Objectives: Sparse data are available about the effects of apheretic graft composition on the clinical transplant outcome in allotransplanted patients who have hematologic malignant disease. Major obstacles in recent studies have included hetero-geneity of patient populations and differences in the conditioning regimens used.
Materials and Methods: This prospective study included 50 patients who had acute myeloblastic leukemia and received busulfan-fludarabine-antithymocyte globulin-based conditioning for peripheral allogeneic stem cell transplant. The concentration of CD34+ cells, T-cell subsets, B cells, and natural killer cells in the graft were analyzed by flow cytometry in the donors who were matched for human leukocyte antigen.
Results: In univariate analysis, infusion with a higher dose of natural killer cells (> 1.55 × 106/kg) was associated with improved survival (P = .007 for disease-free survival; P = .024 for overall survival) in patients with acute myeloblastic leukemia. Cox regression models revealed that increased concentration of natural killer cells and CD34+ cells positively affected the clinical outcome of allotransplanted patients (P = .005 for both cell types). According to univariate analysis, these findings were dependent on minimal residual disease and acute graft-versus-host disease. Graft-versus-host disease (acute and chronic forms) was not affected by graft composition.
Conclusions: Our results suggest that increased concentration of natural killer cells and CD34+ cells in the apheretic product may predict better survival. In contrast, busulfan-fludarabine-antithymocyte globulin-based conditioning eliminates the disadvantages that resulted from the high content of T-cell subsets and B cells, and the course of the transplant and clinical parameters were not affected by the amount of T and B cells.
Key words : Hematology, Malignancy, Natural killer cells, Stem cell transplant
The success of hematopoietic stem cell transplant is measured by the early observation of engraftment, fast immune recovery, acceptable complication and pro-cedure-related mortality rates, and positive survival parameters. The source and composition of the trans-planted cellular product can play a role in transplant success.1-3 In particular, the number of CD34+ cells above the critical threshold in the cellular product is of vital importance.4,5 The importance of the T-lymphocyte count in the product is understood for the results of T-lymphocyte-depleted hematopoietic stem cell transplant. In this type of transplant, well known issues include the development of engraftment failure and infectious complications due to late immune recovery and high relapse rate. Conversely, the risk of graft-versus-host disease (GVHD) is reduced.6
Stem cell products vary markedly in terms of quantity and quality, depending on the source of cells. The importance of stromal cells from bone marrow is understood from studies comparing bone marrow and peripheral blood used as a stem cell source in stem cell transplant procedures. These studies have shown that hematopoietic stem cells prepared from peripheral blood increased the risk of GVHD, and this effect adversely affected survival parameters.7,8
There are several reports about the effect of lymphocyte subtype counts observed in the apheretic graft on transplant success. These reports have shown that cells involved in the adoptive immune system, such as CD4+/CD25+/FoxP3+T-regulatory cells, B cells, CD8+ cells, and natural killer (NK) cells, may determine posttransplant immune recovery, sugges-ting that these cells are associated with 2 fundamental objectives: immune tolerance and the structuring of effector cells.9,10 The results from studies conducted to assess the effects of apheretic graft composition on transplant success are sparse and contradictory, possibly because of differences in the conditioning transplant regimen used, another key factor affecting transplant success. The recently developed non-myeloablative conditioning regimens that include lympholytic drugs, particularly antithymocyte globulin (ATG), have not been adequately assessed in the literature. Additionally, heterogeneity of the patients enrolled in the study groups is an important limitation of previous studies.3,11,12
This work aimed to investigate the effect of apheretic graft composition on transplant outcome in patients who had acute myeloblastic leukemia (AML) and who received a single ATG-conditioning regimen consisting of fludarabine, busulfan, and ATG.
Materials and Methods
A single-center, prospective, and crosssectional study was performed between March 2007 and January 2014. A total of 50 consecutive patients underwent allogeneic peripheral blood stem cell transplant for AML from a sibling donor who was identical in human leukocyte antigen (HLA) typing. In addition to the specific cytogenetic and molecular features associated with high risk of relapse or refractory state, age-independent clinical factors associated with poor prognosis were defined as the presence of leukocyte count at diagnosis > 100 ×109/L and the persistence of minimal residual disease defined by multiparameter flow cytometry after induction therapy.13 The features of the patients were tabulated (Table 1). Patients who had a concomitant secondary cancer, acute coronary syndrome, mismatched transplant, and incomplete analysis of the apheresis product were excluded from the study. All patients signed consent to use their medical records for research purposes, and the study was approved by the Başkent University Faculty of Medicine, Research Ethics Committee, Ankara, Turkey.
Stem cell collection
Granulocyte colony-stimulating factor (G-CSF) (Filgrastim, Neupogen, Amgen-Roche, Thousand Oaks, CA, USA) was used for all donors to mobilize stem cells from the marrow to peripheral blood at a dosage of 10 μg/kg/day subcutaneously (in 2 divided doses daily) for a total of 5 days. All donors underwent stem cell collection using a continuous flow cell separator (Cobe Spectra V. 7.0, Terumo BCT, Lakewood, CO, USA). All apheresis procedures were performed 2 hours after the ninth and eleventh doses of G-CSF. All apheresis products were routinely cultured for microbial identification (SOP; KIT-KY 022). When infection was present at collection, this apheresis product was excluded from the study.
All patients received a busulfan-fludarabine-ATG-based conditioning regimen. The conditioning consisted of a myeloablative regimen (9.6-12.8 mg/kg busulfan, 150 mg/kg fludarabine, and 7.5-30 mg/kg rabbit ATG). All patients received cyclosporine and methotrexate for GVHD prophylaxis. Cyclosporine was administered at a dose of 3 mg/kg/day intravenously from 2 days before transplant, with target trough concentrations ≥ 200 ng/mL until it was tapered. Methotrexate was administered at a dose of 10 mg/m2 intravenously on day 2 and 8 mg/m2 intravenously on days 4 and 8, supported by folinic acid 24 hours after each methotrexate dose. Acute GVHD was scored by standard criteria and initially treated by increasing doses of methylprednisolone (2 mg/kg).14 Patients surviving in remission ≥ 100 days after transplant, with full donor chimerism, were considered evaluable for chronic GVHD, which was graded according to established criteria.15
The patients were placed into laminar airflow rooms, and irradiated and leukocyte-depleted blood products were used for all patients. Patients were transfused with red blood cells or platelets when their hemoglobin was < 80 g/L and platelet count < 20 × 109/L. Viral and fungal prophylactic drugs were given according to standard operating procedures compatible with Joint Accreditation Committee of International Society of Cellular Therapy and European Blood and Marrow Transplantation (JACIE) standards (SOP; KIT-TU-002). The patients received a preemptive treatment with ganciclovir for cytomegalovirus reactivation based on the results of molecular screening.
All flow cytometry procedures were performed with blue (wavelength, 488 nm) and red (wavelength, 633 nm) lasers and an 8-parameter fluorescence-activated cell sorting device (Becton Dickinson FACS CANTO II, BD Biosciences, San Jose, CA, USA). The CD34+ cell counts were performed following the ISHAGE sequential gating strategy.16
Apheresis samples were diluted in phosphate-buffered saline (pH 7.2) to obtain a leukocyte count 10 to 20 × 109/L, as previously recommended.17 All samples were counted in duplicate. Briefly, a 100-μL apheresis sample was incubated with 10 μL anti-CD45 conjugated with fluorescein isothiocyanate (FITC) and phycoerythrin (PE) and PE-conjugated anti-CD34 (clone 8G12, BD Biosciences) for 20 minutes at 4°C. Initial experiments included samples incubated with an isotype control PE-labeled immunoglobulin G1 (IgG1) as a negative control for nonspecific binding. Red blood cells were lysed with 2.0 mL bicarbonate-buffered ammonium chloride solution (lysis solution; 0.15 M ammonium chloride, 0.01 M sodium bicarbonate, 1.0 mM ethylenedi-aminetetraacetic acid) for 15 minutes at room temperature. The cells were washed twice, resuspended in 1.0 mL phosphate-buffered saline, stored on ice in the dark, and analyzed within 1 hour. The original dual-platform ISHAGE gating strategy and calculation to obtain the number of CD34+ cells per kilogram was performed as described previously.16
Additionally, leukocyte subtypes in all apheresis products were analyzed using anti-CD45 (allophycocyanin APC-H7, a new APC-cyanine tandem dye), anti-CD19 (APC), anti-CD16/56 (PE), anti-CD3 (Alexa Fluor), anti-CD4 (PE-Cy7, an APC-cyanine tandem dye), and anti-CD8 (FITC). All antibodies were obtained from a commercial supplier (BD Biosciences). The NK cells phenotype mixture CD16+/CD56+/CD3- was used.
The number of CD34+ cells and the leukocyte subtype of the apheresis products were determined. The data analysis was performed using computer software (FACS DIVA, BD Biosciences) after acquisition. The United Kingdom National External Quality Assessment Site (NEQAS) Program for routine CD34+ cell analysis was maintained bimonthly (Laboratory number 42074).
Absolute cell numbers in each leukocyte subpopulation were defined by multiplying by the percentage of each cell subset (determined by electronic gating during analysis) and by the total white blood cell count (determined by automated cell counting) using an analyzer (Cell-Dyn 3700 analyzer, Abbott Diagnostics, Chicago, IL, USA).
Statistical analyses were performed using a statistical package (SPSS for Windows, Version 17.0, SPSS Inc., Chicago, IL, USA). When continuous variables were normally distributed, they were reported as mean ± standard deviation (P > .05 in Kolmogorov-Smirnov test or Shapiro-Wilk test; n < 30). When the continuous variables were not normally distributed, they were reported as the median. Comparisons between groups were evaluated using Mann-Whitney test. The categorical variables between groups were analyzed using chi-square test or Fisher exact test.
Overall survival time was defined as the number of months from the date of transplant to death as a result of disease or last follow-up. The disease-free survival was defined as the number of months from the date of transplant to relapse. The receiver operating characteristic curves failed to reveal a cutoff value for the cell dose; therefore, continuous covariates were encoded as binary covariates after dichotomization, using the median as the cutoff. Survival curves were generated using Kaplan-Meier method. Differences between 2 curves were analyzed with log-rank test.
The association of the infused cell count with overall survival was analyzed using Cox proportional hazard model. Cox regression model with stepwise selection was used to identify variables of the infused cell value. Thereafter, the treatment effect adjusted for these selected variables was calculated. The Cox model also was used to evaluate the interaction of treatment effect with subgroup status in an exploratory analysis. Differences in the best overall response rates between the treatment groups were analyzed with Cochran-Mantel-Haenszel test.
There were 50 patients who underwent stem cell transplant using myeloablative conditioning regimens (Table 1). There were 42 patients who had primary AML, and the remaining patients had various forms of secondary AML before inclusion in the trial. There were sufficient samples for genetic testing of all patients. There were 8 patients who had single adverse genetics (2 patients had 7q deletion, and 1 patient each had monosomy 7, trisomy 21, trisomy 8, t(9:22), t(3:6), and hypodiploidy) and noncomplex genotypes. There was FLT3-ITD-positivity observed in 2 of the measured 20 samples. Patient characteristics are shown in Table 1.
Follow-up data were obtained from all 50 patients. The patients received single conditioning and a single prophylactic regimen against GVHD. All patients were in complete remission at follow-up. Median follow-up in survivors was 25.7 months (range, 2.4-76.9 mo). No patients were lost to follow-up.
Clinical factors and outcome
The median times to neutrophil and platelet engraftment were 12 days (range, 9-17 d) and 12 days (range, 7-19 d). No early or late engraftment failure was documented. The median disease-free and overall survival times were 24.8 months (range, 0.7-76.9 mo) and 25.7 months (range, 2.4-76.9 mo).
Univariate analysis of clinical factors affecting disease free survival and overall survival showed that age, sex, age-independent poor-risk features, and cytomegalovirus antigenemia did not affect transplant outcome (Table 2). The presence of minimal residual disease in clinical remission negatively affected 2-year disease-free and overall survival (Table 2). Acute GVHD had a negative effect on transplant outcome. The dose of busulfan or ATG administered in myeloablative conditioning did not affect transplant outcome.
Apheretic graft composition and outcome
Regarding cell dose in the graft, CD34+ cells were administered at a dose of 5.5 × 106/kg (range, 3.5-10.5 × 106/kg), CD3+ cells 16.0 × 107/kg (4.8-45.1 × 107/kg), CD4+T cells 8.8 × 107/kg (1.1-23.6 × 107/kg), CD8+ T cells 4.2 × 107/kg (0.4-14.8 ×107/kg), B cells 2.8 × 107/kg (0.3-9.8 × 107/kg), and NK cells 1.6 × 107/kg (0.1-8.0 × 107/kg). Statistical analysis showed that the infused number of NK cells affected the probability of disease-free (P = .005) and overall survival (P = .024) (Figure1). A dose of NK cells > 1.5 × 107/kg was associated with an improved outcome.
Cox regression model for graft composition confirmed that NK cell count in the graft was significantly associated with overall survival (P = .005). Increased CD34+ cell concentration in the graft also positively affected survival of transplant recipients (P = .005) (Table 3).
Recent advances in the clinical application of hematopoietic stem cell transplant may benefit many patients who have leukemia. Many factors may affect transplant outcome, such as disease type, pretransplant disease burden, comorbidities of the recipients, age, Karnofsky performance score of the recipients, donor selection, and sources of stem cells.2,7,18 Among these factors, the stem cell content of the cellular product is the most important.4 However, there were limitations in all existing research investigating the effect of cell content in the graft on clinical outcome. First, enrolled patient populations were heterogeneous in disease type and pretransplant disease status. Second, a wide variety of conditioning regimens were administered to the recipients, even though conditioning is associated strongly with engraftment, disease control, complications such as acute and chronic GVHD, toxicities, and morbidity. Third, standardized exclusion criteria have not been defined, such as exclusion from the study when infection is present at the time of collection. However, the effects of infections developing during other periods of transplant treatment on the association between cell dose and study endpoints have not been analyzed widely.19-21
To overcome the limitations identified in previous studies, we recruited only patients who had AML in hematologic remission. Patients received a single induction treatment with cytosine arabinoside and idarubicin and salvage chemotherapy when necessary. Standard supportive treatment was implemented in an accredited center. All donors were an HLA-full match with recipients. For stem cell collection, a standard procedure was applied (SOP: KIT-TU-002). The cyclosporine and methotrexate for GVHD prophylaxis regimen was used. More importantly, this study also used a conditioning regimen, consisting of busulfan, fludarabine, and ATG, and this was not used in other relevant studies. The efficacy of this regimen for AML patients has been demonstrated in several studies. This regimen is a strategy to limit the toxicity previously observed with myeloablative conditioning regimens, while maintaining a full dose of the myeloablative agent without increasing nonrelapse mortality22-24; it can reduce the incidence of acute and chronic GVHD, because both are important contributing factors for mortality.25,26 Using a single regimen for all our patients may overcome the difficulty of interpreting the results. To prevent the effect of competing factors, Cox proportional hazard model was used in evaluation of transplant outcome analysis.
Providing a sufficient number of CD34+ cells in the cellular product is the primary requirement to ensure adequate engraftment. Lymphocytes in the product also are required for rapid immune reconstitution and facilitation of engraftment.3,4 The results of the present study showed that neutrophil or platelet recovery was not delayed, suggesting the viable CD34+ cells and sufficient lymphocyte count in the unmanipulated cellular product were enough to maintain sustained engraftment.
The effects of other cells in the graft on the clinical outcome after allogeneic transplant are less clear. The T-lymphocyte count in the product is associated with acute and chronic GVHD. Therefore, for patients at high risk of GVHD occurrence, bone marrow, rather than peripheral blood, is preferred. When in vivo T-cell depletion with pretransplant ATG was performed, even with a high CD3+ cell content, no increase in the incidence of GVHD was reported.27,28 We demonstrated that CD3+ T cell count was not associated with transplant-related mortality and clinical outcome. The explanation of our observation was based on the following information from the literature. The rabbit ATG used in the conditioning regimen may have played a role via a lympholytic effect extending up to several weeks. Previous studies have shown that ATG induces a marked long-term decline in the CD4+ and CD8+ T cell counts and CD4+/CD8+ ratio. Regeneration the effector cells such as CD8+ T cells was delayed, possibly due to an additive effect of drugs used for GVHD prophylaxis up to several months after transplant. These effects can translate clinically as effective prevention of GVHD, promotion of engraftment, or otherwise an increased risk of malignant disease relapse and infectious complications. However, if ATG was not used in the conditioning, busulfan-fludarabine may be as potent as busulfan-fludarabine-ATG in reducing mortality rate (reported as only 5.7%) due to acute GVHD.12 It is probable that the long-acting lympholytic effect of the drugs used in the conditioning regimen for elimination of the malignant clone was compensated for by drug efficacy, and the results showing the effect of fludarabine cannot be ignored.
The NK cells may eliminate allogeneic tumor cells via cell-to-cell contact and lytic effects. Better survival is associated with the dose of the NK cells.2,19 Previous reports also have reported a short-lived effect of ATG-Fresenius on NK cells, but not B cells, when provided at low doses.29 The T-cell–depleting effects of conditioning and drugs used for prophylaxis of GVHD, with the contribution of slow thymic regeneration, persist ≥ 1 year after transplant. However, NK cells can regenerate and reach baseline values a few weeks after transplant.29,30 The NK cell counts also are higher 1 month after ATG- than non-ATG-conditioned transplant, suggesting that ATG may exert a stimulatory effect on NK lympho-poiesis.31 This theoretically suggests that NK cells should be useful posttransplant, because they have anti-GVHD, antitumor, and anti-infection effects, as do other early recovered innate immune cells such as monocytes, neutrophils, and basophils.30,31 This study showed that NK cell concentration in the graft affected survival of transplant recipients. The results presented here are congruous with results reported in the literature, because the increased dose of NK cells in the graft was associated with a low relapse rate.
Regarding B cells, some studies showed an increased risk of transplant-related mortality in patients receiving a dose of B cells > median value. This was attributed to a potential direct consequence of the increased risk of acute GVHD.11,32 In the present study, the dose of infused B cells did not affect the posttransplant outcome of allotransplanted patients. The variability of these results may be associated with the drugs used in the conditioning regimen. In this context, both naive and memory B cells have been observed at significantly higher levels in ATG- than non-ATG-conditioned patients.33 Moreover, improved reconstitution of B cells was due to the lower incidence of GVHD after ATG rather than the ATG-conditioning.31 Our observations may suggest that the B-cell count in the graft had an insignificant effect on the clinical transplant outcome when a busulfan-fludarabine-ATG-based conditioning regimen was used.
Analysis of memory cells and dendritic cells and the assessment of lymphocyte subsets after transplant were not included in this study, and this may be considered a limitation of the present research.
In conclusion, in allogeneic transplant, the composition of T or B lymphocytes in a graft does not affect the clinical outcome in patients with AML when a busulfan-fludarabine-ATG-based conditioning regimen is used due to in vivo lympholytic effects of the drugs. However, NK cells may be affected, maintaining a graft-versus-leukemia effect, as supported by multivariate analysis for this patient population.
Volume : 13
Issue : 5
Pages : 453 - 460
DOI : 10.6002/ect.2014.0223
From the 1Baskent University Adana Adult Bone Marrow
Transplantation Center and 2Baskent University Adana Teaching and
Medical Research Center, Biostatistics, Ankara, Turkey
Acknowledgements: The authors have no conflicts of interest to declare. No funding was received for this study.
Corresponding author: Mahmut Yeral, Baskent University Adana Adult Bone Marrow Transplantation Center, Yuregir, 01250, Adana, Turkey
Phone: +90 322 327 2727- 2023
Fax: +90 322 327 1274
Table 1. Patient Characteristics
Table 2. Univariate Analysis of Clinical Factors Affecting Disease-Free Survival and Overall Survival
Table 3. Cox Regression Model for Graft Composition Affecting Overall Survival: Analysis of Maximum Likelihood Estimates
Figure 1. Effects of (NK) Cell Dose on Probability of Disease-Free Survival and Overall Survival in Patients Who Had Acute Myeloblastic Leukemia After Human Leukocyte Antigen-Identical Peripheral Stem Cell Transplant