Efficacy of Nonpreserved Sodium Hyaluronate Artificial Tears in Dry Eye Disease Patients Treated with Prostaglandin Analogs for Primary Open-Angle Glaucoma: A Prospective, Nonrandomized, Open-Label Pilot Study

Purpose Dry eye disease (DED) can be triggered using preserved ophthalmic formulations or prostaglandin analogs. In this prospective, nonrandomized, open-label pilot study, we evaluated the efficacy of a 0.15% hyaluronic acid (HA) nonpreserved ophthalmic formulation in decreasing DED symptoms in patients with open-angle glaucoma treated with prostaglandin analogs. Methods 30 patients with DED receiving chronic treatment with prostaglandin analogs for primary open-angle glaucoma or ocular hypertension were administered ophthalmic formulations 3 times daily for 12 weeks. Foreign body sensation, burning, stinging, dryness, pain, frequency of symptoms, Ocular Surface Disease Index (OSDI), conjunctival hyperaemia, corneal fluorescein staining (CFS), tear film break-up time (TBUT), best-corrected visual acuity, Schirmer test results, and 25-item National Eye Institute Visual Function Questionnaire score between the baseline and 4 and 12 weeks were evaluated. Results The analysis shows that all primary endpoints improved; in particular, burning sensation and the frequency of symptoms after 4 and 12 weeks of treatment (p < 0.001) and dryness and pain after 12 weeks of treatment (p < 0.001 and p=0.03, respectively) were reduced significantly. Secondary outcomes confirmed the positive results, with a statistically significant change in the OSDI score and CFS between the baseline and 4 (p=0.02 and p < 0.001, respectively) or 12 weeks (both p < 0.001) and TBUT after 4 weeks (p=0.01). Conjunctival hyperaemia improved in both eyes in >90% of cases at 12 weeks of treatment. Conclusion The present study shows that the ophthalmic formulation containing 0.15% HA has a promising beneficial effect on reducing the signs and symptoms of DED in patients treated with prostaglandin analogs.


Introduction
Te ocular surface consists of a continuous epithelium hydrated by the tear flm, which also contains protective antimicrobial factors (e.g., defensins, immunoglobulin A, lactoferrin, and lysozyme). Tear flm stability is crucial for maintaining homeostasis of the ocular surface and is ensured in particular by the mucin-rich gel produced by epithelial cells [1]. In addition, goblet cells in the epithelium produce cytokines, epidermal growth factor, and retinoic acid, which together maintain immune tolerance [2]. When these protective mechanisms fail, tear defciency results in alterations in the tear flm and hyperosmolar stress, which lead to increased friction and mechanical irritation of the ocular surface [3][4][5]. In addition to these phenomena, activation of infammatory processes further increases ocular discomfort [6][7][8][9].
According to the 2017 International Dry Eye Workshop II report, dry eye disease (DED) is a multifactorial condition characterized by increased osmolarity of the tear flm and infammation of the ocular surface [10][11][12]. Te prevalence of DED has a wide range (5%-50%) and is estimated to be higher in women, with a tendency to increase with age [13]. Most of the symptoms associated with DED are nonspecifc and common to other ocular diseases and include redness, burning, stinging, foreign body sensation, pruritus, and, in some cases, photophobia [13]. Te clinical signs of ocular surface infammation are a loss of conjunctival goblet cells and corneal epitheliopathy [14]. Te course of the disease is persistent and characterized by an episodic pattern of symptoms (fares) [15].
Because of the enormous variability of clinical signs, there is no consensus on the diagnosis of DED; nevertheless, self-reported questionnaires, such as the Ocular Surface Disease Index (OSDI), are commonly used as tools for assessing the severity of the disease. In addition, other clinical tests are used by physicians, including the Schirmer test, tear break-up time (TBUT), corneal and conjunctival staining, and tear osmolarity [16].
Many causes underlie the occurrence of DED. Te most common are the use of contact lenses, refractive laser cataract surgery, and the use of topical formulations containing preservatives and prostaglandin (PG) analogs for glaucoma and ocular hypertension [17,18]. Recent studies suggest that the use of topical formulations containing benzalkonium chloride (BAK) as a preservative may have adverse efects on the ocular surface [19]. BAK is a quaternary ammonium compound used in a variety of formulations [20]; although BAK destabilizes cell membranes, leading to bacterial death, its efect is nonspecifc and may also afect mammalian cells, resulting in local side efects that are cumulative and become more severe with repeated exposure [21,22].
Instead, PG analogs have become the frst-line therapy for treating patients with glaucoma due to their efcacy in lowering intraocular pressure (IOP) [18]. While reducing IOP, PG analogs are associated with ocular side efects, such as a prominent feature of ocular irritation associated with dry eye disease and an increase in conjunctival hyperaemia [23]. A recent meta-analysis of glaucoma patients showed that the risk of conjunctival hyperaemia increases in patients treated with PG analogs compared to patients treated with other classes of drugs [18,24]. Conjunctival hyperaemia is thought to be caused by nitric oxide-mediated vasodilation in the conjunctiva. However, the relationship between PG analogs and ocular surface changes is complicated and remains unclear [18].
To date, treatment options for DED have been based on avoidance of triggering factors, such as cigarette smoking, adverse environments, and others, in conjunction with the use of topical nonpreserved formulations, such as artifcial tears and corticosteroids or cyclosporin A-based eye drops [11,25]. Due to the side efects of chronic use of preserved ophthalmic formulations, the commercial trend is increasingly toward preservative-free eye drops.
Hyaluronic acid (HA) is a linear polymer composed of N-acetyl-glucosamine and glucuronate units. Its use in ophthalmology has been studied since the early 1990s, and HA is known to increase tear flm stability by stimulating mucin production [26]. Consistent with this notion, existing studies suggest that HA is able to signifcantly alleviate the symptoms of DED and reduce ocular infammation [27,28]. Based on these fndings and aiming to keep on providing the literature with increasing clinical data, the present prospective, nonrandomized, open-label pilot study evaluated the efcacy of the formulation containing 0.15% sodium hyaluronate (as the main component), 0.2% Echinacea extract, and amino acids in improving DED symptoms.

Aim of the Study.
Te aim of this prospective, nonrandomized, open-label pilot study was to evaluate the effcacy of a topical HA-based formulation also containing amino acids and 0.2% Echinacea extract (Iridium A Free; Fidia Farmaceutici, Padova, Italy) in improving DED symptoms as an adjunctive treatment in patients with primary open-angle glaucoma or ocular hypertension undergoing treatment with PG analogs. Tis ophthalmic formulation is specifcally designed to protect the corneal epithelium and helps increase the biological defence of the tear flm by better stabilizing and preserving its properties. Te inclusion criteria were as follows: (1) an age of ≥18 years, (2) a diagnosis of primary open-angle glaucoma or ocular hypertension and current treatment with PG analogs as monotherapy or in fxed combination/association with betablockers for ≥6 months before enrolment, (3) DED symptoms defned by an Ocular Surface Disease Index (OSDI) score of ≥13 points, (4) a Schirmer test I result of ≥5 mm to avoid the inclusion of dry eye patients due to decreased tear production, and (5) conjunctival hyperaemia of ≥2. Patients were excluded if they (1) used artifcial tear substitutes in 2 weeks before the start of the study, (2) had a history of ocular trauma, (3) had an active ocular surface infection of any type, (4) had an ocular allergy, (5) had undergone ocular surgery within 30 days prior to enrolment, (6) had another concurrent eye disease associated with ocular surface infammation (e.g., pinguecula, pterygium, or corneal scarring associated with corneal irregularities), or (7) were pregnant or breastfeeding. We also excluded patients with DED linked to a systemic disease or therapeutic used to treat a systemic disease and patients with known hypersensitivity to any of the components of the study eye drops. Te study protocol was assessed and approved by the Internal Commission of the Clinic; the research was conducted in accordance with the Helsinki Declaration, and patients provided informed consent.

Treatment and Evaluations.
Patients, already undergoing treatment with PG analog therapy, were administered 1 drop of ancillary topical therapy containing HA, amino acids, and Echinacea 3 times daily for 12 consecutive weeks. Te time points considered were baseline, 4 (±1) weeks, and 12 (±1) weeks. At each visit, as per the TFOS DEWS II Diagnostic Methodology report [29], the parameters evaluated were (1) best-corrected visual acuity (BCVA) using a logMAR chart; (2) IOP by Goldmann applanation tonometry; (3) conjunctival hyperaemia measured on a 4-point scale (0 � none, 1 � mild, 2 � moderate, and 3 � severe); (4) tear flm break-up time (TBUT) measured after instillation of 1 drop of fuorescein sodium; specifcally, one single drop of balanced salt solution (BSS) was applied at the tip of fuorescein strips (AKti-fu fuorescein strips 1 mg sodium fuorescein in each strip; Aktive S.r.l., Italy) and then instilled into the inferior fornix of the patients' eye; patients were then instructed to blink normally for approximately three times and then to stop blinking while TBUT was measured; (5) corneal fuorescein staining (CFS) measured after TBUT according to the National Eye Institute/Industry (NEI) scoring system; (6) ocular surface symptoms using a 10-point visual analog scale (0-10 points) for foreign body sensation, burning, stinging, dryness, pain, and frequency of symptoms; (7) OSDI score; and (8) vision-related quality of life using the 25-item NEI Visual Function Questionnaire (NEI-VFQ-25). Te last parameter was evaluated only at the baseline and at 12 ± 1 weeks. All exams were performed in the same environmental settings to avoid potential DED evaluation bias [30]: thermostat-regulated room, dim room light, maximum slit-lamp illumination, same amount of fuorescein, and patients were all evaluated by the same observer. Questionnaires were administered before clinical tests.

Outcomes.
Te primary outcome of this study was the change in ocular surface infammatory symptoms for each item of the visual analog scale (foreign body sensation, burning, stinging, dryness, pain, and frequency of symptoms) after 4 and 12 weeks of treatment. Te secondary outcomes were the mean change in OSDI score, VFQ-NEI-25, conjunctival hyperaemia, CFS, TBUT, BCVA, and Schirmer test result between baseline, 4 weeks, and 12 weeks after the start of the study.

Sample Size and Statistical Analysis.
A medium clinically relevant efect size equal to −0.50 at 12 weeks for the dryness symptom has been considered primary outcome of this study; all other ocular surface DED symptoms of the visual analog scale, such as foreign body sensation, burning, pain, and frequency of symptoms, were also primary outcomes but were not considered for sample size analysis. A sample size of 27 data pairs achieved a minimum of 80% power to reject the null hypothesis of zero efect size at 12 weeks at a signifcance level (alpha) of 0.10 using the two-sided paired t-test. As a rule of thumb, an anticipated 10% dropout rate has been assumed, and thus, the minimum number of evaluable subjects included in the study was N � 30.
To select the most appropriate statistical analysis, a preliminary between-eye correlation analysis was performed for all eye-specifc outcomes; depending on the value of the Spearman correlation coefcient and visual inspection of between-eye scatterplots, the outcome data were analysed using the average of values from the right and left eyes, or on a per-eye basis, with the exception of symptom scores (VAS and OSDI) which were considered on a per-patient basis.
Continuous variables were summarized by count, mean, standard deviation, and/or interquartile range values by time and analysed with a repeated measures-mixed model. Categorical variables were instead analysed with repeated measures logistic regression analysis, using a subject identifer as a random efect. Regression parameters (β) of univariate and multivariate analyses were tabulated as point estimates along with standard errors and p values for comparisons between the follow-up and baseline. Multiplicity adjustment was performed using the Student maximum modulus method. Individual profles and boxplots were created for each primary outcome and for each domain of the NEI-VFQ-25 quality of life questionnaire. A normality test was performed for all variables using the Shapiro-Wilk test. All tests were two-tailed and considered signifcant at the 5% level. All analyses were conducted using SAS version 9.4 (SAS Institute, Cary, NC, USA).

Demographics.
Tirty patients were enrolled in this study; of these, 15 (50%) were female, and the mean age at the baseline was 64.2 years (standard deviation, 10.5 years). Participating patients had been treated for an average of 9.1 years with PG analogs or a fxed combination thereof in the form of ophthalmic drugs with preservatives for glaucoma. All patients were diagnosed with primary open-angle glaucoma or ocular hypertension, and all had ocular surface infammation and DED symptoms, such as foreign body sensation, burning, stinging, dryness, and pain at the baseline (mean OSDI value, 37.7 points). All patients included in the study completed the entire treatment period; demographic and baseline diagnostic and treatment details are reported in Table 1.

Correlation Analysis.
Te left and right eyes were signifcantly correlated for all primary and secondary outcomes (p < 0.001), with Spearman's correlation coefcients ranging from 0.12-0.75. A per-eye analysis was carried out only for conjunctival hyperaemia and BCVA score changes with respect to the baseline.

Secondary Outcomes.
Te OSDI score showed a monotonic and signifcant mean change between the baseline and 4 and 12 weeks. Te mean change scores were −7.8 points (p � 0.02) and −8.1 points (p � 0.04) at 4 and 12 weeks, respectively (Table 2). Interestingly, TBUT increased   (Table 2). A monotonic and signifcant decrease in CFS values by 2.1 at 4 weeks (p < 0.001) and 4.5 at 12 weeks (p < 0.001), respectively, was observed compared to the baseline ( Table 2). Te results of the Schirmer test showed an increase from the baseline to 12 weeks (from 13.6 to 15.7 mm), which was not clearly statistically signifcant (p � 0.05) ( Table 2). Due to their distributional properties, a per-eye analysis was performed for conjunctival hyperaemia and BCVA changes at the baseline, with results categorized as worsening, no change, or improvement. Conjunctival hyperaemia improved in both eyes at 12 weeks of treatment in >90% of cases, while BCVA did not change from the baseline for most patients. Indeed, BCVA improvement was observed in the left eye of only 1 (3.3%) patient and the right eyes of 2 (6.7%) patients at both 4 and 12 weeks, and BCVA was worse in the right eye of 1 patient (3.3%) at 4 and 12 weeks. Te composite visual quality of life score (NEI-VFQ-25) and corresponding subscales showed no signifcant changes, except for a decrease in eye pain between the baseline (60.8 points) and 12 weeks (67.1 points) (p � 0.02) (Figure 2). Another secondary fnding was a signifcant decrease by −2.0 mmHg in IOP between the baseline and 12 weeks (p < 0.001), which was included in the clinical parameters as a reference value to detect any clinical worsening or any infuence of the HAbased formulation on the efect of PG analog therapy.

Discussion
Te results of the present study support the efcacy of an ophthalmic formulation containing HA, amino acids, and Echinacea in the treatment of ocular symptoms associated with DED in patients undergoing treatment with PG analogs. Te present study examined several outcomes, all of which may be representative of ocular surface changes. Indeed, daily use of the study ophthalmic formulation resulted in rapid improvement in infammation-related symptoms and their frequency, with a signifcant decrease in burning sensation after only 4 weeks of treatment; foreign body sensation, dryness, and pain scores were also statistically signifcantly lower after 12 weeks of treatment than at the baseline. Te improvement in these symptoms is crucial, since lower levels of infammation help break or at least mitigate the typical vicious cycle of DED, in which infammation is not only caused by the ocular surface but also becomes a key factor in damage to the eye [31]. Tese positive outcomes were confrmed by the self-completed OSDI questionnaire, in which scores improved signifcantly after both 4 and 12 weeks of treatment. Changes in objective parameters, such as tear stability (TBUT test), which is severely impaired in DED patients and is one of the concomitant phenomena leading to ocular surface stress,also mirrored the clinical results; conjunctival hyperaemia and basal tear secretion (Schirmer test) data reinforced the observed improvement in questionnaire scores, although not with statistically signifcant diferences. Considering the chronic nature of the disease, it is important to highlight that most of the efects promoted by the HA-based ophthalmic formulation in this study were visible after 4 weeks of treatment and persisted throughout the study period. Tese efects can be attributed to the distinctive composition of the ophthalmic formulation studied, as similar results have been observed in several previous clinical studies with HA-based eye drops [32][33][34]. Indeed, Molina-Solana et al. conducted a prospective, single-arm longitudinal intervention study to evaluate the efcacy of a preservative-free artifcial tear containing 0.4% HA and found a signifcant improvement in signs and symptoms, such as hyperaemia, CFS, and OSDI score, after 1 week and 1 month of treatment [33]. Similar results were also obtained in the study by Sanchez-Gonzalez et al., who recorded an improvement in Schirmer test results, TBUT, and OSDI score after artifcial tears containing diferent concentrations of HA were administered [34]. Among others, Roberti et al. conducted a prospective, randomized, single-masked, parallel study to evaluate the efcacy of a preservative-free solution containing 0.4% HA and 0.5% taurine in glaucoma patients undergoing longterm treatment with preserved hypotensive therapy. Teir results showed that the formulation greatly improved the signs and symptoms associated with DED [35]. Finally, the safety and efcacy of HA-based artifcial tears were thoroughly investigated by Aragona et al. in a randomized, controlled, multicenter, 3-month study involving >460 patients [36]. An interesting point that emerged from this study is the possible synergistic efect of formulation components. Indeed, the formulation tested here was highly efective compared to those investigated in the aforementioned studies despite the lower concentration of HA [28]. Tus, such an efect could be due to the combination of HA with amino acids. Indeed, supplementation with amino acids, especially L-proline, L-lysine, L-glycine, and Lleucine, is known to support the metabolism of the corneal epithelium, which is damaged in DED patients [37]. In the present study, the efect of treatment with the HA-based study formulation on the patients' quality of life was also evaluated using the NEI-VFQ-25 questionnaire. Te results showed a signifcant improvement in the ocular pain score, further extending the clinical efectiveness of preparation. Although this domain was the only one in the questionnaire that showed a statistically signifcant improvement, the result was clinically relevant as pain is the main reason for the impairment in quality of life documented in DED patients [38]. However, these results are still preliminary due to the limited sample of the study, and a future, more in-depth analysis of quality of life involving diferent types of questionnaires would be of great interest. Interestingly, a statistically signifcant decrease in IOP was observed in patients. It seems that improving ocular surface health in glaucoma patients allows for a better control of IOP values. However, the change in IOP could be due to (i) better adherence to glaucoma treatment, which is common in patients participating in a scientifc study, (ii) better adherence to glaucoma treatment due to the patient's perceived improvement in symptoms of ocular discomfort, or (iii) treatment of the ocular surface disease that allows to reduce infammation, thus improving both ocular surface health and IOP values [33]. In future studies, it would be useful to extend the follow-up period to assess how long the efect of the study formulation lasts. Although HA-based eyedrops have been used for many years, longer observation would allow further detection and monitoring of potential adverse efects. In addition, the presence of a control group and, as mentioned above, a larger cohort of patients would be useful to better study the efects of treatment on the patients' quality of life. We acknowledge that the presence of a placebo efect might have infuenced the subjective results of our study given the lack of a control group [40]; however, this was meant to be a pilot study, and in addition, the relevant improvements in clinical signs seem to support the role of the studied supplementation in improving DED in patients treated with PG analogs. Despite the limitations mentioned above and its openlabel nature, the results of this study show that the benefcial efects of eye drops containing HA, amino acids, and Echinacea extract are rapid and persist throughout the treatment period (12 weeks). Future randomized controlled trials would be necessary to better defne the application regimen and extend the use of the ophthalmic formulation to other ocular conditions.

Data Availability
Te datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.