Background: Seropositive RA is characterised by autoantibodies to citrullinated self-antigens (cit-peptides). CD4 T-cells are presumed key to the immunopathology given a strong HLA class II association and efficacy of co-stimulation blockade[1]. Measuring CD4 T-cells specific to disease relevant autoantigens will be vital to understand the immune dysregulation underpinning RA and in the assessment of novel, antigen specific therapies that aim to restore immune tolerance[2].

Objectives: We sought to characterise cit-peptide specific CD4 T-cells from RA patients at diagnosis and during follow up; individuals with ACPA positive arthralgia (APA) and healthy controls (HC); aiming to identify features of ‘healthy’ and ‘pathologic’ autoreactivity that could function as biomarkers in early phase trials of tolerogenic therapies.

Methods: HLA DRB1 0401 positive RA and APA donors were recruited from the Newcastle Early Arthritis Clinic, HC samples were obtained from LRS cones. Antigen specific T-cells were analysed by staining cryopreserved PBMC with HLA class II tetramers containing epitopes from the candidate autoantigens citrullinated tenascin and citrullinated cartilage intermediate layer protein. To assess cell phenotype, samples were stained with surface antibodies and analysed by spectral cytometry.

Results: 8 HCs, 16 RA-diagnosis, 13 RA-follow up and 5 APA samples were analysed. Overall, the frequency of antigen specific T-cells did not differ between groups. Amongst the cit-peptide specific T-cell compartment, RA-diagnosis donors had a significantly lower proportion of naïve cells (CCR7+CD45RA+) compared to HCs (28% vs 44%, p=0.017). There was a non-specific trend towards increased antigen specific effector memory and TEMRA cells amongst the RA-diagnosis group. Amongst RA-follow up and APA donors, the proportion of naïve cells was similar to HCs (47% and 38%, respectively). In paired analysis (n=7), the proportion of antigen specific naïve cells increased during follow up (20% to 47%, p=0.047). These discrepancies were not a result of differences in global T-cell population, no significant differences were observed when all CD4 T-cells were analysed. Amongst the cit-peptide specific memory compartment (CD45RA-CD45RO+), surface phenotype was compared between groups. No significant differences were observed between major T-helper subsets (Th1, Th2, Th17 and Tfh). Interestingly, Tregs (CD25+CD127-) were more common amongst HCs compared to RA-diagnosis donors (20% vs 9%, p=0.025), the frequency was slightly higher in the follow up samples (12%) and similar to HCs for APA donors (21%). RA-diagnosis donors had a significantly higher proportion of CD27- antigen specific memory CD4 T-cells compared to HCs (23% vs 9%, p=0.05), with lower levels observed in the RA-follow up and APA donors (13% and 11%, respectively). Again, these differences were not simply a reflection of the global T-cell population.

Conclusion: CD4 T-cells specific for cit-peptides amongst RA-diagnosis donors were less likely to be naïve or regulatory and more likely to have down regulated CD27, in keeping with an antigen experienced effector phenotype. The RA-follow up and APA groups appeared more similar to HCs, though Tregs were still relatively low in the former group, suggesting immune dysregulation is not completely reversed by current treatments. These observations suggest persistent activation and a relative loss of antigen specific Tregs are features of pathologic autoreactivity and could be leveraged as biomarkers in trials of novel tolerogenic therapies.

REFERENCES: [1] McInnes, I. B. & Schett, G. The Pathogenesis of Rheumatoid Arthritis. N. Engl. J. Med. 365 , 2205–2219 (2011).

[2] Stanway, J. A. & Isaacs, J. D. Tolerance-inducing medicines in autoimmunity: rheumatology and beyond. Lancet Rheumatol. 2 , e565–e575 (2020).

Acknowledgements: NIL.

Disclosure of Interests: James Stanway GSK/EMINENT clinical fellowship, Amy E Anderson: None declared, Arthur Pratt EMINENT/GSK grant, John Isaacs GSK EMINENT funding.