Phase 1/1b Dose Escalation and Expansion of CPX-351 in Combination with Gemtuzumab Ozogamicin in Newly Diagnosed Acute Myeloid Leukemia

Introduction: Initial response to intensive chemotherapy in acute myeloid leukemia (AML) is about 50-60% and relapses are common. CPX-351, a dual-drug liposomal formulation of daunorubicin and cytarabine at the synergistic 1:5 molar ratio, is approved by the Food and Drug Administration (FDA) for newly diagnosed (ND) therapy-related or myelodysplasia-related AML (Lancet et al. JCO. 2018). Gemtuzumab ozogamicin (GO) is a CD33-directed monoclonal antibody covalently linked to the cytotoxic agent calicheamicin approved by the FDA for CD33-positive AML and in combination with standard 7+3 (Castaigne et al. Lancet. 2012). Measurable residual disease (MRD) and the persistence of leukemia stem cells during the post-remission phase are thought to be drivers of relapse. In this study, we hypothesized the combination of CPX-351 and GO would yield higher and deeper responses in ND AML patients. Methods: This is a phase 1/1b clinical trial (NCT05558124) using CPX-351 (daunorubicin 44 mg/m2 and cytarabine 100 mg/m2) (Day 1, 3, 5) combined with different dose levels of GO (Cohort A: 3mg/m2 on Day 1; Cohort B: 3mg/m2 on Day 1, 4; Cohort C: 3mg/m2 on Day 1, 4, 7) for ND CD33-positive AML patients between age 18-70. The primary objective is to determine the safety and maximal tolerated dose/recommended phase 2 dose for this combination. Using Bayesian optimal interval design to guide dose escalation/de-escalation, 18 patients are planned for the study. Reinduction of CPX-351 (Day 1, 3) is allowed if indicated by Day 14-24 bone marrow biopsy. Consolidation, up to 4 cycles, consists of cytarabine (3g/m2 for patients < 60 and 1.5g/m2 for ≥ 60) plus GO (Day 1) (required for C1, optional for C2, and none beyond C2). Patients with FLT3-ITD/TKD mutations are excluded. Correlative studies for resistance prediction were obtained. Results: At the time of data cut-off, 9 patients (6M/3F) with a median age of 65 (range: 22-69) were enrolled between 1/2023 and 6/2024. By European LeukemiaNet (ELN) 2022 classification, 1 (11%), 2 (22%), and 6 (67%) were favorable, intermediate, and adverse risk, respectively. Mutations defining myelodysplasia-related AML according to the World Health Organization (WHO) 2022 occurred in 5/9 (56%) patients. Most common mutations include ASXL1 (44%), TET2 (44%), RAS (44%), RUNX1 (22%), and DDX41 (22%). There were no patients with TP53 mutation. The median follow-up was 5.9 months (range: 0.8-16.7). At data cut-off, 8 patients were evaluable for response (6 in Cohort A and 2 in Cohort B), and 1 patient is pending recovery marrow. The CR/CRi rate was 75% (6/8) with 5 CR and 1 CRi. For responding patients, 83% (5/6) achieved negative MRD by multiparametric flow cytometry. No patient received more than 1 cycle of consolidation, and 3 patients proceeded to allogeneic hematopoietic cell transplant (alloHCT). Of all treated patients, only 1 patient relapsed after a cycle of consolidation, and this patient never achieved MRD negativity during remission. All patient experienced grade ≥ 3 cytopenias. The median time to neutrophil and platelet recovery to meet the best response (CR: ANC ≥ 1k and platelets ≥ 100k; CRi: ANC ≥ 1k or platelets ≥ 100k) was 45 days (range: 34-76). The only grade 3/4 adverse events (AEs) ≥ 20% was febrile neutropenia (67%). Most common non-hematologic AEs include maculopapular rash (77%) and nausea/vomiting (33%). There were 4 serious adverse events (SAEs) that occurred in Cohort A including grade 3 febrile neutropenia in 3 patients and a grade 3 thrush. DLTs were observed in 2 patients (1 in Cohort A and 1 in Cohort B) where both had grade 4 thrombocytopenia beyond Day 42 of induction. One of these patients (biallelic DDX41 and TET2) proceeded to alloHCT over 1 year ago after induction and maintained CR with negative MRD. The other patient with multiple adverse risk mutations including SRSF2, ASXL1, RUNX1, BCOR, and STAG2 had progressive disease and is on salvage therapy. Conclusions: The combination of CPX-351 and GO induced high CR/CRi rates, with nearly all responding patients achieving MRD negativity in these preliminary results. The safety profile is comparable to prior experience with CPX-351 with the exception of a longer count recovery time. The optimal schedule of GO is still to be determined with final phase 1 results expected at the time of presentation.