Molecular Vision 2013; 19:1198-1210
Received 11 January 2012 | Accepted 28 May 2013 | Published 30 May 2013
1Nanomedicine and Drug Delivery Laboratory, Department of Pharmaceutical Sciences, University of Colorado Anschutz Medical Campus, Aurora, CO; 2Emory Eye Center, Emory University, Atlanta, GA
Correspondence to: Uday B. Kompella, Department of Pharmaceutical Sciences, University of Colorado Anschutz Medical Campus 12850 E. Montview Blvd., Aurora, CO 80045 Phone: (303) 724 4028, Fax: (303) 724-4666, email: email@example.com
Purpose: First, to determine the influence of drug lipophilicity (using eight beta-blockers) and molecular weight (using 4 kDa and 40 kDa fluoroscein isothiocyanate [FITC]-dextrans) on suprachoroidal delivery to the posterior segment of the eye by using a rabbit ex vivo eye model. Second, to determine whether drug distribution differs between the dosed and undosed side of the eye following suprachoroidal delivery. Third, to compare the suprachoroidal delivery of sodium fluorescein (NaF) with the intracameral and intravitreal routes by using noninvasive fluorophotometry.
Methods: Using a small hypodermic 26G needle (3/8”) with a short bevel (250 µm), location of the suprachoroidal injection in an ex vivo New Zealand white rabbit eye model was confirmed with India ink. Ocular tissue distribution of NaF (25 µl of 1.5 µg/ml) at 37 °C was monitored noninvasively using the Fluorotron MasterTM at 0, 1, and 3 h following suprachoroidal, intravitreal, or intracameral injections in ex vivo rabbit eyes. For assessing the influence of lipophilicity and molecular size, 25 µl of a mixture of eight beta-blockers (250 µg/ml each) or FITC-dextran (4 kDa and 40 kDa, 30 mg/ml) was injected into the suprachoroidal space of excised rabbit eyes and incubated at 37 °C. Eyes were incubated for 1 and 3 h, and frozen at the end of incubation. Ocular tissues were isolated in frozen condition. Beta-blocker and FITC-dextran levels in excised ocular tissue were measured by liquid chromatography–tandem mass spectrometry and spectrofluorometry, respectively.
Results: Histological sections of India ink-injected albino rabbit eye showed the localization of dye as a black line in the suprachoroidal space. Suprachoroidal injection of NaF showed signal localization to the choroid and retina at 1 and 3 h post injection when compared with intravitreal and intracameral injections. Drug delivery to the vitreous after suprachoroidal injection decreased with an increase in solute lipophilicity and molecular weight. With an increase in drug lipophilicity, drug levels in the choroid–retinal pigment epithelium (RPE) and retina generally increased with some exceptions. Beta-blockers and FITC-dextrans were localized more to the dosed side when compared to the opposite side of the sclera, choroid–RPE, retina, and vitreous. These differences were greater for FITC-dextrans as compared to the beta-blockers.
Conclusions: The suprachoroidal route of injection allows localized delivery to the choroid–RPE and retina for small as well as large molecules. Suprachoroidal drug delivery to the vitreous declines with an increase in drug lipophilicity and molecular weight. Drug delivery differs between the dosed and opposite sides following suprachoroidal injection, at least up to 3 h.
Age- and life style-related ocular disease, including age-related macular degeneration and diabetic retinopathy, are the major cause of visual impairment and blindness in the industrialized developed world . Recent changes in demographics indicate an increase in the aged population . According to the latest release by the United Nations, there were 759 million people aged above 60 in 2010, and this number is expected to increase to 2 billion by 2050 . Rapid progress in the biologic sciences is leading to the development of various small and large therapeutic molecules, including tyrosine kinase inhibitors, monoclonal antibodies, small interfering RNAs, and aptamers, to combat these back of the eye diseases. However, convenient and safe delivery of therapeutic agents to the target posterior ocular tissues remains a major pharmacotherapeutic challenge. Drug delivery to the target tissues, such as the choroid and retina, is hindered by unique anatomic and physiologic barriers, including the outer and inner blood–retinal barriers, and blood circulation that prevent the entry of foreign molecules . Development of drug delivery systems that are safe, minimally invasive, and effective in local delivery of drug to the target tissues in a therapeutically effective concentration is an unmet need for back of the eye drugs.
Conventional topical ophthalmic dosing is effective in delivery of drugs to the anterior segment ocular tissues such as the cornea, iris–ciliary body, and aqueous humor, but typically fails to deliver adequate quantities of therapeutic agents to the back of the eye. Topical eye drops result in less than 5% bioavailability in the anterior segment eye tissues and substantially lower delivery to the posterior segment eye tissues, including the retina [5,6]. Systemic delivery to the eye is restricted by the requirement of large doses of a drug, nonspecific distribution, increased systemic side effects, and limited delivery to the choroid–retina due to outer and inner blood–retinal barriers.
Current clinical methods for local drug delivery to the posterior segments are periocular (off-label) and intravitreal (approved for some products) injections . Although both of these methods are effective in delivery of drugs to the choroid and retina, they are associated with their own limitations. With periocular injection, when drug is placed adjacent to the sclera, the drug has to diffuse across the sclera and the choroid–RPE barriers before it reaches the retina. Although periocular routes are less invasive than intravitreal or suprachoroidal injections, significant amount of drug gets eliminated by the conjunctival and episcleral circulations . Permeability of drug across the sclera is anticipated to be passive, while that across the choroid–RPE may be passive or carrier/receptor mediated. Although the sclera is permeable for high molecular compounds , it offers about 40%–50% resistance for permeability of beta-blockers across the sclera–choroid–RPE in rabbits . The intravitreal route is an effective mode to localize drug in the vicinity of retinal tissue. The intravitreal route allows sustained drug delivery using implants, drug suspensions, and macromolecule solutions . However, intravitreal injection is invasive and carries the risk of retinal detachment and endophthalmitis .
Our long-term goal is to explore suprachoroidal injection as an alternative route for localized drug delivery to the posterior ocular tissues. Following suprachoroidal delivery or placement of drug molecules in the potential space between the sclera and the choroid–RPE, drug is expected to be exposed at high concentrations to the choroid followed by the retina. For the first time in 2002, Einhmal et al. showed the feasibility of suprachoroidal delivery in sustained ocular drug delivery in a rabbit model . Olsen et al. showed the safety and delivery ability of suprachoroidal injection for triamcinolone acetonide and bevacizumab by using microcannulation in pigs [13,14]. Recently Patel et al. showed the applicability of glass microneedles for suprachoroidal delivery of micro- and nanoparticles in rabbit, pig, and human eyes . Although these previous studies have shown the application of suprachoroidal injection for intraocular drug delivery, there is no report available evaluating the effect of solute physicochemical properties, such as lipophilicity and molecular weight, on suprachoroidal drug delivery. Thus, the first objective of this study was to determine the effect of drug physicochemical properties, such as lipophilicity (beta-blockers) and molecular weight (fluoroscein isothiocyanate [FITC]-dextran), on suprachoroidal delivery in ex vivo rabbit eyes. Since other local routes of delivery, such as the transscleral route, are known to distribute drug more on the dosed side and less on the opposite side , the second objective of this study was to evaluate the differences in regional distribution of drug molecules after suprachoroidal injection. Since intravitreal injections are commonly used and because intracamerally injected materials drain via the uveoscleral pathway , which includes the choroid region, the third objective of this study was to compare suprachoroidal injection with intracameral and intravitreal injections for drug delivery to the retina and choroid region.
Atenolol (99.0%), sotalol hydrochloride (~98%), nadolol (~98%), pindolol (98%), timolol maleate (98%), metoprolol tartrate (99%), betaxolol (~98%), labetalol hydrochloride (99%), propranolol hydrochloride (99%), triethyl amine (99.5%), sodium hydroxide, formic acid (88%), sodium fluorescein (NaF), ammonium formate (99.9%), and 4 kDa and 40 kDa FITC-dextrans were purchased from Sigma-Aldrich (St. Louis, MO). High performance liquid chromatography-grade acetonitrile and methanol were purchased from Fisher Scientific (Fair Lawn, NJ). All chemicals and reagents used in this study were of analytic reagent grade.
New Zealand white rabbit eyes were obtained from Pel Freez Biologicals (Rogers, AR) within 24 h of harvesting and shipped to the laboratory in Hank’s balanced salt solution (Catalogue # H1641, Sigma Aldrich, St Louis, MO) on wet ice. Eyes were used immediately after arrival. For the suprachoroidal injection, excised rabbit eyes were placed on a ceramic tile and held with forceps to avoid movement of the eye during injection. Suprachoroidal injections were performed using a 50 μl Hamilton syringe (Hamilton Company, Reno, NV) and a small hypodermic needle (26G, 3/8” in length, 250-µm bevel) commonly used for intradermal injections (BD Biosciences, Franklin Lake, NJ; Product no. 305110). The needle was inserted at a 45° angle through the conjunctiva and sclera, 5 to 7 mm posterior from the limbus, in the posterior direction with the bevel facing downward. A small bevel oriented inward allows delivery of a solution to the suprachoroidal space with minimal leakage in the sclera. When the needle was in the suprachoroidal space, there was little resistance to the injection, whereas there was significant resistance to solution injection when the needle was in the sclera. Once the needle was in the suprachoroidal space, 25 µl of drug solution was injected. The accuracy of the injection technique was validated using an injection of India ink in the suprachoroidal space and dissection of the eyes immediately after injection. Prior pilot studies indicated that increasing the volume of injection to 50 µl resulted in an obvious leakage of the drug solution back through the needle track. Immediately after injection, each eye was placed in 1 ml of assay buffer (122 mM NaCl, 25 mM NaHCO3, 1.2 mM MgSO4, 0.4M K2HPO4, 1.4 mM CaCl2, 10 mM HEPES, and 10 mM glucose) in a 12-well tissue culture plate maintained at 37 °C in a shaking water bath. Eyes were incubated vertically for 1 or 3 h with the cornea facing upward. During incubation, the eye surface was kept moist by applying 30 microliters of assay buffer every 30 min. Although this study used bevel facing downward to reduce egress of drug solution during suprachoroidal injection, for in vivo studies, bevel facing upward towards sclera might be safer in preventing any choroid blood vessel rupture.
To visualize the spread of the injected solution and the accuracy of injection techniques, histological examination of the suprachoroidal injection of India ink was performed in albino rabbit eyes. We used the hypodermic needle to inject 25 µl of 5% India ink in the suprachoroidal space. Eyes were transferred to Davidson’s fixative  immediately after suprachoroidal injection. After 48 h in Davidson’s fixative, eyes were embedded in paraffin and 5-µm sections were obtained. Sections were further stained with hematoxylin and eosin (H & E). Pigmented rabbit and albino rabbit eye sections without India ink injections were used as controls. Cryosections (5 µm thick) were obtained and stained with hematoxylin and eosin (H&E). The stained sections were examined with a light microscope (Zeiss Axioscope 2; Carl Zeiss, Inc. Jena, Germany), and images were captured with an AxioCamMrC5 camera (Carl Zeiss, Inc.) using 10X and 20X objectives.
Ocular tissue distribution of NaF in ex vivo rabbit eyes after the suprachoroidal, intravitreal, and intracameral injection was monitored noninvasively with the Fluorotron MasterTM (OcuMetrics, Mountain View, CA). Ex vivo rabbit eyes were injected with 25 µl of NaF (1.5 µg/ml) either by suprachoroidal, intravitreal, or intracameral injections. Fluorotron scans were acquired as described previously . Fluorotron Master reports sodium fluorescein or its equivalent concentrations (ng/ml) at 149 data points representing various distances along the visual axis in the posterior to anterior direction of the eye. Rabbit eyes were held on a specially designed adjustable stand in such a way that the cornea faced the fluorophotometer lens. Fluorotron scans were acquired immediately after injection and at 1 and 3 h post injection to study the distribution of NaF in ex vivo rabbit eyes. Blank albino rabbit eye Fluorotron scans were acquired as baseline spectra. Fluorotron scans were acquired for the whole eye and assigned to four different regions: choroid–retina, vitreous, aqueous humor, and cornea as described previously . In this study, the location of fluorescence peak immediately after intracameral injection was assigned to aqueous humor (between data points 105 and 115). The peak corresponding to intravitreal injection between data points 55 and 80 was assigned to vitreous humor. The fluorescence peak between data points 20 and 30 was assigned to choroid-retina.
To evaluate the effect of solute molecular weight on suprachoroidal delivery, ex vivo rabbit eyes (n=4) were injected with 25 µl of a 30-mg/ml solution of FITC-dextrans (4 or 40 kDa) in PBS (8.5 gm/l NaCl, 0.95 gm/l Na2HPO4, and 0.455 gm/l NaH2PO4) in the suprachoroidal space of rabbit eyes (0.75 mg of FITC-dextran per eye). To evaluate the effect of solute lipophilicity on suprachoroidal delivery, ex vivo rabbit eyes (n=4) were injected with 25 µl of a beta-blocker cocktail (250-µg/ml solution of each beta-blocker, i.e., 6.25 µg of each beta-blocker per eye) in PBS in the suprachoroidal space. Immediately after injection, each eye was placed in 1 ml of assay buffer in a 12-well tissue culture plate maintained at 37 °C in a shaking water bath. Eyes were incubated vertically for 1 or 3 h with the cornea facing upward. At the end of incubation, the eyes were snap frozen in a dry ice–isopentane bath and stored at −80 °C until further processing.
Ex vivo rabbit eyes injected with FITC-dextran and beta-blockers were dissected in frozen condition to isolate various ocular tissues. All dissection procedures were performed on a cooled ceramic tile placed on a dry ice–isopentane bath (<0 °C) to avoid thawing of the eye during dissection. After separation of the anterior part of the eye, the remaining posterior globe was cut longitudinally along the visual axis into two parts: the injected side and the opposite side. The retina, choroid, vitreous, and sclera were collected from the injected as well as opposite side. A new surgical blade was used for each eye. To prevent the transfer of drug between the tissues of each eye, surgical accessories were thoroughly rinsed with saline followed by methanol followed by saline and blotted dry. All samples were weighed and stored at −80 °C until further processing.
Concentrations of FITC-dextran (4 and 40 kDa) in rabbit ocular tissues were measured using a spectrofluorometer after aqueous extraction of drugs from the tissues. Briefly, weighed amounts of ocular tissues were mixed with 1 ml of PBS (pH 7.4) in a 2-ml microcentrifuge tube, vortexed for 10 min, and then homogenized using a hand homogenizer in an ice bath. The aqueous layer was separated from the tissue matrix by centrifugation at 10,000 × g for 10 min, and the separated aqueous layer (100 µl) was transferred into clear 96-well plates. Fluorescence was measured using a spectrofluorometer (SpectraMax M5; Molecular Devices, Sunnyvale, CA) at excitation and emission wavelengths of 485 and 525 nm, respectively. Blank rabbit ocular tissues were processed similarly to actual tissue samples in order to measure tissue background.
Concentrations of beta-blockers in rabbit ocular tissue samples were measured by means of liquid chromatography–tandem mass spectrometry as described previously . An API-3000 triple quadrupole mass spectrometer (Applied Biosystems, Foster City, CA) coupled with a PerkinElmer series-200 liquid chromatography (PerkinElmer, Inc., Waltham, MA) system was used for analysis. The concentration of beta-blockers in rabbit ocular tissue samples was measured after liquid–liquid extraction using a dichloromethane and ethyl acetate (1:1 volume [v]/v) mixture. The calibration curve for analysis was prepared using blank rabbit ocular tissues samples as the matrix. Blank rabbit ocular tissue samples without beta-blocker injections were processed and analyzed to measure the tissue background. The mobile phase was a mixture of 5 mM ammonium formate adjusted to pH 3.5 (A) and acetonitrile–methanol (75:25) containing 0.02% triethylamine, adjusted to pH 3.5 (B). The linear gradient elution at a flow rate of 0.4 ml/min with a total run time of 13 min was as follows: 90% A (0–1.0 min), 10% A (6–9 min), and 90% A (10–13 min). The following transitions were monitored: 276/145 (atenolol); 273/255 (sotalol); 310/254 (nadolol); 249/116 (pindolol); 317/261 (timolol); 268/133 (metoprolol); 308/116 (betaxolol); 260/116 (propranolol); and 327/162 (labetalol).
The data were presented as mean ± standard deviation (SD). For statistical comparisons between injected versus opposite side tissues, the dependent samples Student t test was used. Comparison between 4 kDa and 40 kDa FITC-dextran was performed using the independent sample Student t test. Comparison of the mean between multiple beta-blockers was performed using one-way analysis of variance followed by Tukey’s post hoc analysis (SPSS, ver. 11.5; IBM, Armonk, NY). The results were considered statistically significant at p<0.05.
Histological sectioning of India ink-injected albino rabbit eye after the suprachoroidal injections was performed to confirm the accuracy of the suprachoroidal injection technique. India ink was used as a dye of choice because of its black color, which is easy to visualize in the albino rabbit eye that is devoid of melanin pigment. A histological cross-section of the India ink-injected albino rabbit eye showed a black line of India ink (Figure 1C,D) between the sclera and choroid–RPE, and a collaged light microscopic image showed the insertion of the needle through the sclera into the suprachoroidal space without any damage or penetration into the choroid–retina, thus confirming the accuracy of the injection. The injection site in the sclera was filled with India ink (Figure 1), and no black stain was observed in the sclera (except the injection site) or choroid–RPE, confirming that the injection specifically occurred in the suprachoroidal space. An albino rabbit eye without any injection of India ink was used as a negative control, and no black color was observed in any part of this eye. A histological cross-section of the albino rabbit eye showed both sclera–choroid and retina with a pink color and the choroid was attached to the sclera without any gap between these tissues. (Figure 1A).
Readers may misinterpret the black color from India ink as melanin pigment in the choroid–RPE. Therefore, for comparative purposes pigmented rabbit eyes were included as a positive control to show the appearance of pigment in the choroid–RPE. Histological cross-section of a pigmented rabbit eye showed the distribution of melanin pigment throughout the choroid–RPE (Figure 1, middle panel), which is distinct in its pattern from the India ink-injection.
Ocular tissue distribution of NaF after suprachoroidal injection was compared with intravitreal and intracameral injections in ex vivo rabbit eyes using ocular fluorophotometry. A baseline spectrum for the albino rabbit eye without NaF injection was included as a control. As shown in Figure 2, tissue background showed negligible readings, with fluorescence intensity below 8 ng/ml. Immediately after intracameral and intravitreal injections of NaF, sharp peaks were observed in the anterior segment and vitreous, respectively. After suprachoroidal injection, a broad dispersed peak of NaF was observed posterior to the vitreous in the sclera–choroid–retina region of the eye. Due to low spatial resolution (980 µm) of the Fluorotron Master fluorophotometer, it was difficult to distinguish the thin ocular tissues, i.e., choroid from retina and sclera. A time course of Fluorotron scans of NaF in ex vivo rabbit eyes showed that after intracameral and intravitreal injections, the peak intensity rapidly diminished over 3 h due to dilution or distribution of the drug to adjacent tissues. With intravitreal injection, the NaF peak broadened during the 3-h study due to the spread of the signal within the vitreous followed by distribution of NaF to the anterior and posterior segment ocular tissues. With the suprachoroidal injection, the peak intensity remained unchanged for 3 h, possibly due to adsorption or binding of NaF to the sclera–choroid–retina and lack of vascular clearance mechanisms in the ex vivo model.
Ocular tissue distribution of FITC-dextrans after suprachoroidal injection in ex vivo rabbit eyes was assessed to evaluate the effect of solute molecular weight on tissue distribution. Blank rabbit ocular tissues showed no interference at the excitation and emission wavelengths employed. As shown in Figure 3, retinal and vitreal (total tissue) delivery was significantly higher for 4 kDa FITC-dextran than 40 kDa FITC-dextran at both 1 and 3 h. Sclera and choroid–RPE levels were higher for 40 kDa FITC-dextran than 4 kDa FITC-dextran because of the low diffusivity of 40 kDa FITC-dextran (Figure 3).
Both 4 and 40 kDa FITC-dextran showed regional differences in ocular tissue distribution. Drug levels in the injected side sclera, choroid–RPE, retina, and vitreous were several fold higher than in the opposite side (Figure 4 and Table 1). These differences in the regional distribution of FITC-dextran decreased significantly at 3 h as compared to 1 h, possibly due to convection/diffusion of the drug to the opposite side as well as adjacent tissues over a given incubation time. Drug levels were higher in the sclera and choroid–RPE and lower in the retina and vitreous. Further, drug levels were higher in tissues adjacent to the site of injection than in those in the anterior segment.
As the suprachoroidal space is closer to the exterior surface of the eye, diffusion of FITC-dextran from the suprachoroidal space into the incubation medium was measured. As shown in Figure 5, both 4 and 40 kDa FITC-dextran showed continuous entry into the incubation medium. The 3 h cumulative percentage release of FITC-dextran into the medium was 4.0% of the administered dose for 4 kDa FITC-dextran and 2.9% for 40 kDa FITC-dextran.
The effect of solute lipophilicity on ocular tissue distribution after suprachoroidal injection was evaluated in ex vivo rabbit eyes using a cocktail of beta-blockers. As shown in Figure 6, choroidal and retinal delivery of beta-blockers increased with an increase in solute lipophilicity, with tissue levels being the highest for lipophilic beta-blocker propranolol and the least for hydrophilic beta-blocker atenolol. For vitreous, solute lipophilicity showed an inverse relation with the vitreal concentration, with an increase in solute lipophilicity resulting in a decrease in vitreal concentration (Figure 6E). Correlation of beta-blocker lipophilicity (Log D, pH 7.4) showed a linear positive correlation with choroidal and retinal concentrations, with the correlation coefficient (R2) being 0.753 or greater for both the choroid–RPE and retina (Figure 6B,D). In the vitreous, solute lipophilicity showed an inverse correlation with vitreal concentrations with R2 ≥ 0.923 (Figure 6F). These differences were more evident at 3 h compared to 1 h.
All beta-blockers showed regional differences in distribution after suprachoroidal injection in posterior ocular tissues. Drug levels on the injected side of posterior ocular tissues were higher than the opposite side tissues (Figure 7). For beta-blockers, tissue concentrations were 1.5- to 7.9-fold higher on the injected side choroid–RPE, retina, and vitreous than the opposite side (Figure 7). These differences in the regional distribution of beta-blockers decreased at 3 h compared to 1 h. Beta-blocker levels were lower in the anterior segment tissues than in the posterior segment tissues.
In the current study we successfully showed the feasibility of suprachoroidal injection in ex vivo rabbit eyes by using a simple short-bevel hypodermic needle and syringe. We observed that suprachoroidal injection results in enhanced exposure of NaF to the choroid and retina compared to intracameral and intravitreal injections. Drug physiochemical properties, including molecular weight and lipophilicity, influenced suprachoroidal drug delivery. Delivery to the vitreous after suprachoroidal injection decreased with an increase in molecular weight and lipophilicity. Suprachoroidal injection of beta-blockers and FITC-dextran showed regional differences in distribution in a given tissue, with drug levels being higher on the dosed side than the opposite side for the choroid, retina, and the vitreous. Differences in regional distribution decreased with an increase in exposure time.
The first objective of this study was to show the feasibility of suprachoroidal injection with a short bevel hypodermic needle and syringe. Previous studies showed the use of the potential space between the sclera and choroid for sustained drug delivery to the posterior ocular tissues [12,13]. However, these earlier studies employed a surgical procedure to insert a cannula or a specially designed microneedle for suprachoroidal injections [12-15]. Insertion of a cannula in the suprachoroidal space is too complicated a procedure to perform routinely in the laboratory setting and requires surgical work. Although suprachoroidal injection using microneedles is more user friendly, this approach needs special tailoring of needles for different species, and no commercial source is available for microneedles intended for the rabbit. In this study we showed the feasibility of suprachoroidal injection using a simple needle and syringe assembly. We used a small tapering head 26G hypodermic needle (3/8”) with an ultra short bevel (250 µm); this allows injection of material specifically in the suprachoroidal space with little leakage. Accuracy of the injection technique was confirmed by histological sectioning of the albino rabbit eyes immediately after suprachoroidal injection of India ink. These histological sections showed a black line of India ink in the space between the sclera and choroid–RPE, without any black stain in the sclera and choroid–RPE (Figure 1). The mean particle size of India ink particles was 150.73±0.76 nm, and these were retained at the site of injection without further movement across the tissue barriers. India ink particles cannot readily move across intact tissue barriers, and hence they are used as a marker to test the corneal barrier integrity after cataract surgery . Injection of drug material in the suprachoroidal space results in expansion of the suprachoroidal space between the sclera and choroid–RPE (Figure 1). Seiler et al.  and Kim et al.  also reported the expansion of the suprachoroidal space to accommodate the injection volume. A similar expansion of the suprachoroidal space was seen after suprachoroidal injection of particles ex vivo in pig eyes .
The second objective of our study was to compare ocular tissue distribution of NaF in ex vivo albino rabbit eyes after suprachoroidal, intravitreal, and intracameral injections. As shown in Figure 2, suprachoroidal injection of NaF sustained signal intensity for 3 h. However, intravitreal and intracameral injections resulted in a steep decline in signal intensity within 3 h. We injected 37.5 ng of NaF (25 µl of 1.5 µg/ml) in the suprachoroidal space. Such a small amount of NaF might largely remain bound to the tissue at the site of injection, accounting for the sustained signal intensity over the duration of the study. Previous reports indicated that NaF binds to sclera with a maximum binding capacity of 80±5 mM . In the case of intravitreal and intracameral injections, NaF was injected in the ocular fluid matrix, which is composed of 99% water. This allows rapid diffusion of NaF to adjacent tissues, resulting in a rapid decrease in signal intensity within 3 h. In ex vivo rabbit eyes, suprachoroidal injection of NaF showed sustained signal intensity in the choroid–retina region for 3 h. However, NaF signal intensity is expected to be influenced by choroidal and retinal blood clearance in the live animal. Kim et al. showed rapid clearance of gadolinium diethylene triamine pentacetate (Gd-DTPA) from the suprachoroidal space within an hour, with an elimination half-life of 11.8 min . The suprachoroidal injection of NaF showed a broad peak in the posterior ocular tissues; this can be explained by the focal diamond theory of fluorophotometry . As noted by McCarey, the suprachoroidal injection of NaF showed a broad peak in posterior ocular tissues possibly due to the halation of the choroid–retina response . Halation or secondary fluorescence occurs when light passing through the choroid–retina is reflected back by the choroid and scattered beyond its boundaries. This causes the fluorescence to bleed through, resulting in tailing of the choroid–retina response .
The third objective of our study was to study the effect of drug physicochemical properties, such as solute molecular weight and lipophilicity (log D), on ocular tissue distribution after suprachoroidal injection. To evaluate the effect of molecular weight, we used 4 and 40 kDa FITC-dextran. Different molecular weights of FITC-dextran were previously used to evaluate the effect of solute molecular weight on permeability across the sclera as well as the choroid–RPE [9,28]. Solute molecular weight showed a significant influence on retinal and vitreal delivery of FITC-dextran after suprachoroidal injection. The 40 kDa FITC-dextran showed significantly lower retinal and vitreal levels than the 4 kDa FITC-dextran (Figure 3). Pitkanen et al. reported that the permeability of 40 kDa FITC-dextran was fivefold lower than 4 kDa FITC-dextran across bovine choroid–RPE. Possibly due to lower diffusivity and permeability, 40 kDa FITC-dextran was retained to a greater extent at the site of injection, resulting in significantly higher levels in the sclera and choroid–RPE compared to 4 kDa FITC-dextran. Because of its lower diffusivity, 40 kDa FITC-dextran showed 1.38-fold lower egress into the incubation medium than 4 kDa FITC-dextran (Figure 5). We have not observed any significant effect of molecular weight on tissue distribution in the anterior segment tissues, such as cornea, aqueous humor, and conjunctiva. This might be due to the low solute levels observed in anterior segment tissues (15- to1500-fold lower than choroid–RPE levels). At such low levels, assay variability is high, making it difficult to discern any molecular weight-related differences. Solute levels in the anterior segment ocular tissues were most likely due to diffusion in conjunction with convection along the suprachoroidal space or needle track. Despite the difference in molecular weight and the extent of delivery, FITC-dextran (both 4 and 40 kDa) showed similar trends in tissue distribution after suprachoroidal injection. Drug distribution in posterior ocular tissues was in the following order: choroid–RPE~sclera>retina>vitreous. For anterior segment tissues, the general trend was conjunctiva>iris–ciliary body~cornea>aqueous humor.
The effect of solute lipophilicity on suprachoroidal delivery was evaluated using a cassette of beta-blockers with a broad range of lipophilicity . Beta-blockers are the most commonly used series of molecules to evaluate the effect of solute lipophilicity because of their narrow range of molecular weight and pKa and wide range of lipophilicity . As shown in Figure 6, solute lipophilicity influenced drug distribution to the choroid–RPE, retina, and vitreous after suprachoroidal delivery. Total drug delivery to the choroid–RPE and retina increased with an increase in solute lipophilicity, whereas vitreal delivery decreased with an increase in lipophilicity (Figure 6). Our previous results with in vitro tissue partitioning of beta-blockers in the bovine choroid–RPE and retina also showed that tissue partitioning of beta-blockers increases with an increase in lipophilicity . As shown in Figure 6E,F, solute lipophilicity showed an inverse correlation with vitreal delivery of beta-blockers after suprachoroidal injection. Reduced vitreal delivery for lipophilic beta-blockers is possibly due to retention of lipophilic molecules in the choroid–RPE and retina and low partitioning into the vitreous. We previously observed a decrease in sclera–choroid–RPE transport with an increase in solute lipophilicity due to accumulation of lipophilic drugs in the sclera–choroid–RPE .
The last objective of our study was to evaluate the differences in the regional distribution of drug molecules after suprachoroidal injection. We observed that drug levels on the injected side of the choroid, retina, and vitreous were significantly higher than the opposite side for both high and low molecular weight compounds (Figure 4 and Figure 7). Higher drug levels on the injected side tissue were due to a rapid uptake of the injected drug by tissues near to the site of injection and then slow diffusion to the adjacent tissues. The upward orientation of the eye during incubation was also a possible contributor to regional differences in solute distribution. The orientation of eyes in this study favors convection of the drug solution toward the posterior pole of the eye. Subconjunctival injection of FITC-dextran (70 kDa) in mice showed regional differences in distribution, with the injected-side sclera and choroid levels being significantly higher than the opposite side at initial time points . Differences in regional tissue distribution decrease over time due to the diffusion of molecules to tissues on the other side. Previous reports of the suprachoroidal injection of a contrast agent in ex vivo porcine eyes also indicated regional differences in distribution, with mean pixel intensity of the contrast agent being significantly higher on the injected side than the opposite side . In our study, these differences were higher for FITC-dextran than beta-blockers. Lower diffusivity of macromolecules might contribute to greater differences between dosed and opposite sides (Figure 4 and Figure 7). These differences decrease with an increase in exposure duration (Table 1).
One of the limitations of the current study is that we used ex vivo rabbit eyes, which were devoid of blood circulation and other clearance mechanisms. The main objective of this study, however, was to show the effect of drug physicochemical properties on ocular tissue distribution after suprachoroidal injection in the absence of vascular clearance mechanisms. In our previous study we observed a good correlation between in vitro tissue partitioning and in vivo delivery for beta-blockers. Further, in vitro transport across isolated tissues, which was indirectly measured in the present study, is commonly used to predict in vivo delivery. Another limitation of the current study is that the eyes were used 18–24 h post harvesting. The time gap between enucleation of the eyes and the actual experiment may have altered ocular barrier properties. Majumdar et al. showed that the preservation of eyes for 24 h in Hank’s balanced salt solution results in no significant alteration in the transport of diazepam and L-arginine across cornea . However, the permeability of passive permeability marker mannitol was increased threefold in corneas preserved for 24 h compared to freshly isolated corneas, indicating an increase in permeability of the corneal epithelial barrier after death . The half-life of zona occludens-1 (ZO-1) in Madin Darby Canine Kidney (MDCK) cells is about 5 h based on metabolic radiolabeling studies . In our study we cannot rule out the possibility of an increase in permeability of ocular barriers at 18–24 h after harvesting. However, the trends observed for delivery may hold true in vivo with some differences in the extent of delivery.
In summary, this study showed the feasibility of the suprachoroidal space as a targeted delivery location for drug delivery to the choroid and retina. Suprachoroidal injection of NaF resulted in better signal localization to the choroid–retina when compared to intravitreal and intracameral injections. Suprachoroidal delivery to the vitreous decreases with an increase in molecular weight and lipophilicity of drug molecules. Suprachoroidal injection results in region-selective distribution of drug molecules, with drug levels being higher on the injected side than the opposite side. However, these differences are likely to be less prominent than those reported for periocular injections.
This work was supported by NIH grants R24EY017045 (through Emory University), R01EY018940, and R01EY017533, and an unrestricted grant from RPB.