Chapter 45 Aqueous Humor Dynamics J. CAMERON MILLAR, B'ANN TRUE GABELT and PAUL L. KAUFMAN Table Of Contents |
AQUEOUS HUMOR FORMATION AND COMPOSITION AQUEOUS HUMOR DRAINAGE OUTFLOW BIOMECHANICS REFERENCES |
In the healthy eye, flow of aqueous humor against resistance generates
an intraocular pressure of about 15 mmHg, which is necessary for the proper
shape and optical properties of the globe.1 The circulating aqueous humor nourishes the cornea and lens, structures
which must be transparent and therefore devoid of blood vessels;1 the aqueous also provides a transparent and colorless medium of refractive
index 1.3333 between them,2 and thus is an important component of the eye's optical system.1 The basic anatomy of the primate anterior ocular segment and the normal
pathways of aqueous humor flow are shown schematically in Figures 1 and 2.3 Aqueous humor is secreted by the ciliary epithelium lining the ciliary
processes (by active ionic transport across the ciliary epithelium and
hydrostatic and osmotic gradients between the posterior chamber and
the ciliary process vasculature and stroma). It enters the posterior chamber
and then flows around the lens and through the pupil into the anterior
chamber, from which it leaves the eye by passive bulk flow through
two pathways at the anterior chamber angle: (1) through the trabecular
meshwork (TM), across the inner wall of Schlemm's canal and
into its lumen, and then into collector channels, aqueous veins, and
the general episcleral venous circulation—the trabecular or conventional
route; and (2) across the iris root, uveal meshwork, and anterior
face of the ciliary muscle, through the connective tissue between
the muscle bundles, and the suprachoroidal space, and then out through
the sclera—the uveoscleral, posterior, or unconventional route. In
certain monkey species, total aqueous drainage is relatively evenly
divided between the two pathways.4–7 In the normal human eye, the importance of the uveoscleral pathway has
not been well determined. In elderly eyes with posterior-segment tumors, it
accounts for about 10% of total aqueous humor drainage, but in
young individuals it may account for more than 30%.8–10 There is no significant net fluid movement across the cornea, iris vasculature, or
vitreoretinal interface, although ion fluxes exist.11,12 Because of autonomic innervation and the receptors of the relevant structures, adrenergic and cholinergic mechanisms play major roles in aqueous humor formation and drainage in terms of both normal physiology and glaucoma therapeutics13–32; evidence for other mechanisms, including serotonergic,33–36 dopaminergic,37–46 adenosinergic,47–52 and prostaglandinergic mechanisms 6,53–67 is growing. |
AQUEOUS HUMOR FORMATION AND COMPOSITION |
PHYSIOLOGY Until the early twentieth century, aqueous humor was regarded as a stagnant fluid.68 Since that time, however, it has been shown to be continuously formed and drained,69 and the associated anatomic drainage portals (Schlemm's canal, collector channels, aqueous veins, and ciliary muscle interstices) have been described.70–73 Three physiologic processes contribute to the formation and chemical composition of the aqueous humor1: diffusion, ultrafiltration (and related dialysis), and active secretion. The first two processes are passive and therefore require no active cellular participation. Diffusion of solutes across cell membranes occurs down a concentration gradient, and substances with high lipid solubility coefficients that can easily penetrate biological membranes move readily in this way. Ultrafiltration is the term used to describe the bulk flow of blood plasma across the fenestrated ciliary capillary endothelia into the ciliary stroma; it can be increased by augmentation of the hydrostatic driving force (see Fig. 2). These processes (diffusion and ultrafiltration) are responsible for the formation of the “reservoir” of the plasma ultrafiltrate in the stroma, from which the posterior chamber aqueous is derived through active secretion across the ciliary epithelium. Active secretion requires energy, normally provided by the hydrolysis of adenosine triphosphate (ATP). The energy is used to secrete substances against a concentration gradient. Energy-dependent active transport of sodium into the posterior chamber by the nonpigmented ciliary epithelium (NPE) (see Fig. 2) results in water movement from the stromal pool into the posterior chamber. Although controversy has surrounded the relative quantitative roles of ultrafiltration and active secretion, it seems fairly certain that under normal conditions active secretion accounts for perhaps 80% to 90% of total aqueous humor formation.1,11,74–81 The observation that moderate alterations in systemic blood pressure and ciliary process blood flow have little effect on aqueous formation rate supports this notion.11,74,82 Moreover, Bill noted that the hydrostatic and oncotic forces that exist across the ciliary epithelium—posterior aqueous interface favor resorption, not secretion, of aqueous humor.74 Active secretion is essentially pressure-insensitive at near-physiologic intraocular pressure (IOP). However, the ultrafiltration component of aqueous humor formation is sensitive to changes in IOP, decreasing with increasing IOP. This phenomenon is quantifiable and is termed facility of inflow or pseudofacility (Cps)—the latter because a pressure-induced decrease in inflow appears as an increase in outflow when techniques such as tonography and constant-pressure perfusion are used to measure outflow facility.1,83–89 Cps quantities in monkeys87,90 and humans91 are approximately 0.02 μl and 0.06 μl × min-1 × mmHg-1, respectively, although the latter may be an overestimate. In most mammalian species, the turnover constant of the anterior chamber aqueous humor is approximately 0.01 × min-1, that is, the rate of aqueous humor formation and drainage is about 1% of the anterior chamber volume per minute.76,92–96 This is true also in the normal human eye, in which the aqueous formation rate is approximately 2.0 to 2.5 μl × min-1.1,8,97–101 More comprehensive theoretic analyses of the fluid mechanics of aqueous production can be found elsewhere.84,99,102–105 BIOCHEMISTRY The active process of aqueous secretion is mediated through selective transport of certain ions and substances across the basolateral membrane of the NPE against a concentration gradient. Two enzymes abundantly present in the NPE are intimately involved in this process: sodium-potassium-activated adenosine triphosphatase (Na+ -K+ -ATPase) and carbonic anhydrase (CA).78,106–117 Na+ -K+ -ATPase is found predominantly bound to the plasma membrane of the basolateral infoldings of the NPE. 78,108–111,113,114116 The enzyme provides the energy for the metabolic pump, which transports sodium into the posterior chamber, by catalyzing the reaction ATP → ADP + Pi + energy,118,119 where ATP = adenosine triphosphate, ADP = adenosine diphosphate, and Pi = inorganic phosphate. As a result of active transport, aqueous humor in humans exhibits greater levels of ascorbate, some amino acids, and certain ions such as Cl-.99,101 There is also a passive transporter for HCO3-.1,120 A summary of the biochemistry of aqueous secretion is shown in Figure 3. Inhibition of the ciliary process Na+ -K+ -ATPase by cardiac glycosides (e.g., ouabain) or vanadate (VO3-, VO43-) significantly reduces the rate of aqueous humor formation and consequently IOP in experimental animals108,113,119,121–124 and humans.123 Cardiac glycosides appear to act at the extracellular aspect of the membrane-bound enzyme, but vanadate acts at the cytoplasmic surface.124,125 Cardiac glycosides given topically are ineffective as ocular hypotensives and may cause corneal edema by interfering with the Na+ -K+ -ATPase—dependent sodium pump in the corneal endothelium.121 Intravitreal and systemic administration are effective but carry unacceptable ocular and cardiovascular risks, respectively.121,126 Vanadate is effective topically in both rabbits and monkeys, apparently without producing acute corneal edema.119,122,123 CA is abundantly present in erythrocytes, renal tubules, and the basal and lateral membranes and cytoplasm of the pigmented epithelium and NPE of the ciliary processes.80,127–132 CA catalyzes reaction I of the sequence CO2 + H2O →I H2CO3 →II H+ + HCO3- Reaction II is a spontaneous, virtually instantaneous ionic dissociation.118,131 The actual sequence may be far more complex, involving energy-dependent separation of H+ and OH- at membrane boundaries within the NPE cell, and formation of HCO3- by CA-catalyzed association of OH- with CO2.80,133 The reaction sequence shown above provides the HCO3-, which is essential for the active secretion of aqueous humor. Although the exact roles of CA and HCO3- are still debated,76,78,80,120,133–136 it has been demonstrated that inhibition of the production of HCO3- also leads to an inhibition of the active transport of Na+ across the NPE into the first-formed aqueous, thereby reducing active aqueous humor formation (AHF). Several hypotheses explain the relationship between reduction in NPE intracellular HCO3- and inhibition of Na+ transport,1 including: (1) inhibition of CA causes a decrease in HCO3- available for transport with Na+ from the cytosol of the NPE to the aqueous, which is required to maintain electroneutrality; (2) reduction in intracellular pH inhibits Na+ -K+ -ATPase, and (3) decreased availability of H+ produced by reaction II decreases H+ /Na+ exchange and reduces the availability of intracellular Na+ for transport into the intercellular channel. In addition, inhibition of renal and erythrocyte CA leads to a systemic acidosis that promotes inhibition of AHF.137 CA inhibitors (e.g., acetazolamide,138 methoxazolamide, ethoxzolamide, dichlorphenamide,129,139 aminozolamide,140 trifluormethazolamide141) given systemically can reduce secretion by as much as 50%127–129,132,139–148 and have been in use for clinical glaucoma therapy for more than 40 years.135,136,139,149 With low doses of certain CA inhibitors (e.g., methazolamide), it is possible to inhibit the ocular enzyme without affecting the renal and erythrocyte enzymes, thus producing submaximal secretory suppression without systemic acidosis.80,129,139,150,151 It was once thought that the drug concentration at the ciliary epithelium required to produce the almost continuous, total ciliary CA inhibition necessary to achieve adequate and sustained reduction in AHF (over 99% of the ciliary enzyme must be inhibited to achieve significant secretory suppression) might never be attainable through the eye-drop route,80,129,136,139,151, but the topically effective CA inhibitor dorzolamide,146,147,152–155 which may achieve nearly the same reduction in IOP as the earlier CA inhibitors but without their systemic side effects, is currently available.137 AQUEOUS HUMOR COMPOSITION The composition of aqueous humor differs from that of plasma as a result of two important physiological characteristics of the anterior segment: a mechanical epithelial/endothelial blood-aqueous barrier, and active transport of various organic and inorganic substances by the ciliary epithelium. The greatest differences are the low protein and high ascorbate concentrations in the aqueous relative to plasma (about 200 times less and 20 times greater, respectively).156–162 The high ascorbate concentration may help protect the anterior ocular structures from ultraviolet light-induced oxidative damage. When the aqueous protein concentration rises significantly above its normal level of approximately 20 mg/100 ml,76 as in uveitis, the resultant light scattering (Tyndall effect) makes visible the slit-lamp beam as it traverses the anterior chamber (a phenomenon known as “flare”). Lactate also is normally in excess in the aqueous, presumably as a result of glycolytic activity of the lens, cornea, and other ocular structures.76,143 Other compounds or ions in excess in the aqueous relative to the plasma are Cl- and certain amino acids.101 BLOOD-AQUEOUS BARRIER The blood-aqueous barrier (BAB)is a functional concept, rather than a discrete structure, invoked to explain the degree to which various solutes are relatively restricted in travel from the ocular vasculature into the aqueous humor. The capillaries of the ciliary processes and choroid are fenestrated, but the interdigitating surfaces of the retinal pigment epithelia and the ciliary process NPE respectively are joined to each other by tight junctions (zonulae occludens) and constitute an effective barrier to intermediate- and high-molecular weight substances, such as proteins.163–173 The endothelia of the inner wall of Schlemm's canal are similarly joined,166,174 preventing retrograde movement of solutes and fluid from the canal lumen into the TM and anterior chamber. The iris and retina have no similar epithelium between their vasculature and the ocular fluids, but their capillaries are of the nonfenestrated, impermeable type.11,76,101,165,168 For present purposes, one may say that the BAB comprises the tight junctions of the ciliary process NPE, the inner wall endothelium of Schlemm's canal, the iris vasculature, and the outward-directed active transport systems of the ciliary processes. A more universal concept of the BAB must explain the movement of smaller molecules, lipid-soluble substances, and water into the eye.76 With disease-, drug-, or trauma-induced breakdown of the BAB (Table 1), plasma components enter the aqueous humor. Net fluid movement from blood to aqueous increases, but so does its IOP-dependence (Cps).175 Total facility, as measured by IOP-altering techniques, cannot distinguish Cps from total outflow facility (Ctot) and therefore erroneously records the Cps component as increased Ctot (hence the term, ‘pseudofacility’) and underestimates the extent to which the outflow pathways have been compromised by the insult. Under these circumstances, increased Cps provides some protection against a precipitous rise in IOP; as IOP rises, aqueous inflow by ultrafiltration is partly suppressed, blunting (but not completely suppressing)88,176 further IOP elevation. Additionally, the inflammatory process that occurs during BAB breakdown leads to a reduction in active secretion of aqueous humor, possibly through interference with active transport mechanisms.177 This in turn may actually produce ocular hypotony, despite compromised outflow pathways (because of plasma protein blockage of the TM). Prostaglandin release during inflammation may contribute to the hypotony by increasing aqueous outflow through the uveoscleral route.178,179 When the noxious stimulus is removed, however, the ciliary body may recover before the TM, and the resulting normalization of AHF rate in the face of still-compromised outflow pathways leads to elevated IOP, as seen from the modified Goldmann equation IOP = [(F - U)/Ctrab] + Pe, where F = aqueous humor flow, Ctrab = facility of outflow from the anterior chamber through the trabecular meshwork and Schlemm's canal, IOP = intraocular pressure, Pe = episcleral venous pressure (the pressure against which fluid leaving the anterior chamber through the trabecular-canalicular route must drain), and U = uveoscleral outflow.180
TABLE 45-1. Factors Interrupting the Blood-Aqueous Barrier Traumatic Paracentesis Physical X-ray Chemical Alkali Pathophysiologic Vasodilation Histamine Corneal and intraocular infections Intraocular inflammation Prostaglandins Anterior segment ischemia Pharmacologic Melanocyte-stimulating hormone Nitrogen mustard Cholinergic drugs, especially cholinesterase inhibitors Plasma hyperosmolality *(Millar C, Kaufman PL: Aqueous humor: Secretion and dynamics. In Duane's Foundation of Clinical Ophthalmology, p 1. Vol 2. Philadelphia: Lippincott-Raven, 1995)
ACTIVE TRANSPORT The ciliary processes possess the ability to actively transport (or exclude) a variety of organic and inorganic compounds and ions from the eye, that is, to move them from the aqueous or vitreous to the blood against a concentration gradient. Para-aminohippurate (PAH), diodrast, and penicillin are examples of large anions that are actively transported out of the eye. These systems are similar to those in the renal tubules and satisfy all the criteria for active transport, including saturability, energy and temperature dependence, Michaelis-Menten kinetics, and inhibition by ouabain and probenecid among others.101,181–192 In addition, another system actively excretes injected iodide from the aqueous, resembling iodide transport in the thyroid and salivary glands.187 The physiologic role of these outward-directed systems is unknown. With the discovery that prostaglandins may be actively transported out of the eye,193,194 some workers have suggested that such outward-directed mechanisms may rid the eye of biologically active substances that are no longer needed or may even be detrimental.186,188,195–197 Other outwardly-directed ion-uptake mechanisms are present in the eye. The anterior uvea of the rabbit eye, for example, accumulates the anions cholate, glycocholate, deoxycholate, chenodeoxycholate, iodipamide, and o-iodohippurate.181,185 At least one outwardly-directed cationic pump also has been reported; iris-ciliary body (ICB) preparations accumulate the cation emepronium,198 although one report199 questioned whether any other cations are actively eliminated from the eye. Bárány,183 making an analogy to the ion pump located at the renal peritubular cell border adjacent to the blood, which pumps simple cations from the blood to the kidney tubule, investigated whether there are any inwardly-directed complex cation pump systems from blood to aqueous, but concluded that such mechanisms probably do not exist, at least in the rabbit. All of these transport systems (inward- and outward-directed) are thought to be located in the NPE.76 PHARMACOLOGY AND REGULATION Sympathetic and parasympathetic nerve terminals are present in the ciliary body200–204 and arise from branches of the long and short posterior ciliary nerves. These nerve fibers are of both myelinated and nonmyelinated variety. Parasympathetic fibers originate in the Edinger-Westphal nucleus of the third cranial nerve, run with the inferior division of this nerve in the orbit, and synapse in the ciliary ganglion.205 Sympathetic fibers synapse in the superior cervical ganglion and are distributed to the muscles and blood vessels of the ciliary body. Numerous unmyelinated nerve fibers surround the stromal vessels of the ciliary processes; these are most likely noradrenergic and subserve vasomotion.76 Sensory fibers arise from the ophthalmic division of the trigeminal nerve and enter the ciliary body, but their distribution and function have not been studied thoroughly. No innervation of the ciliary epithelium has been found anatomically,166 but stimulation of the ciliary ganglion leads to an increase in AHF in the enucleated arterially perfused cat eye,206 suggesting that neurotransmitters released in the ciliary stroma might diffuse toward the epithelia. Cholinergic Mechanisms The effects of cholinergic drugs on AHF and composition and on the BAB are unclear. In general, cholinergic drugs cause vasodilation207–209 in the anterior segment,210–213 resulting in increased blood flow to the choroid, iris, ciliary processes, and ciliary muscle.210,211,214 However, cholinergic drugs also may promote vasoconstriction—in rat coronary arteries208 and rat outer descending vasa recta perfused in vitro,215 for example, and in the rabbit eye.216 These responses are mediated by muscarinic receptors in the anterior uveal arterioles,210 perhaps associated with facial parasympathetic nerve terminals.217,218 Congestion in the iris and ciliary body is a well-recognized clinical side effect of topical cholinomimetics, especially the anticholinesterases.219 The presence of flare and cells in the aqueous humor seen with biomicroscopy indicates that these agents also can cause breakdown of the BAB and perhaps frank inflammation. 219 Pilocarpine increases BAB permeability to iodide220 and inulin.221 Cholinergic drug-induced vasodilation may cause a loss of tight junctions in anterior uveal blood vessels, perhaps contributing to BAB breakdown.222 Cholinergic drugs may alter the aqueous humor concentration of inorganic ions223 and the movement of certain amino acids from the blood into the aqueous humor and may also influence the outward-directed transport systems of the ciliary processes.224,225 Under certain conditions, pilocarpine may increase Cps.84,91 With a variety of species, conditions, and experimental techniques, cholinergic agents or parasympathetic nerve stimulation has been reported to increase, decrease, or not alter the AHF rate and to increase slightly the episcleral venous pressure.91,206,225–239 These apparently confusing results may indicate that cholinergic drug effects on these parameters are extremely dependent on species- and technique-related factors and on the ambient neurovascular milieu. In any event, the effects on the rate of AHF and episcleral venous pressure are surely minor in most instances, and not responsible for the drug-induced decrease in IOP that forms the basis of pilocarpine's therapeutic efficacy in chronic glaucoma; the latter resides in its ability to decrease outflow resistance through its effect on the ciliary muscle. Adrenergic Mechanisms The precise role and receptor specificity of adrenergic mechanisms in regulating the rate of AHF are unclear. At one time it was generally believed that long-term topical administration of epinephrine, a combined α1-, β1-, β2-adrenergic agonist, would decrease the rate of AHF.240 This effect was thought to be mediated by β-adrenergic receptors in the NPE, through activation of a membrane adenylate cyclase,115,241–243 although the consequent biochemical events were (and still are) unknown. In support of these observations, activation of ocular adenylate cyclase through its Gs protein in the rabbit by close arterial infusion of cholera toxin decreases AHF and lowers IOP.241 Further, forskolin, a naturally occurring diterpene derivative of the coleus plant (Coleus forskohlii) that directly and irreversibly activates intracellular adenylate cyclase in some244–249 but not all250 studies, decreases the rate of AHF when given topically or intravitreally. Although vascular phenomena could theoretically be involved in the secretion and pressure responses, these observations are consistent with the primary epithelial action of these drugs. The relationship between ocular vascular events and IOP modulation, if any, has yet to be established.103,251–256 Recent fluorophotometric studies have shown that short-term topical administration of epinephrine increases AHF8,13,29; studies with other adrenergic agonists, including salbutamol,257 isoproterenol (isoprenaline),258 and terbutaline,100 have supported this finding and are consistent with many studies showing that β-adrenergic antagonists unequivocally decrease AHF.14,18–21,23–27,29,31,32,97 The ocular hypotensive action of β-antagonists has led to their becoming mainstays of clinical glaucoma therapy; they include nonselective β1, β2 antagonists timolol,20,21,29,97 levobunolol259, and metipranolol;24 the nonselective β1, β2 partial agonist carteolol14; and the relatively selective β1 antagonist betaxolol.31 Adrenergic receptors in the ciliary epithelium are of the β2 subtype,260–263 but antagonists that are relatively selective for β1 receptors (e.g., betaxolol) are effective (although less potent and efficacious)264,265 in suppressing AHF.31,266–269 However, the apparent β1 efficacy may be related to a sufficiently high concentration reaching the ciliary epithelium so that nonselective blockade of β2 receptors may occur. Whether β antagonists suppress AHF through their effect on ciliary epithelial β receptors has been questioned.119,233,241,244–246,248–250,270–280 There is evidence that classical β-blockade may not be involved, and that other receptor types such as 5-HT1A, may be relevant (see later).33,35,36 Furthermore, inhibition of Na/K/2Cl cotransport by β2 receptor antagonists can eliminate the increase in AHF stimulated by epinephrine and isoproterenol.281,282 AHF is reduced by nearly 50% during sleep,76,101 which is comparable to the daytime reduction induced by β-adrenergic antagonists, or to the daytime or nighttime reduction induced by the CA inhibitors.124 β-antagonists produce little additional decrease in AHF during sleep98 or in pentobarbital-anesthetized monkeys.283 Because sympathetic tone is reduced during both sleep and barbiturate anesthesia,284 these findings provide further indirect evidence that β-adrenergic tone/stimulation enhances AHF but β-adrenergic blockade decreases it. Topically applied α1-adrenergic agonists and antagonists appear to have little effect on fluorophotometrically-determined AHF in the normal intact human eye,285,286 although these compounds do reduce AHF in rabbits.287–291 However, in eyes under β-adrenergic blockade, topical epinephrine acutely produces a small decrease, rather than increase, in AHF, perhaps indicating a weak α-adrenergic influence, possibly but not necessarily mediated by local vasoconstriction.292, 293 Clonidine, which has both α1-antagonist and α2-agonist properties, decreases AHF and ocular blood flow.294–297 Therefore epinephrine may have a dual effect on AHF: stimulation through β-adrenoreceptors, and inhibition through α2-adrenoreceptors.298–301 α2-Adrenergic agonists such as apraclonidine HCl (AP) and brimonidine tartrate (BR) are powerful ocular hypotensive agents when applied topically. Both AP and BR are believed to lower IOP primarily by decreasing AHF.302 There are many other ways in which AHF can be reduced pharmacologically (Table 2). Effective compounds include: the guanylate cyclase activators, atrial natriuretic factor (ANF),303 the nitrovasodilators, sodium nitroprusside,304,305 sodium azide,304 and nitroglycerin.306 8-Bromo cyclic GMP also reduces the AHF rate by 15% to 20% in the monkey.248 ANF injected intravitreally reduced IOP and AHF in rabbits and monkeys.307,308 The calcium channel antagonists verapamil and nifedipine have been reported to reduce AHF in rabbits,309 although topical diltiazem and verapamil have been reported to increase AHF in human volunteers.310 The serotonergic antagonist ketanserin reduces the AHF rate in rabbits, cats, and monkeys.311,312 The endogenous agonist, serotonin (5-HT) may also reduce AHF; topical application results in decreased AHF in the rabbit,312 but intracameral injection in the same species increases AHF.312 Serotonergic receptors of a 5-HT1A-like subtype have been reported to exist in the ICB of rabbits and humans.33, 36 It has been suggested that these receptors may be antagonized by timolol and other β-blockers. However, the precise nature of the putative 5-HT1A-like receptor subtype in the ciliary epithelium is still in question. Angiotensin converting enzyme (ACE) is present in human and rabbit aqueous humor24–26 and also in the feline ciliary body.27 Several ACE inhibitors, when topically applied to rabbit or human eyes, reduce secretion. H1-antihistamines (such as antazoline and pyrilamine) decrease AHF in rabbits,313 although the effect may be unrelated to histamine receptor binding. Δ9-tetrahydrocannabinol (a component of marijuana) may reduce AHF in humans when injected intravenously or inhaled (as in marijuana smoking).314 Metabolic inhibitors such as dinitrophenol (DNP) and fluoracetamide decrease AHF,50 as do the cardiac glycosides ouabain and digoxin, which inhibit the ciliary epithelial Na+ /K+ ATPase enzyme.315 None of these have yet been shown to have any clinical relevance for the human.
TABLE 45-2. Factors Causing Reduced Aqueous Secretion General
Various other nonpharmacologic phenomena also reduce AHF rate, including local, systemic, and surgical factors, as well as age and exercise (Table 2). With respect to age, in a study of 300 normal volunteers, ages 5 to 83 years, there was an average decline in aqueous flow of 25% between the ages of 10 and 80 years.316 |
AQUEOUS HUMOR DRAINAGE |
FLUID MECHANICS The tissues of the anterior chamber angle normally offer a certain resistance to fluid outflow. IOP builds up, in response to the inflow of aqueous humor, to the level sufficient to drive fluid across that resistance at the same rate it is produced by the ciliary body; this is the steady-state IOP. In the glaucomatous eye, this resistance is unusually high, causing elevated IOP. Understanding the factors governing normal and abnormal aqueous humor formation, aqueous humor outflow, IOP, and their interrelationships and manipulation is vital in understanding and treating glaucoma. Briefly, let F = flow (μl/min) Then: Fin = Fs + Ff At steady-state: F = Fin = Fout The simplest hydraulic model, represented by the classic Goldmann equation, views aqueous flow as passive non—energy-dependent bulk fluid movement down a pressure gradient, with aqueous leaving the eye only through the trabecular route, where ΔP = Pi - Pe, so that F = Ctrab(Pi - Pe). This relationship is correct as far as it goes, but it is vastly oversimplified. Because there is no complete endothelial layer covering the anterior surface of the ciliary body and no delimitation of the spaces between the trabecular beams and the spaces between the ciliary muscle bundles,166 fluid can pass from the chamber angle into the tissue spaces within the ciliary muscle. These spaces in turn open into the suprachoroid, from which fluid can pass through the scleral substance or the perivascular/perineural scleral spaces into the episcleral tissues. Along this route (the uveoscleral route), the fluid mixes with tissue fluid from the ciliary muscle, ciliary processes, and choroid. Thus, this flow pathway may be analagous to lymphatic drainage of tissue fluid in other organs (there are no ocular lymphatics) and provide an important means of ridding the eye of potentially toxic tissue metabolites.4,10–12,86,96,177,178,228,229,313,321–323 Flow from the anterior chamber across the TM into Schlemm's canal is pressure dependent, but drainage through the uveoscleral pathway is virtually independent of pressure at IOP levels greater than 7 to 10 mmHg.11,86,322 Although the actual drainage rates (μL/min) through the trabecular and uveoscleral routes in the monkey may be approximately equal, measured facility of uveoscleral outflow (Cu, determined by measuring Fu at two different IOP levels) is only about 0.02 μL/min/mmHg, or less than one-twentieth the facility of trabecular outflow.322 The reasons for the pressure-independence of the uveoscleral pathway are not entirely clear but might be consequent to the complex nature of the pressure and resistance relationships among the various fluid compartments within the soft intraocular tissues along the route.11 For instance, pressure in the potential suprachoroidal space (Ps) is directly dependent on IOP, such that at any IOP level, Ps is considerably but constantly less than IOP.11,324 Because the pressure gradient between the anterior chamber and suprachoroid is thus independent of IOP, bulk fluid flow between these compartments will also be IOP independent. Intraorbital pressure is such that under normal circumstances there is always a positive pressure gradient between the suprachoroidal and intraorbital spaces.11 Fluid and solutes, including large protein molecules, can thus easily exit the eye by passing through the spaces surrounding the neural or vascular scleral emissaria or through the scleral substance itself.4,96,325 At very low IOP levels, the net pressure gradient across the uveoscleral pathways is apparently so low that uveoscleral drainage decreases.86 The absence of an outflow gradient from the suprachoroid may contribute to the development of choroidal detachments seen during the ocular hypotony that sometimes follows intraocular surgery.11 Because Cps and Cu are so low compared with Ctrab under normal steady-state conditions, the hydraulics of aqueous dynamics may be reasonably approximated for clinical purposes by the equation Fin = Fout = Ctrab(Pi - Pe) + Fu. Clinically significant increases in inflow occur only in situations involving breakdown of the BAB. The pressure sensitivity of the ultrafiltration component of aqueous secretion blunts the tendency for IOP to rise under such conditions; that is, Cps is increased. Elevated episcleral venous pressure, such as may occur with arteriovenous communications resulting from congenital malformations or trauma, causes a nearly mmHg for mmHg increase in IOP.326 Pharmacologic agents may exert small and probably clinically insignificant effects on Cps, Cu, or Pe.15, 17, 84, 91, 213, 228, 229, 232, 327–329 Major, clinically relevant alterations in outflow physiology are achieved by iatrogenic manipulation of Ctrab or Fu. PHARMACOLOGY AND REGULATION Cholinergic Mechanisms CONVENTIONAL (TRABECULAR) OUTFLOW In primates the iris root inserts into the ciliary muscle and the uveal meshwork just posterior to the scleral spur, and the ciliary muscle inserts at the scleral spur and the posterior inner aspect of the TM.165,330 The influence of these two contractile, cholinergically innervated structures on resistance to aqueous humor outflow has long been a source of speculation. Voluntary accommodation (human),331 electrical stimulation of the third cranial nerve (cat),332–334 topical, intracameral, or systemically administered cholinergic agonists (monkey and human),335,336 and, in enucleated eyes (monkey and human), pushing the lens posteriorly with a plunger through a corneal fitting337 all decrease outflow resistance, but ganglionic blocking agents and cholinergic antagonists increase resistance.335,338–341 Furthermore, the resistance-decreasing effect of intravenous pilocarpine in monkeys is virtually instantaneous, implying that the effect is mediated by an arterially perfused structure or structures.342 These findings collectively suggest that iris sphincter and/or ciliary muscle contraction physically alters meshwork configuration so as to decrease resistance, but muscle relaxation deforms it so as to increase resistance.342 However, not all experimental evidence supports this strictly mechanical view of cholinergic and anticholinergic effects on meshwork function. For example, in monkeys, intravenous atropine rapidly reverses some but not all of the pilocarpine-induced resistance decrease,343,344 and topical pilocarpine causes a much greater resistance decrease per diopter of induced accommodation than does systemic pilocarpine (monkey)344 or voluntary accommodation (human).345 The inability of atropine to rapidly and completely reverse the pilocarpine-induced facility increase in normal eyes could be because of mechanical hysteresis of the meshwork; ciliary muscle contraction forces a rapid structural change on the meshwork, but muscle relaxation cannot itself reverse this change; elasticity of the meshwork is involved, and its effect may be slow.346 The variation in the relative magnitude of pilocarpine-induced accommodation and resistance decrease when the drug is administered by different routes might reflect differences in bioavailability of the drug to different regions of the muscle. The possible existence of very slowly developing (weeks or longer) primary cholinomimetic effects directly on the endothelium of the TM or Schlemm's canal is especially intriguing.343,344 Secondary mechanical effects of drugs may be distinguished from primary pharmacologic ones in the living monkey eye by totally removing the iris at its root347 and by disinserting the anterior end of the ciliary muscle over its entire circumference and retrodisplacing it to a more posterior position on the inner scleral wall.348 Total removal of the iris has no effect on IOP, resting outflow resistance, or resistance responses to intravenous or intracameral pilocarpine.349 However, following ciliary muscle disinsertion and total iris removal, there is virtually no acute resistance response to intravenous or intracameral pilocarpine348 and no response to topical pilocarpine given at 6-hour intervals for 18 to 24 hours.346 Thus it seems virtually certain that the acute resistance-decreasing action of pilocarpine, and presumably other cholinomimetics, is mediated entirely by drug-induced ciliary muscle contraction, with no direct pharmacologic effect on the meshwork itself. This is consistent with the relative paucity of cholinergic nerve endings in the meshwork, most of which are located posteriorly in proximity to the ciliary muscle insertion and are probably of no significance.204,346,350,351 Muscarinic receptors and contractile elements, however, are present in the TM. The m3 mRNA muscarinic receptor transcript was detected in human TM of cadaver eyes.352 Carbachol-induced mobilization of Ca2+ and phosphoinositide production in human TM cells in culture has been associated with the M3 muscarinic receptor.353 Pharmacologically, the functional muscarinic receptors in isolated bovine TM strips also were shown to be of the M3 subtype.354 Smooth muscle-specific contractile proteins have been discovered in cells within the human TM and adjacent to the outer wall of Schlemm's canal and the collector channels.355–358 Transformed TM cells from a glaucoma patient also demonstrated vimentin, tubulin, and smooth muscle-specific alpha-actin.359 Cultured human TM cells showed electrophysiologic responses typical for smooth muscle cells in response to endothelin-1 and cholinergic agonists.360 Isolated bovine TM stips contracted isometrically in response to carbachol, pilocarpine, aceclidine, acetylcholine, and endothelin-1.354,361,362 However, in the organ-cultured, perfused bovine anterior segment, endothelin-1- and carbachol-induced contractions resulted in an increase in resistance and a reduction of the outflow rate.363 Also, low (10-8 to 10-6 M) but not high (10-4 to 10-2 M) doses of pilocarpine, aceclidine, or carbachol induce increased outflow facility in human-perfused anterior ocular segments devoid of ciliary muscle.364,365 Light and electron microscopic studies of the trabecular meshwork and Schlemm's canal have demonstrated pilocarpine-induced alterations in the size and shape of the intertrabecular spaces and in various characteristics, including vacuolization of the endothelium of the inner canal wall.366–370 The area of empty spaces beneath the inner wall endothelium in monkeys may represent flow pathways through the meshwork, and is positively correlated with outflow facility under pilocarpine or hexamethonium.371 However, these alterations are all considered to be secondary to pilocarpine-induced ciliary muscle contraction and augmented transtrabecular outflow.372 Furthermore, we have no idea what anatomical alteration in the meshwork accounts for the ciliary muscle contraction-induced decrease in resistance to passive bulk fluid outflow. Whether opening entirely new channels, decreasing the resistance of some or all existing channels, widening Schlemm's canal, or some other alteration is critical is unknown, as is any understanding of how ciliary muscle traction on the meshwork brings about the critical change. In short, we are rather ignorant of the physics behind the physiology. At least two different subtypes of muscarinic receptors, M2 and M3, are present in the ciliary muscle.352,373,374 The M3 subtype appears to mediate the outflow facility and accommodative responses to pilocarpine and aceclidine in monkeys.375,376 The M2 receptor shows preferential localization to the longitudinal,37 putatively more facility-relevant portion of the ciliary muscle, but to date no functional role for this subtype has been elucidated. In monkeys, the outer longitudinal region of the ciliary muscle differs ultrastructurally and histochemically from the inner reticular and circular portions.377 Aceclidine increases outflow facility with little accommodative response after intracameral administration in monkey eyes376,378 or topical administration to humans.379–383 No differences were found between the contraction responses to aceclidine in the longitudinal and circular vectors of the isolated rhesus monkey ciliary muscle.384 Muscarinic receptor subtype differences also do not appear to play a role in the dissociation of the accommodative and outflow facility responses.376 Aceclidine may act directly on the TM, as evidenced by studies showing that low doses of aceclidine can enhance outflow facility in organ-cultured human eyes in vitro and in monkey eyes in vivo where the ciliary muscle has been disinserted from the scleral spur.364,365,385 UNCONVENTIONAL (UVEOSCLERAL) OUTFLOW When the ciliary muscle contracts in response to exogenous pilocarpine, the spaces between the muscle bundles are essentially obliterated.330,335 Conversely, during atropine-induced muscle relaxation, the spaces are widened.335 If mock aqueous humor containing albumin labeled with iodine-125 or iodine-131 (which under resting conditions leave the anterior chamber essentially by bulk flow through the trabecular and uveoscleral drainage routes) is perfused through the anterior chamber, autoradiographs may be made to show qualitatively the distribution of the flow.10,229 In the pilocarpinized eye, radioactivity is present in the iris stroma, the iris root, the region of Schlemm's canal and surrounding sclera, and the most anterior portion of the ciliary muscle. In the atropinized eye, radioactivity is found in all these tissues, as well as throughout the entire ciliary muscle, and even further posteriorly in the choroid/sclera.10,229 In other perfusion experiments quantifying uveoscleral drainage, pilocarpinized eyes demonstrate but a fraction of the uveoscleral flow in atropinized eyes.228,229,327 Thus, to generalize in the primate eye, pilocarpine (and presumably all cholinergic agonists) augments aqueous humor drainage through the trabecular route and diminishes drainage through the uveoscleral route. In most instances in the human, the former apparently exceeds the latter, and the net result is enhanced aqueous drainage and decreased IOP.386 In the normal monkey eye, where drainage through the trabecular and uveoscleral routes is more nearly equal, pilocarpine may sometimes induce a slight rise in IOP, perhaps by inhibiting uveoscleral drainage more than it enhances trabecular drainage.228,229 ALTERATIONS IN CHOLINERGIC SENSITIVITY OF THE OUTFLOW APPARATUS Given the long-term use of cholinergic agonists in glaucoma therapy and the vital role of ciliary muscle tone in regulating outflow resistance, it is important to note that in the monkey topical administration of the cholinesterase inhibitor echothiophate or the direct acting agonist pilocarpine can induce subsensitivity of the outflow facility and accommodative responses to pilocarpine, accompanied by decreased numbers of muscarinic receptors in the ciliary muscle.387–394 Even a single dose of pilocarpine or carbachol reduces receptor number.389 The molecular mechanisms involved may bear on potentially important clinical questions. Will patients receiving long-term cholinergic drug treatment for glaucoma eventually become refractory to therapy? How can drug-induced refractoriness be distinguished clinically from progression of the disease? Will certain agonists be more likely to induce profound refractoriness than others? Can the problem be alleviated by periodically switching from one cholinomimetic to another, or by alternating periods of cholinomimetic therapy and abstinence? Will the obvious clinical advantages of low-dose sustained-release systems over pulsed topical eye drop delivery be offset by a greater tendency to induce subsensitivity?388 Will the induced subsensitivity be as reversible in the diseased human eye as it apparently is in the healthy animal eye?387,395 Can noniatrogenic abnormalities in cholinergic systems be causally related to glaucoma? These questions will be difficult to answer, especially because the parameter of greatest clinical interest, IOP, is influenced by so many anatomical structures and physiological processes. Clinical experience to date with sustained-release pilocarpine delivery systems does not appear to consistently demonstrate a progressive loss of IOP-lowering efficacy analogous to the experimental results in the monkey, although some late therapeutic failures certainly occur.396,397 Adrenergic Mechanisms CONVENTIONAL (TRABECULAR) OUTFLOW Topical and intracameral epinephrine increase outflow facility in rabbit and primate eyes.13,15,16,17,30,314, 398–407 Much work has attempted to define the time course, type of receptors (e.g., α, β), and biochemical pathways (e.g., prostaglandins, cyclic adenosine monophosphate [cAMP]) involved in these responses. Adrenergic agonists affect smooth-muscle tone in the iris and ciliary body.408–410 CM receptors are of the β2-subtype.411 Adrenergic receptor stimulation may alter intraocular, intrascleral, and extrascleral vascular tone, as well as have possible direct effects on the endothelium lining the outflow pathways, all of which may alter facility. These potential sites of action are not mutually exclusive, and indeed this may account for much of the variability and confusion in the literature. The facility increases and dose-response relationships for epinephrine and norepinephrine are virtually identical in surgically untouched and totally iridectomized monkey eyes, and in eyes with the ciliary muscle disinserted from the TM and scleral spur, indicating that neither the iris nor the ciliary muscle is involved in the responses.412,413 It has been proposed that the facility-increasing action of adrenergic agonists is related primarily to their effects on intrascleral and extrascleral vasculature.414 However, intracameral infusion of the noncatecholamine vasoconstrictors ergotamine and angiotensin II, and the vasoactive agents histamine, serotonin, and bradykinin, which may have either vasoconstricting or vasodilating effects depending on the particular vascular bed, decreased facility in surgically-untouched, aniridic, and ciliary muscle-disinserted eyes.412,413,415 Although various vascular beds may react differently to any given vasoconstrictor or vasodilator, these results did not support the contention that the facility-increasing effects of catecholamines such as epinephrine (a vasoconstrictor or vasodilator, depending on ambient condition of vascular tone) and norepinephrine (a vasoconstrictor) are a result of their vascular actions. In surgically untouched, aniridic, and ciliary muscle-disinserted monkey eyes with widely varying starting facilities, epinephrine and norepinephrine increased facility by a constant percentage of the starting facility, indicating that the drugs exert their effects on whatever is responsible for the major part of the variation in starting facility. Attributing entirely to pseudofacility or facility of uveoscleral routes the constant percentage increase in facility observed in eyes with starting facilities varying over a four- to five-fold range would require that virtually all of the starting facility be pseudofacility or uveoscleral facility,412,413 which of course is not so.15,322,416 Similarly, because pilocarpine exerts its resistance-decreasing effect by way of ciliary muscle traction on the TM,348 and because in the normal monkey eye pilocarpine markedly decreases and ganglionic blockade with hexamethonium markedly increases both outflow resistance and its interindividual variability,417 most of the variability in starting resistance must reside at sites other than the sclera.412,413 This suggests as a possible site the TM/inner wall of Schlemm's canal. Epinephrine and norepinephrine exert their action on whatever characteristic of the meshwork/canal accounts for interindividual variability in starting facility. One possibility would be an increase in the hydraulic conductivity per unit filtering area.412,413 Biochemical evidence also points to the meshwork as the target tissue. It appears that the facility-increasing effect of epinephrine and norepinephrine is mediated by β2-adrenergic receptors on the trabecular endothelial cells, and the subsequent G-protein-adenylate cyclase-cyclic AMP cascade.104 The evidence for this is as follows: trabecular cells in culture express β-adrenergic receptors (as measured by 125I-hydroxybenzylpindolol [HYP] binding), which have been characterized through competition studies as being of the β2 subtype.418 Trabecular outflow facility increases in response to β-adrenergic agonists.8,13,17, 419 Trabecular cells synthesize cyclic AMP in response to stimulation with β-adrenoceptor-selective agonists.30,242,314 The increase in cyclic AMP synthesis by TM cells in response to epinephrine can be blocked by timolol,30 although not by betaxolol (a β1-receptor antagonist), which is consistent with the hypothesis that there are only β2-receptors present in the TM. Topically applied adrenergic agonists elevate aqueous humor cAMP levels, and intracameral injection of cAMP or its analogs (but not the inactive metabolite 5'AMP) lowers IOP and increases outflow facility.242,405,420–422 The cAMP-induced facility increase is not additive to that induced by adrenergic agonists, and vice-versa.423 Epinephrine increases facility and perfusate cyclic AMP levels in the organ-cultured perfused human anterior segment, effects that are blocked by timolol and the selective β2-antagonist ICI 118,551.424 Thus, the adrenergic agonist-induced facility increase seems to be mediated through the adenylate cyclase-cAMP pathway. The facility-increasing effect of epinephrine is blocked by timolol425,426 but not betaxolol,265,427,428 in both humans and monkeys, which is consistent with the hypothesis that there are no β1 receptors present in the primate TM. Adrenergic innervation of the primate TM is sparse and concentrated mainly in the region of the meshwork near the ciliary muscle tendons. No functional significance can as yet be ascribed to these terminals.201,350,351 However, even noninnervated cells may express autonomic receptors.429–432 The nature of the physiologic change in the meshwork responsible for the decreased flow resistance remains uncertain. Recent studies suggest an epinephrine-induced disruption of actin filaments within the TM cells, consequent alterations in cell shape, and cell-cell and cell-extracellular matrix (C-ECM) adhesions within the meshwork, resulting in altered meshwork geometry and increased hydraulic conductivity across the meshwork. Thus cytochalasin B (a disruptor of actin filament formation) potentiates the facility-increasing effect of epinephrine,433 but phalloidin (a stabilizer of actin filaments), inhibits it.434 Continuous exposure to epinephrine at a concentration of 10 μM produced arrest of normal cytokinetic cell movements, inhibition of mitotic and phagocytic activity, marked cell retraction, separation from the substrate, and cellular degeneration after 4 to 5 days in cultured human trabecular cells.435 Similarly, the hydraulic conductivity of trabecular cell monolayer cultures grown on filters was increased by epinephrine and was associated with changes in cell shape and with separation between cells.419 These actions of epinephrine were all partly blocked by pretreatment with timolol. Epinephrine is known to alter cell adhesion and the actin cytoskeleton in other cell systems. Epinephrine co-stimulates tyrosine phosphorylation of pp 125FAK (focal adhesion kinase, a tyrosine kinase involved in the formation of C-ECM adherens junctions) in human platelets436; induces a dose-dependent decrease in macrophage spreading associated with changes in the distribution of F-actin437; and promotes detachment of natural killer cells from umbilical cord vein endothelial cells438; these effects are all likely mediated by β2-adrenergic receptors acting through a cAMP-dependent mechanism. cAMP itself promotes F-actin disassembly and inhibits actin polymerization in macrophages under certain conditions.439 It is hoped that the coming years will see the integration of physiologic, pharmacologic, biochemical, morphologic, and cell biologic data into a comprehensive scheme of adrenergic modulation of aqueous outflow. UNCONVENTIONAL (UVEOSCLERAL) OUTFLOW β-Adrenergic receptors are present in the primate ciliary muscle, and their physiologic or pharmacologic stimulation relaxes the muscle.408–410,440 Epinephrine, in addition to increasing trabecular outflow, also increases uveoscleral outflow in monkeys16 and humans.8,29 The mechanism is unknown, but may be in part caused by the mildly relaxant effect of epinephrine on the ciliary muscle, which presumably acts by means of its β-adrenergic receptors.200,408–410 However, adrenergic agonists also stimulate prostaglandin biosynthesis in several tissues, including rabbit441 and bovine442 iris. Pretreatment with the cyclo-oxygenase inhibitor indomethacin inhibits the ocular hypotensive effect of topically applied epinephrine in rabbits443 and humans,444 suggesting that the IOP-lowering action of epinephrine may be mediated at least in part by prostaglandins or other cyclo-oxygenase products.443,444 Numerous clinical studies in humans claim that topical application of timolol, a nonselective β1-, β2-adrenergic receptor antagonist, induces no change in distance refraction.240 However, a single topical application of 0.5% timolol may increase myopia by nearly 1 diopter, presumably by blocking the effect of endogenous ciliary muscle-relaxing, sympathetic neuronal tone.445 Indirect fluorophotometric estimates have failed to demonstrate any effect of timolol on uveoscleral outflow per se.29 However, timolol may reduce epinephrine-induced increases in uveoscleral outflow when the two drugs are applied concurrently.29 These findings are consistent with the data concerning adrenergic influences on ciliary muscle contractile tone and also illustrate the importance of the ambient neuronal and pharmacologic adrenergic tone in determining the response of a target tissue to an exogenous adrenergic agent. In addition to suppressing aqueous formation, α2-adrenergic agonists may enhance uveoscleral outflow. Brimonidine apparently increased uveoscleral outflow in rabbits in the treated eye and lowered IOP in both the treated and contralateral eyes,446 although measurements of uveoscleral flow in the rabbit are difficult to interpret because of the substantially different anatomy of the rabbit outflow apparatus compared with the monkey.447 In humans apraclonidine- and brimonidine-induced IOP reductions are associated with decreased AHF and decreased (apraclonidine) or increased (brimonidine) uveoscleral outflow.448,449 Prostaglandin Mechanisms Nilsson and colleagues6 reported a 60% increase in aqueous outflow through the uveoscleral pathway in monkeys after a single submaximal dose of prostaglandin F2a-1-isopropyl ester (PGF2a-IE). After multiple submaximal dosing, there was a greater than 100% increase.57 In both instances, aqueous outflow was substantially re-directed from the trabecular to the uveoscleral pathway. The effect is consequent to an initial PGF2a-induced relaxation of the ciliary muscle (probably consequent to PGF2a action at specific prostanoid receptors), and subsequent dissolution of type I and type III collagen in the connective tissue-filled spaces between the muscle bundles (probably consequent to PGF2a-induced activation of transcription/translation of ciliary muscle fiber genes encoding stromolysin and collagenase enzymes).57,64,450 PG-induced increased turnover and remodeling of extracellular matrix also has been demonstrated in ciliary muscle cells in vitro, possibly associated with activation of the proto-oncogene c-fos.304,451–453 This system likely evolved to protect the eye in several ways during inflammation. The TM may become compromised by inflammation or obstructed by inflammatory debris, and the choroid may be overloaded with debris and extravasated proteins that must be removed from the eye.454 In this situation, prostaglandins would be released and, as autocoids or hormones that are synthesized, released, and locally acting, would induce the changes described. Because the eye has no lymphatics, uveoscleral outflow may serve as an analogue to an intraocular lymphatic drainage system.11 The normal low flow rate that is sufficient to remove normal levels of extravascular protein may be inadequate when protein levels are increased as in uveitis. Redirection of aqueous outflow from the trabecular to the uveoscleral pathway would both rid the eye of excess proteins and maintain physiologic IOP. This could also explain the very low IOP that often accompanies uveitis; during experimental iridocyclitis in monkeys, uveoscleral outflow is increased approximately four-fold.455 The increase in uveoscleral outflow in response to these compounds is so great that a larger reduction in IOP is possible than with any other known substance. It has yet to be established whether or not endogenous prostaglandins have a physiologic or only a pathophysiologic role in regulating uveoscleral flow. Nonetheless, PGF2a analogues and metabolites show great promise as clinically useful ocular hypotensive agents, despite some undesirable side effects (conjunctival foreign body sensation, conjunctival hyperemia, stinging pain, photophobia,58,456 and altered iris pigmentation in some instances457).54,66,458–464 Corticosteroid Mechanisms Topical or systemic glucocorticoids may induce elevation of IOP in susceptible persons,465–468 consequent to decreased outflow facility. Recent preliminary evidence suggests that glucocorticoids may play a major role in the normal physiologic regulation of outflow facility and intraocular pressure. Although glucocorticoid receptors have been identified in the cells of the outflow pathways,469–471 the biochemical and consequent physical processes causing the facility decrease are unclear. However, the mechanism may be consequent to modulation of macromolecular metabolism, or prostaglandin/adrenergic interactions involving the outflow system. For example, the elevated IOP of albino rabbits, caused by the administration of either dexamethasone or dexamethasone plus 5β-dihydrocortisol, is reduced after 3 to 7 days of dosing with 0.1% 3α,5β-tetrahydrocortisol, a cortisol metabolite. This metabolite has no effect on the IOP of untreated, normotensive rabbits472 or on IOP or outflow facility in normotensive monkeys.306 The oral administration of the glucocorticoid biosynthesis inhibitor metyrapone in glaucoma patients has been observed to elicit a small, transient decline in IOP.473 A generalized cellular hypersensitivity to glucocorticoids is not intrinsic to primary open-angle glaucoma.474 Glaucoma patients have increased plasma levels of cortisol compared with normal individuals.475,476 The glucocorticoid effect on the physiology of the outflow pathways may be more rapid than previously believed—certainly less than the 3 to 6 weeks classically described after administration of topical eye drops477—and perhaps as rapid as a few hours. Dexamethasone alters complex carbohydrate, hyaluronic acid, protein, and collagen synthesis and distribution in the cells and tissues of the rabbit and human aqueous humor outflow systems.477–485 Prostaglandins are produced by human trabecular endothelial cells in culture; dexamethasone inhibits trabecular cell prostaglandin synthesis by up to 90%. In cultured human trabecular cells, dexamethasone reduces phagocytic486,487 and extracellular protease activity, changes gene expression,488–490 and increases nuclear DNA content.470 In cultured TM cells obtained from patients with primary open-angle glaucoma, cortisol metabolism is altered. These cells accumulate 5β-dihydrocortisol and, to a lesser extent, 5α-dihydrocortisol metabolites, which were not found in cells derived from normal individuals. This difference is because of a marked increase (more than 100-fold) in Δ4-reductase activity and to a decrease (more than 4-fold) in 3-oxidoreductase activity, enzymes responsible for the catalysis of the following reactions, I and II, respectively: In cultured TM cells obtained from normal human eyes, there is no accumulation of the active dihydrocortisol intermediates; all cortisol is rapidly metabolized to the inactive tetrahydrocortisols. Peripheral lymphocytes from glaucoma patients did not show these abnormalites, indicating that the defects are not found in all cortisol-metabolizing cells.491,492 3α,5β-Tetrahydrocortisol applied topically decreased IOP and increased outflow facility in glaucomatous human eyes,493 and antagonized dexamethasone-induced cytoskeletal reorganization in normal human-cultured TM cells493,494 but had no effect on outflow facility in normotensive monkeys after intracameral injection or 10 days of topical administration.306 Interestingly, cultured TM cells from patients with primary open-angle glaucoma (POAG) metabolized cortisol predominantly to 5β-dihydrocortisol, which potentiates the facility-decreasing and IOP-increasing effects of dexamethasone; these cells produce relatively little 3α,5β-tetrahydrocortisol from cortisol.493 Identification of the mRNAs for these and other relevant steroid-induced effects on TM cell biology is needed.489,490,495 A stress- and glucocorticoid-inducible response gene product (TIGR protein) whose progressive induction over time matches the time course of clinical steroid effects on intraocular pressure and outflow facility has been associated with a role in outflow obstruction and glaucoma pathogenesis.310,496 This molecule appears to be a secreted glycoprotein with aggregation- and extracellular matrix—binding groups.489,497 Combined with recent work on changes in laminin and fibronectin and alterations in phagocytosis and surface binding properties, studies of dexamethasone effects in cultured human TM cells may provide clues to the changes in pressure-dependent outflow seen with corticosteroid use. Cultured human TM cells exposed to dexamethasone also exhibit increased cell and nuclear size, an unusual stacked arrangement of smooth and rough endoplasmic reticulum, proliferation of the Golgi apparatus, pleomorphic nuclei,498 and perhaps most intriguing in terms of outflow resistance, increased amounts of extracellular matrix material and unusual geodesic dome-like cross-linked actin networks.498,499 Human TM cell monolayers grown on filters exhibit enhanced tight junction formation and decreased hydraulic conductivity in the presence of dexamethasone.500 Glucocorticoid glaucoma and primary open-angle glaucoma eyes both exhibit increased amounts of extracellular matrix material in the meshwork.501–504 Precisely how steroids affect aqueous humor dynamics remains controversial, with much work required to resolve all the issues. Other Mechanisms In the primate eye, approximately 60% to 80% of the resistance to aqueous outflow resides in the tissues between the anterior chamber and the lumen of Schlemm's canal.11,371,408–410,505–510 The major resistance site within the trabecular structures has still not been identified, but most investigators believe that it resides in the cribriform portion of the meshwork (juxtacanalicular tissue or JXT);11, 371,505–509,511 this outermost part of the mesh consists of several layers of endothelial cells embedded in a ground substance comprising a wide variety of macromolecules, including hyaluronic acid, other glycosaminoglycans, collagen, fibronectin, and other glycoproteins.512–516 These macromolecules are presumably produced by meshwork endothelial cells,481 and their synthesis and turnover may be one means by which outflow resistance is modulated;517 however, if this is the case, the control mechanisms are largely unknown despite new evidence suggesting that modulation of TIGR gene expression and TIGR protein production may play an important role.310,496 Some investigators believe the main resistance lies slightly proximal to the JXT.518,519 A small percentage of the resistance, perhaps 10% to 25%, resides in the inner wall of Schlemm's canal.505,506,520 Fluid movement across the inner canal wall endothelium itself appears to be predominantly through passive pressure-dependent transcellular pathways,507 but interest in flow between the cells has recently been revived.521 In eyes with POAG, there appears to be deposition of an as yet only partially characterized complex, electron-dense substance (possibly incorporating glycosaminoglycan (GAG) material within its structure) in the JXT, although the incidence of the confounding variables of age and prior medical therapy needs to be assessed.522,523 Exactly how important this substance is in increasing resistance to aqueous outflow (and thus increasing IOP) in open-angle glaucoma is uncertain. The endothelial cells of the TM have phagocytic capabilities.481,524,525 It has been proposed that the TM is in effect a self-cleaning filter, and that in most of the open-angle glaucomas, the self-cleaning (i.e., phagocytic) function is deficient or at least inadequate to cope with the amount of material present.526,527 Phagocytosis, especially of particulate matter and red blood cells, is also carried out by trabecular macrophages. Although this process may be important in clearing the anterior chamber of some inflammatory and other debris, it is probably not significant for bulk outflow of aqueous. Interestingly, however, artificial perfusion (even with pooled homologous aqueous humor or an artificial solution closely resembling it528,529) of the anterior chamber in primates is associated with a progressive time-dependent decrease in outflow resistance.17,327,343,412,413,528–532 This occurs even when the ciliary muscle has been detached from the scleral spur.413 Although the precise mechanism responsible for this phenomenon is not known, washout of resistance-contributing extracellular material from the TM has been a leading theory.528,529 Combining the clogged-filter concept of glaucoma with the washout concept of perfusion-induced resistance decrease inevitably led to interest in compounds (such as cytochalasins B433,532–534 and D,535 calcium chelators,536 ethacrynic acid,537–539 H-7,540 staurosporine, and latrunculin A541,542) that might disrupt the structure of the meshwork and canal inner wall so as to promote washout of normal and pathologic resistance-producing extracellular material. Such compounds might provide insights into cellular and extracellular mechanisms governing outflow resistance in normal and glaucomatous states. Additionally, if normal or pathologic extracellular material required many years to accumulate to the extent that would cause IOP to become elevated, perhaps a one-time washout would provide years of normalized outflow resistance and IOP.12,534,543 Hyaluronidase and Protease-Induced Facility Increases Intracameral infusion of hyaluronidase markedly increases facility in the bovine eye, presumably as the result of washout of acid mucopolysaccharide-rich extracellular material in the chamber angle tissues.544 Effects in primates are much more variable.508,510,545–547 In the enucleated human eye perfused at room temperature, α-chymotrypsin has little effect on facility.508 However, effects of trypsin may be masked at low temperatures,548 and a combination of trypsin and ethylenediaminetetraacetate (EDTA) may have a marked effect on dissociating cultured cells not easily dissociated by either agent alone.549 Perfusion of the anterior chamber of living monkeys with 50 U/ml α-chymotrypsin gives a large facility increase that persists for several hours even after the enzyme is removed from the infusate.550 The facility increase induced by intracameral 0.5 mmol Na2EDTA is augmented and prolonged by α-chymotrypsin.550 Exposure of porcine TM cells to growth factors such as TGF-β induced increases in matrix metalloproteinases (MMPs) such as stromelysin, gelatinase B, and collagenase, suggesting a role in the regulation of ECM turnover by TM cells.551 Purified MMPs increased outflow facility in organ-cultured human anterior segments by 160% for at least 125 hours.552 Cytoskeletal and Cell Junctional Mechanisms The adhesion of cells to their neighbors or to the extracellular matrix has multiple effects on cell shape and dynamics. Cell-cell and cell-extracellular matrix adherens junctions are complex and dynamic in nature, comprise many proteins, and are modulated by the ambient physical (pressure and shear stress) and chemical (endogenous hormonal and biochemical, exogenous pharmacologic) milieu. They play a role in signalling to the cell the state of its external environment. Actin filaments play a central role as the “backbone” of the submembrane plaque in both types of junctions, with the coupling of actin and myosin being essential for cell contactility.553 Monkey and human trabecular and Schlemm's canal endothelial cells are no exception to the above, and contain an abundant actin filament network.554–559 Cytochalasin B is a fungal metabolite that interferes with polymerization of cytoplasmic actin to form actin microfilaments, causing changes in cell shape and motility properties, and also inhibits hexose transport across the cell membrane.560–563 Cytochalasin D, a structural analog of cytochalasin B, is many times more potent in interfering with the formation of actin microfilaments but has virtually no effect on hexose transport. Cytochalasin D thus permits distinction between actin filament disruption and hexose transport inhibition as a possible basis for cytochalasin effects on a given biologic system.561–564 Anterior chamber infusion of microgram doses of cytochalsin B and D in cynomolgus and rhesus monkeys markedly increases total outflow facility within minutes.433,532,565 Tracer studies demonstrate that the increase in total facility represents increased facility across the meshwork and inner canal wall533 and is not related to contraction of the ciliary muscle, because the effect is similar in eyes with and without surgically disinserted ciliary muscles.532,533 Rather, the total facility increase is caused by distension of the cribriform meshwork and separation of its cells, ruptures of the inner wall endothelium of Schlemm's canal, and washout of extracellular material.534,565,566 Cytochalasin D is about 25 times more potent than cytochalasin B, although the maximal response (an approximate doubling of starting facility) is about the same for the two analogs,535 indicating an effect on actin filaments. Anterior chamber perfusion with other cytoskeletal agents, such as ethacrynic537,539,567,568 and tienilic504,569 acids (phenoxyacetic acids that induce shape changes in cultured TM cells), H-7540 and staurosporine541 (inhibitors of various protein kinases), and latrunculins A and B (marine sponge macrolides that bind monomeric G-actin and inhibit actin polymerization)541,542 also increase outflow facility more than two-fold. Perfusion of the monkey anterior chamber with calcium and magnesium-free mock aqueous humor containing 4 to 6 mmol Na2EDTA or with calcium-free mock aqueous containing 4 mmol ethyleneglycol bis (aminoethylether) tetraacetate (EGTA) also causes large facility increases and ultrastructural changes similar to those described earlier.536 Because EDTA chelates both calcium and magnesium, and EGTA is much more specific for calcium,570 calcium would appear to be a critical cation in maintaining the structural and functional integrity of the conventional outflow pathway. Although neither the precise subcellular events nor the exact pathophysiological sequence responsible for the chamber angle alterations produced by these agents has yet been elucidated, it seems that agents that alter the cytoskeleton, cell junctions, contractile proteins, or extracellular material are capable of, in effect, producing a “pharmacologic trabeculectomy.” Much work remains to be done before clinical use of such agents can be considered, but the prospect is enormously exciting. Cell- and Other Particulate-Induced Facility Decreases Normal erythrocytes are deformable and pass easily from the anterior chamber through the tortuous pathways of the TM and the inner wall of Schlemm's canal.507 However, non-deformable erythrocytes such as sickled or clastic (ghost) cells may become trapped within and obstruct the TM, elevating outflow resistance and IOP.571–573 Similarly, macrophages that have swollen after having ingested lens proteins leaking from a hypermature cataract,574 breakdown products from intraocular erythrocytes,575 or pigmented tumors (or the tumor cells themselves)576 may produce meshwork obstruction. Pigment liberated from the iris spontaneously (pigmentary dispersion syndrome) or iatrogenically (after argon-laser iridotomy) may clog the meshwork, presumably without prior ingestion by wandering macrophages,508,577,578 as may zonular fragments after iatrogenic enzymatic zonulolysis579,580 or lens capsular fragments after neodymium:yttrium-aluminum-garnet (Nd:YAG) laser posterior capsulotomy.581 Perfusion of the anterior segment with cross-linked polyacrylamide microgels, similar to those contained in flawed batches of the viscoelastic Orcolon (which produced severe glaucoma in some patients), produced persistent elevated IOP in monkeys and increased outflow facility in both monkeys and organ-cultured human and calf anterior segments because of obstruction of trabecular drainage.582 Protein- and Other Macromolecule-Induced Facility Decreases Glaucoma secondary to hypermature cataract (phacolytic glaucoma) or uveitis has long been ascribed to trabecular obstruction. In the former entity, the presence of protein-laden macrophages lining the chamber angle seemed adequate to account for the increased outflow resistance.574 The uveitis-related glaucomas comprise many different entities, and the etiology of the increased outflow resistance seems less clear; postulated mechanisms include trabecular involvement by the primary inflammatory process, trabecular obstruction by inflammatory cells, or secondary alteration of trabecular cellular physiology by inflammatory mediators or by-products released elsewhere in the eye. Small amounts of purified high-molecular-weight, soluble lens proteins583,584 or serum itself,585 when perfused through the anterior chamber of freshly enucleated human eyes, cause an acute and marked increase in outflow resistance. Thus, it may be that specific proteins, protein subfragments, or other macromolecules are themselves capable of obstructing or altering the meshwork so as to increase outflow resistance, perhaps contributing to the elevated IOP in entities such as the phacolytic, uveitic, exfoliation,586 and hemolytic575 glaucomas. The protein concentration in the peripheral portion of the anterior chamber, close to the meshwork, may be much higher than in the more central region, because of protein entry from the ciliary stroma through the peripheral iris.587–590 Experimental perfusion of the enucleated calf eye591,592 or monkey eye in vivo593 with medium containing higher protein concentrations than found in normal aqueous humor has indeed reduced or eliminated resistance washout. Perhaps protein in the TM is essential for maintenance of normal resistance either by providing resistance itself, or by signaling or modifying some property of the meshwork such as stimulation of focal adhesion and stress fiber formation, which would enhance adhesion of TM cells to the extracellular matrix.593–595 Hyaluronate- and chondroitin sulfate-based agents used as tissue spacers during intraocular surgery (“viscoelastic agents”) may raise IOP in human eyes if not completely removed from the eye by irrigation/aspiration at the conclusion of a procedure, presumably by obstructing trabecular outflow.596–600 |
OUTFLOW BIOMECHANICS |
Aqueous outflow through the conventional drainage pathway is a physical
process that can be altered pharmacologically. In every instance an alteration
of some physical characteristic of the meshwork/canal must account
for the altered outflow. Understanding the physics of pharmacologically-induced
alterations in outflow is crucial in developing new pharmacologic
approaches to manipulating meshwork physics to improve aqueous
outflow in glaucomatous states. A nonpharmacologic approach to altering physical characteristics of the meshwork/canal has been the application of laser energy to the TM. Initial efforts were directed at punching holes in the meshwork and inner canal wall. Such holes could be produced with resultant increases in outflow facility and decreases in IOP, but patency was nearly always transient; the openings scarred over, and facility and IOP usually returned to their former levels.601–606 Indeed, in monkeys intensive circumferential treatment of the meshwork with argon-laser energy can produce sufficient scarring to significantly decrease facility and increase IOP on a long-term basis.607 However, spacing 50 to 100 small (50 μm), less intensive argon, krypton, or diode laser burns evenly around the circumference of the meshwork of the glaucomatous human eye can result in a significant and long-term increase in outflow facility and decrease in IOP, apparently without actually producing a “hole” in the meshwork or inner canal wall.608–614 Indeed, histologic studies demonstrate the expected scar formation at the lasered sites.610 Although the presence of an undetected puncture leading from the anterior chamber into the canal lumen cannot be unequivocally excluded, it may be that contracture of the laser-produced scars tightens and narrows the trabecular ring, and the distortion somehow expands the meshwork, opens aqueous channels, and improves hydraulic conductance.605,610,611 This may be somewhat analagous to the effect of ciliary muscle contraction. Other data suggest that laser energy-induced alterations in trabecular cell biosynthetic, biodegradative, or phagocytic functions result in less hydraulic resistance. Increases in matrix metalloproteinases, gelatinase B, and stromelysin, which can degrade trabecular proteoglycans, have been demonstrated in human anterior segment organ cultures after argon laser trabeculoplasty.615,616 Definitive proof of the facility-increasing mechanism of laser trabeculoplasty remains to be established. |