Does Xylinas' nomogram accurately predict intravesical recurrence risk after radical nephroureterectomy for primary upper urinary tract urothelial carcinoma when applied to Asian populations?

OBJECTIVE To validate a prognostic nomogram (Xylinas' nomogram) for intravesical recurrence after radical nephroureterectomy for primary upper urinary tract urothelial carcinoma patients of Asian descent. METHODS Clinicopathological and survival data from 243 primary urinary tract urothelial carcinoma patients who underwent radical nephroureterectomy with bladder cuff excision between January 2004 and May 2017 were collated. Univariate and multivariable Cox regression analyses were performed to identify independent risk factors associated with intravesical recurrence-free survival. External validation was determined using regression coefficients abstracted from previously published data. Performance was then quantified through calibration and discrimination, according to concordance indexes (c-index) in receiver operating characteristic curves. RESULTS 163 patients met our eligibility criteria and were finally included in this study. At a median follow-up of 60 months, intravesical recurrence occurred in 29.4% (n = 48). Multivariable analysis revealed that being male, ureteral tumor location, tumor multifocality and previous bladder cancer were independent prognostic factors of intravesical recurrence-free survival. When Xylinas' nomogram was applied to our cohort, the discriminatory power was found to be roughly equivalent with a c-index of 68.3% for the reduced model and 68.4% for the full model. Calibration plots also revealed intravesical recurrence predictions at 3, 6, 12, 18, 24 and 36 months had relative concordance. Contrasting the respective performances of the reduced and full model suggests there is no significant difference between the two (all P > 0.05). CONCLUSIONS This nomogram appears accurate at predicting intravesical recurrence after radical nephroureterectomy for primary urinary tract urothelial carcinoma in Asian populations. However, it remains necessary to data mine for unknown prognostic factors for optimization. Further external validation is required across larger, ethically diverse populations before applying this nomogram in clinical practice.