Background: SAKK 06/17 investigated the addition of perioperative immunotherapy with the anti-PD-L1 antibody durvalumab (Durva) in the multimodality treatment of resectable MIUC. While most similar trials had a primary endpoint of pathological complete remission rate, this study evaluated the clinically more relevant primary endpoint of event-free survival (EFS) at 2 years (yrs). Methods: SAKK 06/17 was an open-label, single-arm phase II study including 61 cisplatin-fit patients (pts) with stage cT2-T4a cN0-1 operable MIUC. Pts received four cycles of neoadjuvant Cis/Gem q3w in combination with 4 cycles Durva 1500mg q3w followed by complete resection. Adjuvant Durva 1500mg q4w was given for 10 cycles or a maximum of 40 weeks. The primary endpoint was EFS at 2 yrs after neoadjuvant trial treatment (NAT) start. An event was defined as progression during NAT, appearance of metastases, locoregional recurrence after surgery or death from any cause. 58 pts were needed based on one-sided type I error 10% and power 80% for H₁ EFS at 2 yrs ≥ 65% compared to H₀ EFS at 2 yrs ≤ 50%. Secondary endpoints included pathological response, recurrence free survival after R0 resection (RFS), overall survival (OS) and safety. We report the primary analysis of the full analysis set (FAS, received at least one dose of Durva). Results: 61 pts were included between July 2018 and September 2019 at 12 sites. Median follow up is 28.1 months (95%CI 27.8-28.4). FAS consisted of 58 pts (79% male, median age 68 yrs) with bladder cancer (95%) or upper urinary tract/urethral cancer (5%). Clinical T2, T3, T4 stage was present at diagnosis in 69%, 21%, 10%, respectively, and 17% had cN1. Resection was performed in 53 pts (91%; 4 refused, 1 unresectable) with R0 resection in 52 pts (98%). 48 (91%) of resected pts started adjuvant Durva and 32 (67%) completed it. Pathological response < ypT2 ypN0 was achieved in 32 pts (18 pts ypT0 and 14 pts ypT1/a/is), corresponding to 60% of resected pts and 55% of the FAS. EFS at 2 yrs was overall 76.1% (one-sided 90% CI (lower bound): 67.6%; 95% CI 62.3% - 85.3%), for ypT1/a/is 92.9% and for ypT0 100%. RFS at 2 yrs after R0 resection (N=52) was 83.5% (95% CI 69.6% - 91.4%) and OS at 2 yrs for the FAS population was 87.3% (95% CI 73.8% - 94.1%). Grade 1, 2, 3, 4 adverse events attributed to Durva during overall treatment were 14%, 35%, 19%, 7%, respectively. Conclusions: The addition of perioperative Durvalumab to the standard of care for pts with resectable MIUC results in a high EFS, RFS and OS at 2 yrs, especially for pts with downstaging to <ypT2. The null hypothesis for the primary endpoint was clearly rejected. More in-depth analyses for biomarkers (PD-L1, ctDNA) will be presented at the meeting. Clinical trial information: NCT03406650.