Metabolomics for contrast-induced nephropathy risk prediction?

Abstract

Contrast-induced nephropathy (CIN) is a serious and sometimes life-threatening complication since it is associated with increased incidence of adverse outcome and even mortality [1–3]. Pre-existing chronic kidney disease, diabetes mellitus, hemodynamic instability, as well as high osmolality agents, high volume of contrast medium, and intraarterial iodine administration are well-known risk factors for CIN [3]. Although some models have been constructed for CIN risk stratification, these scores include variables that may be unknown before the procedure, such as intercurrent hemodynamic instability and the volume of contrast media which is needed for imaging [3]. There is, therefore, a growing interest in achieving reliable scores for CIN risk determination, especially in patients with pre-existing predisposing factors or overt disease. The potential of quantitative measurement of metabolites in biological samples by nuclear magnetic resonance (NMR) spectroscopy or mass spectrometry-based methods, known as “metabolomics” [4], is considered a useful tool for risk prediction in clinical practice [5]. Besides the diagnostic and prognostic role, this technique allows the clinician to identify relevant risk conditions through recognition of a specific “molecular fingerprint” and eventually to prevent the occurrence of disease using a non-invasive examination. Recently, an in-depth review reported available evidence on metabolomics in renal disease, also exploring the therapeutic potential of metabolites analysis [6]. In an animal model of ischemic acute kidney injury (AKI), for example, a decrease in kidney tissue niacinamide (NAM) was identified as a key metabolite factor possibly associated with kidney damage in mice, and a reno-protective action was indeed shown after NAM administration [7]. Metabolomics also proved to be useful in the setting of renal transplantation, where analysis of changes in urine metabolites was adopted to study the effect of calcineurin inhibitors [8] and to predict the occurrence of acute cellular rejection [9]. It is, therefore, intriguing to ascertain whether “metabolomics” systematic dosage adoption could become a powerful tool to assess and quantify the risk of CIN, with the aim of preventing and possibly even avoiding this common and harmful complication. In this issue of “Internal and Emergency Medicine”, Dalili et al. [10] identified a pattern of urine and serum metabolites which was effective in predicting the occurrence of CIN in a cohort of patients with normal renal function undergoing elective coronary angiography for coronary artery disease (CAD). CIN was defined as an acute increase in serum creatinine greater than 0.5 mg/dl within 48–72 h from contrast medium administration. Urine and blood samples were collected right before the procedure and after 72 h, and were analysed using NMR spectroscopy to identify predictive biomarkers of kidney damage: great accuracy in CIN prediction was reached by a panel of six urinary metabolites, namely taurodeoxycholic acid (TDCA), 3-methylhistidine, tyrosine, phenylalanine, argininosuccinic acid, and epi-coprostanol. The evaluation of these molecules altogether led to a 100% sensitivity and an 83% specificity, representing a useful and optimal tool to assess the risk of subsequent CIN. Moreover, the following serum biomarkers were also detected: glutamic acid, uridine diphosphate (UDP), glutamine and tyrosine decreasing in serum were correlated with subsequent renal dysfunction. Furthermore, this study emphasizes the role of altered amino acids metabolism as a hallmark of susceptibility to kidney damage, especially by identifying a dysregulation in phenylalanine metabolic pathway in patients prone to develop CIN after contrast media administration. Strength of this study is the use of absolute changes in serum creatinine to define CIN, which were reportedly more accurate for CIN prediction than percent changes as shown in a cohort of thousand patients who underwent coronary angiography and in whom 6-month mortality rate * Ernesto Paoletti [email protected]