Background and Goal of Study:Vasoplegia after Left Ventricular Assist Device (LVAD) insertion occurs frequently and is associated with increased mortality and morbidity. In search of clinical factors involved in the development of vasoplegia we have previously identified previous cardiothoracic surgery, preoperative
dopamine use, increased bilirubine levels and decreased creatinine clearance as independent predictors.
The current work focuses on fundamental biological mechanisms and explores the hypothesis that preoperative molecular events are associated with and help to predict post-LVAD vasoplegia.
Materials and Methods:Vasoplegia was defined on the basis of vasodilation, hemodynamic criteria and high vasopressor requirements during the first 48 hours after first longterm LVAD insertion. We have selected 25 bilogical determinants (pro- and anti-inflammatory cytokines, markers of endothelial activation,
apoptosis, and copeptin and thyroxin). R-Studio for supervised cluster analysis
was used to evaluate the relation between biomarkers and post-LVAD vasoplegia and SPSS to explore differences in biomarkers between both groups.
Results and Discussion:In 127 patients operated on between 2006 and 2016, the incidence of postoperative vasoplegia was 29.1%. Both cytokines TNFR1 and TNFR2 and endothelial markers ANG2 and sVCAM were higher in vasoplegia patients (p<0.05). TIMP1 and FAS were higher in vasoplegia patients implicating potential involvement of apoptosis and extracellular matrix in vasoplegia (p<0.05). ROC
analysis was used to investigate if these preoperative biomarkers could predict future vasoplegia. ROCs varied between 0.62 and 0.71. Inclusion of the biomarkers into the initial model (4 clinical markers) resulted in a predictive value of 0.80 (95% CI 0.72-0.89) compared to 0.76 (95%CI 0.67-0.86) for the initial model.
Conclusion(s):Preoperative plasma levels of certain biomarkers of inflammation, apoptosis and endothelial activation were significantly higher in patients developing post-LVAD insertion vasoplegia compared to controls. However, these biomarkers fail to predict postoperative vasoplegia on their own, and  give a non significant improvement of the preoperative prediction model based solely on clinical parameters. These findings seem to support the inflammation and endothelial activation concept of vasoplegia. Further studies are needed to validate our
findings and expand the evaluation of biological events in the intra- and postoperative period.

04 / Preoperative biological determinants of postoperative vasoplegia after first time long term LVAD insertion