Author Affiliations: Bascom Palmer Eye Institute,University of Miami School of Medicine, Miami, Fla (Drs Stoiber, Fernandez,Dubovy, Alfonso, and Parel and Ms Lamar); Department of Ophthalmology andOptometry, Paracelsus Private Medical University, Salzburg, Austria (Dr Stoiber);and Department of Ophthalmology and Optometry, University of Vienna, Vienna,Austria (Dr Kaminski).
To evaluate the biocompatibility of a novel nonpenetrating keratoprosthesis(supraDescemetic synthetic cornea) in a rabbit model.
Seven rabbits received a supraDescemetic synthetic cornea (7-mm diameter,350-μm-thick optical zone, 100-μm-thick peripheral flange) in theirhealthy right eyes. A surgical technique was developed that allowed implantationof the device on top of the bare Descemet membrane. Three rabbits receiveda supraDescemetic synthetic cornea made of hydroxyethyl methacrylate–methylmethacrylate26, 1 received a hydroxyethyl methacrylate–N-vinyl pyrrolidone75 mesoplant, and 3 wereimplanted with devices made of polymethyl methacrylate. All rabbits were euthanizedafter 8 weeks; the eyes were enucleated and examined by conventional histologicaland immunohistochemical evaluations.
All eyes became quiet within several days. The Descemet membrane remainedtransparent during the observation period. Indirect ophthalmoscopy performedthrough the prosthesis allowed accurate examination of the posterior pole.Histological evaluation of the implanted corneas displayed no signs of anacute or chronic inflammatory reaction to the supraDescemetic synthetic corneain 5 eyes; a few inflammatory cells were detected in the corneas of 2 rabbits.The interface between the Descemet membrane and the mesoplant displayed ingrowthof very thin (<10-μm) tissues colonized by keratocytes in 3 of the7 corneas.
This study validates the biocompatibility of this new type of nonpenetratingkeratoprosthesis. Because opening of the anterior chamber is not requiredwith the supraDescemetic synthetic cornea, the risk for intraocular infectionis minimal, and the implantation procedure is less traumatic compared witha penetrating device.
The implantation of a keratoprosthesis (KPro) is performed as a lastclinical attempt to restore vision in patients not amenable to conventionalcorneal transplantation or ocular surface reconstruction procedures. Varioustypes and materials of KPro’s with different methods of insertion havebeen tested and implanted in patients during the last decades, with varyingbut growing success.1- 11 Despiteimproved retention rates, postoperative complications, such as glaucoma, endophthalmitis,retroprosthetic membranes, expulsion of the implant, or sometimes total visionloss, remain problems with those penetrating devices.12 Recently,new soft and flexible materials for an artificial cornea were evaluated, tested,and further developed to avoid stress at the points of attachment that couldlead to stromal melting.13- 18 However,a certain risk for complications still persists because of the penetratingnature of these devices. A lamellar supraDescemetic synthetic cornea (sDSC)implant would theoretically minimize such risks because there is no need toenter the anterior chamber, leaving the Descemet membrane (DM) and the endotheliumintact.19 Thus, potentially disastrous complications,such as epithelial downgrowth leading to aqueous humor leaks and intraocularinfection, could be prevented. The results of a pilot animal study with anintrastromal, pre-Descemetic implant developed by our institute were encouraging.20,21 However, opacification of the remainingstroma underneath the synthetic cornea developed with time, compromising apotentially good visual outcome. Consequently, the implantation techniquewas changed, inserting the sDSC at maximal depth on a completely exposed DM.To our knowledge, the reaction of the DM to the direct contact of the polymerhas not been evaluated. Bioincompatibility of the KPro material leading toexcessive scarring and opacification in the DM-polymer interface would reducethe potential visual benefit, especially with this type of KPro. In this midtermstudy, we tested the response of rabbit corneal tissue to implantation ofsuch a synthetic cornea.
Three different materials were used for the synthetic corneas tested:polymethyl methacrylate (PMMA; n = 3), hydroxyethyl methacrylate–methylmethacrylate (HEMA-MMA; n = 3), and hydroxyethyl methacrylate–N-vinyl pyrrolidone (HEMA-NVP; n = 1), the latter2 with a water content of 26% and 75%, respectively. All implants were 7 mmin diameter and had the same design, adjusted to the dimensions of the rabbit’scornea (Figure 1): a thicker centraloptical zone was surrounded by a thinner outer flange with perforations, whichwould allow nutrition transfer and tissue ingrowth for improved fixation ofthe implant. The central part had a diameter of 4.5 mm and a thickness of350 μm. Both anterior and posterior curvatures measured 8 mm. The outerflange showed a thickness of 100 μm. The transition zone between the centraloptical part and the peripheral part was made conical. All implants were producedand provided by Cornéal Laboratoires (Paris, France).
Optical comparator shadow photographof a supraDescemetic synthetic cornea. A, Front view; the prosthesis skirtdisplays openings for nutrition transfer and tissue ingrowth. B, Side view.
The study was conducted in accordance with the Association for Researchin Vision and Ophthalmology Statement for the Use of Animals in Ophthalmicand Vision Research, and the study protocol was approved by the Universityof Miami School of Medicine Animal Care and Use Review Board.
Implantation of an artificial cornea was performed in the right eyesof 7 female New Zealand White rabbits with 3.0 to 4.7 kg in body weight. Theanimals were anesthetized with an intramuscular injection of a ketamine-xylazine-acepromazinemixture (35 mg/kg, 5 mg/kg, and 0.75 mg/kg, respectively).
Ultrasonic pachymetry was performed in the outermost corneal peripherynext to the limbus. A 7-mm curved incision along the limbus was performedfrom 9 to 11 o’clock in the clear cornea, at a calculated 85% of thecorneal depth using a diamond knife with double footplate and adjustable blade.An intralamellar dissection was made using a disposable knife (SatinCrescentKnife, bevel down; Alcon Laboratories Inc, Fort Worth, Tex) and a blunt curvedspatula (K3-3000; Katena Products Inc, Denville, NJ). A custom-made, round-shapedmetal plate with a diameter of 6.5 mm was inserted in the created pocket,followed by trephination of the corneal tissue above the plate using a handheldcustom-made 3.7-mm vacuum trephine (Figure 2A).The remaining stromal layers were removed by gentle tearing and cutting. Oncea small patch of the DM was exposed, a fine spatula shaped like blunt wire(Straight spatula, 0.25 mm; Rumex International Co, Miami, Fla) was carefullyslid along the plane between the DM and the remaining stroma. The stroma wasthen lifted and cut away until the DM was completely exposed to the edgesof the trephination (Figure 2B). Thesynthetic cornea was inserted via the incision; the optical part was positionedon top of the DM and the peripheral skirt placed in deep corneal stroma. ThreePMMA, 3 HEMA-MMA26, and 1 HEMA-NVP75 devices were implantedin this series. The incision was closed using 3 interrupted 10-0 nylon sutures.Surgical procedures were performed by 2 of the authors (J.S., V.F.). Dexamethasonewith neomycin and polymyxin B ointment (Maxitrol; Alcon Laboratories Inc)was topically applied twice daily for 3 days, starting at the end of surgery.No local or systemic treatment was administered after the third postoperativeday. Slitlamp examination was performed on the first 3 postoperative daysand on a weekly basis thereafter. Slitlamp photography was done during eachexamination. All sutures were removed 1 week after surgery. Prior to euthanasiaof the rabbits, the clarity of the remaining tissue was judged, and indirectophthalmoscopy of the posterior segment was performed, using a 90-diopterlens.
Surgical procedure. A, A thin metalplate is inserted in a corneal pocket, and trephination is performed witha 3.7-mm trephine. B, The Descemet membrane is exposed, and the remainingstromal layers are gently removed.
Eight weeks after surgery, all animals were euthanized with a lethaldose of pentobarbital and phenytoin (Eutasol; Diamond Animal Health Inc, DesMoines, Iowa) administered intravenously through the marginal ear vein.
The operated eye of each rabbit was immediately enucleated and placedin 10% buffered formaldehyde for at least 24 hours. Corneas, including themesoplant, were removed and further processed for histological evaluation.Sections were stained with hematoxylin-eosin, periodic acid-Schiff, and Massontrichrome techniques.
To determine the actual extent of epithelial downgrowth into the cornea-polymerinterface in each animal, immunohistochemical evaluation was performed usinga pan-specific cocktail of antibodies displaying primary reactivity with cytokeratins:AE1/AE3, 34βE12 (Dako Corp, Carpinteria, Calif), and CAM5.2 (BD Biosciences,San Jose, Calif).
To determine if the DM plane was actually reached using the techniquepreviously described, surgery was performed also on the contralateral lefteyes of 4 rabbits immediately before planned euthanasia, the fragility ofthe bare DM not permitting prolonged follow-up unless protected by the syntheticcornea. Three of those corneas were processed for light microscopic analysis.The surface characteristics of the exposed membrane were studied on the fourtheye using scanning electron microscopy.
All rabbits showed an uneventful initial postoperative phase. All eyesbecame quiet on the first or second postoperative day. The DM was found tobe firmly attached to the posterior surface of the implant and was detectableby careful slitlamp examination. The corneal tissue overlying the outer flangeshowed slight edema postoperatively that resolved within 2 weeks, resultingin a more continuous transition between the cornea and the implant’soptic. Neovascularization of the cornea did not occur in any of the operatedeyes within the observation period. The implants maintained their opticaltransparency and displayed no alterations in their outer surface (Figure 3). The last layer of the corneal stromacould not be removed completely in the area of the trephination edge at thetime of surgery in each of the cases. Those remnants were detectable postoperativelybetween the implant and the DM, appearing as a narrow circular band next tothe edge of the trephination. With time, opacification of these remnants slightlyincreased, displaying a very slow progression toward the optical center. Thecentral part of the denuded DM remained transparent in all rabbits that receivedPMMA or HEMA-MMA26 implants. When we inserted the HEMA-NVP75 implant, a small bundle of loose stromal tissue was accidentallyimplanted between the synthetic cornea and the DM, which resulted in minoropacification in its surrounding areas with time.
Implanted supraDescemetic syntheticcornea. A, Oblique view: clear optical center and remnants of stromal tissueunderneath optics periphery; hydroxyethyl methacrylate–methyl methacrylate26, postoperative day 23. B, Front view: polymethyl methacrylate, postoperativeday 38. C, Side view: hydroxyethyl methacrylate–methyl methacrylate26, postoperative day 3.
The cornea above the sDSC flange showed continuous retraction with time,leaving the inner part of the flange uncovered. This was the case for bothPMMA and the softer hydrophilic materials tested, occurring mostly at the6- and 12-o’clock positions. Tissue retraction did not reach the flangeopenings in any of the cases.
Indirect ophthalmoscopy performed prior to euthanasia allowed very accurateexamination of the posterior pole with vascular details discernible throughthe artificial cornea (Figure 4), evenin those few rabbits that displayed minor opacification of the DM-sDSC interface.
Indirect ophthalmoscopy (90-diopterlens) performed through the supraDescemetic synthetic cornea: detailed viewof posterior pole, optic nerve.
All mesoplants appeared to be well tolerated by the corneal tissue.A few inflammatory cells (neutrophils) could be detected at the trephinationedge of 1 cornea that had received a mesoplant; 1 cornea with a HEMA-NVP75 mesoplant showed some clusters of inflammatory cells on the endothelium.None of the other corneas displayed any noticeable signs of acute or chronicinflammatory reaction. Epithelial thinning up to 2 layers was found on topof areas above the sDSC flange. Trephination edges, on the other hand, showedmarked thickening of the epithelial layer; however, epithelial coverage couldnot be detected on the anterior surface of the optic itself. Immunohistochemicalexamination displayed a tendency of the corneal epithelium to grow backwardsonto the upper surface of the flange (Figure 5A). In 1 of the eyes, epithelial ingrowth reached the posteriorsurface of the flange (Figure 5B) butdid not extend into the interface between the DM and the sDSC optic. The flangeopenings were filled with newly synthesized collagen tissue, lined partiallyby epithelial cells (Figure 6A). Theinterface between the DM and the mesoplant displayed ingrowth of very thin(<10-μm) keratocyte-containing tissue in 3 of the 7 corneas (1 ×PMMA, 1 × HEMA-MMA26, and 1 × HEMA-NVP75).In all others, the DM was still found to be in direct contact with the polymer(Figure 6B, C). The endothelium wasfound to be healthy and normal in all of the rabbits.
Immunohistochemical analysis of epithelialingrowth with anticytokeratin antibodies; corneal tissue overlying the supraDescemeticsynthetic cornea flange shows thickening of the epithelium at the trephinationedge; arrows indicate the frontier of epithelial ingrowth. A, Hydroxyethylmethacrylate–N-vinyl pyrrolidone75.B, Hydroxyethyl methacrylate–methyl methacrylate26.
Histological evaluation. A, Areaof flange opening, filled with new collagen (asterisk) and epithelium (arrows);the mesoplant is partly dissolved because of histologic processing (circle);hydroxyethyl methacrylate–N-vinyl pyrrolidone75, Masson trichrome × 100. B, Bare Descemet membrane with endothelium;hydroxyethyl methacrylate–methyl methacrylate26, periodicacid-Schiff × 400. C, Ingrowth of a thin layer of fibrous tissue (arrow)in DM-mesoplant interface; hydroxyethyl methacrylate–N-vinyl pyrrolidone75 × 400.
Histological examination showed a completely exposed DM, without anyresidual stromal layer remaining on top. With scanning electron microscopy,the outer surface of the DM appears as an extremely smooth and even structure(Figure 7); even with high magnification(×10 000), no structural roughness could be discovered.
Scanning electron microscopy of arabbit cornea following surgical procedure. The Descemet membrane is completelyexposed.
The concept of implanting a lamellar KPro was described by Stone fivedecades ago.22,23 His deviceswere made of PMMA and were implanted in rabbit eyes using a 2-stage procedure.The KPro’s were inserted in a corneal pocket, trephining the centralupper part of the cornea after a sufficient time interval when firm fibrosishad developed at the implant’s perforated periphery. In this series,a 1-stage procedure was used; no mesoplant was lost spontaneously in the earlypostoperative phase as fibrous ingrowth in the flange openings occurred withina few weeks, providing excellent fixation of the sDSC within the host cornea.However, ongoing retraction of the corneal tissue above the peripheral flangewas noticed, jeopardizing the stability of the mesoplant with time. With theKPro periphery acting as a barrier, a disrupted or restricted nutrient flowto keratocytes anterior to the flange could eventually compromise their homeostasis.In addition, enzymatic degradation of the collagen24 couldfinally lead to the observed tissue reduction. This phenomenon was found lessfrequently and was delayed in cases where the trephination edge became vascularized(J.S., P.D.L., J-M.P., and E.A., unpublished data, 2002). This suggests apotential benefit of neovascularization for the long-term survival of cornealtissue located above a material with restricted permeability. Vascularizedcorneas were found to be much less likely to melt after KPro implantationcompared with avascular corneas,25 probablybecause vessels provide proteolytic enzyme inhibitors and nutrients essentialfor tissue preservation. However, mesoplant presence has not stimulated neovascularizationwithin the observation period in this series. Because vascularization of thefixation material seems to be desirable, coverage with a conjunctival flapmight prevent melting in cases of nonvascularized host corneas. A centralopening above the KPro optics could then be performed following integrationof the conjunctiva, similar to the procedure proposed for the AlphaCor.18 The material itself was tolerated very well by thehost tissue, and a toxic reaction, as previously reported for other polymers,26 could not be detected. Biocompatibility of the softpolymers used in this study was tested in an earlier study.27 Smalldisks of the materials implanted in intrastromal pockets demonstrated somecellular reactions at the sites of mechanical stress, but no major cellularreaction against the implant material itself could be detected. Although PMMAhas good optical property, it has disadvantages when used as material forsynthetic corneas because of its rigid nature. Both HEMA-MMA26 andHEMA-NVP75 are found to be soft and flexible. Biocompatibilityhas been defined as the ability of a material to perform with an appropriateresponse in a specific application. Although the International Organizationfor Standardization recommends longer test periods, up to 78 weeks for long-termbiocompatibility testing of novel biomaterials, our results strongly indicatethat those hydrophilic polymers are well tolerated by the corneal stroma andare suitable as material for sDSC KPro.
Ingrowth of a thin membrane of fibrous tissue underneath the opticsoccurred in several rabbits of our series even within the relatively shortobservation period of 8 weeks. One might also speculate on the incidence andintensity of fibrous ingrowth after a longer period. However, because theregenerative capacity of the rabbit’s eye tissue is found to be greatercompared with that of human eye tissue,28,29 scarringof the DM-polymer interface might actually be less frequent when the sDSCis implanted in patients. Following opacification underneath a synthetic polymer,the actual visual acuity that could be achieved with such a lamellar KPromight therefore be lower compared with those with penetrating devices. Buton the other hand, the increase in safety in the long run—as there isno need to enter the anterior chamber of the eye—might compensate fora potentially lower but still useful visual acuity. Retinal detachments couldnot be found in any of our rabbits at the end of the observation period, incontrast to animal studies on perforating KPro’s, which reported retinaldetachments in a high number of rabbits following implantation.30 Incases of severe interface scarring, removal of this membrane with both theDM and the endothelial layer might be an option for regaining visual acuity,using instruments similar to those already described for deep lamellar endothelialkeratoplasty.31 Another surgical approach forsDSC implantation was tested in an earlier study, inserting the device insuch a way that its complete base (base and flange)was located on the bare DM.32 This techniquecould theoretically postpone fibroblast ingrowth into the DM-polymer interfacefrom the side. However, those mesoplants showed a postoperative tendency fordecentration and instability, most likely because tissue ingrowth into flangeopenings could just occur for one (the stromal) side and might therefore notbe as strong as with the implantation technique previously described.
Although the results of this study are promising, one has to considerthat all mesoplants have so far been implanted in healthy and clear corneas.Because the biological response might differ in diseased and vascularizedcorneas (eg, after corneal burns), it seems advisable to further study sDSCreliability and long-term stability in an appropriate animal model for theseconditions before proceeding to human trials.
Correspondence: Jean-Marie Parel, PhD, OphthalmicBiophysics Center, Bascom Palmer Eye Institute, 1638 NW 10th Ave, Miami, FL33136 (firstname.lastname@example.org).
Submitted for Publication: August 25, 2003;final revision received March 9, 2004; accepted April 29, 2004.
Funding/Support: The study was supported bygrant P30 EY14801 from the National Institutes of Health, Bethesda, Md; grantBMH4-CT97-9507 from the European Project, Brussels, Belgium; the AustrianScience Fund, Vienna, Austria; the Henri and Flore Lesieur Foundation, WestPalm Beach, Fla; the Florida Lions Eye Bank, Miami, Fla; and by an unrestrictedgrant from Research to Prevent Blindness, New York, NY.
Financial Disclosure: None.
Previous Presentation: This study was presentedin part at the annual meeting of the Association for Research in Vision andOphthalmology, Ft Lauderdale, Fla, May 8, 2003, and at the 5th KPro–8thIOSS Joint Meeting, Miami, Fla, May 9, 2003.
Acknowledgment: Emmanuel Lacombe, MD, and BernardDuchesne, MD, participated in the design of the supraDescemetic syntheticcornea and preliminary pilot studies; Franck Villain, PhD, headed the polymerchemistry synthesis and analysis; Izuru Nose, BSEE, William Lee, and DavidDenham, MSME, PE, fabricated the new surgical instruments and apparatuses;David Chin-Yee provided the shadow photography and optical analysis; MagdaCeldran and Sue Decker, BA, did the tissue preparation for light microscopyand scanning electron microscopy; Eleut Hernandez, LAT, provided daily animalcare; Reva Hurtes, MA, very kindly edited the manuscript; and Waldemar Kitaof Cornéal Laboratoires, Paris, France, generously provided all syntheticcorneas and technical support free of charge.
Thank you for submitting a comment on this article. It will be reviewed by JAMA Ophthalmology editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Register and get free email Table of Contents alerts, saved searches, PowerPoint downloads, CME quizzes, and more
Subscribe for full-text access to content from 1998 forward and a host of useful features
Activate your current subscription (AMA members and current subscribers)
Purchase Online Access to this article for 24 hours
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 9
Customize your page view by dragging & repositioning the boxes below.
and access these and other features:
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.