Investigative & Visual Science, Vol. 31, No. 10, October 1990 Copyright © Association for Research in Vision and Ophthalmology

Retinal Point Location in the Foveal Avascular Zone

Bernard 5. Zeffren,* Raymond A. Applegare,* Arthur Bradley,f and W. A. J. van Heuven"

The site of normal fixation is often assumed to be centered in the foveal avascular zone (FAZ). This assumed anatomic relationship is used during photocoagulation therapy as an objective guide to avoid damaging critical retinal structures on or near fixation. With laser therapy being directed closer and closer to the center of the FAZ, the accuracy with which the center of the FAZ locates the retinal point of fixation becomes an important therapeutic issue. Using an optimized technique for visualizing the retinal vasculature entoptically, the authors determined the location of the retinal point of fixation with respect to the foveal area vasculature in 26 of 14 healthy subjects. In 23 eyes (12 subjects), a traditional FAZ was observed, the other three eyes (two subjects) had capillaries near or crossing the center of fixation. Of the 23 eyes with a traditional FAZ, 20 had centers of fixation located eccentric to the center but in the FAZ, (average deviation from the center of the FAZ, 66.5 ± 49.5 /xm) with the direction of deviation from the FAZ center appearing random. Consequently, when following protocols that advocate photocoagulation treatment with spot centers closer to the FAZ center than 300 nm, the center of the FAZ is a poor locator of a subject's retinal point of fixation. When using the FAZ as a reference, the resulting uncertainty in the location of the subject's retinal point of fixationincrease s the probability of significant damage to the actual point of fixation by up to 20%. Invest Ophthalmol Vis Sci 31:2099-2105, 1990

The posterior pole of the contains the anatomic In a normal observer, visual performance is maxi- fovea and . These structures are highly special- mized at the subject's point of fixation. That is, when ized to maximize . In the center of the asked to fixate a point in space, normal observers anatomic fovea, the inner retinal layers down to the move their eyes such that the point of interest is are displaced, forming a pit, the imaged on the retinal region providing the highest foveola, which contains the highest density of cones resolution.3-4 The connection between fixation and in the (147,000 cones per mm2).1 The retinal optimal resolution, together with the histologic spe- capillaries in the area of the fovea typically form con- cialization of the retina, suggests that a normal indi- centrically arranged channels ending in a capillary vidual will move the eye to place the of the loop approximately 0.5 mm across which outlines a fixation point on the foveola. Thus the retinal point capillary-free zone known as the "foveal avascular of fixation and the foveola are assumed to be coinci- zone" (FAZ).1 The lateral displacement of inner reti- dent and reasonably centered in the FAZ. nal structures, including the retinal vasculature, pre- Given current therapeutic trends to photocoagu- sumably exists to leave an unobstructed path to late closer and closer to the point of fixation using the the site of phototransduction, thereby enhancing center of the FAZ as a guiding landmark,5"9 the accu- image quality.2 racy with which the center of the FAZ locates the point of fixation becomes critical. That is, when using the FAZ as an anatomic landmark for locating the From the *Department of Ophthalmology, University of Texas retinal point of fixation, significant deviation of the Health Science Center at San Antonio, San Antonio, Texas, and point of fixation from the center of the FAZ could the fDepartment of Visual Science, School of Optometry, Indiana lead to placement of laser burns closer to, or further University, Bloomington, Indiana. Supported by NIH grant EYO8OO5 to RAA, NIH grant EY07863 from, the fovea than anticipated. As a result, burns to AB, and an unrestricted research grant to the Department of may unintentionally lead to inappropriate laser-in- Ophthalmology, University of Texas Health Science Center at San duced damage or incomplete coverage or nontreat- Antonio from the Research to Prevent Blindness, Inc. New York. ment of a (sub)retinal neovascular lesion due to its Arthur Bradley is also supported by the Indiana Institute for the assumed (as opposed to known) proximity to the reti- Study of Human Capabilities grant AFOSA #870089. nal point of fixation. Reprint requests: Raymond A. Applegate, OD, PhD, Depart- 10 ment of Ophthalmology, University of Texas Health Science We used optimization techniques for visualizing Center at San Antonio, San Antonio, TX 78284-7779. the smallest foveal capillaries entoptically to deter-

2099 2100 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / Ocrober 1990 Vol. 31 mine whether or not the retinal point used for fixa- Source S, a 1-mm pinhole, back-illuminated by tion actually lies in the center of the FAZ, and if not, light from a fiber optic passing through a blue filter whether or not the deviations are significant enough (3M, Visual Products Division, 3 M Center, St. Paul, to warrant clinical consideration. MN, filter part #47 with peak transmittance at 470 nm), was set to rotate at a speed of 3.5 hertz in a Materials and Methods circular path 2 mm from, and concentric with, the optical axis of the instrument. Light from S was first A number of techniques have been used to exam- collimated by L, and then imaged into the plane ine the FAZ and its component capillaries in living of the subject's with unit magnification by lens, humans, including ,""14 and L . An diaphragm ( A), imaged at optical two noninvasive subjective psychophysical tech- 2 1516 infinity by L , served as the field stop for the eye-ap- niques: the "blue-field entoptic effect" and the 2 paratus system. entoptic visualization of the retinal vasculature (commonly known as the Purkinje image)."1718 The A second channel provided the subject a view of blue-field entoptic effect provides the viewer an two dim point sources (PSi and PS2) optically conju- image of the white corpuscles as they flow through gate with aperture A through a beam splitter, Bi. the retinal capillaries. The capillaries remain unseen, Point source, PS,, was attached to a motorized X-Y and the actual form of the vascular tree must be con- positioning plate and served as a guidelight for mark- structed in the mind of the viewer. An optimized ing positions of interest in the . Point Purkinje image allows simultaneous evaluation of a source, PS2, served as a stationary fixation point cen- persistent, easily visible, high-contrast image of the tered on the optical axis of the apparatus. The subject foveal capillaries and the point of fixation.10 controlled the position of PSi with a joystick and could thereby locate, with respect to the point of fix- ation, any point on the capillary loop defining the Instrumentation FAZ. A variable voltage signal reflecting the location optimization for the entoptic visualiza- of moveable point source, PSl5 was sent from the X-Y tion of the macular area retinal vasculature and FAZ position plate to an X-Y plotter for hard-copy docu- was achieved by modifying a standard Maxwellian- mentation of the border of the FAZ. view system.19 We call the system, schematically il- Alignment of the subject's pupil to the optical axis lustrated in Figure 1, the Vascular Entoptoscope. of the apparatus and stabilization of the subject's

Monitor

Subject Fig. 1. Schematic repre- sentation of Maxwellian view system. S = source, a rotating 1-mm pinhole, peak wavelength 470 nm; L, = collimating lens; A = variable iris diaphragm; B! = beam splitter; L2 = Maxwellian view lens; B2 = beam splitter; VC = video X-Y ; FSM = front sur- Joystick face ; AR = align- ment ring; PSi, PS2 = point X-Y Plotter sources. A toB^Bj toPSj

FSM to B2 = B2 to Subject

S to L j = A to L2 = L2to Subject = 22cm PS is mobile

PS 2 is stationary No. 10 RETINAL FIXATION POINT LOCATION / Zeffren er ol 2101 was obtained with a dental bite-bar mounted on near their center of fixation and subsequently con- a compound vice mobile in three planes. To insure nect the dots free hand again, adding as much detail the subject was properly aligned to the apparatus, a as possible to form their final tracing. third channel provided a closed-circuit video view of the pupil entry location of the rotating beam and the Results corneal reflection of alignment ring AR (a circle of light-emitting diodes concentric with the op- All of the subjects easily saw the Purkinje image of tical axis of the apparatus). The video view of the their retinal capillaries. Ten of the 14 subjects rotating beam was obtained by deflecting some of the graphed details of the shadow of their FAZ in both beam from source S at beam splitter B2 onto a front eyes and two (due to personal time constraints), in surface mirror (FSM; optically conjugate with the only one eye. Another subject observed a traditional subject's ) and back through beam FAZ in one eye but saw capillaries running through splitter B2 into the video camera. The video view of what should have been the FAZ in the other, and one alignment ring, AR, was obtained by reflection off the subject saw. capillaries running through the fixation subject's back into the apparatus and reflec- point in both eyes. Figure 2 depicts a sample of the tion at beam splitter B2 into the video camera. By various FAZ tracings obtained. Panel A displays the continually observing the information on the moni- tracing of one of only three eyes with a retinal point tor, the experimenter could maintain subject/appara- of fixation located in the geographic center of the tus alignment by adjusting the position of the com- FAZ as classically described anatomically. Panel B pound vice. displays a tracing from an eye with the retinal point of fixation located a typical distance from the geo- Subjects graphic center of the FAZ, and Panel C displays the tracing of the subject with the largest distance (189 We assessed 26 eyes in 14 healthy subjects ranging p) between the retinal point of fixation and the in age from 11-54 yr. There were 11 male and three geographic center of the FAZ. Panel D displays the female subjects, all with corrected vision of 20/25 or tracing of one of three eyes with vessels in the retinal better. area more commonly occupied by the FAZ. Even by casual observation, it becomes clear that the FAZ Protocol boundaries are not always concentric with the fixa- The purpose and methods of the experiment were tion point (Panel B and C). explained to the subject, and informed consent was Although classically described as circular, the FAZ obtained. To familiarize the subjects with the experi- is irregular in size and shape (Fig. 2). This irregularity mental task, each was shown enlarged prints of fluo- in size and shape complicates the objective determi- rescein angiograms of several having: (1) clas- nation of the center of the FAZ. To solve this prob- sic FAZs and (2) "variant FAZs" containing capillar- lem, a geographic "center" was defined in the follow- ies crossing the fovea in various patterns.11"12 After ing manner (Fig. 3). First, the largest possible rectan- the subject was familiar with what they might see in gle having all four corners located inside the borders their own eyes, a bite bar was made, and the subject of the FAZ tracing was aligned such that the horizon- was aligned to the optical axis of the apparatus. tal dimension of the rectangle was parallel to the hori- Subjects, while carefully fixating point source PS2 zontal axis of the tracing. Second, as illustrated in used a joystick to move the guidelight (point source, Figure 3A, distances from each side of the rectangle PSi) to at least 12 locations along the borders of their (labeled h, h', v, and v') to the traced FAZ border were FAZ. At each location the subject centered the guide- measured at each of five equally spaced locations light on the capillary defining the FAZ border and along each side of the rectangle (lines labeled v,-v5, pushed a button signaling the pen to drop on the X-Y v'i-v'5, hj-h5, and h',-h'5). Third, as illustrated in Fig- plotter, thus making a dot outline of the FAZ. After ure 3B, each of the four subsets of distances were exiting the optical system, subjects were shown their averaged [eg, (h, + h2 4- h3 + h4 + h5)/5 = hx and (h', "dot pattern" and instructed to connect the dots by + h'2 + h'3 + h'4 + h'5)/5 = h'x], and these averages hand adding as much detail as possible to form their were added to the appropriate end of the horizontal final FAZ tracing. The subject then marked on the or vertical dimension of the initial rectangle (dashed FAZ tracing the location of the fixation point. We rectangle) to yield a new rectangle (solid rectangle) thus obtained a hard-copy replica of the subject's with a total horizontal dimension of H and total ver- FAZ to serve as the raw data for our future analyses. tical dimension of V (ie, H = h + hx + h'x and V = v If there were vessels crossing the "FAZ," the subject + vx + v'x)- The geographic center of the FAZ was, in was asked to mark dots along the vessels they saw turn, defined as the point GC formed by the crossing 2102 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Ocrober 1990 Vol. 31

100 \L

Fig. 2. Four FAZ trac- ings. (A) Classic FAZ trac- ing with centered fixation point; (B) FAZ tracing showing a typical eccentric location for the retinal point of fixation; (C) FAZ tracing A. CENTERED B. TYPICAL showing the largest eccen- tric distance between the geographic center of the FAZ and the retinal point of fixation; (D) Tracing show- ing no FAZ and retinal cap- illaries near the retinal point of fixation. ""Center of fixa- tion; 4- geographic center of FAZ.

C. LARGEST D. ATYPICAL

of a vertical line bisecting H and a horizontal line compensating for the optical magnification factor of bisecting V (Fig. 3b). the Maxwellian view optical system and the gain of All 23 eyes with FAZs had retinal fixation points the X-Y plotter. Figure 4 uses a polar-coordinate sys- located in the FAZ. However, only three eyes from tem to illustrate the location of the retinal point of three different subjects had their retinal points of fix- fixation relative to geographic center of the FAZ for ation located at the geographic center of the FAZ. each eye tested. Although the data as a whole tends to Vectors defining the distance from the geographic cluster near the origin (ie, the retinal point used for center of the FAZ to the subject's fixation point and fixation tended to be nearer the center of the FAZ the direction of deviation (with 0° being horizontal to compared with the edge of the FAZ), the distribution the right) were determined for each FAZ tracing. of directions of deviation appear random. The largest These distances were then converted to retinal dis- deviation of the retinal point of fixation from the tances using the Gullstrand reduced model eye with a geographic center of the FAZ was 189 /*m. The aver- nodal-point to retina distance of 16.67 mm, after age deviation from the geographic center across all No. 10 RETINAL FIXATION POINT LOCATION / Zeffren er ol 2103

A B

V5

\

V

"2 H h,

\ .__ 1

=—^ v V '2 3 4

Fig. 3. Schematic illustrating the method for determining the geographic center of the FAZ. Panel A illustrates a hypothetical FAZ containing the largest possible rectangle (sides labeled h, h', v, and V) and distances to the FAZ border measured at 5 equally spaced points along each side of the rectangle. Panel B illustrates the solid rectangle defined by adding the average distance to the FAZ border for each side (hx, hx, vx, and v'x) to h, h', v, and v* denning a new rectangle with sides of length H and V. The geographic center of the FAZ(GC) is defined as the intersection of bisectors of H and V. subjects was 66.50 Aim. There was no tendency for the Discussion eccentricity of the retinal point of fixation to increase Our data indicate that the retinal point of fixation with increasing FAZ diameter. deviates from the geographic center of the FAZ by about 65 nm (standard deviation, ± 50 jum) with a

120

210 330

300 270 Fig. 4. Location of the retinal point of fixation from the geo- Fig. 5. Location of the retinal point of fixation (solid circles) from graphic center of the FAZ tracing for each eye tested. Left eyes are the geographic center of the FAZ tracing for each eye tested. Lightly represented by open circles; right eyes are represented by closed shaded area represents the area at risk of damage from a single circles. Irregular circular pattern represents a typical FAZ border. 200-/xm or 100-/xm laser burn (darkly shaded circles) placed to Angles are shown in degrees; distances are given in microns. overlap a lesion 200-jum from the center of the FAZ by 100-jum. 2104 INVESTIGATIVE OPHTHALMOLOGY b VISUAL SCIENCE / October 1990 Vol. 31 range of 0-190 /urn. These findings suggest that laser with lesions within 375 ^m of the FAZ center be- burns centered on retinal points less than 300 ^m tween 8%—33% of the eyes successfully treated lost six from the geographic center of the FAZ run a signifi- lines or more at first follow-up depending on the par- cantly higher risk of falling directly on or nearer to ticular study.6"9 Although this could be accounted for the retinal point of fixation than intended. Further- by assuming that pathology within 375 ixm is more more, the risk of burning the point of fixation can likely to affect the retinal point of fixation, our data markedly increase as burns are placed closer to the suggest another possibility. Given the uncertainties of geographic center of the FAZ. The implications of thermal spread, dose specification, actual spot size in this finding are profound and find support in current the plane of the retina, variation in absorp- literature. tion, and accuracy of burn placement with respect to To illustrate, assume that, as in recent clinical desired location, the speculative estimate of 4%-8% trials,6"9 retinal lesions up to 200 ^m from the FAZ may indeed be an underestimate of the number of center are treated with burns which overlap the lesion retinal fixation points at risk. Simply allowing for 50 by up to 100+ jum. Furthermore, grant that our data /im of uncertainty would raise the number of retinal form a representative sample of the population for fixation points at risk from 4%-8% up to 16%-20%. the location of the retinal point of fixation with re- To the extent this analysis is correct, it suggests that spect to the geographic center of the FAZ (data therapeutic failures at first visit for eyes with retinal points, Fig. 5). With these assumptions, six of 24 eyes lesions between 200-375 jum could be reduced by as (25%) have retinal points of fixation which are po- much as 20% by using the actual retinal point of tentially vulnerable to being burned (Fig. 5, lightly fixation as a reference as opposed to the geographic shaded area). However, since photocoagulation treat- center of the FAZ. This, in turn, raises the issue of ment is generally limited to or slightly overlaps the developing a clinically accurate technique for locat- area of frank pathology (eg, neovascular membrane ing the retinal point of fixation. or histospot), it is likely that a series of burns will be In the unparalyzed eye having good visual acuity placed only in the sector of the macular area contain- the retinal point used for fixation is relatively easy to ing the site of the lesion (Fig. 5, darkly shaded area identify.21 If techniques are adopted to paralyze the schematically shows both a 200- and 100-/um burn). eye which do not compromise vision, then it would Under these criteria and if the treatment sector of the be a relatively easy to have the patient move a small macular area is limited to 90°, it is likely that for any fixation point (eg, with a joy stick), which is also one particular series of burns one or two eyes of our visible to the clinician superimposed on the patient's sample of 24 (4%-8%) would have their retinal point , until the fixation point is exactly centered of fixation adversely effected by photocoagulation over the patient's retinal point of fixation. therapy. Unfortunately, with a visual acuity loss the patient The obvious question arises as to what percentage may or may not be using their "normal" (prepatho- of treatment failures (loss of six lines of visual acuity logy) retinal point of fixation, and the procedures or more at first follow-up) can be accounted for by outlined above may or may not be appropriate. Yet, variations in the location of the point of fixation with since visual acuity has been reported to recover, it respect to the geographic center of the FAZ. Although remains important to identify the retinal point of fix- we cannot answer this question definitively with the ation used before loss of visual acuity and avoid this data collected to date, it is interesting to note that retinal location during therapy. The uncertainty can with argon treatment it has been reported that 9% of be reduced, if not eliminated, by identifying the loca- eyes treated for neovascular ,6 9% of the tion of the retinal point of fixation in at-risk eyes (eg, eyes treated for macular area ocular histoplasmosis,7 confluent drusen,22"26 loss of absolute sensitivity,25"29 and 10% of those eyes treated for macular area idio- acquired tritan errors,30"32 or loss of vision in fellow pathic neovascularization8 lost six or more lines of eye3334 in age-related ) before vision visual acuity at first follow-up despite "successful" loss and using this point as the reference should ther- treatment of the pathology. This order of magnitude apy be needed at a later date. of initial follow-up failure is not unique to argon Our data indicate the retinal point of fixation is not treatment. Studies using krypton therapy for histo- always centered in the FAZ. Furthermore, the devia- plasmosis have reported a similar percentage of pa- tions of the retinal point of fixation from the center of tients (8%) with a six-line loss in visual acuity at first the FAZ can be large enough to jeopardize the retinal follow-up.9 The argument is further fueled by the fact point of fixation during foveal area laser photocoagu- that the best predictor of visual acuity loss, despite lation therapy which avoids the center of the FAZ adequate therapy, is treatment proximity to the compared with locating and avoiding the retinal center of the FAZ.20 When analysis is limited to eyes point of fixation. This risk can be reduced by using No. 10 RETINAL FIXATION POINT LOCATION / Zeffren er ol 2105

the retinal point of fixation as the reference for foveal 16. Yap M, Gilchrist J, and Weatherill J: Psychophysical measure- area photocoagulation rather than the center of ment of the foveal avascular zone. Ophthalmolic Physiol Opt 7:405, 1987. the FAZ. 17. Bradley A, Applegate RA, Zeffren BS, and van Heuven WAJ: Key words: retinal fixation point, foveal avascular zone Psychophysical evaluation of retinal vessels. In Technical Di- (FAZ), fovea, photocoagulation, Purkinje image gest, Noninvasive Assessment of the , Vol 7. Washington, DC, Optical Society of America, 1989, pp. Acknowledgments 162-164. 18. Zeffren BS, Applegate RA, Bradley A, and van Heuven WAJ: The authors thank Sherard Dorroh for his assistance in Psychophysical evaluation of the foveal avascular zone (FAZ) data collection and Jana Harvey for her assistance in data size and foveola location. ARVO Abstracts. Invest Ophthalmol collection, analysis, word processing, and illustration prep- VisSci30(Suppl):410, 1989. aration. 19. Applegate RA and Bonds AB: Induced movement of receptor References alignment toward a new pupillary aperture. Invest Ophthalmol Vis Sci 21:869, 1981. 1. Sigelman J and Ozanics V: Retina. In Biomedical Foundations 20. Han DP, Folk JC, and Bratton AR: Visual loss after successful of Ophthalmology, Vol 1, Duane TD and Jaeger EA, editors. photocoagulation of choroidal neovascularization. Ophthal- Philadelphia, JB Lippincott, 1988, p. 7. mology 95:1380, 1988. 2. Weale RA: Why does the human retina possess a fovea? Na- 21. L'Esperance FA Jr: Ophthalmic Lasers. St. Louis, CV Mosby, ture 212:255, 1966. 1989, p. 81, 102-105. 3. Weymouth FW, Hines DC, Acres LH, Raaf JE, and Wheeler 22. Gass JDM: Drusen and disciform macular detachment and MC: Visual acuity within the area centralis and its relation to degeneration. Arch Ophthalmol 90:206, 1973. eye movements and fixation. Am J Ophthalmol 11:947, 1928. 23. Sarks SH: Drusen and their relationship to senile macular de- 4. McKee SP: The spatial requirements for fine stereoacuity. Vi- generation. Aust J Ophthalmol 8:117, 1980. sion Res 23:191, 1983. 24. Smiddy WE and Fine SL: Prognosis of patients with bilateral 5. Macular Photocoagulation Study Group: Manual of Opera- macular drusen. Ophthalmology 91:271, 1984. tions. National Technical Information Service, Springfield 25. Sunness JS, Johnson MA, Massof RW, and Marcuss: Retinal VA, accession No. PB82-221-482. sensitivity over drusen and nondrusen areas: A study using 6. Macular Photocoagulation Study Group: Argon laser photo- fundus perimetry. Arch Ophthalmol 106:1081, 1988. coagulation for neovascular maculopathy: Three year results 26. Massof RW, Choy D, Sunness JS, Johnson MA, Rubin GS, from randomized clinical trials. Arch Ophthalmol 104:694, and Fine SL: Foveal threshold elevations associated with age- 1986. related drusen. Clin Vis Sci 4:221, 1989. 7. Macular Photocoagulation Study Group: Argon laser photo- 27. Eisner A, Fleming SA, Klein ML, and Mauldin WM: Sensitivi- coagulation for ocular histoplasmosis: Results of a randomized ties in older eyes with good acuity: Eyes whose fellow eye has clinical trial. Arch Ophthalmol 101:1347, 1983. exudative AMD. Invest Ophthalmol Vis Sci 28:1832, 1987. 8. Macular Photocoagulation Study Group: Argon laser photo- 28. Sunness JS, Massof RW, Johnson MA, Bressler NM, Bressler coagulation for idiopathic neovascularization: Results of a ran- SB, and Fine SL: Diminished foveal sensitivity may predict the domized clinical trial. Arch Ophthalmol 101:1358, 1983. development of advanced age-related . 9. Macular Photocoagulation Study Group: Krypton laser pho- Ophthalmology 96:375, 1989. tocoagulation for neovascular lesions of ocular histoplasmosis: 29. Brown B, Adams AT, Coletta N, and Haegerstrom-Portnoy G: Results of a randomized clinical trial. Arch Ophthalmol Dark adaption in age related maculopathy. Ophthalmic Phys- 105:1499, 1987. iol Opt 6:81, 1986. 10. Applegate RA, Bradley A, and van Heuven WAJ: Entoptic 30. Applegate RA, Adams AJ, Cavender JC, and Zisman F: Early visualization of the retinal vasculature near fixation. Invest colorvision changes in age-related maculopathy. Applied Ophthalmol Vis Sci 31:0000, 1990. 26:1458, 1987. 11. Bird AC and Weale RA: On the retinal vasculature of the 31. Smith VC, Pokorny J, and Diddie KR: Color matching and the human fovea. Exp Eye Res 19:409, 1974. Stiles-Crawford effect in observers with early age-related macu- 12. Yeung J, Crock G, and Billson F: Macular-foveal capillaries in lar changes. J Opt Soc Am [A]5:2113, 1988. the human retina. Aust J Ophthalmol 1:17, 1973. 32. Haegerstrom-Portnoy G and Brown B: Two-color increment 13. Bresnick GH, Condit R, Syrjala S, Palta M, Groo A, and Korth thresholds in early age related maculopathy. Clin Vis Sci 4:165, K: Abnormalities of the foveal avascular zone in diabetic reti- 1989. nopathy. Arch Ophthalmol 102:1286, 1984. 33. Gregor Z, Bird AC, and Chisholm IH: Senile disciform macu- 14. Laatikainen L and Larinkari J: Capillary-free area of the fovea lar degeneration in the second eye. Br J Ophthalmol 61:141, with advancing age. Invest Ophthalmol Vis Sci 16:1154, 1977. 1977. 15. Riva CE and Petrig B: Blue field and 34. Strahlman ER, Fine SL, and Hillis A: The second eye of pa- blood velocity in the retinal capillaries. J Opt Soc Am 70:10, tients with senile macular degeneration. Arch Ophthalmol 1980. 101:1191, 1983.