Granulocyte colony-stimulating factor (G-CSF)–mobilized peripheral blood mononuclear cells (G-PBMCs) are increasingly used as a stem cell source for allogeneic transplantation. The in vivo administration of G-CSF to the donor results in an increase of circulating CD34+ hematopoieitic progenitor cells, which are responsible for engraftment after myeloablative therapy.
In a recent report by Zaucha et al,1 an association between the CD34 cell dose in G-PBMCs collected from human leukocyte antigen (HLA)–identical siblings and the development of extensive chronic graft-versus-host disease (cGvHD) is shown. The number of CD3 T lymphocytes or CD14 monocytes in the graft were not significantly associated with extensve cGvHD. Therefore, the authors concluded that increasing the CD34 cell number in G-PBMC products may be counterproductive and suggest limiting the number of CD34 cells in the unmodified G-PBMC grafts.
Rather than using unmodified G-PBMC grafts, we have transplantated highly purified G-CSF–mobilized peripheral CD34 stem cells from HLA-matched unrelated2 and HLA-mismatched haploidentical donors3 in 76 pediatric patients with malignant and nonmalignant diseases. The CD34 cells were purified to a median purity of at least 98% and transplanted without any posttransplantation GvHD prophylaxis. The median number of transplanted CD34 stem cells was 12.1 × 106/kg recipient body weight with a wide range of 1 to 54.6 × 106/kg. The median number of cotransplanted CD3 T lymphocytes was 7.6 × 103/kg (range, 0.5-130 × 103/kg). The median follow-up time for our patients is 2.9 years (range, 0.9 years to 5.7 years). Only patients who survived at least 100 days after transplantation were analyzed.
In contrast to the study by Zaucha et al, we have not seen any correlation between the number of transplanted CD34 stem cells and the development of extensive chronic GvHD. A transient, limited cGvHD was observed in 6 patients. Although 25 of the 76 patients received megadoses of CD34+ cells (at least 20 × 106/kg), no extensive cGvHD was seen in any of the 76 patients. We observed, however, a positive correlation between the number of transplanted purified CD34 stem cells and the speed of T-cell reconstitution.3
Comparing the Zaucha et al study with our results, we conclude that the development of extensive cGvHD is not directly associated with the CD34 stem cell dose but, rather, with the graft composition after G-CSF mobilization. The number of transplanted CD34 stem cells is directly correlated with the number of mobilized CD34 stem cells and therefore reflects the donor's responsiveness to G-CSF. It is not known why some donors are poor or good CD34 mobilizers, and there is a high degree of variability among donors. It is known, however, that G-CSF has an effect on the induction of a T helper 2 (Th2) cell response4 and mobilizes Th2-inducing DC2 dendritic cells.5 Since cGvHD is associated with excessive Th2 activity, one possibility is that donors who are good mobilizers develop a more pronounced Th2 or DC2 response. This response may be reflected not in the absolute numbers of CD3 T lymphocytes or CD14 monocytes but, rather, in relative differences in their subset composition. Patients who have a good mobilizer as a stem cell donor will receive transplants with higher CD34 stem cell doses and, therefore, also with higher numbers of Th2 or DC2 subsets. This could be an explanation for the high incidence of extensive cGvHD in the Zaucha et al study and the complete lack of extensive cGvHD in our studies.
The high incidence of extensive cGvHD after transplantation with unmanipulated G-PBMCs can be reduced by the reduction of T cells in the graft.6 Due to improved technologies for stem cell purification,7 it is possible to exactly define the graft composition. From our experience with the transplantation of purified CD34+ stem cells, we suggest increasing rather than limiting the CD34 stem cell dose and adjusting the number of cotransplanted T lymphocytes or other subsets to defined numbers. This could result in an improved immunoreconstitution without increasing the risk of extensive chronic GvHD.
Association of CD34 cell number with chronic graft-versus-host disease in adults
We reported a correlation between high numbers of CD34 cells and incidence of chronic graft-versus-host disease (GVHD) in adult patients (median age, 43 years) given unmodified peripheral blood stem cell grafts from HLA-identical siblings.1-1Handgretinger et al reported that, by removing T cells from the PBSC inoculum, chronic GVHD can be prevented in children (median age, 7 years) regardless of the number of CD34 cells transplanted from HLA-haploidentical or matched unrelated donors.1-2These statements are true, true, and unrelated. We know from studies published almost 20 years ago in patients given unmodified grafts that chronic GVHD is rare in children, compared with adults over 30 years old.1-3 In the absence of donor T cells, we would not expect to find any correlation between the number of CD34 cells in the graft and the risk of chronic GVHD in children, even those receiving HLA-mismatched grafts.
We agree that CD34 cells per se are not sufficient to cause chronic GVHD, but our results suggest that they or their progeny interact directly or indirectly with donor T-cell populations to increase the risk of the disease, at least in adults. Validation through experimental testing will be needed to substantiate the interesting suggestion that high-yield mobilization of CD34 cells might be associated with increased Th2 activity, thereby explaining the correlation between CD34 cell dose and risk of chronic GVHD.
Having said this, the following point should also be considered. The ultimate success of hematopoietic stem cell transplantation is measured by patient survival. All of the 181 adult patients in our report had malignant diseases, and 77% of them had advanced disease. Their 3-year progression-free survival was 44%. The corresponding figures for 2-year survival for the 31 pediatric patients with malignant diseases in Handgretinger et al's report1-2 were 28% overall and 14% for patients with advanced disease. The 14 patients in Lang et al1-4 are difficult to evaluate because the median follow-up time was only 11 months.
Many reports have shown increased risks of graft failure and disease recurrence after T-cell depletion; nevertheless, the question of whether to deplete T cells or not remains open, particularly in the HLA-mismatched setting. To us, however, induction of graft-host tolerance by transient postgrafting immunosuppression is currently more attractive than wholesale removal of T cells from the grafts, especially in adults who have impaired thymic function. Given that, restricting the number of transplanted CD34 cells seems an expedient way to reduce the risk of severe chronic GVHD.
