Acute kidney injury: Detection, risk stratification, and predictive biomarkers.

BACKGROUND: Acute kidney injury (AKI) is a complex multifactorial syndrome characterized by a rapid decline in kidney function, frequently observed in hospitalized and critically ill patients. Despite its high morbidity and mortality, current diagnostic criteria-based primarily on serum creatinine and urine output-are delayed and non-specific, limiting early detection and timely intervention. CONTENT: Over the past two decades, extensive research has led to the discovery and validation of a diverse array of AKI biomarkers that reflect distinct pathophysiological mechanisms, including tubular cell injury (e.g., NGAL, KIM-1), inflammation (e.g., IL-18, CCL-2 and CCL14), oxidative stress (e.g., L-FABP), cell cycle arrest (e.g., TIMP-2·IGFBP7), and endothelial dysfunction (e.g., suPAR). Additional functional biomarkers, such as urinary cystatin C and low-molecular-weight proteins like α1-microglobulin and β2-microglobulin, have emerged as sensitive indicators of proximal tubular dysfunction. These biomarkers offer promise for detecting subclinical AKI, improving risk stratification, guiding therapy, and serving as surrogate endpoints in clinical trials. However, clinical implementation remains constrained by assay variability, lack of harmonized cut-offs, cost and reimbursement barriers, and limited validation in pediatric and other specialized populations. Newer platforms, such as urinary tubular enzymes (e.g., NAG, LDH, ALP) and extracellular vehicles (EVs), continue to expand the diagnostic landscape. SUMMARY: Integration of complementary multi-marker panels with artificial intelligence-based models offers significant potential to advance precision nephrology by enabling earlier risk prediction, individualized management, and optimized clinical trial enrichment. Ongoing international standardization initiatives and multicenter validation efforts are critical to accelerating the translation of AKI biomarkers from research tools to routine clinical practice.