Disrupting monocytic AML and monocytic leukemia stem cells with a novel cladribine regimen: From bench to bedside

Abstract

Abstract Recent studies, including our own, have demonstrated that the heterogeneity of leukemia stem cells (LSCs) drives therapeutic resistance in AML. Specifically, venetoclax (Ven) resistance arises from a monocytic LSC subset (mono-LSCs), which shifts its survival dependence from BCL2 to MCL1, distinguishing it from conventional primitive LSCs (prim-LSCs). Here we propose a novel triple-drug regimen designed to target mono-LSCs and monocytic AML (mono-AML) on three fronts: (1) Cladribine (CdA), which we previously showed selectively eradicates mono-LSCs (Cancer Discov. 2023); (2) Homoharringtonine (HHT), a clinical agent widely used in China that efficiently degrades MCL1 highly expressed in monocytic blasts; and (3) Cytarabine (AraC), the standard chemo backbone. We present results from cell lines, primary AMLs, xenograft models, and clinical data from AML patients treated with the CHA regimen. We first compared the efficacy of the CHA combination and venetoclax/azacitidine (Ven/Aza) in the Ven-resistant mono-AML cell line THP1 in vitro and in THP1-derived xenograft models in vivo. In both settings, CHA demonstrated a synergistic anti-leukemic effect, whereas Ven/Aza exhibited resistance. Further studies in primary AMLs revealed that mono-AMLs were more sensitive to CdA and HHT, while primitive AMLs (prim-AMLs) showed greater sensitivity to Ven, indicating lineage-specific vulnerabilities. To elucidate the mechanism of the selectivity, we analyzed AML specimens with a mixed monocytic-primitive (MMP) phenotype, in which prim-AML and mono-AML subclones with identical oncogenic mutations (verified by whole exome sequencing) coexist. After 48 hours of treatment with Ven/Aza and CHA, tSNE-based flow analysis and single-cell RNA-seq of residual cells revealed that Ven/Aza selectively cleared the prim subpopulation, while CdA preferentially targeted the mono counterpart. HHT primarily affected more differentiated monocytic blasts, and AraC exhibited no significant lineage preferences. To assess the selectivity at the LSC level, functionally validated prim-LSC-driven prim-AML and mono-LSC-driven mono-AML samples were treated ex vivo with vehicle, Ven/Aza, or CHA, followed by transplantation into NSG-S mice. This allowed direct measurement of the impact of Ven/Aza and CHA on LSC sub-compartments, using engraftment as a functional readout. We found that Ven/Aza diminished the engraftment capacity of prim-AML while sparing mono-AML, consistent with its reported activity against BCL2-dependent prim-LSCs. In striking contrast, CHA caused a dramatic loss of engraftment potential in mono-AML but left prim-AML relatively unaffected, highlighting its unique ability to eradicate Ven/Aza-resistant mono-LSCs. Notably, the anti-mono-LSC activity was primarily attributed to CdA, with HHT and AraC playing a supportive role in eradicating more mature monocytic blasts. Given the promising preclinical results, we initiated a prospective single-center, single-arm clinical trial (NCT05906914) to evaluate the clinical efficacy and safety of CHA. By May 31, 2025, 71 Asian Chinese patients with newly diagnosed AML (median age: 42; range: 18–60; 50.7% male) had been enrolled. CHA demonstrated an excellent CR rate and median overall survival (mOS) in the mono-AML (FAB M4/M5) subgroup, noninferior to the prim-AML (FAB M0/M1/M2) subgroup (76.2% vs. 63.0%, p =0.24; mOS was not reached for either subgroup, p =0.92). In contrast, a retrospective cohort treated with Ven/HMA during the same period in our center showed a significantly lower CR rate and mOS in the mono subgroup than in the prim (56.9% vs. 75.6%, p=0.02; 11.3 vs. 16.3 months, p=0.08), consistent with a recent US/European multicenter analysis (Blood Cancer Discov. 2025). In a genetic subgroup (RAS/FLT3-mut, TP53/NPM1/IDH-wt), CHA was even more superior than Ven/HMA (78.8% vs. 34.8% CR). CHA regimen was also well tolerated, with a 30-day treatment-related mortality rate of 1.4%. Single-cell RNA-seq analysis of specimens from CHA-treated patients is ongoing to further elucidate the mechanism. In summary, in the Ven/HMA-dominated era of AML targeted therapies, effective treatments for mono-LSCs and mono-AML remain an unmet clinical need. We propose a novel and effective CHA regimen that is readily applicable in the clinic and demonstrates the ability to disrupt mono-LSCs and mono-AML. A clinical trial evaluating an alternating Ven/Aza and CHA regimen is currently in development.