Yang et al. (1) established a histology-based early detection of diabetic retinopathy (DR), a complication of diabetes affecting vision. Before the clinical symptoms of DR become apparent, histological biomarkers could provide insights into the cellular and molecular changes occurring in the retina. The study was conducted on postmortem retinas from a limited sample size of 10 donors with diabetes, 3 control participants, and a single case subject with DR. In the group of people with diabetes, the examination of tissue sections and quantitative investigation of vasculature exposed capillary dropout in the deeper vascular plexuses and increased capillary diameter, along with the loss of inner nuclear layer cells. In early-stage DR, one school of thought is microvascular loss precedes neurodegeneration of retina. This study challenges this conventional view, indicating independent effects of diabetes on vascular and neural components.
Previous research has suggested retinal neurodegeneration may precede microvasculopathy in DR, with studies presenting that it is linked to diabetes duration rather than vascular capillary or pericyte loss (2). However, the study by Yang et al. (1) lacks a straightforward early DR diagnosis approach and has several significant shortcomings, including a small sample size, lack of longitudinal data, and absence of controls for confounding factors such as the duration of diabetes. Subclinical, progressive macular vasculature change is associated with longer diabetes duration (3). Another major pitfall of the study is the longer fixation delays (>20 h). This has affected the staining pattern or expression of specific markers (glutamine synthase [GS] and glial fibrillary acidic protein [GFAP] expression) and resulted in poor quality of vasculature in the image. Further, causes of death related to systemic conditions like cancer could skew their results and render conclusions unreliable, highlighting the need for further research on biological mechanisms and additional biomarkers.
Early detection of DR is increasingly reinforced by favorable artificial intelligence–based solutions. Artificial intelligence can help people with diabetes avoid vision defects by identifying high-risk populations and improving screening affordability, efficiency, and accuracy (4). However, the best patient outcomes depend on integrating automated risk stratification and clinical characteristics, and achieving this goal is still a long way off.
Another approach, we argue, is focusing on the genetic underpinnings of DR. It is becoming clear how common and rare genetic variations influence distinct DR symptoms and the response to anti-vascular endothelial growth factor therapy (5). Preliminary research indicates that new genetic variations related to inflammation and angiogenesis may protect against or speed up DR progression. Although genetic factors are implicated, their exact roles remain unclear. Further research using computational genetics–based methods is essential, as current findings are just the beginning.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
