Driver Gene Mutations in Stools of Colorectal Carcinoma Patients Detected by Targeted Next-Generation Sequencing
2016; Elsevier BV; Volume: 18; Issue: 4 Linguagem: Inglês
10.1016/j.jmoldx.2016.01.008
ISSN1943-7811
AutoresGemma Armengol, Virinder Kaur Sarhadi, Reza Ghanbari, Masoud Doghaei-Moghaddam, Reza Ansari, Masoud Sotoudeh, Pauli Puolakkainen, Arto Kokkola, Reza Malekzadeh, Sakari Knuutila,
Tópico(s)Cancer Genomics and Diagnostics
ResumoDetection of driver gene mutations in stool DNA represents a promising noninvasive approach for screening colorectal cancer (CRC). Amplicon-based next-generation sequencing (NGS) is a good option to study mutations in many cancer genes simultaneously and from a low amount of DNA. Our aim was to assess the feasibility of identifying mutations in 22 cancer driver genes with Ion Torrent technology in stool DNA from a series of 65 CRC patients. The assay was successful in 80% of stool DNA samples. NGS results showed 83 mutations in cancer driver genes, 29 hotspot and 54 novel mutations. One to five genes were mutated in 75% of cases. TP53, KRAS, FBXW7, and SMAD4 were the top mutated genes, consistent with previous studies. Of samples with mutations, 54% presented concomitant mutations in different genes. Phosphatidylinositol 3-kinase/mitogen-activated protein kinase pathway genes were mutated in 70% of samples, with 58% having alterations in KRAS, NRAS, or BRAF. Because mutations in these genes can compromise the efficacy of epidermal growth factor receptor blockade in CRC patients, identifying mutations that confer resistance to some targeted treatments may be useful to guide therapeutic decisions. In conclusion, the data presented herein show that NGS procedures on stool DNA represent a promising tool to detect genetic mutations that could be used in the future for diagnosis, monitoring, or treating CRC. Detection of driver gene mutations in stool DNA represents a promising noninvasive approach for screening colorectal cancer (CRC). Amplicon-based next-generation sequencing (NGS) is a good option to study mutations in many cancer genes simultaneously and from a low amount of DNA. Our aim was to assess the feasibility of identifying mutations in 22 cancer driver genes with Ion Torrent technology in stool DNA from a series of 65 CRC patients. The assay was successful in 80% of stool DNA samples. NGS results showed 83 mutations in cancer driver genes, 29 hotspot and 54 novel mutations. One to five genes were mutated in 75% of cases. TP53, KRAS, FBXW7, and SMAD4 were the top mutated genes, consistent with previous studies. Of samples with mutations, 54% presented concomitant mutations in different genes. Phosphatidylinositol 3-kinase/mitogen-activated protein kinase pathway genes were mutated in 70% of samples, with 58% having alterations in KRAS, NRAS, or BRAF. Because mutations in these genes can compromise the efficacy of epidermal growth factor receptor blockade in CRC patients, identifying mutations that confer resistance to some targeted treatments may be useful to guide therapeutic decisions. In conclusion, the data presented herein show that NGS procedures on stool DNA represent a promising tool to detect genetic mutations that could be used in the future for diagnosis, monitoring, or treating CRC. Colorectal cancer (CRC) represents approximately 10% of all cancers diagnosed worldwide and is the fourth leading cause of death from cancer in both men and women, with 694,000 deaths in 2012. Despite its decreased mortality in the last decades, death rates still remain high, with more deaths in the less developed regions of the world.1Ferlay J. Soerjomataram I. Dikshit R. Eser S. Mathers C. Rebelo M. Parkin D.M. Forman D.D. Bray F. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012.Int J Cancer. 2014; 136: E359-E386Crossref PubMed Scopus (21348) Google Scholar Epidemiological transition in the low- and middle-income countries that are witnessing socioeconomic progress with the adoption of a Western life style with higher intake of fat, meat, and total calories, along with increasing life expectancy, results in a significant increase in the burden of CRC. This cancer type can be curable if detected in stage I or II, usually asymptomatic stages; however, cancer is often detected a few years later when it has reached an advanced stage. Therefore, detecting early CRC is vital to reduce mortality at a population level.2Kanthan R. Senger J.L. Kanthan S.C. Fecal molecular markers for colorectal cancer screening.Gastroenterol Res Pract. 2012; 2012: 184343Crossref PubMed Scopus (26) Google Scholar So far, colonoscopy is the most sensitive and specific test available. Nonetheless, it has some limitations, such as risk of complications, limited availability, need of visibly detectable lesions, manpower requirements, and probably the most important, low patient compliance for cathartic preparation and dietary restrictions, especially in healthy individuals.2Kanthan R. Senger J.L. Kanthan S.C. Fecal molecular markers for colorectal cancer screening.Gastroenterol Res Pract. 2012; 2012: 184343Crossref PubMed Scopus (26) Google Scholar Stool-based assays raise the possibility of avoiding these pitfalls. The currently used fecal immunochemical test detects occult blood in the stool, but it is also limited because bleeding from CRC is more common at late stages and can be intermittent, and moreover, other lesions can cause fecal blood, giving some false negatives and false positives and resulting in the need for repetitive testing.3Yang H. Xia B.Q. Jiang B. Wang G. Yang Y.P. Chen H. Li B.S. Xu A.G. Huang Y.B. Wang X.Y. Diagnostic value of stool DNA testing for multiple markers of colorectal cancer and advanced adenoma: a meta-analysis.Can J Gastroenterol. 2013; 27: 467-475Crossref PubMed Scopus (35) Google Scholar, 4Lee J.K. Liles E.G. Bent S. Levin T.R. Corley D.A. Accuracy of fecal immunochemical tests for colorectal cancer: systematic review and meta-analysis.Ann Intern Med. 2014; 160: 171Crossref PubMed Google Scholar On the other hand, stool testing for DNA abnormalities is potentially more attractive for population-based screening programs. This noninvasive procedure is based on tumor exfoliation: colonocytes are continuously shed into the large-bowel lumen. The number of colonocytes exfoliated from malignant lesions is greater than from normal tissue.5Davies R.J. Miller R. Coleman N. Colorectal cancer screening: prospects for molecular stool analysis.Nat Rev Cancer. 2005; 5: 199-209Crossref PubMed Scopus (193) Google Scholar The genetic alterations that are detected come from neoplastic lesions and represent the tumor cells themselves.6Bosch L.J. Carvalho B. Fijneman R.J. Jimenez C.R. Pinedo H.M. van Engeland M. Meijer G.A. Molecular tests for colorectal cancer screening.Clin Colorectal Cancer. 2011; 10: 8-23Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar Up to now, single markers or combinations of different DNA mutations in APC, KRAS, and/or TP53 genes, aberrant methylation, microsatellite instability, and/or DNA integrity have been tested for early detection of CRC with good sensitivities and specificities.6Bosch L.J. Carvalho B. Fijneman R.J. Jimenez C.R. Pinedo H.M. van Engeland M. Meijer G.A. Molecular tests for colorectal cancer screening.Clin Colorectal Cancer. 2011; 10: 8-23Abstract Full Text Full Text PDF PubMed Scopus (87) Google Scholar, 7Imperiale T.F. Ransohoff D.F. Itzkowitz S.H. Turnbull B.A. Ross M.E. Fecal DNA versus fecal occult blood for colorectal-cancer screening in an average-risk population.N Engl J Med. 2004; 351: 2704-2714Crossref PubMed Scopus (697) Google Scholar, 8Imperiale T.F. Ransohoff D.F. Itzkowitz S.H. Levin T.R. Lavin P. Lidgard G.P. Ahlquist D.A. Berger B.M. Multitarget stool DNA testing for colorectal-cancer screening.N Engl J Med. 2014; 370: 1287-1297Crossref PubMed Scopus (1038) Google Scholar Stool DNA testing has been shown to be more sensitive than the fecal occult blood test7Imperiale T.F. Ransohoff D.F. Itzkowitz S.H. Turnbull B.A. Ross M.E. Fecal DNA versus fecal occult blood for colorectal-cancer screening in an average-risk population.N Engl J Med. 2004; 351: 2704-2714Crossref PubMed Scopus (697) Google Scholar, 8Imperiale T.F. Ransohoff D.F. Itzkowitz S.H. Levin T.R. Lavin P. Lidgard G.P. Ahlquist D.A. Berger B.M. Multitarget stool DNA testing for colorectal-cancer screening.N Engl J Med. 2014; 370: 1287-1297Crossref PubMed Scopus (1038) Google Scholar and than testing on other circulating DNA, such as plasma DNA.9Ahlquist D.A. Taylor W.R. Mahoney D.W. Zou H. Domanico M. Thibodeau S.N. Boardman L.A. Berger B.M. Lidgard G.P. The stool DNA test is more accurate than the plasma septin 9 test in detecting colorectal neoplasia.Clin Gastroenterol Hepatol. 2012; 10: 272-277.e1Abstract Full Text Full Text PDF PubMed Scopus (120) Google Scholar, 10Diehl F. Schmidt K. Durkee K.H. Moore K.J. Goodman S.N. Shuber A.P. Kinzler K.W. Vogelstein B. Analysis of mutations in DNA isolated from plasma and stool of colorectal cancer patients.Gastroenterology. 2008; 135: 489-498Abstract Full Text Full Text PDF PubMed Scopus (184) Google Scholar Furthermore, a meta-analysis highlighted the diagnostic value of multiple markers over single markers.3Yang H. Xia B.Q. Jiang B. Wang G. Yang Y.P. Chen H. Li B.S. Xu A.G. Huang Y.B. Wang X.Y. Diagnostic value of stool DNA testing for multiple markers of colorectal cancer and advanced adenoma: a meta-analysis.Can J Gastroenterol. 2013; 27: 467-475Crossref PubMed Scopus (35) Google Scholar The potential of stool DNA testing is not only for CRC screening, nevertheless. It can be a helpful diagnostic and prognostic tool, and in addition, it can be used as a method for the early detection of disease relapse after initial therapy, or even as a guide for preoperative treatment selection (eg, through risk stratification on the basis of DNA mutations).11Lansdorp-Vogelaar I. Knudsen A.B. Brenner H. Cost-effectiveness of colorectal cancer screening: an overview.Best Pract Res Clin Gastroenterol. 2010; 24: 439-449Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar, 12Avital I. Langan R.C. Summers T.A. Steele S.R. Waldman S.A. Backman V. Yee J. Nissan A. Young P. Womeldorph C. Mancusco P. Mueller R. Noto K. Grundfest W. Bilchik A.J. Protic M. Daumer M. Eberhardt J. Man Y.G. Brücher B.L. Stojadinovic A. Evidence-based guidelines for precision risk stratification-based screening (PRSBS) for colorectal cancer: lessons learned from the US armed forces: consensus and future directions.J Cancer. 2013; 4: 172-192Crossref PubMed Scopus (13) Google Scholar However, for such purposes, more than three genes should be tested. Recent genetic advances, such as massive parallel sequencing, could be applied to stool DNA to search for mutations in many cancer genes simultaneously. Moreover, in the past few years, sequencing information has helped to develop targeted therapy drugs on the basis of specific molecular aberrations of neoplastic cells. Three such drugs (all monoclonal antibodies) are approved to treat CRC, mainly metastatic CRC: two of them inhibit epidermal growth factor receptor (EGFR), and the other inhibits the angiogenesis-promoting vascular endothelial growth factor. However, some CRC patients become refractory to anti-EGFR therapies because of genetic mutations in tumor cells (eg, KRAS, NRAS, and BRAF) that activate mitogen-activated protein kinase kinase and extracellular signal–regulated kinase signaling, which persists after EGFR blockade.13Misale S. Arena S. Lamba S. Siravegna G. Lallo A. Hobor S. Russo M. Buscarino M. Lazzari L. Sartore-Bianchi A. Bencardino K. Amatu A. Lauricella C. Valtorta E. Siena S. Di Nicolantonio F. Bardelli A. Blockade of EGFR and MEK intercepts heterogeneous mechanisms of acquired resistance to anti-EGFR therapies in colorectal cancer.Sci Transl Med. 2014; 6: 224ra26Crossref PubMed Scopus (219) Google Scholar Therefore, sequencing of stool DNA mutations could allow us to identify patients who may benefit or not from such alternative treatment strategies, even before surgery. In addition, stool DNA mutation analysis could help in detecting early evidence of mutations that confer acquired resistance to targeted therapies and that may arise during treatment. Interestingly, two independent genome-scale studies aimed at identifying genetic changes in CRC by next-generation sequencing (NGS) concluded that five pathways are involved in CRC development: KRAS/NRAS/BRAF, p53, transforming growth factor-β, phosphatidylinositol 3-kinase, and WNT.14Seshagiri S. Stawiski E.W. Durinck S. Modrusan Z. Storm E.E. Conboy C.B. Chaudhuri S. Guan Y. Janakiraman V. Jaiswal B.S. Guillory J. Ha C. Dijkgraaf G.J. Stinson J. Gnad F. Huntley M.A. Degenhardt J.D. Haverty P.M. Bourgon R. Wang W. Koeppen H. Gentleman R. Starr T.K. Zhang Z. Largaespada D.A. Wu T.D. de Sauvage F.J. Recurrent R-spondin fusions in colon cancer.Nature. 2012; 488: 660-664Crossref PubMed Scopus (753) Google Scholar, 15The Cancer Genome Atlas Network Comprehensive molecular characterization of human colon and rectal cancer.Nature. 2012; 487: 330-337Crossref PubMed Scopus (5933) Google Scholar Moreover, after performing a hierarchical classification of >3000 tumors from 12 tumor types on the basis of genomic alterations, Ciriello et al16Ciriello G. Miller M.L. Aksoy B.A. Senbabaoglu Y. Schultz N. Sander C. Emerging landscape of oncogenic signatures across human cancers.Nat Genet. 2013; 45: 1127-1133Crossref PubMed Scopus (905) Google Scholar observed that the oncogenic signature of CRC is dominated by point mutations over copy number changes. Therefore, using NGS technology to search for point mutations in known colon cancer genes, such as targeted sequencing by Ion Torrent technology, would be suitable for the above-mentioned purposes, and much cheaper and faster than whole genome sequencing. Two previous studies have demonstrated that the Ion AmpliSeq platform provides highly sensitive and quantitative mutation detection for cancer genes in CRC samples, even in formalin-fixed, paraffin-embedded specimens.17Zhang L. Chen L. Sah S. Latham G.J. Patel R. Song Q. Koeppen H. Tam R. Schleifman E. Mashhedi H. Chalasani S. Fu L. Sumiyoshi T. Raja R. Forrest W. Hampton G.M. Lackner M.R. Hegde P. Jia S. Profiling cancer gene mutations in clinical formalin-fixed, paraffin-embedded colorectal tumor specimens using targeted next-generation sequencing.Oncologist. 2014; 19: 336-343Crossref PubMed Scopus (48) Google Scholar, 18Malapelle U. Vigliar E. Sgariglia R. Bellevicine C. Colarossi L. Vitale D. Pallante P. Troncone G. Ion Torrent next-generation sequencing for routine identification of clinically relevant mutations in colorectal cancer patients.J Clin Pathol. 2015; 68: 64-68Crossref PubMed Scopus (69) Google Scholar Our aim was to assess the feasibility of identifying gene mutations in 22 cancer driver genes with the semiconductor-based Ion Torrent technology in stool DNA from a series of patients with CRC. We analyzed stool DNA from samples of both colon and rectum adenocarcinomas, because NGS analysis has revealed that colon and rectum cancers have similar patterns of genomic alteration.15The Cancer Genome Atlas Network Comprehensive molecular characterization of human colon and rectal cancer.Nature. 2012; 487: 330-337Crossref PubMed Scopus (5933) Google Scholar The study was performed on a series of 65 stool samples from CRC patients, obtained at the time of diagnosis: 13 had the tumor located at the rectum and 52 at the colon. Sampling was performed before any treatment. Patients were diagnosed and treated at Tehran University–affiliated Shariati Hospital (Tehran, Iran), from 2012 to 2014. The study protocol was approved by the ethics committee of Digestive Disease Research Institute at Shariati Hospital, and an informed consent was obtained from all patients before being enrolled into the study. A list of tumor samples and their clinical characteristics is shown in Supplemental Table S1. None of the patients received EGFR antibody treatment. Stool DNA samples were sequenced with Ion Torrent technology (Life Technologies, Carlsbad, CA), which has a low input DNA requirement (as little as 10 ng of DNA). For a screening of CRC mutations, we used Ion AmpliSeq Colon and Lung Cancer Panel, designed and validated by eight cancer research groups from different European institutions; moreover, its accuracy has been proved recently in CRC patients.18Malapelle U. Vigliar E. Sgariglia R. Bellevicine C. Colarossi L. Vitale D. Pallante P. Troncone G. Ion Torrent next-generation sequencing for routine identification of clinically relevant mutations in colorectal cancer patients.J Clin Pathol. 2015; 68: 64-68Crossref PubMed Scopus (69) Google Scholar DNA was isolated using the QIAamp DNA Stool Mini Kit (Qiagen GmbH, Hilden, Germany), following manufacturer's instructions. DNA concentration was assessed using Qubit 2.0 Fluorometer (Life Technologies) and the Qubit dsDNA HS Assay kit. Five samples did not show enough DNA quality to be sequenced. Therefore, deep sequencing was performed for the remaining 60 samples using the Ion Torrent platform, as follows. Genomic DNA (10 ng) was used to prepare the library with the Ion AmpliSeq Library kit 2.0 (Life Technologies) and with a primer pool to analyze 504 mutational hotspots and targeted regions (totaling 14.6 kb) in 22 genes implicated in colon and lung cancer: AKT1, ALK, BRAF, CTNNB1, DDR2, EGFR, ERBB2, ERBB4, FBXW7, FGFR1, FGFR2, FGFR3, KRAS, MAP2K1, MET, NOTCH1, NRAS, PIK3CA, PTEN, SMAD4, STK11, and TP53. Amplified products were purified with Agencourt AMPure XP beads (Beckman Coulter Genomics, High Wycombe, UK). Concentrations of amplified and barcoded libraries were measured using the Qubit 2.0 Fluorometer. After this, template preparation was performed with the Ion OneTouch 2 System (Life Technologies), an automated system for emulsion PCR, recovery of Ion Sphere Particles, and enrichment of template-positive particles. Finally, sequencing was performed using Ion 316 chips on the Ion Personal Genome Machine System (Life Technologies) and with the Ion Personal Genome Machine Sequencing 200 kit version 2. Torrent Suite software version 4.0.2 (Life Technologies) was used to assess run performance and data analysis. Of 60 samples, 52 were successfully sequenced, with a mean depth >20. Then, alignment to the hg19 human reference genome and variant calling was performed, according to the parameters set in the Torrent Suite software version 4.0.2 (Life Technologies). Only those variants with the following criteria were considered positive: variant frequency >5%, homopolymer length 15 (a minimum base call accuracy of 97%), coverage >50, strand bias <0.9, common signal shift 45. The variant frequency threshold was set at 5% to avoid false-positive results. Recent reports have demonstrated that the Ion Torrent platform is able to consistently detect mutations at the 5% level of mutant alleles, pinpointing true positive gene mutations.18Malapelle U. Vigliar E. Sgariglia R. Bellevicine C. Colarossi L. Vitale D. Pallante P. Troncone G. Ion Torrent next-generation sequencing for routine identification of clinically relevant mutations in colorectal cancer patients.J Clin Pathol. 2015; 68: 64-68Crossref PubMed Scopus (69) Google Scholar, 19Tops B. Normanno N. Kurth H. Amato E. Mafficini A. Rieber N. Le Corre D. Rachiglio A. Reiman A. Sheils O. Noppen C. Lacroix L. Cree I.A. Scarpa A. Ligtenberg M. Laurent-Puig P. Development of a semi-conductor sequencing-based panel for genotyping of colon and lung cancer by the Onconetwork consortium.BMC Cancer. 2015; 15: 26Crossref PubMed Scopus (42) Google Scholar Alignments were visually verified with the Integrative Genomics Viewer (IGV version 2.3.34; Broad Institute, Cambridge, MA). The previously identified single-nucleotide polymorphisms were determined using the National Center for Biotechnology Information's Single Nucleotide Polymorphism, HapMap, or 1000 Human Genome Project databases, and known polymorphic sites with a minor allele frequency >1% were removed from the analysis. Moreover, we filtered against synonymous and intron variants, as well as those variants that were considered to have both a neutral effect and a tolerated effect according to PROVEAN and SIFT prediction tools, respectively.20Choi Y. Sims G.E. Murphy S. Miller J.R. Chan A.P. Predicting the functional effect of amino acid substitutions and indels.PLoS One. 2012; 7: e46688Crossref PubMed Scopus (1959) Google Scholar, 21Kumar P. Henikoff S. Ng P.C. Predicting the effects of coding non-synonymous variants on protein function using the SIFT algorithm.Nat Protoc. 2009; 4: 1073-1081Crossref PubMed Scopus (5025) Google Scholar PCR was performed on DNA from the samples that showed KRAS codon 12, 13, or 20 or EGFR codon 706 or 757 mutations by NGS. Primers used and annealing temperature for PCR were as follows: reverse, 5′-AGGCCTGCTGAAAATGACTGAATATAA-3′ and forward, 5′-CAAAGAATGGTCCTGCACCAGTAATAT-3′ (52°C) for KRAS and forward, 5′-TGTGGAGCCTCTTACACCCA-3′ and reverse, 5′-GTGCCAGGGACCTTACCTTATAC-3′ (53°C) for EGFR. The PCR products were cleaned and sequenced using Big Dye Terminator version 3.1 sequencing kit (Applied Biosystems, Waltham, MA) and run on a 3130 XL Sequencer (Applied Biosystems). The NGS assay was successfully performed on 52 of 65 stool DNA samples (80% success): five samples were removed from the analysis because of bad DNA quality, and eight samples because of unsatisfactory sequencing quality. Results of gene mutation analysis showed both hotspot and novel mutations fulfilling the selection criteria. Overall, 83 mutations in cancer driver genes were identified (Supplemental Table S2). Variant coverage ranged from 50 to 7324, with a mean value of 394 and an SEM of 101. Approximately 65% (54/83) of mutations were new and had not been reported in COSMIC version 71 (http://cancer.sanger.ac.uk/cosmic, last accessed March 4, 2015).22Forbes S.A. Tang G. Bindal N. Bamford S. Dawson E. Cole C. Kok C.Y. Jia M. Ewing R. Menzies A. Teague J.W. Stratton M.R. Futreal P.A. COSMIC (the Catalogue of Somatic Mutations In Cancer): a resource to investigate acquired mutations in human cancer.Nucleic Acids Res. 2010; 38: D652-D657Crossref PubMed Scopus (446) Google Scholar All mutations were protein-altering single-nucleotide substitutions: 87% transitions and 13% transversions. Only nine mutations corresponded to nonsense mutations, whereas all of the others were missense mutations. The NGS analysis revealed that one to five genes were mutated in 39 of 52 cases (75%), whereas no mutations in the 22 tested genes were found in the 13 remaining cases (Supplemental Table S3). Among the 22 genes, mutations were found in 18 genes (all except ALK, PIK3CA, NOTCH1, and FGFR1). There were some cases with two or even three different coexisting mutations in the same gene (ie, TP53, KRAS, SMAD4, ERBB2, FGFR3, MET, and MAP2K1). No correlation could be found between the number and type of mutations with tumor stage, site of tumor (colon/rectum), or patient overall survival. Although quality control was performed to prevent false-positive results, we performed some validation experiments to check how the results were compared with Sanger sequencing. We selected samples with KRAS mutations in exon 2 (codons 12, 13, and 20) and samples with EGFR mutations (codons 706 and 757) for validation by PCR and Sanger sequencing (Table 1). Only those samples with a good PCR amplification product were sequenced. All samples that had mutant allele frequency >10% were confirmed by Sanger sequencing (it is known that Sanger sequencing is not sensitive enough to detect low-frequency mutations with a sensitivity threshold of approximately 10%23Tsiatis A.C. Norris-Kirby A. Rich R.G. Hafez M.J. Gocke C.D. Eshleman J.R. Murphy K.M. Comparison of Sanger sequencing, pyrosequencing, and melting curve analysis for the detection of KRAS mutations: diagnostic and clinical implications.J Mol Diagn. 2010; 12: 425-432Abstract Full Text Full Text PDF PubMed Scopus (368) Google Scholar). Figure 1 shows the mutations detected by NGS and Sanger sequencing. Moreover, we could perform the sequencing analysis of a tumor formalin-fixed, paraffin-embedded sample from case 44, and we were able to identify the same mutations that were present in the corresponding stool sample (both in KRAS and TP53 genes) (Figure 2). Therefore, mutations identified in the present study by NGS are most probably bona fide mutations, and not false-positive mutations.Table 1Validation of NGS-Identified Mutations (at Specific Codons of KRAS and EGFR Genes) by PCR and Sanger SequencingSampleMutation by NGSMutant allele frequency, %PCR band qualityMutation by Sanger sequencing9KRAS p.G13D11GoodKRAS p.G13D27KRAS p.T20M8BadNot sequenced32KRAS p.G12D9BadNot sequenced44KRAS p.G12A24GoodKRAS p.G12A50KRAS p.G12D26GoodKRAS p.G12D52KRAS p.G12S5BadNot sequenced56KRAS p.G12S & p.E31K7GoodNo mutation65KRAS p.G13D & p.Q25*7 and 8GoodNo mutation11EGFR p.I706-576-653-661T7BadNot sequenced34EGFR p.K757-627-704-712E11BadNot sequenced59EGFR p.I706-576-653-661T8GoodNo mutationNGS, next-generation sequencing. Open table in a new tab Figure 2KRAS (A and C) and TP53 (B and D) mutations observed by next-generation sequencing (NGS) in a tumorous sample (A and B) and the corresponding stool sample (C and D) from case 44. The two NGS-identified mutations (p.G12A in KRAS and p.G244A in TP53) in stool from case 44 are also detected in the corresponding tumoral tissue from the same patient.View Large Image Figure ViewerDownload Hi-res image Download (PPT) NGS, next-generation sequencing. The most frequently mutated genes were TP53, KRAS, FBXW7, EGFR, and SMAD4, all of them with at least 10% of samples with mutations. Unfortunately, the AmpliSeq Cancer Panel used in the present study does not include the APC gene or the ATM gene, and therefore mutations in these genes, which are frequent in CRC, could not be analyzed. TP53 gene mutations occurred in 13 of 52 cases (25%); most of them (nine) were hotspot mutations at the DNA binding domain (codons 94 to 292), and one was a nonsense mutation at the oligomerization domain (codon 342). The other three were novel mutations spread across all TP53 domains. KRAS gene was mutated in 12 of 52 cases (23%). Surprisingly, KRAS hotspot mutations in exon 2 (codon 12 or 13) were identified in only 7 of these 12 samples (58%). Two patients had hotspot mutations in other codons (eg, 117 or 146, both in exon 4), and the remaining three had novel mutations in exons 2 and 3 (codons 20, 43, and 63). Mutations in FBXW7 were detected in 9 of 52 cases (17%). Three FBXW7 mutations were hotspot mutations, and the other novel mutations. Moreover, eight of the FBXW7 mutations were located at the WD repeat domain. EGFR gene was mutated in eight samples (15% of cases), two of them with hotspot mutations. Interestingly, seven of these eight mutations were located at the tyrosine kinase domain. SMAD4 gene mutations were all novel and occurred in 7 of 52 cases (13%), one of them at the MH1 domain, five at the MH2 domain, and one at the linker between these two domains. Less frequent mutations (4% to 8% of cases) were found in the other seven genes [ie, PTEN, BRAF (different from V600E), ERBB2, FGFR3, MET, DDR2, and CTNNB1]. Mutations in the ERBB4, FGFR2, STK11, AKT1, MAP2K1, and NRAS genes occurred rarely (only one case each). Of the 39 samples with mutations, 21 (54%) presented concomitant mutations in different genes. In some of the cases with concomitant mutations, no big differences were found in the variant frequency for the distinct genes, whereas in other cases variant frequencies were different (Supplemental Table S2). For example, case 9 had 15% of alleles with a TP53 p.C176Y mutation coexisting with a 12% of KRAS p.G13D mutation. On the other hand, for example, in case 22 the KRAS variant frequency was 6% and the TP53 variant frequency was 23%. Moreover, the analysis revealed that 12 of the 18 cases (67%) with only one mutation harbored mutations in one of the most mutated genes in our series: TP53, KRAS, FBXW7, EGFR, or SMAD4 genes. Table 2 shows the concomitant mutations in every case ordered by pathways affected. Interestingly, the combined phosphatidylinositol 3-kinase/mitogen-activated protein kinase pathway was altered in approximately 70% of samples. Among these, we found that more than half (58%) had alterations in KRAS, NRAS, or BRAF, with a pattern of mutual exclusivity (only one sample harbored mutations in KRAS and BRAF simultaneously). Characteristics of mutations in these genes (with probable predictive value for anti-EGFR treatment) are summarized in Table 3.Table 2Mutation Pattern in the Pathways Typically Altered in CRC for the 39 Samples with MutationsSample16791011151819212224272830323334353739414446474950515253545556575859606365Combined PI3K and MAPK pathway KRAS●●●●●●●●●●●● NRAS● BRAF●●● EGFR●●●●●●●● ERBB2●●● ERBB4● FGFR2● FGFR3●● MET●● PTEN●●●● ALK PIK3CA MAP2K1● AKT1● Counts010111112010120202120010111010112123005WNT pathway FBXW7●●●●●●●●● CTNNB1●● Counts101000000120010000001001001010000100000TGF-β pathway SMAD4●●●●●●●TP53 pathway TP53●●●●●●●●●●●●●For each sample, it is shown with a black circle if at least one mutation was identified for that particular gene. For the sake of clarity, number of genes mutated per sample were scored (counts) in the two pathways with more than one gene analyzed.CRC, colorectal cancer; MAPK, mitogen-activated protein kinase; PI3K, phosphatidylinositol 3-kinase; TGF-β, transforming growth factor-β. Open table in a new tab Table 3Characteristics of Mutations in Genes Predictive of EGFR Antibody Treatment Identified from Stool DNA of Colorectal Cancer PatientsGeneExonMutationCOSMIC ID/novelResistant/sensitive mutationNo. of samplesKRAS2p.G12ACOSM522Resistant12p.G12DCOSM521Resistant22p.G12SCOSM517Resistan
Referência(s)