Rate and Predictors of Misclassification of Active Diabetic Macular Edema as Detected by an Automated Retinal Image Analysis System.

INTRODUCTION: The aim of this work is to estimate the sensitivity, specificity, and misclassification rate of an automated retinal image analysis system (ARIAS) in diagnosing active diabetic macular edema (DME) and to identify factors associated with true and false positives. METHODS: We conducted a cross-sectional study of prospectively enrolled patients with diabetes mellitus (DM) referred to a tertiary medical retina center for screening or management of DME. All patients underwent two-field fundus photography (macula- and disc-centered) with a true-color confocal camera; images were processed by EyeArt V.2.1.0 (Woodland Hills, CA, USA). Active DME was defined as the presence of intraretinal or subretinal fluid on spectral-domain optical coherence tomography (SD-OCT). Sensitivity and specificity and their 95% confidence intervals (CIs) were calculated. Variables associated with true (i.e., DME labeled as present by ARIAS + fluid on SD-OCT) and false positives (i.e., DME labeled as present by ARIAS + no fluid on SD-OCT) of active DME were explored. RESULTS: A total of 298 eyes were included; 92 eyes (31%) had active DME. ARIAS sensitivity and specificity were 82.61% (95% CI 72.37-89.60) and 84.47% (95% CI 78.34-89.10). The misclassification rate was 16%. Factors associated with true positives included younger age (p = 0.01), shorter DM duration (p = 0.006), presence of hard exudates (p = 0.005), and microaneurysms (p = 0.002). Factors associated with false positives included longer DM duration (p = 0.01), worse diabetic retinopathy severity (p = 0.008), history of inactivated DME (p < 0.001), and presence of hard exudates (p < 0.001), microaneurysms (p < 0.001), or epiretinal membrane (p = 0.06). CONCLUSIONS: The sensitivity of ARIAS was diminished in older patients and those without DME-related fundus lesions, while the specificity was reduced in cases with a history of inactivated DME. ARIAS performed well in screening for naïve DME but is not effective in surveillance inactivated DME.

Assessing Diabetic Retinopathy Staging With AI: A Comparative Analysis Between Pseudocolor and LED Imaging.

Purpose: To compare the diagnostic performance of artificial intelligence (AI)-based diabetic retinopathy (DR) staging system across pseudocolor, simulated white light (SWL), and light-emitting diode (LED) camera imaging modalities. Methods: A cross-sectional investigation involved patients with diabetes undergoing imaging with an iCare DRSplus confocal LED camera and an Optos confocal, ultra-widefield pseudocolor camera, with and without SWL. Macula-centered and optic nerve-centered 45 × 45-degree photographs were processed using EyeArt v2.1. Human graders established the ground truth (GT) for DR severity on dilated fundus exams. Sensitivity and weighted Cohen's weighted kappa (wκ) were calculated. An ordinal generalized linear mixed model identified factors influencing accurate DR staging. Results: The study included 362 eyes from 189 patients. The LED camera excelled in identifying sight-threatening DR stages (sensitivity = 0.83, specificity = 0.95 for proliferative DR) and had the highest agreement with the GT (wκ = 0.71). The addition of SWL to pseudocolor imaging resulted in decreased performance (sensitivity = 0.33, specificity = 0.98 for proliferative DR; wκ = 0.55). Peripheral lesions reduced the likelihood of being staged in the same or higher DR category by 80% (P < 0.001). Conclusions: Pseudocolor and LED cameras, although proficient, demonstrated non-interchangeable performance, with the LED camera exhibiting superior accuracy in identifying advanced DR stages. These findings underscore the importance of implementing AI systems trained for ultra-widefield imaging, considering the impact of peripheral lesions on correct DR staging. Translational Relevance: This study underscores the need for artificial intelligence-based systems specifically trained for ultra-widefield imaging in diabetic retinopathy assessment.

Rate and associations of inner retinal thinning in eyes with retinal vein occlusion and regressed macular oedema.

PURPOSE: To investigate the associations between visual acuity (VA) and structural optical coherence tomography (OCT) features in retinal vein occlusion (RVO) eyes after cystoid macular oedema (CMO) regression and to assess whether inner retinal thinning is progressive. METHODS: Retrospective observational study of RVO eyes with regressed CMO for at least 6 months. OCT scans at CMO regression were analysed, and features were correlated with VA at that visit. The inner retinal thickness was longitudinally compared between RVO and unaffected fellow eyes (controls) with linear mixed models. The rate of inner retinal thinning was obtained as the interaction term between disease status and time. Associations between inner retinal thinning and clinical characteristics were explored. RESULTS: Thirty-six RVO eyes were followed for 34.2 ± 21.1 months after CMO regression. The presence of ellipsoid zone disruption (regression estimate[standard error(SE)] = 0.16[0.04] LogMAR vs. intact, p < 0.001) and lower inner retinal thickness (regression estimate[SE] = -0.25[0.12] LogMAR for 100-µm increase, p = 0.01) were associated with worse VA. The inner retinal thickness decreased faster in RVO than controls (rate of retinal thinning -0.27 ± 0.09 µm/month vs. -0.08 ± 0.11 µm/month, p = 0.01). Macular ischaemia was associated with a faster rate of retinal thinning (interaction term macular ischaemia*follow-up time, p = 0.04). CONCLUSION: Inner retinal and photoreceptors' layers integrity are associated with better visual acuity once CMO resolves. RVO eyes undergo progressive inner retinal thinning after CMO regression, faster in eyes with macular ischaemia.

Fixation Location and Stability in Best Vitelliform Macular Dystrophy.

Purpose: To analyze fixation location and stability in best vitelliform macular dystrophy (BVMD) and test their association with best-corrected visual acuity (BCVA). Design: Observational, cross-sectional study. Participants: Thirty patients (55 eyes) affected by genetically confirmed BVMD were followed up at the Retinal Heredodystrophies Unit of IRCCS San Raffaele Scientific Institute, Milan. Methods: Patients underwent testing with macular integrity assessment (MAIA) microperimeter. Fixation location was measured as distance in degrees (°) between preferred retinal locus (PRL) and estimated fovea location (EFL); fixation was defined as eccentric when the distance between PRL and EFL exceeded 2°. Fixation stability was graded as stable, relatively unstable, or unstable and expressed as bivariate contour ellipse area (BCEA, °2). Main Outcome Measures: Fixation location and stability. Results: The median distance of the PRL from the anatomic fovea was 0.7°, and fixation location was eccentric in 27% of eyes. Fixation was graded as stable in 64% of eyes, relatively unstable in 13%, and unstable in 24%, with a median 95% BCEA of 6.2°2. The atrophic/fibrotic stage was associated with worse fixation parameters (all P < 0.01). Both PRL eccentricity and fixation stability were linearly associated with BCVA: every 1° increase in PRL eccentricity was associated with a 0.07 logarithm of the minimum angle of resolution (logMAR) worse BCVA (P < 0.0001) while every 1°2 increase in 95% BCEA was associated with a 0.01 logMAR worse BCVA (P < 0.001). No significant intereye correlation was found for PRL eccentricity and fixation stability, as well as no association between the patient's age and fixation parameters. Conclusions: We demonstrated that most eyes affected by BVMD retain a central stable fixation and provided evidence that both fixation eccentricity and stability are strongly associated with visual acuity in BVMD. These parameters may serve as secondary end points for future clinical trials. Financial Disclosures: Proprietary or commercial disclosure may be found after the references.

Peripheral and central capillary non-perfusion in diabetic retinopathy: An updated overview.

Capillary non-perfusion (CNP) is one of the key hallmarks of diabetic retinopathy (DR), which may develop both in the periphery and at the posterior pole. Our perspectives on CNP have extended with the introduction of optical coherence tomography angiography (OCTA) and ultra-widefield imaging, and the clinical consequences of peripheral and macular CNP have been well characterized. Fluorescein angiography (FA) continues to be the gold standard for detecting and measuring CNP, particularly when ultra-widefield imaging is available. OCTA, on the other hand, is a quicker, non-invasive approach that allows for a three-dimensional examination of CNP and may soon be regarded as an useful alternative to FA. In this review, we provide an updated scenario regarding the characteristics, clinical impact, and management of central and peripheral CNP in DR.

Tocilizumab in Juvenile Idiopathic Arthritis Associated Uveitis, a Narrative Review.

Juvenile idiopathic arthritis (JIA) associated uveitis (JIA-U) is the most common extra-articular manifestation of JIA, affecting 10-15% of patients, especially in oligoarticular JIA where its course may be faint. Therefore, JIA-U is one of the most challenging pediatric uveitis, associated with major ocular morbidity and possibly leading to irreversible structural ocular damage and to vision-threatening complications. Adequate management is crucial for avoiding visual impairment complications. Since the introduction of biologic disease modifying anti-rheumatic drugs (bDMARDS), the visual prognosis of JIA-U has dramatically improved over the decades. Tumor necrosis factor-α (TNF-α) blockers are the most used bDMARDs in treating JIA-U with large evidence of efficacy. However, inadequate response to these agents, either due to intolerance or inefficacy, may be observed, requiring a swap to other classes of immunosuppressive agents, including anti-IL-6, anti-CD20, and, more recently, JAK inhibitors. Tocilizumab is a humanized monoclonal antibody to the interelukin-6 receptor preventing IL-6 from binding to its soluble and membrane-bound receptors. A growing body of literature provides promising results about the efficacy of intravenous and subcutaneous tocilizumab in the treatment of JIA-U. A narrative review of the literature on this topic will improve our knowledge on the potential use of tocilizumab in JIA-U.

FUNDUS AUTOFLUORESCENCE IN EXTENSIVE MACULAR ATROPHY WITH PSEUDODRUSEN AND DIFFUSE TRICKLING GEOGRAPHIC ATROPHY.

PURPOSE: To establish whether extensive macular atrophy with pseudodrusen (EMAP) can be distinguished from the diffuse-trickling phenotype of geographic atrophy (DTGA) secondary to age-related macular degeneration on the basis of its features on blue-light autofluorescence. METHODS: The authors reviewed our prospectively maintained database to enroll patients with a diagnosis of EMAP, DTGA, and non-DTGA with a minimum follow-up of 1 year. Atrophic areas and growth rates were measured on blue-light autofluorescence images, using the Heidelberg Region Finder tool. Circularity and roundness were chosen as atrophy shape descriptors, extracted using ImageJ, and compared between disease groups. RESULTS: A total of 28 EMAP, 27 DTGA, and 30 non-DTGA eyes were included in the analysis. The median follow-up time was around 3.5 years. Extensive macular atrophy with pseudodrusen was characterized by an irregular and elongated shape (low circularity and low roundness) and associated with a fast atrophy growth rate (3.6 mm 2 /year), compared with non-DTGA. However, these parameters were not significantly different between EMAP and DTGA. CONCLUSION: Our study found that EMAP and DTGA cannot be effectively differentiated on fundus autofluorescence. In both diseases, the macular atrophic area has a major vertical axis, fringed borders, and fast progression.

Incidence and Risk Factors of Visual Impairment in Patients with Angioid Streaks and Macular Neovascularization.

OBJECTIVE: To estimate the incidence and risk factors of visual impairment and complications in eyes with macular neovascularization (MNV) because of angioid streaks (ASs). DESIGN: Longitudinal multicenter retrospective cohort study. SUBJECTS: Patients with AS-associated MNV treated with anti-VEGF agents and a follow-up of > 3 months. METHODS: Clinical and MNV characteristics were collected at baseline. Visual acuity (VA) values and the presence of atrophy or fibrosis were collected at each visit. MAIN OUTCOME MEASURES: Rate of VA change over time and associated factors; the incidence rate of moderate-to-severe visual impairment (MSVI) and blindness and hazard ratio (HR) of candidate risk factors for MSVI; the incidence rate of fibrosis and macular atrophy. RESULTS: Overall, 84 eyes of 66 patients (39 men, 58%) with a mean (standard deviation) age of 55.7 (13.8) years were followed for a mean (standard deviation) of 67.7 (48.5) months. The median number of anti-VEGF doses per eye was 13. The average rate (95% confidence interval [CI]) of visual loss was +0.04 (0.02-0.06) logarithm of the minimum angle of resolution/year (P < 0.001); the visual loss was faster in nonnaive eyes (P = 0.007) and those with better baseline VA (P < 0.001); it was slower in eyes with pattern dystrophy-like features (P = 0.04). The incidence rates (95% CI) of MSVI and blindness were 10.4 (6.88-15)/100-eye-years and 2.33 (1.12-4.29)/100-eye-years. A higher number of injections (HR [95% CI] = 0.45 [0.19-0.94] for receiving ≥ 13 injections vs. < 13; P = 0.03) was protective against MSVI. The incidence rates (95% CI) of fibrosis and macular atrophy were 24.1 (17.5-32.3)/100-eye-years and 14.3 (10.1-19.6)/100-eye-years. CONCLUSIONS: Eyes with MNV-related AS had a high rate of visual impairment and propensity to macular fibrosis and atrophy. A higher number of injections yielded better chances of maintaining good VA, suggesting the need for intensive treatment. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found after the references.

RISK FACTORS OF VISION LOSS AND MULTIPLE RECURRENCES IN MYOPIC MACULAR NEOVASCULARIZATION.

PURPOSE: To investigate the factors associated with maximum visual improvement (peak vision) gain and the risk factors of peak vision loss and multiple recurrences in myopic macular neovascularization undergoing antivascular endothelial growth factor therapy. METHODS: Retrospective study of 310 eyes with active myopic macular neovascularization and median follow-up of 3.5 years. We defined peak vision gain as the maximum best-corrected visual acuity value reached under treatment and peak vision loss as best-corrected visual acuity never scoring as peak vision. We used multiple-event Prentice, Williams, and Peterson models to compute recurrences' incidence and Cox regression to identify risk factors for peak vision gain, peak vision loss, and multiple recurrences. RESULTS: Eyes with worse baseline best-corrected visual acuity {hazard ratio (HR) = 2.59 (95% confidence interval [CI]: 1.63-4.11) for 0.1 logMAR increase, P < 0.001} had higher chance to achieve peak vision. Peak vision was lost in 162 eyes (52%). Older age (HR = 1.22 [95% CI: 1.02-1.43] for 10-year increase, P = 0.02) and recurrences (HR = 1.10 [95% CI: 1.01-1.22] for event, P = 0.04) predicted nonsustained peak vision. Older age (HR = 1.13 [95% CI: 1.04-1.27] for 10-year increase, P = 0.006), larger myopic macular neovascularization (HR = 1.06 [95% CI: 1.01-1.13] for 1-mm 2 increase, P = 0.04), and juxtafoveal location (HR = 1.88 [95% CI: 1.28-2.77] vs. extrafoveal, P = 0.001) predicted multiple recurrences. CONCLUSION: Myopic macular neovascularization eyes lose vision mainly because of multiple recurrences. Patients at risk for recurrences should undergo more attentive monitoring to avoid vision loss.

Long-term Incidence and Risk Factors of Macular Fibrosis, Macular Atrophy, and Macular Hole in Eyes with Myopic Neovascularization.

PURPOSE: To identify the risk factors associated with myopic macular neovascularization (mMNV)-related complications in patients treated with intravitreal anti-VEGF agents. DESIGN: Longitudinal cohort study. PARTICIPANTS: Myopic eyes (n = 313) with active mMNV and median (interquartile range) follow-up of 42 months (interquartile range, 18-68 months) after initiation of anti-VEGF treatment. METHODS: Data regarding patients' clinical and mMNV-related characteristics were collected at baseline. Subsequent OCT scans were inspected for mMNV-related complications. Best-measured visual acuity (BMVA) values were retrieved from each visit. MAIN OUTCOME MEASURES: Incidence rate and hazard ratio (HR, with 95% confidence interval [CI]) of risk factors for fibrosis and macular atrophy calculated with Kaplan-Meier curves and Cox regression models. Crude incidence of macular hole (MH). Longitudinal BMVA changes. RESULTS: Five-year incidence of fibrosis, atrophy, and MH were 34%, 26%, and 8%, respectively. The rate of fibrosis was 10.3 (95% CI, 8.25-12.6) per 100 person-years. Risk factors were subfoveal mMNV location (HR [95% CI] = 12.7 [2.70-56.7] vs. extrafoveal, P = 0.001) and intraretinal fluid at baseline (HR [95% CI] = 1.75 [1.05-2.98], P = 0.03). The rate of macular atrophy was 6.5 (95% CI, 5-8.3) per 100 person-years. Risk factors were diffuse (HR, 2.20 vs. tessellated fundus; 95% CI, 1.13-5.45; P = 0.02) or patchy chorioretinal atrophy (HR, 3.17 vs. tessellated fundus; 95% CI, 1.32-7.64; P = 0.01) at baseline and more numerous anti-VEGF injections before baseline (HR, 1.21; 95% CI, 1.06-1.38 for each treatment; P = 0.005). Eyes with fibrosis and macular atrophy had faster BMVA decay over follow-up. Twenty eyes (6%) developed MH. Two subtypes of MH were identified: "atrophic" and "tractional." CONCLUSIONS: Myopic MNV-related complications are common in the long term despite initially successful treatment and have detrimental effects on visual acuity. Insights into their incidence and risk factors may help for future treatments to mitigate sight-threatening outcomes.