SWOG S1011: A phase III surgical trial to evaluate the benefit of a standard versus an extended lymphadenectomy performed at time of radical cystectomy for muscle invasive urothelial cancer.

4508 Background: S1011 tested the hypothesis that an extended lymphadenectomy (ELND) is associated with improved disease-free and overall survival (DFS, OS) compared to standard (S) LND in patients with localized muscle invasive bladder cancer (MIBC) undergoing radical cystectomy (RC)(NCT01224665). Methods: Eligible patients with cT2-4a N0-2 were stratified by receipt and type of neoadjuvant chemotherapy (NAC), T2 vs T3-4a and PS 0-1 vs 2. Patients were randomized 1:1 after intraoperative exploration determined they did not have disease outside the pelvis. All patients then underwent a standard bilateral pelvic LND including external and internal iliac and obturator LNs. If randomized to the experimental arm additional ELND up to at least the aortic bifurcation including common iliac (CI), pre-sciatic, and pre-sacral nodes was performed. We hypothesized that patients in the ELND arm would have a 10% improvement in 3-year DFS compared to an estimated 55% for patients in the SLND arm (HR = 0.72). Assuming a 1-sided a=0.025 and 85% power, 564 eligible randomized patients were required. Final analysis was to occur at 184 DFS events in the SLND arm or after max follow-up (6 yrs) using a stratified logrank(LR) test with a=0.022 to account for interim testing. Hazard ratios from Cox model are adjusted for strat factors. Secondary endpoints included OS and safety. Results: 36 surgeons at 27 sites in US and Canada were credentialed prior to enrolling patients, 658 were registered from 8/11-2/17, and 618 eligible patients were randomized to ELND (n=292) or SLND (n=300). Median f/up was 6.1 years in both arms. Median age was 69, 21% female, and 9% non-White. Clinical stage was balanced in both arms: T2 (71%) and T3-4a (29%). NAC was given to 57% in both. Pathologic T stage was <T2 in 39% in S and 37% in E and ³ T2 in 61% and 63%, respectively. Median lymph nodes removed was higher in ELND compared to SLND (41 vs 25), but there was no difference in node metastasis with 26% vs 24%, respectively. More ELND patients had N2 or N3 disease. ELND was associated with increased G3-4 AEs compared to SLND: 16% vs 8%. Deaths within 90 days of RC occurred in 26 (4.4%) patients, including 16 in ELND vs 9 in SLND. There was no difference in DFS between E vs S arms (HR 1.10; 95% CI 0.87, 1.42), 1-sided LR p=0.82. OS results were similar (HR 1.15 95% CI 0.89, 1.48), 1-sided LR p=0.87. Conclusions: Patients with MIBC undergoing RC and ELND had increased node yield and higher pathologic N stage, but no significant DFS or OS benefit compared to patients undergoing SLND. ELND was also associated with greater morbidity and higher peri-operative mortality. Clinical trial information: NCT01224665 . Table: see text