Summary:Patients with delayed platelet recovery post-PBPC transplant (PBPCT) are a high-risk group for thrombocytopenic bleeding and platelet transfusion dependence. Total CD34 ؉ cell dosage has been proposed as the most important factor influencing the rate of platelet recovery. To achieve the shortest time to platelet engraftment, a minimum leukapheresis target of 10 ؋ 10 6 CD34 ؉ cells/kg was established for 30 patients. Of the 29 evaluable patients, 62% had rapid (group I: time to platelets Ͼ20 ؋ 10 9 /l р10 days and 50 × 10 9 /l р14 days) platelet recoveries while 38% had delayed (group II: 20 ؋ 10 9 /l Ͼ10 days and 50 × 10 9 /l Ͼ14 days) recoveries. Groups I and II were compared for: (1) pretreatment variables; (2) mobilizing capability of CD34 ؉ cells and subsets including megakaryocyte (Mk) progenitors; (3) infused dose of these cells at transplant; (4) changes in endogenous levels of Mpl ligand (or TPO) during mobilization and myeloablative chemotherapy. Group II patients received significantly more platelet transfusions (6 vs 2.1, P ؍ 0.002) post-PBPCT, had a higher proportion of patients with a prior history of BM disease (64% vs 6%, P ؍ 0.001), and showed a reduced ability to mobilize differentiated (CD34 ؉ /38 ؉ , CD34 ؉ /DR ؉ ) and Mk progenitors (CD34 ؉ /42a ؉ , CD34 ؉ /61 ؉ ). Only the number of Mk progenitors reinfused at transplant was significantly different between the groups (group II vs group I: CD34 ؉ /42a ؉ ؍ 1.02 vs 2.56 ؋ 10 6 /kg, P ؍ 0.013; CD34 ؉ /61 ؉ ؍ 1.12 vs 2.70 ؋ 10 6 /kg, P ؍ 0.015). The ability to mobilize Mk progenitors correlated with percentage changes in endogenous levels of TPO from baseline to platelet nadir during mobilization chemotherapy (CD34 ؉ /42a ؉ : r ؍ 0.684, P ؍ 0.007; CD34 ؉ /61 ؉ : r ؍ 0.684, P ؍ 0.007), with group II patients experiencing lower percentage changes. An inverse trend but no correlation was observed between serial TPO levels and platelet counts. TPO levels remained elevated in group II patients throughout a prolonged period of thrombocytopenia (median days to 50 × 10 9 /l = 25 vs 11 for group I), indicating that delayed engraftment was not due to a