Introduction: Immune thrombocytopenia (ITP) is a bleeding disorder caused by an autoimmune response against platelets. In the majority of cases, ITP is thought to be caused by the presence of autoreactive B cells that produce anti-platelet autoantibodies and target platelets for destruction by phagocytic cells. However, in about 40% of ITP patients platelet autoantibodies cannot be detected and there is some evidence that cytotoxic cells might also be responsible for platelet death. Indeed, many patients repeatedly fail to respond to current immunosuppressive therapies that target B cells and their autoantibodies. As a consequence, these patients retain very low platelet counts with increased bleeding diathesis. In this study we have immunophenotyped a group of adult chronic ITP patients that have not responded to traditional immunosuppressive therapies and we identified 2 subgroups of patients with either an increase or decrease in the frequency of CD8+ T effector memory CD45RA+ cells (CD8TEMRA) compared to healthy controls.Methods: PBMCs were isolated from blood samples of 14 ITP patients with platelet counts <100x109/L and 14 matched healthy controls. The cells were phenotyped using a variety of antibodies including: CD3, CD4, CD8, CD45RA, CCR7, CD127, CD25, CD14, CD16 and CD19. In addition, at least 5x106 PBMCs were stimulated with PMA (50ng/ml) and ionomycin (1µg/ml) for 5 hours at 37°C, 5% CO2 and stained with antibodies against CD3 and CD8, then fixed and permeabilised before staining with antibodies specific to Granzyme B and Interferon-γ.Results and discussion: In our cohort of ITP patients we were able to identify two subgroups of patients based on their frequency of CD8TEMRA cells, identified as CD45RA+ CCR7- cells, gated on CD3+ CD8+ cells. Compared to healthy controls (mean=16.33%), 6/14 patients had significantly lower frequencies of CD8TEMRA cells (mean=11.31%) and 8/14 patients showed a significant increase (mean=31.50%). Interestingly, these two groups of patients also show significant differences between them in the frequency of CD19+ B cells (gated on CD3- cells), as the group with the lowest CD8TEMRA frequency showed a significant increase in B cells compared to the high CD8TEMRA group. Considering that CD8TEMRA cells are described as highly differentiated cytotoxic T cells, these results suggest that in patients with active ITP in which the CD8TEMRA population is more prevalent and the frequency of B cells is reduced, cytotoxic T cells might play an important role in platelet destruction. Although an increase in the frequency of CD8TEMRA with age has been described we did not find a correlation between these two variables in our cohort of patients. In the low CD8TEMRA group we also observed a significant increase in the frequency of T regulatory cells (Tregs) and monocytes when compared to healthy controls, whereas the trend in the high CD8TEMRA group was for frequencies closer to controls. In addition, when analysing the production of Granzyme B and Interferon-γ after a short in vitro stimulation, we found that the trend was for the CD8+ T cells in the high CD8TEMRA group to produce higher levels of both Granzyme B and Interferon-γ when compared to the patients in the low CD8TEMRA group. This would support the hypothesis that in patients with increased frequency of CD8TEMRA there has been an expansion of cells with cytotoxic properties. Further work will be required to confirm that in this cohort of patients there is a CD8+ T cell population that can specifically target and lyse platelets, thus contributing to ITP pathogenesis. DisclosuresProvan:UCB: Consultancy; GSK: Equity Ownership, Honoraria, Research Funding; Amgen: Honoraria, Research Funding; Medimmune: Consultancy.