Changes in intraocular pressure following intravitreal dexamethasone implant in patients with history of glaucoma filtration surgery.

PURPOSE: This study aimed to evaluate changes in intraocular pressure following intravitreal dexamethasone implant injection, specifically in patients undergoing glaucoma filtration surgery. METHODS: The degree of increase in intraocular pressure was compared retrospectively among three groups. Group 1 comprised patients who underwent prior glaucoma filtration surgery (54 eyes). Group 2 included patients with or suspected glaucoma without such surgical history (20 eyes). Group 3 included patients without glaucoma (33 eyes). Pressure measurements were taken before the injection and at 1, 2, 3, and 6 months post-injection. A subgroup analysis was performed for pressure > 35 mmHg, > 30 mmHg, > 25 mmHg, and a difference > 10 mmHg between the peak and baseline pressure. RESULTS: Group 1 consistently displayed lower pressures compared with Group 2, with significant difference at both 1- and 6-month post-injections (15.09 mmHg vs. 18.10 mmHg, P = 0.042 and 13.91 mg vs. 17.25 mmHg, P = 0.040). The proportion of patients in Group 1 and Group 3 with pressures > 25 mmHg, > 30 mmHg, and a difference > 10 mmHg did not significantly differ (15.6% vs. 9.5%, P = 0.231; 3.1% vs. 2.3%, P = 0.867; and 17.1% vs. 7.1%, P = 0.231). Notably, Group 2 exhibited a significantly higher proportion within each category (> 25 mmHg, 24.0%; > 30 mmHg, 20.0%; > 10 mmHg difference, 28.0%). CONCLUSION: Intravitreal dexamethasone implant did not increase the risk of elevated intraocular pressure in patients with a history of glaucoma filtration surgery compared with patients with suspected glaucoma; the risk was similar to those without glaucoma.

Exploring the Challenges of Distinguishing Punctate Inner Choroidopathy from Multifocal Choroiditis and Panuveitis.

PURPOSE: This retrospective case series aimed to assess the concordance between clinical diagnoses of punctate inner choroidopathy (PIC) and multifocal choroiditis and panuveitis (MCP) using the 2021 Standardization of Uveitis Nomenclature (SUN) Working Group criteria. METHODS: Using the medical records of the patients, we reevaluated 100 eyes of 75 patients with idiopathic multifocal chorioretinal inflammatory lesions based on SUN criteria and compared the result to the clinical diagnosis. RESULTS: Of 100 eyes, 29 eyes (29%) were diagnosed as PIC and 15 eyes (15%) were diagnosed as MCP using SUN criteria, and 56 (56%) eyes could not be diagnosed as either. Clinically diagnosed PIC eyes were significantly more myopic than the clinically diagnosed MCP eyes (mean spherical equivalent -6.65 ± 4.63 vs. -3.85 ± 2.31, P = 0.01). Sixteen eyes with vitreous inflammation were all clinically diagnosed as MCP, but four (25%) could not be diagnosed as MCP using SUN criteria. CONCLUSIONS: The existing diagnostic criteria showed limitations in capturing all clinical cases of PIC or MCP, and adding or revising criteria on features such as vitreous inflammation or myopia, could be considered to enhance diagnostic accuracy.

Multitask Deep Learning for Joint Detection of Necrotizing Viral and Noninfectious Retinitis From Common Blood and Serology Test Data.

Purpose: Necrotizing viral retinitis is a serious eye infection that requires immediate treatment to prevent permanent vision loss. Uncertain clinical suspicion can result in delayed diagnosis, inappropriate administration of corticosteroids, or repeated intraocular sampling. To quickly and accurately distinguish between viral and noninfectious retinitis, we aimed to develop deep learning (DL) models solely using noninvasive blood test data. Methods: This cross-sectional study trained DL models using common blood and serology test data from 3080 patients (noninfectious uveitis of the posterior segment [NIU-PS] = 2858, acute retinal necrosis [ARN] = 66, cytomegalovirus [CMV], retinitis = 156). Following the development of separate base DL models for ARN and CMV retinitis, multitask learning (MTL) was employed to enable simultaneous discrimination. Advanced MTL models incorporating adversarial training were used to enhance DL feature extraction from the small, imbalanced data. We evaluated model performance, disease-specific important features, and the causal relationship between DL features and detection results. Results: The presented models all achieved excellent detection performances, with the adversarial MTL model achieving the highest receiver operating characteristic curves (0.932 for ARN and 0.982 for CMV retinitis). Significant features for ARN detection included varicella-zoster virus (VZV) immunoglobulin M (IgM), herpes simplex virus immunoglobulin G, and neutrophil count, while for CMV retinitis, they encompassed VZV IgM, CMV IgM, and lymphocyte count. The adversarial MTL model exhibited substantial changes in detection outcomes when the key features were contaminated, indicating stronger causality between DL features and detection results. Conclusions: The adversarial MTL model, using blood test data, may serve as a reliable adjunct for the expedited diagnosis of ARN, CMV retinitis, and NIU-PS simultaneously in real clinical settings.

Factors predicting redislocation or suture-break in eyes after scleral-fixated intraocular lens.

PURPOSE: To investigate predictive factors for redislocation in patients with recurrent intraocular lens (IOL) dislocation following secondary scleral-fixated IOL (SFIOL) surgery. SETTING: Two tertiary referral hospitals. DESIGN: Retrospective case series. METHODS: Patients undergoing SFIOL surgery were grouped into redislocation and no-redislocation groups. Medical records of consecutive patients who underwent SFIOL surgery between June 2014 and December 2019 at two tertiary referral centers were reviewed. Data regarding patient demographics, treatment factors, anatomical and functional outcomes, and postoperative complications were recorded. RESULTS: We included 237 eyes of 225 patients (169 [75.1%] men). The redislocation group was more likely to have a younger mean age at the initial SFIOL surgery (redislocation vs no-redislocation, 55.4 vs 62.0 years, respectively; P=0.008), have a prior history of a previous suture-break (23 eyes, 52.3% vs 1 eye, 0.5%; P<0.001), and have undergone the initial SFIOL surgery using <1 mm-sized side-port incisions (17 eyes, 38.6% vs 32 eyes, 16.5%; P=0.002) than was the no-redislocation group. Additionally, the redislocation group had a higher occurrence of complications (P<0.001). Multivariable regression revealed that younger age, left eye involvement, aphakic status prior to the surgery, unremarkable primary IOL dislocation cause, need for ocular hypertension treatment and glaucoma surgery, and no large incision during the initial surgery were significantly (all P<0.05) associated with redislocation. CONCLUSION: Younger age, left eye involvement, postoperative complications like ocular hypertension and glaucoma, and techniques without large incisions increase the risk of redislocation. Conversely, lower risk factors include unremarkable surgery causes and a history of aphakic conditions.

Long-term Outcomes of Surgeries for Retinal Detachment Secondary to Parasitic or Viral Infectious Retinitis.

PURPOSE: This study sought to compare the long-term outcomes of surgeries for retinal detachment (RD) secondary to viral or parasitic infectious retinitis. METHODS: A total of 47 eyes that received pars plana vitrectomy with or without scleral buckling due to RD secondary to polymerase chain reaction-proven viral (cytomegalovirus, varicella zoster virus, and herpes zoster virus) or parasitic (toxoplasma and toxocara) retinitis from October 1, 2006, to June 30, 2023, in a single medical center were retrospectively enrolled. RESULTS: Mean follow-up period was 59.03 ± 55.24 months in viral retinitis and 34.80 ± 33.78 months in parasitic retinitis after primary reattachment surgery. During follow-up, nine eyes (24.3%) with viral retinitis and five eyes (50.0%) with parasitic retinitis developed retinal redetachment. Visual acuity success at final follow-up was achieved in 19 eyes (51.4%) with viral retinitis and six eyes (60.0%) with parasitic retinitis (p = 0.64). The incidence of retinal redetachment during the 1st postoperative year was significantly higher in parasitic retinitis compared with viral retinitis (crude incidence, 0.21 vs. 0.85; p = 0.02). Hazard ratio analysis adjusted for age and sex showed 4.58-fold (95% confidence interval, 1.22-17.27; p = 0.03) increased risk of retinal redetachment in parasitic retinitis compared with viral retinitis during the 1st postoperative year. Tamponade with silicone oil and preoperative diagnostic vitrectomy were associated with significantly decreased risk of retinal redetachment in patients with parasitic retinitis. CONCLUSIONS: Compared with RD secondary to viral retinitis, RD secondary to parasitic retinitis showed higher incidence of retinal redetachment during the 1st postoperative year. Tamponade with silicone oil and preoperative diagnostic vitrectomy were associated with significantly decreased risk of retinal redetachment in patients with parasitic retinitis.

Uveitis Risk After the First Dose of COVID-19 Vaccination Based on Uveitis History: Matched Cohort and Crossover Case Series Study.

PURPOSE: To investigate the risk of noninfectious uveitis following the first dose of coronavirus disease 2019 (COVID-19) vaccination based on the uveitis history. DESIGN: Retrospective matched cohort and crossover case series study. METHODS: A random sample of 7 917 457 individuals who received COVID-19 vaccine between January 2021 and March 2022 in Korea, and had no recorded history of COVID-19 were categorized into the control and uveitis groups based on their uveitis history. After performing 3:1 propensity score matching, we assessed the cumulative incidence and risk of noninfectious uveitis in the 180 days after COVID-19 vaccination. Additionally, we performed a crossover case series analysis to compare the pre- and postvaccination incidence rate ratios (IRRs) of uveitis in individuals with and without a history of uveitis. RESULTS: In the matched cohort analysis, uveitis group had a significantly higher cumulative incidence of uveitis (15.4%) than control group (0.10%). The uveitis group exhibited increased risks of all uveitis types, anterior, and nonanterior uveitis in the first 60 days (hazard ratio [HR]: 169, 158, and 253, respectively) and in days 61 to 180 (HR: 166, 164, and 143, respectively) after vaccination. In the crossover case series analysis, uveitis occurred with relatively equal frequency in 20-day intervals during the 180 days before and after vaccination, regardless of uveitis history. For uveitis group, the adjusted IRRs for early and late postvaccination events were 0.92 (95% CI, 0.88-0.96) and 0.83 (95% CI, 0.80-0.85), respectively. CONCLUSIONS: COVID-19 vaccination did not increase the risk of uveitis, regardless of uveitis history.

Pachychoroid Spectrum Disease: Comparison of Patients with Central Serous Chorioretinopathy Complicated with Pachychoroid Neovasculopathy.

PURPOSE: This study aimed to determine the incidence and visual outcomes of pachychoroid neovasculopathy (PNV) in patients initially diagnosed with central serous chorioretinopathy (CSC). METHODS: In this study, 144 patients aged 20 to 55 years with treatment-naive chronic CSC, defined as the persistence of subretinal fluid (SRF) for ≥6 months, were retrospectively enrolled. Patients with PNV at the initial evaluation were categorized as group 1, whereas those who developed new-onset PNV during follow-up were categorized as group 2. Patients without PNV until the end of the follow-up were categorized as group 3. RESULTS: Over a mean follow-up period of 49.9 ± 39.9 months, new-onset PNV was diagnosed in 11.8% of patients with CSC. The time taken to reach the initial resolution was longest in group 1 (group 1, 11.13 ± 10.70 months; group 2, 8.14 ± 7.90 months; group 3, 7.32 ± 9.55 months), although these differences were not statistically significant. The numbers of injections needed to achieve initial resolution were 3.76 ± 5.90, 1.64 ± 2.06, and 1.74 ± 4.33 in groups 1, 2, and 3, respectively, with no significant differences. SRF recurrence was recorded in seven patients (29.2%) in group 1, nine (64.3%) in group 2, and 28 (26.7%) in group 3. The recurrence rates were significantly higher in group 2 than those in group 1 or 3. At the end of the follow-up period, significant improvements in best-corrected visual acuity were achieved in groups 1 and 3, compared with baseline, but not in group 2. CONCLUSIONS: Patients with chronic CSC with new-onset PNV exhibited higher SRF recurrence and worse visual outcomes compared to those with initial PNV or those with chronic CSC without PNV. Our study emphasizes the importance of routine screening for prompt diagnoses of new-onset PNV in individuals with chronic CSC.

Real-World Incidence of Incident Noninfectious Uveitis in Patients Treated With BRAF Inhibitors: A Nationwide Clinical Cohort Study.

PURPOSE: To compare the incidence of noninfectious uveitis in skin melanoma or lung cancer patients who received BRAF inhibitors with that in those who received immune checkpoint inhibitors (ICIs) or conventional cytotoxic chemotherapy. DESIGN: Nationwide population-based retrospective clinical cohort study METHODS: From the Health Insurance Review and Assessment Service database of South Korea, we retrospectively defined 77,323 patients with skin melanoma or lung cancer who received BRAF inhibitor therapy (BRAF inhibitor-exposed group; n = 396), ICIs (ICI-exposed group; n = 22,474), or conventional cytotoxic chemotherapy (unexposed group; n = 54,453). We calculated the 1-year cumulative incidence of noninfectious uveitis in each group from the first day of BRAF inhibitor, ICI, or cytotoxic agent administration. RESULTS: During the first year of treatment initiation, the cumulative incidence of uveitis was 0.33%, 0.35%, and 2.27% in the unexposed, ICI-exposed, and BRAF inhibitor-exposed groups, respectively. Adjusted hazard ratios (aHR) indicated a 7.52-fold and 5.68-fold increased risk of uveitis in the BRAF inhibitor-exposed group compared with that in the unexposed and ICI-exposed groups (95% confidence interval [CI] 3.83-14.75, P < .001 and 95% CI 2.81-11.47, P < .001, respectively). After 1:4 propensity score matching, aHRs showed a 35.51-fold and 15.80-fold increased risk (95% CI 4.49-280.48, P = .001 and 95% CI 1.76-141.00, P = .014) of uveitis and severe uveitis, respectively, in the BRAF inhibitor-exposed versus unexposed patients. Crossover analysis within the BRAF inhibitor-exposed group showed a 3.71-fold increase in uveitis risk during 1-year post index date in comparison with 1-year prior to index date (95% CI 1.03-13.40, P = .046). In the BRAF inhibitor-exposed group, female sex, chronic kidney disease, and melanoma were associated with a trend of increased, albeit nonsignificant, risk of uveitis. CONCLUSIONS: Melanoma or lung cancer patients treated with BRAF inhibitors showed significantly higher risk of noninfectious uveitis than patients treated with conventional cytotoxic drugs or ICIs. These findings emphasize the importance of pretreatment patient education on BRAF-inhibitor-associated uveitis risk to enable prompt ophthalmic evaluation and treatment if symptoms arise during drug administration.

Real-world incidences and risk factors of immune-related adverse events in patients treated with immune checkpoint inhibitors: A nationwide retrospective cohort study.

Immune-related adverse events (irAEs) caused by immune checkpoint inhibitors (ICIs) are rare but fatal, requiring systemic steroid use. Therefore, to examine the outcomes, incidence, timing, and risk factors of ICI-associated steroid-requiring severe irAEs, we conducted a nationwide, retrospective, cohort study utilizing the Korean Health Insurance and Review Assessment database. We identified 357,010 patients with lung cancer, bladder cancer, or skin melanoma, eligible for ICI reimbursement in Korea between January 2012 to June 2020. Steroid-requiring severe irAEs following ICI treatment or treatment-emergent AEs following cytotoxic chemotherapy were defined as moderate- or high-dose steroid administration for over 2 consecutive days, along with corresponding ICD-10 codes indicating affected organ systems. The ICI-exposed group (N = 10,118) was compared to a matched cohort of 55,436 ICI-unexposed patients treated with cytotoxic chemotherapy. Incidences of acute severe irAEs requiring moderate- and high-dose steroids were higher in the ICI-exposed group (1.95% and 6.42%, respectively). The ICI-exposed group also had a higher risk of developing delayed severe irAEs requiring moderate- and high-dose steroid use (3.89% and 7.39%). Male sex, high comorbidity index, or previously diagnosed autoimmune diseases were associated with an increased risk of severe irAEs. Notably, 27.4-38.8% of the patients experienced recurrent severe irAEs after re-challenge with ICIs following moderate- or high-dose steroid use, with the severity matching the initial episode. Steroid-requiring severe irAEs were significantly more prevalent among patients exposed to ICIs than among those treated with chemotherapy in acute and delayed periods.

Clinician-Driven Reanalysis of Exome Sequencing Data From Patients With Inherited Retinal Diseases.

Importance: Despite advances in next-generation sequencing (NGS), a significant proportion of patients with inherited retinal disease (IRD) remain undiagnosed after initial genetic testing. Exome sequencing (ES) reanalysis in the clinical setting has been suggested as one method for improving diagnosis of IRD. Objective: To investigate the association of clinician-led reanalysis of ES data, which incorporates updated clinical information and comprehensive bioinformatic analysis, with the diagnostic yield in a cohort of patients with IRDs in Korea. Design, Setting, and Participants: This was a multicenter prospective cohort study involving 264 unrelated patients with IRDs, conducted in Korea between March 2018 and February 2020. Comprehensive ophthalmologic examinations and ES analyses were performed, and ES data were reanalyzed by an IRD specialist for single nucleotide variants, copy number variants, mobile element insertions, and mitochondrial variants. Data were analyzed from March to July 2023. Main Outcomes and Measures: Diagnostic rate of conventional bioinformatic analysis and clinician-driven ES reanalysis. Results: A total of 264 participants (151 [57.2%] male; mean [SD] age at genetic testing, 33.6 [18.9] years) were enrolled, including 129 patients (48.9%) with retinitis pigmentosa and 26 patients (9.8%) with Stargardt disease or macular dystrophy. Initial bioinformatic analysis diagnosed 166 patients (62.9%). Clinician-driven reanalysis identified the molecular cause of diseases in an additional 22 patients, corresponding to an 8.3-percentage point increase in diagnostic rate. Key factors associated with new molecular diagnoses included clinical phenotype updates (4 patients) and detection of previously overlooked variation, such as structural variants (9 patients), mitochondrial variants (3 patients), filtered or not captured variants (4 patients), and noncanonical splicing variants (2 patients). Among the 22 patients, variants in 7 patients (31.8%) were observed in the initial analysis but not reported to patients, while those in the remaining 15 patients (68.2%) were newly detected by the ES reanalysis. Conclusions and Relevance: In this cohort study, clinician-centered reanalysis of ES data was associated with improved molecular diagnostic yields in patients with IRD. This approach is important for uncovering missed genetic causes of retinal disease.

Long-term efficacy and safety of brolucizumab in neovascular age-related macular degeneration: A multicentre retrospective real-world study.

PURPOSE: To investigate the long-term efficacy and safety of intravitreal brolucizumab (BRZ) injections in patients with typical neovascular age-related macular degeneration (typical nAMD) and polypoidal choroidal vasculopathy (PCV). METHODS: This multicentre retrospective study included 401 eyes of 398 patients with nAMD who received BRZ injection(s), with a follow-up duration of ≥12 months. Changes in best-corrected visual acuity (BCVA), retinal fluid evaluation and central subfield thickness (CST) on optical coherence tomography were assessed. The efficacy of BRZ was compared between typical nAMD and PCV groups. RESULTS: Analyses were conducted with 280 eyes of 278 patients with typical nAMD and 121 eyes of 120 patients with PCV (mean age, 71.1 ± 8.6 years). 29 eyes (7.2%) were treatment naïve. The mean follow-up period was 15.3 ± 2.8 months; the mean number of BRZ injections within 1 year was 4.5 ± 1.7. BCVA was maintained during the follow-up period, and CST significantly improved from the first injection month and was maintained for 12 months in both the typical nAMD and PCV groups. The dry macula proportion increased from 2.7% at baseline to 56.1% at 1 month and 42.9% at 12 months. Among the 18 eyes that underwent indocyanine green angiography both before and after treatment, 10 (55.6%) showed polyp regression. Overall, the incidence of intraocular inflammation (IOI), retinal vasculitis and occlusive retinal vasculitis was 9.4% (38 eyes), 1.2% (5 eyes) and 0.5% (2 eyes), respectively. IOI occurred from the first to the sixth injections, with an average IOI onset of 28.5 ± 1.4 days. All eyes achieved IOI resolution, although the two eyes with occlusive retinal vasculitis showed a severe visual decline after IOI resolution. CONCLUSION: Brolucizumab was effective in maintaining BCVA and managing fluid in eyes with nAMD for up to 1 year, exhibiting a high polyp regression rate. However, the not uncommon incidence of IOI and the severe visual decline caused by the rare occlusive retinal vasculitis following BRZ treatment underscore the importance of careful monitoring and timely management.