Abstract
Acute kidney injury (AKI) is a frequent complication of sepsis, major surgery or nephrotoxic medication use. It is associated with high morbidity and mortality. In an effort to identify novel biomarkers capable of predicting the development of AKI after cardiac surgery with cardiopulmonary bypass use, urine specimens were collected before and at 4 and 24 h after surgery from 106 patients and ...
Abstract
Acute kidney injury (AKI) is a frequent complication of sepsis, major surgery or nephrotoxic medication use. It is associated with high morbidity and mortality. In an effort to identify novel biomarkers capable of predicting the development of AKI after cardiac surgery with cardiopulmonary bypass use, urine specimens were collected before and at 4 and 24 h after surgery from 106 patients and analyzed by means of nuclear magnetic resonance spectroscopy. Postoperative AKI of stage 1–3 as defined by the Acute Kidney Injury Network (AKIN) developed in 34 cases. Employing Quantile Normalization and support vector machine based classification, spectra of the 24-hour postoperative urine specimens were found to predict AKI across all stages with an average accuracy of 76.0 % and a corresponding area under the receiver operating characteristic curve of 0.83. Considering only AKIN-stage 2 and 3 patients, prediction accuracy increased to 81.7 % and 100 %, respectively. Among the small set of predictive biomarkers identified was carnitine, the urinary concentration of which was elevated significantly in AKI-free patients only, and tranexamic acid, which is routinely applied as an antifibrinolytic agent at the end of surgery, and whose renal excretion was delayed in AKI patients. The study underscores the power of NMR and bioinformatics in identifying novel biomarkers of disease and in gaining new insights into pathomechanisms.
Altmetric