Retroprosthetic membrane: A complication of keratoprosthesis with broad consequences
Introduction
Many advances in the field of keratoprosthesis have been made since the first description of an artificial cornea in 1789 by French ophthalmologist Guillaume Pellier de Quengsy [1]. Various types of keratoprosthesis are successfully used today to restore vision to individuals with end-stage corneal blindness not amenable to traditional penetrating keratoplasty. Current penetrating artificial cornea devices can be divided into those best suited for patients with a wet blinking eye (type 1) and those for patients with severe surface dryness (type 2). Within the type 1 devices, the most commonly used is the Boston type 1 keratoprosthesis (BKPro), with over 14,000 implanted worldwide. Type 2 devices are much less commonly implanted, with only a few active centers around the world. Among type 2 devices, the osteo-odonto keratoprosthesis (OOKP) has shown the best long-term results.
Despite relatively good outcomes with keratoprosthesis implantation, reports by multiple institutions indicate that postoperative complications can lead to vision loss over time. One of the most common complications is the growth of fibrovascular tissue behind the prosthetic device, or retroprosthetic membrane (RPM) (Fig. 1). While retrocorneal membranes can be found in patients after penetrating keratoplasty, they occur at a much higher rate in implanted artificial corneas to the point that it has become a key complication in the field of keratoprosthesis. RPM is likely due to the host's physiologic response to implantation of a foreign device, which induces scarring and fibrosis. Initially, RPM was not considered a common cause of permanent vision loss, but evidence is now suggesting that it may have a contributory if not causative role in several other complications encountered in these patients, including glaucoma, corneal melt, retinal detachment, and hypotony.
Although retroprosthetic membranes appear to be less common in biological devices such as the OOKP, they remain a universal problem across all populations and in all types of keratoprosthesis used today. This review will examine current evidence on RPM etiology, risk factors, and management, comparing different devices and surgical techniques and focusing on their potential contribution to other complications and patient outcomes.
Section snippets
Incidence
Table 1 demonstrates the incidence of RPMs in different types of KPro according to large published studies. The table also includes information on how the RPMs were treated and the incidence of KPro complications associated with RPM.
Mechanism and histopathology of RPM
Retrocorneal membranes can develop after various types of intraocular surgery, including penetrating keratoplasty, DSAEK, and glaucoma drainage implants [18]. A histopathologic study examining 28 retrocorneal membranes from keratoplasty specimens showed that many were composed of stromal keratocytes or fibroblasts [18]. These membranes had strong α-smooth muscle actin (α-SMA) staining. An animal study on transplantation of porcine corneas into rhesus monkeys has shown that the retrocorneal
Prevention
Intraocular inflammation is thought to contribute to the development of RPM, and steroids have been proposed as a preventive measure. It has been shown that intracameral steroids delay the time of onset of RPMs but do not alter the overall incidence of RPM [6].
Secondary complications
Because myofibroblasts secreting α-smooth muscle actin are a major component of RPMs, they possess contractile activity and can cause traction when they grow over intraocular structures such as the iris, anterior chamber angle, ciliary body, and retina. RPMs can therefore lead to serious complications, including occlusion of the visual axis, corneal melt and device extrusion, glaucoma exacerbation due to angle closure, hypotony, and retinal detachment (Fig. 2).
Titanium vs PMMA
Titanium back plates are a newer option for the Boston KPro Type 1. Titanium demonstrates excellent tissue tolerance in other biological implants and can be made thinner than its PMMA counterpart (edge thickness 0.25Â mm versus 0.9Â mm with PMMA), thereby reducing anterior chamber crowding [36]. It has been proposed that titanium back plates could lead to reduced RPM formation. An in vitro study that showed improved corneal cell proliferation and decreased cell death when cells were in contact
Treatment
RPMs that do not obstruct the visual axis can be observed. Visually significant RPMs can be successfully treated with the Nd:Yag laser. However, thick and/or vascularized RPMs must be excised via surgical membranectomy and sometimes KPro explantation or exchange. Given that the field of view is limited by the diameter of the optical cylinder of the KPro, endoscopic visualization can be used when performing vitreoretinal surgery for membranectomy [14]. AS-OCT, which can quantify the thickness of
Future directions
As mentioned previously, intracameral steroids delay the time of onset of RPM but do not prevent the eventual development of RPM [6]. This is not surprising, given that the effect of intraocular steroid injections is limited by the medication's short half-life. Sustained release steroid formulations could be a solution to this issue, but with this comes the risk of steroid-induced glaucoma exacerbation in a population already plagued with severe glaucoma.
Intraocular administration of
Conclusions
Obtaining highly powered studies on KPro, specifically devices other than the BKPro Type 1, is challenging. This is due to the relatively small number of patients receiving these implants. From the studies that we have so far, RPM seems to be more common in the Boston Type 1 KPro. Examining the similarities and differences among the various KPro models can provide clues for this discrepancy. Studies focusing on RPM are limited, but it is clear that RPMs are involved in vision-threatening
Funding sources
NEI Center Core Grant for Vision Research (P30 EY001792) and Unrestricted grant from Research to Prevent Blindness.
Acknowledgments
The authors thank Lauren Kalinoski for her assistance in creating Fig. 2.
References (48)
- et al.
Boston keratoprosthesis: outcomes and complications
Ophthalmology
(2015) - et al.
Risk factors for the development of retroprosthetic membranes with Boston keratoprosthesis type 1: multicenter study results
Ophthalmology
(2012) - et al.
Long-term visual outcomes and complications of Boston keratoprosthesis type II implantation
Ophthalmology
(2017) - et al.
Boston type 2 keratoprosthesis- mid term outcomes from a tertiary eye care centre in India
Ocul Surf
(2019 Jan) - et al.
Vitreoretinal complications and outcomes in 92 eyes undergoing surgery for modified osteo-odonto-keratoprosthesis: a 10-year review
Ophthalmology
(2018) - et al.
Comparison of long-term results between osteo-odonto-keratoprosthesis and tibial bone keratoprosthesis
Ocul Surf
(2018) - et al.
Retrocorneal membranes: a comparative immunohistochemical analysis of keratocytic, endothelial, and epithelial origins
Am J Ophthalmol
(2010) - et al.
A progressive anterior fibrosis syndrome in patients with postsurgical congenital aniridia
Am J Ophthalmol
(2005) - et al.
Chronic hypotony associated with the Boston type 1 keratoprosthesis
Am J Ophthalmol
(2012) - et al.
Impact of total pars plana vitrectomy on postoperative complications in aphakic, snap-on, type 1 Boston keratoprosthesis
Ophthalmology
(2017)