Author's Accepted Manuscript The impact of PBRM1 expression on prognostic and predictive marker in metastatic renal cell carcinoma Ji-Yeon Kim , Se-Hoon Lee , Kyung Chul Moon , Cheol Kwak , Hyeon Hoe Kim , Bhumsuk Keam , Tae Min Kim , Dae Seog Heo
PII: DOI: Reference:
S0022-5347(15)04012-4 10.1016/j.juro.2015.04.114 JURO 12639
To appear in: The Journal of Urology Accepted Date: 22 April 2015 Please cite this article as: Kim JY, Lee SH, Moon KC, Kwak C, Kim HH, Keam B, Kim TM, Heo DS, The impact of PBRM1 expression on prognostic and predictive marker in metastatic renal cell carcinoma, The Journal of Urology® (2015), doi: 10.1016/j.juro.2015.04.114. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain.
Embargo Policy All article content is under embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date. Questions regarding embargo should be directed to [email protected]
The impact of PBRM1 expression on prognostic and predictive marker in metastatic renal cell carcinoma
Keam1, 2, Tae Min Kim1,2, Dae Seog Heo1,2
Ji-Yeon Kim1,2, Se-Hoon Lee1,2,5, Kyung Chul Moon3,5, Cheol Kwak4, Hyeon Hoe Kim4, Bhumsuk
Department of Internal Medicine; 2Cancer Research Institute;
Department of Urology, Seoul National University Hospital, Seoul, Korea
Department of pathology;
Article category: Original Article
M AN U
Running head: PBRM1 expression in metastatic RCC Corresponding authors
Se-Hoon Lee, M.D, Ph.D,
Division of Hematology and Medical Oncology, Department of Internal Medicine, Seoul National
University Hospital, 101 Daehak-ro, Jongno-gu, Seoul 110-744, Korea. Tel: +82-2-2072-0832; Fax: +82-2-764-2199; E-mail: [email protected]
Kyung Chul Moon, M.D, Ph.D,
Division of Pathology, Department of Internal Medicine, Seoul National University Hospital, 101 Daehak-ro, Jongno-gu, Seoul 110-744, Korea. Tel: +82-2-2072-1767; Fax: +82-2-743-5530; E-mail: [email protected]
Key words: PBRM1; Renal cell carcinoma; Prognostic marker; Predictive marker; mTOR inhibitor
ACCEPTED MANUSCRIPT Abstract Background PBRM1, a SWI/SNF chromatin remodeling complex gene, is the second most frequently mutated gene in clear cell renal cell carcinoma (ccRCC). Although previous studies showed loss of PBRM1
expression was associated with poor prognosis of renal cell carcinoma(RCC) patients, these studies were conducted on earlier stage, surgically resected RCC. Accordingly, we investigated the PBRM1expression profile in stage IV RCC patients to determine the prognostic and predictive value
M AN U
Fifty three patients with stage IV or recurred RCC were included. Immunohistochemistry for PBRM1 of formalin-fixed paraffin-embedded tissue microarrays was performed.
Overall, 25 of 53 patients (47%) had high PBRM1 expression in tumor. In analysis comparing survival and treatment response of RCC patients, high PBRM1 expression in the tumor was associated with reduced survival ([median overall survival(OS), 45.0±4.8 vs. 23.0±8.3 months; P
= .022],[median OS in ccRCC, 46.0±4.1 vs. 27.0±6.7 months; P=.053]). Regarding treatment outcome, the higher PBRM1 expression tended to the poorer response to the mTOR inhibitor (median progression free survival(PFS), 3.0±0.2 vs. 1.9±2.3 months; P = .101). By subgroup analysis
according to Heng’s score, this trend was showed more significantly in higher risk group rather than low risk group ([median PFS at low risk, 11.6±1.2 vs. 7.4±7.4 months; P= .157], [median PFS at intermediate risk, 7.1±24.7 vs. 2.9±0.9 months; P= .060]).
Conclusions Our study showed PBRM1 expression level is a potential prognostic marker for advanced RCC. We suggest that determination of tumoral PBRM1 expression level may be used to inform prognosis and treatment of metastatic RCC.
ACCEPTED MANUSCRIPT Introduction Renal cell carcinoma (RCC) is the eighth most common cancer in the world.1 Although most kidney cancer patients receiving curative surgical resection have relatively good prognosis, patients with the most advanced kidney cancers have a dismal prognosis.
Among all types of kidney cancer, three-quarters are clear cell renal cell carcinomas (ccRCC), which is the most common pathologic subtype.2 Genetic alteration of Von Hippel Lindau (VHL), a wellknown tumor suppressor gene, is the traditional hallmark of ccRCC. Loss-of-function of the VHL gene promotes Hypoxia Inducible Factor 1 and 2 stabilization and consequently facilitates cell
proliferation and angiogenesis.3, 4
According to recent comprehensive genetic analyses of next generation sequencing data in ccRCC including The Cancer Genome Atlas (TCGA) cohort, PBRM1, a SWI/SNF chromatin remodeling
M AN U
complex gene, is the second most frequently mutated gene.5, 6 Follow-up study of the TCGA cohort investigating the relationship between PBRM1 and BAP1 mutation and tumor progression show that ccRCC patients with mutant PBRM1 had a better prognosis than those with mutant BAP1.7 On contrary to above study, other investigators report that negative PBRM1 expression is associated with poor prognosis in RCC.8, 9 However, these studies mostly involved patients with earlier stage,
surgically resected cancers. Only 17% of ccRCC cases were stage IV in the TCGA cohort 6, 7, and the study that demonstrated a relationship between PBRM1 expression and prognosis using ccRCC tissue microarray (TMA) utilized a sample consisting of only 10% stage IV ccRCC patients.9 In contrast, RCC at stages I-III has very good prognosis without recurrence, the 5-year overall survival of stage
IV RCC is only 12.1%.1 Therefore, these earlier biomarker studies could not elucidate the impact of PBRM1 expression on clinical outcome in advanced stage metastatic RCC.
Accordingly, we determined the protein expression profile of PBRM1 in stage IV RCC patients and investigated its role as a prognostic and predictive marker.
Materials and Methods Study Population
We used prospectively collected RCC tumors to perform TMA in the department of pathology in Seoul National University Hospital. Additional demographic and medical history data were collected retrospectively. TMA consisted of formalin-fixed paraffin-embedded kidney tissue and was reviewed by a qualified pathologist.
ACCEPTED MANUSCRIPT Among the samples included in the TMA were 53 tumors collected from patients with stage IV or metastatic RCC between 1 January 2003 and 31 December 2012. All patients were followed up and medically treated according to standard protocols. The following variables were included in the data collection: age at diagnosis, gender, smoking, histology of RCC, nephrectomy, stage at diagnosis, and Heng’s score 10 consisting of white blood cell
sequence, response, and survival were also collected retrospectively.
count, hemoglobin, platelet count, serum calcium, LDH, and performance status. Data about treatment
This study was reviewed and approved by the Institutional Review Board of Seoul National
University Hospital (IRB No: 1110-024-380).
M AN U
TMA and immunohistochemistry (IHC)
TMA blocks consisting of a representative tumor core sections (2 mm in diameter) from each formalin-fixed paraffin block (SuperBioChips Laboratories, Seoul, Korea) were manufactured for IHC analysis. IHC staining was performed using the BenchMark XT Slide Preparation System (Ventana Medical Systems, Inc., Tucson, AZ, USA) and a polyclonal rabbit anti-human PB1/BAF180
antibody (1:100, Bethyl Laboratories, Inc., Montgomery, TX, USA).
The IHC scores for PBRM1 expression were calculated as a quantitative measurement at 200× magnification. We only scored the intensity of PBRM1 nuclear staining as previous studies.
Categories were based on the percentage of positive cells per field and scored as follows: negative
(score = 0), 1-10% (score = 1), 11-25% (score = 2), 26-50% (score = 3), or above 50% of immunoreactive cells (score = 4). Three TMA slides per sample were examined and the mean IHC
score was calculated.
Differences in clinical characteristics were compared using Mann-Whitney U tests. Progression-free survival (PFS) was defined as the elapsed time from the first date of treatment to disease progression or death from any cause and overall survival (OS) was defined as the interval from the date of stage IV diagnosis or recurrent RCC to death from any cause. The Kaplan–Meier method was used to analyze PFS and OS. Univariate and multivariate analyses of PFS or OS were conducted using logrank tests and Cox proportional hazards regression tests, respectively. Two-tailed P-values of < 0.05 were considered statistically significant. For evaluating the prognostic value of expression level of
ACCEPTED MANUSCRIPT PBRM1, time-dependent ROC analysis was performed. Time-dependent ROC analysis was analyzed adding weighted value of PBRM1 IHC score on pre-existing known prognostic marker. All data analyses were conducted using IBM SPSS Statistics 21 for Windows (IBM Corp., Armonk, NY, USA).
Results Baseline characteristics
The baseline characteristics of the 53 patients are described in Table 1. The median age of patients
diagnosed with stage IV RCC was 62.3 years (range, 24.8–78.6 years). Twenty-three patients (43.4%) were initially diagnosed with stage IV disease. Forty-four patients (83.0%) were diagnosed with ccRCC, and all patients underwent nephrectomy. Most patients (n = 32, 60.4%) had favorable risk
M AN U
scores according to Heng’s risk criteria.10 Thirty-seven patients (69.8%) were treated with everolimus, and thirty-three patients (57.9%) were treated with sunitinib. Twenty-five patients (47.2%) were treated with both everolimus and sunitinib (Table 1C). Median follow-up duration was 29 months (range, 3–111 months).
IHC for PBRM1 was successful in all patients (n = 53). We calculated the average PBRM1 IHC score from three separate tests, which was considered the representative value for each tumor. We set
score = 2.5).
the median IHC score of PBRM1 as the cut-off value to divide patients into two groups (median IHC
Impact of baseline characteristics including PBRM1 IHC on overall survival Among the baseline data collected, non-clear cell histology, high Heng’s risk score, and high
PBRM1 IHC score were associated with short median OS ([clear vs. non-clear cell histology, 42.0 vs. 17.0 months, P = .015], [favorable vs. intermediate vs. poor, 42.0± vs. 12.0 vs. 3.0, P < .001], [low vs. high, 46.0 vs. 24.0, P = .022]) (Table 1A, Figure 1A). By subgroup analysis, PBRM1 IHC score still affected to median OS in patients with clear cell renal cell carcinoma (n= 44) although statistically significance decreased (low vs. high, 46.0 vs. 27.0, P = .053) (Table 1B, Figure 1B). PBRM1 expression was not significantly correlated with any other clinical baseline characteristic (Table 2). Histology, Heng’s risk score, and IHC score for PBRM1 gave P values < .05 in each univariate analysis and, therefore, were included in the multivariate analysis. In multivariate analysis, these three factors were still related to OS ([clear vs. non-clear cell histology, hazard ratio (HR) = 4.91 (95% confidence interval (CI): 1.95–12.39); P = .001], [favorable vs. intermediate vs. poor, HRintermediate =
ACCEPTED MANUSCRIPT 3.03 (95% CI: 1.38–6.62}, HRpoor = 37.18 (95% CI: 6.17–224.09); P =.002], [low vs. high, HR = 2.29 (95% CI: 1.04–5.04), P = .039]) (Table 3). However, IHC score for PBRM1 was not statistically significant in patients with clear cell renal cell carcinoma (low vs. high, HR = 2.22 (95% CI: 0.94– 5.21; P = .068). In time-dependent ROC analysis, adding the value of the level of PBRM1 expression potentiated
Heng’s risk score to predict patients’ prognosis (Figure 4). We performed these analyses at the 12 and 36-month follow-up. Our results showed that the value of PBRM1 expression level strengthen the predictive efficacy of Heng’s risk score regardless of ccRCC or not. However, the value of prediction was more accurate at 12-month follow-up rather than at 36-month follow-up ([AUC of PBRM1 with
Heng’s score (P+H) = .847, P