Bone Marrow Transplantation (2011) 46, 650–658 & 2011 Macmillan Publishers Limited All rights reserved 0268-3369/11
www.nature.com/bmt
ORIGINAL ARTICLE
Graft source determines human hematopoietic progenitor distribution pattern within the CD34 þ compartment C Arber1, J Halter, M Stern, A Rovo´, A Gratwohl and A Tichelli Department of Medicine, Hematology, University Hospital Basel, Basel, Switzerland
The CD34 þ compartment of grafts for clinical allogeneic hematopoietic cell transplantation (HCT) is very heterogeneous. It contains hematopoietic stem cells and several different progenitor cell populations. This study assesses (1) the content of these populations in clinical grafts from G-CSF-mobilized PBMCs, BM and cord blood, (2) the functional correlation of the graft composition with time to engraftment of neutrophils, platelets and reticulocytes and (3) donor age-related changes. Quantitative flow cytometry showed that the distribution of the progenitor subsets differed significantly between the graft sources and that donor age-related changes occur. In patients after myeloablative allogeneic HCT, accelerated platelet and reticulocyte engraftment correlated with the content of common myeloid and/or megakaryocyte erythroid progenitors in the graft. These findings show that a better understanding of the progenitor compartment in human hematopoietic grafts could lead to improved strategies for the development of cellular therapies, for example in situations where platelet engraftment is delayed. Bone Marrow Transplantation (2011) 46, 650–658; doi:10.1038/bmt.2010.193; published online 16 August 2010 Keywords: human hematopoietic stem cells; common myeloid progenitors; graft composition; engraftment; allogeneic hematopoietic cell transplantation; donor age
Introduction In clinical hematopoietic cell transplantation (HCT), CD34 is used as a surrogate marker to determine the content of hematopoietic stem and progenitor cells in the graft. Increasing numbers of CD34 þ cells in the graft were significantly associated with accelerated neutrophil and platelet engraftment after HLA-matched related or unrelated
Correspondence: Dr C Arber, Hematology, University Hospital Basel, Petersgraben 4, Basel 4031, Switzerland. E-mail:
[email protected] or
[email protected] 1 Current address: Center for Cell and Gene Therapy, Baylor College of Medicine, The Methodist Hospital and Texas Children’s Hospital, 1102 Bates Street, Suite 1770, Houston, TX 77030, USA. Received 15 January 2010; revised 24 June 2010; accepted 29 June 2010; published online 16 August 2010
myeloablative allogeneic HCT with G-CSF-mobilized PBMCs (G-PBMC).1–5 When G-CSF became available clinically as a mobilizing agent, the differences in graft composition between G-PBMC and BM products, as well as the different engraftment kinetics, were clearly documented.6 Influence of the CD34 þ cell dose on incidence and severity of acute and/or chronic GVHD and decreased relapse rates were discussed. However, the CD34 þ compartment is very heterogeneous. It contains several hematopoietic progenitor subpopulations, such as hematopoietic stem cells (HSCs), common myeloid progenitors (CMPs), granulocyte-macrophage progenitors (GMPs), megakaryocyte erythroid progenitors (MEPs), common lymphoid progenitors (CLPs), T/natural killer (NK) cell progenitors and pro-B cells.7–14 These populations have been prospectively isolated in mice and men.8 The hallmarks of these committed progenitors are a limited life span of several weeks, a loss of self-renewal potential and the in vivo production of a single wave of mature hematopoietic cells of the respective cellular lineage. Functionally, myelo-erythroid progenitors were shown to confer radioprotection,15 as well as protection against invasive fungal infections,16 lymphoid progenitors were shown to protect against viral infections post transplant in mice.17 Furthermore, murine studies of hematopoietic aging showed a shift in the mouse BM progenitor compartment because of an increased percentage of myeloid progenitors in aged mice as compared with young mice.18,19 To date, the human progenitor counterparts have not been quantified and functionally correlated with time to engraftment in vivo in patients after allogeneic HCT and donor age-related changes have not been assessed. This study was designed (1) to assess the distribution of the CD34 þ subpopulations in G-PBMC, BM and cord blood (CB) products, (2) to functionally correlate the quantitative content of these CD34 þ subpopulations with time to neutrophil, platelet and reticulocyte engraftment in patients after myeloablative allogeneic HCT and (3) to assess donor age-related changes.
Materials and methods Hematopoietic stem/progenitor cell products G-PBMC (n ¼ 33), BM (n ¼ 7) and CB (n ¼ 7) products from healthy donors were analyzed by fluorescenceactivated cell sorting (FACS) after written informed consent at the University Hospital Basel from October
CD34 subpopulations in allograft sources C Arber et al
651
2007 to August 2009. Only CB donations too small for CB banking (with a total nucleated cell count of o100 107) were used for the study after informed consent from the mother. The local ethics committee approved the study. Mobilization of G-PBMC was performed by application of G-CSF for five consecutive days at a dose of 10 mg per kg body weight per day.
Table 1
Patient selection A total of 26 patients undergoing myeloablative allogeneic HCT at the University Hospital Basel both with G-PBMC (n ¼ 23) or BM (n ¼ 3) and with available FACS analysis of the graft product were included. Patients were transplanted from HLA-identical siblings (n ¼ 18) or matched unrelated donors (n ¼ 8) for a variety of diseases (Table 1). No growth factors were applied post transplant. Additional 10 G-PBMC and 4 BM products were analyzed, but patients were excluded from the engraftment analysis because they received either a non-myeloablative conditioning regimen (eight G-PBMC, 1 BM), were HLA mismatched (two G-PBMC) or the products were harvested for other transplantation centers (three BM). The CB products were not used for transplantation.
Gender distribution Male/female
Flow cytometry evaluation of graft composition CD34 þ cells were enumerated using a single platform bead method according to the guidelines of the international society for hematotherapy and graft engineering.20 The analysis of the CD34 þ progenitor subpopulations was performed on lineage (Lin)-negative CD34 þ cells. Linpositive cells were excluded by scatter profile and staining with CD3, CD56 and CD19 except for the analysis of Pro-B cells. Within the LinCD34 þ fraction the following populations were assessed: CD38 HSCs,10 CD38 þ IL3Ralo CD45RA CMPs, CD38 þ IL3RaloCD45RA þ GMPs, CD38 þ IL3RaCD45RA MEPs,11 CD10 þ CD19 CLPs,13 CD10 þ CD19 þ Pro-B cells and CD10CD7 þ CD45RA þ T/NK cell progenitors.12 The following reagents were used from BD Biosciences (Altschwil, Switzerland) unless otherwise indicated: TruCount reagents for CD34 þ enumeration using CD45 (2D1) FITC, CD34 (8G12) PE and 7-AAD; CD123 (7G3) and CD10 (MEM-78, Invitrogen, Basel, Switzerland) in FITC, CD45RA (HI100) and CD7 (M-T701) in PE, CD38 (HIT2), CD3 (SK7), CD19 (SJ25C1) and CD56 (HCD56, BioLegend, Luzern, Switzerland) in PerCP-Cy5.5, CD34 (8G12) in APC. Lymphocyte populations were analyzed by staining for T cells (CD3 þ , CD3 þ CD4 þ and CD3 þ CD8 þ ), NK cells (CD3CD56 þ ) and B cells (CD19 þ ). One million events were acquired for analysis of progenitor populations on a FACSCalibur using Cellquest Software. Data were analyzed with FlowJo Software (Treestar, California, USA). Representative FACS plots for each graft source are shown in Figures 1a–c. Percentages of CD34 þ subpopulations were multiplied with the absolute CD34 þ count to obtain absolute counts of the CD34 þ subpopulations. Definitions of cut-off levels for ‘low’ and ‘high’ progenitor cell content To evaluate the correlation between the content of progenitor cells within a product and time to engraftment,
Patient characteristics G-PBMC
BM
Total number of patients
23
3
Age at HCT (in years) Median Range
51 22–69
26 12–63
17/6
3/0
Diagnosis Acute myeloid leukemia Acute biphenotypic leukemia B-acute lymphoblastic leukemia T-acute lymphoblastic leukemia Myelodysplastic syndrome Chronic myeloid leukemia Chronic lymhocytic leukemia Extranodal NK-T-cell lymphoma Severe aplastic anemia Non-malignant metabolic disease
11 1 1 1 2 3 3 1 0 0
0 1 0 0 0 0 0 0 1 1
Disease status at HCT CR 1 CR 2 Primary induction failure Relapse 1 CML chronic phase 2 or higher Partial remission 1 or higher upfront
11 2 1 2 3 3 1
0 1 0 0 0 0 2
Donor type Matched related (identical sibling) 10/10 matched unrelated
16 7
2 1
Conditioning regimen CY/BU CY/12 Gy TBI CY/etoposide/12 Gy TBI BEAM/2 Gy TBI CY/ATG Cy/fludarabine BU/fludarabine
13 4 1 4 0 1 0
1 0 0 0 1 0 1
GVHD prophylaxis CsA/MTX CsA/mycophenolate mofetil
18 5
2 1
Abbreviations: ATG ¼ anti-thymoglobulin; BEAM ¼ carmustine/etoposide/Ara-C/melphalan; CB ¼ cord blood; G-PBMC ¼ G-CSF mobilizedPBMCs; Gy ¼ Gray; HCT ¼ hematopoietic cell transplantation.
we categorized the transplants into groups containing ‘low’ or ‘high’ numbers of the respective population. All patients were transplanted with a minimum of 4 106 total CD34 þ cells per kg body weight. Patients receiving 4–6 106 CD34 þ /kg were considered as receiving a ‘low’ dose of total CD34 þ cells (n ¼ 4).1–3 The cut-off for ‘low’ or ‘high’ CMP graft content was calculated on the basis of the median percentage of CMP (73.2%) contained in the total CD34 þ compartment, thus for this cohort at 4.4 106/kg (n ¼ 7). For MEP content, two different cut-off points were investigated because of the overall very low MEP content in G-PBMC grafts: first at the median MEP content (0.009 106/kg), and second at the fourth quartile of MEP content (0.050 106/kg). Bone Marrow Transplantation
CD34 subpopulations in allograft sources C Arber et al
652
Definitions of engraftment Time to engraftment was measured in days. Complete blood counts were performed daily, reticulocytes were measured three times per week using the ADVIA120 system (Siemens, Zu¨rich, Switzerland). All patients have engrafted, none of them died before engraftment. Neutrophil engraftment was defined as the first of three consecutive days with an ANC 40.5 109/L (ANC500), platelet engraftment as the first of three consecutive days with platelets 420 109/L (Plt20) in the absence of platelet transfusions, and reticulocyte engraftment as the first of three measurements with reticulocytes 420 109/L (Reti20) in patients with no or minor ABO blood group barrier. An internationally established value for the reticulocyte engraftment after allogeneic HCT does not exist. On the basis of our own experience and by analogy to patients with severe aplastic anemia we retained the value of 420 109/L.21
Statistical analysis The Mann–Whitney test was used for comparison of progenitor content in the different graft sources. For analysis of correlations between progenitor subsets and influence of donor age on graft composition, the Spearman’s rank test was used. For correlation of progenitor content with time to engraftment, a univariate Cox regression model was used. Cumulative incidences of engraftment were compared by log-rank statistics (SPSS Software version 16.0).
Results Distribution and quantification of the CD34 þ subpopulations in G-PBMC, BM and CB products The relative distribution of the CD34 þ subpopulations in % was determined in G-PBMC, BM and CB products.
BM
Gated on Lin–CD34+
104
104 103
103 102 104
800
3
10
100
101
2
400 200
101
103
101 10
104
Gated on Lin–CD34+ without CD19
104
R1 0 200 400 600 800 1000
102
102 CMP
0
GMP MEP
100
101 102 CD45RA
CD38
10 CD34
SSC
101 0
10
600
0
Lin–CD34+
HSC
IL-3 Rα
1000
Gated on R1
CD34
No gate
Gated on Lin–CD34+
100 0 10
FSC
104
103
104
– + Gated on Lin CD34 without CD19
CLP and Pro-B 101
102
103
104
103
103
2
102
CD3, (CD19), CD56
CD19
CD10
10
101 T/NK-P 100 100
G- PBMC
101 CD7
102
103
104
Pro-B
101
CLP
100 0 10
Gated on Lin–CD34+
104
101 102 CD10
103
104
Gated on Lin–CD34+ 104
103
103 CMP
2
104
800
103
600
101 100 0 10
200
R1
0 0 200 400 600 800 1000 FSC
CD34
102
400 SSC
Gated on R1 Lin-CD34+
CD34
No gate 1000
IL-3 Rα
HSC
10
101
4
101 102 103 104 CD38 Gated on Lin–CD34+ without CD19
GMP
101 MEP 100 0 10 101 102 CD45RA 4
10
10 100 100
102
103
104
Gated on Lin–CD34+ without CD19
CLP and Pro-B 101 102 103 104 CD3, (CD19), CD56
3
103
2
2
10
10 CD19
10 CD10
Pro-B
1
10
101
CLP
T/NK-P 0
0
10
Figure 1
100
101 CD7
102
103
104
10
100
101 102 CD10
103
104
Representative FACS plots and distribution of progenitor subpopulations according to graft source. (a–c) Representative FACS plots for the analysis of the CD34 þ progenitor subpopulations in BM, G-CSF-mobilized PBMCs (G-PBMC) and cord blood (CB). (d) Box plots representing the distribution (percentages) of the CD34 þ progenitor subpopulations as median and s.d. according to the graft source. P-values were calculated using the Mann–Whitney test, G-PBMC (n ¼ 33), BM (n ¼ 7) and CB (n ¼ 7). (e) Three-dimensional visualization showing a clear clustering of the progenitor subpopulations according to the graft sources G-PBMC (blue circles), BM (black crosses) and CB (red triangles). The one G-PBMC graft clustering together with the CB products was collected from a donor with poor mobilization.
Bone Marrow Transplantation
CD34 subpopulations in allograft sources C Arber et al
653 Gated on Lin–CD34+ 104
104
103
103
102 CD34
No gate
Gated on R1 104 Lin–CD34+
1000 800
101 100 0 10
103
101 102 103 104 CD38 Gated on Lin–CD34+ without CD19
R1
0 0 200 400 600 800 1000
GMP
101
101
100 0 10
10 10
3
103
10
2
2
10
1
10
CD10
101 102 103 104 CD3, (CD19), CD56
10
T/NK-P
100 0 10
4
101
102
103
104
1
10
0
0
100 0 10
101 102 CD10
103
104
% MEP
P=NS
*
4 2 CB
C G
-P
BM
100
CMP %
80 60 40 20
10
0 0 CB
BM
C
20 40 60
GMP %
80
0 40 20 0 60
100 8
MEP %
G
-P
BM
CB
C BM -P G
CLP
6
CB
* 20
0
CB
G
-P
BM
BM
C BM G
20
BM
*
2
P=NS P=0.002
30
30
Pro-B
0
P