Introduction: Patients with acute leukemia harbor a population of chemotherapy-resistant leukemia stem cells (LSC) which ultimately leads to disease relapse in many patients. In animal studies, we demonstrated that AKT-β-catenin interaction, indicated by Akt phosphorylation of β-catenin at serine 552 (pS552), was critical to LSC self-renewal and LSCs increased after cytotoxic chemotherapy. High throughput screening of over 1,100 FDA-approved compounds identified doxorubicin (DOX) and daunorubicin (DNR) as potent inhibitors of pS552. To better understand the role of DOX and DNR in specifically targeting LSC through pS552, we experimented with lower dosing schedules, recognizing that most AML patients receive high dose DNR, yet LSC persist and result in disease relapse. In a murine model, 1/40thof the conventional dose of DOX for 5 consecutive days (=1/8 total dose) decreased LSCs, increased hematopoetic stem cells and inhibited S552 phosphorylation in leukemic cells compared to baseline or post-chemotherapy. Using a phospho-specific monoclonal antibody, we tested 30 anonymized human leukemia specimens for immunohistochemical staining of β-catenin S552. Positive staining was seen in 18/30 patients with AML or ALL and supported inclusion of both AML and ALL in a pilot trial to test an inhibitor of β-catenin S552 phosphorylation.

Methods: This pilot study (NCT02914977) was designed to determine the molecular pharmacodynamic effects of low dose DNR on β-catenin phosphorylation in serial bone marrow samples in adult patients with relapsed or refractory acute leukemia. Exploratory objectives included measurement of treatment effects on the leukemia stem cell population in serial bone marrow specimens. At least two prior induction attempts were required for fit patients with primary refractory leukemia; unfit or relapsed patients were allowed entry with one prior therapy. Patients had an ECOG performance score of 0-3, adequate hepatic and renal function and cardiac ejection fraction ≥45%. Exclusion criteria included presence of acute promyelocytic leukemia, CNS leukemia or total lifetime anthracycline exposure exceeding the equivalent of 900 mg/m2 of DNR. Treatment consisted of bone marrow aspiration for correlative studies followed by one cycle of low dose DNR (6.75mg/m2 x 5 consecutive days, days 1-5) and second bone marrow aspiration on day 8. Pharmacokinetics of low dose DNR were measured. Bone marrow samples from pre- and post-treatment were analyzed by flow cytometry for the frequency (%) of total cells that are pS552-β-catenin positive AML LSCs (defined as CD34+ CD38- cells that express TIM3, which distinguishes LSCs from normal stem cells).

Results: To date, 9 patients with relapsed/refractory leukemia have been enrolled. Here we report results of 8 patients with AML. Patient characteristics are summarized in Table 1. Patients had received a median of 2.25 prior regimens (1-3) and 6/8 had previous DNR (2/8 were treated with hypomethylating agent alone or in combination with an investigational agent). AML risk classification based on diagnostic cytogenetic and molecular testing was favorable (2/8) and poor risk (6/8). Two patients had an allogeneic stem cell transplant (SCT) in first complete remission with subsequent relapse. Non-hematologic treatment-related AEs included fatigue, nausea and febrile neutropenia. Plasma samples were analyzed for pharmacokinetics (Table 2). Pre- and post-treatment bone marrow cells were analyzed by flow cytometry for the frequency (%) of total cells that are pS552-beta-catenin positive AML LSCs (Figure 1). 4/8 patients had decreased pS552+ LSCs, 2/8 were unchanged and 2/8 had increased pS552+ LSCs. Of four patients with decreased LSCs, one had prior SCT and 2/4 had prior anthracycline. Two patients with increased LSC number had received prior anthracycline induction; one had prior SCT.

Conclusions: Low dose DNR was safe and well-tolerated in this pilot trial of a single cycle of therapy. In 50% of patients, low dose DNR targeting pS552 in LSC resulted in a decrease in the number of LSC as measured by flow cytometry. Further analysis is needed to better identify biomarkers which may predict LSC responsiveness to low dose DNR. This proof of concept pilot study has validated pS552+ as a LSC target clinically in patients with AML. Additional studies to incorporate this LSC-targeted therapy for patients with AML in remission are planned.


Lin: Jazz Pharmaceuticals: Consultancy. Perez: Bristol-Myers Squibb: Employment.

Author notes


Asterisk with author names denotes non-ASH members.