Expression of epithelial–mesenchymal transition markers in renal cell carcinoma: impact on prognostic outcomes in patients undergoing radical nephrectomy

Authors


Hideaki Miyake, Division of Urology, Kobe University Graduate School of Medicine, 7-5-1 Kusunoki-cho, Chuo-ku, Kobe 650-0017, Japan. e-mail: hideakimiyake@hotmail.com

Abstract

What's known on the subject? and What does the study add?

There have been few studies evaluating the prognostic value of epithelial-mesenchymal transition markers in renal cell carcinoma (RCC); therefore, the significance of these markers in the prognosis of patients with RCC, particularly that in those with localized disease, remains largely unknown.

Consideration of the expression levels of potential epithelial-mesenchymal transition markers, particularly clusterin and Twist in addition to conventional prognostic parameters, would contribute to the prediction of disease recurrence after radical nephrectomy for localized renal cell carcinoma.

OBJECTIVE

  • • To evaluate the expression of multiple molecular markers involved in the process of epithelial–mesenchymal transition (EMT) in renal cell carcinoma (RCC) with the aim of clarifying the prognostic significance of these markers in patients undergoing radical nephrectomy.

PATIENTS AND METHODS

  • • The expression levels of 11 EMT markers, including E-cadherin, N-cadherin, β-catenin, γ-catenin, clusterin, Slug, Snail, Twist, vimentin, ZEB1 and ZEB2, in radical nephrectomy specimens from 122 patients with clinically localized RCC were measured by immunohistochemical staining.

RESULTS

  • • In this series, disease recurrence occurred in 39 (32.0%) patients, with a 5-year recurrence-free survival rate of 64.4%.
  • • Univariate analysis identified expression levels of E-cadherin, clusterin, Twist and vimentin, in addition to C-reactive protein (CRP) level, pathological stage and microvascular invasion, as significant predictors for disease recurrence.
  • • Of these, expression levels of clusterin and Twist, CRP levels and microvascular invasion appeared to be independently related to disease recurrence on multivariate analysis.
  • • There were significant differences in recurrence-free survival according to positive numbers of these four independent factors: disease recurrence occurred in two of 26 patients negative for any risk factor (7.7%), 23 of 73 pateints positive for one or two risk factors (31.5%) and 14 of 23 patients positive for three or four risk factors (60.9%).

CONCLUSION

  • • Consideration of the expression levels of potential EMT markers, particularly clusterin and Twist, in RCC specimens, in addition to conventional prognostic parameters, contributes to the accurate prediction of disease recurrence after radical nephrectomy for localized RCC.
Abbreviation
EMT

epithelial–mesenchymal transition

INTRODUCTION

RCC is the most common malignancy of the adult kidney, and annual estimates of newly diagnosed cases have been increasing gradually over recent years. Because RCC has a highly resistant phenotype to conventional therapeutic modalities, including chemotherapy and radiation, surgical resection of localized disease has been regarded as the only curative treatment; however, ≈20–30% of these patients experience local and/or distant disease recurrence [1]. Although the recent introduction of molecular-targeted agents has modestly improved the prognosis of such patients, most metastatic diseases remain incurable [2]. Therefore, it is important to characterize the molecular biological events that drive the growth and metastasis of RCC with the aim of developing an optimal therapeutic strategy against recurrent diseases.

Epithelial–mesenchymal transition (EMT) is the process by which polarized epithelial cells are converted into motile mesenchymal cells. Alterations in adhesion, morphology, cellular architecture and migration potential comprise the major events observed during this process [3]. Initially, EMT was intensively investigated as a crucial process in embryonic development in a number of animal models [4]. In recent years, this process has also been shown to apply to the progression and metastasis of a wide variety of malignant tumours, including RCC [5–15]. For example, Chuang et al. [14] reported the enhanced invasive potential in human RCC 786-O cells by tumour-derived tumour necrosis factor-α-induced EMT, as characterized by the repression of E-cadherin, as well as the promotion of vimentin and matrix metalloproteinase 9 [14].

Multiple complex signalling pathways are needed for the induction of EMT in tumour cells [16]; thus, several types of molecular marker mediating the execution of EMT have been identified, such as epithelial markers, mesenchymal markers and transcriptional factors [17], and various studies have reported the usefulness of these markers as prognostic predictors in some types of malignant tumour [5–10]. However, few studies have evaluated the prognostic value of EMT markers in RCC [15,18]; therefore, the significance of these markers in the prognosis of patients with RCC, particularly in those with clinically localized disease, remains largely unknown. In the present study, we evaluated the expression of multiple EMT markers, including E-cadherin, N-cadherin, β-catenin, γ-catenin, clusterin, Slug, Snail, Twist, vimentin, ZEB1 and ZEB2, in radical nephrectomy specimens from patients with clinically organ-confined RCC, aiming to clarify the prognostic significance of these markers in this category of patients.

PATIENTS AND METHODS

The present study included a total of 122 consecutive patients who were diagnosed with clinically organ-confined RCC, and subsequently underwent radical nephrectomy between 1998 and 2008 at our institution. Pathological examinations were performed by a single pathologist in accordance with the 2009 American Joint Committee on Cancer TNM classification system. Information on the clinicopathological characteristics of the patients was retrieved from their medical records. The median (range) duration of postoperative follow-up of this series was 44 (8–148) months. Generally, all patients were followed by laboratory and radiological evaluations, as well as physical examination, every 6 months, to monitor recurrence and metastasis. In the absence of a relapse of RCC 5 years after surgery, the interval between re-examinations was increased. All patients provided their informed consent to participate in the present study. The study design was approved by the Research Ethics Committee of our institution.

Immunohistochemical staining of radical nephrectomy specimens was performed as described previously [19]. Briefly, sections from formaldehyde-fixed, paraffin-embedded tissue obtained from 122 specimens were deparaffinized by xylene and rehydrated in decreasing concentrations of ethanol. After blocking endogenous peroxidase with 3% hydrogen peroxidase in methanol, sections were boiled in 0.01 m citrate buffer for 10 min and incubated with 5% normal blocking serum in Tris-buffered saline for 20 min. The sections were then incubated with anti-human antibodies: E-cadherin mouse monoclonal antibody, N-cadherin mouse monoclonal antibody (Dako, Carpinteria, CA, USA); β-catenin mouse monoclonal antibody (BD Transduction Laboratories, Franklin Lakes, NJ, USA); γ-catenin mouse monoclonal antibody, clusterin rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA); Slug rabbit polyclonal antibody, Snail rabbit polyclonal antibody, Twist rabbit polyclonal antibody (Abcam, Cambridge, UK); vimentin mouse monoclonal antibody, ZEB1 rabbit polyclonal antibody (Santa Cruz Biotechnology); and ZEB2 rabbit polyclonal antibody (Assay Biotechnology, Sunnyvale, CA, USA). The sections were then incubated with biotinylated goat anti-mouse or rabbit IgG (Vector Laboratories, Burlingame, CA, USA). After incubation in an avidin–biotin peroxidase complex for 30 min, the samples were exposed to diaminobenzidine tetrahydrochloride solution and counterstained with haematoxylin.

Staining results were interpreted by two independent observers who were blind to the clinicopathological data. In accordance with a previous study [13], the staining intensity was scored visually as 0 (no staining at all), 1 (weak), 2 (medium) or 3 (strong), and the staining extent was also scored as 0 (0–5%), 1 (5–25%), 2 (26–75%) or 3 (75–100%). The intensity and extent scores were summed to obtain a composite score, and scores ≤3 and >3 were considered to indicate weak and strong expression, respectively.

All statistical analyses were performed using Statview, version 5.0 (Abacus Concepts, Inc., Berkeley, CA, USA). P < 0.05 was considered statistically significant. A chi-squared test was used to analyze the association between several clinicopathological factors and expression levels of molecular markers. Recurrence-free survival rates were calculated employing the Kaplan–Meier method, and differences were determined by the log-rank test. The prognostic significance of certain factors was assessed using the Cox proportional hazards regression model.

RESULTS

The clinicopathological characteristics of the 122 patients included in the present study are provided in Table 1. The association of several conventional clinicopathological factors and expression levels of the 11 EMT markers in radical nephrectomy specimens was then analyzed. As shown in Table 2, CRP level, pathological stage, microvascular invasion and histological subtype were significantly associated with the expression levels of E-cadherin, clusterin and vimentin; those of N-cadherin and γ-catenin; that of γ-catenin; and those of vimentin and ZEB2, respectively. However, the remaining combinations between EMT markers and conventional factors showed no significant correlations.

Table 1. Patient characteristics
CharacteristicValue
Age (years), median (range)65.0 (32–84)
Sex, n (%) 
 Male87 (71.3)
 Female35 (28.7)
Mode of presentation, n (%)
 Incidental90 (73.8)
 Symptomatic32 (26.2)
Karnofsky performance status, n (%)
 <8097 (79.5)
 ≥8025 (20.5)
Preoperative C-reactive protein level, n (%)
 Normal77 (63.1)
 Abnormal45 (36.9)
Pathological stage, n (%) 
 pT164 (52.5)
 pT210 (8.2)
 pT348 (39.3)
Grade, n (%) 
 19 (7.4)
 290 (73.8)
 323 (18.8)
Microvascular invasion, n (%) 
 Negative74 (60.7)
 Positive48 (39.3)
Histological subtype, n (%) 
 Clear cell carcinoma104 (85.2)
 Non-clear cell carcinoma18 (14.8)
Table 2. Correlation of expression levels of epithelial–mesenchymal transition markers in radical nephrectomy specimens with conventional prognostic factors
Molecular marker*Age (years)Mode of presentationKarnofsky performance status (%)C-reactive proteinPathological stageGradeMicrovascular InvasionHistological subtype
<70≥70 P IncidentalSymptomatic P <80≥80 P NormalAbnormal P pT1 or pT2pT3 P 1 or 23 P NegativePositive P CCCNon-CCC P
  1. CCC, clear cell carcinoma. *Values are expressed as the number of patients with a strong expression of each molecule.

E-cadherin22160.8725130.183080.933080.04822160.672990.3624140.703440.38
N-cadherin32300.09148140.3549130.8939230.412834<0.00148140.2834280.185660.11
β-catenin14140.321990.422440.3517110.3418100.652440.4816120.192350.28
γ-catenin31290.1045150.7647130.7541190.6631290.04648120.7530300.01850100.56
Clusterin25180.9933100.583670.5523200.03823200.233580.9622210.113850.80
Slug27220.4737120.7233160.3329200.4625240.0744090.9128210.524270.90
Snail22160.1028100.992990.7724140.6127110.423440.1122160.673530.15
Twist26210.5136110.5737100.8727200.1025220.1837100.5928190.854250.31
Vimentin36290.3851140.2150150.4535300.001736290.8753120.9238270.606140.0042
ZEB110120.151570.511930.3812100.191390.871920.2211110.262020.41
ZEB234190.3140130.714670.006736170.7534190.494490.6431220.6741120.031
Overall725090329725774574489923744810418

During the observation period of the present study, 39 of the 122 (32.0%) patients developed disease recurrence, and the 1-, 3- and 5-year recurrence-free survival rates were 83.6%, 74.7% and 64.4%, respectively (Fig. 1). By univariate analysis using the Cox proportional hazards regression model, the weak expression of E-cadherin, as well as the strong expression of clusterin, Twist and vimentin, were identified as significant factors associated with recurrence-free survival, whereas CRP level, pathological stage and microvascular invasion were also significant among the clinicopathological factors examined (Table 3). Furthermore, a multivariate analysis of seven significant factors determined by univariate analysis was performed to evaluate the predictive value for disease recurrence, showing that the overexpression of clusterin and Twist, an abnormal CRP level, and positive microvascular invasion, were independently associated with disease recurrence, irrespective of the other factors included (Table 3). Representative findings of an immunohistochemical study for detecting clusterin and Twist expression are shown in Fig. 2. Recurrence-free survival curves according to the expression levels of clusterin and Twist, CRP level and microvascular invasion are presented in Fig. 3. There were significant differences in recurrence-free survival with respect to all of these four factors.

Figure 1.

Recurrence-free survival of the 122 patients with clinically localized RCC who underwent radical nephrectomy.

Table 3. Univariate and multivariate analyses of association between various parameters with recurrence-free survival in 122 patients with renal cell carcinoma who underwent radical nephrectomy
VariablesNumber of patients (number of patients with disease recurrence)Univariate analysisMultivariate analysis
Hazard ratio P Hazard ratio P
  1. CCC, clear cell carcinoma.

Age (years) (<70 vs ≥70)71 (23) vs 51 (16)0.990.99
Mode of presentation (incidental vs symptomatic)90 (33) vs 32 (6)2.320.059
Karnofsky performance status (%) (<80 vs ≥80)97 (31) vs 25 (8)0.810.69
C-reactive protein level (normal vs abnormal)77 (10) vs 45 (29)5.59<0.0012.770.010
Pathological stage (pT1 or pT2 vs pT3)74 (15) vs 48 (24)3.860.00811.950.089
Grade (1 or 2 vs 3)99 (32) vs 23 (7)1.120.81
Microvascular invasion (negative vs positive)74 (14) vs 48 (25)4.33<0.0012.680.016
Histological subtype (CCC vs non-CCC)104 (35) vs 18 (4)1.430.53
E-cadherin (weak vs strong expression)84 (35) vs 38 (4)3.940.0100.490.21
N-cadherin (weak vs strong expression)60 (16) vs 62 (23)1.680.11
β-catenin (weak vs strong expression)94 (27) vs 28 (12)1.610.67
γ-catenin (weak vs strong expression)62 (19) vs 60 (20)1.040.56
Clusterin (weak vs strong expression)79 (17) vs 43 (22)5.320.00292.890.020
Slug (weak vs strong expression)73 (17) vs 49 (22)1.880.051
Snail (weak vs strong expression)84 (24) vs 38 (15)1.150.67
Twist (weak vs strong expression)75 (13) vs 47 (26)3.110.00722.100.048
Vimentin (weak vs strong expression)57 (11) vs 65 (28)2.030.0381.420.71
ZEB1 (weak vs strong expression)100 (26) vs 22 (10)1.570.24
ZEB2 (weak vs strong expression)69 (19) vs 53 (20)1.230.52
Figure 2.

Representative findings of immunohistochemical staining of radical nephrectomy specimens with a clusterin or Twist antibody. A, RCC with weak expression of clusterin. B, RCC with strong expression of clusterin. C, RCC with weak expression of Twist. D, RCC with strong expression of Twist.

Figure 3.

Recurrence-free survival of patients with clinically localized RCC who underwent radical nephrectomy according to the expression level of clusterin (A), that of Twist (B), preoperative c-reactive protein level (C) and microvascular invasion (D).

To more precisely characterize the prognostic features after radical nephrectomy, we categorized patients according to positive numbers of four independent risk factors for disease recurrence. Disease recurrence occurred in two of 26 patients negative for any risk factor (7.7%), 23 of 73 positive for a single or two risk factors (31.5%) and 14 of 23 positive for three or four risk factors (60.9%). As shown in Fig. 4, there were significant differences in recurrence-free survival among these three groups.

Figure 4.

Recurrence-free survival of patients with clinically localized RCC who underwent radical nephrectomy according to the number of independent risk factors for postoperative disease recurrence, including strong expression of clusterin, that of Twist, abnormal level of preoperative C-reactive protein and positive microvascular invasion.

DISCUSSION

It is important to precisely predict the postoperative clinical course of patients undergoing surgical resection of malignant tumours using a widely accepted consensus that allows physicians to make a decision on the follow-up schedule, as well as additional treatment. This concept is particularly true in RCC patients, considering that its phenotype is highly resistant to conventional non-surgical treatments, even after the introduction of novel molecular-targeted agents [1,2]. To date, staging based on the TNM classification system has been regarded as the standard for stratifying patients into risk groups with respect to prognosis; however, studies have suggested limitations for predicting postoperative disease recurrence in patients with localized RCC using a staging system alone [18]. Accordingly, the prognostic significance of a number of molecular markers associated with the pathogenesis of RCC has been investigated [18,20,21]. In recent years, it has been shown that EMT plays crucial roles in the progression of malignant tumours [5–15], and that several molecules involved in EMT are able to be used as molecular markers predicting the prognosis in patients with some types of malignant tumour [5–10]. However, there are few data available on the significance of EMT markers in RCC patients [15,18]. In the present study, we analyzed the value of conventional clinicopathological prognostic factors, as well as multiple potential EMT markers, for predicting clinical outcome in 122 patients undergoing radical nephrectomy for clinically localized RCC.

EMT, consisting of multiple steps, could be involved in a wide variety of events driving malignant tumour progression, such as carcinogenesis, invasion, metastasis and resistance to pro-apoptotic stimuli; thus, several types of molecule, including epithelial markers, mesenchymal markers and transcriptional factors, are implicated in the execution of this process [5,6]. The present study focused on 11 potential EMT markers whose expression could be detected in most RCC tissues; however, there were varied patterns regarding the relationship between the expression levels of each molecule and conventional prognostic parameters. This could be explained by the spatial and/or temporal heterogeneity of EMT at a comparatively early stage of malignant disease [17]; therefore, cancer cells undergoing oncologically relevant EMT in clinical specimens need to be identified, such as those at the invasive tumour front, even if it is a small cell cluster [22].

We subsequently compared the impact of established prognostic parameters and potential molecular markers on recurrence-free survival after radical nephrectomy. Univariate analysis showed that CRP level, pathological stage, microvascular invasion, and expression levels of E-cadherin, clusterin, Twist and vimentin, were significantly associated with postoperative disease recurrence. Of these factors, only CRP level, microvascular invasion and expression levels of clusterin and Twist appeared to be independent prognostic predictors on multivariate analysis. Although a higher CRP level, presence of microvascular invasion and overexpression of clusterin were shown to correlate well with the prognosis of RCC patients after radical nephrectomy [23–25], there are no studies available showing the prognostic relevance of Twist, a basic helix-loop-helix transcriptional factor [26], in patients with RCC. However, considering the function of Twist, which transcriptionally represses the expression of E-cadherin, leading to a loss of E-cadherin-mediated cell–cell adhesion and a gain of cell motility [26], the present findings could be supported theoretically. A significant inverse association between the expression levels of Twist and E-cadherin was observed in the present series (data not shown).

It would be of interest to evaluate the ability to predict disease recurrence after radical nephrectomy by combining powerful prognostic indicators allowing a better individualization of follow-up, as well as postoperative treatment schedules, for patients with RCC. Accordingly, based on the multivariate analyses identifying four independent risk factors for disease recurrence, the 122 patients were classified into three groups: negative for any risk factor, positive for a single or two risk factors and positive for three or four risk factors, and significant differences in recurrence-free survivals among these three groups were observed. Collectively, these findings suggest that a consideration of the four main risk factors identified by multivariate analysis (i.e. clusterin and Twist expression levels, CRP level and microvascular invasion) may contribute to the development of a novel system that can more precisely predict disease recurrence after radical nephrectomy.

There are limitations to the present study that need to be considered. The present study is retrospective study, including a small number of patients, and the follow-up period is not yet sufficiently long to allow definitive conclusions to be made on prognostic issues. In addition, as a result of recent progress in the field of EMT research, it is currently hypothesized that novel concepts, such as cancer stem cells and microRNA, could be involved in EMT [27,28]; therefore, additional investigations focusing on molecules associated with these concepts are warranted. Finally, 11 of the markers included in the present study were selected by a subjective rather than scientifically objective criterion; hence, there may be other molecules that more closely reflect the postoperative prognosis in RCC patients.

In the present series, we evaluated the significance of multiple molecular markers involved in the process of EMT, in addition to that of several conventional prognostic parameters, as predictors of postoperative disease recurrence in 122 patients with clinically organ-confined RCC. The expression levels of clusterin and Twist, preoperative CRP level and microvascular invasion appeared to be independently associated with recurrence-free survival after radical nephrectomy. Moreover, the usefulness of the combined application of these four independent factors is suggested to contribute to a further refinement of the system predicting postoperative disease recurrence. However, a prospective study including additional potential molecular markers underlying the malignant progression of RCC will be required before any definitive conclusions can be made.

CONFLICT OF INTEREST

None declared.

Ancillary