ORIGINAL PAPER Preservative-free tafluprost in the treatment of open-angle or in India: a phase III clinical trial

A. Chabi, C. Baranak, R. Lupinacci, W. J. Herring

SUMMARY Merck & Co., Inc., Kenilworth, What’s known NJ, USA Aim: The aim of this study was to evaluate the efficacy and safety of preserva- • analogues (PGAs) such as tafluprost reduce intraocular pressure (IOP) and Correspondence to: tive-free (PF) tafluprost compared with PF in Indian subjects with open- W. Joseph Herring, Merck & delay progression of open-angle glaucoma angle glaucoma (OAG) or ocular hypertension. Methods: This was a randomised, Co., Inc., North Wales, PA (OAG). multicentre, double-masked, phase III trial. Subjects aged 18–80 years, following 19454, USA • Many PGAs contain preservatives to maintain washout of current medication, with intraocular pressure (IOP) ≥ 24 and Tel.: + 267 305 7933 sterility; however, preservatives may be Fax: + 267-305-6454 ≤ 36 mmHg in at least one eye were randomised in a 1:1 ratio to 0.0015% PF associated with toxic side effects which may Email: tafluprost or 0.5% PF timolol for 4 weeks. IOP was measured at 08:00, 10:00 compromise tolerability. [email protected] and 16:00 hours at baseline and at weeks 2 and 4. The primary efficacy end-point • Preservative-free (PF) and preservative-containing was the mean diurnal IOP change from baseline at week 4, and PF tafluprost was formulations of tafluprost show similar reductions considered non-inferior to PF timolol if the upper bound of the 95% confidence in IOP, and PF tafluprost was non-inferior to PF Disclosures interval (CI) for between-treatment differences was ≤ 1.5 mmHg. The secondary timolol, the accepted standard for glaucoma Financial support for the end-point was the proportion of subjects with ≥ 25% reduction in IOP from base- treatment, in a study in US and European conduct of the research and line at week 4. Results: In total, 190 subjects were randomised, 95 each, to PF subjects. the preparation of this article was provided by Merck & Co., tafluprost and PF timolol treatment. PF tafluprost was non-inferior to PF timolol What’s new Inc., Kenilworth, NJ. The with respect to diurnal IOP changes from baseline over 4 weeks. The mean PF • This is the first study to determine if the efficacy sponsor was involved in study À À design; in the collection, tafluprost-PF timolol difference in the diurnal IOP change was 1.7 (95% CI 2.6 of PF tafluprost is comparable to that of PF analysis and interpretation of to À0.7), suggestive of superiority for PF tafluprost. The secondary end-point was timolol in the treatment of OAG in an Indian data; in the writing of the achieved in a higher proportion of PF tafluprost group subjects. Both PF tafluprost population. report; and in the decision to and PF timolol were well-tolerated with similar incidences of adverse events. Con- • PF tafluprost was non-inferior to PF timolol in submit the article for clusions: PF tafluprost was safe and efficacious in reducing IOP in Indian sub- IOP reduction after 4 weeks. Our results also publication. WJH, RL, and CB jects. suggest that PF tafluprost might have greater are employees of Merck and IOP-lowering efficacy than PF timolol. own stock/stock options in Merck. AC was an employee of Merck and owned stock/stock options in Merck at the time the study was performed. Introduction to the effect of drugs that lower IOP through other mechanisms (13–15). Although PGAs are generally Glaucoma is characterised by progressive optic neu- well-tolerated, commonly observed local side effects ropathy; it affects millions of people worldwide and include ocular hyperaemia, changes to the eyelashes, is the second leading cause of blindness (1). A major and increased iris pigmentation (4,5). The sterility of modifiable risk factor in the progression of open- multi-dose ophthalmic formulations is maintained angle glaucoma (OAG) is elevated intraocular pres- by the inclusion of preservatives that prevent micro- sure (IOP). Hence, reduction in IOP with ocular bial growth and lower the risk of eye infections. hypotensive agents plays a key role in delaying OAG Numerous studies have reported toxic side effects of progression and subsequent loss of vision (2,3). these preservatives on the ocular surface (16–18), Prostaglandin analogues (PGAs) are topical ocular suggesting that preservatives may influence the toxic- hypotensive drugs that are considered first-line ther- ity of ophthalmic preparations. apy for the medical management of glaucoma (4,5). Many topical eye medications contain variable PGAs have proven effective in lowering the IOP in concentrations (0.004–0.02%) of the preservative subjects with OAG and those with ocular hyperten- benzalkonium chloride (BAK). BAK may compro- sion (OHT) (6–10) by increasing uveoscleral outflow, mise the tolerability of these medications over time, with minimal effects on aqueous inflow (11,12). The particularly in patients with comorbid ocular surface IOP reductions achieved with PGAs may be additive diseases such as chronic dry eye (9,18). Findings

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586. doi: 10.1111/ijcp.12815 577 578 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension

from previous preclinical and clinical studies have relatively short treatment duration of 4 weeks was suggested that in addition to its allergenic properties, based on evaluation of data from previous studies, BAK may be toxic to various tissues on the ocular which indicated that IOP-lowering effects for both surface (19–21). BAK may induce corneal damage as treatments would be apparent after this time. To evidenced by loss of microvilli, increased number of ensure masking, the PF tafluprost group received epithelial holes and loss of hexagonal shape, and masked PF vehicle in the morning and active PF conjunctival lymphocytic infiltration (21–23). tafluprost in the evening, while the PF timolol group Chronic subclinical ocular inflammation may affect received PF timolol maleate in the morning and eve- the tolerability of these medications and potentially ning. Subjects were enrolled at 12 centres in India limit the success of future glaucoma surgery (24,25). from 14 December 2011 through 7 May 2013. The Preservative-free (PF) ophthalmic formulations, or study was initiated following approval from Institu- formulations with a negligible concentration of BAK, tional Review Board/Independent Ethics Committee. offer a potential advantage over their preservative- Written informed consent was obtained from each containing (PC) counterparts as they may reduce patient, and the trial was conducted in accordance impairment of the ocular surface without compro- with the ethical principles that have their origin in mising efficacy (16,18,26–29). the Declaration of Helsinki, and consistent with Tafluprost 0.0015% (Zioptan , Merck & Co., Inc., Good Clinical Practice and all applicable local regula- Kenilworth, NJ) was the first PF PGA in the US mar- tory requirements. ket; PF tafluprost was approved in February 2012 (30,31) and has since been used in the treatment of Patients OHT and OAG. Tafluprost is chemically and struc- Men and women aged between 18 and 80 years diag- turally similar to other PGAs (32) and has been mar- nosed with primary OAG, pigmentary glaucoma, cap- keted in Japan and several European countries since sular glaucoma/pseudoexfoliation or OHT with a best 2008 (13). corrected Early Treatment of Diabetic Retinopathy Numerous studies have analysed the safety and Study (ETDRS) visual acuity score of +0.6 log MAR efficacy of tafluprost in comparison with those of (Snellen equivalent of 20/80) or better in each eye placebo and other hypotensive agents (33–35). Simi- and willing and able to avoid wearing contact lenses lar reductions in IOP have been achieved with PF throughout the study were eligible. Subjects pre- and PC formulations of tafluprost in subjects with scribed ocular hypotensive medication with a stable glaucoma and OHT (36). A previous study com- treatment regimen at least 30 days prior to the pared the safety and efficacy of PF tafluprost with screening or were treatment na€ıve were eligible. Sub- those of PF timolol, an accepted standard for glau- jects on glaucoma treatment at the time of screening coma treatment with well-established safety and effi- visit (visit 1) underwent a washout of previous glau- cacy. The results of this study conducted across coma treatment; following this, at the baseline visit centres in the USA, Spain and Switzerland showed (visit 2), the subjects’ mean (or median) IOP had to that PF tafluprost was well-tolerated and non-inferior be ≥ 24 and ≤ 36 mmHg in at least one eye at the to PF timolol at lowering IOP in subjects with OHT 08:00 hour time point, and there had to be a and OAG (13). < 5 mmHg difference in mean (or median) IOP This study aimed to determine if the efficacy of between the eyes at the 08:00, 10:00 and 16:00 hour PF tafluprost is comparable to that of PF timolol in time points. Subjects were excluded if they had a his- the treatment of OHT and OAG in an Indian popu- tory of, or current, abnormal corneal sensation or lation. The safety and tolerability of these two agents any abnormality in either eye preventing reliable were also compared. PF timolol is a non-selective tonometry, evidence of acute ocular injection, embed- b-adrenergic antagonist (beta-blocker) for ded corneal foreign body or clinically significant ocu- treating glaucoma. lar inflammation or infection in either eye within 3 months of the screening, or had undergone intraoc- Materials and methods ular surgery in either eye within 4 months prior to visit 1. Subjects who had experienced significant This multicentre, randomised, double-masked paral- visual field loss, used ocular medications other than lel-group, phase III clinical trial (clinicaltrials.gov antiglaucoma medications and topical ocular lubri- NCT01254604, Merck Protocol 002-01) was designed cants within 1 week prior to screening, or had a his- to compare the efficacy and safety of PF tafluprost tory of anterior or posterior uveitis in either eye (0.0015%) once-daily and PF timolol maleate (0.5%) within 6 months prior to screening, were excluded. twice-daily ophthalmic solutions in subjects with Subjects with a history of retinal disease, cardiovascu- OAG or OHT over a 4-week treatment period. The lar disorder or pulmonary disease were also excluded.

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension 579

Procedure Safety end-points The total duration of a patient’s participation in the Safety and tolerability were measured by clinical study was approximately 2.5 months, inclusive of a review of all applicable parameters including adverse pretreatment screening/washout period, 4-week treat- events (AEs) and vital signs. Ocular assessments for ment period, and posttreatment telephone follow-up safety included ETDRS visual acuity test, automated safety assessments. The washout period was up to perimetry, external and adnexa ocular examination, 4 weeks depending upon the prior glaucoma treat- conjunctival hyperaemia (the degree of conjunctival ment at the time of screening (4 weeks for PGAs or hyperaemia was graded in each eye using a 0–4 scale and b-adrenergic antagonists; 2 weeks of graded-colour photographs with 0.5-increments), for a-orb-adrenergic ; 1 week for choliner- ocular surface examination using fluorescein staining, gic agonists and carbonic anhydrase inhibitors and slit-lamp microscopy, gonioscopy and dilated fundus 5 days for miotics). During the washout period, the examination. Ocular safety findings that were suffi- subjects were permitted to instil investigator-supplied ciently pronounced were reported by the investigator hydrochloride at the investigator’s dis- as an AE. Safety outcomes were evaluated from the ‘all cretion until 7 days prior to randomisation. patients as treated’ set that included all randomised Subjects who met the eligibility criteria were ran- subjects who received at least one dose of study treat- domised in a 1:1 ratio to PF tafluprost (0.0015%) or ment. Conjunctival hyperaemia, ocular pain/stinging/ PF timolol maleate (0.5%) groups using a stratified irritation and ocular pruritus were a priori identified randomisation method. Baseline IOP (< 26 or AEs used for analysing between-group differences. ≥ 26 mmHg at 08:00 hours) and ocular diagnosis AEs were counted only once for each patient even if (OAG or OHT) were considered for stratification. The the AE occurred more than once during the study or eye with the highest IOP at baseline was designated as in more than one eye, and a threshold count of at least the ‘study eye’. Subjects received masked kits for four was chosen for statistical analyses. 2 weeks of study medication via an interactive voice response system using a computer-generated random Statistical analysis allocation schedule. Subjects randomised to the PF The primary hypothesis was that PF tafluprost would tafluprost group received masked PF vehicle in the be non-inferior to PF timolol if the upper bound of morning pouches and active PF tafluprost in the eve- the two-sided 95% confidence interval (CI) for the ning pouches, while the PF timolol group received PF between-group difference in mean diurnal IOP change timolol maleate in both pouches. Subjects were from baseline was ≤ 1.5 mmHg at week 4. An analysis instructed to instil one drop in each eye at of covariance (ANCOVA) model was used to estimate 08:00 hours (Æ1 h) from the unit dose (UD) marked the CI; the model included terms for treatment, base- for morning and similarly at 20:00 hours (Æ1 h) from line IOP, and ocular diagnosis (OAG or OHT). A the UD marked for evening on the day following the constrained longitudinal data analysis (cLDA) model baseline visit. Other study visits were at week 2 (visit was used to perform sensitivity analysis and included 3) and week 4 (visit 4) when IOP was measured at terms for treatment, week, ocular diagnosis and the three time points, 08:00, 10:00 and 16:00 hours, and interaction of week by treatment (37). For the sec- the morning eye drops were administered by site per- ondary end-point, the number and proportion of sub- sonnel immediately after the 08:00 hours IOP mea- jects for each treatment group were provided, and a surements. At least 14 days after the last dose of study 95% CI for the between-group difference was medication, subjects were contacted by telephone and calculated using the Miettinen and Nurminen assessed for serious adverse events (SAEs). method.p-values for the between-group differences in Any other ocular medication that might interfere safety outcomes were also calculated using the Mietti- with study parameters was not allowed during the nen and Nurminen method (38). Missing data were study duration. Use of steroid and CYP2D6 inhibitor handled using last-observation-carried-forward was also not permitted. method for the efficacy analysis. All analyses were per- formed using SAS version 9.1.3 (SAS Institute, Cary, Efficacy end-points NC, USA). The primary efficacy end-point evaluated the mean diurnal IOP change from baseline at week 4 in the Treatment compliance study eye. IOP was measured using a Goldmann Percentage compliance was calculated as follows: applanation tonometer. The secondary efficacy end- (number of days on therapy/number of days should point was the proportion of subjects with ≥ 25% be on therapy) 9 100. Subjects were considered to reduction in IOP from baseline to week 4. be ‘on therapy’ if both the morning and evening

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 580 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension

drops of study medication were administered. The treatment, had at least one efficacy measurement number of days a patient ‘should be on therapy’ was available for the analysis end-point, and did not defined as the total number of days from baseline to commit any protocol violations that might have sub- the last scheduled/discontinued day of treatment stantially affected the results of the primary efficacy administration. Subjects were provided medication end-point data. The full analysis set (FAS) provided dosing diaries to record daily instillation of the eye supportive analysis for the primary efficacy end-point drops. Compliance was assessed at each postbaseline and comprised all randomised subjects who received visit by the site personnel who checked the used at least one dose of study treatment and had at least (empty) and unused (full) UDs and medication dos- one efficacy measurement available for the analysis ing diaries of each patient. end-point. Subjects were included in the treatment group to which they were randomised for analysis of Study populations and sample size efficacy data using the PPS and the FAS. The study The per protocol set (PPS) comprised the primary planned to enrol 248 subjects (124 per treatment population for analysis of efficacy; it included sub- group) to yield approximately 230 evaluable subjects jects who received at least one dose of study (115 per treatment group) for the PPS. This sample

Assessed for eligibility, n = 224

Excluded, n = 34

Randomised, n = 190

Allocated to preservative-free tafluprost, n = 95 Allocated to preservative-free timolol, n = 95

Completed, n = 87 (91.6%) Completed, n = 86 (90.5%) Discontinued Discontinued • Adverse event, n = 3 (3.2%) • Adverse event, n = 2 (2.1%) • Lost to follow-up, n = 1 (1.1%) • Lost to follow-up, n = 1 (1.1%) • Protocol violation, n = 1 (1.1%) • Protocol violation, n = 2 (2.1%) • Patient withdrawal, n = 2 (2.1%) • Patient withdrawal, n = 4 (4.2%) Excluded Excluded • Protocol violation, n = 1 (1.1%) • Protocol violation, n = 0

Analysed for efficacy - Per protocol set Analysed for efficacy - Per protocol set Included, n = 84 (88.4%) Included, n = 84 (88.4%) Excluded,a n = 11 (11.6%) Excluded,a n = 11 (11.6%) • Inclusion/exclusion criteria not met, • Inclusion/exclusion criteria not met, n = 4 (4.2%) n = 4 (4.2%) • No postbaseline IOP, n = 6 (6.3%) • No postbaseline IOP, n = 5 (5.3%) • No study medication, n = 2 • No study medication, n = 1 (1.1%) • Prior medication washout time < 2/3 • Prior medication washout time of required, n = 0 < 2/3 of required, n = 1 (1.1%) • Noncompliance with study therapy, • Noncompliance with study therapy, n = 1 (1.1%) n = 1 (1.1%) • Use of prohibited concomitant • Use of prohibited concomitant medication, n = 2 (2.1%) medication, n = 3 (3.2%)

Figure 1 Disposition of subjects. aSubjects may have more than one reason for exclusion from the per protocol population. IOP, intraocular pressure

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension 581 size provided 90% power to establish that PF taflu- 4 in both treatment groups. Table 2 presents the prost was non-inferior to PF timolol. The power and results of the primary end-point analysis of mean sample size were based on the results from a previ- IOP change from baseline at 4 weeks. The estimated ous phase III non-inferiority study vs. timolol con- between-treatment difference (PF tafluprost-PF timo- ducted by Santen, Inc. The assumptions that were lol) was À1.7 mmHg in the PPS with a 95% CI from applied at week 4 were as follows: a = 0.025 (one- À2.6 to À0.7 mmHg. PF tafluprost was thus non- sided), non-inferiority margin = 1.5 mmHg, true inferior to PF timolol as the criterion for the primary treatment difference = 0 mmHg and standard devia- hypothesis was met; i.e. the upper bound of the two- tion = 3.5 mmHg. sided 95% CI for the between-group difference in mean diurnal IOP change from baseline was ≤ Results 1.5 mmHg. In fact, both the lower and upper lim- its of the CI were < 0, suggesting superiority of PF Patient disposition and demographics tafluprost over PF timolol with respect to the pri- Of the 224 screened subjects, 190 met the inclusion mary efficacy variable. Furthermore, the reduction in criteria and were equally randomised to the treat- IOP from baseline was greater in the PF tafluprost ment groups. Of these, 173 (91.1%) completed the group compared with the PF timolol group, and the study. Reasons for discontinuation were gener- upper bounds of the 95% CI for between-group dif- ally similar between the treatment groups (Fig- ferences was < 0 at each time point of both study ure 1). visits (Table 3, Figure 3). The ANCOVA conducted Demographical and baseline characteristics for the in the FAS and the sensitivity analysis based on the subjects by treatment group are presented in Table 1. cLDA model for the PPS and FAS provided analo- There was a small gender imbalance between the gous results. Table 4 presents the results of the sec- groups with more men/fewer women in the PF timo- ondary end-point analysis. More than half of the lol group vs. the PF tafluprost group. The character- subjects in both treatment groups achieved a istics related to baseline IOP and ocular diagnosis decrease of ≥ 25% in diurnal IOP from baseline, and were comparable between both treatment groups. this proportion was higher in the PF tafluprost group than in the PF timolol group at both visits. Treatment compliance The lower bounds of the 95% CIs for the between- Mean treatment compliance was 98.4% and 98.6% group differences at weeks 2 and 4 were > 0.0%, sug- for PF tafluprost and PF timolol groups, respectively. gesting superiority of PF tafluprost over PF timolol for this secondary variable. In addition, the results of Efficacy subgroup analyses by age, sex, baseline IOP and ocu- As shown in Figure 2, treatment effect was observed lar diagnosis were generally consistent across the sub- at the first posttreatment visit (week 2) through week groups, suggesting that the IOP-lowering effect of PF

Table 1 Subject demographics and baseline characteristics

Preservative-free Preservative-free Characteristic tafluprost (n = 93) timolol (n = 94) Difference* 95% CI

Sex, n (%) Men 59 (63.4) 72 (76.6) À13.2 À26.2 to À0.2 Women 34 (36.6) 22 (23.4) 13.2 0.2–26.2 Age Mean (SD), years 56.7 (11.5) 54.9 (13.4) 1.8 À1.8 to 5.4 Baseline IOP, n (%) < 26 mmHg 53 (57.0) 52 (55.3) 1.7 À12.6 to 15.9 ≥ 26 mmHg 40 (43.0) 42 (44.7) À1.7 À15.9 to 12.6 Ocular diagnosis, n (%) Ocular hypertension 33 (35.5) 33 (35.1) 0.4 À13.3 to 14.1 Open-angle glaucoma† 60 (64.5) 61 (64.9) À0.4 À14.1 to 13.3

† CI, confidence interval; IOP, intraocular pressure; SD, standard deviation. Includes subjects with primary open-angle glaucoma, pseudoexfoliative glaucoma and pigmentary glaucoma. *Difference = preservative-free tafluprostÀpreservative-free timolol; post hoc analysis.

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 582 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension

27 26 PF Tafluprost 25 PF Timolol 24

23 22 21 20 19 IOP (mmHg) 18 17 16 15 14 Baseline Week 2 Week 4 Time

Figure 2 Mean (Æ95% confidence interval) intraocular pressure at each visit in subjects with open-angle glaucoma or ocular hypertension receiving PF tafluprost or timolol – per protocol analysis. IOP, intraocular pressure; PF, preservative-free

Table 2 Mean (95% confidence interval) intraocular pressure* change from baseline in subjects receiving preservative- free tafluprost or timolol by visit – per protocol analysis

† Preservative-free tafluprost Preservative-free timolol Difference

Least squares Least squares Least squares Visit n mean (mmHg) 95% CI n mean (mmHg) 95% CI mean (mmHg) 95% CI

Week 2 84 À8.1 À8.8 to À7.4 82 À6.1 À6.8 to À5.4 À2.0 À3.0 to À1.1 Week 4 84 À8.3 À9.0 to À7.6 83 À6.6 À7.3 to À5.9 À1.7 À2.6 to À0.7‡

CI, confidence interval; n, number of subjects included in the analysis at the corresponding visit. *Average of the intraocular measurements within a visit. Mean (SD) at baseline = 24.8 (3.0) for PF tafluprost and 24.9 (2.6) for PF timolol. †Difference = preservative-free tafluprostÀpreservative-free timolol. ‡Non-inferiority of preservative-free tafluprost based on the prespecified non-inferiority margin of 1.5 mmHg.

Table 3 Intraocular pressure change from baseline* by visit and time point in subjects receiving preservative-free tafluprost or timolol – per protocol analysis

† Preservative-free tafluprost Preservative-free timolol Difference

Time Least squares Least squares Least squares Visit point n mean (mmHg) 95% CI n mean (mmHg) 95% CI mean (mmHg) 95% CI‡

Week 2 08:00 82 À8.7 À9.5 to À7.9 80 À7.0 À7.8 to À6.2 À1.7 À2.8 to À0.6 10:00 84 À8.1 À8.9 to À7.3 81 À5.8 À6.6 to À4.9 À2.3 À3.5 to À1.2 16:00 84 À7.5 À8.3 to À6.6 81 À5.5 À6.4 to À4.7 À2.0 À3.2 to À0.8 Week 4 08:00 84 À9.0 À9.7 to À8.2 83 À7.4 À8.1 to À6.6 À1.6 À2.6 to À0.6§ 10:00 84 À8.0 À8.8 to À7.2 82 À6.4 À7.2 to À5.6 À1.6 À2.7 to À0.5§ § 16:00 84 À7.7 À8.6 to À6.8 82 À6.0 À6.9 to À5.1 À1.7 À3.0 to À0.5

CI, confidence interval; n, number of subjects included in the analysis at the corresponding visit and time point. *Change from baseline = Week xÀbaseline. †Difference = preservative-free tafluprostÀpreservative-free timolol maleate. ‡Based on analysis of covariance with terms of treatment, baseline intraocular pressure and ocular diagnosis (open-angle glaucoma or ocular hypertension). § Non-inferiority of preservative-free tafluprost based upon prespecified non-inferiority margin of 1.5 mmHg.

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension 583

2

Non-inferiority of PF tafluprost 1

0 Superiority of PF tafluprost

–1

–2

–3

–4 Mean difference in IOP and 95 % CIs (mmHg) 8 : 00 10 : 00 16 : 00 8 : 00 10 : 00 16 : 00 2 weeks 4 weeks

Figure 3 Between-group differences (95% confidence intervals) for the mean changes from baseline in intraocular pressure at each visit and time point in subjects receiving preservative-free tafluprost or timolol – per protocol analysis. CI, confidence interval; IOP, intraocular pressure; PF, preservative-free

Table 4 Proportion (95% confidence interval) of subjects with ≥ 25% reduction in diurnal intraocular pressure* from baseline with preservative-free tafluprost or timolol – per protocol analysis

‡ Preservative-free tafluprost Preservative-free timolol Treatment difference

Visit m/N† % 95% CI m/N† % 95% CI % 95% CI

Week 2 67/84 79.8 71.2 to 88.4 45/82 54.9 44.1 to 65.6 24.9 11.1 to 38.7 Week 4 65/84 77.4 68.4 to 86.3 48/83 57.8 47.2 to 68.5 19.5 5.7 to 33.4

CI, confidence interval. *Average of intraocular measurements within a visit. †m/N = number of subjects with ≥ 25% reduction in ‡ diurnal intraocular pressure from baseline/number of subjects in the analysis. Stratified Miettinen and Nurminen method. tafluprost (vs. PF timolol) was independent of these most frequently reported drug-related AE was con- variables. junctivitis in five (5.4%) subjects with PF tafluprost and six subjects (6.4%) with PF timolol. Tolerability and safety The average level of conjunctival hyperaemia (in Table 5 summarises the AEs observed in this study. study eye), assessed using slit-lamp microscopy, was PF tafluprost and PF timolol were both well-toler- below half a grade for both treatment groups at base- ated throughout the 4-week treatment period. Simi- line. Increase in hyperaemia with at least one grade lar incidences of AEs (95% CI for the between-group (in either eye) was observed in 6 (6.8%) and 4 difference: À17.3% to 9.4%) and discontinuations (4.7%) subjects for the PF tafluprost and PF timolol because of AEs (95% CI for between-group differ- groups, respectively. No other clinically relevant ence: À3.5% to 5.7%) were reported in both treat- changes were observed in the slit-lamp microscopy, ment groups. There were no deaths in the study. No visual acuity, visual field and ophthalmoscopy exami- serious AEs were reported in the PF tafluprost group, nations in either treatment group. but one subdural haematoma (not related to study drug) occurred in the PF timolol group. Of the pre- Discussion specified AEs, ocular pruritus and ocular hyperaemia were reported with a slightly higher incidence in the These study results in Indian subjects demonstrated PF tafluprost than the PF timolol group, although that the reduction in mean IOP with PF tafluprost none of these events were significantly different and PF timolol was evident after 2 weeks of treat- between the treatment groups (ocular pruritus, ment and sustained at 4 weeks. The IOP-lowering p = 0.188; conjunctival hyperaemia, p = 0.145). The effect of PF tafluprost was non-inferior to that of PF

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 584 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension

Table 5 Summary of adverse events in subjects receiving preservative-free tafluprost or timolol – all patients as treated analysis

Preservative- Preservative- Difference in preservative-free free tafluprost free timolol tafluprost and timolol

n % n % Estimate (95% CI)* p-value*

Subjects in population 93 94 One or more adverse events 28 30.1 32 34.0 À3.9 (À17.3 to 9.4) † Drug-related adverse events 15 16.1 22 23.4 À7.3 (À18.6 to 4.1) Serious adverse events 0 0.0 1 1.1 À1.1 (À3.1 to 1.0) Discontinued due to adverse event 3 3.2 2 2.1 1.1 (À3.5 to 5.7) Specific adverse events with an incidence of ≥ 4 subjects in ≥ 1 treatment groups ‡ Conjunctival hyperaemia 9 9.7 4 4.3 5.4 (À1.8 to 12.7) 0.145 Ocular/eye pain/stinging/irritation 6 6.5 5 5.3 1.1 (À5.6 to 7.9) 0.742 Ocular/eye pruritis 7 7.5 3 3.2 4.3 (À2.1 to 10.8) 0.188

† ‡ CI, confidence interval. *Based on Miettinen and Nurminen method. Determined by investigator to be drug-related. Conjunctival hyperaemia = conjunctival hyperaemia, erythema of the eyelid and ocular hyperaemia.

timolol. Moreover, both lower and upper limits of study align well with the findings of our earlier large the 95% CI of the mean IOP difference were < 0, 12-week phase III clinical trial conducted in the USA suggesting that PF tafluprost conferred a superior and Europe in patients with OAG or OHT who were IOP-lowering effect compared with PF timolol. Dif- predominantly white (75%) or black (23%), with ferences of around 2 mmHg in favour of PF taflu- < 2% being Asian (13). This study results also con- prost were witnessed at both weeks 2 and 4. formed to those published by Hamacher et al. who Similarly, the proportion of subjects with ≥ 25% compared the 4-week efficacy and safety of PC and decrease in diurnal IOP from baseline was higher in PF tafluprost formulations. PF tafluprost was well- the PF tafluprost than the PF timolol group at weeks tolerated with no unexpected AEs and showed equiv- 2 and 4. alent decreases in IOP compared with PC tafluprost PF tafluprost and PF timolol were generally well- (36). Similar results have also been obtained in an tolerated. The AE profiles were similar for both observational study of data from 544 patients treated groups, although fewer AEs, serious AEs and drug- with PF tafluprost, which showed that PF tafluprost related AEs were observed in the PF tafluprost group. effectively lowered IOP and could be safely used in a Consistent with previous findings (13), conjunctival patient population with poor IOP control and/or tol- hyperaemia occurred in a slightly higher proportion erability issues (9). of subjects in the PF tafluprost group than in the PF The short duration of the study naturally limits an timolol group. Conjunctival hyperaemia has been in-depth safety assessment of these long-term medi- reported as an ocular side effect of PGAs, with no cations. Although a time frame of 4 weeks is suffi- apparent serious consequences (4). cient to evaluate changes in IOP and assess AEs, a The prevalence of glaucoma is increasing world- longer follow-up may be required to assess specific wide, and the burden is substantially higher in Asia AEs such as iris pigmentation changes. than in the rest of the world (1,39). Global estimates The fact that only 77% of the planned number of suggest that the burden of OAG and angle-closure enrolled patients could be achieved is a second topic glaucoma was 60.5 million in 2010, with Asians rep- of consideration. Importantly, 80% power was main- resenting almost half (47%) of this patient tained for the primary hypothesis with the eventual population. Furthermore, the burden of OAG and sample of 190 patients and the assumptions prespeci- angle-closure glaucoma is expected to increase to fied in the study protocol. Hence, we do not believe 79.6 million by 2020, with 74% of patients having that the difference in sample size would have materi- OAG (1). Estimates per 2010 statistics suggest that, ally affected the outcome of this study. in India, approximately 6.5 million people have pri- mary OAG, and the condition remains undetected or Conclusions untreated in a high proportion of patients (40). Tafluprost, similar to other PGAs, has a favourable The results of this study demonstrate that the IOP- safety and efficacy profile (33–35). The results of this lowering efficacy of PF tafluprost was non-inferior to

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension 585 that of PF timolol in the treatment of OHT and and Ophthalmology (ARVO), 4–8 May 2014, OAG in the Indian population. Furthermore, the Orlando, FL, USA. results suggested that PF tafluprost might have Medical writing assistance was provided by greater IOP-lowering efficacy than PF timolol. Both Elphine Telles, PhD, of Cactus Communications. study medications had an acceptable safety profile, This assistance was funded by Merck & Co., Inc., and only a modest number of treatment-induced Kenilworth, NJ, USA. Christopher Lines, PhD from AEs typical for these two (classes of) medications Merck & Co., Inc., provided comments and edits. were reported. The list of investigators who participated in the trial is provided in Appendix 1. Author contributions Funding AC, CB and RL conceived, designed and planned the study. AC collected the data. AC and RL performed the Financial support for the conduct of the research and data analysis. AC, CB, RL and WJH were involved in the preparation of this article was provided by Merck interpretation of the results. All authors critically & Co., Inc., Kenilworth, NJ. The sponsor was involved reviewed or revised the manuscript for important intel- in study design; in the collection, analysis and inter- lectual content and all approved it for submission. pretation of data; in the writing of the report; and in the decision to submit the article for publication. Acknowledgements WJH, RL and CB are employees of Merck and own stock/stock options in Merck. AC was an employee of These data were previously presented at the annual Merck and owned stock/stock options in Merck at the meeting of the Association for Research in Vision time the study was performed.

hypertension. The Study Group. Oph- benzalkonium chloride on growth and survival of References thalmology 1996; 103: 126–37. Chang conjunctival cells. Invest Ophthalmol Vis Sci 11 Linden C, Alm A. Prostaglandin analogues in the 1999; 40: 619–30. 1 Quigley HA, Broman AT. The number of people treatment of glaucoma. Drugs Aging 1999; 14: 387– 21 Noecker RJ, Herrygers LA, Anwaruddin R. Corneal with glaucoma worldwide in 2010 and 2020. Br J 98. and conjunctival changes caused by commonly used Ophthalmol 2006; 90: 262–7. 12 Toris CB, Gabelt BT, Kaufman PL. Update on the glaucoma medications. Cornea 2004; 23: 490–6. 2 Heijl A, Leske MC, Bengtsson B et al. Reduction of mechanism of action of topical for 22 Becquet F, Goldschild M, Moldovan MS, Ettaiche M, intraocular pressure and glaucoma progression: intraocular pressure reduction. Surv Ophthalmol Gastaud P, Baudouin C. Histopathological effects of results from the Early Manifest Glaucoma Trial. 2008; 53(Suppl. 1): S107–20. topical ophthalmic preservatives on rat corneocon- Arch Ophthalmol 2002; 120: 1268–79. 13 Chabi A, Varma R, Tsai JC et al. Randomized clin- junctival surface. Curr Eye Res 1998; 17: 419–25. 3 Leske MC, Heijl A, Hussein M et al. Factors for ical trial of the efficacy and safety of preservative- 23 Burstein NL. Preservative cytotoxic threshold for glaucoma progression and the effect of treatment: free tafluprost and timolol in patients with open- benzalkonium chloride and chlorhexidine diglu- the early manifest glaucoma trial. Arch Ophthalmol angle glaucoma or ocular hypertension. Am J Oph- conate in cat and rabbit corneas. Invest Ophthalmol 2003; 121:48–56. thalmol 2012; 153: 1187–96. Vis Sci 1980; 19: 308–13. 4 Alm A, Grierson I, Shields MB. Side effects associ- 14 Tabet R, Stewart WC, Feldman R, Konstas AG. A 24 Baudouin C. Side effects of antiglaucomatous drugs ated with prostaglandin analog therapy. Surv Oph- review of additivity to prostaglandin analogs: fixed on the ocular surface. Curr Opin Ophthalmol 1996; thalmol 2008; 53(Suppl. 1): S93–105. and unfixed combinations. Surv Ophthalmol 2008; 7:80–6. 5 Weinreb RN, Khaw PT. Primary open-angle glau- 53: S85–S92. 25 Broadway D, Grierson I, Hitchings R. Adverse coma. Lancet 2004; 363: 1711–20. 15 Webers CA, Beckers HJ, Zeegers MP, Nuijts RM, Hen- effects of topical antiglaucomatous medications on 6 Aptel F, Cucherat M, Denis P. Efficacy and tolera- drikse F, Schouten JS. The intraocular pressure-lower- the conjunctiva. Br J Ophthalmol 1993; 77: 590–6. bility of prostaglandin analogs: a meta-analysis of ing effect of prostaglandin analogs combined with 26 Gross RL, Peace JH, Smith SE et al. Duration of randomized controlled clinical trials. J Glaucoma topical b-blocker therapy: a systematic review and IOP reduction with BAK-free solution. 2008; 17: 667–73. meta-analysis. Ophthalmology 2010; 117: 2067–74 e6. J Glaucoma 2008; 17: 217–22. 7 Denis P, Lafuma A, Khoshnood B, Mimaud V, Ber- 16 Kuppens EV, de Jong CA, Stolwijk TR, de Keizer 27 Lewis RA, Katz GJ, Weiss MJ et al. Travoprost 0.004% deaux G. A meta-analysis of topical prostaglandin RJ, van Best JA. Effect of timolol with and without with and without benzalkonium chloride: a comparison analogues intra-ocular pressure lowering in glau- preservative on the basal tear turnover in glau- of safety and efficacy. JGlaucoma2007; 16:98–103. coma therapy. Curr Med Res Opin 2007; 23: 601–8. coma. Br J Ophthalmol 1995; 79: 339–42. 28 Thakkar BV, Hirsch AT, Satran D et al. The effi- 8 Eyawo O, Nachega J, Lefebvre P et al. Efficacy and 17 Leung EW, Medeiros FA, Weinreb RN. Prevalence cacy and safety of enhanced external counterpulsa- safety of prostaglandin analogues in patients with of ocular surface disease in glaucoma patients. J tion in patients with peripheral arterial disease. predominantly primary open-angle glaucoma or Glaucoma 2008; 17: 350–5. Vasc Med 2010; 15:15–20. ocular hypertension: a meta-analysis. Clin Ophthal- 18 Pisella PJ, Pouliquen P, Baudouin C. Prevalence of 29 Niwano Y, Iwasawa A, Ayaki M. Ocular surface mol 2009; 3: 447–56. ocular symptoms and signs with preserved and cytotoxicity and safety evaluation of tafluprost, a 9 Hommer A, Mohammed Ramez O, Burchert M, preservative free glaucoma medication. Br J Oph- recently developed anti-glaucoma prostaglandin Kimmich F. IOP-lowering efficacy and tolerability thalmol 2002; 86:418–23. analog. Ophthalmol Eye Dis 2014; 6:5. of preservative-free tafluprost 0.0015% among 19 Baudouin C, Labbe A, Liang H, Pauly A, Brignole- 30 U.S. Food and Drug Administration. FDA approves patients with ocular hypertension or glaucoma. Baudouin F. Preservatives in eyedrops: the good, the Zioptan to treat elevated eye pressure, 2012. http:// Curr Med Res Opin 2010; 26: 1905–13. bad and the ugly. Prog Retin Eye Res 2010; 29: 312–34. www.fda.gov/NewsEvents/Newsroom/ 10 Watson P, Stjernschantz J. A six-month, random- 20 De Saint Jean M, Brignole F, Bringuier AF, Bauchet PressAnnouncements/ucm291966.htm (updated 15 ized, double-masked study comparing latanoprost A, Feldmann G, Baudouin C. Effects of February 2012; accessed 18 January 2016). with timolol in open-angle glaucoma and ocular

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586 586 Preservative-free tafluprost in the treatment of open-angle glaucoma or ocular hypertension

31 Swymer C, Neville MW. Tafluprost: the first latanoprost in healthy males. J Ocul Pharmacol designs. Sankhya: Indian J Stat Ser B 2000; 62: preservative-free prostaglandin to treat open-angle Ther 2007; 23: 359–65. 134–48. glaucoma and ocular hypertension. Ann Pharma- 35 Uusitalo H, Pillunat LE, Ropo A. Efficacy and 38 Miettinen O, Nurminen M. Comparative analysis cother 2012; 46: 1506–10. safety of tafluprost 0.0015% versus latanoprost of two rates. Stat Med 1985; 4: 213–26. 32 Nakajima T, Matsugi T, Goto W et al. New fluoro- 0.005% eye drops in open-angle glaucoma and 39 Wong TY, Loon SC, Saw SM. The epidemiology of prostaglandin F2. ALPHA. Derivatives with prosta- ocular hypertension: 24-month results of a ran- age related eye diseases in Asia. Br J Ophthalmol noid FP-receptor agonistic activity as potent domized, double-masked phase III study. Acta 2006; 90: 506–11. ocular-hypotensive agents. Biol Pharm Bull 2003; Ophthalmol 2010; 88:12–9. 40 George R, Ve RS, Vijaya L. Glaucoma in India: 26: 1691–5. 36 Hamacher T, Airaksinen J, Saarela V, Liinamaa MJ, estimated burden of disease. J Glaucoma 2010; 19: 33 Egorov E, Ropo A. Adjunctive use of tafluprost Richter U, Ropo A. Efficacy and safety levels of 391–7. with timolol provides additive effects for reduction preserved and preservative-free tafluprost are of intraocular pressure in patients with glaucoma. equivalent in patients with glaucoma or ocular 19 – Eur J Ophthalmol 2009; : 214 22. hypertension: results from a Paper received February 2016, accepted April 2016 34 Sutton A, Gilvarry A, Ropo A. A comparative, pla- analysis. Acta Ophthalmol Suppl 2008; 242:14–9. cebo-controlled study of fluoro- 37 Liang K-Y, Zeger SL. Longitudinal data analysis of prostaglandin-receptor agonists tafluprost and continuous and discrete responses for pre-post

Appendix 1: Investigators participating in the trial The following investigators participated in the trial: Jain Hospital, Bengaluru; Dr Balekudaru Shantha, Dr Himadri Dutta, Regional Institute of Ophthal- Glaucoma Services, Sankara Nethralaya Vision mology, Kolkota; Dr Rahim Muljiani, Conwest & Research Foundation, Chennai; Dr P Sathyan and Dr Manjula S. Badani Jain Hospital, Mumbai; Dr Ruvit Ganesh Venkataraman, Aravind Eye Hospital, Coim- Nikam, Aditya Jyot Eye Hospital, Mumbai; Dr Rajul batore; Dr Prashanth Ranganath, Aravind Eye Hospi- Parekh, Bombay City Eye Institute, Mumbai; Dr S B tal, Madurai; Dr R Ramkrishnan, Aravind Eye Kelkar, National Institute of Ophthalmology, Pune; Hospital, Tirunelveli; and Dr Sasikumar Rajesh, Nar- Dr Senthil Sirisha, National Institute of Ophthalmol- ayana Netralaya, Bengaluru. ogy, Pune; Dr Rajesh Parekh, Bhagwan Mahaveer

ª 2016 John Wiley & Sons Ltd Int J Clin Pract, July 2016, 70, 7, 577–586