Definition of a Diagnostic Window Prior to the Onset of Clinically Apparent Acute Graft-Versus-Host Disease.

Abstract 3746

Acute graft-versus-host disease (aGvHD) is an immune syndrome after allogeneic hematopoietic cell transplantation (allo-HCT) caused by alloreactive donor T cells that attack the gastrointestinal tract, liver and skin. Thus, early T cell migration patterns to these organs could provide first cues for the onset of aGvHD. Hence, a unique surface marker profile of donor T cells at early time points after allo-HCT may be an indicator for patients at risk of aGVHD. Therefore, we analyzed the course of donor T cell activation, proliferation and homing in a clinical relevant murine MHC minor mismatch (miHAg) allo-HCT model to define critical time points and marker profiles for the detection of alloreactive T cells.

Luciferase-labeled C57Bl/6 (H-2b) T cells plus bone marrow cells were transplanted into conditioned (8 Gy) MHC major mismatched Balb/c (H-2d) or miHAg Balb/b (H-2b) recipients. Donor T cell migration was visualized by in vivo bioluminescence imaging (BLI) and cells were characterized by multiparameter flow cytometry for 30 consecutive days after allo-HCT. GVHD scoring was performed by histopathology. Donor T cells proliferated exclusively in secondary lymphoid organs until day+3 (initiation phase) before migrating via the peripheral blood into target organs (effector phase). This occured in both models, MHC major mismatch and miHAg allo-HCT, which resulted in hyper-acute (starting at day+6) or acute GVHD (starting at day+21), respectively. In the hyper-acute scenario one wave of T cell migration starting at day+4 sufficed to cause lethal aGVHD. We detected a 4000-fold increase in CD4 and a 1500-fold increase in CD8 donor T cell numbers in the peripheral blood between day+3 and day+6 in this model. In contrast, in the more clinical relevant miHAg allo-HCT model we found 3 waves of T cell migration with peaks at days +6, +11 and +15 after allo-HCT. In the peripheral blood CD4 T cells increased 20-fold, CD8 T cells 50-fold between day+3 and day+6, but more than 40-fold (CD4) and 400-fold (CD8) between day+3 and day+11. After the third peak on day+15 a period followed when we could only detect very few migrating donor T cells in the peripheral blood before aGvHD became clinically apparent on day+21. Next, we asked whether we could identify alloreactive T cells by testing a large panel of surface markers at the defined migration peaks. Indeed, allogeneic T cells upregulated certain homing receptors at these peaks (e.g. at day+11: α4β7 integrin: 27% of CD4 T cells, 3.4×104/ml, 60% of CD8 T cells, 1.6×105/ml; P-selectin ligand: 28% of CD4 T cells, 3.5×104/ml, 35% of CD8 T cells, 9.1×104/ml). In contrast, syngeneic transplanted mice only showed a constant low expression level of those receptors (e.g. at day+11: α4β7 integrin: 20% of CD4 T cells, 9.6×103/ml, 5% of CD8 T cells, 3.1×103/ml; P-selectin ligand: 17% of CD4 T cells, 8.5×103/ml, 10% of CD8 T cells, 6.6×103/ml). However, other markers such as CD44 could be found on more than 80% of all donor T cells in allogeneic or syngeneic recipients.

Our results in this clinical relevant mouse model show accelerating waves of T cell migration consistent with an enhancing feedback loop model of aGvHD pathogenesis. The homing receptor expression profile of donor T cells correlated with critical migration waves and clearly differed between mice with or without aGvHD. The assessment of critical time points frame a diagnostic window for a potential predictive test based on the dynamic change of the T cell homing receptor profile after allo-HCT. This preclinical study now awaits to be evaluated in patients undergoing allo-HCT.