The Role of Prognostic and Predictive Biomarkers for Assessing Cardiovascular Risk in Chronic Kidney Disease Patients

Table 1

Summary of the principal prognostic and predictive biomarkers of cardiovascular risk in chronic kidney disease patients.

Biomarkers	Characteristics	Prognostic value	Predictive value
Cystatin C	Protein produced by all nucleated cells mainly used as marker of kidney function	Cystatin C improves the estimation of eGFR and risk prediction of CV events; it also allows to reclassify patients into more accurate CV risk categories [52]	—
β2-Microglobulin	Component of MHC class I molecules and expressed on all nucleated cells in humans	Improves risk prediction in CKD patients beyond traditional risk factors [53]	—
hs-cTnT	Regulatory protein that is integral to cardiac and skeletal muscle contraction	Improves the risk prediction of CV events, particularly heart failure regardless of the level of kidney function [54–56]	—
NT-proBNP	Prohormone with a 76-amino acid N-terminal inactive protein	Improves the risk prediction of CV events, particularly heart failure regardless of the level of kidney function [54–56]	It has been used as predictive biomarker in the SONAR trial during the run-in phase, in order to exclude patients with sodium retention after treatment with atrasentan [83].
sST2	Member of the IL-1 receptor family, which is produced by cardiomyocytes and cardiac fibroblasts	It is delivered in response to mechanical stress conditions and showed incremental prediction ability (over NT-proBNP) for HF-related death and hospitalizations [58]	—
Galectin-3	30 kDa protein that contains a carbohydrate-recognition-binding domain that enables the linkage of β-galactosides	In patients with already established CV disease, galectin-3 is an independent predictor of hospitalizations and death due to CV causes [59, 60]	—
MMPs	Six families of zinc-containing endopeptidases that are involved in regulating tissue development and homeostasis	Serum MMP-2, MMP-8, MMP-9, and TIMP-1 are associated with atherogenesis, the severity of kidney damage, and the onset of left ventricular hypertrophy and peripheral vascular disease [61–66]	MMP levels are modified by selective and nonselective drugs. Changes in MMP levels have been associated with a reduction of CV risk [72, 73].
CAC	CAC is a score measured at cardiac TC based on the entity of calcium depositions on artery plaques.	Improves risk prediction in CKD patients beyond traditional risk factors [48, 68]	—
eGFRcrea	eGFRcrea is an estimation of the kidney function level based on serum creatinine, age, gender, and race.	A reduction of eGFR is a potent predictor of CV endpoints, regardless of age, gender, and other risk factors [1, 2, 5, 8, 22, 23]	Although a treatment-induced reduction of eGFR is considered a surrogate endpoint of ESKD, the predictive role of eGFR change for CV risk is still controversial [67].
Proteinuria	Presence of an abnormal quantity of proteins in urine; it is considered the principal marker of kidney damage.	The increase in proteinuria is strongly associated with the onset of fatal and nonfatal CV events [1, 2, 5, 8, 21, 22]	In clinical trials, patients who develop a significant reduction in proteinuria during the first months after treatment were protected against CV events over time [12–15, 69–71].
RI	Renal resistive index is a sonographic index of intrarenal arteries defined as .	Raised RI levels above have been shown to predict CV events in hypertensive and CKD patients [75, 76]	Medications as RAAS inhibitors and SGLT-2i reduce RI levels over time and improve vascular damage [77, 78].
ACE ID/DD	Insertion (I)/deletion (D) polymorphism of the angiotensin-converting enzyme (ACE) gene influences the circulating and renal activity of RAAS.	The D allele patients showed a poor CV prognosis in the RENAAL trial [79]	Patients with DD genotype, despite being at high risk of CV events, showed the better response to losartan in the RENAAL study [79].
Classifiers	A classifier is the combination of the informative markers which is able to classify patients according to their risk of developing an outcome or likelihood of response to a treatment.	—	A panel of 185 metabolites and a proteomic-based classifier have shown to predict the proteinuric response to RAAS inhibitors [81, 82].