Integrating kidney imaging for risk prediction, therapeutic monitoring, and prognostication across the kidney disease spectrum: a review of emerging evidence
TL;DR
Kidney imaging advances are transforming the field of nephrology by allowing non-invasive examination of renal structure, function, and pathology, and imaging biomarkers show significant predictive value for progression to kidney failure in CKD, including diabetic kidney disease, independent of traditional measurements.
Abstract
ABSTRACT Kidney imaging advances are transforming the field of nephrology by allowing non-invasive examination of renal structure, function, and pathology. Traditional measures such as albuminuria and estimated glomerular filtration rate (eGFR) are only partially effective in identifying early or regionally variable kidney damage. As an example, imaging identifies chronic kidney disease (CKD) in autosomal dominant polycystic kidney disease (ADPKD) decades earlier than eGFR or albuminuria, addressing the “blind spot” in CKD, and providing a criterion to start early therapy. Currently, quantitative imaging techniques such as advanced ultrasound (Doppler sonography, photoacoustic imaging [PAI], contrast-enhanced ultrasound [CEUS], and ultrasound-based elastography), multiparametric MRI, and some CT methods provide insights into fibrosis, lipid infiltration, oxygenation, and microvascular integrity. Imaging biomarkers, including cortical R2*, apparent diffusion coefficient (ADC), arterial spin labeling (ASL), and shear-wave elastography (SWE), show significant predictive value for progression to kidney failure in CKD, including diabetic kidney disease, independent of traditional measurements. Imaging detects fibrotic remodeling and microvascular impairment in hypertensive and obese kidney phenotypes before laboratory indicators appear. Emerging technologies such as AI-augmented radiomics, photoacoustic oxygen mapping, and H-scan ultrasound may facilitate the early detection of graft fibrosis and functional decline in kidney transplantation. Most methods, however, lack optimal clinical integration, standardized protocols, and comprehensive validation. Future directions may include longitudinal cohort studies, multi-omics data integration, imaging-informed clinical trial outcomes, and machine learning-based prognostic models. Imaging biomarkers have the potential to profoundly transform precision nephrology by providing organ- and tissue-level insights that enhance transplant surveillance, therapy monitoring, and personalized risk classification.