Blood-Aqueous Barrier Changes After the Use of Prostaglandin Analogues in Patients with Pseudophakia and Aphakia a 6-Month Randomized Trial

CLINICAL SCIENCES Blood-Aqueous Barrier Changes After the Use of Prostaglandin Analogues in Patients With Pseudophakia and Aphakia A 6-Month Randomized Trial

Enyr S. Arcieri, MD; Alessandro Santana, MD; Fabiano N. Rocha, MD; Gustavo L. Guapo, MD; Vital P. Costa, MD

Objectives: To investigate the effects of prostaglandin throughout follow-up (P Ͻ .02). Four latanoprost- analogues on the blood-aqueous barrier and to evaluate treated eyes, 1 bimatoprost-treated eye, and 1 travoprost- the occurrence of cystoid macular edema in aphakic or treated eye developed cystoid macular edema; all cases pseudophakic patients with glaucoma. resolved after discontinuation of the prostaglandin analogue and treatment with topical diclofenac sodium. Mean Methods: In this randomized, masked-observer, 6-month intraocular pressure reductions after 6 months were higher clinical trial, patients with primary open-angle, pseudo- for the latanoprost (26%), bimatoprost (28%), and tra- phakic, or aphakic glaucoma were treated once daily with voprost (29%) groups than for the control (3%) and uno- bimatoprost (n=16), latanoprost (n=15), or travoprost prostone (14%) groups (PϽ.05). Bimatoprost induced (n=17) or twice daily with unoprostone (n=16) or lu- significantly higher hyperemia scores than latanoprost, bricant drops (control group) (n=16). Blood-aqueous bar- unoprostone, and placebo (PϽ.01). rier status, which was assessed using a laser flare meter; intraocular pressure; the occurrence of angiographic cystoid macular edema; and conjunctival hyperemia were Conclusion: Bimatoprost, latanoprost, and travoprost use evaluated. may lead to disruption of the blood-aqueous barrier in patients with pseudophakia and aphakia. Results: Mean flare values were significantly higher in the bimatoprost, latanoprost, and travoprost groups Arch Ophthalmol. 2005;123:186-192

EVERAL YEARS AFTER THE OB- CME has also been reported25 in patients servation that prostaglandin with pseudophakic or aphakic glaucoma (PG) F2␣ was a potent ocular being treated with bimatoprost, travo- hypotensive substance,1-5 la- prost, or unoprostone. tanoprost, unoprostone, tra- To the best of our knowledge, there are voprost,S and bimatoprost were devel- no published clinical studies that com- oped and became widely used in the pare the safety of topical PG analogues in treatment of primary open-angle glau- patients with pseudophakia or aphakia. coma and ocular hypertension.5-7 Al- The primary objectives of this study are though these drugs have structural differ- to investigate the effects of PG analogues ences, they share similar characteristics and on the blood-aqueous barrier and to evalu- Author Affiliations: Glaucoma 5 Service, Department of are often referred to as PG analogues. ate the occurrence of angiographic CME Ophthalmology, University of Many adverse effects have been re- in patients with aphakic or pseudopha- Campinas, Campinas, São ported with PG analogues, including con- kic glaucoma. Paulo, Brazil (Drs Arcieri, junctival hyperemia, iris hyperpigmenta- Santana, Rocha, and Guapo); tion, and eyelash growth.3,8-17 However, METHODS Department of Ophthalmology, among the serious PG-induced adverse ef- School of Medicine, Federal fects are the disruption of the blood- University of Uberlândia, aqueous barrier and the development of This 6-month, randomized, masked-observer Uberlândia, Minas Gerais, Brazil cystoid macular edema (CME).18-25 Al- clinical trial was conducted at the Glaucoma (Dr Arcieri); and Department of though there have been several stud- Service of the University of Campinas. The Ophthalmology, University of 18-23 study was performed in accordance with the São Paulo (Dr Costa). ies of CME and anterior uveitis asso- Declaration of Helsinki after receiving ap- Financial Disclosure: Dr Costa ciated with latanoprost use in patients with proval from the ethics committee of the Uni- has received research grants pseudophakia and aphakia, the inci- versity of Campinas. Written informed con- from Alcon Inc, Novartis, and dence of these adverse effects is un- sent was obtained from each patient before Pfizer in the past. known. Angiographically documented inclusion in the study.

(REPRINTED) ARCH OPHTHALMOL / VOL 123, FEB 2005 WWW.ARCHOPHTHALMOL.COM 186

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 Patients were eligible for participation if they met the fol- lowing inclusion criteria: age older than 18 years, pseudophakia or aphakia, intraocular pressure (IOP) greater than the tar- get level (determined by V.P.C.), and a diagnosis of primary open-angle, pseudophakic, or aphakic glaucoma. Pseudopha- kic and aphakic glaucoma were defined in individuals with pseudophakia and aphakia with no history of glaucoma before cataract surgery. Patients were excluded from the study if they had Grade 0 (None) Grade 1 (Mild) a history of uveitis or CME, substantial ocular irritation at baseline, or a history of intraocular surgery or a laser procedure within 6 months of baseline. We also excluded individuals who had been treated with PG analogues in the past and those who had undergone other ocular surgery except for cataract or glaucoma. Finally, the presence of systemic disorders that could be associated with uveitis or CME (ie, diabetes mellitus and rheu- matologic diseases), pregnancy, lactation, and inadequate con- Grade 2 (Moderate) Grade 3 (Severe) traception (in females) were also exclusion criteria. If patients were eligible but were using any antiglaucoma Figure 1. Standard photographic chart used to grade conjuctival hyperemia. medications (except PG analogues), hypotensive therapy was discontinued. Required washout periods before the baseline visit were 4 weeks for ␤-adrenergic antagonists, 2 weeks for adren- ments, and ophthalmoscopy. The measurements were per- ergic agonists, and 5 days for cholinergic agonists and car- formed at the same time (10 AM) at all visits by a masked ob- bonic anhydrase inhibitors. A safety check with IOP measure- server. ment was required after 2 weeks for all patients undergoing a A laser flare meter (FM 500; Kowa Co Ltd, Tokyo, Japan) 4-week washout. At that time, patients whose IOPs had risen was used to determine the status of the blood-aqueous barrier to levels deemed to be detrimental were excluded from the study. at all follow-up visits by the same masked investigator (E.S.A.). No other IOP-reducing therapy was permitted during the study. The flare measurements were repeated 7 times, the highest and If both eyes of a patient were eligible for study inclusion, the lowest values were excluded, and the mean of the 5 remaining same medication was prescribed for both eyes, although only values was adopted as a “flare value” for statistical analysis. Ac- 1 eye per patient was included in the analysis. cording to information provided by the manufacturer, flare read- Study medications, packaged in commercially available, la- ings greater than 26 photon counts per millisecond (p/ms) are beled containers, were as follows: 0.005% latanoprost (Xala- indicative of a disruption in the blood-aqueous barrier. tan; Pfizer Inc, New York, NY), 0.03% bimatoprost (Lumigan; The IOP was measured using a Goldmann applanation to- Allergan Inc, Irvine, Calif), 0.004% travoprost (Travatan; Al- nometer by the same investigator (E.S.A.) at all visits. Three con Inc, Ft Worth, Tex), and 0.12% unoprostone (Rescula; measurements were performed in each eye, and the mean of 3 Novartis AG, Basel, Switzerland). Before dispensing, latano- values was used for statistical analysis. At baseline and during prost and unoprostone were stored refrigerated at 2°C to 8°C, follow-up, a masked investigator (A.S.) graded the conjuncti- and patients were instructed to conserve them under refrig- val hyperemia according to a scale. Each eye was compared with eration. Bimatoprost and travoprost were stored at room tem- standard photographs showing conjunctival hyperemia of grades perature. 0, 1, 2, and 3 (none, mild, moderate, and severe, respectively) To investigate the effects of benzalkonium chloride, the pre- (Figure 1); the scale included values of 0, 0.5, 1.0, 1.5, 2.0, servative used in all PG analogues, we included in the study a 2.5, and 3.0. control group of patients with pseudophakic and aphakic glau- Fluorescein angiography (FA) was performed to investi- coma who had reached IOP control after trabeculectomy with- gate the occurrence of CME at baseline and at 1 and 6 months out the need for medication. These patients received a lubri- of follow-up, or if a patient showed decreased visual acuity at cant drop (Tears Naturale; Alcon Inc) containing 0.01% any time during follow-up. If CME was detected, the patient benzalkonium chloride. was instructed to discontinue taking the medication, and a non- To preserve masking, a designated unmasked coordinator steroidal anti-inflammatory drug, diclofenac sodium (Voltaren; (A.S.)—who did not perform any study evaluations or assess- Novartis), was prescribed (4 times a day for 4 weeks). Then, ments—received randomization codes, dispensed the medica- FA was again performed to evaluate whether the CME had re- tions, and instructed the patients on how to use and store the solved. The FAs were analyzed by a single masked retina spe- containers. The Web program Research Randomizer v3.0 (http: cialist, who graded them as normal (no fluorescein leakage) or //www.randomizer.org) was used for random assignment. Pa- abnormal (fluorescein leakage compatible with CME). tients were randomized to the once-daily use (at 8 PM)ofbi- Before the study, it was determined that a sample size of 14 matoprost, latanoprost, or travoprost or to the twice-daily use patients in each treatment group had 90% power to detect a (at 8 AM and 8 PM) of unoprostone or placebo. New bottles of 20% difference in aqueous flare measurements between the study medications were dispensed by the unmasked coordina- groups at a significance level of ␣=.05, using the estimated vari- tor to all treatment groups every month to ensure that patients ability determined in a previous study.26 were using medications within the label-specified timeframe Categorical variables were analyzed using the Fisher exact and to preserve masking. test or the ␹2 test. Continuous variables were analyzed using Study visits occurred at baseline; after 7 and 15 days; and analysis of variance when the values were normally distrib- after 1, 2, 3, 4, 5, and 6 months of treatment. At baseline, the uted. If statistically significant differences were detected, the medical records of each patient were analyzed, and the follow- Tukey test for multiple comparisons was used to identify the ing data were obtained: age, race, diagnosis, the presence of an site(s) of difference. When values were not normally distrib- intraocular lens, posterior capsule status, previous ocular pro- uted and the variances were not homogeneous, we used the cedures, and the interval between cataract surgery and base- Kruskal-Wallis test, followed by the Wilcoxon 2-sample test if line. At each visit, all patients underwent Snellen visual acuity statistically significant differences were found. Within-group measurement, slitlamp biomicroscopy, IOP and flare measure- changes from baseline were analyzed using paired t tests. The

(REPRINTED) ARCH OPHTHALMOL / VOL 123, FEB 2005 WWW.ARCHOPHTHALMOL.COM 187

16 Patients Randomized to Daily 15 Patients Randomized to Daily 17 Patients Randomized to Daily 16 Patients Randomized to 16 Patients Randomized to Bimatoprost Use Latanoprost Use Travoprost Use Twice-Daily Unoprostone Use Twice-Daily Placebo Use

15 Completed 11 Completed 16 Completed 16 Completed 16 Completed 1 Discontinued Owing to 3 Discontinued Owing to CME 1 Discontinued Owing to CME 0 Discontinued 0 Discontinued Uveitis and CME 1 Discontinued Owing to Uveitis and CME

Figure 2. Trial profile. For eligible patients taking any antiglaucoma medications, the required washout periods before the baseline visit were 4 weeks for ␤-adrenergic antagonists, 2 weeks for adrenergic agonists, and 5 days for cholinergic agonists and carbonic anhydrous inhibitors. CME indicates cystoid macular edema.

Table 1. Demographic and Clinical Characteristics of the Treatment Groups

Placebo Unoprostone Bimatoprost Latanoprost Travoprost P Characteristic Group (n = 16) Group (n = 16) Group (n = 16) Group (n = 15) Group (n = 17) Value Age, mean ± SD, y 64.7 ± 14.2 68.9 ± 13.5 69.6 ± 9.2 66.7 ± 15.4 65.8 ± 13.6 .82 Sex, M/F, No. 9/7 10/6 8/8 7/8 9/8 .92 Race, white/black, No. 16/0 11/5 10/6 12/3 13/4 .11 Cup-disc ratio, mean ± SD 0.61 ± 0.19 0.57 ± 0.18 0.70 ± 0.20 0.61 ± 0.18 0.68 ± 0.19 .29 Diagnosis, POAG/PG, No. 9/7 6/10 11/5 7/8 10/7 .45 No. of previous intraocular operations, mean ± SD 1.43 ± 0.51 1.12 ± 0.34 1.12 ± 0.34 1.13 ± 0.35 1.11 ± 0.33 .08 Lens status, pseudophakic/aphakic, No. 15/1 15/1 14/2 13/2 15/2 .93 Posterior capsule status, intact/ruptured, No. 9/7 9/7 8/8 8/7 8/9 .98 ⌬t, Mean ± SD, y 3.23 ± 3.23 3.54 ± 2.48 2.01 ± 2.48 3.39 ± 2.58 3.29 ± 3.09 .70

Abbreviations: PG, pseudophakic/aphakic glaucoma; POAG, primary open-angle glaucoma; ⌬t, interval between cataract surgery and baseline.

Table 2. Aqueous Flare Values for Each Treatment Group at Baseline and Throughout Follow-up

Aqueous Flare Value, Mean ± SD, p/ms P Time Placebo Group Unoprostone Group Bimatoprost Group Latanoprost Group Travoprost Group Value Baseline 10.45 ± 4.33 8.36 ± 2.87 8.80 ± 3.51 6.81 ± 2.75 7.96 ± 3.59 .07 7 d 9.23 ± 3.55 8.63 ± 2.91 18.55 ± 16.77 11.08 ± 6.43 14.75 ± 7.32 .01* 15 d 9.53 ± 3.85 8.82 ± 2.79 20.81 ± 17.03 12.75 ± 8.82 15.27 ± 7.56 .003* 1 mo 9.66 ± 4.27 8.60 ± 2.78 20.35 ± 20.28 12.58 ± 9.76 14.10 ± 6.77 .01* 2 mo 9.11 ± 3.71 8.21 ± 2.24 23.69 ± 37.64 10.46 ± 6.94 13.76 ± 6.90 .003* 3 mo 9.69 ± 3.93 8.21 ± 2.84 15.08 ± 7.60 9.80 ± 5.12 15.03 ± 7.71 .002* 4 mo 9.50 ± 3.34 8.61 ± 2.91 14.97 ± 7.78 9.56 ± 5.19 15.18 ± 7.23 Ͻ.001* 5 mo 9.20 ± 3.85 7.28 ± 2.69 14.16 ± 6.73 9.67 ± 4.27 15.58 ± 7.76 Ͻ.001* 6 mo 9.39 ± 3.71 8.19 ± 2.71 14.36 ± 6.85 9.68 ± 4.49 14.80 ± 6.83 Ͻ.001*

*Statistically significant difference between groups (analysis of variance).

nonparametric Spearman rank correlation test was used to evalu- cant differences among groups regarding sex, age, race, ate the relationship between mean flare values and hyperemia cup-disc ratio, diagnosis, number of previous intraocu- Ͻ scores. P .05 was considered statistically significant. lar procedures, lens status, posterior capsule status, and interval between cataract surgery and baseline. RESULTS Table 2 lists the amount of aqueous flare at baseline and at each follow-up visit for the treatment groups. In Eighty patients were included in the study: 16 in the bi- the control group, the mean flare value was signifi- matoprost group, 15 in the latanoprost group, 17 in the cantly lower compared with baseline throughout follow- travoprost group, 16 in the unoprostone group, and 16 up, except at 1 month (P=.14). Comparing the PG ana- in the control group. Full details on patient flow through logues, although there were no significant differences at the study and exit status are given in Figure 2. Demo- baseline (P=.38), mean flare values were significantly graphic and clinical characteristics of the treatment groups higher in the bimatoprost, latanoprost, and travoprost are given in Table 1. There were no statistically signifi- groups throughout follow-up compared with the uno-

(REPRINTED) ARCH OPHTHALMOL / VOL 123, FEB 2005 WWW.ARCHOPHTHALMOL.COM 188

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 Table 3. Intraocular Pressures for Each Treatment Group at Baseline and Throughout Follow-up*

Intraocular Pressure, Mean ± SD, mm Hg

Time Placebo Group Unoprostone Group Bimatoprost Group Latanoprost Group Travoprost Group Baseline 14.8 ± 2.5 21.2 ± 2.4 20.2 ± 2.3 20.3 ± 2.3 20.5 ± 2.7 7 d 14.4 ± 2.7 18.7 ± 2.2 15.2 ± 2.4 15.7 ± 2.0 15.1 ± 2.2 15 d 14.0 ± 2.4 18.2 ± 2.3 14.6 ± 2.1 15.1 ± 1.5 14.7 ± 2.1 1 mo 14.3 ± 2.4 18.2 ± 2.2 14.5 ± 1.9 14.7 ± 1.6 14.5 ± 1.7 2 mo 14.7 ± 2.8 17.9 ± 1.9 14.3 ± 2.2 14.9 ± 1.6 14.2 ± 1.9 3 mo 14.4 ± 3.0 17.7 ± 2.0 14.4 ± 1.9 15.0 ± 1.8 14.3 ± 1.9 4 mo 14.4 ± 2.6 18.1 ± 1.9 14.5 ± 2.3 15.0 ± 1.7 14.3 ± 2.0 5 mo 14.4 ± 2.2 18.1 ± 2.1 14.5 ± 1.9 15.0 ± 1.6 14.5 ± 2.0 6 mo 14.3 ± 2.4 17.9 ± 1.9 14.3 ± 2.2 14.9 ± 1.7 14.2 ± 1.8

*P Ͻ .001 for all; statistically significant difference between groups (analysis of variance).

prostone group (PϽ.02). However, there were no significant differences in mean flare measurements among Placebo Unoprostone Bimatoprost Latanoprost Travoprost eyes receiving bimatoprost, latanoprost, and travoprost Group Group Group Group Group at all times (PϾ.07). Within-group analyses showed no 22 significant differences in flare values among patients 21 treated with unoprostone (PϾ.24), whereas patients re- 20 ceiving bimatoprost (PϽ.03), latanoprost (PϽ.02), and 19 travoprost (PϽ.001) showed significant increases in mean 18 flare values compared with baseline at all times. 17 Although patients treated with bimatoprost, latano- 16 prost, and travoprost demonstrated statistically signifi- 15 14 Intraocular Pressure, mm Hg cant increases in mean flare values, only 1 patient in the 13 latanoprost group (7%) and 1 in the bimatoprost group (6%) had anterior uveitis detected by slitlamp examina- Baseline 7 d 15 d 1 mo 2 mo 3 mo 4 mo 5 mo 6 mo Follow-up tion (P=.52). Both patients also developed CME. If we consider aqueous flare values greater than 26 p/ms as in- Figure 3. Mean intraocular pressures from baseline to month 6 for each dicative of uveitis, 2 patients in the bimatoprost group treatment group. (12%), 2 in the latanoprost group (13%), and 2 in the travoprost group (12%) had this complication (P=.36). When we evaluated the effect of posterior capsule sta- oped CME at 6 months. Only 1 patient treated with la- tus on mean flare values during follow-up, we observed tanoprost had CME associated with visual acuity loss (vi- that among patients treated with bimatoprost, the mean sual acuity decreased from 0.2 to 0.05). After resolution aqueous flare value was significantly higher in eyes with of the CME, his visual acuity improved to 0.2. All cases an absent or ruptured posterior capsule (18.70 vs 12.37 of CME resolved after discontinuation of the PG ana- p/ms; P=.04). Levels of aqueous flare did not differ in logue and the use of diclofenac for 1 month. eyes with vs without a broken posterior capsule in the Of the 6 patients who developed CME, 5 (83%) had latanoprost group (12.11 vs 9.81 p/ms; P=.48), the tra- an absent or ruptured posterior capsule. However, there voprost group (13.87 vs 15.79 p/ms; P=.51), the uno- was no significant difference between the incidence of prostone group (7.68 vs 8.97 p/ms; P=.66), or the pla- CME in eyes with a ruptured or absent posterior cap- cebo group (7.28 vs 11.07 p/ms; P=.20). sule (13.2%) and eyes with an intact posterior capsule No patient had angiographic CME at baseline. Four (2.4%) (P=.10). The mean±SD flare values were signifi- latanoprost-treated eyes (27%), 1 bimatoprost-treated eye cantly higher in patients who developed CME (30.23±30.32 (6%), and 1 travoprost-treated eye (6%) developed CME. p/ms) compared with those who did not (11.69±6.92 p/ms) The incidence was significantly higher in eyes receiving (P=.03). latanoprost than in the other groups (P=.03). Compar- At baseline, mean IOP was significantly lower in the ing latanoprost vs unoprostone and latanoprost vs pla- control group (PϽ.001), but the PG analogue groups had cebo, the incidence of angiographic CME was signifi- similar mean IOP levels (P=.71) (Table 3 and Figure 3). cantly higher with latanoprost (P=.04). However, there Within-group analyses showed no significant differ- was no significant difference between the incidence of ences in mean IOPs among eyes receiving placebo, ex- CME when latanoprost-treated eyes were compared with cept at 15 days, when there was a significant decrease in those receiving bimatoprost (P =.17) or travoprost mean IOP (P=.03). Patients treated with PG analogues (P=.16). Three patients treated with latanoprost and 1 showed significant IOP reductions at all times com- receiving travoprost developed CME at 1 month. One pa- pared with baseline (PϽ.001). tient in the bimatoprost group developed CME at 2 Overall, mean IOP reductions after 6 months were 3% months, and 1 patient in the latanoprost group devel- (0.4 mm Hg) for the placebo group, 14% (3.1 mm Hg)

(REPRINTED) ARCH OPHTHALMOL / VOL 123, FEB 2005 WWW.ARCHOPHTHALMOL.COM 189

Hyperemia Grade, Mean ± SD (Median) P Time Placebo Group Unoprostone Group Bimatoprost Group Latanoprost Group Travoprost Group Value* 7 d 0.03 ± 0.12 (0) 0.15 ± 0.30 (0) 0.68 ± 0.75 (0.5) 0.20 ± 0.31 (0) 0.35 ± 0.52 (0) .005 15 d 0 (0) 0.06 ± 0.17 (0) 0.71 ± 0.63 (0.5) 0.16 ± 0.24 (0) 0.35 ± 0.52 (0) Ͻ.001 1 mo 0 (0) 0.03 ± 0.12 (0) 0.56 ± 0.57 (0.5) 0.13 ± 0.22 (0) 0.20 ± 0.35 (0) Ͻ.001 2 mo 0 (0) 0 (0) 0.46 ± 0.76 (0.25) 0 (0) 0.09 ± 0.20 (0) .001 3 mo 0 (0) 0 (0) 0.16 ± 0.30 (0) 0 (0) 0.06 ± 0.17 (0) .02 4 mo 0 (0) 0 (0) 0.13 ± 0.22 (0) 0 (0) 0.03 ± 0.12 (0) .01 5 mo 0 (0) 0 (0) 0.13 ± 0.22 (0) 0 (0) 0.03 ± 0.12 (0) .01 6 mo 0 (0) 0 (0) 0.13 ± 0.22 (0) 0 (0) 0.03 ± 0.12 (0) .01

*All P values had a statistically significant difference between groups (Kruskal-Wallis).

in patients treated with bimatoprost compared with those Placebo Unoprostone Bimatoprost Latanoprost Travoprost Group Group Group Group Group treated with unoprostone or placebo throughout follow-up (PϽ.04). Bimatoprost-treated eyes had signifi- 0.8 cantly higher grades of hyperemia than travoprost- 0.7 treated eyes at 1 month (P=.04) and 2 months (P=.045). 0.6 Compared with the latanoprost group, eyes receiving bi- 0.5 matoprost showed significantly higher hyperemia scores 0.4 from 7 days to 2 months (PϽ.04). Patients treated with 0.3 travoprost had significantly higher hyperemia scores than 0.2 those treated with unoprostone at 15 days (P=.03) and 0.1 than the control group at 7 days (P=.009), 15 days Conjunctival Hyperemia Score 0 (P=.01), and 1 month (P=.03). Control eyes showed sig-

Baseline 7 d 15 d 1 mo 2 mo 3 mo 4 mo 5 mo 6 mo nificantly lower hyperemia scores than latanoprost- Follow-up treated eyes at 15 days (P=.01) and 1 month (P=.03). There were no significant differences regarding hyperemia scores Figure 4. Mean conjunctival hyperemia scores from baseline to month 6 for between the unoprostone and placebo (PϾ.14), latano- each treatment group. prost and travoprost (PϾ.12), and latanoprost and unoprostone (PϾ.13) groups. The bimatoprost group had a significantly higher number of eyes (n=8, 50%) with peak for the unoprostone group, 26% (5.4 mm Hg) for the la- hyperemia scores of 1 or greater than the latanoprost (n=1, tanoprost group, 28% (5.8 mm Hg) for the bimatoprost 7%) (P=.02), travoprost (n=2, 12%) (P=.03), unopros- group, and 29% (5.9 mm Hg) for the travoprost group. tone (n=1, 6%) (P=.02), and control (n=0) (P=.002) Patients receiving unoprostone had significantly higher Ͻ groups. There was a weak correlation between mean flare IOP reductions than the control group (P .05), but the values and hyperemia scores according to the Spearman mean hypotensive effect was significantly lower than that rank correlation test (r=0.143; P=.005). of the other PG analogues at all times (PϽ.05). Mean IOP reductions did not differ among the bimatoprost, latanoprost, and travoprost groups throughout follow-up COMMENT (PϾ.05). Changes in hyperemia scores throughout follow-up are In this study, all the PG analogues statistically signifi- displayed in Table 4 and Figure 4. At baseline, all pa- cantly reduced IOP in pseudophakic or aphakic eyes with tients were classified as having no hyperemia. There was a glaucoma. Our findings suggest that bimatoprost, latano- significant increase in hyperemia scores in the latanoprost, and travoprost administered once daily, however, prost, bimatoprost, and travoprost groups 1 week after base- are significantly more effective in reducing IOP than uno- line. Hyperemia scores reached their peak 15 days after base- prostone administered twice daily. Twice-daily unopros- line and started to decrease 1 month after therapy was tone use was associated with a 14% reduction in IOP from initiated. Twelve patients in the bimatoprost group (75%), baseline after 6 months, which is in accordance with pre- 8 in the travoprost group (47%), 6 in the latanoprost group vious studies.10,27 The ocular hypotensive effects of la- (40%), 4 in the unoprostone group (25%), and 1 in the con- tanoprost (26%), bimatoprost (28%), and travoprost trol group (6%) showed conjunctival hyperemia during fol- (29%) did not differ statistically. Although the study was low-up (P=.001). Median (range) hyperemia scores dur- not primarily designed to compare the hypotensive ef- ing follow-up were 0 (0-0.5) for placebo, 0 (0-1.0) for fects of PG analogues, our sample had 80% power to de- unoprostone, 0 (0-1.0) for latanoprost, 0 (0-2.0) for tra- tect IOP differences greater than 1.07 mm Hg among the voprost, and 0.28 (0-3.0) for bimatoprost (PϽ.001). groups at a significance level of ␣=.05. Comparing the treatment groups at all times, we ob- The intensity of ocular hyperemia (the number of eyes served that the hyperemia scores were significantly higher with a peak hyperemia score Ն1) was also greater in the

(REPRINTED) ARCH OPHTHALMOL / VOL 123, FEB 2005 WWW.ARCHOPHTHALMOL.COM 190

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 bimatoprost group compared with the latanoprost, uno- sule, the presence of an anterior chamber intraocular lens, prostone, and control groups. This finding agrees with pre- and diabetes mellitus. Wand and Gaudio25 described 3 vious comparative studies8,9,11,28 that evaluated hyperemia cases of CME related to PG analogues. All 3 patients had after the use of topical PG analogues. Stewart et al8 evalu- previous cataract and filtration surgery and an absent or ated conjunctival hyperemia after short-term use of latano- open posterior capsule. Decreased vision of at least 2 prost, bimatoprost, and travoprost in 28 healthy adults in Snellen lines caused by angiographically confirmed CME a crossover study with a washout interval of 1 week. The was noted in each of the 3 patients, who were receiving results suggested that latanoprost therapy may cause sig- unoprostone, travoprost, and bimatoprost. nificantly less short-term conjunctival hyperemia on av- Apart from these isolated case reports, only 1 short- erage than bimatoprost and travoprost use in healthy adults. term, prospective study20 describes the occurrence of CME Parrish et al9 compared latanoprost, bimatoprost, and tra- in latanoprost-treated eyes early after cataract surgery. voprost use in 410 patients with primary open-angle glau- Of the 37 eyes receiving latanoprost and fluorometho- coma or ocular hypertension. Masked investigators’ assess- lone in the postoperative period, 30 (81%) developed an- ments of hyperemia were similar across treatments at giographic CME. In previous retrospective stud- baseline. However, at weeks 2 and 12, mean hyperemia ies18,19,21-23,25 that evidenced PG-related CME, FA was scores were significantly lower for latanoprost-treated than performed when the patient complained about visual acu- for bimatoprost-treated patients. Hyperemia was consis- ity loss. Hence, it is possible that some of these patients tently rated lowest in the latanoprost group and highest in already had angiographic CME or a disruption in the the bimatoprost group, with patients in the travoprost group blood-aqueous barrier before starting PG analogue receiving intermediate mean ratings. therapy. To investigate the effects of PG analogues on the blood- The present study is the first to prospectively evalu- aqueous barrier, we used the laser flare meter, an objec- ate the long-term effects of all commercially available PG tive, noninvasive, and reproducible technique,26,29 to quan- analogues on the blood-aqueous barrier in aphakic and titatively measure flare in the anterior chamber in vivo.30 pseudophakic eyes with glaucoma. Pretreatment mea- In a randomized, double-masked trial, Miyake et al20 stud- surements of aqueous flare and FAs were performed to ied the occurrence of blood-aqueous barrier changes and ensure that no patient had a disruption of the blood- angiographic CME in eyes undergoing phacoemulsifica- aqueous barrier before taking the PG analogue, and no tion and intraocular lens implantation that received la- patient with a history of uveitis or CME was allowed to tanoprost in the early postoperative period. The authors participate in the study. Notwithstanding, we demon- also used the laser flare meter and reported that latano- strated that latanoprost, bimatoprost, and travoprost are prost therapy led to a disruption of the blood-aqueous capable of inducing CME in pseudophakic or aphakic eyes barrier and significantly increased the incidence of an- even in lower risk cases. The fact that only 1 of the 6 pa- giographic CME. Furthermore, they demonstrated that tients with angiographic CME had visual complaints sug- these adverse effects were prevented when nonsteroidal gests that the incidence of CME in patients with pseu- anti-inflammatory drops were given concurrently. dophakia and aphakia treated with PG analogues may be The present study demonstrates that the use of bima- higher than previously recognized. toprost, travoprost, and latanoprost in pseudophakic or Although the precise pathogenesis of CME is not yet aphakic eyes may lead to a disruption in the blood- fully understood, a variety of mechanisms are proposed aqueous barrier. These abnormalities were not detected to explain the disruption of the blood-retinal barrier.31 in the control and unoprostone groups. We also dem- The concentration of benzalkonium chloride is 0.02% in onstrated that disruption of the blood-aqueous barrier latanoprost, 0.015% in travoprost and unoprostone, 0.01% is not associated with the degree of conjunctival hyper- in placebo, and 0.005% in bimatoprost. The facts that bi- emia but is significantly associated with the develop- matoprost induced blood-aqueous barrier changes with ment of CME. a lower concentration of benzalkonium chloride and that Several isolated studies18-23,25 have retrospectively de- unoprostone did not result in blood-aqueous barrier scribed the development of CME in patients with pseudo- changes despite a higher concentration of benzalko- phakia or aphakia using PG analogues. Lima et al18 de- nium chloride suggest that the preservative was not re- scribed 3 of 185 patients with pseudophakia or aphakia who sponsible for the changes we observed. Numerous clini- experienced visually significant CME while using latano- cal trials support the role of inflammatory mediators, such prost for a mean of 10 months. Watanabe et al23 reported as PGs, in this disruption, with evidence that anti- the development of visually significant CME 1 month af- inflammatory agents may prevent or affect the recovery ter the use of topical latanoprost in 1 pseudophakic eye with of CME after cataract surgery.32 It has been shown that glaucoma and a ruptured posterior capsule. The CME dis- PGF2␣ may induce the release of endogenous PGs from appeared 2 weeks after the discontinuation of latanoprost the iris and ciliary muscles of mammals, including hu- therapy and the use of topical 0.1% betamethasone diso- mans.33 The facts that patients who developed CME had dium and 0.1% diclofenac sodium. significantly higher mean flare values than those who did Most reported cases of latanoprost-associated not develop this complication and that the discontinu- CME18,19,21-23 occurred in patients with coexisting ocular ation of the PG analogues associated with the use of di- or systemic conditions that increase the risk of CME. clofenac induced CME recovery are compatible with a These conditions include a history of CME or anterior PG-mediated break in the blood-retinal barrier. Never- uveitis, epiretinal membrane, vein occlusion, compli- theless, there must be other factors involved because most cated cataract surgery, absent or ruptured posterior cap- of even the high-risk eyes receiving ocular PG ana-

(REPRINTED) ARCH OPHTHALMOL / VOL 123, FEB 2005 WWW.ARCHOPHTHALMOL.COM 191

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/30/2021 logues do not develop CME.24 Broken or absent poste- 4. Hurvitz LM, Kaufman PL, Robin AL, Weinreb RN, Crawford K, Shaw B. New developments in the drug treatment of glaucoma. Drugs. 1991;41:514-532. rior capsules are known as risk factors for the develop- 5. Alexander CL, Miller SJ, Abel SR. Prostaglandin analog treatment of glaucoma and ocular ment of CME in pseudophakic or aphakic eyes.24,34 Our hypertension. Ann Pharmacother. 2002;36:504-511. study indicated borderline statistical significance in the 6. Linden C. Therapeutic potential of prostaglandin analogues in glaucoma. Expert Opin Investig Drugs. 2001;10:679-694. association between CME and broken posterior cap- 7. Stamper RL, Wigginton SA, Higginbotham EJ. Primary drug treatment for glaucoma: ␤-block- sules, possibly a consequence of the small sample sizes ers versus other medications. Surv Ophthalmol. 2002;47:63-73. 8. Stewart WC, Kolker AE, Stewart JA, Leech J, Jackson AL. Conjunctival hyperemia in healthy included in the analysis. subjects after short-term dosing with latanoprost, bimatoprost, and travoprost. Am J Ophthalmol. 35-37 Receptor binding studies have demonstrated that 2003;135:314-320. bimatoprost, latanoprost, and travoprost have high af- 9. Parrish RK, Palmberg P, Sheu WP; XLT Study Group. A comparison of latanoprost, bimatoprost, and travoprost in patients with elevated intraocular pressure: a 12-week, randomized, finity for the FP receptor, whereas the affinity of uno- masked-evaluator multicenter study. Am J Ophthalmol. 2003;135:688-703. prostone for the FP receptor is 100-fold lower. It is not 10. Sponsel WE, Paris G, Trigo Y, Pena M. Comparative effects of latanoprost (Xalatan) and uno- clear whether binding to the PG FP receptor may be in- prostone (Rescula) in patients with open-angle glaucoma and suspected glaucoma. Am J Ophthalmol. 2002;134:552-559. volved in the pathogenesis of PG-induced CME, but this 11. Netland PA, Landry T, Sullivan EK, et al; Travoprost Study Group. Travoprost compared with could explain the higher incidence of CME and the sig- latanoprost and timolol in patients with open-angle glaucoma or ocular hypertension. Am J nificant increase in aqueous flare in patients treated with Ophthalmol. 2001;132:472-484. 12. Noecker RS, Dirks MS, Choplin NT, Bernstein P, Batoosingh AL, Whitcup SM; Bimatoprost/ bimatoprost, latanoprost, and travoprost compared with Latanoprost Study Group. A six-month randomized clinical trial comparing the intraocular pres- unoprostone-treated patients. sure–lowering efficacy of bimatoprost and latanoprost in patients with ocular hypertension or There is not a single medication or surgical treat- glaucoma. Am J Ophthalmol. 2003;135:55-63. 13. DuBiner H, Cooke D, Dirks M, Stewart WC, VanDenburgh AM, Felix C. Efficacy and safety of ment without potential adverse effects and complica- bimatoprost in patients with elevated intraocular pressure: a 30-day comparison with latanoprost. tions. It is fundamental to be aware of the potential risks Surv Ophthalmol. 2001;45(suppl 4):S353-S360. 14. Higginbotham EJ, Schuman JS, Goldberg I, et al. Bimatoprost Study Groups 1 and 2. One- of any therapy we recommend and to choose an ad- year, randomized study comparing bimatoprost and timolol in glaucoma and ocular hypertension. equate treatment plan with the best benefit-risk ratio for Arch Ophthalmol. 2002;120:1286-1293. our patients.24 In this study, we confirmed that the use 15. Eisenberg DL, Toris CB, Camras CB. Bimatoprost and travoprost: a review of recent studies of two new glaucoma drugs. Surv Ophthalmol. 2002;47(suppl 1):S105-S115. of bimatoprost, latanoprost, and travoprost may lead to 16. Camras CB, Wax MB, Ritch R, et al. Latanoprost treatment for glaucoma: effects of treating for the disruption of the blood-aqueous barrier and the de- 1 year and of switching from timolol. Am J Ophthalmol. 1998;126:390-399. velopment of CME in pseudophakic and aphakic eyes, 17. Camras CB, Toris CB, Tamesis RR. Efficacy and adverse effects of medications used in the treatment of glaucoma. Drugs Aging. 1999;15:377-388. even in the absence of a history of uveitis or CME. We 18. Lima MC, Paranhos A Jr, Salim S, et al. Visually significant cystoid macular edema in pseu- recommend caution when prescribing bimatoprost, tra- dophakic and aphakic patients with glaucoma receiving latanoprost. J Glaucoma. 2000;9: voprost, and latanoprost to patients with aphakic or pseu- 317-321. 19. Warwar RE, Bullock JD, Ballal D. Cystoid macular edema and anterior uveitis associated with dophakic eyes with glaucoma. When therapy with one latanoprost use: experience and incidence in a retrospective review of 94 patients. Ophthalmology. of these PG analogues is started in such a patient, care- 1998;105:263-268. ful monitoring of the blood-aqueous barrier status (by 20. Miyake K, Ota I, Maekubo K, Ichihashi S, Miyake S. Latanoprost accelerates disruption of the blood-aqueous barrier and the incidence of angiographic cystoid macular edema in early post- slitlamp examination) and visual acuity is advocated. operative pseudophakias. Arch Ophthalmol. 1999;117:34-40. 21. Ayyala RS, Cruz DA, Margo CE, et al. Cystoid macular edema associated with latanoprost in aphakic and pseudophakic eyes. Am J Ophthalmol. 1998;126:602-604. Submitted for Publication: December 8, 2003; final re- 22. Moroi SE, Gottfredsdottir MS, Schteingart MT, et al. Cystoid macular edema associated with vision received XX 2004; accepted June 24, 2004. latanoprost therapy in a case series of patients with glaucoma and ocular hypertension. Correspondence: Enyr S. Arcieri, MD, Department of Ophthalmology. 1999;106:1024-1029. 23. Watanabe K, Hayasaka S, Hayasaka Y, Nagaki Y, Watanabe K. Cystoid macular edema asso- Ophthalmology, School of Medicine, Federal Univer- ciated with latanoprost use in a pseudophakic eye with a history of surgical complications. sity of Uberlândia, Rua Corcovado, No. 155, ap 402, Uber- Jpn J Ophthalmol. 2003;47:110-112. lândia, MG, 38411-092, Brazil ([email protected]). 24. Wand M, Shields BM. Cystoid macular edema in the era of ocular hypotensive lipids. Am J Ophthalmol. 2002;133:393-397. Funding/Support: This study was supported by grant 25. Wand M, Gaudio AR. Cystoid macular edema associated with ocular hypotensive lipids. Am J 2001/09520-0 from Fundação de Amparo à Pesquisa do Ophthalmol. 2002;133:403-405. 26. Oshika T, Araie M, Masuda K. Diurnal variation of aqueous flare in normal human eyes mea- Estado de São Paulo, São Paulo, Brazil. sured with laser flare-cell meter. Jpn J Ophthalmol. 1988;32:143-150. Previous Presentation: This study was presented in part 27. Susanna R Jr, Giampani J Jr, Borges AS, Vessani RM, Jordao ML. A double-masked, random- as a poster at the Association for Research in Vision and ized clinical trial comparing latanoprost with unoprostone in patients with open-angle glaucoma or ocular hypertension. Ophthalmology. 2001;108:259-263. Ophthalmology Annual Meeting Scientific Program; April 28. Gandolfi S, Simmons ST, Sturm R, Chen K, VanDenburgh AM; Bimatoprost Study Group 3. 25, 2004; Fort Lauderdale, Fla. Three-month comparison of bimatoprost and latanoprost in patients with glaucoma and ocu- Acknowledgment: We thank Heyder Diniz Silva, PhD, lar hypertension. Adv Ther. 2001;18:110-121. 29. Sawa M. Clinical application of laser flare-cell meter. Jpn J Ophthalmol. 1990;34:346-363. Mathematic School, Federal University of Uberlândia, 30. Sen HN, Uusitalo R, Laatikainen L. Subclinical inflammation after laser in situ keratomileusis for statistical consultation and assistance, and Magno in corneal grafts. J Cataract Refract Surg. 2002;28:782-787. 31. Jampol LM. Macular edema. In: Ryan SJ, ed-in-chief. Retina. Vol 2. St Louis, Mo: Mosby– Antônio Ferreira, MD, Retina and Vitreous Service, Year Book Inc; 1989:81-88. Department of Ophthalmology, Federal University of 32. Rossetti L, Chaudhuri J, Dickersin K. Medical prophylaxis and treatment of cystoid macular edema Uberlândia, for evaluation of fluorescein angiography. after cataract surgery: the results of a meta-analysis. Ophthalmology. 1998;105:397-405. 33. Yousufzai SY, Ye Z, Abdel-Latif AA. Prostaglandin F2␣ and its analogs induce release of endogenous prostaglandins in iris and ciliary muscles isolated from cat and other mammalian REFERENCES species. Exp Eye Res. 1996;63:305-310. 34. Schumer RA, Camras CB, Mandahl AK. Putative side effects of prostaglandin analogs. Surv Ophthalmol. 2002;47(suppl 1):S219-S230.

1. Giuffrè G. The effects of prostaglandin F2␣ in the human eye. Graefes Arch Clin Exp Ophthalmol. 35. Sharif NA, Williams GW, Kelly CR. Bimatoprost and its free acid are prostaglandin FP receptor 1985;222:139-141. agonists. Eur J Pharmacol. 2001;432:211-213. 2. Resul B, Stjernschantz J, No K, et al. Phenyl-substituted prostaglandins: potent and selective 36. Hellberg MR, Sallee VL, McLaughlin MA, et al. Preclinical efficacy of travoprost, a potent and antiglaucoma agents [published correction appears in J Med Chem. 1993;36:2242]. J Med Chem. selective FP prostaglandin receptor agonist. J Ocul Pharmacol Ther. 2001;17:421-432. 1993;36:243-248. 37. Sharif NA, Kelly CR, Crider JY. Agonist activity of bimatoprost, travoprost, latanoprost, uno- 3. Alm A. Prostaglandin derivates as ocular hypotensive agents. Prog Retin Eye Res. 1998;17:291- prostone isopropyl ester and other prostaglandin analogs at the cloned human ciliary body FP 312. prostaglandin receptor. J Ocul Pharmacol Ther. 2002;18:313-324.

(REPRINTED) ARCH OPHTHALMOL / VOL 123, FEB 2005 WWW.ARCHOPHTHALMOL.COM 192