Chapter 217 – Retinal Nerve Fiber Layer Analysis
NEIL T. CHOPLIN
Patients at risk for glaucoma may have optic nerves and visual fields of normal appearance, but they may still have nerve fiber layer defects that are indicative of early, undetected glaucomatous damage. Thus, assessment of the retinal nerve fiber layer is important in the initial evaluation of the patient suspected of having glaucoma and in follow-up for detection of early damage or progression.
The normal human optic nerve is made up of 1.0-1.2 million axons of retinal ganglion cells, which converge at the optic disc. These fibers make up the retinal nerve fiber layer and lie in the inner retina, just below the internal limiting membrane. The distribution of nerve fibers within the nerve fiber layer is illustrated in Figure 217-1 . Fibers from the superior and inferior halves of the retina do not cross the horizontal midline and are separated from each other by a horizontal raphe. Macular fibers are oriented horizontally and make up the papillomacular bundle, which enters the optic nerve on the temporal side. Fibers on the temporal side of the disc that arise peripheral to the papillomacular bundle have to arch over the bundle to reach the optic nerve and are thus known as arcuate fibers. Fibers from the nasal side of the disc are oriented in a more radial fashion.
Visual field defects that arise from loss of discrete nerve fiber bundles take on the shape of the bundle that was lost, which gives rise to arcuate scotomas (from loss of arcuate fibers) and wedge-shaped
Figure 217-1 Normal anatomy of the retinal nerve fiber layer (right eye). The characteristic organization of the nerve fiber layer and the existence of a horizontal raphe explain the observed visual field defects that occur in glaucoma as bundles of nerve fibers are lost.
defects (from loss of radially oriented nasal fibers). Because glaucoma damage usually affects the temporal areas of the superior and inferior poles of the optic nerve early on, nasal defects and arcuate defects tend to occur first. The papillomacular bundle and nasal fibers that subserve the temporal field are affected relatively late in the disease process, which accounts for the preservation of central and temporal islands until the end stage is reached.
The anterior-posterior orientation of the nerve fibers with regard to the positions within the optic nerve head is not clear. Fibers from the more peripheral portions of the retina occupy more peripheral locations in the nerve head, and those from the more central locations are more central within the nerve. Fibers from the peripheral retina are thought to occupy more superficial positions within the nerve fiber layer (i.e., closer to the vitreous), while more central fibers are thought to lie closer to the sclera. The fibers are thought to cross each other (i.e., the more superficial peripheral fibers cross the deeper central fibers to become more peripheral in the optic nerve, while the deeper fibers become more central in the nerve) somewhere in the anterior portion of the nerve head. The layer of nerve fibers is expected to be thickest just before the fibers make the 90-degree turn into the nerve head, and progressively thinner peripherally. The distribution of fibers is not uniform around the nerve head, because the nerve fiber layer is thicker at the superior and inferior poles and thinner nasally and temporally.
The appearance of the nerve fiber layer is dependent upon the method used to visualize it. Ophthalmoscopically, this layer is seen most easily in eyes that are darkly pigmented and have clear media, particularly if the red-free filter is used. The nerve fiber layer pattern appears as bright striations, most obvious where the layer is thickest. The bright striations of the fiber bundles are offset by darker, elongated processes of Müller cell origin that surround the bundles. The striations of the nerve fiber layer are seen clearly in the superior region of the red-free fundus view shown in Figure 217-2 .
NERVE FIBER LAYER IN GLAUCOMA
Localized defects in the nerve fiber layer appear as wedge-shaped (not spindle-shaped), clearly defined areas that radiate from the optic disc and widen peripherally. They appear on red-free photographs or ophthalmoscopically as dark areas between the bright striations of an otherwise normal nerve fiber layer. Figure 217-2 illustrates a nerve fiber layer defect in the inferior bundle of a patient who has early glaucoma. Such localized nerve fiber layer defects have been reported in up to 20% of patients who have glaucoma. Diffuse loss of the retinal nerve fiber layer also may occur and results in decreased visibility of the layer, which may be difficult to detect, particularly in eyes in which the media is not clear. Comparison with the fellow eye may help discern any diffuse loss of fibers in one eye. Another indication of the loss of nerve fibers is increased visibility of retinal blood vessels, which normally are embedded within the nerve fiber layer and, thus. partially obscured. If the nerve fiber layer is lost diffusely, the vessels become more visible and sharper in appearance.
Figure 217-2 Red-free view in a patient who has early glaucomatous loss (left eye). The bright striations of the normal nerve fiber layer are visible in the superior one half of the fundus. Inferiorly, there is a clearly defined, wedge-shaped nerve fiber bundle defect (arrows). Note how the defect touches the optic disc and widens as it extends peripherally.
Techniques for Nerve Fiber Layer Analysis
No technique is available to count the number of axons within the optic nerve or nerve fiber layer. Because axons cannot be counted in vivo, indirect measures of axon “count” must be used. Various techniques have been used to evaluate the optic nerve and nerve fiber layer in glaucoma, as summarized in Table 217-1 .
Stereo photographs of the optic nerve are not very useful for evaluation of the nerve fiber layer, because this tissue is not seen well, particularly in a color photograph. Because the nerve fiber layer is the innermost retinal layer (after the internal limiting membrane), a short-wavelength light that focuses more anteriorly helps bring the layer into focus.
Red-free (green) light used in either direct ophthalmoscopy or slit-lamp biomicroscopy of the fundus with a 78D or 90D lens makes the nerve fiber layer easier to visualize. The nerve fiber layer may be examined using high-contrast, black-and-white photographs obtained with red-free light; grading systems are used to quantify the amount of detectable loss. 
Measurement of Retinal Contour
Examination of the nerve fiber layer by ophthalmoscopy or photographic grading is subjective and prone to variability. A more objective technique involves the measurement of the height of the retinal surface above a reference plane. The topography of the retinal surface may be determined by tomographic scans using a confocal scanning laser ophthalmoscope, such as the
TABLE 217-1 — SUMMARY OF TECHNIQUES FOR RETINAL NERVE FIBER LAYER ANALYSIS
Direct ophthalmoscope or slit lamp and 78D or 90D lens
Nerve fiber layer visibility is enhanced with short-wavelength light
Easy to perform using readily available equipment
May be difficult without clear media
Nerve fiber layer not easily seen in lightly pigmented fundi
Red-free, high-contrast fundus photography
Fundus camera with red-free filter
High-contrast black-and-white film and paper
Nerve fiber layer visibility is enhanced with short-wavelength light
Nerve fiber layer defects may be easy to detect
Requires skilled photographer and dilated pupil Requires dilated pupil Limitations of ophthalmoscopy apply
Retinal contour analysis
Scanning laser ophthalmoscope that can perform tomographic topography
Three-dimensional construction of retinal surface can measure retinal height above a reference plane— height is related to nerve fiber layer thickness
Easy to perform through undilated pupil No discomfort to patient Can image through most media opacities unless very dense
Equipment is expensive Height measurements depend upon location of reference plane Retinal thickness may not be true indirect measure of nerve fiber layer thickness
Optical coherence tomography
Optical coherence tomography unit
Uses reflected and backscattered light to create images of various retinal layers (analogous to the useof sound waves in ultrasonography)
Can differentiate layers within the retina, including the nerve fiber layer, with a 10?µm resolution Correlates with known histology
Equipment is expensive Requires dilated pupil Resolution may not be high enough to detect small changes
Scanning laser polarimetry
Scanning laser polarimeter
Birefringent properties of the nerve fiber layer cause a measurable phase shift of an incident polarized light proportional to the tissue thickness
Easy to perform through undilated pupil
No discomfort to patient
Can image through most media opacities, unless very dense
Resolution limited to size of a pixel (possibly as small as 1?µm)
Equipment is expensive
Measurements not correlated histologically in humans
Requires compensation for other polarizing media (e.g., cornea)
Heidelberg Retinal Tomograph. This instruments determine the retinal contour in three dimensions; the retina is imaged in multiple (usually 32) planes, and the images are combined into a contour map. A reference plane is set below the retinal surface, and the height of the peripapillary retina (theoretically directly proportional to the nerve fiber layer thickness) above the plane is measured. This method has a sensitivity of 73% for the detection of glaucomatous optic nerve damage and can be used to detect, over time, changes attributed to glaucoma progression.
Optical Coherence Tomography
High-resolution, tomographic, cross-sectional images of the retina may be obtained using a technique called optical coherence tomography. This imaging technique is similar in concept to ultrasonographic imaging, except that reflected and backscattered light is used to create the image, rather than sound waves. Interfaces occur at tissues of different optical densities, and layers within a tissue may be differentiated from each other, with a resolution of approximately 5–8?µm. The retinal nerve fiber layer is visualized easily using this technique, and thickness measurements are determined by computer analysis of the resultant image. Measurements of nerve fiber layer thickness using optical coherence tomography correlate well with structural (histological nerve fiber layer thickness measurements) and functional (visual fields) parameters in normal individuals and in patients who have glaucoma. 
Scanning Laser Polarimetry
When polarized light passes through the retinal nerve fiber layer, the birefringent property of the axons (attributed to microtubules within the axons) causes the polarized light to undergo a phase shift. Such polarized laser light that passes through the nerve fiber layer is reflected off of the outer eye layers, and the degree of phase shift in the light that returns is measured by a detector. The amount of phase shift, also known as retardation, is directly proportional to the amount of nerve fiber layer (microtubules) through which the incident light has passed; this gives an indirect measurement of the thickness of the tissue. Rapid scans across an area of retina (by moving the polarized laser light) enable measurements to be obtained for a given retinal area. This technique is known as scanning laser polarimetry.
The Nerve Fiber Analyzer is a scanning laser polarimeter that employs this measurement technique. Using a diode laser in the near infrared, nerve fiber layer measurements are obtained at 65,536 retinal points in a 15° × 15° grid centered on the optic nerve head. The measurement may be performed through an undilated pupil of at least 2?mm diameter, takes approximately 0.7 seconds to perform, and is not apparent to the patient (no flashes of light or other sensation). The acquired image is processed rapidly by the instrument’s software, which runs under Microsoft Windows on a personal computer. Normally, three images are obtained for each eye and then averaged to create a mean or baseline image. The reproducibility of the measurements is of the order of 5–8?µm per measured pixel.
Normative data for the Nerve Fiber Analyzer measurements have been obtained from many centers around the world and incorporated into the analytical software of the instrument, and the instrument has been renamed the GDx. Figure 217-3 is an example of a printout obtained from a normal eye using the GDx software. The color printout includes a picture of the area scanned (obtained by reflectance of the laser light) at the upper left. Next to the reflectance image is a color-coded “thickness” map of the retinal nerve fiber layer for the area scanned. Colors in the blue and black represent areas or lower retardation (below 60?µm), and red, orange, and yellow are used to represent regions of higher retardation, up to 140?µm. The ellipse denotes an area 1.75 disc diameters in size, centered on the optic nerve (determined by the edge of the nerve head as marked by the operator). The graph below the reflectance image (Nerve Fiber Layer) represents the retardation measurements along the ellipse, from temporal to superior to nasal to inferior and back to temporal. Note the typical “double-hump” distribution of nerve fibers found in normal individuals, with highest values superiorly and inferiorly and lower values nasally and temporally, which corresponds to the known anatomy of normal nerve fiber layers. The shaded area includes 95% of normal values, which enables ready identification of abnormal nerve fiber areas along the eclipse. The box below the retardation map, labeled “Deviation from Normal,” shows the difference within each quadrant between the measured values and those of age-corrected normal individuals, in microns. Shading indicates probabilities, with the legend given to the right of the box. The bottom of the printout summarizes various parameters, along with the probability of finding such values in the normal database. The latest version of the instrument, called the GDx VCC, incorporates individualized compensation for the birefringent of the anterior segment, and the software can display a pixel-by-pixel deviation for normal map.
Histopathological examination of the nerve fiber layer performed after scanning laser polarimetry in monkeys shows good correlation between retardation values and thickness measurements. In the monkey model, one degree of retardation corresponded to approximately 7.4?µm of thickness. Human measurements
Figure 217-3 Printout from the GDx software of the Nerve Fiber Analyzer; normal eye. A reflectance image, as well as the color-coded “thickness” map, is provided. Various measurement parameters are shown in the lower box.
Figure 217-4 Nerve fiber layer analysis of the eyes of a patient with suspected glaucoma, which shows asymmetrical loss of fibers in the superior bundle of the right eye.
have not yet been obtained. Retardation values obtained by scanning laser polarimetry correspond to known properties of the retinal nerve fiber layer in normal and glaucomatous eyes and may be used to detect glaucoma, with a reported sensitivity of 96% and specificity of 93%. Retardation values also correlate with some measures of visual field loss in patients who have glaucoma. Measurements of mean nerve fiber layer thickness in patients affected by ocular hypertension are statistically significantly lower than those of normal patients, with considerable overlap in the values. 
Scanning laser polarimetry may be used to detect glaucoma damage before standard automated perimetry does. The nerve fiber layer analysis ( Fig. 217-4 ) of a 66-year-old African-American man who has optic nerves of normal appearance and ocular hypertension demonstrates asymmetrical loss of nerve fibers from the superior bundle of the right eye compared with the left. In this symmetry analysis printout from the GDx software, the thickness curves from the right and left eyes are superimposed upon each other (Nerve Fiber Layer Both) and the loss of the superior peak in the right eye is readily visible (blue line). The serial analysis ( Fig. 217-5 ) may demonstrate change over time. In this example, a new nerve fiber layer defect developed over a 2.5-year period (left side of figure). A corresponding visual field defect also developed (right side of figure).
Assessment of the retinal nerve fiber layer appears to be more sensitive and specific for the evaluation of glaucoma damage, especially early in the course of the disease, than estimations of cup-to-disc ratio or other measures of the optic nerve head. Quantitative assessment of nerve fiber layer thickness holds
Figure 217-5 Change over time with the development of a new nerve fiber layer defect and corresponding visual field defect. The serial analysis of this patient with glaucoma shows the development of a new nerve fiber layer defect over a 2.5-year period. The blue pixels show the change from the initial image. A superior arcuate scotoma developed over the same period, corresponding to the area of nerve fiber layer loss.
promise for early detection of glaucoma and objective follow-up of disease progression. As measurement techniques become more refined and the characteristics of the retinal nerve fiber layer in normal individuals and patients with glaucoma become better understood, objective measures, such as optical coherence tomography and scanning laser polarimetry, may prove to be better than visual field tests for diagnosis and follow-up in patients who have glaucoma and for the recognition of early damage in patients at risk for glaucoma.
1. Tuulonen A, Lehtola J, Airaksinen J. Nerve fiber layer defects with normal visual fields: Do normal optic discs and normal visual field indicate absence of glaucomatous abnormality? Ophthalmology. 1993;100:587–98.
2. Migdal C. Optic nerve head in primary open angle glaucoma, pathogenesis and pathophysiology. In: Tasman W, Jaeger EA, eds. Duane’s ophthalmology, clinical vol. 3, ch. 52, record 35188-35223 on CD-ROM. Philadelphia: JB Lippincott; 1996.
3. Ogden TE. Nerve fiber layer of the macaque retina: retinotopic organization. Invest Ophthalmol Vis Sci. 1983;24:85–98.
4. Jonas JB, Dichtl A. Evaluation of the retinal nerve fiber layer. Surv Ophthalmol. 1996;40(5):369–78.
5. Jonas JB, Schiro D. Localised wedge shaped defects of the retinal nerve fiber layer in glaucoma. Br J Ophthalmol. 1994;78:285–90.
6. Niessen AG, van den Berg TJ, Langerhorst CT, Bossuyt PM. Grading of retinal nerve fiber layer with a photographic reference set. Am J Ophthalmol. 1995;120(5):57–86.
7. Quigley HA, Reacher M, Katz J, et al. Quantitative grading of nerve fiber layer photographs. Ophthalmology. 1993;100(12):1800–7.
8. Caprioli J, Miller JM. Measurement of relative nerve fiber layer surface height in glaucoma. Ophthalmology. 1989;96:633–41.
9. Caprioli J, Prum B, Zeyen T. Comparison of methods to evaluate the optic nerve head and nerve fiber layer for glaucomatous change. Am J Ophthalmol. 1996;121(6):659–67.
10. O’Connor DJ, Zeyen T, Caprioli J. Comparison of methods to detect glaucomatous optic nerve damage. Ophthalmology. 1993;100(10):1498–503.
11. Shuman JS, Hee MR, Puliafito CA, et al. Quantification of nerve fiber layer thickness in normal and glaucomatous eyes using optical coherence tomography. Arch Ophthalmol. 1995;113:586–96.
12. Dreher AW, Reiter K, Weinreb RN. Spatially resolved birefringence of the retinal nerve fiber layer assessed with a retinal laser ellipsometer. Appl Opt. 1992;31: 3730–5.
13. Weinreb RN, Dreher AW, Coleman A, et al. Histopathologic validation of Fourier ellipsometry measurements of retinal nerve fiber layer thickness. Arch Ophthalmol. 1990;108:557–60.
14. Weinreb RN, Shakiba S, Zangwill L. Scanning laser polarimetry to measure the nerve fiber layer of normal and glaucomatous eyes. Am J Ophthalmol. 1995; 195(5):627–36.
15. 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. 1997;123(1):62–9.
16. Weinreb RN, Shakiba S, Sample PA, et al. Association between quantitative nerve fiber layer measurement and visual field loss in glaucoma. Am J Ophthalmol. 1995;120(12):732–8.
17. 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. 1996;122(8):220–7.
18. Anton A, Zangwill L, Emdadi A, Weinreb RN. Nerve fiber layer measurements with scanning laser polarimetry in ocular hypertension. Arch Ophthalmol. 1997;115:331–4.