Assessing Cancer Progression and Stable Disease After Neoadjuvant Chemotherapy for Organ-confined Muscle-invasive Bladder Cancer

Objective To propose and validate a new approach to stratify clinically staged, organ-confined, muscle-invasive bladder cancer patients (cT2N0M0) who are pathologic non-responders to neoadjuvant chemotherapy (NAC) to better characterize NAC non-response. Methods We retrospectively identified radical cystectomy patients with cT2N0M0 disease at our institution (2005-2014) and in the National Cancer Database (2004-2012) for external validation. Patients were stratified as stable (pT2N0M0) or progressors (>pT2 or pN+). The primary end points were cancer-specific survival (CSS), overall survival (OS), and recurrence-free survival (RFS). Results In the institutional cohort, NAC stable patients (n = 17) had better OS (P =.05) and RFS (P =.04) than NAC progressors (n = 50), and had comparable OS (P =.7) and CSS (P =.09) with non-NAC stable patients (n = 27). Multivariable Cox proportional hazard models showed that larger tumor size predicted worse OS (hazard ratio [HR] = 1.20 per centimeter, 95% confidence interval [CI: 1.07, 1.35]), CSS (HR = 1.27, 95% CI [1.11, 1.45]), and RFS (HR = 1.24, 95% CI [1.09, 1.42]). Similarly, in the National Cancer Database, NAC stable patients (n = 223) had improved OS (P <.0001) compared with NAC progressors (n = 232) and comparable (P =.4) OS with non-NAC stable patients (n = 950). Multivariable Cox proportional hazard model showed that larger tumor size (HR = 1.03 per centimeter, 95% CI [1.02, 1.03]) and progression (HR = 2.69, 95% CI [2.40, 3.01]) predicted worse OS. Conclusion Distinct survival outcomes suggest that NAC non-responders should be further stratified into stable disease and progressors. Comparable survival between non-NAC and NAC stable disease patients suggests that NAC stable disease may represent a chemoresistant but more indolent phenotype on the disease spectrum. Moreover, tumor size is an important prognostic biomarker in NAC non-responders. Clinical predictors of disease progression on NAC were not identified, highlighting the need to explore molecular and genomic subtyping determinants of disease progression.