Eiji Kikuchi, Department of Urology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-8582, Japan. e-mail: email@example.com
Study Type – Therapy (case series)
Level of Evidence 4
What's known on the subject? and What does the study add?
So far, few previous reports have analysed the risk factors for tumour recurrence and stage progression with a special focus on BCG-relapsing disease, defined as the recurrence after achieving a disease-free status by initial BCG instillations for 6 months. There are no guidelines outlining a specific treatment strategy for BCG-relapsing disease, although many BCG failure cases are attributable to BCG-relapsing disease.
In this study, additional BCG instillation was shown to decrease the subsequent tumour recurrence rate against BCG-relapsing tumours with intermediate pathological risk features; however, a BCG-relapsing tumour with a pathologically high risk was a significant risk factor for both subsequent tumour recurrence and stage progression. This information might identify a therapeutic strategy for BCG-relapsing tumours.
• To investigate the risk of subsequent tumour recurrence and stage progression in bacillus Calmette-Guérin (BCG)-relapsing non-muscle-invasive bladder cancer, defined as recurrence after achieving a disease-free status for 6 months.
PATIENTS AND METHODS
• A total of 183 patients with BCG-relapsing tumours were treated with conservative therapy between 1985 and 2008 at our three institutions.
• We analysed the association between their clinicopathological parameters and subsequent tumour recurrence or stage progression.
• Additional induction courses of BCG or anticancer drug (mitomycin C or epirubicin) instillations were performed in 119 patients and 24 patients, respectively. The remaining 40 patients did not undergo any adjuvant therapy.
• Multivariate analysis showed that a relapsing tumour with a pathologically high risk (defined as tumours with G3 and/or pT1 and/or concomitant carcinoma in situ) was a significant risk factor for subsequent tumour recurrence (P= 0.002; hazard ratio [HR] 2.15). Additional BCG instillation significantly decreased the subsequent tumour recurrence rate (P < 0.001; HR 0.41).
• Multivariate analysis also showed that a relapsing tumour with a pathologically high risk was also significantly associated with stage progression (P < 0.001; HR 8.05).
• An additional course of BCG instillation might be effective in patients with BCG-relapsing tumours with pathologically intermediate risk.
• Nevertheless, some patients with high-risk pathological features developed subsequent stage progression. Such patients should be followed up closely and counselled on the need for aggressive therapeutic options, such as radical cystectomy.
Initially diagnosed intermediate- or high-risk non-muscle-invasive bladder cancer (NMIBC) with stage Ta or T1 papillary carcinoma and/or carcinoma in situ (CIS) are usually treated by transurethral resection of bladder tumour (TURBT) and adjuvant BCG immunotherapy to prevent tumour recurrence. Based on the results of several randomized trials, it has become obvious that intravesical BCG instillation, with or without maintenance therapy, is the effective therapy of choice for intermediate- or high-risk NMIBC, and BCG is now known to be superior to other intravesical agents for the prevention of tumour recurrence [1–4]. However, some patients treated with BCG therapy still suffer from recurrence and, in the most general sense, any disease appearance after BCG therapy can be referred to as ‘BCG failure’. Current therapeutic options for BCG failure cases are radical cystectomy or additional intravesical therapy (repeat BCG treatment or other intravesical chemotherapy) . In previous reviews of patients who underwent radical cystectomy for high-risk NMIBC after BCG failure, 5-year disease-specific survival rates ranged from 60 to 79% [6–8]. Nevertheless, despite the potentially curative nature of cystectomy, many physicians and patients may prefer to delay cystectomy in favour of preserving bladder function. O'Donnell and Boehle  reviewed previous published studies and calculated that ≈35% (76/157) of patients failing the first course of BCG achieve durable success with another BCG cycle. In other studies that examined additional chemo-instillation for BCG failure cases with shorter follow-up periods, the recurrence-free rates at 24 months ranged from ≈20 to 30% with gemcitabine  or mitomycin C (MMC) . Steinberg et al.  performed intravesical instillation of valrubicin in 90 CIS patients with BCG failure and reported a complete response rate at 6 months of 21%.
In the above-mentioned studies, BCG failure was inconsistently defined and included very heterogeneous populations with various natural histories. Herr and Dalbagni  noted that comparisons between therapies have been hampered by the lack of a standard definition for BCG failure. To provide more uniformity in reporting, Nieder et al.  advocated that the following alternative descriptive terms for specific types of BCG failure should be used whenever possible. They divided BCG failure into four different defined types. BCG-refractory disease is the term used when there is failure to achieve a disease-free state by 6 months after initial BCG therapy because of either persistent or rapidly recurrent disease. BCG-resistant disease is recurrence or persistence of disease at 3 months after the induction cycle. It is of lesser degree, stage or grade, and is no longer present at 6 months from BCG retreatment. BCG-relapsing disease is defined as recurrence of disease after achieving a disease-free state by 6 months. BCG-intolerant disease is the term used when disease recurs after a less-than-adequate course of therapy is applied because of a serious adverse event or symptomatic intolerance.
There is growing consensus that early cystectomy is beneficial for BCG-refractory disease . By contrast, it is unclear how we should treat BCG-relapsing disease, although many BCG failure cases are attributable to BCG-relapsing disease. In the present study, we reviewed and focused on 183 patients with BCG-relapsing tumours, who were treated with conservative therapy, to analyse their subsequent tumour recurrence and stage progression. These data should assist with the establishment of an appropriate therapeutic strategy for patients with BCG-relapsing disease.
PATIENTS AND METHODS
We identified a total of 280 patients with BCG failure with NMIBC between 1985 and 2008 at our three institutions, 253 of whom were treated with conservative therapy (45 patients with BCG-refractory, three patients with BCG-resistant, 22 patients with BCG-intolerant, and 183 patients with BCG-relapsing disease). A retrospective study was then conducted in the 183 patients with BCG-relapsing disease (149 men and 34 women) with diagnosed NMIBC (pTa or pT1) recurrence after a previous complete response to a single induction course of BCG therapy (6–8 instillations). The median (range) follow-up interval after BCG relapsing was 5.1 (0.4–15.2) years and the mean (range) age of the patients at BCG relapsing was 67.3 (31.8–89.7) years.
TREATMENT AND FOLLOW-UP
These cases were routinely assessed by urine cytology and cystoscopy every 3 months for the first 2 years after TURBT, every 6 months for the next 3 years, and every 6–12 months thereafter. I.v. urography, ultrasonography, and/or CT were used to evaluate the bladder and upper urinary tract every year for the first 5 years and then every 1 or 2 years thereafter. BCG treatment at a dose of 81 mg (Connaught strain), 80 or 40 mg (Tokyo 172 strain), MMC at a dose of 30 mg, or epirubicin at a dose of 50 mg was begun 4 to 5 weeks after TURBT and continued weekly for 6–8 weeks. The starting point of this study was the timing of TURBT for a relapsing tumour and the endpoint was subsequent tumour recurrence or stage progression. Recurrence was defined as a new tumour appearing in the bladder after clearance. Progression was defined as confirmed histological muscle invasion (pathological stage T2 or higher disease) or detectable distant metastases.
CLINICOPATHOLOGICAL RISK STRATIFICATION
The characteristics of tumour grade, number of tumours, pathological stage and the presence of CIS were determined. Referring to the current published AUA guidelines  we divided the patients into three clinicopathological risk groups. We constructed this risk stratification by interpreting low grade as G1-2 and high grade as G3. Tumour size was excluded from the analysis because some of the data contained inaccurate measurements. The pathological high risk cases were defined as tumours with G3 and/or pT1 and/or concomitant CIS. Multiple and/or recurrent G1-2 pTa tumours were considered to be intermediate-risk tumours. Initial and solitary tumours with G1-2 pTa were considered to be low-risk tumours; therefore, the BCG-relapsing tumours were inevitably classified as high- or intermediate-risk tumours because they were all recurrent tumours.
The chi-squared test was used to analyse the difference between two groups. Recurrence-free and progression-free survival curves were constructed using the Kaplan–Meier method, and were compared with the log-lank test. Differences among groups were regarded as significant when P < 0.05. To determine risk factors for subsequent recurrence or progression, univariate and multivariate analyses were performed using the Cox proportional hazards model with stepwise forward selection.
PATIENT CHARACTERISTICS AND TREATMENT
For patients with BCG relapsing, additional induction courses of BCG instillations were performed in 119 patients, and 24 patients received intravesical chemotherapy (18 MMC and six epirubicin). The remaining 40 patients did not undergo any adjuvant therapy. BCG relapsing was further defined by time of recurrence as early (within 12 months), intermediate (12 to 24 months) or late (>24 months) . The clinical characteristics of the patients with BCG-relapsing tumours are listed in Table 1. No adjuvant therapy tended to be chosen in patients with a clinicopathological intermediate-risk tumour in the relapsing tumour group, compared with the BCG and chemotherapy groups. In the additional BCG therapy group for BCG-relapsing tumours (119 patients), 60 patients (50.4%) had clinicopathologically high risk tumours in the previous tumours and 66 patients (55.5%) had clinicopathologically high risk tumours in the BCG-relapsing tumours. Among the 24 patients treated with additional chemotherapy instillation for a BCG-relapsing tumour, 18 (75.0%) had clinicopathologically high risk tumours in the previous tumours and 13 (54.2%) had clinicopathologically high risk tumours in the BCG-relapsing tumours.
Table 1. Clinicopathological backgrounds of patients with BCG-relapsing disease
Total no. of patients
Intravesical therapy for BCG-relapsing tumours
Statistical difference of populations; *BCG group vs chemotherapy group P < 0.05; †BCG group vs no adjuvant group P < 0.05; ‡Chemotherapy group vs no adjuvant group P < 0.05.
Mean age at BCG relapsing, years
Gender, n (%)
Clinicopathological risk group of the previous tumour, n (%)*‡
Overall, 82 patients with BCG-relapsing disease (44.8%) experienced subsequent tumour recurrence during the follow-up periods. To identify significant risk factors for subsequent tumour recurrence, we performed univariate and multivariate analyses using the Cox proportional hazard model with stepwise forward selection. As shown in Table 2, additional BCG instillation was shown to significantly decrease the rate of subsequent tumour recurrence (P < 0.001; hazard ratio [HR], 0.41; 95% CI, 0.23–0.72), and a clinicopathologically high risk of the relapsing tumour (P= 0.002; HR, 2.15; 95% CI, 1.33–3.49) was selected as a significant unfavourable risk factor for subsequent tumour recurrence. As seen in Fig. 1a, the 5-year recurrence-free survival rates were 59.8% in the BCG group, 21.6% in the chemotherapy group, and 43.1% in the no adjuvant therapy group. In the 119 patients who underwent an additional induction course of BCG therapy, the 5-year recurrence-free survival rate was 77.6% in the intermediate-risk group and 46.3% in the high-risk group (P= 0.025 [Fig. 1b]).
Table 2. Univariate and multivariate analyses for tumour recurrence and stage progression in the BCG-relapsing tumour group
Male (n= 149) vs female (n= 34)
Clinicopathological risk group of the prior tumour
High (n= 95) vs intermediate /low (n= 88)
Timing of relapsing
Early (n= 56) vs intermediate/late (n= 127)
Clinicopathological risk group of the relapsing tumour
High (n = 89) vs
intermediate (n = 94)
BCG (n = 119) vs
chemotherapy (n = 24) vs
not done (n = 40)
SUBSEQUENT STAGE PROGRESSION
In the follow-up periods, subsequent stage progression was observed in 25 patients (13.7%). In 19 (76.0%) of these 25 patients, stage progression was found with muscle-invasive bladder tumours without any metastases. Radical cystectomy was performed in 12 patients, three of whom later died from bladder cancer. Three other patients were treated with local radiation therapy; two patients were still alive, and one died from bladder cancer. One elderly patient also died from bladder cancer after being treated with supportive care alone. The treatments and prognoses of the remaining three patients who had local invasive disease without distant metastases and were treated at other hospitals are unknown. Meanwhile, six of 25 stage progression cases (24.0%) were diagnosed with metastatic disease. Five were treated with systemic chemotherapy and one elderly patient was treated with supportive care alone; however, all six eventually died from bladder cancer. Univariate and multivariate analyses were performed again to evaluate the risk factors for stage progression (Table 2). The results indicated that a clinicopathologically high risk relapsing tumour (P < 0.001; HR, 8.05; 95% CI, 2.74–23.6) was significantly associated with subsequent stage progression. The 5-year progression-free survival rate was 97.7% in the intermediate-risk group and 72.2% in the high-risk group (P < 0.001 [Fig. 1c]). Kaplan–Meier curves of the cancer-specific survival rates showed the 5-year cancer-specific survival rates were 98.9% and 87.9% in the intermediate- and high-risk groups, respectively (P= 0.006 [Fig. 1d]). However, we could not find significant differences in the progression-free survival rates (Fig. 1e) and cancer-specific survival rates (Fig. 1f) among the BCG group, chemotherapy group and no adjuvant group.
In the present study, special attention was focused on BCG relapsing, which is defined as the recurrence of disease after achieving a disease-free status by 6 months. Until now, few previous reports have analysed risk factors for subsequent tumour recurrence and stage progression with a special focus on BCG-relapsing disease. In the present study, we evaluated 183 patients with BCG-relapsing disease treated by conservative therapy with a long follow-up. An important finding was that a BCG-relapsing tumour with a pathologically high risk was a significant unfavourable factor for both subsequent tumour recurrence and stage progression, and it could be used to clearly differentiate the prognosis. Additional BCG instillation proved to decrease the subsequent tumour recurrence rate compared with the chemotherapy group and the no adjuvant group.
A few previous reports have described treatment options for BCG-relapsing disease. Repeat BCG treatment for BCG-relapsing tumours was reported by Brake et al.  who found that 19 of 24 patients who had BCG-relapsing tumours remained tumour free, while four patients experienced stage progression with a median interval of 2 years, although their definition of a BCG-relapsing tumour was recurrence after achieving disease-free status 3 months after the first BCG therapy. Bui et al.  also described 11 patients with BCG-relapsing tumours (recurrence 9–74 months after the first BCG therapy) who received a second induction course of BCG, all of whom experienced tumour recurrence after a 6-month disease-free status. Five of their 11 patients continued to be disease-free until the last follow-up at a median of 87 months.
There are no guidelines outlining a specific treatment strategy for BCG-relapsing disease. The AUA guidelines recommend radical cystectomy for BCG failure cases, but they also state that repeat BCG intravesical therapy may be appropriate in patients who develop a late recurrence after a previous complete response to an intravesical agent . The National Comprehensive Cancer Network guidelines recommend radical cystectomy, changing the intravesical agent to MMC, valrubicin, gemcitabine or BCG plus interferon, or maintenance BCG to treat tumour recurrence after BCG therapy . The European Association of Urology guidelines strongly recommend immediate cystectomy for BCG failure because of the high risk for stage progression. They also state that changing from BCG to intravesical chemotherapy or device-assisted chemotherapy instillations can yield responses in selected cases . Boström et al.  commented, in their review, that if there is a significant tumour-free period, re-induction of BCG may be considered. However, clinical data to support these recommendations are insufficient and further investigation and validation of results seem to be warranted to establish the efficacy of these second-line therapies, especially for BCG-relapsing cases.
An additional course of BCG instillation might be effective in patients with BCG-relapsing tumours that have intermediate-risk pathological features as this may help prevent subsequent tumour recurrence to a significant degree compared with the chemotherapy group and the no adjuvant group. We concluded that they could be relatively safely treated by conservative therapy with additional BCG instillations, and few patients experienced subsequent stage progression. However, some patients with BCG-relapsing tumours with high-risk pathological features treated with conservative therapy developed subsequent stage progression, and the 5-year cancer-specific death rate was 12.1%. Such patients should be followed up closely and counselled on the possible need for aggressive therapeutic options, such as radical cystectomy.
The present study has several limitations. First, it was performed in a retrospective fashion and in a limited number of patients. The preference of the attending doctor dictated what kind of intravesical adjuvant therapy was to be performed. Disparities in the backgrounds of the patients among the groups of intravesical therapies could prevent a fair comparison between these groups and may have introduced bias into the results. Another limitation is that a second TUR was not performed in this period. Furthermore, maintenance BCG therapy was not performed in patients in the present study, which is increasingly accepted owing to its positive impact on both recurrence and progression , and this may limit the generalizability of the results. The results of the meta-analysis performed by Sylvester et al.  showed that the effect of BCG on progression was limited to the trials in which some form of BCG maintenance was used, and no reduction was seen in the trials where maintenance BCG was not used. Therefore, if maintenance BCG therapy is performed for a BCG-relapsing tumour, we might be able to reduce the subsequent tumour recurrence rate and even the stage progression rate more than the present results. In a future study, it would be of great interest to evaluate the effect of maintenance BCG therapy in patients with a BCG-relapsing tumour for reducing subsequent tumour recurrence and stage progression.
CONFLICT OF INTEREST
There are no funding sources. There are no financial disclosures from any authors.