Methods

A retrospective review was performed on 57 eyes of 47 patients with the clinical diagnosis of either wet or dry AMD seen between November 2009 and January 2010 at the New England Eye Center, Tufts Medical Center, Boston, Massachusetts. All patients included in the study underwent a comprehensive ophthalmologic examination with fundus biomicroscopy, color fundus photography, best-corrected Snellen visual acuity, fluorescein angiography, and OCT. OCT imaging was performed using Cirrus-HD OCT software version 4.5. The software version allows for the acquisition of high-definition 1-line raster scans that are constructed from 20 B-scans obtained at the same location and processed using a unique Selective Pixel Profiling system (Cirrus HD-OCT; Carl Zeiss Meditec). The 1-line raster is a 6-mm line consisting of 4096 A-scans, with an axial resolution of approximately 5 to 6 μm and a transverse resolution of approximately 15 to 20 μm. These high-definition images provide increased definition of retinal layers, as well as more posterior structures, such as the choroid–sclera junction. The images were not inverted to bring the choroid in closer proximity to the 0 delay line, as has been described previously as a tool to visualize better the choroid. Nevertheless, in most cases, good-quality images were obtainable, allowing choroidal thickness measurements to be performed. All images included in this study were at least 6 of 10 in intensity, taken as close to the fovea as possible, and able to demonstrate clear visualization of the choroid–sclera boundary.

Choroidal thickness measurements were obtained manually using the Cirrus linear measurement tool at 500-μm intervals, 2500 μm temporal and nasal to the fovea, perpendicular from the base of the hyperreflective retinal pigment epithelium to the choroid–sclera junction. In 9 eyes with choroidal neovascularization, measurements were performed from the edge of the Bruch membrane to the top of the sclera ( Figure 1 ). The 11 measurements were obtained from all 57 eyes by 2 independent observers (V.M., J.G.) to correlate the thickness measurements. Chart review was performed to collect information regarding duration of disease, treatments, number of intravitreal anti-VEGF injections, visual acuity (VA), lens status, and concomitant retinal pathologic features. Eyes with more than −6 diopters of refractive error were excluded because of known choroidal thinning associated with high myopia.

Choroidal thickness measurement in age-related macular degeneration. Cirrus HD optical coherence tomography (OCT) 1-line raster scan demonstrating measurement technique used to obtain choroidal thickness across the macula. This OCT scan also demonstrates the measurement from the edge of the Bruch membrane to the choroid–sclera junction in eyes with wet age-related macular degeneration and pigment epithelial detachments.

All statistics were calculated using SPSS software version 17.0 for Windows (SPSS, Inc, Chicago, Illinois, USA). Pearson correlation and the Student t test were used to correlate interobserver measurements and to correlate choroidal thickness values with various disease factors. A P value of < .05 was considered statistically significant in this study.

Results

The study group included 23 males (40%) and 34 females (60%). Forty eyes (70%) had the diagnosis of wet AMD with a mean age of 78.6 years (standard deviation [SD], 7.0 years). Seventeen eyes (30%) had dry AMD with a mean age of 78.2 years (SD, 7.9 years). AMD is defined as a chronic, progressive degenerative disorder of the macula that affects older individuals and features central visual loss as a result of drusen deposition, geographic atrophy, serous detachment of the retinal pigment epithelium, and neovascularization. Patients were subclassified as having dry AMD if there was no evidence of neovascularization such as intraretinal or subretinal fluid at any visit time point. This group included patients with geographic atrophy, drusen, and pigment epithelial detachments. Patients were subclassified as having wet AMD if there was evidence of neovascularization such as intraretinal or subretinal fluid at any visit time point. Patients were treated by different attending physicians with a combination of treat-and-extend and as-needed regimens with standard anti-VEGF therapy. Of the 40 eyes with wet AMD, 7 of these patients had undergone prior photodynamic therapy at least once in the study eye. The average duration of disease was 3.8 years. Subfoveal choroidal thickness measurements were used for comparative purposes in this study. The pattern of choroidal thickness in the macula in the 57 eyes with AMD in this study demonstrated the thickest choroid located subfoveally, with nasal and temporal thinning, maintaining a similar pattern seen in normal eyes ( Figure 2 ).

Graph showing mean macular choroidal thickness in eyes with age-related macular degeneration. Mean thickness at each of the 11 locations measured at 500-μm (0.5-mm) intervals temporal (T) and nasal (N) to the fovea (F) in all 57 eyes. The graph illustrates variation in choroidal thickness across the macula with the choroid thinnest nasally, thickening subfoveally, and thinning again temporally.

The authors’ recent paper demonstrated a mean subfoveal choroidal thickness of 272 ± 81 μm in normal eyes using the Cirrus HD-OCT 1-line raster scan in patients with an average age of 51.1 years. In a study utilizing the Heidelberg Spectralis, normal subfoveal choroidal thickness of 287 ± 76 μm in patients with a mean age of 50.4 years was reported. The mean age of the 57 eyes in this study was 78.4 years (SD, 7.3 years); therefore, an age-adjusted normal subfoveal choroidal thickness of 229 μm was used for comparison because of an approximate 1.56-μm decline in choroidal thickness per year of life. A strong interobserver correlation of choroidal thickness measurements in all 57 eyes was found at each of the 11 locations measured along the horizontal scans of the macula. The correlation values of interobserver measurements ranged from 0.909 to 0.981, with P values of < .01 at each of the 11 measured points. The subfoveal choroidal measurements performed on the Cirrus HD-OCT 1-line raster images by the 2 separate observers (V.M., J.G.) were averaged.

The results of the choroidal thickness measurements in the 57 eyes with wet and dry AMD demonstrated considerable variation in thickness. The measurements were plotted according to SDs from the mean age-adjusted normal subfoveal choroidal thickness ( Figure 3 ). Twenty-seven eyes (47.3%) fell outside the range of 1 SD from the mean and 2 eyes (3.5%) had significantly thickened choroids, measuring more than 2 SDs above the mean. Of the 2 eyes more than 2 SDs above the mean, 1 eye was categorized as having dry AMD and 1 was categorized as having wet AMD ( Figure 4 ).

Bar graph showing the distribution of subfoveal choroidal thickness in eyes with wet and dry age-related macular degeneration. The distribution of mean subfoveal choroidal thickness measurements in the studied eyes with wet and dry age-related macular degeneration is compared with a normal age-adjusted mean of 229 μm, with a standard deviation of 81 μm, based on recently published data. Dark gray column = wet age-related macular degeneration; light gray column = dry age-related macular degeneration.

Choroidal thickness measured to be >2 standard deviation above the mean in age-related macular degeneration. (Top left) Color fundus photograph and (Top right) high-definition Cirrus HD-OCT 1-line raster scan demonstrating thickened choroid in an 84-year-old woman with a history of wet age-related macular degeneration. (Bottom left) Color fundus photograph and (Bottom right) high-definition Cirrus HD-OCT 1-line raster scan demonstrating increased choroidal thickness in a 71-year-old woman with a history of dry age-related macular degeneration.

Mean subfoveal choroidal thickness was found to be 194.6 μm (SD, 88.4; n = 40) in the wet AMD group and 213.4 μm (SD, 92.2; n = 17) in the dry AMD group. A weak correlation ( r = −0.367; P = .005) was found between subfoveal choroidal thickness and age when all 57 eyes were considered together. However, when analyzed separately, eyes with wet AMD showed a weak correlation with age ( r = −0.203; P = .208) that was not statistically significant, whereas eyes with dry AMD demonstrated a statistically significant correlation ( r = −0.703; P = .002). This relationship in dry eyes was found to be stronger than the correlation between subfoveal choroidal thickness and age found in normal eyes ( r = −0.61; P = .0001).

Analysis of the relationship between the number of intravitreal anti-VEFG injections (mean number of injections, 5.4; SD, 4.6) and subfoveal choroidal thickness in the eyes with wet AMD demonstrated no correlation ( r = −0.037; P = .821). Correlation of disease duration defined as the number of years since initial diagnosis of macular degeneration (mean number of years of disease, 3.8; SD, 2.7 years) also demonstrated no relationship ( r = −0.046; P = .737) with subfoveal choroidal thickness. In subgroup analysis, there was no significant correlation found between duration of disease and choroidal thickness within the wet AMD group ( r = −0.066; P = .684) or the dry AMD group ( r = 0.134; P = .608) with relation to disease duration and choroidal thickness.

Analysis also was performed to examine the relationship between VA and subfoveal choroidal thickness. All Snellen visual acuities were converted to logarithm of the minimal angle of resolution (logMAR) for statistical analysis. Overall mean VA was 20/65 (logMAR units, 0.52). Within the subgroups, the wet eyes had a mean VA of 20/71 (logMAR units, 0.55), whereas the dry eyes had a mean VA of 20/54 (logMAR units, 0.43). No correlation was found overall ( r = −0.009; P = .948) between choroidal thickness and VA, nor within the wet AMD eyes ( r = −0.126; P = .446) and dry AMD eyes ( r = 0.245; P = .327) when analyzed separately.