- Email: [email protected]
Three-Dimensional Topography of the Central Visual Field
Three-Dimensio nal
Topography of the Central Visual Field
Sparing of Foveal Sensitivity in Macular Disease WILLIAM M. HART, JR., MD, PhD, RONALD M. BURDE, MD
Abstract: Threshold static perimetry was performed using test object patterns that covered contiguous areas of the central visual field. Computer imaging methods were used to display a three-dimensional surface that was interpolated between the sensitivity values at each of the test object locations. The examinations covered the area out to and including 10° of eccentricity from the point of fixation, corresponding to the same area of the visual field covered by the Amsler grid. The normal visual field surface appears as a high plateau with a smoothly rising level of sensitivity forming a peak at the point of fixation . It was found that in a variety of macular diseases, including those caused by vascular, as well as primary degenerative disorders, central scotomas were characterized by relative sparing of visual sensitivity at the point of fixation. The pattern thus produced was one of a ring-shaped depression within the central 10° of the visual field. This phenomenon was present in 20% of cases with central scotomas resulting from macular disease, but was not found in any eye of 64 patients suffering from central scotomas as a result of optic nerve disease. This pattern of visual field loss may be common, though not frequently recognized. It is proposed that the phenomenon of preservation of foveal sensitivity may be a marker for macular disease, as distinct from central visual field defects arising from optic nerve disease. [Key words: annular, computer, fovea, macula, optic nerve, perimetry, ring, scotoma.] Ophthalmology
90:1 028-1 038, 1983
From the Department of Ophthalmology, Washington University School of Medicine, St. Louis, Missouri. Supported by grants EY-02044 from the National Eye Institute, and RR01380 from the Division of Research Resources. Presented at the Eighty-seventh Annual Meeting of the American Academy of Ophthalmology, San Francisco, California, October 30-November 5, 1982. Reprint requests to William M. Hart, Jr., MD, PhD, Department of Oph· thalmology-Box 8096, Washington University School of Medicine, 660 So. Euclid Avenue, St. Louis, MO 63110.
1028
The techniques of static perimetry have undergone significant changes with the development of computerassisted instruments for control of the examination, and subsequent processing of the data obtained. One fundamental change that this technology has allowed has been the use of patterns of targets that cover contiguous areas of the visual field. 1-4 These patterns are commonly quadratic grids, and their advantages, when compared to conventional linear meridional patterns, include improved probability of success during the detection phase of perimetry, and increased topographic resolution dur-
0161-6420/83/0800/1028/$1.35 ©American Academy of Ophthalmology
HART AND BURDE •
CENTRAL SCOTOMAS OF MACULAR DISEASE
p4
@ •••••
p2 pl p3
@ tS&\R!"\Ri~ ® tr\!0!'\R!'R!l ® Ll tl L2 Fig 1. Diagram of relationship between points of image matrix and test object locations. L 1 • • • L3 (points in circles): test object locations. Closed points: point elements of image matrix. T 1 : horizontal trabecula. T 2 : vertical trabecula. S: set points comprising one intrabecular area of image matrix. P 1 • • • P5 : group of five points whose arithmetic mean would be assigned to point P 1 during one pass of smoothing algorithm.
ing the analysis phase. 5 The value of the increased topographic resolution is illustrated by the visual field findings reported in this study. It has been found that in a diverse variety of macular diseases there is a tendency for visual field defects to form in the perifoveal region of the visual field with a relative sparing of visual sensitivity at the point of fixation. This pattern is contrasted with that of central visual field defects that occur as a result of optic nerve disease, in which there is no comparable tendency towards perservation of foveal sensitivity.
MATERIALS AND METHODS Threshold static perimetry was performed with a modified Goldmann perimeter as previously reported. 5 Test object patterns were regular, quadratic grids. Each grid was made up of 49 target locations in a 7 X 7 square. The intertarget distance was 3.3° such that the area covered by the square pattern was 20° on each side. When examining the central visual field, the pattern was positioned so that its central target was located at fixation. Thus, the area examined included all points within 1oo of fixation. Background luminance was 31.5 abs and
brightness increment thresholds were determined at each target location by conventional manual technique. Foveal thresholds were determined while using the standard eccentric fixation device. The threshold value determined for each target location was recorded locally by the instrument, and the final results of each examination were subsequently transmitted to a minicomputer for image processing. The threshold values (recorded on a logarithmic scale) were inverted prior to interpolation of the display surface so that results were expressed as sensitivity values in units of decibels. Following interpolation of a three-dimensional display surface between the sensitivity values, images were drawn on oscillographic display terminals and permanent images were generated with an electrostatic plotting device. A display surface, interpolated between the sensitivity values at each of the target locations, was represented by a three-dimensional image matrix. The matrix consisted of 1,225 points arrayed as a series of 25 parallel lines, each line being composed of 49 evenly spaced points. Like the target locations, each point of the image matrix was associated with three variables: horizontal and vertical position and visual sensitivity. The relationship between the points of the image matrix and the target locations is diagrammed in Figure 1. The matrix was registered in position with and matched to the size of the target pattern used for the examination. Its lines were positioned parallel to the "front" of the area examined, which was taken to be the inferior border of the object pattern. Following interpolation of the surface between the sensitivity values, images of the three-dimensional surface, drawn in two dimensions, were produced by a hidden line algorithm, which used superposition of foreground elements to produce the illusion of a three-dimensional structure. Drawing of the image matrix was accomplished by sequentially tracing each of its lines, starting at the front and proceeding toward the rear. An illusion of perspective was created by offsetting the starting position for the drawing of each line. The image matrix was fitted to the perimetric data and smoothed, using a technique that preserved the sensitivity values at each of the 49 object locations. A portion of the image matrix is diagrammed in Figure 1. For the standard object pattern the inferior line of the image matrix lay directly over the seven locations comprising the inferior border of the examination area, and every fourth line of the matrix was matched to subsequent rows of target locations. Thus, three lines of the image matrix fell into each of the inter-row spaces, and there were a total of 25 lines. For a given row of seven target locations, every eighth point of the matching matrix line was coincident with a target location, and seven matrix points fell into each interlocation space for a total of 49 points in each line. Matrix points coincident with target locations were assigned the corresponding sensitivity values. Points that fell along interlocation lines were thought of as forming linear trabeculae. Groups of seven points formed trabeculae in the horizontal dimension,
1029
OPHTHALMOLOGY
•
AUGUST 1983
•
VOLUME 90
•
NUMBER 8
8.
A. 44
38~~~~~~~ 32 26
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY (Degrees)
c.
D.
-a) 44 -o
;
38
~
32
!::::
1-
44
k=~~~~~~~l!!=
~ 26 ILl
"'
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY
(Degrees)
(Degrees)
E.
F.
44
-
a)
;
1-
~
38
k="!!!!!!~~~~~~
32
>-
1-
>
;::
z 26 1-
co
-o
-
z
ILl
"'
44
38 32 26
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY (Degrees)
and groups of three formed interlocation trabeculae in the vertical dimension. (Horizontal and vertical trabeculae are diagrammed in Figure I.) Each point within a trabecula was assigned a sensitivity value by linear interpolation between the corresponding values of the two locations connected by the trabecula. Quadratic groups of 21 points formed intratrabecular portions of the image matrix, as indicated by the set of points labeled "S" in Figure 1. Each point within these groups was given a sensitivity value by linear interpolation between the two values of the matrix points of the vertical trabeculae lying to either side in the horizontal dimension.
1030
Fig 2. Examples of three-dimensional visual field surface following alternative methods of image smoothing. A: test object pattern for left eye used to cover point of fixation and include physiologic blind spot. B: results of three-dimensional interpolation without smoothing. (Arrow #I : point of fixation. Arrow #2: physiologic blind spot, tic marks indicate position of vertical meridian). C: results of excessive smoothing of surface image (32 passes of algorithm). D: image appearance following 32 passes of smoothing algorithm while restraining points located at test object locations. E: results of 32 passes of smoothing alogorithm while restraining all points along horizontal trabeculae. S: surface image appearance following two passes of smoothing algorithm while restraining all points of horizontal trabeculae.
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY (Degrees)
Following interpolation of the image matrix, its intertrabecular portions were smoothed by iteratively averaging each point with its immediate neighbors. The sensitivity vaues of all points lying within horizontal trabeculae were held constant during this process, so that the interpolated surface was constrained by the sensitivity values at the test object locations. The results of this surface interpolation technique are illustrated in Figure 2. All of the three-dimensional visual field surfaces produced for this work employed the linear interpolation technique followed by two passes of the restrained smoothing algorithm.
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
NORMAL
_44 co
~
>!::::
>
II/')
z
w
1/')
-10
0
+10
HORIZONTAL (Degrees)
CASE 1
44
co ~33 >1-
I 1e
~ 22
1-
1/')
z 11
w
1/')
0
+10
HORIZONTAL (Degrees) Fig 3. Top left, test object pattern used for threshold static perimetry of central visual field. Top right, appearance of normal central visual field. Tic marks indicate position of vertical meridian. Bottom left, kinetic visual field, case I. Bottom right, three-dimensional static visual field, case I, demonstrating central scotoma with relative preservation of foveal sensitivity.
RESULTS The central visual field was examined, using the object location pattern shown in Figure 3 (top left). The normal visual field surface with this technique of imaging has the appearance of a high plateau that rises through a gently sloping aspect to a small, but distinct peak located at the point of fixation. The point of fixation is located directly at the center of the area examined, and the vertical meridian is indicated on the surface by a row of tic marks. The visual fields shown in the bottom of Figure 3 (case 1) were obtained from the right eye of a 41-year-
old woman, who gave a history of having had a venous stasis retinopathy 10 years earlier. A fundus photograph from the patient's records confirmed the diagnosis. Although the visual acuity of the eye was 20/25 the patient complained of considerable difficulty reading due to a blind spot in the center of the visual field of the right eye. There was no afferent pupillary defect, and the ophthalmoscopic appearance of the macula (Fig 4) showed only dilation and tortuosity of the perifoveal capillaries (best seen by contact lens examination). Kinetic perimetry, however, (Fig 3, bottom left), demonstrated a central scotoma, approximately go wide in its horizontal dimension, and having a maximum density plottable with the l3e target. How visual acuity could be
1031
OPHTHALMOLOGY
•
AUGUST 1983
•
VOLUME 90
•
NUMBER 8
Fig 4. Case I : fundus photograph of macula, showing abnormalities of perifoveal capillaries.
Fig S. Case 2: appearance of macula, showing subretinal hemorrhage.
maintained at 20/25 with such a dense central scotoma, was explained by the three-dimensional contour of the visual field surface, which clearly demonstrated the presence of a small but distinct peak of higher visual sensitivity located at the fovea. A second case demonstrated a similar visual field pattern. A 64-year-old man, who had suffered loss of vision in the right eye 2 months before being seen, was found to have visual acuity of 20/50 in the affected eye. Ophthalmoscopic examination revealed a disciform macular degeneration with a prominent subretinal hemorrhage in the macula (Fig 5). Kinetic perimetry (case 2, Fig 6) demonstrated a large central scotoma with sloping borders. It appeared to be maximally dense at its center where it was present to the l4e target. The threedimensional visual field surface, however (top right, Fig 6), demonstrated an irregular perifoveal depression of the visual field with distinct preservation of a peak of sensitivity at the point of fixation. The third case noted to demonstrate this type of finding was that of a 73-year-old man who had undergone intracapsular cataract extraction with implantation of an iris fixation intraocular lens in the right eye 2 years ago. He complained of difficulty reading due to a central scotoma, and the right eye was found to have visual acuity of 20/60. Kinetic perimetry (bottom left, Fig 6) showed a small, moderately dense central scotoma. In this as in the first two cases, no hint of preserved visual sensitivity at the point of fixation was obtained during kinetic perimetry. The three-dimensional surface of the central visual field, however, clearly demonstrated the presence of relatively preserved visual sensitivity at the point of fixation (bottom right, Fig 6). Ophthalmoscopic examination disclosed the presence of cystoid macular edema. The fourth case was that of a 69-year-old man referred with complaints of difficulty reading due to the presence of central scotomas in both eyes. Visual acuity in both
eyes, however, was 20/25, and the results of kinetic perimetry, which were similar in both eyes, demonstrated the presence of irregular though dense central scotomas. Examination of both eyes (right eye shown in Fig 7) demonstrated geographic atrophy of the retinal pigment epithelium, making the underlying choroidal vessels prominently visible. There was, however, no visible scar, hemorrhage, or edema. The three-dimensional topography of the central visual field surface (bottom half Fig 8) demonstrated symmetrical, perifoveal central scotomas with a distinct tendency towards preservation of visual sensitivity at the point of fixation in both eyes. The fifth case involved a 33-year-old woman who complained of the sudden onset of central scotomas in both eyes. Visual acuity was 20/40 in each eye. The patient had undergone an emergency cholecystectomy during the third trimester of pregnancy, followed a short time later by a Cesarean section. After the second surgical procedure, the onset of reduced visual acuity was noted. Ophthalmoscopic examination demonstrated the presence of tiny, glistening white retinal inclusions in the maculae. There were also retinal pigment epithelial defects in the same area (right eye, Fig 9). These window defects were confirmed by fluorescein angiography, and had the effect of producing as so-called target maculopathy. The appearance was compatible with Stargardt's disease, although there was no family history of visual disorder. There was no history of drug abuse or use of antimalarial drugs. The visual fields (Fig 10) showed small, bilateral central scotomas by kinetic perimetry. Static perimetry demonstrated small, very dense central visual field defects, with marked sparing of visual sensitivity at the point of fixation in both eyes. This had the appearance of sharp spikes in the visual field surface, located at the center of the visual field surfaces. The subsequent course of the patient showed no change in the visual acuity or in the visual field findings. The sixth case was that of a 25-year-old woman who
1032
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
CASE 2
-44
.,co >......
33
> 22
......
z
IJ)
11
UJ
IJ)
-10
0 HORIZONTAL
+10
(Degrees)
CASE 3
-44
.,co >......
33
>
22
z
11
...... IJ)
UJ
IJ)
0 HORIZONTAL
+10
(Degrees) Fig 6. Top left, kinetic visual field, case 2. Top right, three-dimensional static visual field, case 2, demonstrating broad area of central visual field defect with relative preservation of foveal sensitivity. Bottom left, kinetic visual field, case 3. Bottom right, three-dimensional static visual field, case 3, demonstrating perifoveal depression of visual field surface with relative preservation of foveal sensitivity.
noted visual field defects in both eyes immediately following emergency Cesarean section for an eclamptic pregnancy. The patient, however, was not seen until 2 weeks after the onset of visual loss, at which time the appearance of the fundus was unremarkable (Fig 11 ). Fluorescein angiography showed no abnormality. Kinetic perimetry of the left eye demonstrated relative central scotomas generally surrounding the point of fixation. Static perimetry of the central visual field (top right of Fig 12) demonstrated definite sparing of visual sensitivity at the point of fixation. Note that the visual field surface has been rotated 180° in this figure to allow better visualization of the point of fixation. The visual field defect in the patient's right eye was also associated with a normal appearing fundus, but was cecocentral in distribution involving the physiologic blind spot and the temporal half of the macula, but not including the point of fixation.
Fig 7. Case 4: appearance of macula, right eye, showing geographic atrophy of retinal pigment epithelium.
1033
OPHTHALMOLOGY
•
AUGUST 1983
•
VOLUME 90
•
NUMBER 8
CASE 4
RIGHT EYE
LEFT EYE
-44
-44
-33
-33
I-
> 22
I-
I-
I-
..,
..,
a::l
a::l
>-
>-
1.1)
zLU
~ 1.1)
11
z
22 11
LU
1.1)
1.1)
v~
o
fD Rr,c eg,.ee -<1( S)
-10
0 HORIZONTAL (Degrees)
+10
-10
0 HORIZONTAL (Degrees)
+10
Fig 8. Case 4: kinetic visual fields {top) demonstrate bilateral central scotomas plotted with the 1.. target with small islands of greatest density in a perifoveal area. Three-dimensional static visual fields {bottom) show bilateral, dense central scotomas with marked preservation of visual sensitivity at the fovea.
Fig 9. Case 5: appearance of macula, right eye, demonstrating retinal pigment epithelial defects in a perifoveal, ring-shaped pattern. A few small white intraretinal dots can be seen.
1034
No instance offoveal sparing in the central visual field was observed for any patient with a central scotoma due to optic nerve disease. Case 7 (lower half, Fig 12) was that of a 30-year-old man being followed for multiple sclerosis. The kinetic visual field was obtained 8 weeks after the abrupt onset of retrobulbar optic neuritis in the right eye. There was a cecocentral scotoma with its greatest density in the superior half of the central visual field. The three-dimensional visual field surface (bottom right, Fig 12) was reversed to allow the point of fixation to be seen more easily. It clearly demonstrates a depression of the foveal sensitivity that falls well within the margins of the central visual field defect. This finding is representative of central and cecocentral scotomas arising from optic nerve disease. The constellation of findings in patients with selective preservation of foveal sensitivity has included relatively preserved visual acuity (20/60 or better) in the presence of signs and symptoms of a central scotoma and/or
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
CASE 5
------~ -- m4 e
l3e
I2e
LEFT EYE
RIGHT EYE
44
lil
co ~ 33 >-
~
>- 33
I-
t:::
I-
I-
> 22
> 22
V)
z UJ
44
V)
11
z
11
UJ
V)
V)
-10
0 HORIZONTAL (Degrees)
+10
-10
0 HORIZONTAL (Degrees)
+10
Fig 10. Case 5: kinetic visual fields (top) show small, dense central scotomas plotted with the l3e target. Three-dimensional static visual fields (bottom) demonstrate small, very dense central scotomas with preservation of visual sensitivity at fovea in both eyes.
macular disease. The common complaint was of inability to read, and to a degree greater than would be expected on the basis of acuity impairment alone. The prevalence of foveal preservation as a finding in central scotomas arising from macular disease has not been determined. We have used contiguous area static perimetry to examine 30 eyes in 38 patients who had signs or symptoms of central scotomas. Of these, nine eyes in six patients showed foveal sparing within central scotomas (20% of cases). This finding was not demonstrated in any patient in our series whose visual acuity had fallen to 20/80 or less. Thus, foveal sparing is not likely to help in differentiating macular and optic nerve disorders when acuity is below this level. To date we have used contiguous area static perimetry to examine 84 eyes in 64 patients with central visual field defects arising from optic nerve disease, including ischemic, toxic, and demyelinating neuropathies. In none of these
Fig 11. Case 6: fundus, left eye, showing unremarkable appearance of macula. A fluorescein angiogram was likewise normal.
1035
OPHTHALMOLOGY
•
AUGUST 1983 •
VOLUME 90
•
NUMBER 8
CASE 6
CASE 7
+10
0
-10
HORIZONTAL (Degrees) Fig 12. Cases 6 and 7: top left, kinetic visual field, case 6, demonstrating variably dense central scotoma with perifoveal distribution. Top right, three-dimensional static visual field, case 6. Note that orientation of surface has been rotated 180° to allow better appreciation of shape of surface at foveal position (at center of pattern). The central scotoma is seen to be very dense, small, and perifoveal in distribution. Bottom left, kinetic visual field, case 7, shows dense, cecocentral scotoma with a superior arcuate pattern. Bottom right, three-dimensional static visual field, case 7, demonstrating altitudinal distribution of depressed visual sensitivity, and definite depression of sensitivity at fovea (indicated by arrow). Note that this surface has also been rotated 180° to allow better visualization of point of fixation.
cases has there been evidence of preferential sparing of foveal sensitivity.
DISCUSSION Relative preservation of visual sensitivity at the fovea (within a central scotoma) may be a more common accompaniment of macular disease than has been previously recognized. Prior reports of ring scotomas occurring within 10° of fixation have been associated with chloroquine retinopathy. 6 Ring maculopathies not associated with chloroquine use are not uncommon. 7- 9 and are usually associated with dominantly inherited
1036
cone dystrophy, Stargardt's disease, early disciform macular degeneration and possibly with late stages of central serous retinopathy. Central scotomas in some of these cases7 are known to be annular. The more commonly recognized ·ring-shaped visual field defect is that associated with pigmentary retinal degenerations. This type of defect, however, occurs in the midperiphery of the visual field and is usually associated with some degree of sparing of the macula. The cases we report were detected at random during examination of a variety of patients known to have macular and optic nerve disorders. The fact that this phenomenon has not been more widely recognized is probably due to the limitations of the techniques of perimetry customarily used
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
to explore the central visual field. Most of the patients with this finding have not subjectively appreciated the fact that foveal sensitivity was preferentially spared, nor were they ordinarily capable of describing this phenomenon when using the Amsler grid. Use of the Goldmann perimeter precludes easy detection of the phenomenon unless the eccentric fixation target is used while exploring the visual field at and around the point of fixation. Otherwise the small area of preserved sensitivity will be covered by the standard fixation target. If the eccentric fixation device is used, it may be possible with kinetic perimetry to demonstrate foveal preservation. However, static perimetry remains the most sensitive and definitive technique. The use of contiguous area static perimetry in which quadratic grids of test object locations are used to blanket an area of the visual field, has made demonstration of the foveal preservation phenomenon quite easy. The disease processes associated with relative foveal preservation are predominantly vascular and degenerative disorders of the macula. One case was associated with a prior venous stasis retinopathy (case 1), one was found to have cystoid macular edema (case 3), three had degenerative diseases of the macula (cases 2, 4, and 5), and one (case 6) was unexplained. It would be expected that toxic maculopathies, such as those resulting from the use of antimalarial agents that produce "target" lesions within the macula, might also be associated with ring depressions of the central visual field, showing relative preservation of foveal sensitivity. However, we have examined only one case in which a fully developed target lesion of the macula had resulted from long-term use of chloroquine phosphate. There was severely reduced visual acuity (20/200), and the visual field surface failed to show preservation of foveal sensitivity. The macula, a highly specialized region of the retina, is adapted to the tasks of high resolution vision, included high contrast sensitivity, and color perception. The unique anatomic structures associated with this specialization provide a number of possible explanations for the phenomenon of relative preservation of foveal sensitivity in macular disease. The fact that the fovea is rod-free, and that the anatomic density of rods is greatest in the perifoveal macula, would explain a perifoveal depression of the visual field in diseases for which rods are selectively sensitive. Thus, a macular degenerative process that preferentially disrupts rod function, would be expected to preserve visual sensitivity at the fovea. Also, the fact that the foveal retina is avascular may provide an explanation for preservation of its sensitivity in diseases involving the perifoveal capillaries. Unless such an insult were predominately felt towards the outer layers of the retina however, it would be difficult to use this as an explanation, since the afferent processes of foveal cones must pass through the perifoveal macula to reach their respective bipolar cells. The radial distribution of these processes within Henle's fiber layer, and their absence in the fovea, may account for relative pres-
ervation offoveal sensitivity in macular edema. Selective sequestration of intraretinal edema in this specialized region of the outer plexiform layer could produce damage to retinal function in the extrafoveal macula, while leaving foveal function relatively intact. This would be a functional correlate of the common clinical observation of a "petaloid" or radial distribution of loculated fluid, frequently found in macular edema. Other perimetric tests (of increasing complexity) have been found to differentiate between visual field defects arising from macular and optic nerve disorders. Acuity profile perimetry in macular disease has shown a dissociation from brightness increment sensitivity, whereas optic nerve disorders tend to affect both functions to a more equal extent. 10- 12 Clinical adaptations of perimetric tests that mimic retinal receptive field properties have been used to study visual field defects arising from retinal and optic nerve disorders. 13- 14 With damage to the choroid or outer layer of the retina, reduced brightness sensitivity, decreased dark adaptation, and an altered Stiles-Crawford function can be found with preserved receptive field-like properties (including sustained-like and transient-like functions). 15- 16 Repetitive sensitivity testing to demonstrate a " fatigue" effect (the so-called flashing repeat static test) has reportedly produced no such phenomenon in retinal diseases, although it is apparently characteristic of retrolaminar disorders of the optic nerve. 17 The finding of relative preservation of foveal sensitivity in a variety of macular diseases, serves to underscore the relative insensitivity of visual acuity measurement as a determinant of the presence or extent of functional damage to the macula. It is common clinical experience to observe patients with extensive ophthalmoscopic evidence of macular disease, who none-theless have no impairment of visual acuity. It is likewise a not infrequent experience to encounter patients, who in spite of undiminished visual acuity, complain of significant visual dysfunction due to central scotomas ar metamorphopsia. Although no cases have yet been observed of foveal preservation in central scotomas that have arisen as a result of optic nerve disease, more experience will have to be gathered before we can definitely say that this finding is uniform. If that proves to be the case, however, the demonstration of relative sparing of foveal sensitivity in central visual field defects may serve as a convenient marker for macular disease, as distinct from visual field defects resulting from optic nerve disorders.
REFERENCES 1. Spahr J . Zur Automatisierung der Perimetrie. I. Die Anwendung eines computergesteuerten Perimeters. Albrecht von Graefes Arch Klin Exp Ophthalmol 1973; 188:323-38. 2. Weber 8, Spahr J . Zur Automatisierung der Perimetrie; Darstellungs-
1037
OPHTHALMOLOGY
3. 4.
5. 6.
7. 8. 9. 10.
•
AUGUST 1983 •
methoden perimetrischer Untersuchungsergebnisse. Acta Ophthalmol1976; 54:349-62. Heijl A Studies on computerized perimetry. Acta Ophthalmol 1977; Suppl132. Frankhauser F, Bebie H: Threshold fluctuations, interpolations and spatial resolution in perimetry. Doc Ophthalmol Proc Ser 1979; 19:295-309. Hart, WM Jr. Hartz RK. Computer-generated display for three-dimensional static perimetry. Arch Ophthalmol 1982; 100:312-8. Okun E, Gouras P, Bernstein H, von Sallmann L. Chloroquine retinopathy; a report of eight cases with ERG and dark-adaptation findings. Arch Ophthalrnol 1963; 69:59-71 . Grey RHB, Blach RK, Barnard WM. Bull's eye rnaculopathy with early cone degeneration. Br J Ophthalmol1977; 61 :702-18. Weise EE, Yannuzzi LA. Ring maculopathies mimicking chloroquine retinopathy. Arn J Ophthalmol 1974; 78:204-10. Krill AE , Deutman AF. Dominant macular degenerations; the cone dystrophies. Am J Ophthalmol1972; 73:352-69. Johnson CA. Keltner JL, Balestrery F. Effects of target size and eccentricity on visual detection and resolution . Vision Res 1978; 18:1217- 22.
1038
VOLUME 90 •
NUMBER 8
11 . Johnson CA, Keltner JL, Balestrery FG. Acuity profile perimetry; description of technique and preliminary clinical trials. Arch Ophthalmol 1979; 97:684-9. 12. Keltner JL, Johnson CA, Cowley IJ. Acuity profile perimetry in a unique case of bilateral central serous retinopathy. Ann Ophthalmol 1980; 12:726-32. 13. Sunga RN, Enoch JM. A static perimetric technique believed to test receptive field properties. Ill. Clinical trials. Am J Ophthalmol 1970; 70:244- 72. 14. Sunga RN, Enoch JM. Further perimetric analysis of patients with lesions of the visual pathways. Am J Ophthalmol 1970; 70:403-22. 15. Fitzgerald CR, Enoch JM, Campos EC, Bedell HE. Comparison of visual function studies in two cases of senile macular degeneration. Albrecht von Graefes Arch Klin Exp Ophthalmol1979; 210:79-91 . 16. Campos EC, Bedell HE, Enoch JM, Fitzgerald CR. Retinal receptive field-like properties and Stiles-Crawford effect in a patient with traumatic choroidal rupture. Doc Ophthalmol 1978; 45:381-95. 17. Enoch JM, Fitzgerald CR, Campos EC. Quantitative Layer-by-Layer Perimetry; An Extended Analysis. New York: Grune and Stratton, 1981; 195-223.
Topography of the Central Visual Field
Sparing of Foveal Sensitivity in Macular Disease WILLIAM M. HART, JR., MD, PhD, RONALD M. BURDE, MD
Abstract: Threshold static perimetry was performed using test object patterns that covered contiguous areas of the central visual field. Computer imaging methods were used to display a three-dimensional surface that was interpolated between the sensitivity values at each of the test object locations. The examinations covered the area out to and including 10° of eccentricity from the point of fixation, corresponding to the same area of the visual field covered by the Amsler grid. The normal visual field surface appears as a high plateau with a smoothly rising level of sensitivity forming a peak at the point of fixation . It was found that in a variety of macular diseases, including those caused by vascular, as well as primary degenerative disorders, central scotomas were characterized by relative sparing of visual sensitivity at the point of fixation. The pattern thus produced was one of a ring-shaped depression within the central 10° of the visual field. This phenomenon was present in 20% of cases with central scotomas resulting from macular disease, but was not found in any eye of 64 patients suffering from central scotomas as a result of optic nerve disease. This pattern of visual field loss may be common, though not frequently recognized. It is proposed that the phenomenon of preservation of foveal sensitivity may be a marker for macular disease, as distinct from central visual field defects arising from optic nerve disease. [Key words: annular, computer, fovea, macula, optic nerve, perimetry, ring, scotoma.] Ophthalmology
90:1 028-1 038, 1983
From the Department of Ophthalmology, Washington University School of Medicine, St. Louis, Missouri. Supported by grants EY-02044 from the National Eye Institute, and RR01380 from the Division of Research Resources. Presented at the Eighty-seventh Annual Meeting of the American Academy of Ophthalmology, San Francisco, California, October 30-November 5, 1982. Reprint requests to William M. Hart, Jr., MD, PhD, Department of Oph· thalmology-Box 8096, Washington University School of Medicine, 660 So. Euclid Avenue, St. Louis, MO 63110.
1028
The techniques of static perimetry have undergone significant changes with the development of computerassisted instruments for control of the examination, and subsequent processing of the data obtained. One fundamental change that this technology has allowed has been the use of patterns of targets that cover contiguous areas of the visual field. 1-4 These patterns are commonly quadratic grids, and their advantages, when compared to conventional linear meridional patterns, include improved probability of success during the detection phase of perimetry, and increased topographic resolution dur-
0161-6420/83/0800/1028/$1.35 ©American Academy of Ophthalmology
HART AND BURDE •
CENTRAL SCOTOMAS OF MACULAR DISEASE
p4
@ •••••
p2 pl p3
@ tS&\R!"\Ri~ ® tr\!0!'\R!'R!l ® Ll tl L2 Fig 1. Diagram of relationship between points of image matrix and test object locations. L 1 • • • L3 (points in circles): test object locations. Closed points: point elements of image matrix. T 1 : horizontal trabecula. T 2 : vertical trabecula. S: set points comprising one intrabecular area of image matrix. P 1 • • • P5 : group of five points whose arithmetic mean would be assigned to point P 1 during one pass of smoothing algorithm.
ing the analysis phase. 5 The value of the increased topographic resolution is illustrated by the visual field findings reported in this study. It has been found that in a diverse variety of macular diseases there is a tendency for visual field defects to form in the perifoveal region of the visual field with a relative sparing of visual sensitivity at the point of fixation. This pattern is contrasted with that of central visual field defects that occur as a result of optic nerve disease, in which there is no comparable tendency towards perservation of foveal sensitivity.
MATERIALS AND METHODS Threshold static perimetry was performed with a modified Goldmann perimeter as previously reported. 5 Test object patterns were regular, quadratic grids. Each grid was made up of 49 target locations in a 7 X 7 square. The intertarget distance was 3.3° such that the area covered by the square pattern was 20° on each side. When examining the central visual field, the pattern was positioned so that its central target was located at fixation. Thus, the area examined included all points within 1oo of fixation. Background luminance was 31.5 abs and
brightness increment thresholds were determined at each target location by conventional manual technique. Foveal thresholds were determined while using the standard eccentric fixation device. The threshold value determined for each target location was recorded locally by the instrument, and the final results of each examination were subsequently transmitted to a minicomputer for image processing. The threshold values (recorded on a logarithmic scale) were inverted prior to interpolation of the display surface so that results were expressed as sensitivity values in units of decibels. Following interpolation of a three-dimensional display surface between the sensitivity values, images were drawn on oscillographic display terminals and permanent images were generated with an electrostatic plotting device. A display surface, interpolated between the sensitivity values at each of the target locations, was represented by a three-dimensional image matrix. The matrix consisted of 1,225 points arrayed as a series of 25 parallel lines, each line being composed of 49 evenly spaced points. Like the target locations, each point of the image matrix was associated with three variables: horizontal and vertical position and visual sensitivity. The relationship between the points of the image matrix and the target locations is diagrammed in Figure 1. The matrix was registered in position with and matched to the size of the target pattern used for the examination. Its lines were positioned parallel to the "front" of the area examined, which was taken to be the inferior border of the object pattern. Following interpolation of the surface between the sensitivity values, images of the three-dimensional surface, drawn in two dimensions, were produced by a hidden line algorithm, which used superposition of foreground elements to produce the illusion of a three-dimensional structure. Drawing of the image matrix was accomplished by sequentially tracing each of its lines, starting at the front and proceeding toward the rear. An illusion of perspective was created by offsetting the starting position for the drawing of each line. The image matrix was fitted to the perimetric data and smoothed, using a technique that preserved the sensitivity values at each of the 49 object locations. A portion of the image matrix is diagrammed in Figure 1. For the standard object pattern the inferior line of the image matrix lay directly over the seven locations comprising the inferior border of the examination area, and every fourth line of the matrix was matched to subsequent rows of target locations. Thus, three lines of the image matrix fell into each of the inter-row spaces, and there were a total of 25 lines. For a given row of seven target locations, every eighth point of the matching matrix line was coincident with a target location, and seven matrix points fell into each interlocation space for a total of 49 points in each line. Matrix points coincident with target locations were assigned the corresponding sensitivity values. Points that fell along interlocation lines were thought of as forming linear trabeculae. Groups of seven points formed trabeculae in the horizontal dimension,
1029
OPHTHALMOLOGY
•
AUGUST 1983
•
VOLUME 90
•
NUMBER 8
8.
A. 44
38~~~~~~~ 32 26
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY (Degrees)
c.
D.
-a) 44 -o
;
38
~
32
!::::
1-
44
k=~~~~~~~l!!=
~ 26 ILl
"'
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY
(Degrees)
(Degrees)
E.
F.
44
-
a)
;
1-
~
38
k="!!!!!!~~~~~~
32
>-
1-
>
;::
z 26 1-
co
-o
-
z
ILl
"'
44
38 32 26
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY (Degrees)
and groups of three formed interlocation trabeculae in the vertical dimension. (Horizontal and vertical trabeculae are diagrammed in Figure I.) Each point within a trabecula was assigned a sensitivity value by linear interpolation between the corresponding values of the two locations connected by the trabecula. Quadratic groups of 21 points formed intratrabecular portions of the image matrix, as indicated by the set of points labeled "S" in Figure 1. Each point within these groups was given a sensitivity value by linear interpolation between the two values of the matrix points of the vertical trabeculae lying to either side in the horizontal dimension.
1030
Fig 2. Examples of three-dimensional visual field surface following alternative methods of image smoothing. A: test object pattern for left eye used to cover point of fixation and include physiologic blind spot. B: results of three-dimensional interpolation without smoothing. (Arrow #I : point of fixation. Arrow #2: physiologic blind spot, tic marks indicate position of vertical meridian). C: results of excessive smoothing of surface image (32 passes of algorithm). D: image appearance following 32 passes of smoothing algorithm while restraining points located at test object locations. E: results of 32 passes of smoothing alogorithm while restraining all points along horizontal trabeculae. S: surface image appearance following two passes of smoothing algorithm while restraining all points of horizontal trabeculae.
-17 -13 -9 -5 -1 3 HORIZONTAL ECCENTRICITY (Degrees)
Following interpolation of the image matrix, its intertrabecular portions were smoothed by iteratively averaging each point with its immediate neighbors. The sensitivity vaues of all points lying within horizontal trabeculae were held constant during this process, so that the interpolated surface was constrained by the sensitivity values at the test object locations. The results of this surface interpolation technique are illustrated in Figure 2. All of the three-dimensional visual field surfaces produced for this work employed the linear interpolation technique followed by two passes of the restrained smoothing algorithm.
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
NORMAL
_44 co
~
>!::::
>
II/')
z
w
1/')
-10
0
+10
HORIZONTAL (Degrees)
CASE 1
44
co ~33 >1-
I 1e
~ 22
1-
1/')
z 11
w
1/')
0
+10
HORIZONTAL (Degrees) Fig 3. Top left, test object pattern used for threshold static perimetry of central visual field. Top right, appearance of normal central visual field. Tic marks indicate position of vertical meridian. Bottom left, kinetic visual field, case I. Bottom right, three-dimensional static visual field, case I, demonstrating central scotoma with relative preservation of foveal sensitivity.
RESULTS The central visual field was examined, using the object location pattern shown in Figure 3 (top left). The normal visual field surface with this technique of imaging has the appearance of a high plateau that rises through a gently sloping aspect to a small, but distinct peak located at the point of fixation. The point of fixation is located directly at the center of the area examined, and the vertical meridian is indicated on the surface by a row of tic marks. The visual fields shown in the bottom of Figure 3 (case 1) were obtained from the right eye of a 41-year-
old woman, who gave a history of having had a venous stasis retinopathy 10 years earlier. A fundus photograph from the patient's records confirmed the diagnosis. Although the visual acuity of the eye was 20/25 the patient complained of considerable difficulty reading due to a blind spot in the center of the visual field of the right eye. There was no afferent pupillary defect, and the ophthalmoscopic appearance of the macula (Fig 4) showed only dilation and tortuosity of the perifoveal capillaries (best seen by contact lens examination). Kinetic perimetry, however, (Fig 3, bottom left), demonstrated a central scotoma, approximately go wide in its horizontal dimension, and having a maximum density plottable with the l3e target. How visual acuity could be
1031
OPHTHALMOLOGY
•
AUGUST 1983
•
VOLUME 90
•
NUMBER 8
Fig 4. Case I : fundus photograph of macula, showing abnormalities of perifoveal capillaries.
Fig S. Case 2: appearance of macula, showing subretinal hemorrhage.
maintained at 20/25 with such a dense central scotoma, was explained by the three-dimensional contour of the visual field surface, which clearly demonstrated the presence of a small but distinct peak of higher visual sensitivity located at the fovea. A second case demonstrated a similar visual field pattern. A 64-year-old man, who had suffered loss of vision in the right eye 2 months before being seen, was found to have visual acuity of 20/50 in the affected eye. Ophthalmoscopic examination revealed a disciform macular degeneration with a prominent subretinal hemorrhage in the macula (Fig 5). Kinetic perimetry (case 2, Fig 6) demonstrated a large central scotoma with sloping borders. It appeared to be maximally dense at its center where it was present to the l4e target. The threedimensional visual field surface, however (top right, Fig 6), demonstrated an irregular perifoveal depression of the visual field with distinct preservation of a peak of sensitivity at the point of fixation. The third case noted to demonstrate this type of finding was that of a 73-year-old man who had undergone intracapsular cataract extraction with implantation of an iris fixation intraocular lens in the right eye 2 years ago. He complained of difficulty reading due to a central scotoma, and the right eye was found to have visual acuity of 20/60. Kinetic perimetry (bottom left, Fig 6) showed a small, moderately dense central scotoma. In this as in the first two cases, no hint of preserved visual sensitivity at the point of fixation was obtained during kinetic perimetry. The three-dimensional surface of the central visual field, however, clearly demonstrated the presence of relatively preserved visual sensitivity at the point of fixation (bottom right, Fig 6). Ophthalmoscopic examination disclosed the presence of cystoid macular edema. The fourth case was that of a 69-year-old man referred with complaints of difficulty reading due to the presence of central scotomas in both eyes. Visual acuity in both
eyes, however, was 20/25, and the results of kinetic perimetry, which were similar in both eyes, demonstrated the presence of irregular though dense central scotomas. Examination of both eyes (right eye shown in Fig 7) demonstrated geographic atrophy of the retinal pigment epithelium, making the underlying choroidal vessels prominently visible. There was, however, no visible scar, hemorrhage, or edema. The three-dimensional topography of the central visual field surface (bottom half Fig 8) demonstrated symmetrical, perifoveal central scotomas with a distinct tendency towards preservation of visual sensitivity at the point of fixation in both eyes. The fifth case involved a 33-year-old woman who complained of the sudden onset of central scotomas in both eyes. Visual acuity was 20/40 in each eye. The patient had undergone an emergency cholecystectomy during the third trimester of pregnancy, followed a short time later by a Cesarean section. After the second surgical procedure, the onset of reduced visual acuity was noted. Ophthalmoscopic examination demonstrated the presence of tiny, glistening white retinal inclusions in the maculae. There were also retinal pigment epithelial defects in the same area (right eye, Fig 9). These window defects were confirmed by fluorescein angiography, and had the effect of producing as so-called target maculopathy. The appearance was compatible with Stargardt's disease, although there was no family history of visual disorder. There was no history of drug abuse or use of antimalarial drugs. The visual fields (Fig 10) showed small, bilateral central scotomas by kinetic perimetry. Static perimetry demonstrated small, very dense central visual field defects, with marked sparing of visual sensitivity at the point of fixation in both eyes. This had the appearance of sharp spikes in the visual field surface, located at the center of the visual field surfaces. The subsequent course of the patient showed no change in the visual acuity or in the visual field findings. The sixth case was that of a 25-year-old woman who
1032
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
CASE 2
-44
.,co >......
33
> 22
......
z
IJ)
11
UJ
IJ)
-10
0 HORIZONTAL
+10
(Degrees)
CASE 3
-44
.,co >......
33
>
22
z
11
...... IJ)
UJ
IJ)
0 HORIZONTAL
+10
(Degrees) Fig 6. Top left, kinetic visual field, case 2. Top right, three-dimensional static visual field, case 2, demonstrating broad area of central visual field defect with relative preservation of foveal sensitivity. Bottom left, kinetic visual field, case 3. Bottom right, three-dimensional static visual field, case 3, demonstrating perifoveal depression of visual field surface with relative preservation of foveal sensitivity.
noted visual field defects in both eyes immediately following emergency Cesarean section for an eclamptic pregnancy. The patient, however, was not seen until 2 weeks after the onset of visual loss, at which time the appearance of the fundus was unremarkable (Fig 11 ). Fluorescein angiography showed no abnormality. Kinetic perimetry of the left eye demonstrated relative central scotomas generally surrounding the point of fixation. Static perimetry of the central visual field (top right of Fig 12) demonstrated definite sparing of visual sensitivity at the point of fixation. Note that the visual field surface has been rotated 180° in this figure to allow better visualization of the point of fixation. The visual field defect in the patient's right eye was also associated with a normal appearing fundus, but was cecocentral in distribution involving the physiologic blind spot and the temporal half of the macula, but not including the point of fixation.
Fig 7. Case 4: appearance of macula, right eye, showing geographic atrophy of retinal pigment epithelium.
1033
OPHTHALMOLOGY
•
AUGUST 1983
•
VOLUME 90
•
NUMBER 8
CASE 4
RIGHT EYE
LEFT EYE
-44
-44
-33
-33
I-
> 22
I-
I-
I-
..,
..,
a::l
a::l
>-
>-
1.1)
zLU
~ 1.1)
11
z
22 11
LU
1.1)
1.1)
v~
o
fD Rr,c eg,.ee -<1( S)
-10
0 HORIZONTAL (Degrees)
+10
-10
0 HORIZONTAL (Degrees)
+10
Fig 8. Case 4: kinetic visual fields {top) demonstrate bilateral central scotomas plotted with the 1.. target with small islands of greatest density in a perifoveal area. Three-dimensional static visual fields {bottom) show bilateral, dense central scotomas with marked preservation of visual sensitivity at the fovea.
Fig 9. Case 5: appearance of macula, right eye, demonstrating retinal pigment epithelial defects in a perifoveal, ring-shaped pattern. A few small white intraretinal dots can be seen.
1034
No instance offoveal sparing in the central visual field was observed for any patient with a central scotoma due to optic nerve disease. Case 7 (lower half, Fig 12) was that of a 30-year-old man being followed for multiple sclerosis. The kinetic visual field was obtained 8 weeks after the abrupt onset of retrobulbar optic neuritis in the right eye. There was a cecocentral scotoma with its greatest density in the superior half of the central visual field. The three-dimensional visual field surface (bottom right, Fig 12) was reversed to allow the point of fixation to be seen more easily. It clearly demonstrates a depression of the foveal sensitivity that falls well within the margins of the central visual field defect. This finding is representative of central and cecocentral scotomas arising from optic nerve disease. The constellation of findings in patients with selective preservation of foveal sensitivity has included relatively preserved visual acuity (20/60 or better) in the presence of signs and symptoms of a central scotoma and/or
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
CASE 5
------~ -- m4 e
l3e
I2e
LEFT EYE
RIGHT EYE
44
lil
co ~ 33 >-
~
>- 33
I-
t:::
I-
I-
> 22
> 22
V)
z UJ
44
V)
11
z
11
UJ
V)
V)
-10
0 HORIZONTAL (Degrees)
+10
-10
0 HORIZONTAL (Degrees)
+10
Fig 10. Case 5: kinetic visual fields (top) show small, dense central scotomas plotted with the l3e target. Three-dimensional static visual fields (bottom) demonstrate small, very dense central scotomas with preservation of visual sensitivity at fovea in both eyes.
macular disease. The common complaint was of inability to read, and to a degree greater than would be expected on the basis of acuity impairment alone. The prevalence of foveal preservation as a finding in central scotomas arising from macular disease has not been determined. We have used contiguous area static perimetry to examine 30 eyes in 38 patients who had signs or symptoms of central scotomas. Of these, nine eyes in six patients showed foveal sparing within central scotomas (20% of cases). This finding was not demonstrated in any patient in our series whose visual acuity had fallen to 20/80 or less. Thus, foveal sparing is not likely to help in differentiating macular and optic nerve disorders when acuity is below this level. To date we have used contiguous area static perimetry to examine 84 eyes in 64 patients with central visual field defects arising from optic nerve disease, including ischemic, toxic, and demyelinating neuropathies. In none of these
Fig 11. Case 6: fundus, left eye, showing unremarkable appearance of macula. A fluorescein angiogram was likewise normal.
1035
OPHTHALMOLOGY
•
AUGUST 1983 •
VOLUME 90
•
NUMBER 8
CASE 6
CASE 7
+10
0
-10
HORIZONTAL (Degrees) Fig 12. Cases 6 and 7: top left, kinetic visual field, case 6, demonstrating variably dense central scotoma with perifoveal distribution. Top right, three-dimensional static visual field, case 6. Note that orientation of surface has been rotated 180° to allow better appreciation of shape of surface at foveal position (at center of pattern). The central scotoma is seen to be very dense, small, and perifoveal in distribution. Bottom left, kinetic visual field, case 7, shows dense, cecocentral scotoma with a superior arcuate pattern. Bottom right, three-dimensional static visual field, case 7, demonstrating altitudinal distribution of depressed visual sensitivity, and definite depression of sensitivity at fovea (indicated by arrow). Note that this surface has also been rotated 180° to allow better visualization of point of fixation.
cases has there been evidence of preferential sparing of foveal sensitivity.
DISCUSSION Relative preservation of visual sensitivity at the fovea (within a central scotoma) may be a more common accompaniment of macular disease than has been previously recognized. Prior reports of ring scotomas occurring within 10° of fixation have been associated with chloroquine retinopathy. 6 Ring maculopathies not associated with chloroquine use are not uncommon. 7- 9 and are usually associated with dominantly inherited
1036
cone dystrophy, Stargardt's disease, early disciform macular degeneration and possibly with late stages of central serous retinopathy. Central scotomas in some of these cases7 are known to be annular. The more commonly recognized ·ring-shaped visual field defect is that associated with pigmentary retinal degenerations. This type of defect, however, occurs in the midperiphery of the visual field and is usually associated with some degree of sparing of the macula. The cases we report were detected at random during examination of a variety of patients known to have macular and optic nerve disorders. The fact that this phenomenon has not been more widely recognized is probably due to the limitations of the techniques of perimetry customarily used
HART AND BURDE • CENTRAL SCOTOMAS OF MACULAR DISEASE
to explore the central visual field. Most of the patients with this finding have not subjectively appreciated the fact that foveal sensitivity was preferentially spared, nor were they ordinarily capable of describing this phenomenon when using the Amsler grid. Use of the Goldmann perimeter precludes easy detection of the phenomenon unless the eccentric fixation target is used while exploring the visual field at and around the point of fixation. Otherwise the small area of preserved sensitivity will be covered by the standard fixation target. If the eccentric fixation device is used, it may be possible with kinetic perimetry to demonstrate foveal preservation. However, static perimetry remains the most sensitive and definitive technique. The use of contiguous area static perimetry in which quadratic grids of test object locations are used to blanket an area of the visual field, has made demonstration of the foveal preservation phenomenon quite easy. The disease processes associated with relative foveal preservation are predominantly vascular and degenerative disorders of the macula. One case was associated with a prior venous stasis retinopathy (case 1), one was found to have cystoid macular edema (case 3), three had degenerative diseases of the macula (cases 2, 4, and 5), and one (case 6) was unexplained. It would be expected that toxic maculopathies, such as those resulting from the use of antimalarial agents that produce "target" lesions within the macula, might also be associated with ring depressions of the central visual field, showing relative preservation of foveal sensitivity. However, we have examined only one case in which a fully developed target lesion of the macula had resulted from long-term use of chloroquine phosphate. There was severely reduced visual acuity (20/200), and the visual field surface failed to show preservation of foveal sensitivity. The macula, a highly specialized region of the retina, is adapted to the tasks of high resolution vision, included high contrast sensitivity, and color perception. The unique anatomic structures associated with this specialization provide a number of possible explanations for the phenomenon of relative preservation of foveal sensitivity in macular disease. The fact that the fovea is rod-free, and that the anatomic density of rods is greatest in the perifoveal macula, would explain a perifoveal depression of the visual field in diseases for which rods are selectively sensitive. Thus, a macular degenerative process that preferentially disrupts rod function, would be expected to preserve visual sensitivity at the fovea. Also, the fact that the foveal retina is avascular may provide an explanation for preservation of its sensitivity in diseases involving the perifoveal capillaries. Unless such an insult were predominately felt towards the outer layers of the retina however, it would be difficult to use this as an explanation, since the afferent processes of foveal cones must pass through the perifoveal macula to reach their respective bipolar cells. The radial distribution of these processes within Henle's fiber layer, and their absence in the fovea, may account for relative pres-
ervation offoveal sensitivity in macular edema. Selective sequestration of intraretinal edema in this specialized region of the outer plexiform layer could produce damage to retinal function in the extrafoveal macula, while leaving foveal function relatively intact. This would be a functional correlate of the common clinical observation of a "petaloid" or radial distribution of loculated fluid, frequently found in macular edema. Other perimetric tests (of increasing complexity) have been found to differentiate between visual field defects arising from macular and optic nerve disorders. Acuity profile perimetry in macular disease has shown a dissociation from brightness increment sensitivity, whereas optic nerve disorders tend to affect both functions to a more equal extent. 10- 12 Clinical adaptations of perimetric tests that mimic retinal receptive field properties have been used to study visual field defects arising from retinal and optic nerve disorders. 13- 14 With damage to the choroid or outer layer of the retina, reduced brightness sensitivity, decreased dark adaptation, and an altered Stiles-Crawford function can be found with preserved receptive field-like properties (including sustained-like and transient-like functions). 15- 16 Repetitive sensitivity testing to demonstrate a " fatigue" effect (the so-called flashing repeat static test) has reportedly produced no such phenomenon in retinal diseases, although it is apparently characteristic of retrolaminar disorders of the optic nerve. 17 The finding of relative preservation of foveal sensitivity in a variety of macular diseases, serves to underscore the relative insensitivity of visual acuity measurement as a determinant of the presence or extent of functional damage to the macula. It is common clinical experience to observe patients with extensive ophthalmoscopic evidence of macular disease, who none-theless have no impairment of visual acuity. It is likewise a not infrequent experience to encounter patients, who in spite of undiminished visual acuity, complain of significant visual dysfunction due to central scotomas ar metamorphopsia. Although no cases have yet been observed of foveal preservation in central scotomas that have arisen as a result of optic nerve disease, more experience will have to be gathered before we can definitely say that this finding is uniform. If that proves to be the case, however, the demonstration of relative sparing of foveal sensitivity in central visual field defects may serve as a convenient marker for macular disease, as distinct from visual field defects resulting from optic nerve disorders.
REFERENCES 1. Spahr J . Zur Automatisierung der Perimetrie. I. Die Anwendung eines computergesteuerten Perimeters. Albrecht von Graefes Arch Klin Exp Ophthalmol 1973; 188:323-38. 2. Weber 8, Spahr J . Zur Automatisierung der Perimetrie; Darstellungs-
1037
OPHTHALMOLOGY
3. 4.
5. 6.
7. 8. 9. 10.
•
AUGUST 1983 •
methoden perimetrischer Untersuchungsergebnisse. Acta Ophthalmol1976; 54:349-62. Heijl A Studies on computerized perimetry. Acta Ophthalmol 1977; Suppl132. Frankhauser F, Bebie H: Threshold fluctuations, interpolations and spatial resolution in perimetry. Doc Ophthalmol Proc Ser 1979; 19:295-309. Hart, WM Jr. Hartz RK. Computer-generated display for three-dimensional static perimetry. Arch Ophthalmol 1982; 100:312-8. Okun E, Gouras P, Bernstein H, von Sallmann L. Chloroquine retinopathy; a report of eight cases with ERG and dark-adaptation findings. Arch Ophthalrnol 1963; 69:59-71 . Grey RHB, Blach RK, Barnard WM. Bull's eye rnaculopathy with early cone degeneration. Br J Ophthalmol1977; 61 :702-18. Weise EE, Yannuzzi LA. Ring maculopathies mimicking chloroquine retinopathy. Arn J Ophthalmol 1974; 78:204-10. Krill AE , Deutman AF. Dominant macular degenerations; the cone dystrophies. Am J Ophthalmol1972; 73:352-69. Johnson CA. Keltner JL, Balestrery F. Effects of target size and eccentricity on visual detection and resolution . Vision Res 1978; 18:1217- 22.
1038
VOLUME 90 •
NUMBER 8
11 . Johnson CA, Keltner JL, Balestrery FG. Acuity profile perimetry; description of technique and preliminary clinical trials. Arch Ophthalmol 1979; 97:684-9. 12. Keltner JL, Johnson CA, Cowley IJ. Acuity profile perimetry in a unique case of bilateral central serous retinopathy. Ann Ophthalmol 1980; 12:726-32. 13. Sunga RN, Enoch JM. A static perimetric technique believed to test receptive field properties. Ill. Clinical trials. Am J Ophthalmol 1970; 70:244- 72. 14. Sunga RN, Enoch JM. Further perimetric analysis of patients with lesions of the visual pathways. Am J Ophthalmol 1970; 70:403-22. 15. Fitzgerald CR, Enoch JM, Campos EC, Bedell HE. Comparison of visual function studies in two cases of senile macular degeneration. Albrecht von Graefes Arch Klin Exp Ophthalmol1979; 210:79-91 . 16. Campos EC, Bedell HE, Enoch JM, Fitzgerald CR. Retinal receptive field-like properties and Stiles-Crawford effect in a patient with traumatic choroidal rupture. Doc Ophthalmol 1978; 45:381-95. 17. Enoch JM, Fitzgerald CR, Campos EC. Quantitative Layer-by-Layer Perimetry; An Extended Analysis. New York: Grune and Stratton, 1981; 195-223.