Retrospective Analysis Using Statistical Modeling on the Effect of Repeated Stimulated Granulocyte Donation on Hematopoietic Indices in Granulocytapheresis Donors

Abstract

BACKGROUNDShort and long term clinical effects of mobilization regimens in hematopoietic stem cell and granulocytapheresis donors have been characterized. The longitudinal hematopoietic changes in RBC, MCV, platelet, and Hgb values related to repeat stimulated granulocytapheresis donations were examined STUDY DESIGN AND METHODSPre-apheresis CBCs for consecutive GCSF and dexamethasone or GCSF- alone stimulated healthy volunteer granulocytapheresis donors (n=496) between Oct 1994- May 2017 were compared to unstimulated healthy granulocytapheresis donors (n=526). Plateletapheresis donors served as controls (n=3743). The longitudinal change in RBC, platelet, Hgb and MCV for these donor groups were modeled using a linear mixed effects model to control for the intra-subject correlation of repeat donations. Variables in the analysis included: donation category, age, race, gender, and the number of previous donations for each category by time interval: 0-1, 1-2, 2-3, 3-4, and 4+ years ago. The models were built using the nlme package, R version 3.3.1, and used the computational resources of the high performance computing Biowulf cluster. Contrasts were done to illustrate the effect of donation number on pre-collection counts. In the contrast models, a 60-year old donor was set as baseline to place emphasis on age related effects. Male gender, with race set to Caucasian was selected to coincide with our center's majority of granulocyte donors. RESULTSThe pseudo-R2 for the pre-RBC model was 0.795; for the pre-platelet model it was 0.765 and for pre-Hgb it was 0.748.Pre-RBC modelAge was found to have a minimal negative effect on pre-apheresis RBC counts (b=-0.003, P<0.001). This minor effect was found in all donor groups. This translated to a difference of 0.003 M/µL (P<0.0001) in the marginal mean between a 40 year GCSF-Dex donor and a 60 year GCSF-Dex donor.Pre-Plt modelAfrican American race, was associated with a minor increased platelet count (Black: b=14.3, P=0.003 vs White b=-1.7 P=0.671). The Dex and the GCSF-Dex donor groups had statistically significantly higher platelet counts of 30.7 K/uL (P<0.001) and 35.8 K/uL (P<0.001), respectively. Age was not found to have a significant effect. Male gender had a 25K/uL lower platelet level (P<0.001) across all groups.Pre-Hgb modelAfrican American race, was associated with a decrease in pre-apheresis Hgb (b=0.537,P<0.001) across all groups, including the platelet control. Notably, African Americans had a smaller GCSF-Dex donor group increase in Hgb (White donors: b=1.56, P<0.001 vs Black donors: b=1.02, P=0.012).Effect of donation number (Figure 1)For the GCSF-Dex stimulated group the platelet, RBC counts and Hgb decreased 22.6 K/µL (-8.8%, P<0.001), 0.16 M/µL (-3.3%, P<0.001) and 0.45 g/dL (-3%, P<0.001) respectively between donations 1 and 20. This effect was still present at the 2-3 year interval for each (pre-PLT: b=-1.01, SE=0.21, P<0.001) (pre-RBC: b=-0.013, SE=0.0014, P<0.001) (pre-Hgb: b=-0.022, SE=0.004, P<0.001).For the unstimulated granulocytapheresis donor group, between donations 1 and 20, the platelet and RBC count decreased by 24.8 K/µL (-11.8%, P<0.001) and 0.305 M/µL, respectively (-6.4%, P<0.001). Hgb decreased 0.75 g/dL (-5.2%, P<0.01). The effect in the unstimulated donor group was statistically significant up to year 3 post-donation for RBC (b=-0.013, SE=0.001, P=0.001) and PLT counts (b=-2.06, SE=0.89, P=0.021). CONCLUSIONSThough the absolute changes in CBCs were small; past granulocytapheresis donations were found to have a statistically strong negative effect on precollection RBC counts, platelet counts, and Hgb values. These effects were evident in granulocyte apheresis donor groups but not plateletpheresis donors. In this statistical model, for both these groups, the effect of past donations on RBC, Platelet counts and hemoglobin were detectable 2 years post-donation. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.