CLINICAL PREDICTORS OF CHRONIC KIDNEY DISEASE PROGRESSION IN PATIENTS WITH LONG‑STANDING DIABETES

Abstract

Chronic kidney disease has been found to be a significant comorbidity in patients with long-term diabetes and is seen in around 35% of patients with type 2 diabetes and is a significant contributor to cardiovascular morbidity and mortality. Although the introduction of the use of estimated glomerular filtration rate and albuminuria as prognostic variables are helpful, the reality that the predictive value of the disease progression especially in the non-proteinuric phenotypes is low leaves the need to seek new predictors that can explain the intricate pathophysiology of diabetic kidney disease. The review is an overview of the evidence on the foundation of 147 studies that comprised 894,327 patients to establish the presence or absence of clinical predictors of acquiring chronic kidney disease in particular patients with established diabetes. The results show the stratification of base risk by the use of the traditional clinical predictors i.e. glycemic control, blood pressure and albuminuria with an area under the curve of between 0.712 to 0.745. However, a combination of a tubular injury biomarker (kidney injury molecule-1), an inflammatory biomarker (soluble tumor necrosis factor receptors), and a fibrotic biomarker are significantly better predictive than the old logistic regression and stacked ensemble methods are the best in terms of calibration Network meta-analysis supports the use of sodium-glucose cotransporter-2 inhibitor as the most effective intervention to reduce renal damage (relative risk 0.67 composite renal outcome) then glucagon-like peptide-1 receptor agonists and non-steroidal mineralocortin receptor antagonists. Multi-omics types of classes that contain CKD273 proteomic classesifier and integrated metabolomic-microRNA classes have been reported to be the most predictive classes (0.951 area under the curve) with high calibration validity in validation cohorts. This evidence indicates that a multi-domain model of clinical, glomerular, tubular, inflammatory, fibrotic and molecular biomarker is the most effective model of chronic kidney disease progression to prognosticate the disease and respond to therapeutic interventions in the entire continuum of disease pathophysiology.