The category “BCG-unresponsive disease”, formulated by experts at the request of the United States Food and Drug Administration, denotes a group of patients with recurrent non-muscle-invasive bladder cancer for whom continued BCG treatment is unlikely to provide benefit. Although quickly adopted for trial design, many of the nuances within the definition lack validation. In this study, we evaluated the prognostic value of BCG unresponsive designation (i.e. recurrence after induction plus at least 1 maintenance course of BCG) by comparing the oncologic outcomes of these patients with those recurring after induction BCG alone. We confirm that appropriately defined, BCG-unresponsive patients are more likely to require salvage radical cystectomy (54.5% vs 17.9%, p = 0.002). Moreover, those opting for second-line bladder-sparing therapies are less likely to remain free of tumor recurrence (23% vs 69.2%, p = 0.003). On multivariate analysis, BCG-unresponsive disease independently predicts inferior high-grade recurrence-free survival (hazard ratio [HR]: 6.25, 95% confidence interval [CI]: 2.27–16.67; p < 0.001) and cystectomy-free survival (HR: 3.85, 95% CI: 1.49–10.0; p = 0.006). Our data confirm the prognostic implication of the BCG unresponsive definition i.e. recurrence of high grade disease after induction and one course of maintenance BCG, and support its use in counseling and risk stratification of patients with tumor recurrence after BCG.
Patients who have BCG-unresponsive disease, that is, high-grade non-muscle-invasive bladder cancer recurring after BCG induction and maintenance, have a low likelihood to respond to further BCG treatment and should consider radical cystectomy or clinical trial enrollment.