Predicting nonmuscle invasive bladder cancer recurrence and progression in a United States population

PURPOSE: We assessed the performance of the EORTC (European Organisation for Research and Treatment of Cancer) and CUETO (Club Urológico Español de Tratamiento Oncológico) nonmuscle invasive bladder cancer predictive models compared to current United States NCCN Guidelines® in an American population. MATERIALS AND METHODS: We retrospectively analyzed the electronic medical records of patients with nonmuscle invasive bladder cancer in a multicenter population in the United States. We evaluated recurrence-free and progression-free survival according to EORTC and CUETO, and assessed discriminative performance with the c-index at 1 and 5 years. We then compared the discrimination of EORTC and CUETO to the discrimination of the 4 nonmuscle invasive bladder cancer treatment groups described in NCCN Guidelines. RESULTS: We identified 1,333 patients with nonmuscle invasive bladder cancer and a median followup of 37 months. At 5 years the recurrence c-index of EORTC and CUETO was 0.59 and 0.56 while for progression it was higher at 0.74 and 0.72, respectively. NCCN Guidelines demonstrated a similar c-index of 0.56 and 0.75, respectively. The discrimination of all 3 risk models decreased in patients who received bacillus Calmette-Guérin. EORTC was better able to identify patients at low risk for recurrence or progression but it overestimated the 5-year risk of progression in patients at high risk. This study was limited by its retrospective design. CONCLUSIONS: Our work illustrates the need for improved predictive tools for clinicians who treat patients with nonmuscle invasive bladder cancer. However, until new tools are developed NCCN Guidelines are a simple option for clinicians who treat patients with nonmuscle invasive bladder cancer. Those guidelines provide predictive power comparable to that of the EORTC and CUETO models