Classification of Exudative Age-Related Macular Degeneration With Pachyvessels on En Face Swept-Source Optical Coherence Tomography

Abstract

Purpose: The purpose of this study was to classify exudative maculopathy by the presence of pachyvessels on en face swept-source optical coherence tomography (SSOCT).

Methods: Consecutive patients with signs of exudative maculopathy underwent SSOCT, fluorescein and indocyanine green angiography (ICGA), ultra-widefield fundus color photography, and autofluorescence examinations. Images were analyzed in a masked fashion by two sets of four examiners in different sessions: (1) the presence of pachyvessels in en face OCT and (2) features of exudative maculopathy in conventional imaging modalities. Quantitative data obtained were subfoveal choroidal thickness (SFCT) and choroidal vascularity index (CVI), which was the ratio of choroidal vessels lumen area to a specified choroidal area from binarized cross-sectional OCT scans.

Results: Pachyvessels was observed in 38 (52.1%) of 73 eyes. The pachyvessels group was associated with younger age (69.1 ± 9.4 years, odds ratio [OR] = 0.95, 95% confidence interval [95% CI] = 0.90–0.97, P = 0.04), presence of polypoidal lesions (OR = 3.27, 95% CI = 1.24–8.62, P = 0.01), increased SFCT (OR = 1.08, 95% CI = 1.02–1.14, P < 0.01), and increased CVI (65.4 ± 5.3, OR = 1.12, 95% CI = 1.02–1.23, P = 0.01). In multivariate regression, CVI significantly correlated with pachyvessels (OR = 1.24, 95% CI = 1.03–1.55, P = 0.04).

Conclusions: Exudative maculopathy could be classified based on differences in choroidal vasculature morphology. Current results implied that choroidal hemodynamics may be relevant to variable natural history and treatment response in neovascular AMD and polypoidal choroidal vasculopathy.

Age-related macular degeneration (AMD) is one of the leading causes of vision loss worldwide.¹ The prevalence of polypoidal choroidal vasculopathy (PCV), a presumed subtype of AMD, was reported to be between 22.3% and 61.6% among Asians and 8% and 13% among Caucasians.² Whether PCV is a distinct disease entity or lies within the spectrum of AMD is still a subject of controversy.³ Yannuzzi et al. first described PCV as a distinct form of exudative maculopathy associated with recurrent serosanguinous detachments of the RPE and neurosensory retina.⁴ Although AMD and PCV patients share some common genetic and clinical features, they also have different histopathology, natural history, risk factors, and response to treatment.^5–8 Appropriate classification of AMD and PCV phenotypes is paramount to optimization of treatment outcomes. 

Recently, the term pachychoroid has been used to describe the configuration of a thick choroid that can be associated with a number of retinal conditions including central serous chorioretinopathy (CSC), PCV, and retinal pigment epitheliopathy.^9–13 Associated clinical features of pachychoroid include reduced tessellation overlying the area of thick choroid in fundoscopy examination, choroidal vascular hyperpermeability (CVH) in indocyanine green angiography (ICGA), increased choroidal thickness in optical coherence tomography (OCT) B-scans, and focal or diffuse dilatation of outer choroidal vessels sometimes with club-shaped posterior termination or “pachyvessels,” observed using en face OCT scans. 

Polypoidal lesions can occur in eyes with or without pachychoroid background. Dansingani et al. observed PCV occurring in eyes with pachychoroid configuration.¹⁴ On the other hand, polypoidal lesions have been observed in eyes with tilted disc syndrome and high myopia with posterior staphyloma, which do not have pachychoroid background.¹⁵ Hence, Balaratnasingam et al. proposed an alternative classification for patients with polypoidal lesions.¹³ Polypoidal choroidal neovascularization (CNV) describes the presence of all neovascular complexes associated with polyps but without pachychoroid features.^11,16 Pachychoroid neovasculopathy describes the occurrence of type 1 CNV, with or without polypoidal lesions, occurring on a pachychoroid background.¹¹ 

The advent of multimodal imaging has revealed new structural information in AMD and PCV patients. Long wavelength swept-source OCT (SSOCT) enables deep imaging of the choroid and the choroid–scleral interface (CSI) within a short acquisition time.¹⁷ The high scanning speed also allows for raster scans to be taken with sufficient density to produce volume maps of the choroid for the construction of en face (c-scans) sections. High-resolution images enable visualization of distinct deep choroidal vessels. 

We hypothesize that, in patients who present with exudative maculopathy, a clinical distinction could be made based on the presence or absence of pachychoroid phenotype. Using en face OCT, a number of observational studies have reported the presence of pachyvessels in PCV patients.^13,17–19 However, most studies did not correlate the findings to the wider clinical manifestations in neovascular AMD and PCV.^13,18–20 The purpose of this study was to classify exudative AMD by choroidal morphology, determined by the presence of pachyvessels on en face SSOCT and quantitative measurements of choroidal vascularity index (CVI) from binarized OCT B-scan images, as well as the subfoveal choroidal thickness (SFCT). We then systematically evaluated for the correlation of choroidal morphology with detail clinical and imaging features. 

Methods

Consecutive patients older than 50 years with new onset or recurrence of symptomatic, exudative maculopathy and their fellow eyes were prospectively recruited from the retina clinics at the University Eye Centre at the Hong Kong Eye Hospital and the Prince of Wales Hospital from November 2015 to May 2016. The exclusion criteria included the following: spherical equivalent refractive error ≤−6.00 diopters (D) or axial length ≥26 millimeters (mm); obscuration of choroidal images by existence of significant media opacity or thick subfoveal hemorrhage, history of ocular inflammation, history of retinal detachment, previous vitrectomy, intraocular surgery (including cataract surgery) in the study eye within 1 year, history of ocular trauma, and glaucoma in the study eye. Eyes with a history of intravitreal triamcinolone acetonide injection were also excluded; however, eyes that had undergone anti-VEGF therapy or photodynamic therapy (PDT) more than 3 months ago were included in this study. The study was approved by institutional review board and adhered to the tenets of the Declaration of Helsinki. Informed consent was signed by all subjects before the investigative procedures. Each patient was interviewed using a structured questionnaire to obtain their demographic data and medical history, including smoking status, hypertension, diabetes, hyperlipidemia, ischemic heart disease, cerebrovascular accident, systemic steroid use, and past episodes of CSC. 

All patients received comprehensive ophthalmic examinations, including best-corrected visual acuity (BCVA), refraction, intraocular pressure (IOP) measurement with Goldmann tonometry, axial length measurement with a laser interferometry (IOL master; Carl Zeiss Meditec, Inc., Dublin, CA, USA), slit-lamp biomicroscopy, dilated fundal examination, ultra-widefield color fundus photography and autofluorescence (Optos Camera; Daytona, Dunfermline, Scotland, United Kingdom), fluorescence angiography (FA) and ICGA with a confocal scanning laser system (HRA Spectralis; Heidelberg Engineering, Dossenheim, Germany), and SSOCT (DRI Triton; Topcon, Tokyo, Japan) for B-scan and en face images. 

Two masked retinal specialists, experienced in the assessment and management of AMD and PCV, independently reviewed the ultra-widefield fundus color photos and autofluorescence and early to late frames of FA and ICGA images. The presence and types of drusen and presence of hemorrhage were graded according to color fundus and autofluorescence photographs. In FA images, type I CNV is characterized by early well-demarcated hyperfluorescence with progressive leakage. Type II CNV is characterized by an ill-defined area of stippled hyperfluorescence or late leakage of undetermined source. In ICGA, the graders evaluated for presence of polypoidal lesions and their configurations (solitary or cluster). A branching vascular network (BVN) was distinguished from a CNV as it is perfused simultaneously with the choroid on ICGA without leakage on FA, and CVH was defined as multifocal areas of hyperfluorescence with blurred margins within the choroid, followed by minimal extension of focal hyperfluorescent area in middle-to-late ICGA phases.²¹ Any discrepancy in the grading of the photographic or angiographic features between the two examiners was resolved by open adjudication with a senior retinal specialist. 

Swept-Source OCT Image Acquisition

All patients underwent SSOCT imaging, which incorporates a light source at 1050-nm wavelength and acquires 100,000 A-scans/s. It has an axial resolution of 8 μm and lateral resolution of 20 μm. En face images were generated by volume data of the posterior pole over a 12- × 9-mm area consisting of 256 horizontal B-scans. The data were subjected to automated topology normalization (flattening) using the inbuilt software (IMAGEnet 6; Topcon Corporation, Tokyo, Japan) with Bruch's membrane as a reference plane for flattening. The modified volume data were further segmented over a configurable slab thickness to produce en face images at various pixel depths below the reference plane, with a depth scale of 2.6 μm/pixel.¹³ 

Another set of two observers, who were masked to the clinical features on FA and ICGA, independently identified the presence or absence of pachyvessels in the en face OCT images. The observers were also blinded to the pairing of fellow eye to its contralateral study eye. To date, no large, normative database is available on the morphology of choroidal vessels in en face images. Hence, we obtained en face SSOCT images of the deep choroid from 60 age-matched healthy subjects as control. Fifteen eyes of 15 control subjects for each decade of age ranged from 50 to 89 years were used for comparison. Because there is no distinct boundary between the inner and outer choroidal layers, segmentation at the Haller's layer was confirmed when the characteristic pattern of outer choroidal vessels appeared.^22–24 The presence of pachyvessels is characterized by focal or diffuse dilatation of outer choroidal vessels, sometimes with evidence of club-shaped posterior termination. These signs were consistent with the features identified in the case series with pachychoroid spectrum diseases reported previously.^14,18–20 One of the observers studied all the images in another session for assessment of intraobserver variability. Figure 1 illustrates the representative en face SSOCT choroidal images of eyes from healthy control subjects compared with eyes with and without pachyvessels identified. 

Figure 1

 

En face SSOCT images of the Haller's layer of the choroid from control eyes (left column) compared with disease eyes with pachyvessels (center column) and without pachyvessels (right column). Images are categorized according to age ranges: (first row) 50 to 59 years old, (second row) 60 to 69 years old, (third row) 70 to 79 years old, and (fourth row) 80 to 90 years old. All of the en face images are segmented in the outer choroid where vessels are discrete, wider in diameter compared with the inner choroidal vessels, and organized in a radial distribution outward from the center of the macula toward the equator. In the control group (left column, from first to fourth rows), the vessel caliber and density appeared to reduce with increasing age. (Left column, first row, arrowhead) Retinal arcuate veins are projected on the en face image, which can be discerned by its course from the optic nerve head traveling through the inferior watershed zone void of choroidal vessels. Pathologically dilated outer choroidal vessels termed pachyvessels (center column) are much larger in caliber in comparison with their age-matched control. (Center column, first row, arrowheads) Pachyvessels are distributed in superior quadrants. (Center column, second row) Pachyvessels are diffusely distributed in all four quadrants. (Center column, third row, white arrowhead) Pachyvessels are displaced from a zone of gravitating atrophy following photodynamic therapy in the macula. Red arrowhead indicates the abrupt, club-shaped posterior termination of the pachyvessels. Normally, outer choroidal vessels gradually taper (left column, third row, red arrowhead) toward the posterior pole. (Right column, second row, arrowhead) The dark round shadow is a projection artifact from a hemorrhagic or serous pigment epithelial defect.

The macular region was scanned using a 12 raster radial lines scan (30° × 5°) 1 clock hour apart centered on the fovea, with 16 frames averaged in each B-scan. Each scan was 8.9 mm in length. Bruch's membrane and the choroid–scleral interface were delineated with the machine's in-built autosegmentation software, with manual fine adjustments when necessary, and the subfoveal choroidal thickness was automatically measured by built-in caliper. 

Choroidal Vascularity Quantitative Analysis

Raster scans through the fovea in B-scan OCT were binarized using the Niblack autolocal threshold tool.²⁵ Dark pixels represented lumen of blood vessels and light pixels represented the stromal tissue between blood vessels.²⁶ The area of dark pixels was called the lumen area (LA), and the area of light pixels was called the stromal area (SA). The total choroidal area (TCA) of interest was calculated by multiplying the standard width of 150 μm (75 μm on either side of the fovea) by the subfoveal choroidal thickness (between the Bruch's membrane and choroid–scleral interface; Fig. 2). The CVI was calculated as the ratio of LA over the TCA. 

Figure 2

 

Imaging processing for obtaining CVI. (Top) Original SSOCT image of a raster scan through the fovea. (Middle) After binarization, the choroidal vessels lumen is represented by dark pixels and the stromal tissue by light pixels. (Bottom) The CVI is calculated by dividing the LA by the TCA. The TCA is obtained by multiplying the subfoveal choroidal thickness (distance from the Bruch's membrane [red line] to the choroid–sclera interface (yellow line) by a width of 1.5 mm (0.75 mm on either side of fovea, bounded by the purple lines).

Statistical Analysis

The obtained data from each observer were statistically analyzed in relation to frequency and description. Inter- and intraobserver agreements were evaluated by the intraclass correlation coefficient (ICC) value. An ICC within 0.81 to 1.0 indicated good agreement and less than 0.4 indicated poor agreement. Best-corrected visual acuity was converted to logarithm of the minimal angle of resolution (LogMAR) units before the analysis. A χ² test or Fisher exact test was used for categorical analysis, and means were compared using a nonparametric Mann-Whitney U test. Univariate regression analysis was performed to evaluate the associations of age, sex, ocular parameters (BCVA, axial length, and refraction), fundus and autofluorescence photos findings (drusen and hemorrhage), FA (CNV), ICGA (polypoidal lesions, BVN, CVH), and OCT B-scan (pigment epithelial detachment [PED], CRT, SFCT, and CVI) with the presence and absence of pachyvessels. Multivariate regression analysis was performed to evaluate the association of age, sex, axial length, and the significant fundus photographic and angiographic features with the presence of pachyvessels. To determine whether pachyvessels were also present in the uninvolved fellow eyes, κ analysis was used. The agreement between subfoveal choroidal thickness and choroidal vascularity was evaluated by Pearson correlation analysis. P < 0.05 was statistically significant. All of the statistical analyses were performed with SPSS software version 18.0 (SPSS, Inc., Chicago, IL, USA). 

Results

During the study period, 79 consecutive patients with symptomatic, exudative maculopathy in at least one eye were recruited. Six patients were excluded from the analysis due to massive subretinal hemorrhage or media opacities that obscured the SSOCT acquisition. None of the patients had both eyes symptom onset at the same time during enrollment, and therefore 73 eyes of 73 patients were included in the final analysis. The mean ± SD age was 71.2 ± 8.9 years (range, 54–89 years). Forty-nine (67.1%) were male, and 22 (30.1%) had bilateral disease with a previous history of PCV or AMD in the fellow eye. Three (4.1%) eyes had extrafoveal lesions, and the remaining had subfoveal or juxtafoveal lesions. The demographic and clinical history data of the 73 patients are summarized in Table 1. The presence of pachyvessels was observed in 38 (52.1%) of 73 eyes. The ICC for observation of pachyvessels was 0.81. There were no statistically significant differences by χ² test in sex and other clinical history data, except for the higher number of patients in the pachyvessels group who received previous PDT (34/38 [89.5%] in pachyvessels group versus 5/35 [14.3%] in the no pachyvessels group, P = 0.03). Sixty-seven eyes (93.1%) had received previous anti-VEGF intravitreal injections, and the remaining five disease eyes were treatment naïve. 

Table 1

 

Demographic and Clinical History Data of 73 Patients With Exudative Maculopathy

 

Univariate regression analysis for the association of presence or absence of pachyvessels with age, sex, ocular factors, and multimodal imaging features are presented in Table 2. The ICC for grading of the fundus photographic and angiographic features was 0.73. The pachyvessels group was significantly younger in age (69.1 ± 9.4 vs. 73.7 ± 8.8 years, odds ratio [OR] = 0.95, 95% confidence interval [95% CI] = 0.90–0.97, P = 0.04). The presence of pachyvessels was also significantly correlated with presence of polypoidal lesions (64.3% vs 37.5%, OR = 3.27, 95% CI = 1.24–8.62, P = 0.01), increased SFCT (225.8 ± 93.8 vs. 157.3 ± 97.8 μm, OR = 1.08, 95% CI = 1.02–1.14, P ≤ 0.01), and increased CVI (65.4 ± 5.3 vs. 58.3 ± 15.5, OR = 1.12, 95% CI = 1.02–1.23, P = 0.01). In the multivariate regression model (Table 3), which included age, sex, axial length, presence of polypoidal lesions, SFCT, and CVI, only CVI was significantly associated with the presence of pachyvessels (OR = 1.24, 95% CI = 1.03–1.55, P = 0.04). Figures 3 to 5 display the representative images of patients' en face SSOCT scans with their corresponding ultra-widefield fundus color and autofluorescence photos, FA, and ICGA images. 

Table 2

 

Univariate Regression Analysis for the Association of Presence or Absence of Pachyvessels With Age, Sex, Ocular Factors, and Multimodal Imaging Features in 73 Eyes of 73 Patients With Exudative Maculopathy

 

Table 3

 

Multivariate Regression Analysis of Demographics, Ocular Parameters, and Imaging Features and Their Association With the Presence of Pachyvessels (n = 38)

 

Figure 3

 

(Top left) Ultra-widefield color fundus photo of the right eye of a 77-year-old man showed macular haemorrhage and exudates. (Top right) Fundus autofluorescence showed granular hypoatuofluorescence and mixed stippled hyper- and hypoautofluorescence. (Middle left) Staining and small amount of leakage on late phase fluorescein angiography. (Middle right) Polypoidal lesions in cluster on indocyanine green angiography. (Bottom left) Choroidal vascular hyperpermeability in late indocyanine green angiography. (Bottom right) Diffuse pachyvessels on en face SSOCT. The subfoveal choroidal thickness was 241 μm and the CVI was 67.2.

Figure 4

 

(Top left) Ultra-widefield fundus photo of the left eye of a 73-year-old man with macular hemorrhage and exudation. (Top right) Fundus autofluorescence showed granular hypoatuofluorescence and mixed stippled hyper- and hypoautofluorescence. (Middle left) Leakage of undetermined origin on fluorescein angiography late frame. (Middle right) A solitary polypoidal lesion on indocyanine green angiography. (Bottom left) No choroidal vascular hyperpermeability in late indocyanine green angiography. (Bottom right) There were no pachyvessels on en face SSOCT. The subfoveal choroidal thickness was 199 μm, and the CVI was 57.5.

Figure 5

 

(Top right) A 79-year-old woman with left eye polypoidal lesion and branching vascular network on indocyanine green angiography. (Top left) Right eye had no hyperfluorescce and no branching vascular network on indocyanaine green angiography. (Bottom left and right) The corresponding en face SSOCT saw diffuse pachyvessels in both eyes. Subfoveal choroidal thicknesses and CVIs of the disease (right) eye was 309 μm and 66.9 and for the asymptomatic fellow (left) eye was 312 μm and 66.3.

Correlation analysis was performed between 26 disease eyes with pachyvessels and their fellow uninvolved eyes (Table 4). The presence of pachyvessels, SFCT, and CVI found high agreement between study eyes and their asymptomatic fellow eyes (coefficient of κ = 0.67, Pearson's correlation coefficient = 0.73 and 0.85, respectively). 

Table 4

 

Correlation Analysis for the Presence of Pachyvessels, SFCT, and CVI Between 26 Disease Eyes and Their Uninvolved Fellow Eyes

 

Discussion

There is a myriad of overlapping clinical manifestations between neovascular AMD and PCV, particularly among Asians.² With the advent of SSOCT, near-histologic resolution of the choroidal structures can be visualized in vivo. Phenotypic distinction based on choroidal morphology has important implications for the understanding of histopathology of AMD and PCV, which may lead to significant consequences to treatment outcomes. Currently, there is no consensus on the definition of pachychoroid phenotype. In this study, we characterized in detail choroidal morphologic features that could specifically define the disease and systematically evaluated the possibilities to assess the differences in demographics, ocular parameters, fundus photographic, and angiographic information. 

We attempted to classify exudative maculopathy based on the presence or absence of pachyvessels on structural en face OCT scans. Dansingani et al. studied 33 patients who had pachychoroid features and noted that pachyvessels in en face OCT was the most consistently observed feature.¹³ Other studies that used en face OCT as a noninvasive method to examine patients with thick choroid also frequently detected pachyvessels.^13,18–20 However, a well-validated, normative database is not yet available for the morphology of choroidal vessels on en face images. We recruited healthy, age-matched volunteers and performed SSOCT to obtain en face images of the choroid for qualitative comparison with disease eyes in our study. Segmentation at the Haller's layer was confirmed when the characteristic pattern of outer choroidal vessels appeared. Several studies have previously studied the morphology of choroidal vasculature in normal subjects using en face high penetration OCT.^22–24 All of these studies failed to identify a distinct boundary between the inner and outer choroidal layers, but reported that the two layers can be differentiated from each other by observing the distribution and pattern of the vasculature. In the outer choroid, vessels are wider in diameter, straighter and fewer in number, and organized in a radial distribution outward from the center of the macula toward the equator. Vessels belonging to the inner choroid are thinner, more tortuous, more evenly distributed, and interwoven in the macular area. These characteristic differences between the inner and outer choroidal vessels do not seem to vary with age.²⁷ 

In our study, the presence of pachyvessels was correlated with increased SFCT, but its significance was lost in multivariate analysis when other confounders are included. In a cohort of 200 Japanese neovascular AMD patients, Miyaki et al. distinguished that 19.5% had pachychoroid neovasculopathy based on an arbitrary cutoff value for SFCT above 200 μm.²⁸ Nevertheless, SFCT has been found to be highly variable in the normal population, and the normative value has not yet been determined.^29–31 Choroidal thickness can be influenced by physiologic and ocular factors including age, sex, and axial length.^27,32 Although many eyes with pachychoroid will have quantifiably greater choroidal thickness measurements than age-matched controlled eyes, it is possible for an eye to be defined as pachychoroid to have normal choroidal thickness.^33–36 The discordance occurs when the increased luminal volume secondary to choroidal vessels dilation is offset by the reduction in tissue volume from the stroma. Therefore, pachychoroid is not simply a thick choroid; rather, it implicates the structural and functional alterations of the choroid leading to increased choroidal vascularity. This reflects the limitation of using choroidal thickness as a marker for choroidal vascularity as both vascular and stromal elements contribute toward thickness.³⁷ 

After controlling for age, sex, axial length, presence of polypoidal lesions, and SFCT in our multivariate analysis, it was shown that increased CVI was still a significant predictor for the presence of pachyvessels. For every 1 unit increase in CVI, the odds for presence of pachyvessels increase by a factor of 1.24 (P = 0.04). The CVI is the proportion of blood vessel lumen within a designated area of choroid. In a population-based study of 354 healthy eyes, physiologic and ocular factors have been shown to influence SFCT measurements, whereas CVI had less variability.³⁸ Thus, CVI was proposed to be a more robust surrogate marker to monitor diseases including AMD, CSC, uveitis, and diabetes mellitus.^26,39–41 Clinically, pachychoroid configuration could be characterized by a high CVI value, together with pachyvessels present on en face OCT image. 

In our study, eyes with pachyvessels were significantly younger than those without pachyvessels. Recently, Dansingani et al. reported that patients with pachychoroid neovasculopathy with or without polypoidal lesions were younger than patients with neovascular AMD.⁴² The pachychoroid spectrum diseases including CSC and PCV were known to have earlier onset than AMD, so it is plausible that pachyvessels could be observed more often in younger patients. Conversely, reduction in choroidal thickness, choroidal vessels density, and diameter has been reported in eyes with AMD.^43–45 Younger subjects also tend to have larger calibre and more densely spaced outer choroidal vessels than older subjects. Adhi et al. performed en face SSOCT in healthy eyes and also observed that reduction in size and density of choroidal vessels is associated with increasing age and is correlated with reduced choroidal thickness.⁴⁶ Histologic studies of the choroid have also suggested that reduction in size and density of choroidal vessels might occur with physiologic aging.⁴⁷ 

Our study found that polypoidal lesions in ICGA were associated with pachyvessels. Ferrara et al. performed en face SSOCT to study the choroidal vasculature in 15 eyes of CSC patients and reported the common finding of enlarged choroidal vessels at Sattler's and Haller's layers, which implied that choroidal venous stasis, choroidal hyperpermeability, and subsequent RPE dysfunction in the pathogenesis of CSC.^9,48,49 Although PCV and CSC are distinct phenotypically, many parallels have been found between these conditions to suggest that they may lie on the same pathogenic spectrum, including type I CNV, CVH, and increased choroidal thickness.⁵⁰ The abnormal histologic hallmark of PCV is abnormal dilatation of choroidal vessels with hyalinization, arteriosclerosis, massive exudation of fibrin and blood plasma, and loss of smooth muscle.^51–53 In addition, there is loss of choriocapillaris with disruption of the continuity of the RPE layer.⁵⁴ In contrast, neovascular AMD eyes consisted of smaller vascular channels and loss of choroidal vascular density underscoring the role of choroidal ischemia.^55,56 In vivo choroidal imaging using en face high-penetration OCT in PCV eyes of dilated choroidal vessels and thinning of choriocapillaris appeared to be consistent with the histopathologic findings.^9,13 

The detection rate of CVH in ICGA has been reported to be 9.8% to 34.7% among AMD and PCV patients, which may imply underlying pachychoroid configuration.^14,21,57,58 The CVH implies underlying pachychoroid configuration.¹⁴ However, we did not find significant association of CVH with other pachychoroid features in our study. Furthermore, we could not identify associations of CNV, BVN, and drusen with qualitative and quantitative configurations of the choroidal in our study. Often, a neovascular complex (type I CNV or BVN) may appear as a late-staining plaque that is often difficult to distinguish from CVH in ICGA.¹³ On the other hand, CVH may not be detectable when there is extensive atrophy of choriocapillaris in which leakage is minimal. 

Bilaterality has been reported at a wide range of rates between 10% and 50% in cross-sectional or retrospective studies of PCV patients.⁵⁹ Prospective studies have yet to confirm the incidence of bilateral symptomatic PCV. A number of observational studies in patients with pachychoroid disease spectrum detected similar morphologic changes bilaterally.^9,13,18 In our study, we observed a high level of agreement between the disease eyes with their uninvolved fellow eyes in terms of pachyvessels, SFCT, and CVI measurements. We hypothesize that pachyvessels may be a forme fruste change preceding sight-threatening serosanguinous maculopathy, and OCT angiography has detected quiescent type I CNV preceding leakage on FA in eyes with pachychoroid configuration.⁶⁰ En face OCT will also enable noninvasive and regular longitudinal examinations of the outer choroid morphologic changes in asymptomatic eyes to elucidate the pathogenesis and prognosis of the eyes with pachychoroid spectrum diseases. 

There are several limitations associated with our study. First, we excluded six eyes due to the presence of massive subretinal hemorrhage or media opacity and might have introduced selection bias. Second, imaging artifacts might occur in en face OCT especially in eyes with PED. An overlying large serous or hemorrhagic PED may cast a shadow obscuring the visualization of the choroidal vessels immediately underneath the lesion. (Fig. 1, third row, middle column) Nevertheless, features of pachyvessels can still be observed adjacent to the PED shadow on the choroidal layer. Third, retinal vessels may project on the en face choroidal image. Hence, comparison must be made with en face image of the retinal layer to not mix up the projected retinal vessels from choroidal vessels. Fourth, we only measured the subfoveal choroidal thickness and CVI without topographic correlation with en face images. Currently, there are no normative data on the morphologic appearance of choroidal vasculature; therefore, we used standard en face OCT images from control patients from different ages for comparison. 

The majority of our patients (93.1%) had received anti-VEGF treatments with or without combination PDT. Subfoveal choroidal thickness and vascularity have been demonstrated to decrease following anti-VEGF therapy and PDT.^61,62 The mean CVI for the pachyvessels group in our study was 65.4, which was quite close to the mean CVI of 65.6 reported by Agrawal et al. in a population-based study.³⁸ Nonetheless, eyes without pachyvessels in our study had significantly lower CVI (58.3), even though fewer patients in this group had previous PDT (14.3% vs. 89.5% in the pachyvessels group). Wei et al. recently reported reduced CVI (60.1) in neovascular AMD and PCV eyes.³⁹ In addition to CVI in the subfoveal region, our study examined en face OCT images that allowed a panoramic view of the extramacular regions where pachychoroid morphology could still be apparent. A representative en face OCT image in Figure 1 (middle column, third row) demonstrates atrophic changes in a region with previous PDT, in which the pachyvessels appeared to be displaced peripherally. Previous treatments did not appear to influence the qualitative analysis for pachyvessels because the effect of PDT on the choroidal morphologic changes seemed to be localized. 

In summary, exudative maculopathy can be clinically distinguished by choroidal morphologic features. En face OCT allows panoramic, topographic, and three-dimensional analysis to facilitate the characterization of the morphology and extent of choroidal changes. In clinical practice, point-to-point correlation of the en face OCT findings with cross-sectional OCT can supplement other imaging modalities. En face OCT has several advantages over ICGA as leakage of dye will mask the pachyvessels and ICGA lacks depth resolution. The presence of pachyvessels is significantly correlated with a quantitative increase in SFCT and CVI measurements. The pachychoroid configuration is associated with younger disease onset and polypoidal lesions detected on ICGA. Investigation of choroidal morphology has significant implication on our understanding of the histopathology of AMD and PCV, which could ultimately lead to individualized treatment with optimal vision outcomes. 

Acknowledgments

The authors thank Antony K. Law, MSc in Epidemiology and Biostatistics, Research Assistant, Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, for contribution in revising and editing the statistical analysis and presentation of the results of this study. 

Supported, in part, by The Chinese University of Hong Kong Direct Grant 2015/2016, reference 2015.1.044. 

Disclosure: D.S.-C. Ng, None; M. Bakthavatsalam, None; F.H.-P. Lai, None; C.Y.-L. Cheung, None; G.C.-M. Cheung, None; F.Y. Tang, None; C.W. Tsang, None; T.Y.-Y. Lai, None; T.Y. Wong, None; M.E. Brelén, None