|Home | About | Journals | Submit | Contact Us | Français|
All blood components undergo loss of potency during storage. These loss-of-potency storage lesions are important in trauma resuscitation because they reduce the haemostatic capacity of mixtures of components that attempt to reconstitute whole blood. Even red cell storage-related loss of potency, which averages 17% with modern additive solutions, is important because 6 units of red cells must be given to achieve the effect of 5 fully potent units.
Loss of potency of stored units of red blood cells, plasma, platelets, and cryoprecipitate were summed for dilutional, storage-related, pathogen reduction-related, and splenic sequestration-related causes and expressed as fractional plasma coagulation factor concentrations and platelet counts.
Production of reconstituted whole blood from 1:1:1 unit ratios of red cells:plasma:platelets is associated with a 38% loss of plasma coagulation factor concentration and 56% loss of platelets. Storage losses of 17% for red cells, 10% for coagulation factors, and 30% for platelets are additive to pathogen reduction-related losses of 18% for coagulation factors and 30% for platelets.
Component preparation and storage-related losses of potency for all blood components are serious problems for trauma resuscitation. Even red cell storage contributes to this problem and this can be made better in ways that can save many lives each year.
Patients suffering severe injury frequently have an accompanying coagulopathy characterised by reduced concentrations of coagulation factors and platelets that arises though haemorrhagic loss, dilution of remaining blood elements with asanguineous fluids, and consumption on damaged vascular surfaces. Acidosis and hypothermia can further impair the function of the remaining coagulation factors and platelets. Treatment is needed to address the underlying causes. Surgical control of all bleeding is desirable but not always fully achievable, especially in damaged tissues. Volume resuscitation to correct haemodynamics and oxygen transport can simultaneously address acidosis and hypothermia, while timely repletion of haemostatic components minimises further blood loss. In this setting, replacement of lost blood volume with crystalloid fluids and red blood cells (RBCs) can lead to dilution of platelets and coagulation factors. This insight provides the basis for balanced resuscitation with an approximate 1:1:1 ratio of RBCs, platelets, and fresh frozen plasma (FFP).
Blood that is stored as separate components and then given to a patient will not attain the original concentration of the separate blood elements because of losses during processing and dilution of components with anticoagulant and additive solutions. With the effects of storage lesions on these products, the in vivo recovery of RBCs, platelets, and plasma coagulation factors can be substantially lower. For patients undergoing operative management, general triggers for transfusion are to maintain a haematocrit greater than 21%, a platelet count above 50×109/L, and an International Normalised Ratio (INR) below 1.5 or 2.01. The recovered portion of transfused blood products is sufficiently low to be barely above these traditional transfusion triggers. Patients with severe trauma are among those who can least afford to have a borderline haemostatic profile; therefore, the minimisation of storage lesions is critical for adequate resuscitation in this patient population.
The act of collecting and storing blood as components dilutes the original donation. A typical whole blood donation of 450 mL will start with an average haematocrit of 42%, coagulation factor concentrations of 100%, and a platelet count of 250×109/L. This original whole blood volume is collected into 63 mL of anticoagulant solution and the whole blood unit is split into its components. The cellular components are leucoreduced in filters that retain some of the initial components and volume. Finally, 100 mL of an additive solution is added to the RBCs.
The resulting RBC concentrate contains about 180 mL of RBCs in 9 mL of anticoagulant, 40 mL of plasma, and 100 mL of additive solution for a total volume of about 330 mL and a storage haematocrit of 55%. After collection and processing, a single whole-blood-derived platelet contains approximately 55×109/L platelets in 9 mL of anticoagulant and 40 mL of plasma. The platelets occupy about 0.5 mL of a 50 mL total volume. The plasma unit typically contains about 200 mL of plasma and 45 mL of anticoagulant, giving a total volume of 245 mL of 80% plasma in anticoagulant. Due to this loss and dilution, an idealised pool of the three components in a 1:1:1 unit ratio of plasma, platelets, and RBCs will have a haematocrit of 29%, a factor concentration of 65%, and a platelet concentration of 88×109/L2.
The above calculations represent the upper limit of attainable haematocrit in a trauma patient receiving nothing but balanced transfusions of units with average sizes and concentrations of fresh products. Barring the removal of storage solutions, the use of unusually concentrated products, or the transfusion of additional amounts of one product that displace the other blood components, a further increase is not possible. Adding a storage lesion to any of the components decreases the recovery even further. In a retrospective study of radiolabelled RBC recoveries, Dumont et al. reported that mean recoveries for radiolabelled RBCs that were transfused autologously after 42 days of refrigerated storage averaged 82.4±6.7% for 641 units 3. The unrecovered fraction of 17.6% represents the best estimate of storage-related loss among current US Food and Drug Administration-approved RBC storage systems.
It appears that it is possible to make better RBC storage solutions. A second-generation additive solution (AS-7) evaluated by Cancelas et al. found a 24-hour recovery at 42 days storage of 88±5%4. Incremental improvements in storage solutions can thus improve recovery and allow for successful balanced resuscitation while allowing more “room” for intravenous fluids for drug delivery.
Table I demonstrates the effect of applying storage lesions of varying size to blood components transfused in a 1:1:1 ratio. These values were calculated with the included equations (Figure 1) after allowing for a storage lesion of the indicated size. According to these estimates, transfused units with a recovery of about 70% produce a haematocrit approximately equal to suggested transfusion triggers.
Recovery of plasma coagulation factors is also highly dependent on the conditions of processing and storage. Factors V and VIII are heat labile, and these factors will quickly degrade if plasma is stored in a non-frozen state. Because Factor VIII is synthesised in endothelial cells and it is an acute phase reactant, it is likely less clinically relevant within the context of severe trauma. In contrast, insufficient repletion of Factor V is likely to be clinically relevant, particularly in severely injured patients with activation of the Protein C system.
A unit of FFP contains about 70% of the plasma from one unit of whole blood. This 200 mL of plasma is diluted with 50 mL of anticoagulant solution to obtain a typical unit volume of 250 mL. The dilution produces a product with 80% of the original factor concentration of the donated plasma. This FFP is added to the remaining plasma present in RBCs and platelets that has been diluted with anticoagulant and red cell storage solution; the combined factor concentration in this reconstituted whole blood (RWB) is about 65%. In a previous study of RWB wherein these products were combined in a 1:1:1 fashion, the pooled products were found to have an average INR of 1.31 and PTT of 425. The clotting factor concentrations measured in these products were similar to the expected concentrations calculated for RWB in this present report.
This concentration, however, assumes perfect storage and recovery of coagulation factors, and it assumes that the donor has factor concentrations at the population median. While clotting factor levels actually vary widely and reference ranges generally fall from 50 to 150%, this detail likely has a limited impact on average factor levels in massive transfusions where many units of plasma are given. In contrast, storage conditions do decrease factor levels across all units. Cardigan et al. found that room temperature storage of whole blood for 24 hours caused a 23% decrease of Factor VIII and small but statistically significant decreases of Factors II, IX, and X. These units also had PTT prolongation of about two seconds when compared to plasma frozen within 8 hours of collection6.
In addition to loss of factors during the component production process, prolonged storage of plasma in the thawed state also leads to degradation. Downes et al. measured the factor content of refrigerated thawed plasma over a period of five days. As confirmed by prior studies, Factor VIII was the most labile factor, losing 40% of its concentration from day 1 to day 5. Factors V and VII saw decreases of 16 and 20% respectively, and other measured factors had negligible decreases7. Because pre-thawed plasma is disproportionately used in massive transfusions, degradation during storage for these factors disproportionately affects patients with severe trauma and it likely reduces the effectiveness of the replacement product for some patients.
This degradation is compounded by the fact that 30% of the plasma being given to the patient, that which is present in RBCs and platelets, is not stored in a way meant to preserve coagulation factors. While the amount of degradation varies by factor, it is reasonable to estimate that Factor V and VIII levels are low in RBCs stored for six weeks, and that more stable factors also suffer some degradation. Similarly, the approximately 40 mL of plasma present in a unit of whole blood platelets has been stored at room temperature; it is reasonable to estimate that the coagulation factors in this plasma have degraded at least as quickly as indicated by the data of Downes et al., and it is likely that further losses have been sustained because platelets are stored at room temperature. Pathogen reduction adds further degradation of factors: plasma treated with methylene blue contains approximately 20% less fibrinogen, and solvent/detergent treated plasma has been shown to have 10% reductions of all factor levels8. A graph of the coagulation factor content and platelet count of RWB at varying recovery levels can be seen in Figure 2. The figure also includes estimates for pathogen-reduced products.
Among blood components, platelets have the greatest discrepancy between the amount of product recovered after transfusion and its theoretical expectation. An apheresis collection or a 6 unit pool of whole blood-derived platelets containing over 300×109 platelets/L would be expected to produce an increase of circulating platelets of over 60×109/L. However, as about one-third of these transfused platelets become sequestered in the spleen and 30% of the remainder are lost to storage lesions, the average 4-hour increment in patients transfused with platelets in the PLADO trial was about 28×109/L9. While part of this reflects losses due to patient disease states, most is degeneration during storage, as further analysis found that only about 1% of those patients in the trial had alloimmunisation-based refractoriness10. With an average increment of 28×109/L, attributing half of the unrealised gain to consumption and half to a storage lesion implies that storage losses are nearly the same magnitude as the total increase in circulating platelets after transfusion.
As with pathogen-reduced FFP, pathogen-reduced (PR) platelets have been shown to improve product safety at the cost of decreased platelet recovery and activity. Prior studies have shown a reduction in total platelet recovery to the amount of approximately 25–44%11. One representative study found autologous recoveries in PR platelets of 50.0±18.9%, compared to 66.5±13.4% for control platelets12. Applying a 30% reduction to an expected platelet count in RWB would reduce the platelet count from a maximum of 88×109/L platelets to a maximum of 62×109/L for pathogen-reduced platelets. The cumulative effect of both pathogen reduction and storage lesions is demonstrated in Figure 2, using the equations in Figure 1. This, if transfused, is further reduced by splenic sequestration of one-third to a platelet count of 41×109/L before taking into account any storage lesions on the product. An additional storage lesion of 25% would reduce the platelet count of that RWB to, at best, about 30×109/L. In a patient with severe trauma, pathogen reduction reduces platelet counts from what is already a tenuously low replacement product.
For patients with severe trauma, correction of coagulopathy is accomplished by surgical control of bleeding and balanced resuscitation with stored blood products in ratios that attempt to restore haemostasis. This reconstituted whole blood is more diluted than whole blood because of the addition of storage solution and anticoagulant. The end product is a mixture that is, at best, only slightly better than traditional haematological indices associated with unfavourable outcomes; storage lesions exacerbate this problem. Red cells, FFP, and platelets all have storage lesions of a clinically significant magnitude to a patient with severe trauma because reconstituted whole blood is already limited in its haemostatic profile by the dilution of stored components. Minimisation of storage lesions, therefore, represents an excellent opportunity for improving resuscitation in these patients.
JAM wrote the article, performed the calculations, prepared the illustrations, and reviewed the final product. JRH provided the basic idea, gave suggestions on citations and design of graphics, edited the article, and reviewed the final product.
The Authors declare no conflicts of interest.