Abstract 1844

Acute myeloid leukemia (AML) can display de novo or acquired resistance to cytosine arabinoside (Ara-C), a primary component of induction chemotherapy. We hypothesized that the changes required by the cells to become Ara-C resistant might also make them more susceptible to elimination by other drugs. We previously developed two highly resistant murine AML cell lines, B117H and B140H, by introducing increasing concentrations of Ara-C to their parental cell lines, B117P and B140P, respectively. The Ara-C resistant derivatives, B117H and B140H, can tolerate Ara-C concentrations ∼1000X that of their drug sensitive parental cell lines. We previously demonstrated that the down-regulation of deoxycytidine kinase was one mechanism of resistance in the Ara-C resistant B117H and B140H cells, a mechanism that has also been documented in human cases of refractory disease. A high throughput drug screen was conducted to determine if an increase in Ara-C resistance would result in increased sensitivity to other drugs. The screen evaluated 446 FDA approved drugs, which included a variety of drug types, including chemotherapy drugs, analgesics, steroids, antidepressants, immunosuppressives, and antipsychotics. Despite the diverse nature of the parental cell lines and their Ara-C resistant derivatives, two drugs (prednisolone and AM404), showed an increase in sensitivity in both Ara-C resistant cell lines. Follow-up drug assays were performed on both prednisolone and AM404 to establish the 50% inhibitory concentration (IC50) in all four cell lines. Both Ara-C resistant cell lines displayed a greater than 3-fold reduction in the IC50 of prednisolone as compared to their parental lines, and the IC50 of AM404 demonstrated a greater than 2-fold reduction. Prednisolone has been shown to induce differentiation in murine AML cells in vitro, and high-dose methylprednisolone has been found to induce differentiation in various subtypes of childhood AML. The mechanism of differentiation by prednisolone is theorized to be independent of glucocorticoid receptor binding. We believe this model will be useful in elucidating the role of prednisolone in leukemic differentiation, and in suggesting new mechanisms to target refractory AML cells.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.