TRAIL/DR5 Interactions Are Important For Mediating Thymic Damage After Allogeneic Bone Marrow Transplantation

Abstract

Conclusions: Progressive improvement in cutaneous manifestations of cGVHD was observed following a 24-week course of ECP in pts who previously had exhibited no clinical response or worsening of cGVHD while receiving standard immunosuppressive therapy with corticosteroids. ECP improved skin GVHD and allowed for tapering of corticosteroid doses to ''low risk'' levels of \10 mg prednisone-equivalents/day. Donor T cell alloreactivity can be co-opted to deliver a graft-versustumor (GVT) response following blood or bone marrow transplantation (BMT). However, the major reason for treatment failure following BMT is tumor recurrence, suggesting a long-term failure of GVT immunity. In this study, we have considered the role of non-hematopoietic antigen in determining the fate of donor CD8 cells in a model of delayed DLI to partially MHC-mismatched chimeras, where antigen was either expressed ubiquitously or restricted to the hematopoietic compartment. We observed that donor CD8 cytotoxic and cytokine responses were poorly sustained (at .8 weeks) in the presence of non-hematopoietic antigen and this was associated with a failure to establish a central memory (T CM ) population. This effect occurred at two distinct levels. Firstly, residual donor CD8 cells demonstrated a classic PD-1 high CD127 NON-HEMATOPOIETIC ANTIGEN BLOCKS EARLY MEMORY IMPRINTING OF GRAFT-VERSUS-HOST REACTIVE CD8 CELLSlow 'exhaustion' signature and consistent with this, CD8 functions were partially restored by blockade of the PD-1 pathway. Secondly, during the initial phase of the primary response (at day 5-8 following transfer of CFSE-labelled splenocytes), we observed that the proportion of post-mitotic, CFSE low CD8 cells expressing a CD44 high CD62Lhigh phenotype was much lower in chimeras where antigen was ubiquitous. These findings suggested the possibility that early CD8 encounter with non-hematopoietic antigen might severely disable memory precursor formation during the initial phase of the response. In order to test this concept, we purified CD8 T cells using congenic markers from the two sets of chimeras 14 days after initial T cell transfer, and then parked them in secondary, antigen free hosts. After 21 days, we tested recall responses to alloantigen. We found that priming in the presence of non-hematopoietic antigen severely impaired the establishment of recall immunity in the absence of antigen. Furthermore, while a significant proportion of CD8 T cells primed in the absence of non-hematopoietic antigen had a (T CM ) phenotype in secondary hosts, this population was almost entirely absent when initial priming occurred in primary hosts where antigen was ubiquitous. These data demonstrate that alloantigen within non-hematopoietic tissues is not ignored, but rather actively inhibits both the formation and maintenance of CD8 T cell memory. While these effects may lessen the risk of GVHD, they might also impair the durability of the GVT response. Thymic GVHD (tGVHD) is associated with prolonged immunodeficiency. We previously found that thymic ...