Dexamethasone diffusion across contact lenses is inhibited by Staphylococcus epidermidis biofilms in vitro

Abstract

Purpose

This study was designed to measure the impact of bacterial biofilms on diffusion of an ocular therapeutic through silicone hydrogel bandage lenses in vitro.

Methods

An assay was designed to study the passage of a commonly used steroid dexamethasone through the silicone hydrogel soft contact lenses. Diffused dexamethasone was measured using a spectrophotometer over a period of 18 hours and quantified using a standard curve. This assay was performed with control and Staphylococcus epidermidis biofilm-coated contact lenses composed of lotrafilcon A and methafilcon. Biofilms were formed in brain heart infusion broth supplemented with D-glucose.

Results

The presented data validate a simple in vitro model that can be used to measure penetration of a topical therapeutic through silicone hydrogel soft contact lenses. Using this model we measured a reduction of dexamethasone diffusion by up to 88% through S. epidermidis biofilm-coated silicon hydrogel lenses compared to control lenses.

Conclusions

The results of this in vitro study demonstrate that bacterial biofilms impede dexamethasone diffusion through silicon hydrogel contact lenses, and warrant future studies regarding the clinical benefit of using ocular therapeutics in the setting of bandage contact lens use for corneal epithelial defects.

Introduction

The use of contact lenses to protect the ocular surface is a common and well-established practice. With many clinical conditions such as large, non-healing corneal epithelial defects or corneal ulcers, topical medications are co-administered with bandage contact lens use ^1-3. In some situations, such as corneal transplant rejection with concurrent non-healing epithelial defects, use of both steroids and anti-infectives are instilled into eyes with bandage lenses ^4-7.

Biofilms are surface associated aggregations of microbes that are highly tolerant to antimicrobials and the immune system ⁸. Extended wear of contact lenses has been associated with the development of biofilms on lens surfaces ^9-12. Bacteria readily adhere to silicone hydrogel lenses ^9-12. In addition, mucus and protein accumulation occurs on lenses when worn for extended periods that are recalcitrant to cleaning ^13-15.

Staphylococcus epidermidis is a bacterium frequently isolated from the ocular surface and commonly isolated from contact lenses ^16-22. The ability to form biofilms is considered to be one of the key virulence factors of S. epidermidis ²³. This bacterium can cause ocular infections such as chronic blepharitis, conjunctivitis, endophthalmitis, and keratitis ^{16, 24-26}.

A glucocorticoid steroid dexamethasone is commonly prescribed to relieve inflammation, swelling, redness, and pain caused by chemicals, bacterial infection, and/or severe allergies ^{27, 28}. Many commonly prescribed ocular therapeutics are hydrophobic and similar in size to dexamethasone, making this steroid an ideal model compound to explore diffusion in bandage contact lenses ²⁷. While numerous references point to the use of contact lenses as a drug reservoir, it is not known whether the presence of such surface contaminants can impede the transmission of ocular therapeutics through soft bandage contact lenses ^29-31. This gap in knowledge prevents the rational selection and dosing of topical ocular medications when a bandage lens is used for corneal diseases. To begin to answer this question, we measured the diffusion of dexamethasone through two commonly used silicone hydrogel bandage lens materials in vitro.

Materials and Methods

Bacterial Strains and Growth Conditions

Staphylococcus epidermidis (ATCC 35984) cultures were grown overnight at 30°C in Lysogeny Broth (LB, per liter: 5 g yeast extract, 10 g tryptone, 5 g NaCl) in test tubes on a tissue culture rotor with aeration (New Brunswick model TC-7). Cultures were subcultured 1:10 into brain heart infusion medium containing 0.2% D-glucose (11 mM). To generate biofilms, contact lenses Air Optix^® Night and Day™ Aqua power -0.50, diameter, 13.8 mm, composed of lotrafilcon A (Ciba Vision, Deluth GA) and Kontur Precision Sphere lenses with a base curve of 8.90, PLANO sphere, and a diameter of 14.0 mm, composed of methafilcon (Kontur Kontact Lens Co, Inc, Hercules, CA) were added to the medium concave side up in a 12 well dish (Costar # 3513) and incubated at 37°C overnight. Control lenses, without bacteria, were also placed in brain heart infusion medium and incubated overnight at 37°C. To enumerate bacteria composing the biofilms, lenses were sonicated in 1 ml sterile PBS for a total of 3 × 5-second pulses with a CL-188 microtip using a Q Sonica ultrasonic processor (setting 22.5). A dilution series was plated on LB agar plates and incubated at 37°C overnight. Colony forming units were quantified. Experiments were conducted on at least two different days yielding similar results.

Contact Lens Diffusion Experiments

Biofilm-coated and control lenses composed of lotrafilcon A and methafilcon were placed concave side up in a 6-well tissue culture dish containing 0.45 μm filter inserts (Costar # 3412) (Figure 1A and B) with 1 ml of sterile phosphate buffered saline (PBS) below the insert. Forty microliters of 6 mM dexamethasone (Sigma product number D4902 dissolved in DMSO) was placed in the concave side of the biofilm and non-biofilm lenses and incubated overnight at room temperature. An aliquot (20 microliters) of the PBS solution below the insert was collected and absorbance was read at 239 nm the absorbance maxima for dexamethasone previously reported by ³² and verified under our experimental conditions using a Biotek Synergy 2 plate reader (Biotek). Dexamethasone concentration was determined using a standard curve. Experiments were also conducted with 40 μl of 6 mM fluorescein-conjugated dexamethasone (Sigma product number D1383 dissolved in DMSO) on lotrafilcon A (Air Optix^® Night and Day™ Aqua) lenses (+/- biofilm) and fluorescence was measured at 485 nm excitation and 515 nm emission with a Biotek Synergy 2 plate reader. Experiments were conducted in triplicate on at least two different days yielding similar results. Data was analyzed using Prism Software (GraphPad) using a two-tailed Student's T-test with significance set to P < 0.05.

Figure 1. Conceptual design of dexamethasone diffusion experiments.

A) Non-biofilm and B) biofilm covered lenses were placed on 0.45 μm filters into 6 well dishes. Dexamethasone was added to the concave side of the lens. PBS below the filter insert was collected and the concentration of dexamethasone that diffused into the chamber below the filter insert was calculated. C) Confocal image of a Staphylococcus epidermidis biofilm on a lotrafilcon A contact lens. Scale bar = 10 μm. Biofilms were identical on methafilcon contact lenses.

Microscopic Imaging of Biofilms

Biofilms were grown on lotrafilcon A and methafilcon contact lenses as described above and stained with 1.67 μM Syto-9, a fluorescent dye (Molecular Probes). Lotrafilcon A and methafilcon lenses were placed concave side down into MatTek (#P06G-1.5-20-F) glass bottom dishes. Lens biofilms were imaged at 60X magnification objective using an Olympus IX-81 inverted microscope equipped with an FV-1000 laser scanning confocal system (Olympus) with FluoView FV10-ASW 3.1 imaging software. The experiment was repeated on two days.

Results

Use of a Contact Lens Model to Measure Dexamethasone Diffusion

The model described above was used to measure diffusion of a therapeutic through commonly used bandage lens materials. Dexamethasone diffusion through lotrafilcon A contact lenses (Air Optix^® Night and Day™ Aqua) could be detected in vitro according to the method described in Figure 1. Absorbance readings (239 nm) for dexamethasone were taken at 1, 5, 10, 30 minutes and 1, 1.5, 2, 4 and 18 hours after addition of dexamethasone. Dexamethasone in this assay was not detectable until 18 hours after incubation (average of 1.09 mM with a standard deviation of 0.18). Similar results were found with contact lenses composed of methafilcon (Kontur Precision Sphere), with dexamethasone not detectable until 18 hours after incubation (average of 0.70 mM with a standard deviation of 0.0002). Results were nearly identical using fluorescein-conjugated dexamethasone (average of 0.71 mM with a standard deviation of 0.01).

Biofilm Formation on Lotrafilcon A and Methafilcon Lenses Impedes Passage of Dexamethasone

To determine the impact of biofilms on dexamethasone diffusion through bandage contact lenses, S. epidermidis biofilms were grown on lotrafilcon A lens to 3.0 × 10⁵ ± 3.4 × 10³ CFU per lens (n=4). Biofilms were also assessed using confocal microscopy after staining with the fluorescent dye Syto-9. Robust biofilms were observed on both lotrafilcon A and methafilcon lenses (Figure 1C and data not shown). Biofilm-coated and control lotrafilcon A lenses were tested and resulted in reduced diffusion of fluorescent and non-fluorescent dexamethasone through biofilm-coated lenses relative to control lenses (Figure 2A-B, Table 1).

Figure 2. Dexamethasone diffusion through lotrafilcon A contact lenses.

A) Concentration of fluorescent dexamethasone that diffused through lotrafilcon A contact lenses. p < 0.05 by Student's T-test. Fluor control n = 8 biofilm n = 6 Graph is from 3 independent experiments conducted on 3 different days B) Concentration of non-fluorescent dexamethasone that diffused through lotrafilcon A contact lenses. p < 0.05 by Student's t-test. Control n = 15 biofilm n = 14. Graph depicts mean and standard deviation from 5 independent experiments conducted on 5 different days.

Table 1. Dexamethasone diffusion through Lotrafilcon A (Air Optix Night and Day Aqua) and methafilcon (Kontur Precision Sphere) contact lenses.

Druga	CLb	No Biofilmc	Biofilmc	Biofilm-Percent Reduction	p Value	N
Dexamethasone	Lotrafilcon A	0.080 ± 0.055	0.032 ± 0.023	60	p = 0.0108	Control n = 15 Biofilm n = 14
Dexamethasone	Methafilcon	0.089 ± 0.015	0.036 ± 0.004	60	p < 0.0001	Control n = 10 Biofilm n = 6
Dexamethasone fluorescein	Lotrafilcon A	0.041 ± 0.025	0.005 ± 0.006	88	p < 0.0001	Control n = 8 Biofilm n = 6

^a)Dose 40 ul 6 mM Dexamethasone

^b)Contact lens material

^c)mM Dexamethasone Mean +/- standard deviation

To determine the impact of biofilms on a second commonly prescribed bandage lens type, S. epidermidis biofilms were grown on methafilcon lenses to 3.0 × 10⁷ ± 4.9 × 10⁶ CFU per lens (n = 3). Dexamethasone diffusion was nearly identical to experiments conducted with lotrafilcon A lenses (Figure 3, Table 1). Biofilm formation impeded dexamethasone passage through methafilcon lens in comparison to control lenses by 60% (p < 0.05).

Figure 3. Biofilms grown on methafilcon also impede dexamethasone diffusion.

p < 0.05 by student's T-test. Control n = 10, Biofilm n = 6. Graph depicts mean and standard deviation from 3 independent experiments conducted on three different days.

Discussion

Topical administration of ocular pharmaceuticals is convenient but results in low bioavailability. The ocular surface is a dynamic environment, and factors such as reflex tearing, blinking, tear film quality and surface characteristics all play a role in the residence time of topically administered drugs ^{33, 34}. It has been shown that only a small portion of an instilled dose of topical drops are retained in the cul-de-sac ³⁵. Actions of the tear film and tear pump reduce contact times of therapeutics with the cornea because they distribute the compound and remove it from the ocular milieu through dilution, dispersion and drainage ^{36, 37}. The amount of time for which the drug is able to interact with the target tissue is also limited, hence the extensive use of suspensions, emulsions and ointments to increase contact times ^{38, 39, 40}. Furthermore, contact lenses separate the tear film into pre-lens and post-lens environments. The post-lens tear film has reduced tear exchange and with topical drops on a bandage contact lens should have even less exposure to therapeutics ^{41, 42}. However, therapeutics that get into the post-lens tear film by diffusion through the lens or movement around the lens may reside there longer and result in increased diffusion into the cornea ⁴³.

In this study, dexamethasone penetration through the tested bandage lens types was slower than expected - a potential cause for concern regarding dosing frequency and duration for patients with bandage lenses. Experiments which explore the posterior lens concentration of dexamethasone after topical application will be needed, but even this in vitro study suggests that if one is attempting to reach the corneal surface using a solution formulation, that the most commonly used contact lens polymers will prevent the drug from reaching the target tissue.

It is commonly believed that administration of an anti-infective in the setting of extended lens use would protect against inflammatory events. However, Ozkan et al. showed that in spite of a reduction in bioburden, the number of inflammatory events in persons taking prophylactic antibiotics is not less than the control group ⁴⁴. To make matters worse, patients who are prescribed bandage lenses are instructed not to remove or clean them, sometimes for extended periods. The ensuing mucus accumulation forms a receptive and protective surface for bacterial attachment. Thus, in spite of chronic antibiotic use, biofilms readily form on extended wear lenses. In fact, Kaplan and colleagues recently showed that staphylococcal biofilm formation is stimulated by subinhibitory concentrations of several types of antibiotics ^{45, 46}. Our results indicated that even a simple single-species biofilm significantly increase the time required for dexamethasone to penetrate both tested lens materials by up to 88% (Table 1). A potential endpoint for the dexamethasone in our study is suggested by recent work by Parsek and colleagues who demonstrated the biofilm extracellular matrix is sufficient to limit penetration of the antibiotic tobramycin, although all other tested antibiotics quickly penetrated the biofilm matrix ⁴⁷. Bacteria have also demonstrated the capacity to metabolize and degradade steroids ^{48, 49}. In our studies, it is possible that dexamethasone could be sequestered in the biofilm matrix or converted or degraded by S. epidermidis in our biofilm coated lenses resulting in the observed reduction in the detectable concentration of dexamethasone through biofilm coated lenses.

Our study did not test the importance of protein and mucus co-contaminants on lenses, nor did it assess the ability of bacterial biofilms to inhibit diffusion of other ocular therapeutics. Future projects are underway which explore these issues, and additional work is required to understand how contact lenses affect the pharmacokinetics and bioavailability of topically delivered ocular medications in vivo.

The findings presented here using a commonly prescribed steroid and an in vitro model system suggest that biofilms on contact lenses may have a major impact on the efficacy of therapeutics, and that therapeutics may not readily diffuse through bandage contact lenses. Future work testing the impact of biofilms on diffusion of other therapeutics is warranted.