Evaluating the Persistence of Large Choroidal Hypertransmission Defects using SS-OCT Imaging

Abstract

In age-related macular degeneration (AMD), large choroidal hypertransmission defects (hyperTDs) are identified on en face optical coherence tomography (OCT) images as bright lesions measuring at least 250 μm in greatest linear dimension (GLD). These choroidal hyperTDs arise from focal attenuation or loss of the retinal pigment epithelium (RPE). We previously reported that once large hyperTDs formed, they were likely to persist compared with smaller lesions that were more likely to be transient. Due to their relative persistence, these large persistent choroidal hyperTDs are a point-of-no-return in the progression of intermediate AMD to the late stage of atrophic AMD. Moreover, the onset of these large choroidal hyperTDs can serve as a clinical trial endpoint when studying therapies that might slow disease progression from intermediate AMD to late atrophic AMD. To confirm the persistence of these large choroidal hyperTDs, we studied an independent dataset of AMD eyes enrolled in an ongoing prospective swept-source OCT (SS-OCT) natural history study to determine their overall persistence. We identified a total of 202 eyes with large choroidal hyperTDs containing 1725 hyperTDs followed for an average of 46.6 months. Of the 1725 large hyperTDs, we found that 1718 (99.6%) persisted while only 7 hyperTDs (0.4%) were non-persistent. Of the 7 non-persistent large hyperTDs in 6 eyes, their average GLD at baseline was 385 μm. Of the large hyperTDs ranging in size between 250-300 μm when first detected, only one was not persistent with a baseline GLD of 283 μm. In 6 of the non-persistent hyperTDs, the loss of a detectable large hyperTD was due to the accumulation of hyperreflective material along the retinal pigment epithelium (RPE) and in the retina over the area where the hyperTD was located. This hyperreflective material is thought to represent the migration and aggregation of RPE cells into this focal region where the choroidal hyperTD arose due to attenuated or lost RPE.

1.1. INTRODUCTION

Age-related macular degeneration (AMD), a leading cause of irreversible blindness worldwide among the elderly (Fleckenstein et al., 2021) has several stages of progressive disease severity defined by fundus exam and by using different imaging strategies. (Ferris et al., 2013; Jaffe et al., 2021; Sadda et al., 2018; Spaide et al., 2020) Historically, color fundus imaging (CFI) was the gold standard used to classify the stages of AMD, (Ferris et al., 2013) and these stages include early, intermediate, and late AMD. The hallmarks of early and intermediate AMD are the size and distribution of drusen along with hyperpigmentation, while the features of late AMD include macular exudation caused by neovascularization and geographic atrophy (GA).(Ferris et al., 2013)

Currently, optical coherence tomography (OCT) imaging is the gold standard imaging strategy for diagnosing, staging, and following patients with AMD. (Ferris et al., 2013; Jaffe et al., 2021; Sadda et al., 2018; Spaide et al., 2020) Each of the stages of AMD have been redefined based on their characteristic OCT features. While the OCT features are associated with the typical anatomic changes associated with drusen, hyperpigmentation, geographic atrophy, and exudation, there has been a great deal of interest in using OCT imaging to identify the earliest irreversible changes that herald the transition from intermediate AMD (iAMD) to late AMD with the goal of using this early irreversible transition as a clinical trial endpoint in order to shorten the duration of these studies. The OCT candidate lesions that have been identified as the earliest irreversible changes that herald the onset of atrophy include nascent GA (nGA), incomplete retinal pigment epithelium and outer retinal atrophy (iRORA), complete RORA (cRORA), and large choroidal hypertransmission defects (hyperTDs). (Guymer et al., 2020; Jaffe et al., 2021; Laiginhas et al., 2022b; Liu et al., 2022; Liu et al., 2023a; Sadda et al., 2018; Wu et al., 2020) However, for these lesions to be used as clinical trial endpoints, they need to be irreversible, easily identified, accurately measured, and shown to progress to the more typical form of late AMD known as GA.

Of all these early OCT features that serve as harbingers of disease progression to geographic atrophy, the appearance of large choroidal hyperTDs is the only one that has been accepted currently by the Food and Drug Administration (FDA) for use as a clinical trial endpoint when studying the progression of intermediate AMD to late AMD characterized by the appearance of macular atrophy. Liu et al. (Liu et al., 2023a) showed the usefulness of this OCT biomarker in their proposed clinical trial design that combines not only the onset but also the progression of the endpoint. Firstly, by using en face OCT imaging, they are able to detect the onset of hyperTDs, and secondly, they use the same imaging strategy to measure area growth of the hyperTDs, thereby satisfying the FDA requirement for documenting the rate of change in the growth of the endpoint. (Laiginhas et al., 2022b; Liu et al., 2022; Liu et al., 2023a) Other early OCT biomarkers of atrophic disease, such as nGA, iRORA, and cRORA, are defined by their onset using individual OCT B-scans, but their growth can only be defined along the horizontal dimension unless these lesions are annotated on closely spaced B-scans and compiled for en face imaging, which then becomes the more typical en face imaging strategy used to detect large choroidal hyperTDs.

The detection of large hyperTDs requires the use of a dense raster scan pattern so that the OCT features can be viewed reliably in three dimensions.(Laiginhas et al., 2022b; Liu et al., 2022; Liu et al., 2023a) We found that en face viewing of specific slabs with boundaries positioned 64 μm to 400 μm under Bruch’s membrane allowed for the detection of focal areas where the RPE is lost or attenuated, resulting in increased transmission of light into the choroid and the increased reflectivity from the choroid in these areas appeared as bright spots. We previously reported that once the GLD of these lesions reached at least 250 μm, they tended to persist; however, the likelihood of this persistence was similar when the lesions developed reached a GLD size range of 250 μm to 300 μm. (Shi et al., 2021a) This cut-off for persistence was later confirmed by Nanegrungsunk et al.(Nanegrungsunk et al., 2024) The added advantage of this en face strategy is that once a hyperTD is identified on the en face image, the corresponding B-scan can be easily positioned over the hyperTD to confirm the diagnosis using either the review software on the OCT instrument or a specialized review station. This combined use of en face and B-scan imaging allows for a three-dimensional view of the disease as it progresses.

Recently, we proposed a clinical trial design for the study of treatments for iAMD using the onset and progression of large hyperTDs.(Berni et al., 2024; Liu et al., 2023a) This clinical trial design enrolls eyes with iAMD without any evidence of large hyperTDs. This research by Berni et al. and Liu et al. identified drusen volume and area measurements of hyperreflective foci as important predictors for the onset of hyperTDs, hence serving as OCT biomarkers for the selection of iAMD patients at high risk for progression. The primary endpoint for this trial is the onset and growth of these large hyperTDs. However, to support the use of hyperTDs to serve as a clinical trial endpoint, it was important to further validate that these lesions were likely to persist once detected. To further validate the persistence of these hyperTDs, we longitudinally studied an independent dataset of eyes with large hyperTDs to document whether they persisted and why any of these lesions became non-persistent.

1.2. METHODS

Patients with AMD were enrolled in an ongoing prospective, observational, SS-OCT imaging study at the Bascom Palmer Eye Institute. The study was approved by the Institutional Review Board of the University of Miami Miller School of Medicine, and all patients signed an informed consent for this prospective SS-OCT study. The study was performed in accordance with the tenets of the Declaration of Helsinki and complied with the Health Insurance Portability and Accountability Act of 1996.

1.2.1. Inclusion and Exclusion Criteria

A retrospective review of the prospectively enrolled subjects was performed to identify eyes with treatment-naive, nonexudative AMD and at least one large choroidal hyperTD measuring ≥250 μm in greatest linear dimension (GLD). The earliest visit when this large choroidal hyperTD first appeared was designated as the baseline visit. In addition, to be included in this study, eyes were required to have at least 12 ± 1 months of follow-up from their baseline visit to their last available visit. Exclusion criteria at baseline included a history of exudative MNV, diabetic retinopathy, vascular occlusions, central serous chorioretinopathy, or other retinal diseases associated with drusen-like deposits, such as Stargardt disease or vitelliform dystrophy. We also excluded patients with pathological myopia and any vitreoretinal interface diseases that significantly distorted macular anatomy or those who had previously undergone vitreoretinal surgery. The eye was censored from the study once the eye developed symptomatic exudation due to MNV, which was defined by the appearance of any subretinal or intraretinal fluid on structural OCT B-scans and on the retinal thickness maps with subsequent intravitreal (IV) injections of a vascular endothelial growth factor inhibitor. In addition, eyes developing tears of the RPE were excluded, and if IV pegcetacoplan therapy was initiated, the eye was censored, and the last scan prior to the first IV injection was included.

1.2.2. Imaging Protocols

All patients underwent SS-OCTA imaging (PLEX Elite 9000, Carl Zeiss, Meditec Inc., Dublin, CA) during the study and scans were acquired by one of two trained imaging technicians. The SS-OCTA instrument is powered by a swept source laser with a central wavelength of 1050 nm and a bandwidth of 100 nm, corresponding to a full width at half-maximum axial resolution in tissue of ~5 μm in tissue and an estimated transverse resolution of ~15 μm at the retinal surface. The instrument can operate at scanning speeds of 100 kHz and 200 kHz, whereas the scans evaluated for this study were all acquired at a scanning rate of 100 kHz. All eyes were imaged using the 6x6 mm and 12x12 mm angiographic scan patterns centered on the fovea. Both SS-OCTA scan patterns consisted of 500 A-scans per B-scan and 500 B-scans, with each B-scan repeated twice at each position, resulting in a uniform 12 μm spacing between A-scans in the 6x6 mm scan, and 24 μm in the 12x12 scan pattern. Each volumetric scan was reviewed for quality and signal strength to guarantee the best quality scans to assess hyperTDs, and scans with a signal strength less than 7 based on the instrument’s output as well as those with significant motion artifacts were excluded. Based on our calculations, we estimate that less than 1% of all collected scans were excluded after this quality review step.

1.2.3. Grading of HyperTDs

Large hyperTDs were detected on en face structural images generated from a custom slab positioned 64 to 400 μm beneath Bruch’s membrane (sub-RPE slab) and were identified as areas with increased focal brightness corresponding to the hypertransmission of light into the choroid with a GLD measuring at least 250 μm on the en face image. Review of the corresponding B-scans through the lesions confirmed the presence of choroidal hyperTDs due to the attenuation/loss of the RPE, as previously reported.(Laiginhas et al., 2022b; Lu et al., 2023; Nanegrungsunk et al., 2024; Shi et al., 2021b) Two independent graders (S.B., G.H.) evaluated the en face images for the presence of large hyperTDs. Graders used a semiautomated algorithm developed by Chu et al.(Chu et al., 2022) and set up a threshold with a GLD of ≥250 μm to generate the initial outlines of the hyperTD lesions. A built-in editing tool was used to arrive at the final outlines according to the consensus agreement of the two graders. Any remaining disagreement between the two graders was adjudicated by a senior grader (P.J.R.). The GLD measurements of each choroidal hyperTD and the number of lesions per eye were provided from the algorithm output.

For each eye identified with a large choroidal hyperTD at baseline, all available visits were reviewed, annotated, and processed by the algorithm to identify and measure all hyperTDs ≥ 250 μm in GLD. Multiple hyperTDs could be identified in an eye at baseline and at follow-up visits, and they were each given a unique number in their order of appearance to be tracked over time as the lesions enlarged and merged into larger choroidal hyperTDs. Whenever lesions grew or new ones developed outside of the 6x6 mm field of view, the 12x12 mm scan was used to continue with the tracking process throughout the follow-up visits.(Thulliez et al., 2019) HyperTDs were classified into two groups according to their GLD at first appearance: those with GLD between 250-300 μm, and those ≥301 μm, and their number and GLD measurements were noted. Lesions were further divided into persistent and non-persistent, depending on their consistent detection at the same location on the en face image throughout the follow-up process. These non-persistent hyperTDs were identified and their GLD at baseline described.

1.3. RESULTS

A total of 202 eyes from 147 patients with nonexudative AMD and large hyperTDs were identified for this study. Two graders evaluated all visits for every eye and reached an agreement through discussion. Only 10 out of 202 eyes (5%) presented a disagreement among graders, requiring a third, senior grader for adjudication. The mean age of the subjects at baseline was 76.5 ± 8.2 years (median, 76 years; range, 47-92 years), and 95 subjects (65%) were female. The subjects had undergone SS-OCT imaging between April 27, 2016, and May 23, 2024. Eyes were monitored for a mean of 46.6 ± 23.6 months (median, 46.5 months; range, 11-94 months). A total of 69 eyes were censored at a given timepoint in their follow-up. Of these, 49 were censored on the visit after receiving their first pegcetacoplan injection, 18 were censored when progression to exudative MNV was detected on the scan, one that developed macular edema due to radiation retinopathy, and one that developed an RPE tear. Patients received imaging during their routine visits, which resulted in variable follow-up intervals. The minimum follow-up for all eyes included in this study was 11 months. Of these, 41 eyes (20%) were followed for at least 12 months, 29 eyes (14%) were followed for at least 24 months, 28 eyes (14%) were followed for at least 36 months, 33 eyes (16%) were followed for at least 48 months, 30 eyes (15%) were followed for at least 60 months, 20 eyes (10%) were followed for at least 72 months and 15 eyes (7%) were followed for at least 84 months. The average number of visits per year was 3.60 ± 1.14 visits (median, 3.69; range, 1.09-7.33). When categorized by average number of visits per year, during the duration of follow-up, there were between 1-2 visits for 17 eyes (8%), 2-3 visits for 39 eyes (19%), 3-4 visits for 65 eyes (32%), 4-5 visits for 65 eyes (32%), 5-6 visits for 10 eyes (5%), and 6 or more visits for 6 eyes (3%). These values can be found in tabular format in Table 1 and Table 2.

Table 1.

Tabular representation of follow-up length for study eyes

Follow up length	Number of Eyes (Total: 202)	%
at least 11 months	6	3%
1 - 2 years	41	20%
2 - 3 years	29	14%
3 - 4 years	28	14%
4 - 5 years	33	16%
5 - 6 years	30	15%
6 - 7 years	20	10%
> 7 years	15	7%

Table 2.

Tabular representation of number of visits per year for study eyes

Number of visits per year	Number of Eyes (Total: 202)	%
1-2	17	8%
2-3	39	19%
3-4	65	32%
4-5	65	32%
5-6	10	5%
>6	6	3%

Throughout the monitoring interval, a total of 202 eyes with 1725 hyperTDs with GLD ≥ 250 μm were identified, with an average of 8.5 lesions per eye (standard deviation, 8.6 lesions; range, 1 - 44 lesions per eye). Of these, 1718 hyperTDs (99.6%) from 196 eyes persisted throughout all of the follow-up visits. Seven hyperTDs (0.4%) from 6 eyes did not persist in one or more follow-up visits and are referred to as non-persistent hyperTDs. These 7 non-persistent hyperTDs were followed for an average of 39.1 ± 14.8 months (median, 40 months; range, 19-57 months).

Previously, Shi et al. (Shi et al., 2021a) analyzed the effect of baseline GLD on the persistence of hyperTDs over time. They found that hyperTDs with a minimal GLD threshold cutoff in the range of 250 μm to 300 μm were most likely to be persistent while smaller lesions were more likely to be transient. In the following post-hoc analysis, we considered the behavior of hyperTDs with a GLD ≥ 250 μm. Of the total 1725 large hyperTDs, 347 (20.1%) had a GLD at baseline measuring between 250 μm to 300 μm with an average of 277 ± 15 μm (median, 279 μm; range, 250-300 μm). Only one lesion in this size range was non-persistent with a baseline GLD value of 283 μm. This comprised 0.29% of hyperTDs with baseline GLD between 250 μm to 300 μm. The remaining 6 non-persistent hyperTDs had a GLD at baseline greater than 300 μm, with an average of 402 ± 69 μm (median, 390 μm; range, 328-532 μm). These non-persistent lesions comprised 0.44% of hyperTDs with baseline GLD greater than 300 μm.

Figures 1, 2, 3, and 4 show the typical progression for these large persistent and non-persistent hyperTDs in the study. Figure 1 shows an example of an eye developing a large persistent hyperTD with a GLD at baseline of 447 μm. Throughout the duration of follow-up, the area of this hyperTD continued to enlarge, while the B-scan views show the choroidal hyperTD with progressive loss of the outer retina and RPE.

Figure 1.

An example of a large persistent choroidal hypertransmission defect (hyperTD) that appears and grows in an eye with age-related macular degeneration (AMD). Panels A1-D1 show en face swept-source OCT structural images that were created using a slab positioned from 64 μm to 400 μm under the Bruch’s membrane referred to as the subRPE slab. The areas of interest are identified by color-coded arrows. In A1, the orange arrow identifies a region that will develop into a large persistent hyperTD with a greatest linear dimension (GLD) of at least 250 μm as identified by the green arrows in panels B1 to D1. Panels A2-D2 show the same en face structural slab images with a purple line identifying the position of the B-scans shown in panels A3-D3. The dashed yellow lines in A3-D3 identify the segmentation boundaries of the subRPE slab used to generate the en face images. In A3, the orange arrow identifies a region that will develop into a large persistent hyperTD as identified by the green arrows in panels B3 to D3. Panels B1-3 show the follow-up visit 20 months after the visit shown in panels A1-A3, and the large persistent hyperTD is identified by the green arrow. The B-scan views confirm the grading by showing the presence of the choroidal hyperTD associated with outer retinal attenuation. In B1, the GLD of the hyperTD measures 447 μm. Panels C1-C3 and D1-D3 correspond to follow-up visits at 23 months and 36 months, and these panels show enlargement of this lesion over time. The GLD of the lesion in C1-C3 measures 489 μm, and the GLD of the lesion D1-D3 measures 628 μm.

Figure 2.

An example of a large non-persistent choroidal hypertransmission defect (hyperTD) in an eye with age-related macular degeneration (AMD). Panels A1-D1 show en face swept-source OCT structural images that were created using a slab positioned from 64 μm to 400 μm under the Bruch’s membrane referred to as the subRPE slab. The areas of interest are identified by color-coded arrows. In A1, the orange arrow identifies a region that will develop into a large hyperTD with a greatest linear dimension (GLD) of at least 250 μm as identified by the green arrow in panel B1. Panels A2 to D2 show the same en face structural slab images with a blue and purple line identifying the position of the B-scans shown in panels A3-D3 and A4-D4 respectively, corresponding to the color of the B-scan frame. The dashed yellow lines in A3-D3 and A4-D4 identify the segmentation boundaries of the subRPE slab used to generate the en face images. In A3 and A4, the orange arrow identifies a region that will develop into a large hyperTD as identified by the green arrow in panels B3 and B4. Panels B1-B4 show the follow-up visit 9 months after the visit shown in panels A1-A4, and the large hyperTD is identified by the green arrow. The B-scan view in B3 confirms the grading by showing the presence of the choroidal hyperTD associated with outer retinal attenuation. In B1, the GLD of the hyperTD measures 400 μm. In addition, the en face view in B1 shows dark foci or hypotransmission defects (hypoTD) representing hyperreflective foci (HRF) in the retina and along the RPE, which correspond to the observed findings in the B-scan (B4). Panels C1-C4 show the follow up visit 44 months after the visit shown in panels A1-A4, and the orange arrow in C1, C3 and C4 identifies the region where the hyperTD is no longer evident. In the same region, areas of hypoTDs can be observed on the en face image (C1), which correspond with the B-scan views in C3 and C4 that show an increase in HRF in the retina that have migrated most likely from the thickened RPE seen in the previous visit. Panels D1-D4 correspond to the follow up visit at 50 months from the visit shown in A1-A4, and the orange arrow in D1, D3 and D4 identifies the region where the hyperTD is no longer evident. The hypoTDs remain and are seen on the en face image (D1) and correspond to the B-scans in D3 and D4 that show intraretinal HRF.

Figure 3.

An example of a large non-persistent choroidal hypertransmission defect (hyperTD) in an eye with age-related macular degeneration (AMD). Panels A1-D1 show en face swept-source OCT structural images that were created using a slab positioned from 64 μm to 400 μm under the Bruch’s membrane and referred to as the sub-retinal pigment epithelium (sub-RPE) slab. The areas of interest are identified by color-coded arrows. In A1, the green arrow identifies a large hyperTD with a greatest linear dimension (GLD) of at least 250 μm. Panels A2-D2 show the same en face structural slab images with a purple line identifying the position of the B-scans shown in panels A3-D3. The dashed yellow lines in A3-D3 identify the segmentation boundaries of the subRPE slab used to generate the en face images. The large hyperTD identified by the green arrow in A1 is confirmed by the B-scan view in A3 showing the presence of the choroidal hyperTD associated with outer retinal attenuation. In A1, the GLD for the hyperTD measures 395 μm. Panels B1-B3 show the follow-up visit 25 months after the visit shown in panels A1-A3. The orange arrow in B1 and B3 points to the region where the hyperTD is no longer apparent. In B1, dark foci or hypotransmission defects (hypoTD) represent hyperreflective foci (HRF) along the RPE that are seen in the B-scan view in B3. Panels C1-C3 show the follow-up visit 46 months after the visit shown in A1-A3, where the green arrows in C1 and C3 identify the hyperTD that reappears. The en face view in C1 corresponds to the choroidal hyperTD associated with RPE and outer retinal attenuation seen in the B-scan in C3. In C1, the GLD of the hyperTD measures 280 μm. Panels D1-D3 show the last follow-up visit at 57 months from the visit shown in A1-A3. D1 now shows persistence and enlargement of the lesion over time, with a GLD of 531 μm. The B-scan view in D3 shows a choroidal hyperTD, attenuation of the RPE, and attenuation of the outer retina.

Figure 4.

An example of a large non-persistent choroidal hypertransmission defect (hyperTD) in an eye with age-related macular degeneration (AMD). Panels A1-E1 show en face swept-source OCT structural images that were created using a slab positioned from 64 μm to 400 μm under the Bruch’s membrane referred to as the sub-retinal pigment epithelium (sub-RPE) slab. The areas of interest are identified by color-coded arrows. In A1, the orange arrow identifies a region that will develop into a large hyperTD with a greatest linear dimension (GLD) of at least 250 μm as identified by the green arrow in panel B1. Panels A2-E2 show the same en face structural slab images with a purple line identifying the position of the B-scans shown in panels A3-E3. The dashed yellow lines in A3-E3 identify the segmentation boundaries of the subRPE slab used to generate the en face images. In A3, the orange arrow identifies a region that will develop into a large hyperTD as identified by the green arrow in panel B3. Panels B1-B3 show the follow-up visit 1 month after the visit shown in panels A1-A3, and the large hyperTD is identified by the green arrow. The B-scan views in B3 confirm the grading by showing the presence of the choroidal hyperTD associated with RPE and outer retinal attenuation. In B1, the GLD of the hyperTD measures 532 μm, presenting the largest GLD value at baseline among all non-persistent lesions. Additionally, hypotransmission defects (hypoTDs) are located in the region of the hyperTD on the en face image (B1) but not observed in the selected B-scan in B3. Panels C1-C3 show the follow-up visit 19 months from the visit shown in A1-A3, and the orange arrow identifies the region where the hyperTD is no longer evident. HypoTDs are present at the same region in the en face view in C1, consistent with a thickened RPE due to hyperreflective foci (HRF) in the corresponding B-scan in C3. Panels D1-D3 show the follow-up visit at 47 months from the visit in panels A1-A3, where the green arrow indicates that the hyperTD has reappeared. The en face view in D1 shows a hyperTD, which corresponds to the choroidal hyperTD and disruption of the outer retinal layers seen in the B-scan. In D1, the GLD of the hyperTD is 699 μm. In addition, a large druse associated with the hyperTD can be observed in the B-scan shown in D3. Panels E1-E3 present the follow-up visit 59 months from the visit in A1-A3, showing the enlargement and merging of hyperTDs over time, with a GLD of 3101 μm in E1. The B-scan in E3 shows the collapse of the druse observed in the previous visit, a choroidal hyperTD, and the apparent loss of the RPE and outer retinal layers.

Figure 2 shows an example of an eye with a large non-persistent hyperTD. At baseline, the hyperTD is observed on the en face image with a GLD of 400 μm and corresponding B-scan images showing a choroidal hyperTD. However, at the follow-up visits, the hyperTD is no longer evident and is replaced by areas of choroidal hypotransmission defects (hypoTDs), which represent HRF in the retina and along the RPE as seen on the B-scan images.

Figure 3 shows another example of a large non-persistent hyperTD, which has a baseline GLD of 395 μm. At the follow-up visit, the hyperTD is not evident and hypoTDs are now located in the same region on the en face view and seen on the B-scan as FIRF along the RPE. More recent follow-up visits show the reappearance of the hyperTD on the en face images, corresponding to a larger choroidal hyperTD with attenuation of the RPE and outer retina as seen on the B-scan view.

Figure 4 shows an example of a large non-persistent hyperTD with a GLD of 532 μm, which is the largest GLD at baseline value among the non-persistent lesions. At the follow-up visit, this hyperTD was not apparent and shows a similar pattern as the previous two examples with hypoTDs occupying the same en face region. At later follow-up visits, the hyperTD re-appears and the area of the lesion enlarges and merges with adjacent hyperTDs.

1.4. DISCUSSION

Large choroidal hyperTDs were defined as bright lesions having a GLD of at least 250 μm when viewed on en face OCT images and confirmed using corresponding B-scans.(Shi et al., 2021a) This size cut-off was established based on the high likelihood that these lesions would persist over 3 years compared with smaller lesions that tended to be transient. In this previous report, Shi et al.(Shi et al., 2021a) showed that the threshold cutoff for persistence was between 250 μm to 300 μm, and 250 μm was chosen as a minimum size cut-off. Subsequently, Nanegrunsunk et al. (Nanegrungsunk et al., 2024) investigated and confirmed this size cut-off for persistence using an independent dataset of eyes followed through 2 years. However, we wanted to further investigate why some of these large hyperTDs might become undetectable during follow-up given our understanding that these lesions developed due to the attenuation or loss of the RPE, and the RPE was not thought to have regenerative capability.(Bearelly et al., 2009; Laiginhas et al., 2022b; Lujan et al., 2009) To investigate this question of non-persistence, we studied an additional cohort of eyes with large hyperTDs with the goal of explaining how these hyperTDs became undetectable.

This current cohort of eyes with large hyperTDs confirms our previous definition of persistence by showing a rate of non-persistence of only 0.41%, which corresponded to seven lesions from a total of 1725 hyperTDs with a GLD of 250 μm or greater. Of the seven non-persistent lesions, six of the lesions had a GLD of over 300 μm at the time of their disappearance with these lesions having a range of GLD measurements from 328 μm to 532 μm. The case representing the upper end of this range is shown in Figure 4 and demonstrates that even larger hyperTDs may become undetectable. However, it should be noted that these 6 cases represent a very small sample of exceptions (0.44%) from the total 1378 lesions that measured above 300 μm. Of note, when we followed hyperTDs with GLD in the range of 250-300 μm at baseline, only one (0.29%) of the 347 lesions became non-persistent after the GLD reached a size of 283 μm. In their study showing the progression of large hyperTDs to GA, Laiginhas et al.(Laiginhas et al., 2022b) observed a small group of cases in which the hyperTD did not persist and evolved into hypoTDs. HypoTDs arise when melanin containing cells, most likely RPE cells, proliferate, aggregate, and migrate along the RPE or into the retina space and block the transmission of light into the choroid due to their non-transparency. (Berni et al., 2024; Laiginhas et al., 2022a) Figures 2, 3 and 4 show a similar clinical course as hyperTDs become no longer detectable. This pattern was observed in 6 of our 7 non-persistent cases, where only 1 hyperTD appears to fade without presence of prominent hypoTDs. When the large hyperTDs become non-persistent, they are replaced with hyperreflective foci along the RPE and in the retina. The hyper-reflective material in these foci is optically opaque and prevents the transmission of light into the choroid, casting a shadow on the subRPE slab, which causes the appearance of dark lesions on en face images known as hypoTDs. (Augustin et al., 2023; Laiginhas et al., 2022a; Laiginhas et al., 2022b) These hyperreflective foci have been associated in many reports as risk factors for the progression of iAMD to geographic atrophy in AMD. (Berni et al., 2024; Hirabayashi et al., 2023; Jaffe et al., 2021; Sadda et al., 2020) Based on our natural history data, it is reasonable to assume that some of these regions with HRF actually arose from foci of hyperTDs that were later concealed by the migration of HRF into these regions. In support of this possibility, in Figures 3 and 4, we showed the reappearance of the hyperTDs in the same area in later visits. Therefore, the likely scenario is that these lesions were temporarily hidden from view while the RPE atrophy evolves, and eventually, the hyperTD reappears. This raises the question of whether eyes with a large burden of HRF should be excluded from clinical trials designed to enroll eyes without large hyperTDs, with the endpoint being the formation of hyperTDs.(Liu et al., 2023b) If the HRF are concealing eyes that have already developed hyperTDs, then these eyes may progress to detectable hyperTDs at a more rapid rate. However, due to the small number of large hyperTDs that became non-persistent (0.41%), the likelihood is remote that these eyes would impact the results from a prospective, randomized clinical trial.

The advantages of using en face OCT imaging to detect and monitor hyperTDs in this study include the ease and reproducibility of this method for early atrophy detection.(Laiginhas et al., 2022b; Liu et al., 2022; Schaal et al., 2017; Shi et al., 2021a; Yehoshua et al., 2013; Yehoshua et al., 2011) En face imaging enables rapid screening for hyperTDs while also being able to detect other relevant imaging features such as hypoTDs. (Laiginhas et al., 2022a) In addition, these en face views facilitated the accurate quantification of the GLD considering the vertical, horizontal and diagonal dimensions, compared with only the horizontal dimension offered on B-scans. As noted previously, once a lesion is detected on an en face image, B-scans can then be used to confirm the finding. On B-scans, these large hyperTDs associated with documented outer retinal and RPE atrophy would define cRORA in any B-scan dimension, so all cRORA are large hyperTDs. (Corvi et al., 2023; Wang et al., 2023) However, the en face presence of a large hyperTD can precede the formation of typical cRORA. Furthermore, by using the 12x12 mm scan pattern, we can be assured that all foci of hyperTDs were identified in our analysis, since those atrophic lesions that existed beyond the field-of-view limits of a 6x6 mm scan pattern, were detected on the 12X12 mm scans during follow-up. In previous studies, when the same regions of atrophy were scanned using these two scan patterns, the en face measurements of GA were comparable. (Thulliez et al., 2019) Thus, we always perform both scan patterns. Another advantage of our study was a sample size of 1725 hyperTD lesions in 202 eyes from 147 patients with non-exudative AMD, which allowed for an adequate sampling of both persistent and non-persistent lesions. In addition, the average follow-up of 46.6 months exceeded the 36 months from our previous study. (Shi et al., 2021a)

Limitations of this study include the variability in follow-up intervals. Nevertheless, our dataset’s average follow-up of 46.6 months with an average of 3.60 visits per year, provided adequate data to support our conclusions. Another limitation is that we performed our imaging and analysis using SS-OCT scans. A research report (Herrera et al., 2024) evaluated the identification and measurement of hyperTDs with a GLD ≥ 250 μm on en face sub-RPE images generated from SS-OCTA and SD-OCTA scans from the same eye taken on the same day. Comparison of mean square root area and number of hyperTDs from manual annotations by independent graders showed high inter-device concordance and yielded nearly identical results. For the quantification of hyperTD area and GLD, we used a semi-automated segmentation algorithm to initialize the annotation of subRPE images, which required that every hyperTD had to be reviewed and manually corrected when necessary by the graders. While a fully automated algorithm to detect hyperTDs would be ideal, it is not yet available. Lastly, while we have observed in two cases that large non-persistent hyperTDs may reappear, we are unable to specify with certainty whether this was indeed the same lesion or rather the extension of a new adjacent hyperTD. Nevertheless, both the en face and B-scan views confirmed the involvement of the same area as the original lesion.

In summary, we confirmed the persistence of large hyperTDs with a GLD equal to or greater than 250 μm using a larger cohort of eyes with longer follow-up than previously reported.(Nanegrungsunk et al., 2024; Shi et al., 2021a) Moreover, we observed that the phenomenon of non-persistence among large hyperTDs develops due to migration of HRF into the region of hypertransmission.

HIGHLIGHTS.

• En face OCT imaging identifies large hypertransmission defects.

• Large hypertransmission defects have a greatest linear dimension of at least 250μm.

• The overall persistence rate for large hypertransmission defects was 99.6%.

• Non-persistence is observed once the lesion becomes occupied by hyperreflective foci.

Abbreviations:

AMD

Age-related Macular Degeneration

BM

Bruch’s membrane

CAM

Classification of Atrophy Meeting

CFI

Color Fundus Imaging

cRORA

Complete Retinal Pigment Epithelium and Outer Retinal Atrophy

FAF

Fundus Autofluorescence

GA

Geographic Atrophy

GLD

Greatest Linear Dimension

HyperTD

Hypertransmission Defect

MNV

Macular neovascularization

NIR

Near-infrared Reflectance

RPE

Retinal Pigment Epithelium

SS-OCT

Swept-Source Optical Coherence Tomography

SS-OCTA

Swept-Source Optical Coherence Tomography Angiography