Graft Composition and Outcomes in Unrelated Donor Transplantation.

Abstract

We examined the effect of graft composition parameters including total white blood cell (WBC), myeloid cell dose (CD34+, CD34+/CD38+) lymphoid cell dose (CD3+, CD4+, CD8+), and activated lymphoid cells expressing activation antigens (CD25, CD69, HLA-DR) in 123 bone marrow (BM) and 198 peripheral blood progenitor cell (PBPC) transplants from unrelated donors (URD) between July 2003-March 2005. Most recipients (85% BM, 94% PBPC) had a hematologic malignancy. Samples were shipped from 33 participating centers for analysis at a central laboratory. Most BM grafts (67%) were processed: 24% T-cell depleted, 26% red blood cell (RBC) depleted, 12% plasma depleted and 5% mononuclear cell concentrated. Fewer PBPC grafts (29%) were processed: 19% plasma removal, 9% CD34+ selection and 1% RBC depletion. All samples had acceptable viability (≥70%). CD34 recovery was high for BM grafts after plasma or RBC depletion (84% and 62%, respectively) but low after T-cell depletion (26%). CD34 recovery was high for plasma depleted PBPC (83%). Plasma or RBC depletion had minimal effect on lymphoid cell content of either BM or PBPC, while T- cell depletion or CD34 selection resulted in 3–4 log10 depletion of CD3+ and HLA-DR+ lymphoid cells. We did not observe differences in transplant outcomes among those receiving processed vs. non-processed BM or PBPC grafts. The overall or subset cellular composition of BM grafts was not associated with the likelihood of hematopoietic recovery, acute and chronic graft-versus-host disease (GVHD) or 100-day survival. After adjusting for other clinically significant factors, risks of overall mortality were significantly lower when BM grafts contained ≥2.36 × 106/kg CD25 cells (relative risk [RR] 0.55, p=0.048) and ≥8.20 × 104/kg CD 34+/38+ (RR 0.58, p=0.043); no other subsets were associated with overall survival. With PBPC grafts, day-28 neutrophil recovery was higher with grafts containing ≥3.30 × 106/kg CD69 cells (97% vs. 89%, p=0.036). Platelet recovery by day-90 was higher with PBPC grafts containing: ≥5.44 × 106/kg CD34+ cells (87% vs. 74%, p=0.027), ≥2.20 × 108/kg CD3+ cells (87% vs. 74%, p=0.028) and ≥8.15 × 107/kg CD8+ cells (88% vs. 73%, p=0.011). The likelihood of acute GVHD was not associated with the composition of PBPC grafts, but chronic GVHD was higher with PBPC grafts containing ≥2.20 × 108/kg CD3+ cells (RR 1.66, p=0.029). Early mortality rates were lower with PBPC grafts containing ≥8.15 × 107/kg CD 8+ cells (RR 0.47, p=0.025) and overall mortality rates were lower with PBPC grafts containing ≥5.44 × 106/kg CD 34+ cells (RR 0.58, p=0.003) and ≥8.15 × 107/kg CD 8+ cells (RR 0.67, p=0.029). These data suggest that adequate CD25 and CD34+/38+ are necessary for optimal survival after URD BM and adequate CD34+ and CD8+ after PBPC transplantation. Further analysis of the ideal graft composition associated with transplant outcome will be needed to plan graft engineering manipulations to improve patients' survival and could be particularly relevant for PBPC transplants which are associated with higher chronic GVHD.