Objective To examine whether the outcome of cystectomy for invasive transitional cell carcinoma (TCC) of the bladder was influenced by the type of disease at initial presentation.
Patients and methods The charts of 76 patients treated for TCC by radical cystectomy from 1987 to 1997 in our unit were reviewed. The patients were divided into three groups: group 1 comprised 43 patients with primary invasive disease; group 2 included 12 patients with progression of an initial superficial bladder tumour after failure of conservative treatment; and group 3 comprised 21 patients who had a radical cystectomy for superficial TCC, with a high risk of progression after attempts at conservative treatment. The pathological findings on transurethral resection and cystectomy specimens, cancer-specific survival and the time to progression were compared among the three groups.
Results The rate of pT0 in cystectomy specimens was 16%, 41% and 24% in groups 1, 2 and 3, respectively. Under-staging occurred in 24% of cases in group 3. The 10-year cancer-specific survival rates were 48%, 47% and 82% in groups 1, 2 and 3, respectively. The cancer-specific survival rate and progression rate were not significantly different between groups 1 and 2, but were significantly lower/higher in these patients than in group 3 (P < 0.01).
Conclusions These data suggest that the prognosis of superficial TCC which progresses despite conservative management is no better than that of invasive TCC at initial presentation, despite the closer follow-up received by the former patients. Early identification of this group of patients may improve the cancer-specific survival, as early cystectomy for high-risk superficial TCC yields better results.
TCC of the bladder has a wide spectrum; epidemiological studies clearly delineate several patterns of occurrence and progression [1–3]. Since the pioneering work of Jewett, this spectrum has been divided into two entities, i.e. superficial and invasive bladder cancer. Since then, the disease has been better characterized at presentation in a given patient with progress in clinicopathological staging and the advent of molecular markers, but the main practical issue remains the identification of those patients whose disease is amenable to conservative management and those who require radical therapy.
Radical cystectomy remains the standard treatment for patients with invasive TCC of the bladder. Chemo-radiotherapy is an alternative for infiltrating tumours  but the long-term results have yet to be confirmed. TURBT with close surveillance is only feasible for a specific subset of patients [5–7]. In the urological community there is a broad consensus supporting radical cystectomy for invasive TCC, but the approach to superficial disease is more controversial, the initial management being critical for the outcome. Most superficial bladder cancers at a low and intermediate risk of progression are well managed by TURBT and intravesical instillations. However, in the special subgroup of patients with high-risk superficial disease, conservative treatment is the preferred option, but requires life-long surveillance, with a 30% risk of cancer-specific death and a 50% cystectomy rate . Radical cystectomy for high-grade superficial bladder cancer is an aggressive but highly effective alternative, with a 5-year survival rate of 80% [9,10]. Difficulties arise in identifying patients whose superficial disease will progress and who are thus candidates for radical cystectomy. As failure can only be identified when evaluating the response to intravesical therapy , the fate of this subgroup of patient is not well documented.
We thus compared the outcome of patients treated by radical cystectomy for TCC according to their mode of presentation. Three groups were defined, i.e. patients with primary invasive tumours, those in progression despite conservative treatment and those with persistent high-risk superficial TCC after attempts at conservative treatment.
Patients and methods
The files of 76 patients who underwent radical cystectomy with bilateral pelvic node dissection for TCC of the bladder in our department between 1987 and 1997 were reviewed. Radical cystectomies were performed by the two senior surgeons (D.K.C. and C.C.A.). During this period the change in surgery was mainly in the type of urinary diversion, with an increasing number of ileal neobladders constructed compared with cutaneous trans-intestinal derivation.
Three groups of TCC were defined according to the clinical history before radical cystectomy. Group 1 included 43 patients presenting with primary invasive tumours and no history of superficial TCC; the clinical staging was based on TURBT before cystectomy and invasive tumours defined as T2. Group 2 included 12 patients with invasive TCC and a previous history of superficial TCC which had progressed, despite conservative treatment based on repeated TUR and prophylactic intravesical BCG and/or mitomycin instillations. Progression was defined as muscle invasion on TUR and was the indication for radical cystectomy. In our department, BCG and mitomycin C instillations have been used since 1984. Group 3 comprised 21 patients who had radical cystectomy for superficial TCC, indicated by the persistence of high-risk tumour after attempts at conservative treatment. High-risk tumours were grade 3 and associated with lamina propria invasion, carcinoma in situ and multifocal lesions.
During the study period conservative treatment of intermediate- and high-risk TCC consisted of TURBT followed by a weekly intravesical instillation of 150 mg of BCG (Pasteur strain) for 6 weeks. If there was a late recurrence a second series of intravesical instillations was given. Radical cystectomy was considered if there was early recurrence (< 3 months) or progression. If there was intolerance to BCG, mitomycin C (40 mg) was used weekly for 8 weeks.
In patients with initially superficial disease (group 2 and 3), clinical staging was recorded at the first and last presentation before radical cystectomy. In group 2 the mean time to progression was defined as the interval between the first evidence of superficial TCC and the last endoscopic resection before radical cystectomy. The interval between the first and the last resection before radical cystectomy was also noted in group 3. In the three groups the histopathological aspect of the cystectomy specimen was analysed, including pathological tumour stage (pT), grade, and nodal or vascular involvement. The number of stage pT0 tumours on the cystectomy specimens was noted in the three patient groups. Histological grading followed the WHO criteria .
The clinicopathological variables of three groups were compared using the chi-square test. The disease-free survival time was defined as the interval between the date of surgery and the onset of local or metastatic recurrence. Cancer-specific survival was defined by death from bladder cancer. Censored survival values represent patients who were still alive at the last follow-up or those who died from unrelated causes. The cancer-specific and tumour-free survival rates in the three groups were estimated using the Kaplan-Meier method, and statistical differences determined using the log-rank test, with P 0.05 considered to indicate significant differences.
The results of clinical staging and histopathological analysis are shown in Table 1. There were no significant differences in histopathological findings at first presentation in groups 2 and 3 (Table 1) nor between groups 1 and 2. In group 2 the mean (range) time to progression was 57 (6–154) months. In group 3 the interval between the first evidence of TCC and the decision for radical cystectomy was 29 (4–132) months. The respective mean follow-up was 49, 55.3 and 52.5 months in groups 1, 2 and 3. The 10-year cancer-specific survival rates were, respectively, 48%, 47% and 82% in group 1, 2 and 3 (Fig. 1a). The progression-free survival rates at 10 years were 49% in group 1 and 2 and 77% in group 3 (Fig. 1b). There was no statistical difference in the progression-free and cancer-specific survival rates between groups 1 and 2, but both rates were significantly lower/higher in group 3 than in groups 1 and 2 (P < 0.01).
Table 1. Histopathological features of the three groups, with the histopathological features of the TCC of groups 2 and 3 at initial presentation
Mean age, years
Clinical stage/pathological stage, n (%)/n (%)
2 (10)/1 (5)
3 (14)/3 (14)
16 (76)/7 (33)
30 (70)/3 (7)
12 (28)/15 (35)
1 (2)/10 (23)
N + Initial presentation, n
In group 1, two patients died on the first day after surgery (venous thrombosis of an ileal loop and massive cardiac infarction). Six patients died with metastases within 6 months after surgery. One of these patients was staged pT0N0 on the cystectomy specimen. The remaining five patients had high tumour stages (pT3 to pT4) with nodal involvement on the cystectomy specimen. None of these patients had metastases at the evaluation before cystectomy. There were no deaths soon after surgery in groups 2 and 3; one patient in group 2 died with metastases 3 months after surgery; the tumour had been staged pT4 on the cystectomy specimen, with positive lymph nodes.
The course of TCC of the bladder is unpredictable [1–3,8]; superficial TCC of the bladder (pT1, pTa and carcinoma in situ, CIS) may recur in the same form for many years or change into invasive or metastatic tumour. The profile of development is usually estimated by initial endoscopy and histopathological analysis; superficial bladder cancers are classified as low, intermediate or high risk, according to the grade, size, number of foci and time to recurrence. High-risk tumours are managed conservatively, with complete TUR and prophylactic intravesical BCG instillations. However, superficial tumours with similar histopathological risks of progression may have completely different outcomes, and some patients with initially superficial TCC of the bladder continue to develop invasive cancer. This subgroup of patients represent only a small minority of patients with invasive TCC, and the histopathological features of these tumours are poorly documented. Brawn  showed that invasive carcinoma developing in a patient with a history of superficial tumours arises at a different location from the initial site in half the cases. Kaye and Lange  reported that only 16% of patients with invasive TCC had a previous history of superficial disease and provided evidence that the depth of invasion and the grade of the invasive TCC were greater in primary invasive TCC. However, the TCC was analysed histologically before cystectomy. In the present study, 12 (22%) of the 55 patients treated by radical cystectomy for invasive TCC had previous superficial disease. Despite their closer follow-up, patients in group 2 had similar clinical stages to patients in group 1. There was an important rate of down-staging in group 2, with five of the 12 patients found to be pT0. According to Thrasher et al., a pT0 cystectomy specimen confers no survival advantage for patients operated for invasive TCC. The present data support this view, as survival in group 2 was similar to that in group 1, despite more frequent down-staging.
The overall cancer-specific survival rate in patients with invasive TCC is in keeping with those reported previously [10,16,17]. The cancer-specific survival rate in patients with superficial TCC treated by radical cystectomy may differ slightly among studies [9,10,18], probably because of the difficulty in accurately identifying muscle invasion in this group of patients. Pagano et al. underlined the high rate of clinical understaging (35%) among patients treated by cystectomy for T1 tumours. In the present study the rate of understaging was lower (24%).
Cancer-specific survival among patients with invasive tumours and a previous history of superficial disease is poorly documented, and published data are conflicting. Cookson et al. and Dinney et al. studied the long-term outcome of high-risk superficial TCC and reported 10-year cancer-specific survival rates of 64% and 28%, respectively, if there was progression. The outcome has not been compared with that of primary invasive tumours treated by radical cystectomy. In the present study the initial survival curve in group 2 was intermediate between groups 1 and 3, but finally joined that of primary invasive tumours. Although there were few patients, the long-term survival of groups 1 and 2 appeared to be similar, with superimposed curves. This suggests that despite a closer follow-up, the outcome of invasive TCC in patients with a previous history of superficial tumours is no better than that in patients with primary invasive tumours, and highlights the poor prognosis of invasive bladder cancer treated by radical cystectomy. Identifying patients with initially superficial tumours who will go on to develop invasive disease, and the correct timing of radical cystectomy, are major challenges, as cancer-specific survival decreases dramatically once the tumour has invaded the muscle.
The effects of intravesical BCG on progression and long-term survival of patients with superficial TCC is still debated. While numerous studies have shown a significant decrease in the recurrence rate, the prevention of stage progression is not widely reported. In a 15-year randomized study, Cookson et al. found no significant difference in survival advantage or progression rate among patients treated for superficial TCC by TUR plus BCG than in those treated with TUR only. Similarly, adjuvant chemotherapy for superficial TCC has been shown to reduce the rate of recurrence without influencing the risk of progression and the specific survival .
The conclusions of the present retrospective study are limited by the size of the sample and cannot be used to induce changes in the management of high-risk superficial cancer, but they stress the need to better predict clinical aggressiveness. New molecular markers could be used to predict the progression of high-risk superficial TCC and pT2 tumours [20–24]. Recent advances in the molecular characterization of bladder cancer have identified potential markers of tumour progression. Mutation of p53 or p53 nuclear overexpression is one of the most promising biological determinants. However, these markers have yet to be tested prospectively to provide concrete clinical recommendations. For the time being, the main indication for radical cystectomy for high-risk superficial TCC remains early recurrence after attempts at conservative treatment.
In conclusion, the high rate of understaging in patients with high-risk superficial bladder cancer, and the poor survival rate of patients with tumours that become invasive despite conservative management, highlights the need for effective criteria of clinical aggressiveness in bladder cancer at initial presentation.
D.K. Chopin, Service d'Urologie, CHU Henri Mondor, 51 Avenue du Maréchal de Lattre de Tassigny, Créteil 94010 France. e-mail: email@example.com