Intracorneal Inlay Complicated by Intrastromal Epithelial Opacification

Objective  To report epithelial perilenticular opacity as a new complication ofintracorneal inlay implantation for the correction of hyperopia.

Design  Prospective observational case series.

Participants  Eleven eyes of 7 patients underwent intracorneal inlay implantationfor the correction of hyperopia.

Methods  Intracorneal inlays were implanted onto the stromal bed by using a microkeratomecut to create an inferior hinged corneal flap.

Main Outcome Measures  Postoperative complication occurrence of intracorneal perilenticularopacity, microbiological laboratory analysis, histopathological analysis,and confocal microscopy study.

Results  Of 11 implanted eyes, 5 showed diffuse perilenticular opacity of varyingintensity that was unresponsive to steroid use following intracorneal inlayimplantation. All patients had moderate to severe loss of best-corrected visualacuity. The inlays showed deposits at the edge and on the surface. Confocalmicroscopy in all eyes produced images compatible with the confocal morphologicfeatures of epithelial cells. Explantation of inlays was performed in 5 eyes.The histopathologic study showed the presence of epithelial cells, and microbiologicalanalysis and cultures were negative for bacteria, fungi, and mycobacteria.

Conclusion  Epithelial perilenticular opacity is a new and serious complicationin patients with intracorneal inlay implantation for the correction of hyperopia.

Many methods for the surgical correction of hyperopia have been usedin recent years. Some of the most common corneal refractive procedures usedfor the correction of hyperopia are laser-assisted in situ keratomileusis,photorefractive radial keratotomy, laser thermokeratoplasty, and conductivekeratoplasty. Now the numbers of clinical indications are increasing, withhundreds of thousands of patients operated on around the world. One of therecently proposed methods is additive refractive keratoplasty. This term refersto a procedure in which a foreign material, either biological or synthetic,is added to the corneal tissue to modify the refractive condition of the eye.This method creates the potential for reversibility; if necessary the implantmay be removed, and other treatment may still be available to the patient.

Synthetic stromal inlays or intracorneal implants have been investigatedfor nearly half a century. Barraquer¹ was thefirst, in 1949, followed by many researchers who used an implantable inlayto modify the refraction on the cornea.^2-4 Theyused flint glass and plexiglass in their studies. The implants manufacturedfrom these materials caused anterior stromal necrosis followed by extrusionin eyes implanted with this inlay. The limitations of this impermeable membranedeveloped in previous studies could be avoided by the use of hydrogel. Withits permeability, hydrogel is similar to the corneal stroma, allowing theexchange of water and nutrients between the posterior and anterior layersof the cornea to maintain normal corneal physiologic characteristics.^5,6 The first hydrogel to be evaluatedfor refractive keratoplasty was hydroxyethyl methacrylate, by Dohlman⁷ in 1967 and later on by other researchers in the areaof refractive keratoplasty.^8-11

Currently, the method by which intracorneal inlays are implanted withinthe cornea consists of creating a corneal flap with an automated microkeratomefollowed by inlay implantation onto the corneal stroma. It is much easierto predict and maintain a hinged flap of constant thickness and a desiredcorneal depth for inlay placement. Previous experimental studies have demonstratedthat hydrogel lenses need to be placed at a depth between 36% and 60% of thecorneal thickness for success.^12,13

To our knowledge, no published studies report the implantation of thisinlay for the correction of hyperopia in human eyes. The only published studieshave been performed on primate eyes. A variety of complications occurred postoperativelyfollowing implantation of the hydrogel model in animals. These included cornealopacification,^12,14 epithelialand stromal thinning,^9,15,16 keratocyteand fibroblastic changes with or without intracorneal deposits,^15,16 epithelialingrowth,¹⁷ decentration,¹⁸ anteriorcorneal necrosis, and lens extrusion.¹⁹

This article describes a new complication arising after implantationof the latest generation of hydrogel corneal inlays for the correction ofhyperopia in human eyes. We have termed this complication epithelial perilenticular opacity.

The hydrogel intracorneal inlays were implanted in 11 hyperopic eyesof 7 patients (4 women and 3 men). The mean ± SD age was 42.3 ±8.1 years (range, 24-60 years). Mean ± SD preoperative cycloplegichyperopia was 4.6± 1.1 diopters (D) (range, 2.5-6.0 D). All patientshad less than 1 D of keratometric astigmatism. Mean ± SD preoperativeuncorrected visual acuity (UCVA) was 20/80 ± 20/200 (range, 20/40-20/200).Mean ± SD preoperative best spectacle-corrected visual acuity (BSCVA)was 20/25 ± 20/100 (range, 20/20-20/40). The patients showed no systemicor ocular health problems. Results of preoperative biomicroscopy examinationof the anterior segment and fundus were normal. The preoperative evaluationalso included corneal pachymetry using ultrasonic pachymetry (DGH-500 pachymeter;DGH Technology, Inc, Exton, Pa) and the determination of scotopic pupil size(Colvard Pupillometer; Oasis Medical, Inc, Glendora, Calif). Approval fromthe Ethical Board Committee was obtained, and all patients read and signedan informed consent document explaining the surgical procedure and possiblerisks in accordance with the Declaration of Helsinki.

Nine eyes (5 patients) were implanted with inlays (Permalens; PermaVision,Anamed Inc, Lake Forest, Calif). The Permalens inlay is made of hydrogel witha water content of more than 70% and a refractive index of 1.3. The thicknessis from 20 to 50 µm, and its diameter is 4.5 mm. The power of the inlayranges from +3.00 D to +6.00 D. The other 2 eyes (2 patients) were implantedwith another inlay design made of hydrogel with a water content of 78% anda refractive index of approximately 1.3. The thickness is between 48 and 92µm, and the diameter is from 4.75 to 5.25 mm. The power of the inlayranges from +2.00 D to +8.00 D.

All patients underwent the same surgical protocol to implant the cornealinlay. An automated microkeratome (M2; Moria, Antony, France) was used totentatively create a 180-µm corneal flap with a diameter of 8.5 mm anda 4-mm inferior hinged corneal flap. During the procedure, corneal pachymetrywas used to measure the cornea and residual stromal bed by using an ultrasonicpachymeter. Following the manufacturer's indications, a "dry technique" wasused for the implantation of the inlay. Hence, the interface was not irrigatedafter the microkeratome cut or the implantation. Immediately after the microkeratomecut was performed, the stromal bed was dried using a sponge (Merocel; OasisMedical, Inc), and the inlay was placed on the pupil zone by means of a specificmanual vacuum device as recommended by the manufacturer. The hinged cornealflap was replaced onto the bed using a nonstitch technique. The gutter aroundthe edge of the flap was also dried with a sponge. After the fluid was squeezedfrom the interface, the flap was left to settle for 2 minutes. All surgicalprocedures were successful. At the end of the procedure, the eye was occludedfor 24 hours. Postoperative treatment included 0.3% ofloxacin 4 times perday for 1 week and combined tobramycin and 0.1% dexamethasone 4 times a dayfor 1 week. Four cases required inlay reposition owing to different degreesof decentration. Follow-up was carried out at 1 to 2 days, 1 week, 1 month,3 months, and 6 months. Figure 1 showsa successful inlay implant.

Of 11 implanted eyes, 5 postoperatively developed a form of perilenticularcorneal opacity thought to be directly related to the refractive inlay. Theopacity was evident after 1 week of follow-up in all cases. The inlays developeddeposits in and around the surface. The appearance was very similar to thatof diffuse lamellar keratitis (DLK), leading to the initial diagnosis, butthe corneal opacity was limited to the edges of the inlay. Otherwise, therest of the cornea was not affected by opacity (Figure 2, Figure 3, Figure 4, and Figure 5). Three cases had previously required flap lifting andinlay repositioning because of inlay decentration shortly after the firstimplantation. All cases were symptomatic. In no case was there evidence ofepithelial ingrowth from the edge of the flap. Results of fluorescein stainingof the gutter were positive for the first 3 days but became negative afterward.Peripheral flap interface epithelial nets or sheets of cells were not observedin any of the cases in the present series, including those in which perilenticularopacification did not develop.

All patients complained of similar symptoms: night glare, moderate photophobia,starbursts, and blurry vision. They received treatment with 0.3% ofloxacinand with tobramycin and dexamethasone 4 times a day. A study of the cornealstroma using confocal microscopy (ASL model 500; Advanced Scanning, New Orleans,La) was performed the second postoperative week in all eyes (5/5) with similarresults. Scanning was performed from the epithelium to the endothelium, payingspecial attention to the stroma interface and inlay edges. The corneal epitheliumand the stroma behind the basal membrane were normal. The keratocytes of theanterior stroma were activated, and a zone of apoptotic keratocytes was foundon the anterior inlay surface. At this level, stromal edema, nonmetallic particles,and the extracellular matrix were clearly seen using this technique (Figure 6). Just behind the inlay, epithelialcells were found lying over the stromal bed with nonmetallic particles andepithelioid cells (Figure 7). Weobserved many epithelial cells in the posterior inlay and around the edgeand epithelioid cells. The edge of the corneal inlay is clearly seen in Figure 8. The posterior stroma and endotheliumwere normal, and this finding was similar in all cases. No improvement inbiomicroscopic appearance or in the patient's symptoms was noticed after 3weeks of treatment. Central flap thickness was deducted by subtraction ofthe perspective pachymetry measurement from the stromal measurement obtainedimmediately after flap lifting. The mean ± SD thickness was 169 ±20 µm. There were no significant differences in flap thickness betweenpatients who did or did not develop postoperative perilenticular opacity.

Explantation of the inlay was performed after 1 month of follow-up inall 5 cases. During the explantation procedure, when the flap was lifted,a thin membrane was observed between the posterior inlay surface and the stromalbed in 5 of 5 eyes (Figure 9). Thismembrane was carefully removed and sent for histopathologic and microbiologicalstudy. After explantation, an intensive topical steroid (0.1% dexamethasone)and an antibiotic (ofloxacin) were used for 5 days. Corneal transparency improvedin all eyes, although 2 of 5 eyes still showed mild stromal peripheral opacityaround the central cornea; however, this did not interfere with central cornealtransparency. The membrane and inlay were inoculated using several media includingblood, chocolate, MacConkey agar, Thayer-Martin agar, Lowenstein-Jensen medium,and Sabouraud agar for bacterial, mycobacterial, and fungal cultures. Gramand Giemsa stains and histopathologic studies were also carried out. Microbiologicalanalysis and cultures of corneal specimens from all patients were negativefor bacteria, fungi, and mycobacteria. Although pathologic study was performedfor all explanted inlays together with adherent tissues attached to the inlayobtained during surgery, only 2 pathologic results were obtained because theother 3 samples had insufficient material to be processed. There was 1 caseof abnormalities with the Permalens inlay and 1 for the other type of inlay.In all cases, the postoperative BCVA and refraction were nearly the same asinitial preoperative levels after 3 months of follow-up. Removal of the inlayshowed a positive effect on the recovery of corneal transparency, which returnedto nearly normal levels in all cases. In 3 cases a faint ring-shaped peripheralopacity, which did not affect the central cornea, was still visible 6 monthsafter explantation.

One case involved a 39-year-old man who had hyperopia in his right eye.The UCVA was 20/125 OD, and the BSCVA was 20/25 OD. The refraction was +6.00D of spherical defect and −0.5 D × 50° of cylinder defect,and his scotopic pupil size was 4.0 mm. A +6.00 D intracorneal hydrogel inlaywas selected for insertion in this case. The surgery was performed as previouslydescribed.

Five days after surgery the intracorneal inlay had migrated 1.0 mm temporally,so the patient underwent surgery to reposition the inlay by lifting the flap.One week later, the patient complained of photophobia, starbursts, blurryvision, and night glare. The UCVA was 20/100 OD, and the BSCVA was 20/50 ODwith +3.50 D of spherical defect and –0.5 D × 90° of cylinderdefect. There were several whitish deposits on the edge of the inlay, whichwas decentrated 1.0 mm temporally (Figure3). Mild corneal edema and moderate diffuse haze were also apparent.There was no anterior chamber reaction. Initially the patient was diagnosedas having DLK, and treatment was started with 0.1% fluorometholone every 4hours and cyclopentolate hydrochloride every 8 hours. The patient showed noimprovement, and we decided to explant the intracorneal inlay after 3 weeksof observation.

Confocal microscopy was performed using a tandem scanning confocal microscope(ASL model 500). The corneal epithelium and the stroma behind the basal membranewere normal. The inlay was found to be between 134 and 173 µm. Keratocyteactivation and apoptosis anterior to the inlay surface were found as describedpreviously. Epithelial cells were found lying over the stromal bed with fewnonmetallic particles and epithelioid cells (Figure 7). We observed many epithelial cells in the posterior inlayand around the edge and epithelioid cells. The posterior stroma and endotheliumwere normal.

The patient improved rapidly, with regression of the ocular inflammationand corneal deposits; the topical steroids were slowly reduced. After 2 weeksof treatment the UCVA was 20/125 OD, and the patient's BSCVA was 20/40 OD.The refraction was +6.00 D of spherical defect and −2.00 D × 75°of cylinder defect. Biomicroscopy examination showed peripheral leukoma aroundthe corneal flap. Because of the loss of 4 lines of BSCVA as well as the patient'sreported symptoms, the inlay was finally removed and sent for microbiologicaland pathologic study.

Microbiological analysis and cultures of the corneal specimen showedno abnormalities. Histopathologic studies showed that the membrane consistedof epithelial cells (Figure 10).There was no presence of granulocytes such as polymorphonuclear cells, eosinophils,or basophils, ruling out acute infection. The inlay dissolved during the histopathologicprocess.

The correction of moderate and high hyperopia with corneal refractivesurgical procedures presents substantial challenges for refractive surgeons.Phakic intraocular lenses, both anterior and posterior chamber, have potentialintraocular complications such as secondary glaucoma, endothelial cell loss,complicated cataract, and uveitis and are still being investigated in hyperopiceyes.^20-23

This study describes a new and unique complication termed epithelial perilenticular opacity, which occurred in 5 eyes followingintracorneal inlay implantation. Symptoms such as photophobia, blurry vision,and starbursts as well as clinical course were similar in all patients. Thesesymptoms began in the first week, and biomicroscopy examination results weresimilar in all cases. Patients did not improve with topical steroids, withprogression of the opacity and worsening of the BSCVA, leading to inlay explantation.The differential diagnosis of this complication was DLK. This consists ofnoninfectious diffuse interface inflammation after corneal lamellar surgery,characterized by infiltration of inflammatory cells at the interface.²⁴ As in DLK, the diffuse aspect of the infiltrate (theabsence of a single focus and the confinement of the infiltrate to the interface)suggests a noninfectious etiology. During the inlay explantation procedure,samples were obtained in all 5 eyes for microbiological analysis and producednegative results in all cases. In DLK cases, an allergic or toxic inflammatoryreaction is the most likely cause; in general, such cases respond well totopical steroids.²⁴

The complication of epithelial perilenticular opacity is distinctlydifferent from DLK: the infiltration was confined to the limits of the inlay,did not respond to steroids, and had a different clinical evolution from DLK.A possibility that could be considered in our cases is a hypersensitivityreaction type 4.²⁵ Immunological rejectiondepends on whether the host recognizes the implanted material as foreign andproduces specific persistent antigens, as in the case of the intracornealinlay. This might be the stimulus for macrophage cells to migrate, surroundingthe foreign object, and then to transform into epithelial-like cells (knownas epithelioid cells), causing an inflammatory response on the stromal bed.This was ruled out because our histopathologic studies did not show the presenceof this kind of cell.

According to pathologic and confocal microscopy analysis, the implantationof epithelial cells and their further ingrowth on the inlay surface was thecause of the perilenticular opacity. The implantation of epithelial cellsin the interface occurs by mechanical dragging of the keratome blade duringkeratectomy; backflow during irrigation, carrying floating epithelial cells;ingrowth at the junction of the epithelium and keratotomy; and migration underthe flap.²⁶ When already present at the interface,the adequate condition of the inlay surface and its isolation from other tissueswill probably promote the ingrowth of the layer formed on the hydrophilicsurface of the inlay. The migrating epithelial cells may further contributeto stromal melting of the flap, but no evidence of corneal melting was observedin any of our cases. Epithelial ingrowth at the interface is more common afterenhancement procedures because the lifting of the flap can induce adjacentepithelial abrasions with increased cell proliferation.²⁷ Inour study, 3 of 5 eyes had the flap lifted to reposition the inlay as a resultof decentration. Epithelial cells were on the posterior inlay surface andaround the edge. This induced a hyperopic shift with loss of BSCVA causedby the whitish deposits. No evidence was observed in any of the cases thatthe epithelial cells were growing around the inlay or invading the interfacefrom the edge of the flap. This supports the idea that such cells were implantedat the interface by the microkeratome cut. The dry technique used in thesecases likely influenced the development of this complication. Use of theseinlays should be adequately reevaluated in the future as a potential sourceof epithelial perilenticular opacity.

This article forms part of a multicenter study of the correction ofhyperopia with intracorneal inlays. The new generation of soft intracornealinlays offers an alternative for the correction of hyperopia with potentialreversibility and implant removal.²⁸ Therefore,we report a new specific complication of intracorneal epithelial inlay implantationcalled intracorneal perilenticular opacity, a finding that could have a significanteffect on the future consideration of inlays in refractive surgery. This complicationshould be distinctly differentiated from DLK. The lack of a positive responseto steroids and the confinement of the interface opacity to the limits ofthe implanted inlay are the most remarkable differentiating clinical signs.A careful review of the surgical process involved in inlay implantation shouldbe conducted. The causes and prevention of epithelial perilenticular opacityshould be carefully investigated in future research into additive refractivekeratoplasty with intracorneal inlay implantation.

Correspondence: Jorge L. Alió, MD, PhD, Instituto Oftalmológicode Alicante, Avenida de Denia S/N, Edificio Vissum, 03016 Alicante, Spain(rdioa@oftalio.com).

Submitted for publication July 10, 2003; final revision received November17, 2003; accepted May 14, 2004.

The study has been supported in part by grant C03/13 from the SpanishMinistry of Health, Red Temática de Investigación en Oftalmología,Subproyecto de Cirugía Refractiva y Calidad Visual (Madrid).