Flow cytometry analysis of circulating mononuclear phagocytes in LCH patients. (A) Overview of pediatric LCH cohorts and experimental design. (B) Box plots (i) showing the distribution of plasma cfBRAF^V600E load between SS low-risk, MS low-risk, and MS high-risk patients. Wilcoxon rank sum test was used to test the significance of difference, and P values are indicated for the comparison. P < .05 is considered statistically significant. Dashed line represents the boundary of the first quarter (≥3.1%) of 63 positive samples defined as high mutation group. The median cfBRAF^V600E level was 0 for both SS low-risk and MS low-risk patients and 0.45% for MS high-risk patients, whereas the mean cfBRAF^V600E level was 0.11%, 0.71%, and 3.81% for SS low-risk, MS low-risk, and MS high-risk groups, respectively. Table (ii) showing the number of samples in different groups. Fisher's exact text was conducted and P < .05, suggesting the correlation between 2 groupings. (C) Gating strategy for distinguishing 3 cell subsets within the Lin⁻HLA⁻DR⁺ quadrant: CD14⁺ monocyte (pink box), CD14⁻CD11C⁺BDCA2⁻ subset containing cDC and CD16⁺⁺ monocyte (yellow box), and CD14⁻BDCA2⁺CD11c⁻ pDC (blue box). (D-E) Box plots showing the proportions of CD14⁺ monocytes, CD11c⁺BDCA2^–CD14^– cells, and BDCA2⁺CD11c^–CD14^– plasmacytoid dendritic cells (pDCs) in the Lin^–HLA-DR⁺ population from different groups. (D) Normal control (Control, n = 9), SS low-risk (n = 24), MS low-risk (n = 14) and MS high-risk patients (n = 11). (E) Normal control (Control, n = 9), cfBRAF^V600E-negative, lesion BRAF^V600E-positive LCH patients (cfBRAF^V600E−, n = 7), low-mutation (n = 10), and high-mutation (n = 6). Kruskal-Wallis test was used for significance verification. Nonparametric Wilcoxon test P values are shown on the top of box plots. ns, P = 1; *P < .05; **P < .01; ***P < .001; ****P = .0001.