APPLICATIONS OF MULTIMODAL GENERATIVE ARTIFICIAL INTELLIGENCE IN A REAL-WORLD RETINA CLINIC SETTING.

PURPOSE: This study evaluates a large language model, Generative Pre-trained Transformer 4 with vision, for diagnosing vitreoretinal diseases in real-world ophthalmology settings. METHODS: A retrospective cross-sectional study at Bascom Palmer Eye Clinic, analyzing patient data from January 2010 to March 2023, assesses Generative Pre-trained Transformer 4 with vision's performance on retinal image analysis and International Classification of Diseases 10th revision coding across 2 patient groups: simpler cases (Group A) and complex cases (Group B) requiring more in-depth analysis. Diagnostic accuracy was assessed through open-ended questions and multiple-choice questions independently verified by three retina specialists. RESULTS: In 256 eyes from 143 patients, Generative Pre-trained Transformer 4-V demonstrated a 13.7% accuracy for open-ended questions and 31.3% for multiple-choice questions, with International Classification of Diseases 10th revision code accuracies at 5.5% and 31.3%, respectively. Accurately diagnosed posterior vitreous detachment, nonexudative age-related macular degeneration, and retinal detachment. International Classification of Diseases 10th revision coding was most accurate for nonexudative age-related macular degeneration, central retinal vein occlusion, and macular hole in OEQs, and for posterior vitreous detachment, nonexudative age-related macular degeneration, and retinal detachment in multiple-choice questions. No significant difference in diagnostic or coding accuracy was found in Groups A and B. CONCLUSION: Generative Pre-trained Transformer 4 with vision has potential in clinical care and record keeping, particularly with standardized questions. Its effectiveness in open-ended scenarios is limited, indicating a significant limitation in providing complex medical advice.

Efficacy and Safety of Anti-VEGF Injections and Surgery for Age-Related Macular Degeneration-Related Submacular Hemorrhage: A Systematic Review and Meta-Analysis.

TOPIC: This systematic review and meta-analysis investigates the efficacy and safety of anti-VEGF injections compared with surgical intervention in improving visual acuity (VA) and reducing complications for patients with submacular hemorrhage (SMH) due to neovascular age-related macular degeneration (AMD). CLINICAL RELEVANCE: Determining the optimal intervention for SMH in AMD is crucial for patient care. METHODS: We included studies on anti-VEGF injections or surgical interventions for SMH in AMD from 7 databases, searched up to May 2024. Data extraction and quality assessment were done by 2 independent reviewers. Certainty of evidence was assessed using the Grading of Recommendations, Assessment, Development and Evaluation (GRADE) approach. Meta-analysis employed random-effects models. Primary outcomes were pooled mean logarithm of the minimum angle of resolution VA difference (initial examination minus last follow-up VA) and adverse events rates. RESULTS: A total of 43 observational studies were included: 21 (960 eyes) on anti-VEGF and 22 (455 eyes) on surgery. Comparisons were made across separate studies due to lack of head-to-head studies. Meta-analysis included 11 anti-VEGF studies (444 eyes) and 12 surgical studies (195 eyes) for VA outcomes. The mean difference in VA was -0.16 (95% confidence interval (CI), -0.24 to -0.08) for anti-VEGF and -0.36 (95% CI, -0.68 to -0.04) for surgery, with no significant difference between groups (chi-square = 1.70, df = 1, P = 0.19). Heterogeneity was high in surgical studies (I2 = 96.2%, τ2 = 0.23, P < 0.01) and negligible in anti-VEGF studies (I2 = 7%, τ2 = 0.003, P = 0.38). The GRADE certainty was moderate for anti-VEGF and low for surgery. Anti-VEGF had lower rates of cataract (0% vs. 4.6%), proliferative vitreoretinopathy (0.1% vs. 2.0%), and retinal detachment (0.1% vs. 10.6%), but similar rates of recurrent hemorrhage (5.4% vs. 5.3%). Complications were summarized descriptively due to zero-cell problem. CONCLUSION: Both anti-VEGF and surgery treat SMH in AMD with similar VA outcomes but different safety profiles. Anti-VEGF is preferred for less severe hemorrhage, whereas surgery is suited for extensive hemorrhage. Despite uncertain comparative VA outcomes, treatment should be guided by clinical judgment and patient factors. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

SURGICAL OUTCOMES IN CONCURRENT SICKLE CELL AND DIABETIC RETINOPATHY.

PURPOSE: To report the clinical features and outcomes in eyes that underwent vitreoretinal surgery for complications of concurrent sickle cell retinopathy and diabetic retinopathy. METHODS: Retrospective, consecutive case series of all eyes that underwent vitreoretinal surgery for complications secondary to concurrent sickle cell retinopathy and diabetic retinopathy between January 01, 2014, and December 31, 2021. RESULTS: The study included 20 eyes of 14 patients. Indications for surgery included tractional retinal detachment in 12 eyes (60%), combined tractional retinal detachment/rhegmatogenous retinal detachment in 6 eyes (30%), and vitreous hemorrhage in 2 eyes (10%). Pars plana vitrectomy was performed in all eyes. One eye received a scleral buckle at the same time as pars plana vitrectomy. There was no change in mean best-corrected visual acuity at the last follow-up examination (1.5 [20/678]) when compared with mean preoperative best-corrected visual acuity (1.6 [20/762], P = 0.83). Preoperative best-corrected visual acuity correlated with postoperative best-corrected visual acuity at the last follow-up examination in eyes with retinal detachment (r = 0.49, P = 0.04). Single operation anatomic success was achieved in 11 of 17 eyes (65%) with retinal detachment. CONCLUSION: Functional and anatomic outcomes after surgery in eyes with concurrent sickle cell retinopathy and diabetic retinopathy are relatively poor.

Potential Risk Factors for Ocular Pain in Patients Undergoing Multiple Intravitreal Injections of Anti-Vascular Endothelial Growth Factor.

Purpose: To assess ocular pain in patients undergoing multiple intravitreal injections of anti-vascular endothelial growth factor (anti-VEGF) who have previous factors that may influence pain sensitivity. Methodology: This is a prospective, observational, case series study involving patients who underwent multiple (≥3) pro re nata intravitreal injections of ranibizumab or aflibercept to treat any cause of chorioretinal vascular disease. Ocular pain was assessed by the numerical analog scale during intravitreal injection. For this study, the main variable was ocular pain and the secondary variables included age, sex, previous history of glaucoma, primary retinal vascular disease, severe dry eye history, trigeminal pain, scleral buckle surgery, collagen diseases, fibromyalgia, severe migraine history, pars plana vitrectomy, scleral thickness measurements, and type of anti-VEGF. Results: In a total of 894 patients, 948 eyes (4822 intravitreal injections), 793 patients (88.6%) had ocular pain sensitivity between no pain to mild pain, 80 patients (8.9%) had moderate ocular pain, 15 patients (1.6%) had severe ocular pain, and 6 patients (0.7%) had extremely severe ocular pain. Patients with severe dry eye (p = 0.01) and previous history of scleral buckle surgery (p = 0.01) showed a significant correlation with ocular pain during intravitreal injection. Pars plana scleral thickness (>550 um) and diabetic neuropathy were associated with ocular pain but did not meet the criteria for statistical significance (p = 0.09 and p = 0.06, respectively). Conclusion: Dry eye and prior scleral buckle surgery may contribute to pain associated with intravitreal injection. These issues should be taken into consideration in patients undergoing multiple intravitreal injections.

Faricimab in Previously Treated Eyes With Neovascular Age-Related Macular Degeneration: An Assessment of Durability and Treatment Outcomes.

BACKGROUND AND OBJECTIVE: This study evaluated the efficacy and durability of faricimab in patients with neovascular age-related macular degeneration (nAMD) who were previously treated with anti-vascular endothelial growth factor (anti-VEGF) agents. PATIENTS AND METHODS: This retrospective case series was conducted at a single tertiary center in the United States. It focused on nAMD patients who transitioned to faricimab after initial anti-VEGF therapy, with a follow-up period of at least 9 months. "Complete dryness" was defined as the absence of intra- and/or subretinal fluid on optical coherence tomography. Durability was gauged by the extension of treatment intervals relative to the injection frequency of the previous agent. RESULTS: Sixty-two eyes from 62 patients were included. Treatment interval ranged from 5 to 10 weeks; 10 (16%) patients were able to be extended by 2 or more weeks compared to their previous regimen. Median (interquartile range [IQR]) central field thickness was 310 µm (254, 376) on initiating faricimab and declined by the ninth month (P values at 3, 6, and 9 months were 0.01, 0.02, and 0.07, respectively). Median (IQR) visual acuity at initiation of faricimab was 0.4 (0.20, 0.50) and did not change by the ninth month. Complete anatomical dryness was present in 10 (16%) eyes before switching; 90% remained dry at 9 months. Of 52 (84%) incompletely dry eyes before switching, 15% achieved complete dryness by 9 months on faricimab. CONCLUSIONS: Faricimab modestly improved the treatment intervals for a small proportion of previously treated patients on anti-VEGF therapy. [Ophthalmic Surg Lasers Imaging Retina 2024;55:504-509.].

A Cost Effectiveness Analysis of Avacincaptad Pegol for the Treatment of Geographic Atrophy with Comparison to Pegcetacoplan.

PURPOSE: The purpose of this study was to evaluate the cost effectiveness of the treatment of geographic atrophy (GA) with intravitreal avacincaptad pegol (ACP) and to compare it with pegcetacoplan (PEG). DESIGN: Cost analysis based on data from published studies. SUBJECTS: None; based on data from published sham control compared with 2 treatment groups in each of the index studies. METHODS: Costs were based on 2022 Medicare reimbursement data for both facility (hospital-based) and nonfacility settings in Miami. Specific usage and outcomes were derived from the GATHER2 study as well as DERBY and OAKS trials. For ACP, all patients were treated every month (EM) in year 1 then randomized to every other month (EOM) or EM in year 2. Two-year models were created for patients in the facility setting for extrafoveal (ACP and PEG) and all patients (PEG). MAIN OUTCOME MEASURES: Cost, cost utility, and cost per area of GA (in United States dollars). RESULTS: The cost to treat GA with ACP in EM and EOM treatment groups over the 2 years as reported was $67 400 and $40 600, respectively. With ACP treatment over 2 years, the daily cost of delaying GA 3.4 months (EM) and 4.5 months (EOM) was $649 (EM) and $356 (EOM). The (facility-based) costs per unit area of retinal pigment epithelium saved for patients with extrafoveal GA over the 2-year period were $119 000/mm2 (EM ACP) versus $54 000/mm2 (EM PEG) (P < 0.001), $57 100/mm2 (EOM ACP) versus $31 400/mm2 (EOM PEG) (P < 0.001), and $45 300/mm2 (hypothetical EOM from outset ACP). CONCLUSION: Treatment of GA with intravitreal ACP EOM was more cost effective than EM. When assessing extrafoveal lesions, ACP was less cost effective than PEG for both EM and EOM treatment. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Phylogenetic and Recombination Analysis of Clinical Vitreous Humor-Derived Adenovirus Isolates Reveals Discordance Between Serotype and Phylogeny.

Purpose: To sequence, identify, and perform phylogenetic and recombination analysis on three clinical adenovirus samples taken from the vitreous humor at the Bascom Palmer Eye Institute. Methods: The PacBio Sequel II was used to sequence the genomes of the three clinical adenovirus isolates. To identify the isolates, a full genome-based multiple sequence alignment (MSA) of 722 mastadenoviruses was generated using multiple alignment using fast Fourier transform (MAFFT). MAFFT was also used to generate genome-based human adenovirus B (HAdV-B) MSAs, as well as HAdV-B fiber, hexon, and penton protein-based MSAs. To examine recombination within HAdV-B, RF-Net 2 and Bootscan software programs were used. Results: In the course of classifying three new atypical ocular adenovirus samples, taken from the vitreous humor, we found that all three isolates were HAdV-B species. The three Bascom Palmer HAdV-B genomes were then combined with over 300 HAdV-B genome sequences, including nine ocular HAdV-B genome sequences. Attempts to categorize the penton, hexon, and fiber serotypes using phylogeny of the three Bascom Palmer samples were inconclusive due to incongruence between serotype and phylogeny in the dataset. Recombination analysis using a subset of HAdV-B strains to generate a hybridization network detected recombination between nonhuman primate and human-derived strains, recombination between one HAdV-B strain and the HAdV-E outgroup, and limited recombination between the B1 and B2 clades. Conclusions: The discordance between serotype and phylogeny detected in this study suggests that the current classification system does not accurately describe the natural history and phylogenetic relationships among adenoviruses.

OUTCOMES OF INFECTIOUS PANUVEITIS ASSOCIATED WITH SIMULTANEOUS MULTI-POSITIVE OCULAR FLUID POLYMERASE CHAIN REACTION.

PURPOSE: To evaluate features of infectious panuveitis associated with multiple pathogens detected by ocular fluid sampling. METHODS: Single-center, retrospective, consecutive case series of patients with aqueous/vitreous polymerase chain reaction testing with >1 positive result in a single sample from 2001 to 2021. RESULTS: Of 1,588 polymerase chain reaction samples, 28 (1.76%) were positive for two pathogens. Most common pathogens were cytomegalovirus (n = 16, 57.1%) and Epstein-Barr virus (n = 13, 46.4%), followed by varicella zoster virus (n = 8, 28.6%), Toxoplasma gondii (n = 6, 21.4%), herpes simplex virus 2 (n = 6, 21.4%), herpes simplex virus 1 (n = 6, 21.4%), and Toxocara (n = 1, 3.6%). Mean initial and final visual acuity (logarithm of the minimum angle of resolution) were 1.3 ± 0.9 (Snellen ∼20/400) and 1.3 ± 1.1 (Snellen ∼20/400), respectively. Cytomegalovirus-positive eyes (n = 16, 61.5%) had a mean final visual acuity of 0.94 ± 1.1 (Snellen ∼20/175), whereas cytomegalovirus-negative eyes (n = 10, 38%) had a final visual acuity of 1.82 ± 1.0 (Snellen ∼20/1,320) ( P < 0.05). Main clinical features included intraocular inflammation (100%), retinal whitening (84.6%), immunosuppression (65.4%), retinal hemorrhage (38.5%), and retinal detachment (34.6%). CONCLUSION: Cytomegalovirus or Epstein-Barr virus were common unique pathogens identified in multi-PCR-positive samples. Most patients with co-infection were immunosuppressed with a high rate of retinal detachment and poor final visual acuity. Cytomegalovirus-positive eyes had better visual outcomes compared with cytomegalovirus-negative eyes.

A Cost-Effectiveness Analysis of Pegcetacoplan for the Treatment of Geographic Atrophy.

PURPOSE: To evaluate the cost-effectiveness of the treatment of geography atrophy (GA) with intravitreal pegcetacoplan and to identify utility-measurement surrogates. DESIGN: Cost analysis based on data from a published study. SUBJECTS: None; based on data from published sham control compared with 2 treatment groups in the index study. METHODS: Costs were based on 2022 Medicare reimbursement data. Specific outcomes were extrapolated from the DERBY and OAKS trials. Assumptions were made for the lifetime analysis based on a theoretical logistic growth model of the atrophy. OUTCOME MEASURES: Cost, cost utility, cost per quality-adjusted life-year, and cost per area of GA (in US$). RESULTS: The costs to treat GA in every month (EM) and every-other-month (EOM) treatment groups over the 2 years as reported were $70 000 and $34 600, respectively. The costs per area of delaying GA for 2 years in all patients were $87 300/mm2 (EM) and $49 200/mm2 (EOM), and in initially extrafoveal patients, $53 900/mm2 (EM) and $32 100/mm2 (EOM). The costs per day of delaying GA for 2 years were $295 (EM) and $170 (EOM); the marginal cost (EM vs. EOM) per retinal pigment epithelium cell saved was $30. The modeled lifetime costs were $350 000 (EM) and $172 000 (EOM), or $309 000/mm2 (EM) and $180 000 (EOM) /mm2. The modeled time to 95% atrophy at 13 years was delayed by 2.5 years (EM) and 2.1 years (EOM). The costs/quality-adjusted life-year gained based on modeled visual loss with 95% atrophy were $706 000 (EM) and $397 000 (EOM). CONCLUSION: Treatment of GA with intravitreal pegcetacoplan EOM was more cost effective than EM. Treatment of extrafoveal lesions yielded greater utility than the treatment of the entire group. As atrophy progression approaches an upper limit, the marginal cost/benefit ratios increase. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Perifoveal exudative vascular anomalous complex and resembling lesions treated with intravitreal triamcinolone.

PURPOSE: To report the use of intravitreal triamcinolone to treat macular edema associated with isolated perifoveal exudative vascular anomalous complex (PVAC) and resembling lesion (PVAC-RL). METHODS: In this case series, three diabetic patients (3 eyes) with PVAC-RLs and one healthy patient (1 eye) with PVAC lesion associated with cystic spaces underwent three intravitreal injections of aflibercept before switching to one intravitreal triamcinolone injection in each patient. RESULTS: Macular edema improved from 297.5 ± 8.10 µm, at baseline, to 269.2 ± 8.89 µm, after triamcinolone; whereas visual acuity improved from 20/38 to 20/26 (ETDRS). CONCLUSION: PVAC and PVAC-RL are rare and often misdiagnosed lesions that may be associated with decreased vision. Our outcomes suggest that intravitreal injection of triamcinolone may be an effective and affordable treatment for PVAC and PVAC-RL with intraretinal fluid.

RETINAL VASCULAR DISEASE IN LIMB-GIRDLE MUSCULAR DYSTROPHY.

PURPOSE: To report bilateral retinal vascular occlusive disease in limb-girdle muscular dystrophy (LGMD). METHODS: Case report. RESULTS: A 34-year-old Asian woman was referred for evaluation and management of central retinal vein occlusion. Ultra-wide-field fluorescein angiography showed resolving initial peripheral retinal vein occlusion in one eye and peripheral venular segmental staining in the fellow asymmetric eye. Genetic testing established the diagnosis of LGMD. CONCLUSION: Similar to other forms of muscular dystrophy, LGMD is caused by genetic abnormalities in sarcolemma proteins, a key structural component that connects the intracellular cytoskeleton of a myofiber to the extracellular matrix. Like other muscular dystrophies, LGMD may be associated with retinal vascular abnormalities noted. In this case, retinal vascular smooth muscle dysfunction was seen in LGMD, analogous to reported vascular abnormalities in other muscular dystrophies such as facioscapulohumeral dystrophy and Duchenne muscular dystrophy.

Unexplained Vision Loss Associated With Intraocular Silicone Oil Tamponade in Rhegmatogenous Retinal Detachment Repair.

Purpose: To evaluate the visual outcomes with unexplained vision loss during or after silicone oil (SO) tamponade. Methods: This multicenter retrospective case series comprised patients with unexplained vision loss associated with SO tamponade or its removal. Eyes with other clear secondary identifiable causes of vision loss were excluded. Results: Twenty-nine eyes of 28 patients (64% male) were identified. The mean age was 50 ± 13 years (range, 13-78 years). The mean duration of SO tamponade was 148 ± 38 days. Eighteen eyes (62%) developed unexplained vision loss while under SO; 11 (38%) had vision loss after SO removal. The most common optical coherence tomography (OCT) finding was ganglion cell layer (GCL) thinning (55%). Eyes with vision loss after SO removal had a mean logMAR best-corrected visual acuity (BCVA) of 0.6 ± 0.7 (Snellen 20/85) before SO tamponade and 1.2 ± 0.4 (20/340) before SO removal. By the last follow-up after SO removal, the BCVA had improved to 1.1 ± 0.4 (20/235). In eyes with vision loss after SO removal, the BCVA before SO removal was 0.7 ± 0.7 (20/104), which deteriorated to 1.4 ± 0.4 (20/458) 1 month after SO removal. By the last follow-up, the BCVA had improved to 1.0 ± 0.5 (20/219). Conclusions: Unexplained vision loss can occur during SO tamponade or after SO removal. Vision loss was associated with 1000-centistoke and 5000-centistoke oil and occurred in macula-off and macula-on retinal detachments. The duration of tamponade was 3 months or longer in the majority of eyes. Most eyes had GCL thinning on OCT. Gradual visual recovery can occur yet is often incomplete.

Familial Exudative Vitreoretinopathy-Like Phenotype in a Patient With Microcephaly and TUBGCP6 Mutations.

Purpose: To describe a case of microcephaly, unilateral retinal fold, and familial exudative vitreoretinopathy (FEVR)-like phenotype in the context of 2 TUBGCP6 variants. Methods: A case and its findings were analyzed. Results: A 4-month-old boy with no family history of eye disease presented by referral for management of presumed persistent fetal vasculature in the left eye. An external examination showed microcephaly. The patient grimaced to light in both eyes, and the anterior segments were unremarkable. On dilated fundus examination, diffuse chorioretinal atrophy was present bilaterally. In the left eye, a retinal fold emanated from the optic nerve head. There was early termination of retinal vasculature, especially in zone 3 in the left eye, resembling a FEVR-like phenotype. Panel-based genetic testing was performed and found 2 mutations in TUBGCP6. Conclusions: Microcephaly, chorioretinopathy, and retinal folds may be associated with TUBGCP6 mutations and masquerade as PFV.

Trends in Immunosuppressive Agent Use for Non-Infectious Uveitis by US Ophthalmologists in Medicare Beneficiaries and Association with Physician-Industry Interactions.

PURPOSE: To report the trends of immunosuppressive drug use for non-infectious uveitis and explore their relationship with industry payments. METHODS: A retrospective review of ophthalmologists reimbursed by Medicare for the administration of adalimumab (ADA), repository corticotropin (RCI), methotrexate (MTX), and mycophenolate mofetil (MMF) between 2014 and 2018. RESULTS: A total of 316 ophthalmologists were reimbursed by Medicare for 1567 ADA, 465 RCI, 1752 MTX, and 12 333 MMF administrations. The number and dollar amount of industry payments were positively associated with ADA and RCI use (P < 0.001). From 2014 to 2018, there was a positive trend in the proportion of ADA (P = 0.007) and RCI (P = 0.007) used and negative trend in the proportion of MMF (P = 0.025) used. CONCLUSION: From 2014 to 2018, the use of ADA and RCI increased while MMF decreased and MTX remained stable. There was a positive association between ADA and RCI use and physician-industry interactions. A causal relationship is not determined.

Endophthalmitis caused by Burkholderia cepacia complex (BCC): clinical characteristics, antibiotic susceptibilities, and treatment outcomes.

PURPOSE: To report the clinical characteristics, antibiotic susceptibilities, and review the literature of Burkholderia cepacia complex (BCC) associated endophthalmitis. STUDY DESIGN: Retrospective, observational case series. METHODS: Clinical and microbiology records were reviewed for patients evaluated at the Bascom Palmer Eye Institute and diagnosed wisth culture-confirmed endophthalmitis due to BCC. Antibiotic susceptibility profiles were generated using standard microbiologic protocols via an automated VITEK system. RESULTS: Endophthalmitis associated with BCC was diagnosed in three patients. Infection occurred in the setting of post-penetrating keratoplasty (PKP), glaucoma filtering surgery, and suspected trauma. All isolates demonstrated in vitro susceptibility to ceftazidime and meropenem. Presenting visual acuity (VA) ranged from hand motion to light perception. Initial treatment strategies included intravitreal ceftazidime (2.25 mg/0.1 mL) and vancomycin (1.0 mg/0.1 mL) injections with fortified topical antibiotics in 2 patients, and surgical debridement of a corneoscleral melt with patch graft along with both topical fortified antibiotics oral antibiotics in the third patient. In all 3 patients, there was no VA improvement at last follow-up, as 2 eyes ultimately underwent enucleation and 1 eye exhibited phthisis bulbi at last follow-up. BCC related endophthalmitis was reviewed among 13 reports. Treatment outcomes were generally poor and antibiotic resistance was common. These BCC isolates cases demonstrated broad resistance patterns, with susceptibilities to ceftazidime (58%), ciprofloxacin (53%), and gentamicin (33%). CONCLUSIONS: Endophthalmitis caused by B. cepacia is a rare clinical entity with generally poor visual outcomes despite prompt treatment with appropriate antibiotics.