Mesopic and Dark-Adapted Two-Color Fundus-Controlled Perimetry in Choroidal Neovascularization Secondary to Age-Related Macular Degeneration

CORRELATION BETWEEN MICROPERIMETRY AND IMAGING IN EXTENSIVE MACULAR ATROPHY WITH PSEUDODRUSEN-LIKE APPEARANCE

PURPOSE

To determine the correlation between microperimetry and imaging findings in extensive macular atrophy with pseudodrusen-like appearance (EMAP).

METHODS

This cross-sectional, observational study included 44 consecutive patients with EMAP (88 eyes) and 30 healthy subjects (60 eyes). Both groups underwent visual acuity assessment, mesopic and scotopic microperimetry, fundus photography, autofluorescence, optical coherence tomography, and optical coherence tomography angiography. Retinal sensitivity was also subdivided in macular (0-4°) and paramacular areas (8-10°). Scotopic sensitivity loss was defined as the difference between scotopic and mesopic sensitivities for each tested point. Eyes with EMAP were further classified into the three stages described by Romano et al: 19 eyes in Stage 1, 31 in Stage 2, and 38 in Stage 3.

RESULTS

Mesopic and scotopic retinal sensitivity were significantly reduced in patients with EMAP compared with controls, particularly in the macular area (all P < 0.001). Mesopic retinal sensitivity progressively declined in more advanced EMAP stages (all P < 0.01), but no scotopic differences were observed between Stages 2 and 3 ( P = 0.08). Remarkably, scotopic sensitivity loss was significantly higher in Stage 1 ( P < 0.05).On multivariate analysis, mesopic dysfunction was associated with larger atrophic areas ( P < 0.01), foveal involvement ( P = 0.03), and fibrosis ( P = 0.02). Conversely, no independent variable was associated with a reduced scotopic retinal sensitivity (all P > 0.05).

CONCLUSION

The findings highlight that patients with EMAP suffer from a severe cone- and rod-mediated dysfunction on microperimetry. The predominant rod impairment in the early cases (Stage 1) emphasizes the importance of dark-adapted scotopic microperimetry as a clinical end point and suggests defective transportation across the RPE-Bruch membrane complex in its pathogenesis.

Establishing Fully-Automated Fundus-Controlled Dark Adaptometry: A Validation and Retest-Reliability Study

Purpose

The purpose of this study was to establish and validate a novel fundus-controlled dark-adaptometry method.

Methods

We developed a custom dark-adaptometry software for the S-MAIA device using the open-perimetry-interface. In the validation-substudy, participants underwent dark-adaptometry testing with a comparator device (MonCvONE, 59% rhodopsin bleach, cyan and red stimuli centered at 2 degrees, 4 degrees, and 6 degrees eccentricity). Following a brief break (approximately 5 minutes), the participants were bleached again and underwent dark-adaptometry testing with the S-MAIA device (same loci). In the retest reliability-substudy, participants were tested twice with the S-MAIA device (same loci as above). Nonlinear curve fitting was applied to extract dark-adaptation curve parameters. Validity and repeatability were summarized in terms of the mean bias and 95% limits of agreement (LoAs).

Results

In the validation-substudy (N = 20 participants, median age interquartile range [IQR] 31.5 years [IQR = 25.8, 62.0]), measures of rod-mediated dark-adaptation showed little to no between method differences for the cone-rod-break-time (bias 95% confidence interval [95% CI] of +0.1 minutes [95% CI = -0.6 to 0.8]), rod-intercept-time (-0.23 minutes [95% CI = -1.38 to 0.93]), and S2 slope (-0.01 LogUnits/minutes [95% CI = -0.02 to -0.01]). In the retest reliability-substudy (N = 10 participants, 32.0 years [95% CI = 27.0, 57.5]), the corresponding LoAs were (cone-rod-break-time) -3.94 to 2.78 minutes, (rod-intercept-time) -4.55 to 3.11 minutes, and (S2 slope [rate-limited component of rod recovery]) -0.03 to 0.03 LogUnits/minutes. The LoAs for the steady-state cone and rod thresholds were -0.28 to 0.33 LogUnits and -0.34 to 0.28 LogUnits.

Conclusions

The devised fundus-controlled dark-adaptometry method yields valid and reliable results.

Translational Relevance

Fundus-controlled dark-adaptometry solves the critical need for localized testing of the visual cycle and retinoid transfer in eyes with unstable fixation.

Retest variability and patient reliability indices of quantitative fundus autofluorescence in age-related macular degeneration: a MACUSTAR study report

This study aimed to determine the retest variability of quantitative fundus autofluorescence (QAF) in patients with and without age-related macular degeneration (AMD) and evaluate the predictive value of patient reliability indices on retest reliability. A total of 132 eyes from 68 patients were examined, including healthy individuals and those with various stages of AMD. Duplicate QAF imaging was conducted at baseline and 2 weeks later across six study sites. Intraclass correlation (ICC) analysis was used to evaluate the consistency of imaging, and mean opinion scores (MOS) of image quality were generated by two researchers. The contribution of MOS and other factors to retest variation was assessed using mixed-effect linear models. Additionally, a Random Forest Regressor was trained to evaluate the extent to which manual image grading of image quality could be replaced by automated assessment (inferred MOS). The results showed that ICC values were high for all QAF images, with slightly lower values in AMD-affected eyes. The average inter-day ICC was found to be 0.77 for QAF segments within the QAF8 ring and 0.74 for peripheral segments. Image quality was predicted with a mean absolute error of 0.27 on a 5-point scale, and of all evaluated reliability indices, MOS/inferred MOS proved most important. The findings suggest that QAF allows for reliable testing of autofluorescence levels at the posterior pole in patients with AMD in a multicenter, multioperator setting. Patient reliability indices could serve as eligibility criteria for clinical trials, helping identify patients with adequate retest reliability.

Assessment of Scotopic Function in Rod-Cone Inherited Retinal Degeneration With the Scotopic Macular Integrity Assessment

Purpose

The scotopic macular integrity assessment (S-MAIA) can perform scotopic assessment to detect localized changes to scotopic rod and cone function. This study is an exploratory investigation of the feasibility of using the S-MAIA in a rod-cone dystrophy population to identify the pattern of loss in scotopic photoreceptor function.

Methods

Twenty patients diagnosed with a rod-cone dystrophy underwent visual acuity testing, full-field stimulus threshold assessment, and multiple S-MAIA tests after dark adaptation periods of 20 minutes and 45 minutes performed separately. Only right eyes were tested. Three tests were performed following a learning test. A Bland-Altman analysis was used to assess repeatability and agreement between tests after the two time periods. Spatial interpolation maps were created from the group plots to display the pattern of rod and cone loss.

Results

Learning effects took place between testing sessions 1 and 2 but not 2 and 3. Limits of agreement were larger in the patient eyes than control eyes, but within previously reported values. Using longer adaptation time of 45 minutes did not offer a significant advantage over 20 minutes. Patterns for the cyan and red sensitivities were different, indicating different patterns of loss for rods and cones.

Conclusions

A dark adaptation time of 20 minutes before testing is sufficient for thresholding. The S-MAIA is suitable for use in patients with a logarithm of the minimum angle of resolution vision of at least 0.7 and provides a viable outcome measure for patients with rod-cone dystrophies and preserved central vision. The spatial information about scotopic function from the S-MAIA provides information about disease processes and progression.

Translational Relevance

There is a need for scotopic measures for use in clinical trials. Scotopic microperimetry works well in patients with early disease, allowing the extension of recruitment criteria for novel therapies of rod-cone dystrophies.

Evaluation of Cone- and Rod-Mediated Parameters in Dark Adaptation Testing as Outcome Measures in Age-Related Macular Degeneration

PURPOSE

To investigate the suitability of 6 rod- or cone-mediated dark adaptation (DA) parameters as outcome measures for clinical trials in age-related macular degeneration (AMD), including their retest reliability, association with age and disease severity, and measurable longitudinal change over time.

DESIGN

Prospective, longitudinal study (Clinicaltrials.gov: NCT01352975).

PARTICIPANTS

A total of 191 patients with AMD and older participants followed longitudinally over 5 years.

METHODS

Dark adaptation testing was performed using the AdaptDx dark adaptometer with a maximum test time of 40 minutes. A 2-part exponential-linear curve was fitted to obtain values for cone decay, cone plateau, time to rod-cone break, rod intercept time (RIT), rod adaptation rate (S2), and area under the curve. Intersession retest reliability was assessed in tests performed within 2 weeks using the Bland-Altman analysis. The relationship of DA parameters with age, AMD severity, and reticular pseudodrusen (RPD) presence was evaluated using linear mixed models.

MAIN OUTCOME MEASURES

Retest reliability, association with disease severity, and longitudinal change of 6 DA parameters.

RESULTS

A total of 1329 DA curves were analyzed. Rod intercept time was the parameter that showed the greatest reliability (intraclass correlation coefficient of 0.88) and greatest association with age, AMD severity, and RPD (marginal R² of 0.38), followed by the rod-mediated parameters area under the curve and rod-cone break. Cone plateau appeared constant at lower RIT values but increased with progressive rod dysfunction (RIT > 22.8 minutes) with a slope of 0.07 log units per 10 minutes RIT prolongation. Therefore, it might provide additional information in the advanced stages of AMD.

CONCLUSIONS

Age-related macular degeneration severity and RPD presence are each associated with large differences in multiple DA curve parameters. In addition, substantial differences in some parameters occur with age, even accounting for AMD severity and RPD status. This supports the 2-hit hypothesis of age and disease status on DA (and perhaps AMD pathophysiology itself). Of the DA parameters, RIT has the highest retest reliability, closest correlation with AMD severity and RPD, and largest longitudinal changes. This underscores the suitability of RIT as an outcome measure in clinical trials. The cone plateau increases only in advanced stages of kinetic rod dysfunction, indicating rod dysfunction preceding cone dysfunction and degeneration in the temporal sequence of pathology in AMD.

Visual Dysfunction and Structural Correlates in Sorsby Fundus Dystrophy

PURPOSE

To elucidate morphological determinants of rod and cone dysfunction in Sorsby fundus dystrophy (SFD), and to systematically compare visual function tests for interventional trials.

DESIGN

Prospective cross-sectional study.

METHODS

Patients with SFD (n = 16) and controls (n = 20) underwent visual function testing (best-corrected visual acuity [BCVA] and low luminance visual acuity [LLVA], contrast sensitivity, mesopic and dark-adapted (DA) fundus-controlled perimetry [FCP], rod-mediated dark adaptation [RMDA]), and multimodal imaging. Vision-related quality of life was evaluated. FCP and RMDA thresholds were analyzed using mixed models and structure-function correlation using machine learning (ML). Longitudinal data of 1 patient with high-dose vitamin A supplementation were available.

RESULTS

Although photopic BCVA was normative in SFD, LLVA was impaired (0.30 LogMAR [0.20; 0.45] vs 0.20 LogMAR [0.03; 0.28], P < .05). Scotopic visual function exhibited a delayed rod-intercept time (21 minutes [12.15; 21] vs 4.05 minutes [3.22; 5.36], P < .001), and marked DA cyan mean sensitivity loss (-11.80 dB [-3.47; -19.85]), paralleled by a reduced vision-related quality of life. ML-based structure-function correlation allowed prediction of mesopic, DA cyan, and red sensitivity with high accuracy (cross-validated mean absolute error: 4.36, 7.77, and 5.31 dB, respectively), whereas RMDA could be slowed even in the absence of fundus alterations on multimodal imaging. After high-dose vitamin A supplementation, RMDA and DA thresholds improved markedly.

CONCLUSIONS

Patients with SFD exhibit severely impaired scotopic visual function even in the absence of funduscopic alterations on multimodal imaging. In contrast to BCVA, scotopic visual function tests are suitable to quantify dysfunction in the early stages. Improvement of scotopic dysfunction after (off-label) high-dose vitamin A intake, as observed in one patient in our study, is compatible with the hypothesized local deficiency of vitamin A secondary to Bruch's membrane alterations.

Scotopic microperimetry: evolution, applications and future directions

For many inherited and acquired retinal diseases, reduced night vision is a primary symptom. Despite this, the clinical testing options for spatially resolved scotopic vision have until recently been limited. Scotopic microperimetry is a relatively new visual function test that combines two-colour perimetry with fundus-controlled perimetry performed in scotopic luminance conditions. The technique enables spatially resolved mapping of central retinal sensitivity alongside the ability to distinguish between rod and cone photoreceptor sensitivities. Two companies produce commercially available scotopic microperimeters - Nidek (Nidek Technologies Srl, Padova, Italy) and CenterVue (CenterVue S.p.A., Padova, Italy). Scotopic microperimetry is a promising technology capable of detecting changes in retinal sensitivity before changes in other measures of visual function. Scotopic microperimetry is a promising functional biomarker that has the potential as a useful clinical trial outcome measure. This review summarises the evolution and applications of scotopic microperimetry, and discusses testing options, including testing grid selection, dark-adaptation time and threshold sensitivity analyses.

Role of microperimetry in evaluating disease progression in age-related macular degeneration: a scoping review

Purpose

Recent research has found variable evidence on the role of mesopic and dark-adapted scotopic microperimetry assessment in age-related macular degeneration. This scoping review summarises how mesopic and scotopic microperimetry can be used to assess disease progression in age-related macular degeneration and identifies gaps in the literature.

Methods

A population, concept, and context approach was used to develop the search strategy. Ovid MEDLINE, EMBASE, Cochrane Library, PubMed, CINAHL Plus, Web of Science, and SCOPUS databases were used to conduct the literature search. The key search terms used in the databases were age-related macular degeneration and microperimetry.

Results

Twelve studies were eligible and included in the review. All the studies (n = 12) were conducted in European countries [Germany (9), Italy (2), and the United Kingdom (1)]. The mesopic and scotopic sensitivities were measured using the Nidek scotopic microperimeter (MP1-S) (n = 6), scotopic Macular Integrity Assessment device (S-MAIA) (n = 5), and both MP1-s and S MAIA (n = 1). 83.3% (n = 10) studied (cross-sectional design) on mesopic, scotopic microperimetry and found reduced rod (scotopic) photoreceptors sensitivities compared to cone (mesopic) photoreceptors sensitivities in patients with small and reticular pseudodrusen despite having good visual acuity. Only 16.7% (n = 2) of studies followed participants with reticular drusen/large drusen for three years (longitudinal design) and found reduced scotopic over mesopic sensitivity at baseline and localized mesopic with profound scotopic sensitivity loss during follow-ups.

Conclusion

Scotopic sensitivity is a better functional indicator than mesopic sensitivity to understand early and intermediate age-related macular degeneration progression. The evidence from longitudinal studies is debatable due to the limited stimuli range of existing microperimeters, smaller sample size, and lost follow-ups.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10792-021-02170-9.

Autofluorescent Organelles Within the Retinal Pigment Epithelium in Human Donor Eyes With and Without Age-Related Macular Degeneration

Purpose

Human retinal pigment epithelium (RPE) cells contain lipofuscin, melanolipofuscin, and melanosome organelles that impact clinical autofluorescence (AF) imaging. Here, we quantified the effect of age-related macular degeneration (AMD) on granule count and histologic AF of RPE cell bodies.

Methods

Seven AMD-affected human RPE-Bruch's membrane flatmounts (early and intermediate = 3, late dry = 1, and neovascular = 3) were imaged at fovea, perifovea, and near periphery using structured illumination and confocal AF microscopy (excitation 488 nm) and compared to RPE-flatmounts with unremarkable macula (n = 7, >80 years). Subsequently, granules were marked with computer assistance, and classified by their AF properties. The AF/cell was calculated from confocal images. The total number of granules and AF/cell was analyzed implementing a mixed effect analysis of covariance (ANCOVA).

Results

A total of 152 AMD-affected RPE cells were analyzed (fovea = 22, perifovea = 60, and near-periphery = 70). AMD-affected RPE cells showed increased variability in size and a significantly increased granule load independent of the retinal location (fovea: P = 0.02, perifovea: P = 0.04, and near periphery: P < 0.01). The lipofuscin fraction of total organelles decreased and the melanolipofuscin fraction increased in AMD, at all locations (especially the fovea). AF was significantly lower in AMD-affected cells (fovea: <0.01, perifovea: <0.01, and near periphery: 0.02).

Conclusions

In AMD RPE, lipofuscin was proportionately lowest in the fovea, a location also known to be affected by accumulation of soft drusen and preservation of cone-mediated visual acuity. Enlarged RPE cell bodies displayed increased net granule count but diminished total AF. Future studies should also assess the impact on AF imaging of RPE apical processes containing melanosomes.

Detecting Diabetic Retinal Neuropathy Using Fundus Perimetry

Fundus perimetry is a new technique for evaluating the light sense in the retina in a point-to-point manner. Light sense is fundamentally different from visual acuity, which measures the threshold for discriminating and perceiving two points or lines, called the minimum cognoscible. The quality of measurement of retinal sensitivity has dramatically increased in the last decade, and the use of fundus perimetry is now gaining popularity. The latest model of fundus perimetry, MP-3, can be used for a wide range of measurements and has an advanced eye tracking system. High background illumination enables accurate measurement of mesopic retail sensitivity. Recent investigations have shown that neuronal damage precedes vascular abnormalities in diabetic retinopathy. The loss of retinal function has also been reported prior to morphological changes in the retina. In this review, the importance of measuring retinal sensitivity to evaluate visual function in the early stages of diabetic retinopathy was discussed. The usefulness of retinal sensitivity as an outcome measure in clinical trials for treatment modalities is also presented. The importance of fundus perimetry is promising and should be considered by both diabetes researchers and clinical ophthalmologists.

Age-related macular degeneration

Age-related macular degeneration (AMD) is the leading cause of legal blindness in the industrialized world. AMD is characterized by accumulation of extracellular deposits, namely drusen, along with progressive degeneration of photoreceptors and adjacent tissues. AMD is a multifactorial disease encompassing a complex interplay between ageing, environmental risk factors and genetic susceptibility. Chronic inflammation, lipid deposition, oxidative stress and impaired extracellular matrix maintenance are strongly implicated in AMD pathogenesis. However, the exact interactions of pathophysiological events that culminate in drusen formation and the associated degeneration processes remain to be elucidated. Despite tremendous advances in clinical care and in unravelling pathophysiological mechanisms, the unmet medical need related to AMD remains substantial. Although there have been major breakthroughs in the treatment of exudative AMD, no efficacious treatment is yet available to prevent progressive irreversible photoreceptor degeneration, which leads to central vision loss. Compelling progress in high-resolution retinal imaging has enabled refined phenotyping of AMD in vivo. These insights, in combination with clinicopathological and genetic correlations, have underscored the heterogeneity of AMD. Hence, our current understanding promotes the view that AMD represents a disease spectrum comprising distinct phenotypes with different mechanisms of pathogenesis. Hence, tailoring therapeutics to specific phenotypes and stages may, in the future, be the key to preventing irreversible vision loss.

AI-based structure-function correlation in age-related macular degeneration

Sensitive and robust outcome measures of retinal function are pivotal for clinical trials in age-related macular degeneration (AMD). A recent development is the implementation of artificial intelligence (AI) to infer results of psychophysical examinations based on findings derived from multimodal imaging. We conducted a review of the current literature referenced in PubMed and Web of Science among others with the keywords ‘artificial intelligence’ and ‘machine learning’ in combination with ‘perimetry’, ‘best-corrected visual acuity (BCVA)’, ‘retinal function’ and ‘age-related macular degeneration’. So far AI-based structure-function correlations have been applied to infer conventional visual field, fundus-controlled perimetry, and electroretinography data, as well as BCVA, and patient-reported outcome measures (PROM). In neovascular AMD, inference of BCVA (hereafter termed inferred BCVA) can estimate BCVA results with a root mean squared error of ~7–11 letters, which is comparable to the accuracy of actual visual acuity assessment. Further, AI-based structure-function correlation can successfully infer fundus-controlled perimetry (FCP) results both for mesopic as well as dark-adapted (DA) cyan and red testing (hereafter termed inferred sensitivity). Accuracy of inferred sensitivity can be augmented by adding short FCP examinations and reach mean absolute errors (MAE) of ~3–5 dB for mesopic, DA cyan and DA red testing. Inferred BCVA, and inferred retinal sensitivity, based on multimodal imaging, may be considered as a quasi-functional surrogate endpoint for future interventional clinical trials in the future.

Mesopic and Scotopic Light Sensitivity and Its Microstructural Correlates in Pseudoxanthoma Elasticum

Importance

Correlates for Bruch membrane alterations are needed for interventional trials targeting the Bruch membrane in pseudoxanthoma elasticum (PXE).

Objectives

To quantify mesopic and scotopic light sensitivity and identify its microstructural correlates associated with a diseased Bruch membrane in patients with PXE.

Design, Setting, and Participants

A prospective, single-center, cross-sectional case-control study was conducted at a tertiary referral center from January 31, 2018, to February 20, 2020. Twenty-two eyes of 22 patients with PXE and 40 eyes of 40 healthy individuals were included. Data analysis was completed March 15, 2020.

Exposures

Mesopic and dark-adapted 2-color fundus-controlled perimetry (microperimetry) and multimodal retinal imaging including spectral-domain optical coherence tomography (SD-OCT) and OCT angiography were performed. Perimetry thresholds were analyzed using mixed models, and structure-function correlation with SD-OCT data was performed using machine learning.

Main Outcomes and Measures

Observed dark-adapted cyan sensitivity loss as measure of rod photoreceptor dysfunction, as well as mean absolute error between predicted and observed retinal sensitivity to assess the accuracy of structure-function correlation.

Results

Of the 22 patients with PXE included in this study, 15 were women (68%); median age was 56.5 years (interquartile range, 50.4-61.2). These patients exhibited mesopic (estimate, 5.13 dB; 95% CI, 2.89-7.38 dB), dark-adapted cyan (estimate, 9.08 dB; 95% CI, 6.34-11.82 dB), and dark-adapted red (estimate, 7.05 dB; 95% CI, 4.83-9.27 dB) sensitivity losses. This sensitivity loss was also evident in 9 eyes with nonneovascular PXE (mesopic: estimate, 3.21 dB; 95% CI, 1.28-5.14 dB; dark-adapted cyan: 5.93 dB; 95% CI, 3.59-8.27 dB; and dark-adapted red testing: 4.84 dB; 95% CI, 2.88-6.80 dB), showing a distinct centrifugal pattern of sensitivity loss with preserved function toward the periphery. Retinal function could be predicted from microstructure with high accuracy (mean absolute errors, of 4.91 dB for mesopic, 5.44 dB for dark-adapted cyan, and 4.99 dB for dark-adapted red). The machine learning-based analysis highlighted an association of a thinned inner retina and putative separation of the pigment-epithelium-photoreceptor complex with sensitivity loss.

Conclusions and Relevance

In this study, among 22 patients with PXE, those with and without choroidal neovascularization exhibited reductions of retinal sensitivity being most pronounced in dark-adapted cyan testing. This finding suggests that pathologic characteristics of this Bruch membrane disease may be dominated by rod photoreceptor degeneration and/or dysfunction. A putative pigment-epithelium-photoreceptor separation may further impair rod function, while inner retinal abnormalities appear to be correlated with overall dysfunction.

Clinical Perspectives and Trends: Microperimetry as a trial endpoint in retinal disease

Endpoint development trials are underway across the spectrum of retinal disease. New validated endpoints are urgently required for the assessment of emerging gene therapies and in preparation for the arrival of novel therapeutics targeting early stages of common sight-threatening conditions such as age-related macular degeneration. Visual function measures are likely to be key candidates in this search. Over the last two decades, microperimetry has been used extensively to characterize functional vision in a wide range of retinal conditions, detecting subtle defects in retinal sensitivity that precede visual acuity loss and tracking disease progression over relatively short periods. Given these appealing features, microperimetry has already been adopted as an endpoint in interventional studies, including multicenter trials, on a modest scale. A review of its use to date shows a concurrent lack of consensus in test strategy and a wealth of innovative disease and treatment-specific metrics which may show promise as clinical trial endpoints. There are practical issues to consider, but these have not held back its popularity and it remains a widely used psychophysical test in research. Endpoint development trials will undoubtedly be key in understanding the validity of microperimetry as a clinical trial endpoint, but existing signs are promising.

[Artificial intelligence in ophthalmology : Guidelines for physicians for the critical evaluation of studies]

BACKGROUND

Empirical models have been an integral part of everyday clinical practice in ophthalmology since the introduction of the Sanders-Retzlaff-Kraff (SRK) formula. Recent developments in the field of statistical learning (artificial intelligence, AI) now enable an empirical approach to a wide range of ophthalmological questions with an unprecedented precision.

OBJECTIVE

Which criteria must be considered for the evaluation of AI-related studies in ophthalmology?

MATERIAL AND METHODS

Exemplary prediction of visual acuity (continuous outcome) and classification of healthy and diseased eyes (discrete outcome) using retrospectively compiled optical coherence tomography data (50 eyes of 50 patients, 50 healthy eyes of 50 subjects). The data were analyzed with nested cross-validation (for learning algorithm selection and hyperparameter optimization).

RESULTS

Based on nested cross-validation for training, visual acuity could be predicted in the separate test data-set with a mean absolute error (MAE, 95% confidence interval, CI of 0.142 LogMAR [0.077; 0.207]). Healthy versus diseased eyes could be classified in the test data-set with an agreement of 0.92 (Cohen's kappa). The exemplary incorrect learning algorithm and variable selection resulted in an MAE for visual acuity prediction of 0.229 LogMAR [0.150; 0.309] for the test data-set. The drastic overfitting became obvious on comparison of the MAE with the null model MAE (0.235 LogMAR [0.148; 0.322]).

CONCLUSION

Selection of an unsuitable measure of the goodness-of-fit, inadequate validation, or withholding of a null or reference model can obscure the actual goodness-of-fit of AI models. The illustrated pitfalls can help clinicians to identify such shortcomings.

Fundus-controlled perimetry (microperimetry): Application as outcome measure in clinical trials

Fundus-controlled perimetry (FCP, also called 'microperimetry') allows for spatially-resolved mapping of visual sensitivity and measurement of fixation stability, both in clinical practice as well as research. The accurate spatial characterization of visual function enabled by FCP can provide insightful information about disease severity and progression not reflected by best-corrected visual acuity in a large range of disorders. This is especially important for monitoring of retinal diseases that initially spare the central retina in earlier disease stages. Improved intra- and inter-session retest-variability through fundus-tracking and precise point-wise follow-up examinations even in patients with unstable fixation represent key advantages of these technique. The design of disease-specific test patterns and protocols reduces the burden of extensive and time-consuming FCP testing, permitting a more meaningful and focused application. Recent developments also allow for photoreceptor-specific testing through implementation of dark-adapted chromatic and photopic testing. A detailed understanding of the variety of available devices and test settings is a key prerequisite for the design and optimization of FCP protocols in future natural history studies and clinical trials. Accordingly, this review describes the theoretical and technical background of FCP, its prior application in clinical and research settings, data that qualify the application of FCP as an outcome measure in clinical trials as well as ongoing and future developments.

Prediction of Function in ABCA4-Related Retinopathy Using Ensemble Machine Learning

Full-field electroretinogram (ERG) and best corrected visual acuity (BCVA) measures have been shown to have prognostic value for recessive Stargardt disease (also called “ABCA4-related retinopathy”). These functional tests may serve as a performance-outcome-measure (PerfO) in emerging interventional clinical trials, but utility is limited by variability and patient burden. To address these limitations, an ensemble machine-learning-based approach was evaluated to differentiate patients from controls, and predict disease categories depending on ERG (‘inferred ERG’) and visual impairment (‘inferred visual impairment’) as well as BCVA values (‘inferred BCVA’) based on microstructural imaging (utilizing spectral-domain optical coherence tomography) and patient data. The accuracy for ‘inferred ERG’ and ‘inferred visual impairment’ was up to 99.53 ± 1.02%. Prediction of BCVA values (‘inferred BCVA’) achieved a precision of ±0.3LogMAR in up to 85.31% of eyes. Analysis of the permutation importance revealed that foveal status was the most important feature for BCVA prediction, while the thickness of outer nuclear layer and photoreceptor inner and outer segments as well as age of onset highly ranked for all predictions. ‘Inferred ERG’, ‘inferred visual impairment’, and ‘inferred BCVA’, herein, represent accurate estimates of differential functional effects of retinal microstructure, and offer quasi-functional parameters with the potential for a refined patient assessment, and investigation of potential future treatment effects or disease progression.

Presymptomatic Retinal Sensitivity Changes in Intermediate Age-Related Macular Degeneration Associated With New Retinal Fluid

Purpose

To determine whether change in retinal sensitivity in areas with subretinal or intraretinal fluid secondary to age-related macular degeneration (AMD) precedes visual symptoms. If confirmed, retinal sensitivity testing could be used for home monitoring in AMD.

Methods

Individuals with intermediate AMD enrolled in a longitudinal study were seen every 6 months and underwent best-corrected visual acuity testing (BCVA), spectral domain–optical coherence tomography (SD-OCT), and microperimetry. Asymptomatic individuals who developed incidental, reading center–determined retinal fluid detected on SD-OCT were identified. The point-wise sensitivity (PWS) at the time of fluid detection was compared with 6 and 12 months prior.

Results

Fourteen of 161 individuals developed fluid without symptoms. PWS over fluid areas at detection was reduced compared with 6 (difference −2.04 dB, P < 0.001) and 12 months (−2.27 dB, P < 0.001) prior. PWS over fluid areas was reduced compared with perifluid areas (difference −1.02 dB, P = 0.03), peripheral areas (−1.51 dB, P < 0.001), nonprogressed fellow eyes (−1.49 dB, P = 0.006), and nonprogressed age-matched intermediate AMD eyes (−2.29 dB, P = 0.001). No difference in BCVA was observed in eyes developing fluid compared to eyes that do not develop fluid (P = 0.76).

Conclusions

Retinal areas with fluid on SD-OCT had a corresponding reduction in retinal sensitivity at the time of fluid detection compared with 6 and 12 months prior, in asymptomatic intermediate AMD without change in BCVA.

Translational Relevance

Development of self-monitoring tools to detect changes in retinal sensitivity may be helpful for early detection of retinal fluid suggestive of progression to neovascular AMD before acuity is affected.

Assessment of Exudative Activity of Choroidal Neovascularization in Age-Related Macular Degeneration by OCT Angiography

PURPOSE

Based on exudative activity, choroidal neovascularization (CNV) in age-related macular degeneration (AMD) can be classified as "active" aCNV, pretherapied "silent" sCNV (i.e., a treatment-free interval >12 weeks), or treatment-naïve "quiescent" qCNV. We evaluated the qualitative and quantitative optical coherence tomography angiography (OCTA) features of these CNV subgroups.

METHODS

The presence of small-caliber vessels, peripheral arcades, and a -perilesional OCTA signal attenuation as well as values for vessel length, density, and branching index were evaluated for each CNV network in a 6 × 6 mm OCTA scan pattern.

RESULTS

Fifty-one eyes of 51 patients with AMD (age 75.9 ± 7.5 years; 20 males [39.2%]) were included. The qCNV subgroup (n = 8) showed the highest prevalence of qualitative and quantitative values for OCTA activity criteria, reaching significance with regard to small-caliber vessels (p = 0.003), peripheral arcades (p = 0.039), vessel length (p = 0.020), and branching index (p < 0.001) when compared to the aCNV (n = 32) and sCNV (n = 11) subgroups. Qualitative criteria were inversely associated with the number of previous anti-VEGF injections (each p < 0.03), while quantitative metrics also suggested lower values.

CONCLUSIONS

These findings suggest that OCTA may be supportive in the phenotypical differentiation of CNV lesions secondary to AMD, while the assessed structural changes appeared to be more indicative of previously administered anti-VEGF therapy than current exudative activity.

Artificial intelligence for morphology-based function prediction in neovascular age-related macular degeneration

Spatially-resolved mapping of rod- and cone-function may facilitate monitoring of macular diseases and serve as a functional outcome parameter. However, mesopic and dark-adapted two-color fundus-controlled perimetry (FCP, also called "microperimetry") constitute laborious examinations. We have devised a machine-learning-based approach to predict mesopic and dark-adapted (DA) retinal sensitivity in eyes with neovascular age-related macular degeneration (nAMD). Extensive psychophysical testing and volumetric multimodal retinal imaging data were acquired including mesopic, DA red and DA cyan FCP, spectral-domain optical coherence tomography and confocal scanning laser ophthalmoscopy infrared reflectance and fundus autofluorescence imaging. With patient-wise leave-one-out cross-validation, we have been able to achieve prediction accuracies of (mean absolute error, MAE [95% CI]) 3.94 dB [3.38, 4.5] for mesopic, 4.93 dB [4.59, 5.27] for DA cyan and 4.02 dB [3.63, 4.42] for DA red testing. Partial addition of patient-specific sensitivity data decreased the cross-validated MAE to 2.8 dB [2.51, 3.09], 3.71 dB [3.46, 3.96], and 2.85 dB [2.62, 3.08]. The most important predictive feature was outer nuclear layer thickness. This artificial intelligence-based analysis strategy, termed "inferred sensitivity", herein, enables to estimate differential effects of retinal structural abnormalities on cone- and rod-function in nAMD, and may be used as quasi-functional surrogate endpoint in future clinical trials.