Umbilical Cord Blood Transplantation: Basic Biology and Clinical Challenges to Immune Reconstitution

Abstract

Allogeneic stem cell transplantation has continued to evolve as a common procedure for the treatment of hematological malignancies and bone marrow failure. Donor bone marrow and mobilized peripheral stem cells are routinely employed for the reconstitution of immune function in leukemia and lymphoma patients following radiation and/or chemotherapy. Unfortunately, only 30% of patients have an HLA identical sibling donor and the identification of matched unrelated donors, particularly for minorities, can present an exceptional challenge. The transplantation of umbilical cord blood (UCB) represents the most recent strategy to expand the potential donor pool while maintaining an acceptable level of treatment related complications. First utilized in children, UCB transplantation permits a higher degree of HLA disparity while demonstrating a reduction in the incidence and severity of graft versus host disease (GvHD) compared to previous transplantation modalities. Despite the apparent decrease in GvHD, relapse rates remain comparable to transplantation with bone marrow or mobilized peripheral blood suggesting a strong graft versus leukemia/lymphoma (GvL) effect. However, several issues complicate the use of UCB transplantation and its extension to the treatment of adults. Many infections that afflict transplant patients are particularly frequent and more severe in the context of UCB transplantation. UCB T cells are naïve and therefore display less proliferation and IFN-γ production in response to cognate antigen and also appear to demonstrate defects in signal transduction mechanisms. In addition, UCB contain T regulatory cells (Treg) with more potent suppressor function than adult Treg. Furthermore, adult patients often require more total cells and CD34+ progenitors for transplantation than a single UCB unit can provide. Thus, strategies to expand selected subpopulations from UCB and the use of multi-unit transplantation are areas of active research. This review will provide a condensed summary of the clinical history of UCB transplantation and emphasize the advantages and disadvantages of this approach to hematological malignancies in comparison to other methods of hematopoietic stem cell transplantation. Subsequently, it will mainly focus on the current challenges to immune reconstitution presented by UCB transplantation, recent research into their cellular and molecular mechanisms, and experimental approaches to overcome them.

Introduction

The transplantation of hematopoietic stem cells (HSC) in the context of treatment for high-risk hematologic malignancies has evolved into a standard procedure. Although therapeutic regimens vary depending upon the type of malignancy, HSC are infused into cancer patients following an intense course of chemotherapy and radiation. Traditionally, HSC are harvested from the bone marrow (BM) or peripheral blood of an HLA-matched sibling or unrelated donor. However, only 30% of patients have a potential sibling donor who can meet the stringent requirement of a 6/6 or 5/6 match with the patient’s HLA loci (HLA-A, HLA-B, HLA-DRB1) [1]. In the absence of a matched sibling donor, patients must rely on the worldwide network of bone marrow registries to find an HLA-matched donor. Complicating this process is the fact that the majority of registered donors are Caucasian, thereby making the selection process extremely difficult for patients of different or mixed races. The search process for a suitable donor is often lengthy; a recent report cites an average time of 4 months [1]. In the interim, the progression of the patient’s malignancy and the toxicity of the chemotherapy required while a matched donor is sought can result in a worsening prognosis.

After HSC transplantation, the additional complications of susceptibility to infection and graft versus host disease (GvHD) provide substantial challenges to immune reconstitution. In particular, acute GvHD frequently develops in the context of allogeneic HSC transplantation. Mature allogeneic T cells that accompany the HSC graft are activated by MHC class I and II antigens expressed by the recipient, resulting in an aggressive T helper type I response that principally targets the skin and gut. Furthermore, chronic GvHD, which is related to acute GvHD but more closely resembles an autoimmune disorder with the development of autoreactive T cells, can result in debilitating and life-threatening disease many months post-transplantation. GvHD and opportunistic infections cause considerable morbidity in transplant patients and treatment protocols for the resolution of each will frequently exacerbate the other (i.e., steroid treatment for GvHD diminishes the immune response to infection).

In an attempt to better address these issues, considerable interest has focused on the use of umbilical cord blood (UCB) as an alternative source of HSC for hematopoietic reconstitution (Table I). Early studies by Knudtzon [2] demonstrated that granulocytic colony-forming cells could be grown in vitro from UCB. Further in vitro studies by Broxmeyer et al. [3] established that UCB contains a sufficient number of hematopoietic stem/progenitor cells to be used for autologous or allogeneic hematopoietic reconstitution. In 1989, Gluckman and colleagues published results from the first successful UCB transplant using cord blood from an HLA-identical sibling in a patient with Fanconii’s anemia [4]. Large-scale clinical trials to evaluate UCB transplantation for a variety of hematologic malignancies, bone marrow failure syndromes, and congenital errors of metabolism were soon established.

Table I.

Umbilical cord blood vs. bone marrow or mobilized stem cell transplantation (HSC)

	UCB	HSC
Advantages:
Availibility of donor grafts	Large Supply	Limited supply
Availibility of graft	Rapid	Prolonged
Optimal degree of match	4/6 or Higher	9/10 or 10/10
Risk of GvHD	Lower risk	Higher risk
Risk of viral transmission	Very low risk	Higher risk
Risk to donor	No risk	Higher risk
Disadvantages
Number of nucleated cells	Limited numbers	Higher numbers (1–2 logs)
Risk of infection	Higher than HSC	High
Speed of neutrophil/platelet recovery	Prolonged	Faster than UCB
Possibility of donor lymphocyte infusion	Not possible	Possible
Risk of genetic disease transmission	Higher than HSC	Very low

Transplantation with UCB : clinical studies

Since UCB units typically contain 1–2 logs fewer nucleated cells than a unit of bone marrow [5], pediatric patients were primarily chosen as the first recipients of UCB transplants (Table II). Rocha and colleagues from the Eurocord and International Bone Marrow Transplant Registry [6] compared the incidence of GvHD in children who received either cord blood (113 patients) or bone marrow (2052 patients) from an HLA-identical sibling for the treatment of either malignant or non-malignant disease. Using a multivariate analysis and adjustment for relevant risk factors, the authors found that the risk of acute or chronic GvHD was significantly lower in children who received UCB compared to those who received BM. However, in the first month post-transplantation, UCB recipients demonstrated slower recovery of neutrophil and platelet counts than the BM transplant patients. UCB transplant patients were also more likely to die of infection. Notably, the two groups had similar rates of death due to relapse [6].

Table II.

Umbilical cord blood transplantation in pediatric patients - selected clinical trials

	Gluckman et al. [7]	Rocha et al. [6]	Rubinstein et al. [8]	Wagner et al. [9]	Dalle et al. [10]	Rocha et al. [12]
Dates of clinical trial (yrs)	1988–1996	1990–1997	1992–1998	1994–1995	1996–2002	1994–1998
Patients enrolled	143	113	562	18	36**	99
Age in yrs (median)# (range)	0.2–45*	5 (<1–15)	< 2 - ≥18	2.7 (0.1–21.3)	7.5 (0.1–19.5)	6 (2.5–10)
Patients with hematologic malignancies	95	61	378	13	30	99
Patients without hematologic malignancies	48	52	184	5	6	0
Myeloablative (M) or nonmyeloablative (N)	M	M	M	M	M	M
Degree of HLA match (%)
matched (6/6)	69	100 (HLA-ID sibs)	7	7	6	8
1 or 2/6	62		86	10	94	84
3/6 or more	12		7	1		8
Median cell dose (range):
Nucleated cells/kg (×108)	0.37 (0.07–3.0)	0.47 (1.0–3.6)	range: 0.07 - >1.0	0.41 (0.14–4.0)	0.25 (na)	0.38 (0.24–3.6)
CD34+ cells/kg (×105)	2 (0–45)	na	na	na	na	na
Days to neutrophil recovery (median) (range)	30 (8–56)	26 (na)	28 (10–120)	24 (16–53)	28 (16–49)	32 (11–56)
Days to platelet recovery (median) (range)	56 (9–180)	44 (na)	90 (16–250)	54 (39–130)	43 (18–59)	81 (16–159)
No. given hematopoietic growth factors	104	45	na	7	23	54
Incidence of acute GvHD:	In 143 patients:	In 107 patients	In 399 patients:	In 13 patients	In 36 patients	In 99 patients
Grade I–II (%)	related=13; unrelated=12	29	78	61	3	35
Grade III–IV (%)	related=5; unrelated=20	2	22	11	11	21
Patients with cGvHD/surviving patients	related 8/56; unrelated 0/23	5/93	39/158	na	3/36	5/43
Incidence of relapse (%)	related=14; unrelated=8	na	26	11	17	38
Treatment related mortality (TRM) (%)	related=14; unrelated=45	31	39	28	27 at day 100	39 at day 100
Overall survival (2–3yrs) (%)	63 (related donor)	46 (malignant)	61% at 100 days	65% at 6 months	59	35
	73 (unrelated; 6/6 match)	86 (nonmalignant)

na = not available

^*84% of patients ≤ 15 yrs.

^#when indicated

^**data given is for UCB transplants only

In another study, case reports of related or unrelated UCB transplantation in children from 45 centers revealed the importance of cell dose as a predictor of neutrophil and platelet engraftment [7]. Patients transplanted with unrelated UCB containing less that 37 million nucleated cells/kg required a median of 34 days to reach an absolute neutrophil count of 500 per cubic millimeter and a median of 134 days to obtain a platelet count of 20,000 per cubic millimeter. In patients who received 37 million or more nucleated cells/kg, these numbers were reduced to 25 days and 47 days, respectively. HLA mismatching was also found to be related to the delay in engraftment times. The incidence of GvHD was low in this study, with the degree of HLA mismatch correlated with the incidence of GvHD in patients with related donors. Interestingly, the number of HLA mismatches did not affect the incidence of GvHD in patients given UCB from unrelated donors.

In 1998, Rubinstein and colleagues reported on 562 recipients of UCB transplants from unrelated donors at 98 transplantation centers [8]. Similar to Gluckman et al.[7], this study demonstrated that the number of leukocytes per kilogram in the UCB graft correlated with the time to myeloid engraftment. The kinetics of myeloid engraftment were also related to the degree of HLA mismatch. Conversely, the time to platelet engraftment in this study was not delayed but comparable to that seen in studies of BM transplantation. Event-free survival post-engraftment correlated significantly with the age of the patient and not to the number of leukocytes per kilogram in the graft. Significantly, acute and chronic GvHD in this study was less severe than in transplant patients that received BM from unrelated donors. This observation was seen even in patients that received UCB grafts with up to two HLA mismatched antigens. The authors concluded that UCB is a viable source of HSC for bone marrow reconstitution. Another study during this period commented on the rapid availability of UCB units for transplantation. Wagner et al. [9] found a median time of 3.5 months to receive BM from an unrelated donor, whereas the search for a suitable UCB unit required only 69 days. A more recent study [10] has also demonstrated that the time interval from decision to actual transplantation was 1 month shorter for UCB compared to BM.

In further studies involving UCB transplant in children, Barker et al. [11] and Rocha et al. [12] continued to see a delay in neutrophil engraftment but encouraging results regarding acute and chronic GvHD. In a retrospective multicenter study, Rocha et al. [12] also observed a delay in platelet engraftment and re-emphasized the finding of other groups [7, 8] that a nucleated cell dose above 0.37×10⁸/kg is required to increase the probability of engraftment. This study also demonstrated, similar to the results of Rubinstein et al. [8], no correlation between the outcome of UCB transplantation and the number of HLA mismatches between the UCB unit and recipient. In a single institute matched pair analysis comparing the outcomes of patients receiving HLA-mismatched UCB versus HLA-matched BM, Barker et al. [11] found that the probabilities of engraftment, GvHD and survival were actually comparable between the two groups. These studies suggested that UCB T cells demonstrate a lower level of alloreactivity than T cells in BM transplants. Although beneficial in terms of the reduced incidence of GvHD, this continued finding presented the concern that UCB transplants would mediate a poor graft versus tumor or graft versus leukemia (GvL) effect. However, Rocha et al. [12, 13] reported that after 100 days post-transplant, the risk of relapse was comparable in patients receiving UCB versus BM transplant. This finding has been confirmed by several other groups [11, 14, 15].

Following these favorable comparisons with BM transplantation in pediatric patients, many clinical studies began to focus on the adoption of UCB transplantation for the treatment of adult malignancies (Table III). A major concern regarding adult patients is the restricted number of HSC in the UCB graft. In a study involving both children and adults, Wagner et al. [16] followed 102 unrelated donor UCB transplant patients to assess the effect of CD34⁺ cell dose and HLA mismatch on treatment-related mortality (TRM) and survival. They concluded that the number of CD34⁺ cells infused per kilogram was significantly related to the rate of engraftment, TRM, and overall survival. In this study, 29% of patients that received 1.7 to 2.7×10⁵ CD34⁺/kg experienced TRM while 68% of those that received less than 1.7×10⁵ CD34⁺/kg had TRM. Furthermore, for each degree of HLA disparity, the authors suggested that a critical number of CD34⁺ cells/kg must be infused; below this threshold, the probability of survival is significantly decreased. This finding was found to be particularly applicable to recipients of UCB units with 2 HLA mismatches. As a consequence of these results, the authors concluded that UCB graft selection should be based on the CD34⁺ cell dose when several UCB units are available with an HLA mismatch of 2 or less. In another multi-center study of adult transplant patients, UCB grafts with a relatively high number of nucleated cells were associated with a faster rate of neutrophil recovery [15]. Additionally, infusion of UCB units containing relatively high numbers of CD34⁺ cells was correlated with improved event-free survival.

Table III.

Umbilical cord blood transplantation in adult patients - selected clinical trials

	Rocha et al. [13]	Laughlin et al. [14]	Wagner et al. [16]	Barker et al. [24]	Ballen et al. [25}	Brunstein et al. [26]
Dates of clinical trial (yrs)	1998–2002	1996–2001	1994–2001	2000–2003	2003–2005	2001–2005
Patients enrolled	98	150	102	23	21	110
Age in yrs (median)#(range)	24.5 (15–55)	na (16–60)	7.4 (0.2–56.9)	24 (13–53)	49 (24–63)	51 (17–69)
Patients with hematologic malignancies	98	150	65	23	19	106
Patients without hematologic malignancies	0	0	37	0	2	4
Myeloablative (M) or nonmyeloablative (N)	M	M	M	M	N	N
Degree of HLA match (%)				**Double UCB trial: each cord 4-6/6 matched to each other and to the recipient	**Double UCB trial: each cord 4-6/6 matched to each other and to the recipient	**Double UCB trial: each cord 4-6/6 matched to each other and to the recipient
matched (6/6)	6	0	14
1 or 2/6	90	100	84
3/6 or more	4	0	2
Median cell dose (range):
Nucleated cells/kg (×108)	0.23 (0.09-0.6)	0.22 (0.10–0.65)	0.31 (0.07–5.79)	0.35 (0.11–0.63)	0.4 (2.9–5.1)	0.37 (0.15–0.68)
CD34+ cells/kg (×105)	1.1 (0.8–8.8)	na	2.8 (0.4–39.1)	4.9 (1.2–14.5)	1.9 (0.6–9.7)	4.9 (0.7–16.6)
Days to neutrophil recovery (median) (range)	26 (14–80)	27 (25–29)	23 (9–54)	23 (15–41)	20 (15–34)	12 (0–32)
Days to platelet recovery (median) (range)	na	60 (54–71)	86 (29–276)	na	41 (21–55)	49 (0–134)
No. given hematopoietic growth factors	na	na	80	23	na	110
Incidence of acute GvHD:	In 98 patients:	In 150 patients:	In 63 patients:	In 23 patients:	In 19 patients:	In 110 patients
Grade I–II (%)	13	Grades II, III, IV: 41	51	52	53	37
Grade III–IV (%)	13		17	13	5	22
Patients with cGvHD/surviving patients	18/61	35/69	9/59	5/23	5/16	23/110
Incidence of relapse (%)	23	17	37	13	10	31
Treatment related mortality (TRM) (%)	44 (2 yrs)	63	30 (1 yr)	22 (6 months)	14 (100 days)	19 (180 days)
Overall survival (2–3yrs) (%)	63	19	58 (1 yr)	57 (1 yr)	71 (1 yr)	45 (3 yrs)

na = not available

Additional studies comparing UCB versus BM transplantation in adults found that patients transplanted with mismatched UCB had similar rates of TRM, treatment failure, overall mortality, and relapse rates compared to patients who received mismatched BM [13, 14]. Furthermore, Laughlin et al. [14] reported that the rate of acute GvHD in mismatched UCB transplant patients was comparable to the rate found in HLA-matched BM transplant patients. In a registry-based retrospective study of 682 patients, Rocha et al. [13] found both the rate of acute and chronic GvHD in mismatched UCB patients to be lower than the rates in matched BM recipients. However, the relationship between the relatively small cell dose found in UCB units and the delay in engraftment times and subsequent immune reconstitution for neutrophils and platelets continued to be problematic [13, 15, 17, 18]. Long et al. [17] noted that infection was the primary cause of death in their study of 57 adults with high-risk disease who were transplanted with unrelated UCB. A large multicenter prospective study of UCB transplantation in adults found that in 34 patients 90% experienced infections; two thirds of those studied reported at least 3 infections during the first 6 months post-transplant [19]. These authors suggested that the delay in immune reconstitution in UCB transplant patients versus BMT recipients could play a major role in their increased susceptibility to infection.

Given the critical importance of CD34⁺ cell dose and total nucleated cell dose to engraftment and immune reconstitution, new strategies have evolved to increase the total cell number provided in UCB transplants. Ex vivo expanded UCB has demonstrated potential for improved neutrophil engraftment despite the infusion of a less than optimal nucleated cell dose [20]. McNiece et al. [20] have developed a two-step ex vivo culture procedure that results in the further expansion of CD34⁺ cells with a resulting enrichment for mature neutrophils comparable to cultures of ex vivo expanded peripheral blood progenitor cells. Furthermore, human CD34⁺ UCB cells expanded ex vivo without a murine stromal cell feeder layer have “long-term” engrafting potential as demonstrated by their ability to reconstitute primary and secondary nonobese diabetic-severe combined immunodeficiency (NOD-SCID) mice or preimmune fetal sheep [21]. Such “stroma free” systems are more adaptable for clinical use by avoiding complications from murine feeder cells. Ex vivo expansion of UCB cells to facilitate engraftment and immune reconstitution continues to be an area of active investigation [22].

Another strategy for increasing the total nucleated cell dose of UCB transplants involves the infusion of a single recipient with multiple cord blood units. Studies in NOD-SCID mice have demonstrated that transplantation with two unrelated UCB units can increase engraftment compared to single unit transplants [23]. Barker et al. [18] reported the use of double UCB transplantation in the context of a clinical trial designed to evaluate the efficacy of reduced-intensity conditioning with UCB transplantation. This initial report was followed by a study devoted exclusively to evaluating the safety of double UCB transplantation for hematological malignancy in adults [24]. Following myeloablative conditioning, 23 patients with high-risk hematologic malignancy were infused with 2 partially HLA-matched UCB units containing a median cell dose of 3.5 × 10⁷ nucleated cells/kg. All of the 21 patients available for analysis demonstrated donor neutrophil engraftment at a median of 23 days post-transplant. Despite the infusion of 2 HLA disparate grafts, these patients did not demonstrate an increase in the incidence of GvHD observed with single unit transplants. Disease-free survival in this high-risk group was 57% at 1 year. Most notably, 24% of patients displayed engraftment from both UCB units and 76% of patients from 1 unit at day 21 while all patients demonstrated reconstitution with only 1 unit by day 100. The predominating unit was found to contain a significantly higher dose of CD3⁺ cells.

Additional clinical studies of double UCB transplantation have demonstrated the safety and efficacy of this approach in the treatment of hematological malignancies. In a study of 21 patients treated with a nonmyeloablative approach, Ballen et al. [25] found that TRM at day 100 post-transplant was only 14% and the occurrence of severe acute GvHD remained low. As in the study by Barker et al. [24], all of the patients engrafted with 1 UCB unit by 3 months posttransplantation. Relapse rates among these patients were low suggesting that the GvL effect was present despite the low rates of GvHD. Interestingly, patients who displayed mixed chimerism with both cord blood units at 6 weeks post-transplant presented with a higher incidence of chronic GvHD, a finding that may reflect an ongoing graft-versus-graft interaction. A recent report by Brunstein et al. [26] further substantiates the low rates of severe GvHD, reduced TRM, and low relapse rates seen with the double UCB transplant approach. In this study of 110 patients, neutrophil engraftment was notably rapid and occurred at a median of 12 days in 92% of patients. Although many parameters were examined including total CD34⁺ and CD3⁺ numbers, the authors found no predictive characteristic for determining which UCB unit would eventually predominate in these patients. Ting et al. [27] has reported that in a group of 10 UCB double transplant patients, the predominant cord blood unit contained a higher number of NK cells compared to the rejected unit.

Summary

UCB transplantation has been rigorously examined in the setting of many clinical trials and continues to develop as a viable alternative to the use of bone marrow or mobilized peripheral stem cells for the treatment of hematological malignancies and bone marrow failure. The use of UCB as a stem cell source has revealed certain advantages for transplantation recipients. Relative to other donor sources, UCB units can be obtained quickly and mediate less severe GvHD without compromising GvL effects. However, the total cell numbers in UCB units are at least one log lower than bone marrow or peripheral stem cell sources. New approaches, such as the ex vivo expansion of UCB stem cells and the use of double unit transplantation, have been employed to avoid a prolonged delay in immune reconstitution.

UCB basic biology and implications for the development of GvHD post-UCB transplantation

One of the early clinical observations in the outcome of UCB transplantation has been the fact that UCB permits a greater degree of HLA mismatching without an unacceptably high incidence of GvHD. HLA disparity between the donor and recipient is an important determinant of acute and subsequent chronic GVHD. Notably, a higher incidence of acute and chronic GVHD has been observed in patients transplanted with HLA-matched unrelated grafts when compared with histocompatible sibling grafts, possibly attributable to reactivity of donor T cells with recipient minor histocompatibility antigens. Minor histocompatibility antigen disparity is expectedly greater between unrelated individuals. Despite the use of HLA class I and II disparate grafts, the incidence and severity of acute GVHD observed in pediatric and adult recipients of UCB grafts is lower when compared to recipients of unrelated adult donor grafts [6, 9, 11, 13, 16, 28].

Multiple factors intrinsic to the UCB graft may contribute to the reduced GVHD observed after UCB transplantation. These may include reduced graft lymphocyte numbers, altered recognition of recipient self antigens by UCB donor T cells interacting with recipient’s antigen-presenting cells (APC) and limited response of these naive donor T cells activated by recipient alloantigen. Subsequently, these changes will result in impaired cytokine production, limited cellular activation and lack of clonal expansion of alloreactive T cells [29–31].

UCB T lymphocytes are typically CD45RA+ and express low levels of activation markers. Several in vitro studies point to the inherent lack of full expression of immunomodulatory cytokines by alloreactive T cells contained in UCB grafts. In primary mixed lymphocyte culture UCB T cells demonstrate proliferative responses to allogeneic stimulation, but less cytotoxic effector function, less proliferation and greater activation-induced cell death (AICD). Further mechanisms potentially underlying UCB immune tolerance includes altered toll-like receptors and adhesion molecule expression on donor graft antigen-presenting cells [31]. Early recovery of NK cells able to activate the granzyme/perforin lytic pathway and Fas/Fas ligand (FasL) activity has also been proposed as contributing to the low incidence of acute GVHD observed after UCB transplantation [32]. Other studies have suggested that UCB graft T cells display reduced expression of nuclear factor of activated T cells-1 (NFAT1), which may be one important molecular mechanism underlying their reduced capacity for cytokine production [33]. The reduced GVHD after UCB transplant reported in clinical studies may be related to these in vitro observations that T cells in the UCB graft respond less fully than mature alloreactive lymphocytes.

The long-term clinical implications of the reduced GvHD in UCB transplantation are currently not well understood. From the present clinical experience it is clear that, despite the decreased incidence and severity of GvHD associated with UCB grafts, GvL immunologic effects are maintained [34]. Clinical reports of allogeneic UCB recipients have not identified increased relapse rates [7–9, 14, 15]. Interestingly, studies have reported that successful immune reconstitution after UCB transplantation, as determined by development of detectable CMV, VZV and HSV immunity, is associated with reduced relapse rate [35]. These results indicate that despite reduced incidence of GvHD, lymphocytes in the UCB graft are capable of providing significant immunity against leukemia and viral antigens.

UCB basic biology and implications for hematopoietic reconstitution post-UCB transplantation

Graft characteristics known to positively correlate with rapid donor engraftment in recipients of conventional allografts include cell dose, CD34 content, and HLA matching. The rate of donor hematopoietic reconstitution is lower and kinetics of engraftment and myeloid recovery are delayed using UCB compared to bone marrow grafts [14]. This event is thought to be secondary to the low stem cell dose in the UCB graft compared with adult donor grafts [6, 36]. UCB graft characteristics shown to have a predictive value for time to myeloid engraftment include total nucleated graft cell content, CD34 content and colony-forming units (CFU) [15, 37]. UCB HLA compatibility is also predictive of engraftment [15]. Minor histocompatibility disparity in unrelated allogeneic transplantation may contribute to graft rejection and to graft-versus-leukemia effects [38–41]. HLA class I mismatching including HLA-C and NK epitope mismatching are associated with higher rates of graft rejection and severe acute GVHD after unrelated donor transplantation [42, 43]. Allele matching for adult unrelated blood and marrow grafting has improved rates of engraftment and GVHD. Effects of the number and type of graft HLA disparities on UCB donor engraftment have not been fully studied. Perhaps such evaluation for UCB selection will improve engraftment time.

Various groups have attempted to compare HSC in UCB to those of peripheral blood grafts (Table IV). These studies have revealed a less mature phenotype of UCB CD34+ progenitors compared to adult marrow and peripheral blood grafts [12, 44]. Interestingly, these studies have shown that the frequency of early stem cells is similar in adult bone marrow grafts, mobilized peripheral blood HSC, and UCB, but the proliferative potential of UCB early stem cells is significantly higher. Cobblestone area-forming cell (CAFC) assays have also shown that UCB CD34+ cells contain the highest frequency of CAFC (3.6- to 10-fold higher than BM CD34+ cells and peripheral blood stem cells, respectively). In addition, the engraftment capacity in vivo, as determined by nonobese diabetic/severe combined immunodeficiency (NOD/SCID) repopulation assay, is also significantly greater than BM CD34+ cells [45, 46]. From the above studies it becomes apparent that UCB may have lower CD34+ cell numbers, which have a higher proliferation potential than adult CD34+ cells. Thus, other characteristics of UCB progenitor cells may be responsible for the delayed engraftment of UCB stem cells. Such characteristics may include adhesion molecules, homing properties and maturational stage of UCB progenitor cells [22, 47–49]. These factors may have a major contribution in the kinetics and efficiency of engraftment and may explain why CD34 quantification in UCB has not been consistently predictive of time to hematopoietic reconstitution.

Table IV.

Biology of umbilical cord blood cells versus bone marrow or mobilized peripheral stem cells (HSC)

	UCB		HSC
T cell response to alloantigens/mitogens	Modest response		Vigorous response
Proportion of CD4+ CD45RA+ cells	Majority of T cells		Dual population of CD45RA+/CD45RO+
Proportion of hematopoietic progenitor cells	Greater than HSC		Reduced
CD34+ proliferation rate in response to cytokines in vitro	Greater than HSC		Reduced
Production of hematopoietic factors* from activated mononuclear cells	Significantly less than HSC		Higher amounts
Production of IL-2, IFN-γ, TNF-α from activated T cells	Significantly less than HSC		Higher amounts
Allogeneic cytotoxic activity of activated T cells	Significantly less than HSC		Higher amounts
Expression of NFAT1	Reduced		Higher
Natural killer cell activity	——	Comparable	——

^*G-CSF, GM-CSF, IL-3, M-CSF, TGF-β1, MIF-1α [78]

Several approaches have been employed to improve rates and kinetics of hematopoietic engraftment in adult UCB recipients. A successful strategy involves infusion of more than one UCB graft. As mentioned in the previous sections, initially, the Minnesota group reported their experience with adults infused with two UCB units in whom 92% demonstrated sustained donor engraftment [18] . Interestingly, although patients developed a transient state of mixed chimerism, eventually one donor dominated over time. Neither cell dose nor HLA disparity were predictive for selection of engraftment. GvHD incidence paralleled that of historic controls and 1-year survival was 33%. These observations were encouraging and further studies using two UCB units in adult recipients followed immediately thereafter. The Dana-Farber/MGH Partners Cancer care group reported their preliminary experience using two UCB units in adult patients following a nonmyeloablative conditioning regimen [25]. This study used a reduced-intensity conditioning regimen of fludarabine, melphalan, and antithymocyte globulin followed by 2 partially matched UCB units. The UCB units were a 4/6 HLA match or better with each other and with the patient and achieved a minimum precryopreservation cell dose of 3.7 × 10⁷ nucleated cells/kg. The median time to an absolute neutrophil count >0.5 × 10⁹/L was 20 days, and the median time to an unsupported platelet count >20 × 10⁹/L was 41 days. Two patients experienced primary graft failure and underwent a second UCB transplant. One patient had a late graft failure. Acute graft-versus-host disease (GVHD) grade II–IV occurred in 40% of patients. The 100-day TRM was 14%, and the 1-year disease-free survival was 67%. Mixed chimerism was associated with a higher risk of chronic GVHD.

A second interesting approach to enhance hematopoietic recovery in adult UCB recipients has been proposed by Fernandez et al. [50], who attempted simultaneous transplantation of a low number of highly purified peripheral blood CD34+ cells from an HLA haplo-identical donor along with UCB. All patients had prompt myeloid recovery and the median time to an absolute neutrophil count >0.5 × 10⁹/L was 10 days. One interesting observation was that after a transient initial phase of mixed chimerism and engraftment of haplo-identical HSC there was a subsequent progressive replacement by UCB. However, a major concern with this strategy is the potential development of severe acute GvHD that could be triggered by the haplo-identical CD34+ cells in this setting.

A third strategy was proposed in an attempt to shorten the time interval to attain donor-derived hematopoietic recovery after UCB transplantation, and allow transplantation of adult recipients. This approach involves expansion of UCB in cytokines in vitro prior to infusion [51–53]. Early clinical trials based on infusion of a proportion of the UCB graft expanded ex vivo in cytokine-based conditions have failed to demonstrate more rapid hematopoietic recovery in UCB recipients, suggesting that cytokine based expansion may result in differentiation of early self-renewable stem cells. Reports from the Duke University program have shown that infusion of ex-vivo expanded UCB did not alter the time to myeloid, erythroid or platelet engraftment [53]. Shpall et al. reported their experience at the University of Colorado with infusion of ex vivo expanded cells along with the UCB [51]. Although this approach allowed engraftment in all patients, the median time to reach an absolute neutrophil count 0.5 × 10⁹/L was 28 days and did not differ from that observed in adult patients infused with non-expanded UCB [15]. These observations suggest that not simply the number of CD34+ cells but other components or properties of the UCB govern engraftment.

UCB basic biology and implications for immune reconstitution post-UCB transplantation

Following allogeneic transplantation, all patients experience a period of profound immunodeficiency. Immune reconstitution of T- and B-cell compartments following allogeneic transplantation may require as long as 12–24 months. The slow process of immune reconstitution together with post-engraftment immunosuppression creates an immunologic environment whereby the host is susceptible to opportunistic infections [54– 56] as well as virally induced malignancies [57, 58]. Recipients of unrelated donor or HLA non-identical transplants appear to have a higher rate of infectious complications than recipients of matched sibling allogeneic grafts [59–61]. GvHD and its treatment further delays immune recovery after allogeneic transplantation and may account for the higher rate of delayed immune reconstitution in recipients of unrelated versus related donor grafts.

Several reports outlining the kinetics of immune recovery after UCB transplantation in adults and children point to slow T-cell recovery [60],[62–66] compared with published reports after conventional graft allogeneic transplantation [66–71]. Recent studies have evaluated the immune function of adult UCB patients during the first two years post-transplant. Immune recovery in these early series of adult UCB transplant recipients was marked by profound lymphopenia and immunodeficiency during the first 6–12 months after transplant. However, when immune recovery was attained, generally 9–12 months after transplant, recovery of both T- and B-cell function was noted. Enumeration of T, B and NK cells revealed absolute lymphocyte counts measured in the early post-transplant time period were very low but absolute numbers of circulating NK cells did not differ from that of adult controls, a finding consistent with previous observations in BMT recipients [72, 73]. B cells (both proportions and absolute numbers) increased beginning 6 months post-UCB transplant and persisted at 1 year follow-up. The mechanism and significance of this B-cell ‘rebound’ phenomenon observed post-transplant in adult UCB recipients is unclear. Work from the Duke University program reported that T-cell recovery was comparable in adults and children following UCB transplant [64]. This event was associated with delayed emergence of thymic recent emigrants cells (TREC) in adults (18 months) vs. children (12 months) as well as lower levels than expected for age. That study also reported skewing of T-cell repertoires in these patients persisting until 2–3 years after transplant.

The immediate consequence of the impaired rate and quality of immune recovery after UCB transplantation is directly associated with higher infection rates in adult UCB recipients. The Case Western Reserve University transplant group evaluated patients with hematologic disorders treated with myeloablative regimen and transplantation with either UCB or HLA-matched unrelated (MUD) grafts, for life-threatening infections, hematologic reconstitution, GvHD, relapse and event-free survival [60]. The median duration of neutropenia after transplantation was longer (29 vs. 14 days) in the UCB group. Overall infection rates were higher in UCB recipients, particularly at early time points (prior to day 50) after transplantation. Event-free survival at 3-year follow-up was 0.25 in UCB and 0.35 in MUD recipients.

Recent reports indicate that the rate of infections during the early post-transplantation period is higher in adult patients transplanted with HLA mismatched unrelated UCB, while overall rates of infections at later time points are similar to that observed in unrelated adult donor transplant recipients [60]. Mechanisms responsible for this event may be related to the prolonged duration of neutropenia and lymphopenia after infusion of smaller numbers of total graft nucleated cells and CD34+ cells. Saavedra et al. reported a high incidence of bacteremia (55%) in 27 adults at early time points after UCB transplant. Ten patients (37%) died prior to day 100. Infection was a direct cause of death in 4 patients [74]. Investigators from Japan reported cytomegalovirus (CMV) infection following UCB in 28 adults compared with sibling matched (R-BMT) and URD BM recipients. CMV antigenemia was observed in 19 (79%) of UCB patients at median 42 days [75]. A higher proportion of UCB patients treated with preemptive gancyclovir therapy required a second course of treatment compared with R-BMT and URD BM patients, suggesting that CMV-specific immunity after UCB may be delayed.

In addition to the delayed engraftment and cellular reconstitution, increased infection rates in UCB recipients may be related with other cell populations that are present in the UCB. Extensive studies have shown that UCB is a superior source for isolation of T regulatory cells (Treg) that have a more potent suppressor function that Treg isolated from adult peripheral blood [76]. Fresh CD4⁺CD25⁺ cells isolated from UCB expressed high levels of Foxp3 mRNA and protein that was further upregulated after in vitro culture. Cell lines were readily generated from CD4⁺CD25⁺ (Treg) isolated from UCB, by repetitive stimulation with anti-CD3/CD28 coated beads and IL-2 culture. Moreover, UCB-Treg cell lines displayed impaired activation of Ras, MEK1/2 and Erk1/2, whereas activation of Akt was retained and activation of Rap1 was enhanced [77]. Treg also displayed increased expression of p27^kip1 and had high susceptibility to apoptosis. This event was reversed by IL-2, which induced activation of Erk1/2, upregulation of Bcl-x_L and phosphorylation of Bad at Ser¹¹², a site specifically phosphorylated by Erk. These UCB-derived Treg cell lines had potent suppressive function when added in HLA-mismatched allo-MLRs, in which CD4⁺ responder cells were stimulated with allogeneic DC that generate a very robust proliferation. Interestingly, the inhibitory activity of UCB-derived Treg cell lines was persistently superior to the inhibitory activity of PBMC-derived Treg in all responder: stimulator ratios tested. UCB Treg cell lines also markedly reduced cytokine and chemokine accumulation in MLRs. These observations may have potentially significant clinical implications because these highly suppressive UCB Treg may be able to inhibit responses of effector cells in UCB recipients.

Summary

The unique biology of the HSC and lymphocytes found in UCB grafts provides the explanation for the variations in clinical outcomes seen in HSC versus UCB transplantation. The lower incidence and severity of GvHD seen in UCB recipients is a direct consequence of the reduced proliferation, cytokine production, and cytotoxicity to alloantigens displayed by UCB lymphocytes. The prolonged delay to neutrophil and platelet reconstitution seen in UCB patients can be attributed to relatively small numbers of total nucleated cells and CD34+ stem cells found in UCB grafts. However, the substantial proliferative and engraftment potential of UCB CD34+ cells can offset these disadvantages to a degree that allows UCB to be a viable source of HSC for transplantation. In terms of immune function, UCB recipients remain highly vulnerable to infection. Further basic research and new methodologies are necessary to address the prolonged times to engraftment and the possible influence of T regulatory cells on mechanisms of immune suppression.

Concluding remarks and future directions

Lower incidence of acute GVHD in UCB transplant recipients would be expected to be associated with higher rates of malignancy relapse, similarly to the highest rates of infectious complications. However, relapse rates after UCB transplant remain low. The mechanisms underlying the strong graft versus-leukemia (GVL) effects mediated by UCB have not been delineated. Clinical reports of allogeneic UCB recipients have not identified increased relapse rates, despite the fact that the majority of patients have advanced disease at the time of transplant. The unique immunologic features of UCB that allow HSC transplantation in the absence of full HLA compatibility, resulting in low incidence of GvHD but efficient GvL, make UCB an attractive approach for patients requiring allogeneic stem cell grafts. Further understanding of the mechanisms responsible for delayed engraftment and immune reconstitution will improve the outcome of UCB transplantation and will possibly alter our choice priority between peripheral blood mobilized stem cells and UCB as a HSC source in adult patients.