Analysis of relapse after transplantation in acute leukemia: A comparative on second allogeneic hematopoietic cell transplantation and donor lymphocyte infusions
2018; Elsevier BV; Volume: 62; Linguagem: Inglês
10.1016/j.exphem.2018.03.002
ISSN1873-2399
AutoresGuillermo Ortí, Jaime Sanz, Irene García‐Cadenas, Isabel Sánchez‐Ortega, Laura Alonso, María José Jiménez, Luisa Sisinni, Carmen Azqueta, Olga Salamero, Isabel Badell, Christelle Ferrà, Cristina Díaz de Heredia, Rocío Parody, Miguel Á. Sanz, Jorge Sierra, José Luís Piñana, Sergi Querol, David Valcárcel,
Tópico(s)Acute Myeloid Leukemia Research
Resumo•In this cohort of acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) patients, second allogeneic hematopoietic cell transplantation (Allo-HCT) and donor lymphocyte infusion (DLI) showed comparable results.•The time interval from Allo-HCT to relapse is a strong predictor of outcomes for DLI and second Allo-HCT, a shorter time associated with lower overall survival (OS) and disease-free survival (DFS).•Within the DLI cohort, patients receiving T-cell depletion at Allo-HCT had higher OS, DFS, and lower relapse incidence. Relapse of acute leukemia (AL) after allogeneic hematopoietic cell transplantation (Allo-HCT) entails a dismal prognosis. In this scenario, donor lymphocyte infusions (DLIs) and second Allo-HCT are two major approaches. We compared outcomes of AL patients treated for relapse with DLI or second Allo-HCT after receiving debulking therapy. In total, 46 patients were included in the study; 30 (65%) had acute myeloid leukemia and 16 (35%) had acute lymphoblastic leukemia. The median age was 38 years (range 4–66). Twenty-seven patients received a second Allo-HCT and 19 patients received DLI. The median follow-up of the cohort was 273 days (range 9–7013). Overall survival (OS), disease-free survival (DFS), nonrelapse mortality, and cumulative incidence (CI) of relapse were calculated from DLI or second Allo-HCT date. In univariate analysis, second Allo-HCT was associated with higher OS (p = 0.021) and a trend to higher DFS (p = 0.097) and CI of relapse (p = 0.094) on univariate analysis. However, multivariate analysis showed comparable outcomes between DLI and second Allo-HCT, with the time interval to relapse before DLI or second Allo-HCT the only statistically significant factor with an impact on OS and DFS. Within the DLI cohort, T-cell-depleted Allo-HCT was associated with higher OS (p = 0.003) and DFS (p < 0.001) and lower CI of relapse (p = 0.002) than T-cell-replete Allo-HCT. Overall, in this cohort of AL patients, second Allo-HCT and DLI associated similar outcomes. As in other relapse studies, the length of remission (time to relapse) was identified as a factor with statistical impact on survival. Further studies are warranted. Relapse of acute leukemia (AL) after allogeneic hematopoietic cell transplantation (Allo-HCT) entails a dismal prognosis. In this scenario, donor lymphocyte infusions (DLIs) and second Allo-HCT are two major approaches. We compared outcomes of AL patients treated for relapse with DLI or second Allo-HCT after receiving debulking therapy. In total, 46 patients were included in the study; 30 (65%) had acute myeloid leukemia and 16 (35%) had acute lymphoblastic leukemia. The median age was 38 years (range 4–66). Twenty-seven patients received a second Allo-HCT and 19 patients received DLI. The median follow-up of the cohort was 273 days (range 9–7013). Overall survival (OS), disease-free survival (DFS), nonrelapse mortality, and cumulative incidence (CI) of relapse were calculated from DLI or second Allo-HCT date. In univariate analysis, second Allo-HCT was associated with higher OS (p = 0.021) and a trend to higher DFS (p = 0.097) and CI of relapse (p = 0.094) on univariate analysis. However, multivariate analysis showed comparable outcomes between DLI and second Allo-HCT, with the time interval to relapse before DLI or second Allo-HCT the only statistically significant factor with an impact on OS and DFS. Within the DLI cohort, T-cell-depleted Allo-HCT was associated with higher OS (p = 0.003) and DFS (p < 0.001) and lower CI of relapse (p = 0.002) than T-cell-replete Allo-HCT. Overall, in this cohort of AL patients, second Allo-HCT and DLI associated similar outcomes. As in other relapse studies, the length of remission (time to relapse) was identified as a factor with statistical impact on survival. Further studies are warranted. Relapse of acute leukemia (AL) after allogeneic hematopoietic cell transplantation (Allo-HCT) entails a dismal prognosis and its approach remains challenging. The relapse rate of acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) after transplantation vary between 30% and 50%, depending on a wide range of factors such as the transplantation platform [1Scott B.L. Pasquini M.C. Logan B.R. et al.Myeloablative versus reduced-intensity hematopoietic cell transplantation for acute myeloid leukemia and myelodysplastic syndromes.J Clin Oncol. 2017; 35: 1154-1161Crossref PubMed Scopus (352) Google Scholar], disease risk [2Armand P. Gibson C.J. Cutler C. et al.A disease risk index for patients undergoing allogeneic stem cell transplantation.Blood. 2012; 120: 905-913Crossref PubMed Scopus (261) Google Scholar], and the intensity of T-cell depletion (TCD) [3Baron F. Labopin M. Blaise D. et al.Impact of in vivo T-cell depletion on outcome of AML patients in first CR given peripheral blood stem cells and reduced-intensity conditioning allo-SCT from a HLA-identical sibling donor: a report from the Acute Leukemia Working Party of the European Group for Blood and Marrow Transplantation.Bone Marrow Transplant. 2014; 49: 389-396Crossref PubMed Scopus (75) Google Scholar], among others. Posttransplantation relapse is associated with poor survival rates [4Bachanova V. Verneris M.R. DeFor T. et al.Prolonged survival in adults with acute lymphoblastic leukemia after reduced-intensity conditioning with cord blood or sibling donor transplantation.Blood. 2009; 113: 2902-2905Crossref PubMed Scopus (84) Google Scholar, 5Schmid C. Labopin M. Nagler A. et al.Treatment, risk factors, and outcome of adults with relapsed AML after reduced intensity conditioning for allogeneic stem cell transplantation.Blood. 2012; 119: 1599-1606Crossref PubMed Scopus (199) Google Scholar, 6Devillier R. Crocchiolo R. Etienne A. et al.Outcome of relapse after allogeneic stem cell transplant in patients with acute myeloid leukemia.Leuk Lymphoma. 2013; 54: 1228-1234Crossref PubMed Scopus (33) Google Scholar]. In this scenario, reducing the tumor burden before further immunotherapy, whether via a second Allo-HCT or donor lymphocyte infusion (DLI), seems essential to improve overall survival (OS) and disease-free survival (DFS). Consistent with this, patients responding to debulking therapy before second Allo-HCT seem to have superior outcomes [7Bejanyan N. Weisdorf D.J. Logan B.R. et al.Survival of patients with acute myeloid leukemia relapsing after allogeneic hematopoietic cell transplantation: a center for international blood and marrow transplant research study.Biol Blood Marrow Transplant. 2015; 21: 454-459Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar, 8Orti G. Sanz J. Bermudez A. et al.Outcome of second allogeneic hematopoietic cell transplantation after relapse of myeloid malignancies following allogeneic hematopoietic cell transplantation: a retrospective cohort on behalf of the Grupo Español de Trasplante Hematopoyetico.Biol Blood Marrow Transplant. 2016; 22: 584-588Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar]. Further, responses have been reported to be better when DLI is received for low disease burden (cytogenetic relapse) compared with hematologic relapse [9Dazzi F. Szydio R.M. Craddock C. et al.Comparison of single-dose and escalating-dose regimens of donor lymphocyte infusion for relapse after allografting for chronic myeloid leukemia.Blood. 2000; 95: 67-71PubMed Google Scholar], strongly suggesting that the tumor burden or disease status is fundamental to predicting outcomes. Although no prospective comparative trials have been performed, data from retrospective studies suggest that relapsed patients treated only with chemotherapy seem to have worse outcomes compared with patients receiving chemotherapy plus immunotherapy (DLI or second Allo-HCT) [5Schmid C. Labopin M. Nagler A. et al.Treatment, risk factors, and outcome of adults with relapsed AML after reduced intensity conditioning for allogeneic stem cell transplantation.Blood. 2012; 119: 1599-1606Crossref PubMed Scopus (199) Google Scholar, 7Bejanyan N. Weisdorf D.J. Logan B.R. et al.Survival of patients with acute myeloid leukemia relapsing after allogeneic hematopoietic cell transplantation: a center for international blood and marrow transplant research study.Biol Blood Marrow Transplant. 2015; 21: 454-459Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar]. In this particular situation, data are scarce and physicians decide on an individualized basis according to factors such as the presence of active graft-versus-host disease (GvHD) at relapse or the patient's performance status (PS). Consistent with this, patients with poor PS would not generally be considered for a second Allo-HCT. Further, with active GvHD at relapse, DLIs are not usually offered given the risk of worsening due to GvHD. Conversely, the absence of GvHD at relapse and before relapse suggests the use of DLI. Therefore, the question of who benefits most from each approach remains elusive. We have performed a comparative of patients treated either with a second Allo-HCT or DLI for relapse. We focused on AL patients treated with salvage therapy for transplantation relapse before further enhancement of the graft-versus-leukemia effect. Herein, we present results for a cohort of relapsed AL patients treated to reduce the disease load, who subsequently received a second transplantation or DLI. Patients from five European Society for Blood and Marrow Transplantation centers were included consecutively, and data were collected retrospectively. Informed consent in accordance with the Declaration of Helsinki was obtained before transplantation. Inclusion criteria were as follows: (1) post-Allo-HCT relapse of AL (AML or ALL), (2) morphological remission or postchemotherapy aplasia with absence of blasts on bone marrow after salvage treatment, and (3) DLI or second Allo-HCT after salvage treatment. Inclusion of patients was consecutive, not including patients who had second Allo-HCT after receiving DLI and patients who were treated with DLI after a second Allo-HCT at last follow-up date. Patients who did not receive treatment for relapse were not included. The inclusion selection was done as per physician's choice individualizing each patient, although donor availability could have influenced this decision in some patients. Morphological remission was defined as a blast count of <5% on a bone marrow aspirate (BMA) performed at postchemotherapy peripheral blood recovery. Post-chemotherapy aplasia was defined as a significant absence of the three hematopoietic cell lines observed on a BMA sample to pancytopenia observed on a peripheral blood sample at any point after salvage therapy. Response and disease status were assessed by BMA. Data regarding immunophenotyping, karyotyping analysis, or minimal residual disease by molecular genetics at relapse were not available. DLI was defined as infusion of CD3+ cells from the donor at any time after salvage therapy with no further immunosuppressive therapy. Second Allo-HCT was defined the infusion CD34+ cells/kg after conditioning regimen followed by immunosuppressive therapy (IST). All patients received IST after infusion day and no CD34+ selected grafts were used in this study. For the statistical analysis, day zero was set as the second Allo-HCT or first DLI day and the analysis was performed from this point. The analysis of time to second Allo-HCT or time to DLI was performed as a dichotomic variable, dividing in two groups according above and below 6 months. For the descriptive results analysis, the clinical manifestations of acute and chronic GvHD were graded according to the Keystone 1994 consensus criteria [10Przepiorka D. Weisdorf D. Martin P. et al.1994 consensus conference on acute GVHD.Bone Marrow Transplant. 1995; 15: 825-828PubMed Google Scholar] and the historical criteria [11Atkinson K. Chronic graft-versus-host disease.Bone Marrow Transplant. 1990; 5: 69-82PubMed Google Scholar]. If the patients were diagnosed with GvHD beyond day 100, then GvHD was defined as chronic. The second Allo-HCT cohort was compared against the DLI cohort. A chi-squared test was performed to identify significant differences between groups. The following variables were considered for its prognostic value and assessed for OS, DFS, nonrelapse mortality (NRM), and the CI of relapse univariate analysis: second Allo-HCT or DLI, patient's age, disease, patient's and donor's gender, donor type, level of mismatch, TCD, myeloablative versus reduced intensity conditioning (RIC), source of stem cells, first Allo-HCT date, disease status at first HCT, number of mismatches, time to relapse, type of immunosuppressive therapy, and, only for the second Allo-HCT cohort, same versus different donor at second Allo-HCT. A subanalysis of the DLI cohort also included: total DLI dose, total first DLI dose, and DLI collection method (DLI cryopreserved vs. DLI obtained by lymphapheresis). A subanalysis of the second Allo-HCT cohort was performed according to the donor used (same vs. other). The probabilities of OS and DFS were calculated using Kaplan–Meier estimates. Cumulative incidence of NRM and CI of relapse were calculated to accommodate for competing risks and results were presented according to the Fine and Gray model. Log-rank and Breslow tests were used for univariate comparisons for all variables considered. Univariate analysis and multivariate analysis used the Cox proportional hazards regression model and the analysis of variance, respectively. For multivariate analysis, we included all independent covariates with p values < 0.1 in univariate analysis. The p value was set at < 0.05 for statistical significance. Statistical analyses were performed with the statistical package SPSS version 17 and R software version 3.4.1. From 1995 to 2017, 46 patients were in total included. Thirty patients (65%) were diagnosed with AML and16 (35%) were diagnosed with ALL. All lymphoid leukemias were B-cell ALL (B-ALL) except one, which was a pro-T-cell ALL. Four ALL patients carried the BCR/ABL rearrangement. Regarding the AML cohort, one patient had myeloid sarcoma and 10 of 19 evaluable AML patients carried poor prognostic cytogenetic and molecular features (seven complex karyotype, two poor prognosis karyotype abnormality, and one FLT3-ITD mutation). One additional AML patient had inv16. Twenty-seven patients (59%) underwent a second Allo-HCT, and 19 patients (41%) received DLI. The median patient age was 38 years (range 4–66 years): 28 years (range 4–54 years) and 42 years (range 22–66 years) for the second Allo-HCT and the DLI cohorts, respectively (p = 0.015). Median time to relapse was 285 days (range 35–3956 days), the median time to relapse of the second Allo-HCT was 378 days (range 61–1508 days), and the median time to relapse for the DLI cohort was 152 days (range 35–3956 days) (p = 0.019). The median time from relapse to second Allo-HCT was 118 days (range 30–902 days), and the median time from relapse to DLI was 34 days (range 6–63 days) (p < 0.001). Myeloablative conditioning (MAC) was used in 13 DLI patients (68%) and in 16 second Allo-HCT patients (59%). Nineteen of 29 MAC transplantations (62%) were based on total body irradiation, and nine were busulfan based (missing data in one patient). RIC transplantations were all fludarabine based, in combination with either busulfan (11 patients) or melphalan (3 patients) (missing data in 3 patients). IST was cyclosporine based in the vast majority of transplantations (85% and 78% in the second Allo-HCT and DLI cohorts, respectively). Further information regarding significant differences between the second Allo-HCT and DLI cohorts can be found in Table 1, which compares baseline patient's characteristics. Table 2, Table 3 describe characteristics of the second transplantation (second Allo-HCT cohort) and the characteristics of the first transplantation of the DLI cohort, respectively.Table 1Comparative of baseline patient's characteristicsSecond Allo-HCTDLIp valueMedian Age, Years (range)28 (4–54)42 (22–61)0.015Disease, n (%)AML20 (74%)10 (52%)0.112ALL7 (26%)9 (42%)First HCT Date, years (%)<20047 (26%)4 (21%)0.492≥200420 (74%)15 (79%)Median Time to Relapse, days (range)378 (61–1508)152 (35–3956)0.019Median Time from Relapse; days (range)118 (30–902)34 (6–63)<0.001 Open table in a new tab Table 2Description of the second Allo-HCT characteristics of the second Allo-HCT cohortSecond Allo-HCT, n (%)Donor/Recipient GenderFemale → Male8 (38%)Other15 (62%)Missing4Stem Cell SourcePB16 (59%)Other11 (41%)HLA MismatchYes3 (12%)No22 (88%)Missing2MyeloablationMAC16 (59%)RIC11 (41%)ConditioningCyTBI10 (40%)FluBu6 (24%)FluMel3 (12%)BuCy1 (4%)Other5 (20%)Missing2Immunosuppressive TherapyCsA-Based23 (88%)Other3 (12%)Missing1TCDYes7 (26%)No20 (74%)Type of DonorRelated20 (80%)Unrelated5 (20%)Missing2Donor Change from First Allo-HCTSame Donor15 (68%)Other Donor7 (32%)Missing5PB = peripheral Blood; CyTBI = cyclophosphamide and total body irradiation; FluBu = fludarabine and busulfan; FluMel = fludarabine and melphalan; BuCy = busulfan and cyclofosfamide; CsA = cyclosporine. Open table in a new tab Table 3Description of the Allo-HCT characteristics of the DLI cohortDLI, n (%)Donor/Recipient GenderFemale → Male7 (39%)Other12 (61%)Stem Cell SourcePB18 (95%)Other1 (5%)HLA MismatchYes6 (31%)No13 (69%)MyeloablationMAC13 (68%)RIC6 (32%)ConditioningCyTBI8 (42%)FluBu7 (37%)BuCy3 (16%)Other1 (5%)Pre-DLI TreatmentCytarabine-Based7 (50%)Chemo + TKI/Sorafenib3 (22%)/1Cyclophosphamide-Based2 (14%)Other2 (14%)Missing5Immunosuppressive TherapyCsA-Based15 (78%)Other4 (22%)TCDYes7 (37%)No12 (63%)DonorRelated/Haploidentical14 (74%)/1Unrelated5 (26%)PB = peripheral Blood; CyTBI = cyclophosphamide and total body irradiation; FluBu = fludarabine and busulfan; BuCy = busulfan and cyclophosphamide; Chemo = chemotherapy; TKI = tyrosine kinase inhibitor; CsA = cyclosporine. Open table in a new tab PB = peripheral Blood; CyTBI = cyclophosphamide and total body irradiation; FluBu = fludarabine and busulfan; FluMel = fludarabine and melphalan; BuCy = busulfan and cyclofosfamide; CsA = cyclosporine. PB = peripheral Blood; CyTBI = cyclophosphamide and total body irradiation; FluBu = fludarabine and busulfan; BuCy = busulfan and cyclophosphamide; Chemo = chemotherapy; TKI = tyrosine kinase inhibitor; CsA = cyclosporine. The median follow-up of the whole cohort was 273 days (range 9–7013 days), and the median follow-up of the patients alive was 692 days (range 28–7013 days). The median follow-up of the second Allo-HCT and DLI cohorts were 404 days (range 22–7013 days) and 66 days (range 9–4249 days), respectively. Twenty-one patients (46%) had died at last follow-up. Of these, eight (17%) died due to NRM: six patients due to GvHD and its complications, two patients due to secondary malignancy, and one patient due to pulmonary disease. Thirteen patients died after relapse or progressive disease, including the patient diagnosed with myeloid sarcoma, who was treated with DLI at relapse. Grades II–IV acute GvHD was diagnosed in 10 (37%) and 5 (26%) patients after second Allo-HCT and DLI, respectively. Chronic GvHD was diagnosed in 10 (four extensive) and three patients after a second Allo-HCT and DLI, respectively. For the DLI patients, the median time to GvHD development was 54 days (range 7–465 days). The 2-year CI of relapse was 34% (standard error [SE] ± 7%). The 2-year CI of relapse for second Allo-HCT and DLI was 24% (SE ± 8%) and 49% (SE ± 13%), respectively (Table 4). Univariate analysis identified DLI (p = 0.093) as being statistically significantly linked to higher RI; however, time to relapse was not found to be significant (p = 0.310) (Fig. 1). Only the variable second Allo-HCT versus DLI was included in the CI relapse multivariate analysis, which did not confirm its statistical significance. However, when including time to relapse to multivariate analysis along to second Allo-HCT versus DLI, a shorter time to relapse was linked to higher relapse (p = 0.007).Table 4Univariate analysis comparing second Allo-HCT patients with DLI patientsSecond Allo-HCTDLIp value2-year OS60% (SE ± 10%)22% (SE ± 12%)0.0212-year DFS56% (SE ± 10%)20% (SE ± 12%)0.0972-year NRM16% (SE ± 7%)31% (SE ± 13%)0.4602-year relapse incidence24% (SE ± 8%)49% (SE ± 13%)0.093 Open table in a new tab The 2-year NRM of the whole cohort was 21% (SE ± 7%). The 2-year NRM for second Allo-HCT and DLI was 16% (SE ± 7%) and 31% (SE ± 13%), respectively (hazard ratio 0.59; confidence interval, 0.15–2.39), p = 0.460) (Table 4). In NRM univariate analysis, no variable was statistically significant at p < 0.1, so multivariate analysis was not performed for NRM. The 2-year DFS of the whole cohort was 38% (SE ± 8%). The 2-year DFS for second Allo-HCT and DLI was 56% (SE ± 10%) and 20% (SE ± 12%), respectively (Table 4). The DFS univariate analysis linked a longer time to relapse (p = 0.007) (Fig. 2) and second Allo-HCT (p = 0.097) to higher DFS. However, DFS univariate analysis performed on patients with late relapse (>6 months) showed no differences between DLI and second Allo-HCT (p = 0.910). The DFS multivariate analysis confirmed the association of longer time to relapse to higher DFS (p = 0.037). The 2-year OS of the whole cohort was 44% (SE ± 9%). The 2-year OS for second Allo-HCT and DLI was 60% (SE ± 10%) and 22% (SE ± 12%), respectively (Table 4). OS univariate analysis linked second Allo-HCT (p = 0.021) and a long time to relapse (p = 0.004) (Fig. 3) to better OS. However, OS univariate analysis performed on patients with late relapse (>6 months) showed no differences between DLI and second Allo-HCT (p = 0.653) (Fig. 4). OS multivariate analysis only confirmed time to relapse as statistically significant (p = 0.007). Multivariate analysis results can be found in Table 5.Figure 4OS comparing DLI (continuous line) and second Allo-HCT (dotted line) for patients with late relapses (p = 0.653).View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table 5Multivariate analysesVariablep valueOSSecond Allo-HCT vs. DLI0.169Time to Relapse0.005DFSSecond Allo-HCT vs. DLI0.616Time to Relapse0.037NRMNo Variable p < 0.1 in Univariate AnalysisN/ARelapse IncidenceSecond Allo-HCT vs. DLI0.259Time to Relapse0.007 Open table in a new tab The median DLI dose per patient was 5 × 107 CD3+/kg (0.1–19.4 × 107/kg) cells, the median first DLI dose was 1.03 × 107 CD3+/kg (0.1–19.4 × 107/kg) cells, and the mean number of infused DLI was 1.4/patient. The median dose of CD34+ cells infused within the first DLI was 0.3 × 106 CD34+ cells/kg (range 0.02–2). Ten patients (52%) received DLI from available cryopreserved cells, whereas the remaining 9 (42%) received DLI after requesting a donor's lymphapheresis. DLI–CI relapse univariate analysis showed that the use of TCD in the previous Allo-HCT was linked to higher relapse after DLI (p = 0.005), which was also confirmed on multivariate analysis (p = 0.003). DLI–DFS univariate analysis linked the use of TCD in the previous Allo-HCT (p = 0.002) and a trend of the use of DLI from unrelated donor (p = 0.071) to better DFS. DFS multivariate analysis only confirmed the use of TCD in the previous Allo-HCT (p < 0.001). Similar findings in DLI–OS univariate analysis, which linked TCD in Allo-HCT (p = 0.010) and the use of DLI from UD (log-rank, p = 0.107, Breslow, p = 0.043) to better OS. OS multivariate analysis only confirmed the TCD in Allo-HCT before DLI to higher OS (p = 0.002). Despite the available approaches, the prognosis for AL patients relapsing after Allo-HCT remains poor. Among patients who are candidates for intensive therapy, the treatment often entails the administration of chemotherapy plus a consolidation treatment. At that time, further immunotherapy-based treatment such as a second transplantation or DLI is a commonly used approach. We describe a cohort of patients treated at relapse to reduce the tumor burden, who subsequently received a second Allo-HCT or DLI. Outcomes between DLI and second Allo-HCT cohorts were comparable, as shown by multivariate analysis. Nevertheless, we found some differences in univariate analysis that are worthy of mentioning. We observed higher OS and DFS toward second Allo-HCT (p = 0.021 and p = 0.097, respectively) and a trend to higher CI of relapse (p = 0.093) in the DLI cohort (Table 4). Interestingly, when analyzing OS and DFS only for patients who had a late relapse (>6 months), we did not find any difference between DLI and second Allo-HCT and could not identify the trend in CI relapse observed in the whole cohort. The retrospective and nonrandomized inclusion of patients might have influenced these results. Conversely, it is important to take into account that the second Allo-HCT cohort included patients with a longer time interval from first Allo-HCT to relapse (time to relapse), which might suggest that the second transplantation cohort included patients with diseases with less aggressive biologic characteristics. In this context, the 2-year OS of 68% observed in the second Allo-HCT cohort appears to be high for patients in this context. However, these patients were young (median age 38 years) and on complete remission (CR) at second Allo-HCT, and the median time interval to relapse was >12 months. According to what has been published regarding second Allo-HCT for relapse, this cohort reflects a group of patients with good prognosis criteria, which might be reflected in the survival outcomes. We also calculated the time from relapse to second Allo-HCT and DLI and found that it was statistically significantly longer in the second Allo-HCT group. This result might be partially explained by the fact that the preparation for Allo-HCT takes more time compared with DLI, particularly if DLI is already cryopreserved. There are limited data comparing the use of DLI or second Allo-HCT for hematological malignancy relapse [12Willasch A.M. Salzmann-Manrique E. Krenn T. et al.Treatment of relapse after allogeneic stem cell transplantation in children and adolescents with ALL: the Frankfurt experience.Bone Marrow Transplant. 2017; 52: 201-208Crossref PubMed Scopus (17) Google Scholar, 13Bär B.M. Schattenberg A. Mensink E.J. et al.Donor leukocyte infusions for chronic myeloid leukemia relapsed after allogeneic bone marrow transplantation.J Clin Oncol. 1993; 11: 513-519Crossref PubMed Google Scholar, 14Pollyea D.A. Artz A.S. Stock W. et al.Outcomes of patients with AML and MDS who relapse or progress after reduced intensity allogeneic hematopoietic cell transplantation.Bone Marrow Transplant. 2007; 40: 1027-1032Crossref PubMed Scopus (60) Google Scholar] because it has not been decided which approach should be used. Because it would be of utmost complexity to develop a randomized study, available data come from retrospective studies. The Center for International Blood & Marrow Transplant Research (CIBMTR) [7Bejanyan N. Weisdorf D.J. Logan B.R. et al.Survival of patients with acute myeloid leukemia relapsing after allogeneic hematopoietic cell transplantation: a center for international blood and marrow transplant research study.Biol Blood Marrow Transplant. 2015; 21: 454-459Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar] retrospectively reported outcomes of relapsed AML patients. They found that an approach with chemotherapy alone (comparing chemo plus DLI or chemo plus second Allo-HCT) was associated with worse survival, with CR rates of 16% after chemotherapy, whereas CR rates after second Allo-HCT or DLI were 44% and 37%, respectively. Furthermore, median survival was inferior among patients receiving DLI comparing second transplantation at 7 and 12 months, respectively. The CIBMTR study also observed a 3-year OS of 4% for early relapsing patients (within 6 months of Allo-HCT). Consistent with this, we also found a lower OS and DFS in patients with shorter time to relapse. Significantly, others have also reported the relation of early relapse with poorer outcomes for immunotherapy after Allo-HCT relapse, with data having been published at several time cut-off points [15Bar M. Sandmaier B.M. Inamoto Y. et al.Donor lymphocyte infusion for relapsed hematological malignancies after allogeneic hematopoietic cell transplantation: prognostic relevance of the initial CD3+ T cell dose.Biol Blood Marrow Transplant. 2013; 19: 949-957Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar, 16Schmid C. Labopin M. Nagler A. et al.Donor lymphocyte infusion in the treatment of first hematological relapse after allogeneic stem-cell transplantation in adults with acute myeloid leukemia: a retrospective risk factors analysis and comparison with other strategies by the EBMT Acute Leukemia Working Party.J Clin Oncol. 2007; 25: 4938-4945Crossref PubMed Scopus (367) Google Scholar, 17Takami A. Okumura H. Yamazaki H. et al.Prospective trial of high-dose chemotherapy followed by infusions of peripheral blood stem cells and dose-escalated donor lymphocytes for relapsed leukemia after allogeneic stem cell transplantation.Int J Hematol. 2005; 82: 449-455Crossref PubMed Scopus (25) Google Scholar, 18Eapen M. Giralt S.A. Horowitz M.M. et al.Second transplant for acute and chronic leukemia relapsing after first HLA-identical sibling transplant.Bone Marrow Transplant. 2004; 34: 721-727Crossref PubMed Scopus (181) Google Scholar]. Further, although AML and ALL showed similar outcomes in this study, the effect of DLI in ALL seems to be of less benefit compared with AML. In ALL, 2-year OS <15% has been reported [19Collins Jr, R.H. Goldstein S. Giralt S. et al.Donor leukocyte infusions in acute lymphocytic leukemia.Bone Marrow Transplant. 2000; 26: 511-516Crossref PubMed Scopus (198) Google Scholar], whereas for AML, the number of patients who seem to benefit from DLI appears to be higher [16Schmid C. Labopin M. Nagler A. et al.Donor lymphocyte infusion in the treatment of first hematological relapse after allogeneic stem-cell transplantation in adults with acute myeloid leukemia: a retrospective risk factors analysis and comparison with other strategies by the EBMT Acute Leukemia Working Party.J Clin Oncol. 2007; 25: 4938-4945Crossref PubMed Scopus (367) Google Scholar]. In the current study, the median initial lymphocyte dose of the patients receiving DLI was 1.03 × 107 CD3+/kg. Initial doses of 1 × 107 CD3+/kg have been linked to a GvHD rate of 21% [15Bar M. Sandmaier B.M. Inamoto Y. et al.Donor lymphocyte infusion for relapsed hematological malignancies after allogeneic hematopoietic cell transplantation: prognostic relevance of the initial CD3+ T cell dose.Biol Blood Marrow Transplant. 2013; 19: 949-957Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar], although higher GvHD rates have also been reported [20Levine J.E. Braun T. Penza S.L. et al.Prospective trial of chemotherapy and donor leukocyte infusions for relapse of advanced myeloid malignancies after allogeneic stem-cell transplantation.J Clin Oncol. 2002; 20: 405-412Crossref PubMed Scopus (224) Google Scholar]. In the DLI cohort, GvHD was diagnosed in 42% of patients. Only three of the 17 evaluable DLI patients developed GvHD before receiving DLI and the median time from Allo-HCT to DLI was 179 days. Interestingly, a time from transplantation to DLI of <2 years has been linked to the development of GvHD [21Chalandon Y. Passweg J.R. Schmid C. et al.Outcome of patients developing GVHD after DLI given to treat CML relapse: a study by the Chronic Leukemia Working Party of the EBMT.Bone Marrow Transplant. 2010; 45: 558-564Crossref PubMed Scopus (52) Google Scholar]. One of the reasons to explain the GvHD rate reported here could be the fact that most DLI transplantations (63%) were T-cell replete (TCR), although there might be many factors influencing this result. GvHD could not be analyzed in a time-dependent manner, but the NRM univariate analysis of the DLI cohort showed comparable results between TCD and TCR transplantations (data not shown, p = 0.163). To the best of our knowledge, few data compare the use of DLI in both in vivo TCD and TCR transplantations [22Sehn L.H. Alyea E.P. Weller E. et al.Comparative outcomes of T-cell-depleted and non-T-cell-depleted allogeneic bone marrow transplantation for chronic myelogenous leukemia: impact of donor lymphocyte infusion.J Clin Oncol. 1999; 17: 561-568Crossref PubMed Google Scholar]. In this former study, only three of 23 patients had DLI in the TCR setting. In our study, DLI patients who received TCD conditioning associated favorable outcomes (OS, DFS, and relapse incidence). The number of patients in this subgroup is small, and this finding requires further confirmation, but we speculate that enhancement of the T-cell compartment in TCD transplantations (either pre- or postrelapse) might be of more benefit for lowering the relapse risk in AL patients than TCR transplantations (in which the T-cell reservoir is previously expanded), where DLI appear to provide less benefit. Regarding the statistically non-significant NRM difference between DLI and second Allo-HCT cohorts, the second Allo-HCT patients were younger compared with patients receiving DLI, which might be relevant as to why the NRM was lower (although not statistically significant) in the second Allo-HCT group compared with the DLI group. The selection bias of fitter patients for secondary transplantation may also partially explain these results. Conversely, 37% of the patients in the DLI cohort received a TCD-based Allo-HCT, whereas only 26% of the second transplantation cohort had in vivo TCD. Patients receiving a second Allo-HCT presented with more GvHD than did patients after DLI. The use of TCD has been associated with less GvHD and is therefore a more tolerable procedure [23Kröger N. Solano C. Wolschke C. et al.Antilymphocyte globulin for prevention of chronic graft-versus-host disease.N Engl J Med. 2016; 374: 43-53Crossref PubMed Scopus (245) Google Scholar]. In our study, 11 (41%) and 6 (32%) patients received a RIC conditioning regimen in the second Allo-HCT and DLI cohort, respectively. These data might have also influenced the final results; although it has been reported that the use of a RIC as second Allo-HCT conditioning does not appear to improve NRM rates significantly [24Shaw B.E. Mufti G.J. Mackinnon S. et al.Outcome of second allogeneic transplants using reduced-intensity conditioning following relapse of haematological malignancy after an initial allogeneic transplant.Bone Marrow Transplant. 2008; 42: 783-789Crossref PubMed Scopus (87) Google Scholar]. We recognize that this study has limitations, some related to its retrospective nature and others to its design itself. There is an obvious bias regarding patient selection to those who received second transplantation or DLI. The only way to overcome this bias would be by performing a randomized trial. In terms of GvHD, it would have been interesting to have data on the presence of GvHD at the time of relapse or therapy (DLI or second Allo-HCT) because others have linked the presence of GvHD at those time points to inferior survival [7Bejanyan N. Weisdorf D.J. Logan B.R. et al.Survival of patients with acute myeloid leukemia relapsing after allogeneic hematopoietic cell transplantation: a center for international blood and marrow transplant research study.Biol Blood Marrow Transplant. 2015; 21: 454-459Abstract Full Text Full Text PDF PubMed Scopus (173) Google Scholar]. Conversely, although data regarding outcomes were mostly complete, data about the variable time to relapse was unavailable in five patients. We calculated the time from first Allo-HCT to second Allo-HCT and time from first Allo-HCT to DLI and found comparable outcomes to time to relapse (data not shown). Finally, the GvHD analysis is incomplete because there were missing data on GvHD dates; however, the development of GvHD after DLI in AL seems to have little impact on survival outcomes [25Verdonck L.F. Petersen E.J. Lokhorst H.M. et al.Donor leukocyte infusions for recurrent hematologic malignancies after allogeneic bone marrow transplantation: impact of infused and residual donor T cells.Bone Marrow Transplant. 1998; 22: 1057-1063Crossref PubMed Scopus (66) Google Scholar]. Overall, outcomes of second Allo-HCT and DLI appear to be comparable in this study, in which the time to relapse affects OS and DFS in both DLI and second Allo-HCT patients. Results from a prospective randomized comparative study would be of great interest to discern what procedure would provide better results and less toxicity. However, variables such as the disease's characteristics, the donor's availability, or the patient's performance status would be very difficult to control in such a study, which renders this an unlikely approach. Until then, retrospective trials including a higher number of patients are warranted. The authors declare no competing financial interests.
Referência(s)