Hepatic metabolism reduces the blood level of most drugs absorbed from the gastrointestinal tract. However, drugs absorbed into the blood vessels of the conjunctiva, lacrimal drainage apparatus, and nasopharynx avoid this “first pass” effect of the liver. As a result, surprisingly high plasma levels may be achieved from an eye drop despite loss of a large portion from overflow onto the lids. The maximum plasma level after oral timolol maleate, 5 mg, was 17 ng/ml of plasma.³ The amount of β-blocker in a 40-μL eye drop is from 0.1 mg in a 0.25% solution (e.g., timolol maleate) to 0.4 mg in a 1% solution (e.g., carteolol). Two drops of timolol 0.5% given bilaterally twice a day for 2 weeks produced plasma concentrations of <1 ng/mL to >5 ng/mL.⁴ Timolol could be detected in all the urine samples assayed. In another study, timolol 0.5%, two drops twice a day to both eyes, was given for 5 days; the maximum plasma level was 9.6 ng/mL.⁵ Because urine levels were the same after both the first and ninth dose, drug accumulation did not appear to occur. Timolol 0.5%, one drop to each eye, was given to patients maintained on one drop bilaterally twice a day.⁶ Before the test set of drops, the baseline mean timolol level was 0.34 ± 0.3 ng/mL plasma. One hour after the test set of drops, the maximum mean value of 1.39 ng/mL was reached; the single highest value was 2.45 ng/mL. Children on maintenance timolol 0.25% or 0.5% to one or both eyes had their blood samples drawn randomly during the course of the day. The lowest timolol level, 3.5 ng/mL, was in a 5-year-old child, and the highest, 34 ng/mL, was in a 3-week-old infant.

Levobunolol, 0.5%, one drop bilaterally, produced detectable blood levels within 1 hour, at which time the mean plasma levels were 0.3 ng/mL. The highest plasma level was 1.1 ng/mL.⁷ Forty microliters of 0.5% betaxolol was instilled into the inferior cul-de-sacs bilaterally in 15 subjects⁸. The mean peak plasma concentration was 1.1 ± 0.8 ng/mL and the mean time to peak value was 89 ± 79 minutes.

Twenty microliters of betaxolol 0.5% was applied bilaterally into the eyes of nine glaucoma patients at 12-hour intervals. After the second dose, there were two plasma peaks. The first was at 8±4 minutes and the second at 210 ± 132 minutes. The respective mean ± SDng/mL plasma levels were 1.1 ± 0.3 and 2.0 ± 1.1 ng/mL. Twelve hours after the first dose, just before the second dose, the mean plasma level was 0.4 ± 0.2 mg/mL.⁹

These levels of β-blockers are capable of producing pharmacologic effects. The β-adrenergic stimulatory effects of intravenous isoproterenol are inhibited by plasma timolol levels well below 1 ng/mL.^10,11 Isoproterenol was infused in healthy volunteers at increasing rates until the heart rate was raised 25 beats per minute (bpm) above baseline. The isoproterenol infusion was then discontinued. A single dose of timolol 0.25 mg was then infused during a 15-minute period. Plasma timolol levels were 2.3 ± 0.4 ng/mL at the end of the infusion and 0.7 ± 0.1 and 0.4 ± 0.1 ng/mL at 1 and 4 hours, respectively. The isoproterenol infusions were repeated. Timolol decreased the sensitivity to isoproterenol. The dose of isoproterenol had to be increased 6.4 ± 1.2-fold at 1 hour and 3.9 ± 0.9-fold at 4.5 hours after timolol to achieve the same pulse rate elevation of 25 bpm. By Schild’s equation (log of isoproterenol infusion plotted against the minus log of timolol plasma concentration), 0.17 ng of timolol per milliliter of plasma was the pA2 (i.e., the plasma level of timolol requiring a doubling of the amount of isoproterenol infused to produce the same response as in timolol’s absence).

Timolol’s plasma half-life is 4.1 ± 1.1 hours. The drug is metabolized extensively by the liver, primarily by microsomal oxidative reactions that result in the cleaving of its morpholino ring¹² Debrisoquin has been used as a probe to determine the genetic make-up of individuals with regard to oxidative metabolism of drugs. About 9% of the white British and North American populations are poor metabolizers of debrisoquin, and the defect is transmitted as an autosomal recessive trait¹³ The abnormal phenotype has been found in <1% of Asians.^14,15 The enzyme involved is cytochrome P450 (CYP2D6).¹⁶

Timolol metabolism is impaired in subjects with slow debrisoquin metabolism.^17,18 These subjects, given a single oral dose of timolol, 20 mg, developed significantly higher plasma levels than normal metabolizers of debrisoquin. The maximum timolol plasma concentrations, timolol plasma concentrations at 24 hours and timolol plasma half lives, respectively, for slow and rapid debrisoquin metabolizers were 113.8 ± 20.6 versus 60.7 ± 38.6 ng/mL, 11.3 ± 9.6 ng/mL versus none detectable, and 7.5 ± 3 versus 3.7 ± 1.7 hours. Several other β-blockers (e.g., alprenolol, metoprolol, and propranolol) have been identified as being inactivated primarily by oxidation reactions involving the same microsomal enzyme as debrisoquin.¹⁹ Furthermore, other drugs may compete for or otherwise alter this pathway, resulting in elevated and prolonged plasma drug levels in both normal and defective debrisoquin metabolizers. For example, quinidine inhibits the cytochrome P450 isoenzyme responsible for debrisoquin metabolism even though quinidine is not metabolized by this enzyme.²⁰ Quinidine has been implicated in a toxicity reaction with ophthalmic timolol.²¹ Bradycardia (36 bpm) was found in a 70-year-old man taking oral quinidine bisulfite, 250-mg sustained-release tablets, and timolol, 0.5% eye drops. The serum quinidine level, 3.5 mcg/mL, was at a therapeutic, not toxic, level. The bradycardia resolved when both drugs were discontinued, and it did not recur when ocular timolol alone was administered. However, when oral quinidine was added, the bradycardia returned; and when it was discontinued a second time, the bradycardia resolved. In a well-controlled study,¹⁰ the cardiac response to intravenous isoproterenol of healthy, normal debrisoquin metabolizers was compared after 0.25 mg of intravenous timolol with and without pretreatment with 100 mg of oral quinidine. Quinidine was shown to increase both the timolol plasma level and the timolol cardiac β-blocking effect by 10% to 40%. Cimetidine, an H₂ histamine receptor antagonist and diphenhydramine, an H₁ histamine receptor antagonist also inhibit cytochrome P450 (CYP2D6).²² The cardiovascular effects from timolol ½% eye drops were not significantly enhanced by oral use of 400 mg cimetidine daily.²³ However, with 13 subjects whose CYP2D6 status was known, eight rapid debrisoquin metabolizers and five slow, a convincing correlation was shown.²⁴ Two drops of 0.5% timolol ophthalmic or placebo were instilled into each nostril. The exercise heart rate was significantly lower (p = 0.01) and the plasma timolol concentration was significantly higher (p = 0.03) in the poor metabolizers. Oral quinidine, 50 mg, given 30 minutes before the timolol eye drops produced significant further reductions in heart rate (p = 0.02) and increases in plasma timolol levels (p = 0 .04) in the rapid metabolizers. The quinidine-associated increase in timolol plasma levels in fast metabolizers made them insignificantly different from slow metabolizers.

Studies in the ophthalmic literature tend to evaluate cardiac β₁-blocking activity by monitoring the daytime pulse at rest. The result is that systemic absorption of eye drops seems to produce little or no effect. At rest, healthy individuals have low levels of circulating epinephrine and of sympathetic neural activity; as a result the heart rate is controlled primarily by vagal tone. When the heart is stressed, a reversal of this situation occurs. Only then may the degree of β-blockade become evident. β-Blocking effects that appear trivial in healthy subjects at rest may be of serious consequence in the sick and elderly when they are stressed. A single oral dose of timolol maleate, 10 mg, will reduce the pulse rate in healthy volunteers by 13 to 16 bpm.²⁵ However, the stroke volume increases, resulting in maintenance of cardiac out-put at rest and largely compensating for the pulse rate reduction during exercise.²⁶ Healthy male volunteers given timolol 0.5% twice a day bilaterally showed no changes in their resting pulses after the first and ninth doses.⁵ However, exercise tachycardias were significantly reduced. Approximately 1 and 4 hours after the first and ninth drops, exercise resulted in mean pulse rates of 153 (placebo, 1 hour after first drop) versus 143 (timolol, 1 hour after first drop), 158 (placebo, 4 hours after first drop) versus 150 (timolol, 4 hours after first drop), 149 (placebo, 1 hour after ninth drop) versus 141 (timolol, 1 hour after ninth drop), and 152 (placebo 4 hours after ninth drop) versus 147 (timolol 4 hours after ninth drop). In other studies, 15 months of timolol 0.5% or levobunolol 0.5%, twice daily bilaterally,²⁷ produced significant reductions in the resting pulses of 5 to 10 bpm; betaxolol 0.25% eye drops twice a day for 6 weeks did not change resting pulse;²⁸ metipranolol 0.3%, twice a day bilaterally for 6 weeks, did not significantly alter resting pulse rates;²⁹ and after a 3-month trial of timolol 0.5% or carteolol 1%, there were no significant reductions in pulse rate from pretreatment levels.³⁰ No significant differences were found in the resting pulse effects of betaxolol 0.5% and levobunolol 0.5%³¹ or between timolol 0.25% and carteolol 1%.³² In a crossover, masked, placebo-controlled study, a single drop of timolol 0.5% or betaxolol 1% had no effect on the resting pulse.³³ However, after a 10-minute stress test on the treadmill, the mean pulse rate was 141 bpm after the placebo and betaxolol drops, but was significantly reduced to 132 bpm by the timolol drop. In another crossover study,³⁴ one drop of bilateral timolol 0.5%, betaxolol 0.5%, and carteolol 2% was given in at least three different sessions. Forty-five minutes later, the β-agonist isoproterenol was infused. The β-blockade by timolol and carteolol was sufficient to require that a fourfold higher dose of isoproterenol was needed to produce the same increase in heart rate as when the placebo drop had been given. The betaxolol drop had no more effect than placebo. Timolol 0.5%, carteolol 2%, or metipranolol 0.6% was given once to each eye at different times.³⁵ Although there were no changes in the resulting blood pressure or pulse, the doses of intravenous isoproterenol needed to increase the pulse 25 bpm were increased significantly for all three drugs: after placebo, 3.1 ± 0.5 mcg isoproterenol; after timolol, 10.9 ± 1.9 mcg isoproterenol; after carteolol, 39.6 ± 5.4 mcg isoproterenol; and after metipranolol, 5.2 ± 0.9 mcg isoproterenol.

The opposite extreme from exercise is sleep. The mean pulse rate is decreased. Different topical β-blockers have varying effects during sleep, usually small but often statistically significant. Occasionally the effect is clinically significant. In a randomized, double-masked prospective study 82 patients received carteolol 1% and 87 patients received timolol 0.5% eye drops twice a day for 4 weeks.³⁶ At the conclusion of this period, both drugs reduced the daytime pulse rate similarly, 4 to 6 bpm. However, from midnight to 4:00 AM, the mean ±SEM carteolol-treated pulse rate (70.7 ± 0.7) was slightly higher than the corresponding pretreatment period (69.1 ± 1.1), whereas the corresponding timolol treated pulse rate fell (66.0 ± 0.7 posttreatment vs. 68.4 ± 1.1 pretreatment). The different effects of the two drugs were significant, p <0.001, each of the 4 hours. Four times as many patients became bradycardic (<60 bpm) on timolol during this 4-hour period (18.4% vs. 4.5%).

The reduction in pulse rate produced by β₁-blockers was formerly believed to contraindicate their use in chronic congestive failure. In a sea-change, β₁-blockers have become a lynch-pin in its treatment. The antiarrhythmic and antirenin effects of β₁-blockers provide major benefits. Sudden death from ventricular fibrillation is reduced³⁷ Although eye drops with β₁-blocking activity need not be feared in chronic congestive heart failure, their use may be dangerous in acute cardiac decompensation and bradycardia. Pre-existing bradycardia may indicate disease of the cardiac conduction system. A superimposed β₁-blocking effect can lead to asystole.

The mechanisms by which β-blocking agents reduce blood pressure are not understood. Blood vessel smooth muscle relaxes when stimulated by β₂-agonists; a β₂-antagonist would, therefore, be expected to increase, not decrease, blood pressure. Angiotensin II production by the juxtaglomerular cells is reduced by β-adrenergic blockade, lowering peripheral arterial resistance and blood pressure. However, β-blockers are effective even when hypertension is not associated with increased levels of angiotensin.

Although systemic absorption of ophthalmic β-blockers significantly affects postexercise heart rate, there appears to be less of an effect on blood pressure. Single drops given bilaterally of timolol 0.5%, carteolol 2%, or metipranolol 0.6% produced no changes in the resting blood pressures of healthy volunteers.³⁵ One drop of timolol 0.5%, betaxolol 1%, or placebo did not affect the blood pressure at rest or during treadmill testing.³³ Six weeks of metipranolol 0.3% twice a day bilaterally did not affect resting mean systemic systolic or diastolic blood pressures.²⁹ Six weeks of betaxolol 0.25% eye drops twice a day also failed to alter the resting blood pressure.²⁸ Three months of timolol 0.5% or carteolol 1% or 2% eye drops failed to cause a significant difference in resting blood pressure.³⁰ Fifteen months of timolol 0.5% or levobunolol 0.5% twice a day bilaterally reduced the resting systolic and diastolic blood pressures by <4 mmHg.²⁷ Bilateral timolol 0.5% drops failed to significantly alter the postexercise systolic or diastolic blood pressure after the first or ninth dose.⁵