Bladder Cancer as a Prognostic Factor for Upper Tract Transitional Cell Carcinoma

0022-5347/04/1726-2177/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 172, 2177–2181, December 2004 Printed in U.S.A.

DOI: 10.1097/01.ju.0000144505.40915.98

BLADDER CANCER AS A PROGNOSTIC FACTOR FOR UPPER TRACT TRANSITIONAL CELL CARCINOMA MICHAEL MULLERAD, PAUL RUSSO, DRAGAN GOLIJANIN, HUI NI CHEN, HAROLD H. TSAI,* S. MACHELE DONAT, BERNARD H. BOCHNER, HARRY W. HERR, JOEL SHEINFELD, PRAMOD C. SOGANI, MICHAEL W. KATTAN AND GUIDO DALBAGNI† From the Departments of Urology and Radiology (HNC), Memorial Sloan-Kettering Cancer Center, New York, New York

ABSTRACT

Purpose: Upper tract transitional cell carcinoma (UTTCC) is a relatively rare tumor. Overall 5-year disease specific survival is in the range of 16.5% to 95% depending on stage. In this study we evaluated predictors associated with disease recurrence and disease specific survival. Materials and Methods: We report on 129 patients with a median age of 68 years who underwent nephroureterectomy for UTTCC between July 1989 and June 2002. A total of 67 patients had primary UTTCC and 62 had previous (52) or synchronous (10) transitional cell carcinoma of the bladder (BTCC). Medical records were reviewed and analyzed for possible prognostic predictors (primary tumor stage, grade, multifocality, carcinoma in situ, symptoms and signs at presentation, sex, and history of smoking). Disease specific survival and freedom from bladder recurrence were assessed with the Kaplan-Meier method, and differences between the groups were compared using the log rank test. The Cox proportional hazards regression model was used to assess the significance of each predictor. Results: Disease specific death was reported for 44 patients. In a multivariate analysis using previous BTCC as a predictor (categorized as superficial, invasive or none), primary disease stage and history of BTCC were associated with disease specific survival (p ⫽ 0.001 and p ⫽ 0.018). History of BTCC grouped the patients into distinct populations in terms of disease specific survival and freedom from bladder recurrence. Conclusions: This study demonstrates that a history of BTCC (invasive or superficial) has an adverse effect on the prognosis of patients diagnosed with UTTCC independent of primary tumor stage. KEY WORDS: carcinoma, transitional cell; carcinoma, bladder, urinary tract, prognosis

Upper tract transitional cell carcinoma (UTTCC) is a relatively rare tumor. About 5% of primary urothelial tumors originate from the upper urinary collecting system.1 Munoz and Ellison used the National Cancer Institute Surveillance, Epidemiology and End Results database to evaluate the changes in incidence of and survival of the disease.2 They found the incidence of ureteral and pelvic tumors to be 0.73 per 100,000 and about 1.0 per 100,000 person-years, respectively, from 1985 to 1994. Overall 5-year disease specific survival was 75%, ranging from 95% to 88.9% for in situ and localized disease, and decreasing to 62.6% and 16.5% for regional and distant disease spread, respectively. Hall et al evaluated predictors affecting prognosis in a cohort of 252 patients with upper tract tumors followed for a median of 64 months.3 Nephroureterectomy was the treatment of choice in 77% of the patients whereas 17% were treated with the parenchymal sparing approach. Disease recurred in 27% of the patients and the most frequent site was the bladder, accounting for 51% of the recurrence. Retroperitoneal and remaining upper tract recurrence accounted for 27% of events. Of disease recurrence events 22% presented as metastatic disease. Predictors significantly associated with disease recurrence and disease specific survival were tumor grade and disease stage. A multivariate analysis demonstrated that stage and type of sur-

gical procedure were significantly associated with disease recurrence. Hisataki et al found that the 5-year rate of bladder recurrence after treatment of primary upper tract disease was 35%, and that independent predictors for recurrence included tumor multifocality as well as pathological stage.4 Although several clinical variables have been implicated in the adverse prognosis for patients with upper tract disease, it is not known whether a history of transitional cell carcinoma of the bladder determines the natural course of upper tract disease. In this study we evaluated whether a history of bladder cancer was associated with recurrence and disease specific survival after controlling for known prognostic factors. MATERIALS AND METHODS

From July 1989 to June 2002, 86 men and 43 women with a median age of 68 years (range 41 to 86) underwent nephroureterectomy for upper tract TCC at the Memorial Sloan-Kettering Cancer Center. A total of 67 patients had primary UTTCC and 62 had previous (52) or synchronous (10) transitional cell carcinoma of the bladder (BTCC). Of the 62 patients with bladder cancer, 33 had superficial UTTCC and 29 had invasive muscle disease (fig. 1). A total of 22 patients had invasive bladder disease, of whom superficial UTTCC developed in 10 and invasive UTTCC developed in 12 (15 were treated with radical cystectomy and 7 with bladder sparing surgery). There was a history of contralateral UTTCC in 2 patients and 7 had unilateral UTTCC treated with distal ureterectomy. All 9 patients had a history of BTCC. Patient medical records were reviewed and the data ana-

Accepted for publication June 11, 2004. * Financial interest and/or other relationship with Bayer/GlaxoSmithKline, Sanofi-Synthelabo, Novartis and Johnson & Johnson. † Correspondence: Department of Urology, Memorial SloanKettering Cancer Center, 1275 York Ave., New York, New York 10021 (FAX: 212-988-0760; e-mail: [email protected]). 2177

2178

HISTORY OF BLADDER CANCER ALTERS PROGNOSIS FOR UPPER TRACT DISEASE RESULTS

FIG. 1. Patient population and events of disease recurrence and death. Bladder Hx, evidence of bladder TCC before diagnosis with upper tract TCC. Primary, no history of bladder TCC before diagnosis with upper tract TCC. Superf., tumor stage Cis, Ta, T1. Invasive, tumor stage T2-4. DOD, events of disease specific mortality. Bladder rec, events of bladder recurrence after surgery for upper tract disease. Kidney rec, events of kidney surgical bed recurrence after surgery for upper tract disease.

lyzed for possible prognostic predictors such as primary tumor stage, tumor grade, multifocality, carcinoma in situ, evidence of hematuria, presentation with flank pain, sex and history of smoking. Survival time was calculated from definitive treatment of the upper tract disease (time of nephroureterectomy) to the last followup or disease specific death. Median followup was 43 months for the entire group, 48 months for patients presenting with primary UTTCC and 34 months for those with a history of BTCC. Disease specific actuarial survival and freedom from bladder recurrence were performed with the Kaplan-Meier method, and differences among the groups were compared with the log rank test using SPSS statistical software (SPSS, Inc., Chicago, Illinois). For a multivariate analysis of all the predictors we used the Cox proportional hazards regression model and constructed 3 models. Model 1 contained all prognostic factors except history of bladder cancer. Model 2 added history of bladder disease categorized as positive or negative. Model 3 added bladder history categorized as invasive (T2 and above), superficial (T1 or below) or absent. The sample size simulates analysis of codings or interactive examination. The concordance index for each model was calculated using bootstrapping with 200 resamplings. Two sided p values of less than 0.05 denoted statistical significance.

Disease specific death was reported for 44 patients, 12 with invasive primary UTTCC, 1 with superficial primary UTTCC, 19 with invasive UTTCC and history of bladder TCC, and 12 with superficial UTTCC and history of bladder cancer. When the records of all 62 patients with invasive and superficial bladder cancer history were examined 12 of 22 and 19 of 40, respectively, had died of the disease. Bladder recurrence was noted in 32 of all 129 patients. Of the patients with recurrence, 8 had primary invasive UTTCC, 5 had primary superficial UTTCC, 7 had invasive UTTCC and history of bladder cancer, and 12 had superficial UTTCC and history of bladder cancer (fig. 1). Distribution of predictors between primary UTTCC (invasive and superficial) with and without history of bladder cancer was similar, with the exception of tumor multifocality, with 56% and 45% of patients having invasive and superficial UTTCC and history of bladder cancer, respectively, compared to 13% and 14% with primary invasive and superficial UTTCC only (table 1). Using the Kaplan-Meier method for actuarial assessment of disease specific survival, we demonstrated that history of bladder disease was an adverse prognostic factor in patients with invasive and superficial UTTCC (fig. 2, A and B, p ⫽ 0.038 and p ⫽ 0.012, correspondingly). When patients with superficial UTTCC were evaluated, history of bladder cancer appeared as an adverse prognostic factor when there was a history of invasive and superficial bladder tumors (fig. 2, C, p ⫽ 0.0017). Kaplan-Meier analyses of freedom from bladder recurrence (fig. 3) demonstrated that history of bladder disease was significantly associated with bladder recurrence in patients with noninvasive upper tract disease (fig. 3, B, p ⫽ 0.04), whereas in patients with invasive upper tract disease it did not reach statistical significance (fig. 3, A, p ⫽ 0.917). We evaluated the prognostic power of all predictors in multivariate analysis. In model 1 primary tumor stage was significantly associated with disease specific survival (p ⫽ 0.001), as was evidence of hematuria (gross or microscopic) (p ⫽ 0.047). None of the other predictors was significantly associated with disease specific survival (table 2). The overall concordance index for the base model (model 1) was 0.67. Model 2 (table 2), with previous bladder disease included as a predictor for disease specific survival (grouped as positive or negative history), yielded an improved concordance index of 0.70 with primary tumor stage and history of bladder disease significantly associated with disease specific survival (p ⫽ 0.001 and p ⫽ 0.014, respectively). Because the outcome for patients with history of invasive bladder disease might be the result of bladder tumor, we examined model 3 in which history of bladder disease was grouped as invasive, superficial and no bladder disease (table 2). The area under the curve for the model was 0.70 and, again, primary tumor stage and history of bladder disease were significantly asso-

TABLE 1. Distribution of predictors Invasive UTTCC

Superficial UTTCC

% Cis (No. pos/total No.) 31 (11/35) 22 (7/32) % High grade (No. high/low)* 75 (25/8) 74 (23/8) % Multifocality (No. pos/total No.) 13 (4/30) 14 (4/29) No. pos lymph nodes/total No. lymph 5/12 2/13 node dissections % Cytology (No. pos/total No.)† 39 (13/34) 52 (16/31) % Hematuria (No. pos/total No.) 83 (29/35) 69 (22/32) % Flank pain (No. pos/total No.) 14 (5/35) 19 (6/32) % Smoking history (No. pos/total No.) 71 (25/35) 50 (16/32) * High grade— grade 3 of 3 or 4 and grade 4 of 4. Low grade— grades 1 and 2 of 3 or 4. † Grouped as positive vs negative (suspicious and no evidence of malignant cells).

Invasive UTTCC ⫹ Bladder Ca History

Superficial UTTCC ⫹ Bladder Ca History

38 (11/29) 89 (17/2) 56 (14/25) 4/15

27 (9/33) 58 (14/10) 45 (15/33) 1/19

83 (24/29) 55 (16/29) 10 (3/29) 66 (19/29)

64 (21/33) 48 (16/33) 11 (3/33) 85 (28/33)

HISTORY OF BLADDER CANCER ALTERS PROGNOSIS FOR UPPER TRACT DISEASE

2179

FIG. 2. Kaplan-Meier analysis for disease specific survival. A, patients with invasive upper tract disease (stage T2 or greater) grouped by previous history of bladder disease. Patients without history of bladder disease have significantly better survival (log rank p ⫽ 0.038 with bladder history in 29, without in 35, numbers along x-axis demonstrate patients at risk at various times). B, patients with superficial upper tract disease (less than stage T2) grouped by previous history of bladder disease. Patients without history of bladder disease have better survival (log rank p ⫽ 0.0012 with bladder history in 33, without in 31 [1 patient lost to followup], numbers along x-axis demonstrate patients at risk at various times). C, patients with superficial upper tract disease. History of bladder disease categorized as negative, positive superficial disease and positive invasive disease. Categorization divided patients in terms of survival (log rank p ⫽ 0.0017 with invasive bladder history in 10, noninvasive bladder disease in 23, without history of bladder disease in 31 [1 patient lost to followup], numbers along x-axis demonstrate patients at risk at various times).

FIG. 3. Kaplan-Meier analysis for freedom from bladder recurrence. A, patients with invasive upper tract disease. B, patients with superficial upper tract disease. Patients without history of bladder disease have lower rate of bladder recurrence (log rank p ⫽ 0.04, with bladder history in 33, without in 31 [1 patient lost to followup], numbers along x-axis demonstrate patients at risk at various times).

ciated with disease specific survival (p ⫽ 0.001 and p ⫽ 0.018, respectively). DISCUSSION

Although upper tract transitional cell carcinoma is an uncommon tumor,1 the high rate of disease recurrence after treatment of primary disease is significant and represents a challenge to the treating urologist. Bladder recurrence can reach up to 51%,3 while 5-year disease specific survival ranges between 16.5% and 95% depending on stage.2 Predictors implicated in conveying adverse prognosis are tumor grade and disease stage, type of surgical procedure performed3 and tumor multifocality.4 Similar to these studies,2, 3 we found disease stage to be persistently associated with

disease specific survival. However, neither multifocality nor tumor grade reached statistical significance. Several authors have examined the risk of development of upper tract disease after primary bladder disease. The cumulative risk depends on the duration of followup as well as the degree of surveillance. Studies suggest a low risk of about 0.7% to 3% for the development of upper tract disease with a median followup of about 4 years.5– 8 Herr demonstrated that when long-term followup is available for patients who were previously treated with bacillus Calmette-Guerin (307 patients with a median followup of 12 years), the cumulative incidence of upper tract disease is 13% for less than 5 years, 28% for 5 to 10 years and as high as 38% for 10 to 15 years.9 These high incidences of recurrence necessitate a strict fol-

2180

HISTORY OF BLADDER CANCER ALTERS PROGNOSIS FOR UPPER TRACT DISEASE TABLE 2. Multivariate analysis of predictors associated with disease specific survival Model 1

Model 2

Model 3

Predictor

p Value

CI

p Value

CI

p Value

CI

Bladder disease Stage Sex Smoking Hematuria Flank pain Cytology Grade Cis Multifocality

— 0.001 0.195 0.825 0.047 0.965 0.866 0.329 0.936 0.190

0.670

0.014 0.001 0.130 0.692 0.311 0.273 0.644 0.408 0.694 0.526

0.700

0.018 0.001 0.081 0.997 0.127 0.336 0.576 0.458 0.655 0.432

0.70

lowup schedule. In this series of patients nearly 48% had synchronous or previous bladder TCC. Hurle et al reported on patients followed after diagnosis of primary bladder TCC.10 They categorized patients as low risk—low grade, low stage tumor treated with resection only, intermediate risk— recurrent or multifocal disease treated with resection and mitomycin C, and high risk— carcinoma in situ (Cis) or high grade tumor and failure after intravesical chemotherapy. They found that patients in the high risk group had a significantly higher incidence of UTTCC developing compared to low and intermediate risk groups. Similar results were obtained by Milla´n-Rodrı´guez7 and Solsona et al11 demonstrated that history of Cis in a previous bladder TCC event had an impact on the further development of UTTCC. This finding was especially true for those events in the distal ureter, as well as for bilateral disease and urethral extension of the disease. They concluded that Cis is an excellent marker for diffuse disease that is more likely to involve multiple urothelial sites. These reports illustrate the theory that TCC is a panurothelial phenomenon which can yield multifocal tumors and transcend from the lower to the upper collecting system. However, the theory of upper tract cell implantation through vesicoureteral reflux is a possibility as well.12, 13 The development of bladder TCC after UTTCC might be the result of reimplantation of tumor cells as well as of tumor multifocality. This might explain the high incidence of bladder recurrence in UTTCC compared to the lower incidence of UTTCC after primary bladder TCC. In our entire cohort of patients bladder recurrence developed in 25%. If history of bladder disease in a patient diagnosed with UTTCC is a marker for a panurothelial disease, it might explain the higher incidence of bladder recurrence that we found in these patients. Models 2 and 3 of the multivariate analysis in this study demonstrated that history of bladder cancer was significantly associated with adverse prognosis. It can be argued that the results might be swayed by a subpopulation of patients in whom upper tract disease was insignificant compared to bladder disease. We would assume that these patients would be represented by superficial upper tract disease and history of invasive lower tract disease. Overall we had 10 of these cases. We further examined the patients who had a history of superficial bladder cancer and, while controlling for primary upper tract tumor grade and stage, we found that these patients had a significantly decreased disease specific survival compared to those without bladder disease (p ⬍0.001). Similar results were obtained by the Kaplan-Meier analysis curves. Disease specific survival was better for patients with invasive and superficial UTTCC if they did not have a previous event of bladder TCC (fig. 2, A and B, p ⫽ 0.038 and p ⫽ 0.0012, respectively). To eliminate cases of aggressive bladder disease that would influence survival, we further examined the population with superficial upper tract disease and divided them into those with history of invasive or superficial bladder disease and those with no history of bladder

disease. Kaplan-Meier analysis demonstrated that the patient population was grouped into 3 distinctly different populations in terms of disease specific survival (fig. 2, C, p ⫽ 0.0017). History of bladder disease did not alter freedom from bladder recurrence for patients with invasive UTTCC (fig. 3, A, p ⫽ 0.917). The short survival of these patients most likely diminishes the prognostic value of the long-term effect of previous bladder cancer. Patients with superficial UTTCC and history of bladder disease had significantly worse freedom from bladder recurrence (fig. 3, B, p ⫽ 0.04). Since we assume that the rate of tumor recurrence in the bladder due to reimplantation would be similar, it is more likely that the difference in bladder recurrence would be due to the panurothelial nature of the disease. Despite that to the best of our knowledge there has been no previous report on the history of bladder cancer as an adverse prognostic factor for patients with UTTCC, a recent study demonstrated that in a multivariate analysis of those diagnosed with this disease, synchronous bladder cancer was not only an adverse prognostic factor but also a predictor for the further development of contralateral UTTCC.14 Our study is a retrospective review of the patients treated at our institution and, as such, it is prone to all the biases of a retrospective report. Because of the infrequent nature of this disease, prospective studies, although much in need, are scarce. Additional possible predictors such as exact tumor location, bladder tumor in situ and bladder recurrence were not included in the multivariate analysis because they would render it out of sufficient statistical power because of the small sample size and incomplete data. History of bladder disease in a patient diagnosed with UTTCC can suggest a tendency of tumors to develop in the urothelium. The possibility of finding tumor multifocality, which had been previously described as an adverse prognostic factor, caused patients with TCC to be assigned to a high risk group.7, 10 Our findings demonstrate that a history of bladder TCC, even as superficial disease, has an adverse effect on bladder recurrence and patient survival. However, further confirmatory studies with larger sample sizes are needed to establish that history of bladder TCC is indeed a prognostic factor for newly diagnosed UTTCC. CONCLUSIONS

This study suggests that a history of bladder TCC has an adverse effect on the prognosis of patients diagnosed with UTTCC. Disease specific survival and freedom from bladder recurrence are decreased in patients with a history of bladder disease. Until further data are made available we recommend that these patients be considered for more aggressive treatment and a tighter followup schedule. REFERENCES

1. Tawfiek, E. R. and Bagley, D. H.: Upper-tract transitional cell carcinoma. Urology, 50: 321, 1997

HISTORY OF BLADDER CANCER ALTERS PROGNOSIS FOR UPPER TRACT DISEASE 2. Munoz, J. J. and Ellison, L. M.: Upper tract urothelial neoplasms: incidence and survival during the last 2 decades. J Urol, 164: 1523, 2000 3. Hall, M. C., Womack, S., Sagalowsky, A. I., Carmody, T., Erickstad, M. D. and Roehrborn, C. G.: Prognostic factors, recurrence, and survival in transitional cell carcinoma of the upper urinary tract: a 30-year experience in 252 patients. Urology, 52: 594, 1998 4. Hisataki, T., Miyao, N., Masumori, N., Takahashi, A., Sasai, M., Yanase, M. et al: Risk factors for the development of bladder cancer after upper tract urothelial cancer. Urology, 55: 663, 2000 5. Rabbani, F., Perrotti, M., Russo, P. and Herr, H. W.: Upper-tract tumors after an initial diagnosis of bladder cancer: argument for long-term surveillance. J Clin Oncol, 19: 94, 2001 6. Millan-Rodrı´guez, F., Che´chile-Toniolo, G., Salvador-Bayarri, J., Palou, J. and Vicente-Rodrı´guez, J.: Multivariate analysis of the prognostic factors of primary superficial bladder cancer. J Urol, 163: 73, 2000 7. Milla´n-Rodrı´guez, F., Che´chile-Toniolo, G., Salvador-Bayarri, J., Huguet-Pe´rez, J. and Vicente-Rodrı´guez, J.: Upper urinary tract tumors after primary superficial bladder tumors: prognostic factors and risk groups. J Urol, 164: 1183, 2000 8. Balaji, K. C., McGuire, M., Grotas, J., Grimaldi, G. and Russo, P.: Upper tract recurrences following radical cystectomy: an

9. 10.

11.

12.

13.

14.

2181

analysis of prognostic factors, recurrence pattern and stage at presentation. J Urol, 162: 1603, 1999 Herr, H. W.: Extravesical tumor relapse in patients with superficial bladder tumors. J Clin Oncol, 16: 1099, 1998 Hurle, R., Losa, A., Manzetti, A. and Lembo, A.: Upper urinary tract tumors developing after treatment of superficial bladder cancer: 7-year follow-up of 591 consecutive patients. Urology, 53: 1144, 1999 Solsona, E., Iborra, I., Ricos, J. V., Dumont, R., Casanova, J. L. and Calabuig, C.: Upper urinary tract involvement in patients with bladder carcinoma in situ (Tis): its impact on management. Urology, 49: 347, 1997 Amar, A. D. and Das, S.: Upper urinary tract transitional cell carcinoma in patients with bladder carcinoma and associated vesicoureteral reflux. J Urol, 133: 468, 1985 De Torres Mateos, J. A., Banu´s Gassol, J. M., Palou Redorta, J. and Morote Robles, J.: Vesicorenal reflux and upper urinary tract transitional cell carcinoma after transurethral resection of recurrent superficial bladder carcinoma. J Urol, 138: 49, 1987 Kang, C. H., Yu, T. J., Hsieh, H. H., Yang, J. W., Shu, K., Huang, C. C. et al: The development of bladder tumors and contralateral upper urinary tract tumors after primary transitional cell carcinoma of the upper urinary tract. Cancer, 98: 1620, 2003