Background: Graft-vs-host disease (GVHD) is a major complication of allogenic bone marrow transplantation (BMT). Targeting costimulatory molecules with antagonist antibodies could dampen the excessive immune response that occurs, while preserving the beneficial graft vs leukemia (GVL) of the allogeneic response. Previous studies using a mouse model of GVHD have shown that targeting the T-cell Inducible COStimulator (ICOS, CD278) molecule is beneficial, but it is unclear whether the same applies to human cells.
Methods: Here, we assessed whether a monoclonal antibody (mAb) to human ICOS was able to antagonize the costimulatory signal delivered in vivo to human T cells. To test this hypothesis, we used a xenogeneic model of GVHD where human peripheral blood mononuclear cells were adoptively transferred in immunocompromised NOD.SCID.gc-null mice (NSG).
Results: In this model, control mice invariably lost weight and died by day 50. In contrast, 65% of the mice receiving a single injection of the anti-hICOS mAb survived beyond 100 days. Moreover, a significant improvement in survival was obtained in a curative xeno-GVHD setting. Mechanistically, administration of the anti-hICOS mAb was associated with a strong reduction in perivascular infiltrates in liver and lungs and reduction in frequencies and numbers of human T cells in the spleen. In addition, the mAb prevented T-cell expansion in the blood during xeno-GVHD. Importantly, GVHD-protected mice retained the ability to control the P815 mastocytoma cell line, mimicking GVL in humans.
Conclusion: A mAb-targeting human ICOS alleviated GVHD without impairing GVL in a xenograft murine model. Thus, ICOS represents a promising target in the management of BMT, preventing GVHD while preserving GVL.
Keywords: CD278; ICOS; NSG; antibodies; graft vs leukemia effect; graft-vs-host disease; mice; monoclonal.