Journal of Ophthalmology (Aug 2024)

Prognostic biomarkers of non-proliferative diabetic retinopathy progression in type 2 diabetes mellitus

  • S. Iu. Mogilevskii,
  • A. V. Serdiuk,
  • S. V. Zyablitsev

DOI
https://doi.org/10.31288/oftalmolzh202443845
Journal volume & issue
no. 4
pp. 38 – 45

Abstract

Read online

Purpose: To establish prognostic biomarkers of non-prolefarative diabetic retinopathy (NPDR) progression in T2DM on the basis of examination of clinical, ophthalmological and laboratory (carbohydrate and lipid metabolism and coagulation and hemostasis) characteristics. Methods: This study included 358 T2DM2 patients with diabetic retinopathy (DR); for each patient, the eye with more severe DR was included in the study. Cases were dichotomized into three groups based on DR severity: group 1 (NPDR; 189 eyes), group 2 (preproliferative DR or PPDR; 96 eyes), and group 3 (proliferative DR or PDR; 73 eyes). Central retinal thickness (CRT; µm) and central retinal volume (CRV) were assessed by optical coherence tomography (OCT; mm3). Serum samples were taken to assess fasting serum glucose, glycosylated hemoglobin (HbA1c), total cholesterol, high-density lipoprotein (HDL), low-density lipoprotein (LDL), very low-density lipoprotein (VLDL), triglycerides, and fibrinogen. Coagulation and hemostasis parameters were assessed. Statistical analyses were performed with EZR version 1.54. Results: Associations of the variables with the risk of NPDR progression were confirmed by a univariate logistic regression analysis. The area under curve (AUC) was largest for HbA1c (AUC = 0.84), followed by CRT and CRV (AUC = 0.79). In addition, the AUC ranged from 0.56 to 0.67 for the rest of variables. Age, gender, diabetes duration, and blood HDL level were not associated with NPDR progression. Our multivatiate logistic regression analysis found six biomarkers (age, HbA1c, LDL, VLDL, prothrombin index and CRT) directly associated with the risk of NPDR progression (AUC = 0.91; p < 0.001). The established association was mostly due to the three independent variables, HbA1c, VLDL and CRT (AUC = 0.90; p < 0.001). Conclusion: Independent effects on the risk of NPDR progression were confirmed for HbA1c, VLDL and CRT. The built model was found to be of very high predictive quality, which allowed recommending it for confirming high risk for NPDR progression in equivocal clinical cases or as a criterion for conclusive prognosis in the relevant expert systems.

Keywords