Intraocular Pressure-Lowering Combination Therapies With

Home , Dorzolamide, Glaucoma medication, Prostaglandin, Tafluprost

Drugs 2012; 72 (10): 1355-1371 REVIEW ARTICLE 0012-6667/12/0010-1355/$55.55/0

Intraocular Pressure-Lowering Combination Therapies with Prostaglandin Analogues Florent Aptel,1,2 Christophe Chiquet1,2 and Jean-Paul Romanet1,2 1 Universite´ Joseph Fourier (UJF), Grenoble, France 2 Department of Ophthalmology, University Hospital, CHU Grenoble, Grenoble, France

Contents

Abstract...... 1355 1. Introduction ...... 1356 1.1 Literature Search Strategy ...... 1357 2. Combination Therapies with Prostaglandin Analogues (PGAs): Unfixed Combinations ...... 1357 2.1 PGA–b-Blocker ...... 1357 – – 2.2 PGA a2-Adrenergic Agonist and PGA Carbonic Anhydrase Inhibitor ...... 1358 3. Combination Therapies with PGAs: Fixed Combinations ...... 1360 3.1 Rationale for Using Fixed Combinations ...... 1360 3.2 Comparative Efficacy and Tolerability vs Monotherapies ...... 1361 3.3 Comparative Efficacy and Tolerability vs Unfixed Combinations ...... 1363 3.3.1 Results of Clinical Trials ...... 1363 3.3.2 Limitations of the Currently Available Clinical Trials ...... 1363 3.4 Direct Comparisons of the Three PGA–b-Blocker Fixed Combinations...... 1364 3.5 Comparative Efficacy and Tolerability with Other b-Blocker-Based Fixed Combinations . . . . . 1365 4. Discussion and Conclusions ...... 1366 5. Key Practical Points for Clinicians...... 1367

Abstract Intraocular pressure (IOP) reduction is currently the only therapeutic approach demonstrated to preserve visual function in patients with glaucoma. The first line of glaucoma treatment consists of topical IOP-lowering medications, usually initiated as monotherapy. A significant proportion of patients require more than one medication to reach a target IOP at which optic nerve damage will not progress. As prostaglandin analogues (PGAs) are the most effective class for reducing IOP, one of the other commonly used classes (b-adrenoceptor antagonist [b-blocker], carbonic anhydrase inhibitor or a2-adrenoceptor agonist) is frequently combined with a PGA. In the last decade, the use of fixed combinations containing two medications in a single bottle has steadily increased. Fixed combinations have the potential to simplify the dosing regimen, increase patient adherence, avoid the washout effect of the second drop on the first medication instilled, decrease exposure to preservatives and, sometimes, reduce the cost of treatment. Clinical trials have evaluated PGA-based fixed combinations versus unfixed combinations 1356 Aptel et al.

(individual components administered concomitantly) or versus individual monotherapies; however, any advantage that the fixed combinations may have in terms of IOP-lowering efficacy is still debated. For these reasons, the PGA-based fixed combinations are not approved by regulatory authorities in some countries, such as the US. We review the published studies evaluating the efficacy and tolerability of the IOP-lowering unfixed and fixed combination therapies with PGAs. Regarding unfixed combinations, the review shows that a2-adrenergic agonists–PGA and carbonic anhydrase inhibitor–PGA combinations seem to be at least as effective at reducing IOP as the b-blocker–PGA combinations. As for the fixed combinations, the review shows that the three PGA–timolol fixed combinations are more effective than their component medications used separately as monotherapy and are better tolerated than the three respective prostaglandins. The three PGA–timolol fixed combinations are less effective at reducing IOP than the unfixed combinations but are better tolerated. The advantage of the fixed combinations in terms of patient adherence and persistence is supported by a very small number of studies and remains to be more accurately determined. Most studies, but not all, seem to show that PGA–timolol fixed combinations are more effective than other available b-blocker fixed combinations (dorzolamide–timolol fixed combinations) at reducing IOP and are similarly tolerated.

1. Introduction open-angle glaucoma. Because of their efficacy, convenience (once-daily application) and minimal Glaucoma is a progressive optic neuropathy systemic side effects, prostaglandin analogues caused by the degeneration and death of retinal (PGAs) launched in the last decade have nowadays ganglion cells and their axons that form the optic become the first line of treatment. Many clinical nerve, causing visual field reduction until blind- trials and meta-analyses have found that PGAs ness.[1] Primary open-angle glaucoma is the most have a greater ability to reduce IOP than the other common form of glaucoma and the second lead- prescribed therapeutic classes, including b-block- ing cause of blindness in developed countries.[2] ers.[7-9] For example, a meta-analysis has shown The biological mechanisms of the retinal ganglion that the three available PGAs can provide an IOP cell degeneration are not precisely known; however, reduction ranging from -31% to -33% at peak and many risk factors of glaucoma occurrence or pro- from -28% to -29% at trough.[7] As a class, the gression have been identified, elevated intraocular PGAswerefoundtobemoreeffectiveatreduc- pressure (IOP) being the only proven treatable risk ing IOP than the other classes of medications factor. Lowering IOP is currently the only ther- (b-blockers, carbonic anhydrase inhibitors, a2- apeutic approach demonstrated to preserve visual adrenoceptor agonist) both at peak and trough. function. Over the past decade, several large clinical For this reason, and also because of the limited studies have proven that IOP reduction can delay systemic side effects and regimen convenience, or prevent the progression of ocular hypertension PGAs are recommended as first-line therapy by to glaucoma and can delay disease worsening in medical society guidelines (European Glaucoma established glaucoma.[3-6] Society guideline,[10] American Academy of For many years, b-adrenoceptor antagonists Ophthalmology preferred practice[11]). (b-blockers) have been the first choice in medical Despite the efficacy of the PGAs, a significant treatment of ocular hypertension and primary proportion of patients require more than one

medication to reach an IOP at which optic nerve ‘adrenergic agonist’, ‘combination’, ‘association’ damage will not progress (target IOP). The Ocular and ‘medical treatment’, both alone and in com- Hypertension Study found that almost 40% of binations. All studies were identified using a two- patients required two or more medications to level search strategy: (i) databases in the public reach a given target IOP.[3] Similarly, the results domain were searched using web-based search of the Collaborative Initial Glaucoma Treatment engines; and (ii) relevant studies were identified Study showed that medically treated patients re- through a manual search of secondary sources, quired a median of two medications to control including the references of articles found from the the disease 5 years after treatment initiation.[12] databases and information obtained directly To achieve a better IOP reduction, combination from experts and pharmaceutical companies. All therapy with drugs of differing but complemen- references were downloaded for consolidation, tary mechanisms of action are preferred. As PGAs elimination of duplicates and further scrutiny. are the most effective therapeutic class, one of the Additional searches were conducted for known other classes is frequently combined with a PGA. trial acronyms cited in review articles. Several unfixed drug combinations including a PGA are available. However, unfixed combina- 2. Combination Therapies with tions increase regimen complexity, potentially Prostaglandin Analogues (PGAs): causing decreased adherence and persistence, and Unfixed Combinations require a waiting time between two instillations in order to avoid drug washout.[13-22] For these rea- Four PGAs are currently available: latano- sons, fixed combinations containing two medica- prost (Xalatan, Pfizer, New York, NY, USA), tions in a single bottle have been developed and travoprost (Travatan, Alcon, Fort Worth, TX, marketed. Fixed combinations simplify the dos- USA), bimatoprost (Lumigan, Allergan, Irvine, ing regimen and may thus improve compliance CA, USA) and tafluprost (Taflotan, Santen, and persistence. Fixed combinations also decrease Osaka, Japan; or Saflotan, Merck, White House the daily exposure to preservatives such as ben- Station, NJ, USA). Tafluprost is only authorized zalkonium chloride, which is known to have toxic in a few countries (Japan and recently some effects for the ocular surface and to increase treat- European countries). To date, tafluprost is not ment side effects.[23] Finally,thecostofafixed available as a fixed combination. To the best of combination is frequently less than the combined our knowledge, only one published study has evalu- cost of the two ingredients prescribed separately. ated the efficacy and tolerability of an unfixed We undertook a review of the published stud- combination of tafluprost with timolol.[24] ies evaluating the efficacy and tolerability of the IOP-lowering unfixed and fixed combina- 2.1 PGA–b-Blocker tion therapies with PGAs. We particularly reviewed the studies comparing the efficacy and Many previous studies have evaluated the ad- tolerability of the unfixed and fixed combinations ditive effect of combining a b-blocker with a PGA with monotherapies. in an unfixed manner.[25-31] Most of these studies evaluating the unfixed combination of latano- 1.1 Literature Search Strategy prost and timolol have been summarized in published meta-analyses. We searched the MEDLINE and Google Scholar In 2007, Webers et al.[32] analysed three studies databases up to January 2012, with the following evaluating the IOP-lowering effect of 0.005% search terms: ‘glaucoma’, ‘ocular hypertension’, ‘la- latanoprost added to 0.5% timolol.[25-27] IOP tanoprost’, ‘travoprost’, ‘bimatoprost’, ‘tafluprost’, reduction 12 hours after latanoprost instillation ‘timolol’, ‘dorzolamide’, ‘brinzolamide’, ‘brimoni- and 0 hours after timolol instillation was -5.7 dine’, ‘apraclonidine’, ‘prostaglandin’, ‘prostamide’, mmHg (95% CI -4.0, -7.4); relative IOP reduc- ‘beta-blocker’, ‘carbonic anhydrase inhibitor’, tion was -24.6% (95% CI -17.4, -31.8). Over the

diurnal curve, IOP reduction was -3.0 (95% CI event and was more frequent in the groups re- -2.2, -3.8); relative IOP reduction was -13.4% ceiving tafluprost plus timolol than in patients (95% CI -10.8, -16.0). In another meta-analysis, treated with timolol alone. Webers et al.[33] evaluated three studies and % – found that adding 0.005 latanoprost once daily 2.2 PGA a2-Adrenergic Agonist and in the evening to 0.5% timolol twice daily resulted PGA–Carbonic Anhydrase Inhibitor in a pooled change of -6.3 mmHg (95% CI -5.5, -7.1); relative change -28.3% (95% CI -23.7, The additive efficacy of carbonic anhydrase - [26-28] 32.8) in mean IOP curve. inhibitors or a2-adrenergic agonists with PGAs In a large multicentre, double-blind, parallel- has also been investigated.[34-54] The character- group trial, Schuman et al.[29] found that a con- istics of the studies identified are summarized in comitant regimen of 0.5% timolol given every day table I. Most of these studies found that both a % in the morning, plus 0.004 travoprost every day carbonic inhibitor and an a2-adrenergic agonist in the evening, was significantly more effective in added to a PGA provided a significant additional reducing IOP than 0.5% timolol given twice daily. IOP reduction. In a retrospective study, O’Connor Mean IOP change at month 3 was -7.8 mmHg et al.[34] evaluated three classes of additive ther- (-32.5%) versus -6.4 (-25.8%) with 0.5% timolol apy in 73 eyes with IOP not sufficiently lowered alone. Hyperaemia was the most frequent treat- by latanoprost monotherapy. Added to latano- ment-related ocular event and was more frequent prost, 2% dorzolamide (twice or three-times daily) in the groups receiving travoprost plus timolol lowered IOP an additional 3.9 mmHg (19.7%), than in patients treated with timolol alone (23.4% and brimonidine an additional 2.0 mmHg (9.3%). vs 2.4%). Other treatment-related ocular adverse In this study, the additive effect of timolol or effects were more frequent in patients receiving other b-blockers was also evaluated. From a la- travoprost plus timolol, such as ocular discomfort, tanoprost baseline, b-blockers further reduced foreign body sensation and pruritus (7.6% vs 4.8%, IOP by 2.0 mmHg (12.3%). The IOP reduction 4.4% vs 2.4% and 5.1% vs 0%, respectively). provided by the adjunction of dorzolamide to In another large multicentre, double-blind, latanoprost was statistically significantly higher parallel-group trial, Hommer et al.[30] found that than those provided by b-blockers and brimoni- a concomitant regimen of 0.03% bimatoprost dine (p < 0.0015). No significant difference in given every day in the evening plus 0.5% timolol IOP reduction was noted between twice-daily and every day in the morning was significantly more three-times-daily dorzolamide dosing regimens effective in reducing IOP than 0.03% bimatoprost (p = 0.92). alone given in the evening. The mean IOP change A few studies have directly compared the effi- at week 3 was -9.6 mmHg (-38%) with the un- cacy of a b-blocker (usually 0.5% timolol) with fixed combination versus -7.9 (-31.6%) with the efficacy of a carbonic anhydrase inhibitor or bimatoprost alone. Hyperaemia was the most an a2-adrenergic agonist as an unfixed adjunctive frequent treatment-related ocular event in both therapy to a PGA.[36,37,41-45,48-51] Most studies groups, with no significant differences (25.6% demonstrated that a carbonic anhydrase inhibitor % with the combination vs 27.8 in patients treated or an a2-adrenergic agonist is usually equally or with bimatoprost alone). The rate of the other more effective at reducing IOP than a b-blocker treatment-related side effects was comparable when added to a PGA. Some studies have found between the two groups. that carbonic anhydrase inhibitors could be a In the only clinical trial evaluating the ad- more effective class to use adjunctively with PGA [24] [36,49,51] junctive effect of tafluprost, Egorov et al. found compared with an a2-adrenergic agonist, that tafluprost given as adjunctive therapy to whereas others have found an opposite trend.[42,48] 0.5% timolol twice a day provides an additional In a prospective parallel-group clinical trial, IOP reduction: -6.2 to -6.8 mmHg. Hyperaemia Bournias and Lai[48] evaluated the efficacy of was the most frequent treatment-related ocular brimonidine, dorzolamide and brinzolamide in

Adis ª 2012 Springer International Publishing AG. All rights reserved. Drugs 2012; 72 (10) Adis Table I. Trials comparing the prostaglandin analogue-based unfixed combinations with monotherapies or other unfixed combinations Glaucoma in Therapies Combination Prostaglandin-Based

ª Study Study arms Efficacy Pts Methodology Results 02Srne nentoa ulsigA.Alrgt eevd rg 02 2(10) 72 2012; Drugs reserved. rights All AG. Publishing International Springer 2012 assessment (n) period O’Connor et al.[34] LDUFC vs LBrimUFC vs 1 y 73 Retrospective interventional case series Added to latanoprost, 19.7% additional IOP LTUFC reduction with dorzolamide, 12.3% with b-blockers and 9.3% with brimonidine Erdogan et al.[35] LBrimUFC vs latanoprost 5 d 32 PRO, SC, R, DB, CO; 4 wk WO period Additional 2.5- to 3.1-mmHg decrease in IOP in the and then 5 d RI period with latanoprost latanoprost+brimonidine group vs no additional IOP reduction in the latanoprost+placebo group Konstas et al.[36] LBrimUFC vs LDUFC 6 wk 31 PRO, double-centre, R, DB, CO; 6 wk Mean 24 h IOP reduction was 11.0% with latanoprost RI LBrimUFC and 11.6% with LDUFC Tsukamoto et al.[37] LBrinUFC vs 8 wk 58 PRO, SC, R, OP, PG study; Mean diurnal IOP reduction was 0.1% with LDUFC LDUFC latanoprost+dorzolamide RI and 1.7% with LBrinUFC Shoji et al.[38] LBrinUFC 3 mo 14 PRO, MC, OP, NC study; latanoprost RI Mean IOP reduction was 24.6% with LBrinUFC Mundorf et al.[39] LBrimUFC 2 mo 43 PRO, MC, OP, NC study; latanoprost RI Mean 10am IOP reduction was 23% at peak 2 mo after brimonidine adjunction Nakamoto and LBrinUFC vs latanoprost 8 wk 22 PRO, SC, R, OP, PG (1 eye vs other) Diurnal mean IOP reduction: latanoprost+ Yasuda[40] study; at least 4 wk WO period and then brinzolamide 19.8%, latanoprost 14.1%; nocturnal 8 wk latanoprost RI mean IOP reduction latanoprost+brinzolamide 13.4%, latanoprost 10.0% Tamer and Oksu¨z[41] LTUFC vs LDUFC 4 wk 36 PRO, SC, R, DB, CO study; 24 h mean IOP reduction was 14.2% with LDUFC latanoprost RI and 10.6% with LTUFC Day and Hollander[42] LBrimUFC vs LBrinUFC 3 mo 40 PRO, SC, R, investigator-blind, Diurnal mean IOP reduction was 16.8% with PG study; latanoprost RI LBrimUFC and 10.1% with LBrinUFC Ishikawa and LBrinUFC vs LTUFC 3 mo 30 PRO, SC, NR, OP, CO study; 4 wk WO 24 h mean IOP reduction was 33.3% with LTUFC Yoshitomi[43] period and 37.0% with LBrinUFC Liu et al.[44] LBrinUFC vs LTUFC 8 wk 26 PRO, SC, R, OP, CO study; Mean diurnal IOP reduction was 4.6% with latanoprost RI LBrinUFC and 0.5% with LTUFC Nakamura et al.[45] LDUFC vs LBrinUFC 4 wk 20 PRO, SC, R, CO study; latanoprost RI Mean diurnal IOP reduction was 19.3% with LDUFC bid, 21.1% and 18.1% with LBrinUFC Inoue et al.[46] LDUFC 12 mo 46 PRO, MC, NC, OP study; latanoprost RI Mean diurnal IOP reduction was 15.1% with LDUFC Li and Montgomery[47] LBrinUFC 4 mo 93 Retrospective chart review; latanoprost Mean diurnal IOP reduction was 19.7% with RI LBrinUFC at 4.0 mo median follow-up Bournias and Lai[48] PGA+dorzolamide or 4 mo 120 PRO, SC, R, investigator-blind, Mean IOP reduction at 10am and 4pm was 21% brinzolamide or PG study; 6 wk PGA RI and 19% with brimonidine–PGA UFC, 16% and brimonidine 14% with dorzolamide–PGA UFC, and 16% and 13% with brinzolamide–PGA UFC

Continued next page 1359 1360 Aptel et al. = = reducing IOP when used as adjunctive therapy

WO to latanoprost, travoprost or bimatoprost. After with 4 months of adjunctive treatment, the mean IOP % with with multicentre; single-blind; with LBrimUFC = % % % = reduction from baseline at 10am and 4pm was % SB MC % % 4.8 mmHg (21%) and 3.8 mmHg (19%) with brimonidine, 3.4 mmHg (16%) and 2.8 mmHg (14%)

run in; with dorzolamide and 3.4 mmHg (16%) and = with TBrinUFC unfixed combination; RI 2.6 mmHg (13%) with brinzolamide (p < 0.001 for = % with TTUFC

timolol UFC; brimonidine vs dorzolamide and brinzolamide at % intraocular pressure; – UFC with TBrinUFC and 13.4 = each timepoint). All three adjunctive therapies %

IOP seem to be similarly well tolerated, with no pa- randomized;

= tient discontinuation. R latanoprost = timolol UFC; – TBrimUFC TTUFC, 22.7 TBrinUFC and 15 Mean diurnal IOP reduction was 20.2 Mean diurnal IOP reduction was 15.8 TBrimUFC and 12.8 Mean diurnal IOP reduction was 9.7 Mean diurnal IOP reduction was 35.1 Mean 24 h IOP reduction was 5.1

double-blind; 3. Combination Therapies with PGAs: LTUFC = Fixed Combinations prospective; DB = travoprost = 3.1 Rationale for Using Fixed Combinations PRO crossover; TTUFC The use of fixed combination medications has = several potential advantages over using the in- CO dorzolamide UFC;

– dividual components separately. Firstly, combining two or more medications in a single bottle simplifies the dosing regimen and twice daily; = latanoprost increases convenience, and could therefore po- prostaglandin analogue; = brinzolamide UFC; = – bid tentially increase both compliance and patient qual- investigator-blind, PG study; at least 8travoprost wk RI travoprost RI Methodology Results travoprost RI study; drugs WO bimatoprost RI

PGA ity of life. Similarly to chronic systemic diseases LDUFC that can be slowed or stabilized by a therapeutic travoprost 28 PRO, MC,25 R, investigator-blind, PG 89 PRO, SC, NC, SB study; bimatoprost PRO, RI MC, NC, SB study; 1 mo Mean 24 h IOP reduction was 4.7 = 192 PRO, SC, R,163 DB, PG study; 4 wk PRO, MC, R, DB, PG study; at least 4 wk Pts (n) drug followed over a long period and often for life, the effectiveness of glaucoma treatment de- dorzolamide UFC;

parallel group; pends largely on the patient’s adherence to the – =

TBrinUFC prescribed treatment. Many studies have estab- PG lished a relationship between poor adherence and brinzolamide UFC; 4 wk 32 PRO, double-centre, R, 12 wk 2 mo 2 mo assessment period – poor IOP control, increasing the risk of glaucoma bimatoprost [14,17-20] = development or progression. Moreover, open-label;

= many studies have found a relationship between latanoprost OP BDUFC brimonidine UFC; complex dosing regimens and poor patient ad- = – herence.[13,15,16,21] By reducing the number of bottles to be used and the number of drops to be travoprost

LBrinUFC administered, a fixed combination combining = two medications in a single bottle may increase

nonrandomized; [55,56] TTUFC vs TBrinUFC vs TBrimUFC TTUFC vs TBrinUFCTBrimUFC vs TBrinUFC 3 mo BBrimUFC 3 mo BDUFC BDUFC brimonidine UFC; = adherence, persistence and quality of life. In –

NR a study of a retail pharmacy claims database in TBrimUFC the US, three cohorts of glaucoma patients were [53] [54]

[51] defined according to the number of bottles (not bimatoprost [52]

brimonidine UFC; [55] = –

[50] medications) prescribed. Patients in cohort 1 [49] Contd (n = 14 742) were prescribed a single fixed com- et al. noncomparative; single-centre;

´ = = = bination, those in cohort 2 (n 18 411) received Hollo Feldman et al. Netland et al. Table I. Study Study arms Efficacy Reis et al. Stankiewicz et al. Stankiewicz et al. BBrimUFC NC latanoprost washout. SC two bottles (a b-blocker and another ocular hy-

potensive) and those in cohort 3 (n = 4826) were 3.2 Comparative Efficacy and Tolerability vs prescribed three bottles. The proportion of pa- Monotherapies tients completing the year without discontinuing was 35% in cohort 1, 27.2% in cohort 2 and 23.9% Currently, all marketed PGA-based fixed com- in cohort 3. As the number of separate products binations combine a PGA with a b-blocker: 0.005% used for glaucoma therapy increased, patient latanoprost with 0.5% timolol (LTFC, Pfizer), persistence decreased. In another study based on the 0.004% travoprost with 0.5% timolol (TTFC; a pharmacy claims database, Schwartz et al.[56] Alcon Inc.) and the 0.03% bimatoprost with compared the persistence of patients newly treated 0.5% timolol (BTFC, Allergan). Several studies with the fixed combination of dorzolamide and ti- have compared these three fixed combinations molol with patients newly treated with various un- with each of the component medications used as fixed combinations, including b-blockers–carbonic monotherapy.[29,30,64-77] The characteristics of the anhydrase inhibitors and PGA-based unfixed com- studies identified are summarized in table II. binations. Kaplan-Meier analysis showed that, over It should be noted that results sometimes con- a 12-month period, 62.2% of patients remained on flicted, with studies failing to demonstrate a signi- the fixed combination compared with 18.9–26.1% ficantly greater efficacy of the fixed combinations with the unfixed combinations. than the PGAs alone. For this reason, PGA– Secondly, fixed combinations also avoid a timolol fixed combinations were not approved by potential washout effect and avoid having to wait regulatory authorities in some countries, includ- between two separate instillations, thus poten- ing the US. tially increasing the efficiency of the treatment Meta-analyses have summarized the growing over separate component medications. Instilla- literature and increase the power so as to focus on tion of a reduced number of drops could also questions that individual studies cannot ad- likely make use of eyelid closure and nasolacrimal dress.[33,78] In a recent meta-analysis including occlusion more acceptable to patients, thus en- studies published up to the end of August 2010 hancing intraocular absorption and reducing and comparing the three PGA–b-blocker fixed systemic absorption of topically applied glauco- combinations with each of their component ma drugs.[57] medications used as monotherapy, we found that Thirdly, the reduced number of drops instilled the three PGA–timolol fixed combinations usu- reduces the exposure to preservatives such as ally provide greater IOP reduction than the three benzalkonium chloride, which is known to have PGAs or the 0.5% timolol alone, and a lower in- toxic effects on the ocular surface. In vitro, in vivo cidence of hyperaemia than the three PGAs animal and in vivo human studies have shown alone.[78] Depending on the measurement time that the effect of benzalkonium chloride on the point (9am, noon, 4pm or mean diurnal curve), ocular surface may be related to the concentra- the mean differences in IOP lowering were tion and total amount administered.[23,58-60] 1.31–2.01 mmHg (LTFC vs latanoprost; p < 0.01), Finally, the cost of fixed combinations is less 2.02–2.59 mmHg (TTFC vs travoprost; p < 0.001) than the combined cost of the component medi- and 0.00–1.14 mmHg (BTFC vs bimatoprost; cations in many countries.[61] Studies have shown p > 0.1 to p < 0.001). The incidence of hyperaemia that the direct costs (visits, surgeries, medica- was significantly less with LTFC and BTFC than tions, etc.) and indirect costs (visual impairment, with the individual PGAs (LTFC vs latanoprost social rehabilitation, etc.) of glaucoma increase relative risk [RR] = 0.66; p = 0.05; BTFC vs bi- with elevated IOP, disease progression, late matoprost RR = 0.61; p < 0.001). In another meta- stages of the disease, and poor adherence and analysis and systematic review conducted with persistence.[62,63] By increasing the patient’s studies published up to August 2009, Webers compliance and IOP control, fixed combinations et al.[33] found that switching from a 0.5% ti- could also potentially decrease the direct and in- molol twice-daily run-in phase to LTFC once direct costs of glaucoma. daily in the morning resulted in a pooled change

ª Trial Study arms Efficacy Pts Methodology Results 02Srne nentoa ulsigA.Alrgt eevd rg 02 2(10) 72 2012; Drugs reserved. rights All AG. Publishing International Springer 2012 assessment (n) period Hommer et al.[30] BTFC vs bimatoprost 3 wk 268 PRO, MC, R, DB, PG study; Mean diurnal IOP reduction was 35.7% with BTFC and 31.6% with tx-naive pts bimatoprost

Brandt et al.[64] BTFC vs bimatoprost 3 mo 1061 PRO, MC, R, DB, PG study; 4 d to Mean diurnal IOP reduction was 33.1%, 32.1% and 25.8% with vs timolol 6wkWO BTFC, bimatoprost and timolol

Konstas et al.[65] BTFC vs bimatoprost 3 mo 60 PRO, SC, R, observer-blind, CO; Mean 24 h IOP reduction was 27.8% with bimatoprost (35.3% 3- to 6-wk WO evening, 33.8% morning)

Palmberg et al.[66] LTFC vs latanoprost 3 mo 335 PRO, MC, R, DB, PG study; 5 d to Mean diurnal IOP reduction was 35.6% with LTFC, 33% with vs timolol 4wkWO latanoprost and 26.9% with timolol

Diestelhorst and LTFC vs latanoprost 1 mo 92 PRO, MC, R, DB, PG study; 1 wk Mean diurnal IOP reduction was 25.3% with LTFC, 19.4% with Almega˚rd[67] vs timolol timolol RI latanoprost and 8.5% with timolol

Higginbotham et al.[68] LTFC vs latanoprost 6 mo 418 PRO, MC, R, DB, PG study; 2–4wk Mean diurnal IOP change was 5.3% with timolol, 14.2% with vs timolol timolol RI latanoprost and 14.9% with LTFC

Pfeiffer et al.[69] LTFC vs latanoprost 6 mo 436 PRO, MC, R, DB, PG study; 2–4wk Mean diurnal IOP reduction was 12% with LTFC, 9.3% with vs timolol timolol RI latanoprost and 4.9% with timolol

Olander at al.[70] LTFC vs latanoprost 1 mo 350 PRO, MC, R, DB, PG study; 4 wk Mean diurnal IOP reduction was 16.5% with LTFC and 8.2% with latanoprost RI latanoprost

Konstas et al.[71] LTFC vs latanoprost 3 mo 37 PRO, SC, R, DB, CO study; 6 wk Mean 24 h IOP reduction was 31% with LTFC and 20.6% with WO latanoprost

Magacho et al.[72] LTFC vs latanoprost 1 mo 28 PRO, SC, R, observer-blind, PG 10am mean IOP reduction was 35.8% with LTFC and 25.6% with study; 30 d WO latanoprost

Konstas et al.[73] LTFC vs timolol 2 mo 33 R, SC, CO study Mean IOP change was 8.6 mmHg (34.4%) with LTFC and 5.7 mmHg (22.8%) with timolol

Konstas et al.[74] LTFC vs timolol 2 mo 29 PRO, SC, R, observer-blind, CO Mean 24 h IOP reduction was 34.2% with LTFC and 23.9% with study; 4–6wkWO timolol

Higginbotham et al.[75] LTFC vs latanoprost 3 mo 394 PRO, MC, R, DB, PG study; 5 d to Mean diurnal IOP reduction was 36.4% with LTFC, 31.6% with vs timolol 4wkWO latanoprost and 25.6% with timolol

Barnebey et al.[76] TTFC vs travoprost 3 mo 263 PRO, MC, R, DB, PG study; 5–28 d Mean diurnal IOP reduction was 35.6% with TTFC, 30.3% with vs timolol WO travoprost and 27.8% with timolol

Schuman et al.[29] TTFC vs timolol 3 mo 236 PRO, MC, R, DB, PG study; 5–28 d Mean diurnal IOP reduction was 31% with TTFC and 25.8% with WO timolol

Konstas et al.[77] TTFC vs travoprost 2 mo 34 PRO, SC, R, DB, CO study; 6 wk Mean 24 h IOP reduction was 35.3% with TTFC and 26.3% with al. et Aptel WO travoprost BTFC = bimatoprost–timolol fixed combination; CO = crossover; DB = double-blind; IOP = intraocular pressure; LTFC = latanoprost–timolol fixed combination; MC = multicentre; PG = parallel group; PRO = prospective; pts = patients; R = randomized; RI = run in; SC = single-centre; TTFC = travoprost–timolol fixed combination; tx = treatment; WO = washout. Prostaglandin-Based Combination Therapies in Glaucoma 1363

- % - - % % of 2.8 mmHg (95 CI 3.3, 2.3) relative % with % - % % - - with change 13.1 (95 CI 14.9, 11.2) in mean % IOP curve. % timolol fixed – 3.3 Comparative Efficacy and Tolerability vs Unfixed Combinations with TTFC and 34.6 with TTFC and 36.1 with LTFC and 35.7 with BTFC and 38 latanoprost % % % % 3.3.1 Results of Clinical Trials timolol fixed combination; with LTFC and 7.1 = – with TTFC and 32.5

A few studies have compared the PGA–b- % blocker fixed combinations with the corresponding LTFC [29,30,65,79-82] unfixed combinations. The character- travoprost = istics of the studies identified are summarized in

table III. Regarding the fixed and unfixed combi- TTFC nation of latanoprost and 0.5% timolol, a meta- analysis summarized several of the studies.[33] In run in; = intraocular pressure; = terms of efficacy, all studies showed equivalent or RI

greater IOP lowering with the unfixed combina- IOP with TTUFC with TTUFC with LTUFC with BTUFC Mean diurnal IOP reduction was 32.2 Mean diurnal IOP reduction was 0 % tions compared with the corresponding fixed com- LTUFC TTUFC binations. In agreement, the meta-analysis found randomized; that adding 0.005% latanoprost once daily in the = R % double-blind; evening to 0.5 timolol twice daily resulted in a = - % - - DB 4wk pooled change of 6.3 mmHg (95 CI 7.1, 5.5) 28 d WO28 d Mean WO diurnal IOP change was 31 Mean diurnal IOP reduction was 34.2 relative change of -28.3% (95% CI -32.8, -23.7) in mean IOP curve, whereas switching from a 0.5% prospective; =

timolol twice daily run-in phase to the fixed com- crossover; = bination of 0.5% timolol and 0.005% latanoprost PRO once daily in the morning resulted in a pooled CO change of -2.8 mmHg (95% CI -3.3, -2.3) relative change of -13.1% (95% CI -14.9, -11.2). [27,72] parallel group; timolol UFC; = – LTUFC RI Pooled data from two similarlystudies designed Methodology Results 3.3.2 Limitations of the Currently Available washout. = Clinical Trials PG WO It should be mentioned that 0.5% timolol was + bimatoprost 403 Pts (n) timolol fixed vs unfixed combinations

used only once daily with the fixed combinations, – = multicentre;

whereas it was used twice daily with the unfixed =

combinations in almost all the studies examined. MC Similarly, it should be mentioned that the PGA–b- BTUFC 3 mo3 mo 1903 mo PRO, 316 MC, R, DB, CO study; 2 – 3 mo PRO, 403 MC, R, DB, PG study; WO period3 mo PRO, 517 MC, R, DB, PG study; 5 – Mean1 mo diurnal IOP reduction was PRO, 316 33.3 MC, R, DB, PG study; 5 – 268 PRO, MC, R, DB, PG study Mean diurnal IOP reduction was 35.7 assessment period

blocker fixed combinations were given in the unfixed combination; morning in most studies, and thus the diurnal time- = timolol UFC; – points of IOP measurements did not include the UFC peak IOP reduction of the PG (about 12 hours after instillation). As the hypotensive effect of the LTFC vs LTUFC TTFC vs TTUFC TTFC vs TTUFC LTFC vs LTUFC TTFC vs TTUFC BTFC vs BTUFC

PGA is likely predominant compared with the latanoprost = timolol fixed combination; timolol UFC; – effect of the b-blocker (given once daily in the – [79] [81] fixed combination formulation), diurnal evalua- [29] [30] LTUFC [80]

– [82] tion of a PGA timolol fixed combination may Trials comparing the prostaglandin analogue travoprost bimatoprost lead to underestimation of the differences in IOP = = reduction ability compared with a PGA given in TTUFC combination; Hommer et al. Schuman et al. Gross et al. BTFC Trial Study arms Efficacy Table III. Hughes et al. Diestelhorst et al. the evening. Twenty-four-hour evaluations would Diestelhorst et al.

probably have led to a smaller difference between with with with with the IOP reduction ability of the fixed combina- with % % % % % tions and the unfixed combinations. A recently with LTFC multicentre; % published study has compared the morning and =

evening dosing of BTFC, showing that evening MC dosing gives rise to statistically better 24-hour IOP control than morning dosing.[65] In terms of with BTFC and 15.9 with BTFC and 13.7 with BTFC and 9.4 with BTFC and 13.7 with TTFC and 34.1 tolerability, most studies found a lower incidence with BTFC, 33.3 % % % % % of systemic and ocular treatment-related adverse % events with the fixed combinations than with the unfixed combinations. washout. There could be other reasons that most studies = timolol fixed combination; WO comparing fixed and unfixed combinations of a – PGA and timolol show that the unfixed combinations reduce IOP more. Thus, it could be pos- latanoprost with TTFC tulated that some of the potential advantages of = the fixed combination compared with using the % individual components separately may be under- LTFC LTFC LTFC LTFC LTFC LTFC Mean diurnal IOP reduction was 19.5 Mean diurnal IOP reduction was 21.4 Mean diurnal IOP reduction was 12.6 Mean diurnal IOP reduction was 21.4 Mean diurnal IOP reduction was 35.4 Mean diurnal IOP reduction was 31.8 and 25 estimated in clinical trials. Randomized controlled timolol fixed combination; trials, although recognized as the gold standard for – establishing efficacy, operate in an idealized en- 28 d vironment and could for many reasons limit the travoprost problem of poor compliance and persistence = intraocular pressure;

(patients know that they are involved in a clinical = trial, they often receive more information about TTFC IOP their disease, they are sometimes trained in drop instillation and informed about the risk of washout, they are often monitored more frequently, single-centre; = they sometimes receive financial compensation, double-blind; = etc.). In real-life practice, it is likely that the im- SC DB 6 wk timolol RI no WO period WO period 6 wk LTFC RI WO Methodology Results proved adherence and persistence due to the 1moWO timolol fixed combinations – run in; 54 PRO, SC, R, investigator-blind, CO study; 82 PRO, MC, R, investigator-blind, PG study; 36 PRO, SC, R, investigator-blind, study; no 91 PRO, MC, R, investigator-blind, CO study; simplified dosing regimen of the fixed combina- = Pts (n) tion could play a more important role. RI crossover; =

3.4 Direct Comparisons of the Three CO randomized; assessment period PGA–b-Blocker Fixed Combinations 12 mo 128 PRO, SC, R, investigator-blind, PG study; = R Very few studies have directly compared the efficacy and tolerability of the three PGA–b- blocker fixed combinations.[83-88] The character- prospective;

istics of the studies identified are summarized in =

table IV: three studies compared LTFC and BTFC vs LTFC 3 mo BTFC vs LTFC 3 mo BTFC vs LTFC 1 mo TTFC vs LTFC 12 moTTFC vs BTFC 3 mo 407TTFC vs BTFC vs PRO,LTFC MC, R, DB, MC, PG study; 5 – timolol fixed combination; PRO BTFC, one study compared LTFC and TTFC, – [84] [87] one study compared TTFC and BTFC, and one [86] [88]

recently published study compared the three fixed Trials comparing the three prostaglandin analogue [83] [85] combinations. bimatoprost = parallel group;

In a meta-analysis comparing the three PGA- = Centofanti et al. Martinez and Sanchez Rigollet et al. Topouzis et al. Centofanti et al. Table IV. Trial Study arms Efficacy Martinez and Sanchez based fixed combinations with each other over BTFC PG

3 months, and including the above-mentioned –

studies except the most recent, we found that timolol – with DTFC with DTFC with DTFC BTFC was generally slightly superior to either with DTFC % % % % % % latanoprost with DTFC

[78] with DTFC

LTFC or TTFC. BTFC lowered IOP sig- = % %

nificantly more than did LTFC or TTFC over the travoprost = diurnal curve (-1.24 and -0.66 mmHg, respect- LTFC ively), at all timepoints during the day versus TTFC TTFC and at most timepoints during the day with LTFC and 15.8 with BTFC and 35.4 with LTFC and 38.9 with LTFC and 26.1 with LTFC and 34.8 with TTFC and 34.2 % % with LTFC and 9.9 % % % % with LTFC and 5.8 versus LTFC. Comparing LTFC with TTFC, run in; % = there was no clear result: TTFC was slightly % RI better at some times of the day, LTFC at others, intraocular pressure; and there was no significant difference over the =

diurnal curve. The incidence of hyperaemia was IOP slightly – and nonsignificantly – higher with randomized; =

BTFC than with LTFC. TTFC was associated R with a higher risk of hyperaemia than LTFC (RR for TTFC vs LFTC: 5.99 [95% CI 2.38, 15.10]). Mean diurnal IOP reduction was 9.4 mmHg with LTFC and 8.4 mmHg with DTFC Mean diurnal IOP reduction was 26.1 Mean 24 h IOP reduction was 11.8 Results On this point, only one clinical trial provided with DTFC with DTFC prospective; valid data for this comparison, and the results = [89] timolol fixed combination; should probably be interpreted with caution. PRO An objective assessment of conjunctival hyper- – 28 d WO Mean diurnal IOP reduction was 36.5 aemia was also performed by the investigator in – this double-blind study, by grading conjunctival dorzolamide = hyperaemia based on a scale ranging from 0 to 3. parallel group; =

Differences in average hyperaemia scores were 28 d WO Mean diurnal IOP reduction was 36.9 – PG always small and not statistically significant at DTFC eight of the ten study visits. open-label; =

3.5 Comparative Efficacy and Tolerability with double-blind; = Other b-Blocker-Based Fixed Combinations OP DB timolol fixed combinations with other timolol-based fixed combinations Studies have compared the efficacy and toler- – no WO period no WO period ability of the three PGA–b-blocker fixed combi- Methodology nations with that of the dorzolamide–timolol crossover; 33 PRO, double-centre, R, DB, CO study; 432 wk timolol RI PRO, single-centre, R, investigator-blind, CO study; 31 Mean diurnal IOP reduction was PRO, 13.9 single-centre, R, investigator-blind, CO study; 32 PRO, single-centre, R, DB, CO study; 6 wk timolol RI Mean48 diurnal IOP change was 8.7 PRO, single-centre, R, OP, PG study; WO period Mean diurnal IOP reduction was 45.8 = 178 PRO, single-centre, NR, OP, PG study; no WO period Mean diurnal IOP reduction was 34.1 Pts (n)

[90-98] nonrandomized;

fixed combination (DTFC). The character- = CO

istics of the studies identified are summarized in NR table V. Of the seven studies comparing LTFC and DTFC, three found greater IOP lowering assessment period with LTFC, two found equivalent IOP lowering multicentre; with LTFC and DTFC, and one found greater =

IOP lowering with DTFC. In most cases, the MC washout. differences in IOP reduction were modest. One = timolol fixed combination; – WO LTFC vs DTFC 3 moLTFC vs DTFC 2 mo LTFC vs DTFC 253 1 PRO, mo MC,LTFC R, vs investigator-blind, DTFC PG study 3 mo LTFC vs DTFCLTFC vs DTFC 6 wk LTFC vs DTFC 4 y TTFC vs DTFC 3 moBTFC vs DTFC 6 wk 8 wk 270 PRO, MC, R, 319 investigator-blind, PG study; 5 PRO, MC, R, DB, PG study; 5 study compared TTFC and DTFC, and found Study arms Efficacy TTFC to be more effective. Similarly, one study [91] [93] [94] compared BTFC and DTFC, and found greater [96] Trials comparing the three prostaglandin analogue [97] [95] [98] [90] [92] efficacy for BTFC. In terms of tolerability, the bimatoprost = rate of systemic and ocular treatment-related Konstas et al. Jothi et al. Cvenkel et al. Pajic et al. Table V. Trial BTFC Shin et al. Miglior et al. Martinez and Sanchez Konstas et al. timolol fixed combination; Teus et al. adverse events were usually comparable with fixed combination;

the PGA–b-blocker fixed combinations and the treated patients. The timepoint of treatment ad- DTFC. ministration (particularly morning or evening dosing of fixed combinations) and the timepoint 4. Discussion and Conclusions of IOP measurements (diurnal or 24-hour IOP measurements) also likely influenced the results We reviewed the published studies evaluating as the methodology of some studies may make it the efficacy and tolerability of the IOP-lowering impossible to identify the peak and trough effects unfixed and fixed combinations with PGAs. Re- of some medications. We were careful to mention garding unfixed combinations, a2-adrenergic agonist the methodological characteristics and the schedule – and carbonic anhydrase inhibitor–PGA unfixed of IOP measurements of the studies identified in combinations seem to be at least as effective in the tables summarizing the study characteristics. reducing IOP and provide a similar rate of side It should also be mentioned that 24-hour studies effects to the b-blocker–PGA unfixed combina- actually often include only a limited number of tions.[34-54] Regarding the fixed combinations, the night-time measurements, at intervals of 2, 3 or three PGA–timolol fixed combinations are more even 4 hours. Moreover, IOP is measured in the effective than their component medications used sitting position and therefore nocturnal IOP is separately as monotherapy and are better toler- likely underestimated. In the future, continuous ated than the three respective PGAs.[29,30,64-77] noninvasive monitoring of IOP will make it pos- The three PGA–timolol fixed combinations re- sible to better evaluate and compare the circadian duce IOP to a lesser degree than the correspond- effect of glaucoma medications. ing unfixed combinations.[29,30,65,79-82] Because of Clinical trials and a recent meta-analysis show the simplified dosing regimen, reduced exposure that PGA-based fixed combinations provide greater to preservative, reduced rates of side effects and IOP reduction and a lower incidence of hyper- lower medication costs, fixed combinations may aemia than the three respective prostaglandin provide improved adherence and persistence monotherapies. PGA–timolol fixed combina- compared with concomitant use of individual tions may therefore be considered as second-line components. However, limited data are currently therapy in patients who have not attained target available on these points.[55-57] Most studies, but IOP with PGA or timolol monotherapy. In some not all, found that the three PGA–timolol fixed studies, the differences in IOP reduction may seem combinations were more effective than DTFCs at modest. From pharmacokinetic/pharmacodynamic reducing IOP and were similarly tolerated.[90-98] studies, it is usually considered that the peak ac- It should be mentioned that the ability to re- tion of an topically applied ocular b-blocker is duce IOP expressed as relative or absolute IOP achieved 2 hours after administration and the reduction from baseline varied, sometimes con- trough action 12 hours after administration.[99] siderably, between studies testing the same med- Similarly, it is considered that the peak action of ications. This may be due, at least in part, to a topically applied ocular PGA is achieved 12 hours methodological differences in the studies pub- after administration and the trough action 24 hours lished, particularly in terms of inclusion criteria after administration.[99] It should be mentioned and washout period and/or the medication run-in that PGA–b-blocker combinations are usually given period, which varied greatly among studies. In in the morning in the included studies, and thus terms of diagnosis, some studies included a sig- the diurnal timepoints of IOP measurements nificant proportion of ocular hypertension usually did not include the peak IOP reduction of patients (up to >40% of patients in some studies) the PGA (about 8pm to 10pm depending on in- or patients with pseudoexfoliation glaucoma and/ stillation time). Twenty-four-hour IOP measure- or pigment dispersion glaucoma (up to >50% of ments would likely have resulted in a greater patients in some studies). In terms of medica- difference in IOP reduction ability between the tions, some studies included only treatment-naive PGA given in the evening and the PGA–b-blocker patients, whereas others included previously fixed combination given in the morning.

Studies comparing fixed and unfixed com- in the treatment of glaucoma and ocular binations of a PGA and timolol show that the hypertension. unfixed combinations reduce IOP to a greater More than one-third of patients with extent. Similarly, the design of the studies could glaucoma require more than one medication have disadvantaged the fixed combination. Fixed to control the disease. combinations were usually given in the morning; Fixed combinations could have the potential therefore diurnal IOP measurement timepoints to simplify the dosing regimen and decrease did not include the peak action of the PGA. By side effects due to drugs and preservatives, contrast, when using unfixed combinations, the thereby increasing patient adherence to and PGA was given in the evening, resulting in a persistence with the prescribed therapy. maximal effect occurring within the diurnal hours Fixed combinations avoid the washout effect of IOP measurement. Actually, it is likely that the and decrease treatment costs in many countries. improved adherence due to the better tolerability The exact value of the fixed combinations on of the fixed combinations may lead to a smaller patient adherence and persistence remains to difference between the efficacy of the fixed and be determined in further studies. unfixed combinations. Further studies could more The three PGA–timolol fixed combinations accurately evaluate the benefits of the fixed com- are more effective than their component binations in terms of patient convenience, adher- medications used separately as monotherapy ence and related IOP control, as little evidence is and are better tolerated than the three respec- currently available. Similarly, it would be inter- tive prostaglandins. esting to evaluate the impact of the reduced The three PGA–timolol fixed combinations are preservative exposure provided by the fixed com- less effective in reducing IOP than the unfixed binations (usually a reduction by a factor of three combinations, but are better tolerated. of the total amount of benzalkonium chloride instilled when a PGA associated with 0.5% ti- Acknowledgements molol twice daily is switched to the corresponding fixed combination) on the conjunctiva and No sources of funding were used to conduct this study or the filtering surgery bleb failure rate. prepare this manuscript. Currently, all marketed PGA-based fixed com- Florent Aptel and Christophe Chiquet have received sup- binations combine a PGA with a b-blocker. port for travel from Alcon, Allergan, Pfizer. b-blockers have numerous systemic side effects, References particularly bronchospasm mediated by b2 receptors, 1. Weinreb RN, Khaw PT. Primary open-angle glaucoma. which may limit their use in patients with chronic Lancet 2004 May; 363 (9422): 1711-20 obstructive pulmonary disease – a rather com- 2. Quigley HA. Number of people with glaucoma worldwide. mon condition in patients with glaucoma since Br J Ophthalmol 1996; 80: 389-93 – 3. Kass MA, Heuer DK, Higginbotham EJ, et al. The Ocular they are often elderly and in patients with Hypertension Treatment Study: a randomized trial de- asthma. In diabetic patients, b-blockers can also pre- termines that topical ocular hypotensive medication delays vent tachycardia, which serves as a warning sign for or prevents the onset of primary open-angle glaucoma. insulin-induced hypoglycaemia, and should there- Arch Ophthalmol 2002 Jun; 120 (6): 701-13 4. Heijl A, Leske MC, Bengtsson B, et al. Reduction of in- fore be used with caution. Systemic side effects of traocular pressure and glaucoma progression: results from b-blockers are a major reason why it would be useful the Early Manifest Glaucoma Trial. Arch Ophthalmol to have fixed combinations of a PGA and an a-adre- 2002 Oct; 120 (10): 1268-79 nergic agonist or carbonic anhydrase inhibitor. 5. Collaborative Normal-Tension Glaucoma Study Group. Comparison of glaucomatous progression between un- treated patients with normal-tension glaucoma and 5. Key Practical Points for Clinicians patients with therapeutically reduced intraocular pressures. Collaborative Normal-Tension Glaucoma Study Group. Am J Ophthalmol 1998 Oct; 126 (4): 487-97 Prostaglandin analogues and b-blockers are 6. Advanced Glaucoma Intervention Study. The Advanced the most commonly used therapeutic classes Glaucoma Intervention Study (AGIS): the relationship

between control of intraocular pressure and visual field 22. Djafari F, Lesk MR, Harasymowycz PJ, et al. Determinants deterioration. The AGIS Investigators. Am J Ophthalmol of adherence to glaucoma medical therapy in a long-term 2000 Oct; 130 (4): 429-40 patient population. J Glaucoma 2009 Mar; 18 (3): 238-43 7. Van der Valk R, Webers CA, Schouten JS, et al. Intraocular 23. Baudouin C, Labbe´A, Liang H, et al. Preservatives in eye- pressure-lowering effects of all commonly used glaucoma drops: the good, the bad and the ugly. Prog Retin Eye Res drugs: a meta-analysis of randomized clinical trials. Oph- 2010 Jul; 29 (4): 312-34 thalmology 2005 Jul; 112 (7): 1177-85 24. Egorov E, Ropo A, Investigators. Adjunctive use of tafluprost 8. Stewart WC, Konstas AG, Nelson LA, et al. Meta-analysis with timolol provides additive effects for reduction of in- of 24-hour intraocular pressure studies evaluating the effi- traocular pressure in patients with glaucoma. Eur J Oph- cacy of glaucoma medicines. Ophthalmology 2008 Jul; 115 thalmol 2009 Mar-Apr; 19 (2): 214-22 (7): 1117-22 25. Bron AM, Denis P, Nordmann JP, et al. Additive IOP- 9. Van der Valk R, Webers CA, Lumley T, et al. A network reducing effect of latanoprost in patients insufficiently meta-analysis combined direct and indirect comparisons controlled on timolol. Acta Ophthalmol Scand 2001 Jun; between glaucoma drugs to rank effectiveness in lowering 79 (3): 289-93 intraocular pressure. J Clin Epidemiol 2009 Dec; 62 (12): 26. Bucci MG, Italian Latanoprost Study Group. Intraocular 1279-83 pressure-lowering effects of latanoprost monotherapy 10. European Glaucoma Society. Terminology and guidelines versus latanoprost or pilocarpine in combination with ti- for glaucoma. 3rd ed. Savona: Editrice Dogma, 2008 molol: a randomized, observer-masked multicenter study in patients with open-angle glaucoma. J Glaucoma 1999 11. American Academy of Ophthalmology Glaucoma Panel. Preferred Practice Pattern Guidelines. Primary open- Feb; 8 (1): 24-30 angle glaucoma. San Francisco (CA): American Academy 27. Diestelhorst M, German Latanoprost Study Group. The of Ophthalmology, 2010 [online]. Available from URL: additive intraocular pressure-lowering effect of latanoprost http://www.aao.org/ppp [Accessed 2012 May 31] 0.005% daily once and pilocarpine 2% t.i.d. in patients with open-angle glaucoma or ocular hypertension: a 6-month, 12. Feiner L, Piltz-Seymour JR, Collaborative Initial Glaucoma randomized, multicenter study. Graefes Arch Clin Exp Treatment Study. Collaborative Initial Glaucoma Treat- Ophthalmol 2000 May; 238 (5): 433-9 ment Study: a summary of results to date. Curr Opin Ophthalmol 2003 Apr; 14 (2): 106-11 28. Petounis A, Mylopoulos N, Kandarakis A, et al. Compar- ison of the additive intraocular pressure-lowering effect of 13. Stewart WC, Konstas AG, Pfeiffer N. Patient and ophthal- latanoprost and dorzolamide when added to timolol in mologist attitudes concerning compliance and dosing in patients with openangle glaucoma or ocular hypertension: glaucoma treatment. J Ocul Pharmacol Ther 2004 Dec; 20 a randomized, openlabel, multicenter study in Greece. (6): 461-9 J Glaucoma 2001 Aug; 10 (4): 316-24 14. Schwartz GF, Quigley HA. Adherence and persistence with 29. Schuman JS, Katz GJ, Lewis RA, et al. Efficacy and safety glaucoma therapy. Surv Ophthalmol 2008 Nov; 53 Suppl. of a fixed combination of travoprost 0.004%/timolol 0.5% 1: S57-68 ophthalmic solution once daily for open-angle glaucoma or 15. Robin AL, Covert D. Does adjunctive glaucoma therapy ocular hypertension. Am J Ophthalmol 2005 Aug; 140 (2): affect adherence to the initial primary therapy? Ophthal- 242-50 mology 2005 May; 112 (5): 863-8 30. Hommer A, Ganfort Investigators Group I. A double- 16. Robin AL, Novack GD, Covert DW, et al. Adherence in masked, randomized, parallel comparison of a fixed combi- glaucoma: objective measurements of once-daily and ad- nation of bimatoprost 0.03%/timolol 0.5% with non-fixed junctive medication use. Am J Ophthalmol 2007 Oct; 144 combination use in patients with glaucoma or ocular hy- (4): 533-40 pertension. Eur J Ophthalmol 2007 Jan-Feb; 17 (1): 53-62 17. Olthoff CM, Schouten JS, van de Borne BW, et al. Non- 31. Higginbotham EJ, Diestelhorst M, Pfeiffer N, et al. The compliance with ocular hypotensive treatment in patients efficacy and safety of unfixed and fixed combinations of with glaucoma or ocular hypertension an evidence-based latanoprost and other antiglaucoma medications. Surv review. Ophthalmology 2005 Jun; 112 (6): 953-61 Ophthalmol 2002 Aug; 47: 133-40 18. Okeke CO, Quigley HA, Jampel HD, et al. Adherence with 32. Webers CA, van der Valk R, Schouten JS, et al. Intraocular topical glaucoma medication monitored electronically: the pressure-lowering effect of adding dorzolamide or latano- Travatan Dosing Aid study. Ophthalmology 2009 Feb; 116 prost to timolol: a meta-analysis of randomized clinical (2): 191-9 trials. Ophthalmology 2007 Jan; 114 (1): 40-6 19. Patel SC, Spaeth GL. Compliance in patients prescribed 33. Webers CA, Beckers HJ, Zeegers MP, et al. The intraocular eyedrops for glaucoma. Ophthalmic Surg 1995 May-Jun; pressure-lowering effect of prostaglandin analogs combin- 26 (3): 233-6 ed with topical b-blocker therapy: a systematic review 20. Higginbotham EJ. Considerations in glaucoma therapy: and meta-analysis. Ophthalmology 2010 Nov; 117 (11): fixed combinations versus their component medications. 2067-74 Clin Ophthalmolol 2010 Feb; 4: 1-9 34. O’Connor DJ, Martone JF, Mead A. Additive intraocular 21. Taylor SA, Galbraith SM, Mills RP. Causes of non- pressure lowering effect of various medications with lata- compliance with drug regimens in glaucoma patients: a noprost. Am J Ophthalmol 2002 Jun; 133 (6): 836-7 qualitative study. J Ocul Pharmacol Ther 2002 Oct; 18 (5): 35. Erdogan H, Toker MI, Arici MK, et al. Effect of latanoprost 401-9 0.005% and brimonidine 0.2% on intraocular pressure after

phacoemulsification and intraocular lens implantation 50. Hollo´G, Chiselita D, Petkova N, et al. The efficacy and surgery. Jpn J Ophthalmol 2004 Nov-Dec; 48 (6): 600-1 safety of timolol maleate versus brinzolamide each given 36. Konstas AG, Karabatsas CH, Lallos N, et al. 24-hour intra- twice daily added to travoprost in patients with ocular hy- ocular pressures with brimonidine purite versus dorzola- pertension or primary open-angle glaucoma. Eur J Oph- mide added to latanoprost in primary open-angle glaucoma thalmol 2006 Nov-Dec; 16 (6): 816-23 subjects. Ophthalmology 2005 Apr; 112 (4): 603-8 51. Feldman RM, Tanna AP, Gross RL, et al., Additivity Study 37. Tsukamoto H, Noma H, Mukai S, et al. The efficacy and Group. Comparison of the ocular hypotensive efficacy of ocular discomfort of substituting brinzolamide for dorzo- adjunctive brimonidine 0.15% or brinzolamide 1% in lamide in combination therapy with latanoprost, timolol, combination with travoprost 0.004%. Ophthalmology 2007 and dorzolamide. J Ocul Pharmacol Ther 2005 Oct; 21 (5): Jul; 114 (7): 1248-54 395-9 52. Netland PA, Michael M, Rosner SA, et al. Brimonidine 38. Shoji N, Ogata H, Suyama H, et al. Intraocular pressure purite and bimatoprost compared with timolol and lata- lowering effect of brinzolamide 1.0% as adjunctive therapy noprost in patients with glaucoma and ocular hyperten- to latanoprost 0.005% in patients with open angle glauco- sion. Adv Ther 2003 Jan-Feb; 20 (1): 20-30 ma or ocular hypertension: an uncontrolled, open-label 53. Stankiewicz A, Wierzbowska J, Siemiatkowska A, et al. The study. Curr Med Res Opin 2005 Apr; 21 (4): 503-8 additive effect of dorzolamide hydrochloride (Trusopt) and 39. Mundorf T, Noecker RJ, Earl M. Ocular hypotensive effi- a morning dose of bimatoprost (Lumigan) on intraocular cacy of brimonidine 0.15% as adjunctive therapy with la- pressure and retrobulbar blood flow in patients with pri- tanoprost 0.005% in patients with open-angle glaucoma or mary open-angle glaucoma. Br J Ophthalmol 2010 Oct; 94 ocular hypertension. Adv Ther 2007 Mar-Apr; 24 (2): 302-9 (10): 1307-11 40. Nakamoto K, Yasuda N. Effect of concomitant use of la- 54. Stankiewicz A, Misiuk-Hojło M, Grabska-Liberek I, et al. tanoprost and brinzolamide on 24-hour variation of IOP in Intraocular pressure and ocular hemodynamics in patients normal-tension glaucoma. J Glaucoma 2007 Jun-Jul; 16 with primary open-angle glaucoma treated with the com- (4): 352-7 bination of morning dosing of bimatoprost and dorzola- 41. Tamer C, Oksu¨z H. Circadian intraocular pressure control mide hydrochloride. Acta Ophthalmol 2011 Feb; 89 (1): with dorzolamide versus timolol maleate add-on treat- e57-63 ments in primary open-angle glaucoma patients using la- 55. Higginbotham EJ, Hansen J, Davis EJ, et al. Glaucoma tanoprost. Ophthalmic Res 2007; 39 (1): 24-31 medication persistence with a fixed combination versus 42. Day DG, Hollander DA. Brimonidine purite 0.1% versus multiple bottles. Curr Med Res Opin 2009 Oct; 25 (10): brinzolamide 1% as adjunctive therapy to latanoprost in 2543-7 patients with glaucoma or ocular hypertension. Curr Med 56. Schwartz GF, Pate VD, Burk C, et al. Patient persistency Res Opin 2008 May; 24 (5): 1435-42 with topical glaucoma therapy is higher with fixed dose 43. Ishikawa M, Yoshitomi T. Effects of brinzolamide vs timo- brimonidine/timolol than with two-bottle combinations lol as an adjunctive medication to latanoprost on circadian [abstract no. 185; plus poster A408]. Association for Re- intraocular pressure control in primary open-angle glau- search in Vision and Ophthalmology; 2010 May 2-6; Fort coma Japanese patients. Clin Ophthalmol 2009; 3: 493-500 Lauderdale (FL) 44. Liu JH, Medeiros FA, Slight JR, et al. Comparing diurnal 57. Flach AJ. The importance of eyelid closure and nasolacrimal and nocturnal effects of brinzolamide and timolol on in- occlusion following the ocular instillation of topical glau- traocular pressure in patients receiving latanoprost mono- coma medications, and the need for the universal inclu- therapy. Ophthalmology 2009 Mar; 116 (3): 449-54 sion of one of these techniques in all patient treatments 45. Nakamura Y, Ishikawa S, Nakamura Y, et al. 24-hour in- and clinical studies. Trans Am Ophthalmol Soc 2008; 106: traocular pressure in glaucoma patients randomized to re- 138-45 ceive dorzolamide or brinzolamide in combination with 58. Rolando M, Brezzo V, Giordano G, et al. The effect of dif- latanoprost. Clin Ophthalmol 2009;3: 395-400 ferent benzalkonium chloride concentrations on human 46. Inoue K, Masumoto M, Wakakura M, et al. Ocular hypo- normal ocular surface. In: Van Bijsterveld O, Lemp M, tensive effect, preservation of visual fields, and safety of Spinelli D, editors. The lacrimal system. Amsterdam, Ber- adding dorzolamide to prostaglandin therapy for twelve kely, Milano: Kugler Publications, 1991 months. Clin Ophthalmol 2011; 5: 393-6 59. Nuzzi R, Finazzo C, Cerruti A. Adverse effects of topical 47. Li Yim JF, Montgomery DM. The intraocular pressure re- antiglaucomatous medications on the conjunctiva and ducing effect of brinzolamide as adjunctive therapy to la- the lachrymal (Brit. Engl) response. Int Ophthalmol 1998; tanoprost. Curr Med Res Opin 2011 Aug; 27 (8): 1499-502 22 (1): 31-5 48. Bournias TE, Lai J. Brimonidine tartrate 0.15%, dorzola- 60. Guenoun JM, Baudouin C, Rat P, et al. In vitro comparison mide hydrochloride 2%, and brinzolamide 1% compared as of cytoprotective and antioxidative effects of latanoprost, adjunctive therapy to prostaglandin analogs. Ophthal- travoprost, and bimatoprost on conjunctiva-derived epi- mology 2009 Sep; 116 (9): 1719-24 thelial cells. Invest Ophthalmol Vis Sci 2005 Dec; 46 (12): 4594-9 49. Reis R, Queiroz CF, Santos LC, et al. A randomized, investigator-masked, 4-week study comparing timolol male- 61. Rylander NR, Vold SD. Cost analysis of glaucoma medi- ate 0.5%, brinzolamide 1%, and brimonidine tartrate 0.2% cations. Am J Ophthalmol 2008 Jan; 145 (1): 106-13 as adjunctive therapies to travoprost 0.004% in adults with 62. Oostenbrink JB, Rutten-van Mo¨lken MP, Sluyter- primary open-angle glaucoma or ocular hypertension. Clin Opdenoordt TS. Resource use and costs of patients with Ther 2006 Apr; 28 (4): 552-9 glaucoma or ocular hypertension: a one-year study based

on retrospective chart review in the Netherlands. J Glau- for primary open-angle glaucoma or ocular hypertension: a coma 2001 Jun; 10 (3): 184-91 randomized, double-masked study. Arch Ophthalmol 2010 63. Lee PP, Walt JG, Doyle JJ, et al. A multicenter, retro- Feb; 128 (12): 165-72 spective pilot study of resource use and costs associated 76. Barnebey HS, Orengo-Nania S, Flowers BE, et al. The safety with severity of disease in glaucoma. Arch Ophthalmol and efficacy of travoprost 0.004%/timolol 0.5% fixed 2006 Jan; 124 (1): 12-9 combination ophthalmic solution. Am J Ophthalmol 2005 64. Brandt JD, Cantor LB, Katz LJ, et al., Ganfort In- Jul; 140 (1): 1-7 vestigators Group II. Bimatoprost/timolol fixed combina- 77. Konstas AG, Mikropoulos D, Haidich AB, et al. Twenty- tion: a 3-month double-masked, randomized parallel four-hour intraocular pressure control with the travo- comparison to its individual components in patients with prost/timolol maleate fixed combination compared with glaucoma or ocular hypertension. J Glaucoma 2008 Apr- travoprost when both are dosed in the evening in primary May; 17 (3): 211-6 open-angle glaucoma. Br J Ophthalmol 2009 Apr; 93 (4): 65. Konstas AG, Hollo´G, Mikropoulos D, et al. Twenty-four- 481-5 hour intraocular pressure control with bimatoprost and the 78. Aptel F, Cucherat M, Denis P. Efficacy and tolerability of bimatoprost/timolol fixed combination administered in the prostaglandin-timolol fixed combinations: a meta-analysis morning, or evening in exfoliative glaucoma. Br J Oph- of randomized clinical trials. Eur J Ophthalmol 2012 Jan- thalmol 2010 Feb; 94 (2): 209-13 Feb; 22 (1): 5-18 66. Palmberg P, Kim EE, Kwok KK, et al., for the Canada and 79. Diestelhorst M, Larsson LI, European Latanoprost Fixed United States Fixed combination Latanoprost/Timolol Combination Study Group. A 12 week study comparing Study Gro. A 12-week, randomized, double-masked study the fixed combination of latanoprost and timolol with the of fixed combination latanoprost/timolol versus latano- concomitant use of the individual components in patients prost or timolol monotherapy. Eur J Ophthalmol 2010 Jul- with open angle glaucoma and ocular hypertension. Br J Aug; 20 (4): 708-18 Ophthalmol 2004 Feb; 88 (2): 199-203 67. Diestelhorst M, Almega˚rd B. Comparison of two fixed 80. Hughes BA, Bacharach J, Craven ER, et al. A three-month, combinations of latanoprost and timolol in open-angle multicenter, double-masked study of the safety and efficacy glaucoma. Graefes Arch Clin Exp Ophthalmol 1998 Aug; of travoprost 0.004%/ timolol 0.5% ophthalmic solution 236 (8): 577-81 compared to travoprost 0.004% ophthalmic solution and 68. Higginbotham EJ, Feldman R, Stiles M, et al., Fixed Com- timolol 0.5% dosed concomitantly in subjects with open bination Investigative Group. Latanoprost and timolol angle glaucoma or ocular hypertension. J Glaucoma 2005 combination therapy vs monotherapy: one-year random- Oct; 14 (5): 392-9 ized trial. Arch Ophthalmol 2002 Jul; 120 (7): 915-22 81. Diestelhorst M, Larsson LI, European-Canadian Latano- 69. Pfeiffer N, European Latanoprost Fixed Combination prost Fixed Combination Study Group. A 12-week, ran- Study Group. A comparison of the fixed combination of domized, double-masked, multicenter study of the fixed latanoprost and timolol with its individual components. combination of latanoprost and timolol in the evening Graefes Arch Clin Exp Ophthalmol 2002 Nov; 240 (11): versus the individual components. Ophthalmology 2006 893-9 Jan; 113 (1): 70-6 70. Olander K, Zimmerman TJ, Downes N, et al., Xalacom/ 82. Gross RL, Sullivan EK, Wells DT, et al. Pooled results of Latanoprost Study Group. Switching from latanoprost to two randomized clinical trials comparing the efficacy and fixed-combination latanoprost-timolol: a 21-day, ran- safety of travoprost 0.004%/timolol 0.5% in fixed combi- domized, double-masked, active-control study in patients nation versus concomitant travoprost 0.004% and timolol with glaucoma and ocular hypertension. Clin Ther 2004 0.5%. Clin Ophthalmol 2007 Sep; 1 (3): 317-22 Oct; 26 (10): 1619-29 83. Martinez A, Sanchez M. Bimatoprost/timolol fixed combi- 71. Konstas AG, Boboridis K, Tzetzi D, et al. Twenty-four- nation vs latanoprost/timolol fixed combination in open- hour control with latanoprost-timolol-fixed combination angle glaucoma patients. Eye 2009 Apr; 23 (4): 810-8 therapy vs latanoprost therapy. Arch Ophthalmol 2005 Jul; 84. Centofanti M, Oddone F, Vetrugno M, et al. Efficacy of the 123 (7): 898-902 fixed combinations of bimatoprost or latanoprost plus ti- 72. Magacho L, Reis R, Shetty RK, et al. Efficacy of latano- molol in patients uncontrolled with prostaglandin mono- prost or fixed-combination latanoprost-timolol in patients therapy: a multicenter, randomized, investigator-masked, switched from a combination of timolol and a non- clinical study. Eur J Ophthalmol 2009 Jan-Feb; 19 (1): prostaglandin medication. Ophthalmology 2006 Mar; 113 66-71 (3): 442-5 85. Martinez A, Sanchez M. A comparison of the safety and 73. Konstas AG, Lake S, Economou AI, et al. 24-Hour control intraocular pressure lowering of bimatoprost/timolol fixed with a latanoprost-timolol fixed combination vs timolol combination versus latanoprost/timolol fixed combination alone. Arch Ophthalmol 2006 Nov; 124 (11): 1553-7 in patients with open-angle glaucoma. Curr Med Res Opin 74. Konstas AG, Pikilidou MI, Tsironi S, et al. 24-hour in- 2007 May; 23 (5): 1025-32 traocular pressure and blood pressure levels with latano- 86. Topouzis F, Melamed S, Danesh-Meyer H, et al. A 1-year prost/timolol fixed combination versus timolol. Curr Eye study to compare the efficacy and safety of once-daily Res 2009 May; 34 (5): 369-77 travoprost 0.004%/timolol 0.5% to once-daily latanoprost 75. Higginbotham EJ, Olander KW, Kim EE, et al., United 0.005%/timolol 0.5% in patients with open-angle glaucoma States Fixed-Combination Study Group. Fixed combina- or ocular hypertension. Eur J Ophthalmol 2007 Mar-Apr; tion of latanoprost and timolol vs individual components 17 (2): 183-90

87. Centofanti M, Oddone F, Gandolfi S, et al. Comparing ma or ocular hypertension. Curr Eye Res 2008 Feb; 33 (2): prostaglandin analogues plus timolol fixed combinations 163-8 for the treatment of open angle glaucoma: the Glaucoma 95. Pajic B, Pajic-Eggspuehler B, Ha¨fliger IO. Comparison of Randomized European Assessment Trial. Am J Ophthal- the effects of dorzolamide/timolol and latanoprost/timolol mol 2010 Oct; 150 (4): 575-80 fixed combinations upon intraocular pressure and pro- 88. Rigollet JP, Ondategui JA, Pasto A, et al. Randomized trial gression of visual field damage in primary open-angle comparing three fixed combinations of prostaglandins/pros- glaucoma. Curr Med Res Opin 2010 Sep; 26 (9): 2213-9 tamide with timolol maleate. Clin Ophthalmol 2011; 5: 187-91 96. Miglior S, Grunden JW, Kwok K, Xalacom/Cosopt 89. Aptel F, Denis P. Balancing efficacy and tolerability of European Study Group. Efficacy and safety of fixed com- prostaglandin analogues and prostaglandin-timolol fixed binations of latanoprost/timolol and dorzolamide/timolol combinations in primary open-angle glaucoma. Curr Med in open-angle glaucoma or ocular hypertension. Eye 2010 Res Opin 2011 Oct; 27 (10): 1949-58 Jul; 24 (7): 1234-42 90. Shin DH, Feldman RM, Sheu WP, Fixed Combination 97. Teus MA, Miglior S, Laganovska G, et al. Efficacy and Latanoprost/Timolol Study Group. Efficacy and safety of safety of travoprost/timolol vs dorzolamide/timolol in the fixed combinations latanoprost/timolol versus dorzo- patients with open-angle glaucoma or ocular hypertension. lamide/timolol in patients with elevated intraocular pres- Clin Ophthalmol 2009; 3: 629-36 sure. Ophthalmology 2004 Feb; 111 (2): 276-82 98. Jothi R, Ismail AM, Senthamarai R, et al. A comparative 91. Konstas AG, Kozobolis VP, Lallos N, et al. Daytime diur- study on the efficacy, safety, and cost-effectiveness of bi- nal curve comparison between the fixed combinations of matoprost/timolol and dorzolamide/timolol combinations latanoprost 0.005%/timolol maleate 0.5% and dorzolamide in glaucoma patients. Indian J Pharmacol 2010 Dec; 42 (6): 2%/timolol maleate 0.5%. Eye 2004 Dec; 18 (12): 1264-9 362-5 92. Martinez A, Sanchez M. Retrobulbar haemodynamic effects 99. Liesegang TJ, Skuta GL, Cantor LB, et al. 2003-2004 Basic of the latanoprost/timolol and the dorzolamide/timolol and clinical science course (BCSC) section 10: glaucoma. fixed combinations in newly diagnosed glaucoma patients. San Francisco (CA): American Academy of Ophthalmol- Int J Clin Pract 2007 May; 61 (5): 815-25 ogy, 2003 93. Konstas AG, Mikropoulos D, Dimopoulos AT, et al. Second-line therapy with dorzolamide/timolol or latanoprost/timolol fixed combination versus adding dorzolamide/ timolol fixed combination to latanoprost monotherapy. Br J Ophthalmol 2008 Nov; 92 (11): 1498-502 Correspondence: Dr Florent Aptel, Clinique Universitaire 94. Cvenkel B, Stewart JA, Nelson LA, et al. Dorzolamide/ d’Ophtalmologie, Centre Hospitalo-universitaire de Gre- timolol fixed combination versus latanoprost/timolol fixed noble, BP217, 38043 Grenoble cedex 09, France combination in patients with primary open-angle glauco- E-mail: [email protected]