Chronic GVHD following hematopoietic cell transplantation is associated with reduced relapse incidence in patients with leukemias. This impact has been investigated in myelodysplastic syndrome, showing a beneficial impact of limited chronic GVHD on transplant outcomes in a cohort of more than 3,000 patients.
See related article by Konuma et al., p. 6483
In this issue of Clinical Cancer Research, Konuma and colleagues (1) retrospectively look at data from over 3,000 patients with low-risk and high-risk myelodysplastic syndrome (MDS) to evaluate the impact of acute and chronic GVHD on hematopoietic stem cell transplant outcomes.
MDS is a clonal hematopoietic stem cell neoplasm with a broad spectrum of presentations united by peripheral blood cytopenias and ineffective production of normal mature blood cells. Currently, allogeneic hematopoietic stem cell transplantation (HSCT), from a related or unrelated donor, is the only potentially curative therapy for MDS. However, HSCT comes not only with the risk of posttransplant relapse, but also the significant risk of inducing GVHD, in which donor-derived immune cells attack the patient (alloreactivity), even with the best HLA matching (2).
Studies in the 1950s and 1960s by Barnes and Loutit used murine models to characterize effects of GVHD on recipient animals. They showed that this alloreactivity of donor cells against the recipient, while triggering GVHD, can exert antileukemic or antitumor effects (3). This desirable antileukemic effect of GVHD was subsequently demonstrated in patients with acute leukemia in two classic papers in the field published by Weiden and colleagues around 1980 (4, 5). These patients were conditioned with total body irradiation and chemotherapy, then underwent transplantation from related, generally HLA matched, sibling donors. Weiden and colleagues showed that patients who developed GVHD had a lower incidence of disease relapse. While GVHD was often associated with higher mortality from nonrelapse causes, it was of note that, in particular, the development of chronic GVHD resulted in superior survival. This observation has been confirmed in various subsequent studies.
Konuma and colleagues focus on these very same questions but restrict their analyses to patients who have undergone HSCT for MDS, rather than acute leukemias. They provide data from one of the largest cohorts of patients with MDS undergoing transplant in which GVHD has been studied to date. In their study, the effect of GVHD is relatively clear: the presence of acute GVHD, particularly in its more severe stages III and IV, had little to no beneficial effects, with substantially reduced overall survival for patients with grade III–IV acute GVHD in both low- and high-risk MDS. This was associated with a significant increase in nonrelapse mortality, which is often related to infections that complicate GVHD. On the other hand, chronic GVHD, specifically in its limited form (generally defined as involving only one organ of the body), had a clear effect on relapse prevention and benefit to overall mortality in patients with high-risk MDS (Fig. 1).
MDS encompasses a broader spectrum of disease presentations than acute leukemia; some patients have very low-risk disease and have a relatively protracted course, sometimes extending over a decade, whereas patients with high-risk disease may succumb to their disease within a year. Calculation of risks, benefits, and timing of HSCT is particularly challenging in patients with MDS. It is primarily the higher risk patients for whom HSCT is of particular relevance, while treatment decisions are more difficult in patients with low-risk disease. One area of consideration is whether to use low- or high-intensity conditioning regimens in patients with low-risk disease; high-intensity regimens tend to produce rapid engraftment of donor cells but also increase the frequency and severity of GVHD (6). Interestingly, Konuma and colleagues find that in contrast to patients with high-risk MDS, there was no significant benefit of any form of GVHD in preventing disease relapse in patients with low-risk MDS conditioned with high-intensity regimens. Perhaps, it is possible that cells in lower risk MDS are less immunogenic, that is, expose fewer or less avid targets for donor cells to initiate a graft-versus-tumor effect. It is, however, also likely that results are related to the fact that fewer relapses occur in patients with low-risk disease, adding weight to the significance of nonrelapse mortality for these patients.
The authors present many subanalyses of the data, which is reflective of the clinical heterogeneity of MDS. Some questions probably cannot be fully addressed by the available data as patient numbers for certain subgroup analyses become limiting. However, there are several interesting subpoints that raise intriguing issues for the field. For example, some studies have shown that the relapse rate is independent of donor type in patients with MDS undergoing allogenic HSCT. Interestingly, the authors identified some associated benefit (reduced relapse and improved overall mortality) associated with the development of grades I–II acute GVHD, as well as borderline benefit of limited chronic GVHD, in patients who were transplanted with umbilical cord blood (UCB). These findings suggest that graft-versus-tumor effects associated with GVHD in patients with MDS may differ according to donor type. One might conceive that this benefit in UCB recipients may be related to greater adaptability of “immature” cord blood cells to the recipient microenvironment, and/or variable cell-type immune reconstitution, pointing to avenues for further research exploration.
By necessity, based upon dates of patient inclusion in the study, Konuma and colleagues utilized long established criteria as determined by the French-American-British (FAB) and the World Health Organization (WHO) for MDS disease classification. While cytogenetic data were included, mutational data were not available. In recent years, the mutational spectrum in MDS has become more associated with clinical phenotype and prognosis, and selection of a patient with MDS for HSCT has begun to incorporate the mutational landscape of the patient's disease. Thus, one will have to be cautious in extrapolating the current data to ongoing and future trials, which have begun to incorporate molecular information. It will be very informative to see similar studies, perhaps work from this group analyzing their more recent patients, that will incorporate novel classifications such as the 2012 Revised International Prognostic Scoring System for MDS as well as mutational data.
Overall, there is a clear basic message, consistent with data in other disease groups and now backed by one of the largest studies in patients with MDS: mild GVHD, in particular, limited chronic GVHD, significantly reduces the incidence of relapse in patients with high-risk MDS as determined by FAB and WHO classifications.
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.