A, Slitlamp photograph of donorcornea with epithelial side down in a vial with Optisol-GS. Retroilluminationreveals a laser-assisted in situ keratomileusis (LASIK) flap edge rated as"obvious." Cornea is from a 54-year-old female donor with LASIK performed4½ years before death. B, An example of a "subtle" LASIK flap edgeby slitlamp examination from a 60-year-old male donor who had the procedure2½ years before death.
A, Specular microscopy image focused255 µm from the endothelium reveals a severe amount of highly reflectivestromal particles. This 47-year-old donor cornea underwent laser-assistedin situ keratomileusis (LASIK) 9 months before death. The field of view is0.2 × 0.4 mm. B, A more representative specular micrograph of highlyreflective stromal particles. The maximum number of highly reflective particlesobserved was 3 particles, found a distance of 489 µm from the endothelium.This 47-year-old donor had LASIK 2 years before death. The particles rangein size from approximately 5 to 20 µm.
A, Specular micrograph of endotheliumfrom a 46-year-old laser-assisted in situ keratomileusis (LASIK) donor cornea1 day after death. Cell density and morphologic structure appear normal. B,Specular micrograph of a 45-year-old LASIK donor cornea endothelium 5 daysafter death. Moderate and diffuse changes are found in cell appearance andmorphologic structure, with notable distortion of hexagonal shape and scatteredareas of blackening consistent with endothelial cell edema.
Example of pyknosis and vacuolizationof keratocytes in corneal interface (hematoxylin-eosin, original magnification×400).
V. Vinod Mootha, Dan Dawson, Amit Kumar, Joel Gleiser, Clifford Qualls, Daniel M. Albert. Slitlamp, Specular, and Light Microscopic Findings of Human Donor CorneasAfter Laser-assisted In Situ Keratomileusis. Arch Ophthalmol. 2004;122(5):686–692. doi:10.1001/archopht.122.5.686
To examine slitlamp, specular, and light microscopic features of humandonor corneas known to have undergone laser-assisted in situ keratomileusis(LASIK).
Twenty-six donor corneas known to have undergone LASIK prospectivelyunderwent slitlamp examination with particular attention to the presence ofa flap edge, as well as specular microscopy with particular attention to thepresence of highly reflective particles in the stroma corresponding to theLASIK interface. Central endothelial cell density and pachymetery were obtained.They were compared with 26 control donor corneas without LASIK. Eleven LASIKdonor corneas were processed for histology. Twenty-six donor corneas withno known prior keratorefractive surgery also underwent similar slitlamp examinationand specular microscopy to serve as controls.
Twelve (46%) of 26 LASIK donor corneas had an obvious flap edge, and10 (39%) had a subtle flap edge by slitlamp examination. Four (15%) had infiltratesby slitlamp examination, of which 3 were confirmed by histopathologic examination.Highly reflective particles were seen by specular microscopy in the stromaof 23 (88%) of 26 LASIK donor corneas, but only 1 (4%) of 26 control donorcorneas had a single highly reflective particle in the stroma (P<.001). The mean central endothelial cell counts were similar:2138 cells/mm2 in the LASIK group compared with 2250 cells/mm2 in the controls (P = .39). Vacuolizationand pyknosis of keratocytes was a consistent histopathologic finding afterLASIK. Metallic particles at the interface were not detected by histology.
Detection of a flap edge by slitlamp examination may detect at leasthalf of the donor corneas that may have undergone LASIK. The detection ofhighly reflective stromal particles may form an effective basis for screeningfor LASIK donor corneas using specular microscopy and requires further study.
In 2001, the 80 eye banks of the Eye Bank Association of America providedtissue for 46 532 corneal transplantation procedures.1 Withmillions of procedures already performed in the United States, the increasingpopularity of laser-assisted in situ keratomileusis (LASIK) threatens thedonor cornea pool. Corneas that have had any previous refractive surgery arenot suitable for transplantation because of structural and optical compromiseto the tissue. Michaeli-Cohen et al2 described2 patients who underwent penetrating keratoplasty by surgeons who were unawarethat the donor eyes had previous LASIK. In one of these cases, separationof the corneal lamellae was noted at the time of surgery.
Eye banks rely primarily on the medical and social history interviewto detect donors who have had previous refractive surgery. New screening techniquesare required to protect the donor cornea pool. Corneal topography and pachymetry,3,4 videokeratography,5,6 dark-fieldbiomicroscopy,7 wavefront analysis,8 and optical coherence tomography9 havebeen proposed as screening tools for the detection of previous refractivesurgery. However, none of these instruments are readily available at mosteye banks.
The in vivo confocal microscope consistently demonstrates the presenceof variable reflectivity particles in the corneal interfaces of all patientsafter LASIK.10,11 Pisella et al10 speculate that the highly reflective interface particlesconsist of metallic particles originating from the microkeratome blade andthe less reflective particles may represent cellular debris or plastic particles.Gokmen et al11 used the presence of the highlyreflective particles found in all examined patients with the in vivo confocalmicroscope as a basis to measure the thickness of the LASIK flap. In anotherrecent confocal microscope study,12 highlyreflective particles were found in the anterior stroma in 78 of 85 confocalscans performed after LASIK. Like the confocal microscope, the specular microscopeis a reflected light microscope. The specular microscope is readily availableat eye banks and is used to study the endothelial cell count density and morphologicstructure in donor corneas. This study attempted to determine the usefulnessof the specular microscope to detect the presence of the highly reflectiveparticles in the region of the central stroma corresponding to an interfacein donor corneas known to have had LASIK. To our knowledge, this study isthe first to examine the endothelium of donor corneas after LASIK surgery.The specular microscope was also used to assess central endothelial cell densityand central corneal pachymetry measurements. This study also attempted todetermine the usefulness of the slitlamp to identify the LASIK flap edge inpostmortem eyes.
In this study, we looked at the histologic findings of 11 of these LASIKdonor corneas. One other histologic study13 ofhuman corneas after successful LASIK included only 4 corneas. Other reportsof histologic changes in human corneas involved corneal button specimens afterpenetrating keratoplasty for LASIK-related complications14- 16 orphthisical blind eyes in which the procedure was performed just before enucleation.17
This project was approved by the Human Research Review Committee ofThe University of New Mexico Health Sciences Center. With use of the medicaland social history interview, participating eye banks of Tissue Banks International(Baltimore, Md) identified potential donor corneas that had undergone LASIK.After consent for research was obtained, the corneas were removed by in situexcision and were placed in vials with Optisol-GS corneal storage medium (ChironOphthalmics, Irvine, Calif) by eye bank technicians using established protocolsof the Eye Bank Association of America.18 Centralendothelial cell counts were obtained by the participating eye bank responsiblefor harvesting the cornea. The corneas were sent to the New Mexico Lions EyeBank, Albuquerque, using protocols established by the Eye Bank Associationof America.18 The donor corneas were storedat 4°C until further examination.
Twenty-six donor corneas known to have undergone LASIK prospectivelyunderwent slitlamp examination with particular attention to the presence ofa flap edge. A cornea fellowship–trained investigator (V.V.M.) usedthe slitlamp examination to rate the LASIK flap edge as "obvious," "subtle,"or "absent." The LASIK donor corneas with an absent flap edge were confirmedto have had the refractive surgery by histopathologic examination. These corneasthen underwent thorough specular microscopy with the Konan Eyebank Keratoanalyzer(Konan Medical, Tokyo, Japan) with attention to the presence of highly reflectiveparticles in the region of the stroma that would correspond to an interface.The field of view of the specular microscope used in this study was 0.2 ×0.4 mm. The vial with the cornea positioned with its endothelial surface downwas placed in the vial adapter and chamber holder assembly. The z-axis knobadjustment was used to focus on the central endothelium. Endothelial cellcount measurements (using the center method technique with the built-in softwareof the Keratoanalyzer) were obtained in each donor cornea if not obtainedat the harvesting eye bank. The coefficient of variation (standard deviationof the endothelial cell area divided by the mean cell area ×100) andpercentage of hexagonal cells were calculated by the software of the Keratoanalyzerspecular microscope. Central pachymetry measurements were obtained on eachcornea optically by focusing from the endothelium to the epithelium usingthe z-axis knob. The z-axis knob was used to search for highly reflectivecentral stromal particles starting from a depth of 150 µm from the endothelium.The paracentral stroma was examined by use of the x-axis and y-axis knobson the microscope platform. The central stroma was examined at 25-µmincrements toward the epithelium. The specular microscope was sharply focusedon any highly reflective particle in the stroma, and its depth from the endotheliumwas noted. The maximum number of particles in any field observed was countedin each cornea. Twenty-six donor corneas already present at the New MexicoLions Eye Bank with no known history of prior keratorefractive surgery underwentsimilar slitlamp examination and specular microscopy to serve as controls.
Eleven LASIK donor corneas were placed in 10% neutral buffered formalinand processed for permanent sections. Five-micron serial sections from thecenter of the donor corneas were obtained. They were examined under lightmicroscopy (original magnification, ×20 to ×400) after being stainedwith hematoxylin-eosin and periodic acid–Schiff.
The refractive data of the LASIK donors, including the preoperativemanifest refraction, preoperative pachymetry, date of LASIK, microkeratomeused, footplate depth, and amount of ablation, were obtained if possible.The collected data were entered into an Excel spreadsheet (Microsoft Corp,Redmond, Wash), and analyses were performed with SAS statistical software,version 8.02 (SAS Institute, Cary, NC).
Demographic information on the LASIK donor corneas and controls is givenin Table 1. The mean donor ageamong the LASIK corneas (50 years) was younger than that among the controls(66 years) (P<.001). On average, specular microscopicexamination of the endothelium occurred 4.8 days after death in the LASIKgroup compared with 2.5 days in the controls (P =.01). The available preoperative refractive and operative data on the LASIKdonor corneas are given in Table 2.
Slitlamp examination of the LASIK donor corneas revealed that 12 (46%)of 26 had an obvious flap edge (Figure 1A),10 (39%) had a subtle LASIK flap edge (Figure1B), and 4 (15%) had an absent flap edge. Among the controls, anabsent flap edge was observed in 26 (100%) (P<.001compared with controls, Fisher exact test). Four donor corneas (15%) had evidenceof infiltrates by slitlamp examination.
Specular microscopy findings of the LASIK donor corneas and controlsare presented in Table 3 and Table 4. The maximum number of highly reflectiveparticles observed in any observed field in 26 LASIK donor corneas was asfollows: none (0 particles), 3 (12%); mild (1-2 particles), 10 (39%); moderate(3-6 particles), 12 (46%); and severe (≥7 particles), 1 (4%). Twenty-three(88%) of 26 LASIK donor corneas revealed highly reflective stromal particlesin the region of the interface (Figure 2).Only 1 control cornea (4%) had an isolated highly reflective intrastromalparticle. The particles were found at a mean ± SD minimal depth of303 ± 95 µm to a maximum depth of 362 ± 101 µm fromthe endothelium. The endothelial cell density was 2138 cells/mm2 inthe LASIK corneas compared with 2250 cells/mm2 in the control group(P = .39 [P = .41 adjustedfor age]) (Figure 3). The coefficientof variation was 42 in the LASIK corneas and 33 in the controls (P<.001). The percentage of hexagonal cells was 45% in the LASIKcorneas compared with 55% in the controls (P = .001)(Table 3). The mean central pachymetrymeasurement was 506 µm in the LASIK group compared with 528 µmin the control group (P = .06).
Histopathologic results are summarized in Table 5. Pathological findings confirmed the presence of a lamellarflap in all examined LASIK donor corneas. All interface scars showed vacuolizationand pyknosis of keratocytes along the lamellar cut (Figure 4). Metallic particles were not found on the serial sectionstaken from the center of the donor corneas in any of the examined interfaces.Epithelial facets (minor epithelial down-growth) were occasionally seen atthe entrance to the lamellar cut. Three of the 4 donor corneas that had peripheralinfiltrates had evidence of nongranulomatous inflammatory anterior stromalinfiltrates. One cornea stained for gram-positive cocci on the surface epitheliumby Brown-Brenn stain and another cornea with Gomori methenamine silver stainedpositive for a small number of yeast organisms on the surface epithelium.The endothelium was unremarkable in all examined cases except for 1 pair ofdonor corneas with decreased numbers of endothelial cells.
Until better screening technology is developed and implemented by eyebanks, the slitlamp and specular microscope may be useful in the detectionof donor corneas with previous LASIK surgery. The results of this study suggestthat 46% (12/26) of LASIK donor corneas have an obvious flap edge. In thisunmasked study, 39% (10/26) of the LASIK donor corneas had a subtle flap edgeand 15% (4/26) had no detectable flap edge; these would probably be missedby routine slitlamp examination. Eye bank technicians, however, must be trainedto recognize a LASIK flap edge, as nearly half of LASIK donor corneas couldbe potentially screened by the simple use of the slitlamp. Retroilluminationslitlamp techniques are useful in the detection of a LASIK flap. In this study,clear cornea cataract wounds, astigmatic keratotomy scars, arcus senilis changes,and posterior embryotoxon changes were seen that could mimic a LASIK flapto an inexperienced slitlamp observer. Careful slitlamp examination of theendothelial surface can help distinguish a clear cornea cataract wound andposterior embryotoxon from a true microkeratome-induced flap edge.
In this study, 15% (4/26) of the LASIK donor corneas were noted to haveinfiltrates by slitlamp examination, suggesting that the well-documented neurotrophicstatus of the cornea following LASIK19 mayfurther predispose to exposure-related keratitis in intubated or sedated patients.The details of these 4 cases will be published in a separate article.
The results of this study suggest that the standard specular microscopeavailable in eye banks can detect highly reflective intrastromal particlesin donor corneas after LASIK. In published reports using the in vivo confocalmicroscope, the presence of highly reflective interface particles varies between92%12 and 100%10,11 inpatients after LASIK. In our study, specular microscopy revealed highly reflectivestromal particles in 88% (23/26) of the LASIK donor corneas. The stromal depthof the particles correlates to where microkeratome-induced changes would beseen.
The mean ± SD density of variable reflectivity particles (allparticles, including highly reflective and less reflective) at the interfaceby in vivo confocal microscopy has been reported to be 504 ± 101 particles/mm2 at 8 days after the LASIK procedure and decreases during the first3 months after surgery to stabilize at 332 ± 100 particles/mm2.10 Another in vivo confocal microscopestudy12 reported the mean ± SD densityof the highly reflective particles at the interface to be 19.8 ± 24.6particles/mm2 at 1 week after the LASIK procedure and did not changesignificantly with time. In our investigations with the specular microscope,density measurements of the highly reflective particles were not performed.However, the maximum number of highly reflective particles in any observedfield (0.08 mm2) averaged only 3 particles. In comparison to theconfocal microscope, the resolution of standard eye bank specular microscopesis far less. In our study, we were probably detecting only a small numberof the most highly reflective of the variable reflectivity particles seenin vivo by the confocal microscope. The poorer resolution of the specularmicroscope may also account for why 12% (3/26) of the LASIK donor corneashad no detectable highly reflective stromal particles, despite a thoroughsearch. Also, the eye bank specular microscope used in this study was designedfor examination of a small area of the central corneal endothelium (0.2 ×0.4 mm) and may miss peripheral interface particles in a typical 8- or 9-mmLASIK flap.
Specular microscope reflectivity is a function of a change in indexof refraction at a particular interface in the tissue examined. Air bubbleson the corneal surface, the interface between the epithelium and corneal storagemedium, and epithelial irregularities or debris may induce specular reflectivity.Care must be taken not to mistake these artifacts for highly reflective particleswithin the substance of the stroma seen at the LASIK interface. Less reflectivestromal debris is not a reliable indicator of a microkeratome cut, as thiswas occasionally seen in the control corneas. Interestingly, we cannot accountfor the presence of a single, highly reflective particle in the stroma ofa control cornea from a 70-year-old female donor who died of cancer. The fellowcornea from this patient had a few highly reflective particles at the levelof the epithelium.
In addition to evaluating the donor cornea endothelium, the specularmicroscope may be a useful instrument to screen for previous LASIK surgeryby the detection of the highly reflective stromal particles correspondingto the interface. However, the manual search for highly reflective stromalparticles can be tedious, as 50% (13/26) of the LASIK donor corneas had amild amount of particles or none. Masked studies are required to determinethe true sensitivity and specificity of existing specular microscopes to screenfor donor corneas that have had LASIK. The results of this study suggest thatthe detection of highly reflective interface debris may form an effectivebasis for screening using existing specular microscopes until better screeningtechnologies are developed.
In vivo specular microscopic investigations in patients who have hadLASIK have demonstrated no significant change in central endothelial cellcounts.20 However, a study21 reporteda transient endothelial dysfunction marked by increased pleomorphism withdefinite loss of hexagonality immediately after LASIK surgery. In our study,the mean central endothelial cell count density was slightly lower in theLASIK group (2138 cells/mm2) compared with the control donor corneas(2250 cells/mm2). This small difference was, however, not statisticallysignificant even when the age discrepancy in the 2 groups was considered (P = .39 [P = .41 adjusted forage]). However, the coefficient of variation (indicator of polymegathism)was slightly higher in the LASIK group at 42 compared with 33 in the controls(P<.001). The percentage of hexagonal cells (indicatorof pleomorphism) was slightly lower in the LASIK group (45%) compared withthe controls (55%) (P = .001). On average, the endotheliumwas analyzed with the specular microscope 4.8 days after death in the LASIKdonor corneas compared with 2.5 days in the controls (P = .01). This 2-day delay was attributable to the processing and shippingtime required to send the tissue from participating eye banks to the New MexicoLions Eye Bank. This project was designed primarily for the specular microscopicexamination of the stroma in donor corneas. Further studies with proper matchingof age, death to preservation time, and delay before specular microscopy ofthe endothelium are required to assess the effects of LASIK surgery on centralendothelial cell count density, morphologic structure, and viability.
Although the LASIK donor corneas were slightly thinner (506 µm)compared with the control corneas (528 µm), this difference was notstatistically significant (P = .06). Pachymetry measurementswith use of specular microscopy may not be as reliable as ultrasound technology.However, the pachymetry results in this study and the potential artifactsinduced by epithelial defects and variable stromal edema of corneas in thestorage media suggest that pachymetry using specular microscopy may not bean effective basis for screening for previous LASIK surgery.
The histologic findings on the LASIK donor corneas failed to identifymetallic particles in any of the interfaces examined to be the source of thehighly reflective particles seen by specular and confocal microscopy. However,the consistent histopathologic finding of vacuolization and pyknosis of keratocytesalong the lamellar cut may account for some of the variable reflectivity particlesseen by confocal and specular microscopy. The pyknotic nuclei may be relatedto the previous LASIK surgery, but postmortem changes cannot be ruled out.Vacuolization and pyknosis of keratocytes have been confirmed by other histologicstudies13 and may account for the decreasedkeratocyte density in the stroma adjacent to the LASIK lamellar cut demonstratedby the in vivo confocal microscope.12 Furtherresearch is required to elucidate the nature and clinical significance ofthe keratocyte changes.
Corresponding author and reprints: V. Vinod Mootha, MD, Divisionof Ophthalmology, Department of Surgery, The University of New Mexico HealthSciences Center and School of Medicine, 2211 Lomas Blvd NE, 2-Ambulatory CareCenter, Albuquerque, NM 87131 (e-mail: email@example.com).
Submitted for publication February 3, 2003; final revision receivedSeptember 24, 2003; accepted November 21, 2003.
This project was supported in part by the Dedicated Health ResearchFunds of The University of New Mexico School of Medicine, Albuquerque.
The study was presented in part at the 40th Annual Scientific Sessionof Eye Bank Association of America; November 10, 2001; New Orleans, La; andthe 41st Annual Meeting of the Eye Bank of Association of America; June 22,2002; Ponte Vedra Beach, Fla.
We thank Tissue Banks International, Kestrel Corporation, Albuquerque,Gregory S. H. Ogawa, MD, and Curt Vavra, RN, for the collection of donor corneas.We also thank Edward Otero and Chris Miguel-Nathan for their assistance withpreparation of the manuscript.