Optic atrophy after anti-vascular endothelial growth factor injection in diabetic patients with proliferative diabetic retinopathy.

OBJECTIVE: To study the prevalence of optic atrophy in patients with proliferative diabetic retinopathy (PDR) who underwent intravitreal bevacizumab injection and risk factors associated with optic atrophy. MATERIAL AND METHOD: A retrospective case control study enrolled 269 cases (394 eyes) of patients with PDR, in which 166 cases (219 eyes) received intravitreal bevacizumab injection. Associated factors such as type of DM, hemoglobin A1c level, hypertension, hypercholesterolemia, chronic kidney disease, previous intravitreal surgery retinal detachment, and vitreous hemorrhage were recorded. Criteria for diagnosis of optic atrophy were decreased visual acuity, pale optic disc and decreased nerve fiber layer thickness, which was measured by Stratus optical coherence tomography (OCT). The association between intravitreal bevacizumab injection and optic atrophy was analyzed by multiple logistic regression. RESULTS: Two hundred sixty nine patients with PDR, consisting of 166 patients with intravitreal bevacizumab injection and 103 cases without bevacizumab injection. Optic atrophy was found in 11.4% (25/219 eyes) and 8% (14/175 eyes) respectively. There was no evidence that intravitreal bevacizumab injection and associated systemic diseases were related to optic atrophy. The risk factor that was related to optic atrophy was previous intravitreal surgery (adjusted odds ratio (OR), 2.57 [95% CI, 1.13, 5.84], p = 0.024). CONCLUSION: Anti-VEGF (bevacizumab) does not increase the risk of optic atrophy. The ophthalmologists should be aware of subsequent optic atrophy development in patients with PDR who undergo surgical intervention.

Accuracy of the postnatal growth and retinopathy of prematurity screening criteria in predicting prethreshold retinopathy of prematurity in the tertiary hospital, Bangkok, Thailand.

PURPOSE: The postnatal growth and retinopathy of prematurity (G-ROP), retinopathy of prematurity (ROP) predictive model, was developed in North America with high sensitivity and fewer infants examined. This study aimed to validate this model in Thai infants by assessing sensitivity and comparing it to the current American Academy of Ophthalmology (AAO) screening guideline. MATERIALS AND METHODS: The records of infants screened for ROP were retrospectively reviewed from 2015 to 2020. G-ROP model was applied to calculate sensitivity for prethreshold type 1 and 2 ROP and the reduction of the number of infants examined. RESULTS: Of 129 infants screened, there were 102 infants who met G-ROP criteria. The mean gestational age at birth was 29.7 ± 2.7 weeks. The mean birth weight was 1177.8 ± 401.3 g. Both G-ROP and AAO detected prethreshold type 1 ROP in 24 of 24 infants (sensitivity, 100%; 95% confidence interval [CI], 85.8%-100%). Furthermore, they detected all four infants prethreshold type 2 ROP with 100% of sensitivity (95% CI, 39.8-100.0). The reduction in infants receiving examinations using G-ROP was 20.9%. CONCLUSIONS: G-ROP model provided high sensitivity and lessen unnecessary examinations for ROP screening in Thai infants.

Factors associated with the development of blepharoptosis after pars plana vitrectomy surgery.

PURPOSE: To determine the incidence and risk factors for developing blepharoptosis after vitrectomy surgery. METHODS: This prospective observational study conducted in patients who had vitrectomy surgery at the University of California, San Francisco. The patients' eyelids were photographed before, surgery and follow-up visits at 1 day, 1 week, 1 month, 3 months, and 6 months after surgery. Levator excursion (LE), palpebral fissure (PF) height and marginal reflex distance 1 (MRD1) changes from baseline were evaluated. Operative parameters associated with lid parameter changes were analyzed. RESULTS: Thirty-eight eyes were enrolled in the study. Multivariate analysis performed for PF changes from pre-operative were -0.47 mm, 0.33 mm, and 0.09 mm at 1 week, 1 month and 6 months after surgery, respectively (p-value 0.1, 0.2, and 0.8). The mean change of LE from preoperative was -0.44 mm, -0.15 mm, and 0.35 mm at 1 week, 1 month and 6 months after surgery, respectively (p-value 0.3, 0.7 and 0.4). The reduction of MRD1 at 1 week, 1 month and 6 months after surgery were -0.08 mm, -0.13 mm and -0.01 mm, respectively (p = .003, p = .6 and 0.9). Triamcinolone usage was associated with reduction of MRD1 and LE. CONCLUSION: Blepharoptosis presents most during the first week after surgery. The possibility of developing transient changes in eyelid position after vitrectomy surgery should be discussed with patients.

Real-time diabetic retinopathy screening by deep learning in a multisite national screening programme: a prospective interventional cohort study.

BACKGROUND: Diabetic retinopathy is a leading cause of preventable blindness, especially in low-income and middle-income countries (LMICs). Deep-learning systems have the potential to enhance diabetic retinopathy screenings in these settings, yet prospective studies assessing their usability and performance are scarce. METHODS: We did a prospective interventional cohort study to evaluate the real-world performance and feasibility of deploying a deep-learning system into the health-care system of Thailand. Patients with diabetes and listed on the national diabetes registry, aged 18 years or older, able to have their fundus photograph taken for at least one eye, and due for screening as per the Thai Ministry of Public Health guidelines were eligible for inclusion. Eligible patients were screened with the deep-learning system at nine primary care sites under Thailand's national diabetic retinopathy screening programme. Patients with a previous diagnosis of diabetic macular oedema, severe non-proliferative diabetic retinopathy, or proliferative diabetic retinopathy; previous laser treatment of the retina or retinal surgery; other non-diabetic retinopathy eye disease requiring referral to an ophthalmologist; or inability to have fundus photograph taken of both eyes for any reason were excluded. Deep-learning system-based interpretations of patient fundus images and referral recommendations were provided in real time. As a safety mechanism, regional retina specialists over-read each image. Performance of the deep-learning system (accuracy, sensitivity, specificity, positive predictive value [PPV], and negative predictive value [NPV]) were measured against an adjudicated reference standard, provided by fellowship-trained retina specialists. This study is registered with the Thai national clinical trials registry, TCRT20190902002. FINDINGS: Between Dec 12, 2018, and March 29, 2020, 7940 patients were screened for inclusion. 7651 (96·3%) patients were eligible for study analysis, and 2412 (31·5%) patients were referred for diabetic retinopathy, diabetic macular oedema, ungradable images, or low visual acuity. For vision-threatening diabetic retinopathy, the deep-learning system had an accuracy of 94·7% (95% CI 93·0-96·2), sensitivity of 91·4% (87·1-95·0), and specificity of 95·4% (94·1-96·7). The retina specialist over-readers had an accuracy of 93·5 (91·7-95·0; p=0·17), a sensitivity of 84·8% (79·4-90·0; p=0·024), and specificity of 95·5% (94·1-96·7; p=0·98). The PPV for the deep-learning system was 79·2 (95% CI 73·8-84·3) compared with 75·6 (69·8-81·1) for the over-readers. The NPV for the deep-learning system was 95·5 (92·8-97·9) compared with 92·4 (89·3-95·5) for the over-readers. INTERPRETATION: A deep-learning system can deliver real-time diabetic retinopathy detection capability similar to retina specialists in community-based screening settings. Socioenvironmental factors and workflows must be taken into consideration when implementing a deep-learning system within a large-scale screening programme in LMICs. FUNDING: Google and Rajavithi Hospital, Bangkok, Thailand. TRANSLATION: For the Thai translation of the abstract see Supplementary Materials section.

Erratum: Author Correction: Deep learning versus human graders for classifying diabetic retinopathy severity in a nationwide screening program.

[This corrects the article DOI: 10.1038/s41746-019-0099-8.].

Deep learning versus human graders for classifying diabetic retinopathy severity in a nationwide screening program.

Deep learning algorithms have been used to detect diabetic retinopathy (DR) with specialist-level accuracy. This study aims to validate one such algorithm on a large-scale clinical population, and compare the algorithm performance with that of human graders. A total of 25,326 gradable retinal images of patients with diabetes from the community-based, nationwide screening program of DR in Thailand were analyzed for DR severity and referable diabetic macular edema (DME). Grades adjudicated by a panel of international retinal specialists served as the reference standard. Relative to human graders, for detecting referable DR (moderate NPDR or worse), the deep learning algorithm had significantly higher sensitivity (0.97 vs. 0.74, p < 0.001), and a slightly lower specificity (0.96 vs. 0.98, p < 0.001). Higher sensitivity of the algorithm was also observed for each of the categories of severe or worse NPDR, PDR, and DME (p < 0.001 for all comparisons). The quadratic-weighted kappa for determination of DR severity levels by the algorithm and human graders was 0.85 and 0.78 respectively (p < 0.001 for the difference). Across different severity levels of DR for determining referable disease, deep learning significantly reduced the false negative rate (by 23%) at the cost of slightly higher false positive rates (2%). Deep learning algorithms may serve as a valuable tool for DR screening.

Secondary Ocular Hypertension and the Risk of Glaucoma Surgery After Dexamethasone Intravitreal Implant in Routine Clinical Practice.

BACKGROUND AND OBJECTIVES: To determine the rate of ocular hypertension (OHT) after dexamethasone intravitreal implant in routine clinical practice and identify patient characteristics associated with a risk for glaucoma surgery. PATIENTS AND METHODS: The charts of 260 eyes from 221 patients with diabetic macular edema, retinal vein occlusion, uveitis, and macular edema secondary to various causes treated with one or more dexamethasone implants were reviewed. Intraocular pressure (IOP), medications, and glaucoma interventions were collected before and after implantation. RESULTS: The mean baseline IOP was 14.3 mm Hg ± 3.6 mm Hg, and after receiving dexamethasone implant(s), 26.2% and 7.7% of patients had IOP greater than 25 mm Hg and 35 mm Hg, respectively. There was evidence (P < .001) of an association between preexisting glaucoma or glaucoma suspect status (103 eyes) and need for glaucoma surgery, and 4.62% (12 eyes) required glaucoma surgery. CONCLUSIONS: Secondary OHT induced by the dexamethasone implant can usually be controlled by medications, but the incidence of OHT requiring glaucoma surgery was high (4.62%) in our study relative to rates previously reported in the literature. All patients, especially those with preexisting glaucoma, should be advised of the possible need for glaucoma surgery prior to undergoing treatment with the dexamethasone implant. [Ophthalmic Surg Lasers Imaging Retina. 2018;49:680-685.].

A CASE OF POSTERIOR UVEAL EFFUSION SYNDROME MASQUERADING AS UVEITIS.

PURPOSE: To report a case of posterior uveal effusion syndrome presented and treated as posterior uveitis. METHODS: We report one case of a 73-year-old woman with posterior uveal effusion syndrome who developed bilateral chorioretinal folds, treated as uveitis, who did not resolve on oral steroids. RESULTS: Nasal choroidal thickening and intraretinal fluid were present in both eyes, along with choroidal folds and yellowish pigmentary changes in the macula. CONCLUSION: Posterior uveal effusion syndrome can present with features that can mimic uveitis.

OUTER RETINAL ATROPHY ON SPECTRAL DOMAIN OPTICAL COHERENCE TOMOGRAPHY AFTER RESOLUTION OF GLAUCOMA SURGERY-ASSOCIATED HYPOTONY MACULOPATHY.

PURPOSE: To report two cases of persistent central vision loss after resolution of filtration surgery-associated hypotony, corresponding with foveal outer retinal atrophy on spectral domain optical coherence tomography. METHODS: Retrospective observational case series. A retrospective chart review was conducted of two patients with persistent vision loss after normalization of intraocular pressure and resolution of glaucoma surgery-associated hypotony maculopathy. RESULTS: Several months after glaucoma surgery and resolution of hypotony, both patients suffered persistent vision loss and new foveal photoreceptor loss on spectral domain optical coherence tomography. CONCLUSION: This is the first case series demonstrating foveal outer retinal atrophy on spectral domain optical coherence tomography after resolution of glaucoma surgery-associated hypotony maculopathy. Both patients had previous uveitis that was controlled at the time of hypotony. Eyes with previous hypotony and compromised ciliary body function may be at greater risk of photoreceptor damage.

Conversion back to bevacizumab or ranibizumab for recurrent neovascular activity with aflibercept in age-related macular degeneration: a case series.

BACKGROUND: Neovascular age-related macular degeneration often requires chronic therapy with anti-VEGF agents, and patients with recurrent disease are challenging to manage. METHODS: This retrospective case series evaluates patients who were switched from bevacizumab or ranibizumab to aflibercept and then back again because of recurrent fluid on optical coherence tomography (OCT) by reporting changes in OCT measurements over the course of medication changes. RESULTS: Twenty-one eyes in nineteen patients received an average of 20.7 bevacizumab and/or ranibizumab injections and then an average of 7.2 aflibercept injections before being switched back to bevacizumab or ranibizumab because of recurrent fluid on OCT. Median central macular thickness improved on transition from bevacizumab or ranibizumab (317 µm) to aflibercept (285 µm; p = 0.034), then worsened over the course of aflibercept treatment (296 µm; p = 0.080), but improved again with transition from aflibercept back to bevacizumab or ranibizumab (283 µm; p = 0.016). The total volume of subretinal fluid, intraretinal fluid, and pigment epithelial detachments also decreased on transition from bevacizumab or ranibizumab (2.56 mm3) to aflibercept (2.44 mm3; p = 0.080), then worsened over the course of aflibercept treatment (3.18 mm3; p = 0.019), and improved again on transition back to bevacizumab or ranibizumab (2.11 mm3; p = 0.016). CONCLUSIONS: While aflibercept appears initially effective, some patients develop recurrent fluid with aflibercept that improves with transition back to bevacizumab or ranibizumab. Rotating anti-VEGF agents may be beneficial with recurrent neovascular activity.