CHARACTERISTICS OF CYSTOID SPACES IN TYPE 2 IDIOPATHIC MACULAR TELANGIECTASIA ON SPECTRAL DOMAIN OPTICAL COHERENCE TOMOGRAPHY IMAGES JONG-HYUN OH, MD, PHD,* JAERYUNG OH, MD, PHD,† ARIUNAA TOGLOOM, PHD,† SEONG-WOO KIM, MD, PHD,† KUHL HUH, MD, PHD† Purpose: To investigate the morphologic and topographic characteristics of intraretinal cystoid spaces in eyes with Type 2 idiopathic macular telangiectasia (MacTel 2). Methods: Using B-scan and en face images of eyes with MacTel 2 taken from a spectral domain optical coherence tomography database, the circularities and mean gray values of the cystoid spaces were measured to determine their boundaries and reflectivity. The characteristics of cystoid spaces in MacTel 2 eyes were compared with those in eyes with Type 1 idiopathic macular telangiectasia (MacTel 1), retinal vein occlusion, and diabetic macular edema, which are caused by vascular leakage. The cystoid spaces of en face optical coherence tomography images were matched with fluorescein angiographic images. Results: The circularity of the cystoid spaces in B-scan and en face optical coherence tomography images of 16 eyes with MacTel 2 was lower than that of eyes with MacTel 1 (P = 0.004 and P = 0.003, respectively), retinal vein occlusion (P , 0.001 and P , 0.001, respectively), and diabetic macular edema (P , 0.001 and P , 0.001, respectively). The mean gray value ratio of the cystoid spaces of eyes with MacTel 2 was lower than that of eyes with MacTel 1 (P = 0.002) and diabetic macular edema (P , 0.001). In eyes with MacTel 2, the cystoid spaces were located in the foveal center or parafoveal area. Conclusion: Characteristics of cystoid spaces of eyes with MacTel 2 were different from those in eyes with MacTel 1, retinal vein occlusion, and diabetic macular edema. The irregular boundaries and low reflectivity of the cystoid spaces in MacTel 2 may represent the degenerative origin of the disease. RETINA 34:1123–1131, 2014
T
a neurodegenerative disease.4 However, its cause and pathogenesis remain unclear. The diagnosis of MacTel 2 is based on biomicroscopic and fluorescein angiographic findings, including a loss of retinal transparency, retinal crystals, right-angled venules, and angiographic diffuse leakage.1,2 The introduction of optical coherence tomography (OCT) imaging has provided a valuable tool for identifying various changes in MacTel 2. Studies using OCT images have revealed details such as intraretinal cystoid spaces, hyperreflective spots, and disruption of the line representing the junction of the photoreceptor inner and outer segments (IS/OS line).5–14 Yannuzzi et al2 suggested that the progression of cystoid spaces is the result of Müller cell degeneration and cystic macular degeneration in the fovea. In
ype 2 idiopathic macular telangiectasia (MacTel 2) is an uncommon retinal disease that affects the juxtafoveal region of both eyes.1–3 Type 2 idiopathic macular telangiectasia usually manifests as a slow decrease in visual acuity in the fifth to seventh decades of life. Recently, MacTel 2 has been suggested to be From the *Department of Ophthalmology, Dongguk University Ilsan Hospital, Goyang, South Korea; and †Department of Ophthalmology, Korea University College of Medicine, Seoul, South Korea. Supported by a grant from the Korean Health Technology R & D Project, Ministry for Health, Welfare and Family Affairs, Korea (A102024). None of the authors have any conflicting interests to disclose. Reprint requests: Jaeryung Oh, MD, PhD, Department of Ophthalmology, Korea University College of Medicine, 126-1 Anam-dong 5-ga, Sungbuk-gu, Seoul 136-705, Korea; e-mail: [email protected]
1123
1124 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
a histopathologic study, Powner et al15 demonstrated Müller cell depletion in the macula of an eye donor with MacTel 2. Determination of the morphologic and topographic characteristics of eyes with MacTel 2 may help elucidate the pathogenesis of the disease, which is still poorly understood. In this study, we investigated the morphologic characteristics of the intraretinal cystoid spaces of eyes with MacTel 2 using images obtained by spectral domain optical coherence tomography (SD-OCT). We hypothesize that degenerative cystoid spaces have different shapes and reflectivities compared with exudative cystoid spaces. We measured the circularity and reflectivity of the cystoid spaces of eyes with MacTel 2 and compared these characteristics with those of eyes with Type 1 idiopathic macular telangiectasia (MacTel 1), retinal vein occlusion (RVO), and diabetic macular edema (DME). Type 1 idiopathic macular telangiectasia, RVO, and DME have abnormal intraretinal cystoid spaces similar to those of MacTel 2 but are caused by vascular leakage because of disruption of the inner blood–retinal barrier (BRB). We also investigated the topographic relationships of cystoid spaces to other intraretinal abnormalities in eyes with MacTel 2.
2014 VOLUME 34 NUMBER 6
eye was selected randomly. Type 1 idiopathic macular telangiectasia was diagnosed on the basis of the results of the fundus examination, fluorescein angiography, and OCT, after excluding age-related macular degeneration, polypoidal choroidal vasculopathy, pathologic myopia, idiopathic choroidal neovascularization, angioid streaks, RVO, radiation retinopathy, and diabetes. Eyes with history of treatment, including macular photocoagulation, photodynamic therapy, pars plana vitrectomy, intravitreal anti–vascular endothelial growth factor injection within 3 months, intravitreal triamcinolone acetonide injection within 3 months, and/or cataract surgery within 1 year, were also excluded from the study sample. Spectral Domain Optical Coherence Tomography Imaging The SD-OCT (3D OCT-1000 Mark II, software version 3.20; Topcon Corp, Tokyo, Japan) device used in this study had a wavelength of 840 nm, horizontal resolution of #20 mm, and axial resolution of up to 5 mm. Its imaging speed was 27,000 axial scans per second. Patients underwent SD-OCT evaluations using
Subjects and Methods Institutional review board approval was obtained from the Korea University Medical Center, Seoul, Korea. All research adhered to the tenets of the Declaration of Helsinki. Subject Selection and Data Collection We obtained and reviewed the medical records of all patients with both diagnosis of MacTel 2 and available SD-OCT images in our database between 2009 and 2011. Each patient underwent a full ophthalmic examination, including fluorescein angiography using a fundus camera (Zeiss FF 450 Plus; Carl Zeiss Meditec AG, Jena, Germany). Diagnoses of MacTel 2 were based on biomicroscopic, fluorescein angiographic, and OCT evidence. The clinical signs for diagnosing MacTel 2 were loss of retinal transparency, retinal crystals, intraretinal cystoid spaces, right-angled venules, and angiographic diffuse leakage.1,2 All eyes with MacTel 2 were enrolled in this study and classified according to the staging system proposed by Gass and Blodi.1 The study included consecutive MacTel 1, RVO, or DME patients with cystoid spaces visible on SD-OCT images in our database. Images were used to compare the morphologic characteristics of the cystoid spaces of eyes with MacTel 2 with those with MacTel 1, RVO, and DME. If images for both eyes were available, one
Fig. 1. Spectral domain optical coherence tomography images in an eye with MacTel 2. The B-scan image (top) shows a cystoid space. En face images (bottom) were visualized using OCT En Face Viewer software (Enview, software version 1.00; Topcon Corp).
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL
3-dimensional scanning protocols with 128 B-scans (512 A-scans per B-scan) of a 6 mm · 6 mm area, and line-scanning protocols with an average of 50 B-scans (1024 A-scans per B-scan) with 6 mm lengths. The B-scan images were visualized using the B-scan module of the three-dimensional OCT image. Using the B-scan OCT images, we investigated which retinal layers were occupied by the cystoid spaces in eyes with MacTel 2. En face images were visualized using En Face Viewer software (Enview,
1125
software version 1.00; Topcon Corp) (Figure 1). Enview facilitates en face frontal scanning to adapt to the shape of the retinal pigment epithelium using the three-dimensional scanning data of SD-OCT. Measurement of the Circularity and Reflectivity of Cystoid Spaces The cystoid spaces in OCT images were defined as relative hyporeflective areas with surrounding
Fig. 2. Optical coherence tomographic images of cystoid spaces in eyes with MacTel 2 (A), MacTel 1 (B), RVO (C), and DME (D). Using ImageJ software, a white line was drawn along the boundary of each cystoid space in the OCT images.
1126 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
hyperreflective boundaries that could be distinguished or imaged. A single investigator (J.O.) selected the B-scan and en face OCT image with the largest cystoid space. Optical coherence tomography images were obtained as gray-scale images and were numbered irrespective of diagnosis. Each image was loaded into the ImageJ software program (version 1.44p; National Institutes of Health, Bethesda, MD). In ImageJ, a white line was drawn along the boundary of the cystoid space using the Eraser tool with a width of 2 pixels (Figure 2). Circularity was measured in ImageJ to determine the shape of the boundary of the cystoid space, and the mean gray value was measured to determine reflectivity. Circularity values ranged from 0 (infinitely elongated polygon) to 1 (perfect circle).16 The mean gray value of the vitreous cavity was obtained to determine differences in signal strength among eyes. The mean gray value of the vitreous cavity was calculated as the average of mean gray values measured in 3 different places with a size of 100 pixels · 100 pixels using the Rectangle tool. Two independent observers (J.-H.O. and A.T.) who were masked to the diagnoses analyzed images in ImageJ.
Fig. 3. Topographic mapping of en face OCT images and late-phase fluorescein angiographic images. Outlines of foveolar (red) cystoid spaces, parafoveolar (yellow) cystoid spaces, clusters of intraretinal hyperreflective spots (green), retinal pigmentations (pink), and disruption of the line representing the junction of photoreceptor IS/OS line (blue) were drawn on a late-phase angiographic image.
2014 VOLUME 34 NUMBER 6
Tomographic and Topographic Mapping of Optical Coherence Tomography Images and Fluorescein Angiography After delineating the location of cystoid spaces in B-scan images, we performed topographic mapping of cystoid spaces using en face and fluorescein angiographic images. To determine the topographic relationships of cystoid spaces to other intraretinal abnormalities in eyes with MacTel 2, we drew outlines around cystoid spaces, clusters of intraretinal hyperreflective spots, retinal pigmentations, and/or IS/OS disruptions of IS/OS line on late-phase fluorescein angiographic images using Adobe Photoshop CS3 (Adobe Systems Inc, San Jose, CA) (Figure 3). En face OCT images with the largest cystoid space cross-sections were selected and manually overlapped onto a late-phase fluorescein angiographic image about the retinal vessels. Outlines of each cystoid space were drawn on a separate layer on the background of en face OCT images. The en face OCT image was then hidden. Intraretinal hyperreflective spots were defined as hyperreflective lesions that could be distinguished from the reflectivity of surrounding tissue. En face OCT images with the largest cluster of hyperreflective spots were selected and outlined. Areas of retinal
1127
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL Table 1. Characteristics of 12 Patients With MacTel 2 24 eyes of 12 patients
Total Sex (male/female) Age, years BCVA, logMAR Stage 1 2 3 4 5 Central point retinal thickness, mm Central subfield retinal thickness, mm Less than normal range Normal range (5%–95%) Greater than normal range
2/10 57.8 ± 11.9 0.2 ± 0.2 3 (12.5%) 13 (54.2%) 2 (8.3%) 5 (20.8%) 1 (4.2%) 152 ± 43 187 ± 38 13 (54.2%) 10 (41.7%) 1 (4.2%)
Results are presented as mean ± SD or number of eyes (percentage). BCVA, best-corrected visual acuity; logMAR, logarithm of the minimum angle of resolution.
pigmentation were outlined in the same way on color fundus photographs. Inner and outer segment disruptions were visualized using the B-scan module of SD-OCT, and both ends of the disruption were marked on color fundus SD-OCT photographs. Outlines were drawn along the marked points on several photographs. Statistical Analyses All data were analyzed using SPSS software version 20.0 (SPSS Inc, Chicago, IL). P , 0.05 were considered statistically significant. Results are expressed as mean ± SD. Intraobserver and interobserver reproducibilities were assessed using an intraclass correlation coefficient. The Mann–Whitney U test was used for comparison between groups. Wilcoxon signed-rank test was used to identify the statistical significance of differences in mean gray values between the cystoid space and the vitreous cavity in each group.
Results Patient Characteristics Twenty-four eyes of 12 patients with MacTel 2 were included. Their characteristics are shown in Table 1. Two of the 12 patients with MacTel 2 did not have cystoid spaces. Sixteen eyes of 10 patients with MacTel 2 and cystoid spaces were compared with 4 eyes of 4 patients with MacTel 1, 20 eyes of 20 patients with RVO, and 21 eyes of 21 patients with DME. The characteristics of subjects with cystoid spaces are shown in Table 2. Circularity and Reflectivity of Cystoid Spaces in Optical Coherence Tomography Images Intraobserver and interobserver intraclass correlation coefficients for measurements of circularity and mean gray values were all .0.9 (P , 0.001) (Table 3). The mean circularity of cystoid spaces in eyes with MacTel 2 was significantly different from that of eyes with MacTel 1, RVO, and DME (Table 4). Although the circularities of cystoid spaces in eyes with MacTel 1, RVO, and DME were .0.7, the circularity was ,0.5 in eyes with MacTel 2 in both B-scan and en face OCT images (Table 4). The mean gray value of cystoid spaces in eyes with MacTel 2 was significantly lower than that of eyes with MacTel 1 (P = 0.003 and P = 0.002) and DME (P , 0.001 and P , 0.001) in both B-scan and en face OCT images, respectively (Table 5). However, the mean gray value of cystoid spaces in eyes with MacTel 2 was not different from that of eyes with RVO in either the B-scan or en face OCT images (P = 0.162 and P = 0.114, respectively). The mean gray values of cystoid spaces were greater than those of the vitreous cavities in all groups (P , 0.001). Tomographic and Topographic Characteristics of Cystoid Spaces In B-scan images, the cystoid space was localized within the inner retina (Henle fiber layer) in all five
Table 2. Characteristics of Subjects With Cystoid Spaces
Total patients Mean age, years Men/women Diabetes Hypertension Enrolled eyes OD/OS BCVA, logMAR Snellen equivalent
MacTel 2
MacTel 1
RVO
DME
10 60.9 ± 10.2 2/8 2/10 (20%) 7/10 (70%) 16 10/6 0.3 ± 0.2 20/35
4 51.3 ± 8.6 4/0 0/4 (0%) 1/4 (25%) 4 2/2 0.2 ± 0.2 20/30
20 62.1 ± 11.5 8/12 0/20 (0%) 11/20 (55%) 20 10/10 0.8 ± 0.5 20/125
21 55.8 ± 10.2 8/13 21/21 (100%) 10/21 (47.6%) 21 10/11 0.6 ± 0.5 20/80
Results are presented as mean ± SD or number (percentage). BCVA, best-corrected visual acuity; logMAR, logarithm of the minimum angle of resolution.
1128 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2014 VOLUME 34 NUMBER 6
Table 3. Intraobserver and Interobserver Intraclass Correlation Coefficients for Measurements of the Circularity and Mean Gray Value of Cystoid Spaces in OCT Images Circularity
Intraobserver 1 Intraobserver 2 Interobserver
Mean Gray Value
ICC
95% CI
P
ICC
95% CI
P
0.947 0.935 0.912
0.910–0.969 0.888–0.962 0.854–0.948
,0.001 ,0.001 ,0.001
0.994 0.995 0.988
0.990–0.997 0.996–0.999 0.980–0.993
,0.001 ,0.001 ,0.001
CI, confidence interval; ICC, intraclass correlation coefficient.
eyes with only foveolar cystoid spaces. In 11 eyes with both foveolar and parafoveolar cystoid spaces, 4 foveolar cystoid spaces were localized within the inner retina, 4 foveolar cystoid spaces were localized within the inner retina and outer nuclear layer, and 3 foveolar cystoid spaces extended from the inner retina to the IS/OS. Nine parafoveolar cystoid spaces were localized within the inner retina, one parafoveolar cystoid space was localized within the inner retina and outer nuclear layer, and one parafoveolar cystoid space extended from the inner retina to the IS/OS. There were two Stage 1 cases with cystoid spaces. These 2 eyes had only foveolar cystoid spaces, and these were localized within the inner retina (Figure 4). The cystoid spaces identified in en face OCT images were matched with angiographic images. In the en face OCT images of eyes with MacTel 2, cystoid spaces were located in the foveal center or the parafoveal area. Among 16 eyes with MacTel 2, 11 eyes had both foveolar and parafoveolar cystoid spaces. In the remaining 5 eyes, the cystoid space was located only in the foveal center (Table 6). In 7 of 11 eyes with both foveolar and parafoveolar cystoid spaces, the parafoveolar cystoid spaces had a similar location to the angiographic leakage but were not filled with the dye. A different grouping of 7 of the 11 eyes with both foveolar and parafoveolar cystoid spaces had clusters of hyperreflective spots that were in a similar location to the parafoveolar cystoid space, except for 1 eye. All 11 eyes with foveolar and parafoveolar cystoid spaces had disruptions of IS/OS line. The disruptions of IS/OS line were in similar locations to the foveolar (1 eye), parafoveolar (3 eyes), or both cystoid spaces (6 eyes), except for 1 eye. Angiographic leakage was detected in all five eyes with only foveolar cystoid spaces. However, the
locations of cystoid spaces did not match with those of the angiographic leakages. Eleven of 24 eyes with MacTel 2 had clusters of hyperreflective spots. These clusters had similar locations to the retinal pigmentations, parafoveolar cystoid spaces, and/or sites of angiographic leakage. Six of 24 eyes with MacTel 2 had pigmentation. All six of these eyes had clusters of hyperreflective spots in similar locations to retinal pigmentations. Retinal pigmentations and clusters of hyperreflective spots were only encountered in the parafoveal area.
Discussion In this study, the cystoid spaces of eyes with MacTel 2 were less circular than the cystoid spaces of eyes with MacTel 1, RVO, and DME, which are caused by vascular leakage because of disruption of the inner BRB. These findings support the suggestion that cystoid spaces in eyes with MacTel 2 are caused by degenerative changes rather than vascular leakage. Because the suggestion that the primary abnormality may reside in parafoveolar retinal neural cells, Müller cells, or both,17 many studies have suggested that Müller cells are involved in the pathogenesis of MacTel 2.2,4,10,15 Unlike natural optical fibers, such as the IS/OS of photoreceptor cells, Müller cells have many complex side branches that enclose neuronal compartments.18 The degenerative involvement of adjacent Müller cells may be responsible for irregular cystoid space boundaries observed in eyes with MacTel 2. Recently, Ooto et al19 reported the cone abnormality differences between MacTel subtypes related to the different pathogenesis of these subtypes. In this study, we used a novel approach, the analysis of circularity, to
Table 4. Circularity of Cystoid Spaces in OCT Images
B-scan images En face images
MacTel 2
MacTel 1 (P*)
RVO (P*)
DME (P*)
0.43 ± 0.13 0.48 ± 0.09
0.78 ± 0.08 (0.004) 0.70 ± 0.08 (0.003)
0.74 ± 0.08 (,0.001) 0.73 ± 0.07 (,0.001)
0.75 ± 0.05 (,0.001) 0.77 ± 0.05 (,0.001)
Results are presented as mean ± SD. P values were calculated using the Mann–Whitney U test. *Compared with MacTel 2.
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL
1129
Table 5. Reflectivity of Cystoid Spaces in OCT Images
B-scan images Mean gray value Mean gray value Mean gray value En face images Mean gray value Mean gray value Mean gray value
MacTel 2
MacTel 1 (P*)
RVO (P*)
DME (P*)
of cystoid space of vitreous cavity ratio
11.5 ± 2.4 9.3 ± 1.4 1.2 ± 0.2
20.5 ± 3.3 (0.003) 9.8 ± 1.5 (0.832) 2.1 ± 0.1 (0.003)
12.7 ± 2.6 (0.162) 10.4 ± 1.3 (0.019) 1.2 ± 0.2 (0.675)
18.1 ± 6.3 (,0.001) 10.4 ± 1.2 (0.008) 1.8 ± 0.7 (0.017)
of cystoid space of vitreous cavity ratio
12.2 ± 1.9 10.6 ± 0.9 1.2 ± 0.2
21.3 ± 2.0 (0.002) 10.9 ± 0.7 (0.537) 1.9 ± 0.1 (0.002)
12.7 ± 2.2 (0.114) 11.5 ± 1.3 (0.006) 1.1 ± 0.3 (0.080)
20.1 ± 6.6 (,0.001) 11.3 ± 1.3 (0.029) 1.8 ± 0.7 (,0.001)
Results are presented as mean ± standard deviation. P values were calculated using the Mann–Whitney U test. *Compared with MacTel 2.
quantify the value of the shape of the cystoid spaces and demonstrate significant differences in the shape of cystoid space in eyes with MacTel 2 compared with MacTel 1 and other diseases with vascular exudation. Circularity is not measured in routine settings, and the shape of cystoid spaces can be qualitatively observed on OCT. However, the quantitative measurement used in this study would be included in the differential diagnosis of diseases with vascular exudation. The reflectivity of cystoid spaces in eyes with MacTel 2 was lower than the reflectivity in eyes with DME. The gray-scale analysis of images used in this study is a relatively insensitive measure that is, probably not as sensitive as analysis of the OCT waveforms, such as that performed by Barthelmes et al.20 However, our findings are consistent with their results.20 Although the cystoid spaces of eyes with RVO or DME are known to result from leakage, we found that the reflectivity of cystoid spaces in eyes with RVO was lower than that of eyes with DME (P , 0.001), but was not different from that of eyes with MacTel 2. It is not clear why there was
Fig. 4. Late-phase fluorescein angiographic and OCT images of an eye with MacTel 2 in Stage 1. Fluorescein angiography shows mild hyperfluorescence temporal to the foveola. The foveolar cystoid space is localized within the inner retina in the B-scan OCT image.
a difference in the reflectivity of cystoid spaces between eyes with RVO or DME. Differences in plasma constituents, pathophysiology, size of BRB breakdown, and/or hydrostatic pressure through the BRB may affect the reflectivity of cystoid spaces. To gain a better understanding, we topographically matched OCT findings on en face images to late-phase angiographic images.21 In this study, there was topographic discordance between cystoid space and angiographic leakage, as in previous study.5 There were no foveolar cystoid spaces that had similar locations to sites of vascular leakage. This may be because of the geographic characteristics of foveal vasculature, that is, the foveal avascular zone. Some parafoveolar cystoid spaces had a similar location to the angiographic leakage but were not filled with the dye. Baumüller et al6 reported that hyperreflective spots in the outer retina were found in all stages of the disease. Because of the relatively low resolution of our OCT images and difficulty in differentiating them from noise, we did not identify hyperreflective spots as they did in the study by Baumüller et al,6 but we analyzed clustering spots in OCT images. Except for six eyes with retinal pigmentation, the clusters of hyperreflective spots were similar in location to the sites of angiographic leakage. This finding suggests that some hyperreflective spots may be related to vascular abnormalities or extravasated deposits, as proposed by Baumüller et al.6 No eyes had only parafoveolar cystoid spaces. The five eyes with only foveolar cystoid spaces did not show clusters of hyperreflective spots, disruptions of IS/OS line, or retinal pigmentations. This finding suggests that cystoid spaces may occur in the foveolar area first, and that foveolar-only cystoid spaces may be early findings. Inner and outer segment disruptions, retinal pigmentation, and clusters of hyperreflective spots may be secondary to cystoid spaces or may be late findings of MacTel 2. Cystoid spaces may be related to the loss of photoreceptor cells and/or Müller cell processes, because they are only compositions in the foveolar area.22
1130 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2014 VOLUME 34 NUMBER 6
Table 6. Cystoid Spaces and Other Abnormalities in 16 Eyes With MacTel 2 Cystoid Spaces Eyes with both foveolar and parafoveolar cystoid spaces (n = 11 eyes) Eyes with only foveolar cystoid spaces (n = 5 eyes)
Cluster of Disruptions of Angiographic Leakage Hyperreflective Spots IS/OS Line Retinal Pigmentations 11 eyes
7 eyes
11 eyes
3 eyes
5 eyes
None
None
None
The IS/OS line represents the junction of the inner and outer photoreceptor segments.
There were several limitations to this study. En face images are not readily available in every OCT device. Therefore, the circularity of cystoid spaces on B-scan images was also measured. The results were not different from those of en face images. The diagnosis of MacTel 2 was not confirmed with other imaging techniques. More recently, it has been recognized that focal loss of macular pigments, consisting of lutein and zeaxanthin on fundus autofluorescence, is a more sensitive and specific marker.23–26 This study has other limitations such as a retrospective design and few cases. In this study, only images with the largest cystoid spaces were used for comparison. Further comparison, including cystoid spaces localized in different layers, would be interesting because the anatomical differences between different retinal layers or between the fovea and parafovea may influence the shape of cystoid spaces. However, the limited number of MacTel 2 cases prevented further analysis. Using the B-scan OCT images, we showed which retinal layers were occupied by the cystoid spaces in eyes with MacTel 2. However, it is difficult to draw conclusions from the relatively low number of patients included in this study. Klein et al27 reported that the prevalence of MacTel 2 was higher than previously estimated, but MacTel 2 is still rare.28 Large-scale or multicenter studies are required to increase the number of cases that can be analyzed. Another limitation of our study is that the images were analyzed manually. However, intraclass correlation coefficients for the circularity and mean gray value measurements of cystoid spaces were all .0.9. In conclusion, the morphologic characteristics of cystoid spaces in eyes with MacTel 2 were different from those in eyes with MacTel 1, RVO, and DME, which are caused by vascular leakage because of disruption of the inner BRB. Cystoid spaces in eyes with MacTel 2 have more irregular boundaries than those in eyes with MacTel 1, RVO, and DME, and lower reflectivity than those in eyes with MacTel 1 and DME. Our results are consistent with the observations
of previous studies, which have suggested that MacTel 2 is related to degenerative changes. Key words: macular telangiectasia, en face images, optical coherence tomography, intraretinal cystoid spaces. References 1. Gass JD, Blodi BA. Idiopathic juxtafoveolar retinal telangiectasis. Update of classification and follow-up study. Ophthalmology 1993;100:1536–1546. 2. Yannuzzi LA, Bardal AM, Freund KB, et al. Idiopathic macular telangiectasia. Arch Ophthalmol 2006;124:450–460. 3. Gass JD, Oyakawa RT. Idiopathic juxtafoveolar retinal telangiectasis. Arch Ophthalmol 1982;100:769–780. 4. Issa PC, Gillies MC, Chew EY, et al. Macular telangiectasia type 2. Prog Retin Eye Res 2013;34:49–77. 5. Koizumi H, Iida T, Maruko I. Morphologic features of group 2A idiopathic juxtafoveolar retinal telangiectasis in threedimensional optical coherence tomography. Am J Ophthalmol 2006;142:340–343. 6. Baumüller S, Charbel Issa P, Scholl HP, et al. Outer retinal hyperreflective spots on spectral-domain optical coherence tomography in macular telangiectasia type 2. Ophthalmology 2010;117:2162–2168. 7. Gupta V, Gupta A, Dogra MR, Agarwal A. Optical coherence tomography in group 2A idiopathic juxtafoveolar telangiectasis. Ophthalmic Surg Lasers Imaging 2005;36:482–486. 8. Gaudric A, Ducos de Lahitte G, Cohen SY, et al. Optical coherence tomography in group 2A idiopathic juxtafoveolar retinal telangiectasis. Arch Ophthalmol 2006;124:1410– 1419. 9. Maruko I, Iida T, Sekiryu T, Fujiwara T. Early morphological changes and functional abnormalities in group 2A idiopathic juxtafoveolar retinal telangiectasis using spectral domain optical coherence tomography and microperimetry. Br J Ophthalmol 2008;92:1488–1491. 10. Sallo FB, Leung I, Chung M, et al. Retinal crystals in type 2 idiopathic macular telangiectasia. Ophthalmology 2011;118: 2461–2467. 11. Surguch V, Gamulescu MA, Gabel VP. Optical coherence tomography findings in idiopathic juxtafoveal retinal telangiectasis. Graefes Arch Clin Exp Ophthalmol 2007;245: 783–788. 12. Paunescu LA, Ko TH, Duker JS, et al. Idiopathic juxtafoveal retinal telangiectasis: new findings by ultrahigh-resolution optical coherence tomography. Ophthalmology 2006;113:48–57.
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL 13. Sallo FB, Peto T, Egan C, et al. The IS/OS junction layer in the natural history of type 2 idiopathic macular telangiectasia. Invest Ophthalmol Vis Sci 2012;53:7889–7895. 14. Sallo FB, Peto T, Egan C, et al. “En face” OCT imaging of the IS/OS junction line in type 2 idiopathic macular telangiectasia. Invest Ophthalmol Vis Sci 2012;53:6145–6152. 15. Powner MB, Gillies MC, Tretiach M, et al. Perifoveal müller cell depletion in a case of macular telangiectasia type 2. Ophthalmology 2010;117:2407–2416. 16. Blott SJ, Pye K. Particle shape: a review and new methods of characterization and classification. Sedimentology 2008;55: 31–63. 17. Gass JD. Histopathologic study of presumed parafoveal telangiectasis. Retina 2000;20:226–227. 18. Franze K, Grosche J, Skatchkov SN, et al. Müller cells are living optical fibers in the vertebrate retina. Proc Natl Acad Sci U S A 2007;104:8287–8292. 19. Ooto S, Hangai M, Takayama K, et al. Comparison of cone pathologic changes in idiopathic macular telangiectasia types 1 and 2 using adaptive optics scanning laser ophthalmoscopy. Am J Ophthalmol 2013;155:1045– 1057. 20. Barthelmes D, Sutter FK, Gillies MC. Differential optical densities of intraretinal spaces. Invest Ophthalmol Vis Sci 2008; 49:3529–3534.
1131
21. Wanek J, Zelkha R, Lim JI, Shahidi M. Feasibility of a method for en face imaging of photoreceptor cell integrity. Am J Ophthalmol 2011;152:807–814. 22. Massey SC. Functional anatomy of the mammalian retina. In: Ryan SJ, ed. Retina. Philadelphia, PA: Elsevier/Mosby; 2006: 43–82. 23. Charbel Issa P, van der Veen RL, Stijfs A, et al. Quantification of reduced macular pigment optical density in the central retina in macular telangiectasia type 2. Exp Eye Res 2009;89:25–31. 24. Zeimer MB, Padge B, Heimes B, Pauleikhoff D. Idiopathic macular telangiectasia type 2: distribution of macular pigment and functional investigations. Retina 2010;30:586–595. 25. Helb HM, Charbel Issa P, van der Veen RL, et al. Abnormal macular pigment distribution in type 2 idiopathic macular telangiectasia. Retina 2008;28:808–816. 26. Chin EK, Kim DY, Hunter AA III, et al. Staging of macular telangiectasia: power-Doppler optical coherence tomography and macular pigment optical density. Invest Ophthalmol Vis Sci 2013;54:4459–4470. 27. Klein R, Blodi BA, Meuer SM, et al. The prevalence of macular telangiectasia type 2 in the Beaver Dam eye study. Am J Ophthalmol 2010;150:55–62. 28. Aung KZ, Wickremasinghe SS, Makeyeva G, et al. The prevalence estimates of macular telangiectasia type 2: the Melbourne Collaborative Cohort Study. Retina 2010;30:473–478.
T
a neurodegenerative disease.4 However, its cause and pathogenesis remain unclear. The diagnosis of MacTel 2 is based on biomicroscopic and fluorescein angiographic findings, including a loss of retinal transparency, retinal crystals, right-angled venules, and angiographic diffuse leakage.1,2 The introduction of optical coherence tomography (OCT) imaging has provided a valuable tool for identifying various changes in MacTel 2. Studies using OCT images have revealed details such as intraretinal cystoid spaces, hyperreflective spots, and disruption of the line representing the junction of the photoreceptor inner and outer segments (IS/OS line).5–14 Yannuzzi et al2 suggested that the progression of cystoid spaces is the result of Müller cell degeneration and cystic macular degeneration in the fovea. In
ype 2 idiopathic macular telangiectasia (MacTel 2) is an uncommon retinal disease that affects the juxtafoveal region of both eyes.1–3 Type 2 idiopathic macular telangiectasia usually manifests as a slow decrease in visual acuity in the fifth to seventh decades of life. Recently, MacTel 2 has been suggested to be From the *Department of Ophthalmology, Dongguk University Ilsan Hospital, Goyang, South Korea; and †Department of Ophthalmology, Korea University College of Medicine, Seoul, South Korea. Supported by a grant from the Korean Health Technology R & D Project, Ministry for Health, Welfare and Family Affairs, Korea (A102024). None of the authors have any conflicting interests to disclose. Reprint requests: Jaeryung Oh, MD, PhD, Department of Ophthalmology, Korea University College of Medicine, 126-1 Anam-dong 5-ga, Sungbuk-gu, Seoul 136-705, Korea; e-mail: [email protected]
1123
1124 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
a histopathologic study, Powner et al15 demonstrated Müller cell depletion in the macula of an eye donor with MacTel 2. Determination of the morphologic and topographic characteristics of eyes with MacTel 2 may help elucidate the pathogenesis of the disease, which is still poorly understood. In this study, we investigated the morphologic characteristics of the intraretinal cystoid spaces of eyes with MacTel 2 using images obtained by spectral domain optical coherence tomography (SD-OCT). We hypothesize that degenerative cystoid spaces have different shapes and reflectivities compared with exudative cystoid spaces. We measured the circularity and reflectivity of the cystoid spaces of eyes with MacTel 2 and compared these characteristics with those of eyes with Type 1 idiopathic macular telangiectasia (MacTel 1), retinal vein occlusion (RVO), and diabetic macular edema (DME). Type 1 idiopathic macular telangiectasia, RVO, and DME have abnormal intraretinal cystoid spaces similar to those of MacTel 2 but are caused by vascular leakage because of disruption of the inner blood–retinal barrier (BRB). We also investigated the topographic relationships of cystoid spaces to other intraretinal abnormalities in eyes with MacTel 2.
2014 VOLUME 34 NUMBER 6
eye was selected randomly. Type 1 idiopathic macular telangiectasia was diagnosed on the basis of the results of the fundus examination, fluorescein angiography, and OCT, after excluding age-related macular degeneration, polypoidal choroidal vasculopathy, pathologic myopia, idiopathic choroidal neovascularization, angioid streaks, RVO, radiation retinopathy, and diabetes. Eyes with history of treatment, including macular photocoagulation, photodynamic therapy, pars plana vitrectomy, intravitreal anti–vascular endothelial growth factor injection within 3 months, intravitreal triamcinolone acetonide injection within 3 months, and/or cataract surgery within 1 year, were also excluded from the study sample. Spectral Domain Optical Coherence Tomography Imaging The SD-OCT (3D OCT-1000 Mark II, software version 3.20; Topcon Corp, Tokyo, Japan) device used in this study had a wavelength of 840 nm, horizontal resolution of #20 mm, and axial resolution of up to 5 mm. Its imaging speed was 27,000 axial scans per second. Patients underwent SD-OCT evaluations using
Subjects and Methods Institutional review board approval was obtained from the Korea University Medical Center, Seoul, Korea. All research adhered to the tenets of the Declaration of Helsinki. Subject Selection and Data Collection We obtained and reviewed the medical records of all patients with both diagnosis of MacTel 2 and available SD-OCT images in our database between 2009 and 2011. Each patient underwent a full ophthalmic examination, including fluorescein angiography using a fundus camera (Zeiss FF 450 Plus; Carl Zeiss Meditec AG, Jena, Germany). Diagnoses of MacTel 2 were based on biomicroscopic, fluorescein angiographic, and OCT evidence. The clinical signs for diagnosing MacTel 2 were loss of retinal transparency, retinal crystals, intraretinal cystoid spaces, right-angled venules, and angiographic diffuse leakage.1,2 All eyes with MacTel 2 were enrolled in this study and classified according to the staging system proposed by Gass and Blodi.1 The study included consecutive MacTel 1, RVO, or DME patients with cystoid spaces visible on SD-OCT images in our database. Images were used to compare the morphologic characteristics of the cystoid spaces of eyes with MacTel 2 with those with MacTel 1, RVO, and DME. If images for both eyes were available, one
Fig. 1. Spectral domain optical coherence tomography images in an eye with MacTel 2. The B-scan image (top) shows a cystoid space. En face images (bottom) were visualized using OCT En Face Viewer software (Enview, software version 1.00; Topcon Corp).
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL
3-dimensional scanning protocols with 128 B-scans (512 A-scans per B-scan) of a 6 mm · 6 mm area, and line-scanning protocols with an average of 50 B-scans (1024 A-scans per B-scan) with 6 mm lengths. The B-scan images were visualized using the B-scan module of the three-dimensional OCT image. Using the B-scan OCT images, we investigated which retinal layers were occupied by the cystoid spaces in eyes with MacTel 2. En face images were visualized using En Face Viewer software (Enview,
1125
software version 1.00; Topcon Corp) (Figure 1). Enview facilitates en face frontal scanning to adapt to the shape of the retinal pigment epithelium using the three-dimensional scanning data of SD-OCT. Measurement of the Circularity and Reflectivity of Cystoid Spaces The cystoid spaces in OCT images were defined as relative hyporeflective areas with surrounding
Fig. 2. Optical coherence tomographic images of cystoid spaces in eyes with MacTel 2 (A), MacTel 1 (B), RVO (C), and DME (D). Using ImageJ software, a white line was drawn along the boundary of each cystoid space in the OCT images.
1126 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
hyperreflective boundaries that could be distinguished or imaged. A single investigator (J.O.) selected the B-scan and en face OCT image with the largest cystoid space. Optical coherence tomography images were obtained as gray-scale images and were numbered irrespective of diagnosis. Each image was loaded into the ImageJ software program (version 1.44p; National Institutes of Health, Bethesda, MD). In ImageJ, a white line was drawn along the boundary of the cystoid space using the Eraser tool with a width of 2 pixels (Figure 2). Circularity was measured in ImageJ to determine the shape of the boundary of the cystoid space, and the mean gray value was measured to determine reflectivity. Circularity values ranged from 0 (infinitely elongated polygon) to 1 (perfect circle).16 The mean gray value of the vitreous cavity was obtained to determine differences in signal strength among eyes. The mean gray value of the vitreous cavity was calculated as the average of mean gray values measured in 3 different places with a size of 100 pixels · 100 pixels using the Rectangle tool. Two independent observers (J.-H.O. and A.T.) who were masked to the diagnoses analyzed images in ImageJ.
Fig. 3. Topographic mapping of en face OCT images and late-phase fluorescein angiographic images. Outlines of foveolar (red) cystoid spaces, parafoveolar (yellow) cystoid spaces, clusters of intraretinal hyperreflective spots (green), retinal pigmentations (pink), and disruption of the line representing the junction of photoreceptor IS/OS line (blue) were drawn on a late-phase angiographic image.
2014 VOLUME 34 NUMBER 6
Tomographic and Topographic Mapping of Optical Coherence Tomography Images and Fluorescein Angiography After delineating the location of cystoid spaces in B-scan images, we performed topographic mapping of cystoid spaces using en face and fluorescein angiographic images. To determine the topographic relationships of cystoid spaces to other intraretinal abnormalities in eyes with MacTel 2, we drew outlines around cystoid spaces, clusters of intraretinal hyperreflective spots, retinal pigmentations, and/or IS/OS disruptions of IS/OS line on late-phase fluorescein angiographic images using Adobe Photoshop CS3 (Adobe Systems Inc, San Jose, CA) (Figure 3). En face OCT images with the largest cystoid space cross-sections were selected and manually overlapped onto a late-phase fluorescein angiographic image about the retinal vessels. Outlines of each cystoid space were drawn on a separate layer on the background of en face OCT images. The en face OCT image was then hidden. Intraretinal hyperreflective spots were defined as hyperreflective lesions that could be distinguished from the reflectivity of surrounding tissue. En face OCT images with the largest cluster of hyperreflective spots were selected and outlined. Areas of retinal
1127
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL Table 1. Characteristics of 12 Patients With MacTel 2 24 eyes of 12 patients
Total Sex (male/female) Age, years BCVA, logMAR Stage 1 2 3 4 5 Central point retinal thickness, mm Central subfield retinal thickness, mm Less than normal range Normal range (5%–95%) Greater than normal range
2/10 57.8 ± 11.9 0.2 ± 0.2 3 (12.5%) 13 (54.2%) 2 (8.3%) 5 (20.8%) 1 (4.2%) 152 ± 43 187 ± 38 13 (54.2%) 10 (41.7%) 1 (4.2%)
Results are presented as mean ± SD or number of eyes (percentage). BCVA, best-corrected visual acuity; logMAR, logarithm of the minimum angle of resolution.
pigmentation were outlined in the same way on color fundus photographs. Inner and outer segment disruptions were visualized using the B-scan module of SD-OCT, and both ends of the disruption were marked on color fundus SD-OCT photographs. Outlines were drawn along the marked points on several photographs. Statistical Analyses All data were analyzed using SPSS software version 20.0 (SPSS Inc, Chicago, IL). P , 0.05 were considered statistically significant. Results are expressed as mean ± SD. Intraobserver and interobserver reproducibilities were assessed using an intraclass correlation coefficient. The Mann–Whitney U test was used for comparison between groups. Wilcoxon signed-rank test was used to identify the statistical significance of differences in mean gray values between the cystoid space and the vitreous cavity in each group.
Results Patient Characteristics Twenty-four eyes of 12 patients with MacTel 2 were included. Their characteristics are shown in Table 1. Two of the 12 patients with MacTel 2 did not have cystoid spaces. Sixteen eyes of 10 patients with MacTel 2 and cystoid spaces were compared with 4 eyes of 4 patients with MacTel 1, 20 eyes of 20 patients with RVO, and 21 eyes of 21 patients with DME. The characteristics of subjects with cystoid spaces are shown in Table 2. Circularity and Reflectivity of Cystoid Spaces in Optical Coherence Tomography Images Intraobserver and interobserver intraclass correlation coefficients for measurements of circularity and mean gray values were all .0.9 (P , 0.001) (Table 3). The mean circularity of cystoid spaces in eyes with MacTel 2 was significantly different from that of eyes with MacTel 1, RVO, and DME (Table 4). Although the circularities of cystoid spaces in eyes with MacTel 1, RVO, and DME were .0.7, the circularity was ,0.5 in eyes with MacTel 2 in both B-scan and en face OCT images (Table 4). The mean gray value of cystoid spaces in eyes with MacTel 2 was significantly lower than that of eyes with MacTel 1 (P = 0.003 and P = 0.002) and DME (P , 0.001 and P , 0.001) in both B-scan and en face OCT images, respectively (Table 5). However, the mean gray value of cystoid spaces in eyes with MacTel 2 was not different from that of eyes with RVO in either the B-scan or en face OCT images (P = 0.162 and P = 0.114, respectively). The mean gray values of cystoid spaces were greater than those of the vitreous cavities in all groups (P , 0.001). Tomographic and Topographic Characteristics of Cystoid Spaces In B-scan images, the cystoid space was localized within the inner retina (Henle fiber layer) in all five
Table 2. Characteristics of Subjects With Cystoid Spaces
Total patients Mean age, years Men/women Diabetes Hypertension Enrolled eyes OD/OS BCVA, logMAR Snellen equivalent
MacTel 2
MacTel 1
RVO
DME
10 60.9 ± 10.2 2/8 2/10 (20%) 7/10 (70%) 16 10/6 0.3 ± 0.2 20/35
4 51.3 ± 8.6 4/0 0/4 (0%) 1/4 (25%) 4 2/2 0.2 ± 0.2 20/30
20 62.1 ± 11.5 8/12 0/20 (0%) 11/20 (55%) 20 10/10 0.8 ± 0.5 20/125
21 55.8 ± 10.2 8/13 21/21 (100%) 10/21 (47.6%) 21 10/11 0.6 ± 0.5 20/80
Results are presented as mean ± SD or number (percentage). BCVA, best-corrected visual acuity; logMAR, logarithm of the minimum angle of resolution.
1128 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2014 VOLUME 34 NUMBER 6
Table 3. Intraobserver and Interobserver Intraclass Correlation Coefficients for Measurements of the Circularity and Mean Gray Value of Cystoid Spaces in OCT Images Circularity
Intraobserver 1 Intraobserver 2 Interobserver
Mean Gray Value
ICC
95% CI
P
ICC
95% CI
P
0.947 0.935 0.912
0.910–0.969 0.888–0.962 0.854–0.948
,0.001 ,0.001 ,0.001
0.994 0.995 0.988
0.990–0.997 0.996–0.999 0.980–0.993
,0.001 ,0.001 ,0.001
CI, confidence interval; ICC, intraclass correlation coefficient.
eyes with only foveolar cystoid spaces. In 11 eyes with both foveolar and parafoveolar cystoid spaces, 4 foveolar cystoid spaces were localized within the inner retina, 4 foveolar cystoid spaces were localized within the inner retina and outer nuclear layer, and 3 foveolar cystoid spaces extended from the inner retina to the IS/OS. Nine parafoveolar cystoid spaces were localized within the inner retina, one parafoveolar cystoid space was localized within the inner retina and outer nuclear layer, and one parafoveolar cystoid space extended from the inner retina to the IS/OS. There were two Stage 1 cases with cystoid spaces. These 2 eyes had only foveolar cystoid spaces, and these were localized within the inner retina (Figure 4). The cystoid spaces identified in en face OCT images were matched with angiographic images. In the en face OCT images of eyes with MacTel 2, cystoid spaces were located in the foveal center or the parafoveal area. Among 16 eyes with MacTel 2, 11 eyes had both foveolar and parafoveolar cystoid spaces. In the remaining 5 eyes, the cystoid space was located only in the foveal center (Table 6). In 7 of 11 eyes with both foveolar and parafoveolar cystoid spaces, the parafoveolar cystoid spaces had a similar location to the angiographic leakage but were not filled with the dye. A different grouping of 7 of the 11 eyes with both foveolar and parafoveolar cystoid spaces had clusters of hyperreflective spots that were in a similar location to the parafoveolar cystoid space, except for 1 eye. All 11 eyes with foveolar and parafoveolar cystoid spaces had disruptions of IS/OS line. The disruptions of IS/OS line were in similar locations to the foveolar (1 eye), parafoveolar (3 eyes), or both cystoid spaces (6 eyes), except for 1 eye. Angiographic leakage was detected in all five eyes with only foveolar cystoid spaces. However, the
locations of cystoid spaces did not match with those of the angiographic leakages. Eleven of 24 eyes with MacTel 2 had clusters of hyperreflective spots. These clusters had similar locations to the retinal pigmentations, parafoveolar cystoid spaces, and/or sites of angiographic leakage. Six of 24 eyes with MacTel 2 had pigmentation. All six of these eyes had clusters of hyperreflective spots in similar locations to retinal pigmentations. Retinal pigmentations and clusters of hyperreflective spots were only encountered in the parafoveal area.
Discussion In this study, the cystoid spaces of eyes with MacTel 2 were less circular than the cystoid spaces of eyes with MacTel 1, RVO, and DME, which are caused by vascular leakage because of disruption of the inner BRB. These findings support the suggestion that cystoid spaces in eyes with MacTel 2 are caused by degenerative changes rather than vascular leakage. Because the suggestion that the primary abnormality may reside in parafoveolar retinal neural cells, Müller cells, or both,17 many studies have suggested that Müller cells are involved in the pathogenesis of MacTel 2.2,4,10,15 Unlike natural optical fibers, such as the IS/OS of photoreceptor cells, Müller cells have many complex side branches that enclose neuronal compartments.18 The degenerative involvement of adjacent Müller cells may be responsible for irregular cystoid space boundaries observed in eyes with MacTel 2. Recently, Ooto et al19 reported the cone abnormality differences between MacTel subtypes related to the different pathogenesis of these subtypes. In this study, we used a novel approach, the analysis of circularity, to
Table 4. Circularity of Cystoid Spaces in OCT Images
B-scan images En face images
MacTel 2
MacTel 1 (P*)
RVO (P*)
DME (P*)
0.43 ± 0.13 0.48 ± 0.09
0.78 ± 0.08 (0.004) 0.70 ± 0.08 (0.003)
0.74 ± 0.08 (,0.001) 0.73 ± 0.07 (,0.001)
0.75 ± 0.05 (,0.001) 0.77 ± 0.05 (,0.001)
Results are presented as mean ± SD. P values were calculated using the Mann–Whitney U test. *Compared with MacTel 2.
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL
1129
Table 5. Reflectivity of Cystoid Spaces in OCT Images
B-scan images Mean gray value Mean gray value Mean gray value En face images Mean gray value Mean gray value Mean gray value
MacTel 2
MacTel 1 (P*)
RVO (P*)
DME (P*)
of cystoid space of vitreous cavity ratio
11.5 ± 2.4 9.3 ± 1.4 1.2 ± 0.2
20.5 ± 3.3 (0.003) 9.8 ± 1.5 (0.832) 2.1 ± 0.1 (0.003)
12.7 ± 2.6 (0.162) 10.4 ± 1.3 (0.019) 1.2 ± 0.2 (0.675)
18.1 ± 6.3 (,0.001) 10.4 ± 1.2 (0.008) 1.8 ± 0.7 (0.017)
of cystoid space of vitreous cavity ratio
12.2 ± 1.9 10.6 ± 0.9 1.2 ± 0.2
21.3 ± 2.0 (0.002) 10.9 ± 0.7 (0.537) 1.9 ± 0.1 (0.002)
12.7 ± 2.2 (0.114) 11.5 ± 1.3 (0.006) 1.1 ± 0.3 (0.080)
20.1 ± 6.6 (,0.001) 11.3 ± 1.3 (0.029) 1.8 ± 0.7 (,0.001)
Results are presented as mean ± standard deviation. P values were calculated using the Mann–Whitney U test. *Compared with MacTel 2.
quantify the value of the shape of the cystoid spaces and demonstrate significant differences in the shape of cystoid space in eyes with MacTel 2 compared with MacTel 1 and other diseases with vascular exudation. Circularity is not measured in routine settings, and the shape of cystoid spaces can be qualitatively observed on OCT. However, the quantitative measurement used in this study would be included in the differential diagnosis of diseases with vascular exudation. The reflectivity of cystoid spaces in eyes with MacTel 2 was lower than the reflectivity in eyes with DME. The gray-scale analysis of images used in this study is a relatively insensitive measure that is, probably not as sensitive as analysis of the OCT waveforms, such as that performed by Barthelmes et al.20 However, our findings are consistent with their results.20 Although the cystoid spaces of eyes with RVO or DME are known to result from leakage, we found that the reflectivity of cystoid spaces in eyes with RVO was lower than that of eyes with DME (P , 0.001), but was not different from that of eyes with MacTel 2. It is not clear why there was
Fig. 4. Late-phase fluorescein angiographic and OCT images of an eye with MacTel 2 in Stage 1. Fluorescein angiography shows mild hyperfluorescence temporal to the foveola. The foveolar cystoid space is localized within the inner retina in the B-scan OCT image.
a difference in the reflectivity of cystoid spaces between eyes with RVO or DME. Differences in plasma constituents, pathophysiology, size of BRB breakdown, and/or hydrostatic pressure through the BRB may affect the reflectivity of cystoid spaces. To gain a better understanding, we topographically matched OCT findings on en face images to late-phase angiographic images.21 In this study, there was topographic discordance between cystoid space and angiographic leakage, as in previous study.5 There were no foveolar cystoid spaces that had similar locations to sites of vascular leakage. This may be because of the geographic characteristics of foveal vasculature, that is, the foveal avascular zone. Some parafoveolar cystoid spaces had a similar location to the angiographic leakage but were not filled with the dye. Baumüller et al6 reported that hyperreflective spots in the outer retina were found in all stages of the disease. Because of the relatively low resolution of our OCT images and difficulty in differentiating them from noise, we did not identify hyperreflective spots as they did in the study by Baumüller et al,6 but we analyzed clustering spots in OCT images. Except for six eyes with retinal pigmentation, the clusters of hyperreflective spots were similar in location to the sites of angiographic leakage. This finding suggests that some hyperreflective spots may be related to vascular abnormalities or extravasated deposits, as proposed by Baumüller et al.6 No eyes had only parafoveolar cystoid spaces. The five eyes with only foveolar cystoid spaces did not show clusters of hyperreflective spots, disruptions of IS/OS line, or retinal pigmentations. This finding suggests that cystoid spaces may occur in the foveolar area first, and that foveolar-only cystoid spaces may be early findings. Inner and outer segment disruptions, retinal pigmentation, and clusters of hyperreflective spots may be secondary to cystoid spaces or may be late findings of MacTel 2. Cystoid spaces may be related to the loss of photoreceptor cells and/or Müller cell processes, because they are only compositions in the foveolar area.22
1130 RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES
2014 VOLUME 34 NUMBER 6
Table 6. Cystoid Spaces and Other Abnormalities in 16 Eyes With MacTel 2 Cystoid Spaces Eyes with both foveolar and parafoveolar cystoid spaces (n = 11 eyes) Eyes with only foveolar cystoid spaces (n = 5 eyes)
Cluster of Disruptions of Angiographic Leakage Hyperreflective Spots IS/OS Line Retinal Pigmentations 11 eyes
7 eyes
11 eyes
3 eyes
5 eyes
None
None
None
The IS/OS line represents the junction of the inner and outer photoreceptor segments.
There were several limitations to this study. En face images are not readily available in every OCT device. Therefore, the circularity of cystoid spaces on B-scan images was also measured. The results were not different from those of en face images. The diagnosis of MacTel 2 was not confirmed with other imaging techniques. More recently, it has been recognized that focal loss of macular pigments, consisting of lutein and zeaxanthin on fundus autofluorescence, is a more sensitive and specific marker.23–26 This study has other limitations such as a retrospective design and few cases. In this study, only images with the largest cystoid spaces were used for comparison. Further comparison, including cystoid spaces localized in different layers, would be interesting because the anatomical differences between different retinal layers or between the fovea and parafovea may influence the shape of cystoid spaces. However, the limited number of MacTel 2 cases prevented further analysis. Using the B-scan OCT images, we showed which retinal layers were occupied by the cystoid spaces in eyes with MacTel 2. However, it is difficult to draw conclusions from the relatively low number of patients included in this study. Klein et al27 reported that the prevalence of MacTel 2 was higher than previously estimated, but MacTel 2 is still rare.28 Large-scale or multicenter studies are required to increase the number of cases that can be analyzed. Another limitation of our study is that the images were analyzed manually. However, intraclass correlation coefficients for the circularity and mean gray value measurements of cystoid spaces were all .0.9. In conclusion, the morphologic characteristics of cystoid spaces in eyes with MacTel 2 were different from those in eyes with MacTel 1, RVO, and DME, which are caused by vascular leakage because of disruption of the inner BRB. Cystoid spaces in eyes with MacTel 2 have more irregular boundaries than those in eyes with MacTel 1, RVO, and DME, and lower reflectivity than those in eyes with MacTel 1 and DME. Our results are consistent with the observations
of previous studies, which have suggested that MacTel 2 is related to degenerative changes. Key words: macular telangiectasia, en face images, optical coherence tomography, intraretinal cystoid spaces. References 1. Gass JD, Blodi BA. Idiopathic juxtafoveolar retinal telangiectasis. Update of classification and follow-up study. Ophthalmology 1993;100:1536–1546. 2. Yannuzzi LA, Bardal AM, Freund KB, et al. Idiopathic macular telangiectasia. Arch Ophthalmol 2006;124:450–460. 3. Gass JD, Oyakawa RT. Idiopathic juxtafoveolar retinal telangiectasis. Arch Ophthalmol 1982;100:769–780. 4. Issa PC, Gillies MC, Chew EY, et al. Macular telangiectasia type 2. Prog Retin Eye Res 2013;34:49–77. 5. Koizumi H, Iida T, Maruko I. Morphologic features of group 2A idiopathic juxtafoveolar retinal telangiectasis in threedimensional optical coherence tomography. Am J Ophthalmol 2006;142:340–343. 6. Baumüller S, Charbel Issa P, Scholl HP, et al. Outer retinal hyperreflective spots on spectral-domain optical coherence tomography in macular telangiectasia type 2. Ophthalmology 2010;117:2162–2168. 7. Gupta V, Gupta A, Dogra MR, Agarwal A. Optical coherence tomography in group 2A idiopathic juxtafoveolar telangiectasis. Ophthalmic Surg Lasers Imaging 2005;36:482–486. 8. Gaudric A, Ducos de Lahitte G, Cohen SY, et al. Optical coherence tomography in group 2A idiopathic juxtafoveolar retinal telangiectasis. Arch Ophthalmol 2006;124:1410– 1419. 9. Maruko I, Iida T, Sekiryu T, Fujiwara T. Early morphological changes and functional abnormalities in group 2A idiopathic juxtafoveolar retinal telangiectasis using spectral domain optical coherence tomography and microperimetry. Br J Ophthalmol 2008;92:1488–1491. 10. Sallo FB, Leung I, Chung M, et al. Retinal crystals in type 2 idiopathic macular telangiectasia. Ophthalmology 2011;118: 2461–2467. 11. Surguch V, Gamulescu MA, Gabel VP. Optical coherence tomography findings in idiopathic juxtafoveal retinal telangiectasis. Graefes Arch Clin Exp Ophthalmol 2007;245: 783–788. 12. Paunescu LA, Ko TH, Duker JS, et al. Idiopathic juxtafoveal retinal telangiectasis: new findings by ultrahigh-resolution optical coherence tomography. Ophthalmology 2006;113:48–57.
CYSTOID SPACES IN TYPE 2 MACTEL OH ET AL 13. Sallo FB, Peto T, Egan C, et al. The IS/OS junction layer in the natural history of type 2 idiopathic macular telangiectasia. Invest Ophthalmol Vis Sci 2012;53:7889–7895. 14. Sallo FB, Peto T, Egan C, et al. “En face” OCT imaging of the IS/OS junction line in type 2 idiopathic macular telangiectasia. Invest Ophthalmol Vis Sci 2012;53:6145–6152. 15. Powner MB, Gillies MC, Tretiach M, et al. Perifoveal müller cell depletion in a case of macular telangiectasia type 2. Ophthalmology 2010;117:2407–2416. 16. Blott SJ, Pye K. Particle shape: a review and new methods of characterization and classification. Sedimentology 2008;55: 31–63. 17. Gass JD. Histopathologic study of presumed parafoveal telangiectasis. Retina 2000;20:226–227. 18. Franze K, Grosche J, Skatchkov SN, et al. Müller cells are living optical fibers in the vertebrate retina. Proc Natl Acad Sci U S A 2007;104:8287–8292. 19. Ooto S, Hangai M, Takayama K, et al. Comparison of cone pathologic changes in idiopathic macular telangiectasia types 1 and 2 using adaptive optics scanning laser ophthalmoscopy. Am J Ophthalmol 2013;155:1045– 1057. 20. Barthelmes D, Sutter FK, Gillies MC. Differential optical densities of intraretinal spaces. Invest Ophthalmol Vis Sci 2008; 49:3529–3534.
1131
21. Wanek J, Zelkha R, Lim JI, Shahidi M. Feasibility of a method for en face imaging of photoreceptor cell integrity. Am J Ophthalmol 2011;152:807–814. 22. Massey SC. Functional anatomy of the mammalian retina. In: Ryan SJ, ed. Retina. Philadelphia, PA: Elsevier/Mosby; 2006: 43–82. 23. Charbel Issa P, van der Veen RL, Stijfs A, et al. Quantification of reduced macular pigment optical density in the central retina in macular telangiectasia type 2. Exp Eye Res 2009;89:25–31. 24. Zeimer MB, Padge B, Heimes B, Pauleikhoff D. Idiopathic macular telangiectasia type 2: distribution of macular pigment and functional investigations. Retina 2010;30:586–595. 25. Helb HM, Charbel Issa P, van der Veen RL, et al. Abnormal macular pigment distribution in type 2 idiopathic macular telangiectasia. Retina 2008;28:808–816. 26. Chin EK, Kim DY, Hunter AA III, et al. Staging of macular telangiectasia: power-Doppler optical coherence tomography and macular pigment optical density. Invest Ophthalmol Vis Sci 2013;54:4459–4470. 27. Klein R, Blodi BA, Meuer SM, et al. The prevalence of macular telangiectasia type 2 in the Beaver Dam eye study. Am J Ophthalmol 2010;150:55–62. 28. Aung KZ, Wickremasinghe SS, Makeyeva G, et al. The prevalence estimates of macular telangiectasia type 2: the Melbourne Collaborative Cohort Study. Retina 2010;30:473–478.
