In the future, prenatal stem cell therapy will probably occupy a prominent role in fetal therapy.1 Whereas fetal surgical intervention remains limited to a few structural anomalies, prenatal stem cell therapy is a potential therapeutic approach for a large number of genetic disorders. Realization of this potential would expand fetal therapy far beyond its current focus of treating the compromised fetus. If there are biological advantages that favour prenatal over post-natal therapy, fetal therapy may become the preferred strategy for the treatment of many anticipated paediatric and adult diseases. This review will cover the rationale, current status and potential future applications of prenatal stem cell therapy.
Rationale for in utero stem cell transplantation
A stem cell can be defined as ‘a cell that can self-replicate and can give rise to more than one type of mature daughter cell’. In recent years, there have been many cell populations characterized as ‘stem cells’, many of which may ultimately be useful in fetal therapy. However, the best characterized stem cell and the first that will be applicable to fetal therapy is the haematopoietic stem cell (HSC). The HSC is a multipotent stem cell that maintains functional haematopoiesis by generation of all haematopoietic lineages throughout fetal and adult life. It can therefore be used to treat a broad range of haematopoietic disorders as has been demonstrated by the success of post-natal HSC transplantation (HSCT). However, in the absence of a matched donor, standard protocols for HSCT entail considerable morbidity and mortality. In utero haematopoietic stem cell transplantation (IUHCT) is a potential non-myeloablative alternative to HSCT for congenital haematological disorders that can be diagnosed early in gestation. Through advances in prenatal screening and molecular-based diagnostics, the opportunity for fetal intervention is greater than ever before and will undoubtedly continue to increase.
The rationale for IUHCT is based on unique events that occur during normal haematological and immunological development that favour the successful engraftment of transplanted allogeneic HSCs. The phenomenon of fetal immunological tolerance, first described by Billingham et al.,2 is perhaps the most important advantage of IUHCT over post-natal stem cell transplantation. The fetal thymic microenvironment plays a primary role in the determination of self-recognition and the repertoire of responses to foreign antigens. Pre-T cells undergo positive and negative selection during a series of maturational steps in the fetal thymus that are controlled by thymic stromal and dendritic cells.3 The end result is the deletion of T-cell clones with high-affinity recognition of self antigen, and preservation of a T-cell repertoire against foreign antigen. Therefore, introduction of allogeneic HSC prior to thymic processing should, in theory, result in presentation of donor antigen in the thymus with resultant life-long donor-specific tolerance.
Another biological opportunity unique to the fetus is the normal developmental migrations of HSCs to form haematopoietic compartments. Haematopoiesis starts in the yolk sac and aorto-gonado-mesonephric region, migrates to the fetal liver and finally resides in the bone marrow.4 Although the original theory was that development of new niches would facilitate engraftment of donor cells after IUHCT without the need for myeloablation, it is now recognized, as will be discussed below, that the fetal haematopoietic system is highly competitive with a relative excess of circulating HSCs.5,6 However, if the regulatory signals controlling the migrations of HSCs can be understood, it may be possible to manipulate them to favour the engraftment of donor cells.7 Finally, the very small size of a fetus allows the transplantation of much larger cell doses on a per kilogram basis than can be delivered after birth. In combination, these biological advantages that exist only in the fetus may provide the opportunity for engraftment and induction of associated donor-specific tolerance to allogeneic cells. The phenomenon of fetal tolerance can potentially eliminate the requirement for immunosuppression with its associated morbidity. The potential clinical impact of IUHCT is enormous when we consider the possibility that any disorder that can be prenatally diagnosed and can be treated by HSCT might be optimally treated by IUHCT.
Experimental results supporting the efficacy of in utero haematopoietic stem cell transplantation
The potential for fetal tolerance to facilitate clinical transplantation has been recognized since Billingham et al.'s original description of ‘actively acquired tolerance’ in 1953.2 Additional support for the concept was provided by observations in several species of haematopoietic chimerism and associated tolerance in dizygotic twins who share placental circulation. Finally, mechanistic insight into tolerance for self antigens (and by inference foreign antigen), and the central role of the thymus in this process, has been elucidated over the past two decades.3,8 Given these observations, the administration of allogeneic HSCs with appropriate timing to the pre-immune fetus should theoretically result in engraftment of donor cells and consistent donor-specific tolerance. In the ovine model, this appeared to be the case and this was the first model in which the potential therapeutic efficacy of IUHCT was investigated. Early gestational transplantation of allogeneic HSCs into normal sheep fetuses results in sustained multilineage haematopoietic chimerism.9 The fetal sheep model is also permissive to xenogeneic engraftment, as persistent, multilineage haematopoietic chimerism has been documented after transplantation of human-derived HSCs.10–12 In contrast to the ovine model, other normal animal models such as the primate, goat, canine and rat have shown much greater resistance to engraftment with significantly lower levels of chimerism (microchimerism) or a complete lack of engraftment. These less encouraging results, along with clinical failures by other groups,13 led us to develop the murine model of allogeneic IUHCT, and systematically analyse the requirements for engraftment and tolerance and to identify and overcome the barriers to engraftment after IUHCT.
Barriers to prenatal engraftment
It is clear from the preceding discussion that despite the unique opportunities offered by the fetal microenvironment, there are also unique challenges to overcome. These barriers to engraftment after IUHCT can best be understood in the context of two broad categories: (1) receptivity or competition of the host haematopoietic compartment; and ( 2) immunological barriers to engraftment.14
Host haematopoietic receptivity or competition
Perhaps the most important barrier to engraftment after IUHCT is host cell competition. In contrast to post-natal bone marrow transplantation in which myeloablation is used to condition the recipient prior to transplantation, IUHCT at the present time must be performed without any myeloablative preconditioning. Because of the developmental status of the fetus and the potential repercussions of any pharmacological toxicity on the fetus and mother, standard conditioning agents cannot be utilized. Therefore, the fetus has an intact and vigorous haematopoietic compartment to compete with the donor cells and the success of IUHCT will depend on the ability of donor HSCs to effectively engraft and then compete with host haematopoiesis. The concepts of receptivity (available space) and host cell competition are interlinked and therefore will be discussed together.
There is abundant experimental evidence that competition from the host haematopoietic compartment is a formidable barrier to successful engraftment. When donor cells have a competitive advantage, even the engraftment of a relatively limited number of cells can ultimately reconstitute the recipient. The high level of donor haematopoiesis achieved in c-kit-deficient mouse strains, in which there is a proliferative defect in host HSCs, is an extreme example. In this model, as few as one or two normal HSCs were shown to fully reconstitute the haematopoietic compartment after IUHCT.15,16 Studies of IUHCT performed in the mouse severe combined immunodeficiency (SCID) model also illustrate the importance of host cell competition.17,18 In this model, donor lymphoid cells have a survival and proliferative advantage. IUHCT results in complete reconstitution of the competitively deficient lymphoid compartment with minimal engraftment of other lineages, where the non-lymphoid progenitors maintain their competitive capacity. The converse is also true, i.e. when host cells are competitively superior, very minimal engraftment will be achieved. Although a relationship between donor cell dose and levels of engraftment clearly exists, transplantation of even massive doses of donor cells (2 × 1011 cells/kg) in a congenic strain combination (where the immune barrier is not a factor) results in average levels of chimerism of around 10%.19 Fortunately, donor-specific tolerance does not require high levels of chimerism.
Our studies suggest that the threshold for consistent induction of donor-specific tolerance is between 1% and 2% in the murine model, a clinically achievable range.20 However, that was not always the case. For many years the murine model of IUHCT was extremely difficult to engraft with donor cell engraftment of 0% to 1% at best (referred to as microchimerism).21,22 However, mixed haematopoietic chimerism across full MHC barriers with associated donor-specific tolerance is now observed in 100% of transplanted animals in our laboratory. This is partially attributable to overcoming the competitive barrier by transplantation of extremely high doses of donor HSC by the intravascular route and partially due to overcoming the immunological barrier described below. With the intravascular approach we can deliver approximately six times the number of HSCs that was possible with intraperitoneal transplantation directly into the intravascular compartment.23 This has resulted in average levels of chimerism in the allogeneic IUHCT model of over 25%. This progress and data discussed below from the preclinical canine model clearly demonstrate the ability to at least partially overcome the competitive barrier to IUHCT by intravascular delivery of very large doses of donor cells.
Other methods to achieve a competitive advantage for donor cells have also been investigated. For instance, we have investigated the use of CD26/dipeptidylpeptidase IV (CD26) blockade by preincubation of donor cells with Diprotin A (a CD26 blocker) to selectively improve homing of donor cells.7 CD26 cleaves dipeptides from the N-terminus of polypeptide chains that contain the N-terminal X-Pro or X-Ala motif.24,25 Among its many substrates, CD26 has been shown to have selectivity for stromal cell-derived factor 1α(SDF-1α)/CXCL12, a primary chemokine interaction involved in homing of HSCs to the fetal liver.26 We demonstrated that CD26 blockade increases the availability of homing receptors on HSCs and progenitors, and markedly improves homing and subsequent donor cell engraftment in the murine model of IUHCT. This represents only one strategy that would provide a selective advantage for donor cells. We have investigated other approaches including up-regulation of homing receptors by preincubation of donor cells in haematopoietic growth factors,27 and immunologically based methods that attempt to induce a graft-versus-host haematopoiesis effect without systemic graft-versus-host disease.28
Of course, the ultimate method to provide a donor cell advantage over host cells would be a highly selective and non-toxic approach to myeloablation in the fetus, but this ‘holy grail’ of IUHCT has not yet been developed.
Immunological barriers to engraftment
Although the theoretical basis for fetal tolerance appeared sound, experimental results from various laboratories were mixed on the ability of IUHCT to induce donor-specific tolerance. Our early mechanistic analysis of tolerance in chimeric mice supported a primary mechanism of deletion of donor-reactive lymphocytes, although deletion was not complete, implicating the presence of peripheral tolerance mechanisms as well.21,22 Thus, IUHCT appeared to result in ‘normal’ immunological processing of donor cells with high-level deletion of donor-reactive lymphocytes in the thymus. However, despite the accomplishment of easily measurable levels of engraftment in some animals and associated donor-specific tolerance by the documented mechanisms of clonal deletion, there were unexplained observations that suggested an additional barrier to engraftment beyond host cell competition. First, despite what appeared to be consistent delivery of donor cells, long-term donor chimerism occurred in only approximately one-third of recipients. Second, engraftment differed significantly between strain combinations. By performing early tracking of donor cells and longterm assessment of donor chimerism, we were able to document that 100% of allogeneic and congenic recipients maintained high levels of engraftment up to 3 weeks after IUHCT. However, between three and 5 weeks, 70% of allogeneic animals lost their engraftment, whereas 100% of congenic animals remained chimeric. The difference in the incidence of chimerism between congenic and allogeneic donors clearly supported the presence of an adaptive immune barrier to engraftment after IUHCT.19 We have now confirmed that there is an allospecific cellular and humoral response that is quantitatively higher in non-chimeric than in chimeric animals. This finding was inconsistent with our previous demonstration of long-term chimerism in some animals and the presence of deletional tolerance, and cast doubt on the validity of the primary rationale of IUHCT.
The pivotal observation that explains this contradiction, and should be considered in all subsequent studies of IUHCT, is our recent observation that the immune response is in reality a maternal immune response that is transferred to the neonate via maternal breastmilk.29 The proof of this was that if pups after IUHCT are fostered with a surrogate mother who has not been exposed to donor antigen, the frequency of chimerism remains 100%. We demonstrated unequivocally that the transfer of allospecific antibodies via maternal breastmilk leads to activation of an adaptive immune response in the pup with subsequent loss of donor chimerism. We also investigated why pups within the same litter could be chimeric or non-chimeric. The explanation relates to the self-reactive T cells that are known to escape thymic deletion in significant numbers as a result of inadequate or late presentation of antigen in the thymus, and to be controlled by regulatory mechanisms, including T-regulatory (T-reg) cell populations, which are essential for prevention of autoimmune disease.30,31 It is also known that maternal–fetal cell trafficking in humans results in the generation of tolerogenic fetal T-reg cells.32 This suggests that donor cells would induce T-reg cells in our chimeric pups and that these could potentially counteract a low-level alloimmune response. Therefore, we examined the level and suppressive capacity of CD4+CD25+ T-reg cells in chimeric compared with non-chimeric pups, and found that there does appear to be a more prominent T-reg response in the non-fostered chimeric pups.28,29 We speculate that in the absence of an overwhelming maternal response, it is the balance of immune-activating and regulatory response that determines whether or not a pup remains chimeric. Thus, the mechanisms of tolerance after IUHCT do appear to recapitulate the normal mechanisms of self-tolerance. The most important and central component is high level, but not complete, deletion of donor-reactive T-cells orchestrated by the fetal thymus. However, it is also essential that cells that escape thymic deletion and are donor cell reactive are suppressed in the periphery by an adequate T-reg response.
We feel that the most important finding in these studies is not the identification of the maternal immune response as the key factor in loss of chimerism, but rather the observation that in the absence of maternal influence allogeneic engraftment and long-term chimerism uniformly occur. This confirms the absence of an adaptive immune barrier in the preimmune fetus and validates the potential for practical application of ‘actively acquired tolerance’ to facilitate allogeneic cellular and/or organ transplantation. It raises the question of whether or not maternal immunization is an issue in large-animal models and clinical circumstances, and whether or not it is a limitation to engraftment after clinical IUHCT. It also raises the question of the importance of the innate immune system as a barrier to engraftment. Although natural killer (NK) cells have recently been implicated in loss of low-level engraftment after IUHCT,33 their effect appears to be lost in circumstances of absence of maternal influence. It is possible that the presence of maternal allospecific antibody can directly activate NK cells via the mechanism of antibody-dependent cell-mediated cytotoxicity. These are important questions but, in any case, an obvious strategy to avoid any potential maternally derived immune barrier would be the use of maternal donor cells when appropriate.
The virtues of tolerance
The achievement of donor-specific tolerance allowed us to perform proof-in-principle studies of the promising clinical strategy of prenatal tolerance induction by IUHCT followed by post-natal non-toxic bone marrow transplantation (BMT) to increase low levels of chimerism to levels that would be therapeutic for diseases such as the haemoglobinopathies. The value of this strategy is that it lowers the threshold of chimerism that must be obtained by IUHCT to proceed with clinical application. It is well established that HSCT from a syngeneic or identical twin donor requires a very minimal conditioning regimen (no immunosuppression and minimal myelosuppression) to achieve engraftment.34,35 As IUHCT theoretically produces, from the perspective of the immune system, a perfectly matched donor, this strategy should provide high-level engraftment with very minimal or no toxicity. We have demonstrated three different non-toxic post-natal strategies to be effective in the murine model after IUHCT: (1) preparative low-dose total-body irradiation followed by T-cell-depleted BMT;36 (2) post-natal donor-specific lymphocyte infusion (DLI) without BMT37; and (3) low-dose busulfan as a single-agent preparative regimen, followed by T-cell-depleted BMT.20 In each study, complete or near-complete replacement of host haematopoiesis by donor cells was achieved, essentially without toxicity or graft-versus-host disease (GVHD). These studies form the basis for what we believe will be the first successful clinical strategy for application of IUHCT to competitive haematological disorders.
Of course, the murine model may not be representative of what will occur clinically and a better preclinical model was needed. With the exception of sheep, there has been very limited success after IUHCT in large-animal models, although recently that has begun to change. Successful achievement of measurable multilineage chimerism after IUHCT with associated donor-specific tolerance for swine leucocyte antigen (SLA)-matched kidney transplants has been demonstrated in limited experiments in the SLA inbred pig model.38,39 In order to validate our success in the murine model, we have developed the canine model of IUHCT. The canine model has been used extensively in the preclinical testing of post-natal HSCT regimens and has been a reliable predictor of clinical results40,41 For instance, many strategies to prevent or treat GVHD were first evaluated in the canine model prior to their use in humans.40–46 The canine model also offers biological and practical advantages specific to the evaluation of IUHCT. The ontogeny of the canine immune system appears relatively similar to that of humans,47 and from a technical perspective, the canine pregnancy allows ultrasound-guided injection of pups prior to immune maturation. Finally, the canine model offers the advantage of the availability of disease models that are analogous to human disorders.
We initiated our canine studies using dogs that have the canine analogue of human leucocyte adhesion deficiency [canine leucocyte adhesion deficiency (CLAD)]. CLAD-affected dogs have a severe immunodeficiency that results in death prior to 6 months of age, whereas the CLAD carrier is phenotypically normal. Neither the affected nor carrier dogs have a significant competitive defect in the HSC compartment or in any of the haematopoietic lineages.48 Therefore, the CLAD model should be representative of the degree of host cell competition expected for most target diseases. Historically, the canine model has been difficult to engraft by IUHCT, supporting the competitive capacity of the fetal haematopoietic system.49 In our initial studies in the canine model, we demonstrated that low-level chimerism can be achieved by IUHCT and that these levels of chimerism can (1) ameliorate or cure the clinical phenotype of CLAD and (2) result in associated donor-specific tolerance in some animals that is adequate to facilitate post-natal enhancement of chimerism to potentially therapeutic levels using the single-agent, low-dose busulfanconditioning regimen, followed by transplantation of T-cell-depleted bone marrow from the same donor.50 In this study, we saw no significant toxicity and no GVHD. We have recently developed techniques that result in engraftment of > 2% in over 90% of recipient dogs after IUHCT with an average level of engraftment of around 8%. These levels are well within the range of chimerism required to induce donor-specific tolerance and confirm that the barriers to engraftment can be overcome in a large-animal, clinically relevant model (Flake AW, manuscript in preparation). We are encouraged that the results of IUHCT in the canine model appear remarkably similar to our results in the murine model, suggesting that our results in the murine model can be translated to clinical application.
Experience with clinical application of in utero haematopoietic stem cell transplantation
There have been approximately 50 reported cases of IUHCT in humans over the past 20 years.51 Drawing conclusions based on clinical experience in humans has been difficult because of a large variety of target diseases, donor cell sources and transplantation protocols. Not surprisingly, success has largely been limited to cases of immunodeficiency syndromes in which donor cells have a clear selective advantage over host cells, i.e. X-linked SCID (XSCID). In utero therapy for XSCID has been successful, with at least 10 documented cases of cellular reconstitution with functional T cells.52–57 However, recipients manifest split chimerism, with only the T-cell compartment engrafted, similar to the results of non-myeloablative post-natal HSCT. Thus far, there is no proven advantage for prenatal treatment of XSCID over neonatal transplantation but not enough cases have been performed to enable meaningful comparisons.58 Attempts to treat other immunodeficiency disorders, such as chronic granulomatous disease (CGD) or Chediak–Higashi syndrome, have been unsuccessful thus far as all subjects were born without detectable engraftment.59–61 The use of IUHCT for haemoglobinopathies has also been attempted, but has thus far been largely unsuccessful. There have been 12 attempts to treat β-thalassaemia in utero, with only two investigators reporting detectable post-natal engraftment,62–64 at least one of whom subsequently lost engraftment.54 There have been three reported attempts to treat ⟨-thalassaemia by IUHCT, with one patient exhibiting microchimerism and tolerance to donor antigen by mixed lymphocyte reaction;65 however, all three patients remained transfusion dependent. There have also been three reported attempts to treat sickle cell anaemia; however, none has resulted in detectable engraftment.66,67 There have been seven reported attempts to treat metabolic storage diseases by IUHCT, with two reports of engraftment,53,68 one of which led to no clinical improvement and the other resulted in prenatal death, probably because of GVHD. Given this history, few recent attempts at IUHCT have been reported and many investigators have been discouraged. However, the rationale remains compelling and there are lessons to be learned from this experience that may help guide future efforts. Many of the historic attempts were ill-advised for reasons that are now recognized. Many of the transplants were performed too late in gestation, or with donor cell sources that would not be expected to succeed. For instance, the use of highly enriched HSCs as a donor source has been unsuccessful in allogeneic experimental systems and has also been clinically unsuccessful. In addition, the expectation that one could achieve therapeutic levels of engraftment after conventional IUHCT alone for diseases such as the haemoglobinopathies was somewhat naive given what we now understand about the barriers to engraftment. However, given the experimental progress described above we feel that it is now time to revisit clinical application of IUHCT.
Considerations for future clinical application of in utero haematopoietic stem cell transplantation
Based on data from the combined murine and preclinical canine models of IUHCT we now have the expectation that clinical application of IUHCT can be successful either alone or in combination with a post-natal minimally conditioned same-donor transplant. However, it is essential that future trials of IUHCT be performed in centres with a vested interest in this therapy and preferably proven success in a preclinical animal model. Our data support an optimized protocol that includes the use of maternal cells and intravascular administration. Based on our current understanding of human immune and haematopoietic ontogeny, the ideal timing of at least the first IUHCT would be at 13 to 14 weeks' gestation. During this time the fetal liver is the primary haematopoietic organ and thymic selection is ongoing with very few mature lymphocytes present in the thymus or peripheral circulation. Also, at this time the fetus is very small, < 35 g in weight, allowing the opportunity to maximize the dose of donor cells. At the present time, the only disorders that this strategy can be contemplated for are disorders that offer either a competitive advantage for donor cells, or perhaps disorders that require only minimal levels of engraftment for therapeutic success. At the present time, there are two clinical strategies that may be successful in clinical application. The first is IUHCT alone, which may be successful for selected biologically favourable target disorders. The second is IUHCT for donor-specific tolerance induction followed by post-natal minimally conditioned HSCT from the same donor. The latter approach holds the most immediate promise for broad clinical application of IUHCT because it requires only a minimal level of chimerism to be successful, and as it would be applicable to the majority of disorders that can be prenatally diagnosed and treated by post-natal IUHCT.
Favourable disease targets for in utero haematopoietic stem cell transplantation alone
Clearly, the most biologically favourable disease for treatment by IUHCT alone remains XSCID. Other characterized mutations in cytokine receptor signalling pathways [i.e. Janus kinase 3 (Jak3) or zeta-associated protein 70 (ZAP-70)] resulting in SCID should also be favourable candidate diseases for IUHCT. Based on the available clinical and experimental evidence, it is likely that any member of this group of disorders can be effectively treated by IUHCT alone as even minimal levels of engraftment should provide adequate T- and B-cell function to provide immune protection. If B-cell engraftment occurs, then there would be a strong rationale favouring IUHCT over the current standard for comparison of neonatal non-myeloablative haploidentical transplantation in which only T-cell engraftment occurs and the children must be supported with supplemental IgG.69 Another group of diseases that could benefit from IUHCT alone are those in which somatic mosaicism and in vivo selection have been documented to occur. In these diseases there is presumably a survival advantage for the spontaneously corrected cells.70 Such correction has been noted in adenosine deaminase SCID,71 Fanconi anaemia72 and Bloom syndrome,73 the last two of which are chromosomal breakage syndromes. In both Fanconi anaemia and Bloom syndrome mitotic recombination was documented as the molecular mechanism of somatic reversion. This represents an experiment of nature documenting the improvement in a disease by clonal expansion of a single spontaneously corrected HSC and suggests that even low-level engraftment achieved by IUHCT could eventually replace host haematopoiesis as progressive bone marrow failure occurred. True clinical cure of either disease is unlikely, as they are associated with other pleiotropic manifestations, such as an increased rate of malignancy, that are unlikely to be reversed by haematopoietic reconstitution alone. Other diseases that are known to be treatable with low levels of chimerism include CGD, hyper-IgM syndrome and LAD. It has been well documented that CGD can be corrected by as few as 5% normal neutrophils,74 and in X-linked hyper-IgM syndrome, phenotypically normal carriers have been identified in whom the normal gene has been predominantly silenced.75 LAD results from mutations in the leucocyte integrin CD18, which inhibits the expression of the CD18/CD11 complex on the cell surface and thus the ability of leucocytes to adhere to the vessel wall and migrate to sites of infection.76 Recent studies in the analogous CLAD model have demonstrated that even low levels of donor CD18+ cell engraftment following non-myeloablative matched littermate BMT can reverse the lethal disease phenotype in CLAD.48,77 We have recently demonstrated correction of the CLAD phenotype by IUHCT of haploidentical adult BM-derived cells in the canine model.50 Specific non-haematopoietic disorders of bone metabolism may also be attractive target disorders for IUHCT. A recent report of rescue of osteopetrotic mice with the same mutation as approximately half of human patients with the autosomal recessive disease by IUHCT is intriguing.78 In this study, complete phenotypic correction associated with osteoclast engraftment was achieved, despite the fact that an abundance of host osteoclasts remained present that were not functional. There is also interest in treatment of osteogenesis imperfecta by prenatal replacement of mesenchymal stem cells or stromal progenitor cells79 and clinical cases have been reported with somewhat promising results.80 The experimental basis for application of IUHCT towards this disease, however, needs further development. For all of these disorders, even if curative levels of engraftment are not achieved, levels adequate for donor-specific tolerance induction would allow conversion of the patient to the second strategy described below.
In utero haematopoietic stem cell transplantation for donor-specific tolerance induction followed by post-natal minimal conditioning same-donor haematopoietic stem cell transplantation
As discussed earlier, low levels of mixed haematopoietic chimerism after IUHCT are associated with donor-specific tolerance. The exact level of chimerism required may vary slightly with species, but is in the 1–2% range. In the murine system, we have observed that in the presence of microchimerism (< 0.5%, donor cells detectable only by amplification by polymerase chain reaction) approximately onethird of animals are tolerant of donor skin grafts and non-reactive by mixed lymphocyte reaction.21,22 In a separate study, using the ability to enhance chimerism after IUHCT by minimal conditioning post-natal IUHCT from the same donor strain as the definition of tolerance, only 60% of animals with flow cytometrically detectable chimerism of <7thinsp;1% were tolerant, whereas 100% of animals with chimerism of > 1% were tolerant.20 Given the relatively high levels of chimerism achieved in the canine model, we feel that there is a very high likelihood that this strategy can be successful. If so, it would be applicable to all disorders that are currently treatable by HSCT and that can be diagnosed prenatally. The most important of these are the haemoglobinopathies, i.e. sickle cell disease and β-thalassaemia. These are two of the most common genetic disorders in the world and there is at present no satisfactory post-natal treatment. They are favourable disorders to treat by IUHCT from the perspective of their post-natal biology, in that levels of mixed haematopoietic chimerism of 15–25% phenotypically correct the disease. The short half-life of diseased red cells in both disorders allows amplification of myelolymphoid engraftment in the bone marrow in the circulating red cell compartment.81 However, both disorders have normal prenatal haematopoiesis and are highly competitive targets for IUHCT. Thus it is likely that a two-step approach will be required. We plan to initiate a clinical trial for SCD at our centre in the near future.