Use of a Polyurethane Patch for Temporary Closure of a Sterile Corneal Perforation

Treatments for corneal perforation after thinning processes include acute penetrating keratoplasty, lamellar keratoplasty, grafting of conjunctival flaps, suturing of a scleral lamella into the perforation, tarsorrhaphy, and sealing of the perforation site with tissue adhesives.¹ However, the use of homologous tissue is often followed by immunological allograft rejection and may be technically demanding in a soft eye. Tissue adhesives are especially suitable for perforations smaller than 1.5 mm, but the application is not always easy.¹ For larger perforations, a patching material may be used. In this study we demonstrate that Neuro-Patch (B Braun Melsungen AG, Melsungen, Germany), a microporous, purified polyurethane material with excellent biocompatibility properties, can be effective in achieving temporary closure of corneal perforations. Neuro-Patch is primarily used as a dural substitute in neurosurgery. According to the product information, the material allows rapid immigration of connective tissue cells and is biostable and biocompatible. The material has a thickness of 0.45 mm, bends easily, and has a variable pore size ranging from 1 to 50 µm. The polyurethane fibers have a thickness of approximately 5 to 10 µm.

An 84-year-old woman with rheumatoid arthritis developed a 2×3-mm central corneal perforation 2 weeks after phacoemulsification of the left eye. Since donor tissue was not available at that time for emergency grafting and the perforation was too large to apply tissue adhesive, it was decided to use Neuro-Patch for closure of the perforation.

After all necrotic corneal tissue had been removed, the edges of the piece of Neuro-Patch were matched to the shape of the wound bed. A 2×3-mm patch was cut and sutured into the defect with six 10-0 nylon sutures, and the cornea was covered with a bandage contact lens. Six months later, the patch was still perfectly positioned with neither aqueous leakage nor any signs of anterior chamber inflammation (Figure 1). Meanwhile the patient had been treated with prednisone, 30 mg daily, and methotrexate, 10 mg weekly, and a penetrating keratoplasty was performed. Findings from histologic examination of the corneal button revealed epithelialization of the patch and infiltration of fibroblastlike cells into the deeper layers of the patch (Figure 2). No inflammatory cells were seen in the stroma or patch. Findings from transmission electron microscopy revealed invagination of epithelial cells into the patch and the formation of a new basal membrane by the epithelial cells. In addition, there was collagen formation of the keratocytes around the fibrillar architecture of the patch.

Despite the systemic immunosuppression and topical treatment with serum drops, the patient developed marginal thinning of the graft. A conjunctival flap ended in necrosis, and ultimately tarsorrhaphy was performed. Nineteen months postkeratoplasty, the graft was relatively clear with mild punctate epitheliopathy, and the patient's visual acuity was counting fingers OS.

A 5-year-old boy with Smith-Lemli-Opitz syndrome, an autosomal recessive disorder with severe psychomotor retardation and microcephaly, was seen for a 2×1.5-mm corneal perforation in the right cornea owing to neurotrophic keratopathy caused by infrequent blinking. A 2×1.5-mm piece of Neuro-Patch was sutured into the perforation with six 10-0 nylon sutures (Figure 3). Prior to treatment, informed consent was obtained from the patient's parents. One month later, the sutures were removed, and the patch was tightly fastened to the corneal wound margins. Postoperatively, neither anterior chamber leakage nor signs of anterior chamber inflammation were noted. Nevertheless, after tapering the topical preservative-free 0.5% prednisolone, the patch was lost. Findings from an examination with the patient under general anesthesia revealed that the stromal thickness had increased to 50% of the original thickness (Figure 4). Twelve months after the placement of the patch, the eye was quiet, and the epithelial surface was well controlled with artificial tears and ointments.

The closure of corneal perforations with human tissue in noninfectious corneal thinning is hampered by rejection of the corneal graft and by the recurrence of corneal thinning in the donor. Factors include the timing of the perforation, which often necessitates immediate closure in a patient at risk of immune reactions with high levels of corneal polymorphonuclear cells, collagenases, and proteases that destroy corneal collagen and proteoglycans.¹ Nobe et al² recently reported that all grafts failed in 4 patients with rheumatoid arthritis and corneal perforation. Palay et al³ demonstrated a survival probability of only 32% 2 years after the first keratoplasty in patients with rheumatoid arthritis. In a study by Bernauer et al,⁴ penetrating keratoplasty in corneal thinning resulted in an 80% failure of the grafts 6 months postoperatively. Notable improvement in graft survival could be achieved with immunosuppression, suggesting that a delay of penetrating keratoplasty until the inflammation has subsided would be beneficial.

The use of tissue adhesives has been advocated for perforations that are less than 1.5 to 2 mm and have a small amount of surrounding stromal ulceration.¹ The application of the smallest amount of glue to create a smooth surface may be technically demanding, and multiple applications may be needed.⁵ In a large study⁶ of 80 patients with corneal perforation or impending perforation, N-butyl-cyanoacrylate remained in place for an average of 50 days, and 44% of patients healed without further treatment. With respect to impending or actual perforation associated with rheumatoid arthritis, use of tissue adhesive was successful in 6 of 12 cases.

We sutured Neuro-Patch into 2 corneal perforations and observed good adhesion of the patch to the host tissue. Epithelialization of the patch occurred gradually without any signs of inflammation of the anterior eye chamber. Findings from histopathologic examination demonstrated the development of an epithelial cell layer above the patch, and findings from electron microscopy showed the beginning of the deposition of the basal membrane. Legeais et al⁷ reported a study of 6 patients treated with a temporary polytetrafluoroethylene graft 0.7 to 1.0 mm in thickness (Gore-Tex). Although this material was well tolerated, no epithelialization of the implant or cellular ingrowth into the porous polymer was seen on histologic examination. Portnoy et al¹ described the successful use of lyophilized donor tissue (Kerato-Patch; Allergan Medical Optics, Irvine, Calif) as a planolamellar button to manage central stromal ulceration. Recently, Neuro-Patch also has been used for the closure of a scleral defect after the spread of infection that was the result of an explosion with multiple shrapnel.⁸

We believe that Neuro-Patch offers an additional method for the temporary closure of corneal perforations. Its advantages inlcude the ease in technical handling and the excellent biocompatibility as proven by these 2 cases. In addition, the use of Neuro-Patch allows the surgeon to schedule graft surgery until the time corneal disease is adequately controlled. Finally, in one case, graft surgery was avoided since stromal regeneration occurred underneath the patch.

The authors have neither commercial, proprietary, or financial interest in Neuro-Patch nor did they receive payment as consultants.

Corresponding author: Rudy M. M. A. Nuyts, MD, PhD, Department of Ophthalmology, Academic Hospital Maastricht, Postbus 5800, 6202 AZ Maastricht, the Netherlands.