Phase II Trial of Atezolizumab in BCG-Unresponsive Non-Muscle Invasive Bladder Cancer (SWOG 1605)

National Cancer Institute/Canadian Cancer Trials Group and Southwest Oncology Group
Study Design: 
This is a single-arm Phase II trial in which patients with BCG-unresponsive high-grade non-muscle muscle invasive bladder cancer (CIS ± Ta/T1 OR Ta/T1) received one year of atezolizumab anti PDL-1 therapy intravenously every 3 weeks x 17 cycles. Patients were evaluated every 3 months with cystoscopy and cytology with a mandatory biopsy done at 6 months.
Immunotherapy and in particular anti PD-1 and anti PD-L1 therapies have demonstrated efficacy in metastatic urothelial cancer and in muscle-invasive bladder cancer. Therefore, these drugs are now being evaluated alone or in combination with other agents in BCG-unresponsive non-muscle invasive bladder cancer. The PD1 inhibitor pembrolizumab was approved in January 2020 and this trial is the first reporting results of a PD-L1 inhibitor (atezolizumab) in this high risk patient population.
The primary endpoint of the trial was the pathological CR rate at 6 months in patients with BCG-unresponsive CIS determined by a mandatory biopsy. The secondary endpoints included: 18-month event-free survival in the Ta/T1 cohort, progression-free survival in all patients, cystectomy-free survival in all patients, bladder-cancer specific survival in all patients, and overall survival in all patients.
The trial is a herculean effort to treat non-muscle invasive bladder cancer with systemic immunotherapy. Unfortunately, the strict statistical criteria for closure of this trial led to its falling short of its primary endpoint as the expectation of novel therapies in this disease space should provide a higher 6-month CR rate in patients with CIS. However, the 42% 3-month CR noted was similar to that seen with pembrolizumab which has been FDA-approved although the duration of response for atezolizumab is still pending at this time.
The trial launched in 2017 but had a required futility analysis after 25 eligible CIS patients reached the 6 month-endpoint. In order to proceed, at least 7 patients had to have a CR and only 5 patients actually achieved a CR, so the trial was closed to accrual. Ultimately, 172 patients at 68 centers were enrolled of which 128 were eligible: 74 CIS ± Ta/T1 and 54 Ta/T1 without CIS. The CR rate at 6 months in CIS patients was 27% which fell below the null hypothesis of 30%. Of note, the CR rate at 3 months was 42% but this was an unplanned secondary endpoint. Toxicity data is yet unavailable but at least 9 patients in the CIS ± Ta/T1 cohort encountered grade 3-5 AEs during the course of the trial.