Prediction of 90-day mortality after radical cystectomy for bladder cancer in a prospective European multicenter cohort

Background: Despite recent improvements, radical cystectomy (RC) is still associated with adverse rates for 90-d mortality.

Objective: To validate the performance of the Isbarn nomogram incorporating age and postoperative tumor characteristics for predicting 90-d RC mortality in a multicenter series and to generate a new nomogram based strictly on preoperative parameters.

Design, setting, and participants: Data of 679 bladder cancer (BCa) patients treated with RC at 18 institutions in 2011 were prospectively collected, from which 597 patients were eligible for final analysis.

Intervention: RC for BCa.

Outcome measurements and statistical analysis: An established prediction tool, the Isbarn nomogram, was applied to our cohort. For the purpose of external validation, model discrimination was measured using the receiver operating characteristics-derived area under the curve. Calibration plots examined the relationship between predicted and observed probabilities. Univariable and multivariable logistic regression models were fitted to assess the impact of preoperative characteristics on 90-d mortality.

Results and limitations: The 30-, 60-, and 90-d mortality rates in the development cohort (n=597) were 2.7%, 6.7%, and 9.0%, respectively. The Isbarn nomogram predicted individual 90-d mortality with an accuracy of 68.6%. Our preoperative multivariable model identified age (odds ratio [OR]:1.052), American Society of Anesthesiologists score (OR: 2.274), hospital volume (OR: 0.982), clinically lymphatic metastases (OR: 4.111), and clinically distant metastases (OR: 7.788) (all p<0.05) as independent predictors of 90-d mortality (predictive accuracy: 78.8%). Our conclusions are limited by the lack of an external validation of the preoperative model.

Conclusions: The Isbarn nomogram was validated with moderate discrimination. Our newly developed model consisting of preoperative characteristics might outperform existing models. Our model might be particularly suitable for preoperative patient counseling.

Patient summary: The current report validated an established nomogram predicting 90-d mortality in patients with bladder cancer after radical cystectomy (RC). We developed a new prediction tool consisting of strictly preoperative parameters, thus allowing clinicians an optimal consultation for RC candidates.