The impact of venovenous extracorporeal membrane oxygenation on cytokine levels in patients with severe acute respiratory distress syndrome: a prospective, observational study.

OBJECTIVE: The immunoinflammatory response is central to the pathogenesis of acute respiratory distress syndrome (ARDS). However, little is known how this is affected by venovenous (VV) extracorporeal membrane oxygenation (ECMO). Our objective was to investigate the factors that influence the inflammatory response of patients with ARDS undergoing VV ECMO, and to analyse the impact of this response on hospital mortality. DESIGN AND SETTING: A prospective observational study of all consecutive patients with severe ARDS who had VV ECMO at a tertiary German ECMO centre from 2009 to 2015. Patients without complete datasets were excluded. Cytokines (interleukin [IL]6, IL8 and tissue necrosis factor [TNF]α) and inflammatory markers (white cell count and C-reactive protein) were assessed before ECMO initiation and on Days 1, 5 and 10, before explantation and at explantation. RESULTS: A total of 262 adult patients undergoing VV ECMO were analysed. Their median Sequential Organ Failure Assessment score was 12, PaO2/FiO2 ratio was 64 mmHg, and overall in-hospital mortality was 34%. Cytokine levels fell quickly within 24 hours and fell further over the first 5 days. Extra-pulmonary ARDS was associated with higher IL6 and IL8 levels compared with pulmonary ARDS. Mechanical ventilation with positive end-expiratory pressure ≥ 15 cmH2O before ECMO was associated with higher IL6, IL8 and TNFα levels. Driving pressures ≥ 19 cmH2O before ECMO were associated with higher IL8 levels. Non-survivors had higher IL6 and IL8 levels for the duration of ECMO. CONCLUSION: Cytokine levels, on average, fall rapidly after initiation of VV ECMO, which may be related to the reduction of invasiveness of mechanical ventilation. Higher cytokine levels are associated with extrapulmonary causes of ARDS, more aggressive mechanical ventilation before VV ECMO, and mortality.

Genetic and clinical risk factors for fluid overload following open-heart surgery.

BACKGROUND: Post-operative fluid overload following cardiac surgery is associated with increased morbidity and mortality. We hypothesised that genetic variations and pre-operative clinical factors predispose some patients to post-operative fluid overload. METHODS: Perioperative variables were collected prospectively for 1026 consecutive adults undergoing open-heart surgery at St. Olavs University Hospital, Norway from 2008-2010. Post-operative fluid overload was defined as a post-operative fluid balance/kg ≥ the 90th percentile of the study population. Genotyping was performed for 31 single-nucleotide polymorphisms related to inflammatory/vascular responses or previously associated with complications following open-heart surgery. Data were analysed using logistic regression modelling, and the findings were internally validated by bootstrapping (n = 100). RESULTS: Homozygous carriers of the common G allele of rs12917707 in the UMOD gene had a 2.2 times greater risk of post-operative fluid overload (P = 0.005) after adjustment for significant clinical variables (age, duration of cardiopulmonary bypass, and intraoperative red cell transfusion). A genetic risk score including 14 single-nucleotide polymorphisms was independently associated with post-operative fluid overload (P = 0.001). The number of risk alleles was linearly associated with the frequency of fluid overload (odds ratio per risk allele 1.153, 95 % confidence interval 1.056-1.258). Nagelkerke's R(2) increased with 7.5% to a total of 25% for the combined clinical and genetic model. Hemofiltration did not reduce the risk. CONCLUSION: A common variation in the UMOD gene previously shown to be related to renal function was associated with increased risk of post-operative fluid overload following cardiac surgery. Our findings support a genetic susceptibility to disturbed fluid handling following cardiac surgery.