ABSTRACT
Acute Kidney Injury (AKI) affects up to 30% of the patients who undergo cardiac surgery (CVS) and is related to higher mortality. We aim to investigate molecular features associated with in-hospital AKI development and determine the predictive value of these features when analyzed preoperatively. This is a case-control study. From an initial cohort of 110 recruited subjects, a total of 60 patients undergoing cardiac surgery were included: 20 (33%) developed in-hospital AKI (CVS-AKI) and 40 did not (controls, CVS-C). Pre- and post-surgery samples were collected and a prospective study was carried out. A total of 312 serum samples and 258 urine samples were analyzed by nuclear magnetic resonance, mass spectrometry and ELISA. Six features predicted AKI development in pre-surgery samples: urinary kidney functional loss marker kidney injury molecule-1 (uKIM-1), 2-hydroxybutyric acid, 2-hydroxyphenylacetic acid, hippuric acid, phosphoethanolamine and spermidine. Two of them stood out as powerful predictors. Pre-surgery uKIM-1 levels were increased in CVS-AKI vs. CVS-C (AUC = 0.721, p-value = 0.0392) and associated strongly with the outcome (OR = 5.333, p-value = 0.0264). Spermidine showed higher concentration in CVS-AKI (p-value < 0.0001, AUC = 0.970) and had a strong association with the outcome (OR = 69.75, p-value < 0.0001). uKIM-1 and particularly spermidine predict in-hospital AKI associated with CVS in preoperative samples. These findings may aid in preventing postoperative AKI and improve prognosis of CVS.
ABSTRACT
Acute kidney injury (AKI) is one of the main complications in acute care medicine and a risk factor for chronic kidney disease (CKD). AKI incidence has increased; however, its diagnosis has limitations and physiopathological mechanisms are underexplored. We investigated urine samples, aiming to identify major metabolite changes during human AKI evolution. Metabolic signatures found were further explored for a potential link to severity of injury. Twenty-four control subjects and 38 hospitalized patients with AKI were recruited and urine samples were collected at the time of diagnosis, during follow-up and at discharge. Nuclear magnetic resonance (NMR) was used in a first discovery phase for identifying potential metabolic differences. Target metabolites of interest were confirmed by liquid chromatography-mass spectrometry (LC-MS/MS) in an independent group. Underlying metabolic defects were further explored by kidney transcriptomics of murine toxic AKI. Urinary 2-hydroxybutyric acid, pantothenic acid, and hippuric acid were significantly downregulated and urinary N-acetylneuraminic acid, phosphoethanolamine, and serine were upregulated during AKI. Hippuric acid, phosphoethanolamine, and serine showed further downregulation/upregulation depending on the metabolite in acute tubular necrosis (ATN) AKI compared to prerenal AKI. Kidney transcriptomics disclosed decreased expression of cystathionase, cystathionine-ß-synthase, and ethanolamine-phosphate cytidylyltransferase, and increased N-acetylneuraminate synthase as the potentially underlying cause of changes in urinary metabolites. A urinary metabolite panel identified AKI patients and provided insight into intrarenal events. A urine fingerprint made up of six metabolites may be related to pathophysiological changes in oxidative stress, energy generation, and H2S availability associated with AKI. KEY MESSAGES: The urinary metabolome reflects AKI evolution and severity of injury. Kidney transcriptomics revealed enzymatic expression changes. Enzymatic expression changes may be the potentially underlying cause of changes in urine metabolites. Identified metabolite changes link oxidative stress, energy generation, and H2S availability to AKI.
Subject(s)
Acute Kidney Injury/metabolism , Acute Kidney Injury/urine , Energy Metabolism , Hydrogen Sulfide/metabolism , Metabolomics , Oxidative Stress , Acute Kidney Injury/genetics , Acute Kidney Injury/pathology , Adult , Aged , Aged, 80 and over , Female , Gene Expression Regulation , Humans , Linear Models , Male , Middle Aged , ROC Curve , Severity of Illness Index , Transcriptome/geneticsABSTRACT
Implementation of therapy for acute kidney injury (AKI) depends on successful prediction of individual patient prognosis. Clinical markers as serum creatinine (sCr) have limitations in sensitivity and early response. The aim of the study was to identify novel molecules in urine which show altered levels in response to AKI and investigate their value as predictors of recovery. Changes in the urinary proteome were here investigated in a cohort of 88 subjects (55 AKI patients and 33 healthy donors) grouped in discovery and validation independent cohorts. Patients' urine was collected at three time points: within the first 48 h after diagnosis(T1), at 7 days of follow-up(T2) and at discharge of Nephrology(T3). Differential gel electrophoresis was performed and data were confirmed by Western blot (WB), liquid chromatography/mass spectrometry (LC-MS/MS) and enzyme-linked immunosorbent assay (ELISA). Retinol binding protein 4 (RBP4) and kininogen-1 (KNG1) were found significantly altered following AKI. RBP4 increased at T1, and progressively decreased towards normalization. Maintained decrease was observed for KNG1 from T1. Individual patient response along time revealed RBP4 responds to recovery earlier than sCr. In conclusion, KNG1 and RBP4 respond to AKI. By monitoring RBP4, patient's recovery can be anticipated pointing to a role of RBP4 in prognosis evaluation.