For more than 30 years, physicians have transplanted allogeneic hematopoietic cells (HCs) harvested from the bone marrow cavity without any specific HC stimulation. During the past decade, blood HCs collected by apheresis after stimulation with granulocyte-colony stimulating factor (G-CSF) of the donor have been explored as an alternative source and it was demonstrated in a phase III trial that G-CSF–mobilized blood HCs are the preferred inoculum, compared with marrow HCs that had been harvested from “steady state” donors, at least for patients in advanced stages of their hematologic malignancy (Bensinger et al. N Engl J Med. 2001;344:175-181).

Last year, intriguing phase II data were reported from 3 independent groups (Couban et al, Biol Blood Marrow Transpl. 2000;6:422-427; Isola et al, Biol Blood Marrow Transpl. 2000;6:428-433; Serody et al, Biol Blood Marrow Transpl. 2000;6:434-440) indicating that G-CSF–treated donors provide marrow HCs that lead to rapid engraftment and a relatively low rate of graft-versus-host disease (GVHD). Here, Morton and colleagues (page 3186) report phase III trial results showing that G-CSF–mobilized HCs derived from marrow or blood have virtually equivalent engraftment kinetics for granulocytes and platelets but that marrow HCs are associated with significantly less steroid refractory acute GVHD and a lower incidence of chronic GVHD. One would like to see this important observation confirmed in a uniform patient population conditioned with a single preparatory regimen.

Existing preclinical data could explain the interesting observation described by Morton and colleagues. These studies in murine models indicate that bone marrow T cells induce less GVHD than equivalent amounts of blood T cells (189:1073-1081; Lan et al, Blood. 2001;97:3458-3465; Zeng et al, Blood. In press). Likewise, human T cells from marrow or blood not only differ quantitatively but also may have vastly different qualitative features and functions. In the days of microarray assays, such differences should soon become evident.

Progress toward optimizing the allogeneic graft continues to be made. Ultimately, ex vivo manipulation of allogeneic HCs, regardless of the source, will provide the patient with the best result: a graft that leads to rapid and durable engraftment with preserved graft-versus-tumor effects and without GVHD.