Article Text

Umbilical cord blood transplantation
  1. A M WILL, Consultant Paediatric Haematologist
  1. Manchester Children’s Hospital
  2. Pendlbury, Manchester M27 4HA, UK

    Statistics from

    Since Gluckman et al’s first description of successful haemopoietic stem cell transplantation using umbilical cord blood (UCB) as the source of marrow progenitors in 1989,1more than 500 umbilical cord blood stem cell transplants have been performed. Umbilical cord blood banks have been set up in the USA and Europe to store these cells, which in the past have been considered a waste product of reproduction. Although most UCB haemopoietic stem cell transplants have taken place in the past three years, recently published reports2-5 have given an important insight into the clinical potential of UCB as a source of haemopoietic progenitor cells for transplantation.

    Allogeneic bone marrow transplantation (BMT) can potentially be used to cure a variety of diseases including haematological malignancies, bone marrow failure syndromes, haemoglobinopathies, immunodeficiencies, and some inborn errors of metabolism.6 The use of allogeneic BMT is limited by the need for adequate tissue matching of host and donor cells to reduce the risks of rejection and the severity of graft versus host disease (GVHD) in the short term, while allowing immune reconstitution in the longer term. Many patients who might benefit from allogeneic BMT are prevented from doing so because there is no adequately matched donor available. In part, this problem has been addressed by the establishment of large panels of unrelated adult donors who are prepared to donate their bone marrow. Approximately five million donors are available worldwide. Despite these large numbers the need for precise tissue matching compounded by the predominantly white European ethnicity of the donor panels means that a significant number of patients remain unable to benefit from BMT. There is a need for a source of haematopoietic stem cells with a less rigorous requirement for tissue matching that could be used for transplantation into patients who do not have a conventional donor. UCB cells are possible candidates for this clinical role.

    Umbilical cord blood

    Laboratory studies7-9 have demonstrated and clinical use confirmed that UCB is a rich source of haemopoietic stem cells. UCB contains an increased proportion of the earliest progenitors, and per nucleated cell UCB has approximately 10 times the repopulation potential of bone marrow. Moreover, in vitro studies10-13have suggested that naive UCB lymphocytes are potentially less immunologically active than those usually found in the blood or marrow and may therefore produce fewer problems with GVHD than functionally mature lymphocytes harvested from live donors.

    Harvesting umbilical cord cells

    Umbilical cord cells can be harvested following vaginal delivery by cannulation of the placental vessels with the placenta still in utero or by direct expression of cord blood from the placenta following its delivery (usually in a room adjacent to the delivery suit) or at the time of caesarean section. There appears to be little difference in the quality of cells harvested by either method following vaginal delivery but cells from caesarean section may be less satisfactory14; however, these recent data are not borne out by an earlier study.15 A significant minority of postplacental delivery harvests provide an inadequate volume of UCB and are discarded. In Europe it is considered best practice that where possible UCB is harvested by specially trained staff working for the cord blood banks. However, in the USA harvesting by different obstetric teams has not been associated with a reduction in UCB quality.16

    For directed family donations, cannulation of the placental vessels in utero is the preferred option to reduce the risk of an inadequate harvest. Where UCB is being donated for unrelated use postplacental delivery harvesting is more appropriate as it is less intrusive and reduces the need for donor counselling. Only mothers who produce adequate UCB harvests need to be approached for permission to store and test the UCB and to have a medical history taken. This history is relied on to ensure that the cord blood is unlikely to carry any known genetic disorders. The chances of a fetus carrying an unknown but clinically significant congenital disorder is negligible.

    A very important part of the procedure is to ensure that the midwife or obstetrician attending the delivery does not clamp the cord earlier than usual just because the UCB is to be harvested. This prevents any potential harm to the fetus. Harvests over 40 ml are considered suitable for use. Usually between 40 and 150 ml are obtained, which approximates to between 4 and 11 × 108 nucleated cells available for transplantation.

    Samples are taken at the time of harvesting from the mother for microbiological screening, including HIV I and II, hepatitis B and C, and syphilis. Retesting of the donor for HIV at a later date is not performed because the risk of silent HIV infection at the time of delivery is extremely small, and failure to locate donors would lead to an unacceptable waste of stored harvested cord bloods.17As well as being tissue typed, a specimen of cord blood is often kept aside for HIV retesting should the UCB be required for transplantation. Further tests at the time of transplantation are probably unnecessary and are not performed on other types of stem cells stored in liquid nitrogen before use in transplantation. The UCB may be volume reduced but is most often directly mixed with DMSO to allow for long term storage in liquid nitrogen following a brief period in a quarantine freezer awaiting the results of virology tests. The UCB is also tested for bacteria. Improved harvest techniques have reduced the incidence of bacterial contamination.18

    The costs involved in harvesting and banking UCB are not insignificant. If it is to be done properly unrelated cord blood banking requires a considerable outlay in equipment, storage facilities, and trained staff. Each cord harvest costs approximately £2500.

    Theoretical advantages and disadvantages of UCB transplantation

    The advantages to the donor are obvious. As long as care is taken to clamp the cord at the normal time the fetus will come to no harm. There is no need to anaesthetise the donor. The donor feels no discomfort and does not require time off work or school.

    Potential advantage for the recipient is crucially dependent on the theoretical immune tolerance of UCB lymphocytes that might allow for transplantation across presently insurmountable HLA barriers. If successful engraftment is possible across HLA barriers and GVHD is controllable, the potential donor pool will be greatly increased and many more patients will be able to undergo curative haemopoietic stem cell transplantation. There are other possible advantages. UCB cells are potentially immediately available. Thus the time from consideration of transplantation to the actual procedure should be reduced compared to donation from an unrelated panel donor. This may be important for patients with malignancies and some inborn errors of metabolism where delays in transplantation may adversely affect outcome. The risk of transmissible infection should also be less. The UCB comes already virologically tested and will be free from postnatally acquired infections such as cytomegalovirus.

    There are disadvantages. The reduced number of stem cells present in the UCB may limit their use to small recipients. The theoretical immune naivete of the transplanted lymphocytes may reduce the graft versus leukaemia effect of the UCB transplants with an increased risk of relapse in patients transplanted for malignancy. Where relapse does occur the donor is not available to donate peripheral blood lymphocytes that could be used to treat the relapse.

    Clinical results


    Eurocord4 has recently reported the results of 78 related UCB transplants performed in Europe between October 1988 and December 1996 with a median follow up of 29 months. The patients were mainly children; median age was 5 years (range 0.2 to 20) and median weight was 19 kg (range 5 to 50). Thirty eight patients had leukaemia. In 60 cases the donor was an HLA identical sibling and in 18 there was a degree of HLA disparity. Induction regimens were standard with 36 patients receiving cyclophosphamide and total body irradiation, and 40 chemotherapy alone, which included busulphan. Post-transplant immunosuppression was with cyclosporin alone or in combination with prednisolone or methotrexate or both. Only four patients received serotherapy with either antithymocyte globulin or monoclonal antibodies.

    Overall 53 patients survived one year; 44 in the HLA matched group and six in the HLA mismatched group. In 10 patients receiving UCB mismatched at three or more HLA antigens only one survived for 12 months. Neutrophil engraftment was better in patients younger than 6 years of age and weighing less than 20 kg. Thirty five of 48 patients receiving < 3.7 × 107 cells/kg had neutrophil engraftment compared with 34 of 40 who received > 3.7 × 107 cells/kg. Platelet engraftment was also favoured by age under 6 years and weight more than 20 kg. HLA disparity was significantly associated with poor platelet engraftment. Acute GVHD developed to grade II or more in 9% of the fully matched recipients and in nine of the mismatched transplants. 14.3% of evaluable patients went on to develop chronic GVHD. Ten of the 38 patients with leukaemia relapsed during the median follow up of 29 months. Survival was also significantly affected by recipient cytomegalovirus status: 31 of 40 cytomegalovirus negative recipients but only 16 of 36 cytomegalovirus positive recipients survived for 12 months.

    Results for related UCB grafts reported from the USA are broadly similar but with less acute GVHD. This may be related to the use of more intensive GVHD prophylaxis.3


    The results for 65 unrelated donor UCB transplantation from the Eurocord registry have also recently been reported.5 The patients had a median age of 9 years (range 0.3 to 45) and a median weight of 30 kg (range 4 to 90). Forty one patients had acute or chronic leukaemia. The median follow up was 10 months (range 1 to 30). Seven patients had an HLA “identical” donor at HLA-A, -B, and -DRB1; 43 had a one antigen mismatch; 11 had two mismatches; and two had three HLA differences.

    Nineteen patients survived for one year. Patients receiving > 3.7 × 107 nucleated cells/kg fared better with a survival of 41% at one year compared with 22% in those infused with fewer cells. Cytomegalovirus serology was important to outcome: 42% of cytomegalovirus negative recipients survived 12 months compared with 20% of cytomegalovirus positive patients. HLA differences did not affect survival. Engraftment was slow and related to the number of nucleated cells/kg infused. The probability by day 60 of a neutrophil count > 0.5 × 109 per litre was 87% and for a platelet count of > 20 × 109 per litre the probability was 39%. Acute GVHD of grade II or more was seen in 21 patients: eight grade II, nine grade III, and four grade IV. The incidence of acute GVHD was significantly less in cytomegalovirus negative recipients. Seven of the 41 patients with leukaemia relapsed.

    Results from the USA3 19 20 are better with two year survival of approximately 45%, and more rapid engraftment probably because of the greater use of growth factors post-transplant. There was a significant incidence of acute GVHD. Duke and Minneapolis20 reported an incidence of acute GVHD of ⩾ grade II at 35%; Cairo and Wagner,3 in a recent review, reported an incidence of acute GVHD ⩾ grade II of 57% with 9% of grade III or IV. In the largest study reported so far involving 272 cases,21 Rubinstein et al reported an incidence of acute GVHD of grade III or IV of 23%. The incidence of GVHD did not appear to be related to the apparent degree of HLA mismatch.


    The results of unrelated UCB transplantation compare favourably with results of unmanipulated unrelated BMT.22 However, in the UK at least, unmanipulated unrelated donor BMT is rarely if ever performed. The vast majority of British children treated with unmanipulated unrelated BMT receive serotherapy with antithymocyte globulin or monoclonal antibodies, and where appropriate the graft is also lymphodepleted before infusion. In 1996 the Bristol group23 reported results for 50 patients with relapsed acute lymphoblastic leukaemia transplanted with manipulated unrelated donor grafts in second complete remission. Event free survival at two years was 53%; 94% of patients engrafted; acute GVHD ⩾ grade II was seen in 12% (only three patients had grade III or IV acute GVHD); 26% relapsed. The same group have more recently reported manipulated unrelated BMT in 15 children and adolescents with poor prognosis Philadelphia positive acute lymphoblastic leukaemia.24 All initially engrafted; the incidence of acute GVHD was 13%; two year event free survival was 37%.


    UCB transplantation is feasible. There are apparent limitations owing to the small numbers of progenitor cells present in the harvests. Immunological naivete of cord blood has been demonstrated by the relatively reduced incidence of GVHD in umbilical cord transplants compared with unmanipulated unrelated donor BMT and by the poor outcome in patients who are cytomegalovirus positive before transplantation. However, despite routine prophylaxis, clinically significant acute GVHD was reported in all studies. Follow up is as yet too short to determine whether relapse rates are higher with cord blood than with bone marrow.

    Initial results are favourable for matched related UCB transplantation in patients younger than 6 years of age weighing less than 20 kg who are cytomegalovirus negative. For suitable patients UCB offers the opportunity of prompt haemopoietic stem cell transplantation without any danger or discomfort for the donor. Results in the unrelated setting are not as good as those for children undergoing lymphocyte depleted unrelated donor BMT. However, patients from non-white European populations for whom bone marrow donors are difficult to find on the existing donor panels may benefit from the greater proportion of ethnic minority donors represented in the cord blood banks.

    Future results of UCB transplantation will no doubt be better than those summarised above. In vitro expansion of the cord blood progenitor cells may allow heavier patients (> 20 kg) to benefit from the immunological naivete of umbilical lymphocytes.25 The clinical outcomes need to improve to ensure the future of unrelated cord blood transplantation and to justify the continued investment in cord blood banking. New alternatives are on the horizon. In particular the initial results of mismatched CD34 selected stem cell transplantation are encouraging.26


    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.