- Email: [email protected]
Application of next-generation sequencing in gastrointestinal and liver tumors
Cancer Letters 374 (2016) 187–191
Contents lists available at ScienceDirect
Cancer Letters j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c a n l e t
Mini-review
Application of next-generation sequencing in gastrointestinal and liver tumors Sameh Mikhail a,*, Bishoy Faltas b, Mohamed E. Salem c, Tanios Bekaii-Saab a a b c
Ohio State University Comprehensive Cancer Center–James Cancer Hospital and Solove Research Institute, Columbus, OH 43221, USA Weill Cornell Medical College, Washington, DC, USA Medstar Georgetown University Hospital-Lombardi Comprehensive Cancer Center, Washington, DC, USA
A R T I C L E
I N F O
Article history: Received 31 October 2015 Received in revised form 11 February 2016 Accepted 16 February 2016 Keywords: Next generation sequencing Clinical application Predictive markers Gastrointestinal cancer Challenges
A B S T R A C T
Malignant transformation of normal cells is associated with the evolution of genomic alterations. This concept has led to the development of molecular testing platforms to identify genomic alterations that can be targeted with novel therapies. Next generation sequencing (NGS) has heralded a new era in precision medicine in which tumor genes can be studied efficiently. Recent developments in NGS have allowed investigators to identify genomic predictive makers and hereditary mutations to guide treatment decision. The application of NGS in gastrointestinal cancers is being extensively studied but continues to face substantial challenges. In our review, we discuss various NGS platforms and highlight their role in identifying familial mutations and markers of response or resistance to cancer therapy. We also provide a balanced discussion of the challenges that limit the routine use of NGS in clinical practice. © 2016 Published by Elsevier Ireland Ltd.
Introduction
Molecular testing platforms
Malignant transformation of normal cells is associated with the evolution of genomic alterations. This finding has led to the development of molecular testing platforms to find “Achilles heel” that can be targeted with novel molecularly-directed therapies. Molecular testing techniques have, therefore, evolved over the years and slowly transitioned from being a research concept to a modality that is readily available in routine clinic practice. The advent of next generation sequencing (NGS) has heralded a new era in clinical genomics. Several massive parallel sequencing approaches have dramatically decreased the cost of sequencing over the last decade [1]. These technologies resulted in a tremendous increase in throughput by sequencing millions of DNA fragments in parallel. The evolution of molecular diagnostic techniques paralleled by advances in drug development has translated into significant improvements in outcomes of certain patient subgroups. Notable examples include the introduction of ALK inhibitors in patients with lung cancer and ALK rearrangement [2], epidermal growth factor receptor (EGFR) inhibitors in patients with lung adenocarcinoma and EGFR mutations [3,4], or BRAF inhibitors in patients with melanoma who harbor a BRAFv600E mutation [5]. These advances have not been uniformly seen in all cancer types. Herein, we discuss the application of NGS in gastrointestinal and liver cancers and highlight recent advances, ongoing challenges and future directions.
The application of NGS in clinical settings requires careful interpretation of the output of NGS data in the context of “actionability”. It is prudent to understand that the definition of “actionable” molecular alterations is based on several molecular, patientspecific and practical factors such as the availability of appropriate clinical trials or standard therapies. It is also critical to understand the performance characteristics and analytical validity of each sequencing method as well as the details of each variant considered as a basis for further clinical action. There are several sequencing platforms and protocols (Table 1), but the basic steps start with library preparation, which usually involves fragmentation of nucleic acids into small fragments that are subsequently amplified and “bar-coded”. The final actual “sequencing” step provides an output that is encoded by changes in fluorescent labels captured by an ultrasensitive camera or changes in pH captured by an ion-sensitive detector [6]. These data are then decoded into “reads” of nucleotide sequences strung together. These reads are subsequently aligned and mapped to their respective reference genomic regions. Variations from the reference genome can then be detected; this so called “variant calling” is achieved using computational algorithms that factor in the probability of each variant being a true variant based on the known sequencing errors and polymorphisms. Called variants have to be annotated afterward in order to infer the potential for functional significance. Several factors need to be considered for determining the potential functional impact of a mutation. These include the prevalence of a particular variant in databases such as COSMIC (Catalogue of Somatic
* Corresponding author. Tel.: +614 293 9863; fax: +614 293 7520. E-mail address: [email protected] (S. Mikhail). http://dx.doi.org/10.1016/j.canlet.2016.02.029 0304-3835/© 2016 Published by Elsevier Ireland Ltd.
188
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
Table 1 Next generation sequencing technologies. NGS method
Detected molecular alterations
Description
Whole sequencing Targeted sequencing
Single nucleotide variants, Indels, Somatic copy number alterations Single nucleotide variants, Indels, and Somatic copy number alterations Single nucleotide variants, Indels, somatic copy number alterations, translocations and chromosomal rearrangements Expression levels of mRNA of different genes, detection of fusion transcripts from translocations, identification of splice variants and non-coding RNA DNA methylation patterns
Sequencing of all protein coding regions (exons) of all genes Sequencing of exons of pre-selected list of genes; usually assembled in a “panel” Sequencing of all regions of the genome including exonic, intronic and intergenic regions Sequencing provides a “snapshot” of the RNA transcribed from the genome at a give time point
Whole genome sequencing RNA sequencing
Bisulfite sequencing: Methylseq and Reduced representation bisulfite sequencing ChIP-Seq
16S Ribosomal RNA sequencing
Identifying binding sites of DNA-associated proteins
Bacterial phylogenetic classification for microbiome studies
Mutations in Cancer) and the location of a variant in reference to coding genes and its predicted function on the corresponding functional protein domains and three-dimensional structures [7]. The role of NGS in identifying predictive markers for response to cancer therapy Several large projects such as The Cancer Genome Atlas (TCGA) have provided insight into genomic alterations that are prevalent in gastrointestinal (GI) cancers (Table 2) and can potentially be targeted through novel therapies [8–12]. Notable examples of progress that has resulted in identifying such genomic targets include treating gastrointestinal stromal tumors with c-kit mutations with imatinib [13] and gastric cancer with HER2 amplification with trastuzumab (Table 3). Targeted therapies offer the potential for reducing side effects and potentially improving outcomes of treatment of GI cancer that harbor targetable alterations. Determining the feasibility NGS is a multifaceted process that involves evaluating its applicability to routine clinical practice, its yield, availability of targeted therapies and impact on patient outcomes. Several studies have, therefore, been launched and/or completed to further evaluate the feasibility of NGS in identifying patients with GI and other cancers who may benefit from targeted therapies [14–16]. A recent study has demonstrated that NGS is feasible in gastroesophageal cancer in clinical practice [14]. Eighty nine percent (50/56) of patients undergoing NGS had at least one actionable molecular alteration. The most prevalent alterations included cell cycle abnormalities (58%), HER2 amplification (30%), PI3KCA mutations (14%), MCL1 amplification (11%), PTEN loss (9%) and MET amplification (5%). These results are intriguing as several cell cycle inhibitors are currently in clinical development [17,18]. It is important to note, however, that NGS should not replace standard immunohistochem-
Utilizes the bisulfite reaction which converts cytosine residues to uracil while leaving 5-methylcytosine residues unconverted
ChIP (chromatin-immunoprecipitation) enriches cross-linked DNA–protein complexes selected using an antibody against the protein of interest Based on sequencing of hypervariable regions in the 16S to identify bacteria
istry (IHC) and fluorescent in situ hybridization (FISH) techniques for HER2 testing in gastroesophageal cancer. Only 12/18 (66%) patients, positive for HER2 by IHC and/or FISH, demonstrated HER 2 amplification by NGS. These results suggest that NGS should be added to IHC/FISH testing for HER2 overexpression or amplification rather than replace it. Recently, investigators from the University of Texas MD Anderson Cancer Center published their experience in evaluating the use of NGS to facilitate enrollment onto genomicallymatched clinical trials [16]. Their cohort included 2000 patients, of whom 19% had gastrointestinal malignancy. The prevalence of actionable alterations including KRAS was 79%, 67% and 16% in pancreas, colorectal and gastroesophageal cancers, respectively. When KRAS mutations were excluded, the prevalence of actionable mutations was 31%, 16%, 16% and 11% in colorectal, esophageal, pancreatic and stomach cancers, respectively. In the whole cohort, 789 (39%) patients had at least one actionable mutation. However, only 83 patients (11% of those with actionable mutations; 4% of the total cohort) were enrolled in genotype-matched trials. The median time from consent to obtaining genomic test results was 26 days. The authors cited patient preference for local or non-investigational treatment, poor performance status, lack of trials or trial slots, insurance denial and/or trial ineligibility as the main challenges to clinical trial enrollment. This study is extremely valuable in understanding the current landscape of genomic testing and clinical trial enrollment and provides a platform to address the major challenges to the application of widespread genomic testing in cancer treatment. This study, however, did not report the outcome of patients who received targeted therapy. The outcome of patients with actionable molecular alterations that received targeted therapies has, however, been studies in the SHIVA trial [15]. The SHIVA trial is a randomized controlled open-label phase 2 trial that was launched in eight French academic centers to evaluate the efficacy of 11
Table 2 Notable molecular alterations in gastrointestinal cancers. Tumor
Notable molecular alterations
References
Gastroesophageal
TP53, CDKN2A, CCNE, CDK 4/6, CCND, EGFR, ERBB2, ERBB3, PI3KCA, PIK3R1, MCL1, PTEN, CDH1, JAK2, PD-L1/2, VEGFA, KRAS/NRAS and MET IDH1, CDH1, KIT, FGFR2, FLT3, NPM1, PTEN, MET, AKT1, RET, NOTCH1 ERBB4, ERBB2, KRRAS, BRAF and FBXW7 APC, TP53, SMAD4, PIK3CA, PIK3R1, PTEN, KRAS, NRAS, BRAF, ARID1A, SOX9, FAM123B/WTX, ERBB2, IGF2 NAV2/TCF7L, ACVR2A, APC, TGFBR2, MLH1, MSH3, MSH6, SLC9A9, SMAD2, SMAD3 and SMAD4 TP53, SMAD4, CDKN2A, ARID1A, ROBO2, KDM6A, PREX2, ERBB2, MET, FGFR1, CDK6, PIK3R3, PIK3CA, BRCA1, BRCA2, PALB2, KRAS, TGFBR2, BRAF, PREX2, MLL2, MLH2 and MLH2 TP53, KRAS, ERBB3, EGFR, ERBB2, ERBB4, FGFR2, PRKACA and PRKACB TP53, CTNNB1, MET, PTEN, CDKN2A, AXIN1, PTEN, PIK3CA, KRAS, NRAS, MYC, MET, CCND2, RB1, ARID1A, ARID1B, ARID2, IRF2, NFE2L2, ERRFl1, RPSKA3 and MLL3
[14,37]
Small bowel Colorectal Pancreas Biliary Hepatocellular
[38,39] [40] [8] [9,10] [41,42]
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
189
Table 3 Frequency of common (≥5) actionable molecular alterations and potential targeted therapy. Tumor
molecular alterations
Frequency (%)
Therapy
References
Gastroesophageal
ERBB2 CDKN2A, CCNE, CDK 4/6, CCND PIK3CA PTEN loss MET KRAS PIK3CA ERBB2 BRAF FBXW7 PTEN(loss) PIK3CA FBXW7 BRAF CDKN2A BRCA2 FANCC KRAS PIK3CA BRAF CCND1 FGFR2 FGFR1 FGFR3 CDKN2A PTEN MET
15–30 58 14 9 5 40 10 8 6 6 34 15–20 14 8 95 7–10 6 18 7 6 6 5 4 4 10–60 5–10 1–5
Trastuzumab CDK 4/6 inhibitors PIK3CA, AKT, MTOR inhibitors PIK3CA, AKT, MTOR inhibitors MET inhibitors MEK, RAF, AKT inhibitors
[14,17,43]
PIK3CA, AKT, MTOR inhibitors
[33,44–47]
Small bowel
Colorectal
Pancreas
Biliary
Hepatocellular
molecularly targeted agents versus investigator’s choice of conventional therapy when used in patients with advanced cancer who have failed standard therapies and underwent fresh tumor biopsies and NGS. At the time of data cut off, the study had enrolled 195 patients (99 in the experimental arm and 96 in the control arm), of whom 17% had a gastrointestinal malignancy. Median progression free survival (PFS) was 2.3 months (95% CI 1.7–3.8) in the experimental group versus 2.0 months (1.8–2.1) in the control group (hazard ratio 0.88, 95% CI 0.65–1.19, p = 0.41). Of note, there was no statistically significant difference in the rate of grade 3/4 adverse events between the two groups. These results were disappointing and suggest that further research is needed to investigate the utility of genomic predictive markers and targeted agents that may be effective as single agents or in combination against those markers. Taken together, these data suggest that NGS can be routinely performed in clinical practice. Although actionable mutations are identified in a significant number of patients with cancer and gastrointestinal malignancies, the majority of patients, however, do not receive molecularly directed therapies on the basis of NGS results. Available trials have yet to show improved outcomes with NGS and molecularly directed therapies as compared to conventional treatment. This field is, however, progressing rapidly. As molecular testing techniques evolve and our understanding of genomic alterations and molecular pathways develops further, we suspect that NGS will play an integral role of linking patients to appropriate targeted therapies and improving outcomes compared to conventional therapy. The role of NGS in identifying predictive markers for resistance to cancer therapy Genomic testing is used to identify molecular alterations that confer resistance to targeted therapy. The most notable example is identification of RAS mutations to determine the benefit from EGFR inhibitors [19]. Two recent studies reported by Van Cutsem et al. [20] and Douillard et al. [21], respectively, demonstrated that extended RAS testing identifies more patients that are resistant to EGFR inhibitors compared to only KRAS exon 2 (codons 12 and 13) testing. The study by Doulliard et al. [21] included testing for KRAS and NRAS
MTOR inhibitors RAF, MEK, AKT inhibitors CDK 4/6 inhibitors PARP inhibitors PARP inhibitors MEK, RAF, AKT inhibitors PIK3CA inhibitors BRAF inhibitors CDK 4/6 inhibitors FGFR inhibitors FGFR inhibitors FGFR inhibitors CDK 4/6 inhibitors PIK3CA, AKT, MTOR inhibitors MET inhibitors
[39]
[8,17,48]
[9,10,17,34,49]
[50–54]
exons 2 (codons 12 and 13), 3 (codon 61) and 4 (codons 117 and 146), while the study by Van Cutsem et al. [20] involved testing for KRAS exon 3 (codons 59 and 61), KRAS exon 4 (codons 117 and 146), NRAS exon 2 (codons 12 and 13), NRAS exon 3 (codons 59 and 61), and NRAS exon 4 (codons 117 and 146). Other alterations such as BRAF, PIK3CA exon 20 mutations, PTEN loss and MET amplification may also be associated with resistance to EGFR inhibitors [22–24]. NGS has emerged as a practical platform to study these mutations and their effect on outcomes of EGFR therapy [25]. It remains a matter of debate which NGS technology is most feasible in detecting such mutations in clinical practice. A study by Guedes et al. compared direct sequencing of KRAS codons 12 and 13 (sensitivity = 20%) with newer techniques, with sensitivities of up to 0.1% such as high resolution melting (HRM) analysis followed by sequencing for positive cases [26]. The incidence of KRAS mutations increased from 45% by direct sequencing to 60% by HRM followed by sequencing. When other mutations that may confer resistance to KRAS such as BRAF, PI3K and PTEN were added to the analysis, up to 87% of non-responders to anti-EGFR were found to have mutations that may have predicted their resistance to anti-EGFR therapy. These results, however, are hypothesis generating and need to be confirmed in larger studies. Additionally, it is important to note that other mutations not included in this study may also influence the response of patients to anti-EGFR therapy. Similarly, MET amplification and PTEN loss may constitute a mechanism of resistance to HER 2 directed therapy in gastroesophageal cancer [27,28]. The widespread availability of NGS is currently playing a role toward improving our knowledge of the impact of such molecular alterations and in further personalization of treatment for patients with GI cancers [14]. The role of NGS in identifying familial gastrointestinal cancers Approximately 5% of patients with CRC have identifiable familial CRC syndromes [29]. Additionally up to 30% of patients diagnosed with CRC have evidence of a familial component [29]. The most frequent CRC syndromes are: (1) Familial adenomatous polyposis (FAP) associated with adenomatous polyposis (APC) gene mutations, (2)
190
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
polyposis associated with MUTYH gene mutations and (3) hereditary nonpolyposis CRC associated with alterations in genes responsible for DNA mismatch repair such as MLH1, MSH2, MSH6 or PMS2. NGS represents a viable option to identify molecular alterations associated with familial CRC syndromes. When compared to Sanger sequencing, however, approximately 4% of genomic alterations may be missed by NGS but detected by Sanger sequencing. A study by Simbolo et al. [29] suggested that sequences close or within homopolymetric stretches of DNA represent a group of genetic mutations that may better characterized by Sanger sequencing versus NGS. Given the lower cost of NGS and its ability to simultaneously analyze multiple genes in multiple samples, it may be feasible to adopt an approach of using NGS for routine diagnosis of familial CRC syndromes and limiting the use of costly Sanger sequencing techniques to regions that are not adequately covered by NGS. Similarly, in gastroesophageal cancers, NGS is a feasible technique to uncover mutations such as CDH1 or BRCA2 mutations to identify patients that need further testing for germline mutations [14,30–32]. Challenges to the routine use of NGS in clinical practice The routine use of NGS in clinical practice continues to face several challenges. Although NGS techniques may be less costly than other molecular diagnostic techniques, their cost effectiveness remains uncertain in patients with GI cancers. A large number of patients who undergo NGS testing do not receive molecularly directed therapy [16]. Challenges that limit the clinical utility of NGS testing include the long turnaround time (2–4 weeks) of NGS, rapid decline of patients’ performance status as NGS is being performed, lack of actionable mutations, lack of available clinical trials or patient refusal to enroll in clinical trials. Additionally it remains unclear if molecularly directed therapy that is used based on data obtained from NGS is associated with improved outcomes [15]. Altered genes detected by NGS may not represent driver mutations that would result in meaningful responses to targeted therapy. This case was clearly demonstrated when vemurafenib, a proven effective therapy in BRAF mutated melanoma [5], was evaluated in a phase II trial in patients with CRC and a BRAFV600 mutation [33]. Treatment responses were significantly less impressive in CRC compared to those observed in patients with melanoma that harbor the same mutation and were treated with the same agent. This finding indicates that MAPK/BRAF pathway may contain alternate feedback and resistance loops in CRC when compared to melanoma [33]. Another challenge that limits clinical utility of NGS and other molecular testing techniques is tumor heterogeneity [34,35]. Heterogeneity may be related to different subclones within the same tumor or between primary and metastatic tumor deposits. Such heterogeneity may result in differential treatment responses and resistance mechanisms depending on selective pressures in each tumor deposit. This constitutes both a diagnostic and a therapeutic challenge. Further research is needed to better understand the mechanism of genomic heterogeneity and identify diagnostic and therapeutic strategies to overcome it. The concordance of NGS testing and other molecular diagnostic techniques also remains undefined [14,29]. It is important to carefully study if NGS should complement or replace other testing modalities such as IHC or FISH. It appears, currently, that NGS may have inferior sensitivity to the aforementioned techniques [14]. However, we believe that NGS technologies will continue to develop and will eventually replace other molecular testing techniques. Therefore, with all the current limitations of NGS, cost effectiveness studies are needed to better evaluate the economic impact and feasibility of more widespread use of molecular testing and targeted therapy in the treatment of patients with GI cancers. Overall, we suspect that further refinement of NGS technologies will result in lower cost, faster turnaround times for testing and better characterization of molecular abnor-
malities and consequently will improve its clinical utility, cost effectiveness and use in routine clinical practice. Future directions Research in NGS requires a commitment to providing funding and philanthropic support to continue the progress that has been already achieved and help transition NGS from the bench to the bedside. Improving turnaround time is crucial to provide clinical data in a timely fashion to guide clinical decisions. We believe that repeating biopsies to detect molecular evolution of tumors may be necessary. This approach will be facilitated by advances such as testing circulating tumor cells or circulating tumor DNA by NGS without obtaining a tissue biopsy. Several issues will need to be resolved such as the interplay between tumor histology and the various genomic alterations. This will require accurate characterization of feedback and resistance loops of the various molecular abnormalities based on tissue histology and better study their response to targeted therapy. Such progress can be potentially made through clinical trials with adaptive designs that aim to identify mutations and tumor types that have exceptional responses to targeted therapies. The National Cancer institute MATCH trial [36] represents a step in this direction and more trials are likely to follow to help define the role of NGS in clinical oncology. The advances in molecular diagnostics will not result in meaningful clinical benefits unless they are paralleled with advances in developing novel therapeutics that are both clinically effective and safe. Although the challenges remain substantial, the progress that has been made is enormous and the opportunity to continue this progress has never been greater. Conflict of interest The authors report no relevant conflicts of interest. References [1] E.L. van Dijk, H. Auger, Y. Jaszczyszyn, C. Thermes, Ten years of next-generation sequencing technology, Trends Genet. 30 (2014) 418–426. [2] A.T. Shaw, D.W. Kim, K. Nakagawa, T. Seto, L. Crino, M.J. Ahn, et al., Crizotinib versus chemotherapy in advanced ALK-positive lung cancer, N. Engl. J. Med. 368 (2013) 2385–2394. [3] M. Fukuoka, Y.L. Wu, S. Thongprasert, P. Sunpaweravong, S.S. Leong, V. Sriuranpong, et al., Biomarker analyses and final overall survival results from a phase III, randomized, open-label, first-line study of gefitinib versus carboplatin/paclitaxel in clinically selected patients with advanced non-smallcell lung cancer in Asia (IPASS), J. Clin. Oncol. 29 (2011) 2866–2874. [4] C. Zhou, Y.L. Wu, G. Chen, J. Feng, X.Q. Liu, C. Wang, et al., Final overall survival results from a randomised, phase III study of erlotinib versus chemotherapy as first-line treatment of EGFR mutation-positive advanced non-small-cell lung cancer (OPTIMAL, CTONG-0802), Ann. Oncol. 26 (2015) 1877–1883. [5] J.A. Sosman, K.B. Kim, L. Schuchter, R. Gonzalez, A.C. Pavlick, J.S. Weber, et al., Survival in BRAF V600-mutant advanced melanoma treated with vemurafenib, N. Engl. J. Med. 366 (2012) 707–714. [6] B. Tran, J.E. Dancey, S. Kamel-Reid, J.D. McPherson, P.L. Bedard, A.M. Brown, et al., Cancer genomics: technology, discovery, and translation, J. Clin. Oncol. 30 (2012) 647–660. [7] E.M. Van Allen, N. Wagle, M.A. Levy, Clinical analysis and interpretation of cancer genome data, J. Clin. Oncol. 31 (2013) 1825–1833. [8] N. Waddell, M. Pajic, A.-M. Patch, D.K. Chang, K.S. Kassahn, P. Bailey, et al., Whole genomes redefine the mutational landscape of pancreatic cancer, Nature 518 (2015) 495–501. [9] M. Li, Z. Zhang, X. Li, J. Ye, X. Wu, Z. Tan, et al., Whole-exome and targeted gene sequencing of gallbladder carcinoma identifies recurrent mutations in the ErbB pathway, Nat. Genet. 46 (2014) 872–876. [10] H. Nakamura, Y. Arai, Y. Totoki, T. Shirota, A. Elzawahry, M. Kato, et al., Genomic spectra of biliary tract cancer, Nat. Genet. 47 (2015) 1003–1010. [11] J.C. Haan, M. Labots, C. Rausch, M. Koopman, J. Tol, L.J.M. Mekenkamp, et al., Genomic landscape of metastatic colorectal cancer, Nat. Commun. 5 (2014). [12] B.M. Wolpin, C. Rizzato, P. Kraft, C. Kooperberg, G.M. Petersen, Z. Wang, et al., Genome-wide association study identifies multiple susceptibility loci for pancreatic cancer, Nat. Genet. 46 (2014) 994–1000. [13] M.C. Heinrich, C.L. Corless, G.D. Demetri, C.D. Blanke, M. von Mehren, H. Joensuu, et al., Kinase mutations and imatinib response in patients with metastatic gastrointestinal stromal tumor, J. Clin. Oncol. 21 (2003) 4342–4349.
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
[14] S. Mikhail, K. Ciombor, A. Noonan, C. Wu, R. Goldberg, W. Zhao, et al., Upfront molecular testing in patients with advanced gastro-esophageal cancer: Is it time yet?, Oncotarget 6 (2015) 22206–22213. [15] C. Le Tourneau, J.P. Delord, A. Goncalves, C. Gavoille, C. Dubot, N. Isambert, et al., Molecularly targeted therapy based on tumour molecular profiling versus conventional therapy for advanced cancer (SHIVA): a multicentre, open-label, proof-of-concept, randomised, controlled phase 2 trial, Lancet Oncol. 16 (2015) 1324–1334. [16] F. Meric-Bernstam, L. Brusco, K. Shaw, C. Horombe, S. Kopetz, M.A. Davies, et al., Feasibility of large-scale genomic testing to facilitate enrollment onto genomically matched clinical trials, J. Clin. Oncol. 33 (2015) 2753–2762. [17] S. Mikhail, C. Albanese, M.J. Pishvaian, Cyclin-dependent kinase inhibitors and the treatment of gastrointestinal cancers, Am. J. Pathol. 185 (2015) 1185–1197. [18] P.P.-S.M. Gopalan, A. Chiappori, A. Ivey, A. Godillo-Villegas, F. Kaye, A phase II clinical trial of CDK 4/6 inhibitor palbociblib (PD 0332991) in previously treated, advanced non-small cell lung cancer patients with inactivated CDKN2A, in: ASCO Annual Meeting, Chicago, IL, 2014. [19] C.J. Allegra, R.B. Rumble, S.R. Hamilton, P.B. Mangu, N. Roach, A. Hantel, et al., Extended RAS gene mutation testing in metastatic colorectal carcinoma to predict response to anti-epidermal growth factor receptor monoclonal antibody therapy: American Society of Clinical Oncology provisional clinical opinion update 2015, J. Clin. Oncol. 34 (2016) 179–185. [20] E. Van Cutsem, H.J. Lenz, C.H. Kohne, V. Heinemann, S. Tejpar, I. Melezinek, et al., Fluorouracil, leucovorin, and irinotecan plus cetuximab treatment and RAS mutations in colorectal cancer, J. Clin. Oncol. 33 (2015) 692–700. [21] J.Y. Douillard, K.S. Oliner, S. Siena, J. Tabernero, R. Burkes, M. Barugel, et al., Panitumumab-FOLFOX4 treatment and RAS mutations in colorectal cancer, N. Engl. J. Med. 369 (2013) 1023–1034. [22] L. Barault, N. Veyrie, V. Jooste, D. Lecorre, C. Chapusot, J.M. Ferraz, et al., Mutations in the RAS-MAPK, PI(3)K (phosphatidylinositol-3-OH kinase) signaling network correlate with poor survival in a population-based series of colon cancers, Int. J. Cancer 122 (2008) 2255. [23] A. Bardelli, S. Corso, A. Bertotti, S. Hobor, E. Valtorta, G. Siravegna, et al., Amplification of the MET receptor drives resistance to anti-EGFR therapies in colorectal cancer, Cancer Discov. 3 (2013) 658. [24] W. De Roock, B. Claes, D. Bernasconi, J. De Schutter, B. Biesmans, G. Fountzilas, et al., Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy of cetuximab plus chemotherapy in chemotherapy-refractory metastatic colorectal cancer: a retrospective consortium analysis, Lancet Oncol. 11 (2010) 753. [25] K.K. Ciombor, S. Haraldsdottir, R.M. Goldberg, How can next-generation sequencing (genomics) help us in treating colorectal cancer?, Curr. Colorectal Cancer Rep. 10 (2014) 372–379. [26] J.G. Guedes, I. Veiga, P. Rocha, P. Pinto, C. Pinto, M. Pinheiro, et al., High resolution melting analysis of KRAS, BRAF and PIK3CA in KRAS exon 2 wild-type metastatic colorectal cancer, BMC Cancer 13 (2013) 169. [27] A.K. Paulson, E.S. Linklater, B.D. Berghuis, C.A. App, L.D. Oostendorp, J.E. Paulson, et al., Graveel, MET and ERBB2 are coexpressed in ERBB2+ breast cancer and contribute to innate resistance, Mol. Cancer Res. 11 (2013) 1112–1121. [28] X. Zhang, J.S. Park, K.H. Park, K.H. Kim, M. Jung, H.C. Chung, et al., PTEN deficiency as a predictive biomarker of resistance to HER2-targeted therapy in advanced gastric cancer, Oncology 88 (2015) 76–85. [29] M. Simbolo, A. Mafficini, M. Agostini, C. Pedrazzani, C. Bedin, E.D. Urso, et al., Next-generation sequencing for genetic testing of familial colorectal cancer syndromes, Heredit. Cancer Clin. Pract. 13 (2015) 18. [30] R.C. Grant, I. Selander, A.A. Connor, S. Selvarajah, A. Borgida, L. Briollais, et al., Prevalence of germline mutations in cancer predisposition genes in patients with pancreatic cancer, Gastroenterology 148 (2015) 556–564. [31] Y. Hirotsu, H. Nakagomi, I. Sakamoto, K. Amemiya, H. Mochizuki, M. Omata, Detection of BRCA1 and BRCA2 germline mutations in Japanese population using next-generation sequencing, Mol. Genet. Genomic Med. 3 (2015) 121– 129. [32] Y. Hirotsu, H. Nakagomi, I. Sakamoto, K. Amemiya, T. Oyama, H. Mochizuki, et al., Multigene panel analysis identified germline mutations of DNA repair genes in breast and ovarian cancer, Mol. Genet. Genomic Med. 3 (2015) 459–466.
191
[33] S. Kopetz, J. Desai, E. Chan, J.R. Hecht, P.J. O’Dwyer, D. Maru, et al., Pilot study of vemurafenib in patients with metastatic BRAF-mutated colorectal cancer, J. Clin. Oncol. 33 (2015) 4032–4038. [34] R. Simon, S. Roychowdhury, Implementing personalized cancer genomics in clinical trials, Nat. Rev. Drug Discov. 12 (2013) 358–369. [35] M. Gerlinger, A.J. Rowan, S. Horswell, J. Larkin, D. Endesfelder, E. Gronroos, et al., Intratumor heterogeneity and branched evolution revealed by multiregion sequencing, N. Engl. J. Med. 366 (2012) 883–892. [36] C. McNeil, NCI-MATCH launch highlights new trial design in precision-medicine era, J. Nat. Cancer Inst. 107 (2015). [37] TCGA Network, Comprehensive molecular characterization of gastric adenocarcinoma, Nature 513 (2014) 202–209. [38] M.A. Alvi, D.G. McArt, P. Kelly, M.A. Fuchs, M. Alderdice, C.M. McCabe, et al., Comprehensive molecular pathology analysis of small bowel adenocarcinoma reveals novel targets with potential for clinical utility, Oncotarget 6 (2015) 20863–20874. [39] A. Laforest, T. Aparicio, A. Zaanan, F.P. Silva, A. Didelot, A. Desbeaux, et al., ERBB2 gene as a potential therapeutic target in small bowel adenocarcinoma, Eur. J. Cancer 50 (2014) 1740–1746. [40] Cancer Genome Atlas Network, Comprehensive molecular characterization of human colon and rectal cancer, Nature 487 (2012) 330. [41] C. Ang, C. Gamblin, S. Millis, J. Xiu, Z. Gatalica, S. Reddy, et al., Molecular characterization of 350 hepatocellular carcinomas to identify biomarker aberrations with potential novel therapeutic options, in: ASCO Annual Meeting, Chicago, IL, 2015. [42] H. Nakagawa, T. Shibata, Comprehensive genome sequencing of the liver cancer genome, Cancer Lett. 340 (2013) 234–240. [43] Y.J. Bang, E. Van Cutsem, A. Feyereislova, H.C. Chung, L. Shen, A. Sawaki, et al., Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial, Lancet 376 (2010) 687–697. [44] R.B. Corcoran, C.E. Atreya, G.S. Falchook, E.L. Kwak, D.P. Ryan, J.C. Bendell, et al., Combined BRAF and MEK inhibition with dabrafenib and trametinib in BRAF V600-mutant colorectal cancer, J. Clin. Oncol. 33 (2015) 4023–4031. [45] D.L. Jardim, J.J. Wheler, K. Hess, A.M. Tsimberidou, R. Zinner, F. Janku, et al., FBXW7 mutations in patients with advanced cancers: clinical and molecular characteristics and outcomes with mTOR inhibitors, PLoS ONE 9 (2014) e89388. [46] P.C. Lin, J.K. Lin, H.H. Lin, Y.T. Lan, C.C. Lin, S.H. Yang, et al., A comprehensive analysis of phosphatase and tensin homolog deleted on chromosome 10 (PTEN) loss in colorectal cancer, World J. Surg. Oncol. 13 (2015) 186. [47] S. Ogino, P. Lochhead, E. Giovannucci, J.A. Meyerhardt, C.S. Fuchs, A.T. Chan, Discovery of colorectal cancer PIK3CA mutation as potential predictive biomarker: power and promise of molecular pathological epidemiology, Oncogene 33 (2014) 2949–2955. [48] J.M. Winter, A. Maitra, C.J. Yeo, Genetics and pathology of pancreatic cancer, HPB 8 (2006) 324–336. [49] T. Bekaii-Saab, M.A. Phelps, X. Li, M. Saji, L. Goff, J.S. Kauh, et al., Multiinstitutional phase II study of selumetinib in patients with metastatic biliary cancers, J. Clin. Oncol. 29 (2011) 2357–2363. [50] P. Laurent-Puig, P. Legoix, O. Bluteau, J. Belghiti, D. Franco, F. Binot, et al., Genetic alterations associated with hepatocellular carcinomas define distinct pathways of hepatocarcinogenesis, Gastroenterology 120 (2001) 1763–1773. [51] P. Laurent-Puig, J. Zucman-Rossi, Genetics of hepatocellular tumors, Oncogene 25 (2006) 3778. [52] M.J. Pishvaian, G. Upadhyay, Z. Yao, Y. Kim, A. Said, J. Mendelson, et al., CDK4 is an important target for hepatocellular cancer with inactivation of TGF-beta signaling through Smad and ELF, in: 99th AACR Annual Meeting, San Diego, CA, 2008. [53] J. Zucman-Rossi, Molecular classification of hepatocellular carcinoma, Digest. Liver Dis. 42 (Suppl. 3) (2010) S235–S241. [54] J. Zucman-Rossi, P. Laurent-Puig, Genetic diversity of hepatocellular carcinomas and its potential impact on targeted therapies, Pharmacogenomics 8 (2007) 997–1003.
Contents lists available at ScienceDirect
Cancer Letters j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / c a n l e t
Mini-review
Application of next-generation sequencing in gastrointestinal and liver tumors Sameh Mikhail a,*, Bishoy Faltas b, Mohamed E. Salem c, Tanios Bekaii-Saab a a b c
Ohio State University Comprehensive Cancer Center–James Cancer Hospital and Solove Research Institute, Columbus, OH 43221, USA Weill Cornell Medical College, Washington, DC, USA Medstar Georgetown University Hospital-Lombardi Comprehensive Cancer Center, Washington, DC, USA
A R T I C L E
I N F O
Article history: Received 31 October 2015 Received in revised form 11 February 2016 Accepted 16 February 2016 Keywords: Next generation sequencing Clinical application Predictive markers Gastrointestinal cancer Challenges
A B S T R A C T
Malignant transformation of normal cells is associated with the evolution of genomic alterations. This concept has led to the development of molecular testing platforms to identify genomic alterations that can be targeted with novel therapies. Next generation sequencing (NGS) has heralded a new era in precision medicine in which tumor genes can be studied efficiently. Recent developments in NGS have allowed investigators to identify genomic predictive makers and hereditary mutations to guide treatment decision. The application of NGS in gastrointestinal cancers is being extensively studied but continues to face substantial challenges. In our review, we discuss various NGS platforms and highlight their role in identifying familial mutations and markers of response or resistance to cancer therapy. We also provide a balanced discussion of the challenges that limit the routine use of NGS in clinical practice. © 2016 Published by Elsevier Ireland Ltd.
Introduction
Molecular testing platforms
Malignant transformation of normal cells is associated with the evolution of genomic alterations. This finding has led to the development of molecular testing platforms to find “Achilles heel” that can be targeted with novel molecularly-directed therapies. Molecular testing techniques have, therefore, evolved over the years and slowly transitioned from being a research concept to a modality that is readily available in routine clinic practice. The advent of next generation sequencing (NGS) has heralded a new era in clinical genomics. Several massive parallel sequencing approaches have dramatically decreased the cost of sequencing over the last decade [1]. These technologies resulted in a tremendous increase in throughput by sequencing millions of DNA fragments in parallel. The evolution of molecular diagnostic techniques paralleled by advances in drug development has translated into significant improvements in outcomes of certain patient subgroups. Notable examples include the introduction of ALK inhibitors in patients with lung cancer and ALK rearrangement [2], epidermal growth factor receptor (EGFR) inhibitors in patients with lung adenocarcinoma and EGFR mutations [3,4], or BRAF inhibitors in patients with melanoma who harbor a BRAFv600E mutation [5]. These advances have not been uniformly seen in all cancer types. Herein, we discuss the application of NGS in gastrointestinal and liver cancers and highlight recent advances, ongoing challenges and future directions.
The application of NGS in clinical settings requires careful interpretation of the output of NGS data in the context of “actionability”. It is prudent to understand that the definition of “actionable” molecular alterations is based on several molecular, patientspecific and practical factors such as the availability of appropriate clinical trials or standard therapies. It is also critical to understand the performance characteristics and analytical validity of each sequencing method as well as the details of each variant considered as a basis for further clinical action. There are several sequencing platforms and protocols (Table 1), but the basic steps start with library preparation, which usually involves fragmentation of nucleic acids into small fragments that are subsequently amplified and “bar-coded”. The final actual “sequencing” step provides an output that is encoded by changes in fluorescent labels captured by an ultrasensitive camera or changes in pH captured by an ion-sensitive detector [6]. These data are then decoded into “reads” of nucleotide sequences strung together. These reads are subsequently aligned and mapped to their respective reference genomic regions. Variations from the reference genome can then be detected; this so called “variant calling” is achieved using computational algorithms that factor in the probability of each variant being a true variant based on the known sequencing errors and polymorphisms. Called variants have to be annotated afterward in order to infer the potential for functional significance. Several factors need to be considered for determining the potential functional impact of a mutation. These include the prevalence of a particular variant in databases such as COSMIC (Catalogue of Somatic
* Corresponding author. Tel.: +614 293 9863; fax: +614 293 7520. E-mail address: [email protected] (S. Mikhail). http://dx.doi.org/10.1016/j.canlet.2016.02.029 0304-3835/© 2016 Published by Elsevier Ireland Ltd.
188
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
Table 1 Next generation sequencing technologies. NGS method
Detected molecular alterations
Description
Whole sequencing Targeted sequencing
Single nucleotide variants, Indels, Somatic copy number alterations Single nucleotide variants, Indels, and Somatic copy number alterations Single nucleotide variants, Indels, somatic copy number alterations, translocations and chromosomal rearrangements Expression levels of mRNA of different genes, detection of fusion transcripts from translocations, identification of splice variants and non-coding RNA DNA methylation patterns
Sequencing of all protein coding regions (exons) of all genes Sequencing of exons of pre-selected list of genes; usually assembled in a “panel” Sequencing of all regions of the genome including exonic, intronic and intergenic regions Sequencing provides a “snapshot” of the RNA transcribed from the genome at a give time point
Whole genome sequencing RNA sequencing
Bisulfite sequencing: Methylseq and Reduced representation bisulfite sequencing ChIP-Seq
16S Ribosomal RNA sequencing
Identifying binding sites of DNA-associated proteins
Bacterial phylogenetic classification for microbiome studies
Mutations in Cancer) and the location of a variant in reference to coding genes and its predicted function on the corresponding functional protein domains and three-dimensional structures [7]. The role of NGS in identifying predictive markers for response to cancer therapy Several large projects such as The Cancer Genome Atlas (TCGA) have provided insight into genomic alterations that are prevalent in gastrointestinal (GI) cancers (Table 2) and can potentially be targeted through novel therapies [8–12]. Notable examples of progress that has resulted in identifying such genomic targets include treating gastrointestinal stromal tumors with c-kit mutations with imatinib [13] and gastric cancer with HER2 amplification with trastuzumab (Table 3). Targeted therapies offer the potential for reducing side effects and potentially improving outcomes of treatment of GI cancer that harbor targetable alterations. Determining the feasibility NGS is a multifaceted process that involves evaluating its applicability to routine clinical practice, its yield, availability of targeted therapies and impact on patient outcomes. Several studies have, therefore, been launched and/or completed to further evaluate the feasibility of NGS in identifying patients with GI and other cancers who may benefit from targeted therapies [14–16]. A recent study has demonstrated that NGS is feasible in gastroesophageal cancer in clinical practice [14]. Eighty nine percent (50/56) of patients undergoing NGS had at least one actionable molecular alteration. The most prevalent alterations included cell cycle abnormalities (58%), HER2 amplification (30%), PI3KCA mutations (14%), MCL1 amplification (11%), PTEN loss (9%) and MET amplification (5%). These results are intriguing as several cell cycle inhibitors are currently in clinical development [17,18]. It is important to note, however, that NGS should not replace standard immunohistochem-
Utilizes the bisulfite reaction which converts cytosine residues to uracil while leaving 5-methylcytosine residues unconverted
ChIP (chromatin-immunoprecipitation) enriches cross-linked DNA–protein complexes selected using an antibody against the protein of interest Based on sequencing of hypervariable regions in the 16S to identify bacteria
istry (IHC) and fluorescent in situ hybridization (FISH) techniques for HER2 testing in gastroesophageal cancer. Only 12/18 (66%) patients, positive for HER2 by IHC and/or FISH, demonstrated HER 2 amplification by NGS. These results suggest that NGS should be added to IHC/FISH testing for HER2 overexpression or amplification rather than replace it. Recently, investigators from the University of Texas MD Anderson Cancer Center published their experience in evaluating the use of NGS to facilitate enrollment onto genomicallymatched clinical trials [16]. Their cohort included 2000 patients, of whom 19% had gastrointestinal malignancy. The prevalence of actionable alterations including KRAS was 79%, 67% and 16% in pancreas, colorectal and gastroesophageal cancers, respectively. When KRAS mutations were excluded, the prevalence of actionable mutations was 31%, 16%, 16% and 11% in colorectal, esophageal, pancreatic and stomach cancers, respectively. In the whole cohort, 789 (39%) patients had at least one actionable mutation. However, only 83 patients (11% of those with actionable mutations; 4% of the total cohort) were enrolled in genotype-matched trials. The median time from consent to obtaining genomic test results was 26 days. The authors cited patient preference for local or non-investigational treatment, poor performance status, lack of trials or trial slots, insurance denial and/or trial ineligibility as the main challenges to clinical trial enrollment. This study is extremely valuable in understanding the current landscape of genomic testing and clinical trial enrollment and provides a platform to address the major challenges to the application of widespread genomic testing in cancer treatment. This study, however, did not report the outcome of patients who received targeted therapy. The outcome of patients with actionable molecular alterations that received targeted therapies has, however, been studies in the SHIVA trial [15]. The SHIVA trial is a randomized controlled open-label phase 2 trial that was launched in eight French academic centers to evaluate the efficacy of 11
Table 2 Notable molecular alterations in gastrointestinal cancers. Tumor
Notable molecular alterations
References
Gastroesophageal
TP53, CDKN2A, CCNE, CDK 4/6, CCND, EGFR, ERBB2, ERBB3, PI3KCA, PIK3R1, MCL1, PTEN, CDH1, JAK2, PD-L1/2, VEGFA, KRAS/NRAS and MET IDH1, CDH1, KIT, FGFR2, FLT3, NPM1, PTEN, MET, AKT1, RET, NOTCH1 ERBB4, ERBB2, KRRAS, BRAF and FBXW7 APC, TP53, SMAD4, PIK3CA, PIK3R1, PTEN, KRAS, NRAS, BRAF, ARID1A, SOX9, FAM123B/WTX, ERBB2, IGF2 NAV2/TCF7L, ACVR2A, APC, TGFBR2, MLH1, MSH3, MSH6, SLC9A9, SMAD2, SMAD3 and SMAD4 TP53, SMAD4, CDKN2A, ARID1A, ROBO2, KDM6A, PREX2, ERBB2, MET, FGFR1, CDK6, PIK3R3, PIK3CA, BRCA1, BRCA2, PALB2, KRAS, TGFBR2, BRAF, PREX2, MLL2, MLH2 and MLH2 TP53, KRAS, ERBB3, EGFR, ERBB2, ERBB4, FGFR2, PRKACA and PRKACB TP53, CTNNB1, MET, PTEN, CDKN2A, AXIN1, PTEN, PIK3CA, KRAS, NRAS, MYC, MET, CCND2, RB1, ARID1A, ARID1B, ARID2, IRF2, NFE2L2, ERRFl1, RPSKA3 and MLL3
[14,37]
Small bowel Colorectal Pancreas Biliary Hepatocellular
[38,39] [40] [8] [9,10] [41,42]
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
189
Table 3 Frequency of common (≥5) actionable molecular alterations and potential targeted therapy. Tumor
molecular alterations
Frequency (%)
Therapy
References
Gastroesophageal
ERBB2 CDKN2A, CCNE, CDK 4/6, CCND PIK3CA PTEN loss MET KRAS PIK3CA ERBB2 BRAF FBXW7 PTEN(loss) PIK3CA FBXW7 BRAF CDKN2A BRCA2 FANCC KRAS PIK3CA BRAF CCND1 FGFR2 FGFR1 FGFR3 CDKN2A PTEN MET
15–30 58 14 9 5 40 10 8 6 6 34 15–20 14 8 95 7–10 6 18 7 6 6 5 4 4 10–60 5–10 1–5
Trastuzumab CDK 4/6 inhibitors PIK3CA, AKT, MTOR inhibitors PIK3CA, AKT, MTOR inhibitors MET inhibitors MEK, RAF, AKT inhibitors
[14,17,43]
PIK3CA, AKT, MTOR inhibitors
[33,44–47]
Small bowel
Colorectal
Pancreas
Biliary
Hepatocellular
molecularly targeted agents versus investigator’s choice of conventional therapy when used in patients with advanced cancer who have failed standard therapies and underwent fresh tumor biopsies and NGS. At the time of data cut off, the study had enrolled 195 patients (99 in the experimental arm and 96 in the control arm), of whom 17% had a gastrointestinal malignancy. Median progression free survival (PFS) was 2.3 months (95% CI 1.7–3.8) in the experimental group versus 2.0 months (1.8–2.1) in the control group (hazard ratio 0.88, 95% CI 0.65–1.19, p = 0.41). Of note, there was no statistically significant difference in the rate of grade 3/4 adverse events between the two groups. These results were disappointing and suggest that further research is needed to investigate the utility of genomic predictive markers and targeted agents that may be effective as single agents or in combination against those markers. Taken together, these data suggest that NGS can be routinely performed in clinical practice. Although actionable mutations are identified in a significant number of patients with cancer and gastrointestinal malignancies, the majority of patients, however, do not receive molecularly directed therapies on the basis of NGS results. Available trials have yet to show improved outcomes with NGS and molecularly directed therapies as compared to conventional treatment. This field is, however, progressing rapidly. As molecular testing techniques evolve and our understanding of genomic alterations and molecular pathways develops further, we suspect that NGS will play an integral role of linking patients to appropriate targeted therapies and improving outcomes compared to conventional therapy. The role of NGS in identifying predictive markers for resistance to cancer therapy Genomic testing is used to identify molecular alterations that confer resistance to targeted therapy. The most notable example is identification of RAS mutations to determine the benefit from EGFR inhibitors [19]. Two recent studies reported by Van Cutsem et al. [20] and Douillard et al. [21], respectively, demonstrated that extended RAS testing identifies more patients that are resistant to EGFR inhibitors compared to only KRAS exon 2 (codons 12 and 13) testing. The study by Doulliard et al. [21] included testing for KRAS and NRAS
MTOR inhibitors RAF, MEK, AKT inhibitors CDK 4/6 inhibitors PARP inhibitors PARP inhibitors MEK, RAF, AKT inhibitors PIK3CA inhibitors BRAF inhibitors CDK 4/6 inhibitors FGFR inhibitors FGFR inhibitors FGFR inhibitors CDK 4/6 inhibitors PIK3CA, AKT, MTOR inhibitors MET inhibitors
[39]
[8,17,48]
[9,10,17,34,49]
[50–54]
exons 2 (codons 12 and 13), 3 (codon 61) and 4 (codons 117 and 146), while the study by Van Cutsem et al. [20] involved testing for KRAS exon 3 (codons 59 and 61), KRAS exon 4 (codons 117 and 146), NRAS exon 2 (codons 12 and 13), NRAS exon 3 (codons 59 and 61), and NRAS exon 4 (codons 117 and 146). Other alterations such as BRAF, PIK3CA exon 20 mutations, PTEN loss and MET amplification may also be associated with resistance to EGFR inhibitors [22–24]. NGS has emerged as a practical platform to study these mutations and their effect on outcomes of EGFR therapy [25]. It remains a matter of debate which NGS technology is most feasible in detecting such mutations in clinical practice. A study by Guedes et al. compared direct sequencing of KRAS codons 12 and 13 (sensitivity = 20%) with newer techniques, with sensitivities of up to 0.1% such as high resolution melting (HRM) analysis followed by sequencing for positive cases [26]. The incidence of KRAS mutations increased from 45% by direct sequencing to 60% by HRM followed by sequencing. When other mutations that may confer resistance to KRAS such as BRAF, PI3K and PTEN were added to the analysis, up to 87% of non-responders to anti-EGFR were found to have mutations that may have predicted their resistance to anti-EGFR therapy. These results, however, are hypothesis generating and need to be confirmed in larger studies. Additionally, it is important to note that other mutations not included in this study may also influence the response of patients to anti-EGFR therapy. Similarly, MET amplification and PTEN loss may constitute a mechanism of resistance to HER 2 directed therapy in gastroesophageal cancer [27,28]. The widespread availability of NGS is currently playing a role toward improving our knowledge of the impact of such molecular alterations and in further personalization of treatment for patients with GI cancers [14]. The role of NGS in identifying familial gastrointestinal cancers Approximately 5% of patients with CRC have identifiable familial CRC syndromes [29]. Additionally up to 30% of patients diagnosed with CRC have evidence of a familial component [29]. The most frequent CRC syndromes are: (1) Familial adenomatous polyposis (FAP) associated with adenomatous polyposis (APC) gene mutations, (2)
190
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
polyposis associated with MUTYH gene mutations and (3) hereditary nonpolyposis CRC associated with alterations in genes responsible for DNA mismatch repair such as MLH1, MSH2, MSH6 or PMS2. NGS represents a viable option to identify molecular alterations associated with familial CRC syndromes. When compared to Sanger sequencing, however, approximately 4% of genomic alterations may be missed by NGS but detected by Sanger sequencing. A study by Simbolo et al. [29] suggested that sequences close or within homopolymetric stretches of DNA represent a group of genetic mutations that may better characterized by Sanger sequencing versus NGS. Given the lower cost of NGS and its ability to simultaneously analyze multiple genes in multiple samples, it may be feasible to adopt an approach of using NGS for routine diagnosis of familial CRC syndromes and limiting the use of costly Sanger sequencing techniques to regions that are not adequately covered by NGS. Similarly, in gastroesophageal cancers, NGS is a feasible technique to uncover mutations such as CDH1 or BRCA2 mutations to identify patients that need further testing for germline mutations [14,30–32]. Challenges to the routine use of NGS in clinical practice The routine use of NGS in clinical practice continues to face several challenges. Although NGS techniques may be less costly than other molecular diagnostic techniques, their cost effectiveness remains uncertain in patients with GI cancers. A large number of patients who undergo NGS testing do not receive molecularly directed therapy [16]. Challenges that limit the clinical utility of NGS testing include the long turnaround time (2–4 weeks) of NGS, rapid decline of patients’ performance status as NGS is being performed, lack of actionable mutations, lack of available clinical trials or patient refusal to enroll in clinical trials. Additionally it remains unclear if molecularly directed therapy that is used based on data obtained from NGS is associated with improved outcomes [15]. Altered genes detected by NGS may not represent driver mutations that would result in meaningful responses to targeted therapy. This case was clearly demonstrated when vemurafenib, a proven effective therapy in BRAF mutated melanoma [5], was evaluated in a phase II trial in patients with CRC and a BRAFV600 mutation [33]. Treatment responses were significantly less impressive in CRC compared to those observed in patients with melanoma that harbor the same mutation and were treated with the same agent. This finding indicates that MAPK/BRAF pathway may contain alternate feedback and resistance loops in CRC when compared to melanoma [33]. Another challenge that limits clinical utility of NGS and other molecular testing techniques is tumor heterogeneity [34,35]. Heterogeneity may be related to different subclones within the same tumor or between primary and metastatic tumor deposits. Such heterogeneity may result in differential treatment responses and resistance mechanisms depending on selective pressures in each tumor deposit. This constitutes both a diagnostic and a therapeutic challenge. Further research is needed to better understand the mechanism of genomic heterogeneity and identify diagnostic and therapeutic strategies to overcome it. The concordance of NGS testing and other molecular diagnostic techniques also remains undefined [14,29]. It is important to carefully study if NGS should complement or replace other testing modalities such as IHC or FISH. It appears, currently, that NGS may have inferior sensitivity to the aforementioned techniques [14]. However, we believe that NGS technologies will continue to develop and will eventually replace other molecular testing techniques. Therefore, with all the current limitations of NGS, cost effectiveness studies are needed to better evaluate the economic impact and feasibility of more widespread use of molecular testing and targeted therapy in the treatment of patients with GI cancers. Overall, we suspect that further refinement of NGS technologies will result in lower cost, faster turnaround times for testing and better characterization of molecular abnor-
malities and consequently will improve its clinical utility, cost effectiveness and use in routine clinical practice. Future directions Research in NGS requires a commitment to providing funding and philanthropic support to continue the progress that has been already achieved and help transition NGS from the bench to the bedside. Improving turnaround time is crucial to provide clinical data in a timely fashion to guide clinical decisions. We believe that repeating biopsies to detect molecular evolution of tumors may be necessary. This approach will be facilitated by advances such as testing circulating tumor cells or circulating tumor DNA by NGS without obtaining a tissue biopsy. Several issues will need to be resolved such as the interplay between tumor histology and the various genomic alterations. This will require accurate characterization of feedback and resistance loops of the various molecular abnormalities based on tissue histology and better study their response to targeted therapy. Such progress can be potentially made through clinical trials with adaptive designs that aim to identify mutations and tumor types that have exceptional responses to targeted therapies. The National Cancer institute MATCH trial [36] represents a step in this direction and more trials are likely to follow to help define the role of NGS in clinical oncology. The advances in molecular diagnostics will not result in meaningful clinical benefits unless they are paralleled with advances in developing novel therapeutics that are both clinically effective and safe. Although the challenges remain substantial, the progress that has been made is enormous and the opportunity to continue this progress has never been greater. Conflict of interest The authors report no relevant conflicts of interest. References [1] E.L. van Dijk, H. Auger, Y. Jaszczyszyn, C. Thermes, Ten years of next-generation sequencing technology, Trends Genet. 30 (2014) 418–426. [2] A.T. Shaw, D.W. Kim, K. Nakagawa, T. Seto, L. Crino, M.J. Ahn, et al., Crizotinib versus chemotherapy in advanced ALK-positive lung cancer, N. Engl. J. Med. 368 (2013) 2385–2394. [3] M. Fukuoka, Y.L. Wu, S. Thongprasert, P. Sunpaweravong, S.S. Leong, V. Sriuranpong, et al., Biomarker analyses and final overall survival results from a phase III, randomized, open-label, first-line study of gefitinib versus carboplatin/paclitaxel in clinically selected patients with advanced non-smallcell lung cancer in Asia (IPASS), J. Clin. Oncol. 29 (2011) 2866–2874. [4] C. Zhou, Y.L. Wu, G. Chen, J. Feng, X.Q. Liu, C. Wang, et al., Final overall survival results from a randomised, phase III study of erlotinib versus chemotherapy as first-line treatment of EGFR mutation-positive advanced non-small-cell lung cancer (OPTIMAL, CTONG-0802), Ann. Oncol. 26 (2015) 1877–1883. [5] J.A. Sosman, K.B. Kim, L. Schuchter, R. Gonzalez, A.C. Pavlick, J.S. Weber, et al., Survival in BRAF V600-mutant advanced melanoma treated with vemurafenib, N. Engl. J. Med. 366 (2012) 707–714. [6] B. Tran, J.E. Dancey, S. Kamel-Reid, J.D. McPherson, P.L. Bedard, A.M. Brown, et al., Cancer genomics: technology, discovery, and translation, J. Clin. Oncol. 30 (2012) 647–660. [7] E.M. Van Allen, N. Wagle, M.A. Levy, Clinical analysis and interpretation of cancer genome data, J. Clin. Oncol. 31 (2013) 1825–1833. [8] N. Waddell, M. Pajic, A.-M. Patch, D.K. Chang, K.S. Kassahn, P. Bailey, et al., Whole genomes redefine the mutational landscape of pancreatic cancer, Nature 518 (2015) 495–501. [9] M. Li, Z. Zhang, X. Li, J. Ye, X. Wu, Z. Tan, et al., Whole-exome and targeted gene sequencing of gallbladder carcinoma identifies recurrent mutations in the ErbB pathway, Nat. Genet. 46 (2014) 872–876. [10] H. Nakamura, Y. Arai, Y. Totoki, T. Shirota, A. Elzawahry, M. Kato, et al., Genomic spectra of biliary tract cancer, Nat. Genet. 47 (2015) 1003–1010. [11] J.C. Haan, M. Labots, C. Rausch, M. Koopman, J. Tol, L.J.M. Mekenkamp, et al., Genomic landscape of metastatic colorectal cancer, Nat. Commun. 5 (2014). [12] B.M. Wolpin, C. Rizzato, P. Kraft, C. Kooperberg, G.M. Petersen, Z. Wang, et al., Genome-wide association study identifies multiple susceptibility loci for pancreatic cancer, Nat. Genet. 46 (2014) 994–1000. [13] M.C. Heinrich, C.L. Corless, G.D. Demetri, C.D. Blanke, M. von Mehren, H. Joensuu, et al., Kinase mutations and imatinib response in patients with metastatic gastrointestinal stromal tumor, J. Clin. Oncol. 21 (2003) 4342–4349.
S. Mikhail et al./Cancer Letters 374 (2016) 187–191
[14] S. Mikhail, K. Ciombor, A. Noonan, C. Wu, R. Goldberg, W. Zhao, et al., Upfront molecular testing in patients with advanced gastro-esophageal cancer: Is it time yet?, Oncotarget 6 (2015) 22206–22213. [15] C. Le Tourneau, J.P. Delord, A. Goncalves, C. Gavoille, C. Dubot, N. Isambert, et al., Molecularly targeted therapy based on tumour molecular profiling versus conventional therapy for advanced cancer (SHIVA): a multicentre, open-label, proof-of-concept, randomised, controlled phase 2 trial, Lancet Oncol. 16 (2015) 1324–1334. [16] F. Meric-Bernstam, L. Brusco, K. Shaw, C. Horombe, S. Kopetz, M.A. Davies, et al., Feasibility of large-scale genomic testing to facilitate enrollment onto genomically matched clinical trials, J. Clin. Oncol. 33 (2015) 2753–2762. [17] S. Mikhail, C. Albanese, M.J. Pishvaian, Cyclin-dependent kinase inhibitors and the treatment of gastrointestinal cancers, Am. J. Pathol. 185 (2015) 1185–1197. [18] P.P.-S.M. Gopalan, A. Chiappori, A. Ivey, A. Godillo-Villegas, F. Kaye, A phase II clinical trial of CDK 4/6 inhibitor palbociblib (PD 0332991) in previously treated, advanced non-small cell lung cancer patients with inactivated CDKN2A, in: ASCO Annual Meeting, Chicago, IL, 2014. [19] C.J. Allegra, R.B. Rumble, S.R. Hamilton, P.B. Mangu, N. Roach, A. Hantel, et al., Extended RAS gene mutation testing in metastatic colorectal carcinoma to predict response to anti-epidermal growth factor receptor monoclonal antibody therapy: American Society of Clinical Oncology provisional clinical opinion update 2015, J. Clin. Oncol. 34 (2016) 179–185. [20] E. Van Cutsem, H.J. Lenz, C.H. Kohne, V. Heinemann, S. Tejpar, I. Melezinek, et al., Fluorouracil, leucovorin, and irinotecan plus cetuximab treatment and RAS mutations in colorectal cancer, J. Clin. Oncol. 33 (2015) 692–700. [21] J.Y. Douillard, K.S. Oliner, S. Siena, J. Tabernero, R. Burkes, M. Barugel, et al., Panitumumab-FOLFOX4 treatment and RAS mutations in colorectal cancer, N. Engl. J. Med. 369 (2013) 1023–1034. [22] L. Barault, N. Veyrie, V. Jooste, D. Lecorre, C. Chapusot, J.M. Ferraz, et al., Mutations in the RAS-MAPK, PI(3)K (phosphatidylinositol-3-OH kinase) signaling network correlate with poor survival in a population-based series of colon cancers, Int. J. Cancer 122 (2008) 2255. [23] A. Bardelli, S. Corso, A. Bertotti, S. Hobor, E. Valtorta, G. Siravegna, et al., Amplification of the MET receptor drives resistance to anti-EGFR therapies in colorectal cancer, Cancer Discov. 3 (2013) 658. [24] W. De Roock, B. Claes, D. Bernasconi, J. De Schutter, B. Biesmans, G. Fountzilas, et al., Effects of KRAS, BRAF, NRAS, and PIK3CA mutations on the efficacy of cetuximab plus chemotherapy in chemotherapy-refractory metastatic colorectal cancer: a retrospective consortium analysis, Lancet Oncol. 11 (2010) 753. [25] K.K. Ciombor, S. Haraldsdottir, R.M. Goldberg, How can next-generation sequencing (genomics) help us in treating colorectal cancer?, Curr. Colorectal Cancer Rep. 10 (2014) 372–379. [26] J.G. Guedes, I. Veiga, P. Rocha, P. Pinto, C. Pinto, M. Pinheiro, et al., High resolution melting analysis of KRAS, BRAF and PIK3CA in KRAS exon 2 wild-type metastatic colorectal cancer, BMC Cancer 13 (2013) 169. [27] A.K. Paulson, E.S. Linklater, B.D. Berghuis, C.A. App, L.D. Oostendorp, J.E. Paulson, et al., Graveel, MET and ERBB2 are coexpressed in ERBB2+ breast cancer and contribute to innate resistance, Mol. Cancer Res. 11 (2013) 1112–1121. [28] X. Zhang, J.S. Park, K.H. Park, K.H. Kim, M. Jung, H.C. Chung, et al., PTEN deficiency as a predictive biomarker of resistance to HER2-targeted therapy in advanced gastric cancer, Oncology 88 (2015) 76–85. [29] M. Simbolo, A. Mafficini, M. Agostini, C. Pedrazzani, C. Bedin, E.D. Urso, et al., Next-generation sequencing for genetic testing of familial colorectal cancer syndromes, Heredit. Cancer Clin. Pract. 13 (2015) 18. [30] R.C. Grant, I. Selander, A.A. Connor, S. Selvarajah, A. Borgida, L. Briollais, et al., Prevalence of germline mutations in cancer predisposition genes in patients with pancreatic cancer, Gastroenterology 148 (2015) 556–564. [31] Y. Hirotsu, H. Nakagomi, I. Sakamoto, K. Amemiya, H. Mochizuki, M. Omata, Detection of BRCA1 and BRCA2 germline mutations in Japanese population using next-generation sequencing, Mol. Genet. Genomic Med. 3 (2015) 121– 129. [32] Y. Hirotsu, H. Nakagomi, I. Sakamoto, K. Amemiya, T. Oyama, H. Mochizuki, et al., Multigene panel analysis identified germline mutations of DNA repair genes in breast and ovarian cancer, Mol. Genet. Genomic Med. 3 (2015) 459–466.
191
[33] S. Kopetz, J. Desai, E. Chan, J.R. Hecht, P.J. O’Dwyer, D. Maru, et al., Pilot study of vemurafenib in patients with metastatic BRAF-mutated colorectal cancer, J. Clin. Oncol. 33 (2015) 4032–4038. [34] R. Simon, S. Roychowdhury, Implementing personalized cancer genomics in clinical trials, Nat. Rev. Drug Discov. 12 (2013) 358–369. [35] M. Gerlinger, A.J. Rowan, S. Horswell, J. Larkin, D. Endesfelder, E. Gronroos, et al., Intratumor heterogeneity and branched evolution revealed by multiregion sequencing, N. Engl. J. Med. 366 (2012) 883–892. [36] C. McNeil, NCI-MATCH launch highlights new trial design in precision-medicine era, J. Nat. Cancer Inst. 107 (2015). [37] TCGA Network, Comprehensive molecular characterization of gastric adenocarcinoma, Nature 513 (2014) 202–209. [38] M.A. Alvi, D.G. McArt, P. Kelly, M.A. Fuchs, M. Alderdice, C.M. McCabe, et al., Comprehensive molecular pathology analysis of small bowel adenocarcinoma reveals novel targets with potential for clinical utility, Oncotarget 6 (2015) 20863–20874. [39] A. Laforest, T. Aparicio, A. Zaanan, F.P. Silva, A. Didelot, A. Desbeaux, et al., ERBB2 gene as a potential therapeutic target in small bowel adenocarcinoma, Eur. J. Cancer 50 (2014) 1740–1746. [40] Cancer Genome Atlas Network, Comprehensive molecular characterization of human colon and rectal cancer, Nature 487 (2012) 330. [41] C. Ang, C. Gamblin, S. Millis, J. Xiu, Z. Gatalica, S. Reddy, et al., Molecular characterization of 350 hepatocellular carcinomas to identify biomarker aberrations with potential novel therapeutic options, in: ASCO Annual Meeting, Chicago, IL, 2015. [42] H. Nakagawa, T. Shibata, Comprehensive genome sequencing of the liver cancer genome, Cancer Lett. 340 (2013) 234–240. [43] Y.J. Bang, E. Van Cutsem, A. Feyereislova, H.C. Chung, L. Shen, A. Sawaki, et al., Trastuzumab in combination with chemotherapy versus chemotherapy alone for treatment of HER2-positive advanced gastric or gastro-oesophageal junction cancer (ToGA): a phase 3, open-label, randomised controlled trial, Lancet 376 (2010) 687–697. [44] R.B. Corcoran, C.E. Atreya, G.S. Falchook, E.L. Kwak, D.P. Ryan, J.C. Bendell, et al., Combined BRAF and MEK inhibition with dabrafenib and trametinib in BRAF V600-mutant colorectal cancer, J. Clin. Oncol. 33 (2015) 4023–4031. [45] D.L. Jardim, J.J. Wheler, K. Hess, A.M. Tsimberidou, R. Zinner, F. Janku, et al., FBXW7 mutations in patients with advanced cancers: clinical and molecular characteristics and outcomes with mTOR inhibitors, PLoS ONE 9 (2014) e89388. [46] P.C. Lin, J.K. Lin, H.H. Lin, Y.T. Lan, C.C. Lin, S.H. Yang, et al., A comprehensive analysis of phosphatase and tensin homolog deleted on chromosome 10 (PTEN) loss in colorectal cancer, World J. Surg. Oncol. 13 (2015) 186. [47] S. Ogino, P. Lochhead, E. Giovannucci, J.A. Meyerhardt, C.S. Fuchs, A.T. Chan, Discovery of colorectal cancer PIK3CA mutation as potential predictive biomarker: power and promise of molecular pathological epidemiology, Oncogene 33 (2014) 2949–2955. [48] J.M. Winter, A. Maitra, C.J. Yeo, Genetics and pathology of pancreatic cancer, HPB 8 (2006) 324–336. [49] T. Bekaii-Saab, M.A. Phelps, X. Li, M. Saji, L. Goff, J.S. Kauh, et al., Multiinstitutional phase II study of selumetinib in patients with metastatic biliary cancers, J. Clin. Oncol. 29 (2011) 2357–2363. [50] P. Laurent-Puig, P. Legoix, O. Bluteau, J. Belghiti, D. Franco, F. Binot, et al., Genetic alterations associated with hepatocellular carcinomas define distinct pathways of hepatocarcinogenesis, Gastroenterology 120 (2001) 1763–1773. [51] P. Laurent-Puig, J. Zucman-Rossi, Genetics of hepatocellular tumors, Oncogene 25 (2006) 3778. [52] M.J. Pishvaian, G. Upadhyay, Z. Yao, Y. Kim, A. Said, J. Mendelson, et al., CDK4 is an important target for hepatocellular cancer with inactivation of TGF-beta signaling through Smad and ELF, in: 99th AACR Annual Meeting, San Diego, CA, 2008. [53] J. Zucman-Rossi, Molecular classification of hepatocellular carcinoma, Digest. Liver Dis. 42 (Suppl. 3) (2010) S235–S241. [54] J. Zucman-Rossi, P. Laurent-Puig, Genetic diversity of hepatocellular carcinomas and its potential impact on targeted therapies, Pharmacogenomics 8 (2007) 997–1003.