LAUSR

In the inflammatory myeloid neoplasia called Langerhans cell histiocytosis (LCH), leukocytes accumulate in tissues causing granulomatous lesions. The disease ranges from mild (singlesystem, SS) to severe (multisystem, MS) manifestations with potential involvement of risk organs (RO; liver, spleen or haematopoietic system) and worse prognosis. Previous studies identified disease-causing mutations in mitogen-activated protein kinases (MAPK) in many LCH patients, with BRAF being the most common. BRAF correlates with increased PD-L1 expression and regulatory T-cell (T-reg) accumulation in LCH lesions, thus linked to immunosuppressive microenvironment and reduced disease-free survival in LCH. Besides the reported hypercytokinaemia in LCH lesions, blood and CSF, the way immune cells affect the pathogenesis of LCH and response to therapy is not well understood. Low lymphocyte counts in blood of children with LCH were reported in a few previous studies, but the reasons and implications of this were not fully evaluated. In addition, age or treatment were not always taken into account when grouping the patients, while both factors affect lymphocyte counts. If patients with certain disease characteristics tend to be lymphopenic, lymphopenia may warrant special consideration. Since lymphocyte counts of children with LCH in our clinic were often below reference values, we evaluated this systematically and investigated possible implications for LCH pathogenesis. First, we studied the medical records of the 40 children with LCH treated in Karolinska Children’s hospital over a 15-year period (Supporting Information; Tables SI–SIII), which showed a significantly higher prevalence of lymphopenia at LCH diagnosis (i.e. treatment-na€ıve) than in the general paediatric population (Table SIV, Fig 1A). Children < 2 years old at diagnosis had higher prevalence of lymphopenia, both at diagnosis and at periods off treatment, i.e. at least four months (0 3–14 3 y) after the end of therapy for patients receiving any treatment (Fig 1B). All patients lymphopenic at diagnosis had at least one lymphopenic sample later off treatment (Fig 1C). Patients at diagnosis with MS-LCH had significantly higher prevalence of lymphopenia than SS-LCH patients (Fig 1C). Patients with RO MS-LCH at maximal extent of disease had significantly higher prevalence of lymphopenia at periods off treatment than RO MS-LCH patients. Overall, 15/40 patients were lymphopenic at least once while off treatment and/or at LCH diagnosis (Figs 1B-C). Lymphocyte counts at periods off treatment were frequently close to the lowest reference limit and dipped below on some occasions for several patients, as shown for two representative patients (Fig 1D). In non-treated LCH patients, lymphopenia could be the result of multiple factors, with one possible confounder being the redistribution/accumulation of lymphocytes in affected tissues, similar to other diseases. Lymphopenia should not be solely defined based on lymphocyte counts, as ongoing lymphocyte depletion could be obscured by compensating chronic homeostatic proliferation of high-affinity, self-reactive T cells. To evaluate if the absolute counts of specific lymphocyte subsets differed between LCH patients and controls, we analyzed 19 additional LCH patients (patients 1–6, 9, 11–22 described in) and seven paediatric controls (Supporting Information; Figs 1E–G) and found that 18% of the nontreated patients were lymphopenic, similar to the treatmentna€ıve retrospective cohort. As shown in Fig 2A, even nontreated LCH patients had significantly lower total lymphocyte and T-cell counts, compared to controls. In addition, although none of the patients with non-active disease (NAD) was lymphopenic (Fig 1E), the median value for total lymphocyte or T-cell counts was lower even in NAD patients (Fig 2B). Monocyte counts were significantly lower only in patients on treatment (Fig S1). Inflammation and lymphopenia may result in high Interleukin (IL)-7 production, recruitment of lymphocytes to inflamed tissue and increased IL-21 production by T-effector cells in an autocrine manner, further supporting their survival. In our cohort, significantly higher IL-7 and IL-21 plasma levels were observed compared to paediatric controls (Figs 2C-D), as well as significant positive correlation between IL-7 and IL-21 plasma levels (Fig 2E). To examine whether this was an LCH-specific finding or an inflammation/granuloma-related characteristic, we measured the plasma levels of IL-7 in 16 additional LCH patients (Table SV) and additional paediatric groups. This analysis revealed even larger difference between LCH patients and paediatric controls, but also between LCH patients and other paediatric patients with inflammation and/ or granulomas (Fig 2F; Supporting Information) indicating that the observed high cytokine production is not the result of inflammation or granulomas per se, but is rather due to a combination of factors linked to LCH pathogenesis. The high IL-7 levels in LCH may be important for the lesional T-reg expansion, despite the fact that activated conventional T cells are also present. IL-21 binding to its receptor (IL-21R) can activate the MAPK pathway, so the correspondence