To study peripapillary morphologic changes in highly myopic eyes using swept-source optical coherence tomography at a longer wavelength.
Prospective cross-sectional study.
Peripapillary regions of 196 eyes of 107 patients with high myopia (refractive error, <−8.0 diopters or axial length, >26.0 mm) were analyzed quantitatively and qualitatively with an swept-source optical coherence tomography prototype system that uses a tunable laser light source operated at a 100,000-Hz A-scan repetition rate in the 1-μm wavelength region. The visual field was evaluated by standard automated perimetry. Area of peripapillary atrophy β and presence of scleral protrusion temporal to the optic disc were assessed.
Peripapillary atrophy β area, but not disc area, was significantly larger in eyes with visual field defect (3.16 ± 2.70 mm 2 ; range, 0.00 to 12.85 mm 2 ) than those without visual field defect (2.31 ± 2.83 mm 2 ; range, 0.00 to 17.70 mm 2 ). Temporal scleral protrusion was detected by color stereo disc photography in 22 (19.5%) of 113 eyes with visual field defect and in 4 (4.8%) of 83 eyes without visual field defect. Scleral bending demonstrated a wide range of angles (mean, 31.0 ± 21.1 degrees; range, 2 to 80 degrees). The angle of scleral bending, but not the distances from scleral bend to disc margin or foveal center, correlated significantly with retinal nerve fiber layer thickness above the bend ( r = −0.557, P = .007) and visual field defect severity ( r = −0.445, P = .038).
Swept-source optical coherence tomography visualizes peripapillary deep structures in high myopia. Some cases of high myopia may be affected by direct scleral compression or stretching at the peripapillary region.
Myopia long has been identified as a risk factor for open-angle glaucoma and is a serious public health concern that continues to increase in prevalence, especially in certain young Asian populations. Myopia has been associated with a high risk of visual impairment caused by macular diseases such as chorioretinal atrophy, choroidal neovascularization, macular retinoschisis, macular hole, and retinal detachment. Recently, a high incidence of visual field defects that could not be explained by myopic macular lesions was reported in patients with pathologic myopia. The pathogenesis of these visual field defects has been poorly understood to date. Further investigations are needed to determine whether these visual field defects are the result of high myopia alone or overlap with glaucoma.
Highly myopic eyes demonstrate various types of posterior staphyloma. Recently, magnetic resonance imaging revealed that eyes with high myopia demonstrate highly deformed shapes. Peripapillary structural changes in eyes with high myopia have been the focus of several studies, and changes in peripapillary scleral curvature identified on color fundus photographs were reported to relate to progression of visual field defects. Further detailed investigations focused on peripapillary scleral deformation may elucidate the mechanisms underlying the presence and progression of visual field defects that cannot be explained by myopic macular lesions in high myopia.
Conventional spectral-domain optical coherence tomography (SD OCT) was developed using Fourier-domain detection technology and enables high-axial resolution imaging around the optic disc, but its depth penetration is limited because of absorption and scattering at the retinal epithelial membrane with a probe light operated at approximately 800 nm. Enhanced depth imaging (EDI) was developed for SD OCT to improve image quality of the deep structures of the posterior segment, and this technique is useful for evaluating deep structures of the optic nerve complex. However, although EDI is an effective method for visualizing the deep structures of the optic disc, it is disadvantageous for observing axially extended structures in highly myopic eyes in their entirety because its signal intensity decays with axial distance.
Swept-source (SS) OCT is a next-generation Fourier-domain OCT that demonstrates less signal decay over depth compared with the current SD OCT. Our prototype SS OCT device uses an SS probe light with a center wavelength of 1040 to 1060 nm, which allows high-penetration imaging of deep retinal tissues such as the choroid and sclera. SS OCT enables improved visualization of the deep structure of the human optic disc. Compared with SD OCT, SS OCT is characterized by a higher-speed scan rate and relatively lower sensitivity roll-off versus depth; therefore, SS OCT along with a longer wavelength probe light may be advantageous for the investigation of axially extended peripapillary scleral structures in highly myopic eyes.
In this study, we quantitatively investigated characteristics of peripapillary lesions, including peripapillary atrophy (PPA)-β and peripapillary scleral protrusion, in highly myopic eyes. These biometric parameters were correlated to visual field defects and neighboring retinal nerve fiber layer (RNFL) thickness using a prototype SS OCT device at a longer wavelength.
This prospective, cross-sectional study was carried out with the approval of the Institutional Review Board and Ethics Committee of Kyoto University Graduate School of Medicine and in adherence to the tenets of the Declaration of Helsinki. Written informed consent for the research was obtained from the patients after providing an explanation of the nature and possible consequences of the study.
Patients with high myopia were recruited from Kyoto University Hospital from July 2010 through July 2011. To be included, patients were required to demonstrate a refractive error (spherical equivalent) of less than −8.00 diopters (D) or axial length of more than 26.0 mm and to agree to undergo examinations using a prototype SS OCT instrument (Topcon Corp, Tokyo, Japan). Only phakic eyes were included for evaluating the mean refractive errors to avoid the effects of intraocular lens implantation. Exclusion criteria included evidence of history of optic neuritis or other neuro-ophthalmologic diseases and history of vitreoretinal surgery, uveitis, or any type of myopic macular or peripheral pathologic features that could cause visual field defects. Study subjects were assigned to 2 groups according to the presence or absence of visual field defects.
Study participants underwent a comprehensive ophthalmic examination, including measurement of uncorrected and best-corrected visual acuity with the 5-m Landolt chart, slit-lamp examination, intraocular pressure (IOP) measurement with a Goldmann applanation tonometer, gonioscopy, axial length measurement with an IOLMaster biometer (Carl Zeiss Meditec Inc, Dublin, California, USA), dilated stereoscopic examination of the fundus, stereo disc photography (3-Dx simultaneous stereo disc camera; Nidek, Gamagori, Japan), color fundus photography, SD OCT examination (Spectralis HRA+OCT; Heidelberg Engineering, Dossenheim, Germany), Heidelberg Retina Tomograph II analysis (Heidelberg Engineering), and visual field testing. All IOP measurements were collected between 9:00 am and 3:00 pm .
Visual field testing was performed with optical correction by contact lenses or trial lenses by standard automated perimetry (SAP) using the Humphrey Visual Field Analyzer with the 24-2 Swedish interactive threshold algorithm standard (Carl Zeiss Meditec, Inc.). To determine visual field defects, patients had to present corresponding abnormalities in 2 reliable visual field tests (≤20% fixation loss, ≤15% false-positive results, and ≤33% false-negative results). An abnormal visual field result was defined as abnormal range in a glaucoma hemifield test or the presence of at least 3 contiguous test points within the same hemifield on the pattern deviation plot at P < .05, with at least 1 point at P < .01, on at least 2 consecutive visual field tests.
Assessment of the Optic Disc and Peripapillary Lesions
Optic disc and PPA-β areas were measured using the Heidelberg Retina Tomograph II. The presence of an abrupt change of scleral curvature temporal to the optic disc on biomicroscopy was determined based on complete agreement of 2 observers (T.A. and A.N.).
Swept-Source Optical Coherence Tomography at 1050 nm
We used a prototype SS OCT system originally developed by Topcon Corp. (Tokyo, Japan) with an axial scan rate of 100 000 Hz, operated as reported previously. This SS OCT system uses a wavelength-sweeping laser with an approximately 100-nm tuning range centered at 1050 nm, yielding 8-μm axial resolution in tissue. Transverse resolution was set to approximately 20 μm. A single OCT image consisted of 1024 A lines, which can be acquired in 10 ms. SS OCT imaging at 1050 nm was conducted with approximately 1-mW light intensity on the cornea, which is well below the safe retinal exposure limit established by the American National Standards Institute. Sensitivity was approximately 98 dB at this input power.
All SS OCT examinations were performed through a dilated pupil by trained examiners. Six radial B-scan images through the center of the optic disc were obtained for each eye. Line scans passing through the foveal center and the center of the optic disc also were obtained, centered at the halfway point between them. Speckle noises were decreased by averaging 50 B-scans obtained at an identical location of interest to improve image quality.
Measurements Obtained Using 1050-nm Swept-Source Optical Coherence Tomography
The angle of scleral bending at the edge of scleral bending was measured on SS OCT images using ImageJ (National Institutes of Health, Bethesda, Maryland). First, SS OCT images were scaled to an actual ratio between axial and lateral length. Then, 2 tangent lines were drawn along the surface of the sclera on both sides of the scleral bending, and the angle between the 2 lines was measured as the angle of scleral bending. RNFL thickness at the edge of scleral bending, scleral thickness at the edge of scleral bending and beneath the central fovea, distance from the bending angle to the edge of the optic disc and foveal center, and distance from the edge of the optic disc to the foveal center were measured using the built-in software of the prototype SS OCT device produced by Topcon. RNFL thickness and scleral thickness at the edge of scleral bending were measured on B-scans that passed through the centers of the optic disc and fovea along the axial line that passed the intersection of the 2 tangent lines. In the current study, temporal scleral bending was located inside the PPA in all eyes, and the retina above the scleral bending contained only RNFL. Therefore, the RNFL thickness was measured as the distance between the vitreoretinal interface and the anterior boundary of the scleral bend along the axial line. Scleral thickness beneath the central fovea was determined along the vertical line of the SS OCT image that passed through the foveal center. In this software, the lateral length measurement was corrected based on Littmann’s formula.
Interobserver and Intraobserver Reproducibility
To evaluate the interobserver reproducibility of our method, 2 authors (T.A. and Y.K.) blinded to clinical information independently measured the angle of scleral bending and RNFL thickness and scleral thickness at the bending angle; intraclass correlation coefficients (ICCs [2,1]) were calculated for the 3 parameters. To evaluate the intraobserver reproducibility and variability in SS OCT measurements, the 3 parameters were evaluated in triplicate by 1 author (T.A.) to calculate ICC (1,1) and coefficient of variation.
All statistical evaluations were performed using commercially available software (SPSS software version 20; International Business Machines Corp., Armonk, New York, USA). The 2-sample t test or chi-square test was used to compare mean age, visual field mean deviation (MD), axial length, refraction, IOP, and the presence of temporal protrusion between groups. To determine the ocular characteristics associated with temporal scleral protrusion and to examine the relationship between the angle of scleral bending and ocular parameters, the Pearson correlation coefficient and multiple logistic regression analysis were used. Multiple logistic regression analysis was applied to correlate age, IOPs, and existence of temporal protrusion with the presence of visual field defects. Multiple linear regression analysis was used to evaluate the correlation between RNFL thickness or MD value and angle of scleral bending, age, and axial length. P values less than .05 were considered statistically significant; the data are presented as the mean ± standard deviation.
One hundred seven patients underwent SS OCT examinations. Five eyes were excluded because of difficulties in performing the OCT examinations, including poor visual fixation or convergent strabismus fixus. Twelve patients had unilateral high myopia, and their fellow eyes did not meet the criteria for high myopia and were excluded. One eye was excluded because reliable visual field results could not be obtained. Finally, the remaining 196 eyes (107 patients) with high myopia were evaluated. The cohort included 52 men and 56 women with mean age of 46.5 ± 14.3 years (range, 21 to 80 years). The mean refractive error was −9.79 ± 3.44 diopters (D; range, −23.25 to −4.25 D), mean axial length was 27.65 ± 1.28 mm (range, 24.39 to 32.58 mm), and IOP at the initial visit was 17.0 ± 4.3 mm Hg (range, 8 to 40 mm Hg).
Advantages of Imaging With 1050-nm Swept-Source Optical Coherence Tomography
First, we compared visualization of the peripapillary structures among B-scan images obtained using standard Spectralis, Spectralis EDI OCT, and our prototype SS OCT in highly myopic eyes. Visualization of the sclera, particularly its posterior boundary, was not clear on standard SD OCT images ( Figure 1 , Top middle). EDI OCT images improved visualization of the posterior boundary of the sclera in the inferior half of the imaging frame, but demonstrated a marked signal intensity decrease in the superior half ( Figure 1 , Top right). Such poor visualization in the inferior half of the standard Spectralis imaging frame and superior half of the Spectralis EDI imaging frame was observed in all eyes. This signal decay was very low in all eyes using SS OCT; SS OCT provided well-defined images of the neural retina, sclera, lamina cribrosa, and structures beneath the sclera, such as orbital fat and subarachnoid spaces with subarachnoid trabeculae, from the top to bottom of the imaging frame ( Figure 1 , Bottom left and right).
Imaging of Peripapillary Sclera With 1050-nm Swept-Source Optical Coherence Tomography
On biomicroscopy, ridge-like anterior protrusion of the sclera temporal to the optic disc was found in 26 (13.3%) of 196 study eyes. SS OCT images in 22 (84.6%) of 26 eyes with scleral protrusion showed abrupt scleral bending, with the sclera on both sides almost straight ( Figures 2 and 3 ), whereas those in the 4 (15.4%) remaining eyes showed gradual curvature with a rounded angle edge (2 eyes) or more than 2 peaks of scleral protrusion (2 eyes) ( Figure 2 , Top row near left) in the region corresponding to the scleral protrusion on biomicroscopy. Such abnormal scleral bending or round curvature on SS OCT images was not found in the region nasal or superior to the optic disc in any eyes. Two of 26 eyes with temporal scleral bending showed abnormal scleral bending from the temporal to inferior regions of the optic disc. The degree of scleral bending varied; sharp angles with the sclera were observed in some cases ( Figure 2 , Top row near left), and obtuse angles were observed in others ( Figure 2 , Second row near left). The sclera nasal to the protrusion sloped steeply and ended in the scleral ring, resulting in a large difference in height between the top of the protrusion and the scleral ring. By contrast, the sclera temporal to the protrusion varied in slope from steep to almost flat. The neural retina over the scleral bending demonstrated variable thinning, extremely diminished in some eyes ( Figure 2 , Top row near left and Third row near left) and thick in others ( Figure 2 , Second row near left).
In eyes with scleral bending, other unusual peripapillary features were seen on SS OCT images. In 14 (53.8%) of 26 eyes with scleral bending, highly reflective tissue was seen just beneath the edge of the scleral bending ( Figure 2 ). This structure had high reflectivity similar to that of the sclera, but was separated from the overlying sclera. Retinal detachment, outer retinal splitting, or both were observed inside PPA-β regions in 4 (2.0%) of 196 highly myopic eyes ( Figure 2 , Second row near left and near right).
Comparison of Peripapillary Structures Between Eyes With and Without Visual Field Defects
We compared demographics, IOP, and peripapillary lesions between eyes with and without visual field defects ( Table 1 ). A significant difference in age, but not sex, refraction, or axial length, was found between groups. Eyes with visual field defects had higher initial and maximum IOPs, and statistical differences were found only for maximum IOPs ( Table 1 ).
|Characteristic||With Visual Field Defects (n = 113)||Without Visual Field Defects (n = 83)||P Value|
|Age (y)||51.3 ± 13.5 (23 to 80)||39.9 ± 12.9 (21 to 74)||<.001 a|
|Gender (female/male)||59/54||42/41||.824 b|
|Refraction (diopters)||−9.92 ± 3.48 (−20.75 to −5.25)||−9.60 ± 2.76 (−16.75 to −4.25)||.498|
|Axial length (mm)||27.72 ± 1.38 (24.39 to 32.58)||27.51 ± 1.02 (25.99 to 29.99)||.246|
|IOP, first visit (mm Hg)||17.4 ± 5.1 (8 to 40)||16.5 ± 2.7 (9 to 25)||.091|
|IOP, maximum (mm Hg)||18.8 ± 5.6 (11 to 43)||16.9 ± 2.8 (11 to 25)||.002 a|
|MD (dB)||−11.0 ± 8.5 (0.04 to −32.44)||−1.5 ± 1.3 (0.75 to −4.91)||<.001 a|
|Papillary and peripapillary lesions|
|Disc area (mm 2 )||2.33 ± 1.15 (0.51 to 9.40)||2.37 ± 0.69 (1.00 to 5.22)||.810|
|PPA-β area (mm 2 )||3.16 ± 2.70 (0.00 to 12.85)||2.31 ± 2.83 (0.00 to 17.70)||.036 a|
|Temporal scleral protrusion, n (%)||22 (19.5%)||4 (4.8%)||.003 a , b|
On biomicroscopy, we found PPA-β in 188 (96.4%) of 196 eyes; the incidence of PPA-β was comparable between eyes with (110/113 [97.3%]) and without (78/83 [94.0%]) visual field defects ( P = .207, chi-square test). However, PPA-β area in eyes with visual field defects was significantly larger than that in eyes without visual field defects, whereas disc area was almost equivalent between groups ( Table 1 ). On biomicroscopy, scleral protrusion in a vertical ridge temporal to the optic disc was found in 26 (13.3%) of 196 eyes. This feature was significantly more frequent in eyes with visual field defects (19.5%) than in eyes without visual field defects (4.8%; Table 1 ). Multiple logistic regression analysis performed to correlate age, IOP, and the existence of temporal protrusion with the presence of visual field defects showed that age and maximum IOP, but not the existence of temporal protrusion, significantly differed between the 2 groups (age, P < .001; maximum IOP, P = .026; temporal protrusion, P = .144).
Comparison Between Eyes With and Without Temporal Scleral Protrusion
Significantly older age and a higher frequency of women were observed among patients with temporal scleral protrusion on biomicroscopy ( Table 2 ). Eyes with temporal scleral protrusion demonstrated poorer MD values, lesser spherical equivalent refractive errors, and longer axial lengths than eyes without temporal scleral protrusion ( Table 2 ).