Background Although the mainstream of pathogenesis of systemic lupus erythematosus (SLE) is thought to be interactions between antigen-presenting cells like dendritic cells, helper T cells, B cells and cytokines, previous reports suggested CD8 T cells also involve in the pathogenesis of SLE^1,2. However the associations between subsets of CD8 T cells and clinical manifestations remains unclear.

Objectives We conducted standard immunophenotyping analysis with peripheral blood from SLE patients and focused on CD8 T cell subsets to elucidate the association with clinical phenotype and serological markers.

Methods Peripheral blood was obtained from inactive SLE patients and healthy subjects as controls and also from active SLE before and 3 months after treatment. CD8 T cell subsets were measured by flow cytometry with fresh whole blood samples.

Results Thirty-four active SLE patients and 38 inactive patients and 22 healthy controls (HCs) whose age and sex were matched with those in SLE patients were enrolled. Mean SLE disease activity index (SLEDAI) was 14.2 and 1.8 in active and inactive patients, respectively. Among CD8 T cell subsets, the proportion of HLA-DR+ cells was significantly higher in SLE patients than HCs and positively correlated with SLEDAI (p=0.016, rho =0.283), and was also higher in patients with nephritis than patients without nephritis (p=0.074), though it did not reach statistical significance. The proportion of naive CD8 T cells positively correlated with the titer of anti-dsDNA antibody (p=0.011, rho =0.30) and C1q immune complex levels (p=0.043, rho =0.25), and negatively correlated with serum complement levels (p=0.019, rho = -0.34). The proportion of central memory CD8 T cells (Tcm) negatively correlated with SLEDAI (p<0.001, rho = -0.43), the titer of dsDNA antibody (p<0.001, rho = -0.51) and C1q immune complex levels (p<0.001, rho = -0.44), and positively correlated with serum complement levels (p<0.001, rho =0.49), and was lower in the patients with nephritis (p=0.041), skin rash (p<0.001), and fever (p=0.002) than the patients without them. The proportion of effector memory CD8 T cells (Tem) was lower in SLE and negatively correlated with the titer of anti-dsDNA antibody (p=0.028, rho = -0.26). Sixteen active patients were treated with prednisolone (PSL) 50mg (mean) with concomitant immunosuppressant; eight were treated with cyclophosphamide. Mean SLEDAI decreased from 15.9 to 3.9 at 3 months after the treatments. The proportion of naïve CD8 T cells significantly decreased (p=0.003), and the proportions of HLA-DR+ cells and Tem increased at 3 months (p=0.011 and p=0.007, respectively). The proportion of Tem increased and negatively correlated with SLEDAI only with those who were treated with cyclophosphamide.

Conclusions Pathological state of SLE positively correlated with the proportion of naïve CD8 T cells and negatively correlated with the proportion of Tem and these subsets were affected by treatment with PSL and immunosuppressant particularly cyclophosphamide. These results indicate that CD8 T cells are involved in pathophysiology and could potentially become a biomarker or a treatment target in SLE patients.

References

1. Blanco P, et al. Arthritis Rheum 2005; 52: 201–11.

2. Kim JS, et al. Rheumatology (Oxford) 2012; 51: 1587–94.

References

Disclosure of Interest J. Kikuchi: None declared, M. Ushikubo: None declared, S. Saito: None declared, K. Yamaoka: None declared, K. Sugahara Employee of: Mitsubishi Tanabe Pharma Corporation, K. Chiba Employee of: Mitsubishi Tanabe Pharma Corporation, T. Takeuchi: None declared