An artificial cornea (Boston Keratoprosthesis type 1; Massachusetts Eye and Ear Infirmary, Boston) made of polymethyl methacrylate with a front plate, a clear optical stem, and a back plate with holes.
Patient 7 at 39 months' follow-up after artificial cornea (Boston Keratoprosthesis; Massachusetts Eye and Ear Infirmary, Boston) implantation for herpetic keratitis, with a visual acuity of 20/40.
Patient 1, with 6 months of inflammation and a persistent epithelial defect secondary to herpetic keratitis (A) and the same patient 1 month after artificial cornea (Boston Keratoprosthesis; Massachusetts Eye and Ear Infirmary, Boston) implantation, with a visual acuity of 20/30, which was maintained for more than 12 months of follow-up (a soft contact lens can also be seen) (B).
Khan BF, Harissi-Dagher M, Pavan-Langston D, Aquavella JV, Dohlman CH. The Boston Keratoprosthesis in Herpetic Keratitis. Arch Ophthalmol. 2007;125(6):745-749. doi:10.1001/archopht.125.6.745
To evaluate the role of an artificial cornea (Boston Keratoprosthesis [KPro]) in herpetic corneal graft failures.
A series of 17 herpetic eyes of 14 patients with failed transplantations, treated with implantation of a corneal graft with a KPro, was reviewed. Outcomes measured were retention of the device, visual acuity at 1 week, best visual acuity, last assessed visual acuity, time of follow-up, and complications.
Follow-up ranged between 6 and 72 months (median, 14 months). All patients had improvement in postoperative visual acuity, with 16 (94%) of 17 eyes getting better than preoperative visual acuity within 1 week. Of the 17 eyes, 15 (88%) achieved a best visual acuity of 20/25 to 20/70 and, at the last examination, 11 eyes (73%) were still in that range. Four patients with prolonged preoperative inflammation and ulceration at KPro placement had resolution of inflammation within days. The KPro had no extrusions. Complications included retroprosthesis membrane in 3 eyes and 1 tissue melt in an early case.
The KPro can be of value in patients with graft failure from herpetic keratitis, even in the inflamed stage. We believe that herpetic keratitis is not a contraindication for KPro use.
Standard corneal transplantation traditionally carries a poorer long-term prognosis in patients with herpetic keratitis compared with several other categories of corneal recipients, even after the arrival of antiviral agents and better surgical techniques.1,2 Graft failures are not uncommon, and subsequent corneal transplantations carry an even lower chance of survival.3 An artificial cornea (Boston Keratoprosthesis [KPro]; Massachusetts Eye and Ear Infirmary) may be the next alternative. Historically, KPros carry a questionable prognosis in herpetic keratitis; this diagnosis is also considered a contraindication for another available prosthetic cornea (AlphaCor prosthesis; Argus Biomedical, Perth, Australia).4
In this report, we looked at our experience at the Massachusetts Eye and Ear Infirmary and the University of Rochester Eye Institute of implanting a fresh corneal graft with the KPro type 1 in patients with corneal graft failures who experienced scarring of the primary graft for herpetic keratitis or who developed infection in the graft later. Retention of the device, early and best visual acuity, and complications were evaluated.
A retrospective case review was done of 17 eyes in 14 patients who had the KPro implanted in Boston and Rochester by 2 surgeons (patients 1-9 by C.H.D. and patients 10-14 by J.V.A.) with similar surgical techniques.
Preoperative assessment included detailed ocular history, including the number of previous failed grafts and the last herpetic flare-up. Special emphasis is given to lens status, glaucoma, inflammation, macular and optic nerve status, and best visual acuity achieved when the last graft was clear. The Snellen chart was used for visual testing when possible. Other techniques included counting fingers, hand motions, and central fixation and projection, determined with a point light. Intraocular pressure measurement was performed using Goldmann applanation tonometry or an easy-to-use, portable, handheld instrument (Tono-Pen; Medtronic, Minneapolis, Minn), or by digital palpation depending on the status of the corneal disease. A detailed slitlamp examination was performed, and when the posterior pole could not be visualized, B-scan ultrasonography was done. Axial length measurements were obtained to decide on the dioptric power of the KPro. Anterior segment photographs were taken. When there were questions about the glaucoma damage or the retinal status, consultations were obtained from the appropriate department. Detailed informed consent was obtained using special KPro consent forms, and the patients and relatives were shown photographs illustrating postoperative cosmetic outcomes.
The type 1 KPros are made of polymethyl methacrylate and shaped as a double-plated collar button. History,5,6 design7,8 (Figure 1), and surgical techniques9- 11 have been previously described.
All of the patients except 2 (patients 11 and 14, whose original diagnosis was keratoconus and later had herpetic keratitis in the graft) had herpetic keratitis scarring as the initial pathological feature for the primary corneal transplantation. Seventeen devices were implanted over 13 years (Figure 2). One patient had bilateral KPros and 2 other patients needed KPro replacements. There were 6 women and 8 men, with an age distribution between 49 and 89 years. All patients had 2 or more previous penetrating keratoplasties (mean, 2.7). Patient 4 had 2 KPros placed in his right eye after developing a retrokeratoprosthesis membrane within 12 months. Patient 6 had a second KPro implanted in her right eye after developing a fusarium fungal infection in it. A KPro was later implanted in her left eye, also at the patient's request.
Preoperative comorbid conditions included glaucoma in 10 (66%) of the 15 eyes in which an artificial cornea was implanted, chronic cystoid macular edema in 2 eyes (13%), age-related macular degeneration in 3 eyes (20%), and scleral buckle in 1 eye (7%). Patient 4 had an Ahmed glaucoma valve preoperatively, and patient 2 had the Ahmed valve placed at KPro implantation.
The patients received lifelong prophylactic antibiotics (moxifloxacin, 20 μL twice daily) and oral antiviral therapy (acyclovir, 400 mg twice daily). They also had soft contact lenses placed (Kontur Lens Co, Richmond, Calif) (except for patient 5) for indefinite durations to prevent dryness and exposure damage to the corneal tissue around the KPro, which might lead to melt and necrosis.12
The patients' preoperative conditions are detailed in Table 1. Outcomes were measured as retention of the device, visual acuity at 1 week, best and last assessed visual acuity, time (in months) of follow-up, and complications.
Follow-up ranged between 6 and 72 months (median, 14 months) (Table 2). There were no new cases of glaucoma other than those already diagnosed preoperatively as of last follow-up.
In the 4 patients (patients 1, 6C, 8, and 9) who had active inflammation, corneal melt, and pain when receiving the KPro, the inflammation resolved within a few days of surgery (Table 3 and Figure 3A and B).
A retroprosthesis membrane developed in patient 3, which was successfully opened with an Nd:YAG laser. Patient 4 developed a mild tissue melt around the stem (no holes in the back plate), but no extrusion resulted. He also developed a thicker retroprosthesis membrane within 12 months, reducing his visual acuity from 20/50 to hand motions. This required removal of the KPro for the membrane removal and placement of a new KPro. This patient now has a visual acuity of 20/70, after a 60- month follow-up. He has also had 2 episodes of mild sterile vitritis, which were rapidly controlled with topical corticosteroids and peribulbar injections of triamcinolone. Patient 9 also developed a retroprosthesis membrane and did not want to proceed with further surgery. Patient 6 developed fusarium fungal keratitis in her right eye and, after complete infection control, she underwent another KPro placement in the same eye; she currently has a visual acuity of 20/60 at 14 months' follow up. Patient 5 eventually lost all vision in his eye because of glaucoma. He was noncompliant with his antiglaucoma medication. However, he had good vision for 2 years and his KPro was anatomically intact at 72 months' follow-up. The outcomes are summarized in Table 2 and Table 3.
During the 1980s and 1990s, changes in the medical and surgical management of ocular herpetic disease have reduced the need for penetrating keratoplasty and have improved the prognosis for surgical intervention when required. Standard corneal transplantation is traditionally the procedure of choice in the visual restoration of patients with significantly scarred herpetic eyes.
It has been reported that in herpetic eyes that have active inflammation at grafting, only 45% of grafts survive, compared with 85% survival in eyes that were quiet at surgery.13 In addition, the recurrence of herpes simplex virus infection in the graft was a significant risk factor for graft failure.14 The 5-year success rate of grafts in patients with herpetic keratitis is notably less than in other corneal diseases, such as keratoconus and dystrophies.15- 18 The prognosis for each subsequent new graft is worse for all subgroups.3,19,20
Considering the likelihood of failure in repeat grafts for herpes, an artificial cornea might be considered as an alternative. Several models of KPros are available, or are being developed.21 However, the literature on the outcome of KPros in herpes is scarce. A recent study4 showed a high failure rate with a different prosthesis (AlphaCor), and the procedure was, therefore, not indicated in herpetic keratitis.
We do not share this pessimistic outlook based on our results with the KPro provided herein. The assembly and surgical technique with the KPro are simple, and visual rehabilitation is relatively quick. At 1 week after surgery, 15 of 17 eyes had already achieved better visual acuity than their preoperative status, and most remained in the range of 20/25 to 20/70. Comorbid conditions were usually the cause of reduced vision, especially glaucoma. Progression of preoperative glaucoma was occasionally observed, as in series22,23 of standard regrafts, but there were no new cases of glaucoma after surgery.
All eyes maintained anatomical integrity without any KPro extrusions, although 1 early case developed tissue melt and required revision. The KPro back plate in this particular patient had no holes that could allow for nutrition. This may have contributed to the melt24 (Table 1). The polished polymethyl methacrylate maintained its clarity over time (Figure 3B). There were no bacterial infections after surgery, nor any visible herpetic recurrences in the cornea surrounding the KPro. One fungal infection was successfully treated, and that patient currently has a visual acuity of 20/60 at 14 months' follow-up.
In herpetic keratitis, the anatomical success of penetrating keratoplasty has been good but graft clarity has generally been low. In our 4 patients with chronic inflammation and ulcerated herpetic disease (patients 1, 6C, 8, and 9), a standard corneal transplantation without a KPro might have initially achieved the same therapeutic result. However, given the neurotrophic nature of these corneas, the likelihood of repeat ulceration would probably be high. The addition of a central clear window via a KPro should result in better postoperative vision. In our series, the eyes calmed down quickly within days, with immediate visual rehabilitation, and were without inflammation and ulceration at last follow-up. We postulate that the removal of the central corneal button reduces the release of inflammatory mediators. It is possible that this debulking approach, with a graft-KPro combination, should have wider use in severe keratitis.
In summary, this case series provides evidence for use of the KPro in patients with corneal grafts in whom a repeat corneal transplantation would be unlikely to restore long-term vision. We believe that herpetic keratitis is not a contraindication for such KPro use and that the KPro may be indicated in a number of patients.
Correspondence: Claes H. Dohlman, MD, PhD, 243 Charles St, Boston, MA 02114 (firstname.lastname@example.org).
Submitted for Publication: August 26, 2006; final revision received November 6, 2006; accepted November 7, 2006.
Financial Disclosure: None reported.
Funding/Support: This study was supported by Alcon Research Institute and by the Massachusetts Eye and Ear Infirmary.
Previous Presentation: This study was presented at the 2006 Association for Research in Vision and Ophthalmology Meeting; May 4, 2006; Fort Lauderdale, Fla.