P691: MULTIOMIC SINGLE-CELL APPROACH REVEALED CLONAL ARCHITECTURE OF KIT MUTATIONS IN CHRONIC MYELOMONOCYTIC LEUKAEMIA PATIENTS WITH/WITHOUT SYSTEMIC MASTOCYTOSIS

Background: Systemic mastocytosis (SM) involves neoplastic expansion of mast cells associated with activating KIT-D816V mutations. Most Advanced SM patients harbor another hematological neoplasm (“SM-AHN”): often chronic myelomonocytic leukaemia (SM-CMML). Monocytes in SM-CMML can be KIT mutated (KIT^MT), as can hematopoietic stem cells (HSCs), confirming KIT^MT as not restricted to the mast cell lineage. Conversely, >5% of CMMLs harbor KIT mutations without overt SM. Treatment for SM targets KIT^MT with tyrosine kinase inhibitors. However, the relative clonal placement of KIT^MT, how this contrasts with KIT^MT in other CMMLs, and longitudinal dynamics of competing clones remain unknown.

Aims: We investigated the clonal architecture and hierarchical differentiation landscape of KIT mutations in CMML, with/without concurrent SM, through disease evolution and response to therapy.

Methods: Ten bone marrow (BM) samples from 8 KIT^MT CMML patients were processed: (a) SM-CMML, n=4; (b) non-SM KIT^MT CMML (“KIT-CMML”), n=3; and (c) SM-CMML post-treatment, n=3 (Fig 1A). BM mononuclear cells were genotyped on the single cell (sc) Tapestri platform, using a custom DNA panel (263 amplicons) and antibody cocktail against 42 surface proteins. Mutations were called, clones resolved and cells annotated using a novel custom pipeline.

Results: In total 42,061 cells were genotyped. Median 5 clones and 4 mutations were detected per sample. Median age of SM-CMML was higher than KIT-CMML (76 v 60y, p=0.057). SM-CMML was borderline associated with fewer Ras-pathway mutations (p=0.045) and more transcription factor-mutated clones (p=0.052). All cases harbored a solitary KIT^MT clone. Interestingly, all KIT^MT occurred alongside concurrent, pre-existing “CHIP” mutations (most commonly TET2). Number of concurrent mutations in KIT^MT clones was higher in SM-CMML (median 3 v 2, p=0.036). The fraction of KIT^MT cells was lower in SM-CMML than KIT-CMML (median 5.4% v 50.7%, p=0.029). KIT mutations were subclonal, in a minority clone, in all SM-CMML samples. By contrast, in every KIT-CMML case the KIT mutation occurred in the dominant (or near-dominant) clone.

Among 19 populations separated immunophenotypically (Fig 1B), as expected SM-CMML carried higher mast cell proportions (median 2.8% v 0.9%, p=0.002). Interestingly, the proportion of KIT^MT mast cells was similar in SM-CMML and KIT-CMML. Meanwhile, the proportion of KIT^MT HSCs was markedly higher in KIT-CMML (median 58.6% v 0.6%, p=0.008), consistent with origination in a more apical HSC with multilineage differentiation potential.

Post-treatment samples all displayed suppression (or loss) of KIT^MT clones, but marked expansion of clones carrying epigenetic (ASXL1/IDH1/IDH2) and/or signalling (NRAS) genes (Fig 1C): commensurate with observed patterns of non-mast cell disease coming to dominate the clinical picture. Lymphoid lineages were enriched post-treatment, while primitive fractions were also suppressed.

Conclusion: We characterize the clonal architecture of KIT^MT-CMML with/without SM, noting distinct patterns of KIT^MT origination and propagation. KIT mutations were always detected in apical HSCs, accompanied by CHIP mutations: contrasting with sporadic SM, in which KIT is typically the sole mutation. The fraction of KIT^MT cells was much lower in SM-CMML, indicating SM as a subclonal branch of a broader disease complex. Further studies are needed to decipher the cell-of-origin, pathogenesis trajectories and clonal relationships within SM-AHN. However, we show progression clearly driven by KIT-WTwild type clones, highlighting the importance of therapeutic targeting beyond the KIT^MT clone and potential for serial sc genotyping to personalize adaptive therapy.

Figure 1.

Keywords: Kit, Systemic mastocytosis, CMML, Clonal expansion