#2177 It is possible to predict the clinical course of chronic kidney disease in patients with type 2 diabetes: the significance of selected biomarkers

Abstract Background and Aims Clinical practice indicates that the course of diabetic kidney disease (DKD) in patients with type 2 diabetes mellitus (T2DM) and the prognosis of this condition are highly heterogeneous, often difficult to predict, and at the same time crucial for diagnostic and therapeutic decision-making. Therefore, identifying patients at high risk of developing and progressing DKD is of critical importance. The aim of this retrospective observational study was to evaluate the utility of selected biomarkers correlated with pathophysiology in assessing the progression and phenotype of DKD. Method The study recruited 97 (32♂, 49♀) patients with T2DM at an average age of 67 ± 11.5 years from nephrology outpatient clinics and a dialysis center. The patients were assigned by a nephrologist to DKD phenotypic groups based on clinical history, eGFR values (ml/min/1.73 m²), and uACR (mg/g). The groups were defined as follows: The division into DKD phenotypes is illustrated in Fig. 1. Analyses of biomarkers associated with DKD pathogenesis were performed using urine and blood samples from these patients. The Partial Least Squares Structural Equation Modeling (PLS-SEM) method was used to evaluate complex relationships among independent variables (age, sex, biomarkers), mediating variables (angiotensin-converting enzyme concentration - ACE, atherogenic index of plasma - AIP), and the dependent variable – estimated glomerular filtration rate (eGFR). Discriminant analysis was performed to differentiate three groups (F1, F2, F3) using four variables (log API, age, serum ACE protein, and uric acid levels), achieving 86.2% correct classifications with statistically significant results (Wilks' Lambda, P < 0.05). Results We found that 61.9% of the variance in eGFR was explained by regression on ACE protein levels, age, and the AIP as well as hsa-miR-423-5p (positively correlated). Figure 2 shows the final PLS model with all significant paths. The model fit accuracy was very good (SRMR = 0.024). All specific indirect effects of age, BMI, via AIP and ACE, male sex and mir-126.3p were significant (P < 0.05). All in all, eGFR was associated (in descending order of importance) with age, ACE, AIP, male sex, mir-126-3p and mir-423-5p. Discriminant analysis was conducted to differentiate between phenotypic groups. The classification results were visualized in the Fig. 3. Finally, based on the conducted analysis of clinical and laboratory data, the DKD phenotypes were characterized according to the following statistically significant features: Conclusion We identified new modifiable risk factors for the adverse progression of DKD, such as serum ACE, uric acid concentration, AIP, hsa-miR-423-5p and hsa-miR-126-3p. These, when correlated with clinical and epidemiological data, appear to be useful for determining the phenotype of CKD and assessing prognosis in the course of T2DM.