Introduction: The appropriate extent of lymph node dissection (LND) at radical cystectomy for bladder cancer (BCa) is still controversial. In particular, patients with clinical evidence of locoregional lymph node metastasis (cN+) may benefit from an extended LND. In this study, we investigated the effect of extended versus standard LND on survival outcomes in patients with cN+ BCa. Methods: We queried our retrospective, multi-institutional database to identify all patients who underwent upfront RC with standard (internal, external iliac, and obturator lymph nodes) or extended LND (additionally common iliac and presacral lymph nodes) for cT2-4N1-3M0 BCa. We excluded censored patients with a follow-up <3 months and all patients who received preoperative chemotherapy. The primary objective was to assess recurrence-free survival (RFS) after RC and the distribution of recurrences (local vs. distant). Secondary endpoints comprised overall (OM) and cancer-specific mortality (CSM). The Kaplan-Meier method and cumulative incidence curves accounting for competing risks were used to delineate survival outcomes graphically. We assessed the impact of the LND template on RFS, OM, and CSM on uni- and multivariable regression analyses. This study was IRB-approved (ref. no. 1480/2022). Results: Of 370 patients, 232 underwent standard and 138 extended LND. In total, 182 experienced cancer recurrence, 190 died with 140 dying from BCa. Median follow-up of patients alive was 32 months, and median time to recurrence was 8 months. Patients who underwent an extended LND were significantly younger (mean 73 vs. 68 years, p=0.001), had more frequent cN3-stage (1.7% vs. 9.4%), and more positive lymph nodes removed (median 1.0 vs. 2.0, p=0.02). 114 patients received adjuvant treatment with no difference between LND groups. There was no difference in the distribution of recurrences (in total, 32 locoregional, 103 distant, and 18 both) between the groups. On survival analyses, we did not find a difference in RFS (p=0.17), OM (p=0.91), or CSM (p=0.07) when stratified by the LND template. On multivariable analyses, adjusted for age, gender, comorbidities, pathological T- and N-stage, type of urinary diversion, surgical margins, and receipt of adjuvant treatment, the LND template was not independently associated with survival outcomes. Conclusions: In patients with cN+ BCa, extended LND did not results in an improved RFS, OM, or CSM compared to standard LND. The results of the ongoing prospective SWOG 1101 trial are urgently awaited. SOURCE OF Funding: None
