The effect of p53 protein expression and MIB-1 proliferative activity on survival and chemotherapeutic response in patients with lymph node (LN)-positive transitional cell carcinoma (TCC) of the urinary bladder remains unclear. The objective of this study was to assess the ability of these markers to predict disease-associated outcomes and response to chemotherapy in a cohort of patients with LN-positive TCC.
The authors examined the expression of p53 and MIB-1 in the LN metastases from 139 patients who underwent cystectomy for TCC at their institution. P53 and MIB-1 nuclear staining were quantified using an image-analysis system. Cox proportional hazards regression models were used to test associations of these markers with death from TCC, distant metastases, and local recurrence for all patients and in the subset of patients who were treated with adjuvant chemotherapy.
The median p53 and MIB-1 indices were 45.2% and 30.3%, respectively. The median follow-up was 4.5 years (range, 0.1–10 years). There were no statistically significant associations noted between the p53 and MIB-1 indices and the outcomes studied. When the analysis was limited to patients who were treated with adjuvant chemotherapy (n = 37 patients), the p53 index was found to have no prognostic value; however, there was a significant association between MIB-1 and distant metastases (P = 0.049). When disease-specific survival rates were stratified according to p53 index and chemotherapy, patients exhibited a response to chemotherapy regardless of p53 index.
Patients with carcinoma of the urinary bladder who have regional involvement of lymph nodes are at high risk for disease progression. Recent evidence indicates that platinum-based chemotherapy in an adjuvant setting prolongs survival.1, 2 There is a need for better indicators of failure after cystectomy. These indicators can be used to develop multivariate models for predicting disease progression of bladder carcinoma and stratifying patients for adjuvant chemotherapy.
The tumor suppressor gene p53 is one of the most frequently altered genes in human malignancies. The gene is located on the short arm of chromosome 17 and functions to suppress cell proliferation and to direct DNA-damaged cells toward apoptosis. p53 mutations are associated with genomic instability and the development of carcinoma. Previous studies have shown a correlation between increased p53 immunohistochemical staining and increased recurrence rates as well as reduced survival in superficial3 and invasive4 bladder carcinomas. However, to our knowledge, the prognostic effect of abnormal p53 protein expression in patients with lymph node-positive transitional cell carcinoma (TCC) of the urinary bladder who undergo surgery with or without adjuvant chemotherapy is unknown. In addition, there are conflicting data regarding the effect of positive p53 status on response to chemotherapy.5–7
Ki-67 is a nuclear protein that is part of the DNA replicase complex encoded by a gene located on chromosome 10. The gene is expressed in the active phases (G1, S, and G2) of the cell cycle, making this protein a marker of cell proliferation. MIB-1 is a murine monoclonal antibody that is used to detect Ki-67. MIB-1 immunostaining has been correlated with increased recurrence rates and reduced survival in patients with superficial and invasive TCC.8–12 However, similar to the p53 protein, the significance of MIB-1 expression in lymph node-positive TCC is largely unknown. The objectives of the current study were to determine whether p53 and MIB-1 protein expression have prognostic value and to assess their ability to predict response to chemotherapy using a large cohort of patients who had TCC of the urinary bladder metastatic to regional lymph nodes and who underwent cystectomy.
MATERIALS AND METHODS
One hundred eighty-six patients underwent a radical or partial cystectomy or an anterior exenteration for lymph node-positive TCC of the urinary bladder between 1970 and 1998 at the study institution. Of these, 154 patients underwent conventional pelvic lymphadenectomy, including the obturator and iliac lymph nodes. Patients with evidence of prior or concurrent systemic disease (pM1) were excluded. Fifteen patients with inadequate tissue blocks for immunohistochemical staining also were excluded, leaving 139 patients for the analysis.
MIB-1 tumor proliferation and p53 protein expression were determined by digital image analysis with a CAS 200 image analyzer (Bacus Laboratories, Lombard, IL) using 6-μm-thick sections of paraffin embedded tissue blocks from the largest lymph node metastasis. All analyses were performed by board-certified cytotechnologists who were trained specifically to perform digital image analysis. MIB-1 and p53 expression levels were assessed with the nuclear antibodies MIB-1 (Immunotech, Marseilles, France; 1:400 dilution) and DO-7 (Dako Corporation, Carpinteria, CA; 1:200 dilution), respectively, using a standard avidin-biotin complex method. The slides were processed with an automatic stainer (BioTek®, Ventana Medical Systems, Tucson, AZ), and diluted hematoxylin was used as the nuclear counterstain. The MIB-1 index was expressed as the percentage of the nuclear area that was positive for MIB-1, as measured by the Quantitative Proliferation Index program (Bacus Laboratories). The total optical density of the nuclei expressing the antigen that reacted with MIB-1 (identified with the brown chromogen diaminobenzidine) was divided by the total optical density of all measured nuclear images at 40× magnification (identified with diluted hematoxylin). The p53 index was evaluated in the same manner. The highest index values for MIB-1 and p53 were used for patients who had more than one tissue block available for analysis. The correlation between p53 and MIB-1 was assessed using a Spearman rank correlation coefficient.
Disease-specific survival after cystectomy was estimated using the Kaplan–Meier method, and log-rank tests were used to assess differences in disease-specific survival between patient groups. Univariate Cox proportional hazards regression models were fit to assess the associations of the p53 and MIB-1 indices with death from TCC, distant metastases, and local recurrence. P53 and MIB-1 were analyzed as continuously scaled variables. In addition, the p53 index was analyzed using the commonly cited cut-off level of 20%. The correlations of the p53 and MIB-1 indices with outcomes were summarized with risk ratios and 95% confidence intervals. P values < 0.05 were considered statistically significant.
Among the 139 patients studied, 69 patients died from TCC at a median of 1.3 years (range, 0.2–6.9 years) after they underwent cystectomy. Of the 37 patients who remained alive at last follow-up, the median time from cystectomy to last follow-up was 4.6 years (range, 0.1–10.3 years). Seventy-eight patients developed metastases at a median of 0.7 years (range, 0.1–6.4 years) after cystectomy. Twenty-four patients experienced local recurrence at a median of 0.7 years (range, 0.1–5.8 years) after cystectomy. A summary of the p53 and MIB-1 indices is shown in Table 1. The Spearman rank correlation coefficient for the association between the p53 and MIB-1 indices was 0.04 (P = 0.672). The univariate associations of the p53 and MIB-1 indices with death from TCC, distant metastases, and local recurrence are summarized in Table 2. The indices for p53 and MIB-1 were not associated significantly with disease-related outcomes in a univariate setting.
Table 1. p53 and MIB-1 Indices in Patients with Lymph Node-Positive Transitional Cell Carcinoma of the Urinary Bladder
Entire cohort (n = 139 patients)
Adjuvant chemotherapy (n = 37 patients)
SD: standard deviation.
Mean ± SD
45.2 ± 29.5
47.8 ± 29.5
Mean ± SD
30.3 ± 19.8
32.0 ± 20.1
p53 > 20%
Table 2. Univariate Associations with Disease-Related Outcomes in Patients with Lymph Node-Positive Transitional Cell Carcinoma of the Urinary Bladder
Thirty-seven patients (27%) with lymph node-positive TCC were treated with adjuvant chemotherapy. Chemotherapy regimens used in the course of this study are summarized in Table 3. Among these 37 patients, 15 patients died from TCC, 20 patients developed metastases, and 5 patients experienced a local recurrence. The p53 and MIB-1 indices for this subset of patients are summarized in Table 1. The univariate associations of the p53 and MIB-1 indices with death from TCC, distant metastases, and local recurrence for patients who underwent cystectomy and received adjuvant chemotherapy are summarized in Table 2. For each 10-unit increase in MIB-1, the risk of distant metastases increased 21% (P = 0.049).
Methotrexate, vinblastine, doxorubicin, and cisplatin
Taxol and cisplatin
Etoposide and cisplatin
Cytoxan, doxorubicin, and cisplatin
Disease-specific survival stratified by p53 index (low [≤ 20%] vs. high [> 20%]) and adjuvant chemotherapy is shown in Figure 1. Within the chemotherapy-treated and nonchemotherapy-treated groups, no statistically significant difference in disease-specific survival between patients with low and high p53 indices was observed. However, among patients who had a high p53 index, there was a statistically significant difference in disease-specific survival between patients who did or did not receive adjuvant chemotherapy (P = 0.007). A benefit from chemotherapy was observed among patients with low p53, but this difference did not reach statistical significance, most likely because there were only 7 patients with a low p53 index who received adjuvant chemotherapy (P = 0.495).
Patients with TCC of the urinary bladder who have metastases to the regional lymph nodes have a poorer disease-specific survival compared with similarly treated patients who have tumors confined to the urinary bladder. However, some patients with regional lymph node metastases are cured with surgery alone, whereas others benefit from adjuvant chemotherapy in addition to surgery. Recently, in a retrospective study, we observed that patients who had TCC of the bladder with regional lymph node involvement who underwent cystectomy and pelvic lymphadenectomy benefited from adjuvant chemotherapy.13 Patients who received chemotherapy had a 5-year disease-specific survival rate of 54.8%, compared with 32.4% in patients who did not receive chemotherapy. However, currently, there is no marker with which to stratify patients reliably according to risk for disease progression and need for aggressive adjuvant chemotherapy. In addition, there are no predictors of chemotherapeutic response. A reliable predictor of survival would enhance our ability to target patients who are at risk and to reduce the toxicity of treatment to patients who have a low risk of disease progression.
The expression of p53 and MIB-1 proteins has been studied extensively in superficial and invasive bladder carcinomas. These markers have been correlated with disease-specific survival; and, in many studies, increased expression of p53 and MIB-1 is identified as an independent predictor of disease-specific, metastasis-free, and progression-free survival.8–11, 14–17 However, to our knowledge there are limited data regarding the significance of p53 and MIB-1 protein expression in TCC of the urinary bladder metastatic to regional lymph nodes. The objective of the current study was to examine the ability of p53 protein expression and MIB-1 to predict disease-specific outcomes and chemotherapeutic response in a large cohort of surgically treated patients with lymph node-positive TCC of the urinary bladder.
To our knowledge, there are only two studies in the literature examining the importance of p53 in this cohort. Fleshner et al.18 examined 59 patients with TCC of the bladder metastatic to lymph nodes and found that p53 expression was not predictive of outcome. In the only other report of p53 in metastatic TCC, Esrig et al.4 examined 42 patients and reported a 5-year recurrence rate of 91% in patients who had tumors that were considered p53-positive compared with 69% in patients who had tumors that were considered p53-negative (P = 0.054). Both studies were limited by a small sample size.
The current study revealed no statistically significant association between p53 and death from TCC, distant metastases, or local recurrence. It is possible that abnormalities in the p53 gene or protein occur early in urothelial carcinogenesis, and the development of these abnormalities is not related to more aggressive behavior in metastatic TCC. This notion also is supported indirectly by high p53 indices in our study, suggesting that a majority of the clones that spread to lymph nodes have mutated p53. A similar finding was noted by Fleshner et al.18 Other studies have shown a strong correlation between high p53 immunoreactivity and high tumor stage and grade,19–21 again supporting the hypothesis stated above. The behavior of MIB-1 is similar, with a very strong correlation between increasing tissue staining and tumor grade and stage.9
The appropriate cut-off level for p53 or MIB-1 is a subject of much debate. A number of previous studies have used an arbitrary cut-off level of 20%,15, 18, 21, 22 whereas other cut-off levels have been used as well.15, 21 The use of the median number of nuclei stained also has been suggested.9, 17, 22, 23 Currently, to our knowledge, there is no consensus about the ideal cut-off level for assessing p53 or MIB-1 immunohistochemical staining and outcome. We attempted to avoid this potential pitfall by examining p53 protein expression and MIB-1 as continuously scaled variables. In addition, we assessed p53 protein expression using the most commonly cited cut-off level of 20%. Neither approach identified a statistically significant association between disease-associated outcomes and p53 protein expression and MIB-1.
It should be noted that lymph node specimens—not primary tumor specimens—were used for immunohistochemical staining in this study. However, complete correlation between p53 staining in lymph node specimens and primary tumor specimens has been reported previously.18 In addition, we believe that lymph node tumor deposits likely represent the most aggressive clones within the primary tumor and correlate more closely with biologic behavior.
The role of p53 in predicting response to chemotherapy is controversial. Some investigators believe that p53-positive cells are more resistant to chemotherapy,5, 6, 24 whereas others suggest that p53-positive cells are more sensitive to chemotherapy.7 However, some of the more recent studies suggest that patients with p53 nuclear immunoreactivity in the primary bladder tumor do not have a lower response rate or poorer outcome after chemotherapy for metastatic disease compared with patients without p53 nuclear reactivity.25 The role of MIB-1 is even less clear. This issue is of critical importance, because it has been suggested that p53 status may be used to stratify patients for chemotherapy trials and even for bladder preservation.26 The results of the current study suggest that p53 protein expression in lymph node metastasis is not associated significantly with outcome and that chemotherapy appears to be effective regardless of p53 status.
In assessing chemotherapeutic response, the current study was limited by its retrospective design, in which we examined 139 patients who were treated over a span of nearly 3 decades. The chemotherapy regimens used over the course of the study differed significantly because new chemotherapeutic agents were being introduced. However, > 80% of patients received chemotherapy either with combined cisplatin, methotrexate, and vinblastine or with combined methotrexate, vinblastine, doxorubicin, and cisplatin; whereas only small numbers of patients were treated with alternative types of chemotherapy. With newer and more effective chemotherapeutic regimens, an even more pronounced survival advantage may be expected. Furthermore, given the relative rarity of lymph node-positive TCC in surgically treated bladder carcinoma, a single institution is unlikely to have sufficient numbers with which to conduct a prospective study within a reasonable period. To better address the association of p53 and MIB-1 and chemotherapeutic response, a large, prospective, multiinstitutional study will be necessary.
The current study findings confirmed that multiple factors are involved in TCC disease progression and that an intricate combination of genetic events results in disease development and progression. Relying on a single factor to predict behavior is unlikely to yield results. Although p53 undoubtedly is important in the initial stages of TCC, its role in advanced disease likely is augmented by other downstream factors. A combination of markers, rather than a single marker, most likely will provide important prognostic information. Currently, using p53 to stratify patients for chemotherapy appears to be unfounded, especially in the cohort with lymph node-positive TCC.
The role of MIB-1 in predicting response to chemotherapy is encouraging. There was a statistically significant association found between MIB-1 and distant metastases among patients who were treated with chemotherapy. Further studies of MIB-1 alone or in combination with other markers will be needed to substantiate our results.
In conclusion, the data from the current study suggest that p53 has no role in predicting death from TCC in patients with lymph node-positive TCC or in predicting response to chemotherapy. The results of this retrospective study also suggest that chemotherapy appears to be effective regardless of p53 status. The association between MIB-1 and distant metastases in the subset of patients who were treated with adjuvant chemotherapy was found to be statistically significant, suggesting a possible role for MIB-1 in predicting response to adjuvant chemotherapy. Further multiinstitutional, prospective, randomized studies will be required to substantiate our results.