• Cornea/External Disease

    This article discusses the kinds of cases for which the Rochester Eye Institute Team used the Boston KPro device, and it addresses some of the underlying reasons for the selection of this modality. This article does not report in detail on the risks, benefits, and complications associated with keratoprosthesis surgery in general or those specifically related to the Boston device.

    Recent Keratoprosthesis History

    For more than half a century, keratoprosthesis (KPro) has been perceived by corneal surgeons as well as by the general ophthalmological community as an infrequently performed procedure, indicated only in bilateral blindness where no alternatives were possible, and associated with poor prognosis and a high rate of complication. Over the last decade, only a handful of these procedures were performed annually throughout the world. Recent reports seem to indicate an altered utilization pattern associated with the modified Boston KPro device.

    As a result of a number of changes in the design of the Boston device,1 the techniques for surgical implantation,2 as well as  routines of postoperative management, the procedure has taken on a new identity.3  A recent review of more than 250 cases performed at the University of Rochester Eye Institute over a period of 48 months revealed a wide set of indications and a low rate of complications.

    Current Design

    The Boston KPro is composed of polymethylmethacrylate and consists of a 3.2 mm optical cylinder that is passed through a central 3 mm aperture in an 8.5 mm or 8.75 mm donor corneal button. A fenestrated back plate clips on to secure donor tissue to the front plate (optical cylinder). A locking titanium ring prevents uncoupling of the device (diagram 1).

    Courtesy of James V. Aquavella, MD
    Diagram 1. Depicts the donor cornea situated between the front optical cylinder and back retaining plate.

    Coupling the device with a piece of donor tissue assists in the assembly process and enables rapid implantation in the style of a cornea transplant utilizing 10-0 interrupted nylon sutures. The fenestrations allow for the free passage of metabolites, a modification perceived to be primarily responsible for the new stability of the implanted device.

    A large-diameter hydrophilic bandage contact lens is placed over the implanted device and is worn continuously, affording protection to the ocular surface. Long-term prophylactic antibiotic coverage is thought to be responsible for the paucity of postoperative infections, in contrast to the problems associated with previous iterations of this procedure.3

    Current Indications for Keratoprosthesis

    In the current series of cases performed at the University of Rochester Eye Institute, device utilization was found to fall into 3 broad categories.

    The first category consists of adult cases in which routine corneal transplantation was considered complicated and associated with a guarded prognosis. The second category consists of primary procedures performed in lieu of routine corneal transplantation. The third category consists of infants and young children (Table 1).

    Table 1

    Indications for Keratoprosthesis

    1. Poor prognosis keratoplasty

      a. Autoimmune

      b. Nonautoimmune

    2. Primary keratoprosthesis in cases with good prognosis for corneal transplant

    3. Pediatric keratoprosthesis

    Category 1 can be divided into groups with and without associated autoimmune involvement. In the absence of autoimmune disease, there were no extrusions, and there was 100% retention despite a preoperative diagnostic array that included keratitis sicca, neurotropic and chemical keratitis, repeated graft failure, herpes simplex, graft vs. host disease, stem-cell deficiency, glaucoma, and aniridia.

    The prime motivating factor in this group of patients was to avoid the prospect of allograft rejection. The initial cases fell into this category, and all of these as well as subsequent cases were characterized by a return to full visual potential within a matter of weeks after surgery.

    The absence of astigmatism and the ability to essentially ignore neovascularization were benefits realized in the early postoperative period. The researchers were not confronted with persistent epithelial defects or other evidence of ocular surface disease, which frequently threaten graft survival following penetrating keratoplasty. Management of glaucoma was straightforward and unimpeded by the presence of the device. Fundus evaluation and photography were unimpeded as were routine visual fields.

    Some patients had undergone earlier glaucoma surgery including trabeculectomy and aqueous shunts. In 3 patients, shunts were implanted at the time of prosthesis surgery; in 3 others, implantation followed prosthesis surgery. Figures 1 and 2 show the pre- and postoperative appearance of a patient with xerophthalmia secondary to trachoma.

    Courtesy of James V. Aquavella, MD
    Figure 1. Preoperative appearance of severe xerophthalmia and trachoma.

    Courtesy of James V. Aquavella, MD
    Figure 2. Same eye, 3 years following KPro.

    The autoimmune-associated cases included primary diagnoses of pemphigoid, Stevens-Johnson syndrome and graft-versus-host disease. Complications were anticipated for these cases, and all were instances of bilateral, longstanding involvement. One of the major difficulties is the fitting of a contact lens in the absence of an adequate fornix. Some of these cases required additional surgery; in 3 cases, the prostheses were replaced. There was 1 instance of endophthalmitis. Systemic immunosuppression is an important adjunct. Dense retroprosthetic membranes were observed in some instances, not all amenable to YAG laser therapy. While the majority of these patients have done well, this is the sole area in which the potential for complications will persist indefinitely. The researchers utilized the type 1 Boston device exclusively for these complicated cases as well.

    Category 2 consisted of cases for which traditional penetrating keratoplasty was thought to offer a good prognosis, and it would have been the procedure of choice as recently as 4 years ago. These cases included pseudophakic bullous keratopathy, Fuchs dystrophy, keratoconus, lattice dystrophy, previous allograft rejection, and herpetic and other nonvascularized scars.

    The lure of KPro in these cases is the very rapid rehabilitation with the absence of astigmatism. In a few of the earlier cases in this category, the researchers did not remove the natural crystalline lens if it was clear. In other cases, they removed both clear and cataractous lenses, implanted an intraocular lens (IOL), and used a pseudophakic KPro. In still another subgroup, they elected not to implant an IOL and used an aphakic device with appropriate refractive power. The rationale and methodology associated with these decisions are beyond the scope of this article. Figures 3 and 4 represent pre- and postoperative appearance of a patient with secondary high astigmatism and keratoconus in an otherwise clear graft.

    Courtesy of James V. Aquavella, MD
    Figure 3. Preoperative appearance of recurrent keratoconus and high astigmatism following corneal transplant.

    Courtesy of James V. Aquavella, MD
    Figure 4. Same eye, 2 years following KPro.

    Category 3 consisted of 45 eyes in infants and children. The perception that traditional cornea transplantation in infants and children is associated with a high rate of allograft rejection4,5 coupled with the success in the adult series led me to postulate that keratoprosthesis might prove more suited to quickly provide the clear retinal image necessary for amblyopia prevention and therapy. All of these eyes are functioning with retention of the prosthesis without infection, while the severity of associated vitreoretinal disease has resulted in no light perception in 3 of the 45 eyes. The researchers believe they are functioning at the maximum level of acuity that the visual apparatus can perceive. They await visual acuity and amblyopia prevention and therapy data, which will require years of additional observation.

    About half the cases in this category consisted of virgin eyes with congenital opacity (mostly Peters anomaly and congenital glaucoma). The absence of postoperative inflammation and the immediate quality of the optics enable useful vision in the immediate postoperative period. This coupled with the absence of allograft rejection concerns have reinforced my contention that this is the procedure of choice in infants. The subjective visual results in the eyes with multiple previous surgical interventions (repeated graft failure, glaucoma procedures, cataract/IOL, vitrectomy, and retinal detachment) have led to the conclusion that useful acuity can be restored even in the face of severe pathology, often following long periods of visual deprivation.

    Congenital Cornea Opacity Is Associated With multiple Pathology

    In working with congenital disease, a team approach is mandatory, since frequently multiple pathological processes must be addressed concurrently in this population and are further complicated by the propensity for a heightened inflammatory response. Figures 5 and 6 demonstrate pre- and postoperative appearance in a 3-year-old with multiple failed transplants.

    Courtesy of James V. Aquavella, MD
    Figure 5. Three-year-old child following multiple failed transplants.

    Courtesy of James V. Aquavella, MD
    Figure 6. Same eye, 3 years following KPro.

    In conclusion this review indicates our current acceptance of this new keratoprosthesis technology for a variety of indications in which traditional cornea transplantations is complicated.  In addition our level of comfort with the procedure is also demonstrated by our use of keratoprosthesis as an alternative to routine cornea transplantation with good prognosis.  The most important and unusual finding has been our preference of keratoprosthesis in infants and children over traditional keratoplasty.


    1. Dohlman CH, Abad JC, Dudenhoefer EJ, Graney JM. Keratoprosthesis: beyond corneal graft failure. In: Spaeth G, ed. Ophthalmic Surgery. Principles and Practice. Philadelphia: WB Saunders; 2003:199-207.

    2. Aquavella JV, Qian Y, McCormick GJ, Palakuru JR. Keratoprosthesis: the Dohlman-Doane device. Am J Ophthalmol. 2005; 140(6):1032-1038.

    3. Aquavella JV, Qian Y, McCormick GJ, Palakuru JR. Keratoprosthesis: current techniques. Cornea. 2006;25(6):656-662.

    4. Taylor D, Wright KW, Amaya L, Cassidy L, et al. Should we aggressively treat unilateral congenital cataracts? Br J Ophthalmol. 2001;85(9):1120-1126.

    5. Aquavella JV, Gearinger MD, Akpek EK, McCormick GJ. Pediatric keratoprosthesis. Ophthalmology. 2007;114(5):989-994.

    Author Disclosure

    The author states that he has no financial interest, affiliation, or other relationship with the manufacture of any commercial project discussed or with the manufacture of any competing commercial project.

    The corneal surgeons in our group (James V. Aquavella, MD, Steven ST Ching, MD, Holly B. Hindman, MD, and Ronald D. Plotnik, MD) are grateful to our colleagues in pediatrics, glaucoma, and vitreoretinal disease (Mina Chung, MD, David DiLoreto, MD, Matthew D. Gearinger, MD, Shakeel Shareef, MD, Rajeev Ramchandran, MD,) for their assistance.