Proposed international clinical diabetic retinopathy and diabetic macular edema disease severity scales

Diabetic Retinopathy Assessment through Multitask Learning Approach on Heterogeneous Fundus Image Datasets

Objective

To develop and validate an artificial intelligence (AI)-based system, Diabetic Retinopathy Analysis Model Assistant (DRAMA), for diagnosing diabetic retinopathy (DR) across multisource heterogeneous datasets and aimed at improving the diagnostic accuracy and efficiency.

Design

This was a cross-sectional study conducted at Zhejiang University Eye Hospital and approved by the ethics committee.

Subjects

The study included 1500 retinal images from 957 participants aged 18 to 83 years. The dataset was divided into 3 subdatasets: color fundus photography, ultra-widefield imaging, and portable fundus camera. Images were annotated by 3 experienced ophthalmologists.

Methods

The AI system was built using EfficientNet-B2, pretrained on the ImageNet dataset. It performed 11 multilabel tasks, including image type identification, quality assessment, lesion detection, and diabetic macular edema (DME) detection. The model used LabelSmoothingCrossEntropy and AdamP optimizer to enhance robustness and convergence. The system's performance was evaluated using metrics such as accuracy, sensitivity, specificity, and area under the curve (AUC). External validation was conducted using datasets from different clinical centers.

Main Outcome Measures

The primary outcomes measured were the accuracy, sensitivity, specificity, and AUC of the AI system in diagnosing DR.

Results

After excluding 218 poor-quality images, DRAMA demonstrated high diagnostic accuracy, with EfficientNet-B2 achieving 87.02% accuracy in quality assessment and 91.60% accuracy in lesion detection. Area under the curves were >0.95 for most tasks, with 0.93 for grading and DME detection. External validation showed slightly lower accuracy in some tasks but outperformed in identifying hemorrhages and DME. Diabetic Retinopathy Analysis Model Assistant diagnosed the entire test set in 86 ms, significantly faster than the 90 to 100 minutes required by humans.

Conclusions

Diabetic Retinopathy Analysis Model Assistant, an AI-based multitask model, showed high potential for clinical integration, significantly improving the diagnostic efficiency and accuracy, particularly in resource-limited settings.

Financial Disclosures

The author(s) have no proprietary or commercial interest in any materials discussed in this article.

Vision-Threatening Diabetic Macular Ischemia Based on Inferred Progression Pathways in OCT Angiography

Purpose

To elucidate the progression pathways of diabetic macular ischemia (DMI) using OCT angiography (OCTA) images and to assess changes in visual acuity (VA) associated with each pathway.

Design

A single-center, prospective case series study.

Participants

One hundred fifty-one eyes from 151 patients with a 3-year follow-up period.

Methods

We obtained 3 × 3 mm swept-source OCTA images and conducted analyses of en face images within a central 2.5 mm diameter circle. Nonperfusion squares (NPSs) were defined as 15 × 15-pixel squares without retinal vessels. Each eye at baseline and after 3 years was embedded into a 2-dimensional uniform manifold approximation and projection space and assigned to 1 of 5 severity grades-Initial, Mild, Superficial, Moderate, and Severe-using the k-nearest neighbors method. We assessed major transitions (involving ≥4 cases) during 3 years. Subsequent probabilistic analyses enabled the construction of a graphical model, wherein directed arrows represented inferred pathways of DMI progression. From this cohort, 103 eyes of 103 patients who did not receive any ocular treatments during the follow-up period were subsequently evaluated for VA changes.

Main Outcome Measures

Inference of DMI progression pathways.

Results

In most cases, NPS counts increased in both the superficial and deep layers. The major transitions between these severity groups at 3 years displayed a unique distribution, and probabilistic analyses suggested a directed graphical model comprising 7 inferred pathways of DMI progression: Initial to Mild, Initial to Superficial, Mild to Superficial, Mild to Moderate, Superficial to Moderate, Superficial to Severe, and Moderate to Severe. Eyes of the Mild and Superficial groups had greater increases in superficial NPS within the central sector than those of the Severe group. Additionally, deep NPS counts within the central sector decreased more in the eyes of the Initial group than in those of the Superficial and Moderate groups. Notably, the eyes of the Superficial and Moderate groups exhibited greater VA deterioration at 3 years compared with those in the Initial group.

Conclusions

A directed graphical model of DMI progression may serve as a useful tool for inferring progression pathways and predicting VA deterioration.

Financial Disclosures

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Retinal vessel tortuosity and fractal dimension in diabetic retinopathy

BACKGROUND

Retinal vessel geometry characteristic have been studied as one of the signs of microvascular changes in diabetic retinopathy (DR) that necessitates early screening. This study aimed to investigate the differences in retinal vessel tortuosity (VT) and fractal dimension (FD) between patients with and without DR.

METHODS

This retrospective study analyzed medical records and OCT-A images of DR and No-DR patients. DR severity was graded by a vitreoretinal specialist following the International Clinical Diabetic Retinopathy and Diabetic Macular Edema Severity Scales. Retinal VT and FD were quantified using ImageJ software. Comparison between groups using non-parametric and Generalized Estimating Equations (GEE) statistical analysis combined with cluster bootstrapping.

RESULTS

We analyzed 96 (161 eyes) with the mean age of 52.7 ± 9.9 years. Compared to No-DR, VT was significantly higher in all DR groups (p < 0.05). Mild non proliferative DR (β = +0.0621), Moderate NPDR (β = +0.0412), Severe NPDR (β = +0.0441), and proliferative DR (β = +0.0404). FD of the superficial capillary plexus (SCP) showed no significant difference among the groups and a significantly lower FD of the deep capillary plexus (DCP) compared to the No-DR groups (moderate NPDR (β = -0.0131), severe NPDR ( β = -0.0316) and PDR ( β = -0.0326)).

CONCLUSION

Compared to No-DR group, VT was found significantly higher in DR group, and FD of the DCP found significantly lower in the DR group. These parameters offer unique insights beyond simple vessel loss and complementary information into the geometric complexity and structural alterations of the retinal microvasculature in DR.

Posterior Retinal Ischemia Correlates With Vision in Patients With Diabetes

Purpose

The purpose of this study was to investigate whether the topographic location of retinal non-perfusion influences visual function, specifically low luminance visual acuity (LLVA) and best-corrected visual acuity (BCVA), in individuals with diabetes.

Methods

In this cross-sectional observational study, individuals with diabetes across the spectrum of diabetic retinopathy (DR) severity were enrolled. LLVA and BCVA were measured according to the Early Treatment Diabetic Retinopathy Study (ETDRS) protocol, with a 2.0-log unit neutral density filter for LLVA. Retinal ischemia was evaluated using ultra-widefield fluorescein angiography (UWF-FA) to manually quantify non-perfusion within (posterior ischemia) and outside (peripheral ischemia) the ETDRS seven fields. Macular ischemia was assessed by optical coherence tomography angiography (OCTA) using geometric perfusion deficits (GPDs) in both the superficial and deep capillary plexus (DCP). Associations between visual acuity and various explanatory variables, focusing on retinal ischemic parameters were assessed with linear mixed models.

Results

A total of 181 eyes from 126 patients without diabetic macular edema were analyzed. Increasing DR severity reduced both BCVA and LLVA. After adjusting other explanatory variables, age and posterior ischemia (estimate = -0.46, P = 0.046) were significant for LLVA. In contrast, age, sex, posterior ischemia (estimate = -0.50, P = 0.009), and GPD-DCP (estimate = -0.25, P = 0.049) were statistically significant for BCVA.

Conclusions

Retinal ischemia's topographic location differentially affects visual function in diabetes. Posterior ischemia predominantly impacts LLVA, whereas both macular and posterior ischemia contribute to BCVA decline. These results highlight the importance of assessing retinal ischemia beyond the macula to better understand visual function deficits in patients with diabetes.

Blood omega-3 biomarkers, diabetic retinopathy and retinal vessel status in patients with type 1 diabetes

BACKGROUND/OBJECTIVES

Clinical research on dietary omega-3 fatty acids and retinal health in type 1 diabetes (T1D) is scarce. In patients with T1D, we examined the associations between blood biomarkers of marine omega-3 (which reflect their dietary intake) and prevalent diabetic retinopathy (DR), and retinal microvascular data obtained through optical coherence tomography angiography (OCTA).

SUBJECTS/METHODS

Exploratory, cross-sectional sub-study of a prospective, consecutive, large-scale OCTA study conducted in a longitudinal cohort (ClinicalTrials.gov NCT03422965). We used baseline data from 188 consecutive patients with T1D and 88 controls. We determined blood omega-3 biomarkers (eicosapentaenoic [EPA], docosapentaenoic [DPA] and docosahexaenoic [DHA] acids) by gas-chromatography. Ocular data included DR grading, and 6 × 6 mm OCTA scans to obtain macular vessel density and perfusion density, and foveal avascular zone area, perimeter, and circularity.

RESULTS

Patients with T1D, regardless of DR stage, showed significantly lower blood levels of EPA, DHA, DHA, and EPA + DHA than non-diabetic controls (P < 0.001, all cases). In multivariate models in patients with T1D, higher EPA was associated with a lower prevalence of DR (P = 0.044); and increasing proportions of DPA, DHA, EPA + DHA, and total marine omega-3 fatty acids related to a higher vessel and perfusion densities in the macula (P values from 0.014 to 0.050).

CONCLUSIONS

In patients with T1D, higher blood omega-3 status related to lower DR grades and preserved retinal perfusion. Our results, which are consistent with the current model of the pathogenesis of DR and data from experimental models, add to the notion of marine-derived omega-3 fatty acids as a healthy fat.

miR-124, miR-126-3p, and miR-200b: Potential therapeutic targets for VEGF-mediated complications in proliferative diabetic retinopathy

INTRODUCTION

This study aimed to investigate alterations in intravitreal microRNA and vascular endothelial growth factor (VEGF) levels in patients with proliferative diabetic retinopathy (PDR) as these factors are implicated in PDR pathogenesis.

METHODS

Fifty-two participants, including 26 patients with PDR and 26 controls without diabetes, were included in this study. VEGF levels were assessed using ELISA, and seven microRNAs (miRNAs) (miR-19a, miR-20b, miR-27a, miR-124, miR-126-3p, miR-146a, and miR-200b) were analyzed using quantitative real-time PCR.

RESULTS

PDR patients exhibited significantly higher miR-124 and miR-126-3p levels in the vitreous material compared to controls ( P < 0.05). Conversely, miR-200b levels were significantly lower in the PDR group ( P < 0.05). VEGF-A levels were markedly elevated in PDR patients compared with controls ( P < 0.05). A nonsignificant positive correlation was found between miR-124 and miR-126-3p levels and VEGF levels (r = 0.361, P = 0.076 and r = 0.168, P = 0.422, respectively), whereas a nonsignificant negative correlation was observed between miR-200b and VEGF levels (r = -0.145, P = 0.488).

CONCLUSION

Our study demonstrated a significant upregulation of miR-124 and miR-126-3p, along with a downregulation of miR-200b, in vitreous samples from patients with PDR, accompanied by elevated VEGF-A levels. These findings provide valuable insights into the pathogenesis of PDR. Further research is needed to evaluate the potential diagnostic and therapeutic implications of these molecular changes and to explore their viability as potential therapeutic targets.

Macular perfusion alterations in people with recent-onset diabetes and novel diabetes subtypes

AIMS/HYPOTHESIS

Our aim was to detect early structural and functional changes in the macular capillaries using optical coherence tomography angiography during the course of type 1 or 2 diabetes mellitus.

METHODS

In this cross-sectional study, individuals with type 1 diabetes (n=143) or type 2 diabetes (n=197) from the German Diabetes Study (ClinicalTrials.gov registration no. NCT01055093) underwent clinical examination and cluster analysis to identify phenotype-based diabetes subtypes, using BMI, age, HbA1c, homoeostasis model estimates and islet autoantibodies. Colour fundus photography, optical coherence tomography and optical coherence tomography angiography were performed within the first year of diabetes diagnosis (baseline) and at 5 year intervals up to year 10. Age- and sex-adjusted participants served as control participants (n=105). Perfusion density, vessel density, presence of retinal microaneurysms in superficial, intermediate and deep capillary plexus (SCP, ICP, DCP), choriocapillaris flow deficit density (CC FD) and the foveal avascular zone (FAZ) of the macula as well as retinal layer thickness, visual acuity and contrast sensitivity were analysed.

RESULTS

Perfusion density and vessel density of SCP were already reduced at baseline in type 2 diabetes (expected difference compared with control participants: -0.0071, p=0.0276, expected difference: -0.0034, p=0.0184, respectively), especially in participants with severe insulin-deficient and mild obesity-related diabetes. At year 10 only perfusion density of the SCP and DCP was reduced in both type 1 and 2 diabetes (p=0.0365, p=0.0062, respectively). The FAZ was enlarged and the CC FD within the first year increased in type 1 (p=0.0327, p=0.0474, respectively) and more markedly in type 2 diabetes (p=0.0006, p<0.0001). The occurrence of microaneurysms in SCP and DCP was significant at year 5 (p=0.0209, p=0.0279, respectively) and year 10 (p=0.0220, p=0.0007). Presence of microaneurysms in SCP and DCP was associated with decreases in perfusion density and vessel density in both SCP and ICP. Furthermore, microaneurysms were associated with decreased ganglion cell layer and inner plexiform layer thickness.

CONCLUSIONS/INTERPRETATION

Type 2 diabetes already reduces macular perfusion SCP at time of clinical diagnosis, while long-standing diabetes affects both SCP and DCP. The FAZ of the SCP and the CC FD are early indicators of diabetic alterations, with more pronounced changes observed in type 2 diabetes. Microaneurysms in the macular plexus are associated with a decrease of ganglion cell layer and inner plexiform layer. Subclinical microangiopathy occurs prior to manifestation of diabetic retinopathy, disease-related visual acuity impairment or inner retinal layer thinning.

Evaluating the efficacy of AI systems in diabetic retinopathy detection: A comparative analysis of Mona DR and IDx-DR

PURPOSE

To compare two artificial intelligence (AI)-based Automated Diabetic Retinopathy Image Assessment (ARIA) softwares in terms of concordance with specialist human graders and referable diabetic retinopathy (DR) diagnostic capacity.

METHODS

Retrospective comparative study including 750 consecutive diabetes mellitus patients imaged for non-mydriatic fundus photographs. For each patient four images (45 degrees field of view) were captured, centered on the optic disc and macula. Images were manually graded for severity of DR as no DR, any DR (mild non-proliferative diabetic retinopathy [NPDR] or more), referable DR (RDR (more than mild DR)), or sight-threatening DR (severe NPDR or more severe disease and/or clinically significant diabetic macular edema [CSDME]). IDx-DR and MONA DR output was compared with manual grading and with each other.

RESULTS

Total sample size was 750 patients, of which 55 were excluded due to ungradable images. Out of the remaining 695 patients 522 (75%) were considered as having no DR by manual consensus grading, and 106 (15%) as having RDR. Agreement between raters varied between moderate to substantial. IDx-DR showed moderate agreement with human grading (k = 0.4285) while MONA DR had substantial agreement (k = 0.6797). Out of 106 patients with a ground truth of RDR, IDx-DR identified 105 and MONA DR identified 99. The sensitivity and specificity rates for RDR detection of IDx-DR were 99.1 and 71.5% compared with MONA DR of 93.4 and 89.3% respectively. Of note, both ARIAs had 100% sensitivity for the detection of STDR.

CONCLUSION

Both ARIAs performed well in this study population, both with sensitivity for RDR screening over 90%, with IDx-DR showing higher sensitivity and MONA DR higher specificity. MONA DR showed superior agreement with human certified graders.

Flow and ischemic changes in retina and choroid across diabetic retinopathy spectrum: a SS-OCTA study

PURPOSE

To examine changes in retinal and choroidal vasculature in diabetes mellitus across the range of diabetic retinopathy (DR) severities using optical coherence tomography angiography (OCTA) and compare the patterns of vascular changes.

METHODS

We conducted a cross-sectional study enrolling 296 patients (498 eyes) with diabetes mellitus. Swept-Source OCT Angiography variables in both retina and choroid, including perfusion density (PD), vessel density (VD), large vessel density (LVD) in both superficial and deep layer of retina and CC flow voids (FD) density of the choroid were quantified. Correlations between OCTA parameters and DR severity, visual acuity and studied factors were performed.

RESULTS

Totally 498 eyes including 176 had no DR, 160 had mild NPDR, 98 had moderate NPDR, 11 had severe NPDR, 41 had PDR with PRP, and 12 had PDR without PRP. Choriocapillaris (CC) flow voids density increased with increasing DR severity (17.06% vs 17.41% vs 17.60% vs 17.62% vs 18.05% vs 18.41%, p-trend = 0.0004), FAZ area increased with DR severity in both superficial and deep layer (superficial layer p trend=0.0027; deep layer p trend=0.0022). Visual acuity correlated negatively with CC flow voids (Pearson's ρ = 0.09, p = 0.04) and superficial FAZ area (Pearson's ρ = 0.22, p < 0.001), while inversely correlated with SCP PD (Pearson's ρ = -0.15, p < 0.001) and VD (Pearson's ρ = -0.15, p < 0.001), as well as DCP PD (Pearson's ρ = -0.21, p < 0.001) and VD (Pearson's ρ = -0.19, p < 0.001).

CONCLUSION

Choriocapillaris ischemia increased, FAZ area enlarged, and total retina perfusion density decreased with increasing DR severity. The deep layer and large vessels may change in early stage before DR progresses to PDR. More ischemia and vessel tortuosity are correlated with worse visual acuity and higher HbA1c level. OCTA can be utilized to detect both large and small vascular changes in both the retina and choroid in DR patients.

Brief report: harmonic analysis of the 30 Hz flicker ERG in early-stage diabetic retinopathy

PURPOSE

To determine if harmonic components of the 30 Hz flicker ERG are useful for detecting neural dysfunction in diabetics who have mild or no non-proliferative diabetic retinopathy (NPDR).

METHODS

Previously reported light-adapted flicker ERG data recorded from 20 diabetics who had no clinically-apparent retinopathy (NDR), 20 who had mild NPDR (MDR), and 20 non-diabetic controls were reanalyzed. From this dataset, the amplitude and phase of the 31.25 Hz flicker ERG fundamental and second harmonic were extracted. The 62.5 Hz flicker ERG fundamental was also extracted. Similar responses were also acquired prospectively from 10 controls, 5 NDR, and 5 MDR subjects, comprising a second dataset.

RESULTS

Analysis of variance indicated that both diabetic groups had normal amplitudes elicited by the 31.25 Hz stimulus (fundamental and second harmonic), whereas the 62.5 Hz amplitude was reduced significantly in both diabetic groups. This pattern was found in both the retrospective and prospective analyses.

CONCLUSIONS

The second harmonic of the 31.25 Hz flicker response (equivalent to 62.5 Hz) was normal in early-stage DR, whereas the response to 62.5 Hz flicker stimuli was abnormal. The second harmonic of the ISCEV standard 30 Hz flicker ERG does not appear to be a useful indicator of neural dysfunction in early DR.

Association between hemodialysis and surgical outcomes of pars plana vitrectomy in patients with chronic renal failure and proliferative diabetic retinopathy: a retrospective analysis

BACKGROUND

To investigate the association between Hemodialysis (HD) and postoperative outcomes in patients diagnosed with chronic renal failure (CRF) who underwent pars plana vitrectomy (PPV) for proliferative diabetic retinopathy (PDR).

METHODS

This was a retrospective study. Fifty-three eyes were divided into HD (25 eyes) and non-HD (28 eyes) groups. Follow-up assessments were performed preoperatively and at 1 week, 2 weeks, 1 month, 2 months, 4 months, 8 months and 12 months postoperatively. Independent risk factors affecting visual recovery following PPV were also determined.

RESULTS

The HD group exhibited significantly lower intraocular pressure (IOP) from 1 week to 2 months postoperatively compared to the non-HD group (P < 0.05). While non-HD patients had better preoperative best-corrected visual acuity (BCVA), HD patients achieved superior BCVA at 2 months (P < 0.05), with higher rates of visual improvement at 12 months (84.00% vs. 78.57%). Prolonged diabetes duration and elevated total cholesterol levels negatively correlated with postoperative VA prognosis (P < 0.05). The preoperative eye conditions, lens status, postoperative complications were not significantly different (P > 0.05).

CONCLUSIONS

Preoperative HD is associated with short-term reductions in IOP and improvements in BCVA after PPV, with long-term data correlating HD to better visual acuity outcomes. However, unmeasured confounders may influence these results. Prolonged diabetes duration and elevated total cholesterol levels also appear associated with slower visual recovery, warranting further validation.

Early detection of retinal and choroidal microvascular impairments in diabetic patients with myopia

Purpose

To evaluate and quantify diabetes-related retinal and choroid perfusion changes in individuals with and without high myopia and explore their associations with diabetes risk factors.

Methods

Diabetic patients [n = 133; 43 without diabetic retinopathy in group DM; 48 non-proliferative diabetic retinopathies in group DR; 42 without DR but with high myopia in group HM] underwent ophthalmological and endocrinological examinations. Swept-source optical coherence tomography angiography (SS-OCTA) was used to image the retinal vessel density (RVD), retinal thickness (RT), choroidal thickness (CT), choriocapillaris vessel perfusion (CPV) and choroidal vascularity index (CVI). Automatic segmentation of retinal and choroidal layers was performed using a deep learning-based U-Net architecture. A ResNet-50 convolutional neural network was further applied to analyze vascular density patterns and assist in DR grading. Univariate and multiple linear regression analyses explored the associations between perfusion and risk factors.

Results

The inner ring retinal vessel density and CVI in all areas were significantly different between groups (P < 0.05); CPV was not significantly changed except for the inferotemporal area among the groups. CT was decreased in all areas between groups (P < 0.05). The visual impairments in HM group was more obvious correlation with the retinal and choroidal structural changes. The AI-driven analysis revealed that decreased CVI and CT were significantly associated with age and spherical equivalent (SE), highlighting the utility of automated algorithms in identifying early microvascular impairments.

Conclusion

Diabetic patients with high myopia exhibited significantly lower CVI compared to those with diabetic retinopathy, indicating that CVI monitoring could facilitate risk stratification of diabetic retinopathy progression. The integration of SS-OCTA with artificial intelligence-enhanced segmentation and vascular analysis provides a refined method for early detection of retinal and choroidal microvascular impairments in diabetic populations.

Correlation Between Macular Microstructural Changes with Disease Staging and Visual Acuity in Diabetic Retinopathy

Purpose

To investigate the changes in macular microvascular structure at different stages of diabetic retinopathy (DR) and the correlation between macular ischemia and visual acuity.

Patients and Methods

A prospective cross-sectional study was conducted. A total of 173 patients with DR were enrolled and divided into three groups according to DR stage. The control group consisted of 29 gender and age matched healthy individuals. Macular perfusion indexes were measured by optical coherence tomography angiography (OCTA) and compared.

Results

The p-values of central foveal thickness (CFT), focal avascular zone (FAZ) area, and vessel density were less than 0.05 in DR patients and healthy individuals. As the severity of DR increased, there was a corresponding decline in visual acuity, the logMAR best corrected visual acuity (BCVA) was 0.40±0.30 in mild-moderate NPDR, then worsened to 0.48 ± 0.30 (p=0.059) in severe NPDR and further to 0.60 ± 0.34 (p=0.043) in PDR. Superficial capillary plexus (SCP) vessel density correlated negatively with logMAR BCVA (p<0.001, R=-0.267), whereas the severity of DR correlated positively with logMAR BCVA (p<0.001, R=0.199). And increased DR stage was associated with significant reductions in foveal density at 300 μm (FD-300 μm; p=0.004) and deep capillary plexus (DCP) vessel density (p=0.009).

Conclusion

Macular perfusion status decreases as DR progresses. Early changes of macular microvascular structure in different capillary plexus may indicate progression of DR severity and affect visual acuity.

SMART (artificial intelligence enabled) DROP (diabetic retinopathy outcomes and pathways): Study protocol for diabetic retinopathy management

INTRODUCTION

Delayed diagnosis of diabetic retinopathy (DR) remains a significant challenge, often leading to preventable blindness and visual impairment. Given that physicians are frequently the first point of contact for people with diabetes, there is a critical need for integrated screening programs within diabetes clinics to enhance DR management and reduce the risk of severe vision loss.

METHODS AND ANALYSIS

We will conduct a prospective cohort study comparing (i) the intervention cohort, screened at diabetes clinics and referred to eye clinics per the proposed pathway, and (ii) the standard-of-care (SOC) eye clinic cohort. The study will be conducted in Hyderabad, India, at LV Prasad Eye Institute and four IDEA (Institute of Diabetes, Endocrinology, and Adiposity) Clinics. The primary objective is to evaluate the effectiveness of a systematic diabetic retinopathy screening program in achieving earlier detection and reducing visual impairment among People With Diabetes (PWD) attending IDEA clinics compared to routine care at eye care settings. The screening program will be operationalized using AI-enabled tools and supported by trained non-medical technicians. We will perform visual acuity tests and non-mydriatic fundus photography using AI-assisted cameras. DR-positive patients will be referred for treatment and follow-up. We aim to achieve high accuracy (>90%) in appropriate referral of DR and high screening coverage (>80%) of eligible PWD. Success metrics include screening uptake, AI diagnostic accuracy, referral rates, cost-effectiveness, patient satisfaction, follow-up adherence, and long-term outcomes.

CONCLUSION

This study aims to enhance diabetic retinopathy screening and management through an AI-enabled approach at diabetes clinics, improving early detection and care pathways. The findings will contribute to evidence-based strategies for optimizing DR screening and management, with results disseminated through peer-reviewed publications to inform policy and practice.

TRIAL REGISTRATION

Trial registration number: CTRI/2024/03/064518 [Registered on: 20/03/2024] (https://ctri.nic.in/).

U-Shaped Association of Visceral Fat Area With Diabetic Peripheral Neuropathy

Background: The impact of visceral fat area (VFA) on diabetic peripheral neuropathy (DPN) remains controversial in Type 2 diabetes mellitus (T2DM), with conflicting evidence. Methods: We conducted a cross-sectional study at the National Metabolic Management Center of Beijing Luhe Hospital between October 2017 and May 2024. VFA was quantified using bioelectrical impedance analysis, and DPN was diagnosed according to standardized clinical criteria. The association between VFA and DPN was examined using multiple logistic regression models with comprehensive confounder adjustment. Nonlinear relationships were investigated through generalized additive models and threshold effect analyses. Results: Among 7436 T2DM patients (3044 females), the median VFA was 104 cm2 (interquartile range: 79.5-132 cm2), with DPN present in 26.55% of participants. Generalized additive models revealed a significant U-shaped association between VFA and DPN, with an inflection point at 133 cm2 (log-likelihood ratio test, p < 0.001). In the segmented regression analysis, each 10 cm2 increase in VFA below this threshold was associated with decreased DPN risk (OR: 0.95, 95% CI: 0.93-0.97, p < 0.05), while increases above the threshold were associated with elevated risk (OR: 1.06, 95% CI: 1.03-1.09, p < 0.05). Conclusions: A U-shaped association exists between VFA and DPN prevalence in T2DM patients, suggesting clinical relevance of moderate visceral adiposity.

Efficacy and Safety of Qiming Granule for Nerve Injury Associated with Non-proliferative Diabetic Retinopathy: A Multicenter, Randomized, Non-inferiority, Active-Controlled Clinical Trial

OBJECTIVE

To evaluate the efficacy and safety of Qiming Granule as an early treatment for patients with nerve injury associated with non-proliferative diabetic retinopathy (NPDR).

METHODS

This was a multicenter, randomized, non-inferiority, active-controlled clinical trial. Patients with NPDR complicated with nerve injury, regardless of whether they presented with fundus abnormalities, were randomly assigned in a 1:1 ratio via a randomized number table to orally receive either 4.5 g of Qiming Granule or 0.5 g of calcium dobesilate (CaD), both 3 times daily for 24 weeks. The primary endpoints were changes in retinal nerve fiber layer (RNFL) thickness and foveal avascular zone (FAZ) area from baseline to week 24. The secondary endpoints included changes in RNFL thickness and FAZ area from baseline to week 12, and visual function questionnaire (NEI-VFQ-25) and health survey questionnaire (SF-36 scale), CM syndrome element scale score and the rates of abnormal full-field electroretinogram (ERG), abnormal dilated fundus, and abnormal visual acuity at treatment of weeks 12 and 24. Adverse drug reactions (ADRs) were detected.

RESULTS

A total of 82 patients were enrolled in the study. Changes in RNFL thickness from baseline to week 24 in the Qiming Granule and CaD groups were -1.53 ± 9.88 µm and -4.61 ± 9.23 µm, respectively (a difference of 3.08 µm [97.5% CI: -2.11 to 8.25]). Changes in FAZ area from baseline to week 24 were -0.08 ± 0.39 mm2 and 0.01 ± 0.05 mm2, respectively (a difference of -0.09 mm2 [97.5% CI: -0.26 to 0.08]). Non-inferiority was achieved for both primary endpoints. There were no significant differences between the two groups in secondary endpoints, including changes in RNFL thickness and FAZ area from baseline to week 12, rates of abnormal ERG, dilated fundus, and visual acuity results at weeks 12 and 24, as well as NEI-VFQ-25, SF-36 scale, and CM syndrome element scale scores at week 24. ADRs were detected in 4 (9.76%) and 1 (2.44%) patients in the Qiming Granule and CaD groups, respectively. No serious ADRs occurred.

CONCLUSION

Qiming Granule demonstrates non-inferiority in terms of efficacy and safety as an early treatment for nerve injury associated with NPDR. (Registration No. ISRCTN39825773).

Analysis of risk factors for vitreous hemorrhage and recurrent hemorrhage after vitrectomy in patients with diabetic retinopathy

PURPOSE

To investigate the risk factors associated with recurrent hemorrhage following vitrectomy surgical intervention for diabetic retinopathy (DR).

METHODS

A retrospective analysis was conducted on 579 eyes diagnosed with DR necessitating surgical intervention. These cases were categorized into two groups: recurrent hemorrhage and non-recurrent hemorrhage. Comparative, random forest (RF), and regression analyses were subsequently performed to evaluate variables pertaining to patients' demographic information, clinical examination and blood test results, treatment approaches, lifestyle habits, and overall health status.

RESULTS

This study compared patients with recurrent and non-recurrent hemorrhages, revealing significant differences in factors such as endodiathermy, anticoagulant use, cerebrovascular diseases, smoking status, glycosylated hemoglobin levels and BMI. Patients with no recurrent hemorrhage have faster vision recovery. The univariable logistic regression analysis indicated that cerebrovascular disease (OR = 7.87, P < 0.001), anticoagulant use (OR = 16.72, P < 0.001), and elevated glycated hemoglobin levels (OR = 21.22, P < 0.001) exhibited strong associations with recurrent hemorrhage. The multivariable logistic regression analysis indicated that recurrent hemorrhage risk factors include anticoagulant use (OR = 120.77, P = 0.020) and glycated hemoglobin levels (OR = 18.41, P = 0.001).

CONCLUSIONS

Recurrent postoperative hemorrhage is influenced by several factors, notably the use of intraoperative endodiathermy, adjustments in ocular therapy, and management of the patient's systemic condition. In clinical practice, careful consideration of these factors is essential to mitigate postoperative hemorrhage in patients.

Diabetic foveal neovascularization is associated with diminished subfoveal choroidal flow on optical coherence tomography angiography

OBJECTIVE

To evaluate the role of retinal and choriocapillaris blood flow in the development of diabetic foveal neovascularization (DFN) using optical coherence tomography angiography (OCTA).

METHODS

Eyes with proliferative diabetic retinopathy (PDR) and DFN underwent OCTA imaging; defined as surface retinal neovascularization within the central foveal 1 mm diameter circle. 3 × 3 and 6 × 6 mm2 choriocapillaris and superficial and deep retinal capillary plexus (SCP and DCP) slabs were extracted to evaluate adjusted flow index (AFI) as a surrogate for blood flow. For choriocapillaris flow; total, subfoveal and extrafoveal AFI were assessed, while only total AFI was calculated for SCP and DCP. These findings were compared to healthy controls and eyes with PDR with no DFN.

RESULTS

18 eyes of 18 patients were included in each of the 3 groups: healthy controls, PDR with and without DFN. Choriocapillaris AFI was significantly lower in PDR with DFN than healthy controls in all but the 6 × 6 mm2 extrafoveal AFI (p < 0.01). PDR with DFN also showed a lower AFI compared to eyes without DFN, but only in the 3 × 3 mm2 total and subfoveal AFI (p = 0.01). SCP and DCP AFI were not statistically significant.

CONCLUSIONS

Our findings suggest that choroidal hypoperfusion may be a potential driving factor for the development of DFN. The detection of these changes in the smaller scans of the total and subfoveal areas suggests a rather exaggerated and localized subfoveal distribution of ischaemia. Larger longitudinal studies are needed to explore the use of subfoveal choroidal AFI as a prognostic sign for DFN.

Genetic mechanisms of hemispheric functional connectivity in diabetic retinopathy: a joint neuroimaging and transcriptomic study

Background

DR represents a major cause of global vision loss; however, the genetic basis of functional homotopy,a critical neurobiological metric reflecting interhemispheric functional synchronization, remains largely unexplored. Emerging evidence suggests that DR patients exhibiting aberrant VMHC may potentially associate with distinct transcriptional profiles. These findings could provide novel mechanistic insights into the neuropathological substrates underlying DR-related visual and cognitive dysfunction.

Methods

Resting-state fMRI data from 46 DR patients and 43 HCs were analyzed to compute VMHC for assessing interhemispheric functional connectivity. Spatial transcriptomic-neuroimaging associations were examined using AHBA, revealing genes significantly correlated with VMHC alterations. Subsequent analyses included functional enrichment assessment and PPI network construction.

Results

DR patients demonstrated significantly lower VMHC in bilateral LING, PoCG, and PreCG versus controls, indicating impaired interhemispheric connectivity in visual-sensorimotor networks. VMHC variations spatially correlated with 4,000 genes (2,000 positive/negative each), enriched in transcriptional regulation, mitochondrial function, synaptic activity (BP/CC/MF), and lipid metabolism/N-glycan biosynthesis (KEGG). PPI network identified hub genes (ACTB/MRPL9/MRPS6,positive; H4C6/NDUFAB1/H3C12,negative) regulating mitochondrial dynamics, cytoskeleton, and epigenetics.

Conclusion

This study represents the first integration of fMRI and transcriptomics to elucidate the genetic determinants underlying VMHC disruption in DR. The findings demonstrate that impaired interhemispheric connectivity in DR involves complex interactions among genes regulating neurovascular, metabolic, and neurodegenerative pathways. These results significantly advance the understanding of neurological manifestations in DR and identify potential therapeutic targets for clinical intervention.

What do You Need to Know after Diabetes and before Diabetic Retinopathy?

Diabetic retinopathy (DR) is a leading cause of vision impairment and blindness among individuals with diabetes mellitus. Current clinical diagnostic criteria mainly base on visible vascular structure changes, which are insufficient to identify diabetic patients without clinical DR (NDR) but with dysfunctional retinopathy. This review focuses on retinal endothelial cells (RECs), the first cells to sense and respond to elevated blood glucose. As blood glucose rises, RECs undergo compensatory and transitional phases, and the correspondingly altered molecules are likely to become biomarkers and targets for early prediction and treatment of NDR with dysfunctional retinopathy. This article elaborated the possible pathophysiological processes focusing on RECs and summarized recently published and reliable biomarkers for early screening and emerging intervention strategies for NDR patients with dysfunctional retinopathy. Additionally, references for clinical medication selection and lifestyle recommendations for this population are provided. This review aims to deepen the understanding of REC biology and NDR pathophysiology, emphasizes the importance of early detection and intervention, and points out future directions to improve the diagnosis and treatment of NDR with dysfunctional retinopathy and to reduce the occurrence of DR.

Ocular drug delivery systems based on nanotechnology: a comprehensive review for the treatment of eye diseases

Ocular drug delivery is a significant challenge due to the intricate anatomy of the eye and the various physiological barriers. Conventional therapeutic approaches, while effective to some extent, often fall short in effectively targeting ocular diseases, resulting in suboptimal therapeutic outcomes due to factors such as poor ocular bioavailability, frequent dosing requirements, systemic side effects, and limited penetration through ocular barriers. This review elucidates the eye's intricate anatomy and physiology, prevalent ocular diseases, traditional therapeutic modalities, and the inherent pharmacokinetic and pharmacodynamic limitations associated with these modalities. Subsequently, it delves into nanotechnology-based solutions, presenting breakthroughs in nanoformulations such as nanocrystals, liposomes, dendrimers, and nanoemulsions that have demonstrated enhanced drug stability, controlled release, and deeper ocular penetration. Additionally, it explores a range of nanosized carriers, including nano-structured lipid carriers, hydrogels, nanogels, nanoenzymes, microparticles, conjugates, exosomes, nanosuspensions, viral vectors, and polymeric nanoparticles, and their applications. Unique insights include emerging innovations such as nanowafers and transcorneal iontophoresis, which indicate paradigm shifts in non-invasive ocular drug delivery. Furthermore, it sheds light on the advantages and limitations of these nanotechnology-based platforms in addressing the challenges of ocular drug delivery. Though nano-based drug delivery systems are drawing increasing attention due to their potential to enhance bioavailability and therapeutic efficacy, the review ends up emphasizing the imperative need for further research to drive innovation and improve patient outcomes in ophthalmology.

Epiretinal Membrane Is Associated with Diabetic Retinopathy Severity and Cumulative Anti-VEGF Injections

Objective

To identify and assess the risk factors for epiretinal membrane (ERM) formation in patients with diabetes mellitus (DM).

Design

A cross-sectional observational study.

Participants

Diabetes mellitus subjects with varying severities of diabetic retinopathy (DR).

Methods

Epiretinal membrane severity was graded using OCT macular volumetric scans. After comparing the features of eyes with and without ERM, a logistic regression model was employed to investigate risk factors for the presence of ERM formation, with explanatory variables that were statistically significant between 2 groups.

Main Outcome Measures

The risk factors for ERM formation and their odds ratios.

Results

The study analyzed 207 eyes of 207 patients without center-involving diabetic macular edema, including 44 with DM but no DR, 23 with mild DR, 42 with moderate DR, 28 with severe nonproliferative DR, and 70 with proliferative DR. Among these, 49 eyes had ERM. Eyes with ERM demonstrated significantly higher DR severity, more frequent history of cataract surgery, a greater number of anti-VEGF injections, less frequent hypertension, and higher geometric perfusion deficits in the deep capillary plexus, as calculated by OCT angiography, compared to eyes without ERM. After adjusting for explanatory variables, the logistic regression model identified DR severity and the number of injections as significant risk factors for ERM formation. The adjusted odds ratios were 2.388 (95% confidence interval [CI]: 1.694-3.566, P < 0.0001) for DR severity and 1.196 (95% CI: 1.013-1.444, P = 0.0449) for the number of injections, respectively.

Conclusions

In this cross-sectional study, DR severity as well as the number of anti-VEGF injections were identified as significant risk factors for ERM formation.

Financial Disclosures

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

An inherently interpretable AI model improves screening speed and accuracy for early diabetic retinopathy

Diabetic retinopathy (DR) is a frequent complication of diabetes, affecting millions worldwide. Screening for this disease based on fundus images has been one of the first successful use cases for modern artificial intelligence in medicine. However, current state-of-the-art systems typically use black-box models to make referral decisions, requiring post-hoc methods for AI-human interaction and clinical decision support. We developed and evaluated an inherently interpretable deep learning model, which explicitly models the local evidence of DR as part of its network architecture, for clinical decision support in early DR screening. We trained the network on 34,350 high-quality fundus images from a publicly available dataset and validated its performance on a large range of ten external datasets. The inherently interpretable model was compared to post-hoc explainability techniques applied to a standard DNN architecture. For comparison, we obtained detailed lesion annotations from ophthalmologists on 65 images to study if the class evidence maps highlight clinically relevant information. We tested the clinical usefulness of our model in a retrospective reader study, where we compared screening for DR without AI support to screening with AI support with and without AI explanations. The inherently interpretable deep learning model obtained an accuracy of .906 [.900-.913] (95%-confidence interval) and an AUC of .904 [.894-.913] on the internal test set and similar performance on external datasets, comparable to the standard DNN. High evidence regions directly extracted from the model contained clinically relevant lesions such as microaneurysms or hemorrhages with a high precision of .960 [.941-.976], surpassing post-hoc techniques applied to a standard DNN. Decision support by the model highlighting high-evidence regions in the image improved screening accuracy for difficult decisions and improved screening speed. This shows that inherently interpretable deep learning models can provide clinical decision support while obtaining state-of-the-art performance improving human-AI collaboration.

Non-invasive biopsy diagnosis of diabetic kidney disease via deep learning applied to retinal images: a population-based study

BACKGROUND

Improving the accessibility of screening diabetic kidney disease (DKD) and differentiating isolated diabetic nephropathy from non-diabetic kidney disease (NDKD) are two major challenges in the field of diabetes care. We aimed to develop and validate an artificial intelligence (AI) deep learning system to detect DKD and isolated diabetic nephropathy from retinal fundus images.

METHODS

In this population-based study, we developed a retinal image-based AI-deep learning system, DeepDKD, pretrained using 734 084 retinal fundus images. First, for DKD detection, we used 486 312 retinal images from 121 578 participants in the Shanghai Integrated Diabetes Prevention and Care System for development and internal validation, and ten multi-ethnic datasets from China, Singapore, Malaysia, Australia, and the UK (65 406 participants) for external validation. Second, to differentiate isolated diabetic nephropathy from NDKD, we used 1068 retinal images from 267 participants for development and internal validation, and three multi-ethnic datasets from China, Malaysia, and the UK (244 participants) for external validation. Finally, we conducted two proof-of-concept studies: a prospective real-world study with 3 months' follow-up to evaluate the effectiveness of DeepDKD in screening DKD; and a longitudinal analysis of the effectiveness of DeepDKD in differentiating isolated diabetic nephropathy from NDKD on renal function changes with 4·6 years' follow-up.

FINDINGS

For detecting DKD, DeepDKD achieved an area under the receiver operating characteristic curve (AUC) of 0·842 (95% CI 0·838-0·846) on the internal validation dataset and AUCs of 0·791-0·826 across external validation datasets. For differentiating isolated diabetic nephropathy from NDKD, DeepDKD achieved an AUC of 0·906 (0·825-0·966) on the internal validation dataset and AUCs of 0·733-0·844 across external validation datasets. In the prospective study, compared with the metadata model, DeepDKD could detect DKD with higher sensitivity (89·8% vs 66·3%, p<0·0001). In the longitudinal study, participants with isolated diabetic nephropathy and participants with NDKD identified by DeepDKD had a significant difference in renal function outcomes (proportion of estimated glomerular filtration rate decline: 27·45% vs 52·56%, p=0·0010).

INTERPRETATION

Among diverse multi-ethnic populations with diabetes, a retinal image-based AI-deep learning system showed its potential for detecting DKD and differentiating isolated diabetic nephropathy from NDKD in clinical practice.

FUNDING

National Key R & D Program of China, National Natural Science Foundation of China, Beijing Natural Science Foundation, Shanghai Municipal Key Clinical Specialty, Shanghai Research Centre for Endocrine and Metabolic Diseases, Innovative research team of high-level local universities in Shanghai, Noncommunicable Chronic Diseases-National Science and Technology Major Project, Clinical Special Program of Shanghai Municipal Health Commission, and the three-year action plan to strengthen the construction of public health system in Shanghai.

Development and validation of an AI algorithm to generate realistic and meaningful counterfactuals for retinal imaging based on diffusion models

Counterfactual reasoning is often used by humans in clinical settings. For imaging based specialties such as ophthalmology, it would be beneficial to have an AI model that can create counterfactual images, illustrating answers to questions like "If the subject had had diabetic retinopathy, how would the fundus image have looked?". Such an AI model could aid in training of clinicians or in patient education through visuals that answer counterfactual queries. We used large-scale retinal image datasets containing color fundus photography (CFP) and optical coherence tomography (OCT) images to train ordinary and adversarially robust classifiers that classify healthy and disease categories. In addition, we trained an unconditional diffusion model to generate diverse retinal images including ones with disease lesions. During sampling, we then combined the diffusion model with classifier guidance to achieve realistic and meaningful counterfactual images maintaining the subject's retinal image structure. We found that our method generated counterfactuals by introducing or removing the necessary disease-related features. We conducted an expert study to validate that generated counterfactuals are realistic and clinically meaningful. Generated color fundus images were indistinguishable from real images and were shown to contain clinically meaningful lesions. Generated OCT images appeared realistic, but could be identified by experts with higher than chance probability. This shows that combining diffusion models with classifier guidance can achieve realistic and meaningful counterfactuals even for high-resolution medical images such as CFP images. Such images could be used for patient education or training of medical professionals.

Diagnostic utility of artificial intelligence software through non-mydriatic digital retinography in the screening of diabetic retinopathy: an overview of reviews

OBJECTIVE

To evaluate the capability of artificial intelligence (AI) in screening for diabetic retinopathy (DR) utilizing digital retinography captured by non-mydriatic (NM) ≥45° cameras, focusing on diagnosis accuracy, effectiveness, and clinical safety.

METHODS

We performed an overview of systematic reviews (SRs) up to May 2023 in Medline, Embase, CINAHL, and Web of Science. We used AMSTAR-2 tool to assess the reliability of each SR. We reported meta-analysis estimates or ranges of diagnostic performance figures.

RESULTS

Out of 1336 records, ten SRs were selected, most deemed low or critically low quality. Eight primary studies were included in at least five of the ten SRs and 125 in less than five SRs. No SR reported efficacy, effectiveness, or safety outcomes. The sensitivity and specificity for referable DR were 68-100% and 20-100%, respectively, with an AUROC range of 88 to 99%. For detecting DR at any stage, sensitivity was 79-100%, and specificity was 50-100%, with an AUROC range of 93 to 98%.

CONCLUSIONS

AI demonstrates strong diagnostic potential for DR screening using NM cameras, with adequate sensitivity but variable specificity. While AI is increasingly integrated into routine practice, this overview highlights significant heterogeneity in AI models and the cameras used. Additionally, our study enlightens the low quality of existing systematic reviews and the significant challenge of integrating the rapidly growing volume of emerging evidence in this field. Policymakers should carefully evaluate AI tools in specific contexts, and future research must generate updated high-quality evidence to optimize their application and improve patient outcomes.

Examining the Visual Search Behaviour of Experts When Screening for the Presence of Diabetic Retinopathy in Fundus Images

Objectives: This study investigated the visual search behaviour of optometrists and fellowship-trained ophthalmologists when screening for diabetic retinopathy in retinal photographs. Methods: Participants assessed and graded retinal photographs on a computer screen while a Gazepoint GP3 HD eye tracker recorded their eye movements. Areas of interest were derived from the raw data using Hidden Markov modelling. Fixation strings were extracted by matching raw fixation data to areas of interest and resolving ambiguities with graph search algorithms. Fixation strings were clustered using Affinity Propagation to determine search behaviours characteristic of the correct and incorrect response groups. Results: A total of 23 participants (15 optometrists and 8 ophthalmologists) completed the grading task, with each assessing 20 images. Visual search behaviour differed between correct and incorrect responses, with data suggesting correct responses followed a visual search strategy incorporating the optic disc, macula, superior arcade, and inferior arcade as areas of interest. Data from incorrect responses suggest search behaviour driven by saliency or a search pattern unrelated to anatomical landmarks. Referable diabetic retinopathy was correctly identified in 86% of cases. Grader accuracy was 64.8% with good inter-grader agreement (α = 0.818). Conclusions: Our study suggests that a structured visual search strategy is correlated with higher accuracy when assessing retinal photographs for diabetic retinopathy. Referable diabetic retinopathy is detected at high rates; however, there is disagreement between clinicians when determining a precise severity grade.

The rs3844492/ARHGAP22 and rs741301/ELMO1 polymorphisms are associated with changes in laboratory markers of renal damage among patients with type 2 diabetes mellitus

OBJECTIVE

To investigate the association between the rs3844492/ARHGAP22 and rs741301/ELMO1 polymorphisms and diabetic kidney disease in patients with type 2 diabetes mellitus.

METHODS

The sample consisted of 740 patients with type 2 diabetes mellitus and diabetic kidney disease (cases) and 303 patients with type 2 diabetes mellitus, but no diabetic kidney disease (controls). The genotyping of the polymorphisms was conducted using real-time polymerase chain reaction with Taqman probes.

RESULTS

The frequency of the rs3844492/ARHGAP22 G/G genotype was 16.8% in the control group and 15.7% in cases (p = 0.069). After adjusting for covariables, the presence of the G allele was associated with risk for diabetic kidney disease (OR = 1.435, 95% CI 1.023 - 2.011; p = 0.036), as well as with a decreased estimated glomerular filtration rate (p = 0.012) and elevated creatinine levels (p = 0.009). No difference was observed in the rs741301/ELMO1 genotype frequencies between groups (p = 0.800). However, the presence of the C allele appears to be associated with higher creatinine levels in patients with type 2 diabetes mellitus (p = 0.064).

CONCLUSION

The rs3844492/ARHGAP22 and rs741301/ELMO1 polymorphisms are associated with alterations in renal function markers among patients with type 2 diabetes mellitus.

Adjusted glycated albumin is a novel indicator of glycemic control in patients with macroalbuminuria in diabetic kidney disease

Glycated albumin (GA), a blood glucose monitoring biomarker, is impacted by variables such as albumin turnover and is not entirely relevant throughout diabetic kidney disease (DKD). There is insufficient data to routinely adjust GA measurements. We examined how albuminuria affected clinically measured GA (mGA) and adjusted GA (adjGA). We included 195 patients with DKD, 108 with non-macroalbuminuria and 87 with macroalbuminuria, and adjusted GA based on albumin, albuminuria, and body weight. Subgroups were divided to two groups according to albuminuria and serum albumin levels. The relationship between mGA, adjGA, and glucose was investigated. The optimum GA correction method based on albumin turnover metabolism was investigated: adjGA = mGA×[1+(8×K×UP) ÷ (11×V×SA)]. where K represents the standard metabolic days of albumin (15 days), UP is 24-hour urine protein excretion (g/24 h), V is plasma volume (calculated as 5% of body weight in liters), and SA is serum albumin concentration (g/L). In non-macroalbuminuria, mGA was 19.75% and adjGA was 22.32%, and in macroalbuminuria, mGA was 13.20% and adjGA was 22.45%, the mGA was substantially different across albuminuria categories (P < 0.001), but adjGA was not. HbA1c, 24-h urine protein(24hUP) and serum albumin (ALB) were influencing variables for mGA (P < 0.001), while 24hUP and ALB had no effect on adjGA (P > 0.05). The adjGA had stronger cor-relation with blood glucose than mGA, especially in the context of macroalbuminuria. Macroalbuminuria lowers mGA accuracy. In DKD patients with macroalbuminuria, adjusted GA is a novel indication of glucose monitoring.

Increased fatty acid-binding protein 4 levels are associated with the risk of developing retinopathy in type 2 diabetes mellitus patients

AIM

Fatty acid-binding protein 4 (FABP4) is associated with the risk of developing diabetes and its microvascular complications. We aimed to explore the association between serum FABP4 levels and the risk of diabetic retinopathy (DR) in patients with type 2 diabetes mellitus (T2DM).

METHODS

Serum FABP4 levels were measured via enzyme-linked immunosorbent assay in 275 individuals (78 healthy controls, 197 T2DM patients). DR severity was determined via fundus fluorescence angiography. Multivariate analyses were performed via logistic regression models. The diagnostic value of these measures was assessed via receiver operating characteristic curve analysis.

RESULTS

Serum FABP4 levels were significantly greater in the proliferative DR (PDR) group than in the ZeroDR (ZDR) and non-proliferative DR (NPDR) groups, and the FABP4 levels positively with DR severity (r = 0.328, P < 0.001). Logistic regression analysis revealed that after adjusting for potential confounders, increased FABP4, fasting plasma glucose (FPG) and hemoglobin A1c (HbA1c) levels were risk factors for DR development, and FABP4 was an independent risk factor for PDR development. A multivariate logistic regression model that included FABP4 as a categorical binary variable with a cutoff value of 20.57 ng/ml revealed that a level of FABP4 above the cutoff value increased the DR risk (OR=6.394; 95 % CI=3.18;13.58; P < 0.001). Similarly, a FABP4 concentration above the cutoff value of 24.40 ng/ml increased the PDR risk (OR=4.686; 95 % CI=1.907;12.34; P = 0.001).

CONCLUSION

The FABP4 level is associated with DR severity and has the potential as a serum biomarker for DR prediction.

The C allele of the rs741301 polymorphism in the ELMO1 gene is associated with increased risk of diabetic retinopathy in patients with type 2 diabetes mellitus

OBJECTIVE

To investigate the association of the rs741301 polymorphism in the ELMO1 gene with diabetic retinopathy (DR) in patients with type 2 diabetes mellitus (T2DM).

MATERIALS AND METHODS

This study analyzed 350 patients withT2DM and DR (cases) and 234 patients withT2DM without this complication but with more than 10 years of diabetes mellitus (DM) (controls). DR was diagnosed by indirect fundoscopy. Genotyping was performed by allelic discrimination real-time PCR.

RESULTS

The frequency of the C/C genotype of the rs741301 polymorphism in the ELMO1 gene was 26.9% in cases and 17.9% in controls (P = 0.011). After adjustment for covariables, the C/C genotype was associated with an increased risk of DR [odds ratio (OR) = 1.805, 95%CI 1.101-2.961; P = 0.019]. This association remained significant in dominant and additive inheritance models after adjustment for the same variables [OR = 1.597, 95%CI 1.089-2.343; P = 0.017; and OR = 1.818, 95%CI 1.099-3.007; P = 0.020].

CONCLUSION

This study demonstrated an association between the presence of the C allele of the ELMO1 rs741301 polymorphism and an increased risk of DR in patients with T2DM from Southern Brazil.

Diabetic retinopathy: a comprehensive review of pathophysiology and emerging treatments

Diabetic retinopathy constitutes a major complication associated with diabetes mellitus, resulting in visual impairment and blindness on a global scale. The pathophysiology of DR is characterized by intricate interactions among metabolic, hemodynamic, and inflammatory pathways, which include the activation of the polyol pathway, the accumulation of advanced glycation end products, the overactivation of protein kinase C, dysregulation of the renin-angiotensin-aldosterone system, and retinal neurodegeneration. This review investigates the classification, complex pathophysiology, and therapeutic modalities for DR, encompassing conventional interventions such as anti-VEGF agents, aldose reductase inhibitors, angiotensin receptor blockers, laser photocoagulation, and vitrectomy. Innovative treatments, including advanced anti-VEGF agents, neuroprotective strategies, gene and stem cell therapies, and advancements in drug delivery systems, exhibit considerable transformative potential. Furthermore, integrating artificial intelligence for early detection and modulation of inflammatory pathways signifies cutting-edge progress in the field. By integrating contemporary knowledge and prospective avenues, this review underscores the significance of comprehending the multifaceted nature of DR and the advancements in its therapeutic approaches. The objective is to bridge the gaps between research findings and clinical application, thereby providing a comprehensive resource to enhance outcomes and quality of life for individuals impacted by DR.

The diagnostic value of platelet-to-neutrophil ratio in diabetic macular edema

PURPOSE

To evaluate the diagnostic value of platelet-to-neutrophil ratio (PNR) in the occurrence of diabetic macular edema (DME) in patients with diabetic retinopathy (DR).

METHODS

This cross-sectional study included 366 participants categorized into four groups: DME group (n = 96), DR group (n = 90, DR without DME), diabetes mellitus (DM) group (n = 90, without DR), and healthy control group (n = 90). PNR was calculated by dividing the platelet count by the neutrophil count. Each subject was classified as one of three DME types according to the optical coherence tomography (OCT) features: diffuse retinal thickening (DRT), cystoid macular edema (CME), serous retinal detachment (SRD). The correlations between the PNR and the occurrence of DME, as well as the DME subtypes based on OCT were investigated. Multivariate logistic regression analysis was employed to determine the risk factors for DME. Receiver operating characteristic (ROC) curve analysis was conducted to assess the predictive value of PNR for DME.

RESULTS

DME group exhibited significantly lower PNR level compared to the other three groups [50.73 (38.92, 65.20) in DME group, 95.63 (68.83, 120.19) in DR group, 92.39 (72.38, 130.61) in DM group, and 100.66 (75.26, 152.77) in healthy control group, respectively, p < 0.001], but did not differ across the DME subtypes based on OCT (p = 0.548). The ROC curve demonstrated that the PNR could better predict DME (area under the curve = 0.832, 95% confidence interval: 0.773 - 0.891, p < 0.001). When the cut-off value of the PNR was 68.51, the sensitivity was 80.2%, and the specificity was 75.6%. Multivariate regression analysis indicated that PNR ≤ 68.51 was an independent risk factor for DME occurrence in DR patients (Odds ratio = 12.05, 95% confidence interval: 5.93 - 24.47, p < 0.001).

CONCLUSION

PNR ≤ 68.51 was strongly associated with the development of DME in DR patients, while no significant differences in PNR levels were observed across the different OCT morphological groups. Hence, PNR may serve as a valuable diagnostic biomarker for identifying DME, thereby enhancing risk stratification and management strategies for patients with DR.

OCTA and Microperimetry Changes Preceding the Appearance of Diabetic Retinopathy in Patients with Type 1 Diabetes

PURPOSE

to evaluate changes in retinal microvasculature and sensitivity (RS) preceding the appearance of diabetic retinopathy (DR) among patients with type 1 diabetes (T1D).

METHODS

in this observational cross-sectional cohort study, vascular parameters measured by OCTA and RS evaluated by microperimetry were assessed in patients with T1D without DR (no-DR), T1D with mild DR (m-DR), and healthy controls.

RESULTS

Sixty-two eyes of 31 patients with T1D and 40 eyes of 20 healthy patients were included. OCTA examinations did not yield any significant differences in terms of perfusion density (PD), vascular density (VD), foveal avascular zone (FAZ) area, FAZ perimeter or FAZ circularity between patients with diabetes (no-DR vs. m-DR). However, comparisons between healthy controls and patients with diabetes (both no-DR and m-DR groups) revealed statistically significant differences in PD, VD, and FAZ area. Similarly, no significant differences were observed between no-DR and m-DR groups regarding RS, gaze fixation stability (GFS), or macular integrity (MI). Nevertheless, mean RS and MI were significantly impaired in patients with T1D, both in no-DR and m-DR groups, compared to healthy controls. A statistically significant positive correlation was observed between RS and PD and between FAZ area and RS.

CONCLUSION

although no differences were found between patients with diabetes without DR and those with mild DR, these patients already demonstrated some degree of retinal impairment, both structural and functional, when compared to healthy controls. Our data support the hypothesis that neurodegeneration occurs together with microvascular damage at early stages of diabetes.

PREVALENCE AND ASSOCIATED FACTORS OF CHOROIDAL CAVERN IN PATIENTS WITH TYPE 2 DIABETES MELLITUS

PURPOSE

To study the prevalence of choroidal caverns (CCs) in patients with Type 2 diabetes (T2DM) and their association with demographic and clinical data.

METHODS

Patients with T2DM and nondiabetic control subjects were included in a cross-sectional, monocenter study. The main outcome measure was the presence of CCs. Exploratory variables included age, sex, diabetic retinopathy status, duration of T2DM, visual acuity, arterial hypertension, hyperlipidemia, HbA1c, choroidal thickness and choroidal vascularity index. Group differences between eyes with and without CCs and associations with demographic and clinical variables were assessed.

RESULTS

After including a total of 205 eyes of 116 patients, the prevalence of CCs was 25% in the control, 28% in the T2DM without diabetic retinopathy, and 9.5% in the diabetic retinopathy groups. The locations of the caverns were choriocapillaris (N = 2, 4.2%), Sattler layer (N = 29, 60.4%), and Haller layer (N = 17, 35.4%). In total, 34.3% of patients had caverns in both eyes. No significant differences were found in systemic conditions (including arterial hypertension, hyperlipidemia, HbA1c, or T2DM disease duration). Choroidal vascularity index was the only parameter independently associated with the presence of CCs (OR 1.37).

CONCLUSION

The authors found CCs in almost a quarter of patients with and without T2DM. Choroidal caverns were mainly located in the Sattler layer. Choroidal vascularity index was independently associated with the presence of CCs. These findings suggest that the presence of CCs may be a sign of a higher metabolic activity within the choroidal microenvironment, irrespective of T2DM status.

Association between multiple metals mixture and diabetic retinopathy in older adults with diabetes mellitus: a cross-sectional study in China

Previous studies have linked single metal with diabetic retinopathy (DR), but information about the combined effects of multiple metals mixture was scarce. Thus, we performed this cross-sectional study to investigate the single and joint associations between multiple metals mixture and DR risk among elderly diabetic population in China. A total of 1127 elderly adults (aged ≥ 60) with diabetes mellitus from a large-scale DR screening program in southern China included. Metals (beryllium, magnesium, chromium, manganese, iron, nickel, copper, arsenic, thallium and lead) in serum were quantified by inductively coupled plasma mass spectrometer. DR was diagnosed according to the consensus of the global DR project group. The relationships between metals and DR risks were estimated by logistic regression, Bayesian kernel machine regression (BKMR) and weighted quantile sum (WQS) regression. Of 1127 older adults with diabetes mellitus, there were 324 DR and 803 non-DR participants. Logistic regression models found serum magnesium and iron were negatively related to DR risks. Both BKMR model and WQS regression revealed that higher serum levels of multiple metals mixture were associated with lower risks of DR, with Be contributing the most to the overall effect. Additionally, in subgroup analyses, the interaction between beryllium and blood pressure on DR risk was also observed (Pinteraction = 0.008). Overall, these results provided new evidence of direct association between multiple metals mixture and DR risk among elderly diabetic population in China.

Urinary titin as a biomarker of sarcopenia in diabetes: a propensity score matching analysis

PURPOSE

Measuring urinary titin levels is expected to be useful in screening for muscle damage or injury in various diseases. We evaluated whether urinary titin levels were elevated in individuals with type 2 diabetes mellitus (T2DM) and how urinary titin levels were associated with the diagnosis of sarcopenia in T2DM.

METHODS

We performed a cross-sectional analysis of 114 controls and 515 patients with T2DM. Multivariate-adjusted models were used to determine the odds ratios (OR) of urinary titin cutoff values for diagnosing sarcopenia.

RESULTS

Urinary titin levels were higher in the T2DM group than in the non-diabetes group after propensity score matching (median [IQR] 3.2 [2.3, 4.6] vs. 4.4 [2.7, 6.9] pmol/mg·creatinine). T2DM was associated with high titin levels after correction for comorbidities (odds ratio 2.46, 95% confidence interval (CI) 1.29-4.70, P = 0.006) but not after correction for sarcopenia-associated factors. Urinary titin levels above the cutoff value showed an odd ratio of 6.61 (age- and body mass index-adjusted, 1.26-34.6, P = 0.021) for the diagnosis of sarcopenia in men with T2DM aged ≥ 75 years.

CONCLUSION

Results indicated that T2DM was associated with a high-titin state and that the urinary titin cutoff value could be useful for identifying candidates at high risk for sarcopenia, such as elderly men.

Magnesium, fibrinolysis and clotting interplay among children and adolescents with type 1 diabetes mellitus; potential mediators of diabetic microangiopathy

BACKGROUND AND AIM

Hypomagnesemia and clotting disorders have been reported among people with diabetes especially those with type 2 diabetes (T2DM). Magnesium plays a crucial role in hemostasis and hypomagnesemia was found to increase the thrombotic risk. The patho-mechanism linking magnesium, clotting disorders, and diabetic microangiopathy in T1DM remains to be unraveled. Hence this study aimed to assess the magnesium level among children and adolescents with T1DM compared to healthy controls and to correlate it with coagulopathy markers and diabetic microangiopathy.

METHODS

Forty-six children and adolescents with T1DM & 46 controls were assessed for serum magnesium, prothrombin time (PT), activated-partial thromboplastin time (aPTT), plasminogen activator inhibitor-1 (PAI-1) and HbA1c. The Toronto clinical scoring system, fundus, urinary microalbumin, and serum fasting lipids were used to assess diabetic microangiopathy.

RESULTS

Children and adolescents with T1DM have significantly lower magnesium, PT, aPTT, and significantly higher PAI-1 than controls (p<0.001), this is more evident in those having microangiopathy than those without (p<0.001). Serum magnesium is positively correlated with PT, aPTT, and HDL and negatively correlated with insulin daily dose, PAI-1, HbA1c, triglycerides, and urinary microalbumin. Multivariate-logistic regression revealed that diabetes duration, HbA1c, PT, aPTT, PAI-1, and urinary microalbumin were independently associated with serum magnesium among children and adolescents with T1DM (p<0.05).

CONCLUSION

Children and adolescents with T1DM have lower magnesium levels than controls; that is more pronounced among those having microangiopathy. Low serum magnesium is associated with poor glycemic control, coagulopathy, and diabetic microangiopathy among children and adolescents with T1DM. Magnesium supplementation combined with standard insulin therapy in pediatric patients with T1DM is recommended for better glycemic control and prevention of diabetic microangiopathy.

The evolution of diabetic retinopathy screening

Diabetic retinopathy (DR) is a leading cause of preventable blindness and has emerged as a global health challenge, necessitating the development of robust management strategies. As DR prevalence continues to rise, advancements in screening methods have become increasingly critical for timely detection and intervention. This review examines three key advancements in DR screening: a shift from specialist to generalist approach, the adoption of telemedicine strategies for expanded access and enhanced efficiency, and the integration of artificial intelligence (AI). In particular, AI offers unprecedented benefits in the form of sustainability and scalability for not only DR screening but other aspects of eye health and the medical field as a whole. Though there remain barriers to address, AI holds vast potential for reshaping DR screening and significantly improving patient outcomes globally.

Predict and prevent microvascular complications of type 2 diabetes: a cross-sectional and longitudinal study in Chinese communities

Purpose

This study investigates the incidence, predictors, and preventive strategies for microvascular complications in type 2 diabetes patients in community settings.

Methods

Data were collected from 3,008 type 2 diabetes patients enrolled across 31 clinics in Beijing and Hebei. Prevalence and incidence of diabetic kidney disease (DKD), diabetic retinopathy (DR), and diabetic peripheral neuropathy (DPN) were assessed. Predictors were identified using XGBoost and Cox regression, and the impact of lifestyle and multifactorial interventions (MFI) was analyzed.

Results

The prevalence of DKD, DR, and DPN were 39.5%, 26.2%, and 27.1%, respectively, with incidences of 74, 21, and 28 per 1000-person year. XGBoost identified that diabetes duration, age, HbA1c, FBG, triglyceride, BP, serum creatinine, proteinuria, aspirin and statin use were associated with those microvascular complications. The risk of DKD increased more rapidly as HbA1c exceeded 7.5% and decreased as blood pressure was maintained below 120/70 mmHg. Cox regression models showed that community-based intervention, including lifestyle modifications, were associated with a lower risk of DR and DPN. The study also found that higher variability in HbA1c and albumin-to-creatinine ratio (ACR) was associated with an increased risk of microvascular complications.

Conclusions

Community-based interventions significantly reduce the of DR and DPN, highlighting the need for individualized glycemic and BP management in primary care. The findings emphasize the importance of comprehensive management strategies to prevent the development and progression of microvascular complications in type 2 diabetes patients.

Clinical trial registration

http://www.chictr.org.cn/, identifier ChiCTR-TRC-13003222.

Changes in peripapillary microvasculature and retinal nerve fibre layer in diabetes and diabetic retinopathy using optical coherence tomographic angiography: a community-based, cross-sectional study

OBJECTIVE

To evaluate changes in the peripapillary retinal microvasculature and retinal nerve fibre layer (RNFL) in diabetic participants with various stages of diabetic retinopathy (DR) using swept-source optical coherence tomographic angiography (OCTA).

DESIGN

Community-based, cross-sectional study.

SETTING

This study was conducted in a tertiary teaching hospital in Guangzhou, China.

PARTICIPANTS

A total of 1325 ocular-treatment-naive participants, of whom 1115 had no DR and 210 had DR, were recruited in a community in Guangzhou, China.

PRIMARY AND SECONDARY OUTCOME MEASURES

A commercially available OCTA device was used to obtain various peripapillary retinal microvascular metrics centred on the optic disc, including vessel density (VD), vessel length density (VLD) and vessel diameter index (VDI). The peripapillary RNFL thickness was automatically obtained using built-in software. Linear regression analyses were used to evaluate the association of the peripapillary OCTA parameters (VD, VLD and VDI), RNFL thickness with various DR stages and average RNFL thickness with peripapillary OCTA parameters.

RESULTS

Moderate and severe DR had progressively decreased VD in the peripapillary ring (β = -0.72, 95% CI = -1.31 to -0.14 and -1.79, 95% CI = -2.81 to -0.77, respectively) and other regions (all p<0.05). Similar changes were observed between peripapillary VLD and moderate and severe DR (all p<0.05). Moderate (β = -4.56, 95% CI = -8.97 to -0.15, p=0.043) and severe DR (β = -10.12, 95% CI = -18.29 to -1.95, p=0.015) had significant thinner peripapillary RNFL in the inferior quadrant. VD and VLD were linearly associated with the average RNFL in the peripapillary ring and average peripapillary area (all p<0.05).

CONCLUSIONS

The peripapillary retinal microvasculature and RNFL were significantly reduced with the progression of DR, which suggests that monitoring differences in peripapillary microvasculature and the RNFL may be a promising approach to detecting DR progression.

Correlation of point-wise retinal sensitivity with localized features of diabetic macular edema using deep learning

OBJECTIVE

To evaluate the association between localized features of diabetic macular edema (DME) and point-wise retinal sensitivity (RS) assessed with microperimetry (MP) using deep learning (DL)-based automated quantification on optical coherence tomography (OCT) scans.

DESIGN

Cross-sectional study.

PARTICIPANTS

Twenty eyes of 20 subjects with clinically significant DME were included in this study.

METHODS

Patients with DME visible on OCT scans (Spectralis Heidelberg Retina Angiograph [HRA]+OCT) completed 2 MP examinations using a custom 45 stimuli grid on MAIA (CenterVue). MP stimuli were coregistered with the corresponding OCT location using image registration algorithms. DL-based algorithms were used to quantify intraretinal fluid (IRF) and ellipsoid zone (EZ) thickness. Hard exudates (HEs) were quantified semiautomatically. Multivariable mixed-effect models were calculated to investigate the association between DME-specific OCT features and point-wise RS. As EZ thickness values below HEs were excluded, the models included either EZ thickness or HEs.

RESULTS

A total of 1800 MP stimuli from 20 eyes of 20 patients were analyzed. Stimuli with IRF (n = 568) showed significantly decreased RS compared to areas without (estimate [95% CI]: -1.11 dB [-1.69, -0.52]; p = 0.0002). IRF volume was significantly negatively (-0.45 dB/nL [-0.71; -0.18]; p = 0.001) and EZ thickness positively (0.14 dB/µm [0.1; 0.19]; p < 0.0001) associated with localized point-wise RS. In the multivariable mixed model, including HE volume instead of EZ thickness, a negative impact on RS was observed (-0.43/0.1 nL [-0.81; -0.05]; p = 0.027).

CONCLUSIONS

DME-specific features, as analyzed on OCT, have a significant impact on point-wise RS. IRF and HE volume showed a negative and EZ thickness, a positive association with localized RS.

Plasma metabolomic profiling of diabetic macular edema

Diabetic macular edema (DME), a sight-threatening retinopathy, is a leading cause of vision loss in persons with diabetes mellitus. Despite strict control of systemic risk factors, a fraction of patients with diabetes developed DME, suggesting the existence of other potential pathogenic factors underlying DME. This study aimed to investigate the plasma metabotype of patients with DME and to identify novel metabolite markers for DME. Biomarkers identified from this study will provide scientific insight and new strategies for the early diagnosis and intervention of DME. To match clinical parameters between case and control subjects, patients with DME (DME, n = 30) or those with diabetes but without DME (Control, n = 30) were assigned to the present case-control study. Distinct metabolite profiles of serum were examined using liquid chromatography-mass spectrometry (LC-MS). A total of 190 distinct metabolites between DME and Control groups were identified (VIP > 1, Fold Change > 1.5 or < 0.667, and P < 0.05). The distinct metabolites between DME and Control groups were enriched in 4 KEGG pathways, namely, Glutamate Metabolism, Carnitine Synthesis, Oxidation of Branched Chain Fatty Acids, and Phytanic Acid Peroxisomal Oxidation. Finally, 4 metabolites were selected as candidate biomarkers for DME, namely, 5-Phospho-beta-D-ribosylamine, Succinic acid, Ascorbyl glucoside, and Glutathione disulfide. The area under the curve for these biomarkers were 0.693, 0.772, 0.762, and 0.771, respectively. This study suggested that impairment in the metabolism and 4 potential metabolites were identified as metabolic dysregulation associated with DME, which might provide insights into potential new pathogenic pathways for DME. 5-Phospho-beta-D-ribosylamine was first identified as a novel metabolite marker, with no previous reports linking it to diabetes or DME. This discovery may offer valuable insights into potential new pathogenic pathways associated with DME.

Unraveling the Role of Midkine in Proliferative Diabetic Retinopathy: Implications from Hypoxia-Induced Angiogenesis

Objectives

This study aimed to compare the expression of midkine (MK) in the vitreous of patients with proliferative diabetic retinopathy (PDR) and non-diabetic individuals, elucidating its potential role in the pathogenesis of the disease.

Methods

This prospective cross-sectional study included three groups of patients who underwent pars plana vitrectomy (PPV) surgery. The first group (control) consisted of patients who underwent PPV for epiretinal membrane and macular hole and did not have diabetes mellitus (DM). The second group included patients who underwent PPV for vitreous hemorrhage (VH) and tractional retinal detachment (TRD) secondary to PDR without prior anti-VEGF treatment (No preoperative anti-VEGF application: NPa-VEGF). The third group comprised patients who underwent PPV for VH and TRD secondary to PDR and received a preoperative anti-VEGF injection one week before surgery (preoperative anti-VEGF application: Pa-VEGF). Vitreous samples were collected intraoperatively, and the concentrations of MK, interleukin (IL)-6, and IL-8 were measured using specific Enzyme-Linked Immunosorbent Assay (ELISA) kits.

Results

The study included a total of 49 eyes from 49 patients undergoing PPV. The concentrations of IL-6 and IL-8 in vitreous samples from the NPa-VEGF group (n=15) and the Pa-VEGF group (n=14) were not significantly different compared to the control group (n=20) (p>0.05). However, the vitreous fluid of patients in the NPa-VEGF group exhibited significantly higher MK concentrations compared to the control group (p<0.007). Similarly, MK concentrations were significantly elevated in the Pa-VEGF group compared to the control group (p<0.046). No significant difference in MK levels was detected between the NPa-VEGF and Pa-VEGF groups (p>0.05).

Conclusion

These findings suggest that increased MK expression in the vitreous may be associated with the pathogenesis of PDR. Further studies are warranted to elucidate the precise mechanisms underlying this association and to explore the potential of MK as a therapeutic target for PDR management.

Predicting diabetic retinopathy based on routine laboratory tests by machine learning algorithms

OBJECTIVES

This study aimed to identify risk factors for diabetic retinopathy (DR) and develop machine learning (ML)-based predictive models using routine laboratory data in patients with type 2 diabetes mellitus (T2DM).

METHODS

Clinical data from 4259 T2DM inpatients at Beijing Tongren Hospital were analyzed, divided into a model construction data set (N = 3936) and an external validation data set (N = 323). Using 39 optimal variables, a prediction model was constructed using the eXtreme Gradient Boosting (XGBoost) algorithm and compared with four other algorithms: support vector machine (SVM), gradient boosting decision tree (GBDT), neural network (NN), and logistic regression (LR). The Shapley Additive exPlanation (SHAP) method was employed to interpret the XGBoost model. External validation was performed to assess model performance.

RESULTS

DR was present in 47.69% (N = 1877) of T2DM patients in the model construction data set. Among the models tested, the XGBoost model performed best with an AUC of 0.831, accuracy of 0.757, sensitivity of 0.754, specificity of 0.759, and F1-score of 0.752. SHAP explained feature importance for XGBoost model and identified key risk factors for DR. External validation yielded an accuracy of 0.650 for the XGBoost model.

CONCLUSIONS

The XGBoost-based prediction model effectively assesses DR risk in T2DM patients using routine laboratory data, aiding clinicians in identifying high-risk individuals and guiding personalized management strategies, especially in medically underserved areas.

Correlation between Systemic Inflammation and Morphological Changes of Retinal Neurovascular Unit in Patients with Early Signs of Diabetic Retinopathy: An OCT and OCT-Angiography Study

INTRODUCTION

The aim of the study was to investigate the correlation between systemic inflammation biomarkers and morphological changes of retinal neurovascular unit (RNVU) under optical coherence tomography (OCT) and OCT angiography (OCTA) in type 2 diabetic patients with early signs of diabetic retinopathy (DR).

METHODS

This cross-sectional study was carried out among 93 type 2 diabetic patients with early signs of DR (170 eyes), ranging from level 10 to level 35 based on ETDRS DR severity scale score. Age-, sex-, and axial length-matched normal subjects were enrolled as controls. Systemic inflammation biomarkers including neutrophil-to-lymphocyte ratio (NLR), monocyte-to-lymphocyte ratio (MLR), and systemic immune-inflammatory index (SII) were calculated based on peripheral blood results. Retinal neuronal changes of RNVU were identified by accessing the thickness of macular retinal nerve fiber layer (RNFL) and ganglion cell layer (GCL) using OCT. Retinal microvascular alterations of RNVU were evaluated by measuring macular vessel density (VD) and size of foveal avascular zone (FAZ) using OCTA.

RESULTS

GCL thickness was significantly correlated with NLR (r = -0.183, p = 0.017) and MLR (r = -0.235, p = 0.002), RNFL thickness was significantly associated with MLR (r = -0.210, p = 0.008), FAZp was positively correlated with NLR (r = 0.153, p = 0.046) and MLR (r = 0.187, p = 0.014), FAZa was positively correlated with MLR (r = 0.189, p = 0.014), and VD was significantly correlated with NLR (r = -0.188, p = 0.014) on spearman correlation analysis. Additionally, VD was independently associated with SII in both univariable and multivariable GLM analysis (p < 0.05). This difference still remained statistically significant during subgroup analysis after controlling DM duration.

CONCLUSION

Systemic inflammation biomarkers including NLR, MLR, and SII are significantly associated with not only retinal microvascular alterations but also retinal neuronal changes, providing evidence that systemic inflammation may play a crucial role on the early morphological changes of RNVU and early DR pathogenesis. SII is independently associated with VD, which supports SII may serve as a potential biomarker for monitoring early microvascular changes of DR.

Apolipoprotein and menopausal status are significant influencing factors for diabetic retinopathy in type II diabetes mellitus women

Diabetic retinopathy (DR) is a common complication of type II diabetes mellitus (T2DM) and a leading cause of blindness in the working population. Apolipoprotein levels have been reported to be associated with the risk of DR. This study aimed to develop a predictive model for DR based on apolipoproteins (apoA and apoB) and menopausal status in Chinese Han women with T2DM and to evaluate the model's effectiveness. Data from 2339 T2DM women were collected between January 2018 and June 2022. Multilevel regression was used to explore the independent effect of apolipoproteins and interaction between apolipoproteins and menopausal status on DR and proliferative diabetic retinopathy (PDR). Receiver operating characteristic (ROC) analysis was performed to compare the fitting degree and predictive efficiency of different models. Results showed that both apoA and apoB were independent influencing factors for DR and PDR and interacted significantly with menopausal status. The interaction between apoA and menopausal status had a protective effect on DR [OR (95% CI) = 0.925 (0.858-0.996), P = 0.040] and PDR [OR (95% CI) = 0.937 (0.895-0.981), P = 0.006]. In contrast, the interaction between apoB and menopausal status was a risk factor for DR [OR (95% CI) = 1.684 (1.141-2.379), P = 0.008)] and PDR [OR (95% CI) = 3.377 (1.148-9.937), P = 0.027]. ROC analysis demonstrated that the interaction model outperformed models without interaction terms (P < 0.01). The area under the curve for the interaction model was 0.879 (95% CI 0.864-0.893) for DR and 0.930 (95% CI 0.915-0.945) for PDR. These findings suggest that the interaction model is highly efficient in predicting DR, particularly PDR, in Chinese Han women with T2DM.

Non-ocular biomarkers for early diagnosis of diabetic retinopathy by non-invasive methods

Diabetic retinopathy (DR) is the predominant vision-threatening complication in individuals with diabetes mellitus. Timely diagnosis and intervention facilitate the prevention of diabetes-associated visual impairment. Classical imaging methods may prevent the timely detection of DR due to shortages of specialized facilities and retinal specialists, particularly in remote areas. In recent years, research on biomarkers related to DR has rapidly developed, playing an important role in risk assessment and early detection of the disease. Some ocular biomarkers from the vitreous body or aqueous humor were invasive, which hampered their application in clinical practice. Meanwhile, biomarkers based on omics were limited by their uneasily accessible use and complicated variables with a relatively low degree of reproducibility. As modern technology progresses, advanced non-ocular biomarkers of DR have established a comprehensive platform for the prompt identification of DR, independent of ophthalmic professionals or devices and accessible to non-ophthalmologists during community screenings. This review focuses on biomarkers derived from non-ocular sample sources, such as nailfold and skin, accessible through non-invasive methods, to reveal if they can be considered as an effective option for the early identification of DR by non-ophthalmologists in community screening initiatives.

Choroidal thickness as predictor of subclinical carotid atherosclerosis in adults with type 1 diabetes

Patients with Type 1 Diabetes (T1DM) have a higher risk of cardiovascular disease. This study used carotid ultrasound to identify subclinical carotid plaques and Optical Coherence Tomography (OCT) to evaluate ophthalmological markers as predictors of carotid plaque presence in 242 adults with T1DM, employing machine learning models for early risk assessment. Individuals with carotid plaques (N = 67) did not show significant differences in retinal nerve fiber layer (RNFL) and ganglion cell layer (GCL) and inner plexiform layer (IPL) complex compared to those without (N = 175). However, subfoveal and temporal choroidal area thickness significantly decreased in individuals with plaques (P ≤ 0.01). Machine learning identified age, hypertension, dyslipidemia, smoking, and diabetic retinopathy as key predictors for plaque presence, while ophthalmological measures made a modest contribution. Choroidal thickness exhibited an inverse relationship with plaque risk. Despite robust accuracy and high specificity (82-85% and 92-98%, respectively), the models were overly conservative in predicting positive instances (balanced accuracy of 0.60 for the left eye and 0.71 for the right eye). If further validated, choroidal thickness could complement traditional risk factors as an early marker of CV risk in T1DM patients. Integrating this measure in specialized clinical settings could help identify individuals who may need additional cardiovascular assessments.

Elevated Triglyceride-Glucose Index as a Risk Stratification Marker for Diabetic Retinopathy in Type 2 Diabetes Mellitus: The Influence of Glycemic Control

Background

Diabetic retinopathy (DR) is a common microvascular complication observed in people with type 2 diabetes mellitus (T2DM). The triglyceride-glucose (TyG) index, an indicator of insulin resistance, has an uncertain role in glycemic management in T2DM patients at risk for DR. This study aimed to evaluate the association between the TyG index and DR risk across different glycemic control status.

Methods

A total of 4,372 T2DM in-patients from Beijing Tongren Hospital (2013-2024) were included in this study. The patients were categorized into four groups based on TyG index quartiles (Q1-Q4). Participants were stratified by TyG index quartiles (Q1-Q4) and glycemic control status (HbA1c <7%, 7-9%, and >9%). Restricted cubic spline (RCS) analysis, logistic regression, and subgroup analyses were employed to explore the associations among the TyG index, glycemic control and DR risk.

Results

Overall, there were positive association of the TyG index and DR risk. The highest TyG Q4 was significantly associated with an increased risk of DR (odds ratio [OR]: 1.289; 95% confidence interval [CI]: 1.008-1.648, P<0.05), compared with the lowest TyG Q1. Significant interaction was observed between the TyG index and glycemic control (P for interaction <0.05). Notably, the association was strongest in patients with HbA1c >9%.

Conclusion

The relationship between high TyG index and DR risk was more pronounced in patients with HbA1c >9% group, indicating that TyG index may serve as a useful tool for making risk stratification on glycemic control in T2DM patients.