The rationale for combining anticancer drugs has not been applied consistently to use of intravesical agents for treatment of superficial bladder cancer, for which immunotherapeutic BCG and chemotherapeutic mitomycin seem to be a potentially effective combination. We aimed to do a prospective, randomised comparison of BCG alone with that of sequential BCG and electromotive mitomycin in patients with stage pT1 bladder cancer.

After transurethral resection and multiple biopsies, 212 patients with stage pT1 bladder cancer were randomly assigned to: 81 mg BCG infused over 120 min once a week for 6 weeks (n=105); or to 81 mg BCG infused over 120 min once a week for 2 weeks, followed by 40 mg electromotive mitomycin (intravesical electric current 20 mA for 30 min) once a week as one cycle for three cycles (n=107). Complete responders underwent maintenance treatment: those assigned BCG alone had one infusion of 81 mg BCG once a month for 10 months, and those assigned BCG and mitomycin had 40 mg electromotive mitomycin once a month for 2 months, followed by 81 mg BCG once a month as one cycle for three cycles. The primary endpoint was disease-free interval; secondary endpoints were time to progression; overall survival; and disease-specific survival. Analyses were done by intention to treat. This trial has been submitted for registration at the US National Cancer Institute website clinicaltrials.gov.

Median follow-up was 88 months (IQR 63–110). Patients assigned sequential BCG and electromotive mitomycin had higher disease-free interval than did those assigned BCG alone (69 months [95% CI 55–86] vs 21 months [15–54]; difference between groups 48 months [42–54], log-rank p=0·0012). Patients assigned sequential BCG and electromotive mitomycin also had lower recurrence (41·9% [32·7–51·5] vs 57·9% [48·7–67·5]; difference between groups 16·0% [2·7–29·3], log-rank p=0·0012); progression (9·3% [3·8–14·8] vs 21·9% [17·9–25·9]; difference between groups 12·6% [3·0–22·2], log-rank p=0·004); overall mortality (21·5% [13·5–29·5] vs 32·4% [23·4–41·4], difference between groups 10·9% [0·6–21·2], log-rank p=0·045); and disease-specific mortality (5·6% [1·2–10·0] vs 16·2% [6·1–23·3], difference between groups 10·6% [2·5–18·7], log-rank p=0·01). Side-effects were mainly localised to the bladder.

BCG-induced inflammation might increase the permeability of the bladder mucosa such that mitomycin can reach the target tissue more easily and exert its anticancer effect.