*3.1. Preservative-Free Drops*

Adverse drug reactions are a major barrier to adherence and persistence. Local side effects can vary from minor dry eye symptoms to allergic and toxic-inflammatory responses [20]. The use of preservatives, especially benzalkonium chloride (BAK), is a known cause of ocular surface disease (OSD) in patients taking topical IOP-lowering eye drops [21]. In a survey study in French glaucoma patients, 62% of the patients cited at least one OSD side effect and 19% of patients at least four such side effects [22]. The reported prevalence of OSD among glaucoma patients in other studies is similar, between 60% and 70%, which is much higher than in age-matched subjects without glaucoma (between 15% and 33%) [23–25]. Self-reported nonadherence over 9.4 years, defined as missing ≥5% of the prescribed eye drops, was reported by 30% of participants [26]. Individuals who experienced side effects reported significantly higher levels of nonadherence than those who did not (37.6% vs. 18.4%) [26]. The side effects of OSD were associated with a reduced quality of life and worsening of quality-of-life scores correlated with reduced adherence captured by a questionnaire [22]. Several studies reported that the severity and prevalence of OSD in glaucoma patients correlated with the number of preserved drops per day and duration of treatment [24,27,28].

The improvement of symptoms and signs of OSD after switching from preserved to preservative-free eye drops has been shown in many studies [29–33]. In a large prospective survey including 4107 participants, Pisella et al. reported that patients taking preserved eye drops had significantly more symptoms and signs of OSD than those taking preservativefree eye drops [29]. For patients experiencing more pronounced signs and symptoms of ocular irritation, a treatment change from preserved to preservative-free eye drops significantly decreased the prevalence of all symptoms and signs. Jaenen et al. in a crosssectional study including 9658 patients assessed subjective symptoms and signs of ocular irritation [30]. Each symptom and all the signs (blepharitis, eczema, hyperemia, and fluorescein staining) were significantly more frequent in patients taking preserved than in patients on preservative-free eye drops [30]. At the time of these two studies, most patients used preserved hypotensive drops, and the choice of preservative-free formulations on the market was limited. Following the launch of preservative-free tafluprost and a preservativefree fixed combination of timolol with dorzolamide, preservative-free latanoprost and bimatoprost and later their fixed combinations with timolol became available. Several studies have compared the efficacy and tolerability of preservative-free prostaglandin analogues and their fixed combinations to their preserved counterpart and found that preservative-free formulations are noninferior in their IOP-lowering effect and associated with less signs and symptoms of OSD. Patients with ocular symptoms or signs of OSD on preserved latanoprost (Xalatan®; Pfizer, New York, NY, USA) were switched to preservativefree tafluprost (Taflotan®; Santen Oy, Finland) which had similar IOP-lowering effect as preserved latanoprost but was better tolerated and resulted in a decrease in symptoms and signs, and improved quality of life and patients' satisfaction [31,34]. The same efficacy and better tolerability have been shown for a preservative-free latanoprost (Monopost®, Thea Pharmaceuticals, France) compared with preserved latanoprost (Xalatan®; Pfizer, New York, NY, USA) [35]. Pillunat et al. [33] evaluated in an open-label study efficacy and tolerability of preservative-free fixed combination of tafluprost/timolol (Taptiqom®; Santen Oy, Tampere, Finland) in 1157 patients. Preservative-free fixed combination lowered IOP in all subgroups of patients: treatment naïve, prior monotherapy and prior fixed combinations. At the final visit at 16 weeks, symptoms, and signs of OSD improved in patients with prior medical therapy and, using a simple questionnaire, 90% of patients rated treatment comfort as very good or good [33].

## *3.2. Sustained Drug Delivery Systems*

Development of sustained drug delivery vehicles has been an ongoing search to improve adherence among patients. These drug delivery systems may be applied onto the ocular surface (contact lenses, nanoparticles, microspheres, extraocular inserts), in the puncta (punctal plugs) or injected into the eye. Different novel drug delivery techniques are in different stages of development and only one, the intracameral bimatoprost SR ocular implant (Durysta®; Allergan, Irvine, CA, USA) has been recently approved by the FDA for sustained IOP reduction [36].

#### 3.2.1. Nanoparticles

To overcome the limitations of topical antiglaucoma medications, nanoparticulate (NP) delivery systems may improve solubility of the drug and corneal penetration, increase concentration at the target tissue, reduce irritation and systemic side effects and provide dose accuracy and sustained release of the drug [37]. Nanoparticles, tiny structures ranging from 1 to 100 nm in size, can bypass biological barriers and deliver drug to the target tissue. Nanoparticles are used in different shapes such as nanoemulsions, dendrimers, liposomes, nanospheres, hydrogels, nanocrystals, nanodiamonds, microspheres, niosomes, nanofibers, and nanocapsules [37]. Lipid and polymer nanoparticles are usually used to carry the drugs, isolate their contents from degradation and regulate their release. Several drugs using a NP delivery system are under investigation, but currently none has been approved for clinical use [38].

#### 3.2.2. Contact Lenses

The contact lens drug delivery system is also very attractive. As the bioavailability of drugs with drop instillations is very low, the incorporation of drugs into the contact lens matrix increases the drug residence time on the cornea and improves drug bioavailability by more than 50% compared to eye drops [39]. Achieving sustained or prolonged release of the drug from the contact lens allows for reduced frequency of drop instillation and potentially improved adherence. Different drug loading methods are used to incorporate drugs into polymeric support [39,40]. At present, most of the glaucoma drug-loaded therapeutic contact lenses are in the preclinical or clinical stages and data regarding safety, efficacy and pharmacokinetics are required [41].

#### 3.2.3. Extraocular Inserts

Ocular inserts of different forms and sizes are shaped to fit into the conjunctival fornices. These inserts increase the ocular surface contact time of the drug, improve its bioavailability and reduce the need for frequent drop instillation. Among the first approved ocular inserts, Ocusert™ was placed under the eyelid and released pilocarpine over one week [42]. Although it was effective in reducing IOP, its use was limited by device dislodgement and high cost. Ocusert™ is not available on the market since 1993.

Another insert, bimatoprost sustained-release fornix ring-type insert, is in the late stage of development for clinical use. The insert achieved IOP reduction over 6 months similar to timolol 0.5% BID drops, was safe and well tolerated [43]. In a 13 months openlabel extension study the bimatoprost ring showed good retention rate with a median IOP reduction of 4 mmHg (interquartile range 2–6) [44]. The most frequently reported adverse events from both studies were mucous eye discharge (16%), conjunctival hyperemia (14%) and punctate keratitis (12%).

#### 3.2.4. Punctal Delivery Systems

Different solutions, suspensions, emulsions, nanoparticle or microparticle or liposome suspensions can be loaded into the core of the plug [45]. The latanoprost punctal plug delivery system (Mati Therapeutics) was loaded with 70.5 μg of latanoprost per device. When two such plugs were inserted in the upper and lower puncta, the mean IOP was reduced by 5.7 mmHg (22.3%) from baseline after 4 weeks [46]. The latanoprost-loaded punctal plug was well tolerated, with tearing reported as the most frequent adverse event.

The travoprost punctal plug (OTX-TP, Ocular Therapeutics, Bedford, MA, USA) is a rod-shaped hydrogel rod that swells in the canalliculus, thus preventing extrusion. Travoprost is encapsulated in polylactic acid microparticles for sustained release to the tear film over 90 days [47]. In the double-masked phase 2b study (NCT02312544) comparing the safety and efficacy of sustained-release travoprost plug delivery to timolol eyedrops in patients with open-angle glaucoma or ocular hypertension, OTX-TP plugs reduced IOP by 4.5–5.7 mmHg, whereas timolol reduced IOP by 6.4 mmHg–7.6 mmHg. The superb efficacy of timolol eye drops is likely the effect of decreased wash-out through the nasolacrimal

ducts by inert punctal plugs. In an Asian population, sustained-release travoprost reduced IOP by 24% at 10 days, and by 15.6% at day 30 [48].

Among the major limitations of the punctal drug delivery system is that only low drug doses, typically required for potent drugs (e.g., prostaglandins and corticosteroids), can be embedded into the plug core matrix. At present the only punctal delivery system available on the market is dexamethasone 0.4 mg insert (Dextenza™, Ocular Therapeutix) approved by the FDA in 2018 for the treatment of ocular inflammation and pain following ophthalmic surgery.

#### 3.2.5. The Periocular Drug Delivery System

For the subconjunctival route, several delivery systems can be used such as implants, microspheres, nanospheres, liposomes, and gels [45]. Most of the studies were performed in rabbits by injecting a subconjunctival formulation of timolol, brimonidine, latanoprost and carbonic anhydrase inhibitors and achieved good IOP reduction without signs of inflammation for up to 90 days, depending on the delivery system used [49–54]. In a pilot study (NCT01987323) including six patients, a nanoliposome-based latanoprost delivery system was well tolerated, and in five out of six patients IOP reduction achieved at 3 months was as effective as previous reports of latanoprost ophthalmic solution [55]. A recently completed randomized trial including 80 participants (NCT02466399) comparing the efficacy of subconjunctival liposomal latanoprost injection to latanoprost ophthalmic solution reported a mean change in IOP at month 3 of −2.3 mmHg (SD 4.6) and of −6.4 mmHg (SD 2.9), respectively. Adverse events were reported in the liposomal latanoprost group only, with the most frequent being conjunctival hemorrhage (26.4%), foreign body sensation (17.0%), and conjunctival hyperemia (13.2%). To date, no subconjunctival delivery system has been approved, suggesting inherent delivery and efficiency limitations associated with these delivery systems.
