Higher-Order Assessment of OCT in Diabetic Macular Edema from the VISTA Study: Ellipsoid Zone Dynamics and the Retinal Fluid Index

Ellipsoid Zone Integrity and Visual Function in Dry Age-Related Macular Degeneration

In this longitudinal retrospective image analysis, conducted on patients diagnosed with dry age-related macular degeneration (AMD) and 5 years of follow-up imaging data, the study aimed to investigate the relationship between ellipsoid zone (EZ) integrity on spectral domain optical coherence tomography (SD-OCT) and visual acuity (VA). Using a machine learning-enabled feature extraction tool, quantitative EZ parameters were derived from SD-OCT images. The analysis revealed significant correlations between EZ integrity metrics and VA. Eyes with excellent VA (≥20/25 Snellen) exhibited higher EZ integrity, including less EZ attenuation, thicker ellipsoid zone-retinal pigment epithelium (EZ-RPE) thickness, and higher EZ intensity, in contrast to eyes with worse VA (≤20/40 Snellen). Additionally, eyes with geographic atrophy (GA) in the foveal region displayed compromised EZ integrity compared to those without GA. Notably, baseline EZ integrity metrics were predictive of future VA loss. These findings suggest that quantitative SD-OCT measurements of EZ integrity could potentially detect early changes in dry AMD and serve as valuable indicators for predicting future functional outcomes. Furthermore, these measurements hold promise for use in clinical trial screenings, offering insights into the progression of the disease and its impact on visual acuity. This study underscores the importance of EZ integrity assessment in understanding and managing dry AMD.

Automated Evaluation of Ellipsoid Zone At-Risk Burden for Detection of Hydroxychloroquine Retinopathy

BACKGROUND

Screening for hydroxychloroquine (HCQ) retinopathy is crucial to detecting early disease. A novel machine-learning-based optical coherence tomography (OCT) biomarker, Ellipsoid Zone (EZ) At-Risk, can quantitatively measure EZ alterations and at-risk areas for progressive EZ loss in a fully automated fashion. The purpose of this analysis was to compare the EZ At-Risk burden in eyes with HCQ toxicity to eyes without toxicity.

METHODS

IRB-approved image analysis study of 83 subjects on HCQ and 44 age-matched normal subjects. SD-OCT images were reviewed for evidence of HCQ retinopathy. A ML-based, fully automatic measurement of the percentage of the macular area with EZ At-Risk was performed.

RESULTS

The mean age for HCQ subjects was 67.1 ± 13.2 years and 64.2 ± 14.3 years for normal subjects. The mean EZ At-Risk macular burden in the "toxic" group (n = 38) was significantly higher (10.7%) compared to the "non-toxic" group (n = 45; 2.2%; p = 0.023) and the "normal" group (1.4%; p = 0.012). Additionally, the amount of EZ At-Risk burden was significantly correlated with the HCQ dose based on the actual (p = 0.016) and ideal body weight (p = 0.033).

CONCLUSIONS

The novel biomarker EZ-At Risk was significantly higher in subjects with evidence of HCQ retinopathy as well as significantly associated with HCQ dose. This novel biomarker should be further evaluated as a potential screening tool for subjects on HCQ.

Compartmental Exudative Dynamics in Neovascular Age-Related Macular Degeneration: Volumetric Outcomes and Impact of Volatility in a Phase III Clinical Trial

PURPOSE

To examine retinal feature dynamics in eyes with neovascular age-related macular degeneration (nAMD) treated with anti-VEGF therapy and the relationship of these features with visual acuity.

DESIGN

Post hoc analysis of the phase III, randomized, HAWK nAMD clinical trial.

PARTICIPANTS

Participants randomized to the brolucizumab 6 mg or aflibercept 2 mg arms of the trial.

METHODS

Spectral-domain OCT scans collected at 4-week intervals were analyzed using an automated machine learning-enhanced segmentation and feature-extraction platform with manual verification. Quantitative volumetric measures of retinal and exudative features were exported at multiple timepoints over 48 weeks. Volatility of exudative features was calculated as the standard deviation of each feature value during the maintenance phase (week 12-48) of treatment. These features were examined for their associations with anatomic and functional outcomes.

MAIN OUTCOME MEASURES

Longitudinal intraretinal fluid (IRF) and subretinal fluid (SRF) volume, subretinal hyperreflective material (SHRM) volume, ellipsoid zone (EZ) integrity (EZ-retinal pigment epithelium [RPE] volume/thickness), and correlation with best-corrected visual acuity (BCVA).

RESULTS

Intraretinal fluid, SRF, and SHRM demonstrated significant volumetric reduction from baseline with anti-VEGF therapy (P < 0.001 at each timepoint). Ellipsoid zone integrity measures demonstrated significant improvement from baseline (P < 0.001 at each timepoint). Both EZ integrity and SHRM measures correlated significantly with BCVA at all timepoints (EZ-RPE volume: 0.38 ≤ r ≤ 0.47; EZ-RPE central subfield thickness: 0.22 ≤ r ≤ 0.41; SHRM volume: -0.33 ≤ r ≤ -0.44). After treatment initiation, correlations of IRF and SRF volume with BCVA were weak or nonsignificant. Eyes with lower volatility of IRF, SRF, and SHRM volumes during the maintenance phase showed greater improvements in EZ integrity (all P < 0.01) and greater gains in BCVA (all P < 0.01) at week 48 compared with eyes with higher volatility in those exudative parameters.

CONCLUSIONS

Quantitative measures of SHRM volume and EZ integrity correlated more strongly with BCVA than retinal fluid volumes during treatment. High volatility of exudative parameters, including SRF, during the maintenance phase of treatment was associated with loss of EZ integrity and BCVA.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Texture-Based Radiomic SD-OCT Features Associated With Response to Anti-VEGF Therapy in a Phase III Neovascular AMD Clinical Trial

Purpose

The goal of this study was to evaluate the role of texture-based baseline radiomic features (Fr) and dynamic radiomics alterations (delta, FΔr) within multiple targeted compartments on optical coherence tomography (OCT) scans to predict response to anti-vascular endothelial growth factor (VEGF) therapy in neovascular age-related macular degeneration (nAMD).

Methods

HAWK is a phase 3 clinical trial data set of active nAMD patients (N = 1082) comparing brolucizumab and aflibercept. This analysis included patients receiving 6 mg brolucizumab or 2 mg aflibercept and categorized as complete responders (n = 280) and incomplete responders (n = 239) based on whether or not the eyes achieved/maintained fluid resolution on OCT. A total of 481 Fr were extracted from each of the fluid, subretinal hyperreflective material (SHRM), retinal tissue, and sub-retinal pigment epithelium (RPE) compartments. Most discriminating eight baseline features, selected by the minimum redundancy, maximum relevance feature selection, were evaluated using a quadratic discriminant analysis (QDA) classifier on the training set (Str, n = 363) to differentiate between the two patient groups. Classifier performance was subsequently validated on independent test set (St, n = 156).

Results

In total, 519 participants were included in this analysis from the HAWK phase 3 study. There were 280 complete responders and 219 incomplete responders. Compartmental analysis of radiomics featured identified the sub-RPE and SHRM compartments as the most distinguishing between the two response groups. The QDA classifier yielded areas under the curve of 0.78, 0.79, and 0.84, respectively, using Fr, FΔr, and combined Fr, FΔr, and Fc on St.

Conclusions

Utilizing compartmental static and dynamic radiomics features, unique differences were identified between eyes that respond differently to anti-VEGF therapy in a large phase 3 trial that may provide important predictive value.

Translational Relevance

Imaging biomarkers, such as radiomics features identified in this analysis, for predicting treatment response are needed to enhanced precision medicine in the management of nAMD.

Optical coherence tomography in the management of diabetic macular oedema

Diabetic macular oedema (DMO) is the major cause of visual impairment in people with diabetes. Optical coherence tomography (OCT) is now the most widely used modality to assess presence and severity of DMO. DMO is currently broadly classified based on the involvement to the central 1 mm of the macula into non-centre or centre involved DMO (CI-DMO) and DMO can occur with or without visual acuity (VA) loss. This classification forms the basis of management strategies of DMO. Despite years of research on quantitative and qualitative DMO related features assessed by OCT, these do not fully inform physicians of the prognosis and severity of DMO relative to visual function. Having said that, recent research on novel OCT biomarkers development and re-defined classification of DMO show better correlation with visual function and treatment response. This review summarises the current evidence of the association of OCT biomarkers in DMO management and its potential clinical importance in predicting VA and anatomical treatment response. The review also discusses some future directions in this field, such as the use of artificial intelligence to quantify and monitor OCT biomarkers and retinal fluid and identify phenotypes of DMO, and the need for standardisation and classification of OCT biomarkers to use in future clinical trials and clinical practice settings as prognostic markers and secondary treatment outcome measures in the management of DMO.

Functional and Morphological Responses to Fluocinolone Acetonide 0.19 mg in Noninfectious Uveitic Macular Edema Evaluated as the Area-Under-the-Curve

Purpose: This study investigated the impact of baseline clinical and optical coherence tomography (OCT) factors on the response to a 0.19-mg fluocinolone acetonide (FAc) implant in patients with noninfectious uveitic macular edema evaluated by the area under the curve over 24 months. Methods: A retrospective study was conducted of eyes of patients with noninfectious uveitic macular edema undergoing FAc treatment, with follow-up from baseline to 24 months. The area under the curve (AUC) of best-corrected visual acuity (BCVA) and the central macular thickness (CMT) were calculated using the trapezoidal rule. Clinical and OCT data at the time of FAc administration were collected, and associations with AUC of BCVA and CMT changes were investigated. Results: Twenty-three patients were enrolled. BCVA and CMT significantly improved after FAc implantation (P < 0.05). AUCBCVA and AUCCMT were 0.41 ± 0.33 logarithm of minimal angle of resolution/6 months and 320.15 ± 321.64 μm/6 months, respectively. Better baseline BCVA (coefficient [coef.] = 0.83, P < 0.001) and macular thickness reduction after FAc administration (coef. = -0.0001, P < 0.05) were associated with better BCVA after FAc treatment. In contrast, baseline OCT biomarkers such as ellipsoid zone reflectivity and choroidal vascularity index, sex, or disease duration before FAc injection showed no correlation with AUCBCVA and AUCCMT (P > 0.05). The younger the patient at the time of FAc injection, the greater the reduction in CMT (coef. = 1.76, P < 0.05). Conclusions: Among all clinical and morphological baseline factors, Baseline BCVA was the strongest predictor for AUCBCVA, while no association with baseline OCT features was observed. Overall, improvement of BCVA and CMT after FAc injection was maintained over 24 months. This study is registered in the German Clinical Trials Register under the DRKS-ID: DRKS00024399.

Predictors of treatment outcomes following treat-and-extend regimen with aflibercept for branch retinal vein occlusion: post-hoc analysis of the PLATON trial

We investigated predictors of visual outcomes and injection interval in macular edema (ME) secondary to branch retinal vein occlusion (BRVO) treated with a treat-and-extend (TAE) regimen. All 48 patients in a multicenter study were followed for 52 weeks and received three monthly intravitreal aflibercept injections before the TAE regimen, with treatment intervals adjusted by 4 weeks, up to a maximum of 16 weeks. Various laboratory biomarkers and optical coherence tomography parameters were evaluated. Patients were classified into the extension failure group if they had ≥ 1 treatment interval decreased due to an increase in the central macular thickness compared to the previous visit and 18 patients were assigned to this group. In multivariate logistic analyses, presence of microaneurysms and prominent middle limiting membrane (p-MLM) sign, increased initial external limiting membrane (ELM) disruption, and higher total cholesterol were correlated with inhibiting a sustained extension in the injection interval (P = 0.015, P = 0.032, P = 0.037, P = 0.009, respectively). Therefore, in the patients with ME secondary to BRVO with these risk factors, early consideration of frequent injection may improve treatment outcome.

Efficacy and Safety of Brolucizumab, Aflibercept, and Ranibizumab for the Treatment of Patients with Visual Impairment Due to Diabetic Macular Oedema: A Systematic Review and Network Meta-Analysis

INTRODUCTION

Key clinical guidelines recommend anti-vascular endothelial growth factor (VEGF) therapy as first-line treatment for visual impairment due to diabetic macular oedema (DMO). A systematic literature review (SLR) and network meta-analysis (NMA) were conducted comparing the relative efficacy of the anti-VEGF brolucizumab with a focused network of the most relevant comparator dosing regimens approved in countries other than the USA (aflibercept, ranibizumab). The safety and tolerability of brolucizumab were also assessed.

METHODS

A broad SLR was conducted to identify randomised controlled trials to ensure all relevant potential comparators were captured. Identified studies were refined to those appropriate for inclusion in the NMA. A Bayesian NMA was conducted comparing brolucizumab 6 mg (every 12 [Q12W]/every 8 weeks [Q8W]) with relevant aflibercept 2 mg and ranibizumab 0.5 mg regimens.

RESULTS

Fourteen studies were included in the NMA. At 1-year follow-up, the various aflibercept 2 mg and ranibizumab 0.5 mg regimens were mostly comparable with brolucizumab 6 mg Q12W/Q8W across key visual and anatomical outcomes, except brolucizumab 6 mg was favoured over ranibizumab 0.5 mg every 4 weeks (Q4W) for the change from baseline (CFB) in best-corrected visual acuity (BCVA), and BCVA loss/gain of pre-specified numbers of letters, and over ranibizumab 0.5 mg pro re nata for CFB in diabetic retinopathy severity scale, and retinal thickness. At year 2, where data were available, brolucizumab 6 mg showed similar results across efficacy outcomes versus all other anti-VEGFs. In most cases, discontinuation rates (all cause, and due to adverse events [AE]) and serious and overall rates of AEs excluding ocular inflammatory events were similar (in unpooled and pooled-treatment analyses) versus comparators.

CONCLUSION

Brolucizumab 6 mg Q12W/Q8W was comparable or superior to aflibercept 2 mg and ranibizumab 0.5 mg regimens for various visual and anatomical efficacy outcomes and discontinuation rates.

Pathological Neurovascular Unit Mapping onto Multimodal Imaging in Diabetic Macular Edema

Diabetic retinopathy is a form of diabetic microangiopathy, and vascular hyperpermeability in the macula leads to retinal thickening and concomitant reduction of visual acuity in diabetic macular edema (DME). In this review, we discuss multimodal fundus imaging, comparing the pathogenesis and interventions. Clinicians diagnose DME using two major criteria, clinically significant macular edema by fundus examination and center-involving diabetic macular edema using optical coherence tomography (OCT), to determine the appropriate treatment. In addition to fundus photography, fluorescein angiography (FA) is a classical modality to evaluate morphological and functional changes in retinal capillaries, e.g., microaneurysms, capillary nonperfusion, and fluorescein leakage. Recently, optical coherence tomography angiography (OCTA) has allowed us to evaluate the three-dimensional structure of the retinal vasculature and newly demonstrated that lamellar capillary nonperfusion in the deep layer is associated with retinal edema. The clinical application of OCT has accelerated our understanding of various neuronal damages in DME. Retinal thickness measured by OCT enables us to quantitatively assess therapeutic effects. Sectional OCT images depict the deformation of neural tissues, e.g., cystoid macular edema, serous retinal detachment, and sponge-like retinal swelling. The disorganization of retinal inner layers (DRIL) and foveal photoreceptor damage, biomarkers of neurodegeneration, are associated with visual impairment. Fundus autofluorescence derives from the retinal pigment epithelium (RPE) and its qualitative and quantitative changes suggest that the RPE damage contributes to the neuronal changes in DME. These clinical findings on multimodal imaging help to elucidate the pathology in the neurovascular units and lead to the next generation of clinical and translational research in DME.

Optical Coherence Tomography Biomarkers in Predicting Treatment Outcomes of Diabetic Macular Edema after Ranibizumab Injections

Background and Objectives: The identification of possible biomarkers that can predict treatment response among DME eyes is important for the individualization of treatment plans. We investigated optical coherence tomography (OCT)-based biomarkers that may predict the one-year real-life outcomes among diabetic macular edema (DME) eyes following treatment by intravitreal ranibizumab (IVR) injections. Materials and Methods: A total of 65 eyes from 35 treatment-naïve patients with DME treated with ranibizumab injection were recruited. Best-corrected visual acuity (BCVA), central retinal thickness (CRT), intraocular pressure (IOP), and OCT scans were retrospectively recorded at baseline before treatment and at 3 months, 6 months, and 12 months after treatment. The OCT scans were evaluated for biomarkers of interest, which included central retinal thickness (CRT), amount and locations of hyperreflective foci (HRF), subretinal fluid (SRF), intraretinal cysts (IRC), large outer nuclear layer cyst (LONLC), ellipsoid zone disruption (EZD), disorganization of retinal inner layers (DRIL), hard exudates (HE), epiretinal membrane (ERM), and vitreomacular interface (VMI). Correlations between these OCT biomarkers and outcome measures (visual and structural) were statistically analyzed. Results: A total of 65 eyes from 35 patients with DME were enrolled. The mean age was 64.2 ± 10.9 years old. Significant improvement in terms of mean BCVA (p < 0.005) and mean CRT was seen at final follow-up compared to baseline. The biomarkers of DRIL, LONLC, and SRF were found to be predictive for at least 50 μm CRT reduction after treatment (with odds ratio of 8.69, 8.5, and 17.58, respectively). The biomarkers of IRC, LONLC, and SRF were predictive for significant improvement in terms of BCVA and CRT after treatment. Finally, the number of HRF was predictive for both BCVA improvement and a CRT reduction of less than 100 μm after treatment. No serious complications were reported during the study. Conclusion: Our study demonstrated the utility of OCT biomarkers as therapeutic predictors of ranibizumab treatment among DME eyes.

Automated Identification and Segmentation of Ellipsoid Zone At-Risk Using Deep Learning on SD-OCT for Predicting Progression in Dry AMD

BACKGROUND

The development and testing of a deep learning (DL)-based approach for detection and measurement of regions of Ellipsoid Zone (EZ) At-Risk to study progression in nonexudative age-related macular degeneration (AMD).

METHODS

Used in DL model training and testing were 341 subjects with nonexudative AMD with or without geographic atrophy (GA). An independent dataset of 120 subjects were used for testing model performance for prediction of GA progression. Accuracy, specificity, sensitivity, and intraclass correlation coefficient (ICC) for DL-based EZ At-Risk percentage area measurement was calculated. Random forest-based feature ranking of EZ At-Risk was compared to previously validated quantitative OCT-based biomarkers.

RESULTS

The model achieved a detection accuracy of 99% (sensitivity = 99%; specificity = 100%) for EZ At-Risk. Automatic EZ At-Risk measurement achieved an accuracy of 90% (sensitivity = 90%; specificity = 84%) and the ICC compared to ground truth was high (0.83). In the independent dataset, higher baseline mean EZ At-Risk correlated with higher progression to GA at year 5 (p < 0.001). EZ At-Risk was a top ranked feature in the random forest assessment for GA prediction.

CONCLUSIONS

This report describes a novel high performance DL-based model for the detection and measurement of EZ At-Risk. This biomarker showed promising results in predicting progression in nonexudative AMD patients.

Longitudinal Quantitative Ultrawide-field Fluorescein Angiography Dynamics in the RUBY Diabetic Macular Edema Study

OBJECTIVE

The purpose of this study was to evaluate the longitudinal change in quantitative ultrawide-field angiographic (UWFA) parameters and correlate them with functional outcomes and spectral domain-OCT metrics.

DESIGN

This study is a post hoc analysis of the phase II RUBY study: a prospective, randomized trial of patients with diabetic macular edema (DME) treated with either intravitreal aflibercept injection (IAI) or combined IAI/nesvacumab (antiangiopoietin 2 mAb).

SUBJECTS

Subjects with DME that underwent UWFA across all treatment groups (n = 44).

METHODS

A machine learning-enabled feature extraction system generated panretinal quantitative UWFA metrics, including leakage, ischemia, and microaneurysm (MA) burden. Zonal assessments were performed corresponding to the macula, midperiphery, and far periphery.

MAIN OUTCOME MEASURES

Changes in ischemic area and index (proportion of nonperfusion in analyzable retina), leakage area and index (proportion of leakage in analyzable retina), and MA count at baseline, week 12, week 24, and week 36 were analyzed. Spectral-domain-OCT quantitative metrics, such as central subfield thickness, ellipsoid zone (EZ) integrity parameters, intraretinal fluid (IRF) volume, and subretinal fluid (SRF) volume were extracted via a machine learning-enhanced OCT feature extraction platform and analyzed. Additionally, the effect of these changes on best-corrected visual acuity (BCVA) was evaluated.

RESULTS

Mean panretinal leakage index, zonal leakage area, and panretinal MA count improved significantly between baseline and week 36. Panretinal ischemic index decreased between baseline and week 36, with some aspects showing significant improvement. Mean BCVA significantly improved from baseline to week 36. There was a significant inverse correlation between change in BCVA and change in macular leakage area. A direct correlation was observed between both baseline macular leakage area and panretinal leakage index with IRF volume, SRF volume, and EZ disruption on OCT.

CONCLUSIONS

Assessment of UWFA parameters demonstrates a significant improvement in panretinal leakage index, leakage area, and MA burden in eyes treated with IAI with or without nesvacumab. A numeric reduction in panretinal ischemic index and area was noted. The analysis also shows the critical association of leakage with visual and OCT features. This highlights the potential role of UWFA in disease burden assessment, with leakage parameters serving as a primary end point.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found after the references.

Time to Resolution of Diabetic Macular Edema after Treatment with Intravitreal Aflibercept Injection or Laser in VISTA and VIVID

OBJECTIVE

To assess the relationship between baseline factors and time to diabetic macular edema (DME) resolution.

DESIGN

Post hoc analysis of VISTA and VIVID.

PARTICIPANTS

Eyes with baseline central subfield thickness (CST) of ≥ 290 μm.

INTERVENTION

Eyes were treated with intravitreal aflibercept injection (IAI) 2 mg (n = 558; every 4 weeks or every 8 weeks after 5 monthly doses) or laser control (n = 274). The effect of baseline factors on the time to DME resolution (CST < 290 μm) was assessed in univariable and multivariable models and further evaluated by the Kaplan-Meier method.

MAIN OUTCOME MEASURES

Time to and cumulative incidence of DME resolution.

RESULTS

Eyes treated with IAI had a 2.5-fold higher DME resolution rate, with median time of 33.0 weeks (95% confidence interval [CI], 28.1-40.0), compared with DME resolution not being achieved in 50% of eyes treated with laser control. Multivariable analysis demonstrated that a lower DME resolution rate was associated with a thicker baseline CST (hazard ratio [HR] [95% CI] per 100-μm CST increase, 0.79 [0.72-0.86]) and better baseline best-corrected visual acuity (BCVA) (HR [95% CI] per 5-letter increase, 0.87 [0.83-0.92]) with IAI. Tertiles of increasing CST (T1_CST ≤ 419 μm; T2_CST > 419 to ≤ 541; T3_CST > 541) were associated with longer median times to DME resolution (20.1, 39.1, and 49.1 weeks, respectively; P < 0.001 for T2_CST and T3_CST versus T1_CST) and lower cumulative incidence of events (HR, 1.0, 0.6, and 0.6, respectively; P < 0.001 for T2_CST and T3_CST versus T1_CST) with IAI. Tertiles of increasing BCVA (T1_BCVA ≤ 57 letters; T2_BCVA > 57 to ≤ 66; T3_BCVA >66) were associated with longer median times to DME resolution (28.4, 31.7, and 44.1 weeks, respectively; P < 0.05 for T3_BCVA versus T1_BCVA) and lower cumulative incidence of events (HR, 1.0, 0.9, and 0.8, respectively; P < 0.05 for T3_BCVA versus T1_BCVA) with IAI. No other baseline factor was associated with the time to DME resolution.

CONCLUSIONS

The median time to DME resolution was 33 weeks among IAI-treated eyes. A thicker baseline CST and better baseline BCVA in the IAI group were associated with a longer time to and a lower rate of DME resolution.

Machine Learning-Based Automated Detection and Quantification of Geographic Atrophy and Hypertransmission Defects Using Spectral Domain Optical Coherence Tomography

The current study describes the development and assessment of innovative, machine learning (ML)-based approaches for automated detection and pixel-accurate measurements of regions with geographic atrophy (GA) in late-stage age-related macular degeneration (AMD) using optical coherence tomography systems. 900 OCT volumes, 100266 B-scans, and en face OCT images from 341 non-exudative AMD patients with or without GA were included in this study from both Cirrus (Zeiss) and Spectralis (Heidelberg) OCT systems. B-scan and en face level ground truth GA masks were created on OCT B-scan where the segmented ellipsoid zone (EZ) line, retinal pigment epithelium (RPE) line, and bruchs membrane (BM) line overlapped. Two deep learning-based approaches, B-scan level and en face level, were trained. The OCT B-scan model had detection accuracy of 91% and GA area measurement accuracy of 94%. The en face OCT model had detection accuracy of 82% and GA area measurement accuracy of 96% with primary target of hypertransmission on en face OCT. Accuracy was good for both devices tested (92-97%). Automated lesion size stratification for CAM cRORA definition of 250um minimum lesion size was feasible. High-performance models for automatic detection and segmentation of GA area were achieved using OCT systems and deep learning. The automatic measurements showed high correlation with the ground truth. The en face model excelled at identification of hypertransmission defects. The models performance generalized well across device types tested. Future development will include integration of both models to enhance feature detection across GA lesions as well as isolating hypertransmission defects without GA for pre-GA biomarker extraction.

Anatomic Biomarkers of Macular Edema Associated with Retinal Vein Occlusion

PURPOSE

To assess the relationship between best-corrected visual acuity (BCVA) and anatomic features in patients with macular edema (ME) related to retinal vein occlusion (RVO).

DESIGN

Post hoc analysis of 3 clinical trials, which included verified diagnoses, protocol refractions, and the assessment of OCT and fluorescein angiography (FA) images at a masked reading center.

PARTICIPANTS

Patients diagnosed with RVO-ME.

METHODS

Correlation analyses were performed to determine the correlation between BCVA and macular anatomy at baseline and at 12 and 24 weeks and between changes from baseline to 12 and 24 weeks.

MAIN OUTCOME MEASURES

The correlations between BCVA and central subfield thickness (CST), ellipsoid zone (EZ) integrity, intraretinal fluid (IRF), subretinal fluid, central leakage, and ischemia were assessed.

RESULTS

In a total of 828 eyes with RVO-ME, the mean age, BCVA, and CST at baseline was 64.7 years, 51.1 letters, and 656.9 μm, respectively. At baseline, a moderate negative correlation was observed between BCVA and CST (r = - 0.56, P < 0.001). At weeks 12 and 24, the mean BCVA of eyes with definitely abnormal (absent) EZ was statistically significantly worse than that of eyes with normal EZ. At week 12, a moderate negative correlation was observed between changes in BCVA and changes in CST (r = - 0.35, P < 0.001), with a similar degree of association noted at week 24 (r = - 0.35, P < 0.001). At weeks 12 and 24, eyes that showed any improvement in central IRF showed a greater improvement in BCVA than eyes that showed no improvement worsening (week 12: 463 eyes, 18.3 letters vs. 177 eyes, 13.0 letters, respectively, P < 0.001) and (week 24: 332 eyes, 20.2 letters vs. 131 eyes, 13.3 letters, respectively, P < 0.001). With respect to the correlation between baseline BCVA and fluorescein leakage or capillary nonperfusion, the Pearson correlation coefficients were - 0.41 (P < 0.001) and - 0.16 (P = 0.060), respectively.

CONCLUSIONS

In addition to CST, there are important clinically relevant relationships between BCVA and both OCT and FA anatomic features in patients with RVO-ME.

Machine Learning-Based Automated Detection of Hydroxychloroquine Toxicity and Prediction of Future Toxicity Using Higher-Order OCT Biomarkers

OBJECTIVE

Despite guidelines for hydroxychloroquine (HCQ) toxicity screening, there are clear challenges to accurate detection and interpretation. In the current report, the feasibility of automated machine learning (ML)-based detection of HCQ retinopathy and prediction of progression to toxicity in eyes without preexisting toxicity has been described.

DESIGN

Retrospective, longitudinal cohort study.

SUBJECTS

Subjects on HCQ therapy.

METHODS

This was an institutional review board-approved, retrospective, longitudinal image analysis of 388 subjects on HCQ. Multilayer, compartmental, retinal segmentation with ellipsoid zone (EZ) mapping was used to harvest quantitative spectral-domain (SD)-OCT biomarkers. Using a combination of clinical features (i.e., cumulative HCQ dose and the duration of therapy) and quantitative imaging biomarkers (e.g., volumetric EZ integrity and compartmental measurements), ML models were created to detect toxicity and predict progression based on ground-truth OCT-based toxicity readings by 2 masked retina specialists. Furthermore, 10-fold cross-validation was performed.

MAIN OUTCOME MEASURES

The model performance was visualized using receiver operator curves and calculating the area under the curve (AUC). The corresponding sensitivity and specificity values were evaluated for the feasibility of HCQ toxicity screening and prediction.

RESULTS

The prevalence of HCQ toxicity in this cohort of 388 patients was 9.8% (n = 38). Twenty-one eyes progressed to toxicity during follow-up. OCT-based features (i.e., partial EZ attenuation, EZ volume, outer nuclear layer volume, and compartmental thicknesses) and clinical features (i.e., HCQ daily dose, HCQ cumulative dose, and duration of therapy) showed significant differences between the toxic and nontoxic groups. Percentage area with partial EZ attenuation (i.e., percentage of the macula with an EZ-retinal pigment epithelium thickness of ≤ 20 μm) was the most discriminating single feature (toxic, 35.7 ± 46.5%; nontoxic, 1.8 ± 4.4%; P < 0.0001). Using a random forest model, high-performance, automated toxicity detection was achieved, with a mean AUC of 0.97, sensitivity of 95%, and specificity of 91%. Furthermore, the toxicity progression prediction model had a mean AUC of 0.89, with a sensitivity and specificity of 90% and 80%, respectively.

CONCLUSIONS

This report described the feasibility of high-performance automated classification models that used a combination of clinical and quantitative SD-OCT biomarkers to detect HCQ retinal toxicity and predict progression to toxicity in cases without toxicity. Future work is needed to validate these findings in an independent dataset.

OCT-Derived Radiomic Features Predict Anti-VEGF Response and Durability in Neovascular Age-Related Macular Degeneration

PURPOSE

No established biomarkers currently exist for therapeutic efficacy and durability of anti-VEGF therapy in neovascular age-related macular degeneration (nAMD). This study evaluated radiomic-based quantitative OCT biomarkers that may be predictive of anti-VEGF treatment response and durability.

DESIGN

Assessment of baseline biomarkers using machine learning (ML) classifiers to predict tolerance to anti-VEGF therapy.

PARTICIPANTS

Eighty-one participants with treatment-naïve nAMD from the OSPREY study, including 15 super responders (patients who achieved and maintained retinal fluid resolution) and 66 non-super responders (patients who did not achieve or maintain retinal fluid resolution).

METHODS

A total of 962 texture-based radiomic features were extracted from fluid, subretinal hyperreflective material (SHRM), and different retinal tissue compartments of OCT scans. The top 8 features, chosen by the minimum redundancy maximum relevance feature selection method, were evaluated using 4 ML classifiers in a cross-validated approach to distinguish between the 2 patient groups. Longitudinal assessment of changes in different texture-based radiomic descriptors (delta-texture features) between baseline and month 3 also was performed to evaluate their association with treatment response. Additionally, 8 baseline clinical parameters and a combination of baseline OCT, delta-texture features, and the clinical parameters were evaluated in a cross-validated approach in terms of association with therapeutic response.

MAIN OUTCOME MEASURES

The cross-validated area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity were calculated to validate the classifier performance.

RESULTS

The cross-validated AUC by the quadratic discriminant analysis classifier was 0.75 ± 0.09 using texture-based baseline OCT features. The delta-texture features within different OCT compartments between baseline and month 3 yielded an AUC of 0.78 ± 0.08. The baseline clinical parameters sub-retinal pigment epithelium volume and intraretinal fluid volume yielded an AUC of 0.62 ± 0.07. When all the baseline, delta, and clinical features were combined, a statistically significant improvement in the classifier performance (AUC, 0.81 ± 0.07) was obtained.

CONCLUSIONS

Radiomic-based quantitative assessment of OCT images was shown to distinguish between super responders and non-super responders to anti-VEGF therapy in nAMD. The baseline fluid and SHRM delta-texture features were found to be most discriminating across groups.

Bacillary Detachment in Neovascular Age-Related Macular Degeneration: Incidence, Clinical Features, and Response to Anti-VEGF Therapy

OBJECTIVE

To evaluate the incidence of bacillary layer detachment among patients with neovascular age-related macular degeneration (nAMD) and their response to anti-VEGF therapy.

DESIGN

Post hoc analysis of the OSPREY clinical trial, a prospective, double-masked, phase II study comparing 6-mg brolucizumab with 2-mg aflibercept over 56 weeks.

PARTICIPANTS

Participants with treatment-naive nAMD at the initiation of the trial were included in the analysis (n = 81).

METHODS

Spectral-domain OCT (SD-OCT) scans were obtained at 4-week intervals throughout the OSPREY study and were segmented automatically using a proprietary, machine learning-enabled higher-order feature-extraction platform.

MAIN OUTCOME MEASURES

The presence of bacillary detachment, and in these eyes the effect of anti-VEGF therapy on change from baseline in visual acuity (VA), central subfield thickness (CST), retinal fluid volumes, subretinal hyper-reflective material (SHRM) volume, subretinal pigment epithelium (sub-RPE) fluid volume, and ellipsoid zone (EZ) integrity at week 56.

RESULTS

Bacillary detachment was identified in 7.4% (6 of 81) eyes, which had higher fluid volumes, increased CST, EZ attenuation, and increased sub-RPE volume at baseline compared with eyes without bacillary detachment. Anti-VEGF treatment resulted in the resolution of bacillary detachment in 100% of the eyes. In eyes with bacillary detachment at baseline, the anti-VEGF treatment decreased CST, fluid burden, and SHRM volumes throughout the treatment course; however, there was no significant change from baseline in VA, sub-RPE volume, or EZ integrity throughout the 56-week course of anti-VEGF treatment.

CONCLUSIONS

Bacillary detachment is an OCT signature that is identifiable in a notable proportion of nAMD eyes. Anti-VEGF therapy resulted in 100% resolution of bacillary detachment and significant decreases in CST and SHRM volume; however, improvements in VA may have been limited by persistent EZ attenuation.

Impact of Baseline Quantitative OCT Features on Response to Risuteganib for the Treatment of Dry Age-Related Macular Degeneration: The Importance of Outer Retinal Integrity

OBJECTIVE

The purpose of the study was to perform a post hoc analysis to explore the effect of baseline anatomic characteristics identified on OCT on best-corrected visual acuity (BCVA) responses to risuteganib from the completed phase II study in subjects with dry age-related macular degeneration (AMD).

DESIGN

Post hoc analysis of a randomized, double-masked, placebo-controlled, phase II study.

SUBJECTS

Eyes with intermediate dry AMD with BCVA between 20/40 and 20/200. Patients with concurrent vision-influencing or macula-obscuring ocular pathologies were excluded.

METHODS

Patients were randomized to receive a 1-mg intravitreal risuteganib injection or a sham injection at baseline. A second 1-mg intravitreal injection of risuteganib was given at week 16 to those in the treatment arm. Two independent, masked reading centers evaluated the baseline anatomic characteristics on OCT to explore features associated with positive responses to risuteganib.

MAIN OUTCOME MEASURES

Treatment response was defined as a gain of ≥ 8 letters in BCVA from baseline to week 28 in the treatment arm, compared with baseline to week 12 in the sham group. Anatomic parameters, measured by retinal segmentation platforms, including measures of retinal thickness were compared between the responders and nonresponders to risuteganib.

RESULTS

Thirty-nine patients completed the study and underwent analysis. In the treatment arm, 48% of eyes demonstrated treatment responses, compared with 7% in the sham group. In the quantitative anatomic assessment, enhanced ellipsoid integrity, greater outer retinal thickness, and decreased geographic atrophy were associated with increased BCVA gains to risuteganib.

CONCLUSIONS

This post hoc analysis demonstrated that baseline OCT features may help determine the likelihood of a functional response to risuteganib. The characterization of higher-order OCT features may provide important information regarding biomarkers for treatment response and could facilitate optimized clinical trial enrollment and enrichment.

The Association of Fluid Volatility With Subretinal Hyperreflective Material and Ellipsoid Zone Integrity in Neovascular AMD

Purpose

To evaluate the association of fluid volatility with ellipsoid zone (EZ) integrity and subretinal hyperreflective material (SHRM) volume during anti–vascular endothelial growth factor (VEGF) therapy in neovascular age-related macular degeneration (nAMD).

Methods

This study was a post hoc analysis of the OSPREY study. Retinal volatility was quantified as the standard deviation across weeks 12 to 56 for six optical coherence tomography (OCT) metrics: central subfield thickness (CST), total fluid (TF) volume, subretinal fluid (SRF) volume, intraretinal fluid (IRF) volume, macular total retinal fluid index (TRFI), and central macular TRFI. Eyes with volatility ≤ 25th or ≥ 75th percentile values were compared.

Results

Eyes with low volatility in several exudative metrics showed greater change from baseline in SHRM volume at week 12 than eyes with high volatility. During the maintenance phase (weeks 12–56), eyes exhibiting high SRF volatility demonstrated increased SHRM volume compared to eyes with low SRF volatility (P = 0.027). Eyes exhibiting high volatility in CST, TF, and SRF demonstrated less improvement in EZ total attenuation (P < 0.001, P = 0.033, and P = 0.043, respectively) than eyes with low volatility. Early exudative instability (i.e., between weeks 4–8 or weeks 8–12) in multiple parameters (i.e., CST, TF, IRF, macular TRFI, or central macular TRFI) was associated with greater volatility during the maintenance phase (P < 0.05).

Conclusions

Greater volatility in exudative OCT metrics, particularly SRF volatility, was associated with a greater increase in SHRM and less improvement in EZ integrity, suggesting that volatility is detrimental to multiple anatomic features in nAMD. Early exudative instability during the loading phase of treatment was associated with longer-term volatility in exudation.

Computational Imaging Biomarker Correlation with Intraocular Cytokine Expression in Diabetic Macular Edema

Purpose

Design

Participants

Methods

Main Outcome Measures

Results

Conclusions

Optical Coherence Tomography Anatomic and Temporal Biomarkers in Uveitic Macular Edema

PURPOSE

To assess the relationship between best-corrected visual acuity (BCVA) and optical coherence tomography (OCT) features in noninfectious uveitis (NIU)-related macular edema.

DESIGN

Clinical cohort study from post hoc analysis of 2 phase 3 clinical trials.

METHODS

Correlation and longitudinal treatment analyses were performed. Of 198 patients with NIU, 134 received suprachoroidal CLS-TA (proprietary formulation of a triamcinolone acetonide injectable suspension), and 64 received sham, with 12.9% and 72%, respectively, receiving rescue therapy.

RESULTS

At baseline, mean BCVA progressively worsened with each ordinal drop in central subfield ellipsoid zone (EZ) integrity. Eyes with normal baseline EZ experienced greater 24-week change in BCVA versus those with some degree of baseline EZ disruption (11.9 vs 9.4 letters, P = .006). In contrast, eyes with baseline central subfield cystoid spaces and/or subretinal fluid showed more improvement (13.7 or 17.2 letters, respectively) at 24 weeks, versus those without such findings (5.5 [P = .012] or 9.5 letters [P < .001], respectively). Longitudinal modeling for CLS-TA-treated eyes showed that central subfield thickness (CST) reached 90% of maximal improvement by week 3, whereas 90% maximal response in BCVA was not reached until week 9. CLS-TA-treated eyes that showed CST reduction of ≥50 µm at 4 weeks experienced a greater 24-week improvement in BCVA versus those without such an early response (14.6 vs 6.5 letters, P = .006 for difference).

CONCLUSIONS

Pretreatment EZ integrity and the presence of central subfield cystoid spaces or subretinal fluid each predict improved therapeutic response to treatment in eyes with NIU. In CLS-TA treated eyes, longitudinal modeling shows CST improvement preceding BCVA improvement.

LONGITUDINAL ELLIPSOID ZONE AND OUTER RETINAL INTEGRITY DYNAMICS AFTER EPIRETINAL MEMBRANE SURGERY

PURPOSE

To quantify ellipsoid zone (EZ) changes in integrity after epiretinal membrane (ERM) surgery, correlate findings to visual acuity, and determine predictors for prognosis.

METHODS

A post hoc analysis of eyes undergoing ERM surgery pooled from the prospective DISCOVER intraoperative optical coherence tomography study and eyes undergoing conventional ERM surgery without intraoperative optical coherence tomography. Quantitative EZ features were extracted using a multilayer machine learning enabled automated segmentation platform after image analyst review/correction for segmentation accuracy. Visual acuity and EZ integrity were quantitatively assessed and correlated before and after ERM surgery. Multiple linear regression was performed to assess preoperative visual acuity and EZ features as predictors for improvement in visual acuity or EZ integrity.

RESULTS

There were 177 eyes from 177 subjects that underwent ERM surgery from the DISCOVER and conventional arms. Improvement in visual acuity and multiple EZ integrity features was noted after ERM surgery, including EZ partial attenuation and EZ-retinal pigment epithelium (RPE) volume (P < 0.05). A reduction in EZ partial attenuation and increase in EZ-RPE central subfield thickness (EZ-RPE CST) was significantly correlated with improved visual acuity after ERM surgery (P < 0.05). More robust EZ-RPE CST at baseline predicted visual acuity improvement after ERM peel in regression modeling (β = 0.005, P < 0.05).

CONCLUSIONS

Longitudinal assessment of EZ features demonstrates significant postoperative improvement in multiple EZ integrity metrics after ERM surgery. Improving EZ integrity was correlated to improving the visual acuity. Ellipsoid zone integrity and visual acuity were significant predictors in regression modeling and may have value in clinical prognostication.

Quantitative Imaging Biomarkers in Age-Related Macular Degeneration and Diabetic Eye Disease: A Step Closer to Precision Medicine

The management of retinal diseases relies heavily on digital imaging data, including optical coherence tomography (OCT) and fluorescein angiography (FA). Targeted feature extraction and the objective quantification of features provide important opportunities in biomarker discovery, disease burden assessment, and predicting treatment response. Additional important advantages include increased objectivity in interpretation, longitudinal tracking, and ability to incorporate computational models to create automated diagnostic and clinical decision support systems. Advances in computational technology, including deep learning and radiomics, open new doors for developing an imaging phenotype that may provide in-depth personalized disease characterization and enhance opportunities in precision medicine. In this review, we summarize current quantitative and radiomic imaging biomarkers described in the literature for age-related macular degeneration and diabetic eye disease using imaging modalities such as OCT, FA, and OCT angiography (OCTA). Various approaches used to identify and extract these biomarkers that utilize artificial intelligence and deep learning are also summarized in this review. These quantifiable biomarkers and automated approaches have unleashed new frontiers of personalized medicine where treatments are tailored, based on patient-specific longitudinally trackable biomarkers, and response monitoring can be achieved with a high degree of accuracy.

Ellipsoid Zone Integrity and Visual Acuity Changes during Diabetic Macular Edema Therapy: A Longitudinal Study

PURPOSE

Ellipsoid zone (EZ) integrity is identified as a potential biomarker for therapy surveillance and outcome prediction of visual acuity (VA). However, only a few studies report long-term results of over 1 year of clinical and anatomical changes in patients with diabetic macular edema (DME). This study is aimed at describing the long-term VA and anatomical outcomes in spectral domain optical coherence tomography (OCT) (relative ellipsoid zone reflectivity ratio, central macular thickness, and volume) in patients with DME treated with antivascular endothelial growth factor (anti-VEGF) therapy. Furthermore, we studied the correlation between EZ integrity and changes in visual acuity.

METHODS

71 eyes of 71 patients were included in this retrospective study. Clinical characteristics were reviewed yearly. OCT data were assessed at baseline and after 1, 3, and 5 years. EZ parameters were quantified automatically. OCT parameters and visual outcome were correlated and analyzed in multivariable regression models.

RESULTS

EZ reflectivity ratio correlated with functional outcome in DME patients from baseline to fifth year at all time points (for all p < 0.05). EZ reflectivity improved the most in the first year of treatment (0.68 to 0.75; p < 0.05) and declined gradually until year 5 of therapy (0.71; compared to baseline p > 0.05). Similarly, best VA was achieved after 1 year (0.40 logarithm of the minimum angle of resolution (logMAR) to 0.28 logMAR; p < 0.001) and declined gradually until year 5. Final VA in year 5 was comparable to baseline (0.45 logMAR, compared to baseline p > 0.05). Together with baseline VA, baseline EZ parameters did predict VA outcome after 1 year (p < 0.05). Concordantly, VA and EZ parameters from year 1 were associated with VA outcome in year 2.

CONCLUSION

This study described the long-term course of EZ changes during anti-VEGF treatment in DME patients. In addition, our results underlined the potential of EZ parameters as novel OCT biomarkers for prediction of VA outcomes during therapy.

Multi-Compartment Spatially-Derived Radiomics From Optical Coherence Tomography Predict Anti-VEGF Treatment Durability in Macular Edema Secondary to Retinal Vascular Disease: Preliminary Findings

OBJECTIVE

Diabetic macular edema (DME) and retinal vein occlusion (RVO) are the leading causes of visual impairments across the world. Vascular endothelial growth factor (VEGF) stimulates breakdown of blood-retinal barrier that causes accumulation of fluid within macula. Anti-VEGF therapy is the first-line treatment for both the diseases; however, the degree of response varies for individual patients. The main objective of this work was to identify the (i) texture-based radiomics features within individual fluid and retinal tissue compartments of baseline spectral-domain optical coherence tomography (SD-OCT) images and (ii) the specific spatial compartments that contribute most pertinent features for predicting therapeutic response.

METHODS

A total of 962 texture-based radiomics features were extracted from each of the fluid and retinal tissue compartments of OCT images, obtained from the PERMEATE study. Top-performing features selected from the consensus of different feature selection methods were evaluated in conjunction with four different machine learning classifiers: Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), Random Forest (RF), and Support Vector Machine (SVM) in a cross-validated approach to distinguish eyes tolerating extended interval dosing (non-rebounders) and those requiring more frequent dosing (rebounders).

RESULTS

Combination of fluid and retinal tissue features yielded a cross-validated area under receiver operating characteristic curve (AUC) of 0.78±0.08 in distinguishing rebounders from non-rebounders.

CONCLUSIONS

This study revealed that the texture-based radiomics features pertaining to IRF subcompartment were most discriminating between rebounders and non-rebounders to anti-VEGF therapy. Clinical Impact: With further validation, OCT-based imaging biomarkers could be used for treatment management of DME patients.

Longitudinal Higher-Order OCT Assessment of Quantitative Fluid Dynamics and the Total Retinal Fluid Index in Neovascular AMD

Purpose

The purpose of this study was to evaluate the feasibility of assessing quantitative longitudinal fluid dynamics and total retinal fluid indices (TRFIs) with higher-order optical coherence tomography (OCT) for neovascular age-related macular degeneration (nAMD).

Methods

A post hoc image analysis study was performed using the phase II OSPREY clinical trial comparing brolucizumab and aflibercept in nAMD. Higher-order OCT analysis using a machine learning−enabled fluid feature extraction platform was used to segment intraretinal fluid (IRF) and subretinal fluid (SRF) volumetric components. TRFI, the proportion of fluid volume against total retinal volume, was calculated. Longitudinal fluid metrics were evaluated for the following groups: all subjects (i.e. treatment agnostic), brolucizumab, and aflibercept.

Results

Mean IRF and SRF volumes were significantly reduced from baseline at each timepoint for all groups. Fluid feature extraction allowed high-resolution assessment of quantitative fluid burden. A greater proportion of brolucizumab participants achieved true zero and minimal fluid (total fluid volume between 0.0001 and 0.001mm³) versus aflibercept participants at week 40. True zero fluid during q12 brolucizumab dosing was achieved in 36.6% to 38.5%, similar to the 25.6% to 38.5% during the corresponding q8 aflibercept cycles. TRFI was significantly reduced from baseline in all groups.

Conclusions

Higher-order OCT analysis demonstrates the feasibility of fluid feature extraction and longitudinal volumetric fluid burden and TRFI characterization in nAMD, supporting a unique opportunity for fluid burden assessment and the impact on outcomes.

Translational Relevance

Detection and characterization of disease activity is vital for optimal treatment of nAMD. Longitudinal assessment of fluid dynamics and the TRFI provide important proof of concept for future automated tools in characterizing disease activity.

AI-based monitoring of retinal fluid in disease activity and under therapy

Retinal fluid as the major biomarker in exudative macular disease is accurately visualized by high-resolution three-dimensional optical coherence tomography (OCT), which is used world-wide as a diagnostic gold standard largely replacing clinical examination. Artificial intelligence (AI) with its capability to objectively identify, localize and quantify fluid introduces fully automated tools into OCT imaging for personalized disease management. Deep learning performance has already proven superior to human experts, including physicians and certified readers, in terms of accuracy and speed. Reproducible measurement of retinal fluid relies on precise AI-based segmentation methods that assign a label to each OCT voxel denoting its fluid type such as intraretinal fluid (IRF) and subretinal fluid (SRF) or pigment epithelial detachment (PED) and its location within the central 1-, 3- and 6-mm macular area. Such reliable analysis is most relevant to reflect differences in pathophysiological mechanisms and impacts on retinal function, and the dynamics of fluid resolution during therapy with different regimens and substances. Yet, an in-depth understanding of the mode of action of supervised and unsupervised learning, the functionality of a convolutional neural net (CNN) and various network architectures is needed. Greater insight regarding adequate methods for performance, validation assessment, and device- and scanning-pattern-dependent variations is necessary to empower ophthalmologists to become qualified AI users. Fluid/function correlation can lead to a better definition of valid fluid variables relevant for optimal outcomes on an individual and a population level. AI-based fluid analysis opens the way for precision medicine in real-world practice of the leading retinal diseases of modern times.

Imaging Features of Vessels and Leakage Patterns Predict Extended Interval Aflibercept Dosing Using Ultra-Widefield Angiography in Retinal Vascular Disease: Findings From the PERMEATE Study

Diabetic Macular Edema (DME) and macular edema secondary to retinal occlusion (RVO) are the two most common retinal vascular causes of visual impairment and leading cause of worldwide vision loss. The blood-retinal barrier is the key barrier for maintaining fluid balance within the retinal tissue. Vascular Endothelial Growth Factor (VEGF) has a significant role in the permeability of the blood-retinal barrier, which also leads to appearance of leakage foci. Intravitreal anti-VEGF therapy is the current gold standard treatment and has been demonstrated to improve macular thickening, improve vision acuity and reduce vascular leakage. However, treatment response and required dosing interval can vary widely across patients. Given the role of the blood-retinal barrier and vascular leakage in the pathogenesis of these disorders, the goal of this study was to present and evaluate new computer extracted features relating to morphology, spatial architecture and tortuosity of vessels and leakages from baseline ultra-widefield fluorescein angiography (UWFA) images. Specifically, we sought to evaluate the role of these computer extracted features from baseline UWFA images. Notably, these UWFA images were obtained from IRB-approved PERMEATE clinical trial [1], [2] to distinguish eyes tolerating extended dosing intervals (n = 16) who are referred to as non-rebounders and those who require more frequent dosing (n = 12) and are called rebounders based on visual acuity loss with extended dosing challenges. A total of 64 features encapsulating different morphological and geometrical attributes of leakage patches including the anatomical (shape, size, density, area, minor and major axis, orientation, area, extent ratio, perimeter, radii) and geometrical characteristics (the proximity of each leakage foci to main vessels, to other leakage foci and to optical disc) as well as 54 tortuosity features (tortuosity of whole vessel network, local tortuosity of vessels in the vicinity of leakage foci) were extracted. The most significant and predictive biomarkers related to treatment response were proximity of leakage nodes to major and minor eye vessels as well as local vasculature tortuosity in the vicinity of the leakages. The imaging features were then used in conjunction with a Linear Discriminant Analysis (LDA) classifier to distinguish rebounders from non-rebounders. The 3-fold cross-validated Area Under Curve (AUC) was found to be 0.82 for the morphological based features and 0.85 for the tortuosity based features. Our findings suggest higher variation in leakage node proximity to retinal vessels in eyes tolerating extended interval dosing. In contrast, eyes with increased local vascular tortuosity demonstrated less tolerance of increased dosing interval. Moreover, a class activation map generated by a deep learning model identified regions that corresponded to regions of leakages proximal to the vessels, providing confirmation of the validity of predictive image features extracted from these regions in this study.

Indicators of Visual Prognosis in Diabetic Macular Oedema

Diabetic macular oedema (DMO) is an important cause of moderate vision loss in people with diabetes. Advances in imaging technology have shown that a significant proportion of patients with DMO respond sub-optimally to existing treatment options. Identifying associations and predictors of response before treatment is initiated may help in explaining visual prognosis to patients and aid the development of personalized treatment strategies. Imaging features, such as central subfoveal thickness, photoreceptor integrity, disorganization of retinal inner layers, choroidal changes, and macular perfusion, have been reported to be prognostic factors of visual acuity (VA) in DMO. In this review we evaluated each risk factor to understand their relative importance in visual prognostication of DMO eyes post-treatment. Although individually, some of these factors may not be significant predictors, in combination they may form phenotypes that can inform visual prognosis. Stratification based on these phenotypes needs to be developed to progress to personalized medicine for DMO.

Retinal Fluid Volatility Associated With Interval Tolerance and Visual Outcomes in Diabetic Macular Edema in the VISTA Phase III Trial

PURPOSE

To describe longitudinal retinal fluid dynamics on spectral domain OCT and to identify imaging biomarkers that predict the worsening of DME with interval extension during anti-vascular endothelial growth factor (VEGF) therapy.

DESIGN

A post hoc sub-analysis of phase III, VISTA-DME study.

METHODS

Eyes received either intravitreal aflibercept injection 2 mg every 4 weeks (2q4) or every 8 weeks after 5 initial monthly injections (2q8), and eyes imaged with the Cirrus HD-OCT system were included. The macular cube was analyzed for 10 time-points from baseline through week 100. Retinal OCT images were evaluated using a novel software platform to extract retinal fluid features for calculation of volumetric fluid parameters, including the retinal fluid index (RFI): the percentage of retinal volume that was occupied by intraretinal fluid.

RESULTS

Fifty-five eyes were included in the 2q4 group, and 58 eyes were included in the 2q8 group. Early RFI volatility with a central macular RFI increase by ≥5 points from week 4 to 8 (P = .004, odds ratio [OR] 31.3, 95% confidence interval [CI] 3.0 to 329) and cumulative RFI volatility with an aggregate increase in macular RFI by ≥10 points from those timepoints with increased RFI between baseline to week 20, P = .005, OR 10.2, 95% CI 2.1 to 51.3) were both significant predictors for the worsening of DME and visual acuity when the treatment interval was extended to 8 weeks in the 2q8 group.

CONCLUSIONS

Early fluid dynamics as measured by (1) early RFI volatility and (2) cumulative RFI instability with aggregate increased RFI were associated with intolerance of interval extension.

Patterns of ellipsoid zone change associated with visual outcome for diabetic macular oedema

Clinical relevance: Optical coherence tomography biomarkers are dynamic and possibly change over time. The dynamic biomarkers may better predict improvement of final vision than those at baseline for diabetic macular oedema after treatment.Background: To investigate predictors for ellipsoid zone (EZ) dynamics and the association with visual acuity after treatment for diabetic macular oedemaMethods: A total of 135 eyes from 135 patients with anti-vascular endothelial growth factor treatment for diabetic macular oedema were retrospectively enrolled in this study. EZ status was evaluated by optical coherence tomography at baseline, 1, 3, 6, and 12 months post-operatively. Macular perfusion status was analysed by fluorescein angiography. Duration of EZ disruption was quantified by giving scores from 0 to 5 based on the sum of visits with evidence of EZ disruption. Linear regression models were used to predict factors for scores of EZ disruption. Binary logistic regression was used to predict various EZ biomarkers for final visual acuity improvement.Results: Thirty of 87 eyes (34.4%) with intact EZ at baseline developed subsequent EZ disruption. For those with EZ disruption at baseline, 17 of 48 (35.4%) patients had restoration of disrupted EZ over time. Non-perfused macula was associated with both earlier development of EZ disruption and poor restoration (p < 0.001 and p = 0.011, respectively). Non-perfused macula, absence of epiretinal membrane, and EZ disruption at baseline predicted higher EZ scores (p < 0.001, p = 0.022 and p < 0.001, respectively). Higher EZ scores (p = 0.016), but not baseline EZ disruption (p = 0.56), were less likely to have improvement of final vision 5 letters or more.Conclusion: Macular perfusion status played an independent role in EZ dynamics. Duration of EZ disruption could be more predictive than baseline status for improvement of vision over 12 months.

Epiretinal Membrane Surgery Using Intraoperative OCT-Guided Membrane Removal in the DISCOVER Study versus Conventional Membrane Removal

PURPOSE

To provide a comparative assessment of clinical outcomes between patients undergoing intraoperative OCT (iOCT) and conventional surgery for pars plana vitrectomy (PPV) with epiretinal membrane (ERM) peel.

DESIGN

Case-control retrospective, comparative assessment.

PARTICIPANTS

Patients undergoing PPV with membrane peel for ERM with eyes pooled from the prospective Determination of Feasibility of Intraoperative Spectral Domain Microscope Combined/Integrated OCT Visualization During En Face Retinal and Ophthalmic Surgery (DISCOVER) iOCT study and eyes undergoing conventional ERM surgery without iOCT.

METHODS

Visual acuity and OCT assessment before ERM surgery and at 1-, 3-, 6-, and 12-month follow-up after standard small-gauge PPV with iOCT feedback (iOCT DISCOVER group) or PPV with compulsory internal limiting membrane (ILM) peeling (conventional group). Visual acuity, central subfield thickness (CST), reoperation rate, and ERM recurrence were determined by record review and post hoc assessment of clinical OCTs after ERM peel.

MAIN OUTCOME MEASURES

Visual acuity and ERM recurrence.

RESULTS

A total of 262 eyes were included. Visual acuity (VA) improved 11.9 letters in the iOCT group (P < 0.0001) and 12.1 letters in the conventional group (P < 0.0001) at 12 months after ERM surgery. Visual acuity improvement did not differ between the iOCT and conventional groups at 1, 3, 6, or 12 months after surgery (P > 0.05 for each time point). Preoperative mean CST decreased in the iOCT group (P < 0.0001) and conventional group (P < 0.0001) with no difference between groups in CST reduction at 12 months (P = 0.36). No reoperations or visually significant recurrent ERMs occurred in either cohort.

CONCLUSIONS

Intraoperative OCT-guided ERM removal without mandated ILM peeling provided similar VA and anatomic results to conventional ILM peeling for ERM. Future randomized prospective studies are needed to assess fully the possible role of iOCT in ERM surgery and to evaluate the potential impact of nonfoveal ERM persistence or recurrence in comparison with conventional surgery.

Longitudinal Assessment of Ellipsoid Zone Integrity, Subretinal Hyperreflective Material, and Subretinal Pigment Epithelium Disease in Neovascular Age-Related Macular Degeneration

PURPOSE

To assess longitudinally the effect of anti-vascular endothelial growth factor (VEGF) treatment on ellipsoid zone (EZ) integrity, subretinal hyperreflective material (SHRM), and the sub-retinal pigment epithelium (sub-RPE) compartment in eyes with neovascular age-related macular degeneration (nAMD).

DESIGN

Post hoc analysis of the OSPREY clinical trial, a prospective, double-masked, phase 2 study comparing brolucizumab 6 mg with aflibercept 2 mg over 56 weeks.

PARTICIPANTS

Participants with treatment-naïve nAMD at the initiation of the trial were included in the analysis.

METHODS

Eyes were evaluated with spectral-domain OCT at 4-week intervals in the OSPREY trial (n = 81). Spectral-domain OCT scans collected from each visit were segmented automatically using a proprietary, machine learning-enabled higher-order feature-extraction platform for retinal layer, SHRM, and sub-RPE boundary lines, which were evaluated and corrected as needed by masked trained graders. The current analysis focused only on patients evaluated with the Cirrus (Zeiss) platform (n = 28).

MAIN OUTCOME MEASURES

Outcome measures included change from baseline in EZ-RPE (i.e., photoreceptor outer segment) volume, EZ-RPE central subfield thickness (CST), total EZ attenuation, SHRM volume, SHRM CST, and total sub-RPE volume. The correlation between each of these measures and best-corrected visual acuity (BCVA) at each visit was evaluated.

RESULTS

EZ-RPE volume and EZ-RPE CST showed significant increases, and total EZ attenuation, SHRM volume, SHRM CST, and total sub-RPE volume showed significant decreases from baseline at each visit from weeks 4 through 56 (P < 0.05 at each visit). Ellipsoid zone integrity measures and SHRM volume correlated significantly with BCVA at most visits (P < 0.05). No significant correlation was found between total sub-RPE volume and BCVA.

CONCLUSIONS

EZ integrity, SHRM, and sub-RPE disease features in eyes with nAMD showed improvement as early as week 4 of anti-VEGF treatment. EZ integrity measures and SHRM volume were predictors of visual acuity over the first year of treatment.

Aqueous Cytokine Expression and Higher Order OCT Biomarkers: Assessment of the Anatomic-Biologic Bridge in the IMAGINE DME Study

PURPOSE

To identify biomarkers for predicting response to anti-vascular endothelial growth factor (VEGF) therapy in diabetic macular edema (DME) and evaluate any links between cytokine expression and optical coherence tomography (OCT) phenotype.

DESIGN

The IMAGINE is a post hoc image analysis and cytokine expression assessment of the Efficacy & Safety Trial of Intravitreal Injections Combined With PRP for CSME Secondary to Diabetes Mellitus (DAVE) randomized clinical trial.

METHODS

Subjects were categorized as anatomical responders or nonresponders, and within the responder group as rebounders and non-rebounders based on quantitative, longitudinal OCT criteria. Retinal layer and fluid features were extracted using an OCT machine-learning augmented segmentation platform. Responders were further sub-classified by rapidity of response. Aqueous concentrations of 54 cytokines were measured at multiple timepoints. Expression was compared between responder groups and correlated with OCT imaging biomarkers.

RESULTS

Of the 24 eyes studied, 79% were anatomical responders with 38% super responders, 17% early responders, and 25% slow responders. Twenty-one percent were nonresponders. Super responders had increased baseline vascular endothelial growth factor (VEGF) (880.0 pg/mL vs 245.4 pg/mL; P = .012) and decreased monocyte chemotactic protein-1 (MCP-1) (513.3 pg/mL vs 809.5 pg/mL; P = .0.042) concentrations compared with nonresponders. Interleukin-6 (-24.9 pg/mL vs 442.8 pg/mL; P = .032) concentrations increased among nonresponders during therapy. VEGF concentrations correlated with central subfield thickness (r = 0.49; P = .01). Panmacular retinal volume correlated with increased interleuckin-6 (r = 0.47; P = .02) and decreased MCP-1 (r = -0.45; P = .03). Matrix metallopeptidase-1 correlated with subretinal fluid volume (r = 0.50; P = .01).

CONCLUSIONS

OCT imaging biomarkers correlated with both intraocular cytokines and responsiveness to anti-VEGF therapy, which indicated a possible link to underlying pathways and their relevance to DME prognosis. Baseline concentrations of VEGF and MCP-1 are associated with anatomic response to anti-VEGF therapy.

Visual Acuity in Retinal Vein Occlusion, Diabetic, and Uveitic Macular Edema: Central Subfield Thickness and Ellipsoid Zone Analysis

PURPOSE

This study assessed relationships between best-corrected visual acuity (BCVA), central subfield thickness (CST), and ellipsoid zone (EZ) integrity in macular edema (ME) patients.

DESIGN

Post hoc analysis of 6 clinical trials, which included verified diagnoses, protocol refractions, and reading center assessment of OCT images.

PARTICIPANTS

Participants (n = 1063) were diagnosed with ME from retinal vein occlusion (RVO), diabetic retinopathy (DR; diabetic macular edema, DME), or noninfectious uveitis (NIU).

METHODS

For CST, 2 clinical trials for each disorder were analyzed. For EZ, 3 studies across 2 disorders were analyzed.

MAIN OUTCOME MEASURES

Primary outcomes were correlations between BCVA and CST, and between BCVA and 4 central subfield EZ grades.

RESULTS

For baseline BCVA and CST, Pearson correlation coefficients were: ME from RVO, -0.56 (774 eyes; 95% confidence interval [CI], -0.61 to -0.51; P < 0.001); DME, -0.50 (91 eyes; 95% CI, -0.64 to -0.33; P < 0.001); and ME from NIU, -0.38 (198 eyes; 95% CI, -0.49 to -0.26; P < 0.001). Regarding change from baseline to 24 weeks for both BCVA and CST, Pearson correlation coefficients were: ME from RVO, -0.35 (95% CI, -0.43 to -0.27; P < 0.001); DME, -0.30 (95% CI, -0.48 to -0.09; P = 0.006); and ME from NIU, -0.42 (95% CI, -0.53 to -0.29; P < 0.001). Acute and chronic ME showed similar baseline and 24-week change linear correlations. With lower baseline CST, a trend of decreased baseline and 24-week change correlations was found. For central subfield EZ at baseline, mean BCVA progressively worsened with each of 4 EZ grades in 185 eyes with gradable EZ (DME, 41 eyes; NIU, 144 eyes; P ≤ 0.050 for all pairwise comparisons except between normal and questionably abnormal EZ grades). Eyes with normal baseline central subfield EZ showed greater 24-week change in BCVA than those with abnormal baseline EZ (15.00 letters vs. 8.16 letters; P = 0.0005, with baseline BCVA, CST, and age as covariates).

CONCLUSIONS

Despite these correlations, CST and EZ integrity, as graded herein, account for the minority of BCVA variation in patients with ME resulting from RVO, DR, and NIU.