Optic nerve head imaging

CONFIRMATORY RESULTS IN SUSPECT GLAUCOMA PATIENTS WITH NORMAL VISUAL FIELD AND ABNORMAL RETINAL NERVE FIBER LAYER FINDINGS

ANNE BJERRE, DAVID B. HENSON, ANNA J. KWARTZ, SAJ MAHMOOD, JOHN R. GRIGG and ANNE F. SPENCER

Ophthalmology, School of Medicine, University of Manchester, Manchester, UK

Abstract

Purpose: To investigate eyes with a normal visual field and an abnormal retinal nerve fiber layer, as measured with the Scanning Laser Polarimeter (GDx), with a battery of diagnostic tests to assess the agreement between the different diagnostic tests in this subset of eyes. Methods: Thirty-two eyes of 21 patients were selected. All eyes had, on at least three consecutive visits, demonstrated normal visual fields (Humphrey Field Analyzer 24-2 full threshold program, GHT within normal limits) and loss of retinal nerve fiber layer (GDx parameters outside normal limits). Each eye was further examined with the Heidelberg Retina Tomograph (HRT) and multifocal visual evoked potential (mVEP). Color stereoscopic photographs of the optic nerve head were classified by three, masked, glaucoma specialists. Each test classified the eyes as normal, borderline or glaucomatous. The classification was tabulated and the percentage of agreement between the individual tests was calculated. Results: Using a polling system of the five diagnostic tests to decide if an eye was normal or suspect/glaucomatous resulted in 63% of the eyes being falsely labeled suspect/glaucomatous by GDx and 37% being falsely labeled normal by the visual field test. Excluding parameters sensitive to variations in corneal birefringence improved the diagnostic performance of GDx to 43%, and the visual field decreased to 57%. The highest level of agreement was found between the visual field and mVEP data (74%). GDx showed relatively poor agreement with all the tests, even when potential errors from a fixed corneal compensator were largely removed. Conclusions: GDx demonstrated relatively poor agreement with the other diagnostic tests in this small proportion of patients. When disparities exist between diagnostic tests for glaucoma, it is important to do a battery of tests before making clinical decisions.

Introduction

Glaucomatous neuropathy causes structural damage to the optic nerve head and functional damage to the visual field.1 Thinning of the retinal nerve fiber layer (RNFL) is an early indicator of glaucoma and up to 50% of the nerve fibers may be lost prior to detectable visual field loss.2,3 Hence, early recognition of structural damage is important in the detection of glaucoma.

Address for correspondence: Anne Bjerre, Academic Department of Ophthalmology, Manchester Royal Eye Hospital, Oxford Road, Manchester M13 9WH, UK. Email: annelbjerre@yahoo.co.uk

Perimetry Update 2002/2003, pp. 277–286

Proceedings of the XVth International Perimetric Society Meeting, Stratford-upon-Avon, England, June 26–29, 2002

edited by David B. Henson and Michael Wall

© 2004 Kugler Publications, The Hague, The Netherlands

It is not unusual to get disparate results between diagnostic tests in the early stages of glaucoma. A detailed investigation of these cases can often lead to a better understanding of the advantages/disadvantages of a particular diagnostic test and also point to where technological developments need to be made.

Instruments such as the Scanning Laser Polarimeter (GDx) and Heidelberg Retina Tomograph (HRT) have been developed to quantitatively measure the thickness of the RNFL and optic disc topography respectively. Color stereoscopic photographs of the optic nerve head provide a qualitative means of evaluating structural damage or change.

GDx is a scanning laser polarimeter that uses the polarization properties of the RNFL to measure its thickness. It is a scanning laser ophthalmoscope with an integrated polarization device that measures the birefringence of the RNFL, and it also has a corneal polarization compensator with a fixed optic axis and magnitude of retardation.4 A 780-nm diode laser beam scans the retina, which penetrates the birefringent RNFL and is partly reflected back from the deeper retinal layers. The light emerging from the eye is separated from the incident light beam by a nonpolarizing beam splitter and the polarization state of the light is analyzed by a polarimeter. The operating principles of GDx have previously been described by Weinreb et al.5

HRT is a confocal scanning laser ophthalmoscope that can measure the topography of the optic nerve head (optic disc and peripapillary retina). A 670-nm diode laser scans a retinal area set to 10 × 10°, 15 × 15° or 20 × 20°. The topographical image is derived from 32 optical sections equally spaced along the z-axis (perpendicular to the optical axis). The image resolution is 256 × 256 pixels and the software produces a topographic map of the retinal surface height (maximum reflectivity) at each pixel. The operating principles of the HRT have previously been described by Chauhan et al.1

Visual function is commonly assessed using the full threshold strategy of the Humphrey Field Analyzer (HFA). A new method, measuring functional loss, utilizes multifocal visual evoked potentials (mVEP). Baseler et al.6 first recorded mVEP. The technique used pseudorandom stimulus presentation (M-sequences) that were cortically scaled and stimulated multiple locations of the visual field simultaneously. They concluded that it would not be a valuable test for clinical assessment due to the high inter-subject variability caused by variation in cortical anatomy.

The ObjectiVision system was developed to reduce the problems of variability. It employs an EEG-based scaling algorithm during recording, which decreases the inter-subject variability.7 Further modifications in electrode placement (an array of four bipolar occipital electrodes placed above, below and to either side of the inion) along with multichannel recording has led to results that show a strong relationship between subjective visual field loss and cortical response amplitudes.8

An important issue to consider when using GDx is that there is considerable intersubject variability in the axis of corneal birefringence, which may induce an error in the absolute measure of the RNFL thickness measurements.4 Garway-Heath et al.9 established that correcting the peripapillary retardation measurements with macular retardation measurements produced a tighter normal distribution and better differentiation between normal and glaucomatous eyes. They also highlighted that parameters based on ratios and differences between different peripapillary segments have been consistently better at discriminating between normal and glaucomatous eyes, and that this improved performance is likely to be due to their insensitivity to absolute thickness values.

This study reports on a cohort of eyes with repeatable normal visual field results and an abnormal RNFL, as determined with GDx. Each eye was examined with a battery of additional tests including HRT, mVEP, and ophthalmic examination of stereo optic nerve head photographs in order to assess the agreement between the different diagnostic tests in this subset of patients.

Methods

Thirty-two eyes of 21 patients were selected from a population of 547 patients enrolled in a glaucoma imaging study at Manchester Royal Eye Hospital. Criteria for inclusion in the glaucoma imaging study were that patients had to be Caucasian and over the age of 40 years. Their spherical refractive error had to be within ±5.00 DS and the cylinder component less than 3.00 DC. Visual acuity was better than 0.5 (LogMAR) and the included eyes had no other ocular pathology. Ocular hypertensive patients had untreated intraocular pressures (IOPs) of >23 mmHg.

All the patients had prior experience of full threshold perimetry. The study followed the tenets of the Declaration of Helsinki for research involving human subjects, and informed consent was obtained from all participants before enrolment in the study.

The 32 eyes selected had, on at least three consecutive visits, demonstrated normal visual fields (HFA 24-2 full threshold program, GHT ‘within normal limits’) and an abnormal RNFL (GDx, software version 2.0.09, Laser Diagnostic Technologies, San Diego, CA).

Five images were taken with GDx and the three of highest quality were chosen to produce a mean image from which the data were taken. GDx data for each eye were classified as borderline or outside normal limits, on the basis of one or more of the 15 parameters automatically calculated by GDx software.10

Visual field testing was performed using HFA (24-2 full threshold program) and an appropriate near correction was worn. Reliability criteria employed were false positives <25%, false negatives <25% and fixation losses <25%. A normal visual field was taken to be one with a HFA-GHT result ‘within normal limits’.

Each eye was further examined with HRT, mVEP, and color stereoscopic optic nerve head photographs. The scan area of HRT (software version 2.01, Heidelberg Engineering, Heidelberg, Germany) was set at 10 × 10° for all patients. Five images were taken and the three of highest quality were chosen to produce a mean image. One experienced operator drew a contour line around the disc margin. In order to determine whether HRT images were normal, borderline or glaucomatous, the Moorfield’s linear regression analysis between optic disc area and log of the neuroretinal rim area was utilized.12

mVEP was recorded using AccuMap (ObjectiVision, Sydney). The AccuMap uses a spread spectrum technique with families of binary sequences to drive the visual stimulus. This system was a pseudo-random cortically scaled pattern stimulus consisting of 58 segments (out to an eccentricity of 32°) with each segment containing a checkerboard pattern (16 checks). This stimulus was presented to the patient at a distance of 30 cm. The central area of 1° was used as fixation monitor. Each segment (stimulation site) is modulated in time according to a different time sequence. The resulting signal is processed by a computer by cross-correlation of the response evoked by the sequence stimulation with the sequence itself. During the recording, EEG scaling of the VEP amplitude is auto-

matically carried out to reduce inter-subject variability. Four gold disc Grass electrodes placed in a custom designed occipital cross electrode holder were used for recording.8 Patients wore an appropriate near correction. This technique has been described in detail previously.7,11 For mVEP, the AccuMap Glaucoma Severity Index Score was used to classify eyes as normal, borderline, or glaucomatous.7

Prior to taking color stereoscopic photographs of the optic nerve head, the patients’ pupils were dilated with 1% tropicamide. Three glaucoma specialists examined the optic nerve head photographs of the selected eyes using a stereoscopic viewer. These three glaucoma specialists classified the color stereoscopic photographs of the optic nerve head as normal, borderline or glaucomatous, and were masked to each other’s classification and the patient identification. The final diagnosis was based on a minimum of two specialists giving the same classification (on no occasion did all the specialists give different classifications).

The classification of normal, borderline or glaucoma of each eye provided by the different diagnostic tests was tabulated, and the percentage of agreement between the individual tests calculated.

A polling system of the five diagnostic tests was utilized in order to determine whether the eyes were normal or suspect/glaucomatous. Eyes were classified as suspect/ glaucomatous if at least two of the additional three tests (mVEP, HRT and/or stereoscopic photographs) gave a borderline/glaucomatous result.

Results

The sample consisted of 12 females and nine males (age range, 46-77 years). The classification of normal, borderline or glaucoma given by the diagnostic tests for each eye is presented in Table 1. Eyes 1-5 represent cases where only GDx classified the eyes as being borderline/glaucomatous, eyes 6-20 where one of the additional tests (mVEP, HRT or stereoscopic photographs) classified the eyes as being borderline/ glaucomatous, eyes 21-28 where two of the additional tests classified the eyes as being borderline/glaucomatous and eyes 29-32 where all the additional tests classified the eyes as being borderline/glaucomatous. Using a polling system of the five diagnostic tests to decide whether or not an eye is normal or suspect/glaucomatous would result in 63% (20/32) of the eyes being falsely labeled as suspect/glaucomatous by GDx and 37% (12/32) being falsely labeled as normal by HFA-GHT.

The percentage of agreement between the individual tests is presented in Table 2. The best agreement was found between HFA-GHT and mVEP data (69%). mVEP showed less than 50% agreement with HRT (47%) and GDx (34%). HRT data showed similar agreement with HFA-GHT (56%) and stereoscopic photographs (53%). In general, all tests demonstrated relatively poor agreement with GDx findings. GDx data agreed best with the stereo-photographs (38%).

In order to establish whether the performance of GDx could be improved, by looking solely at parameters based on ratios and differences between different peripapillary segments, which are less sensitive to errors in corneal compensation, the GDx inclusion criteria were modified. Eyes were only included if they had borderline or glaucomatous results from one or more of the following parameters: the Number, symmetry, superior ratio, inferior ratio, superior-nasal ratio, maximum modulation,

Table 1. The classification given by the different diagnostic tests for each eye; 32 eyes were included

N: normal; B: borderline; G: glaucoma; HFA-GHT: Humphrey Field Analyzer-Glaucoma Hemifield Test; mVEP: multifocal visual evoked potential; HRT: Heidelberg Retina Tomograph; Stereo-photo: color stereoscopic photographs of the optic disc

Table 2. Agreement between the individual tests. Agreement means that two tests both define an eye as normal, borderline, or glaucomatous

and ellipse modulation. The results are presented in Table 3. The sample was reduced to 23 eyes, and nine of the eyes originally classified by GDx as glaucomatous changed to borderline. A similar pattern to Table 1 was found when errors sensitive to the corneal compensator were largely removed.

Eyes 1-3 represent cases where only GDx classified the eyes as borderline/glaucomatous, eyes 4-13 where one of the additional tests (mVEP, HRT or stereoscopic photographs) classified the eyes as borderline/glaucomatous, eyes 14-20 where two of the additional tests classified the eyes as borderline/glaucomatous, and eyes 21-23 where all the additional tests classified the eyes as borderline/glaucomatous. Using the polling system of the five diagnostic tests again, the diagnostic performance of GDx increased to 43% (10/23) and HFA-GHT decreased to 57% (13/23).

The percentage of agreement between the individual tests for the refined inclusion criteria are presented in Table 4. The agreement between GDx and mVEP (17%) improved slightly, but the agreement remained similar with HRT (26%) and worse with the stereo-photographs (30%). The agreement between HFA-GHT and mVEP

Table 3. The classification given by the different diagnostic tests for each eye using the refined inclusion criteria. The eyes had to be borderline or outside normal limits in one or more of the following GDx parameters: the Number, symmetry, superior ratio, inferior ratio, superior-nasal ratio, maximum modulation, and ellipse modulation. Twenty-five eyes were included

Table 4. Agreement between the individual tests. Agreement means that two tests both define an eye as normal, borderline, or glaucomatous. The eyes had to be borderline or outside normal limits in one or more of the following GDx parameters: the Number, symmetry, superior ratio, inferior ratio, superiornasal ratio, maximum modulation, and ellipse modulation

See Table 1 for abbreviations

was also slightly better (74%), but the agreement was slightly poorer with HRT (48%) and the stereo-photographs (35%). mVEP showed similar agreement with HRT and the stereo-photographs as with the non-refined inclusion criteria.

Discussion

The results of this study emphasize how disparity between different diagnostic tests can occur in the early stages of glaucoma. This disparity can pose a dilemma in decision-making and in the clinical management of a suspected glaucoma patient.

Using a polling system of five tests to decide whether or not an eye is normal or

suspect/glaucomatous resulted in 69% (22/32) of the eyes being falsely labeled as suspect/glaucomatous by GDx and 32% (10/32) being falsely labeled as normal by the visual field test. These results suggest that GDx overcalled abnormality in our selected group of patients.

The GDx instrument used in this study had a fixed corneal compensator and Greenfield et al.4 established that considerable inter-subject variability in the axis of corneal birefringence exists, which may affect the RNFL thickness parameters. We determined whether the diagnostic performance of GDx could be improved by excluding parameters sensitive to variations in corneal birefringence. The number of eyes falsely labeled by GDx decreased from 63% to 57%, while the number of eyes falsely labeled as normal on the basis of the visual field went up from 37% to 43%. We confirmed previous study findings that GDx was better at separating normal from glaucomatous eyes when using parameters insensitive to variations in corneal birefrin-

gence.10,13-20

Recently, mVEP recording has been advocated for objective assessment of the visual field in suspect and glaucoma patients.8,11,21,22 Goldberg et al.11 demonstrated good agreement between subjective visual field loss and objective visual field loss, as measured with HFA and objective perimetry, respectively, in patients with glaucoma. This study found that the highest level of agreement was between HFA-GHT and mVEP. This result is not surprising given that they both measure visual function. However, approximately 30% of the eyes were classified as suspect/glaucomatous by mVEP, while subjective perimetry classified them as ‘within normal limits’. The majority of these eyes were also classified as suspect/glaucomatous by HRT and/or stereo-photographs as well as by GDx, suggesting that these eyes have ‘pre-perimetric’ glaucoma. Similar findings were reported by Goldberg et al.11 in a study in which 22 of their 37 glaucoma patients with no scotoma in the fellow eye had suspicious or glaucomatous optic disc appearance on stereo-photographs, and were classified as suspect/glaucomatous by mVEP. They proposed that multifocal objective perimetry might detect functional changes before conventional subjective perimetry.

The three structural measures (HRT, GDx, and stereo-photographs of the optic nerve head) showed a reasonable level of agreement with each other, although it was not as high as that between HFA-GHT and mVEP. A recent study demonstrated better agreement between mVEP and HRT classification of normal, borderline and glaucoma (71%) in suspected glaucoma patients than was found in this study.23 The number of patients included in this study is small in comparison to the 126 glaucoma suspects examined by Graham et al.23 Better agreement might be observed with a larger sample.

A significant linear correlation has been reported between RNFL thickness values, measured by GDx, and equivalent structural parameters measured with other imaging techniques.24,25 The best correlation of HRT and GDx parameters was found to be between optic disc area and inferior quadrant integral of RNFL thickness (r = 0.468).25 Nevertheless, the correlation coefficient is relatively low between these two parameters. The agreement level between GDx using parameters insensitive to variations in corneal birefringence and HRT using the classification system extracted from Moorfield’s linear regression analysis in a large sample remains to be determined.

Zangwill et al.26 compared GDx, HRT and optical coherence tomograph (OCT) parameters and qualitative assessment of stereo-photographs using receiver operating characteristic (ROC) curves. They showed that stereo-photographs, HRT and OCT

284 A. Bjerre et al.

measurements generally had greater sensitivities than GDx measurements. Our inclusion criteria differed by the fact we had not categorized the patients’ eyes into normal, suspect or glaucomatous, instead we performed the five diagnostic tests to determine the diagnosis of our patients’ eyes. Considerable overlap existed between the diagnosis of normal, borderline and glaucoma given by HRT, stereo-photographs and GDx, confirming previous studies.10,14,16,26-28 The fact that the instruments do not identify the same eyes as glaucomatous, borderline, or normal may suggest that they measure different features.

Wollstein et al.29 compared HRT using Moorfield’s linear regression analysis, with qualitative stereoscopic optic nerve head photographs assessed by glaucoma experts, and demonstrated that Moorfield’s analysis was more sensitive in differentiating between normal and early glaucomatous eyes. Of the 32 eyes included in our study (see Table 1), the stereo-photographs appear to have incorrectly diagnosed seven eyes as suspect/glaucomatous and the HRT analysis only incorrectly diagnosed four eyes as suspect/glaucomatous, suggesting that HRT may be better at separating normal from suspect/glaucomatous eyes. Nevertheless, the HRT analysis diagnosed two eyes as normal, while mVEP, stereo-photographs and GDx diagnosed these eyes as being suspect/glaucomatous, so overall HRT does not appear to be better at separating normal from suspect/glaucomatous eyes than the stereo-photographs. The inclusion criterion for early glaucoma patients in the Wollstein et al.29 study was a reproducible visual field defect. In our study, all patients had repeatedly normal GHT results. Therefore, the two studies are not directly comparable.

A study examining a larger number of suspect glaucoma patients with normal HFAGHT is required to determine whether Moorfield’s analysis is more sensitive for identifying early glaucoma than qualitative assessment of stereoscopic optic nerve head photographs.

The patients in this study were a selected group with unusual findings (consistent HFA-GHT within normal limits and borderline or outside normal limits GDx parameters) who were detected in a large population of patients attending the Manchester Royal Eye Hospital. This group does not represent a random sample of glaucoma patients and the agreement demonstrated between the tests will vary with altered inclusion criteria.

In summary, disparities in diagnostic tests for glaucoma were established in a small but important proportion of patients. The highest level of agreement was found between HFA-GHT and mVEP results. The diagnostic performance of GDx increased slightly when the potential errors from a fixed corneal compensator were taken into account, although the agreement with the other diagnostic tests remained fairly poor. HFA-GHT agreed better with the ‘majority-decision’ of the different diagnostic tests than GDx, and thus GDx findings should be given less weight in decision-making.

When disparities exist between different diagnostic tests for glaucoma, a battery of additional tests can help in clinical decision-making.

References

1.Chauhan BC, LeBlanc RP, McCormick TA, Rogers JB: Test-retest variability of topographic measurements with confocal scanning laser tomography in patients with glaucoma and control subjects. Am J Ophthalmol 118:9-15, 1994

2.Sommer A, Katz J, Quigley HA, Miller NR, Robin AL, Richter RC, Witt KA: Clinically detectable nerve fiber atrophy precedes the onset of glaucomatous field loss. Arch Ophthalmol 109:77-83, 1991

3.Quigley HA, Katz J, Derick RJ, Gilbert D, Sommer A: An evaluation of optic disc and nerve fiber layer examinations in monitoring progression of early glaucoma damage. Ophthalmology 99:19-28, 1992

4.Greenfield DS, Knighton RW, Huang X: Effect of corneal polarization axis on assessment of retinal nerve fiber layer thickness by scanning laser polarimetry. Am J Ophthalmol 129:715-722, 2000

5.Weinreb RN, Shakiba S, Sample PA, Shahrokni S, Van Horn S, Garden VS, Asawaphureekorn S, Zangwill L: Association between quantitative nerve fiber layer measurement and visual field loss in glaucoma. Am J Ophthalmol 120:732-738, 1995

6.Baseler HA, Sutter EE, Klein SA, Carney T: The topography of visual evoked response properties across the visual field. Electroencephal Clin Neurophysiol 90:65-81, 1994

7.Klistorner AI, Graham SL: Electroencephalogram-based scaling of multifocal visual evoked potentials: effect on intersubject amplitude variability. Invest Ophthalmol Vis Sci 42:2145-2152, 2001

8.Klistorner AI, Graham SL: Objective perimetry in glaucoma. Ophthalmology 107:2283-2299, 2000

9.Garway-Heath DF, Greaney MJ, Caprioli J: Correction for the erroneous compensation of anterior segment birefringence with the scanning laser polarimeter for glaucoma diagnosis. Invest Ophthalmol Vis Sci 43:1465-1474, 2002

10.Weinreb RN, Zangwill L, Berry CC, Bathija R, Sample PA: Detection of glaucoma with scanning laser polarimetry. Arch Ophthalmol 116:1583-1589, 1998

11.Goldberg I, Graham SL, Klistorner AI: Multifocal objective perimetry in the detection of glaucomatous field loss. Am J Ophthalmol 133:29-39, 2002

12.Wollstein G, Garway-Heath DF, Hitchings RA: Identification of early glaucoma cases with the scanning laser ophthalmoscope. Ophthalmology 105:1557-1563, 1998

13.Tjon-Fo-Sang MJ, De Vries J, Lemij HG: Measurement by nerve fiber analyzer of retinal nerve fiber layer thickness in normal subjects and patients with ocular hypertension. Am J Ophthalmol 122:220227, 1996

14.Tjon-Fo-Sang MJ, Lemij HG: The sensitivity and specificity of nerve fiber layer measurements in glaucoma as determined with scanning laser polarimetry. Am J Ophthalmol 123:62-69, 1997

15.Niessen AG, Van Den Berg TJ, Langerhorst CT, Greve EL: Retinal nerve fiber layer assessment by scanning laser polarimetry and standardized photography. Am J Ophthalmol 121:484-493, 1996

16.Choplin NT, Lundy DC, Dreher AW: Differentiating patients with glaucoma from glaucoma suspects and normal subjects by nerve fiber layer assessment with scanning laser polarimetry. Ophthalmology 105:2068-2076, 1998

17.Xu L, Chen PP, Chen YY, Takahashi Y, Wang L, Mills RP: Quantitative nerve fiber layer measurement using scanning laser polarimetry and modulation parameters in the detection of glaucoma. J Glaucoma 7:270-277, 1998

18.Lee VW, Mok KH: Retinal nerve fiber layer measurement by nerve fiber analyzer in normal subjects and patients with glaucoma. Ophthalmology 106:1006-1008, 1999

19.Trible JR, Schultz RO, Robinson JC, Rothe TL: Accuracy of scanning laser polarimetry in the diagnosis of glaucoma. Arch Ophthalmol 117:1298-1304, 1999

20.Sinai MJ, Essock EA, Fechtner RD, Srinivasan N: Diffuse and localized nerve fiber layer loss measured with a scanning laser polarimeter: sensitivity and specificity of detecting glaucoma. J Glaucoma 9:154-162, 2000

21.Klistorner AI, Graham SL, Grigg JR, Billson FA: Multifocal topographic visual evoked potential: improving objective detection of local visual field defects. Invest Ophthalmol Vis Sci 39:937-950, 1998

22.Graham SL, Klistorner A, Grigg JR, Billson FA: Objective perimetry in glaucoma: recent advances with multifocal stimuli. Surv Ophthalmol 43 (Suppl 1):S199-209, 1999

23.Graham SL, Klistorner AI, Balachandran C, Goldberg I: Comparison of objective tests in glaucoma: multifocal objective perimetry and Heidelberg Retina Tomography. ARVO abstract. Invest Ophthalmol Vis Sci 43:2157, 2002

24.Hoh ST, Greenfield DS, Mistlberger A, Liebmann JM, Ishikawa H, Ritch R: Optical coherence tomography and scanning laser polarimetry in normal, ocular hypertensive, and glaucomatous eyes. Am J Ophthalmol 129:129-135, 2000

25.Toprak AB, Yilmaz OF: Relation of optic disc topography and age to thickness of retinal nerve fibre layer as measured using scanning laser polarimetry, in normal subjects. Br J Ophthalmol 84:473478, 2000

26.Zangwill LM, Bowd C, Berry CC, Williams J, Blumenthal EZ, Sánchez-Galeana CA, Vasile C, Weinreb RN: Discriminating between normal and glaucomatous eyes using the Heidelberg Retina Tomograph, GDx Nerve Fiber Analyzer, and Optical Coherence Tomograph. Arch Ophthalmol 119:985-993, 2001

27.Schuman JS, Hee MR, Puliafito CA, Wong C, Pedut-Kloizman T, Lin CP, Hertzmark E, Izatt JA, Swanson EA, Fujimoto JG: Quantification of nerve fiber layer thickness in normal and glaucomatous eyes using optical coherence tomography. Arch Ophthalmol 113:586-596, 1995

28.Bathija R, Zangwill L, Berry CC, Sample PA, Weinreb RN: Detection of early glaucomatous structural damage with confocal scanning laser tomography. J Glaucoma 7:121-127, 1998

29.Wollstein G, Garway-Heath DF, Fontana L, Hitchings RA: Identifying early glaucomatous changes. Comparison between expert clinical assessment of optic disc photographs and confocal scanning ophthalmoscopy. Ophthalmology 107:2272-2277, 2000