- © 2004 by American Society of Clinical Oncology
p53, p21, pRB, and p16 Expression Predict Clinical Outcome in Cystectomy With Bladder Cancer
- Shahrokh F. Shariat,
- Hideo Tokunaga,
- JainHua Zhou,
- JaHong Kim,
- Gustavo E. Ayala,
- William F. Benedict and
- Seth P. Lerner
- From the Scott Department of Urology and Department of Pathology, Baylor College of Medicine; the Methodist Hospital; and the Department of Genitourinary Medical Oncology, The University of Texas M.D. Anderson Cancer Center, Houston; and the Department of Urology, University of Texas Southwestern Medical Center, Dallas, TX.
- Address reprint requests to Seth P. Lerner, MD, Scott Department of Urology, Baylor College of Medicine, 6560 Fannin, Suite 2100, Houston, TX 77030; e-mail: slerner{at}bcm.tmc.edu
Abstract
Purpose To determine whether p53, p21, pRB, and/or p16 expression is associated with bladder cancer stage, progression, and prognosis.
Patients and Methods Immunohistochemical staining for p53, p21, pRB, and p16 was carried out on serial sections from archival specimens of 80 patients who underwent bilateral pelvic lymphadenectomy and radical cystectomy for bladder cancer (median follow-up, 101 months).
Results p53, p21, and pRB or p16 expression was altered in 45 (56%), 39 (49%), and 43 (54%) tumors, respectively. Sixty-six patients (83%) had at least one marker altered, and 21 patients (26%) had all three altered. Abnormal expressions of p53, p21, and pRB/p16 expression were associated with muscle-invasive disease (P = .007, P = .003, and P = .003, respectively). The alteration of each marker was independently associated with disease progression (P ≤ .038) and disease-specific survival (P ≤ .039). In multivariable models that included standard pathologic features and p53 with p21 or p53 with pRB/p16, only p53 and lymph node metastases were associated with bladder cancer progression (P ≤ .026) and death (P ≤ .028). In models that included p21 and pRB/p16, only p21 and lymph node metastases were associated with bladder cancer progression (P ≤ .022) and death (P ≤ .028). In a model that included the combined variables p53/p21 and pRB/p16, only p53/p21 and lymph node status were associated with bladder cancer progression (P ≤ .047) and death (P ≤ .036). The incremental number of altered markers was independently associated with an increased risk of bladder cancer progression (P = .005) and mortality (P = .007).
Conclusion Although altered expression of each of the four cell cycle regulators is associated with bladder cancer outcome in patients undergoing radical cystectomy, p53 is the strongest predictor, followed by p21, suggesting a more pivotal role of the p53/p21 pathway in bladder cancer progression.
INTRODUCTION
Mutations of cell cycle regulatory genes are the genetic alterations most commonly found in human neoplasia, including bladder cancer [1-4]. Alterations in the p53 and retinoblastoma (RB) tumor suppressor genes are a predominant component in the development of transitional cell carcinoma (TCC) of the bladder, but the downstream pathways that contribute to urothelial transformation are not completely defined [5,6]. Recently, a new family of negative cell cycle regulators, the cyclin-dependent kinase inhibitor genes INK4 and KIP, has been identified. p21WAF1/CIP1 is both a p53-inducible and p53-independent cyclin-dependent kinase inhibitor that can arrest the cell by inhibiting DNA replication [7]. Loss of p16 expression, a protein encoded by the INK4 family member INK4A, has been associated with hyperphosphorylation of RB [8]. In addition to p16 protein, INK4A encodes for a second product, termed p19ARF, which has been shown to interact with mdm2 and to block mdm2-induced p53 degradation and transactivational silencing [9,10]. We and others have reported that alteration of p53, pRB, p16, and p21 expression have prognostic significance in patients treated with radical cystectomy for bladder cancer [4,5,7,8,11-15]. Combinations of independent, complementary markers may provide a more accurate prediction of outcome compared with a single marker. For example, patients with p53 wild-type or pRB wild-type tumors have been shown to be at significantly decreased risk of bladder cancer recurrence and mortality compared with patients who have both proteins altered [5,6,16,17].
The aims of the present study were to determine the association of p53, p21, pRB, and p16 expression with pathologic characteristics and clinical outcome of bladder TCC in a well-defined set of patients who underwent radical cystectomy and had long-term follow-up. The availability of the expression status of all four markers in the same set of patients provided a unique opportunity to determine whether alterations in p53, p21, pRB, and p16 expression exert a cooperative or synergistic effect on bladder cancer progression, metastasis, and survival.
PATIENTS AND METHODS
Patient Population
All studies were undertaken with the approval and institutional oversight of the Institutional Review Board for the Protection of Human Subjects at Baylor College of Medicine. We studied 80 consecutive patients with TCC of the urinary bladder who had undergone radical cystectomy and pelvic and iliac lymphadenectomy at two of our private teaching hospitals and for whom we had adequate archival material. Patients were identified from department billing records and the tumor registry. We excluded patients who had received neoadjuvant chemotherapy or radiation therapy. Histology, grade, stage, and presence of carcinoma in situ were confirmed by blinded review of the original pathology slides. The 1997 tumor-node-metastasis system classification was used for pathologic staging, and the WHO classification was used for pathologic grading. No patient was treated preoperatively with either radiation or systemic chemotherapy, and no patient had distant metastatic disease at the time of cystectomy. There were 65 males (81%) and 15 females (19%). The median age was 65 years (range, 41 to 80 years). We collected formalin-fixed, paraffin-embedded blocks representing the most invasive areas of each tumor. The median follow-up was 101 months (range, 62.4 to 177.3 months) for those patients alive at the time of analysis. When a patient died, the cause of death was determined by the treating physician, by chart review corroborated by death certificate, or by death certificate alone. All patients who were identified as having died of bladder cancer had progressive, widely disseminated, and often highly symptomatic metastases at the time of death.
Immunohistochemistry and Scoring
We performed p53, p21, pRB, and p16, immunohistochemical staining using serial sections from the same paraffin-embedded blocks. Staining and scoring protocols for p53 (anti-p53-monoclonal antibody, DO-7; Dako Corp, Carpinteria, CA) [18], p21 (clone designation-Waf1[Ab1]; Oncogene, Cambridge, MA) [7,17], pRB (anti-RB polyclonal antibody, RB-WL-1) [8], and p16 (anti-p16 monoclonal antibody, NCL-p16, clone DCS-50; Vector Laboratories, Burlingame, CA) [8] have been described elsewhere. All markers were placed in one of two categories, altered or normal, by at least two investigators who were blinded to clinical outcome. Nuclear p53 immunoreactivity was considered altered when samples demonstrated at least 10% nuclear reactivity because it has been shown that accumulation of p53 protein in 10% or more of the tumor-cell nuclei strongly correlates with mutations in the p53 gene [4,19]. Borderline p53-positive–stained tumors were re-evaluated, and percentage of cells staining positive was verified using the Quantitative Proliferation Index Analysis Program in the CAS 200 (Becton Dickinson, Franklin Lakes, NJ). Tumors were classified as p21 negative, defined as those tumors with no detectable or only very low levels of p21 nuclear immunoreactivity (0% to 10% of tumor cells showing p21 expression, p21 altered status), or p21 positive (p21 wild-type status), defined as those tumors with moderate to high levels of p21 nuclear immunoreactivity (> 10% of tumor cells showing p21 expression). pRB and p16 immunoreactivity were assigned to one of the following three categories of nuclear staining in the tumor cells: no nuclear reactivity; normal heterogeneous; and strong homogeneous (> 50%). Tumors with no pRB expression and those with a strong homogeneous staining pattern were categorized as having altered pRB status because it has been shown that overexpression of pRB in bladder cancer is also indicative of dysfunctional pRB status [8]. Similarly, tumors without or with overexpression of p16 were categorized as altered because p16 expression is directly related in an inverse manner to RB protein expression in bladder cancer [8].
Statistical Analyses
For purposes of analysis, tumor pathologic stage (≤ pT1 v ≥ pT2), grade (grades 1 and 2 v grade 3), and lymph node status (N0 v N1 and N2) were evaluated as dichotomized variables. The Fisher's exact test and the χ2 test were used to evaluate the association between p53, p21, pRB, p16, pathologic stage, pathologic grade, lymph node status, and lymphovascular invasion. The concordance rate was determined using the following equation: (number of cases with wild type or altered for both markers ÷ total number of cases) × 100. Differences in variables with a continuous distribution across ranked categories were assessed using the Mann Whitney U test and the Kruskal-Wallis nonparametric analysis of variance, respectively. The Kaplan-Meier method was used to calculate survival functions, and differences were assessed with the log-rank test. Multivariable survival analysis was performed with the Cox proportional hazard regression model. The lowest quartile was used as the reference category when calculating the hazard ratios. Statistical significance in this study was set at P < .05. All reported P values are two-sided. All analyses were performed with the SPSS statistical package (SPSS version 10.0 for Windows; SPSS, Inc, Chicago, IL).
RESULTS
Expression of p53, p21, pRB, and p16 in Normal Bladder and Bladder TCC
Normal bladder epithelium showed heterogeneous pRB (wild type), heterogeneous p16 (wild type), absent p53 (wild type), and positive p21 (wild type) immunoreactivity. p53 and pRB staining patterns were exclusively nuclear. In most cases with p16 nuclear staining, p16 staining was also observed in the cytoplasm, especially in those tumors with strong p16 immunoreactivity. p16-negative tumors, however, exhibited no cytoplasmic p16 staining.
Of the 80 tumors studied, 45 (56%) had altered expression of p53, and 39 (49%) had altered expression of p21 (Table 1). Thirty-seven tumors (46%) had normal heterogeneous pRB staining; 20 tumors (25%) had no nuclear reactivity, and 23 tumors (29%) had strong homogeneous RB staining. We observed heterogeneous positive p16 expression in 38 tumors (47%), strong positive staining in 20 tumors (25%), and negative staining in 22 tumors (28%). All tumors, except one, with heterogeneous p16 staining also had a heterogeneous pRB staining pattern. Tumors that were pRB negative consistently showed strong p16 staining. There was an inverse relationship between pRB and p16, with all tumors with strong homogeneous pRB staining, except one, showing absence of p16 staining. Altered pRB status was thus observed in 43 tumors (54%), and altered p16 status was observed in 42 tumors (53%). pRB staining was strongly concordant with p16 expression (concordance rate, 99%; P < .001). For further analyses, pRB and p16 staining were considered as a combined variable, and the discrepant case was categorized as altered for the pRB/p16 pathway.
Association of the Expression of p53, p21, and pRB/p16 With Clinicopathologic Characteristics and Bladder Cancer Progression and Survival Estimates of 80 Patients Who Underwent Radical Cystectomy for TCC of the Urinary Bladder
Fourteen (18%) of the 80 tumors exhibited wild-type status of all markers. Twenty-six (33%) of 80 tumors showed alteration of one marker, 19 (24%) of 80 tumors showed alteration of two markers, and 21 (26%) of 80 tumors showed alterations of all three markers.
Association of p53, p21, and pRB/p16 As Individual Variables With Pathologic Characteristics
Molecular and pathologic characteristics of the 80 cystectomy patients and association with p53, p21, and pRB/p16 expression are shown in Table 1. The expression of p53 and pRB/p16 were associated with muscle invasive disease (P = .007 and P = .003, respectively) but not presence of carcinoma in situ, pathologic grade, lymph node metastases, or lymphovascular invasion. There was a moderate but statistically significant concordance (both markers altered or normal) between p53 and p21 immunoreactivity (concordance rate, 65%; P = .008) and p53 and pRB/p16 (concordance rate, 63%; P = .042). Age at time of cystectomy was not different between patients who were p53 (P = .708), p21 (P = .139), and pRB/p16 (P = .223) positive and negative.
Association of p53, p21, and pRB/p16 As Individual Variables With Disease Outcomes
Disease progression occurred in 30 patients, and 49 patients were dead at the time of analysis. Of these 49 patients, 25 patients died of metastatic bladder cancer, and 24 patients died of other causes without evidence of disease progression. With a median follow-up of 101.1 months, Kaplan-Meier analyses showed that the altered expression of p53, p21, and pRB/p16 were each significantly associated with an increased probability of tumor progression (P = .004, P = .013, and P = .007, respectively; Table 1) and survival from bladder cancer (P = .005, P = .019, and P = .008, respectively; Table 1).
In multivariable Cox proportional hazards regression analyses that adjusted for the effects of pathologic stage, grade, lymphovascular invasion, lymph node status, and one of the three markers, each marker was independently associated with disease progression (P ≤ .038) and disease-specific survival (P ≤ .039). Similarly, in the subgroup of patients with muscle-invasive disease, all three markers were independently associated with bladder cancer progression (P ≤ .035) after adjusting for the effects of pathologic grade, lymphovascular invasion, and lymph node status. However, only p53 and p21 were independently associated with bladder cancer–specific survival (P = .003 and P = .034, respectively) but not pRB/p16 (P = .083).
When p53 and standard pathologic features were modeled with either p21 or pRB/p16, p53 and lymph node metastases were the sole predictors of bladder cancer progression (P ≤ .026) and death (P ≤ .028). When p21 was modeled with pRB/p16 and standard pathologic features, only p21 and lymph node metastases were associated with bladder cancer progression (P = .038 and P = .022, respectively) and death (P = .041 and P = .028, respectively).
Association of p53, p21, and pRB/p16 As Combined Variables With Disease Outcomes
We repeated the analyses using the following three potential combinations of combined variables: p53 and pRB/p16, p53 and p21, and pRB/p16 and p21. We created three categories within each combination (both wild type, one altered and one wild type, and both altered). Kaplan-Meier analysis demonstrated that patients could be stratified as low, medium, and high risk for progression and death based on the combined status of p53 and pRB/p16 (Fig 1A and 1B) and p53 and p21 (Fig 2A and 2B). However, pRB/p16 and p21 only stratified patients into low- and high-risk groups (Fig 3A and 3B). The risk of disease progression and mortality did not differ between patients with no alterations of pRB/p16 and p21 and patients with alterations of either pRB/p16 or p21.
(A) Bladder cancer progression and (B) mortality in 80 patients who underwent radical cystectomy according to combined p53 and pRB/p16 status. wt, wild type; alt, alteration.
(A) Bladder cancer progression and (B) mortality in 80 patients who underwent radical cystectomy according to combined p53 and p21 status. wt, wild type; alt, alteration.
(A) Bladder cancer progression and (B) mortality in 80 patients who underwent radical cystectomy according to combined pRB/p16 and p21 status. wt, wild type; alt, alteration.
These findings were confirmed by univariable Cox proportional hazards regression analysis. The combined p53 and pRB/p16 status was associated with bladder cancer progression (P for trend = .005) and survival (P for trend = .008). Similarly, p53/p21 status was associated with bladder cancer progression (P for trend = .009) and survival (P for trend = .015). In contrast, although overall pRB/p16 and p21 status was associated with bladder cancer progression and survival (P for trend = .004 and P for trend = .005, respectively), this association was significant only for patients with altered expression of both pRB/p16 and p21 (P = .005 and P = .011, respectively) and not for patients with altered expression of only one variable (P = .451 and P = .765, respectively).
Cox Proportional Hazards Models of Combined Variables Incorporating Pathologic Features
The combined variables p53 and pRb/p16, p53 and p21, and pRB/p16 and p21 were then analyzed in separate multivariable Cox proportional hazards regression analyses that adjusted for effects of lymphovascular invasion, lymph node metastases, pathologic grade, and stage. The combination of p53 and pRB/p16 status was associated with disease progression (P for trend = .003) and bladder cancer survival (P for trend = .003). Similarly, the combined p53 and p21 status was associated with disease progression (P for trend = .004) and bladder cancer–specific survival (P for trend = .006). Again, although the overall pRB/p16 and p21 status was associated with bladder cancer progression and survival (P for trend = .007 and P for trend = .006, respectively), only patients with altered expression of both pRB/p16 and p21 (P = .007 and P = .015, respectively), but not patients with altered expression of only one (P = .430 and P = .818, respectively), were at increased risk for disease progression and cancer-related death. In a model that included both p53/p21 and pRB/p16 (model 1, Table 2), only p53/p21 status and lymph node status were independently associated with bladder cancer progression (P = .003 and P = .047, respectively) and survival (P = .002 and P = .036, respectively).
Multivariable Cox Regression Analyses of Pathologic Features and Molecular Markers for the Prediction of Disease Progression and Disease-Specific Survival of 80 Patients Treated With Radical Cystectomy for TCC of the Bladder
Association of the Number of Altered Markers As Categoric Variable With Disease Outcomes
Twenty-nine (44%) of the 66 patients who had altered expression of at least one cell cycle regulator experienced disease progression, and 25 (38%) of the 66 patients died of bladder cancer. Only two (14%) of 14 patients with wild-type status of all three markers experienced disease progression after cystectomy. Seven (27%) of 26 patients with one marker altered, eight (42%) of 19 with two markers altered, and 14 (67%) of 21 with all three markers altered experienced disease progression. Kaplan-Meier analysis demonstrated that patients could be stratified into different risk groups for bladder cancer progression and outcome for survival based on the number of markers altered (Fig 4A and 4B). When analyzed as categoric variables in univariable Cox proportional hazards regression analyses, the risk of experiencing bladder cancer progression increased incrementally in proportion to the number of markers altered (P for trend = .007; one altered v none, P = .219; two altered v none, P = .034; and three altered v none, P = .006) as did decrease in survival (P for trend = .009; one altered v none, P = .125; two altered v none, P = .049; and three altered v none, P = .014). In a multivariable model that adjusted for the effects of lymphovascular invasion, pathologic grade, and stage (model 2, Table 2), the incremental number of altered markers and lymph node involvement were associated with an increased risk of bladder cancer progression (P for trend = .005 and P = .048, respectively) and decreased survival (P for trend = .007 and P = .028, respectively).
(A) Bladder cancer progression and (B) mortality in 80 patients who underwent radical cystectomy according to number of altered markers. wt, wild type; alt, alteration.
DISCUSSION
We confirmed that altered expression of p53, pRB/p16, and p21 is associated with an increased risk of bladder cancer progression and death after adjusting for the effects of pathologic stage, grade, lymphovascular invasion, and lymph node metastases [4,5,7,8,11-15]. This is the first study, to our knowledge, to investigate all four cell cycle regulators in serial tumor sections from a single, well-defined cohort of bladder TCC patients who underwent radical cystectomy and had long-term follow-up. Although all four cell cycle regulators were independently associated with bladder cancer outcomes, we identified definite distinctions in the associations of these markers with pathologic features of the local tumor and disease prognosis. For example, p21 was the sole cell cycle regulator not associated with pathologic stage. Alteration of p21 occurred in more than half of pT1 bladder tumors and remained equally frequent throughout all stages. In contrast, in concordance with our previous findings and the findings of others, alterations of p53 and pRB/p16 were associated with muscle-invasive disease [4,11-16,20]. These differences in the proportion of alterations along the progression model may reflect the role of each cell cycle regulator and their cumulative effect in the steps associated with bladder cancer development and progression.
We found that p53 was the strongest predictor of disease outcomes in postoperative multivariable models that adjusted for the effects of other cell cycle regulators and standard pathologic features, followed by p21. In a standard multivariable model that included the combined pRB/p16 and p53/p21 variables, only p53/p21 was associated with bladder cancer progression and decrease in overall survival. It should be mentioned, however, that this study truly examines only three cell cycle regulators with partially independent biologic activity. As previously shown [8], the p16 and pRB expression patterns were almost completely concordant, suggesting that they are reflecting the same pathway activity, and, therefore, do not add independent statistical and prognostic information to each other. In addition, strong homogeneous staining of p16 or pRB is a direct reflection of the loss of function of the other. Nevertheless, these findings suggest a more critical role for the p53 tumor suppressor checkpoint pathway than the pRB/p16 tumor suppressor checkpoint pathway in the clinically relevant progression of bladder cancer. Moreover, p21WAF1/CIP1 can also be upregulated in a p53-dependent and -independent manner by various growth factors and regulators of cell differentiation [21]. The sample size may also have limited our ability to detect small differences attributed to the other variables.
p53 is an established tumor suppressor gene that sits at the crossroad of complex cellular signaling networks integrating information from a wide range of molecules that participate in the control of numerous cellular processes. Aberrations of p53 expression are known to play a central role in bladder cancer oncogenesis, resulting in loss of p53 control over apoptosis, cell cycle progression, and transcription of genes involved in identification of DNA damage. Despite being the strongest predictor of bladder cancer outcomes, only 51% of patients with p53 overexpression experienced disease progression, and as many as 23% of patients with p53 wild-type status also experienced disease progression. In addition, although 44% of patients with p53 overexpression died from bladder cancer at 8.5 years after surgery, 17% of patients with p53 wild type also died of their disease. These data, together with the current paradigm that bladder cancer develops along multiple molecular pathways, suggest that including alterations of additional markers in models designed to predict outcomes after cystectomy could enhance the predictive power of single markers.
We also confirmed that p53 and p21, as well as p53 and pRB/p16, had cooperative and synergistic effects on bladder cancer outcomes, which stratified patients into high-, intermediate-, and low-risk groups [5,6,16,17]. Patients with tumors exhibiting altered expression of both p53 and p21, representing 35% of patients, had the poorest progression and survival rates, whereas patients who had tumors exhibiting wild-type status for both p53 and p21 had an encouraging clinical outcome. Patients who had alterations of either p53 or p21 had a significantly worse prognosis than patients with wild-type p53 and p21, but they had a better prognosis than patients with alterations of both p53 and p21. Similarly, patients possessing altered p53 and pRB/p16 were at significantly higher risk for unfavorable clinical outcome than patients with only one of the two markers altered.
Although pRB/p16 and p21 also had cooperative and synergistic effects on disease progression and cancer-specific survival, they only stratified patients into two prognostic groups. Patients with alterations of both p21 and pRB/p16 expression had a poorer prognosis and lower disease-specific survival than patients with both wild-type pRB/p16 and p21 who had outcomes comparable to patients with alterations of only one of the two markers. Recent studies, which have shown that RB is essential for G1/S arrest in response to DNA damage, have provided a potential link between pRB and p21. p21 is induced transcriptionally on treatment of cells with a variety of DNA-damaging agents and is believed to mediate G1/S arrest in DNA-damage–induced checkpoint control [22,23]. Interestingly, ionizing irradiation inhibits the specific phosphorylation of RB by cdk2 in a p21-dependent manner [24], suggesting that pRB is a component of the p21-induced checkpoint control pathway. In addition, p21 acts to prevent pRB phosphorylation by inhibiting activation of cyclinE/cdk2 complexes that are required for pRB phosphorylation [25]. p21 also acts as an assembly factor for cyclinD1/cdk4 complexes at low concentrations and is required for the cyclinD1/cdk4 enzymatic activity necessary for pRB phosphorylation [26].
Finally, we found that the expression of cell cycle regulators is commonly altered in bladder TCC patients, with 83% of patients having altered expression of at least one of the three regulators and with 26% exhibiting altered expression of p53, p21, and pRB/p16. Higher total numbers of altered markers were associated with an increased risk of experiencing disease progression and death after data were adjusted for the effects of standard pathologic features. Of the 66 patients who had altered expression of at least one cell cycle regulator, 44% experienced disease progression, and 38% died of bladder cancer. Of the patients who had altered expression of all three markers, 67% experienced disease progression, and 62% died of bladder cancer. Although the hazard ratios of experiencing disease progression and dying of bladder TCC were approximately three to four times greater for patients with alterations of one cell cycle regulator than for those with wild-type status of all three markers, this difference was not statistically significant when adjusted for the effects of standard pathologic features. In contrast, patients with alterations of two or three cell cycle regulators had significantly higher hazard rates (range, 5 to 16) for disease progression and mortality than patients with wild-type status of all cell cycle regulators after adjusting for the effects of standard pathologic features. These findings suggest that alterations of at least two of the three cell cycle regulators are needed to confer a statistically and prognostically substantial increase in risk of cancer progression and death after radical cystectomy beyond the information available from standard pathologic features.
In conclusion, the p53 phenotype seems to be the strongest predictor of bladder cancer outcome in patients undergoing radical cystectomy. The association of the p53/p21 tumor suppressor pathway with bladder cancer outcome was stronger than the association of the pRB/p16 tumor suppressor pathway with bladder cancer outcome. Combining markers that work at least partly through independent pathways increases the prognostic accuracy for the individual patient. In concordance with recent in vitro evidence, we found that combining pRB/p16 with p21 staining status has a cooperative and synergistic effect, stratifying patients into two risk groups. The role of cell cycle regulators in bladder cancer progression seems to be a complex accumulation of genetic alterations, from which p21 seems to be associated with the early stages of bladder cancer clinical progression and p53 and pRB/p16 seem to be associated with the later stages of bladder cancer clinical progression. The mechanisms for these interactions remain to be elucidated. These findings should be evaluated in larger, multicenter, prospective trials and considered as stratification variables in clinical trials involving patients with invasive bladder cancer treated with cystectomy.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
- Received March 19, 2003.
- Accepted September 2, 2003.
