Intensity of Nuclear Staining for Ki-67, p53 and Survivin as a New Prognostic Factor in Non-muscle Invasive Bladder Cancer

The basic treatment for bladder cancer diagnosed at the stage of non-muscle invasive bladder cancer (NMIBC) is transurethral resection of the bladder tumor (TURBT). However, observational studies indicate that 48–61% of patients undergoing this treatment have a recurrence of the disease, and thereby require close monitoring for many years [1‐3]. The well-known risk factors for cancer recurrence include the number and size of neoplastic lesions, the presence of coexisting carcinoma in-situ (CIS), the number of recurrences, T stage according to the TNM classification, the histopathological grade of the tumor, sex, and age [2, 4]. Based on selected risk factors, prognostic tables have been developed to determine the risk of NMIBC recurrence; however, as demonstrated by later analyses, they overestimate the risk of recurrence in some groups of patients [4, 5]. Adding new elements to existing tables may improve accuracy and, at the same time, increase the usability of prognostic models. Factors of potential prognostic significance are: p53, Ki-67, survivin, β-catenin, E-cadherin and N-cadherin.

The p53 protein is an important element regulating the cell cycle. Activation of p53 leads to the induction or inhibition of the expression of genes involved in the regulation of the cell cycle at G1 and G2 checkpoints, regulation of apoptosis, DNA repair and cell aging [6, 7]. Mutation of the TP53 gene may cause urothelial hyperplasia and dysplasia, which are early stages of neoplastic transformation [8]. Survivin, a member of the inhibitor of apoptosis family, plays a role in the chromosomal segregation by regulating the transition from the G2 to M phase of the cell cycle. The survivin/CDK4 complex also activates the p21 protein that inhibits the apoptosis [9, 10]. Studies carried out by Zhang et al. showed that survivin silencing results in inhibition of tumor cell growth [11]. The Ki-67 protein, regarded as a proliferation marker, can be detected in the cell nuclei in the G1, S and G2 phases of the cell cycle and in mitosis. The function of this protein in the regulation of the cell cycle has not been clearly determined, but it has been shown that its presence is essential for the proliferation process [12]. E-cadherin, N-cadherin and β-catenin are involved in maintaining intercellular connections. Decreased expression of E-cadherin and increased expression of N-cadherin are elements of the epithelial-mesenchymal transition (EMT) process that increase epithelial cell migration [13]. β-catenin, which binds to actin cytoskeleton elements (a reaction mediated by cadherins), is also involved in the control of adherence and migration of cells [14, 15]. Studies on transgenic mice have shown that increased activation of β-catenin causes hyperplasia of the bladder epithelium, which may precede the development of cancer [16].

The aim of the study was to determine the prognostic value of the expression level of biomarkers selected on the basis of the literature: p53, Ki-67, survivin, β-catenin, E-cadherin and N-cadherin. The data collected in this study allow for more precise stratification of the risk of recurrence and identification of a group of patients at high risk of cancer recurrence.

The study group consisted of 134 patients (113 men and 21 women) with papillary non-muscle invasive bladder cancer. The overall median follow-up period was 36 months (minimum 12 months; maximum 93 months). During the analyzed period, the recurrence of neoplastic disease occurred in 74 patients, which constituted 55.22% of the examined group. The median time to recurrence was 9 (range: 3–48) months. Detailed characteristics of the study participants are presented in Table 2.

The commonly accepted risk factors of NMIBC recurrence do not allow for precise estimation of the risk of recurrence; therefore, markers are sought for identification of high-risk patients. This study evaluated the expression of proteins involved in the pathogenesis of papillary carcinoma of the bladder and their significance in relation to selected clinical and pathological factors.

Positive staining for β-catenin and E-cadherin was obtained in all analyzed patients. The percentage of NMIBC patients with positive β-catenin staining in published studies was lower than that obtained in the presented study – 75% in the study by Elsherif et al., 60% in the study by Koletsas et al. and 40% in the study by Senol et al. [18‐20]. The obtained results of E-cadherin assays are confirmed by reports by Singh et al., whereas studies performed by Muramaki et al. and Abufaraj et al. showed the expression of E-cadherin in 50–60% of analyzed tissues [21‐23]. The expression of N-cadherin, which indicates that the cell acquires migratory abilities, was only visible in 3.73% of participants of this study. Similar results were obtained by Singh et al. who observed a positive staining for N-cadherin in 13.3% of cases [21]. These data differ from the results reported by Muramaki et al. and Abufaraj et al. who confirmed the expression of N-cadherin in 40% of cases [22, 23]. As for Ki-67 expression, the result of immunohistochemical tests obtained in our study was similar to that of the Chen et al.’ study in which low Ki-67 expression was confirmed in 53% of cases and high in the remaining 47% [24]. In our study, p53 expression was found in more than 95% of patients, and in 30% of cases the percentage of stained cell nuclei was over 33%. Similar results were obtained by Senol et al. and Vetterlein et al. who showed positive staining in 30–40% of cases [19, 25]. The level of survivin expression in our patients was consistent with the results obtained by Xi et al. who confirmed protein expression in 84.5% of cases [26]. Different data were reported by Skagias et al., Jeon et al. and Senol S. et al., who observed protein expression in 61%, 48.2% and 37% of cases [19, 27, 28], respectively.

Discrepancies between the above data may result from various characteristics of patients included in the study due to genetic differences between Ta and T1 tumors. In our study, the percentage of patients with stage Ta cancer was several times higher than that of patients with T1 tumors (40% and 20%, respectively). In the quoted studies performed by other authors, patients with T1 tumors constituted 40–60% of the examined group [18, 22, 23]. In the study by Koletsas et al., patients with muscle invasive bladder cancer were also included in the analysis [20]. There were also differences regarding the interpretation of color reactions. In our study, the maximum score in the evaluation of membrane staining for protein expression was given for total staining of the cell membrane in at least 30% of cells, and in the study by Singh et al. the expression of E-cadherin was considered high if the percentage of stained cells exceeded 67% [21]. In the studies quoted above, a positive result for p53 was staining of at least 80% of cells and a positive result for survivin – at least 10% of cells, whereas in our study any color reaction was considered positive [19, 25, 27]. Differences in the results obtained may be a consequence of the use of different sets of diagnostic antibodies because, as shown by analyzes performed by van de Vijver, the percentage of positive responses may range from 2 to 30% depending on the set of antibodies used [29].

Statistical analysis showed a significant relationship between the expression of proteins involved in the regulation of the cell cycle and the clinical and pathological features of the tumor. The expression level of the Ki-67 protein expressed by the percentage of stained cell nuclei correlated with the degree of histopathological differentiation of the tumor, which is consistent with data from the literature [30‐32]. This study did not confirm the relationship between Ki-67 expression and bladder wall invasion; neither did the study performed by Abdelkri et al. [93]. This relationship was, however, demonstrated in the studies of Wang et al. and Thakur et al. which also included patients in whom the tumor invaded the muscle wall of the bladder [30, 32]. The analysis of the p53 protein expression showed a statistically significant relationship between the histopathological differentiation of the tumor and the percentage of stained cell nuclei and staining intensity, which is consistent with literature reports [19, 25, 33]. Different results were obtained by Wang et al., which may be due to differences between the analyzed groups [30]. Our own study also showed a relationship between the level of the p53 protein expression and the number and size of neoplastic lesions – the assessment of this relationship is not reported by the authors cited, but may confirm the important role of this protein in the development of bladder cancer. The evaluation of the survivin expression confirms data from the literature, which suggests the association of survivin, a protein involved in the inhibition of apoptosis, with the pathogenesis of bladder cancer [26‐28]. However the results of a meta-analysis performed by Lv et al. did not confirm a statistically significant relationship between protein expression and histopathological grade of neoplasm, which indicates the need for further research in this area [34].

The present study confirmed the prognostic significance of known risk factors for tumor recurrence: the diameter of the largest neoplastic lesion (p = 0.001) and lesions found during cystoscopy (p = 0.01). A statistically significant effect on recurrence-free survival was also demonstrated for expression levels of p53, Ki-67 and survivin. The demonstrated relationship between the recurrence-free survival and the survivin expression is consistent with the results of the meta-analysis by Lv et al. (HR 1.831, 95% CI, 1.344–2.493, p = 0.009) [34]. Our study also confirms the analyzes performed by He et al. which revealed the relationship between the level of Ki-67 protein expression and the recurrence-free survival in the Caucasian population (HR 1.441, 95% CI, 1.014–2.047) [35]. The results of presented study differ from those of the meta-analysis by Zhou et al. who did not show a significant relationship between p53 expression and recurrence-free survival in BCG-treated patients (HR 1.400, 95% CI: 0.910–2.160) [36]. The difference between the obtained results may be related to, among other things, various methodologies used in the discussed studies (in our study, protein expression was determined by assessing the staining intensity, while in the Zhou’s meta-analysis the level of p53 expression was categorized by the percentage of stained cell nuclei).

Our investigated group of patients differs significantly from this considered in EORTS. The main reason of this discrepancy is that we included patients with primary NMIBC in opposite to patients with recurrent NMIBC considered in EORTS (system based on data containing the number of tumors, the largest tumor diameter, prior recurrence rate, T category, the presence of concomitant CIS and tumor grade). Thus one of the six conditions could not be assessed. Our observation is confirmed by a lack of significant influence on time to tumor recurrence of some EORTS conditions in our study (Table 6). Nevertheless, the inclusion of any of the proposed new prognostic factors to the EORTS system seems to improve this tool, because the higher R², the higher influence on the prediction variability. On the other hand, the model is better if AIC (Akaike’s Information Criteria) and SBC (Schwartz Bayesian Criterion) are lower because these indexes penalize the inclusion of additional variables to a model (Table 7).

The study has several limitations. It was a single-center, retrospective study in which the prognostic value of selected proteins was evaluated on a small group of patients. In addition, included in the study were only patients who underwent TURBT and were qualified for adjuvant BCG therapy. Nevertheless, the applied assessment system allowed for a precise evaluation of the expression of the proteins of interest and the identification of variables of significant importance in the assessment of the recurrence probability.

Evaluation of the expression of Ki-67, p53 and survivin proteins provides additional information on tumor aggressiveness and may allow more accurate stratification of the risk of cancer recurrence. However, further studies assessing the prognostic value of these proteins are needed.