CD8+ T Cell Profiles in Patients With Rheumatoid Arthritis—the Effects of Costimulation Blockade: Comment on the Article by Carvalheiro et al

To the Editor: We read with great interest the article by Carvalheiro and colleagues (1), in which they present data showing that CD81 T cells in patients with rheumatoid arthritis (RA) exhibit an effector phenotype, with expression of proinflammatory cytokines and cytolytic enzymes, including interferon-c (IFNc), interleukin-17 (IL-17), and granzyme B. These phenotypic and functional features of peripheral blood CD81 T cells were correlated with those observed in paired samples of synovial fluid and with the Disease Activity Score in 28 joints using the C-reactive protein level (DAS28-CRP), demonstrating the activated proinflammatory status of these lymphocytes in RA and suggesting a role for them in this disease. In their cross-sectional analysis, the authors did not identify any effect of concomitant antirheumatic drugs on the profiles of CD81 T cells. However, only 5 of 96 patients were treated with biologic drugs (all with tumor necrosis factor [TNF]–blocking agents), and this issue might be better addressed via a longitudinal followup study of CD81 T cell characteristics. We considered the T cell costimulatory blocker abatacept (CTLA-4Ig) the most appropriate candidate to modify the profile of T cells in RA (2), and report herein our observations on the effect of this treatment on CD81 T cells. The chemokine receptor CCR7 plays a crucial role with regard to homing in the T cell areas of secondary lymphoid organs. When CCR7 is down-modulated, as a result of repeated T cell stimulations, CCR7– T cells can be recruited in sites of inflammation where they can display their effector activities (3). Carvalheiro et al described a lower frequency of the CCR71CD81 T cells, with a phenotype characteristic of central memory T cells, in patients with RA as compared with healthy controls (1). This abnormality can be modified by treatment with abatacept: in 2010, we reported that in 20 patients with RA, the proportion of central memory (CCR71 CD45RA–) circulating CD81 T cells increased from a median of 11.5% of the total circulating CD81 T cells to 22.3% after 1 year of treatment (P 5 0.002) (4). The variation of CCR7 expression was correlated with that of DAS28-CRP (4). It was unclear, however, whether the results observed were related to a direct effect of abatacept or an indirect effect due to the reduction of disease activity. To evaluate this issue we studied the modification of the T cell phenotype after in vitro stimulation in 5 healthy controls and 10 RA patients before and after 12 months of abatacept therapy (9 of the RA patients were women, the median age was 58 years, 80% were rheumatoid factor positive, 80% were anti–citrullinated protein antibody positive, 7 had received previous treatment with TNF-blocking agents, and the median methotrexate dosage was 15 mg/week). Peripheral blood mononuclear cells (PBMCs) obtained after FicollHypaque gradient centrifugation were cryopreserved in liquid nitrogenous compounds. Paired samples obtained from the same patient before the start of abatacept therapy and after 12 months of abatacept therapy were thawed and cultured in the same experiment. PBMCs were left unstimulated or were stimulated for 5 days with CD3 (20 IU/ml) plus CD28 (1 lg/ml) (Beckman Coulter) at 378C temperature and 5% CO2 pressure, before membrane staining and flow cytometry analysis as previously described (4). As shown in Figure 1, after stimulation, a greater number of cells with the central memory phenotype (CCR71CD45RA–) were observed among CD81 and CD41 T cells in patients treated with abatacept for 12 months as compared with the same patients before the start of therapy and as compared with healthy controls. These results indicate that down-modulation of CCR7 expression after in vitro stimulation is reduced in T cells exposed in vivo to abatacept and suggest that this may be a direct effect of the drug. This observation helps to clarify the mechanism of action of abatacept on T cells, suggesting that in vivo it may act in the secondary lymphoid organ preventing down-modulation of CCR7 and cell efflux in the synovium. Analogous data were found in a study of experimental animals in which abatacept reduced the down-regulation of CD62L, another event needed to exit lymphoid organs (5). Accordingly, results from a study of an RA synovium/SCID mouse model suggested that abatacept acts by preventing T cell activation at a systemic level and does not act directly on the synovium (6). Moreover, our previous studies demonstrated that blocking the CD28 costimulatory pathway can prevent differentiation into the CD81 T cells effector phenotype: in our evaluation of 24 patients with RA we showed that the percentage of CD81 T cells producing IL-17 and IFNc significantly decreased