Background: Nearly half of the patients (pts) diagnosed with non-metastatic muscle invasive bladder cancer are unfit for cisplatin therapy and no alternative options of neoadjuvant treatment exist. Avelumab (A), a monoclonal antibody directed against PD-L1, showed efficacy in advanced urothelial cancer. We report preliminary data assessing preoperative avelumab associated with paclitaxel and gemcitabine or avelumab alone in the cisplatin ineligible cohort. The results of the cisplatin eligible cohort were reported at ESMO 2021. Methods: AURA is a prospective, multicenter, randomized, phase II trial for pts with cT2-4aN0-2M0 bladder carcinoma. Pts were enrolled in two separated cohorts based on eligibility for cisplatin chemotherapy. Cisplatin ineligible pts received 4 cycles of paclitaxel-gemcitabine (PG) plus A every 2 weeks or 4 cycles of A every 2 weeks (1:1). Primary endpoint was pCR rate (ypT0/isN0) with the objective, in each arm, to show pCR rate > 5% (90% power reached in case of pCR rate > 25%). Two-step design was used with planned interim analysis after 12 evaluable pts per arm. Secondary endpoints were pathologic downstaging rate ( < ypT2N0) and safety. Results: A total of 56 cisplatin-ineligible pts were evaluable. For PG + A arm (n = 28): median age was 72 years (41-80), 93% male. For A alone arm (n = 28): median age was 75 years (49-89), 93% male. One patient did not undergo surgery due to progression in the A arm but was included in intention to treat analysis. Five pts did not receive the 4 cycles of treatment; reasons included toxicity (n = 3) for PG + A arm and patient/physician decision (n = 2) for A arm. pCR was achieved in 5 pts (18%; 95%CI 6%-37%) treated with PG + A and 10 pts (36%; 95%CI 19%-56%) treated with A. Downstaging to < ypT2N0 was achieved in 6 pts (21%; 95%CI 8%-41%) and 11 pts (39%; 95%CI 22%-59%), respectively. Median time from treatment initiation to surgery was 82 days (55-144) for PG + A arm and 67 days (38-89) for A arm. Most common irAEs of any grade were asthenia (15%), skin toxicity and myalgia/arthralgia (each 5%). There were 2 pts in the PG + A arm with grade 3 irAEs (hepatitis and pneumonitis) that caused A discontinuation for 1 patient. No treatment-related deaths were reported. Conclusions: Neoadjuvant single agent avelumab resulted in high pCR and was safely administered without compromising surgical resection. Our results suggest that the addition of taxane-gemcitabine regimen to avelumab may reduce avelumab efficacy with low pCR rate. Survival analysis and correlative studies are underway. Clinical trial information: NCT03674424.