Discussion
Liquid biopsy provides enhanced sensitivity for biomarker discovery and ease of repeated sampling throughout treatment in a much more convenient and noninvasive way13,14. Recent advances in mass spectrometry (MS)-based proteomics have greatly extended its reach in biomedical and clinical research, and mass spectrometry (MS) is now poised to characterize the plasma proteome in great depth15. Plasma proteomic analysis has proven to be a promising tool for identifying new and effective biomarkers that can be used for evaluating the prognosis and treatment response in patients with various cancers16.
Patients with RO+ LCH have a poorer prognosis than patients with RO-LCH17. Recent evidence from the myeloid dendritic cell (DC) model suggested that MS-RO+ LCH results from a driver mutation in a bone marrow (BM)-resident multipotent hematopoietic progenitor18. However, the exact mechanism is unknown. Herein, we focused on the differences in plasma proteome profiles between RO+ LCH patients and SS LCH patients, and our study is the first, to our knowledge, to identify a novel soluble form of CSF1R in the plasma of pediatric LCH patients.
The results of the ELISA validation showed that plasma sCSF1R levels were associated with disease extent, and high levels of sCSF1R correlated with more severe disease, as indicated by a younger age; involvement of the RO, skin, and lung; and the presence of the BRAF-V600E mutation in lesions or cfDNA. Moreover, a high sCSF1R level at diagnosis independently predicted inferior PFS and was sustained in patients who experienced disease progression after 6 weeks of treatment. Notably, dynamic monitoring of sCSF1R levels provided an early warning of relapse in patients receiving BRAF inhibitor treatment. Therefore, the sCSF1R level should be closely monitored, especially during treatment with targeted inhibitors of BRAF or other proteins in the MAPK pathway.
IF staining in this study showed that high CSF1R protein expression in LCH lesions was associated with RO involvement, which was in line with the plasma sCSF1R levels. We also noted that sCSF1R levels apparently decreased after dabrafenib administration initiation, suggesting that extracellular CSF1R in plasma was mainly secreted by cells that carried the BRAF -V600E driver mutation. In vitro drug sensitivity data showed that sCSF1R increased resistance to Ara-C in THP-1 cells expressing ectopic BRAF -V600E. However, the exact biological function of sCSF1R is unclear. The release of soluble forms may represent a mechanism for counter regulation of CSF1R. A recent study revealed that overexpression of sCSF1R significantly decreased the extent of microgliosis in both the dorsal and ventral horns, indicating that sCSF1R can reduce the activation of native CSF1R on microglia19. By acting as a decoy receptor, sCSF1R can bind to colony-stimulating Factor 1 (CSF1) and prevent it from interacting with the membrane-bound CSF1R on target cells, possibly modulating CSF1 signaling and affecting the recruitment, survival, and differentiation of cells. In contrast, our data showed that sCSF1R was closely linked to severe disease and relapse in LCH patients, suggesting that sCSF1R contributes to disease progression. Previous studies have suggested that several oxidative stress proteins, including thioredoxin and heat shock proteins, are released from stressed, transformed cells and act as “endogenous” danger signals by binding TLR4 in the extracellular microenvironment, which results in the activation of downstream pathways and the secretion of proinflammatory cytokines20,21. We thus speculated that sCSF1R may act as an endogenous danger signal by binding to danger signal sensors/receptors, which is independent of CSF1. However, further investigations are needed to explore the underlying mechanisms involved.
Several recent studies have shown the good short-term effectiveness and tolerability of BRAF or MEK inhibitors in recurrent/refractory LCH22-25. However, MAPK pathway inhibition did not appear to be curative and was associated with a high risk of reactivation following drug discontinuation 26-28. Pharmacological inhibition of CSF1R has emerged as a promising antitumor strategy, and several small-molecule CSF1R inhibitors have been developed in clinical and preclinical studies29. Thus, combined inhibition of BRAF/MAPK signaling and CSF1R may be beneficial for recurrent/refractory LCH treatment. However, most clinical results on the safety and efficacy of CSF1R inhibition are limited 30, which may be caused, in part, by a high level of plasma sCSF1R, as shown in this study. Therefore, monoclonal antibodies engineered to recognize the extracellular domains of CSF1R may inspire new ideas for recurrent/refractory LCH treatment.
One limitation of this study was the lack of a validation cohort for the prognostic analyses of sCSF1R. Additional studies in independent cohorts are needed to establish the validity of our findings.
In conclusion, our findings identify plasma sCSF1R as a promising prognostic indicator for pediatric LCH. Accurate measurements of plasma sCSF1R at diagnosis and during follow-up have potential academic and clinical importance, and exploring such effective biomarkers to facilitate risk stratification and precision medicine is the major challenge for LCH treatment.