There are potentially three main sources of haematopoietic stem cells for transplantation – bone marrow (BM), peripheral blood stem cells (PBSC) and umbilical cord blood stem cells.

Bone Marrow

Bone marrow (BM) is traditionally harvested from multiple well, spaced sites on the iliac crests under general anaesthetic. Multiple aspirates are taken to obtain a transplantable size. The median number of nucleated cells collected is normally 2 x 10⁸/kg patient weight. Collection usually aims for a minimum CD34+ count of 2 x 10⁶/kg patient weight.

BM is sourced from related donors or adult volunteers on registries. Registry searches and confirmation of donor type and medical clearance take time and so there is a lead up time (search time of 3 – 6 months) required when using BM compared to using cord blood for example. Bone marrow does however have the advantage of being able to return to the donor for future donor lymphocyte infusions (DLIs). BM is collected under general anaesthetic and so although a safe procedure, carries the same risks as any process that requires general anaesthetic. BM stem cells collected have a lower nucleated cell count than peripheral blood stem cells but much larger nucleated and CD34+ stem cell count than cord blood stem cells. Engraftment with bone marrow haematopoietic stem cells is medium to fast, taking a median of around 21 days to neutrophil and platelet engraftment. BM transplantation carries a risk of causing GvHD and requires that patient and donor be well matched for HLA class I and II. BM transplantation carries a risk of transmitting infectious disease or acquired or congenital disorders if not uncovered as part of the donor work up.

Peripheral Blood Stem Cells (PBSC)

Peripheral blood stem cells (PBSCs) are collected after being mobilised from the marrow with granulocyte colony stimulating factor (G-CSF). G-CSF is administered over several days prior to collection. The stem cells are collected by apheresis. The median number of nucleated cells collected is normally 9 x 10⁸/kg patient weight. Collection usually aims for a minimum CD34+ count of 7 x 10⁶/kg patient weight. One advantage of PBSC over BM is that the donor can return for further apheresis if the number of cells collected is not sufficient.

Like BM, PBSC’s are also sourced from related donors or adult volunteers on registries and have the same issues of long lead up time for donor work up as BM. The donor is available for future collections for donor lymphocyte infusions. The collection process for PBSC is however easier and safer as it does not require general anaesthetic. G-CSF has not been shown to have any harmful short-term effects though the long-term effects are unknown. Total nucleated cells collected by PBSC are generally much higher than BM collections. Engraftment with PBSC is faster than with BM, taking a median of around 15 days to neutrophil and platelet engraftment. PBSC carries a risk of causing GvHD and requires that patient and donor be well matched for HLA class I and II. PBSC, like BM, also carries a risk of transmitting infectious disease or acquired or congenital disorders if not uncovered as part of the donor work up.

Cord Blood

Cord Blood Overview

Umbilical cord blood (CB) stem cells are collected post-partum either in utero in the delivery room during the third stage of labour before the placenta is delivered or outside the delivery room, ex utero from the freshly delivered placenta. In general, the in-utero collection method yields a larger volume and higher total nucleated cell count, though more recent studies have shown that with appropriate training it is possible to obtain high collection volumes and cell counts ex utero. Cord blood can be sourced from cord blood banks either in single or double units.

CB stem cells have the advantage of being immediately available. Collection is easy and harmless. A disadvantage is that the donor is not available for further collection should a donor lymphocyte infusion be required to rescue the patient from a failing graft. Total nucleated (median 3 x 10⁷/kg patient weight) and CD34+ cell counts (median 1.5 x 10⁵/kg patient weight) are much lower than those obtained from BM or PBSC collection. For this reason, CB was initially used predominantly for children. Use of double dose CB has increased the usage in adults. CB has a lower risk of transmitting infectious diseases compared to BM and PBSC but the risk of transmitting congenital disorders is unknown. Engraftment with CB haematopoietic stem cells is slower than with either BM or PBSC, taking up to 35 days to neutrophil engraftment and platelet engraftment can take even longer. Transplantation with CB derived stem cells does lead to a reduced incidence and severity of GvHD due to the relative immaturity of the immune system at birth. This allows less stringent HLA matching criteria for CB transplantation with one or two HLA gene mismatches at high resolution at HLA-A, B, C and DRB1 tolerated. Cell dose may be the more important factor for CB transplantation.

Cord Blood Banking

The first successful cord blood transplant was performed in a patient with Fanconi Anaemia by Gluckman et al in 1988 using cord blood stem cells from a HLA matched sibling. The patient is still alive and well today. This success prompted the setting up of the first unrelated cord blood bank in New York by Rubinstein et al in 1991. The number of unrelated cord blood banks around the world has steadily expanded since then and today (checked Jan 2020) there are over 730,000 CB units in public cord blood banks worldwide. Cord blood banking involves the collection, processing, testing, banking, registration, selection and release of cord blood unit under strict quality-controlled conditions for ultimate transplantation.

Organisations setting up CB banks do so either in or associated with hospitals with large maternity units, often aiming for units with more than 5,000 births a year. A typical strategy in countries with a predominantly Caucasian population is to work with hospitals based in regions with a large percentage of Black, Asian and Minority Ethnic (BAME) mothers. This increases the genetic diversity of the CB bank over that typically seen in the adult stem cell donor registries. Collection of CB from mothers requires full informed consent. In European Union member countries, the European Union Tissues and Cells Directive applies. This states under the heading of Consent that ‘The procurement of human tissues or cells shall be authorised only after all mandatory consent or authorisation requirements in force in the Member State concerned have been met’. Umbilical cord blood stem cells are collected post-partum either in utero in the delivery room during the third stage of labour before the placenta is delivered or outside the delivery room ex utero from the freshly delivered placenta. Collection ex utero does have a higher risk of introducing microbial contamination. In general, the in utero collection method yields a larger volume and higher total nucleated cell count, though more recent studies have shown that with appropriate training it is possible to obtain high collection volumes and cell counts ex utero.

Processing mainly involves volume reduction to optimise freezer storage facilities. Volume reduction involves removal of red blood cells and plasma, leaving the stem cell in a buffy coat of around 21ml. Semi-automated closed systems such as Sepax and other systems are available for processing of cord blood units. Volume reduced units have the additional advantage over whole units of requiring much less DMSO for freezing and can, depending on the size of the recipient, be transfused without having to first wash the unit.

The current practice is to perform a number of tests prior to banking with further tests carried out if and when a unit is reserved for a potential patient. Pre storage tests include full cell counts, especially nucleated red cell counts, TNC and CD34+ counts, HLA typing, ABO Rh testing and bacteriology testing of the cord blood unit. Haemoglobinopathy tests may also be undertaken. Also pre storage, microbiology markers including HIV, HCV, HBV, HTLV and CMV testing is undertaken on the mother. At reservation, the additional tests carried out are in part driven by the requirements of the transplant centre making the reservation and the county that Centre is based in. Additional tests typically include maternal HLA type, confirmatory HLA type of the cord blood unit and microbiological tests on the cord blood unit.

Banking of a cord blood unit typically follows a full medical review, a review of all processing data and all test results on the mother and the unit. Cleared units are control rate frozen and stored in liquid nitrogen. Units suitable for transplantation are registered with national and international stem cell or dedicated cord blood registries. In the UK, the NHS cord blood bank, which is part of NHSBT and the Anthony Nolan, collect and bank cord blood. The NHS cord blood units are also registered with NETCORD and with the World Marrow Donor Association (WMDA). NETCORD runs an accreditation program for cord blood banks which many banks around the world are either accredited to or are seeking accreditation to.

CB banking has many advantages, one of the principle ones being the ready availability of CB units for stem cell transplantation. Sourcing of adult stem cells from registries can take anything from 3 – 6 months for the search, selection of potential donors, contact with those donors to confirm willingness to donate, performance of confirmatory typing and any additional tests and medical clearance of the donor. CB stem cells have the advantage of being immediately available. Collection is easy and harmless to the donor compared to adult bone marrow or stem cell donation which involve either a general anaesthetic and the risks that poses or the use of GCSF and the unknown long term risks of that process.

The targeting of hospitals with a high percentage of BAME mother means that CB banks typically have a higher representation of donations from BAME donors and a higher proportion ‘rare’ HLA types compared to adult stem cell registries. The British Bone Marrow Registry (BBMR) for instance has around 5% of donors registered as being from minority ethnic backgrounds whilst 20% of the cord blood units registered for searches by the NHS cord blood bank are from minority ethnic donors. In terms of HLA type, the NHS CB bank has a much higher HLA-A/B/DRB1 allele frequency of the rarer HLA types compared to the BBMR.

Cord Blood Transplantation

Transplantation with CB derived stem cells does lead to a reduced incidence and severity of GvHD due to the relative immaturity of the immune system at birth. This allows less stringent HLA matching criteria for cord blood transplantation with one or two HLA gene mismatches tolerated. Cell dose may be the more important factor for cord blood transplantation.

CB units have a lower risk of transmitting infectious diseases compared to BM and PBSC due to the lack of exposure of the donor.

Transplantation with Cord Blood derived stem cells has a much lower incidence and severity of GvHD compared to BM and PBSC due to the relative immaturity of the immune system at birth. This can be explained by the lower cell numbers and the mostly naive repertoire of cord blood T cells. Lower GvHD is one of the main advantages of cord blood banking as the reduced incidence and severity of GvHD allows for a less stringent HLA match thereby increasing the pool of potential donations for a given patient. The Graft versus Leukemia (GvL) effect does appear to be preserved.

The main disadvantage of CB transplantation is that CB units have a much lower Total Nucleated Cell (TNC) count than adult stem cells from either BM or PBSC. Total nucleated cell counts have a typical median of 0.3 x 10⁸/kg patient weight and CD34+ cell counts have a typical median of 0.2 x 10⁶/kg patient weight. This is tenfold lower than is typically collected from adult stem cells. PBSC also has the advantage of being able to collect from the donor over several days until an adequate volume is collected. This option is not available to for CB. This disadvantage can be overcome to some extent by the use of double CB transplants. Double CB transplantation involves the transplantation of two cord units together in series to a patient. Data to date shows no increase in GvHD as a result of using two cord units rather than one. Interestingly, sustained haematopoiesis after double cord blood transplantation is usually by a single cord unit though the criteria that make one unit prevail over the other remains to be clarified.

The other disadvantage of the using a cord unit is that the donor is not available for DLI should one be required to rescue the patient from a failing graft. In addition, there is a lack of follow up of the donor, so it is not known if the donor later develops a congenital disorder. The haemoglobinopathy tests obtained at banking do help in this regard.

Engraftment with CB haematopoietic stem cells is also slower than with either BM or PBSC, taking up to 35 days to neutrophil engraftment and platelet engraftment can take even longer. During this time the patient has a delayed immune reconstitution with associated risk of infectious morbidity and mortality. CB grafts also have a higher incidence of graft failure compared to BM and PBSC.

A number of studies comparing CB and adult stem cell transplant in children with AML showed similar 5-year disease free survival. In adults transplanted for malignant diseases, the mortality due to delayed engraftment appears to be counterbalanced by a reduction in acute GvHD so that the overall incidence of transplant related mortality, treatment failure and overall mortality are similar in patients receiving cord blood and adult stem cells. Despite this, CB transplantation in adults is still generally performed as a last resort if adult stem cells were not available. CB transplantation is used more frequently in children.

When sourcing donors for children, transplant units generally ask for a cord blood unit search at the same time they are undertaking an adult stem cell donor search. The indications for CB transplant are the same as those for adult stem cell transplant provided cell counts are adequate, though results with bone marrow failure syndromes are less satisfactory.

Guidelines for selecting cord blood units for transplantation recommend the following:

• Select a CB unit high-resolution HLA matched 8/8 at HLA-A, B, C and DRB1. TNC dose should be > 3 x 10⁷/kg at freezing or 2.0-2.5 x 10⁷/kg after thawing
• If selecting a CB unit high-resolution HLA matched 5/8, 6/8 or 7/8 at HLA-A, B, C and DRB1, TNC dose should be > 5 x 10⁷/kg. HLA antibody testing must be undertaken to avoid donor specific antibodies (DSAs)
• Use of CB units < 3 x 10⁷/kg patient weight at freezing and/or 4 or more HLA mismatches at HLA-A, B, C and DRB1 is not recommended
• CD34 counts should be treated with caution due to variations in technique. Aim for 1.5 x 10⁵/kg at freezing or 1 x 10⁵/kg after thawing
• If several units with the same degree of HLA match are available, the one with the highest cell dose should be chosen
• Cell viability and microbiology results must be taken into account
• In the UK a graft advisory panel is available to help with selection of cord units for ‘complicated’ patients

Future challenges for CB transplantation are to develop strategies to reduce the time to engraftment. Use of double cords is helping in that regard. Other strategies under investigation include the co- transplantation of stem cells from a haploidentical sibling and the ex vivo expansion of CD34+ cells. Aggressive early and pre-emptive therapy has been suggested as a means of overcoming the infection rates prior to engraftment.