Purpose: Aberrant expression of T-cell coregulatory molecules has been investigated as a mechanism by which certain cancers may evade host immune surveillance. We evaluated expression of the T-cell coregulators B7-H1, B7-H3, and PD-1 in urothelial cell carcinoma (UCC) of the bladder.

Experimental Design: Immunohistochemistry for B7-H1, B7-H3, and PD-1 was done on paraffin-embedded sections from 318 consecutive patients with UCC who underwent radical cystectomy. Expression was correlated with clinicopathologic outcomes and postoperative survival.

Results: B7-H3 was widely expressed in UCC, as 222 of 314 (70.7%) tumors showed positive staining. Expression of B7-H3 in UCC was significantly increased compared with adjacent, nontumor urothelium, as a median of 70% of tumor cells expressed B7-H3, compared with 20% of cells in nontumor specimens (P < 0.001). The increase in B7-H3 expression was independent of tumor stage (P = 0.13). Expression of B7-H1 by UCC tumors (P < 0.001) and PD-1 by tumor-infiltrating lymphocytes (P = 0.012) were significantly associated with increased pathologic stage. Patients who had received intravesical bacillus Calmette-Guerin before cystectomy tended to show increased expression of B7-H3 (P = 0.023) and PD-1 (P = 0.071) but were less likely to express B7-H1 (P = 0.027). Moreover, for the subset of patients with organ-confined disease (n = 167), B7-H1 expression independently predicted all-cause mortality after cystectomy (hazard ratio, 3.18; 95% confidence interval, 1.74-5.79; P < 0.001).

Conclusions: B7-H3 is highly expressed in UCC across tumor stages, whereas B7-H1 and PD-1 expression are associated with advanced disease. B7-H1 expression predicts mortality after cystectomy for patients with organ-confined tumors. These molecules may represent novel diagnostic or prognostic markers, as well as therapeutic targets, for patients with UCC.

Translational Relevance

The data here represent an analysis of >300 patients with urothelial cell carcinoma of the bladder treated with radical cystectomy with >10 years of postoperative follow-up. Because survival rates for patients with muscle-invasive bladder cancer have not appreciably increased in recent years, better methods of risk stratification to characterize those patients who would most benefit from an aggressive multimodal treatment approach, as well as identification novel therapeutic targets, are needed to improve patient outcomes.

We evaluated protein expression of the T-cell coregulatory molecules B7-H1, B7-H3, and PD-1 in urothelial cell carcinoma tumors and tumor-infiltrating lymphocytes using immunohistochemistry. Our results provide the first report of B7-H3 expression in urothelial cell carcinoma, describe the association of B7-H1 and PD-1 expression with localized stage progression, and show the independent prognostic value of B7-H1 for predicting survival in patients with organ-confined bladder cancer. Together, these data indicate potentially important diagnostic, prognostic, and therapeutic implications of T-cell coregulatory molecule expression by urothelial cell carcinoma tumors.

Urothelial cell carcinoma (UCC) of the bladder is the second most common malignancy of the genitourinary tract, and the second most common cause of death among genitourinary tumors, with an estimated 67,160 new cases and 13,750 deaths in 2007 (1). Interestingly, patients with UCC show evidence of an acquired immune dysfunction, particularly affecting lymphocytes (2, 3). In fact, circulating T cells from bladder cancer patients have been found to be unresponsive to polyclonal T-cell activation compared with healthy donor cells (4). Nevertheless, the etiology of UCC-related immune dysfunction remains to be established.

One mechanism that has been described for tumor-associated immune suppression in certain cancers is the aberrant expression of inhibitory T-cell coregulatory molecules by tumor cells (516). Inhibitory T-cell coregulators, which are normally expressed only on lymphoid cells, may down-regulate the antigen-specific T-cell response by inducing T-cell anergy or apoptosis (15). Because T cells play an integral role in mediating antitumoral immunity, expression of inhibitory T-cell coregulators by tumor cells has been postulated as a mechanism by which these cancers may evade the host immune response.

In particular, the B7 family of immune coregulatory proteins contains several ligands that are expressed by various human cancers and have been implicated as inhibitors of T-lymphocyte function (517). We previously showed that clear cell renal cell carcinoma tumors express the T-cell coregulator B7-H1, also known as Programmed Death-Ligand 1 (1012). We further reported that renal cell carcinoma B7-H1–positive tumors are associated with a significantly increased risk of disease progression, cancer death, and overall mortality (1012). Moreover, we noted that expression of the B7-H1 receptor Programmed Death-1 (PD-1) by mononuclear immune cells infiltrating renal cell carcinoma tumors similarly predicted adverse tumor pathology and clinical outcome (5).

Recently, we found that B7-H3, another member of the B7 family, which shares an ∼25% sequence homology with B7-H1 (18), is expressed by normal human prostate epithelial cells, and is expressed even more intensely by prostate cancer cells (19). In addition, we showed that B7-H3 expression correlated with adverse pathologic features in prostate cancers, and with an increased risk of disease recurrence for patients after radical prostatectomy (19).

Given evidence for an immunosuppressive state in bladder cancer patients (24), we thus hypothesized that one mechanism for disease progression in UCC might be impairment of the host immune response through aberrant expression of inhibitory T-cell coregulatory molecules. However, limited data exist on the expression of T-cell coregulatory molecules in UCC to date. A previous study by our laboratory using tissue microarray analysis of an external, multi-institutional cohort of bladder cancer patients showed that B7-H1 expression was associated with advanced tumor stage (7). This association of B7-H1 expression with stage progression was confirmed by a separate investigation of 65 patients with UCC, which included cancers of the renal pelvis, ureter, and bladder (6). Moreover, expression of the B7-H1 receptor PD-1 on tumor-infiltrating lymphocytes (TIL) in UCC was noted in a subset of 13 randomly selected patients from this study (6). The expression of B7-H3 has not, to our knowledge, been reported in human bladder cancer.

Herein, we examined B7-H1, B7-H3, and PD-1 protein expression in a large, consecutive series of patients with UCC treated with radical cystectomy at our institution. We report that B7-H3 expression is significantly up-regulated in UCC compared with noncancerous urothelium across all pathologic tumor stages. In addition, we validate the association of B7-H1 expression with stage progression in bladder cancer and extend these findings to show that expression predicts postoperative mortality for patients with organ-confined tumors on long-term follow-up. Lastly, we report that expression of PD-1 by TILs within bladder cancers correlates with advanced tumor stage and with tumor expression of B7-H1. Together, these data suggest that T-cell coregulators may represent novel diagnostic or prognostic markers, as well as potential targets for alternative immunotherapy approaches, in patients with UCC of the bladder.

Patient selection. Upon approval from our Institutional Review Board, we reviewed the Mayo Clinic Cystectomy Registry to identify 410 consecutive patients treated with radical cystectomy between 1990 and 1994. This cohort was chosen to allow a minimum of 10 y of follow-up when assessing the association of expression with clinical outcomes. Patients with non-UCC histology, patients who underwent cystectomy for benign indications, and patients in whom no tissue was available for study were excluded from analysis.

Clinical characteristics collected included age, gender, smoking history, Eastern Cooperative Oncology Group (ECOG) performance status, receipt of preoperative intravesical bacillus Calmette-Guerin (BCG) therapy, and receipt of chemotherapy before cystectomy (neoadjuvant) or within 3 mo of surgery (adjuvant). The pathologic features examined included tumor stage, tumor grade, lymph node status, and the presence of lymphocytic infiltration within the bladder tumor. For follow-up, a registered nurse abstractor has updated patients' disease status and vital status annually through a review of the Mayo medical record, crosslinks with the Mayo Tumor Registry, and letters to both patients and local physicians. For those patients who have died, the cause of death was determined using the underlying cause of death listed on the death certificate or from direct correspondence with the local physician.

To assess the expression of T-cell coregulatory molecules in metastatic lymph nodes from patients with UCC, a second cohort of 66 patients who underwent radical cystectomy for lymph node–positive disease between 1996 and 2004 at the Mayo Clinic was identified. As nodal tissue was frequently not available in patients treated from 1990 to 1994, this more recent group allowed for a matched comparison of molecule expression in metastatic lymph nodes versus in the corresponding primary bladder tumors.

Immunohistochemistry. Formalin-fixed, paraffin-embedded tissues were cut into 5-μm sections, deparaffinized, and rehydrated in a graded series of ethanols. Staining of paraffin-embedded sections for B7-H1 was done as previously described (10). For B7-H3 and PD-1, antigen retrieval was done by heating tissue sections in 1 mmol/L EDTA (pH 8) to 121°C using a Digital Decloaking Chamber (Biocare Medical), cooling to 90°C, and incubating for 5 min. Sections were washed in Wash Buffer (Dako) before being placed onto the Autostainer Plus (Dako).

Sections were then blocked for endogenous peroxidase for 5 min using Endogenous Blocking Solution (Dako), washed twice, and incubated for 5 min in Serum-Free Protein Block (Dako), followed by incubation for 60 min in purified goat anti-human B7-H3 antibody (R&D Systems) diluted 1:400 with DaVinci Green antibody diluent (Biocare Medical) or purified goat anti-human PD-1 antibody (R&D Systems) diluted 1:40. Sections were incubated for 15 min in probe from Goat horseradish peroxidase Polymer kit (Biocare Medical #GHP516L), washed, and incubated for 15 min with polymer from Goat horseradish peroxidase Polymer kit. Sections were visualized by incubating in Betazoid 3,3′-diaminobenzidine (Biocare Medical) for 5 min. Sections were then counterstained with hematoxylin, dehydrated in ethanol, cleared in xylene, and coverslipped. The specificity of the antibodies used here has previously been validated (10, 18).

To further characterize the TILs noted to have high levels of PD-1 expression, serial sections from 12 patients with marked TIL PD-1 expression were prepared as above. The additional sections were then either incubated with mouse monoclonal anti-human CD4 antibody (Novocastra Laboratory) diluted 1:200 and visualized with Betazoid 3,3′-diaminobenzidine (Biocare Medical), or were double stained by first incubating with mouse monoclonal anti-human CD8 antibody (Dako), diluted 1:100 and visualized with Vulcan Fast Red chromogen (Biocare Medical), and then with mouse monoclonal anti-human granzyme B antibody (Dako), diluted 1:50 and visualized with Bajoran Purple chromogen (Biocare Medical).

Staining quantification. In accordance with our prior studies in renal cell carcinoma (10, 12, 18), the percentage of tumor cells that stained positive for B7-H1 was quantified in 5% increments. Tumors were considered positive for B7-H1 if there was histologic evidence of cell surface membrane staining in 5% or more of cells. Cases with <5% tumor staining were considered negative. The percentages of tumor and adjacent, nontumor cells that stained positive for B7-H3 were quantified in 10% increments, consistent with our previous evaluation of B7-H3 expression in prostate cancer (19). Cases with <10% of cells staining for B7-H3 were considered negative. Staining for PD-1 on TILs was recorded as absent, focal, moderate, or marked. All quantitation was done by a urologic pathologist (Y.S.) without knowledge of patient outcome.

Statistical methods. Associations of B7-H1, B7-H3, and PD-1 staining with clinicopathologic features were evaluated using Kruskal-Wallis, Wilcoxon rank-sum, χ2, and Fisher's exact tests, as appropriate. The duration of patient follow-up was calculated from the date of cystectomy to the date of disease progression, death, or last known follow-up, depending on the outcome measure of interest. Comparisons of outcome between patients with and without archived tissue available for study were evaluated using a log-rank test. The associations of B7-H1, B7-H3, and PD-1 expression with patient outcome were evaluated using Cox proportional hazards regression models and were summarized using hazard ratios (HR) and 95% confidence intervals (CI). Survival after cystectomy was estimated using the Kaplan Meier method. Statistical analyses were done using the SAS software package (SAS Institute). All tests were two-sided, and P values of <0.05 were considered statistically significant.

Clinicopathologic features of the study population. We identified 410 patients treated with radical cystectomy between 1990 and 1994, of whom 318 (77.6%) had paraffin-embedded tissue blocks available for analysis. There was no statistically significant difference in death from any cause (P = 0.21) or death from bladder cancer (P = 0.26) after cystectomy between patients with and without tissue available for study. Clinicopathologic features for the 318 patients studied here are summarized in Table 1. Median age at cystectomy was 69 years (range, 37-90 years).

Table 1.

Patient demographics

Featuren (%)
Gender  
    Male 259 (81.4) 
    Female 59 (18.6) 
Smoking history (n = 317)  
    Never 66 (20.8) 
    Former 151 (47.6) 
    Current 100 (31.6) 
ECOG performance status  
    0 241 (75.8) 
    1 64 (20.1) 
    2 12 (3.8) 
    3 1 (0.3) 
Preoperative BCG treatment  
    Yes 57 (17.9) 
    No 261 (82.1) 
Neoadjuvant chemotherapy  
    Yes 19 (6.0) 
    No 299 (94.0) 
Tumor stage (n = 316)  
    Ta/T1 55 (17.4) 
    CIS 43 (13.6) 
    T2 91 (28.8) 
    T3 88 (27.9) 
    T4 39 (12.3) 
Tumor grade (n = 290)  
    1 
    2 28 (9.7) 
    3/4 262 (90.3) 
Lymph node status  
    NX 44 (13.8) 
    N0 215 (67.6) 
    N1 25 (7.9) 
    N2 34 (10.7) 
Pathologically organ-confined disease (n = 316)  
    Yes 167 (52.8) 
    No 149 (47.2) 
Lymphocytic infiltration within tumor  
    Absent 4 (1.3) 
    Focal 147 (46.2) 
    Moderate 138 (43.4) 
    Marked 29 (9.1) 
Adjuvant chemotherapy (n = 310)  
    Yes 26 (8.4) 
    No 284 (91.6) 
Featuren (%)
Gender  
    Male 259 (81.4) 
    Female 59 (18.6) 
Smoking history (n = 317)  
    Never 66 (20.8) 
    Former 151 (47.6) 
    Current 100 (31.6) 
ECOG performance status  
    0 241 (75.8) 
    1 64 (20.1) 
    2 12 (3.8) 
    3 1 (0.3) 
Preoperative BCG treatment  
    Yes 57 (17.9) 
    No 261 (82.1) 
Neoadjuvant chemotherapy  
    Yes 19 (6.0) 
    No 299 (94.0) 
Tumor stage (n = 316)  
    Ta/T1 55 (17.4) 
    CIS 43 (13.6) 
    T2 91 (28.8) 
    T3 88 (27.9) 
    T4 39 (12.3) 
Tumor grade (n = 290)  
    1 
    2 28 (9.7) 
    3/4 262 (90.3) 
Lymph node status  
    NX 44 (13.8) 
    N0 215 (67.6) 
    N1 25 (7.9) 
    N2 34 (10.7) 
Pathologically organ-confined disease (n = 316)  
    Yes 167 (52.8) 
    No 149 (47.2) 
Lymphocytic infiltration within tumor  
    Absent 4 (1.3) 
    Focal 147 (46.2) 
    Moderate 138 (43.4) 
    Marked 29 (9.1) 
Adjuvant chemotherapy (n = 310)  
    Yes 26 (8.4) 
    No 284 (91.6) 

Abbreviation: CIS, carcinoma in situ.

B7-H3 expression in UCC. We found that B7-H3 was expressed by 222 of 314 (70.7%) of the bladder tumors evaluated and was uniformly expressed by the basal, middle, and luminal cell layers of the urothelium (Fig. 1). The intensity of expression was noted to be weak, moderate, and diffuse in 36 of 314 (11.5%), 134 of 314 (42.7%), and 52 of 314 (16.6%) cases, respectively. Adjacent, nontumor urothelium was available for analysis in 146 patients, and B7-H3 expression was detected in 107 (73.3%) of these specimens as well, primarily localized to the luminal, umbrella cells of the urothelium. The level of B7-H3 expression was significantly increased in UCC compared with nontumor urothelium, such that, among the 222 B7-H3+ tumors, a median of 70% (range, 10-100%) of cancer cells showed expression, versus a median of 20% (range, 10-70%) of urothelial cells in the 107 patients with B7-H3+ nontumor bladder specimens (P < 0.001; Wilcoxon signed-rank test).

Fig. 1.

Representative microphotographs showing T-cell coregulatory molecule expression in UCC of bladder. A, noninvasive UCC positive for B7-H3. B, invasive UCC expressing B7-H3. C, noninvasive UCC tumor negative for B7-H1. D, expression of B7-H1 by an invasive UCC. E, noninvasive UCC negative for PD-1+ lymphocytes. F, invasive UCC infiltrated by immune cells expressing PD-1. Original magnification, ×100; insets, ×400.

Fig. 1.

Representative microphotographs showing T-cell coregulatory molecule expression in UCC of bladder. A, noninvasive UCC positive for B7-H3. B, invasive UCC expressing B7-H3. C, noninvasive UCC tumor negative for B7-H1. D, expression of B7-H1 by an invasive UCC. E, noninvasive UCC negative for PD-1+ lymphocytes. F, invasive UCC infiltrated by immune cells expressing PD-1. Original magnification, ×100; insets, ×400.

Close modal

B7-H3 expression in UCC occurred across all tumor stages, as B7-H3+ tumor cells were noted in 44 of 54 (81.5%) pTa/T1 lesions, 30 of 42 (71.4%) cases of carcinoma in situ, and 146 of 216 (67.6%) pT2-T4 tumors (P = 0.13). Similarly, the intensity of B7-H3 staining was equivalent among these stages as well (P = 0.41). B7-H3 intensity was, however, noted to be significantly increased in the UCC specimens from patients who had received preoperative BCG therapy, as 15 of 56 (26.8%) patients treated preoperatively with BCG had diffuse B7-H3 staining, compared with 37 of 258 (14.3%) patients who were not treated with BCG before cystectomy (P = 0.023). B7-H3 expression was not associated with gender (P = 0.59), smoking history (P = 0.67), ECOG performance status (P = 0.097), or with the receipt of neoadjuvant chemotherapy (P = 0.15), although the median age at cystectomy for patients with B7-H3+ tumors was 67 years (range, 37-86 years) versus 71 years (range, 43-90 years) for patients with B7-H3 tumors (P = 0.019). As nearly all (90.3%) of the tumors from the study cohort here were classified as grade 3/4, reflecting the adverse pathology consistent with a population of cystectomy specimens, we did not separately test for the association of T-cell coregulatory molecule expression with tumor grade.

B7-H1 expression in UCC. B7-H1 was expressed by 39 of 314 (12.4%) of the UCC tumors evaluated. Among these cases, a median of 60% (range, 10-90%) of tumor cells expressed B7-H1. B7-H1+ cells within tumors were found diffusely throughout the cell layers of the urothelium (Fig. 1). Expression was noted primarily on the cell membrane, although additional, cytoplasmic staining was occasionally detected. B7-H1 expression was tumor-specific, as only rare, focal staining was noted on adjacent, nontumor urothelium. Consistent with previous reports (6, 7), B7-H1 expression was significantly associated with advanced tumor stage at cystectomy (P < 0.001; Table 2). We also noted here a significant increase in B7-H1 expression among tumors with a greater degree of TILs (P < 0.001). Moreover, contrary to the direct association noted between prior BCG treatment and B7-H3 staining, patients who received preoperative BCG were found to be less likely to express B7-H1 in their tumors at the time of cystectomy (P = 0.027).

Table 2.

Association of B7-H1 and PD-1 expression with clinical and pathological features

FeatureTumor B7-H1 expression
PTIL PD-1 expression
P
Negative (n = 275)
Positive (n = 39)
Absent/focal (n = 180)
Moderate (n = 113)
Marked (n = 21)
n (%)n (%)
Gender   0.11    0.85 
    Male 227 (82.6) 28 (71.8)  145 (80.6) 94 (83.2) 17 (81.0)  
    Female 48 (17.5) 11 (28.2)  35 (19.4) 19 (16.8) 4 (19.1)  
Smoking history (n = 313)   0.098    0.72 
    Never 57 (20.8) 8 (20.5)  37 (20.6) 21 (18.6) 6 (30.0)  
    Former 135 (49.3) 13 (33.3)  85 (47.2) 54 (47.8) 10 (50.0)  
    Current 82 (29.9) 18 (46.2)  58 (32.2) 38 (33.6) 4 (20.0)  
ECOG performance status >0 67 (24.4) 10 (25.6) 0.86 48 (26.7) 22 (19.5) 7 (33.3) 0.24 
Receipt of preoperative BCG 54 (19.6) 2 (5.1) 0.027 39 (1.7) 16 (14.2) 1 (4.8) 0.071 
Receipt of neoadjuvant chemotherapy 15 (5.5) 3 (7.7) 0.48 13 (7.2) 6 (5.3) 0.39 
Tumor stage (n = 312)   <0.001    0.012 
    Ta/T1 53 (19.4) 2 (5.1)  29 (16.1) 24 (21.6) 2 (9.5)  
    CIS 42 (15.4)  34 (18.9) 7 (6.3) 1 (4.8)  
    T2/T3/T4 178 (65.2) 37 (94.9)  117 (65.0) 80 (72.1) 18 (85.7)  
Lymph node status   0.93    0.79 
    NX/N0 224 (81.5) 32 (82.1)  147 (81.7) 90 (79.7) 18 (85.7)  
    N1/N2 51 (18.6) 7 (18)  33 (18.3) 23 (20.4) 3 (14.3)  
TIL   <0.001    <0.001 
    Absent 4 (1.5)  4 (2.2)  
    Focal 139 (50.6) 7 (18.0)  132 (73.3) 13 (11.5)  
    Moderate 112 (40.7) 24 (61.5)  44 (24.4) 87 (77.0) 5 (23.8)  
    Marked 20 (7.3) 8 (20.5)  13 (11.5) 16 (76.2)  
FeatureTumor B7-H1 expression
PTIL PD-1 expression
P
Negative (n = 275)
Positive (n = 39)
Absent/focal (n = 180)
Moderate (n = 113)
Marked (n = 21)
n (%)n (%)
Gender   0.11    0.85 
    Male 227 (82.6) 28 (71.8)  145 (80.6) 94 (83.2) 17 (81.0)  
    Female 48 (17.5) 11 (28.2)  35 (19.4) 19 (16.8) 4 (19.1)  
Smoking history (n = 313)   0.098    0.72 
    Never 57 (20.8) 8 (20.5)  37 (20.6) 21 (18.6) 6 (30.0)  
    Former 135 (49.3) 13 (33.3)  85 (47.2) 54 (47.8) 10 (50.0)  
    Current 82 (29.9) 18 (46.2)  58 (32.2) 38 (33.6) 4 (20.0)  
ECOG performance status >0 67 (24.4) 10 (25.6) 0.86 48 (26.7) 22 (19.5) 7 (33.3) 0.24 
Receipt of preoperative BCG 54 (19.6) 2 (5.1) 0.027 39 (1.7) 16 (14.2) 1 (4.8) 0.071 
Receipt of neoadjuvant chemotherapy 15 (5.5) 3 (7.7) 0.48 13 (7.2) 6 (5.3) 0.39 
Tumor stage (n = 312)   <0.001    0.012 
    Ta/T1 53 (19.4) 2 (5.1)  29 (16.1) 24 (21.6) 2 (9.5)  
    CIS 42 (15.4)  34 (18.9) 7 (6.3) 1 (4.8)  
    T2/T3/T4 178 (65.2) 37 (94.9)  117 (65.0) 80 (72.1) 18 (85.7)  
Lymph node status   0.93    0.79 
    NX/N0 224 (81.5) 32 (82.1)  147 (81.7) 90 (79.7) 18 (85.7)  
    N1/N2 51 (18.6) 7 (18)  33 (18.3) 23 (20.4) 3 (14.3)  
TIL   <0.001    <0.001 
    Absent 4 (1.5)  4 (2.2)  
    Focal 139 (50.6) 7 (18.0)  132 (73.3) 13 (11.5)  
    Moderate 112 (40.7) 24 (61.5)  44 (24.4) 87 (77.0) 5 (23.8)  
    Marked 20 (7.3) 8 (20.5)  13 (11.5) 16 (76.2)  

PD-1 expression in UCC. Expression of the B7-H1 receptor PD-1 was detected on TILs in 300 of 314 (95.5%) patients (Fig. 1). Expression was focal, moderate, and marked in 166 (52.9%), 113 (36.0%), and 21 (6.7%) cases, respectively. Expression of PD-1 by tumor cells or nontumorous urothelium was not observed. UCC tumors in those patients with PD-1+ immune cells were noted to be of higher pathologic stage (P = 0.012) and to contain a greater degree of lymphocytic infiltration (P < 0.001; Table 2). As noted with tumor expression of B7-H3, patients with moderate or marked expression of PD-1 on TILs tended to be more likely to have received BCG before cystectomy (P = 0.071). Interestingly, we also found that TIL PD-1 expression correlated with tumor expression of B7-H1, as 28 of 37 (75.7%) patients with B7-H1+ tumors had moderate or marked PD-1 expression on TILs, whereas only 105 of 274 (38.3%) patients with B7-H1 tumors had moderate or marked PD-1 expression on TILs (P < 0.001).

To further characterize the population of PD-1+ TILs, we evaluated serial section CD4, CD8, and granzyme B expression in a subset of 12 patient tumors exhibiting marked TIL PD-1 expression. We observed that nearly all intratumoral PD-1+ lymphocytes were CD8+ (representative case illustrated in Supplementary Fig. S1), with few PD-1+ CD4+ cells. In addition, we noted that <1% of the PD-1+ CD8+ T cells expressed granzyme B, a marker of active CD8+ T cells.

Associations of T-cell coregulatory molecule expression with clinical outcome. At last follow-up, 235 patients had died at a median of 2.7 years after cystectomy (range, 0.01-14.8). Among these 235 deaths, 102 patients died from UCC, at a median of 1.6 years after cystectomy (range, 0.08-14.1), 79 died from other causes, and 54 died from unknown causes. Among the 83 patients still alive at last follow-up, the median duration of follow-up was 13.7 years (range, 0.08-17.7). In this overall cohort of patients, we found no association between B7-H1, B7-H3, or PD-1 expression and disease progression after cystectomy, all-cause mortality, or cancer-specific mortality (Table 3). Similarly, no significant associations with survival were noted when we assessed the effect of combined expression of the ligand B7-H1 and its receptor PD-1 on outcome after cystectomy.

Table 3.

Associations of T-cell coregulatory molecule expression with outcome following cystectomy

FeatureDeath from any cause
Death from UCC
Disease progression
HR (95% CI)PHR (95% CI)PHR (95% CI)P
 All patients (n = 318)      
Positive tumor B7-H1 expression 1.06 (0.71-1.58) 0.79 0.82 (0.42-1.63) 0.58 0.83 (0.43-1.58) 0.56 
Diffuse tumor B7-H3 expression 1.11 (0.79-1.55) 0.55 1.14 (0.69-1.90) 0.61 0.83 (0.48-1.43) 0.50 
Moderate/marked TIL PD-1 expression 0.85 (0.65-1.10) 0.21 0.69 (0.46-1.03) 0.07 0.63 (0.43-0.93) 0.02 
Combined B7-H1 and PD-1 expression 0.91 (0.56-1.46) 0.68 0.83 (0.39-1.79) 0.64 0.83 (0.40-1.70) 0.61 
 Patients with organ-confined disease (n = 167)      
Positive tumor B7-H1 expression 2.18 (1.26-3.77) 0.005 1.59 (0.56-4.49) 0.38 1.24 (0.44-3.45) 0.69 
Diffuse tumor B7-H3 expression 1.03 (0.64-1.65) 0.92 1.03 (0.45-2.35) 0.94 0.94 (0.44-2.01) 0.87 
Moderate/marked TIL PD-1 expression 1.06 (0.73-1.54) 0.75 0.70 (0.36-1.36) 0.29 0.60 (0.32-1.10) 0.10 
Combined B7-H1 and PD-1 expression 2.02 (1.08-3.77) 0.028 1.49 (0.46-4.87) 0.51 1.06 (0.33-3.42) 0.92 
FeatureDeath from any cause
Death from UCC
Disease progression
HR (95% CI)PHR (95% CI)PHR (95% CI)P
 All patients (n = 318)      
Positive tumor B7-H1 expression 1.06 (0.71-1.58) 0.79 0.82 (0.42-1.63) 0.58 0.83 (0.43-1.58) 0.56 
Diffuse tumor B7-H3 expression 1.11 (0.79-1.55) 0.55 1.14 (0.69-1.90) 0.61 0.83 (0.48-1.43) 0.50 
Moderate/marked TIL PD-1 expression 0.85 (0.65-1.10) 0.21 0.69 (0.46-1.03) 0.07 0.63 (0.43-0.93) 0.02 
Combined B7-H1 and PD-1 expression 0.91 (0.56-1.46) 0.68 0.83 (0.39-1.79) 0.64 0.83 (0.40-1.70) 0.61 
 Patients with organ-confined disease (n = 167)      
Positive tumor B7-H1 expression 2.18 (1.26-3.77) 0.005 1.59 (0.56-4.49) 0.38 1.24 (0.44-3.45) 0.69 
Diffuse tumor B7-H3 expression 1.03 (0.64-1.65) 0.92 1.03 (0.45-2.35) 0.94 0.94 (0.44-2.01) 0.87 
Moderate/marked TIL PD-1 expression 1.06 (0.73-1.54) 0.75 0.70 (0.36-1.36) 0.29 0.60 (0.32-1.10) 0.10 
Combined B7-H1 and PD-1 expression 2.02 (1.08-3.77) 0.028 1.49 (0.46-4.87) 0.51 1.06 (0.33-3.42) 0.92 

Given the associations of B7-H1 and PD-1 staining with tumor stage, however, we then separately evaluated the prognostic value of T-cell coregulatory molecule expression in patients found to have organ-confined (pTa/pT1/CIS/pT2 NX/N0 M0) versus nonorgan-confined (pT3/T4NanyMany or pTanyN+Many or pTanyNanyM+) disease at cystectomy. In total, 167 of 316 (52.8%) patients had pathologically organ-confined tumors. From this cohort, 48 patients experienced disease progression at a median of 2.3 years after cystectomy (range, 0.17-14.2), and 113 patients died after surgery at a median of 5.0 years postoperatively (range, 0.01-14.8), including 39 who died from UCC at a median of 3.3 years after cystectomy (range, 0.08-14.1). Cause of death in the remaining 74 patients with organ-confined disease who died after radical cystectomy included cardiovascular disease (n = 20), other cancers (n = 17), pulmonary disease (n = 6), renal failure (n = 5), other medical conditions (n = 5), and was unknown in 21 patients. Among the 54 patients still alive at last follow-up, the median duration of follow-up was 13.6 years (range 0.02-17.5).

For the patients with organ-confined disease, we found that tumor B7-H1 expression was a significant predictor of all-cause mortality after cystectomy on univariate analysis (HR, 2.18; 95% CI, 1.26-3.77; P = 0.005; Table 3). Overall survival of patients with organ-confined disease after cystectomy, stratified by tumor B7-H1 expression, is depicted in Fig. 2. Based on these findings, we then assessed the ability of B7-H1 expression to predict postoperative mortality for patients with organ-confined disease on a multivariate analysis controlling for age at surgery, ECOG performance status, smoking history, and tumor stage (Table 4). Lymphocytic infiltration and receipt of BCG therapy before cystectomy were found not to predict mortality on univariate analysis (data not shown) and were not included in the multivariate model. Furthermore, only seven patients in this group received neoadjuvant chemotherapy, and no patient received adjuvant chemotherapy; therefore, the effect of these variables could not be evaluated here. In addition, as 90% of the tumors in our series were classified as grade 3/4, and no grade 1 cancers were recorded, the independent effect of tumor grade was likewise not assessed. Importantly, B7-H1 remained highly associated with death from any cause after surgery on multivariate analysis, as patients with organ-confined B7-H1+ tumors were 3.18 times more likely to die after cystectomy than patients with organ-confined B7-H1 tumors (HR, 3.18; 95% CI, 1.74-5.79; P < 0.001).

Fig. 2.

Association of B7-H1 expression with overall survival for the 167 patients with organ-confined UCC of the bladder. Overall survival rates (SE, number still at risk) at 1, 5, and 10 y after cystectomy were 93.8% (2.0%, 135), 68.6% (3.9%, 96), and 50.0% (4.2%, 70), respectively, for patients with organ-confined B7-H1 tumors compared with 58.8% (11.9%, 10), 41.2% (11.9%, 7), and 29.4% (11.1%, 5), respectively, for patients with organ-confined B7-H1+ tumors (P = 0.005).

Fig. 2.

Association of B7-H1 expression with overall survival for the 167 patients with organ-confined UCC of the bladder. Overall survival rates (SE, number still at risk) at 1, 5, and 10 y after cystectomy were 93.8% (2.0%, 135), 68.6% (3.9%, 96), and 50.0% (4.2%, 70), respectively, for patients with organ-confined B7-H1 tumors compared with 58.8% (11.9%, 10), 41.2% (11.9%, 7), and 29.4% (11.1%, 5), respectively, for patients with organ-confined B7-H1+ tumors (P = 0.005).

Close modal
Table 4.

Multivariate analysis of factors associated with death from any cause following radical cystectomy for patients with pathologically organ-confined disease

FeatureHR (95% CI)P
Age at cystectomy 1.05 (1.03-1.08)* <0.001 
ECOG (>0 vs 0) 2.63 (1.68-4.13) <0.001 
Smoking status (former/current vs never) 1.69 (1.04-2.76) 0.04 
Tumor stage   
    Ta/T1 1.0 (Reference)  
    CIS 0.80 (0.48-1.36) 0.41 
    T2 0.95 (0.61-1.49) 0.83 
Positive tumor B7-H1 expression 3.18 (1.74-5.79) <0.001 
FeatureHR (95% CI)P
Age at cystectomy 1.05 (1.03-1.08)* <0.001 
ECOG (>0 vs 0) 2.63 (1.68-4.13) <0.001 
Smoking status (former/current vs never) 1.69 (1.04-2.76) 0.04 
Tumor stage   
    Ta/T1 1.0 (Reference)  
    CIS 0.80 (0.48-1.36) 0.41 
    T2 0.95 (0.61-1.49) 0.83 
Positive tumor B7-H1 expression 3.18 (1.74-5.79) <0.001 
*

Represents increased risk of death from any cause for each 1-year increase in age.

At the same time, although patients with organ-confined B7-H1+ tumors were found to be 59% more likely to die from UCC compared with patients with organ-confined B7-H1 tumors, this difference did not reach statistical significance (P = 0.38), likely due to the few deaths from UCC in patients with organ-confined disease, which limited the statistical power needed to detect this difference as statistically significant. Similarly, B7-H1 expression was not a significant predictor of disease-free survival for patients with organ-confined disease (P = 0.69; data provided in Table 3 and Supplementary Fig. S2). Moreover, analyzing the combined expression of B7-H1 and its receptor PD-1 yielded no additive prognostic value. In addition, no significant associations of B7-H1, B7-H3, or PD-1 expression with cancer death were noted among the 149 patients diagnosed with nonorgan confined disease (data not shown).

Matched expression in bladder tumors and metastatic lymph nodes. Expression of B7-H1, B7-H3, and PD-1 were evaluated from matched bladder tumor and metastatic lymph node specimens from 66 patients with TanyN+ UCC of the bladder who underwent radical cystectomy and bilateral pelvic lymph node dissection at our institution between 1996 and 2004. We found that expression of each of these three molecules was highly correlated between the primary tumors and metastatic nodes (staining illustrated in Supplementary Fig. S3), such that 51 of 53 (96.2%) of these patients with B7-H3+ bladder tumors also expressed B7-H3 in the nodal metastases, with no significant difference in the distribution of intensity of staining between the lymph node and bladder specimens. Likewise, 6 of 7 (85.7%) patients with B7-H1+ bladder tumors also showed expression in the corresponding metastatic lymph nodes. PD-1 expression in metastatic lymph nodes also matched the expression pattern of the primary bladder tumors, such that 36 of 45 (80%) patients with PD-1+ immune cells detected in nodal metastases were found to have PD-1+ TILs within their bladder tumors.

We present what is to our knowledge the first report of expression of the inhibitory T-cell coregulator B7-H3 in UCC of the bladder. We show significantly increased expression of B7-H3 in UCC compared with nontumor urothelium, and note that this up-regulation of expression occurred even in low-stage lesions, suggesting that increased B7-H3 expression may represent an early event in the malignant transformation of urothelial cells. At the same time, we confirm the association of tumor B7-H1 and TIL PD-1 expression with advanced tumor stage, using the largest series of UCC specimens, which have been evaluated for T-cell coregulator expression to date. Moreover, we describe a significant correlation between expression of B7-H1 by UCC tumor cells and expression of the B7-H1 receptor PD-1 on TILs. Furthermore, we report that B7-H1 expression by UCC tumors is an independent predictor of mortality after cystectomy in patients with organ-confined disease.

Over the last decade, increasing evidence has suggested that T-cell function and survival are governed by a balance between positive and negative T-cell coregulatory signaling molecules (20). The central role of the B7 family of coregulatory ligands in regulation of the antigen-specific T cell–mediated immune response is well-established (21). For example, B7-H1 and its receptor, PD-1, are members of the B7 family, which govern the activation of naïve T cells as well as the deletion of activated T cells (16, 20). In vitro and in vivo studies show that binding of PD-1 by B7-H1 inhibits T-cell proliferation, survival, or function, whereas blockade of PD-1 promotes autoreactive T-cell and autoantibody production (22). B7-H3 is also thought to serve as an accessory coregulator of T-cell responses after initial antigen priming (23). Although the precise role of B7-H3 remains undefined, most recent studies support its function as an immune inhibitor (17, 24). However, unlike B7-H1, the receptor for B7-H3 has yet to be identified.

Although the expression of B7 molecules is typically restricted to monocyte-derived cells, aberrant expression of these ligands by various human cancers has been described (516). In particular, tumor-associated B7-H1 has been shown to inhibit antitumoral T-cell immunity by interacting with T-cell PD-1 or a putative non–PD-1 receptor to induce tumor-specific T-cell apoptosis or to impair T-cell cytokine production and cytotoxicity (15, 25, 26).

Bladder cancer is a particularly immunogenic malignancy, as pathologic UCC specimens have been shown to harbor TILs (2730), and in fact the presence of CD8+ TILs has been shown to predict survival in muscle-invasive UCC (30). At the same time, however, patients with bladder cancer exhibit a tumor-associated immunologic suppression, particularly evident as an impaired T-cell response (24, 31). Indeed, bladder tumors have been shown to be infiltrated by T regulatory cells, and to express high levels of inhibitory cytokines (4), which may in part explain the immune dysfunction. Thus, we investigated the aberrant expression of T-cell coregulatory molecules, which have been shown in other model systems to inhibit the immune response (14, 15, 17, 22, 25, 26), by UCC cells and TILs.

We show aberrant T-cell coregulatory molecule expression at various stages of UCC in the bladder. The finding of a significant up-regulation of B7-H3 expression in tumor tissue versus nontumor urothelium, for example, together with the fact that this increased expression occurs across all pathologic stages of UCC, suggest a possible role for B7-H3 up-regulation early in the malignant transformation of urothelial cells. In addition, as we noted that tumor expression of B7-H1 was significantly associated with TIL expression of the B7-H1 receptor PD-1, similar to what we have observed in renal cell carcinoma (5), the potential exists for B7-H1 on UCC cells to interact with PD-1 on the TILs and thereby impair the function and survival of host immune cells. Consistent with this hypothesis, we observed that the majority of infiltrating intratumoral PD-1+ lymphocytes were CD8+ granzyme B T cells, suggesting the possibility that PD-1 expression, or the PD-1/B7-H1 interaction, suppresses the ability of tumor-infiltrating CD8+ T cells to acquire or maintain full cytolytic capacity. Interestingly, blockade of the PD-1/B7-H1 interaction has been shown to restore granzyme B expression in CD8+ T cells with impaired activity in the setting of hepatitis C infection (32). Moreover, as has been shown with p53 expression (33), we found that the expression pattern of B7-H1, B7-H3, and PD-1 match closely between the primary bladder tumor and nodal metastases from patients with lymph node–positive UCC. Additional study is required to establish whether aberrant T-cell coregulatory molecule expression functions to facilitate the metastatic process by aiding in host immune evasion.

Our observation that B7-H1 expression predicts postoperative mortality in patients with organ-confined tumors may have important implications for the management of patients after cystectomy as well. To date, adjuvant chemotherapy trials in bladder cancer have shown a survival advantage primarily for patients with pathologic evidence of extravesical or lymph node–positive disease (34, 35). Nevertheless, despite improvements in surgical technique and perioperative care, the survival rates of patients undergoing cystectomy for organ-confined bladder cancer have not significantly improved in recent years (36). Indeed, the 5-year overall survival for patients with muscle-invasive, organ-confined bladder cancer from multiple current series has been reported between 60% and 75% (3739). Thus, select patients even with what is considered relatively favorable pathology may benefit from treatment in addition to surgery.

However, given the potential morbidity associated with perioperative chemotherapy, identifying those patients with organ-confined tumors at highest risk for disease progression remains important to avoid overtreatment. Accurate risk stratification may guide patient counseling and surveillance regimens, and may facilitate the enrollment of high-risk patients in adjuvant therapy trials. Molecular biomarkers have been investigated as a means to add additional prognostic information to the recognized clinical predictors of tumor biology, and thereby help characterize an individual patient's risk of disease progression after surgery (4042). With external validation, B7-H1 may represent such a marker for patients with organ-confined disease.

Although we did not find an association between B7-H1 expression and death from UCC, B7-H1 did predict all-cause mortality after cystectomy on univariate and multivariate analysis for patients with organ-confined disease. As our institution serves as a tertiary referral center, patients are often followed locally after undergoing cystectomy; therefore, although vital status may be reliably established, determining the cause of death through documentation such as death certificates can be much more problematic, and has been shown to introduce error into the reporting of cancer mortality rates (43). Here, for example, in 21 of 74 (28.4%) patients with organ-confined disease who were reported to have died after radical cystectomy, but in whom death was not attributed to UCC, the cause of death was unknown. As such, previous studies that have used databases from large referral centers or population-based registries to explore the effect of neoadjuvant chemotherapy (44) or the extent of lymphadenectomy (45) on survival after radical cystectomy have used overall mortality as a primary end point as well. Overall, then, we believe that demonstrating the association of B7-H1 expression with death from any cause does, pending external validation, provide valuable initial data regarding the potential use of this molecule as a biomarker in UCC of the bladder.

Defining the mechanisms that underlie the observed immune dysfunction in patients with UCC may also help identify the optimal candidates for intravesical instillation of BCG. That is, although BCG treatment in bladder cancer is considered among the most successful examples of immunotherapies to date (46), ∼30% of patients initially fail to respond, and up to 74% of those patients who do respond to BCG ultimately relapse (47). Therefore, it has been proposed that certain populations of UCC cells may possess mechanisms that protect them from immunotherapy (48). As such, one previous investigation noted lower levels of ligands for natural killer cells in BCG-resistant bladder tumors, and suggested that analysis of the expression of these ligands may be used as a predictive assay for BCG response (48). We noted here a significantly decreased expression of B7-H1, and increased expression of B7-H3 and PD-1, in the specimens from patients who had received BCG before cystectomy. With the shown effects of coregulator expression on T-cell function, we believe that further evaluation of a role for these molecules in determining patients' response to BCG, ideally conducted in a prospective setting, is warranted.

In addition to the potential role of T-cell coregulators as biomarkers in UCC, these molecules may also represent promising targets for antitumoral immunotherapy. Indeed, previous experience with in vivo blockade of the T-cell costimulatory receptor CTLA-4 has provided compelling evidence of the value of abrogating inhibitory T-cell signaling with immunotherapy (49, 50). Moreover, in vivo blockade of B7-H1 has been shown to potentiate antitumor T-cell responses directed against immunogenic murine tumors expressing B7-H1 either endogenously or after B7-H1 gene transduction (9, 15, 26). Thus, targeting these ligands to enhance host antitumor immunity may prove useful, either alone or in combination with current conventional UCC therapies, to improve the effectiveness of bladder cancer treatment.

We acknowledge that this study represents an evaluation of the expression pattern of T-cell coregulatory molecules, and, as such, further preclinical investigation is required to establish functionality of this pathway in UCC of the bladder. The potential shortcomings of immunohistochemistry, including dependence upon fixation techniques and variability of interpretation, are noted as well. In addition, we recognize the limitations of retrospective data analysis as done here. Nevertheless, we show a strong association of B7-H3 expression with UCC, as well as significant correlations of B7-H1 and PD-1 expression with localized stage progression. Moreover, we found that B7-H1 predicts survival after cystectomy for patients with organ-confined disease on long-term follow-up. Thus, we believe that aberrant T-cell coregulatory molecule expression may have important diagnostic, prognostic, and potentially therapeutic implications for patients with UCC.

Several of the authors have filed patent applications pertaining to cancer prognostic markers, including B7-H1 and B7-H3. Dr. Kwon has received royalties from the licensing of this technology to Medarex.

Grant support: The Richard M. Schulze Family Foundation, the Helen and Martin Kimmel Foundation (E.D. Kwon), and by a Ferdinand C. Valentine Fellowship from the New York Academy of Medicine (S.A. Boorjian).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/).

E.D. Kwon and I. Frank contributed equally to this work.

1
Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007.
CA Cancer J Clin
2007
;
57
:
43
–66.
2
Herr HW, Bean MA, Whitmore WF, Jr. Decreased ability of blood leukocytes from patients with tumors of the urinary bladder to act as stimulator cells in mixed leukocyte culture.
Cancer Res
1976
;
36
:
2754
–60.
3
Mukamel E, Shohat B, Servadio C. Immunological profile of patients with transitional cell carcinoma of the bladder.
Br J Urol
1982
;
54
:
11
–5.
4
Loskog A, Ninalga C, Paul-Wetterberg G, de la Torre M, Malmstrom PU, Totterman TH. Human bladder carcinoma is dominated by T-regulatory cells and Th1 inhibitory cytokines.
J Urol
2007
;
177
:
353
–8.
5
Thompson RH, Dong H, Lohse CM, et al. PD-1 is expressed by tumor-infiltrating immune cells and is associated with poor outcome for patients with renal cell carcinoma.
Clin Cancer Res
2007
;
13
:
1757
–61.
6
Nakanishi J, Wada Y, Matsumoto K, Azuma M, Kikuchi K, Ueda S. Overexpression of B7–1 (PD-L1) significantly associates with tumor grade and postoperative prognosis in human urothelial cancers.
Cancer Immunol Immunother
2007
;
56
:
1173
–82.
7
Inman BA, Sebo TJ, Frigola X, et al. PD-L1 (B7–1) expression by urothelial carcinoma of the bladder and BCG-induced granulomata: associations with localized stage progression.
Cancer
2007
;
109
:
1499
–505.
8
Krambeck AE, Thompson RH, Dong H, et al. B7–4 expression in renal cell carcinoma and tumor vasculature: associations with cancer progression and survival.
Proc Natl Acad Sci U S A
2006
;
103
:
10391
–6.
9
Hirano F, Kaneko K, Tamura H, et al. Blockade of B7–1 and PD-1 by monoclonal antibodies potentiates cancer therapeutic immunity.
Cancer Res
2005
;
65
:
1089
–96.
10
Thompson RH, Kuntz SM, Leibovich BC, et al. Tumor B7–1 is associated with poor prognosis in renal cell carcinoma patients with long-term follow-up.
Cancer Res
2006
;
66
:
3381
–5.
11
Thompson RH, Webster WS, Cheville JC, et al. B7–1 glycoprotein blockade: a novel strategy to enhance immunotherapy in patients with renal cell carcinoma.
Urology
2005
;
66
:
10
–4.
12
Thompson RH, Gillett MD, Cheville JC, et al. Costimulatory B7–1 in renal cell carcinoma patients: indicator of tumor aggressiveness and potential therapeutic target.
Proc Natl Acad Sci U S A
2004
;
101
:
17174
–9.
13
Konishi J, Yamazaki K, Azuma M, Kinoshita I, Dosaka-Akita H, Nishimura M. B7–1 expression on non-small cell lung cancer cells and its relationship with tumor-infiltrating lymphocytes and their PD-1 expression.
Clin Cancer Res
2004
;
10
:
5094
–100.
14
Zang X, Loke P, Kim J, Murphy K, Waitz R, Allison JP. B7x: a widely expressed B7 family member that inhibits T cell activation.
Proc Natl Acad Sci U S A
2003
;
100
:
10388
–92.
15
Dong H, Strome SE, Salomao DR, et al. Tumor-associated B7–1 promotes T-cell apoptosis: a potential mechanism of immune evasion.
Nat Med
2002
;
8
:
793
–800.
16
Dong H, Zhu G, Tamada K, Chen L. B7–1, a third member of the B7 family, co-stimulates T-cell proliferation and interleukin-10 secretion.
Nat Med
1999
;
5
:
1365
–9.
17
Suh WK, Gajewska BU, Okada H, et al. The B7 family member B7–3 preferentially down-regulates T helper type 1-mediated immune responses.
Nat Immunol
2003
;
4
:
899
–906.
18
Ling V, Wu PW, Spaulding V, et al. Duplication of primate and rodent B7–3 immunoglobulin V- and C-like domains: divergent history of functional redundancy and exon loss.
Genomics
2003
;
82
:
365
–77.
19
Roth TJ, Sheinin Y, Lohse CM, et al. B7–3 ligand expression by prostate cancer: a novel marker of prognosis and potential target for therapy.
Cancer Res
2007
;
67
:
7893
–900.
20
Chen L. Co-inhibitory molecules of the B7–28 family in the control of T-cell immunity.
Nat Rev Immunol
2004
;
4
:
336
–47.
21
Carreno BM, Collins M. BTLA: a new inhibitory receptor with a B7-like ligand.
Trends Immunol
2003
;
24
:
524
–7.
22
Zha Y, Blank C, Gajewski TF. Negative regulation of T-cell function by PD-1.
Crit Rev Immunol
2004
;
24
:
229
–37.
23
Chapoval AI, Ni J, Lau JS, et al. B7–3: a costimulatory molecule for T cell activation and IFN-γ production.
Nat Immunol
2001
;
2
:
269
–74.
24
Zang X, Allison JP. The B7 family and cancer therapy: costimulation and coinhibition.
Clin Cancer Res
2007
;
13
:
5271
–9.
25
Blank C, Brown I, Peterson AC, et al. PD-L1/B7H-1 inhibits the effector phase of tumor rejection by T cell receptor (TCR) transgenic CD8+ T cells.
Cancer Res
2004
;
64
:
1140
–5.
26
Iwai Y, Ishida M, Tanaka Y, Okazaki T, Honjo T, Minato N. Involvement of PD-L1 on tumor cells in the escape from host immune system and tumor immunotherapy by PD-L1 blockade.
Proc Natl Acad Sci U S A
2002
;
99
:
12293
–7.
27
Stavropoulos NE, Ioachim E, Hastazeris K, Goussia A, Agnantis NJ. Lymphocyte subsets in superficial bladder cancer: correlation with biological activity of the tumours.
Anticancer Res
1998
;
18
:
2037
–40.
28
Hilmy M, Campbell R, Bartlett JM, McNicol AM, Underwood MA, McMillan DC. The relationship between the systemic inflammatory response, tumour proliferative activity, T-lymphocytic infiltration and COX-2 expression and survival in patients with transitional cell carcinoma of the urinary bladder.
Br J Cancer
2006
;
95
:
1234
–8.
29
Bevers RF, Kurth KH, Schamhart DH. Role of urothelial cells in BCG immunotherapy for superficial bladder cancer.
Br J Cancer
2004
;
91
:
607
–12.
30
Sharma P, Shen Y, Wen S, et al. CD8 tumor-infiltrating lymphocytes are predictive of survival in muscle-invasive urothelial carcinoma.
Proc Natl Acad Sci U S A
2007
;
104
:
3967
–72.
31
Kastelan Z, Lukac J, Derezic D, et al. Lymphocyte subsets, lymphocyte reactivity to mitogens, NK cell activity and neutrophil and monocyte phagocytic functions in patients with bladder carcinoma.
Anticancer Res
2003
;
23
:
5185
–9.
32
Lukens JR, Cruise MW, Lassen MG, Hahn YS. Blockade of PD-1/B7–1 interaction restores effector CD8+ T cell responses in a hepatitis C virus core murine model.
J Immunol
2008
;
180
:
4875
–84.
33
Fleshner N, Kapusta L, Ezer D, Herschorn S, Klotz L. p53 nuclear accumulation is not associated with decreased disease-free survival in patients with node positive transitional cell carcinoma of the bladder.
J Urol
2000
;
164
:
1177
–82.
34
Freiha F, Reese J, Torti FM. A randomized trial of radical cystectomy versus radical cystectomy plus cisplatin, vinblastine and methotrexate chemotherapy for muscle invasive bladder cancer.
J Urol
1996
;
155
:
495
–9; discussion 9–500.
35
Skinner DG, Daniels JR, Russell CA, et al. The role of adjuvant chemotherapy following cystectomy for invasive bladder cancer: a prospective comparative trial.
J Urol
1991
;
145
:
459
–64; discussion 64–7.
36
Herr HW, Dotan Z, Donat SM, Bajorin DF. Defining optimal therapy for muscle invasive bladder cancer.
J Urol
2007
;
177
:
437
–43.
37
Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients.
J Clin Oncol
2001
;
19
:
666
–75.
38
Shariat SF, Karakiewicz PI, Palapattu GS, et al. Outcomes of radical cystectomy for transitional cell carcinoma of the bladder: a contemporary series from the Bladder Cancer Research Consortium.
J Urol
2006
;
176
:
2414
–22; discussion 22.
39
Madersbacher S, Hochreiter W, Burkhard F, et al. Radical cystectomy for bladder cancer today-a homogeneous series without neoadjuvant therapy.
J Clin Oncol
2003
;
21
:
690
–6.
40
Shariat SF, Karakiewicz PI, Ashfaq R, et al. Multiple biomarkers improve prediction of bladder cancer recurrence and mortality in patients undergoing cystectomy.
Cancer
2008
;
112
:
315
–25.
41
Silva Neto B, Smith GL, Mandeville JA, et al. Prognostic significance of altered p120 ctn expression in bladder cancer.
BJU Int
2008
;
101
:
746
–52.
42
Baumgart E, Cohen MS, Silva Neto B, et al. Identification and prognostic significance of an epithelial-mesenchymal transition expression profile in human bladder tumors.
Clin Cancer Res
2007
;
13
:
1685
–94.
43
Hoel DG, Ron E, Carter R, Mabuchi K. Influence of death certificate errors on cancer mortality trends.
J Natl Cancer Inst
1993
;
85
:
1063
–8.
44
Grossman HB, Natale RB, Tangen CM, et al. Neoadjuvant chemotherapy plus cystectomy compared with cystectomy alone for locally advanced bladder cancer.
N Engl J Med
2003
;
349
:
859
–66.
45
Hollenbeck BK, Ye Z, Wong SL, Montie JE, Birkmeyer JD. Hospital lymph node counts and survival after radical cystectomy.
Cancer
2008
;
112
:
806
–12.
46
Chapman PB, Houghton AN. Non-antibody immunotherapy of cancer.
Curr Opin Immunol
1993
;
5
:
726
–31.
47
Soloway MS, Sofer M, Vaidya A. Contemporary management of stage T1 transitional cell carcinoma of the bladder.
J Urol
2002
;
167
:
1573
–83.
48
Yutkin V, Pode D, Pikarsky E, Mandelboim O. The expression level of ligands for natural killer cell receptors predicts response to bacillus Calmette-Guerin therapy: a pilot study.
J Urol
2007
;
178
:
2660
–4.
49
Phan GQ, Yang JC, Sherry RM, et al. Cancer regression and autoimmunity induced by cytotoxic T lymphocyte-associated antigen 4 blockade in patients with metastatic melanoma.
Proc Natl Acad Sci U S A
2003
;
100
:
8372
–7.
50
Egen JG, Kuhns MS, Allison JP. CTLA-4: new insights into its biological function and use in tumor immunotherapy.
Nat Immunol
2002
;
3
:
611
–8.

Supplementary data