Pathophysiologic mechanisms of acute graft-vs.-host disease.

Graft-vs.-host disease (GVHD) remains the major toxicity of allogeneic bone marrow transplantation. Mechanistic studies in experimental animal models provide a better understanding of the complex relationships and cascade of events mediated by cellular and inflammatory factors. Also, advances in basic immunology have cleared the way for a more precise view of allogeneic reactions between donor and host. In addition, the use of mutant mice lacking critical cytolytic proteins has helped map out the molecular pathways by which GVHD targets organ damage. In this article, these mechanisms are reviewed and synthesized into a coherent conceptual framework, providing a state-of-the-art summary of the pathophysiology of acute GVHD.     

Cyclosporine, methotrexate, and prednisone compared with cyclosporine and prednisone for prevention of acute graft-vs.-host disease: effect on chronic graft-vs.-host disease and long-term survival.

Graft-vs.-host disease (GVHD) is a major predictor of outcome following allogeneic bone marrow transplantation (BMT). For patients alive at day 100 after BMT, the presence or absence of chronic GVHD is one of the most important determinants of survival and quality of life. We wished to determine the effects on chronic GVHD of two regimens used for the prophylaxis of acute GVHD: cyclosporine, methotrexate, and prednisone (CSA/MTX/PSE) and cyclosporine and prednisone (CSA/PSE). One hundred forty-nine evaluable patients were entered into the acute GVHD study. As of 31 March 1997, 63 months after the last patient underwent BMT, the median survival time was 4.5 years (range 0.09-9.9). The incidence of chronic GVHD was independent of the prophylactic regimen (55 vs. 54%), and extensive chronic GVHD occurred in 25 and 24% of patients receiving CSA/MTX/PSE and CSA/PSE, respectively. Of note, the median Karnofsky performance status of both groups was 100% (range 70-100%), reflecting the low incidence of extensive chronic GVHD. Survival rates free of chronic GVHD were 52 vs. 42% (p = 0.29) for patients receiving CSA/MTX/PSE vs. CSA/PSE. The incidence of relapse was also similar in both groups of patients. These data suggest that the combinations of CSA/MTX/PSE and CSA/PSE result in comparable chronic GVHD-free survival without an increase in leukemic relapse.     

Separation of graft-vs.-tumor effects from graft-vs.-host disease in allogeneic hematopoietic cell transplantation

Allogeneic hematopoietic cell transplantation (HCT) is an increasingly widely used treatment modality in hematological malignancies. Alloreactivity mediated by donor T cells (and, in some settings, by donor natural killer cells) can produce durable immunologic control or eradication of residual malignancy after allogeneic HCT. However, graft-vs.-tumor (GVT) effects are variably effective and are often accompanied by deleterious alloreactivity against normal host tissue, manifesting as graft-vs.-host disease (GVHD). A major focus of current research in HCT is the separation of beneficial GVT effects from GVHD. Here we review a number of approaches currently under investigation to specifically augment GVT effects, including the identification of minor histocompatibility antigens (mHA), adoptive immunotherapy with tumor-specific or mHA-specific cytotoxic T lymphocytes, vaccination of the donor or recipient to stimulate tumor-specific immunity, and adoptive transfer of natural killer cells. In addition, we review strategies being investigated to specifically suppress GVHD while sparing GVT, including the manipulation and infusion of regulatory T cells, the use of novel pharmacologic and biologic agents, and the use of mesenchymal stem cells. Ultimately, advances in separation of GVT from GVHD will further enhance the potential of allogeneic HCT as a curative treatment for hematological malignancies.     

Nitric oxide mediates intestinal pathology in graft-vs.-host disease.

We have investigated the involvement of nitric oxide (NO) in intestinal graft-vs.-host reaction (GvHR) in mice. Treatment of mice with L-NG-monomethyl arginine (L-NMMA), a specific inhibitor of NO synthesis, abolished the mucosal pathology of intestinal GvHR and reduced the associated lymphocytic infiltration of the epithelium. L-NMMA had no effect on splenomegaly in GvHR, nor did it interfere with the growth of an undifferentiated crypt stem cell line, or the production of tumor necrosis factor-alpha by activated macrophages in vitro. In contrast, L-NMMA inhibited the enhanced activity of natural killer (NK) cells which occurs in GvHR. We conclude that a NO-dependent mechanism is essential for intestinal immunopathology in GvHR and that this may reflect a role for NO in NK cell function.     

Translational opportunities for targeting the Th17 axis in acute graft-vs.-host disease

Allogeneic stem cell transplantation (allo-SCT) is a curative therapy for different life-threatening malignant and non-malignant hematologic disorders. Acute graft-vs.-host disease (aGVHD) and particularly gastrointestinal aGVHD remains a major source of morbidity and mortality following allo-SCT, which limits the use of this treatment in a broader spectrum of patients. Better understanding of aGVHD pathophysiology is indispensable to identify new therapeutic targets for aGVHD prevention and therapy. Growing amount of data suggest a role for T helper (Th)17 cells in aGVHD pathophysiology. In this review, we will discuss the current knowledge in this area in animal models and in humans. We will then describe new potential treatments for aGVHD along the Th17 axis.     

Protection from lethal graft-vs.-host disease by donor stem cell repopulation.

Graft-vs.-host reaction (GVHR) induced in non-irradiated F1 mice with DBA/2J parental spleen and lymph node (LN) cells usually does not lead to acute GVH disease (GVHD). This contrasts with the GVHR induced in other parent-F1 combinations involving both major histocompatibility complex (MHC) class I and class II differences between donor and host. Most signs of acute GVHD in non-irradiated F1 mice relate to immunodeficiency following destruction of the lymphohemopoietic system of the host, which leads to wasting and death due to infections. This sequence of events is prevented when donor lymphoid cells, originating from grafted stem cells, repopulate the destroyed lymphohemopoietic system of the host. To examine whether a "silent" repopulation of the F1 host by donor stem cells might underly the absence of clinical signs of acute GVHD when GVHR is induced with DBA/2J lymphoid cells, GVHR was induced with LN cells, which do not contain stem cells. Indeed, GVHR induced in (C57BL/10 x DBA/2J)F1 (BDF1) mice with 80 x 10(6) DBA/2J LN cells led to acute GVHD. Signs of acute GVHD such as wasting and death did not occur when donor stem cells, from an inoculum of DBA/2J spleen and LN cells, were allowed to repopulate the lymphohemopoietic system of the host. The effect of donor stem cells on clinical signs of acute GVHD was more apparent when (B10.D2 x DBA/2J)F1, instead of DBA/2J, lymphoid cells were used to induce GVHR. The detection of alloreactive anti-host cytotoxic T lymphocyte (CTL) activity during acute GVHD induced with DBA/2J donor lymphoid cells supports the hypothesis that such CTL contribute to the destruction of the host immune system in acute GVHD.     

The induction of organ-specific antibodies during the graft-vs.-host reaction

During the parent (P) into F1 hybrid graft-vs.-host reaction (GVHR), nuclear, leukocyte and erythrocyte autoantibodies are commonly seen. The specificity of these autoantibodies is reminiscent of those found in systemic lupus erythematosus (SLE) patients and SLE-prone mice. Organ-specific antibodies, however, including thyro-globulin (Tg) antibodies do not arise spontaneously. There have been conflicting reports about the ability of exogenous Tg to induce an anti-Tg response during the GVHR. We have re-examined this question in greater detail. Using the murine P----F1 GVHR system, the results of this work demonstrate that mouse thyroglobulin (MTg)-specific antibodies can be induced during a GVHR. However, mice must both be undergoing a GVHR, and have received exogenous MTg. The highest autoantibody response occurs if mice are injected with mouse thyroid extract or purified MTg at the time of P----F1 cell transfer. The anti-MTg response is MTg dose dependent. The ability to induce anti-MTg antibody was not major histocompatibility complex restricted, for both the DBA/2----B6D2F1 (low responder H-2 haplotypes to MTg), and AKR or DBA/2----AKD2F1 (high/low responder----high responder haplotype) GVHR gave similar responses. The anti-MTg titers peaked between days 7-10 and declined thereafter. In contrast, antibodies to dsDNA were not present at this early time, but developed after several weeks. We conclude that organ-specific autoantibodies can be induced during a GVHR if the appropriate antigen(s) are presented near the time of GVHR induction.     

Adoptive transfer of murine syngeneic graft-vs.-host disease by CD4+ T cells

Syngeneic graft-vs.-host disease (SGVHD) develops in rodents following the treatment of lethally irradiated, bone marrow (BM) reconstituted animals with a short course of the immunosuppressive agent cyclosporine A (CsA). Using an in vivo depletion approach, we recently demonstrated that CD4(+), but not CD8(+), T cells participated in inducing SGVHD. Studies were therefore undertaken to adoptively transfer SGVHD into lethally irradiated, syngeneic BM reconstituted secondary recipients. Whole T cell populations as well as purified CD4(+)T cells isolated from SGVHD, but not normal or transplant control, animals mediated the transfer of SGVHD into secondary recipients. These cells have an apparent specificity for enteric bacterial antigens. The pathologic process that developed was identical to that observed in the animals with de novo SGVHD after syngeneic BMT and CsA therapy. It was shown that a radiation-sensitive mechanism prevented the transfer of SGVHD into normal, nonirradiated secondary recipients. The ability to reproducibly transfer SGVHD into secondary recipients will enhance our ability to study regulatory mechanisms that are altered during CsA therapy and permit the development of murine CsA-induced SGVHD.     

Graft-facilitating doses of ex vivo activated gammadelta T cells do not cause lethal murine graft-vs.-host disease.

The purpose of this study was to examine the ability of gamma(delta) T cells to cause graft-vs.-host disease (GVHD) after allogeneic bone marrow transplantation (BMT) and to determine whether these cells offered any therapeutic advantages relative to alphabeta T cells. Due to the paucity of naive gamma(delta) T cells in mice and humans, gamma(delta), T cells (obtained from alpha(beta) T cell-deficient murine donors) were ex vivo activated and expanded in interleukin (IL)-2 so as to achieve sufficient cell numbers and to serve as a more clinically feasible strategy. After transplantation into lethally irradiated hosts, donor gamma(delta) T cells were detected in target organs of GVHD such as the spleen and intestines 2 weeks after BMT and constituted the primary T cell subpopulation. Large doses (150 x 10(6)) of activated gamma(delta) T cells, which we have previously shown capable of facilitating engraftment in MHC-disparate recipients, failed to cause fatal GVHD in lethally irradiated recipients of MHC-incompatible donor marrow grafts (C57BL/6 [H-2b]-->B10.BR [H-2k] and C57BL/6 [H-2b]-B6D2F1[H-2b/d]). The absence of GVHD was confirmed by histologic analysis of target organs, splenic B cell reconstitution, and appropriate negative selection in the thymus, that were all comparable to those observed in mice transplanted with T cell-depleted BM only. While early splenic reconstitution was attributable to donor gamma(delta) T cells, analysis of durably engrafted chimeras 2 months posttransplant revealed that the vast majority of donor splenic T cells expressed the alpha(beta) T cell receptor. The results of secondary adoptive transfer assays showed that these cells were tolerant of recipient alloantigens in vivo, demonstrating that gamma(delta) T cells did not prevent the subsequent development of donor anti-host tolerance in BM-derived alpha(beta) T cells. When comparatively evaluated, the minimal number of naive alpha(beta) T cells necessary for donor engraftment caused significantly more fatal GVHD than the corresponding minimal dose of activated gamma(delta) T cells and thus had a superior therapeutic index. These studies indicate that doses of activated gamma(delta) T cells that are able to promote alloengraftment do not cause lethal GVHD in mice transplanted with MHC-incompatible marrow grafts.     

Rat anti-glomerular basement membrane antibodies in toxin-induced autoimmunity and in chronic graft-vs.-host reaction share recurrent idiotypes

Cross-reactive idiotypes (CRId) borne on autoanti-glomerular basement membrane antibodies of Brown-Norway (BN) rats with mercury-induced glomerulonephritis have been described in the preceding study (Guéry, J.-C. et al., Eur. J. Immunol. 1990. 20:93). BN rats treated with sodium aurothiopropanol sulfonate or D-penicillamine, as well as (LEW X BN)F1 hybrids transferred with BN rat spleen cells, developed quite similar autoimmune abnormalities. In the present study, it is shown that immunoglobulins bearing such "public" idiotypes are also produced and deposited in the kidney in these three models. The CRId here described may, therefore, be considered as a marker of sets of recurrently expressed V region genes during the course of these autoimmune disorders. Anti-self class II T cells are present in the three models of toxin-induced autoimmunity and anti-allo class II T cells are responsible for the chronic graft-vs.-host reaction. The same B cell clones are probably triggered during these processes as a consequence of a polyclonal B cell activation mediated by anti-class II T cells.     

The effect of T lymphocyte depletion on neonatal tolerance induction,graft-vs.-host disease and cellular chimerism

Treatment with a monoclonal anti-Thy-1 antibody and complement completely prevented C57BL spleen cells from causing graft-vs.-host disease following their inoculation into newborn CBA mice. The proportion of mice that became tolerant to C57BL antigens, as measured by skin grafting, was significantly less compared with mice given (CBA × C57BL)F₁ hybrid cells. This was not due to the elimination of T cells, for antibody-treated F₁ cells induced tolerance as readily as complement-treated control F₁ cells. To investigate whether the apparent superiority of F₁ cells over C57BL cells is attributable to differences in the mechanism inducing and maintaining unresponsiveness, two approaches were followed. First, the level of donor cell chimerism in the spleens of tolerant animals was studied. Though no difference between F₁ and C57BL cells was uncovered, the presence of T cells in the donor inocula favored the establishment of chimerism. Second, the involvement of suppressor T cells was examined in adoptive transfer experiments. Splenic suppressor T cells were associated with tolerance regardless of how it was elicited. Preliminary results with F₁ cells show that the tolerogenic property is not confined to any one cell type. It is proposed that the greater tolerogenicity of F₁ cells is brought about by the presence of host-type (self) antigens, which enable the tolerogenic signals to operate without recourse to antigen processing by host cells.     

Both apoptosis and complement membrane attack complex deposition are major features of murine acute graft-vs.-host disease

The parent-into-F1 mouse model (P-->F1) of acute graft-vs.-host disease (GVHD) is a useful model of human acute GVHD because it allows the study of the T cell contribution to pathology without the complicating effects of conditioning regimens. To determine the similarity of this model to human GVHD, we assessed injury in organs typically involved in human acute GVHD (skin, liver) and less typically involved organs (spleen, kidney, lung). Mice were assessed histologically at early (2 weeks), intermediate (3 months) and late (6 month) time points. Based on the emerging roles of Fas ligand killing and complement deposition in allograft rejection, we correlated the amount of tissue specific TUNEL positive apoptosis and deposition of complement (C5b-9) with histopathologic changes. Our results indicate a striking similarity histologically between acute GVHD occurring in this model and in humans following bone marrow transplant. Moreover, C5b-9 deposition and apoptotic cell accumulation were found to parallel tissue injury in major organs of acute GVHD mice, although not all organs exhibited the same kinetic pattern. These results indicate a role for both adaptive immunity and innate immunity in this model of GVHD and support its use in modeling human acute GVHD in the nonmyeloablative setting.     

Characterization of a subpopulation in neonatal thymus which suppress the graft-vs.-host reaction

Thymus cells from neonatal and infant mice were found to have a high capacity to prevent mortality from acute graft-vs.-host disease as compared with spleen cells from stable radiation chimeras. This suppressive capacity of thymocytes decreases with age after birth as was demonstrated by semi-quantitative cell titrations. This suppressor activity is restricted to syngeneity of the graft-vs.-host disease-including cells. The thymic suppressor cells are Thy-1+ and Lyt-1+ and IgG- and IgM-. They do not agglutinate with peanut agglutinin and have a high electrophoretic mobility. In vitro irradiation experiments showed that the suppressor cells are radiation sensitive. These results are compared with the available information on cells suppressing delayed-type hypersensitivity reactions and those suppressing B cell responses.     

Peripheral blood leukocyte grafts that induce human to mouse graft-vs.-host disease reject allogeneic human skin grafts.

Allogeneic graft-versus-host disease is characterized by skin, gut, and bile duct destruction by relatively few donor type lymphocytes. In contrast, we can now show that human-to-mouse xenogeneic graft-versus-host disease is characterized by vasculitis and tumor-like infiltrations of the murine lymphohemopoietic organs with many human CD25+, HLA-DR+, CD4+ lymphoblasts. Using the technique of serial transplantation, it appears that at least 90% of the human lymphoblasts were unreactive to murine tissues. It is demonstrated consistently that the donor type lymphoblasts induced typical allogeneic rejection of distantly located full thickness human unmatched fetal skin grafts. The fact that the human grafts show primary immune responses in vivo indicates that the graft-versus-host disease murine model may be suitable for vaccination studies.     

Graft-vs.-host and graft-vs.-leukemia reactions after delayed infusions of donor T-subsets.

Infusions of donor leukocytes have been given to allogeneic bone marrow recipients after transplant to treat leukemia relapse. Treatment with these delayed infusions of donor cells has been called delayed or donor leukocyte infusion (DLI). While graft-vs.-host disease (GVHD) has typically been less severe than expected after DLI, it still remains a significant risk factor. Recently, we used a full major histocompatibility complex (MHC)-mismatched model (C57BL/6 into AKR) to determine how increased immunogenetic disparity affects GVH and graft-vs.-leukemia (GVL) reactions after DLI. In contrast to an MHC-matched model (B10.BR into AKR), GVHD was still observed when MHC-mismatched donor T cells were infused 3 weeks posttransplant. Limiting dilution analysis was used to determine the frequency of alloreactive cytotoxic T lymphocytes (CTL) and interleukin (IL)-2-secreting T helper cells in the spleens of MHC-mismatched recipients 7 days after DLI treatment. GVHD correlated with elevated frequencies of alloreactive T-helper cells. One strategy for reducing the severity of GVHD after DLI is the selective administration of CD4 or CD8 T-subsets. Delayed infusion of purified T-subsets 3 weeks posttransplant resulted in significantly less GVHD than infusion of a mixture of the two subsets. No GVH-associated mortality was observed after DLI with purified donor CD4+ T cells. In GVL studies, MHC-mismatched CD8+ T cells were the most potent antitumor effectors against an acute T cell leukemia. The GVL effect of MHC-mismatched T-subsets was compared with that of MHC-matched subsets. When naive MHC-matched cells were given as DLI, depletion of either T-subset eliminated the GVL effect. CD8+ T cells from MHC-matched donors primed against host alloantigens, however, mediated a CD4 (T-helper)-independent GVL reaction. Together, these results suggest that administration of T-subsets can significantly reduce GVHD after DLI without loss of the beneficial GVL effect.     

Ly and Ia phenotype of suppressor T cells induced by graft-vs.-host reaction.

Suppressor T cells were generated in (CBA X C57BL)F1 mice undergoing graft-vs.-host (GvH) reaction and treated with anti-Ly and anti-Ia antisera to determine their membrane antigen phenotype. Pretreatment of CBA-induced GvH suppressor T cells with anti-Ly-1.1, anti-Ly-2.1 or anti-Iak antisera plus complement abrogated their suppressive properties when tested against primed F1 spleen cells in vitro. In contrast, concanavalin A (Con A)-induced suppressor T cells were insensitive to anti-Ly-1.1 serum. It is concluded that GvH-induced suppressor T cells are Ly-1+,2+,3+,Ia+, and thus distinct from Con A or antigen-induced suppressor T cells.     

The effects of cyclosporin on lymphocyte activation in a systemic graft-vs.-host reaction

We have investigated the effects of cyclosporin (CsA) on each of three stages of lymphocyte activation in vivo viz. sequestration of alloantigen-reactive lymphocytes from the circulation into the spleen and lymph nodes, blast transformation and induction of DNA synthesis in the activated cells and release of these cells and their progeny into the circulation. Parental strain lymphocytes injected i.v. into semi-allogeneic rats and recovered from the thoracic duct within 36 h are profoundly unresponsive in a local graft-vs.-host assay to the alloantigens of the F1 hybrid but have normal activity against unrelated alloantigens (negative selection). CsA treatment of the F1 hybrid recipients did not prevent this selective sequestration of antigen-reactive cells. In the untreated F1 hybrid, from 36 h after injection, large numbers of dividing blast cells were released into the lymph. These cells did not appear in the lymph of recipients treated with CsA. However, CsA did not prevent the activation of cells sequestered in the spleen or lymph nodes as assessed by [3H] thymidine incorporation and autoradiography. This unexpected finding suggests that CsA inhibits lymphocyte responses to alloantigens in vivo after DNA synthesis which is a later stage than the in vitro studies have shown.     

Prevention of lethal graft-vs.-host disease by a single low dose injection of anti-T cell monoclonal antibody to the allograft recipients.

We investigated the capacity of monoclonal antibody (mAb) treatment to prevent graft-vs.-host disease (GVHD) in lethally irradiated, allogeneically reconstituted mice, employing anti-T cell (subset) mAb and a fully allogeneic strain combination. In this strain combination, purified CD4+ cells were able to induce a lethal GVH reaction, whereas purified CD8+ cells were not. In the same strain combination, a single intraperitoneal injection of IgG2b anti-Thy-1 mAb, one day after reconstitution, caused a dose-dependent improvement of the survival. A single injection of a dose as low as 12.5 micrograms per mouse was already effective. Intravenous and intraperitoneal administration of the mAb appeared equally effective. For effective prevention of GVHD the treatment could be postponed until the 4th day after transplantation, but treatment delayed until day 6 was no longer effective. Treatment with IgG2b mAb specific for either helper or cytotoxic T cells also led to improvement of GVHD and survival, but was less effective than treatment with anti-Thy-1 mAb. Clinically, there was a difference in the effectiveness of anti-CD4 and anti-CD8 treatment, since symptoms of GVHD started earlier in the anti-CD8 treated group and the survival was better in the anti-CD4 treated group. These results press for prospective clinical studies employing anti-T cell mAb treatment early after allogeneic bone marrow transplantation, especially in HLA mismatched cases.