Updated efficacy and safety data from the AGILE study in patients with newly diagnosed acute myeloid leukemia treated with ivosidenib + azacitidine compared to placebo + azacitidine.

7012 Background: Ivosidenib (IVO) is a potent oral targeted inhibitor of mutant IDH1. In the AGILE study in patients (pts) with newly diagnosed IDH1-mutated acute myeloid leukemia, IVO plus azacitidine (AZA) significantly improved event-free survival (EFS), overall survival (OS), complete remission (CR), and CR or CR with partial hematologic recovery (CR+CRh) rates v placebo (PBO) plus AZA. As of March 2021, median OS (mOS) was 24 months (IVO+AZA) v 7.9 months (PBO+AZA; HR: 0.44; p=0.0005). Long-term follow-up data are presented. Methods: In the double-blind AGILE study, pts were randomized 1:1 to IVO 500 mg QD + AZA 75 mg/m 2 SC or IV for 7 days in 28-day cycles, or PBO+AZA. Long-term follow-up data (June 2022) for OS, blood count recovery, transfusion independence and safety are described here. Results: 148 pts were randomized: 73 to IVO+AZA; 75 to PBO+AZA. Median treatment duration was 10.8 months (IVO+AZA) v 3.2 months (PBO+AZA). Five PBO+AZA pts crossed over to IVO+AZA after March 2021 (Table), and no adjustment was made for crossover in the updated OS analysis. At a median follow-up of 28.6 months, mOS was 29.3 months (95% CI 13.2, not reached) for IVO+AZA v 7.9 months (95% CI 4.1, 11.3) for PBO+AZA (HR 0.42 [0.27, 0.65]; p<0.0001). OS rates were 62.9% and 38.3% at 12 months and 53.1% and 17.4% at 24 months, with IVO+AZA and PBO+AZA, respectively. In the IVO+AZA arm, hemoglobin levels steadily increased from baseline (BL; 88.8 g/L) to cycle 8, and then stabilized; mean platelet count recovered from BL values (72.7 x 10 9 /L) as early as week 8 (171.9 x 10 9 /L) and remained stable; and mean neutrophil counts rapidly increased from BL (0.98 x 10 9 /L) to week 3 (3.99 x 10 9 /L) and week 4 (4.36 x 10 9 /L), and then stabilized to within the normal range. Conversion from BL transfusion dependence (red blood cell and/or platelet transfusion dependence) to post-BL transfusion independence was significantly higher with IVO+AZA than PBO+AZA (53.8% v 17.1%, respectively; p=0.0004). There were fewer neutropenic fever events (27.8% v 33.8%) and infections (34.7% v 51.4%) with IVO+AZA than with PBO+AZA. TEAEs led to discontinuation of IVO+AZA or PBO+AZA in 26.4% and 25.7% of pts, respectively. Conclusions: Updated data on OS (>5 months longer mOS and a greater risk reduction in deaths compared to the initial analysis), transfusion independence, blood count recovery and safety confirms the clinically and statistically robust benefit in favor of IVO+AZA at long-term follow up. Patient disposition. Clinical trial information: NCT03173248 . [Table: see text]