Prospective Randomized and Crossover Comparison of Two Apheresis Machines for Peripheral Blood Stem and Progenitor Cell Collection: A Multicenter Study

Peripheral blood stem and progenitor cell (PBSPC) collection imposes apheresis-related risks including thrombocytopenia, hypocalcemia and thromboembolism in donors for PBSPC transplantation. Efficient PBSC collection with fewer apheresis risks and less contamination in the collected products is crucial and we have shown some distinct characteristics among apheresis devices (Transfusion, 2003, 2007 and 2014). A recently introduced apheresis device for PBSPC collections, Spectra-Optia, has not yet established supremacy in safety and efficiency over other established previous apparatuses/programs. To compare the Optia with the widely used automated program for a former apheresis machine, Spectra-Auto (Software version 6.1), we for the first time carried out a prospective randomized multicenter and crossover study for a total of 233 apheresis done from November 2013 to May 2015 in adult (≥18 years) and G-CSF-given patients/donors. The device for the first-day apheresis was randomly assigned, and a total of 152, consisting of 47 for autologous (auto-) patients and 105 for allogeneic (allo-) donors, were studied (1st-day analysis). Among them, the 81 subjects who required the second-day apheresis to collect more cells, the second-day apheresis was performed with the other machine than the one used in the first day, and apheresis collections for 2 consecutive days using both machines were compared by a paired-t test (crossover analysis). In the both devices, apheresis was performed based on the manufacturer's protocols, with 12:1 to 15:1 ratios of whole blood to acid citrate dextrose without heparin. There was no difference between the Optia (n = 74) and the Spectra-Auto (n = 78) in the pre-apheresis parameters including age, sex, circulating blood volume (CBV), and peripheral complete blood counts (CBCs). In the crossover comparison, age, sex and CBV were technically same and CBCs were similar between the two machines (n = 81 in each). The Spectra-Auto processed larger volumes compared with the Optia [P = .02 (1st-day analysis) and .08 (crossover analysis)], although run time was similar with the two devices [P = .57 (1st-day) and .52 (crossover)]. Mild apheresis-related reactions occurred in both machines with similar frequencies of reactions, and volumes of ACD used were not different between them. Volumes of harvested products were greater in the Optia compared with the Spectra-Auto [158 vs. 95 mL, P < .01 (1st-day) and 137 vs. 86 mL, P < .01 (crossover)]. Yields of mononuclear cells [P = .27 (1st-day) and .22 (crossover)] and CD34+ cells [P = .10 (1st-day) and .19 (crossover)] were not different between them. However, regardless of program, pre-apheresis peripheral CD34+ cell counts strongly correlated with the number of CD34+ cells collected, and collection efficiencies of CD34+ cells were significantly higher in the Optia compared with the Spectra-Auto [82.2% vs. 66.3%, P < .01 (1st-day) and 80.5% vs. 63.1%, P < .01 (crossover)]. The products collected using the Optia contained more contaminating red blood cells [64 vs. 36 x 10⁶ cells, P < .01 (1st-day) and 56 vs. 33 x 10⁶ cells, P < .01 (crossover)], compared with the Spectra-Auto, without corresponding anemia. Only in crossover analysis, there was a trend that contaminated platelets were higher with the Optia compared with the Spectra-Auto (154 vs. 127 x 10⁹ cells, P = .06). In conclusion, the Optia is tolerable, and has no significant inferiority in the yield or adverse events in apheresis collection of PBSPCs compared with Spectra-Auto. This trial was registered at http://www.umin.ac.jp/ctr/index.htm as #UMIN000012095.