Primary Open-Angle Glaucoma
YOUNG H. KWON and JOSEPH CAPRIOLI
Table Of Contents
NEUROPROTECTION: POTENTIAL NEW AVENUES FOR GLAUCOMA TREATMENT
GENERAL MANAGEMENT GUIDELINES
|Primary open-angle glaucoma (POAG) is the most common form of adult glaucoma. It
is an acquired condition characterized by (1) open anterior
chamber angle, with normal gonioscopic appearance; (2) chronic progressive
loss of retinal ganglion cells, manifest by a characteristic optic
neuropathy; and (3) typical patterns of visual field loss in the more
advanced stages. The more classic requirement for elevated intraocular
pressure (IOP) is usually dropped from the modern definition. Normal (mean ± standard
deviation) IOP is 15.5 (±2.57) mmHg.1 Assuming a gaussian distribution in which two standard deviations include
the values of about 95% of the population, an IOP of 20.5 mmHg (mean
plus 2 standard deviations) could be used as the upper limit of normal
IOP. Although the distribution of normal IOP is skewed toward the
higher range, this statistical definition was widely accepted and previously
used in the definition of POAG.|
POAG is an acquired condition, with onset typically after age 40. It affects both eyes but often asymmetrically. In early and even in moderate stages, the patient is usually asymptomatic. Abnormalities of the visual field occur insidiously and initially involve the midperiphery. In more advanced stages, the patient may become aware of an enlarging scotoma, particularly when it encroaches on fixation. If left untreated, vision may be lost. The natural history of the disease is one of a slowly progressive optic neuropathy. It has been estimated that untreated glaucoma can take an average of 14.4 years to progress from early to end stage at IOP of 21 to 25 mmHg, 6.5 years at 25 to 30 mmHg, and 2.9 years at more than 30 mmHg.2
Normal tension glaucoma occurs in a subset of patients who have POAG and can be described as a primary, acquired, progressive optic neuropathy characterized by typical patterns of cupping and visual field changes. It is distinguished from POAG by the absence of IOP elevated above the statistically normal range. There are those who believe that normal-tension glaucoma and POAG do not represent two separate entities but different places in the spectrum of the same disease process. Proponents of the terminology “normal-tension glaucoma” disagree and note some significant clinical differences.3–5
Ocular hypertension refers to IOP elevated above the statistically normal range without evidence of characteristic glaucomatous optic nerve damage or visual field abnormalities. Those with ocular hypertension or optic discs with enlarged cups are often referred to as “glaucoma suspects” because the risk of glaucoma increases with elevated IOP6 or an abnormally large cup (see Epidemiology).
|The pathogenesis of POAG is incompletely understood. It is believed that
the elevated IOP in glaucoma results from increased resistance within
the aqueous drainage system.7 Histopathologic studies of the outflow structures have uncovered potential
mechanisms.8 Other studies have addressed mechanisms of the optic nerve atrophy and
cupping in glaucoma.9 The mechanical theory proposes a direct pressure induced-damage to the
retinal ganglion cell axons at the level of the lamina cribrosa, whereas
the vascular theory proposes microvascular changes and resultant ischemia
in the optic nerve head. Cellular and molecular events that could
lead to glaucomatous retinal ganglion cell death have also been proposed.10|
MECHANISMS OF ELEVATED INTRAOCULAR PRESSURE
Morphologic alterations in the extracellular matrix of the aqueous outflow system in patients with glaucoma have been described in detail.8,11,12 Briefly, these changes include nodular proliferation of extracellular collagen, fragmentation, and “curling” of the collagen fiber bundles. There is an increase in glycosaminoglycan content8 but an overall decrease in hyaluronic acid.13 The endothelial cells lining the trabecular meshwork show “foamy” degeneration with basement membrane thickening.11 Ultrastructural changes in the juxatacanalicular tissue—the outermost aspect of the trabecular meshwork believed to be the most likely site of obstruction in glaucoma—have also been described.8,14–16 There is accumulation of nonfibrillar material with characteristics of basement membrane, curly collagen, and chondroitin sulfate protein complex. Changes in matrix vesicles (extracellular lysosomes), sheath material from subendothelial elastic-like fibers, extracellular glycoprotein, fibronectin, and elastin have been reported.16–20 Specificity of some of the morphologic changes has been questioned because similar findings have been noted in normal, aged eyes without glaucoma.15 This has led some to speculate that glaucomatous changes in the outflow pathway may represent an accelerated aging process.21
In addition to the changes in the trabecular meshwork, collapse of Schlemm's canal has been invoked as another mechanism of outflow obstruction.22,23 To support this hypothesis, adhesions between the inner and outer walls of Schlemm's canal have been shown.8,22 There is a certain amount of segmental variability in histopathologic specimens, however.24
Finally, differences in composition of the aqueous have been suggested as another mechanism for increased outflow resistance. Transforming growth factors (TGF) are polypeptides with multiple cellular regulatory functions. TGF can inhibit epithelial cell proliferation, induce extracellular matrix protein synthesis, and stimulate mesenchymal cell growth. Elevated levels of TGF-β2 have been found in the aqueous of glaucoma eyes.25 The study speculated that increased TGF-β2 levels may be responsible for the decreased cellularity of the trabecular meshwork and may lead to increased debris and resistance to outflow. Others report decreased collagenase activity, increased collagen synthesis, and elevated levels of metalloproteinase-1 inhibitor in the aqueous of glaucoma eyes.26 The study suggests that the decrease in collagen degradation may lead to excess deposition of collagen and loss of the trabecular meshwork cells in glaucoma. Despite these studies, the detailed cellular events and molecular substrates that lead to abnormalities of outflow resistance in glaucoma remain poorly understood.
Molecular genetic studies of large families with juvenile open-angle glaucoma have led to identification of the first glaucoma gene (GLC1A) in chromosome 1.27 Interestingly, about 3% of patients with typical adult-onset POAG also have a mutation in the GLC1A gene.28 This suggests gene mutation is responsible for a small but significant portion of POAG. Cellular and molecular events that lead a defective GLC1A gene and cause elevated IOP and glaucoma remain an active area of research.
MECHANISM OF OPTIC NERVE DAMAGE
Historically, glaucomatous optic nerve damage has been attributed to either a mechanical or vascular etiology. It is unlikely, however, that either theory alone will fully explain the optic nerve damage in glaucoma. POAG likely represents a diverse group of diseases, each involving one or more mechanisms. For the purpose of discussion, these broad mechanisms are considered separately.
Both in vitro and in vivo studies have shown that elevated IOP can cause posterior bowing of the lamina cribrosa, the collagenous structure that supports the retinal ganglion cell axons as they exit the eye.29–31 The lamina cribrosa is made up of about 10 parallel plates, each with various-sized pores that allow bundles of axons to pass through and yet maintain the competence of the eye to hold pressure. Evidence suggests that the plates of the lamina cribrosa are compressed in POAG and may even be entirely collapsed in some cases.32 Such physical distortion of the lamina cribrosa is thought to damage the passing axons by distortion or kinking. Other studies have shown elongation of the pores within the lamina cribrosa, suggesting mechanical forces that may stretch and fragment smaller beams.33 Changes in the extracellular matrix have been described that may lead to the loss of structural support in the lamina cribrosa.34–37 These changes include basement membrane thickening, disorganized and fragmented laminar beams, increased level of certain types of collagen, and structural changes in elastin. Interpretation of these morphologic changes within the lamina cribrosa should be done cautiously because they may represent secondary rather than primary changes in glaucoma.
There is evidence that elevated IOP can impede axoplasmic flow within the retinal ganglion cell axon.38–41 Axonal transport is vital to the normal functioning of neurons; retrograde axonal transport of target-derived neurotrophic factors may be essential for cell survival.42,43 It has been suggested that elevated IOP may lead to the degeneration of retinal ganglion cells by interfering with retrograde axoplasmic flow of essential neurotrophic factors. Lack of neurotrophic factors may trigger apoptosis (programmed cell death) in the retinal ganglion cell (see Cellular Mechanisms of Ganglion Cell Death).
Proponents of the vascular theory argue that microvascular changes in the optic nerve head are responsible for glaucomatous optic nerve damage.44 Blood supply to the prelaminar and laminar areas of the optic nerve is derived from the peripapillary choroid and short posterior ciliary arteries.45 The vascular supply to the anterior optic nerve may be compromised in glaucoma by several different mechanisms:
CELLULAR MECHANISMS OF GANGLION CELL DEATH
There is increasing interest in elucidating the cellular and molecular events that lead to retinal ganglion cell death in glaucoma. Apoptosis is a process by which excess neurons undergo spontaneous degeneration during normal development. Apoptosis has been demonstrated in primate55 and rat models of glaucoma.56,57 These studies suggest that elevated IOP may trigger cellular events leading to apoptosis. One hypothesis is that elevated IOP impairs the retrograde axonal transport of essential neurotrophic factors58,59 and in turn triggers apoptosis of the retinal ganglion cell.
Glutamate is an excitotoxic amino acid that normally functions as a neurotransmitter in the retina. Ischemia can produce excess levels of extracellular glutamate, which may lead to cell death through a complex series of cellular events that involves glutamate receptors and Ca+ + influx into the cell.60–62 Elevated levels of glutamate in the vitreous have been demonstrated in glaucomatous monkeys and humans, garnering support for this theory.63 It is unclear whether the accumulation of vitreal glutamate is a primary or secondary process in glaucoma.
|Glaucoma is a significant public health problem. It is the second leading
cause of blindness in the U.S.64,65 An estimated 2.25 million Americans have glaucoma and about 150,000 are
legally blind.66,67 In the older African-American population, glaucoma is the leading cause
of blindness, and the risk of blindness from glaucoma is 6.6 times greater
in that population, compared with white Americans.67 In the world, glaucoma is the third leading cause of blindness68; an estimated 13.5 million people may have glaucoma and 5.2 million of
those may be blind.68 In Nigeria, up to 34% of glaucoma patients may be blind in both eyes and 91% blind
in at least one eye.69|
PREVALENCE AND INCIDENCE OF POAG
Population-based studies show that the prevalence of POAG ranges from 0.4% to 8.8% in those older than age of 40 (Table 1). On average, POAG is found in 1.9% of white and 0.58% of Asian populations. In black populations however, the prevalence is significantly higher at 6.7%. Although some of the difference can be attributed to epidemiologic study design and the precise definition of POAG, the significantly higher rates observed in Western African populations probably reflect a fundamental risk factor associated with race (see Risk Factors).
The incidence rate of POAG is not precisely known. An annual incidence rate of 0.24% was reported in a Swedish population.78 More recently, population-based studies have addressed this issue (Barbados Eye Study and Baltimore Eye Survey) and are ongoing (1999). Conversely, several studies have reported incidence rates of glaucoma in patients with ocular hypertension (Table 2). Ocular hypertension is a well-known risk factor for the development of glaucoma (see Risk Factors). Reported annual incidence rates vary from 0% to 7%; about 1.7% of ocular hypertensive patients become on average, glaucomatous annually.
POAG = primary open-angle glaucoma.
Several risk factors for the development of POAG have been identified based on statistical analysis of population-based prevalence studies. Of these, elevated IOP, older age, black race, and positive family history are most strongly correlated with POAG. Other factors such as myopia, diabetes mellitus, systemic hypertension, and migraine or vasospasm are less strongly associated or their association is not clearly established.
Major Risk Factors
Both clinical and experimental studies suggest a strong correlation between elevated IOP and glaucoma. Clinically, secondary glaucoma with elevated IOP such as chronic angle-closure glaucoma produce optic nerve cupping and atrophy, with visual field loss often indistinguishable from that produced by POAG. Experimentally, elevated IOP in animal models of glaucoma can cause optic nerve cupping and atrophy similar to that seen in POAG.89 Furthermore, population-based studies have demonstrated a strong positive correlation between IOP and POAG (Fig. 1).6 The higher the IOP, the greater the prevalence of POAG.
The Baltimore Eye Survey has also shown a positive correlation between older age and POAG in both white and black Americans (Table 3).66 IOP is a confounding factor, however, because it also increases with age in Western populations.90 It is possible that the higher prevalence of glaucoma seen in older age groups may be due to an increased IOP. The proportion of ocular hypertensive patients who have glaucoma increases with age, however, implying that older age itself is a risk factor (Table 4). How aging predisposes to the development of glaucoma remains unclear but it may contribute to the vulnerability of the optic nerve to damage.
TABLE 3. Prevalence of POAG by Age and Race66
POAG = primary open-angle glaucoma.
TABLE 4. Prevalence of POAG and OcularHypertension by Age 90
POAG = primary open-angle glaucoma.
Black race is another important risk factor (see Tables 1 and 3, Fig. 1). In Baltimore, the prevalence of POAG is 6.6 times higher in African-Americans compared with white Americans.67 Collectively, several population-based studies have also shown that the prevalence of POAG among blacks is consistently higher than among whites (see Table 1). This is true even when glaucoma prevalence is adjusted for IOP (see Fig. 1) and age (see Table 3). The data strongly suggest an inherent predisposition of Western African descent to develop POAG.
Positive family history is another important risk factor. The Baltimore Eye Survey found 3.7 times the risk of developing POAG with positive first-degree relatives, 2.17 times with positive parents, and 1.12 times with positive children.91
MINOR RISK FACTORS
Several clinic-based studies suggest that myopia is more frequent in ocular hypertension and glaucoma than would be expected in a normal population.92–94 The elevated IOP associated with myopia does not fully explain the higher prevalence of glaucoma in myopia, suggesting that myopia itself is a risk factor.95 The mechanisms by which myopia predisposes to the development of POAG remain unclear.
Several studies have reported higher prevalence of diabetes mellitus in POAG.96–100 Other studies such as the Framingham Eye Study and Baltimore Eye Survey, however, failed to find such an association.51,52 The Framingham Eye Study determined diabetes by the presence of diabetic retinopathy, whereas the Baltimore Eye Survey relied on a history of diabetes provided by the patient. Both may have underestimated the true prevalence of diabetes. More recently, the Rotterdam Study reported a significant association between POAG and newly diagnosed diabetes mellitus.50 Although it seems plausible that microvascular changes in diabetes could predispose the optic nerve to the glaucomatous damage, direct experimental evidence for this is still lacking.
Systemic hypertension is another risk factor for POAG.101,102 The Rotterdam study found a significant association between elevated systolic blood pressure and POAG but not with normal-tension glaucoma.50 The Baltimore Eye Survey suggested that this relation may be more complex.103 Although both diastolic and systolic blood pressures are modestly associated, the lower perfusion pressure (blood pressure minus IOP) was most strongly associated with POAG.
The Blue Mountains Eye Study found a weak association between typical migraine and POAG in one age group (age 70 to 79).104 A Japanese study failed to find any association between migraine and POAG.105 In contrast, association between migraine and normal-tension glaucoma has been reported.106 Ischemia from periodic vasospasm leading to glaucomatous optic nerve atrophy remains an attractive hypothesis.
A large optic nerve cup may be a risk factor for the development of POAG. In one study, ocular hypertensive patients who developed visual field defects over a 5-year period had significantly larger cup-to-disc ratios (CDRs), compared with those who did not.107 The concept that a large cup is a risk factor for glaucoma remains somewhat problematic, however, because a large cup can also be a manifestation of glaucomatous damage.
A gene responsible for a hereditary form of juvenile open-angle glaucoma (GLC1A) has been identified.27 Interestingly, about 3% of patients with typical adult-onset POAG also show a mutation in the GLC1A gene.28 As more glaucoma genes are identified, it may be possible in the future to assess the risk of glaucoma based on genetic testing.
There is evidence that IOP is not only a risk factor for glaucoma but also a prognostic factor. Higher IOP is associated with faster progression of glaucoma.2 Indeed, there is evidence to show that lowering IOP slows or halts progression of the disease (see Treatment section).
Black race is another prognostic factor. At the initial time of diagnosis, blacks tend to be younger and have more advanced disease than whites.66,108 Glaucoma progression is more rapid109 and the rate of blindness from glaucoma is higher in blacks than whites.67 It is generally believed that the differences are only partially explained by socioeconomic factors or accessibility to medical care.
Disc hemorrhage is another important prognostic factor.110 In one study of unilateral disc hemorrhage, the eye with the hemorrhage showed greater visual field progression than the fellow eye.111 A new disc hemorrhage in a patient with glaucoma is considered to be a sign of progressive optic nerve damage.
|Clinical diagnosis of POAG is based on history, clinical examination, and
visual field testing. Previous ocular and medical histories are important
in assessing the risk of having or developing POAG (see Risk Factors). Ocular
examination may reveal elevated IOP (more than 21 mmHg). Otherwise, the
anterior segment examination is unremarkable with an
open normal-appearing anterior chamber angle. The optic nerve shows cupping
and there may be associated visual field abnormalities. Other open-angle
forms of glaucoma, such as those associated with pseudoexfoliation, pigment
dispersion, or trauma should be ruled out.|
ELEVATED INTRAOCULAR PRESSURE
The IOP is subject to normal diurnal fluctuation of 3 to 6 mmHg.112 Diurnal variation of more than 10 mmHg is unusual and should raise suspicion for glaucoma.113,114 The most common diurnal pattern is an early morning peak.115 The early morning peak has been correlated with the endogenous adrenocortical steroid level.116 Others have challenged this association by pointing out that IOP decreases during sleep despite absence of similar reduction in the plasma cortisol level.117 The clinical significance of diurnal variation is that it is possible to miss the elevated IOP in patients with POAG with a single, isolated measurement. A diurnal curve or multiple measurements of IOP can be carried out throughout the day to confirm the diagnosis of POAG and explore the possibility of normal pressure glaucoma.
OPTIC DISC AND VISUAL FIELD CHANGES
Normal optic discs show healthy intact neural rims,118,119 with corresponding full visual fields. In the normal disc, the average horizontal CDR is about 0.4 when viewed stereoscopically.120 Estimates of the CDR vary by as much as 0.2 when the same nerve is examined multiple times.121 The CDR in fellow eyes tends to be similar (Fig. 2).122,123 There is a significant correlation of the disc appearance among members of the same family.124,125 Thus, significant asymmetry of the optic disc rim between fellow eyes should increase suspicion of a potentially glaucomatous process (Fig. 3). Estimates of CDR alone do not adequately describe the status of the neural rim, however. Focal thinning of the neural rim is often an early sign of glaucoma. Such focal changes are difficult to describe with CDR estimates. Instead, they are better documented by careful disc drawings and color stereoscopic optic disc photography. Disc drawings should emphasize the status of the neural rim, asymmetry between fellow eyes, and asymmetry between the superior and inferior rims of the same eye. Ultimately, stereoscopic disc photography provides a more objective documentation.
At least four different patterns of glaucomatous optic disc changes have been described.126 They include the focal ischemic, myopic, senile sclerotic, and generalized enlargement of the cup.
The focal ischemic disc shows localized, discrete loss of the neural rim most commonly in the inferotemporal but also in the superotemporal quadrant (Fig. 4). This has also been referred to as “polar notching,” 127 “focal notching,”128 or “pitlike changes.”129 The rest of the neural rim may be relatively intact. The focal changes in the disc have corresponding visual field defects. Dense superior paracentral scotomas are common and are accompanied by pitlike changes in the inferotemporal neural rim. Clinical factors associated with this type of disc are (1) middle- to older age, (2) female gender, (3) normal or elevated IOP, and (4) migraine.126 A third of the focal ischemic type may show disc hemorrhages.
Myopic discs are tilted discs with temporal crescents and glaucomatous damage to the rim, characterized by superior and inferior thinning of the neural rim (Fig. 5). Their visual field changes are similar to the focal ischemic group but are less likely to threaten fixation or preferentially involve the superior field. These discs tend to be associated with (1) younger age group, (2) male gender, and (3) people of Asian descent.126
Senile sclerotic discs show saucerized and shallow cups with a “moth-eaten” appearance (Fig. 6). They are often accompanied by peripapillary atrophy and choroidal sclerosis, with some pallor of the neural rim. Their visual fields are characterized by relative scotomas with diffuse loss. The senile sclerotic disc is associated with (1) advanced age, (2) normal or elevated IOP, and (3) microvascular diseases such as ischemic heart disease or systemic hypertension.126
Generalized enlargement of the cup is manifest by diffusely enlarged round cups, without focal loss of the neural rim (Fig. 7). Visual fields show diffuse generalized loss without localized defects. These discs tend to be associated with (1) the younger group, and (2) markedly elevated IOP.126 Disc hemorrhages are infrequently found in this type of disc.
The types of glaucomatous disc changes described generally apply to early and moderate stages of glaucoma. These are examples of relatively “pure” types; many discs have an intermediate appearance. In advanced stages, there is loss of the entire neural rim, and distinctions among different disc types may not be possible (Fig. 8). Histologic cross-section of advanced cupping or “bean-pot” cup shows extreme posterior displacement of the lamina cribrosa and undermining of the disc margin.127,128 The visual field loss in advanced glaucoma may leave only a central or temporal island of remaining vision. In the end, loss of all sight is possible with complete cupping, so-called absolute glaucoma.
Splinter hemorrhages at the disc margin are a common feature of glaucoma.130 They occur in all types of glaucomas but are more commonly associated with normal-tension glaucoma.131,324 They frequently occur in the inferior quadrant and are usually seen in the early and moderate stages of glaucoma. They become rare in advanced stages, wherein there is a complete loss of the neural rim.130 Disc hemorrhage is a clinically important sign because its presence has been correlated with an increased rate of optic nerve damage.133 It may also be an early sign of glaucoma because it frequently precedes nerve fiber layer defects,134 focal notches in the neural rim,135 and glaucomatous visual field defects136 (Fig. 9). Loss of retinal ganglion cells and their axons in glaucoma also produces defects in the nerve fiber layer. On ophthalmoscopy, the normal nerve fiber layer appears as fine striations extending temporally in an arcuate fashion from the superior and inferior poles of the disc. With the loss of ganglion cell axons in glaucoma, nerve fiber layer loss can appear as either diffuse or discrete wedge-shaped defects (Fig. 10).137,138 They may follow a disc hemorrhage139 and correlate well with visual field changes.140,141 This finding can be a sensitive and early indicator of glaucoma damage.142–144
Peripapillary atrophy is also frequently seen in glaucoma (Fig. 11). Peripapillary changes occur more frequently and are more extensive in eyes with glaucoma.145,146 Peripapillary atrophy can also progressively enlarge with progression of glaucoma.147 It is not a specific sign of glaucoma, however, because it is also frequently seen in myopia and aging.
Glaucoma is not the only condition with optic nerve cupping and visual field loss. Other conditions, both congenital and acquired, can mimic glaucoma. Congenital disc anomalies such as optic nerve coloboma148 (Fig. 12), congenital pit149,150 (Fig. 13), and tilted disc syndrome151 (Fig. 14) can produce optic nerve and visual field changes that are similar to those found in glaucoma. In addition, these anomalies can interfere with recognition of glaucomatous damage when they coexist with glaucoma.151 Acquired conditions such as anterior ischemic optic neuropathy (of the arteritic variety)152,153 and compressive lesions such as intracranial aneurysm154 can produce disc appearance and visual field defects that resemble glaucoma. Young age, highly asymmetric or unilateral disc changes, atypical visual fields, or visual fields that do not correspond to the disc changes should increase the examiner's suspicion of a nonglaucomatous cause.
The most widely accepted method for assessment of optic nerve function in glaucoma is perimetry or visual field testing. Both kinetic (e.g., Goldmann perimeter) and static automated perimetry (e.g., Humphrey, Octopus) are standard means for peripheral vision testing. There is evidence that a significant amount of optic nerve damage can occur before standard visual fields become abnormal.155,156 Thus, other functional tests have been developed that can detect glaucoma at earlier stages. These include short-wavelength automated perimetry,157,158 motion detection perimetry (Fig. 15),159 contrast sensitivity,160 and the pattern electroretinogram.161,162
Sophisticated imaging instrumentation is available to objectively evaluate the status of the optic nerve and nerve fiber layer. It can provide three-dimensional tomographic analysis of the disc based on confocal scanning laser ophthalmoscopy163 and measure the nerve fiber layer thickness with the techniques of scanning laser polarimetry164,165 (Fig. 16) or optical coherence tomography.166 Whether these new instruments will be clinically useful in early detection and accurate monitoring of glaucoma remains to be shown.167
|Treatment of POAG is limited to reduction of IOP. This is based on the
concept that elevated IOP is at least partly responsible for the optic
nerve damage in glaucoma and that IOP reduction can halt or slow the
progression of the disease. Numerous clinical studies support this concept.168–173 A recent 5-year prospective study showed a better outcome in eyes treated
initially with surgery, compared with eyes treated either with medicine
or laser.174 Surgical treatment resulted in lower mean IOP (14.1 mmHg), compared with
laser or medical treatment (18.5 mmHg), suggesting that lower IOP helped
to protect against disease progression. Additional evidence comes
from the Glaucoma Laser Trial and Glaucoma Laser Trail Follow-up Study.174 The lower IOP in the initial laser trabeculoplasty group was associated
with less disease progression, compared with the initial medicine group. In 1992, a
prospective 4-year clinical study was begun (Early Manifest
Glaucoma Trial) to directly address the question of whether IOP
reduction, compared with no treatment, makes any difference in newly diagnosed
Treatment of POAG involves medications, laser surgery, and incisional surgery.
Medications used to treat POAG include many classes of drugs, all designed to lower IOP. They include beta-adrenergic antagonists, nonselective adrenergic agonists, selective alpha-2 adrenergic agonists, cholinergic agonists, carbonic anhydrase inhibitors, prostaglandin analogs, and hyperosmotic agents (Table 5).
Topical nonselective beta-adrenergic antagonists such as timolol lower IOP by suppressing aqueous production.176 They inhibit synthesis of cyclic adenosine monophosphate (c-AMP) in the ciliary epithelium and lead to a decrease in aqueous secretion.177 Long-term trials with topical timolol in glaucoma patients have shown a sustained reduction of IOP over time.178,179 Ocular side effects of topical beta-blockers are minor and include burning and decreased corneal sensation. Conversely, systemic side effects can be serious. These include bradycardia; arrhythmia; heart failure; heart block; syncope; bronchospasm or airway obstruction; central nervous system effects (depression, anxiety, weakness, fatigue, or hallucinations); and elevation of blood cholesterol levels.180 A beta-1 selective antagonist, betaxolol, has fewer pulmonary side effects but is also less effective in lowering IOP than the nonselective beta antagonists.181,182 Because of long clinical experience and proved efficacy, topical beta-blockers have assumed a central role in the medical treatment of POAG.
Nonselective adrenergic agonists such as epinephrine lower IOP by several different mechanisms.183 Initially, a vasoconstrictive effect decreases aqueous production; another early effect is an increase in the outflow facility by stimulating c-AMP synthesis. Long-term effects may include a further increase of outflow facility. Clinical efficacy of topical epinephrine appears to be similar184 to or slightly less185 than that of timolol. There are numerous side effects associated with topical adrenergic agonists, including both ocular (burning, reactive hyperemia, adrenochrome deposits, mydriasis, maculopathy in aphakic eyes, corneal endothelial damage, and ocular hypoxia) and systemic (hypertension, tachycardia and arrhythmia) symptoms.186 Dipivefrin is a prodrug that is hydrolyzed to epinephrine as it traverses the cornea.187 It has significantly fewer systemic side effects than epinephrine.188 Generally, the use of adrenergic agonists in the treatment of POAG has declined in recent years with the availability of newer medications that show comparable or better efficacy and have fewer side effects.
Selective topical alpha-2 agonists such as apraclonidine decrease aqueous production.189 In addition, a second generation—the highly selective alpha-2 agonist, brimonidine—appears to increase uveoscleral outflow.190 Clinical studies show that topical apraclonidine and brimonidine are as effective as timolol in reducing IOP (20% to 25%) in patients with glaucoma.191,192 Common ocular side effects include allergic reaction, follicular conjunctivitis, eyelid retraction, mydriasis, and conjunctival blanching.193 Systemically, they can cause headache, dry mouth, fatigue, bradycardia, and hypotension. Long-term use of topical apraclonidine is frequently associated with allergy and tachyphylaxis (decreased effectiveness over time).194 Generally, brimonidine seems to produce fewer ocular side effects than apraclonidine.192
Topical cholinergic agonists such as pilocarpine increase the trabecular outflow by contraction of the longitudinal ciliary muscle.195 The same action may also decrease uveoscleral outflow through the ciliary muscle.196 There are two types of cholinergic agonists: direct and indirect. The direct agents (e.g., pilocarpine) are cholinergic receptor agonists; the indirect agents (e.g., echothiophate iodide) inhibit cholinesterase and prolong the action of native acetylcholine. Clinical efficacy of pilocarpine is comparable to timolol,197 whereas that of echothiophate iodide may be superior to timolol in aphakic patients.198 Systemic side effects of pilocarpine are rare. In contrast, ocular side effects are common and include brow ache, induced myopia, miosis (leading to decreased vision), shallowing of the anterior chamber, retinal detachment, corneal endothelial toxicity, breakdown of the blood-brain barrier (which can exacerbate inflammation), hypersensitivity or toxic reaction, cicatricial pemphigoid of the conjunctiva, and atypical band keratopathy. The indirect agents have ocular side effects that are generally more intense than those of the direct agents. In addition, indirect agents can cause iris cysts in children and cataract in adults. Finally, prolonged respiratory paralysis may occur during general anesthesia in patients who are on cholinesterase inhibitors because of their inability to metabolize paralytic agents such as succinylcholine.199 Use of cholinergic agents (as with nonselective adrenergic agonists) has declined in recent years with the availability of newer medications that have comparable efficacy and fewer side effects.
In the ciliary processes, formation of bicarbonate is linked to Na+ secretion and aqueous humor production. Carbonic anhydrase inhibitors such as acetazolamide reduce aqueous production by decreasing bicarbonate production.200 Another proposed mechanism is related to the metabolic acidosis produced by carbonic anhydrase inhibitors that can reduce aqueous production.201 Acetazolamide, a systemic carbonic anhydrase inhibitor, can decrease aqueous flow by 27% in human eyes.202 Many side effects are associated with systemic carbonic anhydrase inhibitors, including transient myopia; parasthesia of the fingers, toes, and perioral area; urinary frequency; metabolic acidosis; malaise; fatigue; weight loss; depression; potassium depletion; gastrointestinal symptoms; renal calculi formation; and rarely, blood dyscrasia.203 Dorzolamide, a topical carbonic anhydrase inhibitor, has recently become available. It has significantly fewer systemic side effects than oral carbonic anhydrase inhibitors and still has clinical efficacy comparable to that of timolol.204
A prostaglandin analogue, latanoprost, represents the newest class of medications. Latanoprost increases uveoscleral outflow by changing the structure of the ciliary muscle.205 In clinical trials, topical latanoprost was more effective in IOP reduction (by 35%) than timolol (27% reduction).206 Ocular and systemic side effects were well-tolerated except for a 12% incidence of increased iris pigmentation. This curious side effect tends to occur in eyes with mixed green-brown or blue-brown irides.207 The clinical significance of this remains unclear; further studies are needed to understand the cellular nature of this side effect. Use of latanoprost has also been associated with anterior uveitis and cystoid macular edema in susceptible individuals.208
Finally, hyperosmotic agents such as oral isosorbide can rapidly lower IOP by decreasing vitreous volume. They do not cross the blood-ocular barrier and therefore exert oncotic pressure that dehydrates the vitreous. Both oral and intravenous agents are available. Side effects associated with the hyperosmotic agents can be severe and include headache, back pain, diuresis, circulatory overload with angina, pulmonary edema and heart failure, and central nervous system effects such as obtundation, seizure, and cerebral hemorrhage.209 Because of the frequent and potentially serious side effects, they are not used as a long-term agent. They are often used to temporarily reduce high IOP until more definitive treatments can be rendered.
Surgical treatment of POAG includes laser trabeculoplasty, filtering procedures (full-thickness and guarded procedures), aqueous drainage implants, and cyclodestructive procedures.
Argon laser application to the trabecular meshwork (argon laser trabeculoplasty) has been shown to significantly lower IOP (Fig. 17).210 The mechanism by which laser trabeculoplasty lowers IOP is not completely understood. Collagen shrinkage and scarring of the trabecular meshwork may open adjacent intertrabecular spaces and decrease the overall outflow resistance.211 Other studies suggest that an increase in mitosis and phagocytosis of the trabecular endothelial cells after laser trabeculoplasty may enhance outflow.212 Argon laser trabeculoplasty can lower IOP in 50% of eyes at 5 years with an attrition rate of 6% to 10% per year.213,214 Prognostic factors for favorable outcome include preoperative IOP of 20 to 29 mmHg, phakic eyes, and older age (older than 40).197,198
Complications of argon laser trabeculoplasty include discomfort, acutely elevated IOP, progressive visual field loss, peripheral anterior synechiae, iritis, sector palsy of the pupillary sphincter, corneal abrasion, corneal edema, endothelial damage, and vasovagal reaction.217 Transiently elevated IOP of less than 10 mmHg can occur in up to 50% of those treated with laser trabeculoplasty.218 Preoperative treatment with apraclonidine219 or brimonidine220 can significantly reduce the rate of postoperative IOP elevation. Rarely, persistently elevated IOP after laser trabeculoplasty may require trabeculectomy.
Full-thickness filtering procedures were designed to create a direct fistula between the anterior chamber and the subconjunctival space, thus bypassing the eye's outflow structures. This can be achieved by thermal cautery,221,222 scleral punch (posterior lip sclerectomy),223 or external trephination.224 Full-thickness filtering procedures effectively lower IOP but are associated with significant postoperative complications, including flat anterior chamber, hypotony, choroidal detachment, endophthalmitis, and cataract formation.
In 1968, a successful partial-thickness guarded-filtering procedure (or trabeculectomy) was first reported.225 Since then, trabeculectomy has become the surgical procedure of choice because of the significantly lower incidence of postoperative complications and comparable efficacy. Clinical success of trabeculectomy approaches 85% to 95% at 2 years.226 Introduction of antifibrotic agents such as 5-fluorouracil and mitomycin C have improved the outcome of trabeculectomy. Antifibrotic agents inhibit wound healing and promote the patency of the fistula (Fig. 18). Trabeculectomy with adjunctive 5-fluorouracil results in significantly lower IOP, compared with trabeculectomy without 5-fluorouracil.227 Adjunctive use of mitomycin C in trabeculectomy produces a similar rate of success as 5-fluorouracil.228 Complications of trabeculectomy are similar to those of full-thickness procedures; they include decreased vision, choroidal effusion, shallow or flat anterior chamber, persistent inflammation, filtration failure, corneal dellen, suprachoroidal hemorrhage, endophthalmitis, chronic hypotony, and maculopathy. Generally, however, trabeculectomy is associated with fewer complications than full-thickness procedures. The risk of persistent hypotony and associated complications such as maculopathy, late-onset bleb leaks, infection of the bleb, and endophthalmitis are increased with the adjunctive use of antifibrotic agents.229,230 The risk of hypotony maculopathy is particularly high in young myopic patients with low scleral rigidity.231 Corneal and conjunctival epithelial toxicity are associated with repeated subconjunctival injections of 5-fluorouracil.232
AQUEOUS DRAINAGE IMPLANTS
Use of aqueous drainage implants (or setons) are generally reserved for patients who have complicated secondary glaucomas such as uveitic glaucoma, neovascular glaucoma, or glaucoma after other ocular procedures. In POAG, the drainage devices are typically used when previous trabeculectomy has failed or in an aphakic or pseudophakic eye in which the conjunctiva is excessively scarred for successful trabeculectomy. Broadly, there are two types of drainage implants: nonrestrictive and restrictive (Fig. 19). Molteno and Baerveldt setons are nonrestrictive drainage implants that permit free flow of aqueous from the anterior chamber to a scleral plate through silicone tubing. Restrictive drainage implants such as the Krupin and Ahmed devices incorporate a valvular mechanism that provides some resistance to aqueous outflow and can reduce the incidence of early postoperative hypotony and shallow anterior chamber.
Surgical outcome of the Molteno seton has a success rate of 60% to 80% at 6 months.233 There is evidence that the double-plate Molteno implant with larger surface area lowers IOP more than the single-plate implant.234 The Baerveldt seton produces surgical results similar to the Molteno.235 Both the Krupin and Ahmed valves achieved similar surgical results at 1 year.236,237 Complications of the aqueous drainage implants include hypotony, choroidal detachment and flat anterior chamber.238 Temporary reduction of flow with the application of a dissolvable suture ligature around the tube until scleral plate encapsulation occurs can reduce complications related to early postoperative hypotony. Early postoperative hypotony and shallow anterior chamber seem less frequently associated with restrictive devices such as the Ahmed,237 although long-term IOP control may be better with the large-plate nonrestrictive devices. Other complications include inflammation, tube obstruction, elevated IOP, tube migration, implant and tube erosion, corneal decompensation, cataract, endophthalmitis, strabismus, and epithelial downgrowth.
If the patient is a poor candidate for trabeculectomy or drainage implant surgery, a cyclodestructive procedure can be considered. Cyclodestructive procedures decrease aqueous production by ablating the portion of the ciliary body that produces aqueous. This can be achieved by freezing the ciliary epithelium and capillaries within the ciliary body (cyclocryotherapy).239,240 Cyclocryotherapy was shown to be effective in controlling IOP in aphakic glaucoma and glaucoma after penetrating keratoplasty.241 Laser has also been used to selectively destroy the ciliary processes (cyclophotocoagulation). Two commonly used lasers for this purpose are transscleral Nd:YAG and semiconductor-diode lasers. These lasers can produce thermal damage to the ciliary processes and decrease aqueous production. Nd:YAG cyclophotocoagulation has been shown to reduce IOP by 44% to 68%.242–244 Diode cyclophotocoagulation can achieve IOP control in 60% to 80% of eyes treated (Fig. 20).245 Finally, transpupillary cyclophotocoagulation246 and more recently, intraocular cyclophotocoagulation with endoscopic visualization, have also been described.247 Complications of cyclodestructive procedures include pain, reduction of visual acuity, inflammation, transient IOP rise, hyphema, vitreous hemorrhage, cataract, hypotony, choroidal detachment, flat anterior chamber, and phthisis.248,249 Generally, cyclophotocoagulation seems to be associated with better tolerance and fewer complications than cyclocryotherapy.
|NEUROPROTECTION: POTENTIAL NEW AVENUES FOR GLAUCOMA TREATMENT|
|The search for neuroprotective agents for glaucoma treatment is grounded
in desperation: the desperation of continuing visual loss in some patients
despite IOP reduction to quite low levels. Some cases continue
to progress despite dramatic IOP lowering to 5 to 10 mmHg. We present
possible avenues for neuroprotection, given the likely players in the
damage process. This topic is the subject of a recent review.61|
CALCIUM CHANNEL BLOCKADE
Calcium channel blockers have been used empirically to treat low-tension glaucoma. Patients with vasospastic conditions and normal pressure glaucoma have been particularly targeted.250,251 Nifedipine and nimodipine have both been used for treatment of normal-tension glaucoma. Blockade of calcium channels at the neuronal cellular level—by interrupting the cascade of events that lead to death from ischemia—is also a reasonable rationale. The systemic lowering of blood pressure by calcium channel blockers is a concern because it could reduce perfusion pressure to the anterior optic nerve head and lead to ischemia. A retrospective clinical study compared 56 patients with glaucoma who were concurrently taking calcium channel blockers to a control group not taking such medications for a mean follow-up period of 3.4 years, suggested that calcium channel blockers may be useful in the management of low-tension glaucoma.252
Neuronal injury from glutamate receptor-mediated neurotoxicity has been implicated as a central mechanism in a wide variety of central nervous system diseases, including ischemia, trauma, and some chronic neurodegenerative diseases. Excitotoxicity may also interact with other pathophysiologic processes to enhance or facilitate neuronal damage. The possibility that excitotoxicity may play a role in the chronic neurodegeneration of glaucomatous damage has been suggested. Recent reports of elevated levels of glutamate in the vitreous of glaucomatous monkeys and humans have provided additional fuel for this hypothesis.253,254 Whether the high vitreous levels of glutamate are a cause or result of damage is undetermined but high concentrations of this neurotoxin is toxic to the inner retina.
The viability of retinal ganglion cells depends on the retrograde flow of neurotrophic compounds from the target tissue to the cell body. Interruption of retrograde axonal transport could interrupt this supply and trigger apoptotic cell death. This concept has recently been reviewed.10 Restoration of these neurotrophic factors by exogenous administration—or more likely, through forms of gene therapy—could restore “sick” retinal ganglion cells back to health.
Recent advances in the understanding of the biochemical and molecular biologic events that lead to neuronal cell death have suggested novel therapeutic approaches. Relatively little attention has been drawn to the importance of intrinsic neuroprotective events in the modulation of cell injury. In this context, heat-shock proteins are likely to play an important role in cell survival after a variety of metabolic insults. A recent study showed that retinal ganglion cells express the 72-kd heat-shock protein after hyperthermia, sublethal hypoxia, and glutamate exposure in vitro.255 Furthermore, retinal ganglion cells in culture treated with hyperthermia or sublethal hypoxia were less to susceptible to subsequent damage from excitotoxicity and anoxia. The neuroprotective effect of the induction of heat-shock protein synthesis by hyperthermia and sublethal hypoxia suggests a role for heat-shock protein as a protective mechanism against ischemic and excitotoxic retinal ganglion cell death.
NITRIC OXIDE SYNTHASE INHIBITION
Nitric oxide is a rapidly diffusing gas with a short half-life in vivo. It has a vasodilatory action and may act as a nonconventional neurotransmitter in the brain. Nitric oxide in sufficient concentrations is a potent neurotoxin. The exact place of nitric oxide in the cascade of events associated with ischemic central nervous system damage is not known but it likely plays an important role. Inhibitors of nitric oxide synthase can protect neurons from nitric oxide toxicity.256
The reperfusion phase after ischemic injury produces highly reactive compounds called free radicals. These oxygen-containing molecules have unpaired electrons and react with lipids, nucleic acids, and proteins. They are thought to be important mediators of reperfusion injury. Free radicals may also facilitate the release of excitotoxins, and both may work together to bring about cellular death from ischemia.257 Free radicals recently have been implicated in the slow chronic neurodegeneration of amyotrophic lateral sclerosis,258 so their role in a chronic neural degeneration like glaucoma is entirely feasible. Free-radical scavengers include endogenous enzymes such as catalase and superoxide dismutase and the antioxidant vitamins, especially C and E. Therapy could take the form of turning on the synthesis of endogenous compounds or providing exogenous ones. Some level of antioxidation can be achieved through vitamin therapy but requires well-controlled clinical studies to determine efficacy.
Apoptosis is a term applied to suicidal cell death. This is a programmed gene-directed self-destruction that is a normal occurrence in neural development and differentiation. In the adult nervous system of mammals, it is a mechanism of pathologic cell death. Apoptosis appears to occur—at least to some degree—in most models of neural cell death, whether it is caused by axotomy, ischemia, excitotoxicity, or deprivation of neurotrophins. Apoptosis has been shown to be at least one of the mechanisms for retinal ganglion cell death in animal models of pressure-induced glaucoma in the monkey55 and the rat.57 Evidence of necrotic cell death has not been found in human glaucoma or in the monkey model and is the basis for the hypothesis that apoptosis is the dominant mechanism of retinal ganglion cell death in glaucoma. Because the rate of glaucomatous damage is so slow, however, few cells are expected to be found in the agonal phase.
Drugs are being developed to block apoptosis. Deprenyl, originally developed as a monoamine oxidase inhibitor, increases the gene expression that inhibits apoptosis. Other drugs inhibit the later steps of apoptosis, which include the action of proteases on cell proteins. Several antiapoptosis agents have already been evaluated in the rat retina to prevent light-induced death of photoreceptor cells. Intravitreal injections of flunarizine and aurintricarboxylic acid259 can delay apoptotic death of photoreceptors. Much remains to be learned about the mechanisms that initiate and regulate the process of apoptosis in adult mammals.
OPTIC NERVE REGENERATION
Impressive results reported by Aguayo and coworkers demonstrate the feasibility of central nervous system regeneration. Implantation of peripheral nerve sheath grafts into the eyes of rats promotes the regrowth of axotomized retinal ganglion cells into the graft.260 These regenerated axons also have the ability to establish synaptic connections at target cells.261 The peripheral nerve sheath appears to confer on the central neurons the ability to regenerate by providing a suitable environment and growth factors. This approach may yield important molecular insights into neuroprotection or neuroregeneration, although it is unlikely to yield any clinically applicable therapies in the near future.
|GENERAL MANAGEMENT GUIDELINES|
|Once the diagnosis of POAG is made, it is usually helpful to set a target
IOP. The target IOP is the level at which no further glaucoma progression
is expected. Generally, the more advanced the glaucoma damage, the
lower the target IOP should be. Therefore, the target IOP is not an
absolute value. It varies among patients and even in a specific patient, depending
on the previous ocular history, age, status of the optic
nerve, associated risk factors, and response to treatment. For example, in
mild glaucoma with minimal optic nerve changes and a normal visual
field, the target IOP may be 21 mmHg; in advanced glaucoma with marked
disc cupping and severe visual field loss, it may be 12 mmHg or even
Once the target IOP is set, treatment to lower IOP is rendered. Glaucoma medications are often prescribed first and may be a beta-blocker, prostaglandin analog, alpha-2 agonist, or topical carbonic anhydrase inhibitor. Other medications such as cholinergic agonists, nonselective adrenergic agonists, or oral carbonic anhydrase inhibitors may also be chosen. Selection of a particular medication depends on efficacy, side effects, patient tolerance, cost, and compliance. Periodic regular follow-up is needed to evaluate these factors. Ineffective medications should be stopped and if necessary, additional medications added until the target pressure is reached. Laser trabeculoplasty can be considered when multiple medications have failed to control IOP. Trabeculectomy can be considered if the laser trabeculoplasty or medications failed to control IOP. Adjunctive antifibrotic agents may be used if the trabeculectomy has a high risk of failure or if a low IOP is desired. If the patient is not a candidate for trabeculectomy, an aqueous drainage implant can be considered. Finally, a cyclodestructive procedure may be necessary if other forms of therapy fail or visual prognosis is poor.
The particular sequence of treatment modalities outlined above represent the authors' bias. Others have suggested that initial treatment with laser174,262 or surgery169,263 may be more effective than medications. Most practitioners, however, still choose at least a trial of medications over laser or surgery as initial therapy in most patients.264,265
The glaucoma patient requires regular follow-up examinations, from 2 to 4 times a year or more often, as indicated by the clinical picture. Formal visual field testing is repeated every 4 to 12 months, depending on the stability of the disease. If unstable or deteriorating, the target IOP may need to be lowered and more aggressive treatment applied.
It is important that the physician treating glaucoma keep the whole patient in mind. Treatment for glaucoma, whether medical or surgical, can significantly impact the patient's quality of life. Treating glaucoma at the expense of the patient's overall health may be counterproductive. Therefore, treatment of glaucoma should be individualized according to the needs and desires of a properly informed patient.
26. González-Avila G, Ginebra M, Hayakawa T et al: Collagen metabolism in human aqueous humor from primary open-angle glaucoma. Decreased degradation and increased biosynthesis play a role in its pathogenesis. Arch Ophthalmol 113:1319, 1995
36. Hernandez MR: Ultrastructural immunocytochemical analysis of elastin in the human lamina cribrosa. Changes in elastic fibers in primary open-angle glaucoma. Invest Ophthalmol Vis Sci 33:2891, 1992
50. Dielemans I, de Jong PTVM, Stolk R et al: Primary open-angle glaucoma, intraocular pressure, and diabetes mellitus in the general elderly population. The Rotterdam Study. Ophthalmology 103:1271, 1996
56. Büchi ER: Cell death in the rat retina after a pressure-induced ischaemia-reperfusion insult: An electron microscopic study. I. Ganglion cell layer and inner nuclear layer. Exp Eye Res 55:605, 1992
59. Dandona L, Hendrickson A, Quigley HA: Selective effects of experimental glaucoma on axonal transport by retinal ganglion cells to the dorsal lateral geniculate nucleus. Invest Ophthalmol Vis Sci 32:1593, 1991
65. Leibowitz HM, Krueger DE, Maunder LR et al: The Framingham Eye Study monograph: An ophthalmological and epidemiological study of cataract, glaucoma, diabetic retinopathy, macular degeneration, and visual acuity in a general population of 2631 adults, 1973-1975. Surv Ophthalmol 24(suppl):335, 1980
89. Quigley HA, Addicks EM: Chronic experimental glaucoma in primates. II. Effect of extended intraocular pressure elevation on optic nerve head and axonal transport. Invest Ophthalmol Vis Sci 19:137, 1980
102. Dielemans I, Vingerling JR, Algra D et al: Primary open-angle glaucoma, intraocular pressure, and systemic blood pressure in the general elderly population. The Rotterdam Study. Ophthalmology 102:54, 1995
116. Weitzman ED, Henkind P, Leitman M et al: Correlative 24-hour relationships between intraocular pressure and plasma cortisol in normal subjects and patients with glaucoma. Br J Ophthalmol 59:566, 1975
155. Quigley HA, Addicks EM, Green WR: Optic nerve damage in human glaucoma. III. Quantitative correlation of nerve fiber loss and visual field defect in glaucoma, ischemic neuropathy, papilledema, and toxic neuropathy. Arch Ophthalmol 100:135, 1982
195. Kaufman PL, Bárány EH: Loss of acute pilocarpine effect on outflow facility following surgical disinsertion and retrodisplacement of the ciliary muscle from the scleral spur in the cynomolgus monkey. Invest Ophthalmol 15:793, 1976
207. Alm A, Stjernschantz J. Scandinavian Latanoprost Study Group: Effect on intraocular pressure and side effects of 0.005% latanoprost applied once daily, evening or morning. A comparison with timolol. Ophthalmology 102:1743, 1995
208. Warwar RE, Bullock JD, Ballal D: Cystoid macular edema and anterior uveitis associated with latanoprost use: Experience and incidence in a retrospective review of 94 patients. Ophthalmology 105:263, 1998
234. Heuer DK, Lloyd MA, Abrams DA et al: Which is better? One or Two? A randomized clinical trial of single-plate versus double-plate Molteno implantation for glaucomas in aphakia and pseudophakia. Ophthalmology 99:1512, 1992
245. Kosoko O, Gaasterland DE, Pollack IP et al: Long-term outcome of initial ciliary ablation with contact diode laser transscleral cyclophotocoagulation for severe glaucoma. Ophthalmology 103:1294, 1996
250. Flammer J, Guthauser U, Mahler F: Do ocular vasospasms help cause low tension glaucoma? In Greve EL, Heijl A (eds): Seventh International Visual Field Symposium, Amsterdam, September 1986, Martinus Nijhoff/Dr W. Junk Publishers, Dordrecht, Netherlands, 1987
251. Rojanapongpun P, Drance SM: The response of blood flow velocity in the ophthalmic artery and blood flow of the finger to warm and cold stimuli in glaucomatous patients. Graefes Arch Clin Exp Ophthalmol 231(7):375, 1993
257. Pellegrini-Giampietro D, Cherici G, Alesiani M et al: Excitatory amino acid release and free radical formation may cooperate in the genesis of ischemia-induced neuronal damage. J Neurosci 10:1035, 1990