Malik Y. Kahook,
Robert J. Noecker,
Gadi Wollstein,
Joel S. Schuman
INTRODUCTION AND HISTORICAL PERSPECTIVE
Helmholtz first wrote of visualization of the optic nerve head (ONH) in 1851 using his invention, the direct ophthalmoscope.[1] Only 4 years later, von Graefe described the optic atrophy seen in glaucoma as 'amaurosis with excavation of the optic nerve'.[2] The first pathologic description of glaucomatous ONH changes are credited to Schnabel, who noted that damaged nerve fibers accompanied this type of ONH damage.[3] Fuchs and Elliot provided a detailed description of the glaucomatous changes in the 1920s. Fuchs described the sequential loss of the anterior and then deeper glial fibers with associated bowing backward and thinning of the lamina cribrosa due to elevated intraocular pressure. Fuchs believed that these changes were significant and preceded visual field loss.[4] Quigley's fidings of significant nerve fiber loss prior to the detection of glaucomatous field defects, as well as the continued field loss occurring during therapy prior to progressive changes on automated perimetry supports Fuchs' assertions.[5,6]
The diagnosis of glaucoma is dependent on the clinician detecting the requisite (ONH) damage along with the characteristic visual field loss. Automated and static perimetry can reliably and accurately detect visual field damage. A significant number of patients are unable to perform these tests due to disabilities or cognitive problems. This mode of testing is also limited in that as much as 40% of nerve fiber layer (NFL) may be lost prior to detection of a visual field defect by perimetric testing.[5,6] The status of the ONH has traditionally been left to subjective methods, such as funduscopic examination and evaluation of ONH photographs, even though accuracy and reliability in ONH assessment is essential to the diagnosis of glaucoma. The subjective nature of ONH evaluation has led to a significant number of patients being undetected and others followed as glaucoma suspects, and others being suboptimally treated during times of progression of glaucomatous damage. These facts have been behind the impetus to develop new techniques to objectively and reproducibly evaluate the ONH and NFL.
The need for more sensitive, accurate and reproducible methods of assessment has led to development of more sophisticated technologies. With the aid of these new techniques, clinicians are afforded additional tools with which to monitor and make clinical decisions for their patients. Early generations of ONH and NFL analyzers, though accurate, were impractical for widespread office use. Significant advancements in ONH imaging have evolved over the past quarter century with the development of sophisticated computer technology. These technologies have met the challenge of making ONH assessment more sensitive, accurate and reproducible while becoming more reasonably priced, user friendly and practical for office use. This chapter will summarize the important developments in imaging of the ONH and NFL using fundus photography, Heidelberg retinal tomography (HRT), GDx, and optical coherence tomography (OCT).
FUNDUS PHOTOGRAPHY OF THE ONH AND NFL
Fundus photography, the current gold standard for documentation of ONH structure, was first introduced in the early twentieth century, using the principles of the indirect ophthalmoscope. Stereoscopic fundus photography was performed using nonsimultaneous exposures with lateral shifts of the fundus camera or the Allen separator that creates a more consistent stereo base. Though stereophotography produces an objective image of the ONH, its interpretation still involves subjectivity, which in turn increases variability and limits its usefulness in the long-term evaluation of glaucoma patients.[7]Photogrammetric techniques have been used in an attempt to quantitate topographic cup parameters using stereophotographs. Though results may be variable, these techniques are limited in their applicability in routine clinical practice, primarily due to the need for an experienced observer.[8]
CONFOCAL SCANNING LASER OPHTHALMOSCOPE
|
Key Features |
||||||||||||
|
The HRT (Heidelberg Engineering, Heidelberg, Germany) is a confocal scanning laser ophthalmoscope (CSLO) that provides quantitative measurements of the ONH and posterior segment. While appropriate for the ONH, HRT is not used clinically for RNFL imaging. A retinal module, although rarely used clinically, allows for quantitative assessment of the macula. The HRT projects a 670nm wavelength light from a diode laser through a pinhole toward the posterior pole. Back-reflected light passes through a second conjugated pinhole to a detector, ensuring that only light reflected from a defied focal plane will reach the detector to be analyzed. A set of 16-64 sequential two-dimensional 384 × 384 pixel scans are then reconstructed to form a three-dimensional data set of reflection intensities. A topographical map is created representing the location of maximal intensity of light at each pixel (Fig. 197.1a,b).
|
|
|
|
FIGURE 197.1 (a) HRT II analysis of a normal left optic nerve showing green checkmarks indicating normal NFL thickness with MRA. (b) HRT II analysis of a glaucomatous right optic nerve showing red X's in areas of NFL thinning, and yellow in areas with borderline NFL thickness per MRA. |
After acquisition of the image, an observer must draw a 'contour line' along the margin of the disk in the reconstructed image. The contour line should be placed at the inner border of the scleral ring (Elschnig's ring) that can be identified by using the color difference between retina and ONH, bending of blood vessels, and by the border of peripapillary atrophy if present. The instrument then calculates the mean height along 6° of the contour line in the temporal inferior region. A plane at 50 ?m posterior to the mean height is designated as the cutoff between the neuroretinal rim and the cup of the ONH. The HRT3 has the added option of using stereometric data for the creation of a three-dimensional model as described by Swindale and colleagues.[9]
Glaucoma Detection
Twelve stereometric parameters are automatically obtained with the HRT2 after reconstruction of the data as explained above. HRT parameters are able to discriminate between glaucomatous and healthy ONHs and correlate well with location of visual defects.[10-14] Several analytic strategies, using data obtained from stereometric analysis, are used in combination to defie ONH measurements (see Table 197.1).
TABLE 197.1 -- HRT Analysis Strategies[15-20]
|
Analysis Strategy |
Description |
|
Moorfields regression analysis (MRA) |
Uses linear regression analysis between optic disc area and neuroretinal rim area in both global and localized segments |
|
Sensitivity is between 74% and 84% with a specificity between 81% and 96% |
|
|
Swindale analysis |
Allows for ONH analysis without manual outlining of optic disk |
|
Reliable and repeatable independent of subjective operator estimates |
Reproducibility and Progression Analysis
Studies of HRT reveal high reproducibility. Reproducibility is improved when multiple scans are averaged leading to the current practice of combining data from three sacns.[21] When multiple scans were averaged, the standard deviation (SD) of all image elements was improved from 35.5 to 25.7 ?m. These fidings were even more impressive in glaucoma patients who improved SD from 40.2 ?m with one scan to 28.5 ?m and 24.1 ?m with three and five scans respectively.
The glaucoma change probability analysis allows for the assessment of HRT data compared to a baseline examination. This program is based on the analysis of follow up testing compared to the baseline contour line that is exported by the user. A second method for analyzing progression is called the topographic change analysis. This method calculates a probability map based on the variability observed in three baseline scans and is illustrated through a color coded map.[22]
Kamal et al reported the capability of HRT to detect structural changes prior to the appearance of changes in perimetry.[23] They found that, in a group of ocular hypertensive patients, early HRT abnormalities later matched reproducible visual field defects. In a longitudinal study using analysis of variance, Chauhan et al showed that the HRT documented a higher rate of progression than the rate observed by perimetry.[22] Zangwill et al found that CSLO analysis of the optic nerves of patients enrolled in the ocular hypertension treatment study was associated with the future development of primary open-angle glaucoma (see Table 197.2).[24]
TABLE 197.2 -- Strengths and Weaknesses of HRT
|
HRT Strengths |
HRT Weaknesses |
||||||||||||||||||
|
|
SCANNING LASER POLARIMETRY
Scanning laser polarimetry (SLP), used clinically with GDx machines, utilizes the birefringence of the retinal NFL to measure thickness. The axons of the ganglion cells contain microtubules arranged in parallel, which change the polarization of passing light. The change in polarization, known as retardation, is given a numerical value which correlates with thickness of the NFL.[25] The current model of GDx uses a 780nm diode laser and measures a 15 × 15° area of NFL surrounding the ONH. The disk is defied by the operator manually at the beginning of the test. The quantitative information is then generated onto a 256 × 256 pixel color coded image with blue colors signifying thinner areas and yellow/red areas signifying areas of thicker NFL (Fig. 197.2a,b).
|
|
|
|
FIGURE 197.2 (a) GDx of normal optic nerves with thicker NFL in the superior and inferior quadrants illustrated by yellow and red colors in a bow-tie distribution. (b) GDx of a patient with bilateral glaucoma showing loss of normal bow-tie NFL distribution. The lower images apply color coding to illustrate the significance of tissue loss compared to normal controls. |
One of the problems with this technology is that the eye contains many birefringent tissues, including the cornea, lens, vitreous, and sclera. The main area of birefringence other than the NFL exists in the cornea. Studies have shown that there is a large variation in the polarization properties of the cornea among tested subjects.[26,27] For this reason, the new GDx model (fourth generation) contains a variable corneal compensator (VCC) which subtracts the retardation contributed by the anterior segment and theoretically provides data only from NFL retardation.
Glaucoma Detection
The GDx with VCC provides a printout containing information for both eyes side by side. The conversion of retardance to thickness is done by a fixed factor of 0.67 nm/?m. Multiple parameters are given in table format with color coded boxes indicating the presence of abnormal or normal values based on a comparison with 540 normal patients contained in the normative database. The most important value to recognize is the nerve fiber indicator (NFI) which correlates best with the existence of glaucomatous damage.[28] This value is calculated using a mathematical algorithm based on several NFL parameters and assigns a number between 0 and 100 for each eye. A number between 0 and 30 is considered normal, between 30 and 50 is borderline, and a value of 50-100 is considered highly suspicious for glaucoma.
Other important values of importance include the inferior normalized area and temporal, superior, nasal, inferior, temporal (TSNIT) average, both of which have been highly correlated with the presence or absence of glaucoma.[28] A graph comparing the NFL of the patient to that of the normative database is included at the bottom of the printout to provide added information. Dilation has not been found to increase the ability of disease detection or quality of obtained scans in previous reports.[29,30]
Reproducibility and Progression Analysis
Many studies have been published on the reproducibility of GDx; however, most of these reports did not include the VCC. Without VCC, GDx has been shown to be reproducible in both normal and glaucomatous eyes.[30] Zangwill and colleagues reported mean coefficient of variance of 4.2% for interoperator reproducibility using an older model.[31] More studies are needed to describe the reproducibility of the GDx-VCC (see Table 197.3).
TABLE 197.3 -- Strengths+ and Weaknesses of GDx
|
GDx Strengths |
GDx Weaknesses |
|
No pupil dilation needed in general |
Few published studies using VCC technology |
|
Provides rapid data on the NFL |
Depends on macula as internal reference for VCC to work |
|
Large normative database |
Evaluates the peripapillary region only |
OPTICAL COHERENCE TOMOGRAPHY
Stratus OCT (Carl Zeiss Meditec, Inc, Dublin, CA) permits high-resolution real-time cross-sectional imaging of the retina. OCT obtains images by measuring the echo time delay and intensity of back-reflected light from retinal tissues in a similar manner to B-mode ultrasonography. Differences in retinal layers can be illustrated due to the unique time delay of the reflections from the various tissue components. The light source used is from a near-infrared, low-coherence superluminescent diode at a wavelength of 810nm. The exiting light is split into two arms with one beam entering the eye and the other sent to a reference mirror. If the path length of the two arms is closely matched to within the coherence length, an interference signal is detected and translated into a two dimensional color coded representation of the retinal layers.
Glaucoma Detection
NFL scanning involves a circular scan pattern with a 3.4 mm diameter centered on the ONH which is then displayed as a flat cross-sectional image (Fig. 197.3a,b). Characteristic changes in reflectivity observed at the inner and outer retinal boundaries are automatically converted to quantitative NFL thickness data by a computer algorithm. The data are then divided into and displayed in four quadrants: superior, inferior, nasal and temporal, as well as in 12 clock hour sections of 30° each. A program to analyze the ONH structure is also available. The boundary of the ONH is defied by the termination of the retinal pigment epithelium, thus minimizing or eliminating the need for the user to identify the disk margin.
|
|
|
|
FIGURE 197.3 (a) Normal OCT showing the patient's NFL (black line) falling within normal thickness (green). Note the signal strength (upper right) recorded as nine and eight for the right and left eyes respectively. Signal strength should be above six in a quality study. (b) Abnormal OCT nerve fiber analysis in a patient (black line) falling in the yellow (borderline) and red (abnormal) compared to the normative database. The signal strength is eight for both eyes. The middle table color codes the fidings for various parameters and records average nerve fiber thickness (bottom of table) which in this case is marked red (abnormal) in the left eye and yellow (borderline) for the right eye. |
OCT is unique among nerve fiber analysis instruments in that it also can provide detailed information about the macula. The machine is also able to scan the macula using six equally spaced 6 mm linear scans in a spoke-like radial configuration. A color coded map is then created, illustrating the retinal thickness for specific regions in the macula. Real-time computer generated images representing data obtained from reflectance also provide information about tissue structures and swelling in all layers of the retina and subretina up to the choriocapillaris.
Reproducibility and Progression Analysis
Stratus OCT demonstrated reproducible measurements of NFL thickness, macular thickness, and ONH parameters in a study by Paunescu and colleagues.[32] They reported intervisit SD of 2.4 ?m and an intravisit SD of 2.2 ?m for mean macular scans using high-density scanning (512 A-scans) with pupil dilation. Mean NFL thickness scanning revealed intravisit SD of 1.6 ?m and intervisit SD of 2.5 ?m also with dilated pupils. While dilation is not necessary to obtain quality images with OCT, it did appear to improve reproducibility in this study.
Longitudinal glaucoma progression studies show that OCT may be effective at discovering pathology prior to standard automated perimetry. Wollstein et al assessed the NFL thickness measurements changes in glaucoma patients and glaucoma suspects in a longitudinal study and found a greater likelihood of progression with OCT as compared to VF.[33]
Recently a new iteration of the OCT, using Fourier-domain technology, has been developed with high resolution and greatly increased speed of image acquisition.[34] This device represents the next step in obtaining detailed retinal cross sections and allows for new scanning patterns. Still in the testing phase, this technology requires further evaluation to better understand its capabilities (see Table 197.4).
TABLE 197.4 -- Strength and Weaknesses of OCT
|
OCT Strengths |
OCT Weaknesses |
|||||||||||||||
|
|
REFERENCES
1. Helmholtz J: Beschreiburg eines Augenspiegels zur Untersuchung der Netzhaut in lebenden Augi, Berlin, A Forstner, 1851.
2. Von Graefe A: Ueber die wirkug der Iridectomie bei Glaucom. Arch Ophthalmol 1857; 3:456.
3. Schnabel I: Die Entwicklungsgeschichte der glaukomatosen. Exkavation Z Augenheilkd 1905; 14:1.
4. Elliot RU: Treatise on glaucoma, London: Henry Fraude and Hodder & Stroughton LTD; 1922:195.
5. Quigley HA, Addicks EM, Green WR: Optic nerve damage in human glaucoma. Arch Ophthalmol 1982; 100:135.
6. Quigley HA, Miller NR, George T: Clinical evaluation of nerve fiber layer atrophy as an indicator of glaucomatous optic nerve damage. Arch Ophthalmol 1980; 98:1564.
7. Tielsch JM, Katz J, Quigley HA: Intraobserver and interobserver agreement in measurement of optic disc characteristics. Ophthalmology 1988; 95:350.
8. Takamoto T, Schwartz B: Reproducibility of photogrammetric optic disc cup measurements. Invest Ophthalmol Vis Sci 1985; 26:814.
9. Swindale NV, Stjepanovic G, Chin A, Mikelberg FS: Automated analysis of normal and glaucomatous optic nerve head topography images. Invest Ophthalmol Vis Sci 2000; 41:1730-1742.
10. Uchida H, Brigatti L, Caprioli J: Detection of structural damage from glaucoma with confocal laser image analysis. Invest Ophthalmol Vis Sci 1996; 37:2393-2401.
11. Zangwill LM, van Horn S, de Souza Lima M, et al: Optic nerve head topography in ocular hypertensive eyes using confocal scanning laser ophthalmoscopy. Am J Ophthalmol 1996; 122:520-525.
12. Iester M, De Ferrari R, Zanini M: Topographic analysis to discriminate glaucomatous from normal optic nerve heads with a confocal scanning laser: new optic disk analysis without any observer input. Surv Ophthalmol 1999; 44(Suppl 1):S33-S40.
13. Iester M, Mikelberg FS, Courtright P, Drance SM: Correlation between the visual field indices and Heidelberg retina tomograph parameters. J Glaucoma 1997; 6:78-82.
14. Gulati V, Agarwal HC, Sihota R, Saxena R: Correlation analysis of visual field thresholds and scanning laser ophthalmoscopic optic nerve head measurements in glaucoma. Ophthalmic Physiol Opt 2003; 23:233-242.
15. Miglior S, Casula M, Guareschi M, et al: Clinical ability of Heidelberg retinal tomograph examination to detect glaucomatous visual field changes. Ophthalmology 2001; 108:1621-1627.
16. Miglior S, Guareschi M, Albe E, et al: Detection of glaucomatous visual field changes using the Moorfields regression analysis of the Heidelberg retina tomograph. Am J Ophthalmol 2003; 136:26-33.
17. Zangwill LM, Bowd C, Berry CC, et al: Discriminating between normal and glaucomatous eyes using the Heidelberg retina tomograph, GDx nerve fiber analyzer, and optical coherence tomograph. Arch Ophthalmol 2001; 119:985-993.
18. Wollstein G, Garway-Heath DF, Hitchings RA: Identification of early glaucoma cases with the scanning laser ophthalmoscope. Ophthalmology 1998; 105:1557-1563.
19. Ford BA, Artes PH, McCormick TA, et al: Comparison of data analysis tools for detection of glaucoma with the Heidelberg retina tomograph. Ophthalmology 2003; 110:1145-1150.
20. Swindale NV, Stjepanovic G, Chin A, Mikelberg FS: Automated analysis of normal and glaucomatous optic nerve head topography images. Invest Ophthalmol Vis Sci 2000; 41:1730-1742.
21. Weinreb RN, Lusky M, Barsch DU, Morsman D: Effect of repetitive imaging on topographic measurements of the optic nerve head. Arch Ophthalmol 1993; 111:636-638.
22. Chauhan BC, Blanchard JW, Hamilton DC, LeBlanc RP: Technique for detecting serial topographic changes in the optic disc and peripapillary retina using scanning laser tomography. Invest Ophthalmol Vis Sci 2000; 41:775-782.
23. Kamal DS, Viswanathan AC, Garway-Heath DF, et al: Detection of optic disc change with the Heidelberg retina tomograph before confirmed visual field change in ocular hypertensives converting to early glaucoma. Br J Ophthalmol 1999; 83:290-294.
24. Zangwill LM, Weinreb RN, Beiser JA, et al: Baseline topographic optic disc measurements are associated with the development of primary open-angle glaucoma: the Confocal Scanning Laser Ophthalmoscopy Ancillary Study to the Ocular Hypertension Treatment Study. Arch Ophthalmol 2005; 123:1188-1197.
25. Weinreb RN, Shakiba S, Zangwill L: Scanning laser polarimetry to measure the nerve fiber layer of normal and glaucomatous eyes. Am J Ophthalmol 1995; 119:627-636.
26. Greenfield D, Knighton RW, Huang X: Effect of corneal polarization axis on assessment of retinal nerve fiber layer thickness by scanning laser polarimetry. Am J Ophthalmol 2000; 129:715-722.
27. Knighton RW, Huang X: Linear birefringence of the central human cornea. Invest Ophthalmol Vis Sci 2002; 43:82-86.
28. Medeiros FA, Zangwill LM, Bowd C, Weinreb RN: Comparison of the GDx VCC scanning laser polarimeter, HRT II confocal scanning laser ophthalmoscope, and stratus OCT optical coherence tomograph for the detection of glaucoma. Arch Ophthalmol 2004; 122:827-837.
29. Hoh S, Greenfield D, Liebmann J, et al: Effect of papillary dilation on retinal nerve fiber layer thickness measured by scanning laser polarimetry in eyes with and without cataract. J Glaucoma 1999; 8:159-163.
30. Hollo G, Suveges I, Nagymihaly A, Vargha P: Scanning laser polarimetry of the retinal nerve fiber layer in primary open angle and capsular glaucoma. Br J Ophthalmol 1997; 81:857-861.
31. Zangwill L, Berry CA, Garden VS, Weinreb RN: Reproducibility of retardation measurements with the nerve fiber analyzer II. J Glaucoma 1997; 6:384-389.
32. Paunescu LA, Schuman JS, Price LL, et al: Reproducibility of nerve fiber thickness, macular thickness, and optic nerve head measurements using StratusOCT. Invest Ophthalmol Vis Sci 2004; 45:1716-1724.
33. 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.
34. Wojtkowski M, Bajraszewski T, Gorczynska I, et al: Ophthalmic imaging by spectral optical coherence tomography. Am. J Ophthalmol 2004; 138:412-419.
