Advantages of Efficacy and Safety of Fixed-Dose Tafluprost/Timolol Combination Over Fixed-Dose Latanoprost/Timolol Combination

Purpose To compare the safety and efficacy of fixed-dose tafluprost/timolol combination (Taf/T-FDC) with those of fixed-dose latanoprost/timolol combination (Lat/T-FDC) by measuring the intraocular pressure (IOP)-lowering effect, ocular pharmacokinetics, and ocular surface toxicity. Methods The IOP-lowering effect of Taf/T-FDC and Lat/T-FDC in ocular normotensive monkeys was evaluated at 4 and 8 h after instillation in study A, at 12, 14, 16, and 18 h after instillation in study B, and at 24, 26, 28, and 30 h after instillation in study C. Drug penetration into the eye was evaluated by measuring the concentrations of timolol, tafluprost acid (active metabolic form of tafluprost), and latanoprost acid (active metabolic form of latanoprost) using liquid chromatography coupled with tandem mass spectrometry after single instillation of Taf/T-FDC or Lat/T-FDC to Sprague Dawley rats. Cytotoxicity following 1–30 min exposure of SV40-transformed human corneal epithelial cells to Taf/T-FDC or Lat/T-FDC was analyzed using 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium assays. Undiluted and 10-fold diluted solutions of each FDC were evaluated. Results The IOP-lowering effect of Taf/T-FDC was almost equivalent to that of Lat/T-FDC at 4–8 h after instillation. The peak IOP reduction of Taf/T-FDC and Lat/T-FDC was observed at 8 h after instillation, and there is no difference between the two. The difference between them was observed at 24–30 h after instillation, and Taf/T-FDC demonstrated a significantly greater IOP-lowering effect than Lat/T-FDC at 24–30 h after instillation. The IOP-lowering effect of Taf/T-FDC was sustained up to 30 h after instillation, while that of Lat/T-FDC had almost disappeared at 28 h after instillation. Timolol concentrations in aqueous humor after Taf/T-FDC instillation were higher than those after Lat/T-FDC instillation (Cmax, 3870 ng/mL vs 1330 ng/mL; AUCinf, 3970 ng·h/mL vs 1250 ng·h/mL). The concentrations of tafluprost acid and latanoprost acid in aqueous humor after instillation of Taf/T-FDC and Lat/T-FDC, respectively, were similar to those after instillation of mono-preparations of tafluprost and latanoprost, respectively. The cytotoxic effect of Taf/T-FDC to the human corneal epithelial cells was significantly lower than that of Lat/T-FDC at all evaluated time points in both undiluted and 10-fold diluted FDCs. Conclusion Taf/T-FDC provides increased IOP-lowering effect duration and lower potential ocular surface toxicity than Lat/T-FDC.


Introduction
Glaucoma, a neurodegenerative ocular disease characterized by selective loss of retinal ganglion cells (RGCs) and resultant visual field defects, is the second leading cause of blindness worldwide. Elevated intraocular pressure (IOP) is a major risk factor for glaucoma progression [1,2], and IOP reduction is the main target of glaucoma treatment [2,3].
Glaucoma patients are initially treated with pharmacological monotherapy, such as prostaglandin analogue (PGA) or beta-adrenergic antagonist; however, in more than 40% patients, additional IOP-lowering drugs are required to control their IOP; hence, two or more drugs are used in the treatment [4][5][6]. However, the addition of a second drug increases the complexity of treatment regimens and incidence of dosing errors [7,8]. As a result, separate instillation times (by at least 5-10 min) are recommended to avoid diluting and/or washing of either drug from cul-de-sac [9]; this approach is very inconvenient for patients. Therefore, there is a significant clinical concern regarding reduction in the pharmacological effects of instilled drugs in patients who do not follow these recommendations. Fixed-dose combination (FDC) therapy is a solution for improving patient compliance [10]. Francis et al. indicated that FDC therapies provide better IOP control than unfixed combinations therapies in real-life settings [11].
In recent years, several FDCs of routinely-used IOP-lowering drugs have been developed [12], such as beta-adrenergic antagonist or alpha-2-adrenergic agonist with carbonic anhydrase inhibitor and PGA with beta-adrenergic antagonist, with the aim of maximizing patient medication adherence. Among the recently developed FDCs, those comprising a PGA and betaadrenergic antagonist (PGA/beta-FDCs) are attracting considerable attention because these therapies combine two different mechanism of action and are most commonly used to treat glaucoma, expecting strong IOP reduction [13]. Ocular surface disease (OSD) is a common complication of chronic treatment with preserved topical IOP-lowering drugs in patients with glaucoma or ocular hypertension. The reported prevalence of OSD in glaucoma patients varies between 45% and 60% [14][15][16]. FDCs decrease daily exposure to preservatives, such as benzalkonium chloride (BAK), which potentially have toxic effects on the ocular surface and increase treatment side effects [17]. Therefore, in addition to the medication adherence, FDCs are expected to reduce the possibility developing OSD.
An FDC of tafluprost and timolol (Taf/T-FDC) is a recently developed PGA/beta-FDC and is expected to become a new treatment choice for glaucoma. Taf/T-FDC demonstrated an almost equivalent IOP-lowering effect to concomitant instillation of 0.0015% tafluprost and 0.5% timolol in a recent clinical trial [18]. Latanoprost/timolol FDC (Lat/T-FDC), an FDC that combines 0.005% latanoprost and 0.5% timolol, was the first PGA/beta-FDC to be marketed worldwide; however, the IOP-lowering effect of Lat/T-FDC was considered lower than that of concomitant administration of timolol and latanoprost [19,20]. Diestelhorst et al. reported that a mean IOP at baseline was 16.9 mmHg in both study groups (Lat/T-FDC group and latanoprost/timolol unfixed combination group) and the mean diurnal IOP was 17.0 mmHg after Lat/T-FDC treatment and 15.9 mmHg after latanoprost/timolol unfixed combination therapy (P < 0.0001) [20]. Therefore, there is a possibility that the efficacy of Taf/T-FDC and Lat/ T-FDC is different; however, differences in the profile of efficacy and safety of these FDCs have yet to be fully elucidated.
In the present study, the IOP-lowering effect (strength and duration), ocular pharmacokinetics, and ocular surface toxicity of Taf/T-FDC and Lat/T-FDC were compared to clarify the characteristics of efficacy and safety of these two FDCs.
Timolol and tafluprost acid, used as standards in quantitative analyses, were provided by Santen Pharmaceutical Co., Ltd. (Osaka, Japan). Latanoprost acid was purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.).

Animals
Female Sprague Dawley rats (Charles River Laboratories Japan, Inc., Yokohama, Japan) weighing 155-214 g and male cynomolgus monkeys (Keari Co., Ltd., Osaka, Japan and Shin Nippon Biomedical Laboratories, Ltd, Tokyo, Japan) weighing 6.3-10.8 kg were used in the present study. Animals were housed under a 12-h light-dark cycle. All monkeys were housed individually, were provided 100 g of Monkey Bit (Nippon Nosan, Kanagawa, Japan) per day and water ad libitum, and received environmental enrichment including toys such as mirrors. They were monitored once daily by breeding stuff belonging to the Animal Care Team at Santen Pharmaceutical Co., Ltd. Breeding environment was set as follows: temperature and humidity were controlled at 23°C (acceptable range: 18-28°C) and 50% (acceptable range: 30%-70%), respectively. The monkeys used for IOP measurement were included following acclimations to prevent stress from measuring IOP; following was the acclimatization protocol: 1 week acclimation for sitting Monkey chair and 3-5 months tonometry acclimation. Local anesthesia was administered prior to measuring IOP. After the end of the study, all monkeys were continuously kept for use in other experiments. For sacrifice, the rats were euthanized by inhalation of isoflurane anesthesia (ISOFLURANE inhalation solution; Pfizer Inc., New York, NY, U.S.A.) and were exsanguinated from the abdominal aorta. No animals became ill, showed abnormalities in health conditions, or died prior to the experimental endpoint, therefore there were no animals that received medical treatment or humane euthanasia.
All animal care and experimental procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved and monitored by the Animal Care and Use Committee at Santen Pharmaceutical

IOP measurement
Ocular normotensive male cynomolgus monkeys were used in all studies. Before the study, all monkeys were trained in being restrained in a monkey chair (CL-4535; Primate Products, Miami, FL, U.S.A.) and undergoing IOP measurements without any general anesthesia or sedation. For IOP measurements, each monkey was kept in a sitting position in the monkey chair and IOP was measured with a pneumatonograph (Model 30 Classic TM Pneumatonometer; Reichert Technologies, NY, U.S.A.). For local corneal anesthesia, 0.4% oxybuprocaine solution (Benoxil 1 ophthalmic solution 0.4%; Santen Pharmaceutical Co., Ltd., Osaka, Japan) was topically applied before IOP measurements. One drop of oxybuprocaine was administered per eye per animal. We administered it twice before IOP measurement (3-20 min before and just before the measurement). All IOP measurement studies were performed under blind conditions.

IOP-lowering effects of Taf/T-FDC and Lat/T-FDC
Three independent studies were performed according to the schedules described in Fig 1. In all studies, drugs or saline were instilled to the right eye of each monkey between 10:00 AM and 11:00 AM, with the contralateral eye remaining untreated. IOP measurements were performed immediately before, and at 4 and 8 h after, drug instillation in study A; immediately before, and at 12, 14, 16, and 18 h after, drug instillation in study B; and immediately before, and at 24, 26, 28, and 30 h after, drug instillation in study C. Dark-phase IOP measurements were performed under a safelight so as to not influence the circadian rhythm.
For studies A and B, 18 monkeys were used with 6 monkeys assigned to each of three groups in such a way that IOP values immediately before the initiation of the study were approximately equal among the three groups. In regard to the use of 9 animals in study C, the monkeys were firstly divided into 3 groups based on IOP values at baseline. Saline, Taf/T-FDC, and Lat/ T-FDC were administered to each group, and IOP was measured at 24, 26, 28, and 30 h after treatment. Monkeys received each treatment in differing orders such that all monkeys received saline, Taf/T-FDC, and Lat/T-FDC. The mean changes in IOP were expressed as the means ± standard error of the mean (SEM).

Ocular penetration after ocular instillation of ophthalmic solutions
Thirty-six rats were used for the timolol ocular penetration study of Taf/T-FDC and Lat/ T-FDC. In each group, aqueous humor was collected at 5, 15, and 30 min, and 1, 2, and 4 h after instillation (n = 6 eyes per time point). Sixty rats were used for the ocular penetration study of FDCs and mono-preparations of tafluprost and latanoprost. In each group, aqueous Three independent IOP studies were performed for the comparison of the IOP-lowering effects of Taf/T-FDC and Lat/T-FDC in ocular normotensive monkeys. In all studies, drug instillations were performed at 10:00 AM to 11:00 AM. IOP measurements were performed immediately before, and at 4 and 8 h after drug instillation in study A, immediately before, and at 12, 14, 16, and 18 h after drug instillation in study B, and immediately before, and at 24, 26, 28, and 30 h after drug instillation in study C. Dark-phase IOP measurements were performed under a safelight. humor was collected at 5, 15, and 30 min, and 1 and 2 h after instillation (n = 6 eyes per time point).
Timolol, tafluprost acid (active metabolic form of tafluprost), and latanoprost acid (active metabolic form of latanoprost) in the collected aqueous humor were quantified by liquid chromatography (Nexera; Shimadzu Corporation, Kyoto, Japan) coupled with tandem mass spectrometry (QTRAP5500; AB Sciex, Framingham, MA, U.S.A.). All quantitative analyses were conducted using validated methods.
Maximum concentrations of timolol, tafluprost acid, and latanoprost acid in the aqueous humor (C max ) were calculated from mean concentrations. T max was defined as the time to reach C max . Areas under time-aqueous humor concentration curves up to 2-4 h after instillation (AUC 0-2h or AUC 0-4h ) were calculated from mean concentrations using the linear trapezoidal method. Elimination half-life (T 1/2 ) values were obtained by dividing the logarithm of 2 by the terminal elimination rate constant. AUC values were extrapolated to infinity by dividing the last measurable concentration by the terminal elimination rate constant. All parameters were determined using non-compartmental pharmacokinetic analyses (Phoenix WinNonlin Version 6.3; Certara LP, Princeton, NJ, U.S.A.).

Statistical analysis
All statistical analyses were performed using EXSUS software version 8.0.0 (CAC EXICARE Corporation, Tokyo, Japan). Tukey's multiple comparison test was used for comparisons among all groups in IOP study and in vitro cytotoxicity study.

IOP lowering effect of Taf/T-FDC versus Lat/T-FDC
Three independent studies were performed based on the schedules shown in Fig 1 in order to compare the IOP-lowering effects of Taf/T-FDC and Lat/T-FDC over early (0-8 h), middle (12-18 h), and late (24-30 h) terms. Changes in IOP are shown in Fig 2 (see data in S1 File). No statistically significant differences in initial IOP values were observed in any of three studies. The mean initial IOP values of each group were 16.7 mmHg to 17.6 mmHg in study A (n = 5-6), 17.9 mmHg to 18.1 mmHg in study B (n = 5-6), and 17.6 mmHg to 17.8 mmHg in study C (n = 9). During this study, no significant abnormalities in ocular condition or general health were observed. In study A and B, one monkey in the Lat/T-FDC group was excluded as accurate measurements could not be obtained from this animal.
Taf  (Fig 2A and S1 File). After appearance of the peak IOP-lowering effect, Taf/T-FDC group continued lowering IOP until 16 h after instillation. On the other hand, Lat/T-FDC group started to return at 12 h after instillation (Fig 2B and S1 File).
When comparing the duration of the IOP-lowering effect between the Taf  almost disappeared at 28 h after instillation; however, the IOP-lowering effect of Taf/T-FDC remained at 30 h after instillation.
The results of three studies demonstrated Taf/T-FDC had greater efficacy than Lat/T-FDC in terms of the duration of the IOP-lowering effect.

Drug penetration to anterior chamber after ocular instillation of Taf/ T-FDC and Lat/T-FDC
Ocular penetration of timolol was compared between Taf/T-FDC and Lat/T-FDC (Fig 3 and  S2 File). After instillation of Taf/T-FDC, timolol concentrations reached C max at 0.5 h and declined with a T 1/2 of 0.454 h. Concentrations were consistently higher after instillation of Taf/T-FDC than after instillation of Lat/T-FDC. The C max and AUC inf of timolol after instillation of Taf/T-FDC were 2.9 times and 3.2 times higher than after instillation of Lat/T-FDC, respectively (Table 1).
In addition, the ocular penetration of PGAs was compared between FDCs and mono-preparations. After instillation of FDCs and mono-preparations, the aqueous humor concentrations of acid forms, the pharmacologically active forms of PGA generated by hydrolysis in the cornea, were measured and compared (Fig 4 and S3 File). The pharmacokinetic parameters of acid forms are shown in Table 2  respectively) however, these decreases were significantly smaller than with Lat/T-FDC (Fig 5A  and S4 File).
Taf/T-FDC demonstrated significantly lower cytotoxicity than Lat/T-FDC at all evaluated time points and two concentrations.

Discussion
The strength and durability of the IOP-lowering effects of Lat/T-FDC and Taf/T-FDC were examined in the present study (Fig 2 and S1 File). In those experiments, Taf/T-FDC demonstrated greater IOP-lowering effect duration that was maintained until 30 h than Lat/T-FDC. These data demonstrated the superiority of Taf/T-FDC over Lat/T-FDC in terms of efficacy duration. Two factors that affect the duration of IOP-lowering effects were considered in the present study: the efficacy of each mono-preparation, tafluprost and latanoprost, and pharmacokinetics based on drug formulation. From the view point of the efficacy of each mono-preparation, it is thought that the difference in efficacy duration is attributable to the difference between the tafluprost and latanoprost compositions of Taf/T-FDC and Lat/T-FDC, respectively. A previous study [22] showed no significant difference between tafluprost and latanoprost in IOP reduction 24 h after a single instillation in monkeys. In the present study, concentrations of PGA acid forms in aqueous humor were measured after instillation of mono-preparations and FDCs containing tafluprost or latanoprost to examine the possibility that the difference of ocular penetration of PGAs between mono-preparation and FDCs caused the superior efficacy duration of Taf/T-FDC. The result showed there were no differences of  ocular penetrations of PGAs between mono-preparation and FDCs (Fig 4 and S3 File). Therefore, it is considered that the difference in efficacy duration between Taf/T-FDC and Lat/ T-FDC is unlikely to be attributable to the efficacy of tafluprost and latanoprost in FDCs.
From the view point of the pharmacokinetics based on drug formulation, Taf/T-FDC formulation was designed to maintain IOP-lowering effect when timolol instillation is changed from twice to once daily with FDC, on the bases that timolol ocular penetration increases with increased pH [23,24], and the stability of tafluprost decreases with increased pH. As a result, timolol penetration into the eye with Taf/T-FDC has been reported higher than that with timolol mono-preparation and close to that with timolol gel-forming formulation [24]. In the present study, we measured timolol concentrations in aqueous humor after instillation of FDCs to examine the difference between Taf/T-FDC and Lat/T-FDC in timolol ocular penetration. As a result, the timolol ocular penetration was found to be greater for Taf/T-FDC than Lat/T-FDC (Fig 3 and S2 File), and it is probably due to difference in pH between Taf/T-FDC and Lat/ T-FDC (6.7-7.2 and 5.8-6.2, respectively). These findings suggest that superior timolol penetration into the eye with Taf/T-FDC was possibly attribute to the long duration of the IOP-lowering effects with Taf/T-FDC.
In addition to pH, there are some formulations (e.g., gellan gum and sorbate formulation) that increase timolol concentration in aqueous humor. Higashiyama et al. showed that the ocular penetration of timolol was improved by the presence of sorbic acid in an ophthalmic solution and concluded that the improvement was caused by increasing the timolol distribution to corneal epithelium owing to the formation of more lipophilic ion pair complexes with sorbic acid [25]. Rozier et al. reported that the ocular penetration of timolol in an ophthalmic solution containing gellan gum was improved by the prolongation of precorneal residence time caused by gel formation in the conjunctival sac [26]. Once-daily timolol ophthalmic solutions  containing gellan gum (Timoptic-XE1) or sorbate (Istalol1) has been reported to be at least as effective as timolol maleate solution given twice daily [27][28][29]. These reports indicate the similar efficacy of once-daily and twice-daily formulations is attributable to higher timolol concentrations in the aqueous humor following instillation of once-daily formulation. Therefore, the greater duration of the IOP-lowering effect of Taf/T-FDC compared to Lat/T-FDC is likely to be due to a higher timolol penetration with Taf/T-FDC.
In fact, Taf/T-FDC was shown to have a comparable IOP-lowering effect to concomitant instillation of 0.5% timolol and 0.0015% tafluprost in a recent clinical trial [18]. On the other hand, Lat/T-FDC reportedly has a slightly weaker effect than concomitant instillation of timolol and latanoprost [19,20]. Consequently, Taf/T-FDC may be superior to Lat/T-FDC in terms of efficacy duration due to formulation improvement of appropriate pH setting, leading to improved efficacy.
Single instillation of Taf/T-FDC was found to lower IOP for greater than 30 h. IOP reduction is less affected by neglect of instillation for IOP-lowering drugs with greater efficacy duration. The prolonged duration of action indicates that an occasional missed dose may not induce substantial IOP variations. Such IOP fluctuations have been identified as a risk factor for glaucoma progression [30]. Therefore, the longer duration of IOP reduction by Taf/T-FDC may contribute to stable IOP control in patients and reduce the progression of glaucoma. The results of the present study demonstrated the superiority of Taf/T-FDC compared to Lat/ T-FDC, corroborating the findings of previous clinical trials [18,20,31,32].
We also compared the effect of the two FDCs, Taf/T-FDC and Lat/T-FDC, on the health of the ocular surface using in vitro cytotoxicity assays. Chronic use of topical anti-glaucoma drugs induces side effects, such as ocular inflammation, allergy, dry eye syndrome, and failure of filtration surgery [33][34][35], which can be caused by medication components such as additives, and there is a clinical need for IOP-lowering drugs without corneal toxicity or significantly reduced such effects. In the present study, Taf/T-FDC demonstrated significantly lower cytotoxicity than Lat/T-FDC when both undiluted drugs and 10-fold dilutions were used (Fig 5 and S4 File). Nelson reported the tear turnover rate of 0-5 minutes and 5-20 minutes estimated by the instillation of disodium fluorescein (DSF) and carboxyfluorescein (CF) [36]. We have calculated the area under the tear concentration-time curve (AUC) using the tear turnover rate reported by Nelson and tear volume [36]. When eye drops at a concentration of 100 are instilled, AUC inf are estimated to be 215 and 141 using tear turnover rate of DSF and CF, respectively. These values are equal to the AUC with 22-minute and 14-minute exposure, respectively, of 10-fold-diluted eye drops. Therefore, we set the assay concentration and exposure time for MTS assay. When 10-fold dilutions were used, Lat/T-FDC demonstrated a significantly higher cytotoxicity than Taf/T-FDC. We considered two factors that may affect the cytotoxicity; the toxicity of the active pharmaceutical ingredient (API) and the cytotoxicity of additives. In the previous study, Pauly et al. reported that cytotoxicity was caused not by latanoprost but by BAK, when comparing a BAK-preserved solution of latanoprost with a preservative-free latanoprost solution [37]. Therefore, preservatives are considered a major factor underlying the observed difference in cytotoxicity as Lat/T-FDC is preserved with a high concentration of BAK (0.02%). Taf/T-FDC includes lower concentrations of BAK (0.001%) than that of Lat/T-FDC, which reportedly do not affect cell viability [38], compared to Lat/T-FDC as demonstrated by the lower observed cytotoxicity of Taf/T-FDC compared to Lat/T-FDC. These findings indicate that Taf/T-FDC may be associated with lower ocular surface damage compared to Lat/T-FDC in clinical use, and superiority of Taf/T-FDC is thought to be associated in low BAK concentration of the formulation. It should be mentioned that this finding is based on in vitro assays. Further evaluations, including studies in animals, are needed to confirm its clinical relavence.
In conclusion, the results of both in vivo and in vitro studies demonstrate Taf/T-FDC as superior to Lat/T-FDC in terms of IOP-lowering effect duration and ocular surface toxicity. These results suggest that Taf/T-FDC is a promising FDC option in the treatment of glaucoma.