Gain-of-Function KIT Mutations Sensitize the Mutant Isoform to the Type I Tyrosine Kinase Inhibitor Crenolanib: A Rationale for the Therapeutic Use in Systemic Mastocytosis (SM) and Core Binding Factor Leukemias (CBFL)

Activating mutations of the class III receptor tyrosine kinases FLT3 and KIT are associated with certain human neoplasms, including hematologic malignancies, i.e. the majority of patients with systemic mast cell disorders (KIT) and subsets of patients with acute myelogenous leukemia (FLT3 and KIT). Crenolanib is a potent selective FLT3 inhibitor with high efficacy against internal tandem dupliction mutations (ITD) – but also secondary kinase domain mutations conferring resistance towards other TKI. Interestingly, crenolanib does not target the wildtype KIT isoform, which is believed to reduce clinical side effects such as prolonged myelosuppression observed with other TKI. Clinical studies are currently enrolling.

We now show that gain-of-function mutations of KIT, including codon D816 alterations as the most prevalent mutation in SM and CBFL, sensitize the mutant isoform to crenolanib. Several mast cell and leukemia cell lines harboring autoactivating KIT or FLT3 isoforms were treated with crenolanib in dose dilution series (MOLM14, MV4;11, HMC1.1/1.2, p815). To minimize cell-type specific off-target effects, an isogenic cell model was established. The murine pro B-cell line Ba/F3 was retrovirally transduced with either a FLT3 ITD or a KIT D816 isoform. Apoptosis induction was analyzed by annexin V-based assays. FLT3/KIT tyrosine phosphorylation was assessed by western immunoblots.

As previously described, the FLT3 ITD positive cell line MOLM14 revealed high sensitivity towards crenolanib with IC50s in the lowest nanomolar range. We also confirmed high sensitivity towards crenolanib ex vivo in the low nanomolar range in a native sample of a heavily pretreated patient. This patient relapsed with FLT3 ITD positive leukemia harboring a secondary D835H mutation in a subclone. Interestingly, leukemia cells in the relapse situation were much more oncogene-addicted than cells at primary diagnosis, which is in line with previous findings by others. Due to the structural homology of FLT3 D835 and KIT D816 mutations, we extended our studies to mutant-KIT mastocytosis and leukemia cell models and confirm clinically relevant antiproliferative as well as proapoptotic sensitivities towards crenolanib: for HMC mastocytosis cells harboring a KIT V560G and/or a D816V mutation, potent induction of apoptosis was observed with IC50s of 100-250nM. The murine p815 mastocytosis cell line (harboring a D814Y mutation corresponding to D816Y in humans) demonstrated a proapoptotic effect of crenolanib with an IC50 of 60 nM. Treatment of corresponding KIT or FLT3 isoform-transduced Ba/F3 cells confirmed similar IC50s in the leukemia cell lines. Parental Ba/F3cells did not show any sensitivity towards crenolanib up to concentrations of 1000 nM. Additionally, potent dephosphorylation at 100 nM of KIT D816V in Ba/F3 and HMC cells after exposure to crenolanib confirmed mutant-KIT as a target of the drug. Evaluation of a broader range of native mast cell and leukemia patient samples as well as additional leukemia cell lines and isogenic Ba/F3 KIT or FLT3 transfectants is ongoing. First results demonstrate activity of crenolanib in native cells of a subset of patient samples with SM or CBFL treated ex vivo. Even more, combination of crenolanib with anthracyclines revealed additive to superadditive proapoptotic effects. Moreover, combination of crenolanib with cladribine, a hallmark agent in the treatment of systemic mastocytosis, resulted in potent induction of apoptosis already at doses that did not display any proapoptotic effects when administered as single agents, thereby providing a rationale for combinatorial therapeutic approaches.

In summary, crenolanib is effective against the KIT D816V isoform associated with several hematologic malignancies. Notably, while not as effective towards mutant-KIT compared to the FLT3 ITD isoform, the observed estimated IC50 of crenolanib is well in the range of achievable plasma concentrations and in the range of the potent KIT inhibitor dasatinib, which is successfully under clinical investigation in CBFL. Our data provide a rationale to test crenolanib as a potent inhibitor of mutant-KIT isoforms in KIT-associated neoplasms.