Predicting local failure after radical cystectomy in patients with bladder cancer: Implications for the selection of candidates at adjuvant radiation therapy

Objective To evaluate incidence and predictors of local failure (LF) after radical cystectomy (RC) due to bladder cancer. Methods We focused on 1,112 patients treated with RC, between 1990 and 2012, at a single center. LF was defined as imaging evidence of recurrence in the pelvic soft tissues or nodes below the aortic bifurcation at least 3 months before the detection of distant metastases. Competing risk analyses tested the relationship between clinical and pathological factors and the risk to develop LF. Regression tree analysis stratified patients into risk-groups based on their characteristics and the corresponding LF rate. Results Overall, 50 (4.5%) patients developed LF during a median follow-up period of 62 (35–92) months. On univariable competing risk regression analyses, pathological T stage (pT4 vs. pT3; hazard ratio [HR] = 2.55, P = 0.003), soft tissue surgical margin (STSM; HR = 2.95, P = 0.005), and variant histology (HR = 1.79, P = 0.03) were associated with an increased risk of developing LF. The cohort was stratified into 4 risk groups: very low (≤pT3a disease and pure urothelial histology), low (≤pT3a disease and variant histology), intermediate (pT4 disease), and high (positive STSM). Conclusions LF is an important event in RC patients. We developed a new risk model based on bladder cancer characteristics. Our findings could help with the identification of the best candidate for consideration of adjuvant radiotherapy.