherein NGS techniques demonstrated enhanced able variants from Tier II who can potentially benefit sensitivity [17]. Other studies by de Leng et al. and Jing from targeted therapy. The broader genomic picture et al. also shows good agreement of NGS and single- promotes the understanding of mutational interactions gene assays for detecting driver-gene mutations in in determining the response to targeted therapies. The NSCLC, exhibiting high sensitivity and specificity, sug- smaller number of cases and limited targeted panel in- gesting the possibility of routine use of NGS assays to stead of a panel with all significantly mutated genes cap- guide clinical decisions [29, 32]. In the latter study, the able of finding all high confidence drivers is the performance of NGS and digital droplet PCR, a tech- limitation of this study along with limited follow-up and nique with high analytical sensitivity (< 1%), were com- survival data. The strength of this work is the prospect- parable for detecting EGFR mutations. ive nature of the study and the use of all FDA approved Previous studies have shown that compound EGFR test methodologies (wherever available) for single-gene mutations and concurrent genomic alterations with assays. Mehta et al. Applied Cancer Research (2020) 40:4 Page 11 of 12 Supplementary information Received: 13 November 2019 Accepted: 14 May 2020 Supplementary information accompanies this paper at https://doi.org/10. 1186/s41241-020-00089-8. References Additional file 1. Venn diagram depicting the 54 genes covered by the 1. Thunnissen E, van der Oord K, den Bakker M. Prognostic and predictive respective panel. The genes in which alterations were detected by NGS biomarkers in lung cancer. A review. Virchows Arch. 2014;464:347–58. sequencing are indicated in red. (N = 100, lung cancer cases). 2. Lovly C, L. Horn, W. Pao. Molecular profiling of lung cancer. My Cancer Genome. 2018. https://www.mycancergenome.org/content/disease/lung- Additional file 2. Pattern and distribution of the TP53 mutations in the cancer/. cohort according to (A) Mutation type (B) Mutation effect on TP53 3. Kris MG, Johnson BE, Berry LD, Kwiatkowski DJ, Iafrate AJ, Wistuba II, et al. protein sequence (C) Exon/ Intron distribution (D) Functional effect (E) Using multiplexed assays of oncogenic drivers in lung cancers to select SIFT prediction (IARC TP53 Database). (N = 100, lung cancer subjects). targeted drugs. Jama. 2014;311:1998–2006. Additional file 3. Clinically actionable and non-actionable alterations 4. Metzker ML. Sequencing technologies - the next generation. Nat rev Genet. detected by NGS in the study group. 2010;11:31–46. Additional file 4. Analysis of Progression-Free Survival (PFS), targeted 5. Diaz Z, Aguilar-Mahecha A, Paquet ER, Basik M, Orain M, Camlioglu E, et al. therapy and mutation profile of EGFR single (S) and co-mutation (C) cases Next-generation biobanking of metastases to enable multidimensional (N = 23, Lung cancer patients). molecular profiling in personalized medicine. Mod Pathol. 2013;26:1413–24. 6. Yamamoto G, Kikuchi M, Kobayashi S, Arai Y, Fujiyoshi K, Wakatsuki T, et al. Additional file 5. Comparison of the prevalence of frequent pathogenic Routine genetic testing of lung cancer specimens derived from surgery, mutations observed in our cohort against the TCGA data and a Japanese bronchoscopy and fluid aspiration by next generation sequencing. Int J cohort study [19–21] Oncol. 2017;50:1579–89. 7. Legras A, Barritault M, Tallet A, Fabre E, Guyard A, Rance B, et al. Validity of Abbreviations targeted next-generation sequencing in routine care for identifying clinically EGFR: Epidermal Growth Factor Receptor; FDA: Food and Drug relevant molecular profiles in non-small-cell lung cancer: results of a 2-year Administration; FFPE: Formalin-fixed and paraffin-embedded; experience on 1343 samples. J Mol Diagn. 2018;20:550–64. FISH: Fluorescence in situ hybridization; IHC: Immunohistochemistry; 8. Garinet S, Laurent-Puig P, Blons H, Oudart JB. Current and future molecular LUAD: Lung adenocarcinoma; NGS: Next-generation sequencing; testing in NSCLC, what can we expect from new sequencing technologies? NSCLC: Non-small-cell lung cancer; PFS: Progression-free survival; J Clin Med. 2018. https://doi.org/10.3390/jcm7060144. RECIST: Response Evaluation Criteria In Solid Tumors; SNP: Single nucleotide 9. Li MM, Datto M, Duncavage EJ, Kulkarni S, Lindeman NI, Roy S, et al. polymorphism; TKI: Tyrosine kinase inhibitor; TCGA: The Cancer Genome Standards and guidelines for the interpretation and reporting of sequence Atlas variants in cancer: a joint consensus recommendation of the Association for Molecular Pathology, American Society of Clinical Oncology, and College of Acknowledgments American Pathologists. J Mol Diagn. 2017;19:4–23. To the Histopathology Laboratory personnel for the IHC experiments. 10. Kopanos C, Tsiolkas V, Kouris A, Chapple CE, Albarca Aguilera M, Meyer R, et al. VarSome: The Human Genomic Variant Search Engine. Bioinformatics. 2018. https://doi.org/10.1093/bioinformatics/bty897. Authors’ contributions 11. Landrum MJ, Lee JM, Benson M, Brown GR, Chao C, Chitipiralla S, et al. AM conceived the idea, supervised experiments, test analysis and manuscript ClinVar: improving access to variant interpretations and supporting writing. SV analysed data and wrote the manuscript. SKS performed NGS evidence. Nucleic Acids Res. 2018;46:D1062–D7. experiments and analysis. MP did Real-Time PCR and analysis. MS supervised 12. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBio experiments and analysed test results. MS performed FISH experiments. UB cancer genomics portal: an open platform for exploring multidimensional guided and treated the patients. The authors read and approved the final cancer genomics data. Cancer Discov. 2012;2:401–4. manuscript. 13. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. Integrative analysis of complex cancer genomics and clinical profiles using Funding the cBioPortal. Sci Signal. 2013;6:pl1. The study was not supported by external agency funds. 14. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res. 2018;46:D8-D13. Availability of data and materials 15. Forbes SA, Beare D, Boutselakis H, Bamford S, Bindal N, Tate J, et al. COSMIC: The data generated and analysed in the study are included in the published somatic cancer genetics at high-resolution. Nucleic Acids Res. 2017;45: article/ as supplementary information. Additional datasets will be made D777–D83. available from the corresponding author on a reasonable request. 16. Bouaoun L, Sonkin D, Ardin M, Hollstein M, Byrnes G, Zavadil J, et al. TP53 variations in human cancers: new lessons from the IARC TP53 database and Ethics approval genomics data. Hum Mutat. 2016;37:865–76. The study was approved by ‘Institutional Review Board’ (RGCIRC/ IRB/ 277/ 17. Tsoulos N, Papadopoulou E, Metaxa-Mariatou V, Tsaousis G, Efstathiadou C, 2019), with a waiver from patient consenting. Tounta G, et al. Tumor molecular profiling of NSCLC patients using next generation sequencing. Oncol Rep. 2017;38:3419–29. Consent for publication 18. Kandoth C, McLellan MD, Vandin F, Ye K, Niu B, Lu C, et al. Mutational Not applicable. landscape and significance across 12 major cancer types. Nature. 2013;502: 333–9. Competing interests 19. Comprehensive molecular profiling of lung adenocarcinoma. Nature. 2014; The authors report no conflict of interest in this work. 543–50. 20. Sato S, Nagahashi M, Koike T, Ichikawa H, Shimada Y, Watanabe S, et al. Author details Impact of concurrent genomic alterations detected by comprehensive Department of Laboratory, Transfusion and Molecular Diagnostic Services, genomic sequencing on clinical outcomes in east-Asian patients with EGFR- Rajiv Gandhi Cancer Institute & Research Centre, Sector-V, Rohini, Delhi mutated lung adenocarcinoma. Sci Rep. 2018;8:1005. 110085, India. Department of Research, Rajiv Gandhi Cancer Institute & 21. Shi H, Tan SJ, Zhong H, Hu W, Levine A, Xiao CJ, et al. Winter temperature Research Centre, Rohini, Delhi 110085, India. Department of Laboratory and and UV are tightly linked to genetic changes in the p53 tumor suppressor Transfusion Services, Molecular Laboratory, Rajiv Gandhi Cancer Institute & pathway in eastern Asia. Am J Hum Genet. 2009;84:534–41. Research Centre, Rohini, Delhi 110085, India. Department of Medical 22. Dumont P, Leu JI, Della Pietra AC, George DL, Murphy M. The codon 72 Oncology, Rajiv Gandhi Cancer Institute & Research Centre, Rohini, Delhi polymorphic variants of p53 have markedly different apoptotic potential. 110085, India. Nat Genet. 2003;33:357–65. Mehta et al. Applied Cancer Research (2020) 40:4 Page 12 of 12 23. Wang J, Liu Q, Yuan S, Xie W, Liu Y, Xiang Y, et al. Genetic predisposition to lung cancer: comprehensive literature integration, meta-analysis, and multiple evidence assessment of candidate-gene association studies. Sci Rep. 2017;7:8371. 24. Thomas M, Kalita A, Labrecque S, Pim D, Banks L, Matlashewski G. Two polymorphic variants of wild-type p53 differ biochemically and biologically. Mol Cell Biol. 1999;19:1092–100. 25. Whibley C, Pharoah PD, Hollstein M. p53 polymorphisms: cancer implications. Nat Rev Cancer. 2009;9:95–107. 26. Chougule A, Prabhash K, Noronha V, Joshi A, Thavamani A, Chandrani P, et al. Frequency of EGFR mutations in 907 lung adenocarcioma patients of Indian ethnicity. PLoS One. 2013;8:e76164. 27. Doval D, Prabhash K, Patil S, Chaturvedi H, Goswami C, Vaid A, et al. Clinical and epidemiological study of EGFR mutations and EML4-ALK fusion genes among Indian patients with adenocarcinoma of the lung. Onco Targets Ther. 2015;8:117–23. 28. Kate S, Chougule A, Joshi A, Noronha V, Patil V, Dusane R, et al. Outcome of uncommon EGFR mutation positive newly diagnosed advanced non-small cell lung cancer patients: a single center retrospective analysis. Lung Cancer (Auckl). 2019;10:1–10. 29. Jing C, Mao X, Wang Z, Sun K, Ma R, Wu J, et al. Next-generation sequencing-based detection of EGFR, KRAS, BRAF, NRAS, PIK3CA, Her2 and TP53 mutations in patients with nonsmall cell lung cancer. Mol Med Rep. 2018;18:2191–7. 30. Lindeman NI, Cagle PT, Aisner DL, Arcila ME, Beasley MB, Bernicker EH, et al. Updated molecular testing guideline for the selection of lung cancer patients for treatment with targeted tyrosine kinase inhibitors: guideline from the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology. Arch Pathol Lab Med. 2018;142:321–46. 31. Ettinger DS, Aisner DL, Wood DE, Akerley W, Bauman J, Chang JY, et al. NCCN guidelines insights: non-small cell lung cancer, version 5.2018. J Natl Compr Cancer Netw. 2018;16:807–21. 32. de Leng WW, Gadellaa-van Hooijdonk CG, Barendregt-Smouter FA, Koudijs MJ, Nijman I, Hinrichs JW, et al. Targeted next generation sequencing as a reliable diagnostic assay for the detection of somatic mutations in Tumours using minimal DNA amounts from formalin fixed paraffin embedded material. PLoS One. 2016;11:e0149405. 33. Kim EY, Cho EN, Park HS, Hong JY, Lim S, Youn JP, et al. Compound EGFR mutation is frequently detected with co-mutations of actionable genes and associated with poor clinical outcome in lung adenocarcinoma. Cancer Biol Ther. 2016;17:237–45. 34. Barnet MB, O'Toole S, Horvath LG, Selinger C, Yu B, Ng CC, et al. EGFR-co- mutated advanced NSCLC and response to EGFR tyrosine kinase inhibitors. J Thorac Oncol. 2017;12:585–90. 35. Kojima K, Konopleva M, Samudio IJ, Ruvolo V, Andreeff M. Mitogen- activated protein kinase kinase inhibition enhances nuclear proapoptotic function of p53 in acute myelogenous leukemia cells. Cancer Res. 2007;67: 3210–9. 36. Zwang Y, Sas-Chen A, Drier Y, Shay T, Avraham R, Lauriola M, et al. Two phases of mitogenic signaling unveil roles for p53 and EGR1 in elimination of inconsistent growth signals. Mol Cell. 2011;42:524–35. 37. Goodwin S, McPherson JD, McCombie WR. Coming of age: ten years of next-generation sequencing technologies. Nature Rev Genet. 2016;17:333–51. 38. Kim J, Park WY, Kim NKD, Jang SJ, Chun SM, Sung CO, et al. Good laboratory standards for clinical next-generation sequencing cancer panel tests. J Pathol Transl Med. 2017;51:191–204. 39. Lee A, Lee SH, Jung CK, Park G, Lee KY, Choi HJ, et al. Use of the Ion AmpliSeq Cancer hotspot panel in clinical molecular pathology laboratories for analysis of solid tumours: with emphasis on validation with relevant single molecular pathology tests and the Oncomine focus assay. Pathol Res Pract. 2018;214:713–9. Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Applied Cancer Research Springer Journals

Biomarker testing for advanced lung cancer by next-generation sequencing; a valid method to achieve a comprehensive glimpse at mutational landscape

Loading next page...
 
/lp/springer-journals/biomarker-testing-for-advanced-lung-cancer-by-next-generation-cCy03AV8B7

References (42)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2020
eISSN
1980-5578
DOI
10.1186/s41241-020-00089-8
Publisher site
See Article on Publisher Site

Abstract

Background: Next-generation sequencing (NGS) based assay for finding an actionable driver in non-small-cell lung cancer is a less used modality in clinical practice. With a long list of actionable targets, limited tissue, arduous single-gene assays, the alternative of NGS for broad testing in one experiment looks attractive. We report here our experience with NGS for biomarker testing in hundred advanced lung cancer patients. Methods: Predictive biomarker testing was performed using the Ion AmpliSeq™ Cancer Hotspot Panel V2 (30 tumors) and Oncomine™ Solid Tumor DNA and Oncomine™ Solid Tumor Fusion Transcript kit (70 tumors) on Ion- Torrent sequencing platform. Results: One-seventeen distinct aberrations were detected across 29 genes in eighty-six tumors. The most commonly mutated genes were TP53 (43% cases), EGFR (23% cases) and KRAS (17% cases). Thirty-four patients presented an actionable genetic variant for which targeted therapy is presently available, and fifty-two cases harbored non-actionable variants with the possibility of recruitment in clinical trials. NGS results were validated by individual tests for detecting EGFR mutation, ALK1 rearrangement, ROS1 fusion, and c-MET amplification. Compared to single test, NGS exhibited good agreement for detecting EGFR mutations and ALK1 fusion (sensitivity- 88.89%, specificity- 100%, Kappa-score 0.92 and sensitivity- 80%, specificity- 100%, Kappa-score 0.88; respectively). Further, the response of patients harboring tyrosine kinase inhibitor (TKI) sensitizing EGFR mutations was assessed. The progression-free-survival of EGFR positive patients on TKI therapy, harboring a concomitant mutation in PIK3CA- mTOR and/or RAS-RAF-MAPK pathway gene and/or TP53 gene was inferior to those with sole-sensitizing EGFR mutation (2 months vs. 9.5 months, P = 0.015). (Continued on next page) * Correspondence: anumehta11@gmail.com Department of Laboratory, Transfusion and Molecular Diagnostic Services, Rajiv Gandhi Cancer Institute & Research Centre, Sector-V, Rohini, Delhi 110085, India Department of Research, Rajiv Gandhi Cancer Institute & Research Centre, Rohini, Delhi 110085, India Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Mehta et al. Applied Cancer Research (2020) 40:4 Page 2 of 12 (Continued from previous page) Conclusions: This is the first study from South Asia looking into the analytical validity of NGS and describing the mutational landscape of lung cancer patients to study the impact of co-mutations on cancer biology and treatment outcome. Our study demonstrates the clinical utility of NGS testing for identifying actionable variants and making treatment decisions in advanced lung cancer. Keywords: Lung cancer, Driver mutations, Single-gene assay, High-throughput sequencing, Compound mutations Background otherwise specified and a few squamous cell carcinoma Cancer is recognized as a genetic disorder. Genetic alter- patients selected by physician based on clinical features ations in the lung adenocarcinoma (LUAD) have been with a high likelihood of finding a driver mutation. Pre- well documented and are observed in more than 60% of dictive biomarker testing by NGS and single gene tests the cases. Identifying a driver mutation in one of the were performed at the time of diagnosis. several oncogenes like EGFR, ALK, ROS1, BRAF, KRAS, A total of hundred patients were tested on NGS using AKT1, HER2, MEK1, MET, NRAS, PIK3CA and RET can ‘Ion AmpliSeq™ Cancer Hotspot Panel V2’ (4475346, have therapeutic bearings in LUAD [1, 2]. Driver Thermo Fisher Scientific) for the first thirty cases, and mutation-based targeted therapies, wherever possible, ‘Oncomine™ Solid Tumor DNA and Oncomine™ Solid improve the median overall survival of patients with Tumor Fusion Transcript Kit’ (A26761, A26762, metastatic LUAD by at least one year [3]. Single-gene as- Thermo Fisher Scientific) for the later seventy cases. says to identify actionable mutations is the current The genes interrogated by Ion AmpliSeq™ Cancer Hot- standard of care in advanced LUAD. However, for spot Panel v2 (hotspot regions of 50 oncogenes/ tumor identifying actionable driver mutations beyond EGFR suppressor genes) and the Oncomine™ Solid Tumor sensitizing mutations, ALK1 and ROS1 fusion rearrange- DNA & Oncomine™ Solid Tumor Fusion Transcript kits ments, next-generation sequencing (NGS) is the most (DNA somatic variants in 22 key solid tumor genes and practical option, given the limited availability of biopsy RNA fusion transcript in ALK, RET, ROS1 and NTRK1) material and arduous single-gene assays. NGS based has been shown in Additional file 1. genetic profiling of advanced solid tumors is a relatively new technique and allows comprehensive search for pre- Isolation of nucleic acid dictive biomarkers in a resource and tissue proficient Formalin-fixed and paraffin-embedded (FFPE) tumor bi- manner [4–8]. Moreover, the broader molecular profile opsies were reviewed by AM. Slides with ~ 10% or more by NGS allows evaluation of the variants of potential tumor area were selected or obtained after macrodissec- clinical significance (Tier II genetic alterations) leading tion. Genomic DNA was isolated from FFPE sections to many additional patients drawing benefits of targeted (5x, 10 μm) using the Promega ReliaPrep™ FFPE gDNA therapy [9]. Miniprep System (A2352, USA) and RNA was extracted To establish the theoretical advantages of NGS over using the Promega ReliaPrep™ FFPE Total RNA Mini- single-gene assay in the real world, we undertook bio- prep System (Z1002, USA). Nucleic acid was quantitated marker testing for advanced non-small-cell lung cancer by Qubit® 3.0 Fluorometer (Invitrogen Life Technolo- (NSCLC) patients by NGS and compared the results gies). cDNA was generated from 10 ng of RNA using the with single-gene assays to determine its accuracy, re- SuperScript™ VILO™ cDNA synthesis kit (11754050, liability, and benefits in understanding the cancer biology USA). The protocol followed was according to the in relation to the effect of co-mutations on treatment vendor’s insert. results and survival statistics. Library preparation and ion-torrent based NGS Methods Amplicon library was prepared using 10 ng of DNA/ The study was approved by our ‘Institutional Review cDNA using the respective targeted panel. Primers were Board’ (RGCIRC/ IRB/ 277/ 2019) and conducted in ac- partially digested and the amplicons were phosphory- cordance with the Declaration of Helsinki. Clinical char- lated with the FuPa reagent. Sample barcoding was per- acteristics, treatment details, and outcome were curated formed using Ion Dx barcodes and the samples were from the electronic medical record of the institute. adaptor-ligated. The amplified product was purified and The study is single institutional; the time period spans the sequencing library was prepared with Ion AmpliSeq™ from January 2015 to December 2018. The cases in- Library Kit Plus (A35907, Thermo Fisher Scientific). cluded were consecutive advanced lung cancer patients Consequently, emulsion PCR was performed using Ion diagnosed as NSCLC- adenocarcinoma, NSCLC- not PI™ Hi-Q™ One Touch2 (OT2) 200 template kit Mehta et al. Applied Cancer Research (2020) 40:4 Page 3 of 12 (A26434) and the library was enriched on Ion One Immunohistochemistry for ALK1 protein Touch™ ES instrument. Sequencing was performed ALK1 protein was tested by immunohistochemistry employing Ion PGM™ Hi-Q™ Sequencing kit (A25592) (IHC) using anti-ALK (D5F3) rabbit monoclonal primary on Ion 318™ Chip v2 (8–10 samples on a single chip for antibody with other proprietary components of the each sequencing run), on the Ion Personal Genome Ma- VENTANA ALK (D5F3) CDx assay (790–4796, Roche) chine™ (PGM™) System (Thermo Fisher Scientific). The on Ventana BenchMark XT automated slide stainer protocol followed at each step was in line with manufac- (using Ventana’s OptiView DAB IHC Detection kit and turers’ instructions without any modifications. OptiView Amplification kit); performed according to the manufacturer’s recommendations. NGS data analysis Sequencing data were checked for quality metrics using Detection of ROS1 rearrangement and c-MET gene the Torrent Suite version 5.0.2 (Thermo Fisher Scien- amplification by fluorescence in situ hybridization tific). Somatic variants and fusion transcripts were called Fluorescence in situ hybridization (FISH) was performed by Ion Reporter using the specific BED files employing on FFPE lung tissue sections of 4–5 μm, placed on posi- appropriate filters. Integrative Genomic Viewer (IGV) tively charged slides. The specimens used for this study version 2.3 (or higher) was used to verify the variants were hybridized using break-apart probe set (ZytoLight® called and to identify short reads with potential mis- SPEC ROS1 Dual Color Break Apart Probe, ZytoVision, priming events. The Oncomine™ Knowledgebase Re- GmbH, Germany), according to the manufacturer’s porter Software (Thermo Fisher Scientific) was used for instructions. FISH measurements were performed using the final report generation, based on the sequence align- fluorescence microscope Leica DM6000B (Leica, Japan). ment with the reference genome hg 19. The threshold of The hybridized sections were examined under 100x the mutation frequency was 2% at a median coverage magnification for break-apart signals. A distance of depth of >1000X. Variants of unknown significance were more than 1 signal diameter between red and green checked on the VarSome search engine which allows ac- signals was considered positive. Lesser than 5 split cess to publications, ClinVar and all in silico prediction signals were reported negative and > 25 split signals tools on one single site [10, 11]. Further, the OncoPrint were considered positive on a count of 50 cells. In heat map and mutation plots were generated using the the case of 6–24 split signals, a second operator online customized tools OncoPrinter and MutationMap- repeated the count. An average of ≥15% signals was per, respectively, at cBioPortal for Cancer Genomics [12, considered positive. 13]. All the pathogenic mutations were checked in the c-MET in situ hybridization was done as per the NCBI and COSMIC databases [14, 15]. TP53 mutations manufacturer’s protocol (ZytoLight directly labeled were compared and analyzed in the IARC TP53 database LSI MET DNA probe; green and CEN-7 probe; (version R19) [16]. orange). A centromeric 7 probe to MET signal ratio > 2.5 with an average number of ≥5MET signals Single tests was considered positive. For determination of analytical and clinical validity of NGS vis-à-vis reference method of single-gene assay, Targeted therapy response in EGFR mutated tumors four single-gene analyses for sensitizing mutations in Therapeutic decision was made according to the NGS the EGFR gene, rearrangement of ALK1 and ROS1 test results. Based on the NGS profile, the EGFR mu- and amplification of c-MET were performed on all tated patients were separated into two groups as those cases where adequate tissue or cytology material was with (i) isolated EGFR mutations and (ii) compound mu- available. tations with a concurrent mutation in PIK3CA – mTOR pathway/ MAPK pathway and/ or TP53 mutation. The EGFR mutation analysis response of the patients put on small molecule Tyrosine Mutational analysis for EGFR was done using Qiagen Kinase Inhibitor (TKI) was assessed. Those patients who EGFR Therascreen® RGQ PCR Kit (870111). Five sec- showed complete/ partial response were classified as tions of 4 μm each were collected in an Eppendorf tube responders. While those with stable/ progressive with manual macro-dissection to enrich tumor fraction disease were grouped as non-responders to the treat- wherever necessary. DNA was extracted using Qiagen ment. The radiological response was evaluated by the DNeasy blood and tissue kit (69504). Multiplex Real- treating physician according to the Response Evalu- Time PCR was carried out on Rotor-Gene Q thermal ation Criteria In Solid Tumors (RECIST). The interval cycler (Qiagen) in 8 tubes along with positive and no- between the computed tomography scan was once in template controls. The interpretation was done as per 3–4 months/ a smaller interval as directed by the the vendor’s insert. treating physician. Mehta et al. Applied Cancer Research (2020) 40:4 Page 4 of 12 Statistical analysis for ALK1, RET, ROS1 and NTRK1 were tested in 70 tu- Descriptive statistics were used to summarize data. As- mors. The variant data summary as visualized on the sociation between the EGFR mutation status and gen- OncoPrint heat map has been shown in Fig. 1. Most pa- der/ smoking history was analyzed using Pearson’s chi- tients (44 cases) presented single-gene mutation, 27 squared test. The diagnostic test evaluation was per- cases presented mutations in two genes, 10 cases har- formed on MedCalc for Windows, version 15.8 (Med- bored mutations in three genes and a lesser number (5 Calc Software bvba, Ostend, Belgium). The sensitivity, cases) presented more than three mutations. Among specificity, and positive and negative predictive values these, the commonest genetic alteration was in the TP53 (NGS versus single tests) were computed by taking the gene (43% cases), followed by EGFR (23%) and KRAS single-gene assay results as a reference. Further, the (17%). The frequency of mutation in other genes ranged concordance of the two techniques was measured by from 1 to 7% (Fig. 1). The mutational plots for TP53, Cohen’s Kappa statistics. EGFR and KRAS genes have been displayed in Fig. 2 (A- The Progression-Free Survival (PFS) was calculated C). Most of the pathogenic TP53 mutations localized to from the date of the start of TKI till the date of radio- the DNA-binding domain. In total 40 distinct TP53 mu- logical progression/ death. Kaplan-Meier survival curves tations were detected, among them the p.P72R was the for the single and compound EGFR gene mutation most frequent variant (n = 10). Missense TP53 mutations groups were plotted and compared by the log-rank test. were common (35 of the 40 distinct TP53 variants, The limit of statistical significance was set as 0.05 (5% 87.5%), while 2 microindels (p.G293fs and p.P301fs) level). Statistical analyses were performed using the SPSS leading to a frameshift (5%) and three protein-truncating version 23.0 software package (IBM Corp, Armonk, NY). nonsense mutations p.R213*, p.R306* and p.E349* (7.5%) were observed (Fig. 2A). The mutations were further an- Results alyzed on the IARC TP53 database (Additional file 2). Patient characteristics All the TP53 mutations identified in the cohort have The baseline characteristics of the 100 advanced NSCLC been previously reported. One significant observation patients tested on the multigene panel have been pre- that emerged from the IARC TP53 database was that 32 sented in Table 1. Majority tumors (98%) were adeno- of the 35 missense TP53 mutations were pathogenic as carcinoma and two (2%) were squamous cell carcinoma. per the SIFT predictions (Additional file 2E). EGFR gene alterations were detected in 23/100 cases. Distribution of the oncogenic driver mutations in the Among the adenocarcinoma patients, the EGFR muta- cohort tions were more frequent in females than in the males Targeted sequencing identified pathogenic alterations in (32.6% versus 15.4%, P = 0.045) and in never-smokers 28 different genes (Additional file 1). Moreover, copy than the ever-smokers (32.3% versus 9.5%, P = 0.041). number change in the gene CDK4 (not included in the Deletion in exon 19 was the commonest mutation (Fig. panel) was called for in a single case. The gene fusions 2B). A coexistent p.T790M mutation was observed in three tumors. Four uncommon EGFR Exon 20 insertion mutations were detected (p.D770delinsES, p.D770_ Table 1 Baseline characteristics of the cohort (N = 100, lung 771insG, p.A767_S768insSVD and p.P772_H773insHV). cancer subjects) Of interest was a single subject who showed a gamut of Patient characteristics N = 100 (%) EGFR alterations that included inframe exon 19 deletion, Age EvIII fusion and copy number gain. Median (Range) 57 (26–85) Somatic mutation in the KRAS gene occurred in 17/100 Mean ± SE 55.3 ± 13.5 subjects. These mutations were missense substitutions that changed the amino acid glycine in codon 12/13 (Fig. 2C). Gender In total NGS detected 117 distinct pathogenic alter- Male 52 ations in 29 different cancer-driver genes. To correlate Female 48 these aberrations clinically, they were categorized into Stage three groups (i) Tier I: driver mutations that are actionable IIIB 11 by Food and Drug Administration (FDA) approved tar- IV 89 geted therapies (EGFR, ALK1 fusions, ROS1 fusions and BRAF (p.V600E) (ii) Tier II: alterations in well-known can- Smoking history cer oncogenes, actionable by targeted agents not-yet- Ever-smokers 21 approved by FDA (ERBB2, RET and MET amplification, Never-smokers 64 and MET Exon 14 skipping mutation) (iii) Tier III: Clinic- Not assessed 15 ally significant, non-actionable variants (all other genes). Mehta et al. Applied Cancer Research (2020) 40:4 Page 5 of 12 Fig. 1 OncoPrint showing the distribution of genomic alterations in 29 genes and 100 lung cancer cases. Note: For cases 1–30 the IonAmpliSeq™ Cancer Hotspot Panel V2 was used, and in cases 31–100 the Oncomine™ Solid Tumor DNA and Oncomine™ Solid Tumor Fusion Transcript kit was employed. Frequency (%) for the genes APC, ATM, FLT3, GNAQ, IDH2, JAK3, KDR, KIT, SMARCB1 and SMO have been calculated in 30 patients; and for fusion transcripts in ALK, RET and ROS1 has been calculated among the 70 tested cases. The frequency for all other genes, common to both the panels has been calculated in 100 cases. The red, green and blue asterisk symbol (*) indicates patients with ≥3, 2 and 1 pathogenic mutation(s), respectively. No mutation was detected in 14 cases Clinically actionable and non-actionable variants detected III) (Fig. 3). Two novel Tier III variants in the genes by sequencing FGFR3 (p.G90del) and IDH2 (p.T138A) were identi- The alterations observed related to approved/ emerging fied which have not been reported previously (Additional treatments have been enlisted in Additional file 3.In file 3). Twenty-two patients harbored more than one total, fourteen different variants/ mutation subtypes were mutation belonging to Tiers I & II/ I & III/ II & III / I, seen in Tier I genes (12%) and six in Tier II genes (5%). II & III. Ninety-seven alterations (83%) were detected in Tier III genes (Fig. 3, Additional file 3). In terms of patient Comparison of NGS and single testing methods for population showing Tier I and Tier II gene mutation detecting genomic alterations in EGFR, ALK1, ROS1 and type, 29 cases (29%) and 5 cases (5%), respectively, were c-MET identified. There were 14 cases (14%) without any We next validated the NGS results for four genes for somatic mutation in the tested genes. While 52 cases which single assay as mentioned in the methods section (52%) had non-actionable genomic alterations (Tier were performed to study the concordance and analytical Mehta et al. Applied Cancer Research (2020) 40:4 Page 6 of 12 Fig. 2 Lollipop plots depicting the distribution of (A) TP53 (B) EGFR (C) KRAS mutations detected on NGS in the study group (N = 100, carcinoma lung subjects). Please note: The four uncommon EGFR gene mutations (p.D770delinsES, p.D770_771insG, p.A767_S768insSVD, p.P772_H773insHV) have been depicted as ‘U’ in the panel B. (GenBank Reference TP53: NM_000546, EGFR: NM_005228, KRAS: NM_033360) validity of NGS. Due to limited tissue availability, single assay was 88.89 and 100%, respectively. Both methods tests were not done in all the samples. Also, NGS test achieved almost perfect agreement (Kappa-score = 0.92) for identifying rearranged ALK1 and ROS1 were per- (Table 3). Similarly, for the detection of ALK1 rearrange- formed in 70 cases (those tested on the Oncomine™ ment, both NGS and IHC methods positively confirmed 4 Solid Tumor Fusion panel). Therefore, in total 75, 49, 43 cases, whereas the result was discordant for a single patient. and 62 cases each have been compared for EGFR The sensitivity and specificity estimates were 80 and 100%, mutation, ALK1 rearrangement, ROS1 fusion and c-MET respectively (Kappa-score = 0.88). With respect to the de- amplification, respectively, for the single test and NGS tectionof ROS1fusionand c-MET expression, the specifi- outcome (Table 2). city of both diagnostic methods was high, but sensitivity Among the seventy-five EGFR gene mutation tested was not determined. The single ROS1 fusion variant patients, results were concordant in all except for two detected by NGS went undetected on FISH. Also, cases. Discordance was observed in two subjects harbor- results varied for c-MET amplification. NGS missed the ing p.L858R mutation. Compared to multiplexed real- four c-MET amplified cases that were detected on FISH time PCR, the sensitivity and specificity of the NGS (Table 3). Mehta et al. Applied Cancer Research (2020) 40:4 Page 7 of 12 Fig. 3 Dough-nut plot showing the percentage of patients with clinically actionable (Tier I and Tier II) and non-actionable (Tier III) genetic alterations detected by NGS in the study group (N = 100, Lung cancer patients) Table 2 Summary of single test and NGS performed in the cohort Total Total Single test vs. NGS Single test NGS test Single test NGS EGFR mutation n = 75 (%) N = 100 (%) n = 75 (%) Mutant 18 (24) 23 (23) 18 (24) 20 (26.7) Exon 19 deletion 9 10 9 9 p.L858R 6 6 6 4 Exon 19 deletion and p.T790M 2 2 2 2 p.L858R and p.T790M 1 1 1 1 Uncommon mutations _ 4 _ 4 Wild 57 (76) 77 (77) 57 (76) 55 (73.3) Concordant cases 73 (97.3) Discordant cases 2 (2.7) ALK1 rearrangement n = 73 (%) n = 70 (%) n = 49 (%) Mutant 5 (6.8) 5 (7.1) 5 (10.2) 4 (8.2) Wild 68 (93.2) 65 (92.9) 44 (89.8) 45 (91.8) Concordant cases 48 (98) Discordant cases 1 (2) ROS1 fusion n = 58 (%) n = 70 (%) n = 43 (%) Mutant 0 (0) 3 (4.3) 0 (0) 1 (2.3) Wild 58 (58) 67 (95.7) 43 (100) 42 (97.7) Concordant cases 42 (97.7) Discordant cases 1 (2.3) c-MET amplification n = 62 (%) N = 100 (%) n = 62 (%) Mutant 4 (6.5) 2 (2) 4 (6.5) 0 (0) Wild 58 (93.5) 98 (98) 58 (93.5) 62 (100) Concordant cases 58 (93.5) Discordant cases 4 (6.5) Mehta et al. Applied Cancer Research (2020) 40:4 Page 8 of 12 Table 3 Comparison of the clinical performance of NGS and single testing platforms for detecting genomic alterations; taking single test as the reference method NGS Single Sensitivity (%) Specificity (%) PPV NPV (%) Accuracy (%) Cohen’s kappa (κ) test (%) (+) (−) EGFR mutation (n = 75) Multiplex Real-Time PCR *(+) 16 0 88.89 (65.29– 100 (93.73–100) 100 96.61 (88.53– 97.33 (90.70–99.68) 0.92 (0.82–1.00) 98.62) 99.06) (−)2 57 Almost perfect agreement ALK1 rearrangement IHC D5F3 (n = 49) assay (+) 4 0 80 (28.36– 100 (91.96–100) 100 97.78 (88.4– 97.96 (89.15–99.95) 0.88 (0.64–1.00) 99.49) 99.61) (−)1 44 Almost perfect agreement ROS1 fusion (n = 43) FISH (+) 0 1 – 97.67 (87.71–99.94) – 100 97.67 (87.71–99.94) – (−)0 42 c-MET amplification FISH (n = 62) (+) 0 0 – 100 (93.84–100) – 93.55 93.55 (84.30–98.21) – (−)4 58 Please Note: 95% CI values have been bracketed PPV: Positive Predictive Value NPV: Negative Predictive Value *The four uncommon mutations in EGFR have been excluded from the comparison Response to targeted therapy in patients with single and NSCLC. The standard single-gene assays are demanding compound EGFR gene mutations in terms of both tissue and time. Next-generation se- To determine the impact of co-mutations on treatment quencing techniques interrogate several cancer-driver- response, we examined the PFS of the twenty-three cases gene alterations, thereby providing a mutational portrait harboring isolated EGFR mutations, or EGFR mutations even in those tumors which have a low tumor fraction. along with mutations in genes involved in the RAS/RAF/ Short turnaround time can be another advantage if the MEK/ERK/MAPK or PIK3CA/AKT/mTOR pathway volume of tests available is optimal for chip/ flow cell and/ or concomitant TP53 mutation (Additional file 4). usage. Despite these benefits, molecular testing requires The response was not evaluated in three cases (two ex- performance characteristics of the NGS techniques ac- ternal outpatient cases and one terminally ill patient). ceptable in terms of analytical validity, clinical validity Also, the four cases with uncommon EGFR gene and clinical utility vis-a-vis the single-gene assays. This mutations were excluded from the PFS analysis, as these is the first study from South Asia looking into the ana- patients were not treated by TKI inhibitors. lytical validity of NGS and describing the mutational TP53 gene mutations were observed in 8/23 (34%) sub- landscape of lung cancer patients to study the impact of jects. While 6/23 (26%) subjects each showed co-mutations co-mutations on cancer biology and treatment outcome. of EGFR with KRAS-BRAF-MAPK pathway genes or with NGS technique was applied to FFPE tumor blocks of PIK3CA-mTOR pathway genes. In total, 16 patients re- hundred lung cancer subjects. The most commonly mu- ceived TKI therapy. Eight patients (50%) harbored co- tated gene was TP53 (43%). This frequency is similar to mutations, and showed a significantly shorter PFS than a study by Tsoulos N. et al. on 502 NSCLC patients but those with single EGFR gene mutation [median PFS = 2 is lower to ‘The Cancer Genome Atlas’ (TCGA) data months, 95% CI (0.00–5.46) versus 9.5 months, 95% CI (51.8%) [17–20]. About 88% of the TP53 mutations (0.52–18.5), respectively; P =0.015] [Fig. 4 (A-C)]. identified in our study were missense mutations and yet Log Rank were pathogenic as per the IARC database. The high rate Discussion of missense pathogenic mutations is unique to TP53 Predictive biomarker identification with cognate targeted where unlike other genes the frameshift and nonsense therapy has improved the treatment outcomes in mutations are infrequently pathogenic. Also, we Mehta et al. Applied Cancer Research (2020) 40:4 Page 9 of 12 Fig. 4 (a) Kaplan-Meier survival curve showing Progression-Free Survival (PFS) in patients with isolated single and compound EGFR mutations (n = 16). (b) Pathway specific co-mutations (encircled) as observed in the compound EGFR mutant group. (c) OncoPrint presenting the summary of the twenty-three EGFR mutated tumors for the single [1–11] and compound [12–23] mutation groups and the patients’ response to TKI therapy. The response could not evaluated in the cases 9, 10, 11, 20, 21, 22 and 23 as they were outpatients/ not treated by TKI. A: Afatinib, E: Erlotinib, G: Gefitinib, O: Osimertinib observed a high prevalence TP53 alteration ‘p.P72R’ (10 meta-analysis evaluating “thestrengthofevidenceof cases, 19.2%) at an allele frequency of approximately published candidate-genes association studies in lung 50% suggesting germline status of this variant. Shi et al. cancer”, to have a significant association with lung reported a high allele frequency of the p.P72R in Eastern cancer susceptibility [23]. This SNP has been widely Asia [21]. This SNP is known as PEX4 (pleomorphism researched [22, 24, 25] and has been linked with a in exon 4), has been reported at reasonably high rates small and definitive risk of several sporadic cancers worldwide though its prevalence in the Indian popula- due to reduced functional efficiency of p53 protein tion is unknown because of a lack of an available coded on this SNP. database. The residue 72 in the TP53 protein is not Around 23% of the cases presented aberrations in conserved residue to severely hamper protein struc- the EGFR gene. This frequency matches a large ture and function. Besides, proline is physiochemically cohort study from India by Chougule et al. [26]. Also, not too different from arginine [22]; yet this SNP has consistent with previous studies, the EGFR gene been shown by Wang and colleagues in their large mutations were more frequent in women and in non- Mehta et al. Applied Cancer Research (2020) 40:4 Page 10 of 12 smokers in the LUAD patients [27]. Four rare EGFR EGFR active mutations are associated with inferior clin- Exon 20 insertion mutations were detected by NGS, ical outcome in EGFR mutated LUAD patients [20, 33, and it’sratein EGFR mutated patients 17.4% (4/23) 34]. The study by Barnet et al. observed significantly was comparable to a recent study from India that has re- shorter PFS in the NSCLC patients harboring compound ported the rate as 18.07% (15/83) [28]. Also, similar to mutations (dual EGFR/ PIK3CA mutations) than the sin- previous studies all the KRAS mutations detected in our gle EGFR mutated cases [34]. Some studies indicate that study were located on Exon 2 [29]. A comparison of the TP53 effects P13K/AKT and ERK pathways and mutated prevalence of frequent driver gene alterations observed in TP53 fails to induce apoptosis in response to TKI [35, this study with the data from the TCGA database for 36]. A similar observation was made in the current study LUAD and that of a recent study in East-Asian patients also with 5/16 EGFR positive patients with a co- has been presented in Additional file 5, where a significant mutation in the TP53 gene showed an inferior response difference in the prevalence of EGFR and KRAS mutations to TKI therapy. were observed. Patients with sole EGFR sensitizing mutations Among the NGS tested cases, 34% (34/100) patients showed significantly longer PFS to those who showed presented an actionable genomic variant according to additional mutations in any one of the three alluded the NCCN guidelines [30, 31]. Among them, the results pathways. Future studies dedicated to the compound of NGS and single-gene assays for identifying EGFR mu- gene mutation group of the patients including the un- tations and ALK1 fusion demonstrated good agreement, common mutation as a group are required to further while the results were more discordant for MET and establish the usage of current targeted therapy in such ROS1 genes. Copy number gain false negativity is likely patients. Also, the imaging performed early in non- due to failure to get adequate amplicons during library responders could lead to a shorter response interval in preparation compared to normalization genes. FISH the group. A prospective study comparing cohorts with assay for ROS1 is fraught with technical inconsistencies isolated sensitizing EGFR mutations against those with like the inability of probes to hybridize and difficult in- dual or several concurrent mutations with a predeter- terpretations and may have contributed to false-negative mined interval for response evaluation will bring fur- results on the FISH assay. Significantly, patients detected ther clarity on this issue. by NGS to have ROS1 fusions which were negative by Lastly, it must be admitted that though NGS allow FISH responded to Crizotinib and prove the superiority multiplexing and offer broad coverage for oncogenes/ of the NGS platform to detect this biomarker. In a retro- tumor suppressor genes, it is associated with variable spective study by Legras et al., TaqMan probes and NGS error rate (0.1–15%) that is often encountered for short were compared for their ability to detect EGFR and reads obtained from FFPE samples [37–39]. This neces- KRAS mutations, and NGS mutation profiles were stud- sitates a larger prospective study with objective response ied on a large series of NSCLC patients (n = 1343) [7]. rates and survival being the study endpoint apart from The results showed a high concordance of the two tech- analytical validity of the NGS. niques, with a Kappa-score for the detection of EGFR gene mutation as 0.99 (95% CI: 0.97–1.00) [7]. The Kappa-value observed in the present study 0.92 (95% CI: Conclusions 0.82–1.00) agrees with the aforementioned study. Study Our study demonstrates the value of NGS in biomarker by Tsoulos et al. showed 100% concordance between testing for advanced/ metastatic NSCLC. The high con- high resolution melting curve analysis and NGS for de- cordance of NGS and single-gene assay results establish tecting mutations in EGFR gene (exons 18, 19, 20 and the analytical validity of NGS. Further, NGS allows the 21), KRAS (exon 2, 3 and 4), and NRAS (exons 11 and identification of additional patients with clinically action- 15), wherein NGS techniques demonstrated enhanced able variants from Tier II who can potentially benefit sensitivity [17]. Other studies by de Leng et al. and Jing from targeted therapy. The broader genomic picture et al. also shows good agreement of NGS and single- promotes the understanding of mutational interactions gene assays for detecting driver-gene mutations in in determining the response to targeted therapies. The NSCLC, exhibiting high sensitivity and specificity, sug- smaller number of cases and limited targeted panel in- gesting the possibility of routine use of NGS assays to stead of a panel with all significantly mutated genes cap- guide clinical decisions [29, 32]. In the latter study, the able of finding all high confidence drivers is the performance of NGS and digital droplet PCR, a tech- limitation of this study along with limited follow-up and nique with high analytical sensitivity (< 1%), were com- survival data. The strength of this work is the prospect- parable for detecting EGFR mutations. ive nature of the study and the use of all FDA approved Previous studies have shown that compound EGFR test methodologies (wherever available) for single-gene mutations and concurrent genomic alterations with assays. Mehta et al. Applied Cancer Research (2020) 40:4 Page 11 of 12 Supplementary information Received: 13 November 2019 Accepted: 14 May 2020 Supplementary information accompanies this paper at https://doi.org/10. 1186/s41241-020-00089-8. References Additional file 1. Venn diagram depicting the 54 genes covered by the 1. Thunnissen E, van der Oord K, den Bakker M. Prognostic and predictive respective panel. The genes in which alterations were detected by NGS biomarkers in lung cancer. A review. Virchows Arch. 2014;464:347–58. sequencing are indicated in red. (N = 100, lung cancer cases). 2. Lovly C, L. Horn, W. Pao. Molecular profiling of lung cancer. My Cancer Genome. 2018. https://www.mycancergenome.org/content/disease/lung- Additional file 2. Pattern and distribution of the TP53 mutations in the cancer/. cohort according to (A) Mutation type (B) Mutation effect on TP53 3. Kris MG, Johnson BE, Berry LD, Kwiatkowski DJ, Iafrate AJ, Wistuba II, et al. protein sequence (C) Exon/ Intron distribution (D) Functional effect (E) Using multiplexed assays of oncogenic drivers in lung cancers to select SIFT prediction (IARC TP53 Database). (N = 100, lung cancer subjects). targeted drugs. Jama. 2014;311:1998–2006. Additional file 3. Clinically actionable and non-actionable alterations 4. Metzker ML. Sequencing technologies - the next generation. Nat rev Genet. detected by NGS in the study group. 2010;11:31–46. Additional file 4. Analysis of Progression-Free Survival (PFS), targeted 5. Diaz Z, Aguilar-Mahecha A, Paquet ER, Basik M, Orain M, Camlioglu E, et al. therapy and mutation profile of EGFR single (S) and co-mutation (C) cases Next-generation biobanking of metastases to enable multidimensional (N = 23, Lung cancer patients). molecular profiling in personalized medicine. Mod Pathol. 2013;26:1413–24. 6. Yamamoto G, Kikuchi M, Kobayashi S, Arai Y, Fujiyoshi K, Wakatsuki T, et al. Additional file 5. Comparison of the prevalence of frequent pathogenic Routine genetic testing of lung cancer specimens derived from surgery, mutations observed in our cohort against the TCGA data and a Japanese bronchoscopy and fluid aspiration by next generation sequencing. Int J cohort study [19–21] Oncol. 2017;50:1579–89. 7. Legras A, Barritault M, Tallet A, Fabre E, Guyard A, Rance B, et al. Validity of Abbreviations targeted next-generation sequencing in routine care for identifying clinically EGFR: Epidermal Growth Factor Receptor; FDA: Food and Drug relevant molecular profiles in non-small-cell lung cancer: results of a 2-year Administration; FFPE: Formalin-fixed and paraffin-embedded; experience on 1343 samples. J Mol Diagn. 2018;20:550–64. FISH: Fluorescence in situ hybridization; IHC: Immunohistochemistry; 8. Garinet S, Laurent-Puig P, Blons H, Oudart JB. Current and future molecular LUAD: Lung adenocarcinoma; NGS: Next-generation sequencing; testing in NSCLC, what can we expect from new sequencing technologies? NSCLC: Non-small-cell lung cancer; PFS: Progression-free survival; J Clin Med. 2018. https://doi.org/10.3390/jcm7060144. RECIST: Response Evaluation Criteria In Solid Tumors; SNP: Single nucleotide 9. Li MM, Datto M, Duncavage EJ, Kulkarni S, Lindeman NI, Roy S, et al. polymorphism; TKI: Tyrosine kinase inhibitor; TCGA: The Cancer Genome Standards and guidelines for the interpretation and reporting of sequence Atlas variants in cancer: a joint consensus recommendation of the Association for Molecular Pathology, American Society of Clinical Oncology, and College of Acknowledgments American Pathologists. J Mol Diagn. 2017;19:4–23. To the Histopathology Laboratory personnel for the IHC experiments. 10. Kopanos C, Tsiolkas V, Kouris A, Chapple CE, Albarca Aguilera M, Meyer R, et al. VarSome: The Human Genomic Variant Search Engine. Bioinformatics. 2018. https://doi.org/10.1093/bioinformatics/bty897. Authors’ contributions 11. Landrum MJ, Lee JM, Benson M, Brown GR, Chao C, Chitipiralla S, et al. AM conceived the idea, supervised experiments, test analysis and manuscript ClinVar: improving access to variant interpretations and supporting writing. SV analysed data and wrote the manuscript. SKS performed NGS evidence. Nucleic Acids Res. 2018;46:D1062–D7. experiments and analysis. MP did Real-Time PCR and analysis. MS supervised 12. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. The cBio experiments and analysed test results. MS performed FISH experiments. UB cancer genomics portal: an open platform for exploring multidimensional guided and treated the patients. The authors read and approved the final cancer genomics data. Cancer Discov. 2012;2:401–4. manuscript. 13. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. Integrative analysis of complex cancer genomics and clinical profiles using Funding the cBioPortal. Sci Signal. 2013;6:pl1. The study was not supported by external agency funds. 14. Database resources of the National Center for Biotechnology Information. Nucleic Acids Res. 2018;46:D8-D13. Availability of data and materials 15. Forbes SA, Beare D, Boutselakis H, Bamford S, Bindal N, Tate J, et al. COSMIC: The data generated and analysed in the study are included in the published somatic cancer genetics at high-resolution. Nucleic Acids Res. 2017;45: article/ as supplementary information. Additional datasets will be made D777–D83. available from the corresponding author on a reasonable request. 16. Bouaoun L, Sonkin D, Ardin M, Hollstein M, Byrnes G, Zavadil J, et al. TP53 variations in human cancers: new lessons from the IARC TP53 database and Ethics approval genomics data. Hum Mutat. 2016;37:865–76. The study was approved by ‘Institutional Review Board’ (RGCIRC/ IRB/ 277/ 17. Tsoulos N, Papadopoulou E, Metaxa-Mariatou V, Tsaousis G, Efstathiadou C, 2019), with a waiver from patient consenting. Tounta G, et al. Tumor molecular profiling of NSCLC patients using next generation sequencing. Oncol Rep. 2017;38:3419–29. Consent for publication 18. Kandoth C, McLellan MD, Vandin F, Ye K, Niu B, Lu C, et al. Mutational Not applicable. landscape and significance across 12 major cancer types. Nature. 2013;502: 333–9. Competing interests 19. Comprehensive molecular profiling of lung adenocarcinoma. Nature. 2014; The authors report no conflict of interest in this work. 543–50. 20. Sato S, Nagahashi M, Koike T, Ichikawa H, Shimada Y, Watanabe S, et al. Author details Impact of concurrent genomic alterations detected by comprehensive Department of Laboratory, Transfusion and Molecular Diagnostic Services, genomic sequencing on clinical outcomes in east-Asian patients with EGFR- Rajiv Gandhi Cancer Institute & Research Centre, Sector-V, Rohini, Delhi mutated lung adenocarcinoma. Sci Rep. 2018;8:1005. 110085, India. Department of Research, Rajiv Gandhi Cancer Institute & 21. Shi H, Tan SJ, Zhong H, Hu W, Levine A, Xiao CJ, et al. Winter temperature Research Centre, Rohini, Delhi 110085, India. Department of Laboratory and and UV are tightly linked to genetic changes in the p53 tumor suppressor Transfusion Services, Molecular Laboratory, Rajiv Gandhi Cancer Institute & pathway in eastern Asia. Am J Hum Genet. 2009;84:534–41. Research Centre, Rohini, Delhi 110085, India. Department of Medical 22. Dumont P, Leu JI, Della Pietra AC, George DL, Murphy M. The codon 72 Oncology, Rajiv Gandhi Cancer Institute & Research Centre, Rohini, Delhi polymorphic variants of p53 have markedly different apoptotic potential. 110085, India. Nat Genet. 2003;33:357–65. Mehta et al. Applied Cancer Research (2020) 40:4 Page 12 of 12 23. Wang J, Liu Q, Yuan S, Xie W, Liu Y, Xiang Y, et al. Genetic predisposition to lung cancer: comprehensive literature integration, meta-analysis, and multiple evidence assessment of candidate-gene association studies. Sci Rep. 2017;7:8371. 24. Thomas M, Kalita A, Labrecque S, Pim D, Banks L, Matlashewski G. Two polymorphic variants of wild-type p53 differ biochemically and biologically. Mol Cell Biol. 1999;19:1092–100. 25. Whibley C, Pharoah PD, Hollstein M. p53 polymorphisms: cancer implications. Nat Rev Cancer. 2009;9:95–107. 26. Chougule A, Prabhash K, Noronha V, Joshi A, Thavamani A, Chandrani P, et al. Frequency of EGFR mutations in 907 lung adenocarcioma patients of Indian ethnicity. PLoS One. 2013;8:e76164. 27. Doval D, Prabhash K, Patil S, Chaturvedi H, Goswami C, Vaid A, et al. Clinical and epidemiological study of EGFR mutations and EML4-ALK fusion genes among Indian patients with adenocarcinoma of the lung. Onco Targets Ther. 2015;8:117–23. 28. Kate S, Chougule A, Joshi A, Noronha V, Patil V, Dusane R, et al. Outcome of uncommon EGFR mutation positive newly diagnosed advanced non-small cell lung cancer patients: a single center retrospective analysis. Lung Cancer (Auckl). 2019;10:1–10. 29. Jing C, Mao X, Wang Z, Sun K, Ma R, Wu J, et al. Next-generation sequencing-based detection of EGFR, KRAS, BRAF, NRAS, PIK3CA, Her2 and TP53 mutations in patients with nonsmall cell lung cancer. Mol Med Rep. 2018;18:2191–7. 30. Lindeman NI, Cagle PT, Aisner DL, Arcila ME, Beasley MB, Bernicker EH, et al. Updated molecular testing guideline for the selection of lung cancer patients for treatment with targeted tyrosine kinase inhibitors: guideline from the College of American Pathologists, the International Association for the Study of Lung Cancer, and the Association for Molecular Pathology. Arch Pathol Lab Med. 2018;142:321–46. 31. Ettinger DS, Aisner DL, Wood DE, Akerley W, Bauman J, Chang JY, et al. NCCN guidelines insights: non-small cell lung cancer, version 5.2018. J Natl Compr Cancer Netw. 2018;16:807–21. 32. de Leng WW, Gadellaa-van Hooijdonk CG, Barendregt-Smouter FA, Koudijs MJ, Nijman I, Hinrichs JW, et al. Targeted next generation sequencing as a reliable diagnostic assay for the detection of somatic mutations in Tumours using minimal DNA amounts from formalin fixed paraffin embedded material. PLoS One. 2016;11:e0149405. 33. Kim EY, Cho EN, Park HS, Hong JY, Lim S, Youn JP, et al. Compound EGFR mutation is frequently detected with co-mutations of actionable genes and associated with poor clinical outcome in lung adenocarcinoma. Cancer Biol Ther. 2016;17:237–45. 34. Barnet MB, O'Toole S, Horvath LG, Selinger C, Yu B, Ng CC, et al. EGFR-co- mutated advanced NSCLC and response to EGFR tyrosine kinase inhibitors. J Thorac Oncol. 2017;12:585–90. 35. Kojima K, Konopleva M, Samudio IJ, Ruvolo V, Andreeff M. Mitogen- activated protein kinase kinase inhibition enhances nuclear proapoptotic function of p53 in acute myelogenous leukemia cells. Cancer Res. 2007;67: 3210–9. 36. Zwang Y, Sas-Chen A, Drier Y, Shay T, Avraham R, Lauriola M, et al. Two phases of mitogenic signaling unveil roles for p53 and EGR1 in elimination of inconsistent growth signals. Mol Cell. 2011;42:524–35. 37. Goodwin S, McPherson JD, McCombie WR. Coming of age: ten years of next-generation sequencing technologies. Nature Rev Genet. 2016;17:333–51. 38. Kim J, Park WY, Kim NKD, Jang SJ, Chun SM, Sung CO, et al. Good laboratory standards for clinical next-generation sequencing cancer panel tests. J Pathol Transl Med. 2017;51:191–204. 39. Lee A, Lee SH, Jung CK, Park G, Lee KY, Choi HJ, et al. Use of the Ion AmpliSeq Cancer hotspot panel in clinical molecular pathology laboratories for analysis of solid tumours: with emphasis on validation with relevant single molecular pathology tests and the Oncomine focus assay. Pathol Res Pract. 2018;214:713–9. Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Journal

Applied Cancer ResearchSpringer Journals

Published: Jun 1, 2020

There are no references for this article.