Evidence (intravesical BCG with maintenance BCG treatments):
An individual patient meta-analysis of nine randomized trials (2,820 subjects with Ta or T1 bladder cancer) that compared intravesical BCG with intravesical MMC was published.
Among trials in which the BCG treatment included a maintenance component, there was a 32% reduction in risk of recurrence (P < .0001) compared with MMC; BCG was associated with a 28% increase in the risk of recurrence when no maintenance BCG was given compared with MMC.
There were no differences in progression or death.
A meta-analysis of nine randomized, controlled trials (2,410 subjects) that compared intravesical BCG to MMC was published.
Progression was seen in 7.67% of the BCG subjects and 9.44% of MMC subjects at a median follow-up of 26 months (P = .08).
When the analysis was limited to trials in which the BCG arm included a maintenance component, the progression rate was significantly lower in the BCG subjects (OR, 0.66; 95% confidence interval, 0.47–0.94; P = .02).
A meta-analysis of the published results of nine randomized, controlled trials that compared intravesical BCG with intravesical chemotherapy in 700 patients with carcinoma in situ of the bladder was published.
With a median follow-up of 3.6 years, 47% of the BCG group had no evidence of disease and 26% of the chemotherapy group had no evidence of disease.
In this meta-analysis, BCG was superior to MMC at preventing recurrence only when maintenance BCG was part of the treatment.
A controlled trial evaluated 384 patients randomly assigned to induction intravesical BCG or induction intravesical BCG followed by maintenance intravesical BCG.
Median recurrence-free survival was 36 months without maintenance BCG and 77 months in the maintenance arm (P < .0001). The risk of disease worsening (progression to T2 or greater disease, use of cystectomy, systemic chemotherapy, or radiation therapy) was greater in the induction arm than in the maintenance arm (P = .04).
Overall 5-year survival was 78% in the induction-only arm versus 83% in the maintenance arm, but this difference was not statistically significant.