Optimal numbers of CD34+ cells to be reinfused in patients undergoing peripheral blood progenitor cell (PBPC) transplantation after high-dose chemotherapy are still unknown. Hematologic reconstitution of 168 transplantations performed in patients with lymphoproliferative diseases was analyzed according to the number of CD34+ cells reinfused. The number of days from PBPC reinfusion until neutrophil recovery (>1.0 × 109/L) and unsustained platelet recovery (>50 × 109/L) were analyzed in three groups defined by the number of CD34+ cells reinfused: a low group with less than or equal to 2.5 × 106 CD34+ cells/kg, a high group with greater than 15 × 106 CD34+cells/kg, and an intermediate group to which the former two groups were compared. The 22 low-group patients had a significantly delayed neutrophil (P < .0001) and platelet recovery (P < .0001). The 41 high-group patients experienced significantly shorter engraftment compared with the intermediate group with a median of 11 (range, 8 to 16) versus 12 (range, 7 to 17) days for neutrophil recovery (P = .003), and a median of 11 (range, 7 to 24) versus 14 (range, 8 to 180+) days for platelet recovery (P< .0001). These patients required significantly less platelet transfusions (P = .002). In a multivariate analysis, the amount of CD34+ cells reinfused was the only variable showing significance for neutrophil and platelet recovery. High-group patients had a shorter hospital stay (P = .01) and tended to need fewer days of antibotic administration (P = .12). In conclusion, these results suggest that reinfusion of greater than 15 × 106 CD34+ cells/kg after high-dose chemotherapy for lymphoproliferative diseases further shortens hematopoietic reconstitution, reduces platelet requirements, and may improve patients' quality of life.

BACKGROUND Chronic red blood cell exchanges (RBCXs) are frequently used to prevent complications in patients with sickle cell anemia, but the scarcity of matched red blood cell packs (RBCPs) is a serious concern. The main goal of this study was to compare the number of RBCPs used during RBCXs between the Spectra Optia (SO) device (with the automatic depletion step) and the former Cobe Spectra (CSP) device. STUDY DESIGN AND METHODS The performances and safety of 300 SO sessions using the automatic depletion step (SO/DE) in 50 patients with sickle cell anemia under a chronic transfusion program over a 1‐year period were prospectively analyzed. The numbers of RBCPs saved using this protocol compared to the SO device without depletion and to the CSP device were determined. RESULTS The SO/DE protocol appeared to be safe, as only 5% and 17% of the sessions were characterized by a significant decrease in blood pressure and increase in heart rate (grade 2 adverse events), respectively. Postapheresis hematocrit and fraction of cells remaining reached expected values. The SO/DE protocol required 16% fewer RBCPs compared to SO without depletion, allowing a mean saving of 12 RBCPs per patient and per year and 13% fewer compared to CSP device. Interestingly, the saving was more important for patients with high total blood volume and/or high preapheresis hematocrit. CONCLUSION The SO/DE protocol is an efficient, safe and cost‐effective procedure for patients with sickle cell anemia under a chronic transfusion program.

Summary:The aim of the study was to evaluate the outcome of unrelated bone marrow donor (UBMD) searches initiated for 174 children between 1986 and 1997. Seven patients were registered twice so that a total of 181 UBMD searches took place. At the time of registration, patients suffered from hematological malignancies (n = 121), non-malignant hemopathies (n = 26) and inborn errors (n = 34). Forty-five of the patients (26%) were given transplants from unrelated donors of whom 26 (58%) were HLA-mismatched transplants. Our strategy accepted HLA mismatches at the time of donor selection, using Thymoglobuline as part of the conditioning regimen. Of the 45 patients given unrelated donor transplants, overall survival was 60% at 3 years and concerned 27 patients of whom 14 were from HLAmismatched donors. Disease-free survival for hematological malignancies was 65% in HLA-matched transplants and 50% in HLA-mismatched transplants. For some patients (16%) urgency led us to use alternative options: non-identical related donor (n = 14), autograft (n = 10), related cord blood transplant (n = 4). For others, UBMD searches were stopped because of favorable evolution (n = 29), death (n = 24), disease progression (n = 22) or other reasons (n = 21). By the end of the follow-up period, 88 patients had died (50%), 75 (43%) are currently alive with or without being transplanted of whom eight are still having active searches and 11 are no longer contactable. In conclusion, in severe disease in children, an immediate transplant from a partially matched donor might be preferable to a prolonged search for a full match. Consequently, this strategy increases the number of patients for whom a suitable donor can be found. We have chosen this option in order not to delay BMT; in so doing we have obtained encouraging results which include high overall survival, low incidence of acute GVHD grade III-IV and low percentage of relapse even in mismatched pairs. Bone Marrow Transplantation (2000) 26, 31-43.

Blood transfusions can modify host immunity and clinical outcomes in hematological malignancies. One thousand sixty-seven patients with acute myeloid leukemia (AML) were studied for their transfusion dependency at initial presentation and transfusion frequency during induction chemotherapy. Three hundred five patients (29 %) showed initial dependence to red blood cell (RBC) transfusion and 109 (10 %) to platelet transfusion. Transfusion dependency at presentation was associated with a poorer prognosis. Both initial RBC and platelet transfusion needs were associated with lower response rates (P = 0.04 and P = 0.03). Median overall survival (OS) was 10.8 months for patients with RBC need vs 18.8 months for the other patients (P = 0.02) and 6.8 months for patients with platelet transfusion need vs 13.6 months for the others (P = 0.01). Similarly, transfusion intensity during induction therapy influenced negatively treatment outcome. Median transfusion burden per week was 2.5 (range 0-25.7) RBC units and 1.6 (range 0-15.7) platelet concentrates (PCs). Both high RBC and PC transfusion intensities were associated with lower response rates (P = 0.003 and P < 0.0001). Median OS was 9.08 months for patients with RBC transfusions >3/week vs 18.29 months for those with RBC transfusions ≤3/week (P = 0.0003) and 10.75 months for patients with PC transfusions >2/week vs 19.96 months for those with PC ≤2/week (P = 0.0003). RBC and platelet transfusion intensities during induction therapy remained of prognostic value in multivariate analysis. Transfusion need at presentation and the frequency of transfusions during induction chemotherapy appear as strong prognostic factors.