|
|
||||||||
1From the Singapore National Eye Center, Singapore; the 3Biostatistics Unit and the 8Department of Ophthalmology, Yong Loo Lin School of Medicine, and the 6Department of Community, Occupational, and Family Medicine, National University of Singapore, Singapore; the 4Glaucoma Research Unit, Moorfields Eye Hospital, Institute of Ophthalmology, London, United Kingdom; the 2Singapore Eye Research Institute, Singapore; the 5Department of Ophthalmology, National University Hospital, Singapore; the 7Center for Vision Research, Westmead Millennium Institute, University of Sydney, Westmead, New South Wales, Australia; and the 9Center for Eye Research Australia, University of Melbourne, Melbourne, Australia.
| Abstract |
|---|
|
|
|---|
METHODS. This was a cross-sectional study involving 316 children 11 and 12 years of age (163 girls and 153 boys) selected randomly from one of the three schools in the Singapore Cohort study of Myopia. A total of 13 optic disc parameters were obtained from HRT images acquired before cycloplegia. Refractive errors were measured by cycloplegic autorefraction. The presence of optic disc tilt or otherwise was determined by two independent assessors using stereoscopic viewing of retinal photographs.
RESULTS. Of the 316 children, 142 had tilted discs. The tilting of optic discs was associated with a smaller disc, rim or cup area measurements, cup-to-disc area ratios, cup volumes or cup depths, but with a larger measured rim volume, rim-to-disc area ratios, height variation of the contour, retinal nerve fiber layer thicknesses or volumes, and a more negative cup shape measure (all P < 0.001). Decreased maximum cup depths were significantly associated with longer axial lengths (P < 0.001), but were not associated with spherical equivalent (P = 0.693). These associations remained only in children without tilted discs, but were no longer significant in those with tilted discs. Other HRT parameters were not associated with axial lengths or myopic status.
CONCLUSIONS. Optic nerve head parameters and retinal nerve fiber layer thickness measured by the current HRT algorithms are strongly influenced by the tilting of the optic nerve head, but not by refractive errors or axial length.
However, the impact of refractive error on HRT measurements is unclear. A previous study using the HRT for disc measurements in subjects at an outpatient ophthalmology clinic (144 Japanese adults) has concluded that the diagnostic precision of the HRT with respect to individual parameters such as cup, rim volumes, and RNFL volumes was lower in myopic discs compared with nonmyopic discs when used to discriminate for glaucoma.10 However, the classification of myopic discs was based on photographic appearance and not on the severity of the refractive error. In Asian populations,11 with a significant proportion of myopes, the reference criteria for screening optic discs for glaucomatous change may differ from those in a largely nonmyopic population. This is all the more important when one considers the higher rates of glaucoma in myopes.12 13 14
In addition, pathologic myopia is associated with optic disc tilting15 and peripapillary atrophy.16 Recently, we have shown that even young Singapore children (7–9 years of age) with relatively low degrees of myopia may exhibit optic disc differences in retinal photographs.17 The optic discs of myopic children have greater horizontal and vertical cup-to-disc ratios, the disc was more oval, and there was a tendency for the upper pole of the discs to rotate toward the fovea.17 However, these associations were based on a qualitative assessment of optic disc parameters from retinal photographs and without stereoscopic photographs, the tilting of the optic discs was not assessed. There have been no previous reports in children on the use of the HRT to document neuroretinal thickness and volumes or cup and disc areas and volumes with variations in myopic optic discs features such as tilting of discs and with severity of refractive error.
We will describe the various HRT parameters of optic nerve heads and retinal nerve fiber layer thickness in Singapore school children and evaluate the associations of these parameters with characteristics of myopic discs such as optic disc tilt and refraction.
| Methods |
|---|
|
|
|---|
The age, sex composition and mean SE of this random sample were not significantly different from those not selected for the study.
The research adhered to the tenets of Declaration of Helsinki. Informed written consent was obtained from the parents of subjects. The study was approved by the Institutional Review Board of the Singapore Eye Research Institute.
Imaging and Analysis with HRT
The eye examinations were performed at a school visit. The Heidelberg Retinal Tomograph II (HRT II; Heidelberg Engineering) measurements were performed in the subjects before cycloplegia in a dim room. The HRT II cylindrical lenses were adapted for children who had astigmatism greater than or equal to 1 D. The previous years measurements of refractive error were used for this purpose. All HRT examinations were performed by a single ophthalmologist. After the baseline images were captured by the HRT II, the optic disc margin was manually defined by a trained ophthalmologist (SCL). This critical step was accomplished by plotting a series of dots around the margin of the disc on the reflectance image provided by the computer. The disc margin was defined as the inner edge of the scleral (Elschnigs) ring. Then, data were analyzed with version 2.0 software. The HRT II optic nerve head (ONH) scan protocol was adopted, automatically repeated three times and combined to produce a pseudo three-dimensional image of the nerve head in each child. Image quality was statistically graded by the HRT II so that any images that had fallen below acceptable criteria were deleted. Each image was accompanied by a topography standard deviation (TSD). A TSD under 30 was used as the image inclusion criterion. The image was acquired point by point within a 15° x 15° area centered on the optic nerve head. A total of 13 parameters were obtained: the optic disc area, cup area, cup-to-disc area ratio, rim area, rim-to-disc area ratio, cup volume, rim volume, mean cup depth, maximum cup depth, cup shape measure, height variation contour, mean RNFL thickness, and RNFL cross-sectional area.
Cycloplegic Refraction
After the HRT examination, cycloplegic refraction was performed after instillation of three drops of 1% cyclopentolate 5 minutes apart. At least 30 minutes after the last drop, five consecutive refraction and keratometry readings were obtained with one of two calibrated autokeratorefractometers (model RK5; Canon, Inc. Ltd., Tochigiken, Japan). Axial length measurements were obtained with one of two contact ultrasound biometry machines (Echoscan model US-800, probe frequency of 10 mHz; Nidek Co., Ltd., Tokyo, Japan) after 1 drop of 0.5% proparacaine was administered. The average of six measurements was taken if the standard deviation was <0.12 mm. If the standard deviation of six measurements was
0.12 mm, the data were not included, and the measurements were repeated until SD < 0.12 mm was obtained.
Optic Disc Tilt
After pupil dilation with cyclopentolate 1% repeated twice at 5-minute intervals, stereopairs of images centered on the optic discs were obtained with a nonmydriatic retinal camera (CR-DGi; Canon, Tokyo, Japan) during the same visit. Photographs were taken with the optic disc centered with a left angulation (30°) and the optic disc centered with a right angulation (30°). The presence of tilt or otherwise in the optic discs was assessed from stereo retinal photographs independently by two ophthalmologists. The assessment of tilt was performed in a masked fashion without knowledge of HRT or any other study data. These assessors were experienced in the assessment of optic discs in clinical scenarios. In the event of differing opinions between the first two observers, the third observer assessed the photographs independently.
Data Analysis and Definitions
Low myopia was defined as having spherical equivalent (SE) smaller than and including –0.5 D but greater than –3.0 D, and higher myopia was defined as SE
–3.0 D.
Analysis of variance was used to evaluate the means of the 13 parameters and comparison was made between the three refractive error groups. Further analysis was performed using multiple linear regression with axial length or SE as the dependent variable and the 13 parameters as independent variables. For multiple comparisons among the 13 parameters measured, a Bonferroni correction was applied with a resultant significance of P < 0.0038.
| Results |
|---|
|
|
|---|
|
When stratified according to the presence or absence of disc tilt, none of the HRT parameters was associated with myopia status, after the Bonferroni correction (Table 2) .
|
|
| Discussion |
|---|
|
|
|---|
This study concurs with a previous evaluation of 66 normal and 78 glaucomatous subjects10 in Japan in which a larger rim volume, height variation contour, mean RNFL thickness, and cross-sectional area were found in eyes with myopic disc shapes. Eyes with myopic disc shapes were defined as tilted (obliquely implanted) with temporal crescents of peripapillary chorioretinal atrophy, without considering refractive errors.10 In that study, it is more likely that the disc parameters were related to myopic optic disc changes rather than independently to the presence of myopia.
Compared with a previous study using the HRT in 77 adults in Japan,21 the children in our study had a smaller disc (P < 0.001), cup (P = 0.008) and rim (P < 0.001) areas, and smaller cup (P = 0.01) and rim (P < 0.001) volumes. The mean cup depth (P = 0.027), height variation of the contour (P < 0.001), RNFL thickness (P < 0.001), and cross-sectional area (P = 0.016) were significantly larger in our study. The difference in cup shape measure was marginally significant (P = 0.04), whereas the cup-to-disc area ratios (P = 0.14) and the maximum cup depth (P = 0.06) were not significantly different between the two studies (assuming the validity of Students t-test, using the published means ± SD). However, among the parameters, only the mean rim area and cup volumes differed by above 30% between studies. These differences are probably due to the older age and Japanese ethnicity of the subjects of the other study21 compared with our subjects.
Nakamura et al.,21 in a study of 77 adults, used multiple linear regression with age, SE, and disc area as the independent variables, and the HRT parameters used in the present study in turn as the dependent variable. Reduction in RNFL thickness and cross-sectional areas were found with increasing age (P < 0.05) and cup depth; maximum cup depths were significantly deeper in myopic subjects (P < 0.05); and larger discs had larger cup area, cup-to-disc area ratio, rim area, cup volume, mean cup depth, cup shape measure (P < 0.01), and maximum cup depth (P < 0.05). The RNFL thickness was smaller in large discs (P < 0.01). Rim volume was unaffected by age, refraction, or disc area. However, optic disc tilt was not evaluated.
The present findings also appear to contradict previous findings using color fundus imaging,17 22 including findings in children recruited from the same population.17 These studies found greater cup-to-disc area ratios in myopes than in nonmyopes. However, it is unlikely that HRT parameters of the nerve head rim and cup represent the same spatial entities determined by clinical ophthalmoscopy and photographic methods, because clinical methods usually require subjective delineation of rim-to-cup border and disc margins, compared with more objective assessments by the HRT machine. In contrast to our results, studies using scanning laser polarimetry have found a reduction of RNFL in myopic subjects.23 24 However, there was no evaluation of the effect of optic disc tilt. In the presence of significant disc tilting in myopes,15 the discrepancy between confocal laser ophthalmoscopic and optical coherence tomographic measurements increases.25 In our study, thicker measured RNFL may be artifactual. In a tilted disc, the reference plane could have been placed at a falsely posterior position, increasing the measured distance between the reference plane and the nerve fiber layer surface.
In the analysis of HRT images, the calculated volumes and areas depend on the placement of the reference plane.26 For example, the cup volume is a space between the arbitrarily defined flat reference plane and the surface of the nerve head topography.26 Without any change in nerve head topography, a relatively lower (more posterior) reference plane will arbitrarily result in a smaller cup volume and vice versa. In the current version of the HRT software, the reference plane was automatically set10 at 50 µm below the mean peripapillary vertical height along the temporal sector between 350° and 356°. Taking into consideration the interindividual variability of optic nerve head topography, the value of 50 µm has been used to ensure that the optic disc cup border is lower than the optic disc contour line in most cases.26 The temporal sector between 350° and 356° was chosen because the average optic disc surface was inclined at –7° ± 3° below the horizontal meridian. Because of the tilting of the discs in myopes,15 these premises may not hold true. The result of a greater rim area volume and smaller cup area volume can simply be due to an inappropriately posterior placement of the reference plane. Other factors, such as chorioretinal atrophy16 and parapapillary atrophy,27 may also affect reference plane positions.
In myopic individuals with tilted discs, the optimal definition for the reference plane in the assessment of cup and rim parameters should employ a strategy beyond the currently used narrow temporal sector. In myopic children, a more realistic HRT analysis (one that correlates to the clinical or ophthalmoscopic impression of the cup size and depth) may be possible if sophisticated software allows for recalculation of areas and volumes after three-dimensional rotation and repositioning of the measured profiles (removing the requirement for the reference surface to be perpendicular to the z-axis of the image).
Our studies allowed the assessment of optic disc and retinal nerve fiber characteristics in a healthy population without the influences of ocular diseases such as glaucoma or optic neuropathies. Thus, the data provide a normative database for future reference. The HRT is a useful noninvasive tool for the determination of optic disc morphology in children. Our results show that tilting of the optic disc impacts on HRT measurements but that refractive error–axial length per se does not. Thus, we can use HRT measurements in myopic eyes as long as there is no tilt. However, the HRT may require modification for use in eyes with tilted optic discs. In addition, the criteria for using the HRT as a screening tool for glaucoma may have to be readjusted to take into account the tilting of discs.
There is a need to determine normative limits for eyes with abnormal shape. This will require additional data from specific groups, focusing on a handful of parameters that can discriminate normal from glaucomatous eyes and developing a method of quantifying the extent of disc tilt. Such data would be anticipated to explain the slightly different findings of the other studies mentioned herein, and provide a method for adjusting normative data according to the tilt.
In conclusion, we report that HRT determination of optic nerve parameters are affected by optic disc tilt but not refractive errors in young Singapore subjects. Thus, these relationships may provide insights into the measurement of optic disc morphology and the clinical usefulness of the HRT system.
| Footnotes |
|---|
Submitted for publication May 17, 2007; revised June 20, 2007; accepted August 7, 2007.
Disclosure: L. Tong, None; Y.-H. Chan, None; G. Gazzard, None; S.-C. Loon, None; A. Fong, None; P. Selvaraj, None; P.R. Healey, None; D. Tan, None; T.Y. Wong, None; S.M. Saw, None
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked "advertisement" in accordance with 18 U.S.C.
1734 solely to indicate this fact.
Corresponding author: Louis Tong, Singapore National Eye Center, 11, Third Hospital Avenue, Singapore 168751; louistong{at}hotmail.com.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
L Lim, G Gazzard, Y-H Chan, A Fong, A Kotecha, E-L Sim, D Tan, L Tong, and S-M Saw Corneal biomechanics, thickness and optic disc morphology in children with optic disc tilt Br. J. Ophthalmol., November 1, 2008; 92(11): 1461 - 1466. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |