A comparison of same setting versus delayed vitrectomy in the management of retained lens fragments after cataract surgery.

PURPOSE: Intravitreal retained lens fragments are a rare but potentially serious complication of phacoemulsification. The purpose of this study was to compare same setting ("no wait") vitrectomy with delayed surgery in the management of retained lens fragments in a single academic setting. METHODS: This study is a retrospective nonrandomized study of all patients undergoing pars plana vitrectomy for retained lens fragments after cataract surgery from 2007 to 2012. Outcomes included visual acuity and the development of various complications such as retinal detachment, elevated intraocular pressure >30 mmHg, and cystoid macular edema. Multivariate analysis was performed to adjust for potentially confounding variables such as age and preoperative visual acuity. RESULTS: Twenty-eight consecutive eyes (13 same setting, 15 delayed setting) were included in the analysis. Patients in the same setting group were older than in the delayed group (81.00 vs. 72.87 years, P = 0.053). No other preoperative differences existed between the groups (axial length, preoperative vision, and intraocular pressure). The mean time to pars plana vitrectomy in the delayed group was 26.6 days (range, 1-91 days). The mean follow-up time was 363 days (same setting) and 643 days (delayed). At the most recent follow-up, no significant difference existed in mean vision between the same setting (logMAR, 0.42) and the delayed group (logMAR, 0.57) (P = 0.132). Multivariate analysis showed no difference in final vision when adjusting for age and preoperative vision. Although there was a trend for eyes in the same setting group to obtain good vision (≥ 20/40) faster, a higher percentage of eyes in the delayed group obtained good vision at the most recent follow-up (66.7 vs. 23.1%, P = 0.02). More eyes in the delayed group had an intraocular pressure >30 at any point (P = 0.055). There was no significant difference between the groups in any other complications such as retinal detachment, choroidal detachment, and cystoid macular edema during the follow-up. CONCLUSION: In this cohort, same setting pars plana vitrectomy offers no significant visual acuity advantage over delayed pars plana vitrectomy in patients with retained lens fragments. Fewer eyes in the same setting group "ever" had an intraocular pressure ≥ 30 during follow-up, whereas no other complication differences were seen between the groups.

Federated Learning for Multicenter Collaboration in Ophthalmology: Improving Classification Performance in Retinopathy of Prematurity.

OBJECTIVE: To compare the performance of deep learning classifiers for the diagnosis of plus disease in retinopathy of prematurity (ROP) trained using 2 methods for developing models on multi-institutional data sets: centralizing data versus federated learning (FL) in which no data leave each institution. DESIGN: Evaluation of a diagnostic test or technology. SUBJECTS: Deep learning models were trained, validated, and tested on 5255 wide-angle retinal images in the neonatal intensive care units of 7 institutions as part of the Imaging and Informatics in ROP study. All images were labeled for the presence of plus, preplus, or no plus disease with a clinical label and a reference standard diagnosis (RSD) determined by 3 image-based ROP graders and the clinical diagnosis. METHODS: We compared the area under the receiver operating characteristic curve (AUROC) for models developed on multi-institutional data, using a central approach initially, followed by FL, and compared locally trained models with both approaches. We compared the model performance (κ) with the label agreement (between clinical and RSD), data set size, and number of plus disease cases in each training cohort using the Spearman correlation coefficient (CC). MAIN OUTCOME MEASURES: Model performance using AUROC and linearly weighted κ. RESULTS: Four settings of experiment were used: FL trained on RSD against central trained on RSD, FL trained on clinical labels against central trained on clinical labels, FL trained on RSD against central trained on clinical labels, and FL trained on clinical labels against central trained on RSD (P = 0.046, P = 0.126, P = 0.224, and P = 0.0173, respectively). Four of the 7 (57%) models trained on local institutional data performed inferiorly to the FL models. The model performance for local models was positively correlated with the label agreement (between clinical and RSD labels, CC = 0.389, P = 0.387), total number of plus cases (CC = 0.759, P = 0.047), and overall training set size (CC = 0.924, P = 0.002). CONCLUSIONS: We found that a trained FL model performs comparably to a centralized model, confirming that FL may provide an effective, more feasible solution for interinstitutional learning. Smaller institutions benefit more from collaboration than larger institutions, showing the potential of FL for addressing disparities in resource access.

Federated Learning for Multicenter Collaboration in Ophthalmology: Implications for Clinical Diagnosis and Disease Epidemiology.

OBJECTIVE: To utilize a deep learning (DL) model trained via federated learning (FL), a method of collaborative training without sharing patient data, to delineate institutional differences in clinician diagnostic paradigms and disease epidemiology in retinopathy of prematurity (ROP). DESIGN: Evaluation of a diagnostic test or technology. SUBJECTS AND CONTROLS: We included 5245 patients with wide-angle retinal imaging from the neonatal intensive care units of 7 institutions as part of the Imaging and Informatics in ROP study. Images were labeled with the clinical diagnoses of plus disease (plus, preplus, no plus), which were documented in the chart, and a reference standard diagnosis was determined by 3 image-based ROP graders and the clinical diagnosis. METHODS: Demographics (birth weight, gestational age) and clinical diagnoses for all eye examinations were recorded from each institution. Using an FL approach, a DL model for plus disease classification was trained using only the clinical labels. The 3 class probabilities were then converted into a vascular severity score (VSS) for each eye examination, as well as an "institutional VSS," in which the average of the VSS values assigned to patients' higher severity ("worse") eyes at each examination was calculated for each institution. MAIN OUTCOME MEASURES: We compared demographics, clinical diagnoses of plus disease, and institutional VSSs between institutions using the McNemar-Bowker test, 2-proportion Z test, and 1-way analysis of variance with post hoc analysis by the Tukey-Kramer test. Single regression analysis was performed to explore the relationship between demographics and VSSs. RESULTS: We found that the proportion of patients diagnosed with preplus disease varied significantly between institutions (P < 0.001). Using the DL-derived VSS trained on the data from all institutions using FL, we observed differences in the institutional VSS and the level of vascular severity diagnosed as no plus (P < 0.001) across institutions. A significant, inverse relationship between the institutional VSS and mean gestational age was found (P = 0.049, adjusted R2 = 0.49). CONCLUSIONS: A DL-derived ROP VSS developed without sharing data between institutions using FL identified differences in the clinical diagnoses of plus disease and overall levels of ROP severity between institutions. Federated learning may represent a method to standardize clinical diagnoses and provide objective measurements of disease for image-based diseases.

Ocular Outcomes after Treatment of Cytomegalovirus Retinitis Using Adoptive Immunotherapy with Cytomegalovirus-Specific Cytotoxic T Lymphocytes.

PURPOSE: To describe ocular outcomes in eyes with cytomegalovirus (CMV) retinitis treated with adoptive immunotherapy using systemic administration of CMV-specific cytotoxic Tlymphocytes (CMV-specific CTLs). DESIGN: Retrospective cohort study. PARTICIPANTS: Patients with active CMV retinitis evaluated at a tertiary care academic center. METHODS: Treatment of CMV retinitis with standard-of-care therapy (systemic or intravitreal antivirals) or CMV-specific CTLs (with or without concurrent standard-of-care therapies). MAIN OUTCOME MEASURES: The electronic medical record was reviewed to determine baseline characteristics, treatment course, and ocular outcomes, including best-corrected visual acuity (BCVA), treatments administered (CMV-specific CTLs, systemic antivirals, intravitreal antivirals), resolution of CMV retinitis, any occurrence of immune recovery uveitis, cystoid macular edema, retinal detachment, or a combination thereof. RESULTS: Seven patients (3 of whom had bilateral disease [n = 10 eyes]) were treated with CMV-specific CTLs, whereas 20 patients (6 of whom had bilateral disease [n = 26 eyes]) received standard-of-care treatment. Indications for CMV-specific CTL therapy included persistent or progressive CMV retinitis (71.4% of patients); CMV UL54 or UL97 antiviral resistance mutations (42.9%); side effects or toxicity from antiviral agents (57.1%); patient intolerance to longstanding, frequent antiviral therapy for persistent retinitis (28.6%); or a combination thereof. Two patients (28.6%; 4 eyes [40%]) received CMV-specific CTL therapy without concurrent systemic or intravitreal antiviral therapy for active CMV retinitis, whereas 5 patients (71.4%; 6 eyes [60%]) continued to receive concurrent antiviral therapies. Resolution of CMV retinitis was achieved in 9 eyes (90%) treated with CMV-specific CTLs, with BCVA stabilizing (4 eyes [40%]) or improving (4 eyes [40%]) in 80% of eyes over an average follow-up of 33.4 months. Rates of immune recovery uveitis, new-onset cystoid macular edema, and retinal detachment were 0%, 10% (1 eye), and 20% (2 eyes), respectively. These outcomes compared favorably with a nonrandomized cohort of eyes treated with standard-of-care therapy alone, despite potentially worse baseline characteristics. CONCLUSIONS: CMV-specific CTL therapy may represent a novel monotherapy or adjunctive therapy, or both, for CMV retinitis, especially in eyes that are resistant, refractory, or intolerant of standard-of-care antiviral therapies. More generally, adoptive cell transfer and adoptive immunotherapy may have a role in refractory CMV retinitis. Larger prospective, randomized trials are necessary.

Comorbid hepatitis C does not modulate prevalence or severity of diabetic retinopathy.

PURPOSE: There are limited and conflicting data regarding the impact of comorbid hepatitis C virus (HCV) infection on diabetic retinopathy (DR). This study sought to compare the prevalence and severity of DR among patients with diabetes mellitus (DM) with and without HCV. PATIENTS AND METHODS: This was a retrospective, case-control study of patients with DM comparing 120 patients with comorbid HCV and 120 age-matched controls. DR prevalence and several measures of severity were compared between groups. Subgroup analyses were performed among HCV patients with cirrhosis, comorbid HIV, or history of treatment with interferon. Statistical analysis for between-group comparisons utilized both univariate and multivariate analyses. RESULTS: Cases and controls exhibited similar baseline characteristics: average hemoglobin A1c, DM duration, and age (p>0.05). Among cases and controls, there was no difference in DR prevalence (35.8% versus 42.5%, respectively, p=0.29) or severity (p>0.05). Within the HCV subgroup, DR severity was reduced in patients with HIV or cirrhosis. However, multivariate analysis identified reduced DM duration in these subgroups as the primary contributor to lesser DR severity, rather than HIV or cirrhosis. CONCLUSION: In this study, comorbid HCV did not modulate the prevalence or severity of DR among patients with DM. These findings may inform clinical monitoring among HCV-positive diabetics undergoing ophthalmic evaluation.

Evaluation of tissue interactions with mechanical elements of a transscleral drug delivery device.

The goal of this work was to evaluate tissue-device interactions due to implantation of a mechanically operated drug delivery system onto the posterior sclera. Two test devices were designed and fabricated to model elements of the drug delivery device-one containing a free-spinning ball bearing and the other encasing two articulating gears. Openings in the base of test devices modeled ports for drug passage from device to sclera. Porous poly(tetrafluoroethylene) (PTFE) membranes were attached to half of the gear devices to minimize tissue ingrowth through these ports. Test devices were sutured onto rabbit eyes for 10 weeks. Tissue-device interactions were evaluated histologically and mechanically after removal to determine effects on device function and changes in surrounding tissue. Test devices were generally well-tolerated during residence in the animal. All devices encouraged fibrous tissue formation between the sclera and the device, fibrous tissue encapsulation and invasion around the device, and inflammation of the conjunctiva. Gear devices encouraged significantly greater inflammation in all cases and a larger rate of tissue ingrowth. PTFE membranes prevented tissue invasion through the covered drug ports, though tissue migrated in through other smaller openings. The torque required to turn the mechanical elements increased over 1000 times for gear devices, but only on the order of 100 times for membrane-covered gear devices and less than 100 times for ball bearing devices. Maintaining a lower device profile, minimizing microscale motion on the eye surface and covering drug ports with a porous membrane may minimize inflammation, decreasing the risk of damage to surrounding tissues and minimizing disruption of device operation.

Training fellows for retinopathy of prematurity care: a Web-based survey.

PURPOSE: To characterize the training received by pediatric ophthalmology and retina fellows in retinopathy of prematurity (ROP) management. METHODS: Pediatric ophthalmology and retina fellowship programs were emailed a Web-based survey to assess fellowship training in ROP management. RESULTS: Of 140 programs contacted, 42 (30%) participated, resulting in 87 surveys for analysis. Of the 87 respondents, 25 (29%) reported that two-thirds or less of ROP examinations performed by fellows were also seen by an attending. When stratified by specialty, this trend was statistically different between pediatric ophthalmology and retina fellows (P = 0.03). Additionally, pediatric ophthalmology fellows performed fewer laser photocoagulation procedures than retina fellows (P < 0.001). Regarding fellows' perceived competency in ROP management, 3 of 51 (6%) felt competent at the start of their fellowship and 43 of 51 (84%) felt competent at the time of the survey. Only 7% of respondents reported the use of formal evaluations at their programs to assess fellow competence in ROP examination. CONCLUSIONS: Training programs for fellows in pediatric ophthalmology and retina vary greatly with respect to ROP training and the quality of clinical care. Many clinical ROP examinations are being performed by pediatric ophthalmology and retina fellows without involvement and/or direct supervision by attending ophthalmologists. Our findings have important implications for the development of a future workforce for ROP management.