Storage duration of packed red blood cells transfused during veno-venous extracorporeal membrane oxygenation is associated with elevated pulmonary artery pressure and lung injury in a sheep model

Abstract

Background: Veno-venous extracorporeal membrane oxygenation (VV-ECMO) is associated with a high transfusion burden. While trials have concluded that red blood cell (RBC) storage does not impact patient morbidity and mortality in the critically ill or cardiac surgical cohorts, evidence is sparse for ECMO cohorts. A sheep model was to investigate this question. On an underlying injury of smoke inhalation, we compared fresh (< 5 days) or stored (35-42 days) RBC transfusion with no transfusion. Clinically relevant outcomes included pulmonary artery pressure as well as biochemical or histopathological markers of lung, liver, and renal injury.

Methods: Twenty-four female Merino/Samm Border Leicester Cross sheep were anaesthetised and placed on pressure-controlled ventilation. They were instrumented for continuous haemodynamic monitoring. They then underwent smoke inhalation injury, followed two hours later by veno-venous ECMO. Sheep were randomised to three groups: control (no transfusion (n = 8)), fresh RBC (n = 8) and stored RBC (n = 8) transfusion occurring six hours after ECMO initiation. Blood samples were taken regularly, and 24 h post-ECMO, animals were sacrificed. Post-mortem samples of lung and kidney were collected for post-mortem analyses.

Results: Following ECMO initiation and transfusion, pulmonary artery pressure increased in the stored RBC group. Histopathological analysis also demonstrated a significantly elevated lung injury in this group. This lung injury was characterised by greater extravasation of inflammatory cells as well as bronchiole damage and oedema. Renal histopathology showed no significant differences between groups. Alanine aminotransferase, aspartate aminotransferase and bilirubin levels increased in a time dependent manner post-transfusion but there were no treatment-associated differences. During experimentation, the coagulation profile changed with firmer clots forming more quickly. Differences were observed between both transfusion groups and controls.

Conclusions: Transfusion of stored RBC elevated pulmonary artery pressure when compared to fresh RBC transfusion and controls. This change was correlated with a greater post-transfusion lung injury in the stored RBC group. Further investigation assessing whether insufficient nitric oxide had a role in these findings is warranted, as is consideration of longer ECMO durations and greater transfusion volumes.

Keywords: Extracorporeal membrane oxygenation; Lung injury; Pulmonary artery pressure; Red blood cell storage; Transfusion.

Fig. 1

Experimental Intervention and Outcome Timeline. Experimental timeline with measurements and samples taken. FBC = full blood count, ROTEM = rotational thromboelastography, ELISA = enzyme linked immunoassay, RT-PCR = real-time polymerase chain reaction, s-ALI = smoke-induced acute lung injury

Fig. 2

Mean Pulmonary Artery Pressure during ECMO Initiation, Transfusion, and Pre-Euthanasia. The initiation of ECMO (A and B) resulted in a large but highly variable increase in pulmonary artery pressure (PAP) in the stored RBC transfusion group (A). However, this was no longer evident when adjusted to fold change (B). After transfusion (C and D), an increase is again seen in the stored RBC group (C). When viewing fold change, this increase was less pronounced but more consistent than that which happened after ECMO initiation (D). PAP was stable for the hour preceding euthanasia (E and F). Control (black/circle, n = 8), fresh RBC (green/square, n = 7) and stored RBC (red/circle, n = 8). ECMO = extracorporeal membrane oxygenation and PAP = pulmonary artery pressure. Mixed effects model with Bonferroni’s multiple comparison. p < 0.05: * = Control vs Stored, ^ = Fresh vs Stored

Fig. 3

Post-Mortem Analysis of Histopathology; Injury, Cell Counts and Oedema (A) and summary descriptions (B) of histopathological grading showed increased bronchiole damage and extravascular neutrophil infiltration after stored RBC transfusion. Stored RBC transfusion was associated with no change in Lung parenchyma PMN (C), but decreased Macrophages (D) were observed. There were no differences present in Wet to Dry ratio (E) or fluid balance (F). Control (black/circle, n = 8), fresh RBC transfused (green/square, n = 7) and stored RBC transfused (red/circle, n = 8). PMN = polymorphonuclear leukocytes. One-Way ANOVA with Bonferroni’s multiple comparison. p values < 0.05 are bolded

Fig. 4

Serum Creatinine, Bilirubin, ALT, and AST. Serum creatinine levels (A) were initially different between transfusion groups and control. This difference was no longer observed for fresh RBC by 6H but persisted for stored RBC until 18H. Total serum bilirubin (B) was initially similar between groups, with an increase compared to the control group evident in the fresh RBC group at 12H and in the stored RBC group at 6H and 12H. Direct serum bilirubin (C) was initially similar between groups, with an increase compared to the control group evident in the fresh RBC and stored RBC groups at 12H. Alanine aminotransferase (ALT) (D) and aspartate aminotransferase (AST) (E) increased in a time dependent manner for all treatment groups but not treatment associated differences were present. Control (black/circle, n = 8), fresh RBC transfused (green/square, n = 7) and stored RBC transfused (red/circle, n = 8). Mixed effects models with Bonferroni’s multiple comparison. * p < 0.05 Control vs Stored: # p < 0.05 Control vs Fresh