Inhibitory Effects of Midostaurin and Blu-285 on Myeloid Progenitor Cells Derived from Patients with Multi-Mutated KIT D816V+ Advanced Systemic Mastocytosis

Recent data have highlighted that the molecular pathogenesis of advanced systemic mastocytosis (advSM) is complex. In addition to the phenotypically most important mutations in KIT, e.g. KIT D816V in 80-90% of patients, one or more additional mutations, e.g. in SRSF2, ASXL1, RUNX1, CBL, JAK2 and others, are present in 60-70% of patients (Jawhar et al., Leukemia 30, 2016). In individual patients, a complex mutational profile is detected not only in mature mast cells (MCs) but also in myeloid progenitors derived from granulocyte-macrophage colony-forming progenitor cells (CFU-GM), indicating multi-lineage involvement of all identified mutations in the vast majority of patients (Jawhar et al., Leukemia 29, 2015). Midostaurin, a multi-targeted kinase inhibitor, has demonstrated an overall response rate of 60% in advSM patients (Gotlib et al., NEJM 374, 2016). BLU-285 is a highly selective KIT D816V kinase inhibitor which has demonstrated biochemical activity on the mutated KIT enzyme (KIT D816V IC₅₀ = 0.27 nM). In the current study, we sought to a) investigate the inhibitory effects of midostaurin and BLU-285 on single-cell-derived CFU-GM from bone marrow mononuclear cells derived from multi-mutated KIT D816V+ advSM patients and b) correlate the midostaurin CFU-GM data with clinical and various response parameters in midostaurin-treated advSM patients. The mutational status of CFU-GM colonies (median colonies per patient, n=20; range 10-30) was analyzed for KIT D816V and additional mutations by PCR followed by Sanger Sequencing. In 10 multi-mutated advSM patients (aggressive SM [n=8] or mast cell leukemia [n=2] with an associated hematological neoplasm), CFU-GM colonies were screened prior to midostaurin (month 0, n=10) and, if available, at month 6 on midostaurin (n=8). At month 0, a median of 90% (range, 40-100) CFU-GM colonies were KIT D816V+, while at month 6 a median of 70% (range, 5-100) CFU-GM colonies were KIT D816V+. A significant relative reduction (≥50%) in the proportion of KIT D816V+ colonies at month 6 was observed in 4/8 (50%) patients. Midostaurin-naïve CFU-GM were incubated with midostaurin at concentrations up to 1000 nM and showed a dose-dependent significant reduction (≥50%) of KIT D816V+ colonies in 1/7 (14%) patients. Overall, the in vitro effects correlated with the in vivo effects of midostaurin on CFU-GM and established IWG-MRT-ECNM response criteria (e.g. mast cell infiltration in BM, serum tryptase level) and KIT D816V allele burden in peripheral blood. Midostaurin-naïve CFU-GM from 7/10 (70%) patients were also incubated with different concentrations of BLU-285 ranging from 0 to 75 nM. A dose-dependent, significant relative reduction (≥50%) of KIT D816V+ CFU-GM colonies was observed at concentrations between 45 and 75nM in 5/7 (71%) patients. Of interest, 3/5 (60%) in vitro responders to BLU-285 were resistant to midostaurin (in vivo and in vitro) while CFU-GM colonies from 2 patients resistant to BLU-285 were also resistant to midostaurin. In addition to KIT D816V, recurrent molecular aberrations (median 2/patient, range 1-3) were identified in all patients, most frequently in SRSF2 (n=9), TET2 (n=7) and ASXL1 (n=4). Neither drug had an effect on the relative frequency of additional mutations in CFU-GM colonies. In summary, we conclude that a) the relative reduction of KIT D816V+ CFU-GM colonies between month 0 and month 6 on midostaurin correlates with clinical response, b) the CFU-GM colony assays may provide useful information for prediction of response to midostaurin, c) the highly selective KIT D816V inhibitor BLU-285 has significant activity against KIT D816V, even in cases which are resistant to midostaurin, and d) neither drug had an effect on the prognostically relevant additional mutations.