Limited versus extended pelvic lymphadenectomy in patients with bladder cancer undergoing radical cystectomy: Survival results from a prospective, randomized trial (LEA AUO AB 25/02).

Abstract

4503 Background: The extent of pelvic lymph node dissection (PLND) in bladder cancer patients (pts.) undergoing radical cystectomy may affect survival according to retrospective studies. The German Urologic Oncology Group (AUO) reports mature data of the first prospective, randomized clinical trial to evaluate the impact of a limited versus an extended PLND. Methods: Pts. with high-grade T1 or invasive urothelial bladder cancer (cT2-T4a) from 16 German centers were randomized 1:1 to receive a limited versus an extended PLND at the time of radical cystectomy. Limited PLND included 6 fields (bilateral obturator, internal and external iliac nodes) and extended PLND defined 14 fields (in addition bilateral deep obturator fossa, presacral, paracaval, interaortocaval and paraaortal nodes up to the inferior mesenteric artery). Pts. with neoadjuvant chemo- or radiotherapy were excluded, adjuvant chemotherapy was allowed. The primary endpoint was recurrence-free survival (RFS). Cancer-specific survival (CSS) was a secondary endpoint. The planned total sample size was 400 to detect an improvement in 5-year RFS from 50% to 65% (90% power, 2-sided test, p < 0.05) in the extended arm. Results: In total, 375 of 437 pts. were randomized from 02/2006 to 08/2010 and eligible for intention-to-treat analysis (191 limited and 184 extended PLND). The tumor was locally confined (≤ pT2 pN0) in 49.6% of patients and 24.0% were node positive (pN+). The median number of dissected nodes was 19 in the limited and 32 in the extended arm. The 5-year RFS rate was 62.0% in the limited compared to 69.3% in the extended arm which was statistically not significant (Hazard ratio (HR) = 0.80, 95% Confidence Interval (CI) (0.54-1.19); log-rank p = 0.28). The 5-year CSS rate was improved from 66.2% in the limited to 77.5% in the extended arm which was statistically not significant (HR = 0.70, 95%CI 0.45-1.10; log-rank p = 0.13). Conclusions: We observed a trend but no significant difference toward improved RFS and CSS with an extended PLND. The rate of recurrence was lower than expected in the limited arm, which might be due to the high number of resected nodes in this arm. Clinical trial information: NCT01215071.