Crossover Comparison of Timolol and Latanoprost in Chronic PrimaryAngle-closure Glaucoma

Objective  To compare latanoprost and timolol maleate as primary therapy in 60eyes with chronic primary angle-closure glaucoma after a laser iridotomy.

Methods  We performed a prospective, randomized, crossover study of 60 eyes of30 patients with chronic primary angle-closure glaucoma after laser iridotomy.Patients were randomized to 2 groups: those taking latanoprost once dailyor those taking timolol twice daily. Three months after treatment with thefirst drug, the second drug was substituted. The circadian rhythm of intraocularpressure (IOP) was recorded before the start of therapy, at 3 months, andat 7 months. The fourth month was the washout period for the first drug.

Results  The mean baseline IOP was 23.5 ± 2.1 mm Hg, which decreased by8.2 ± 2.0 mm Hg with latanoprost (P<.001)and by 6.1 ± 1.7 mm Hg with timolol (P = .01).The decrease in IOP was greater for patients taking latanoprost (P<.001). Latanoprost was significantly more effective in eyes havingmorning and afternoon peaks of IOP. A total of 43 eyes (72%) of patients takinglatanoprost and 26 (43%) on timolol achieved a reduction of more than 30%from baseline IOP.

Conclusion  There were greater mean and peak IOP reductions achieved with 0.005%latanoprost once daily compared with 0.5% timolol twice daily.

Latanoprost, a phenyl-substituted prostaglandin analogue, has been studiedextensively for its intraocular pressure (IOP)-lowering efficacy and adverseeffects in cases of primary open-angle glaucoma.^1-3 Numerousstudies^4,5 have also comparedits effect with that of timolol maleate in cases of primary open-angle glaucoma.Chronic primary angle-closure glaucoma constitutes a significant proportionof all glaucomas, especially in Asia. β-Blockers and pilocarpine nitratehave been routinely used for the medical therapy of chronic primary angle-closureglaucoma after an iridotomy; however, the role of newer antiglaucoma medicationslike latanoprost needs evaluation. The formation of peripheral anterior synechiaein chronic primary angle-closure glaucoma could diminish access to the uveoscleraloutflow, the major mechanism of action of latanoprost. To our knowledge, nolong-term, crossover study has evaluated latanoprost and timolol with respectto their efficacy in cases of chronic primary angle-closure glaucoma. Theaim of the present study was to compare the effect of latanoprost administeredonce at night with timolol administered twice daily in a masked, crossoverstudy in chronic primary angle-closure glaucoma eyes.

The study was designed as a prospective, randomized, crossover trialcomparing the efficacy and adverse effects of timolol and latanoprost as monotherapyin freshly diagnosed cases of eyes having chronic primary angle-closure glaucomaafter an Nd:YAG laser iridotomy. Consecutive adult patients with bilateral,untreated, chronic primary angle-closure glaucoma were included in the studyafter providing informed consent. Human investigation guidelines were adheredto. All patients had an occludable angle, with peripheral anterior synechiaeinvolving more than 180°, a baseline IOP of more than 21 mm Hg withoutany antiglaucoma medication on more than 2 occasions, and optic nerve headand visual field changes commensurate with the diagnosis of glaucoma. Allpatients had a patent peripheral iridotomy.

Exclusion criteria included prior medical or surgical intervention forthe control of IOP, any previous ocular surgery, any other intraocular disorder,or any condition preventing reliable applanation tonometry. Patients usingsystemic β-blockers were excluded. Known hypersensitivity to any componentof the drugs to be used, patients who were unable to adhere to the follow-upprotocol, or those with systemic or ocular problems that contraindicated theuse of either of the 2 study drugs were also excluded. Patients with a baselineIOP of more than 35 mm Hg or cases with advanced glaucoma (defined as cuppingratio of 0.9 and/or perimetric evidence of visual field loss within 10°of fixation in one or more quadrants) were also excluded from the study.

At the time of enrollment in the study, a complete medical and ocularhistory was taken, and any concurrent medical therapies were recorded. A systemicexamination was performed to evaluate the cardiovascular and respiratory status.A comprehensive ocular examination was performed, including best-correctedvisual acuity, slitlamp examination, biomicroscopic fundus evaluation, 3-hourlyapplanation tonometry from 7 AM to 10 PM on a single day usinga Goldmann applanation tonometer, and full threshold automated perimetry onthe 30-2 program of Humphrey field analyzer.

The patients were then randomized into 2 parallel study groups: onegroup received 0.005% latanoprost at 10 PM once daily, and the other groupreceived 0.5% timolol maleate at 8 AM and 8 PM.Follow-up examination was performed at 3 weeks, 6 weeks, and 3 months afterthe start of therapy. Best-corrected visual acuity, IOP recording, and fundusevaluation were performed at each follow-up visit. After 3 months, the secondmedication was substituted (ie, patients in the latanoprost group startedtaking timolol and vice versa). The first month of treatment with the seconddrug (ie, the fourth month of the trial) was deemed the washout period forthe first drug used during the first 3 months of therapy. Additional follow-upwas performed at 3 weeks, 6 weeks, and 3 months after the washout period.

A circadian recording of IOP and full threshold automated perimetrywere repeated at 3 and 7 months after enrollment in the study (ie, 3 monthsfollowing the use of each drug). The timings of the diurnal recording of IOPwere 7 AM, 10 AM, 1 PM, 4 PM, 7 PM, and 10 PM; on all occasionsthe IOP recorded was performed by applanation tonometry in the sitting positionby a single masked observer. Patients were recalled on an outpatient basisevery 3 hours on 1 day.

At each of the follow-up visits, the patient's eye was examined withslitlamp biomicroscopy to rule out any uveitis, iris color changes, or anyeyelash changes. The patients were also asked in detail about any adverseocular and systemic events that occurred during the treatment. A subsequentsystemic examination with heart rate and blood pressure measurement was performedat each follow-up visit.

The efficacy of the 2 drugs was evaluated with respect to the dampeningof the range of diurnal variation in IOP, its effect on the different typesof circadian cycles, and drug efficacy regarding baseline peak IOP. The effectivenessof the 2 drugs was also evaluated with respect to age, sex, and the presenceor absence of diabetes and hypertension.

Peak pressure was defined as the highest pressure recorded in each individualcircadian rhythm. Trough pressure was defined as the lowest pressure recordedin each individual circadian rhythm. A change in the timing of the peak pressuresrecorded at baseline and on diurnal measurements of IOP at 3 and 7 monthswas recorded in each individual. It was considered negative if the peak IOPin the circadian rhythm on treatment was previously earlier compared withthe timing of the baseline peak and positive if it occurred later.

Each baseline circadian rhythm was classified as morning type, noontype, and evening type, depending on the timing of the peak pressures recordedin that diurnal curve. Morning type was defined as peak pressures at 7 AMor 10 AM. Noon type was defined as peak pressures at 1 PM or4 PM. Evening type was defined as peak pressures at 7 PM or10 PM.

Statistical analysis was performed with STATA Intercooled statisticalsoftware, version 6.0 (Stata Corp, College Station, Tex) using the unpaired t test and 2-way analysis of variance. Data are presentedas mean ± SD.

Sixty eyes of 30 patients were enrolled in the study during an enrollmentperiod of 3 months. The mean age of the patients was 57.7 ± 7.4 years(age range, 46-76 years). There were 18 men and 12 women. The prevalence ofdiabetes mellitus was 30%, and the prevalence of hypertension was 13%. Boththe diseases were controlled with oral medication. None of the patients weretaking oral β-blockers for control of elevated blood pressures. The meancup-disc ratio was 0.6 ± 0.8. Peripheral anterior synechiae extendedfrom 180° to 270° in all the eyes studied.

The mean of the baseline IOP was 23.5 ± 2.1 mm Hg and was decreasedto 15.3 ± 1.8 mm Hg (34.9%) with latanoprost (P<.01)and to 17.4 ± 1.7 mm Hg (26.0%) with timolol (P<.01).The baseline circadian rhythm of IOP and the IOP with latanoprost and timololare detailed in Table 1 and Figure 1. Both the drugs significantlyreduced IOP compared with baseline at all points on the diurnal curve. Latanoprostwas significantly more effective in lowering IOP than timolol at 7 AM, 10 AM, 1 PM, 4 PM,and 7 PM. At 10 PM patients taking latanoprostrecorded lower IOPs compared with those taking timolol, but the differencebetween the 2 drugs was not statistically significant (P = .25).

The average of the peak pressures recorded in the 60 individual baselineIOP curves was 27.5 ± 2.7 mm Hg. The average of the highest IOP recordedin each of the subsequent individual circadian rhythms was 18.1 ± 2.4mm Hg after treatment with latanoprost for 3 months and 20.4 ± 2.5mm Hg after treatment with timolol for 3 months; both the drugs showed a significantreduction compared with baseline peak pressures (P<.01).The peak IOP reduction was greater for latanoprost compared with timolol (P<.01).

The average of the trough IOP recorded in each of the individual baselineIOP curves was 20.2 ± 2.1 mm Hg. The average of the trough IOP thatwas recorded at any point of time on each of the diurnal curves was 12.5 ±2.1 mm Hg with latanoprost (P<.01) and 14.4 ±2.9 mm Hg with timolol (P<.01). The IOP reductionwas greater for patients taking latanoprost compared with those taking timolol(P<.01).

In each of the individual diurnal curves, the alteration in timing ofIOP peaks on therapy with both the drugs was compared with the time of thepeak IOP on the baseline diurnal curve of the same patient. The average ofthis shift in the time of the peak pressures was recorded for both the drugsused (Table 2). The results showedthat with timolol there was no significant time shift of the peak IOP comparedwith baseline in all the 3 types of circadian rhythms, and overall the peakIOP was recorded 0.15 ± 0.4 hours later than that recorded in the baselinecurve. Latanoprost caused the peak IOP to occur 3.3 ± 1.79 hours laterin the day, especially for those with a morning or afternoon peak.

The circadian rhythm recorded on baseline evaluation was divided into3 categories: those with peaks in the morning (30 eyes, 50%), noon (20 eyes,33.3%), or evening (10 eyes, 16.7%). Evaluating measurements of IOP at thetime of the baseline peak, the decrease in the IOP after the use of latanoprostor timolol was analyzed in each of the 3 types of circadian patterns (Table 3). Timolol caused a similar decreasein IOP in all the 3 types of circadian rhythms, whereas latanoprost causeda mean percentage decrease in IOP of 40.9% ± 5.5% in those with a morningpeak, 34.8% ± 6.3% for afternoon peaks, and 31.9% ± 6.5% ineyes having a peak at night. This difference in the efficacy of latanoprostin patients with a night peak (31.9%) compared with patients with a morningpeak (40.9%) was statistically significant (P<.01).

Different age groups, the sex of the patient, the presence or absenceof hypertension or diabetes, and the height of the baseline peak, trough,and mean IOP did not affect the pressure reduction achieved with either drug.The effect of the 2 drugs was similar whether used first or second in thestudy.

The numbers of patients who did not achieve a mean IOP of 21 mm Hg orless with treatment were 6 (10%) in the timolol group and 1 in the latanoprostgroup. A peak IOP of more than 21 mm Hg was seen in 15 cases (25%) in thetimolol group and 4 cases (7%) in the latanoprost group. An IOP reductionof less than 20% was noted in 5 cases (8%) after latanoprost treatment andin 13 (22%) after timolol treatment. A pressure reduction of 30% or more frombaseline was observed in 43 eyes (72%) of patients taking latanoprost comparedwith 26 (43%) of those taking timolol (P<.001)(Table 4). The patient whose conditionwas uncontrolled with latanoprost showed a mild elevation of IOP to 25 mmHg, which was confirmed on stopping use of and rechallenging with latanoprost.

No significant adverse effects were observed during the study period.Four patients taking latanoprost complained of discomfort during the studyperiod, 2 patients taking latanoprost complained of conjunctival hyperemia,and 2 patients reported foreign body sensation after instilling timolol drops.None of the adverse effects were significant enough to result in discontinuationof therapy. There was no significant change in the pulse rate and blood pressureof patients taking any of the medications used in our study compared withthe baseline.

In chronic primary angle-closure glaucoma, a peripheral iridotomy helpsto relieve the pupillary block and further attacks of angle closure; however,medication is required to control the chronically elevated IOP caused by obstructionof aqueous outflow secondary to synechial angle closure and trabecular meshworkdamage.⁶ Previous studies^7,8 havenoted that patients with chronic angle-closure glaucoma in whom disc and fieldchanges had occurred were not likely to have their condition controlled withan iridectomy but required a trabeculectomy to control their IOP. Pilocarpineand β-blockers have been the mainstay of therapy for chronic angle-closureglaucoma^9,10 and were seen tocontrol the IOP in 30% of chronic primary angle-closure glaucoma eyes.¹⁰

There are many new antiglaucoma drugs now available, the most efficaciousbeing prostaglandin analogues, which act by increasing the uveoscleral outflow.Access to the uveoscleral pathway is through the ciliary muscle bundles intothe supraciliary and suprachoroidal spaces from which it is drained throughsclera. The presence of peripheral anterior synechiae could possibly hamperthe flow of aqueous into this pathway and perhaps decrease the efficacy ofdrugs such as prostaglandin analogues.

The efficacy of latanoprost has been compared with timolol in multiplestudies, but to the best of our knowledge, there has been no previous long-term,crossover study regarding their effect on peak IOPs in cases of primary angle-closureglaucoma. The current study was undertaken to evaluate and compare the efficacyof latanoprost with timolol in eyes with chronic primary angle-closure glaucoma.The 2 medications have a different viscosity and different administrationregimens that made masking difficult. Because the same patients were testedwith both drugs, confounding factors were minimized.

We noted the mean reduction of IOP to be 8.2 ± 0.4 mm Hg withlatanoprost and 6.1 ± 0.2 mm Hg with timolol. A similar reduction wasreported by Aung et al¹¹ in a small numberof eyes studied for 2 weeks; latanoprost reduced the IOP by 8.8 ± 1.1mm Hg compared with 5.7 ± 0.7 mm Hg by timolol. Hedman and Alm¹² reported a mean decrease of 7.7 ± 0.1 mm Hgin a meta-analysis of primary open-angle glaucoma eyes compared with a meandecrease of 8.7 ± 2.2 mm Hg in our study, starting from similar baselineIOPs.

A previous study¹³ has shown that patientswith IOP consistently below 15 mm Hg had a higher chance of remaining stable.However, because different patients have different baseline pressures anddifferent target pressures, it may be reasonable to lower the IOP by at least30% of baseline pressures to prevent progression of field loss.¹⁴ Theeffect of latanoprost on mean IOP was clinically more significant than timololin our study in causing a reduction of more than 30% from baseline. This decreaseof more than 30% from baseline was seen in 72% of primary angle-closure glaucomaeyes of patients taking latanoprost and 43% of primary angle-closure glaucomaeyes of patients taking timolol.

We evaluated the circadian rhythm in eyes with chronic primary angle-closureglaucoma before and after therapy with latanoprost and timolol. Our studynoted a significantly higher reduction of peak and trough IOP with latanoprostcompared with timolol. There was thus a greater dampening of the circadianrhythm of IOP with latanoprost in chronic primary angle-closure glaucoma eyes.The evening efficacy of latanoprost and timolol was similar. Orzalesi et al¹⁵ and Racz et al¹⁶ havealso noted that latanoprost leads to a more uniform circadian rhythm in primaryopen-angle glaucoma eyes. Hedman and Alm¹² foundlatanoprost to be more effective than timolol in primary open-angle glaucomaeyes when morning, noon, and afternoon IOPs were averaged to determine thediurnal IOP. In our study, latanoprost used at night was less effective inthe control of baseline IOP peaks that occurred in the evening (7-10 PM). Timolol was equally effective in all types of circadian rhythmsand appeared to work around the clock, albeit to a lesser extent.

It is important to schedule the follow-up of glaucoma patients aroundthe time that the highest IOP is expected. We studied the temporal changeof the highest recorded IOP in eyes after the use of latanoprost and timolol.The time shift of peak IOPs in eyes of patients taking timolol was insignificantin all types of circadian rhythms. Of patients taking latanoprost, the peakIOP in eyes with morning and afternoon baseline peaks were shifted 3 to 7hours later in the day. This could be due to its efficacy waning with time.The time difference noted by us could be kept in mind when scheduling follow-upvisits. In conclusion, latanoprost was clinically more effective than timololbecause it lowered the IOP to a greater extent and dampened the circadianrhythm of IOP.

Corresponding author: Ramanjit Sihota, MD, FRCS, Dr Rajendra PrasadCentre for Ophthalmic Sciences, All India Institute of Medical Sciences, NewDelhi 110029, India (e-mail: rjsihota@hotmail.com).

Submitted for publication June 6, 2003; final revision received August17, 2003; accepted September 15, 2003.

The principal author takes full responsibility for the integrity ofthe data and the reliability of the data analysis.