Investigative & Visual Science, Vol. 29, No. 1, January 1988 Copyright © Association for Research in Vision and Ophthalmology

Central Visual Fields For Short Wavelength Sensitive Pathways in and Ocular Hypertension

Gordon Heron,*j- Anthony J. Adorns,* and Roger Husred^:§

While conventional clinical visual acuity and kinetic visual fields may be essentially normal in ocular hypertension and early stages of glaucoma, other foveal aspects of vision (eg color, spatial and temporal contrast sensitivity) may be quite abnormal. Specifically, a selective vulnerability of the short wavelength sensitive (SWS) visual pathways in these conditions has previously been noted. Here we studied the central visual fields of 33 primary open angle glaucoma (POAG) patients, 32 ocular hypertensives (OHT), and 24 age-matched normal controls using blue and yellow test flashes on bright yellow backgrounds. SWS cone and MWS and/or LWS cone pathway sensitivities were measured at the fovea and at 2.5°, 5°, 10° and 15° eccentricities, in either the inferior temporal (for OHT) or horizontal nasal (for POAG). As expected, all groups had normal sensitivity to yellow flashes —detected by LWS and/or MWS cones—in these meridians. By comparison, for the blue flashes— detected by the SWS cones—the POAG and OHT groups had sensitivity deficits, uniformly across the central visual field, of about 6X and 1.8X, respectively, compared to normals. While six of 31 (19%) OHT subjects had localized glaucomatous field defects (>0.4 log units) in the non-foveal inferior temporal retina, none of the 12 OHT subjects who were also tested in the horizontal nasal retina showed loss in this meridian. Finally, while no POAG subjects had localized sensitivity loss for yellow flashes in the horizontal nasal retina, four did show local field defects with blue test flashes. These results are consistent with significant local and diffuse ganglion cell loss involving the blue- sensitive visual pathways in both OHT and POAG patients. Invest Ophthalmol Vis Sci 29:64-72,1988

The loss of nerve fibers in glaucoma is thought to Traditionally, greater emphasis has been placed on occur in two ways. A localized loss of axons, usually evaluating the localized loss of nerve fibers, using vi- occurring at the poles of the head,1 pro- sual field evaluation. The earliest localized visual duces the familiar changes of asymmetric cupping of field loss typically occurs outside of the central macu- the disk and notching of the neuroretinal rim. This lar area and, in recent times, much work has been loss may be evident as a defect of the retinal nerve carried out on improving methods of visual field ex- fiber layer2"4 and give rise to the typical nerve fiber amination to provide increasingly more sensitive bundle defects found in the Bjerrum areas of the vi- tests of the earliest signs of functional loss. Whether sual fields. The second type of axon loss produces a the localized loss of nerve fibers can first be detected generalized reduction in retinal sensitivity, and may functionally by the visual fields,o r structurally by ex- be observed at the nerve head as an overall enlarge- amination of the optic nerve head and retinal nerve ment of the optic cup with a reduction in the neuro- fiber layer, is currently a matter of some debate.7'8 retinal rim area.4"6 The recent reports by Quigley1'8"10 which suggest that patients with mild or moderate visual field defects From the "School of Optometry, University of California at may have lost as many as 50% of their axons at the Berkeley, Berkeley, California, and the ^Ophthalmology Clinic, optic nerve head have stimulated interest in identify- Silas Hayes Hospital, Fort Ord, California. ing the earliest possible changes in glaucoma. More f On leave from Glasgow College of Technology, Glasgow, provocative is the evidence from Quigley's studies Scotland. § Lt. Col., MC, USA. The views expressed in this article are those suggesting that even ocular hypertensive eyes, with of the authors and do not reflect official United States Army or no clinically manifested visual field loss or change in Department of Defense policy. the appearance of the nerve head, may have as many Supported by NIH giant EY-02271 to AJA. as 35% of the axons of the nerve head missing. Submitted for publication: December 3, 1986; accepted July 6, 1987. A generalized reduction of retinal sensitivity from Reprint requests: Anthony J. Adams, School of Optometry, Uni- the diffuse loss of axons, although present, is difficult 4 5 versity of California, Berkeley, CA 94720. to demonstrate by visual field examination. - How-

64

Downloaded from iovs.arvojournals.org on 09/23/2021 No. 1 SWS CONE FIELDS IN GLAUCOMA AND OCULAR HYPERTENSION / Heron er ol. 65

ever, a number of changes of visual function at the inferior retina with about 30% of them also tested fovea have been reported and are usually cited as along the nasal horizontal retina. Here we were inter- evidence of diffuse loss. Much of this evidence comes ested in revealing diffuse or localized field loss that from color vision studies, although studies of spa- clearly had not been identified by conventional clini- tial""13 and temporal14 contrast sensitivity in the cal field testing. (The 40° meridian in the temporal central macula also contribute to this conclusion. In inferior retina is thought to represent one of the earli- fact, even though visual acuity is normally thought to est locations for localized loss in the development of be unaffected in the earlier stages of glaucoma, care- glaucoma.) A similar sized group of age-matched ful assessment often shows that it may be mildly re- control subjects was used for measurements in both duced.5 meridians. A number of studies of foveal color vision can now be cited as showing that, in glaucoma, hue discrimi- Materials and Methods nation is reduced selectively in the blue-yellow axis of Glaucoma and ocular hypertensive patients were color discrimination.15"23 More specifically, both recruited for this study from the Eye Clinic of the achromatic and chromatic sensitivity is reduced, with Silas Hayes Hospital at Fort Ord, California. Glau- the greatest sensitivity loss occurring in the blue end coma patients (n = 33) were of the primary open of the spectrum.24"30 Surprisingly, similar results have angle type (POAG), had visual acuities of 20/40 or been reported among many ocular hypertensive pa- better in the tested eye (all except two had 20/30 or tients where no localized loss of axons has been pre- better), were under reasonable control and had not sumed.20-23'24'28'31 Now there is evidence linking fo- undergone any laser or filtration surgery. Seventeen veal color vision loss to diffuse retinal nerve fiber of the POAG patients had small paracentral or ar- loss.32 Such findings, along with reports of loss of cuate only or an isolated nasal step, five had other central vision function in many ocular hyper- large paracentral or arcuate scotomas including nasal tensives, raises the possibility that diffuse loss may steps, six had marked nasal or temporal depression precede localized loss in the evolution of glaucoma- with large paracentral or arcuate scotomas, and five tous field loss. Of course such a conclusion must were categorized as endstage with small temporal or always be tempered by the fact that the relative sensi- central islands of vision. All subjects consented to tivity of a particular test paradigm can confound any participation after details of the study and its proce- conclusions about whether central diffuse loss of vi- dures were explained. sion precedes or follows localized field defects in The ocular hypertensives (OHT, n = 32) were de- glaucoma. fined as having IOP (by applanation) of greater than Against bright yellow backgrounds, which depress 22 mm Hg on two or more occasions, no visual field the sensitivity of MWS and LWS cones, SWS cones defects (Humphrey Instruments, San Leandro, CA; are responsible for the detection of deep blue flashes Central Threshold Test: Stimulus III white and back- of light in the normal eye. We have shown pre- ground 31.5 apostilbs), normal appearance of the viously33 that the actual conditions we used in this optic nerve head, and acuity better than 20/30 in the study isolate short wavelength sensitive (SWS) cones tested eye. in just this way. SWS cones play no role in detection The age-matched normal control group (n = 24) of yellow flashes superimposed on a bright yellow had acuity better than 20/30, pressures less than 21 background; this is done by MWS and/or LWS mm Hg, no glaucomatous visual fielddefects , no field cones. In this study we use the test of SWS cone loss, and no pathological cupping, pallor, or edema of thresholds to examine the foveal and central visual the disk. All of the subjects in the study had only field sensitivity in groups of glaucoma, ocular hyper- minor cortical and nuclear changes, and were tension and age-matched normal subjects. We were free of retinal disease by fundus examination and particularly interested in identifying both local and fundus photography. diffuse sensitivity loss. In the glaucoma patients sensi- SWS cone thresholds were determined using an tivity profiles were determined primarily along the 11° yellow (Schott, Jena, East Germany; filter nasal retina horizontal raphe with about 30% of the OG530) background onto which a 1° blue (Kodak, subjects also being tested along the 40° meridian in Rochester, NY; Wratten filter 47B) test spot was pro- the temporal inferior retina. The nasal retina hori- jected. For comparison, a threshold was obtained zontal meridian was chosen as representative of a using a 1 ° yellow spot projected onto the same back- "low risk" site for glaucomatous field loss by conven- ground. (This spot is detected by MWS and/or LWS, tional field testing techniques. Ocular hypertensives but not SWS cones.) A viewing distance of 33 cm was were tested along the 40° meridian in the temporal used and a suitable optical correction worn when ap-

Downloaded from iovs.arvojournals.org on 09/23/2021 66 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / January 1988 Vol. 29

••- Normal (24) °- OHT (32) ••- POAG (33) tify threshold at each of the retinal eccentricities tested. For the glaucoma and normal groups mea- 4.20 surements were taken at the fovea and at points 2.5°, 5° and 10° horizontally into the nasal retina. For the ocular hypertensive and normal groups, thresholds were established at the fovea and at points 2.5°, 5°, 10° and 15° on the inferior temporal retina along the 40° meridian. Testing of the different retinal regions was achieved by varying the position of the fixation target. Eleven of the glaucoma patients and 12 of the 1.80 ocular hypertensive patients were also tested along 4 6 8 10 16 Eccentricity (degrees) the inferior temporal retinal meridian (40° from hori- zontal) and horizontally on the nasal retinal merid- Fig. 1. Log sensitivity to 1° yellow test flashes(30 0 msec) on an ian, respectively. 11° yellow background at the fovea and at 2.5°, 5°, 10° and 15° eccentricity. Normal and ocular hypertensive groups measured along the inferior temporal retina, glaucoma group measured along Results the nasal retina to 10° eccentricity. Sensitivities determined by method of limits. Mean ages, numbers of subjects, test and back- Yellow on Yellow Test Sensitivities ground dimensions, and light levels as in Table 1. ± One standard The sensitivity to yellow test flashesi s highest at the deviation is indicated for the normal group. fovea and gradually drops off out to 15 degrees eccen- tricity in the normal eye (Fig. 1). This is consistent propriate. The background luminance for the yellow with the profile of the "hill of vision" for white pho- field was 500 cd/m2. The background and test beams topic backgrounds normally encountered in clinical were calibrated in the instrument by a simple inex- perimetry. Furthermore, the sensitivity falls at essen- pensive circuit and photocell arrangement prior to tially the same rate in the inferior temporal retina as it each test session.33 After 2 min of adaptation to the does in the nasal retina. It can be seen from Figure 1 background, subjects were asked to fixate a marker that the sensitivity profiles for the glaucoma and ocu- while test flashes were introduced for 300 msec, each lar hypertension groups do not differ from that of the second, by a rotating sector disk. The test spot inten- normal group either in shape or absolute sensitivity. sity was varied in 0.1 log unit steps by Kodak (Roch- In fact there is no statistically significant difference ester, NY) carousel rear projection through calibrated between the three groups for any of the eccentricities Wratten neutral density filters. A descending and as- measured. This result is not surprising since ocular cending staircase method of limits was used to iden- hypertensives were included in this study because of their normal fields, and most glaucoma patients would not be expected to have sensitivity loss in the

••- Normal (24) -°- OHT (32) ••- POAG (33) horizontal nasal retina unless their visual field defects could be described as severe. 4.20 Blue on Yellow Test Sensitivities Quite a different result is seen for sensitivity pro- files determined with blue test flasheswher e the SWS cone mechanisms are responsible for detection. Fig- ure 2 shows that, in the normal eye, the sensitivity to blue test flashes is greatest at about 2° retinal eccen- tricity and falls off gradually towards the fovea and 1.80 towards greater eccentricities. This result could have 4 6 8 10 Eccentricity (degrees) been anticipated from the known distribution of SWS Fig. 2. Log sensitivity to 1 ° blue test flashes(30 0 msec) on an 11 ° receptors and the presence of the yellow macular pig- 34 yellow background at the fovea and at 2.5°, 5°, 10° and 15° eccen- ment in the central 2°-4°. The ocular hypertensive tricity. Normal and ocular hypertensive groups measured along the and glaucoma patients are significantly less sensitive

Downloaded from iovs.arvojournals.org on 09/23/2021 No. 1 SWS CONE FIELDS IN GLAUCOMA AND OCULAR HYPERTENSION / Heron er ol. 67

Table 1. Log sensitivities ±1 sd for central visual field

B test flash Y test flash Retinal location n Mean age Fovea 2.5° 5° 10° J5° Fovea 2.5° 5° J0° 15°

Glaucoma nasal 32 65.2 2.42 2.67 2.58 2.38 — 3.90 3.69 3.55 3.35 — n = 33 ± 9.0 ±0.57 ±0.55 ±0.63 ±0.59 ±0.20 ±0.21 ±0.23 ±0.28 mean age inferio- 11 66.5 2.47 2.62 2.43 2.20 1.98 4.09 3.82 3.65 3.20 3.06 65.4 temporal ± 7.2. ±0.49 ±0.45 ±0.61 ±0.55 ±0.80 ±0.15 ±0.13 ±0.21 ±0.88 ±0.94 OHT nasal 12 63.8 2.82 3.06 2.93 2.63 — 3.94 3.72 3.61 3.48 — n = 32 ± 14.0 ±0.27 ±0.24 ±0.25 ±0.27 ±0.13 ±0.14 ±0.20 ±0.24 mean age inferio- 31 60.9 3.02 3.12 2.99 2.77 2.65 3.95 3.71 3.56 3.41 3.31 61.3 temporal ± 15.6 ±0.42 ±0.37 ±0.42 ±0.48 ±0.46 ±0.13 ±0.16 ±0.17 ±0.16 ±0.16 Normals nasal 24 59.8 3.21 3.39 3.30 3.08 — 3.95 3.73 3.59 3.41 — n = 24 ± 9.8 ±0.26 ±0.31 ±0.33 ±0.34 ±0.09 ±0.12 ±0.12 ±0.12 mean age inferio- 23 60.4 3.17 3.34 3.26 3.08 2.91 3.93 3.67 3.51 3.39 3.27 59.8 temporal ± 9.4 ±0.26 ±0.21 ±0.21 ±0.29 ±0.29 ±0.09 ±0.11 ±0.12 ±0.15 ±0.14

Log sensitivities ± 1 standard deviation for glaucoma, ocular hypertension, meridian. Y = yellow test flash, B = blue test flash. All thresholds measured and normals for the central 15° of retina (method of limits). Inferior tem- for a 1° spot (300 msec) against an 11° yellow background (Schott OG530) poral retina = 40° below horizontal meridian. Nasal retina = horizontal set at 500 cd/m2. See text for details.

alized loss of sensitivity is between 0.2 and 0.3 log less than the 0.8 log unit difference seen between units. Even this clinically small difference between normals and glaucoma patients. Second, in an at- the ocular hypertensive group and the normal group tempt to analyze our data with a tighter control on is statistically significant (P < 0.05) at each eccentric- age matching, we paired 21 glaucoma patients indi- ity. The generalized sensitivity loss for the glaucoma vidually age-matched to 21 normals. Here the mean group is more marked, being about six times less sen- age differed by less than 1 year with the normals being sitive (0.8 log units) than the normal control group slightly older. The difference in SWS cone thresholds and four times less sensitive (0.6 log units) than the for the two groups was 0.7 log units and was statisti- ocular hypertensive group. The results for both yel- cally significant (P < 0.001, two-tailed t test). Clearly low and blue test flash threshold measures in the var- age effects cannot account for the differences seen in ious retinal locations for all three groups are shown in our results between normals and the other two study Table 1. groups.

Age-Matched Comparisons Size

It can be seen from Table 1 that the mean age for Pupil size differences between the two groups could each of the three groups is very similar. However, also conceivably influence the test results. Theoreti- since we know that the ocular media becomes more cally, a small pupil could lower the test sensitivity by yellow with age and could be expected to influence creating an effectively longer path length through the the detection of blue flashes, it is important to estab- thicker portion of the crystalline lens, or it might re- lish that the differences we see between the three duce the effective background to bring the patient off groups cannot be attributed to age. Two comments the Weber portion of the threshold versus intensity bear directly on this issue. First, in a separate study curve. In normals, we had previously established that (Adams, unpublished) using the same instrument, at the background light levels used for this test even one of the authors has shown that there is no signifi- relatively large changes in effective background in- cant change in sensitivity to yellow spots on a yellow tensity—due to ocular media attenuation of pupil background with age for normal eyes, but an approxi- size—should not significantly influence test scores.33 mately 0.2 log unit decline per decade for the detec- Nevertheless, in an attempt to directly test the possi- tion of blue spots by the same eyes. Since the differ- ble influence of pupil size on test results for our glau- ence in mean age for the three groups is less than 1 coma patients we compared the sensitivities of those decade, we would anticipate that normal aging effects glaucoma patients whose pupil diameter was less would account for less than 0.2 log units; this is far than 3 mm (n = 12) to those whose pupil diameter

Downloaded from iovs.arvojournals.org on 09/23/2021 68 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Jonuory 1988 Vol. 29

o = WITHIN 0.4 LOG UNIT OF EXPECTED NO. = OEFECT DEPTH IN DB four retinal eccentricities out to 15° in the inferior 32 DB = CENTRAL REF LEVEL temporal retina. By normalizing each individual's re- f <. e 6 o o sults at the fovea and looking only for deviations greater than 0.4 log units from the normal profile of sensitivity across the retina, we found that six of 31

f17_o+_o_ » 30* ocular hypertensives had field defects for at least one retinal eccentricity. While only 12 of the ocular hy- pertensive patients were tested in the nasal retina, none of those showed comparable loss at any of the eccentricities tested in that retinal meridian. We were also interested in whether or not glaucomatous field defects in glaucoma patients were either more exten- sive or of greater magnitude for the blue test flash condition than for the detection of yellow flashes. None of the 33 glaucoma patients had localized sensi- — Y-Oblique B-Oblique — Normal sens. tivity loss (greater than 0.4 log units) for yellow

Downloaded from iovs.arvojournals.org on 09/23/2021 No. 1 SWS CONE FIELDS IN GLAUCOMA AND OCULAR HYPERTENSION / Heron er ol. 69

be detected selectively by the SWS cones, may be a point to the use of color perimetry for blue targets on sensitive perimetric test for functional loss in glau- a bright yellow background. The usefulness of color coma and ocular hypertension. perimetry has been questioned in the past; despite the We have already noted that a number of studies interest it generated in the 1920s,3637 it was largely suggest that foveal color discrimination losses, partic- discredited by the publication of DuBois-Poulsen's ularly in the blue-yellow axis, are frequently reported book in 1952.38 Many of the problems of color perim- for patients with glaucoma. Some have suggested that etry appear to have been related to the reliable isola- the acquired dyschromatopsia is related to the sever- tion and control of chromatic and luminance param- ity of the visual field defect.182027'28 Drance et al21 eters of the presenting stimuli. The biggest problem, suggested that error scores on the FM 100-Hue test or however, may have been the lack of a clear and com- abnormalities in the yellow-blue or green-blue spec- mon purpose for the use of color to define the visual trum on the anomaloscope were predictive of subse- field. In many cases the color target acted only as a quent development of localized nerve fiber bundle reduced luminance test target; in other instances pa- field defects in suspected glaucoma. However, in spite tients were required to name the color as it moved of this a significant number (about 25%) of eyes with through the visual field. Beginning in the mid-1970s advanced glaucomatous visual field defects show no color has been used for a quite different purpose in abnormality in color discrimination.32 This lack of a perimetry. Modern color perimetry manipulates "clear-cut relationship" between the color discrimi- color in an attempt to isolate detection by different nation and the severity of the glaucomatous visual underlying vision mechanisms.39"46 Most of these ap- field defect suggested to Airaksinen et al32 that a sepa- proaches have attempted to isolate each of the three rate mechanism may be responsible for the color vi- cone mechanisms in the visual field measurement. It sion and visual field deficits. They found that color is along these lines that we suggest that perimetry for vision losses on the anomaloscope could significantly the SWS cone mechanism may be most useful in predict diffuse loss, but not localized loss, in the reti- early glaucoma detection. The conditions we use, a nal nerve fiber layer. This result lends further support deep blue test target against a bright yellow back- to the suggestion by Flammer35 that the diffuse loss ground, were specifically designed to isolate visual and the localized loss may have separate, but not detection by the SWS cones. mutually exclusive, origins. He suggests that the dif- There have been at least four recent studies that fuse loss occurs as a result of direct mechanical dam- have used colored test lights in testing the visual fields age related to raised and is re- of glaucoma patients, including the use of blue test vealed functionally, for example, by abnormal color spots.47"50 Two reports47'48 provide preliminary evi- vision and depressed contrast sensitivity. On the dence for blue test stimuli being more useful than other hand, he proposes that localized loss occurs white stimuli in glaucoma visual field testing. In par- primarily as a vascular disorder and may, for exam- ticular, a case report for a woman with chronic open ple, result from a microinfarction at the optic nerve angle glaucoma shows that, with white stimuli, the head producing the familiar nerve fiber bundle de- abnormality is of the order of 0.4 log units while with fect. This loss may be much less directly related to a blue stimulus the abnormality is of the order of 1.0 intraocular pressure. In his scheme raised intraocular log units. In a more extensive study, Logan and An- pressure can also be a cause of localized loss, espe- derson49 tested 60 eyes and concluded that the blue cially in cases where there are other vascular risk fac- stimulus was not more sensitive in detecting glauco- tors. Consequently, in individual cases, axon loss matous defects than the white stimulus. However, the may result from more than one cause. Unfortunately, white background light levels that they used were too our results do not provide any evidence, one way or low to isolate SWS cones with the blue test stimulus. another, for separate mechanisms involved in the Consequently, the test target was probably detected diffuse and localized axon loss. They do show that by other cone types (LWS and/or MWS) which, from loss of foveal sensitivity for the SWS cone pathways is our results, would not be expected to show any selec- associated with a generalized loss of sensitivity across tive sensitivity loss when compared to white or yellow the entire central field, consistent with the proposi- test stimuli. In the fourth study of early glaucomatous tion that color discrimination losses at the fovea are cases, using a color campimeter, Abe et al50 found related to diffuse retinal nerve fiber loss.32 greater reduction in sensitivity in the Bjerrum area Our results also show that localized glaucomatous- for the "blue-sensitive" mechanism than for the "red- like field defects may be present for blue test targets or green-sensitive" mechanisms or for a white stim- which are detected by the SWS cone mechanisms, ulus. when the same defect is not apparent with more The selective vulnerability of the SWS cone path- standard yellow or white test targets. These results ways in glaucoma has been suggested by four pre-

Downloaded from iovs.arvojournals.org on 09/23/2021 70 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / January 1988 Vol. 29

vious studies24'2951"'1 and a similar loss has been re- shown that this approach reveals defects not apparent ported for some ocular hypertensives in two of these from conventional field and acuity testing.6162 These studies.24-51 These latter studies also revealed loss of visual field defects were revealed using low spatial sensitivity to fast-flickering stimuli at the fovea for frequency targets which were flickered at moderate both glaucoma and ocular hypertensive patients. A flicker rates. Such test conditions are ideally suited growing body of physiological evidence from both for "isolating" pathways involving larger axons to the human and primate studies may help explain this magnocellular layers of the lateral geniculate nu- selective loss of the SWS pathways and pathways cleus.63 Also related is the growing body of evidence which carry high temporal and low spatial frequency that the pattern ERG is reduced in ocular hyperten- signals. sion and glaucoma while the flash ERG is unaf- 64 66 Color signals appear to be carried in two separate fected. " This result has been confirmed recently in 67 parallel pathways with origins in the retina of the a study of experimental glaucoma in monkeys. Of primate ; one carries signals relaying significance is the fact that the pattern ERG ampli- blue and yellow and the other relays signals for red tude was particularly reduced for low spatial fre- 67 and green.52 The red-green cells may also carry quency patterns; Marx et al suggest that this dys- achromatic information, multiplexed with the red- function may be associated with Y ganglion cells green color information.53'54 The axons of the gan- whose larger fibers terminate in the magnocellular glion cells of both these pathways, which project to layers of the lateral geniculate nucleus. Finally, Drum 68 the parvocellular layers of the lateral geniculate nu- et al have shown relative elevation of diffuse scoto- cleus, are smaller than those carrying the signals to pic perimetric thresholds, compared to photopic the magnocellular layers.55 The SWS cones appear to thresholds, in both glaucoma and ocular hypertensive send signals only to the chromatic pathways56 and patients and they cite evidence that this is consistent provide little or no basis for everyday vision of high with selective vulnerability of large ganglion cells. resolution objects (visual acuity) or brightness con- Together, the studies of color, scotopic perimetric trast detection. Instead, they appear to provide im- thresholds, contrast sensitivity for low spatial fre- portant signals to the blue-yellow color pathway. quencies, and fast flicker detection are consistent There is some evidence, based on relative somata with the selective and progressive loss of the larger to size, conduction latencies and receptive field sizes for smaller axons of retinal ganglion cells. Whether this macaque monkey retinal ganglion cells at compara- results from direct pressure at the optic nerve head ble retinal eccentricities, that SWS cones send signals and/or the retina, or from a pressure-related ischemia to larger axons than those processing red-green infor- at the optic nerve head producing axon death, is un- 57 mation. These observations and the more recent clear. observations by Quigley et al58 that there is a sequen- tial loss of large, medium and then small axons, when Key words: glaucoma, ocular hypertension, visual fields, glaucoma was experimentally induced in the monkey SWS cone sensitivity retina by chronically raised intraocular pressure, may help explain our results and those of earlier studies. Acknowledgments (It has been known for some time that large nerve The assistance of Ken Huie in the data analysis and Nina fibers are more susceptible than small fibers to in- Friedman in the manuscript preparation is gratefully ac- 59 knowledged. The authors also thank Stanley Ainsworth creased pressure, and this has recently been used to and Anne Collins for making the clinical visual field mea- selectively block the larger Y optic nerve fibers in the surements. cat.60) The loss of sensitivity to high frequency flicker (thought to be mediated by the largest ganglion cell References axons to the magnocellular layers) and the selective 1. Quigley HA, Addicks EM, and Green WR: Optic nerve dam- loss of sensitivity to blue light for the chromatic path- age in human glaucoma: III. Quantitative correlation of nerve ways (carried by the larger of the axons to the parvo- fiber loss and visual fielddefec t in glaucoma, ischemic neurop- cellular layers), with relative sparing of the chromatic athy, papillodema and toxic neuropathy. Arch Ophthalmol pathways for red and green light (carried by the small- 100:135, 1982. 2. Iwata K: Retinal nerve fiber layer, optic cupping and visual est axons to the parvocellular layers), is consistent field changes in glaucoma. In Glaucoma: Contemporary Inter- with the sequential loss of large to small axons in the national Concepts, Bellows JG, editor. New York, Masson, monkey retina. 1979, pp. 139-146. Studies of other vision functions in glaucoma pa- 3. Sommer A, Miller NR, Pollack I, Maumenee AE, and George T: The nerve fiber layer in the diagnosis of glaucoma. Arch tients are also consistent with the sequential loss of Ophthalmol 95:2149, 1977. the larger axons. For example, studies in which grat- 4. Anderson DR: What happens to the and retina in ing test targets were used to measure visual fields have glaucoma? Ophthalmology 90:766, 1983.

Downloaded from iovs.arvojournals.org on 09/23/2021 No. 1 SWS CONE FIELDS IN GLAUCOMA AND OCULAR HYPERTENSION / Heron er ol. 71

5. Anctil JL and Anderson DR: Early foveal involvement and 26. Adams AJ, Zisman F, Rodic R, and Cavender JC: Chromatic- generalized depression of the visual field in glaucoma. Arch ity and luminosity changes in glaucoma and diabetes. Doc Ophthalmol 102:363, 1984. Ophthalmol Proc Ser 33:413, 1982. 6. Balazsi AG, Drance SM, Schulzer M, and Douglas GR: Neuro- 27. Steinschneider T, Ticho U, and Adler D: Correlation between retinal rim area in suspected glaucoma and early chronic color vision deficiency and results of clinical examination in open-angle glaucoma. Arch Ophthalmol 102:1011, 1984. glaucomatous patients. Doc Ophthalmol Proc Ser 39:407, 7. Lewis RA and Johnson CA: Early detection of glaucomatous 1984. damage: I. Psychophysical disturbances. Surv Ophthalmology 28. Zwas F, Shin DH, and McKinnon PF: Early diagnosis of glau- 30:111-117, 1985. coma in ocular hypertensive patients. ARVO Abstracts. Invest 8. Quigley HA: Early detection of glaucomatous damage: II. Ophthalmol Vis Sci 25(Suppl):193, 1984. Changes in the appearance of the optic disc. Surv Ophthalmol 29. Drum B: Sources of short wavelength sensitivity loss in glau- 30:111 and 117, 1985. coma. Doc Ophthalmol Proc Ser 46:413, 1987. 9. Quigley HA, Hohman RM, Addicks EM, Massof RW, and 30. Adams AJ, Heron G, and Husted R: Clinical measures of Green WR: Morphological changes in the lamina cribrosa central vision function. Arch Ophthalmol 105:782, 1987. correlated with neural loss in open-angle glaucoma. Am J Oph- 31. Hamil TR, Post RB, Johnson CA, and Keltner JL: Correlation thalmol 95:673, 1983. of color vision deficits and observable changes in the optic disc 10. Quigley HA, Hohman RM, and Addicks EM: Quantitative in a population of ocular hypertensives. Arch Ophthalmol studies of optic disc capillaries in chronic glaucoma in the 102:1637, 1984. primate and human eye. ARVO Abstracts. Invest Ophthalmol 32. Airaksinen PJ, Lakowski R, Drance SM, and Price M: Color VisSci24(Suppl):102, 1983. vision and retinal nerve fiber layer in early glaucoma. Am J 11. Atkin A, Podos SM, and Bodis-Wollner I: Abnormalities of the Ophthalmol 101:208, 1986. visual system in ocular hypertension and glaucoma: Seeing 33. Adams AJ, Schefrin BE, and Huie K: New clinical color beyond routine perimetry. In Glaucoma Update II, Krieglstein threshold test for . Am J Optom Physiol Opt 64:29, GK and Leydhecker W, editors. Berlin, Springer Verlag, 1983, 1987. pp. 107-116. 34. Castano JA and Sperling HG: Sensitivity of the blue-sensitive 12. Ross JE, Bron AJ, and Clark DD: Contrast sensitivity and cones across the central retina. Vision Res 22:661, 1982. visual disability in chronic simple glaucoma. Br J Ophthalmol 35. Flammer J: Psychophysics in glaucoma, a modified concept of 68:821, 1984. the disease. Doc Ophthalmol Proc Ser 43:11, 1985. 13. Marmion VJ: A comparison of contrast sensitivity function 36. Feree CE and Rand G: Effect of brightness of preexposure and and field loss in glaucoma. Doc Ophthalmol Proc Ser 42:443, surrounding field on breadth and shape of the color fields for 1985. stimuli of different sizes. Am J Ophthalmol 7:843, 1924. 14. Tyler CW: Specific deficits and flicker sensitivity in glaucoma 37. Engelking E and Eckstein A: Physiologische Bestimmung der and ocular hypertension. Invest Ophthalmol Vis Sci 20:204, Musterfarben fiir die KJinische Perimetrie. KJin Monatsbl Au- 1981. genheilkd 64:88, 1920. 15. Francois J and Verriest G: Les dyschromatopsies acquises dans 38. DuBois-Poulsen A: Le champ visuel: topographie normale et le glaucome primaire. Ann Oculistique 192:191, 1959. pathologique de ses sensibilites. Paris, Masson and Cie, 1952. 16. Grutzner P and Schleicher S: Acquired colour vision defects in 39. Hedin A and Verriest G: Is clinical colour perimetry useful? glaucoma patients. Mod Prob Ophthalmol 11:136, 1972. Doc Ophthalmol Proc Ser 26:161, 1981. 17. Fishman GA, Krill AE, and Fishman M: Acquired colour vi- 40. Hansen E: The colour receptors studied by increment thresh- sion defects in patients with open-angle glaucoma and ocular old measurements during chromatic adaptation in the Gold- hypertension. Mod Prob Ophthalmol 13:335, 1974. mann perimeter. Acta Ophthalmologica 52:490, 1974. 18. Austin DJ: Acquired colour vision defects in patients suffering 41. Greve EL, Verduin WM, and Ledeboer M: Two-color thresh- from chronic simple glaucoma. Trans Ophthalmol Soc UK old in static perimetry. Mod Prob Ophthalmol 13:113, 1974. 94:880, 1974. 42. Vola JL, Gastaut P, and Gondois B: Considerations on Stiles pi 19. Lakowski R and Drance SM: Acquired dyschromatopsias: The mechanisms in glaucomatous Bjerrum area. Doc Ophthalmol earliest functional losses in glaucoma. Doc Ophthalmol Proc Proc Ser 42:269, 1985. Ser 19:159, 1979. 43. Hansen E and Seim T: Calibration of the Goldmann perimeter 20. Poinoosawmy D, Nagasubramanian S, and Gloster J: Colour and accessories used in specific quantitative perimetry. Acta vision in patients with chronic simple glaucoma and ocular Ophthalmologica 56:241, 1978. hypertension. Br J Ophthalmol 64:852, 1980. 44. Moreland JD, Maione M, Carta F, Barberini E, Scoccianti L, 21. Drance SM, Lakowski R, Schulzer M, and Douglas GR: Ac- and Lettieri S: The clinical assessment of the chromatic mecha- quired colour vision changes in glaucoma: Use of 100-Hue test nisms of the retinal periphery. Doc Ophthalmol Proc Ser and Pickford anomaloscope as predictors of glaucomatous 14:413, 1977. field change. Arch Ophthalmol 99:829, 1981. 45. Hart WM and Gordon MO: Colour perimetry of glaucoma- 22. Adams AJ and Rodic R: Use of desaturated and saturated tous visual field defects. Ophthalmology 91:338, 1984. versions of the D-15 test in glaucoma and glaucoma-suspect 46. Krastel H, Jaeger W, and Braun S: The contribution of spectral patients. Doc Ophthalmol Proc Ser 33:419, 1982. increment thresholds to the interpretation of colour perimetry. 23. Flammer J and Drance SM: Correlation between color vision Dev Ophthalmol 9:171, 1984. scores and quantitative perimetry in suspected glaucoma. Arch 47. Friedmann AI: A preliminary report on the use of colour filters Ophthalmol 102:38, 1984. in the Mark II visual field analyzer. In Colour Vision Defi- 24. Adams AJ, Rodic R, Husted R, and Stamper R: Spectral sensi- ciencies V, Verriest G, editor. Bristol, Adam Hilger Ltd, 1980, tivity and color discrimination changes in glaucoma and glau- pp. 221-225. coma-suspect patients. Invest Ophthalmol Vis Sci 23:516, 48. Genio C and Friedmann AI: A comparison between white light 1982. and blue light on about 70 eyes of patients with early glaucoma 25. Adams AJ: Chromatic and luminosity processing in retinal using the Mark II visual field analyzer. Doc Ophthalmol Proc disease. Am J Optom Physiol Opt 59:954, 1982. Ser 26:207, 1981.

Downloaded from iovs.arvojournals.org on 09/23/2021 72 INVESTIGATIVE OPHTHALMOLOGY G VISUAL SCIENCE / January 1988 Vol. 29

49. Logan N and Anderson DR: Detecting early glaucomatous lishment of a nerve block by pressure or cocaine. Am J Physiol visual field changes with a blue stimulus. Am J Ophthalmol 88:581, 1929. 95:432, 1983. 60. Burke W, Burne JA, and Martin PR: Selective block of Y optic 50. Abe H, Sakai R, and Yamazaki Y: The selective impairment of nerve fibers in the cat and the occurrence of inhibition in the the three color mechanisms (red-, green-, and blue-sensitive lateral geniculate nucleus. J Physiol 364:81, 1985. mechanisms) isolated by the new color campimeter in patho- 61. Regan D and Neima D: Balance between pattern and flicker logical eyes with fundus disease: II. Studies of static threshold sensitivities in the visual fieldso f ophthalmological patients. Br campimetry in early glaucoma. Acta Soc Ophthalmol Jpn J Ophthalmol 68:310, 1984. 87:950, 1983. 62. Neima D, Leblanc R, and Regan D: Visual field defects in 51. Alvarez SL and Mills KB: Spectral and flicker sensitivity in ocular hypertension and glaucoma. Arch Ophthalmol ocular hypertension and glaucoma. Research and Clinical 102:1042, 1984. Forums 7:83, 1985. 63. Kaplan E and Shapley RM: X and Y cells in the lateral genic- 52. DeValois RL, Abramov I, and Jacobs GH: Analysis of re- ulate nucleus of macaque monkeys. J Physiol 330:125, 1982. sponse patterns of LGN cells. J Opt Soc Am 56:966, 1966. 64. Markoff JI, Breton ME, Shakin E, and Franz J: Pattern rever- 53. Blakemore C and Vital-Durard F: Development of the neural sal electroretinogram in glaucoma, glaucoma suspects and basis of visual acuity in monkeys. Trans Ophthalmol Soc UK normals. ARVO Abstracts. Invest Ophthalmol Vis Sci 99:363, 1979. 24(Suppl):104, 1983. 54. Ingling C and Martinez E: The spatiochromatic signal of the r-g channel. In Colour Vision Mollon JD and Sharpe LT, 65. Wanger P and Persson HE: Pattern-reversal electroretino- editors. London, Academic Press, 1980, pp. 433-444. grams in unilateral glaucoma. Invest Ophthalmol Vis Sci 55. Gouras P: Antidromic responses of orthodromically identified 24:749, 1983. ganglion cells in monkey retina. J Physiol 204:407, 1969. 66. Atkin A, Bodis-Wollner I, Podos SM, Wolkstein M, Mylin L, 56. Eisner A and MacLeod DIA: Blue-sensitive cones do not con- and Nitzberg S: Flicker threshold and pattern VEP latency in tribute to luminance. J Opt Soc Am 70:121, 1980. ocular hypertension and glaucoma. Invest Ophthalmol Vis Sci 57. de Monasterio FM: Asymmetry of on- and off-pathways of 24:1524, 1983. blue-sensitive cones of the retina of macaques. Brain Res 67. Marx MS, Podos SM, Bodis-Wollner I, Howard-Williams JR, 166:39, 1979. Siegel MJ, Teitelbaum CS, Maclin EL, and Severin C: Flash 58. Quigley HA, Dunkelberger GR, and Sanchez RM: Chronic and pattern electroretinograms in normal and laser-induced experimental glaucoma causes selectively greater loss of larger glaucomatous primate eyes. Invest Ophthalmol Vis Sci 27:378, optic nerve fibers. ARVO Abstracts. Invest Ophthalmol Vis Sci 1986. 27(Suppl):42, 1986. 68. Drum B, Armaly MF, and Huppert W: Scotopic sensitivity loss 59. Gasser HS and Erlanger J: The role of fiber size in the estab- in glaucoma. Arch Ophthalmol 104:712, 1986.

Downloaded from iovs.arvojournals.org on 09/23/2021