Risk-factor Analysis And CTL Treatment Of EBV Infection After Allo-HSCT

It takes a long time for immune system to return normal after HSCT due to the use of conditioning regimens and immunosuppressive. Infection is the significant cause of non-relapsed mortality post HSCT. Epstein-Barr virus(EBV) is one of the most common viral infection. EBV presents with a latent infection, with the prevalence>90% among populations. Because of the loss of immune surveillance, there is a reactivation of EBV, which could cause viremia, pneumonia, encephalitis, and post-transplantation lymphoproliferative disorders(PTLD). PTLD is characterized by rapid development and high mortality, with the influence on long term survival. There is no effective treatment for EBV infection at present. As a novel therapeutic method, cellular immunotherapy helps to rebuild the anti-viral immunity, and presents with high efficacy and low toxicity, which has a good prospect of clinical use.Objective: Risk factors and prognosis of patients with EBV infection post HSCT were analyzed. The clinical efficacy and safety of adoptive transfer of EBV specific CTLs were also investigated in order to provide new thoughts of EBV prevention and treatment.Contents: This study was consist of three parts. 1. The characteristic of patients, such as age of recipients, conditioning regimens, type of transplantation, HLA match and acute graft versus host disease(a GVHD) occurrence was retrospectively studied to find out the risk factors and prognosis of EBV infection. 2. The G-CSF mobilized donor hematopoietic stem/progenitor cell products were used as cell source to manufacture virus specific T lymphocytes(G-CTLs). The phenotype and function of CTLs were tested and compared with the non-mobilized donor derived CTLs(N-CTLs).Methods: 1. Patients receiving HSCT from November 2011 to November 2014 were retrospectively studied. The cumulative incidence(CI) of EBV infection and survival rate were analyzed by Kaplan-Meier method, while risk factors were estimated by logistic regression model. 2. The distribution of HLA alleles were studied to find out those with high frequency. The EBV specific peptides were chosen according to the HLA distribution characteristics. The PBMCs were collected by density gradient centrifugation from G-CSF mobilized donors and then separated into two different parts. The adherent cells were differentiate into DCs with GM-CSF, IL-4 and TNF-α. On day 7, the matured DCs were loaded with EBV specific peptides and then co-cultured with non-adherent cells, and expanded in the presence of IL-2. The intracellular IFN-γ after peptide stimulation were detected by intracellular cytokine staining(ICS) method, while the amount of IFN-γ, TNF-α and Granzyme B in the supernatant were measured by Cytometric Beads Array(CBA). The results were compared with N-CTLs manufactured by the same way. 3. Patients with risk factors of EBV reactivation were selected and EBV specific G-CTLs were generated, then cryopreserved in advance. When they had EBV reactivation after HSCT, EBV-CTLs were adoptively transferred at an interval of one week for several times. The cell dose and frequency were determined by the total cell number of CTLs and clinical efficiency. The occurrence of fever and rash was carefully supervised within the first 24 h, and a GVHD were supervised in the first 3 months. The level of EBV-DNA in the peripheral blood was detected by PCR, and the response rate were calculated according to the change of virus load after infusion.Results:1. 402 patients who underwent HSCT were analyzed in this study. One-year CI of EBV-emia and PTLD were 42% and 1.45%, and the median time of occurrence were 50 and 65 days, respectively. 83.5% of patients developed EBV reactivation within the first 100 days after HSCT. ATG use(P<0.001,HR 9.92) and a GVHD grade Ⅲ to Ⅳ(P=0.0016,HR 2.42) were risk factors for EBV-emia. Patients with two risk factors had higher CI of EBV-emia compared with those without risk factors(87.5% vs 24.6%,P<0.001). Patients with EBV reactivation showed a worse 3-year OS as compared with those without EBV reactivation(58.5% vs 75.4%,P<0.001).2. A total of 89 EBV specific peptides were synthesized, mainly including HLA-Ⅰ and a part of HLA-Ⅱ epitope peptides. These peptides were principally from immune-dominant antigens of EBV such as LMP1, LMP2, EBNA1, EBNA3, BZLF1 and BRLF1. The peptide pools covered 90% of HLA epitopes among Chinese. After 7 days of culture, the adherent cells were presented with typical characteristic of DCs, with high expression of CD11c(97.8%±1.8%), CD80(68.35%±21.3%), CD86(90.4%±10.5%), and CD83(13.4%±12.3%). The expansion fold of G-CTLs were lower than N-CTLs(1.7 vs 3.1, p<0.01). CD3+ cells in G-CTLs(84.3%±8.8%) included CD4+ cells(35.3%±11%) and CD8+ cells(51.4%±12.6%). The percentage of CD3-CD56+ cells was 10.9%±6.7%, showing no significance as compared with the other group. After peptide stimulation, G-CTLs produced large amount of granzyme B(5807.8±2926.2pg/ml), IFN-γ(450.4±332.9pg/ml), and TNF-α(34±32.2pg/ml). ICS assay result showed the percentage of CD3+IFN-γ+(20.7%±10.5% vs. 28.2%±17.5%) and CD8+IFN-γ+(18.4%±10.6% vs. 22.5%±17%) of the two groups, and showed no statistical significance.3. A total of 8 patients with EBV reactivation post HSCT received CTLs, 7 of whom were infused with G-CTLs, and the other one were transferred with N-CTLs. The cell number of each dose was(2.3~35.1)×105/kg and the total cell number was(0.87~7.89)×106/kg. The infusion times ranged from 2 to 8. There were no immediate toxicity within the first 24 h. One patient developed severe intestinal a GVHD, which was considered to be associated with the withdrawal of immunosuppressive and the increase of cell dose. 6 of the 8 patients had significant reduction of virus load, while the other 2 patients were non responders. The response rate were 75%.Conclusion: In this study, the risk factors and prognosis of EBV infection post HSCT were first analyzed at our hospital. The results showed a high CI and poor prognosis, with risk factors including ATG use and a GVHD. Then EBV specific peptides covering most HLA epitope among Chinese population were synthesized. EBV specific CTLs were successfully generated from G-CSF mobilized donors. Though its expansion capacity were lower than N-CTLs, they had similar phenotype and function of cytokine secretion. Finally 8 patients with EBV reaction were treated with CTLs and presented with high response rate. Except for one patient with a GVHD, there was no other side effects. In conclusion, G-CTLs was a safe and efficient for the treatment of EBV reactivation after HSCT and was convenient for clinical application, yet the decision on exact dose of infusion should be investigated in further studies.