The Relationship between Host Antigen-Presenting Cells and Graft-versus-Host Responses Following Human Nonmyeloablative Transplantation.

Host-derived antigen presenting cells (APCs) are required for the induction of graft-versus-host disease (GVHD) in murine models of allogeneic hematopoietic stem cell transplantation. Host APCs are also important for graft-versus-tumor (GVT) responses that occur without GVHD in pre-clinical models of delayed donor leucocyte infusions (DLI). Nonmyeloablative transplantation has the theoretical advantage that host APCs will persist at the time of DLI or withdrawal of immunosuppression, permitting the induction of GVT without GVHD. To evaluate this, we have prospectively evaluated peripheral blood (pb) dendritic cell (DC) chimerism in 16 adult patients following nonmyeloablative transplantation with or without subsequent DLI or rapid withdrawal of immunosuppression (IS). T-cell (CD3+), granulocyte, CD1c+ myeloid DC (mDC) and BDCA-4+ plasmacytoid DC (pDC) precursors were isolated to high purity by immuno-magnetic selection from pb at 1, 2, 3, 6, 9, 12, 18 and 24 months (m) post-transplant. Short tandem repeat PCR or FISH XY analysis of sorted fractions was performed. 15 patients received in vivo CAMPATH 1H (Cam). Conditioning regimens were BEAM/Cam (n=8), Fludarabine/Melphalan/Cam (n=7) and Fludarabine/Cyclophosphamide (n=1). GVHD prophylaxis was cyclosporine (n=8) or cyclosporine/methotrexate (n=8). Diagnoses were lymphoproliferative disease (n=13), AML/MDS (n=2), and renal cell carcinoma (n=1). All patients received unmanipulated donor PBSC (HLA-matched sibling n=15, HLA-matched unrelated n=1). Median follow-up is 9 m (range 1–24m). Host-derived pb DC precursors were present in 50% of patients at 1m, 30% at 3m and 33% at 6m. Mean ±SEM % host mDC were 6.4 ±4 at 1m, 1.9 ±1.3 at 3m and 3.1 ±2.3 at 6m. Mean ±SEM % host pDC were 5.6 ±3.6 at 1m, 0.85 ±0.4 at 3m and 2.8 ±1.6 at 6m. The origin of both DC precursor subsets closely paralleled that of the granulocyte fraction, even in the case of a significant lag in donor T-cell engraftment. To date, 8 patients have received either DLI (n=5) or rapid tapering of IS (n=3) for disease progression or relapse (n=6), mixed T-cell chimerism (n=1) or persistent disease (n=1). Of 7 evaluable patients, 4 had detectable pb host DC precursors prior to the intervention. Of these, 3 developed acute or chronic GVHD following DLI or IS taper, 2 converted to full donor chimerism and 1 patient with T-NHL entered complete remission. Of the 3 patients without detectable host DC, no GVHD was observed, 1 patient with persistent mixed T cell chimerism switched to full donor but no GVT responses were observed. We conclude that persistence of pb host DC precursors is evident up to 6 months in a significant proportion of patients following nonmyeloablative transplantation, although there is substantial inter-patient variability. Changes in pb DC precursor chimerism closely follow myeloid chimerism and are independent of changes in T cell chimerism. This is consistent with a myeloid ontogeny of pDC precursors in humans. These preliminary findings suggest that the persistence of host DC may be associated with a risk of GVHD following DLI or IS withdrawal. Further studies that assess in parallel the origin of tissue DCs and the relationship between the presence of host APCs and GVH responses post-transplantation are required.