CD34+CD38-HLA-DR- correlate more closely than CD34+CD38- with engraftment post-PBSC transplantation.

Abstract

Henon et al (1998) described the role of CD34+CD38− cells in early and late trilineage engraftment after PBSC transplantation, and identified better correlation with haemopoietic recovery of CD34+CD38− than total CD34+ cells. We report similar findings in a group of patients undergoing peripheral blood stem cell (PBSC) transplantation in Belfast City Hospital, Belfast, Northern Ireland, and in addition include analysis of the relationship with post-transplant recovery of the CD34+CD38−HLA-DR− population. 29 patients with haematological malignancies were primed for PBSC harvesting with cyclophosphamide 1.5 g/m2 plus G-CSF, with leukapheresis on days 8–10 post-conditioning. CD34+ subpopulations were estimated in post-thaw cryopreserved pilot vials. A previous study in 10 patients (30 leukapheresis products) failed to demonstrate any significant difference in percentage of CD34+ cells expressing CD38 or HLA-DR between pre-freeze and post-thaw samples. Viability assessed by trypan blue exclusion was consistently high (median 94.5%) but numbers of mononuclear cells were reduced (median ratio post/pre-cryopreservation 0.49). Subpopulation analysis was carried out by simultaneous staining of 1 × 106 cells with anti-CD34 phycoerythrin (PE), CD38 fluorescein isothiocyanate (FITC), and HLA-DR peridinin chlorophyll protein (PerCP) or appropriate negative controls, followed by lysis of erythrocytes by incubation for 10 min with FACS lysing solution (1:10 dilution in distilled water), washing and resuspension in CellFIX (1:10 dilution in distilled water). Analysis was carried out using a FACScan flow cytometer and Lysis II software (all reagents and equipment; Becton Dickinson). Subpopulations were estimated as a percentage of CD34+ cells and total numbers calculated using the cell percentage, mononuclear count and product volume. 14 patients progressed to high-dose chemotherapy and PBSC transplantation; G-CSF was used post-transplantation in all patients. Neutrophil recovery (> 1.0 × 109/l) occurred in a median of 11 d and platelet recovery (> 50 × 109/l) in a median of 18 d. Correlation was assessed between cell dose and haemopoietic recovery; correlation coefficients for comparison between neutrophil recovery > 1.0 × 109/l were −0.497, P = 0.07 (CD34+), −0.683, P = 0.007 (CD34+CD38−) and −0.719, P = 0.004 (CD34+CD38−HLA-DR−), indicating an increasing correlation with myeloid recovery of CD34+, CD34+ CD38− and CD34+CD38−HLA-DR− populations respectively. Significant correlation was not identified between any population and platelet recovery. These findings are consistent with those of Henon et al (1998), including the suggested threshold dose of 0.05 × 106/kg CD34+CD38− cells for rapid engraftment: 4/4 patients transplanted with at least this dose achieved rapid neutrophil engraftment (< 11 d), compared to only 4/10 patients transplanted with less than this. Several authors have subdivided the CD34+CD38− population on the basis of HLA-DR expression (Huang & Terstappen, 1994; Rusten et al, 1994). Henon et al (1998) postulate that HLA-DR expression might identify committed and uncommitted progenitors, with the less differentiated CD34+CD38−HLA-DR− cells contributing to late engraftment. Our study does not support this hypothesis and demonstrated a significant contribution of CD34+CD38−HLA-DR− cells to early engraftment and better correlation than with the CD34+ or CD34+CD38− populations. We also found that yield of CD34+CD38− and CD34+CD38−HLA-DR− cells was negatively related to prior chemotherapy quantified using the scoring system previously described by our group (Drake et al, 1997). Geometric mean CD34+CD38− and CD34+CD38−HLA-DR− yields in 29 patients were significantly lower in patients with a score > 25th percentile (15.75, roughly equivalent to 3.5 cycles of CHOP) using the Student's t-test (P = 0.047 and 0.033 respectively). In contrast, yields of CD34+CD38−HLA-DR+-cells were not significantly reduced (P = 0.454). These findings suggest that the CD34+CD38−HLA-DR− population contains committed cells involved in early myeloid engraftment and that CD34+ subpopulations may be affected in a disproportionate manner by previous chemotherapy. Our findings emphasize the importance of early identification of patients likely to require PBSC transplantation as the deleterious effect of prior chemotherapy is detectable after relatively minor drug exposure.