Prognostic value of apoptotic markers in squamous cell carcinoma of the urinary bladder

Authors


Yair Lotan, UT Southwestern Department of Urology, Moss Bldg, 8th Fl, Ste 112, 5323 Harry Hines Blvd, Dallas, TX 75390-9110, USA. e-mail: yair.lotan@utsouthwestern.edu

Abstract

Study Type – Prognosis (case series)

Level of Evidence 4

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

Apoptotic pathways are important in carcinogenesis. Many studies, involving small numbers of patients, have found an association between one or two apoptotic markers and some of the pathological features of squamous cell carcinoma (SCC).

This study included a large number of patients who had undergone radical cystectomy (RC) for SCC with long-term follow-up, allowing us to study biomarker alterations and their prognostic role. This is the first study on the prognostic role of a panel of apoptotic-related markers in SCC of the urinary bladder, introducing the novel concept of a prognostic marker score based on the number of altered markers. We found that apoptotic markers can improve prediction of oncological outcomes after RC for SCC and might potentially help in patient selection for adjunct therapies.

OBJECTIVE

  • • To evaluate the association of cleaved caspase-3 (CC-3), Bax, COX-2, and p53 expression with pathological features and clinical outcomes in patients with squamous cell carcinoma (SCC) of the urinary bladder.

METHODS

  • • Immunohistochemistry for CC-3, Bax, COX-2, and p53 was performed on tissue microarray sections of radical cystectomy specimens with pure SCC from 1997 to 2003. The relationship between the expression of these markers and pathological features was assessed.
  • • A prognostic marker score (PS) was defined as favourable if ≤2 biomarkers were altered and unfavourable if >2 biomarkers were altered and the association of the PS with oncological outcomes was examined.

RESULTS

  • • The study included 151 patients, of whom 98 were men and 53 were women, with a mean age of 52 years. SCC was associated with schistosomiasis (bilharziasis) in 122 (81%) patients.
  • • Pathological stage was T2 in 50%, T3 in 38%, T1 in 6% and T4 in 6% of patients. Tumours were low grade in 53%, lymph node metastasis was found in 30.5% and lymphovascular invasion was found in 16% of patients.
  • • Median follow-up was 63.2 months.
  • • Advanced stage was associated with COX-2, p53 and CC-3 alterations and high grade was associated with COX-2 alterations (P < 0.05). The total number of altered markers and unfavourable PS were associated with both disease recurrence and bladder cancer-specific mortality in Kaplan–Meier analyses (P < 0.05). Unfavourable PS was an independent predictor of disease recurrence (hazard ratio [HR] 2.694, 95% confidence interval [CI] 1.386–5.235, P= 0. 003) and bladder cancer-specific mortality (HR 2.868, 95% CI 1.209–6.802, P= 0. 017) in multivariable Cox regression analysis.

CONCLUSION

  • • Markers of apoptosis pathways may play an important role in the prognosis of SCC of the bladder. An increased number of altered markers and an unfavourable PS may identify patients who might benefit from multimodal therapies.
Abbreviations
SCC

squamous cell carcinoma

CC-3

cleaved caspase-3

HR

hazard ratio

BC

bladder cancer

UCB

urothelial carcinoma of the bladder

RC

radical cystectomy

PS

prognostic marker score

LN

lymph node

DFS

disease-free survival

CFS

cancer-specific survival

LVI

lymphovascular invasion

PG

prostaglandin

INTRODUCTION

Bladder cancer (BC) is the fourth most common cancer in males in the USA, and the most prevalent malignancy in Egyptian males [1,2]. Egyptian males have the highest mortality rates from BC (16.3 per 100 000), twice as high as the highest rates in Europe and >4 times higher than that in the USA (3.7 per 100 000) [3]. While smoking and occupational exposures are the major risk factors in Western countries, the high incidence of BC in Egypt is attributable to schistosomiasis (bilharziasis) which is also associated with an increased incidence of squamous cell carcinoma (SCC) [4,5]. Just as in urothelial carcinoma of the bladder (UCB), conventional pathological features are the main tools used to predict outcomes for SCC, but these tools might not be accurate enough to identify patients at highest risk for recurrence after radical cystectomy (RC), the current ‘gold standard’ for treatment [4]. The integration of molecular markers associated with the biological behaviour of SCC might not only improve prognostication but might also help with decisions about the incorporation of neoadjuvant and adjuvant therapies (chemotherapy and radiotherapy).

Apoptosis, programmed cell death, is influenced by intrinsic and extrinsic signals that converge into a common downstream effector pathway. Alterations in the apoptotic pathway are important beyond tumorigenesis as they allow cancer cells to survive longer, promoting progression and metastasis [6,7]. p53 guards against genetic instability by inducing cell cycle arrest and apoptosis [6,8]. One of the p53 pathways involves activation of Bax, a member of the Bcl-2 family [8,9], which enhances apoptosis by antagonizing the apoptosis inhibitor Bcl-2 [10,11]. Combined alterations of p53 and Bax have been associated with poor prognosis, while an intact p53-to-Bax pathway and low Bax expression have been reported as predictors of survival in many cancers [12,13]. Bax was reported to be a prognostic marker for schistosomiasis-related UCB but not for SCC [14]. Bax alterations change the mitochondrial membrane potential, causing the release of stored calcium and cytochrome C, thereby activating effector caspases, leading to apoptosis [15]. Caspases are the executionary arm of apoptosis and cleaved caspase-3 (CC-3) is best correlated with the initiation of apoptosis because of its location as the most downstream enzyme in the apoptosis pathway [10].

Previous studies have shown that both COX-2 and p53 are implicated in similar cancer pathways including apoptosis, angiogenesis, cellular proliferation and invasion. Moreover, their prognostic role has been studied in other SCCs [16,17] and we recently published their prognostic role in SCC of the urinary bladder [18,19].

Many studies addressing molecular prognostic markers in SCC have included only one or two apoptotic markers. Furthermore, these studies included a small number of patients and did not study the combined prognostic role of apoptotic markers [14,20–23] that was reported in UCB a few years ago [6]. Recently, we introduced the novel concept of a prognostic marker score (PS), defining an unfavourable score based on a greater number of altered markers [24]. In the present study, we investigate the prognostic role of a panel of apoptosis-related markers including CC-3, Bax, p53 and COX-2 in SCC of the urinary bladder. We also evaluated the association of these markers with pathological characteristics as well as with clinical outcome after RC.

MATERIAL AND METHODS

PATIENT POPULATION AND FOLLOW-UP

We reviewed the records and pathological specimens of patients with BC treated by RC and pelvic lymphadenectomy in Mansoura, Egypt from 1997 to 2003. We identified 151 pure SCC cases with sufficient paraffin-embedded archival material from cystectomy specimens available for extensive immunohistochemical evaluation and tissue microarray construction. We excluded patients with mixed histology. Comprehensive clinico-pathological data were collected and entered into an institutional review board-approved database. All patients were followed for disease progression after RC every 2 months in the first 6 months and at 6-month intervals thereafter. A detailed description of the follow-up protocols used has been given previously [18,19].

PATHOLOGICAL EVALUATION

Pathological stage was reassigned according to the 2002 American Joint Committee on Cancer TNM staging system. Tumour grade was based on the 1998 WHO/International Society of Urologic Pathology consensus classification into low-grade and high-grade tumours. The details of pathological evaluation have been described previously [18,19].

CONSTRUCTION OF TISSUE MICROARRAY BLOCKS AND IMMUNOHISTOCHEMISTRY STAINING AND SCORING

Each case was represented using three replicates placed on separate randomly arranged spaces in tissue microarray blocks as described previously [18,19]. Immunohistochemical staining using serial sections from paraffin-embedded tissue microarray blocks was performed. Immunostaining for CC-3 and p53 was performed on Ventana Benchmark XT automatic immunostainer (Ventana, Tucson, AZ, USA). Heat-induced epitope retrieval was performed using citrate or EDTA buffer and a modified pressure cooker. Optimum primary antibody dilutions were predetermined for CC-3 (polyclonal rabbit, 9661, Cell Signaling [Danvers, MA, USA]; dilution 1:50) and p53 (monoclonal mouse, M7001, Dako, Carpinteria, CA, USA; dilution 1:16 000) using known positive control tissues. Immunostaining for Bax and COX-2 was performed at room temperature on Dako Autostainer (Dako). Reagents were used as supplied in the EnVision + System-HRP-labelled Polymer, anti-mouse (Dako). Target buffer, ph 6.1 was used for antigen retrieval. Optimum primary antibody dilutions were predetermined for Bax (polyclonal rabbit, A3533, Dako; dilution 1:100) and COX-2 (polyclonal rabbit, RB-9072-P1, Thermo Fisher Scientific, Waltham, MA, USA; dilution 1:200) using known positive control tissues. Appropriate positive and negative controls were used.

Evaluation of immunohistochemical results was performed using bright field microscopy imaging coupled with advanced colour detection software (Automated Cellular Imaging System, ChromaVision Medical Systems Inc, San Juan Capistrano, CA, USA). We obtained the percentage of positive cells by using 4–6 random hot spots in each punch of every case. An investigator (P.K.), blinded to the sample tracking system, confirmed all automatically generated scores microscopically. The mean of the triplicate cores was calculated for data analysis. All markers were placed in one of two categories, altered or normal. P53 immunoreactivity was considered altered if ≥ 10% cells and COX-2 was considered altered if >20% cells were stained positive [18,19]. CC-3 and Bax immunoreactivities were considered altered if there was no or <10% weakly stained cells.

STATISTICAL ANALYSIS

Pearson's chi-squared test was performed to examine the relationships of biomarker expression with pathological features. The PS was defined as favourable if ≤2 biomarkers were altered and unfavourable if >2 biomarkers were altered. Outcomes were measured by time to disease recurrence or to bladder cancer-specific mortality. Disease recurrence was defined as local failure in the cystectomy bed or regional lymph nodes (LNs) or distant metastasis after RC. The period of disease-free survival (DFS) was defined as the time between the date of RC and the development of local recurrence or distant metastasis. Censored survival values represent patients who were alive without clinical evidence of disease at the last follow-up. Cause of death was determined by chart review or by treating physicians. The period of cancer-specific survival (CSS) was defined as the time between the date of RC and death from cancer. Univariate recurrence and survival probabilities after RC were estimated using the Kaplan–Meier method and differences were assessed using the log-rank statistic. Multivariate Cox regression analysis was used for time to recurrence and the cancer-specific mortality after RC. All reported P values are two-sided and a P value of <0.05 was considered to indicate statistical significance. All statistical tests were performed using spss version 17.0.

RESULTS

PATIENT DEMOGRAPHICS AND CLINICO-PATHOLOGICAL FINDINGS

Of the 151 patients in the present study, 98 were men and 53 were women. Their mean (range) age at diagnosis was 52 (36–74) years. Pathological stage was T2 in 50%, T3 in 38% and T1 and T4 in 6% of the patients and 53% of the tumours were low grade. Gross and cystoscopic descriptions were available for all tumours. Nodular or fungating tumours were seen in 92% of cases (n= 139) and the rest were ulcerating (n= 4), fibrillary (n= 3), papillary (n= 3) or had another configuration (n= 2). Schistosomiasis (bilharzial infection) was found in 81% of patients (n= 122). A mean (range) of 22 (4–70) LNs were removed during RC. LN invasion was found in 30.5% (n= 46) and lymphovascular invasion (LVI) was found in 16% of patients (n= 24). Table 1 shows the patients' characteristics, including clinico-pathological features. The median (range) follow-up was 63.2 months (0–100) months.

Table 1.  Patients' characteristics, including clinco-pathological features and marker alterations
VariableNo. of patients (%)
Total151
Median age, years51.83
Age range, years36–74
Gender 
 Female53 (35.1 )
 Male98 (64.9 )
Schistosomiasis 
 Absent29 (19.2 )
 Present122 (80.8 )
Tumour grade 
 Low80 (53.0 )
 High71 (47.0 )
Pathological T stage 
 pT110 (6.6 )
 pT275 (49.7 )
 pT357 (37.7 )
 pT49 (6.0 )
LN involvement 
 Absent105 (69.5 )
 Present46 (30.5 )
Organ confinement 
 Confined (T1-2/N0)68 (45 )
 Non-confined (T3-4/N+)83 (55 )
LVI 
 Absent127 (84.1 )
 Present24 (15.9 )
 Variables 
Total markers altered 
 01 (0.7 )
 159 (39.1 )
 258 (38.4 )
 325 (16.6 )
 48 (5.3 )
PS 
 ≤2 (favourable)118 (78.1 )
 >2 (unfavourable)33 (21.9 )

ASSOCIATION OF MARKERS WITH PATHOLOGICAL CHARACTERISTICS AND OTHER MARKERS

Figure 1 shows representative immunohistochemistry staining for CC-3, Bax, p53 and COX-2. One or more marker was altered in every patient. Overall, CC-3 was the most altered marker, being altered in 135 patients (89.4%) and Bax was the least altered marker, being altered in 21 patients (13.9%). Table 2 shows the association of the studied markers with clinico-pathological characteristics. There was no difference in the number of LNs removed or in age between patients with normal and patients with altered expression of any of the markers (P > 0.05). Advanced stage was associated with altered COX-2, p53 and CC-3, the number of total markers altered and unfavourable PS. Altered COX-2 was also associated with high grade (P= 0.019). CC-3, p53 and COX-2 alterations were associated with each other (P < 0.05).

Figure 1.

Representative immunohistochemical expression for cases of SCC of the urinary bladder exhibiting altered and unaltered CC-3, Bax, p53 and COX-2. Original magnification, x100.

Table 2.  Association of the studied markers with clinico-pathological characteristics
 No. of patients (%)COX-2 alterationsp53 alterationCaspase alterationsBax alterations
Alterations, n (%) P Alterations, n (%) P Alterations, n (%) P Alterations, n (%) P
Total15174 (49) 49 (32.5) 135 (89.4) 21 (13.9) 
Tumour grade  0.019 0.495 0.181 0.377
 Low80 (53.0)32 (40.0) 24 (30.0) 69 (86.3) 13 (16.3) 
 High71 (47.0)42 (59.2) 25 (35.0) 66 (93.0) 8 (11.37) 
Pathological stage  0.047 0.104 0.015 0.900
 pT110 (6.6)3 (30.0) 1 (10.0) 10 (100.0) 1 (10) 
 pT275 (49.7)34 (45.3) 21 (28.0) 61 (81.3) 12(16) 
 pT357 (37.7)35 (61.4) 22 (38.6) 65 (96.5) 7 (12.3) 
 pT49 (6.0)2 (22.2) 5 (55.6) 9 (100.0) 1 (11.1) 
Extravesical Extension  0.127 0.050 0.008 0.576
 ≤T285 (56.3)37 (43.5) 22 (25.9) 71 (83.5) 13 (15.3) 
 >T266 (43.7)37 (56.1) 27 (40.9) 64 (97.0) 8 (12.1) 
LN involvement  0.368 0.269 0.282 0.758
 Absent105 (69.5)54 (51.4) 37 (35.2) 92 (87.6) 14 (13.3) 
 Present46 (30.5)20 (43.5) 12 (26.1) 43 (93.5) 7 (15.2) 
LVI  0.915 0.565 0.695 0.828
 Absent127 (84.1)62 (48.8) 40 (31.5) 113 (89) 18 (14.2) 
 Present24 (15.9)12 (50.0) 9 (37.5) 22 (91.7) 3 (12.5) 

ASSOCIATION OF MARKERS AND PS WITH DISEASE OUTCOME

The 5-year DFS and CSS rates for the 151 patients included in the study were 67% and 78%, respectively. When markers were tested individually, altered Bax showed the strongest association with disease recurrence and cancer-specific mortality (P= 0.02 and P < 0.001, respectively). Kaplan–Meier analyses showed that the total number of altered markers and unfavourable PS were associated with both disease recurrence and bladder cancer-specific mortality (P < 0.001 and P= 0.003, respectively) (Fig. 2).

Figure 2.

A, DFS probabilities in 151 cases treated by RC for SCC of the urinary bladder. P values calculated using the log-rank test. B, CSS probabilities in 151 cases treated by RC for SCC of the urinary bladder. P values calculated using the log-rank test.

Multivariable Cox proportional hazards regression analyses that adjusted for the effects of pathological tumour stage, grade, LN metastasis and LVI (Table 3) showed that unfavourable PS was an independent predictor of disease recurrence (HR 2.694, 95% CI 1.386–5.235, P= 0.003) and bladder cancer-specific mortality (HR 2.868, 95% CI 1.209–6.802, P= 0.017). LVI was the only other predictor of cancer-specific mortality (HR 3.464, 95% CI 1.299–9.238, P= 0.013).

Table 3.  Multivariable Cox proportional hazards regression analysis addressing disease recurrence and cancer-specific mortality in 151 cases treated by RC for SCC of the urinary bladder
All patientsDisease recurrenceCancer-specific mortality
HR95% CI P HR95% CI P
T stage      
 ≤T2ReferentReferent
 >T22.0650.266–16.0270.2851.6580.012–3.0170.560
Grade1.4220.673–3.0020.3561.2350.495–3.0800.651
LN+1.7760.844–3.7350.1301.6320.644–4.1340.302
LVI1.9110.844–4.3290.1203.4641.299–9.2380.013
PS2.6941.386–5.2350.0032.8681.209–6.8020.017

DISCUSSION

Bladder cancer is the most common cancer in Egypt and there is a high incidence of SCC attributable to schistosomiasis [2,4,5], but the mechanisms of squamous cell carcinogenesis in the urinary bladder are still not fully understood. Dysregulation of apoptosis is one of the important pathways associated with carcinogenesis. The prognostic role of combined alterations of apoptotic markers has been shown in UCB [6]. We have previously defined a PS that predicts outcome in UCB based on the number of altered markers (favourable if ≤2 biomarkers were altered; unfavourable if >2 biomarkers were altered) [24]. In the present study, an unfavourable PS was associated with advanced stage of SCC at RC and was an independent predictor of disease recurrence and bladder cancer-specific mortality after RC. Kaplan–Meier analyses showed a significant difference in oncological outcomes based on the number of altered markers and the PS. Patients with an unfavourable PS had ≈ 2x the chance of disease recurrence after RC. Multivariate analyses showed that an unfavourable PS is associated with an ≈ 2.5-fold risk of disease recurrence and cancer-specific mortality compared with a favourable PS.

One or more markers were altered in almost every patient included in the study showing the importance of molecular alterations in tumorigenesis and progression of SCC. CC-3 was the most frequently altered marker (≈90%) and its altered expression was defined based on a decreased expression that has been associated with a higher probability of disease recurrence after RC for UCB [6]. CC-3 alterations were associated with high stage and grade of SCC in the present study, similar to UCB, but SCC showed more frequent CC-3 alterations than previously reported in UCB [6]. Activated caspase-3 constitutes an important downstream step in both the intrinsic and extrinsic apoptotic pathways. In addition to their direct proteolytic activity, effector caspases like caspase-3 can activate initiator caspases, amplifying the original signal [6,7].

By constrast to other bladder SCC studies, p53 was less frequently altered in this study (≈33% of patients vs ≈ 70% patients in other studies) [21,23]; however, we recently reported p53 to be the only prognostic marker for SCC among a panel of cell cycle regulatory markers [19]. The clinical significance of p53 alterations and their role as an adjunct to classic pathological features had also been identified in previous bladder SCC studies [14,20]. Generally, p53 remains the most frequently altered and studied marker in cancer. It has been associated with different molecular pathways of carcinogenesis including apoptotic pathways in BC [25]. It is a key player in genetic stability and DNA repair and its role cannot be fully interpreted in an isolated manner [9,19].

Interestingly, p53, CC-3 and COX-2 alterations were associated with each other in the present study. The relationship between p53 and COX-2 was suggested previously in many studies that showed that p53 alterations can increase the level of COX-2 expression in malignant tissue via cytokines, oncogenes and growth factors [26,27]. Similarly to the diverse network of p53 [9],COX-2–mediated prostaglandin (PG) signalling pathways can promote tumour growth in several ways, including inhibition of apoptosis, enhancement of cellular proliferation, promotion of angiogenesis, stimulation of invasion/motility and suppression of immune responses. Bcl-2–mediated pathways and Akt signalling may be involved in the downstream effects of COX-2 on cell proliferation and apoptosis [28,29]. Activation of COX-2 increases PG-E2 and decreases arachidonic acid. PG-E2 can increase resistance to apoptosis by increasing the production of anti-apoptotic proteins like Bcl-2, and decreased arachidonic acid inhibits apoptosis through its effect on caspase-3 [28,29]. We and others have previously shown the prognostic role of COX-2 in SCC and its association with schistosomiasis – associated BC grade and stage [18,30,31].

Bax is an apoptosis-inducing member of the Bcl-2 family, which can be activated through p53, and it has been found to play important prognostic role in many cancers [9]. [12,13], Bax was the least frequently altered marker in the present study, which is probably why it did not correlate with clinico-pathological features. Despite being infrequently altered, it was associated with an increased risk of disease recurrence and cancer-specific mortality.

The present study further confirms the distinctive features reported previously in SCC, especially in SCC caused by schistosomiasis: young age, low grade and high stage [4,5]; however, we were unable to define a prognostic role for the previously reported prognostic pathological features, namely T stage, grade and LN metastasis [4]. LVI was the only pathological factor associated with cancer-specific mortality, consistent with previous studies on BC [18,19].

The present study is a step toward understanding the biological behaviour of SCC of the urinary bladder. The PS, based on the number of altered apoptotic markers, was found to be a useful tool for predicting the aggressiveness of SCC and poor outcome after RC. Our findings support the need for further evaluation of apoptotic pathways in bladder SCC.

The present study has some limitations, particularly with regard to retrospective data collection and the variability of immunohistochemistry techniques. To limit this variability, we used a combination of automated methods and an experienced pathologist to confirm all immunohistochemistry readings.

In conclusion, apoptosis is an important pathway altered in squamous cell carcinogenesis and progression. Alterations in apoptotic markers can improve prediction of oncological outcomes after RC for SCC. Patients with an unfavourable PS might be considered for adjunct therapies.

ACKNOWLEDGEMENTS

The patient care was carried out in the Urology and Nephrology Centre of Mansoura University, Egypt. The construction of tissue microarray blocks was completed at the immunohistochemistry laboratory of University of Southern California. The immunohistochemical staining and reading was performed at University of Texas Southwestern Medical Centre, Dallas, Texas, USA.

CONFLICT OF INTEREST

None declared. This study was supported by a grant from the Egyptian Ministry of Higher Education via the Egyptian Cultural and Educational Bureau (ECEB) in Washington DC.

Ancillary