Classification of the Four Main Types of Lung Cancer Using a MicroRNA-Based Diagnostic Assay
2012; Elsevier BV; Volume: 14; Issue: 5 Linguagem: Inglês
10.1016/j.jmoldx.2012.03.004
ISSN1943-7811
AutoresShlomit Gilad, Gila Lithwick‐Yanai, Iris Barshack, Sima Benjamin, Irit Krivitsky, Tina Bocker Edmonston, Marluce Bibbo, Craig Thurm, Laurie Horowitz, Yajue Huang, Meora Feinmesser, J. Steve Hou, Brianna Cyr, Ilanit Burnstein, Hadas Gibori, Nir Dromi, Mats Sanden, Michal Kushnir, Ranit Aharonov,
Tópico(s)RNA modifications and cancer
ResumoFor patients with primary lung cancer, accurate determination of the tumor type significantly influences treatment decisions. However, techniques and methods for lung cancer typing lack standardization. In particular, owing to limited tumor sample amounts and the poor quality of some samples, the classification of primary lung cancers using small preoperative biopsy specimens presents a diagnostic challenge using current tools. We previously described a microRNA-based assay (miRview squamous; Rosetta Genomics Ltd., Rehovot, Israel) that accurately differentiates between squamous and nonsquamous non–small cell lung cancer. Herein, we describe the development and validation of an assay that differentiates between the four main types of lung cancer: squamous cell carcinoma, nonsquamous non–small cell lung cancer, carcinoid, and small cell carcinoma. The assay, miRview lung (Rosetta Genomics Ltd.), is based on the expression levels of eight microRNAs, measured using a sensitive quantitative RT-PCR platform. It was validated on an independent set of 451 samples, more than half of which were preoperative cytologic samples (fine-needle aspiration and bronchial brushing and washing). The assay returned a result for more than 90% of the samples with overall accuracy of 94% (95% CI, 91% to 96%), with similar performance observed in pathologic and cytologic samples. Thus, miRview lung is a simple and reliable diagnostic assay that offers an accurate and standardized classification tool for primary lung cancer using pathologic and cytologic samples. For patients with primary lung cancer, accurate determination of the tumor type significantly influences treatment decisions. However, techniques and methods for lung cancer typing lack standardization. In particular, owing to limited tumor sample amounts and the poor quality of some samples, the classification of primary lung cancers using small preoperative biopsy specimens presents a diagnostic challenge using current tools. We previously described a microRNA-based assay (miRview squamous; Rosetta Genomics Ltd., Rehovot, Israel) that accurately differentiates between squamous and nonsquamous non–small cell lung cancer. Herein, we describe the development and validation of an assay that differentiates between the four main types of lung cancer: squamous cell carcinoma, nonsquamous non–small cell lung cancer, carcinoid, and small cell carcinoma. The assay, miRview lung (Rosetta Genomics Ltd.), is based on the expression levels of eight microRNAs, measured using a sensitive quantitative RT-PCR platform. It was validated on an independent set of 451 samples, more than half of which were preoperative cytologic samples (fine-needle aspiration and bronchial brushing and washing). The assay returned a result for more than 90% of the samples with overall accuracy of 94% (95% CI, 91% to 96%), with similar performance observed in pathologic and cytologic samples. Thus, miRview lung is a simple and reliable diagnostic assay that offers an accurate and standardized classification tool for primary lung cancer using pathologic and cytologic samples. Lung cancer is the leading cause of cancer-related death in men and women worldwide. In the United States, there were estimated to be more than 220,000 new lung cancer cases in 2011 and more than 156,000 estimated deaths (American Cancer Society: Cancer Facts and Figures 2011, http://www.cancer.org/Research/CancerFactsFigures/CancerFactsFigures/cancer-facts-figures-2011, last accessed September 12, 2011). Eighty-five percent to 90% of these lung cancers are non–small cell lung cancers (NSCLCs). This heterogeneous group is subclassified into squamous cell carcinoma, which constitutes 25% to 30% of lung cancers; adenocarcinoma, which constitutes approximately 40% of lung cancers; and large cell carcinoma, which constitutes 10% to 15% of lung cancers (http://www.cancer.org/Cancer/LungCancer-Non-SmallCell/DetailedGuide/non-small-cell-lung-cancer-what-is-non-small-cell-lung-cancer, last accessed September 12, 2011). The second main type of lung cancer is small cell lung cancer (SCLC), which accounts for 10% to 15% of all lung cancers. Along with the two main types of lung cancer, other tumors can occur in the lungs, including carcinoid tumors, that account for 1% to 2% of lung tumors. In many cases, a cytologic sample is the only available material for the diagnosis of lung cancer. Cytologic methods are widely used in the preoperative setting because of the ease of sample acquisition and minimal trauma to patients. These methods include bronchoscopy methods (bronchial brushing and washing) and the fine-needle aspiration (FNA) procedure, which is performed for tumors that cannot be reached by bronchoscopy. Small diagnostic samples are more likely to be misclassified than are samples obtained via resection due to the paucity of tumor cells, the absence of tissue architecture, and sampling errors.1Field R.W. Smith B.J. Platz C.E. Robinson R.A. Neuberger J.S. Brus C.P. Lynch C.F. Lung cancer histologic type in the surveillance, epidemiology, and end results registry versus independent review.J Natl Cancer Inst. 2004; 96: 1105-1107Crossref PubMed Scopus (95) Google Scholar, 2Jorda M. Gomez-Fernandez C. Garcia M. Mousavi F. Walker G. Mejias A. Fernandez-Castro G. Ganjei-Azar P. P63 differentiates subtypes of nonsmall cell carcinomas of lung in cytologic samples: implications in treatment selection.Cancer. 2009; 117: 46-50Crossref PubMed Scopus (26) Google Scholar, 3Khayyata S. Yun S. Pasha T. Jian B. McGrath C. Yu G. Gupta P. Baloch Z. Value of P63 and CK5/6 in distinguishing squamous cell carcinoma from adenocarcinoma in lung fine-needle aspiration specimens.Diagn Cytopathol. 2009; 37: 178-183Crossref PubMed Scopus (114) Google Scholar, 4Stoll L.M. Johnson M.W. Burroughs F. Li Q.K. Cytologic diagnosis and differential diagnosis of lung carcinoid tumors: a retrospective study of 63 cases with histologic correlation.Cancer Cytopathol. 2010; 118: 457-467Crossref Scopus (20) Google Scholar Therefore, an assay that can accurately classify cancer using cytologic samples is necessary. Securing a clear diagnosis in patients with lung cancer is particularly important since different treatment modalities exist. The distinction between lung neuroendocrine tumors that include SCLC and carcinoid tumors is critical for determining the correct choice of therapy. SCLCs show a highly aggressive behavior and are usually treated with chemotherapy without surgery, whereas typical carcinoids are considered low-grade malignancies that may benefit from surgical resection.5Tanca A. Addis M.F. Pagnozzi D. Cossu-Rocca P. Tonelli R. Falchi G. Eccher A. Roggio T. Fanciulli G. Uzzau S. Proteomic analysis of formalin-fixed, paraffin-embedded lung neuroendocrine tumor samples from hospital archives.J Proteomics. 2011; 74: 359-370Crossref PubMed Scopus (39) Google Scholar The precise subclassification of NSCLC into squamous and nonsquamous tumors is becoming increasingly important due to its effect on therapeutic decisions.6Clark G.M. Zborowski D.M. Santabarbara P. Ding K. Whitehead M. Seymour L. Shepherd F.A. Smoking history and epidermal growth factor receptor expression as predictors of survival benefit from erlotinib for patients with non-small-cell lung cancer in the National Cancer Institute of Canada Clinical Trials Group study BR. 21.Clin Lung Cancer. 2006; 7: 389-394Abstract Full Text PDF PubMed Scopus (155) Google Scholar, 7Scagliotti G.V. Parikh P. von Pawel J. Biesma B. Vansteenkiste J. Manegold C. Serwatowski P. Gatzemeier U. Digumarti R. Zukin M. Lee J.S. Mellemgaard A. Park K. Patil S. Rolski J. Goksel T. de Marinis F. Simms L. Sugarman K.P. Gandara D. Phase III study comparing cisplatin plus gemcitabine with cisplatin plus pemetrexed in chemotherapy-naive patients with advanced-stage non-small-cell lung cancer.J Clin Oncol. 2008; 26: 3543-3551Crossref PubMed Scopus (2884) Google Scholar In addition, some agents may be associated with a greater risk of adverse effects in patients with squamous NSCLC compared with those with nonsquamous NSCLC.8Johnson D.H. Fehrenbacher L. Novotny W.F. Herbst R.S. Nemunaitis J.J. Jablons D.M. Langer C.J. DeVore III, R.F. Gaudreault J. Damico L.A. Holmgren E. Kabbinavar F. Randomized phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated locally advanced or metastatic non-small-cell lung cancer.J Clin Oncol. 2004; 22: 2184-2191Crossref PubMed Scopus (1825) Google Scholar For example, patients with squamous NSCLC should not receive bevacizumab (Avastin; Genetech Inc., South San Fransicso, CA) because of 30% mortality due to fatal hemoptysis.9Herbst R.S. O'Neill V.J. Fehrenbacher L. Belani C.P. Bonomi P.D. Hart L. Melnyk O. Ramies D. Lin M. Sandler A. Phase II study of efficacy and safety of bevacizumab in combination with chemotherapy or erlotinib compared with chemotherapy alone for treatment of recurrent or refractory non small-cell lung cancer.J Clin Oncol. 2007; 25: 4743-4750Crossref PubMed Scopus (375) Google Scholar, 10Sandler A. Bevacizumab in non small cell lung cancer.Clin Cancer Res. 2007; 13: s4613-s4616Crossref PubMed Scopus (59) Google Scholar Routine histopathologic analysis is the current standard of lung tumor classification but has demonstrated interobserver variability among pathologists.1Field R.W. Smith B.J. Platz C.E. Robinson R.A. Neuberger J.S. Brus C.P. Lynch C.F. Lung cancer histologic type in the surveillance, epidemiology, and end results registry versus independent review.J Natl Cancer Inst. 2004; 96: 1105-1107Crossref PubMed Scopus (95) Google Scholar, 3Khayyata S. Yun S. Pasha T. Jian B. McGrath C. Yu G. Gupta P. Baloch Z. Value of P63 and CK5/6 in distinguishing squamous cell carcinoma from adenocarcinoma in lung fine-needle aspiration specimens.Diagn Cytopathol. 2009; 37: 178-183Crossref PubMed Scopus (114) Google Scholar, 11Feinstein A.R. Gelfman N.A. Yesner R. Observer variability in the histopathologic diagnosis of lung cancer.Am Rev Respir Dis. 1970; 101: 671-684PubMed Google Scholar, 12Stang A. Pohlabeln H. Muller K.M. Jahn I. Giersiepen K. Jockel K.H. Diagnostic agreement in the histopathological evaluation of lung cancer tissue in a population-based case-control study.Lung Cancer. 2006; 52: 29-36Abstract Full Text Full Text PDF PubMed Scopus (66) Google Scholar In one study, up to 40% of squamous cell carcinoma or adenocarcinoma of the lung was reclassified when evaluated by a second pathologist.12Stang A. Pohlabeln H. Muller K.M. Jahn I. Giersiepen K. Jockel K.H. Diagnostic agreement in the histopathological evaluation of lung cancer tissue in a population-based case-control study.Lung Cancer. 2006; 52: 29-36Abstract Full Text Full Text PDF PubMed Scopus (66) Google Scholar In another study, the number of cases classified correctly by a panel of cytopathologists was 66% for adenocarcinoma and 53% for squamous cell carcinoma.3Khayyata S. Yun S. Pasha T. Jian B. McGrath C. Yu G. Gupta P. Baloch Z. Value of P63 and CK5/6 in distinguishing squamous cell carcinoma from adenocarcinoma in lung fine-needle aspiration specimens.Diagn Cytopathol. 2009; 37: 178-183Crossref PubMed Scopus (114) Google Scholar Immunohistochemical (IHC) performance is limited by the variable sensitivity and specificity of each marker.13Maeshima A.M. Omatsu M. Tsuta K. Asamura H. Matsuno Y. Immunohistochemical expression of TTF-1 in various cytological subtypes of primary lung adenocarcinoma, with special reference to intratumoral heterogeneity.Pathol Int. 2008; 58: 31-37Crossref PubMed Scopus (23) Google Scholar, 14Wang B.Y. Gil J. Kaufman D. Gan L. Kohtz D.S. Burstein D.E. P63 in pulmonary epithelium, pulmonary squamous neoplasms, and other pulmonary tumors.Hum Pathol. 2002; 33: 921-926Abstract Full Text Full Text PDF PubMed Scopus (155) Google Scholar For example, whereas p63 is a sensitive marker for squamous cells, it is not highly specific; up to 30% of adenocarcinomas and 37% of large cell undifferentiated carcinomas also express p63, probably owing to the expression of p63 in basal cells. The shortcomings associated with current lung cancer classification methods point to the need for a highly accurate, objective, reproducible, and standardized classification tool for the diagnosis of lung cancer. MicroRNAs (miRNAs) are small, 20- to 22-nucleotide, noncoding RNA genes that regulate gene expression at the translational level.15Bartel D.P. MicroRNAs: genomics, biogenesis, mechanism, and function.Cell. 2004; 116: 281-297Abstract Full Text Full Text PDF PubMed Scopus (29561) Google Scholar The mature miRNA is incorporated into a large protein complex, the RNA-induced silencing complex, where it binds a corresponding mRNA and inhibits protein translation. During the past decade, our knowledge about the role of miRNAs in human diseases, including cancer, has grown exponentially.16Calin G.A. Croce C.M. MicroRNA signatures in human cancers.Nat Rev Cancer. 2006; 6: 857-866Crossref PubMed Scopus (6616) Google Scholar, 17Farazi T.A. Spitzer J.I. Morozov P. Tuschl T. miRNAs in human cancer.J Pathol. 2011; 223: 102-115Crossref PubMed Scopus (834) Google Scholar Alteration in the expression of miRNA genes contributes to the pathogenesis of most, if not all, human malignancies.16Calin G.A. Croce C.M. MicroRNA signatures in human cancers.Nat Rev Cancer. 2006; 6: 857-866Crossref PubMed Scopus (6616) Google Scholar, 17Farazi T.A. Spitzer J.I. Morozov P. Tuschl T. miRNAs in human cancer.J Pathol. 2011; 223: 102-115Crossref PubMed Scopus (834) Google Scholar, 18Schetter A.J. Heegaard N.H. Harris C.C. Inflammation and cancer: interweaving microRNA, free radical, cytokine and p53 pathways.Carcinogenesis. 2010; 31: 37-49Crossref PubMed Scopus (519) Google Scholar miRNAs can act as oncogenes or tumor suppressors,19Ha T.Y. MicroRNAs in human diseases: from cancer to cardiovascular disease.Immune Netw. 2011; 11: 135-154Crossref PubMed Google Scholar and a subset of these miRNAs show considerable tissue specificity.20Lu J. Getz G. Miska E.A. Alvarez-Saavedra E. Lamb J. Peck D. Sweet-Cordero A. Ebert B.L. Mak R.H. Ferrando A.A. Downing J.R. Jacks T. Horvitz H.R. Golub T.R. MicroRNA expression profiles classify human cancers.Nature. 2005; 435: 834-838Crossref PubMed Scopus (8225) Google Scholar Therefore, miRNAs are excellent molecular candidates for tumor classification. In previous studies conducted in our lab (Rosetta Genomics, Ltd., Rehovot, Israel), the differential expression of hsa-miR-205 between lung squamous cell carcinoma and nonsquamous NSCLC accurately classified these cancer types with a high degree of sensitivity and specificity.21Bishop J.A. Benjamin H. Cholakh H. Chajut A. Clark D.P. Westra W.H. Accurate classification of non-small cell lung carcinoma using a novel microRNA-based approach.Clin Cancer Res. 2010; 16: 610-619Crossref PubMed Scopus (207) Google Scholar, 22Lebanony D. Benjamin H. Gilad S. Ezagouri M. Dov A. Ashkenazi K. Gefen N. Izraeli S. Rechavi G. Pass H. Nonaka D. Li J. Spector Y. Rosenfeld N. Chajut A. Cohen D. Aharonov R. Mansukhani M. Diagnostic assay based on hsa-miR-205 expression distinguishes squamous from nonsquamous non-small-cell lung carcinoma.J Clin Oncol. 2009; 27: 2030-2037Crossref PubMed Scopus (346) Google Scholar The performance of a mir-205–based assay (miRview squamous; Rosetta Genomics Ltd., Rehovot, Israel) was demonstrated for various clinical samples, such as formalin-fixed, paraffin-embedded samples and cytologic samples.21Bishop J.A. Benjamin H. Cholakh H. Chajut A. Clark D.P. Westra W.H. Accurate classification of non-small cell lung carcinoma using a novel microRNA-based approach.Clin Cancer Res. 2010; 16: 610-619Crossref PubMed Scopus (207) Google Scholar, 22Lebanony D. Benjamin H. Gilad S. Ezagouri M. Dov A. Ashkenazi K. Gefen N. Izraeli S. Rechavi G. Pass H. Nonaka D. Li J. Spector Y. Rosenfeld N. Chajut A. Cohen D. Aharonov R. Mansukhani M. Diagnostic assay based on hsa-miR-205 expression distinguishes squamous from nonsquamous non-small-cell lung carcinoma.J Clin Oncol. 2009; 27: 2030-2037Crossref PubMed Scopus (346) Google Scholar Twenty-one cytologic samples were also compared with matched resection samples from the same patient. The sensitivity for these 21 cases was 95.2%, and the concordance between cytologic and resection samples was 100%.21Bishop J.A. Benjamin H. Cholakh H. Chajut A. Clark D.P. Westra W.H. Accurate classification of non-small cell lung carcinoma using a novel microRNA-based approach.Clin Cancer Res. 2010; 16: 610-619Crossref PubMed Scopus (207) Google Scholar In the present study, we extended this previously developed assay to distinguish between four lung cancer types. We show that the quantification of eight miRNA biomarkers, using the quantitative RT-PCR (RT-qPCR) platform, differentiates patients with SCLC, carcinoid, squamous NSCLC, and nonsquamous NSCLC. We tested the current assay, miRview lung (Rosetta Genomics Ltd.), on a set of 204 pathologic samples and 247 cytologic samples (FNA and bronchial brushing and washing samples) and demonstrated that we can classify these four main lung cancer types with a high level of accuracy. Deidentified retrospective samples were obtained from several sources (Temple University, Philadelphia, PA; Thomas Jefferson University, Philadelphia, PA; Rabin Medical Center, Petah Tikva, Israel; Sheba Medical Center, Tel Hashomer, Israel; ABS Inc., Wilmington, DE; Jamaica Hospital Medical Center, Jamaica, NY; Drexel University College of Medicine, Philadelphia, PA; ProteoGenex, Culver City, CA; BioServe, Beltsville, MD; and Oncomatrix, San Marcos, CA). Approvals were obtained for all the samples in accordance with each institution's institutional review board or equivalent guidelines. Samples included formalin-fixed, paraffin-embedded blocks from resection or biopsies and cell blocks from cytologic procedures, including FNA, bronchial brushing, and bronchial washing (Table 1). Each sample was conveyed with its clinical diagnosis (reference diagnosis) and results of IHC stains when available, as well as data from pathologic/cytologic evaluation. Representative blocks were sectioned into 1.5-mL microcentrifuge tubes or were mounted on slides (three to five 10-μm sections), and H&E-stained slides were obtained with each block to evaluate percentage of tumor cellular content at sectioning. Validation samples were blinded to the investigators performing the validation assays.Table 1Tumor Samples Used in the StudySampleDiscovery phase⁎The study phases are described in Results.Assay development phase⁎The study phases are described in Results.Validation phase⁎The study phases are described in Results.Custom microarrayAgilent microarrayPCR discoveryFinal probe list selectionClassifier determinationPathologic32 (32)26 (18)36 (16)141 (98)138 (6)204 (204) Resection32 (32)26 (18)36 (16)120 (77)118 (6)125 (125) Squamous7 (7)2 (2)8 (6)38 (27)35 (0)19 (19) Nonsquamous12 (12)3 (3)4 (4)40 (28)41 (4)58 (58) Carcinoid6 (6)13 (9)12 (3)26 (13)27 (2)35 (35) Small7 (7)8 (4)12 (3)16 (9)15 (0)13 (13) Biopsy0 (0)0 (0)0 (0)21 (21)20 (0)79 (79) Squamous0 (0)0 (0)0 (0)12 (12)12 (0)18 (18) Nonsquamous0 (0)0 (0)0 (0)4 (4)4 (0)19 (19) Carcinoid†Cytologic procedures on carcinoid tumors are rarely performed, and, hence, we could not locate cytologic carcinoid samples to include in this study.0 (0)0 (0)0 (0)0 (0)0 (0)0 (0) Small0 (0)0 (0)0 (0)5 (5)4 (0)42 (42)Cytologic0 (0)0 (0)0 (0)85 (85)78 (4)247 (247) FNA0 (0)0 (0)0 (0)85 (85)78 (4)207 (207) Squamous0 (0)0 (0)0 (0)28 (28)30 (3)90 (90) Nonsquamous0 (0)0 (0)0 (0)33 (33)28 (0)95 (95) Carcinoid0 (0)0 (0)0 (0)0 (0)0 (0)0 (0) Small0 (0)0 (0)0 (0)24 (24)20 (1)22 (22) Brushing and washing0 (0)0 (0)0 (0)0 (0)0 (0)40 (40) Squamous0 (0)0 (0)0 (0)0 (0)0 (0)21 (21) Nonsquamous0 (0)0 (0)0 (0)0 (0)0 (0)14 (14) Carcinoid0 (0)0 (0)0 (0)0 (0)0 (0)0 (0) Small0 (0)0 (0)0 (0)0 (0)0 (0)5 (5)In parentheses are the numbers of new samples relative to the samples in all preceding columns. The study phases are described in Results.† Cytologic procedures on carcinoid tumors are rarely performed, and, hence, we could not locate cytologic carcinoid samples to include in this study. Open table in a new tab In parentheses are the numbers of new samples relative to the samples in all preceding columns. Four histologic diagnoses were included in this study: SCLC, carcinoid, squamous NSCLC, and nonsquamous NSCLC. A diagnosis of nonsquamous NSCLC includes adenocarcinoma and large cell NSCLC. Samples with other diagnoses (eg, large cell neuroendocrine carcinoma) were excluded. For samples used in the biomarker discovery and assay development phases, tumor tissue diagnosis was based on the medical records, and all pathologic slides were reevaluated by an independent pathologist. For samples used for validation, the reference diagnosis was verified by an independent pathologist based on H&E slides and results of IHC stains, if available (IHC staining was performed on 38% overall and on 43% of cytologic samples). Formalin-fixed, paraffin-embedded blocks from resections or biopsies were microdissected, if needed, to increase the tumor cellular content to the minimal requirement of 50%. Cell blocks were used if the tumor cellular content of the nonerythrocyte cells reached 50%. Two independent pathologists (T.B.E. and M.S.) reviewed the bronchial brushing and washing samples used for validation. For these samples, at least one of the two independent pathologists must have assigned the same diagnosis as the reference diagnosis for them to be used in the validation. Total RNA was extracted from pathologic samples as previously described.23Nass D. Rosenwald S. Meiri E. Gilad S. Tabibian-Keissar H. Schlosberg A. Kuker H. Sion-Vardy N. Tobar A. Kharenko O. Sitbon E. Lithwick Yanai G. Elyakim E. Cholakh H. Gibori H. Spector Y. Bentwich Z. Barshack I. Rosenfeld N. MiR-92b and miR-9/9* are specifically expressed in brain primary tumors and can be used to differentiate primary from metastatic brain tumors.Brain Pathol. 2009; 19: 375-383Crossref PubMed Scopus (169) Google Scholar, 24Rosenfeld N. Aharonov R. Meiri E. Rosenwald S. Spector Y. Zepeniuk M. Benjamin H. Shabes N. Tabak S. Levy A. Lebanony D. Goren Y. Silberschein E. Targan N. Ben-Ari A. Gilad S. Sion-Vardy N. Tobar A. Feinmesser M. Kharenko O. Nativ O. Nass D. Perelman M. Yosepovich A. Shalmon B. Polak-Charcon S. Fridman E. Avniel A. Bentwich I. Bentwich Z. Cohen D. Chajut A. Barshack I. MicroRNAs accurately identify cancer tissue origin.Nature Biotechnol. 2008; 26: 462-469Crossref PubMed Scopus (859) Google Scholar Formalin-fixed, paraffin-embedded samples were deparaffinized with xylene, washed in ethanol, and digested with proteinase K. RNA was extracted using acid phenol:chloroform followed by ethanol precipitation and DNase. Custom microarrays were prepared as previously described.24Rosenfeld N. Aharonov R. Meiri E. Rosenwald S. Spector Y. Zepeniuk M. Benjamin H. Shabes N. Tabak S. Levy A. Lebanony D. Goren Y. Silberschein E. Targan N. Ben-Ari A. Gilad S. Sion-Vardy N. Tobar A. Feinmesser M. Kharenko O. Nativ O. Nass D. Perelman M. Yosepovich A. Shalmon B. Polak-Charcon S. Fridman E. Avniel A. Bentwich I. Bentwich Z. Cohen D. Chajut A. Barshack I. MicroRNAs accurately identify cancer tissue origin.Nature Biotechnol. 2008; 26: 462-469Crossref PubMed Scopus (859) Google Scholar Six hundred seventy-five DNA oligonucleotide probes representing human miRNAs were spotted in triplicate on coated microarray slides (Nexterion Slide E; Schott, Mainz, Germany). Total RNA (3.5 μg) was labeled by ligation of an RNA-linker, p-rCrU-Cy/dye (Cy3 or Cy5; Eurogentec, San Diego, CA), to the 3′ end. Slides were incubated with the labeled RNA for 12 to 16 hours at 42°C and then were washed twice. Arrays were scanned at a resolution of 10 μm, and images were analyzed using SpotReader software version 1.3.1.0 (Niles Scientific, Portola Valley, CA). In addition, Agilent custom-designed arrays were used (Agilent Technologies, Santa Clara, CA). The arrays included probes for more than 900 known human miRNAs25Griffiths-Jones S. Saini H.K. van Dongen S. Enright A.J. miRBase: tools for microRNA genomics.Nucleic Acids Res. 2008; 36: D154-D158Crossref PubMed Scopus (3622) Google Scholar printed in triplicate, as well as control probes and probes for putative miRNA sequences. Total RNA (1.7 μg) was labeled for these arrays. For the custom microarrays and the Agilent custom-designed arrays, microarray spots were combined and signals normalized as described previously.24Rosenfeld N. Aharonov R. Meiri E. Rosenwald S. Spector Y. Zepeniuk M. Benjamin H. Shabes N. Tabak S. Levy A. Lebanony D. Goren Y. Silberschein E. Targan N. Ben-Ari A. Gilad S. Sion-Vardy N. Tobar A. Feinmesser M. Kharenko O. Nativ O. Nass D. Perelman M. Yosepovich A. Shalmon B. Polak-Charcon S. Fridman E. Avniel A. Bentwich I. Bentwich Z. Cohen D. Chajut A. Barshack I. MicroRNAs accurately identify cancer tissue origin.Nature Biotechnol. 2008; 26: 462-469Crossref PubMed Scopus (859) Google Scholar miRNA amounts were quantified using an RT-qPCR method described previously.22Lebanony D. Benjamin H. Gilad S. Ezagouri M. Dov A. Ashkenazi K. Gefen N. Izraeli S. Rechavi G. Pass H. Nonaka D. Li J. Spector Y. Rosenfeld N. Chajut A. Cohen D. Aharonov R. Mansukhani M. Diagnostic assay based on hsa-miR-205 expression distinguishes squamous from nonsquamous non-small-cell lung carcinoma.J Clin Oncol. 2009; 27: 2030-2037Crossref PubMed Scopus (346) Google Scholar, 26Gilad S. Meiri E. Yogev Y. Benjamin S. Lebanony D. Yerushalmi N. Benjamin H. Kushnir M. Cholakh H. Melamed N. Bentwich Z. Hod M. Goren Y. Chajut A. Serum microRNAs are promising novel biomarkers.PLoS One. 2008; 3: e3148Crossref PubMed Scopus (1179) Google Scholar For the discovery experiments, samples were assayed for the expression of 110 probes (109 miRNA probes and a probe for the small RNA U6). For the final assay, raw CTs were used (ie, no signal normalization was performed). In Figure 1, to visualize the differential expression of miRNAs, normalized CTs are shown: Each sample was scaled by subtracting from the CT of each miRNA the average CT of two normalizer miRNAs (two of the initial miRNA probes chosen for normalization that were measured for all the samples) for the sample, and adding back a scaling constant (the average CT for the normalizers over the entire sample set). The CT scale was then inverted such that high values represent high expression by subtracting the resulting value from 50. The final assay is based on the CTs of eight miRNAs (Table 2). RNA is extracted and cDNA is created in batches of up to 11 RNAs together with a negative control and a positive control. The negative control serves to detect potential contaminations and should not give any signal in the PCR. It is created for each batch by performing the cDNA creation process on double-distilled water. The positive control is a specific mix of RNA samples that should meet defined CT ranges in the assay. If either the positive or negative control fails, the process is repeated. Each of the eight tested miRNAs comprises a row in a 96-well plate, and each sample is run in duplicate (two columns per sample). Quality assessment (QA) of each well is based on the fluorescence amplification curve using thresholds on the maximal fluorescence and the linear slope as a function of the measured CT. For each miRNA, the CT value used is that calculated by taking the average CT of the two repeats. If one or both repeats fail, or if the signal difference of the probe duplicate wells exceeds a threshold, data for the sample are discarded and the sample is rerun. QA of the PCR experiment for the whole sample is based on maximal CT expression for the miRNAs hsa-miR-21, hsa-miR-375, and hsa-miR-205, whereby if all three miRNAs have high CT values, the sample is considered to have insufficient expression.Table 2miRNAs Used in the AssaymiRNASequenceForward primer sequence⁎The reverse primer is universal and its sequence is 5′-GCGAGCACAGAATTAATACGAC-3′.hsa-miR-106a5′-AAAAGTGCTTACAGTGCAGGTAG-3′5′-CAGTCATTTGGAAAAGTGCTTACAGTGCA-3′hsa-miR-125a-5p5′-TCCCTGAGACCCTTTAACCTGTGA-3′5′-GCTCCCTGAGACCCTTTAACCTGT-3′hsa-miR-129-3p5′-AAGCCCTTACCCCAAAAAGCAT-3′5′-GCAAGCCCTTACCCCAAAAAGCAT-3′hsa-miR-2055′-TCCTTCATTCCACCGGAGTCTG-3′5′-CAGTCATTTGGCTCCTTCATTCCACCGGA-3′hsa-miR-215′-TAGCTTATCAGACTGATGTTGA-3′5′-CAGTCATTTGGCTAGCTTATCAGACTGA-3′hsa-miR-29b5′-TAGCACCATTTGAAATCAGTGTT-3′5′-CATTTGGTAGCACCATTTGAAATCAGTGTT-3′hsa-miR-3755′-TTTGTTCGTTCGGCTCGCGTGA-3′5′-CAGTCATTTGGGTTTGTTCGTTCGGCTC-3′hsa-miR-75′-TGGAAGACTAGTGATTTTGTTGT-3′5′-CAGTCATTTGGCTGGAAGACTAGTGATT-3′ The reverse primer is universal and its sequence is 5′-GCGAGCACAGAATTAATACGAC-3′. Open table in a new tab The classifier used is a K-nearest neighbor (KNN) classifier (K = 9), using a Pearson correlation distance metric over the CT values of the eight assay miRNAs. The KNN prediction is accompanied by a confidence measure, v, which is the number of neighbors (up to 9) agreeing with the KNN-reported result. If v ≥ 7, the sample is considered to be classified with high confidence. For samples classified with low confidence (v < 7), no result is generated. One hundred twelve samples were tested in two laboratories: the Rosetta Genomics Israel laboratory (RG-IL) and the Rosetta Genomics US laboratory (RGL-US). Sixty-nine samples were extracted at the RG-IL, and the resulting RNA was tested in both laboratories. In addition, for 43 samples, each laboratory extracted sections from the same paraffin block and classified the sample using the assay protocol. Positive predictive value (PPV) was calculated in two ways. First, as is widely do
Referência(s)