(T0930-07-43) Voriconazole Nanosuspension-Loaded Microneedle Patch for Fungal Keratitis

Purpose: Fungal keratitis (FK), primarily caused by Fusarium, Aspergillus, and Candida, affects around 1,051,800 people annually [1]. Voriconazole (VOR) is effective against FK causing fungi. Although systemic VOR is FDA-approved, no ophthalmic formulation exists, leading to off-label use of 1% eye drops from IV preparations. Topical delivery has < 5% bioavailability due to less residence time, tear dilution and drainage, while intrastromal injections offer site-specific delivery but are invasive and risk corneal damage. To address these issues, dissolving microneedle array patches (dMAPs) offer a minimally invasive alternative, enabling localised delivery with improved retention and reduced complications. However, loading free VOR into dMAPs is limited by its poor solubility and the presence of large, irregular particles, resulting in low drug loading, poor structural integrity, and reduced drug delivery efficiency [2]. To improve this, a VOR nanosuspension (NS) was prepared via aqueous bead milling and incorporated into 8×3 dMAPs containing 24 needles and a drug-free baseplate. The patches were fabricated using a modified two-layer casting method and evaluated for their potential in FK treatment.

Methods: The VOR-NS was prepared using an aqueous bead milling technique with polyvinyl alcohol (PVA) and polyvinylpyrrolidone (PVP) as stabilizers. The prepared NS was subsequently loaded into the MAPs. To fabricate MAPs, master templates were initially designed using Autodesk Fusion 360 and fabricated via two-photon polymerisation (2PP) 3D printing. The printing was performed on a Nanoscribe Photonic Professional GT system, which uses a femtosecond laser to precisely polymerise a photoresin on an indium tin oxide-coated glass slide. After printing, the master templates were mounted onto custom polylactic acid (PLA) holders produced using an Ultimaker 3D printer. Silicone moulds were prepared by casting MED-4011 silicone (10:1 ratio) onto the master template, degassed by centrifugation, and cured at 55°C for 4 h. These moulds were subsequently used for the fabrication of VOR-NS-loaded dMAPs with a two-layer casting method to create drug-loaded needles and a drug-free baseplate. The dMAPs comprised 24 needles in an 8×3 array, each measuring 600 µm in height, 300 µm in base width, with 100 µm spacing. Unlike the earlier 3×3 array MNs [2], the 8×3 array MNs was developed to accommodate a higher drug load. The rectangular shape, as opposed to a square, was chosen to enhance complete needle insertion into the cornea, which is often hindered by the eye’s curved surface and the flat MAP baseplate. Furthermore, for comparison, dMAPs loaded with free VOR were prepared using the same formulation and fabrication process as those loaded with VOR-NS. The developed dMAPs were evaluated for mechanical strength, insertion efficiency, drug content in the needle tips, ex vivo dissolution behavior using porcine eye globes, ex vivo corneal deposition using excised porcine corneal tissue, and antifungal activity via disc diffusion assay.

Results: Mastersizer^® measurements and SEM images confirmed that the free VOR particles were large and irregular in shape. The optimized formulation yielded a milky white VOR-NS with a mean particle diameter of 270.11 ± 5.82 nm and a PDI of 0.217 ± 0.019 before freeze-drying, and 290.1 ± 7.49 nm with a PDI of 0.314 ± 0.04 after freeze-drying, indicating a significant reduction in VOR particle size and a uniform size distribution. TEM images of the VOR-NS corroborated the DLS findings. The freeze-dried VOR-NS exhibited increased solubility than free VOR (Fig. 1). Microscopy and SEM showed that VOR-NS-loaded dMAPs had smooth, intact tips and uniform drug distribution, while free VOR-loaded dMAPs exhibited clumping and surface defects due to larger particles and poor packing. Free VOR-loaded dMAPs showed greater needle height reduction (21.7%) than VOR-NS-loaded dMAPs (13.23%). VOR-NS-loaded dMAPs achieved higher insertion efficiency, with 47.22% reaching ~390 µm depth versus 22.22% for free VOR-loaded dMAPs (Fig. 2). Additionally, VOR-NS-loaded dMAPs fully dissolved within 5 min, while free VOR-loaded dMAPs showed only 25.32% dissolution over the same period in excised porcine eye (cornea). Furthermore, VOR-NS-loaded dMAPs contained a higher drug amount in the needle tips (118.84 µg vs. 83.08 µg) and achieved a 2.31-fold increase in drug deposition in excised porcine corneas. Finally, the disc diffusion assay showed that VOR-NS and its dMAPs had stronger antifungal activity against Candida albicans and Aspergillus fumigatus than free VOR and its dMAPs (Fig. 3).

Conclusion: This study developed and evaluated a bilayer dMAP incorporating VOR-NS for localized ocular delivery as a potential FK treatment. The bilayer design enabled drug localization in the needle tips while keeping a drug-free baseplate to reduce drug loss. VOR-NS, prepared via aqueous bead milling, improved solubility and ensured uniform incorporation in MAP tips. Compared to free VOR-loaded dMAPs, VOR-NS-loaded dMAPs showed enhanced mechanical strength, insertion, and drug loading. In vitro and ex vivo studies demonstrated that VOR-NS-loaded dMAPs exhibited faster needle dissolution, greater corneal drug deposition, and superior antifungal activity, highlighting their potential for FK treatment.

References: [1] Brown L, Leck AK, Gichangi M, Burton MJ, Denning DW. The global incidence and diagnosis of fungal keratitis. Lancet Infect Dis. 2021;21(3):e49-e57. 10.1016/S1473-3099(20)30448-5
[2] Wu Y, Vora LK, Mishra D, Adrianto MF, Gade S, Paredes AJ, et al. Nanosuspension-loaded dissolving bilayer microneedles for hydrophobic drug delivery to the posterior segment of the eye. Biomater Adv. 2022;137:212767. https://doi.org/10.1016/j.bioadv.2022.212767

Acknowledgements: B.H. Jaswanth Gowda thanks the Commonwealth Scholarship Commission, Foreign, Commonwealth & Development Office, UK Government, for granting him the Commonwealth Split-site PhD Scholarship 2023/24 to conduct research at Queen’s University Belfast for one year. The authors thank Professor Gordon Ramage (Glasgow Caledonian University, UK) for providing the Candida albicans SC5314 and Aspergillus fumigatus AF293 strains used in this study.


Fig. 1 (a) Particle size distribution of free VOR powder (mean ± SD, n = 3). (b) SEM image of free VOR powder. (c–f) TEM images of freeze-dried VOR NS. (g) Apparent and saturation solubility of free VOR, physical mixture, and freeze-dried VOR NS (mean ± SD, n = 3). Dispersion of (h) free VOR and (i) freeze-dried VOR NS in water.


Fig. 2 Digital microscopy and SEM images of (a–c) free VOR-loaded dMAP and (d–f) VOR NS-loaded dMAP. Drug distribution shown in (g) free VOR-loaded and (h) VOR NS-loaded dMAPs. Digital images of (i) free VOR-loaded and (j) VOR NS-loaded dMAPs after 3 N/array compression for 30 s using a TA-XT2 texture analyser. (k) Needle height reduction post-compression. (l) Insertion profiles of dMAPs (mean ± SD, n = 3). (m) Drug content in dMAP tips (mean ± SD, n = 3).


Fig. 3 (a,b) Administration of dMAPs to a porcine eye globe positioned in a custom 3D-printed holder. Digital microscopy images of free VOR-loaded dMAP dissolution after (c) 1 min, (d) 3 min, and (e) 5 min; and VOR NS-loaded dMAP dissolution after (f) 1 min, (g) 3 min, and (h) 5 min. (i) Percentage dissolution of dMAP tips in porcine eye over time. (j) VOR deposited in the cornea after 5 min insertion (mean ± SD, n = 6). Antifungal activity of (k, l) free VOR, (m, n) PVA-PVP mixture, (o, p) freeze-dried VOR NS, (q, r) free VOR-loaded dMAP, and (s, t) VOR NS-loaded dMAP against Candida albicans and Aspergillus fumigatus (mean ± SD, n = 3).