Pediatric Annals

Bone Marrow Transplants in Patients Lacking an HLA-Matched Sibling Donor

Abstract

During the 22-year period since the first successful application of a human leukocyte antigen (HLA) 'Compatible bone marrow graft for the treatment of severe combined immunodeficiency,1 allogeneic marrow transplants from HLA-matched donors have evolved from an experimental approach used only for patients with lethal refractory disorders of hematopoiesis to an established treatment of choice for acute leukemias that have relapsed early in first remission, chronic myelogenous leukemia, aplastic anemia, and a series of genetic disorders of lymphoid development, hematopoiesis, and metabolism.2'3 As recognition of the curative potential of allogeneic marrow transplants has grown, the need to develop approaches that would permit transplants of normal marrow in the 60% to 70% of patients who lack an HLA-matched sibling has increased markedly. Several approaches to such patients are actively being explored, particularly the use of unmodified marrow grafts from partially matched related or HLAcompatible unrelated donors and the use of HLA nonidentical marrow grafts sufficiently depleted of T-lympHocytes so that they can be administered without the risk of lethal graft-versus-host disease (GVHD). Each of these approaches has shown considerable promise, but each is also associated with increased risks of either graft rejection or severe GVHD. This article describes the progress made with each approach, defines the limitations that are currently recognized, and presents approaches currently being explored to circumvent these limitations.

IDENTIFYING HISTOCOMPATIBLE DONORS FOR PATIENTS LACKING AN HLA-IDENTICAL SIBLING

Definition of the H-2 system, the major histocompatibility complex in mice, and recognition of the importance of matching for genetic determinants within H-2 for the engraftment of a marrow transplant and the prevention of lethal graft versus host reactions, coupled with the definition of an analogous system, the HLA system in man, led to the first successful applications of HLA-compatible marrow grafts for the treatment of lethal genetic immunodeficiencies in children.1·4 The HLA-gene complex (Figure 1 ) is located on the short arm of chromosome 6 and is composed of a series of highly polymorphic genes encoding proteins differentially expressed on the surface of nucleated cells.5 The HLA class I genes encode proteins (HLA-A, HLA-B, and HLA-C) that are expressed on the surface of all nucleated cells. In contrast, HLA class Il genes (HLA-Dt; HLA-Dq, and HLA-Dp) encode surface proteins on a more limited group of cell types including macrophages and monocytes, endothelial cells, B-cells, activated T-cells, and early hematopoietic progenitors. Ine HLA class III genes differ from class I and class II genes in that they encode secreted proteins such as the complement components C2 and C4, 21-hydroxylases, tumor necrosis factor, and the heat shock proteins.6

Human leukocyte antigen genes are expressed codominantly, permitting identification of haplotypes inherited from each parent by serologie typing of the HLA-A1 HLA-B, HLA-C, and HLA-Dr determinants. Compatibility for determinants within the entire HLA-D region can be ascertained by the demonstration of mutual unresponsiveness between a donor and recipient in mixed lymphocyte cultures (MLC). Specialized functional tests or molecular hybridization techniques are used to identify HLA-Dq and Dp alíeles. Normally, the inheritance of HLAhaplotypes follows mendelian genetics, so the likelihood that any two siblings will receive the same parental haplotypes is one in four. With increasing family size, the prospects for identifying a donor increase. Currently, matched siblings are identified for no more than 35% to 40% of patients.

The genes within the HLA complex exhibit an extraordinary degree of polymorphism.5 As illustrated in Figure 1, there are 27 serologically identified HLA-B alíeles alone. Molecular analyses further indicate that the alleltc polymorphism detected by serology is markedly incomplete because several molecular variants of certain class I and class II alíeles can be distinguished by isoelectric focusing,…

During the 22-year period since the first successful application of a human leukocyte antigen (HLA) 'Compatible bone marrow graft for the treatment of severe combined immunodeficiency,1 allogeneic marrow transplants from HLA-matched donors have evolved from an experimental approach used only for patients with lethal refractory disorders of hematopoiesis to an established treatment of choice for acute leukemias that have relapsed early in first remission, chronic myelogenous leukemia, aplastic anemia, and a series of genetic disorders of lymphoid development, hematopoiesis, and metabolism.2'3 As recognition of the curative potential of allogeneic marrow transplants has grown, the need to develop approaches that would permit transplants of normal marrow in the 60% to 70% of patients who lack an HLA-matched sibling has increased markedly. Several approaches to such patients are actively being explored, particularly the use of unmodified marrow grafts from partially matched related or HLAcompatible unrelated donors and the use of HLA nonidentical marrow grafts sufficiently depleted of T-lympHocytes so that they can be administered without the risk of lethal graft-versus-host disease (GVHD). Each of these approaches has shown considerable promise, but each is also associated with increased risks of either graft rejection or severe GVHD. This article describes the progress made with each approach, defines the limitations that are currently recognized, and presents approaches currently being explored to circumvent these limitations.

IDENTIFYING HISTOCOMPATIBLE DONORS FOR PATIENTS LACKING AN HLA-IDENTICAL SIBLING

Definition of the H-2 system, the major histocompatibility complex in mice, and recognition of the importance of matching for genetic determinants within H-2 for the engraftment of a marrow transplant and the prevention of lethal graft versus host reactions, coupled with the definition of an analogous system, the HLA system in man, led to the first successful applications of HLA-compatible marrow grafts for the treatment of lethal genetic immunodeficiencies in children.1·4 The HLA-gene complex (Figure 1 ) is located on the short arm of chromosome 6 and is composed of a series of highly polymorphic genes encoding proteins differentially expressed on the surface of nucleated cells.5 The HLA class I genes encode proteins (HLA-A, HLA-B, and HLA-C) that are expressed on the surface of all nucleated cells. In contrast, HLA class Il genes (HLA-Dt; HLA-Dq, and HLA-Dp) encode surface proteins on a more limited group of cell types including macrophages and monocytes, endothelial cells, B-cells, activated T-cells, and early hematopoietic progenitors. Ine HLA class III genes differ from class I and class II genes in that they encode secreted proteins such as the complement components C2 and C4, 21-hydroxylases, tumor necrosis factor, and the heat shock proteins.6

Human leukocyte antigen genes are expressed codominantly, permitting identification of haplotypes inherited from each parent by serologie typing of the HLA-A1 HLA-B, HLA-C, and HLA-Dr determinants. Compatibility for determinants within the entire HLA-D region can be ascertained by the demonstration of mutual unresponsiveness between a donor and recipient in mixed lymphocyte cultures (MLC). Specialized functional tests or molecular hybridization techniques are used to identify HLA-Dq and Dp alíeles. Normally, the inheritance of HLAhaplotypes follows mendelian genetics, so the likelihood that any two siblings will receive the same parental haplotypes is one in four. With increasing family size, the prospects for identifying a donor increase. Currently, matched siblings are identified for no more than 35% to 40% of patients.

The genes within the HLA complex exhibit an extraordinary degree of polymorphism.5 As illustrated in Figure 1, there are 27 serologically identified HLA-B alíeles alone. Molecular analyses further indicate that the alleltc polymorphism detected by serology is markedly incomplete because several molecular variants of certain class I and class II alíeles can be distinguished by isoelectric focusing, restriction fragment length polymorphism, or hybridization with sequence specific oligonucleotide probes.7·8 Given the large number of different alíeles that exist for each of the class I and class II genes of the HLA system and the assumption that the genes within the HLA region have been recombined at random over the course of evolution, it could be expected that the identification of a suitably compatible relative other than an HLA-matched sibling would be a rare event. In fact, however, HLA phenotypically matched or partially matched related donors have been identified for as many as 5% to 10% of patients. This is due, in part, to the high frequency of several HLA class I and class II alleles and, in some cases, to the co-association of certain HLA-A, HLA-B, and HLA-D alíeles to form specific haplotypes in certain ethnogeographic groups - a phenomenon known as genetic linkage dysequilibrium.9

For patients with certain lethal genetic disorders of hematopoiesis known to be common in specified ethnogeographic groups, such as Fanconi's anemia, Kostmann's agranulocytosis, and certain forms of severe combined immunodeficiencies, unrelated parents often share one or more HLA-alleles that are inherited on one of the patient's HLA- haplotypes. In such circumstances, a parent may be identified who is HLA-compatible with a given patient. This situation may also be observed for patients with acquired blood diseases when both parents originate from a relatively restricted ethnogeographic background. Related donors also may be identified at significant frequency for patients who inherit an HLA-A, HLA-B, or HLA-Dr haplotype known to be in genetic linkage dysequilibrium. In such cases, a search is made among the offspring of relatives who share the uncommon or nonlinked haplotype with the patient. Through searches of this kind, a cousin often can be found who derives the common haplotype from a totally different pedigree together with the uncommon haplotype shared by the patient.

These two types of inheritance are illustrated in Figure 2. In the pedigrees illustrated, the haplotype HLA-Al, HLA-B8, and HLA-Dr3 inherited by the patient is representative of a haplotype containing three HLA alíeles in strong genetic dysequilibrium in northern European Caucasian populations. Furthermore, the alíeles within this haplotype, HLA-Al and HLA-B8 are, of themselves, closely linked as are HLA-B8 and HLA-Dr3. In the two pedigrees presented, the father differs from the patient at a single HLA-A alíele on one haplotype, while the maternal uncle is HLA phenotypically matched for HLA-Al, HLA-B8, and HLA-Dr3 and HLA-genotypically matched for HLA-A2, HLA-B35, and HLA-DrI. The patient could receive a transplant from either donor that would be associated with an increased but acceptable risk of graft rejection or severe GVHD.

For a proportion of patients, particularly patients inheriting two HLA-haplotypes containing alíeles in genetic dysequilibrium, unrelated donors also may be identified. For patients with two HLA haplotypes in linkage dysequilibrium, the probability of identifying a suitably compatible donor from a relatively small pool of 10 000 individuals is high. In. contrast, for patients inheriting haplotypes containing rare HLA alleles that are not linked to the other determinants in the haplotype, the probability of identifying a donor, even within a pool of a million donors, is remote.10 Based on the frequency of individual HLA-A, HLAB, and HLA-Dr alíeles and haplotypes in linkage dysequilibrium, it has been estimated that a donor registry containing 1 million individuals would be sufficient to provide an HLA-A, HLA-B, and HLA-Dr matched donor for at least an additional 35% of individuals who lack an HLA-matched sibling.11

Figure 1. Schematic diagram of the express genes of the HLA complex with the different alíeles currently identified by serology.

Figure 1. Schematic diagram of the express genes of the HLA complex with the different alíeles currently identified by serology.

To establish a donor pool of this size, the National Bone Marrow Donor Program has, over the last 6 years, conducted an intensive recruiting effort. At present, this registry maintains a computerized bank of more than 300 000 typed donors. Furthermore, this registry is now connected to several other donor registries in Europe. Principal among these is the Anthony Nolan Foundation, a registry of over 200 000 donors based in London. Additional, smaller registries have been established in France, Holland, Italy, the Scandinavian countries, and Russia. Currently, these registries are identifying serologically matched, MLC compatible donors for up to 20% to 30% of patients who are of European Caucasian background and particularly for patients inheriting two common HLAA, HLA-B, and HLA-D haplotypes. However, the proportion of successful searches for patients who are black, Oriental, or Native American is very low because of the limited number of donors from these minority groups in the donor registries.

Figure 2. Geneitc pedigree of HLA inheritance in a patient (A) with leukemia. Father and patient differ at a single HLA-A allele while the maternal uncle is HLA phenotypicaily identical.

Figure 2. Geneitc pedigree of HLA inheritance in a patient (A) with leukemia. Father and patient differ at a single HLA-A allele while the maternal uncle is HLA phenotypicaily identical.

MARROW GRAFT REJECTION AND GVHD: OBSTACLES TO THE SUCCESS OF ALLOGENEIC MARROW GRAFTS REFLECTING HISTOINCOMPATIBILITIES BETWEEN DONOR AND HOST

Transplants of bone marrow from allogeneic donors can be distinguished from transplants of other organs by their exquisite susceptibility to rejection. TTius, in contrast to organ grafts, which can be maintained in an allogeneic host by chronic administration of moderately immunosuppressive agents such as azathioprine or cyclosporine, marrow transplants are regularly rejected within 25 days unless the host's immune system has been completely ablated. Cyclophosphamide at a total dose of 200 mg/kg administered over 4 days is sufficiently immunosuppressive to ensure engraftment of unmodified HLA-matched marrow in at least 70% of patients transplanted for aplastic anemia.12 On the other hand, total body irradiation at doses of 7.5 to 10 Gray [Gy] radiation units coupled with cyclophosphamide or another chemotherapeutic agent has ensured durable engraftment of HLAmatched donor marrow in almost all leukemia patients, irrespective of their prior transfusion history.13 Even with this level of immunosuppression, however, T-ce!l depleted marrow transplants from HLAmatched siblings may be rejected in 5% to 20% of cases, depending on the method of T-cell depletion used.14,15

While the failure to achieve sustained engraftment may be ascribed to several etiologies, recent evidence suggests that the majority of graft failures observed in patients undergoing transplants for acquired disorders of hematopoiesis are due to an immune rejection mediated by host T-lymphocytes surviving preparative immunosuppression that exhibit donor specific reactivity.16,17

If an allogeneic marrow transplant achieves sustained engraftment, it may expand rapidly, restoring marrow cellularity and peripheral platelet and neutrophil counts within 25 to 35 days posttransplant. Once engraftment has been achieved, however, the patient becomes susceptible to the development of acute GVHD. Graft-versus-host disease is initiated by donor T- lymphocytes responding to alloantigens expressed on host cells, particularly cells of hematopoietic origin. 18 Following allosensitization, these Tlymphocytes replicate, generate host reactive cytolytic T-cells, and recruit other effector cells, particularly natural killer cells and monocytes. These cell populations can then attack the skin, the intestinal tract, the liver, and the lymphoreticular system, inducing necrosis of cells in infiltrated tissues.19 This process is reflected clinically by the development of a maculopapular rash, hepatitis, and damage to the mucosa of the intestinal tract producing diarrhea and intestinal ulcération of varying severity.20

In addition, patients with acute GVHD experience delayed hematopoietic reconstitution particularly associated with sustained thrombocytopenia and prolonged immunodeficiency, which especially involves the humoral immune system. In recipients of marrow derived from HLA-matched siblings, the incidence of acute GVHD is 60%, with moderate to severe forms observed in 30% to 40% of cases. In these donor recipient pairings, the incidence of severe GVHD is proportional to the age of the patient. Among older individuals, acute GVHD contributes directly or indirectly to death in 20% to 40% of affected individuals. The cumulative experience with HLA partially matched related marrow grafts clearly indicates that the incidence of graft rejection and severe GVHD is markedly increased following transplants between individuals disparate for a major histocompatibility determinant and that the risk of these potentially lethal complications is further augmented as the number of allelic disparities is increased.

The minor alloantigens that contribute to the activation of host or donor T-cells participating in graft rejections or GVHD in HLA-matched recipients are still largely unknown. Although H-Y specific T-cells have been implicated in rejections and in GVHD in sex-mismatched donor recipient pairs,21 this minor alloantigen likely contributes to only a portion of the reactions observed. In contrast, patients rejecting HLA partially matched related marrow grafts frequently develop, at the time of rejection, host T-cells exhibiting selective reactivity against unique class I or, less frequently, class Il HLA-determinants expressed on donor cells.16 The discriminatory power of the residual host T-cells to reject a marrow graft may be extraordinary. In one recently described patient, rejection was associated with the development of cytotoxic host T-lymphocytes exhibiting a specificity for a unique microvariant of HLA-B44 expressed on donor cells. In this case, the difference between the donor and host determinants was limited to a single amino acid on the alpha 2 helix of the B44 antigen.22 Human leukocyte antigen-specific cytolytic donor T-cells also have been detected in patients developing GVHD following HLA partially matched marrow grafts.

UNMODIFIED MARROW GRAFTS FROM PARTIALLY MATCHED RELATED DONORS

Since the early 1970s, several groups have conducted limited explorations of the use of HLA partially matched related donors in an attempt to identify tolerable histoincompatibilities within the HLA region. These early studies concentrated on the use of HLA-haplotype identical but MLC-compatible related donors because results in murine transplantation models indicated that marrow grafts from MHC class II disparate donors were most likely to induce lethal GVHD.23 Initial studies, which were conducted in patients with severe combine immunodeficiency (SCID), tended to support this approach to donor selection, since disparities for HLA-A and HLA-B on one haplotype were tolerated without lethal GVHD. In a review of 10 such cases,24 eight developed GVHD, but it was severe in only two patients. However, five of these patients died of infections early posttransplantation, which prevented an assessment of the incidence or severity of chronic GVHD. Of three patients grafted with HLA-D incompatible marrow, none survived long enough to assess graft-versushost reactions. Patients with SClD transplanted from donors incompatible for both class I and class II determinants had a poor outcome. Of 19 évaluable patients, three failed to engraft, and 10 developed severe GVHD. None of these patients survived.

Experience with partially matched related donor marrow grafts in the treatment of aplastic anemia underscores the incremental risk of marrow graft rejection when HLA-disparate grafts are used. In one large series of patients prepared with cyclophosphamide alone, HLA phenotypically matched nonsibling donor transplants were associated with consistent engraftment while seven of 1 1 grafts from one locus disparate donors and three of three transplants from two HLA-locus disparate donors were rejected.25 Among the 13 patients who ultimately achieved sustained engraftment, five of eight phenotypically matched and each of five single HLA-allele disparate recipients developed grade II to IV GVHD.

In 1985, Beatty et al26 reported that transplants administered to leukemic patients differing from their donors at a single HLA-allele developed grade II to IV GVHD in more than 75% of cases. However, longterm disease-free survival was comparable to that observed following HLA-matched sibling marrow transplants. In this series, no one allelic disparity (ie, HLA-A, HLA-B, or HLA-D) could be identified that placed the patient at greater risk for severe GVHD. Graft rejections were observed in 9% of such cases, an incidence higher than that following HLA-matched sibling grafts. As with transplants for SCID, leukemia patients transplanted from donors differing at more than one HLA-allele on the unshared haplotype were more prone to suffer graft rejection (21%) or to develop severe acute and chronic GVHD (80% to 85%). Overall survival of such patients (less than 15%) was markedly inferior to that observed following HLA phenotypically matched or HLA-single allele disparate marrow transplants (40% to 45%). Taken together, these data indicate that engraftment of unmodified bone marrow from a donor exhibiting unique HLA disparities requires a level of pretransplant immunosuppression at least comparable to that induced by the combination of cyclophosphamide and supralethal total body irradiation (TBI) and that even with such immunosuppression, the incidence of graft rejection following transplants from donors disparate for more than one HLA allele is prohibitive. Furthermore, such transplants are associated with a significantly higher incidence of severe acute and chronic GVHD. However, the complications of singleallele disparate marrow grafts, at least in patients transplanted for leukemia and, certainly, in patients transplanted for SCID, are tolerable and may be associated with long-term disease-free survival rates comparable to those achieved following HLAmatched sibling grafts.

TRANSPLANTS OF UNMODIFIED MARROW FROM UNRELATED DONORS

In 1977, our group reported a successful reconstitution of immunologie function in a child with SCID engrafted with marrow from an HLA-compatible unrelated donor.27 Subsequent case reports demonstrated that unrelated marrow grafts could also reconstitute hematopoiesis in patients transplanted for leukemia and aplastic anemia. In 1988, Gingrich et al reported a series of 40 patients who had refractory forms of leukemia and aplastic anemia and who received marrow grafts from unrelated, HLA-matched or partially mismatched donors obtained from a statewide registry developed in Iowa.28 Of these patients, 1 5% survived disease-free for periods of more than a year. Acute severe GVHD was observed in 67%, and all but one of the surviving patients also had chronic GVHD. Consistent engraftment was observed in leukemic patients prepared with TBI and cyclophosphamide. However, of four patients transplanted for aplastic anemia, two suffered graft failures despite preparation with total lymphoid irradiation (TLI) and cyclophosphamide.28 Of five aplastic patients reported by Gajewski et al,29 one also suffered graft rejection after preparation with TLI and cyclophosphamide. The incidence of graft failure following unrelated marrow transplants in patients with aplasia prepared with less immunosuppressive regimens, incorporating cyclophosphamide alone, is prohibitively high. Thus, these early experiences indicated an increased risk of both GVHD and graft failure following unrelated marrow transplants.

Results of unrelated marrow grafts applied to the treatment of chronic myelogenous leukemia have been considerably more encouraging. In a series of 102 recipients of unrelated marrow grafts transplanted at four centers for chronic myelogenous leukemia, 29% achieved extended disease-free survival.30 In this series, patients receiving transplants from unrelated HLA phenotypically matched donors achieved a somewhat better long-term disease-free survival (39%) than those transplanted with marrow from single alíele disparate donors (27%). Among recipients of unmodified HLA-matched unrelated marrow grafts, the incidence of grade II to IV GVHD was 80%. In this series, recipients of T-cell depleted marrow transplants developed severe GVHD less frequently without an increased risk of graft failure. However, longterm disease-free survival was comparable to that observed following unmodified marrow transplants.

In a subsequent report from Seattle, 52 patients received HLA phenotypically matched unrelated marrow transplants as treatment for leukemia.31 In this series of patients, who were prepared for transplantation with cyclophosphamide and TBI, the incidence of graft failure was low and not different from that observed following transplants from matched sibling donors. However, the incidence of grade II to IV acute GVHD was 79% compared with 36% for patients who received transplants from HLA-matched siblings. This incidence of severe acute GVHD is comparable to that seen following transplants of marrow from one or two HLA-allele disparate related donors.26

The basis for the marked increase in the incidence of severe acute GVHD in recipients of HLA phenotypically matched unrelated marrow when compared with that observed in recipients of HLA genotypically matched sibling marrow is still poorly understood. It may reflect alloreactions generated against molecular differences in class I and class II alíeles that can be detected only by more discriminatory molecular approaches such as isoelectric focusing or hybridization with HLA-microvariant sequence specific oglionucleotides. In a recent study, up to 30% of HLA serologically matched MLC-compatible unrelated individuals selected as potential donors were found to differ from their intended recipient by one or two alíeles distinguishable by isoelectric focusing.32 Preliminary evidence suggests that microvariant disparities for class II determinants also occur at significant frequency among HLA- and MLC-matched unrelated donor recipient pairs. These molecular différences in HLA class I and class II alíeles likely also explain the increased frequency of allocytotoxic T-cells that can be generated in mixed lymphocyte cultures between HLA phenotypically matched unrelated donor recipient pairs in comparison with the low frequency of such cells generated in mixed lymphocyte cultures between HLA-matched siblings.32

T-CELL DEPLETED HLA-NONIDENTICAL RELATED AND HLA-COMPATIBLE UNRELATED MARROW GRAFTS

A central limitation to the use of single HLA-disparate related donors and suitably matched unrelated donors is that such donors are realistically available for only 20% to 30% of individuals lacking an HLA-matched sibling. Furthermore, it is already clear that restricted donor availability will persist even when registries in excess of 1 million unrelated donors have been recruited.10 Development of more sensitive techniques for the selection of donors matched for HLA class I and class II microvariants, while important to our understanding of incompatibilities contributing to graft rejection, GVHD, and impaired immunologie reconstitution posttransplantation, will nevertheless only serve to further restrict the proportion of patients for whom an adequately compatible donor will be identified. Thus, there is a continuing need for the development of transplantation approaches whereby consistent engraftment and functional reconstitution can be achieved in HLAdisparate recipients without severe or lethal GVHD.

The development of techniques for depleting T' lymphocytes from a marrow allograft has provided one such approach to this dilemma. In 1981, our group showed that transplants of HLA-A, HLA-B or HLAhaplotype disparate parental marrow depleted of Tcells by agglutination with a soybean lectin followed by ?-rosette depletion could reconstitute hematopoietic and lymphoid function in children with leukemia or SCID without GVHD.33 ín another series of 40 patients with SCID who were transplanted with T-cell depleted HLA-haplotype disparate parental marrow over the last 10 years, 30 currently survive with reconstitution of immunity and stable donor lymphoid chimerism. In this series, only three patients have developed grade I acute graft-versus-host reactions. The actuarial disease-free survival for this group (74%) is not different from that achieved following unmodified HLA-matched grafts in severe combined immunodeficiency. Other centers incorporating this and other approaches to T-cell depletion have reported similar results.34 These studies of T-cell depleted HLA disparate marrow grafts applied to children with SCID demonstrate the potential of this approach and illustrate the feasibility of broad application of marrow grafts to patients lacking a matched sibling donor.

Unfortunately, when T-cell depleted HLA' nonidentical marrow grafts have been used for the treatment of other genetic and acquired diseases of hematopoiesis or leukemia, they have been considerably less effective. While techniques of T-cell depletion achieving a 3 Iog10 depletion of clonable T-cells can consistently prevent severe acute and chronic GVHD in both HLA-marched and HLA-disparate recipients, such transplants are associated with a high incidence of graft rejection.

In a report from the EORTC, of 23 children with genetic immunodeficiencies other than SCID transplanted with T-cell depleted HLA-nonidentical marrow, 11 failed to engraft despite conditioning with busulfan and cyclophosphamide.15 For these patients, 2- to 4-year disease-free survival was 29% compared with 47% for recipients of HLA-matched grafts.

Among patients transplanted with HLAnonidentical T-cell depleted marrow after cytoreduction with TBl and cyclophosphamide, the incidence of graft failures or rejection has ranged from 10% to 50% depending on the number of disparate HLAalleles unique to the donor, the efficiency of the T'Cell depletion technique used, and the intensity of the preparative cytoreduction administered prior to and immediately posttransplantation. Recently, howeven new techniques employing more intensive cytoreductive measures, coupled with less stringent or more selective T-cell depletion methods and the administration of T-cell specific immunotoxins or antithymocyte globulin in the early posttransplant period have reduced the incidence of rejection without unduly increasing the risk of severe GVHD.

For example, in a series of patients transplanted with T-cell depleted marrow from unrelated donors for hématologie malignancies,36 patients transplanted for acute leukemia in early remission or chronic phase CML achieved a 48% extended disease-free survival. In this series, employing a technique achieving a 2xlog10 level of T-cell depletion, HLA-matched unrelated marrow grafts were associated with a 20% incidence of grade II to IV GVHD, a result that is markedly lower than that observed among unmodified marrow transplants. A similarly low incidence of grade II to IV GVHD was observed among recipients of HLA phenotypicaily matched T-cell depleted unrelated marrow grafts in a four-center study of unrelated marrow transplants applied to the treatment of CML.30

In another compendium of the results of unrelated marrow transplants from four centers in the United Kingdom by Howard et al,37 such novel preparative regimens resulted in an incidence of engraftment comparable to that observed following unmodified marrow transplants. As a result, the T-cell depleted grafts, which were associated with the markedly reduced incidence of acute GVHD, also were associated with significant improvement in patients' diseasefree survival (60% versus 29% at 3 months). This was not observed among recipients of HLA-matched sibling marrow grafts.

The results of T-cell depleted partially matched related and HLA phenotypicaily compatible unrelated marrow transplants in children have yielded particularly promising results. In a series of 10 children with aplastic anemia, five have achieved longterm reconstitution.38 Similarly, in a report of 31 children transplanted with HLA-nonidentical T-cell depleted marrow, 54% were able to secure extended disease-free survival-39 In addition, a 50% extended disease-free survival has been reported in patients transplanted for leukemia with partial T-cell depleted marrow after more intensive cytoreduction and treatment posttransplant with a T-cell specific ricin-A immunotoxin.40 In these series, results of T-cell depleted transplants from related, one or two alíele disparate donors and unrelated HLA-matched donors have been similar.

While the results of trials incorporating techniques that permit more consistent engraftment while still achieving a low incidence of severe GVHD are encouraging, HLA partially matched related or HLAmatched unrelated marrow grafts still have not attained rates of extended disease-free survival comparable to those obtained with HLA-matched sibling grafts. In most of these series, the difference is due to a disturbingly high frequency of infectious complications following HLA partially matched related or unrelated marrow grafts, complications that may reflect the relatively profound and protracted state of immunodeficiency observed following these marrow transplants. The basis for these immunodeficiencies is not clear, but they may indicate limitations to the redevelopment or reeducation of donor cells in a partially HLA-disparate environment. Research in this critical area is urgently needed to identify and potentially circumvent such limitations to immune reconstitution. On the positive side, however, the incidence of relapse following such transplants, when applied to the treatment of patients with acute and chronic leukemias, has been strikingly low, possibly reflecting advantages of increased genetic disparity between donor and host for the expression of the antileukemic effects of a marrow allograft.

CONCLUSION

Considerable progress has been made in the development of transplantation approaches for patients lacking an HLA-matched sibling. As further improvements are accrued through the exploration of new cytoreductive regimens ensuring engraftment, more selective techniques for T-cell depletion, and the incorporation of novel approaches such as the use of cytokines to potentiate immunologie and hemopoietic reconstitution in the posttransplant period, further strides toward the general application of curative transplants will be made.

REFERENCES

1. Gatti RA, Meuwiissen JH, Alien HU et al. Immunologies! reconstitution al sex-linked lymphopenia immunological deficiency Lancet. 1968;ii:1366-1369.

2. O'Reilly RJ. Allogeneic bone manow transplantation: current status and future directions. Blood. 1983;62:941-964.

3. O'Reilly RJ, Biochstein J, Dinsmore R, Kirkpatrick D. Marrow transplantation for congenital disorders. SemM Memorai. 1984;21:188-221.

4. Bach FH, Albertini RJ, Joo P, Andersen JL, Bertin MM. Bone marrow transplantation in a patient with the Wiskott-Aldrich syndrome. Lancet. 1968;ii: 1364-1366.

5. Bodmer WF, Albert E, Bodmer JG, et al. Nomenclature fot factors of the HLA system. 1987. In: Dupont B, ed. Immunobiology of HLA, Histocompatibility Testing. Vol 1. Nw York, NY: Springer-Verlag; 1989:71-79.

6. White PC. Molecular genetics of the class III region of the HLA complex in immunobiology of HLA. In: Dupont B, rd. Immunogenetics and Histocompatibility. Vol 2. New York, NY: Springer-Verlag; 1989;62-69.

7. Choo SY, Antonelli P, Nisperos B, Nepom GT, Hansen JA. Six variants of HLA-B27 identified by isoelectric focusing. Immunogenetics. 1986;23:24-29.

8. Thomsen M, Alvarei C, Carpenter CB. Analysis of HLA recombinant families by Southern blots. In: Dupont B, ed. immunobiology of HLA, Histocompatibility Testing. Vol 1. New York. NY: Springer- Verlag; 1989:900-908.

9. Bauer MP, Neugebauer M, Albert ED. Reference tables of three-locus haplotype frequencies and delta values in Caucasians, Orientals, and negroids in histocompatibiliry testing 1984. In: Albert ED ed. Histocompatability Testing 1984. Berlin Heidelberg: Springer-Verlag; 1984:756-760.

10. Sonnenberg FA, Eckman MH, Pauker SG. Bone marrow donor registries: the relation between registry size and probability of finding complete and partial matches. Blood. 1989;74:Z569-2578.

11. Beatty PG, Dahlberg S, Mickelson EM, et al. Probability of finding HLA-matched unrelated marrow donors, Transplantation- 1988;45;714-718.

12. Storb R, Prentice RL, Thomas ED. Marrow transplantation for treatment of aplasie anemia: an analysis of factors associated with graft rejection. N Engl J Med. 1977;296:61-66.

13. Thomas ED, Buckner CD,A Banaji M, et al. One hundred patients with acute leukemia treated by chemotherapy, total body irradiation, and allogeneic marrow transplantation. Biood. 1977;49:511-533.

14. Keman NA, Bordignon C, Heller G, et al. Graft failure aftei T-cell depleted human leukocyte antigen identical marrow transplants for leukemia, I: analysis of risk factors and results of secondary transplants. Blood 1989;74:2Z27-2236.

15. Martin PJ, Hansen JA, Torok-Storb B, et al. Graft failure in patients receiving T-cell depleted HLA-identical allogeneic marrow transplants. Bone Marma Transplant. 1988;3:445-456.

16. Keman NA, Flomenberg N, Duponi B, O'Reilly RJ. Graft rejection in recipients T'Cell depleted HLA-non-idenlical marrow transplants for leukemia. Transplantation. 1987;43:842-847.

17. Bordignon C, Keever CA, Small TN, Flomenberg N, Dupont B. O'Reilìy RJ. Graft failure after T-cell depleted human leukocyte antigen identical marrow transplants for leukemia: in vitro analyses of host effector mechanisms. Biood. 1989;74:2227-2236.

18. Streilein WJ, Billingham RE, An analysis of graft-versus-host disease in Syrian hamsters. J Exp Med 1970;132:163-180.

19. Slavin RE, Woodruff JM. The pathology of bone marrow transplantation. Pathol Annu. 1974;91:291-344.

20. Glucksberg H, Storb R, Fefer A. et al. Clinical manifestations of GVHD in human recipients of marrow from HLA-matched sibling donors. Transplantation. 1974;18:295-304.

21. Goulmy E. Termijtzlen A, Bradley BA, van Rood JJ. Y-amigen killing by T-cells woman is restricted by HLA. Nature. 1977;2Z6:544-545.

22. Fleischhauet K, Keman NA, O'Reilly RJ, Dupont B, Yang SY. Bone marrow allograft rejection by host-derived allocytotoxic T lymphocytes recognizing a single acid at position 156 of the HLA-B44 class 1 antigen. N Engl J Med. 1990;23:1818-1822.

23. Klein J, Park JM. Graft-versus-host reaction across different regions of the H-2 complex of the mouse. J Exp Med. 1973;137:1213-1225.

24. Kenny AB, Hitzig WH. Bone marrow transplantation for severe combined immunodeficiency. Eur J Pdiatr. 1979;131:155-177.

25. Beatty PG, Bartolomeo P, Storb R, et al. Treatment of aplastic anemia with marrow grafts from related donors other than HLA genotypically-matched siblings. Clinical Transplantation. 1987;1:117.

26. Beatty PG, Clift RA, Michelson EM, et al. Marrow transplantation from related donors other than identical siblings. N Engl J Med. 1985;313:765-771.

27. O'Reilly RJ, Dupont B, Pahwa S, et al. Reconstitution in severe combined Immunodeficiency by transplantation of marrow from an unrelated donor. N Engl J Med. 1977;297:1311-1318.

28. Gingrich RD, Ginder CO, Goeken D, et al. Allogeneic marrow grafting with partially mismatched, unrelated donors. Blood. 1988;71:1375-1381.

29. Gajewski J, Ho WG, Feig SA, Hunt L, Kaufman N, Champlin RE. Bone manow transplantation using unrelated donors for patienta with advanced leukemia or bone marrow failure. Transplantation. 1990;50:244-249.

30. McGlave PB, Beatty PG, Ash R, Hows JM. Therapy for chronic myelogenous leukemia with unrelated donor bone marrow transplantation: results in 102 cases. Blood 1990;75:1728-1732.

31. Beatty PG, Hansen JA, Longton GM, et al. Marrow transplantation form HLA-matched unrelated donors for treatment of hematologic malignancies. TransfttnMtion. 1991;51;443-447.

32. Keman NA. Khan R, Landrey C, O1ReUIy RJ1 Dupont B. Yuig SY. Identification of unrelated bone manow donors based on matching for class I IEF subtypes. Blood. 1990;76:548a.

33. Reisner Y, Kapoor N. Kirkpatrick D, et aL Transplantation for acute leukemia with HLA-A and B non-identical parental marrow cells fractionated with soybean agglutini« and sheep blood cells. Lancet. 1981;ii:3Z7-33I.

34. O'Reilly RJ, Keever CA, Small T, Brochstein J. The use of HLA-non- identical T-cell -depleted marrow transplants for correction of severe combined immunodeficiency disease, immunodeficiency Reviews. 1989:1:273-309.

35. Fischer A, Griscelli C, Friedrich W, et al. Bone marrow lransplaniation for immunodefkiencies and osteopetnsis. European survey. 1968-1985. Lancio. 1986;iii:1080-1084.

36. Ash RC, Casper JT, Chitambar CR, et al. Successful allogeneic transplantation on T-cell depleted bone marrow from closely HLA-matched unrelated donors. N Engl J Med. 1990;332:485.

37. Howard MR, Hows JM, Gore SM, et al. Unrelated donor marrew transplantation between 1977 and 1987 at four centers in the United Kingdom. Transplantation. 1990;49:547-553.

36. Camitra B1 Ash R, Menilove J, et al. Bone marrow transplantation for children with severe aplastic anemia: use of donors other than HLA-identical siblings. Blood. 1989;74:1852-1857.

39. Trigg ME, Gingrich R, Goeken N, et al. Low rejection rale when using unrelated or haploidentical donors for children with leukemia undergoing marrow transplantation. Bone Marrow Transplant. 1989:4:431-437.

40. Henslee PJ, MacDonald JS, Messina MJ. freedom from relapse following histoincompatiblc marrow transplantation in patients with high risk acute lymphoblastic leukemia. Eq. Hematol. 1989;17:547a.

10.3928/0090-4481-19911201-08

Sign up to receive

Journal E-contents