Background/Purpose: Translocator protein (TSPO) is a mitochondrial cholesterol transporter, utilised in diseases as a Positron Emission Tomography (PET) imaging marker of macrophage infiltration. Recent observations support its utility as a PET marker in rheumatoid arthritis (RA), predictive of persistent synovitis and response to treatment. Here, we use synovial tissue taken from areas of known increased TSPO PET tracer uptake in vivo, to profile the expression of TSPO within cellular components found within synovium. We further investigated the potential of TSPO as a peripheral blood biomarker of disease activity in RA. 

Methods: 10 patients with RA and clinical evidence of knee synovitis were included. Each patient underwent PET-CT of knees with PBR28, a PET tracer targeted to TSPO followed by synovial biopsy within 7 days. Formalin fixed paraffin embedded tissue sections were generated, and stained for macrophage markers CD68, CD163 and the fibroblast activation marker gp38. Peripheral blood mononuclear cells were collected just prior to PET, and lysed for RNA and protein. From surgically obtained synovium in patients with active RA, Fibroblast-like synoviocytes (FLS), monocytes and lymphocytes, were isolated and lysed for RNA and protein analysis. Macrophages were generated by stimulation of synovial monocytes with M-CSF for 7 days. FLS and monocyte-derived macrophages were further stimulated with TNF alpha before being harvested for RNA and protein analysis. Cells from synovium, as well as monocyte derived macrophages, were also incubated with PBR28 tracer to assess tracer uptake in each cell type.

Results: A positive correlation between PBR28 tracer uptake, clinical synovitis, and semi-quantitative staining scores for TSPO, CD68, CD163, and gp38 was observed (see table). At both RNA and protein level, TSPO was expressed most highly in monocyte derived macrophages, and FLS treated with TNF for 24 hours, being negligible in synovial derived lymphocytes, monocytes and unstimulated FLS; findings that were mirrored by PBR28 uptake radioassay of these cell groups. There was an average 3 fold increase in TSPO expression at mRNA level in peripheral blood mononuclear cells of treatment naïve arthritis patients with clinically active disease, independent of CRP, compared with healthy controls, with TSPO expression correlating with tender and swollen joint count.  

Conclusion: Our work demonstrates that TSPO imaging is likely to reflect FLS activation in the synovium as much as it is macrophage infiltration.  For treatment-naïve patients, there appears to be a statistically significant, higher expression of TSPO in peripheral blood compared to healthy controls, at both RNA and protein level. Further studies are needed to profile TSPO cell expression in inflammation and ascertain whether TSPO may be a sensitive peripheral blood marker in patients with rheumatoid arthritis.