Effects of Topical J8-Blockers on the Diameter of the Isolated Porcine Short Posterior Ciliary Artery

Effects of Topical j8-Blockers on the Diameter of the Isolated Porcine Short Posterior Ciliary Artery

Robert Braakman,1 Peter van der Linden, and Pieter Sipkema2

PURPOSE. Based on diameter measurements on the short posterior ciliary artery, this study was intended to determine the direct pharmacologic effect of /3-blockers; to determine the differences among a selective /3-blocker betaxolol, a j3-blocker with intrinsic sympathetic activity befunolol, and a nonselective /3-blocker timolol; and to find experimental evidence for the indirect hemodynamic effect of /3-blockers. METHODS. A segment of isolated porcine short posterior ciliary artery was cannulated at both ends and mounted in a pressurized vessel chamber. Vessel diameter was measured as a function of /3-blocker concentration and as a function of change in transmural pressure.

RESULTS. In the absence of flow, the mean effective doses (ED50) were 0.8 ± 0.3 mM, 1.0 ± 0.3 mM, and 11.6 ± 6.6 mM (SEM) for betaxolol, befunolol, and timolol, respectively. In the presence of flow, vessel diameter increased with an increase of transmural pressure. The mean relative diameter increased 4.2% ± 1.0% (SEM) at a transmural pressure step from 30 mm Hg to 60 mm Hg. This increase was not significantly dependent on the presence of any of the /3-blockers. CONCLUSIONS. Only at concentrations far exceeding their expected plasma concentrations, betaxolol, befunolol, and timolol increased the diameter of the isolated porcine short posterior ciliary artery, as a result of their direct pharmacologic effect. Only the difference between the vasodilatory potency of the selective and the nonselective /3-blocker was significant: ED50 of betaxolol was 15 times smaller than ED50 of timolol. There was a positive correlation between the diameter of the isolated porcine short posterior artery (when used as a model for an intraocular artery) and the transmural pressure, which corroborates the indirect hemodynamic effect of /3-blockers. It is speculated that instillation of topical /3-blockers into the conjunctival sac may increase the perfusion of the optic nerve head by an indirect hemodynamic mechanism, but not by a direct pharmacologic mechanism. (Invest Ophthalmol Vis Sci. 1999;40:370-377)

he pathogenesis of primary open angle glaucoma is still the transmural pressure (intravascular minus intraocular pres- unclear. There is increasing evidence, however, that not sure) and the diameter of an intraocular vessel. It has recently only the intraocular pressure but also vascular factors been found that /3-blockers may also affect ocular blood flow T 1 play a role in the glaucomatous process. Many stimuli that (velocity) by a direct mechanism. This direct mechanism in- change the vessel diameter affect the flow through the vessel, volves their pharmacologic action on the smooth muscle of the and a small change in diameter results in a large change in flow vessel wall. The effect of various types of /3-blockers on ocular (Poiseuille's Law). blood flow (velocity), however, is equivocal.5 Some cause Topically applied /3-blockers are the basis of medical anti- vasoconstriction,6'7 whereas others cause vasodilatation.4'8"10 glaucoma therapy. Although their pharmacologic action at the It has been shown that in primary open angle glaucoma, cellular level is only partly understood, it is known that all and especially in normal-tension glaucoma, the blood supply to 2 4 |3-blockers lower intraocular pressure " decreasing the me- the optic nerve head is compromised.11 Furthermore, it has chanical load on the retina. By lowering the intraocular pres- been shown that the perfusion of the optic nerve head is sure, /3-blockers may affect ocular blood flow by an indirect inversely proportional to the intraocular pressure.1112 Bill13 mechanism that involves a hemodynamic relationship between showed that, in the stop-flow situation, intraocular pressure changes can be recorded as intravascular pressure changes in

2 the short posterior ciliary artery. The vascular supply to the From the 'Spaarne Ziekenhuis Haarlem; and the Physiological optic nerve head is largely provided by the short posterior Laboratory Free University, Amsterdam, The Netherlands. 114 Supported in part by Bournonville Pharma BV, The Netherlands; ciliary arteries. Schematically, these arteries form an ex- and Merck, Sharpe and Dohme BV, The Netherlands. The j3-blockers traocular tract in series with an intraocular tract. The question used in this study, befunolol (Glauconex), timolol (Timoptol), and is whether /3-blockers affect the perfusion of the optic nerve betaxolol (Betoptic) were provided free of charge by Bournonville head by their intraocular pressure-lowering hemodynamic ef- Pharma BV, The Netherlands, Merck, Sharpe and Dohme BV, The Netherlands, and Alcon BV, The Netherlands, respectively. fect on the intraocular tract, or by their vessel wall-relaxing Submitted for publication July 9, 1998; accepted September 21, pharmacologic effect on the entire tract, or by both. 1998. To gain insight in the possible mechanisms by which Proprietary interest category: N. /3-blockers may affect the perfusion of the optic nerve head, we Reprint requests: Pieter Sipkema, Physiological Laboratory, Free University, Van der Boechorststraat 7, 1081 BT Amsterdam, The Neth- performed an in vitro study, using a segment of extraocular erlands. porcine short posterior ciliary artery placed in a vessel cham-

Downloaded from iovs.arvojournals.org on 10/01/2021 JOVS, February 1999, Vol. 40, No. 2 Hemodynamic and Pharmacologic Effects of /J-Blockers 371

ber. We studied the direct pharmacologic effect of three types PERSONAL COMPUTER VIDEO MONITOR of |3-blocker on the diameter of this extraocular tract of the short posterior ciliary artery in the open vessel chamber. To study the indirect hemodynamic effect of j3-blockers on the intraocular tract of the short posterior ciliary artery we used OUT the same extraocular tract as a model for the intraocular tract in the closed vessel chamber so that extravascular pressure could be applied, as though it were the intraocular pressure, Based on diameter measurements of the short posterior ciliary artery we sought to determine the direct pharmacologic effect of j3-blockers; to determine the differences among a selective j3-blocker betaxolol, a j3-blocker with intrinsic sympathetic activity befunolol, and a nonselective /3-bIocker timolol; and to find experimental evidence for the indirect hemodynamic effect of j8-blockers.

METHODS FIGURE 1. Experimental setup. Vessel diameter was measured at the Preparation and Experimental Setup adjustable position of the horizontal line in the video monitor. The Porcine eyes were obtained from a local abattoir. Directly after superfusion outflow column is not shown. removal from the donor the eyes were placed in an MOPS buffer (Merck, Darmstadt, Germany), consisting of (in millimolar) 145 NaCl, 5.0 KC1, 2.0 CaCl2, 1.0 MgSO4, 1.0 NaH2PO4, 5.0 dextrose, 2.0 pyruvate, 0.02 EDTA, and 3.0 3-(iV-morpholino) Gilson, Villers-Le-Bel, France) was used to superfuse the prep- propanesulfonic acid (pH 7.4, 4°C). Experiments were per- aration with Krebs buffer. The pump rate was set to 5 ml/min, formed within 24 hours. thus replacing the fluid volume of the vessel chamber every For the experiments a ±2-mm segment of the short pos- minute. The outlet of the vessel chamber was connected to a terior ciliary artery was carefully dissected from the eye and second overflow column (not shown). Thus, the external pres- transferred to a vessel chamber filled with Krebs buffer sure to the vessel could be set, simulating intraocular pressure. (Merck) at a constant temperature of 37°C and sealed with a glass cover. The Krebs buffer consisted of (in millimolar) 110 Protocol NaCl, 5.0 KC1, 2.5 CaCl2, 1.0 MgSO4, 24.0 NaHCO3, 1.0 Vessels were equilibrated for at least 30 minutes. During equil- KH2PO,,, 0.02 EDTA, and 10.0 dextrose, equilibrated with 5% ibration, vessels developed spontaneous tone. Vessels without CO2, 21% O2, and 74% N2 at 37°C (pH 7.4, Po2 150 mm Hg and spontaneous tone were discarded. Vessels with a tow sponta- Pco2 35 mm Hg). The vessel chamber, simulating the intraoc- neous tone were additionally precontracted with 0.01 to 4.0 ular space in the second series of experiments, contained two (JLM prostaglandin F2a (Sigma-Aldrich, Zwyndrecht, The Neth- glass cannulae, opposite to each other, a circular heating coil, erlands) or with 20 mM to 40 mM KC1 (Merck). The intention and a thermistor. Above the vessel chamber a video camera was to attain a tone level corresponding to a diameter of ±60% (MXR HCS; S.E.I. Benelux B.V., Breda, The Netherlands) was of the maximum diameter at the start of the experiment. Tone mounted on a microscope and connected to an electronic was necessary to study the vasodilating effect of jB-blockers. measuring system (developed in the electronic workshop of Possible interactions between sympathetic nerves in the vessel the physiology laboratory and tested for linearity) to monitor wall and jS-blockers were not expected.5 Each vessel was inner vessel diameter continuously. Perfusion pressure and tested for its endothelium functionality by supervising it with inner diameter were continuously sampled at a rate of 5 Hz. 0.3 /xM bradykinin (Sigma).17 A dilatation to more than 90% of (model M290S; Olivetti, Italy). This method has been described the maximum diameter indicated that the endothelium was earlier.'5 One side of the artery segment was mounted on one intact (Fig. 2). At the end of each experiment the passive, of two glass cannulae (outer diameter of the tip, approximately maximum-dilated vessel diameter was assessed by adding 0.1 100 jitm) and secured with a 20-/xm monofi lament nylon su- mM papaverine (Sigma) to the superfusate. ture. Then, by raising the perfusion pressure to 7.5 mm Hg, the In a first series of experiments (n = 6) the direct (i.e., vessel was flushed gently to remove blood cells. The other end pharmacologic) effect of the j3-blockers on vasomotor tone of the vessel segment was mounted and secured to the oppo- was studied. Dose-diameter relations were established for all site glass cannula. Care was taken to maintain the in situ vessel three j3-blockers. Betaxolol, befunolol, and timolol were ad- segment length (Fig. 1). The proximal cannula was connected ministered by superfusion, simulating topical instillation. There to a pump (Unita S; Braun, Germany) and the distal cannula to was no perfusion flow through the vessel segment. The intra- the overflow column filled with Krebs buffer. Perfusion flow vascular pressure was set at 75 mm Hg, and the chamber was set to zero or to a standard value of 50 jal/min. When the pressure was 0 mm Hg. Six vessels were each superfusecl with overflow column was raised, the pressure was increased in two timolol, befunolol, and betaxolol in random order to exclude steps to 75 mm Hg which was assumed to be an appropriate time effects. The concentrations were increased stepwise from perfusion pressure.l6 Any backflow in the column was consid- 10~6 mM to 10 mM. Diameter was measured when a steady ered an indication of a leak. If a leak was detected, another state was reached. After each drug application, the vessel was vessel segment was prepared. A second pump (Minipulse 3; washed with Krebs buffer until it resumed its diameter at the

Downloaded from iovs.arvojournals.org on 10/01/2021 372 Braakman et al. IOVS, February 1999, Vol. 40, No. 2

250 approximately normal value for these vessel sizes) by multiply- ing the observed diameter changes by the ratio of die actual shear stress divided by this reference shear stress. To compare the diameter changes of the different experiments, the diameter was normalized. The normalized diameter (%) was denned as follows: (actual diameter/maximum-dilated diameter) X 100%. Nonparametric statistic tests (one-sample, Friedman repeated measures, and paired) were used. P < 0.05 was considered significant.

RESULTS First Series: Pharmacologic Effects The inner diameter of the porcine short posterior ciliary arteries with vasomotor tone, measured at 75 mm Hg and in the absence of flow, ranged from 98 jam to 221 ju,m (mean, 158 ± 100 9 /xtri). The mean maximum-dilated diameter, measured under 100 200 300 the same conditions, ranged from 204 jam to 358 jam (mean, 283 ± 22 /am). The vasomotor tone resulted in a mean diam- Time (s) eter reduction of 44% ± 4% of the maximum-dilated diameter, throughout the experiment. FIGURE 2. Original tracing of a typical experiment. Bradykinin (0.3 6 /xM) in superfusion was used to test the integrity of the endothelium of Within a range of 10~ mM to 10 mM, timolol produced the isolated vessel. The resultant dilatation was evidence of intact vasodilatation at a threshold concentration of 1 mM and bet- functional endothelium. axolol and befunolol at a threshold concentration of 10~' mM (Figure 3). At less than the threshold concentration vessel diameter did not differ significantly from baseline value. The outset (control). Then the vessel was superfused with the next baseline values throughout the experiment were not signifi- drug. The mean effective dose (ED50) was calculated as the cantly different. The ED50 values (means ± SEM in millimolar) concentration at which 50% of the vasodilation trajectory was were 0.8 ± 0.3, 1.0 ± 0.3, and 11.6 ± 6.6 for betaxolol, reached (i.e., diameter at maximum vasodilatation minus diam- befunolol, and timolol, respectively. Only the ED50s of timolol eter at resting vasomotor tone). and betaxolol were significantly different. In a second series of experiments (n = 6) the indirect (i.e., hemodynamic) effect of the /3-blockers on vessel diameter was Second Series: Hemodynamic Effects studied. A situation was simulated in which an intraocular The inner diameter of the porcine short posterior ciliary arter- vessel perfused at a physiologic level was exposed to an in- ies with vasomotor tone, measured at 75 mm Hg and in the traocular pressure change from high (pathologic) to normal. absence of flow, ranged from 100 /xm to 218 /am (mean, 147 ± Transmural pressure-diameter relations were established for 18 jam). The mean maximum-dilated diameter, measured un- all three /3-blockers at a concentration of 10~3 mM (for the der the same conditions, ranged from 121 /xm to 287 /xm choice of this concentration, see the Discussion section). (mean, 218 ± 26 jam). The vasomotor tone resulted in a mean Transmural pressure is the difference between perfusion pres- diameter reduction of 38% ± 3% of the maximum-dilated sure and vessel chamber pressure, the latter simulating intraoc- diameter, throughout the experiment. ular pressure. First the transmural pressure was varied (without A diameter tracing of a step in transmural pressure from perfusion flow and with intravascular pressure set at 75 mm 30 mm Hg to 60 mm Hg (by reducing the chamber pressure Hg) from 30 mm Hg to 60 mm Hg, by changing vessel chamber from 45 mm Hg to 15 mm Hg) is shown in Figure 4. In this pressure from 45 mm Hg to 15 mm Hg. Then, perfusion flow example, this step resulted in an immediate diameter increase (Krebs buffer) was set at 50 /xl/min, intravascular pressure of 23 /xm (from 241 jam to 264 /xm). In the course of approx- remained at 75 mm Hg, and the transmural pressure was again imately 2.5 minutes, however, the diameter decreased because varied from 30 mm Hg to 60 mm Hg by changing vessel of autoregulation and reached a steady state value of 253 jam, chamber pressure from 45 mm Hg to 15 mm Hg. so that the net increase in diameter was 12 jam. From another experiment the steady state relation be- Calculations tween diameter and transmural pressure is shown (Fig. 5). A Because all vessels were perfused with the same standard flow, transmural pressure change from 30 mm Hg to 60 mm Hg, regardless of diameter, the endothelium of each vessel was effectuated by a vessel chamber pressure change from 45 mm exposed to a different shear stimulus, to which the diameter Hg to 15 mm Hg, simulated a change from pathologic intraoc- response could be different. To correct for these differences, ular pressure to normal intraocular pressure. The perfusion first shear stress for each vessel was calculated using the flow was set to zero, and no /3-blocker was present in the formula 4-q-Q/TrR3', (where rj is viscosity, Q is flow, and R is superfusate. As can be seen in this example, a transmural inner radius) from standard flow (50 jal/min) and actual inner pressure step from 30 mm Hg to 60 mm Hg increased vessel diameter. Then the diameter responses due to actual shear diameter from 126 jam to 128 jam. After the transmural pres- stress were normalized to a shear stress of 25 Pa (which is the sure was set back to 30 mm Hg, perfusion flow was begun and

Downloaded from iovs.arvojournals.org on 10/01/2021 IOVS, February 1999, Vol. 40, No. 2 Hemodynamic and Pharmacologic Effects of /3-Blockers 373

A B C

100 -

.2 80 - Q •o (D N 15 60 - E o 40 -10 -8 -6 -4 -2 -10 -8 -6 -4 -2 -10 -8 -6 -4 -2 Log[betaxolol] (M) Log[befunolol] (M) Log[timolol] (M)

FIGURE 3- Semilogarithmic relation between normalized diameter (%) and j3-blocker concentration (M): (A) betaxolol, (B) befiinolol, and (C) timolol. The baseline value is at the right side of each panel and represents the control experiment (c) in which no j3-blocker was present in the superfusate. The plotted points are mean values ± SEM.

set to 50 |al/min. The vessel diameter then increased to 134 The effect of the standard shear stimulus on the vessel /Am. Repeating the transmural pressure step from 30 mm Hg to diameter for all experiments is summarized in Figure 6. At a 60 mm Hg resulted in a vessel diameter increase to 146 \xm. transmural pressure of 60 mm Hg the mean relative diameter Thus, the slope (Adiameter/Atransmural pressure) and offset of increase differed significantly from zero (P = 0.03). The rela- the diameter-transmural pressure relation both increased com- tive diameter increase at a transmural pressure of 30 mm Hg pared with those parameters in the no-flow situation. The was not significantly different from the relative diameter in- transmural pressure step was subsequently repeated (still in crease at 60 mm Hg. the presence of perfusion flow) with the three j3-blockers, each The relative diameter change, without and with /3-block- in a concentration of 10~3 mM. In this experiment the slope ers, as a result of the transmural pressure change (from 30 mm did not change, and the diameter offset was decreased by each Hg to 60 mm Hg), in the presence of standard perfusion flow, of the 0-blockers. This offset was variable per experiment and is summarized for all experiments in Figure 7. The relative on the average not different from the diameter offset without diameter increase without j3-blockers (i.e., the control situation /3-blockers (see Fig. 7).

150

270 -

260 - E

250 - CD

CO Q 240 -

230 - 120 20 30 40 50 200 p transmural (mm H9) Time (s) FIGURE 5- Relation between diameter (in micrometers) and transmu- FIGURE 4. Original tracing of a typical experiment. A stepwise in- ral pressure (in millimeters of mercury) of a typical experiment. (•) No crease in transmural pressure (from 30 mm Hg to 60 mm Hg) initially standard perfusion flow, no /3-blocker. (•) Standard perfusion flow, no produced an increase in diameter of the vessel and then a decrease, as /3-blocker. (A) Standard perfusion flow, 10~3 mM betaxolol. (T) Stan- a result of myogenic autoregulation. In the steady state a net increase dard perfusion flow, 10~3 mM befiinolol. (•) Standard perfusion flow, in diameter remained. 10~3 mM timolol.

Downloaded from iovs.arvojournals.org on 10/01/2021 374 Braakman et al. IOVS, February 1999, Vol. 40, No. 2

found a positive correlation between transmural pressure and vessel diameter.

Critique of Methods CD -t—< CD We used the short posterior ciliary artery to study the effect of topical j3-blockers on vessel diameter, because the optic nerve head is perfused largely by these arteries14 and thus depends on the lumen diameter of these vessels. To study blood flow in "CO E human small ocular arteries, measuring devices such as laser x 5 - and Doppler ultrasound are used that provide information on CO blood flow velocity, but not on blood flow. Without a simul- taneously measured diameter, nothing can be concluded about CD •t—• blood flow. Moreover, in clinical studies it is almost impossible CD to assess whether a change of ocular blood flow by topical CO j3-blockers is caused by their direct pharmacologic effect TJ (through direct action on the smooth muscle of the vascular CO wall) or by their indirect hemodynamic effect (indirectly,

•*—> CD through lowering of extravascular, or intraocular, pressure). Q We have chosen for an in vitro approach, which has the advantage that perfusion flow, perfusion pressure, and ex- 30 60 travascular pressure can be varied independently and measured directly. As far as we are aware, never before has the ° transmural (mm short posterior ciliary artery been cannulated, nor has an ex- FIGURE 6. Diagram of the effect of starting the standard perfusion perimental attempt been made to separate the pharmacologic flow (from 0 to 50 p-l/min) on the (shear corrected) diameter at 30 mm effect from the hemodynamic effect of /3-blockers on vessel Hg to 60 mm Hg transmural pressure. Delta diameter is the vertical diameter. A disadvantage of our in vitro approach is that we difference in Figure 5 between (•) and (•), divided by the maximum must extrapolate our results to the human condition. diameter. To study the pharmacology and physiology of small arteries, vessel cannulation is a more physiological approach than the wire myograph technique, because flow through the lu- with shear only) was significantly different from zero (P = men is possible, the cross-section is circular, and the vessel 0.03). With each of the j3-blockers the relative diameter in- length can be adjusted to the in vivo length.22 In our first series crease was not significantly different from the relative diameter of experiments, we kept perfusion flow at zero, so that the increase without a /3-blocker. major difference between our experimental setup and the wire

DISCUSSION Previous studies have indicated that vascular factors play a role in the pathogenesis of primary open angle glaucoma, especially in normal-tension glaucoma.l8 There is evidence that the blood supply from retrobulbar vessels to the optic nerve head is compromised in glaucoma."19'20 Today, lowering intraocular pressure is still the only choice for therapy. Intraocular pressure reduction alone can cause hemodynamic changes that increase intraocular blood flow."12 However, new therapeutic approaches are emerging: for example, neuroprotection21 and vasodilatation. Regarding the latter, it has been shown that topical j3-blockers have a direct relaxing action on the vessel wall of the external iliac, the retinal, and the long posterior ciliary arteries.4'810 We have shown that, in the absence of flow, a nonselective, a selective, and an intrinsic sympathetic activity j3-blocker all increased the diameter of the short posterior ciliary artery by a direct pharmacologic effect, but only at very high concentrations. Timolol seemed to be a weaker vasodilator than betaxolol by a factor of 15. In the presence of flow and at a lower concentration, all three /3-blockers induced vasodilatation or control betaxolol befunolol timolol vasoconstriction in individual experiments. The overall effect, FIGURE 7. Diagram of the effect of a transmural pressure step from 30 however, was not significantly different from zero. Finally, we mm Hg to 60 mm Hg on the (shear corrected) diameter in the presence have shown that, in using the extraocular tract of the short of standard perfusion flow. The difference in diameter between trans- posterior artery as a model for the intraocular tract of the short mural pressure of 60 mm Hg and 30 mm Hg is calculated and divided posterior ciliary artery (or an intraocular vessel in general), we by the maximum diameter.

Downloaded from iovs.arvojournals.org on 10/01/2021 IOVS, February 1999, Vol. 40, No. 2 Hemodynamic and Pharmacologic Effects of /3-Blockers 375

160 fixed flow rate and a fixed bolus volume), whereas we used continuous drug delivery. Because the resistance of the open- ended side branch depends on the local intravascular pressure and on the concentration of the /3-blocker, their results may contain nonlinear effects. They found that /3-blockers dilate precontracted porcine retinal vessels at threshold concentrations of approximately 10~7 mM. With the wire myograph technique they found that /3-blockers relax precontracted porcine long posterior ciliary arteries at threshold concentrations of approximately 10~3 mM. They ascribe the difference in sensitivity by a factor of 104 to the relatively small size of the retinal vessel and to the use of the cannulated vessel technique. Therefore, the presence of flow may explain the difference in sensitivity. Furthermore, because the cell membrane is strongly depolarized by the high potassium solution that they used to constrict the retinal vessel, the latter may have been rendered extremely sensitive to the supposedly calcium entry-blocking effect of j3-blockers.

120 Results of the First Series: Pharmacologic Effects 200 400 200 400 600 These results showed that die /3-blockers dilated the short Time (s) posterior ciliary artery at threshold concentrations ranging FIGURE 8. Original tracing of a diameter response on beginning the from 10~' mM to 1 mM, with ED50 ranging from 0.8 mM to standard perfusion flow at a transmural pressure of 75 mm Hg. (A) 11.6 mM. Flow was set to zero in these experiments to avoid Without L-NNA in the superfusate, the diameter response was from 130 interference of diameter changes caused by /3-blocker effects /am to 154 /xm. (B) With 10~' mM L-NNA supervision, the diameter and those caused by shear stress effects. The difference be- response was from 123 p,m to 128 /xm. The diameter response was tween our results and those in the literature should be inter- reduced from 17% to 4% by L-NNA. The decrease in initial diameter preted with care because of the different methods used. At the (i.e., before the step response) was caused by adding L-NNA and is the first glance it could be concluded that the porcine short pos- result of blocking the basal nitric oxide production. terior ciliary artery is 100 (a ratio of threshold concentration values) to 10 (a ratio of ED50 values) times less sensitive to /3-blockers than the porcine long posterior ciliary artery or the myograph technique was the geometry of the preparation. We bovine retinal artery. Falloon et al.,22 however, have pointed used a preparation prestretched to its original length with a out that the marked difference in the vascular geometry of circular lumen. In the wire myograph setup, the length of the arteries mounted in the wire myograph and in the cannulae vessel is not the in vivo length, and the lumen is collapsed. system may have implications for the relative contribution of Moreover, in our second series of experiments vessels were wall elements in the two systems. They state that the abnormal perfused with fluid, so that shear stress was applied to the geometry of wire-mounted arteries may be responsible for an endothelium, which plays an important role in the regulation 5 23 25 enhanced sensitivity to norepinephrine, so that the reduced of the vessel diameter.' - ~ As can be seen from the results, sensitivity to norepinephrine found in the cannulae system flow alone had already increased the diameter. In one experi- may reflect the in vivo situation more closely. When instilled in ment A^-nitro-L-arginine (L-NNA) was added to the superfusate the conjunctival sac, the final concentration of a topical (see Figs. 8A, 8B). Although not conclusive, the strongly dimin- j3-blocker at the level of the short posterior arteries will at best ished increase of the diameter of the vessel after application of 19 17 24 be reflected by the plasma concentration. From experiments L-NNA supports the work of Haefliger et al., Kuo et al., and 25 on humans and rabbits the plasma concentration was esti- Meyer et al. showing that the mediator of the flow-induced mated to be 10~3 mM to 10~5 mM.27"29 In our experiments response is nitric oxide. Finally, external pressure was applied the threshold concentration for the short posterior ciliary ar- to the cannulated vessels (all of which showed vasomotor tery to dilate was found to be 10~' mM to 1 mM. Thus, it is tone), so that a physiological situation was simulated, as nearly unlikely that j3-blockers, when topically applied, reach a suffi- as possible, in the pressurized vessel chamber. This interven- ciently high concentration in the retrobulbar space to increase tion is not possible in the wire myograph. Several investiga- 410 26 the perfusion of the optic nerve head by dilating the short tors ' have used the wire myograph technique and found posterior ciliary arteries. /3-blockers to relax precontracted ring segments of bovine retinal and porcine long posterior ciliary artery at threshold concentrations ranging between 10~3 mM and 10~' mM and Results of the Second Series: Hemodynamic Effects (estimated) ED50s ranging between 0.1 mM and 1 mM, with timolol a less potent vasodilator than betaxolol by a factor of In our second series of experiments fluid flow was applied, l0 10. Yu et al. also used a modified cannulated vessel technique simulating reality as much as possible. Because we wanted to with an open side branch between the cannulae. Because of a apply a concentration of /3-blocker, as estimated above, to the connection between the side-branch lumen and the bath, they short posterior ciliary artery, we applied 10~3 mM /3-blocker, could not study the effects of transmural pressure changes. although the threshold concentrations were 10~' mM to 1 mM Furthermore, they used a cumulative bolus drug deliver}' (as a in the first series. In the first series, however, flow was set to

Downloaded from iovs.arvojournals.org on 10/01/2021 376 Braakman et al. IOVS, February 1999, Vol. 40, No. 2

zero. The presence of flow may render an ocular vessel more and may also change the level of venous collapse, or vascular sensitive to a /3-blocker.l0 In individual experiments we found waterfall, that Attariwala et al.32 postulated to be downstream that /3-blockers affected vessel diameter. However, diameters in the central retinal vein, possibly at the level of the lamina were observed to increase in some experiments but to de- cribrosa. crease in others (see, for instance, the diameter reduction in Fig. 5). The diameter changes by the /3-blockers therefore did not add significantly to the effect of transmural pressure on CONCLUSIONS vessel diameter. In all our results the effect of autoregulation on vessel Only at concentrations, far exceeding their estimated plasma diameter was included. The autoregulatory response time in concentrations, betaxolol, befunolol, and timolol increased the our experiments was approximately 2.5 minutes. Values diameter of the isolated porcine short posterior ciliary artery, within a range of ±2 to ±5 minutes are considered physio- as a result of their direct pharmacologic effect. The difference logic.3031 From our results it can be inferred that autoregula- between the vasodilatory potency of the selective and the 3 tion is not influenced by j8-blockers in a concentration of 10~ nonselective /3-blocker was significant: the ED50 of betaxolol mM, because they did not alter the slope of the diameter- was 15 times lower than the ED50 of timolol. There was a transmural pressure relation (Tig. 7). This is in agreement with positive correlation between the diameter of the isolated por- results published by Hoste et al.26 cine short posterior artery (when used to simulate an intraoc- There is a negative correlation between intraocular pres- ular artery) and the transmural pressure, which corroborates sure and optic nerve head perfusion.11'12 We corroborated this the indirect hemodynamic effect of /3-blockers. It is speculated finding by showing that there is a positive correlation between that instillation of topical /3-blockers into the conjunctival sac transmural pressure and vessel diameter. Because flow was may increase the perfusion of the optic nerve head by an inversely proportional to the fourth power of the diameter it indirect, hemodynamic mechanism, but not by a direct, phar- can be inferred that transmural pressure changes, even if they macologic mechanism. cause small diameter changes, may result in considerable flow changes. Instilling a /3-blocker in the eye may thus lead to an Acknowledgments increase of the flow through intraocular vessels. We used the The authors thank Jaap van Best, head of the laboratory of the Eye extraocular tract of the porcine short posterior ciliary artery in Clinic of the University Hospital Leiden, for his cooperation with the the vessel chamber to model its intraocular tract (or any in- fluorophotometry and Robert Krams, Department of Hemodynamics, traocular artery of comparable size) in the intraocular space. Erasmus University Rotterdam, for critically reading the manuscript. Supposing that their specific compliance is comparable, it is not unreasonable to assume that intraocular arteries behave References comparably to transmural pressure changes as extraocular arteries, if the ratio of vessel radius and vessel wall thickness is 1. Flammer J. To what extent are vascular factors involved in the pathogenesis of glaucoma? In: Kaiser HJ, Flammer J, Hendrickson fairly constant, whether porcine or human. The adventitia of PH, eds. Ocular Blood Flow, New Insights into the Pathogenesis the retinal artery is thinner than that of the short posterior of Ocular Diseases. Basel, Switzerland: Karger; 1996;12-39. ciliary artery.10 But this would only result in an underestima- 2. Zimmerman TJ, Kaufman HE. Timolol: a beta-adrenergic blocking tion of the effect of transmural pressure on diameter, because agent for the treatment of glaucoma. Arch Ophthalmol. 1977;95: wall stretch is inversely proportional to wall thickness at the 601-604. 3. Cyander MS, Gupta N, McAllister R, et al. Mapping the distribution same diameter. of neurotransmitter receptors, second messenger molecules, and ion channels in the human eye. In: Anderson DR, Drance SM, eds. Considerations Encounters in Glaucoma Research. Receptor Biology and Glau- The results of our in vitro study on isolated porcine vessels coma. Vol. 1. Milan, Italy: Fogliazza; 1994;301-325. 4. Hester RK, Chen Z, Becker EJ, et al. The direct vascular relaxing with regard to the direct pharmacologic effect of j3-blockers action of betaxolol, carteolol, and timolol in porcine long posterior may at best qualify as circumstantial evidence when applied to ciliary artery. Surv Ophthalmol. 1994;38:S125-S133. the human situation. The indirect hemodynamic effect may be 5. Duijm HFA, Thoe Schwartzenberg-Terpstra ACE, van de Berg applicable to both the porcine and the human situation. If we TLTP, et al. Present findings regarding hemodynamic effects of extrapolate our results to the human situation, we speculate /3-blockers in glaucoma. In: Drance SM, ed. Vascular Risk Factors and Neuroprotection in Glaucoma: Update 1996. Amsterdam: that instillation of topical /3-blockers in the eye does not in- Kugler; 1997;175-180. crease the perfusion of the optic nerve head by a direct relax- 6. Martin XD, Rabineau PA. Vasoconstrictive effect of topical timolol ing action on the smooth muscle of the vessel wall of the short on human retinal arteries. Graefes Arch Clin Exp Ophthalmol. posterior ciliary arteries, because /3-blockers do not reach a 1989:227:526-530. high enough concentration in the retrobulbar space. Their 7. Yoshida A, Feke GT, Ogasawara H, et al. Effects of timolol on value as antiglaucomatous drugs lies primarily in their intraoc- human retinal, choroidal and optic nerve head circulation. Oph- ular pressure-lowering effect. Thus, they seem the logical thalmic Res. 1991:23:162-170. 8. GrunwaldJE. Effect of timolol maleate on the retinal circulation of therapeutic consequence of the pressure theory. Their effect the human eyes with ocular hypertension. Invest Ophthalmol Vis on intraocular pressure, however, affects the hemodynamic Sci. 1990;31:521-526. state in the eye, but in a indirect way. /3-Blockers induce an 9. Harris A, Spaeth GL, Sergott RC, et al. Retrobulbar arterial hemo- increase of the transmural pressure over the intraocular vessel dynamic effects of betaxolol and timolol in normal-tension glaucoma. Am J Ophthalmol. 1995;120:l68-175. walls, thereby increasing their diameter and enhancing the 2+ perfusion flow into the eye, including the optic nerve head. In 10. Yu DY, Su EN, Cringle SJ, et al. Effect of /3-blockers and Ca entry blockers on ocular vessels. In: Drance SM, ed. Vascular Risk the eye /3-blockers thus lower the back pressure (or down- Factors and Neuroprotection in Glaucoma: Update 1996. stream pressure) with respect to the central blood pressure Amsterdam: Kugler; 1997:123-134.

Downloaded from iovs.arvojournals.org on 10/01/2021 IOVS, February 1999, Vol. 40, No. 2 Hemodynamic and Pharmacologic Effects of /3-Blockers 377

11. Greve EL, Buimer R, Duijm HFA, et al. SLO fluoresceinangiograph y 22. Falloon BJ, Stevens N, Tulip JR, et al. Comparison of small artery of the peripapillary choroid before and after trabeculectomy sensitivity and morphology in pressurized and wire-mounted prep- [ARVO Abstract]. Invest Ophthalmol Vis Sci. 1995;36(4):S436. arations. AmJ Physiol. 1995;267:H670-H678. Abstract nr 2001. 23. Mostaghim R, Maddox YT, Ram well PW. Endothelial potentiation 12. Kimura Y, Nitta A, Takayama H, Shimizu R. The effect of raised of relaxation response to beta adrenoceptor blocking agents. intraocular pressures on blood flow in the optic nerve head in J Pharmacol Exp Ther. 1986;239:797-801. monkeys. Chibret Int J Ophthalmol. 1987;5:24-31. 24. Kuo L, Davis MJ, Chilian WM. Endothelium-dependent, flow-in- 13. Bill A. Circulation in the eye. In: Renkin EM, Michel CC, eds. The duced dilation of isolated coronary arterioles. AmJ Physiol. 1990; Handbook of Physiology: Cardiovascular System. Vol. 4. Be- 259:H1O63-H1O7O. thesda, MD: American Physiological Society; 1984:1001-1029- 25. Meyer P, Flammer J, Liischer TF. Endothelium-dependent regula- 14. Biichi ER. The blood supply to the optic nerve head. In: Kaiser HJ, tion of the ophthalmic microcirculation in the perfused porcine Flammer J, Hendrickson PH, eds. Ocular Blood Flow: New In- eye: role of nitric oxide and endothelins. Invest Ophthalmol Vis sights into the Pathogenesis of Ocular Diseases. Basel, Sci. 1993;34:36l4-3621. Switzerland: Karger; 1996:1-8. 26. Hoste AM, Boels PJ, Andries LJ, et al. Effects of beta-antagonists on 15. Sipkema P, van der linden PJW, Hoogerwerf N, Westerhof N. Does contraction of bovine retinal microarteries in vitro. Invest Oph- the endothelium play a role in flow-dependent constriction? A study thalmol Vis Sci. 1990;31:1231-1237. in the isolated rabbit femoral artery. Blood Vessels. 1989;26:368-376. 27. Ahmed I, Patton TF. Importance of the noncorneal absorption 16. Glucksberg MR, Dunn R. Direct measurement of retinal microvas- cular pressures in the live, anesthetized cat. Microvasc Res. 1993; route in topical ophthalmic drug delivery. Invest Ophthalmol Vis 45:158-165. Sci. 1985;26:584-587. 17. Haefliger IO, Flammer J, Liischer TF. Nitric oxide and endothelin-1 28. Vuori ML, Ali-Melkkila T, Kaila T, et al. Plasma and aqueous humor are important regulators of human ophthalmic artery. Invest Oph- concentration and systemic effects of topical betaxolol and timolol thalmol Vis Sci. 1992;33:2340-2343. in man. Ada Ophthalmol. 1993;71:201-206. 18. Geijssen HC, Greve EL. Vascular concepts in glaucoma. Curr Opin 29. Schmidt CJ, Lotti VJ, LeDouarec JC. Penetration of timolol into the Ophthalmol. 1995;6:71-77. rabbit eye. Arch Ophthalmol. 1980;98:547-551. 19- Aim A. Etiological and Pharmacological Aspects of Blood Floiv 30. Braakman R, Sipkema P, Westerhof N. A dynamic non-linear and Glaucoma: Update 1996. Amsterdam: Kugler; 1997:167-173. lumped parameter model for skeletal muscle circulation. Ann 20. Kaiser HJ, Flammer J, Schoetzau A. Blood flow in retrobulbar BiomedEng. 1989;17:593-6l6. vessels in glaucoma and normals. In: Drance SM, ed. Vascular Risk 31. Braakman R, Sipkema P, Westerhof N. Two zero-flow pressure Factors and Neuroprotection in Glaucoma: Update 1996. intercepts exist in autoregulating isolated skeletal muscle. Am J Amsterdam: Kugler; 1997:135-138. Physiol. 1990;258:H1806-H18l4. 21. Osbone NN. Neuroprotection to the retina: relevance in glaucoma. 32. Attariwala R, Giebs CP, Glucksberg MR. The influence of elevated In: Drance SM, ed. Vascular Risk Factors and Neuroprotection in intraocular pressure on vascular pressures in die cat retina. Invest Glaucoma: Update 1996. Amsterdam: Kugler; 1997:139-155. Ophthalmol Vis Sci. 1994;35:1O19-1O25.

Downloaded from iovs.arvojournals.org on 10/01/2021