Exciting new murine model of cGVHD

In this issue of Blood, Srinivasan et al report that mice conditioned for allogeneic hematopoietic stem cell transplantation with cyclophosphamide and total body irradiation develop wide-spectrum manifestations of chronic graft-versus-host disease (cGVHD).¹  This represents an exciting new preclinical model that can be used to uncover mechanisms of cGVHD and test interventional therapies.

While preclinical murine animal models of human cGVHD have provided insights to clinical disease, each model usually only simulates a limited number of the complex manifestations that are observed in patients (reviewed in Chu and Gress² ). In addition, some of the models do not use any pretransplantation conditioning. When the model used by Srinivasan et al here was introduced in 2007, where mice preconditioned with cyclophosphamide and total body irradiation were transplanted with small numbers of T cells and bone marrow, the only manifestations reported were pathology and tissue fibrosis in the lungs.³  Lung tissue damage in these mice resulted in the development of bronchiolitis obliterans, which is pathopneumonic for cGVHD.⁴  On further examination, Srinivasan et al now report that these mice also develop pathologic manifestations in several other organs including liver, tongue, thymus, colon, and spleen.¹  This model provides a new preclinical system to study cGVHD pathogenesis and treatment strategies.

The most interesting observations in this paper are that donor-derived alloantibody (IgG) is required for the development of cGVHD (and bronchiolitis obliterans) and that disruption of germinal center formation, by inhibiting lymphotoxin-β (LTβ) signaling, is capable of treating “established” cGVHD. Because cGVHD is associated with the presence of host-reactive antibodies,⁵  and some cGVHD patients have responded to treatment with anti-CD20 monoclonal antibodies,⁶  the murine model further characterized in this issue could prove to be of value for assessing various therapies that directly or indirectly target antibody production.

Srinivasan et al focused much of their work on examining two of the organs most severely affected in this model, the lungs and liver, where they observed IgG deposition as well as the co-infiltration of CD4 T cells and B cells. These observations prompt several follow-up questions regarding this model (please refer to figure). First, is IgG deposition limited to the lungs and liver or does it occur in other organs? Second, what is the mechanism of IgG-mediated pathology? Is the tissue damage because of complement fixation or through Fc receptor-expressing immune effector cells? Third, is the tissue-reactive IgG specific for alloantigens, autoantigens, or both? Autoantibodies have been observed in cGVHD patients and in other cGVHD mouse models^2,5  but the role of alloantigen-specific Ig is unclear. Fourth, because co-localization of CD4 T cells and B cells was observed, is it possible that tertiary lymphoid tissues develop in the affected tissues, contributing to the alloantibody production? The formation of tertiary lymphoid tissues has been observed at sites of inflammation in mice and humans,^7,8  and LTβ is also important for the formation and maintenance of these tissues. Finally, because CD40/CD40L and ICOS/ICOSL interactions are also known to be important for germinal center formation,^9,10  is it possible that these pathways could also be targeted to treat cGVHD in these mice?

In summary, while there are no perfect models of clinical cGVHD, murine models such as the one reported by Srinivasan et al in this issue can provide important insights to clinical disease. This exciting model could prove to be an ideal system for testing new cGVHD treatment strategies.