Optimal Timing for Response Assessment in Acute Myeloid Leukemia Patients Treated with CPX-351 Induction: Real World Experience from 8 Spanish Hospitals

INTRODUCTION: CPX-351 is a liposomal formulation of daunorubicin and cytarabine approved for induction therapy in patients (pts) with high-risk acute myeloid leukemia (AML) and secondary AML. Pre-clinical data suggested increased and prolonged drug accumulation and growth suppression of CPX-351 in leukemic cells in bone marrow (BM) (Lim et al, Leukemia Research, 2010). This pharmacodynamic singularity challenges the usefulness of early BM (eBM) assessment. In the pivotal study, eBM was required at day 14- 21 after every induction and at recovery to confirm response. In case of unclear eBM at day 14-21, a new BM was recommended to determine the need for a 2nd cycle. However, given the different pharmacodynamic of CPX-351 and as the experience of using this drug increases, there are currently concerns about the value of this eBM and the optimal time to confirm a response (Ronnacker et al. EHA 2024). We aim to investigate the clinical utility of eBM evaluation after induction in pts treated with CPX-351. METHODS: We retrospectively review the evolution of pts treated with CPX-351 between June 2018 and June 2024, from 8 different Spanish centers. Only 83 pts were included as they reported information about all BM responses. Blast clearance (BC) was defined as <5% blasts in eBM. Induction response (IR) was defined according to ELN2022 criteria, ncluding complete remission (CR) or CR with incomplete hematologic recovery (CRi), morphologic leukemia free state (MLFS) and partial response (PR). RESULTS: 83 pts received at least one induction cycle with CPX-351. Median age was 64 y (20- 78). Fifty-two male and 31 female. At diagnosis median leukocytes, LDH and blasts in BM was 2.9·10 9 /L (IQR 1-11), 277 (200-476) and 27 % (20-50,7). We found 13 pts with dry aspirate: 12/13 with fibrosis, grade 1(3),2(2) and 3 (7). TP53 was mutated in 15 (18%) and 12 (14%) had complex karyotype. The majority of pts had adverse risk according to ELN2022 criteria: 59 (71%) adverse risk,18 (22%) intermediate and 6 (7%) favorable. Overall response rate (ORR) was 45 (54%). Three of those pts required 2 cycles to obtain a response. Measurable residual disease (MRD) assessed by multiparametric flow cytometry (MFC) was available in 40 pts, of which 23(52%) showed MRD negativity at the end of induction (22 after cycle 1 and, 1 after 2nd). Median time (days) to neutrophil (>500 /µL) and platelet (>20000 platelets/µL) recovery was 34 (28-47,5) and 33,5(IQR 27-44) respectively, which is consistent with increased drug exposure in the BM of CPX-351-treated pts. After first induction, all pts had at least one BM performed. Forty-five (54%) pts had more than 1 BM (1-3). Median time to 1st evaluation was 25 days (IQR14-33). All pts in which the 1st BM assessment was performed before neutrophil recovery (48), needed at least one more evaluation. In those performed after neutrophil recovery (35), only 5/35 needed repetition of BM study, and in only 1 case response status was changed after repetition (from MLFS to persistent disease). We identified 32 (39%) pts with eBM evaluation (between days 12 and 21). About 2/3 of pts were in BC (22, 69%) and 1/3 with persistent leukemia (10, 31%). All 32 pts were re-evaluated with a 2nd BM, and we found the following results: 13(41%) finally achieved response (9 CR, 3 MLFS and 1 PR) while 19(59%) did not. Although 13 (41%) pts with BC in the 1st BM achieve IR in the 2nd BM, 9(59%) with BC in the 1st BM, failed to achieve IR in the 2nd one. On the other hand, of the 51 pts in which no eBM was performed, ORR was 32 (63%) (30 RC/RCi and 2PR). Baseline blast count was higher in those not achieving BC than in those who achieved BC (63 vs 45%, p<0,05). However, no differences were found in terms of age, baseline leukocytes, gender, ELN risk group, TP53 mutations or myelodysplasia defining cytogenetics or molecular abnormalities. Early BC had a sensitivity for IR of 100%, but specificity of 52%. BC had negative predictive value of 100% and a positive predictive value of 59%. CONCLUSION Early blast clearance with CPX-351 showed low predictive value for response (as other roups recently published). Thus, we conclude that early BM evaluation can be confusing in the clinical management and decision making of patients treated with CPX-351 and it is recommended to perform it later. The results should be validated in a larger number of patients and if so, new studies should be proposed to address the optimal time to perform the 1st evaluation after induction with CPX-351.