Purpose
To evaluate hypotheses about the role of acquired vitelliform lesion (AVL) in age-related macular degeneration pathophysiology.
Design
Laboratory histology study; retrospective, observational case series.
Methods
Two donor eyes in a research archive with AVL and age-related macular degeneration were analyzed with light and electron microscopy for AVL content at locations matched to ex vivo B-scans. A retrospective, observational clinical cohort study of 42 eyes of 30 patients at 2 referral clinics determined the frequency of optical coherence tomography features stratified by AVL fate.
Results
Histologic and clinical cases showed subretinal drusenoid deposit and drusen. Ultrastructural AVL components in 2 donor eyes included retinal pigment epithelium (RPE) organelles (3%-22% of volume), outer segments (2%-10%), lipid droplets (0.2%-12%), and a flocculent material (57%-59%). Of 48 AVLs (mean follow-up 46 ± 39 months), 50% collapsed to complete RPE and outer retinal atrophy, 38% were stable, 10% resorbed, and 2% developed neovascularization. The Early Treatment Diabetic Retinopathy Study grid central subfield contained 77% of AVLs. Hyperreflective foci, ellipsoid zone disruption, and hyperreflective thickening of the RPE-basal lamina–Bruch membrane band were common at maximum AVL expansion. Collapsing and noncollapsing AVLs had different growth rates (rapid vs slow, respectively).
Conclusions
AVL deposits contain unexpectedly low levels of RPE organelles and outer segments. Subfoveal predilection, reflectivity on optical coherence tomography, hyperautofluorescence, yellow color, and growth–regression phases suggest dysregulation of lipid transfer pathways specific to cone photoreceptors and supporting cells in formation of AVL deposit, analogous to drusen and subretinal drusenoid deposit. Prediction of AVL outcomes via growth rates should be confirmed in larger clinical studies.
To refine the progression sequence of age-related macular degeneration (AMD), an international consensus group has cataloged features of potential prognostic value and visible by optical coherence tomography (OCT). These features include acquired vitelliform lesion (AVL), the subject of this report.
Gass and later investigators described AVLs as solid yellow, slightly elevated, and hyperautofluorescent macular lesions approximately one-third disc diameter in size, often subfoveal in location, containing a pigmented spot. On OCT, , , the yellow material correlates to a distinct layer of hyperreflective material between the photoreceptors and an irregularly contoured retinal pigment epithelium (RPE) band, sometimes containing reflective puncta resembling cells. Occurring in several conditions, AVL is differentiable from Best disease, , caused by mutant VMD2 encoding a calcium channel and leading to multifocal disruptions of the RPE–photoreceptor interface. AMD pathology associated with AVL include drusen, subretinal drusenoid deposits (SDDs), hyperreflective foci (HRF), subretinal fluid (SRF), pigment epithelial detachments (PEDs), and macular neovascularization (MNV). , Histologic analyses of AVLs further established associations with basal laminar deposit (BLamD) and photoreceptor (PR) degeneration, and a subretinal deposit containing outer segment (OS) debris and loose organelles from RPE. , , In a large series, 12% of AVL in eyes with AMD led to type 1 MNV and 44% led to atrophy, the main predictor of poor visual outcome.
We hypothesize that AVLs share commonalities with the characteristic extracellular deposits of AMD as refined by recent research. Soft drusen, the best studied intraocular risk factor for progression on a population level, are lipid-rich, remarkably confined to the central macula, , and lead to type 1 MNV , and atrophy. , BLamD between RPE cell bodies and native RPE basal lamina is ubiquitous to AMD and is crossed by soft drusen material in transit to the circulation. SDDs between the RPE and PRs first appear near the vascular arcades where rods are numerous and confer risk for end stages that are distinct from drusen. Drusen and SDDs are proposed to represent dysregulated lipid trafficking within cellular ecosystems supporting cone- and rod-specific physiologies, yielding parallel routes to advanced disease, a radically simplifying notion if proven. , In addition, a model to explain a center of foveal cone resilience, due to sustenance by Müller glia, atop a wider region of parafoveal rod vulnerability, due to impaired transport from the choroidal circulation, has been articulated. , A strong (but not exclusive) preference of AVL for subfoveal locations suggests a relationship with cones and their supporting cells.
The role of AVLs in AMD pathophysiology could be clarified by new quantitative data on deposit ultrastructure, dimensions, and distribution, associations with other tissue features, and clinical lifecycle. To validate OCT features of AVL, our current study used high-resolution histology and transmission electron microscopy (TEM) matched to ex vivo OCT B-scans from postmortem donor eyes. To identify diagnostic biomarkers and evaluate biologic hypotheses about the clinical appearance of the deposit, patients with AMD in 2 clinical centers were retrospectively reviewed. We also developed a lifecycle model to guide practitioners in counseling patients with AVL.
METHODS
HISTOPATHOLOGY STUDY
Case identification, tissue preparation, and imaging
We used published methods that are included in our Supplemental Material. AVL cases were identified among preserved donor eyes in a research archive by ex vivo OCT imaging after anterior segment removal. Of 3 eyes of 2 donors containing AVLs (Supplemental Figure 1), 2 eyes of 2 donors are reported herein. Methods and rationale for eye selection, choice of fixatives, histologic methods, photodocumentation, and analysis approaches were those used for the Project MACULA (MACulopathy Unveiled by Laminar Analysis) online resource for AMD histopathology. , Tissue samples were prepared using electron microscopy methods (osmium-containing postfixation and epoxy embedding) and viewed by light microscopy. Eight-mm–wide 0.8-µm thick sections were scanned in their entirety with oil immersion objectives for high resolution. At the AVL center, one 30-µm–thick section was re-embedded for TEM examination.
Image analysis software
For reviewing and analyzing images (histological and clinical), customized plugins in ImageJ ( https://imagej.nih.gov/ij/download.html ) were developed (by K.R.S.; Creative Computation).
Histopathologic review
Submicrometer sections of 2 cases were imaged by light microscopy and matched with corresponding ex vivo OCT B-scans using overall tissue contour and patterns of reflective material as landmarks (Supplemental Figure 2). AVL volume was calculated using OCT B-scans and the Cavalieri principle, using the plug-in “Draw Volume OCT” (available at https://sites.imagej.net/CreativeComputation/ ).
AVL content was quantified in TEM images via point-counting stereology. A custom ImageJ plug-in navigated through an image marking 54 points on a predefined sampling grid (Supplemental Figure 3). The examiner (M.B.) identified tissue features underlying each grid point via dropdown menu (details in the Results section). The proportion of scored grid points relative to total grid points is an unbiased estimator of tissue component volume.
CLINICAL IMAGING STUDY
At 2 referral centers for retinal disease (University Hospital Zurich [USZ] and Vitreous Retina Macula Consultants of New York [VRMNY]) we sought patients with AVL exclusively in the setting of AMD. Detailed methods for patient selection and multimodal imaging are provided in the Supplemental Material. AVLs contained subretinal hyperreflective material on OCT that was yellow in CFP and hyperautofluorescent in fundus autofluorescence (FAF). Either CFP or FAF alone was sufficient if only one modality was available. The subretinal dome-shaped accumulations of amorphous reflective material was considered a deposit (in the sense of drusen and SDD); the entire outer retinal complex was the AVL.
AMD was defined as presence of ≥1 druse with a minimum width of 125 µm on OCT and/or ≥5 SDDs visible on ≥2 OCT B-scans. AMD eyes with hemorrhage or MNV at baseline were excluded, as were eyes with AVL-associated conditions such as pseudoxanthoma elasticum or central serous chorioretinopathy. Best-corrected visual acuity (BCVA) was recorded.
We use consensus OCT nomenclature with these elaborations. For the RPE–Bruch membrane (BrM) complex, we use the term RPE–basal lamina (BL)–BrM (RPE+BL-BrM). This designation incorporates the appearance of BLamD, a layer of mostly extracellular matrix material between RPE and its basal lamina, as well as drusen and sequela between the RPE–BL and the inner collagenous layer of BrM (sub-RPE–BL space). , The reflective band atop drusen is called RPE+BL, where BL means basal lamina, BLamD, or both.
The anterior AVL boundary was the external limiting membrane (ELM). The posterior AVL boundary was RPE+BL-BrM, unless the AVL was above a PED, in which case the posterior boundary was RPE-BL.
Analysis of morphologic and topographic features in OCT volumes
OCT, near-infrared reflectance (NIR), FAF (488 nm), and CFP images of identified cases were reviewed by 1 grader from each center (M.B. and T.B. from USZ and VRMNY, respectively). To ensure between-grader consistency, areas and volumes from 6 VRMNY cases and 9 USZ cases were measured by both graders and compared. A mixed effects model analysis did not show a significant difference. Thus, data from the 2 centers were combined. AVL locations were documented using the Early Treatment Diabetic Retinopathy Study (ETDRS) grid system using Spectralis software.
AVL dimensions were documented for use in several analyses. Cross-sectional areas were measured in the B-scan containing the maximum AVL extension, using the Draw Region tool (n = 48 AVLs). For AVLs appearing on ≥5 adjacent B-scans (n = 27 AVLs), volume was also measured, using the ImageJ plug-in described above for histology. To define a clinical lifecycle, these measures were assessed in 6-month steps for patients with follow-up ≥6 months (n = 27 patients). To assess AVL growth in eyes with ≥2 visits, areas (n = 42) and volumes (n = 23) were plotted relative to the date of maximal expansion, excluding eyes with maximal area at baseline. Trendlines for growth were determined using eyes with >2 visits 6 months apart (n = 27 patients) using RStudio software (PBC). To compare regional differences in AVL dimensions with regional differences in the distribution of cone photoreceptors, a ratio of mean maximal cross-sectional areas of AVLs centered in the ETDRS central subfield (n = 37 AVLs) vs those in the ETDRS inner ring (n = 11 AVLs) was computed. Areas were raised to the 3/2 power for computing a volume ratio and compared with the ratio of mean cone densities (cells/mm 2 ) in ETDRS subfields found by histology. , To determine the impact of AVLs on vision, the relationship of area at maximal expansion with BCVA (n = 41) was probed with Spearman correlation statistics.
Over the follow-up period, cases were evaluated for OCT characteristics, as demonstrated for 3 representative cases in Figure 1 . AVL deposits were graded as homogenous when 75% of the cross-sectional area had a single reflectivity characteristic and was not obviously part of either continuous layers of inner segments and OS internally or RPE–BL band externally Figure 1 ., A shows an inhomogeneous AVL. For AVLs located atop a drusenoid PED, the PED was not included in the measurements. PEDs were defined as an elevation of hyperreflective RPE+BL band, ≥350 µm in the narrowest diameter. , HRF had reflectivity similar to or greater than the RPE and a minimum area of 3 pixels, as shown in Figure 1 , A and C Figure 1 ., B and C show focal hyperreflective thickening of the RPE+BL band, ie, twice as thick as the normal band Figure 1 ., C shows focal hyperreflective thickening of the RPE+BL band with shadowing of posterior structures. All panels contain areas of hyporeflective splitting of the RPE+BL-BrM band, which correlates to BlamD. , Figure 1 , A shows an intact ellipsoid zone (EZ); the EZ in Figure 1 , B and C is discontinuous.
To determine tissue-level associations of AVLs, fundus features were assessed using published descriptions (Supplemental Table 1): by OCT, SDD, drusen, cuticular drusen, complete RPE and outer retinal atrophy (cRORA), SRF, epiretinal membranes, MNV, and subfoveal choroidal thickness; by CFP, pigment.
RESULTS
HISTOPATHOLOGY STUDY
Supplemental Figure 1 shows ex vivo imaging of eyes from 2 white male donors (90 and 88 years of age, respectively). Using OCT, the AVL of case 1 (Supplemental Figure 1, C) has a large and distinctive dome, with retina attached. The internal reflectivity is homogenous with self-shadowing, ie, brighter anteriorly than posteriorly, and evenly distributed choroidal shadowing posterior to it. This correspondence of B-scans and histology sections through the subretinal dome is shown in Supplemental Figure 2. The AVL of case 2 (Supplemental Figure 1, D) is small and compact relative to that of case 1, with a dome-shaped hyperreflective area, artifactually detached overlying retina, and choroidal shadowing. Although ex vivo CFP is not conclusive in establishing a yellow lesion in either case (Supplemental Figure 1, A and B), case 2 did exhibit a heavily pigmented spot in the foveal center (Supplemental Figure 1, B).
Figures 2 and 3 show microscopic analyses of cases 1 and 2, respectively. Figure 2 , A and Figure 3 , A are overviews showing locations of higher magnification views in other panels. Characteristic AMD features include soft druse material (drusen in Figure 2 , A and B 27,28 and basal mounds in Figure 3 , C and G), thick BLamD that thins directly under the AVL ( Figure 2 , A and Figure 3 , A), and dysmorphic but continuous RPE ( Figure 2 , A, C, and H; Figure 3 , A and G). SDD was abundant extrafoveally in case 1 (94% of glass slides) and near the fovea for case 2 ( Figure 3 , A, D, and H). Figure 2 , G shows absent OS, short inner segments, and thinned outer nuclear layer above and next to the AVL in case 1. Case 2 had shortened OS over SDD ( Figure 3 , H and I); PR directly over the AVL could not be assessed. Neither AVL had evidence of invading cells. Case 1 had a diffuse epiretinal membrane with associated distortions of the inner nuclear and outer plexiform layers (Supplemental Figure 2).
Cases 1 and 2 differed in the nature and degree of RPE disruption. Figure 2 , A, C, and H show for case 1 the transition between a continuous layer of intact RPE cells and an extensive collection of RPE organelles, liberated into the deposit. Organelles localized to the deposit base ( Figure 2 , K) and were confined to its center under the fovea (Supplemental Figure 2). This collection was dominated by typically electron-dense RPE melanosomes, lipofuscin, and melanolipofuscin and included some mitochondria; nuclei were not detected. One cell at the transition shows widely spaced organelles as if it were expanding (Supplemental Figure 4). Deep within the deposit ( Figure 2 , D), a spherical group of lightly stained pigment granules resembled an RPE cell. A cell with typically stained granules was seen crossing the ELM in Figure 2 , E. Out-of-layer RPE cells like Figure 2 , E were observed on 72% of glass slides and seen to be intraretinal on 20% of slides. Case 2 had a massive, nonnucleated tower of liberated RPE granules atop a layer that was still continuous but unhealthy ( Figure 3 , A and K).
Quantification of deposit ultrastructure is presented in Figure 4 and Table 1 and is also shown in Figures 2 and 3 . Five content categories were defined ( Figure 4 ). The largest component (59% for case 1, 57% for case 2) was a flocculent accumulation of small heterogeneous shapes ( Figure 3 , J). The second largest component was distinct from the 4 named components and called “other” (16% for case 1, 20% for case 2). A third component was RPE granules, which accounted for 3% of the deposit in case 1 ( Figure 2 , K) and 22% in case 2 ( Figure 3 , K). A fourth component was smooth spherical profiles with homogenous interiors resembling lipid droplets (12% in case 1 [ Figure 3 , J]; undetectable in case 2). The fifth and smallest component (10% case 1, 2% case 2) was OS-derived material, ie, stacks of lamellar disc–like structures resembling OS in unaffected areas ( Figure 2 , I through K and Figure 3 , L). Neither deposit stained well with toluidine blue or osmium (for saturated fatty acids).
| Acquired Vitelliform Lesion Dimensions | Specimen 1 | Specimen 2 |
|---|---|---|
| Diameter (µm) T–N axis | 1446 | 711 |
| Diameter (µm) S–I axis | 974 | 600 |
| Height (µm) | 286 | 141 |
| Area of base (mm 2 ) | 4.42 | 1.34 |
| Volume (mm 3 ) | 0.23 | 0.03 |
| Acquired vitelliform lesion content | % of Overall Content, Mean ± SD | |
| Unidentified flocculent material | 59 ± 1.7 | 57 ± 4.1 |
| Other | 16 ± 2.2 | 20 ± 4.0 |
| Smoothly structured spherical profiles | 12 ± 1.8 | 0 ± 0.2 |
| Outer segment derived material | 10 ± 0.8 | 2 ± 0.1 |
| RPE granules | 3 ± 0.9 | 22 ± 0.4 |
CLINICAL CASE SERIES
In 2 centers, 62 eyes with AVL were identified. Of these, 42 eyes of 30 patients (16 [53%] females, 14 [47%] males) aged 78.0 ± 10.3 years, seen at 385 visits, had AMD and thus met the criteria for analysis ( Table 2 ). The mean BCVA at the baseline visit was 0.18 ± 0.16 logarithm of the minimum angle of resolution (logMAR; range 0.0-0.7 logMAR; mean Snellen equivalent 20/20 to 20/100). The mean BCVA at final visit was 0.51 ± 0.34 logMAR (range 0.0-1.3 logMAR; mean Snellen equivalent 20/20 to 20/400). Overall, BCVA worsened between baseline and final visits (0.34 ± 0.33; range −0.2 to 1.1) and was unrelated to AVL area at maximal expansion ( P > .45).
