Prediction of the Need for an Extended Lymphadenectomy at the Time of Radical Cystectomy in Patients with Bladder Cancer

Abstract

BackgroundA prospective randomized trial (LEA AUO AB 25/02) found no survival benefit in extended compared with limited pelvic lymph node dissection (PLND) templates in bladder cancer (BCa) patients treated with radical cystectomy (RC). However, the rate of lymph node invasion (LNI) in the standard and extended templates was lower than estimated. ObjectiveTo assess the accuracy of preoperative clinical and pathological parameters to predict LNI and to develop a model to preoperatively select candidates for the extended PLND templates. Design, setting, and participantsA total of 903 BCa patients treated at a single institution were retrospectively identified. The primary outcome was to identify preoperatively the risk of LNI to tailor the type of PLND. The extended PLND templates consisted in the removal of pelvic lymph nodes together with the common iliac, presacral, para-aortocaval, interaortocaval, and paracaval sites up to the inferior mesenteric artery. InterventionA total of 903 BCa patients were treated with RC and bilateral extended PLND templates. Outcome measurements and statistical analysisSeveral models predicting LNI were evaluated using the area under the receiver operating characteristic curve (AUC), calibration plots and decision curve analyses. A nomogram predicting LNI in the extended pattern was developed and validated internally. Results and limitationsOverall, 55 patients (6.1%) had LNI in the extended PLND templates at RC. The median number of nodes removed was 19 (interquartile range: 13–26). A model including age, clinical T stage, clinical node stage, lymphovascular invasion, and presence of carcinoma in situ at the last transurethral resection before RC was developed. The AUC of this model is 73%. Using a cutoff of 3%, 108 extended PLNDs (12%) would be spared and only two LNIs (3%) would be missed. The main limitations of our model are the retrospective nature of the data, lack of external validation, and low rate of LNI. ConclusionsThis is the first proposed model to predict LNI in the extended PLND templates. This model might help urologists identify which patients might benefit from an extended PLND at the time of RC, reserving a standard PLND for all the others. Patient summaryWe developed the first nomogram to predict lymph node invasion (LNI) in the extended pelvic lymph node dissection templates in bladder cancer patients treated with radical cystectomy. The adoption of our model to identify candidates for the extended pelvic lymph node dissection templates could avoid up to 12% of these procedures at the cost of missing only 3% of patients with LNI.