Resistance to FLT3 Inhibitors in FLT3-Mutated AML Is Associated with CD45RA+CD200- leukemic Cells and Can be Overcome By Combination with SMAC Mimetics

Acute myeloid leukemia (AML) is a disease caused by abnormal proliferations of myeloid progenitor cells in the bone marrow. About 25-30% of AML patients have a mutation in the FLT3 gene, which is associated with a poor prognosis. Although FLT3-inhibitors (FLT3i) that can target this mutation are clinically approved, approximately 40% of FLT3^mut patients do not respond to FLT3i.

In this study we performed high-throughput functional ex vivo drug testing (n=528) in 63 FLT3^mut patients, with paired MS-proteomics (n=20), RNA-seq (n=20), plasma proteome (n=16), spatial single-cell proteomics (n=6) and mass cytometry (n=3) data with the aim to uncover the functional and molecular landscape linked to FLT3i response.

First, we compared clinical and ex vivo response to FLT3i (midostaurin) to determine the clinical predictability of our ex vivo drug testing. All five ex vivo responders had a complete remission (CR) after first treatment with conventional induction therapy and midostaurin, while four out of five non-responders eventually relapsed or had no CR.

Furthermore, non-responders had an increased immune activation and displayed immature stem cell phenotypes, as well as decreased expression of surface markers associated with mature myeloid cells (e.g., CD64, CD11c, CD33) compared to responders. Moreover, soluble CD40, CD244, PD-L1, CD4, and IL12RB1 were increased in responders, indicating an immune suppressive environment consistent with RNA and protein data. Surprisingly, the T cell receptor proteins CD200 and CD45RA were found to be increased in FLT3i responders and non-responders, respectively.

None of the other 527 drugs tested ex vivo were more effective in the non-responder group, including other FLT3i. Thus, we performed drug combination screening on FLT3^mut and FLT3i resistant cell lines and patient cells to identify effective drug combinations. Combinatorial drug screening in resistant FLT3^mut cells revealed an increase in drug sensitivity to apoptotic modulators, as well as PI3K/AKT inhibitors in combination with midostaurin. Subsequent validation confirmed synergy between the midostaurin and the SMAC-mimetics birinapant and LCL161, while the BH3 mimetic venetoclax and the PI3K inhibitor idelalisib only showed an additive effect.

Our data shows specific differences in myeloid maturation and LSC (Leukemic Stem Cell) phenotype between midostaurin responders and non-responders, together with a potential functional shift in cell signaling in immune signaling and anti-apoptotic pathways. Moreover, ex vivo drug testing data demonstrates that while there is less overall sensitivity to drug treatment in non-responders, combination therapies including apoptotic modulators such as the SMAC mimetics could overcome FLT3i resistance and improve FLT3^mut patient outcomes.