Conditional analyses of recurrence and progression in patients with TaG1 non-muscle-invasive bladder cancer.

Overview

abstract

OBJECTIVE: To determine conditional recurrence-free survival (RFS) and progression-free survival (PFS) and improve decision-making toward surveillance protocols and scheduling. Furthermore, evaluating the evolution of predictors for disease recurrence over time, because TaG1 non-muscle-invasive bladder cancer harbors a risk of disease recurrence and progression. MATERIAL AND METHODS: The retrospective multicenter design study includes 1,245 TaG1 bladder cancer patients with median follow-up of 62.7 (interquartile range: 34.3-91.1) months. Conditional RFS and PFS estimates were calculated using the Kaplan-Meier method. Multivariable Cox regression model was calculated proportional for the prediction of recurrence and progression (covariables: age, tumor size, multiple tumors, prior recurrence, and immediate postoperative instillation of chemotherapy). RESULTS: After 3 months without event, the conditional RFS and PFS (to ≥pT2) rates for 5 additional years without event were 57.5% and 93.4%, respectively. Given a 1-, 2-, 3-, and 5-year survival, the conditional RFS rates for 5 additional years without event improved by +9.8 (67.3%), +5.2 (72.5%), +6.5 (79.0%), +2.0 (81.0%), and +1.0% (82.0%), respectively. In contrast, the 5-year conditional PFS rates were more or less stable with 94.3% after 1 year to 94.1% after 5 years. Multivariable analyses showed decreasing impact of risk parameters on RFS estimates over time. Based on these findings, we suggest a risk stratification to individualize follow-up for intermediate risk TaG1. Main limitation was the retrospective design. CONCLUSIONS: Conditional-survival analyses demonstrates that the patient risk profile changes over time. RFS rates rise with increasing survival whereas PFS rates were stable. The impact of prognostic features decreases over time. Our findings can be used for patient counseling and planning of personalized follow-up.