Maintenance bacillus Calmette-Guerin immunotherapy for recurrent Ta, T1 and carcinoma in situ transitional cell carcinoma of the bladder: A randomized Southwest Oncology Group study

Purpose: Bacillus Calmette-Guerin (BCG) immunotherapy has been widely accep ted as the optimal treatment for carcinoma in situ and high grade superfici al transitional cell carcinoma. However, controversy remains regarding the role of maintenance therapy, and its long-term effect on recurrence and pro gression. Materials and Methods: All patients in the study had transitional cell carc inoma of the bladder with carcinoma in situ or an increased risk of recurre nce. The criteria for increased risk were 2 or more episodes of tumor withi n the most recent year, or 3 or more tumors within 6 months. At least 1 wee k following biopsy of carcinoma in situ and resection of any stage Ta or T1 transitional cell tumors 660 patients were started on a 6-week induction c ourse of intravesical and percutaneous Connaught BCG. Three months followin g initiation of BCG induction therapy 550 consenting patients were stratifi ed by purified protein derivative skin test and the presence of carcinoma i n situ, and then randomized by central computer to receive BCG maintenance therapy (maintenance arm) or no BCG maintenance therapy (no maintenance arm ). Maintenance therapy consisted of intravesical and percutaneous BCG each week for 3 weeks given 3, 6, 12, 18, 24, 30 and 36 months from initiation o f induction therapy. The 384 eligible patients who were disease-free at ran domization constitute the primary intent to treat analytic group because th ey could be followed for disease recurrence. All patients were followed for adverse effects of treatment, recurrence, disease worsening and survival. Results: No toxicities above grade 3 were noted in the 243 maintenance arm patients. The policy of withholding maintenance BCG from patients with incr eased side effects may have diminished the opportunity to observe severe to xicity. Estimated median recurrence-free survival was 35.7 months (95% conf idence interval 25.1 to 56.8) in the no maintenance and 76.8 months (64.3 t o 93.2) in the maintenance arm Clog rank. p <0.0001). Estimated median time for worsening-free survival, defined as no evidence of progression includi ng pathological stage T2 disease or greater, or the use of cystectomy, syst emic chemotherapy or radiation therapy, was 111.5 months in the no maintena nce and not estimable in the maintenance arm (log rank. p = 0.04). Overall 5-year survival was 78% in the no maintenance compared to 83% in the mainte nance arm. Conclusions: Compared to standard induction therapy maintenance BCG immunot herapy was beneficial in patients with carcinoma in situ and select patient s with Ta, T1 bladder cancer. Median recurrence-free survival time was twic e as long in the 3-week maintenance arm compared to the no maintenance arm, and patients had significantly longer worsening-free survival.