A randomized phase 2 trial of CPX-351 vs. CLAG-m (Cladribine, High-Dose Cytarabine, G-CSF, and Mitoxantrone) in medically less-fit adults with previously untreated acute myeloid leukemia (AML) or other high-grade myeloid neoplasms Free

Background: While treatment options have significantly expanded for medically less-fit adults with newly diagnosed AML over the last decade, the optimal therapeutic approach remains uncertain. Retrospective data from the pre-venetoclax era have suggested intensive therapies may lead to better outcomes than non-intensive therapies in such patients, but few prospective comparisons have been conducted and the relative merits of individual treatment regimens are largely unknown. A previous non-randomized trial at our center suggested that CLAG-M, our institutional standard induction regimen for fit adults with AML, compared to dose-reduced CLAG-M, led to higher rates of response with similar toxicity in less fit patients. We therefore conducted a single-institution, randomized, non-blinded trial (NCT04195945) comparing CPX-351 with CLAG-M in less-fit adults with untreated AML or other high-grade myeloid neoplasm (≥10% blasts in blood or marrow).

Methods: Adults ≥18 years were eligible provided they were medically less fit defined by a Treatment-Related Mortality (TRM) score of ≥13.1, corresponding to >10-15% 28-day mortality with intensive chemotherapy (Walter RB et al. JCO 2011; 29 [33]). The only organ function exclusion was left ventricular ejection fraction ≤45%. CPX-351 doses were: daunorubicin 44 mg/m2 and cytarabine 100 mg/m2 on days (D) 1, 3, and 5 during induction and D1 and 3 for post-remission. CLAG-M doses were: cladribine 5 mg/m2/day (D1-5), cytarabine 2,000 mg/m2/day (D1-5), G-CSF 300 or 480 µcg/day for weight </≥76kg (D0-5), and mitoxantrone 18 mg/m2/day (D1-3). Cytarabine 500 mg/m2 given daily D1-5 was used for post-remission therapy. Up to 2 cycles of induction and up to 4 cycles of post-remission therapy were allowed. Commercially available FLT3-inhibitors could be added in both arms. The primary endpoint was 3-month overall survival (OS). Secondary outcomes included rate of complete remission (CR) negative for measurable residual disease (MRD), overall response rate (ORR) of CR plus CR with incomplete hematologic recovery (CR+CRi), response duration, day 28 and 56 mortality, toxicity profile, relapse-free survival (RFS), and OS.Results were analyzed as intention to treat.Data cutoff was June 9, 2025.

Results: 60 patients were enrolled from 5/2020-3/2025 and randomized 1:1 to CPX-351 or CLAG-M. All patients were treated per assigned arm and received at least 1 dose of study therapy, though 2 patients on CPX-351 and 3 patients on CLAG-M did not complete the full course of induction therapy due to rapid disease progression, patient choice, or death. Baseline characteristics were overall well balanced with a median age of 72 in each arm, ECOG performance status of 3-4 in 77% (CPX-351) and 60% (CLAG-M; p=0.11), and TRM score of 28 (CPX-351) and 23 (CLAG-M; p=0.16). About 2/3 of patients in each arm had ELN2017 adverse risk disease, including 27% in each arm with a TP53 alteration. Favorable risk disease was present in 20% (CPX-351) vs. 13% (CLAG-M); secondary disease was 37% (CPX-351) vs. 43% (CLAG-M). Response rates were numerically but not statistically significantly higher with CLAG-M vs. CPX-351 (MRDneg CR: 37% vs 13%; MRDpos CR: 13% vs. 13%; CRi: 23% vs. 20%, for an ORR of 73% vs. 47% [p=0.07 for MRDneg CR and p=0.21 for ORR]). The rate of resistant disease was lower with CLAG-M (7% vs. 23%). Early mortality was similar between CLAG-M and CPX-351 at day 28 (20% vs. 23%) and day 56 (23% vs. 33%). Only 7 (23%) of patients on the CPX-351 arm and 3 (10%) on the CLAG-M arm received >1 cycle of study treatment. In the CPX-351 arm, post protocol therapy was lower intensity in 37%, no therapy in 50%, and high-dose cytarabine in 13%. In the CLAG-M arm, post protocol therapy was lower intensity in 43%, no therapy in 33%, and high-dose cytarabine in 20%. The most common causes of death in both groups were leukemia and infection; 1 patient died in remission in the CPX-351 group vs. 2 in CLAG-M. Median RFS and OS were non-significantly higher with CLAG-M compared to CPX-351 (RFS: 8.9 vs. 4.3 months; p=0.70; OS: 10.5 vs. 5.8 months, p=0.97). 3-month OS trended to be higher with CLAG-M than CPX-351 (70% vs. 60%, p=0.09).

Conclusions: In this objectively frail, comorbid AML population, CLAG-M led to numerically but not statistically significantly higher remission rates and survival than CPX-351. A randomized comparison of intensive vs. lower-intensity therapy in a similar patient population is planned.