Autoregulation of Human Optic Nerve Head Circulation in Response to Increased Intraocular Pressure

Exp. Eye Res. (1997) 64, 737–744

Autoregulation of Human Optic Nerve Head Circulation in Response to Increased Intraocular Pressure L U T Z E. P I L L U N A T*†, D O U G L A S R. A N D E R S O N§, R O B E R T W. K N I G H T ON, K A R E N M. J O O S‡    W I L L I A M J. F E U E R Department of Ophthalmology, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, Florida, U.S.A. (Received Seattle 5 August 1996 and accepted in revised form 20 November 1996) The following experiments were undertaken to determine if blood flow is maintained by autoregulation in the human optic nerve head when circulation is challenged by elevated intraocular pressure, and to determine if the presence or absence of autoregulation is universal. Laser Doppler flowmetry was used to determine the average velocity, the number of moving erythrocytes, and the volume of flow in the capillary bed of the optic disc. These parameters were measured in 10 subjects at spontaneous levels of intraocular pressure (IOP), and at pressures artificially elevated to 25, 35, 45 and 55 mm Hg with a scleral suction cup. Four subjects (two who showed autoregulation and two who did not) were studied on six additional occasions to determine consistency of the findings. In these same four subjects a second location on the disc was also measured on six occasions to determine if the IOP-effect on blood flow varied by location. Of the 10 subjects initially studied, seven maintained the baseline level of blood flow over the lower part of the range of elevated intraocular pressure (evidence of autoregulation), but showed a decline in flow by the time IOP reached 45 or 55 mm Hg. Two subjects showed a linear decline in blood flow beginning with the smallest increment of elevation of IOP (no autoregulation), and one showed an uninterpretable result. The two individuals who showed the linear decline and two of those who showed efficient autoregulation were remeasured, and each showed consistently the same pattern as before when restudied on six different occasions each. However, at a different location on their discs, autoregulation was manifest in all of these four individuals. When challenged by elevated IOP, the optic nerve head typically maintains a steady-blood flow over a range of IOP, but fails to maintain the same flow by the time IOP reaches 45 or 55 mm Hg. Some disc locations, at least in some individuals, do not show this autoregulation, but exhibit a decline in blood flow linearly related to IOP, even with the modest elevation of IOP. # 1997 Academic Press Limited Key words : blood flow ; optic nerve head ; Laser Doppler Flowmetry (LDF) ; autoregulation ; intraocular pressure.

1. Introduction In many organs, the blood flow increases passively with increasing arterial pressure. In some tissues, however, particularly within the kidney, brain and heart, if the arterial pressure changes, the vascular bed will adapt within 30–60 seconds to maintain the local blood flow relatively constant in spite of a changing perfusion pressure. During this ‘ autoregulation ’, relaxation of the vascular smooth muscle maintains the flow at low arterial pressures, and vasoconstriction prevents excessive flow (and excess capillary intraluminal pressure) at higher arterial pressure (Smith and Kampine, 1990). * Present address of Lutz E. Pillunat, University Eye Hospital, University of Hamburg School of Medicine, Martinistr. 52, 20246 Hamburg, Germany. ‡ Present address of Karen M. Joos, Department of Ophthalmology, Vanderbilt University School of Medicine, Nashville, TN, U.S.A. § Douglas R. Anderson, M.D., Bascom Palmer Eye Institute, University of Miami School of Medicine, P.O. Box 016880 (900 NW 17 Street), Miami, FL 33101, U.S.A. Reprint requests to : † Lutz E. Pillunat, University Eye Hospital, University of Hamburg School of Medicine, Martinistr. 52, 20246 Hamburg, Germany.

0014–4835}97}050737­08 $25.00}0}ey960263

Autoregulation of the optic nerve circulation has been shown in animals (Geijer and Bill, 1979 ; Sossi and Anderson, 1983a ; 1983b ; Weinstein et al., 1982 ; Weinstein et al., 1983). In humans, autoregulation of blood flow in the optic nerve head has not been studied extensively, although evidence of a regulatory response after a challenge of artificially elevated intraocular pressure has been found with respect to erythrocyte velocity studied by laser Doppler velocimetry (Riva, Grunwald and Sinclair, 1982), ‘ redness of the papilla ’ (perhaps representing volume of blood in the capillary bed) studied by photopapillometry (Robert, Steiner and Hendrickson, 1989), and a stable VER-amplitude despite an increase of intraocular pressure (Pillunat et al., 1985 ; Pillunat, Stodtmeister and Wilmanns, 1987). These studies suggest that autoregulation of blood flow is likely present in the human optic nerve vasculature, as it is in the more thoroughly documented regulation in the retinal vasculature of humans (Grunwald et al., 1984 ; Riva, Grunwald and Petrig, 1986 ; 1990). The existence of effective autoregulation in human optic nerve circulation, the range over which regulation is possible, and possible variation in the # 1997 Academic Press Limited

738

autoregulatory capacity are important to document in more detail, because a deficient autoregulatory response of optic nerve blood flow to increased intraocular pressures in some parts of the disc or in some individuals may be a potential pathogenic factor leading to glaucomatous optic nerve damage (Ernest, 1975 ; Sossi and Anderson, 1983a). 2. Materials and Methods Subjects The study was performed on ten volunteer healthy subjects (7 men, 3 women, mean age 44 years, ranging from 23–56 years) without evidence of ocular disease, except that one subject (No. 9) was known to have a variable, sometimes abnormal, intraocular pressure (18–26 mm Hg) without any signs of glaucomatous optic nerve damage or visual field loss. None were using any vasoactive medication. In keeping with the tenets of the Declaration of Helsinki, written documentation of informed consent was obtained after the nature of the study was explained to the subjects, and the research was approved by the University of Miami Institutional Review Board (Medical Sciences Subcommittee for the Protection of Human Subjects). Laser-Doppler-Flowmetry (LDF) Optic nerve head blood flow was measured by Laser Doppler Flowmetry (LDF) (Bonner and Nossal, 1990 ; Riva et al., 1992 ; Petrig, Riva and Hayreh, 1992 ; Petrig and Riva, 1994) in one eye of the 10 volunteer subjects. In LDF, a laser beam illuminates a small volume of tissue and some of the light scattered by the tissue and by the red blood cells is detected by a photodetector. Relative tissue blood flow is obtained by electronic processing of the photocurrent, converted to a power spectrum by means of Fourier transform. Laser Doppler Flowmetry is based on the Doppler effect : laser light scattered by a moving particle is shifted in frequency by an amount dependent on the velocity of the moving particle and the angle of incidence. In a capillary bed within a translucent tissue, the incident light strikes the moving red blood cells at random angles. The result is a broadening of the scattered spectrum rather than a shift (which would occur if the laser beam had encountered a column of blood at a fixed angle with all particles moving at the same speed). The broadened spectrum of light is analysed to obtain an average velocity. The relative volume of blood in the tissue region is obtained from the proportion of light with shifted frequency (scattered by moving particles) compared to that with unshifted frequency (scattered by non-moving tissue elements). From these, the relative blood flow is calculated as the product of velocity and volume (assuming that any volume change represents change in vessel calibre with no change in the pathlength for movement of blood through the sampled volume of

L. E. P I L L U N A T E T A L.

tissue). The shape and size of the volume of tissue within which blood flow is measured depends on the scattering properties of the tissue and the positioning of the detector probe relative to the illuminating beam. The exact region sampled is not known, but is reasonably assumed to be some volume of disc tissue below the illuminated spot, as we routinely positioned the detector probe coincident with the illuminated spot on the disc surface. For our studies of the optic nerve head, we used a commercially available unit (LDF 4000, Oculix, Inc., Philadelphia, Pennsylvania, U.S.A. (Riva et al., 1992 ; Petrig and Riva, 1994). A 670 nm laser probe built into a Topcon fundus camera projects a beam approximately 160 µm in diameter on the optic nerve head, and the spectral shift of the reflected light is analysed. As described above, the analysis yields information about the relative volume of blood (number of moving red blood cells) in a tissue, the average speed (velocity) of the red blood cells, and the relative blood flow through the tissue. The reproducibility of the method and components of variability of this instrument has been documented (Joos et al., 1994). At the beginning of each experimental session, the heart rate and brachial arterial pressure were measured. Diastolic and systolic ophthalmic artery pressures were estimated by taking two third of the measured brachial pressure values (Weigelin and Lobstein, 1963). The average (³..) brachial systolic blood pressure was 114(³11) mm Hg and diastolic blood pressure was 70(³8) mm Hg. Stepwise Increase of Intraocular Pressure Optic nerve photographs of every individual were taken. A location was selected for measurement to be as remote as possible from visible vessels, which in all cases turned out to be on the temporal neuroretinal rim, and the location of the measurements marked on the photograph (Fig. 1). These marks were used to achieve measurements of the same spot when repeat measurements were taken. The spontaneous intraocular pressure was measured on each subject before each session with a Goldmann applanation tonometer. As known by history, one subject (No. 9) showed variable, sometimes elevated baseline intraocular pressures but all others had spontaneous baseline intraocular pressures in the normal range measured on each occasion. The average (³..) baseline IOP of the subjects was 14±4 (³3±38) mm Hg. LDF measurements were made at spontaneous IOP, and then at an induced IOP of 25 mm Hg, 35 mm Hg, 45 mm Hg, and 55 mm Hg in rapid sequence (2 minutes between increments of elevation of IOP). The instrument obtained spectra 20 times per second, each of which underwent computational analysis. The results were saved in digital format and reviewed from

AUTOREGULATION OF OPTIC NERVE BLOOD FLOW

739

F. 1. Optic disc of subject No. 4 with location of the measurement on the nasal neuroretinal rim indicated with a black circle. The illuminating laser spot is of this approximate size, but with an indistinct border due to light scattering properties of the disc tissue. 10

Flow

8 6 4 2

0

20 40 60 80 Intraocular pressure (mm Hg)

100

F. 2. Flow (expressed in the arbitrary units of the LDF apparatus) as a function of intraocular pressure, average of 10 subjects, with standard errors of the mean indicated. The two curves show flow measured immediately after pressure elevation (D) and 60 seconds later (E). The widened bar along the x-axis represents the range between the average diastolic and average systolic ophthalmic artery pressure (OAP) estimated from the brachial artery pressure.

a graphical display. A ‘ measurement ’ in our study consisted of the averaged first 5-second segment that contained neither blink artifacts (voltage spikes twice the amplitude of the base oscillations), nor ocular movements that were detected by the examiner. At all pressures, the measurements were taken at the same spot of the optic disc of the individual subject. After each IOP increase, measurements were made

immediately, and again 60 sec later. The IOP was raised by a 11 mm diameter, standardized suction cup placed on the temporal sclera with the anterior edge at least 1 mm from the limbus. The conversion of negative pressure (suction) in the cup into IOPincrease was derived from conversion curves (Stodtmeister et al., 1989). On a separate occasion, we confirmed in each subject individual measurements of IOP with a Goldmann tonometer at different negative pressures in the suction cup, waiting the same time between incremental IOP elevations as when the flowmetry measurements were made. We thus ensured the accuracy of the conversion nomogram and the absence of a sizeable ‘ tonographic ’ effect during the 8 minutes of incrementally increased IOP. Repeatability and Influence of Spot Location In four subjects (two selected because they showed autoregulation in the initial experiment, and two selected because they did not), the relative ONH blood flow during a stepwise increase in IOP was measured an additional twelve times on twelve different days. For six of the twelve measurements, the relative ONH blood flow was measured at the same location on the optic disc as before. On the other six occasions, a second location was chosen, this time in a location selected to be in a distant sector from the initial measurement, hence in all cases on the nasal neuroretinal rim.

740

L. E. P I L L U N A T E T A L. 1

6

2

7

3

8

4

9

5

10

120

80

40

0 120

80

40

0

Relative flow (%)

120

80

40

0 120

80

40

0 120

80

40

0

20

40

60

80

0

20

40

60

80

100

Intraocular pressure (mm Hg)

F. 3. Flow as a percent of baseline versus IOP in a single session of all ten subjects. The two curves for each subject show flow measured immediately after pressure elevation (D) and 60 seconds later (E). The widened bar along the x-axis represents the range between the diastolic and systolic ophthalmic artery pressure estimated from the brachial artery pressure. Number indicates subject.

3. Results Stepwise Increase of Intraocular Pressure The baseline flow was 7±6³3±4 (mean³standard deviation) arbitrary units. As shown in Fig. 2, which

gives the average of all 10 subjects, the blood flow was maintained very well as IOP became elevated until the IOP was elevated to the range of 45–55 mm Hg. However, this regulation was not evident in all individuals. As shown in Fig. 3, seven out of ten subjects (designated as subjects 1 through 7) showed

AUTOREGULATION OF OPTIC NERVE BLOOD FLOW

Repeatability and Spot Location When we repeated the measurements in four subjects six times each, the initial pattern (reflecting either the presence or absence of autoregulation) was repeatable for each individual. Specifically, subjects No. 4 and No. 6 (male age 56 years and female age 29 years, respectively) showed in all recordings little or no change in relative optic nerve blood flow through the

Average flow (au)

12 10 8 6 4 2 0

Average volume (au)

0.5 0.4 0.3 0.2 0.1 0.0 0.5 Average velocity (au)

a clear decline in ONH blood flow compared to baseline only at the highest IOP level tested. The 10 mm Hg interval between the steps in IOP elevation is large enough that it is impossible to determine the exact IOP at which there began a rapid decline in blood flow ; but the shape of the curves suggests that at some point after exceeding an IOP of 40 or 50 mm Hg, when the IOP approached the estimated diastolic ophthalmic artery pressure (OAP), which means that the perfusion pressure (mean OAP minus IOP) approached zero mm Hg, the relative ONH blood flow began to drop linearly toward zero. Zero flow was not actually reached because in this experiment the IOP was not raised above 55 mm Hg. With each incremental rise in IOP, the measurements taken immediately after raising IOP showed no consistent difference from a measurement taken 60 seconds later (open versus closed symptoms in Figs 2 and 3), with the immediate measurement sometimes higher and sometimes lower. Two out of the ten individuals (No. 8 and No. 9, the latter of whom is the subject who had intermittent ocular hypertension) did not maintain a constant or nearly constant relative optic nerve blood flow as IOP was elevated. In these subjects, the relative ONH blood flow dropped immediately after the first incremental increase in intraocular pressure, and continued to decline more-or-less linearly with further increase of IOP. The linear decline was again such that if it continued, flow would have reached zero when the IOP was within the range of the estimated ophthalmic artery pressure. The occurrence of this immediate drop in flow could not be related to age or gender of the subject, not did it correlate with any of the obtained baseline values (velocity, volume, flow, IOP, blood pressure) compared to the other subjects of the group examined. In one subject (No. 10) there was found no recognizable drop in ONH blood flow under IOP load, perhaps because the IOP had not been raised to his limit, perhaps representing a random falsely high measurement at the highest IOP because of variability of measurement, or perhaps because of technical variation, for example undetected malposition of the suction cup. Unfortunately, we were unable to retest this subject. As we did not actually observe any technical fault to justify discarding the data of this subject, it was retained in the analysis. The observations on this subject neither add to nor detract from the conclusions to be drawn from the rest of the data.

741

0.4 0.3 0.2 0.1 0

10

20 30 40 50 60 70 Intraocular pressure (mm Hg)

80

90

F. 4. Average of six sessions, showing flow, velocity and volume (mean in bold line³.. in fine line) versus IOP, temporal neuroretinal rim of subject No. 6. Regulation of flow (with increased volume in the company of reduced velocity) over lower range of IOP is illustrated.

range of IOP until higher intraocular pressures were reached. In contrast, subjects No. 8 and No. 9 (males, ages 35 and 51 years) showed an approximately linear drop of relative optic nerve blood flow with increasing IOP, starting with the first incremental rise in IOP. An example of each pattern with velocity, volume, and flow parameters is shown in Figs 4 and 5. To test the hypothesis that the two groups (subjects 4 and 6 vs. 8 and 9) behaved differently (specifically that the blood flow decline is linear in the nonregulated locations, but non-linear in the regulated locations), the data were subjected to a stepwise least square regression analysis of blood flow dependent on IOP, including both a linear and a quadratic term. As shown in Table I, the two non-regulated spots fit the linear term better. In all other cases the fit was more dependent on the second order term. The statistically significant coefficients of the second order term had negative values, which shows the blood flow was declining more rapidly with the higher IOP levels. Thus, blood flow is significantly less affected by IOP through the low IOP range than at the higher levels of IOP, after the capacity for regulation is exceeded.

742

L. E. P I L L U N A T E T A L.

measured at a second location, on the nasal neuroretinal rim, well away from the initial measurement site. This location showed values at baseline IOP comparable to the temporal side in each individual. All four subjects, each measured six times, showed constant relative blood flow at least until an IOP of 35 mm Hg was reached at the new site. Thus this newly selected location in subjects No. 8 and No. 9 did not show the absence regulation of blood flow had been shown at the initial site.

Average flow (au)

10 8 6 4 2 0

Average volume (au)

0.4

4. Discussion

0.3 0.2 0.1 0

Average velocity (au)

0.4 0.3 0.2 0.1

0

10

20 30 40 50 60 70 Intraocular pressure (mm Hg)

80

90

F. 5. Average of six sessions, showing flow, velocity, and volume (mean in bold line³.. in fine line) versus IOP, temporal neuroretinal rim of subject No. 8. Absence of regulation with increasing IOP is illustrated.

T I Statistical significance (P-values) of contributions of the linear and quadratic terms for each of two spots in four individuals in multiple regression analysis Subject (No. 4) (No. 6) (No. 8) (No. 9)

Spot 1 2 1 2 1 2 1 2

(temp) (nas) (temp) (nas) (temp) (nas) (temp) (nas)

Fit to IOP

Fit to IOP#

0±037 0±5 0±2 0±3 0±001 0±5 0±001 0±2

0±0054 0±001 0±001 0±013 0±9 0±001 0±3 0±004

The average of multiple readings also served to reveal, as illustrated in Fig. 4 for subject No. 6, that the constancy of the optic nerve blood flow was mainly due to a constant or slightly increasing volume signal upon elevation of IOP, as the relative velocity signal shows a descending slope. For each individual who went through the repeated measurements, the optic nerve blood flow was also

The results show in these healthy subjects that the optic nerve head (ONH) blood flow is typically kept nearly constant despite substantial elevation of the intraocular pressure. The well maintained flow with initial increments of IOP elevation represents autoregulation of blood flow, which keeps the blood flow nearly constant despite the challenge of a reduced vascular perfusion pressure (the result of a higher venous pressure caused by elevating the IOP). A subsequent linear decline toward the estimated mean ophthalmic artery pressure would be the expected pattern if the vascular bed reaches full dilation ; because with no further biological reactivity, the flow becomes passively dependent simply on the perfusion pressure. The point at which ONH blood flow declines most prominently is after the IOP is above 45 mm Hg, where the IOP approaches the estimated diastolic ophthalmic artery pressure in our subjects. Because the subjects all had blood pressure in a relatively narrow normal range (neither hypotension nor hypertension), our data do not indicate definitively if the break point relates simply to the IOP, or if it is governed by the perfusion pressure (blood pressure minus IOP) and is therefore influenced by the blood pressure. The autoregulation seen most of the time in human optic discs seems to be absent at some disc locations of some individuals. These disc locations seem unable to maintain the full level of blood flow under even a minimal IOP challenge. It is apparent that ultimately to understand the pathogenic mechanisms of glaucomatous optic nerve damage, the existence of localized and individual differences in vascular physiology must be taken into account. For now, we can’t yet conclude if all individuals may have a region without autoregulation, or if regional absence of regulation has any pathophysiologic implications for glaucoma. It is also not possible to determine if any localized or individual differences relate dichotomously to the presence or absence of regulation, or if the differences represent a continuous distribution with respect to the position of the break point, with the break point below the spontaneous IOP in subjects No. 8 and No. 9 of our study. The basis for this seeming occasional absence of regulation is likewise not yet

AUTOREGULATION OF OPTIC NERVE BLOOD FLOW

743

known. Such details about the nature, basis, and individual or regional variation of vascular regulation and physiology (and pathophysiology) in the optic nerve head will require further studies for full elucidation. For now, we may conclude only that there is evidence that regulation of blood flow in the optic nerve head is common, but it seems is sometimes locally absent. We are confident that these results are not due to astigmatism variably induced by the suction cap, on the basis of previous experience (Pillunat et al., 1985). Moreover, in the same individual, one position of the disc showed a curve suggesting autoregulation and another position, which would have had the same induced refractive change, did not. The laser beam projected onto the disc was not observed to be distorted during the experiment. Finally, in a separate set of experiments, keratometry did not reveal meaningful astigmatic changes from the same suction cup device used in the same manner (Trible, personal communication). When considering the implications of these results, it is important to keep in mind that the depth in the disc to which the measuring beam penetrates before it returns to the photodetector is still under investigation. The human optic nerve head is divided into three layers : the superficial nerve fiber layer, nourished by the retinal circulation, the prelaminar layer and the laminar layer. The latter two are supplied by branches of the short posterior ciliary arteries, and pathophysiology of the circulation in this vascular bed is generally considered more relevant to the mechanism of glaucomatous optic nerve damage than circulation in the superficial layers (Hayreh, 1970 ; 1985 ; 1993a ; 1993b). If the scattering coefficient is such that a high proportion of the light penetrates to the lamina cribrosa and is reflected back through the tissue (Delori and Pflibsen, 1989), the measurement should include the full thickness of the optic nerve head. Koelle et al. (1993) performed LDF on a simulated blood vessel in a model eye through optic nerve sections in order to ascertain the ability to detect flow through them. It was found that relative blood flow was detected by LDF through tissue sections of 1000 µm thickness. Those results suggest that the ONH blood flow measured by LDF should include not only the capillary flow in the superficial nerve fiber layer, but also in the deeper capillary bed of the prelaminar and laminar layer of the optic nerve. In an effort to address this question in another way, Petrig et al. (1992) measured ONH blood flow by LDF in monkeys before and during total central retinal artery occlusion (CRA). A total CRA occlusion resulted in a highly significant decrease in relative ONH blood flow (®32³11 %). This finding implies, as might be expected, that the superficial layers supplied by the CRA contribute heavily to the measurement. However, the flow parameter did not fall to zero, which suggests

that the sampling volume goes deeper than the superficial nerve fiber layer of ONH tissue supplied by the CRA, unless the observations are explained by a compensatory perfusion of the superficial retinal microvasculature by cilio-retinal anastomoses under this experimental condition. Taken together, these experiments and the knowledge that light penetrates the translucent optic disc tissue easily make us believe that the results of the present study represent an autoregulation pattern of ONH circulation of the deeper optic disc tissue in addition to the superficial layer supplied by the retinal vasculature. Not only the depth, but also the lateral extent over which rays may travel and return to the detector, needs clarification. The light-scattering properties of the disc (demonstrated simply and directly as the entire disc glows when one edge is illuminated) suggest that certain rays may return to the detector after traversing many regions of the disc, so that blood flow of the entire disc is potentially represented in the measurement. However, the proportionate representation of distant locations may be quite small. The longer the path the more likely the light will be scattered in a direction away from the detector, so that remote areas are proportionately less represented than regions on a direct path along which a light ray has only a single backscattering encounter with either a moving red cell or non-moving tissue component. That distinctly different results were obtained from study of two locations in the same disc confirms the expectation that the region measured is predominantly that which underlies the illuminated spot.

Acknowledgements Supported in part by a grant from The Glaucoma Research Foundation, San Francisco, California ; in part by the National Research Service Award T32 EY-07021 and by Clinical Vision Research Development Award R21 EY10900, awarded by the National Eye Institute, Bethesda, Maryland ; in part by Humphrey Instruments, Inc., San Leandro, California, a company of the Carl Zeiss Group ; and in part by a grant of the Deutsche Forschungsgemeinschaft Pi 181}3-1, Bonn, Germany. Each author states that s}he has no proprietary interest in the development or marketing of any instrumentation or method used in this study or competing product.

References Bonner, R. F. and Nossal, R. (1990). Principles of laserDoppler flowmetry. In Laser-Doppler blood flowmetry (Eds Shepherd, A. P. and O> berg, P. AI .). Pp. 17–46. Kluwer Academic : Boston. Delori, F. C. and Pflibsen, K. P. (1989). Reflectance properties of the optic disc. In Noninvasive assessment of the visual system ; 1989 Technical Digest Series, volume 7. Pp. 154–7. Optical Society of America : Washington, D.C. Ernest, J. T. (1975). Pathogenesis of glaucomatous optic nerve disease. Trans. Am. Ophthalmol. Soc. 73, 366–88. Geijer, C. and Bill, A. (1979). Effects of raised intraocular

744

pressure on retinal, prelaminar, laminar and retrolaminar optic nerve blood flow in monkeys. Invest. Ophthalmol. Vis. Sci. 18, 1030–42. Grunwald, J. E., Riva, C. E., Stone, R. A., Keates, E. U. and Petrig, B. L. (1984). Retinal autoregulation in openangle glaucoma. Ophthalmology 91, 1690–4. Hayreh, S. S. (1970). Pathogenesis of visual field defects ; role of the ciliary circulation. Br. J. Ophthalmol. 54, 289–311. Hayreh, S. S. (1985). Inter-individual variation in blood supply of the optic nerve head. Doc. Ophthalmol. 59, 217–46. Hayreh, S. S. (1993a). Anatomie und Blutversorgung des Sehnerven. In Das Glaukom – Aspekte aus der Forschung fuX r die Praxis (Eds. Pillunat, L. E. and Stodtmeister, R.). Pp. 1–20. Springer : Berlin. Hayreh, S. S. (1993b). Die Bedeutung vaskula$ rer Faktoren in der Pathogenese des Glaukomschadens. In Das Glaukom – Aspekte aus der Forschung fuX r die Praxis (Eds. Pillunat, L. E. and Stodtmeister, R.). Pp. 34–58. Springer : Berlin. Joos, K. M., Pillunat, L. E., Knighton, R. W., Anderson, D. R. and Feuer, W. J. (1994). Reproducibility of laser Doppler flowmetry (LDF) measurements of the human optic nerve. ARVO Abstract n. 1870. Invest. Ophthalmol. Vis. Sci. 35 (suppl.), 1659. Koelle, J. S., Riva, C. E., Petrig, B. L. and Cranstoun, S. D. (1993). Depth of tissue sampling in the optic nerve head using laser Doppler flowmetry. Lasers Med. Sci. 8, 49–54. Petrig, B. L. and Riva, C. E. (1994). New continuous real time analysis system for laser Doppler flowmetry and velocimetry in the ocular fundus using a digital signal processor. In Noninvasive assessment of the visual system ; 1994 Technical Digest Series, volume 3. Pp. 238–241. Optical Society of America : Washington, D.C. Petrig, B. L., Riva, C. E. and Hayreh, S. S. (1992). Optic nerve blood flow in the rhesus monkey measured by laser Doppler flowmetry. In Noninvasive assessment of the visual system ; 1992 Technical Digest Series, volume 1. Pp. 114–7. Optical Society of America : Washington, D.C. Pillunat, L. E., Stodtmeister, R. and Wilmanns, I. (1987). Pressure compliance of the optic nerve head in low tension glaucoma. Br. J. Ophthalmol. 71, 181–7. Pillunat, L. E., Stodtmeister, R., Wilmanns, I. and Christ, T.

L. E. P I L L U N A T E T A L.

(1985). Autoregulation of ocular blood flow during changes in intraocular pressure ; preliminary results. Graefes Arch. Clin. Exp. Ophthalmol. 223, 219–23. Riva, C. E., Grunwald, J. E. and Petrig, B. L. (1986). Autoregulation of human retinal blood flow ; an investigation with laser Doppler velocimetry. Invest. Ophthalmol. Vis. Sci. 27, 1706–12. Riva, C. E., Grunwald, J. E. and Sinclair, S. H. (1982). Laser Doppler measurement of relative blood velocity in the human optic nerve head. Invest. Ophthalmol. Vis. Sci. 22, 241–8. Riva, C. E., Harino, S., Petrig, B. L. and Shonat, R. D. (1992). Laser Doppler flowmetry in the optic nerve. Exp. Eye Res. 55, 499–506. Riva, C. E., Petrig, B. L. and Grunwald, J. E. (1990). Retinal blood flow. In Laser-Doppler blood flowmetry (Eds Shepherd, A. P. and O> berg, P. AI .). Pp. 349–384. Kluwer Academic : Boston. Robert, Y., Steiner, D. and Hendrickson, P. (1989). Papillary circulation dynamics in glaucoma. Graefes Arch. Clin. Exp. Ophthalmol. 227, 436–9. Smith, J. J. and Kampine, J. P. (1990). Circulatory physiology ; the essentials, 3rd ed. Pp. 140–60. Williams & Wilkins : Baltimore. Sossi, N. and Anderson, D. R. (1983a). Blockage of axonal transport in optic nerve induced by elevation of intraocular pressure ; effect of arterial hypertension induced by angiotensin I. Arch. Ophthalmol. 101, 94–7. Sossi, N. and Anderson, D. R. (1983b). Effect of elevated intraocular pressure on blood flow ; occurrence in cat optic nerve head studied with iodoantipyrine I 125. Arch. Ophthalmol. 101, 98–101. Stodtmeister, R., Pillunat, L., Mattern, A. and Polly, E. (1989). Die Beziehung zwischen negativer Druckdifferenz und ku$ nstlich erho$ htem Augeninnendruck bei der Saugnapfmethode. Klin. Monatsbl. Augenheilkd. 194, 178–83. Weigelin E, Lobstein, A. (1963). Ophthalmodynamometry. Basel-New York, S. Karger, 1963. Weinstein, J. M., Duckrow, R. B., Beard, D. and Brennan, R. W., (1983). Regional optic nerve blood flow and its autoregulation. Invest. Ophthalmol. Vis. Sci. 24, 1559– 65. Weinstein, J. M., Funsch, D., Page, R. B. and Brennan, R. W. (1982). Optic nerve blood flow and its regulation. Invest. Ophthalmol. Vis. Sci. 23, 640–5.