Clonal persistence of BRAFV600E under MEK inhibition in histiocytosis

Abstract

Abstract BackgroundLangerhans cell histiocytosis (LCH) and Erdheim-Chester disease (ECD) are rare histiocytic disorders driven by activation of the Mitogen-Activated Protein Kinase (MAPK) pathway, mostly BRAFV600E. Although MEK inhibitors (MEKi) induce clinical responses, relapses following discontinuation suggest incomplete clonal clearance. We investigated BRAFV600E persistence and its immune cell distribution under MEKi therapy in LCH and ECD blood and bone marrow samples. MethodsWe analyzed 181 blood or bone marrow samples from 149 histiocytosis patients (72 ECD, 66 LCH and 11 mixed) enrolled in the Histiotarget protocol (NCT04437381). Bone marrow mononuclear cells (BMMCs) were collected from 62 patients (39 ECD, 23 LCH), and peripheral blood mononuclear cells (PBMCs) from 97 patients (43 ECD, 54 LCH). BRAFV600E variant allele frequency (VAF) was quantified by droplet digital PCR (ddPCR) on enriched CD34+ bone marrow progenitors and sorted major immune cell subsets from peripheral blood (B cells, T cells, monocytes, conventional dendritic cells, plasmacytoid dendritic cells and other HLA-DR+ myeloid cells). Ten CD34⁺ samples were profiled using the TAPESTRI platform for single-cell DNA and protein analysis and/or PromethION Nanopore long-read sequencing.In parallel, we used a conditional BRAFV600Ehet-SclCre+ murine model induced by tamoxifen (200 mg/kg/day for 4 days) and treated with trametinib (0.3 mg/kg/day) for one month. Disease burden and residual mutated cells were assessed by organ weight, flow cytometry, immunohistochemistry, and molecular quantification. Results BRAFV600E was detected in CD34⁺ bone marrow progenitors at diagnosis (VAFdiag = 0.03% ±0.02, n=37) and persisted under MEKi (VAFMEKi = 0.04% ±0.04, n=8; p=0.621). Single-cell TAPESTRI profiling of BMMCs from 7 patients identified 9 distinct clusters based on protein expression, with BRAFV600E⁺ cells predominantly located in the monocyte cluster in untreated samples. Under MEKi, we observed a notable shift, with BRAFV600E mutation emerging in two B cell clusters (CD10+ and CD1c+ B cells),suggesting lineage plasticity or therapy-induced reprogramming. Interestingly, the TET2 mutation previously identified in one patient was also detected in monocytes, similar to BRAFV600E, but no clone harboring both mutations was observed. Long-read sequencing is ongoing to investigate this B cell shift.In the BRAFV600E-driven murine model, trametinib significantly reduced spleen weight (Untreated: 0.67 ±0.28 g vs. trametinib: 0.25 ±0.09 g; p=0.002) and liver weight (Untreated: 1.80 ±0.67 g vs. trametinib: 1.15 ±0.19 g; p=0.04), reflecting clinical benefit. However, we did not observe significant reduction in YFP+mutated cells in the bone marrow (% YFP+ among live cells: untreated: 10.13% ±5.68 vs. trametinib: 11.11% ±18.94; p=0.88). Immunohistochemistry also did not reveal significant differences in the number of mutated cells in the brain (Untreated: 143.60 ±105.10 vs trametinib: 129.60 ±41.10; p=0.48) and liver (Untreated: 542.30 ±76.50 vs 500.60 ±216.50; p=0.17) under MEKi, but did in the spleen (Untreated: 298 ±76.15 vs. trametinib: 194 ±51.80; p=0.01) suggesting incomplete elimination of BRAFV600E+ cells in key disease sites. ConclusionMEKi induces clinical responses but fails to eradicate BRAFV600E-mutated clones in histiocytosis. Residual disease persists in hematopoietic progenitors and displays a therapy-induced shift from monocytes to B cells, suggesting lineage plasticity and identifying B cells as a potential reservoir. Our BRAFV600E murine model faithfully recapitulates clonal persistence under MEKi, underscoring the need for combinatorial strategies to achieve complete molecular remission.