Use of Laboratory Data to Enhance Donor Experience and Implement a New Collection Device for HPC, Apheresis

Background: About 80% of unrelated, allogeneic donations for bone marrow transplant are collected as HPC, Apheresis from peripheral blood using the National Marrow Donor Program mobilization protocol,  BB-IND #6821-0297 for filgrastim (Amgen, Thousand Oaks, CA). The aging COBE Spectra (Terumo, Inc., Golden, CO) was widely used for decades. This collection method pre-dated most predictive CD34⁺ counting, and collection orders were written requesting a run of 24 liters of whole blood. Over-collections of CD34⁺cells are seen with this method (mean=225.17%, n=247), often as high as 3X to 6X.  Terumo introduced a replacement apheresis collection device, the Optia, with updated electronic controls and a different method of mononuclear cells extraction. The COBE will no longer be supported in the near future, forcing collection centers to make a device change.

Problem Statement: The collection efficiency of the COBE device and operators was learned through experience. The relationship of the previous collection estimation using “liters processed” as a measure for the new device was unknown. A process to allow training and use on the Optia device while assuring the patient's CD34⁺ cell collection goal was required.

Materials and Methods: An implementation plan used the Optia according to manufacturer's instructions, using only ACDA as the anticoagulant (no heparin, no aspirin), after staff training by the vendor. The cell therapy laboratory supported the implementation by testing the apheresis product  at 2 hours into the procedure in order to calculate the size of the collected product.

Results:

Determined by 2-hour HPC, Apheresis product sample (“Mid”) CD34+ value N=48*		Average	Range
	Actual Collection Volume (Inlet Volume)	14.655 L ± 4.848 L	5.596 L – 27.207 L
	% to target	142.44 % ± 60.26 %	61.40 % - 335.17 %
	Collection Efficiency (CE)	53.14 % ± 17.40 %	20.52 % - 90.00 %
Prospective prediction by peripheral blood CD34+ (“Pre”) and Predictive Equation	Collection Volume	16.368 L ± 6.003 L	5.477 L – 24.000 L
	% to target	149.08 % ± 49.46 %	56.88 % - 267.67 %

*Excluded two poorly mobilized outliers, each with a peripheral blood CD34⁺ counts of <15x10³/mL

Conclusion: Concurrent CD34+ cell counting facilitated implementation of the Optia collection method to meet the collection goals. Frequent over-collection was corrected with yields collections closer to prescription targets (with slight over collection), mean of 142%. Use of yield information improved collection efficiencies from around 20% to 60% with device adjustment. A predictive equation using pre-CD34+ donor counts was mathematically tested with the 48 collections and found to be highly predictive. The average inlet volume was reduced by >40% and the use of ACDA as the sole anticoagulant further improved safety for the donor.