Abstract
Background:
Refractory/relapsed acute myeloid leukemia (RR-AML) remains a significant therapeutic challenge, with long-term survival rates remaining poor despite advances in supportive care and targeted agents. There is currently no universally accepted standard of care for adult RR-AML; treatment selection is influenced by patient age, comorbidities, and duration of first remission. For medically fit patients, high-intensity salvage regimens such as CLAG or FLAG-IDA incorporate purine analogues (cladribine or fludarabine) with high-dose cytarabine and granulocyte colony-stimulating factor (G-CSF), achieving complete remission (CR) rates between 30% and 60% in the relapsed setting. The addition of venetoclax to purine analogue-based regimens has shown promising efficacy in RR-AML, with MD Anderson experience demonstrating a composite complete remission rate (CRc) of 67% overall and an impressive 79% overall response rate (ORR) in first salvage patients with wild-type TP53, along with high rates of measurable residual disease (MRD) negativity and encouraging survival outcomes (DiNardo et al. Leukemia, 2025). However, the optimal approach and long-term benefit of venetoclax-enhanced purine analogue-based therapy in high-intensity salvage contexts continue to be defined in real-world practice.
Methods:
We retrospectively reviewed patients with AML who received purine analogue-based regimens in the refractory relapsed setting at the Medical University of South Carolina, including FLAG-IDA +/- venetoclax, CLAG-IDA +/- venetoclax. Primary endpoints: CR, ORR (CR + CRi), and MRD-negativity by multiparametric flow cytometry (if available). Secondary endpoints: relapse rates, allo-SCT rates, infection rates, median duration of hospitalization, and additional G-CSF use after using the purine analogue-based regimens.
Results:
We identified 18 patients with RR-AML diagnosed between 2020 and 2025. Median age at diagnosis was 52 years (range 26-69), 65% were males, and 65% were white. Nine patients (53%) had complex cytogenetics, 1 (6%) had secondary treated AML, 3 (18%) had bi-allelic TP53 mutations, 3 (18%) had FLT3-ITD mutations (all received midostaurin on first induction), 1 had IDH1 mutation (6%), and 12 (67%) had adverse risk by ELN-2022 risk stratification. The median number of prior AML treatment lines was 1 line (range 1-2). Five patients (29%) received FLAG-IDA (2 of them received a concomitant FLT3 inhibitor: midostaurin and quizartinib), 7 patients (41%) received FLAG-IDA + venetoclax, 4 (24%) received CLAG-IDA + Venetoclax, and 1 patient (6%) received CLAG + VEN. The ORR rate after receiving purine-analogue containing regimens was 82.3%, and MRD-CR rate was 47%. Six patients (35%) were able to have an allo-SCT. A total of 5 relapses occurred, 3 of them post-allo-SCT. The median duration of hospitalization while receiving the purine analogue regimens was 26 days (range 7-64) and 65% of patients had a documented infection. Three patients (18%) required additional G-CSF use due to severe neutropenia.
Conclusion:
Our single-institution experience demonstrates that high-intensity purine analogue-based therapy is safe and active in patients with RR-AML and can achieve meaningful remission rates in a heavily pretreated, predominantly adverse risk population. These outcomes compare favorably to historic expectations and support the use of purine analogue regimens as an effective salvage strategy. Prospective studies and continued real-world reporting are needed to refine risk stratification and optimize post-remission management, including timely allogeneic transplantation or transition to clinical trials.
