European Urology

1. Introduction

Intravesical instillation of Bacillus Calmette-Guerin (BCG) is the most effective adjuvant therapy after transurethral resection of intermediate- or high-risk superficial bladder tumours and is the first-line treatment for patients with carcinoma in situ. The intravesical instillation of BCG reduces the risk of recurrence and delays the time to recurrence compared with transurethral resection alone or other drugs given intravesically. It also significantly reduces the risk of progression after transurethral resection in patients with this kind of tumour who receive maintenance treatment. Nevertheless, treatment with BCG is not free from complications. Irritative local symptomatology can appear in up to 91% of patients as well as other complications at different rates [1], [2], [3], [4], and [5].

BCG efficacy and toxicity are dose-dependent; the problem lies in finding a very low BCG dose that is effective and has low toxicity. In 1991the results of a randomised study [6] to evaluate the effectiveness and toxicity of a 75 mg dose of Pasteur strain BCG in the treatment of superficial bladder cancer were reported; the standard dose of Pasteur strain BCG is 150 mg. The conclusions were that half of the standard BCG dose is effective as adjuvant treatment against recurrent superficial papillary tumours and as treatment of carcinoma in situ. Complete response rates were similar to those achieved using standard dose. Furthermore, treatment-related toxicity appeared to be significantly lower than that reported in previous studies with higher doses. In 1992 the results of a randomised study [7] comparing the standard 120 mg dose of Armand Frappier BCG with a 60 mg dose, were reported. This study suggested that high doses of BCG, and therefore high toxicity, were necessary to eradicate carcinoma in situ; on the other hand lower doses, with less toxicity, might be adequate to prevent tumour recurrences. In 2002 the results of the CUETO (Club Urológico Español de Tratamiento Oncológico) study 90.008 [8] were published. This randomised study compared a standard dose of 81 mg of BCG Connaught strain with a low dose of 27 mg as adjuvant treatment in superficial bladder tumours. The toxicity was lower in the group of patients who had received the 27 mg BCG dose. There were no statistically significant differences related to recurrence and progression between the two groups; nevertheless, the study left some doubts with regards to the low-dose efficacy in high-risk tumours.

To complement the CUETO study 90.008, the same group designed and performed two further studies: CUETO 95.011 and CUETO 95.012.

The randomised study 95.012 [9] compared a standard 81 mg dose of BCG Connaught strain with a low dose of 27 mg in high-risk superficial bladder tumours: carcinoma in situ and T1G3 tumours. The results suggested that a third of the dose of intravesical BCG was as effective as the standard dose against progression in patients with high-risk stages of superficial bladder carcinoma, but with significantly less toxicity.

The primary aim of the present multicentre, randomised CUETO study 95011 is to continue searching for lower doses of BCG that are effective and have low toxicity as adjuvant treatment for superficial bladder cancer. To this end, the outcome of intravesical instillations with 27 mg of BCG Connaught strain was compared with that of instillations with 13.5 mg of BCG Connaught strain, using a group treated with mitomycin C (MMC) 30 mg as a third arm of comparison.

2. Materials and methods

3. Results

3.2. Recurrences

The recurrence rate and the number of recurrences are shown in Table 3. There was a significantly longer disease-free interval for BCG 27 mg versus MMC 30 mg. The Kaplan-Meier curves for the time to first recurrence were given by treatment group in Fig. 1. An overall log-rank test comparing the three groups was done simultaneously (p = 0.002). Unadjusted pair-wise comparisons yielded a significant difference between BCG 27 mg versus MMC 30 mg (p = 0.006); there were no statistically significant differences between BCG 27 mg and BCG 13.5 mg (p = 0.165) or between BCG 13.5 mg and MMC 30 mg (p = 0.183). The results of stepwise Cox proportional hazards regression univariate and multivariate analyses to evaluate the relative importance of individual factors for recurrence and progression are shown in Table 4. The disease-free interval in the univariate analysis was significantly longer for primary disease, unifocal tumour, and treatment with BCG 27 mg. The disease-free interval in the multivariate analysis was significantly better for primary disease and treatment with BCG 27 mg.

Table 3 Patients with recurrences, progression, and cancer-specific death

Results	MMC 30 mg	BCG 27 mg	BCG 13.5 mg
No. of patients with recurrences	58 (38.9%)	38 (26.8%)	50 (36.0%)
No. of recurrences	75	47	63
Recurrence index/100 patients per month	0.957	0.575	0.740
Progression	14 (9.4%)	14 (9.9%)	18 (12.9%)
Cancer-specific death	7 (4.7%)	3 (2.1%)	5 (3.6%)

MMC, mitomycin C; BCG, Bacillus Calmette-Guerin.

Create Platinum Slide Series

Fig. 1 Kaplan-Meier curves for the disease-free interval.

Table 4 Disease-free interval and time to progression (Cox models) by univariate and multivariate analyses

Variable	Disease-free interval		Time to progression
	Univariate (unadjusted) HR (95%CI) [p]	Multivariate (adjusted) HR (95%CI) [p]	Univariate (unadjusted) HR (95%CI) [p]	Multivariate (adjusted) HR (95%CI) [p]
Age
≤65 yr	1		1
>65 yr	1.20 (0.87–1.67) [0.262]		1.34 (0.75–2.42) [0.322]
Disease status
Primary	1	1	1	1
Recurrent	2.43 (1.74–3.39) [<0.001]	2.53 (1.81–3.54) [<0.001]	2.33 (1.29–4.21) [0.005]	2.66 (1.47–4.83) [0.001]
No. of tumours
1	1		1
2–3	1.88 (1.23–2.87) [0.004]		0.82 (0.32–2.11) [0.677]
>3	2.91 (1.98–4.17) [<0.001]		3.00 (1.57–5.75) [<0.001]
Size
≤2 cm	1		1
>2 cm	0.97 (0.70–1.36) [0.880]		0.73 (0.40–1.35) [0.320]
TG Category
TaG2	1		1	1
T1G1	1.69 (0.92–3.10) [0.091]		3.16 (0.66–15.10) [0.150]	3.15 (0.69–15.06) [0.151]
T1G2	1.50 (0.86–2.59) [0.152]		4.79 (1.14–20.06) [0.032]	5.57 (1.32–23.40) [0.019]
Treatment
MMC 30 mg	1.74 (1.15–2.62) [0.008]	1.86 (1.23–2.81) [0.003]	0.98 (0.47–2.06) [0.960]
BCG 27 mg	1	1	1
BCG 13.5 mg	1.35 (0.89–2.06) [0.163]	1.49 (0.97–2.28) [0.065]	1.16 (0.57–2.34) [0.681]

HR, hazard ratio; 95%CI, 95% confidence interval; MMC, mitomycin C; BCG, Bacillus Calmette-Guerin.

4. Discussion

High-risk superficial bladder tumour is the major European indication for BCG to avoid the recurrence and/or progression. The most important prognostic factors for recurrence are the number of tumours, their size, and the prior recurrence rate. The most important prognostic factors for progression are the T category, grade, and the presence of carcinoma in situ—factors that represent the biological aggressiveness of the disease. BCG is relatively less effective for low-grade tumours, independently of the age, and for patients older than 80 yr. In case of BCG failure, the outcome is bad. Standard treatment in failing patients remains cystoprostatectomy. Conservative but investigative alternatives are interferon-alpha, intravesical hyperthermia, or photodynamic therapy. However, treatment with BCG is not free from complication [10], [11], [12], [13], and [14].

Previous prospective studies and review articles have suggested that one third of the standard dose of BCG is as effective as the standard dose, whereas the number of local and systemic side-effects is decreased [6], [7], [8], [9], [15], [16], [17], [18], and [19]. In 1995 a phase 2 study [16] evaluated the feasibility, response, and toxicity of 27 mg BCG dose (Connaught strain) in patients with high-risk superficial transitional cell carcinoma. The results of the study suggested that one third of the standard BCG dose could be successfully used as adjuvant therapy for high-risk patients. Toxicity included local reactions such as severe dysuria, urinary frequency, and significant haematuria. In 2001 the European Organisation for the Research and Treatment of Cancer (EORTC) Genitourinary Group [17] examined the ablative activity and incidence of side-effects of intravesical 30 mg Connaught BCG given for a papillary marker lesion of the bladder. They concluded that 30 mg BCG has a clear ablative effect on superficial bladder cancer with a 61% response rate. Local side-effects included dysuria in 54% of the patients and macroscopic haematuria in 39%. In 2002 a study [18] investigated whether a proinflammatory cytokine interleukin-8 (IL-8) could be used as a predictor for response to a standard dose of 120 mg BCG Danish 1331 strain and to one third the standard BCG dose of 40 mg in patients with superficial bladder cancer. There was no difference with respect to recurrence and progression between the two BCG treatment groups. The quantitative IL-8 response to standard dose and to the one-third dose was similar. In 2002 and 2005 the previously mentioned CUETO 90.008 [8] and 95.012 [9] studies also showed that one third of the standard BCG dose was as effective as the standard dose and was less toxic. In 2004 another phase 2 study [19] also evaluated one third of the standard BCG dose. The conclusion was that this low-dose BCG could be an option as adjuvant therapy for superficial bladder cancer, with acceptable toxicity and good compliance.

In the present study, the MMC 30 mg dose and treatment regimen of 18 wk were a usual treatment at the moment of starting the study. For this reason, we chose the same treatment regimen, induction plus maintenance, for BCG. The most important prognostic factor for recurrence is disease status recurrent, not BCG treatment. The most important prognostic factors for progression are disease status recurrent and T1G2 category. Age is not predictive of response. We want to remark that there are no tumours of high risk and that the comparison is between patients older and younger than 65 yr.

The present study suggests that one third of the standard BCG dose could be recommended as adjuvant treatment for superficial bladder tumours of intermediate risk and is more effective than MMC 30 mg. The disease-free interval in the univariate and multivariate analyses was significantly longer for treatment with BCG 27 mg. The number of recurrences was lower in the BCG 27 mg group than in the MMC 30 mg group: 47 and 75, respectively. A cost assessment revealed that instillations of one third of the standard dose of BCG is less expensive than instillations of MMC 30 mg. The cost of a transurethral resection of a bladder tumour, estimated by the Spanish National Health System for the year 2006, was &z.euro;2194.73 (DRG-311 [Diagnosis Related Group-311] transurethral procedure without complications). The amount spent on each “intravesical instillation procedure” was estimated by the Spanish National Health System at &z.euro;323.82 for the year 2006. Multiplying the recurrence index/100 patients per month by the cost of each transurethral procedure without complications—without considering the effect that transurethral resection has on work, family, and personal aspects—shows that the MMC 30 mg group is more expensive than the BCG group. In the MMC 30 mg group, the cost for 100 patients per month is 0.957 × 2194.73 (DRG 311) = 2099.65&z.euro;. In the BCG 27 mg group, the cost for 100 patients per month is 0.575 × 2194.73 = 1261.96&z.euro;. MMC is more expensive in the long run because there are more recurrences and hence more additional treatments, including transurethral resections, have to be done in the MMC group. It is true that one third of the standard dose of BCG is more toxic than MMC 30 mg, but it is very difficult to estimate the impact of toxicity on the cost of the procedure because the side-effects during the instillation stage are mostly slight, irritative local symptoms, or general symptoms, which, in most cases, cease spontaneously or with the help of anticholinergic medication or paracetamol.

The present study suggests that the minimum effective dose of BCG is one third of the standard dose. We consider that one sixth of the standard dose of BCG is not indicated as adjuvant treatment for superficial bladder cancer of intermediate risk; it has the same efficacy as MMC 30 mg but is more toxic. In fact, the toxicity level of one sixth of the standard dose is similar to that observed with one third of the standard dose.

5. Conclusions

• 1.One sixth of the standard dose, BCG 13.5 mg, is suboptimal compared with one third of the standard dose, BCG 27 mg. Although the one-sixth dose has the same efficacy as MMC, it is more toxic and therefore should not be recommended as initial adjuvant treatment for intermediate-risk superficial bladder tumours.

• 2.One third of the standard dose, BCG 27 mg, seems to be the minimum effective dose as adjuvant treatment for intermediate-risk superficial bladder tumours.

Conflicts of interest

The authors have nothing to disclose.