Choriocapillaris Impairment, Visual Function, and Distance to Fovea in Aging and Age-Related Macular Degeneration: ALSTAR2 Baseline

Abstract

Purpose: In aging and early-intermediate age-related macular degeneration (AMD), rod-mediated dark adaptation (RMDA) slows more at 5° superior than at 12°. Using optical coherence tomography angiography (OCTA), we asked whether choriocapillaris flow deficits are related to distance from the fovea.

Methods: Persons ≥60 years stratified for AMD via the Age-Related Eye Disease Study's nine-step system underwent RMDA testing. Two adjacent 4.4° × 4.4° choriocapillaris OCTA slabs were centered on the fovea and 12° superior. Flow signal deficits (FD%) in concentric arcs (outer radii in mm, 0.5, 1.5, 2.2, 4.0, and 5.0 superior) were correlated with rod intercept time (RIT) and best-corrected visual acuity (BCVA).

Results: In 366 eyes (170 normal, 111 early AMD, 85 intermediate AMD), FD% was significantly worse with greater AMD severity in all regions (overall P < 0.05) and poorest under the fovea (P < 0.0001). In pairwise comparisons, FD% worsened with greater AMD severity (P < 0.05) at distances <2.2 mm. At greater distances, eyes with intermediate, but not early AMD differed from normal eyes. Foveal FD% was more strongly associated with longer RIT at 5° (r = 0.52) than RIT at 12° (r = 0.39) and BCVA (r = 0.21; all P < 0.0001). Choroidal thickness was weakly associated with longer RIT at 5° and 12° (r = 0.10–0.20, P < 0.05) and not associated with AMD severity.

Conclusions: Reduced transport across the choriocapillaris–Bruch's membrane–retinal pigment epithelium complex, which contributes to drusen formation under the macula lutea (and fovea), may also reduce retinoid resupply to rods encircling the high-risk area. FD% has potential as a functionally validated imaging biomarker for AMD emergence.

Age-related macular degeneration (AMD) degrades central vision in older adults worldwide.¹ In 2023 two drugs were approved in the United States for treating geographic atrophy,^2,3 the end-stage of non-neovascular AMD. These agents inhibit key proteins in the complement cascade, the largest pathway implicated by genetic associations.^4,5 To enable treatments or preventions before the expansion of pre-existing atrophy, the Alabama Study on Early Age-related Macular Degeneration 2 (ALSTAR2) prospectively seeks functional and structural biomarkers.⁶ This report focuses on rod-mediated dark adaptation (RMDA) and assessment of choriocapillaris flow signal via optical coherence tomography angiography (OCTA), two technologies well suited for AMD's beginnings. 

RMDA, the first functional risk factor for AMD onset,⁷ is a delayed return of light sensitivity after exposure to bright light. RMDA tested at 5° superior to the fovea probes where histologic rod densities decline in aging and AMD.^8,9 RMDA comprehensively assesses steps of retinoid resupply at multiple tissue layers between the circulation and rod outer segments, including the choriocapillaris endothelium, a microvasculature supporting outer retina.^10–14 RMDA delay, a dynamic measure, precedes reductions in steady-state scotopic sensitivity^10,15 and slows with each decade of healthy aging.¹⁶ Among functional measures assessed in the ALSTAR2 baseline cohort, RMDA was the only visual functional measure that separated normal aging from early AMD.¹⁵ 

OCTA imaging allows dye-free and noninvasive visualization of blood flow in vessels. Choriocapillaris rarefaction is a histologic hallmark of aging¹⁷ and AMD,^18–20 with the central 3-mm diameter most affected.²¹ Indeed, among cellular-level changes in AMD-affected tissue layers, choriocapillary attenuation in central retina quantitatively parallels the loss of rods in aging.¹⁴ OCTA generates depth-resolved flow signal through motion contrast derived from reflectivity changes in sequential B-scans.^22,23 Improved scan speed, signal compensation, projection artifact removal, and analytic tools make quantification achievable.²² In early- and intermediate-AMD eyes, OCTA shows reduced choriocapillaris flow signal compared to age-matched normal eyes,^24,25 worsening with stages of non-neovascular AMD,²⁶ association with extracellular deposits,^27–30 and association with poorer steady-state scotopic sensitivity.³¹ In ALSTAR2 baseline (410 eyes), flow signal deficit (FD%) in a 4.4 mm-wide square fovea-centered region was found to be higher in early AMD eyes than normal, and higher still in intermediate AMD.³² Of visual functions tested, delayed RMDA was the most highly correlated with choriocapillaris impairment seen by OCTA.³² 

Cross-sectional studies using swept-source OCTA in healthy eyes^33,34 showed that choriocapillaris FD% increases during adulthood, especially under the fovea. About 90% of photoreceptor oxygen needs is supplied by the choriocapillaris,³⁵ and foveal cones have very high ATP demands.³⁶ As reviewed,¹⁴ longitudinal data from community-dwelling and clinic populations,^37–39 plus histology,⁴⁰ reveals that drusen-related progression risk for advanced AMD is concentrated within the 3-mm-diameter macula lutea, especially the central 1 mm. Perhaps the life-long burden of serving foveal cells exhausts the underlying microvascular endothelium, by undefined mechanisms,^41–43 contributing to deposit formation under the fovea and poor vision by rods encircling the high-risk area. Herein we ask whether choriocapillaris FD% and its associations with visual functions are related to distance from the fovea. In a coordinate system centered on the point of highest cone density, distance from the fovea (eccentricity) has much greater dynamic range than angle around the fovea, due to steep gradients of photoreceptor density.¹⁴ 

Methods

Study Population

This cross-sectional study analyzed baseline data from the Alabama Study on Early Age-Related Macular Degeneration 2 (ALSTAR2; ClinicalTrials.gov Identifier NCT04112667). Participants ≥60 years old were recruited through the comprehensive care clinics at the Callahan Eye Hospital at the University of Alabama at Birmingham between October 2019 through September 2021 and evaluated at the Clinical Research Unit of the Department of Ophthalmology and Visual Sciences. Institutional Review Board approval was granted by the University of Alabama at Birmingham. Research adhered to the principles of the Declaration of Helsinki. All study participants provided written informed consent. 

The clinic electronic health record system was searched for patients with early or intermediate AMD using ICD-10 codes (H35.30*; H35.31*; H35.36*). Records were screened for eligibility by author CO. Exclusion criteria were (1) any eye condition or disease in either eye (other than early cataract) that can impair vision including diabetic retinopathy, glaucoma, ocular hypertension, history of retinal diseases (e.g., retinal vein occlusion, retinal degeneration), optic neuritis, corneal disease, previous ocular trauma or surgery, or refractive error ≥ 6 diopters; (2) neurological conditions that can impair vision or judgment including multiple sclerosis, Parkinson's disease, stroke, Alzheimer's disease, seizure disorders, brain tumor, or traumatic brain injury; (3) psychiatric disorders that could impair the ability to follow directions, answer questions about health and functioning, or to provide informed consent; (4) diabetes; or (5) any medical condition that causes significant frailty or was thought to be terminal. Persons in normal macular health from the same clinic met the same eligibility criteria except for the ICD-10 codes indicative of AMD. 

AMD Classification

In each participant, the eye with better acuity was tested for imaging and visual function. If both eyes had the same acuity, then the study eye was selected randomly. Classification into diagnostic groups was based on evaluation of three-field, color fundus photographs taken with a digital camera (450+; Carl Zeiss Meditec, Dublin, CA, USA) after dilation with 1% tropicamide and 2.5% phenylephrine hydrochloride. Presence and severity of AMD was assessed using the Age-Related Eye Disease Study (AREDS) 9-step classification system⁴⁴ by a trained grader (author MEC) masked to other participant characteristics. Groups were defined as follows: eyes with normal macular health were classified as AREDS grade 1, early AMD as grades 2 to 4, and intermediate AMD as grades 5 to 8. The presence or absence of subretinal drusenoid deposit (SDD), which is associated with poor RMDA,^45,46 was determined by multimodal imaging, as described.³² Birthdate, gender, race/ ethnicity, and smoking status were obtained through a self-administered questionnaire. 

Visual Function Testing

Vision tests were administered under photopic (cone-mediated), mesopic (rod- and cone-mediated), and scotopic conditions (rod-mediated), as described^15,47 Herein we focus on RMDA at 5° and 12° (rod intercept time, or RIT)¹⁵ and best-corrected photopic letter acuity (logarithm₁₀ of the minimum angle of resolution, BCVA logMAR).^48,49 BCVA was assessed before pupillary dilation using the Electronic Visual Acuity tester (JAEB Center, Tampa, FL, USA).⁴⁸ RMDA was assessed after dilation using the AdaptDx (LumiThera, Poulsbo, WA, USA). Test targets were located at 5° on the superior vertical meridian of the retina, because rod loss is proportionally maximal in aging and AMD at 5°^8,9 as compared to more distant (eccentric) locations.¹⁴ We use a flash-based equivalent bleach paradigm established by Pugh,⁵⁰ to be distinguished from a long exposure (e.g., 30–60 seconds) to a steady bright light.⁵¹ The procedure began with a photo-bleach exposure to a 6° flash centered at the test target location (50 ms duration, 58,000 scotopic cd/m² s intensity)⁵⁰ while the participant focused on the fixation light at a distance of 30 cm. Threshold measurement for a 2° diameter, 500 nm circular target began 15 seconds after bleach offset. Log thresholds were previous expressed as sensitivity in decibel units as a function of time after bleach offset. Threshold measurement continued until RIT was reached. The RIT is the duration in minutes required for sensitivity to recover to a criterion value of 5.0 × 10⁻³ scotopic cd/m²,^7,52 located in the latter one-half of the second component of rod-mediated recovery.^12,53 If the RIT was not reached, threshold measurement procedure stopped at 45 minutes. Participants with fixation errors of more than 30% were excluded from analysis. 

OCTA

We expanded on the methods used in our previous report to include an additional scanned area.³² From each study eye two consecutive 15° × 15° (∼4.4 × 4.4 mm) OCTA volumes centered on the fovea and at 12° superior (four total for most participants), were captured using the OCT Angiography Module of the Spectralis HRA+OCT (Heidelberg Engineering, Germany, HEYEX software version 6.10.6.0).⁵⁴ Each OCTA volume comprised 384 B-scans 12 µm apart, and 3.9 µm axial and 5.7 µm transverse resolution. The two volumes were positioned using the point of fixation as specified by the instrument. Projection-resolved choriocapillaris slabs were segmented using the manufacturer-recommended level of 10–30 µm below Bruch's membrane.⁵⁵ From the two volumes captured at each location, the slabs with fewer artifacts were included for analysis. Absence of OCTA artifacts like shadowing and scan quality ≤25 dB signal-to-noise ratio^56,57 were assessed by a masked, trained grader (author DK). 

Scans were batch processed using custom software (MATLAB 2022b, The MathWorks, Inc., Natick, MA; ImageJ, National Institutes of Health, Bethesda, MD, USA). To compensate for signal attenuation under large drusen and hyperreflective foci, each OCTA image was multiplied by the inverse of the structural OCT image extracted from identically segmented OCTA volumes.^58,59 Compensated images were binarized using the Phansalkar local thresholding method using a 20 µm radius.⁶⁰ Choriocapillaris areas directly beneath major retinal vessels were excluded.⁶⁰ Within each 15° × 15° scan area, we quantified choriocapillaris flow signal deficits (FD%) (i.e., the percentage of area lacking detectable flow signal seen as dark pixel areas in a binary mask).⁶¹ We did not enforce a minimum size cutoff for FD regions, as our previous histologic analysis indicated that >75% of intercapillary spacings are <24 µm wide,³² a commonly used limit. 

Our previous OCTA study of this cohort³² addressed several aspects of quality control and repeatability of the FD% metric. We determined that FD% test-retest repeatability in older normal eyes was excellent (interclass correlation coefficient = 0.96). We determined that compensation methods developed for the longer wavelength and greater depth penetration of swept-source OCT could be applied to spectral domain OCTA. In the same study, we demonstrated that reduced flow signal was not confined to areas under drusen and that the mean difference in FD% when including drusen vs excluding drusen was negligible. Drusen, especially if calcified, may reduce FD% detectability,^62,63 a pertinent concern for AMD eyes at more severe stages of intermediate AMD than those included in ALSTAR2. 

Regional Analysis of FD%

To compare choriocapillaris integrity to visual tests, we expressed FD% within defined retinal regions illustrated in Figure 1. Two adjacent 4.4-mm-wide OCTA slabs centered at the fovea and at 12° were montaged to create a vertically oriented rectangular region that extended into the superior near-periphery⁴⁴ and included the two RMDA test target locations. FD% was quantified in five concentric regions that completely tiled this rectangle (Fig. 1), with the following outer radii: 0.5 mm (1 mm circle), 1.5 mm annulus, 2.2 mm annulus, 4.0 mm arc, and 5.0 mm arc. The 1 mm circle and 1.5 mm annulus are equivalent to the ETDRS central subfield and inner ring, respectively.¹⁴ For correspondence to vision, we used a mean conversion of 0.288 mm/°, with the 1 mm circle centered on 0° and outer radii at 5.2°, 7.7°, 14.0°, and 17.5°, respectively, for the annuli and arcs. 

Figure 1.

 

Retinal landmarks and regions of interest used for choriocapillaris flow signal deficit quantification. Two en face choriocapillaris OCTA scans centered on 0° and 12° eccentricities overlaid on confocal fundus photograph. Arrowhead denotes the superior edge of the 0° centered OCTA scan (see Methods). Gray circles indicate the 5° and 12° dark adaptation test locations used in the study.

Measurement of Choroidal Thickness

To assess the impact of macrovascular integrity in the same eyes assessed for OCTA, we measured choroidal thickness at 25 locations that corresponded with vision assessments (0°, 5°, and 12° in all directions from fovea). A custom plug-in in FIJI was written for manual quantification of thickness (www.fiji.sc; available at: https://sites.imagej.net/CreativeComputation/) Using the “Segmented Line” tool, an observer precisely delineated the scleral-choroidal border by placing points along the interface. To evaluate intra- and inter-rater agreement, choroidal thickness was measurement in a group of 20 cases by authors MA and DK. Following this confirmation, comprehensive choroidal thickness measurements were conducted across the broader cohort of 366 cases by author MA. 

Statistical Analyses

Demographic and visual function characteristics of participants were summarized using means and standard deviations or number (percent) for continuous and categorical data, respectively. FD% was summarized with means and standard deviations. Analysis of covariance (ANCOVA) with pairwise comparisons was used to compare continuous data by AMD severity and SDD status adjusting for age. As ANCOVA requires a continuous dependent variable and a categorical independent variable, AMD severity was the independent variable, and FD%, visual function measures, and choroidal thickness measures, and SDD status were the dependent variables with age. Age is the covariable in these models, which adjusts for the confounding effects of age on FD% and visual functions when comparing by AMD severity. Age-adjusted associations for FD% and visual function parameters were computed with Spearman correlations (r). In addition to age, correlations were adjusted for the presence of SDD. The level of significance was P ≤ 0.05 (two-sided). All analyses were performed in SAS version 9.4. 

Results

A total of 366 eyes from 366 participants (mean [SD] age, 71.3 [5.5] years) were included in the study. Table 1 shows the demographic characteristics of these participants. The sample included 170 older normal (46.4%), 111 early AMD (30.3%) and 85 intermediate AMD (23.2%) based on the AREDS 9-step classification. 

Table 1.

 

Demographics of 366 Study Eyes of 366 Persons

 

Table 2 shows mean and standard deviation of FD% in the five retinal regions described in the Methods, stratified by the three diagnostic groups. Overall mean FD% values ranged from a low of 46.4% (region 4, normal) to a high of 63.5% (region 1, intermediate AMD). Eyes with intermediate AMD had worse FD% compared to normal and early AMD eyes in all regions measured (P < 0.01). Eyes with early AMD had worse FD% than normal eyes (P < 0.05) at eccentricities <2.2 mm. At eccentricities ≥2.2 mm, represented by regions 4 and 5, eyes with intermediate, but not early AMD, differed significantly from older normal eyes (P = 0.0022, 0.0033 vs. 0.4025, 0.9029, respectively). Thus, FD% was able to discriminate normal eyes from early AMD in regions closer to the foveal center than 2.2 mm and not at further distances. 

Table 2.

 

Choriocapillaris Flow Signal Deficits by AMD Presence and Severity

 

The higher vulnerability of choriocapillaris in intermediate AMD is underscored by graphical representations of FD% in Figures 2 and 3. Dot-and-whisker plots in Figure 2 show that FD% in intermediate AMD is proportionately higher relative to early AMD in and near the fovea (regions 1–3). Color-coded heatmaps in Figure 3 show a smooth and somewhat asymmetric area of maximal FD% covering the central subfield (fovea) and extending just beyond. Figure 3 also shows the RMDA 5° test target location at the rim of the area most affected by choriocapillaris FD%. RMDA 12° is well outside the most affected area. 

Figure 2.

 

Regional choriocapillaris flow signal deficits by AMD severity. Dot-and-whisker plot demonstrates choriocapillaris FD% (mean and standard deviation) in eccentricity regions 1–5 of Figure 1. Region 1 includes the fovea. AMD severity groups are assigned based on AREDS 9-step grading system. Asterisks indicate the level of significance. n.s.; non-significant (P > 0.05).

Figure 3.

 

Heatmaps of choriocapillaris demonstrate regional flow signal deficits (FD%) by AMD severity. (A–C) Heatmaps demonstrate greater choriocapillaris FD% in foveal and perifoveal regions, and these worsen with disease severity. Gray circles indicate the 5° and 12° RMDA test locations. (B) In early AMD, worse FD% is roughly confined to the parafoveal regions, near the 5° dark adaptation test target location. (C) In intermediate AMD, there is significant worsening of FD% extending to the perifoveal regions. White reference outlines show the regions of interest defined in Figure 1. AMD severity groups are assigned based on AREDS nine-step grading system.

Table 3 documents RIT at 5° and 12° and BCVA measured at the fovea in the 366 eyes assessed for OCTA in this study. The relationships of vision to AMD presence and severity are like those previously reported.³² In brief, all three visual functions worsened with AMD presence and severity, with RMDA 5° exhibiting the largest change between mean values (2.35-fold difference) for normal and intermediate AMD (vs. 1.85-fold for RMDA 12° and 1.12-fold for BCVA). 

Table 3.

 

Visual Functions in 366 Eyes of 366 Persons Assessed for Choriocapillaris Integrity Using OCT Angiography

 

Table 4 shows correlations between choriocapillaris FD%, RIT at 5°, RIT at 12°, and logMAR BCVA measured at the fovea. Regional FD% was correlated with delayed RMDA at both 5° and 12° locations, and more strongly at 5° than at 12°. Spearman correlations (r) for FD% vs RIT 5° ranged from 0.14–0.52 (all regions P < 0.01). FD% versus RIT 12° ranged from 0.15–0.39 (all regions P < 0.05). FD% vs logMAR BCVA showed the lowest correlations at 0.03–0.21 (P < 0.05 in regions at <2.2 mm eccentricity only). 

Table 4.

 

Age-Adjusted Regional Associations of Flow Deficit % With Cone- and Rod-Mediated Vision

 

Figure 4 graphically displays structure-function correlations between regional FD%, represented by color saturation, and measures of vision. RIT at 5° was more strongly associated with worse FD% in all regions, as compared to RIT at 12° and BCVA. Structure-function correlations became weaker as eccentricity increased for all function tests. FD% under the fovea was most strongly associated with each visual function. 

Figure 4.

 

Structure-function correlations between regional choriocapillaris flow signal deficits (FD%) and rod- versus cone-mediated visual function. Age-adjusted Spearman correlations between regional choriocapillaris FD% with RIT at 5° and 12°, and BCVA in 366 eyes of 366 participants. In all regions, delayed RMDA at 5° was more strongly associated with worse FD%, than RMDA at 12° and BCVA. Correlations decreased with increasing eccentricity across all visual function tests. Asterisks indicate the level of significance. n.s., nonsignificant (P > 0.05). NSTI graphic shows retinal directions (Nasal, Superior, Temporal, Inferior). Gray spots indicate test target locations for RMDA at 5° and 12° (lower and upper, respectively).

Table 5 shows choroidal thickness at 0°, 5°, 12° locations, along with associations between RMDA at 5° and 12°. Compared to FD%, choroidal thicknesses were associated with delayed RMDA at 5° and 12° (r = 0.10–0.20, P < 0.05), and these correlations were weaker than observed for choriocapillaris FD% (Table 4). Choroidal thickness was not associated with AMD severity (all P > 0.10). Intra-rater and inter-rater agreement for thickness measurements were good (intraclass correlation coefficient >0.7 at 17/25 and 20/25 measurement locations, respectively) 

Table 5.

 

Choroidal Thickness and Associations With RMDA by Diagnostic Groups in 362 Eyes

 

Discussion

Main Results

We previously reported for ALSTAR2 baseline that choriocapillaris FD% is worse in early AMD than in normal eyes and worse still in intermediate AMD. Here we show that these differences are related to distance from the fovea. In both normal and early AMD eyes, we observed that FD% is similar within 2.2 mm of the foveal center. In intermediate AMD, FD% is markedly worse than early AMD under the fovea, and this difference gradually diminishes with increasing distance from the fovea (Fig. 2). Notably, the correlation of RMDA 5° with FD% changes smoothly with eccentricity, at all AMD stages (Fig. 4). In contrast RMDA 12° shows a similar but weaker relationship despite rods being more abundant at 12° than at 5° (rod/cone ratio 11–13 vs. 3–4). As summarized,¹⁴ impairment of RMDA near the fovea is robust, having been seen in multiple studies, study populations, and technologies, for two decades.^51,64–66 

Relation to OCTA Studies of Choriocapillaris

Our findings using spectral domain OCTA to explore choriocapillaris FD% can be compared to data at the aging-AMD interface gathered with swept-source OCTA and analyzed with various approaches. For this comparison, we converted anatomic terminology of other publications to fovea-centered circles and annuli, with the annuli specified by outer radius in mm. In 75 eyes of 75 persons (23–80 years) Nassisi et al measured FD% in a series of area-matched fovea-centered rings and found higher FD% within the central 200 µm diameter.³³ Zheng et al studied 164 eyes of 164 persons (19 to 88 years) and quantified FD% in 1-mm circles, 1.5-mm annuli, and a 2.5-mm circles centered on the fovea, finding an age-related increase in FD%, largely driven by the central 1-mm.³⁴ Sacconi et al.⁶⁷ checked 72 eyes of 72 persons (20–80 years) in the 1 mm central subfield, 1.5 mm annuli, and 3 mm annuli and found slightly poorer flow signal density with age under the fovea and less of a difference between fovea and perifovea overall. Braun et al.²⁶ analyzed 56 eyes from 41 persons with early, intermediate, and advanced AMD in 1 mm circle, 3 mm annuli, and 5 mm annuli. They did not see an expected poorer value in the central 1 mm circle, which they attributed to a strong age effect degrading choriocapillaris integrity prior to AMD onset.^6,26 

It should be mentioned that these studies are all cross-sectional, and none including ours produce a range of FD% that coincide exactly with the most defensible estimates of en face choriocapillaris vascular density from histology.^17,68 This reflects both technical limitations of OCTA, assumptions of various analytic approaches, and the few published histologic measurements directly comparable to FD%. As summarized,⁶⁹ many histologic studies lack details about tissue provenance and eye donor characteristics. Furthermore, choriocapillaris endothelium quantification requires optimization of preparation and microscopy techniques.^{20,21,68,70–72} More data are needed. 

Relation to Studies of Choroid

Our data indicate that a measure of macrovascular health (choroidal thickness) is associated with vision but less so than a measure of microvascular health (FD%). Furthermore in this cohort, choroidal thickness did not differ with AMD presence or severity. In some studies, the choroid is thicker in early and intermediate AMD,^73–75 whereas others demonstrated significant choroidal thinning.^76–78 Some reports have failed to show that choroidal thickness significantly changes in AMD.^79,80 Inconsistent results from published studies suggest that choroidal thickness may not show a specific relationship with AMD stage and may instead be influenced by factors such as smoking, age, and associated pachychoroid spectrum disorders.⁸¹ 

Rod Vision Near the Fovea

To explain why rod vision is slowed when choriocapillaris under the macula lutea is impaired, we invoke a Center-Surround model of cone resilience and rod vulnerability (Fig. 5) that 3-dimensionally integrates foveal neuroscience and drusen biology.^6,14 We interpret parafoveal rod dysfunction as occurring on the edge of a high-risk area created by the lifelong sustenance of foveal cones. This model assumes that evolutionary adaptations for detailed vision that are advantageous in reproductive years may become disadvantageous in late adulthood as the support system for foveal cones degrades.⁸² 

Figure 5.

 

Why rod vision is influenced by FD% near the fovea. Cone resilience and rod vulnerability in aging and AMD can be modeled as difference of 2-dimensional Gaussian surfaces, a mathematical relationship familiar to vision science.^97,98 In the top row is an en face view of help, epitomized by macular xanthophyll pigment (orange) and harm via soft drusen/basal linear deposit and sequelae. In the bottom row help and harm are plotted on one vertical axis, positive and negative directions, respectively. (A) The distribution of xanthophyll carotenoids, as shown in Figure 2, is a focused center of help in the macula lutea. (B) The distribution of soft druse material and sequela is shown as a broad circular area of harm. (C) Together, help and harm make a narrow center of foveal cone resilience on top of a broad surround of parafoveal and perifoveal rod vulnerability. Reprinted from¹⁴ under Creative Commons No Derivatives License.

Spatially concentric, opposing mechanisms can be assembled mathematically to form a narrow central region of positive-going influence atop a broader and shallower surround of a negative-going influence. Figure 5 shows a focused positive-going center of cone sustenance, represented by Müller glia, a rich reservoir of xanthophyll carotenoids (lutein and zeaxanthin) and other essential support services to cones (Fig. 5A). A broad negative-going valley of harm affecting cones in the center and rods around the edge is represented by drusen and their precursors in BrM (Fig. 5B). This valley is broad, because xanthophylls extend beyond the foveal center into the plexiform layers.⁸³ In this model, xanthophyll delivery to the retinal pigment epithelium (RPE) by plasma lipoproteins for transfer to the fovea and subsequent repackaging of the carriers as local lipoproteins disposed into BrM is one reason dictated by evolution as to why AMD occurs under the macula lutea. 

Rod Vision and the Choriocapillaris

ALSTAR2 study design and the analysis herein also incorporates the idea that delayed RMDA throughout adulthood and early stages of AMD involves multiple tissue layers between the circulation and the photoreceptor outer segments.¹⁴ These layers include choriocapillaris endothelium, BrM, a potential space between the RPE basal lamina and inner collagenous layer of BrM (sub-RPE-BL space) where drusen may form, and membrane specializations for bidirectional transport at the basolateral and apical processes of RPE. We previously emphasized the age-related accumulation of lipoprotein-derived lipids in aging BrM of normal eyes.^84–86 However, age- and disease-related changes in other layers also contribute and are in fact related. For example, in 476 eyes of ALSTAR2 baseline cohort, a fovea-centered area of interdigitation zone (IZ) non-discernibility increases with AMD severity and before changes in the EZ band.⁸⁷ The IZ is of interest because it represents where RPE apical processes directly contact photoreceptor outer segments⁸⁸ for delivery of retinoids for phototransduction, among other essentials.⁸⁹ From the perspective of drusen formation, the largest population-level risk factor for AMD progression and a visible marker of aging, a straightforward and well-supported hypothesis is that the RPE is relatively functional, constitutively releasing lipoproteins as part of physiologic lipid trafficking, which are impeded from transfer across choriocapillaris and BrM to circulation. Insofar as choriocapillary endothelium is the metabolic gatekeeper to the RPE,⁹⁰ degradation of this interface will impact vision too. 

Strengths, Limitations, Future Directions, and Conclusions

Strengths of this study include a very large sample with well-defined AMD status, highly repeatable measures of RMDA,⁹¹ OCTA studies excluding common OCTA artifacts and incorporating new histologic information about choriocapillaris, and a clearly articulated hypothesis linking a cogent vision test to underlying cell and tissue mechanisms in human retina. Limitations are the small regions analyzed due to constraints of the imaging device and the geometry of the foveal singularity, lack of scaling for differences in axial length, and a sample consisting of many individuals of European continental ancestry. Another limitation is that this report did not analyze in detail the relationship of RMDA and choriocapillaris FD% to SDD, which is present in many AMD eyes with especially poor RMDA.^45,46 These deposits are prominent in highly rod-dominant regions of superior perifovea and near-periphery, with far fewer deposits occurring in the cone-dominant fovea.^92–94 Thus our current data showing a relative vulnerability of subfoveal choriocapillaris are consistent with others also showing effects removed from the primary region of SDD accumulation.^95,96 Our ongoing work directly addresses RMDA in eyes of known SDD abundance. 

In conclusion, we have shown a relationship between choriocapillaris integrity and visual function at the transition of aging to AMD that is related to distance from the fovea. These data can support continued efforts to harmonize, systematize, and facilitate means of measuring choriocapillaris flow signal deficits as a continuous-variable imaging biomarker suitable for clinical trials. These efforts will be aided by new tissue-level understanding of choriocapillaris anatomy and hemodynamics.^68,69,72 This microvasculature uniquely provides metabolic exchange with target retinal tissues across a planar interface. The prognostic value of FD% can be clarified at the three-year follow-up visit of ALSTAR2, which is ongoing. 

Acknowledgments

Presented at the 2023 Annual Meeting of the Association for Research in Vision and Ophthalmology. 

Supported by R01EY029595 (CO, CAC); P30EY03039 (CO); Dorsett Davis Discovery Fund, and Alfreda J. Schueler Trust (CO); Unrestricted funds to the Department of Ophthalmology and Visual Sciences (UAB) from Research to Prevent Blindness, Inc., and EyeSight Foundation of Alabama. 

Disclosure: D. Kar, Apellis Pharmaceuticals Inc (E); M. Amjad, None; G. Corradetti, None; T.A. Swain, None; M.E. Clark, None; G. McGwin, None; K.R. Sloan, None; C. Owsley, Johnson & Johnson Vision (C), dark adaptation device (P); S.R. Sadda, 4DMT (C), Abbvie (C), Alexion (C), Allergan Inc. (C), Alnylam Pharmaceuticals (C), Amgen Inc. (C), Apellis Pharmaceuticals, Inc. (C), Astellas (C), Bayer Healthcare Pharmaceuticals (C), Biogen MA Inc. (C), Boehringer Ingelheim (C), Carl Zeiss Meditec (C), Catalyst Pharmaceuticals Inc. (C), Centervue Inc. (C), Genentech (C), Gyroscope Therapeutics (C), Heidelberg Engineering (C), Hoffman La Roche, Ltd. (C), Iveric Bio (C), Janssen Pharmaceuticals Inc. (C), Nanoscope (C), Notal Vision Inc. (C), Novartis Pharma AG (C), Optos Inc. (C), Oxurion/Thrombogenics (C), Oyster Point Pharma (C), Regeneron Pharmaceuticals Inc. (C), Samsung Bioepis (C), Topcon Medical Systems Inc. (C), (R): Carl Zeiss Meditec (R), Heidelberg Engineering (R), Nidek Incorporated (R), Novartis Pharma AG (R), Topcon Medical Systems Inc. (R), (F): Carl Zeiss Meditec (F), Heidelberg Engineering (F), Optos Inc. (F), Nidek (F), Topcon (F), Centervue (F); C.A. Curcio, Heidelberg Engineering (F), (C) Apellis (C), Astellas (C), Boehringer Ingelheim (C), Character Biosciences (C), Osanni (C), Annexon (C), Mobius (C), Genentech/ Hoffman LaRoche (C), Ripple (C)