- Top of page
- Patients and methods
Transplantation of haematopoietic stem cells from human leucocyte antigen (HLA)-disparate parental donors presents a promising new approach for the treatment of patients lacking a HLA-matched donor. Success against major obstacles such as graft-versus-host disease (GvHD) and graft rejection has recently been demonstrated, so that immune reconstitution is one of the prime factors that determines the long-term prognosis following transplantation. Twenty children transplanted with megadoses of highly purified CD34+ haematopoietic stem cells after rigorous T-cell depletion were prospectively monitored for their immune reconstitution during the first post-transplant year. Natural killer (NK) cells showed a marked increase on d +30. T and B cells began to reconstitute on d +72 and +68 respectively. During extended follow-up, their numbers and proliferative capacity upon mitogen stimulation continually increased. Early reconstituting T cells were predominantly of a primed, activated phenotype with severely skewed T-cell receptor (TCR)-repertoire complexity. Naive T cells emerged 6 months post transplantation, paralleled by an increase in TCR-repertoire diversity. All patients self-maintained sufficient immunoglobulin levels after d +200. This study demonstrates that paediatric recipients of highly purified, haploidentical stem cells are able to reconstitute functioning T-, B- and NK-cell compartments within the first post-transplant year. This, together with the absence of significant GvHD, provides a strong indication for this approach to be considered in children who lack a HLA-matched donor.
Haematopoietic stem cell transplantation (SCT) offers the only curative approach for patients with advanced haematological malignancies. However, the need for an human leucocyte antigen (HLA)-identical donor has presented major limitations to this therapeutic modality, necessitating a search for alternative donors. Recent advances have revolutionized criteria for donor selection (Aversa et al, 1998), with the development of novel immunosuppressive drugs (Chen et al, 2000), approaches towards anergizing antigen-specific T cells (Blazar et al, 1998; Guinan et al, 1999), and techniques for in vitro and in vivo T-cell depletion (Aversa et al, 1994; Handgretinger et al, 1999). Together, these have pushed the limitations of donor selection beyond HLA barriers.
T-cell depletion affects both the incidence and severity of graft-versus-host disease (GvHD) after HLA-mismatched transplantation with a reduction in morbidity rates and even prevention of its onset (Anasetti, 1998; Martin et al, 1999). However, T-cell depletion of marrow grafts has been generally associated with higher rates of graft rejection (Anasetti et al, 1989) and delayed restoration of the immune system (Roux et al, 1996; Small et al, 1999), thus compromising patient response to post-transplant infectious complications.
‘Megadose’usage of highly purified peripheral blood stem cells obtained from a haploidentical (parental) donor with up to three antigen mismatches on different HLA loci is now one of the major bases for alternative donor selection in patients with advanced haematological malignancies (Reisner & Martelli, 1999). Haploidentical SCT was initially used in the treatment of paediatric patients with severe combined immunodeficiency (Friedrich et al, 1985). In immunocompetent recipients, transplantation of T cell-depleted grafts has been associated with higher rejection rates (Martin et al, 1999). Studies in murine models have shown that by increasing stem cell dosage, host alloresistance can be overcome with sustained engraftment of HLA-disparate stem cells (Wang et al, 1997; Reisner & Martelli, 2000). Additionally, doses of > 20 × 106 CD34+ cells have also been shown to promote lymphoid recovery (Handgretinger et al, 1999). Results from the first clinical trial using megadoses of peripheral blood stem cells in adult high-risk leukaemia patients demonstrated safe engraftment but a high mortality rate in the first post-transplant year, owing to infectious complications (Aversa et al, 1994). A second trial with a paediatric cohort confirmed safe engraftment, but with a surprisingly low incidence of infectious complications (Handgretinger et al, 1999). This may be as a result of a more efficient and faster restoration of the immune system in paediatric patients, but data on the pattern of immune reconstitution in children following haploidentical SCT remains scarce.
In the present study, we have carried out a prospective analysis examining the pattern of immune reconstitution in 20 paediatric patients during the first year after haploidentical SCT. Reconstitution has been monitored by: (i) phenotypic analysis of lymphocyte subsets; (ii) functional proliferative capacity of peripheral blood lymphocytes to mitogenic stimulation; and (iii) analysis of the T-cell receptor (TCR)-repertoire diversity using complementarity-determining region 3 (CDR3)-size spectratyping. All patients received megadoses of rigorously T cell-depleted, peripheral CD34+ stem cells after myeloablative therapy. Our data demonstrated that, in children, this innovative approach results in excellent engraftment rates, together with rapid and sustained reconstitution of T-, B- and natural killer (NK)-cell compartments and a complete absence of significant GvHD. These findings indicate that transplantation of haploidentical stem cells is an attractive option for high-risk paediatric patients who lack a conventional HLA-matched donor.
- Top of page
- Patients and methods
Immune reconstitution has been extensively studied following allogeneic bone marrow transplantation in both adult and paediatric patients (Lum, 1987; Foot et al, 1993; Godthelp et al, 1999), and after chemotherapy-induced myelosuppression (Mackall et al, 1995). In haploidentical SCT, however, three major factors may influence the pattern of immune reconstitution. First, HLA disparity of up to three major HLA-alleles between donor and host; second, the high number of stem cells infused to allow for stable engraftment; and third, the profound T-cell depletion by positive selection of donor peripheral stem cells to prevent GvHD.
Following reconstitution of granulocytes, NK cells were the first lymphoid cells to emerge during the first 4 weeks post transplantation. Recent evidence suggests that these early NK cells may play an important role in graft facilitation and graft-versus-leukaemia mediation without causing GvHD (Ruggeri et al, 1999). Thus, in the HLA-disparate transplant setting, NK cells may represent prime candidates for post-transplantation immunotherapy and are currently the subject of intense investigation.
Reconstitution of the T- and B-cell compartment was seen to occur after NK-cell reconstitution. B cells showed a continuous increase in numbers starting around d +60. On these early B cells, expression of CD10 represents an early differentiation marker. Functionality was demonstrated by a progressive increase in proliferative response to PWM and by the ability to self-maintain sufficient IgG levels after d +200.
T-cell reconstitution is the prime determinant in the successful outcome of SCT. Previous studies have shown two separate pathways of T-cell regeneration that contribute to the reconstitution of the T-cell compartment. First, mature T cells inoculated in the graft may, under antigenic stimulation, expand in the host (Mackall et al, 1996). The TCR diversity of these T cells is skewed and limited by the number of T cells in the initial inoculum. This pathway is generally referred to as peripheral expansion and predominates in the early phase following conventional bone marrow transplantation. This has been observed to be a persistent phenomenon in thymectomized hosts (Heitger et al, 1997) and in hosts with decreased thymic function, namely in adults or patients with intensive pretransplant chemotherapy (Mackall et al, 1995). Second, de novo maturation of naive T cells from bone marrow emigrants via passage through the thymus has been shown to occur after 6 months in children (Small et al, 1999). This pathway allows for de novo generation of T-cell responses and may play a part in correcting the initially skewed repertoire (Godthelp et al, 1999). Although it is a prerequisite for the restitutio ad integrum of the T-cell compartment, this pathway is limited to patients with residual thymic function and is therefore dependent upon a variety of factors, notably age and pretransplant chemotherapy (Mackall et al, 1997).
In our study, reconstitution of the T-cell compartment following SCT could be divided into two distinct phases. In the early phase, which was found to start as early as 2 months post transplant, expanded mature T cells of a predominantly CD45RO+ phenotype with limited TCR diversity were found to dominate the T-cell pool. As the minimal number of T cells in our spectratyping assays was > 5–10 times higher than that used by other investigators (Verfuerth et al, 2000), we argue that the limited TCR diversity observed in the early post-transplant phase is not caused by low numbers of T cells available for analysis. Furthermore, in titration experiments with decreasing numbers of T cells (range: 107−103 CD3+ cells) from healthy controls, we had not been able to detect skewing of the TCR repertoire when ≥ 104 T cells were used for the assay (data not shown). The early appearing T cells showed high expression of the activation marker HLA-DR and integrin subunits CD11a and CD29, but only a low capacity to proliferate upon stimulation with various mitogens. The majority of CD8+ T cells in this phase could be identified as memory cells. These early T cells were of donor origin, as demonstrated by chimaerism analysis of the magnetic-activated cell sorted (MACS) CD3+ T cells (data not shown). On the basis of their phenotype, these early T cells are presumably derived from graft-contaminating T cells that have expanded in the antigenic milieu of the host. It should be stressed that the number of co-transplanted T cells in this study was minimal (median 12·5 × 103 CD3+ cells/kg bw, range 1–117 × 103) and that 12 out of 20 patients had received OKT3 as part of their conditioning regimen for prevention of graft rejection. However, although improbable, it is still possible that these T cells may represent new T cells that have rapidly matured from the haematopoietic stem cell and, following antigenic stimulus in the host, acquired a memory phenotype (Murali Krishna & Ahmed, 2000). Although these mature, activated T cells were derived from a HLA-mismatched donor, they did not cause GvHD ≥ grade II and lacked a proliferative response to host cells upon in vitro rechallenge, but proliferated in response to third party cells (data not shown).
Donor leucocyte infusion (DLI) was given in 11 out of 20 patients. Although mixed chimaerism could be converted back to full donor chimaerism in 3 out of 11 patients (Bader et al, 1997), no significant difference in T-cell numbers on d +60, +120, +180 and +360 could be observed in patients with or without T-cell addback (Table III). This is in contrast to results from patients receiving DLI after partial T cell-depleted transplantations from related or unrelated donors (Small et al, 1999). It should be taken into account that the number of infused T cells was more than one log lower (2·5 × 104/kg bw) in our cohort. Despite the low T-cell dose used, one patient (UPN 4147) developed fatal GvHD IV after four T-cell addbacks. Thus, in the haploidentical setting, caution is warranted even with low numbers of unmanipulated donor T cells. The influence of DLI on TCR diversity is currently unknown. A recent study found a variable impact of DLI on T cell-repertoire diversity (Verfuerth et al, 2000). The authors reported a beneficial effect of DLI on TCR-repertoire diversity in only 3 out of 18 patients, whereas spectratypes became even more restricted post DLI in 6 out of 18 patients. In our study, the effect of DLI on TCR diversity was difficult to assess owing to the small number of patients and the fact that most DLIs were given before sufficient levels of peripheral T cells were detectable. Further studies, including comparison of sequence information of T-cell clones in donor and host, are needed to address this question.
A second wave of T-cell regeneration was found to start after 6 months resulting in T cells that expressed a naive phenotype (CD45RA+ for CD4+ or CD45RA+/CD27+ for CD8+ cells). This increase in the number of naive T cells encompassed both CD8+ and CD4+ fractions. A chronic deficiency of naive CD4+ T cells was not seen after haploidentical SCT in our paediatric cohort. These findings are in marked contrast to those from adult recipients of haploidentical grafts whose CD4+ counts remained below 100 and 200/μl for 10 and 16 months respectively (Aversa et al, 1998).
In parallel with the increase in naive T cells, the TCR-repertoire complexity and proliferative responses to mitogens improved. Although normal counts of peripheral T cells were not achieved by the end of the first post-transplant year, CDR3 size distribution of the TCR Vβ-chain had normalized. HLA-disparity with up to three HLA-mismatches did not seem to result in a major long-term skewing of the TCR repertoire. However, it is possible that minor alterations may exist that cannot be detected using primers spanning the Vβ–Cβ region. Further studies using high resolution PCR for analysis of the Vβ–Jβ junction are in progress.
The findings presented in this study have important implications for post-transplant immunotherapy. Caution is warranted using unmanipulated donor T cells for adoptive immunotherapy, as the threshold for induction of severe GvHD appears to be lower than in recipients of HLA-identical or partially mismatched stem cell grafts. Cellular strategies aimed at augmenting the graft-versus-leukaemia effect will potentially focus on the role of NK cells during the immediate post-transplant phase. Alternatively, vaccination-based approaches in the HLA-disparate setting may not be successful if applied before 6 months post transplant. An improvement during this 6-month period in the specific T-cell immunity may be achieved by approaches designed to increase the repertoire diversity of the peripheral T-cell pool. These strategies may include adoptive transfer of either ex vivo-generated T cells with low alloreactivity (Koh et al, 1999) or antigen-specific T cells (Reusser et al, 1997), or T cells that have been engineered to carry a suicide gene as a safety switch (Lal et al, 2000). The striking difference in reconstitution of naive T cells between children and adults (Aversa et al, 1998) after haploidentical SCT underlines the importance of thymic function for rapid and complete restoration of immune function (Weissman & Shizuru, 1999). Although residual thymic function seems to be present even in elderly patients (Douek et al, 1998), thymic output clearly needs improvement in stem cell recipients, particularly in adults. Thymic maturation may be enhanced by the administration of specific cytokines (Abdul-Hai et al, 1996). Alternatively, recent advances in transplantation of thymic tissue have raised new possibilities, providing this could be accomplished in tandem with SCT (Markert et al, 1999).
In summary, the data show that the transplantation of megadoses of haploidentical, haematopoietic stem cells after myeloablative therapy resulted in a rapid and sustained engraftment followed by a fast recovery of the NK-cell compartment around d +30. T and B cells started to reconstitute as early as d +70. The surprisingly low incidence of fatal infectious complications in our paediatric cohort indicates that the early emerging NK-cell compartment, together with the limited T-cell repertoire, seem to have sufficient functional capacity for controlling infections in a majority of patients. Results from this study are highly suggestive that haploidentical peripheral SCT should be considered in paediatric patients in urgent need of a transplantation, but who lack a conventional HLA-matched donor.