Keywords

1 Unique Characteristics of Immune Cells in the UCB Graft

The unique composition and properties of cells contained in UCB units (Table 12.1) may account, at least in part, for the distinct patterns of hematopoietic and immune reconstitution after UCBT. UCB contains higher numbers of CD4+ and CD8+ T-cells compared to adult peripheral blood (PB) (Szabolcs et al. 2003), although the absolute counts of immune cells are lower compared to peripherally mobilized or bone marrow (BM) adult stem cell grafts (Storek 2008). In addition, the majority of UCB CD4+ and CD8+ T-cells are phenotypically naïve (CD45RA+CD62L+), in contrast to adult PB, which contains predominantly memory T-cells (Szabolcs et al. 2003; Chalmers et al. 1998). UCB T-cells produce less IL-2, IL-4, IFN-γ, and TNF-α in response to mitogens (Chalmers et al. 1998). The deficient production of cytokines by stimulated UCB T-cells may be due to the reduced expression of NFAT1 or NFAT-associated genes (Kadereit et al. 1999; Kaminski et al. 2003). While UCB T-cells display vigorous proliferation after allogeneic stimulation, they have decreased antigen-specific cytotoxicity (Risdon et al. 1994). This might be explained by aberrations in the Fas ligand pathway and the decreased perforin expression (Sato et al. 1999; Berthou et al. 1995). Despite an intrinsic Tc2-Th2 bias (Szabolcs and Niedzwiecki 2007), UCB T-cells can rapidly acquire Tc1/Th1 function and mediate enhanced antitumor effects compared to adult PB T-cells in a murine model of B-cell lymphoma (Hiwarkar et al. 2015).

Table 12.1 Properties of hematopoietic cells in UCB grafts versus HSPC grafts from peripherally mobilized adult donor HSPC (pmHSPC)
Full size table

UCB is enriched in CD4+CD25+ T regulatory cells (Tregs), which express Foxp3 and exhibit potent suppression function, independent of IL-10 or TGF-β (Godfrey et al. 2005). UCB Tregs are characterized by higher expression of the naïve markers CD45RA and CD38 and lower expression of activation markers CD45RO and HLA-DR in comparison to adult PB Tregs. Freshly isolated UCB Tregs are incapable of suppressing alloantigen-induced proliferation (Wing et al. 2003; Chang et al. 2005), a property that has been attributed to their naïve state, but can effectively suppress T-cell responses after stimulation (Wing et al. 2003; Takahata et al. 2004). Interestingly, after allogeneic stimulation, CTLA4 expression on Tregs derived from UCB CD4+CD25– T-cells is induced at a higher rate compared to CD4+CD25– T-cells derived from adult HSPC grafts (Chang et al. 2005). UCB Tregs also have a higher expansion capacity to anti-CD3, anti-CD28 and IL-2, which correlates with overexpression of genes involved in proliferation and chromatin remodeling (Torelli et al. 2012).

Similarly to T-cells, UCB B-cells are predominantly naïve and are characterized by the expression of CD5 and sIgM (B1 cells), whereas CD23-expressing B-cells are less frequent compared to adult PB (Szabolcs et al. 2003). These UCB B-cells can produce poly-reactive IgM antibodies but display decreased capacity for Ig class switching and production of IgA or IgG upon stimulation (Hayward and Lawton 1977; Durandy et al. 1990). Furthermore, UCB B-cells have increased susceptibility to spontaneous apoptosis (Kessel et al. 2006). CD56bright and CD56dim NK subsets are present at similar proportions in UCB and PB, although UCB also contains a significant population of NK progenitors, including CD16+CD56– cells (Verneris and Miller 2009; Gaddy and Broxmeyer 1997; Gaddy et al. 1995). UCB NK cells have high-level expression of the inhibitory receptor NKG2A/CD94 and low-level expression of granzyme B, resulting in decreased NK-cell activity compared to adult NK cells (Wang et al. 2007). Decreased expression of adhesion molecules may also account for the decreased NK-cell activity (Tanaka et al. 2003). UCB dendritic cells (DCs) also display unique properties and are characterized by lower surface expression of major histocompatibility complex (MHC) and cell adhesion molecules (Hunt et al. 1994), decreased upregulation of costimulatory molecules or activation markers (CD80, CD86, CD86), and markedly decreased production of IL-12 in response to LPS stimulation (Langrish et al. 2002; Wong et al. 2005). Thus, UCB DCs are functionally immature with impaired ability to stimulate T-cell responses. Instead, UCB DCs efficiently induce CD4+ T-cells with a Treg phenotype (Encabo et al. 2007). UCB is also characterized by a higher ratio of plasmacytoid to myeloid DCs compared to adult PB (Crespo et al. 2004). These observations support the notion that UCB DCs are biased against Th1 responses and toward induction of immune tolerance (Langrish et al. 2002; Wong et al. 2005; Crespo et al. 2004).

2 Mechanisms and Pathways of T-Cell Reconstitution After UCBT

While the innate immune system recovers within weeks after transplantation, the restoration of the adaptive T-cell immunity is delayed for several months. Recovery of T-cells after HSCT follows two different pathways, one peripheral or thymus-independent and one central or thymus-dependent pathway. These two pathways of T-cell reconstitution act in parallel but follow different kinetics (Williams et al. 2007) (Fig. 12.1).

Fig. 12.1
figure 1

Reconstitution of the T-cell compartment after UCBT. Conditioning chemoradiation prior to UCBT results in profound lymphopenia and immunodeficiency of the host. T-cell reconstitution after UCBT is achieved by two independent mechanisms: The thymus-independent pathway of T-cell reconstitution predominates in the early post transplant period and is mediated by adoptively transferred UCB T-cells, which are uniformly naïve and do not transfer protective immune memory, or recipient T-cells that survive conditioning. These T-cell populations undergo peripheral expansion in response to lymphopenia and high cytokine levels (IL-7, IL-15 etc), or oligoclonal proliferation upon interaction with cognate antigen. Overtime, this early peripheral T-cell expansion results in late effector memory T-cell skewing and contraction of the T-cell repertoire diversity, and is associated with impaired immunologic responses to antigens. Reconstitution of a functionally competent T-cell compartment with broad antigenic specificity eventually requires the de novo production of naïve T-cells by the thymus of the USCT recipient. This thymus-dependent pathway of T-cell reconstitution is a prolonged multistep process. LMPs contained in the UCB graft or arising from the engrafted donor-derived HSCs migrate via circulation and repopulate the thymus with thymocyte precursors that can reconstitute thymopoiesis. The thymus provides the essential microenvironment (stroma) that supports T-cell proliferation, selection and differentiation into Recent Thymic Emigrants. Several factors can delay the recovery of thymopoiesis after UCBT, including low number of LMPs (as a result of low nucleated cell dose of UCB and delayed engraftment), advanced recipient age with resultant thymic involution and thymic damage from the conditioning chemoradiation or GvHD. Although slow, the thymus-dependent mechanism is imperative for the renewal of peripheral T-cell pool and constant export of new naïve T-cells with broad TCR repertoire diversity, capable of responding to a great spectrum of antigens. (This research was originally published in Blood. Politikos I and Boussiotis VA. The role of thymus in T-cell immune reconstitution after umbilical cord blood transplantation. Blood. 2014; Vol 124:3201–3211. © the American Society of Hematology)

Full size image

2.1 Peripheral or Thymus-Independent Pathway of T-Cell Reconstitution

In the early posttransplant period, the thymus-independent pathway predominates and is mediated by adoptively transferred donor T-cells contained in the graft or recipient T-cells surviving conditioning chemotherapy. These T-cells proliferate following antigen encounter (Mackall et al. 1996) or undergo lymphopenia-driven homeostatic peripheral expansion (HPE) through low-affinity interactions with self-peptide MHC complexes under the influence of homeostatic cytokines, which are highly elevated during this posttransplant period (Goldrath and Bevan 1999). Among these cytokines, IL-7 is a critical regulator of the thymus-independent pathway (Fry et al. 2001). Studies in HSCT recipients have shown that IL-7 levels rise in the posttransplant period and inversely correlate with T-cell counts, most likely due to decreased consumption (Bolotin et al. 1999; Dean et al. 2008). In double unit UCBT (dUCBT) recipients, IL-7 levels remain elevated for a prolonged time compared to recipients of adult donor HSPC, and inversely correlate with T-cell counts, especially CD4+ T-cells (Politikos et al. 2015).

In contrast to BM or peripherally mobilized HSPC grafts that contain a large number of memory T-cells, UCB grafts contain T-cells that are almost exclusively naïve. As a consequence, UCB T-cells display a complete T-cell receptor (TCR) repertoire at birth (Garderet et al. 1998). However, they are subject to the stringent requirements of naïve T-cells during mounting of primary immune responses which cannot provide efficient protection, in contrast to memory T-cells which develop rapid antigen-specific recall responses. As a result, protection against pathogens during the early posttransplant period is impaired and depends on memory T-cells of host origin that survive conditioning chemotherapy (Chalandon et al. 2006). During HPE, UCB T-cells lose their naïve phenotype and acquire an effector or memory-like phenotype in an antigen-independent manner (Mackall et al. 1993; Cho et al. 2000). At the same time, antigen-driven proliferation of UCB T-cells circulating in the UCBT recipient results in oligoclonal expansion and contraction of TCR repertoire diversity. Moreover, apoptotic cell death of the activated T-cells during homeostatic activation and/or antigen exposure may lead to contraction of the T-cell compartment over time (Hakim et al. 1997). As a result of these concurrent mechanisms, the thymus-independent pathway cannot sustain a diverse TCR repertoire with broad antigenic specificity that ensures immune competence (Roux et al. 1996).

2.2 Central or Thymus-Dependent Pathway of T-Cell Reconstitution

The central or thymus-dependent pathway of T-cell recovery involves the de novo production of T-cells by the thymus, which is critical for successful long-term immune reconstitution (Politikos and Boussiotis 2014). This pathway generates naive T-cells with broad antigenic specificity, and it is essential for the restoration of TCR repertoire diversity (Roux et al. 2000; Dumont-Girard et al. 1998). Restoration of thymopoiesis is a multistep process that begins with the seeding of the thymus with lymphoid-myeloid progenitors (LMPs) derived from the engrafted donor HSPCs. Upon entry in the thymic cortex, LMPs undergo T-cell lineage commitment under the influence of Notch signaling and interactions with cortical thymic epithelial cells (TECs) (De Smedt et al. 2002; Taghon et al. 2009). During intrathymic T-cell development, thymocytes undergo sequential rearrangement of the TCR beta (TCRB) and alpha (TCRA) loci and develop into double-positive (DP) thymocytes, which are subject to positive selection based on their ability to bind self-peptide/self-MHC complexes with appropriate affinity. Positively selected DP thymocytes continue their development into CD4 or CD8 single-positive (SP) thymocytes, migrate to the medulla via the upregulation of the chemokine receptor CCR7, and undergo negative selection by medullary TECs, which have the unique capability to promiscuously express tissue-restricted antigens (TSAs) under the influence of the transcription factor Aire (Anderson et al. 2002, 2005). The negative selection process ensures the elimination of self-reactive lymphocytes. The small number of T-cells that successfully proceed through these intrathymic steps is exported to the periphery and is termed recent thymic emigrants (RTEs).

3 Factors Influencing T-Cell Immune Reconstitution After UCBT

3.1 Graft-Related Factors

Several factors can affect the timing and efficiency of immune reconstitution after UCBT. Graft-related factors such as limited number of stem cells and lymphocytes, T-cell naivety, immaturity, and lack of memory T-cells have the greatest impact on the thymus-independent pathway of T-cell recovery. Specifically, total nucleated cell (TNC) and CD3+ T-cell content in UCB units positively correlate with lymphocyte recovery in the early posttransplant period after UCBT (Niehues et al. 2001; Castillo et al. 2016). It is conceivable that the low stem cell dose of UCB grafts and the delayed engraftment result in delayed emergence of LMPs that repopulate the thymus and restore thymopoiesis. Thus, the properties of the UCB graft may also affect the thymus-dependent pathway of immune reconstitution. Interestingly, although the number of HSPC clones contributing to thymic seeding and intrathymic T-cell development is restricted, this very limited number of HSPC clones can give rise to a fully diverse TCR repertoire. Thus, immunodeficiency after UCBT may not be a consequence of low HSPC numbers but may rather be related to intrathymic events (Brugman et al. 2015). An additional important mechanism regulating immune reconstitution is related to the frequent HLA discordance between the UCB graft and the host, due to the less stringent requirements for HLA matching during UCBT. The intrathymic positive and negative selection is dictated by MHC expression of thymic epithelial cells, which remain of host origin after transplantation, in contrast to the donor-derived LMPs that repopulate the thymus (Krenger and Hollander 2008). As a result, MHC mismatches between graft and the host might adversely impact immune reconstitution, by compromising thymic selection and leading to skewed TCR repertoire diversity (Roux et al. 2000; Krenger and Hollander 2008; Eyrich et al. 2003).

3.2 Host-Related Factors

Host-related factors also influence immune reconstitution. Recipient’s age has a critical impact on T-cell reconstitution. Compared to adults, pediatric patients receive higher TNC and CD3+ cell doses adjusted per body weight resulting in faster neutrophil engraftment and lymphocyte recovery. In addition, the decreased thymic function that occurs with aging as a result of thymic involution (Mackall et al. 1995; Douek et al. 1998) might pose a rate-limiting step in the immune reconstitution after HSCT. In support of this hypothesis, the recipient’s age is inversely associated with the onset and extent of thymopoietic recovery after HSCT or UCBT (Klein et al. 2001; Clave et al. 2009; Uhlin et al. 2012). Furthermore, the pretransplant thymic function of the host plays a role in the rate of thymic recovery (Mackall et al. 1995; Chen et al. 2005; Clave et al. 2005).

3.3 Treatment-Related Factors

Factors related to the transplantation procedure or posttransplant complications may affect the rate and success of immune reconstitution. The use of in vivo T-cell depletion by means of ATG has a detrimental effect on the early thymus-independent T-cell recovery (Admiraal et al. 2015; Lindemans et al. 2014). A negative effect of ATG on the thymus-dependent pathway has also been implied (Sairafi et al. 2012). Reduced intensity conditioning (RIC) regimens may provoke less damage to the peripheral niches in which T-cell proliferation takes place, thus allowing for a faster immune recovery (Uhlin et al. 2012; Sairafi et al. 2012; Chao et al. 2002). Furthermore, radiation-based conditioning regimens may lead to delayed thymus-dependent T-cell reconstitution, likely by inducing greater thymic destruction (Glowala-Kosinska et al. 2016; Chung et al. 2001). It should also be noted that the thymus is a target organ of graft-versus-host disease (GvHD), which affects both its lymphoid and stromal compartments (Krenger and Hollander 2008). Importantly, the development of acute and chronic GvHD has been associated with delayed T-cell reconstitution (Clave et al. 2009; Uhlin et al. 2012; Weinberg et al. 2001) not only because the thymus is a target organ of GvHD, but also because corticosteroid therapy for the treatment of GvHD can lead to lymphodepletion.

4 Assessment of Immune Reconstitution After UCBT

Several tools are available for the study of immune recovery after transplantation, both in the clinical or research settings. Hemograms provide an assessment of myeloid engraftment and quantitative recovery of lymphocytes by measurement of absolute lymphocyte count (ALC). Flow cytometry provides a means for the enumeration of lymphocyte subsets and monitoring of CD4+, CD8+, B, and NK cells is commonly used in the clinical practice to guide decisions regarding the management of transplant recipients, such as discontinuation of prophylactic antiviral therapy and timing of vaccination. Flow cytometry is also a valuable tool for the study of thymopoieisis because it allows quantification of RTEs based on the expression of the CD45 isoform, CD45RA (Mackall et al. 1993). Additional surface markers, such as CCR7, CD31, CD62L, and CD27, have improved the differentiation between naïve and memory T-cells (McFarland et al. 2000). Quantification of T-cell receptor excision circles (TRECs), which are DNA by-products of the intrathymic TCR sequential rearrangement events, also provides an excellent assessment of the thymic function (Politikos and Boussiotis 2014; Douek et al. 1998). Additionally, because diversification of the T-cell repertoire after HSCT depends on the de novo production of naïve T-cells, assessment of the TCR diversity by CDR3 spectratyping or by sequencing of the TCR β chain can also provide information about the ability of thymopoiesis to reconstitute a competent T-cell immune system (Douek et al. 2000; Robins et al. 2009). Because these assays provide only quantitative assessment, approaches to evaluate pathogen-specific T-cell responses such as cytokine flow cytometry (CFC) or enzyme-linked immunosorbent spot (ELISpot) are valuable tools for the assessment of immune competence.

5 Quantitative Immune Reconstitution After UCBT

5.1 Immune Reconstitution in Adults After Single-Unit UCBT

Adult patients undergoing single-unit CBT (sUCBT) after ATG-containing myeloablative conditioning (MAC) have delayed neutrophil engraftment and slower ALC recovery during the first 2 months after transplantation compared to matched unrelated donor (MUD) transplant recipients. However, ALC recovery in the sUCBT group surpassed that of MUD recipients beyond day 180 and reached normal range by day +200. While bacterial infections were more common in the sUCBT group early posttransplant, there was no significant difference in the infection rates between the two groups beyond day 50 (Hamza et al. 2004). However, adult patients who receive sUCBT after ATG-containing conditioning may have protracted lymphopenia for up to 2 years posttransplant (Klein et al. 2001). This is most pronounced for the CD4+ T-cells, which remain below normal range until 2 years, whereas CD8+ T-cells reach low normal levels by 1 year. The kinetics of T-cell recovery in adult patients is markedly delayed compared to pediatric UCBT recipients. Moreover, the majority of T-cells during the first year have a memory phenotype.

An important observation is that TREC levels in adult sUCBT recipients remain below the age-adjusted normal range for up to 3 years after transplantation, suggesting a profound thymic deficiency compared to the pediatric control group (Klein et al. 2001). This finding explains, at least in part, the protracted delay in T-cell recovery. Similarly, sUCBT after ATG-containing RIC is associated with extremely suppressed CD4+ and CD8+ numbers at least through 6 months post-UCBT (Komanduri et al. 2007). Under these conditions, CD8+ T-cells approach near-normal levels at 1 year after transplantation, but CD4+ T-cells remain low. Thymopoiesis, as assessed by sjTRECs, is nearly absent in the first year and is associated with a paucity of naïve and central memory T-cells and expansion of terminally differentiated T effector cells. In contrast, NK and B-cells develop a marked expansion to supranormal levels, as early as 30 days post-UCBT.

5.2 Immune Reconstitution in Adults After Double-Unit UCBT

The use of dUCBT has been employed to circumvent the cell dose limitation in adult patients and to enhance hematopoietic and immune reconstitution. Analysis of immune reconstitution in adult patients undergoing dUCBT (Cutler et al. 2011) and comparison between dUCB and peripherally mobilized HSPC MUD transplantation after RIC (Jacobson et al. 2012) revealed that CD3+, naïve and memory CD4+, CD8+, and CD4+CD25+ Treg numbers were significantly lower in dUCBT recipients during the first 6 months after transplantation. The inclusion of ATG only in the dUCBT group may have contributed to this observation. By 12 months, no significant difference was observed between the two groups, and, by 24 months, the median number of naïve CD4+ T-cells in the dUCBT cohort surpassed that in the MUD cohort. In addition, dUCBT recipients seemed to have higher CD123+ plasmacytoid DCs in the first year compared to MUD, but the difference disappeared at later time points. Infections were more frequent in the dUCBT group, but the non-relapse mortality (NRM) was not different (Jacobson et al. 2012). In addition, ATG may also cause a delay in the emergence of IFN-γ-secreting T-cells in response to immunogenic viral antigens, suggesting that the early post-UCBT period after ATG-containing conditioning is characterized not only by quantitative but also by qualitative/functional immunologic defect in adaptive immunity (Saliba et al. 2015).

Comparison of immune reconstitution patterns between adult patients undergoing dUCBT and patients undergoing HSCT from matched sibling donors (MSD) or HSCT from MUD after ATG-free MAC regimens has provided evidence that ALC is lower in dUCBT recipients at 1.5 months but reaches similar levels to those in MSD/MUD recipients by 3 months (Kanda et al. 2012). CD3+, CD4+, CD8+, CD4+CD25+CD62L+ (Treg), and CD4+CD45RA+CD62L+, which represent RTE counts, are significantly lower after dUCBT during the first 3–6 months, but at 1 year after transplantation, there is no difference in T-cell subsets between dUCBT and MSD/MUD recipients. As a result of the delayed T-cell recovery, dUCBT recipients have a significantly higher incidence of CMV reactivation and CMV disease, compared to MSD/MUD recipients. Notably, at 6 months after transplantation, an increase in TRECs was comparable in dUCBT and MSD/MRD recipients. Moreover, at 12 months, TCRβ repertoire was comparable between the two cohorts. A vigorous recovery of B and NK cells was detected in the dUCBT cohort in this study (Kanda et al. 2012), consistent with previous observations (Jacobson et al. 2012; Brown et al. 2010). UCBT is mainly reserved for patients who lack an MSD or MUD but may have mismatched unrelated donors (MMUD). Comparison between recipients of UCBT and 9/10 MMUD HSCT, after various conditioning regimens, showed largely similar kinetics of lymphocyte recovery and infectious complications (Servais et al. 2014).

Memorial Sloan Kettering Cancer Center (MSKCC) has reported their experience in adult patients after ATG-free dUCBT and RIC to high-intensity MAC (Sauter et al. 2011). Median ALC was within normal range by day 60. Median CD4+ T-cell count >200 cells/uL was reached by day 120, and both CD4+ and CD8+ median counts were at the lower limit of normal range by day 180. The infectious risk was heightened during the first 4 months, especially in the context of GvHD, but no infection-related deaths were seen beyond that time point. The use of deep sequencing technology to assess TCR diversity showed that dUCBT recipients have the highest CD4+ and CD8+ TCR diversity compared to recipients of conventional or T-cell-depleted (TCD) HSCT at 6 and 12 months after transplantation (van Heijst et al. 2013). This increased TCR diversity correlated with a substantially greater fraction of naive CD4+ T-cells. An important observation of the study is that diversity of CD4+ T-cells was 50 times higher than that of CD8+ T-cells in all patients. Notably, acute GvHD and steroid therapy did not restrict repertoire recovery, whereas CMV or EBV infections had a negative impact.

An important benefit of double UCB unit infusion is the recovery of thymopoiesis, as assessed by increasing TREC levels, by 6 months post-UCBT. This is in marked contrast to the thymopoietic failure seen in adult patients who receive single UCB unit infusion, especially after in vivo T-cell depletion by ATG (Komanduri et al. 2007). Early thymopoietic recovery is observed in adult dUCBT recipients both after ATG-containing RIC or MAC regimens (Politikos et al. 2015; Brown et al. 2010; Ballen et al. 2012). In both groups, more than half of the patients had detectable, yet below normal range, TRECs at 6 months. By 1 year, median TREC levels reached low normal values. Interestingly, circulating levels of SCF inversely correlate with TRECs at 6 and 12 months post-dUCBT, as well as CD4+CD45RA+ naïve T-cells at 1 year (Politikos et al. 2015). This outcome is likely due to consumption of SCF in the early intrathymic stages of T-cell development, consistent with the critical role of this cytokine in the thymus-dependent pathway of immune reconstitution.

6 Prognostic Significance of Immune Reconstitution for Clinical Outcomes After UCBT

As a result of the delayed neutrophil engraftment and the lack of adoptive immunoprotection due to the naivety of UCB lymphocytes, UCBT is associated with increased risk of infectious complications, which contribute greatly to the early transplant related mortality (TRM) after UCBT (Rubinstein et al. 1998; Laughlin et al. 2004; Rocha et al. 2004). Reactivation of latent viruses that require intact cellular immunity for their control is a major concern after UCBT. Among them, CMV is one of the most commonly reactivated viruses and is associated with significant morbidity and mortality (Szabolcs and Niedzwiecki 2007; Sauter et al. 2011; Dahi et al. 2015; Beck et al. 2010). The rate of CMV reactivation after UCBT ranges widely from 21 to 100% and the rate of CMV disease between 6 and 27% (Beck et al. 2010; Milano et al. 2011; Matsumura et al. 2007; Takami et al. 2005) depending on the characteristics of the study population, such as the number of seropositive patients at risk for reactivation (Albano et al. 2006; Mikulska et al. 2012); method of monitoring CMV reactivation, e.g., pp65 antigenemia vs. PCR viral load monitoring (Dahi et al. 2015; Milano et al. 2011); type of prophylaxis; and use of preemptive treatment (Milano et al. 2011). With the use of sensitive PCR assays essentially all seropositive UCBT recipients develop CMV viremia, although this can be mitigated with intensified antiviral prophylaxis (Dahi et al. 2015; Milano et al. 2011). Notably, the omission of in vivo T-cell depletion by ATG appears to provide protection against CMV reactivation or CMV disease after UCBT (Sauter et al. 2011; Dahi et al. 2015).

Similarly to CMV reactivation, the use of ATG is associated with increased risk of EBV reactivation and an incidence of post-transplant lymphoproliferative disorder (PTLD) as high as 21% (Cutler et al. 2011; Brunstein et al. 2006; Dumas et al. 2013). The risk can be greatly reduced by the avoidance of ATG, close monitoring of EBV viral load, and preemptive therapy with rituximab (Dumas et al. 2013; Blaes et al. 2010; Barker et al. 2001). An increased frequency of HHV-6 reactivation and related complications has also been observed after UCBT (Scheurer et al. 2013; Hill et al. 2012, 2015; Le Bourgeois et al. 2014), although the incidence of encephalitis appears to be low when ATG is omitted (Olson et al. 2014). Similarly, UCBT is associated with increased rate and severity of VZV disease (Vandenbosch et al. 2008), BK viremia, and increased risk of BKV disease (Rorije et al. 2014; Satyanarayana et al. 2015; Silva Lde et al. 2010), as well as higher risk for disseminated adenovirus infection (Robin et al. 2007).

The increased risk for infectious complications in UCBT recipients is most pronounced in the early posttransplant period, as studies have shown comparable infection-related mortality in recipients of UCBT and HSCT from adult donors beyond the first 3–4 months (Sauter et al. 2011; Laughlin et al. 2004; Parody et al. 2006; Barker et al. 2005a). The restoration of adaptive immunity via the thymus-dependent pathway, as well as the lower rates of GvHD, may account for this observation. Furthermore, dUCBT has been associated with enhanced GvL effect and low relapse risk (Ponce et al. 2011; Verneris et al. 2009; Brunstein et al. 2010; Rodrigues et al. 2009). As a result, long-term outcomes of UCBT are comparable to HSCT from conventional graft sources (Ponce et al. 2011).

The prognostic effect and the pathophysiologic link between different immune reconstitution metrics on the clinical outcomes of UCBT have been extensively investigated both in pediatric and adult patients undergoing UCBT. It has been determined that ALC > 200 × 106/L at day 30 after MAC and at day 42 after RIC is independently associated with superior 2-year overall survival (OS) and progression-free survival (PFS) (Burke et al. 2011). For the MAC cohort, ALC > 200 × 106/L at day 30 was also associated with lower number of infections prior to day 30 and lower TRM. However, no significant relationship between ALC and relapse has been observed. An ALC > 150 × 106/L on day 30 after dUCBT is also highly predictive of TRM and OS (Saliba et al. 2015). However, no difference between patients who did or did not develop viral reactivations or infections was identified. Because ALC offers only a crude assessment of immune reconstitution, other studies have focused on lymphocyte subsets to identify the most clinically relevant cell types for successful transplantation outcomes. Among such subsets, CD4+ T-cells are of great importance, and their impact on TRM and survival outcomes after allo-HCT from conventional donors (Berger et al. 2008) or sUCBT in children (Admiraal et al. 2015; Bartelink et al. 2013) has been established. Furthermore, in adult recipients of sUCBT, it has been observed that preservation of CCR7+ T-cells, which identify naïve and central memory T-cell subsets within the CD4+ T-cell compartment, has been associated with significantly longer survival (Komanduri et al. 2007).

The most objective measure of successful immune reconstitution is the ability of the new immune system to provide protection against infections and leukemia relapse. The presence of a proliferative response against herpes viruses has been associated with decreased risk of infection-related death, but is also associated with marked reduction in leukemia relapse, and improved PFS (Parkman et al. 2006). Detection of CMV-specific CD4+ and CD8+ T-cells by CFC in the early post-UCBT period (Komanduri et al. 2007; McGoldrick et al. 2013) suggests that UCB-derived T-cells are primed early after transplantation and can initiate a primary immune response against CMV. However, the development of antigen-specific T-cell responses does not correlate with the recovery of ALC and lymphocyte subsets (Saliba et al. 2015; Cohen et al. 2006), suggesting that quantitative T-cell recovery is not an optimal surrogate of functional immunity after transplantation. Moreover, CMV-specific T-cells detected early after UCBT fail to control viral reactivation in vivo (McGoldrick et al. 2013).

Our group has examined the immunological mechanisms involved in the restoration of successful anti-CMV immunity and control of CMV viremia in a cohort of dUCBT recipients (Brown et al. 2010). CMV-specific effectors were detected by IFN-γ ELISpot assay as early as 8 weeks posttransplant. In spite of the presence of CMV-specific effectors, clearance of CMV viremia was not achieved at this early time point after transplantation but was increasingly observed after 6 months. Notably, clearance of CMV was associated with the recovery of naïve CD4+CD45RA+ T-cells, a finding consistent with the need of CD4+ T-helper cells for the development of a functional CD8+ T-cell response (Shedlock and Shen 2003). Clearance of CMV viremia was associated with the emergence of thymopoiesis as assessed by TREC levels, and patients who attained normal TREC levels (>2000 copies/ug) were more likely to display absence of CMV viremia (Brown et al. 2010). These findings suggest a critical contribution of the thymopoiesis on CD4+ T-cell recovery, anti-CMV immunity, and in vivo control of the virus. Further analysis was performed to examine whether the development of CMV-specific responses and thymic reconstitution might be linked to improved long-term outcomes after dUCBT, which might correlate with restoration of immune responses against other pathogens or tumor antigens. CMV-specific immune responses were associated with improved PFS and OS. Furthermore, patients with TREC levels >2000 copies/ug DNA had superior OS compared with patients whose TRECs remained below 2000 copies/ug (Brown et al. 2010).

These observations were extended in a subsequent study, which assessed the role of thymic reconstitution on long-term outcomes in adult dUCBT recipients. In time-dependent regression analysis, higher TREC levels were associated with lower TRM and superior PFS and OS (Politikos et al. 2015). In contrast, high plasma levels of SCF, a cytokine critically involved in the early intrathymic T-cell development, inversely correlated with TREC levels and predicted for higher TRM and worse OS. No association of TREC levels with relapse was detected. This is in contrast to a study in pediatric recipients of UCBT or haploidentical HSCT, in which low levels of βTREC and sjTRECs before transplantation and at 3–6 months posttransplant were associated with increased incidence of relapse (Clave et al. 2013). These discrepancies between pediatric and adult recipients of UCBT might be related to an age-dependent effect on the thymus-dependent pathway of immune reconstitution and suggest that impaired thymic reconstitution might have a stronger negative prognostic value for the prediction of long-term clinical outcomes in pediatric than in adult UCBT recipients.

7 Expert Point of View

UCB is a valuable source of HSPC and has extended allograft availability to patients who lack suitable adult donors. UCB grafts contain a limited number of stem cells and lymphocytes with unique proliferative and functional properties compared to their adult counterparts. As a result of the UCB composition and transplantation practices, immune reconstitution following UCBT is characterized by the absence of transferred immune memory and prolonged T-cell lymphopenia, with compensatory marked expansion of NK and B-cells in the early posttransplant period. The omission of pharmacologic in vivo T-cell depletion from the conditioning regimen has led to faster quantitative recovery during the thymus-independent phase of immune reconstitution. However, increased infectious risk and TRM remain a concern early post-UCBT. Recovery of thymopoiesis is critical for the restoration of peripheral T-cell compartment and diversification of T-cell repertoire. The kinetics and efficiency of the thymus-dependent immune reconstitution after sUCBT in children and dUCBT in adults appear to be at least as efficient as that observed after allo-HCT from adult donors and provide excellent reconstitution of long-term T-cell immunity. Whether this is due to inherent properties of UCB-derived lymphoid progenitors or due to the lower rates and severity of GvHD, thereby diminishing the immunosuppression burden, remains unknown. Several measures of quantitative, thymus-specific, and pathogen-specific recovery correlate with infectious complications and long-term clinical outcomes of UCBT and have allowed the identification of patients at greatest risk for poor outcome after UCBT. Newer technologies, such as deep sequencing, may allow a global and more comprehensive assessment of the immune system. Future efforts should focus on the development of strategies to enhance the immune reconstitution process in order to improve outcomes of UCB transplantation.

8 Future Directions

Because immune reconstitution has a significant impact on morbidity and mortality at the early posttransplant stage but also on long-term clinical outcomes of UCBT, several approaches have been developed to shorten the period of immune deficiency and to improve reconstitution of adaptive immunity. The use of two UCB units in adults resulted in improved recovery of innate immunity, by enhancing neutrophil engraftment and reducing the risk of graft failure (Barker et al. 2005b; Ballen et al. 2007). Moreover, dUCBT may lead to faster thymic reconstitution compared to sUCBT (Komanduri et al. 2007; Kanda et al. 2012; Brown et al. 2010). This outcome might be related to a cell-dose effect or yet unidentified mechanisms, considering that the majority of patients display single-unit chimerism within 1–3 months post-dUCBT (Barker et al. 2005b) and long-term hematopoiesis and lymphopoiesis by a single UCB unit.

Several groups have investigated methods to reduce the duration of neutropenia after UCBT, either by ex vivo culture of the UCB graft to increase the stem cell dose or by ex vivo priming of UCB grafts to improve homing (Cutler et al. 2013; Delaney et al. 2010; de Lima et al. 2012; Wagner et al. 2016; Popat et al. 2015). Most of these strategies have been tested in the setting of dUCBT platform, in which patients receive one manipulated and one unmanipulated UCB unit. While these approaches have resulted in faster recovery of neutrophils derived from the ex vivo-manipulated UCB unit, long-term hematopoietic engraftment is provided by the unmanipulated unit. Moreover, a beneficial effect on reconstitution of adaptive immunity has not been observed to date.

Cytokine-based approaches to improve posttransplant immune reconstitution are being investigated in preclinical models or early clinical trials, albeit not yet in the setting of UCBT. KGF is a mitogen of TECs and has been shown to enhance thymopoiesis and peripheral T-cell reconstitution by protecting TECs from damage induced by radiation, chemotherapy, or GvHD (Min et al. 2002). A beneficial effect of KGF administration has been observed in murine experimental UCBT models (Wang et al. 2011). However, in human trials, administration of KGF has not shown a discernible benefit on immune recovery (Rizwan et al. 2011). IL-7 is a lymphopoietic cytokine and in mouse studies of allogeneic BMT it has been shown to promote both the thymus-dependent and thymus-independent pathways of T-cell regeneration, as well as B-cell recovery (Mackall et al. 2001). Administration of rhIL-7 has been tested in a phase I trial of patients undergoing T-cell-depleted HSCT and resulted in expansion of effector memory T-cells, increase of virus-specific T-cell responses, and broader TCR diversity (Perales et al. 2012). Sex steroid ablation with LHRH-A in murine BMT models is associated with increased numbers of LMPs in the bone marrow and enhanced peripheral T-cell recovery, perhaps by reversing thymic atrophy (Goldberg et al. 2009). In humans undergoing autologous or allogeneic transplantation, LHRH-A administration before transplant resulted in enhanced naïve and total CD4+ T-cell recovery, diversification of T-cell repertoire, and improved T-cell function (Sutherland et al. 2008). Other candidate cytokines include FLT3L (Fry et al. 2004), IL-15 (Alpdogan et al. 2005), IL-22 (Dudakov et al. 2012), SCF (Wils et al. 2011), and several combinations, which often display synergistic effects (Wils et al. 2012; Kelly et al. 2008). These approaches may form the basis of new clinical trials that might improve reconstitution of adoptive immunity and provide clinical benefit in UCBT recipients.

A different approach to overcome immunodeficiency due to delayed immune reconstitution after UCBT involves the adoptive transfer of cell products that can confer protection against pathogens. In that regard, multi-specific cytotoxic T lymphocytes (CTLs) recognizing multiple viruses including CMV, EBV, HHV-6, BK, and adenovirus can be generated from UCB T-cell populations (Papadopoulou et al. 2014). These multi-specific CTLs may be highly beneficial for the prevention or treatment of viral disease in UCBT recipients. The use of third party, partially HLA matched, in vitro-expanded EBV-specific CTLs have already been used with success in the treatment of post-UCBT EBV-associated PTLD (Barker et al. 2010). This “off-the-shelf” approach, which is currently being extended to the generation of CTL with specificity for other pathogens, might be particularly beneficial to UCBT recipients, who cannot rely on donor-derived CTLs for transfer of protective immunity. Adoptive transfer of UCB NK cells to prevent relapse is also being evaluated in early studies (Verneris and Miller 2009). As mentioned above, a major complication of UCBT is the prolonged impairment of thymic function. Recently, it was discovered that human progenitor T-cells, differentiated in vitro from UCB stem cells with the use of OP9-DL1 culture system, could be a source of thymus-seeding progenitors when infused in NSG mice. Importantly, a subset of in vitro-derived progenitor human T-cell populations could also restore thymic architecture and, when co-infused with HSPCs, were able to promote HSPC-derived T-cell lymphopoiesis (Awong et al. 2013). Together these ongoing research approaches are expected to form the basis for the development of new clinical trials. Such novel strategies to improve either the thymus-dependent or thymus-independent pathways of immune reconstitution may further enhance the outcomes of UCB transplantation.