Microkeratome engaging a human cornea(long arrow) mounted on an artificial anterior chamber. Saline infusion andmanometer are attached (short arrow).
Our laboratory keratoprosthesis (A)is modeled directly after the AlphaCor keratoprosthesis (B). It has the samedimensions and is made of a similar polymer; however, the rim is not porous.
The flap is reflected (black arrow)and a 2.5- or 3.0-mm central trephination is made through the posterior stromalbed (white arrow).
Two cardinal sutures are placed at180° and 90° from the hinge (black arrow). The model keratoprosthesis(white arrow) is ready to be positioned.
Keratoprosthesis in place after posteriortrephination of 2.5 mm and anterior trephination of 3.5 mm. The edge of thekeratoprosthesis is faintly seen through the corneal flap (tip of arrow).
The area of leak and destabilizationalways occurred at the keratoprosthesis–anterior trephination interface.The keratoprosthesis (long arrow) prolapses through the anterior trephinationedge (short arrow).
Erb MH, Taban M, Barsam CA, Sweet PM, Chuck RS. Mechanical Stability of Microkeratome-Assisted Intracorneal KeratoprosthesisImplantation. Arch Ophthalmol. 2004;122(12):1839-1843. doi:10.1001/archopht.122.12.1839
To develop a laboratory model to study intracorneal keratoprosthesisimplantation.
A combination microkeratome and artificial anterior chamber system wasused to create a hinged lamellar keratectomy on 13 human corneas. After reflectingthe flap, the posterior stroma was trephined at either 2.5 or 3.0 mm. A modelkeratoprosthesis was positioned in the bed. The flap was sutured closed. Intrachamberpressure was increased, and wound leak pressure was recorded. The anteriorcorneal lamella was trephined at either 3.0 or 3.5 mm to expose the keratoprosthesis.Leak pressure was again determined.
After keratoprosthesis placement and prior to anterior trephination,all 13 corneas were watertight at maximum attainable intrachamber pressures.With posterior/anterior trephination combinations of 2.5/3.0 mm, 2.5/3.5 mm,or 3.0/3.5 mm, mean ± SD wound leak pressure occurred at95 ± 12 mm Hg, 32 ± 7 mm Hg, or 59 ± 12mm Hg, respectively (P<.01).
With a posterior trephination of 2.5 mm, there is significant keratoprosthesis-corneainterface destabilization between a 3.0- and 3.5-mm anterior trephination.For an anterior trephination of 3.5 mm, interface destabilization improvesby increasing the posterior trephination to 3.0 mm.
An intracorneal keratoprosthesis may be implanted using microkeratomeassistance. Our laboratory model provides a useful method for examining arange of posterior and anterior trephination diameters and their effects onthe mechanical stability of intracorneal keratoprosthesis placement.
Penetrating keratoplasty is the most frequently performed and most successfulhuman transplantation operation.1 The Eye BankAssociation of America estimates that more than 46 000 transplant procedureswere performed in 2000.2 The success rate ofpenetrating keratoplasty is greater than 90% for certain diseases, such askeratoconus.3 Overall, 5-year survival ratesfor primary corneal transplants range from 46.5% to 93%, depending on thepreoperative diagnosis.3- 18
Failure rates for corneal transplantation vary significantly dependingon the cause of the disease. There remains a subset of patients whose cornealdisorders make a successful outcome from penetrating keratoplasty highly unlikelyand who generally carry a poor prognosis, usually because of previous or ongoingchronic inflammation. Causes of poor outcome include Stevens-Johnson syndrome,chemical burns, ocular cicatricial pemphigoid, severe keratoconjunctivitissicca, stem cell deficiencies, and severe vascularization from other causes.19 Patients with recurrent graft failures also carrya poorer prognosis for favorable outcome following penetrating keratoplasty.20 A potential alternative for restoring vision in thesehigh-risk patients is implantation of an artificial cornea, or keratoprosthesis.21
The concept of an artificial cornea was first suggested in 1789 by GuillaumePellier de Quengsy. More recently, since the 1950s, 2-piece collar-buttonkeratoprostheses, which are maintained in position by anterior and posteriorplates, have been researched, designed, and implanted in a few select patients.The Dohlman-Doane,19 Strampelli,22 andCardona23 keratoprostheses have become themost familiar models. Over the past 25 years, these collar-button keratoprostheseshave had some success, but they are fraught with a high incidence of complications,including extrusion, corneal melting, infection, retroprosthetic membraneformation, and glaucoma.20 In the 1990s, new-generationkeratoprostheses made with biomaterials with improved biological integrationof the prosthetic material were introduced.24 Thesenew-generation keratoprostheses are 1-piece, intracorneal designs.
Recently, Hicks et al21 and Crawfordet al25 have reported clinical trial resultsof their artificial cornea, the Chirila Keratoprosthesis, now renamed AlphaCor(Argus Biomedical, Perth, Australia). The AlphaCor keratoprosthesis has recentlybeen approved by the Food and Drug Administration and is now commerciallyavailable in the United States. The AlphaCor is made from poly(2-hydroxyethylmethacrylate) (PHEMA) and has a fused optic-and-skirt design. The optic coreis transparent. The peripheral sponge skirt is opaque and porous and allowsbiointegration into the host cornea by ingrowth of stromal fibroblasts.20,21,26,27
The AlphaCor keratoprosthesis is implanted in a corneal stromal lamellarpocket.21,24 A 360° conjunctivalperitomy is made, and the epithelium is debrided. A superior 180° paralimbalincision is extended at 50% depth into the corneal stroma to create a cornealflap over the superior half of the cornea. The dissection is continued asan intralamellar pocket of 3.5-mm radius within the inferior cornea. The flapis retracted, and a central trephination is made in the posterior bed. TheAlphaCor is placed in the bed, and the superior paralimbal incision is closedwith interrupted 10-0 nylon sutures. A Gunderson conjunctival flap is createdto cover the entire corneal surface. After 8 to 12 weeks, the conjunctivaand anterior corneal lamella are trephined to expose the AlphaCor optic.
As with any new procedure, the optimal parameters for the surgical techniqueare continually being refined. We have developed a laboratory model to studythe implantation of an intracorneal keratoprosthesis using an artificial anteriorchamber. With this laboratory model, we investigate a new surgical techniquefor keratoprosthesis implantation utilizing microkeratome assistance. Additionally,we examine a range of posterior and anterior trephination diameters and theireffects on the mechanical stability of intracorneal keratoprosthesis placement.
A combination manual microkeratome and artificial anterior chamber system(ALTK System; Moria/Microtech, Doylestown, Pa) was used to perform the hingedanterior lamellar keratectomy (Figure 1).The operation of the system has been previously described.28 Anormal saline solution infusion system was connected via a 3-way connectionto the artificial anterior chamber and a digital manometer (Netech Corp, Hicksville,NY). After a corneoscleral rim was secured onto the artificial anterior chamber,the intrachamber pressure was raised or lowered by raising or lowering theinfusion bottle.
Our laboratory keratoprosthesis (Flexlens X-Cel; Walman Co, Duluth,Ga) has the same dimensions as the AlphaCor keratoprosthesis and is made ofa similar polymer; however, the rim is not porous. The material of our laboratorykeratoprosthesis is methafilcon A, a hydrophilic random copolymer of 2-hydroxyethylmethacrylate (HEMA) and methacrylic acid cross-linked with ethylene glycoldimethacrylate. The model keratoprosthesis’ dimensions are diameterof 7 mm, thickness of 0.5 mm, base curve of 8.5 mm, refractive index of 1.4153,and power of +42 diopters (Figure 2).
After approval by the University of California, Irvine institutionalreview board/ethics committee, 13 corneoscleral rims not suitable for cornealtransplantation and preserved in Optisol were obtained (Donor Network of Arizona,Phoenix). The mean ± SD age at the time of death of the 4women and 9 men was 67 ± 10 years (range, 42-75 years).
The corneoscleral rims were mounted on the artificial anterior chamberwith an attached manometer. Prior to the hinged anterior lamellar keratectomy,the intrachamber pressure was increased to more than 65 mm Hg by raising theinfusion bottle (mean ± SD, 71 ± 3 mm Hg).The hinged-flap anterior lamellar keratectomy was made using a manual microkeratome(LSK One; Moria/Microtech) with a 300-μm-thick head and a blade angle of25°. From our previous work with the combination artificial anterior chamberand manual microkeratome system, the actual mean ± SD depthsof the cut with a 300-μm head are 244.4 ± 38.3 μm and296.5 ± 31.3 μm at intrachamber pressures of 53.5 ± 2.7mm Hg and 95.8 ± 4.8 mm Hg, respectively.28,29
The flap was reflected after the diameter was measured with calipers.A handheld trephine with a diameter of either 2.5 or 3.0 mm was centered inthe posterior stromal bed. The trephine was advanced until perforation. Theposterior lamellar button was excised using forceps and corneal scissors (Figure 3). The flap was replaced, and 2 cardinal10-0 nylon sutures were placed at 180° and 90° from the hinge.
The model keratoprosthesis was then positioned in the bed (Figure 4). A third cardinal 10-0 nylon suturewas placed at 270° from the hinge. The flap was secured with a total of15 interrupted 10-0 nylon sutures to assure watertight wound closure. Intrachamberpressure was increased to a maximum by slowly raising the height of the infusionbottle, which was eventually limited by ceiling height. All 13 corneas werewatertight at the maximum attainable mean ± SD intrachamberpressure of 103 ± 8 mm Hg (range, 81-110 mm Hg).
The intrachamber pressure was then reset to approximate physiologiclevels, 17 to 21 mm Hg, by lowering the height of the infusion bottle. A handheldtrephine with a diameter of either 3.0 or 3.5 mm was centered on the anteriorcorneal lamella and advanced. The anterior lamellar button was excised usingforceps and corneal scissors to expose the keratoprosthesis (Figure 5). Intrachamber pressure was again elevated progressivelyby raising the height of the infusion bottle while the cornea was observedfor leaks. The leak pressure was recorded.
For comparison between groups, t test was used.A P value <.05 was considered statistically significant.
After hinged anterior lamellar keratectomy, posterior lamellar trephination,keratoprosthesis placement, and sutured flap, and prior to anterior trephination,all 13 corneas maintained watertight seals at the maximum attainable mean ± SDintrachamber pressure of 103 ± 8 mm Hg (range, 81-110 mmHg). The maximum pressure was limited by the ceiling height when elevatingthe infusion bottle.
With a posterior trephination of 2.5 mm and an anterior trephinationof 3.0 mm (n = 4), mean ± SD leak pressure occurredat 95 ± 12 mm Hg (range, 81-109 mm Hg). These corneas withkeratoprostheses were extremely stable, with 3 of 4 corneas exhibiting noleakage at the maximum intrachamber pressure of 81 mm Hg or more and 1 of4 corneas leaking at 92 mm Hg. In comparison, with the posterior trephinationmaintained at 2.5 mm and the anterior trephination increased to 3.5 mm (n = 4),the corneas and keratoprostheses were much less stable and the leak pressuremuch lower, occurring at a mean ± SD of 32 ± 7mm Hg (range, 25-40 mm Hg; P<.001). Finally, whenthe anterior trephination was maintained at 3.5 mm while the posterior trephinationwas increased to 3.0 mm (n = 5), the mean ± SDleak pressure increased to 59 ± 12 mm Hg (range, 39-68 mmHg; P = .004). Although the anterior trephinationof 3.5 mm was less stable than 3.0 mm for a posterior trephination of 2.5mm, stability could be improved by increasing the posterior trephination from2.5 to 3.0 mm (Table).
The area of leak and destabilization always occurred at the keratoprosthesis–anteriortrephination interface and not at the flap edge. The keratoprosthesis woulddislocate anteriorly through the anterior trephine opening (Figure 6). The model keratoprosthesis is made of HEMA and is soft,flexible, and deformable. Although the keratoprosthesis diameter was 7 mmand the anterior trephination diameter was 3.0 or 3.5 mm, the increase inintrachamber pressure distorted the keratoprosthesis’ soft, flexibleshape, thus allowing the keratoprosthesis to prolapse through the anteriortrephination. The mean ± SD flap diameter was 9.75 ± 0.4mm. Although corneal sutures served to reduce the effective diameter of thebed, there remained room for a small amount of lateral movement of the 7-mmkeratoprosthesis in the bed. This lateral movement of the keratoprosthesismay have allowed the keratoprosthesis edge to migrate near the anterior trephineopening, thus enabling the keratoprosthesis to prolapse through the anteriortrephine opening.
Over the past 25 years, the rigid collar-button keratoprostheses havehad some success but have been fraught with a high incidence of complications,including extrusion, corneal melting, infection, retroprosthetic membraneformation, and glaucoma.20 In the 1990s, new-generationkeratoprostheses made with biomaterials that improve the flexibility, biocompatibility,and hydrophilicity of the keratoprosthesis were introduced.24 Thesenew-generation keratoprostheses are 1-piece, intracorneal designs.
Intracorneal keratoprostheses may have the following advantages overthe collar-button designs: (1) the biointegration of the intracorneal keratoprosthesiswith the host cornea in combination with a 1-piece optic-and-skirt designmay prevent interface complications such as aqueous leakage, infection, extrusion,or epithelial downgrowth; (2) flexibility of the intracorneal keratoprosthesismay minimize mechanical stresses on the cornea and may allow possible tonometryin the future; (3) a larger optic allows ocular examinations, including fundusand optic nerve ophthalmoscopy, and enables a wider field of view to performvisual field testing; and (4) the nonadhesive nature of PHEMA and intracornealplacement may prevent retroprosthetic membrane formation.20,21,24
The AlphaCor intracorneal keratoprosthesis has recently been approvedby the Food and Drug Administration and is now commercially available. Multicenterclinical trials have been in progress for over 3 years in Australia and SoutheastAsia, with outcomes reported in 40 eyes.21,30 Clinicalimplantation of the AlphaCor keratoprosthesis is just beginning in the UnitedStates. As with any new procedure, the optimal parameters for the surgicaltechnique are continually being refined. In clinical trials, the AlphaCorwas implanted in a corneal stromal lamellar pocket.21,24 Theposterior trephination diameters ranged from 2.0 to 3.5 mm, and the anteriortrephination diameters ranged from 2.0 to 4.0 mm,24,25 withposterior/anterior trephinations of 3.0/3.0 mm in the most recent report.21
The development of an artificial anterior chamber to support human corneoscleralrims has enabled corneal surgeons to experiment with surgical techniques invitro. We have developed a laboratory model to study the implantation of amodel intracorneal keratoprosthesis using an artificial anterior chamber.With this laboratory model, we investigated a new surgical technique for keratoprosthesisimplantation using microkeratome assistance instead of manual dissection ofa lamellar pocket. We found that a model intracorneal keratoprosthesis maybe implanted using microkeratome assistance with relative ease.
We examined a range of posterior and anterior trephination diametersand their effects on the mechanical stability of intracorneal keratoprosthesisplacement. With a posterior trephination of 2.5 mm, there is significant wounddestabilization between a 3.0- and 3.5-mm anterior trephination. Althoughthe anterior trephination of 3.5 mm was less stable, interface stabilizationcould be much improved by increasing the posterior trephination from 2.5 to3.0 mm. Thus, it appears that a smaller mismatch between posterior and anteriortrephination diameters may play a more significant role in stabilization ofthe keratoprosthesis-cornea interface than the absolute anterior trephinationsize. Additional studies of varying trephination diameters are currently underway for confirmation.
The major drawback of this ex vivo model is that it does not allow forbiointegration of the keratoprosthesis. Thus, ex vivo, the keratoprosthesis-corneainterface will have less stability than in vivo. However, our laboratory modeldoes provide insight into the mechanical stability of intralamellar keratoprosthesisplacement, which may be applied to clinical situations. First, if the mechanicalstability of the keratoprosthesis-cornea interface is maximized without anybiointegration ex vivo, then in vivo, there will be reduced mechanical stressesto the system after biointegration. This may be beneficial for long-term woundstability and implant extrusion rates. Second, although the absolute trephinationdiameters may not be directly transferable to a clinical situation, the laboratorymodel gives us a method to easily vary and evaluate anterior and posteriortrephination sizes. We found that mechanical stability decreases as the anteriortrephination size increases. However, with the larger anterior trephinationsize, mechanical stability could be improved by increasing the posterior trephinationsize.
With our model, we can facilitate intelligent keratoprosthesis implantationby varying posterior and anterior trephination diameters and evaluating theireffects on mechanical stability. The eventual goal is to maximize both anteriorand posterior trephination sizes to allow maximal field of vision and adequateocular examinations for optimal ophthalmic management.
Correspondence: Roy S. Chuck, MD, PhD, WilmerOphthalmological Institute, Johns Hopkins University, 3-127 Jefferson Bldg,600 N Wolfe St, Baltimore, MD 21287-9278 (firstname.lastname@example.org).
Submitted for Publication: September 26, 2003;final revision received March 29, 2004; accepted May 18, 2004.
Funding/Support: This study was supported bygrant EY00412-02A1 from the National Institutes of Health, Bethesda, Md (DrChuck).
Financial Disclosure: None.
Previous Presentation: This study was presentedin part at the annual meeting of the American Society of Cataract and RefractiveSurgery; May 2, 2004; San Diego, Calif.
Acknowledgment: We thank the Donor Networkof Arizona, Phoenix, for providing the human corneas.