Polymer-free corticosteroid dimer implants for controlled and sustained drug delivery

Abstract

Polymeric drug carriers are widely used for providing temporal and/or spatial control of drug delivery, with corticosteroids being one class of drugs that have benefitted from their use for the treatment of inflammatory-mediated conditions. However, these polymer-based systems often have limited drug-loading capacity, suboptimal release kinetics, and/or promote adverse inflammatory responses. This manuscript investigates and describes a strategy for achieving controlled delivery of corticosteroids, based on a discovery that low molecular weight corticosteroid dimers can be processed into drug delivery implant materials using a broad range of established fabrication methods, without the use of polymers or excipients. These implants undergo surface erosion, achieving tightly controlled and reproducible drug release kinetics in vitro. As an example, when used as ocular implants in rats, a dexamethasone dimer implant is shown to effectively inhibit inflammation induced by lipopolysaccharide. In a rabbit model, dexamethasone dimer intravitreal implants demonstrate predictable pharmacokinetics and significantly extend drug release duration and efficacy (>6 months) compared to a leading commercial polymeric dexamethasone-releasing implant.

Fig. 1. Processing of drug dimers into stand-alone drug delivery vehicles.

a Schematic demonstrating the general form of the dimers, and the chemical structure of a representative dimer molecule made with the corticosteroid dexamethasone using a triethylene glycol (TEG) linker. (b Transmission electron microscopy (nanoparticles), light micrographs (coatings), and scanning electron microscopy (SEM) (all others) images demonstrating steroid dimers processed into different physical forms through thermal processing (fibers, coatings, extruded rods) and solvent processing (nanoparticles, microparticles, fibers, coatings, fibrous meshes).

Fig. 2. Physical characterization of drug dimers and their processed forms.

a Differential scanning calorimetry thermograms for crystalline dimers compared to solvent (fibrous mesh) and thermal (pellet) processed forms, demonstrating a transition from a crystalline to amorphous state. T_g – glass transition temperature. T_m – melting temperature. b Powder X-ray diffraction diffractogram demonstrating the crystalline nature of the synthesized dimer relative to the amorphous structure of the processed forms. c Experimental set-up, representative stress-strain curve, and flexural modulus for 3-point bend testing of an extruded rod of dexamethasone (Dex) dimer (Dex-TEG-Dex). Gray circles represent the raw data and black circles represent the 10-point moving average. N = 5. Data represent the mean ± standard deviation. d SEM micrograph of a fractured microparticle and light micrograph of an extruded rod following 3-point bend testing. e Polarized light microscopy of a longitudinal view and cross-section of an extruded rod, demonstrating the absence of crystalline structure. f Representative curve demonstrating rheometric characterization of the heat-processed Dex-TEG-Dex.

Fig. 3. Mechanism of drug release from a drug dimer material.

Schematic illustrating the proposed mechanism for the release of free drug from a dimer implant.

Fig. 4. Characterization of the drug release behavior of a corticosteroid dimer implant.

a Method of thermally processing crystalline dimers into pellets for in vitro release testing. b Release kinetics of dexamethasone from thermally processed pellets in 100% fetal bovine serum (FBS), 1% FBS in phosphate-buffered saline (PBS) (pH 7.4), and PBS (pH 7.4) (2 ml). Total drug loading was 1.17 ± 0.24 mg, 1.13 ± 0.17 mg, and 0.85 ± 0.03 mg dexamethasone (Dex) for 100% FBS, 1% FBS, and PBS release conditions, respectively. N = 3 samples. c Light micrographs demonstrating the surface erosion phenomenon, where pellets in FBS decrease in size over time as dimer is solubilized from the surface and degraded into free drug, resulting in surface erosion-based release kinetics. d Comparison of free dexamethasone release profiles in equivalent protein environments with differing hydrolytic enzyme activity (FBS (high) and 3.6% bovine serum albumin (BSA) (low)) in PBS pH 7.4 (2 ml). Total drug loading was 1.28 ± 0.19 mg Dex. N=3 samples for FBS and N=6 samples for BSA. e Demonstration of release rate and its dependence on sample diameter. Total drug loading of 63 ± 16 µg for 30 G × 6 mm implants and 348 ± 141 µg for 23 G x 6 mm implants. N = 6 samples for 23 G and N = 5 samples for 30 G implants. Due to the surface erosion mechanism of release, material diameter is directly correlated to duration of release. Degradation kinetics of solubilized dimer in f fetal bovine serum (N=3 samples) showing full conversion to free drug (g) with and without the presence of a protease inhibitor (Pefabloc, 1 mg/ml) (N = 3 samples). h Inhibition of LPS-induced stimulation of PGE2 production by primary human monocytes in the presence of free dexamethasone and dexamethasone released from Dex-TEG-Dex (N=3). In all panels, data represent the mean ± standard deviation.

Fig. 5. Versatility of the dimer material platform to achieve rational design.

Release kinetics of heat-molded 1 mm×1 mm (height x diameter) pellets for free a prednisolone (Pred) from Pred-TEG-Pred (N=3 samples, TEG – triethylene glycol) in 4 ml phosphate buffered saline (PBS) (0.96 ± 0.1 mg Pred loading), b triamcinolone acetonide (TA) from TA-TEG-TA (N = 5 samples) in 4 ml FBS (0.81 ± 0.18 mg TA loading), and c hydrocortisone (HC) from HC-TEG-HC in 4 ml PBS (0.60 ± 0.08 mg HC loading) (N=3 samples). Scanning electron micrograph (SEM) images and release curves demonstrating free drug release for dimers processed into d coatings (4 ml PBS, 0.78 ± 0.03 mg dexamethasone (Dex) loading, 12 mm diameter circular coating) (N = 3 samples), e extruded rods (4.4 ml PBS, 17.7 ± 4.4 µg Dex loading) (N = 3 samples), and f microparticles (10 ml fetal bovine serum (FBS), 42.2 ± 18.2 µm diameter, 0.12 ± 0.01 mg TA loading) (N = 3 samples). For coatings, the SEM image indicates a smooth coating surface at high magnification. g Release rate in FBS for free Dex from Dex dimers with either a TEG or hexane diol (Hex) linker (2 ml, 1 mm×1 mm height x diameter pellets, 1.39 ± 0.27 mg Dex loading for Dex-Hex-Dex and 0.51 ± 0.05 mg for Dex-TEG-Dex) (N=3 samples). h Correlation between release rate and surface area for implants in FBS (1.5 ml FBS, 23 G × 6 mm−348 ± 141 µg loading, 30 G × 1 mm – 11 ± 2 µg loading, 30 G × 6 mm – 63 ± 16 µg loading). N = 3 samples. In all panels, data represent the mean ± standard deviation.

Fig. 6. Formulation strategies for achieving multi-drug release with corticosteroid dimer materials.

a Release kinetics for Dex and HC from a Dex-TEG-HC heterodimer (4 ml PBS, 1 mm × 1 mm height × diameter pellets, 0.18 ± 0.08 mg HC and 0.19 ± 0.08 mg Dex loading) (N = 3 samples). b Processing of two different corticosteroid homodimers (Dex-TEG-Dex and HC-TEG-HC) into a solid implant that gives controlled release of both drugs. (4 ml PBS, 1 mm × 1 mm height × diameter pellets, 0.51 ± 0.05 mg Dex and 0.35 ± 0.03 mg HC total loading) (N = 3 samples). c Use of a dimer (Dex-TEG-Dex) to provide a matrix for the support, distribution, and controlled release of Dex and a second drug (sunitinib malate). Dex-TEG-Dex:Sunitinib malate 7:1. (4 ml PBS, 1 mm × 1 mm height × diameter pellets, 0.83 ± 0.14 mg Dex and 0.15 ± 0.02 sunitinib malate loading). N = 3 samples. In all panels, data represent the mean ± standard deviation. Orange and blue circles represent different corticosteroids. Dark blue rectangles indicate the linker. Green circles represent free drug contained within the dimer matrix.

Fig. 7. Acute efficacy of a dexamethasone dimer material in a rat model.

a Experimental set-up demonstrating treatment groups, method of implant administration, and methods and timepoints for assessing efficacy. b Fundus micrographs of the Dex-TEG-Dex (Dex – dexamethasone, TEG – triethylene glycol) implant post-implantation in the vitreous humor. (c) Representative optical coherence tomography (OCT) images 24 h post-lipopolysaccharide (LPS) stimulation. d Cell counts in the vitreous humor following LPS stimulation for eyes with an intravitreal dexamethasone dimer implant, eyes receiving topical dexamethasone eyedrops, and eyes receiving a sham injection. Vitreal cell counts were performed on images obtained by OCT. e Quantification of retinal thickness from OCT images following LPS stimulation for eyes that received dexamethasone dimer implants, dexamethasone eye-drops, or a sham injection. Data represent the mean ± standard deviation. N numbers are indicated in the figure for parts d-e. Statistical analysis was performed by ANOVA. *p < 0.05 vs. sham injection.

Fig. 8. Dissolution and pharmacokinetics of a dexamethasone dimer implant following intravitreal administration in the rabbit eye.

a Light micrograph of a 30 G x 6 mm heat-extruded dexamethasone dimer implant and schematic showing its physical location in the eye following intravitreal injection b Quantification of implant diameter over time from IR images for the 30 G × 6 mm Dex-TEG-Dex (Dex – dexamethasone, TEG – triethylene glycol) implants. N = 8 for all data points except N = 7 at 76 days. Data represent the mean ± standard deviation. Quantification of dexamethasone in the c vitreous humor, d retina, and e aqueous humor for 30 G × 6 mm dexamethasone dimer implants compared to experimental and literature^, findings for Ozurdex^® (vitreous humor only). N = 6 for the 30 G × 6 mm implant (except N = 5 for retina at 9 months) and N = 2 for Ozurdex^®. Data represent the mean ± standard deviation. Dashed line indicates the lower limit of quantification for each tissue type.

Fig. 9. Pharmacodynamics of an intravitreal dexamethasone dimer implant in the rabbit eye.

a Fluorescein angiograms of the back of the eye following intravitreal injection of vascular endothelial growth factor (VEGF) for eyes receiving 30 G × 6 mm dexamethasone dimer implants compared to Ozurdex® and a negative control (sham). b Semi-quantitative analysis of scoring for inhibition of VEGF-induced changes in the retinal vasculature. Data represent the mean ± standard deviation. N = 3 for Ozurdex® and N = 8 for 30 G × 6 mm Dex-TEG-Dex implants.

Fig. 10. Ocular histopathology following intravitreal implantation of a dexamethasone dimer implant in the rabbit eye.

Histology (hematoxylin and eosin (H&E)) of a rabbit eye 12-months post-intravitreal sham procedure or administration of Dex-TEG-Dex implants. VH – vitreous humor, L – lens, R – retina.