Poster W3 - AN AUTOMATED SYSTEM TO MAINTAIN A NORMOGLYCEMIC ENVIRONMENT IN RED BLOOD CELL STORAGE

Introduction: Transfusion is a critical component of modern healthcare, providing life-saving support to patients in need with nearly 30,000 units of red blood cells (RBCs) provided daily in the United States alone. However, the storage of RBCs presents significant challenges, as they are susceptible to biochemical and structural changes that can impact their quality and therapeutic efficacy. During storage, RBCs are exposed to hyperglycemic environments that may be a determinant in the chemical and physical properties of the stored cells. To address this issue, the Spence lab has developed an automated feeding system that precisely regulates glucose levels in RBC storage bags and maintains a normoglycemic environment. Importantly, our system improves many properties of the stored RBC without impacting the current RBC storage process.

Materials and

Methods: Whole blood was collected from informed and consenting donors using a procedure approved by the Biomedical and Health Institutional Review Board of Michigan State University. Citrate phosphate dextrose (CPD) and additive solution 1 (AS-1) are the anticoagulant and additive solution most commonly used in the RBC storage process. Both of these solutions were prepared in-house as well as CPD-N and AS-1N, which are the developed normoglycemic alternatives pioneered in our labs. After collection into tubes containing either CPD or CPD-N, the RBCs were processed into either traditional hyperglycemic (AS-1, ~40 mM glucose) or normoglycemic (AS-1N, 4-6 mM glucose) additive solutions.

An automated device was designed to maintain normoglycemic levels in AS-1N. Previously, a gravity-driven system dispensed precise 100 mM glucose volumes into RBC bags every 3 days based on predicted utilization rates using a programmable solenoid valve1. A new peristaltic pump-based design was implemented. Four pumps slot into a housing chamber and connect to a driver board controlled by an Arduino microcontroller. The Arduino receives instructions from control software on a single-board computer over USB. Custom python-based software manages bag feeding based on current glucose levels, storage time, and bag volume. The system is controlled via a web interface. Glucose levels were tracked throughout storage by periodically sampling the RBC units. Various assays were performed at multiple time points including hematocrit, hemolysis, pH, lactate levels, osmotic fragility, and intracellular sorbitol levels. Statistical analysis compared AS-1 and AS-1N storage conditions using two-way ANOVA and Tukey's test across 2-10 donors per assay.

Results, Conclusions, and Discussions: The automated glucose feeding device successfully delivered precise, predictable volumes of concentrated glucose solution to maintain physiological glucose levels (4-6 mM) in the normoglycemic AS-1N stored RBCs for up to 42 days. Although some deviation outside the normoglycemic range occurred at later time points, the overall average was well-controlled. Hemolysis levels and pH values were statistically equivalent between AS-1 and AS-1N stored RBCs throughout storage, remaining below 1% hemolysis, which is the FDA threshold for transfusion. RBCs stored in AS-1N exhibited significantly reduced osmotic fragility compared to AS-1 for the first 23 days of storage, while intracellular sorbitol levels were significantly lower in AS-1N at all time points measured. While the equivalent hemolysis, lactate, and pH suggest normoglycemia did not impair glycolysis, the reduced sorbitol and osmotic fragility indicate less oxidative stress and better membrane integrity preservation. This could translate to improved RBC quality and function after transfusion. Overall, the automated feeding system enabled adequate comparison of normo- and hyperglycemic storage, revealing potential benefits of normoglycemic methods while overcoming a key translation barrier through autonomous, sterile glucose control.

Acknowledgements (Optional): This work was supported by funding from the NIH (HL156440).