Abstract BACKGROUND Epstein-Barr virus (EBV) infection is one of the life-threatening clinical complications in patients who underwent haploidentical hematopoietic stem cell transplantation(haploHSCT). Some studies evaluated the reconstitution of EBV-specific CD8+ T lymphocytes (CTLs) in the cases of EBV reactivation, however, EBV reactivations usually occurred within the first 3 months after HSCT by that time EBV-CTLs were too less to detect in most recipients. It was reported that innate lymphocytes reconstituted faster than adaptive T cells after allogeneic HSCT. Nevertheless, the recovery characteristics of innate immune cells and their associations with EBV infection post-haploHSCT are still unknown to date. We recently compared the influences of different doses of antithymocyte globulin conditioning on the T-cell reconstitution post-haploHSCT and suggested that CD4-CD8-T cells might interact with the occurrence of EBV reactivation. The aim of this study was to confirm the association between the recovery of CD4/CD8 double negative T lymphocytes and occurrence of EBV reactivation in patients who underwent haploHSCT. MATERIALS AND METHODS Consecutive subjects who received haploHSCT at Peking University Institute of Hematology from January 2011 to December 2014 were screened. The EBV positive cohort which comprised 64 recipients testing positive for EBV DNA (>1000 copies/mL) in peripheral blood, and the control cohort tested negative EBV DNA load (< 500 copies/mL) were paired in a 3:1 ratio and age- and gender-matched to the EBV positive cohort. we determined the absolute numbers of 6 T-lymphocyte subpopulations in patients' peripheral blood with immunophenotyping, namely: CD3+, CD4+, CD8+, CD4+CD28+, CD8+CD28+, and CD4-CD8-T cells. The quantitative variables were analyzed with the Mann-Whitney U test, and the qualitative variables were analyzed with the Chi-squared test. Landmark study was performed to evaluate the correlation of double negative T cell recovery with the occurrence of EBV reactivation after haploHSCT. Cox regression multivariate analysis was conducted to evaluate if EBV reactivation and double negative T cell reconstitution remain independent variables to transplantation outcomes. Statistical significance was defined as P ≤ 0.05, based on a two-tailed test. All calculations were carried out with SPSS 16.0 statistical software (SPSS Inc, Chicago, IL, USA). RESULTS The median onset time of EBV reactivation was 45 (13-190) days in the current study. On day 30 after haploHSCT, the median levels of all detected T-cell subsets were significantly lower in the EBV+ group than those in the EBV- group (P <0.001). At 90 and 180 days post-transplantations, the median counts of CD3+, CD4+,CD8+, CD4+CD28+, and CD8+CD28+ T cells were similar between the two groups. In contrast, the absolute numbers of a special T-cell subpopulation defined as CD4/CD8 double negative T cells were significantly lower in the EBV+ group at these two time points, with P values < 0.01 compared to the EBV- group. The normal level of peripheral CD4-CD8- T cells, as detected in the healthy donors, was 57 (26-189) cells/μL (n=20). Thus, the median counts of CD4-CD8- T cells in recipients without EBV reactivation (58 cells/μL) nearly achieved the normal level whereas those in EBV+ group (31 cells/μL) were significantly decreased even at day 180 after haploHSCT. Landmark studies further confirmed that lower counts of CD4-CD8-T cells correlated to the increased risk of EBV infection. Multivariate analysis showed that hampered recovery of CD4-CD8-T cells and EBV reactivation were the independent risk factors to predict transplant-related mortality. DISCUSSION Herein, we found that EBV reactivation occurred in the recipients with hampered recovery of CD4-CD8- T cells after haploHSCT. Moreover, we showed that delayed recovery of CD4-CD8- T cells was associated with the poor outcomes of haploHSCT recipients. These results not only suggest a significant correlation of recovery of CD4/CD8 double negative T cells with EBV infection, but also are under way to better recognize the reconstitution characteristics of this special T-cell subpopulation related to the serious complications of HSCT. Finally, our findings may facilitate the intervention strategies to improve the overall survival of haploHSCT recipients. Disclosures No relevant conflicts of interest to declare.