- Email: [email protected]
High Expression of Karyopherin-α2 Defines Poor Prognosis in Non–Muscle-Invasive Bladder Cancer and in Patients with Invasive Bladder Cancer Undergoing Radical Cystectomy
EUROPEAN UROLOGY 59 (2011) 841–848
available at www.sciencedirect.com journal homepage: www.europeanurology.com
Bladder Cancer
High Expression of Karyopherin-a2 Defines Poor Prognosis in Non–Muscle-Invasive Bladder Cancer and in Patients with Invasive Bladder Cancer Undergoing Radical Cystectomy Jørgen Bjerggaard Jensen a,*, Pia Pinholt Munksgaard b, Christoffer Mørk Sørensen b, Niels Fristrup b, Karin Birkenkamp-Demtroder b, Benedicte Parm Ulhøi c, Klaus Møller-Ernst Jensen a, Torben F. Ørntoft b, Lars Dyrskjøt b a
Department of Urology, Aarhus University Hospital, Skejby, Denmark
b
Department of Molecular Medicine, Aarhus University Hospital, Skejby, Denmark
c
Institute of Pathology, Aarhus University Hospital, Aarhus Sygehus NBG, Denmark
Article info
Abstract
Article history: Accepted January 27, 2011 Published online ahead of print on February 9, 2011
Background: Conventional clinicopathologic risk factors have failed to accurately predict the prognosis of patients with bladder cancer (BC). Objective: To evaluate karyopherin-a2 (KPNA2) expression as a progression marker in patients with non–muscle-invasive BC (NMIBC) treated by conservative methods and as a prognostic marker in patients with invasive BC undergoing radical cystectomy (RC). Design, setting, and participants: Two different tissue microarrays were constructed, one with 234 primary Ta/T1 tumours from patients treated by transurethral resection of the bladder and one with 377 tumours from RC patients. Intervention: KPNA2 expression based on immunohistochemistry. Measurements: Risk of progression of Ta/T1 patients to muscle-invasive BC was estimated in clinical follow-up to progression or a minimum of 53 mo. Risk of recurrent disease and death following RC was estimated in clinical follow-up of a minimum of 24 mo in patients alive. Results and limitations: A high KPNA2 expression in Ta/T1 patients was significantly correlated with a higher risk of progression that was independent of conventional risk factors in multivariate analysis. In patients undergoing RC, a high KPNA2 expression was an independent predictor of poor prognosis. A high KPNA2 expression was correlated with a higher risk of visceral metastasis rather than lymphatic spread. Conclusions: KPNA2 expression is a marker for progression of NMIBC and a prognostic marker in patients undergoing RC.
Keywords: Urothelial carcinoma Bladder cancer KPNA2 Biomarker Prognostic marker Radical cystectomy TMA Tissue microarray
# 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Urology, Aarhus University Hospital, Skejby, Brendstrupgaardsvej 100, DK-8200 Aarhus N, Denmark. Tel. +45 89495950; Fax: +45 89496006. E-mail address: [email protected] (J.B. Jensen).
0302-2838/$ – see back matter # 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.eururo.2011.01.048
842
1.
EUROPEAN UROLOGY 59 (2011) 841–848
Introduction
treatment was given to 56 patients, 38 because of CIS. A regimen without maintenance was used. None of the patients received intravesical
In non–muscle-invasive bladder cancer (NMIBC), patients may progress to muscle-invasive BC (MIBC) despite presumably proper treatment by conservative methods including transurethral resection of the bladder (TURB) and relevant control regimens. These patients have been shown to have a particularly poor prognosis despite radical treatment upon progression [1]. In MIBC, approximately half of the patients with presumably localised disease experience recurrence within 5 yr despite radical treatment. In most patients this leads to cancer-related death. So far, methods to identify these high-risk patients have relied on conventional histopathologic features. These parameters have failed to make the patient selection for earlier and more aggressive treatment sufficient and reliable in a large number of patients [2]. Other prognostic markers are therefore urgently needed to identify these high-risk patients with NMIBC and MIBC who may benefit from earlier radical treatment and/or neoadjuvant or adjuvant therapies. Biomarkers have shown promising results. Thus different molecular profiles have been associated with early progression of NMIBC or poor outcome in patients with MIBC [3–5]. Karyopherin-a2 (KPNA2), a nuclear and cytoplasmatic protein, is a member of the importin family. KPNA2 is thought to play an important role in the nucleocytoplasmatic transport and has also been suggested to be a transporter of tumour suppressors [6–8]. In bladder cancer, KPNA2 was first investigated as a potential classifier of carcinoma in situ (CIS) [9]. In a subsequent multicentre validation study of gene expression classifiers, KPNA2 was identified as involved in a model significantly correlated to poor prognosis [4]. The present study evaluates KPNA2 as a single progression biomarker in a patient cohort with NMIBC undergoing TURB and as a prognostic marker in patients with invasive BC undergoing radical cystectomy (RC). Correlation to the proliferation marker Ki-67 was made to ensure independence of this, and the KPNA2 antibody specificity was validated successfully by Western blotting.
chemotherapy. Patients were followed by control cystoscopies in a routine schedule. Progression was defined as progression to MIBC verified by pathologic examination. Patients undergoing RC before progression to MIBC were not included on the TMA. A single experienced uropathologist reevaluated haematoxylin-eosin stained sections of all tumours to ensure uniform staging and to reclassify grade according to the World Health Organisation 2004 classification [11]. European Organisation for Research and Treatment of Cancer (EORTC) risk score regarding progression was estimated from nomograms when possible [12].
2.3.
Radical cystectomy tissue microarray
The RC TMA consisted of tumour samples from 425 patients undergoing RC between 1992 and 2008 because of invasive urothelial carcinoma. Chemotherapy was administered only at the time of recurrence and not in neoadjuvant or adjuvant settings in the included patients. In two thirds of the patients, lymph node dissection (LND) restricted to the obturator fossa bilaterally (limited LND) was made. In one third of the patients, LND was made by an extended template with cranial limit at the level of the inferior mesenteric artery as previously described [13]. Tumour stage used for analysis was the highest of the pathologically verified TURB stage and pathologic (RC) pT stage. The patients were followed up by a routine schedule including regular radiologic and clinical examinations. Recurrence in each patient was registered as a primary local recurrence, lymph node recurrence, or visceral metastasis. The Danish Central Personal Registry provided the time of death.
2.4.
Immunohistochemistry
Immunohistochemistry (IHC) was performed on formalin-fixed paraffinembedded TMA sections that were deparaffinised and rehydrated in ethanol. After standard demasking, sections were incubated with the primary antibody polyclonal goat anti-KPNA2 (sc-6917, Santa Cruz Biotechnology Inc, Santa Cruz, CA, USA). Two different batches of antiKPNA2 were used for the Ta/T1 and the RC TMAs. Staining of a test TMA identified the optimal antibody dilution, giving the best intensity of each batch (1:100 on the Ta/T1 TMA and 1:250 on the RC TMA). Sections were counterstained with haematoxylin before scoring. By testing different cut-off levels on the Ta/T1 TMA, we found that nuclear staining of 10% of the carcinoma cells (high expression) was optimal. This cut-off level was applied previously in breast cancer [14–16]. Two independent observers scored each core blinded to the clinicopathologic data and clinical outcomes. When there was a discrepancy in the scoring, a
2. 2.1.
Materials and methods Patient cohorts, treatment, and follow-up
Two patient cohorts were used to construct two different tissue
consensus scoring was obtained. Ki-67 IHC staining was made using the MIB-1 antibody (Dako, Glostrup, Denmark). The Ki-67 labelling index was considered to be high when 20% of the carcinoma cells showed nuclear reactivity [17,18].
2.5.
Cloning and plasmid construction
microarrays (TMAs). Both TMAs were constructed as previously described with one core from each tumour [10]. Informed written consent was obtained from all patients, and the Central Denmark Region Committees on Biomedical Research Ethics approved the study.
Wild-type KPNA2 cDNA was polymerase chain reaction amplified from bladder tumour RNA and cloned into the pcDNA 3.1 V5-His plasmid (Invitrogen) using primers sense 5´-UTR-TCTCATCACCATGTCCACCAACGAGAATGC
2.2.
Ta/T1 tissue microarray
The Ta/T1 TMA consisted of 289 primary urothelial tumours (stage Ta
and
antisense
5´-UTR-AAAGTTAAAGGTCCCAGGAGCCC-
CATCC. The DNA sequence was verified by sequencing.
2.6.
Cell culture and transfections
and T1) from patients undergoing primary TURB between 1979 and 2007 as described earlier [10]. Only patients with a minimum of 4 yr of
COS7 cells were cultured in RPMI 1640 medium supplemented with 10%
follow-up without progression were included in the nonprogressing
fetal calf serum (FCS) and 1% penicillin-streptomycin and transfected
group. Routine re-TURB was performed in T1 tumours where no muscle was present in the specimen. Bacillus Calmette-Gue´rin
with plasmid DNA using Lipofectamine (Invitrogen) following the manufacturer’s instructions.
843
EUROPEAN UROLOGY 59 (2011) 841–848
Human urinary bladder transitional cell carcinoma (HT1376) and immortalised human bladder epithelium (HCV29) were cultured in Dulbecco modified Eagle medium supplemented with 10% FCS and 1% penicillin-streptomycin.
2.7.
Western blotting
Total protein samples (25 mg) were run in 12% tris-bis acrylamide gels (Invitrogen) and transferred to nitrocellulose membranes. Membranes
excluded from analysis. Tables 1 and 2 show the clinical and histopathologic characteristics of the included 234 Ta/T1 patients and 377 RC patients. A high Ki-67 labelling index was significantly correlated to high KPNA2 for the Ta/T1 patient cohort ( p < 0.0001) but not for the RC patient cohort ( p = 0.48). The k statistics showed good interobserver agreement of both KPNA2 and Ki-67 scoring (k = 0.79 and 0.80, respectively).
were blocked with 3% wt/vol nonfat powdered milk phosphate-buffered saline. Anti-KPNA2 antibody (Sc-6917; 1:200 dilution) and secondary
3.1.
rabbit antigoat HRP conjugated (P 0449, Dako, Glostrup, Denmark;
bladder cancer
Expression of karyopherin a2 in non–muscle-invasive
1:2000 dilution) were used. The immunoreactive bands were visualised using ECL plus (GE Healthcare, Piscataway, NJ, USA) and a UVP ChemiDoc-It Imaging system (UVP, Upland, CA, USA).
2.8.
Statistical analyses
High expression of KPNA2 was found in 51% of the tissue cores on the Ta/T1 TMA (Fig. 2). High KPNA2 expression was correlated to a significantly higher risk of progression to
Statistical analyses were performed using MedCalc computer software (MedCalc Software, Mariakerke, Belgium). Progression-free survival in the Ta/T1 TMA cohort and survival estimates in the RC TMA cohort were calculated in life table analyses using the Kaplan-Meier method with logrank test for significance and univariate and multivariate Cox regression analysis. Backwards censuring of variables with p > 0.05 was used in multivariate analysis. Harrell’s concordance index was calculated to test the prognostic accuracy of prognostic models with and without KPNA2 [19,20]. Comparison of incidences was assessed using the Fisher exact test. The p values were tested at a 5% significance level.
3.
Results
The KPNA2 antibody specificity was validated successfully by Western blotting of COS7 cells overexpressing KPNA2 and two bladder cancer cell lines HCV29 and HT1376 (Fig. 1). Following KPNA2 staining, 55 cores on the Ta/T1 TMA and 48 cores on the RC TMA were found to have no carcinoma cells on the TMA section. These patients were
[()TD$FIG]
Fig. 1 – Western blot analysis of COS7 cells and human bladder cancer cell lines (HCV29, HT1376). COS7 cells transfected with either a karyopherin-a2 (KPNA2)–V5 construct or an empty construct (Mock) revealed a 55-kDa band corresponding to endogenous KPNA2 and a 60-kDa band corresponding to the KPNA2-V5 protein. The V5-tag corresponds to 5 kDa, which explains the difference between the overexpressed and the endogenous KPNA2.
Table 1 – Clinical and histopathologic characteristics of 234 patients with non–muscle-invasive cancer undergoing transurethral resection of the bladder Median age, yr (range) Male-to-female ratio Median follow-up time for nonprogressing patients, mo (range) Median time to progression, mo (range)
68 (32–86) 4.52 96 (53–232)
– – –
– – –
24 (1–210)
–
–
Low KPNA2
High KPNA2
Stage (%) Ta T1
155 (66) 79 (34)
86 (75) 28 (25)
69 (58) 51 (43)
WHO grading (%) Low grade High grade
155 (66) 79 (34)
86 (75) 28 (25)
69 (58) 51 (43)
Tumour size (%) <3 cm 3 cm Unspecified
142 (61) 65 (28) 27 (12)
71 (62) 31 (27) 12 (11)
71 (59) 34 (28) 15 (13)
Concomitant CIS (%) Yes No Unspecified
72 (31) 159 (68) 3 (1)
28 (25) 84 (74) 2 (2)
44 (37) 75 (63) 1 (1)
EORTC risk of progression score (%) Low 72 (31) Intermediate 52 (22) High 83 (35) Unspecified 27 (12)
48 (42) 35 (31) 29 (25) 2 (2)
24 17 54 25
Adjuvant BCG treatment (%) Yes 44 (19) No 190 (81)
19 (17) 95 (83)
25 (21) 95 (79)
No. of progression events to MIBC (%) All 87 (37) Ta 45 (29) T1 42 (53)
28 (25) 18 (21) 10 (36)
59 (49) 27 (39) 32 (63)
66 (58) 44 (39) 4 (4) 114 (49)
30 (25) 82 (68) 8 (7) 120 (51)
Ki-67 expression (%) Low High Unspecified KPNA2 expression: all patients (%)
96 (41) 126 (54) 12 (5)
(20) (14) (45) (21)
KPNA2 = karyopherin-a2; WHO = World Health Organisation; CIS = carcinoma in situ; EORTC = European Organisation for Research and Treatment of Cancer; BCG = bacillus Calmette-Gue´rin; MIBC = muscleinvasive bladder cancer.
844
EUROPEAN UROLOGY 59 (2011) 841–848
[()TD$FIG]
Table 2 – Clinical and histopathologic characteristics of 377 patients with invasive bladder cancer undergoing radical cystectomy Median age, yr (range) Male-to-female ratio Median follow-up time, mo (range) Median follow-up time for patients alive at the end of follow-up, mo (range)
Tumour stage (%) T1 T2 T3–4 Nodal stage (%) N0 N positive LND (%) Limited Extended Ki-67 expression (%) Low High Unspecified KPNA2 expression: all patients (%)
64 (34–85) 3.77 41 (1–40)
– – –
– – –
70 (23–225)
–
–
Low KPNA2
High KPNA2
41 (11) 112 (30) 224 (59)
12 (10) 40 (33) 68 (57)
29 (11) 72 (28) 156 (61)
286 (76) 91 (24)
90 (75) 30 (25)
196 (76) 61 (24)
256 (68) 121 (32)
77 (64) 43 (36)
179 (70) 78 (30)
75 (20) 252 (67) 50 (13)
20 (17) 80 (67) 20 (17) 120 (32)
55 (21) 172 (67) 30 (12) 257 (68)
KPNA2 = karyopherin-a2; LND = lymph node dissection.
MIBC in univariate analysis ( p < 0.001) and in multivariate analysis when adjusting for significant conventional risk factors (Fig. 3). A high Ki-67 labelling index was correlated to a higher risk of progression in univariate analysis but not in multivariate analysis (Table 3). The concordance index increased slightly but not significantly by applying KPNA2 to a progression model. 3.2.
Expression of karyopherin a2 in radical cystectomy
patients
High expression of KPNA2 was found in 68% of the tissue cores on the RC TMA. Patients with KPNA2-positive tumours
Fig. 2 – High nuclear karyopherin-a2 (KPNA2) expression in (A) non– muscle-invasive bladder cancer and (B) muscle-invasive bladder cancer tumour samples (T40 magnification).
had a significantly poorer prognosis than patients with KPNA2-negative tumours (Fig. 4). KPNA2 expression was independent of conventional risk factors in multivariate analysis (Table 4). Ki-67 expression was not significantly correlated to prognosis in either univariate or multivariate
Table 3 – Univariate and multivariate analyses of association between different risk factors and risk of progression to muscle-invasive bladder cancer in 234 patients with non–muscle-invasive bladder cancer undergoing transurethral resection of the bladder Variable
Progression-free survival Univariate analysis Hazard ratio
T stage (T1 vs Ta) WHO grade (high vs low) CIS (present vs no) BCG treatment (yes vs no) Size of tumour (<3 cm vs 3 cm) Gender (F vs M) Age group (by 5-yr increment) Ki-67 expression (high vs low) KPNA2 expression (high vs low) Concordance index Model without KPNA2 Model with KPNA2
2.59 3.21 1.26 0.54 0.92 1.04 1.18 2.42 2.28
(1.69–3.97) (2.09–4.92) (0.82–1.97) (0.29–1.01) (0.55–1.56) (0.62–1.72) (1.05–1.32) (1.51–3.89) (1.46–3.45)
Multivariate analysis p <0.001 <0.001 0.30 0.05 0.82 0.91 0.006 <0.001 <0.001
Hazard ratio – 3.00 – 0.39 – – 1.18 – 2.59
(1.79–5.01) (0.19–0.83)
(1.02–1.36) (1.49–4.49)
0.67 (0.60–0.72) 0.71 (0.65–0.77)
NS = not significant; WHO = World Health Organisation; CIS = carcinoma in situ; BCG = bacillus Calmette-Gue´rin; KPNA2 = karyopherin-a2.
p NS <0.001 NS 0.02 NS NS 0.03 NS 0.001
EUROPEAN UROLOGY 59 (2011) 841–848
[()TD$FIG]
845
Fig. 3 – Progression-free survival of patients with non–muscle-invasive bladder cancer treated with transurethral resection of the bladder according to karyopherin-a2 (KPNA2) expression in (A) all patients and (B) further stratified for T stage, (C) World Health Organisation 2004 grade, and (D) European Organisation for Research and Treatment of Cancer risk score.
analysis. Thus Ki-67 expression was not included in the prognostic model. Concordance index increased slightly but not significantly by applying KPNA2 to survival models. High KPNA2 expression was a marker of poor prognosis in the patients undergoing extended LND, whereas it was of no significance in patients undergoing limited LND ( p = 0.38) (Fig. 5). Likewise, extended LND was associated with a significantly better prognosis compared with limited LND in patients with KPNA2-negative tumours, whereas it had no significant influence on recurrence-free survival in patients with KPNA2-positive tumours. In patients with KPNA2-positive tumours, risk of visceral metastases was significantly higher than in patients with KPNA2-negative tumours (Table 5). 4.
Discussion
KPNA2 is a transport protein that mediates the nuclear import of NBS1, CKH2, and BRCA1, for example [7,8,21]. These changes in cellular location of key cell cycle regulators and DNA damage response molecules upon alteration of KPNA2 expression may lead to increased malignant transformations. Recent in vitro experiments supported this functional role of KPNA2 by demonstrating
a marked decrease in cell migration capacity and cell viability in lung cancer cells following endogenous KPNA2 knockdown [22]. High KPNA2 expression has previously been suggested as a predictor of poor prognosis in different cancer forms including breast cancer, ovarian and oesophageal carcinoma, and malignant melanoma [16,23–26]. In the present study, KPNA2 expression was an independent prognostic biomarker in NMIBC. Adding KPNA2 expression to conventional clinicopathologic risk factors made it possible to make a better prediction of progression to MIBC. This influence was especially of importance in longterm follow-up. Only patients with a minimum of 4 yr of follow-up without progression were included in the nonprogressing group of NMIBC. This was done to evaluate progression markers in patients with true nonprogressive disease compared with patients with progression after a variable follow-up time. This resulted in a higher risk of progression than other series of NMIBC. Also, the present study registered progression as progression to muscle-invasive disease and not from Ta to T1, which contrasts with the EORTC risk score [12]. This selection of more patients with progression could explain why tumour stage, presence of
846
[()TD$FIG]
EUROPEAN UROLOGY 59 (2011) 841–848
Fig. 4 – Overall survival of patients with invasive bladder cancer undergoing radical cystectomy according to karyopherin-a2 (KPNA2) expression. Log-rank p value is shown.
CIS, and tumour size, for example, were not significant in multivariate analysis. KPNA2 expression was also found to be an independent prognostic marker in patients with invasive BC undergoing RC. The poor prognostic impact of positive KPNA2 expression was independent of the LND template, whereas patients with KPNA2-negative tumours apparently benefitted from an extended LND. There was a significantly higher risk of visceral metastasis in patients with KPNA2-positive tumours. The high risk of distant metastases in patients with KPNA2positive tumours may therefore have minimised the benefits of an extended LND. Theoretically, another explanation could be that KPNA2-positive tumours metastasise more frequently haematogenous than KPNA2-negative tumours. This has been suggested in other cancer forms [25].
Biomarkers involved in the progression process of NMIBC may possibly also be involved in the metastasis process, thereby emphasising the relevance of early radical treatment in high-risk patients with NMIBC. The time span of both patient cohorts included different treatment and control regimens. In the RC cohort, adjustment for different LND templates was made. Extended LND can provide more accurate nodal staging leading to a stronger influence of N stage in this patient cohort [27]. The impact of KPNA2 expression in a prognostic model would possibly have been more evident if more uniform patient material had been available. Concordance index increased only insignificantly by applying KPNA2 expression to prognostic models. This could be a consequence of a poor prognostic correlation of conventional risk factors; thus
[()TD$FIG]
Fig. 5 – Recurrence-free survival of patients with invasive bladder cancer undergoing radical cystectomy. Stratified according to karyopherin-a2 (KPNA2) expression in patients undergoing (A) extended lymph node dissection (LND) and (B) limited LND. Log-rank p values are shown.
847
Table 5 – Recurrence pattern in patients undergoing radical cystectomy stratified according to primary recurrence site within the first 2 yr and KPNA2 expression*
No. of patients No. of patients with recurrence (%) Location Local recurrence (%) Lymph node recurrence (%) Visceral metastasis (%)
(1.38–2.12) (2.25–4.11) (0.33–0.68) (1.08–1.90) (1.01–1.18) (0.97–1.82) (1.03–1.86)
1.47 (1.07–2.03)
0.66 (0.62–0.71) 0.67 (0.61–0.70)
0.001 <0.001 <0.001 NS NS NS 0.02 1.46 (1.15–1.85) 3.30 (2.34–4.67) 0.47 (0.32–0.70)
<0.001 <0.001 <0.001 0.01 0.04 0.08 0.03 1.71 3.04 0.47 1.43 1.09 1.32 1.38
0.68 (0.64–0.74) 0.69 (0.64–0.73)
(1.02–2.26)
<0.001 <0.001 0.003 NS NS NS 0.04 (1.29–2.40) (3.05–6.61) (0.33–0.80)
1.76 4.49 0.52 – – – 1.52 <0.001 <0.001 0.004 0.06 0.68 0.12 0.05 (1.58–2.78) (3.07–6.01) (0.38–0.83) (0.99–1.99) (0.92–1.12) (0.92–2.01) (1.01–2.08) 2.09 4.30 0.56 1.40 1.02 1.36 1.44
NS = not significant; LND = lymph node dissection; KPNA2 = karyopherin-a2.
0.68 (0.63–0.73) 0.69 (0.64–0.74)
(1.17–2.36)
<0.001 <0.001 0.04 NS NS NS 0.005 (1.25–2.16) (3.43–6.77) (0.48–0.97)
(1.62–2.75) (3.50–6.59) (0.55–1.09) (0.99–1.93) (0.95–1.13) (0.85–1.77) (1.02–2.03)
<0.001 <0.001 0.14 0.05 0.45 0.29 0.04
1.64 4.81 0.68 – – – 1.66
Low KPNA2 expression
High KPNA2 expression
p value
377 150 (40)
120 39 (33)
257 97 (38)
0.36
50 (33) 53 (35)
12 (31) 20 (51)
35 (36) 28 (29)
0.69 0.02
68 (45)
12 (31)
49 (50)
0.04
KPNA2 = karyopherin-a2. In some patients, more than one recurrence site was noted because of simultaneous diagnosis. Percentages for location are given from the number of patients with recurrence in each group.
2.11 4.80 0.77 1.39 1.03 1.22 1.44 T stage (T3–4 vs T2 vs T1) N stage (Npos vs N0) LND (extended vs limited) Gender (F vs M) Age group (by 5-yr increment) Ki-67 expression (high vs low) KPNA2 expression (high vs low)
All patients
*
Concordance index Model without KPNA2 Model with KPNA2
p Hazard ratio p Hazard ratio p Hazard ratio p Hazard ratio p Hazard ratio p Hazard ratio
Multivariate analysis Univariate analysis Multivariate analysis Univariate analysis Multivariate analysis Univariate analysis
Overall survival Disease-specific survival Recurrence-free survival Variable
Table 4 – Univariate and multivariate analyses of association between different risk factors and recurrence-free survival, disease-specific survival, and overall survival in 377 patients with invasive bladder cancer undergoing radical cystectomy
EUROPEAN UROLOGY 59 (2011) 841–848
very wide confidence intervals were found in all models. Alternatively, KPNA2 expression alone is not enough, and a combination of molecular markers should be used to increase the prognostic accuracy. Ki-67 has been suggested in previous studies in RC series but was not a significant prognostic maker in the present study [17,18]. Future prospective validation studies should be performed in uniformly treated patients and several prospective prognostic markers evaluated to develop clinically significant nomograms based on traditional and new variables [28]. TMAs can be used to evaluate the presence or absence of a specific biomarker in multiple patients at a large scale expediently. However, because only a minimal amount of tumour tissue from each patient is evaluated, the method is associated with uncertainty of whether the evaluated marker from the TMA is the predominant marker of whole tumour specimens. The supposed monoclonal or oligoclonal origin of urothelial bladder tumours makes this technique somewhat reliable, however [29]. 5.
Conclusions
High KPNA2 expression is associated with a high risk of progression in NMIBC treated with TURB. High KPNA2 expression is also associated with a poor prognosis after RC and a higher risk of organ metastasis. The current study illustrates that new markers should be used together with conventional markers and not replace the latter. Prospective validation studies are needed to estimate whether KPNA2 expression can be used as a supplement to conventional clinicopathologic risk factors in the clinical decision of selecting patients with NMIBC for early RC and also in selecting RC patients suitable for neoadjuvant or adjuvant chemotherapy. Future work should include functional studies to gain insight into the specific function of KPNA2 and to evaluate whether KPNA2 could be a therapeutic target in molecular medicine as part of a multimodality treatment.
848
EUROPEAN UROLOGY 59 (2011) 841–848
Author contributions: Jørgen Bjerggaard Jensen had full access to all the
[10] Egerod FL, Bartels A, Fristrup N, et al. High frequency of tumor cells
data in the study and takes responsibility for the integrity of the data and
with nuclear Egr-1 protein expression in human bladder cancer is
the accuracy of the data analysis.
associated with disease progression. BMC Cancer 2009;9:385. [11] Eble JN, Sauter G, Epstein JI, Sesterhenn IA, editors. WHO classifi-
Study concept and design: Jensen, Munksgaard, Ørntoft, Dyrskjøt. Acquisition of data: Jensen, Munksgaard, Sørenson, Fristrup, BirkenkamDemroder, Ulhøi, Dyrskjøt. Analysis and interpretation of data: Jensen, Munksgaard, Dyrskjøt. Drafting of the manuscript: Jensen, Munksgaard, Dyrskjøt. Critical revision of the manuscript for important intellectual content: Birkenkam-Demroder, Ulhøi, Jensen, Ørntoft. Statistical analysis: Jensen, Munksgaard, Dyrskjøt. Obtaining funding: Jensen, Munksgaard, Ørntoft, Dyrskjøt. Administrative, technical, or material support: None. Supervision: None. Other (specify): None.
cation of tumours: pathology and genetics of tumours of the urinary system and male genital organs. Lyon, France: International Agency for Research on Cancer Press; 2004. [12] Sylvester RJ, van der Meijden APM, Oosterlinck W, et al. Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 2006;49:466–77. [13] Jensen JB, Ulhoi BP, Jensen KM. Lymph node mapping in patients with bladder cancer undergoing radical cystectomy and lymph node dissection to the level of the inferior mesenteric artery. BJU Int 2010;106:199–205. [14] Dahl E, Kristiansen G, Gottlob K, et al. Molecular profiling of laser-
Financial disclosures: I certify that all conflicts of interest, including
microdissected matched tumor and normal breast tissue identifies
specific financial interests and relationships and affiliations relevant
karyopherin alpha2 as a potential novel prognostic marker in
to the subject matter or materials discussed in the manuscript
breast cancer. Clin Cancer Res 2006;12:3950–60.
(eg, employment/affiliation, grants or funding, consultancies, honoraria,
[15] Dankof A, Fritzsche FR, Dahl E, et al. KPNA2 protein expression in
stock ownership or options, expert testimony, royalties, or patents filed,
invasive breast carcinoma and matched peritumoral ductal carci-
received, or pending), are the following: None.
noma in situ. Virchows Arch 2007;451:877–81. [16] Gluz O, Wild P, Meiler R, et al. Nuclear karyopherin alpha2 expression
Funding/Support and role of the sponsor: The study was supported by the
predicts poor survival in patients with advanced breast cancer irre-
John and Birthe Meyer Foundation, the Danish Cancer Society,
spective of treatment intensity. Int J Cancer 2008;123:1433–8.
the Ministry of Technology and Science, the Lundbeck Foundation, the
[17] Margulis V, Shariat SF, Ashfaq R, Sagalowsky AI, Lotan Y. Ki-67 is an
Institute of Clinical Medicine, the University of Aarhus, and the Health
independent predictor of bladder cancer outcome in patients treat-
Research Fund of Central Denmark Region. The research leading up to these results received funding from the European Community´s
ed with radical cystectomy for organ-confined disease. Clin Cancer
Seventh Framework program FP7/2007-2011 under grant agreement
[18] Margulis V, Lotan Y, Karakiewicz PI, et al. Multi-institutional
No. 201663.
Res 2006;12:7369–73. validation of the predictive value of Ki-67 labeling index in patients with urinary bladder cancer. J Natl Cancer Inst 2009;101:114–9. [19] Harrell Jr FE, Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the
References
yield of medical tests. JAMA 1982;247:2543–6. [20] Harrell Jr FE, Lee KL, Mark DB. Multivariable prognostic models:
[1] Knap MM, Lundbeck F, Overgaard J. Prognostic factors, pattern of recurrence and survival in a Danish bladder cancer cohort treated with radical cystectomy. Acta Oncol 2003;42:160–8. [2] van Rhijn BWG, Burger M, Lotan Y, et al. Recurrence and progression of disease in non–muscle-invasive bladder cancer: from epidemiology to treatment strategy. Eur Urol 2009;56:430–42.
issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15:361–87. [21] Thakur S, Zhang HB, Peng Y, et al. Localization of BRCA1 and a splice variant identifies the nuclear localization signal. Mol Cell Biol 1997; 17:444–52. [22] Wang CI, Wang CL, Wang CW, et al. Importin subunit alpha-2 is
[3] Sanchez-Carbayo M, Socci ND, Lozano J, Saint F, Cordon-Cardo C.
identified as a potential biomarker for non-small cell lung cancer by
Defining molecular profiles of poor outcome in patients with
integration of the cancer cell secretome and tissue transcriptome.
invasive bladder cancer using oligonucleotide microarrays. J Clin Oncol 2006;24:778–89. [4] Dyrskjot L, Zieger K, Real FX, et al. Gene expression signatures predict outcome in non–muscle-invasive bladder carcinoma: a multicenter validation study. Clin Cancer Res 2007;13:3545–51. [5] Theodorescu D, Sapinoso LM, Conaway MR, Oxford G, Hampton GM, Frierson Jr HF. Reduced expression of metastasis suppressor RhoGDI2 is associated with decreased survival for patients with bladder cancer. Clin Cancer Res 2004;10:3800–6. [6] Tseng SF, Chang CY, Wu KJ, Teng SC. Importin KPNA2 is required for proper nuclear localization and multiple functions of NBS1. J Biol Chem 2005;280:39594–600. [7] Zannini L, Lecis D, Lisanti S, et al. Karyopherin-alpha2 protein interacts with Chk2 and contributes to its nuclear import. J Biol Chem 2003;278:42346–51. [8] Teng SC, Wu KJ, Tseng SF, Wong CW, Kao L. Importin KPNA2, NBS1, DNA repair and tumorigenesis. J Mol Histol 2006;37:293–9.
Int J Cancer. In press. doi:10.1002/ijc.25568. [23] Wang Y, Klijn JG, Zhang Y, et al. Gene-expression profiles to predict distant metastasis of lymph-node-negative primary breast cancer. Lancet 2005;365:671–9. [24] Zheng M, Tang L, Huang L, et al. Overexpression of karyopherin-2 in epithelial ovarian cancer and correlation with poor prognosis. Obstet Gynecol 2010;116:884–91. [25] Sakai M, Sohda M, Miyazaki T, et al. Significance of karyopherin{alpha} 2 (KPNA2) expression in esophageal squamous cell carcinoma. Anticancer Res 2010;30:851–6. [26] Winnepenninckx V, Lazar V, Michiels S, et al. Gene expression profiling of primary cutaneous melanoma and clinical outcome. J Natl Cancer Inst 2006;98:472–82. [27] Dangle PP, Gong MC, Bahnson RR, Pohar KS. How do commonly performed lymphadenectomy templates influence bladder cancer nodal stage? J Urol 2010;183:499–503. [28] Shariat SF, Lotan Y, Vickers A, et al. Statistical consideration for clinical
[9] Dyrskjot L, Kruhoffer M, Thykjaer T, et al. Gene expression in the
biomarker research in bladder cancer. Urol Oncol 2010;28:389–400.
urinary bladder: a common carcinoma in situ gene expression
[29] Hafner C, Knuechel R, Stoehr R, Hartmann A. Clonality of multifocal
signature exists disregarding histopathological classification. Can-
urothelial carcinomas: 10 years of molecular genetic studies. Int J
cer Res 2004;64:4040–8.
Cancer 2002;101:1–6.
available at www.sciencedirect.com journal homepage: www.europeanurology.com
Bladder Cancer
High Expression of Karyopherin-a2 Defines Poor Prognosis in Non–Muscle-Invasive Bladder Cancer and in Patients with Invasive Bladder Cancer Undergoing Radical Cystectomy Jørgen Bjerggaard Jensen a,*, Pia Pinholt Munksgaard b, Christoffer Mørk Sørensen b, Niels Fristrup b, Karin Birkenkamp-Demtroder b, Benedicte Parm Ulhøi c, Klaus Møller-Ernst Jensen a, Torben F. Ørntoft b, Lars Dyrskjøt b a
Department of Urology, Aarhus University Hospital, Skejby, Denmark
b
Department of Molecular Medicine, Aarhus University Hospital, Skejby, Denmark
c
Institute of Pathology, Aarhus University Hospital, Aarhus Sygehus NBG, Denmark
Article info
Abstract
Article history: Accepted January 27, 2011 Published online ahead of print on February 9, 2011
Background: Conventional clinicopathologic risk factors have failed to accurately predict the prognosis of patients with bladder cancer (BC). Objective: To evaluate karyopherin-a2 (KPNA2) expression as a progression marker in patients with non–muscle-invasive BC (NMIBC) treated by conservative methods and as a prognostic marker in patients with invasive BC undergoing radical cystectomy (RC). Design, setting, and participants: Two different tissue microarrays were constructed, one with 234 primary Ta/T1 tumours from patients treated by transurethral resection of the bladder and one with 377 tumours from RC patients. Intervention: KPNA2 expression based on immunohistochemistry. Measurements: Risk of progression of Ta/T1 patients to muscle-invasive BC was estimated in clinical follow-up to progression or a minimum of 53 mo. Risk of recurrent disease and death following RC was estimated in clinical follow-up of a minimum of 24 mo in patients alive. Results and limitations: A high KPNA2 expression in Ta/T1 patients was significantly correlated with a higher risk of progression that was independent of conventional risk factors in multivariate analysis. In patients undergoing RC, a high KPNA2 expression was an independent predictor of poor prognosis. A high KPNA2 expression was correlated with a higher risk of visceral metastasis rather than lymphatic spread. Conclusions: KPNA2 expression is a marker for progression of NMIBC and a prognostic marker in patients undergoing RC.
Keywords: Urothelial carcinoma Bladder cancer KPNA2 Biomarker Prognostic marker Radical cystectomy TMA Tissue microarray
# 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Urology, Aarhus University Hospital, Skejby, Brendstrupgaardsvej 100, DK-8200 Aarhus N, Denmark. Tel. +45 89495950; Fax: +45 89496006. E-mail address: [email protected] (J.B. Jensen).
0302-2838/$ – see back matter # 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.eururo.2011.01.048
842
1.
EUROPEAN UROLOGY 59 (2011) 841–848
Introduction
treatment was given to 56 patients, 38 because of CIS. A regimen without maintenance was used. None of the patients received intravesical
In non–muscle-invasive bladder cancer (NMIBC), patients may progress to muscle-invasive BC (MIBC) despite presumably proper treatment by conservative methods including transurethral resection of the bladder (TURB) and relevant control regimens. These patients have been shown to have a particularly poor prognosis despite radical treatment upon progression [1]. In MIBC, approximately half of the patients with presumably localised disease experience recurrence within 5 yr despite radical treatment. In most patients this leads to cancer-related death. So far, methods to identify these high-risk patients have relied on conventional histopathologic features. These parameters have failed to make the patient selection for earlier and more aggressive treatment sufficient and reliable in a large number of patients [2]. Other prognostic markers are therefore urgently needed to identify these high-risk patients with NMIBC and MIBC who may benefit from earlier radical treatment and/or neoadjuvant or adjuvant therapies. Biomarkers have shown promising results. Thus different molecular profiles have been associated with early progression of NMIBC or poor outcome in patients with MIBC [3–5]. Karyopherin-a2 (KPNA2), a nuclear and cytoplasmatic protein, is a member of the importin family. KPNA2 is thought to play an important role in the nucleocytoplasmatic transport and has also been suggested to be a transporter of tumour suppressors [6–8]. In bladder cancer, KPNA2 was first investigated as a potential classifier of carcinoma in situ (CIS) [9]. In a subsequent multicentre validation study of gene expression classifiers, KPNA2 was identified as involved in a model significantly correlated to poor prognosis [4]. The present study evaluates KPNA2 as a single progression biomarker in a patient cohort with NMIBC undergoing TURB and as a prognostic marker in patients with invasive BC undergoing radical cystectomy (RC). Correlation to the proliferation marker Ki-67 was made to ensure independence of this, and the KPNA2 antibody specificity was validated successfully by Western blotting.
chemotherapy. Patients were followed by control cystoscopies in a routine schedule. Progression was defined as progression to MIBC verified by pathologic examination. Patients undergoing RC before progression to MIBC were not included on the TMA. A single experienced uropathologist reevaluated haematoxylin-eosin stained sections of all tumours to ensure uniform staging and to reclassify grade according to the World Health Organisation 2004 classification [11]. European Organisation for Research and Treatment of Cancer (EORTC) risk score regarding progression was estimated from nomograms when possible [12].
2.3.
Radical cystectomy tissue microarray
The RC TMA consisted of tumour samples from 425 patients undergoing RC between 1992 and 2008 because of invasive urothelial carcinoma. Chemotherapy was administered only at the time of recurrence and not in neoadjuvant or adjuvant settings in the included patients. In two thirds of the patients, lymph node dissection (LND) restricted to the obturator fossa bilaterally (limited LND) was made. In one third of the patients, LND was made by an extended template with cranial limit at the level of the inferior mesenteric artery as previously described [13]. Tumour stage used for analysis was the highest of the pathologically verified TURB stage and pathologic (RC) pT stage. The patients were followed up by a routine schedule including regular radiologic and clinical examinations. Recurrence in each patient was registered as a primary local recurrence, lymph node recurrence, or visceral metastasis. The Danish Central Personal Registry provided the time of death.
2.4.
Immunohistochemistry
Immunohistochemistry (IHC) was performed on formalin-fixed paraffinembedded TMA sections that were deparaffinised and rehydrated in ethanol. After standard demasking, sections were incubated with the primary antibody polyclonal goat anti-KPNA2 (sc-6917, Santa Cruz Biotechnology Inc, Santa Cruz, CA, USA). Two different batches of antiKPNA2 were used for the Ta/T1 and the RC TMAs. Staining of a test TMA identified the optimal antibody dilution, giving the best intensity of each batch (1:100 on the Ta/T1 TMA and 1:250 on the RC TMA). Sections were counterstained with haematoxylin before scoring. By testing different cut-off levels on the Ta/T1 TMA, we found that nuclear staining of 10% of the carcinoma cells (high expression) was optimal. This cut-off level was applied previously in breast cancer [14–16]. Two independent observers scored each core blinded to the clinicopathologic data and clinical outcomes. When there was a discrepancy in the scoring, a
2. 2.1.
Materials and methods Patient cohorts, treatment, and follow-up
Two patient cohorts were used to construct two different tissue
consensus scoring was obtained. Ki-67 IHC staining was made using the MIB-1 antibody (Dako, Glostrup, Denmark). The Ki-67 labelling index was considered to be high when 20% of the carcinoma cells showed nuclear reactivity [17,18].
2.5.
Cloning and plasmid construction
microarrays (TMAs). Both TMAs were constructed as previously described with one core from each tumour [10]. Informed written consent was obtained from all patients, and the Central Denmark Region Committees on Biomedical Research Ethics approved the study.
Wild-type KPNA2 cDNA was polymerase chain reaction amplified from bladder tumour RNA and cloned into the pcDNA 3.1 V5-His plasmid (Invitrogen) using primers sense 5´-UTR-TCTCATCACCATGTCCACCAACGAGAATGC
2.2.
Ta/T1 tissue microarray
The Ta/T1 TMA consisted of 289 primary urothelial tumours (stage Ta
and
antisense
5´-UTR-AAAGTTAAAGGTCCCAGGAGCCC-
CATCC. The DNA sequence was verified by sequencing.
2.6.
Cell culture and transfections
and T1) from patients undergoing primary TURB between 1979 and 2007 as described earlier [10]. Only patients with a minimum of 4 yr of
COS7 cells were cultured in RPMI 1640 medium supplemented with 10%
follow-up without progression were included in the nonprogressing
fetal calf serum (FCS) and 1% penicillin-streptomycin and transfected
group. Routine re-TURB was performed in T1 tumours where no muscle was present in the specimen. Bacillus Calmette-Gue´rin
with plasmid DNA using Lipofectamine (Invitrogen) following the manufacturer’s instructions.
843
EUROPEAN UROLOGY 59 (2011) 841–848
Human urinary bladder transitional cell carcinoma (HT1376) and immortalised human bladder epithelium (HCV29) were cultured in Dulbecco modified Eagle medium supplemented with 10% FCS and 1% penicillin-streptomycin.
2.7.
Western blotting
Total protein samples (25 mg) were run in 12% tris-bis acrylamide gels (Invitrogen) and transferred to nitrocellulose membranes. Membranes
excluded from analysis. Tables 1 and 2 show the clinical and histopathologic characteristics of the included 234 Ta/T1 patients and 377 RC patients. A high Ki-67 labelling index was significantly correlated to high KPNA2 for the Ta/T1 patient cohort ( p < 0.0001) but not for the RC patient cohort ( p = 0.48). The k statistics showed good interobserver agreement of both KPNA2 and Ki-67 scoring (k = 0.79 and 0.80, respectively).
were blocked with 3% wt/vol nonfat powdered milk phosphate-buffered saline. Anti-KPNA2 antibody (Sc-6917; 1:200 dilution) and secondary
3.1.
rabbit antigoat HRP conjugated (P 0449, Dako, Glostrup, Denmark;
bladder cancer
Expression of karyopherin a2 in non–muscle-invasive
1:2000 dilution) were used. The immunoreactive bands were visualised using ECL plus (GE Healthcare, Piscataway, NJ, USA) and a UVP ChemiDoc-It Imaging system (UVP, Upland, CA, USA).
2.8.
Statistical analyses
High expression of KPNA2 was found in 51% of the tissue cores on the Ta/T1 TMA (Fig. 2). High KPNA2 expression was correlated to a significantly higher risk of progression to
Statistical analyses were performed using MedCalc computer software (MedCalc Software, Mariakerke, Belgium). Progression-free survival in the Ta/T1 TMA cohort and survival estimates in the RC TMA cohort were calculated in life table analyses using the Kaplan-Meier method with logrank test for significance and univariate and multivariate Cox regression analysis. Backwards censuring of variables with p > 0.05 was used in multivariate analysis. Harrell’s concordance index was calculated to test the prognostic accuracy of prognostic models with and without KPNA2 [19,20]. Comparison of incidences was assessed using the Fisher exact test. The p values were tested at a 5% significance level.
3.
Results
The KPNA2 antibody specificity was validated successfully by Western blotting of COS7 cells overexpressing KPNA2 and two bladder cancer cell lines HCV29 and HT1376 (Fig. 1). Following KPNA2 staining, 55 cores on the Ta/T1 TMA and 48 cores on the RC TMA were found to have no carcinoma cells on the TMA section. These patients were
[()TD$FIG]
Fig. 1 – Western blot analysis of COS7 cells and human bladder cancer cell lines (HCV29, HT1376). COS7 cells transfected with either a karyopherin-a2 (KPNA2)–V5 construct or an empty construct (Mock) revealed a 55-kDa band corresponding to endogenous KPNA2 and a 60-kDa band corresponding to the KPNA2-V5 protein. The V5-tag corresponds to 5 kDa, which explains the difference between the overexpressed and the endogenous KPNA2.
Table 1 – Clinical and histopathologic characteristics of 234 patients with non–muscle-invasive cancer undergoing transurethral resection of the bladder Median age, yr (range) Male-to-female ratio Median follow-up time for nonprogressing patients, mo (range) Median time to progression, mo (range)
68 (32–86) 4.52 96 (53–232)
– – –
– – –
24 (1–210)
–
–
Low KPNA2
High KPNA2
Stage (%) Ta T1
155 (66) 79 (34)
86 (75) 28 (25)
69 (58) 51 (43)
WHO grading (%) Low grade High grade
155 (66) 79 (34)
86 (75) 28 (25)
69 (58) 51 (43)
Tumour size (%) <3 cm 3 cm Unspecified
142 (61) 65 (28) 27 (12)
71 (62) 31 (27) 12 (11)
71 (59) 34 (28) 15 (13)
Concomitant CIS (%) Yes No Unspecified
72 (31) 159 (68) 3 (1)
28 (25) 84 (74) 2 (2)
44 (37) 75 (63) 1 (1)
EORTC risk of progression score (%) Low 72 (31) Intermediate 52 (22) High 83 (35) Unspecified 27 (12)
48 (42) 35 (31) 29 (25) 2 (2)
24 17 54 25
Adjuvant BCG treatment (%) Yes 44 (19) No 190 (81)
19 (17) 95 (83)
25 (21) 95 (79)
No. of progression events to MIBC (%) All 87 (37) Ta 45 (29) T1 42 (53)
28 (25) 18 (21) 10 (36)
59 (49) 27 (39) 32 (63)
66 (58) 44 (39) 4 (4) 114 (49)
30 (25) 82 (68) 8 (7) 120 (51)
Ki-67 expression (%) Low High Unspecified KPNA2 expression: all patients (%)
96 (41) 126 (54) 12 (5)
(20) (14) (45) (21)
KPNA2 = karyopherin-a2; WHO = World Health Organisation; CIS = carcinoma in situ; EORTC = European Organisation for Research and Treatment of Cancer; BCG = bacillus Calmette-Gue´rin; MIBC = muscleinvasive bladder cancer.
844
EUROPEAN UROLOGY 59 (2011) 841–848
[()TD$FIG]
Table 2 – Clinical and histopathologic characteristics of 377 patients with invasive bladder cancer undergoing radical cystectomy Median age, yr (range) Male-to-female ratio Median follow-up time, mo (range) Median follow-up time for patients alive at the end of follow-up, mo (range)
Tumour stage (%) T1 T2 T3–4 Nodal stage (%) N0 N positive LND (%) Limited Extended Ki-67 expression (%) Low High Unspecified KPNA2 expression: all patients (%)
64 (34–85) 3.77 41 (1–40)
– – –
– – –
70 (23–225)
–
–
Low KPNA2
High KPNA2
41 (11) 112 (30) 224 (59)
12 (10) 40 (33) 68 (57)
29 (11) 72 (28) 156 (61)
286 (76) 91 (24)
90 (75) 30 (25)
196 (76) 61 (24)
256 (68) 121 (32)
77 (64) 43 (36)
179 (70) 78 (30)
75 (20) 252 (67) 50 (13)
20 (17) 80 (67) 20 (17) 120 (32)
55 (21) 172 (67) 30 (12) 257 (68)
KPNA2 = karyopherin-a2; LND = lymph node dissection.
MIBC in univariate analysis ( p < 0.001) and in multivariate analysis when adjusting for significant conventional risk factors (Fig. 3). A high Ki-67 labelling index was correlated to a higher risk of progression in univariate analysis but not in multivariate analysis (Table 3). The concordance index increased slightly but not significantly by applying KPNA2 to a progression model. 3.2.
Expression of karyopherin a2 in radical cystectomy
patients
High expression of KPNA2 was found in 68% of the tissue cores on the RC TMA. Patients with KPNA2-positive tumours
Fig. 2 – High nuclear karyopherin-a2 (KPNA2) expression in (A) non– muscle-invasive bladder cancer and (B) muscle-invasive bladder cancer tumour samples (T40 magnification).
had a significantly poorer prognosis than patients with KPNA2-negative tumours (Fig. 4). KPNA2 expression was independent of conventional risk factors in multivariate analysis (Table 4). Ki-67 expression was not significantly correlated to prognosis in either univariate or multivariate
Table 3 – Univariate and multivariate analyses of association between different risk factors and risk of progression to muscle-invasive bladder cancer in 234 patients with non–muscle-invasive bladder cancer undergoing transurethral resection of the bladder Variable
Progression-free survival Univariate analysis Hazard ratio
T stage (T1 vs Ta) WHO grade (high vs low) CIS (present vs no) BCG treatment (yes vs no) Size of tumour (<3 cm vs 3 cm) Gender (F vs M) Age group (by 5-yr increment) Ki-67 expression (high vs low) KPNA2 expression (high vs low) Concordance index Model without KPNA2 Model with KPNA2
2.59 3.21 1.26 0.54 0.92 1.04 1.18 2.42 2.28
(1.69–3.97) (2.09–4.92) (0.82–1.97) (0.29–1.01) (0.55–1.56) (0.62–1.72) (1.05–1.32) (1.51–3.89) (1.46–3.45)
Multivariate analysis p <0.001 <0.001 0.30 0.05 0.82 0.91 0.006 <0.001 <0.001
Hazard ratio – 3.00 – 0.39 – – 1.18 – 2.59
(1.79–5.01) (0.19–0.83)
(1.02–1.36) (1.49–4.49)
0.67 (0.60–0.72) 0.71 (0.65–0.77)
NS = not significant; WHO = World Health Organisation; CIS = carcinoma in situ; BCG = bacillus Calmette-Gue´rin; KPNA2 = karyopherin-a2.
p NS <0.001 NS 0.02 NS NS 0.03 NS 0.001
EUROPEAN UROLOGY 59 (2011) 841–848
[()TD$FIG]
845
Fig. 3 – Progression-free survival of patients with non–muscle-invasive bladder cancer treated with transurethral resection of the bladder according to karyopherin-a2 (KPNA2) expression in (A) all patients and (B) further stratified for T stage, (C) World Health Organisation 2004 grade, and (D) European Organisation for Research and Treatment of Cancer risk score.
analysis. Thus Ki-67 expression was not included in the prognostic model. Concordance index increased slightly but not significantly by applying KPNA2 to survival models. High KPNA2 expression was a marker of poor prognosis in the patients undergoing extended LND, whereas it was of no significance in patients undergoing limited LND ( p = 0.38) (Fig. 5). Likewise, extended LND was associated with a significantly better prognosis compared with limited LND in patients with KPNA2-negative tumours, whereas it had no significant influence on recurrence-free survival in patients with KPNA2-positive tumours. In patients with KPNA2-positive tumours, risk of visceral metastases was significantly higher than in patients with KPNA2-negative tumours (Table 5). 4.
Discussion
KPNA2 is a transport protein that mediates the nuclear import of NBS1, CKH2, and BRCA1, for example [7,8,21]. These changes in cellular location of key cell cycle regulators and DNA damage response molecules upon alteration of KPNA2 expression may lead to increased malignant transformations. Recent in vitro experiments supported this functional role of KPNA2 by demonstrating
a marked decrease in cell migration capacity and cell viability in lung cancer cells following endogenous KPNA2 knockdown [22]. High KPNA2 expression has previously been suggested as a predictor of poor prognosis in different cancer forms including breast cancer, ovarian and oesophageal carcinoma, and malignant melanoma [16,23–26]. In the present study, KPNA2 expression was an independent prognostic biomarker in NMIBC. Adding KPNA2 expression to conventional clinicopathologic risk factors made it possible to make a better prediction of progression to MIBC. This influence was especially of importance in longterm follow-up. Only patients with a minimum of 4 yr of follow-up without progression were included in the nonprogressing group of NMIBC. This was done to evaluate progression markers in patients with true nonprogressive disease compared with patients with progression after a variable follow-up time. This resulted in a higher risk of progression than other series of NMIBC. Also, the present study registered progression as progression to muscle-invasive disease and not from Ta to T1, which contrasts with the EORTC risk score [12]. This selection of more patients with progression could explain why tumour stage, presence of
846
[()TD$FIG]
EUROPEAN UROLOGY 59 (2011) 841–848
Fig. 4 – Overall survival of patients with invasive bladder cancer undergoing radical cystectomy according to karyopherin-a2 (KPNA2) expression. Log-rank p value is shown.
CIS, and tumour size, for example, were not significant in multivariate analysis. KPNA2 expression was also found to be an independent prognostic marker in patients with invasive BC undergoing RC. The poor prognostic impact of positive KPNA2 expression was independent of the LND template, whereas patients with KPNA2-negative tumours apparently benefitted from an extended LND. There was a significantly higher risk of visceral metastasis in patients with KPNA2-positive tumours. The high risk of distant metastases in patients with KPNA2positive tumours may therefore have minimised the benefits of an extended LND. Theoretically, another explanation could be that KPNA2-positive tumours metastasise more frequently haematogenous than KPNA2-negative tumours. This has been suggested in other cancer forms [25].
Biomarkers involved in the progression process of NMIBC may possibly also be involved in the metastasis process, thereby emphasising the relevance of early radical treatment in high-risk patients with NMIBC. The time span of both patient cohorts included different treatment and control regimens. In the RC cohort, adjustment for different LND templates was made. Extended LND can provide more accurate nodal staging leading to a stronger influence of N stage in this patient cohort [27]. The impact of KPNA2 expression in a prognostic model would possibly have been more evident if more uniform patient material had been available. Concordance index increased only insignificantly by applying KPNA2 expression to prognostic models. This could be a consequence of a poor prognostic correlation of conventional risk factors; thus
[()TD$FIG]
Fig. 5 – Recurrence-free survival of patients with invasive bladder cancer undergoing radical cystectomy. Stratified according to karyopherin-a2 (KPNA2) expression in patients undergoing (A) extended lymph node dissection (LND) and (B) limited LND. Log-rank p values are shown.
847
Table 5 – Recurrence pattern in patients undergoing radical cystectomy stratified according to primary recurrence site within the first 2 yr and KPNA2 expression*
No. of patients No. of patients with recurrence (%) Location Local recurrence (%) Lymph node recurrence (%) Visceral metastasis (%)
(1.38–2.12) (2.25–4.11) (0.33–0.68) (1.08–1.90) (1.01–1.18) (0.97–1.82) (1.03–1.86)
1.47 (1.07–2.03)
0.66 (0.62–0.71) 0.67 (0.61–0.70)
0.001 <0.001 <0.001 NS NS NS 0.02 1.46 (1.15–1.85) 3.30 (2.34–4.67) 0.47 (0.32–0.70)
<0.001 <0.001 <0.001 0.01 0.04 0.08 0.03 1.71 3.04 0.47 1.43 1.09 1.32 1.38
0.68 (0.64–0.74) 0.69 (0.64–0.73)
(1.02–2.26)
<0.001 <0.001 0.003 NS NS NS 0.04 (1.29–2.40) (3.05–6.61) (0.33–0.80)
1.76 4.49 0.52 – – – 1.52 <0.001 <0.001 0.004 0.06 0.68 0.12 0.05 (1.58–2.78) (3.07–6.01) (0.38–0.83) (0.99–1.99) (0.92–1.12) (0.92–2.01) (1.01–2.08) 2.09 4.30 0.56 1.40 1.02 1.36 1.44
NS = not significant; LND = lymph node dissection; KPNA2 = karyopherin-a2.
0.68 (0.63–0.73) 0.69 (0.64–0.74)
(1.17–2.36)
<0.001 <0.001 0.04 NS NS NS 0.005 (1.25–2.16) (3.43–6.77) (0.48–0.97)
(1.62–2.75) (3.50–6.59) (0.55–1.09) (0.99–1.93) (0.95–1.13) (0.85–1.77) (1.02–2.03)
<0.001 <0.001 0.14 0.05 0.45 0.29 0.04
1.64 4.81 0.68 – – – 1.66
Low KPNA2 expression
High KPNA2 expression
p value
377 150 (40)
120 39 (33)
257 97 (38)
0.36
50 (33) 53 (35)
12 (31) 20 (51)
35 (36) 28 (29)
0.69 0.02
68 (45)
12 (31)
49 (50)
0.04
KPNA2 = karyopherin-a2. In some patients, more than one recurrence site was noted because of simultaneous diagnosis. Percentages for location are given from the number of patients with recurrence in each group.
2.11 4.80 0.77 1.39 1.03 1.22 1.44 T stage (T3–4 vs T2 vs T1) N stage (Npos vs N0) LND (extended vs limited) Gender (F vs M) Age group (by 5-yr increment) Ki-67 expression (high vs low) KPNA2 expression (high vs low)
All patients
*
Concordance index Model without KPNA2 Model with KPNA2
p Hazard ratio p Hazard ratio p Hazard ratio p Hazard ratio p Hazard ratio p Hazard ratio
Multivariate analysis Univariate analysis Multivariate analysis Univariate analysis Multivariate analysis Univariate analysis
Overall survival Disease-specific survival Recurrence-free survival Variable
Table 4 – Univariate and multivariate analyses of association between different risk factors and recurrence-free survival, disease-specific survival, and overall survival in 377 patients with invasive bladder cancer undergoing radical cystectomy
EUROPEAN UROLOGY 59 (2011) 841–848
very wide confidence intervals were found in all models. Alternatively, KPNA2 expression alone is not enough, and a combination of molecular markers should be used to increase the prognostic accuracy. Ki-67 has been suggested in previous studies in RC series but was not a significant prognostic maker in the present study [17,18]. Future prospective validation studies should be performed in uniformly treated patients and several prospective prognostic markers evaluated to develop clinically significant nomograms based on traditional and new variables [28]. TMAs can be used to evaluate the presence or absence of a specific biomarker in multiple patients at a large scale expediently. However, because only a minimal amount of tumour tissue from each patient is evaluated, the method is associated with uncertainty of whether the evaluated marker from the TMA is the predominant marker of whole tumour specimens. The supposed monoclonal or oligoclonal origin of urothelial bladder tumours makes this technique somewhat reliable, however [29]. 5.
Conclusions
High KPNA2 expression is associated with a high risk of progression in NMIBC treated with TURB. High KPNA2 expression is also associated with a poor prognosis after RC and a higher risk of organ metastasis. The current study illustrates that new markers should be used together with conventional markers and not replace the latter. Prospective validation studies are needed to estimate whether KPNA2 expression can be used as a supplement to conventional clinicopathologic risk factors in the clinical decision of selecting patients with NMIBC for early RC and also in selecting RC patients suitable for neoadjuvant or adjuvant chemotherapy. Future work should include functional studies to gain insight into the specific function of KPNA2 and to evaluate whether KPNA2 could be a therapeutic target in molecular medicine as part of a multimodality treatment.
848
EUROPEAN UROLOGY 59 (2011) 841–848
Author contributions: Jørgen Bjerggaard Jensen had full access to all the
[10] Egerod FL, Bartels A, Fristrup N, et al. High frequency of tumor cells
data in the study and takes responsibility for the integrity of the data and
with nuclear Egr-1 protein expression in human bladder cancer is
the accuracy of the data analysis.
associated with disease progression. BMC Cancer 2009;9:385. [11] Eble JN, Sauter G, Epstein JI, Sesterhenn IA, editors. WHO classifi-
Study concept and design: Jensen, Munksgaard, Ørntoft, Dyrskjøt. Acquisition of data: Jensen, Munksgaard, Sørenson, Fristrup, BirkenkamDemroder, Ulhøi, Dyrskjøt. Analysis and interpretation of data: Jensen, Munksgaard, Dyrskjøt. Drafting of the manuscript: Jensen, Munksgaard, Dyrskjøt. Critical revision of the manuscript for important intellectual content: Birkenkam-Demroder, Ulhøi, Jensen, Ørntoft. Statistical analysis: Jensen, Munksgaard, Dyrskjøt. Obtaining funding: Jensen, Munksgaard, Ørntoft, Dyrskjøt. Administrative, technical, or material support: None. Supervision: None. Other (specify): None.
cation of tumours: pathology and genetics of tumours of the urinary system and male genital organs. Lyon, France: International Agency for Research on Cancer Press; 2004. [12] Sylvester RJ, van der Meijden APM, Oosterlinck W, et al. Predicting recurrence and progression in individual patients with stage Ta T1 bladder cancer using EORTC risk tables: a combined analysis of 2596 patients from seven EORTC trials. Eur Urol 2006;49:466–77. [13] Jensen JB, Ulhoi BP, Jensen KM. Lymph node mapping in patients with bladder cancer undergoing radical cystectomy and lymph node dissection to the level of the inferior mesenteric artery. BJU Int 2010;106:199–205. [14] Dahl E, Kristiansen G, Gottlob K, et al. Molecular profiling of laser-
Financial disclosures: I certify that all conflicts of interest, including
microdissected matched tumor and normal breast tissue identifies
specific financial interests and relationships and affiliations relevant
karyopherin alpha2 as a potential novel prognostic marker in
to the subject matter or materials discussed in the manuscript
breast cancer. Clin Cancer Res 2006;12:3950–60.
(eg, employment/affiliation, grants or funding, consultancies, honoraria,
[15] Dankof A, Fritzsche FR, Dahl E, et al. KPNA2 protein expression in
stock ownership or options, expert testimony, royalties, or patents filed,
invasive breast carcinoma and matched peritumoral ductal carci-
received, or pending), are the following: None.
noma in situ. Virchows Arch 2007;451:877–81. [16] Gluz O, Wild P, Meiler R, et al. Nuclear karyopherin alpha2 expression
Funding/Support and role of the sponsor: The study was supported by the
predicts poor survival in patients with advanced breast cancer irre-
John and Birthe Meyer Foundation, the Danish Cancer Society,
spective of treatment intensity. Int J Cancer 2008;123:1433–8.
the Ministry of Technology and Science, the Lundbeck Foundation, the
[17] Margulis V, Shariat SF, Ashfaq R, Sagalowsky AI, Lotan Y. Ki-67 is an
Institute of Clinical Medicine, the University of Aarhus, and the Health
independent predictor of bladder cancer outcome in patients treat-
Research Fund of Central Denmark Region. The research leading up to these results received funding from the European Community´s
ed with radical cystectomy for organ-confined disease. Clin Cancer
Seventh Framework program FP7/2007-2011 under grant agreement
[18] Margulis V, Lotan Y, Karakiewicz PI, et al. Multi-institutional
No. 201663.
Res 2006;12:7369–73. validation of the predictive value of Ki-67 labeling index in patients with urinary bladder cancer. J Natl Cancer Inst 2009;101:114–9. [19] Harrell Jr FE, Califf RM, Pryor DB, Lee KL, Rosati RA. Evaluating the
References
yield of medical tests. JAMA 1982;247:2543–6. [20] Harrell Jr FE, Lee KL, Mark DB. Multivariable prognostic models:
[1] Knap MM, Lundbeck F, Overgaard J. Prognostic factors, pattern of recurrence and survival in a Danish bladder cancer cohort treated with radical cystectomy. Acta Oncol 2003;42:160–8. [2] van Rhijn BWG, Burger M, Lotan Y, et al. Recurrence and progression of disease in non–muscle-invasive bladder cancer: from epidemiology to treatment strategy. Eur Urol 2009;56:430–42.
issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15:361–87. [21] Thakur S, Zhang HB, Peng Y, et al. Localization of BRCA1 and a splice variant identifies the nuclear localization signal. Mol Cell Biol 1997; 17:444–52. [22] Wang CI, Wang CL, Wang CW, et al. Importin subunit alpha-2 is
[3] Sanchez-Carbayo M, Socci ND, Lozano J, Saint F, Cordon-Cardo C.
identified as a potential biomarker for non-small cell lung cancer by
Defining molecular profiles of poor outcome in patients with
integration of the cancer cell secretome and tissue transcriptome.
invasive bladder cancer using oligonucleotide microarrays. J Clin Oncol 2006;24:778–89. [4] Dyrskjot L, Zieger K, Real FX, et al. Gene expression signatures predict outcome in non–muscle-invasive bladder carcinoma: a multicenter validation study. Clin Cancer Res 2007;13:3545–51. [5] Theodorescu D, Sapinoso LM, Conaway MR, Oxford G, Hampton GM, Frierson Jr HF. Reduced expression of metastasis suppressor RhoGDI2 is associated with decreased survival for patients with bladder cancer. Clin Cancer Res 2004;10:3800–6. [6] Tseng SF, Chang CY, Wu KJ, Teng SC. Importin KPNA2 is required for proper nuclear localization and multiple functions of NBS1. J Biol Chem 2005;280:39594–600. [7] Zannini L, Lecis D, Lisanti S, et al. Karyopherin-alpha2 protein interacts with Chk2 and contributes to its nuclear import. J Biol Chem 2003;278:42346–51. [8] Teng SC, Wu KJ, Tseng SF, Wong CW, Kao L. Importin KPNA2, NBS1, DNA repair and tumorigenesis. J Mol Histol 2006;37:293–9.
Int J Cancer. In press. doi:10.1002/ijc.25568. [23] Wang Y, Klijn JG, Zhang Y, et al. Gene-expression profiles to predict distant metastasis of lymph-node-negative primary breast cancer. Lancet 2005;365:671–9. [24] Zheng M, Tang L, Huang L, et al. Overexpression of karyopherin-2 in epithelial ovarian cancer and correlation with poor prognosis. Obstet Gynecol 2010;116:884–91. [25] Sakai M, Sohda M, Miyazaki T, et al. Significance of karyopherin{alpha} 2 (KPNA2) expression in esophageal squamous cell carcinoma. Anticancer Res 2010;30:851–6. [26] Winnepenninckx V, Lazar V, Michiels S, et al. Gene expression profiling of primary cutaneous melanoma and clinical outcome. J Natl Cancer Inst 2006;98:472–82. [27] Dangle PP, Gong MC, Bahnson RR, Pohar KS. How do commonly performed lymphadenectomy templates influence bladder cancer nodal stage? J Urol 2010;183:499–503. [28] Shariat SF, Lotan Y, Vickers A, et al. Statistical consideration for clinical
[9] Dyrskjot L, Kruhoffer M, Thykjaer T, et al. Gene expression in the
biomarker research in bladder cancer. Urol Oncol 2010;28:389–400.
urinary bladder: a common carcinoma in situ gene expression
[29] Hafner C, Knuechel R, Stoehr R, Hartmann A. Clonality of multifocal
signature exists disregarding histopathological classification. Can-
urothelial carcinomas: 10 years of molecular genetic studies. Int J
cer Res 2004;64:4040–8.
Cancer 2002;101:1–6.