Dispase‐assisted vitrectomy for epiretinal prostheses implantation

Abstract

Editor, We previously reported about a new surgical method for the implantation of epiretinal prostheses in porcine eyes (Ivastinovic et al. 2010a). While performing vitrectomy as part of the implantation procedure, we noted a strong vitreoretinal adherence hampering a complete removal of the posterior vitreous cortex. However, close proximity between the electrodes and ganglion cells is crucial for successful induction of visual perception with epiretinal implants (de Balthasar et al. 2008). Although in elderly humans posterior vitreous detachment (PVD) is commonly observed during vitrectomy, vitreoschisis might occur at different levels owing to the multilamellar structure of the posterior hyaloid (Sebag 2008). Hence, considerable amounts of vitreous remnants might still remain attached to the inner limiting lamina (ILL). Intravitreal pharmacological adjuncts, such as plasmin and dispase, have been shown to facilitate PVD. While plasmin hydrolyses laminin and fibronectin, which play key roles in the vitreoretinal attachment, dispase additionally degrades type IV collagen in the ILL, which supports the attachment of the vitreous cortex to the ILL (Wang et al. 2004). The incidence of dispase-related complications including preretinal haemorrhage, vitreous inflammation and morphological retinal damage proved to be dose dependent and more likely to occur after prolonged exposure (Wang et al. 2004). To obviate these adverse effects, the injected dose of dispase should not exceed 50 μg, and the incubation should be limited at 15 min (Oliviera et al. 2001). The aim of our study was to evaluate the efficacy and safety of 50 μg dispase at the proposed exposure in porcine eyes with respect to epiretinal prostheses implantation. We performed an experimental study in eight domestic pigs weighting 20–25 kg. This animal trial was approved by the Austrian legal authorities and the Medical University of Graz. The animals were treated according to guidelines of the Association for Research in Vision and Ophthalmology. Surgeries were performed under sterile conditions in general anaesthesia using standard parameters. One day before surgery, dispase (isolated from Bacillus polymyxa; Roche Diagnostics, Grenzach-Wyhlen, Germany) was diluted with phosphate-buffered saline to 50 μg/0.05 ml (pH = 7.4) under sterile conditions and stored at +4°C. After core vitrectomy, dispase was injected into the midvitreous cavity. The sclerotomies were temporarily closed, and the infusion stopped. After incubation of 15 min, the remaining vitreous was carefully removed. Clinical examination with indirect ophthalmoscopy was performed in general anaesthesia 4–6 weeks after surgery. After painless euthanization, the eyes were enucleated and processed for histological and immunohistochemical examination for glial fibrillary acidic protein (GFAP) to explore the extent of retinal damage. After dispase injection, PVD was induced with suction of the cutter, and the remaining vitreous could easily be removed in all animals. In four eyes, we observed spontaneous preretinal haemorrhage (Fig. 1). After the observation period, no obvious retinal pathology, such as disruption of ILL, was noted clinically. The vitreous cavity was clear in five eyes; three eyes showed a visible vitreous haze. On histological examination, complete vitreoretinal dehiscence was observed in six eyes while two eyes showed slight strands of the posterior hyaloid (Fig. 2). The vitreous haze consisted of a fibrin meshwork with inflammatory cells (Fig. 3). Immunohistochemistry revealed slight GFAP upregulation in the inner retina (Fig. 2C). Spontaneous preretinal capillary haemorrhage after dispase injection. The haemorrhage ceased intraoperatively after increasing the intraocular pressure by raising the infusion bottle. (A) Extreme periphery of the retina and pars plana. Note complete removal of posterior hyaloid. (B) Slight remnants of the posterior vitreous cortex partially attached to inner limiting lamina (arrowhead). (C) Upregulation of glial fibrillary acidic protein (GFAP) in the inner retina (arrowhead). [The samples were stained with Elastica van Gieson and haematoxylin-eosin. Immunohistochemistry was performed with GFAP antibodies. The choroidal and retinal detachment (A) is a processing artefact. Bars measure 100 μm]. Incomplete removal of posterior hyaloid and fibrin meshwork. The rectangle displays a macrophage in higher magnification. The retinal and vascular morphology is intact. [Elastica van Gieson stain. Bar measures 100 μm]. We used dispase in our study because of its commercial availability at relatively low costs. Intravitreal injection of 50 μg dispase (0.025 U) with an exposure of 15 min facilitates the complete removal of the posterior vitreous cortex and would provide a close contact between the epiretinal implant and the inner retina. However, hydrolysis of the extravascular matrix causing preretinal haemorrhage and immune response to dispase as a foreign protein still remain a concern. The extent of GFAP upregulation in our study corresponds to GFAP upregulation after vitrectomy alone (Ivastinovic et al. 2010b). In conclusion, though no structural retinal damage was observed, we do not recommend dispase as pharmacological adjunct in humans undergoing epiretinal prostheses implantation due its potential toxicity.