The Chromatin Remodeling Gene ARID1A Is a New Prognostic Marker in Clear Cell Renal Cell Carcinoma
2013; Elsevier BV; Volume: 182; Issue: 4 Linguagem: Inglês
10.1016/j.ajpath.2013.01.007
ISSN1525-2191
AutoresZsuzsanna Lichner, Andreas Scorilas, Nicole M. White, Andrew H. Girgis, Lora Rotstein, Kimberly C. Wiegand, Ashraf Latif, Christina R. Chow, David G. Huntsman, George M. Yousef,
Tópico(s)Renal cell carcinoma treatment
ResumoClear cell renal cell carcinoma (ccRCC) is the most common tumor of the adult kidney, with an increasing rate of incidence. Recently, exome sequencing studies have revealed that the SWI/SNF (switch/sucrose nonfermentable) members PBRM1 and ARID1A are mutated in ccRCC, and it has also been suggested that aberrant chromatin regulation is a key step in kidney cancer pathogenesis. Herein, we show that down-regulation of another SW/SNF component, ARID1A, occurs frequently in ccRCC. We detected copy number loss of ARID1A in 16% of patients with ccRCC. Immunohistochemistry indicated that 67% of ccRCC (53 of 79) had significantly lower expression of BAF250a, the protein product of ARID1A, than did the matched normal kidney cortex. In parallel, we conducted in silico mRNA expression analysis on 404 ccRCC tumors and 167 normal kidney cortex samples using publicly available databases and confirmed significant down-regulation of ARID1A in 68.8% of patients. We also show that decreased BAF250a protein and ARID1A mRNA expression correlate with tumor stage and grade. Our results indicate that both the protein and mRNA levels of ARID1A are statistically significant prognostic markers for ccRCC. Even after controlling for other confounders in the multivariate analysis, BAF250 retained its prognostic significance. BAF250a IHC is easy to perform and represents a potential biomarker that could be incorporated in laboratory practice to enhance the accuracy of the existing prognostic models. Clear cell renal cell carcinoma (ccRCC) is the most common tumor of the adult kidney, with an increasing rate of incidence. Recently, exome sequencing studies have revealed that the SWI/SNF (switch/sucrose nonfermentable) members PBRM1 and ARID1A are mutated in ccRCC, and it has also been suggested that aberrant chromatin regulation is a key step in kidney cancer pathogenesis. Herein, we show that down-regulation of another SW/SNF component, ARID1A, occurs frequently in ccRCC. We detected copy number loss of ARID1A in 16% of patients with ccRCC. Immunohistochemistry indicated that 67% of ccRCC (53 of 79) had significantly lower expression of BAF250a, the protein product of ARID1A, than did the matched normal kidney cortex. In parallel, we conducted in silico mRNA expression analysis on 404 ccRCC tumors and 167 normal kidney cortex samples using publicly available databases and confirmed significant down-regulation of ARID1A in 68.8% of patients. We also show that decreased BAF250a protein and ARID1A mRNA expression correlate with tumor stage and grade. Our results indicate that both the protein and mRNA levels of ARID1A are statistically significant prognostic markers for ccRCC. Even after controlling for other confounders in the multivariate analysis, BAF250 retained its prognostic significance. BAF250a IHC is easy to perform and represents a potential biomarker that could be incorporated in laboratory practice to enhance the accuracy of the existing prognostic models. Kidney cancer is one of the most prevalent cancers in North America,1Siegel R. Naishadham D. Jemal A. Cancer statistics, 2012.CA Cancer J Clin. 2012; 62: 10-29Crossref PubMed Scopus (10290) Google Scholar, 2Simard E.P. Ward E.M. Siegel R. Jemal A. Cancers with increasing incidence trends in the United States: 1999 through 2008 [published correction appears in CA Cancer J Clin 2012, 62:277].CA Cancer J Clin. 2012; 62: 118-128Crossref Scopus (574) Google Scholar with 85% of renal cell carcinomas (RCCs) falling into the clear cell subtype (ccRCC). Inactivation of the tumor suppressor VHL gene is observed in all inherited ccRCCs and in most sporadic nonfamilial tumors.3Gnarra J.R. Tory K. Weng Y. Schmidt L. Wei M.H. Li H. Latif F. Liu S. Chen F. Duh F.-M. Lubensky I. Duan D.R. Florence C. Pozzatti R. Walther M.M. Bander N.H. Grossman H.B. Branch H. Pomer S. Brooks J.D. Isaacs W.B. Lerman M.I. Zbar B. Linehan W.M. Mutations of the VHL tumour suppressor gene in renal carcinoma.Nat Genet. 1994; 7: 85-90Crossref PubMed Scopus (1469) Google Scholar Despite the central role of the VHL-hypoxia axis in ccRCC pathogenesis, it is presumed that additional genetic alterations are required for tumor initiation.4Kaelin Jr., W.G. The von Hippel-Lindau tumor suppressor gene and kidney cancer.Clin Cancer Res. 2004; 10: 6290S-6295SCrossref PubMed Scopus (252) Google Scholar The mammalian SWI and SNF (switch and sucrose nonfermentable) loci code for proteins that assemble into large multisubunit complexes. These SWI/SNF complexes are part of the ATP-dependent nucleosome remodeling system and can be divided into two major complexes, BAF and PBAF. Both BAF and PBAF aggregates have “core” subunits that are compulsory in all SWI/SNF complexes (SNF5, BAF170, and BAF155) and mutually exclusive subunits such as BAF250a, the protein product of ARID1A; BAF250b, the protein encoded by ARID1B in the BAF complex; and PBRM1 (BAF180) in the PBAF complex.5Hargreaves D.C. Crabtree G.R. ATP-dependent chromatin remodeling: genetics, genomics and mechanisms.Cell Res. 2011; 21: 396-420Crossref PubMed Scopus (573) Google Scholar, 6Wilson B.G. Roberts C.W. SWI/SNF nucleosome remodellers and cancer.Nat Rev Cancer. 2011; 11: 481-492Crossref PubMed Scopus (850) Google Scholar Recurrent mutations of SWI/SNF members have been identified in many cancer types, and their importance has been proved in transgenic mouse models such as the SNF5 transgenic mouse model for lymphomas and rhabdoid tumors.7Roberts C.W. Leroux M.M. Fleming M.D. Orkin S.H. Highly penetrant, rapid tumorigenesis through conditional inversion of the tumor suppressor gene Snf5.Cancer Cell. 2002; 2: 415-425Abstract Full Text Full Text PDF PubMed Scopus (247) Google Scholar ARID1A, coding for the BAF250a subunit of BAF, is mutated in about 50% of ovarian clear cell carcinomas.8Jones S. Wang T.L. Shih IeM. Mao T.L. Nakayama K. Roden R. Glas R. Slamon D. Diaz Jr., L.A. Vogelstein B. Kinzler K.W. Velculescu V.E. Papadopoulos N. Frequent mutations of chromatin remodeling gene ARID1A in ovarian clear cell carcinoma.Science. 2010; 330: 228-231Crossref PubMed Scopus (915) Google Scholar, 9Katagiri A. Nakayama K. Rahman M.T. Rahman M. Katagiri H. Nakayama N. Ishikawa M. Ishibashi T. Iida K. Kobayashi H. Otsuki Y. Nakayama S. Miyazaki K. Loss of ARID1A expression is related to shorter progression-free survival and chemoresistance in ovarian clear cell carcinoma.Mod Pathol. 2012; 25: 282-288PubMed Google Scholar, 10Wiegand K.C. Shah S.P. Al-Agha O.M. Zhao Y. Tse K. Zeng T. Senz J. McConechy M.K. Anglesio M.S. Kalloger S.E. Yang W. Heravi-Moussavi A. Giuliany R. Chow C. Fee J. Zayed A. Prentice L. Melnyk N. Turashvili G. Delaney A.D. Madore J. Yip S. McPherson A.W. Ha G. Bell L. Fereday S. Tam A. Galletta L. Tonin P.N. Provencher D. Miller D. Jones S.J. Moore R.A. Morin G.B. Oloumi A. Boyd N. Aparicio S.A. Shih IeM. Mes-Masson A.M. Bowtell D.D. Hirst M. Gilks B. Marra M.A. Huntsman D.G. ARID1A mutations in endometriosis-associated ovarian carcinomas.N Engl J Med. 2010; 363: 1532-1543Crossref PubMed Scopus (1188) Google Scholar BAF250a is also frequently lost in approximately 30% of endometrial cancers,11Wiegand K.C. Lee A.F. Al-Agha O.M. Chow C. Kalloger S.E. Scott D.W. Steidl C. Wiseman S.M. Gascoyne R.D. Gilks B. Huntsman D.G. Loss of BAF250a (ARID1A) is frequent in high-grade endometrial carcinomas.J Pathol. 2011; 224: 328-333Crossref PubMed Scopus (174) Google Scholar and ARID1A is mutated in some medulloblastomas12Parsons D.W. Li M. Zhang X. Jones S. Leary R.J. Lin J.C. Boca S.M. Carter H. Samayoa J. Bettegowda C. Gallia G.L. Jallo G.I. Binder Z.A. Nikolsky Y. Hartigan J. Smith D.R. Gerhard D.S. Fults D.W. VandenBerg S. Berger M.S. Marie S.K. Shinjo S.M. Clara C. Phillips P.C. Minturn J.E. Biegel J.A. Judkins A.R. Resnick A.C. Storm P.B. Curran T. He Y. Rasheed B.A. Friedman H.S. Keir S.T. McLendon R. Northcott P.A. Taylor M.D. Burger P.C. Riggins G.J. Karchin R. Parmigiani G. Bigner D.D. Yan H. Papadopoulos N. Vogelstein B. Kinzler K.W. Velculescu V.E. The genetic landscape of the childhood cancer medulloblastoma.Science. 2011; 331: 435-439Crossref PubMed Scopus (567) Google Scholar and breast and lung cancers.13Huang J. Zhao Y.L. Li Y. Fletcher J.A. Xiao S. Genomic and functional evidence for an ARID1A tumor suppressor role.Genes Chromosomes Cancer. 2007; 46: 745-750Crossref PubMed Scopus (63) Google Scholar Furthermore, an earlier report observed frequent deficiency of BAF250a (alias p270) in carcinomas of the breast and kidney, and homozygous deletion of ARID1A was detected in the LB1047-RCC ccRCC cell line.14Varela I. Tarpey P. Raine K. Huang D. Ong C.K. Stephens P. Davies H. Jones D. Lin M.L. Teague J. Bignell G. Butler A. Cho J. Dalgliesh G.L. Galappaththige D. Greenman C. Hardy C. Jia M. Latimer C. Lau K.W. Marshall J. McLaren S. Menzies A. Mudie L. Stebbings L. Largaespada D.A. Wessels L.F. Richard S. Kahnoski R.J. Anema J. Tuveson D.A. Perez-Mancera P.A. Mustonen V. Fischer A. Adams D.J. Rust A. Chan-on W. Subimerb C. Dykema K. Furge K. Campbell P.J. Teh B.T. Stratton M.R. Futreal P.A. Exome sequencing identifies frequent mutation of the SWI/SNF complex gene PBRM1 in renal carcinoma [published correction appears in Nature 2012, 484:130].Nature. 2011; 469: 539-542Crossref PubMed Scopus (936) Google Scholar, 15Wang X. Nagl Jr., N.G. Flowers S. Zweitzig D. Dallas P.B. Moran E. Expression of p270 (ARID1A), a component of human SWI/SNF complexes, in human tumors.Int J Cancer. 2004; 112: 636Crossref PubMed Scopus (82) Google Scholar Recent exome sequencing of ccRCC revealed mutations in the SWI/SNF members PBRM1 (41%) and ARID1A (3%). ARID1A mutations usually cause a premature stop codon and are randomly distributed along the gene.8Jones S. Wang T.L. Shih IeM. Mao T.L. Nakayama K. Roden R. Glas R. Slamon D. Diaz Jr., L.A. Vogelstein B. Kinzler K.W. Velculescu V.E. Papadopoulos N. Frequent mutations of chromatin remodeling gene ARID1A in ovarian clear cell carcinoma.Science. 2010; 330: 228-231Crossref PubMed Scopus (915) Google Scholar, 16Wang K. Kan J. Yuen S.T. Shi S.T. Chu K.M. Law S. Chan T.L. Kan Z. Chan A.S. Tsui W.Y. Lee S.P. Ho S.L. Chan A.K. Cheng G.H. Roberts P.C. Rejto P.A. Gibson N.W. Pocalyko D.J. Mao M. Xu J. Leung S.Y. Exome sequencing identifies frequent mutation of ARID1A in molecular subtypes of gastric cancer.Nat Genet. 2011; 43: 1219-1223Crossref PubMed Scopus (575) Google Scholar These findings point to ARID1A as a cancer-associated gene and a possible contributor to ccRCC. Among the large number of potential targets of the SWI/SNF complex, some essential downstream molecules have been identified and validated. SWI/SNF complex activates the p16 INK4A cell cycle checkpoint regulator that blocks the transition to S phase by inhibiting retinoblastoma protein phosphorylation.17Bétous R. Mason A.C. Rambo R.P. Bansbach C.E. Badu-Nkansah A. Sirbu B.M. Eichman B.F. Cortez D. SMARCAL1 catalyzes fork regression and Holliday junction migration to maintain genome stability during DNA replication.Genes Dev. 2012; 26: 151-162Crossref PubMed Scopus (180) Google Scholar, 18Isakoff M.S. Sansam C.G. Tamayo P. Subramanian A. Evans J.A. Fillmore C.M. Wang X. Biegel J.A. Pomeroy S.L. Mesirov J.P. Roberts C.W. Inactivation of the Snf5 tumor suppressor stimulates cell cycle progression and cooperates with p53 loss in oncogenic transformation.Proc Natl Acad Sci U S A. 2005; 102: 17745-17750Crossref PubMed Scopus (160) Google Scholar, 19Oruetxebarria I. Venturini F. Kekarainen T. Houweling A. Zuijderduijn L.M. Mohd-Sarip A. Vries R.G. Hoeben R.C. Verrijzer C.P. P16INK4a is required for hSNF5 chromatin remodeler-induced cellular senescence in malignant rhabdoid tumor cells.J Biol Chem. 2004; 279: 3807-3816Crossref PubMed Scopus (124) Google Scholar, 20Trouche D. Le Chalony C. Muchardt C. Yaniv M. Kouzarides T. RB and hbrm cooperate to repress the activation functions of E2F1.Proc Natl Acad Sci U S A. 1997; 94: 11268-11273Crossref PubMed Scopus (258) Google Scholar It also delays entry to the S phase by directly binding retinoblastoma protein and inhibiting the expression of EF2s and CCND1.20Trouche D. Le Chalony C. Muchardt C. Yaniv M. Kouzarides T. RB and hbrm cooperate to repress the activation functions of E2F1.Proc Natl Acad Sci U S A. 1997; 94: 11268-11273Crossref PubMed Scopus (258) Google Scholar This repression of retinoblastoma protein during differentiation is probably dependent on ARID1A but not ARID1B. Another important difference between the BAF complexes with ARID1A or ARID1B subunit is that the SWI/SNF complex composed with ARID1A represses the oncogenic Myc protein, whereas the ARID1B-containing complexes promote Myc transcription. Knockdown of ARID1A but not ARID1B also interferes with the differentiation-associated cell cycle arrest in osteoblasts.21Nagl Jr., N.G. Patsialou A. Haines D.S. Dallas P.B. Beck Jr., G.R. Moran E. The p270 (ARID1A/SMARCF1) subunit of mammalian SWI/SNF-related complexes is essential for normal cell cycle arrest.Cancer Res. 2005; 65: 9236-9244Crossref PubMed Scopus (103) Google Scholar The SWI/SNF complexes regulate the actin cytoskeleton through inhibition of RhoA, a protein that is responsible for stress fiber formation and cell migration. RhoA has been closely linked to metastasis, and its overexpression is correlated with poor prognosis.22Caramel J. Quignon F. Delattre O. RhoA-dependent regulation of cell migration by the tumor suppressor hSNF5/INI1.Cancer Res. 2008; 68: 6154-6161Crossref PubMed Scopus (43) Google Scholar In the present study, we found that decreased ARID1A expression at both the protein and mRNA levels is common in ccRCC. We also compared ARID1A protein expression between the various kidney cancer subtypes. Moreover, we examined the correlation between ARID1A expression and various clinical parameters, as well as survival in patients with ccRCC. We provide evidence that expression of BAF250a protein and ARID1A mRNA are potential prognostic markers of ccRCC. The present study included 145 kidney cancers (82 ccRCCs, 18 chromophobe RCCs, 25 papillary RCCs, 11 oncocytomas, and 9 transitional cell carcinomas). Of the 79 ccRCCs, matched normal tissues from the same patient were available. Specimens were collected from St. Michael’s Hospital (Toronto, ON, Canada) after obtaining Research Ethics Board approval. Diagnosis and selection of pure tumor areas were performed by a pathologist (G.M.Y.). For tissue microarray (TMA) construction, two 1.0-mm cores were obtained from each tumor using the microarrayer and verified by a pathologist (G.M.Y.). To obtain pure areas of normal and cancerous tissues and to prevent field effect, normal tissue was obtained from the other pole of the kidney, at least 2 cm from the tumor. All normal kidney cortex specimens showed moderate or intense nuclear immunoreactivity for BAF250a. Tumors were scored positive for BAF250a if tumor cells showed definite nuclear staining, and negative if tumor nuclei had no immunoreactivity but endothelial and other nontumor cells from the same sample showed immunoreactivity. Tumors in which neither normal cells in the stroma nor tumor cells were immunoreactive were considered the result of technical failure and were excluded. BAF250 staining was evaluated on the basis of staining frequency (% positive area) (51% to 100%, 3′; 25% to 50%, 2′; and 10% to 24%, 1′) and staining intensity. Nuclear staining intensity was scored as follows: 1 = low, 2 = moderate, and 3 = high. Relative quantification units were calculated as the product of the average staining intensity and frequency. Each tumor was represented in duplicate on the TMA. An average of the two cores was calculated for each. Immunohistochemical analysis was performed as described previously10Wiegand K.C. Shah S.P. Al-Agha O.M. Zhao Y. Tse K. Zeng T. Senz J. McConechy M.K. Anglesio M.S. Kalloger S.E. Yang W. Heravi-Moussavi A. Giuliany R. Chow C. Fee J. Zayed A. Prentice L. Melnyk N. Turashvili G. Delaney A.D. Madore J. Yip S. McPherson A.W. Ha G. Bell L. Fereday S. Tam A. Galletta L. Tonin P.N. Provencher D. Miller D. Jones S.J. Moore R.A. Morin G.B. Oloumi A. Boyd N. Aparicio S.A. Shih IeM. Mes-Masson A.M. Bowtell D.D. Hirst M. Gilks B. Marra M.A. Huntsman D.G. ARID1A mutations in endometriosis-associated ovarian carcinomas.N Engl J Med. 2010; 363: 1532-1543Crossref PubMed Scopus (1188) Google Scholar on 4 μm thick paraffin sections using the semiautomated DiscoveryXT instrument (Ventana Medical Systems, Inc., Tucson, AZ). BAF250a was stained using the ChromoMap DAB kit (Ventana). Antigen retrieval was standard CC1 with a 2-hour primary incubation. BAF250a mouse clone 3H2 (Abgent, Inc., Dan Diego, CA) was applied at a ratio of 1:25, followed by a 16-minute secondary incubation of prediluted UltraMap Mouse horseradish peroxidase (Ventana). Afterward, a tertiary antibody was incubated for 16 minutes using prediluted UltraMap Rabbit horseradish peroxidase (Ventana). Estrogen receptor immunostaining was performed using the Ventana DAB kit with standard CC1. Histologic images were obtained using the ScanScope XT digital scanning system (Aperio Technologies, Inc., Vista, CA). Because the distribution of BAF250a expression levels in the patients with kidney cancer were not gaussian, analysis of the differences in the two or three groups of patients was performed using the nonparametric χ2 test and Fisher’s exact test when appropriate. Relationships between different continuous variables were assessed using the Spearman correlation coefficient. The McNemar, Wilcoxon, and Kruskal-Wallis nonparametric tests were also used. Receiver operating characteristic curves were constructed for BAF250a expression levels by plotting sensitivity versus specificity. The areas under the receiver operating curves were analyzed using the Hanley and McNeil method. The X-tile algorithm was used to generate an optimal cutoff point for BAF250a because they are molecules with no established cutoff points for their expression in kidney cancer. Having corrected for the use of minimum P value statistics, the X-tile software yielded an optimal cutoff, equal to the 80th percentile for BAF250a, with a calculated Monte Carlo P value <0.05. Survival analysis was performed in 78 patients with ccRCC with available survival data. 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 BAF250a expression and all other variables for which the patients were characterized. The multivariate models were adjusted for patient age and for tumor size, grade, and stage. Survival analyses were also performed by constructing Kaplan-Meier disease-free survival and overall survival curves. Any differences between the curves were evaluated using the log-rank test. Level 3 gene expression data (normalized gene expression data derived from the Cancer Genome Characterization Center at the University of North Carolina using the Illumina HiSeq RNA Sequencing platform) for ARID1A in ccRCC and normal kidney and overall survival data were obtained from The Cancer Genome Atlas (TCGA), available through the cBio Cancer Genomics Portal (http://www.cbioportal.org/public-portal). The X-tile algorithm was used to generate a prognostic optimal cutoff point to dichotomize ARID1A mRNA expression as ARID1A-positive and ARID1A-negative using a Monte Carlo P value <0.05. TCGA data types, platforms, and methods have been described previously.23Cerami E. Gao J. Dogrusoz U. Gross B.E. Sumer S.O. Aksoy B.A. Jacobsen A. Byrne C.J. Heuer M.L. Larsson E. Antipin Y. Reva B. Goldberg A.P. Sander C. Schultz N. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data.Cancer Discov. 2012; 2: 401-404Crossref PubMed Scopus (8249) Google Scholar Twofold ARID1A mRNA expression difference cutoffs were generated for tumor versus geometric mean of normal kidney. Level 3 copy number data were queried for ARID1A homozygous and hemizygous deletions in ccRCC from TCGA and correlated with matched mRNA expression data and survival data. TCGA data types, platforms, and methods have been described previously.23Cerami E. Gao J. Dogrusoz U. Gross B.E. Sumer S.O. Aksoy B.A. Jacobsen A. Byrne C.J. Heuer M.L. Larsson E. Antipin Y. Reva B. Goldberg A.P. Sander C. Schultz N. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data.Cancer Discov. 2012; 2: 401-404Crossref PubMed Scopus (8249) Google Scholar To explore potential mechanisms of ARID1A inactivation, we recently conducted genome-wide copy number profiling of ccRCC specimens and identified the 1p36.32-p35.3 region harboring ARID1A as a significantly frequent site of copy number loss (16% of specimens).24Girgis A.H. Iakovlev V.V. Beheshti B. Bayani J. Squire J.A. Bui A. Mankaruos M. Youssef Y. Khalil B. Khella H. Pasic M. Yousef G.M. Multilevel whole-genome analysis reveals candidate biomarkers in clear cell renal cell carcinoma.Cancer Res. 2012; 72: 5273-5284Crossref PubMed Scopus (12) Google Scholar To validate our results, we queried ARID1A copy number loss on the ccRCC data set from TCGA (see Materials and Methods). A total of 436 ccRCCs were analyzed. Median patient age of the TCGA set was 61.0 years (range, 25.0--87.0 years). Median tumor size was 3.50 cm (range, 1.40--24.00 cm). Median follow-up was 49 months (range, 1.0--24.0 months). ARID1A copy number loss was observed in 71 of 436 specimens (16%) (homozygous deletion in 2) (Supplemental Figure S1). These findings confirm our results in this independent data set. Copy number loss of the gene-dense 1p36 region includes a number of other tumor-related genes such as KIF1B, AJAP1, and CASP9 (see Discussion for details). The descriptive statistics of the numerical variables of the immunohistochemistry (IHC) analysis are given in Supplemental Table S1. BAF250a was assessed using IHC on TMA as described previously (see Materials and Methods). BAF250a expression was significantly lower in kidney cancer compared with the matching normal kidney cortex from the same patient in 67% of tumors (P < 0.001), higher in cancer in 28%, and comparable to normal in 5% (Table 1; Figures 1 and 2).Table 1ARID1A/BAF250a Expression in Matched Pairs of Cancerous and Normal Tissues of ccRCC from the Same PatientVariablePatients No. (%)P ValueBAF250a protein (TMA analysis) Lower in cancer vs normal53 (67.1)<0.001∗Wilcoxon signed rank test. Higher in cancer vs normal22 (27.8) Equal4 (5.1)ARID1A mRNA (TCGA data set) Lower in cancer vs normal275 (67.9)<0.001†Mann-Whitney test. Higher in cancer vs normal33 (8.15) Equal97 (23.95)ccRCC, clear cell renal cell carcinoma; TCGA, The Cancer Genome Atlas; TMA, tissue microarray.∗ Wilcoxon signed rank test.† Mann-Whitney test. Open table in a new tab Figure 2BAF250a expression is decreased in ccRCC compared with normal kidney tissue from the same patient as assessed via IHC on TMA. BAF250a level is expressed as relative quantification units (RQ), the product of staining intensity, and the stained area. Solid circles represent the BAF250a RQ values of the ccRCC tumor, and the open circles represent the BAF250a RQ values of the normal kidney tissue from the same patient. P value was calculated using the McNemar nonparametric test.View Large Image Figure ViewerDownload Hi-res image Download (PPT) ccRCC, clear cell renal cell carcinoma; TCGA, The Cancer Genome Atlas; TMA, tissue microarray. To complement and further validate our protein expression analysis, we examined ARID1A expression at the mRNA level. We extracted ARID1A mRNA expression data from the publicly available ccRCC dataset of TCGA (n = 404). In accordance with the protein expression, we observed decreased ARID1A mRNA levels in kidney cancer compared with normal renal cortex in 275 of 404 patients (68%) (P < 0.001) (Table 1). The association between BAF250a expression and clinicopathologic variables was assessed using IHC on TMA; results are given in Table 2. BAF250a expression significantly correlated with tumor size, grade, and stage. Significantly higher BAF250a expression was observed in small tumors ( 7 cm) (P = 0.03). In addition, a higher proportion of grade 1 tumors (65%) were BAF250a-positive compared with grade III and IV tumors (26%), and the difference was statistically significant (P = 0.02). Statistically significant higher BAF250a expression was also associated with stage I disease versus stages II and III disease (48% vs 24%, respectively) (P = 0.004). A weak negative correlation was found between BAF250a expression and tumor size (Rs = −0.21; P = 0.02) (data not shown).Table 2Association Between BAF250a Status and Clinocopathologic VariablesVariablePatients, No. (%)P ValueTotal PatientsBAF250a-NegativeBAF250a-PositiveSex Male5129 (56.9)22 (43.1)0.72∗Fisher’s exact test. Female9357 (61.3)36 (38.7)Tumor size, cm 72722 (81.5)5 (18.5)Tumor grade I207 (35.0)13 (65.0)0.02†χ2 test. II7646 (60.5)30 (39.5) III/IV3123 (74.2)8 (25.8)Tumor stage I9449 (52.1)45 (47.9)0.004†χ2 test. II1816 (88.9)2 (11.1) III2116 (76.2)5 (23.8)BAF250a was assessed using IHC on a TMA of patients with ccRCC. Cutoff point: 6.0 relative quantification units equal to 60th percentile.ccRCC, clear cell renal cell carcinoma; TMA, tissue microarray.∗ Fisher’s exact test.† χ2 test. Open table in a new tab BAF250a was assessed using IHC on a TMA of patients with ccRCC. Cutoff point: 6.0 relative quantification units equal to 60th percentile. ccRCC, clear cell renal cell carcinoma; TMA, tissue microarray. Association between BAF250a protein expression and disease-free survival in our cohort of patients with ccRCC was calculated using univariate and multivariate analyses (Table 3). Specimens were dichotomized on the basis of their BAF250a expression to BAF250a-positive and BAF250a-negative categories (see Materials and Methods). Univariate analysis indicated that patients with BAF250a-positive tumors exhibited a significantly longer disease-free survival than did those with BAF250a-negative tumors [hazard ratio (HR) = 0.11; P = 0.006]. BAF250a expression was also a strong predictor of better survival when analyzed as a continuous variable (HR = 0.64; P = 0.02). When controlling for other variables in the multivariate analysis, BAF250a protein expression retained its clinical significance as a marker of better survival (HR = 0.14; P = 0.02). Similar findings were observed when using BAF250a as a continuous variable (HR = 0.07; P = 0.07). BAF250a showed better performance in predicting survival compared with both stage and grade (Table 3). Kaplan-Meier survival curves demonstrated that patients with ccRCC in the BAF250a-positive expression category have statistically significant longer disease-free survival (P = 0.001) when compared with the BAF250a-negative group (Figure 3A).Table 3Correlation between BAF250a expression and disease-free survival in ccRCCVariableHR∗HR estimated from Cox proportional hazard regression model.95% CI†BAF250a status based on a cutoff point of 2.5 relative quantification units equal to the 80th percentile of the distribution of BAF250a values.P ValueUnivariate analysisBAF250a status Categorical variable‡CI of estimated HR.Negative1.00Positive0.110.021 to 0.520.006 Continuous variable0.640.44 to 0.940.02 Stage (ordinal)2.120.98 to 4.590.06 Grade (ordinal)1.110.35 to 3.490.86Multivariate analysis§Multivariate models were adjusted for tumor grade and stage.BAF250a status Categorical variable‡CI of estimated HR.Negative1.00Positive0.140.027 to −0.770.02 Continuous variable0.700.48 to 1.020.07Stage (ordinal)1.700.71 to 0.410.23Grade (ordinal)0.920.35 to 2.340.85ccRCC, clear cell renal carcinoma; CI, confidence interval; HR, hazard ratio.∗ HR estimated from Cox proportional hazard regression model.† BAF250a status based on a cutoff point of 2.5 relative quantification units equal to the 80th percentile of the distribution of BAF250a values.‡ CI of estimated HR.§ Multivariate models were adjusted for tumor grade and stage. Open table in a new tab ccRCC, clear cell renal carcinoma; CI, confidence interval; HR, hazard ratio. We also examined the prognostic significance of ARID1A expression in ccRCC at the mRNA level using the TCGA dataset. On the basis of ARID1A expression, patients were classified as ARID1A-positive and ARID1A-negative using an optimized cutoff value of the 65th percentile (see Materials and Methods), and the resulting groups included 265 ARID1A-positive patients and 139 ARID1A-negative patients. Similar to the results of IHC analysis, we detected a statistically significant correlation between ARID1A mRNA expression and ccRCC size, grade, and stage (Supplemental Table S2). ARID1A positivity was observed significantly more frequently in small tumors (<7 cm) than in larger tumors (≥7 cm) (P = 0.03). ARID1A-positive tumors were significantly grades I and II enriched, whereas ARID1A-negative tumors were associated with grades III and IV (P = 0.03). The clinicopathologic data of the patient cohort is given in Supplemental Tables S3 and S4. Furthermore, stage I and II tumors were significantly associated with ARID1A-positive tumors, whereas ARID1A-negative tumors were more often associated with stage III and IV tumors (P = 0.03). Kaplan-Meier analysis demonstrated significantly longer overall survival among patients with ARID1A-positive tumors compared with tho
Referência(s)