Subjects

For this prospective cross-sectional study, we recruited patients with reticular pseudodrusen who visited the Macular Service at Kyoto University Hospital, Kyoto, Japan, between July 1, 2010 and February 14, 2013, as well as age-matched normal controls. To be included in the current study, the patients had to be 50 years of age or above.

All of the patients had already been diagnosed with reticular pseudodrusen based on the appearance of reticular patterns according to at least 2 of the following imaging modalities: color fundus photography (blue channel images), infrared reflectance (IR) imaging, fundus autofluorescence (FAF), spectral-domain OCT (SD-OCT), and IA. Eyes with high myopia (axial length ≥26.0 mm) and eyes with other macular abnormalities (eg, idiopathic choroidal neovascularization, angioid streaks, other secondary choroidal neovascularization, central serous chorioretinopathy, epiretinal membranes, or retinal arterial macroaneurysms); any history or signs of retinal surgery, including laser treatment and photodynamic therapy; or evidence of glaucoma or high intraocular pressure (≥22 mm Hg) were also excluded from this study. Subjects with systemic diseases or conditions that might affect the choroidal thickness were also excluded, such as those with diabetes mellitus or malignant hypertension. Eyes whose fellow eyes had any sign of AMD including drusen were excluded from normal controls.

Eyes with reticular pseudodrusen were classified into 3 subgroups: eyes without late AMD (neovascular AMD and geographic atrophy) (Group 1), eyes with neovascular AMD (Group 2), and eyes with geographic atrophy (Group 3). Neovascular AMD was characterized by choroidal neovascularization (CNV) detected by fluorescein angiography (FA) and IA. Geographic atrophy was detected, using color fundus photography as a sharply delineated area (at least 175 μm in diameter) of hypopigmentation, depigmentation, or apparent absence of the RPE due to which the choroidal vessels were clearly visible.

All subjects underwent a comprehensive ocular examination, including measurement of best-corrected visual acuity (BCVA) with a 5-m Landolt chart, intraocular pressure, autorefractometry (ARK1; Nidek, Gamagori, Japan), axial length measurement using ocular biometry (IOL Master; Carl Zeiss Meditec, Jena, Germany), slit-lamp examination, color fundus photography (TRC-NW8F; Topcon Corp, Tokyo, Japan), and prototype SS-OCT. All patients with macular complications underwent simultaneous FA and IA using HRA+OCT (Spectralis; Heidelberg Engineering, Heidelberg, Germany).

Swept-Source Optical Coherence Tomography System and Scanning Protocols

We used an SS-OCT prototype system (Topcon, Tokyo, Japan) with an axial scan rate of 100 000 Hz operated in the 1-μm wavelength region. The details of this system were described elsewhere.

SS-OCT examinations were performed by trained examiners after pupil dilation. First, horizontal and vertical line scans (12 mm) through the fovea were obtained, and ∼50 B-scan images were averaged to reduce speckle noise. Second, a 3D imaging data set was acquired for each subject with a raster scan protocol of 512 (horizontal) × 128 (vertical) A-scans per data set (total: 65 536 axial scans/volume) in 0.8 s. Each 3D scan covered an area of 6 × 6 mm centered on the fovea, which was confirmed by an internal-fixation and fundus camera integrated in the instrument. To reduce speckle noise, each image was averaged by applying the weighted moving average from 3 consecutive images. Owing to the invisible scanning light and high-speed scanning, eye movement during the 3D scan was minimal.

Choroidal Thickness and Volume Measurement Protocol

The choroidal thickness was measured as the distance between the Bruch membrane (or the outer border of the RPE–Bruch membrane complex) and the chorioscleral interface ( Figure 1 ). In each image of the 3D data set, both lines were determined manually by a trained ophthalmologist who was masked to the diagnosis, classification, or angiographic findings. Automated built-in calibration software was used to determine the distance between the 2 lines. From all 128 images of each 3D data set, a choroidal thickness map of 6 × 6 mm was created. False color was determined starting from a cool color and progressing to a warm color, at the range of 0–500 μm. After the choroidal thickness map was created, the ETDRS sectors were applied to it ( Figure 1 ). First, the central fovea on each image was manually registered, if necessary, to coincide with the central circle on the ETDRS segmentation diagram. The mean thickness of each sector was measured in the “center” sector within 0.5 mm from the center of the fovea, in 4 “inner ring” sectors (nasal, superior, temporal, and inferior) 0.5–1.5 mm from the center of the fovea, in 4 “outer ring” sectors (nasal, superior, temporal, and inferior) 1.5–3 mm from the center of the fovea, and as a “whole” within 3 mm from the center of the fovea.

A choroidal thickness map obtained by high-penetration swept-source optical coherence tomography (SS-OCT) in a normal eye of a 74-year-old man. (Top left) Three-dimensional raster scan covering a 6 × 6 mm area centered on the fovea. The Early Treatment Diabetic Retinopathy Study (ETDRS) sectors were applied to the scanned area, and the mean choroidal thickness of 9 sectors was measured. (Top right) The ETDRS sectors consist of a “center” sector within 0.5 mm from the center of the fovea, 4 “inner ring” sectors (nasal, superior, temporal, and inferior) 0.5–1.5 mm from the center of the fovea, and 4 “outer ring” sectors (nasal, superior, temporal, and inferior) 1.5–3.0 mm from the center of the fovea. (Bottom left) B-scan images at different levels of the raster scan. (Bottom right) Green segmentation lines show Bruch membrane and the chorioscleral border.

In B-scans where it was difficult to identify the whole outer choroid, 7-15 points where the chorioscleral interface could be identified were chosen and connected to create a segmentation line. After creating segmentation lines in all 128 B-scans, the choroidal thickness map was checked. If the thickness differed remarkably between adjoining B-scans, the segmentation lines of the B scans were reexamined and corrected as required.

To examine whether the localization of reticular pseudodrusen affected the focal choroidal thickness, area of reticular pseudodrusen was assessed for Group 1 eyes using IR images. With respect to each ETDRS sector, the eyes were divided into 2 groups according to whether reticular pseudodrusen covered more than half of the area or not. The mean choroidal thickness of each ETDRS sector was compared between the 2 groups. The area of reticular pseudodrusen was assessed using the built-in software of HRA-2 (Heidelberg Engineering, Heidelberg, Germany). The data obtained were analyzed to determine the association between the areas covered by reticular pseudodrusen and mean macular choroidal thickness.

Measurement of Choroidal Vascular Area Using En Face Images

En face images were automatically reconstructed from a 3D data set using software developed by Topcon Corporation, which flattened the images at the level of Bruch membrane. Each en face image was created by averaging the consecutive images of the range of 7 μm. To analyze the area of choroidal vasculature, the en face image was adopted at the intermediate level between the Bruch membrane and the deepest site of the chorioscleral interface on the horizontal B-scan image through the fovea. To measure the area of choroidal vasculature of each adopted image, Image J software (National Institutes of Health, Bethesda, Maryland, USA) was used. In Image J, the command path of Image > Adjust > Threshold > Auto was used to differentiate the vascular area from the nonvascular area. To measure the vascular area, the command path of Analyze > Measure was used. If the adopted image contained the chorioscleral border, the image was trimmed minimally so that it contained the choroidal area only. Then, the ratio of choroidal vascular area to total area was calculated.

Statistical Analysis

Statistical analysis was performed using SPSS statistical software (version 20; SPSS Inc, Chicago, Illinois, USA). All values are presented as mean ± standard deviation (SD). We used the unpaired t test or Welch’s t test to compare the data between 2 groups with normal distribution. The Mann-Whitney U test was used to compare data between 2 groups in which normal distributions were not verified. To compare the data between 3 groups, 1-way analysis of variance and the Tukey-Kramer test were used. P values less than .05 were considered to be statistically significant.