Human Pathology (2015) 46, 454–460
www.elsevier.com/locate/humpath
Original contribution
Decreased ARID1A expression correlates with poor prognosis of clear cell renal cell carcinoma☆ Jeong Hwan Park MD a , Cheol Lee MD a , Ja Hee Suh MD a , Ji Yoen Chae MS a , Hwal Woong Kim MD b , Kyung Chul Moon MD, PhD a,c,⁎ a
Department of Pathology, Seoul National University College of Medicine, Seoul, Republic of Korea, 110-799 Department of Pathology, Good Moonhwa Hospital, Busan, Republic of Korea, 601-803 c Kidney Research Institute, Medical Research Center, Seoul National University College of Medicine, Seoul, Republic of Korea, 110-799 b
Received 16 September 2014; revised 3 December 2014; accepted 11 December 2014
Keywords: ARID1A; Clear cell renal cell carcinoma; Progression; Prognosis; Immunohistochemistry
Summary Clear cell renal cell carcinoma (CCRCC) is the most common renal cell carcinoma. It has a relatively unfavorable prognosis compared to other common renal cell carcinomas. Recently, comprehensive molecular studies in CCRCC revealed important genetic alterations, including changes in the VHL, PBRM1, and ARID1A genes. The expression of ARID1A protein is associated with tumor progression and prognosis in many cancers. This study aimed to evaluate the nuclear expression of ARID1A in CCRCC and to assess its expression with the clinical prognosis. The nuclear expression of ARID1A was evaluated in 290 cases of CCRCC by immunohistochemistry. To determine the clinicopathological association with ARID1A, each of the cases was divided into 2 groups, low- and high-expression groups, according to the average proportion of nuclear staining. Decreased ARID1A expression was associated with the higher nuclear grade and higher pTNM stage (P b .001 and P = .013, respectively). The ARID1A low-expression group revealed significantly shorter cancer-specific and progression-free survival times (P = .001 and P b .001, respectively). Furthermore, Cox regression analysis showed that ARID1A expression was an independent prognostic factor for progression-free survival (P = .009). These results suggest that nuclear expression of ARID1A may serve as a new prognostic marker in CCRCC patients. © 2015 Elsevier Inc. All rights reserved.
1. Introduction Kidney cancer ranks among the top 10 most common cancers worldwide [1]. Renal cell carcinoma (RCC) is the most common kidney malignancy in adults and is one of the ☆
Competing interests: The authors declare that they have no conflict of interest ⁎ Corresponding author at: Department of Pathology and Kidney Research Institute, Medical Research Center, Seoul National University College of Medicine, 28 Chongno-gu Yongon-dong, Seoul 110-799, Republic of Korea. E-mail address: [email protected] (K. C. Moon). http://dx.doi.org/10.1016/j.humpath.2014.12.002 0046-8177/© 2015 Elsevier Inc. All rights reserved.
most lethal genitourinary tumors [2]. Among all of the RCCs, clear cell RCC (CCRCC) is the most common subtype and accounts for 60% to 70% of all RCCs. In addition, CCRCC shows a less favorable prognosis than other common subtypes, such as chromophobe and papillary RCC. Recent advances in the molecular pathogenesis of CCRCC revealed several genetic alterations [3-5]. In CCRCC, the Von Hippel–Lindau (VHL) mutation is the most common genetic alteration [3,5], and mutations in the polybromo-1 (PBRM1) gene located in 3p, close to the VHL gene locus, are the second major genetic mutation involved in CCRCC pathogenesis [4]. The VHL tumor
ARID1A expression in CCRCC suppressor gene stabilizes hypoxia-inducible factors (HIF-1α and HIF-2α), and PBRM1 has a role in chromatin remodeling. As mutations in these 2 genes have the main pathogenic role, oncogenic metabolic shift and epigenetic alteration have been regarded as a major pathogenesis of CCRCC. As research extended to epigenetic fields and comprehensive genetic studies became possible, new prognostic markers have been discovered. Recently, the chromatin remodeling gene ARID1A has been suggested as a new prognostic marker in CCRCC [6]. Like PBRM1, ARID1A is a subunit of the switch/sucrose nonfermentable (SWI/SNF) complex [7]. BRG1-associated factor (BAF) and polybromo-associated BAF (PBAF) are human analogs of SWI/SNF-A and SWI/ SNF-B, respectively. Among mutually exclusive subunits of BAF and PBAF, BAF250a and BAF180 are the protein products of ARID1A and PBRM1, respectively. Alterations in the ARID1A gene or loss of ARID1A protein expression have been identified in many cancers. Two forms of endometriosis-associated ovarian carcinomas, clear cell carcinoma and endometrioid carcinoma (approximately 50% and 30% of cases, respectively) [8], high-grade uterine endometrioid carcinomas (46%) [9], gastric carcinoma [10], urothelial bladder tumors [11], and breast cancer [12] were reported to harbor ARID1A mutations or ARID1A loss. Among those cancers with ARID1A mutations or a loss of ARID1A expression, gastric carcinoma, urothelial bladder tumors, and breast cancers had poor prognoses [10-12]. As in other cancers, a loss of ARID1A expression was associated with an unfavorable outcome of CCRCC [6]. In this study, we assessed the nuclear expression of ARID1A in 290 CCRCC samples using immunohistochemistry. We evaluated the clinicopathological correlation with ARID1A expression and assessed its prognostic value.
2. Materials and methods 2.1. Patients and clinicopathological information A total of 290 patients with CCRCC who underwent radical or partial nephrectomy between January 1, 2000, and December 31, 2003, at Seoul National University Hospital were included in this retrospective study. Each CCRCC sample was evaluated with regard to the RCC type, the nuclear grade, and the stage of the tumor. Nuclear grading was reviewed based on the description by Fuhrman et al [13]. Tumor staging was re-evaluated according to the 2010 TNM classification system [14] by reviewing the gross description, gross photos, and representative slides. Tumor recurrence or distant metastasis of CCRCC was determined based on the clinical and radiographic findings. Disease-related deaths were verified by reviewing the patients' medical records. This study was approved by the Institutional Review Board of Seoul National University Hospital.
455
2.2. Tissue microarray The hematoxylin and eosin–stained slides from 290 CCRCC patients were reviewed. For tissue microarray (TMA) construction, a sufficient viable tumor area with no hemorrhage or necrosis was selected in each case (Fig. 1). Three representative core sections (2 mm in diameter) were taken from formalin-fixed paraffin blocks. For each case, 1 core was selected from an area with the worst histologic grade, and the other 2 cores were sampled randomly, considering tumoral heterogeneity. They were embedded in new recipient paraffin blocks (TMA blocks) using a trephine apparatus (Superbiochips Laboratories, Seoul, Republic of Korea). In addition, cortical and medullary portions of 29 nonneoplastic kidney tissues from CCRCC patients were included as positive controls.
2.3. Immunohistochemistry of ARID1A The immunohistochemical staining was conducted on 4-μm-thick sections taken from the TMA slides. Tissue microarray slides were treated to remove wax and rehydrated in a graded series of alcohol solutions. Immunohistochemical staining of ARID1A was performed using the Bond-Max Autostainer (Leica Microsystems, Bannockburn, IL). Monoclonal mouse anti-ARID1A antibody (Santa Cruz Biotechnology, Dallas, TX) was diluted 1:50. After the heat-induced antigen retrieval, the primary antibody was incubated with the samples for 2 hours. The binding of the primary antibody was identified using the Bond Polymer Refine Detection kit (Leica Microsystems) according to the manufacturer's instructions.
2.4. Interpretation of TMA immunohistochemistry Immunoreactivity was sorted semiquantitatively into 4 categories based on the intensity and the extent of nuclear staining (Fig. 1). The intensity was graded as 0 (no staining), weak (granular pattern with light brown), moderate (relatively homogenous pattern with brown), and strong (homogenously dark brown). Tumor cells showing moderate to strong intensity were regarded as having positive nuclear staining. The percent positivity was based on the stained area and was assessed as 0%, less than or equal to 10%, greater than 10 to less than or equal to 50%, and greater than 50%. Finally, we assessed all representative cores of each CCRCC and interpreted them as low expression if an average of less than or equal to 10% of cells stained positive and high expression if an average of greater than 10% stained positive.
2.5. Statistical analysis Correlations between nuclear expression of ARID1A and clinicopathological parameters were analyzed by Pearson's χ2 test and Fisher exact test. The cancer-specific survival
456
J. H. Park et al.
Fig. 1
Flow algorithm for assessment of ARID1A expression in CCRCC.
period was determined by the interval between the primary radical or partial nephrectomy and the final follow-up visit or cancer-related death. The progression-free survival period was defined as the interval between the primary radical or partial nephrectomy and the final follow-up visit or identification of recurrence or metastasis of the CCRCC. The Kaplan-Meier curve and the log-rank test were applied to assess the survival rates and to perform the univariate analysis of overall, cancer-specific, or progression-free survival. A Cox proportional hazards model was used for the multivariate analysis. In all statistical analyses, a 2-tailed P b .05 was regarded as statistically significant. All statistical analyses were performed using IBM SPSS Statistics 21 (IBM SPSS, Chicago, IL).
3. Results 3.1. Clinical and pathologic characteristics of CCRCC patients A total of 290 patients who underwent surgical resection and had confirmed CCRCC were analyzed in this study. The patients were composed of 216 men and 74 women. The mean age was 56 years old (range, 6-82 years), and the average tumor size was 5.50 cm (range, 1.0-22.0). Lymph node metastasis was identified in 6 cases (2.1%), and distant metastasis was found in 30 cases (10.3%). Of the 290 patients, 168 were categorized as stage I (64.4%), 32 were
ARID1A expression in CCRCC
457
Fig. 2 Immunohistochemical findings showing low (A-C) and high (D-F) nuclear expression of ARID1A in CCRCC. Original magnification, ×200.
categorized as stage II (11.1%), 40 were categorized as stage III (13.8%), and 31 were categorized as stage IV (10.7%). The Fuhrman's nuclear grading scale revealed that 20 of the cases were grade 1 (6.9%), 127 cases were grade 2 (43.8%), 106 cases were grade 3 (36.6%), and 37 cases were grade 4 (12.8%). The mean follow-up period was 82.2 months (2-163 months).
3.2. Nuclear expression level of ARID1A in CCRCC The nuclear expression level of ARID1A in tumor cells was evaluated (Fig. 2), and 213 (73.4%) were classified as low expression and 77 (26.6%) were high expression. In nonneoplastic kidney tissues, we identified moderate to strong immunoreactivity of ARID1A on tubular epithelial cells, like previous observation of Lichner et al [6].
3.3. Correlation of ARID1A expression with clinicopathological parameters The correlations of ARID1A expression with clinicopathological parameters are shown in Table 1. Low ARID1A expression was significantly associated with larger tumor size, higher Fuhrman nuclear grade, high pM stage, and advanced pTNM stage.
3.4. Association of ARID1A expression with prognosis The ARID1A low-expression group showed significantly shorter overall, cancer-specific, and progression-free
survival periods than those of the ARID1A high-expression group (Fig. 3). Furthermore, multivariate analysis using the Cox proportional hazards model indicated that ARID1A expression levels were an independent predictor of progression-free survival in patients with CCRCC, when assessed by Fuhrman nuclear grade and the pTNM stage (Table 2). However, ARID1A expression levels failed to show a correlation with cancer-specific survival in a multivariate analysis.
4. Discussion ARID1A mutations and the clinicopathological correlation of ARID1A loss in CCRCC have been studied. Comprehensive molecular studies reveal that ARID1A mutations were identified in 3% of CCRCC [4,5]. One article showed that ARID1A transcript levels were decreased in 30% of renal carcinomas [15]. A recent study revealed lower BAF250a expression in 67% of CCRCC samples and decreased ARID1A messenger RNA (mRNA) levels in 68% compared to normal kidney specimens [6]. The authors identified that a loss of BAF250a expression or lower ARID1A mRNA expression was associated with a larger tumor size, higher Fuhrman nuclear grade, and higher stage. Furthermore, BAF250apositive and ARID1A-positive cancers exhibited a significantly better disease-free and overall survival, respectively. In our study, we identified similar results to these studies. Clear cell renal cell carcinoma with low ARID1A expression showed a larger tumor size, higher Fuhrman nuclear grade,
458
J. H. Park et al.
Table 1 Clinicopathological features of patients with CCRCC and correlation between nuclear expression of ARID1A and clinicopathological parameters Nuclear expression of ARID1a, n (%)
Age (y) ≤55 N55 Sex Female Male Tumor size (cm) ≤7 N7 Nuclear grade Grade 1/2 Grade 3/4 T stage T 1/2 T 3/4 N stage N0/Nx N1 M stage M0 M1 Stage I II III IV
Low (n = 213)
High (n = 77)
99 (46.5) 114 (53.5)
41 (53.2) 36 (46.8)
56 (26.3) 157 (73.7)
18 (23.4) 59 (76.6)
150 (70.8) 62 (29.2)
66 (85.7) 11 (14.3)
93 (43.7) 120 (56.3)
54 (70.1) 23 (29.9)
162 (76.4) 50 (23.6)
66 (85.7) 11 (14.3)
207 (97.2) 6 (2.8)
77 (100.0) 0 (0.0)
185 (86.9) 28 (13.1)
75 (97.4) 2 (2.6)
126 (59.4) 26 (12.3) 31 (14.6) 29 (13.7)
60 6 9 2
P
.352
.651
.009
b.001
.103
.347
.008
.013 (77.9) (7.8) (11.7) (2.6)
high pM stage, and advanced pTNM stage compared to CCRCC with high ARID1A expression. As clinical behavior of T2b tumors are more closely approximating that of T3 tumors, we analyzed the significance of ARID1A expression
on T1/T2a versus T2b/T3/T4 groups. The results showed correlation of ARID1A expression and T stage (P = .034). In addition, the ARID1A low-expression group revealed a worse prognosis with shorter cancer-specific and progression-free survival, compared to the ARID1A highexpression group. Moreover, the ARID1A expression levels were an independent prognostic factor in CCRCC patients with Fuhrman nuclear grade and pTNM stage in progression-free survival. The cutoff value 10% for dividing ARID1A expression level was derived from our results that ARID1A expression was relatively consistent among cores of each case and categorization by 10% positive cells yields the most precise and reliable outcomes (Fig. 1). In addition, our results of low ARID1A expression in 73.4% of CCRCC, by 10% cutoff, were similar to other report of lower expression of BAF250a in 67% [6]. We evaluated each TMA core as score 0, 1, 2, and 3. In many cases, if 1 core showed score 3, 1 of the other 2 cores showed score 2 or 3. The heterogeneity of expression such as score set (3, 1, 0) or (2, 0, 0) was found in 37 (12.8%) of 290 cases. Analyses of correlation between ARID1A expression and clinicopathological features revealed that cases with (2, 0, 0) or (2, 1, 1) showed favorable results and gathering to (2, 2, 1) or (3, 3, 2) cases. So, we decided high-expression cases with (2, 0, 0) or higher score set. The score set (2, 0, 0) means 1 core showed positive nuclear staining with greater than 10% to less than or equal to 50%, on average 30%, and other 2 cores were no staining (0%). Therefore, we described that high expression if an average of greater than 10% stained positive [(30% + 0% + 0%)/3 = 10%]. Several studies have evaluated the impact of ARID1A expression in cancer pathogenesis. First, a cooperative role of ARID1A in the phosphatidylinositol 3-kinase-Aktmammalian target of rapamycin signaling pathway (PI3K/ AKT/mTOR pathway) has been suggested [16]. In ovarian clear cell carcinoma, phosphoinositide-3-kinase, catalytic alpha polypeptide (PIK3CA) mutations occurred more
Fig. 3 Kaplan-Meier curves of overall (A), cancer-specific (B), and progression-free (C) survival in 290 patients with CCRCC according to the 2tiered classification of ARID1A nuclear expression.
ARID1A expression in CCRCC
459
Table 2 Multivariate analysis of cancer-specific and progression-free survival on ARID1A nuclear expression in 290 patients with CCRCC (Cox proportional hazards model) Prognostic factors
Cancer-specific survival Hazard ratio (95% CI)
Nuclear grade 3, 4 vs 1, 2 pTNM stage III, IV vs I, II ARID1A nuclear expression Low vs high
Progression-free survival P
Hazard ratio (95% CI)
b.001 4.551 (1.987-10.422)
.002 2.305 (1.370-3.877)
b.001 11.525 (6.028-22.036)
b.001 7.559 (4.753-12.021)
.243 1.753 (0.684-4.496)
P
.009 2.682 (1.277-5.636)
Abbreviation: CI, confidence interval.
frequently in ARID1A-deficient tumors than in ARID1Aexpressing tumors. The association of ARID1A and TP53 has also been identified [17]. They demonstrated that ARID1A can recruit p53 to the BAF complex and that p53 affects transcriptional regulation of downstream targets of ARID1A, such as CDKN1A and SMAD3, by binding to their promoters. The finding that BAF binds to the promoters and enhances transcriptional activity of SMAD3 in a p53-dependent manner was consistent with the result that low nuclear expression of Smad3 in CCRCC showed significantly shorter cancerspecific and progression-free survival times [18]. The implication of impaired p21 function by ARID1A alterations has been suggested [17,19]. The authors identified that cell cycle arrest failure in BAF250a-deficient cell lines was a result of the impairment of p21 induction and the subsequent up-regulation and repression of the E2F-responsive promoter [19]. Moreover, BAF250a-containing BAF targets c-myc promoter and represses c-myc transcription [20]. This result explains the direct role of BAF on p21 regulation. Other possible mechanisms are associated with homeobox A9 (HOXA9) and hypermethylated in cancer 1 (HIC1) [21,22]. They are 2 of the BAF250a-interacting transcription factors. HIC1 encodes a transcriptional repressor and affects regulatory loops modulating p53-dependent and E2F1dependent cell survival and damage/stress responses [22]. These associations suggest that the loss or dysfunction of ARID1A causes cell cycle alterations and apoptosis evasion. One study presented putative pathogenesis of clear cell carcinoma of the ovary with ARID1A mutations [23]. They noted that ARID1A is part of the ubiquitin ligase complex and is associated with the ubiquitin-mediated degradation of histone H2B. The loss of ARID1A expression resulted in the accumulation of H2B in the nucleus. This accumulation may affect chromatin remodeling, cell cycle regulation, and apoptosis evasion. Epigenetic alterations caused by ARID1A dysfunction may affect proliferation, DNA repair, and tumor suppression [24]. Considering that the chromatin remodeling pathway interacts with a wide range of processes, including immunerelated signaling [25], transcriptional output of genes, such as HIF1A, JUN, FOS, and SP1, BRCA1 function, and TGF-
β and β-catenin signaling, the importance of the chromatin remodeling pathway has been emphasized [5]. In particular, it has been demonstrated that the SWI/SNF chromatin remodeling complex, including PBRM1, PBRM1, and SMARCA4, is the second most frequently mutated subnetwork in CCRCC [5]. Mutations of ARID1A were found in 3% of CCRCCs [4], and a copy number loss of ARID1A was reported in 16% of CCRCCs [6,26]. Other studies, however, reported that the loss or lower expression of ARID1A in CCRCC is more prominent. One study revealed that a deficiency of ARID1A expression was observed in 30% of RCCs [15]. In addition, other studies showed that a significant ARID1A mRNA down-regulation, compared to normal kidney samples was found in 68.8% of CCRCC samples [6]. Moreover, the authors found a significantly lower expression of BAF250a in CCRCCs (67%) when compared to matched normal kidney cortex. These results are similar to our study in which we demonstrate that low expression of ARID1A was present in 73.4% of CCRCCs. The difference between genetic alterations of ARID1A and mRNA or protein expression raises the possibility of regulation in ARID1A inactivation by epigenetic mechanisms, such as hypermethylation of ARID1A promoter regions, interference of mRNA expression, stabilization, or translation, or by micro-RNAs. The peak deletion at 1p36 observed in 14% of CCRCC emphasizes the importance of ARID1A in CCRCC pathogenesis [26]. On the other hand, it generates an interest in other genes located in the same region of CCRCC. Genes located at 1p36 (1p36.32-p35.3) include KIF1B, CASP9, AJAP1, APITD1, and SDHB [26]. KIF1B, CASP9, and AJAP1 were highly associated with hepatocellular carcinoma, non–small cell lung cancer, and glioma, respectively [27-29]. Further studies to assess the pathogenic role and prognostic significance on CCRCC of these genes will be needed. In addition, investigating other SWI/SNF members or chromatin regulators, such as JARID1C, SETD2, UTX, and BAP1, which are mutated or down-regulated, will suggest a more precise mechanism for CCRCC pathogenesis or treatment [4,5,20,30]. The study of ARID1A mutations in endometriosis-associated ovarian carcinomas demonstrated a correlation between
460 ARID1A mutation status and BAF250a immunohistochemical results [8]. Several articles have implicated markers that may be evaluated by immunohistochemistry and have prognostic significance for CCRCC patients. Developing and assessing a molecular scoring system using the aforementioned prognostic markers plus ARID1A expression with regard to prognostic intensity weight and selection would be feasible to perform in a pathology laboratory and helpful for the prediction of CCRCC patients' prognosis, personalized therapy, and gene therapy. In summary, we assessed the clinicopathological correlation and prognostic significance of ARID1A expression by an immunohistochemical study. We demonstrate that low expression level of ARID1A was significantly correlated with higher nuclear grade, advanced pTNM stage, and shorter cancer-specific and progression-free survival times. Multivariate analysis showed ARID1A expression as an independent prognostic factor for progression-free survival and a novel prognostic marker in CCRCC patients.
Acknowledgment This work was supported by the National Research Foundation of Korea grant funded by the Korean government Ministry of Education (MOE) (2013R1A1A2007877).
References [1] Chow WH, Dong LM, Devesa SS. Epidemiology and risk factors for kidney cancer. Nat Rev Urol 2010;7:245-57. [2] Rini BI, Rathmell WK, Godley P. Renal cell carcinoma. Curr Opin Oncol 2008;20:300-6. [3] Linehan WM, Srinivasan R, Schmidt LS. The genetic basis of kidney cancer: a metabolic disease. Nat Rev Urol 2010;7:277-85. [4] Varela I, Tarpey P, Raine K, et al. Exome sequencing identifies frequent mutation of the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature 2011;469:539-42. [5] Cancer Genome Atlas Research Network. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature 2013;499:43-9. [6] Lichner Z, Scorilas A, White NM, et al. The chromatin remodeling gene ARID1A is a new prognostic marker in clear cell renal cell carcinoma. Am J Pathol 2013;182:1163-70. [7] Wilson BG, Roberts CW. SWI/SNF nucleosome remodellers and cancer. Nat Rev Cancer 2011;11:481-92. [8] Wiegand KC, Shah SP, Al-Agha OM, et al. ARID1A mutations in endometriosis-associated ovarian carcinomas. N Engl J Med 2010; 363:1532-43. [9] Allo G, Bernardini MQ, Wu RC, et al. ARID1A loss correlates with mismatch repair deficiency and intact p53 expression in high-grade endometrial carcinomas. Mod Pathol 2014;27:255-61. [10] Wang DD, Chen YB, Pan K, et al. Decreased expression of the ARID1A gene is associated with poor prognosis in primary gastric cancer. PLoS One 2012;7:e40364.
J. H. Park et al. [11] Balbas-Martinez C, Rodriguez-Pinilla M, Casanova A, et al. ARID1A alterations are associated with FGFR3-wild type, poor-prognosis, urothelial bladder tumors. PLoS One 2013;8:e62483. [12] Zhao J, Liu C, Zhao Z. ARID1A: a potential prognostic factor for breast cancer. Tumour Biol 2014;35:4813-9. [13] Fuhrman SA, Lasky LC, Limas C. Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982;6:655-63. [14] Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. AJCC Cancer Staging Manual. 7th ed. New York: Springer; 2009: 479-89. [15] Wang X, Nagl Jr NG, Flowers S, Zweitzig D, Dallas PB, Moran E. Expression of p270 (ARID1A), a component of human SWI/SNF complexes, in human tumors. Int J Cancer 2004;112:636. [16] Yamamoto S, Tsuda H, Takano M, Tamai S, Matsubara O. Loss of ARID1A protein expression occurs as an early event in ovarian clearcell carcinoma development and frequently coexists with PIK3CA mutations. Mod Pathol 2012;25:615-24. [17] Guan B, Wang TL, Shih IeM. ARID1A, a factor that promotes formation of SWI/SNF-mediated chromatin remodeling, is a tumor suppressor in gynecologic cancers. Cancer Res 2011;71: 6718-27. [18] Park JH, Lee C, Suh JH, Chae JY, Moon KC. Nuclear expression of Smad proteins and its prognostic significance in clear cell renal cell carcinoma. HUM PATHOL 2013;44:2047-54. [19] Nagl Jr NG, Patsialou A, Haines DS, Dallas PB, Beck Jr GR, Moran E. The p270 (ARID1A/SMARCF1) subunit of mammalian SWI/SNFrelated complexes is essential for normal cell cycle arrest. Cancer Res 2005;65:9236-44. [20] Nagl Jr NG, Zweitzig DR, Thimmapaya B, Beck Jr GR, Moran E. The c-myc gene is a direct target of mammalian SWI/SNF-related complexes during differentiation-associated cell cycle arrest. Cancer Res 2006;66:1289-93. [21] Gilbert PM, Mouw JK, Unger MA, et al. HOXA9 regulates BRCA1 expression to modulate human breast tumor phenotype. J Clin Invest 2010;120:1535-50. [22] Van Rechem C, Boulay G, Leprince D. HIC1 interacts with a specific subunit of SWI/SNF complexes, ARID1A/BAF250A. Biochem Biophys Res Commun 2009;385:586-90. [23] Shigetomi H, Oonogi A, Tsunemi T, et al. The role of components of the chromatin modification machinery in carcinogenesis of clear cell carcinoma of the ovary (Review). Oncol Lett 2011;2:591-7. [24] Reisman D, Glaros S, Thompson EA. The SWI/SNF complex and cancer. Oncogene 2009;28:1653-68. [25] Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609-15. [26] Girgis AH, Iakovlev VV, Beheshti B, et al. Multilevel whole-genome analysis reveals candidate biomarkers in clear cell renal cell carcinoma. Cancer Res 2012;72:5273-84. [27] Zhang H, Zhai Y, Hu Z, et al. Genome-wide association study identifies 1p36.22 as a new susceptibility locus for hepatocellular carcinoma in chronic hepatitis B virus carriers. Nat Genet 2010;42:755-8. [28] Shultz JC, Goehe RW, Murudkar CS, et al. SRSF1 regulates the alternative splicing of caspase 9 via a novel intronic splicing enhancer affecting the chemotherapeutic sensitivity of non-small cell lung cancer cells. Mol Cancer Res 2011;9:889-900. [29] Lin N, Di C, Bortoff K, et al. Deletion or epigenetic silencing of AJAP1 on 1p36 in glioblastoma. Mol Cancer Res 2012;10:208-17. [30] Dalgliesh GL, Furge K, Greenman C, et al. Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes. Nature 2010;463:360-3.
www.elsevier.com/locate/humpath
Original contribution
Decreased ARID1A expression correlates with poor prognosis of clear cell renal cell carcinoma☆ Jeong Hwan Park MD a , Cheol Lee MD a , Ja Hee Suh MD a , Ji Yoen Chae MS a , Hwal Woong Kim MD b , Kyung Chul Moon MD, PhD a,c,⁎ a
Department of Pathology, Seoul National University College of Medicine, Seoul, Republic of Korea, 110-799 Department of Pathology, Good Moonhwa Hospital, Busan, Republic of Korea, 601-803 c Kidney Research Institute, Medical Research Center, Seoul National University College of Medicine, Seoul, Republic of Korea, 110-799 b
Received 16 September 2014; revised 3 December 2014; accepted 11 December 2014
Keywords: ARID1A; Clear cell renal cell carcinoma; Progression; Prognosis; Immunohistochemistry
Summary Clear cell renal cell carcinoma (CCRCC) is the most common renal cell carcinoma. It has a relatively unfavorable prognosis compared to other common renal cell carcinomas. Recently, comprehensive molecular studies in CCRCC revealed important genetic alterations, including changes in the VHL, PBRM1, and ARID1A genes. The expression of ARID1A protein is associated with tumor progression and prognosis in many cancers. This study aimed to evaluate the nuclear expression of ARID1A in CCRCC and to assess its expression with the clinical prognosis. The nuclear expression of ARID1A was evaluated in 290 cases of CCRCC by immunohistochemistry. To determine the clinicopathological association with ARID1A, each of the cases was divided into 2 groups, low- and high-expression groups, according to the average proportion of nuclear staining. Decreased ARID1A expression was associated with the higher nuclear grade and higher pTNM stage (P b .001 and P = .013, respectively). The ARID1A low-expression group revealed significantly shorter cancer-specific and progression-free survival times (P = .001 and P b .001, respectively). Furthermore, Cox regression analysis showed that ARID1A expression was an independent prognostic factor for progression-free survival (P = .009). These results suggest that nuclear expression of ARID1A may serve as a new prognostic marker in CCRCC patients. © 2015 Elsevier Inc. All rights reserved.
1. Introduction Kidney cancer ranks among the top 10 most common cancers worldwide [1]. Renal cell carcinoma (RCC) is the most common kidney malignancy in adults and is one of the ☆
Competing interests: The authors declare that they have no conflict of interest ⁎ Corresponding author at: Department of Pathology and Kidney Research Institute, Medical Research Center, Seoul National University College of Medicine, 28 Chongno-gu Yongon-dong, Seoul 110-799, Republic of Korea. E-mail address: [email protected] (K. C. Moon). http://dx.doi.org/10.1016/j.humpath.2014.12.002 0046-8177/© 2015 Elsevier Inc. All rights reserved.
most lethal genitourinary tumors [2]. Among all of the RCCs, clear cell RCC (CCRCC) is the most common subtype and accounts for 60% to 70% of all RCCs. In addition, CCRCC shows a less favorable prognosis than other common subtypes, such as chromophobe and papillary RCC. Recent advances in the molecular pathogenesis of CCRCC revealed several genetic alterations [3-5]. In CCRCC, the Von Hippel–Lindau (VHL) mutation is the most common genetic alteration [3,5], and mutations in the polybromo-1 (PBRM1) gene located in 3p, close to the VHL gene locus, are the second major genetic mutation involved in CCRCC pathogenesis [4]. The VHL tumor
ARID1A expression in CCRCC suppressor gene stabilizes hypoxia-inducible factors (HIF-1α and HIF-2α), and PBRM1 has a role in chromatin remodeling. As mutations in these 2 genes have the main pathogenic role, oncogenic metabolic shift and epigenetic alteration have been regarded as a major pathogenesis of CCRCC. As research extended to epigenetic fields and comprehensive genetic studies became possible, new prognostic markers have been discovered. Recently, the chromatin remodeling gene ARID1A has been suggested as a new prognostic marker in CCRCC [6]. Like PBRM1, ARID1A is a subunit of the switch/sucrose nonfermentable (SWI/SNF) complex [7]. BRG1-associated factor (BAF) and polybromo-associated BAF (PBAF) are human analogs of SWI/SNF-A and SWI/ SNF-B, respectively. Among mutually exclusive subunits of BAF and PBAF, BAF250a and BAF180 are the protein products of ARID1A and PBRM1, respectively. Alterations in the ARID1A gene or loss of ARID1A protein expression have been identified in many cancers. Two forms of endometriosis-associated ovarian carcinomas, clear cell carcinoma and endometrioid carcinoma (approximately 50% and 30% of cases, respectively) [8], high-grade uterine endometrioid carcinomas (46%) [9], gastric carcinoma [10], urothelial bladder tumors [11], and breast cancer [12] were reported to harbor ARID1A mutations or ARID1A loss. Among those cancers with ARID1A mutations or a loss of ARID1A expression, gastric carcinoma, urothelial bladder tumors, and breast cancers had poor prognoses [10-12]. As in other cancers, a loss of ARID1A expression was associated with an unfavorable outcome of CCRCC [6]. In this study, we assessed the nuclear expression of ARID1A in 290 CCRCC samples using immunohistochemistry. We evaluated the clinicopathological correlation with ARID1A expression and assessed its prognostic value.
2. Materials and methods 2.1. Patients and clinicopathological information A total of 290 patients with CCRCC who underwent radical or partial nephrectomy between January 1, 2000, and December 31, 2003, at Seoul National University Hospital were included in this retrospective study. Each CCRCC sample was evaluated with regard to the RCC type, the nuclear grade, and the stage of the tumor. Nuclear grading was reviewed based on the description by Fuhrman et al [13]. Tumor staging was re-evaluated according to the 2010 TNM classification system [14] by reviewing the gross description, gross photos, and representative slides. Tumor recurrence or distant metastasis of CCRCC was determined based on the clinical and radiographic findings. Disease-related deaths were verified by reviewing the patients' medical records. This study was approved by the Institutional Review Board of Seoul National University Hospital.
455
2.2. Tissue microarray The hematoxylin and eosin–stained slides from 290 CCRCC patients were reviewed. For tissue microarray (TMA) construction, a sufficient viable tumor area with no hemorrhage or necrosis was selected in each case (Fig. 1). Three representative core sections (2 mm in diameter) were taken from formalin-fixed paraffin blocks. For each case, 1 core was selected from an area with the worst histologic grade, and the other 2 cores were sampled randomly, considering tumoral heterogeneity. They were embedded in new recipient paraffin blocks (TMA blocks) using a trephine apparatus (Superbiochips Laboratories, Seoul, Republic of Korea). In addition, cortical and medullary portions of 29 nonneoplastic kidney tissues from CCRCC patients were included as positive controls.
2.3. Immunohistochemistry of ARID1A The immunohistochemical staining was conducted on 4-μm-thick sections taken from the TMA slides. Tissue microarray slides were treated to remove wax and rehydrated in a graded series of alcohol solutions. Immunohistochemical staining of ARID1A was performed using the Bond-Max Autostainer (Leica Microsystems, Bannockburn, IL). Monoclonal mouse anti-ARID1A antibody (Santa Cruz Biotechnology, Dallas, TX) was diluted 1:50. After the heat-induced antigen retrieval, the primary antibody was incubated with the samples for 2 hours. The binding of the primary antibody was identified using the Bond Polymer Refine Detection kit (Leica Microsystems) according to the manufacturer's instructions.
2.4. Interpretation of TMA immunohistochemistry Immunoreactivity was sorted semiquantitatively into 4 categories based on the intensity and the extent of nuclear staining (Fig. 1). The intensity was graded as 0 (no staining), weak (granular pattern with light brown), moderate (relatively homogenous pattern with brown), and strong (homogenously dark brown). Tumor cells showing moderate to strong intensity were regarded as having positive nuclear staining. The percent positivity was based on the stained area and was assessed as 0%, less than or equal to 10%, greater than 10 to less than or equal to 50%, and greater than 50%. Finally, we assessed all representative cores of each CCRCC and interpreted them as low expression if an average of less than or equal to 10% of cells stained positive and high expression if an average of greater than 10% stained positive.
2.5. Statistical analysis Correlations between nuclear expression of ARID1A and clinicopathological parameters were analyzed by Pearson's χ2 test and Fisher exact test. The cancer-specific survival
456
J. H. Park et al.
Fig. 1
Flow algorithm for assessment of ARID1A expression in CCRCC.
period was determined by the interval between the primary radical or partial nephrectomy and the final follow-up visit or cancer-related death. The progression-free survival period was defined as the interval between the primary radical or partial nephrectomy and the final follow-up visit or identification of recurrence or metastasis of the CCRCC. The Kaplan-Meier curve and the log-rank test were applied to assess the survival rates and to perform the univariate analysis of overall, cancer-specific, or progression-free survival. A Cox proportional hazards model was used for the multivariate analysis. In all statistical analyses, a 2-tailed P b .05 was regarded as statistically significant. All statistical analyses were performed using IBM SPSS Statistics 21 (IBM SPSS, Chicago, IL).
3. Results 3.1. Clinical and pathologic characteristics of CCRCC patients A total of 290 patients who underwent surgical resection and had confirmed CCRCC were analyzed in this study. The patients were composed of 216 men and 74 women. The mean age was 56 years old (range, 6-82 years), and the average tumor size was 5.50 cm (range, 1.0-22.0). Lymph node metastasis was identified in 6 cases (2.1%), and distant metastasis was found in 30 cases (10.3%). Of the 290 patients, 168 were categorized as stage I (64.4%), 32 were
ARID1A expression in CCRCC
457
Fig. 2 Immunohistochemical findings showing low (A-C) and high (D-F) nuclear expression of ARID1A in CCRCC. Original magnification, ×200.
categorized as stage II (11.1%), 40 were categorized as stage III (13.8%), and 31 were categorized as stage IV (10.7%). The Fuhrman's nuclear grading scale revealed that 20 of the cases were grade 1 (6.9%), 127 cases were grade 2 (43.8%), 106 cases were grade 3 (36.6%), and 37 cases were grade 4 (12.8%). The mean follow-up period was 82.2 months (2-163 months).
3.2. Nuclear expression level of ARID1A in CCRCC The nuclear expression level of ARID1A in tumor cells was evaluated (Fig. 2), and 213 (73.4%) were classified as low expression and 77 (26.6%) were high expression. In nonneoplastic kidney tissues, we identified moderate to strong immunoreactivity of ARID1A on tubular epithelial cells, like previous observation of Lichner et al [6].
3.3. Correlation of ARID1A expression with clinicopathological parameters The correlations of ARID1A expression with clinicopathological parameters are shown in Table 1. Low ARID1A expression was significantly associated with larger tumor size, higher Fuhrman nuclear grade, high pM stage, and advanced pTNM stage.
3.4. Association of ARID1A expression with prognosis The ARID1A low-expression group showed significantly shorter overall, cancer-specific, and progression-free
survival periods than those of the ARID1A high-expression group (Fig. 3). Furthermore, multivariate analysis using the Cox proportional hazards model indicated that ARID1A expression levels were an independent predictor of progression-free survival in patients with CCRCC, when assessed by Fuhrman nuclear grade and the pTNM stage (Table 2). However, ARID1A expression levels failed to show a correlation with cancer-specific survival in a multivariate analysis.
4. Discussion ARID1A mutations and the clinicopathological correlation of ARID1A loss in CCRCC have been studied. Comprehensive molecular studies reveal that ARID1A mutations were identified in 3% of CCRCC [4,5]. One article showed that ARID1A transcript levels were decreased in 30% of renal carcinomas [15]. A recent study revealed lower BAF250a expression in 67% of CCRCC samples and decreased ARID1A messenger RNA (mRNA) levels in 68% compared to normal kidney specimens [6]. The authors identified that a loss of BAF250a expression or lower ARID1A mRNA expression was associated with a larger tumor size, higher Fuhrman nuclear grade, and higher stage. Furthermore, BAF250apositive and ARID1A-positive cancers exhibited a significantly better disease-free and overall survival, respectively. In our study, we identified similar results to these studies. Clear cell renal cell carcinoma with low ARID1A expression showed a larger tumor size, higher Fuhrman nuclear grade,
458
J. H. Park et al.
Table 1 Clinicopathological features of patients with CCRCC and correlation between nuclear expression of ARID1A and clinicopathological parameters Nuclear expression of ARID1a, n (%)
Age (y) ≤55 N55 Sex Female Male Tumor size (cm) ≤7 N7 Nuclear grade Grade 1/2 Grade 3/4 T stage T 1/2 T 3/4 N stage N0/Nx N1 M stage M0 M1 Stage I II III IV
Low (n = 213)
High (n = 77)
99 (46.5) 114 (53.5)
41 (53.2) 36 (46.8)
56 (26.3) 157 (73.7)
18 (23.4) 59 (76.6)
150 (70.8) 62 (29.2)
66 (85.7) 11 (14.3)
93 (43.7) 120 (56.3)
54 (70.1) 23 (29.9)
162 (76.4) 50 (23.6)
66 (85.7) 11 (14.3)
207 (97.2) 6 (2.8)
77 (100.0) 0 (0.0)
185 (86.9) 28 (13.1)
75 (97.4) 2 (2.6)
126 (59.4) 26 (12.3) 31 (14.6) 29 (13.7)
60 6 9 2
P
.352
.651
.009
b.001
.103
.347
.008
.013 (77.9) (7.8) (11.7) (2.6)
high pM stage, and advanced pTNM stage compared to CCRCC with high ARID1A expression. As clinical behavior of T2b tumors are more closely approximating that of T3 tumors, we analyzed the significance of ARID1A expression
on T1/T2a versus T2b/T3/T4 groups. The results showed correlation of ARID1A expression and T stage (P = .034). In addition, the ARID1A low-expression group revealed a worse prognosis with shorter cancer-specific and progression-free survival, compared to the ARID1A highexpression group. Moreover, the ARID1A expression levels were an independent prognostic factor in CCRCC patients with Fuhrman nuclear grade and pTNM stage in progression-free survival. The cutoff value 10% for dividing ARID1A expression level was derived from our results that ARID1A expression was relatively consistent among cores of each case and categorization by 10% positive cells yields the most precise and reliable outcomes (Fig. 1). In addition, our results of low ARID1A expression in 73.4% of CCRCC, by 10% cutoff, were similar to other report of lower expression of BAF250a in 67% [6]. We evaluated each TMA core as score 0, 1, 2, and 3. In many cases, if 1 core showed score 3, 1 of the other 2 cores showed score 2 or 3. The heterogeneity of expression such as score set (3, 1, 0) or (2, 0, 0) was found in 37 (12.8%) of 290 cases. Analyses of correlation between ARID1A expression and clinicopathological features revealed that cases with (2, 0, 0) or (2, 1, 1) showed favorable results and gathering to (2, 2, 1) or (3, 3, 2) cases. So, we decided high-expression cases with (2, 0, 0) or higher score set. The score set (2, 0, 0) means 1 core showed positive nuclear staining with greater than 10% to less than or equal to 50%, on average 30%, and other 2 cores were no staining (0%). Therefore, we described that high expression if an average of greater than 10% stained positive [(30% + 0% + 0%)/3 = 10%]. Several studies have evaluated the impact of ARID1A expression in cancer pathogenesis. First, a cooperative role of ARID1A in the phosphatidylinositol 3-kinase-Aktmammalian target of rapamycin signaling pathway (PI3K/ AKT/mTOR pathway) has been suggested [16]. In ovarian clear cell carcinoma, phosphoinositide-3-kinase, catalytic alpha polypeptide (PIK3CA) mutations occurred more
Fig. 3 Kaplan-Meier curves of overall (A), cancer-specific (B), and progression-free (C) survival in 290 patients with CCRCC according to the 2tiered classification of ARID1A nuclear expression.
ARID1A expression in CCRCC
459
Table 2 Multivariate analysis of cancer-specific and progression-free survival on ARID1A nuclear expression in 290 patients with CCRCC (Cox proportional hazards model) Prognostic factors
Cancer-specific survival Hazard ratio (95% CI)
Nuclear grade 3, 4 vs 1, 2 pTNM stage III, IV vs I, II ARID1A nuclear expression Low vs high
Progression-free survival P
Hazard ratio (95% CI)
b.001 4.551 (1.987-10.422)
.002 2.305 (1.370-3.877)
b.001 11.525 (6.028-22.036)
b.001 7.559 (4.753-12.021)
.243 1.753 (0.684-4.496)
P
.009 2.682 (1.277-5.636)
Abbreviation: CI, confidence interval.
frequently in ARID1A-deficient tumors than in ARID1Aexpressing tumors. The association of ARID1A and TP53 has also been identified [17]. They demonstrated that ARID1A can recruit p53 to the BAF complex and that p53 affects transcriptional regulation of downstream targets of ARID1A, such as CDKN1A and SMAD3, by binding to their promoters. The finding that BAF binds to the promoters and enhances transcriptional activity of SMAD3 in a p53-dependent manner was consistent with the result that low nuclear expression of Smad3 in CCRCC showed significantly shorter cancerspecific and progression-free survival times [18]. The implication of impaired p21 function by ARID1A alterations has been suggested [17,19]. The authors identified that cell cycle arrest failure in BAF250a-deficient cell lines was a result of the impairment of p21 induction and the subsequent up-regulation and repression of the E2F-responsive promoter [19]. Moreover, BAF250a-containing BAF targets c-myc promoter and represses c-myc transcription [20]. This result explains the direct role of BAF on p21 regulation. Other possible mechanisms are associated with homeobox A9 (HOXA9) and hypermethylated in cancer 1 (HIC1) [21,22]. They are 2 of the BAF250a-interacting transcription factors. HIC1 encodes a transcriptional repressor and affects regulatory loops modulating p53-dependent and E2F1dependent cell survival and damage/stress responses [22]. These associations suggest that the loss or dysfunction of ARID1A causes cell cycle alterations and apoptosis evasion. One study presented putative pathogenesis of clear cell carcinoma of the ovary with ARID1A mutations [23]. They noted that ARID1A is part of the ubiquitin ligase complex and is associated with the ubiquitin-mediated degradation of histone H2B. The loss of ARID1A expression resulted in the accumulation of H2B in the nucleus. This accumulation may affect chromatin remodeling, cell cycle regulation, and apoptosis evasion. Epigenetic alterations caused by ARID1A dysfunction may affect proliferation, DNA repair, and tumor suppression [24]. Considering that the chromatin remodeling pathway interacts with a wide range of processes, including immunerelated signaling [25], transcriptional output of genes, such as HIF1A, JUN, FOS, and SP1, BRCA1 function, and TGF-
β and β-catenin signaling, the importance of the chromatin remodeling pathway has been emphasized [5]. In particular, it has been demonstrated that the SWI/SNF chromatin remodeling complex, including PBRM1, PBRM1, and SMARCA4, is the second most frequently mutated subnetwork in CCRCC [5]. Mutations of ARID1A were found in 3% of CCRCCs [4], and a copy number loss of ARID1A was reported in 16% of CCRCCs [6,26]. Other studies, however, reported that the loss or lower expression of ARID1A in CCRCC is more prominent. One study revealed that a deficiency of ARID1A expression was observed in 30% of RCCs [15]. In addition, other studies showed that a significant ARID1A mRNA down-regulation, compared to normal kidney samples was found in 68.8% of CCRCC samples [6]. Moreover, the authors found a significantly lower expression of BAF250a in CCRCCs (67%) when compared to matched normal kidney cortex. These results are similar to our study in which we demonstrate that low expression of ARID1A was present in 73.4% of CCRCCs. The difference between genetic alterations of ARID1A and mRNA or protein expression raises the possibility of regulation in ARID1A inactivation by epigenetic mechanisms, such as hypermethylation of ARID1A promoter regions, interference of mRNA expression, stabilization, or translation, or by micro-RNAs. The peak deletion at 1p36 observed in 14% of CCRCC emphasizes the importance of ARID1A in CCRCC pathogenesis [26]. On the other hand, it generates an interest in other genes located in the same region of CCRCC. Genes located at 1p36 (1p36.32-p35.3) include KIF1B, CASP9, AJAP1, APITD1, and SDHB [26]. KIF1B, CASP9, and AJAP1 were highly associated with hepatocellular carcinoma, non–small cell lung cancer, and glioma, respectively [27-29]. Further studies to assess the pathogenic role and prognostic significance on CCRCC of these genes will be needed. In addition, investigating other SWI/SNF members or chromatin regulators, such as JARID1C, SETD2, UTX, and BAP1, which are mutated or down-regulated, will suggest a more precise mechanism for CCRCC pathogenesis or treatment [4,5,20,30]. The study of ARID1A mutations in endometriosis-associated ovarian carcinomas demonstrated a correlation between
460 ARID1A mutation status and BAF250a immunohistochemical results [8]. Several articles have implicated markers that may be evaluated by immunohistochemistry and have prognostic significance for CCRCC patients. Developing and assessing a molecular scoring system using the aforementioned prognostic markers plus ARID1A expression with regard to prognostic intensity weight and selection would be feasible to perform in a pathology laboratory and helpful for the prediction of CCRCC patients' prognosis, personalized therapy, and gene therapy. In summary, we assessed the clinicopathological correlation and prognostic significance of ARID1A expression by an immunohistochemical study. We demonstrate that low expression level of ARID1A was significantly correlated with higher nuclear grade, advanced pTNM stage, and shorter cancer-specific and progression-free survival times. Multivariate analysis showed ARID1A expression as an independent prognostic factor for progression-free survival and a novel prognostic marker in CCRCC patients.
Acknowledgment This work was supported by the National Research Foundation of Korea grant funded by the Korean government Ministry of Education (MOE) (2013R1A1A2007877).
References [1] Chow WH, Dong LM, Devesa SS. Epidemiology and risk factors for kidney cancer. Nat Rev Urol 2010;7:245-57. [2] Rini BI, Rathmell WK, Godley P. Renal cell carcinoma. Curr Opin Oncol 2008;20:300-6. [3] Linehan WM, Srinivasan R, Schmidt LS. The genetic basis of kidney cancer: a metabolic disease. Nat Rev Urol 2010;7:277-85. [4] Varela I, Tarpey P, Raine K, et al. Exome sequencing identifies frequent mutation of the SWI/SNF complex gene PBRM1 in renal carcinoma. Nature 2011;469:539-42. [5] Cancer Genome Atlas Research Network. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature 2013;499:43-9. [6] Lichner Z, Scorilas A, White NM, et al. The chromatin remodeling gene ARID1A is a new prognostic marker in clear cell renal cell carcinoma. Am J Pathol 2013;182:1163-70. [7] Wilson BG, Roberts CW. SWI/SNF nucleosome remodellers and cancer. Nat Rev Cancer 2011;11:481-92. [8] Wiegand KC, Shah SP, Al-Agha OM, et al. ARID1A mutations in endometriosis-associated ovarian carcinomas. N Engl J Med 2010; 363:1532-43. [9] Allo G, Bernardini MQ, Wu RC, et al. ARID1A loss correlates with mismatch repair deficiency and intact p53 expression in high-grade endometrial carcinomas. Mod Pathol 2014;27:255-61. [10] Wang DD, Chen YB, Pan K, et al. Decreased expression of the ARID1A gene is associated with poor prognosis in primary gastric cancer. PLoS One 2012;7:e40364.
J. H. Park et al. [11] Balbas-Martinez C, Rodriguez-Pinilla M, Casanova A, et al. ARID1A alterations are associated with FGFR3-wild type, poor-prognosis, urothelial bladder tumors. PLoS One 2013;8:e62483. [12] Zhao J, Liu C, Zhao Z. ARID1A: a potential prognostic factor for breast cancer. Tumour Biol 2014;35:4813-9. [13] Fuhrman SA, Lasky LC, Limas C. Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982;6:655-63. [14] Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL, Trotti A. AJCC Cancer Staging Manual. 7th ed. New York: Springer; 2009: 479-89. [15] Wang X, Nagl Jr NG, Flowers S, Zweitzig D, Dallas PB, Moran E. Expression of p270 (ARID1A), a component of human SWI/SNF complexes, in human tumors. Int J Cancer 2004;112:636. [16] Yamamoto S, Tsuda H, Takano M, Tamai S, Matsubara O. Loss of ARID1A protein expression occurs as an early event in ovarian clearcell carcinoma development and frequently coexists with PIK3CA mutations. Mod Pathol 2012;25:615-24. [17] Guan B, Wang TL, Shih IeM. ARID1A, a factor that promotes formation of SWI/SNF-mediated chromatin remodeling, is a tumor suppressor in gynecologic cancers. Cancer Res 2011;71: 6718-27. [18] Park JH, Lee C, Suh JH, Chae JY, Moon KC. Nuclear expression of Smad proteins and its prognostic significance in clear cell renal cell carcinoma. HUM PATHOL 2013;44:2047-54. [19] Nagl Jr NG, Patsialou A, Haines DS, Dallas PB, Beck Jr GR, Moran E. The p270 (ARID1A/SMARCF1) subunit of mammalian SWI/SNFrelated complexes is essential for normal cell cycle arrest. Cancer Res 2005;65:9236-44. [20] Nagl Jr NG, Zweitzig DR, Thimmapaya B, Beck Jr GR, Moran E. The c-myc gene is a direct target of mammalian SWI/SNF-related complexes during differentiation-associated cell cycle arrest. Cancer Res 2006;66:1289-93. [21] Gilbert PM, Mouw JK, Unger MA, et al. HOXA9 regulates BRCA1 expression to modulate human breast tumor phenotype. J Clin Invest 2010;120:1535-50. [22] Van Rechem C, Boulay G, Leprince D. HIC1 interacts with a specific subunit of SWI/SNF complexes, ARID1A/BAF250A. Biochem Biophys Res Commun 2009;385:586-90. [23] Shigetomi H, Oonogi A, Tsunemi T, et al. The role of components of the chromatin modification machinery in carcinogenesis of clear cell carcinoma of the ovary (Review). Oncol Lett 2011;2:591-7. [24] Reisman D, Glaros S, Thompson EA. The SWI/SNF complex and cancer. Oncogene 2009;28:1653-68. [25] Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609-15. [26] Girgis AH, Iakovlev VV, Beheshti B, et al. Multilevel whole-genome analysis reveals candidate biomarkers in clear cell renal cell carcinoma. Cancer Res 2012;72:5273-84. [27] Zhang H, Zhai Y, Hu Z, et al. Genome-wide association study identifies 1p36.22 as a new susceptibility locus for hepatocellular carcinoma in chronic hepatitis B virus carriers. Nat Genet 2010;42:755-8. [28] Shultz JC, Goehe RW, Murudkar CS, et al. SRSF1 regulates the alternative splicing of caspase 9 via a novel intronic splicing enhancer affecting the chemotherapeutic sensitivity of non-small cell lung cancer cells. Mol Cancer Res 2011;9:889-900. [29] Lin N, Di C, Bortoff K, et al. Deletion or epigenetic silencing of AJAP1 on 1p36 in glioblastoma. Mol Cancer Res 2012;10:208-17. [30] Dalgliesh GL, Furge K, Greenman C, et al. Systematic sequencing of renal carcinoma reveals inactivation of histone modifying genes. Nature 2010;463:360-3.
