Combination chemotherapy versus Bacillus Calmette-Guérin (BCG) for high-risk non-muscle invasive bladder cancer– a phase III multi-centre randomised controlled trial (COBRA)

Research question Are adjuvant intravesical gemcitabine and docetaxel (Gem-Doce) bladder instillations non-inferior to Bacillus Calmette-Guérin (BCG) in terms of high grade recurrence-free survival (hgRFS) for the treatment of high-risk non-muscle invasive bladder cancer (HR-NMIBC)? Background Approximately 4,800 UK patients are diagnosed with HR-NMIBC annually. For over 40 years, BCG treatment has remained the standard of care after initial surgery to remove the cancer. However, poor tolerability, moderate efficacy, and periodic supply shortages mean identifying an alternative to BCG is a priority for patients and clinicians alike. Aims and objectives We aim to demonstrate non-inferiority, and potentially superiority, of intravesical Gem-Doce compared to standard of care intravesical BCG for HR-NMIBC. The primary objective is to test if Gem-Doce is non-inferior to BCG in terms of hgRFS and if so, whether it is a cost-effective alternative. An exploratory objective will investigate the value of sequential treatment in patients who subsequently have a high grade recurrence. Methods COBRA is an open-label pragmatic phase III randomised controlled trial that will include 520 HR-NMIBC patients from 50 NHS hospitals. Eligible consenting patients, identified at bladder cancer multidisciplinary team meetings, will be randomised 1:1 to bladder instillations of either Gem-Doce induction (6 weekly) & maintenance (once-monthly from 3 to 24 months) or BCG induction (6 weekly) & maintenance (three once-weekly instillations at 3, 6, 12, 18 & 24 months). Participants will have standard of care 3-monthly surveillance cystoscopy and urine cytology for the first 24 months, then 6 monthly in year 3-4 and annually. Those who develop high grade recurrence will be offered the non-allocated treatment (i.e. if allocated BCG will receive Gem-Doce and vice versa). Under the non-inferiority design, developed with patient and public advisors and clinical experts, Gem-Doce will be declared non-inferior if the 2-year hgRFS rate is no more than 10% worse than the assumed 77.5% with BCG (critical hazard ratio 1.542, 185 events for 90% power, 1-sided 5% alpha). If non-inferiority is proven, superiority will be tested. A secondary pre-planned Bayesian analysis will incorporate historic controls to allow for a non-inferiority margin of 7.5% (77.1% power). Secondary endpoints include other cancer outcomes, compliance and treatment tolerability, side effects, patient-reported health related quality of life and health economic evaluation. Timelines for delivery We aim to recruit over 4 years, follow all participants for at least 2 years and report outcomes within 6.5 years of the first participant’s first visit. Anticipated impact and dissemination COBRA will deliver a definitive statement on the effectiveness of Gem-Doce in HR-NMIBC to inform decisions on its adoption in the UK. Widespread dissemination of results, including amongst the patient communities who have long called for alternatives to BCG, will influence clinical practice. If non-inferiority, cost-effectiveness and fewer side effects relative to BCG are confirmed clinical guidelines will be updated and swift UK-wide implementation will be possible due to the use of generic agents and standard equipment for treatment delivery. If Gem-Doce is superior to BCG we envision consequent reductions in cystectomy rates, international impact and rapid practice change worldwide.