Search

Objective: With the increasing prevalence of diabetes, annual screening for diabetic retinopathy (DR) by expert human grading of retinal images is challenging. Automated DR image assessment systems (ARIAS) may provide clinically effective and cost-effective detection of retinopathy. We aimed to determine whether ARIAS can be safely introduced into DR screening pathways to replace human graders., Design: Observational measurement comparison study of human graders following a national screening program for DR versus ARIAS., Participants: Retinal images from 20 258 consecutive patients attending routine annual diabetic eye screening between June 1, 2012, and November 4, 2013., Methods: Retinal images were manually graded following a standard national protocol for DR screening and were processed by 3 ARIAS: iGradingM, Retmarker, and EyeArt. Discrepancies between manual grades and ARIAS results were sent to a reading center for arbitration., Main Outcome Measures: Screening performance (sensitivity, false-positive rate) and diagnostic accuracy (95% confidence intervals of screening-performance measures) were determined. Economic analysis estimated the cost per appropriate screening outcome., Results: Sensitivity point estimates (95% confidence intervals) of the ARIAS were as follows: EyeArt 94.7% (94.2%-95.2%) for any retinopathy, 93.8% (92.9%-94.6%) for referable retinopathy (human graded as either ungradable, maculopathy, preproliferative, or proliferative), 99.6% (97.0%-99.9%) for proliferative retinopathy; Retmarker 73.0% (72.0 %-74.0%) for any retinopathy, 85.0% (83.6%-86.2%) for referable retinopathy, 97.9% (94.9%-99.1%) for proliferative retinopathy. iGradingM classified all images as either having disease or being ungradable. EyeArt and Retmarker saved costs compared with manual grading both as a replacement for initial human grading and as a filter prior to primary human grading, although the latter approach was less cost-effective., Conclusions: Retmarker and EyeArt systems achieved acceptable sensitivity for referable retinopathy when compared with that of human graders and had sufficient specificity to make them cost-effective alternatives to manual grading alone. ARIAS have the potential to reduce costs in developed-world health care economies and to aid delivery of DR screening in developing or remote health care settings., (Copyright © 2016 American Academy of Ophthalmology. All rights reserved.)

Purpose: Most classification systems for age-related macular degeneration (AMD) were developed from patients in clinical trials. We aimed to validate the Age-Related Eye Diseases Study (AREDS) simplified severity scale of AMD classification using 5- and 10-year incident late AMD data from the population-based Blue Mountains Eye Study (BMES) cohort., Design: Comparative study of population-based cohort and clinical trial., Participants: Blue Mountains Eye Study participants 40 to 97 years of age at baseline (n = 2134) and AREDS participants 55 to 80 years of age (n = 3640)., Methods: In the BMES, AMD lesions were graded from stereoscopic color photographs and were classified according to the AREDS simplified severity scale. The AREDS simplified scale calculates a risk score based on the number of early AMD risk factors (large drusen and pigment abnormalities) in both eyes that can range from 0 to 4., Main Outcome Measures: Five- and 10-year incident late AMD (presence of geographic atrophy or choroidal neovascularization)., Results: The AREDS simplified scale performed similarly when applied to both the BMES population-based participants and the AREDS clinical trial-based participants in predicting 5- and 10-year incidence of late AMD. For scores 0 to 4, the 5-year incidence rates for the BMES compared with the AREDS were 0.2% versus 0.4%, 3.1% versus 3.1%, 12.1% versus 11.8%, 13.5% versus 25.9%, and 47.1% versus 47.3%, respectively. The corresponding 10-year incidence rates for the BMES compared with the AREDS were 0.7% versus 1.5%, 7.3% versus 8.4%, 36.6% versus 27.6%, 20.0% versus 52.7%, and 75.0% versus 71.4%, respectively., Conclusions: The AREDS simplified severity scale classified late AMD risk levels similarly when applied to population-based and clinical trial samples. These results support the robustness of the AREDS simplified severity scale., (Copyright © 2016 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Purpose: To study the characteristics of second treated eyes in patients with neovascular age-related macular degeneration (nAMD) treated with ranibizumab in the United Kingdom National Health Service., Design: Multicenter national nAMD database study., Participants: Twelve thousand nine hundred fifty-one treatment-naïve eyes of 11,135 patients receiving 92,976 ranibizumab injections., Methods: Up to 5 years of routinely collected, anonymized data within electronic medical record systems were extracted remotely from 14 centers. Participating centers exclusively used ranibizumab to treat nAMD (loading phase of 3 monthly injections followed by monthly visits and pro re nata re-treatment). The minimum data set included: age, logarithm of the minimum angle of resolution (logMAR) visual acuity (VA) at baseline and at all subsequent visits, and injection episodes., Main Outcome Measures: Baseline, change and actual VA over 3 years, and number of treatments and clinic visits., Results: During the study, 1816 (16.3%) of the 11 135 patients received treatment to the fellow eye. Mean baseline and final VA were 0.66 (standard deviation, 0.32) and 0.65 (0.40) for first treated eyes and 0.41 (0.34) and 0.56 (0.40) for second treated eyes. The rate of VA loss after the loading phase was similar in first and second treated eyes (0.03 and 0.05 logMAR units/year). When fellow eyes with baseline VA worse than 20/200 were excluded to restrict analyses to eyes at risk of nAMD, the rate of second-eye involvement was 14.0% per year (42%/3 years). Mean number of injections/visits in years 1, 2, and 3 were similar for first and second treated eyes (5.6/8.2, 3.9/8.0, 3.8/8.2 and 5.5/8.7, 3.6/9.4, and 3.8/9.1, respectively)., Conclusions: Second treated eyes with nAMD commence treatment with better baseline VA, do not show significant vision gain but maintain better VA than first treated eyes at all time points for at least 3 years, making them the more important eye functionally. These data highlight the high burden of second eye involvement, with almost half of all eyes at risk requiring bilateral treatment by 3 years, and the need for regular monitoring of fellow eyes for best visual outcomes which theoretically may reduce the benefits of extended monitoring regimens., (Copyright © 2014 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Objective: To provide a detailed phenotype/genotype characterization of Bietti crystalline dystrophy (BCD)., Design: Observational case series., Participants: Twenty patients from 17 families recruited from a multiethnic British population., Methods: Patients underwent color fundus photography, near-infrared (NIR) imaging, fundus autofluorescence (FAF) imaging, spectral domain optical coherence tomography (SD-OCT), and electroretinogram (ERG) assessment. The gene CYP4V2 was sequenced., Main Outcome Measures: Clinical, imaging, electrophysiologic, and molecular genetics findings., Results: Patients ranged in age from 19 to 72 years (median, 40 years), with a visual acuity of 6/5 to perception of light (median, 6/12). There was wide intrafamilial and interfamilial variability in clinical severity. The FAF imaging showed well-defined areas of retinal pigment epithelium (RPE) loss that corresponded on SD-OCT to well-demarcated areas of outer retinal atrophy. Retinal crystals were not evident on FAF imaging and were best visualized with NIR imaging. Spectral domain OCT showed them to be principally located on or in the RPE/Bruch's membrane complex. Disappearance of the crystals, revealed by serial recording, was associated with severe disruption and thinning of the RPE/Bruch's membrane complex. Cases with extensive RPE degeneration (N = 5) had ERGs consistent with generalized rod and cone dysfunction, but those with more focal RPE atrophy showed amplitude reduction without delay (N = 3), consistent with restricted loss of function, or that was normal (N = 2). Likely disease-causing variants were identified in 34 chromosomes from 17 families. Seven were novel, including p.Met66Arg, found in all 11 patients from 8 families of South Asian descent. This mutation appears to be associated with earlier onset (median age, 30 years) compared with other substitutions (median age, 41 years). Deletions of exon 7 were associated with more severe disease., Conclusions: The phenotype is highly variable. Several novel variants are reported, including a highly prevalent substitution in patients of South Asian descent that is associated with earlier-onset disease. Autofluorescence showed sharply demarcated areas of RPE loss that coincided with abrupt edges of outer retinal atrophy on SD-OCT; crystals were generally situated on or in the RPE/Bruch's complex but could disappear over time with associated RPE disruption. These results support a role for the RPE in disease pathogenesis., (Copyright © 2014 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Objective: There are limited data on the natural history and longitudinal changes of retinal microvascular lesions. We examined 10-year changes in retinal microvascular lesions, focusing on those related to hypertension and shown to predict development of cardiovascular disease., Design: Prospective cohort., Participants: We included 1120 middle-aged participants without diabetes of the Atherosclerosis Risk in Communities (ARIC) Study in 1993 to 1995 and again 10 years later in 2003 to 2005., Methods: Retinal microvascular lesions were graded from retinal photographs using the same protocol at both examinations, with changes (incidence or disappearance) adjudicated by a side-by-side comparison of photographs. The study sample was stratified by carotid intima media thickness (IMT) and ARIC field center; thus, all analyses were weighted by these factors. Persons with diabetes were excluded because the frequency and pathophysiology of diabetic retinal lesions is different., Main Outcome Measures: Incidence and disappearance rates of lesions., Results: The 10-year incidence of focal arteriolar narrowing, arteriovenous (AV) nicking, and retinopathy in persons without diabetes was 3.4% (95% confidence interval [CI], 2.3-4.9), 2.5% (95% CI, 1.6-3.9), and 2.2% (95% CI, 1.3-3.5) respectively. Over the 10-year period, of 32, 219, and 24 eyes with focal arteriolar narrowing, AV nicking and retinopathy at baseline, 50.3% (95% CI, 28.6-71.9), 40.7% (95% CI, 32.7-49.4), and 65.9% (95% CI, 42.4-83.5), respectively, disappeared. Higher baseline plasma fibrinogen and white cell counts were associated with incident focal arteriolar narrowing; antihypertensive medication use was associated with incident AV nicking, and higher diastolic blood pressure, carotid IMT, and white cell counts were associated with incident retinopathy. Higher fasting serum glucose was not significantly associated with incident retinopathy, although this may be related to the small number of lesions (odds ratio, 5.88; 95% CI, 0.74-46.64 per standard deviation difference)., Conclusions: In this sample of middle-aged adults, new retinal microvascular lesions appeared at a rate between 2% and 4% over 10 years. A high percentage of lesions (≥40%) disappeared over the same period, suggesting considerable remodeling in the retinal microvasculature., Financial Disclosure(s): The authors have no proprietary or commercial interest in any of the materials discussed in this article., (Copyright © 2011 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Purpose: To describe the prevalence and associations of retinopathy in a population-based nondiabetic sample of rural Chinese., Design: Population-based cross-sectional study., Participants: We included 6830 Han Chinese aged > or =30 years from 13 villages of Yongnian County, Handan City, Hebei Province, China., Methods: All participants underwent a standardized interview and extensive examinations including retinal photography, measurement of blood pressure (BP) and fasting plasma glucose (FPG). Diabetes mellitus was defined as either FPG > or = 7.0 mmol/l, use of diabetic medication or a physician diagnosis of diabetes. Photographic grading of retinopathy followed the modified Early Treatment Diabetic Retinopathy Study classification system. Logistic regression models were used to assess associations of retinopathy., Main Outcome Measures: Any retinopathy., Results: The prevalence of retinopathy among participants without diabetes was 13.6% (95% confidence interval [CI], 12.6-14.6%). The age and gender standardized prevalence of retinopathy in the Chinese adult population (aged 30+ years) without diabetes was estimated to be 12.1% (95% CI 11.1-12.9%). Independent risk factors associated with retinopathy were age (odds ratio [OR], 1.02 ; 95% CI 1.01-1.03 per year increase), male gender (male vs. female, OR 1.27; 95% CI 1.08-1.49), higher FPG (OR 1.30; 95% CI 1.11-1.53 per mmol/l increase), higher systolic BP (OR 1.15; 95% CI 1.05-1.27 per 10 mmHg increase) and higher diastolic BP (OR 1.16; 95% CI 1.09-1.22 per 10 mmHg increase)., Conclusions: Retinopathy was common among rural Chinese adults without diabetes. Its association with FPG and BP suggests that early microvascular damage is occurring at "high normal" levels of blood glucose and BP., (Copyright 2010 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Purpose: To assess the relationship of birth weight, birth length, and head circumference as proxy markers of intrauterine growth, cup/disc ratio, and other optic disc parameters measured using optical coherence tomography (OCT)., Design: Population-based cross sectional analysis., Participants: The Sydney Childhood Eye Study examined 2353 primarily 12-year-old children from 21 randomly selected secondary schools during 2003 to 2005., Methods: Of 2353 children examined, 2134 (90.7%) had OCT scans (Zeiss Stratus OCT, Carl Zeiss Meditec, Dublin, CA) and are included in this study. The "fast" optic disc scan protocol was used. Birth weight, birth length, and head circumference were ascertained from health records. Height and weight were measured using standardized protocols, body mass index (BMI) was defined as weight (kilograms)/ height squared (meters), and sociodemographic information was collected in a questionnaire completed by parents. Low birth weight was defined as birth weight

Objective: Fractals represent a type of derived geometric pattern that permits the characterization of the branching pattern of retinal vessels. We examined a new semiautomated method to measure retinal vessel fractals., Design: Methodology study., Participants: Three hundred randomly selected participants from the population-based Blue Mountains Eye Study., Methods: We developed a semiautomated computer program to measure the fractal dimension (D(f)) of the retinal vessels from digitized images of disk-centered retinal photographs. Two trained graders masked to participant characteristics measured D(f) of right eye images of participants. Reliability was determined by repeat grading of the images from 60 participants, and association with systolic and diastolic blood pressure was examined in all 300 participants., Main Outcome Measure: D(f) of the retinal vessels., Results: Mean D(f) was 1.437 with a standard deviation of 0.025. Intragrader and intergrader reliability estimates were high with intraclass correlation ranging from 0.93 to 0.95. D(f) was inversely correlated with age (r = -0.42, P = 0.001) and systolic blood pressure (r = -0.29, P<0.0001). After adjustment for age and sex, mean D(f) was significantly lower in participants with than without hypertension (D(f) difference 0.01, P = 0.02)., Conclusions: The D(f) of the retinal vessels can be reliably measured from photographs and shows a strong inverse correlation with blood pressure. These data suggest that the D(f) may be a measure of early microvascular alterations from elevated blood pressure. Further studies to examine the systemic and ocular correlates of the D(f) of the retinal vessels are needed.

Objective: To examine whether retinal vessel signs are independent predictors of the long-term development of age-related maculopathy (ARM)., Design: Prospective population-based cohort study., Participants: Blue Mountains Eye Study participants aged > or =49 years (n = 3654) were examined during 1992 through 1994; 2335 (75% of survivors) were reexamined after 5 years and 1952 (76% of survivors) were reexamined after 10 years., Methods: Baseline focal arteriolar narrowing and arteriovenous (AV) nicking were assessed and vessel calibers were measured from retinal photographs. A side-by-side grading method was used to assess ARM incidence. Eye-specific data were analyzed using generalized estimating equation models, adjusting for age, gender, smoking, and blood pressure., Main Outcome Measures: Incident early ARM (soft indistinct or reticular drusen or combined soft distinct drusen and retinal pigment abnormality) was defined in eyes free of both early and late ARM at baseline. Incident late ARM (either geographic atrophy or neovascular macular degeneration) was defined in eyes free of these 2 lesions at baseline., Results: Over a 10-year period, incident late ARM developed in 106/4745 eyes at risk of late ARM (2.2%). Eyes with mild (2.7%) or severe (4.6 %) AV nicking were more likely to develop late ARM than eyes without this sign (1.5%). The adjusted odds ratio (OR) was 1.4, 95% confidence interval (CI) 0.8 to 2.4 for mild and OR 2.6, 95% CI 1.2 to 5.5 for severe AV nicking. Eyes with focal narrowing were also more likely to develop late ARM (7.6% vs. 1.8%), adjusted OR 2.2, 95% CI 1.1 to 4.1. Incident early ARM developed in 398/4490 eyes at risk of early ARM (8.9%). Severe AV nicking, but not focal arteriolar narrowing, was associated with an increased long-term risk of early ARM (13.6% vs. 8.2%; OR 1.5; 95% CI 1.0-2.3). Neither arteriolar nor venular caliber was significantly associated with the incidence of either early or late ARM., Conclusions: These 10-year incidence data confirm our previous observation that structural retinal arteriolar changes may either contribute to ARM progression or share common pathologic pathways with ARM, independent of traditional vascular risk factors.