Ten years of cord blood transplantation: from bench to bedside


Professor E. Gluckman, Eurocord, Hospital Saint Louis, 1 Avenue Claude Vellefaux, 75475 Paris cedex10, France. E-mail: eliane.gluckman@sls.aphp.fr


Cord blood is an unlimited source of haematopoietic stem cells for allogeneic haematopoietic stem cell transplant. Since the first human cord blood transplant, cord blood banks have been established worldwide for collection and cryopreservation of cord blood for allogeneic haematopoietic stem cell transplant. More than 400 000 cord blood units are now available for international exchange. Results of unrelated allogeneic cord blood transplants in malignant and non-malignant diseases, in adults and children, show that, compared to human leucocyte antigen (HLA)-matched unrelated bone marrow transplant, cord blood has several advantages, including prompt availability of the transplant, decrease of graft-versus-host disease and better long-term immune recovery resulting in a similar long term survival. Several studies have shown that the number of cells is the most important factor for engraftment while some degree of HLA mismatches is acceptable. Progress is expected to facilitate engraftment and reduce transplant-related mortality and includes reduced intensity conditioning regimen, intra bone injection of cord blood cells and double cord blood transplants. In addition to haematopoietic stem cells, cord blood and placenta contain a high number of non-haematopoietic stem cells that explains the increasing interest of using cord blood for developing regenerative medicine.

Haematopoietic stem cell transplantation (HSCT) can be curative in a large variety of selected malignant and non-malignant diseases. Umbilical cord blood transplantation (UCBT) has extended the availability of allogeneic HSCT to patients who would not otherwise be eligible for this curative approach because of the lack of a human leucocyte antigen (HLA) identical bone marrow (BM) or granulocyte colony-stimulating factor mobilized peripheral blood haematopoietic stem cell (PBSC) donor. The first UCBT was performed in a patient with Fanconi Anaemia, in 1988, (Gluckman et al, 1989). This patient had a healthy HLA-identical sibling who was shown by prenatal testing to be unaffected by the disorder and to have a normal karyotype. Her cord blood was collected at birth, cryopreserved and, after thawing, used for transplantation. Currently, this patient is doing well with complete haematological donor chimaerism and complete recovery of his blood counts.

This first success opened the way to a new field in the domain of allogeneic HSCT as it showed that: a single unit of umbilical cord blood (UCB) contained enough haematopoietic stem cells to reconstitute definitely the host lympho-haematopoietic compartment; an UCB unit could be collected at birth without any harm to the new-born infant, and UCB haematopoietic stem cells could be cryopreserved and transplanted in a host after thawing without losing their repopulating capacity. Since then, our knowledge on the biological characteristics of UCB cells has increased, emphasizing the advantages of using UCB stem cells for transplantation.

Compared to other sources of haematopoietic stem cells, UCB has many theoretical advantages, due to the immaturity of newborn cells. UCB haematopoietic progenitors are enriched with primitive stem/progenitor cells able to produce in vivo long-term repopulating stem cells. The properties of UCB cells should compensate the relatively low number of cells contained in a single UCB unit and, through rapid expansion, reconstitute myeloablated patients. Despite the capacity for UCB cell expansion, clinical results showed that haematopoietic recovery was delayed after UCBT; engraftment was associated with the number of nucleated and CD34+ cells infused and the number of HLA differences (Gluckman et al, 1997, 2004; Wagner et al, 2002). As acute graft-versus-host disease (GVHD) is an early event after allogeneic bone marrow transplantation (BMT) and is partly triggered by cytokine release, it is reasonable to postulate that UCBT induces less frequent and less severe acute and chronic GVHD than adult HSCT which contain a higher number of activated T cells. These properties should lead to less stringent criteria for HLA donor-recipient selection. In comparison with other sources of allogeneic HSCT, UCB offers substantial logistic and clinical advantages such as: significantly faster availability of banked cryopreserved UCB units with patients receiving UCB transplantation in a median of 25–36 days earlier than those receiving BM (Barker et al, 2002); extension of the donor pool due to tolerance of 1–2 HLA mismatches out of six (higher HLA mismatched is associated with lower probability of engraftment); lower incidence and severity of GVHD; lower risk of transmitting infections by latent viruses; lack of donor attrition; lack of risk to the donor and, finally higher frequency of rare haplotypes compared to bone marrow donor registries, as it is easier to target ethnic minorities.

Milestones in the development of UCBT

Development of cord blood banks (CBB)

Umbilical CBB have been established for related or unrelated UCBT. The progress in the field of UCBT has paralleled the huge interest in establishing and developing CBB worldwide. Today, more than 400 000 cord blood grafts are available in more than 50 CBB (http://www.bmdw.org and http://www.wmda.org). This development is due to the organization of international registries for outcome data collection, named Eurocord (http://www.eurocord.org) and the Center for International Blood and Marrow Transplant Research (CIBMTR; http://www.cibmtr.org), and of a CBB networks, named Netcord (http://www.netcord.org) and National Marrow Donor Program (NMDP; http://www.nmdp.org).

Eurocord is an international registry which operates on behalf of the European Blood and Marrow Transplant group (EBMT), and includes European and non-European centres (more than 437 transplants centres in 47 countries), all performing either related (n = 499) or unrelated cord blood transplants (n = 4655). It works in close collaboration with EBMT and Netcord banks to collect clinical data and follow patients transplanted in or outside Europe with Netcord units. During the last 3 years, the number of unrelated UCBT reported to Eurocord has increased as a result of: the increased number of cord blood units stored, the expansion of indications to all diagnoses curable by allogeneic haematopoietic stem cell transplant, publication of criteria of donor choice, the increasing number of adult transplants that use double cord blood units to overcome the problem of the cell dose, and the use of reduced intensity conditioning in older patients or with disabilities (Figs 1–3). Overall results have improved with time due to better choice of indications and better quality of the cord blood unit transplanted (Figs 4 and 5).

Figure 1.

 Eurocord registry data: number of related and unrelated cord blood transplants reported yearly.

Figure 2.

 Eurocord registry data: number of adult and paediatric cord blood transplant.

Figure 3.

 Eurocord registry data: number of patients receiving a double cord transplant (panel A); number of patients receiving a reduced intensity conditioning (panel B).

Figure 4.

 Overall survival according to the period of cord blood transplantation in children.

Figure 5.

 Overall survival according to the period of cord blood transplantation in adults.

In order to promote education and information, Eurocord has recently launched a new European Online Cord Blood Learning Portal, an on-line curriculum on cord blood technology and transplantation, which aims to provide a learning tool for the scientific, technical, clinical and regulatory aspects of cord blood that is easily accessible at a time and language convenient for users (http://www.eurocord-ed.org).

The Netcord group was established in 1998 to provide good practises in UCB storage, facilitate donor search, improve the quality of the grafts, standardize excellence criteria on an international scale and, importantly, establish procedures for bank accreditation in collaboration with FACT (Foundation on Accreditation in Cell Therapy). National regulatory agencies and transplant centres are aware of the need of international standards for cord blood collection, processing, testing, banking, selection and release.

All the practical aspects of cord blood banking, such as mother informed consent, collection techniques, labelling and identification, infectious disease and genetic disease testing, HLA typing, methodology of cell processing, cryopreservation, transportation and release have been extensively published. All these aspects are detailed in the last version of the Netcord-FACT Standards (http://www.factwebsite.org).

Clinical experience with related and unrelated umbilical cord blood transplantation

In a CIBMTR-Eurocord study, comparing paediatric BM and UCBT from an HLA identical sibling, UCBT was associated with delayed granulocyte and platelet engraftment, reduced acute and chronic GVHD and the same survival. This was the first analysis to demonstrate, unambiguously, that acute and chronic GVHD was reduced when CB cells were used instead of BM even when the BM was provided by children (Rocha et al, 2000). This first study was the basis for advocating the use of mismatched UCBT and triggered the development of unrelated CBB.

The second step was the demonstration that unrelated UCBT could be used in all current indications of allogeneic HSCT, including malignant and non-malignant diseases in children and adults and compared favourably to matched unrelated BM or PBSC transplants.

In children with malignant diseases, two studies compared the outcome of unrelated UCBT and BMT. Eurocord published a study comparing the outcome of matched unrelated BMT (HLA six out of six), either unmanipulated or T depleted, to mismatched UCBT: the results showed that after UCBT, engraftment was delayed, GVHD was reduced similarly to T-cell–depleted BMT and that relapse and leukaemia-free survival (LFS) were the same (Rocha et al, 2001). Eapen et al (2007) compared outcomes of 503 children with acute leukaemia given an unrelated mismatched UCBT with 282 unrelated BM transplant recipients (116 HLA alleles matched eight out of eight). HLA allele-mismatched BM recipients had more acute and chronic GVHD without reducing LFS. Importantly, they found that even using an allele-matched BM donor, LFS was not statistically different from one or two HLA disparate UCBT and that an HLA-matched UCBT recipient had better outcomes compared to HLA allele-matched BM recipients. However, an increased transplant-related mortality was observed in children transplanted with a low CB cell dose (<3 × 107/kg) and one HLA-disparate CB graft or in children given a two HLA-disparate UCBT independently of the cell dose infused. Interestingly, use of two HLA-mismatched UCBT was associated with lower incidence of relapse (Eapen et al, 2007).

The same studies were performed in adults with malignancies. A Eurocord study compared adults with acute leukaemia receiving either a matched unrelated BMT (HLA six out of six) or a mismatched cord blood transplant. Results showed that, despite a delay of engraftment, UCBT gave a similar leukaemia survival to BMT (Rocha et al 2004). The CIBMTR and National Cord Blood Program (NCBP) of the New York Blood Center showed that, in adults with malignancies, CBT gave the same LFS as one antigen-mismatched unrelated BMT (UBMT) (Laughlin et al, 2004). At the same time, a Japanese study showed that UCBT gave better results than a matched unrelated donor transplant (Takahashi et al, 2007).

A meta-analysis that combined the published studies, analysed 161 children and 316 adults undergoing UCBT (mostly one or two antigen-mismatched), along with 316 children and 996 adults undergoing UBMT (almost entirely fully matched with the recipient). T-cell–depleted UBMT was excluded; only fully matched UBMT was used in the analysis. Pooled comparisons of studies of UCBT and UBMT in children found that the incidence of chronic GVHD was lower with UCBT, but the incidence of grade III–IV acute GVHD did not differ. There was no difference in 2-year overall survival in children when studies were pooled. For adults, transplantation-related mortality (TRM) and LFS were not statistically different (Hwang et al, 2007).

Taken together, the results of these comparative studies and the meta-analysis showed that (i) UCBT is feasible in adults when a cord blood unit contains a high number of cells and should be considered as an option for an allogeneic stem cell source for patients lacking a HLA-matched bone marrow donor; (ii) Despite increased HLA disparity, UCB from unrelated donors offers sufficiently promising results to matched UBM in adults with haematological malignancies leading to the conclusion, as in children, that the donor search process for BM and UCB from unrelated donors should be started simultaneously, especially in patients with acute leukaemia, where the time factor is crucial.

Algorithm of donor choice

Eurocord and other studies have shown that the main criteria for donor choice are based in the amount of cells infused. All authors agree that the minimum number of nucleated cells (NC)/kg should be 3 × 107/kg or CD34+ 2 × 105/kg. When using two cord blood units, it seems that there is an additive effect: the sum of the two units gives the same engraftment as the same number in one unit (Wagner et al, 2002; Gluckman et al, 2004).

The role of HLA is still unclear. It is important to note that the majority of UCBT reported have utilized matching at the antigen level for HLA-A, and HLA-B and at the allele level for HLA-DRB1. Most of the transplants have one or two HLA differences, transplant centres have a tendency to choose units with better DRB1 matching. Few data are available on allele level typing for HLA-A and -B (Kogler et al, 2005). To summarize several articles (Kamani et al 2008), it seems that increasing the number of HLA differences increases the delay of engraftment, increases TRM, increases the incidence of grade III–IV GVHD and decreases relapse in acute leukaemia. Differences in HLA have a major impact in survival after UCBT for non-malignant diseases but not for leukaemia. It seems that increasing cell dose partially overcome the influence of HLA disparities.

The current recommendations are to choose:

  • 1Cord blood units with ≤2 HLA disparities and ≥3 × 107 nucleated cells/kg or ≥2 × 105 CD34+ cells/kg.
  • 2In non-malignant disease, where the risk of rejection is higher, the dose should be increased and one must avoid units with <3·5 × 107 NC/kg and two or more HLA incompatibilities. If there is no single unit with these characteristics, look for two units with a combined total dose of ≥3 × 107 NC/kg and, if possible, not more than one HLA difference between the two units and the patient.

Role of KIR mismatches

Donor killer cell immunoglobulin-like receptor (KIR)-ligand incompatibility is associated with decreased relapse incidence (RI) and improved LFS after haploidentical and HLA-mismatched unrelated haematopoietic stem cell transplantation. Eurocord assessed outcomes of 218 patients with acute myeloid leukaemia (AML n = 94) or acute lymphoblastic leukaemia (n = 124) in complete remission (CR) who had received a single-unit UCBT from a KIR-ligand-compatible or -incompatible donor. Grafts were HLA-A, -B or -DRB1 matched (n = 21) or mismatched (n = 197). Patients and donors were categorized according to their degree of KIR-ligand compatibility in the graft-versus-host direction by determining whether or not they expressed HLA-C group 1 or 2, HLA-Bw4 or HLA-A3/-A11. Both HLA-C/-B KIR-ligand- and HLA-A-A3/-A11 KIR-ligand-incompatible UCBT showed a trend to improved LFS (P = 0·09 and 0·13, respectively). Sixty-nine donor–patient pairs were HLA-A, -B or -C KIR-ligand incompatible and 149 compatible. KIR-ligand incompatible UCBT showed improved LFS [hazard ratio (HR) = 2·05, P = 0·0016) and overall survival (OS) (HR = 2·0, P = 0·004) and decreased RI (HR = 0·53, P = 0·05). These results were more evident for AML transplant recipients (2-year LFS and RI with or without KIR-ligand incompatibility 73% vs. 38% (P = 0·012) and 5% vs. 36% (P = 0·005), respectively). UCBT for acute leukaemia in CR from KIR-ligand incompatible donors was associated with decreased RI and improved LFS and OS. These results indicate that typing for HLA-C might be helpful, especially in AML patients (Willemze et al, 2009).

These results were not confirmed in a study that analysed 251 recipients of single or double unit cord blood grafts following myeloablative or reduced intensity conditioning (Brunstein et al, 2009). Discrepancy of the results could be explained by small number and bias of selection of patients and the heterogeneity of the transplant protocols. Large multi-institutional studies are necessary to understand the role of histocompatibility in UCBT.

New methods for facilitating engraftment and improving results

Double cord blood transplant

One of the main obstacles for extending the indications of UCBT is that a CD34+ cells dose inferior to 2 × 105/kg or inferior to 3 × 107/kg NC at infusion decreases the chance of engraftment and survival. The minimum number of cells to be infused increases further with increasing number of HLA mismatches and diagnosis of a non-malignant disease. In order to address this question, the Minnesota group proposed a study where two cord blood units were infused when the number of cells in each cord blood unit did not meet the requirements (Barker et al, 2003). The results were quite encouraging as they demonstrated that double UCBT increase the rate of engraftment, gives more graft-versus host effect and less relapse. Interestingly, they showed that despite the contribution of both units to short-term engraftment, in the long term, only one unit engrafts. Barker et al (2003) reported, in a double cord blood transplant study that utilized myeloablative conditioning, that 76% of patients at day 30 and 100% of patients at day 100 had only a single cord unit contributing to haematopoiesis. In a double UCBT study of 92 patients receiving a reduced intensity conditioning regimen, only one unit was detected in 91% of the patients and in all the patients beyond 1 year (Ballen et al, 2007; Brunstein et al, 2009). No studies have reliably identified preinfusion variables that significantly influence cord blood predominance (Delaney et al, 2009).

Since the publication of these preliminary results, the number of double cord blood transplants has increased dramatically, with more than 450 cases reported to Eurocord. Most new cases have been performed in adults with malignant diseases who would not have been transplanted previously because of the absence of a suitable unrelated donor. Consequently, most patients who are candidates for an allogeneic HSCT, currently find a donor, resulting in a dramatic increase of the number of cord blood transplants performed in adults world wide (source Eurocord, EBMT, World Marrow Donor Association, NMDP).

Reduced intensity conditioning

In the last 10 years, in order to reduce transplant-related toxicity in elderly patients or in those with organ impairment, various non-myeloablative or reduced intensity conditioning (RIC) regimens have been used as an alternative to myeloablative conditioning (MAC). The purpose of RIC is to take advantage of the graft-versus-leukaemia effect of the immuno competent cells in the graft, rather than the antitumour effect of high-dose chemo-radiotherapy. The higher graft failure and delayed neutrophil and immune recovery after UCBT in the MAC setting, have suggested that cord blood cells should not be used in the RIC setting. However, phase II trials and preliminary retrospective studies have shown encouraging results in adults with acute leukaemia or other haematological malignant diseases (Barker et al, 2003; Brunstein et al, 2007). The Minnesota group described very encouraging results and analysed risk factors for outcomes after RIC-UCBT (Brunstein et al, 2009). They used a conditioning regimen consisting of Fludarabine (200 mg/m2), cyclophosphamide (50 mg/kg) and a single fraction of 2 Gy total body irradiation with ciclosporin and mycophenolate mofetil for post-transplantation immunosuppression. The target cell dose for the UCB graft was 3·0 × 107 nucleated cells/kg, resulting in the selection of a second, partially HLA-matched UCB unit in 85%. One hundred and ten patients with various haematological diseases were enrolled. Neutrophil recovery was achieved in 92% at a median of 12 days. Incidences of grades III and IV acute and chronic GVHD were 22% and 23%, respectively. TRM was 26% at 3 years. OS and event-free survival (EFS) at 3 years were 45% and 38%, respectively. Favourable risk factors for survival were absence of high-risk clinical features (Karnofsky 50–60, serious organ dysfunction, recent fungal infection, P < 0·01) and absence of severe GVHD (P = 0·04), and favourable risk factors for EFS were absence of high-risk clinical features (P < 0·01) and use of 2 UCB units (P = 0·07). These findings support the use of UCB after a non-myeloablative conditioning as a strategy for extending the availability of transplantation therapy, particularly for older patients.

Results of RIC-UCBT gather with the use of double cord blood transplantation explain the increased use of cord blood cells to treat adults and elderly patients with leukaemia and other haematological malignancies who otherwise would not receive this treatment.

However, many questions have still to be answered regarding RIC-UCBT, such as the importance of cell dose, number of HLA disparities and other donor related factors, the use of single or double UCBT, the type of RIC and GVHD prophylaxis regimen that are associated with improved outcomes. Currently, some groups are trying to answer those questions. Importantly, a longer follow-up is needed to establish the role of cord blood cells in the RIC setting compared to other sources of HSC.

Intrabone infusion

It was postulated that direct intra bone transplantation of cord blood cells could improve haematological recovery as a result of better stem-cell homing. This postulation was made on the basis of the following findings: stem cells recirculate in animals irradiated with limb shielding; only a small fraction of cells injected intra venously migrate to haematopoietic sites, in a mouse model direct intra bone injection of haematopoietic stem cells resulted in repopulation of the marrow of lethally irradiated mice that was 10-times more efficient than when haematopoietic stem cells were injected intravenously. Frassoni et al (2008) performed a phase I/II study on 32 patients with leukaemia who received a single cord blood transplant injected directly in the iliac crest after standard myeloablative conditioning. No complication occurred during the infusion. All patients showed fast neutrophil and platelet engraftment engrafted fast on neutrophils except two patients who died too early. The incidence of severe GVH effect was very low. These preliminary encouraging results justify a larger multi institutional study.

Non-haematopoietic stem cells from cord blood and placenta

Stem cells can be used to restore tissue function either as integrated participants in the target tissue or as vehicle that deliver complex signals to a target tissue without actually integrating into the tissue (Daley & Scadden, 2008). Several sources of stem cells have been explored for potential use in regenerative medicine; they include embryonic stem cells, fetal cells, cord blood cells and adult cells. More recently, induced pluripotent stem cells (iPS) collected from adult cells seem to reprogram embryonic properties but many obstacles remain before their medical and pharmaceutical applications can be fully realized. Human iPS cells have been generated from skin fibroblasts, keratinocytes and blood progenitor cells. Other sources include cell banks, such as those for cord blood; it would be extremely useful if iPS cells could be generated from cord blood cells (Yamanaka, 2009).

In conclusion, much has been learned in a relatively short time regarding the properties of cord blood haematopoietic progenitors and their clinical application. UCBT needs to meet several new challenges. First, several methods to improve of the speed of engraftment and decrease TRM are currently under investigation, such as increasing the donor pool to decrease the number of HLA mismatches or the use of double cord blood transplants. Other methods currently being investigated such as cord blood intra-bone infusion, ex-vivo expansion with cytokine cocktails, mesenchymal stem cells (MSC) or homing factors.

More interestingly, non-haematopoietic stem cells have been isolated from cord blood and placenta. These cells can be grown and differentiate into various tissues including MSC, bone, cartilage, liver, pancreas, neurones, endothelial cells, muscle, keratinocytes etc. They have an advantage over other sources of stem cells embryonic stem cells or iPS cells because the supply is unlimited, they can be used in autologous or allogeneic situations, they need minimal manipulation and they raise no ethical concerns. Future studies will, in the near future, test the potential of cord blood cells for the treatment of several diseases including, among other possibilities, diabetes, arthritis, burns, neurological disorders and myocardial infarction.