In thinking about prophylaxis, we come back to the idea of identifying the optimal donor, also recognizing that, in patients who receive allo-HSCT for a hematologic malignancy, we have the challenges of immune reconstitution and maintaining an adequate GVT response, as well as preventing excessive rates of GVHD, especially severe GVHD and moderate to severe cGVHD. 

There has been a reduction in nonrelapse mortality after allo-HSCT over the years, which probably reflects several influences. There have been advancements in supportive care, and there have also been advancements in donor and recipient genotyping. So, progress has occurred, independent of the modalities that we use for GVHD prevention. 

Given an “ideal matched donor,” how else can we reduce the risk of living with cGVHD without obviating the GVT immune response? Well, in an ideal world, more of the allo-HSCTs performed for malignancy would be bone marrow stem cell grafts. Anasetti and others have shown reduced rates of cGVHD with bone marrow grafts vs peripheral blood stem cell grafts; however, peripheral grafts are often preferred nowadays, for a variety of reasons.

Prophylactic strategies also include the so-called graft-sculpting or graft-shaping protocols. Significant reductions in moderate and severe cGVHD have been seen, but some of these strategies have been associated with trade-offs (eg, risk of cytomegalovirus infection). Infusing regulatory T cells, either in a matched or a haploidentical transplant context, is another approach to try to limit cGVHD.

Of course, post-transplantation cyclophosphamide (PTCy)–based regimens have, over the last 15 years, really revolutionized transplant. PTCy and tacrolimus plus mycophenolate mofetil combination therapy prevents GVHD in peripheral blood transplants with HLA-matched donors. Recent data also suggest that the addition of abatacept can have substantial activity in reducing the development of severe acute GVHD. 

I think that we are eager to see how combinations of some of the aforementioned approaches might advance the goal of reducing cGVHD even further in the future, especially without risking the persistence of the graft-vs-leukemia responses.

I agree that we are dealing with a very complicated situation in which we are trying to achieve multiple goals in patients undergoing allo-HSCT. The first goal is to achieve a successful allo-HSCT while limiting toxicity, as measured by nonrelapse mortality. We are also trying not to compromise immunity, as measured by risk of infection. And, finally, we are trying to reduce GVHD while also maintaining GVT effects. 

Every approach that we pursue must factor in not only the goals of treatment with allo-HSCT but also the possible costs, or consequences, of achieving success with each of these goals (eg, cGVHD, infection, and relapse). These are the things that can only be sorted out through carefully constructed randomized trials. And there are multiple ways to accomplish what we are trying to achieve in allo-HSCT. One way is not necessarily the absolute best, but it comes down to designing and conducting trials that lead to convincing results and moving these findings forward in effective ways. 

This includes performing randomized trials that allow us to make progress by assessing one approach versus another. Those trials have been incredibly impactful. I think that the Blood & Marrow Transplant Clinical Trials Network has done an excellent job with organizing some of these trials, and some drug companies are also doing that. Further, it is imperative that we do this in a large multi-institutional fashion because we have learned from experience that the results from a single institution's, or a few institutions', phase 2 trials are often very difficult to reproduce. 

I agree with my colleagues regarding the continued importance of well-constructed clinical trials. We cannot rest on our laurels. Severe cGVHD is an entity that continues to cause mortality in our patients beyond 2 years. And, despite the success of PTCy, you can expect that it may not perform well in the prevention of cGVHD in the setting of a myeloablative transplant. 

The other element that my colleagues have alluded to is the need to prevent relapse. The current thinking is that the best way to prevent relapses may be to use some form of cellular therapy post transplant. However, in the context of having to have chronic immune suppression in these patients, those cellular therapies, unfortunately, are not likely to work well. I love the concept of graft sculpting. We must find ways to perform transplants that are precise, personalized, and predictable.