Background/Purpose: The cyclic citrullinated peptide (CCP) autoantibody is a highly specific and predictive marker for the clinical diagnosis of RA. Elevation in CCP titers can be detected several years before patients develop clinical arthritis and is also a predictive biomarker for RA risk. CCP+ RA is associated with increased levels of distinct synovial and circulating T cell immunophenotypes, including the expansion of PD-1^hi CXCR5- T peripheral helper (Tph) cells. However, whether these and other immune cell populations are altered in the peripheral blood of CCP+ individuals without RA is unknown.

Methods: We employed time of flight mass cytometry (CyTOF) with a panel of 39 surface markers to characterize T cells in cryopreserved peripheral blood mononuclear cells from a cohort of CCP- controls (n=23), CCP- first-degree relatives (FDRs) of patients with classified RA (n=27), CCP+ individuals at-risk of future RA (n=30), and active CCP+ early RA patients (diagnosis within 1 year, n=20). RA patients fulfilled the 2010 ACR/EULAR criteria. We used CITRUS to define and quantify the clusters of PD-1^hi cells based on their multi-dimensional features. Expanded cell populations were confirmed by biaxial gating, and the statistical significance was determined by Mann-Whitney U test.

Results: Global T cell phenotypes were generally similar in CCP- groups, CCP+ at risk, and CCP+ early RA, including frequencies of naïve and memory CD4 and CD8 T cells and Tregs. By biaxial gating, both Tph cells and Tfh cells (PD-1^hi CXCR5+) were increased in this CCP+ early RA cohort (Figure 1). The number of Tph cells in CCP+ at-risk individuals was increased compared to CCP- controls, although not statistically significant. For a broader assessment of PD-1^hi T cells, we used CITRUS to cluster gated PD-1^hi cells, including both CD4 and CD8 T cells. Among PD-1^hi T cells, 2 specific subpopulations were identified as significantly expanded in CCP+ early RA patients (Figure 2A). Phenotypic assessment highlighted increased CD39 expression on the expanded subclusters. Biaxial gating confirmed the significant expansion of PD-1^hi CD39^hi CD4 memory T cells in CCP+ early RA patients compared to all the other groups (Figure 2B). In addition, the CD39^hi subset of Tfh or Tph cells, but not total Tph and Tfh cells, were significantly higher in CCP+ RA compared to CCP+ at-risk group. The frequency of CD39^hi Tph cells was not increased in CCP+ at-risk individuals compared to CCP- controls; however, the expression of PD-1 on these cells was increased in CCP+ at-risk compared to controls (p=0.003, 1.25-fold change), a finding also seen in CCP+ early RA patients (Figure 3). The mean expression intensity of PD-1 on CD39^hi Tph cells was significantly correlated with CCP titer across all donors (r=0.376, p< 0.0001).

Conclusion: Detailed immunophenotyping analyses highlighted a robust expansion of CD39^hi Tph cells in CCP+ early RA patients. In CCP+ individuals at risk of future RA, CD39^hi Tph cells were not clearly increased in frequency, but expressed higher levels of PD-1 compared to CCP- controls and PD-1 levels were correlated with CCP titer. These results suggest that alterations in T cell phenotypes may be detectable not only early in RA but even preceding clinical onset after CCP elevation.

Figure 1. Tph cells and Tfh cells were expanded in CCP+ early RA cohort.

Figure 2. Expanded PD-1hi CD39hi cells in the blood of CCP+ early RA patients.

Figure 3. Increased PD-1 expression of CD39hi Tph cells in CCP+ at-risk individuals.

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/alterations-in-circulating-cd4-t-cell-phenotypes-in-ccp-early-ra-and-ccp-at-risk-individuals-by-mass-cytometry/