Scholarly article on topic 'Effect of CD8+ Cell Content on Umbilical Cord Blood Transplantation in Adults with Hematological Malignancies'

Effect of CD8+ Cell Content on Umbilical Cord Blood Transplantation in Adults with Hematological Malignancies Academic research paper on "Clinical medicine"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{"Single umbilical cord blood" / Transplantation / Engraftment / "Cell dose" / Adults / "Hematological malignancies"}

Abstract of research paper on Clinical medicine, author of scientific article — Federico Moscardó, Jaime Sanz, Francisco Carbonell, Miguel A. Sanz, Luis Larrea, et al.

Abstract Total nucleated (TNCs) and CD34+ cells are considered major determinants of outcome after umbilical cord blood (UCB) transplantation but the effect of other cell subtypes present in the graft is unknown. This single-center cohort study included patients with hematological malignancies who received UCB transplantation after a myeloablative conditioning regimen. UCB units were primarily selected according to cell content, both TNCs and CD34+ cells, and also according to the degree of HLA matching. Counts of several cell subtypes of the infused UCB unit, together with HLA disparities and other patient- and transplantation-related characteristics, were analyzed by multivariable methodology for their association with myeloid and platelet engraftment, graft-versus-host disease, nonrelapse mortality (NRM), disease-free survival (DFS), and overall survival (OS). Two hundred patients (median age, 32 years) were included in the study. In multivariable analyses, a greater number of CD8+ cells was significantly associated with better results for myeloid (P = .001) and platelet (P = .008) engraftment, NRM (P = .02), DFS (P = .007), and OS (P = .01). CD34+ cell content was predictive of myeloid engraftment (P < .001). This study suggests that the outcome after UCB transplantation in adults with hematological malignancies could be better when UCB grafts had a greater CD8+ cell content.

Academic research paper on topic "Effect of CD8+ Cell Content on Umbilical Cord Blood Transplantation in Adults with Hematological Malignancies"

AS BMI

American Society for Blood and Marrow Transplantation

Biology of Blood and Marrow Transplantation

journal homepage: www.bbmt.org

Effect of CD8+ Cell Content on Umbilical Cord Blood crossMark

Transplantation in Adults with Hematological Malignancies r°ss ar

Federico Moscardó1,*, Jaime Sanz1, Francisco Carbonell2, Miguel A. Sanz1, Luis Larrea2, Pau Montesinos1, Ignacio Lorenzo1, Belén Vera1, Blanca Boluda1, Claudia Salazar1, Carolina Cañigral 1, Dolores Planelles 2, Isidro Jarque 1, Pilar Solves 1, Guillermo Martín 1, Francisca López1, Javier de la Rubia1, Jesús Martínez1, Nelly Carpio1, David Martínez-Cuadrón1, Nieves Puig2, José A. Montoro2, Roberto Roig2, Guillermo F. Sanz1

1 Hematopoietic Cell Transplantation Unit, Hematology Department, Hospital Universitario y Politécnico La Fe, Valencia, Spain

2 Centro de Trasfusión de la Comunidad Valenciana, Valencia, Spain

Article history: Received 13 March 2014 Accepted 30 June 2014

Key Words:

Single umbilical cord blood Transplantation Engraftment Cell dose

Adults

Hematological malignancies

ABSTRACT

Total nucleated (TNCs) and CD34+ cells are considered major determinants of outcome after umbilical cord blood (UCB) transplantation but the effect of other cell subtypes present in the graft is unknown. This singlecenter cohort study included patients with hematological malignancies who received UCB transplantation after a myeloablative conditioning regimen. UCB units were primarily selected according to cell content, both TNCs and CD34+ cells, and also according to the degree of HLA matching. Counts of several cell subtypes of the infused UCB unit, together with HLA disparities and other patient- and transplantation-related characteristics, were analyzed by multivariable methodology for their association with myeloid and platelet engraftment, graft-versus-host disease, nonrelapse mortality (NRM), disease-free survival (DFS), and overall survival (OS). Two hundred patients (median age, 32 years) were included in the study. In multivariable analyses, a greater number of CD8+ cells was significantly associated with better results for myeloid (P = .001 ) and platelet (P = .008) engraftment, NRM (P = .02), DFS (P = .007), and OS (P = .01). CD34+ cell content was predictive of myeloid engraftment (P < .001). This study suggests that the outcome after UCB transplantation in adults with hematological malignancies could be better when UCB grafts had a greater CD8+ cell content.

© 2014 American Society for Blood and Marrow Transplantation.

INTRODUCTION

Umbilical cord blood (UCB) transplantation (UCBT) is a curative procedure for patients with hematological disorders who require an allogeneic hematopoietic cell transplantation but lack a suitable adult donor [1].

Cell dose and degree of HLA matching between the UCB unit and the recipient are critical determinants of the outcome after UCBT and used for guiding the selection of the UCB unit [1-3]. The cell dose measurement universally employed is the number of total nucleated cells (TNCs) because it is reproducible, has shown prognostic impact [4,5], and constitutes a good surrogate of CD34+ cell content and the potency of the graft [3]. However, CD34+ cell dose has only occasionally been associated with long-term

Financial disclosure: See Acknowledgments on page 1750.

* Correspondence and reprint requests: Federico Moscardó, MD, PhD, Hematopoietic Cell Transplantation Unit, Hematology Department, Hospital Universitario y Politécnico La Fe, Valencia, Spain.

E-mail address: moscardo_fed@gva.es (F. Moscardó).

outcome [6] and the relevance of other cell subtypes remains undefined.

The aim of this study was to analyze the impact on outcome of different cell populations of the UCB unit in a large series of adult patients with hematological malignancies undergoing UCBT at a single institution. CD8+ cell content was critical for success.

METHODS

Patient eligibility, graft selection, conditioning regimens, immune suppression, and supportive care have been previously reported [7,8] and are summarized below.

Patients

All patients with hematological malignancies undergoing myeloablative single-unit UCBT (sUCBT) at our institution between May 1997 and January 2012 were included in the study.

Patients with hematological malignancies were eligible for enrolment if they met the following criteria: (1) allogeneic hematopoietic cell transplantation was considered the best therapeutic option, (2) a suitable related donor (HLA identical or 1 antigen mismatched) was not available, (3) there was a lack of a suitable HLA-matched unrelated donor at a reasonable time

1083-8791/$ — see front matter © 2014 American Society for Blood and Marrow Transplantation. http://dx.doi.org/10.1016/j.bbmt.2014.06.038

after the start of the search through international registries, and (4) there was a suitable UCB unit available, as described below.

Patients receiving ex vivo expanded grafts were excluded. The institutional review board approved the protocol and written informed consent was obtained from all patients, according to the Declaration of Helsinki.

Cord Blood Unit Selection

The search of UCB units was conducted by the Spanish Registry of Bone Marrow Donors (Registro Español de Donantes de Médula Ósea). A degree of HLA matching between the UCB unit and the recipient greater or equal to 4 of 6 (considering HLA-A and -B at antigen level and -DRB1 at allele level) was required. Those units fulfilling HLA-matching criteria were ranked according to their content of TNCs and CD34+ cells. The minimum number of TNCs and CD34+ cells required for selection of UCB units changed over time. Until 2005, only a TNC dose >1.5 x 107/kg recipient's body weight was required (71 patients, 36%). During this early period, CD34+ cell content was also closely monitored, but a threshold for UCB unit selection was not established. In 2006 and 2007, the numbers of TNCs and CD34+ cells required were >2 x 107/kg and >1 x 105/kg recipient's body weight (47 patients, 24%), respectively. Since 2008, the minimum cell dose requirements are TNCs >150 x 107 and CD34+ cells >70 x 105, without taking into account recipient's body weight (82 patients, 41%). Cell dose was always considered the most important criteria for UCB unit selection.

Conditioning Regimen, Graft-versus-Host Disease Prophylaxis and Treatment, and Supportive Care

Two myeloablative conditioning regimens were used and have been described in detail elsewhere [7,8]. Briefly, the first 71 patients received thiotepa (10 mg/kg), busulfan (12 mg/kg orally or 9.6 mg/kg i.v.), and cyclophosphamide (120 mg/kg), whereas the last 129 patients received the same preparative regimen but cyclophosphamide was replaced by fludar-abine (150 mg/m2). All patients received antithymocyte globulin (ATG); horse ATG (Lymphoglobuline, Merieux, Lyon, France; total dose, 60 mg/kg) in the first 32 patients (16%) and rabbit ATG (Thymoglobulin, Sangstat/ Genzyme, Lyon, France; total dose, 6 to 8 mg/kg) in the last 168 patients (84%).

All patients received cyclosporine 1.5 mg/kg/12 hours i.v., followed by 3 to 5 mg/kg/12 hours orally, when oral intake was possible, with slow tapering starting between day +90 and +180 and with discontinuation on day +180 or before if feasible. Cyclosporine was combined with prednisone in 150 patients (.5 mg/kg from day +7 to +14,1 mg/kg from day +14 to +28, and then slowly tapered in 120 cases, and 1 mg/kg from day +14 to +28 in 30 patients) or with mycophenolate mofetil (MMF) (15 mg/kg/12 hours until day +28) in the remaining 50 patients. Patients developing acute graft-versus-host disease (GVHD) received high-dose methylprednisolone as initial therapy (2 to 20 mg/kg/day) and ATG was used in refractory cases [7]. Chronic GVHD was treated with prednisone (1 mg/kg/day).

Patients were kept in reverse isolation under high-efficiency particulate air filtration. Ciprofloxacin and cotrimoxazole were given as antibacterial prophylaxis. For antifungal prophylaxis, fluconazole, itraconazole, and vor-iconazole were sequentially used during the study period. The use of intravenous antibiotics and antifungal agents and the transfusion policy were those usually employed when handling neutropenic patients. All transfused products were irradiated and depleted of leukocytes. As cyto-megalovirus (CMV) prophylaxis, all patients received intravenous acyclovir (400 mg/m2 every 12 hours) from day -5 until engraftment, followed by intravenous ganciclovir (5 mg/kg per day 3 to 5 days per week) or oral valganciclovir (900 mg per day) from engraftment until day 100 after transplantation. CMV infection surveillance and treatment have been described in detail elsewhere [8]. Nonspecific intravenous immunoglobulins (500 mg/kg) were administered weekly until day +100 and monthly thereafter during the first year after the transplantation.

Cellular Assays

UCB units were thawed and washed using the standard method of Rubinstein et al., slightly modified [7]. Samples for cell counts, including TNCs, CD34+, CD3+, CD4+, CD8+, CD16+, CD56+, and CD19+ cells, were drawn directly from the UCB unit bag(s) after thawing and washing and before infusion.

The CD34+ cell content was quantified by flow cytometry with CD34 and CD45 monoclonal antibodies conjugated to phycoerythrin and fluorescein, respectively (Becton Dickinson, San Jose, CA). The lymphoid population was analyzed using CD19 and CD3 monoclonal antibodies for total B and T lymphocytes, respectively. T lymphocytes subpopulations were analyzed using CD8, CD4, CD16, and CD56 monoclonal antibodies (Becton Dickinson). Flow cytometry analysis was performed on the FACSCalibur and FACScanto cytometer (Becton Dickinson). Acquisition and analysis of events were done

by using Cellquest and FACSDiva software, respectively (Becton Dickinson). A double platform system was used for cell assays. The automated cell counter used was Sysmex K21 (Roche, Basilea, Switzerland).

Definitions

Neutrophil engraftment was defined as an absolute neutrophil count equal or greater than .5 x 109/L on 3 consecutive days. Platelet engraftment was defined as a platelet count equal or greater than 20 x 109/L without transfusional support for 7 consecutive days. Patients who died 28 days after transplantation without neutrophil engraftment were considered graft failures. Time to neutrophil or platelet engraftment was defined as the time required to reach the first day of engraftment. Acute and chronic graft-versus-host disease (GVHD) were defined and graded according to standard criteria [9-11]. Disease status at the time of transplantation was classified as follows: (1) early stage: acute leukemia, myelodysplastic syndrome, and lymphoma in the first complete remission untreated myelodysplastic syndrome with <5% blasts and/or chronic myelogenous leukemia (CML) in the first chronic phase; (2) intermediate stage: acute leukemia, lymphoma, or myelodysplastic syndrome in a second or further compete remission and CML in a second or further chronic or accelerated phase; and (3) advanced stage: acute leukemia and lymphoma not in remission, CML in blast crisis, and untreated refractory anemia with excess blasts. Nonrelapse mortality (NRM) was defined as death from any cause without evidence of relapse. Disease-free survival (DFS) was defined as survival from the time of transplantation without evidence of disease relapse. Overall survival (OS) was defined as survival from the time of transplantation.

Prognostic Factors

Variables considered for prognostic factor analysis were age, gender, recipient weight, recipient CMV serology, diagnosis, disease status at transplantation, HLA match, ABO blood group compatibility, conditioning regimen, GVHD prophylaxis, and TNCs, CD34+, CD3+, CD4+, CD8+, CD16+/ CD56+, and CD19+ cells infused. Continuous variables were introduced as such in multivariable regression procedures but were dichotomized at their most discriminative cut-off point for depicting their effect on univariable analyses.

Statistical Analysis

Correlation between variables was analyzed by Spearman's correlation test. The variable inflation factor was estimated to quantify the severity of possible colinearity between the different cell subsets. Probabilities of engraftment, NRM, and GVHD, and relapse were estimated by the cumulative incidence method [12]. Unadjusted time-to-event analyses were performed using the Kaplan-Meier estimate, and, for comparisons, the log-rank test was used [13]. Cox proportional hazard regression method and Fine and Gray method for competing events were used for multivariable analysis [14]. For the Cox models, the proportional hazards assumption was tested and confirmed to be satisfied, by correlating scaled Schoenfeld residuals with a transformation of time. Characteristics selected for inclusion in the multivariable model were those with some indication of association in univariable analysis (P < .10). Only cases with complete data (n = 189) were considered in multivariable modeling. The forward stepwise procedure was stopped when the P value for entering an additional variable was above .05. The follow-up of the patients was updated on April 1,2013 and all follow-up data were censored at that point. Two-sided P values <.05 were considered statistically significant and no adjustment for multiple testing was applied. Statistical analysis were conducted using R version 2.12.2 (The CRAN project) with packages car v2.0-16, rms v3.6-3, prodlim v1.3.3, cmprsk v2.2-4, survival v2.37-4 and fmsb v0.4.1.

RESULTS

Patient, Graft, and Transplantation Characteristics

Data from 200 transplantations fulfilling the inclusion criteria were analyzed. Patient, graft, and transplantation characteristics are shown in Table 1. Briefly, median age was 32 years (range, 15 to 55), 145 patients (73%) had acute leukemia, 60 (30%) underwent transplantation while in an advanced phase, and 139 (70%) received UCB units mismatched at 2 HLA loci.

Cell Content of the Umbilical Cord Blood Units

Table 2 summarizes the dose of the different cell populations present in the grafts. The R correlation coefficient between TNCs infused and CD34+, CD3+, CD8+, and CD4+ cells infused was .66, .66, .53, and .65, respectively (P < .01 for

Table 1

Patient and Transplantation Characteristics

Characteristics

Age, median (range), yr 32 15-55)

Patient's gender

Male 127 63.5)

Female 73 36.5)

Diagnosis

Acute myeloid leukemia 73 36.5)

Acute lymphoblastic leukemia 72 (36)

Myelodysplastic syndrome 10 (5)

Chronic myeloid leukemia 26 (13)

Chronic lymphoproliferative disorder 14 (7)

Other 5 (2.5)

Disease status*

Early 88 (44)

Intermediate 52 (26)

Advanced 60 (30)

Recipient's body weight, median (range), kg 72 37-112)

CMV serologic status

Positive 157 (78.5)

Negative 43 21.5)

Previous autologous HCT

Yes 23 11.5)

No 177 88.5)

Conditioning regimen

TT + BU + CY 71 35.5)

TT + BU + FLU 129 64.5)

Type of ATG

Equine 32 16)

Rabbit 168 84)

GVHD prophylaxis

CSA + MMF 50 25)

CSA + PRED 150 75)

Donor-receptor gender disparity

Male-male 64 32)

Male-female 35 18)

Female-female 37 19)

Female-male 63 32)

ABO group incompatibility

No 93 46.5)

Major incompatibility 51 25.5)

Minor incompatibility 56 28)

HLA match

6 out of 6 11 5)

5 out of 6 50 25)

4 out of 6 139 70)

HCT indicates hematopoietic cell transplantation; TT, thiotepa; BU, busulfan; CY, cyclophosphamide; FLU, fludarabine; CSA, cyclosporine; PRED, prednisone.

Data presented are n (%), unless otherwise indicated.

* Definitions of early, intermediate, and advance stage disease status at the time of transplantation are shown in Supplemental Material. y By considering HLA-A and -B at antigen level and -DRB1 at allele level.

all comparisons) and between CD34+ cells infused and CD3+, CD8+, and CD4+ cells infused was .48, .41, and .47, respectively (P < .01 for all comparisons). However, according to

Table 2

Cell Content of the Cord Blood Units

Cell Type

Median (Range)

At time of freezing

TNCs x 107 per kg 2.9 (1.4-7.5)

CD34+ cells x 105 per kg 1.6 .2-6.8)

At time of infusion

TNCs x 107 per kg 2.4 (1-5.9)

CD34+ cells x 105 per kg 1.3 (.1-10.1)

Lymphocytes x 106 per kg 9.9 (2.8-32.7)

CD3+ cells x 106 per kg 5.6(1.5-17.4)

CD4+ cells x 106 per kg 4.1 (.1-12.1)

CD8+ cells x 106 per kg 1.6 (.5-10.2)

CD19+ cells x 106 per kg 1.9 (.2-10.9)

CD16+ cells x 106 per kg 2.1 (.4-9.2)

variable inflation factor results, there was no statistically significant colinearity between TNCs, CD34+ cells, and CD8+ cells.

Predictive Factors of Outcome

A summary of the results of multivariable analyses for the different transplantation outcomes is offered in Table 3. The data of univariable analyses are presented as web extra material (Supplementary material).

Neutrophil engraftment

Eight patients (4%) died before day 28 after transplantation (median time, 16 days; range, 9 to 24) and were considered not evaluable for engraftment. Nine of 192 patients (5%) had graft failure and the remaining 183 patients (95%) achieved neutrophil engraftment at a median of 21 days (range, 11 to 57). The cumulative incidence of neutrophil engraftment at day 60 after transplantation was 92% (Figure 1A). Multivariable analysis showed that higher CD34+ (P < .001 ) and CD8+ (P = .001 ) cell dose, diagnosis of acute leukemia (P = .01), and negative CMV serological status of the recipient (P = .01) were independently associated with improved cumulative incidence of engraft-ment. Figure 1B to 1D depict the relationship between CD8+ and CD34+ cell doses and neutrophil engraftment. The median number of infused CD8+ cells in patients without evidence of engraftment was 1.2 x 106/kg (range, .5 to 3.3).

Platelet engraftment

Platelet recovery was achieved in 135 patients at a median time of 51 days after UCBT (range, 23 to 188). The cumulative incidence of platelet engraftment was 67% (Figure 2A). Multivariable analysis showed that a higher CD8+ cell dose at infusion (P = .008) and use of fludarabine rather than cyclophosphamide in the conditioning regimen (P = .03) were independently associated with improved cumulative incidence of platelet recovery. The effect of CD8+ cell dose on platelet recovery is shown in Figure 2B.

Acute and chronic GVHD

The cumulative incidence of acute GVHD grade II to IV and grade III to IV at 100 days was 43% and 25%, respectively. The cumulative incidence at 4 years of chronic and chronic extensive GVHD was 61% and 41%, respectively. In multivariable analysis, the incidence of acute grade II to IV GVHD was lower in patients diagnosed of acute leukemia (P = .004) and in those receiving rabbit ATG (P = .0003), whereas the incidence of chronic GVHD was lower in patients receiving major ABO incompatible grafts (P = .008) and in those receiving a higher CD19+ cell dose (P = .007).

Relapse risk

Fifty-one patients relapsed after UCBT, with a cumulative incidence of relapse at 4 years of 24%. The 4-year cumulative incidence of relapse risk was higher in patients diagnosed of acute leukemia (P = .0006) and in patients who underwent transplantation in advanced phase (P = .004).

After a median follow-up of79 months (range, 17 to 180), 144 patients (72%) have died, 96 (48%) without relapsing and 48 (24%) after relapse. The 4-year NRM was 46% (95% confidence interval [CI], 39% to 53%). Multivariable analysis

Table 3

Factors Affecting the Outcome after UCBT in Multivariable Analysis

Favorable Variables for Outcomes

HR (95% CI)

P Value

Myeloid engraftment

Diagnosis of acute leukemia 1.53 (1.09-2.14) .01

CMV status negative 1.57 (l.1-2.24) .01

CD34+ cells infused 1.19(l.07-1.32) .001

CD8+ cells infused 3.5 (1.72-7.15) <.001

Platelet engraftment

CD8+ cells infused 3.25 (1.36-7.79) .008

Use of fludarabine 1.5 (1.03-2.16) .03

Nonrelapse mortality

CD8+ cells infused .1 (.02-.7) .02

Diagnosis of acute leukemia .52 (.35-.78) .002

Relapse risk

Diagnosis of acute leukemia 13 (3-56.27) <.001

Early disease stage 2.46 (1.34-4.52) .004

Disease-free survival

CD8+ cells infused 1.8 (1.2-2.7) .006

Nonadvanced disease status* 2.1 (1.5-3) <.001

Overall survival

CD8+ cells infused 1.7 (1.1-2.5) .01

Nonadvanced disease status 2.1 (1.5-3) <.001

HR indicates hazard ratio.

* Nonadvanced disease status includes early and intermediate disease status at transplantation (see definitions in Supplemental Material).

showed that a higher CD8+ cell dose (P = .02) and a diagnosis of acute leukemia (P = .002) were associated with a lower NRM. The impact of CD8+ cell dose on NRM is drawn on Figure 3A.

DFS and OS

The 4-year DFS and OS of the entire cohort were 29% (95% CI, 22% to 35%) and 31% (95% CI, 25% to 38%), respectively. Variables associated with improved DFS and OS in multivariable analysis were a higher CD8+ cell dose (P = .006 for DFS and P = .01 for OS) and nonadvanced disease status (P < .001 for DFS and P < .001 for OS). The effect of CD8+ cell dose on DFS and OS is depicted in Figure 3B and 3C, respectively.

The dose of cryopreserved TNCs and CD34+ cells did not show any significant relationship with GVHD, NRM, relapse risk, DFS, or OS. The cryopreserved CD34+ cells only influenced myeloid recovery in univariable analysis.

DISCUSSION

This study shows that the number of CD8+ cells of the graft is a major and independent determinant of neutrophil and platelet engraftment, NRM, DFS, and OS after myeloa-blative sUCBT in adult patients with hematological malignancies. This novel finding is clinically relevant because the inclusion of CD8+ cell content among the factors used for UCB unit selection could improve the results of UCB transplantations.

Cell dose and HLA match are universally recognized as the most relevant criteria for UCB unit choice and outcome after UCBT [2]. TNC content remains the preferred measurement of cell dose and is used by most transplantation centers for UCB selection, with 2.5 x 107 TNCs per recipient's body weight usually recommended as the critical threshold that should be fulfilled [15]. Nevertheless, as occurred in most studies in adults, TNCs were not associated with any particular outcome. It is likely this finding reflects our current inability to offer a sufficiently high number of cells to adult recipients to improve results after a minimum threshold of TNCs has been transplanted. This is in contrast with data in

children, in whom several large registry series have shown that an additional cut-off point of 5 x 107 TNCs per kg could be beneficial [16,17].

Our data reinforce the importance of CD34+ cell dose on engraftment after UCBT [18,19] and the need to include this characteristic as a major criterion for UCB unit selection [2,3]. Interestingly, the amount of CD34+ cells appeared to affect the speed of myeloid recovery rather than the overall engraftment rate. It should be taken into account that these findings may be applicable only when, as in our center, a minimum number of CD34+ cells has been used for UCB selection (see Supplemental Material). In contrast to 1 series [18] but in accordance with several studies [6], the CD34+ cell dose was not independently associated with other short-or long-term transplantation outcomes. In sharp contrast, CD8+ cell dose emerged as an important factor for most relevant transplantation outcomes. First, the number of CD8+ cells influenced not only the speed of neutrophil recovery but also the overall neutrophil engraftment rate. Further, in contrast to the TNC or CD34+ cell dose, CD8+ cell content also predicted time to platelet recovery. To our knowledge, only 1 previous report has suggested a role of CD8+ cells on neutrophil engraftment after UCBT [20]. However, in that study, this relevance was restricted to patients receiving a low dose of CD34+ cells. The mechanisms by which donor T cells could inhibit the host ability to reject an allograft remain largely undefined. However, the presence of a graft-versus-host immunological response mediated by CD8+ cells from the graft has been suggested in the double-unit UCBT platform, where unit dominance is strongly related to the number of CD8+ cells present in each UCB unit [21,22]. In addition, CD8+ cells have also been shown to facilitate engraftment in murine models [23,24].

The most important finding of this study was that the CD8+ cell dose also had an independent impact on NRM, DFS, and OS. To our knowledge, this observation has never been reported. The only characteristics related to the graft that have previously shown to affect OS in adult recipients have been the TNCs dose after sUCBT [25] and the CD34+ cell dose after double-unit UCBT [18]. Although the basic mechanisms underlying CD8+ cell content effect are unknown, several hypotheses could explain its relevance. Obviously, a greater CD8+ cell dose could have offered a benefit in early NRM by improving the engraftment rate but, noteworthy, a benefit close to 10% on NRM and OS was also observed for patients surviving beyond 100 days after the transplantation. Whether this long-term advantage is due to improved immune recovery or other causes is, at present, unknown. Further, the subpopulations of CD8+ cells of the UCB unit potentially responsible for the relevance of CD8+ cells on engraftment and NRM merit close consideration because the CD8+ cell content of the UCB unit could likely be a surrogate for 1 or some of them. Both the large naive (CD45RA+CCR7+) and the small alloreactive effector-memory (CD45RA~ and CD45RA+CCR7~) CD8+ Tcell subpopulations present in UCB, the latter with potential capacity to recognize maternal antigens, including HLA noninherited antigens as well as different pathogens, could be involved both in engraftment facilitation and immune response to viral infections occurring after the transplantation. The potential role of the latter lymphocyte subset is favored by the close agreement between our results with data showing the positive impact of noninherited antigens match on engraftment and NRM after UCBT [26]. Furthermore, it has been shown that a CMV-specific T cell response derived from the UCB graft is

Figure 1. Unadjusted cumulative incidence of myeloid engraftment in the overall series (A), and according to CD8+ cells infused x 106/kg (B), CD34+ cells infused x 105/kg (C), and both cell populations (D). For patients who received UCB units above and below 1.6 x 106 CD8+ cells/kg, the cumulative incidence of myeloid engraftment at 60 days and the median time to neutrophil recovery were 96% and 19 days and 87%, and 22 days, respectively (P = .007) (B). For patients who received UCB units above and below 1.3 x 105 CD34+ cells/kg, those figures were 92% and 18 days and 91 % and 23 days, respectively (P = .005) (C). The numbers of CD34+ and CD8+ cells were able to define 4 groups regarding the speed and overall rate of myeloid engraftment (D), with group 1 being those grafts with high doses of both CD34+ cells (>1.3 x 105 per kg) and CD8+ cells (>1.6 x 106 per kg), group 2 those with high dose of CD34+ cells (>1.3 x 105 per kg) but low dose of CD8+ cells (<1.6 x 106 per kg), group 3 those with low dose of CD34+ cells (<1.3 x 105 per kg) but high dose of CD8+ cells (>1.6 x 106 per kg), and group 4 those with low doses of both CD34+ cells (<1.3 x 105 per kg) and CD8+ cells (<1.6 x 106 per kg). The cumulative incidence of and median time to myeloid engraftment were 95% and 17 days for group 1 (high rate and fast engraftment), 88% and 19 days for group 2 (low rate but fast engraftment), 97% and 23 days for group 3 (high rate but slow engraftment), and 87% and 23 days for group 4 (low rate and slow engraftment), respectively (P = .009).

primed to viral antigens as early as day 42 after UCBT [27]. These UCB-derived T cells are able to rapidly proliferate in vitro but fail to achieve sufficient numbers in vivo to control CMV reactivation, suggesting that the total amount of this alloreactive T cells in the UCB unit could be clinically relevant to augment immunity against pathogens [27]. Currently, we are analyzing in depth the potential impact of CD8+ cell dose on the incidence of specific bacterial, viral, and fungal infections and on immune reconstitution after UCBT, and we are also studying the specific CD8+ T cell fractions that could account for our findings. It should be taken into account that these results were observed with the use of ATG in the conditioning regimen and might not be applicable in other settings. The lack of influence of TNCs and CD34+ cells, whether cryopreserved or infused, on NRM, DFS,

and OS strongly argues against the hypothesis that the relevance of CD8+ cell dose is merely a surrogate for graft size, reflecting the fact that this cell subtype is better preserved during cryopreservation, storage, and thawing than other cell subsets of the graft.

One caveat of this large single-center study is that it included a heterogeneous patient population and encompassed a long period of time. However, patients received a relatively homogeneous conditioning regimen and supportive care and the criteria used for UCB unit selection were quite uniform. Further, CD8+ cell dose was never used for UCB unit choice, which excludes a potential selection bias in the results. It should be also noted that cell populations were analyzed in samples drawn from the UCB bag infused to the recipient, using the same method, reagents, and equipment

Figure 2. Unadjusted cumulative incidence of platelet engraftment in the overall series (A) and according to CD8+ cells infused x 106 per kg (B). For patients who received UCB units with > 1.6 x 106 CD8+ cells/kg, the cumulative incidence of platelet engraftment at 180 days and median time to platelet engraftment were 77% and 44 days, respectively, whereas for those receiving a lower CD8+ cell dose they were 57% and 62 days, respectively (P = .0001).

throughout the study, and in a single laboratory, overcoming 1 of the most important limitations of registry-based studies. In fact, the cell composition of UCB units in our series was not different than in previous reports in terms of TNCs, CD34+ cells [4,5,28], and lymphocyte subsets [20]. As expected, the number of T cells in UCB was 10 to 100 times lower than in bone marrow or mobilized peripheral blood [29], but it was higher than in T cell—depleted grafts, a strategy resulting in a

high rate of graft failure [30]. Interestingly, no significant colinearity was observed between CD8+ cells and TNCs or CD34+ cells, likely because of the high variability in T cell composition present in UCB grafts, which merits further study.

If the relevance of CD8+ cell number is confirmed by others, this characteristic should be considered an additional criterion for UCB unit selection that would need to be offered

Figure 3. Unadjusted cumulative incidence of NRM (A), DFS (B), and OS (C) according to CD8+ cells infused x 106 per kg. The cumulative incidence of NRM at 100 days, 1 year, and 4 years was 13%, 34%, and 39% for patients with CD8+ cells above 1.6 x 106 CD8+ cells per kg whereas it was 26%, 46%, and 54% for those below that level (P = .03). DFS and OS for patients with CD8+ cells above 2.2 x 106 CD8+ cells per kg were 38% and 42%, respectively compared with 23% and 26% for patients who received a lower CD8+ cell dose (P = .02 and P = .04, respectively).

to transplantation centers. In this sense, quantification of CD8+ cells is standardized, routinely used, and would not represent a technical or economic burden to UCB banks.

In conclusion, this study demonstrates that outcome after UCBT in adults with hematological malignancies was better when UCB grafts had a greater CD8+ cell content. Taking into account CD8+ cell dose for graft selection could have significant clinical implications in the UCB transplantation setting, where measures focused on improving engraftment, reducing NRM, and increasing long-term survival are desperately needed.

ACKNOWLEDGMENTS

The authors thank Alejandro Madrigal and Sergio Querol for their critical review of the manuscript.

Authorship statement: F.M., J.S., and G.F.S. conceived the study, performed data collection, statistical analysis, and wrote the paper. M.A.S. revised the manuscript and approved the final version. L.L prepared UCB for infusion, performed cell counts, and collected the data. F.C., P.M., I.L., C.S., C.C., B.B., G.M., F.L., J.D.L.R., I.J., B.V., J.A.M., D.M.C., N.P., and R.R. provided different kinds of data and critically reviewed the manuscript. P.S. processed UCB at the time of storage and critically reviewed the manuscript. D.P. performed HLA-typing and critically reviewed the data. All authors revised and approved the final manuscript.

Conflict of interest statement: There are no conflicts of interest to report.

Financial disclosure: The authors have nothing to disclose.

SUPPLEMENTARY DATA

Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.bbmt.2014.06.038.

REFERENCES

1. Sanz MA. Cord-blood transplantation in patients with leukemia—a real alternative for adults. N Engl J Med. 2004;351:2328-2330.

2. Ballen KK, Gluckman E, Broxmeyer HE. Umbilical cord blood transplantation: the first 25 years and beyond. Blood. 2013;122:491-498.

3. Rocha V, Gluckman E. Improving outcomes of cord blood transplantation: HLA matching, cell dose and other graft- and transplantation-related factors. Br J Haematol. 2009;147:262-274.

4. Rocha V, Labopin M, Sanz G, et al. Transplants of umbilical-cord blood or bone marrow from unrelated donors in adults with acute leukemia. N Engl J Med. 2004;351:2276-2285.

5. 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;351:2265-2275.

6. Laughlin MJ, Barker J, Bambach B, et al. Hematopoietic engraftment and survival in adult recipients of umbilical-cord blood from unrelated donors. N Engl J Med. 2001;344:1815-1822.

7. Sanz GF, Saavedra S, Planelles D, et al. Standardized, unrelated donor cord blood transplantation in adults with hematologic malignancies. Blood. 2001;98:2332-2338.

8. Sanz J, Sanz MA, Saavedra S, et al. Cord blood transplantation from unrelated donors in adults with high-risk acute myeloid leukemia. Biol Blood Marrow Transplant. 2010;16:86-94.

9. Glucksberg H, Storb R, Fefer A, et al. Clinical manifestations of graft-versus-host disease in human recipients of marrow from HL-A-matched sibling donors. Transplantation. 1974;18:295-304.

10. Przepiorka D, Weisdorf D, Martin P, et al. 1994 Consensus Conference on Acute GVHD Grading. Bone Marrow Transplant. 1995;15:825-828.

11. Shulman HM, Sullivan KM, Weiden PL, et al. Chronic graft-versus-host syndrome in man. A long-term clinicopathologic study of 20 Seattle patients. Am J Med. 1980;69:204-217.

12. Gray RJ. A class of K-sample tests for comparing the cumulative incidence of a competing risk. Ann Stat. 1988;16:1141-1154.

13. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-481.

14. Fine JP, Gray RJ. A proportional hazards model for subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496-509.

15. Barker JN, Byam C, Scaradavou A. How I treat: the selection and acquisition of unrelated cord blood grafts. Blood. 2011;117:2332-2339.

16. Rocha V, Cornish J, Sievers EL, et al. Comparison of outcomes of unrelated bone marrow and umbilical cord blood transplants in children with acute leukemia. Blood. 2001;97:2962-2971.

17. Gluckman E, Rocha V, Ionescu I, et al. Results of unrelated cord blood transplant in fanconi anemia patients: risk factor analysis for engraft-ment and survival. Biol Blood Marrow Transplant. 2007;13:1073-1082.

18. Wagner JE, Barker JN, DeForTE, 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;100:1611-1618.

19. Moscardo F, Sanz J, Senent L, et al. Impact of hematopoietic chimerism at day +14 on engraftment after unrelated donor umbilical cord blood transplantation for hematologic malignancies. Haematologica. 2009;94: 827-832.

20. Terakura S, Azuma E, Murata M, et al. Hematopoietic engraftment in recipients of unrelated donor umbilical cord blood is affected by the CD34+ and CD8+ cell doses. Biol Blood Marrow Transplant. 2007;13: 822-830.

21. Martin PJ. Donor CD8 cells prevent allogeneic marrow graft rejection in mice: potential implications for marrow transplantation in humans. J Exp Med. 1993;178:703-712.

22. Martin PJ, Akatsuka Y, Hahne M, Sale G. Involvement of donor T-cell cytotoxic effector mechanisms in preventing allogeneic marrow graft rejection. Blood. 1998;92:2177-2181.

23. Gutman JA, Turtle CJ, Manley TJ, et al. Single-unit dominance after double-unit umbilical cord blood transplantation coincides with a specific CD8+ T-cell response against the nonengrafted unit. Blood. 2010;115:757-765.

24. Milano F, Heimfeld S, Gooley T, et al. Correlation of infused CD3+CD8+ cells with single-donor dominance after double-unit cord blood transplantation. Biol Blood Marrow Transplant. 2013;19:156-160.

25. Cohen YC, Scaradavou A, Stevens CE, et al. Factors affecting mortality following myeloablative cord blood transplantation in adults: a pooled analysis of three international registries. Bone Marrow Transplant. 2011;46:70-76.

26. van Rood JJ, Stevens CE, Smits J, et al. Reexposure of cord blood to noninherited maternal HLA antigens improves transplant outcome in hematological malignancies. Proc Natl Acad Sci USA. 2009;106: 19952-19957.

27. McGoldrick SM, Bleakley ME, Guerrero A, et al. Cytomegalovirus-spe-cific T cells are primed early after cord blood transplant but fail to control virus in vivo. Blood. 2013;121:2796-2803.

28. Ooi J, Takahashi S, Tomonari A, et al. Unrelated cord blood transplantation after myeloablative conditioning in adults with acute myelogenous leukemia. Biol Blood Marrow Transplant. 2008;14: 1341-1347.

29. Theilgaard-Monch K, Raaschou-Jensen K, Palm H, et al. Flow cytometric assessment of lymphocyte subsets, lymphoid progenitors, and he-matopoietic stem cells in allogeneic stem cell grafts. Bone Marrow Transplant. 2001;28:1073-1082.

30. Urbano-Ispizua A, Rozman C, Pimentel P, et al. The number of donor CD3(+) cells is the most important factor for graft failure after allogeneic transplantation of CD34(+) selected cells from peripheral blood from HLA-identical siblings. Blood. 2001;97:383-387.