© 2000 AlphaMed Press Nonmyeloablative Allogeneic Stem Cell Transplant Strategies and the Role of Mixed ChimerismBone Marrow Transplant Program, Harvard Medical School, Boston, Massachusetts, USA Correspondence: Thomas R. Spitzer, M.D., Director, Bone Marrow Transplant Program, Massachusetts General Hospital, Associate Professor of Medicine, Harvard Medical School, Boston, Massachusetts 02114, USA. Telephone: 617-724-1124; Fax: 617 724-1126; e-mail: spitzer.thomas{at}mgh.harvard.edu
Experimental and clinical experiences have demonstrated successsful donor engraftment following nonmyeloablative preparative regimens. These less toxic conditioning strategies may be better tolerated with diminished transplant-related morbidity and mortality. Importantly, the intentional induction of mixed chimerism can be established following nonmyeloablative conditioning. This approach has the potential advantages of inhibiting graft-versus-host disease, presumably secondary to the persistence of host immunoregulatory cells, and providing a platform for the delivery of adoptive cellular immunotherapy with donor leukocyte infusions for patients with an underlying malignancy. This review will describe the preclinical evolution of nonmyeloablative transplant strategies, the rationale for considering these approaches, and the preliminary clinical experience with this novel allogeneic stem cell therapy. Key Words. Nonmyeloablative therapy • Mixed chimerism • Graft-versus-host disease • Allogeneic transplant
Conventional allogeneic bone marrow or peripheral blood stem cell transplantation has utilized high-dose, allegedly myeloablative chemoradiotherapeutic preparative (conditioning) therapy. This strategy was founded on the principles of maximal tumor cytoreduction according to a well-demonstrated, dose-response relationship of chemoradiotherapy [1] and adequate immunosuppression to allow for engraftment, even following non-HLA-genotypically identical donor transplants [2-4]. While the latter objectives have been largely realized (rejection rates are generally less than 5% with non-T-cell-depleted allogeneic transplants), the former goal remains problematic. Despite maximally tolerated doses of chemoradiotherapy, relapse probabilities, especially for advanced hematologic malignancies, remain disappointingly high [4-6]. High-dose chemoradiotherapy is also associated with substantial transplant-related toxicities and a significant incidence of acute and chronic graft-versus-host disease (GVHD) [7]. The latter complications of allogeneic transplantation are, moreover, enhanced by the increasing intensity of the conditioning regimen [8].
Recently, attempts have been made to diminish transplant-related morbidity and possibly mortality by administering relatively nontoxic, nonmyeloablative doses of chemotherapy or radiation therapy prior to allogeneic transplantation [9-13]. Donor engraftment is feasible even following nonmyeloablative conditioning therapy [14-18], and these less toxic conditioning regimens are generally better tolerated, thus allowing for treatment of older patients and patients with co-morbid disease. The induction of mixed chimerism, moreover, may serve as a platform for later adoptive cellular immunotherapy (and conversion of the mixed chimeric state to one of fully donor hematopoiesis). Figure 1
Transplant strategies that employ nonmyeloablative preparative conditioning therapy have been popularly termed "mini-transplantation" or "transplant-lite" [10]. However, central to the issue of these strategies is what constitutes ablative versus nonablative conditioning therapy. Lymphohematopoietic recovery has been described following putatively ablative doses of whole-body irradiation and other seemingly ablative doses of high-dose chemotherapy, raising the question of whether conventional conditioning regimens are truly ablative of the bone marrow and immune system [19, 20]. Certain preparative regimens, particularly those containing whole-body irradiation in total doses of 900 cGy or high-dose busulfan, are associated with such profound marrow injury that meaningful hematologic recovery generally does not occur following graft rejection. On the other hand, minimal or transient myelosuppression usually follows the relatively low doses of chemotherapy or radiation therapy that are used for these "mini-transplants." Following low-dose total-body irradiation (200 cGy, for example), most patients have not become neutropenic or transfusion dependent [20]. Following the more intensive cyclophosphamide-based regimens that we have employed (which include cyclophosphamide at a dose of 150-200 mg/kg), significant myelosuppression occurs but rapid hematologic recovery has uniformly been seen, even when donor hematopoiesis is not established (with early recovery of host hematopoiesis in these situations). Thus, nonmyeloablative conditioning regimens should be defined by the transient marrow injury that occurs with resultant complete and nondelayed hematologic recovery, even following graft rejection. Another outcome of some nonmyeloablative transplant strategies is the induction of mixed lymphohematopoietic chimerism. This state of mixed chimerism is operationally defined by the stable presence of both host and donor lymphohematopoietic elements in the peripheral blood and/or bone marrow. In some situations, "split-lineage" chimerism has also been observed, in which varying percentages of host and donor lymphoid and nonlymphoid cell subsets are seen following transplantation. These states of mixed chimerism have been intentionally induced in some transplant protocols [12, 13, 21]. This approach has the potential advantage of preventing GVHD by allowing for the persistence of host immunoregulatory cells, and at the same time establishing a platform for delivering adoptive cellular immunotherapy via delayed DLI.
In several animal models, mixed lymphohematopoietic chimerism can be established following nonmyeloablative transplant preparative regimens (Table 1
In large animals, mixed lymphohematopoietic chimerism has also been established following several different nonmyeloablative preparative strategies. In miniature swine, mixed chimerism has been achieved following low-dose whole body radiation therapy (300 cGy), anti-T-cell antibody therapy, thymic irradiation, and swine leukocyte antigen-matched donor BMT [23]. There is evidence in both murine and miniature swine models indicating that stable mixed chimerism can be achieved following regimens that employ only anti-T-cell antibody therapy and thymic radiation, even following stem cell transplants from MHC mismatched donors ([17] and Dr. D.H. Sachs, personal communication). Acute GVHD also appears to be less prominent than following transplants utilizing conventional ablative doses of chemoradiotherapy. In a dog model, an effective regimen for inducing mixed chimerism employs very low-dose whole body radiation (200 cGy) as the sole preparative therapy and post-transplant immunosuppression with mycophenolate mofetil (MMF) and CYA, and dog leukocyte antigen-matched donor BMT [16]. Optimizing post-transplant immunosuppression to inhibit competing graft-versus-host and host-versus-graft reactions is believed to be a central focus of this treatment strategy.
With this positive preclinical experience as background, clinical investigation of nonablative transplant regimens has begun in a number of transplant centers. Several published reports have demonstrated the feasibility of achieving allogeneic engraftment following nonmyeloablative conditioning therapy [9-12]. The diversity of the conditioning regimens, small numbers and heterogeneity of the patient populations, and limited goals of the sentinel experiments (e.g., reduction in transplant-related morbidity versus the intentional induction of mixed lymphohematopoietic chimerism), allow only preliminary interpretation of the data, particularly in terms of its effectiveness in achieving control of malignant disease. Nonetheless, taken together, these reports indicate the tolerability of most of these regimens, and demonstrate that mixed lymphohematopoietic chimerism can be intentionally induced, even across major HLA barriers; in some cases, the mixed chimerism led to potent antitumor response (perhaps enhanced by delayed DLI), a particulary exciting advance in the field of clinical allogeneic stem cell transplantation. Many transplant centers have initiated pilot trials evaluating nonmyeloablative conditioning regimens. The results of most of the trials are too preliminary to report. The clinical data from several trials, representing the diversity of conditioning strategies that have been employed, have been published. The results of the more notable of these preliminary trials are shown in Table 2
Two series from the M.D. Anderson Cancer Center have employed fludarabine-based nonmyeloablative conditioning regimens and recombinant growth factor-mobilized allogeneic peripheral blood stem cell transplants for treatment of hematologic malignancies [9, 10]. Khouri and colleagues [10] described 15 patients with chronic lymphocytic leukemia (CLL) or low to intermediate-grade non-Hodgkin's lymphoma (NHL) who received fludarabine (30 mg/m2/d x 3 days) and cyclophosphamide (300 mg/m2/d x 3 days) or fludarabine (30 mg/m2/d x 2 days), cisplatin (25 mg/m2/d x 4 days), and cytosine arabinoside (500 mg/m2/d x 2 days). Eleven of 15 patients had donor cell engraftment as measured by restriction fragment length polymorphism (RFLP) analysis of the bone marrow. The percentage of donor cells ranged from 50% to 100% at one month post-transplant. Seven patients had greater than or equal to 90% donor cells. Four patients had exclusively recipient marrow cells and recovered autologous hematopoiesis promptly. Only four patients received delayed DLI from 55 to 100 days post-transplant. One of the patients converted from 75% donor cells at six weeks post-transplant to 100% donor cells following the DLI. Five patients developed acute GVHD following the initial transplant. Two patients developed extensive chronic GVHD, which was fatal in one patient. Three patients developed GVHD following DLI; one of these patients had grade IV GVHD that was fatal. Eight patients achieved a complete response. At an initial median follow-up of 180 days, five of six patients (83.3%) with chemosensitive disease were alive compared to only two of nine patients (22.2%) who had chemorefractory or untested disease. Giralt and colleagues [9] treated 15 patients with acute myeloid leukemia or myelodysplastic syndrome. Preparative regimens were either fludarabine (30 mg/m2/d x 4 days) and idarubicin (12 mg/m2/d x 3 days) with either cytosine arabinoside (2,000 mg/m2/d x 4 days) or high-dose melphalan (140 mg/m2/d), or 2-chlorodeoxyadenosine ([2-CDA], 12 mg/m2/d x 5 days) and cytosine arabinoside (1,000 mg/m2/d x 5 days). Thirteen patients received peripheral blood stem cell transplants from HLA-genotypically identical sibling donors and two from a one-antigen mismatched sibling donor. Chemotherapy was generally well tolerated with only one death from multiorgan failure. Thirteen patients had engraftment as evidenced by an absolute neutrophil count of greater than 0.5 x 109 per liter. Chimerism was measured by either conventional cytogenetic or RFLP analysis. Seven patients achieved greater than 90% donor hematopoiesis 14 to 30 days post-transplant. Four patients had no evidence of donor hematopoiesis. Five patients experienced early relapse (between 43 and 127 days) post-transplant. Six of 15 patients were alive between 34 and 175 days post-transplant. Acute GVHD occurred in three patients. None of the five evaluable patients had evidence of chronic GVHD. Utilizing a busulfan-based preparative regimen, Slavin and colleagues [11] demonstrated excellent tolerability and favorable survival probabilities in 22 patients with hematologic malignancies and four patients with genetic diseases. Preparative therapy consisted of busulfan at a dose of 8 mg/kg, plus fludarabine 180 mg/m2 and anti-T-lymphocyte globulin. Fourteen of the 22 patients with hematologic malignancy had low-risk disease (acute myeloid leukemia in first remission or chronic myeloid leukemia [CML] in chronic phase). G-CSF-mobilized peripheral blood stem cells were used as the source of stem cell support. GVHD prophylaxis consisted of CYA. Twenty-five patients received HLA-genotypically identical sibling donor transplants. One patient received stem cells from a donor with a single antigen mismatch at the A and C locus. Chimerism was evaluated by standard cytogenetic analysis, by analysis of residual Ph+ cells in patients with CML, or by polymerase chain reaction (PCR)-based variable number of tandem repeat (VNTR) sequence analysis in sex matched donor recipient combinations. Treatment was generally well tolerated. All patients had evidence of donor engraftment. In 9 of 26 evaluable patients, transient mixed chimerism was observed. Acute GVHD occurred in 12 of 26 patients. Six patients developed grade III-IV GVHD which was the sole cause of mortality in four patients. Limited chronic GVHD developed in 9 of 25 evaluable patients. At a median of eight months post-transplant, 22 of 26 patients (85%) were alive, 21 of whom (81%) were clinically disease free. Based on their dog model showing that mixed chimerism is reliably achieved following low-dose total body irradiation (200 cGy) and post-transplant immunosuppression (MMF and CYA), McSweeney and colleagues have initiated a clinical trial utilizing this treatment. Forty-four patients with hematologic malignancies, ineligible for conventional allografting due to age, prior therapy or organ dysfunction were treated. Of 42 evaluable patients, all had persistent donor chimerism at two months post-transplant [21]. The regimen was remarkable for its excellent tolerability (lack of significant myelosuppression, gastrointestinal toxicities or alopecia) and the achievement of major disease responses in 70% of the patients. Childs and colleagues described the engraftment kinetics after nonmyeloablative therapy with cyclophosphamide (60 mg/kg/d x 2 days) and fludarabine (25 mg/m2/d x 5 days) and HLA-matched or one-antigen mismatched donor peripheral blood stem cell transplantation for hematologic malignancy (n = 8) or solid tumors (melanoma, n = 4; renal cell cancer [RCC], n = 3). Patients who achieved mixed chimerism were eligible for monthly escalating doses of DLI. Full donor T-cell engraftment was more rapid than donor myeloid engraftment and was a prerequisite for acute GVHD and disease regression. Ten patients had a response (three patients with metastatic RCC; two with CML in chronic phase; one with chronic myelomonocytic leukemia; one with myelodysplastic syndrome; one with diffuse large-cell NHL; one with melanoma; and one with extramedullary plasmacytoma), five of whom were in a sustained complete remission (one metastatic RCC, three with CML and one with NHL). Eight were alive between 121 and 409 days post-transplant [24]. Utilizing a similar nonmyeloablative preparative regimen to that used in the murine model of Sykes and colleagues [14, 18], we have reliably induced mixed lymphohematopoietic chimerism in patients with chemoradiotherapy refractory hematologic malignancies [12, 13]. Preparative therapy consisted of cyclophosphamide 150-200 mg/kg, antithymocyte globulin (ATG) 15-30 mg/kg/d on days 2, 1 and +1 or 1, +1, +3 and +5, and thymic irradiation in patients who had not previously received mediastinal radiation therapy. Post-transplant CYA was given as GVHD prophylaxis. Twenty-eight patients received an HLA-matched (HLA genotypically matched in 27 and phenotypically matched in 1) donor transplant while 16 received an HLA-mismatched donor transplant (one HLA 3-antigen, 11 HLA 2-antigen, and four HLA 1-antigen mismatch). Therapy has been well tolerated. Cyclophosphamide-induced cardiac toxicity has been observed in six patients and has been reversible in five. Severe ATG toxicities at a dose of 30 mg/kg in two patients necessitated a dose reduction to 15 mg/kg. Chimerism assays were performed using PCR analysis of VNTR or short tandem repeat sequences. In recipients of HLA-mismatched donor transplants, flow cytometric analysis of chimerism was performed using allele specific anti-HLA monoclonal antibodies (mAbs). Analyses were performed on weekly peripheral blood samples through day 100 post-transplant (then every six months) and on day +28, +100 and yearly bone marrow samples. Of 23 evaluable recipients of HLA-matched donor transplantation, 20 have achieved stable mixed lymphohematopoietic chimerism. Ten patients with stable mixed chimerism, who had no evidence of GVHD, received DLI beginning on day +35 post-transplant. Conversion of mixed chimerism to fully donor hematopoiesis has occurred in six of the ten patients. Full donor T-cell chimerism was not necessary for the development of acute GVHD (or antitumor response). Seven of 10 evaluable recipients of HLA-mismatched donor marrow transplants have achieved stable mixed chimerism. Flow cytometric determinations of cell subsets exhibit discrepancies in the level of chimerism in various lineages. This "split lineage" chimerism has been seen in all evaluable recipients of HLA-2 antigen mismatched donor marrow transplants. Because of the presence of acute GVHD, none of the recipients of HLA-mismatched donor marrow received prophylactic DLI. Striking antitumor responses have been seen in the majority of these patients with refractory hematologic malignancies, including three patients who received a prior autologous stem cell transplant. Of 23 evaluable patients with chemorefractory Hodgkin's disease or NHL, seven (29%) achieved a partial remission and eight (33%) a complete response. Twenty-two patients are presently alive. Thirteen of these 22 patients are evaluable for response, and eight are clinically progression free. Notably, of the 10 patients who received delayed DLIs beginning at day +35 post-transplant for conversion of their chimerism (and enhancement of their graft-versus-malignancy effect), six patients have achieved a complete remission and seven patients remain progression free. Thus, the nonmyeloablative regimen that we have employed has been associated with a low incidence of acute GVHD following HLA-matched donor allogeneic BMT and in approximately one-half of the patients, a state of stable mixed chimerism without GVHD is induced allowing for the early administration of DLI. The majority of these patients have achieved a complete remission and are progression free. Several patients have had conversion of chimerism and achievement of a complete remission without the development of severe GVHD, confirming the findings in the Sykes murine model in which a potent lymphohematopoietic GVHD could be achieved following delayed DLI without clinical GVHD. We have also demonstrated for the first time that stable mixed lymphohematopoietic chimerism can be achieved with this regimen following HLA-mismatched donor allogeneic BMT. The presence of GVHD, although manageable in most cases, has impeded the use of delayed DLI for chimerism conversion and enhancement of GVL effect in this population. This GVHD is believed to result from suboptimal in vivo donor T-cell depletion with ATG. A clinical trial evaluating a novel anti-CD2 mAb, which has been shown in preclinical models to affect a profound and sustained T-cell depletion, is in progress.
Given the excellent tolerance of these nonmyeloablative regimens and the high rates of alloengraftment, even following transplants from HLA-mismatched donors, there has been considerable interest in extending these transplant strategies to patients with nonmalignant disease [11, 25, 26]. These include the genetic diseases described by Slavin et al. (beta-thalassemia major, Fanconi's anemia, Blackfan Diamond anemia and Gaucher's disease) [11]. In a separate report [25] the transplant group from Hadassah University Hospital reported a child with Fanconi's anemia and leukemic transformation who underwent successful transplantation following a nonmyeloablative conditioning regimen consisting of fludarabine, cyclophosphamide and ATG. Two patients with primary T-cell immunodeficiency have been described who received only post-transplant immunosuppression with MMF and CYA [26]. Stable multilineage mixed chimerism was seen in both patients. Both patients developed grade II acute GVHD that responded to prednisone therapy. Studies of immune reconstitution in one patient showed a significant increase in numbers of T-cells, T-cell subsets and T-cell proliferative responses in vitro. We have shown that stable mixed erythroid chimerism is achievable following ABO mismatched BMT, raising the prospect of mixed chimerism transplant approaches for disorders such as sickle cell anemia and the thalassemias. Given the therapeutic dilemmas that have surrounded the application of allogeneic BMT for conditions such as sickle cell anemia and thalassemia major, particularly the early transplant-related mortality risk among a group of patients who may have prolonged survival with medical therapy alone, these nonmyeloablative transplant approaches may have particular benefit. These advantages include a low transplant-related mortality risk, the development of stable mixed erythroid chimerism and the lack of a need to enhance a GVL effect by giving delayed DLI , all of which make this strategy particularly attractive for nonmalignant disorders. Another potential application of nonmyeloablative transplant strategies and induction of mixed chimerism is the induction of donor-specific tolerance for solid organ grafts [27, 28]. We recently treated a 56-year-old patient with end-stage renal disease secondary to multiple myeloma with a combined HLA-matched bone marrow and sibling donor bone marrow and kidney transplant following a preparative regimen consisting of cyclosphosphamide (120 mg/kg), thymic irradiation and peri-transplant ATG [28]. The patient achieved normal renal function, which has been sustained approximately ten months after withdrawal of all immunosuppression. While donor chimerism was lost after day +105 post-transplant, the patient remains without evidence of disease progression approximately one year post-transplant. This patient has, moreover, had no evidence of acute or chronic GVHD. Based on a series of animal models demonstrating that donor-specific tolerance can be induced for solid organ allografts following the achievement of mixed lymphohematopoietic chimerism, several clinical trials are under way to further test this hypothesis. Since sustained tolerance can be induced even in the presence of transient hematopoietic chimerism, delayed DLIs may be necessary only in the situation of an underlying malignant disease (where optimal GVL effects are desirable).
There have been many fascinating observations, both in preclinical models and clinical trials, about the safety and efficacy of nonmyeloablative allogeneic stem cell transplant strategies. The achievement of stable mixed lymphohematopoietic chimerism, even following transplants from HLA-mismatched donors, and the use of this mixed chimerism as a platform for subsequent adoptive cellular immunotherapy, mark significant advances in the field of transplantation, patient biology and immunology. However, there are likely as many questions remaining regarding the utility of these approaches as have been answered. These questions and the studies designed to answer them include:
1.Is there a true distinction between myeloablative and nonmyeloablative regimens?
2.Has the optimal nonmyeloablative regimen been determined?
3.Is mixed chimerism beneficial (or necessary) for GVHD prevention and optimal antitumor effect?
4.What is the optimal source of stem cells?
5.What is the overall efficacy of nonmyeloablative transplant strategies versus conventional allogeneic BMT?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||