A randomized phase 2 trial of azacitidine with or without durvalumab as first-line therapy for higher-risk myelodysplastic syndromes

Azacitidine-mediated hypomethylation promotes tumor cell immune recognition but may increase the expression of inhibitory immune checkpoint molecules. We conducted the first randomized phase 2 study of azacitidine plus the immune checkpoint inhibitor durvalumab vs azacitidine monotherapy as first-line treatment for higher-risk myelodysplastic syndromes (HR-MDS). In all, 84 patients received 75 mg/m² subcutaneous azacitidine (days 1-7 every 4 weeks) combined with 1500 mg intravenous durvalumab on day 1 every 4 weeks (Arm A) for at least 6 cycles or 75 mg/m² subcutaneous azacitidine alone (days 1-7 every 4 weeks) for at least 6 cycles (Arm B). After a median follow-up of 15.25 months, 8 patients in Arm A and 6 in Arm B remained on treatment. Patients in Arm A received a median of 7.9 treatment cycles and those in Arm B received a median of 7.0 treatment cycles with 73.7% and 65.9%, respectively, completing $4 cycles. The overall response rate (primary end point) was 61.9% in Arm A (26 of 42) and 47.6% in Arm B (20 of 42; P 5 .18), and median overall survival was 11.6 months (95% confidence interval, 9.5 months to not evaluable) vs 16.7 months (95% confidence interval, 9.8-23.5 months; P 5 .74). Durvalumab-related adverse events (AEs) were reported by 71.1% of patients; azacitidine-related AEs were reported by 82% (Arm A) and 81% (Arm B). Grade 3 or 4 hematologic AEs were reported in 89.5% (Arm A) vs 68.3% (Arm B) of patients. Patients with TP53 mutations tended to have a worse response than patients without these mutations. Azacitidine increased programmed cell death ligand 1 (PD-L1 [CD274]) surface expression on bone marrow granulocytes and monocytes, but not blasts, in both arms. In summary, combining azacitidine with durvalumab in patients with HR-MDS was feasible but with more toxicities and without significant improvement in clinical outcomes over azacitidine alone.