Portal de Periódicos da CAPES

New insights into haematopoietic stem cell transplantation for patients with haemoglobinopathies

2004; Wiley; Volume: 125; Issue: 1 Linguagem: Inglês

10.1111/j.1365-2141.2004.04842.x

1365-2141

Franco Locatelli, Piero De Stefano,

Prenatal Screening and Diagnostics

Although improvements in conservative treatment have considerably ameliorated the prognosis of haemoglobinopathies, haematopoietic stem cell transplantation (HSCT) still remains the only curative treatment for patients with thalassemia and sickle cell disease (SCD). The majority of patients with thalassaemia given HSCT from a human leucocyte antigen (HLA)-identical sibling can be cured of their disease. Regarding patients with SCD, HSCT is reserved almost exclusively for children who have significant sickle-related morbidity. HSCT from an unrelated volunteer, carefully selected by high-resolution HLA-typing, can be an alternative for patients lacking a compatible family donor, as the results obtained using such donors are comparable with those obtained employing an HLA-identical sibling. Recently, it has been demonstrated that cord blood is as effective as, and possibly safer than, bone marrow for HSCT in paediatric patients. It is possible that, in the future, thalassaemia patients with poor health could benefit from HSCT employing non-myeloablative, less toxic, conditioning regimens. More than 20 years have elapsed since the first successful allogeneic transplant of haematopoietic stem cells (HSCT), performed in Seattle on a young Italian patient with thalassaemia major (Thomas et al, 1982). Since then, hundreds of patients with this disease and with sickle cell disease (SCD) have been cured of their original disorder after having received HSCT, in most cases using a human leucocyte antigen (HLA)-identical sibling as the donor (Johnson et al, 1984; Vermylen et al, 1988, 1998; Giardini & Lucarelli, 1999). All these transplants have enabled the identification of the variables that influence post-transplant outcome, thus permitting more refined prognostic counselling (Lucarelli et al, 1993, 1996; Walters et al, 1996). HSCT still remains the only curative treatment for patients with haemoglobinopathies and thus it represents a most attractive therapeutic alternative, especially for those patients, and their families, who do not accept the prospective of a lifelong disease that requires continuous treatment with blood transfusion and drugs. However, HSCT is associated with a non-negligible risk of both transplant-related mortality and morbidity and, especially in Western countries, only less than 25% of patients have a suitable compatible family donor. In view also of the relevant improvements achieved with conventional treatment, i.e. blood transfusion, inducers of γ-chain synthesis and chelation therapy, it is worthwhile to discuss why patients with haemoglobinopathies should be offered an allograft, how the number of patients that can be treated with HSCT can be increased and how we can render transplantation safer and more effective. Blood transfusion and iron chelation have dramatically improved both survival and quality of life of patients with thalassaemia over the last two decades and have changed a previously fatal disease with early death to a chronic, although progressive, disease compatible with prolonged survival (Piga et al, 1997; Borgna-Pignatti et al, 1998; Modell et al, 2000). In the developed world, the life expectancy of patients with thalassaemia has been reported to be between 25 and 55 years, mainly depending on compliance with medical treatment (Zurlo et al, 1989; Piga et al, 1997; Borgna-Pignatti et al, 1998). In fact, when compliance with chelation therapy is good and consistent, 90% of patients survive well into their 30s, whereas poor compliance results in fewer than 10% of patients reaching their 40th birthday (Piga et al, 1997). Despite the prolonged life expectancy, recently published data from the UK Thalassaemia Registry (Modell et al, 2000) showed a steady decline in survival starting from the second decade, with fewer than 50% of patients remaining alive during their 30s. New iron-chelating agents, which can be taken orally, allow better compliance with chelation (Cohen et al, 2000; Nisbet-Brown et al, 2003). If they will definitively prove to be as effective as deferrioxamine in the long-term, they should have a favourable impact on life expectancy and, possibly, they could further ameliorate also the quality of life. In the developing world, by contrast, most children with this disease die before the age of 20 years because of the unavailability of safe blood products in adequate amounts and/or of expensive iron-chelating drugs. In view of all these considerations the evaluation of the benefit/risk balance of what HSCT can offer profoundly differs for patients with thalassaemia living in different socio-economic contexts. As mentioned before, HSCT has the great appeal of being the only treatment able to definitively cure thalassaemia, although it is associated with a consistent risk of complications that can dramatically shorten the life duration of some patients. The risk of dying of transplant-related complications has been demonstrated to be mainly dependent on patient age, iron overload and liver viral infections. Adults, especially when affected by chronic active hepatitis, have a worse outcome than children (Lucarelli et al, 1999); among children, three classes of risk have been identified on the basis of regularity of previous iron chelation, liver enlargement and presence of portal fibrosis (Lucarelli et al, 1990). In fact, in paediatric patients who did not have liver disease and had received regular iron chelation (i.e. class 1 patients), the probabilities of survival, of rejection, and of event-free survival were reported to be 94%, 0% and 94%, respectively, whereas in patients who had low compliance with iron chelation and showed signs of severe liver damage (i.e. class 3 patients) the same probabilities were 61%, 16% and 53% (Lucarelli et al, 1990). By reducing the dosage of cyclophosphamide in the preparative regimen, it has been possible to document a lower incidence of transplant-related toxicity and mortality for children in class 3, which remained, however, in the order of 25% (Lucarelli et al, 1996). Other studies have confirmed that the majority of patients with thalassaemia given HSCT from an HLA-identical sibling can be cured of their disease and that both iron overload and older age unfavourably affect post-transplant outcome (Ghavamzadeh et al, 1998; Lawson et al, 2003). Moreover, it has been demonstrated that, after the allograft, through regular phlebotomy or even chelation therapy, it is possible to prevent or reverse organ damage due to pre-existing iron overload, particularly in the liver and heart (Angelucci et al, 1998; Mariotti et al, 1998; Muretto et al, 2002). Over the last three decades, many patients with SCD have benefited from advances in supportive and interventional therapies and this disease, rather than being a life-threatening condition in childhood, has become a chronic illness of adulthood, characterized, however, in many cases by a progressive accrual of significant, invalidating health problems (Platt et al, 1994; Wierenga et al, 2001; Walters et al, 2002). HSCT today is reserved almost exclusively for patients with SCD who have clinical features predictive of a poor outcome or significant sickle-related morbidity, such as history of stroke, magnetic resonance imaging of central nervous system (CNS) lesions associated with impaired neuropsychologic function, failure to respond to hydroxyurea with continuous vaso-occlusive crises, severe anaemia, osteonecrosis, or alloimmunization to erythrocyte antigens (Vermylen et al, 1988; Walters et al, 1996). The results of HSCT were best (with a probability of cure in the order of 80–90%) when performed in children who had an HLA-identical sibling (Vermylen et al, 1988; Walters et al, 1996), whereas the anecdotal reports of transplantation in adults suggest a poorer outcome (Hoppe & Walters, 2001). After transplantation, SCD patients with sustained donor engraftment do not have painful vaso-occlusive crises, splenic or hepatic sequestration, bone infarction, and benefit from protection from subsequent CNS vascular events. Disease recurrence remains a major problem to be overcome in SCD, as around 10% of patients have been reported to experience either early or late graft failure (Vermylen et al, 1988; Walters et al, 2001). Pretransplant exposure to blood products has been recognized to be a risk factor for graft rejection (Walters et al, 2001), suggesting that alloimmunization to blood products facilitates the generation of host immune competent cells able to interfere with donor engraftment. The most frequently observed long-term complications of HSCT include chronic graft-versus-host disease (GvHD), infertility and endocrine complications, as well as, in children, growth disturbances (Sanders et al, 1986; Locatelli et al, 1993; Giorgiani et al, 1995). Chronic GvHD has been reported to occur in 8–27% of patients given HSCT from a compatible relative (Gaziev et al, 1997; Ghavamzadeh et al, 1998; Lawson et al, 2003), whereas in patients with SCD chronic GvHD occurs in approximately 12% of subjects at risk (Sullivan et al, 2000). Although only a minority of patients with chronic GvHD have the extensive form of the disease, the quality of life of patients with extensive chronic GvHD may be certainly worse than that of patients with haemoglobinopathies treated with supportive therapy. As radiotherapy is the most important risk factor for development of growth impairment and endocrine sequels after HSCT (Sanders et al, 1986; Locatelli et al, 1993; Giorgiani et al, 1995), both these complications are less frequent in patients who, like those with haemoglobinopathies, are given chemotherapy in preparation for the allograft. No study has systematically compared the incidence of either growth failure or endocrine disturbances in thalassaemia patients treated with transfusions and chelation therapy with that observed in patients treated with HSCT. However, available data (Gaziev et al, 1993; Giorgiani et al, 1995) suggest that HSCT probably does not have more adverse endocrine consequences than conservative therapy, with the exception perhaps of patients who maintain an excellent compliance with iron chelation for the crucial years before and around puberty. Effects on fertility are more worrying as, despite anecdotal reports of successful pregnancy after transplantation (Borgna-Pignatti et al, 1996; Sanders et al, 1996), the vast majority of patients given a standard preparative regimen before HSCT become infertile (Locatelli et al, 1993), whereas a significant number of children born to women with thalassaemia have been reported in recent years (Jensen et al, 1995; Skordis et al, 1998). In conclusion, from the patient's point of view, HSCT still represents the only strategy able to guarantee a definitive cure for both thalassaemia and SCD, providing independence from both transfusions and iron chelation, as well as prevention/reversal of disease-induced organ damage. The best results can be obtained in younger patients with limited iron overload and without severe organ damage. These patients are the 'ideal' candidates for an allograft from a compatible sibling, whereas older, heavily iron-overloaded patients, especially when affected by liver complications for thalassaemia and CNS complications for SCD, run a considerable risk of transplant-related life-threatening complications. It is possible, however, that the latter patients, whenever the transplant is successful, are those who receive the maximum benefit from it, given the much reduced life expectancy associated, in thalassaemia, with continuing transfusions in the presence of reduced compliance with iron chelation and, in SCD, with the probability of fatal vascular accidents increasing over time. Looking at the option of HSCT for treating haemoglobinopathies from the point of view of a health system will show that this procedure is highly cost effective, at least for thalassaemia. The cost of medical treatment for this disease is considerable: recent calculations estimated the lifetime treatment cost in the UK at £803 002 (Karnon et al, 1999), whereas the cost of HSCT ranges between approximately £35 000 and £80 000, depending on the country in which it is performed. In view of this finding, there is no doubt that, for countries with limited economic resources, HSCT can be a convenient option, provided that the resources needed for performing the transplant are immediately available. Compared with adult haematopoietic stem cells, umbilical cord blood haematopoietic stem cells have many theoretical advantages for HSCT: they are enriched with in-vivo long-term repopulating stem cells (Wu et al, 1999; Gluckman & Locatelli, 2000), they can better restore host haematopoietic progenitor cell reservoir after transplantation (Frassoni et al, 2003), they produce larger haematopoietic colonies in vitro (Broxmeyer et al, 1992) are able to expand in in-vitro long-term culture and to engraft severe combined immunodeficient-human mice in the absence of additional human growth factors (Holyoake et al, 1999). Over the past decade, cord blood haematopoietic stem cells have been used increasingly, instead of bone marrow progenitors, for transplanting paediatric patients with either malignant or non-malignant disorders (Wagner et al, 1995; Gluckman et al, 1997). The main limitation of using cord blood for transplantation is that of the limited number of haematopoietic progenitors contained in a unit, which has confined its use mainly to patients with a body weight not exceeding 40 kg, whereas the main advantage, at least for transplants between relatives, is the low risk of GvHD (Gluckman et al, 1997; Gluckman & Locatelli, 2000). An analysis comparing children given cord blood transplantation (CBT) from a compatible sibling with HLA-identical sibling bone marrow transplant (BMT) recipients has formally documented that children transplanted with placental blood have a significantly lower relative risk of both grade II–IV acute and chronic GvHD (Rocha et al, 2000). The low incidence of both acute and chronic GvHD in CBT recipients has made this type of transplant attractive for patients with non-malignant disorders, who, in contrast with leukaemia patients, do not benefit from the graft-versus leukaemia effect associated with development of chronic GvHD. Moreover, this led to the development of a comprehensive infrastructure to make sibling cord blood banking more widely available (Reed et al, 2003). After a few anecdotal reports of successful CBT for patients with thalassaemia (Issaragrisil et al, 1995; Lau et al, 1998), a recent study from the Eurocord cooperative group, analysing the outcome of 44 patients with either thalassaemia or SCD transplanted with cord blood cells from a sibling, has reported that no patient died from transplant-related complications, suggesting that related CBT for haemoglobinopathies is a safe procedure (Locatelli et al, 2003). Eight out of these 44 patients did not have sustained engraftment; graft failure mainly occurred in patients with thalassaemia. Three of these eight patients obtained sustained engraftment after a subsequent BMT from the same donor. Moreover, patients who did not receive methotrexate as part of GvHD prophylaxis and who were treated with thiotepa during the preparative regimen had a probability of cure exceeding 90%, indicating that, in optimal conditions, CBT offers a probability of success at least as good as that of BMT recipients. As the incidence and severity of both acute and chronic GvHD was negligible, this study also supported the concept that reducing the risk of GvHD can also reduce the occurrence of lethal complications and increase HSCT applicability. The results of the Eurocord study and their comparative evaluation with respect to those of previously published reports on BMT recipients encourage the use of CBT for paediatric patients, whenever possible. In this regard, some bioethical thoughts on CBT for hereditary disorders can be made. Prenatal diagnosis of both thalassaemia and SCD is widely available for couples that already have an affected child, enabling a decision to be made regarding the completion or termination of a pregnancy with an affected foetus. The possibility of using cord blood cells from a compatible, healthy sibling for curing the affected child may lead to the request that, at the time of prenatal diagnosis, the HLA-compatibility between the affected child and the foetus be determined. In fact, conceiving a healthy and compatible foetus offers the couple the possibility of curing the affected child, making collection of placental blood mandatory. However, prenatal knowledge of HLA-incompatibility between the affected child and the foetus may also induce the parents to terminate the pregnancy even if tests indicate that the foetus is healthy, an attitude that is questionable from an ethical standpoint. Finally, the recent demonstration that in-vitro fertilization and preimplantation selection of compatible, healthy embryos is feasible (Verlinski et al, 2001) may encourage a couple with an affected child to initiate a pregnancy with the certainty that a source of stem cells will be available for transplantation (Burgio & Locatelli, 2000). This attitude of selecting the 'HLA compatible and not sick' embryo entails the choice of discarding other (equally not sick) embryos just because they are not HLA-compatible with the candidate recipient and it may raise concerns, at least for Catholic ethics (Burgio et al, 2003). On the other hand, in utero HLA typing performed early in the pregnancy offers the opportunity to discuss with the parents the recent report of Orofino et al (2003), where couples at risk of beta thalassaemia that have an affected embryo may reject termination of the pregnancy if the prospect of BMT from a family member (namely an HLA-identical living sibling) is available. Only 25–30% of patients with diseases that are potentially curable with HSCT have an HLA-compatible sibling. Thus, the vast majority of patients who can benefit from an allograft lack a suitable donor and, for many years, the problem of increasing the number of patients who can be treated with an HSCT has been unsolved. During the past 15 years, with the establishment of bone marrow donor registries that now include more than 8 million volunteers world-wide, many patients who need an allogeneic HSCT have been able to locate a suitable unrelated donor, and transplantation from unrelated volunteer donors has been increasingly adopted for haematological malignancies and life-threatening inborn errors (Kernan et al, 1993; Dini et al, 1996). In view of the high incidence of transplant-related immune complications (namely graft failure and GvHD) observed after HSCT from unrelated donors (Kernan et al, 1993; Dini et al, 1996; Gratwohl et al, 1998; Davies et al, 2000), recourse to this type of transplant for patients with thalassaemia or SCD did not meet consensus until recently, due to the unacceptable risk of transplant-related death. Recent reports, however, have demonstrated that more precise characterization of HLA alleles using high-resolution molecular typing for both classes I and II loci can reduce the risk of immune-mediated complications and fatal events (Petersdorf et al, 1998; Sasazuki et al, 1998). These results have provided the rationale for considering the possibility of performing HSCT from unrelated volunteers in patients with haemoglobinopathies. Some anecdotal reports have shown that unrelated donor HSCT can successfully cure patients with thalassaemia major (Contu et al, 1994; De Stefano et al, 1999). More recently, the experience of the Italian cooperative group for bone marrow transplantation has been described in detail (La Nasa et al, 2002). In this study, 32 patients aged 2–28 years were transplanted with bone marrow cells from an unrelated donor that was prospectively selected using high-resolution molecular typing for HLA-A, B, C, DRB1, DRB3, DRB4, DRB5, DQA1 and DQB1 loci. Four and 11 patients were assigned to risk class 1 and to risk class 2, respectively, whereas the remaining 17 patients either belonged to risk class 3 of the Pesaro classification or were adults. Twenty-two of these 32 patients (69%) were alive and transfusion independent after a median follow-up of 30 months, whereas four and six had graft failure or died due to transplant-related complications, respectively. All deaths but one occurred in the risk class 3 or adult group. The results of this study indicate that BMT from unrelated donors may offer results comparable with those obtained in transplants using HLA-identical family donors, particularly for patients with lesser iron overload, provided that stringent criteria of compatibility are employed for donor selection. The main limitation of this experience was that only approximately one-third of thalassaemia patients included in the search found a suitable donor. In order to further widen the applicability of HSCT, it is theoretically possible to select the donor according to criteria of lower resolution matching and to use some sort of in-vivo serotherapy, which has recently been proven to reduce the incidence of both acute and chronic GvHD in a dose-dependent manner, after the allograft from an unrelated donor (Bacigalupo et al, 2001). However, the quality of the results achievable with HSCT from unrelated donors selected using less stringent compatibility criteria is at present unknown. In the same way, the realization of T-cell depleted HSCT from an HLA-partially matched relative is not routinely advisable for patients with haemoglobinopathies, but it can be considered in extreme situations, such as that of a patient completely incompliant with any type of chelation therapy and/or with immunization to allogeneic or autologous erythrocyte antigens that renders transfusion either impossible or life-threatening. Although this type of transplantation is promptly available in almost all cases, and recent approaches based on the infusion of high number of CD34+ cells have documented that sustained donor engraftment can be achieved in the vast majority of patients (Aversa et al, 1998), a prolonged state of immune incompetence affects patients after transplantation and causes a remarkable risk of serious, often fatal, infectious complications. Stable mixed chimerism is not uncommon among patients transplanted for haemoglobinopathies from a compatible sibling (Andreani et al, 1996; Walters et al, 1996), as seen also in patients with acquired aplastic anaemia following BMT (Locatelli et al, 2000). Patients with haemoglobinopathies that develop stable chimerism, even with a low percentage of donor marrow progenitors (20–30%), have been reported to experience marked enrichment for donor cells in the mature red blood cell compartment, which makes them clinically asymptomatic and transfusion-independent (Andreani et al, 1996). The presence of residual host cells may also contribute to reduce the incidence and severity of GvHD, as, in animal models, mixed chimerism is associated with reduced susceptibility to GvHD, probably through mechanisms of central tolerance, with negative selection of host-reactive donor T cells (Sykes et al, 1988; Wekerle & Sykes, 1999). This observation has provided the rationale for using reduced-intensity preparative regimens in patients with thalassaemia or SCD. Recently, this novel transplant approach has been extensively investigated for the treatment of patients with malignancies who are not eligible for a standard allograft either because of age or poor medical condition (Giralt et al, 1997; McSweeney et al, 2001). Non-myeloablative conditioning regimens, mainly based on the use of purine-analogues like fludarabine (Giralt et al, 1997; Chan et al, 1999) can reduce or eliminate toxic effects associated with conventional HSCT and precluding a wider applicability of the procedure. Immunosuppression without myeloablation of the host can be sufficient to achieve donor chimerism, with engraftment of donor T and natural killer (NK) cells, which can provide a graft versus malignancy effect, demonstrated in both haematological and non-haematological tumours (Giralt et al, 1997; McSweeney et al, 2001; Pedrazzoli et al, 2002). Post-transplant donor lymphocyte infusion (DLI) may help eradicate host malignant cells, together with residual host haematopoiesis. In the context of non-malignant disorders, the use of a reduced intensity preparative regimen could enable the engraftment of donor cells, as well. Once the donor T and NK cells have engrafted, they can either completely eliminate residual host cells, thus leading to a state of complete chimerism, or favour the induction of a state of stable mixed chimerism, which, as mentioned before, may be sufficient to cure the patient. Thalassaemia patients with liver/heart damage or SCD patients who have suffered chronic vaso-occlusive organ damage could represent the ideal candidates for initial well-controlled, clinical trials of less toxic conditioning regimens at centres experienced in HSCT. So far, few reports have demonstrated the feasibility of using reduced-intensity preparative regimens for successfully treating patients with haemoglobinopathies (Slavin et al, 1998; Krishnamurti et al, 2001; Hongeng et al, 2002; Schleuning et al, 2002). The use of allogeneic non-myeloablative stem cell transplantation may also help reduce the incidence of late effects, particularly those concerning growth and fertility, that result from high-dose chemotherapy used for conventional HSCT. Should we consider this approach as the gold standard for transplanting patients with haemoglobinopathies for the future years? We believe that an affirmative answer is premature. Successful reduced-intensity HSCT has been obtained only in a limited number of patients and unfavourable outcomes of adults with SCD given a reduced-intensity preparative regimen have been also reported (van Besien et al, 2000). Graft failure is not a negligible problem even when a conventional preparative regimen is administered (Lucarelli et al, 1990, 1993; La Nasa et al, 2002; Locatelli et al, 2003) and the dose reduction of cyclophosphamide in the preparative regimen has been associated with increased rejection (particularly in patients who had received less than 100 red blood cell transfusions before transplant) (Lucarelli et al, 1996). Recently, seven patients, aged 3–20 years, with either SCD (six cases) or thalassaemia (one case), given HSCT after a preparative regimen consisting of fludarabine, low-dose (i.e. 200 cGy) total body irradiation and, in two cases, anti-thymocyte globulin, have been reported (Iannone et al, 2003). Although transplant-related toxicity was mild, none of these patients had sustained engraftment. These results indicate that, using non-myeloablative strategies, stable donor engraftment is more difficult to achieve in patients with haemoglobinopathies than in adults with malignancy and they do not encourage the use of this approach in young and otherwise healthy patients, who deserve more intensive myeloablative treatment. It is also reasonable to hypothesize that, to increase the chance of successful transplants, stem cell doses higher than those attainable with marrow harvests are needed. This means a preferential use of donor-mobilized peripheral blood haematopoietic stem cells, a practice which raises concerns when the donor is a minor, and because of the increased risk of chronic GvHD suggested by some studies (Burgio & Locatelli, 1997; Champlin et al, 2000). Moreover, the use of DLI for obtaining a stable chimerism may trigger the development of severe, life-threatening condition or invalidating GvHD. Thus, controlled trials of reduced intensity preparative regimens are warranted to assess the safety and efficacy of this approach for older patients or those with poor performance status and/or organ dysfunction. In conclusion, despite promising results of gene-therapy in animal models of haemoglobinopathies (Rivella et al, 2003), allogeneic HSCT still remains an important treatment option for children with thalassaemia and SCD and the only radically curative one. The use of cord blood haematopoietic stem cells can open new scenarios that can make the procedure safer, and careful selection of HLA-matched unrelated volunteers can increase the number of patients benefiting from this option. Transplant strategies based on the use of reduced-intensity preparative regimens and the induction of mixed chimerism could be explored to evaluate the possibility of rendering safer allogeneic HSCT also in adult or heavily iron-overloaded patients. This work has been partly supported by grants from AIRC (Associazione Italiana Ricerca sul Cancro), CNR (Consiglio Nazionale delle Ricerche), MURST (Ministero dell'Università e della Ricerca Scientifica e Tecnologica) and IRCCS (Istituto di Ricovero e Cura a Carattere Scientifico) Policlinico S. Matteo to F.L.