| Background:Allogeneic hematopoietic stem cell transplantation(HSCT)is an important way for curing acute leukemia in adults.The outcome of the treatment is usually determined by two critical factors:relapse-related mortality and transplantation-related morbidity or mortality(TRM).Despite recent progress,it is difficult to mitigate one cause of mortality while without compromising the other.Consequently,HSCT cannot be effectively evaluated fully by focusing on TRM or relapse alone.To address this problem,a novel composite outcome of GVHD-free/relapse-free survival(GRFS)is proposed,in which the endpoint events included grade 3-4 acute GVHD(aGVHD),systemic treatment-requiring chronic GVHD(cGVHD),relapse,and death.To this end,GRFS represents real recovery without ongoing morbidity.To further understand the clinical factors which can effectively impact on GRFS in patients with acute leukemia after HSCT,we retrospectively reviewed 312 adult patients with acute leukemia treated with HSCT between 2008 and 2014 in our institution,and further investigated overall GRFS,disease-free survival(DFS),and overall survival(OS)at 1 and 2 years after HSCT.Through this way,we aim to identify reliable prognostic factors which can effectively predict the outcome of acute leukemia patients treated with HSCT in order to optimize therapeutic avenues.MethodsA consecutive series of 312 adult patients(age≥18 years)with acute leukemia receiving allogeneic HSCT between Mar 2008 and October 2014 in our institution was enrolled in this retrospective investigation.All living patients had been routinely followed until October 2016.Clinical data including gender,age,donor type,diagnosis of disease,status of disease,GVHD,conditioning regimen,and other clinical characteristics and complications were gathered.Advanced leukemia was defined as disease status of NR before HSCT was performed.For GRFS events,all data were considered as the first posttransplant event during 12 and 24 months.The impact of each clinical factor on 1-and 2-year GRFS was investigated.Donors’ type included matched sibling donor(MSD),matched unrelated donor(MUD),and haploidentical-related donor(HRD).Five approaches of myeloablative conditioning regimens were adopted,including Bu(busulfan)+Flu(fludarabine),Bu+Cy(cyclophosphamide),TBI(total body irradiation)+ Cy,TBI + Cy+etoposide(intensified myeloablative conditioning),and Flu + cytarabine + TBI + Cy+etoposide(sequential intensified conditioning).CsA(Cyclosporine A)or CsA + MTX(methotrexate)was adopted for patients treated with HLA-matched sibling donor transplants;CsA + MTX + ATG(antithymocyte globulin)with or without mycophenolate was used in patients treated with unrelated donor or haploidentical transplants.Patients with advanced leukemia or whose minimal residual disease(MRD)were detected after transplantation received DLI in the circumstance that they did not suffer from aGVHD of grade 2 or above.For preventing relapse,donor lymphocytes mobilized by G-CSF were infused once to part of patients with advanced leukemia regardless of MRD and were then administered according to their GVHD and MRD status.DLI would be discontinued under the condition that patients suffered from GVHD.Life-table method was used to evaluate 1-and 2-year posttransplant GRFS,OS and DFS.The impact of each clinical factor on 1-and 2-year GRFS was estimated by Kaplan-Meier method in univariate analysis,and differences between the survive curves were evaluated by using log-rank tests.The independent effect of each factor on 1-and 2-year GRFS was examined by using Cox proportional hazard models in multivariate analysis.The subgroup of patients with advanced leukemia was analyzed similarly.The level with statistical significance was fixed at p<0.05.The SPSS 18.0 software(SPSS,Chicago,USA)was used for statistical analysis.Results:1.General characteristics:We retrospectively reviewed 312 patients receiving allogeneic HSCT between Mar 2008 and October 2014.The median age at HSCT was 30 years(range 18～61 years).Donor types included haploidentical related donors(24.0%),HLA-matched sibling donors(44.6%)and unrelated donors(31.4%).The survival rate of 1-and 2-year GRFS was 54.8%and 51.5%,as compared with DFS(1-year:66.0%;2-year:62.5%)and OS(1-year:70.1%;2-year:65.0%),respectively.2.Clinical prognostic factors for GRFS:In multivariate analysis,the impacts of prognostic factors including age of donor and recipient,disease status,and diagnosis were remained significant for 1-and 2-year GRFS.Patients in NR before HSCT had significantly inferior GRFS compared with those in CR(HR 3.061,95%CI:2.043-4.587,p<0.001 for 1-year GRFS;HR 3.173,95%CI:1.233-2.411,p<0.001 for 2-year GRPS).acute biphenotypic leukemia(ABL)patients showed a nearly 3-fold worsening of GRFS(HR 2.957,95%CI:1.195-7.314,p = 0.019 for 1-year GRFS;HR 3.276,95%CI:1.328-9.079,p=0.010 for 2-year GRPS)compared with patients with blast phase of chronic myeloid leukemia(CM-BP).Conditioning intensity did not show an independent association with 1-nor 2-year GRFS.3.Subgroup Analysis for advanced leukemia:The subgroup of patients with advanced leukemia(n=79)was analyzed.The survival rate of 1-and 2-year GRFS of the subgroup was 25.3%and 23.8%,as compared with DFS(1-year:36.7%;2-year:33.9%%)and OS(1-year:40.5;2-year:34.9%),respectively.In multiple regression analysis,the prognostic factors including donor type(p=0.022),diagnosis(p=0.023)and DLI(HR 0.328,p=0.001)were remained significant.Conclusions:1.Our study revealed that GRFS was significantly impacted by clinical factors including age of recipients and donors,diagnosis,disease stage in the whole cohort,and HRD transplant resulted in comparable GRFS to MSD or MUD transplant.2.In the subgroup of advanced leukemia patients,prophylactic DLI contributed to better GRFS.A superior outcome of GRFS was achieved in HRD transplant compared with MSD or MUD transplant,suggesting HRD transplant may associated with a stronger graft-versus-leukemia effect while without a significantly raised risk of GVHD. |
PDF Full Text Request