PREDICTIVE FACTORS FOR DEVELOPMENT OF NEOVASCULAR AGE-RELATED MACULAR DEGENERATION: A Spectral-Domain Optical Coherence Tomography Study

Imaging the eye as a window to brain health: frontier approaches and future directions

Recent years have seen significant advances in diagnostic testing of central nervous system (CNS) function and disease. However, there remain challenges in developing a comprehensive suite of non- or minimally invasive assays of neural health and disease progression. Due to the direct connection with the CNS, structural changes in the neural retina, retinal vasculature and morphological changes in retinal immune cells can occur in parallel with disease conditions in the brain. The retina can also, uniquely, be assessed directly and non-invasively. For these reasons, the retina may prove to be an important "window" for revealing and understanding brain disease. In this review, we discuss the gross anatomy of the eye, focusing on the sensory and non-sensory cells of the retina, especially microglia, that lend themselves to diagnosing brain disease by imaging the retina. We include a history of ocular imaging to describe the different imaging approaches undertaken in the past and outline current and emerging technologies including retinal autofluorescence imaging, Raman spectroscopy, and artificial intelligence image analysis. These new technologies show promising potential for retinal imaging to be used as a tool for the diagnosis of brain disorders such as Alzheimer's disease and others and the assessment of treatment success.

Evaluation of inflammatory hyperreflective foci and plasma EPA as diagnostic and predictive markers for age-related macular degeneration

Objectives

To detect the plasma polyunsaturated fatty acids (PUFAs) concentrations in age-related macular degeneration (AMD) patients and healthy controls. Additionally, advanced studies were conducted to investigate the relationship between PUFAs concentrations and ophthalmological characteristics, including hyperreflective foci (HRF), visual acuity, and anti-vascular endothelial growth factor (anti-VEGF) response in patients with AMD.

Methods

This prospective, single-site study recruited a total of 315 participants, consisting of 105 individuals with dry AMD (early-stage AMD group), 105 individuals with neovascular AMD (late-stage AMD group), and 105 elderly individuals without any fundus diseases (healthy controls). The levels of omega-3 and omega-6 PUFAs in plasma were detected using gas chromatography. Retinal thickness, choroidal thickness, and macular volume were quantified using optical coherence tomography angiography (OCTA) scan with a 6 × 6 mm macular area, and the amounts of HRF were analyzed with OCTA scanning data.

Results

Compared to the control group, AMD patients exhibited significantly lower plasma concentrations of eicosapentaenoic acid (EPA), docosahexaenoic acid (DHA) and alpha linolenic acid. HRF were observed in various retinal layers of AMD patients, particularly those with late-stage AMD. The correlation coefficient matrix and multiple linear regression models demonstrated that HRF played a crucial role in best corrected visual acuity for both early (p < 0.001) and late-stage AMD patients (p = 0.006), while EPA had an inverse effect on the logarithm of the minimum angle of resolution (logMAR) value in patients with early-stage AMD (p < 0.001). As compared to patients with good responses to anti-VEGF therapy, those with poor responses had significantly lower baseline logMAR (p < 0.001), central retina thickness (p = 0.002), macular volume (p = 0.027), HRF (p = 0.024), and plasma EPA (p < 0.001). This study used a ROC curve analysis to identify the combination of HRF and EPA as a potential biomarker for predicting the response to anti-VEGF treatment in late-stage AMD patients, with an area under the curve (AUC) value of 0.775.

Conclusions

Reduced plasma EPA was detected in AMD cases and the lower EPA concentration was related to poorer visual acuity. Additionally, the quantity of HRF combined with concentration of plasma EPA may serve as the prognostic indicator for predicting the effect of anti-VEGF treatment in late-stage AMD patients.

Retinal Pigment Epithelium Curvature Can Predict Late Age-Related Macular Degeneration

Purpose

Outer retinal band integrity strongly predicts late age-related macular degeneration (AMD). However, it is often assessed subjectively as "continuity" using inconsistent definitions. Alternatively, "curvature" of the outer retinal bands is a quantitative metric that strongly correlates with AMD biomarkers and can screen for intermediate AMD. We evaluated the prognostic ability of retinal pigment epithelium (RPE) and ellipsoid zone (EZ) curvature for late AMD against outer retinal band continuity, pigmentary abnormalities, reticular pseudodrusen, and drusen volume.

Methods

Consecutive patients with intermediate AMD who progressed to late AMD (n = 17) or remained stable (n = 42) were recruited. RPE and EZ curvature were quantified as a ratio of their lengths over Bruch's membrane using the sinuosity method of assessing river curvature, where a ratio of ∼1 indicates no outer retinal pathology. RPE, EZ, and Bruch's membrane were manually segmented and their lengths automatically extracted. The primary outcomes were outer retinal sinuosity and the odds ratio of predicting late AMD.

Results

Mean follow-up time for progressors and nonprogressors was 4.4 and 3.6 years. RPE sinuosity was strongly associated with pigmentary abnormalities (P = 0.001) and drusen volume (P = 0.004) but not reticular pseudodrusen (P = 0.28). RPE sinuosity >1.03 was the strongest predictor of late AMD developing within 5 years (15 [2.9-75]) and across the study period (25 [2.3-282]). Drusen volume >0.03 mm3 was the strongest predictor of progression within 2 years (33 [2.5-426]), and RPD could not independently predict progression within any time frame.

Conclusions

RPE curvature is a promising, quantitative outer retinal biomarker that can prognosticate late AMD and potentially enhance prognostic models.

Analysis of prognostic and predictive factors in neovascular age-related macular degeneration Kuopio cohort

PURPOSE

The aim of the study was to explore factors affecting the progression of neovascular age-related macular degeneration (nAMD) and identify predictive factors that can estimate the duration of intravitreal treatments.

METHODS

This retrospective real-world study included 421 nAMD patients treated at the Kuopio University Hospital during years 2007-2021. The collected data included background demographics, treatment history, visual acuity and retinal biomarker analysis. Impact of baseline factors on age at diagnosis, treatment duration, received treatment intensity and visual acuity gains were analysed.

RESULTS

Heavy smoking and high body mass index (BMI) were associated with an earlier onset, while the use of anticoagulation and anti-aggregation medication were associated with a later onset of nAMD. A low number of injections during the first year of treatment and the presence of intraretinal fluid (IRF) at baseline were associated with shorter treatment duration. Interestingly, when IRF only patients were compared to subretinal fluid (SRF) only patients, IRF patients showed higher occurrences of subretinal drusenoid deposits (43.5% vs. 15%, p = 0.04). In addition, when all patients with IRF were compared to SRF only patients, more hyperreflective foci (HRF) and complete RPE and outer retinal atrophy (cRORA; 20.7% vs. 5%, p = 0.02) were observed in patients with IRF.

CONCLUSIONS

Our results reveal that heavy smoking and high BMI are accelerating factors for earlier emergence of nAMD, while the presence of IRF results in a fast-progressing disease. More intriguingly, the link between IRF and appearance of subretinal drusenoid deposits, HRF, and increased retinal atrophy was observed.

First-Year Real-Life Experience with Intravitreal Faricimab for Refractory Neovascular Age-Related Macular Degeneration

Background: To evaluate the outcomes of intravitreal faricimab (IVF) for refractory neovascular age-related macular degeneration (nAMD) and investigate the impact of baseline optical coherence tomography, biomarkers for total IVF injections are needed. Methods: A retrospective analysis of 33 eyes of patients who completed one year (52 W) of treatment with IVF. The eyes received four IVF injections (6 mg/0.05 mL) as the upload phase. Thereafter, the treatment interval was extended to 8 or 12 weeks if disease activity was not recorded. The outcome measures included best-corrected visual acuity (BCVA), central macular thickness (CMT), subfoveal choroidal thickness (SFCT), and retinal fluid distribution. Results: A total of 33 eyes were included. CMT decreased significantly at 52 W (p < 0.01). BCVA and SFCT did not change significantly at 52 W (p > 0.05). The number of eyes with subretinal fluid decreased significantly at 52 W (p < 0.01). Complete fluid resolution was achieved in 20 eyes (60%). The total number of injections was significantly negatively correlated with the presence of hyperreflective dots at baseline (HRDs, p < 0.01) and SFCT at baseline (p < 0.01). Conclusions: IVF led to a significant reduction in CMT with stabilization of BCVA. The total number of injections was lower in eyes with HRDs and increased SFCT at baseline. This might provide clues regarding response to IVF for future studies.

Associations of presenting visual acuity with morphological changes on OCT in neovascular age-related macular degeneration: PRECISE Study Report 2

PURPOSE

To study associations of optical coherence tomography (OCT) features with presenting visual acuity (VA) in treatment naive neovascular age-related macular degeneration (nAMD).

METHODS

Patients with nAMD initiated on aflibercept therapy were recruited from December 2019 to August 2021. Demographic and OCT (Spectralis, Heidelberg Engineering) features associated with good VA (VA ≥ 68 ETDRS letters, Snellen ≥ 6/12) and poor VA (VA < 54 letters, Snellen < 6/18) were analysed using Generalised Estimating Equations to account for inter-eye correlation.

RESULTS

Of 2274 eyes of 2128 patients enrolled, 2039 eyes of 1901 patients with complete data were analysed. Mean age was 79.4 (SD 7.8) years, female:male 3:2 and mean VA 58.0 (SD 14.5) letters. On multivariable analysis VA < 54 letters was associated with increased central subfield thickness (CST) (OR 1.40 per 100 µm; P < 0.001), foveal intraretinal fluid (OR 2.14; P < 0.001), polypoidal vasculopathy (PCV) relative to Type 1 macular neovascularisation (MNV) (OR 1.66; P = 0.049), presence of foveal subretinal hyperreflective material (SHRM) (OR 1.73; P = 0.002), foveal fibrosis (OR 3.85; P < 0.001), foveal atrophy (OR 5.54; P < 0.001), loss of integrity of the foveal ellipsoid zone (EZ) or external limiting membrane (ELM) relative to their preservation (OR 3.83; P < 0.001) and absence of subretinal drusenoid deposits (SDD) (presence vs absence; OR 0.75; P = 0.04). These features were associated with reduced odds of VA ≥ 68 letters except MNV subtypes and SDD.

CONCLUSION

Presence of baseline fovea-involving atrophy, fibrosis, intraretinal fluid, SHRM, PCV EZ/ELM loss and increased CST determine poor presenting VA. This highlights the need for early detection and treatment prior to structural changes that worsen baseline VA.

Therapies for Geographic Atrophy

Diabetic retinopathy (DR) is a leading cause of blindness. Artificial intelligence (AI) has been proposed to provide a novel opportunity to increase screening for DR. While it is paramount to ensure AI has adequate technical capabilities to perform accurate screening, it is also important to assess how to best implement such technology into clinical practice. Human-centered design offers a methodology to understand the real-world context and behaviors of individuals, engage stakeholders, and rapidly prototype and test solutions, enhancing usability and avoiding unintended consequences. This review describes the methodology of human-centered design, examining how it has been used within a variety of health care contexts, with a particular focus on how it has been used to implement an AI-based DR screening program. Further research is needed to understand the best strategies to implement and evaluate AI in health care.

Automated detection of hyperreflective foci in the outer nuclear layer of the retina

PURPOSE

Hyperreflective foci are poorly understood transient elements seen on optical coherence tomography (OCT) of the retina in both healthy and diseased eyes. Systematic studies may benefit from the development of automated tools that can map and track such foci. The outer nuclear layer (ONL) of the retina is an attractive layer in which to study hyperreflective foci as it has no fixed hyperreflective elements in healthy eyes. In this study, we intended to evaluate whether automated image analysis can identify, quantify and visualize hyperreflective foci in the ONL of the retina.

METHODS

This longitudinal exploratory study investigated 14 eyes of seven patients including six patients with optic neuropathy and one with mild non-proliferative diabetic retinopathy. In total, 2596 OCT B-scan were obtained. An image analysis blob detector algorithm was used to detect candidate foci, and a convolutional neural network (CNN) trained on a manually labelled subset of data was then used to select those candidate foci in the ONL that fitted the characteristics of the reference foci best.

RESULTS

In the manually labelled data set, the blob detector found 2548 candidate foci, correctly detecting 350 (89%) out of 391 manually labelled reference foci. The accuracy of CNN classifier was assessed by manually splitting the 2548 candidate foci into a training and validation set. On the validation set, the classifier obtained an accuracy of 96.3%, a sensitivity of 88.4% and a specificity of 97.5% (AUC 0.989).

CONCLUSION

This study demonstrated that automated image analysis and machine learning methods can be used to successfully identify, quantify and visualize hyperreflective foci in the ONL of the retina on OCT scans.

Characteristics of intermediate age-related macular degeneration with hyperreflective foci

Hyperreflective foci (HRF) are the findings observed in optical coherence tomography (OCT) in several retinal diseases and are believed to be associated with the increased risk of atrophy in eyes with age-related macular degeneration (AMD). In this study, we investigated the clinical and genetic characteristics of intermediate AMD with HRF. We reviewed the medical charts for 155 patients with intermediate AMD, in whom macular neovascularization (MNV) was observed in the contralateral eye. The presence or absence of an HRF was evaluated using a spectral-domain OCT volume scan spanning the macular region. Patients were followed longitudinally for at least 12 months, and the maximum follow-up period was 60 months. Genotyping of ARMS2 A69S and CFH I62V was performed in all participants. Of the 155 patients (mean age: 77.8 ± 7.6 years, male/female: 103/52), HRF was observed in 53 eyes (34.2%) and was significantly associated with type-3 MNV (p = 1.0 × 10^-5) in the contralateral eye, pseudodrusen (p = 5.0 × 10^-4), thinner subfoveal choroidal thickness (p = 0.013), and risk of ARMS2 A69S (p = 0.023). During follow-up (40.8 ± 17.5), 38 eyes (24.5%) developed advanced AMD. The mean time to the onset of advanced AMD was 29.8 ± 12.9 months in eyes with intermediate AMD. HRF was associated with MNV (p = 1.0 × 10^-3), but not with atrophy.

Association between inflammatory cytokines in the aqueous humor and hyperreflective foci on optical coherence tomography in patients with neovascular age-related macular degeneration and polypoidal choroidal vasculopathy

Purpose

To investigate the associations between cytokine levels in the aqueous humor (AH) and hyperreflective foci (HF) on spectral-domain optical coherence tomography (SD-OCT) in neovascular age-related macular degeneration (nAMD) and polypoidal choroidal vasculopathy (PCV).

Methods

The prospective study included 63 eyes with nAMD, 44 with PCV, and 43 with cataracts (Controls). AH samples were obtained before anti-vascular endothelial growth factor (VEGF) therapy or cataract surgery. Cytokines interleukin 6 (IL-6), IL-8, IL-10, interferon-inducible protein 10 (IP-10), monocyte chemotactic protein 1 (MCP-1), and VEGF were measured by multiplex bead assay. Best-corrected visual acuity (BCVA), central macular thickness (CMT), and the number of HF were evaluated at baseline and 1 month after anti-VEGF treatment.

Results

No significances difference in IL-6 and IL-8 levels were noted among the three groups (P = 0.370 and P = 0.067). VEGF, IP-10, and IL-10 levels were significantly higher in nAMD and PCV groups than in Controls (all P &lt; 0.05). In nAMD, HF was positively correlated with VEGF (rs = 0.300, P = 0.025) and in eyes with HF group, VEGF and IL-10 were significantly higher than those without HF (P = 0.008 and P = 0.022). In PCV, no correlation was observed between HF and cytokines (all P &gt; 0.05). After anti-VEGF treatment, patients with HF in nAMD and PCV were predisposed to worse visual outcomes (P = 0.022 and P = 0.015) and a significantly greater reduction in CMT (P = 0.001 and P = 0.057). And nAMD patients with HF were more sensitive to anti-VEGF treatment than those without HF (P = 0.029).

Conclusions

In the nAMD group, HF was positively correlated with VEGF. Patients in nAMD with HF had elevated levels of VEGF and IL-10 and responded favorably to anti-VEGF. HF might serve as an inflammatory biomarker and a predictive factor for therapeutic efficacy in patients with nAMD.

Incidence and risk of advanced age-related macular degeneration in eyes with drusenoid pigment epithelial detachment

To investigate the incidence and risk of advanced age-related macular degeneration (AMD), including geographic atrophy (GA) and macular neovascularization (MNV), in eyes with drusenoid pigment epithelial detachment (PED). Eighty-five eyes with drusenoid PED from 85 patients (77.2 ± 7.0 years, male/female: 44/41) were included in this study. Patients were followed up every 1–3 months via spectral-domain optical coherence tomography (SD-OCT) and color fundus photography. If exudation was observed on SD-OCT, fluorescein and indocyanine green angiography were performed to confirm the MNV subtype accordingly. The maximum follow-up period was 60 months. During the study period, GA developed in 8 eyes while MNV also developed in 8 eyes. The Kaplan–Meier estimator revealed that the cumulative incidence for 60 months was 17.9% and 12.2% for GA and MNV, respectively. In eyes developing MNV, retinal angiomatous proliferation was the most common. Cox regression analysis revealed that baseline PED width was the only factor associated with advanced AMD. (p = 0.0026, Cox regression analysis). The 5-year cumulative incidence of advanced AMD, including GA and MNV, was approximately 30% in eyes with drusenoid PED among the Japanese elderly. A larger baseline PED width was the only risk factor for advanced AMD.

Non-invasive testing for early detection of neovascular macular degeneration in unaffected second eyes of older adults: EDNA diagnostic accuracy study

BACKGROUND

Neovascular age-related macular degeneration is a leading cause of sight loss, and early detection and treatment is important. For patients with neovascular age-related macular degeneration in one eye, it is usual practice to monitor the unaffected eye. The test used to diagnose neovascular age-related macular degeneration, fundus fluorescein angiography, is an invasive test. Non-invasive tests are available, but their diagnostic accuracy is unclear.

OBJECTIVES

The primary objective was to determine the diagnostic monitoring performance of tests for neovascular age-related macular degeneration in the second eye of patients with unilateral neovascular age-related macular degeneration. The secondary objectives were the cost-effectiveness of tests and to identify predictive factors of developing neovascular age-related macular degeneration.

DESIGN

This was a multicentre, prospective, cohort, comparative diagnostic accuracy study in a monitoring setting for up to 3 years. A Cox regression risk prediction model and a Markov microsimulation model comparing cost-effectiveness of the index tests over 25 years were used.

SETTING

This took place in hospital eye services.

PARTICIPANTS

Participants were adults (aged 50-95 years) with newly diagnosed (within the previous 6 weeks) neovascular age-related macular degeneration in one eye and an unaffected second (study) eye who were attending for treatment injections in the first eye and who had a study eye baseline visual acuity of ≥ 68 Early Treatment Diabetic Retinopathy Study letters.

INTERVENTIONS

The index tests were Amsler chart (completed by participants), fundus clinical examination, optical coherence tomography, self-reported vision assessment (completed by participants) and visual acuity. The reference standard was fundus fluorescein angiography.

MAIN OUTCOME MEASURES

The main outcome measures were sensitivity and specificity; the performance of the risk predictor model; and costs and quality-adjusted life-years.

RESULTS

In total, 552 out of 578 patients who consented from 24 NHS hospitals (n = 16 ineligible; n = 10 withdrew consent) took part. The mean age of the patients was 77.4 years (standard deviation 7.7 years) and 57.2% were female. For the primary analysis, 464 patients underwent follow-up fundus fluorescein angiography and 120 developed neovascular age-related macular degeneration on fundus fluorescein angiography. The diagnostic accuracy [sensitivity (%) (95% confidence interval); specificity (%) (95% confidence interval)] was as follows: optical coherence tomography 91.7 (85.2 to 95.6); 87.8 (83.8 to 90.9)], fundus clinical examination [53.8 (44.8 to 62.5); 97.6 (95.3 to 98.9)], Amsler [33.7 (25.1 to 43.5); 81.4 (76.4 to 85.5)], visual acuity [30.0 (22.5 to 38.7); 66.3 (61.0 to 71.1)] and self-reported vision [4.2 (1.6 to 9.8); 97.0 (94.6 to 98.5)]. Optical coherence tomography had the highest sensitivity across all secondary analyses. The final prediction model for neovascular age-related macular degeneration in the non-affected eye included smoking status, family history of neovascular age-related macular degeneration, the presence of nodular drusen with or without reticular pseudodrusen, and the presence of pigmentary abnormalities [c-statistic 0.66 (95% confidence interval 0.62 to 0.71)]. Optical coherence tomography monitoring generated the greatest quality-adjusted life-years gained per patient (optical coherence tomography, 5.830; fundus clinical examination, 5.787; Amsler chart, 5.736, self-reported vision, 5.630; and visual acuity, 5.600) for the lowest health-care and social care costs (optical coherence tomography, £19,406; fundus clinical examination, £19,649; Amsler chart, £19,751; self-reported vision, £20,198; and visual acuity, £20,444) over the lifetime of the simulated cohort. Optical coherence tomography dominated the other tests or had an incremental cost-effectiveness ratio below the accepted cost-effectiveness thresholds (£20,000) across the scenarios explored.

LIMITATIONS

The diagnostic performance may be different in an unselected population without any history of neovascular age-related macular degeneration; the prediction model did not include genetic profile data, which might have improved the discriminatory performance.

CONCLUSIONS

Optical coherence tomography was the most accurate in diagnosing conversion to neovascular age-related macular degeneration in the fellow eye of patients with unilateral neovascular age-related macular degeneration. Economic modelling suggests that optical coherence tomography monitoring is cost-effective and leads to earlier diagnosis of and treatment for neovascular age-related macular degeneration in the second eye of patients being treated for neovascular age-related macular degeneration in their first eye.

FUTURE WORK

Future works should investigate the role of home monitoring, improved risk prediction models and impact on long-term visual outcomes.

STUDY REGISTRATION

This study was registered as ISRCTN48855678.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 26, No. 8. See the NIHR Journals Library website for further project information.

Significance of Hyperreflective Foci as an Optical Coherence Tomography Biomarker in Retinal Diseases: Characterization and Clinical Implications

Hyperreflective foci (HRF) is a term coined to depict hyperreflective dots or roundish lesions within retinal layers visualized through optical coherence tomography (OCT). Histopathological correlates of HRF are not univocal, spacing from migrating retinal pigment epithelium cells, lipid-laden macrophages, microglial cells, and extravasated proteinaceous or lipid material. Despite this, HRF can be considered OCT biomarkers for disease progression, treatment response, and prognosis in several retinal diseases, including diabetic macular edema, age-related macular degeneration (AMD), retinal vascular occlusions, and inherited retinal dystrophies. The structural features and topographic location of HRF guide the interpretation of their significance in different pathological conditions. The presence of HRF less than 30 μm with reflectivity comparable to the retinal nerve fiber layer in the absence of posterior shadowing in diabetic macular edema indicates an inflammatory phenotype with a better response to steroidal treatment. In AMD, HRF overlying drusen are associated with the development of macular neovascularization, while parafoveal drusen and HRF predispose to macular atrophy. Thus, HRF can be considered a key biomarker in several common retinal diseases. Their recognition and critical interpretation via multimodal imaging are vital to support clinical strategies and management.

Clinical Manifestations of Cuticular Drusen: Current Perspectives

Cuticular drusen are part of the spectrum of age-related macular degeneration (AMD) with particular clinical and multimodal imaging characteristics. This drusen subpopulation shares several high-risk single nucleotide polymorphisms with AMD. Despite this feature, they can manifest at a relatively young age, presenting with a female preponderance. Multimodal imaging is essential for characterizing such lesions, using a combination of color fundus photographs, optical coherence tomography (OCT), fluorescein angiography (FA), and fundus autofluorescence (FAF). The classic starry-sky pattern visible on FA and the typical central hypoautofluorescent lesion with hyperautofluorescent rim on FAF is considered the result of a central retinal pigment epithelium (RPE) erosion from these triangular elevations of the RPE-basal lamina. This finding may also be responsible for the typical choroidal hypertransmission appreciated through OCT. The clinical course of cuticular drusen may be relatively benign at early stages, with small drusen presenting at a young age. However, the presence of clinical phenotypes characterized by diffuse involvement and/or accompanying large drusen in patients older than 60 years may confer a significant risk for either macular neovascularization or geographic atrophy.

Automated Quantification of Pathological Fluids in Neovascular Age-Related Macular Degeneration, and Its Repeatability Using Deep Learning

Purpose

To develop a reliable algorithm for the automated identification, localization, and volume measurement of exudative manifestations in neovascular age-related macular degeneration (nAMD), including intraretinal (IRF), subretinal fluid (SRF), and pigment epithelium detachment (PED), using a deep-learning approach.

Methods

One hundred seven spectral domain optical coherence tomography (OCT) cube volumes were extracted from nAMD eyes. Manual annotation of IRF, SRF, and PED was performed. Ninety-two OCT volumes served as training and validation set, and 15 OCT volumes from different patients as test set. The performance of our fluid segmentation method was quantified by means of pixel-wise metrics and volume correlations and compared to other methods. Repeatability was tested on 42 other eyes with five OCT volume scans acquired on the same day.

Results

The fully automated algorithm achieved good performance for the detection of IRF, SRF, and PED. The area under the curve for detection, sensitivity, and specificity was 0.97, 0.95, and 0.99, respectively. The correlation coefficients for the fluid volumes were 0.99, 0.99, and 0.91, respectively. The Dice score was 0.73, 0.67, and 0.82, respectively. For the largest volume quartiles the Dice scores were >0.90. Including retinal layer segmentation contributed positively to the performance. The repeatability of volume prediction showed a standard deviations of 4.0 nL, 3.5 nL, and 20.0 nL for IRF, SRF, and PED, respectively.

Conclusions

The deep-learning algorithm can simultaneously acquire a high level of performance for the identification and volume measurements of IRF, SRF, and PED in nAMD, providing accurate and repeatable predictions. Including layer segmentation during training and squeeze-excite block in the network architecture were shown to boost the performance.

Translational Relevance

Potential applications include measurements of specific fluid compartments with high reproducibility, assistance in treatment decisions, and the diagnostic or scientific evaluation of relevant subgroups.

MORPHOLOGICAL AND FUNCTIONAL CHARACTERISTICS AT THE ONSET OF EXUDATIVE CONVERSION IN AGE-RELATED MACULAR DEGENERATION

PURPOSE

To characterize retinal morphology differences among different types of choroidal neovascularization and visual function changes at the onset of exudative age-related macular degeneration.

METHODS

In a post hoc analysis of a prospective clinical study, 1,097 fellow eyes from subjects with choroidal neovascularization in the study eye enrolled in the HARBOR trial were evaluated. The onset of exudation was diagnosed on monthly optical coherence tomography by two masked graders. At conversion as well as 1 month earlier, pigment epithelial detachment, intraretinal cystoid fluid, subretinal fluid, subretinal hyperreflective material, as well as ellipsoid zone and external limiting membrane loss were quantitatively analyzed. Hyperreflective foci, retinal pigment epithelial defects, haze and vitreoretinal interface status were evaluated qualitatively. Main outcome measures included visual acuity and rates of morphologic features at conversion and 1 month earlier.

RESULTS

New-onset exudation was detected in 92 eyes. One month before conversion, hyperreflective foci, pigment epithelial detachment, retinal pigment epithelial defects, and haze were present in the majority of eyes. At the onset of exudation, the volumes of intraretinal cystoid fluid, subretinal fluid, subretinal hyperreflective material and pigment epithelial detachment, and the areas of external limiting membrane and ellipsoid zone loss significantly increased. The mean vision loss was -2.2 letters. Pathognomonic patterns of the different choroidal neovascularization types were already apparent 1 month before conversion.

CONCLUSION

Characteristic choroidal neovascularization-associated morphological changes are preceding disease conversion, while vision loss at the onset of exudation is minimal. Individual lesion types are related to specific changes in optical coherence tomography morphology already before the time of conversion. Our findings may help to elucidate the pathophysiology of neovascular age-related macular degeneration and support the diagnosis of imminent disease conversion.

Association of imaging biomarkers and local activation of complement in aqueous humor of patients with early forms of age-related macular degeneration

Purpose

To investigate a possible correlation between established imaging biomarkers for age-related macular degeneration and local complement system activation, measured in aqueous humor (AH) of patients with early stages of age-related macular degeneration (AMD) and controls.

Methods

This analysis included prospectively acquired AH samples of 106 eyes (35 with early/intermediate AMD, 71 controls). The levels of complement protein 3 (C3), 4 (C4), 5 (C5); activation products of complement factor 3a (C3a) and Ba, C3b/iC3b; complement factors B, D, H, I (CFB, CFD, CFH, CFI); and total protein concentration were analyzed. Quantitative levels of complement factors were correlated to the presence of reticular pseudodrusen (RPD), the presence of hyperreflective foci (HRF), and total drusen volume (DV) graded on imaging by spectral-domain optical coherence tomography and using Spearman’s rank correlation test.

Results

DV correlated with C3b/iC3b (r = 0.285; P = 0.034), C3a (r = 0.200; P = 0.047), Ba (r = 0.262; P = 0.009), and C5 (r = 430; P = 0.005), and showed a tendency towards correlation with C3a (r = 0.198; P = 0.057). HRF correlated significantly with C5 (r = 0.388; P = 0.011) and RPD showed a tendency towards correlation with CFB (r = 0.196; P = 0.050).

Conclusion

In patients with early AMD, HRF and drusen parameters but not RPD show low to fair levels of correlation with local complement activation in patients’ AH. Better understanding of complement activation could provide some insights into the pathogenesis of AMD. Imaging biomarkers could be useful to identify suitable patients for future clinical trials with complement-modulating therapies.

Application of Artificial Intelligence in Targeting Retinal Diseases

Retinal diseases affect an increasing number of patients worldwide because of the aging population. Request for diagnostic imaging in ophthalmology is ramping up, while the number of specialists keeps shrinking. Cutting-edge technology embedding artificial intelligence (AI) algorithms are thus advocated to help ophthalmologists perform their clinical tasks as well as to provide a source for the advancement of novel biomarkers. In particular, optical coherence tomography (OCT) evaluation of the retina can be augmented by algorithms based on machine learning and deep learning to early detect, qualitatively localize and quantitatively measure epi/intra/subretinal abnormalities or pathological features of macular or neural diseases. In this paper, we discuss the use of AI to facilitate efficacy and accuracy of retinal imaging in those diseases increasingly treated by intravitreal vascular endothelial growth factor (VEGF) inhibitors (i.e. anti-VEGF drugs), also including integration and interpretation features in the process. We review recent advances by AI in diabetic retinopathy, age-related macular degeneration, and retinopathy of prematurity that envision a potentially key role of highly automated systems in screening, early diagnosis, grading and individualized therapy. We discuss benefits and critical aspects of automating the evaluation of disease activity, recurrences, the timing of retreatment and therapeutically potential novel targets in ophthalmology. The impact of massive employment of AI to optimize clinical assistance and encourage tailored therapies for distinct patterns of retinal diseases is also discussed.

Characterization of Drusen and Hyperreflective Foci as Biomarkers for Disease Progression in Age-Related Macular Degeneration Using Artificial Intelligence in Optical Coherence Tomography

Importance

The morphologic changes and their pathognomonic distribution in progressing age-related macular degeneration (AMD) are not well understood.

Objectives

To characterize the pathognomonic distribution and time course of morphologic patterns in AMD and to quantify changes distinctive for progression to macular neovascularization (MNV) and macular atrophy (MA).

Design, Setting, and Participants

This cohort study included optical coherence tomography (OCT) volumes from study participants with early or intermediate AMD in the fellow eye in the HARBOR (A Study of Ranibizumab Administered Monthly or on an As-needed Basis in Patients With Subfoveal Neovascular Age-Related Macular Degeneration) trial. Patients underwent imaging monthly for 2 years (July 1, 2009, to August 31, 2012) following a standardized protocol. Data analysis was performed from June 1, 2018, to January 21, 2020.

Main Outcomes and Measures

To obtain topographic correspondence between patients and over time, all scans were mapped into a joint reference frame. The time of progression to MNV and MA was established, and drusen volumes and hyperreflective foci (HRF) volumes were automatically segmented in 3 dimensions using validated artificial intelligence algorithms. Topographically resolved population means of these markers were constructed by averaging quantified drusen and HRF maps in the patient subgroups.

Results

Of 1097 patients enrolled in HARBOR, 518 (mean [SD] age, 78.1 [8.2] years; 309 [59.7%] female) had early or intermediate AMD in the fellow eye at baseline. During the 24-month follow-up period, 135 (26%) eyes developed MNV, 50 eyes (10%) developed MA, and 333 (64%) eyes did not progress to advanced AMD. Drusen and HRF had distinct topographic patterns. Mean drusen thickness at the fovea was 29.6 μm (95% CI, 20.2-39.0 μm) for eyes progressing to MNV, 17.2 μm (95% CI, 9.8-24.6 μm) for eyes progressing to MA, and 17.1 μm (95% CI, 12.5-21.7 μm) for eyes without disease progression. At 0.5-mm eccentricity, mean drusen thickness was 25.8 μm (95% CI, 19.1-32.5 μm) for eyes progressing to MNV, 21.7 μm (95% CI, 14.6-28.8 μm) for eyes progressing to MA, and 14.4 μm (95% CI, 11.2-17.6 μm) for eyes without disease progression. The mean HRF thickness at the foveal center was 0.072 μm (95% CI, 0-0.152 μm) for eyes progressing to MNV, 0.059 μm (95% CI, 0-0.126 μm) for eyes progressing to MA, and 0.044 μm (95% CI, 0.007-0.081) for eyes without disease progression. At 0.5-mm eccentricity, the largest mean HRF thickness was seen in eyes progressing to MA (0.227 μm; 95% CI, 0.104-0.349 μm) followed by eyes progressing to MNV (0.161 μm; 95% CI, 0.101-0.221 μm) and eyes without disease progression (0.085 μm; 95% CI, 0.058-0.112 μm).

Conclusions and Relevance

In this study, drusen and HRF represented imaging biomarkers of disease progression in AMD, demonstrating distinct topographic patterns over time that differed between eyes progressing to MNV, eyes progressing to MA, or eyes without disease progression. Automated localization and precise quantification of these factors may help to develop reliable methods of predicting future disease progression.

Predicting conversion to wet age-related macular degeneration using deep learning

Progression to exudative 'wet' age-related macular degeneration (exAMD) is a major cause of visual deterioration. In patients diagnosed with exAMD in one eye, we introduce an artificial intelligence (AI) system to predict progression to exAMD in the second eye. By combining models based on three-dimensional (3D) optical coherence tomography images and corresponding automatic tissue maps, our system predicts conversion to exAMD within a clinically actionable 6-month time window, achieving a per-volumetric-scan sensitivity of 80% at 55% specificity, and 34% sensitivity at 90% specificity. This level of performance corresponds to true positives in 78% and 41% of individual eyes, and false positives in 56% and 17% of individual eyes at the high sensitivity and high specificity points, respectively. Moreover, we show that automatic tissue segmentation can identify anatomical changes before conversion and high-risk subgroups. This AI system overcomes substantial interobserver variability in expert predictions, performing better than five out of six experts, and demonstrates the potential of using AI to predict disease progression.

Major Predictive Factors for Progression of Early to Late Age-Related Macular Degeneration

INTRODUCTION

We present a prediction model for progression from early/intermediate to advanced age-related macular degeneration (AMD) within 5.9 years.

OBJECTIVES

To evaluate the combined role of genetic, nongenetic, and phenotypic risk factors for conversion from early to late AMD over ≥5 years.

METHODS

Baseline phenotypic characteristics were evaluated based on color fundus photography, spectral-domain optical coherence tomography, and infrared images. Genotyping for 36 single-nucleotide polymorphisms as well as systemic lipid and complement measurements were performed. Multivariable backward logistic regression resulted in a final prediction model.

RESULTS AND CONCLUSIONS

During a mean of 5.9 years of follow-up, 22.4% (n = 52) of the patients (n = 232) showed progression to late AMD. The multivariable prediction model included age, CFH variant rs1061170, pigment abnormalities, drusenoid pigment epithelial detachment (DPED), and hyperreflective foci (HRF). The model showed an area under the curve of 0.969 (95% confidence interval 0.948-0.990) and adequate calibration (Hosmer-Lemeshow test, p = 0.797). In addition to advanced age and carrying a CFH variant, pigment abnormalities, DPED, and HRF are relevant imaging biomarkers for conversion to late AMD. In clinical routine, an intensified monitoring of patients with a high-risk phenotypic profile may be suitable for the early detection of conversion to late AMD.

Risk factors for progression of age-related macular degeneration

Purpose

Age‐related macular degeneration (AMD) is a degenerative disease of the macula, often leading to progressive vision loss. The rate of disease progression can vary among individuals and has been associated with multiple risk factors. In this review, we provide an overview of the current literature investigating phenotypic, demographic, environmental, genetic, and molecular risk factors, and propose the most consistently identified risk factors for disease progression in AMD based on these studies. Finally, we describe the potential use of these risk factors for personalised healthcare.

Recent findings

While phenotypic risk factors such as drusen and pigment abnormalities become more important to predict disease progression during the course of the disease, demographic, environmental, genetic and molecular risk factors are more valuable at earlier disease stages. Demographic and environmental risk factors such as age and smoking are consistently reported to be related to disease progression, while other factors such as sex, body mass index (BMI) and education are less often associated. Of all known AMD variants, variants that are most consistently reported with disease progression are rs10922109 and rs570618 in CFH, rs116503776 in C2/CFB/SKIV2L, rs3750846 in ARMS2/HTRA1 and rs2230199 in C3. However, it seems likely that other AMD variants also contribute to disease progression but to a lesser extent. Rare variants have probably a large effect on disease progression in highly affected families. Furthermore, current prediction models do not include molecular risk factors, while these factors can be measured accurately in the blood. Possible promising molecular risk factors are High‐Density Lipoprotein Cholesterol (HDL‐C), Docosahexaenoic acid (DHA), eicosapentaenoic acid (EPA), zeaxanthin and lutein.

Summary

Phenotypic, demographic, environmental, genetic and molecular risk factors can be combined in prediction models to predict disease progression, but the selection of the proper risk factors for personalised risk prediction will differ among individuals and is dependent on their current disease stage. Future prediction models should include a wider set of genetic variants to determine the genetic risk more accurately, and rare variants should be taken into account in highly affected families. In addition, adding molecular factors in prediction models may lead to preventive strategies and personalised advice.

Prediction of Individual Disease Conversion in Early AMD Using Artificial Intelligence

Purpose

While millions of individuals show early age-related macular degeneration (AMD) signs, yet have excellent vision, the risk of progression to advanced AMD with legal blindness is highly variable. We suggest means of artificial intelligence to individually predict AMD progression.

Methods

In eyes with intermediate AMD, progression to the neovascular type with choroidal neovascularization (CNV) or the dry type with geographic atrophy (GA) was diagnosed based on standardized monthly optical coherence tomography (OCT) images by independent graders. We obtained automated volumetric segmentation of outer neurosensory layers and retinal pigment epithelium, drusen, and hyperreflective foci by spectral domain-OCT image analysis. Using imaging, demographic, and genetic input features, we developed and validated a machine learning-based predictive model assessing the risk of conversion to advanced AMD.

Results

Of a total of 495 eyes, 159 eyes (32%) had converted to advanced AMD within 2 years, 114 eyes progressed to CNV, and 45 to GA. Our predictive model differentiated converting versus nonconverting eyes with a performance of 0.68 and 0.80 for CNV and GA, respectively. The most critical quantitative features for progression were outer retinal thickness, hyperreflective foci, and drusen area. The features for conversion showed pathognomonic patterns that were distinctly different for the neovascular and the atrophic pathways. Predictive hallmarks for CNV were mostly drusen-centric, while GA markers were associated with neurosensory retina and age.

Conclusions

Artificial intelligence with automated analysis of imaging biomarkers allows personalized prediction of AMD progression. Moreover, pathways of progression may be specific in respect to the neovascular/atrophic type.

Artificial intelligence in retina

Major advances in diagnostic technologies are offering unprecedented insight into the condition of the retina and beyond ocular disease. Digital images providing millions of morphological datasets can fast and non-invasively be analyzed in a comprehensive manner using artificial intelligence (AI). Methods based on machine learning (ML) and particularly deep learning (DL) are able to identify, localize and quantify pathological features in almost every macular and retinal disease. Convolutional neural networks thereby mimic the path of the human brain for object recognition through learning of pathological features from training sets, supervised ML, or even extrapolation from patterns recognized independently, unsupervised ML. The methods of AI-based retinal analyses are diverse and differ widely in their applicability, interpretability and reliability in different datasets and diseases. Fully automated AI-based systems have recently been approved for screening of diabetic retinopathy (DR). The overall potential of ML/DL includes screening, diagnostic grading as well as guidance of therapy with automated detection of disease activity, recurrences, quantification of therapeutic effects and identification of relevant targets for novel therapeutic approaches. Prediction and prognostic conclusions further expand the potential benefit of AI in retina which will enable personalized health care as well as large scale management and will empower the ophthalmologist to provide high quality diagnosis/therapy and successfully deal with the complexity of 21st century ophthalmology.