There have been recent advancements in the treatment of cGVHD; however, as with any disease, prevention is preferable to treatment. T cells have a critical role in the pathogenesis of GVHD, and selective T-cell depletion strategies can greatly reduce the incidence of cGVHD. Such strategies aim to reduce alloreactive T cells by manipulating the graft, whether through CD34 selection, the depletion of naive T cells, or the administration of post-transplant cyclophosphamide, which can eliminate alloreactive T cells early after transplantation and before the development of clinical signs of aGVHD or cGVHD.

However, the use of T-cell depletion to prevent cGVHD is associated with certain limitations. T cells are responsible for immune reconstitution after an allogeneic transplant, and immune reconstitution is necessary to both protect patients from opportunistic infections and mediate the graft-vs-leukemia response that is responsible for a significant percentage of the cures that are observed with allogeneic transplants. As such, although T-cell depletion substantially reduces the rates of GVHD, it can also increase the rates of opportunistic infections and relapse.

Ex vivo CD34-selected T-cell depletion is one approach that has been explored. In a study of more than 500 patients with acute leukemia, prophylaxis with CD34+ selection resulted in lower rates of aGVHD and cGVHD compared with antithymocyte globulin–based prophylaxis. Unfortunately, there is a significant delay in immune reconstitution with CD34 selection, along with an increased risk of opportunistic infections. Moreover, the lower rates of moderate and severe cGVHD do not translate to improved survival, as shown in the BMT CTN trial by Luznik et al (NCT02345850).

The most common approach to prophylaxis has probably been the administration of a short pulse of cyclophosphamide after an allogeneic transplant. With this regimen, patients experience some delays in hematopoietic engraftment, but they have remarkably lower levels of cGVHD without compromising the graft-vs-leukemia effect of the transplant. 

Nonetheless, there has been great interest in graft engineering, building on the concept of selective depletion, where the goals include preserving the memory compartment of T cells without retaining the naive compartment of T cells that are less differentiated and contain the specificities of alloreactive T cells. For example, Bleakley et al have shown that the depletion of naive T cells from peripheral blood stem cell grafts is associated with a low incidence of severe aGVHD and cGVHD without an apparent excess risk of relapse or nonrelapse mortality in patients with acute leukemia. 

We are also becoming more aware of the important contribution of other immune cells in the pathogenesis of cGVHD, particularly regulatory T cells and plasmacytoid dendritic cells. For example, in a preplanned analysis of BMT CTN 0201, we showed that the content of donor plasmacytoid dendritic cells in the marrow of graft recipients was inversely correlated with the incidence of cGVHD after transplantation. In this trial, patients with myelodysplastic syndrome or leukemia were randomized to receive either bone marrow or peripheral blood stem cells from unrelated donors. Those who received a marrow graft with more donor plasmacytoid dendritic cells had a lower incidence of cGVHD than similar patients who received a marrow graft with fewer donor plasmacytoid dendritic cells. Such studies are allowing us to dissect the donor cell subsets that are responsible for different functions, such as regulating immune reconstitution and limiting cGVHD, and we are also identifying new molecular targets that could be the subject of therapeutic trials.