Introduction: High relapse rate and transplant-related mortality (TRM) are the major obstacles for the second allogeneic hematopoietic stem cell transplantation (allo-HSCT). Reduced-toxicity conditioning (RTC) can decrease TRM and healthier donor may provide stronger anti-leukemia and anti-infection effects.Objective:In current clinical study, the safety and efficacy of the second allo-HSCT with RTC and donor change were examined.Methods: Between April 2018 and June 2021, total 48 patients with hematological malignancies (B-ALL 26, T-ALL/LBL 4,AML16,MDS 2) who failed to the first allo-HSCT(relapsed 47, secondary rejection 1) and underwent the second allo-HSCT in our hospital were enrolled. The median age was 25(5-55) years old. Male to female was 25:23. Before the second allo-HSCT, 35 (72.9%) patients were in complete remission (CR) (minimal residual disease (MRD) negative in 28, MRD positive in 7), and 13 (27.1%) cases (B-ALL7,T-ALL/LBL2,AML3,MDS1) in non-remission(NR). The median interval between twoallo-HSCTs was 19.5 (8-77) months. For better donor selection, the functions of hematopoiesis and immune (subsets of lymphocyte, immune globulin,and granzyme and perforin of NK cell) as well as hematological and immunological hereditary predisposition gene variants for potential donors were tested. The types of the first allo-HSCT included haploidentical (n=33), unrelated (n=8), identical sibling (n=5), and unrelated cord blood (n=2). All donors were changed for the second allo-HSCT(haploidentical 35, unrelated 13)except one recipient no other donor available. RTC regimens were mainly total body irradiation(TBI)/fludarabine (FLU)-based(n=39) or busulfan (BU)/FLU-based (n=7).Two patients were with total marrow irradiation (TMI)/FLU-basedregimen. For TBI/FLU regimen, fractionated TBI (8Gy in 32, 10Gy in 7), cytarabine 1-2g/m 2 for 3 days,FLU 30mg/m 2 for 5 days, Me-CCNU 250mg/m 2 for 1 day, and ATG were used. For BU/FLU regimen, BU 0.8mg/kg q6h for 3 days, cytarabine 1-2g/m 2 for 3 days,FLU 30mg/m 2 for 5 days, Me-CCNU 250mg/m 2 for 1 day, and ATG were used. For TMI/FLU regimen, the same conditioning as TBI/FLU was used except fractionated TMI 10-12Gy instead of TBI. Decitabine (20mg/m 2, 3 days) or etoposide (200mg/m 2, 2 days) were administrated in some patients in NR or MRD positive. Cyclosporine, mycophenolate mofetil and short-term methotrexate were employed for graft-versus-host disease (GVHD) prophylaxis. Some patients received maintenance therapy post-transplant with targeted medicine based on their fusion genes or gene mutations.Results: ALL patients became full donor chimerism. The median time for neutrophil and platelet recovery was 15 (11-22) days and 16 (10-240) days. The incidences of grade II-IV acute GVHD (aGVHD) and chronic GVHD(cGVHD) were 18.8%, 47.5% (limited 25%, extensive 22.5%), respectively. The incidences of CMV and EBV reactivation were 31.3% and 6.3%. No severe hemorrhagic cystitis occurred. The median follow-up was 13(1-38) months. One-year disease-free survival (DFS) and overall survival (OS) of all patients were 68.7% (95% CI:0.511;0.811) and 78.9% (95% CI:0.62;0.889) . Nine patients died(relapse 7,GVHD and infection 2).The relapse rate was 22.9%.TRM was only 4.2%. NR before transplant and TP53 mutation were high risk for disease recurrence post-HSCT. Relapse rates were 53.8% vs. 11.4% in NR and CR patients, and 50% vs. 17.5% in the patients with or without TP53 mutation. Disease status before transplant was key impact factor for survival after the second allo-HSCT. One-year OS in CR and NR settings were81.4 % and 70%, respectively, p=0.13. One-year DFS in CR and NR settingswere79.6%, and 41.9%, respectively, p=0.005. No significant difference of DFS was seen in haploidentical and unrelated transplants(68.1% vs.71.8%, p=0.627).Conclusions: With our strategy of RTC regimens and donor change, excellent outcomes of the second allo-HSCT in hematological malignancies have been achieved. One-year DFS and OS are 68.7% and 78.9%. Both TRM and relapse rate are low (4.2%, 22.9%). The most important factor on prognosis of the second allo-HSCT is disease status before transplant but not donor type. TP53 mutation also has negative impact on DFS after the second allo-HSCT. DisclosuresNo relevant conflicts of interest to declare.