Clinical Lymphadenopathy in Urothelial Cancer: A Transatlantic Collaboration on Performance of Cross-sectional Imaging and Oncologic Outcomes in Patients Treated with Radical Cystectomy Without Neoadjuvant Chemotherapy

Background: Data regarding clinical node metastases (cN+) in patients undergoing radical cystectomy (RC) are scarce. Objective: To evaluate the performance of conventional imaging in detecting cN+ and analyze the impact of cN+ on survival among patients treated with RC without neoadjuvant chemotherapy (NAC). Design, setting, and participants: Data from three independent centers of consecutive patients with bladder cancer treated with RC without NAC were analyzed. Outcome measurements and statistical analysis: cN+ was defined as pelvic nodes >8. mm or abdominal nodes >10. mm in maximum short-axis diameter as detected via preoperative computed tomography or magnetic resonance imaging. Performance characteristics were evaluated considering pN+ disease as the reference standard. Multivariable Cox regression analyses were performed for prediction of survival. Results and limitations: Overall, 196 patients (7.1%) had cN+ disease before RC and pN+ status was confirmed for 122 of them (62.2%). cN+ status in the overall population had sensitivity of 18% and specificity of 96% with a calculated area under the curve of 57%. The median follow-up was 108 mo. On multivariable analyses, cN+pN+ (hazard ratio [HR] 1.84, 95% confidence interval [CI] 1.26-2.68) and cN-pN+ (HR 2.36, 95% CI 1.90-2.92) were predictors of CSM (both . p . <. 0.001). Conversely, cN+pN- status was not associated with worse survival outcomes (p > 0.2). Conclusions: Our study confirms the poor accuracy of conventional preoperative imaging in assessing nodal disease status. cN status had no independent impact on survival when all confounders were evaluated, and potentially curative treatments should not be withheld on the basis of clinical nodal status alone. Patient summary: The accuracy of conventional imaging techniques for detection of pathologic lymph node-positive disease before radical cystectomy for bladder cancer is suboptimal. The presence of clinical lymph node positivity on preoperative imaging is not an independent predictor of oncologic outcomes, and if the node invasion is not confirmed at radical cystectomy, these patients may have good long-term outcomes. Conventional imaging techniques are suboptimal in detecting preoperative node metastases. The presence of node metastases on preoperative imaging is not a predictor of poor oncologic outcomes if the node invasion is not confirmed pathologically after surgery.