Lipids: Free Fatty Acids, Eicosanoids, and Lysophospholipids and the Pro-Inflammatory Effects of Transfusion

Abstract SCI-48

Stored packed red blood cells (PRBCs) have been implicated in increased morbidity and mortality in critically ill patients, patients requiring cardiothoracic surgery, and injured patients following blunt and/or penetrating trauma. Lipids are generated during the routine storage of cellular blood components and have been implicated in transfusion-related acute lung injury (TRALI) and postinjury multiple organ failure (MOF). These lipids are comprised of two classes, as denoted by their retention times on normal-phase HPLC: 1) nonpolar lipids, including arachidonic acid, 5-hydroxy-eicosotetranoic acid (HETE), 12-HETE, 15-HETE, and 5-oxo-eicosotetranoic acid (5-oxo-ETE); and 2) a mixture of lysophosphatidylcholines (lyso-PCs), including stearoyl-, oleoyl-, or palmitoyl-lyso-PCs as well as C16 or C18 lyso-platelet activating factor. Importantly, all of these lipids were identified by quadrupole LC/MS/MS. The former are derived from red blood cells, because they are generated during the routine storage of prestorage leuko- and platelet-reduced PRBCs, while the latter are platelet-derived, and are generated during storage of platelet concentrates (PC) and unmodified PRBCs but not leukocyte-reduced PRBCs (LR-PRBCs). Generation of the nonpolar lipids requires an active phospholipase, such as peroxiredoxin-6, which accumulates during PRBC and LR-PRBC storage and appears to be active and T-phosphorylated. The nonpolar lipids do not accumulate during PC storage. These lipids are released into the plasma fraction of the stored component and have been shown to: 1) prime the polymorphonuclear leukocytes (PMN) oxidase; 2) activate primed PMNs; and 3) serve as the second event in a two-event model of TRALI in both rodents and sheep. This lipid priming of PMNs occurs through activation of specific receptors, for example G2A, for lyso-PCs, and causes activation of G-protein-linked cellular receptor, G-protein release, and stimulation of kinase cascades. This results in translocation of the cytosolic oxidase components and a change from a resting PMN phenotype to a hyperactive, adherent phenotype. Conversely, stored PRBCs have also been implicated in MOF. However, acute lung injury does not occur until 72 hours postinjury, implicating PRBCs as a first event(s). The nonpolar lipids that accumulate during routine storage activate primary human pulmonary microvascular endothelial cells (HMVECs) and primary human liver sinusoidal endothelial cells (LSECs), as quantified by increased surface expression of intercellular adhesion molecule-1 (ICAM-1) and the synthesis and release of chemokines, for example IL-8. This proinflammatory activation results in PMN adherence, and such HMVEC activation occurs through activation of the BLT2 receptor and activation of a PKC-dependent kinase cascade. Stored PRBCs (day 42) may serve as the first event in a two-event rodent model of acute lung injury. We conclude that lipids generated during the routine storage of cellular blood components have proinflammatory effects in vitro and in vivo, which may explain some of the adverse events of transfusion, including TRALI and postinjury MOF, and that inhibition and avoidance of these compounds may make transfusions safer. Further research is already under way in several prospective trials, including: 1) Age of Blood Components transfused in the PICU (ABC-PICU); 2) red cell storage duration and outcomes in cardiac surgery at the Cleveland Clinic; 3) Red Cell Storage duration Study (RECESS); and 4) the Age of Blood Evaluation trial (ABLE).