Purpose
To determine the incidence of a continuous cone outer segment tips (COST) line at the fovea in spectral-domain optical coherence tomographic (SD-OCT) images of normal eyes.
Design
Prospective, interventional case series.
Methods
Forty-six right eyes of 46 normal individuals with visual acuities (VA) ≥20/20 were studied. SD-OCT images were obtained with a Cirrus HD-OCT instrument with both the standard 5-line raster and the high-definition (HD) 5-line raster scan modes. Images with signal strengths weaker than 5 (on a scale from 0 to 10) were excluded. The appearances of the COST line, photoreceptor inner segment/outer segment (IS/OS) junction, and external limiting membrane (ELM) line were determined in a masked way.
Results
The incidence of eyes with an intact foveal COST line was about 95%, and the incidence was not significantly associated with the age, sex, refractive error, signal strength, horizontal or vertical scans, and the use of either the standard or HD scans. Fragmented COST lines appeared to be attributable to blocking artifacts caused by a hyperreflectivity at the foveal surface because the COST fragmentation had corresponding fragmentation of the ELM and IS/OS junction lines. The correlation between a fragmented COST line and the hyperreflectivity on the retinal surface was significant for the vertical HD scans ( P = .011) but not for the vertical standard, horizontal standard, or horizontal HD scans.
Conclusions
Commercial SD-OCT instruments can detect the COST line. Fragmentations of the COST lines in normal eyes are most likely artifacts caused by a parabolic reflection of a surface hyperreflectivity of the foveal pit.
Optical coherence tomography (OCT) is a noninvasive imaging method that can provide cross-sectional images of the morphology of the retina with high resolution in real time. Quantitative and qualitative evaluations of the retinal microstructures are becoming a common diagnostic procedure in ophthalmology. Spectral-domain OCT (SD-OCT), which has better axial resolution than the conventional time-domain OCT (TD-OCT), has been used to evaluate the microstructures of the photoreceptors in eyes with different retinal pathologies. The restoration of the photoreceptor inner segment/outer segment (IS/OS) junction has been shown to be significantly correlated with the recovery of the visual acuity after surgery. It was suggested that the presence of a continuous IS/OS line was a sign of well-restored photoreceptor cells, and a continuous external limiting membrane (ELM) was a sign of intact photoreceptor cell bodies in the outer nuclear layer and Müller cells.
Srinivasan and associates used high-speed, ultrahigh-resolution OCT to prove that the bright reflecting line located between the IS/OS junction and the retinal pigment epithelium in the ultrahigh-resolution OCT images was the boundary of the cone outer segment tips (COST). This highly reflective line has also been called the Verhoeff membrane; however, histopathologic correlation has been under investigation.
The relationship between the visual acuity and the integrity of the foveal COST line detected by commercial SD-OCTs has been reported in cases of occult macular dystrophy. In addition, we recently reported a significant correlation between the visual acuity and the integrity of foveal COST line after successful macular hole surgery. The recovery of the COST line was observed only in eyes with an intact IS/OS junction and ELM line. Eyes with postoperative visual acuity ≥20/25 at postoperative 12 months were more frequently associated with the eyes with a recovered COST line.
The microstructures of the outer retina detected by commercial OCT instruments in normal subjects have not been well evaluated. We believe that it is essential to determine whether these commercial SD-OCT instruments can detect the foveal microstructures of normal subjects before applying these techniques to study eyes with vitreoretinal pathologies.
Thus, the aim of this study was to determine the incidence of a continuous COST line at the fovea of normal eyes using a commercial SD-OCT instrument. In addition, we determined which parameters of the SD-OCT would be best to use to obtain good-quality images of the COST line. The possible reasons why abnormal COST lines, IS/OS junctions, and ELM signals were present in the OCT images were also investigated.
Patients and Methods
Forty-six eyes of 46 healthy volunteers (19 men and 27 women) without any eye diseases were studied. The right eye of each subject was chosen arbitrarily to study. Each subject had a routine eye examination including measurements of the best-corrected visual acuity (BCVA) and intraocular pressure. All of the subjects had a BCVA of ≥20/20.
SD-OCT images were obtained by 6-mm scans with the Cirrus HD-OCT instrument (OCT 4000, Carl Zeiss Meditec, Dublin, California, USA) with both the standard 5-line raster scan (standard scan) and the high-definition (HD) 5-line raster scan (HD scan). We recorded both 6-mm horizontal and vertical scans to evaluate the integrity of the COST line, the ELM, and the IS/OS junction in the foveal region. The images were magnified with the built-in zoom function to evaluate the foveal microstructures. Three experienced investigators (T.R., Y.I., M.I.) who were masked to the subjects’ ophthalmologic data, including age and refractive errors, evaluated the SD-OCT images. Images with signal strength weaker than 5 (on a scale from 0 to 10) were excluded. The demographic data of the subjects, refractive errors (spherical equivalent) of the eyes, and signal strengths of the OCT images were recorded.
For statistical analysis, Student t tests, Welch t tests, Kruskal-Wallis tests, and 1-way analysis of variance (ANOVA) were used. A P value of .05 was set as significant.
Results
The mean age of the subjects was 33.0 years with a range from 23 to 50 years. The mean decimal BCVA was 1.1 with a range from 1.0 to 1.5. The mean refractive error was −1.49 diopters (D) with a range from −8.0 to +3.25 D. The intraocular pressure ranged from 12 to 18 mm Hg. The mean signal strengths were 8.2 (horizontal standard scan), 8.3 (vertical standard scan), 8.3 (horizontal HD scan), and 8.3 (vertical HD scan); the range of signal strength was from 6 to 10.
In the SD-OCT images, the COST line was recognized as a continuous line located between the IS/OS line and the retinal pigment epithelium layer ( Figure 1 ) . The COST line at the fovea was classified as continuous, fragmented, or absent. In 2 eyes where the COST line was absent and 20 eyes where a continuous COST line was detected, the OCT examination was repeated at different times of the day and on different days in a masked way. In all cases, the findings were the same in all eyes, and the repeatability of detecting a COST line was confirmed. In the 2 eyes with an absent COST line, the foveal COST line was absent for up to 6 months.
The incidence of detecting a COST line (continuous COST and fragmented COST) was 43 of 45 eyes (96%) in the horizontal standard scans, 41 of 43 eyes (95%) in the vertical standard scans, 44 of 46 eyes (96%) in the horizontal HD scans, and 41 of 43 eyes (95%) in the vertical HD scans ( Table 1 ). The range of detecting a continuous COST line was between 88% and 96% and a continuous COST line, but with fragmentation at the fovea (fragmented COST line), was between 0% and 7%. The differences in the incidences of a continuous COST line between the horizontal and vertical scans were not significant in either the standard or the HD scans ( P = .429, P = .475, respectively, Fisher exact probability test). The differences in the incidences of detecting a continuous COST line between the standard scans and the HD scans were also not significant in either the horizontal or the vertical scans ( P > .999, P > .999, respectively). The 2 eyes with an absent COST line were the same eyes in horizontal or vertical scan and standard or HD scan. One of these 2 eyes was that of a 31-year-old woman and the other that of a 34-year-old man. The refractive errors were −2.75 D and −3.75 D and the axial lengths measured by OA1000 (TOMEY Corp, Nagoya, Japan) were 26.1 mm and 26.0 mm. The age, sex, signal strengths, and refractive errors were not significantly different among the eyes with a continuous, fragmented, or absent COST line ( Table 1 ).
| Standard Scan (n = 45) | HD Scan (n = 46) | |||||||
|---|---|---|---|---|---|---|---|---|
| Continuous COST Line | Fragmented COST Line | Absent COST Line | P Value | Continuous COST Line | Fragmented COST Line | Absent COST Line | P Value | |
| Horizontal scan | ||||||||
| No. (%) eyes | 43 (96%) | 0 | 2 (4%) | n/a | 43 (93%) | 1 (2%) | 2 (4%) | n/a |
| Age (mean, y) | 33.3 | n/a | 32.5 | .882 a | 33.1 | 41.0 | 32.5 | .503 b |
| Sex (M:F) | 18:25 | n/a | 1:1 | >.999 c | 18:25 | 0:1 | 1:1 | >.999 c |
| Signal strength (mean) | 8.3 | n/a | 7.0 | .130 a | 8.4 | 9.0 | 7.0 | .225 b |
| Refractive error (mean, diopters) | −1.44 | n/a | −3.25 | .335 a | −1.45 | 0.00 | −3.25 | .262 b |
| Standard Scan (n = 43) | HD Scan (n = 43) | |||||||
|---|---|---|---|---|---|---|---|---|
| Continuous COST Line | Fragmented COST Line | Absent COST Line | P Value | Continuous COST Line | Fragmented COST Line | Absent COST Line | P Value | |
| Vertical scan | ||||||||
| No. (%) eyes | 39 (91%) | 2 (5%) | 2 (5%) | n/a | 38 (88%) | 3 (7%) | 2 (5%) | n/a |
| Age (mean, y) | 34.0 | 25.0 | 32.5 | .144 b | 33.1 | 37.3 | 32.5 | .718 b |
| Sex (M:F) | 16:23 | 0:2 | 1:1 | .761 c | 17:21 | 0:3 | 1:1 | .376 c |
| Signal strength (mean) | 8.3 | 8.5 | 7.5 | .451 b | 8.3 | 8.7 | 8.0 | .724 b |
| Refractive error (mean, diopters) | −1.51 | −0.75 | −3.25 | .369 b | −1.60 | +0.17 | −3.25 | .217 b |
b Kruskal-Wallis 1-way analysis of variance.
c Freeman-Halton extension of the Fisher exact probability test.
In all of the eyes with a fragmented COST line, a hyperreflective signal was seen at the surface of the foveal pit with weaker signals behind. These signals in both the standard and HD scans resembled the “acoustic shadow” artifacts that are observed in ultrasound echography ( Figure 2 ) . The signal intensities of the ELM, IS/OS junction, and COST lines at the fovea behind the hyperreflectivity were weaker and the lines were clearly fragmented. The diameter of the weaker signals behind the hyperreflectivity was wider in the lower (scleral) direction.
A fragmented COST line was always associated with the hyperreflectivity at the foveal pit surface (horizontal HD scans, P = .227; vertical standard scans, P = .055; vertical HD scans, P = .011, Fisher exact probability test, Table 2 ) and fragmented signals of the ELM (horizontal HD scans, P = .098; vertical standard scans, P = .001; vertical HD scans, P < .001) and IS/OS junction (horizontal HD scans, P = .068; vertical standard scans, P = .001; vertical HD scans, P < .001). A fragmented IS/OS junction line was also found in eyes with hyperreflectivity at the foveal surface (horizontal standard scans, P = .017; horizontal HD scans, P = .009; vertical standard scans, P = .055; vertical HD scans, P = .011) and fragmented ELM line (horizontal standard scans, P = .001; horizontal HD scans, P < .001; vertical standard scans, P = .001; vertical HD scans, P < .001). A fragmented ELM line was also associated with the hyperreflectivity (horizontal standard scans, P = .017; horizontal HD scans, P = .002; vertical standard scans, P = .055; vertical HD scans, P < .001).
| Standard Scan (n = 43) | HD Scan (n = 44) | |||||
|---|---|---|---|---|---|---|
| HR+ | HR- | P Value a | HR+ | HR- | P Value a | |
| Horizontal scan | ||||||
| ELM line | ||||||
| Continuous | 4 (9%) | 37 (86%) | .017 | 6 (14%) | 34 (77%) | .002 |
| Fragmented | 2 (5%) | 0 | 4 (9%) | 0 | ||
| IS/OS junction | ||||||
| Continuous | 4 (9%) | 37 (86%) | .017 | 7 (16%) | 34 (77%) | .009 |
| Fragmented | 2 (5%) | 0 | 3 (7%) | 0 | ||
| COST line | ||||||
| Continuous | 6 (14%) | 37 (86%) | >.999 | 9 (20%) | 34 (77%) | .227 |
| Fragmented | 0 | 0 | 1 (2%) | 0 | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
