A phase I, open-label, multicenter study to assess the safety, pharmacokinetics, and preliminary antitumor activity of AZD4635 both as monotherapy and in combination in patients with advanced solid malignancies: Results from prostate cancer patients (NCT02740985).

5518 Background: AZD4635 inhibits adenosine 2a receptor (A2aR) signaling and improves immune activation and anti-tumor activity in preclinical models. This phase I study assessed the safety, pharmacokinetics, pharmacodynamics and efficacy of AZD4635 monotherapy (mono) and in combination (combo) with durvalumab (durva) in patients (pts) with refractory solid tumors. Here we present data for immune checkpoint-naïve pts with metastatic castrate-resistant prostate cancer (mCRPC). Methods: Pts with refractory mCRPC received AZD4635 mono (75 mg or 100 mg QD oral nanosuspension [(RP2D]) or in combination (75 mg or 100mg QD) with durva 1.5g IV q4wk. Results: Between 30Aug16 and 20Jun19 (data cutoff [DCO]) 94 mCRPC pts were treated with mono (n = 49) or combo (n = 45): median age 70.5 yrs; ECOG 0-1 = 99%. The median number of prior treatment regimens was 5 (range = 1-10); 61% of pts (57/94) had prior chemotherapy, 90% had prior new hormonal therapy. PK data suggest AZD4635 concentrations at 75-100 mg QD are above the in vitro IC50 for A2aR inhibition throughout the dosing interval. Modeling predicts 80-90% A2aR occupancy at steady state for doses at ≥75 mg QD. Most common treatment-related AEs ( > 10%) were nausea, vomiting, fatigue, decreased appetite, dizziness, and diarrhea. At the DCO in this ongoing study, 70 pts were evaluable for response by RECIST v1.1 (mono = 33, combo = 37). Confirmed response occurred in 8 pts: mono = ORR 6.1% (2PRs) and combo = 16.2% (2CRs, 4PRs). The duration of response across cohorts ranged from 1-18.5 mo (5 pts ongoing). PSA response (defined as ≥50% decrease from baseline of ≥1ng/ml) was observed in 6.4% (3/47 pts; 95% CI, 1.3-17.5) of mono-treated patients and 20% (9/45 pts; 95% CI, 9.6-34.6) of combo-treated patients. Patients with high adenosine (ADO) gene expression signature (N = 46) in peripheral blood, showed a median PFS of 21 weeks v. 8.7 weeks in ADO signature low patients (N = 46) (HR 0.5, CI 0.3-0.9) (DCO 20Jun19). In addition, baseline TCR clonality and diversity were linked with response. Conclusions: In mCRPC pts, AZD4635 alone or in combination with durva was tolerable and associated with clinical benefit. A mCRPC phase II trial is ongoing with continued exploration of predictors of response to treatment. Clinical trial information: NCT02740985 .