Abstract
Purpose.:
To evaluate the prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots in eyes with choroidal neovascularization (CNV).
Methods.:
This retrospective, observational study included 382 eyes with typical exudative AMD (97 eyes), polypoidal choroidal vasculopathy (PCV, 163 eyes), retinal angiomatous proliferation (RAP, 37 eyes), or myopic CNV (86 eyes). The prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots was estimated based on available indocyanine green angiography (ICGA) images.
Results.:
Choroidal vascular hyperpermeability was noted in 12.4% (12 eyes) and 26.9% (42 eyes) of eyes with typical exudative AMD and PCV, respectively. Choroidal vascular hyperpermeability was not noted in any eye with RAP or myopic CNV. Punctate hyperfluorescent spots were noted in 43.3% (42 eyes), 72.4% (118 eyes), 10.8% (4 eyes), and 4.7% (4 eyes) of eyes with typical exudative AMD, PCV, RAP, and myopic CNV, respectively. Of the 56 eyes with choroidal vascular hyperpermeability, punctate hyperfluorescent spots were noted in 55 eyes (98.2%).
Conclusions.:
Choroidal vascular hyperpermeability and punctate hyperfluorescent spots may have a common pathophysiology. Although choroidal vascular hyperpermeability and punctate hyperfluorescent spots have been thought to be associated with pathologic conditions, the markedly low prevalence of these findings in eyes with RAP and myopic CNV may not be a normal finding. It is possible that compromised choroidal perfusion, with or without associated with choroidal thinning, may lead the low prevalence of these abnormalities in eyes with these two disorders.
Choroidal vascular hyperpermeability and punctate hyperfluorescent spots are peculiar findings on indocyanine green angiography (ICGA) and are thought to be associated with hyperpermeable choroid.
1,2 Choroidal vascular hyperpermeability and punctate hyperfluorescent spots were originally observed in eyes with central serous chorioretinopathy (CSC).
2–7 However, these findings have more recently been found in eyes with choroidal neovascularization (CNV) secondary to AMD.
8–11 In general, these findings were more frequently noted in eyes with polypoidal choroidal vasculopathy (PCV) than in eyes with typical exudative AMD.
8 Eyes exhibiting these findings were found to have a relatively thick choroid.
9,12,13 In addition, the presence of choroidal vascular hyperpermeability has been associated with a poor response to anti-VEGF therapy in eyes with PCV.
9,12
To date, most published studies have examined choroidal vascular hyperpermeability and punctate hyperfluorescent spots in eyes with CSC or PCV, only limited knowledge is available on the prevalence and characteristics of these ICGA findings in other types of CNV.
In the present study, we investigated the prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots in subtypes of CNV, including typical exudative AMD, PCV, retinal angiomatous proliferation (RAP), and myopic CNV. In particular, we examined the prevalence of these two unusual ICGA findings in eyes with RAP and myopic CNV.
This retrospective, observational case series was performed on medical records from a single center. This study was approved by the institutional review board (Kim's Eye Hospital IRB) and all study conduct adhered to the tenets of the Declaration of Helsinki.
We conducted a computerized search for patients newly diagnosed with exudative AMD or myopic CNV between January 2013 and March 2014 at Kim's Eye Hospital (Seoul, South Korea). Eyes were included in analyses if they were treatment naïve and had undergone ICGA. Each subject had undergone a comprehensive ophthalmologic examination. This included measurements of best-corrected visual acuity (BCVA), 90-diopter (D) lens slit-lamp biomicroscopy, fundus photography, and spectral-domain optical coherence tomography (OCT; Spectral OCT/SLO; OTI Ophthalmic Technologies, Inc., Toronto, Canada). Fluorescein angiography and ICGA were also performed with a confocal laser-scanning system (HRA-2; Heidelberg Engineering, Dossenheim, Germany). Enhanced-depth imaging OCT
14 images were obtained using the Spectralis OCT (Heidelberg Engineering) at the physician's discretion. Eyes were excluded from analyses if they had severe media opacity, prior intraocular surgery (except cataract surgery), end stage disease in which the accurate classification of subtypes of exudative AMD was not possible, or retinal vascular disorder signs (e.g., macroaneurysm, proliferative diabetic retinopathy, and central retinal vascular occlusion). Eyes with subretinal hemorrhages greater than 5 disc areas in size were also excluded because large hemorrhages may interfere with ICGA interpretation. If both eyes were diagnosed with CNV, the eye that showed the symptoms first was included in the study.
Two independent examiners (JHK, YSC) interpreted ICGA results and used this information to classify CNV as typical exudative AMD, PCV, RAP, or myopic CNV. Eyes were classified as having myopic CNV if they had a refractive spherical equivalent more severe than −6.0 D or an axial length greater than 26.0 mm. Eyes with branching vascular networks and/or terminating polypoidal lesions were classified as having PCV.
15 Eyes with evidence of retinal–retinal or retinal–choroidal anastomoses were classified as having RAP.
16 All other eyes with exudative AMD were classified as having typical exudative AMD. Eyes were placed into an unclassified group if ICGA images were of poor quality or not available.
The presence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots was independently evaluated by two examiners (JHK, YSC). Both were masked to diagnosis and OCT result. Choroidal vascular hyperpermeability was defined as multifocal hyperfluorescent areas present in late-phase ICGA images (
Fig. 1A). A punctate hyperfluorescent spot was defined as a focal hyperfluorescent spot seen in mid-to-late phase ICGA images (
Fig. 1B).
1 In a previous study, punctate hyperfluorescent spots were classified as either solitary or clustered.
17 We modified the previously suggested classification criteria and classified spots as clustered (≥5 punctate hyperfluorescent spots clustered in one area) or as solitary (<5 punctate clustered hyperfluorescent spots). When clustered spots were observed, the eye was included in the clustered pattern group, even if solitary spots were also present. Eyes with solitary hyperfluorescent spots, but not clustered spots, were included in the solitary pattern group. Hard drusen are often observed as hyperfluorescent lesions in ICGA images.
18 To differentiate hyperfluorescence from hard drusen and from punctate hyperfluorescent spots, additional analysis was performed using fundus photography and infrared images. Eyes exhibiting hyperfluorescent spots only at the location corresponding to that of the hard drusen on fundus photographs or infrared images were excluded from the study. Classification disagreements between the two examiners were settled by discussion between the examiners.
The prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots was estimated and compared among the four CNV subtype groups. The difference in punctate hyperfluorescent spot subtype (i.e., solitary versus clustered) incidence was also compared among the four CNV subtype groups. In addition, subfoveal choroidal thickness was compared between eyes in the solitary and clustered pattern groups for patients who had undergone enhanced-depth imaging OCT. The difference in punctate hyperfluorescent spot prevalence was examined in eyes with and without choroidal vascular hyperpermeability. This was also done for punctate hyperfluorescent spot subtypes.
To evaluate changes in choroidal vascular hyperpermeability and punctate hyperfluorescent spots following anti-VEGF therapy, additional analyses were performed on eyes diagnosed with exudative AMD. In eyes that had undergone ICGA 3 months following anti-VEGF therapy initiation, changes in ICGA findings were analyzed by the two examiners (JHK, YSC). Resolution of old abnormalities and development of new findings were confirmed by examiner consensus.
Statistical analyses were performed with a commercially available software package (SPSS ver. 12.0 for Windows; SPSS, Inc., Chicago, IL, USA). Differences in incidence between groups were tested for statistical significance using either a χ2 test or a Fisher's exact test. Differences in punctate hyperfluorescent spot prevalence in eyes with and without choroidal vascular hyperpermeability were tested for statistical significance using Fisher's exact tests. Other comparisons were performed using χ2 tests. Statistical significance was defined as P less than 0.05.
A total of 382 eyes of 382 patients met all eligibility criteria. Ultimately, 97, 163, 37, and 85 eyes were diagnosed with typical exudative AMD, PCV, RAP, and myopic CNV, respectively. Among the 382 eyes, choroidal vascular hyperpermeability and punctate hyperfluorescent spots were noted in 56 (14.7%) and 168 eyes (43.9%), respectively. Among the 168 eyes with punctate hyperfluorescent spots, 89 (52.9%) and 79 eyes (47.0%) were classified as having solitary and clustered spots, respectively. Fifty-five of 56 eyes (98.2%) with choroidal vascular hyperpermeability also had punctate hyperfluorescent spots. Initial agreement between the two examiners for evaluating the presence of punctate hyperfluorescence spots was achieved in 354 of 382 eyes (92.7%). The presence of these spots in the remaining 28 eyes was determined by discussion between the two examiners. Initial agreement for evaluating the presence of choroidal vascular hyperpermeability was achieved in 329 of 382 eyes (86.1%).
Table 1 summarizes the prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots in eyes with each CNV subtype. Among eyes with typical exudative AMD and PCV, choroidal vascular hyperpermeability was noted in 12 eyes (12.4%) and 44 eyes (26.9%), respectively. Choroidal vascular hyperpermeability was not observed in any eye with RAP or myopic CNV. The prevalence of choroidal vascular hyperpermeability was significantly different among the four CNV subtype groups (
P < 0.001). Punctate hyperfluorescent spots were noted in 42 eyes (43.3%) with typical exudative AMD, 118 eyes (72.4%) with PCV, 4 eyes (10.8%) with RAP, and 4 eyes (4.7%) with myopic CNV. The prevalence was significantly different among the four CNV subtype groups (
P < 0.001). In the 168 eyes with punctate hyperfluorescent spots, a clustered pattern was noted in 12 eyes (28.6%) with typical exudative AMD, 66 eyes (55.9%) with PCV, and 1 eye (25.0%) with RAP. Clustered spots were not observed in any eye with myopic CNV. Differences among the four CNV subtype groups in spot subtype prevalence (solitary versus clustered) were statistically significant (
Table 2,
P = 0.001).
A total of 55 eyes had both punctate hyperfluorescent spots and choroidal vascular hyperpermeability. Of these, solitary and clustered pattern punctate hyperfluorescent spots were observed in 15 (27.3%) and 40 eyes (72.7%), respectively. In the other 113 eyes with punctate hyperfluorescent spots, but no choroidal vascular hyperpermeability, solitary, and clustered punctate hyperfluorescent spots were observed in 74 (65.5%) and 39 eyes (34.5%), respectively. The prevalence of clustered-pattern punctate hyperfluorescent spots was higher in eyes with choroidal vascular hyperpermeability than in eyes without it (
Table 3,
P < 0.001).
Table 4 shows the difference in punctate hyperfluorescent spot incidence between eyes with and without choroidal vascular hyperpermeability in typical exudative AMD and PCV. In eyes with typical exudative AMD, all 12 eyes (100%) with choroidal vascular hyperpermeability also had punctate hyperfluorescent spots. However, only 30 of 85 eyes (35.3%) without choroidal vascular hyperpermeability had punctate hyperfluorescent spots. This large difference was statistically significant (
P < 0.001). In eyes with PCV, punctate hyperfluorescent spots were observed in 43 of 44 eyes (97.7%) with choroidal vascular hyperpermeability and in 75 of 119 eyes (63.0%) without choroidal vascular hyperpermeability. This difference was statistically significant (
P < 0.001).
A total of 56 eyes had choroidal vascular hyperpermeability, of which 49 eyes (87.5%) had enhanced-depth imaging OCT images available. Mean subfoveal choroidal thickness was 389.5 ± 121.1 μm in these 49 eyes. Subfoveal choroidal thickness was less than 200 μm in four eyes (one eye with typical exudative AMD, three eyes with PCV), between 200 and 400 μm in 21 eyes (4 eyes with typical exudative AMD, 17 eyes with PCVs), and greater than 400 μm in 24 eyes (4 eyes with typical exudative AMD, 20 eyes with PCVs,
Fig. 2A). Of the 168 eyes with punctate hyperfluorescent spots, enhanced-depth imaging OCT images were available in 154 eyes (91.7%). Mean subfoveal choroidal thickness was 318.6 ± 127.4 μm in these 154 eyes. Subfoveal choroidal thickness was less than 200 μm in 30 eyes (10 eyes with typical exudative AMD, 16 eyes with PCV, 3 eyes with RAP, 1 eye with myopic CNV), between 200 and 400 μm in 79 eyes (16 eyes with typical exudative AMD, 59 eyes with PCV, 1 eye with RAP, 3 eyes with myopic CNV), and greater than 400 μm in 45 eyes (9 eyes with typical exudative AMD, 36 eyes with PCV,
Fig. 2B). The difference in subfoveal choroidal thickness between eyes in the solitary pattern group (314.5 ± 127.0 μm,
n = 80) and eyes in the clustered pattern group (322.9 ± 128.5 μm,
n = 74) was not significant (
P = 0.682). Among the 326 eyes without choroidal vascular hyperpermeability, 284 eyes (87.1%) had enhanced-depth imaging OCT images available. The mean subfoveal choroidal thickness in these 284 eyes was 215.3 ± 123.9 μm. Among the 214 eyes without punctate hyperfluorescent spot, 179 eyes (83.6%) had enhanced-depth imaging OCT images available. The mean subfoveal choroidal thickness in these 179 eyes was 174.2 ± 108.9 μm.
Seventy-eight of the 297 eyes (26.3%) with exudative AMD had available ICGA images obtained 3 months after initiating intravitreal ranibizumab therapy (Lucentis; Genentech, Inc., South San Francisco, CA, USA). All 78 eyes had received three injections, 1 month apart, during the 3-month period. Neither new development nor resolution of choroidal vascular hyperpermeability or punctate hyperfluorescent spots was observed in any of these eyes.
In the present study, we first evaluated the prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots in RAP and myopic CNV. In addition, the punctate hyperfluorescent spots were first classified into two groups and the prevalence of each subtype of spots were estimated separately. As a result, we did not observe choroidal vascular hyperpermeability in any eye with RAP or myopic CNV. In addition, only a few eyes exhibited punctate hyperfluorescent spots. The prevalence of the two ICGA findings in RAP and myopic CNV was markedly lower than in typical exudative AMD or PCV. Moreover, choroidal vascular hyperpermeability was more frequently noted in eyes with cluster pattern punctate hyperfluorescent spots than in eyes with solitary pattern spots. In addition, cluster pattern spots were more frequently noted in PCV than in typical exudative AMD.
It is generally accepted that choroidal vascular hyperpermeability occurs more frequently in eyes with PCV than in eyes with typical exudative AMD.
8,9,12 However, there was a wide variation in the reported prevalence of choroidal vascular hyperpermeability in eyes with either PCV (9.8%–50.0%)
8–10,12,13,19 or typical exudative AMD (1.9%–37.5%).
8,10 This variation may be the result of diagnosing choroidal vascular hyperpermeability based on qualitative image analysis and patient characteristics. Relatively little information is available on the prevalence of punctate hyperfluorescent spots in eyes with exudative AMD. One recent study reported that punctate hyperfluorescent spot prevalence was 40.0% and 86.0% in eyes with typical exudative AMD and PCV, respectively.
17 Our observed prevalence of vascular hyperpermeability and punctate hyperfluorescent spot was relatively comparable to the previously reported values.
In the present study, choroidal vascular hyperpermeability was not noted in any eye with either RAP or myopic CNV. The presence of choroidal vascular hyperpermeability or punctate hyperfluorescent spots is generally thought to be associated with pathologic conditions. In eyes with active CSC, it has been reported that choroidal vascular hyperpermeability persists, even after fluorescein leakage has resolved.
2 Recurrence of leakage was also noted in areas of choroidal hyperfluorescence.
2 In addition, choroidal vascular hyperpermeability was found to be associated with vortex vein congestion.
20 The presence of punctate hyperfluorescent spots in eyes with CSC may be associated with CSC development.
1 Additionally, punctate hyperfluorescent spots have been shown to be present in 40% of eyes with typical exudative AMD.
17 Therefore, some investigators believe that neovascular AMD includes a variety of choroidal pathologies that have a wide-presentation spectrum.
17 However, we suspect that the presence of choroidal vascular hyperpermeability or punctate hyperfluorescent spots do not always indicate pathologic status, even though the two conditions are almost certainly associated with some retinal disorders. For example, a thick choroid is considered to be associated with CSC
21 and PCV.
6 However, a thin choroid is also associated with AMD,
6,22 suggesting that not thick choroid may not always indicate a normal status. For the same reason, we postulate that the absence of or a markedly low prevalence of choroidal vascular hyperpermeability or punctate hyperfluorescent spots on ICGA may not always indicate a normal choroidal vasculature.
In general, eyes with RAP and eyes with myopic CNV have a thin choroid. Considering that choroidal vascular hyperpermeability and punctate hyperfluorescent spots were usually noted in eyes with thick choroids, a relatively thin choroid may be the underlying reason for the low prevalence of these findings in eyes with RAP and myopic CNV. However, among our patients with typical exudative AMD or PCV, some eyes also had choroidal vascular hyperpermeability or punctate hyperfluorescent spots, even when the choroid was relatively thin. This suggests that a thin choroid may not be the only reason for the low prevalence of these findings observed in eyes with RAP and myopic CNV. Yannuzzi and associates
16 postulated that RAP may develop as a result of a hypoxic environment. Others have also suggested a possible role of choroidal perfusion disturbances
23–25 and retinal pigment epithelial cell hypoxia
26 in the development of RAP. A thin choroid has also been thought to be a contributing cause of the hypoxic condition.
24,26 In highly myopic eyes, a correlation between a thin choroid and poor visual acuity has been demonstrated.
5,27 Decreased choroidal perfusion, due to choroidal thinning, was thought to underlie this association.
5 Similar to RAP, choroidal filling delays on ICGA have been reported in eyes with myopic CNV. This is also suggestive of decreased choroidal perfusion in this condition.
28 We postulate that the markedly low prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots in eyes with RAP and myopic CNV may, at least in part, have originated from decreased choroidal perfusion, regardless of its association with thin choroid.
In the present study, punctate hyperfluorescent spot prevalence was relatively lower in eyes with myopic CNV than in eyes with RAP. In addition, clustered punctate hyperfluorescent spots were not noted in any eye with myopic CNV. We believe that the onset of choroidal thinning may have had some influence. Although it is well-known that eyes with RAP usually have thin choroids,
24,26 the exact timing of choroidal thinning is not known. Because choroidal changes associated with AMD usually develop after middle age, it is possible that some eyes with RAP had a normal or relatively thick choroid prior to the development of age-related changes. In a previous study, the locations of tiny retinal pigment epithelium protrusions on OCT images corresponded to the locations of punctate hyperfluorescent spots on ICGA images.
1 This suggests that the spots on ICGA are closely associated with structural alterations as well as functional abnormalities. It is possible that this anatomical change persists over the long-term, despite the causative underlying condition had resolved. If persistent punctate hyperfluorescent spots had developed before age-related changes had occurred, when choroidal perfusion was not compromised, these spots would be observed, even in eyes with RAP. On the other hand, eyes with high myopia generally have a thin choroid, even in relatively young subjects.
29 This may suggest that choroidal perfusion may already be somewhat compromised, far before myopic CNV develops. Thus, the likelihood of developing punctate hyperfluorescent spots may be relatively low in myopic eyes, which would explain their extremely low prevalence in eyes with myopic CNV.
Although there is some debate,
30,31 several studies have demonstrated a decrease in choroidal thickness in eyes with exudative AMD following anti-VEGF therapy.
32,33 This suggests that anti-VEGF agents may influence the underlying choroid, and that choroidal vascular permeability reduction is a possible mechanism.
33 In a previous study, a decrease in choroidal vascular hyperpermeability was noted in eyes with CSC following intravitreal anti-VEGF therapy,
34 suggesting a possible effect of anti-VEGF agents on choroidal permeability. However, to the best of our knowledge, changes in choroidal vascular hyperpermeability and punctate hyperfluorescent spots following anti-VEGF therapy has not been evaluated in eyes with CNV. In the present study, no definite changes in choroidal vascular hyperpermeability and punctate hyperfluorescent spots were noted after intravitreal ranibizumab therapy. This suggests that the two ICGA findings may not be significantly influenced by short-term anti-VEGF therapy. However, our method of evaluating these changes was limited to the appearance or disappearance of ICGA findings. Additionally, only a small percentage (27%) of our patients with exudate AMD were evaluated for these changes. Further studies on consecutive patients are needed to better examine subtle change in these ICGA findings with anti-VEGF therapy.
In the present study, punctate hyperfluorescent spots were noted in nearly all eyes with choroidal vascular hyperpermeability, suggesting that these ICGA abnormalities share a common pathophysiologic mechanism.
8,9 Although the exact cause of punctate hyperfluorescent spots is not clear, previous studies showed that choroidal vascular hyperpermeability can induce pigment epithelial detachment.
35,36 Thus, we postulated that choroidal vascular hyperpermeability may contribute to the development of retinal pigment epithelium protrusions that are shown as punctate hyperfluorescent spots on ICGA.
In eyes with choroidal vascular hyperpermeability, clustered punctate hyperfluorescent spots were observed more frequently than in eyes without hyperpermeability. It is possible that the relatively high choroidal vascular hyperpermeability in certain regions may induce multiple, clustered retinal pigment epithelium protrusions in the adjacent area; that is, clustered punctate hyperfluorescent spots may reflect greater choroidal vascular hyperpermeability than a few scattered spots. The higher incidence of clustered spots in eyes with PCV than in eyes with other CNV subtypes also supports this theory because PCV is generally associated with choroidal vascular hyperpermeability. However, this remains somewhat controversial. It is well-known that choroidal vascular hyperpermeability is closely associated with a thick choroid. Although eyes with clustered spots tended to have a slightly greater choroidal thickness than eyes with solitary spots, the slight mean difference of 8.4 μm was not statistically significant. Even though some eyes were not evaluated because enhanced-depth imaging OCT images were not available, this small difference is not enough to conclude that clustered spots are associated with choroidal vascular hyperpermeability. However, we suspect that anatomical changes occurred in some eyes with clustered spots long before the included examination date. It is possible that the choroid was thicker and had a more severe vascular hyperpermeability, enough to lead to an anatomical alteration. As age-related changes progressed, the underlying condition that resulted in clustered spots may have resolved, resulting in a more normal choroidal thickness. This would lead to a relatively normal choroidal thickness and the absence of definite choroidal vascular hyperpermeability, even in the presence of clustered spots. Further longitudinal studies are required to reveal more accurate associations between the development of punctate hyperfluorescent spots and the degree of choroidal vascular hyperpermeability.
One drawback to the evaluation of choroidal vascular hyperpermeability and punctate hyperfluorescent spots is that there are no objective guidelines to determine the presence of these findings. In the present study, two independent examiners evaluated ICGA findings. Although a high rate of initial agreement was achieved when evaluating the presence of punctate hyperfluorescence spots, the rate was less satisfactory when evaluating the presence of choroidal vascular hyperpermeability. We hope that further studies establish more objective and sophisticated diagnostic methods for choroidal vascular hyperpermeability and punctate hyperfluorescent spots.
This study had limitations related to its retrospective design. Additionally, all data were collected from the medical records at a single-center and all included patients were Asian. Although a relatively large number of patients were included, data from enhanced-depth imaging OCT were only available from a limited number of patients. In addition, only 26.3% of eyes with exudative AMD had available ICGA images that were obtained 3 months after treatment. In the present study, we focused on the common findings in two different disorders, RAP and myopic CNV. However, it is important to note that the cause of choroidal thinning is different between RAP and myopic CNV. Choroidal thinning in eyes with RAP may be caused by age-related changes,
37,38 whereas thinning is associated with eyeball elongation in eyes with myopic CNV.
39,40 Thus, it is possible that the cause or clinical significance of low prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots differs between the two clinical entities.
In summary, the prevalence of choroidal vascular hyperpermeability and punctate hyperfluorescent spots in eyes with RAP and eyes with myopic CNV was markedly lower than in eyes with typical exudative AMD and eyes with PCV. We suggest that this result may be partially attributable to decreased choroidal perfusion in eyes with RAP and myopic CNV. Choroidal vascular hyperpermeability was more frequently noted in eyes with clustered pattern punctate hyperfluorescent spots than in eyes with solitary pattern spots. In addition, clustered pattern spots were more frequently noted in PCV than in typical exudative AMD. These results may suggest that the cluster pattern spots are more closely associated with increased choroidal vascular hyperpermeability as well as a greater increase in permeability compared with solitary pattern spots. Further population-based studies with longer follow-up are required to better evaluate the cause and prevalence of these findings and to understand how they are associated with pathologic conditions.
Supported by grants from Kim's Eye Hospital Research Center (Seoul, South Korea).
Disclosure: J.H. Kim, None; Y.S. Chang, None; T.G. Lee, None; C.G. Kim, None