Plasma lipidomics reveals potential lipid markers for the prediction of delayed neurocognitive recovery after cardiac surgery with cardiopulmonary bypass

Abstract

BackgroundDelayed neurocognitive recovery (DNR) is a cognitive change that affects patients after anesthesia and surgery, with a high incidence, severely impairing their quality of life. We identified relevant metabolic pathways to determine predictive lipid markers and understand the pathogenesis of DNR. MethodsA non-targeted lipidomic strategy using Waters Acquity I Class UPLC-Xevo G2-S QTOF UPLC-MS was employed to identify different lipids in the 24 h postoperative plasma of patients whose cognitive function was assessed 7–14 days after surgery. Principal component analysis, orthogonal partial least squares discriminant analysis, and receiver operating characteristic (ROC) curves were used to identify potential biomarkers for predicting DNR. Metabolic pathway analysis was performed using the online platform MetaboAnalyst 5.0. ResultsPre-operative hemoglobin was identified as an independent risk factor for DNR. The individual areas under the ROC curves (AUC) of the 5 representative lipids were all > 0.6, and the AUC of the combined predictor was 0.912. Glycerophospholipid metabolic pathway was dysregulated in patients with subsequent DNR. ConclusionThis study identified a series of potential predictive biomarkers for DNR. These data suggested that glycerophospholipid metabolism may be the mechanism underlying the occurrence and progression of DNR.