FRI0249 Circulating follicular helper-like t cells in patients with rheumatoid arthritis treated with abatacept

Background Rheumatoid arthritis (RA) is characterized by synovial inflammation and, in most cases, by autoantibodies, including rheumatoid factor (RF) and anti-cyclic citrullinated peptide antigen (ACPA), whose presence is associated to a more severe disease. The capability of producing autoantibodies is acquired by B cells with the “help” of specialized T lymphocytes, known as Follicular helper T (TFH) cells, in the germinal centers (GC) of secondary lymphoid tissues. TFH express the inducible co-stimulator (ICOS), and are characterized by a high expression of CXC-chemokine receptor 5 (CXCR5), which mediate their migration into the GC, where its ligand CXCL13 is expressed. Here, TFH promote the GC development and B cell maturation. These cells are reported to exist in greater quantities in the peripheral blood (circulating TFH-like cells) of RA patients. Abatacept (ABA) is a fusion protein, which through its CTLA4 portion can bind to CD80 and CD86 on antigen presenting cells, thereby inhibiting CD28 costimulation. Data from animal models and phenotypic analysis of circulating T cells in RA suggest that ABA may act in the secondary lymphoid organ, and not directly on the synovium (1,2). While experimental models show that ABA can block the generation of TFH (3), little is known on the effect of ABA on circulating TFH-like cells of RA patients (4). Objectives To analyze the effect of the blockade of costimulation performed by ABA on the levels of circulating TFH-like cells of RA patients. Methods 25 RA patients (F/M=18/7; median age (25th-75th percentile): 57 (44–63) years; ACPA+: 16 (64%); RF+: 17 (68%)), treated for at least 6 consecutive months with ABA were evaluated. Circulating TFH-like cells were identified by flow-cytometry as CD4+ICOS+CXCR5+.The response to treatment was evaluated with the EULAR Criteria. Results After 6 months of therapy with ABA the percentage of circulating TFH-like cells among total CD4+ T-cells tended to decrease from a median of 1.8 (0.6–3.8) to 0.8 (0.1–2.3) (p:0.07; Wilcoxon signed rank test). The percentage of circulating TFH-like cells at baseline was higher among patient who did not respond to ABA (n.8; 2.6 (2.0–3.6)) than in those who achieved a moderate or good EULAR response (n:17; 1.0 (0.5–5.4)) (p Conclusions ABA tends to reduce the number of circulating TFH-like cells in RA patients, suggesting the relevance of costimulation via CD28 in their generation. A higher percentage of these cells is present in patients not responding to ABA. While this observation suggests their possible use as predictors of clinical response to ABA, further studies are deserved to evaluate whether it reflects different localizations of TFH, or functional variability within this T-cell subset in patients with RA. References Koenders M et al, Arthritis Rheum 2012. Airo9 P et al, Arthritis Rheumatol 2015. Platt AL et al, J Immunol 2010. Iwata S et al, Arthritis Rheumatol 2015. Acknowledgements Bristol-Myers-Squibb Italy provided an unrestricted research grant for the study conduct and did not interfere with the conception and design of the study, acquisition, analysis, interpretation of data, and manuscript drafting. Disclosure of Interest None declared