Bronchial Reactivity in Healthy Individuals Undergoing Long-term Topical Treatment With β-Blockers

Objective  To assess the impact of long-term treatment with topical timolol on bronchial reactivity in healthy individuals.

Methods  Twenty-one otherwise healthy individuals with high-pressure primary open-angle glaucoma were enrolled in a randomized controlled clinical trial. Eleven patients underwent 3 years of topical 0.5% timolol treatment followed by a 1-year washout period; 10 patients underwent primary argon laser trabeculoplasty. Functional variables and bronchial reactivity (forced expiratory volume in 1 second and metacholine challenge test results) were assessed in both groups at enrollment and after 3 and 4 years of follow-up.

Results  After 3 years, a measurable response to metacholine challenge was recorded in 6 of 11 otherwise symptom-free individuals treated with 0.5% timolol twice daily. A detectable response to metacholine challenge was still present in half of these individuals (3 of 6) when further washed out for 1 year from the topical β-blocker. No significant variation in bronchial reactivity was measured in the laser-treated group during 4 years of follow-up.

Conclusions  Healthy individuals who undergo long-term topical application of a nonselective β-blocker (0.5% timolol) can develop a subclinical increase in bronchial reactivity. This phenomenon may not be completely reversible on withdrawal of the β-blocker.

Topical antagonists of β receptors are widely used as first-line therapy in ocular hypertension and glaucoma.¹ Because of the chronic nature of the disease, patient exposure to the drug is expected to last for several years. Adverse effects due to systemic β-blockade are known to occur² This evidence limits the use of topical β-blockers in patients who report a history of constrictive lung disease or serious cardiac arrhythmia. In particular, β-blocker eyedrops has been reported to precipitate latent asthma.^3,4 Replacing topical timolol with either topical betaxolol hydrochloride or dipivefrin was associated with an increase in peak flow rate and forced expiratory volume in 1 second (FEV₁) in a group of elderly patients with glaucoma.⁵ When evaluating lung function in elderly people exposed to topical β-blockers, Diggory et al⁶ reported that 8 of 40 patients discontinued topical β-blockers after 1 month because of a greater than 15% decline in peak flow and a greater than 10% decline in FEV₁. Not every discontinuing patient was symptomatic.⁶ Many people experience unrecognized respiratory impairments when prescribed topical β-antagonists.⁵ In a recent study,⁷ topical timolol therapy caused a reduction of the respiratory function in patients with newly diagnosed glaucoma 12 months after initiation of treatment. Waldock et al^7(p713) pointed out their “concern in view of the potential subclinical reversible airways disease in the elderly glaucoma population.”

We, therefore, planned a long-term randomized controlled clinical trial to verify in otherwise healthy individuals with glaucoma (1) the extent of bronchial reactivity induced by the administration of 0.5% timolol twice daily and (2) the possible reversibility of any change in respiratory functions.

Thirty-two patients (17 women and 15 men; age range, 44-67 years) referred to the Glaucoma Service at the University of Parma for primary open-angle glaucoma were enrolled in this study. The inclusion criteria consisted of (1) intraocular pressure (IOP) greater than 22 mm Hg in both eyes (mean of the 2 highest readings of the daily phasing) and (2) glaucomatous visual field defects in at least 1 eye as assessed by computer-assisted static perimetry (Octopus G1; Interzeag, Koeniz-Berne, Switzerland); in the case of a unilateral field defect, the fellow eye had to show optic disc cupping consistent with glaucomatous optic neuropathy. The exclusion criteria consisted of (1) previous antiglaucoma treatment; (2) smoking; (3) a history of allergic and respiratory disease, including asthma; and (4) atopy as assessed by skin prick testing.

A randomized clinical trial was performed. On enrollment, each patient signed an informed consent form and was randomized according to age (≥50 vs <50 years) and IOP (≥25 vs <25 mm Hg) to undergo either 0.5% timolol treatment twice daily (n = 16) or primary argon laser trabeculoplasty without any further antiglaucoma medication use (n = 16). The study protocol and the informed consent form had previously been approved by the local Ethics Committee of the University of Parma. Follow-up lasted 4 years. In the case of IOP of 22 mm Hg or greater, additional antiglaucoma medications were prescribed, and the patient was excluded from the study. The t test (for continuous variables) and the Fisher exact test (for ordinal or categorical variables) were used for analysis. Statistical significance was set at P ≤ .05.

The following study variables were tested at each follow-up visit: (1) IOP (Goldmann applanation tonometry, average of the 2 highest readings of the daily phasing, from 8 AM to 6 PM, 6 readings at 2-hour intervals) and (2) blood pressure and heart rate (average of the values measured at the beginning and end of each follow-up visit). Respiratory variables were evaluated at recruitment and at the end of each follow-up phase (3 and 4 years after enrollment). Respiratory function was measured using a flow-sensing spirometer connected to a computer for data analysis (Vmax22; Sensor Medics, Yorba Linda, Calif). Each patient performed 3 flow-volume curves, and the best FEV₁ value was used. Metacholine challenge testing was performed according to a standardized procedure. Each participant inhaled doubling increasing concentrations of metacholine (0.03-256 mg/mL), vaporized by a dosimeter with a mean ± SD output of 9.0 ± 0.3 μL/puff (Dosimeter MB3; Mefar, Brescia, Italy), and FEV₁ was measured after each inhalation. Bronchial response to metacholine challenge was defined as an FEV₁ that was reduced by at least 20% from the postsaline value. The provocative concentration of metacholine that caused a 20% decrease in FEV₁ (PC20) was calculated by linear interpolation of the logarithmic dose-response curve. The PC20 values were log-transformed before analysis.

Five patients in the timolol-treated group and 6 in the laser-treated group had an IOP of 22 mm Hg or greater during follow-up and were excluded from the study. Therefore, the number of individuals who completed follow-up dropped to 11 in the timolol-treated group (5 women and 6 men; mean age, 53 years) and 10 in the laser-treated group (5 women and 5 men; mean age, 55 years).

The mean ± SD IOP, blood pressure, heart rate, and FEV₁ values in the 2 groups are given in the Table. Baseline data were not statistically significantly different between the 2 groups. After 3 years of follow-up, mean blood pressure, heart rate, and FEV₁ values showed statistically significant decreases from baseline values in the timolol-treated group (paired samples t test), whereas no significant changes were observed in the laser-treated group. Regarding bronchial reactivity, 6 of 11 timolol-treated patients showed a detectable response to metacholine challenge. All laser-treated patients had a PC20 greater than 256 mg/mL (P<.001 by Fisher exact test). Not a single patient in either group experienced symptoms consistent with chronic bronchial distress.

Thereafter, timolol therapy was withdrawn and laser trabeculoplasty was offered instead, following the same schedule adopted for the laser-treated group. Both groups underwent a further 12 months of follow-up. Final IOP, blood pressure, heart rate, and FEV₁ values for the 2 groups are detailed in the Table. In the timolol-treated group, mean blood pressure and heart rate recovered to baseline levels on withdrawal of timolol therapy. The mean FEV₁ was still lower than the baseline value, with the difference reaching borderline significance (P = .052). In the laser-treated group, no change occurred.

Data regarding reactivity to metacholine challenge for the timolol-treated group are plotted in the Figure. The results indicate that 1 year after timolol withdrawal, bronchial reactivity recovered to baseline levels (PC20>256 mg/mL) in 3 of 6 previous “responders.” Reactivity to metacholine challenge was still present in the remaining 3 patients. The 5 timolol-treated patients with PC20 greater than 256 mg/mL at the 3-year follow-up visit were still showing an undetectable response 1 year later. No response to metacholine challenge was detectable in the laser-treated group (data not shown).

These data show increased bronchial reactivity in 6 (55%) of 11 patients undergoing long-term topical treatment with timolol eyedrops. The enrolled patients had excellent lung function at baseline, with the response to metacholine challenge being undetectable up to 256 mg/mL. No patient experienced clinically significant respiratory impairment. In fact, bronchial symptoms do not develop until the PC20 index declines to less than 16 to 32 mg/mL⁸; the PC20 value, measured at the end of the 3-year timolol phase, was never less than 40 mg/mL (Figure).

In humans, some spontaneous intertest variability of PC20 values has been described.⁹ “Regression-to-the-mean” phenomena can then occur. We addressed this possible source of artifacts as follows:

• In this study, the laser-treated group representing the “controls” to be compared with the timolol-treated group. The change in response to metacholine challenge was detectable in the timolol-treated group only. Patients included in the laser-treated group never showed an appreciable PC20 decrease throughout follow-up. Should a regression-to-the-mean phenomenon have occurred, both groups would have been affected.

• The 2 study groups were age-matched, and care was taken to avoid enrollment of individuals at risk of bronchial pathologies (Table). Possible selection biases are then unlikely to occur.

Therefore, the decrease in PC20 values in long-term timolol users represented a true biological phenomenon in our study cohort.

β-Blocker eyedrops are known to exert systemic effects. On instillation of topical β-blockers, decreases in heart rate and arterial pressure have been extensively reported.^6,10 The treatment protocol adopted in our study involved twice-daily administration of 0.5% timolol in both eyes. Because 80% of the volume of the eyedrop (approximately 60 μL) is drained into the lacrimal pathway,¹¹ 1.0 to 1.2 mg of timolol daily per patient was absorbed by the highly vascularized mucosa of the nasolacrimal ducts. This amount is far lower than the oral dose of timolol maleate (20-60 mg/d) suggested for the treatment of systemic hypertension.¹² However, oral β-blockers undergo substantial “first-pass” hepatic metabolism, with the actual dose that reaches the circulation being less than 10% of the amount absorbed by the gastroenteric tract.¹¹ Because the eyedrop is immediately absorbed by the mucosal vasculature, the liver is bypassed. Therefore, by avoiding first-pass metabolism, “higher drug plasma levels relative to their starting dose is achieved, which may explain the systemic side effects of topical β-blockers despite low dosages.”^13(p268)

The plasma levels of timolol were not measured in the present study. Therefore, we do not know if or how much of the drug was systemically absorbed and whether the response to metacholine challenge was related to the actual concentration of the circulating drug. However, evaluation of heart rate and arterial pressure allowed the detection of a decrease in both at the end of 3 years of follow-up. Heart rate and blood pressure recovered to baseline values on 1-year withdrawal of timolol (Table). This phenomenon suggests that timolol was systemically absorbed.

In conclusion, our data show that in otherwise healthy individuals with glaucoma, long-term treatment with topical β-blockers stimulates bronchial reactivity, leading to subclinical reductions in FEV₁ and the PC20 threshold. This phenomenon may not be completely reversible on withdrawal of the β-blockers.

Correspondence: Stefano A. Gandolfi, MD, Servizio di Glaucoma, Istituto di Oftalmologia, Universitá di Parma, Via Gramsci 14, 43100 Parma, Italy (s.gandolfi@rsadvnet.it).

Submitted for Publication: August 16, 2003; final revision received March 9, 2004; accepted June 1, 2004.

Financial Disclosure: None.

Funding/Support: This study was supported in part by a local academic research grant from the Ministero dell’Università e Ricerca Scientifica e Tecnologia, Rome, Italy.