br Subjects and methods br Results br Discussion

Subjects and methods
Results
Discussion In the present study, we found significantly increased CSF levels of IL-4, IL-5, IL-9, IL-10, TNF-α, CXCL8/IL-8, and VEGF-A together with significantly higher detection rates of IL-6, GM-CSF, PDGF-BB and IL-1ra in IHP patients compared with OND patients. Among these, IL-4, IL-5, IL-6, IL-9 and IL-10 belong to the Th2 cytokine group (Wynn, 2015; Oliphant et al., 2011), suggesting Th2 cell involvement in IHP. The dual cluster analyses disclosed two major clusters: one mainly consisted of IHP patients and the other of OND patients. The cluster analysis of correlations between each CSF cytokine level also detected a large cluster of upregulated cytokines/chemokines comprised mainly of Th2 cytokines, including IL-9, G-CSF, IL-10, IL-13, IL-5, CCL4/MIP-1β, bFGF, VEGF-A, TNF-α, CCL11/eotaxin, and CCL5/RANTES in IHP but not OND patients, supporting the contribution of Th2 cells to dural inflammatory fibrosis observed in most IHP patients. The results of the dual clustering analysis, the PCA, and the k-means clustering analysis consistently showed that all but one IHP patient and one IgG4-related HP patient belonged to the same cluster with the upregulation of Th2 cytokines. The results of CDA also suggested that the levels of IL-4 and IL-5, representative Th2 cytokines, were crucial discriminating factors between the HP and OND patient groups. Although mechanism of IgG4-related disease remains to be established, Th2 cytokines drive plasma cell differentiation and proliferation with class switching to IgG4 (Jeannin et al., 1998; Jelinek, 2000; Lu et al., 2014; De Virgilio et al., 2017). Thus, Th2 cells likely play a key role in this condition. In our animal model of IgG4-related HP, which is driven by overactive Th2 cells (Cui et al., 2018), the massive infiltration of B220-positive B cells, IgG1 (human equivalent of IgG4)-positive cells, and CD138-positive plasma cells in the dura mater was followed by marked fibrotic thickening. Therefore, Th2 cytokines are likely to be involved in plasma cell infiltration and following dural fibrosis of IgG4-related HP. Plasma cell infiltration is also frequently observed in the dura of IHP (Kupersmith et al., 2004; Lu et al., 2014). Plasma cells and Scriptaid promote fibrosis through autoantibodies that activate fibroblasts and cell-cell contact with fibroblasts that induce contact-dependent fibroblast collagen production (Sakkas and Bogdanos, 2016). Accordingly, Th2 cytokines may also be contributory in not all but most patients with IHP, which could be a Th2-driven fibrosis, like IgG4-related HP. Among the elevated cytokines/chemokines and growth factors in IHP patients, IL-4 has been shown to have chemoattractant effects and induce the proliferation of fibroblasts, leading to tissue fibrosis (Postlethwaite and Seyer, 1991; Jakubzick et al., 2003; Brown and O'Reilly, 2018). IL-10 (Sziksz et al., 2015), TNF-α (Connolly et al., 2009), CXCL8/IL-8 (Yang et al., 2018), VEGF-A (Barratt et al., 2018), and PDGF-BB (Antoniades et al., 1990) can also promote fibroblast proliferation, thereby contributing to dural fibrosis. Simultaneously upregulated growth factors are likely to cooperatively promote inflammatory fibrosis of the dura in IHP. In our study, one ANCA-related HP patient and one IHP patient did not belong to the above-mentioned Th2 cytokine-related HP group. Because fibrosis itself is induced by heterogeneous mechanisms including Th1-, Th2-, and even Th17-mediated fibrosis (Wynn, 2004; Yoshizaki et al., 2010; Okamoto et al., 2012; Sciurba et al., 2018), IHP might also be heterogeneous. Yokoseki et al., reported an increase of CXCL10/IP-10 levels in the CSF of MPO-ANCA-related HP (Yokoseki et al., 2014). Although we could not confirm this because of the small sample size in this study, ANCA-related HP may be driven by a Th1-related mechanism, which might also have a role in the pathogenesis of IHP. TGF-β1 and CCL2/ MCP-1 were significantly decreased in IHP patients. IL-10, which was significantly upregulated in our patients, can suppress TGF-β1 production (Nakagome et al., 2006). It might explain the decrease of TGF-β1 in our patients. TGF-β functions as a potent immune suppressor by inhibiting the proliferation, differentiation, activation, and effector functions of immune cells (Mantel and Schmidt-Weber, 2011; Wynn and Ramalingam, 2012). Thus, a decrease in the anti-inflammatory effects of TGF-β might eventually potentiate dural inflammation in IHP. Alternatively, because TGF-β1 is a potent profibrotic cytokine (Fernandez and Eickelberg, 2012; Cui et al., 2018), TGF-β1 might be consumed by highly proliferating fibroblasts in the HP dura. The profibrotic role of CCL2/ MCP-1 is controversial (Kalderén et al., 2014). Indeed, the anti-fibrotic role of CCL2/MCP-1 was previously reported (Mitchell et al., 2009), consistent with the downregulated level of this cytokine in the present study. Further studies are required to clarify the roles of these downregulated cytokines in HP.