Received date: October 21, 2014; Accepted date:November 25, 2014; Published date: December 01, 2014
Citation: Peltier L, Brown R, Monczak Y, Popradi G, Storring J, et al. (2015) Development of a New Third-Party Unit for Adult Stem Cell Transplantation using Clinical-Grade Rejected Cord Blood Units. J Genet Syndr Gene Ther 6:254. doi:10.4172/genetic-syndromes-gene-therapy.1000254
Copyright: © 2015 Peltier L, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Visit for more related articles at Journal of Genetic Syndromes & Gene Therapy
Cord blood is a valuable alternative source of Hematopoietic Stem Cells (HSC) for the transplantation of
patients without a related stem cell donor. Even though the number of banked units has increased in recent years, units containing sufficient stem cells for transplantation into larger adult recipients are still very limited. Different approaches have been used to increase the number of infused HSC. However, most approaches have the disadvantage of increasing the cost of graft procurement or rely on a related haplo-identical donor who may not be available. We developed an alternative approach by creating a third-party unit of enriched CD34+ cells from a pool of multiple, human leukocyte antigen (HLA)-blind, cryopreserved CBUs. These pooled units were rejected by the public cord blood bank due to insufficient volumes and/or low nucleated cell counts. Seven recipients with hematological cancers received myeloablative conditioning followed by co-infusion of a ≥ 4/6 HLA-compatible CBU and a third-party source of CBUs consisting of CD34+ cells selected from a pool of cryopreserved CBUs. Six patients were engrafted with a median neutrophil engraftment time of 19.5 (15-29) days. All engrafted patients had 100% HLA-matched CBU chimerism on day +14. All recipients had grades I-III acute graft-versus-host-disease (GVHD) that responded promptly to treatment, and no patients developed chronic GVHD. Two patients died, one on day +28
due to disease relapse, and one on day +360 due to multisystem organ failure. This new method not only supports the goal of increasing the number of HSCs to enable rapid engraftment, but it also demonstrates that clinical-grade rejected CBUs can be used to create a third-party, HLA-blind source of enriched CD34+ cells to support a small, ≥ 4/6 HLA-compatible CBU.
Stem cells; Cord blood; Bone marrow; Transplantation
Since the first HSC transplantation in 1988, Cord Blood Unit (CBU) have been primarily used to treat patients suffering from hematological and oncologic disorders. This stem cell source not only has the advantage of being rapidly available compared with Bone Marrow (BM) and Peripheral Blood Stem Cells (PBSCs), but it also has a greater proliferative capacity, significantly reducing the number of Nucleated Cells (NCs) required for transplantation. When transplanting a single unit, a count of 2.5 x 107 NCs/kg of recipient weight is recommended for successful engraftment. More recently, researchers have recommended increasing the number of NCs to 5 x 107 NCs/kg in the presence of two Human Leukocyte Antigen (HLA) incompatibilities, to accelerate engraftment, especially for patients with malignancies [1,2].
Based on the minimum number of NCs/kg of the recipient weight, the probability of finding a CBU with a 6/6 and 4/6 HLA compatibility for a patient weighing less than 50 kg is 23% and 100%, respectively. However, the probability of finding a CBU with a 6/6 and 4/6 HLA compatibility for a recipient weighing more than 70 kg is only 2% and 76%, respectively . Additionally, donor availability varies according to ethnicity. Half of all Caucasians will find a donor, while only 35% of Hispanics and 20% of African-Americans will find suitable donors [4,5]. To overcome this limitation, especially in adult recipients, some centers choose to perform Double Cord Blood Units Transplantation (dCBUs). Unfortunately, this method requires two CBUs with at least a 4/6 HLA compatibility with the recipient, where each selected CBU must contain a minimum of 1.5 x 107 NCs/kg for a total of ≥ 3.5 x 107 NCs/kg. This increase in nucleated cells improved donor neutrophil engraftment with a median of 32 days (18-53) and platelet recovery by day 91 (56-381), which compares favorably with the results of single CBU transplantation in lower-weight patients [6-9]. Posttransplantation outcomes also showed that dCBUs resulted in a lower rate of leukemia relapse, as well as a similar incidence of acute Graft- Versus-Host-Disease (GVHD) and chronic GVHD, when compared to single CBU transplants .
Another method that is used to increase the number of nucleated cells without increasing the number of T and B cells is the infusion of CD34+ selected haplo-related PBSCs as a third-party source with ≥ 4/6 HLA-mismatched CBUs containing at least 1.5 x 107 NCs/kg . The average time required to achieve neutrophil counts of ≥ 0.5 x 109/L with this method is 14 days (9-36), which is followed by CBU engraftment on day +21 (12-57 days) with sustained platelet counts of ≥ 20 x 109/L after a median time of 32 days [12-14]. Even when mixed chimerism was observed during the first 97 days (11-186), most of the recipients exhibited full CBU chimerism by day +100, with a low incidence of GVHD . However, this method requires the availability of a haplorelated donor and has a higher cost related to PBSC collection and CD34+ selection, compared to single CBU transplantation.
Studies of both methods showed that the infusion of more than one unit with a similar or lower HLA compatibility does not increase GVHD. In addition, there is a bias for the long-term engraftment of the first infused CBU. The need to increase cell counts was not considered in discussions regarding the appropriate number of CD34+ cells/kg. However, these methods clearly support the consensus that infusing ≥ 2 x 105 CD34+ cells/kg to best enable the engraftment of a single CBU. The availability of units containing a greater number of nucleated cells will require that the public cord blood banks increase the cut-off value of the collected units, which will increase the number of discarded units to more than the current amount of 70%. The present pilot study was performed to demonstrate that clinical-grade rejected CBUs can be pooled to create an HLA-blind third-party unit consisting of purified CD34+ cells that can support a single ≥ 4/6 HLA-matched CBU that alone is too small for transplantation.
Matched cord blood unit
The selection of the “matched” CBU required ≥ 4/6 HLA compatibility with the recipient and NCs criteria. The unit was HLAtyped, at the antigen level, for HLA-A and HLA-B and, at the allele level, for HLA-DRβ1 by the donor cord blood bank. Preference was given to full (6/6) compatibility over one or two mismatches; in the presence of a mismatch, HLA-A and HLA-B mismatches were preferred over HLA-DRβ1 mismatch. The CBU had to contain a minimum of 1.5 x 107 NCs per kg of the recipient’s body weight without exceeding 3 x 107 NCs/kg in the presence of complete HLA compatibility, 4 x 107 NCs/kg in the presence of one HLA mismatch, or 5 x 107 NCs/kg in the presence of two HLA mismatches.
Adult recipients aged from 18 and 60 years old, with high risk of relapse were eligible for the study. All participants had a performance status of Eastern Cooperative Oncology Group (ECOG) <3 or Karnofsky ≥ 60% and hematological malignancies that would normally be considered for myeloablative allogeneic transplantation at the McGill University Health Centre (MUHC). No adult Matched Unrelated Donors (MUDs) were available or potentially available 3 to 6 months after the initiation of the search for donors. The myeloablative conditioning regimen consisted of low-dose i.v.fludarabine (20 mg/m2/day) for three consecutive days immediately prior to the administration of the myeloablative regimens of i.v. busulfan 0.8 mg/kg every 6 hours on days -7 to -4 and i.v. cyclophosphamide 60mg/kg per day on day -3 and -2. All recipients received GVHD prophylaxis consisting of combined therapy with Mycophenolate Mofetil (MMF) 1 gram twice per day on days -3 to +30 and Tacrolimus (FK506) on days -3 to +180 to maintain a serum concentration in the range of 10-15 ng/mL.
Pooled cord blood units
All CBUs were collected in utero from mothers previously consented by an obstetrician at the MUHC Birthing Centre. CBUs were processed and cryopreserved within 48 hours of collection in compliance with Health Canada’s Safety of Human Cells, Tissues and Organ for Transplantation Regulations .
A 1 mL sample of the cord blood was obtained before and after volume reduction using an Opti System as per manufacturer requirement (Fenwal, Il, USA). Both samples were assessed for TNC counts using an electric Beckman Coulter® LH 750 Hematology Analyzer. The sample obtained after volume reduction was diluted 1:1 with a Coulter rinsing solution (BD, cat: 8547167). The TNC was calculated by multiplying the White Blood Cell (WBC) number (109/L) by the volume (mL) of the CBU pre- and post-processing. The percentage of nucleated red blood cells (nRBCs) was determined by a manual differential count of Wright-Giemsa-stained blood (Beckman Coulter cat: 754181). CD34, CD45, and 7-Amino-Actinomycin D (7AAD) viability were assessed in both samples using flow cytometry. In two tubes, 20 μL of CD45 (FITC), 20 μL of phycoerythrin (PE) for CD34, 20 μL of 7-AAD, and 100 μL of Stem-Count fluorospheres were added. In each tube, 100 μL of blood was added. Samples were then incubated at room temperature and protected from ambient light for 20 minutes, followed by lysis of RBCs for 10 minutes of fresh CB using 2 mL of 1x ammonium chloride solution (Beckman Coulter Stemkit). Flow cytometry studies on thawed samples were done in a similar manner as for fresh CB samples, with the exception that cell lysis was done using only 2 mL of Hank’s Balanced Salt Solution (HBSS; Bio Whitteker, Lonza cat: 10-547F) to ensure cell viability and accuracy of the cell count.
Short Tandem Repeats (STRs)
The identity of each cell source (HLA ≥ 4/6 matched CBU, all pooled CBUs and transplant recipients) were molecularly determined by Short Tandem Repeat (STR) pattern analysis using PCR with fluorescent primers and capillary electrophoresis. Each individual cell source was tested at 9 different STR loci (D1S1660, D3S2387, D5S1456, D5S818, D7S820, D8S1179, D13S317, D16S539, and D20S481) using a semi-quantitative (PCR) assay. This method was used to obtain a unique molecular “fingerprint” for each cell source. The STR analysis was performed on DNA extracted from the cord blood leuko-platelet layer cells and on the recipient leucocytes collected from a peripheral blood sample. Chimerism analysis was performed at days +14, +21, +28, +56 and +100 post-transplant.
DNA Extraction for STR determination
A 200-400 μL sample of buffy coat was thawed and transferred to a 15 mL conical tube with 3 mL of RBC Lysis Solution (5 Prime Gmbh Hamburg, Germany). The sample was kept 20 minutes at room temperature to allow lysis of the RBCs, and then centrifuged at 720xg for 2 minutes to pellet the nucleated cells. The supernatant, consisting of lysed RBCs, was discarded; the white pellet containing the nucleated cells was gently disrupted by slow vortex mixing; 3 mL of cell lysis solution was added. All cells were lysed, and the DNA content was released into the suspension. The tube was placed on ice for 2 minutes, after which 1 mL of Protein Precipitation Solution was added, and mixed by vortexing at a high speed for 20-30 seconds; the tube was then left on ice for 5 minutes, precipitating the protein while leaving the DNA in suspension. The tube was then centrifuged at 1440xg at 4ºC for 15 minutes, and the supernatant was transferred into a clean 15 mL conical tube. DNA was precipitated by adding 3ml of isopropanol to the pellet using a sterile transfer pipette; the precipitated DNA was transferred into a clean 15 mL conical tube. The DNA was washed with 3ml of 70% ethanol. The tube was centrifuged at 1440xg and 4ºC for 3 minutes, the supernatant was gently decanted, and the tube was inverted for 5 minutes to drain all excess liquid. The DNA pellet was resuspended in DNA Hydration Solution (10 mMTris, 1 mM EDTA buffer, pH 7.6). The DNA concentration and purity were assessed by measuring the UV absorption of the sample at 260 and 280 nm using 2 μL of the DNA solution on a Nano Drop ND-1000 (Nano Drop, Wilmington, DE, USA) spectrophotometer. DNA was then diluted to a final concentration of 0.1 μg/μL using DNA Hydration Solution.
PCR Amplification of STR loci
Nine STR loci were amplified under identical PCR conditions. The loci used were D1S1660, D3S2387, D5S1456, D5S818, D7S820, D8S1179, D13S317, D16S539, and D20S481. PCR was performed in a 200 μL PCR tube using the manufacturer’s recommended conditions, 4 nM fluorescent primer pair 0.1 μg of DNA, 0.1 U Taq DNA polymerase, 1.5 mM MgCl2, 4 nMdNTPs, and PCR buffer (Perkin-Elmer 10X buffer). The PCR conditions were as follows: 94ºC for 5 minutes followed by 35 cycles at 95°C for 30 seconds, 57°C for 30 seconds, and 72°C for 45 seconds. The PCR reaction was analyzed on an ABI PRISM 3100 Genetic Analyzer (automated capillary electrophoresis sequencer) according to the manufacturer’s recommendations. The fragment lengths were determined by the number of STRs present at each locus for each individual. In a mixture of two or more samples (chimerism) in a post-transplantation event, each CBU contributes a PCR fragment with a unique length.
DNA chimerism interpretation
Calculation of the percentage of DNA chimerism was done using the area under the curve of the STR amplification peaks of the CBU. These peaks were preferably chosen to be 4-8 or up to 12 base pairs apart to facilitate analysis. Different equations were used, depending on peak patterns, which may vary between homozygous and/or heterozygous states. The following formulas were used:
(Fernandez-Avilés et al, Leukemia (2003) 17, 613-620)
Y Chromosome assay
The Quantifiler Y Human Male DNA Qantification Kit (Applied Biosystems, Inc., cat. no. 4343906) was used to identify the presence of any Y chromosomes in the peripheral blood sample of the recipient at days +14 and +21 post-transplantation. Varying DNA concentrations (0.1, 0.01, 0.001, and 0.0001 μg/μL) were used to verify the sensitivity of the analysis. In a 25μL capillary PCR tube, 2 μL of each DNA sample, at different concentrations; sterile water as a negative control; and female DNA as a negative DNA control were mixed with 10.5 μL of the Quantifiler Y Human Male Primer Mix, which contained targetspecific primers and a FAM dye-labeled probe, and with 12.5 μL of the Quantifiler PCR Reaction Mix, containing dNTPs, buffer, and Ampli Taq Gold DNA polymerase (Applied Biosystems) according to the manufacturer’s instructions. Samples were amplified and analyzed on the Roche Light Cycler® 2.0 Instrument. After an initial incubation at 95°C for 10 minutes to denature the genomic DNA, 40 cycles were performed using the following profile: denaturation at 95°C for 15 seconds and amplification at 60°C for 1 minute. After completion, an absolute quantification analysis, using a standard curve, was performed using the Roche Light Cycler® Software 4.05 (Roche).
The number of pooled CBUs was based on the sums of the pre-cryopreservation CD34+ counts, ensuring a minimum of 1 x 106 CD34+ cells/kg of recipient weight. For example, a recipient weighing 80 kg would require a total of ≥ 80 x 106 CD34+ cells before thawing. The pooling method consisted of a semi-closed system using a 600 mL transfer bag (MacoPharma cat: VSE4110XA), one transfer bag containing an equal mixture of Gentran 40 and 5% human albumin, and large tubing (Figure 1). The first cryopreserved CBU was inserted into a sterile plastic bag and thawed in a 37°C water bath for 2 minutes. The unit was pierced, and an equal volume of a dilution solution was slowly infused into the thawed CBU and kept at 4 ± 2°C for a period of at least 5 minutes. The content was then transferred to a 600 mL pooling bag (VES4110XA, Maco Pharma; 4C5628, Fenwal). The unit was rinsed up to three times with the mixed solution of 5% human albumin/Dextran 40 and transferred into the same pooling bag. The second and third units were thawed and diluted using the same method. The 600 mL transfer bag with the pooled CBUs was centrifuged at 600xg at 4°C for 9 minutes. A 600 mL bag was then pierced and joined with the pooling bag in order to expel the supernatant using a Fenwal Optipress blood component extractor up to a final volume of 20 and 30 mL. Each pooled bag of three CBUs was then transferred into another 600-mL bag using 10-in-1 tubing (Maco Pharma, EPE 10177). Forty (40) to 60 mL of phosphate-buffered saline (PBS)/EDTA (Miltenyi Biotech, cat: 700-25), with 0.5% human albumin was used to rinse all pooled bags. A final centrifugation at 600xg at 4°C for 9 minutes was performed, followed by expulsion of the supernatant in the Optipress to reduce the final pooled CBU bag to a volume of 95 ± 5 mL.
|1||B-LPD||41 (M)||81||pos||pos||A/O||Arabic||A and B||2.9||1.1Â + 3.0||4.1||8.6||2||45||33 Cauc;
|2||CML-BC||32 (M)||56||pos||pos||B/A||Asian||A||2.7||0.7 + 2.0||2.67||1.4||1||30||26 Cauc;
|3||AML||30 (F)||88||pos||pos||O/O||Cauc||A and B||2.6||1.6 + 2.5||4.13||0.4||2||42||23 Cauc;
|4||B-ALL||57 (F)||72||pos||pos||O/O||Cauc||A||2.5||1.3 + 1.9||3.19||3.6||1||39||27 Cauc;
|5||B-SLL||45 (M)||76||pos||pos||O/O||African descent||A and B||3||2.1 + 1.7||3.8||2.4||2||42||25 Cauc;
1 African descent;
|6||ALL||37 (F)||48||pos||pos||B/A||Pakistani||A and B||4||1.1 + 3.8||4.9||0.03||2||30||18 Cauc;
|7||CML-BC||53 (M)||143||pos||pos||B/A||Polish||DrB1||1.7||0.3 + 2.7||3.02||0.27||1||30||
|Mean||42||81||-||-||-||-||-||2.8||1.2 + 2.5||3.69||2.39||-||37||-||66.6 min||21.2||70|
|Median||41||76||-||-||-||-||-||2.7||1.1 + 2.5||3.8||1.4||39||-||72||19.5||53|
ND: not done; MM: mismatch; Cauc: Caucasian;
B-LPD = B-liphoproliferative disorder, CML-BC = chronic myelogeneous leukemia in blast crisis, AML = acute myelogenous leukemia, ALL = acute lymphoblastic
leukemia, SLL = small cell lymphocytic lymphoma.
Table 1: Summary of Patients, Graft Characteristics, and Engraftment.
|Pt #||Diagnostic||Age (Sex)||Weight (kg)||DNA Chimerism
(CD3 and CD33 subpopulations)
% of Matched CBU
|GVHD||Length of Hospital Stay||Overall Survival Days post-transplant|
|1||B-LPD||41 (M)||81||CD3||100%||100%||100%||100%||90%||100%||Grade III (S/GI/K)||94||Alive||Died +360||N/A|
|2||CML-BC||32 (M)||56||CD3||100%||100%||100%||100%||100%||Grade II (S/GI)||99||Alive||N/A||Alive +1521|
|3||AML||30 (F)||88||CD3||100%||100%||100%||N/A||N/A||100%||Grade I (S)||27||Died +28||Died +28||N/A|
|4||B-ALL||57 (F)||72||CD3||100%||100%||100%||100%||100%||Grade II (S/GI)||57||Alive||N/A||Alive +1081|
|6||ALL||37 (F)||48||CD3||100%||100%||100%||100%||100%||Grade I (S)||47||Alive||N/A||Alive +597|
|7||CML-BC||53 (M)||143||CD3||100%||100%||100%||100%||N/A||Grade II (S)||64||Alive||N/A||Alive +507|
S = Skin; GI = gastrointestinal; L = Liver; K = Kidney; N/A = Not applicable
B-LPD = B-liphoproliferative disorder, CML-BC = chronic myelogeneous leukemia in blast crisis, AML = acute myelogenous leukemia, ALL = acute lymphoblastic
leukemia, SLL = small cell lymphocytic lymphoma.
Table 2: Summary of Transplant Outcomes.
|1||Neg Control Water||26.91||15||Day +21||0.01||27.15|
|2||Neg Control Water||26.89||16||Day +21||0.01||27.09|
|3||Neg Control Female||27.05||17||Pos Control Male 0.001||29.28|
|4||Neg Control Female||27.22||18||Pos Control Male 0.001||29.37|
|Pos Control Male|
|Pos Control Male|
|7||Day +14||0.1||25.93||21||Day +21||0.001||27.3|
|8||Day +14||0.1||25.92||22||Day +21||0.001||27.27|
|Pos Control Male|
|9||Day +21||0.1||26.58||23||0.0001||> 35|
|Pos Control Male|
|10||Day +21||0.1||26.53||24||0.0001||> 35|
|Pos Control Male|
|Pos Control Male|
|13||Day +14||0.01||26.9||27||Day +21||0.0001||27.01|
|14||Day +14||0.01||26.79||28||Day +21||0.0001||27.07|
CP: crossing point.
Table 3: Absolute Quantification of the Y Chromosome on Days +14 and +21 with Regard to the Chimerism of Recipient #4.
The final pooled CBU bag was injected with 100 mL of air, followed by the slow addition of the 7.5 mL of Clini MACS CD34 Reagent (Miltenyi Biotech cat: 171-01), drawn from the vials using a10-mL syringe with a 16 G needle. The bag was placed on a rocking platform for 45 minutes at room temperature, and turned twice to enhance binding of the magnetic antibodies to CD34. A final wash and centrifugation at 600xg at 4°C for 9 minutes was performed after adding 400 mL of PBS/ EDTA with 0.5% human albumin. The supernatant was expelled on the Optipress to obtain a final volume of ≤ 300 mL. A Cell Separation Column, placed in the Clini MACS Magnetic Separation Unit, was used according to the manufacturer’s instructions to isolate the magnetic antibody-bound CD34-expressing cells contained in the 300 mL bag. The positive volume fraction of 40 mL was transferred by the Clini MACS into a 150 mL transfer bag. At the end of the selection, the 150 mL transfer bag containing the CD34+ cells was taken from the equipment and transferred to the biological safety cabinet, where a 0.5 mL sample was taken for CFU assay, CBC, CD34, CD45, and CD3 count analysis, and 7AAD viability assessment. The negative fraction was verified for its content of CD34+ cells by the retrieval of a 2 mL sample after the insertion, of a sampling site coupler. The negative fraction bag was then centrifuged at 600xg at 4°C for 9 minutes, which was followed by the expulsion of the supernatant using the Optipress to concentrate the leucocytes to a volume varying between 30 and 40 mL. Samples taken for analysis included 10 mL for bacterial and fungal cultures, 5 mL for DNA extraction (for STR by PCR), and the remainder was preserved as two 10-13 mL samples in 15 mL polystyrene tubes at -80°C.
Descriptive statistics were used to compare the TNC and CD34+ cell counts, viability, and cell count recovery. All statistical analyses were conducted using software from the Statistical Package for the Social Sciences (SPSS, Chicago) and with Microsoft Excel version 2010. Probability (p) values of less than 0.05 were considered to be statistically significant. Descriptive statistics were used to compare the TNC and CD34+ cell counts, viability, and Colony-Forming Units (CFUs) pre- and post-processing. Results are presented as the mean and Standard Deviation (SD). Analyses of the variables following the normal t-test distribution were performed with a one-way analysis of variance (ANOVA). In the presence of an asymmetric distribution, non-parametric analyses and correlations were performed. Two independent-sample t-tests ensured the comparison of two groups based on one variable with the Mann-Whitney U test. If ANOVA was not applicable, an independent-samples t-test was used for categorical variables and the Mann-Whitney U-test was used for continuous variables. The power and analysis of the sample size were performed using the software Power and Precision 4.1 (2011).
The research protocol was approved by the Research Ethics Board of the McGill University Health Center (GEN#06-018). Informed consent for the donation of CBUs was obtained from mothers before the onset of active labor by obstetrical nurses with the understanding that units would not be reserved for family use and would be de-identified once selected for use. General transplant and study-specific consent was obtained from eligible transplant candidates by a transplant physician.
Seven patients received 4/6 or 5/6 HLA-matched CBUs. HLA Class I, A, and B loci were analyzed using a low-resolution molecular typing method, and Class II HLA loci were analyzed using high-resolution molecular typing for DRβ1. The HLA typing was performed by an external laboratory. All patients had high-risk disease with no related or unrelated HLA-compatible adult donors available. The CBUs that were available were too small for transplantation as a single CBU.
Patient #1 was a 41-year-old, CMV-positive male with blood type A weighing 81 kg. He was diagnosed in 2007 with an unclassifiable CD5-negative B-cell lymphoproliferative disorder that was initially morphologically similar to CLL. His disease course included successive Partial Remissions (PRs) of short duration after multiple treatment regimens, including rituximab with cyclophosphamide, vincristine and prednisolone (R-CVP), fludarabine, clophosphamide and rituximab (FCR), and high-dose etoposide/cyclophosphamide (VP16/CTX), with late transformation (Richter’s Syndrome). The patient had progressive stage IV disease at the time of allogeneic stem cell transplantation.
Patient #2 was a 32-year-old, CMV-positive male with blood type B who weighed 56 kg. He was diagnosed in 2009 with Philadelphia chromosome-positive chronic myeloid leukemia (CML) in myeloid blast crisis, presenting with extramedullary chloromas. He was initially treated with imatinib mesylate and later switched to nilotinib due to intolerance. He was treated with the “7+3” regimen combined with nilotinib. Six months later he required salvage treatment with FLAGIda for disease progression without evidence of an Abl kinase domain mutation while on nilotinib. Each course of chemotherapy was followed by more than 2 months of pancytopenia, which was complicated by aspergillus infection. At the time of transplantation, the patient was incomplete cytogenetic remission with no evidence of extramedullary disease or active aspergillus infection after treatment, but he had rising BCR/ABL expression measured by RQ-PCR.
Patient #3 was a 30-year-old, CMV-positive female with blood type O who weighed 88 kg. She was initially diagnosed with Acute Promyelocytic Leukemia (APL) in 2007, and was treated with combination chemotherapy and all-trans-retinoic acid. Two years after achieving a CR she developed a therapy-related AML with a non t(15;17) karyotype and without retinoic acid receptor alpha locus rearrangement. At the time of transplantation, she had a near aplastic bone marrow with no evidence of leukemia, following three treatments with high-dose cytarabine (HiDAC) and etoposide (VP16)/arsenic trioxide.
Patient #4 was a 57-year-old, CMV-positive female with blood type O who weighed 72 kg. This patient was diagnosed with highrisk Philadelphia-chromosome negative (Ph-negative) B-acute lymphoblastic leukemia (ALL) in 2010 (high WBC count, high-risk cytogenetics). The patient was transplanted in first complete remission (CR1) after treatment with the Dana Farber Cancer Institute (DFCI) ALL 4 regimen.
Patient #5 was a 46-year-old, CMV-positive male with blood type O who weighed 76 kg. The patient was diagnosed with B-cell Small Cell Lymphoma (SLL) in 2007 and achieved a CR1 after treatment with FCR. He relapsed with bulky multi-site extramedullary disease in 2010. The patient was transplanted in second complete remission (CR2) following treatment with Rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP).
Patient #6 was a 37-year-old, CMV-positive female with blood type B who weighed 48 kg. The patient was diagnosed with T-cell acute lymphoblastic leukemia in 2011. Twenty-eight days post-induction, she had persistent lymphoblasts (8%). She achieved a CR1 after treatment with 6-MP and methotrexate. The patient went on protocol after failure to find an unrelated donor after 6 months.
Patient #7 was a 53-year-old, CMV-positive male with blood type B who weighed 143 kg. He was diagnosed in 2011 with Ph-positive CML. He was initially treated with imatinibmesylate and was then switched to dasatinib after developing resistance with cytogenetic relapse and transformation to CML myeloid blast crisis. He later developed resistance to dasatinib with presence of a T315I Abelson kinase domain mutation. The recipient also suffered from much comorbidity including ischemic heart disease, respiratory syncytial virus, and morbid obesity.
The selection of CBUs for pooling was restricted to those that were ABO blood group compatible with the matched CBU. Additionally, maternal-fetal ABO compatibility was maintained between the pooled CBUs to minimize potential interference by passively transferred maternal anti-A and anti-B IgG in the pool. The clinical-grade CBUs rejected by the public bank contained between 0.20 x 106 and 28.4 x 106 CD34+ cells. The weight of the recipients varied between 48 and 143 kg, requiring a pool of up to 45 cryopreserved CBUs to arrive at a prethaw cell dose of 1 x 106 CD34+ cells/kg recipient weights per protocol. The pooled cords contributed between 25 and 54% of the total CD34+ cells infused into the recipients. The percentage of each CBU within the pool ranged from 0.01% to a maximum of 8% of all CD34+ cells administered to the recipient.
Day of transplantation
The seven recipients of this pilot study were 30-57 years of age with a median weight of 76 (48-143) kg. Four recipients were ABO blood type-incompatible with their transplanted ≥ 4/6 HLA-matched CBU. Two patients had a single HLA class I A locus mismatch, four had both an HLA class I A and B locus mismatch, and only one patient had an HLA class II DR mismatch. On the day of transplantation the matched CBUs were infused first, followed by the pooled CBUs within an average of 66.6 minutes. Up to 45 CBUs of different volumes and different TNC and CD34+ counts were pooled to create the “third party unit”, containing ≥ 1 x 106 CD34+ cells/kg of recipient weight before thawing. As shown in Table 1, the average number of CD34+ cells contained in the matched CBU was 1.2 x 105/kg, and the “third party unit” derived from the pooled CBUs contained a mean of 2.5 x 105/kg for a total of 3.7 x 105 CD34+ cells/kg.
Engraftment and chimerism
The median time to neutrophil recovery (ANC ≥ 0.5 x 109/L) was 19.5 (15-29) days post-infusion for all recipients except one, who had primary graft failure (Figure 2). The median time to platelet recovery (≥ 20 x 109/L) was 53 (29-175) days. One patient had platelet engraftment delayed until day +175, which was probably secondary to a concurrent CMV infection and treatment with gancyclovir. The median time to platelet recovery of the other four recipients was 48 days. On day +14, DNA chimerism analysis revealed that 100% of the CD3+ and CD33+ cells derived from the matched CBU donor chimerism, including the cells in the patient who had primary graft failure (Table 2). Only two recipients had a reduced matched CBU donor chimerism in the first 100 days post-transplantation. The first transplanted recipient had a matched CBU/recipient chimerism at day +100 of 90%, which returned to 100% in the subsequent weeks. In contrast, patient #5 never achieved a 100% matched CBU chimerism. The CD3 and CD33 chimerism of the matched CBU were both found at a level of only 35.5% on day +21, suggesting primary graft failure, which was later confirmed on day +40.Given the apparent absence of bridging engraftment by 3rd party CBUs by STR analysis (Figure 3), a more sensitive method for detecting the engraftment of these units was performed. One patient (recipient #4) was a female who had received a female matched CBU with pooled CBUs composed of 21 male and 18 female donors. Real-time PCR of Y chromosome sequences (Table 3) from day +14 and day +21 peripheral blood samples failed to detect pooled CBUs indicating that these either failed to engraft or were present at levels below the sensitivity of Y-chromosome RT-PCR.
Five of the seven transplanted patients are still alive at the time of writing. Four are off immunosuppression, and only one patient has relapsed post-transplantation. Patient #1 died on day +360 from multi-organ failure following months of antiviral therapy for reactivation of resistant CMV. He was clinically in complete remission at the time of death. Patient #3, who had refractory therapy-related AML, had an acute relapse after day +14 and died on day +28 from disease progression. Patient #5 who had primary graft failure, was later salvaged by double-matched cord blood transplantation and remains in complete remission. Five patients developed transient cutaneous grade I/II acute graft-versus host disease (aGVHD), while one patient (patient #1) developed multi-organ grade III aGVHD (skin, kidney and gut) that responded promptly to steroid therapy. To date, none of the patients have developed chronic GVHD. There were no transplantrelated complications or negative patient outcomes that were felt to be related to the third-party cells infused into this group of very high-risk patients.
Hematopoietic Stem Cell (HSC) transplantation is an accepted treatment for life-threatening hematological diseases. The principal HSC sources for the adult population are bone marrow and peripheral blood. However, only 25 to 30% of recipients have a compatible related donor, and approximately 65% will find an unrelated adult donor through the international registry [17,18]. In the absence of a suitably matched adult donor, alternative graft sources arecord blood stem cells or a related haplo-identical donor. Cord blood is rapidly accessible and may allow for reduced HLA compatibility. However, one constraining factor persists, namely, the limited number of nucleated cells present in a single CBU, which is necessary for ensuring engraftment, especially for larger recipients.
The most frequently used method to overcome this limitation is the infusion of two CBUs each containing a minimum of 1.5 x 107 NCs/kg of recipient weight (i.e. double cord transplant or dCBUs) [5,10]. Using this method, a mean neutrophil recovery at 21 days for 93% of the recipients was obtained by the University of Minnesota with a relapse rate lower than that associated with single CBU transplantation . The selection of a smaller CBU increases the probability of finding a second ≥ 4/6 HLA-matched CBU in 75% of the searches. Despite this increase, 25% of the launched searches fail to find donors due to the difficulty of finding two CBUs with the minimum TNC required that are HLA 4-6/6-compatible with the recipient . The cost of such a method can also be considered a limitation in many countries, raising it from $35,000 (USD) for a single CBU to $70,000 for dCBUs.
As an alternative to using a second CBU to increase the number of CD34+ cells, co-transplantation of a ≥ 4/6 HLA-matched CBU with a third-party unit created by CD34+ selection from a HLA haplo-identical apheresis-related donor has been shown to enhance engraftment . Similar to dCBUs, the matched CBU must contain a minimum TNC count of 1.5 x 107/kg of the recipient weight. Fernandez, the first to use this method, selected matched CBUs with a mean TNC of 2.39 x107/kg (range 1.14-4.3) and CD34 of 1.1 x 105/kg (range 0.35-3.7) of recipient weight . The infused third-party unit contained a mean number of 24 x 105/kg (10.5-33.4) CD34+ cells and a reduced number of CD3+ cells of 2.4 x 103/kg (0.6-15.6) . The neutrophil engraftment time and Overall Survival (OS) were similar to those seen with dCBUs transplantation. Although this method is promising, the availability of a related haplotype donor remains an issue. According to the 2011 Canadian Census, the number of siblings per family was only 1.5 . In addition, the increasingly older age of transplant recipients and an aging populace reduces the probability of finding a healthy parental donor younger than 60 years of age.
The common factor in all of these methods is the supplementation of a single ≥ 4/6 HLA-compatible CBU with an additional unit with equal or lower HLA-compatibility, containing some nucleated and/ or CD34+ cells, to enhance neutrophil engraftment. We investigated another method of achieving this by using clinical-grade rejected CBUs of unknown HLA type. This pilotstudy demonstrates the feasibility and safety of performing CD34+ selection from multiple, clinical-grade rejected, HLA-blind CBUs that were pooled post-thawing to create a supportive third-party unit.
Seven patients, aged 32 to 57 years old (mean of 42 years), with different hematological malignancies were transplanted in this pilot study between 2009 and 2013 at the McGill University Health Centre. All recipients received an initial ≥ 4/6 HLA-matched CBU containing a mean of 2.8 x 107 (1.7 - 4 x 107) TNC/kg of the recipient, followed by infusion of a newly developed third-party unit containing a mean of 2.5 x 105 (1.7-3.0 x 105) CD34+/kg. The creation of this third-party unit required thawing, pooling, and CD34+ selection of a median of 39 (30-45) small clinical-grade rejected CBUs. Median times to neutrophil and platelet engraftment were 19.5 days (15-29) and 53 days (29-175) respectively, similar to those reported with dCBUs and thirdparty haplo-related transplantation. One patient had a longer platelet engraftment time (day +175), which was attributed to early reactivation of CMV and treatment with ganciclovir. One patient, who relapsed and died on day +28, had 100% matched donor chimerism on days +14 and +21 (patient #3). A second patient failed to engraft (patient #5). The neutrophil count of this recipient transiently increased to 0.38 x 109/L on day +20 before declining. He was later successfully salvaged on day +71 with two ≥ 4/6 HLA-matched CBUs and had neutrophil engraftment on day +35 with a single unit.
Recently, Rocha (2010) demonstrated that a high CD34+ cell count post-thaw can improve engraftment depending on the transplant indication . A CD34+ cell count of ≥ 1.7 x 105/kg was recommended post-thaw when recipients had a malignancy, and a count of ≥ 2-2.5 x 105/kg was recommended for non-malignant disorders. In our study, all recipients received an average of 2.5 x 105 CD34+ cells/kg from the third-party unit, to which was added a mean of 1.1 x 105 CD34+ cells/kg pre-thaw from the matched CBU. This infusion of a large number of CD34+ cells (mean 3.6 x 105/Kg) may be the most important determinant of the early neutrophil engraftment that we obtained, one week earlier than typically reported after single CBU transplantation. Only one patient did not show evidence of myeloid or lymphoid engraftment by day +40 and salvaged 71 days after the first transplantation using dCBUs and engrafted.
In our study, the matched CBU was infused before the pooled thirdparty unit. Using this infusion sequence we achieved full matched CBU chimerism on day +14 in all engrafted recipients, which was possibly due to the sequence of cell infusions. According to Haspel, the first unit infused in dCBUs transplants was favored to achieve long-term engraftment in 76% ofcases . This bias in favor of the first infused unit appears to occur when infusion of the second unit is sufficiently delayed to enable engraftment of the first. Quiescent stem cells in G0/G1 preferentially adhere to the bone marrow vascular endothelium within 5 minutes post-infusion and undergo transendothelial migration beginning about 20 to 40 minutes after the initiation of stem cell rolling, continuing for up to 3 hours [24,25]. In our cohort, the third-party unit was infused after 66.6 minutes on average (range 18-96 minutes). A greater CD3+ cell dose and CD34+ cell viability are also determinants of the winning unit when transplanting dCBUs [6,26]. The pooled CD34+-selected third-party unit in our study contained a total of only 2.39 x 103 CD3+ cells/Kg, and the number of CD3+ cells originating from individual units in the pool was even lower by an order of magnitude. The significantly delayed infusion of the third party units and their low CD3+ content compared to the matched CBU may have contributed to the 100% day +14 matched CBU chimerism that we observed [6,27]. The extent of this bias in favor of the matched unit was greater than reported for dCBUs and third-party haplo-donor transplants where transient mixed graft chimerism is generally observed in the early stages of engraftment. The sensitivity of STR analysis to detect thirdparty cells in our study may have been reduced and below the limit of detection given that each unit of the pool represented a relatively small fraction of the total cells infused. We were able to show however that in the one female patient who received a female matched CBU, no third party cells could be detected on day +14 and +21 by Y-chromosome PCR, suggesting that preferential engraftment of the matched CBU may be particularly rapid in our protocol.
An important feature of this study was the utilization of pooled HLA-blinded CBUs to support the HLA ≥ 4/6 matched CBU. The possibility of finding one or several units within the pooled CBUs that have a ≥ 4/6 HLA compatibility with the recipient is extremely unlikely, considering the number of alleles in HLA classes I A, B, and C, which varies from 1788 to 2934 per class, class II DRβ1 (1418 alleles), and DQ (50-323 alleles) (HLA Nomenclature website, 2013). A study performed in 1899 unrelated American-European individuals showed that the DRB1-DQA1-DQB1 haplotypes were shared in less than 1% of these individuals. This low probability of HLA compatibility is further decreased by the presence of non-Caucasian units in the pooled CBUs. The low probability of compatibility and the very low cell dose of each pooled cell unit should translate into a very low chance of thirdparty cell engraftment. Rare exceptions could occur, as reported by Lister (2007), who described long-term engraftment of a cord unit with a single compatible HLA locus (HLA B) in a study of non-CD34+- selected pooled cord unit transplantation.
Acute GVHD is less likely to occur and less severe after the transplantation of mismatched cord blood than after the transplantation of mismatched bone marrow [30,31]. However, the rate of acute GVHD was higher among recipients receiving dCBUs and haplo-related third-party grafts compared with recipients receiving a single unit or the co-transplantation of CBUs [2,20]. All engrafted recipients in our study developed acute GVHD (2 grade I, 3 grade II, 1 grade III) that responded promptly to corticosteroid therapy and no patients developed chronic GVHD. The T cells contained in the ≥ 4/6 HLA-matched CBUs may have eliminated the residual host T cells, and the depletion of CD3+ cells in the pooled unit may have contributed to limiting acute GVHD [32-35]. It is thought that the Graft Versus Leukemia (GVL) effect may be enhanced in dual cord transplants with both units potentially contributing to the overall effect. Even though all 7 of our patients had advanced disease, a high risk of relapse, and low expected survival, only one patient with refractory AML relapsed quickly after engraftment. This may indicate that the third-party units contribute to the GVL effect despite CD3+ celldepletion.
Allogeneic transplantation is a costly procedure with the source of donor cells independently contributing to differences in costeffectiveness. The cost of cord blood transplantation is generally greater than that of bone marrow or peripheral blood stem cell transplantation [36,37]. We are not aware of any study that examines the cost-effectiveness of different types of cord blood transplants though such a consideration could affect the decision to perform dCBUs transplantation, CD34+-selected third-party haplo-related transplantation, or the novel third-party pooled cord method outlined in this pilot study. Although our series is small, the engraftment time, low incidence of acute GVHD, and the median length of hospital stay observed in this pilot study are comparable to those reported for dCBUs. The cost of a 40 unit pooled third-party graft was approximately $24,000, including $350/unit for processing, freezing, serological/ mycological/microbiological testing, and $10,000 for pooling and CD34+-selection. This compares favorably with the purchase of two CBUs at a cost of $35,000 each for a total of $70,000 (USD). Public cord blood banking is an inefficient and expensive endeavor in part due to the cost associated with the collection of units, the majority of which are discarded owing to insufficient cell content, the need to maintain large inventories, and the low annual turn-over of stored units. Using our approach, some of these costs could be recovered and result in significant savings.
The preliminary results of this pilot study suggest that the use of HLA-blind, clinical-grade rejected pooled cord blood CD34+-selected third party cells for allogeneic cord blood stem cell transplantation is feasible, safe and may result in time to neutrophil engraftment comparable to dCBUs transplants. These results need to be confirmed in a larger series of patients and the preparation of pooled CBUs further simplified. Our approach could allow some patients to undergo a transplant who would otherwise not be eligible due to lack of a suitable graft.
This study was funded by the McGill University Health Center Foundation.