In the September 8, 2016, issue of The New England Journal of Medicine, Dr. Filippo Milano and colleagues from Fred Hutchinson Cancer Research Center reported that for patients with acute leukemia and myelodysplastic syndrome who had evidence of minimal residual disease (MRD) prior to transplantation, the risk of relapse was significantly reduced in unrelated cord blood recipients relative to matched or mismatched volunteer unrelated donors.1 These results suggest for the first time that cord blood transplants may be the preferred alternative donor source for patients with evidence of MRD going into transplantation.
This report carries with it all the caveats in interpretation inherent in retrospective data. While the authors carefully adjusted for clinical factors that could affect outcomes to limit bias, there may still be confounding factors that are important to discuss. That said, while perhaps not practice changing, the data are provocative and raise a number of interesting questions: 1) since the majority of patients received two umbilical cord blood units, what if any influence did this have on the risk of relapse? Double cord blood transplantation has been associated with decreased relapse and an enhanced graft-versus-leukemia effect in some,2 though not all, studies. 2) Although the authors do not report how many of the seropositive cordblood transplant patients actually had cytomegalovirus (CMV) reactivation, roughly 60 percent of umbilical cord blood transplant patients would be expected to have CMV reactivation despite intensive strategies to prevent CMV disease.3 CMV reactivation has been independently associated with decreased relapse rates in acute myeloid leukemia,4 so this possible confounding variable cannot be ignored. Again, not all studies are in agreement on this matter, but this potential effect of cord blood transplantation could be explored further. 3) Should we be assessing all of our patients for MRD prior to transplant using a sensitive flow based method as used in this paper or some similarly sensitive method? Importantly, Dr. Milano and colleagues raise this question and suggest that we may need to use better assays for depth of response prior to transplantation to determine the best intervention needed to mitigate the risk of relapse.
Cord blood transplantation has been associated with less chronic graftversus-host disease5 and could enable nearly all patients to have an available donor. On the other hand, haploidentical transplants are being used increasingly, even in older patients, and may be simpler to perform and less expensive relative to cord blood. Which is the preferred source for patients without a matched related or unrelated donor? The Blood and Marrow Transplant Clinical Trials Network is currently addressing this question in its BMT CTN 1101 study (clinicaltrials.gov, NCT01597778) and we wholeheartedly support its importance, perhaps even more so given the recent report from Fred Hutchinson Cancer Research Center. Until those results are available, the available data seem to support the use of cord blood transplant as a reasonable first option in MRD-positive patients with acute leukemia and myelodysplastic syndrome who do not have a suitable matched sibling donor.
-Jennifer N. Saultz, DO, Hematology/Oncology Fellow, The Ohio State University, Columbus, OH-Steven Devine, MD, Professor of Internal Medicine, Program Director, Blood and Marrow Transplant Program, The Ohio State University, Columbus, OH
Dr. Steven Devine and Dr. Jennifer Saultz provide commentary on cord blood transplantation in patients with MRD (in response to a NEJM article by Dr. Filippo Milano and colleagues).