- Email: [email protected]
Low usage rate of banked sibling cord blood units in hematopoietic stem cell transplantation for children with hematological malignancies: Implications for directed cord blood banking policies
Blood Cells, Molecules, and Diseases 46 (2011) 177–181
Contents lists available at ScienceDirect
Blood Cells, Molecules, and Diseases j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y b c m d
Low usage rate of banked sibling cord blood units in hematopoietic stem cell transplantation for children with hematological malignancies: Implications for directed cord blood banking policies Evgenios Goussetis a,⁎, Ioulia Peristeri a, Vasiliki Kitra a, Andreas C. Papassavas b, Maria Theodosaki a, Eftichia Petrakou a, Antonia Spiropoulos a, Anna Paisiou a, Alexandra Soldatou a, Catherine Stavropoulos-Giokas b, Stelios Graphakos a a b
Stem Cell Transplant Unit, Aghia Sophia Children's Hospital, Thivon and Papadiamantopoulou, 11527 Athens, Greece Hellenic Cord Blood Bank, Foundation for Biomedical Research, Academy of Athens, Athens, Greece
a r t i c l e
i n f o
Article history: Submitted 8 October 2010 Revised 18 October 2010 Available online 20 November 2010 (Communicated by G. Stamatoyannopoulos, M.D., Dr. Sci., 22 October 2010) Keywords: Directed cord blood banking Hematological malignancies
a b s t r a c t Directed sibling cord blood banking is indicated in women delivering healthy babies who already have a sibling with a disease that is potentially treatable with an allogeneic cord blood transplant. We evaluated the effectiveness of a national directed cord blood banking program in sibling HLA-identical stem cell transplantation for hematological malignancies and the factors influencing the usage rate of the stored cord blood units. Fifty families were enrolled from which, 48 cord blood units were successfully collected and 2 collections failed due to damaged cord/placenta at delivery. Among enrolled families 4 children needed transplantation; however, only one was successfully transplanted using the collected cord blood unit containing 2 × 107 nucleated cells/kg in conjunction with a small volume of bone marrow from the same HLAidentical donor. Two children received grafts from matched unrelated donors because their sibling cord blood was HLA-haploidentical, while the fourth one received bone marrow from his HLA-identical brother, since cord blood could not be collected due to damaged cord/placenta at delivery. With a median follow-up of 6 years (range, 2–12) for the 9 remaining HLA-matched cord blood units, none from the prospective recipients needed transplantation. The low utilization rate of sibling cord blood in the setting of hematopoietic stem cell transplantation for pediatric hematological malignant diseases necessitates the development of directed cord blood banking programs that limit long-term storage for banked cord blood units with low probability of usage such as non-HLA-identical or identical to patients who are in long-term complete remission. © 2010 Elsevier Inc. All rights reserved.
Introduction Umbilical cord blood transplantation has been successfully used for the treatment of both non-malignant and malignant diseases in pediatric and adult patients. Since the first matched sibling cord blood transplantation in 1988 [1] and the first from unrelated cord blood in 1993 [2], the majority have been performed for malignant diseases [3]. Approximately 2000 children with hematological malignancies have been transplanted with a related (n = 199) or unrelated cord blood transplantation (n = 1663) and reported to Eurocord registry from 1990 to 2008 [4]. The few studies reporting outcomes of related cord blood transplantation in pediatric patients with hematological malignancies suggest that cord blood is as effective as bone marrow in the setting of the related hematopoietic transplantation [5,6]. These ⁎ Corresponding author. SCT-Unit, Aghia Sofia Children's Hospital, Thivon and Mikras Asias, 11527 Athens, Greece. Fax: + 30 210 7792200. E-mail address: [email protected] (E. Goussetis). 1079-9796/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.bcmd.2010.10.017
results support banking of sibling cord blood units for use in transplantation of pediatric patients with hematological malignancies. Several cord blood banks for unrelated transplants also offer directed-donor banking of cord blood. However, in directed sibling cord blood banking parameters such as minimum volume collected or minimum cell number do not apply [7]. In addition, clear policies regarding the duration of storage for cord blood units that are unlikely to be used, such as those that are non-identical to the patients, must be put in place. Recent experience reported by cord blood banks in United States and the United Kingdom has shown that the utilization rate of stored related cord blood units is low, particularly in the setting of malignant diseases [8,9]. We communicate our experience in directed sibling cord blood banking, evaluating the effectiveness of a national program in sibling HLA-matched cord blood transplantation for hematological malignancies in Greece. We also propose storage policies for cord blood units with low likelihood to be used in order to offer a cost-effective national service for directed sibling cod blood banking.
178
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181
Patients and methods
requested that the unit is transferred to a private bank or used in research.
Eligibility and written consent Families were eligible if they met the following criteria: (1) the family was already caring for a child with hematological malignancy considered to be treatable by allogeneic stem cell transplantation, (2) the family was expecting the birth of a healthy full sibling, and (3) the child's physician was supportive of the cord blood banking. Families were enrolled in an approved research protocol; written informed consent was obtained from the donor mother for screening of both the mother and the cord blood for mandatory microbiological markers, for tissue typing, and for the storage of the cord blood unit. At enrolment, the referring physician completed a standardized data form to assess the child's disease status and potential suitability for hematopoietic stem cell transplantation. A blood sample from the sibling for reference HLA-typing was also obtained. Families were not charged for cord blood banking. Enrolment began May 01, 1994 until April 2008. Collection, processing, and storage of sibling cord blood Cord blood was collected before delivery of the placenta using a 250-ml blood bag with 35 ml citratephosphate-dextrose anticoagulant. Cord blood units were considered suitable for banking based on a net volume of ≥20 ml, since even for low volume collections there is the possibility of combining cord blood with a small volume of bone marrow from the same donor. Blood samples for HLA-typing, blood grouping, testing for infectious and genetic diseases, and for total and CD34+ cells counting were removed. All cord blood units collected between 1994 and 2004 (n = 23) were processed without volume reduction and stored in the bone marrow processing laboratory of the Stem Cell Transplant Unit at Aghia Sophia Children's Hospital. Cryoprotectant solution was prepared from dimethyl sulfoxide (DMSO) and 5% (wt/vol) human albumin solution. An equal volume of cooled 20% (vol/vol) DMSO solution was added to the cord blood in a controlled manner. Cord blood cells were cryopreserved in freezing bags (Gabro, Haemofreeze, Amsterdam) using an automated, controlled cell freezer (Cryo 10, CryoMed). In 2004, governmental funding facilitated the establishment of the first national cord blood bank, where 25 additional cord blood units were processed and stored as described by Rubinstein et al. [10]. Briefly, cryoprotectant solution was prepared from dimethyl sulfoxide (DMSO) and 10% (wt/vol) dextran-40/saline solution. An equal volume of sterile, cooled 20% (vol/vol) DMSO solution was added to the cord blood in a controlled manner. The cord blood unit after completion of the freezing program were stored in the liquid phase of a liquid nitrogen freezer at −196 °C. A small volume was taken post-processing for blood culture testing (1.0 ml for aerobic and anaerobic culture) before the cells were frozen in cryocyte bags. Reference samples were cryopreserved with the cord blood unit and stored in the same tank. Long-term storage policy All cord blood units were stored until HLA-typing results. Those units with at least one haplotype match were further stored. The decision on long-term storage was not dependent on volume, cellularity, and patient weight. A written report from the referring physician regarding the disease status and the need for transplantation of his/her patient to the institution banking sibling cord blood unit was required. Patients with a mismatched cord blood unit who needed transplantation and found a matched unrelated donor were also documented. After a 5-year time interval, further storage was reevaluated considering the disease status and the likelihood of using the cord blood unit. For all units decided to not store longer, our policy is to advise the family to donate the unit to the unrelated cord blood unit pool provided it fulfils the criteria for this program; otherwise we
Release for transplantation Release required a specific written request from the transplantation physician. Frozen reference samples were thawed for confirmatory HLA-typing. Only one sibling cord blood unit was released and transplanted. The recipient was a 3-year old boy with ALL in second remission and cord blood from his sister was collected 8 months earlier. We had to use, however, both cord blood and bone marrow from the same HLA-matched sibling as the source of allogeneic hematopoietic stem cells. Bone marrow cells were used because the number of cord blood nucleated cells was considered insufficient to ensure engraftment. Transplantation The preparative regimen included the combination of oral Busulfan 16 mg/kg, etoposide 40 mg/kg, and cyclophosphamide 120 mg/kg. Following parental consent, we harvested a reduced bone marrow volume of 120 ml from the sibling, who was at that time 8 months old and weighed 10 kg. Cord blood unit was thawed in a 37 °C waterbath with gentle agitation and immediately infused into the patient without further processing. Bone marrow cells were given first. Graft versus host disease prophylaxis consisted of cyclosporine and a short course of methotrexate. Supportive care included the prophylactic use of acyclovir, fluconazole, ciprofloxacin, and pentamidine. Neutrophil engraftment was defined as the first of 3 consecutive days with an absolute neutrophil count equal to or higher than 500/μl. Platelet engraftment was defined as the first of 3 consecutive days with a platelet count equal to or higher than 20000/μl with at least 5 days away from the last platelet transfusion. Donor hematopoiesis was documented by PCR study of genetic polymorphism of short tandem repeats. Results Enrolment characteristics Over a 14-year period, 50 requests were received for collection; 48 were successfully performed, while two were not due to damaged placenta at delivery. Diagnoses included acute lymphoblastic leukemia (42), acute myeloblastic leukemia (4), non-Hodgkin lymphoma (1), and chronic myeloid leukemia (1). All 48 cord blood units collected were banked. All siblings were at the time of collection in complete remission. In 5 cases, the mother was pregnant at time of diagnosis, while in the remaining cases the sibling was born at a median time of 2 years (range 1–10) after diagnosis. In one case preimplantation genetic diagnosis for HLA-typing was used to conceive a “saviour sibling” for the sibling with chronic myeloid leukemia (CML). The requests for banking increased the last 4 years since the establishment of a national cord blood bank. Characteristics of cord blood collections Table 1 gives an overview of the collected cord blood characteristics. The volume was ≥50 ml and total nucleated cells ≥3 × 108 in all units. The mean cord blood unit volume without anticoagulant was 82 ml ± 29.1; the mean total nucleated cell count was 7.08 × 108 ± 2.98; and the mean CD34+ cell count was 2.38 × 106 ± 1.36. All banked cord blood units were free of bacterial contamination. HLA-A, HLA-B, and HLA-DR typing was performed serologically. Of the 48 cord blood collections 10 were HLA-identical to the siblings, while two additional units had 2 HLA-antigen mismatches.
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181 Table 1 Laboratory characteristics of banked sibling cord blood units.
Volume minus anticoagulant, ml Nucleated cell count × 108 Total CD34+ cells/collection × 106 Viability %
Mean
SD
Range
82 7.08 2.38 91.6
29.1 2.98. 1.36 7.41
(50–146) (3–22.4) (0.5–6.3) (73.7–100)
Factors influencing the usage rate We found three factors to be associated with the usage of a cord blood unit in transplantation (Fig. 1). These included HLA histocompatibility, disease in long-term remission without need for transplantation, and collection failure. Among enrolled families 4 children needed transplantation; however, only one was used in transplantation in conjunction with a small volume of bone marrow from the same HLA-identical donor. With a median follow-up of 6 years (range, 2–12) for 9 remaining HLA-matched cord blood units, none from the prospective recipients needed transplantation. All of them remain in complete remission for a time longer than 5 years. In contrast, among children with HLA-mismatched cord blood, two needed transplantation and received grafts from matched unrelated donors. One of the 2 failed cord blood collections was the case where the donor was born to be HLA-identical following preimplantation HLA-genotyping. Bone marrow cells from the donor at the age of 18 months were transplanted to the affected (CML) sibling resulting in successful hematopoietic reconstitution. Transplantation characteristics and outcome The cord blood net volume of 85 ml was harvested and transported immediately to processing laboratory. The total nucleated and CD34+cells were 4×108 and 2×106 corresponding to 2×107/kg and 0.75×105/kg, respectively. Patient was at the time of cord blood collection in complete remission out of chemotherapy. Unfortunately, 4 months later he relapsed. Then he received 3 blocks systemic chemotherapy and entered into a second remission. At this point a stem cell transplantation was planned. Having considered the cord blood stem cells insufficient to
179
secure a stable engraftment, we decided to enrich the cord blood unit with bone marrow collected from the 8-month-old sibling weighing 10 kg. Harvesting of 120 ml of bone marrow was performed from the infant providing an additional total number of nucleated cells 2.68×108/kg with CD34 2.5×106/kg. Collection was uneventful without significant decrease in Hb levels. Post-transplantation course was uneventful with minimal transfusion requirements, low-grade mucositis, and only one minor febrile episode during aplasia. The patient demonstrated engraftment with neutrophils above 500/μl on day 18 and platelets above 20000/μl on day 22. Full sustained stable donor-derived hematopoietic chimerism was repeatedly observed by short tandem repeats analysis. At the present time the patient is 12 years post-transplant and remains in complete remission with full hematopoietic reconstitution. Discussion Many studies have shown that hematopoietic stem cells derived from related or unrelated cord blood units are at least as effective as those derived from the bone marrow or growth factor-mobilized peripheral blood in allogeneic transplantation setting for hematologic malignancies [6,11,12]. Specifically for sibling cord blood transplantation the latest analysis of Eurocord-European Bone Marrow Transplantation registry showed that the probability of disease free survival and overall survival, for 147 patients receiving a related cord blood transplant with a median follow-up of 6.7 years, was 44% and 55%, respectively [13]. Most of these cord blood units were banked at the local institution where transplantation took place, in collaboration between highly motivated families, obstetricians and hematopoietic cell transplantation physicians. Our stem cell transplant unit is the only center offering an allogeneic hematopoietic stem cell transplantation program for pediatric patients in Greece. We initiated the service of directed sibling cord blood banking until 2004, where we were succeeded by the first public cord blood bank. This service included the contact with the obstetric delivery unit to arrange collection, processing, testing, storage, and transplantation. Over a 14year period we collected 48 cord blood units from respective families with pediatric hematological malignancies while in 2 additional families the collection failed due to placental damage. Interestingly, in 45 cases out 50, the donor was conceived after diagnosis of the
Fig. 1. Transplant needs in 50 siblings with hematological malignancy and utility of banked sibling cord blood units.
180
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181
sibling's disease and in one case pre-implantation HLA-typing was performed successfully in order to create a “saviour sibling.” The practice of conceiving an additional child with the intention of using it as a hematopoietic stem cell donor for an affected sibling has been done for many years. In the early years, parents planned to have another child by natural conception and hoped for an HLA-matched sibling. These attempts are now being conjoined with reproductive technologies such as in vitro fertilization combined with preimplantation genetic diagnosis to determine HLA tissue matching [14,15]. Both volume and cell counts (nucleated and CD34+ cells) of cord blood unit we collected are comparable to those reported by cord blood banks in several countries involved in collection of cord blood from siblings of patients with malignancies [9,16,17]. Our policy is to store every unit ≥20 ml since even low volume collections could be usable in combination with reduced bone marrow harvest from the same donor, thus resulting in sufficient grafts. In such cases, one may argue that it is quite possible to delay the transplant until the donor is old enough to donate an adequate number of bone marrow stem cells, thus making cord blood storage unnecessary. However, in cases with hematological malignancies, an urgent transplantation may be indicated at an early time point where the donor is not old enough to donate a sufficient number of bone marrow stem cells. In such cases, reduced volume bone marrow collection not exceeding 15 ml/kg can be an attractive option for enhancing the stem cell number in the graft without placing the donor at the same risk as a standard bone marrow collection. Indeed, such an approach has proven valuable in our patient who needed an urgent transplantation and the donor age (8 months) was a limiting factor for donating sufficient bone marrow stem cells. Regarding the usage rate of cord blood units in stem cell transplantation, our results are in line with the experience reported by cord blood banks in the United States and the United Kingdom, confirming the low utility rate even for HLAmatched cord blood units. Reeds et al. [8] described USA national experience of sibling cord blood banking and transplantation for children with malignant disease. Two hundred fifty cord blood units were banked from families with a member suffering from a malignancy. From the stored cord blood units only 4 were used for transplantation. Two of the 4 donor recipient pairs were HLAidentical, one had donor mismatched at 2 HLA antigens and one had donor mismatch at 3 HLA antigens. The patient who received a cord blood unit mismatched at 2 HLA antigens had high-risk leukemia, failed engraftment and died 3.7 months post transplant. The other three patients were in complete remission. Smythe et al. have reported their 10-year experience with directed sibling cord blood banking in the National Blood Service in England and Wales [9]. They have used 3 cord blood units out of 114 collected from families who had a member with hematological malignancy. Two of their patients engrafted but died from relapse 4.5 and 2.5 years post transplant. The remaining one, who received a combined graft of cord blood and bone marrow, is alive free of disease 4.5 years post transplantation. Although neither the English bank nor the American bank reported the number of siblings needing transplantation among the enrolled siblings with hematological malignancies, we assume that HLAdisparity and the high possibility of cure with primary chemotherapy for pediatric hematological malignancies were the two main limiting factors for carrying out cord blood transplantation in these series. In the last years two additional smaller banks reported directed related cord blood banking programs without, however, mentioning the usage rate in transplantation [17,18]. In our experience, among 50 enrolled families 4 children needed transplantation; only one was successfully transplanted, however, in conjunction with a small volume of bone marrow from the same HLAidentical donor. Two additional children received grafts from matched unrelated donors because their sibling cord blood cells were HLAhaploidentical and the last one, despite pre-implantation selection to
be HLA-identical to the affected with CML sibling, cord blood could not be collected due to damaged cord/placenta at delivery. Based on these results, we recommend that both HLA-mismatched and matched sibling cord blood units should be stored independently of cell count and volume. Although it is not clear whether sibling cord blood donors disclosing 1 and 2 HLA-antigen mismatches can be proven effective, encouraging results from unrelated cord blood donors with similar HLA-disparity and the fact of the donor availability make the storage of such cord blood units reasonable. Even haploidentical cord blood may be used in emergency cases [19]. However, cord blood units that do not share at least one HLAhaplotype with the respective recipient should not be further retained. Continued storage should, however, be reviewed at least annually to ensure that cord blood are not stored unnecessarily; in cases where the donor is older than 3 years and recipient in long term complete remission (N5 years) the transplant physician in cooperation with the referring physician may decide to interrupt the long term storage because a such cord blood unit is unlikely to be used. Contrary to sibling cord blood banking for non-malignant diseases, where the usage rate was very high [9,16,20], utilization of sibling cord blood in children with a malignant disease is quite limited. The usage rates of units collected in directed banking programs reported by us and others are within the range of 1.6–2.6%. This low usage rate reflects the fact that most of the children with malignancy needing transplantation will be HLA-mismatched to respective cord blood unit and most of those with an HLA-compatible unit will never be in need for transplantation. In only few children who developed recurrent disease, cord blood transplantation was a considerable treatment option. Despite the fact of limited utility of directed sibling cord blood banking in transplantation for hematological malignancies, we still believe that the cost of storing such cord blood units in a public cord blood bank is reasonable and by limiting the long-term storage of those with extremely low probability to be used, a cost-effective national service for directed cord blood banking for pediatric patients with hematological malignancies could be established.
References [1] E. Gluckman, H.A. Broxmeyer, A.D. Auerbach, H.S. Friedman, G.W. Douglas, A. Devergie, et al., Hematopoietic reconstitution in a patient with Fanconi's anemia by means of umbilical-cord blood from an HLA-identical sibling, N Engl J. Med. 321 (1989) 1174–1178. [2] J. Kurtzberg, M. Graham, J. Casey, J. Olson, C.E. Stevens, P. Rubinstein, The use of umbilical cord blood in mismatched related and unrelated hemopoietic stem cell transplantation, Blood Cells 20 (1994) 275–283. [3] V. Rocha, E. Gluckman, Eurocord-Netcord registry and European Blood and Marrow Transplant group. Improving outcomes of cord blood transplantation: HLA matching, cell dose and other graft- and transplantation-related factors, Br. J. Haematol. 147 (2009) 262–274. [4] V. Rocha, N. Kabbara, I. Ionescu, A. Ruggeri, D. Purtill, E. Gluckman, Pediatric related and unrelated cord blood transplantation for malignant diseases, Bone Marrow Transplant. 44 (2009) 653–659. [5] J.E. Wagner, N.A. Kernan, M. Steinbuch, H.E. Broxmeyer, E. Gluckman, Allogeneic sibling umbilical cord-blood transplantation in children with malignant and nonmalignant disease, Lancet 346 (1995) 214–219. [6] V. Rocha, J.E. Jr Wagner, K.A. Sobocinski, J.P. Klein, M.J. Zhang, M.M. Horowitz, et al., Graft-versus-host disease in children who have received a cord-blood or bone marrow transplant from an HLA-identical sibling. Eurocord and International Bone Marrow Transplant Registry Working Committee on Alternative Donor and Stem Cell Sources, N Engl J. Med. 342 (2000) 1846–1854. [7] C. Navarrete, M. Contreras, Cord blood banking: a historical perspective, Br. J. Haematol. 147 (2009) 236–245. [8] W. Reed, R. Smith, F. Dekovic, J.Y. Lee, J.D. Saba, E. Trachtenberg, et al., Comprehensive banking of sibling donor cord blood for children with malignant and nonmalignant disease, Blood 101 (2003) 351–357. [9] J. Smythe, S. Armitage, D. McDonald, D. Pampphilon, M. Cuttridge, J. Brown, et al., Directed sibling cord blood banking for transplantation: the 10-year experience in the national blood service in England, Stem Cells 25 (2007) 2087–2093. [10] P. Rubinstein, L. Dobrila, R.E. Rosenfield, J.W. Adamson, G. Migliaccio, A.R. Migliaccio, et al., Processing and cryopreservation of placental/umbilical cord blood for unrelated bone marrow reconstitution, Proc. Natl Acad. Sci. USA 92 (1995) 10119–10122. [11] J. Kurtzberg, V.K. Prasad, S.L. Carter, J.E. Wagner, L.A. Baxter-Lowe, D. Wall, et al., Results of the Cord Blood Transplantation Study (COBLT): clinical outcomes of
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181
[12]
[13]
[14] [15]
unrelated donor umbilical cord blood transplantation in pediatric patients with hematologic malignancies, Blood 112 (2008) 4318–4327. M. Eapen, P. Rubistein, M.J. Zhang, C. Stevens, J. Kurtzberg, A. Scaradavou, et al., Outcomes of transplantation of unrelated donor umbilical cord blood and bone marrow in children with acute leukaemia: a comparison study, Lancet 369 (2007) 1947–1954. A.L. Herr, N. Kabbara, C.M. Bonfim, P. Teira, F. Locatelli, K. Tiedemann, et al., Longterm follow-up and factors influencing outcomes after related HLA-identical cord blood transplantation for patients with malignancies: an analysis on behalf of Eurocord-EBMT, Blood 116 (2010) 1849–1856. Y. Verlinsky, S. Rechitsky, T. Sharapova, R. Morris, M. Taranissi, A. Kuliev, Preimplantation HLA testing, JAMA 291 (2004) 2079–2085. E.R. Morgan, J. Girod, J.S. Rinehart, Having a child to save a sibling: reassessing risks and benefits of creating stem cell donors, Pediatr. Blood Cancer 48 (2007) 249–253.
181
[16] M.C. Walters, L. Quirolo, E.T. Trachtenberg, S. Edwards, L. Hale, J. Lee, et al., Sibling donor cord blood transplantation for thalassemia major: experience of the sibling donor cord blood program, Ann. NY Acad. Sci. 1054 (2005) 206–213. [17] T. Brune, H.S.P. Garritsen, Potential of directed cord blood donation for the treatment of patients with hematologic disorders, Transfus. Med. Hemother. 34 (2007) 100–103. [18] M. Screnci, E. Murgi, D. Carmini, L. Piro, N. Cinelli, L. Laurenti, et al., Related cord blood banking for haematopoietic stem cell transplantation, Transfus. Med. 20 (2010) 185–190. [19] H.M. Katzenstein, E.R. Morgan, M. Olsewski, K. Danner-Koptik, M. Kletzel, Haploidentical related umbilical cord blood stem cell transplant in a child with acute non-lymphocytic leukemia, Bone Marrow Transplant. 19 (1997) 765–769. [20] E. Goussetis, E. Petrakou, M. Theodosaki, V. Kitra, I. Peristeri, G. Vessalas, et al., Directed sibling donor cord blood banking for children with β-thalassemia major in Greece: usage rate and outcome of transplantation for HLA-matched units, Blood Cells Mol. Dis. 44 (2010) 107–110.
Contents lists available at ScienceDirect
Blood Cells, Molecules, and Diseases j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y b c m d
Low usage rate of banked sibling cord blood units in hematopoietic stem cell transplantation for children with hematological malignancies: Implications for directed cord blood banking policies Evgenios Goussetis a,⁎, Ioulia Peristeri a, Vasiliki Kitra a, Andreas C. Papassavas b, Maria Theodosaki a, Eftichia Petrakou a, Antonia Spiropoulos a, Anna Paisiou a, Alexandra Soldatou a, Catherine Stavropoulos-Giokas b, Stelios Graphakos a a b
Stem Cell Transplant Unit, Aghia Sophia Children's Hospital, Thivon and Papadiamantopoulou, 11527 Athens, Greece Hellenic Cord Blood Bank, Foundation for Biomedical Research, Academy of Athens, Athens, Greece
a r t i c l e
i n f o
Article history: Submitted 8 October 2010 Revised 18 October 2010 Available online 20 November 2010 (Communicated by G. Stamatoyannopoulos, M.D., Dr. Sci., 22 October 2010) Keywords: Directed cord blood banking Hematological malignancies
a b s t r a c t Directed sibling cord blood banking is indicated in women delivering healthy babies who already have a sibling with a disease that is potentially treatable with an allogeneic cord blood transplant. We evaluated the effectiveness of a national directed cord blood banking program in sibling HLA-identical stem cell transplantation for hematological malignancies and the factors influencing the usage rate of the stored cord blood units. Fifty families were enrolled from which, 48 cord blood units were successfully collected and 2 collections failed due to damaged cord/placenta at delivery. Among enrolled families 4 children needed transplantation; however, only one was successfully transplanted using the collected cord blood unit containing 2 × 107 nucleated cells/kg in conjunction with a small volume of bone marrow from the same HLAidentical donor. Two children received grafts from matched unrelated donors because their sibling cord blood was HLA-haploidentical, while the fourth one received bone marrow from his HLA-identical brother, since cord blood could not be collected due to damaged cord/placenta at delivery. With a median follow-up of 6 years (range, 2–12) for the 9 remaining HLA-matched cord blood units, none from the prospective recipients needed transplantation. The low utilization rate of sibling cord blood in the setting of hematopoietic stem cell transplantation for pediatric hematological malignant diseases necessitates the development of directed cord blood banking programs that limit long-term storage for banked cord blood units with low probability of usage such as non-HLA-identical or identical to patients who are in long-term complete remission. © 2010 Elsevier Inc. All rights reserved.
Introduction Umbilical cord blood transplantation has been successfully used for the treatment of both non-malignant and malignant diseases in pediatric and adult patients. Since the first matched sibling cord blood transplantation in 1988 [1] and the first from unrelated cord blood in 1993 [2], the majority have been performed for malignant diseases [3]. Approximately 2000 children with hematological malignancies have been transplanted with a related (n = 199) or unrelated cord blood transplantation (n = 1663) and reported to Eurocord registry from 1990 to 2008 [4]. The few studies reporting outcomes of related cord blood transplantation in pediatric patients with hematological malignancies suggest that cord blood is as effective as bone marrow in the setting of the related hematopoietic transplantation [5,6]. These ⁎ Corresponding author. SCT-Unit, Aghia Sofia Children's Hospital, Thivon and Mikras Asias, 11527 Athens, Greece. Fax: + 30 210 7792200. E-mail address: [email protected] (E. Goussetis). 1079-9796/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.bcmd.2010.10.017
results support banking of sibling cord blood units for use in transplantation of pediatric patients with hematological malignancies. Several cord blood banks for unrelated transplants also offer directed-donor banking of cord blood. However, in directed sibling cord blood banking parameters such as minimum volume collected or minimum cell number do not apply [7]. In addition, clear policies regarding the duration of storage for cord blood units that are unlikely to be used, such as those that are non-identical to the patients, must be put in place. Recent experience reported by cord blood banks in United States and the United Kingdom has shown that the utilization rate of stored related cord blood units is low, particularly in the setting of malignant diseases [8,9]. We communicate our experience in directed sibling cord blood banking, evaluating the effectiveness of a national program in sibling HLA-matched cord blood transplantation for hematological malignancies in Greece. We also propose storage policies for cord blood units with low likelihood to be used in order to offer a cost-effective national service for directed sibling cod blood banking.
178
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181
Patients and methods
requested that the unit is transferred to a private bank or used in research.
Eligibility and written consent Families were eligible if they met the following criteria: (1) the family was already caring for a child with hematological malignancy considered to be treatable by allogeneic stem cell transplantation, (2) the family was expecting the birth of a healthy full sibling, and (3) the child's physician was supportive of the cord blood banking. Families were enrolled in an approved research protocol; written informed consent was obtained from the donor mother for screening of both the mother and the cord blood for mandatory microbiological markers, for tissue typing, and for the storage of the cord blood unit. At enrolment, the referring physician completed a standardized data form to assess the child's disease status and potential suitability for hematopoietic stem cell transplantation. A blood sample from the sibling for reference HLA-typing was also obtained. Families were not charged for cord blood banking. Enrolment began May 01, 1994 until April 2008. Collection, processing, and storage of sibling cord blood Cord blood was collected before delivery of the placenta using a 250-ml blood bag with 35 ml citratephosphate-dextrose anticoagulant. Cord blood units were considered suitable for banking based on a net volume of ≥20 ml, since even for low volume collections there is the possibility of combining cord blood with a small volume of bone marrow from the same donor. Blood samples for HLA-typing, blood grouping, testing for infectious and genetic diseases, and for total and CD34+ cells counting were removed. All cord blood units collected between 1994 and 2004 (n = 23) were processed without volume reduction and stored in the bone marrow processing laboratory of the Stem Cell Transplant Unit at Aghia Sophia Children's Hospital. Cryoprotectant solution was prepared from dimethyl sulfoxide (DMSO) and 5% (wt/vol) human albumin solution. An equal volume of cooled 20% (vol/vol) DMSO solution was added to the cord blood in a controlled manner. Cord blood cells were cryopreserved in freezing bags (Gabro, Haemofreeze, Amsterdam) using an automated, controlled cell freezer (Cryo 10, CryoMed). In 2004, governmental funding facilitated the establishment of the first national cord blood bank, where 25 additional cord blood units were processed and stored as described by Rubinstein et al. [10]. Briefly, cryoprotectant solution was prepared from dimethyl sulfoxide (DMSO) and 10% (wt/vol) dextran-40/saline solution. An equal volume of sterile, cooled 20% (vol/vol) DMSO solution was added to the cord blood in a controlled manner. The cord blood unit after completion of the freezing program were stored in the liquid phase of a liquid nitrogen freezer at −196 °C. A small volume was taken post-processing for blood culture testing (1.0 ml for aerobic and anaerobic culture) before the cells were frozen in cryocyte bags. Reference samples were cryopreserved with the cord blood unit and stored in the same tank. Long-term storage policy All cord blood units were stored until HLA-typing results. Those units with at least one haplotype match were further stored. The decision on long-term storage was not dependent on volume, cellularity, and patient weight. A written report from the referring physician regarding the disease status and the need for transplantation of his/her patient to the institution banking sibling cord blood unit was required. Patients with a mismatched cord blood unit who needed transplantation and found a matched unrelated donor were also documented. After a 5-year time interval, further storage was reevaluated considering the disease status and the likelihood of using the cord blood unit. For all units decided to not store longer, our policy is to advise the family to donate the unit to the unrelated cord blood unit pool provided it fulfils the criteria for this program; otherwise we
Release for transplantation Release required a specific written request from the transplantation physician. Frozen reference samples were thawed for confirmatory HLA-typing. Only one sibling cord blood unit was released and transplanted. The recipient was a 3-year old boy with ALL in second remission and cord blood from his sister was collected 8 months earlier. We had to use, however, both cord blood and bone marrow from the same HLA-matched sibling as the source of allogeneic hematopoietic stem cells. Bone marrow cells were used because the number of cord blood nucleated cells was considered insufficient to ensure engraftment. Transplantation The preparative regimen included the combination of oral Busulfan 16 mg/kg, etoposide 40 mg/kg, and cyclophosphamide 120 mg/kg. Following parental consent, we harvested a reduced bone marrow volume of 120 ml from the sibling, who was at that time 8 months old and weighed 10 kg. Cord blood unit was thawed in a 37 °C waterbath with gentle agitation and immediately infused into the patient without further processing. Bone marrow cells were given first. Graft versus host disease prophylaxis consisted of cyclosporine and a short course of methotrexate. Supportive care included the prophylactic use of acyclovir, fluconazole, ciprofloxacin, and pentamidine. Neutrophil engraftment was defined as the first of 3 consecutive days with an absolute neutrophil count equal to or higher than 500/μl. Platelet engraftment was defined as the first of 3 consecutive days with a platelet count equal to or higher than 20000/μl with at least 5 days away from the last platelet transfusion. Donor hematopoiesis was documented by PCR study of genetic polymorphism of short tandem repeats. Results Enrolment characteristics Over a 14-year period, 50 requests were received for collection; 48 were successfully performed, while two were not due to damaged placenta at delivery. Diagnoses included acute lymphoblastic leukemia (42), acute myeloblastic leukemia (4), non-Hodgkin lymphoma (1), and chronic myeloid leukemia (1). All 48 cord blood units collected were banked. All siblings were at the time of collection in complete remission. In 5 cases, the mother was pregnant at time of diagnosis, while in the remaining cases the sibling was born at a median time of 2 years (range 1–10) after diagnosis. In one case preimplantation genetic diagnosis for HLA-typing was used to conceive a “saviour sibling” for the sibling with chronic myeloid leukemia (CML). The requests for banking increased the last 4 years since the establishment of a national cord blood bank. Characteristics of cord blood collections Table 1 gives an overview of the collected cord blood characteristics. The volume was ≥50 ml and total nucleated cells ≥3 × 108 in all units. The mean cord blood unit volume without anticoagulant was 82 ml ± 29.1; the mean total nucleated cell count was 7.08 × 108 ± 2.98; and the mean CD34+ cell count was 2.38 × 106 ± 1.36. All banked cord blood units were free of bacterial contamination. HLA-A, HLA-B, and HLA-DR typing was performed serologically. Of the 48 cord blood collections 10 were HLA-identical to the siblings, while two additional units had 2 HLA-antigen mismatches.
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181 Table 1 Laboratory characteristics of banked sibling cord blood units.
Volume minus anticoagulant, ml Nucleated cell count × 108 Total CD34+ cells/collection × 106 Viability %
Mean
SD
Range
82 7.08 2.38 91.6
29.1 2.98. 1.36 7.41
(50–146) (3–22.4) (0.5–6.3) (73.7–100)
Factors influencing the usage rate We found three factors to be associated with the usage of a cord blood unit in transplantation (Fig. 1). These included HLA histocompatibility, disease in long-term remission without need for transplantation, and collection failure. Among enrolled families 4 children needed transplantation; however, only one was used in transplantation in conjunction with a small volume of bone marrow from the same HLA-identical donor. With a median follow-up of 6 years (range, 2–12) for 9 remaining HLA-matched cord blood units, none from the prospective recipients needed transplantation. All of them remain in complete remission for a time longer than 5 years. In contrast, among children with HLA-mismatched cord blood, two needed transplantation and received grafts from matched unrelated donors. One of the 2 failed cord blood collections was the case where the donor was born to be HLA-identical following preimplantation HLA-genotyping. Bone marrow cells from the donor at the age of 18 months were transplanted to the affected (CML) sibling resulting in successful hematopoietic reconstitution. Transplantation characteristics and outcome The cord blood net volume of 85 ml was harvested and transported immediately to processing laboratory. The total nucleated and CD34+cells were 4×108 and 2×106 corresponding to 2×107/kg and 0.75×105/kg, respectively. Patient was at the time of cord blood collection in complete remission out of chemotherapy. Unfortunately, 4 months later he relapsed. Then he received 3 blocks systemic chemotherapy and entered into a second remission. At this point a stem cell transplantation was planned. Having considered the cord blood stem cells insufficient to
179
secure a stable engraftment, we decided to enrich the cord blood unit with bone marrow collected from the 8-month-old sibling weighing 10 kg. Harvesting of 120 ml of bone marrow was performed from the infant providing an additional total number of nucleated cells 2.68×108/kg with CD34 2.5×106/kg. Collection was uneventful without significant decrease in Hb levels. Post-transplantation course was uneventful with minimal transfusion requirements, low-grade mucositis, and only one minor febrile episode during aplasia. The patient demonstrated engraftment with neutrophils above 500/μl on day 18 and platelets above 20000/μl on day 22. Full sustained stable donor-derived hematopoietic chimerism was repeatedly observed by short tandem repeats analysis. At the present time the patient is 12 years post-transplant and remains in complete remission with full hematopoietic reconstitution. Discussion Many studies have shown that hematopoietic stem cells derived from related or unrelated cord blood units are at least as effective as those derived from the bone marrow or growth factor-mobilized peripheral blood in allogeneic transplantation setting for hematologic malignancies [6,11,12]. Specifically for sibling cord blood transplantation the latest analysis of Eurocord-European Bone Marrow Transplantation registry showed that the probability of disease free survival and overall survival, for 147 patients receiving a related cord blood transplant with a median follow-up of 6.7 years, was 44% and 55%, respectively [13]. Most of these cord blood units were banked at the local institution where transplantation took place, in collaboration between highly motivated families, obstetricians and hematopoietic cell transplantation physicians. Our stem cell transplant unit is the only center offering an allogeneic hematopoietic stem cell transplantation program for pediatric patients in Greece. We initiated the service of directed sibling cord blood banking until 2004, where we were succeeded by the first public cord blood bank. This service included the contact with the obstetric delivery unit to arrange collection, processing, testing, storage, and transplantation. Over a 14year period we collected 48 cord blood units from respective families with pediatric hematological malignancies while in 2 additional families the collection failed due to placental damage. Interestingly, in 45 cases out 50, the donor was conceived after diagnosis of the
Fig. 1. Transplant needs in 50 siblings with hematological malignancy and utility of banked sibling cord blood units.
180
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181
sibling's disease and in one case pre-implantation HLA-typing was performed successfully in order to create a “saviour sibling.” The practice of conceiving an additional child with the intention of using it as a hematopoietic stem cell donor for an affected sibling has been done for many years. In the early years, parents planned to have another child by natural conception and hoped for an HLA-matched sibling. These attempts are now being conjoined with reproductive technologies such as in vitro fertilization combined with preimplantation genetic diagnosis to determine HLA tissue matching [14,15]. Both volume and cell counts (nucleated and CD34+ cells) of cord blood unit we collected are comparable to those reported by cord blood banks in several countries involved in collection of cord blood from siblings of patients with malignancies [9,16,17]. Our policy is to store every unit ≥20 ml since even low volume collections could be usable in combination with reduced bone marrow harvest from the same donor, thus resulting in sufficient grafts. In such cases, one may argue that it is quite possible to delay the transplant until the donor is old enough to donate an adequate number of bone marrow stem cells, thus making cord blood storage unnecessary. However, in cases with hematological malignancies, an urgent transplantation may be indicated at an early time point where the donor is not old enough to donate a sufficient number of bone marrow stem cells. In such cases, reduced volume bone marrow collection not exceeding 15 ml/kg can be an attractive option for enhancing the stem cell number in the graft without placing the donor at the same risk as a standard bone marrow collection. Indeed, such an approach has proven valuable in our patient who needed an urgent transplantation and the donor age (8 months) was a limiting factor for donating sufficient bone marrow stem cells. Regarding the usage rate of cord blood units in stem cell transplantation, our results are in line with the experience reported by cord blood banks in the United States and the United Kingdom, confirming the low utility rate even for HLAmatched cord blood units. Reeds et al. [8] described USA national experience of sibling cord blood banking and transplantation for children with malignant disease. Two hundred fifty cord blood units were banked from families with a member suffering from a malignancy. From the stored cord blood units only 4 were used for transplantation. Two of the 4 donor recipient pairs were HLAidentical, one had donor mismatched at 2 HLA antigens and one had donor mismatch at 3 HLA antigens. The patient who received a cord blood unit mismatched at 2 HLA antigens had high-risk leukemia, failed engraftment and died 3.7 months post transplant. The other three patients were in complete remission. Smythe et al. have reported their 10-year experience with directed sibling cord blood banking in the National Blood Service in England and Wales [9]. They have used 3 cord blood units out of 114 collected from families who had a member with hematological malignancy. Two of their patients engrafted but died from relapse 4.5 and 2.5 years post transplant. The remaining one, who received a combined graft of cord blood and bone marrow, is alive free of disease 4.5 years post transplantation. Although neither the English bank nor the American bank reported the number of siblings needing transplantation among the enrolled siblings with hematological malignancies, we assume that HLAdisparity and the high possibility of cure with primary chemotherapy for pediatric hematological malignancies were the two main limiting factors for carrying out cord blood transplantation in these series. In the last years two additional smaller banks reported directed related cord blood banking programs without, however, mentioning the usage rate in transplantation [17,18]. In our experience, among 50 enrolled families 4 children needed transplantation; only one was successfully transplanted, however, in conjunction with a small volume of bone marrow from the same HLAidentical donor. Two additional children received grafts from matched unrelated donors because their sibling cord blood cells were HLAhaploidentical and the last one, despite pre-implantation selection to
be HLA-identical to the affected with CML sibling, cord blood could not be collected due to damaged cord/placenta at delivery. Based on these results, we recommend that both HLA-mismatched and matched sibling cord blood units should be stored independently of cell count and volume. Although it is not clear whether sibling cord blood donors disclosing 1 and 2 HLA-antigen mismatches can be proven effective, encouraging results from unrelated cord blood donors with similar HLA-disparity and the fact of the donor availability make the storage of such cord blood units reasonable. Even haploidentical cord blood may be used in emergency cases [19]. However, cord blood units that do not share at least one HLAhaplotype with the respective recipient should not be further retained. Continued storage should, however, be reviewed at least annually to ensure that cord blood are not stored unnecessarily; in cases where the donor is older than 3 years and recipient in long term complete remission (N5 years) the transplant physician in cooperation with the referring physician may decide to interrupt the long term storage because a such cord blood unit is unlikely to be used. Contrary to sibling cord blood banking for non-malignant diseases, where the usage rate was very high [9,16,20], utilization of sibling cord blood in children with a malignant disease is quite limited. The usage rates of units collected in directed banking programs reported by us and others are within the range of 1.6–2.6%. This low usage rate reflects the fact that most of the children with malignancy needing transplantation will be HLA-mismatched to respective cord blood unit and most of those with an HLA-compatible unit will never be in need for transplantation. In only few children who developed recurrent disease, cord blood transplantation was a considerable treatment option. Despite the fact of limited utility of directed sibling cord blood banking in transplantation for hematological malignancies, we still believe that the cost of storing such cord blood units in a public cord blood bank is reasonable and by limiting the long-term storage of those with extremely low probability to be used, a cost-effective national service for directed cord blood banking for pediatric patients with hematological malignancies could be established.
References [1] E. Gluckman, H.A. Broxmeyer, A.D. Auerbach, H.S. Friedman, G.W. Douglas, A. Devergie, et al., Hematopoietic reconstitution in a patient with Fanconi's anemia by means of umbilical-cord blood from an HLA-identical sibling, N Engl J. Med. 321 (1989) 1174–1178. [2] J. Kurtzberg, M. Graham, J. Casey, J. Olson, C.E. Stevens, P. Rubinstein, The use of umbilical cord blood in mismatched related and unrelated hemopoietic stem cell transplantation, Blood Cells 20 (1994) 275–283. [3] V. Rocha, E. Gluckman, Eurocord-Netcord registry and European Blood and Marrow Transplant group. Improving outcomes of cord blood transplantation: HLA matching, cell dose and other graft- and transplantation-related factors, Br. J. Haematol. 147 (2009) 262–274. [4] V. Rocha, N. Kabbara, I. Ionescu, A. Ruggeri, D. Purtill, E. Gluckman, Pediatric related and unrelated cord blood transplantation for malignant diseases, Bone Marrow Transplant. 44 (2009) 653–659. [5] J.E. Wagner, N.A. Kernan, M. Steinbuch, H.E. Broxmeyer, E. Gluckman, Allogeneic sibling umbilical cord-blood transplantation in children with malignant and nonmalignant disease, Lancet 346 (1995) 214–219. [6] V. Rocha, J.E. Jr Wagner, K.A. Sobocinski, J.P. Klein, M.J. Zhang, M.M. Horowitz, et al., Graft-versus-host disease in children who have received a cord-blood or bone marrow transplant from an HLA-identical sibling. Eurocord and International Bone Marrow Transplant Registry Working Committee on Alternative Donor and Stem Cell Sources, N Engl J. Med. 342 (2000) 1846–1854. [7] C. Navarrete, M. Contreras, Cord blood banking: a historical perspective, Br. J. Haematol. 147 (2009) 236–245. [8] W. Reed, R. Smith, F. Dekovic, J.Y. Lee, J.D. Saba, E. Trachtenberg, et al., Comprehensive banking of sibling donor cord blood for children with malignant and nonmalignant disease, Blood 101 (2003) 351–357. [9] J. Smythe, S. Armitage, D. McDonald, D. Pampphilon, M. Cuttridge, J. Brown, et al., Directed sibling cord blood banking for transplantation: the 10-year experience in the national blood service in England, Stem Cells 25 (2007) 2087–2093. [10] P. Rubinstein, L. Dobrila, R.E. Rosenfield, J.W. Adamson, G. Migliaccio, A.R. Migliaccio, et al., Processing and cryopreservation of placental/umbilical cord blood for unrelated bone marrow reconstitution, Proc. Natl Acad. Sci. USA 92 (1995) 10119–10122. [11] J. Kurtzberg, V.K. Prasad, S.L. Carter, J.E. Wagner, L.A. Baxter-Lowe, D. Wall, et al., Results of the Cord Blood Transplantation Study (COBLT): clinical outcomes of
E. Goussetis et al. / Blood Cells, Molecules, and Diseases 46 (2011) 177–181
[12]
[13]
[14] [15]
unrelated donor umbilical cord blood transplantation in pediatric patients with hematologic malignancies, Blood 112 (2008) 4318–4327. M. Eapen, P. Rubistein, M.J. Zhang, C. Stevens, J. Kurtzberg, A. Scaradavou, et al., Outcomes of transplantation of unrelated donor umbilical cord blood and bone marrow in children with acute leukaemia: a comparison study, Lancet 369 (2007) 1947–1954. A.L. Herr, N. Kabbara, C.M. Bonfim, P. Teira, F. Locatelli, K. Tiedemann, et al., Longterm follow-up and factors influencing outcomes after related HLA-identical cord blood transplantation for patients with malignancies: an analysis on behalf of Eurocord-EBMT, Blood 116 (2010) 1849–1856. Y. Verlinsky, S. Rechitsky, T. Sharapova, R. Morris, M. Taranissi, A. Kuliev, Preimplantation HLA testing, JAMA 291 (2004) 2079–2085. E.R. Morgan, J. Girod, J.S. Rinehart, Having a child to save a sibling: reassessing risks and benefits of creating stem cell donors, Pediatr. Blood Cancer 48 (2007) 249–253.
181
[16] M.C. Walters, L. Quirolo, E.T. Trachtenberg, S. Edwards, L. Hale, J. Lee, et al., Sibling donor cord blood transplantation for thalassemia major: experience of the sibling donor cord blood program, Ann. NY Acad. Sci. 1054 (2005) 206–213. [17] T. Brune, H.S.P. Garritsen, Potential of directed cord blood donation for the treatment of patients with hematologic disorders, Transfus. Med. Hemother. 34 (2007) 100–103. [18] M. Screnci, E. Murgi, D. Carmini, L. Piro, N. Cinelli, L. Laurenti, et al., Related cord blood banking for haematopoietic stem cell transplantation, Transfus. Med. 20 (2010) 185–190. [19] H.M. Katzenstein, E.R. Morgan, M. Olsewski, K. Danner-Koptik, M. Kletzel, Haploidentical related umbilical cord blood stem cell transplant in a child with acute non-lymphocytic leukemia, Bone Marrow Transplant. 19 (1997) 765–769. [20] E. Goussetis, E. Petrakou, M. Theodosaki, V. Kitra, I. Peristeri, G. Vessalas, et al., Directed sibling donor cord blood banking for children with β-thalassemia major in Greece: usage rate and outcome of transplantation for HLA-matched units, Blood Cells Mol. Dis. 44 (2010) 107–110.