Development and Validation of a Prognostic Nomogram Model for Recurrence of Non-Muscle Invasive Bladder Cancer

Abstract BackgroundDifferent recurrence probability of non-muscle invasive bladder cancer (NMIBC) requests different adjuvant treatments and follow-up strategies. However, there is no simple, intuitive, and generally accepted clinical recurrence predictive model available for NMIBC. This study aims to construct a predictive model for the recurrence of NMIBC based on demographics and clinicopathologic characteristics from two independent centers. MethodsDemographics and clinicopathologic characteristics of 511 patients with NMIBC were retrospectively collected. Recurrence free survival (RFS) was estimated using the Kaplan-Meier method and log-rank tests. Univariate Cox proportional hazards regression analysis was used to screen variables associated with RFS, and a multivariate Cox proportional hazards regression model with a stepwise procedure was used to identify those factors of significance. A final nomogram model was built using the multivariable Cox method. The performance of the nomogram model was evaluated with respect to its calibration, discrimination, and clinical usefulness. Internal validation was assessed with bootstrap resampling. X-tile software was used for risk stratification calculated by the nomogram model. ResultsIndependent prognostic factors including tumor stage, recurrence status, and European Association of Urology (EAU) risk stratification group were introduced to the nomogram model. The model showed acceptable calibration and discrimination (area under the receiver operating characteristic [ROC] curve was 0.85; the consistency index [C-index] was 0.79 [95% CI: 0.76 to 0.82]), which was superior to the EAU risk stratification group alone. The decision curve also proved well clinical usefulness. Moreover, all populations could be stratified into three distinct risk groups by the nomogram model. ConclusionsWe established and validated a novel nomogram model that can provide individual prediction of RFS for patients with NMIBC. This intuitively prognostic nomogram model may help clinicians in postoperative treatment and follow-up decision-making.