The Clinical Utility of miR-21 as a Diagnostic and Prognostic Marker for Renal Cell Carcinoma
2012; Elsevier BV; Volume: 14; Issue: 4 Linguagem: Inglês
10.1016/j.jmoldx.2012.02.003
ISSN1943-7811
AutoresHala Faragalla, Youssef M. Youssef, Andreas Scorilas, Bishoy Khalil, Nicole M. White, Salvador Mejia‐Guerrero, Heba Khella, Michael A.S. Jewett, Andrew Evans, Zsuzsanna Lichner, Georg A. Bjarnason, Linda Sugar, Magdy I. Attalah, George M. Yousef,
Tópico(s)Renal and related cancers
ResumoRenal cell carcinoma (RCC) is the most common neoplasm of the kidney. Increasing evidence suggests that microRNAs are dysregulated in RCC and are important factors in RCC pathogenesis. miR-21 is a known oncogene with tumor-promoting effects in many types of cancer. In this study, we analyzed miR-21 in 121 cases of healthy kidney and different RCC subtypes, including clear cell (ccRCC), papillary (pRCC), chromophobe (chRCC), and oncocytoma. Total RNA was extracted, and the expression of miR-21 was measured with real-time quantitative RT-PCR using miR-21–specific probes. The expression of miR-21 was significantly up-regulated in RCC compared with healthy kidney. There was a significant difference in the expression levels between RCC subtypes, with the highest levels of expression in ccRCC and pRCC subtypes. miR-21 expression distinguished ccRCC and pRCC from chRCC and oncocytoma with 90% specificity (95% CI, 63.9% to 98.1%) and 83% sensitivity (95% CI, 53.5% to 97.6%). Significantly higher miR-21 levels were associated with higher stage and grade. Patients who were miR-21 positive had statistically significant shorter disease-free and overall survival rates. Thus, miR-21 is up-regulated in RCC, and its expression levels can be used as a diagnostic marker to distinguish ccRCC and pRCC from chRCC and oncocytoma. Moreover, it has potential as a prognostic marker in RCC, although it is not independent of tumor stage and grade. Renal cell carcinoma (RCC) is the most common neoplasm of the kidney. Increasing evidence suggests that microRNAs are dysregulated in RCC and are important factors in RCC pathogenesis. miR-21 is a known oncogene with tumor-promoting effects in many types of cancer. In this study, we analyzed miR-21 in 121 cases of healthy kidney and different RCC subtypes, including clear cell (ccRCC), papillary (pRCC), chromophobe (chRCC), and oncocytoma. Total RNA was extracted, and the expression of miR-21 was measured with real-time quantitative RT-PCR using miR-21–specific probes. The expression of miR-21 was significantly up-regulated in RCC compared with healthy kidney. There was a significant difference in the expression levels between RCC subtypes, with the highest levels of expression in ccRCC and pRCC subtypes. miR-21 expression distinguished ccRCC and pRCC from chRCC and oncocytoma with 90% specificity (95% CI, 63.9% to 98.1%) and 83% sensitivity (95% CI, 53.5% to 97.6%). Significantly higher miR-21 levels were associated with higher stage and grade. Patients who were miR-21 positive had statistically significant shorter disease-free and overall survival rates. Thus, miR-21 is up-regulated in RCC, and its expression levels can be used as a diagnostic marker to distinguish ccRCC and pRCC from chRCC and oncocytoma. Moreover, it has potential as a prognostic marker in RCC, although it is not independent of tumor stage and grade. Renal cell carcinoma (RCC) is the most common kidney tumor.1Chow W.H. Devesa S.S. Warren J.L. Fraumeni Jr, J.F. Rising incidence of renal cell cancer in the United States.JAMA. 1999; 281: 1628-1631Google Scholar Interestingly, it is one of few cancers that showed an increase in incidence and mortality during the past 20 years, despite earlier detection and introduction of new therapies. RCC encompasses a heterogeneous group of cancers, each with distinct morphological characteristics and associated cytogenetic changes, suggesting that different molecular mechanisms are responsible for their development. The most common is the clear cell subtype (ccRCC), which accounts for approximately 75% to 80% of cases. Other less common subtypes include papillary RCC (pRCC), chromophobe RCC (chRCC), and collecting duct carcinoma. Oncocytoma is one of the common benign neoplasms of the kidney that shares genomic and morphological similarities with chRCC. Characterization of the RCC subtypes relies on histological appearance, which is usually distinct for each type. There are, however, many cases in which the morphological criteria are not conclusive. For instance, distinguishing the eosinophilic variant of ccRCC from either chRCC or oncocytoma can be challenging. Furthermore, 2% to 3% of cases cannot be categorized, and interobserver variations significantly limit the accuracy of histological classification.2Ficarra V. Martignoni G. Galfano A. Novara G. Gobbo S. Brunelli M. Pea M. Zattoni F. Artibani W. Prognostic role of the histologic subtypes of renal cell carcinoma after slide revision.Eur Urol. 2006; 50: 786-793Google Scholar This is particularly important in small biopsy specimens, in which the diagnostic material is limited. Added to this complexity is the presence of hybrid tumors with more than one subtype in the same tumor and the recently recognized entities with features overlapping with classic subtypes.3Srigley J.R. Delahunt B. Uncommon and recently described renal carcinomas.Mod Pathol. 2009; 22: S2-S23Google Scholar This necessitates the generation of the unclassified category of RCC. Accurate classification is important because subtypes differ in their prognosis. Subtypes also show a different response to targeted therapies.4Klatte T. Patard J.J. Wunderlich H. Goel R.H. Lam J.S. Junker K. Schubert J. Bohm M. Allhoff E.P. Kabbinavar F.F. Crepel M. Cindolo L. De La T.A. Tostain J. Mejean A. Soulie M. Bellec L. Bernhard J.C. Ferriere J.M. Pfister C. Albouy B. Colombel M. Zisman A. Belldegrun A.S. Pantuck A.J. Metachronous bilateral renal cell carcinoma: risk assessment, prognosis and relevance of the primary-free interval.J Urol. 2007; 177: 2081-2086Google Scholar A more accurate classification using molecular approaches is, therefore, critical and can help in selecting the optimal treatment option.5Diamandis M. White N.M. Yousef G.M. Personalized medicine: marking a new epoch in cancer patient management.Mol Cancer Res. 2010; 8: 1175-1187Google Scholar Recent evidence showed that molecular signatures can classify RCC subtypes more accurately than morphological characteristics.6Arsanious A. Bjarnason G.A. Yousef G.M. From bench to bedside: current and future applications of molecular profiling in renal cell carcinoma.Mol Cancer. 2009; 8: 20Google Scholar, 7Fridman E. Dotan Z. Barshack I. David M.B. Dov A. Tabak S. Zion O. Benjamin S. Benjamin H. Kuker H. Avivi C. Rosenblatt K. Polak-Charcon S. Ramon J. Rosenfeld N. Spector Y. Accurate molecular classification of renal tumors using microRNA expression.J Mol Diagn. 2010; 12: 687-696Google Scholar, 8Youssef Y.M. White N.M. Grigull J. Krizova A. Samy C. Mejia-Guerrero S. Evans A. Yousef G.M. Accurate molecular classification of kidney cancer subtypes using microRNA signature.Eur Urol. 2011; 59: 721-730Google Scholar The prognosis of RCC can vary widely. Detecting recurrences early can improve patient outcomes because the likelihood of a favorable response to systemic treatment is greater when the metastatic burden is limited, and surgical resection of a single or limited number of metastases can result in longer survival.9Hofmann H.S. Neef H. Krohe K. Andreev P. Silber R.E. Prognostic factors and survival after pulmonary resection of metastatic renal cell carcinoma.Eur Urol. 2005; 48: 77-81Google Scholar The anatomical extent, or stage, of disease is the most useful prognostic factor for patients with RCC, but this is not always accurate. The most commonly used prognostic models for patients with metastatic disease are based on clinical parameters.10Mekhail T.M. Abou-Jawde R.M. Boumerhi G. Malhi S. Wood L. Elson P. Bukowski R. Validation and extension of the Memorial Sloan-Kettering prognostic factors model for survival in patients with previously untreated metastatic renal cell carcinoma.J Clin Oncol. 2005; 23: 832-841Google Scholar A more accurate assessment of RCC prognosis is urgently needed to better guide patient management. MicroRNAs (miRNAs) are small noncoding RNA nucleotides that post-transcriptionally regulate expression of their target proteins. An aberrant pattern of miRNA expression has been observed in RCC.7Fridman E. Dotan Z. Barshack I. David M.B. Dov A. Tabak S. Zion O. Benjamin S. Benjamin H. Kuker H. Avivi C. Rosenblatt K. Polak-Charcon S. Ramon J. Rosenfeld N. Spector Y. Accurate molecular classification of renal tumors using microRNA expression.J Mol Diagn. 2010; 12: 687-696Google Scholar, 11Chow T.F. Youssef Y.M. Lianidou E. Romaschin A.D. Honey R.J. Stewart R. Pace K.T. Yousef G.M. Differential expression profiling of microRNAs and their potential involvement in renal cell carcinoma pathogenesis.Clin Biochem. 2010; 43: 150-158Google Scholar, 12Jung M. Mollenkopf H.J. Grimm C. Wagner I. Albrecht M. Waller T. Pilarsky C. Johannsen M. Stephan C. Lehrach H. Nietfeld W. Rudel T. Jung K. Kristiansen G. MicroRNA profiling of clear cell renal cell cancer identifies a robust signature to define renal malignancy.J Cell Mol Med. 2009; 13: 3918-3928Google Scholar, 13White N.M. Bao T.T. Grigull J. Youssef Y.M. Girgis A. Diamandis M. Fatoohi E. Metias M. Honey R.J. Stewart R. Pace K.T. Bjarnason G.A. Yousef G.M. miRNA profiling for clear cell renal cell carcinoma: biomarker discovery and identification of potential controls and consequences of miRNA dysregulation.J Urol. 2011; 186: 1077-1083Google Scholar miRNAs may be involved in RCC pathogenesis,14White N.M. Yousef G.M. MicroRNAs: exploring a new dimension in the pathogenesis of kidney cancer.BMC Med. 2010; 8: 65Google Scholar, 15White N.M. Yousef G.M. Translating molecular signatures of renal cell carcinoma into clinical practice.J Urol. 2011; 186: 9-11Google Scholar and a few miRNAs were experimentally shown to have an effect on tumor growth and progression in RCC.16Chow T.F. Mankaruos M. Scorilas A. Youssef Y. Girgis A. Mossad S. Metias S. Rofael Y. Honey R.J. Stewart R. Pace K.T. Yousef G.M. The miR-17-92 cluster is over expressed in and has an oncogenic effect on renal cell carcinoma.J Urol. 2010; 183: 743-751Google Scholar, 17Neal C.S. Michael M.Z. Rawlings L.H. Van der Hoek M.B. Gleadle J.M. The VHL-dependent regulation of microRNAs in renal cancer.BMC Med. 2010; 8: 64Google Scholar, 18White N.M. Bui A. Mejia-Guerrero S. Chao J. Soosaipillai A. Youssef Y. Mankaruos M. Honey R.J. Stewart R. Pace K.T. Sugar L. Diamandis E.P. Dore J. Yousef G.M. Dysregulation of kallikrein-related peptidases in renal cell carcinoma: potential targets of miRNAs.Biol Chem. 2010; 391: 411-423Google Scholar Recent reports19Metias S.M. Lianidou E. Yousef G.M. MicroRNAs in clinical oncology: at the crossroads between promises and problems.J Clin Pathol. 2009; 62: 771-776Google Scholar demonstrated the potential for miRNAs to be useful diagnostic, prognostic, and predictive markers. miR-21 is an oncogenic miRNA that is up-regulated in many tumors, including breast, lung, colon, pancreas, prostate, and stomach cancers.20Markou A. Tsaroucha E.G. Kaklamanis L. Fotinou M. Georgoulias V. Lianidou E.S. Prognostic value of mature microRNA-21 and microRNA-205 overexpression in non-small cell lung cancer by quantitative real-time RT-PCR.Clin Chem. 2008; 54: 1696-1704Google Scholar, 21Slaby O. Svoboda M. Fabian P. Smerdova T. Knoflickova D. Bednarikova M. Nenutil R. Vyzula R. Altered expression of miR-21, miR-31, miR-143 and miR-145 is related to clinicopathologic features of colorectal cancer.Oncology. 2007; 72: 397-402Google Scholar, 22Yan L.X. Huang X.F. Shao Q. Huang M.Y. Deng L. Wu Q.L. Zeng Y.X. Shao J.Y. MicroRNA miR-21 overexpression in human breast cancer is associated with advanced clinical stage, lymph node metastasis and patient poor prognosis.RNA. 2008; 14: 2348-2360Google Scholar The clinical utility of miR-21 as a tumor marker is evident for many cancers. Of particular promise is the potential use of miR-21 as a prognostic marker. In breast cancer, high levels of miR-21 significantly correlated with advanced clinical stage, lymph node metastases, and shortened survival.22Yan L.X. Huang X.F. Shao Q. Huang M.Y. Deng L. Wu Q.L. Zeng Y.X. Shao J.Y. MicroRNA miR-21 overexpression in human breast cancer is associated with advanced clinical stage, lymph node metastasis and patient poor prognosis.RNA. 2008; 14: 2348-2360Google Scholar In colon cancer, high expression was associated with the development of metastasis and correlated with clinical stage.21Slaby O. Svoboda M. Fabian P. Smerdova T. Knoflickova D. Bednarikova M. Nenutil R. Vyzula R. Altered expression of miR-21, miR-31, miR-143 and miR-145 is related to clinicopathologic features of colorectal cancer.Oncology. 2007; 72: 397-402Google Scholar Similar results were reported for squamous cell carcinoma of the tongue: miR-21 was overexpressed in advanced tumor stages compared with early stages.23Li J. Huang H. Sun L. Yang M. Pan C. Chen W. Wu D. Lin Z. Zeng C. Yao Y. Zhang P. Song E. MiR-21 indicates poor prognosis in tongue squamous cell carcinomas as an apoptosis inhibitor.Clin Cancer Res. 2009; 15: 3998-4008Google Scholar Markou et al20Markou A. Tsaroucha E.G. Kaklamanis L. Fotinou M. Georgoulias V. Lianidou E.S. Prognostic value of mature microRNA-21 and microRNA-205 overexpression in non-small cell lung cancer by quantitative real-time RT-PCR.Clin Chem. 2008; 54: 1696-1704Google Scholar showed that miR-21 overexpression is an independent negative prognostic factor for overall survival in non–small-cell lung cancer. The effect of miR-21 on carcinogenesis and tumor progression has been experimentally analyzed.24White N.M. Fatoohi E. Metias M. Jung K. Stephan C. Yousef G.M. Metastamirs: a stepping stone towards improved cancer management.Nat Rev Clin Oncol. 2011; 8: 75-84Google Scholar miR-21 is involved in cell growth and inhibition of apoptosis. Inhibition of miR-21 expression led to caspase activation and associated apoptotic cell death in glioblastoma.25Chen Y. Liu W. Chao T. Zhang Y. Yan X. Gong Y. Qiang B. Yuan J. Sun M. Peng X. MicroRNA-21 down-regulates the expression of tumor suppressor PDCD4 in human glioblastoma cell T98G.Cancer Lett. 2008; 272: 197-205Google Scholar Also, suppression of miR-21 in MCF-7 cells was associated with increased apoptosis and decreased cell proliferation.26Lu Z. Liu M. Stribinskis V. Klinge C.M. Ramos K.S. Colburn N.H. Li Y. MicroRNA-21 promotes cell transformation by targeting the programmed cell death 4 gene.Oncogene. 2008; 27: 4373-4379Google Scholar In this study, we compare miR-21 expression between healthy kidney and the different subtypes of RCC and oncocytoma. We show that miR-21 can distinguish ccRCC and pRCC from chRCC and oncocytoma with 90% specificity. We also show that miR-21 has potential prognostic utility in RCC. We analyzed a total of 121 cases of kidney cancer and healthy kidney tissues, including 71 cases with ccRCC, 18 with pRCC, 10 with chRCC, and 8 with oncocytoma and 14 healthy kidney tissues. Malignant, benign, and healthy tissues were collected from St Michael's Hospital, University Health Network, and Sunnybrook Health Sciences Center, Toronto, ON, Canada. Healthy and malignant tissues were identified, and diagnoses were confirmed by two independent genitourinary pathologists (H.F. and G.M.Y.). Samples from cancerous areas were taken from areas with no hemorrhage or necrosis, and multiple sections were submitted from the same tumor to compensate for tumor heterogeneity. Samples of healthy kidney were taken from the kidney cortex, away from the tumor, to avoid cancer field effect. Six samples were obtained from each case through needle core biopsy specimens (1-mm thick). Tumor classification and staging were established according to the 2002 TNM System and the 2004 World Health Organization classification. All procedures were performed according to Research Ethics Board approval from St Michael's Hospital, Sunnybrook Health Sciences Center, and University Health Network. Six cores of pure tumor (or healthy) tissue were obtained from formalin-fixed, paraffin-embedded tissues and pooled for each specimen. Total RNA was extracted using miRNeasy (Qiagen, Mississauga, ON, Canada), according to the manufacturer's protocol, as described in our recent publication.27Khella H.W. White N.M. Faragalla H. Gabril M. Boazak M. Dorian D. Khalil B. Antonios H. Bao T.T. Pasic M.D. Honey R.J. Stewart R. Pace K.T. Bjarnason G.A. Jewett M.A. Yousef G.M. Exploring the role of miRNAs in renal cell carcinoma progression and metastasis through bioinformatic and experimental analyses.Tumour Biol. 2012; 33: 131-140Google Scholar Total RNA concentrations were determined spectrophotometrically (NanoDrop 1000 Spectrophotometer; NanoDrop Technologies Inc., Wilmington, DE). Samples optimal for analysis were stored at −80°C. Quantitative real-time RT-PCR was used to measure miRNA expression with TaqMan MicroRNA Assays (Applied Biosystems, Foster City, CA), as recently described.28White N.M. Chow T.F. Mejia-Guerrero S. Diamandis M. Rofael Y. Faragalla H. Mankaruous M. Gabril M. Girgis A. Yousef G.M. Three dysregulated miRNAs control kallikrein 10 expression and cell proliferation in ovarian cancer.Br J Cancer. 2010; 102: 1244-1253Google Scholar miR-21–specific reverse transcription was performed with 5 ng of total RNA using the TaqMan MicroRNA Reverse Transcription Kit (Applied Biosystems), as recommended by the manufacturer. Quantitative PCR was performed using the TaqMan microRNA Assay Kit on the Step One Plus Real-Time PCR System (Applied Biosystems). Thermal cycling conditions were according to the manufacturer's fast protocol, and all reactions were performed in triplicate. Gene expression analysis was performed using the comparative CT (2−ΔΔCT) method to calculate the relative quantification units of miR-21 in kidney tumors. The comparative CT method (2−ΔΔCΤ) was used for performing relative quantification analysis.29Livak K.J. Schmittgen T.D. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method.Methods. 2001; 25: 402-408Google Scholar, 30Wotschofsky Z. Meyer H.A. Jung M. Fendler A. Wagner I. Stephan C. Busch J. Erbersdobler A. Disch A.C. Mollenkopf H.J. Jung K. Reference genes for the relative quantification of microRNAs in renal cell carcinomas and their metastases.Anal Biochem. 2011; 417: 233-241Google Scholar The normalization of miR-21 expression between different specimens was implemented through RNU44 amplification using one positive sample as a calibrator. By using the formula ΔCT = CT miR-21 − CT RNU44, we normalized the miR-21 expression of each tested sample to the RNU44 endogenous reference expression of the same sample. Consequently, by the formula ΔΔCT = ΔCT, sample − ΔCT,calibrator, the normalized miR-21 expression of each tested sample was determined relative to the normalized miR-21 expression of the calibrator sample. Therefore, the amount of the miR-21 expression levels normalized to the expression of the RNU44 endogenous reference gene and relative to a calibrator is given by the 2−ΔΔCT formula. Given that the distributions of miR-21 expression levels in the patients with kidney cancer were not gaussian, the analysis of the differences in the two or three groups of patients was performed with the nonparametric χ2 test and the Fisher's exact test when it was appropriate. Relationships between different continuous variables were assessed by Spearman correlation coefficient. The X-tile algorithm was used to generate an optimal cutoff point for miR-21, because they are molecules with no established cutoff points regarding their expression in kidney cancer. After correcting for the use of minimum P value statistics, the X-tile software (Yale University School of Medicine; http://medicine.yale.edu/labs/rimm/www/xtilesoftware.html, last accessed May 14, 2012) yielded an optimal cutoff, equal to the 40th percentile for miR-21, with a calculated Monte Carlo P < 0.05. For survival analysis, Cox regression analysis was conducted at both univariate and multivariate levels. Only patients for whom the status of all variables was known were included in the multivariate regression models, which incorporated miR-21 expression and all other variables for which the patients were characterized. Because the tumor subtype can be a significant factor in determining survival (eg, oncocytoma has a much more benign behavior), in our survival analysis, we only included the subgroups of clear cell and papillary subtypes (n = 89) that have comparable behavior and a comparable overall miR-21 dysregulation pattern. Information on sex and laterality was missing for two patients, and information on grade was missing for one patient. The multivariate models were adjusted for tumor size, patient age, tumor grade, and patient stage. Survival analyses were also performed by constructing Kaplan-Meier disease-free survival (DFS) and overall survival curves. Any differences between the curves were evaluated by the log-rank test. DFS was defined as the time between the initial resection of the kidney tumor and the occurrence of recurrence or metastasis. Overall survival was defined as the time between the first surgery for tumor resection and death from any cause. Receiver operating characteristic (ROC) curves were constructed for miR-21 expression levels, by plotting sensitivity versus (1-specificity). The areas under the ROC curves (AUCs) were analyzed by the Hanley and McNeil method. The CIs around sensitivity and specificity were calculated using a web calculator (VassarStats; Vassar College, Poughkeepsie, NY; http://faculty.vassar.edu/lowry/clin1.html, last accessed November 1, 2011).30Wotschofsky Z. Meyer H.A. Jung M. Fendler A. Wagner I. Stephan C. Busch J. Erbersdobler A. Disch A.C. Mollenkopf H.J. Jung K. Reference genes for the relative quantification of microRNAs in renal cell carcinomas and their metastases.Anal Biochem. 2011; 417: 233-241Google Scholar, 31Wilson E.B. Probable inference, the law of succession, and statistical inference.J Am Stat Assoc. 1927; 22: 209-212Google Scholar We used a combination of algorithms, including PicTar, TargetScan 5.1, and miRecords (which compile data from 11 different prediction programs),32Xiao F. Zuo Z. Cai G. Kang S. Gao X. Li T. miRecords: an integrated resource for microRNA-target interactions.Nucleic Acids Res. 2009; 37: D105-D110Google Scholar to identify potential targets for miR-21. These were cross matched with key molecules involved in ccRCC pathogenesis. To ensure accurate prediction, we included only those targets predicted by at least three different algorithms. The descriptive statistics of the numerical variables in the study are shown in Table 1. miR-21 expression was significantly higher in the clear cell and papillary subtypes of kidney cancer compared with chromophobe RCC, oncocytoma, and healthy kidney tissues (Figure 1). Expression levels were comparable in both chromophobe RCC and oncocytoma.Table 1Descriptive Statistics of the Numerical Variables in the StudyVariableMean ± SEMedianRangemiR-21 ccRCC (n = 71)14.74 ± 1.5112.560.26–56.85 pRCC (n = 18)10.16 ± 1.617.262.54–24.71 chRCC (n = 10)3.41 ± 0.722.640.76–7.51 Oncocytoma (n = 8)1.38 ± 0.321.610.08–2.48 Healthy (n = 14)4.11 ± 1.491.180.08–19.18Age (years)62.4 ± 1.364.026.0–87.0Tumor size (cm)5.43 ± 0.324.501.40–20.00Follow-up (months)52.8 ± 2.550.01.0–192.0Data are given as relative quantification units. Open table in a new tab Data are given as relative quantification units. To evaluate the discrimination value of miR-21 expression profiles in RCC, we performed ROC analysis. The ROC curve (Figure 2) illustrated the significant value of miR-21 expression levels in distinguishing ccRCC or pRCC from noncancerous kidney tissues (AUC, 0.687; 95% CI, 0.578 to 0.795; P = 0.002). Moreover, the ROC curve (Figure 3) illustrated the significant value of miR-21 expression levels in distinguishing ccRCC or pRCC from chRCC or oncocytoma (AUC, 0.886; 95% CI, 0.818 to 0.955; P < 0.001). By using a cutoff equal to 3.7 relative quantification units, the specificity was 90% (95% CI, 63.9% to 98.1%), with a sensitivity of 83% (95% CI, 53.5% to 97.6%).Figure 3miR-21 can distinguish clear and papillary subtypes of RCC from chromophobe RCC and oncocytoma with high accuracy. By using a cutoff equal to 3.7 relative quantification units, the specificity is 90%, with a sensitivity of 83%. For ccRCC and pRCC versus chRCC and oncocytoma: AUC, 0.886; 95% CI, 0.818 to 0.955; P = 0.001.View Large Image Figure ViewerDownload Hi-res image Download (PPT) Statistically significant discriminative values between ccRCC and pRCC, chRCC, and oncocytoma or noncancerous tissues were not observed. These results were in agreement with those obtained by an independent technique, microarray analysis of fresh-frozen tissue specimens of primary RCC of different subtypes in an independent cohort of patients who were described in our recent publications.8Youssef Y.M. White N.M. Grigull J. Krizova A. Samy C. Mejia-Guerrero S. Evans A. Yousef G.M. Accurate molecular classification of kidney cancer subtypes using microRNA signature.Eur Urol. 2011; 59: 721-730Google Scholar, 13White N.M. Bao T.T. Grigull J. Youssef Y.M. Girgis A. Diamandis M. Fatoohi E. Metias M. Honey R.J. Stewart R. Pace K.T. Bjarnason G.A. Yousef G.M. miRNA profiling for clear cell renal cell carcinoma: biomarker discovery and identification of potential controls and consequences of miRNA dysregulation.J Urol. 2011; 186: 1077-1083Google Scholar There was a statistically significant increase in expression of miR-21 in late stages (stages II and III) compared with early stage (stage I) (P = 0.018) (Table 2). A higher miR-21 expression level was also associated with higher tumor grades (P = 0.015) (Table 2). No significant difference in miR-21 expression levels was found between males and females. Also, there was a statistically significant positive correlation between miR-21 expression and tumor size (rs = 0.263, P = 0.013) (data not shown). This correlation might be, however, clinically insignificant.Table 2Associations between miR-21 Status and Other VariablesVariableTotalNo. (%) of patientsP valuemiR-21 negativemiR-21 positiveSex Male5423 (42.6)31 (57.4)0.51⁎Fisher's exact test. Female3317 (51.5)16 (48.5)Laterality Left3412 (35.3)22 (64.7)0.13⁎Fisher's exact test. Right5328 (52.8)25 (47.2)Grade 15732 (56.1)25 (43.9)0.018†χ2 Test. 2123 (25.0)9 (75.0) 3/4205 (25.0)15 (75.0)Stage I5732 (56.1)25 (43.9)0.018†χ2 Test. II123 (25.0)9 (75.0) III205 (25.0)15 (75.0)Cutoff point: 5.0 relative quantification units equals the 40th percentile. Fisher's exact test.† χ2 Test. Open table in a new tab Cutoff point: 5.0 relative quantification units equals the 40th percentile. In the univariate analysis, patients who were miR-21 positive showed a significantly shorter DFS [hazard ratio (HR), 2.15; 95% CI, 1.16 to 3.98; P = 0.014] and overall survival (HR, 1.97; 95% CI, 1.04 to 3.73; P = 0.036) (Table 3). The same trend of worse prognosis was seen when using miR-21 as a continuous variable (HR, 1.41), although this did not reach statistical significance (P = 0.15). In the multivariate analysis, miR-21–positive patients had shorter disease-free and overall survival (HRs, 1.64 and 1.75, respectively) compared with those who were miR-21 negative, although this was not statistically significant (P = 0.15 and P = 0.11, respectively).Table 3Associations between miR-21 and Patient SurvivalmiR-21 statusDisease-free survivalOverall survivalHR (95% CI)P valueHR (95% CI)P valueUnivariate Analysis Categorical variable⁎miR-21 status was based on a cutoff point equal to the 40th percentile of the distribution of miR-21 values. Negative1.000.0141.000.036 Positive2.15 (1.16–3.98)1.97 (1.04–3.73) Continuous variable (log)1.41 (0.84–2.37)0.191.36 (0.79–2.37)0.26Multivariate Analysis†Multivariate models were adjusted for tumor size, patient age, tumor grade, and patient stage. Categorical variable⁎miR-21 status was based on a cutoff point equal to the 40th percentile of the distribution of miR-21 values. Negative1.00 (0.83–3.27)0.151.00 (0.87–3.51)0.11 Positive1.64 (0.83–3.27)0.151.75 (0.87–3.51)0.11 Continuous variable (log)1.02 (0.57–1.81)0.951.14 (0.63–2.05)0.65The HR was estimated from a Cox proportional hazard regression model. miR-21 status was based on a cutoff point equal to the 40th percentile of the distribution of miR-21 values.† Multivariate models were adjusted for tumor size, patient age, tumor grade, and patient stage. Open table in a new tab The HR was estimated from a Cox proportional hazard regression model. Kaplan-Meier survival curves showed a clear demarcation between miR-21–positive and miR-21–negative patients; the latter demonstrated a significantly higher DFS (P = 0.008) (Figure 4) and overall survival (P = 0.031) (Figure 5) compared with miR-21–positive patients.Figure 5Kaplan-Meier overall survival plot comparing miR-21 expression (positive or negative) in patients with RCC. miR-21–positive patients show a statistically significant shorter overall survival compared with those who are miR-21 negative (P = 0.031).View Large Image Figure ViewerDownload Hi-res image Download (PPT) A list of genes predicted to be targeted by miR-21 and their biological processes and relevance to cancer pathways, as identified by Gene Ontology analysis, are shown in Supplemental Table S1 (available at http://jmd.amjpathol.org). As shown in Table 4, there were many carcinogenesis-related pathways that included potential miR-21 targets. Furthermore, the tar
Referência(s)