Chronic GVHD of the CNS

In this issue of Blood, Adams et al¹ present a novel mouse model of central nervous system (CNS) chronic graft-versus-host disease (GVHD) and identify unique mediators of this understudied complication after allogeneic hematopoietic stem cell transplantation (HSCT).

Allogeneic HSCT is a curative treatment choice for a wide variety of hematological malignant and nonmalignant diseases. As many as 60% of adult HSCT survivors develop symptoms related to neurocognitive dysfunction.² Even though studies have shown that patients with GVHD have increased risk of neurocognitive dysfunction,^3,4 it is currently controversial whether the CNS is a direct target organ of GVHD in humans.^5,6 Because many drugs used in conditioning regimens, immunosuppressants, and infections of the CNS are associated with increased risk of neurocognitive toxicities, it has been difficult to distinguish the direct effects of GVHD on the CNS from those resulting from other complications of HSCT.³ 

Previously published data from both murine and nonhuman primate models of acute GVHD showed that alloreactive T cells could infiltrate the CNS and cause neuronal damage.^7,8 Using multiple tests, Adams et al detected significant behavioral changes in mice with chronic GVHD compared with GVHD-free controls in 2 different models. Like acute GVHD, increased mobility deficits as detected by the forced swim test and learning deficits as revealed in active place avoidance tasks were observed in mice with chronic GVHD. However, the impaired recognition memory, reduced exploratory behavior, and increased anxiety observed in mice with acute CNS GVHD^8,9 were not detected in mice with chronic GVHD.

Behavioral changes could be due to inflammation in the CNS associated with chronic GVHD, as shown in the study by Adams et al. It was observed that T-cell infiltration in the brain was initially dominated by CD8⁺ T cells and slowly transitioned to CD4⁺ T cells over time in chronic GVHD. Instead of increased expression of tumor necrosis factor (TNF) and interleukin-6 during acute GVHD,^8,10 persistent upregulation of interferon-γ (IFN-γ) and CC motif chemokine ligand 2 was observed in brains of mice with chronic GVHD. Flow cytometric and immunofluorescent analyses revealed that transient microglia cell activation and infiltration of donor major histocompatibility complex (MHC) class II–expressing macrophages were 2 major characteristics of chronic CNS GVHD. RNA sequencing analyses of donor bone marrow–derived macrophages and host microglia from chronic GVHD brains demonstrated distinguishable transcriptional profiles of these 2 cell subsets. Whereas host microglia returned to a homeostatic state, donor bone marrow–derived macrophages were still highly activated 70 days after transplantation. Of interest, genes related to IFN-γ signaling were highly upregulated in donor-derived macrophages but not host microglia. Decreased neuroinflammation and behavioral changes observed in recipients of MHC class II knockout stem cell grafts suggest that MHC class II expression in donor macrophages plays a critical role in the development of chronic CNS GVHD. Even though the critical role of IFN-γ in chronic CNS GVHD should be further investigated by genetic and pharmaceutical approaches, the data presented in this study strongly suggest that IFN-γ is a mediator of MHC class II expression by donor-derived macrophages. The finding that donor bone marrow macrophages are the culprit behind chronic CNS GVHD is in stark contrast to the situation in acute GVHD of the CNS, in which effect cytokines (eg, TNF) are produced by host microglial cells.⁸ 

This is a timely report, because we are just beginning to appreciate the effect of alloresponses on the CNS. The unique findings associated with chronic GVHD as revealed by Adams et al¹ (see figure) suggest that the pathogenesis of the CNS damage in chronic GVHD is different from that in the acute setting. Even though a significant amount of work will be needed for us to understand how exactly these immune attacks lead to synapse damage and neurocognitive dysfunction, the data presented in this report lay the foundation for future studies of chronic CNS GVHD.