Hematopoietic Engraftment in Recipients of Unrelated Donor Umbilical Cord Blood Is Affected by the CD34 and CD8 Cell Doses

Transcription

1 Biology of Blood and Marrow Transplantation 13: (2007) 2007 American Society for Blood and Marrow Transplantation /07/ $32.00/0 doi: /j.bbmt Hematopoietic Engraftment in Recipients of Unrelated Donor Umbilical Cord Blood Is Affected by the CD34 and CD8 Cell Doses Seitaro Terakura, 1,7 Eiichi Azuma, 2,7 Makoto Murata, 1,7 Tadashi Kumamoto, 2,7 Masahiro Hirayama, 2,7 Yoshiko Atsuta, 3,7 Yoshihisa Kodera, 4,7 Makoto Yazaki, 5,7 Tomoki Naoe 1,7 Koji Kato 6,7 1 Department of Hematology and Oncology, Nagoya University Graduate School of Medicine, Nagoya, Japan; 2 Department of Pediatrics and Cell Transplantation, Mie University School of Medicine, Tsu, Japan; 3 Department of Hematopoietic Stem Cell Transplantation Data Management, Nagoya University School of Medicine, Nagoya, Japan; 4 Bone Marrow Transplantation Center, Japanese Red Cross Nagoya First Hospital, Nagoya, Japan; 5 Department of Pediatrics, Nagoya Higashi Municipal Hospital, Nagoya, Japan; 6 Department of Pediatric Hematology/Oncology, Japanese Red Cross Nagoya First Hospital, Nagoya, Japan; and 7 Tokai Cord Blood Bank, Nagoya, Japan The first two authors contributed equally to this study. Correspondence and reprint requests: Makoto Murata, MD, PhD, Department of Hematology and Oncology, Nagoya University Graduate School of Medicine, 65 Tsurumai-cho, Showa-ku, Nagoya, Aichi , Japan ( mmurata@med.nagoya-u.ac.jp). Received December 19, 2006; accepted March 12, 2007 ABSTRACT Umbilical cord blood (UCB) transplantation is limited by the low number of hematopoietic stem cells in UCB units, which results in a low engraftment rate in transplant recipients. Here, we measured the total nucleated cell count and CD34, CD3, CD4, CD8, CD14, and CD16 /56 cell doses in each UCB unit and evaluated their influence on engraftment and other outcomes in 146 recipients. Multivariate analysis showed a significant association between a higher incidence of successful engraftment and a dose of CD34 and CD8 cells above the median ( and cells/kg, respectively). Engraftment occurred 4 days earlier in patients who received UCB with more than the median dose of CD34 cells than those receiving UCB at or below the median. Stratification of the group according to CD34 cell dose revealed a significant influence of the CD8 cell dose on the time to achieve neutrophil engraftment in patients receiving a lower CD34 cell dose, whereas there was no significant influence in the patients receiving a higher CD34 cell dose. These results suggest that consideration of CD34 and CD8 cell doses in UCB units may improve the engraftment in recipients of UCB transplantation American Society for Blood and Marrow Transplantation KEY WORDS Cord blood transplantation Engraftment CD34 CD8 T lymphocyte INTRODUCTION Umbilical cord blood (UCB) from unrelated donors has been established as an alternative source for hematopoietic stem cell (HSC) transplantation in patients who lack a human leukocyte antigen (HLA)- matched bone marrow or a peripheral blood stem cell (PBSC) donor [1-10]. UCB transplantation (UCBT) is limited, however, by the low number of HSCs per unit of UCB, resulting in a low engraftment rate and slow hematopoietic recovery after transplantation. There is a consensus that higher total nucleated cell count (TNCC) per recipient body weight and fewer differences in HLAs between the donor and recipient promote engraftment in recipients of UCBT [3,11-14]. A higher dose of CD34 cells in UCB is also thought to be associated with better engraftment, lower treatment-related mortality (TRM), and higher survival rate in the recipients [13]. The contribution of accessory cells, which are cryopreserved along with HSCs in UCB units, to 822

2 CD34 and CD8 Cell Doses and Engraftment after UCBT 823 engraftment in UCBT recipients has not been well investigated. In bone marrow transplantation (BMT), it is well known that the numbers of donor T cells as well as HSCs can affect the engraftment rate. In humans, when compared with unmanipulated grafts, T cell-depleted bone marrow prevents severe acute graft-versus-host disease (agvhd), but results in a lower probability of engraftment [15-17]. A phase I-II clinical trial of HLA-mismatched bone marrow transplantation demonstrated that the removal of CD8 cells but not CD4 cells from donor marrow increases the risk of rejection [18]. Also, in a transplant model, in which mice were lethally irradiated and subsequently reconstituted with a mixture of T cell-depleted syngeneic bone marrow cells and untreated major histocompatibility complex (MHC)-mismatched allogeneic bone marrow cells, all mice developed complete chimera of the MHC-mismatched donor mice [19]. Another MHC-mismatched mouse transplant model showed that donor CD8 T cells are at least 5-fold more effective than donor CD4 T cells for preventing marrow graft rejection [20]. Thus, CD8 T cells in the graft facilitate optimal engraftment of HSCs in BMT. In this study, we prospectively measured the TNCC and CD34, CD3, CD4, CD8, CD14, CD16 /CD56 cell doses in each UCB unit shipped from a single cord blood registry, and we evaluated how these accessory cell doses affect the engraftment and other outcomes in 146 recipients of UCBT. The analyses demonstrated that a higher incidence of successful engraftment was associated with a dose of CD34 and CD8 cells above the median ( and /kg of recipient body weight, respectively). Stratification of the group according to CD34 cell dose revealed a significant influence of the CD8 cell dose on the time to achieve neutrophil engraftment in the patients receiving less than the median dose of CD34 cells, but there was no significant influence in the patients receiving more than the median dose of CD34 cells. These findings and previous data [12] suggest the number of CD34 cells in the UCB units affects the engraftment in recipients. Additionally, the CD8 cell dose may have a beneficial effect on engraftment in recipients of UCBT. PATIENTS, MATERIALS, AND METHODS Patients Between September 1997 and August 2005, a total of 240 UCB units were shipped from the Tokai Cord Blood Bank to transplant institutions. A total of 146 patient and UCB graft pairs were selected according to the following criteria: (1) patients had acute leukemia, other malignancies, severe aplastic anemia, severe combined immunodeficiency syndrome, or a hereditary metabolic disorder; (2) at least 2 agents were used as a preconditioning regimen; (3) at least 1 immunosuppressant agent other than antithymocyte globulin was used for GVHD prophylaxis; (4) an unmanipulated single UCB graft was transplanted; (5) test vials were available for flow cytometry analysis; and (6) clinical outcome data were available. The characteristics of the patients are summarized in Table 1. One hundred thirty-seven patients (93.8%) were treated for malignant diseases and 100 (68.5%) were classified as having advanced diseases, including acute lymphoblastic leukemia (ALL) beyond first relapse (n 26), acute myelogenous leukemia (AML) beyond first relapse (n 29), acute type adult T cell lymphoma/leukemia (n 7), chronic myelogenous leukemia (CML) in accelerated or blast phase (n 6), myelodysplastic syndrome (MDS) with excess blast (n 8), non-hodgkin s lymphoma refractory or beyond second remission (n 14), and multiple myeloma (n 3), solid tumor (n 2), or other malignancies (n 5) not in complete remission (CR). The preconditioning regimen varied according to disease or stage of disease at transplantation. Thirty-seven patients (25.3%) received a regimen including fludarabine with reduced total body irradiation (TBI) ( 10 Gy) or without TBI, which was defined as a fludarabine-based reduced intensity conditioning (RIC) regimen. As a prophylaxis for GVHD, 95 patients (65.1%) received a cyclosporine A-based regimen, 44 (30.0%) received a tacrolimus-based regimen, and 84 (57.5%) received short-term methotrexate with or without other immunosuppressive agents. One hundred forty (95.9%) patients received granulocyte colony-stimulating factor (G-CSF) after transplantation. HLA Typing and Donor Matching Confirmatory HLA typing and matching of HLA-A, -B, and -DRB1 of the selected UCB unit and recipient were performed at the Tokai Cord Blood Bank prior to transplantation. HLA-A and -B were typed using the standard serologic method, and HLA- DRB1 type was determined by polymerase chain reaction sequencing-based typing. The HLA differences between the patients and the UCB units are summarized in Table 1. Flow Cytometry of UCB Units Flow cytometry was performed for the cells in UCB test vials when they were thawed for confirmatory HLA typing. The doses of each cell type from the UCB units actually used for transplantation were estimated by multiplying the TNCC measured before cryopreservation by the percentage of each cell fraction. For immunophenotyping, cells were incubated for 20 minutes at 4 C with 1 of the following monoclonal antibodies: fluorescein isothiocyanate-conju-

3 824 S. Terakura et al. Table 1. Patient and Transplant Characteristics Patient Characteristics n 146 Age, median, years (range) 15.0 ( ) Body weight, median, kg (range) 41.0 ( ) Recipient sex, n (%) Male 89 (61.0) Female 57 (39.0) Diagnosis, n (%) Malignant disease Acute lymphoblastic leukemia 49 (33.5) Acute myelogenous leukemia 39 (26.7) Adult T cell lymphoma/leukemia 8 (5.5) Chronic myelogenous leukemia 6 (4.1) Myelodysplastic syndrome 11 (7.5) Non-Hodgkin s lymphoma 14 (9.6) Multiple myeloma 3 (2.1) Solid tumor 2 (1.4) Other malignancies 5 (3.4) Nonmalignant disease Severe aplastic anemia 3 (2.1) Severe combined immunodeficiency syndrome 2 (1.4) Other nonmalignancies 4 (2.7) Disease states at transplant, n (%) Standard 46 (31.5) Advanced 100 (68.5) Degree of HLA match in rejection vector, n (%) 6/6 antigens 35 (24.0) 5/6 antigens 65 (44.5) 4/6 antigens 45 (30.8) 3/6 antigens 1 (0.7) Degree of HLA match in GVH vector, n (%) 6/6 antigens 34 (23.3) 5/6 antigens 72 (49.3) 4/6 antigens 38 (26.0) 3/6 antigens 2 (1.4) ABO blood type match, n (%) Match 48 (32.9) Minor mismatch 42 (28.8) Major mismatch 56 (38.3) Conditioning agents, n (%)* Two 51 (34.9) Three 95 (65.1) Total body irradiation dose, n (%) Full (>10 Gy) 75 (51.4) Reduced (<10 Gy) 33 (22.6) No 38 (26.0) GVHD prophylaxis, n (%) Cyclosporine A alone 30 (20.6) Cyclosporine A short-term methotrexate 55 (37.7) Cyclosporine A prednisolone 10 (6.8) Tacrolimus alone 12 (8.2) Tacrolimus short-term methotrexate 28 (19.2) Tacrolimus predonisolone 4 (2.7) Short-term methotrexate alone 1 (0.7) Others 6 (4.1) G-CSF administration after transplantation, n (%) Yes 140 (95.9) No 6 (4.1) G-CSF indicates granulocyte colony-stimulating factor. *Total body irradiation was considered a single agent. gated mouse anti-cd34, -CD3, -CD4, or -CD45 (BD Biosciences, Mountain View, CA), or phycoerythrinconjugated mouse anti-cd8, -CD14, -CD16, or -CD56 (BD Biosciences). The numbers of each cell were calculated as the percentage with a low sidescatter and using the CD45 lymphocyte gate, adjusted for live cells by 7-amino actinomycin D vital dye exclusion. Counting of CD34 cells was performed according to the International Society of Hematotherapy and Graft Engineering guidelines [21]. CD3, CD4, CD8, CD14, and CD16 /CD56 cell numbers were determined using a protocol based on the guidelines from the Centers for Disease Control and Prevention [22]. Stained samples were analyzed on a FACScan flow cytometer (BD Biosciences) as described perviously [23]. Approval of the study was obtained from the institutional review board of Mie University Hospital. Study Definitions The primary study endpoints were the engraftment rate and its kinetics. The secondary study endpoints were GVHD, TRM, relapse-free survival, and overall survival (OS). Hematologic recovery was defined as the time in days to an absolute neutrophil count 500/ L (first of 3 consecutive days) and a platelet count 20,000/ L (without transfusion support). Failure of engraftment was defined as the absence of an absolute neutrophil count recovery at day 100. Patients who survived at least 14 days after transplantation were evaluated for agvhd, and patients who survived at least 100 days after transplantation were evaluated for chronic GVHD (cgvhd). agvhd and cgvhd were graded according to standard criteria [24,25]. TRM was defined as any death that occurred while the patient was in remission. Relapse-free survival was defined as the number of days from transplantation to disease progression or death from any causes other than disease progression. OS was defined as the number of days from transplantation to death from any cause. Statistical Analysis Prospective collection of patient characteristics and outcomes were performed by the Tokai Cord Blood Bank Data Management Center. Variables related to the patients, diseases, grafts, and transplant procedure were compared among the groups by chi-square analysis for categoric variables and the Mann-Whitney test for continuous variables. The Kaplan-Meier method was used to obtain estimates of relapse-free survival and OS, and the log-rank test was used to evaluate differences between patient group strata [26]. Cumulative incidence curves for hematopoietic recovery, GVHD, TRM, and relapse were used, taking account of competing risks. To calculate the probability of neutrophil and platelet recovery and GVHD, death was the competing event; for

4 CD34 and CD8 Cell Doses and Engraftment after UCBT 825 TRM, relapse was the competing event; for relapse, TRM was the competing event [27]. Cox-proportional hazards univariate and multivariate regression models were applied to identify predictors of study endpoints [28]. All variables with P.10 were entered into the multivariate logistic regression using a backward, stepwise method. Day 100 posttransplant was applied as a landmark for analysis of cgvhd as a correlate of study endpoints. P-values of.05 were regarded as indicating statistical significance, and those between.05 and.20 as suggesting a trend. Statistical analyses were performed using STATA version 8.2 software (STATA Corp., College Station, TX). RESULTS UCB Graft Composition We found that the TNCC, CD34, CD3, CD4, CD8, CD14, and CD16 /CD56 cell doses in the individual UCB units varied greatly (Figure 1). The median TNCC was /kg (range: /kg), and UCB containing /kg of TNCC was used in 61 (81.3%) of 75 child patients ( 15 years old) and 12 (16.9%) of 71 adult patients ( 15 years old). UCB containing /kg of TNCC, which is regarded as a minimum dose for successful engraftment [9] was used in 70 (93.3%) of 75 child patients and 27 (38.0%) of 71 adult patients. The median number of CD34 cells Figure 1. UCB graft cell dose. Box plot for cell doses of each cell fraction. The horizontal bar within each box plot indicates the median of the contained cell number. The top and bottom of the box indicates the 25th and 75th percentiles, respectively. The top and bottom of the whisker indicates the 5th and 95th percentiles, respectively. was /kg (range: /kg), and UCB containing /kg of CD34 cell was used in 53 (70.7%) of 75 child patients and 18 (25.4%) of 71 adult patients. UCB containing /kg of CD34 cells, which has been proposed by Wagner et al. [13] as the minimum for engraftment, was used in 42 (56.0%) of 75 child patients and 8 (11.3%) of 71 adult patients. The median numbers and ranges of CD3, CD4, CD8, CD14, and CD16 /CD56 cells are summarized in Figure 1. Engraftment and Hematopoietic Recovery Kinetics The cumulative incidence of engraftment at day 60 was 0.81 (95% confidence interval [CI], ). A total of 26 patients (17.8%) did not exhibit an engraftment. Of these patients, 15 died on or before day 28 and 3 died after day 28 from neutropenia. Three patients died of early progression of their disease, and the remaining 5 patients were censored on the day of the second transplantation because of primary engraftment failure. For the 120 patients who exhibited engraftment, the median time to achieve neutrophil engraftment was 22 days (range: days), and the median time to achieve platelet engraftment was 44 days (range: days). Univariate analysis of engraftment revealed a significant association with a higher incidence of engraftment failure with higher patient age and body weight as a continuous variable and with a median or lower TNCC and CD34, CD3, CD4, CD8, and CD16 /CD56 dose (Table 2). A trend for a higher incidence of engraftment failure (P-value between 0.05 and 0.20) was found in the patients with advanced disease, an HLA mismatch in the rejection vector in 1 or more loci, and a fludarabine-based reduced intensity conditioning (RIC) regimen. Multivariate analysis showed a significant association between a higher incidence of engraftment failure and a lower dose of CD34 cells ( /kg) and CD8 cells ( /kg). The time to achieve neutrophil engraftment in the patients receiving more than a median dose of CD34 cells (median: 21 days; range: days) was significantly earlier than that in patients receiving a median or lower dose of CD34 cells (25 days; days) (P.0001 at day 21 and P.0001 at day 28). The overall incidences of engraftment in each patient group were 90.4% and 72.3% (P.004), respectively. We next evaluated the influence of the CD8 cell dose on the cumulative incidence of neutrophil engraftment in each patient group according to CD34 cell dose (Figure 2). In the group receiving more than the median dose of CD34 cells, the time to achieve neutrophil engraftment was similar (median: 21 days; range: days) in the patients receiving more than the median dose of CD8 cells ( /kg) than

5 826 S. Terakura et al. Table 2. Univariate and Multivariate Analyses for Engraftment Significant Factor Hazard Ratio (95% CI)*** P-value Univariate analysis Background Age* 1.01 ( ).007 Body weight* 1.01 ( ).001 Advanced disease 1.34 ( ).14 HLA mismatch (one or more)** 1.47 ( ).067 Conditioning Two conditioning agents 1.24 ( ).26 Fludarabine-based RIC regimen 1.46 ( ).10 Total body irradiation dose* 1.00 ( ).97 Immunosuppression Tacrolimus (vs. cyclosporine) 0.93 ( ).73 The use of methotrexate 0.84 ( ).37 G-CSF administration Yes 0.76 ( ).55 Graft cell dose TNCC > /kg 0.60 ( ).006 CD34 > /kg 0.46 ( ) <.0001 CD3 > /kg 0.63 ( ).015 CD4 > /kg 0.59 ( ).005 CD8 > /kg 0.50 ( ).0002 CD14 > /kg 1.21 ( ).31 CD16/CD56 > /kg 0.64 ( ).017 Multivariate analysis CD34 dose > /kg 0.56 ( ).007 CD8 dose > /kg 0.65 ( ).047 cell dose ( /kg) (Table 3). A trend for a higher incidence of grades III-IV agvhd was found in the patients who received /kg of CD8 cells (hazard ratio: 2.15; 95% CI, ; P.14). RIC indicates reduced intensity conditioning; G-CSF, granulocyte colony-stimulating factor. *Continuous variable. **HLA mismatch in rejection vector. ***A hazard ratio over 1.00 indicates that the category is a risk factor for an engraftment failure. in those receiving this dose or less (median: 22 days; range: days; P.11 on day 21); however, in the group receiving less than the median dose of CD34 cells, the time to achieve neutrophil engraftment was significantly earlier in the patients receiving more than the median dose of CD8 cells (median: 21 days; range: days) than in those receiving less than the median dose (median: 25 days; range: days) (P.0047 at day 21). Finally, no significant factor was identified by either univariate or multivariate analysis for platelet recovery. GVHD Of 130 patients that could be evaluated, agvhd was scored as grade 0 (n 36; 27.7%), I (n 41; 31.5%), II (n 37; 28.5%), III (n 10; 7.7%), or IV (n 6; 4.6%). The cumulative incidences of grades II-IV and III-IV agvhd were 0.41 (95% CI, ) and 0.14 (95% CI, ), respectively. Cox regression analysis showed a significant association between a higher incidence for grade II-IV agvhd and an HLA mismatch in the graft-versus-host (GVH) vector in 1 or more loci and a higher CD8 Figure 2. Cumulative incidence of engraftment. A, Cumulative incidence of engraftment in the patients who received more than median dose of CD34 cells ( /kg). The solid line indicates the patients receiving more than the median dose of CD8 cells ( /kg), and the dotted line indicates the patients receiving a median or lower dose of CD8 cells ( /kg). The overall incidences of engraftment in each patient group were 0.93 (n 53) and 0.85 (n 20), respectively. B, Cumulative incidence of engraftment in the patients who received a median or lower dose of CD34 cells ( /kg). The solid line indicates the patients receiving more than the median dose of CD8 cells, and the dotted line indicates the patients receiving a median or lower dose of CD8 cells. The overall incidences of engraftment in each patient group were 0.90 (n 20) and 0.66 (n 53), respectively.

6 CD34 and CD8 Cell Doses and Engraftment after UCBT 827 Table 3. Multivariate Analyses for GVHD, Relapse, Treatment- Related Mortality, Relapse-Free Survival, and Overall Survival Outcome and Significant Factor Hazard Ratio (95% CI) P-value Grade II-IV acute GVHD HLA mismatch (one or more)* 2.70 ( ).012 CD8 > /kg 2.02 ( ).014 Relapse Advanced disease 2.71 ( ).012 Treatment-related mortality HLA mismatch (one or more)* 8.33 ( ).003 Fludarabine-based RIC regimen 2.29 ( ).008 Use of short-term methotrexate 0.50 ( ).027 Relapse-free survival Advanced disease 1.73 ( ).031 HLA mismatch (one or more)* 2.08 ( ).010 Use of short-term methotrexate 0.55 ( ).005 Overall survival HLA mismatch (one or more)* 2.63 ( ).003 Fludarabine-based RIC regimen 1.81 ( ).021 Use of short-term methotrexate 0.52 ( ).006 GVHD indicates graft-versus-host disease; RIC, reduced intensity conditioning. *HLA mismatch in GVH vector. Of 116 patients that could be evaluated, the cumulative incidence of cgvhd was 0.16 (95% CI, ). Cox regression analysis showed a trend for a higher incidence of cgvhd in the patients who received UCBT from an HLA mismatched donor in the GVH vector in 1 or more loci (hazard ratio: 3.23; 95% CI, ; P.065). There was no significant association between each cell dose and the incidence of cgvhd. Relapse Relapse was observed in 41 of 137 patients treated for malignant disease (median: day 110; range: day 7-655). The cumulative incidence of relapse at 2 years was 0.31 (95% CI, ) for all patients with malignant disease, 0.31 ( ) for patients with ALL (n 49), and 0.33 ( ) for patients with AML (n 39). Cox regression analysis revealed a significant association between advanced disease at transplantation and a higher incidence of relapse (Table 3). There was no significant association between each cell dose and the incidence of relapse. Treatment-Related Mortality The cumulative incidences of TRM at day 100 and 365 were 0.25 (95% CI, ) and 0.35 ( ), respectively. Cox regression analysis showed a significant association between a higher TRM and an HLA mismatch in the GVH vector in 1 or more loci, fludarabine-based RIC regimen, and not using shortterm methotrexate (Table 3). There was no significant association between each cell dose and TRM. Relapse-Free Survival and Overall Survival The cumulative incidences of relapse-free survival and OS at 3 years after transplantation were 0.33 (95% CI, ) and 0.37 ( ), respectively. The median follow-up time for surviving patients was 24 months (range: 2-62). Cox regression analysis revealed a significant association between a lower relapse-free survival rate with advanced disease, an HLA mismatch in the GVH vector in 1 or more loci, and not using short-term methotrexate (Table 3). There was also a significant association between a lower OS and an HLA mismatch in the GVH vector in 1 or more loci, a fludarabine-based RIC regimen, and not using short-term methotrexate (Table 3). There was no significant association between each cell dose and relapse-free survival or OS. DISCUSSION In UCBT, engraftment in the recipients is an important surrogate marker because of there is a lower engraftment rate and a delayed blood recovery compared to bone marrow or PBSC transplantation [6,7]. We analyzed the influence of CD34 and other accessory cell doses in UCB grafts on the engraftment rate and kinetics in recipients, and we demonstrated that a higher CD34 cell dose is associated with successful engraftment. In patients receiving more than /kg of CD34 cells, the engraftment occurred 4 days earlier than in patients receiving /kg or less CD34 cells. Wagner et al. [13] also reported that the CD34 cell dose is associated with engraftment in recipients of UCBT. They showed that the rate and probability of engraftment were markedly lower in patients receiving /kg of CD34 cells compared with those receiving a higher dose of these cells. These findings suggest that the CD34 cell dose in UCB grafts is a definitive predictor of the engraftment rate and kinetics after UCBT and that the threshold dose for assured and rapid engraftment is approximately /kg, although differences in methodologies for measuring CD34 cells could influence the exact threshold dose. A few studies have attempted to show an influence of the CD34 cell dose on engraftment after UCBT, but most of them have not been able to confirm that there is an effect [11,29]. These studies have determined the CD34 cell dose using different techniques and institutions [11] or in small populations ( 30 patients) [29]. In contrast, Wagner et al. [13] estimated the infused CD34 cell dose by flow cytometry of an aliquot of the infused UCB cells under a fixed condition at a single institution and in a larger study population (n 102). We also estimated the cell dose by measuring the percentage of CD34 cells in UCB test vials at a single institution, although there is no

7 828 S. Terakura et al. assurance that the compositions of the test vial and the unit actually thawed and infused into the patients were identical. It is well known that the measurement of CD34 cells cannot be used for comparative studies between transplant centers because of the absence of standard thawing, sampling, and counting methods. The use of a standardized technique with sufficient reproducibility is essential for accurately evaluating the association between cell components in the UCB graft and their role in UCBT. We showed for the first time that a higher dose of CD8 cells ( /kg) in UCB grafts is associated with successful engraftment. Stratification of the group according to CD34 cell dose revealed that there is a significant influence of CD8 cells on the median time to achieve neutrophil engraftment in the patients receiving lower CD34 cell doses (P.0047), whereas there was no significant influence in the patients receiving a higher CD34 cell dose (P.11). These results suggest CD8 cells have a supportive role in the engraftment of UCB containing fewer HSCs. The mechanism of action by which CD8 T cells facilitate engraftment has been extensively studied in BMT [30]. Earlier studies demonstrated that donor CD8 T cells eliminate residual immune cells of the recipient via perforin or a Fas ligand-dependent mechanism [20,31]. Several subpopulations of CD8 T cells have been reported to facilitate engraftment. Kaufman et al. [32] reported that CD8 T cells expressing T cell receptor (TCR)-associated CD3 molecules but not TCRs can facilitate marrow engraftment without causing GVHD. Gandy et al. [33] later reported that TCR T cells also facilitate engraftment of highly enriched HSCs in allogeneic recipients. It remains possible that other mechanisms such as veto activity as defined by Miller [34] contribute to the graft-enhancing effect. Veto cells can induce apoptosis of cytotoxic T cells that are directed against antigens of the veto cells via a Fas-Fas ligand interaction [35]. Some of the most potent veto cells are of T cell origin, and in particular, CD8 cytotoxic T lymphocytes have a very high veto activity [36,37]. We showed here that each UCB unit contains a considerable number of CD8 cells (median, /kg), which is 1 to 2 orders of magnitude more than cells obtained from T cell-depleted bone marrow by CD34 positive selection (typically 10 5 CD8 cells/kg). T cells in UCB are more naive than adult lymphocytes [38], but, as described above, expression of the TCR is not necessarily required for facilitation of engraftment by T cells. Overall, these results are consistent with the idea that CD8 T cells play an important role in facilitating HSC engraftment in transplantation of UCB. Although both CD34 cells and CD8 cells had a beneficial effect on engraftment, neither cell population affected survival in UCBT recipients. This may result from the heterogeneity of the study population. The patients had a variety of diseases and severity of disease, and they received a variety of preconditioning regimens and prophylaxis for GVHD (Table 1). A study population with a more homogenous background is required to define the contribution of these cell populations on survival after transplantation. Thus, our results do not exclude the possibility that CD34 and/or CD8 cell doses influence survival in UCBT recipients. Interestingly, the use of methotrexate was associated with lower rates of TRM and higher relapse-free survival and OS (Table 3). Although it was not associated with lower incidence of agvhd in this study population (Table 2), we previously demonstrated that use of methotrexate has a significant favorable effect on post-ucbt immune reactions and OS in a different study population [39]. In contrast, Locatelli et al. [40] reported that the use of methotrexate is unfavorable for neutrophil engraftment and event-free survival in a study of 44 pediatric patients with thalassemia and sickle cell disease. Accordingly, the effect of methotrexate on UCBT outcome remains unclear. It might be worth further investigation to determine whether there is an explanation for the finding that the use of methotrexate is associated with a lower risk of TRM and higher rate of survival. A fludarabine-based RIC regimen was a risk factor for TRM. Because we used registry data in this study, the patients who were at high risk for TRM might have tended to receive an RIC regimen rather than a myeloablative regimen. In addition, the RIC regimens used in this study greatly varied. Thus, another study is required to determine whether the RIC regimen decreases TRM and improves the survival rate in the patients who are at high risk for TRM. We studied the effect of the dose of CD34 and other cells on engraftment in patients receiving a single unmanipulated UCB graft and an immunosuppressant other than ATG. Antithymocyte globulin as a preconditioning or posttransplantation immunosuppressive treatment would considerably attenuate the influence of T cells, but the role of natural killer cells could be expected to become more important [41,42]. Thus, the present results should be interpreted with caution for patients treated with ATG or other T cell depletion procedures. In summary, the present study demonstrated that both CD34 and CD8 cells have a beneficial effect on engraftment in UCBT recipients, and our findings raise the possibility that CD8 cells have a supportive role in engraftment of UCB units containing fewer HSCs. Further analysis in a larger population is warranted. If our findings are confirmed, if all of the UCB

8 CD34 and CD8 Cell Doses and Engraftment after UCBT 829 units accompany their test vial, and if the measurements of CD34 and CD8 cells are standardized, it may be worthwhile to consider the doses of both CD34 and CD8 cells as well as TNCC and HLA matching for UCB graft selection strategies. ACKNOWLEDGMENTS This work was supported by a Research Grant on Human Genome, Tissue Engineering from the Ministry of Health, Labour and Welfare of Japan (to E.A. and M.M.). We would like to thank the staff of the Japan Cord Blood Network and its attendant transplant centers. We are especially grateful for the help of Drs. Kiyoaki Suzuki, Tadayuki Ishimaru, Kaoru Ishikawa, and Takami Inoue in collecting the UCB, as well as the assistance of Mses. Tomoko Ito, Makiko Shibayama, Hiromi Ogawa, and Nobuko Ishikawa in processing the UCB units. REFERENCES 1. Kurzberg J, Laughlin M, Graham ML, et al. Placental blood as a source of hematopoietic stem cells for transplantation into unrelated donors. N Engl J Med. 1996;335: Wagner JE, Rosenthal J, Sweetman R, et al. Successful transplantation of HLA-matched and HLA-mismatched umbilical cord blood from unrelated donors: analysis of engraftment and acute graft-versus-host disease. Blood. 1996;88: Gluckman E, Rocha V, Boyer-Chammard A, et al. Outcome of cord-blood transplantation from related and unrelated donors: Eurocord Transplant Group and the European Blood and Marrow Transplantation Group. N Engl J Med. 1997;337: Rubinstein P, Carrier C, Stevens CE, et al. Outcomes among 562 recipients of placental-blood transplants from unrelated donors. N Engl J Med. 1998;339: Laughlin MJ, Eapen M, Rubinstein P, et al. Outcomes after transplantation of cord blood or bone marrow from unrelated donors in adults with leukemia. N Engl J Med. 2004;351: Rocha V, Labopin M, Sanz G, et al. Acute Leukemia Working Party of European Blood and Marrow Transplantation Group; Eurocord-Netcord Registry. Transplants of umbilical-cord blood or bone marrow from unrelated donors in adults with acute leukemia. N Engl J Med. 2004;351: Takahashi S, Iseki T, Asano S, et al. Single-institute comparative analysis of unrelated bone marrow transplantation and cord blood transplantation for adult patients with hematologic malignancies. Blood. 2004;104: Barker JN, Wagner JE. Umbilical cord blood transplantation: current state of the art. Curr Opin Oncol. 2002;14: Rocha V, Gluckman E. Clinical use of umbilical cord blood hematopoietic stem cells. Biol Blood Marrow Transplant. 2006; 12: Barker JN, Krepski TP, Weisdorf DJ, et al. Searching for unrelated donor hematopoietic stem cells: availability and speed of umbilical cord blood versus bone marrow. Biol Blood Marrow Transplant. 2002;8: Laughlin MJ, Barker JN, Bambach B, et al. Hematopoietic engraftment and survival in adult recipients of umbilical-cord blood from unrelated donors. N Engl J Med. 2001;344: Gluckman E, Rocha V, Chevret S, et al. Factors associated with outcomes of unrelated cord blood transplant: guidelines for donor choice. Exp Hematol. 2004;32: Wagner JE, Barker JN, Davies SM, et al. Transplantation of unrelated donor umbilical cord blood in 102 patients with malignant and nonmalignant diseases: influence of CD34 cell dose and HLA disparity on treatment-related mortality and survival. Blood. 2002;100: Sanz MA, Sanz GF. Unrelated donor umbilical cord blood transplantation in adults. Leukemia. 2002;16: Kernan NA, Flomenberg N, O Reilly RJ, et al. Graft rejection in recipients of T cell depleted HLA-nonidentical marrow transplants for leukemia: identification of host derived anti-donor allocytotoxic T lymphocytes. Transplantation. 1987;43: Martin PJ, Hansen JA, Buckner CD, et al. Effect of in vitro depletion of T cells in HLA-identical allogeneic marrow grafts. Blood. 1985;66: Champlin R, Ho W, Gajewski J, et al. Selective depletion of CD8 T lymphocytes for prevention of graft-versus-host disease after allogeneic bone marrow transplantation. Blood. 1990; 76: Martin PJ, Rowley SD, Anasetti C, et al. A phase I-II clinical trial to evaluate removal of CD4 cells and partial depletion of CD8 cells from donor marrow for HLA-mismatched unrelated recipients. Blood. 1999;94: Ildstad ST, Wren SM, Bluestone JA, Barbieri SA, Stephany D, Sachs DH. Effect of selective T-cell depletion of host and/or donor bone marrow on lymphopoietic repopulation, tolerance, and graft versus host disease in mixed allogeneic chimeras (B10 B10.D2 B10). J Immunol. 1986;136: Martin PJ. Donor CD8 cells prevent allogeneic marrow graft rejection in mice: potential implications for marrow transplantation in humans. J Exp Med. 1993;178: Sutherland DR, Anderson L, Chin-Yee I. The ISHAGE guidelines for CD34 cell determination by flow cytometry: International Society of Hematoherapy and Graft Engineering. J Hematother. 1996;5: Centers for Disease Control and Prevention Revised guidelines for performing CD4 T-cell determinations in persons infected with human immunodeficiency virus (HIV). MMWR Recomm Rep. 1997;46: Kobayashi M, Azuma E, Ido M, et al. A pivotal role of Rho GTPase in the regulation of morphology and function of dendritic cells. J Immunol. 2001;167: Przepiorka D, Weisdorf D, Thomas ED, et al Consensus conference on acute GVHD grading. Bone Marrow Transplant. 1995;15: Sullivan KM, Shulman HM, Thomas ED, et al. Chronic graftversus-host disease in 52 patients: adverse natural course and successful treatment with combination immunosuppression. Blood. 1981;57: Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. Am Stat Assoc J. 1958;53: Gooley TA, Leisenring W, Crowley JA, Storer BE. Estimation of failure probability in the presence of competing risks: new representations of old estimators. Stat Med. 1999;18: Cox DR. Regression models and life-tables. J R Stat Soc B. 1972;34:

9 830 S. Terakura et al. 29. Sanz GF, Saavedra S, Planelles D, et al. Standardized, unrelated donor cord blood transplantation in adults with hematologic malignancies. Blood. 2001;98: Martin PJ. Winning the battle of graft versus host. Nat Med. 2000;6: Martin PJ, Akatsuka Y, Hahne M, Sale G. Involvement of donor T-cell cytotoxic effector mechanisms in preventing allogeneic marrow graft rejection. Blood. 1998;92: Kaufman CL, Colson YL, Wren SM, Watkins S, Simmons RL, Ildstad ST. Phenotypic characterization of a novel bone marrow-derived cell that facilitates engraftment of allogeneic bone marrow stem cells. Blood. 1994;84: Gandy KL, Domen J, Aguila H, Weissman IL. CD8 TCR and CD8 TCR-cells in whole bone marrow facilitate the engraftment of hematopoietic stem cells across allogeneic barriers. Immunity. 1999;11: Miller RG. An immunological suppressor cell inactivating cytotoxic T-lymphocyte precursor cells recognizing it. Nature. 1980;287: Reich-Zeliger S, Zhao Y, Krauthgamer R, Bachar-Lustig E, Reisner Y. Anti-third party CD8 CTLs as potent veto cells: coexpression of CD8 and FasL is a prerequisite. Immunity. 2000;13: Fink PJ, Shimonkevitz RP, Bevan MJ. Veto cells. Annu Rev Immunol. 1988;6: Bachar-Lustig E, Reich-Zeliger S, Reisner Y. Anti-third-party veto CTLs overcome rejection of hematopoietic allografts: synergism with rapamycin and BM cell dose. Blood. 2003;102: Cohen SB, Perez-Cruz I, Fallen P, Gluckman E, Madrigal JA. Analysis of the cytokine production by cord and adult blood. Hum Immunol. 1999;60: Narimatsu H, Terakura S, Matsuo K, et al. Short-term methotrexate could reduce early immune reactions and improve outcomes in umbilical cord blood transplantation for adults. Bone Marrow Transplant. 2007;39: Locatelli F, Rocha V, Reed W, et al. Related umbilical cord blood transplantation in patients with thalassemia and sickle cell disease. Blood. 2003;101: Ruggeri L, Capanni M, Velardi A, et al. Effectiveness of donor natural killer cell alloreactivity in mismatched hematopoietic transplants. Science. 2002;295: Farag SS, Fehniger TA, Caligiuri MA, et al. Natural killer cell receptors: new biology and insights into the graft-versus-leukemia effect. Blood. 2002;100: