03 The Contour Line

27

CHAPTER 3 THE CONTOUR LINE

3.1 The optic disc - Elschnig’s Ring The optic disc is the anterior end of the optic nerve and it is formed by the convergence and grouping of the retinal nerve fibers that go through common holes of the three ocular tunics. In an ophthalmological examination it looks like a yellowish pink circle, much lighter than the fundus color. In a binocular biomicroscopic examination, it looks like a disc. The lamina cribrosa coming from the sclera divides the optic disc into: an anterior part where the nerve fibers have no myelin (amyelinic), and a myelinic one posterior to it. This makes the anterior part transparent or translucid, which allows for its examination with the HRT (as it only scans transparent tissue). It is the only part accessible to this examination. The whole constitutes the optic nerve’s bulbar segment. The narrowest part that the fibers go through on their way within the bulbar segment of the optic nerve is located at the level of Bruch’s membrane (choroid vitreous sheet foramen). This foramen divides the intrabulbar segment of the optic nerve into two truncated cones that are joined by their minor bases (figure 3.1). It is thus divided into two portions: a) an anterior retinic portion and, b) a posterior chorioscleral portion. The Retinal Portion (a in figure 3.1) is made up of two protrusions: a nasal one (larger as it has a higher number of axons) and a temporal one (smaller). Between both protrusions there is a depression that may have an umbilical, or at times cylindrical, shape. This depression is a physiological cupping that is not central but eccentric, and which is located at the temporal half. The pigment epithelium is separated from the most peripheral fibers of the optic nerve by a conjunctive tissue known as intermedial Kuhnt tissue. In this part, the retinal central artery trunk crosses the whole retinal portion of the optic disc and is bifurcated. The arterial vessels are always medial to the venous vessels. The vein trunk, in contrast, is formed by the joining of two branches, one superior, and the other, inferior, at the level of the lamina cribrosa, i.e. more towards the back. The vessels are surrounded by glia without interposed retinal fibers. This glia becomes thick at the center where it surrounds the vessels, and it forms the central support meniscus of Kuhnt. The same tissue accompanies the vessels within the optic nerve, where it is called Elschnig’s intercalar tissue. The internal limiting retinal membrane (feet of Muller’s fibers) does not exist in the optic disc, instead there is glia. In the periphery, it continues

28 with the internal limiting membrane. This is the internal retinal limit of the optic disc, which, as mentioned before, is greater than the external one (scleral ring). This accounts for its cone trunk shape. The chorioscleral portion of the optic disc (figure 3.1) is that portion extending from the orifice formed by Bruch’s membrane, towards the back. It consists a truncated cone whose smaller base is anterior and larger base posterior. This portion is the optic nerve canal. The lamina cribrosa, which is at the level of the sclera, divides this chorioscleral portion into two parts: an anterior one and a posterior one. The anterior part is formed by amyelinic fibers and the posterior part by myelinic fibers. The lamina cribrosa is opaque and conjunctival in its posterior part. It is called scleral lamina cribrosa. The choroidal lamina cribrosa, which is glial and transparent, is located in front of the scleral lamina cribrosa, and at the level of the choroid. It should be stressed that the precribiform amyelinic portion has ectodermal glia (astroglia), and that the retrocribiform myelinic portion has ectodermal and mesodermal glia (astrodendroglia, microdendroglia and oligodendroglia ) Elschnig’s mesenchymal tissue forms, in its anterior part, the scleral spur (E in figure 3.1) and as a whole, it constitutes the scleral optic foramen, that when ophthalmoscopically visible, is referred to as Elschnig’s Scleral Ring. In front of it there is a transparent glial ring [1]. Everything inside the scleral ring is the optic disc, and everything outside it is the retina. The parapapillary zone is the one surrounding the scleral ring.

Fig. 3.1: R: retina, EP: pigment epithelium, B: Bruch membrane, C: choroid, E: sclera, l: choroidal lamina cribrosa, 2: scleral lamina cribrosa, p: scleral spur, a: retinal portion, b: chorioscleral portion. At the bottom right: scleral spur in detail; it forms Elschnig’s Ring. Its anterior projection, whose internal margin corresponds with the external limit of the optic disc is clearly seen. On the left side, the correlation with a topographic image.

29

Fig. 3.2 The internal margin of Elschnig’s Ring corresponds anatomically with the external margin of the optic disc, and therefore, with the external margin of the Neuroretinal rim [2, 3]. Figure 3.2 shows a histological section of a normal optic disc, where the scleral spur and the fibers that cross the scleral canal can be clearly seen. Notice the shape of the two cones joined at the vertex. Elschnig’s scleral ring can be most frequently seen in the temporal area of the optic disc, and at times it can be seen in the nasal area as well. However, it is more difficult to see it in the superior or inferior quadrant since there is a greater number of fibers there (figure 3.3). The image on the left shows that the superior and inferior quadrants are darker because it is a topographic image presenting the more elevated structures with darker colors, and the more depressed structures with lighter colors. It is more difficult to see Elschnig’s ring in normal optic discs than in pathological ones [4, 5], because retinal fibers are located on the anterior face of Elschnig’s ring, thus hindering observation. Elschnig’s ring can be seen almost entirely in severely damaged optic discs because all the retinal fibers have disappeared. In figure 3.3, in the image on the right, both the internal and external margins of the ring can be clearly seen in the temporal margin. The ring cannot be seen so clearly in the nasal rim due to the greater number of fibers crossing its anterior face and because of vessel disposition. At the poles the ring is completely covered by the great number of retinal fibers coming into the optic nerve. Figure 3.4 illustrates Elschnig’s ring at 360 degrees. This optic disc belongs to a late congenital glaucoma. This is made evident by the fiber loss in every sector, allowing the ring to be seen clearly.

30

Fig. 3.3

Fig. 3.4 From the center of the papilla towards periphery the following structures can be observed: - external margin of the cup = internal margin of the neuroretinal rim - external margin of the neuroretinal rim = external margin of the optic disc = internal margin of Elschnig’s ring - external margin of the Elschnig’s ring - parapapillary area In figure 3.5 we can see the same optic disc, where these limits are marked in an image similar to the one observed in the eye fundus, which is the summation image (on the right) and in a topographical image (on the left). It must be noted that the limits are accurate in both images and therefore, they are both taken into account when the contour line

31

Fig. 3.5 is drawn. There is a graph at the bottom of this figure that shows the disposition adopted by Elschnig’s ring with regard to HRT sections. As it is not parallel to the retina, it cannot be completely seen in any of the 32 planes. If the planes are displayed one after the other at high speed, the different parts of Elschnig’s ring combine together in such a way that the image of the whole is produced. This can be achieved with the Movie menu [5]. 3.2 Contour line drawing What makes Elschnig’s scleral ring so important is that its internal margin signals the external limit of the optic disc. It has been proven that this ring does not vary in structure nor dimensions throughout glaucoma evolution, because, as it belongs to the sclera it is not an elastic tissue [6]. If we measure the optic disc taking this structure as a limit, we will always be measuring the same thing since the frame of reference does not vary. This does not happen in the optic discs of children with congenital glaucoma [7]. The contour line is a tool for optic disc examination that serves as an aid to determine the external limit of the optic disc, because it is a line drawn at the internal margin of Elschnig’s ring. This is done with a mouse handled by the observer, who must carefully follow the edge of the internal margins of the ring. Most of the quantitative parameters provided by stereometric analysis depend exclusively on this drawing. Because of this, if the contour line is not drawn properly the parameters are not valid for optic disc analysis. Once the 32 planes have been processed (S, A, T), the Processing menu must be selected, then the Analyze Topography submenu must be chosen. Two images will appear on the screen: The summation or extended focus image (on the right) which is the most similar to what is usually seen in the eye fundus; and the topographical image (on the left), which presents the anterior structures in dark and the posterior ones with light colors [5].

32 The next step is to choose the Contour Line menu, within which two submenus must be chosen. The first one is the Draw option, which allows for contour line drawing with the mouse; the second one is the Display Images or Image Selection option which allows you to choose, from the image on the right, any of the 32 images obtained in the series, in order to make the drawing easier [5]. As seen in figure 3.3, the complete scleral ring path cannot be seen in the summation image, only a part of it can be seen. The observer must then draw the contour line with the mouse, on the visible parts, and he must go on drawing it on subsequent planes until completed, with no discontinuity. Figures 3.6a and b show the ring segments appearing in the images of the tomography. They also illustrate three different individual planes. The contour line can be drawn in either image (topographical or summation) since the computer automatically and simultaneously draws the line on the other image. This is very useful because sometimes much information can be found in the topographi-

Fig. 3.6a

Fig. 3.6b

33 cal image. It must be stressed that the contour line tracing should be drawn exactly on the internal rim of Elschnig’s ring, not a pixel inside or outside this structure. Furthermore, the line drawing in the first examination of the patient is very important since this very same contour line may later be exported to the subsequent tomographies for a follow-up, thus eliminating intraobserver variation. As we have mentioned before, this contour line may be drawn on one of the series images and then be exported to the mean. Also, it can be drawn directly on the mean without the aid of the 32 planes. Once it is drawn, the option Accept of the Contour Line menu, can be chosen in order to accept the new contour line. If the line were not complete, the computer would indicate so with a white circle on the place where it is discontinued. If it has been drawn over one of the original series, it should be exported to the mean with the Export function in the Contour Line menu. The correlation between the projection of the scleral ring’s internal rim and the con-

Fig. 3.7

Fig. 3.8

34 tour line drawing can be observed in figure 3.7. Figure 3.8 shows how the computer analyzes everything inside the contour line. The right side of the optic disc was not analyzed in order to show where the contour line was drawn. 3.3 Understanding the contour line Once the contour line is traced and accepted, the computer analyzes any structure inside of it. This includes a list of quantitative parameters, the stereometric analysis, the color surfaces analysis, the parameters studied by quadrants and octants, and various other different displays of results. Much of this information is explained in chapters 4 and 5. From now on we will devote ourselves to the explanation of those topics closely related to the contour line. Once the line is accepted, the computer automatically locates the Reference Plane, which is at 50 µm beneath the retinal surface. It is important to make it clear that it is not located 50 µm beneath the total optic disc surface, but rather beneath a sector of the temporal quadrant. Based on a study by Quigley and Addicks [8], and Airaksinen and Tuulonen [9], it was determined that the sector that least varies in thickness throughout glaucomatous optic disc changes is the temporal sector, more specifically, the angle between -4 and -10 degrees at the center of the papillomacular bundle. This sector becomes narrower only in the final stages of the disease. In a 44 year-old female patient with glaucoma due to goniodysgenesis, with an optic disc in the terminal stage and a severely damaged visual field (stage 3) the retinal thickness in every quadrant and octant was compared with normal values. Thickness decreased by 75% in the nasal quadrant, 50% in the superior temporal octant, 80% in the inferior temporal octant and 6% in the temporal quadrant. This shows that the loss of retinal thickness in the papillomacular bundle sector does not occur until the optic disc is severely damaged. This 6% loss might be due to the fact that the temporal quadrant covers 90 degrees and the sector where the plane is located includes only 6 degrees (see chapter 9). We can then say that the Reference Plane (software version 2.01) is located 50 µm beneath the retinal surface in the temporal sector, within the -4 to -10 degrees angle (papillomacular bundle) (figure 3.9) [5,8,9]. As the contour line goes through the superior and inferior poles, where there is a greater number of fibers than in the rest of the optic disc, its height in both of these sectors in relation to the reference plane increases, whereas it decreases in the nasal and temporal sectors. If we take the contour line and extend it linearly so as to cover the 360 degrees from 0 to 360, we can observe that its course has the appearance of a double camel hump or peak. The first hump represents the height increase of the contour line in relation to the reference plane as it goes through the superior pole of the optic disc. The second hump represents the inferior pole.

35

Fig. 3.9

Fig. 3.10 In 1993, J. Caprioli published a paper where the double hump image is obtained by a different method [10]. Caprioli then showed how the humps’ height decreases in glaucomatous patients when compared to a control group (figure 3.10). Figure 3.11 shows the spatial relationship between the contour line and the reference plane and also shows the location of the humps on the contour line in a threedimensional graph. Figure 3.12 is the same image in a two-dimensional graph referred to as height variation along the contour line. The height variation along the contour line graph represents one of the program’s most important piece of information. Together with the stereometric parameters it helps understand the optic disc condition, and unlike the previous ones, it helps distinguish a generalized defect from a localized one. Also, in this latter case, it allows to locate the defect within the 360 degrees. It also helps correlate the fiber defects in the visual field,

36

Fig. 3.11

Fig. 3.12 in the gray scale as well as in the scale of values. Figure 3.12 shows the contour line diagram exactly as it appears on the screen. Three different color lines can be seen in this figure: The green line is the contour line, the red line represents the reference plane, and the white line, the mean height of the contour line. This last line is often located between the green line and the red line [5]. On the right side of the graph there is a scale of absolute height values (axis z) and on the left, a scale of values relative to the mean height of the contour. The latter gives a negative value to anything located above the mean height of the contour line, and a positive value to anything below it. In other words, it considers the mean height of the contour line as the mean of the retinal surface in the optic disc contour. This is why we say that everything before the retinal surface has a negative value and everything behind it has a positive one.

37

Fig. 3.13

Fig. 3.14 As mentioned before, the line is extended in the direction from 0 to 360 degrees, that is to say that the line goes through the four quadrants in the following order: superior temporal quadrant (STQ), superior nasal quadrant (SNQ), inferior nasal quadrant (INQ) and inferior temporal quadrant (ITQ). On the right of the graph, the word tilted appears on the screen. It indicates that the computer has corrected the tomography in such a way that even if there were a certain inclination in the laser incidence angle, the line curve corresponds to a perfectly perpendicular beam incidence on the retina. The word relative indicates that coordinates along the z axis are relative to the average height of the peripapillary retinal surface. Href indicates the location of the reference plane along the z axis. Fig. 3.13 and 3.14 illustrate the same graphs but belonging to a normal optic disc and a pathological one. Fig. 3.13 shows the two humps in the contour line graph. In this case the superior one is smaller than the inferior one. It is important to note that the green line (contour

38 line) seems normal and is far away from the line representing the reference plane. The white line, which indicates the mean height of the contour line, is also far away from the reference plane, which indicates that the mean retinal thickness below the contour line is well maintained. Fig. 3.14 shows a pathological optic disc. The contour line profile is different from that of a normal optic disc. As in most cases, one of the first changes is the decrease, then the loss, of one or both humps. In this case, no localized depression can be observed throughout the contour line profile, which means that there are no localized defects or bundles of damaged fibers. However, a general loss in height of the contour line can be seen, which might be related to an increase in the mean defect of the visual field. It is also important to note that the white line (mean height of the contour line) is closer to the reference plane than in the above mentioned case (compare both yellow segments). The height of both humps has diminished. During the next examination of the same patient it will be very important to check that the contour line does not show any localized depression. Should this happen, it would be logical to eventually find a scotomatous defect in the visual field. Turning back to figure 3.14, this homogenous decrease of the mean height of the contour line should be related to a damaged neuroretinal rim with a concentric cupping that does not show any notch or localized thinning in all its course. If this were the case, the line would show a significantly worse depression or decrease in height which would be easy to locate in a quadrant or a segment. It should also be stressed that sometimes these localized depressions of the contour line may be due to the fact that though the examiner has drawn the contour line correctly, he has drawn that sector inside the actual optic disc limit, thus letting the contour line fall inside the cup. Similarly, the absence of both humps with a lack of localized depressions may be due to the fact that the examiner has drawn the contour line outside the internal margin of the scleral ring (i.e. outside the external optic disc limit). In light of all of this, we reaffirm the importance of a correct contour line drawing by experienced examiners who are well-acquainted with optic disc anatomy. 3.4 Basics and Fundamentals When the 1.10 software version was used, significant differences were found regarding the parameters resulting from the contour line drawing because it placed the reference plane at 320 µm below the contour line mean height. The differences were smaller in normals and became larger according to the severity of the optic disc glaucomatous damage. We must also bear in mind that every reference plane is relative and parallel to the retinal surface. If we select a point in the temporal optic disc area it may or may not correspond with the scleral level in the nasal zone according to the thickness of the nerve fiber layer. The problem with this reference level lay with the fact that it was chosen on the basis of reflectivity images, then it was written down and used for measurements. The next step was to select a point on the temporal contour line for the measurements. This new reference level was called papillomacular. With this new level or reference plane, good results were obtained from normal optic discs to advanced glaucomas,

39 where the differences between both groups are statistically significant (Airaksinen and Tuulonen). In version 1.11, a modification was introduced in the papillomacular reference level. Instead of choosing a point in the temporal area of the optic disc, an area between -4 and -10 degrees is now chosen. Consequently, the papillomacular bundle is taken from below the zero degree of the horizontal line, since the macula is anatomically slightly below it. A location 50 µm below the retinal surface, which is the reference level used in software versions 1.11 and 2.01, is taken within these 7 degrees. These 50 µm come from Dr. Pickli's studies of the primates' eyes where he discovered that the thickness of the nerve fiber layer in the papillomacular area is 50 µm [11], and from the studies of Quigley and Addicks [8]. We should keep in mind that studies on the location of the reference plane are still in process. When the optic disc is severely damaged, the reference plane may be located even lower than 50 µm due to a decrease in retinal thickness (even in the temporal sector in terminal optic discs). As a consequence of this, the results may show a greater volume of the neuroretinal rim and a lower cup volume in a longitudinal study. Extensive research is being conducted at present to achieve an automatic location of the reference plane, but based on the scleral reflex finding, which should be quantifiable [12]. 3.5 Papillomacular bundle persists normal in the course of optic nerve head deterioration in glaucoma evolution. A case report. In figure 3.15 the right eye of a patient is shown, presenting a damaged optic nerve head, which is in phase 4 of our classification (see chapter 10). As the stereometric analysis shows, the rim area is decreased while the cup area is increased. The contour line diagram shows a small retinal thickness remaining above the reference plane. There is no presence of localized nerve fiber bundle defects. The yellow arrows on the right side indicate the remaining neuroretinal rim, and the arrows on the picture on the left show the same structure in green and blue colors. Figure 3.16 shows the same optic nerve head in a 20 degrees field examination, where the fundus presents a diffuse loss of the RNFL, but just one nerve fiber bundle the papillomacular bundle - is still normal. Its reflectivity is superior in relationship to the rest of the retina, and the values for the contour profile are bigger. On the upper left picture, the topographic image shows that there are no localized defects. Only one bundle is darker because it is more anterior than the rest. On the right side, the papillomacular bundle appears more clearly than the rest. On the bottom left, the three-dimensional structure shows the normal bundle in the temporal quadrant, and on the right side, the reflectivity image shows more clearly this bundle. All these features indicate the fact that the papillomacular bundle remains intact up to the last stage of glaucoma evolution and it helps to hold the reference plane level along this area. The visual field of this patient is in stage 3 with the 15 central degrees conserved.

40 On the left eye of this patient we can observe the same phenomenon as in the right eye, and the visual field is almost the same (figures 3.17 and 3.18). We conclude that the present location of the new standard reference plane is good enough to measure the parameters of the optic nerve head along glaucoma evolution. Nevertheless, in some cases, it is very useful to control the Height Reference values in order to evaluate if the papillomacular bundle is still preserved.

Fig. 3.15

Fig. 3.16

41

Fig. 3.17

Fig. 3.18

42 Bibliography 1.

Sampaolesi R: Glaucoma. Second edition, 1994, p. 299-301. Editorial Panamericana.

2.

Elschnig A: Der normale Sehenerveneintritt des menschlichen Auges. Denkschrift der Kaiserlichen Akademie der Wissenschaften, mathematisch, naturwissenchaftliche. Band LXX : 219/1911.

3.

Jonas JB, Airaksinen PJ, Robert Y: Definitionsentwurf der intra- und parapapillären Parameter für die Biomorphometrie des Nervus Opticus. Klin Mbl Augenheilkd 1988;192-621.

4.

Sampaolesi JR, Sampaolesi R: Lecture: Study of optic nerve head normality. In: Curso y Simposio Argentino de Glaucoma, July 1995, Buenos Aires, Republica Argentina.

5.

Zinser, G: Heidelberg Retina Tomograph Operation Manual, Software Version 2.01. Heidelberg Engineering GmbH, Heidelberg, Germany, January 1997.

6.

Jonas JB: Biomorphometrie des Nervus Optikus. Stuttgart, Enke-Verlag, 1990.

7.

Sampaolesi R, Sampaolesi JR: Elschnig's ring does change. Special Lecture at 5th International Meeting on Scanning Laser Ophthalmoscopy, Tomography and Microscopy, San Antonio, Texas, October 1995 (in press).

8.

Quigley HA, Addiks EM: Quantitative studies of retinal nerve fiber layer in glaucoma. Arch Ophthalmol 1982;100:807-814.

9.

Airaksinen PJ, Burk ROW, Vihanninjoki K, Toulonen A, Alanco HI: Neuroretinal rim volume measurements with the Heidelberg Retina Tomograph. Glaucoma Update, Vol 5, edited by G.K Krieglstein, Kaden Verlag, pp. 91-93, 1993.

10. Boeglin RJ, Caprioli J: Actual clinical evaluation of the optic nerve in glaucoma. In: Clinicas Oftalmologicas de Norteamerica, Actualizaciones en Glaucoma. Editor Stamper RL, Editorial Inter-Medica, Buenos Aires, Argentina 1993, pp. 67-88. 11. Toulonen A: Factors determining the location of the standard reference plane. Imaging of the ONH and retina with the Heidelberg Retina Tomograph and Flowmeter, Inaugural European Symposium, University of Manchester, Dep. of Ophthalmology, February 17th and 18th, 1995. 12. Burk ROW: Analisis de la capa de fibras nerviosas. Imaging of the ONH and Retina with the Heidelberg Retina Tomograph and Flowmeter, Inaugural European Symposium, University of Manchester, Dep. of Ophthalmology, February 17th and 18th, 1995.

43