Cladribine-based regimens for Relapsed/Refractory Acute Myeloid Leukemia – a real-world multicenter retrospective study

Abstract Background: Relapsed/refractory (R/R) AML represents an unmet medical need requiring novel salvage approaches to enhance survival outcomes and preserve transplant eligibility, with particular urgency for patients without targetable mutations. The addition of cladribine (Clad) to low/intermediate-dose cytarabine and venetoclax (Ven) has been previously shown as a valid salvage therapy. Aims: We aimed to evaluate the efficacy of Clad-based regimens as salvage therapy in AML patients, in a real-world setting, including Ven-refractory patients. Methods: This was a multi-center retrospective cohort study conducted across five academic centers in Israel. Demographic, clinical, and treatment-related data from consecutive patients with r/r AML were collected. The decision to treat patients with Clad, the specific regimen, and dose modifications were determined at the discretion of the treating physician. Results: Thirty patients with R/R AML were treated with Clad-based regimens between Nov 2019 and Feb 2025, with a median follow-up time of 3 (range, 0.4- 64) months. Median age at was 72 (range, 21-81) years, 67% male, 53% secondary AML. Fifty percent of the patients had refractory disease, while the other had relapsed disease. Twelve (40%) patients had complex/monosomal karyotype. Median number of previous therapy lines was two (range, 1-5). The majority of patients were priorly exposed to venetoclax and azacitidine (87% and 77%, respectively) and 60% had prior allogeneic hematopoietic stem cell transplantation (AHSCT). Among the latter group, median time from HCT to treatment was 6 (range, 1-74) months. The majority (83%) received Clad-LDAC-Ven triplet, one patient received navitoclax as well, one patient received Clad-Ven with decitabine and three patients received Clad-Ven with intermediate-dose cytarabine. Ven was given for a median of 14 (range, 2-21) days. The majority (n=25, 83%) received one cycle. Overall Response Rate (CR+CRi+MLFS) was 50%; CR/CRi rate was 30%. Notably, the presence of complex cytogenetics did not have a significant impact on ORR. Despite limited numbers, none of the patients who harbored MECOM rearrangements (n= 3) or multi-hit TP53 (n=2) mutations responded, while NPM1(n=4) and IDH1 (n= 2) mutations were associated with achievement of CR/CRi (ORR 75% and 100%, respectively). Median overall survival was 3 (95% CI, 0-6) months. Among those who achieved CR/CRi median survival was 8 months. Among those who were bridged to AHSCT/donor lymphocyte infusion (n=9), median survival was 16 months. Hematologic adverse events were most common with all patients experiencing grade 3/4 neutropenia; Infections were relatively common with 50% of patients experiencing grade 3/4 sepsis and 37% of patients experiencing invasive fungal infections. Median time to platelet recovery (PLT>50,000/µL)was 20 days (IQR, 14-26) and to neutrophil recovery (ANC>500/µL) was 18 days (IQR, 9-27). 30-days death rate was 20% (n=6), and was due to the following reasons: infectious complications (n=3), progressive disease (n=2) and GVHD (n=1). Conclusion: This real-world analysis, conducted in a predominantly elderly with R/R AML, suggests that the Clad-Ven-based regimen may offer transient disease control, primarily in patients who subsequently received either HCT or donor lymphocyte infusions. While the regimen may serve as a potential bridging strategy, its overall efficacy appears limited in the absence of consolidative immunotherapeutic interventions and is associated with significant infectious toxicity.