Abstract
Background The adequacy of HPC collection for BMT is typically assessed by the number of CD34+ cells. However, during a series of leukapheresis procedures (LP) the CD34 value on the final HPC product may not be available for testing until late evening, sometimes resulting in additional, retrospectively unnecessary, LP in order to ensure an adequate HPC collection (>5×106 CD34/kg). We hypothesized that an estimate of the CD34 content of HPC products prior to 16:00 h on the day of LP would permit improved HPC collection planning. We therefore assessed the effectiveness of predicting the total amount of CD34 cells that would be collected in a given LP by either (a) the concentration of CD34 cells/μL in peripheral blood prior to LP (pre-CD34) or (b) the predicted total amount of CD34 cells to be collected based on sampling the LP product at the mid-point of each LP. We also compared the number of LP per patient and total HPC collected for the study group with data from the previous calendar year.
Methods Allogeneic and autologous BMT donors who completed a 20-L HPC collection between September 2002 and February 2003 were eligible. CD34 cells were measured on blood drawn prior to LP and from the HPC product at the mid-point (10 L) of LP. The CD34 content of the final LP was predicted by doubling the value of total CD34 cells at the mid-run (MRp-CD34). The MRp-CD34/kg and the cumulative CD34/kg collected were made available before 16:00 h and used to determine the need for additional LP. The true CD34 content of each HPC collection was also measured from the final product the next day (CD34-FP).
Results A 20-L LP was completed and data were available from 31 patients and nine allogeneic donors who underwent a total of 85 LP for diagnoses, including 11 myeloma, 10 lymphoma, seven HD, three acute leukemia and five others. The mean (range) and correlation (R2) vs. the CD34-FP were, for pre-CD34, 54 CD34/μL (0.3–232), R2=0.66 (P<0.01), and for MRp-CD34, 3.2×106 CD34/kg (0.04–22.48), R2=0.90 (P<0.01). The mean number of CD34/kg collected per LP in the patients/donors was 3.4×106 CD34/kg (0.05–18.94). The median number of CD34 cells employed for transplant in the study group vs. controls (5.7 vs. 5.6×106/kg) and the time to engraftment of neutrophils (12 vs. 11 days) and platelets (12 vs. 12 days) was similar to historical controls. However, the study group had a significantly lower median number of LP (three vs. two; P<0.02) to obtain the required collection of 5×106 CD34 cells/kg.
Discussion Both the pre-CD34 and the MRp-CD34 were significantly correlated with CD34-FP. However, the CD34-FP was more reliably predicted by MRp-CD34. Early availability of mid-run CD34 values was associated with a significant reduction in the number of LP required to collect 5×106 CD34 cells/kg, without reduction in the number of CD34 cells for transplant or prolongation of days to neutrophil or platelet engraftment.