(P-TS-99) Small-Scale Model of RBC Cryopreservation and Transient Warming is Not Predictive of Large-Scale Unit Hemolysis Post-Deglycerolization

Small-scale (SS) models of red blood cell (RBC) cryopreservation have been developed as a tool to predict the outcomes of large-scale (LS) red cell concentrate (RCC) counterparts while avoiding LS processing.  We sought to analyze the effectiveness of a SS RBC cryopreservation model in predicting quality outcomes from transient warming events (TWEs) impacting frozen RBCs.  TWEs pose a challenge to the management of frozen RCC inventories, as the impact on quality is uncertain and leads to discard.

 

Study

Design/Methods:

Six pools of nine ABO/Rh matched RCCs were combined using a pool-split method, glycerolized using the ACP 215, and frozen in a < -65 °C freezer (n=54) (LS RCCs). Prior to glycerolization, an 80 mL sample was removed from each pool, manually glycerolized, divided into 1 mL cryovials (73 vials/pool), and frozen at 1°C/min in a < -65 °C freezer (SS RCCs). Two types of TWEs were applied to both groups: (1) “fast” TWEs (warming to -20 °C at room temperature then back to < -65 °C) and (2) “slow” TWEs (exposure to -20 °C for 2 hours then back to < -65 °C). LS and SS RCCs were exposed to 0 (control), 1, or 10 TWEs for each type of TWE, with an additional room temperature thaw-refreeze condition. LS RCCs were thawed, deglycerolized and resuspended in AS-3 using the ACP 215 and stored hypothermically (2 - 6 °C). SS RCCs were thawed at a comparable rate to LS RCCs, deglycerolized manually with subsequent saline washes, resuspended in AS-3, and stored hypothermically (2 - 6 °C) in 1.5 mL microtubes. RBC hemolysis, and hematocrit were measured post-thaw and at 0 and 24 hours post-deglycerolization (PDG) for both LS and SS RBCs.

Results/Findings:

No significant differences in the hematocrit were identified between the LS and SS groups at either testing point. SS hemolysis values were significantly higher than LS hemolysis values for all groups at both testing points (p< 0.001, Table 1). Variability was highest in the SS experimental group exposed to 10 “fast”/”slow” TWEs at both testing points post-deglycerolization.

Conclusions: Our results indicate that this SS cryopreservation model does not provide an accurate representation of LS RCC hemolysis under any TWE condition examined. Considerable discrepancies among hemolysis values due to tech-tech, replicate, and temporal variability along with the manual nature of the SS method continues to be a challenge. Further optimization of the SS model is necessary before RBC quality results between SS and LS units would be in agreement at extended storage times PDG, or after multiple TWEs.