|
|
||||||||
Commentary |
IntroGene BV, Leiden, The Netherlands
Dr. D.W. van Bekkum, IntroGene BV, P.O. Box 2048, 2301 CA, Leiden, The Netherlands.
| Introduction |
|---|
|
|
|---|
In the past few years, the first attempts were made to treat patients suffering from severe refractive autoimmune diseases with a similar myeloablative regimen and rescue with autologous stem cells. At the time of writing, over 100 patients are known to have been treated this way, with encouraging results thus far. Several meetings devoted to this new treatment have been held, and dedicated working parties as well as a registry have been initiated [1]. A whole new group of indications for autologous stem cell transplantation is about to emerge. The application of a potentially risky procedure to diseases that are generally held to be not immediately life-threatening demands a very cautious approach. This implies that trials should only be carried out in a clinical setting with adequate expertise in BMT.
In this review, the experimental data that contributed to the development of this new treatment strategy will be briefly outlined.
| Rationale for Treatment with Stem Cell Transplants |
|---|
|
|
|---|
In recent years, some patients with severe progressive connective tissue diseases such as lupus erythematosus (LE) and rheumatoid arthritis (RA) have been treated successfully with increasing doses of methotrexate and/or cyclophosphamide [2, 3]. This experience has facilitated the acceptance of a more toxic and potentially lethal regimen followed by rescue with stem cells for refractory patients.
The use of autologous stem cells for hematological rescue does not arise from logical reasoning but is wholly empirical. Logic, as well as some clinical and experimental evidence, dictate the use of allogeneic bone marrow. The prevailing perceptions of autoimmunity imply that stable cures can only be expected when the autoreactive cells of the patient are replaced with cells from a healthy donor who is not susceptible to whatever factors break self-tolerance. As will be discussed in more detail later, animals with full-blown autoimmune disease of either the spontaneous or the induced type can be cured by treatment with high-dose total-body irradiation (TBI) and allogeneic BMT. Furthermore, retrospective analysis of clinical records from long-term survivors of leukemia or aplastic anemia following treatment with allogeneic BMT has revealed 22 patients who were also cured of a coincidental autoimmune disease [4-6].
Allogeneic BMT is preferred for treating leukemias, especially chronic myeloid leukemia, because, apart from not carrying malignant cells, an additional benefit may be obtained from an immunological reaction of the donor-type lymphoid cells against the leukemic cells, the so-called graft-versus-leukemia reaction. An analogous effect of allogeneic BMT has been observed in rats suffering from experimental allergic encephalomyelitis (EAE)the model for multiple sclerosis (MS) [7]. Following conditioning with a lethal dose of TBI, allogeneic bone marrow transplants were more effective than syngeneic marrow grafts in the prevention of relapses. This advantage is ascribed to an immunological reaction of the donor lymphocytes against the residual T cells of the recipient, a reaction that Marmont appropriately named graft-versus-autoimmunity [4]. The reversion of insulitisand prevention of diabetesin nonobese diabetic (NOD) mice treated with sublethal TBI and allogeneic bone marrow has also been tentatively attributed to such a reaction [8].
In spite of these arguments in favor of its use, however, allogeneic BMT is presently not an acceptable option for the majority of patients with severe autoimmune disease. The transplantation-associated mortality of allogeneic BMT from a major histocompatability complex (MHC)-identical sibling donor is presently estimated at 15% [9] for patients with favorable prognostic features. Another obstacle is the scarcity of MHC-identical sibling donors. Currently, such a donor is available for only one-third of the transplant candidates. For others, a matched unrelated donor may be found, but unrelated grafts carry an even higher risk than those from MHC-identical siblings. The unexpected finding in arthritic rats that autologous BMT is as effective as allogeneic BMT [10] has paved the way for the current clinical studies with autologous stem cell transplants for treating autoimmune diseases.
Autologous bone marrow and mobilized peripheral stem cells are frequently employed in combination with high-dose chemotherapy for the treatment of solid tumors and hematological malignancies. The main concern is that tumor cells present in the autograft will give rise to relapse. The obvious remedy consists of purging of the graft. In the case of leukemia, autologous BMT is carried out after the patients have been brought into remission, and in the case of mammary cancer, patients without bone marrow involvement are selected. These precautions do not rule out the presence of small numbers of tumor cells in the graft. To minimize that risk, a variety of techniques have been introduced for the elimination of tumor cells without affecting the repopulation efficacy of the graft. In the case of active autoimmune disease, the patients are not in remission at the time of transplantation, and, therefore, depleting the autograft of autoreactive lymphocytes seems even more advisable. Autoreactive cells cannot as yet be selectively removed. Most autoimmune diseases are T-cell mediated or T-cell dependent, so depletion of the graft should focus on T cells. Purging in general has become easier and more reliable by the introduction of positive stem cell selection technology.
| Animal Models of Autoimmune Disease |
|---|
|
|
|---|
The induced diseases are elicited by immunization in certain susceptible inbred strains of rodents. Susceptibility and resistance to induction are largely, but not absolutely, determined by genetic factors. Representatives of this category are adjuvant arthritis (AA) and streptococcal wall arthritis in rats; collagen arthritis in rats and mice; and EAE in rats, mice, and guinea pigs.
It has been known for more than 20 years that in hemopoietic chimeras between (spontaneous) autoimmune strains and normal strains, the development of the disease is dictated by the origin of the bone marrow graft . These experiments were reviewed elsewhere [5, 11]. Of particular interest are the experiments performed with the BXSB lupus-prone mouse strain [12]. The males develop autoimmune disease around 22 weeks of age, and the females much later, at the age of 65 weeks. Grafting of male bone marrow from still-healthy young animals into lethally irradiated males or females induced early-life disease, whereas female marrow induced late disease in both male and female recipients. In addition, transfer of spleen cells from older male mice with fully developed disease did not induce the disease in female recipients any faster than spleen cells from young, healthy males. Apparently, there was no passive transfer of the disease by the lymphocytes of the spleen in these experiments, and it was concluded that the autoimmune condition is transferred by the hemopoietic stem cells, which had to differentiate into T and B lymphocytes before the disease became manifest.
More recently, Ikehara et al. [13] were able to transfer diabetes as well as a lupus syndrome to lethally irradiated recipients from a normal mouse strain with T-cell-depleted or stem-cell-enriched bone marrow grafts from the NOD mouse strain and the (NZWxBXSB)F1 lupus strain respectively. The stem cell donors were in the predisease stage. These results gave rise to the hypothesis that the cause of autoimmune diseases lies in the hemopoietic stem cell. Accordingly, curing the fully developed disease would not only require the destruction of all autoreactive lymphocytes but also the replacement of the stem cells with those of a normal non-autoimmune prone, i.e., allogeneic, donor. Indeed, transplantation of allogeneic stem cells has induced long-lasting complete recovery of BXSB mice with fully developed LE-like disease [14]. In similarly treated MLR/lpr mice, the disease recurred after initial remission. These relapses were associated with a reversal to recipient-type hemopoiesis, which could be prevented by grafting stromal cells from donor bones [15]. The transfers described above led to the expectation that rescue with autologous stem cells following lymphomyeloablative conditioning will inevitably result in recurrences after the lymphoid cell population has recovered [16]. Unfortunately, these postulates have never been experimentally verified by actually treating sick animals with autologous stem cells. However, abrogation of florid autoimmune disease of the spontaneous category a.o. in MLR/lpr mice has been reported by several authors following treatment with a sublethal dose of TBI, as reviewed by Loor et al. [17]. As regeneration of the immune system after sublethal TBI derives from surviving precursors of the affected animal itself, i.e., autologous cells, a more complete ablation and rescue with autologous stem cells is expected to be even more effective.
In the inducible rat models of autoimmune disease, the susceptibility and the resistance to induction can also be transferred by bone marrow grafting between normal animals of the susceptible and resistant strains. The responses of the resulting chimeras were determined by the genotype of the bone marrow. However, non-stem-cell-related factors may be equally important, as for instance, the levels of stress-related corticosteroids in EAE. Adrenalectomy can turn rats from EAE-resistant into EAE-susceptible animals [18, 19]. The role of the bone marrow genotype has been studied in two inducible mouse models: collagen-induced arthritis in DBA/1 mice [20] and EAE in SJL/J mice [21, 22]. In both cases, the inducibility did not depend on the bone marrow origin. The discrepancy between the results in rat and mouse models has not so far been resolved.
The induced experimental autoimmune diseases appear to be the more realistic models of human disease, as both partially depend on genetic determinants of predisposition and require exposure to external factors for the initiation of the disease. We selected AA and EAE because these diseases have many features in common with various forms of inflammatory arthritis in man and with multiple sclerosis. By selecting the buffalo (BUF) rat from among a number of susceptible strains and by manipulating the dose of antigen for induction, we could reproducibly induce chronic progressive polyarticular arthritis and chronic relapsing encephalomyelitis, respectively. Both diseases take a severe course in this strain of rats, and, if left untreated, cause some mortality.
The objective of our research was to investigate whether the fully developed disease could be favorably influenced by radical lymphomyeloablation and rescue with bone marrow transplantation. In our first experiments with AA, we used allogeneic bone marrow from a nonsusceptible rat strain, expecting this to have a beneficial effect, which it had. In fact, it induced complete and long-lasting remission in all animals [23]. As negative controls, rats were rescued with syngeneic marrow from healthy BUF rats, with the expectation that they would relapse soon. Surprisingly, this was not the case, and this led us to investigate autologous BMT, which proved to be similarly effective [10]. As an explanation, it is postulated that the reconstitution of the immune system from a limited number of hemopoietic stem cells is a recapitulation of ontogenesis with the acquisition of self-tolerance. In the EAE model, autologous BMT is also quite effective, but at the maximal tolerated dose of TBI there is still about a 30% spontaneous relapse rate. These relapses can be reduced by adding anti-T-lymphocytic immunoglobulin to the conditioning regimen [24].
The information obtained from both models should be taken into account when designing a general therapeutic strategy for the human autoimmune diseases. The large variation and complexity of the different human autoimmune diseases and the variation between patients with the same disease preclude a simple translation from one model to one disease. Moreover, although each model has basic features in common with the corresponding human disease, they are certainly not identical. The most important issues for clinical protocols at present are patient selection, the conditioning regimen, and T-cell depletion of the autograft. Regarding patient selection there is reasonable consensus that candidates should have progressive inflammatory disease that is refractory to all conventional treatment and should be without significant organ failure. Experiments in rats have shown that "scar" lesions such as osseous malformations in AA [23] and extensive demyelination in EAE [25] cannot be cured by BMT.
The information regarding conditioning and the necessity of T-cell depletion of the graft as collected from experiments with the two models are discussed below.
| Conditioning and Rationale for T-Cell Depletion |
|---|
|
|
|---|
|
The best curative results were obtained in AA with a TBI of 9 Gy and in EAE with 10 Gy TBI single exposure. This dose amounts to a 3-4 log kill of lymphocytes (D0 = 1 Gy). If a similar reduction were required in patients with severe autoimmune disease, and the total lymphatic cell mass of man is taken as 1012, the residual number of lymphocytes after conditioning is 108-109, of which about 10%-30% are T cells. A bone marrow graft may contain as many as 4 x 109 and a mobilized peripheral blood cell graft up to 2 x 1010 T lymphocytes. Without T-cell depletion, one would reinfuse 10 to 100 times the number of residual T cells that survived conditioning, which does not make sense. In view of the uncertainties in these estimates, it seems wise to aim at 3 log T-cell depletion in case of bone marrow and for 4 log T-cell depletion in case of peripheral blood cells. An even better guideline is to set a maximum at 105 T cells/kg of body weight for both types of grafts, as recommended by the International Meeting on Hemopoietic Stem Cell Therapy in Autoimmune Diseases at Basel in 1996 [1], although in practice, 106/kg seems to be more realistic.
Cyclophosphamide (CP) as a single agent was compared with TBI in rats with AA (van Bekkum, unpublished data). At the highest tolerated dose of 160 mg/kg (divided over two days), CP was clearly less effective than 9 Gy TBI in terms of the proportion of rats attaining complete remissions and the number of recurrences.
In the EAE model, the treatment efficacy in terms of relapse prevention as obtained with 10 Gy TBI was equaled by 120 mg/kg CP plus 7 Gy TBI, which indicates that this dose of CP is equivalent to only 3 Gy TBI. These findings argue against using CP as the sole conditioning agent. Furthermore, a qualitative difference between CP and TBI is suggested by Pestronk et al. [28], who provided evidence in rats with experimental autoimmune myasthenia gravis (EAMG) that CP is less effective than TBI in eliminating memory cells. Admittedly, the memory cells in this model were predominantly B memory cells, but there is some evidence from in vivo experiments in animals that a similar difference holds for T memory cells [29-31]. This notion is supported by the experience with allogeneic BMT for treating patients with severe aplastic anemia. In nonsensitized patients, a low rate of graft failures is seen after conditioning with high-dose CP as a single agent, but in patients who have been sensitized as a result of numerous blood transfusions, allografts are frequently rejected. Take failure or rejection of an allogeneic bone marrow graft is attributed to residual memory T lymphocytes. A combination of high-dose TBI and CP (100 or 120 mg/kg) is needed in these patients to restore the take rate.
Accordingly, a combination of high-dose CP with moderate-dose TBI (to inactivate memory cells) may provide the best chance for achieving complete and relapse-free remission in patients with refractory autoimmune disease.
Conditioning with a combination of busulfan (BU) and CP was equally effective as a lethal dose of TBI in AA but inferior to TBI in EAE. In the latter experiments syngeneic bone marrow was used, which gives essentially the same results in EAE as autologous marrow, presumably because the T-cell content of the rat bone marrow is only 3%. In both models, partial-body irradiation proved to be ineffective.
| Autologous or Allogeneic Stem Cells? |
|---|
|
|
|---|
An exception might be considered for patients with progressive severe MS who have an MHC-matched sibling donorfirst, because in the EAE model the incidence of both spontaneous and induced relapse rates was much lower after allogeneic BMT, and second, because the central nervous system is not affected by graft-versus-host disease.
| Conclusions and Recommendations |
|---|
|
|
|---|
Conditioning should be as intensive as is safely tolerated to minimize residual T lymphocytes in both the inflammatory lesions as well as elsewhere in the body. In our animal models, supralethal TBI (9-10 Gy single exposure) and the combination of CP with a moderate dose of TBI (4-7 Gy) were about equally effective.
The recommendation to include TBI in the conditioning regimen for autoimmune diseases is not attractive to some physicians. They fear that the irradiation will aggravate the fibrotic lesions in the connective-tissue-type diseases, e.g., systemic sclerosis. This fear is not realistic, however, because the dose that is known to cause fibrosis following local irradiations is 5-10 times higher than that employed for TBI. Furthermore, fibrotic changes have not been observed so far in the tissues of long-term survivors of TBI.
Another realistic concern is the increased risk of malignancies following TBI. This late effect of irradiation occurs in the dose range used for TBI and has been well documented. However, CP and the other alkylating agents are expected to carry a similar risk, as does prolonged treatment with immunosuppressive agents. A more detailed discussion of these adverse effects will appear elsewhere [32].
The BEAM regimen consisting of BCNU, etoposide, cytosine-arabinoside, and melphalan, with post-transplant ATG (antilymphocyte globulin) was employed with encouraging results by Fassas et al. [33] for the treatment of MS patients with autologous stem cells. As it is likely to be less lymphoablative than the other regimens discussed above, additional pretransplant administration of ATG is advisable. Burt et al. [34] recently reported that patients with progressive MS tolerated well a conditioning with 120 mg/kg CP, methylprednisolone, and 12 Gy fractionated TBI, the latter dose being equivalent to 8-9 Gy single exposure.
In selecting a conditioning regimen, it should be kept in mind that CP as the sole conditioning agent was inadequate in both the EAE and AA rat models, as well as in the specific B-cell model, EAMG. It may be argued that this seems to be in contrast with the favorable outcome of BMT in autoimmune aplastic anemia following conditioning with high-dose CP alone. In this disease, however, the graft is allogeneic, which allows for a graft-versus-residual host lymphocyte reaction.
In all cases, extensive T-cell depletion of the autologous graft is mandatory. Early relapses occurred when this was omitted [35].
Autologous stem cell grafting of patients suffering from severe refractory autoimmune diseases of many kinds is rapidly emerging [33, 34, 36-40]. Encouraging early results with more than 100 patients with systemic RA, LE, or MS have been reported at various meetings during the past year. Among these are children suffering from severe refractory juvenile chronic arthritis who have responded with a spectacular complete remission [37]. The transplantation-related procedures were generally well tolerated by these severely ill patients. The majority went into remission or, as some of the MS patients, remained stable. Obviously, the results after longer observation periods have to be awaited. In view of the diversity of autoimmune disease manifestations, it is important to adhere to standards of staging, treatment protocols, and criteria of reporting as recommended by EULAR/EBMT [1]. More detailed recommendations for MS are expected in the near future, in particular for patient selection. At this stage, it is important to reconsider the validity of current prognostic factors for some of these diseases so that a reliable estimate of benefits versus risk of the treatment can be made.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T Mondria, C H J Lamers, P A W t. Boekhorst, J W Gratama, and R Q Hintzen Bone-marrow transplantation fails to halt intrathecal lymphocyte activation in multiple sclerosis J. Neurol. Neurosurg. Psychiatry, September 1, 2008; 79(9): 1013 - 1015. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. C. Marty, F. Alliot, J. Rutin, R. Fritz, D. Trisler, and B. Pessac The myelin basic protein gene is expressed in differentiated blood cell lineages and in hemopoietic progenitors PNAS, June 25, 2002; 99(13): 8856 - 8861. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| STEM CELLS | THE ONCOLOGIST | CME | ALPHAMED PRESS JOURNALS |
