Experimental Antibody Filtration Inhibits Antibody-Mediated Neutrophil Priming and TRALI in An In Vivo model

Abstract 41

Transfusion-related acute lung injury (TRALI) is a leading cause of transfusion-related death with a majority of the reported cases secondary to the infusion of antibodies (Abs) contained within the plasma/blood component. An experimental filter that removes IgG was developed. We hypothesize that filtration of plasma with antibodies to leukocyte antigens will decrease both antibody-mediated priming of PMNs and antibody-mediated TRALI in a two-event in vivo model. Methods: Human plasma was drawn from healthy volunteers and IgG concentrations were measured before and after filtration. Plasma was obtained from two multiparous female donors: one with antibodies to HLA-A2 and to DR7 and the other with antibodies against HNA-3a. These plasma samples were filtered (F-Plas) or left as an unmodified control (Plas) and the anti-leukocyte antibodies were measured in a blinded fashion in referral labs using flow cytometry and Luminex™ beads or standard granulocyte antibody detection assays. These plasma samples were then used to prime the fMLP-activated respiratory burst, measured as the SOD-inhibitable reduction of cytochrome c (nmol O₂⁻/min), of PMNs from HNA-3a⁺ donors or donors homozygous donors for HLA-A2, respectively. For the two-event in vivo modeling rats were incubated with 2 μg/ml endotoxin (LPS, S. enteritides) or saline (NS) for 2 hours (first event) and then were transfused with heat-treated human plasma that contained 25 μg/ml of an antibody against the MHC class I antigen OX27 that was either filtered (or left unmodified) prior to infusion (second event) followed by Evans Blue dye (EBD). ALI was measured as %EBD leak from the plasma into the bronchoalveolar lavage fluid. Statistical differences were measured via paired (PMN priming) or independent (in vivo TRALI) ANOVA, and data are reported as the mean ± the standard error of the mean. *=p<.05 vs. all groups (Table). Results: Plasma filtration removed 98±2.1% of IgG from normal plasma and both the antibodies to HNA-3a and HLA-A2 such that they were no longer detected (HLA-A2: 94 Luminex™ units (LU) pre-filtration and 0 LU units post-filtration and DR7: 30 LU pre-filtration and 0 LU post-filtration). In addition, filtration also inhibited the priming activity of the plasma containing antibodies to HNA-3a and HLA-A2 on HNA-3A⁺ PMNs and HLA-A2⁺ PMNs, respectively. Moreover, the plasma spiked with antibodies to OX27 caused ALI in LPS-treated rats, but not NS treated animals, which was inhibited by filtration. We conclude that this experimental filter removes IgG and detectable amounts of specific antibodies to HLA and HNA ligands as well as obviating the priming activity of these antibodies in PMNs which express the cognate antigens. Filtration of plasma spiked with specific antibodies to MHC ligands also abrogated the antibody-induced TRALI in a two-event, in vivo model. Such a filtration step could mitigate antibody-mediated TRALI.