Clinical Implications of a Targeted RNA-Sequencing Panel in the Detection of Gene Fusions in Solid Tumors
2021; Elsevier BV; Volume: 23; Issue: 12 Linguagem: Inglês
10.1016/j.jmoldx.2021.08.009
ISSN1943-7811
AutoresLulu Sun, Samantha N. McNulty, Michael Evenson, Xiaopei Zhu, Joshua A. Robinson, Patrick R. Mann, Eric J. Duncavage, John D. Pfeifer,
Tópico(s)Tumors and Oncological Cases
ResumoThe detection of recurrent gene fusions can help confirm diagnoses in solid tumors, particularly when the morphology and staining are unusual or nonspecific, and can guide therapeutic decisions. Although fluorescence in situ hybridization and PCR are often used to identify fusions, the rearrangement must be suspected, with only a few prioritized probes run. It was hypothesized that the Illumina TruSight RNA Fusion Panel, which detects fusions of 507 genes and their partners, would uncover fusions with greater sensitivity than other approaches, leading to changes in diagnosis, prognosis, or therapy. Targeted RNA sequencing was performed on formalin-fixed, paraffin-embedded sarcoma and carcinoma cases in which fluorescence in situ hybridization, RT-PCR, or DNA-based sequencing was conducted during the diagnostic workup. Of the 153 cases, 138 (90%) were sequenced with adequate quality control metrics. A total of 101 of 138 (73%) cases were concordant by RNA sequencing and the prior test method. RNA sequencing identified an additional 30 cases (22%) with fusions that were not detected by conventional methods. In seven cases (5%), the additional fusion information provided by RNA sequencing would have altered diagnosis and management. A total of 19 novel fusion pairs (not previously described in the literature) were discovered (14%). Overall, the findings show that a targeted RNA-sequencing method can detect gene fusions in formalin-fixed, paraffin-embedded specimens with high sensitivity. The detection of recurrent gene fusions can help confirm diagnoses in solid tumors, particularly when the morphology and staining are unusual or nonspecific, and can guide therapeutic decisions. Although fluorescence in situ hybridization and PCR are often used to identify fusions, the rearrangement must be suspected, with only a few prioritized probes run. It was hypothesized that the Illumina TruSight RNA Fusion Panel, which detects fusions of 507 genes and their partners, would uncover fusions with greater sensitivity than other approaches, leading to changes in diagnosis, prognosis, or therapy. Targeted RNA sequencing was performed on formalin-fixed, paraffin-embedded sarcoma and carcinoma cases in which fluorescence in situ hybridization, RT-PCR, or DNA-based sequencing was conducted during the diagnostic workup. Of the 153 cases, 138 (90%) were sequenced with adequate quality control metrics. A total of 101 of 138 (73%) cases were concordant by RNA sequencing and the prior test method. RNA sequencing identified an additional 30 cases (22%) with fusions that were not detected by conventional methods. In seven cases (5%), the additional fusion information provided by RNA sequencing would have altered diagnosis and management. A total of 19 novel fusion pairs (not previously described in the literature) were discovered (14%). Overall, the findings show that a targeted RNA-sequencing method can detect gene fusions in formalin-fixed, paraffin-embedded specimens with high sensitivity. Increasingly, rearrangements resulting in gene fusions have been revealed in multiple tumor types across diverse organ systems. These gene fusions can drive tumorigenesis by altering gene expression and activity.1Mitelman F. Johansson B. Mertens F. The impact of translocations and gene fusions on cancer causation.Nat Rev Cancer. 2007; 7: 233-245Crossref PubMed Scopus (989) Google Scholar Timely and reliable detection of fusions provides useful information to the pathologist and clinical team. For the surgical pathologist, gene fusions associated with particular pathologic entities can be helpful in confirming a diagnosis, especially in cases where the morphology is unusual or nonspecific, such as in small round blue cell tumors or in high-grade sarcomas and carcinomas.2McConnell L. Houghton O. Stewart P. Gazdova J. Srivastava S. Kim C. Catherwood M. Strobl A. Flanagan A.M. Oniscu A. Kroeze L.I. Groenen P. Taniere P. Salto-Tellez M. Gonzalez D. A novel next generation sequencing approach to improve sarcoma diagnosis.Mod Pathol. 2020; 33: 1350-1359Crossref PubMed Scopus (12) Google ScholarThe protein products of gene fusions can be targeted for therapy, as illustrated by the use of imatinib for BCR-ABL in leukemia3Rossari F. Minutolo F. Orciuolo E. Past, present, and future of Bcr-Abl inhibitors: from chemical development to clinical efficacy.J Hematol Oncol. 2018; 11: 84Crossref PubMed Scopus (155) Google Scholar and crizotinib for ALK rearrangements in lung cancer.4Gridelli C. Peters S. Sgambato A. Casaluce F. Adjei A.A. Ciardiello F. ALK inhibitors in the treatment of advanced NSCLC.Cancer Treat Rev. 2014; 40: 300-306Abstract Full Text Full Text PDF PubMed Scopus (143) Google Scholar Fusions may also have prognostic implications; for example, B-cell lymphomas with concurrent BCL2 and MYC rearrangements (double-hit lymphomas) are a subtype with more aggressive clinical behavior.5Aukema S.M. Siebert R. Schuuring E. Van Imhoff G.W. Kluin-Nelemans H.C. Boerma E.J. Kluin P.M. Double-hit B-cell lymphomas.Blood. 2011; 117: 2319-2331Crossref PubMed Scopus (543) Google ScholarMolecular techniques, such as fluorescence in situ hybridization (FISH), reverse transcription PCR (RT-PCR), and Sanger and next-generation sequencing (NGS), have enabled discovery of characteristic fusions associated with particular tumors. With FISH and RT-PCR, the genes involved must be known or suspected, and usually only a few prioritized assays are run, due to cost and time limitations. These methods are not suitable for the identification of novel fusion partners. In contrast, RNA-based NGS can evaluate hundreds of possible fusion results simultaneously.6Teixidó C. Giménez-Capitán A. Molina-Vila M.Á. Peg V. Karachaliou N. Rodríguez-Capote A. Castellví J. Rosell R. RNA analysis as a tool to determine clinically relevant gene fusions and splice variants.Arch Pathol Lab Med. 2018; 142: 474-479Crossref PubMed Scopus (13) Google Scholar,7Heyer E.E. Deveson I.W. Wooi D. Selinger C.I. Lyons R.J. Hayes V.M. O'Toole S.A. Ballinger M.L. Gill D. Thomas D.M. Mercer T.R. Blackburn J. Diagnosis of fusion genes using targeted RNA sequencing.Nat Commun. 2019; 10: 1388Crossref PubMed Scopus (70) Google ScholarThe study assessed the performance and utility of a commercially available targeted RNA-sequencing panel (TruSight RNA Fusion Panel) that covers 507 genes commonly found in cancer-related gene fusions and their fusion partners. Additionally, a cohort of 153 sarcoma, carcinoma, and lymphoma cases were studied, in which rearrangements were previously assessed by routine clinical FISH, RT-PCR, or DNA-based NGS. The ability of the TruSight RNA Fusion Panel to detect known and previously undetected gene fusions was determined. Finally, the clinical impact (in diagnosis, prognosis, or therapy) of any differences in the molecular results between the RNA fusion panel and the other methods were investigated.Materials and MethodsEthicsThis study was approved by the Institutional Review Board of Washington University School of Medicine (St. Louis, MO) with a waiver of consent (Institutional Review Board number 201102311).Case SelectionSarcoma and carcinoma cases submitted for FISH, RT-PCR, or DNA-based NGS testing during diagnostic workup between 2013 and 2019 were identified retrospectively from our institutional database. These assays were performed in other laboratories as part of clinical care, and therefore, the methods are not included in this study. Hematoxylin and eosin–stained slides and paraffin blocks were screened for adequate tissue (at least enough for a 1-mm punch, with ≥50% tumor cellularity in the punched area). A different block from the same patient was selected in cases where the original block was unavailable or there was inadequate tissue remaining.RNA ExtractionAppropriate tumor areas were circled on the hematoxylin and eosin slides by a pathologist (L.S.), and 1-mm punches were taken from the corresponding areas on the paraffin blocks. Deparaffinization and RNA extraction were performed using RecoverALL Total Nucleic Acid Isolation Kit for FFPE (Thermo Fisher Scientific, Waltham, MA). Quality of RNA and input amount were determined by Bioanalyzer (Agilent, Santa Clara, CA), with DV200 values, per TruSight RNA Fusion Panel kit protocol (Illumina, San Diego, CA). A total of 100 to 200 ng of RNA was used for library preparation based on these values, with higher RNA input used for cases with lower DV200 values, per manufacturer recommendations.TruSight RNA Fusion Panel Library PreparationLibrary preparation was performed using the TruSight RNA Fusion Panel kit, according to the manufacturer's suggested protocol. The TruSight RNA Fusion Panel uses 21,283 probes to target 7690 exonic regions in 507 genes. Briefly, blunt-ended double-stranded cDNA was synthesized from extracted RNA. The 3′ ends were adenylated, and multiple indexing adapters were ligated to each specimen. After PCR amplification, the library was quantified by Bioanalyzer. Probes targeted to the regions of interest were hybridized and then captured. After washing, a second hybridization and capture were performed. The captured library was purified, and a second PCR amplification was performed. The amplified, enriched library was cleaned with AMPure XP beads (Beckman Coulter, Pasadena, CA), and the concentration and quality of the library were measured by Bioanalyzer before sequencing. Library preparation was repeated once on cases with incorrect/low band size.SequencingSequencing was performed on the Illumina MiSeq and NextSeq 550 systems, using paired-end (2 × 75 nucleotides) reads. To maximize sensitivity and specificity, a target of approximately 3 million reads was sequenced per specimen, according to kit recommendations. In cases where total reads greatly exceeded this number (ie, >9 million reads per sample), based on available lanes per flow cell, the reads were randomly downsampled to prevent false positives. Downsampling was applied at the same ratio across all samples run on the same chip.Bioinformatic AnalysisReads were aligned to the human reference assembly (GRCh37) using STAR aligner version 2.7.2b.8Dobin A. Davis C.A. Schlesinger F. Drenkow J. Zaleski C. Jha S. Batut P. Chaisson M. Gingeras T.R. STAR: ultrafast universal RNA-seq aligner.Bioinformatics. 2013; 29: 15-21Crossref PubMed Scopus (17731) Google Scholar Samtools version 1.99Li H. Handsaker B. Wysoker A. Fennell T. Ruan J. Homer N. Marth G. Abecasis G. Durbin R. The sequence alignment/map format and SAMtools.Bioinformatics. 2009; 25: 2078-2079Crossref PubMed Scopus (30186) Google Scholar and bedtools version 2.29.210Quinlan A.R. Hall I.M. BEDTools: a flexible suite of utilities for comparing genomic features.Bioinformatics. 2010; 26: 841-842Crossref PubMed Scopus (11125) Google Scholar were used to tally the number of reads that were aligned to the genome and targeted regions (targeted regions file available from Illumina). Cases with <75% on-target reads aligned to targeted regions were excluded. Gene fusions were predicted using STAR-Fusion version 1.8.1.11Haas B.J. Dobin A. Stransky N. Li B. Yang X. Tickle T. Bankapur A. Ganote C. Doak T.G. Pochet N. Sun J. Wu C.J. Gingeras T.R. Regev A. STAR-fusion: fast and accurate fusion transcript detection from RNA-Seq.bioRxiv. 2017; ([Preprint] doi:)10.1101/120295Google Scholar Fusion events were considered when at least one partner was included in the capture panel, the fusion was supported by both spanning reads (ie, paired reads align to each fusion partner) and junction reads (ie, a single read that aligns across the junction), the fusion was supported by at least five spanning reads or five junction reads, and junctions located on the same chromosome were at least 10,000 bp apart. Fusion calls found in greater than three cases with the exact same breakpoints, not previously verified in the literature, were excluded as false positives/artifacts. Putative fusion transcripts were assembled by Trinity (included in the STAR-Fusion package).Annotation and InterpretationFusionInspector (included in the STAR-Fusion package) was used to verify fusion calls that passed initial filtering. Fusion calls were visually inspected using HTML-based FusionInspector reports and with Integrated Genomics Viewer (Broad Institute, Cambridge, MA).12Robinson J.T. Thorvaldsdóttir H. Winckler W. Guttman M. Lander E.S. Getz G. Mesirov J.P. Integrative genomics viewer.Nat Biotechnol. 2011; 29: 24-26Crossref PubMed Scopus (7268) Google Scholar In-frame and frameshift status was determined using FusionInspector, transcripts generated by Trinity, and manual inspection with Integrated Genomics Viewer and the University of California, Santa Cruz, Genome Browser.13Kent W.J. Sugnet C.W. Furey T.S. Roskin K.M. Pringle T.H. Zahler A.M. Haussler aD. The Human Genome Browser at UCSC.Genome Res. 2002; 12: 996-1006Crossref PubMed Scopus (6552) Google Scholar Discordant cases that had positive fusions by conventional methods and negative results by RNA sequencing were manually inspected with Integrated Genomics Viewer at putative gene breakpoints for read coverage.New fusions (not identified by prior testing methods) underwent a literature search and were checked against The Cancer Genome Atlas Fusion Gene Database and FusionHub14Panigrahi P. Jere A. Anamika K. FusionHub: a unified web platform for annotation and visualization of gene fusion events in human cancer.PLoS One. 2018; 13: e0196588Crossref PubMed Scopus (31) Google Scholar to determine if they had been previously reported. Gene fusion polymorphisms were excluded from further analysis. Gene significance in the relevant tumor type, or other cancer types, was investigated.Reverse Transcription-PCR and Sanger SequencingNovel fusions not previously reported in the literature were verified by reverse transcription (RT)-PCR and Sanger sequencing. RNA was extracted from formalin-fixed, paraffin-embedded (FFPE) blocks using the Absolutely RNA FFPE kit (Agilent, Santa Clara, CA), and converted to cDNA using the High Capacity RNA-to-DNA Kit (Thermo Fisher Scientific, Waltham, MA). Custom primers were designed for each fusion transcript (see Supplemental Table S1 for primer sequences) (Integrated DNA Technologies, Coralville, IA). Nested primers were designed in some instances where initial PCR was unsuccessful. PCR was performed in 20-μL reactions using 20 ng of cDNA, 10 μL of KAPA Hi-Fi Hotstart Polymerase ReadyMix (Roche, Basel, Switzerland), and 0.5 μL each forward and reverse primer. The PCR program was as follows: 95°C for 3 minutes, then 40 cycles of 98°C for 30 seconds, 60°C for 30 seconds, and 72°C for 30 seconds; followed by 72°C for 5 minutes. Resulting PCR products were gel electrophoresed and visualized to verify correct band size. PCR products were purified using the QiaQuick PCR Purification Kit (Qiagen, Hilden, Germany) and Sanger sequenced (GeneWiz, South Plainfield, NJ). Sanger sequencing results were compared with NGS fusion transcripts.ResultsSample CharacteristicsA total of 153 cases corresponding to 153 patients met inclusion criteria. A total of 138 cases were successfully sequenced with adequate quality control metrics (90%) (Figure 1). There were 110 sarcomas, with 31 sarcomas not otherwise specified, 25 Ewing sarcomas, and 13 synovial sarcomas, among others (Table 1 and Supplemental Table S2). There were 13 carcinomas, 9 soft tissue neoplasms not classified as sarcomas, 3 neuroblastomas/Wilms tumors, and 3 neoplasms not further classified. FISH testing was performed in 117 cases, RT-PCR was performed in 3 cases, and DNA-based NGS was performed in 8 cases. In addition, both FISH and DNA-based NGS were performed in eight cases, and both FISH and RT-PCR were performed in two cases. During the clinical workup, a positive fusion was found in 59 specimens (43%), whereas no fusions were found in 79 specimens (57%) (Figure 1 and Supplemental Table S2).Table 1Case DiagnosesDiagnosisNSarcoma NOS31Ewing sarcoma25Synovial sarcoma13Clear cell adenocarcinoma5Phosphaturic mesenchymal tumor5Undifferentiated pleomorphic sarcoma4Low-grade fibromyxoid sarcoma4Alveolar rhabdomyosarcoma3Ewing-like sarcoma3Endometrial stromal sarcoma3Extraskeletal myxoid chondrosarcoma3Alveolar soft part sarcoma2Rhabdomyosarcoma2Clear cell sarcoma2NUT midline carcinoma2Secretory carcinoma2Osteosarcoma1Embryonal rhabdomyosarcoma1Salivary clear cell carcinoma1Fibrosarcoma1Neuroblastoma1Ganglioneuroblastoma1Poorly differentiated neuroendocrine carcinoma1Gliosarcoma1Sclerosing epithelioid fibrosarcoma1High-grade neoplasm NOS1Fibromyxoid neoplasm1Epithelioid myofibroblastoma1Epithelioid hemangioendothelioma1Dedifferentiated liposarcoma1Carcinoma1Wilms tumor1Basal cell adenocarcinoma1Low-grade neoplasm NOS1Chondrosarcoma1Dedifferentiated adamantinoma1DFSP/giant cell fibroblastoma1Myxofibrosarcoma1Uterine sarcoma1Myxoid sarcoma1Myxoid/round cell liposarcoma1Myoepithelioma1Mesenchymal chondrosarcoma1MPNST1Low-grade myxoid sarcoma1Total138DFSP, dermatofibrosarcoma protuberans; MPNST, malignant peripheral nerve sheath tumor; NOS, not otherwise specified; NUT, nuclear protein in testis. Open table in a new tab RNA quality, as measured by DV200 values, ranged from 11% to 96%, with a median of 62% (Supplemental Table S2). There was an average of 4,353,760 downsampled reads per sample, with a range from 2,106,703 to 8,289,134. The median percentage on-target read was 89.1% (range, 77.7% to 95.5%) (Figure 2A ). A positive fusion was detected in 80 of 138 (58%) cases, whereas no gene fusions were detected above threshold in 58 of 138 (42%) cases. Within the 80 cases with positive fusions, the median number of supporting junction reads was 50, and the median number of spanning reads was 16 (Figure 2B-C). A total of 9 of 138 (7%) cases had more than one fusion detected, with up to four in one case. Three fusion artifacts (fusion calls found in greater than three cases with the exact same breakpoints, not found in the literature) were identified in 11 cases and were excluded from further analysis. Fusion artifacts may be introduced during assembly due to sequence-similar genes.11Haas B.J. Dobin A. Stransky N. Li B. Yang X. Tickle T. Bankapur A. Ganote C. Doak T.G. Pochet N. Sun J. Wu C.J. Gingeras T.R. Regev A. STAR-fusion: fast and accurate fusion transcript detection from RNA-Seq.bioRxiv. 2017; ([Preprint] doi:)10.1101/120295Google Scholar,15Haas B.J. Dobin A. Li B. Stransky N. Pochet N. Regev A. Accuracy assessment of fusion transcript detection via read-mapping and de novo fusion transcript assembly-based methods.Genome Biol. 2019; 20: 213Crossref PubMed Scopus (184) Google Scholar A total of 4 of 138 (3%) cases had a known germline fusion polymorphism, TFG-GPR128.16Chase A. Ernst T. Fiebig A. Collins A. Grand F. Erben P. Reiter A. Schreiber S. Cross N.C.P. TFG, a target of chromosome translocations in lymphoma and soft tissue tumors, fuses to GPR128 in healthy individuals.Haematologica. 2010; 95: 20-26Crossref PubMed Scopus (47) Google Scholar This polymorphism was not included in further analyses.Figure 2Quality metrics and supporting read counts. A: Histogram of percentage on-target reads. B and C: Dot plot of number of supporting spanning reads, with horizontal line showing median of 16 (B), and number of supporting junction reads in fusion-positive cases, with horizontal line showing median of 50 (C). Each dot represents one case. D: Stacked bar graph showing mean number of fusions called in a subset of downsampled and nondownsampled cases. Black bar shows mean number of fusions called above pipeline cutoffs for spanning and junction reads, not including artifacts (0.65 ± 0.11 for downsampled and 0.75 ± 0.14 for nondownsampled; P = 0.80 by the Sidak multiple-comparisons test). Gray bar shows other fusions called, including those below pipeline cutoffs, artifacts, and read-through transcripts (1.38 ± 0.19 for downsampled and 2.71 ± 0.31 for nondownsampled; P < 0.0001 by the Sidak multiple-comparisons test). Error bars indicate SEM (D). N = 80 (B and C); N = 48 (D).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fifty-nine cases were randomly downsampled because of elevated numbers of reads, according to manufacturer recommendations. On a subset of these cases, STAR-Fusion was run on both the downsampled and nondownsampled reads for comparison. Downsampling significantly decreased the number of artifacts and fusions below pipeline cutoffs (P < 0.0001), whereas the number of fusions above cutoffs remained similar (Figure 2D). In three cases, fusions were called above cutoff in the nondownsampled reads that were not above cutoff in the downsampled reads. However, in all three of these cases, the downsampled reads contained a dominant fusion with higher numbers of supporting reads compared with these missed fusions. The results support the manufacturer recommendation of a target of 3 million reads per sample.Concordance with Previous FISH, RT-PCR, and DNA-NGS TestingThe overall concordance of the RNA-fusion panel with previous testing results was asssessed (Figure 1). For the 59 patient samples with previously detected gene rearrangements, the RNA-fusion panel concordantly identified 50 fusions (85%). In two cases, the rearranged gene (FN1) was not included in the RNA fusion panel. In the remaining seven discordant cases, quality control metrics for RNA sequencing, including DV200 and percentage on-target reads, were adequate (Table 2). Three of these discordant cases were randomly downsampled; analysis of nondownsampled reads did not demonstrate any previously detected fusions above threshold. Manual inspection using Integrated Genomics Viewer showed adequate read coverage (at least 50×) at putatively rearranged genes (based on prior test method) for all seven cases. This verified that the negative RNA-sequencing results were not due to lack of coverage at these sites. The initial testing modalities varied, including FISH (four cases), RT-PCR (one case), and both FISH and DNA NGS (two cases). In the specimens with positive FISH results, most had a high percentage of nuclei with rearrangements (>50%), although in one case, only 11.5% of nuclei contained a break-apart signal. In one Ewing sarcoma specimen, initial FISH and DNA NGS results were conflicting: FISH was negative for EWSR1 rearrangement (1% rearranged nuclei), whereas DNA NGS revealed an EWSR1-ERG fusion. RNA sequencing in this specimen detected an EWSR1-ERG fusion at low levels, below the pipeline cutoffs, with only two junction reads and one spanning read (Table 2).Table 2Discordant Cases by Initial Testing Modality and RNA SequencingPathologist diagnosisInitial test typeInitial fusions assayedInitial test resultRNA-sequencing resultDV200On-target reads, %Break-apart nuclei by FISH, %Blocks usedPositive by initial testing modality, negative by RNA sequencing Clear cell adenocarcinomaFISH and NGSALK rearrangement (BA)PositiveNo fusions above threshold8189.5UnknownUnknown Clear cell sarcomaFISHEWSR1 rearrangement (BA)PositiveNo fusions above threshold4789.373.5% of 200 nucleiDifferent Endometrial stromal sarcomaFISHJAZF1 rearrangement (BA)PositiveNo fusions above threshold4989.799% of 100 nucleiDifferent Ewing sarcomaRT-PCREWSR1-FLI1PositiveNo fusions above threshold8387.8N/ADifferent Epithelioid myofibroblastomaFISHEWSR1 rearrangement (BA)PositiveNo fusions above threshold6793.211.5% of 200 nucleiDifferent Ewing sarcomaFISH and NGSEWSR1-ERG1 by NGS;EWSR1 rearrangement (BA FISH)Positive by NGS, negative by FISHNo fusions above threshold (EWSR1-ERG detected below threshold)3888.61% of 200 nuclei (below cutoff)NGS: differentFISH: same Synovial sarcomaFISHSS18 rearrangement (BA)PositiveNo fusions found8187.859% of 200 nucleiDifferentNegative by initial testing modality, positive by RNA sequencing CarcinomaFISHETV6 rearrangement (BA)NegativeNTRK3-ETV630<15% of 200 nucleiSameBA, break-apart; FISH, fluorescence in situ hybridization; N/A, not applicable; NGS, next-generation sequencing. Open table in a new tab To assess the possibility of tumor heterogeneity, the blocks used for testing were investigated in the seven discordant cases. In the Ewing sarcoma case with initially conflicting results, the same block was used for FISH and RNA sequencing (both negative for fusion detection), whereas a different block was used for NGS (with a positive EWSR1-ERG1 fusion). In five cases, a different paraffin-embedded block was used for the prior molecular analysis than for the RNA sequencing. In one case, the block used for prior testing was unknown.Previously Undetected Fusions Identified by RNA SequencingOf the 79 samples in which previous molecular testing showed no gene fusions or indeterminate results, the RNA-fusion panel found fusion genes in 30 cases (38%) (Figure 1). Of the 30 cases with previously undetected fusions (false negatives by routine testing), 27 were initially tested by FISH, 2 were tested by FISH and DNA-based NGS, and 1 was tested by DNA-based NGS alone. In one specimen, initial testing by break-apart FISH failed to reveal an ETV6 rearrangement; however, RNA sequencing demonstrated an NTRK3-ETV6 fusion (Table 2). The same block was used for both assays. The FISH report did not indicate the exact percentage of separate signals, but only that the cutoff for a positive break-apart result was 15%.The 30 cases with previously undetected fusion genes were reviewed, and a literature and database search was conducted. In three cases, the previously unknown fusion detected by RNA sequencing confirmed the initial diagnostic impression (BCOR-ZC3H7B in uterine sarcoma, WWTR1-CAMTA1 in epithelioid hemangioendothelioma, and FUS-DDIT3 in myxoid liposarcoma). In seven specimens, information provided by the targeted RNA-sequencing panel suggested a more specific or altered diagnosis (5% of the 138 total cases analyzed) (Table 3). For example, RNA sequencing uncovered NAB2-STAT6 fusions consistent with solitary fibrous tumor in three cases (Figure 3). In two patients, these were signed out as sarcoma not otherwise specified, whereas one case was assigned a diagnosis of synovial sarcoma. The tumors were hypercellular and consisted of round tumor cells with varying amounts of cytoplasm. Classic features of spindled cells and a hemangiopericytic growth pattern were not evident on hematoxylin and eosin sections (Figure 3). IWahile in each of these cases, a different FISH assay was performed during the diagnostic workup (EWSR1, TFE3, and SS18 break-apart), all were negative.Table 3Additional Diagnostic Information Provided by Targeted RNA SequencingInitial pathologist diagnosisInitial test typeInitial fusions assayedInitial test resultRNA-sequencing resultRevised diagnosisCarcinomaFISHETV6 rearrangement (BA)NegativeNTRK3-ETV6Secretory carcinoma17Skálová A. Vanecek T. Sima R. Laco J. Weinreb I. Perez-Ordonez B. Starek I. Geierova M. Simpson R.H. Passador-Santos F. Ryska A. Leivo I. Kinkor Z. Michal M. Mammary analogue secretory carcinoma of salivary glands, containing the ETV6-NTRK3 fusion gene: a hitherto undescribed salivary gland tumor entity.Am J Surg Pathol. 2010; 34: 599-608Crossref PubMed Scopus (664) Google ScholarUndifferentiated round cell sarcomaFISHEWSR1 rearrangement (BA)NegativeBCOR-CCNB3BCOR-CCNB3 (Ewing-like) sarcoma18Pierron G. Tirode F. Lucchesi C. Reynaud S. Ballet S. Cohen-Gogo S. Perrin V. Coindre J.-M. Delattre O. A new subtype of bone sarcoma defined by BCOR-CCNB3 gene fusion.Nat Genet. 2012; 44: 461-466Crossref PubMed Scopus (334) Google ScholarChondrosarcomaFISHSS18 rearrangement (BA)NegativeFUS-NFATC2FUS-NFATC2 sarcoma of bone19Diaz-Perez J.A. Nielsen G.P. Antonescu C. Taylor M.S. Lozano-Calderon S.A. Rosenberg A.E. EWSR1/FUS-NFATc2 rearranged round cell sarcoma: clinicopathological series of 4 cases and literature review.Hum Pathol. 2019; 90: 45-53Crossref PubMed Scopus (38) Google Scholar,20Lacambra M.D. Loong H. To K.F. Feng X. Taylor G. Pleasance E. Laskin J. Marra M. Griffith J. Yeung H.Y.M. Wong K.-C. Chow C. Kumta S. Ng W.H.A. Tse T. Tong C. Ng T. FUS-NFATc2 sarcoma of bone, a novel molecular entity with aggressive behavior: clinical and molecular pathology findings of two cases.Ann Oncol. 2018; 29: ix125Abstract Full Text Full Text PDF Scopus (0) Google ScholarMesenchymal chondrosarcoma/small round cell sarcomaFISHEWSR1, SS18 rearrangement (BA)NegativeFUS-NFATC2FUS-NFATC2 sarcoma of bone19Diaz-Perez J.A. Nielsen G.P. Antonescu C. Taylor M.S. Lozano-Calderon S.A. Rosenberg A.E. EWSR1/FUS-NFATc2 rearranged round cell sarcoma: clinicopathological series of 4 cases and literature review.Hum Pathol. 2019; 90: 45-53Crossref PubMed Scopus (38) Google Scholar,20Lacambra M.D. Loong H. To K.F. Feng X. Taylor G. Pleasance E. Laskin J. Marra M. Griffith J. Yeung H.Y.M. Wong K.-C. Chow C. Kumta S. Ng W.H.A. Tse T. Tong C. Ng T. FUS-NFATc2 sarcoma of bone, a novel molecular entity with aggressive behavior: clinical and molecular pathology findings of two cases.Ann Oncol. 2018; 29: ix125Abstract Full Text Full Text PDF Scopus (0) Google ScholarEwing-like sarcomaFISHEWSR1, SS18 rearrangement (BA)NegativeEP400-PHF1Ossifying fibromyxoid tumor21Hofvander J. Jo V.Y. Fletcher C.D.M. Puls F. Flucke U. Nilsson J. Magnusson L. Mertens F. PHF1 fusions cause distinct gene expression and chromatin accessibility profiles in ossifying fibromyxoid tumors and mesenchymal cells.Mod Pathol. 2020; 33: 1331-1340Crossref PubMed Scopus (13) Google ScholarSarcoma NOSFISHEWSR1 rearrangement (BA)NegativeNAB2-STAT6Solitary fibrous tumor22Robinson D.R. Wu Y.M. Kalyana-Sundaram S. Cao X. Lonigro R.J. Sung
Referência(s)