Study design

This was an open-label, phase IV clinical study (EudraCT registration number: 2014-005273-37) conducted at 16 centres in Finland, Germany, Italy and the UK from June 2015 to May 2016. The study was reviewed and approved by the appropriate Independent Ethics Committees in the participating countries and conducted in accordance with the Good Clinical Practice guidelines of the International Council on Harmonisation and the ethical principles of the Declaration of Helsinki.

Patient population

Patients included in this study were aged ≥18 years, diagnosed with OH or OAG, inclusive of both primary OAG and pseudoexfoliation glaucoma, and treated with bimatoprost–timolol in the evening (BAK-preserved or PF single-dose formulation) in one or both eyes for ≥4 weeks before screening. Patients presented at screening with conjunctival redness/hyperaemia of at least moderate severity (grade ≥2) in at least one treated eye and ≥one ocular symptom of at least mild severity (grade ≥2) in either eye. Exclusion criteria included: use of more than two active medicinal agents to treat OH or OAG in the 6 months prior to screening; anterior chamber angle grade <2 (Shaffer classification) in either treated eye; and any corneal abnormality or other condition preventing applanation tonometry, including prior refractive eye surgery and IOP >21 mm Hg in the treated eye(s) at screening. A full list of inclusion and exclusion criteria is presented in online supplementary table S1.

Treatments and assessments

Eligible patients had used BAK-preserved or PF bimatoprost–timolol (bimatoprost 0.03%–timolol 0.5%) eye drops in the evening for ≥4 weeks prior to screening. Study treatment kits, containing PF tafluprost–timolol eye drops (tafluprost 0.0015%–timolol 0.5%) in unit-dose containers, were dispensed to patients at the screening visit. Patients were not blinded to treatment because an open-label design could not be avoided owing to differences in packaging between BAK-preserved and PF bimatoprost–timolol and PF tafluprost–timolol. Each patient instilled one drop of tafluprost–timolol once daily at 21:00 (±1 hour) in the affected eye(s) for 12 weeks. Drug accountability documentation and dosing data from case report forms were used to assess treatment compliance. Patients were assessed at screening, and at 2, 6 and 12 weeks postscreening. After week 12, a poststudy visit was scheduled, and the investigator was free to prescribe any IOP-lowering medication.

Ocular signs and symptoms

Ocular signs and symptoms were assessed at every visit and were defined by the criteria described in online supplementary table S2. Ocular symptoms were evaluated per patient and treated eyes were considered together. Ocular signs were analysed in the worst eye designated for each sign at screening. The coprimary endpoints were changes from screening in conjunctival hyperaemia and worst ocular symptom at week 12. The severity of conjunctival hyperaemia was assessed from screening through to week 12. Use of the Ora Calibra^TM Redness Scale #6.0 (0–4 scale) was made under licence from Ora, Inc. Patients indicated their perceived worst ocular symptom at screening.

Secondary endpoints were changes from screening in ocular signs and symptoms, other than conjunctival hyperaemia, at week 12. The patient was asked about each symptom by a leading question, with symptoms graded 0 (none), 1 (trace), 2 (mild), 3 (moderate) or 4 (severe). A total symptom score (0–20) was calculated. Fluorescein tear break-up time was assessed by examination of tear film under a slit lamp following instillation of 2 µL of non-preserved 2% sodium fluorescein dye to the eyes. The time taken (in seconds) to form micelles or for dry spots to develop was recorded as the break-up time. Corneal and conjunctival fluorescein staining were also evaluated. Using reference pictures (Oxford Grading scale) the corneal fluorescein staining and nasal and temporal conjunctival fluorescein stainings were scored from 0 to V each. The presence of blepharitis was also evaluated, and the severity was graded 0 (none), 1 (mild), 2 (moderate) or 3 (severe). Tear production was assessed using the Schirmer-I test for 5 min without anaesthesia.

Adverse events

Treatment-emergent ocular and non-ocular AEs were reported at each postscreening visit. The information obtained included event term, report source, the seriousness of the event, onset and resolution date, frequency, severity, relation to study drop instillation, location (left/right eye, both or not applicable), study drug treatment action, and the investigator’s causality assessment of the study treatment and outcome. All AEs were coded using the latest Medical Dictionary for Regulatory Activities.

Ocular safety and quality of life

At each visit, IOP was measured in both eyes using Goldmann applanation tonometry; the right eye was measured first. Two consecutive measurements were taken to determine the mean IOP. If the initial two measurements differed by ≥3 mm Hg, then a third measurement was taken and the median IOP was determined. Other measures of ocular safety and quality of life (QOL) are described in the online supplementary information.

Sample size

A mean change of 0.37 units (SD 1.12) from screening in conjunctival redness/hyperaemia was assumed from the previous tafluprost switch studies.11 Using these estimates, it was determined using a paired t-test that 100 patients would be required for a power of 90% for conjunctival hyperaemia and >99% for worst ocular symptom.11 12 The calculations were done using the nQuery Advisor (V.6.0). The paired t-test was initially used to calculate sample size. However, the Wilcoxon signed rank test was found to perform better for heavy-tailed distributions and was thus used for the primary analysis.

Statistical methods

The intention-to-treat (ITT) dataset included all enrolled patients who received at least one dose of tafluprost–timolol and had at least one postscreening primary outcome measurement available. The safety set included all enrolled patients who had at least one dose of study treatment and had a subsequent safety measurement. The primary outcome measures for ocular signs and symptoms were assessed using the Wilcoxon signed rank test. No imputations for missing data were carried out. However, sensitivity analyses using the last observation carried forward imputation were carried out for the primary outcome measures. The analyses of secondary outcome and IOP measures were completed using standard statistical methods for paired data (eg, McNemar’s test for binary data, Wilcoxon signed rank test for ordinal data and the paired t-test for continuous data). For AEs, both patient and event counts were calculated, and events leading to discontinuations were summarised. Best corrected visual acuity, biomicroscopy, ophthalmoscopy, visual field test, drop discomfort and Comparison of Ophthalmic Medications for Tolerability (COMTol) are ocular safety and QOL outcomes, which were analysed descriptively.

Patient involvement

No patients were involved in setting the research question or the outcome measures, nor were they involved in the design or implementation of this study. There are no plans to involve patients in the dissemination of results as the open-label nature of the study meant that patients were aware of which medication they received.

Patient demographics and baseline characteristics

A total of 126 patients were screened. Of the 123 patients enrolled, two had no postscreening data and were excluded; therefore, 121 (98.4%) patients were included in the ITT analysis, of which 114 (94.2%) patients completed the study (BAK-preserved, n=71; PF, n=43) (figure 1). The safety set comprised 123 patients. The mean (range) age was 66 (36–86) years, and more than half of the patients were female (54.5%) (table 1). Of the patients, ~70% in both BAK-preserved and PF subgroups had used bimatoprost–timolol for at least 6 months; 76 patients had used BAK-preserved (62.8%) and 45 had used PF (37.2%) bimatoprost–timolol. Approximately 20% of patients were diagnosed with OH and 75% with OAG. Most patients (91.7%) required treatment in both eyes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Patient disposition by previous bimatoprost–timolol treatment. After initial screening, three patients did not meet the inclusion criteria. A total of nine (7.4%) patients discontinued the study; five discontinued because of AEs and four withdrew from the study. AE, adverse event; BAK, benzalkonium chloride; PF, preservative-free.

Changes to signs and symptoms

A significant improvement was observed in the severity of both conjunctival hyperaemia and worst ocular symptom compared with screening after switching from bimatoprost–timolol to tafluprost–timolol (p<0.001 at weeks 2, 6 and 12). The mean ±SD deviation grade of conjunctival hyperaemia for all patients decreased from 2.26±0.45 at screening to 0.94±0.64 at week 12 (a mean reduction of 58.5%) (figure 2A). The percentage of patients with conjunctival hyperaemia significantly reduced from 76 (100%) and 45 (100%) patients at screening in BAK-preserved and PF bimatoprost–timolol groups, respectively, to 47 (66.2%) and 31 (72.1%) at week 12 (figure 2B). All patients identified a worst ocular symptom at screening, which was at least mild in severity; the number of patients with the identified symptom was reduced to 47 (41.2%) at week 12. The number of patients with moderate and severe worst ocular symptom decreased from 62 (51.2%) and 12 (9.9%) at screening, to 11 (9.6%) and one (0.9%) at week 12, respectively (figure 2C). In the BAK-preserved and PF bimatoprost–timolol subgroups, the number of patients with moderate and severe worst ocular symptom decreased from 46 (60.5%) to five (7.0%) patients and from 28 (62.2%) to seven (16.3%) patients, respectively (figure 2D).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Comparison of week 12 outcomes with screening in conjunctival hyperaemia and worst ocular symptom after switching from bimatoprost–timolol to tafluprost–timolol (A) change in conjunctival hyperaemia from screening (n=121) to week 12 (n=114); (B) breakdown of changes in conjunctival hyperaemia severity by subgroup at week 12 compared with screening. One patient in the ITT dataset violated inclusion criterion 2 and only had mild conjunctival hyperaemia at screening; (C) severity of worst ocular symptom at screening and week 12 in all patients; and (D) changes in severity of worst ocular symptom by subgroup at week 12 compared with screening. BAK, benzalkonium chloride; ITT, intention-to-treat; PF, preservative-free.

The frequencies of abnormal ocular signs and symptoms were significantly reduced at week 12 after switching from bimatoprost–timolol to tafluprost–timolol (p<0.012 for signs and p<0.001 for symptoms) (figure 3A, B, C and D). For ocular signs, the greatest relative reductions from screening were observed in tear secretion (screening, 62.8%; week 12, 37.7%) and blepharitis (screening, 44.6%; week 12, 27.2%). For ocular symptoms, all relative reductions were over 50%, and the greatest were observed in foreign body sensation (screening, 49.6%; week 12, 20.2%) and itching (screening 51.2%; week 12, 21.9%).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Secondary endpoints (A) abnormal ocular signs at screening (n=121); (B) abnormal ocular signs at week 12 (n=114); (C) abnormal ocular symptoms at screening; (D) abnormal ocular symptoms at week 12. BAK, benzalkonium chloride; PF, preservative-free.

Adverse events

Overall, 70 treatment-emergent AEs based on the safety dataset (n=123) were reported by 41 (33.3%) patients during the study, of which 15 events in 12 (9.8%) patients were ocular and 55 events in 34 (27.6%) patients were non-ocular (table 2). Only 12 AEs in 10 (8.1%) patients were classified as being related to tafluprost–timolol. Two patients had serious AEs: worsening of arterial branch occlusion (resolved after 4 weeks) and paroxysmal atrial flutter with high-grade atrioventricular block (resolved in 2 days); both of which were adjudicated by the investigator and sponsor to be unrelated to tafluprost–timolol treatment. A total of five patients discontinued the study because of AEs, which were: two cases of moderate increase in IOP; moderate pruritus and eye pruritus, a moderate urticaria; and a severe increase in lacrimation. There were no deaths during the study.

Ocular safety and quality of life

At screening, IOP was well controlled with bimatoprost–timolol treatment (n=123; mean IOP 15.9±2.1 mm Hg); this was sustained at week 12 (n=114; mean IOP 16.3±2.3 mm Hg) and was clinically insignificant and statistically non-inferior compared with screening (0.34 mm Hg; 95% upper limit 0.86 mm Hg; p=0.134). IOP was maintained at ≤21 mm Hg for >97% of patients and ≤18 mm Hg for >80% of patients. Other results for ocular safety and QOL are described in the online supplementary information (online supplementary table S3).