JOURNAL OF OCULAR PHARMACOLOGYAND THERAPEUTICSVolume 17, Number 3, 2001Mary Ann Liebert, Inc.

Effect of Brimonidine Tartrate on OcularHemodynamics in Healthy Volunteers

CHRISTIAN P. JONESCU-CUYPERS1, ALON HARRIS1,2, YOKO ISHII1,LARRY KAGEMANN1, HANNA J. GAZOZI3, YGAL ROTENSTREICH1,

HAK SUNG CHUNG1, and BRUCE MARTIN2

Glaucoma Research and Diagnostic Center, Departments of 1Ophthalmology and 2Physiology and

Biophysics, Indiana University Medical Center, Indianapolis, Indiana3Department of Ophthalmology, Central Emek Medical Center, Afula, Israel

ABSTRACT

While a2-adrenergic agonists, such as brimonidine tartrate, significantly reduce theintraocular pressure (IOP), the presence of vasoconstrictor postsynaptic a2 receptors onvascular smooth muscle raise the possibility that brimonidine could potentially compro-mise ocular blood flow. Consequently, the ocular hemodynamic effects of brimonidinewere studied in normal subjects.

Twelve healthy volunteers were included in this prospective, double-masked, placebocontrolled, crossover-designed clinical trial. They received either brimonidine tartrate0.2% or placebo b.i.d. for 2 weeks. Goldmann tonometry and color Doppler imaging (CDI)were performed at baseline, at 2 hr, 1 week, and 2 weeks after the treatment. Fundus an-giography using a scanning laser ophthalmoscope was performed at baseline and 2 weeksafter treatment to determine retinal arteriovenous passage time.

Brimonidine lowered IOP at 2 hr, 1 week, and 2 weeks (p 5 0.058, p 5 0.031, andp 5 0.022, respectively). Brimonidine did not affect the retrobulbar arterial velocities mea-sured by CDI, nor retinal arteriovenous passage time.

In conclusion, two-week treatment with brimonidine reduces IOP and does not re-duce the bulk retinal or retrobulbar arterial perfusion in young healthy volunteers.

INTRODUCTION

Retinal ganglion cell death in glaucoma is caused by apoptosis (1,2). Two major factors trig-gering apoptosis in ganglion cells are withdrawal of neurotrophin and excessive release of glutamate(3). Either mechanical damage by high intraocular pressure (IOP), or an ischemic insult from de-creased perfusion, may accelerate neurotrophin withdrawal and glutamate toxicity. Therefore, in or-der to optimize glaucoma treatment, interventions should maximize neuroprotection either directly orindirectly by reducing pressure and improving ocular blood flow. However, reducing IOP is the onlycurrent clinically available treatment modality for glaucoma patients. It is thus essential to evaluateantihypertensive glaucoma medications for their vascular effects.

199

Brimonidine tartrate 0.2% (Alphagan, Allergan, Inc., Irvine, CA), a potent selective a2-adrenore-ceptor agonist, is widely used in glaucoma treatment (4). Topical brimonidine decreases IOP by sup-pressing aqueous humor production and enhancing uveoscleral outflow. In addition, animal studiessuggest that the agent may also have a direct neuroprotective effect on retinal ganglion cells (5). How-ever, while centrally active a2-adrenergic agonists (e.g., clonidine) reduce sympathetic outflow fromthe brain and are used to treat systemic hypertension (6), peripherally active a2-adrenergic agonists,such as brimonidine, would be expected to constrict vessels supplied with postsynaptic a2 receptors.Vasoconstriction within the retina or fundus could be deleterious to glaucoma patients.

It remains unknown if topical brimonidine can reach the posterior pole or, if having reached thatsite, it can constrict vessels within the retinal nerve fiber layer and optic nerve head. In this study,we evaluated the hemodynamic actions of topical brimonidine by studying its effects on retinal andretrobulbar vessels.

SUBJECTS AND METHODS

We performed a randomized, prospective, double-masked, cross-over clinical trial with placebocontrol. Twelve healthy volunteers (9 women, 3 men; mean age 33 6 9.8 years) were recruited. Nosubject had a history of ocular or systemic disease or was taking topical or systemic medication. Thestudy eye was randomly chosen. Informed consent was obtained from all subjects, and an institutionalreview board approved the protocol. All procedures conformed to the tenets of the Declaration ofHelsinki.

After baseline measurements, subjects were treated with topical brimonidine 0.2% (Alphagan®,Allergan) or placebo, b.i.d. for 2 weeks in the study eye. The vehicle was used as placebo. After 2weeks of drug washout, a second set of baseline measurements was taken. Subsequent to this secondcontrol, the remaining drug regimen was carried out over a 2-week period. During the treatment pe-riod, the measurements were performed after 2 hr, 1 week, and 2 weeks. All measurements were ob-tained at the same time of day for each visit. Subjects avoided caffeine intake, smoking, and exercisefor 3 hr prior to the measurements.

IOP was determined by Goldmann applanation tonometry. Systemic blood pressure was mea-sured using sphygmomanometry; heart rate was determined by palpation. Ocular perfusion pressure(OPP) was calculated from systemic blood pressure and IOP using the formula, OPP 5 2/3 mean ar-terial blood pressure 2 IOP (7). Color Doppler imaging velocity measurements were performed onthe ophthalmic artery (OA), central retinal artery (CRA), and a short posterior ciliary artery (SPCA)with subjects comfortably reclining at a 60° angle. A Siemens Quantum 2000 (Issaquah, Washing-ton) with a 7.5-MHz linear phase transducer was used for imaging. From each vessel, peak systolicvelocity (PSV) and end diastolic velocity (EDV) were recorded. Resistance index (RI) [(PSV-EDV)/PSV] was calculated for each vessel (8). To examine the OA, the sample volume is orientednasally and superior to the optic nerve, just lateral to and abutting the visible hyporeflective striperepresenting the nerve. The CRA is measured anterior to the optic nerve: the sample volume is placedabout 3 mm behind the surface of the optic disk. SPCAs are located temporal or nasal to the opticnerve shadow. Because of their small size (less than 200 mm) relative to the sample volume, it is notalways possible to resolve individual vessels. Nevertheless, the presence of colored pixels in this re-gion and the characteristic Doppler spectrum obtained from them confirm the presence of posteriorciliary artery flow. Fluorescein angiography with a scanning laser ophthalmoscope (SLO 101; Ro-denstock Instruments, Munich, Germany) was performed three times: once at the first baseline, and2 weeks after treatment with either brimonidine or placebo (8). After all other examinations werecomplete, the pupil was dilated with 0.1% tropicamide and 0.5% cyclopentolate before angiography.After antecubital venous injection of 5 cc fluorescein dye, the angiogram was videotaped and ana-lyzed off-line using an HRX digital image analysis system. After digitization of a series of videotapeframes, fluorescence density changes were determined. After selection of 3 3 3 pixel areas on the su-

200

perior and inferior temporal arteries, and corresponding veins, dye dilution curves were plotted to de-termine the appearance time of dye. Arteriovenous passage (AVP) time is determined by finding thedifference between the time of the appearance of the fluorescein dye in the arteries and its appear-ance in the corresponding veins.

Statistical comparisons were made using two-tailed paired t-tests for IOP, heart rate, blood pres-sure, and OPP data. The AVP time and capillary transit velocity were calculated as percentage changesfrom baseline. Changes due to the different treatments were compared by the two-way analysis ofvariance. When the overall analysis of variance was significant (p , 0.05), paired t-tests (with Bon-ferroni adjustments of significance levels) were used to test differences between groups. The samplesize was chosen to provide 90% power to detect a 15% change in flow velocity or RI in the OA, and80% power for a 15% change in CRA or SPCA velocity. The sample provided 80% power to detecta 15% change in AVP time.

RESULTS

Brimonidine reduced IOP at 2 hr, 1 week, and 2 weeks (p 5 0.058, p 5 0.031, and p 5 0.022,respectively) (Table 1). Brimonidine did not change retrobulbar arterial blood flow velocities mea-sured by CDI (Table 2). Arteriovenous passage time remained unchanged by brimonidine (baseline,after placebo treatment, and brimonidine) (Table 3). Neither IOP nor parameters of CDI and fluores-cein angiography in blood flow measurements were altered by placebo. Heart rate, blood pressure,and OPP remained unchanged by brimonidine (Table 4).

DISCUSSION

Brimonidine tartrate is a relatively potent and highly selective a2-adrenoreceptor agonist and hasmarkedly greater affinity for a2-adrenoceptors than apraclonidine (23- to 32-fold) and clonidine (6-to 12-fold), which are also selective a2-adrenergic agonists (9). Brimonidine lowers IOP by reducingaqueous humor production and increasing aqueous humor outflow via the uveoscleral pathway (10,11).The reduction of IOP is mediated by stimulation of ocular a2-adrenoceptors in the ciliary body (10,11).Further, and perhaps equally important, brimonidine has been consistently reported to protect retinalganglion cells from acute insults, including retinal ischemia/reperfusion and calibrated optic nervecompression (5).

Glaucoma is characterized by optic nerve head excavation and death of retinal ganglion cells.Although elevated IOP is an important risk factor for the development of visual field loss in patientswith glaucoma, approximately 20–50% of patients with glaucomatous optic nerve damage have IOPwithin the normal range (12), suggesting that factors apart from ocular tension may contribute to glau-coma initiation and progression. Just as the risk factors for glaucoma development remain unclear,the cellular mechanisms mediating ganglion cell death in glaucoma are not well known. While gan-glion cells primarily die via apoptosis (1,2), the factors regulating this process remain undefined (13).

201

However, several biochemical factors involved in retinal ganglion cell death are identical to thosemediating ischemic cell death in the central nervous system (14), suggesting that ischemia could con-tribute to glaucomatous optic nerve head damage, at least in susceptible individuals. This possibilityargues that interventions designed to protect the retinal ganglion cell should be carefully screened foreffects on retinal and optic nerve head blood flow.

By lowering IOP, brimonidine clearly possesses beneficial properties for the treatment of glau-coma. However, brimonidine, as an a2-adrenergic agonist, belongs to a drug class reported to havesystemic vasoconstrictive effects (6). Thus, it is possible that topical application of the drug could in-

202

duce ocular vasoconstriction. No evidence for reductions of either bulk retinal flow or retrobulbar ar-terial perfusion was found after brimonidine administration. Our results confirm other findings thatfail to document any ocular vasoconstrictor effects of topical brimonidine in animal models or hu-man subjects (15–17).

While we did fail to detect a significant brimonidine-induced increase of ocular perfusion pres-sure (15), this estimate may be inexact. In fact, the fall in IOP at constant arterial blood pressure maybe sufficient to create a greater pressure gradient for perfusion within the perifoveal capillary bed. In-deed, Harris et al. reported that flow velocities in the central retinal arteries are highly IOP depen-dent, implying that decreased IOP can increase flow velocities in the central retinal arteries (18). Al-ternatively, local effects of the drug could predominate. Brimonidine is less lipophilic and less likelyto cross the blood-brain barrier than clonidine; topical application of brimonidine will clearly haveminimal central effects on sympathetic outflow. Using in vitro ligand-binding and autoradiography,a large number of specific brimonidine binding sites have been identified on the human iris and cil-iary epithelium, but a smaller number of binding sites are located on human ciliary muscle, retina,retinal pigmented epithelium, and choroid (19). Additionally, although brimonidine can penetrate tothe posterior segment of the human eye after topical dosing, its concentration reaches only about 1027

M. In vitro studies find brimonidine unable to compromise blood flow in the retina or optic nerve atconcentrations as high as 1025 M, while apraclonidine and clonidine demonstrate vasoconstrictor ac-tivity at concentrations as low as 10211 M (20). Taken together, these findings suggest that, despiteits high potency as an a2 agonist at the ciliary body, brimonidine has a relatively low potency as anocular vasoconstrictor, and with its relatively low concentration at the posterior pole of the eye aftertopical application, the drug may allow subtle changes in local ocular perfusion pressure to emergeas the predominant effect upon macular capillary hemodynamics.

203

In conclusion, brimonidine does not reduce the bulk retinal or retrobulbar arterial perfusion. How-ever, these results must be interpreted with caution, because responses in patients over the long termmay differ from those seen acutely in healthy persons.

REFERENCES

1. Quigley, H.A., Nickells, R.W., Kerrigan, L.A., Pease, M.E., Thibault, D.J., and Zack, D.J. Reti-nal ganglion cell death in experimental glaucoma and after axotomy occurs by apoptosis. In-vest. Ophthalmol. Vis. Sci. 36:774–786, 1995.

2. Pease, M.E., McKinnon, S.J., Quigley, H.A., Kerrigan-Baumrind, L.A., and Zack, D.J. Ob-structed axonal transport of BDNF and its receptor TrkB in experimental glaucoma. Invest. Ophthalmol. Vis. Sci. 41:764–774, 2000.

3. Nickells, R.W. Retinal ganglion cell death in glaucoma: the how, the why, and the maybe. J.Glaucoma 5:345–356, 1996.

4. Adkins, J.C. and Balfour, J.A. Brimonidine. A review of its pharmacological properties and clin-ical potential in the management of open-angle glaucoma and ocular hypertension. Drugs Ag-ing 12:225–241, 1998.

5. Wheeler, L.A., Lai, R., and Woldemussie, E. From the lab to the clinic: activation of an alpha-2 agonist pathway is neuroprotective in models of retinal and optic nerve injury. Eur. J. Oph-thalmol. 9:S17–21, 1999.

6. Hoffman, B.B. and Lefkowitz, R.J. Catecholamines and sympathomimetic drugs. In Goodmanand Gilman’s The Pharmacological Basis of Therapeutics, 8th ed., Gilman, A.G., Rall, T.W.,Nies, A.S., Taylor, P., eds., Pergamon Press, New York, 1990, pp. 187–220.

7. Harris, A., Kagemann, L., and Cioffi, G. Assessment of human ocular hemodynamics. Surv.Ophthalmol. 42:509–533, 1998.

8. Harris, A., Chung, H.S., Ciulla, T.A., and Kagemann, L. Progress in measurement of ocularblood flow and relevance to our understanding of glaucoma and age-related macular degenera-tion. Prog. Retin. Eye Res. 18:669–687, 1999.

9. Chien, D.S., Homsy, J.J., Gluchowski, C., and Tang-Liu, D.D. Corneal and conjunctival/scleralpenetration of p-aminoclonidine, AGN 190342, and clonidine in rabbit eyes. Curr. Eye Res.9:1051–1059, 1990.

10. Burk, J. and Manlapaz, C. Therapeutic use of alpha2-adrenoceptor agonists in glaucoma. In Al-pha2-Adrenergic Receptors: Structure, Function and Therapeutic Implications, Lanier, S., Lim-bird, L., eds., Reading:Harwood Academic Publishers, 1996, pp. 179–187.

11. Toris, C.B., Gleason, M.L., Camras, C.B., and Yablonski, M.E. Effects of brimonidine on aque-ous humor dynamics in human eyes. Arch. Ophthalmol. 113:1514–1517, 1995.

12. Quigley, H.A. Open-angle glaucoma. N. Engl. J. Med. 328:1097–1106, 1993.

204

13. Nickells, R.W. Apoptosis of retinal ganglion cells in glaucoma: An update of the molecularpathways involved in cell death. Surv. Ophthalmol. 43:S151–161, 1999.

14. Endres, M., Fan, G., Hirt, L., Fujii, M., Matsushita, K., Liu, X., Jaenisch, R., and Moskowitz,M.A. Ischemic brain damage in mice after selectively modifying BDNF or NT4 gene expres-sion. J. Cereb. Blood Flow Metab. 20:139–144, 2000.

15. Carlsson, A.M., Chauhan, B.C., Lee, A.A., and LeBlanc, R.P. The effect of brimonidine tartrateon retinal blood flow in patients with ocular hypertension. Am. J. Ophthalmol. 129:297–301,2000.

16. Bhandari, A., Cioffi, G.A., Van Buskirk, E.M., Orgul, S., and Wang, L. Effect of brimonidineon optic nerve blood flow in rabbits. Am. J. Ophthalmol. 128:601–605, 1999.

17. Lachkar, Y., Migdal, C., and Dhanjil, S. Effect of brimonidine tartrate on ocular hemodynamicmeasurements. Arch. Ophthalmol. 116:1591–1594, 1998.

18. Harris, A., Joos, K., Kay, M., Evans, D., Shetty, R., Sponsel, W.E., and Martin, B. Acute IOPelevation with scleral suction: effects on retrobulbar haemodynamics. Br. J. Ophthalmol.80:1055–1059, 1996.

19. Matsuo, T. and Cynader, M.S. Localization of alpha-2 adrenergic receptors in the human eye.Ophthalmic Res. 24:213–219, 1992.

20. Spada, C.S., Nieves, A.L., Burke, J.A., and Woodward, D.F. Comparative effects of brimoni-dine, p-aminoclonidine and clonidine on arteriolar caliber in human retinal tissue. Invest. Oph-thalmol. Vis. Sci. 36:S1041 (Abstract 4817), 1995.

Received: September 29, 2000Accepted for Publication: December 11, 2000

Reprint Requests: Alon Harris, M.S., Ph.D.Department of OphthalmologyIndiana University School of Medicine702 Rotary CircleIndianapolis, Indiana 46202 U.S.A.E-mail: alharris@indiana.edu

205