Font Size: a A A

Relapse Of AML After Allo-HSCT:Risk Factors And Mechanisms

Posted on:2020-05-16Degree:DoctorType:Dissertation
Country:ChinaCandidate:Z W WangFull Text:PDF
GTID:1364330575961595Subject:Internal medicine
Abstract/Summary:PDF Full Text Request
Acute myeloid leukemia?AML?is one of the most common hematological malignancies.And allogeneic hematopoietic stem cell transplantation?allo-HSCT?is a curative treatment due to the immune-mediated graft-versus-leukemia?GVL?effect,thus reducing the risk of relapse compared to conventional chemotherapy and improving outcomes[1].Advances in the transplantation technology have reduced the complications and transplant-related mortality in the past two decades,with few inroads in the treatment or prevention of relapse[2].As a result,the prognosis is very poor for patients relapsing after transplantation.Further analysis of the factors related to relapse after transplantation and post-relapse overall survival?prOS?has significance for identification of high-risk patients and selection of effective treatment.Although patients receiving HSCT have lower risk of relapse as a result of the GVL effect,relapse is still the main cause of death after transplantation.It is believed that relapse maybe initiated from a small part of leukemia cells which is less sensitive for chemotherapy.Lapidot et al transplanted CD34+CD38-cells isolated from AML patients into severe combined immune-deficient mice to establish human leukemia for the first time.These cells have the same immunophenotyping and biological characteristics as normal hematopoietic stem cells and was named"leukemia initiating cells?LIC?"[3].Many studies have shown that LIC involves in leukemia initiation,therapy resistance and relapse,and that the frequency of LIC has crucial influence on prognosis for AML patients[4][5][6][7][8].Therefore,exploration of relationship between LIC and relapse after allo-HSCT has great significance for finding a therapeutic target.Based on the above background,we firstly analyzed prognostic factors for relapse and prOS in AML patients receiving HSCT.And we first used mouse acute myeloid leukemia model to study the relationship between LIC and leukemia recurrence after allogenic transplantation.We found that expression of C-Kit regulated the cell cycle entrance through PD-L1/STAT3 signaling axis which meant that higher expression of C-Kit on LIC induced expression of PD-L1 via the JAK/STAT3 signaling pathway.As a result,LIC exhibited resistance to the proliferation arrest mediated by GVL effect,which may result in more entrance of active cell-cycle states and initiation of relapse.Therapy with tyrosine kinase inhibitor could reduce the expression of PD-L1 and prolong the survival of leukemia mice.Part 1:Prognostic factors for relapse and post-relapse overall survivalMethods:A total of 218 AML patients who received their first allo-HSCT between March 2001and August 2015 in Changhai or Ruijin Hospital and surviving beyond 100 days were enrolled in this retrospective study.We analyzed the relationship between relapse or post-relapse overall survival and factors including age,gender,cytogenetics and molecular stratification,donor-patient sex matching,donor type,disease status before transplantation,conditioning regimen,acute graft-versus-host disease?aGVHD?,chronic graft-versus-host disease?cGVHD?and the number of mononuclear cells?MNC?in the grafts.Cox proportional hazard models were used to estimate hazard ratios with 95%confidence interval for prognostic factors.OS,prOS and event-free survival?EFS?were calculated by the Kaplan-Meier method and were compared using the log-rank test.The cumulative incidences of relapse rate was estimated using competing risk analysis with non-relapse mortality as competing risks.Wilcoxon's rank sum test was used to compare continuous variables,and the chi-squared test was used to compare categorical variables.A P value of0.05 or less was considered statistically significant.Results:1.We retrospectively analyzed 218 patients including 123 males and 95 females.The median age was 36?14-62?years,and the median follow-up for survivors was 21.5 months?range,3.4 months to 179.6 months?.The estimated 3-year OS was 62.0%?95%CI 55.0%-69.0%?.The 3-year EFS was 54.9%?95%CI 47.9%-61.9%?and cumulative incidence of relapse was 30.2%?95%CI 23.7%-36.7%?.Sixty-seven?31.02%?patients developed aGVHD and 103?47.69%?developed cGVHD after transplantation.Sixty-two patients relapsed with a median time of 5.23?2.17-42.2?months.The 1-year prOS was 33.7%?95%CI 20.6%-46.9%?and the 2-year prOS was 16.0%?95%CI 4.9%-27.1%?.2.Prognostic factors for relapse:In the univariate analysis,disease status before transplantation,donor type,donor-patient sex matching,cGVHD and the number of MNC were associated with cumulative incidence of relapse after transplantation.In the multivariate analysis,disease status before transplantation[CR2/3 HR 4.62?2.37-8.99?,P<0.001;NR HR 3.31?1.75-6.24?,P<0.001]and occurrence of cGVHD[limited HR 0.40?0.21-0.77?,P=0.006;extensive HR 0.33?0.13-0.82?,P=0.018]were associated with relapse.3.Prognostic factors for prOS:In the univariate analysis,disease status before transplantation,cytogenetics and molecular stratification,donor type,donor-patient sex matching,conditioning regimen,cGVHD and duration of remission were associated with prOS.In the multivariate analysis,cGVHD[HR 0.23?0.10-0.52?,P<0.001]and duration of remission[more than 2 years HR 0.11?0.03-0.43?,P=0.001]were associated with prOS.Conclusion:1.Late disease status before transplantation was risk factor for relapse2.Occurrence of cGVHD after transplantation was not only associated with lower risk of relapse but also associated with higher prOS3.Longer duration of remission after transplantation was associated with higher prOSPart 2.C-Kithigh leukemia cells isolation and characterizationMethods and materials:1.Mice and cellsBALB/C mice were used as leukemia host.Mouse AML cell line with AML1-ETO fusion gene and C-Kit N882K mutation was used.2.C-Kithigh cells isolation and characterizationSorting for cells identified as Lin-GFP+Sca-1-C-Kithigh and as GFP+Sca-1-C-Kitlow.One hundred thousand cells were cultured with M3234 culture medium and 1×104 cells were seeded in each well of 96-well plates.After 9-10 days,the number of colonies generated by C-Kithigh cells or C-Kitlow cells were counted and compare.Ten?a hundred or a thousand C-Kithigh cells or C-Kitlow cells isolated by flow cytometry were transplanted to the BALB/C mice.Compare the morbidity and survival between mice receiving different type of leukemia cells.3.Apoptosis detectionBone marrow cells?1×106?collected from leukemia mice at different time points were stained with Lin,Sca-1,C-Kit,Annexin V/PI?Annexin V/7-AAD?and then analyzed by flow cytometry.Protein extracted form Lin-GFP+Sca-1-C-Kithigh cells or GFP+Sca-1-C-Kitlow cells was identified and quantified with Western Blot using anti-Caspase3 and anti-Bcl2 antibodies.4.Cell cycle analysisBone marrow cells?1×106?collected from leukemia mice at different times were stained with Lin,Sca-1,C-Kit.After fixation with Lysing Solution and permeabilization with TF Fix/Perm Buffer,the cells were stained with Ki67 and Hoechst33342 and then analyzed by flow cytometry.G0 phase was characterized as Ki67-Hoechst-,G1 phase as Ki67+Hoechst-and S/G2/M phase as double positive.5.Quantitative reverse transcription-polymerase chain reactionRNA was extracted with TRIZol from C-Kithigh cells and C-Kitlow cells sorted by flow cytometry.Reverse transcription was performed using RR036A Takara PrimeScript?RT Master Mix?Perfect Real Time?and PCR was done with SYBR Green qPCR Master Mix.6.Detection of costimulatory molecules on CD3+T cells and CD3+CD4+T cellsPeripheral blood was collected from mice receiving C-Kithigh cells or C-Kitlow cells on day 0,7,14,21 after transplantation.Blood samples were stained with CD3,CD4,CD28,ICOS,PD-1 and CTLA4 after lysis of red blood cells.Then analyzed with flow cytometry.7.Statistical analysis:Continuous data were presented with mean±SEM.Difference of mean between two groups was compared with Student test and between multiple groups with one-way ANOVA.Categorical viarables were presented with number?percentage?.Survival were calculated by Kaplan-Meier method and were compared using the log-rank test.A P value of 0.05 or less was considered statistically significant.We divided the progression of leukemia into three stages according to the percentage of GFP+cells in the bone marrow:less than 10%of GFP+cells are identified as early stage,greater than 10%and less than 30%of GFP+cells are identified as the middle stage,greater than 30%of GFP+cells represent the late stage of disease.results:1.C-Kithigh cells demonstrated feature of more colonies formation and higher leukemia mortalityThe number of colonies generated by C-Kithigh cells was greater than that of C-Kitlowow cells?14.3±8.1/1×104 cells and 2.9±1.8/1×104 cells,P<0.001?.None mice that received ten C-Kithigh or C-Kitlow cells developed leukemia.In the group of 100 cells/mouse,the morbidity was 70.59%?12/17?for mice receiving C-Kithigh cells compared with 33.33%?6/18?for mice receiving C-Kitlow cells.The median survival time was 29?22-48?days for mice receiving Kithigh cells compared with 35?20-70?days for mice receiving Kitlow cells?P=0.011?.In the group of 1000 cells/mouse,the morbidity was 100%for mice receiving C-Kithigh cells compared with 77.27%?17/22?for mice receiving C-Kitlow cells.The median survival time was 27?22-41?days for mice receiving Kithigh cells compared with 31?23-53?days for mice receiving Kitlow cells?P=0.002?.The mice that developed leukemia eventually all died of leukemia infiltration.2.Percentage of apoptosis increased for both type of leukemia cells with the progression of disease.There was no difference of the ratio of apoptosis between C-Kithighigh and C-Kitlow cells.PI+AnnexinV+cells cannot be detected in both type of leukemia cells.In all stages of disease,the apoptosis proportion were the same between C-Kithigh and C-Kitlow cells?early/middle/late stage,P values were 0.479/0.673/0.587?,and the proportion of apoptosis for both groups increased with the progression of the disease.Western Blot analysis demonstrated that the expression of Bcl2 and Caspase3 were similar between C-Kithigh and C-Kitlow cells3.More C-Kithigh cells were in active cell-cycle states than C-Kitlow cellsMore cells entered G0 phases in both of C-Kithigh and C-Kitlow cells with disease progression.The proportion of C-Kithigh cells in quiescent status was lower than C-Kitlowow cells?P value for early/middle/late stages were<0.001/0.005/0.029?.The proportion of phases G1/S/G2/M and S/G2/M decreased as the disease progressed in both type of leukemia cells.And more Kithigh cells were in active cell-cycle states than Kitlow cells all the time?P value for G1/S/G2/M phase were 0.002/0.034/0.029;P value for SG2/M phase were 0.038/0.022/0.029?.4.Expression of genes regulating cell cycle and apoptosis in C-Kithigh and C-Kitlowow cellsCyclinD2 gene and CyclinD3 gene were upregulated in C-Kithigh cells compared with C-Kitlow cells?P values were 0.048 and 0.049?.P16 gene and P57 gene were down-regulated in C-Kithigh cells?P values were 0.036 and 0.050?.The expression of Bcl-2 in Kithigh cells was higher than that in Kitlow cells?P=0.047?.5.Expression of co-stimulatory molecules on the surface of T cells were similar between mice receiving C-Kithigh cells or C-Kitlow cellsThere was no significant difference of expression of CD28,ICOS,PD-1,and CTLA4 on T cells in peripheral blood collected from mice receiving Kithigh cells and Kitlow cells at different time points.However,the proportion of CD3+CD4+CD28+T cells declined with leukemia progress.The proportion of CD3+CD4+CD28+T cells from mice receiving Kithighigh cells decreased from 53.43%to 36.73%?P=0.050?and from 58.60%to 31.58%?P=0.018?in mice receiving Kitlow cells.The expression of PD-1on CD3+CD4+T cells were upregulated with disease progression from 2.97%to 7.40%?P<0.001?in mice receiving Kithigh cells and from 3.71%to 7.88%?P<0.001?in mice receiving Kitlow cells.The expression of PD-1 on CD3+T cells were upregulated from 3.64%to 8.51%?P=0.001?in mice receiving Kithigh cells and from 4.20%to 9.75%?P=0.001?in mice receiving Kitlow cells.Conclusion1.Cells with immunophenotype Lin-GFP+Sca-1-C-Kithigh had stronger ability of colony formation and mouse leukemia reconstitution than cells labelled as GFP+Sca-1-C-Kitlowow which represent the characterization of leukemia-initiating cells.2.With the progression of disease,the apoptosis of both types of leukemia cells increased and more cells entry quiescent status.But the proportion of active cell-cycle states were always higher in C-Kithigh cells.3.There was no significant difference of co-stimulatory moleculars expressing on T cells in peripheral blood collected from mice receiving Kithigh and Kitlow cells at different time points.But CD3+CD4+CD28+T cells decreased and the expression of PD-1 increased with disease progression in both types of host mice.Patr 3.Mechanisms of relationship between relapse after allo-HSCT and C-Kithigh cellsMaterials and methods1.Leukemia mice and cell line used were the same as that in part 2.C57BL/6?H2Kb?mice aged 8-9 weeks were selected as the donor for transplantation and BALB/C?H2Kd?mice aged 10-11 weeks as the recipient.2.GVL mice modelRecipients received a total does of 8Gy TBI 7 hours before transplantation.Recipients were grouped according to the grafts they received.The BL mice model received 5 million T cell depleted bone marrow cells?TCD-BM?and 10 thousand leukemia cells while the GVL mice model received 5 million TCD-BM cells,2.5 million spleen cells and 10thousand leukemia cells3.Mechanisms of relationship between C-Kithigh cells and post-transplant recurrenceSorting for cells identified as Lin-GFP+Sca-1-C-Kithigh and as GFP+Sca-1-C-Kitlowow by flow cytometry following the protocol described in part 2.Recipients received a total does of 8Gy TBI 7 hours before transplantation and divided into four groups according to the grafts received.Recording the morbidity and mortality of these recipients.Kithigh-BL group:TCD-BM?5×106?,Kithigh cell?1×104?Kitlow-BL group:TCD-BM?5×106?,Kitlow cell?1×104?Kithigh-GVL group:TCD-BM?5×106?,spleen cell?2.5×106?,Kithigh cell?1×104?Kitlow-GVL group:TCD-BM?5×106?,spleen cell?2.5×106?,Kitlow cell?1×104?Apoptosis and cell-cycle analysis were performed at different time points using bone marrow cells collected from BL or GVL recipients that received unsorted leukemia cells.The protocols were listed in Part2.Meanwhile,expression of molecules on Kithigh andKitlow cells including PD-L1 and ICOSL were analyzed using the same bone marrow samples by flow cytometry.4.Therapy for leukemia with dasatinib and anti-PD-L1 antibodyHost BALB/C mice received 10 thousand leukemia cells.Anti-PD-L1 antibody was administrated by intraperitoneal injection at a dose of 400ug daily on days 1,5,9,13 and17 after transplantation.Meanwhile,dasatinib was administrated by gavage at a dose of20mg/kg daily,begining at day 1 after transplantation and continuing until the death of control leukemic mice.The control groups included mice receiving carboxymethyl cellulose sodium?CMC-Na?plus PBS,receiving CMC-Na plus anti-PD-L1 antibody and receiving dasatinib plus PBS.5.Expression of phosphorylated STAT3 and phosphorylated MAPK?Erk?detected by Western blotExtract protein from cells sorted as Lin-GFP+Sca-1-C-Kithigh,GFP+Sca-1-C-Kitlow,GFP+PD-L1high and GFP+PD-L1low.Compare the expression of phosphorylated STAT3 and phosphorylated Erk by a Simple Western system?Peggy Sue?.6.Statistical analysis:Statistical analysis was the same as Part2.Results1.Mice in GVL group survived longer than those in BL group,and that in Kithigh-GVL group survived shorter than in Kitlow-GVL group.The morbidity of leukemia after transplantation was 100%with a median survival of19?15-20?days in the BL group and 100%with a median of 39?30-52?days in the GVL group?P<0.001?.There was no significant difference in the survival of mice between Kithigh-BL group[20?19-21?days]and Kitlow-BL group[21?19-24?days]?P=0.076?.However,mice in Kithigh-GVL group survived significantly shorter than that of Kitlow-GVL group[32?24-54?days versus 51?31-58?days,P=0.008].2.Proportion of apoptosis was higher in GVL mice compared with BL mice,but it was similar between C-Kithigh cells and C-Kitlow cellsOnly PI-AnnexinV+cells could be detected in all mice model.The proportions of apoptosis for Kithigh and Kitlow cells were the same both in BL group and in GVL group?P value for early/middle/late stage in BL group:0.674/0.174/0.212;P value for early/middle/late stage in GVL group:0.343/0.877/0.561?.There was more apoptosis in Kithigh cells from GVL group compared with Kithigh cells from BL group in the mediate and late stage of leukemia progression?P value for early/middle/late stage:0.486/0.025/0.011?.Similarly,there was more apoptosis in Kitlow cells from GVL group compared with those from BL group in the mediate stage of leukemia progression?P value for early/middle/late stage:0.943/0.012/0.200?.3.Cell-cycle analysis for C-Kithigh and C-Kitlow cells in BL and GVL groupIn the BL group,there were more C-kitlow cells in G0 phase?P value for early/middle/late stages:0.026/0.028/<0.001?and fewer in G1/S/G2/M phases?P value:0.133/0.032/<0.001?and S/G2/M phases?P value:<0.001/0.002/<0.001?compared with C-Kithigh cells.It was the same in the GVL group,there were more C-kitlow cells in G0phase?P value:0.029/0.004/0.003?and fewer in G1/S/G2/M phase?P value:0.029/0.004/0.003?and S/G2/M phase?P value:<0.001/0.002/<0.001?compared with C-Kithigh cells.Comparing cell-cycle distribution in C-Kithigh cells collected from BL mice and GVL mice at early stage of disease.More Kithigh cells in BL group were in G2/S/M phase than those from GVL group?P=0.016?,while the proportion of G1/S/G2/M phase and G0 phase were similar between Kithigh cells from the two groups.Comparing cell-cycle distribution in C-Kitlow cells collected from BL mice and GVL mice at early stage of disease.More Kitlow cells in BL group were in active cell-cycle states than those from GVL group?P value for G1/S/G2/M phase were 0.028;P value for S/G2/M phase were 0.002;P value for G0 phase were 0.002?.4.The expression of PD-L1 on C-Kithigh cells was significantly higher than C-Kitlowow cells in all stages of disease.In BL group,the expression of PD-L1 on Kithigh cells was higher than Kitlow cells at all stages of disease?P value for early/middle/late stage:0.017/0.038/0.038?.And it was the same in GVL group?P value for early/middle/late stage:0.015/0.002/0.001?.The comparison of moleculars expressed on Kithigh cells collected from BL and GVL group showed no significant difference.Similarly,the expression of PD-L1 and ICOSL on the surface of Kitlow cells from BL and GVL group were also not different significantly.5.The survival time significantly prolonged in mice treated with dasatinibThe median time of survival for mice receiving CMC-Na plus PBS was 23?21-24?days and it was 27?26-28?days for mice receiving dasatinib plus PBS,24?22-26?days for mice receiving CMC-Na plus anti-PD-L1 antibody and 28?26-31?days for mice receiving dasatinib plus anti-PD-L1 antibody.Mice that treated with dasatinib survived much longer than those treated with vehicles?P<0.001?while treated with dasatinib plus anti-PD-L1antibody survived a little longer than those treated with dasatinib only?P=0.027?.6.C-Kithigh cells upregulate the expression of PD-L1 via JAK/STAT3 signaling pathwayThe expression of phosphorylated STAT3?pSTAT3?in Lin-GFP+Sca1-C-Kithighigh cells was much higher than that in GFP+Sca1-C-Kitlowow cells.The expression of pSTAT3 in GFP+PD-L1highigh cells was much higher than that in GFP+PD-L1lowow cells.After treatment of dasatinib,expression PD-L1 on C-Kithigh cells was down regulated from 49.8%±13.5%to26.8%±4.1%?P=0.047?in the early stage of disease,from 40.9%±1.6%to 27.3%±4.6%?P=0.009?in the middle stage and from 36.3%±9.7%to 21.6%±7.7%?P=0.110?in the late stage.Conclusion1.C-Kithigh cells induced the expression of PD-L1 via JAK/STAT3 signaling pathway which resulted in a resistance to the proliferation arrest mediated by GVL effect,so that more of Kithigh cells could enter active cell-cycle states and initiate relapse.2.Treatment with dasatinib reduced the expression of PD-L1 on C-Kithigh cells and prolonged the survival of leukemia mice.
Keywords/Search Tags:AML, allo-HSCT, relapse, GVL effect, LIC
PDF Full Text Request
Related items