Introduction

Glaucoma

Glaucoma is a leading cause of irreversible visual impairment and blindness worldwide [1,2,3]. Individuals with glaucoma were estimated to be 76.0–79.6 million in 2020 and this number may rise to over 111.8 million by 2040 [3, 4]. The global glaucoma prevalence in the population at the age of 40–80 was calculated to be approximately 3.54% [4, 5]. Glaucoma is known as a “silent thief of vision” because warning signs are usually subtle and symptoms only felt in the late stages when the visual field has already been compromised severely [5,6,7].

Glaucoma has been recognized as a multifactorial neurodegenerative disorder and its pathogenesis remains not fully elucidated [8, 9]. It is a group of diseases characterized by structural damage and loss of retinal nerve fibre layer (retinal ganglion cells (RGCs)) in pathology, and progressive defect of the visual field in clinical manifestation [5, 6, 10] (Fig. 1). Although many risk factors have been identified, such as ocular structural predisposition, increased intraocular pressure (IOP) is the only modifiable risk factor at present [11,12,13]. IOP is most often controlled by the daily dose of IOP-lowering eye drops [11, 14]. Current anti-glaucoma management is conducted in a stepwise fashion and starts with single topical hypotensive eye drops [11]. These eye drop medications typically lower the IOP through alteration of aqueous humour dynamics, either reducing its production (beta-blockers, alpha-agonists, carbonic anhydrase inhibitors) or increasing its outflow (pilocarpine, epinephrine, prostaglandin analogues (PGA)) [12, 13]. If initial monotherapies are not sufficient to control the IOP, multi-drug treatments, laser and/or surgical interventions are employed [11, 12]. On the other hand, the concept of “neuroprotection” (i.e. treatments independent of IOP reduction intending to prevent or delay RGCs and axonal death) has also received increasing attention, since the disruption of functional connectivity in the optic nerve is indicated in glaucoma pathophysiology [15,16,17]. Glaucomatous RGC damage is a multifactorial neurodegenerative process, whose possible mechanisms include but are not limited to the aggregation of misfolded proteins, neuroinflammation, oxidative stress, mitochondrial dysfunction, and neurotrophin support reduction [18,19,20,21]. Simple reduction and maintenance of IOP may not be sufficient to prevent the progressive loss of the visual field [9, 10, 22,23,24]. Neuroprotective strategies have shown promising treatment outcomes in animal models, many of which are under clinical trials, but none of them has been applied in clinical practice to date [22].

Fig. 1
figure 1

Progressive defect of the visual field and the loss of retinal nerve fibre layer in glaucoma. A Normal vision and vision in glaucoma patients. Patients usually experience blurry or missing spots in peripheral vision at early-stage disease. At nearly end-stage disease, only a central vision remains and “tunnel vision” is generated. B Visual field tests of glaucomatous left eyes show early (A), moderate (B), and severe (C) stages of functional loss. Reprinted from Ref. [6] with permission from Elsevier. C The ophthalmoscopic photograph of the retinal nerve fibre layer in healthy individuals (A) show a healthy retinal nerve fibre layer without any defect (red arrows). In patients with glaucoma (B), there are localised reduced reflexes of the retinal nerve fibre layer (light blue arrows), indicating the diminution of retinal nerve fibre layer. Reprinted from Ref. [5] with permission from Elsevier. D The optic disc of healthy individuals without glaucoma shows a normal shape of the neuroretinal rim, with its widest part in the inferior region (A). With glaucoma damage, the cup becomes deeper and larger, and the rim is much thinner than in the healthy optic disc (B). Reprinted from Ref. [5] with permission from Elsevier

Issues with current treatment regimens

Topical administration of IOP-lowering eye drops is a relatively non-invasive and simple route for drug delivery, which is the current gold standard for treatment [25, 26]. However, the efficacy of treatment is undermined by patients’ inadequate adherence to medication regimens and limited bioavailability of drugs to target sites [9, 12].

An ideal medication instillation requires the right timing, frequency, dose, and better accompanied with skills to prolong the preservation time on the eye surface (e.g. pressing the dacryocyst area after the instillation) [27,28,29]. However, objective studies have demonstrated poor patient adherence on average. In some cases, more than half of patients have deviated from their prescribed medication regimens [9, 30,31,32,33,34,35,36,37,38,39]. Common barriers to medication adherence include low self-efficacy, forgetfulness, and difficulties with eye drop administration [35]. Taking medications that require more than twice per day, taking adjunctive treatments, or undergoing changes of medications also seem to be the factors decreasing patient adherence [40,41,42,43]. Patient compliance may be optimized when applying monotherapy or electronic monitoring [9, 27, 44], but neither of them is feasible for each patient in a clinical setting at least for now.

Additionally, it is reported that over 60% of patients are struggling with self-administering eye drops [45, 46], and only 39% of patients can complete the instillation properly without touching the ocular surface [28, 45]. These findings have been confirmed by later studies in Asia: less than a half of the patients are able to administrate eye drops on their first attempt; no more than 0.05% of patients are aware of pressing the dacryocyst area after instillations; over 62% of patinets got contact with the ocular surface during the administration [47, 48]. Contact with the ocular surface during instillation contributes to the contamination of eye drop bottles, which is of particular concern in patients who have accepted glaucoma surgeries [13, 49, 50]. It is estimated that 19% of eye drops become contaminated within 8 weeks, and 29–40% for bottles used longer [49, 50].

It is also found that age-related factors (e.g. reduced cognition, arthritis, and paralysis) and poor eyesight are responsible for worse self-administration techniques [28, 47, 51], especially in identifying medications, squeezing drops from bottles, and checking whether drops are delivered [13, 28, 45]. Moreover, the financial burden and adverse effects (AEs) of life-long treatment may add more noncompliance to medical regimens as well [13, 29, 35, 52]. Adherence is critical for the stabilization of the visual field. Studies have shown that patients with 80% adherence are more likely to hinder visual field progression, while those with 21% adherence demonstrate progressive visual field defects [40, 53].

Bioavailability refers to the extent of drug absorption and is commonly described as the percentage of dose absorption [9]. Delivering drugs to intraocular target tissues through topically administered medications is a long-standing challenge due to the presence of anatomical (statics barriers, such as the cornea, blood-aqueous and blood-retinal barriers) and physiological (dynamic barriers, such as tear drainage, conjunctival blood and lymph flow) barriers of the human eyes [12, 54, 55] (Fig. 2). When medication is given topically as eye drops, anatomical barriers retard drug absorption into intraocular tissues and dynamic barriers rapidly drain the drug into the systemic circulation. Meanwhile, secondary factors, such as blinking, tear film turnover, and nasolacrimal drainage accelerate the elimination of the drug [26, 55]. It is estimated that only 10 μL of the instilled formulation remains on the ocular surface after a single eye blink [56], and almost all drug agents are eliminated from the ocular surface after 15–25 minutes [57, 58]. Eventually, only 5% at best of topically administered drug agents may overcome the hindrance and access the anterior segment, thus frequent administration is required [14, 25, 59,60,61,62,63,64,65,66]. These ocular barriers also contribute to the wax and wane drug effect before and after each administration of the eye drops [9]. Pulsatile drug concentrations may lead to IOP fluctuation at different time points of the day, which is likely to be a risk factor in glaucoma progression [9, 14, 67].

Fig. 2
figure 2

Eye structures and ocular barriers

Topical administration of IOP-lowering eye drops certainly remains the cornerstone of anti-glaucoma treatment [25, 26]. However, the aforementioned problems result in poor bioavailability of drugs and non-adherence of patients, which has urged researchers to focus on novel therapeutics with improved treatment efficacy. This need may be met through the employment of nanomedicine [12, 95,96,97]. For instance, silica NPs sized 15 nm display higher retinal cytotoxicity than 50 nm-sized ones in vitro and in vivo [98]. However, properties that may have negative effects in vivo are often what makes these materials attractive as drug carriers [85]. For example, cationic or small-sized carriers with superior abilities of disrupting the cell-lipid bilayer lead to a better interaction between drugs and target tissues at the cellular level [94,95,96,97]. Transfection efficiency will be improved when delivering genes [99, 100]. The balance between the desired capabilities and the accompanying potential negative effects should be addressed.

Physical stabilization

An ideal nanocarrier should have stable characteristics and not change dramatically after being administrated into living tissues. Take nanoparticles (NPs), the most common form of drug carriers, for examples. Small NPs tend to aggregate in vivo because they are unstable thermodynamically [25, 85]. This aggregation may lead to an extremely high accumulation of drugs at certain sites [85]. NPs also tend to adsorb plasma proteins onto the surface [85]. Hence, caution must be paid when performing an intravitreal injection of NPs, because blood-retina barrier impairment may occur during this procedure. Currently, transmission electron microscopy (TEM) is the commonly used strategy to observe the distribution and morphology of nanocarriers in living tissues [101]. For fluorescent-labelled nanocarriers, observation with fluorescent microscopes is also a viable alternative [102]. However, the aforementioned methods can only provide a general trend. There is still a huge gap concerning the exact behaviour of nanocarriers in an intraocular environment, especially for degradation and elimination [25, 103].

Proper sterilization techniques

Regardless of the forms or the materials used to deliver the drug cargo, the assembled DDSs should be sterile before the final administration. However, proper and convenient sterilization techniques have become a limiting requirement when develo** DDSs, as many sterilization methods have been shown to alter the physiochemical properties of carrier materials and drug molecules [25, 104, 105].

Ethylene oxide, gamma irradiation, and autoclaving are the most commonly used sterilization methods for pharmaceutical products and medical devices [106]. During autoclaved sterilization, high temperature and pressure frequently results in physical instability and aggregation of polymers [25, 105]. Gamma irradiation has been proven to be effective with some nanomaterials [107, 108], but free radicals produced in the process can induce structural changes and physical instability [109,110,111], especially when the loading agent is a protein [112]. Accelerated drug release from its carrier after gamma irradiation was also reported before [108].

Ultraviolet (UV) light and filtration are familiar and economical sterilization methods, but UV light may contribute to increased polymer wettability [113]. Filtration utilizing a 0.20–0.22 µm sterile film may be a practical method to expel contaminants without changing the physicochemical properties of nanomaterials [114, 115]. However, this strategy may not be applicable to NPs with larger sizes as they may experience entrapment inside the membrane [25]. It is also worth mentioning that adding antimicrobial agents to drug carriers can be very risky [25]. DDSs are typically designed to continuously release the loading drugs and remain in the eye for a relatively long time. Long-term application of antimicrobial agents such as benzalkonium chloride is associated with serious side effects [116,117,118].

Perhaps there is no universal sterilization process suitable for all nanosystems [115]. Utilizing different sterilization techniques for different components separately and completing manufacturing under aseptic conditions may be a practical way [25]. The sterilization strategy should be validated on a case-by-case basis [115].

Routes of administration

Intracameral delivery versus intravitreal delivery

A unique advantage of delivering drugs via intraocular routes is that ocular barriers are bypassed and drugs are immediately available at target sites, and consequently, bioavailability is improved [9]. The general approaches to drug delivery via intraocular routes are intracameral and intravitreal injections.

Intracameral injection is applied in present clinical practice for anaesthesia and ocular inflammation [61]. Researchers believe that this route may be suitable for delivering anti-glaucoma drugs as well. Intracameral delivery allows for direct contact between drug agents and anterior segment tissues involved in glaucoma pathology (e.g., the ciliary body, trabecular meshwork and uveoscleral outflow pathways), leading to the rapid increase and high concentration of drugs in the anterior chamber [9]. In this way, drug bioavailability is 100% and a much lower total dose of drugs is required compared with topical medications [9, 13, 61]. However, intracameral delivery is inefficient in delivering drugs to the retina [9]. The posterior segment of the eye is better targeted by the intravitreal route of administration [9, 119]. Intravitreal delivery refers to administrating drug solution/suspension into the vitreous humour via pars plana with a sterile needle. Hence, a higher concentration of drugs in the internal eye and more direct contact of drugs with the retinal ganglion cell layer and the optic nerve head can be achieved in this way [9, 61, 63]. Intravitreal injections may be more acceptable for patients since it has been routinely used for various ocular conditions, such as uveitis, neovascular age-related macular degeneration, and diabetic retinopathy [9, 120]. Certainly, there are complications for both approaches, especially with repeated injections, such as intraocular infection, endophthalmitis, cataract, retinal detachment and haemorrhage, corneal and scleral damage [9, 13, 121, 122]. In the study of intracameral implants using rabbit eyes, partial corneal opacification and neovascularization were observed [123]. Cautions must be paid no matter which route is used for administration.

From the pharmacokinetic point of view, intracamerally administrated drugs are predominantly concentrated in the anterior chamber and difficult to reach the retina [9]. Hence, the intracameral route may be more suitable for IOP-lowering treatments than neuroprotection targeting at the retina. In contrast, intravitreal drugs can be cleared both anteriorly and posteriorly due to their access to the ciliary body, aqueous humour outflow, and the retina [9, 124]. Thus, intravitreal delivery can be a viable route for both IOP reduction and RGCs neuroprotection. Nevertheless, the intravitreal route has not been widely explored in IOP control therapy. On the other hand, intravitreal lipophilic drugs tend to be cleared posteriorly via the retina-choroid circulation, while intravitreal hydrophilic drugs are more likely to be cleared anteriorly via the aqueous humour outflow [125,126,127]. In other words, the increase of drug lipophilicity reduces the extent of drugs entering into the anterior segment, resulting in a weaker hypotensive effect [9]. Therefore, treatment goal (IOP control or neuroprotection), routes of administration (intracameral delivery or intravitreal delivery), and physicochemical properties of drugs (the extent of lipophilicity and hydrophilicity) should be considered together when designing DDSs.

Tolerance of intracameral and intravitreal spaces

Drug-loaded nanocarriers are typically administrated into the eye in the form of suspension or as an implant. The volume of injection or the number and size of the implant should be compatible with the model eyes because the tolerance of external suspension/implants that can be administrated is not infinite. In current studies, the suspension is most used when the DDS is administrated intravitreally. The common solution used for dispersing drug-loaded particles to form a nano-formulation includes a balanced solution and an isotonic phosphate buffer solutions of pH 7.4 [105]. Implants are often seen during the use of intracameral delivery. Implants are generally delivered through an incision near the limbus, and typically, only one implant is administrated per eye.

The amount of suspension/implant required for treatment in vivo depends on the therapeutic window of the drug itself, the drug payload in carriers, and the in vivo release kinetics of the drug cargo from its carriers [105, 128]. The upper limit of the dose is limited by (1) the maximum volume/size that does not trigger a spike in IOP [129], and (2) the tolerance of intraocular concentration of the delivered drug and its products [130]. The former is generally determined by the species of animal models (Table 1); the latter is influenced by the solubility and the intraocular metabolism of the drug delivered, as well as its release pattern from the drug carrier [130, 131].

Table 1 Intraocular volumes of different species

Rats and rabbits are common model choices for studies on anti-glaucoma intraocular DDSs. A rat vitreous volume can be considered as approximately 20 µL and an intravitreal injection volume of less than 5 µL is generally considered to have a low risk of AEs [133]. The normal depth of the rabbit anterior chamber is about 2.08 mm [137] and an intracameral injection normally should be within the range of 50 µL [119]. In addition to the volume, the density of materials administrated should also cause no mechanical trauma or severe inflammatory response [130]. In animal studies using rats, intravitreal injection of poly (lactic-co-glycolic acid) (PLGA) microspheres greater than 0.5 mg may induce retinal stress and neuronal cell dysfunctions [138]; 2-µL mix-sized PLGA microspheres of intracameral delivery can form angular aggregation and cause the rise of IOP [139, 140].

Regarding implants, the compatible size and fitness of the implant within the anterior chamber structures are key factors for safety prediction since the implant tends to stay within the confines of the inferior angle after the administration. Otherwise, device migration or restriction, and anterior synechia are likely to happen [123], especially for narrow iridocorneal angles or angles with an anatomical obstruction such as scarring [9, 141, 142].

Feasibility of administration

Syringeability and injectability are two key factors that guarantee the administration of the prescribed dose of DDSs with minimal damage to ocular structures [105]. Syringeability means that DDS can pass and be withdrawn by needles, and the finer needles employed, the less invasiveness to the eyes. Injectability refers to the performance of the DDSs during the injection [105]. If clum** occurs, pseudoplastic polymers such as hyaluronic acid can be used to relieve the blockage and improve the syringeability and injectability [105, 143, 144].

Drug carriers with larger sizes typically have higher drug loading capacity and longer drug release duration [145, 146]. For DDSs as a form of suspension, extensive use of large-sized subjects results in poor injectability, such as the clum** of particles in the needle and more backflow from the injection site [145, 146]. For DDSs as an implant, larger-sized implants require greater access with severer invasiveness to complete the administration. In conclusion, a balance should be made between the loading capacity of the drug carrier and the feasibility of administration when designing a DDS.

Drug carriers

Factors affecting in vivo behaviours of drug carriers

Although the specific behaviour of drug carriers in an intraocular environment has not been elucidated, the size and surface charge of particles are believed to determine their intraocular performance [8, 104, 119]. The vitreous humour is an isotonic clear gel-like network mainly consisting of water (98–99%), hyaluronic acid, collagen and proteoglycans [119]. It has a loose structure with an estimated mesh size of 550 nm [147], making it difficult to act as a severe barrier for particle diffusion, but the increase of particle size reduces intravitreal mobility [104, 124, 148, 149]. From a different angle, restricted particles may be seen as a localized system that provides sustained drug delivery to the retina [104]. Small-sized particles typically possess better retinal cell uptake than large-sized particles, but too small particles may be cleared rapidly in vivo, resulting in unsustained drug release [8, 8]. Compared with a single-originated nano/micro system, a hybrid drug delivery system retains the advantages of its components, while minimizing its respective disadvantages. In addition, the entrapped NPs enlarge the total surface area for attracting drug agents [8]. For instance, when NPs possess relatively poor biocompatibility are incorporated with polymers with high biocompatibility, outer polymer matrixes may protect the embedded NPs and drug cargo in living tissues, consequently ameliorating the drug release profile and reducing the biotoxicity [8]. Another example is MSN, where the payload is easy to diffuse out of the porous channels before reaching the targeted sites from bare particles due to the open porous structure [165]. To protect the drug cargo from early release, Lyu et al. incorporated bevacizumab (BEV)-loaded MSNs into cyclosporine A-loaded PLGA-PEG-PLGA thermogel matrix [164]. In vitro BEV release study showed a burst release of BEV (about 77%) from BEV-loaded MSNs during the first 48 h, while only 33% of BEV was released from BEV-loaded MSNs embedding in thermogel during this period.

Smart stimuli-responsive system

The term “smart” refers to the ability of DDS to provide a controlled release of the drug cargo at the exact time and site required in response to stimuli [158]. The stimuli can be exogenous (e.g. temperature gradient, light, magnetic field, ultrasound, electric field), or endogenous (e.g. pH change, enzyme activity) [8, 158]. Smart stimuli-responsive delivery systems can provide precise site-specific delivery in a controllable manner with minimal side effects or toxicity, which remains challenging for conventional NPs [8]. In addition, programmed sequential release and multi-responsiveness can also be achieved when combining different stimuli-responsive components with NIM strategies [8, 158]. Versatile smart stimuli-responsive DDSs have been well-developed for various diseases (for reviews, refer to [8, 158, 245]), but with few studies on glaucoma.

Concluding remarks

Glaucoma is a sight-threatening disease affecting the all-age population worldwide. The major obstacles to glaucoma treatment with topical eye drops include the non-adherence of patients and limited bioavailability of medications, especially for a chronic disease that requires life-long treatment every single day. Nanomaterial-based drug delivery strategies hold great promise because they are powerful in achieving sustained release, target delivery, improved bioavailability, reduced side effects, and enhanced treatment efficacy. Despite promising prospects and expectations of intraocular drug delivery systems, there remain problems to be addressed, such as reliable and cost-effective scale-up production, safety and efficacy studies throughout their lifecycle in different intraocular environments, before regulatory authority approval and commercialization. To complete the successful bench-to-bedside translation, further extensive investigations are still required to answer the above-mentioned questions. With significant multidisciplinary research efforts, clinicians and patients can look forward to additional therapeutic options that may be available in the coming years.