Comparison between bone marrow and G-CSF-mobilized peripheral blood allografts undergoing clinical scale CD34+ cell selection.

Allogeneic transplantation of selected CD34+ cells, rather than conventional transplantation of bone marrow (BM) harvest or peripheral blood (PB) leukapheresis products, has the advantage of reducing volume, facilitating storage and decreasing the amount of dimethylsulfoxide (DMSO) and cell lysis products, as well as reducing the number of T-lymphocytes responsible for graft-versus-host disease (GVHD). Using biotinavidin immunoaffinity columns (Ceprate SC system, CellPro; Bothell, WA), CD34+ cells were selected from each of 20 allografts (12 G-CSF-mobilized PB and 8 BM) collected from 14 HLA-identical normal healthy donors for transplantation. After the clinical-scale selection, the median concentration of CD34+ cells was 44.6% (range, 13% to 91%) in BM and 50.4% (range, 15% to 77%) in PB. Whereas 75% of the PB allografts had a CD34+ cell yield of more than 65%, only 37.5% of the BM allografts achieved such a yield, p < 0.01. The number of T-lymphocytes in the selected CD34+ cell allografts was reduced by two to three logs from a median of 4.2 x 10(9) to 7.8 x 10(5) CD3+ cells. The enrichment in CD34+ cells was 240-fold (range, 24- to 382-fold) in PB versus only 34-fold (range, 14- to 108-fold) in BM. Also, the enrichment in clonogenic cells was significantly more in PB (median value of 38.6-fold) than in BM (median value of 19.2-fold) and more in allografts from younger (< 50 years old) rather than older (> or = 50 years old) adult donors. A correlation was found between the percentage of CD34 or CD3+ cells before and after selection (r = 0.58 or r = 0.60, respectively, p < 0.05). Selective enrichment of the colony forming units-granulocyte-macrophage (CFU-GM) was found in all 20 allografts. The progenitor cell recovery after freezing and thawing was similar in BM and PB allografts, with a mean of about 60% for the CFU-GM and BFU-E. In the same six donors, the CD34+ cell yield was significantly more in the PB after mobilization (median 78.5%, range 50% to 90%) than in the BM before mobilization (median 41.5%, range 25% to 87%), p < 0.01. Ten patients with hematologic malignancies have been allotransplanted with 14 of the 20 selected CD34+ cells either combined BM + PB (n = 4) or single (n = 6) grafts. Seven patients did not develop acute GVHD, and only two patients developed > or = grade II GVHD, one of whom developed only grade II GVHD that resolved after brief treatment with corticosteriods. Only one patient showed chronic GVHD (skin and liver). The low incidence and severity of GVHD seen in the present study (only 30%) could be due to the two- to three-log reduction of T-lymphocytes in the selected CD34+ cell allotransplants. All 10 patients had stable hematological recovery, and seven had full donor hematopoiesis. In conclusion, G-CSF-mobilized PB leukapheresis products undergoing selection of CD34+ cells have a greater yield and enrichment of progenitor cells than BM harvests collected from HLA-identical normal healthy donors for allogeneic transplantation. The low incidence and severity of both acute and chronic GVHD (30%) seen in the present study are very encouraging.