A novel nomogram for predicting extraurothelial recurrence in patients with upper urinary tract urothelial carcinoma after radical nephroureterectomy.

Abstract

We aimed to establish and validate a nomogram for extraurothelial recurrence (EUR) after radical nephroureterectomy (RNU) for upper urinary tract urothelial carcinoma (UTUC). The data of 521 patients with UTUC after RNU from 2 medical centers were retrospectively studied and were used as training cohort (n = 301) and external validation cohort (n = 220). We used the least absolute shrinkage and selection operator (LASSO) to select variables for multivariable Cox regression, and included independent risk factors into nomogram models predicting EUR-free survival (EURFS). Multiple parameters were used to validate the nomogram, including the concordance index (C-index), the calibration plots, the time-dependent receiver-operator characteristics curve (ROC), and the decision curve analysis (DCA). Patients were stratified into three risk groups according to total points calculated by nomograms. The differences of EURFS in each group were analyzed by the Kaplan-Meier analysis. Four variables were screened through LASSO regression. Bladder cancer history, Ki-67, lymphovascular invasion (LVI), and pathological T stage were shown to be independent predictive factors for EUR. The C-indexes of the model were 0.793 and 0.793 in training and validation cohorts, respectively. In comparison with prediction based on categorized pathological T stage, the DCA curves for 5-year EUR exhibited better performance. The 5-year EURFS rates were 92.2%, 63.8%, and 36.2% in patients stratified to the low-, medium-, and high-risk group. Our study provided a new nomogram to predict the probability of EUR in UTUC patients underwent RNU, with perfect performance in discrimination ability and clinical net benefit. The application of the model may help urologists to choose proper treatment and monitoring.