THU0008 GM-CSF PATHWAY SIGNATURE IDENTIFIED IN TEMPORAL ARTERY BIOPSIES OF PATIENTS WITH GIANT CELL ARTERITIS

Abstract

Background Giant Cell Arteritis (GCA) is a type of large vessel vasculitis that can cause blindness and aortic aneurysms. A significant unmet medical need remains in GCA, as current treatment options are limited, and relapse increases corticosteroid (CS) exposure and toxicity. The primary role of macrophages/dendritic cells (DCs) and TH1/TH17 lymphocytes in GCA pathogenesis has been highlighted previously. Granulocyte-macrophage colony stimulating factor (GM-CSF) may contribute to GCA pathogenesis by stimulating giant cell formation.1 GM-CSF produced by CD4+ T helper TH1 and TH17 cells can stimulate conventional DCs and promote differentiation of monocyte-derived DCs.2 GM-CSF may drive DCs to program naive CD4+ cells to TH1, TH17, and T follicular helper phenotypes (IFNγ/IL-17/IL-21). Notably GM-CSF RNA has been reported in GCA lesions3 and in peripheral blood mononuclear cells of symptomatic patients.4 Objectives We hypothesized elevation of the GM-CSF pathway signature in GCA vessels versus controls. Methods Two independent sources of temporal artery biopsies were utilized. First, GCA (n=17) and control (symptomatic patients suspected for GCA, but with a normal temporary artery biopsy; n=5) biopsies were analyzed for 15 mRNA transcripts representing TH1, TH17, and GM-CSF signaling (RNAscope; RS) and for mRNA transcripts representing the autoimmune panel (Nanostring; NS). Semi-quantitative scoring was performed on RS images, and fold-change of representative TH1, TH17 and GM-CSF related mRNA transcripts were calculated via NS nCounter analysis. Additional GCA and control biopsies were obtained and analyzed by RT-PCR for a subset of transcripts (n=10 each) and by confocal microscopy for GM-CSF and GM-CSF-Rα protein (n=2 each). Results The GM-CSF signaling pathway molecular signature was confirmed to be upregulated by 4 independent analyses. GM-CSF-associated and TH1-associated genes were upregulated in GCA biopsies versus control (GMCSF: 3-4x RS; GM-CSF-Rα: 6.7x NS, 6x RS; and CD83: 3.9x NS, 6x RS; TNFα: 2x NS, 3x RS; IFNγ: 2x RS; IL-1β: 6x RS). TH17 associated genes were not elevated, potentially due to concomitant CS treatment. Upregulation of both GM-CSF (12x) and GM-CSF-Rα (3x) mRNA was confirmed in a separate cohort of biopsies from GCA patients vs. controls by RT-PCR (Figure). GM-CSF and GM-CSFRα proteins were detected in the luminal endothelium, neovessels and inflammatory cells of GCA patients. In normal temporal arteries, GM-CSF protein was not detected, and some GM-CSFRα expression was observed in the luminal endothelium. Pu.1, a transcription factor downstream of GM-CSF signaling, was increased 8x in GCA vs. controls (RS, NS) (Figure). Conclusion GM-CSF and TH1 pathway signatures were demonstrated in GCA patient temporal arteries by independent analytical techniques. Active GM-CSF signaling in diseased tissue is evidenced by increased expression of Pu.1 in the vessel wall. These data implicate the GM-CSF pathway in GCA pathophysiology and increase confidence in rationale for targeting GM-CSF in GCA.