Prognosis of Muscle-Invasive Bladder Cancer: Difference between Primary and ProgressiveTumours and Implications fortherapy

European Urology European Urology 45 (2004) 292 296 Prognosis of Muscle-Invasive Bladder Cancer: Difference between Primary and ProgressiveTumours and Implications fortherapy Barthold Ph. Schrier a, Maarten P. Hollander b, Bas W.G. van Rhijn c, Lambertus A.L.M. Kiemeney a,b, J. Alfred Witjes a,* a Department of Urology 426, University Medical Centre, St Radboud, Geert Grooteplein 10, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands b Department of Epidemiology, University Medical Centre, St Radboud, Nijmegen, The Netherlands c Department of Urology, Erasmus Medical Center, University Hospital Dijkzigt, Dr. Molenwaterplein 40, 3015 GD Rotterdam, The Netherlands Accepted 8 October 2003 Published online 29 October 2003 Abstract Objective: To evaluate the difference in prognosis between progressive and primary muscle-invasive bladder cancer. Materials and Methods: From 1986 to 2000, 74 patients with progressive muscle-invasive bladder cancer were identified. Eighty-nine patients with primary muscle-invasive bladder cancer were frequency matched for stage to these patients with progressive disease. Baseline data including patient and tumour characteristics were collected at the time of diagnosis of the muscle-invasive tumour. Duration of survival was defined as time from muscle-invasive bladder cancer diagnosis until disease-specific death. Kaplan Meier curves were drawn to determine the difference in prognosis between the two study groups. To adjust for potential residual confounding due to differences in treatment, 4 subgroups (T2/3, T4, Nþ and Mþ) were constructed according to the TNM classification. In order to see whether age and gender had any effect on outcome, the four stage groups, age and gender were entered in a Cox s proportional hazard regression model. Results: The 3-year bladder cancer-specific survival was 67% in the primary group and 37% in the progressive group (log rank p ¼ 0:0015). Kaplan Meier curves comparing the different stage groups showed a better prognosis for the patients with primary, i.e. pt2/3 or Nþ, tumours at baseline. Cox regression analysis demonstrated that age and gender had no influence on bladder cancer-specific survival. Conclusions: Patients with muscle-invasive bladder cancer and a history of superficial bladder cancer have a worse prognosis than patients with primary muscle-invasive bladder cancer. # 2003 Elsevier B.V. All rights reserved. Keywords: Bladder neoplasms; Transitional cell carcinoma; Disease progression; Prognosis 1.Introduction Transitional cell carcinoma (TCC) of the urinary bladder can present as a superficial or as a muscleinvasive lesion. The majority of patients (approximately 75%) present with relatively benign superficial tumours, which are limited to the mucosa (Ta) or the lamina propria (T1) [1]. Carcinoma in situ (Cis) can also be * Corresponding author. Tel. þ31-24-3616712; Fax: þ31-24-3541031. E-mail address: f.witjes@uro.umcn.nl (J. Alfred Witjes). considered superficial. However, Cis tends to behave more aggressively and is often found in association with high-grade superficial tumours. Tumour recurrence with superficial tumour is quite common and can be as high as 50% to 70% following therapy [2,3]. About 30% ofthese recurrent tumours will progress to a higher grade or stage, with approximately 10% of these recurrent cancers progressing to muscle-invasive disease [4,5].On the other hand, about 15% of patients who present with muscle-invasive tumours have a history superficial carcinoma. However, most of the diagnosed cases of 0302-2838/$ see front matter # 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2003.10.006

B.Ph. Schrier et al. / European Urology 45 (2004) 292 296 293 muscle-invasive bladder cancer (80% to 90%) present without prior history of superficial disease i.e. primary invasive bladder cancer. Patients with primary and progressive muscle-invasive bladder cancer are treated equally in normal daily practice assuming a similar cancer-specific survival for both patient groups. However, to our knowledge, studies on this subject do not exist. In addition, timing of radical treatment (cystectomy) for patients with high risk superficial disease remains a controversial issue. Therefore, we performed a retrospective study to determine the difference in survival between these two patient populations. 2.Patients and methods A retrospective study was performed in patients treated for muscle-invasive bladder cancer at the Department of Urology of the UMC (University Medical Centre), St Radboud Nijmegen and the Department of Urology of UMC Dijkzigt Rotterdam. Medical records of patients treated between 1986 and 2000 were reviewed. Seventy-four patients treated for muscle-invasive bladder cancer, who had shown progression from a superficial tumour, were identified. Diagnosis of superficial bladder cancer was made by histological judgment of material acquired by fractionated transurethral resection (TUR-T) of all obvious tumour. All tumours were TCC. Muscle was present in all specimens and no tumour invasion was found. Tumours were evaluated using WHO grades 1, 2 and 3. Pathological stage was classified according the TNM staging system [6]. All patients received additional intravesical therapy, consisting of at least two courses of bacillus Calmette-Guérin (BCG) or maintenance BCG. At three months all patients were checked for early recurrence or residual tumour. Patients with invasive tumour at this point were excluded in order to prevent misclassification of invasive tumours, missed during the initial TUR-T. Patients with upper urinary tract tumours were also excluded. Eighty-nine patients with primary muscle-invasive bladder cancer were frequency matched for stage to the patients with progressive disease. Patients with stage II (T2N0 1M0) and III (T3N0 1M0) underwent extended pelvic lymph node dissection and radical cystectomy. Time to cystectomy after diagnosis of muscle-invasive bladder cancer was less than 10 weeks for all patients. No patients with stage II or III were treated by an organ preserving approach. Chemotherapy was offered to patients with stage IV (T4N0M0, TanyN2 3M0, TanyN0M1 3, TanyN2 3M1 3). Baseline data including patient and tumour characteristics were collected at the time of diagnosis of the muscle-invasive tumour. Primary endpoint of the study was disease-specific survival, defined as time from muscle-invasive bladder cancer diagnosis until bladder cancer-specific death. Patients known to be alive or lost to follow-up were censored at the last contact date. To analyze possible differences between the primary and the progressive group survival curves were calculated by the Kaplan Meier product limit method. Differences between survival curves were assessed with the log rank test. Survival curves were drawn for the total study population and the four stage groups separately in order to look for possible effect modification. Multivariate analysis using proportional hazards regression modelling was used to exclude any confounding effect of age, gender and stage group. All analyses were performed using the SPSS software package (Version 9.0, Chicago, IL). 3.Results A total of 163 patients were included in the study, 99 from the UMC Nijmegen and 64 from the UMC Rotterdam. Of the total study population, 89 cases showed a primary invasive bladder tumour and 74 patients had a history of treated superficial disease. One patient with a primary invasive bladder tumour and 2 patients with a progressive tumour were lost to follow-up. The last stage of the superficial tumour before progression to muscle-invasive bladder carcinoma was ptag1 in 6, ptag2 in 22, ptag3 in 3, pt1g2 in 17, pt1g3 in 20 and ptis in 6 patients. Baseline characteristics of the study population are shown in Table 1. The mean age of the progressive group, as well as the percentage of male patients is slightly higher than in the primary group. The distribution of the stage groups pt2/3, Nþ and Mþ is comparable for the primary and the progressive tumours. The primary group has almost twice as much pt4 tumours, but numbers are small. Fig. 1 shows the Kaplan Meier survival curve comparing the disease-specific survival for the primary and progressive study groups. The primary group has a statistically significant better disease-specific survival than the progressive group (p ¼ 0:0015). The 3- and 5- year survival rates are 67% (confidence interval [CI] 95% 57 to 77%) and 55% (95% CI 43 to 67%) respectively for patients with a primary invasive tumour and 37% (95% CI 24 to 50% and 28% (95% CI 15 to 41%) respectively for patients with a progressive invasive tumour. After 3 years 17 patients in the progressive group and 42 in the primary group were alive. The number of patients alive after 5 years Table 1 Baseline characteristics of the study group (n ¼ 163) Primary tumours Progressive tumours N 89 74 Mean age (years) 63.3 68.5 Gender (n, %) Male 65 (73%) 60 (81%) Female 24 (27%) 14 (19%) TNM (n, %) pt2/3 48 (54%) 42 (57%) pt4 7 (8%) 3 (4%) Nþ 27 (30%) 21 (28%) Mþ 7 (8%) 8 (11%)

294 B.Ph. Schrier et al. / European Urology 45 (2004) 292 296 1,0 Bladder cancer specific survival,8,6,4,2 0,0 0 2 4 6 8 10 12 14 16 18 Follow-up (years) after diagnosis N+ Fig. 1. Kaplan Meier survival curve comparing primary ( ) and Fig. 3. Kaplan Meier survival curve comparing Nþ primary ( ) and This means that the risk for disease-specific death in the primary group is about half (0.49) the risk in the progressive group at all times during followup. The second model (II), including age and gender, shows that age and gender are no independent predictors of survival in our study. Model III shows that the pathological stage does not have any significant effect on the results. However, as expected, the presence of nodal or distant metastasis significantly increases disease-specific death risks with a hazard ratio of 2.07 (95% CI: 1.27 to 3.40) and 6.77 (95% CI: 3.52 to 13.04) respectively. Due to the small number of patients in this study, interaction terms between study groups and stage groups could not be studied. Fig. 2. Kaplan Meier survival curve comparing pt2/3 primary ( ) and decreased to 9 in the progressive and 21 in the primary group. In the pt2/3 group, the survival curve shows a similar pattern as the curve for the total study population (p ¼ 0:001) (Fig. 2). For the Nþ group, the curve has a similar trend as the total study population and the pt2/3 group (p ¼ 0:03) (Fig. 3). There was no statistically significant difference in disease-specific survival for the pt4 group and the Mþ group between patients with a progressive tumour and the study group (p ¼ 0:28 and p ¼ 0:49, respectively), but numbers are small for these subgroups. Multivariable analysis (Table 2, Model I) shows a hazard ratio of 0.49 (95% CI 0.32 to 0.77) for the primary group versus the progressive tumour group. Table 2 Multivariate analysis of disease-specific survival in muscle-invasive bladder cancer Model I Study group (primary versus Hazard ratio 95% CI 0.49 0.32 0.77 Model II Study group (primary versus 0.50 0.31 0.78 Gender (male versus female) 0.78 0.48 1.29 Age 1.00 0.98 1.02 Model III Study group (primary versus 0.52 0.33 0.82 Stage group (pt4 versus pt2/3) 0.43 0.10 1.82 Stage group (Nþ versus pt2/3) 2.07 1.27 3.40 Stage group (Mþ versus pt2/3) 6.77 3.52 13.04

B.Ph. Schrier et al. / European Urology 45 (2004) 292 296 295 4.Discussion The current study shows a large and clinically significant difference in disease-specific survival between primary and progressive muscle-invasive bladder cancer patients, favouring the primary group. The diseasespecific survival appears to be approximately twice as high in the primary group at all times during followup. The 3- and 5-year survival rates are 67% and 55% respectively for patients with a primary invasive tumour and 37% and 28% respectively for patients with a progressive invasive tumour. This trend in survival difference between the two study groups was observed in the Nijmegen population as well as in the Rotterdam patients: in both centers patients with progressive invasive tumours had a significantly worse prognosis compared to patients with primary invasive tumours. Although this was a retrospective study, with no standard treatment protocol for primary and/or progressive invasive tumours, in both centers no difference was made with regard to advised therapy in case of a primary invasive tumour or a history of previous invasive disease. Moreover, treatment advises (radical surgery in case of invasive non-metastasized tumours, and chemotherapy in case of metastasized tumours) were comparable in both academic centers and constant over the years of the period of the retrospective analysis. Therefore, potential differences in treatment should have been equally distributed between both study groups. This is illustrated by the observation that overall survival results of primary invasive tumours are comparable between the two centres, with literature, and with the 3- and 5-year survival of invasive bladder cancer according to the Netherlands Cancer Registry (60% and 55%, respectively) [7]. Furthermore, the survival differences between the two study groups were comparable between the two centres, which suggest that pooling these results did not cause bias, but enabled us to draw a firmer conclusion. The mechanism behind this observed difference is not so easily understood. Since survival figures of the primary invasive patients is comparable with literature and the results of the Netherlands cancer registry, the explanation for the significant differences between the two groups has to be found in the worse survival of patients with progressive muscle-invasive bladder cancer. One possible explanation could be that in high risk superficial bladder tumours both therapy-sensitive and -insensitive cells coexist. Intravesical therapy may select resistant clones and more aggressive tumour cells may continue to grow and progress to invasive bladder cancer. Another interesting possible explanation could be a recent finding published by El-Abbady et al. [8]. They compared 16 patients with progressive invasive tumours with 20 patients who were diagnosed with primary invasive tumours, all undergoing cystectomy. On meticulous histopathological examination they found that patients who underwent previous transurethral resections had significantly more local spread of malignant cells into the bladder muscle as compared to patients with primary invasive tumours. Since they could demonstrate that intravesical pressure reaches pressures as high as 80 cm water, they suggested some malignant cells penetrate through the denuded urothelium during resection as a result of high intravesical pressures. Similarly, random biopsies during resection of superficial tumours might cause tumour cell implantation at the site of the damaged mucosa, and influence the prognosis. However, two large series clearly demonstrated that the risk of recurrence and the risk of progression is almost the same comparing a biopsy policy and a no biopsy policy [9,10]. Moreover, as was argued above, since both groups were stage matched, this finding potentially could explain progression of high risk superficial disease, but not the difference between the pathologically matched groups of the current study. To our best of knowledge, the current study is the first published on invasive bladder cancer that demonstrates a statistically significant survival advantage for patients without a history of superficial bladder cancer. The findings of this study emphasize one of the most challenging questions in the treatment of patients with superficial bladder cancer: the choice between conservative and radical therapy. The ideal situation would be if we could recognize those tumours with aggressive potential with prognostic factors. However, although numerous studies have been undertaken to identify better prognostic factors of progression to muscleinvasive disease, a clear consensus has not been reached [11]. The most reliable factors affecting tumour progression remain high grade (G3), stage pt1 and pt1 subclassification, concurrent CIS, high recurrence rate, no complete remission with BCG immunotherapy and p53 expression [2,12,13]. The current conservative approach for a patient with a high risk superficial tumour (e.g. pt1g3 with CIS) remains an adequate TUR followed by intravesical BCG therapy. Three recent reports with approximately 100 patients each and a follow-up of more than 10 years clearly concluded that, especially as primary treatment, complete resection and intravesical BCG in high risk disease is the treatment of choice [14 16]. The recent meta-analysis of Sylvester et al. even suggested that maintenance intravesical BCG significantly reduces or

296 B.Ph. Schrier et al. / European Urology 45 (2004) 292 296 delays the risk of progression, and therefore is the drug of choice in intermediate and high risk patients after resection [17]. However, the observation that progressive patients do bad, can also be concluded from this meta-analysis. Looking at the number of deaths due to bladder cancer and at the progression data of these same trials the chance of death due to bladder cancer in case of progression is around 64% with a median follow-up of 2.5 years, irrespective of BCG treatment. Although these numbers do not completely match, the results appear amazingly similar to the 67% 3-year risk of bladder cancer death rate in our report for progressive patients (personal communication with Sylvester RJ). This again emphasizes the bad prognosis of these progressive patients. This risk of conservative therapy in high risk patients, together with the excellent survival data from early cystectomy in high risk superficial bladder cancer indicates that radical surgery remains a realistic treatment option in these patients [18]. In conclusion, the time window for conservative therapy in high risk superficial bladder cancer is small. On the one hand very good long term results with prevention or delay of progression can be achieved with maintenance intravesical BCG, while on the other hand, as soon as progression has occurred, the change of dying due to bladder cancer significantly increases. Therefore, in high risk patients with recurrent or persisting tumour at initial evaluation, radical surgery should be seriously considered. References [1] Prout Jr GR. Management (control) of early bladder lesions. Cancer Res 1977;37:2891 4. [2] Heney NM, Ahmed S, Flanagan MJ, Frable W, Corder MP, Hafermann MD, et al. Superficial bladder cancer: progression and recurrence. J Urol 1983;130(6):1083 6. [3] Kiemeney LALM, Witjes JA, Heijbroek RP, Verbeek AL, Debruyne FMJ. Predictability of recurrent and progressive disease in individual patients with primary superficial bladder cancer. J Urol 1993;150(1): 60 4. [4] Skinner DG, Lieskovsky G. Management of invasive and high-grade bladder cancer. In: Skinner DG, Lieskovsky G, editors. Diagnosis and management of genitourinary cancer. Philadelphia: WB Saunders; 1988. p. 295 312 [Chapter 16]. [5] Kurth KH, Denis L, Bouffioux C, Sylvester R, Debruyne FMJ, Pavone-Macaluso M, et al. Factors affecting recurrence and progression in superficial bladder tumours. Eur J Cancer 1995; 31A:1840 6. [6] Greene FL, Balch CM, Fleming ID, Fritz A, Haller DG, Morrow M, et al. AJCC Cancer Staging Handbook TNM Classification of Malignant Tumors, 6th edition. Union Internationale Contre le Cancer and the American Joint Committee on Cancer. Heidelberg: Springer-Verlag; 2002. p. 335 40 [Part IX]. [7] Visser O, Schouten LJ, van Dijck JAAM, Coebergh JWW. Urological tumours in the Netherlands 1989 1996. Utrecht: Netherlands Cancer Registry; 1999. p. 25 7. [8] El-Abbady AA, Shoukry MS, Hanno AG, Younis LK, Abdel- Rahman M. Repeated transurethral resection of recurrent superficial bladder tumours does it affect the spread and stage of the tumour? Scand J Urol Nephrol 2002;36(1):60 4. [9] Kiemeney LALM, Witjes JA, Heijbroek RP, Koper NP, Verbeek AL, Debruyne FMJ. Should random urothelial biopsies be taken from patients with primary superficial bladder cancer? A decision analysis. Br J Urol 1994;73(2):164 71. [10] van der Meijden A, Oosterlinck W, Brausi M, Kurth KH, Sylvester R, de Balincourt C. Significance of bladder biopsies in Ta, T1, bladder tumours: a report from EORTC-GU Group. Eur Urol 1999; 35(4):267 71. [11] Zlotta AR, Schulman CC. Biological markers in superficial bladder tumours and their prognostic significance. Urol Clin North Am 2000;27(1):179 89. [12] Stein JP, Grossfeld GD, Ginsberg DA, Esrig D, Freeman JA, Figueroa AJ, et al. Prognostic markers in bladder cancer: a contemporary review of the literature. J Urol 1998;160(3):645 59. [13] Smits G, Schaafsma E, Kiemeney LALM, Caris C, Debruyne FMJ, Witjes JA. Microstaging of pt1 transitional cell carcinoma of the bladder: Identification of subgroups with distinct risks of progression. Urology 1998;52(6):1009 13. [14] Pansadoro V, Emiliozzi P, de Paula F, Scarpone P, Pansadoro A, Sternberg CN. Long-term follow-up of G3T1 transitional cell carcinoma of the bladder treated with intravesical bacille Calmette- Guerin: 18-year experience. Urology 2002;59(2):227 31. [15] Patard J, Moudouni S, Saint F, Rioux-Leclercq N, Manunta A, Guy L, et al. Tumour progression and survival in patients with T1G3 bladder tumours: multicentric retrospective study comparing 94 patients treated during 17 years. Urology 2001;58(4):551 6. [16] Davis JW, Sheth SI, Doviak MJ, Schellhammer PF. Superficial bladder carcinoma treated with bacillus Calmette-Guerin: progression-free and disease specific survival with minimum 10-year follow up. J Urol 2002;167(2):494 500. [17] Sylvester RJ, van der Meijden AP, Lamm DL. Intravesical bacillus Calmette-Guerin reduces the risk of progression in patients with superficial bladder cancer: a meta-analysis of the published results of randomized clinical trials. J Urol 2002;168(5):1964 70. [18] Herr HW, Sogani PC. Does early cystectomy improve the survival of patients with high risk superficial bladder tumours? J Urol 2001;166(4):1296 9.