Optic Disc, Cup and Neurorefinal Rim Size, Configuration ond Correlations in Normal Eyes

Transcription

1 Investigative Ophthalmology & Visual Science, Vol. 29, No. 7, July 1988 Copyright Association for Research in Vision and Ophthalmology Optic Disc, Cup and Neurorefinal Rim Size, Configuration ond Correlations in Normal Eyes Jost Bruno Jonas, Gabriele Charlotte Gusek, and Gottfried Otto Helmut Naumann Four hundred and fifty-seven unselected normal human optic nerve heads of 319 subjects (163 men, 156 women, mean age 42.7 ± 19.6 years) were evaluated by magnification-corrected morphometry of optic disc photographs. Mean optic disc surface measured 2.69 ± 0.70 mm 2 ( mm 2 ), mean diameter horizontally 1.76 ± 0.31 mm ( mm), and vertically 1.92 ± 0.29 mm ( mm). The form was slightly vertically oval. Optic cup area averaged 0.72 ± 0.70 ( mm 2 ), mean horizontal cup diameter 0.83 ± 0.58 mm ( mm) and mean vertical diameter 0.77 ± 0.55 mm ( mm). The cup was significantly (P < ) larger in discs with steep "punched-out" cups (1.37 ± 0.62 mm 2 ) compared to discs having cups with temporal flat slopes (0.59 ± 0.39 mm 2 ). Neuroretinal rim area ranged from 0.80 to 4.66 mm 2 (mean 1.97 ± 0.50 mm 2 ), and was significantly correlated (P < ) to the optic disc area. It was broadest in the inferior optic disc region (P < 0.001), followed by the superior, nasal and temporal (P < 0.001) regions. Horizontal cup/disc ratio (mean 0.39 ± 0.28, minimum 0.00, maximum 0.87) was larger in 426 (93.2%) optic discs than the vertical one (mean 0.34 ± 0.25, minimum 0.00, maximum 0.85). Concerning optic disc area, side differences of 0.10 mm 2 or less were detected in 28%, of 0.2 mm 2 or less in 46% and of 0.50 mm 2 or less in 80% (cumulative frequencies). Concerning neuroretinal rim area, side differences of 0.10 mm 2 or less were found in 31%, of 0.20 mm 2 or less in 52% and of 0.50 mm 2 or less in 84%. There were no significant correlations between these morphometric optic disc data and refraction, age, sex or side. Invest Ophthalmol Vis Sci 29: , 1988 Optic nerve head size has been often regarded as interindividually constant, and there have been only few reports on its absolute dimensions. 1 " 5 The purpose of this study was to measure in absolute units the optic disc, cup and neuroretinal rim area, diameters and width, and the c/d ratios in unselected eyes with no sign of any optic nerve disease, and further to correlate these values with each other and ocular and general parameters. The measurements were performed by magnification-corrected morphometry of optic disc photographs. Materials and Methods Four hundred and fifty-seven eyes (221 right and 236 left eyes) of 319 unselected patients (163 men, 156 women) attending the university eye hospital in Erlangen, 1986 and 1987, were the basis of this prospective study. Informed consent was received from From the University Eye Hospital, Erlangen-Niirnberg, West Germany. Supported by Deutsche Forschungsgemeinschaft DFG Nr NA 55/4-1 and Jo 155/2-1 and "Muck-Foundation." Submitted for publication: August 6, 1987; accepted February 5, Reprint request: Dr. J. B. Jonas, University Eye Hospital, Schwabachanlage 6, D-8520 Erlangen, West Germany. each patient prior to study. In all cases any optic nerve disease, especially glaucoma, had been ruled out by history, slit-lamp biomicroscopy, gonioscopy, and fundus examination. Manual kinetic (Goldmann; Haag-Streit, Bern, Switzerland) and/or computerized static (Octopus, program 32, 34; Interzeag AG, Schlieren, Switzerland) perimetry and retinal nerve fiber layer photography, performed as described by Airaksinen, 6-7 were unremarkable if they had been performed as part of another study. Intraocular pressure ranged between 10 and 21 mm Hg and visual acuity between 10/20 to 20/20 or better. The latter depended on the degree of beginning cataract or age-related macular degeneration. There was no history or sign of any intraocular operation in these eyes. The patients came to the eye hospital because of a regular ocular check-up, prescribing of glasses, or diseases in the contralateral eye not primarily affecting the optic nerve. Mean patient age was 42.7 ± 19.6 years, mean refraction ± 2.35 diopters (-7.50 dpt to dpt). High myopes with a myopic refraction of more than diopters and eyes with acquired lens-induced myopia of more than 1 diopter had been excluded. By morphologic criteria three subgroups were defined: (1) discs without cupping; (2) discs with cups 1151

2 1152 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / July 1988 Vol. 29 NASAL TEMPORAL Fig. 1. Optic disc having a cup with temporalflatslope; optic disc area: 2.81 mm 2 ; horizontal diameter 1.72 mm; vertical diameter: 2.05 mm; optic cup area: 0.84 mm 3 ; horizontal diameter 1.12 mm; vertical diameter 1.02 mm; neuroretinal rim area: 1.97 mm 2 ; horizontal c/d ratio: 0.65; vertical c/d ratio: having a temporal flat slope (Fig. 1); and (3) discs with steep "punched out" cups (Fig. 2). Stereo optic disc photographs (IS degree color slides) were taken with a telecentric Zeiss camera equipped with a twofold magnification adapter and an Allen stereo separator. The slides were projected 10, 15 or 20 times magnified, the optic disc and cup outlines were plotted and morphometrically analyzed (Zeiss Morphomat 30). The optic disc was that area lying inside the inner circumference of the whitish peripapillary scleral rim. The edge of the optic cup was denned using contour changes, not pallor, as the criterion. The neuroretinal rim resulted from subtraction of the cup from disc area. Denning the cup border in the temporal disc region was more difficult in the discs having cups with temporal flat slopes than in the discs Fig. 2. Optic nerve head with steep, punched-out optic cup; optic disc area: 4.51 mm 2 (macrodisc); horizontal diameter 2.30 mm; vertical diameter: 2.45 mm; optic cup area: 2.31 mm 2 ; horizontal diameter 1.80 mm; vertical diameter: 1.65 mm; neuroretinal rim area, total: 2.20 mm 2 ; temporal upper optic disc sector 0.53 mm 2 ; temporal lower sector: 0.72 mm 2 ; horizontal c/d ratio: 0.78; vertical c/d ratio: Fig. 3. Optic disc divided into four sectors. Sectors II and III are right-angled, and their middle axes (double dotted lines) are tilted 13 (angle 0) temporal to the vertical optic disc axis. Sectors I (temporal side, 64 ) and IV (nasal side, 116 ) cover the remaining areas. with steep "punched-out" cups. In every case, however, a differentiation between cup and neuroretinal rim was possible by the use of red-free light for contrast enhancement and by the stereoscopical evaluation of the photographs. The photographic magnification was corrected by Littmann's method. 8 This takes into account the constant magnification factor of a telecentric fundus camera and the individual ocular magnification factor, the latter depends mainly on the anterior corneal curvature, refraction and axial length, respectively, while posterior corneal curvature, anterior chamber depth, anterior and posterior lens curvature and lens thickness are less important. Littmann showed that a modification of 20% of the corneal thickness, of the anterior chamber depth, of the lens thickness, or the lens radii results in a change of the individual ocular magnification factor of 0.07%, 0.6%, 0.5% and 4.6%, respectively. A 20% variation of the anterior corneal curvature, however, causes a modification of the ocular magnification factor of 14.6%." Littmann's method can give false values in the case of an abnormally high lens refractive power (eg, by cataract formation) and it is so far not applicable in aphakic or pseudophakic eyes. In all examined eyes the refraction had been determined, and the anterior corneal curvature had been measured using a Zeiss ophthalmometer. According to Littmann's method the ocular magnification factor was evaluated, taking these two parameters. It was checked by the value of the axial length in those eyes where the axial length of the globe had been measured sonographically in the framework of an extracapsular cataract extraction (in the contralateral eye). To rule out the possible effect of an abnormally high lens refractive power, no eyes with cataract-induced myopization of more than 1 diopter were taken.

3 No. 7 OPTIC DISC, CUP AND RIM AREA AND FORM IN NORMAL EYES / Jonas er ol Concerning optic disc and cup, we measured area, horizontal, vertical, minimal and maximal diameters, quotient of minimal to maximal diameter, the angle between the maximal diameter and the horizontal and a form factor. The latter ranged from 1.0 for an ideal circle to 0.0 for a completely irregular structure. The measurements were performed by one of two investigators (JJ and GG). There was no significant difference in age, sex, refraction and results between the group of optic discs evaluated by the first examiner and the group measured by the second. In optic nerve heads with cupping the disc surface was divided up into four sectors (Fig. 3). The middle axis of two right-angled quadrants was turned 13 to the temporal disc side, and two further sectors covered the rest of the optic disc area (temporal 64, nasal 116 ) (Fig. 3). In these sectors optic disc and neuroretinal rim areas were determined separately. Total neuroretinal rim area was the sum of the rim areas in the four sectors equal to the difference of total disc area minus cup area. Neuroretinal rim width was measured every 30. Horizontal and vertical c/d ratios were calculated according to the determined horizontal and vertical cup and disc diameters. Side differences were evaluated in 138 patients in whom both eyes had been photographed. Total optic disc, cup and neuroretinal rim area, rim area in the four sectors, rim width, horizontal and vertical c/d ratios, refraction, age and sex were correlated with each other. Wilcoxon or Wilcoxon-Mann-Whitney tests were used for proving statistical significance of differences, if not indicated otherwise. Results Optic Disc Optic disc area ranged from 0.80 mm 2 to 5.54 mm 2 (mean and standard deviation: 2.69 ± 0.70 mm 2 ) with an interindividual variability of 1:6.9 (Fig. 4, 0.90 B l.sl SI B1 4. SI S. 81 S.S S. SB 6.00 Optic Disc Area <nn'> Fig. 4. Histogram (Gaussian distribution curve) of the area of 457 unselected, normal human optic nerve heads. Mean value: 2.69 ± 0.85 mm 2, median: 2.56 mm 2. Table 1). The values were arranged in a Gaussian distribution curve (David, Pearson and Stephens test, Alpha < 0.05). Mean optic disc diameter was horizontally 1.76 ± 0.31 mm ( mm), and vertically 1.92 ± 0.29 mm ( mm). The quotient of minimal to maximal diameter (mean 0.88 ± 0.05), the form factor (mean 0.96 ± 0.05) and the angle between the maximal diameter and the horizontal (mean 85.9 ± 30.4 degree) indicated a slightly vertically oval optic disc form with the vertical axis being about 9% longer than the horizontal one. Optic nerve heads with steep, "punched-out" cups (3.29 ± 0.75 mm 2 ) were significantly (P < ) larger than the average one. They were also larger than optic discs having cups with temporal flat slopes (2.57 ± 0.68 mm 2 ) or discs without cupping (2.05 ± 0.59 mm 2 ). Discs having cups with the temporal flat slope differed insignificantly in size from the unselected normal ones, but were significantly (P < ) larger than discs without cupping and smaller than discs with punched-out cupping. Optic nerve heads without cupping were significantly (P < ) smaller than the average and than all other disc types. Table 1. Morphometric optic disc data of 457 unselected, normal human optic nerve heads divided morphologically into three subgroups 0 / // III Number Disc area (mm 2 ) Diameter (mm) Horizontal Vertical Minimal Maximal Minimal/maximal diameter Angle maximal diameter/horizontal Form factor ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± = All optic discs. I = Optic discs without cupping. II = Optic discs having cups with temporal flat slopes. Ill = Optic discs with punched-out cupping.

4 1154 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / July 1988 Vol. 29 Table 2. Morphometric optic cup data of 457 unselected, normal human optic nerve heads divided morphologically into three subgroups Number Cup Area (mm 2 ) Diameter (mm) Horizontal Vertical Minimal Maximal Minimal/maximal diameter Angle maximal diameter/ horizontal Form factor ± ± ± ± ± ± ± ± ± ± ± = All optic nerve heads. II = Optic nerve heads having cups with temporal flat slopes. III = Optic nerve heads with punched-out cups ± ± ± ± ± ± ± 0.02 Differences between the three disc types concerning the quotient of minimal to maximal diameter, the angle between the maximal diameter and the horizontal, and the form factor were significant but small (Table 1). Table 3. Morphometric neuroretinal rim data of 338 unselected, normal human optic nerve heads with physiologic cupping, divided morphologically into optic discs having cups with temporal flat slopes and discs with punched-out cups Number Neuroretinal rim Area (mm 2 ) total Sector I Sector II Sector III Sector IV Width (mm) Position 1 Position 2 Position 3 Position 4 Position 5 Position 6 Position 7 Position 8 Position 9 Position 10 Position 11 Position 12 Area sector III/II I/total NRR area H/total NRR area Ill/total NRR area IV/total NRR area 00 II III ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± = 338 unselected normal optic discs with physiologic cupping. II = Optic discs having cups with temporal flat slopes. HI = Optic discs with punched-out cups. Position numbers: clockwise counted in left eyes and counter-clockwise counted in right eyes. Table 4. Horizontal and vertical c/d ratios and quotients in 457 unselected, normal optic nerve heads, divided into optic discs having cups with temporal flat slopes and discs with punched-out cups Number c/d Ratio Horizontal 0.39 ± ±0.12 Vertical 0.34 ± ±0.11 Quotient horizontal/ vertical 1.17 ± ± = Unselected normal optic nerve heads. II = Normal optic discs having cups with temporal flat slopes. Ill = Normal optic discs with punched-out cups. Optic Cup ± ± ±0.09 Optic cup area ranged from 0.00 to 3.41 mm' (mean 0.72 ± 0.70 mm 2 ) (Table 2). Mean horizontal diameter was 0.83 ± 0.58 mm ( mm), and vertical was 0.77 ± 0.55 mm ( mm). The form was more oblique with a mean quotient of minimal to maximal diameter of 0.86 ± Optic cup area in discs with punched-out cups (1.37 ± 0.62 mm 2 ) was significantly (P < ) larger than in discs having cups with temporal flat slopes (0.59 ± 0.39 mm 2 ) or discs without cupping (0.00 mm 2 ). The punched-out cup type was significantly (P < ) more circular than the cup with a temporal flat slope, as indicated by higher form factor and quotient of minimal to maximal diameter (Table 2). Neuroretinal Rim Neuroretinal rim area averaged 1.97 ± 0.50 mm 2 ( mm 2 ), showing an interindividual variability of 1:5.8 (Table 3). In the inferior optic disc half it was significantly larger than in the superior one. The rim was significantly (P < 0.001) widest at the inferior optic disc pole (position 6 in Table 3), followed by the superior optic disc pole (position 12 in Table 3). It was smaller at the nasal disc border and most narrow at the temporal one. Neuroretinal rim area in the temporal lower disc sector was significantly (P < 0.001) larger than in the temporal upper sector, while there was no significant difference in disc area between these two disc sectors (0.809 ±0.197 mm 2 versus ±0.198 mm 2 ). In only 78 (17%) optic discs was rim area smaller in the temporal lower sector than in the temporal upper one. Cup/Disc Ratios C/D ratio ranged horizontally from 0.00 to 0.87 (mean 0.39 ± 0.28) and vertically from 0.00 to 0.85 (mean 0.34 ± 0.25) (Table 4). The horizontal value

5 No. 7 OPTIC DISC, CUP AND RIM AREA AND FORM IN NORMAL EYES / Jonas er al Table 5. Side differences of optic disc and neuroretinal rim area in 138 unselected pairs of normal human optic nerve heads Side difference (mm 2 ) 0.10 or less 0.20 or less 0.30 or less 0.40 or less 0,50 or less 0.60 or less 0.70 or less 0.80 or less 0.90 or less 1.00 or less 1.25 or less 1.50 or less 2.00 or less Optic disc area 28% 46% 62% 73% 80% 82% 87% 91% 94% 95% 99% 99% Cumulative frequency Neuroretinal rim area 33% 55% 72% 82% 88% 90% 93% 94% 97% 98% 99% was significantly (P < ) larger than the vertical one. Accordingly the mean quotient of horizontal to vertical c/d ratio was 1.17 ± The horizontal and vertical c/d ratios were significantly (P < ) lower in discs having cups with temporal flat slopes compared to discs with punched-out cups (Table 4). In each of these groups, however, the horizontal ratio was significantly (P < ) larger than the vertical one. In only 31 (6.8%) of all 457 optic nerve heads or in 31 (9.2%) of all 338 optic discs with physiologic cupping, was the horizontal c/d ratio smaller than the vertical one. Side Differences Optic disc area side differences of 0.10 mm 2 or less were found in 38 subjects (38/138 resp. 27.5%), of 0.20 mm 2 or less in 64 (46%), and of 0.50 mm 2 or less in 111 (80%) (cumulative frequencies, Table 5). Side differences in neuroretinal rim areas of 0.10 mm 2 or less were detected in 33%, of 0.20 mm 2 or less in 55% and of 0.50 mm 2 or less in 88% (cumulative frequencies, Table 5). Side differences of horizontal c/d ratio of 0.1 or less existed in 84%, of 0.2 or less in 96%, and of 0.3 or less in 99%. Differences of vertical c/d ratios between both eyes showed similar figures (Table 6). Side differences concerning type of cupping were detected in 14 subjects (14/138, 10.1%). Correlations Neuroretinal rim and optic cup areas were significantly (P < ) correlated to the optic disc area (Fig. 5). Scattergrams and correlation coefficients Table 6. Morphometrically determined side differences of horizontal and vertical c/d ratios in 138 unselected subjects with no evidence of optic nerve disease c/d Ratio side difference 0.1 or less 0.2 or less 0.3 or less 0.4 or less Cumulative frequency Horizontal c/d ratio 84% 96% 99% Vertical c/d ratio 88% 96% were different when the three groups of optic nerve heads were evaluated separately: In optic discs without cupping rim area was by definition identical with disc area (correlation coefficient 1.0, slope of regression line 1.0). In optic nerve heads having cups with predominantly temporal flat slopes, rim area increased with disc area (slope of regression line 0.62, Fig. 6) while cup area changed less. In optic discs with punched-out cups rim area showed less clear correlation to disc area (slope of regression line 0.30, Fig. 7), while cup area generally increased with disc size (Fig. 8). No correlations existed between the above-mentioned morphometric optic disc data and side, refraction, sex or age. Discussion Optic nerve anomalies and diseases are often associated with specific alterations of the optic disc topography. Quantification of the optic nerve head surface can thus be helpful in diagnosis, differential diagnosis and follow-up of optic nerve abnormalities. Fig. 5. Scattergram of neuroretinal rim area correlated to optic disc area in 457 normal, unselected optic nerve heads. Slope of regression line, 0.33, correlation coefficient, 0.56.

6 1156 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / July 1988 Vol. 29 v X X n E? 1 A v N 4< (< < < < * xs > < * xx X i?o- A * 3 XXX l( *# * : _ VA *.b 0 - t r i 1 IT i-r-i t i i T T i i i i i i-i i i opt Ic disc urea tnn*1 Fig. 6. Scattergram of neuroretinal rim area correlated to optic disc area in 174 optic nerve heads having cups with temporal flat sloping. Slope of regression line, 0.62, correlation coefficient, opt Ic disc area I ran') Fig. 8. Scattergram of optic cup area correlated to optic disc area in 164 unselected, normal optic nerve heads with punched-out cups. Slope of regression line, 0.70, correlation coefficient, The purpose of this study was the evaluation of the normal optic disc topography in absolute size units (mm or mm 2 ). These values might be the basis for future quantitative comparisons. The absolute optic disc dimensions can be determined by evaluating photographs with correction of the photographic magnification according to Littmann's method. 8 " 12 This study gives indirect evidence for the validity of this method: there was no significant difference (Wilcoxon-Mann-Whitney test) between the mean optic disc area in this study (2.69 ± 0.70 mm 2 ) and the mean size of the optic nerve scleral canal evaluated postmortem in 107 unselected, unfixed human donor eyes (2.59 ± 0.72 mm 2 ). 913 The interindividual variability of the optic disc area, ranging from 0.80 to 5.54 mm 2, was remarkably high. It fits the observations of Franceschetti, Bengts- optic disc area X x X Fig. 7. Scattergram of neuroretinal rim area correlated to optic disc area in 164 normal human optic nerve heads with punchedout cups. Slope of regression line, 0.30, correlation coefficient, son and Jaeger 1 ' 310 obtained by different methods and in smaller series. The optic nerve heads at the two ends of this "size spectrum" can be defined as micro- or macrodiscs 14 ; they have an abnormal size and exceed the limits set by the mean value (2.69 mm 2 ) plus/minus two standard deviations (2 X 0.70 mm 2 ). Only 2.28% of all normal optic discs fall outside of each of these marks (1.29 mm 2 and 4.06 mm 2, respectively), according to the Gaussian distribution curve (Fig. 4). The neuroretinal rim also was not interindividually constant, showing a variability of about 1:5.8. The rim area was significantly correlated to the disc area, in agreement with a recent study by Britton, Drance and coworkers. 5 The correlation coefficients could still be increased if three populations of optic nerve heads were differentiated; in the group of optic discs without cupping rim area was identical to the disc area. In optic nerve heads having cups with temporal flat slopes the rim area increased by a factor of about 0.62 with the disc size. In discs with punched-out cups rim area was enlarged by a smaller factor of 0.30 (slope of the regression line). The difference between the last two groups can be attributed to their different ophthalmoscopic appearance: in discs with steep, punched-out cups the border between the rim and cup will be defined at the upper cup edge. In these discs the nerve fibers bend outward at almost a right angle, and there is no slope a part of which could be regarded as belonging to the rim. In discs with cups having a temporal flat slope the nerve fibers and blood vessels leave the eye in the temporal disc region more obliquely. Close to the temporal disc border the nerve fibers might even lie nearly horizontal before kinking down toward the bottom of the optic cup. That area with the nearly horizontal arrangement of the nerve fibers, however, is considered as part of the

7 No. 7 OPTIC DISC, CUP AND RIM AREA AND FORM IN NORMAL EYES / Jonos er ol neuroretinal rim because, for example, the vessel kinking is situated more centrally of it. Therefore in these discs the rim area measurement will be larger and the cup area smaller than in the discs with the steep, punched-out cups. According to this explanation the difference in the correlation of rim area/disc area between the two disc types cannot be regarded as evidence for a different number of optic nerve fibers. The relationship of rim size/disc size is important in the morphometric evaluation of glaucomatous optic nerve heads; if the total neuroretinal rim area only is taken as the criterion for glaucoma, healthy but small optic discs with a small neuroretinal rim area will be classified as abnormal, and large optic nerve heads with beginning glaucomatous loss of the originally (and still) large neuroretinal rim will be considered as healthy. The neuroretinal rim furthermore showed a characteristic configuration. It was not even at all points but was widest at the inferior disc pole, followed by the superior one. It was smaller at the nasal disc side and smallest in the temporal disc sector. In only two (0.4%) of the 457 optic discs of this study was the neuroretinal rim smallest outside of the temporal horizontal disc sector. This rim configuration corresponds to the form of the optic disc and cup: the disc was slightly vertically oval with the vertical diameter being about 9% longer than the horizontal one. The cup was more horizontally oriented. According to the vertical form of the disc and the horizontal form of the cup, the rim was broadest at the inferior and superior disc poles and smallest on the horizontal disc sides. Accordingly the cup/disc ratios were larger horizontally than vertically in 93.6% of all discs examined. These facts concerning the normal neuroretinal rim configuration might be more important than even the total rim area and the side differences in early glaucoma diagnosis. The topographic neuroretinal rim form might be explained by: (1) the parapapillary retinal nerve fiber layer that in monkeys is thickest at the upper and lower disc poles, followed by the nasal and finally by the temporal disc side; 15 and (2) the macula's location inferior to the optic disc's center 16 so that more retinal nerve fibers might reach the lower optic disc half compared to the upper one. In this group of optic discs with high myopes excluded no correlations existed between disc, cup and neuroretinal rim area on one side and refraction, side, age and sex on the other one. This is in accordance with previous studies 517 " 19 and in partial disagreement with others " 24 In these latter studies, however, high myopps had not always been excluded, and examination techniques, examined populations and type of study 25 were different. The missing correlation between optic nerve head size and refraction can be explained by the fact that growth of the posterior scleral opening has finished at about the second year of life, 26 while growth of the total globe, especially in the anterior-posterior axis, will last up to 15 to 20 years, at which point final refraction is reached. It is as yet unclear which factors determine optic disc, optic nerve, and scleral canal dimensions, respectively. As hypotheses, possible mechanisms are: (1) different number and different caliber of existing optic nerve fibers 4 ; (2) different number and volume of neuroglial cells; (3) different number of formed ganglion cells and/or different percentage of lost ganglion cells during embryogenesis 27 " 30 ; (4) different time during embryogenesis when final fixation of the scleral lamina cribrosa occurs; and/or (5) misalignment between scleral lamina cribrosa fixation and caliber growth of the optic nerve fibers. Axons of ganglion cells lying at the edge of the embryonic optic cup reach the primitive papilla last because of delayed development 31 and longest distance to reach the papilla. They will probably lie in the center of the primitive papilla, close to the hyaloid artery, because the papilla's border will already be occupied by axons coming from more centrally located ganglion cells. If by any circumstances, eg, loss of peripheral ganglion cells in the course of ciliary body differentiation or by failure to reach the lateral geniculate ganglion, these axons get lost, there will be a defect in the center of the primitive papilla. This may be the future optic cup. Indirect evidence for this hypothesis is that in normal eyes lamina cribrosa pores can be detected at the bottom of the optic cup. These pores would not have been formed if there had not been axons or axon-like structures at this place before glial and later scleral cells had modelled the lamina cribrosa meshwork. Thus optic cup and consequently optic disc size might partly depend on the number of these centrally located axons that could get lost during embryogenesis. Not only optic discs with cupping but also discs without cupping showed a considerable interindividual variability of neuroretinal rim area. This may be caused by different developmental stages of axon caliber growth when final fixation of the lamina cribosa occurs; if at this time the intrapapillary axons already have a larger diameter, 32 the individual lamina cribrosa pore and thus the total optic disc will be larger even though the cup might not be present. The nerve fiber density will be less. If scleral lamina cribrosa fixation occurs earlier and/or axon caliber growth is retarded, the future optic nerve head will be smaller with higher optic nerve fiber density per disc area and per individual lamina cribrosa pore area.

8 1158 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / July 1988 Vol. 29 Key words: optic nerve head, optic disc morphometry, papillometry, neuroretinal rim, glaucoma Acknowledgment Mrs. A. Handel provided invaluable help in documentation and statistical analysis. References 1. Franceschetti A and Bock RH: Megalopapilla: A new congenital anomaly. Am J Ophthalmol 33:227, Straatsma BR, Foos RY, and Spencer LM: The retina: Topography and clinical correlation. Symposion on retina and retinal surgery. Trans New Orleans Acad Ophthalmol p. 1, Bengtsson B: The variation and covariation of cup and disc diameters. Acta Ophthalmol 54:804, Betz P, Camps F, Collignon-Brach C, and Weekers R: Photographie stereoscopique et photogrammetrie de l'excavation physiologique de la papille. J Fr Ophthalmol 4(3): 193, Britton RJ, Drance SM, Schulzer M, Douglas GR, and Mawson DK: The area of the neuroretinal rim of the optic nerve in normal eyes. Am J Ophthalmol 103:497, Airaksinen PJ, Nieminen H, and Mustonen E: Retinal nerve fiber photography with a wide angle fundus camera. Acta Ophthalmol 60:362, Airaksinen PJ, Drance SM, Douglas GR, Mawson DK, and Nieminen H: Diffuse and localized nerve fiber loss in glaucoma. Am J Ophthalmol 98:566, Littmann H: Zur Bestimmung der wahren GroBe eines Objektes auf dem Hintergrund des lebenden Auges. Klin Monatsbl Augenheilkd 180:286, Jonas JB, Gusek G, Guggenmoos-Holzmann I, and Naumann GOH: Variability of the absolute optic disc size in human living and donor eyes. ARVO Abstracts. Invest Ophthalmol Vis Sci 28(Suppl.):30, Jaeger W: Ermittlung der wahren PapillengroBe an Patienten (Beitrag zur Diagnose der Mikropapille). Fortschr Ophthalmol 80:527, Airaksinen PJ, Drance SM, Douglas GR, and Schulzer M: Neuroretinal rim areas and visual field indices in glaucoma. Am J Ophthalmol 99:107, Drance SM, Airaksinen PJ, Price M, Schulzer M, Douglas GR, and Tansley BW: The correlation of functional and structural measurements in glaucoma patients and normal subjects. Am J Ophthalmol 102:612, Jonas JB, Gusek GC, Guggenmoos-Holzmann I, and Naumann GOH: Size of the optic nerve scleral canal and comparison to intravital determination of optic disc dimensions. Graefes Arch Clin Exp Ophthalmol, in press. 14. Jonas JB, Gusek G, and Naumann GOH: Makropapillen mit physiologischer Makroexkavation (Pseudo-Glaukompapillen). Klin Monatsbl Augenheilkd 191:452, Quigley HA and Addicks EM: Quantitative studies of retinal nerve fiber layer defects. Arch Ophthalmol 100:807, Hogan MJ, Alvarado JA, and Weddel JE: Histology of the Human Eye: An Atlas and Textbook. Philadelphia, W.B. Saunders, 1971, p Snydacker D: The normal optic disc. Am J Ophthalmol 58:958, Armaly MF: The optic cup in the normal eye: I. Cup width, depth, vessel deplacement, ocular tension and outflow facility. Am J Ophthalmol 68:401, Lotmar WH, Goldmann H, and Bruckner R: Zur Bestimmung zeitlicher Veranderungen der Papille bei normalen Erwachsenen. Klin Monatsbl Augenheilkd 173:480, Pickard R: The alteration in size of the normal optic disc cup. Br J Ophthalmol 32:355, Tomlinson A and Phillips CI: Ratio of optic cup to optic disc in relation of axial length of eye ball and refraction. Br J Ophthalmol 53:765, Ford M and Sarwar M: Features of a clinically normal optic disc. Br J Ophthalmol 47:50, Schwartz B: Cupping and pallor of the optic disc. Arch Ophthalmol 89:272, Carpel EF and Engstrom PF: The normal cup/disc ratio. Am J Ophthalmol 91:588, 1981, p Robert Y: Die klinischen Untersuchungsmethoden der Papille: Ihre Bedeutung fur die Glaukom-Fruhdiagnostik. Stuttgart, Enke-Verlag, Quigley HA: Childhood glaucoma: Results with trabeculectomy and study of reversible cupping. Ophthalmology 89:219, Chalupa LM, Williams RW, and Henderson Z: Binocular interaction in the fetal cat regulates the size of ganglion cell population. Neuroscience 12:1039, Sefton AJ and Lam K: Quantitative and morphologic studies on developing optic axons in normal and enucleated albino rats. Exp Brain Res 57:107, Provis JM, van Driel D, Billson FA, and Russell P: Human fetal optic nerve: Overproduction and elimination of retinal axons during development. J Comp Neurol 238:92, Crespo D, O'Leary DD, and Cowan WM: Changes in the numbers of optic nerve fibers during late prenatal and postnatal development in the albino rat. Brain Res 351:129, Ozanics V and Jacobiec FA: Prenatal development of the eye and its adnexa. In Ocular Anatomy, Embryology and Teratology, Jacobiec FA, editor. Philadelphia, Harper and Row, 1982, p Rhodes RH: Development of the optic nerve. In Ocular Anatomy, Embryology and Teratology, Jacobiec FA, editor. Philadelphia, Harper and Row, 1982, p. 623.