Rationale: Fluid resuscitation is widely considered a life-saving intervention in septic shock; however, recent evidence has brought both its safety and efficacy in sepsis into question. Objectives: In this study, we sought to compare fluid resuscitation with vasopressors with the use of vasopressors alone in a hyperdynamic model of ovine endotoxemia. Methods: Endotoxemic shock was induced in 16 sheep, after which they received fluid resuscitation with 40 ml/kg of 0.9% saline or commenced hemodynamic support with protocolized noradrenaline and vasopressin. Microdialysis catheters were inserted into the arterial circulation, heart, brain, kidney, and liver to monitor local metabolism. Blood samples were recovered to measure serum inflammatory cytokines, creatinine, troponin, atrial natriuretic peptide, brain natriuretic peptide, and hyaluronan. All animals were monitored and supported for 12 hours after fluid resuscitation. Measurements and Main Results: After resuscitation, animals that received fluid resuscitation required significantly more noradrenaline to maintain the same mean arterial pressure in the subsequent 12 hours (68.9 mg vs. 39.6 mg; P = 0.04). Serum cytokines were similar between groups. Atrial natriuretic peptide increased significantly after fluid resuscitation compared with that observed in animals managed without fluid resuscitation (335 ng/ml [256–382] vs. 233 ng/ml [144–292]; P = 0.04). Cross-sectional time-series analysis showed that the rate of increase of the glycocalyx glycosaminoglycan hyaluronan was greater in the fluid-resuscitated group over the course of the study (P = 0.02). Conclusions: Fluid resuscitation resulted in a paradoxical increase in vasopressor requirement. Additionally, it did not result in improvements in any of the measured microcirculatory- or organ-specific markers measured. The increase in vasopressor requirement may have been due to endothelial/glycocalyx damage secondary to atrial natriuretic peptide–mediated glycocalyx shedding.

Extracorporeal membrane oxygenation (ECMO) is a life-saving modality used to manage cardiopulmonary failure refractory to conventional medical and surgical therapies. Despite advances in ECMO equipment, bleeding and thrombosis remain significant complications. While the flow rate for ECMO support is well recognized, less is known about the minimum-rate requirements and haemostasis. We investigated the relationship between different ECMO flow rates, and their effect on haemolysis and coagulation. Ten ex-vivo ECMO circuits were tested using donated, < 24-h-old human whole blood, with two flow rates: high-flow at 4 L/min (normal adult cardiac output; n = 5) and low-flow at 1.5 L/min (weaning; n = 5). Serial blood samples were taken for analysis of haemolysis, von Willebrand factor (vWF) multimers by immunoblotting, rotational thromboelastometry, platelet aggregometry, flow cytometry and routine coagulation laboratory tests. Low-flow rates increased haemolysis after 2 h (p = 0.02), 4 h (p = 0.02) and 6 h (p = 0.02) and the loss of high-molecular-weight vWF multimers (p = 0.01), while reducing ristocetin-induced platelet aggregation (p = 0.0002). Additionally, clot formation times were prolonged (p = 0.006), with a corresponding decrease in maximum clot firmness (p = 0.006). In an ex-vivo model of ECMO, low-flow rate (1.5 L/min) altered haemostatic parameters compared to high-flow (4 L/min). Observed differences in haemolysis, ristocetin-induced platelet aggregation, high-molecular-weight vWF multimers and clot formation time suggest an increased risk of bleeding complications. Since patients are often on ECMO for protracted periods, extended-duration studies are required to characterise long-term ECMO-induced haemostatic changes.

Sepsis is a multi-system syndrome that remains the leading cause of mortality and critical illness worldwide, with hemodynamic support being one of the cornerstones of the acute management of sepsis. We used an ovine model of endotoxemic shock to determine if 0.9% saline resuscitation contributes to lung inflammation and injury in acute respiratory distress syndrome, which is a common complication of sepsis, and investigated the potential role of matrix metalloproteinases in this process. Endotoxemic shock was induced in sheep by administration of an escalating dose of lipopolysaccharide, after which they subsequently received either no fluid bolus resuscitation or a 0.9% saline bolus. Lung tissue, bronchoalveolar fluid (BAL) and plasma were analysed by real-time PCR, ELISA, flow cytometry and immunohistochemical staining to assess inflammatory cells, cytokines, hyaluronan and matrix metalloproteinases. Endotoxemia was associated with decreased serum albumin and total protein levels, with activated neutrophils, while the glycocalyx glycosaminoglycan hyaluronan was significantly increased in BAL. Quantitative real-time PCR studies showed higher expression of IL-6 and IL-8 with saline resuscitation but no difference in matrix metalloproteinase expression. BAL and tissue homogenate levels of IL-6, IL-8 and IL-1β were elevated. This data shows that the inflammatory response is enhanced when a host with endotoxemia is resuscitated with saline, with a comparatively higher release of inflammatory cytokines and endothelial/glycocalyx damage, but no change in matrix metalloproteinase levels.

Heart and lung transplantation is frequently the only therapeutic option for patients with end stage cardio respiratory disease. Organ donation post brain stem death (BSD) is a pre-requisite, yet BSD itself causes such severe damage that many organs offered for donation are unusable, with lung being the organ most affected by BSD. In Australia and New Zealand, less than 50% of lungs offered for donation post BSD are suitable for transplantation, as compared with over 90% of kidneys, resulting in patients dying for lack of suitable lungs. Our group has developed a novel 24 h sheep BSD model to mimic the physiological milieu of the typical human organ donor. Characterisation of the gene expression changes associated with BSD is critical and will assist in determining the aetiology of lung damage post BSD. Real-time PCR is a highly sensitive method involving multiple steps from extraction to processing RNA so the choice of housekeeping genes is important in obtaining reliable results. Little information however, is available on the expression stability of reference genes in the sheep pulmonary artery and lung. We aimed to establish a set of stably expressed reference genes for use as a standard for analysis of gene expression changes in BSD.