Transplantation of 2 cord blood units to a single patient has been enthusiastically adopted as a strategy to increase access to unrelated donor cord blood transplantation (UCBT) for larger pediatric and adult patients. With this strategy, 1 cord blood unit ultimately dominates and confers durable engraftment, but prospective identification of the engrafting unit has not been possible.
The article by Avery and colleagues in this issue of the Blood sheds light on this biologic mystery in double umbilical CBT by demonstrating, in a series of 84 patients, that indicators of cord blood unit potency, for example, total nucleated cell count (TNC), colony-forming unit (CFU), CD3, and viable CD34 cell content, predict the dominating unit.1 Surprising to some, human leukocyte antigen (HLA) matching does not appear to play a role in unit dominance.
Since the demonstration that unrelated donor, banked umbilical cord blood provides a source of cells for hematopoietic reconstitution after myeloablative therapy, more than 500 000 publicly donated units have been banked and more than 25 000 UCBTs have been performed worldwide. The fact that cord blood could be transplanted without full HLA matching also emerged as a major advantage providing access to transplantation to patients unable to find a fully matched adult donor. Early on, the importance of transplanting an adequate cell dose was established with generally accepted thresholds of 25-30 million nucleated cells per kg required for engraftment.2–3 Given the limited volumes of blood available from typical term placentas, less than 10% of collected units will contain sufficient cell doses for patients weighing > 70 kg. In 2005, a group at the University of Minnesota reported enhancement of engraftment in adults with hematologic malignancies receiving 2, sequentially transplanted, partially HLA-matched unrelated donor cord blood units.4 Interestingly, even though engraftment rates were higher than those seen after transplantation of a single cord blood unit, only 1 of the 2 units conferred durable engraftment. Furthermore, the dominating unit was generally declared within the first month after transplantation. However, despite rapid adoption of this strategy by adult transplant physicians, the mechanisms underlying the success of this approach are unknown. Furthermore, the ability to predict the engrafting unit was not previously possible. In early series, obvious candidates such as total nucleated cell dose, HLA match, and CD3 dose did not predict the engrafting unit. Some investigators questioned whether the infusion of 2 cords was really necessary and hypothesized that the engraftment rates were higher because using 2 units gave the patient a better chance of receiving at least 1 unit with high potency. Others thought that 1 of the 2 units facilitated engraftment of the dominating unit. Others thought that concomitant changes in the preparative regimen (substitution of fludarabine for antithymocyte globulin) and/or graft-versus-host disease (GVHD) prophylaxis (substitution of mycophenolic acid for methylprednisonlone) were the reason(s) for success. The latter questions remain unanswered but are under study through the Blood and Marrow Transplant Clinical Trials Network 0501 study of double versus single cord blood transplantation in children with hematologic malignancies.
While traditional selection of cord blood units uses TNC and HLA matching, recent studies in single cord blood transplantation have shown that dosing of CD34 cells and CFUs are better predictors of unit potency as measured by engraftment potential.5–7 The current study supports this concept; in addition, units with higher viable CD34 cells and higher CFUs tended to be associated with higher sustained engraftment and more rapid neutrophil recovery.
The conclusions of the current article are intriguing, but do not prove that 2 cord blood units are better than 1 or, in fact, necessary for engraftment in a larger patient. One must question whether the observation that the more potent unit engrafts proves that there is a benefit to infusing 2 cords for 1 transplant. This is important because some series have shown increased GVHD with double cord transplantation and also because the costs to procure 2 cord blood units per patient are expensive and even prohibitive in some health care models. This question could be addressed in a clinical trial that assessed cord blood potency at the time of unit selection from an attached segment and then randomized patients to receive 1 or 2 potent units. If it is shown that, despite potency, 2 are better than 1, strategic decisions about which unit's engraftment might be favored (eg, the better HLA match) may be possible ranking potency of the preferred unit above the helper unit. We still have a lot to learn about the biology of CBT. These authors have taken us 1 more step along the journey, but we still have a lot more work to do before we master the field and optimize clinical practice.
Conflict-of-interest disclosure: The author declares no competing financial interests. ■