Text of Abstracts

Cristol SM, Lee JH, Kim WC, Chung ES, Tchah H, Kim MS, Nam CM, Cho H-S, Kim EK. Estimating the prevalence of granular corneal dystrophy type 2 (Avellino corneal dystrophy) in the Korean population. ARVO Annual Meeting; 2010 May 2-6; Fort Lauderdale, FL.

Purpose: In this study, we used modeling to estimate the prevalence of granular corneal dystrophy type 2 (GCD2; Avellino corneal dystrophy), a rare disease, in the Korean population.

Methods: Through a collaboration of Korean referral centers for corneal disease, we identified all GCD2 homozygotes that were over the age of three in 2005. The genetic status of the patients and their immediate families were verified by DNA analysis. We developed a model based on the Hardy-Weinberg principle to calculate a lower bound for the gene prevalence. We developed a second, population-based model to correct for known underestimation in the primary model. Both models used population data from the 2005 Korean census and the corrected model also used fertility rates from historical Korean census data.

Results: We identified 21 individuals homozygous for GCD2 (R124H mutation) from 16 Korean families. From this, we estimate that the overall prevalence (combining heterozygotes and homozygotes) is at least 8.25 affected persons/10,000 persons. Based on our corrected estimate, the overall prevalence is likely to be greater than 11.2 affected persons/10,000 persons.

Conclusion: We present the first estimate of the prevalence of GCD2. Although uncommon, the prevalence of GCD2 in Korea is greater than anticipated. We believe that our approach could potentially be applied to estimating the prevalence of other rare diseases.

Kim EK, Cho HJ, Ahn SY, Choi YJ, Kim TI, Cristol S, Choi SI. Lysosomal Traffic and Degradation of βig-h3 Protein. ARVO Annual Meeting; 2007 May 6-10; Fort Lauderdale, FL.

Purpose: To investigate intracellular traffic and degradation of βig-h3 protein, a protein associated with amyloid deposition in granular corneal dystrophy Type II (GCDII).

Methods: Corneal fibroblasts were treated with proteasomal inhibitors, lactacystin and MG132, or lysosomal inhibitors, bafilomycin A1 for 12 h. We used immunoblot analysis to measure βig-h3 protein in cell lysate.

Results: The lysosomal inhibitors, bafilomycin A1, significantly blocked βig-h3 protein degradation in a dose-dependent manner in cultured corneal fibroblasts. In contrast, proteasomal inhibitor, MG132, did not lead to significant accumulation of βig-h3 protein. The internalization of βig-h3 is inhibited by nystatin, which blocks lipid raft endocytosis, but not by chlorpromazine which is known to block clathrin-mediated endocytosis. Moreover, βig-h3 endocytosis was decreased significantly in TGF-β1-exposed cultured corneal fibroblasts.

Conclusions: These results demonstrate that βig-h3 protein degradation is mediated by an autophagy/lysosomes pathway. Alterations both in lysosomal degradation and in expression of βig-h3 induced by TGF-β1 may contribute to amyloid deposition of βig-h3 in GCDII.

Cristol SM, Chung SH, Kim WC, Choi SY, Stulting RD, Lee HK, Kim EK. Treating Avellino corneal dystrophy with a conjunctival flap over the mid-peripheral cornea. ARVO Annual Meeting; 2005 May 1-5; Fort Lauderdale, FL.

Purpose: To assess the short term safety and efficacy of treating Avellino corneal dystrophy (ACD) with a mid–peripheral conjunctival flap. This approach was suggested by observations in a 20 year old Korean female patient with heterozygous ACD. She had numerous fine, white, granular deposits in her right anterior stroma but no deposits in her left cornea which had 120° neovascularization at the limbus with 360° conjunctival injection.

Methods: Three heterozygous ACD patients, each with an eye with accelerated deposition after LASIK, had a conjunctival flap attached to the superior cornea 3 mm from the pupil center with (n=1) or without (n=2) removal of the LASIK flap. One homozygous patient (Patient "A") had phototherapeutic keratectomy (PTK) in each eye and superior and inferior mid–peripheral conjunctival flaps in one eye. Three eyes of two other homozygous patients had conjunctival flaps attached to the superior and inferior cornea 3 mm from the pupil center; all three eyes had previously been treated with PTK.

Results: Within two weeks of the flap placement, all eyes of heterozygous ACD patients showed some reduction in ACD deposits near the conjunctival flap. Patient "A" had minimal recurrence in the eye with the flaps after 5 months while moderate recurrence had been noted in the fellow eye after only 2 months. The three eyes of the other two homozygous patients showed clearance of deposits near the edge of the conjunctival flap.

Conclusions: Mid–peripheral conjunctival flaps can reverse ACD deposits and delay recurrence over short time intervals. It seems likely that this is related to the vascularity of the conjunctiva.

Kim EK, Jung S, Park J, Cristol SM, Kim CY, Seo KY, Lee H. Using an amniotic membrane pressure patch to treat the LASIK flap with a central defect and epithelial ingrowth. ARVO Annual Meeting; 2004 April 25-29; Fort Lauderdale, FL.

Purpose: To report a new successful treatment of epithelial ingrowth occuring under a LASIK flap with a central defect by pressure patching with amniotic membrane.

Methods: Two cases of epithelial ingrowth through a central defect in the LASIK flap were referred after failure of therapeutic contact lens treatment. In patient 1, the flap was bisected by a poor microkeratome incision 5 days earlier. The cut margin was elevated by the epithelium under the flap (Fig., upper left). In patient 2, the microkeratome had produced a very thin flap with only the epithelial layer in the central flap. Laser ablation had been done at the remaining posterior stroma. The central flap had melted and perforated; the epithelium had been growing into the interface for 5 months (Fig., upper right). Each flap was lifted and the epithelium was removed. In patient 2, phototherapeutic keratectomies of 6 mm in diameter were performed on the posterior flap and on the stromal bed for 30 pulses. An amniotic membrane pressure patch (epithelial side against the cornea) was sutured to the episclera under tension, using interrupted 10–0 nylon sutures. The amniotic membrane patch was removed 5–6 days after application.

Results: Both flaps were attached to the remaining posterior stroma without recurrent epithelial ingrowth after removal of amniotic membrane. Visual acuity improved in patient 1 from 20/70 to 20/25 (Fig., lower left) and in patient 2 from 20/100 to 20/25 (Fig., lower right).

Conclusions: Epithelial ingrowth in the central cornea following LASIK can be treated with amniotic membrane pressure patch. The elasticity of the membrane maintains the anatomic relationship between the flap and the stromal bed and prevents recurrence of the epithelial ingrowth.

Cristol SM, Rengarajan K, Mehta M, Nickerson JM. Quantifying Small DNA Concentrations using Fluorometry. ARVO Annual Meeting; 2002 May 5-10; Fort Lauderdale, FL.

Purpose: This study compared the sensitivity of four commercially available fluorophores for rapidly measuring very small concentrations (pg/ml) of DNA by fluorometry. The principle of the assay is to measure enhancement of fluorescence emitted by DNA bound fluorophores. This technique uses inexpensive disposable cuvettes in contrast to A₂₆₀ measurements, which only have a sensitivity on the order of 150 ng/ml.

Methods: Samples ranging from 100 pg/ml to 20 µg/ml of DNA were prepared by a serial dilution of calf thymus DNA and mixed with each dye. Fluorescence measurements were made at room temperature (23-24 °C) using a photon counting spectrofluorometer. The dyes used to measure fluorescence enhancement were ethidium bromide, Hoechst 33258, PicoGreen, and SYBR Green I. For each dye, the concentration of DNA (over the 0-2 ng/ml range) was modelled as a function of fluorescence intensity using linear regression. Intra-assay repeatability was evaluated using SYBR Green I by making six replicates for each dilution and making a single measurement of each. Inter-assay repeatability was evaluated using SYBR Green I from four lot numbers (of three vendors) measured across two days. The stability of fluorescence was measured over a 24 h interval using SYBR Green I obtained from three vendors.

Results: Within an assay, the coefficient of variation was less than 10% for each fluorophore. No significant difference in fluorescence was observed among the three sources of SYBR Green I. There was no variation between the two lots from a single source. Regression models for ethidium bromide and Hoechst 33258 over the 0-2 ng/ml range were not statistically significant. The models for PicoGreen and SYBR Green I were statistically significant and had values of R²>0.96. From this data, we estimated that the resolution of PicoGreen was about 40% smaller than SYBR Green I.

Conclusion: SYBR Green I and PicoGreen are useful fluorophores for measuring small concentrations of DNA; ethidium bromide and Hoechst 33258 are not. Both SYBR Green I and PicoGreen offer a means of measuring small concentrations of DNA rapidly, inexpensively, and with greater sensitivity than comparable techniques. Because PicoGreen is substantially more expensive than SYBR Green I, we recommend SYBR Green I for assaying small concentrations of DNA.

Jones SS, Azar RG, Cristol SM, Geroski DH, Waring GO, Stulting RD, Thompson KP, Edelhauser HF. Short-Term Effects of Laser In-Situ Keratomileusis (LASIK) on the Corneal Endothelium. Invest Ophthalmol Vis Sci 1996; 37(suppl):63.

Purpose: To determine the effects (2 weeks-3 months) of LASER In-Situ Keratomileusis (LASIK) on the human corneal endothelial cell density, coefficient of variation, and percent hexagonal cells.

Methods: The corneal endothelium of 97 eyes (64 patients; 22-66 years) was photographed with the Konan ROBO non-contact specular microscope pre-op and two weeks post-op. Morphometric analysis; cell density (CD), coefficient of variation (CV), and percent hexagonal cells (%H) was performed using 150-200 cells from each image. Current three month post-op analysis is in progress.

Results: We found that there was no significant difference in each of the parameters studied pre-op and 2 week post-op. Pre-op and post-op values for the entire population were as follows: mean CD±SD, 2454±364 cells/mm2 (range 1364-3278) and 2558±360 cells/mm2 (range 1626-3257), CV±SD, 0.31±0.13 (range 0.24-0.64) and 0.31±0.14 (range 0.25-0.63), and %H±SD, 60%±6% (range 45%-74%) and 59%±6% (range 44%-74%). Data were also analyzed in terms of patient age and refractive error, at pre-op and 2 weeks post-op, are as follows: 20-29 y.o. (N=13) 2770±339 cells/mm2 and 2797±301 cells/mm2, CV, 0.32±0.04 and 0.32±0.04, and %H, 65%±5% and 64%±6%, 30-39 y.o. (N=28) 2648±316 cells/mm2 and 2673±312 cells/mm2, CV, 0.34±0.04 and 0.33±0.04, and %H±SD, 60%±6%, 40-49 y.o. (N=39) 2510±354 cells/mm2 and 2511±356 cells/mm2, CV, 0.37±0.08 and 0.38±0.08, and %H±SD, 58%±6% and 57±6%, and 59+ y.o. (N=16) 2282±328 cells/mm2 and 2292±310 cells/mm2, CV, 0.35±0.05 and 0.36±0.05, and %H, 59%±6% and 58%±5%. When the data was analyzed for refractive error (0-4D, 4-8D, 8-12D, 12+D) there was no apparent change in the post-op values for CD, CV, or %H.

Conclusion: Corneal endothelial cell density and morphology remained normal following 2 weeks LASIK. Three month data is in progress.

Cristol SM, Lynn MJ, Waring GO, Edelhauser HF, and the PERK Study Group. Astigmatism One Year After Radial Keratotomy in the PERK Study. Invest Ophthalmol Vis Sci 1994; 35(suppl):1294.

Purpose: To characterize the change in astigmatism after eight incision radial keratotomy in the Prospective Evaluation of Radial Keratotomy (PERK) study. Attempt to statistically model various measures of astigmatism change within the PERK data set.

Methods: Two scalar and five vector methods of quantifying astigmatism change were used to describe the astigmatism at one year following surgery in 399 eyes that underwent a single radial keratotomy operation. Multiple regression and logistic regression modeling were used to identify factors significant in predicting outcome.

Results: Using the six measures of astigmatism change (that can be interpreted as diopters), a change of a half diopter or more was present in 45.4%, 33.3%, 48.6%, 54.4%, 56.6%, and 56.6% of the eyes. One diopter of change was present in 11.8%, 10.3%, 12.5%, 17.8%, 16.8%, and 16.5% of the eyes. Four of the vector techniques were highly correlated and comparable in predicting a change in astigmatism of more than 0.5 D; the remaining three techniques could not be modeled usefully. The diameter of the clear zone was most predictive of outcome. For a 3 mm clear zone, an astigmatism change greater than 0.5 D is 3.9 (2.2, 6.8), 3.2 (1.8, 5.4), 4.7 (2.7, 8.2), or 4.0 (2.3, 7.1), times more likely than for a 4 mm clear zone (odds ratio with 95% confidence interval). A composite variable representing the variability in the incision depth was also predictive. Greater differences among the incision depths were associated with a greater likelihood of a larger astigmatism change.

Conclusions: This work confirms previous findings (Lynn: ARVO, 1991) that the diameter of the clear zone is a predictor of astigmatism change. It also identifies variability of incision depth as a predictor of astigmatism change.

Freund DE, McCally RL, Farrell RA, Cristol SM, Edelhauser HF, L’Hernault NL. Comparison of the Ultrastructure in Anterior and Posterior Stroma of Normal Human and Rabbit Corneas and its Relationship to Transparency. Invest Ophthalmol Vis Sci 1992; 33(suppl):1411.

The ultrastructure of the stroma determines the degree to which the cornea is transparent. Low magnification EMs of normal cornea reveal differences in the lamellar structure between the anterior and posterior regions of the stroma. Additionally, the anterior and posterior regions are known to differ in their state of hydration, the relative amounts of keratan sulfate and dermatan sulfate proteoglycans, their capacity to swell, and the degree of transparency loss with hydration. These anatomic, physiological, and biochemical differences suggest that the ultrastructure, and its effect on transparency, may also be different for the two regions. To address this idea, standard image processing techniques are used to obtain fibril locations, average fibril diameters, and the radial distribution function of fibril positions from high magnification EMs of the anterior, middle and posterior regions of the stroma from several normal human and rabbit corneas. Theoretical light scattering calculations are performed in each case and used to predict the transparency. Fibril positions are more ordered in the posterior human cornea than they are in the anterior. Although both anterior and posterior ultrastructures lead to a high degree of transparency (>96% transmittance of red light and >84% transmittance of violet light) the calculations reveal that the anterior would scatter twice as much red light and 1.7 times as much violet light as the posterior. Differences between anterior and posterior rabbit cornea are less distinct.

Kim EK, Cristol SM, McCarey BE, Geroski DH, Edelhauser HF. Corneal Endothelial Damage from Air Bubbles during Phacoemulsification. Invest Ophthalmol Vis Sci 1992; 33(suppl):1306.

The fast swirling movement of air bubbles during phacoemulsification (phaco) has been reported to cause endothelial cell loss by physical trauma. We hypothesized that endothelial cell loss can also occur by gentle contact with an air bubble following the removal of protective molecules (protein or viscoelastics) from the anterior chamber fluid. Anterior chamber irrigation or bilateral phaco was performed in 2-3 Kg rabbits. Phaco was performed with Viscoat (Alcon) or Healon (Pharmacia) in one eye and no viscoelastics in the other eye. The Cooper Vision model 9001 phaco unit was set with 0.9 power for 60 seconds while infusing 150 ml BSS Plus (Alcon) at 85 cm H20. At the end of the procedure, the rabbit was euthanized, corneas excised and the endothelial F-actin was stained with NBD-Phallacidin to demonstrate cellular borders and cytoskeleton. Results showed no cell damage with anterior chamber irrigation if there was no air bubble. Irrigation of anterior chamber with stationary air bubble for 2 minutes caused endothelial cell damage where the bubble was in contact with the endothelium. Phaco without air bubbles caused isolated cell loss with or without viscoelastics. However, if air bubbles occurred during phaco, many areas of endothelial cells loss were observed with either BSS Plus or Healon. Viscoat prevented endothelial cell damage from the air bubbles. A similar finding occurred during in vitro specular microscope perfusion when an air bubble remained adjacent to the endothelium for 15 min. In a final series, air injected into the anterior chamber without irrigation resulted in no endothelial cell damage. This study demonstrates that air bubble damage during phaco is dependent upon the absence of aqueous protein and viscoelastic substance and not solely upon traumatic contact with the air bubbles.

Grajewski AL, Cristol SM, Wright MM, Parrish RK. Reduced Dose of 5-Fluorouracil in Glaucoma Filtering Surgery. Ophthalmology 1991; 98(suppl):109.

The use of 5-fluorouracil with trabeculectomy in Iridocorneal Endothelial Syndrome (ICE) has not been reported. We reviewed nine cases of ICE treated with trabeculectomy and 5-fluorouracil. Subjects were 30-62 years old. All had previous intraocular surgery, including eight subjects with at least one previous trabeculectomy. Five were successful after an average of 12 months. Three failed within 3-12 months. One subject required repeated Nd:YAG laser treatment of the internal ostomy to control pressure. 5-fluorouracil does not necessarily prevent trabeculectomy failure in ICE patients with previously failed filters.

Cristol SM, Edelhauser HF, L’Hernault NL, Grossniklaus H, Waring GO. Correlation of Histologic and Ultrastructural Changes of Edematous Corneal Stroma. Invest Ophthalmol Vis Sci 1991; 32(suppl):775.

Paraffin embedded histologic sections of corneal stroma show a usual pattern of artifactual clefting. Stromal edema is suggested by a loss of these clefts and an increase in stromal thickness. Often the loss of artifactual stromal clefting is an inhomogeneous phenomenon, but the relation of this inhomogeneity to edema has not been previously studied. To investigate the significance of this phenomenon, human corneal buttons were identified which had loss of artifactual clefting of only the anterior or posterior stroma. The corneas had been excised at keratoplasty for either bullous keratopathy or Fuchs' dystrophy. Thick sections of tissue embedded in epoxy resin for electron microscopy were examined by light microscopy and showed no inhomogeneous stromal changes. Ultrathin sections were examined by transmission electron microscopy for diffuse or focal increases in interfibrillar spacing as these findings have previously been reported in association with chronic stromal edema. Our findings showed that stromal edema correlated with diffuse increases in interfibrillar spacing resulting in more than a 30% decrease in fibril density. Focal increases in interfibrillar spacing was also seen with the formation of "lakes" of extracellular material with largest dimensions greater than 8500 nm. All corneas demonstrated these ultrastructural changes in the anterior, middle, and posterior stroma. While edematous changes were present in each stromal region, they were more pronounced in the region of loss of artifactual clefting.

Wright MM, Cristol SM, Grajewski AL. Use of 5-Fluorouracil Following Trabeculectomy in Patients with Iridocorneal Endothelial (ICE) Syndrome. Ophthalmology 1990; 97(suppl):142.

5-Fluorouracil after trabeculectomy (maximum dose 105 mg in a multicenter trial) has been shown to reduce filter failure but is associated with numerous side effects. In order to minimize side effects while maintaining efficacy, a reduced dose (maximum 50 mg) of 5-FU was given after trabeculectomy to 37 eyes with previous cataract or filtering surgery. Nineteen eyes (51%) developed punctate corneal staining and 2 eyes (5%) had an epithelial defect after the second post-operative week. Eight eyes (22%) failed to maintain intraocular pressure below 22 mmHg at the one year visit. Our success rate is similar to the success rate in the multicenter trial, but with fewer side effects.

Guildford JH, Alfonso E, Cristol SM, Roussel TJ, Culbertson WW, Forster RK. Large Therapeutic Penetrating Keratoplasty. Ophthalmology 1988; 95(suppl):161-2.

Extensive corneal disease can result in frank or impending corneal perforation necessitating a large therapeutic penetrating keratoplasty. In an eight-year period, 33 therapeutic penetrating keratoplasties with recipient beds 9.5 mm or greater were performed on a total of 29 eyes. Microbial keratitis was diagnosed in 22 eyes. Of those, 11 were bacterial, ten were fungal, and one was mixed bacterial and fungal. Other etiologies included sterile stromal melts, epithelial downgrowth and old trauma. Structural integrity of the globe was maintained in 24 of the 29 eyes (83%). The remainder required enucleation or progressed to phthisis bulbi. Graft failure was noted in 23 of 25 grafts (92%). Mean interval to graft failure was 4 weeks in fungal keratitis and 17 weeks in bacterial keratitis. With structural integrity preserved, a future smaller optical penetrating keratoplasty is an option in many of these eyes.

Guildford JH, Alfonso E, Roussel TJ, Culbertson WW, Cristol SM, Forster RK. Large Therapeutic Keratoplasty in Microbial Keratitis. Invest Ophthalmol Vis Sci 1988; 29(suppl):6.

Progressive microbial keratitis can result in frank or impending corneal perforations necessitating therapeutic penetrating keratoplasty. With extensive corneal involvement, large penetrating grafts are often necessary to restore the integrity of the globe and intraocular contents.

From January 1979 through July 1987, 24 therapeutic penetrating keratoplasties with recipient beds 9 mm or greater were performed for microbial keratitis at the Bascom Palmer Eye Institute. These procedures were completed on a total of 20 eyes. Of those, 11 were fungal, 8 were bacterial and one was mixed bacterial and fungal. Fusarium species was identified in 11 eyes. Pseudomonas aeruginosa was cultured from 6 eyes. Recurrence of infection in the graft was noted in five eyes (20%). Four of these underwent repeat keratoplasty with subsequent resolution of the infection in three cases. Structural integrity of the globe was maintained in 19 of 20 eyes (95%), however, graft failure was noted in 17 of 20 eyes (85%). Those grafts remaining clear have follow-up less than three months. Eyes with fungal keratitis failed earlier than those eyes with bacterial keratitis. Mean interval to graft failure was 4 weeks in fungal keratitis and 16 weeks in bacterial keratitis. Of the 20 eyes, 11 (55%) had uncontrolled intraocular pressure elevation above 30 mm Hg despite topical and systemic pressure reducing medication.