Pre-Clinical Modeling of Concomitant Therapy with Azacytidine and Eltrombopag in MDS/AML

Hemorrhage-related morbidity and mortality frequently arises in patients with myelodysplastic syndromes (MDS). A major therapeutic goal in patients with MDS is the alleviation commonly occurring thrombocytopenia. Low platelet counts increase bleeding risk, limit tolerated doses and efficacy of standard therapies, and are associated with an increased risk of leukemic transformation. Preclinical and phase I/II studies on eltrombopag (EPAG), a megakaryopoiesis-stimulating thrombopoietin receptor (TPO-R) agonist, demonstrated safety and efficacy of the small molecule as monotherapy in MDS. However, a phase III study testing the ability of EPAG to increase platelet numbers in patients with intermediate or higher-risk MDS under azacytidine (AZA) therapy (NCT02158936) was prematurely terminated due to inferior performance of EPAG-treated vs. placebo-receiving patients. While the study did not find a difference in overall survival rates between the two groups, concerns pertaining to disease progression under concomitant treatment with EPAG and azacytidine have been raised.

We assessed whether EPAG exerts antagonistic physiological activities in the context of anti-leukemia treatment with AZA exploiting a set of complementary pre-clinical ex vivo assays measuring aberrant cell proliferation, differentiation and self-renewal. MDS and AML patient-derived cell lines (n=4) and primary cell specimen (n=6) were treated with non-cytotoxic, clinically-relevant doses of EPAG and AZA in short-term liquid cultures, myeloid colony-forming assays, and long-term serial-replating assays. To elucidate functional effects of the combination treatment, we determined cell growth, apoptosis, differentiation and long-term colony initiation as a surrogate for self-renewal upon treatment. We also quantified cell surface presence of TPO-R, and assessed pathway activation downstream of the receptor using phosphoflow.

Consistent with previous observations, continuous single agent treatment with either >0.3μM AZA, or 3-20mg/ml EPAG were found to be cytotoxic for leukemic cells. Simultaneous treatment with a non-cytotoxic functionally-relevant dosage of AZA (0.3μM) and EPAG at concentrations above 3mg/ml resulted in a striking decrease of cell viability compared to AZA single agent treatment in short-term cell line cultures. Lower-dose EPAG treatment (0.3 and 0.5mg/ml) along with AZA (0.3μM) did not lead to any significant changes in cell viability and growth in short term cultures (vs. single agent treated controls) of cell lines or primary cells. Apoptosis, cell surface differentiation marker analysis, and morphological assessment revealed no discernable differences upon combination treatment compared to AZA single agent treatment in cell lines. Characterization of aberrant self-renewal using serial replating colony assays revealed a significant decrease in the number of colony-initiating cells upon continuous single agent treatment with AZA, which was not diminished by combination treatment with EPAG. An AZA-resistant cell line also did not show altered long-term colony-forming cell numbers in the presence of both drugs. We found no evidence of increased leukemic self-renewal upon simultaneous treatment using primary cells in 5/6 specimen tested. One sample showed a transient increase in colony formation and average cell numbers per colony in the presence of both drugs, compared to AZA-treated controls, albeit the number of serial replatings was unchanged compared to AZA controls. We tested whether TPO-R/CD110-expressing aberrant cells would expand under EPAG single agent, or combination treatment with AZA, and found no significant changes in the frequency of CD110-positive cells compared to AZA single agent controls. TPO-R downstream signaling in primary mononuclear cells comprised a marked, yet transient impairment in STAT5 and AKT phosphorylation upon combination treatment, compared to single EPAG treatment.

Togehter, our data show that simultaneous treatment of leukemic cells with non-cytotoxic, but clinically-relevant dosages of EPAG and AZA does not lead to a stimulation of leukemic cell growth, even in the presence of TPO-R, in several pre-clinical models. These results do not support a generalized mechanistic antagonism between the two drugs in leukemic cells. Further investigations are underway to understand the impact on non-leukemic hematopoiesis.