Ophthalmic Drug Delivery through Contact Lenses

Abstract

purpose. Currently available ophthalmic drug delivery systems are inefficient and may lead to side effects. To increase efficiency and reduce side effects, the authors propose disposable particle-laden soft contact lenses for ophthalmic drug delivery.

methods. The essential idea is to encapsulate the ophthalmic drug formulations in nanoparticles and to disperse these drug-laden particles in the lens material, such as poly-2-hydroxyethyl methacrylate (p-HEMA) hydrogels. The drug-laden p-HEMA hydrogels were synthesized by free radical solution polymerization of the monomers in presence of nanoparticles. The particle-laden hydrogels were characterized by light-transmission and electron microscopy studies. Release profiles of lidocaine, a model hydrophobic drug, were measured by UV-Vis spectrophotometry.

results. Microemulsions of hexadecane in water stabilized with a silica shell around the particles produced transparent hydrogels. Contact lenses made with particle-laden hydrogels released therapeutic levels of drug for a few days.

conclusions. Particle-laden hydrogels are promising candidates for ophthalmic drug delivery. They are transparent and can release drugs for extended periods. The drug delivery rates can be controlled by varying the loading of nanoparticles in the gel.

Topical delivery through eye drops, which accounts for approximately 90% of all ophthalmic formulations, is very inefficient and in some instances leads to serious side effects. ^{1 2} Only approximately 5% of the drug applied as drops penetrates the cornea and reaches the ocular tissue, whereas the rest is lost to other organs in the body. ³ On instillation, the drug mixes with the fluid present in the tear film where it has a short residence time of approximately 2 to 5 minutes. The drug not absorbed by the cornea is either absorbed by the conjunctiva or flows through the upper and the lower canaliculi into the lacrimal sac. ³ The drug containing tear fluid is carried from the lacrimal sac into the nasolacrimal duct, and there the drug is absorbed into the bloodstream. Absorption of drugs into the bloodstream leads to drug wastage, and more important, the presence of certain drugs in the bloodstream leads to undesirable side effects. For example, β-blockers such as timolol, used in the treatment of wide-angle glaucoma, have a deleterious effect on heart. ⁴ Furthermore, application of ophthalmic drugs as drops results in rapid variation in the drug delivery rates to the cornea, which limits the efficacy of therapeutic systems. ⁵ Thus, there is a need for new ophthalmic drug delivery systems that increase the residence time of the drug in the eye, thereby reducing wastage and eliminating side effects. ¹  

In this study we proposed to develop disposable soft contact lenses as a new vehicle for ophthalmic drug delivery. The essential idea was to encapsulate the ophthalmic drugs in nanoparticles and to disperse these drug-laden particles in the contact lens matrix (Fig. 1) . If the nanoparticle size and loading are sufficiently low (the exact value depends on the refractive index mismatch between the gel and the particles), the particle-loaded lens stays transparent. In this study, we focused on soft poly 2-hydroxyethyl methacrylate (p-HEMA) hydrogel lenses. The p-HEMA hydrogel matrix is synthesized by bulk or solution-free radical polymerization of HEMA monomers in the presence of a cross-linker such as ethylene glycol-di-methacrylate (EGDMA). ⁶ Addition of drug-laden particles in the polymerizing medium results in particle dispersion in the hydrogel matrix. If contact lenses made of this material are placed on the eye, the drug will diffuse from the particles, travel through the lens matrix, and enter the postlens tear film (POLTF), the thin tear film trapped between the cornea and the lens. In the presence of a lens, drug molecules would have a much longer residence time in the postlens tear film than the residence time of approximately 2 to 5 minutes that is the case with topical application of drugs as drops. ^{2 7 8} The longer residence time would presumably result in higher drug flux through the cornea and reduce the drug absorption into the blood stream through the conjunctiva or the nasolacrimal duct. In addition, due to the slow diffusion of the drug molecules through the particles and the lens matrix, drug-laden contact lenses can provide continuous drug release for extended periods. 

Most of the previous attempts at using contact lenses for ophthalmic drug delivery have focused on soaking the contact lens in a drug solution to load the drug. One of the recent studies focused on soaking the lens in eye drop solutions for 1 hour followed by insertion of the lens into the eye. ⁹ Researchers studied five different drugs and they concluded that the amounts of drug released by the lenses are lower or of the same order of magnitude as the amount of drug released by eye drops. In another recent study, researchers studied delivery of timolol maleate and brimonidine tartrate by absorbing it in the contact lens from a dilute solution of the drugs for approximately 3 hours. ¹⁰ They then applied the soaked lenses in patients’ eyes twice daily for periods of 30 minutes each. The contact lenses released most of the drug in this short time, and the treatment was successful in reducing IOP. 

There are several other studies and patents in which developers sought to use soft contact lenses as drug-release agents. In all these studies, drug entrapment was based on the soaking of contact lenses or similar devices in various ophthalmic drug solutions followed by insertion into the eye. ^{11 12 13 14 15 16 17 18 19 20 21 22} Although soaked contact lenses are perhaps more efficient drug delivery systems than eye drops, they have a number of limitations. First, the amount of drug that can be incorporated into the lens matrix by soaking is limited by the equilibrium solubility of the drug in the lens matrix, which is small for most hydrophobic drugs. Second, if the drug is incorporated into the matrix by soaking, the entire drug diffuses in a few hours. Thus, soaked contact lenses cannot provide slow and extended drug release. Also, it takes a few hours to load the lens with the drug from the aqueous solution, and the large fraction of the drug that is left in the solution is wasted. 

Nakada and Sugiyama ²³ developed a contact lens with a hollow cavity by bonding together two separate pieces of lens material. The compound was then soaked in the drug solution and inserted in the eye. Because the concentration of the drug in the cavity was the same as the concentration of the drug in the solution, the compound lens had the same limitations as the drug-soaked lens, and thus, such a lens could supply the drug for only a short period. Furthermore, oxygen and carbon dioxide permeabilities of the compound lens are smaller compared with a regular soft contact lens, due to the presence of two separate sheets of lens material, which may cause an edema in the corneal tissue. 

The crux of our idea is to design a nanocapsule that can entrap a large amount of drug and then disperse the capsules in the lens material during polymerization. Nanocapsules prevent the interaction of drug molecules with the polymerization mixture and also provide additional resistance to drug release. The drug must first diffuse through the nanoparticles and penetrate the particle surface to reach the hydrogel matrix. Thus, nanoparticle-laden lenses can deliver drugs at a slow rate for a long period. 

In this study we evaluated dispersions of drops of stabilized oil-in-water (O/W) microemulsions in p-HEMA gels. Microemulsions are thermodynamically stable isotropic dispersions of nanosized drops in water, stabilized by surfactants. An oil-water (O/W) microemulsion is an effective vehicle for encapsulating a hydrophobic drug because of its ability to dissolve the drug in the oil phase. 

Materials and Methods

HEMA monomer was purchased from Sigma-Aldrich (St. Louis, MO). Ethylene glycol dimethacrylate (EGDMA), azobis-iso-butrylonitrile (AIBN), Brij 97, and octadecyltrimethoxysilane (OTMS) from Aldrich Chemicals (Milwaukee, WI); and hexadecane and 1 N hydrochloric acid (HCl) from Fisher Scientific (Pittsburgh, PA). All other chemicals were of reagent grade, and all were used without further purification. 

Synthesis of the Lidocaine-Loaded Microemulsion

The model drug lidocaine was dissolved in hexadecane (2.6% wt/wt drug loading), and 0.12 g hexadecane loaded with the drug and 40 mg OTMS were mixed in 10 g water and then 1.5 g Brij 97 was added to the mixture. The microemulsion was formed by stirring the mixture at 1000 rpm and heating it at 60°C until the solution became clear. Next, 10 g of 1 N HCl was added to the microemulsion. Addition of HCl causes hydrolysis and condensation of OTMS, to form a silica shell around the particles. The hydrolysis and condensation reactions were performed at 60°C for 6 hours with continuous stirring. 

Synthesis of Particle-Loaded p-HEMA Gels

The particle loaded p-HEMA hydrogels were synthesized by free radical solution polymerization of the monomer with chemical initiation. EGDMA (37 μL) and HEMA (10 mL) were added to 7.4 mL of the microemulsion. The solution was degassed by bubbling nitrogen for 30 minutes. Next, 32 mg of AIBN was added to 25 mL of the polymerization mixture, and the mixture was poured between two glass plates that were separated from each other with 1-mm-thick soft plastic tubing. The polymerization reaction was performed in an oven at 60°C for 24 hours. 

Particle and Gel Characterization Studies

The microemulsions were characterized by light scattering to determine the particle size in a particle size analyzer (Zeta Plus; Brookhaven Instruments, Holtsville, NY). The transparency of the hydrogels was measured by light-transmittance studies in a spectrometer (Genesys 10 UV-Vis; ThermoSpectronic, Rochester, NY) at a visible wavelength of 600 nm. A scanning electron microscope (SEM; JSM6330F Field Emission; JEOL, Tokyo, Japan) was used to study the microstructure of the drug-laden hydrogels. The samples were kept overnight in a vacuum oven to remove any volatile component from the gel. The dried samples were cracked in liquid nitrogen, and the freshly exposed surfaces were studied by SEM. The lowest possible accelerating voltages were used in the experiments, and a very thin carbon coating was applied to prevent charging of the samples. Also, optical images at ×500 magnification were obtained by optical microscope (BX60; Olympus, Tokyo, Japan; with a Spot RT Digital Camera; Diagnostic Instruments, Sterling Heights, MI), both before and after the vacuum treatment, to determine whether any structural changes occurred during the vacuum drying. 

Drug-Release Studies

After gel synthesis, drug-release experiments were performed to establish that the trapped drug can diffuse from the particles. Although the drug lidocaine is hydrophobic, it has a small but finite solubility in water at the experimental pH of ∼6.5, and thus it diffuses from the nanoparticles and through the gel matrix into the water phase. The diffusion process stops when the concentrations in the beaker, in the gel, and in the particles reach equilibrium. In the drug-release experiments the samples were submerged in water, aliquots of water were withdrawn at various times, and the concentration of the drug in the aliquots was determined by measuring the absorbance at 270 nm by spectroscopy. We also prepared blank hydrogels that were identical with the samples with the exception that they did not contain any drug for use as the reference. The absorbance values in the experiments with the blanks were attributed to the diffusion of the unreacted monomer and some of the components of the nanoparticles, such as the surfactant and the oil. The concentration of the drug in the aqueous solution was calculated by determining the difference in the absorbance between the sample and the blank experiments. 

Results

Particle Size Measurements

The microemulsion was transparent and colorless, with a mean particle size of approximately 13 nm. It remained stable after 2 weeks of shelf storage. 

Transparency of Gels

The hydrogels synthesized with the silica-stabilized microemulsion had approximately 79% transmittance, whereas hydrogels synthesized with the same microemulsion but without the silica shell around the drops had a transmittance of 66%. This suggests that the presence of the silica shell enhances the stability of the microemulsion drops. These transmittances are similar to the 88% transmittance of pure p-HEMA hydrogels. The difference in transmittance of a pure p-HEMA and the nanoparticle-loaded p-HEMA hydrogels is smaller with contact lenses that are approximately 10 times thinner than the lenses that were used in transmittance measurements. 

Microstructure of Microemulsion-Laden Gels

Figure 2A shows an SEM image of a cross section of a pure p-HEMA hydrogel. The surface of a p-HEMA hydrogel is smooth, uniform, and nonporous. An SEM image of a hydrogel loaded with the nanoparticles (Fig. 2B) shows that its structure is almost like that of a p-HEMA hydrogel, which is expected in view of the high transparency obtained with this hydrogel. The similarities in the morphologies of particle-laden gel and the HEMA gel suggest that particles remain stable during polymerization. 

However, SEM images at higher magnifications showed that the particle-laden gels had two different kinds of domains. One type did not contain particles, and its morphology looked identical with pure p-HEMA, even at very high magnification. The other type looked rougher and showed the presence of isolated particles and some aggregates. Figure 3A is an image of a particle-containing region, and Figure 3B is an image of a region that does not contain any particles. The area fraction of the particles in the particle-containing regions is 1.3% ± 0.3%. Particles have the major and minor diameters of approximately 50 and 20 nm, respectively. 

Drug-Release Profiles

Drug Release from Microemulsion-Loaded Gels.

Figure 4 shows the drug released from a 1-mm thick hydrogel. The gel was fabricated by mixing a 0.55% (wt/wt) oil microemulsion, and the drug loading in the gel was 0.23 mg/g. At the end of the 10-day period, almost the entire drug initially introduced into the hydrogel was recovered. The experimental error in the data reported in 4 5 Figures 4 to 7 is ±7% to 8%. 

Effect of Particle Loading on Drug-Release Profiles.

To determine the effect of particle loading on the drug-release profiles, we increased the oil fraction in the microemulsion from 0.55% to 3%. Thus, the drug loading in the gel increased from 0.23 to 1.2 mg/g. The drug-release profiles from gels loaded with these two microemulsions are compared in Figure 5 . To facilitate easy comparison, the percentage of drug released is plotted for each gel as a function of time. The release profiles for the two gels were almost identical, and after approximately 8 days, equilibrium was established between the water, gel, and oil phases. 

Effect of Cross-Linking in Gels on Drug-Release Profiles.

To determine the effect of cross-linking on the drug-release profiles, the amount of cross-linker in the polymerizing mixture was changed. Figure 6 shows the release profiles of three different gels with different amounts of cross-linker. Each of these gels was loaded with the 0.55% (wt/wt) oil microemulsion, which corresponds to a drug loading of 0.23 mg/g. The weights of the cross-linker used in the experiments for 10 g of HEMA are listed on the curves. The release profiles were only marginally affected by the degree of cross-linking in the gel. 

Effect of the Silica Shell on Drug-Release Profiles.

In Figure 7 we compare the drug-release profiles from the gels loaded with the silica-stabilized microemulsions and the gels loaded with the microemulsion without the silica shell. The drug loading in the two gels were slightly different and are noted in the figure caption. Again, for ease of comparison the percentage of drug released is plotted as a function of time for the two gels. The release profiles of the two gels were almost identical, and almost the entire drug was released in approximately 7 to 8 days. 

Discussion

To synthesize particle-loaded gels for ophthalmic drug delivery, one must ensure that the gels are transparent and able to release drugs at therapeutic levels for extended periods of time. In this study we entrapped drug-loaded oil-in-water (O/W) microemulsions in p-HEMA gels. The mean particle size of the oil drops was smaller than 20 nm, and the gels were transparent. The particles segregated during polymerization and there were two types of domains in the gels. One type had particles and the other type was devoid of particles (Fig. 3) . The particles in Figure 3A appear elongated, which suggests that these particles were subjected to stresses either during the polymerization process or during the drying step that is needed to prepare the SEM samples. The particles shown in Figure 3A are separated from each other, but their area is more than the cross-sectional area of the drops in the microemulsion. The particles may have partially aggregated either after the addition of the monomer to the microemulsion or during the early stages of the polymerization. The mean area fraction of the particles in the particle-laden regions was approximately 1.3%, but the volume fraction of oil in the dried gel was only 0.33%. This suggests that approximately 0.33/1.3 × 100 ≈ 25% of the gel contained particles. 

The drug trapped in the nanoparticles was released when the gels were soaked in water. Almost the entire drug amount was released in approximately 7 to 8 days (Figs. 4 5 6 7) . There were two separate time scales for the drug release: an initial burst that released approximately 50% of the drug in the first few hours and then a much slower release over a time scale of a few days. The brief initial burst may be due to drug adsorbed on the surface of the nanoparticles and on the gel, and the long-term release perhaps arose from the drug that was trapped inside the oil drops. The drug-release profiles scaled with the total drug loading in the gel, which could be controlled by manipulating the particle loading in the gel (Fig. 5) . The degree of cross-linking did not significantly alter the drug-release profiles, presumably because the drug molecules were much smaller than the pores in the gel (Fig. 6) . Also the presence of the silica shell did not alter the drug-release profiles (Fig. 7) . This could be because the amount of OTMS added to the microemulsion could form only a partial silica shell. 

Our results show that 1-mm-thick hydrogels can be loaded with nanoparticles to deliver drug for several days. However, the drug-release rates from contact lenses in the eyes will be different from those from our gels into a well-stirred beaker. The usual starting dose of timolol which is a hydrophobic ophthalmic drug is 1 drop of 0.25% timolol maleate in the affected eye(s) twice daily. ⁴ Assuming a volume of 25 μL for each drop, the daily dosage of timolol is 0.125 mg each day. Only approximately 5% of this amount actually reaches the cornea. Thus, the dosage that must be delivered to the cornea is approximately 0.0063 mg each day. At a loading of 1.2 mg of drug per gram of gel, which is the loading in our gels for the 3% microemulsion, a contact lens can contain approximately 0.024 mg of drug. Thus, the lens contains enough drug to last approximately 4 days. However, a fraction of the drug released by the lens will still be lost due to tear drainage and absorption through the conjunctiva. Because the residence time of the drug in the postlens tear film is approximately 30 minutes, most of the drug released by the lens into the postlens tear film is taken up by the cornea, and almost all the drug released by the lens into the prelens tear film is lost. After each blink, the prelens tear film breaks up in approximately 1 to 3 seconds, depending on the type of the lens. Because each interblink period is approximately 10 seconds, the amount of drug released into the postlens tear film is approximately five times the amount released into the prelens tear film. Thus, a contact lens would deliver approximately five-sixths of the entrapped drug into the postlens tear film. This rough scaling shows that a particle-laden lens contains enough drug to last approximately 3 to 4 days. The drug loading in the gel can be further increased by maximizing the particle fraction and/or by choosing other types of nanoparticles. 

The drug-release studies and the SEM images show us that we are successful in entrapping drug-filled nanoparticle in p-HEMA hydrogel matrices. These microemulsion-laden gels have a transparency comparable to the pure p-HEMA gels. The amount of drug entrapped is enough to last approximately 3 to 4 days, which is also the period in which most of the drug is released by the lens (Figs. 4 5 6 7) . The main drawback with these particles is the exponentially decaying release rates. However, this release profile is still an improvement over the bolus dosage introduced in drops. 

The present study was only a pilot study. More detailed studies must be conducted to address a number of other questions, such as the effect of water fraction, type of contact lens material (ionic or neutral), and type of nanoparticles on the microstructure of the particle-laden gel and the drug-release profiles. Also, we should develop methodologies to increase drug loading, improve drug-release profiles, and disperse nanoparticles in silicone lenses, which, due to their higher oxygen permeability, are the preferred extended-wear lenses. Furthermore, more detailed modeling studies and in vivo studies should be performed with the particle-laden lenses to determine the fraction of drug that is absorbed by the cornea. We plan to perform these studies in the future, but our current pilot study has demonstrated the feasibility of synthesized particle-laden contact lens that can deliver drugs at a therapeutic dosage for a few days. 

 Supported by Engineering Research Center (ERC) for Particle Science and Technology at the University of Florida, National Science Foundation Grant EEC-94-02989, and the Industrial Partners of the ERC.

 Submitted for publication September 2, 2003; revised January 29, 2004; accepted March 9, 2004.

 Disclosure: D. Gulsen, None; A. Chauhan, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Schematic of the novel particle-laden lens inserted in the eye.

Figure 2.

 

SEM images of (A) a pure p-HEMA hydrogel and (B) hydrogel loaded with drug-laden microemulsion nanodroplets. Magnification, ×10,000.

Figure 3.

 

SEM images of a hydrogel loaded with nanoparticles. Images show different regions of the same hydrogel: (A) a region of the gel containing particles (B) a region with no particles. Magnification, ×60,000.

Figure 4.

 

Drug diffusion from a particle-laden hydrogel loaded with 0.22 mg of drug per gram of hydrogel into a beaker.

Figure 5.

 

Comparison of drug-release profiles for hydrogels with (a) 3% and (b) 0.55% oil loading in the microemulsion. The drug loading for the gels were (a) 1.2 and (b) 0.23 mg/g.

Figure 6.

 

Drug-release profile from a hydrogel loaded with 0.55% oil synthesized with different amounts of cross-linker.

Figure 7.

 

Comparison of drug-release profiles for hydrogels (a) with and (b) without a silica shell around the microemulsion drops. The drug loading for the gels were (a) 0.22 and (b) 0.18 mg/g.