- Email: [email protected]
Retinal Nerve Fiber Layer Thickness in Amblyopic Eyes
Retinal Nerve Fiber Layer Thickness in Amblyopic Eyes MICHAEL X. REPKA, MD, NITZA GOLDENBERG-COHEN, MD, AND ALLISON R. EDWARDS, MS
● PURPOSE: To compare the peripapillary retinal nerve fiber layer (RNFL) thickness of sound and amblyopic eyes. ● DESIGN: Prospective observational case series. ● METHODS: SETTING: Institutional. STUDY POPULATION: Patients with unilateral strabismic, anisometropic, or combined amblyopia. OBSERVATION: Fast RNFL analysis with ocular coherence tomography (OCT) of sound and amblyopic eyes. MEASURE: Mean RNFL thickness. ● RESULTS: For the 17 patients (mean age 10.7 years) in whom both eyes were imaged, the mean thickness of the sound eye was 109.2 m (median 112.7) and of the amblyopic eye was 104.2 m (median 105.0), and the average difference (sound eye less amblyopic eye) was 5.0 m (median 3.0) (95% confidence interval ⴚ2.3, 12.2, P ⴝ .17). The sound eye was 10 m or more thicker than the amblyopic eye in four patients; the amblyopic eye was 10 m or more thicker than the sound eye in one patient; and the difference was within 10 m in 12 patients. Test-retest data were obtained for 23 pairs of sound eyes and 21 pairs of amblyopic eyes, with 75% of the test-retest pairs within 7%. ● CONCLUSIONS: We found a small, but not clinically significant, difference in nerve fiber layer (NFL) thickness between amblyopic and sound eyes. Reliability was excellent, with most eyes testing within 7% of the first test. (Am J Ophthalmol 2006;142:247–251. © 2006 by Elsevier Inc. All rights reserved.)
A
MBLYOPIA IS DEFINED AS REDUCED BEST-COR-
rected visual acuity in one or both eyes caused by abnormal visual experience during the critical period of visual development. It is generally attributed
Accepted for publication Feb 27, 2006. From the Johns Hopkins University School of Medicine, Baltimore, Maryland (M.X.R., N.G.-C.), and Jaeb Center for Health Research, Tampa, Florida (A.R.E.). This study was supported by the Isabel and Zanvyl Krieger Fund, Baltimore, Maryland (N.G.-C.) and through a cooperative agreement by grant EY11751 from the National Eye Institute, National Institute of Health, Bethesda, Maryland (A.R.E.). Inquiries to Michael X. Repka, MD, Wilmer 233, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21287-9028; e-mail: [email protected] 0002-9394/06/$32.00 doi:10.1016/j.ajo.2006.02.030
©
2006 BY
to abnormal development of the visual cortex due to strabismus, image blur from refractive error, form deprivation, or a combination of these factors. However, some have suggested that some eyes diagnosed with amblyopia may also have abnormalities in the afferent visual system anterior to the striate cortex, including the retina, retinal ganglion cell, retinal nerve fiber layer (RNFL), optic nerve, and lateral geniculate body of the thalamus.1 In animal models of visual deprivation amblyopia during the neonatal period, histologic changes were noted in the lateral geniculate body2 and cortex.3 Similar observations were found in humans.4 Using computerized analysis of optic disk photographs from 205 amblyopic subjects, Lempert and Porter identified optic disk hypoplasia in 45% of the amblyopic eyes of patients.1,5 They also noted microphthalmos in the amblyopic eye.1 Retinal studies have failed to demonstrate any difference in cone photoreceptor function.6 Peripapillary nerve fiber layer (NFL) thickness, as a total measurement of both macular and peripheral NFL, has been reported to be the best surrogate marker for the assessment of the optic nerve in patients with glaucoma.7 Wollstein and associates8 applied optical coherence tomography (OCT) to measure peripapillary NFL thickness in glaucoma. They found it to be very sensitive to NFL loss, better than automated visual fields. Prompted by studies of OCT, as well as the observations of Lempert,1,9 we measured peripapillary retinal NFL thickness using OCT in patients with amblyopia. OCT measurement of the RNFL is not affected by refractive error or axial length of the eye, unlike photographs.10 The purpose of this pilot study was to examine the feasibility of OCT testing in amblyopic children, provide some pilot data on amblyopic eyes, and assess test-retest variability in this patient population.
METHODS THE STUDY WAS APPROVED BY THE JOHNS HOPKINS HOSPI-
tal/Johns Hopkins University combined investigational
ELSEVIER INC. ALL
RIGHTS RESERVED.
247
TABLE. Retinal Nerve Fiber Layer Thickness for the 17 Imaged Pairs
Sound eye (n ⫽ 17) Mean (SD) Median Amblyopic eye (n ⫽ 17) Mean (SD) Median Difference (sound ⫺ amblyopic eye) Mean (SD) Median
Thickness By Quadrant (microns)
Average Thickness (microns)
Inferior
Nasal
Temporal
Superior
109.2 (17.3) 112.7
139.8 (31.0) 138.0
92.2 (23.6) 91.0
72.6 (14.8) 72.0
130.8 (27.3) 139.0
104.2 (21.1) 105.0
139.4 (34.1) 137.0
84.5 (30.7) 82.0
66.8 (16.5) 68.0
127.2 (23.1) 127.0
5.0 (14.2) 3.0
0.4 (20.0) 1.0
7.6 (28.3) 11.0
5.8 (17.8) 3.0
3.5 (23.2) 2.0
SD ⫽ standard deviation.
review board. Written informed consent was obtained from each patient or from the parents of participating minors. Children seven to 17 years of age also provided written assent. Regulations of the Health Insurance Portability and Accountability Act (HIPAA) were followed as developed by the Institutional Review Board. Patients five to 30 years of age were prospectively recruited with the diagnosis of strabismic amblyopia, anisometropic amblyopia, or both. We excluded patients with myopia in either eye or hypermetropia greater than ⫹5.00 diopters measured with cycloplegia from topical cyclopentolate 1% to avoid including eyes that might have additional disease. The diagnosis was confirmed with a comprehensive eye examination with pupillary dilation and completion of visual acuity testing with the electronic Early Treatment Diabetic Retinopathy Study (eETDRS) visual acuity chart. OCT images were obtained using the Humphrey-Zeiss Stratus (OCT3) (Carl Zeiss-Humphrey-Meditec, Dublin, California, USA), with software 4.0.3.1. We utilized the rapid RNFL scan as the test most likely to be completed by the children and teens we planned to enroll. In this technique, three circular samples are taken around the optic disk with a diameter of 3.44 mm. The right eye was studied first, followed by the left eye. A retest of both eyes was instituted shortly after data collection had begun to provide some test-retest pilot data. The instrument software calculates average thickness values from the three scans for each quadrant (superior, nasal, inferior, temporal), each clock hour, and the RNFL as a whole. Signal strength is rated on a 10-point scale; thickness values with signal strength five or more are considered acceptable, and these data were then used in this study. Descriptive statistics were calculated for the initial (first) measurements, including mean and median values of RNFL thickness for sound eyes, amblyopic eyes, and intereye difference. The mean values were compared with a paired t test. 248
AMERICAN JOURNAL
Amblyopic eye test-retest data were obtained from six patients, for a total of 19 tests, and 21 test-retest pair combinations (five eyes were retested twice and one retested three times). Sound eye test-retest data were obtained from eight patients, for a total of 23 tests, and 23 test-retest pair combinations (two eyes were retested once, five retested twice, and one retested three times). Each data pair was plotted and a correlation coefficient calculated to describe the reliability. The relative absolute difference was calculated for every test-retest pair, and quartiles were reviewed to assess the degree of agreement. The difference versus the average value of all test-retest pairs was plotted, and the limits of agreement (means ⫾ 2 standard deviation (SD)) were calculated and plotted.11 An analysis of variance (ANOVA) model was fit using RNFL average thickness as the dependent variable and subject as the independent variable to determine (1) the standard error of measurement (square root of mean square error) and (2) the proportion of total variability due to measurement variability.
RESULTS EIGHTEEN AMBLYOPIC PATIENTS FIVE TO 28 YEARS OLD
were studied (mean 11.2 years, six female). The cause of amblyopia was strabismus (n ⫽ 6), anisometropia (n ⫽ 4), and combined anisometropia and strabismus (n ⫽ 8). The eETDRS logMAR acuity was measured in 15 patients using the Electronic Visual Acuity tester.12 The mean acuity of the sound eyes was 0.02 (Snellen approximation 20/20⫺1) and of the amblyopic eyes was 0.49 (Snellen approximation 20/40). The RNFL of all of the sound eyes could be imaged, whereas one amblyopic eye could not be accurately imaged (eETDRS letter count ⫽ 55, 20/80 equivalent). The scan path for that eye could not be centered correctly around the optic nerve image. For the 17 patients in whom both OF
OPHTHALMOLOGY
AUGUST 2006
ment of 4.7 m. This suggests that 2.1% of the total variability was due to the measurement variability. The analysis of sound eye data yielded similar results, with an estimated standard error of measurement equal to 4.0 m, and 2.3% of total variability due to the measurement variability.
DISCUSSION WE MEASURED NFL THICKNESS IN THE PERIPAPILLARY RE-
FIGURE. A Bland-Altman plot of the test-retest difference of retinal nerve fiber layer thickness (RNFL) for amblyopic and sound eyes. The difference of each RNFL thickness test-retest pair is plotted vs the average thickness of the test and the retest. Amblyopic and sound eyes are plotted separately with the mean difference and the limits of agreement shown as horizontal lines (mean ⴞ 2 standard deviation (SD)).
eyes were imaged, the mean age was 10.7 years. The mean thickness of the sound eye RNFL was 109.2 m (median 112.7, range 60.0 to 142.0), the mean thickness of the amblyopic eye RNFL was 104.2 m (median 105.0, range 61.0 to 139.0), and the average difference (sound eye less amblyopic eye) was 5.0 m (median 3.0) (95% confidence interval ⫺2.3, 12.2, P ⫽ .17). The sound eye was 10 or more m thicker than the amblyopic eye in four patients, the amblyopic eye was 10 or more m thicker than the sound eye in one patient, and the difference was within 10 m in 12 patients. RNFL thickness measurements were recorded for the temporal, inferior, nasal, and superior quadrants (Table). These show the greatest difference in the nasal quadrant, with the amblyopic eyes being an average of 7.6 m thinner. Test-retest data were obtained for 23 pairs of sound eyes and 21 pairs of amblyopic eyes. Bland-Altman plots for these data are shown in the Figure. The median relative absolute difference for the amblyopic eyes was 0.05 (half of the test-retests were within 5% of each other), whereas the 75th percentile was 0.07 (three-fourths of the test-retests were within 7% of each other). The median relative absolute difference for the sound eyes was 0.03 (half of the test-retests were within 3% of each other), and the 75th percentile was 0.06. The intraclass correlation coefficient was 0.97 (P ⬍ .0001) for the amblyopic eye pairs and for the sound eye pairs. An ANOVA model for the amblyopic eye test-retest data resulted in an estimated standard error of measureVOL. 142, NO. 2
gion of sound and amblyopic eyes, employing a technique similar to that reported for glaucoma.13,14 In this pilot study, we found a small (mean 5 m) but not clinically significant difference in average NFL thickness between amblyopic and sound eyes. Larger differences were found between patients. Our analytic model found that most of the variability was due to anatomic variation from patient to patient, whereas little of the variability was due to the measurement. Testing was easily performed despite the reduced central acuity and age of the patients. This varies from the findings of Hess and associates,15 who studied children with glaucoma and found some difficulty with fixation. Reliability was excellent with most eyes retesting within 7% of the first test. We found no difference in reliability between sound and amblyopic eyes. The difference we found in sound and amblyopic eyes is similar in magnitude to data reported for adults.16,17 Budenz and associates17 reported 4.7 m variability with rapid RNFL analysis. They noted that this could be reduced by using a test with greater sampling density, such as the standard RNFL analysis, which in their hands does not take substantially more time. Their findings and our preliminary data of good testability would support further investigation of the standard test in older children and teenagers. Our results agree with previous reports of NFL thickness in eyes with strabismic amblyopia using scanning laser polarimetry in which there was no clinically important difference between sound and amblyopic eyes.18,19 Colen and associates20 found an average difference of 1.5%. These research groups did not include anisometropic amblyopia, which has been suggested by Lempert to often involve abnormalities of the optic nerve.1,5 Also using the scanning laser polarimetry (GDx), Leung and associates21 reported average superior and inferior quadrant values of 79.50 and 80.41 m, respectively, in the amblyopic group and 80.75 and 82.75 m, respectively, in the controls. Previous studies of OCT and amblyopia have reported differing results. Yen and associates22 found the RNFL with OCT to be thicker in amblyopic eyes compared with sound eyes of children with anisometropic amblyopia, but no difference in children with strabismic amblyopia. However, Rabbione and associates found no difference from normal in the amblyopic eyes. (Rabbione MM, Roagna B, Tonetti S and associates Optical coherence tomography
RETINAL NERVE FIBER LAYER
AND
AMBLYOPIA
249
measurements of thickness in amblyopic and nonamblyopic eyes [abstract]. Invest Ophthalmol Vis Sci 45:S106, 2004.) Lempert9 has studied photographs of the optic nerves of human eyes with amblyopia and found an association with reduced disk areas and axial lengths in amblyopic eyes. He noted that the optic disks of hypermetropic eyes with strabismus (with and without amblyopia) were disproportionately reduced in size compared with hypermetropic eyes without amblyopia or esotropia.23 He speculated that vision impairment in amblyopia is associated with optic nerve hypoplasia with relative microphthalmos.1 If this were true, we would expect to find relatively thinner RNFL in the amblyopic eyes on OCT, which would suggest a role for anterior visual pathway pathology.9 However, in this study, none of the patients with hypermetropia and anisometropic amblyopia demonstrated detectable anatomic changes in peripapillary NFL thickness. There are several limitations of our study. This was a pilot study to demonstrate feasibility for including this testing in a planned clinical trial (currently under way) and thus does not have the statistical power to detect small differences in thickness or between types of amblyopia. Second, most of the patients enrolled in this study were young and had moderate residual amblyopia. With the small sample and the moderate level of amblyopia, these results should not be generalized to deprivation amblyopia or severe amblyopia. Furthermore, they may be different than the patients reported by Lempert with optic nerve dysplasia and amblyopia.9 Lastly, we did not have a control group of normal children, and thus the absolute thicknesses should not be considered normal even for sound eyes. To the best of our knowledge, normative data have not yet been reported for children with OCT using the rapid RNFL analysis program. Some normative data for children and teens were reported by Hess and associates15 for 104 normal patients aged 3 to 17 years using the OCT3, but they employed a different program sampling two concentric circles at diameters different from those we used. In this pilot study, we found rapid RNFL analysis with OCT to be readily testable in children and teenagers with amblyopia and to have high test-retest correlation. There was no clinically significant interocular difference in RNFL for eyes of patients with amblyopia due to strabismus, anisometropia, or both combined.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
REFERENCES 1. Lempert P. Optic nerve hypoplasia and small eyes in presumed amblyopia. J AAPOS 2000;4:258 –266. 2. von Noorden GK, Middleditch PR. Histology of the monkey lateral geniculate nucleus after unilateral lid closure and
250
AMERICAN JOURNAL
18.
OF
experimental strabismus: further observations. Invest Ophthalmol 1975;14:674 – 683. Kiorpes L, Kiper DC, O’Keefe LP, et al. Neuronal correlates of amblyopia in the visual cortex of macaque monkeys with experimental strabismus and anisometropia. J Neurosci 1998; 18:6411– 6424. von Noorden GK, Crawford ML. The lateral geniculate nucleus in human strabismic amblyopia. Invest Ophthalmol Vis Sci 1992;33:2729 –2732. Lempert P, Porter L. Dysversion of the optic disc and axial length measurements in a presumed amblyopic population. J AAPOS 1998;2:207–213. Delint PJ, Weissenbruch C, Berendschot TT, Norren DV. Photoreceptor function in unilateral amblyopia. Vision Res 1998;38:613– 617. Leung CK, Chan WM, Yung WH, et al. Comparison of macular and peripapillary measurements for the detection of glaucoma: an optical coherence tomography study. Ophthalmology 2005;112:391– 400. Wollstein G, Schuman JS, Price LL, et al. Optical coherence tomography longitudinal evaluation of retinal nerve fiber layer thickness in glaucoma. Arch Ophthalmol 2005;123: 464 – 470. Lempert P. The axial length/disc area ratio in anisometropic hyperopic amblyopia: a hypothesis for decreased unilateral vision associated with hyperopic anisometropia. Ophthalmology 2004;111:304 –308. Schuman JS, Pedut-Kloizman T, Hertzmark E, et al. Reproducibility of nerve fiber layer thickness measurements using optical coherence tomography. Ophthalmology 1996;103: 1889 –1898. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310. Beck RW, Moke PS, Turpin AH, et al. A computerized method of visual acuity testing: adaptation of the early treatment of diabetic retinopathy study testing protocol. Am J Ophthalmol 2003;135:194 –205. Wollstein G, Schuman JS, Price LL, et al. Optical coherence tomography (OCT) macular and peripapillary retinal nerve fiber layer measurements and automated visual fields. Am J Ophthalmol 2004;138:218 –225. Sony P, Sihota R, Tewari HK, et al. Quantification of the retinal nerve fibre layer thickness in normal Indian eyes with optical coherence tomography. Indian J Ophthalmol 2004; 52:303–309. Hess DB, Asrani SG, Bhide MG, et al. Macular and retinal nerve fiber layer analysis of normal and glaucomatous eyes in children using optical coherence tomography. Am J Ophthalmol 2005;139:509 –517. Carpineto P, Ciancaglini M, Zuppardi E, et al. Reliability of nerve fiber layer thickness measurements using optical coherence tomography in normal and glaucomatous eyes. Ophthalmology 2003;110:190 –195. Budenz DL, Chang RT, Huang X, et al. Reproducibility of retinal nerve fiber thickness measurements using the stratus OCT in normal and glaucomatous eyes. Invest Ophthalmol Vis Sci 2005;46:2440 –2443. Colen TP, de Faber JT, Lemij HG. Retinal nerve fiber layer thickness in human strabismic amblyopia. Binocul Vis Strabismus Q 2000;15:141–145.
OPHTHALMOLOGY
AUGUST 2006
19. Baddini-Caramelli C, Hatanaka M, Polati M, et al. Thickness of the retinal nerve fiber layer in amblyopic and normal eyes: a scanning laser polarimetry study. J AAPOS 2001;5:82– 84. 20. Colen TP, de Faber JT, Lemij HG. Retinal nerve fiber layer thickness in human strabismic amblyopia. Binocul Vis Strabismus Q 2000;15:141–146. 21. Leung CK, Chan WM, Hui YL, et al. Analysis of retinal nerve fiber layer and optic nerve head in glaucoma with
VOL. 142, NO. 2
different reference plane offsets, using optical coherence tomography. Invest Ophthalmol Vis Sci 2005;46: 891– 899. 22. Yen M, Cheng C, Wang A. Retinal nerve fiber layer thickness in unilateral amblyopia. Invest Ophthalmol Vis Sci 2004;45: 2224 –2230. 23. Lempert P. Axial length-disc area ratio in esotropic amblyopia. Arch Ophthalmol 2003;121:821– 824.
RETINAL NERVE FIBER LAYER
AND
AMBLYOPIA
251
Biosketch Nitza Goldenberg-Cohen, MD, is currently a Pediatric Neuro-ophthalmologist at Schneider Children’s Medical Center of Israel and head of the Krieger Eye Research Laboratory at FMRC, Tel Aviv University, Israel. Dr Goldenberg-Cohen received her medical degree and ophthalmology residency at the Sackler School of Medicine, Tel Aviv University, Israel. She completed fellowships in Pediatric Ophthalmology and Neuro-Ophthalmology at the Wilmer Eye Institute. Dr Goldenberg-Cohen research interests include amblyopia and strabismus, as well as neuroprotection and stem cell research.
251.e1
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
AUGUST 2006
Biosketch Michael X. Repka, MD, is a Pediatric Ophthalmologist at the Wilmer Institute of the Johns Hopkins University, Baltimore, Maryland. Dr Repka has interests in neuro-ophthalmology in children, retinopathy of prematurity, amblyopia, and strabismus.
VOL. 142, NO. 2
RETINAL NERVE FIBER LAYER
AND
AMBLYOPIA
251.e2
● PURPOSE: To compare the peripapillary retinal nerve fiber layer (RNFL) thickness of sound and amblyopic eyes. ● DESIGN: Prospective observational case series. ● METHODS: SETTING: Institutional. STUDY POPULATION: Patients with unilateral strabismic, anisometropic, or combined amblyopia. OBSERVATION: Fast RNFL analysis with ocular coherence tomography (OCT) of sound and amblyopic eyes. MEASURE: Mean RNFL thickness. ● RESULTS: For the 17 patients (mean age 10.7 years) in whom both eyes were imaged, the mean thickness of the sound eye was 109.2 m (median 112.7) and of the amblyopic eye was 104.2 m (median 105.0), and the average difference (sound eye less amblyopic eye) was 5.0 m (median 3.0) (95% confidence interval ⴚ2.3, 12.2, P ⴝ .17). The sound eye was 10 m or more thicker than the amblyopic eye in four patients; the amblyopic eye was 10 m or more thicker than the sound eye in one patient; and the difference was within 10 m in 12 patients. Test-retest data were obtained for 23 pairs of sound eyes and 21 pairs of amblyopic eyes, with 75% of the test-retest pairs within 7%. ● CONCLUSIONS: We found a small, but not clinically significant, difference in nerve fiber layer (NFL) thickness between amblyopic and sound eyes. Reliability was excellent, with most eyes testing within 7% of the first test. (Am J Ophthalmol 2006;142:247–251. © 2006 by Elsevier Inc. All rights reserved.)
A
MBLYOPIA IS DEFINED AS REDUCED BEST-COR-
rected visual acuity in one or both eyes caused by abnormal visual experience during the critical period of visual development. It is generally attributed
Accepted for publication Feb 27, 2006. From the Johns Hopkins University School of Medicine, Baltimore, Maryland (M.X.R., N.G.-C.), and Jaeb Center for Health Research, Tampa, Florida (A.R.E.). This study was supported by the Isabel and Zanvyl Krieger Fund, Baltimore, Maryland (N.G.-C.) and through a cooperative agreement by grant EY11751 from the National Eye Institute, National Institute of Health, Bethesda, Maryland (A.R.E.). Inquiries to Michael X. Repka, MD, Wilmer 233, The Johns Hopkins Hospital, 600 North Wolfe Street, Baltimore, MD 21287-9028; e-mail: [email protected] 0002-9394/06/$32.00 doi:10.1016/j.ajo.2006.02.030
©
2006 BY
to abnormal development of the visual cortex due to strabismus, image blur from refractive error, form deprivation, or a combination of these factors. However, some have suggested that some eyes diagnosed with amblyopia may also have abnormalities in the afferent visual system anterior to the striate cortex, including the retina, retinal ganglion cell, retinal nerve fiber layer (RNFL), optic nerve, and lateral geniculate body of the thalamus.1 In animal models of visual deprivation amblyopia during the neonatal period, histologic changes were noted in the lateral geniculate body2 and cortex.3 Similar observations were found in humans.4 Using computerized analysis of optic disk photographs from 205 amblyopic subjects, Lempert and Porter identified optic disk hypoplasia in 45% of the amblyopic eyes of patients.1,5 They also noted microphthalmos in the amblyopic eye.1 Retinal studies have failed to demonstrate any difference in cone photoreceptor function.6 Peripapillary nerve fiber layer (NFL) thickness, as a total measurement of both macular and peripheral NFL, has been reported to be the best surrogate marker for the assessment of the optic nerve in patients with glaucoma.7 Wollstein and associates8 applied optical coherence tomography (OCT) to measure peripapillary NFL thickness in glaucoma. They found it to be very sensitive to NFL loss, better than automated visual fields. Prompted by studies of OCT, as well as the observations of Lempert,1,9 we measured peripapillary retinal NFL thickness using OCT in patients with amblyopia. OCT measurement of the RNFL is not affected by refractive error or axial length of the eye, unlike photographs.10 The purpose of this pilot study was to examine the feasibility of OCT testing in amblyopic children, provide some pilot data on amblyopic eyes, and assess test-retest variability in this patient population.
METHODS THE STUDY WAS APPROVED BY THE JOHNS HOPKINS HOSPI-
tal/Johns Hopkins University combined investigational
ELSEVIER INC. ALL
RIGHTS RESERVED.
247
TABLE. Retinal Nerve Fiber Layer Thickness for the 17 Imaged Pairs
Sound eye (n ⫽ 17) Mean (SD) Median Amblyopic eye (n ⫽ 17) Mean (SD) Median Difference (sound ⫺ amblyopic eye) Mean (SD) Median
Thickness By Quadrant (microns)
Average Thickness (microns)
Inferior
Nasal
Temporal
Superior
109.2 (17.3) 112.7
139.8 (31.0) 138.0
92.2 (23.6) 91.0
72.6 (14.8) 72.0
130.8 (27.3) 139.0
104.2 (21.1) 105.0
139.4 (34.1) 137.0
84.5 (30.7) 82.0
66.8 (16.5) 68.0
127.2 (23.1) 127.0
5.0 (14.2) 3.0
0.4 (20.0) 1.0
7.6 (28.3) 11.0
5.8 (17.8) 3.0
3.5 (23.2) 2.0
SD ⫽ standard deviation.
review board. Written informed consent was obtained from each patient or from the parents of participating minors. Children seven to 17 years of age also provided written assent. Regulations of the Health Insurance Portability and Accountability Act (HIPAA) were followed as developed by the Institutional Review Board. Patients five to 30 years of age were prospectively recruited with the diagnosis of strabismic amblyopia, anisometropic amblyopia, or both. We excluded patients with myopia in either eye or hypermetropia greater than ⫹5.00 diopters measured with cycloplegia from topical cyclopentolate 1% to avoid including eyes that might have additional disease. The diagnosis was confirmed with a comprehensive eye examination with pupillary dilation and completion of visual acuity testing with the electronic Early Treatment Diabetic Retinopathy Study (eETDRS) visual acuity chart. OCT images were obtained using the Humphrey-Zeiss Stratus (OCT3) (Carl Zeiss-Humphrey-Meditec, Dublin, California, USA), with software 4.0.3.1. We utilized the rapid RNFL scan as the test most likely to be completed by the children and teens we planned to enroll. In this technique, three circular samples are taken around the optic disk with a diameter of 3.44 mm. The right eye was studied first, followed by the left eye. A retest of both eyes was instituted shortly after data collection had begun to provide some test-retest pilot data. The instrument software calculates average thickness values from the three scans for each quadrant (superior, nasal, inferior, temporal), each clock hour, and the RNFL as a whole. Signal strength is rated on a 10-point scale; thickness values with signal strength five or more are considered acceptable, and these data were then used in this study. Descriptive statistics were calculated for the initial (first) measurements, including mean and median values of RNFL thickness for sound eyes, amblyopic eyes, and intereye difference. The mean values were compared with a paired t test. 248
AMERICAN JOURNAL
Amblyopic eye test-retest data were obtained from six patients, for a total of 19 tests, and 21 test-retest pair combinations (five eyes were retested twice and one retested three times). Sound eye test-retest data were obtained from eight patients, for a total of 23 tests, and 23 test-retest pair combinations (two eyes were retested once, five retested twice, and one retested three times). Each data pair was plotted and a correlation coefficient calculated to describe the reliability. The relative absolute difference was calculated for every test-retest pair, and quartiles were reviewed to assess the degree of agreement. The difference versus the average value of all test-retest pairs was plotted, and the limits of agreement (means ⫾ 2 standard deviation (SD)) were calculated and plotted.11 An analysis of variance (ANOVA) model was fit using RNFL average thickness as the dependent variable and subject as the independent variable to determine (1) the standard error of measurement (square root of mean square error) and (2) the proportion of total variability due to measurement variability.
RESULTS EIGHTEEN AMBLYOPIC PATIENTS FIVE TO 28 YEARS OLD
were studied (mean 11.2 years, six female). The cause of amblyopia was strabismus (n ⫽ 6), anisometropia (n ⫽ 4), and combined anisometropia and strabismus (n ⫽ 8). The eETDRS logMAR acuity was measured in 15 patients using the Electronic Visual Acuity tester.12 The mean acuity of the sound eyes was 0.02 (Snellen approximation 20/20⫺1) and of the amblyopic eyes was 0.49 (Snellen approximation 20/40). The RNFL of all of the sound eyes could be imaged, whereas one amblyopic eye could not be accurately imaged (eETDRS letter count ⫽ 55, 20/80 equivalent). The scan path for that eye could not be centered correctly around the optic nerve image. For the 17 patients in whom both OF
OPHTHALMOLOGY
AUGUST 2006
ment of 4.7 m. This suggests that 2.1% of the total variability was due to the measurement variability. The analysis of sound eye data yielded similar results, with an estimated standard error of measurement equal to 4.0 m, and 2.3% of total variability due to the measurement variability.
DISCUSSION WE MEASURED NFL THICKNESS IN THE PERIPAPILLARY RE-
FIGURE. A Bland-Altman plot of the test-retest difference of retinal nerve fiber layer thickness (RNFL) for amblyopic and sound eyes. The difference of each RNFL thickness test-retest pair is plotted vs the average thickness of the test and the retest. Amblyopic and sound eyes are plotted separately with the mean difference and the limits of agreement shown as horizontal lines (mean ⴞ 2 standard deviation (SD)).
eyes were imaged, the mean age was 10.7 years. The mean thickness of the sound eye RNFL was 109.2 m (median 112.7, range 60.0 to 142.0), the mean thickness of the amblyopic eye RNFL was 104.2 m (median 105.0, range 61.0 to 139.0), and the average difference (sound eye less amblyopic eye) was 5.0 m (median 3.0) (95% confidence interval ⫺2.3, 12.2, P ⫽ .17). The sound eye was 10 or more m thicker than the amblyopic eye in four patients, the amblyopic eye was 10 or more m thicker than the sound eye in one patient, and the difference was within 10 m in 12 patients. RNFL thickness measurements were recorded for the temporal, inferior, nasal, and superior quadrants (Table). These show the greatest difference in the nasal quadrant, with the amblyopic eyes being an average of 7.6 m thinner. Test-retest data were obtained for 23 pairs of sound eyes and 21 pairs of amblyopic eyes. Bland-Altman plots for these data are shown in the Figure. The median relative absolute difference for the amblyopic eyes was 0.05 (half of the test-retests were within 5% of each other), whereas the 75th percentile was 0.07 (three-fourths of the test-retests were within 7% of each other). The median relative absolute difference for the sound eyes was 0.03 (half of the test-retests were within 3% of each other), and the 75th percentile was 0.06. The intraclass correlation coefficient was 0.97 (P ⬍ .0001) for the amblyopic eye pairs and for the sound eye pairs. An ANOVA model for the amblyopic eye test-retest data resulted in an estimated standard error of measureVOL. 142, NO. 2
gion of sound and amblyopic eyes, employing a technique similar to that reported for glaucoma.13,14 In this pilot study, we found a small (mean 5 m) but not clinically significant difference in average NFL thickness between amblyopic and sound eyes. Larger differences were found between patients. Our analytic model found that most of the variability was due to anatomic variation from patient to patient, whereas little of the variability was due to the measurement. Testing was easily performed despite the reduced central acuity and age of the patients. This varies from the findings of Hess and associates,15 who studied children with glaucoma and found some difficulty with fixation. Reliability was excellent with most eyes retesting within 7% of the first test. We found no difference in reliability between sound and amblyopic eyes. The difference we found in sound and amblyopic eyes is similar in magnitude to data reported for adults.16,17 Budenz and associates17 reported 4.7 m variability with rapid RNFL analysis. They noted that this could be reduced by using a test with greater sampling density, such as the standard RNFL analysis, which in their hands does not take substantially more time. Their findings and our preliminary data of good testability would support further investigation of the standard test in older children and teenagers. Our results agree with previous reports of NFL thickness in eyes with strabismic amblyopia using scanning laser polarimetry in which there was no clinically important difference between sound and amblyopic eyes.18,19 Colen and associates20 found an average difference of 1.5%. These research groups did not include anisometropic amblyopia, which has been suggested by Lempert to often involve abnormalities of the optic nerve.1,5 Also using the scanning laser polarimetry (GDx), Leung and associates21 reported average superior and inferior quadrant values of 79.50 and 80.41 m, respectively, in the amblyopic group and 80.75 and 82.75 m, respectively, in the controls. Previous studies of OCT and amblyopia have reported differing results. Yen and associates22 found the RNFL with OCT to be thicker in amblyopic eyes compared with sound eyes of children with anisometropic amblyopia, but no difference in children with strabismic amblyopia. However, Rabbione and associates found no difference from normal in the amblyopic eyes. (Rabbione MM, Roagna B, Tonetti S and associates Optical coherence tomography
RETINAL NERVE FIBER LAYER
AND
AMBLYOPIA
249
measurements of thickness in amblyopic and nonamblyopic eyes [abstract]. Invest Ophthalmol Vis Sci 45:S106, 2004.) Lempert9 has studied photographs of the optic nerves of human eyes with amblyopia and found an association with reduced disk areas and axial lengths in amblyopic eyes. He noted that the optic disks of hypermetropic eyes with strabismus (with and without amblyopia) were disproportionately reduced in size compared with hypermetropic eyes without amblyopia or esotropia.23 He speculated that vision impairment in amblyopia is associated with optic nerve hypoplasia with relative microphthalmos.1 If this were true, we would expect to find relatively thinner RNFL in the amblyopic eyes on OCT, which would suggest a role for anterior visual pathway pathology.9 However, in this study, none of the patients with hypermetropia and anisometropic amblyopia demonstrated detectable anatomic changes in peripapillary NFL thickness. There are several limitations of our study. This was a pilot study to demonstrate feasibility for including this testing in a planned clinical trial (currently under way) and thus does not have the statistical power to detect small differences in thickness or between types of amblyopia. Second, most of the patients enrolled in this study were young and had moderate residual amblyopia. With the small sample and the moderate level of amblyopia, these results should not be generalized to deprivation amblyopia or severe amblyopia. Furthermore, they may be different than the patients reported by Lempert with optic nerve dysplasia and amblyopia.9 Lastly, we did not have a control group of normal children, and thus the absolute thicknesses should not be considered normal even for sound eyes. To the best of our knowledge, normative data have not yet been reported for children with OCT using the rapid RNFL analysis program. Some normative data for children and teens were reported by Hess and associates15 for 104 normal patients aged 3 to 17 years using the OCT3, but they employed a different program sampling two concentric circles at diameters different from those we used. In this pilot study, we found rapid RNFL analysis with OCT to be readily testable in children and teenagers with amblyopia and to have high test-retest correlation. There was no clinically significant interocular difference in RNFL for eyes of patients with amblyopia due to strabismus, anisometropia, or both combined.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
REFERENCES 1. Lempert P. Optic nerve hypoplasia and small eyes in presumed amblyopia. J AAPOS 2000;4:258 –266. 2. von Noorden GK, Middleditch PR. Histology of the monkey lateral geniculate nucleus after unilateral lid closure and
250
AMERICAN JOURNAL
18.
OF
experimental strabismus: further observations. Invest Ophthalmol 1975;14:674 – 683. Kiorpes L, Kiper DC, O’Keefe LP, et al. Neuronal correlates of amblyopia in the visual cortex of macaque monkeys with experimental strabismus and anisometropia. J Neurosci 1998; 18:6411– 6424. von Noorden GK, Crawford ML. The lateral geniculate nucleus in human strabismic amblyopia. Invest Ophthalmol Vis Sci 1992;33:2729 –2732. Lempert P, Porter L. Dysversion of the optic disc and axial length measurements in a presumed amblyopic population. J AAPOS 1998;2:207–213. Delint PJ, Weissenbruch C, Berendschot TT, Norren DV. Photoreceptor function in unilateral amblyopia. Vision Res 1998;38:613– 617. Leung CK, Chan WM, Yung WH, et al. Comparison of macular and peripapillary measurements for the detection of glaucoma: an optical coherence tomography study. Ophthalmology 2005;112:391– 400. Wollstein G, Schuman JS, Price LL, et al. Optical coherence tomography longitudinal evaluation of retinal nerve fiber layer thickness in glaucoma. Arch Ophthalmol 2005;123: 464 – 470. Lempert P. The axial length/disc area ratio in anisometropic hyperopic amblyopia: a hypothesis for decreased unilateral vision associated with hyperopic anisometropia. Ophthalmology 2004;111:304 –308. Schuman JS, Pedut-Kloizman T, Hertzmark E, et al. Reproducibility of nerve fiber layer thickness measurements using optical coherence tomography. Ophthalmology 1996;103: 1889 –1898. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–310. Beck RW, Moke PS, Turpin AH, et al. A computerized method of visual acuity testing: adaptation of the early treatment of diabetic retinopathy study testing protocol. Am J Ophthalmol 2003;135:194 –205. Wollstein G, Schuman JS, Price LL, et al. Optical coherence tomography (OCT) macular and peripapillary retinal nerve fiber layer measurements and automated visual fields. Am J Ophthalmol 2004;138:218 –225. Sony P, Sihota R, Tewari HK, et al. Quantification of the retinal nerve fibre layer thickness in normal Indian eyes with optical coherence tomography. Indian J Ophthalmol 2004; 52:303–309. Hess DB, Asrani SG, Bhide MG, et al. Macular and retinal nerve fiber layer analysis of normal and glaucomatous eyes in children using optical coherence tomography. Am J Ophthalmol 2005;139:509 –517. Carpineto P, Ciancaglini M, Zuppardi E, et al. Reliability of nerve fiber layer thickness measurements using optical coherence tomography in normal and glaucomatous eyes. Ophthalmology 2003;110:190 –195. Budenz DL, Chang RT, Huang X, et al. Reproducibility of retinal nerve fiber thickness measurements using the stratus OCT in normal and glaucomatous eyes. Invest Ophthalmol Vis Sci 2005;46:2440 –2443. Colen TP, de Faber JT, Lemij HG. Retinal nerve fiber layer thickness in human strabismic amblyopia. Binocul Vis Strabismus Q 2000;15:141–145.
OPHTHALMOLOGY
AUGUST 2006
19. Baddini-Caramelli C, Hatanaka M, Polati M, et al. Thickness of the retinal nerve fiber layer in amblyopic and normal eyes: a scanning laser polarimetry study. J AAPOS 2001;5:82– 84. 20. Colen TP, de Faber JT, Lemij HG. Retinal nerve fiber layer thickness in human strabismic amblyopia. Binocul Vis Strabismus Q 2000;15:141–146. 21. Leung CK, Chan WM, Hui YL, et al. Analysis of retinal nerve fiber layer and optic nerve head in glaucoma with
VOL. 142, NO. 2
different reference plane offsets, using optical coherence tomography. Invest Ophthalmol Vis Sci 2005;46: 891– 899. 22. Yen M, Cheng C, Wang A. Retinal nerve fiber layer thickness in unilateral amblyopia. Invest Ophthalmol Vis Sci 2004;45: 2224 –2230. 23. Lempert P. Axial length-disc area ratio in esotropic amblyopia. Arch Ophthalmol 2003;121:821– 824.
RETINAL NERVE FIBER LAYER
AND
AMBLYOPIA
251
Biosketch Nitza Goldenberg-Cohen, MD, is currently a Pediatric Neuro-ophthalmologist at Schneider Children’s Medical Center of Israel and head of the Krieger Eye Research Laboratory at FMRC, Tel Aviv University, Israel. Dr Goldenberg-Cohen received her medical degree and ophthalmology residency at the Sackler School of Medicine, Tel Aviv University, Israel. She completed fellowships in Pediatric Ophthalmology and Neuro-Ophthalmology at the Wilmer Eye Institute. Dr Goldenberg-Cohen research interests include amblyopia and strabismus, as well as neuroprotection and stem cell research.
251.e1
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
AUGUST 2006
Biosketch Michael X. Repka, MD, is a Pediatric Ophthalmologist at the Wilmer Institute of the Johns Hopkins University, Baltimore, Maryland. Dr Repka has interests in neuro-ophthalmology in children, retinopathy of prematurity, amblyopia, and strabismus.
VOL. 142, NO. 2
RETINAL NERVE FIBER LAYER
AND
AMBLYOPIA
251.e2