This study developed the topographic histological mapping of photoreceptor and ganglion cell densities on UWF FA images. This new method has made it possible to measure the number of pixel cells in correlation with metabolic activity across the retina. Retinal ischemia causes dysfunction and death of neuronal cells, including photoreceptors and ganglion cells, through glial cell activation and release of immunomodulatory cytokines19,20,21. Moreover, various retinal cells such as Müller cells, lymph nodes, endothelial cells and pericytes induce overexpression of proangiogenic factors under ischemic conditions.22. In this regard, it is hypothesized that a region containing more cells will lead to greater metabolic alteration and hence greater clinical impact under ischemic conditions.

The distribution of photoreceptors and ganglion cells varied with eccentricity. Approximately 50% of total cones are within 18° of the foveola and approximately 50% of total ganglion cells are within 13° of the foveola23. Additionally, the central foveola, within 1.25° (350 μm in diameter), is free of rods. Rod density increases fastest upwards and slowest nasally from the fovea, and highest at 20° from the foveal center17. Due to the much higher number of cones and rods than image pixels and its uneven distribution, the weighted ISI reflecting differences in regional metabolic activity could be lower than the unweighted ISI, regardless of the retinal area or cell type. In this study, the difference between unweighted and weighted ISI was greatest in the posterior region. This result suggests that the unweighted ISI may be overestimated in the posterior pole, which may lead to limitations in analyzing the association with clinical characteristics. Therefore, there could be a great advantage in weighted ISI, which differs from unweighted ISI in which the pixels of a two-dimensional image are evenly distributed.

In this study, the unweighted and weighted ISI of NPR with leak demonstrated a better correlation with clinical characteristics than NPR without leak. Similarly, Fang et al. reported that the grayish retinal region with leakage on UWF FA was positively correlated with CMT, while the entirely dark region without leakage was negatively correlated with CMT24. They suggested that the leaky grayish region might have more viable cells that could produce cytokines leading to ME. Thus, the results of this study can be explained by a better correlation between NPR and leakage, including cell dysfunction, rather than complete cell death.

Several studies have demonstrated the effects of ischemic insults on the retina2,3,11,12,13. Histologically, reduction in inner retinal thickness occurred before reduction in outer retinal thickness in rats, suggesting that photoreceptors are less sensitive11. However, most studies have focused on the impact of ischemia on the inner retina; some studies focused on photoreceptors11,12,13,25. Recent studies have demonstrated reductions in overall retinal thickness and loss of retinal ganglion cells, glial cells, and photoreceptors after induction of ischemia, suggesting degradation of all retinal layers11. Therefore, it seems reasonable to consider metabolic changes and impairment of ganglion cells and photoreceptors in a long-lasting ischemic state. Although several studies suggest that retinal ganglion cells are most susceptible to ischemic injury11,12,13,12, ganglion cell-weighted ISI showed no correlation with clinical characteristics in this study. These results could be related to the lower density of ganglion cells than cones and rods. The different cell distributions, densities, and energy requirements could explain why the rod-weighted ISI had more correlations with clinical outcomes. In addition, a rod-derived cone viability factor causes a metabolic interaction between rods and cones, suggesting that damage to rods causes extinction of intact cones.26.27.

No consistent association between NPR and DME has been identified. Wessel et al. found an association between the binary classification of peripheral ischemia and the presence of DME6. However, they did not find a correlation between degree of ischemia and CMT. On the other hand, Fan et al. reported that NPR in the mid-periphery was significantly correlated with CMT28. In this study, the presence of ME was not associated with the unweighted ISI of leaky NPR, while it had a positive association with the weighted ISI of leaky NPR. It is hypothesized that because the weighted ISI can reflect the cellular distribution containing metabolic activity, it showed better correlation. Previous studies demonstrated that NPR greater than 23% of total visible retina on UWF FA was more strongly correlated with posterior segment NV8,9,10. This study adds the finding that the presence of basal NV had stronger correlations with the photoreceptor-weighted (specifically rod-weighted) ISI of leaky NPR, while it was not correlated with the Unweighted ISI.

Various cytokines are believed to be the primary mechanism behind disease processes such as DME or NV. Previous studies have shown elevated levels of MCP-1, IL-8, and IL-6 in the aqueous humor of DR eyes, which may affect intercellular gap junctions and intracellular tight junctions, thereby increasing vascular permeability29,30,31,32. PlGF and VEGF are involved in angiogenesis and inflammatory process through enhancement of tissue factor production and chemotaxis33. In this study, the weighted ISI of NPR with leakage was correlated with the levels of MCP-1, IL-8, IL-6, PlGF and VEGF-A. However, the unweighted ISI of leaky NPR only correlated with IL-8 and IL-6 levels. These results suggest that increased cytokines and quantification accounting for metabolic changes may better reflect disease processes. The correlation between weighted ISI and cytokine level was maintained when cones or rods were weighted, but not when ganglion cells were weighted. Due to the low density of ganglion cells, regional metabolic changes in ganglion cells may not affect the results.

This study had certain limitations. First, the sample size was relatively small because the data was collected prospectively. Second, manual NPR segmentation tends to have variability among students. However, the kappa value between raters was 0.92 (range 0.89 to 0.94), with a high level of repeatability. Third, because this method only covered cone, rod, and ganglion cell density maps on UWF FA images, the results might not fully reflect metabolic activity. The weighted ISI might more closely reflect the metabolic relevance of a disease state by incorporating more retinal cells, such as retinal pigment epithelial cells or glial cells, that might show biological activity. Fourth, it is known that there is individual variation in photoreceptors depending on the retinal region. Specifically, the maximum density of foveal cones has been reported to be highly variable, although the total number of foveal cones is similar between individuals.17. Nevertheless, to our knowledge, this is the first study to analyze the clinical significance of weighted values ​​applied to each pixel by histological mapping in patients with DR, which could be supported by cytokine assay.

In conclusion, weighted ISI reflecting both metabolic activity and cellular distribution showed better correlation with clinical characteristics than unweighted ISI, which was classically used in previous studies. Specifically, the weighted ISI was more useful in NPRs with leakage containing more viable tissue than in NPRs without leakage. The presence of baseline ME or NV in patients with DR was associated with weighted ISI, with a stronger association when cones and rods were weighted. This quantification method reflecting metabolic activity at the cellular level will provide new insights into the pathogenesis and prediction of clinical characteristics in patients with DR.