Carbonic anhydrase inhibitors

Carbonic anhydrase inhibitors (CAIs) continue to be the only systemic agents used for the long-term treatment of glaucoma, if only occasionally. A topical version introduced more than 40 years after the introduction of the systemic agent moved the oral agents quite far down on the list of practical options for the chronic treatment of glaucoma. These drugs are all derivatives of sulfonamides and were introduced into clinical practice as diuretics. Even with chronic use, their diuretic action is only effective for 1–2 weeks. CAIs reduce intraocular pressure (IOP) by decreasing aqueous humor formation. As such they are useful in essentially all forms of glaucoma, even when the anterior chamber angle is sealed or the outflow facility is very low.

Acetazolamide was introduced into clinical practice as an antiglaucoma drug in 1954 and remains the prototype for this class of drugs. Other members of this group include the oral agents, methazolamide, ethoxzolamide, and dichlorphenamide, as well as the topical agents, dorzolamide and brinzolamide.

Carbonic anhydrase inhibitors have also been important tools for the study of aqueous humor dynamics. Their ability to alter aqueous humor formation has been useful in understanding the chemical composition, the turnover of substances, and the rate of aqueous humor formation.


The enzyme carbonic anhydrase (CA) catalyzes the following reaction:

<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='CO2+OH-CAHCO3-‘>CO2+OHCAHCO3CO2+OH-CAHCO3-
CO 2 + OH –  CA HCO 3 –

This enzyme is found in many tissues in the body, including the renal cortex, gastric mucosa, red blood cells (RBCs), lung, pancreas, and central nervous system (CNS). It is also found in many tissues in the eye, including the corneal endothelium, non-pigmented iris epithelium, pigmented and non-pigmented epithelium of the ciliary processes, Müller cells, and retinal pigment epithelium. Carbonic anhydrase exists in multiple forms. Although type I and II CAs are both present in the corneal endothelium and lens, the type II isoenzyme (type C in another classification system) appears to be the only one of the two forms present in any quantity in human ciliary epithelium. Recent animal studies have placed both types III and IV CA in the non-pigmented ciliary epithelium of both rabbits and humans and have implicated it as having a role in aqueous production.

The notion that CAIs reduce aqueous humor formation is amply supported by a number of experimental and clinical studies involving tonography, fluorophotometry, fluorescein appearance time in the anterior chamber, photogrammetry, changes in the steady-state concentration of endogenous ions (e.g., ascorbate, phosphate) in the anterior and posterior chambers, turnover of systemically administered substances (e.g., ascorbate, p -aminohippuric acid, urea, iodide, iodopyracet) in the anterior and posterior chambers, and dilution techniques measuring the loss of substances infused into the anterior or posterior chambers (e.g., inulin or isotopically labeled protein). Although all of these techniques have underlying assumptions and shortcomings, there is general agreement that CAIs in full doses reduce aqueous humor formation by about 40%. This means that at least 60% of aqueous humor formation is independent of the enzyme CA. This limits not only the efficacy of the inhibitors but also their potential ocular side effects. More than 99% of the enzyme activity must be inhibited before aqueous production is reduced.

There has been considerable debate about the mechanism by which CAIs decrease aqueous humor formation. This debate reflects our imperfect understanding of aqueous humor formation and the contradictory results of studies performed in different animal species. Among the issues under discussion are whether CAIs play a direct or indirect role in reducing aqueous humor formation and whether the effect is primarily ocular or systemic. With the advent of a successful topical CAI that does not change systemic parameters, that aspect of the debate is largely settled.


The bulk of evidence strongly suggests that CAIs reduce aqueous humor formation by a direct effect on ciliary epithelial CA:

  • 1

    Carbonic anhydrase is found in the non-pigmented epithelium of the ciliary processes. This metabolically active tissue is thought to be responsible for the secretion of aqueous humor.

  • 2

    Acetazolamide inhibits aqueous humor secretion by isolated rabbit ciliary processes.

  • 3

    Intravenous injection of acetazolamide, 125 mg, does not lower IOP in patients with a congenital deficiency of CA II.

  • 4

    Small doses of CAIs can reduce IOP while producing minimal or no systemic acidosis or electrolyte imbalance.

  • 5

    The CAIs are capable of reducing IOP in nephrectomized rabbits. Friedman and co-workers found that intravenous injection of acetazolamide, 5 mg/kg, to nephrectomized rabbits lowered IOP without an effect on arterial pH, partial arterial pressure of carbon dioxide bicarbonate, or base excess. Furthermore, acetazolamide lowers IOP in elasmobranches that lack renal CA.

  • 6

    Topical application of CAIs lowers IOP without inducing changes in systemic acid–base or electrolyte balance. Furthermore, topical CAI administration to one eye with good reduction in IOP fails to significantly lower IOP in the fellow eye. While early studies failed to detect an effect of topical CAIs on IOP, this probably represented a lack of penetration of the drugs to active sites in the ciliary epithelium.

  • 7

    Intracarotid injection of acetazolamide in animals reduces IOP on the ipsilateral but not the contralateral side.

  • 8

    Carbonic anhydrase inhibitor administration reduces bicarbonate movement into the posterior chamber of rabbit, dog, and monkey eyes. It is known that the rabbit eye has a posterior chamber bicarbonate concentration in excess of plasma as reflected by measurement of both cold concentration and C 14 CO 2 . Furthermore, the entry of bicarbonate into the rabbit posterior chamber is slowed by acetazolamide. The concentration of any ion, including bicarbonate, in the posterior chamber is affected by many factors (e.g., secretion, absorption, exchange). Acetazolamide does reduce the movement of bicarbonate from the plasma to the posterior chamber by 30–50%. It is likely that the CAIs also reduce chloride ion transport into the posterior chamber.


A few investigators have proposed that CA plays an indirect role in aqueous humor formation. It has been suggested that in various species the non-pigmented ciliary epithelium secretes either an acidic or a basic aqueous humor and then requires CA to generate hydrogen or bicarbonate ions for an intracellular buffering system. Such a system would help maintain intracellular pH for the enzymes involved in ion transport. Carbonic anhydrase inhibitors would interfere with this buffering system and indirectly reduce aqueous humor formation. Although this proposed mechanism has generated considerable discussion, there is no direct proof to support the hypothesis.

Some authorities propose that CAIs influence aqueous humor formation by altering systemic acid–base or electrolyte balance. Systemic administration of CAIs produces diuresis and metabolic acidosis. However, the diuresis and the concomitant loss of sodium, potassium, and bicarbonate in the urine are transient in nature and cannot explain the long-term lowering of IOP that is produced by CAIs. Other diuretic agents that are more potent and more persistent in their effects do not lower IOP. Furthermore, pretreatment of animals with ammonium chloride prevents acetazolamide-induced diuresis yet does not block drug-induced IOP reduction. Finally, in glaucomatous dogs, systemic administration of methazolamide does not increase the IOP-lowering effect of topical dorzolamide.

Systemic acidosis from either disease (e.g., diabetic coma) or the administration of pharmacologic agents (e.g., ascorbic acid or calcium chloride) is associated with decreased IOP. Some investigators believe that CAI-induced IOP lowering correlates with changes in blood pH. Other authorities dispute this conclusion and believe that the time course and magnitude of the ocular hypotensive response do not correlate well with changes in blood pH ; for example, IOP declines in rabbits before acetazolamide causes any measurable effect on the kidney or blood pH.

A variety of other mechanisms has been proposed to explain the IOP reduction seen after CAI administration. Macri and Cevario postulated that acetazolamide lowers IOP by producing vasoconstriction in the anterior uveal tract. However, using a labeled microsphere technique in rabbits, Bill found no effect of acetazolamide on the blood flow to a number of ocular tissues, including the anterior uveal tract. Lütjen-Drecoll and co-workers noted that CAIs block the development of capillary fenestrations in a variety of tissues. They postulated that this phenomenon might be involved in the therapeutic effect of these drugs. Macri reported decreased episcleral venous pressure after CAI administration. Thomas and Riley postulated that CAIs act through an adrenergic mechanism because the effects of the drugs in animals are altered by adrenalectomy and adrenergic blocking agents.

As stated previously, we lack a complete understanding of the mechanism of action of CAIs. The best evidence suggests that they lower IOP by reducing bicarbonate (and possibly chloride ion) movement and aqueous humor formation via a direct effect on ciliary epithelial CA. All of the CAIs share a common structure of the organic anion SO 2 NH2 ( Fig. 26-1 ). This suggests that CAIs may compete with the OH2 ion at the active site on the enzyme. The ocular hypotensive response is probably augmented by the induced systemic acidosis, which would explain the small increase in effect of systemic over topical agents in humans.

Fig. 26-1

Carbonic anhydrase inhibitors.



Because of the high incidence of systemic side effects with oral CAIs, investigators have long searched for active topical CAIs. Early reports indicated that topical CAIs were ineffective and speculated that circulating RBCs formed a sink that took up any drug reaching the ciliary body. However, more recent investigations indicate that topical CAIs are quite effective in lowering IOP when administered in high concentration, with a soft contact lens, as an analogue, or in an appropriate delivery system. As noted previously, topical CAIs are capable of lowering IOP while producing no effect on systemic electrolyte or acid–base balance. With the demonstration that a topical CAI could work, it was only a matter of time until a clinically useful agent would be developed.

Several sulfonamide-derived topical agents were examined. The original agent (ethoxzolamide gel) showed promise, but both the drug and its vehicle were noted to cause bulbar irritation in a very high percentage of patients. MK-927 and its more potent S -enantiomer MK-417 were given trials and found to have satisfactory but not impressive effectiveness. Finally, dorzolamide (formerly known as L-671, 152, then MK-507) was found to have the right balance of effectiveness and minimal local side effects. Several years after the introduction of dorzolamide, brinzolamide, another topical CAI, was found to have similar clinical efficacy. These agents seem to have an equivalent effect on aqueous humor inhibition and differ only in small ways as far as side effects are concerned.


Dorzolamide (Trusopt, Merck, West Point, Penn) differs from the oral agents in that it has both a free sulfonamide group and a second amine group, which adds the right amount of lipid and aqueous solubility for good corneal penetration. Dorzolamide is effective in inhibiting isoenzymes II and IV, with a somewhat weaker effect on isoenzyme I. There is little crossover effect on the other eye, confirming its local rather than systemic action. The drug is excreted by the kidneys largely unmetabolized, but a small amount is metabolized by the liver to the des -ethyl form.

The des -ethyl form has a significant inhibitory effect on isoenzyme I. Both dorzolamide and its metabolite are largely bound to RBC cholinesterase and are not present to any extent as free molecules in plasma. Systemic effects are minimal, but RBC CA is depressed to 21% of normal levels. After 8 days of topical therapy, virtually all of the RBC CA is inhibited. After 18 months of use, only about 1/200th of the concentration needed to induce significant systemic side effects is found in plasma. The depression of CA in the RBCs may be seen for months after the drug has been discontinued. Despite this profound depression of RBC CA, no systemic symptoms or problems have been attributed to it.

Dose–response studies of 0.7%, 1.4%, and 2.0% have shown that the 2% solution is the most effective, producing peak and trough IOP reductions of 21% and 13%, respectively, with twice-daily usage and slightly better trough reduction with dosing three times daily. The lower doses do have some effect, although they are not commercially available in the United States. Topical dorzolamide reaches a peak effect on IOP at about 3 hours after dosing compared with 2 hours with oral acetazolamide. At its peak, dorzolamide lowers IOP as well as does timolol maleate, but the effect does not last as long. In monkeys, dorzolamide 2% will produce a 38% reduction in aqueous inflow. In a 1-year study, topical dorzolamide 2% appeared to be as effective and as well tolerated as topical betaxolol 0.5%. Long-term studies have shown tolerability similar to or better than other topical agents such as timolol and pilocarpine. Dorzolamide is available commercially as a 2% solution to be used every 8 hours as monotherapy and every 12 hours as adjunctive therapy.

In pediatric practice, dorzolamide, although not quite as effective as oral therapy, clinically worked in most of the children tested, produced much less side effects, and was better tolerated than acetazolamide. In adults with glaucoma, dorzolamide was nearly equivalent to systemic acetazolamide in lowering IOP but was much better tolerated with an improved quality of life compared to the oral agent.

While probably not as effective as the prostaglandin analogs over the 24-hour period, dorzolamide appears comparable in pressure-lowering activity to both timolol and brimonidine when used three times daily in monotherapy. Long term, dorzolamide seems to maintain its effectiveness without tachyphylaxis and is reasonably well tolerated. Topical dorzolamide is generally well tolerated and reasonably effective in children even under 6 years of age.

Surprisingly, dorzolamide is additive to other aqueous suppressants such as timolol, both clinically and in measurements of aqueous formation, and to drugs acting on the outflow system such as pilocarpine and prostaglandins. In fact, dorzolamide showed slightly better additivity in a retrospective study of 73 glaucoma patients to latanoprost than β-blockers or brimonidine. This slight superiority to brimonidine as an adjunctive medication was confirmed in a prospectie study. In another prospective study, dorzolamide and carteolol were equally additive to latanoprost.

Dorzolamide has been used successfully in a fixed combination with timolol (Cosopt, Merck, West Point, Penn). The combination certainly is more convenient than using either alone and seems to improve compliance compared to using both agents separately. While probably not quite as effective as bimatoprost given once daily, the fixed timolol/dorzolamide combination does seem to be as effective as latanoprost. One retrospective study of a large database suggests that the timolol/dorzolamide combination shows better efficacy and persistency than a fixed brimonidine/timolol combination.


Another topical CAI, brinzolamide (Azopt, Alcon Laboratories, Fort Worth, TX), has become available. Brinzolamide was a known CAI that was irritating when used topically in solution because of its low pH and had a disappointing effect due to poor corneal penetration. Placing the compound in a suspension rather than in the traditional solution was found to improve its ability to get across the cornea, reduce the surface irritation, increase its pressure-reducing activity, and prolong the duration of action. In fact, topical brinzolamide 1% given three times daily to patients with glaucoma or ocular hypertension was shown in several masked studies to be at least as effective as dorzolamide 2% administered three times daily, with less discomfort on administration. In addition, twice-daily dosing with brinzolamide was equal in pressure-lowering efficacy to thrice-daily dorzolamide and twice-daily timolol maleate. These findings were duplicated in another randomized, masked study. In all of the studies comparing dorzolamide and brinzolamide to date, dorzolamide has a higher rate of stinging and pain on administration with brinzolamide having a higher rate of transient blurred vision from the suspension. Brinzolamide and latanoprost have been shown to be more effective when used concomitantly in keeping IOP controlled throughout the 24 hours in normal-tension glaucoma patients than either agent alone.

Like dorzolamide, brinzolamide lowers IOP an additional 13–16% when added to twice-daily timolol maleate therapy. Brinzolamide is additive in pressure-lowering activity when given with latanoprost. For all intents and purposes, brinzolamide 1% seems to be the therapeutic equivalent of dorzolamide. Some studies suggest that brinzolamide may be clinically more comfortable for patients than dorzolamide.


All of the commonly used oral CAIs produce similar IOP reductions and similar types of side effects when administered in equipotent doses. The various oral agents differ in potency and to some extent in efficacy; that is, they all produce a similar IOP reduction but at different doses. All of the compounds have a similar range of activity in vitro except for ethoxzolamide, which is five to ten times more active. The difference in potency of the various drugs in vivo reflects differences in lipid solubility and protein binding that affect body distribution ( Table 26-1 ).

Table 26-1

The chemical and pharmacologic properties of the systemic carbonic anhydrase inhibitors in clinical use

Name K 1 × 10 −9 pK a Unionized in plasma pH 7.4 (%) Unbound in plasma (%) Half-life in humans (hours)
Acetazolamide 6 7.4 50 5 4
Methazolamide 8 7.2 39 45 15
Ethoxzolamide 1 8.1 83 4 6
Dichlorphenamide 18 8.3 89 2

Modified from Friedland BR, Maren TH: Carbonic anhydrase: pharmacology of inhibitors and treatment of glaucoma. In: Sears ML, editor: Pharmacology of the eye: handbook of experimental ophthalmology, vol 69, Berlin, Springer-Verlag, 1984.


Acetazolamide (Diamox, Storz Ophthalmics, St Louis; Ak-Zol, Akorn, Inc., Abita Springs, LA) was the first CAI to receive widespread use in ophthalmology, and it remains the agent with which we have the greatest experience. It is supplied in 125 and 250 mg tablets and as a 500 mg sustained-release (SR) preparation. The 250 mg tablet administered four times daily produces an IOP reduction similar to that seen with the 500 mg SR preparation administered twice daily. The SR preparation is more convenient, not significantly more expensive per day, and better tolerated than are the tablets. In fact, the SR capsules are the fastest acting, most effective, and best tolerated at full dosage of the oral CAIs. It is rarely useful to prescribe acetazolamide in doses greater than 1000 mg/day.

After oral tablet administration, IOP begins to drop in 1–2 hours, reaches a minimum in 2–4 hours, and returns to baseline in 4–12 hours. After oral administration of a 500 mg SR preparation, IOP begins to drop in 2–4 hours, reaches a minimum in 8 hours, and returns to baseline in 12–24 hours. However, one study suggested that the SR preparation may be faster acting than previously thought.

Oral acetazolamide can be administered to infants in a dose of 5–10 mg/kg every 6 hours. To prepare the medication, the pharmacist must crush the tablets and suspend the powder in flavored syrup.

Acetazolamide is also available in 500 mg emergency ampules. The drug is dissolved in 5–10 ml of distilled water and then intravenously or intramuscularly administered in a dose of 250–500 mg. After intravenous injection, IOP begins to fall within minutes, reaches a minimum in 15–30 minutes, and returns to baseline in 4–6 hours. Parenteral injection is indicated in conditions associated with nausea and vomiting (e.g., acute angle-closure glaucoma) or when maximum IOP lowering is immediately imperative.

Acetazolamide penetrates the eye poorly because of high plasma binding and ready ionization. The serum half-life of the drug is approximately 4 hours. Acetazolamide is not metabolized and is actively secreted by the renal tubules and then passively resorbed by non-ionic diffusion. Older patients have a lower clearance of unbound acetazolamide, but this is mostly offset by a lower percentage of binding.

The IOP reduction produced by acetazolamide generally parallels the plasma level of the drug. The maximum IOP reduction is generally obtained with plasma acetazolamide concentrations of 4–20 μg/ml. Initially, acetazolamide causes a loss of sodium, potassium, and bicarbonate in the urine. The mild metabolic acidosis that develops with acetazolamide therapy is a consequence of the initial bicarbonate loss. However, the electrolyte balance soon reaches a new steady state despite continued treatment. Bicarbonate resorption independent of carbonic anhydrase prevents further loss of this ion and progressive acidosis.


Methazolamide (Neptazane, Storz Ophthalmics, St Louis; Glauctabs, Akorn, Inc., Abita Springs, LA; MZM, Ciba Vision Ophthalmics, Duluth, GA) is supplied in 25 and 50 mg tablets. It is best to initiate methazolamide therapy with a low dose of the drug (e.g., 25 mg twice daily) and to increase this dose as required to control IOP. Low doses of methazolamide often reduce IOP while producing minimal acidosis and electrolyte disturbance as well as a low incidence of side effects. Higher doses of the drug produce greater IOP reductions but also greater acidosis and a higher incidence of side effects. The most common treatment regimen for methazolamide is 50–100 mg twice daily. Methazolamide, 50 mg twice daily, is slightly less effective than is acetazolamide, 250 mg four times daily or 500 mg SR preparations twice daily. After administration of a 50 mg tablet, IOP begins to drop in 1–2 hours, reaches a minimum in 4–6 hours, and returns to baseline in 12–24 hours. Because methazolamide has a serum half-life of 14 hours, it is unnecessary to administer the drug more frequently than twice daily.

Methazolamide is not actively secreted by the kidneys. Approximately 25% of the drug appears unchanged in the urine. The metabolic fate of the remainder is unknown, although some appears to be converted by glutathione. The metabolism of methazolamide makes it a safer choice than acetazolamide for patients with advanced renal disease (e.g., a diabetic patient with neovascular glaucoma). Methazolamide has some advantages over acetazolamide. First, methazolamide diffuses into the eye more easily than does acetazolamide. This probably reflects in part the fact that methazolamide is less bound to plasma protein. Second, methazolamide is not actively taken up by the renal tubules as is the case with acetazolamide. Third, its duration of action makes it more convenient to use (twice daily) than acetazolamide tablets (four times daily). Some authorities postulate that methazolamide is less likely than is acetazolamide to produce urolithiasis because it produces less suppression of urinary citrate and less urine alkalinization. However, methazolamide therapy has been reported to cause urinary tract stones. Whether methazolamide actually has a lower incidence of urinary calculi than does acetazolamide has not been established, but most authorities prefer methazolamide to acetazolamide in patients with a history of renal lithiasis. Methazolamide diffuses more easily into the eye and CNS. Thus it is more likely than acetazolamide to produce such CNS-related symptoms as fatigue, depression, and drowsiness.


Ethoxzolamide is the most potent of the clinically used CAIs in vitro. However, its in-vivo activity is reduced by high plasma protein binding. Ethoxzolamide is supplied in a 125 mg tablet and prescribed in doses of 62.5–250 mg every 4–8 hours. The most commonly prescribed dose is 125 mg every 6 hours. After administration of a 125 mg tablet, IOP begins to fall in 2 hours, reaches a minimum in 5 hours, and returns to baseline in 12 hours.

Ethoxzolamide is a weak organic acid that is secreted slightly by the renal tubules. Forty per cent of the drug appears unchanged in the urine. The metabolic fate of the remainder is not entirely known, but some is converted by glutathione. Ethoxzolamide was the first CAI to be successfully used topically. This agent has essentially disappeared from current use in the United States.


Dichlorphenamide (Daranide, Merck, West Point, Penn) is supplied in a 50 mg tablet and prescribed in doses of 25–200 mg every 6–8 hours. After oral administration of a 50 mg tablet, IOP begins to fall in 30 minutes, reaches a minimum in 2–4 hours, and returns to baseline in 6–12 hours.

Despite the fact that the dichlorphenamide molecule contains two sulfonamide groups, it is no more effective than the other CAIs. Dichlorphenamide produces less metabolic acidosis because it has inherent chloruretic activity. However, the continued loss of chloride sometimes produces sustained diuresis and potassium depletion. Dichlorphenamide produced more symptoms and side effects than did the other CAIs in one trial. It is little used at this time.



The side effects of the topical CAIs are largely ocular in nature ( Box 26-1 ). Stinging on administration is the most common especially with dorzolamide. Some patients experience actual ocular pain. As noted above, brinzolamide often produces a transient blurring of vision. Presumably because of their sulfonamide derivation, the topical agents are associated with a relatively high rate of allergic reactions (about 10%), as are the oral agents. Contact dermatitis may be seen on the eyelids.

Box 26-1


  • Stinging

  • Allergy

  • Dryness

  • Superficial punctate keratopathy

  • Induced myopia


  • Metallic taste

  • Urticaria

  • Neutropenia

  • Headache

  • Gastrointestinal distress

  • Dizziness

  • Paresthesias

  • ?Aplastic anemia

  • ?Stevens-Johnson syndrome

Side effects of topical carbonic anhydrase inhibitors

Superficial punctate keratopathy is seen in about 10% of cases. Transient myopia has been reported. All of the major topicalantiglaucoma medications can be associated after long-term use with squamous metaplasia of the conjunctiva and the topical CAIs are no exception. What this finding portends clinically is not known, although rare patients may react to long-term topical medication use with a syndrome that looks like ocular cicatricial pemphigoid.

As can happen with any sulfa-derived medication, choroidal detachments with induced myopia and hypotony have been reported. A prospective study on 34 glaucoma patients of the effect of dorzolamide on axial length and refraction failed to show any change in refraction after 2 weeks of use.

The corneal endothelium is rich in CA; this fact caused concern that CA inhibition by topical agents might interfere with endothelial function. Using ultrasonic pachymetry, Wilkerson and co-workers showed that topical dorzolamide administered over 4 weeks increases the corneal thickness very slightly compared with placebo; however, it is unlikely that this small change is clinically significant, as confirmed by Kaminski and co-workers. Furthermore, Serle and co-workers failed to show any change in corneal thickness after 6 weeks of application. In a multicenter study, Lass and co-workers were unable to demonstrate any differences in corneal thickness or specular microscopy after 1 year of topical therapy with either dorzolamide, timolol, or betaxolol. In a controlled study spanning over 6 years, Baratz and co-workers were unable to show any changes in specular or confocal microscopy of the cornea with chronic glaucoma therapy versus no treatment. Finally, one study induced acute corneal edema through contact lens overwear in 19 patients with glaucoma or ocular hypertension without finding any effect of dorzolamide on recovery time. These were otherwise healthy corneas and the results might not be generalizable to those with borderline or worse endothelial function. On the other hand, several small series have shown an increase in corneal thickness after chronic dorzolamide therapy.

Despite the lack of evidence that dorzolamide or brinzolamide affect corneal thickness or microscopic structure in normal corneas, case reports have surfaced of corneal decompensation in patients treated with dorzolamide. Whether this is an effect of the carbonic anhydrase inhibition or whether these corneas had Fuch’s dystrophy or were predisposed to aphakic or pseudophakic bullous keratopathy is not clear at this time. Clearly, this is an uncommon or even rare complication, but until further information is received, caution is indicated regarding the use of topical CAIs in those patients with borderline corneal endothelial function. Other agents should be tried first. However, CAIs are probably less toxic to the endothelium than intraocular surgery so they could be tried with careful monitoring.

Although serious systemic side effects are not common, some do occur. A metallic taste in the mouth, especially associated with carbonated beverages, is relatively common (25%). Obviously, systemic side effects can be reduced by having the patient use simple eyelid closure and punctal occlusion immediately after administering the drops. Some of the more serious side effects associated with the oral agents (e.g., aplastic anemia and Stevens-Johnson syndrome) have not been reported with topical agents; however, anecdotal reports of neutropenia suggest that it may only be a matter of time before the rare case is actually seen. Gastrointestinal distress may occur particularly in the first few days of use. Urticaria and dizziness have also been reported in rare instances as has bullous pemphigoid. Although RBC CA is depressed with topical dorzolamide, no symptoms appear to be associated with this finding. Nephrolithiasis has been reported in one patient. Erythema multiforme has been associated with topical dorzolamide use. Other side effects, especially those associated with systemic agents, may be seen as the topical agents are used in more patients for longer periods.


It is estimated that 50% of glaucoma patients cannot tolerate long-term treatment with oral CAIs because of the associated side effects ( Box 26-2 ). The etiology of many of the side effects is unclear but may be related to acidosis or carbon dioxide retention. It is important to emphasize that the incidence and severity of side effects can be reduced greatly by using the medication in the lowest dose and frequency necessary for IOP control. Patients should be warned of potential side effects and told that many side effects will diminish in severity after a few days to a few weeks of treatment. Patients prepared in this manner are more trusting of the physician and are more ready to accept problems when they occur. Older patients are generally less tolerant of oral CAIs.

Box 26-2

Myopic shift *

Paresthesias of fingers, toes, circumoral region *

  • * Effects are either common or of major clinical concern.

    Decreased dexterity

Electrolyte disturbances

  • Metabolic acidosis *

  • Potassium depletion associated with concomitant use of diuretics or corticosteroid

  • Chloride depletion associated with use of dichlorphenamide

  • Uric acid retention


  • Abdominal cramping/discomfort *

  • Metallic taste to carbonated beverages *

  • Nausea

  • Diarrhea *

  • Anorexia

  • Weight loss *

  • Constipation


  • Nocturia

  • Urolithiasis *

  • Impotence

  • Frequency with polydypsia * (especially in the first week of treatment)

  • Hypersensitivity nephropathy *

Central nervous system

  • Malaise *

  • Excitement

  • Elevated cerebrospinal fluid pressure

  • Fatigue *

  • Confusion

  • Depression *

  • Drowsiness *

  • Headache

  • Decreased libido

  • Vertigo

  • Irritability *

  • Insomnia

  • Tremor

Blood dyscrasias

  • Thrombocytopenia *

  • Agranulocytosis *

  • Hyperchromic anemia

  • Aplastic anemia *

  • Neutropenia

  • Interference with anticholinesterase treatment of myasthenia gravis

  • Exacerbation of effect of diphenylhydantoin on bone demineralization

  • Dyspnea *

  • Leg cramps

  • Hearing loss

  • Birth defects

  • Hypersensitivity hepatitis


  • Rash

  • Exfoliative dermatitis * (Stevens-Johnson syndrome)

  • Pruritis *

  • Hair loss

  • Flushing

  • Hirsutism

Only gold members can continue reading. Log In or Register to continue

Feb 12, 2019 | Posted by in OPHTHALMOLOGY | Comments Off on Carbonic anhydrase inhibitors
Premium Wordpress Themes by UFO Themes