Acute myeloid leukemia (AML) is a clonal hematopoietic disorder characterized by an accumulation of immature leukemic cells. The incidence of AML sharply increases with age. However outcomes for AML in the elderly are dismal with low response rates and high rates of relapse. Primitive AML leukemia stem cells (LSC) are resistant to conventional therapy, and potentially contribute to propensity to relapse. Therefore, new strategies to enhance targeting of AML LSCs and are required to improve outcomes. AML cells demonstrate several epigenetic alterations including enhanced promoter hypermethylation that may contribute to altered growth and self-renewal of LSC. The DNA methyltransferase (DNMT) inhibitor decitabine (DAC) has shown activity against AML LSC in preclinical studies. In clinical studies DAC is well tolerated as monotherapy and induces responses in approximately 20% of AML patients who are not candidates for conventional chemotherapy. Therefore there is considerable interest in developing combinations of DAC with other therapies for elderly patients with AML. We have previously shown that LSC from AML patients with poor risk cytogenetics demonstrate increased JAK2 tyrosine kinase activity and enhanced sensitivity to JAK2 inhibition. Since cross regulation between JAK/STAT and DNMT pathways has been reported, we evaluated the activity of the combination of DAC with the JAK2 tyrosine kinase inhibitor (TKI) Ruxolitinib (RUX) on CD34+ cells from elderly AML patients (>60 years). Exposure to low dose DAC (10 to 200nM) for 72 hours significantly inhibited the growth of primary AML CD34+ cells in a CellTitre Glo assay in a dose dependent manner (n=12, DAC 10nM 88±9% of control, 50nM 77±9%, 100nM 64±11% and 200nM 56±15%). DAC did not induce significant apoptosis in AML CD34+ cells (n=10, control, 22±5% of Annexin V positive cells vs 24±7% for the samples treated with 200nM of DAC) but dramatically and significantly reduced AML CD34+ cell growth in CFC assays (33±3%, 57±14%, 75±3% and 86±7% inhibition with 10, 50, 100 and 200nM of DAC respectively). The combination of RUX (100nM) and DAC (10 and 50nM) significantly enhanced inhibition of AML CD34+ cell growth compared to RUX or DAC alone (DAC 10nM 87±10% of control for, RUX 62±20%, DAC10+RUX 46±20%, DAC50nM 75±10%, and DAC50+RUX 43±20%, n=7, p<0.009). Combination of RUX (100nM) with 10 and 50nM DAC did not significantly enhance apoptosis in AML (n=10) or normal stem/progenitor cells (n=8, Ctrl 20.6±5%, DAC10nM 21.6±5%, RUX100nM 31.5±5%, DAC10-RUX 32.1±5%, DAC50nM 22.6±5% and DAC50-RUX 32.4±8%), but resulted in significantly enhanced inhibition of AML CD34+ CFC compared to RUX or DAC alone (n=4, CFC inhibition with DAC10 34±4%, RUX 46±15%, DAC10+RUX 68±17%, DAC50 56±15%, and DAC50+RUX 77±7%, p<0.004), but did not enhance inhibition of CB CFU-GM (n=4, CFC inhibition with DAC10 0±12%, RUX 24.3±16%, DAC10+RUX 4±9%, DAC50 9±16%, and DAC50+RUX 25±11%). These results indicate that combined targeting of both JAK/STAT and DNMT pathways can selectively inhibit the growth capacity of AML compared with normal stem/progenitor cells. We performed Western blotting on AML CD34+ cells (n=5) treated for 3 days with RUX 100nM, DAC 50nM or the combination of RUX and DAC to evaluate expression of DNMT1, DNMT3a, DNMT3b, phosphorylated STAT3 (Y705) and STAT5 (Y694) and SOCS3. Compared to the individual drugs, the combination resulted in greater reduction of p-STAT3 in 4 out 5 AML patient samples that was associated with increase in SOCS3 in 2 samples, suggesting a crosstalk between the JAK/STAT and DNMT pathways. Interestingly we did not observe enhanced inhibition of p-STAT5 with the combination. Two of five AML samples displayed decrease of DNMT1 expression and 3 of 5 samples decreased DNMT3a and DNMT3b expression upon treatment with the combination. In conclusion, our studies suggest that there may be a crosstalk between the JAK/STAT/SOCS3 signalling and DNA methylation mechanisms in CD34+ cells from elderly AML patients, and indicate that JAK2 TKI in combination with low doses of DAC selectively and significantly enhance inhibition of stem/progenitor cells from elderly AML patients compared to DNMT inhibitors alone. These results support further evaluation of this novel combination for the treatment of AML in the elderly.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.