Open access peer-reviewed chapter

Hematopoietic Stem Cell Source and Storage

By Sinem Civriz Bozdag and Osman İlhan

Submitted: October 14th 2014Reviewed: June 8th 2015Published: December 2nd 2015

DOI: 10.5772/60994

Downloaded: 1034


Hematopoietic stem cell transplantation(HSCT), has been accepted as a feasible treatment option that prolongs survival in hematological malignancies. Stem cell choice during hematopoietic stem cell transplantation can differ according to the experience of physicians, mostly treated hematological diseases in the centers or ongoing clinical trials. In this chapter we will discuss the advantages and disadvantages of three stem cell sources peripheral blood, bone marrow and umbilical cord blood.


  • hematopoietic stem cell

1. Introduction

Hematopoietic stem cell transplantation (HSCT) has been accepted as a feasible treatment option that prolongs survival in hematological malignancies. Bone marrow (BM) has been widely used as stem cell source in the early stem cell transplantation series. Tendency towards peripheral blood (PB) as stem cell source has been started in the beginning of 2000s. Initially, advantages of peripheral blood stem cell collection have been demonstrated in autologous stem cell transplantation (autoSCT). The introduction of peripheral blood into allogeneic stem cell transplantation (alloSCT) has been followed by allogeneic transplantation from unrelated donors. Due to the less stringent HLA matching requirement, cord blood stands out as an option for patients who do not have HLA-matched donors.

Stem cell choice can change according to the experience of physicians, mostly treated hematological diseases in the centers or ongoing clinical trials. In this chapter, we will discuss the advantages and disadvantages of three stem cell sources: peripheral blood, bone marrow, and umbilical cord blood.

2. Peripheral blood versus bone marrow

Although the early transplantation series mostly used bone marrow as stem cell source, the administration of granulocyte colony-stimulating factor allowed physicians to collect stem cells from peripheral blood by apheresis procedure. Studies in autologous stem cell transplantation (autoSCT) have been reported first. Faucher et al. [1,2] compared peripheral blood and bone marrow in a series with small patient numbers and reported an improvement in hematological recovery and a decrease in hospitalization duration. Survival of the patients was found to be similar between these two stem cell sources [3]. Also, the cost-effectiveness of the procedure increased peripheral blood stem cell transplantation rates compared with bone marrow transplantation. In a study, the total cost in peripheral blood stem cell transplantation was found to be 27.5% less than bone marrow transplantation. Faster hematological recovery, fewer hospitalization days, and less antibiotic treatment created this difference in total cost [4].

Afterwards, the comparison of peripheral blood with bone marrow in allogeneic stem cell transplantation (alloSCT) setting was started. Bensinger et al. [5] reported faster hematological recovery, less transfusion numbers, less severe acute graft versus host disease (a gvhd in favor of PB), but similar chronic graft versus host disease (cGVHD) rates. Our centers’ experience also revealed faster engraftment, fewer transfusions with PB [6]. Miflin et al. [7] showed that increasing the infused CD34+ cell number over 4× 10e6/kg can significantly accelerate the engraftment kinetics.

In a multicenter randomized trial performed by the European Group for Blood and Marrow Transplantation (EBMT), transplant-related mortality and leukemia-free survival rates showed no significant difference between PB and BM [8]. Mielcarek et al. [9] reported a better 10 years disease-free survival in favor of PB but similar overall survival between two stem cell sources. In a Cochrane database review, trials including related stem cell donors were analyzed. Both neutrophil and platelet recoveries were faster in PB. Disease-free survival, nonrelapse mortality was not different between PB and BM. The advantage of relapse in PB was recorded [10].

Peripheral blood has more CD34+ cells but has also more T cells in comparison with BM. This reflects to the outcomes as more acute and chronic GVHD rates. In a study, acute grades I–IV GVHD incidences for infused CD34+ cell doses less than 2× 10e6/kg, between 2× 10e6/kg and 4× 10e6/kg, and more than 4× 10e6/kg were 21%, 35%, and 43%, respectively. In the same study, increase in CD3+ cell dose was also identified as an independent factor for acute GVHD [11]. The correlation between infused cell dose and chronic GVHD has also been shown. Zaucha et al. reported CD34+ cell doses over 8× 10e6/kg were found to be associated with increased clinical extensive chronic GVHD. However, this association could not be shown with stem cell transplantation from bone marrow [12]. cGVHD after PB transplantation has to be treated for a longer period with higher immunosuppressive regimens. The presence of cGVHD has been found to be related with fewer relapses but more treatment-related mortality [1316].

The comparison of stem cell sources has been performed according to the diagnosis of hematological malignancies. Pidala et al. [17] performed a Markov model in which PB was found to be the optimum stem cell source for hematological malignancies, which had an advantage of 7 months in comparison with BM. BM was chosen to be superior in patients with 1-year relapse rates lower than 0.05. Patients with high-risk hematological malignancies like acute leukemia in second or later remission, chronic myeloid leukemia (CML) in blastic transformation, refractory anemia with excess blast in transformation, and heavily pretreated lymphoma patients were found to benefit from PB transplants in comparison with BM transplants [16]. A meta-analysis of nine randomized clinical trials showed better disease-free and overall survival with PB in late stage disease [18]. In a randomized study of patients with myeloid malignancies, acute myeloid leukemia AML,CML and myelodysplastic syndrome (MDS), hematological recovery was faster in PB group. Improvement in overall survival with PB was found to be related with reduction in nonrelapse deaths, with no difference in early, late relapses, or deaths in relapses [19]. In another study, which has included chronic myeloid leukemia patients, the incidence of acute and extensive chronic GVHD was similar between BM and PB patients, and there was no significant difference between survival and nonrelapse mortality rates. However, in the subgroup analysis of chronic phase patients, PB patients experienced more chronic GVHD, and BM patients had higher relapse rates [20]. The advantage of BM remains in benign hematological diseases, where chronic GVHD rates affect transplant outcomes negatively. In a study that included 537 adolescent aplastic anemia patients who had alloSCT, the survival advantage for BM recipients was reported to be significant [21]. Bacigalupo et al. [22] showed the advantage of BM also for the patients older than 50 years. The major causes of death were GVHD, infections, and graft rejection in this study. GCSF-stimulated and GCSF-unstimulated bone marrow and peripheral blood have been compared in another study that included aplastic anemia patients. Engraftment rates were not different in three treatment arms. Grades II–IV acute GVHD and chronic GVHD rates were higher with PB transplants in comparison with BM. GCSF-manipulated BM was not superior to BM in terms of mortality rates [23].

Late effects of transplantation were also analyzed in studies. Performance status, return to work, incidence of bronchiolitis obliterans, and hematopoietic functions were found to be similar between PB and BM. Also, there was no significant difference in secondary malignancies between the groups [24].

As HLA-matched siblings can be found in 25% of the patients, either PB or BM of the unrelated donors has been accepted as sources for stem cell transplantation. Eapen et al [25] reported significantly higher acute and chronic GVHD risk but similar transplant-related mortality and survival rates with PB than BM in unrelated allogeneic stem cell transplants. In a phase 3, multicenter randomized trial from 46 transplant centers in the United States and Canada, only chronic GVHD rates were significantly higher in PB transplantations. In long-term outcome analysis, chronic GVHD was graded as extensive in 85% of PB recipients compared with 76% of BM recipients. Neutrophile and platelet engraftment rates were found to be significantly higher in PB transplantations, but no difference was observed in survival outcomes [26]. In a Cochrane database review, trials including unrelated stem cell donors were also revised. Both neutrophile and platelet recovery was faster in PB transplantations. Disease-free survival, nonrelapse mortality of PB, and BM recipients were not different. The relapse advantage of PB was not proved. Acute grades II–IV incidence did not reveal a statistical significance, whereas extensive chronic GVHD rates were in favor of BM [10].

The outcomes of unrelated allogeneic stem cell transplantation were also separately assessed according to the diagnosis of hematological diseases. During an overall survival of 7 years, nonrelapse mortality and relapse rates were similar in acute myeloid leukemia and myelodysplastic syndrome patients who have received PB or BM. The only significant difference was reported in chronic myeloid leukemia patients due to higher nonrelapse mortality rates in PB transplants [27]. In a Korean study, the difference in GVHD incidence could be overcomed by risk adapted GVHD prophylaxis in AML patients [28]. In aplastic anemia patients, the risk of death was higher with unrelated donors, but peripheral blood as a stem cell source was a negative predictor for outcomes [29].

The impact of conditioning regimens in unrelated allogeneic stem cell transplantation has also been assessed in clinical trials. EBMT data showed that when reduced intensity conditioning (RIC) regimens were used in unrelated transplant settings, acute and chronic GVHD rates were higher, and relapse rates were lower in PB transplants of AML patients [30].In contradictory with these results, GVHD, relapse, and survival rates of hematological malignancies who had unrelated stem cell transplantation were reported to be similar in a recent trial [31].

The issues that have to be paid importance during hematopoietic stem cell transplantation should be the benefits of either the donor or the patient.

Both the peripheral blood and the bone marrow have advantages and disadvantages for the donors. Peripheral blood donation is a collection of HSC from peripheral blood after a 5-day course of granulocyte colony-stimulating factor administration on 1–2 days via 4–5 hours of apheresis procedure. The monitorization of circulating CD34+ cells on the first day of apheresis is predictive for stem cell yield [32].

Peripheral blood stem cell collection seems to be an easier collection method, but the central venous line can be a necessity for some of the donors if the standard peripheral venous line is not adequate. Again in NMDP experience, femoral and jugular lines were in same frequency, which was twice as the subclavian line [33]. The collection of peripheral blood stem cells via apheresis procedure has been proven to be an effective and safe method for donors [34]. Bone marrow harvesting is the collection of HSC from posterior iliac crest under anesthesia. Hospital admission can be necessary for postoperative follow-up. In a prospective study, the experience of donors for bone marrow and peripheral blood collection has been compared. Bone marrow donors were found to be less confused and more prepared for donation. They also found the process psychologically beneficial in a short term. However, the long-term health-related quality was similar between both donors [35]. In a survey of 51,024 AHSCT performed by 338 teams, five donor fatalities were observed. Severe adverse events were reported in 37 donors. Hematological malignancy rates were not different from the age and sex-adjusted general population [36]. Bone pain was the most frequent side effect of donors; in bone marrow donors, pain at the collection site and in peripheral blood donors pain at various sites of body during GCSF administration have been reported. Tiredness, light-headedness, nausea, sleeping problems, and chills were the other frequent side effects. At 12 months postdonation, the most common side effects were tiredness and muscle aches [35]. A median time to recovery in PB and BM donors were 1 and 2 weeks, respectively [37]. In a prospective trial from NMDP, females and heavier donors had more III–IV CALGB adverse events [38].

3. Umbilical cord blood

Umbilical cord remains to be a stem cell source option for patients who do not have a matched sibling or unrelated donor. The first umbilical cord blood transplantation has been performed at the end of the 1980s in a child diagnosed with Fanconi anemia [39]. Today, the storage of more than 600,000 cryopreserved cord bloods serves as an alternative option for both pediatric and adult patients. In the U.S. registry, almost all of the patients younger than 20 years and 80% of the patients older than 20 years had cord blood mismatched units in one or two HLA locus [40].

Rapid availability, no requirement for full HLA match, and being an option for ethnic minorities became the advantages of umbilical cord. Also, no harm for donor can be another important feature in stem cell source choice. Lack of sufficient cell doses for adult patients, delayed engraftment, poor immune reconstitution, and high infection rates are the major disadvantages of cord blood transplantation. Cord blood includes more naive T cells and Tregs. The first naive T cells proliferate but show a limited repertoire. Then a thymic-dependent population expands, which can be affected by conditioning regimen, GVHD, or aging [41].

The standard practice for the HLA typing of cord blood units is to analyze A-B antigens and DRB1 alleles with high resolution. The acceptable HLA match is 4 to 6/6 match for performing transplantation, but each mismatch results in increased TRM, increased severity of acute GVHD, and decreased survival [42]. Eapen et al. [43] reported the impact of level matching on TRM and neutrophile engraftment. Neutrophile recovery was found to be delayed in transplantations with more than two mismatches, and nonrelapse mortality was reported to be higher in 1–5 mismatch compared with HLA full-match transplantations. HLA C matching has also been shown to be beneficial in studies [42,44]. Barker et al. [45] analyzed 1691 MDS and AML patients who received single cord blood transplantation and found that regardless of cell dose, HLA A, B, and DR-matched transplantations result in the best outcomes. Total nucleated cell (TNC) doses greater than 2.5× 10e7/kg in one mismatch recipient and 5× 10e7/kg in two mismatch recipients have been shown to be sufficient for better survival. In another study, high-resolution DRB1 match was related with less acute GVHD and better event-free survival rates. Also, in the same study, infusing higher CD34+ cells, CD34+HLA DR+ CD38+ cells, and CD3+ cells resulted in faster engraftment [46]. The impact of HLA mismatch direction between donor and recipient has also been assessed by EUROCORD/EBMT analysis. Neither one to two mismatch in graft versus host direction nor host versus graft direction was found to be related with increased nonrelapse mortality and survival [47].

Graft failure risk remains to be a problem in 10%–20% of patients who had cord blood transplantation [48]. Neutrophile and platelet engraftment is delayed in cord blood transplantations. Engraftment is shown to correlate with cell dose infused [49]. In an analysis of 1268 patients with acute leukemia, 3 years overall survival was 47% and TRM was 16%. Delay in engraftment was associated with increased mortality and shorter survival rates [50].

Survival outcomes have been improved over years with better patient, conditioning regimen selection, and progress in HLA typing [44]. Rubinstein et al. [51] showed that 46% of the patients experienced transplant-related events by posttransplant day 100. Transplant-related events and event-free survival were related with diagnosis, number of leukocytes in the transplant, age, extent of HLA disparity, and transplant center. Cohen et al. [52] reported the outcome of 500 patients who had single-unit cord blood transplantation from 1995 to 2005 and found 1 year survival as 37%. Factors affecting early mortality following the myeloablative single-unit cord blood transplantation were cell dose, advanced disease, older age, cytomegalovirus status, female gender, and limited cord blood center experience. A Japanese study revealed that disease status and cytogenetics had an impact on event-free survival rates in AML patients [53]. For acute lymphoblastic leukemia patients who had cord blood transplantation, the factors associated with better leukemia-free survival were age, advanced disease, and conditioning regimen [54]. Brunstein et al. [55] reported 3 years event-free and overall survival as 38% and 45%, respectively, by performing transplantation with nonmyeloablative conditioning. In a comparative study, on the effect of conditioning regimen intensity, transplant-related mortality was similar with both regimens; lower risk of relapse and longer leukemia-free survival could be achieved after myeloablative regimens [56]. Fludarabine in combination with total body irradiation (TBI) concluded with high treatment-related mortality [57]. In another trial, fludarabine, TBI in combination with busulphan versus cyclophosphamide has been investigated; cyclophosphamide resulted in better transplantation outcome [58]. Also, the Japanese group used a myeloablative conditioning regimen, which included TBI, cytarabine, and cyclophosphamide, and reported 51% overall survival for high-risk hematological malignancies [59].

Graft versus host disease (GVHD) is one of the major early and late morbidity and mortality causes. It has been proven that donor source has an impact on GVHD rates. Although the HLA disparities were higher in cord blood transplantations, GVHD was found to be lower than bone marrow [6062]. Eurocord and EBMT revealed that HLA mismatch increased the acute GVHD risk in cord blood transplantation [63]. Patient age [51], CMV status [63], nonmyeloablative conditioning, and absence of ATG were also factors thought to be related with acute GVHD. Acute GVHD was reported to be higher in double-unit cord transplants than the single-unit transplants [60]. In a study with 1072 patients, chronic GVHD in posttransplant 2 years was 28%. In multivariate analysis, risk factors were identified as myeloablative conditioning regimen, mycophenolate mofetil in GVHD prophylaxis, increased HLA mismatch, higher body weight, and previous acute GVHD [64]. Newell et al. [65] found higher chronic GVHD rates, which were analyzed according to NIH 2005 criteria and had a predominance of acute GVHD features.

Comparative trials of umbilical cord blood with different stem cell sources have been reported. Neutrophile and platelet recovery have been delayed in cord blood transplantations, and chronic GVHD rates have been observed less. Laughlin et al. [66] included patients who had received either an HLA-matched marrow transplant- a marrow transplant with a single HLA mismatch from an unrelated donor or who had received a cord blood transplant with one or two HLA mismatches in their study. Overall mortality was found to be lower in matched related BM recipients. The rate of leukemia recurrence was found to be similar, but 3 years survival for cord blood recipients was 26%, which was lower than matched BM recipients’ survival. In a recent trial, donor types have been investigated in AML patients; survival rates were similar between matched related, matched unrelated, mismatched unrelated, and cord blood transplant. Age and type of conditioning regimens were the major determinants of survival [67].

As a consequence of inadequate engraftment with single-unit cord blood transplantation in adult hematological malignancy patients, double-unit cord transplantation was introduced in the beginning of 2000s. Sustained engraftment could be achieved more than 90% of patients and one of the double-units dominate [68,69]. Barker et al. [70] observed the impact of CD3+ cell dose on engraftment, but not the CD34+ cell dose or HLA match in double cord unit transplantations.

Avery et al. [68] found an association between higher numbers of CD34+ and TNC cell dose of the dominant unit and sustained engraftment. In the same trial, unit–unit HLA match and unit–recipient HLA match were not associated with sustained engraftment. Mixed chimerism can be displayed in follow-up, especially if the HLA match of the both units is close [71]. The HOVON group showed that the unit predominance was observed by posttransplant day 11, and the role of CD4+ lymphocyte-mediated alloreactivity was suggested [72]. In another study, the cord blood bankreported pre cryopreservation and postthaw viable CD34+ cell doses were the important parameters for the engraftment [73].

Macmillan et al. [60] reported double cord transplantation as one of the risk factors for acute GVHD. Ponce et al. [74] demonstrated the acute grades II–IV rates as 53% and predominantly the gut as affected organ. Chronic GVHD rates have been reported around 30% in double cord blood transplantation trials [75,76].

Eurocord and EBMT compared single-unit transplantation with different myeloablative conditioning regimens and double-unit transplantation. Conditioning regimens were either the widely accepted Minnesota protocol, which consisted of TBI/cyclophosphamide/ fludarabine or thiotepa/busulphan/fludarabine (TBF). In this study, 2 years LFS was similar between double-unit cord and single-unit cord if the new TBF protocol was chosen [75,77]. In a prospective multicenter trial, which included 56 acute leukemia and myelodysplasia patients transplanted with double-unit cord, 3 years disease-free survival was 50% and TRM was 39% [76].

Comparative studies of different stem cell sources in the era of double-unit cord have been continued to be reported. In comparison with filgrastim mobilized peripheral blood, delay in immune reconstitution of T cells in cord blood transplantation resulted in increased infection risks. The double-unit cord blood has been compared with unrelated donor grafts; although 3 years survival analysis were similar, double cord blood was associated with less chronic GVHD but more nonrelapse mortality [78]. Brunstein et al. [79] have found comparable 5-year leukemia-free survival after HLA-matched related, unrelated, and double-unit cord blood transplantation.

4. Hematopoietic progenitor cell and umbilical cord blood cryopreservation

Collected hematopoietic stem cells have been cryopreserved by using a cryoprotectant, dimethyl sulfoxide (DMSO), and frozen in liquid nitrogen vapor until reinfusion. DMSO penetrates to cells and binds to water molecules in order to prevent dehydration of the cells [80]. During the cryopreservation process, DMSO dilution, freezing period, storage in vapor phase, or liquid nitrogen are the important factors for optimum results. DMSO and cell suspension should be cooled down to 0°C–4°C, and after the addition of DMSO, the product should be placed in a controlled freezer subsequently. The optimum concentration of DMSO has also been analyzed in different studies because it can be toxic for stem cell viability and also may cause side effects during infusion. Although majority of transplant centers still prefer to use 10% DMSO, lower percentage of DMSO or washing the product before infusion to decrease toxicity has also been used [81]. Reducing DMSO concentration into 7.5% has been revealed as feasible [82]. Up to 10 years cryopreservation with 5% DMSO has also been found not to have a negative impact on cell viability [83]. DMSO concentration has been calculated as ml/kg body weight or ml/min, and in EBMT results from 65 centers, it has been revealed that calculation as milliliters per minute should be the preferred way to reduce side effects [84]. DMSO toxicity can result in nausea, vomiting, fever, or more severe reactions like hepatic dysfunction, cardiac arrhythmia, and neurotoxicity [85,86].

In the freezing period, cells have been mostly frozen using controlled rate freezers. Too fast cooling can result in intracellular crystallization, and too slow cooling can induce extracellular ice formation. Sputtek et al. [87] reported that the cooling rate range can vary from 1 to 5 K/min. Also, the recovery of white blood cell (WBC) recovery was found to be superior in slow rate freezing to fast rate freezing [88]. After the introduction of stem cell storage in liquid nitrogen, risks for microbial contamination of the products concluded with the usage of vapor phase for storage. However, the comparison of the two phase was found to be similar for either WBC recovery or WBC viability in comparative studies [88,89].

The major disadvantage of the umbilical cord blood processing is the potential risk for loss of progenitor cells in the collected product. The techniques that have been used for red cell separation like simple centrifugation lysis with ammonium chloride, filtration through density gradients, or collection from bags to vessels were found to have detrimental effects during cryopreservation [9093]. Thus, Rubinstein et al. [94] initially proved that forming 20 ml cord blood units with uniform volume can be achieved by using rouleaux formation induced by hydroxyethyl starch and centrifugation. Semiautomated top–bottom systems and lately automatic devices like AXP-SEPAX have been developed [95]. Automatic systems could achieve similar cell recovery with less technical influence [96].

After the addition of 10% dimethyl sulfoxide, cord blood samples can be cooled from 4°C to –80°C mostly by controlled rate freezers [97]. It has been recently shown that cord bloods can be transferred into liquid nitrogen vapor phase directly or after storage at –80°C for 18 hours [98]. It has been demonstrated by Broxmeyer et al. that the long-term storage does not have a negative influence on in vitro function of umbilical cord blood progenitor cells. Also, the duration of cryopreservation was found to have no impact on clinical outcome like neutrophile or platelet recovery after cord blood transplantation [99].

5. Conclusion

Hematopoietic stem cell source choice is an important issue to be concerned during stem cell transplantation. Diagnosis and pretransplantation status of hematological diseases and type of conditioning regimens are the major factors in making decision for one type of stem cell source.

Processing before the storage of collected stem cells can show differences according to the source of stem cells. In particular, cord blood processing needs more attention due to the risk of hematopoietic stem cell loss.

© 2015 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

How to cite and reference

Link to this chapter Copy to clipboard

Cite this chapter Copy to clipboard

Sinem Civriz Bozdag and Osman İlhan (December 2nd 2015). Hematopoietic Stem Cell Source and Storage, Progress in Stem Cell Transplantation, Taner Demirer, IntechOpen, DOI: 10.5772/60994. Available from:

chapter statistics

1034total chapter downloads

More statistics for editors and authors

Login to your personal dashboard for more detailed statistics on your publications.

Access personal reporting

Related Content

This Book

Next chapter

Progress in Cryopreservation of Stem Cells and Immune Cells for Cytotherapy

By Zhiquan Shu, Shelly Heimfeld, Zhongping Huang, Carolyn Liu and Dayong Gao

Related Book

First chapter

Generation of Patient Specific Stem Cells: A Human Model System

By Stina Simonsson, Cecilia Borestrom and Julia Asp

We are IntechOpen, the world's leading publisher of Open Access books. Built by scientists, for scientists. Our readership spans scientists, professors, researchers, librarians, and students, as well as business professionals. We share our knowledge and peer-reveiwed research papers with libraries, scientific and engineering societies, and also work with corporate R&D departments and government entities.

More About Us