Introduction

At present, radical cystectomy (RC) remains the gold standard treatment for those with high-risk non-muscle-invasive bladder cancer (NMIBC) and muscle-invasive bladder cancer (MIBC) [1]. However, despite the aggressive local approach, mortality rates are still high, with approximately 50 % of patients succumbing in the first 5-year after the surgery [2]. Therefore, identification of the risk of cancer-specific mortality is critical to guide surveillance and select adjuvant therapies. Numerous studies have investigated potential prognostic factors for urothelial bladder cancer (UBC), in order to guide therapeutic strategies and improve outcomes. We decided to conduct the current study by daily using and commonly accessible clinicopathological markers to investigate their contribution to predict prognosis of UBC patients after RC.

Materials and methods

General information

Between July 2007 and June 2014, a total of 103 patients with UBC who underwent RC with bilateral pelvic lymph node dissection (PLND) at our institution were included in this study. The indication of RC comprised patients with high-risk superficial or MIBC without evidence of extravesical spread or metastatic disease on preoperative imaging. High-risk superficial bladder cancers were defined as pT1 high-grade (HG) disease, frequently recurrent pT1 or HG disease, or any unresectable NMIBC through transurethral resection of bladder tumor (TURBT). All patients underwent RC in 3 months following the first diagnostic TURBT. None of the patients received neoadjuvant chemotherapy or radiotherapy. Adjuvant cisplatin-based combination systemic chemotherapy was routinely administrated in patients with extravesical invasion and/or lymph node positive disease postoperatively. However, a small proportion of patients had to give up systemic chemotherapy due to severe intolerability.

Pathological evaluation

All specimens were examined and evaluated by pathologists with expertise in genitourinary diseases. The specimens were staged and graded according to the Union for International Cancer Control (UICC) TNM classification (2002) [3] and the World Health Organization (WHO)/International Society of Urological Pathology (ISUP) (2004) [4], respectively. The extent of PLND was at the surgeon’s discretion. Pelvic lymph node dissections were examined grossly. All lymphoid tissue was submitted for histological examination.

Immunohistochemistry

All samples were paraffin embedded, and sections were cut into 4 μm and mounted on glass slides. Before further processing, heat-mediated antigen retrieval was performed by boiling the slides in 0.01 M citrate buffer, pH 6.0, for 20 min in a microwave oven. The primary antibodies were diluted, including anti-p53 (Novocastra, Newcastle, UK) at 1:50, anti-Ki67 (Novocastra, Newcastle, UK) at 1:100 and anti-p63 (Novocastra, Newcastle, UK) at 1:50. The slides were stained immunohistochemically using the avidin–biotin complex method for all antibodies. Finally, the slides were dehydrated through graded alcohols to xylene and mounted in a mounting medium. For positive controls, we used colon carcinoma tissue for Ki67 and p53, and human squamous epithelium tissue for p63. For negative controls, all primary antibodies were omitted.

Assessment of Ki67, p53 and p63 staining and scoring

Staining and scoring protocols for Ki67, p53 and p63 were previously described [5–7]. Ki67 was scored by label index, while p53 value was scored semiquantitatively as 0 for negative, 1+ for mild, 2+ for moderate and 3+ for strong. The expression of p63 was scored semiquantitatively as 0 for proportion of stained cells no more than 10 % and + for proportion of stained cells more than 10 %. Ki67 labeling index was considered high when samples demonstrated 40 % or greater positivity.

Follow-up regimen

Postoperatively, patients were seen quarterly in the first year, semiannually in the second year and annually thereafter. Follow-up visits included physical examination and serum routine blood tests. Diagnostic imaging of the upper urinary tract and chest radiography were performed at least annually. The cause of death was determined by the treating physicians through chart review or clinical history. Patients who had widely disseminated metastases at time of death were identified as having died of UBC. Perioperative mortality (death within 30 days after RC) was censored at time of death for cancer-specific survival (CSS) analyses. All patients were followed up retrospectively both through hospital records and through telephone interviews, either with the patients or with their close relatives. Informed consents were obtained from all patients, and the study was approved by the Huashan Institutional Review Board (HIRB).

Statistical analysis

The SPSS 17.0 for Windows was used for statistical analyses. All data were presented as mean ± standard deviation (SD). The Student’s t test was applied to compare scores of Ki67 between two groups, while analysis of variance (ANOVA) was used for comparisons in more than two groups. We applied the Mann–Whitney U test for the comparison of p53 and p63 between two categories and the Kruskal–Wallis test for more than two categories. Multivariate Cox proportional regression analysis was performed to determine the independent contribution of clinicopathological factors for CSS. The end-point variable of interest was cancer-specific deaths. These hazards were estimated with their 95 % confidence interval. A P value of 0.05 was considered statistically significant.

Results

Population characteristics

In our cohort, there were 78 males (75.73 %) and 25 females (24.27 %). The average age of the patients was 65.85 years old (35–86 years). There were 12 specimens (11.65 %) graded as low-grade carcinomas and 91 specimens (88.35 %) graded as high-grade carcinomas. There were 30 cases (29.13 %) staged as pTa/T1, 34 cases (33.01 %) as pT2, 30 cases (29.13 %) as pT3 and 9 cases (8.73 %) as pT4. As for lymph node metastasis, there were 88 cases (85.43 %) staged as pN0, 6 cases (5.83 %) as pN1, 6 cases (5.83 %) as pN2 and 3 cases (2.91 %) as pN3. Lymphovascular invasion (LVI) was discovered in 21 specimens (20.39 %). Thirty-eight patients (44.66 %) had a higher level of preoperative neutrophil/lymphocyte ratio (NLR ≥ 2.7). ABO blood type was O in 29 (28.16 %), A in 40 (38.83 %), B in 22 (21.36 %) and AB in 12 (11.65 %) patients. Twelve cases (11.65 %) have type 2 diabetes mellitus at the time of diagnosis.

Ki67, p53 and p63 expressions in relation to clinicopathological parameters

Expression levels of Ki67, p53 and p63 are summarized in Table 1. Ki67 expression was significantly elevated with the progression of tumor grade (P < 0.001) and tumor pT stage (P < 0.001). Elevated Ki67 was seen in patients with MIBC (P < 0.001) and LVI disease (P = 0.009). The expression of p53 was elevated with the progression of tumor grade (P < 0.001) but not stage. Furthermore, there were significant differences in terms of p53 expression among patients with different ABO blood type. Neither Ki67 nor p53 expression was associated with tumor pN stage or lymph node status. However, the expression of p63 showed no preference in tumor grade, tumor pT stage, pN stage and any other clinicopathological parameters.

Table 1 Ki67, p53 and p63 expressions in relation to clinicopathological parameters (mean ± SD)
Full size table

Cancer-specific survival in all patients underwent RC

The median follow-up was 22 months (range 2–89 months) for the whole cohort. As of January 2015, 70 patients (67.96 %) were still alive at the last follow-up with a median follow-up of 26 months (3–89 months). Twenty-nine patients (28.16 %) died of UBC during follow-up, with a median follow-up of 10 months (2–54 months). Multivariate Cox proportional hazards regression analyses were performed, adjusted for gender, tumor grade, pT stage, pN stage, LVI status, preoperative NLR, ABO blood type, history of diabetes mellitus (DM), expressions of Ki67, p53 and p63. Tumor pT stage, pN stage, ABO blood type, a history of DM (absent vs. present, RR 0.104, 95 % CI 0.014–0.786) and p63 expression (negative vs. positive, RR 3.623, 95 % CI 1.113–11.793) (Fig. 1a) were significant risk factors for CSS. Moreover, a weak positive correlation (P = 0.056) was observed between gender and CSS (Table 2).

Fig. 1
figure 1

a Cancer-specific survival probability of all 103 UBC patients treated with RC based on p63 expression. b Cancer-specific survival probability of 73 MIBC patients treated with RC based on Ki67 expression

Full size image
Table 2 Cox regression in the prediction of cancer-specific survival of all patients who underwent radical cystectomy
Full size table

Cancer-specific survival in MIBC patients

There were 73 cases of MIBC diseases in our cohort, and the median follow-up was 14 months (2–72 months). As of January 2015, 44 patients (60.27 %) were still alive at the last follow-up with a median follow-up of 22 months (3–72 months). There were 26 patients (35.62 %) died of UBC during follow-up, with a median follow-up of 9.5 months (2–54 months). In the same way, multivariate Cox proportional hazards regression analyses were performed, adjusted for the same aforementioned clinicopathological parameters. Gender (female vs. male, RR 0.089, 95 % CI 0.019–0.415), ABO blood type, a history of DM (absent vs. present, RR 0.007, 95 % CI 0.000–0.174), Ki67 expression (≤40 vs. >40 %, RR 4.696, 95 % CI 1.144–19.276) (Fig. 1b) and p63 expression (negative vs. positive, RR 35.926, 95 % CI 3.946–327.069) were significant risk factors for CSS. A significant positive correlation was observed between cancer-specific death and tumor T and N stage. In addition, a weak positive correlation was observed (P = 0.062) between LVI and higher cancer-specific death (Table 3).

Table 3 Cox regression in the prediction of cancer-specific survival of MIBC patient who underwent radical cystectomy
Full size table

Discussion

It is not unexpected that pathological tumor T stage and lymph node status were the most important independent prognostic factors in this study, since these variables are well-established prognostic factors [8, 9]. In addition, we found several unconventional prognostic factors which independently associated with CSS.

Ki67 is an indicator of cell proliferation and a measure of cell growth fraction, present during the G1, S, G2 and M stages of the cell cycle [10]. Proliferative activity of tumors determined by Ki67 labeling index has been found to correlate with aggressive behavior of many tumors. Although tumor grade and stage are considered signs of aggressive behavior for bladder cancer, several reports describe a correlation between Ki67 expressions with those well-known prognostic factors [11]. In our study, Ki67 labeling index was correlated with higher tumor grade, pT stage and LVI status, which corresponds with most reports. Some studies showed that Ki67 overexpression was significantly associated with lymph nodes invasion [12]. But we did not find this correlation in our cohort. Furthermore, multivariate analyses in our study likewise confirmed the independent prognostic value of Ki67 in cancer-specific survival in MIBC patients treated with RC. Several studies likewise reported that Ki67 expression was independently associated with both disease recurrence and cancer-specific mortality in patients with organ-confined disease (<pT3 N0) [12–15]. Therefore, we consider that Ki67 expression may help us to identify such patients who are at increased risk of disease progression and may benefit from adjuvant therapies.

TP63 is expressed at a high level in normal human uroepithelial cells, and disruption of its normal expression is associated with neoplastic change. Altered p63 expression in urothelial cancer may occur with progressive loss of urothelial differentiation as tumors advance both in stage and in grade [16]. Several studies concluded that loss of p63 is associated with shorter survival in patients with bladder cancer [17]. In contrast, however, a recent study claimed that maintenance of p63 expression is associated with adverse outcomes in MIBC patients, and maybe T1 patients. Overall survival was shorter in patients with MIBC who retained p63 [18]. In our study, positive p63 expression was proved to be a significant risk factor for cancer-specific survival both in high-risk UBC and in MIBC patients who underwent RC. p63 expression may be another important factor that helps us to identify those patients with high risk of progression and death after RC. However, the molecular mechanism of TP63 in MIBC is unclear and mandates further investigations.

TP53 is a key tumor suppressor gene involved in the maintenance of genomic stability, response to genotoxic stress and activation of cell cycle apoptosis. Some studies have shown correlations between p53 and tumor stage and grade [19]. Nevertheless, others show a lack of association between p53 expression and progression [20, 21]. In our research, we found that p53 was associated with UBC grade but not with stage, which corresponds with our previous studies [6, 7]. Interestingly, p53 expressions were significantly different among patients with different ABO blood types. Some researchers even assert that p53 accumulation in the tumor cell nuclei detected by immunohistochemical methods is an independent predictor of increased risk of recurrence and death in patients with bladder cancer [22]. In our series, unfortunately, multivariate analysis indicated that p53 expression was not associated with cancer-specific survival for all UBC patients who underwent RC. This discrepancy may result from the semiquantitative scoring pattern that we applied.

In line with previous studies [23–25], female patients who underwent RC for UCB were associated with a higher risk of cancer-specific mortality compared with male patients in our study. This gender disparity might be explained as differences in care and/or tumor biology.

In a study among all patients treated with RC [26], blood type B had slightly more unfavorable pathology than others, which translated in worse cancer-specific mortality in univariable analysis. However, blood type lost its statistical significance in multivariate analysis. In our series, however, ABO blood type, in multivariate analysis, was proved to be an independent significant risk factor in all patients who underwent RC. Patient with blood type B have the most unfavorable prognosis than any other blood types.

Numerous studies have been performed to assess the prognostic value of LVI, but the results are still controversial. A meta-analysis recently indicated that LVI is significantly associated with poor outcome in UBC patients who underwent RC [27]. Nonetheless, in our study, a weak positive correlation was observed between LVI and cancer-specific death in MIBC patients.

Recent studies reported that among patients treated with RC for UBC, DM was associated with higher cancer-specific and any-cause mortality in multivariable analyses, whereas metformin use was associated with decreased risk of disease recurrence, cancer-specific and any-cause mortality [28, 29]. However, we got a reverse result in our study that DM was a protective factor for CSS. This discrepancy may result from the fact that we failed to classify patients into DM taking metformin and DM without metformin due to small sample size in our cohort.

Last but not least, this study is subject to limitations. The retrospective nature of our study renders some of the results biased. The sample size is still small, limiting the ability to distinguish outcomes further stratified. In addition, differences in surgical technique might have influenced outcomes since all RCs were performed by four urologists. Furthermore, some of our statistically significant results remain hard to explain or interpret.

Conclusion

This study demonstrates that apart from tumor pathological T, N stage and gender, immunohistochemical expressions of Ki67 and p63 are potential prognostic factors for UBC patients who underwent RC. Besides that, ABO blood type may have predictive effect on prognosis as well.