Search

Background Cytomegalovirus (CMV) infection is an important infectious complication of allogeneic hematopoietic stem cell transplantation (allo-HSCT) due to the immune-suppressive state of the transplant recipient. Patients are likely to be resistant to traditional antiviral drugs, such as ganciclovir and foscarnet, or have intolerable adverse reactions to these drugs, including bone marrow suppression and nephrotoxicity. CMV-specific CD8 + and CD4 + T-cell functional defects are the main reasons for the emergence of refractory CMV infection. Thus, it is possible to cure refractory CMV infection using CMV-specific TCR-T cell therapy via a direct antiviral effect and indirect T cell mediated immune reconstruction. Objectives and Methods We conducted a single-arm, open-label, phase I clinical trial (https://clinicaltrials.gov, NCT04153279) at the Hebei Yanda Lu Daopei Hospital. This study was designed to assess the safety and feasibility of CMV-specific TCR-T cell therapy in allo-HSCT patients with refractory CMV viremia or viral disease. We screened multiple TCRs with specificity for CMV-PP65 or IE1 polypeptides restricted by HLA-A*11:01, 24:02 and 02:01, and constructed lentiviral vectors containing CMV-specific TCRs. Peripheral blood (PB) stem cells were obtained from healthy HSCT donors for the TCR-T cell preparation. After stimulation for 24 hours, T-cells were infected with lentivirus carrying HLA-matched CMV TCR, then cultured with medium containing IL-7/IL-15 for a total of 7-9 days before harvest. After infusion, CMV TCR-T cell proliferation and persistence in PB was detected by Q-PCR. Results From December 24, 2019 to January 1, 2021, 9 patients were enrolled and as a result 7 patients including 5 with refractory CMV viremia and 2 with CMV disease successfully received the TCR-T therapy. Among the 7 patients who received TCR-T, the median age was 11 years (5-45 years of age) and 5 were male (71.4%). The diagnosis included 5 cases of acute lymphoblastic leukemia (ALL), 1 case of acute myeloid leukemia (AML) and 1 case of mixed phenotype acute leukemia (MPAL) . Patients underwent either haplo-identical transplantation (n=6) or unrelated donor transplantation (n=1). The median time for patients to CMV viremia diagnosis was 28 days (21-231 days). Patients received a median dose of 3 × 10 5 TCR-T cells/kg body weight (1-10 × 10 5 TCR-T cells/kg body weight). The mean proportion of CMV-specific TCR-T in the cultured cell products was 35.5% (18.0-68.9%). Among the 7 patients, successful follow up and data evaluation was available for 6 patients. One patient (P3) withdrew early due to uncontrolled lung infection. The median follow-up time for the entire cohort after HSCT was 199 days (144-479 days). Only 1 patient experienced Grade 1/2 cytokine release syndrome (CRS) with mild hypotension and fever. No immune effector cell-associated neurotoxicity syndrome (ICANS) or TCR-T cell-related graft-versus-host disease (GVHD) occurred in any patient. A total of 6 patients achieved complete response (CMV DNA-negative plasma) after TCR-T cell infusion. The median time from CMV-specific TCR-T infusion to the first CMV clearance was 41 days (19-91 days). The median time for CMV TCR-T cells to reach their first peak was 21 days (10-28 days) with a median copy number of 3.85×10 4 copies/μg genomic DNA (range: 1.93×10 4-7.75×10 4 copies/μg DNA) (Figure 1). CMV TCR T-cells persisted up to 3 months with detectable copy number. The levels of cytokines IL-2, TNF-α, IL-6, IL-8 and IFN-γ in PB were relatively low. Conclusion This study demonstrates the safety and feasibility of CMV-specific TCR-T cells for refractory CMV infection after HSCT. We show that adoptive transfer of CMV-specific TCR-T cells can lead to complete CMV clearance in patients who have undergone hematopoietic stem cell transplantation (HSCT) and with good tolerability and safety. Longer-term observation of these patients and larger patient studies are warranted to demonstrate the efficacy and safety of CMV TCR-T cells. Given that immune defect-related CMV infections are common and life threatening for HSCT patients, we see a potential future for CMV-specific TCR-T cell therapy for the treatment and prevention of CMV infection following HSCT or other organ transplantation. Figure 1 Figure 1. Disclosures No relevant conflicts of interest to declare.

Introduction SIL-TAL1 rearrangement is relatively common among patients with T-cell acute lymphoblastic leukemia (T-ALL). Pediatric patients with SIL-TL1 positive T-ALL generally have a poor response to early intensive treatment and a lower median relapse-free survival (RFS) and overall survival (OS) rates compared to pediatric T-ALL patients who do not harbor the SIL-TAL1 gene fusion. Therefore, we explored the clinical effect of haploid transplantation among pediatric SIL-TAL1-gene fusion positive T-ALL and aimed to identify potential patient prognostic factors. M ethods We performed a retrospective analysis of 21 children with SIL-TAL1 gene fusion positive T-ALL who underwent haploidentical transplantation at the Lu Daopei Hospital in Langfang, China between July 2014 and June 2020. The median patient age was 9 years (range: 1.6 to 18 years), the male to female ratio was 19:2, and the median white blood cell count at initial onset was 193.65×109/L (range: 3.6-582.17×109/L). Prior to transplantation, 18 patients were bone marrow minimal residual disease (MRD) negative, three patients were bone marrow MRD-positive and the overall median value for MRD positivity by flow cytometry was 0.05% (range: 0.03%-0.35%). Eighteen patients were SIL-TAL1 gene fusion negative and three patients were SIL-TAL1 gene fusion positive prior to transplantation. The quantitative median number is 0.04% (range: 0.033%-0.05%). 13 patients with CR1 and 8 patients with CR2. 11 of 21 patients harbored the SIL-TAL1 gene fusion along with at least one other gene mutation. All patients were given myeloablative pretreatment, of which 17 patients received cyclophosphamide/total body irradiation (TBI/Cy) and 4 patients received Busulfan/cyclophosphamide (Bu/Cy). The median follow-up time was 9.5 months (range: 2 to 51 months). Results The white blood cells (WBCs) of 21 patients were all completely engrafted , two patients did not have platelet engraftment . At one month after transplantation, all patients were bone marrow MRD negative and the quantification of SIL-TAL1 fusion gene was 0% among all the patients. For 20 of these 21 patients, the chimerism rate of peripheral blood CD3+ cells and bone marrow was 100% donor type. The 4-year disease-free survival (DFS) and overall survival (OS) rate were 56.5% and 55.1%, respectively (Figure 1). Causes of death were recurrence (5 patients all of whom had bone marrow recurrence), infection (3 patients) and alveolar hemorrhage after reinfusion (1 patient). Six patients relapsed after transplantation with a median time to relapse of 90 days (range: 54-180 days). The OS rate following transplantation in the CR1 group was higher than in the CR2 groups (69.2% vs 31.3%, respectively) but there was no statistical difference between the two groups (P=0.147) (Figure 2). The OS rate of MRD-negative patients and MRD-positive patients prior to transplantation was 58.4% and 33.4% respectively, (P=0.287) (Figure 3). Prior to transplantation, the OS rate of SIL-TAL1 gene fusion-negative patients was 64.3%, while the OS rate of SIL-TAL1 gene fusion-positive patients was 0.0%--the difference between the two groups was statistically significant (P=0.001) (Figure 4). The OS rate of 11 patients that harbored the SIL-TAL1 gene fusion along with another gene mutation was 47.7% (Figure 5). Among these 11 patients, the OS rate of 9 patients who switched from SIL-TAL1 positive to SIL-TAL1 negative prior to transplantation was 58.3%. In contrast, two patients that remained SIL-TAL1 gene fusion positive had an OS rate of 0% (P=0.027) (Figure 6). Conclusion Pediatric T-ALL patients harboring the SIL-TAL1 gene fusion achieved a higher OS rate through haploid transplantation and were able to avoid extramedullary recurrence. We conclude that if patients with the SIL-TAL1 gene fusion do not switch to SIL-TAL1-negative prior to transplantation, we showed that even if the myeloablative pretreatment program is adopted, the survival rate cannot be improved, and the mortality rate is 100%. Harboring another gene mutation besides SIL-TAL1 at the initial onset of disease did not affect the prognosis in our cohort analysis. We conclude that for pediatric T-ALL patients, intensive chemotherapy should be adopted to make the bone marrow MRD negative and SIL-TAL1 gene fusion negative, and allogeneic hematopoietic stem cell transplantation should be performed as soon as possible to improve OS. Disclosures No relevant conflicts of interest to declare.

Introduction Current available treatments are limited once patients with B-ALL relapse following allogeneic hematopoietic stem cell transplantation (allo-HSCT). While chimeric antigen receptor (CAR) T-cell therapy offers a chance of remission, long-term outcomes for these patients remain poor. The benefit of bridging into a second transplant after CAR T-cell therapy remains inconclusive and available data are limited. Here, we report the long-term outcomes of 23 B-ALL patients who chose to undergo a second allo-HSCT after achieving complete remission (CR) from CAR T-cell therapy. Methods From April 2017 to April 2020, 23 R/R B-ALL patients (median age of 20 years, ranging from 3 to 58 years) who relapsed after first allo-HSCT received CAR T-cell therapy. The data were aggregated from seven different clinical trials (www.clinicaltrials.gov NCT03173417, NCT02546739 NCT03825718, NCT03825731, NCT03952923, NCT04100187 and www.chictr.org.cn ChiCTR1800016541). Patients' first transplant sources were HLA-identical sibling (n=5), matched-unrelated donor (MUD) (n=1), and haploidentical donors (haplo) (n=17). Eight of the 23 patients had disease relapse within 6 months following the first transplant. The median time from first transplant to CAR T-cell infusion was 261days (range: 117~2181 days). Before CAR T-cell infusion, patients' median bone marrow (BM) blasts by morphology were about 72.5% (1.5%-94.5%) including 12 patients with BM blasts >70% (5 with BM blasts >90%). Three of the 23 patients (13%) had received at least one prior donor lymphocyte infusion. No patients had active graft-versus-host disease (GVHD) prior to CAR T-cell therapy. Second generation CAR T-cells were generated by using purified T-cells from transplant donors (n=15) or patients (n=8). Twenty-two patients received T-cells modified with CD19-targeting CAR T-cells containing either a 4-1BB (n=18) or a CD28 co-stimulatory domain (n=4), and one patient received CD19-CD22 dual specificity CAR T-cells. All patients received a conditioning regimen of IV fludarabine (30mg/m2/d) and cyclophosphamide (250mg/m2/d) for 3 days followed by a single CAR T-cell infusion with a median dose of 3×105 cells/kg (1×105-6×105 cells/kg) in 21 patients. Two patients received a second CAR T-cell infusion in 2-3 months (1/3×105 cells/kg dose). Post CAR-T therapy, all patients bridged into a consolidation second transplantation with conventional myeloablative pre-transplantation conditioning regimens including 15 patients who received total body irradiation-based and 7 patients that received a busulfan-based conditioning regimen. Cyclosporin A, short-term methotrexate, and mycophenolate mofetil were used for GVHD prophylaxis. Results Patients' characteristics are shown in Table 1. On Day 30 post CAR-T-cell infusion, 23/23 (100%) patients achieved minimal residual disease (MRD)-negative CR. A total of 16/23 (69.6%) patients developed cytokine release syndrome (CRS) of which 14/23 (60.9%) had Grade I-II and 2/23 (8.7%) had Grade III CRS. Two patients had Grade III neurotoxicity. All patients with MRD-negative status subsequently bridged into a second transplant (2 from MUD and 21 from haplo donors) with a median interval time of 67 days (39- 329 days) from CAR T-cell therapy to a second transplant. At a median follow-up time of 258 days (84-978 days), no patients relapsed, which was encouraging. Five of 23 patients (21.7%) died from transplant-related mortality (TRM) at a median time of 295 days (103-372 days) (1 from GVHD and 4 from infection). The 1-year overall survival (OS) was 68.0% and 2-year OS was 54.4% (Fig.1). While there was a trend towards a more efficacious OS for patients whose CAR T-cells were derived from donors rather than from patients themselves but the number are too small to reach statistical significance (1-year OS 83.9% vs. 64.3%, 2-year OS 83.9% vs. 42.9%, P=0.739. Fig.2). After the 2nd transplant, four patients developed GVHD. Conclusions Our study demonstrates that even for R/R B-ALL patients who have relapsed following a first allo-HSCT , an MRD-negative CR status can still be achieved through CAR T-cell cell therapy without increasing CRS or neurotoxicity, making consolidation second allo-HSCT feasible for these patients. CAR T-cell therapy combined with a consolidation second HSCT are effective for these heavily pre-treated patients with an encouraging prospect for long-term survival. Disclosures No relevant conflicts of interest to declare.

Background Post-transplant lymphoproliferative disorder (PTLD) is a serious complication that can occur following an allogenic hematopoietic stem cell transplant (allo-HSCT). PTLD occurs in approximately 0.8% to 20% of patients following an allo-HSCT and is associated with Epstein-Barr virus (EBV) infection in about 60% to 80% of patients. EBV positive (EBV+) PTLD generally arises early, several months following transplantation. The routine methods to diagnose EBV+ PTLD are clinical symptoms, EBV copy number, imageological examination, and pathological diagnosis, of which pathological diagnosis is the gold standard. Yet obtaining a biopsy is not possible with some patients and is not applicable for post-therapy monitoring. Therefore, there is an urgent need to find a simple, highly efficient, feasible, sensitive, and specific diagnostic and monitoring tool. Flow cytometry (FCM) has been established as a highly cost-effective method to diagnose lymphoma over the past several decades, and particularly for screening monoclonal B and/or plasma cells (MC B/P). Several recent publications as well as our own published and unpublished data have found that MC B/P are present in the peripheral blood of most PTLD patients. To this end, FCM detection of MC B/P in PB is a promising screening and monitoring method for EBV(+) PTLD. Objective To investigate the effectiveness of detecting MC B/P in PB by FCM for EBV+ PTLD screening and monitoring. Methods A total of 1470 patients received allo-HSCT at the Hebei Yanda Ludaopei Hospital, China from January 2018 through December 2019. We conducted a retrospective study of 481 patients with fever and lymphadenopathy following allo-HSCT. Patient PB was extracted for FCM MC B/P analysis. Plasma EBV viral loads were detected by PCR. The median fellow-up time was 6 months (range: 2 days to 21 months). The relationships of PB MC B/P, EBV load and clinical EBV-associated PTLD symptoms were investigated. Results: MC B/Ps were detected in the PB of 47 patients, of which 29 cases had monoclonal B cells, 14 cases had detectable co-existence of monoclonal B cells and monoclonal plasma cells, and 4 cases had detectable monoclonal plasma cells. The median time to PTLD onset following allo-HSCT was 70 days (range: 33 days to 491 days). The median percentage of monoclonal B cells was 0.43% (range: 0.1% to 23.41%. The median percentage of monoclonal plasma cells was 0.25%). Forty of 47 patients (85.1%) were finally diagnosed with PTLD using clinical comprehensive examination of symptoms. The incidence of clinical EBV+ PTLD was 2.9% (44 of 1470 cases). By FCM, MC B/Ps were observed in the PB of 91% of the patients (40 of 44 cases). No MC B/Ps were found in the PB of 4 patients. MC B/P detection in PB by FCM screening could effectively predict the incidence of EBV+ PTLD with a sensitivity of 90.91% and specificity of 98.40%. The positive predictive value was 85.17% and the negative predictive value was 99.08%. There was no significant correlation between EBV viral copy number and percentage of MC B/P. The median fellow-up time was 6 months (range: 2 days to 21 months) for all patients. The therapeutic response rate was 87.5% (35 of 40 patients) for MC B/P positive EBV+ PTLD patients. In those patients with a therapeutic response, clinical symptoms improved, lymph nodes significantly shrank, EBV viral loads decreased, MC B/Ps decreased or vanished, and CD20+ cell proportion decreased or vanished. Among the 7 patients who had MC B/Ps in PB but were not diagnosed with PTLD , their clinical symptoms recovered after treatment with immunosuppressive therapy. Conclusion By detecting PB MC B/Ps using FCM, we can successfully screen for EBV+ PTLD and monitor patient therapeutic responses in most cases. FCM to detect MC B/Ps in PB is a new, promising method for the diagnosis and monitoring of EBV+ PTLD. Compared to a biopsy, this appears to be a more simple, easier and more applicable tool to diagnose and monitor EBV+ PTLD. Disclosures No relevant conflicts of interest to declare.

Background: The outcomes of alternative donor haematopoietic stem cell Transplantation(HSCT)for Inherited bone marrow failure syndromes(IBMFS) have not been well described,especially the conditioning regimens are major challenges ,each disease type has different characteristics, whether engraftment can be achieved with less toxicity. Method : With this background,we retrospectively analyzed alternative donor HSCT for 42 patients with IBMFS in our single center from November 2012 to August 2018. Results: 27 cases were Fanconia anemia(FA),7 cases were dyskeratosis congenital(DC), 8 cases were severe congenital neutropenias(SCN). The median age at diagnosis and transplantation were 4(1 to 25 ) years and 10(1.9 to 26)years respectively. Male to female was 28 :14. All patients were confirmed to have BMF and disease-specific pathogenesis-related gene mutations. 16 cases had disease specific congenital anomalies, 10 patients had family history. Chromosomal fragility test was positive in 8 cases of FA group. Indication of HSCT for FA and DC patients which were 30 patients had BMF or transfusion dependency at transplantation;4 cases had clonal disease (2 cases myelodysplasia, 2 cases acute myeloid leukaemia). Indication of HSCT for SCN patients were uncontrollable severe infection .FA received low dose Busulfan (Bu;total dose of 6.4 mg/kg, IV), Fludarabine (Flu; total dose of 120 mg/m2, IV) , Cyclophosphamide (Cy; total dose of 2.0 g/m2, IV) based-reduced intensity conditioning(RIC) ; DC patients received low dose TBI (total 300cGy, Special position, supine) , Flu(total dose of 120 mg/m2, IV) , Cy( total dose of 3.0 g/m2, IV) based-RIC, while SCN patients had Bu(total dose of 12.8 mg/kg, IV),Cy( total dose of 3.6 g/m2, IV) or Flu(total dose of 160 mg/m2, IV) based -myeloablative conditioning(MA); and all patients combinated either of 2 different rabbit ATG ,ATG-T , rabbit anti-human thymocyte immunoglobulin, total dose 5.0 -10 mg/kg in 26 cases or ATG-F ,rabbit anti-human lymphocyte immunoglobulin, total dose 20 mg/kg in 14 cases. Campath-1, Anti-CD52 mAb was accepted with total dose 1mg/kg in 2 cases. Donor types were matched unrelated donor(MUD) in 22 patients ,Haploidentical donor (HID) in 17 patients,unrelated cord blood (UCB) in three cases. Unmanipulated stem cells were used for all patients. The Haplo-HSCT cohort received granulocyte colony-stimulating factor (G-CSF)-primed BM combined with peripheral blood stem cells (PBSCs) , The MUD HSCT cohort only received G-CSF PBSCs. The UCB HSCT cohort received one unit CB . No primary graft failure was observed. The median myeloid engraftment time was 14 (range, 10 to 21) days.Survivor median follow-up time was 38 months (range, 9-63 months), the overall survival in all patients was 76.1% ,in FA,DC,SCN were 72.4% ,100%,53.0% respectively. Cumulative incidence of 100 days acute graft-versus-host disease(GVHD) was 48.1%,Cumulative incidence 1 year and 3 years of chronic GVHD were 35.0% and 69.3% respectively. The positive chromosomal fragility test was the only independent adverse prognostic factor in multivariate analysis for FA patients rather than age ,donor type and graft source. Main causes of death were GVHD (50%) and infection (20%).No secondary malignancies occurred after HSCT till the last follow up time. Conclusion: In our study, alternative donor and disease-specific conditioning regimen HSCT for IBMFS showed promising prognosis especially for DC patients. Chromosomal fragility test positive was the only independent adverse prognostic factor in HSCT for FA patients. Disclosures No relevant conflicts of interest to declare.

Background: DNAM-1 (DNAX accessory molecule-1, CD226) is a costimulatory molecule that is constitutively expressed by NK cells and CD8+ T cells. Interaction between DNAM-1 on NK cells and CD8+ T cells with its ligands on target cells triggers cell-mediated cytotoxicity and cytokine production. Previous mouse model had showed that the expression of DNAM-1 is associated with NK education. Moreover, Monika's group found that DNAM-1 serves as an intrinsic, readout-independent marker for NK cell education in health donor. Our previous study had demonstrated that when both donors and hosts present all the KIR ligands for donor KIRs, reconstituted NK cells would achieve better functional education and contribute to least relapse for the patients post allogeneic hematopoietic stem cell transplantation (allo-HSCT). However, the roles of DNAM-1 in NK education post allo-HSCT were unknown. Aims: In this study, we have investigated the contribution of DNAM-1 expression to NK education post transplantation. Methods :We prospectively enrolled 114 patients undergoing haplo-SCT between May 2016 and April 2017 to explore the NK cell dynamic education at days 30, 90 and 180 post-transplant. Peripheral blood mononuclear cells of each sample were analyzed by 15-colors flow cytometry to evaluate of the phenotype as well as functional recovery of NK cells with different maturation expression levels of NKG2A, KIR, and CD57, as well as of activating receptors (NKp30, NKp46, NKG2D, DNAM-1) and CD25, CD122 as well as CD107a and IFN-gamma on NK cells. To study the correlation between the expression of DNAM-1 and effect of donor and host HLA on NK cell education, the expression of DNAM-1 on single-KIR+ NK cells (exhibiting NKG2A, KIR2DL1, KIR2DL2/KIR2DL3, or KIR3DL1 as their sole receptor) was compared based on the combination between donor/host HLA and donor inhibitory KIR. Results: The DNAM-1 expression on single-KIR+ NK cell ligated by sole donor HLA, or sole host HLA, or both donor and host HLA is higher compared to those single-KIR+ NK cells lacking ligands from donor or host or both. From the donor point of view, when donor only presented C1C1 ligand for KIR2DL2/L3, the MFI of DNAM-1 on single KIR2DL2/L3+ NK cells was significantly higher than KIR2DL1+ NK cells and KIR3DL1+ NK cells at 90day (P Summary: This study demonstrated that no matter from donor or/and host point of view, DNAM-1 expression contributes to optimal NK cells education post allo-HSCT. Disclosures No relevant conflicts of interest to declare.

The level of minimal residual disease (MRD) is one of the most important prognostic indicators for acute lymphoblastic leukemia (ALL). In this study, the data about 390 cases of ALL patients who obtained MRD negative after chemotherapy or CART therapy underwent allogenic hematopoietic stem cell transplantation (HSCT) in our center were retrospectively analyzed. The MRD was detected by flow cytometry or molecular methods such as fusion genes or gene mutations. 235 were males and 155 were females. Median age was 15 years old (range 2-64). According to 2016 WHO classification, the diagnosis were Pro-B-ALL(n=24), Common B-ALL(n=113), Pre-B-ALL(n=17), Hyperdiploid (n=8), BCR-ABL positive B-ALL(n=51), MLL rearranged B-ALL (n=19), TEL-AML positive B-ALL (n=13), E2A-PBX B-ALL (n=16), BCR-ABL1-like B-All (n=4), Pro-T-ALL(n=8), Pre-T-AL(n=21), Cortical T-ALL(n=17), Medullary T-ALL (n=6) and ETP T-ALL(n=1). Other 5 patients have complex karyotypes and 67 patients cannot be grouped because of absence complete information about immunophenotypes or genetic profiles. Total number of B-ALL was 295 (54 of which received CART therapy before transplant.) and T-ALL was 92. Another one was T-B mixed lineage and remaining two's lineages was unknown. 14% of the patients had sibling identical donors (n=54), 66% of the patients had haplo-identical relative donors (n=258) and others are unrelated donor (n=77) or cord blood (n=1) transplantation. Disease status before transplant are CR1 (n=228), CR2(n=134) and ≥CR3(n=28). MNC dose was 8.30(2.50-22.60)×108/kg, CD34+cells dose was 4.51(0.89-19.61) ×106/kg and CD3+cells was 1.58(0.01-37.63) ×108/kg. Preparative regimens were based on TBI (n=352) or Bu (n=38). The median time to neutrophil and platelets engraftment was 14 and 12 days. Five-years OS and disease-free survival (DFS) for all patients were 71.7% and 71.1% . Univariate analysis showed difference of impact on overall survival about the patient' gender, age, T or B immunotype, WHO classification, whether or not receiving CART therapy before transplant, conditioning regimen based on TBI or Bu, years of transplant, and time from diagnosis to transplant (≤1year, >1year, ≤2year, >2year, ≤3year, >3year) was not statistically significant. 5-ys OS for CR1, CR2, >CR2 were 77.4%, 60.9%, 67.8% (p=0.0018). Five-years OS for sibling-identical , unrelated donor and haplo-identical transplant were 74.8%, 68.8% and 79.6% (p=0.026). The incidence of gradeⅠ-Ⅱ aGVHD and grade Ⅲ-ⅣaGVHD within 100 days were 1.5% and 2.1%. Limited cGVHD was 10.3% and extensive cGVHD was 9.2%. 5 years cumulative relapse rate was 12.2%. In summary, when MRD were negative before transplantation, the result of allo-HSCT for ALL is good. The factors affecting the survival rate are disease status and interval between diagnosis and transplant time rather than gender, age, WHO classification, immunotype, whether or not receiving CART therapy and conditioning regimen. Disclosures No relevant conflicts of interest to declare.

Introduction: CD19-CAR-T cells induce high rates of initial response among patients with refractory/relapsed and MRD positive high risk B-cell acute lymphoblastic leukemia (B-ALL). Afterwards allogeneic hematopoietic stem cell transplantation (HSCT) can further reduce relapse rate. Our previous results had shown that CR obtained by CD19 CAR-T had comparable significance to CR by chemotherapy before Allogeneic HSCT in B-ALL. Patients and Methods: Between July 2015 and Mar 2018, consecutive 135 patients with refractory/relapsed or high risk B-ALL obtained CR with CD19-CART therapy followed by allogeneic HSCT were retrospective analyzed. Median follow-up of survivors was 13 months (range, 3-32 months). Results: The median age was 11 (2-49) years. The median disease course before transplant was 21(4-143) months. The median time from CART therapy to HSCT was 69 (35-312) days. Disease status was 108 cases relapsed diseases, 11 cases refractory, and 16 persistent/recurrent measurable residual diseases (MRD). MRD pre-conditioning measured by flow cytometry and QT-PCR was positive in 20(14.8%) subjects. Donor source was haploidentical donors in 107(79.3%), identical sibling in 7(5.1%), and unrelated in 21(15.6%). Most subjects (87.4%) received conventional myeloablative pretransplant conditioning with total body radiation (TBI), the rest with busulfan (Bu). Antithymocyte globulin was used in haploidentical and unrelated transplants. Cyclosporine, short-term methotrexate, and mycophenolate mofetil were employed for GVHD prophylaxis. There were no cases of graft-failure except one early death on Day 0 for septic shock. The median time to neutrophil engraftment was 14 days (10, 26 days), and median time to platelet engraftment 14 days (5, 70 days). The incidences of non-relapse mortality within 100 days were 4.4% (0.8, 7.9%) The incidence of grades II-IV acute graft-versus-host disease (GvHD) were 32.1% (24.3, 39.9%) and grades III-IV GvHD 10.5% (5.4, 15.6%). Chronic GvHD and extensive chronic GvHD were 69.7% (60.7, 78.7%) and17.6% (10.7, 24.5%). Cumulative incidence of relapses (CIRs) at 2-year was 11.1% (5.4, 16.8%). There were totally 14 subjects relapsed after HSCT, among which 8 were CD19 negative relapse, 5 CD19 positive and 1 partial CD19 positive. And among the 8 CD19 negative relapse after transplant, 4 subjects had CD19 negative MRD before conditioning. Leukemia-free survival (LFS) was 76.5% (64.2, 88.8%) and overall survival (OS) was 80.8% (72.6, 89.0%) at two years after transplant. In multivariate analysis subjects who were MRD- positive pre-transplant had a higher 2-year CIR (43.5% [18.4, 68.6%] vs. 5.9% [1.2, 10.6%]; p=0.000) and worse 2-year OS (61.5% [35.6, 87.4%] vs. 83.6% [75, 92.2%]; p=0.034). Conclusions: Our clinical results showed that CART therapy followed by allogeneic HSCT was a promising modality for refractory/relapsed B-ALL. CD19 negative relapse accounted for most relapse after allogeneic HSCT. Figure. Figure. Disclosures No relevant conflicts of interest to declare.

Background: Allogeneic hematopoietic stem cell transplantation (allo-HSCT) is an effective treatment strategy for hematological malignancies. However, graft-versus-host disease (GVHD) is a common complication after allo-HSCT resulting from the activation, amplification and secretion of numerous inflammatory factors related to donor alloreactive T cells that damage host tissues and organs, mainly in the gastrointestinal tract, liver and skin. Notably, natural killer (NK) cells represent the first donor-derived lymphocyte population to recover after allo-HSCT and generally are observed within the first month after allo-HSCT. NK cells express a series of immune receptors that identify relevant ligands on target cells and maintain the immune balance between NK cell activation and tolerance. Previous studies have shown that the number of NKG2A+ cells is decreased in patients with chronic GHVD after HSCT. However, the relationship between NKG2A+ NK cells and aGVHD has not been characterized. In addition, the role of NKG2A+ cells in aGVHD disease progression and the mechanism underlying NKG2A+ cell immunoregulation have not been clearly explained. Objective: In this study, we used peripheral blood from GVHD, non-GVHD paired specimens and healthy donors to address the underlying mechanism by which NKG2A+ NK cells regulate T cells after HSCT. Methods: We detected the specimens using flow cytometry from two independent cohorts, which were prospective cohort and paired cohort. Futher, we performed experiments in vitro to investigate the potential role of NKG2A+ NK cells on T cells. Results: Here, we found that, compared with non-GVHD subset, NKG2A+ NK cells percentage and absolute cell counts were significantly reduced in GVHD patients after HSCT. Moreover, the reduction of NKG2A+ NK cells in GVHD patients was ascribed to its increased apoptosis and decreased proliferation capacity, while retaining a strong graft-versus-leukemia (GVL) effect. In vitro assay showed that when co-cultured T cells with NKG2A+ NK cells, the T cells secreted IFN-r level was significantly reduced, while the IL-4 level was increased. Moreover, CD25 expression level was decreased, while the CD4+CD25+FOXP3+ cells number was increased. In addition, the NKG2A+ NK cells induced T cell apoptosis and decreased T cell proliferation during the coculture process. Significantly, NKG2A+ mainly regulated activated but not resting T cells. In vivo assay showed that serological IL-10 level in GVHD subset was evidently lower than those of non-GVHD subgroup, the IL-1β, IFN-r and TNF-a level was however higher in the GVHD subgroup. Furthermore, the percentage of NKG2A+ NK cells from GVHD patients was markedly increased by the presence of exogenous IL-10, but not by other cytokines. However, this phenotype was not observed at non-GVHD patients. Together, the GVHD might be ascribed to lower IL-10 induced NKG2A+ NK cells reduction, which further overactivate T cells after HSCT. Discussion and Conclusion: Overall, we herein observed reduced proportions and absolute cell counts of NK cells and NKG2A+ subsets in patients with acute GVHD after allo-HSCT. The causative association between NK cell numbers, NKG2A+ subsets and GVHD remains debatable. Based on our results, speculating that reduced proportions of NKG2A+ subsets in patients after allo-HSCT are associated with acute GVHD due to their interplay with the patient's donor-derived alloreactive T cells is tempting. Disclosures No relevant conflicts of interest to declare.

Backgroud: Refractory cytomegalovirus (CMV) infection remains important causes of morbidity and mortality after allogeneic hematopoietic stem cells transplantation (allo-HSCT). Previous researches reported that adaptive immunity, such as CD8+ CMV-CTL, plays an important role in the in the control of refractory CMV infection. In mouse, Lanier et al. found there existed subsets of adaptive NK cells with the features of expanding, contracting after control of mouse CMV, and generating long-lived "memory" NK cells. In human, these adaptive NK cells were initially identified based on the high expression of the NKG2C which against HCMV through their cytotoxic potential and the production of TNF-α and IFN-γ upon Ab-mediated stimuli in vitro. Meanwhile, the expression levels of the NKG2C+ adaptive NK cells has been positively correlated with the NKG2C copy number. Several researchers had found that NKG2C+ adaptive NK cells persistent expanded and were potent producers of IFN-γ during CMV reactivation after solid-organ transplant or allo-HSCT. However, the role of NKG2C+ adaptive NK cells on refractory CMV reactivation were still unknown. Whether the rapid reconstitution of NKG2C+ adaptive NK cells can reduce the refractory CMV reactivation merit to be investigated. Aims:In this research, we had investigated the impacts of the quantity and quality recovery of NKG2C+ adaptive NK cells on the occurrence of refractory CMV infection. Method: At first, continuous 364 patients underwent allo-HSCT since June 2012 to February 2016 were prospectively enrolled and we retrospectively analyzed the correlationship between their donor NKG2C genotype and refractory CMV infection occurrence post transplantation. Secondly, the second cohort comprising continuous 125 patients underwent allo-HSCT since May 2016 to April 2017 were prospectively enrolled to analyzed the effect of donor NKG2C genotype on NKG2C+ adaptive NK cell recovery as well as the effect of NKG2C+ adaptive NK cell recovery on refractory CMV infection. The cytotoxicity of reconstituting NKG2C+ adaptive NK cells were evaluated against K562 cells, AD169 CMV stain infected MRC-5 cells, and UL40 peptide pulsed 721.221 cells to detect the anti-tumor or anti-CMV function of NKG2C+ adaptive NK cells. Results: Firstly, from the first cohort, we found that donor NKG2C gene deletion was an independent prognostic factor for refractory CMV reactivation (P=0.010) through the multivariate analysis. Then, through in-depth investigation from the second cohort, we found that the absolute cell counts recovery and anti-tumor function of NKG2C+ adaptive NK cells were both significantly lower in patients accepting NKG2C+/del donor than those patients accepting NKG2C+/+ donors at day 30, 90, and180, respectively after transplantation. There was no NKG2C+ adaptive NK cell recovery post transplantation in the patients who accepted NKG2Cdel/del donors. Meanwhile, anti-CMV function recovery of NKG2C+ adaptive NK cells in patients with NKG2C+/del donors were significantly lower than those patients with NKG2C+/+ donors at day 30 post transplantation. Furthermore, we further analyzed the relationship between the early reconstitution of NKG2C+ adaptive NK cells and refractory CMV infection occurrence. The patients were divided into three groups: no CMV, CMV reactivation (persistent time of CMV infection 2 weeks). We found that the absolute cell counts of NKG2C+ adaptive NK cells in refractory CMV group was significantly lower than that of other two groups at day 30 transplantation. When the patients were devided into high and low level groups based on the ROC cut-off percentage of NKG2C+ adaptive NK cells (1.42%), the result revealed that the patients with lower level of NKG2C+ adaptive NK cells at day 30 post-HSCT had an higher cumulative incidence rate of refractory CMV infection (81.1%) comparing with the higher one (40.5%) (P=0.0014). Moreover, Cox regression model further demonstrated that the lower level of NKG2C+ adaptive NK cells at day 30 post-HSCT was significantly associated with refractory CMV infection (HR=2.578, 95% CI 1.379-4.21, P=0.003). Summary/Conclusion: Our results indicated that donor NKG2C deletion damaged the reconstitution of NKG2C+ adaptive NK cells after allo-HSCT, therefore increased the occurrence of refractory CMV infection post transplantation. Disclosures No relevant conflicts of interest to declare.

Licensed NK cells have been demonstrated to have anti-cytomegalovirus (CMV) activity. We prospectively analysed the HLA typing of donor-recipient pairs and the KIR typing of the donors in 180 leukaemia patients to assess the predictive roles of licensed NK cells and donor-activating KIR genes on CMV reactivation post-T cell-replete haplo-SCT. Multivariate analysis showed that donor and recipient KIR ligand-ligand graft-versus-host direction or host-versus-graft direction mismatch were associated with the incidence of CMV reactivation [HR=1.663, 95%CI, 1.161-2.383, P=0.006] and refractory CMV infection [HR=2.34, 95%CI, 1.28-4.27, P=0.006] post-haploidentical T cell-replete transplantation. Donor and recipient KIR ligand-ligand match decreased CMV reactivation (51.65% [46.67, 56.62%] vs. 75.28% [70.87, 79.69%], P=0.012), refractory CMV infection (17.58% [13.77, 21.40%] vs. 35.96% [31.09, 40.82%], P=0.004) and CMV disease (3.30% [1.51, 5.08%] vs. 11.24% [8.04, 14.43%], P=0.024), respectively. Meanwhile, there was a significantly increased risk of CMV reactivation in patients who accepted a KIR2DS2-positive donor compared those who accepted a KIR2DS2-negative donor (80% [71.93, 88.07%] vs. 63.87% [60.18, 67.56%], P=0.039), especially in recipient HLA-C1/C2 and donor HLA-C1C2 with KIR2DL1/2DL2/2DL3/2DS2 genotype (100%). Conclusions: These findings suggested that donor and recipient KIR ligand-ligand match might promote the NK cell licensing process, thereby increasing NK cell-mediated protection against CMV reactivation. However, roles of donor positivity for the KIR2DS2 against CMV reactivation need to be further explored in T cell-replete haplo-SCT. Disclosures No relevant conflicts of interest to declare.

Introduction: Cytogenetic abnormality is considered to be an independent prognostic factor in newly diagnosed acute myeloid leukemia (AML). However, recent studies have demonstrated that acquired gene mutations also play an important role in the pathogenesis and prognosis of AML. It has been well known that minimal residual disease (MRD) pre-conditioning has remarkable impact on disease-free survival (DFS) after allogeneic hematopoietic stem cell transplantation (allo-HSCT) in acute lymphoblastic leukemia, but the effect of MRD pre-transplant on allo-HSCT in AML is still unclear. Objective: In present study, the effect of NCCN risk stratification which has integrated gene mutations into cytogenetics as well as MRD pre-transplant on DFS after allo-HSCT in AML was studied in order to learn whether risk-directed conditioning and prevention of relapse are needed. Methods: Between April 2012 and March 2015, consecutive 258 patients with AML in complete remission (CR) (186 cases in CR1 and 72 cases in CR2) who underwent allo-HSCT in our hospital were analyzed retrospectively. The median age was 25 (1.8-64) years. Male (M) to female (F) was 147:111. The median disease course was 6 (1-51) months. According to 2015-NCCN risk stratification, 63 (24.4%) cases were in low risk, 112 (43.4%) cases in intermediated risk, and 83 (32.2%) cases in high risk. MRD in bone marrow pre-conditioning was detected by eight-color flow cytometry. Results: With the median follow up 18 (5-41) months, overall 2-year DFS was78.0%. No significant difference in DFS was found among low-risk (78.6%), intermediated-risk (76.0%) and high-risk (80.3%) patients (P=0.886). 205 (79.5%) cases were MRD- and 53 (20.5%) cases were MRD+ before conditioning. DFS after transplant in MRD+ patients was significant lower than that in MRD- patients(65.0% vs. 81.4%, P=0.003). Univariate analysis showed that DFS was not associated with patient age (≤14years vs.>14years, P=0.292), disease course before HSCT (≤6 months vs.>6months, P=0.532), WBC counts at diagnosis (≤50×109/L vs.>50×109/L, P=0.120), CBC recovery pre-HSCT (yes vs. no, P=0.664), disease status (CR1 vs. CR2, P=0.201), extramedullary leukemia before transplant (yes vs. no, P=0.532), conditioning regimen (BUCy/Flu-based vs. TBICy/Flu-based, P=0.753), donor type (identical sibling vs. unrelated vs. haploidentical, P=0.743), donor-recipient gender (M-M vs. M-F vs. F-M vs. F-F, P=0.245), donor-recipient blood type (compatibility vs. major incompatibility vs. minor incompatibility vs. major and minor incompatibility, P=0.402), mononuclear cells infused (≤8×108/kg vs.>8×108/kg, P=0.583), CD34+ cells infused (≤4×106/kg vs.>4×106/kg, P=0.946), and CD3+ cells infused (≤1.6×108/kg vs.>1.6×108/kg, P=0.143). DFS was significant lower in the patients with secondary AML (79.4% in primary AML vs. 53.5% in secondary AML, P=0.006) and MRD+ cases before transplant (81.4% in MRD- vs. 65.0% in MRD+, P=0.003). Accumulative non-relapse mortality (NRM) was significant higher in secondary AML (11.7% in primary AML vs. 33.3% in secondary AML, P=0.004) and MRD+ patients (10.5% in MRD- vs. 21.9% in MRD+, P=0.010). Accumulative relapse rate was significant higher in CR2 cases (8.0% in CR1 vs. 17.5% in CR2, P=0.046). Multivariate analysis showed that MRD pre-HSCT was the only impact factor on DFS and NRM with higher DFS (P=0.020) and lower NRM (P=0.045) in MRD- cases. Conclusions: Allo-HSCT has attenuated the influence of cytogenetics and gene mutations on DFS in AML. Secondary AML has lower DFS and higher NRM. Although disease status (CR1 vs. CR2) has no significant influence on DFS, relapse rate in CR2 is higher than that in CR1. MRD pre-conditioning was a key impact factor on DFS after allo-HSCT in AML but not conditioning regimen and donor type. Disclosures No relevant conflicts of interest to declare.

Introduction: The prognosis of refractory/relapsed acute lymphoblastic leukemia (ALL) is poor with chemotherapy or even allogeneic hematopoietic stem cell transplantation (HSCT). Although our immunotherapy with autologous anti-CD19 chimeric antigen receptor T cells (CART) has resulted in 88.6% complete remission (CR) in refractory/relapsed B-cell ALL (B-ALL), many patients relapsed at around 2 months after CART therapy (unpublished data). Our current strategy is to perform HSCT for refractory/relapsed B-ALL in CR by CART therapy to attain continuous leukemia-free survival (LFS). However, majority of the patients with CART therapy developed cytokine release syndrome which may increase transplant-related mortality (TRM). Moreover, all patients with CART therapy have very tough diseases which could result in higher relapse rate after transplant. Objective: In current study, the safety and efficacy of HSCT for refractory/relapsed B-ALL after CART therapy were investigated. The patients with B-ALL who received HSCT during the same time period without CART therapy were as control. Patients and Methods: Between July 2015 and May 2016, consecutive 22 patients with refractory/relapsed CD19+ B-ALL in CR by CART therapy followed by allogeneic HSCT in our hospital were analyzed as CART group; and consecutive 89 patients with B-ALL in CR who received allogeneic HSCT in our hospital during the same time period but without previous CART therapy were as control group. Clinical characteristics between two groups was comparable except more CR1 (22.7% vs. 57.3%) in control group and more CD3+ cells infused (1.93x108/kg vs. 1.46 x108/kg, p=0.026) in CART group. The median age was 8 (2-44) years, 15 (2-52) years in CART and control groups (p=0.147). The median disease course was 19.1 (3.9-53.7) months, 10.6 (3.7-123.0) months in CART and control groups (p=0.385). The median time from CART therapy to HSCT was 86 (31-172) days. Disease status was 22.7% cases in CR1, 54.5% in CR2, 18.2% in CR3 and 4.5% in CR4 in CART group; and 57.3% cases in CR1, 36.0% in CR2 and 6.7% in CR3 in control group (p=0.08). Minimal residual disease pre-conditioning by flow cytometry was positive in 22.7%, 31.5% patients in CART and control groups (p=0.422). Donor source was identical sibling (IS) in 13.6%, unrelated (UR) in 31.8% and haploidentical (HI) in 54.5% in CART group; and IS in 14.6%, UR in 16.9% and HI in 68.5% in control group (p=0.313). Myeloablative conditioning regimens were administered with either total body irradiation (TBI) plus cyclophosphamide (Cy)/ fludarabine (Flu)-based in 90.9% cases or busulfan (Bu) plus Cy/ Flu-based in 9.1% cases in CART group; and TBICY/Flu-based in 85.4% cases or BuCy/Flu-based in 14.6% cases in control group (p=0.498). Antithymocyte globulin was used in UR and HI transplants. Cyclosporine, short-term methotrexate, and mycophenolate mofetil were employed for GVHD prophylaxis. Results: The median time to neutrophil engraftment was similar between two groups (14 days vs. 13 days, p=0.196), but platelet engraftment was slower in CART group (14 days vs. 12 days, p=0.031). Cumulative incidence of grade II-IV acute GVHD (aGVHD) was higher in CART group (57.6% vs. 33.1%, p=0.009), which may related to higher CD3+ cells infused in CART cohort; but the incidences of grade III-IV aGVHD were no statistical significance (25.1% vs. 15.7%, p=0.564) in two groups. No remarkable differences were seen in CMV reactivation (45% vs. 51.7%, p=0.601) and transplant-associated thrombotic microangiopathy (13.6% vs. 9.0%, p=0.514) in two groups. No significant difference was found in TRM between CART and control groups (7.1% vs. 15.2%, p=0.808). Relapse rates were similar in two groups (5.0% vs. 6.9%, p=0.888). With a median follow-up 9 (2-12) months, LFS was comparable in CART and control groups (84.8% vs. 80.9%, p=0.937). Conclusions: Our preliminary results have shown that the strategy with CART therapy followed by allogeneic HSCT in refractory/relapsed B-ALL is very safe and effective with similar outcomes in TRM, relapse rate and LFS compared with control group. CART therapy has resulted in very good CR in refractory/relapsed B-ALL and allowed the patients to achieve continuous LFS by subsequent allogeneic HSCT, which is revolutionary modality for those patients who have otherwise incurable diseases. Disclosures No relevant conflicts of interest to declare.

Introduction: Cytogenetics is an independent prognostic factor in acute myeloid leukemia (AML). Molecular genetics including leukemia fusion gene, gene mutation and gene over expression are recognized to have significant impact on survival in patients with AML as well. In present study, the impact of cytogenetic and molecular markers on disease-free survival (DFS) of allogeneic hematopoietic stem cell transplantation (HSCT) for AML was investigated. Methods: Between April 2012 and December 2014, consecutive 345 patients with AML who underwent allogeneic HSCT in our center were analyzed retrospectively. All patients were either in poor-risk or in good-risk/intermediate-risk but with persistent minimal residual disease. The median age was 19 (1.8 to 64) years old. Children (≤14 years) were 96 (27.8%) cases and adults (>14 years) were 249 (72.2%) cases. Male to female was 200:145. The median disease course was 6 (1-64) months. Leukocyte count at diagnosis was < 30 x 109/L in 230 (66.7%) patients (low leukocyte) and ≥30 x 109/L in 115 (33.3%) cases (high leukocyte). Transplants at CR1, ≥CR2, and advanced disease were 168 (48.7%), 53 (15.4%) and 124 (35.9%), respectively. Donor sources were identical sibling (IS) in 45 (13.0%) cases, unrelated (UR) in 71 (20.6%) cases and haploidentical (HI) in 229 (66.4%) cases. Myeloablative conditioning regimens were administered with either Busulfan (Bu) plus Cyclophosphamide (Cy)/Fludarabine (Flu)-based in 285 (82.6%) patients or total body irradiation (TBI) plus Cy/Flu-based in 60 (17.4%) patients. Antithymocyte globulin was used in unrelated and haploidentical HSCT. Unmanipulated bone marrow and peripheral blood stem cells (PBSC) for IS and HI HSCT and PBSC for UR transplant were applied as the grafts. Cyclosporine, short-term Methotrexate, and Mycophenolate mofetil were employed for GVHD prophylaxis. Results: Univariate analysis showed that DFS after allogeneic HSCT in AML was not associated with patient age (children vs. adults, 70.3% vs. 69.4%, p=0.6), leukocyte count at diagnosis (low leukocyte vs. high leukocyte, 68.8% vs. 71.3%, p=0.8), donor source (IS vs. UD vs. HI, 77.3% vs. 76.8% vs. 65.8%, p=0.21), and conditioning regimen (Bu-based vs. TBI-based, 70.1% vs. 67.3%, p=0.45). Multivariate analysis indicated that disease status before HSCT was the only impact factor on DFS (CR1 vs. ≥CR2 vs. advanced disease, 81.6% vs. 70.0% vs. 53.1%, p Disclosures No relevant conflicts of interest to declare.

Introduction: With GIAC regimen, haploidentical blood and marrow transplantation (haplo-BMT) has achieved comparable outcomes with identical sibling transplant (Dao-Pei Lu et al., Blood 2006; 107:3065). Our previous study has shown that the third party cell co-infusion in haplo-BMT (GIAC-3 regimen) could significantly reduce aGVHD and transplant-related mortality (TRM). We have also demonstrated that individualized chemotherapy to decrease leukemia burden followed by conditioning could improve disease-free survival (DFS) in refractory/relapsed AML. Objective: To learn the outcomes of our haplo-BMT with these integrated approaches, all patients who received haplo-BMT for hematologic malignancies in our center were analyzed retrospectively. Methods: Between April 2012 and December 2014, consecutive 514 patients with hematologic malignancies who underwent haplo-BMT were included. The median age was 20 (1.8 to 64) years old. The diagnosis included AML 232 (45.1%), ALL 207 (40.3%), MDS 27(5.3%), CML 14 (2.7%), lymphoma 13 (2.5%) and others 21 (4.1%). Transplants at CR1, ≥CR2 or advanced disease were 216 (42.0%), 114 (22.2%), 184 (35.8%), respectively. All patients received unmanipulated bone marrow (BM) and peripheral blood stem cells as graft after myeloablative conditioning plus ATG. Majority of the patients with AML received BuCy-based conditioning, while most ALL patients received TBICy-based regimen. Fludarabine was substituted for cyclophosphamide in some patients due to impaired organ function or high tumor burden. For refractory/relapsed diseases, individualized chemotherapy followed by conditioning was administered. Cyclosporine/tacrolimus, short-term Methotrexate, and Mycophenolate mofetil were employed for GVHD prophylaxis. Either 1ml/kg (recipient's body weight) haploidentical BM from the second haploidentical donor or one unit of unrelated cord blood was infused right after haplo-BMT as the third party cells. Minimal residual disease (MRD) was monitored routinely by quantitative PCR or flow cytometry. The patients with persistent MRD were interfered by immunosuppressant withdrew, adoptive immunotherapy with cytokine induced killer or NK cells or donor lymphocyte infusion. Results: All patients but 5 achieved durable engraftment. The cumulative incidences of grade II to IV aGVHD and grade III to IV aGVHD were 32.2%, 19.8%, respectively. The cumulative incidences of cGVHD and extensive cGVHD were 48.3%, 18.4%, respectively. 100-day TRM and 2-year TRM were 4.1%, 14.9%, respectively. Two-year relapse rate was 22.8%. With the median follow up 17 (6 to 38) months, overall 2-year DFS rates in CR1, ≥CR2 and advanced disease were 75.6%, 70.9%, 49.2%, respectively. For AML, two-year DFS rates in CR1, ≥CR2 and advanced disease were 74.1%, 76.9%, 48.2% (CR1 vs. ≥CR2, p=0.84; CR vs. advanced disease, p=0.000). For ALL, two-year DFS rates in CR1, ≥CR2 and advanced disease were 78.9%, 56.6%, 38%, respectively (CR1 vs. ≥CR2, p=0.018; CR1 vs. NR, p=0.000; ≥CR2 vs. NR P=0.02 ). Conclusions: With our strategies, overall outcomes of haplo-BMT have been improved remarkably and very encouraging. Therefore, haplo-BMT should be an important way to save life for the patients with hematologic malignancies who need urgent BMT but without matched either sibling or unrelated donor. Disclosures No relevant conflicts of interest to declare.

Introduction: Relapse remains the main cause of failure of hematopoietic stem cell transplantation (HSCT) in acute leukemia. NK cells have the property of killing leukemia cells without GVHD aggravation theoretically. Moreover, in some cases, leukemia cells may lost HLA-I and/HLA-II antigens which would result in poor response to the immunotherapy except NK-based adoptive effectors. Objective: In present study, the safety and efficacy of donor-derived ex-vivo activated NK cells in management of relapse after allogeneic HSCT in high-risk acute leukemia were examined. Patients and methods: Between July 2012 and July 2014, 29 patients with acute leukemia who received NK cell infusion after HSCT were analyzed retrospectively. Some cases failed to chemotherapy combined with donor lymphocyte infusion (DLI) before NK cell therapy. The diagnosis were ALL (10 cases), AML (18 cases) and mixed acute leukemia (1 case). All patients were high-risk leukemia. The disease status before transplant was CR1 in 8 cases, CR2 in 7, CR3 in 1 and non-remission in 13. The types of donor included identical sibling (5 cases), haploientical family member (21 cases) and unrelated donor (3 cases). The conditioning and GVHD prophylactic regimens were reported previously (Lu DP et al., Blood 2006; 107:3065). Minimal residual disease (MRD) was detected by either quantitative RT-PCR for fusion genes or flow cytometry or both. The expression of HLA-I and HLA-II antigens in leukemia cells was evaluated by flow cytometry. Donor-derived either peripheral blood stem cells or lymphocytes were cultured for 6 days using original culture system (AIM-V medium with IL-2, IL-12, IL-15 and IL-21) or modified culture system (SCGM medium with IL-2, IL-12, IL-15, IL-18 and IL-21). Escalated dosage of NK cells were infused starting with 1×105 cells/kg (recipient’s body weight) with or without IL-2 injection. Nine patients were in prevention group and 20 cases were in treatment group. The patients with hematologic relapse received NK cells 3 days later after chemotherapy. Results: Compared with our original culture system, the modified culture system enhanced approximately 10% to 20% of the purity and 4 to 8 fold in number of NK cells by day 6. Furthermore, our modified culture system elevated the expression of function phenotype including TRAIL, NKG2D and CD62L on NK cells in approximately 8 to 10 folds at day 6 and simultaneously stimulated higher level of IFN-γ. One to 4 NK cell infusions were given in each case with two week interval. Two of 29 cases developed mild skin GVHD. No transfusion-related side effects were noted. In prevention group, four of 9 cases remain complete remission, and the other 5 patients became MRD positive or relapse. In treatment group, seven of 20 cases have response to NK cell therapy, and two out of 7 cases who response to NK cells had failed to chemotherapy plus DLI before. Among 11 patients who had response to NK cells, eight of them are AML, and the remaining 3 patients are ALL. Higher response rate (10/23 cases) was seen with NK cell therapy by our modified culture system compared with the one (1/6 cases) by our original culture system. Conclusions: Our preliminary results have demonstrated that donor-derived ex-vivo activated NK cells are safe and effective modality in the management of relapse after allogeneic HSCT in high-risk acute leukemia even failed to chemotherapy combined with DLI. Optimal culture system has improved not only NK cell’s purity, number and function phenotype but also clinical efficacy. Disclosures No relevant conflicts of interest to declare.

Introduction: With GIAC regimen (G: G-CSF priming; I: intensified immune-suppression; A: ATG application; C: combination of peripheral blood stem cell and bone marrow (BM) as graft), haploidentical blood and marrow transplantation (haplo-BMT) has been an important alternative option for patients with hematological malignancies who need urgent transplant but without HLA matched either sibling or unrelated donor (Dao-Pei Lu et al., Blood 2006; 107:3065). In a murine model, we have first demonstrated that the animals transplanted with three mixed cells (A+B+C to A) were survived much longer due to milder acute GVHD (aGVHD) compared with the mice transplanted with single allogeneic BM (B to A). Our previous clinical study has shown that with GIAC-3 protocol (GIAC with the third party cells in order to induce immune-tolerance) that unrelated cord blood (UCB) as the third party cells, the incidences of aGVHD, especially for severe aGVHD, and also treatment-related mortality in haplo-BMT have been significantly reduced. Objective: In present clinical study, we examine whether haploidentical bone marrow (BM) from the second donor that substitutes for UCB as the third party cells could also reduce aGVHD and improve survival in haplo-BMT setting. Patients and Methods: Between April 2012 and June 2013, total 158 haplo-BMT patients with hematological malignancies were enrolled. The median age was 18 (1.8 to 56) years old. The diagnosis included AML (46.2%), ALL (42.4%), lymphoma (3.8%), CML (2.5%), MDS (1.9%), acute mix leukemia (1.9%), JCMML(0.6%)and plasma cell leukemia (0.6%). Transplants at CR1, ≥CR2 or CML-AP, and advanced disease (refractory/relapsed acute leukemia or CML-BC) were 34.8%, 24.7% and 40.5%. All patients received unmanipulated blood and marrow transplant after BUCy2 or CyTBI plus antithymocyte globulin as pre-conditioning. Fludarabine was substituted for cyclophosphamide in some patients due to impaired organ function or high tumor burden. Cyclosporine, short-term Methotrexate, and Mycophenolate mofetil were employed for GVHD prophylaxis. 1ml/kg (recipient’s body weight) haploidentical BM from the second donor was infused right after haplo-BMT as the third party cells. Results: All patients but two achieved engraftment. One case had primary graft failure and the other one had secondary graft failure. Both patients obtained durable hematopoietic reconstitution after the second BMT. Low levels of the third party cells were detected in a few patients at early stage after transplantation. No long-term third party cell engraftment was found. The cumulative incidences of grade II to IV aGVHD and grade III to IV aGVHD were 30.5%, 14.0%, respectively. With the median follow-up time of 17 months, two-year leukemia-free survival (LFS) rates in CR and NR cases were 73.2%, 41%, respectively, and 2-year overall survival (OS) rates in CR and NR patients were 86.9%%, 55.5%, respectively. Conclusions: Our preliminary clinical results have shown that with current GIAC-3 protocol,the outcomes of haplo-BMT have been improved remarkably with lower aGVHD and better LFS and OS compared with our historic control. Haploidentical BM from the second donor as the third party cells is more feasible and cost effective compared with UCB. The mechanism of this strategy need to be further investigated. Disclosures No relevant conflicts of interest to declare.

Introduction Allogeneic hematopoietic stem cell transplantation (HSCT) is an only curative modality currently for myelodysplastic syndrome (MDS). High-risk MDS usually has lower complete remission (CR) rate and higher chemotherapy-related mortality compared with de novo acute myeloid leukemia (AML). To examine whether CR before HSCT has survival benefit for MDS treated by HSCT, we retrospectively analyzed the data during 11 years from our center. The clinical outcomes of MDS after HSCT from different donor sources have also been evaluated. Objective In present clinical study, the effects of disease status and donor sources on disease-free survival (DFS) of MDS after HSCT were studied. Methods From August 2001 to December 2012, total 122 patients with MDS that underwent HSCT in our center were enrolled. Male to Female was 76: 46. The median age was 35 (8 to 57) years old. The median blasts in bone marrow (BM) before conditioning were 9% (1% to 65%). According to 2008 WHO classification, the patients were diagnosed as refractory cytopenias with unilineage dysplasia (RCUD) in 12, refractory anemia with ring sideroblasts (RARS) in 2, 5q- in 1, refractory cytopenias with multilineage dysplasia (RCMD) in 15, refractory anemia with excess blasts (RAEB) -1/RAEB-2 in 36 and transformed AML in 56. For International Prognostic Scoring System (IPSS), 12 patients were in low-risk, 27 in intermediate-1, 24 in intermediate-2, and 59 in high-risk. Based on BM blast percentage pre-conditioning, 47 cases were less than 5%, 43 patients were between 5% to 20%, and 32 cases were more than 20%. The stem cells were from identical siblings (45) or unrelated donor (24) or haploidentical family members (53). Conditioning regimens were BUCY/BUFLU for identical sibling HSCT, and BUCY/BUFLU plus ATG (Thymoglobuline, 8-10mg/kg) for unrelated or haploidentical transplants. Graft-versus-host disease prophylaxis was employed by Cyclosporin A, Methotrexate and Mycophenolate mofetil as reported previously (DP Lu et al., Blood 2006; 107:3065). Results: With median follow-up 31 (1-144) months, DFS was 73.8%. Fourteen patients (11.4%) relapsed. Transplant-related mortality was 14.8%. No significant differences on DFS were found among RCUD/RARS/5q- (68.8%), RCMD (85.7%), RAEB-1/RAEB-2 (72.2%) and transformed AML (73.2%) (p=0.761). A similar DFS was seen in different risk categories (73.3% in low-risk, 79.2% in intermediate-1, 75.0% in intermediate-2 and 71.2% in high-risk; p=0.861). Moreover, CR or not before HSCT has no remarkable effect on DFS (blasts 20%, 75.0%; p=0.342). Donor sources have also no significant effects on DFS (identical sibling 75.6%, unrelated donor 79.2%, haploidentical family member 69.8%; p=0.651). Conclusions Our clinical results have shown that under current protocol, DFS of MDS after allogeneic HSCT is quite encouraging no matter the disease status and stem cell donor sources. Therefore, it is not necessary that complete remission is achieved by chemotherapy before transplant. Haploidentical family member is an important alternative donor for patients with MDS when matched either identical sibling or unrelated donor is not available. Disclosures: No relevant conflicts of interest to declare.

Abstract 228 Haploidentical hematopoietic stem cell transplantation (haplo-HSCT) is an important alternative option for patients with hematological malignancies who need urgent transplant but without HLA matched either sibling or unrelated donor (Dao-Pei Lu et al., Blood 2006; 107:3065). Each patient usually has several haploidentical family members who could be selected as a donor. To determine the principal of donor selection among all available related haploidentical candidates, the clinical outcomes of a large series of haplo-HSCT in our hospital are analyzed. From April 2002 to April 2010, consecutive 440 patients with hematological malignancies who underwent haplo-HSCT were included. The median age of patients was 23 (3-59) years old. The diagnosis included AML (39.8%), ALL (35.9%), MDS (3.6%), CML (16.1%), and others (4.6%). Transplants at CR1 or CML-CP1, ≥CR2 or CML-CP2/AP, and advanced disease (refractory/relapsed acute leukemia or CML-BC) were 33.4%, 40.9% and 25.7%. HLA mismatched at 1, 2, 3 loci was 13.2 %, 27.5%, 59.3%, respectively. Transplants in sex-matched donor-recipient pair, female donor to male recipient, and male donor to female recipient were 55.1%, 33.0% and 13.9%. Major clinical characteristics among these three arms were similar. All patients received unmanipulated combined marrow and peripheral blood stem cells for transplant after BUCy2/CyTBI plus ATG conditioning. Fludarabine was substituted for cyclophosphamide in some patients due to impaired organ function. Prophylaxis and treatment of GVHD were reported previously. Steady hematopoietic reconstitution was seen in 98.6% of recipients. The cumulative incidences of grade II to IV acute GVHD and chronic GVHD were 32.6%, 61.3%, respectively. 100-day mortality was 10.5%. With the median follow-up of 32 (3-99) months, 2-year and 5-year overall survival (OS) rates for patients who were in different disease status were 57.9% and 52.9%. Univariate and multivariate analysis all showed that disease status before transplant, CD34+ cell dosage infused and sex-matched or not between donor and recipient but not HLA disparity and age were pivotal impact factors on survival. Two-year OS of transplants in sex-matched donor-recipient pair, female donor to male recipient, male donor to female recipient were 65.5%, 55.3%, 37.6, respectively (p=0.000). No significant differences were found on OS of transplants among haploidentical donors from sibling or parent or offspring or other relatives. In conclusion, our clinical results from a large series of transplants demonstrate that haplo-HSCT in sex-matched donor-recipient pair has survival advantage. Therefore, in haplo-HSCT, sex-matched donor should be the first choice, if not available, then female donor to male recipient could be the second option. Disclosures: No relevant conflicts of interest to declare.

Abstract 4095 Poster Board III-1030 The clinical outcomes of refractory/recurrent leukemia that salvaged by allogeneic hematopoietic cell transplantation (HCT) is usually poor. It is crucial for the management of those high-risk patients with appropriate conditioning regimens that balance the efficacy and safety well and with enhanced anti-leukemia effect post-HCT. Our previous study has shown that donor's dendritic cell-primed cytokine-induced killer cells (DC-CIK) is a safe and effective therapy in the management of early leukemia recurrence after allogeneic HCT, which fail to or ineligible for current standard treatment. Objective In present clinical study, we examine the efficacy of refractory/recurrent myelogenous leukemia salvaged by allogeneic HCT and prophylactic immunotherapy. Methods From September 2006 to May 2008, 30 patients with refractory/recurrent myelogenous leukemia (AML 29, CML-BC 1) were enrolled. The median age was 32 (12 to 55) years old. The median blasts in bone marrow were 36% (20% to 87%) prior to conditioning. The grafts were from HLA identical siblings (5), unrelated donors (7), and haploidentical family members (18). Conditioning regiments were individualized according to patients' status as following. Generally, the regimen with high-dose cytarabine plus BUCY2 was used (13 cases). The patients with impaired organ function received above regimen except with fludarabine instead of cyclophosphamide (11 cases). For the recipients with > 40% blasts in bone marrow, melphalan (2 cases) or aclarubicin (1 case) was added into the regimen or FLAG followed by reduced-intensified BUCY2 (3 cases) was employed in order to reduce leukemia burden. Cyclosporine A, methotrexate and mycophenolate mofetil were administrated for GVHD prophylaxis. To prevent leukemia relapse, immunosuppressants were tapered off early post-HCT. Prophylactic immunotherapy including cellular (DLI, DC-CIK, NK cells), and humoral (IL-2, IFN-a, thymosin) was used selectively in the patients who had no GVHD 120 days after HCT. Results The median mononuclear cells in the graft were 7.36 (3.49 to 11.5) ×108/kg. The median CD34+ cells were 4.06 (1.57 to 11.4) ×106/kg. The median CD3+ cells were 1.42 (0.75 to 3.61) ×108/kg. Twenty-nine patients attained sustained engraftment. One died of multi-organ failure before hematopoietic reconstitution. The median time for white blood cells > 1.0 × 109/L, and platelets > 20 × 109/L was 14 (10 to 21) days, 15 (12 to 26) days, respectively. Thirteen patients developed acute GVHD (grade I in 5, grade II in 7, and grade III in 1). Thirteen patients developed chronic GVHD after immunosuppressants' reduction or withdrawal. In addition, 13 patients received prophylactic immunotherapy due to lack of chronic GVHD 120 days post-HCT, then 7 of 13 developed chronic GVHD. With the median follow-up of 15 (3 to 35) months, two (6.7%) patients with AML had hematological recurrence. One patient attained durable complete remission again after treatment with chemotherapy followed by immunotherapy with DLI, DC-CIK, and NK cells. Five (16.7%) patients died (infections in 2, hematological relapse in 1, chronic GVHD in 1, and multi-organ failure in 1). 25 of 30 (83.3%) patients have been in continuous complete remission since salvaged HCT. Conclusion Our preliminary clinical results have shown that the combination of salvaged allogeneic HCT and prophylactic immunotherapy is a promising modality for the treatment of myelogenous leukemia in refractory/recurrent status, even with high leukemia burden. Disclosures: No relevant conflicts of interest to declare.

More alternative donor stem cell transplantation (SCT) has been performed in recent years in China due to lack of HLA identical siblings. Higher risk for developing acute graft-versus-host disease (aGVHD) has been seen in patients who receive haploidentical stem cell transplantation (haplo-SCT) compared with identical sibling SCT patients (Lu DP et al, Blood2006; 107: 3065). In a murine model, we have first demonstrated that the animals transplanted with three mixed bone marrow (A+B+C to A) were survived longer due to milder aGVHD compared with the mice transplanted with single allogeneic bone marrow (B to A). In current clinical trial, we examine whether cord blood (CB) as the third party cells could reduce aGVHD in haplo-SCT setting under the same principle investigator in two hospitals. Between January 2006 and April 2008, total 133 haplo-SCT patients with hematological malignancies were enrolled. The patients with advanced diseases were excluded. All patients received unmanipulated blood and marrow transplant after BUCy2 or CyTBI plus antithymocyte globulin (ATG, Genzyme 10mg/kg) as preconditioning. Cyclosporine, short-term Methotrexate, and Mycophenolate mofetil were employed for GVHD prophylaxis. Fifty-six patients received one unit of HLA 3–6/6 matched CB one day before SCT as CB group, and 77 cases did not receive CB as control group. The main clinical characteristics in both groups are comparable. All patients in both groups achieved full donor chimerism. Low levels of CB chimerism were detected in a few patients at early stage after SCT. No long-term CB engraftment was found. The cumulative incidences of grade II to IV aGVHD were 16.4% versus 38.4% (p=0.008) in CB group, control group respectively. The cumulative incidences of grades III to IV aGVHD for CB group versus control group were 9.2% versus 22.4% with p=0.043. The incidences of 100-day treatment-related mortality were 1.8% versus 10.4% (p=0.053) for CB group, control group respectively. Our preliminary clinical study has shown that CB as the third party cells could significantly reduce the incidences of aGVHD, especially for severe aGVHD, and also treatment-related mortality in haplo-SCT. The mechanism of this strategy need to be further investigated.

Invasive aspergillosis has high mortality in patients with hematological malignancies after chemotherapy or allogeneic stem cell transplantation (SCT) due to usually low response rate to current antifungal agents and relatively poor tolerance to the toxicities of the antifungal medicines. In this pilot clinical study, we explore to manage the patients with refractory invasive aspergillosis with aspergillus-specific cytotoxic T lymphocytes (aspergillus CTL). Ten patients who were either no longer response to antifungal medicines or have to stop antifungal agents due to severe toxicities were included in this clinical study. Diagnosis are ALL (2 cases), AML (5 cases), and CML (3 cases). The median age was 37 (10 to 62) years old. Seven of them received allogeneic SCT (unrelated 3 cases, haploidentical 3 cases, identical sibling 1 case), and 3 patients were after chemotherapy. Aspergillus CTLs were either donor origin (5 cases) or patient origin (5 cases) if original donor was not available. Patients received 1 to 3 infusions of aspergillus CTLs. The median CTLs infused were 3.61 (0.28–40.00) x 105/kg. No infusion-related reactions were noted. No pre-existed graft-versus-host disease deteriorated. With aspergillus CTL therapy, one of them (10%) achieved complete response, 6 patients (60%) improved significantly, 1 case was stable, and 2 patients became worse. Our preliminary data has shown that aspergillus-specific cytotoxic T lymphocyte is a safe and promising mean to manage refractory invasive aspergillosis in patients with hematological malignancies.

Leukemia relapse after allogeneic hematopoietic stem cell transplantation (HSCT) is one of the main obstacles for survival. Immunosuppressant withdraw, chemotherapy, and donor lymphocyte infusion(DLI) are usually employed for management of recurrence after HSCT, but some patients have poor response to above therapy or have contraindications for DLI due to relapse at early stage after transplant or with active graft-versus-host disease (GVHD). Dendritic cell-primed cytokine-induced killer cells (DC-CIK) have been successfully applied in treatment of minimal residual disease (MRD) in our center for 12 years. In present pilot clinical study, we explore to manage early leukemia relapse after HSCT with donor’s DC-CIK in appropriate patients. The patients who relapsed in hematological (5–20% blasts in BM) or molecular or immunological (MRD>0.1% by flow cytometry) with at least one of the following criteria were included in this clinical trial. 1. No response to DLI; 2. Relapsed within 60 days after HSCT; 3. Relapsed with active GVHD. Total 18 patients (male 9, female 9) with median age 26 (4 to 42) years were eligible to this clinical study. The diagnosis included acute myeloid leukemia (AML 13), acute lymphoblastic leukemia (ALL 4) and chronic myeloid leukemia (CML 1) who failed to reach molecular remission with imatinib before transplant. The types of donor were HLA identical sibling (11), haploidentical family member (5), and unrelated donor (2). Six of 18 patients had either molecular or immunological recurrence, while 12 of 18 cases relapsed hematologically. The median cell dosage of DC-CIK infused was 2.34×109 (0.2–44×109). With DC-CIK treatment, overall 11 of 18 (61.1%) patients achieved complete remission (CR, molecular or immunological or hematological based on the disease status before DC-CIK). Among 6 cases in molecular or immunological recurrence, five of them (83.3%) obtained CR, while 12 patients with early hematological relapse, 6 of them (50%) returned to CR. Seven of 13 patients with AML, and all 4 cases with ALL responded to DC-CIK treatment, while a patient with CML had no therapeutic benefit from it. Among 7 cases without response to DC-CIK, one patient with CML achieved molecular remission with high-dose Imatinib, 1 case obtained CR after DLI later on, 1 patient survived with primary disease so far, and the remaining 4 patients died from leukemia recurrence. In 11 cases who responded to DC-CIK, 10 of them survived with median 359(164 to 1233) days. One patient died from transplant-related complications. Four patients developed GVHD after DC-CIK infusion and controlled completely with Cyclosporin A and Methylprednisolone. Our encouraging results indicate that Donor’s DC-CIK is a safe and effective therapeutic option in management of early leukemia recurrence after allogeneic HSCT, especially for the patients who fail to or ineligible to current standard practice.