Development and external validation of nomograms predicting disease-free and cancer specific survival after radical cystetcomy

==inizio abstract==

Introduction: Available nomograms predicting survival probabilities after radical cystectomy (RC) and pelvic lymph node dissection (PLND) are based on a lot of variables. External validation of nomograms is an essential step to confirm their discrimination accuracy among different Institutions. We developed two nomograms predicting disease-free (DFS) and cancer-specific survival (CSS) and we externally validated them in multiple series.

Material & Methods: Prospectively collected data of a single-centre series of 818 consecutive patients who underwent RC and PLND for muscle-invasive urothelial carcinoma were used for nomogram building. External validation was performed in 3173 patients from 7 centres worldwide. All patients underwent RC and PLND. Time to disease recurrence and to cancer-death were addressed with univariable and multivariable Cox analyses. Nomograms were built to predict 2-yr, 5-yr and 8-yr DFS and CSS probabilities. Predictive accuracy was quantified using the concordance index.

Results: Age, pT stage, Lymph-node density and extent of PLND were independent predictors of DFS and CSS (p<0.05). Two nomograms were built to predict DFS and CSS probability, respectively. (Figure 1 and Figure 2) Discrimination accuracies for DFS and CSS at 2-yr, 5-yr and 8-yr were 0.81, 0.8, 0.79 and 0.82, 0.81, 0.8, respectively, with a slight overestimation at calibration plots beyond 24 months. In the external series predictive accuracies for DFS and CSS at 2-yr, 5-yr and 8-yr were 0.83, 0.82, 0.82 and 0.85, 0.85, 0.83 for European centres; 0.73, 0.72, 0.71 and 0.80, 0.74, 0.68 for African series; 0.76, 0.74, 0.71 and 0.79, 0.76, 0.73 for American series. Limitations included the inapplicability of these nomograms in patients receiving a limited PLND or a neoadjuvant chemotherapy. Conclusions: These nomograms developed on a contemporary series are easy prediction tools and provided optimal DFS and CSS discrimination in all external cohorts. ==fine abstract==