ARVO 2 Annual Meeting Abstracts 273 Corneal Extracellular ... · Program Number: 2357 Poster Board Number: A0086 Presentation Time: 3:45 PM–5:30 PM p38/TGFβ-signaling pathway involved

ARVO 2016 Annual Meeting Abstracts

These abstracts are licensed under a Creative Commons Attribution-NonCommercial-No Derivatives 4.0 International License. Go to http://iovs.arvojournals.org/ to access the versions of record.

273 Corneal Extracellular matrix and biomechanicsMonday, May 02, 2016 3:45 PM–5:30 PMExhibit/Poster Hall Poster SessionProgram #/Board # Range: 2355–2400/A0084–A0129Organizing Section: Cornea

Program Number: 2355 Poster Board Number: A0084Presentation Time: 3:45 PM–5:30 PMIdentifying the role of PDGFRα in the fibrotic pathway of TGF -β3Sriniwas Sriram1, 2, Jennifer A. Tran1, Xiaoqing Q. Guo1, 2, Hetian Lei1, 2, Andrius Kazlauskas1, 2, James D. Zieske1, 2. 1Schepens Eye Research Institute/MEE, Boston, MA; 2Ophthalmology, Harvard Medical School, Boston, MA.Purpose: The difference in function between two highly similar transforming growth factor beta (TGF-β) isoforms, TGF-β1 and -β3 (T1 and T3, respectively), still has not been clearly elucidated in the literature. The purpose of this study was to develop a 3-dimensional (3D) cell culture model that aimed to identify this difference in function in vitro. Our hypothesis was that T1 and T3 show varied effects on fibrosis by differentially activating the platelet derived growth factor receptor alpha (PDGFRα).Methods: Human corneal fibroblasts (HCF) were infected with lentivirus containing optimized short-hairpin RNA (shRNA) sequences for PDGFRα, which permanently knocked down the PDGFRα in these cells (HCF-P). 3D constructs of HCF or HCF-P were cultured on polycarbonate inserts for 4 weeks ± 0.1ng/ml T1 or T3. At the end of 4 weeks, the constructs were processed for immunofluorescence (IF) and RT-qPCR analysis. RNA from these constructs also was analyzed with a custom mRNA array that looked at 86 key fibrosis genes.Results: IF and RT-qPCR results showed that when compared to HCF no growth factor control, SMA was not stimulated as much by T3 as T1; however, in HCF-P, SMA expression remarkably increased with the presence of T3. The mRNA array results showed that in HCF, T3 stimulated a majority of the fibrotic markers that T1 stimulated but to a lesser degree—ITGB6, CTGF, PDGF and SERPINA1. There also were some signaling factors that were downregulated by T3 when compared to T1—Integrin’s (ITGAV, ITGA2), MMPs (MMP3, MMP14), STATs (STAT1, 6) and SMADs (SMAD2-7). When we looked at the difference between T3 treatment in HCF and HCF-P to identify the source of T3’s fibrotic response, STAT3, SRC and ITGAV were some of the signaling growth factors that were highly upregulated in HCF-P.Conclusions: The presence or absence of PDGFRα elicited contrasting fibrotic responses to the same T3 treatment. Identifying the signaling factors that make T3 cause fibrosis may help in understanding how it differs from T1, and may potentially help in developing a treatment that stimulates wounding with reduced levels of scarring.Commercial Relationships: Sriniwas Sriram, None; Jennifer A. Tran, None; Xiaoqing Q. Guo, None; Hetian Lei, None; Andrius Kazlauskas, None; James D. Zieske, NoneSupport: NIH Grant R01EY005665 (JDZ), NIH Grant P30EY03790 (core), and Massachusetts Lions Research Fund 2014-15

Program Number: 2356 Poster Board Number: A0085Presentation Time: 3:45 PM–5:30 PMUtility of TGFBI screening in the diagnosis of suspected paraproteinemic keratopathyMarina Zakharevich1, Tova Mannis2, Gabriel Mannis2, Jennifer Rose -Nussbaumer2, Anthony J. Aldave1. 1Stein Eye Institute, Los Angeles, CA; 2University of California San Francisco, San Francisco, CA.Purpose: To describe the clinical utility of molecular genetic analysis in the diagnosis or exclusion of suspected corneal dystrophies.Methods: Slit lamp examination and optical coherence tomography (OCT) were performed on a 52 year old female referred for evaluation of bilateral corneal stromal opacities. After informed consent was obtained, genomic DNA was collected, and PCR amplification and cycle sequencing of all 17 exons and exon-intron boundaries of TGFBI were performed.Results: Slit lamp examination demonstrated bilateral discrete anterior stromal opacities, which were also demonstrated by OCT to be confined to a single plane in the anterior corneal stroma. While the appearance of the opacities was consistent with granular corneal dystrophy, they were located in only the temporal cornea in a crescentic pattern, sparing the central cornea. This pattern, the late onset and the absence of a family history were atypical of GCD. Screening of TGFBI did not reveal a pathogenic mutation, excluding GCD1 and other TGFBI-related corneal dystrophies as the cause of the observed corneal phenotype. The patient was subsequently referred for evaluation to exclude paraproteinemic keratopathy, which identified chronic lymphocytic leukemia (CLL) as the potential systemic cause of the corneal findings.Conclusions: Paraproteinemic keratopathies may present with phenotypic similarities to a variety of corneal dystrophies. Genetic testing, which is available for most corneal dystrophies, provides the clinician with a definitive means of differentiating between inherited and acquired disorders.Commercial Relationships: Marina Zakharevich; Tova Mannis, None; Gabriel Mannis, None; Jennifer Rose-Nussbaumer, None; Anthony J. Aldave, NoneSupport: NEI Grant 1R01 EY022082 (A.J.A.), NEI Grant P30 EY000331 (core grant), the Walton Li Chair, the Stottr fund and an unrestricted grant from Research to Prevent Blindness

Program Number: 2357 Poster Board Number: A0086Presentation Time: 3:45 PM–5:30 PMp38/TGFβ-signaling pathway involved in induction of thrombospondin-1 in human corneal fibroblastsAudrey E. Hutcheon1, 2, Xiaoqing Q. Guo1, 2, Sriniwas Sriram1, 2, Jennifer A. Tran1, James D. Zieske1, 2, Audrey E. Hutcheon1, 2. 1Schepens Eye Research Institute/MEE, Boston, MA; 2Ophthalmology, Harvard Medical School, Boston, MA.Purpose: Previous studies have shown that competitive inhibition of the SMAD/TGFβ-signaling pathway does not reduce the expression of Thrombospondin-1 (TSP1), a known downstream TGFβ-target protein, in human corneal fibroblasts (HCF). In addition, when HCF were stimulated with TGF-β1 (T1), they were induced to produce more TSP1, as well as, fibrotic markers, cellular fibronectin (cFN) and collagen type III (Col III). Therefore, in this study, we hypothesized that an alternate TGFβ-signaling pathway—the p38 pathway—induces TSP1 in HCF, and may also affect cFN and Col III.Methods: Human corneal fibroblasts (HCF) were grown in EMEM + 10% FBS until they were 60-70% confluent, at which time the cells were cultured overnight in EMEM only, to serum starve the cells. The next day, HCF were cultured as follows for 24 hours: 1) Control: EMEM only; 2) T1: EMEM + 2ng/ml T1; 3) p38inh: EMEM



+ 10μM of p38 inhibitor (SB202190); and 4) Both: EMEM + 2ng/ml T1 + 10μM p38 inhibitor. Cells then were harvested and processed for western blot or quantitative real-time polymerase chain reaction (qRT-PCR) for TSP1, cFN, and Col III. Experiments were repeated at least three times.Results: TSP1 protein and mRNA is present in serum starved HCF (control), and as expected, upon treating these cells with T1, TSP1 protein increased; however, the mRNA remained relatively similar to control. In the p38inh samples, both TSP1 protein and mRNA dramatically decreased when compared with both the control and T1 samples. When both T1 and p38 inhibitor were introduced, TSP1 mRNA levels dramatically decreased, similar to the p38inh samples; however, TSP1 protein levels only decreased compared to T1 samples. Col III and cFN responded in a similar manner as TSP1.Conclusions: Interestingly, the p38 inhibitor did decrease the protein and/or mRNA levels of TSP1, Col III, and cFN in HCF, thus showing that in HCF these proteins are downstream from the p38/TGFβ-signaling pathway. Also, the discrepancy between mRNA and protein levels for TSP1 suggests that there are other factors involved in either the stabilization or degradation of TSP1 in HCF.Commercial Relationships: Audrey E. Hutcheon, None; Xiaoqing Q. Guo, MEE 2015-575 - Zieske; ML 36770-549P01US (P); Sriniwas Sriram, None; Jennifer A. Tran, None; James D. Zieske, MEE 2015-575 - Zieske; ML 36770-549P01US (P); Audrey E. Hutcheon, NoneSupport: NIH Grant R01EY005665 (JDZ), NIH Grant P30EY03790 (core)

Program Number: 2358 Poster Board Number: A0087Presentation Time: 3:45 PM–5:30 PMComparison of human corneal fibroblasts cultured in human or fetal bovine serum in an in vitro 3D modelWenjing Wu1, 2, Shrestha Priyadarsini3, James D. Zieske1, 2, Dimitrios Karamichos3, 4. 1Schepens Eye Research Institute/MEE, Boston, MA; 2Ophthalmology, Harvard Medical School, Boston, MA; 3Ophthalmology/Dean McGee Eye Institute, University of Oklahoma Health Sciences Center, Oklahoma City, OK; 4Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK.Purpose: TGF-β isoforms are known for their effects on corneal fibrosis with TGF-β1 and -β2 being pro-fibrotic and TGF-β3 being anti-fibrotic. The purpose of this study was to investigate the differences between human serum (HS) and fetal bovine serum (FBS) on human corneal fibroblasts (HCF) in an in vitro 3D culture model, as well as, examine HS’s influence on corneal stromal fibrosis.Methods: HCF were cultured for 2 weeks in EMEM + 0.5mM Vitamin C (2-O-α-D-glucopyranosyl-L-ascorbic acid) with either HS or FBS ± TGF-β1, -β2, or -β3. After 2 weeks, samples were collected, processed for qRT-PCR and/or western blot, and examined for alterations in fibrotic markers—α-smooth muscle actin (α-SMA) and collagen types III (Col III) and V (Col V)—as well as, focal adhesion kinase (FAK) and matrix metalloproteinases 1 (MMP1) and 3 (MMP3).Results: In HCF controls, Col III was significantly downregulated (p<0.05) and MMPs 1 and 3 (p<0.0001 for both) were upregulated in HS compared to FBS; however, there was no difference in α-SMA or Col V expression. For all TGFβ isoforms, α-SMA showed no significant difference between serums; however, Col III was significantly decreased (p<0.004) and Col V was significantly increased (p<0.0005) in HS as compared with FBS samples. When compared with their respective serum controls, both MMP1 and 3 in HS were significantly downregulated by the TGFβ isoforms, especially TGF-β3 (p<0.05); however, in the FBS samples, MMPs

1 and 3 were both upregulated with TGFβ stimulation. Interestingly, TGF-β1 stimulated FAK (p<0.05), while TGF-β3 appeared to have no effect.Conclusions: Our results show that HS could potentially replace FBS in our in vitro culture model, and offer a more translational in vitro corneal stroma equivalent. The use of HS may also provide insights into the development of a fibrotic phenotype. For example, HCF cultured with HS and TGF-β3 appeared to significantly inhibit degradation enzymes (ie. MMPs 1 and 3) and promote a non-fibrotic phenotype.Commercial Relationships: Wenjing Wu, None; Shrestha Priyadarsini, None; James D. Zieske, None; Dimitrios Karamichos, NoneSupport: NIH Grant R01EY020886 (JDZ and DK), NIH Grant R01EY005665 (JDZ), NIH Grant P30EY03790 (core)

Program Number: 2359 Poster Board Number: A0088Presentation Time: 3:45 PM–5:30 PMRAFT tissue equivalent model of keratoconus: A novel 3D system useful for interrogating disease phenotype and functional differences of corneal stromal cellsAlvena Kureshi1, Christine Putri1, James Phillips2, James L. Funderburgh3, Julie T. Daniels1. 1Institute of Ophthalmology, University College London, London, United Kingdom; 2UCL Eastman Dental Institute, London, United Kingdom; 3University of Pittsburgh, Pittsburgh, PA.Purpose: Keratoconus (KC) is a corneal disorder characterized by progressive structural thinning, resulting in a conical-shaped cornea causing loss of vision. Cell-extracellular matrix interactions are thought to control normal tissue organization and transparency and it is expected that these interactions may be perturbed in KC. The aim of this study was to develop a 3D RAFT tissue equivalent (TE) model of KC to compare the phenotypic and functional behavior of stromal cells.Methods: Stromal cells were characterised by phenotypic assessment using a variety of stromal cell markers and a contraction assay to compare their contractile capacity. RAFT TE were seeded with normal and KC stromal cells (n=3) and cultured for 3 weeks. Functional characterization of RAFT TE included investigations of material properties using an Electroforce Bose tensile testing device; MMP activity in stromal cells using gelatin zymography and ELISA; collagen production using a Sircol collagen assay and assessment of collagen fibril organization using scanning electron microscopy.Results: A 3D in vitro model of KC was achieved using RAFT TE. Normal and KC stromal cells both exhibited stromal cell markers in culture. KC stromal cells exhibited a significantly greater contractile capacity at 72 hours of culture with 75 ± 3% compared to normal cells at 29 ± 7% and acellular gels at 3 ± 1% (n=3) (p<0.0001). Statistically significant differences in RAFT TE material properties were observed. Mean break strength of KC and normal RAFT TE were 0.17 ± 0.02 MPa and 0.25 ± 0.03 MPa respectively (n=3) (p<0.05). A reduced stiffness of 0.01 ± 0.002 MPa in KC RAFT TE compared to 0.02 ± 0.002 MPa in normal RAFT TE was observed. Increased activity of MMP-2 in normal RAFT TE was observed but no differences in pro-MMP-1 or collagen production. Collagen fibril ultrastructure in KC RAFT TE appeared markedly thinner and loosely packed compared to normal RAFT TE.Conclusions: To our knowledge, this is the first 3D tissue equivalent model of KC. RAFT TE serves as an ideal biomimetic model to investigate normal and abnormal tissue organization in the corneal stroma. Significant differences in contractile capacity of KC cells and strength and stiffness of KC RAFT TE suggests the remodeling



capacity of stromal cells is perturbed in KC, leading to ultrastructural changes of the corneal stroma.Commercial Relationships: Alvena Kureshi, None; Christine Putri, None; James Phillips, None; James L. Funderburgh, None; Julie T. Daniels, NoneSupport: Special Trustees of Moorfields Eye Hospital

Program Number: 2360 Poster Board Number: A0089Presentation Time: 3:45 PM–5:30 PMThe distribution of elastic fibres in keratoconic corneas using en bloc tannic acid and orcein stainingTomas L. White2, Philip N. Lewis2, Koji Kitazawa1, Tsutomu Inatomi1, Shigeru Kinoshita1, Keith M. Meek2. 1Opthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan; 2Optometry and Vision Science, Cardiff Univeristy, CARDIFF, United Kingdom.Purpose: We have recently demonstrated the presence and distribution of elastic fibres in normal human cornea using serial block face scanning electron microscopy (SBF-SEM). These fibres formed sheets at the limbus before becoming long fibres that run transversely and longitudinally, along the plane of the cornea, concentrated in the ‘pre-Descemet’s layer’ (PDL). The aim of the current study was to use SBF-SEM and transmission electron microscopy (TEM) to determine if these findings are consistent, or differ, in human keratoconic corneas.Methods: Keratoconic corneal buttons were fixed in 4% paraformaldehyde (PFA) following penetrating keratoplasty. Buttons were then dissected, fixed in modified Karnovsky’s fixative and en bloc stained using 2 different protocols. The first involved using multiple staining solutions including osmium tetroxide, tannic acid, uranyl acetate and lead acetate, prior to dehydration and embedding in CY212 resin. The second protocol substituted tannic acid with orcein, with normal cornea and sclera samples being stained as controls. Polymerised blocks were attached to Gatan specimen pins and transferred to a Zeiss Sigma FEG equipped with a Gatan 3View system. Images of the block face surface were acquired every 50 nm by automated serial sectioning. 100nm gold sections were obtained from the same specimen pins and visualised using a Joel1010 TEM.Results: No fibres could be clearly identified using SEM due to a lack of contrast. However, fibres were clearly visible above Descemet’s membrane in the orcein stained control samples. TEM confirmed the presence of elastic fibres in the control samples and also revealed that there were very few fibres in the PDL region of both central and peripheral keratoconic samples. Many dark stained fibres were seen below the epithelium, in the anterior stroma of the thinned cone region in keratoconus tissue. Furthermore, the central scarred region of cornea contained a distinct mass of staining where numerous amounts of fibres were evident.Conclusions: Elastic fibre specific en bloc staining using tannic acid and orcein protocol has revealed a lack of fibres in the PDL region in keratoconic corneal buttons and an increased presence of fibres in the anterior stroma, below the epithelium. These findings are directly opposite to those that we have reported in normal cornea and may influence keratoconus morphology and/or pathogenesis.Commercial Relationships: Tomas L. White, None; Philip N. Lewis, None; Koji Kitazawa, None; Tsutomu Inatomi, None; Shigeru Kinoshita, None; Keith M. Meek, NoneSupport: MRC

Program Number: 2361 Poster Board Number: A0090Presentation Time: 3:45 PM–5:30 PMCellularization of 3D printed Recombinant Human Collagen type III scaffolds using corneal mesenchymal stem cellsSteffi Matthyssen1, Kurt Coppens2, Eleonora Ferraris2, Jennifer Patterson3, Marie-José Tassignon1, 4, Nadia Zakaria1, 4. 1Ophthalmology, University of Antwerp, Antwerp, Belgium; 2Advanced Manufacturing laboratory, Faculty of Engineering Technology, KU Leuven, Leuven, Belgium; 3Materials for Living Systems, Department of Materials Engineering, KU Leuven, Leuven, Belgium; 4Ophthalmology, Antwerp University Hospital, Edegem, Belgium.Purpose: To investigate cellularization of 3D printed human recombinant collagen type III (RHC III) scaffolds in vitro, using corneal mesenchymal stem cells (MSCs).Methods: Corneal MSC cultures were established using either a collagenase digestion (1.5hrs vs 4hrs) of the stroma or by plating stromal explants. Cells were seeded in either DMEM + 10% FBS or Epilife +5% FBS. At passage three, cells were characterized using Flow Cytometry and trilineage differentiation was performed using specific differentiation media. MSCs were seeded onto 3D printed (n=6) and non-printed (n=6) samples at a density of 20,000 cells/sample. For controls, cells were seeded on glass (n=6). At day 10, scaffolds were processed for immunocytochemistry (n=15) and Scanning Electron Microscopy (SEM) (n=3).Results: Flow cytometry showed that corneal MSCs were positive for CD73, CD90, CD105, CD13, CD29, CD44, CD166 and negative for CD11b, CD14, CD19, CD34, CD45, CD79a and HLA-DR. Corneal MSCs were able to differentiate into osteocytes, chondrocytes and adipocytes. There was no significant difference between the MSC phenotypes using the different isolation or culture protocols. Light microscopy showed that corneal MSCs proliferated on both printed and non-printed RHC III scaffolds and SEM showed the circular print pattern of the RHC III. On immunocytochemistry we observed collagen type III fibril alignment in the printed samples but not in the non-printed samples. The MSCs actin filaments of the cytoskeleton stained positive for phalloidin and we observed alignment orthogonally to the direction of the collagen.Conclusions: Our results demonstrate that 3D printed RHC III is a suitable substrate for cultivating corneal MSCs.Commercial Relationships: Steffi Matthyssen, None; Kurt Coppens, None; Eleonora Ferraris; Jennifer Patterson, None; Marie-José Tassignon, None; Nadia Zakaria, None

Program Number: 2362 Poster Board Number: A0091Presentation Time: 3:45 PM–5:30 PMCollagen-Proteoglycan Interactions in the Corneal Stroma: A Computer SimulationCarlo Knupp, Rob D. Young, Andrew J. Quantock, Keith M. Meek. Cardiff University, Cardiff, United Kingdom.Purpose: Recent three-dimensional electron microscopy studies carried out by our group (e.g. Lewis et al, Structure,18, 1-7, 2010) confirmed that the spatial distribution of collagen fibrils in the corneal stroma is regulated by the interaction of proteoglycans with the collagen fibrils. From these studies, we suggested that the relatively flexible proteoglycan chains make temporally transient links between adjacent collagen fibrils in a non-systematic, asymmetric way and that they give rise to forces that are ultimately responsible for the spatial distribution of the collagen fibrils in the cornea. Our purpose is to simulate these interactions computationally and verify whether they can explain the most salient structural and functional characteristics of the cornea.



Methods: Ad-hoc software was written to simulate the behaviour of the collagen fibrils under the influence of external forces brought about by the action of the proteoglycans.Results: Our computer simulation covers a 1 second-long time-dependent evolution of a portion of the corneal stroma (with up to 1000 fibrils) in which the collagen fibrils are subject to thermal motion and proteoglycan induced forces. The appearance of the simulated corneas is remarkably similar to what is seen in electron micrographs, as confirmed by the calculation of Fourier transforms and radial projection functions for the simulations and the micrographs. The calculation of transparency and X-ray diffraction patterns from the simulated corneas are also in good agreement with the data collected from real corneas. In addition, we have probed the possibility of other proteoglycan unrelated forces acting on the collagen fibrils, such as the pressure exercised by adjacent corneal lamellae, which shows that they may contribute substantially to the overall collagen fibril distribution in the cornea.Conclusions: The fibril distributions obtained in our simulations compare favourably with electron microscopical, X-ray diffraction and corneal transparency data, indicating that the local interactions between collagen and proteoglycans can explain some of the main structural and functional aspects of the cornea.Commercial Relationships: Carlo Knupp, None; Rob D. Young, None; Andrew J. Quantock, None; Keith M. Meek, NoneSupport: MRC Grant MR/K000837/1

Program Number: 2363 Poster Board Number: A0092Presentation Time: 3:45 PM–5:30 PMCytosporone B inhibits the TGF-β–induced expression of a–smooth muscle actin and contractility in human corneal fibroblastsNaoyuki Morishige, Yukiko Morita, Shizuka Murata. Ophthalmology, Yamaguchi Univ Grad Sch of Med, Ube, Japan.Purpose: Cytosporone B (CsnB) is an antagonist for the nuclear receptor transcription factor NR4A1, known as Nur77, which contributes to the regulation of apoptosis and glucose homeostasis. Recently, it is reported that NR4A1 is related to tissue scarring by regulating transforming growth factor–b (TGF-β) signaling. To investigate the role of NR4A1 in human corneal fibroblasts (HCFs), we examined the effects of CsnB on the expression of a–smooth muscle actin (α-SMA) and contractility in these cells induced by stimulation with TGF-β.Methods: Cultured HCFs were stimulated with TGF-β (10 ng/ml) in the presence of various concentrations (0 to 10 μM) of CsnB for 3 days, after which the cells were lysed and subjected to immunoblot analysis with antibodies to α-SMA. HCFs were also cultured in collagen gels and exposed to CsnB (1 μM), TGF-β (10 ng/ml), or neither or both agents for 3 days, after which the change in gel diameter was determined as a measure of cell contractility.Results: The TGF-β–induced expression of α-SMA in HCFs was inhibited by CsnB in a concentration-dependent manner. The diameter of HCF-containing collagen gels after culture for 3 days was 7.8 ± 0.8, 8.3 ± 0.3, 7.0 ± 0.3, or 8.0 ± 0.9 mm in the absence of additions or in the presence of CsnB, TGF-β, or both agents, respectively.Conclusions: CsnB inhibited TGF-β–induced contraction of HCFs, likely as a result of its attenuation of the up-regulation of α-SMA expression. Our results suggest that NR4A1 mediates these effects of TGF-β in HCFs, and that CsnB warrants further investigation as a potential therapeutic modulator of corneal stromal contraction and scarring.Commercial Relationships: Naoyuki Morishige, None; Yukiko Morita; Shizuka Murata, None

Support: JSPS KAKENHI grant number 15K20264

Program Number: 2364 Poster Board Number: A0093Presentation Time: 3:45 PM–5:30 PMChitosan membrane-cultivated keratocyte spheres suppress stromal fibrosis via inhibition of epithelial mesenchymal transition (EMT)Chi-Chin Sun1, 2, Kuo-Ying Lee1, Jong-Hwei S. Pang3. 1Department of Ophthalmology, Chang Gung Memorial Hospital, Keelung, Keelung, Taiwan; 2Department of Chinese Medicine, School of Medicine, Chang Gung University, Taoyuan, Taiwan; 3Institute of Clinical Medical Sciences, Chang Gung University, Taoyuan, Taiwan.Purpose: Previously, we demonstrated the anti-fibrotic effects of keratocyte spheres cultivated on chitosan membrane. The purpose of this study is to investigate the underlying mechanisms.Methods: Human corneas stripped of both endothelium and epithelium were placed in a solution of collagenase overnight at 37°C. Confluent cell layers were treated with trypsin-EDTA and seeded onto the chitosan coatings or plastic dishes. We used Affymetrix oligonucleotide microarrays to analyze common transcriptomes and thereby learn about the core gene expression profile in human keratocytes cultured on chitosan membrane on plastic dishes. The advantage of microarray analysis of gene expression is that it can be used to disclose associating genes that were previously thought to be unrelated to a physiological or pathological event. However, interpreting complex biological processes from gene expression profiles often requires extensive knowledge mining in biomedical literature. Therefore, we adopted a commercially available biological database and software program, MetaCore (GeneGo Inc.), for functional network analysisto reveal the common functions of the interacting genes. Real-time PCR, western blotting analysis and immunofluorescence staining were performed to validate the expressions of these genes at RNA and protein levels.Results: Analysis of gene expression by microarray and MetaCore pathway mapping revealed that regulation of epithelial-to-mesenchymal transition (EMT) was the most important pathway involved in keratocyte spheres cultured on chitosan membrane as compared with on plastic dishes (4.193E-12). Transforming growth factor beta receptor type 1 was significantly down-regulated in chitosan-cultivated keratocyte spheres. Cell motility-associated genes including caldesmon, tropomyosin-1, ACTA2 and ACTB were all down-regulated in keratocyte spheres. Real-time PCR, western blotting analysis and immunofluorescence staining confirmed the findings observed at transcriptional level.Conclusions: Our studies demonstrated that chitosan membrane-cultivated keratocyte spheres repaired corneal scarring through inhibition of transforming growth factor beta-induced epithelial mesenchymal transition (EMT).Commercial Relationships: Chi-Chin Sun, None; Kuo-Ying Lee, None; Jong-Hwei S. Pang, NoneSupport: NSC99-2314-B-182A-025-MY3

Program Number: 2365 Poster Board Number: A0094Presentation Time: 3:45 PM–5:30 PMConfocal microscopy findings after 3 different techniques: epi-off, trans-epithelial and iontophoresis-assisted cross-linkingManuel Ramirez, Emmanuel Cabrera, Everardo Hernandez-Quintela. Cornea and Refractive Surgery, APEC, Mexico City, Mexico.Purpose: To evaluate the stromal corneal findings in 3 different cross-linking techniques (epi-off, trans-epithelial and iontophoresis-assisted).Methods: 18 eyes of 18 keratoconus patients (group A) were treated by epi-off corss-linking technique, 15 eyes of 15 keratoconus



patients (group B) were treated by trans-epithelial crosslinking technique and 22 eyes of 26 keratoconus patients (group C) were treated by iontophoresis-assisted cross-linking technique. All eyes were examined with the confocal microscope (Confoscan 4, Fortune Technologies, Italy), a central scan of the total corneal thickness was taken before, 1 week and 1 month after cross-linking treatment. Corneal stroma was evaluated by using the Navis v. 3.5.0. Software (NIDEK, Multi-Instrument Diagnostic System, Japan).Results: In group A we found, after cross-linking treatment, activated keratocites from superficial stroma to 275.1 ± 85.9 µm stromal depth at 1 week and from superficial stroma to 324.9 ± 66.0 µm stromal depth at 1 month after treatment. In group B and group C we did not found activated keratocites in all stromal depth at 1 week and 1 month after cross-linking treatment.Conclusions: Activated keratocites were found at the corneal stroma in patients treated after epi-off cross-linking technique, in contrast patients treated with trans-epithelial and iontophoresis-assisted crosslinking techniques did not show activated keratocites at the corneal stroma after treatment.Commercial Relationships: Manuel Ramirez, None; Emmanuel Cabrera, None; Everardo Hernandez-Quintela, None

Program Number: 2366 Poster Board Number: A0095Presentation Time: 3:45 PM–5:30 PMHuman Stromal Lamellar Morphology and Its Relationship to Central to Peripheral Thickness ChangeJan P. Bergmanson, Carmen L Mosqueda, Alan R Burns. College of Optometry, University of Houston, Houston, TX.Purpose: The human cornea is thicker peripherally than centrally, but is this due to a greater number of lamellae in the peripheral cornea or an increase in lamellar thickness, or both? The purpose of this study was to determine the total number of lamellae present and the average lamellar thickness in the anterior, middle and posterior stroma, both in the central and extreme peripheral normal cornea.Methods: Three eye bank corneas were processed and examined with a Tecnai G2 12 Twin transmission electron microscope (FEI Company Hillsboro, OR) operating at 100 kV and magnified at 4200X. Montages of overlapping images from epithelium to endothelium were assembled for each cornea for extreme periphery and central regions. Lamellae on each montage were marked and counted using an established protocol. The stroma was divided into anterior, mid and posterior segments. Data were expressed as means ± SD and differences assessed using Analysis of Variance (ANOVA) and Tukey post-hoc tests with significance set at P ≤ 0.05.Results: The central/peripheral lamellar counts for each cornea (260/260; 231/235; 247/257) were remarkably similar. As the lamellae approach the central region they were frequently observed to branch. However, peripheral posterior lamellae were on average twice as thick as those in the central cornea (3.2 ± 0.4 and 1.6 ± 0.3 µm, respectively), with some lamellae reaching 10 µm in thickness. Anterior lamellae were thinner than mid-stromal lamellae, but only at the periphery.Conclusions: The human cornea is thicker in the periphery not because it has more lamellae but because it has thicker posterior lamellae. While lamellar branching is common, the total number of lamellae (~250) does not increase centrally. Therefore, some lamellae must stop short of crossing the cornea. Given that keratoconic cornea has considerably more lamellae (~350) at the cone apex, it will be important to establish whether this increase is also present peripherally.Commercial Relationships: Jan P. Bergmanson, None; Carmen L Mosqueda, None; Alan R BurnsSupport: NIH/NEI Grants EY018239 and P30EY007551

Program Number: 2367 Poster Board Number: A0096Presentation Time: 3:45 PM–5:30 PMDifferentiation of Normal Human Corneal Induced Pluripotent Stem Cells to Retinal Progenitor CellsRoy Joseph2, Om P. Srivastava2, Roswell R. Pfister1. 1Ophthalmology, Univ of Alabama at Birmingham, Birmingham, AL; 2Vision Sciences, University Of Alabama at Birmingham, Birmingham, AL.Purpose: To differentiate human corneal induced pluripotent stem cells (iPSC) to retinal progenitor cells.Methods: Normal human corneal fibroblasts from three donors were reprogramed directly to pluripotent stem cells (PSC) by delivering reprogramming factors in a single virus using 2A “self-cleaving” peptides, using a single polycistronic lentiviral vector co-expressing four transcription factors (Oct 4, Sox2, Klf4 and Myc) to yield induced pluripotent stem cells (iPSC). Phase-contrast images were obtained using our Olympus 1X71 inverted microscope. These iPS cells were immunofluorescentally characterized for stem cell markers (SSEA4, Oct4 and Sox2). The iPS cells were grown in embryoid medium (EM) [(DMEM F-12 medium containing 10% knockout serum replacement, 2% B27 supplement, 1% N2 supplement, 1% L-glutamine, 1% 100X NEAA, 1% penicillin/streptomycin, 1 ng/ml noggin, 1 ng/ml DKK-1, 1 ng/ml IGF-1 and 0.5 ng/ml bFGF], and plated at a density of 50 cell clumps/cm2 in ultralow attachment plates. Cell clumps were cultured for 5 days, and then embryoid bodies were removed and differentiated in EM +10 ng/ml noggin, 10 ng/ml DKK-1, 10 ng/ml IGF-1 and 1 ng/ml bFGF (Differentiation medium 1) in 6-well culture plates coated with poly-D-lysine, collagen, laminin and fibronectin. Media were changed every other day for 10 days with Differentiation media 1, then every other day for an additional 18 days with Differentiation media 1 +10 nm DAPT (N-[N-(3,5-Difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester, a γ- secretase inhibitor).Results: Immunofluorescence analysis of the colonies showed that they were positive for all the three stem cell markers (SSEA4, Oct4 and Sox2). The iPS cells were grown to differentiate for 33 days, and then immunocytochemical analysis of the cells showed expression of retinal photoreceptor markers, recoverin and rhodopsin.Conclusions: We successfully reprogramed normal human corneal fibroblasts using single polycistronic virus co-expressing four transcription factors. We were also able to differentiate the iPSC to retinal precursor cells which expressed photoreceptor markers, rhodopsin and recoverin. These retinal precursor cells are a valuable source which could be used for transplantation in retinal degenerative diseases.Commercial Relationships: Roy Joseph; Om P. Srivastava, None; Roswell R. Pfister, None

Program Number: 2368 Poster Board Number: A0097Presentation Time: 3:45 PM–5:30 PMCellular Factor XIII is Present in the Corneal StromaZsuzsanna Z. Orosz1, 2, Amir H. Shemirani2, Helga Bárdos3, Bence Nagy4, Andrea Facskó1, András Berta5, Róza Ádány3, László Muszbek2, 6. 1Department of Ophthalmology, University of Szeged, Szeged, Hungary; 2Division of Clinical Laboratory Science, Department of Laboratory Medicine, University of Debrecen, Debrecen, Hungary; 3Department of Preventive Medicine, University of Debrecen, Debrecen, Hungary; 4Department of Pathology, University of Szeged, Szeged, Hungary; 5Department of Ophthalmology, University of Debrecen, Debrecen, Hungary; 6Thrombosis, Hemostasis and Vascular Biology Research Group of the Hungarian Academy of Sciences, University of Debrecen, Debrecen, Hungary.



Purpose: Transglutaminases (TGs) are a family of enzymes that cross-link proteins by e(g-glutamyl)lysyl bonds and most of them have been implicated in the modulation of extracellular matrix. Here we investigated the presence of three TGs, keratinocyte TG (TG-1), tissue transglutaminase (TG-2) and the cellular form of blood coagulation factor XIII (cFXIII) in the corneal tissue.Methods: Frozen sections of normal human cornea obtained from enucleated bulbus were stained for cFXIII, TG-1 and TG-2 using poly-, or monoclonal antibodies. Detection of cFXIII was also combined with labeling for CD11b, CD34, CD45, CD68 and CD163 using double immunofluorescent staining. FITC-labeled or biotinylated secondary antibodies with Texas red-labeled streptavidin were used for the visualization of immunoreactions. Western blot analysis was also performed on corneal stroma, and the expression of cFXIII mRNA was analyzed by real-time qPCR.Results: A significant part of keratocytes showed intensive staining for cFXIII, but not for TG-1 and TG-2. Neither epithelial nor endothelial cells were labeled by anti-cFXIII antibody. cFXIII positive keratocytes were unevenly distributed in the corneal stroma; they were abundant in the subepithelial tertile of stroma (120±10/visual field), while they were sparse (38±6/visual field) in the subendothelial tertile. cFXIII+ cells showed co-staining for CD34, however, a significant number of CD34+ cells were negative for cFXIII. CD34+ cells were evenly distributed throughout the stroma. Only a few cells were stained for CD11b and CD45, they were also labeled by anti-cFXIII antibody. No cell showed positivity for CD68 and CD163. cFXIII expression was confirmed by Western blotting, and real-time qPCR demonstrated the presence of cFXIII mRNA in the corneal stroma.Conclusions: This is the first report demonstrating the presence of cFXIII in the cornea. A significant part of CD34+ keratocytes contained cFXIII, their transglutaminase activity might play a role in wound healing of the cornea. Further investigations are needed to explore the role of FXIII in corneal wound healing.Commercial Relationships: Zsuzsanna Z. Orosz; Amir H. Shemirani, None; Helga Bárdos, None; Bence Nagy, None; Andrea Facskó, None; András Berta, None; Róza Ádány, None; László Muszbek, NoneSupport: TÁMOP 4.2.2/A-11/1/KONV-2012-0045; TÁMOP 4.2.2/B-10/1-2010-0024

Program Number: 2369 Poster Board Number: A0098Presentation Time: 3:45 PM–5:30 PMInvestigation of the Corneal Biomechanics in Healthy Chinese Subjects Using a Corneal Visualization Scheimpflug TechnologyMiao He1, Hui Ding2, Xingwu Zhong2, 1. 1Zhongshan Ophthalmic Center, Sun Yat-sen University, Guangzhou, China; 2Hainan Eye Hospital, Zhongshan Ophthalmic Center, Sun Yat-senUniversity, Haikou, China.Purpose: In vivo quantification of corneal biomechanics became local focus since the introduction of Corneal Visualization Scheimpflug Technology (CorVis ST) in recent years. The aim of this study was to evaluate corneal biomechanical properties in a Chinese healthy population using this device.Methods: Healthy subjects with 20 years and older were enrolled. All eyes underwent complete ocular examinations, including slit-lamp examination, auto refractometer, intraocular pressure (IOP), and CorVis ST. Corneal biomechanical parameters included maximum deformation amplitude (MDA), first applanation time (A1T), first applanation length (A1L), first applanation velocity (A1V), second applanation time (A2T), second applanation length (A2L), second applanation velocity (A2V), highest concavity time (HCT), peak distance (PD), and central radius of curvature (CCR). The Pearson

correlation analysis was used to evaluate the correlations of corneal biomechanical parameters with age, central corneal thickness (CCT), spherical equivalent of refractive error (SE), gender, and IOP.Results: 150 subjects were included with a mean age was 56.4±6.5 years. Mean SE was -1.65±1.07 diopters, the mean IOP was 14.5±2. 4 mmHg. Mean CCT was 549.72±29.96mm. There were no significant differences in SE, IOP, or CCT between male and female. Mean deformation amplitude was 1.07 ±0.04 mm. Highest concavity time was 17.38 ± 0.10 mm. A1T, A2T were 7.16 ± 0.46 ms and 22.10 ±0.37 ms, respectively. Most (9/10) CorVis ST parameters were independently associated with IOP, but only 4 parameters (MDA, A1L, A2L, A2V) were significantly correlated with CCT or age. Gender and SE did not influence CorVis ST parameters.Conclusions: The CorVis ST provided informative measurements of corneal biomechanical properties. No significant difference of corneal biomechanics existed between male and female. IOP, age, and CCT were correlated to corneal biomechanics.Commercial Relationships: Miao He, None; Hui Ding, None; Xingwu Zhong, NoneSupport: Science and Technology Planning Project of Guangdong and Hainan Province (ZDXM2014077,2013B090200057)

Program Number: 2370 Poster Board Number: A0099Presentation Time: 3:45 PM–5:30 PMComparative Study of the Effect of Pilocarpine and Tropicamide Eye Drops on Corneal TopographyAbbas Azimikhorasani, Shahrzad Mirzahoseini, Javad Heravian, Azamalsadat Parizad. Mashhad University of Medical Sciences, Mashhad, Iran (the Islamic Republic of).Purpose: This study was conducted to evaluate the effect of contraction and relaxation of ciliary muscle caused by pilocarpine and tropicamide eye drops on corneal radius, central corneal power and corneal astigmatism.Methods: This prospective study was performed on sixty normal and healthy eyes of sixty volunteers with a mean age of 38.19 year (range 18 to 49 years) and without any ocular pathology. Volunteers divided into 2 groups of thirty, in the first group corneal topography of both eye were measured before and 30 minutes after instillation of topical tropicamide 0.5% in only one eye, and the other eye was the control eye and no drop was given. In the second group, the same routine was performed, except that subject received on drop of pilocarpine 1% in one eye. Statistical comparison between groups for the central corneal power, corneal radius and corneal astigmatism were performed using paired t test.Results: In group 1, no significant changes were found in corneal radius, power and astigmatism, however, in group 2 subjects who have received pilocarpine eye drop, mean corneal radius value decreased significantly by 0.05 mm and mean corneal power increased by a plus power of +0.32 D. There was no significant change in corneal astigmatism in both groups.Conclusions: It seems that pilocarpine induced ciliary muscle contraction which may cause pressure on corneal limbus of cornea and scleral spur that resulted in changes in corneal curvature. But tropicamide eye drop did not affect corneal radius and other corneal parameters and corneal topography can be carried out after instillation of tropicamide eye drop.Commercial Relationships: Abbas Azimikhorasani, None; Shahrzad Mirzahoseini, None; Javad Heravian, None; Azamalsadat Parizad, None



Program Number: 2371 Poster Board Number: A0100Presentation Time: 3:45 PM–5:30 PMDepth-dependent extrafibrillar matrix stiffness of the human cornea predicts spherical aberration induced by laser refractive surgeryMengchen Xu1, 2, Mark Buckley3, Amy L. Lerner1, 3, Geunyoung Yoon2, 3. 1Mechanical Engineering, University of Rochester, Rochester, NY; 2Flaum Eye Institute, University of Rochester, Rochester, NY; 3Biomedical Engineering, University of Rochester, Rochester, NY.Purpose: To investigate the influence of depth-dependent corneal extrafibrillar matrix stiffness on wavefront aberrations induced by laser refractive surgery.Methods: Three-dimensional anisotropic hyperelastic corneal models with a regional fibril distribution (Pandolfi, 2008) were implemented in finite element software (ABAQUS 6.14). Two material cases were compared: (1) averaged uniform extrafibrillar matrix stiffness along depth, (2) five material layers with depth-dependent extrafibrillar matrix stiffness based on previous experimental data (Sloan, 2014). Three preoperative (preOP) model eyes with different axial lengths to induce 5, 7 and 9 diopters (D) myopic refractive error were generated for each material case. An iterative algorithm was used to identify the stress-free configurations of preOP eyes. Myopic corrections were simulated by removing tissue with ablation thickness calculated by the Munnerlyn equation for a 6 mm optical zone. Anterior and posterior corneal surface profiles were exported from each preOP and postoperative (postOP) model eye and used to calculate wavefront aberrations for a 6 mm pupil using a custom-developed surface fitting and ray-tracing program.Results: For the uniform matrix stiffness model, the trend observed for induced spherical aberration (SA) was opposite to the clinical observation. In contrast, the depth-dependent extrafibrillar matrix stiffness model showed a significant positive linear correlation between the induced positive SA and the amount of attempted diopter correction, in accordance with clinical data. Induced SA predicted after 5, 7 and 9D correction was 0.35, 0.48 and 0.63μm, respectively. PostOP refractive power results predicted that the biomechanical response to the tissue ablation would cause an under-correction in myopia of 15 – 20%. An increasing trend was also found for other aberrations such as astigmatism and quadrafoil with increasing amount of myopic correction.Conclusions: The uniform and depth-dependent extrafibrillar matrix stiffness of the human cornea lead to different trends of wavefront aberrations induced by laser refractive surgery when fibril distribution remains unchanged. Characterization of the depth-dependent matrix stiffness improves reliability in predicting clinical trends and may be key to understanding the role of biomechanics in optical outcomes after laser refractive surgery.Commercial Relationships: Mengchen Xu, None; Mark Buckley, None; Amy L. Lerner, None; Geunyoung Yoon, NoneSupport: NIH EY014999

Program Number: 2372 Poster Board Number: A0101Presentation Time: 3:45 PM–5:30 PMAnisotropic Diffusion in the Corneal Stroma Measured by Three-Dimensional Fluorescence Recovery After PhotobleachingR Glenn Hepfer1, Peng Chen1, George O Waring2, 1, Hai Yao1. 1Bioengineering, Clemson University, Charleston, SC; 2Ophthalmology, Medical University of South Carolina, Charleston, SC.Purpose: This study investigates transport properties within the corneal stroma. The corneal stroma is a highly organized tissue, with collagen fibers that run primarily in the lateral direction. Such

organization may give rise to anisotropic diffusion, i.e. diffusion that is dependent on direction. We hypothesize that anisotropic diffusion can be detected in the corneal tissue of whole ex-vivo eyes through three-dimensional fluorescence recovery after photobleaching (3D FRAP).Methods: Fresh, intact porcine eyes were obtained from a local slaughterhouse. The epithelia of the eyes were removed, and a solution containing a fluorescent probe was applied to the cornea. 3D FRAP was performed on the corneas and a custom Matlab program was used to calculate the diffusion coefficient in three directions: nasal-temporal (X), superior-inferior (Y) and anterior-posterior (Z). 3D FRAP was also performed on glutaraldehyde-treated corneas.Results: The diffusion coefficient in the Z direction was significantly decreased when compared to the diffusion coefficients in the X (P=0.004, ANOVA) and Y (P<0.0001, ANOVA) directions. There was no significant difference between the diffusion coefficients in the X and Y directions. The corneas treated with glutaraldehyde had significantly reduced diffusion coefficients in all directions when compared to normal corneas (P<0.0001, ANOVA). Furthermore, there was no significant difference between diffusion coefficients in different directions in the glutaraldehyde-treated corneas.Conclusions: Native porcine corneas exhibit anisotropic diffusion; the diffusion in directions parallel to collagen fibers is faster than the diffusion in directions orthogonal to collagen fibers. The glutaraldehyde treatment decreases the diffusion in the cornea and eliminates the anisotropy. The technique presented will be useful in examining the effects of certain corneal treatments (e.g. corneal cross-linking) as well as in developing tissue-engineered corneas that mimic the material properties of the native tissue.Commercial Relationships: R Glenn Hepfer, None; Peng Chen, None; George O Waring, None; Hai Yao, None

Program Number: 2373 Poster Board Number: A0102Presentation Time: 3:45 PM–5:30 PMCorneal Thickness is Positively Correlated with Tissue Elasticity by Elastography ImagingMichael D. Twa1, Srilatha Vantipalli2, Manmohan Singh3, Jiasong Li3, Kirill Larin3. 1Vision Science, University of Alabama at Birmingham, Birmingham, AL; 2Optometry, University of Houston, Houston, TX; 3Biomedical Engineering, University of Houston, Houston, TX.Purpose: Corneal thickness and hydration varies considerably in normal and diseased eyes. There are conflicting reports from previous studies regarding the influence of tissue hydration on corneal biomechanics. Here we evaluated corneal elasticity as a function of corneal thickness using optical coherence elastography (OCE).Methods: Fresh whole rabbit eyes (n=10) were deepithelialized, then treated with 0.9% saline every 5 minutes for 60 minutes, followed by 20% Dextran every 5 minutes for 60 minutes. A focused air pulse system was used to induce low amplitude (≤ 10 μm) tissue deformation at the corneal apex that was imaged using a phase-stabilized swept source optical coherence tomography system. Tissue elasticity was measured along with total corneal thickness every 20 minutes. Tissue elasticity was quantified by the dynamic natural frequency, which previous work has shown is linearly related to the square root of Young’s modulus.Results: Although corneal thickness changed little over the saline treatment period, measured elasticity changed by 4% (P > .05). Dextran treatment reduced corneal thickness by 250 μm and reduced measured corneal elasticity by 30% (P < .0001). These two parameters were highly correlated (R2 = .980; P = .0001).Conclusions: Corneal thickness changes due to variations in tissue hydration are positively correlated with measured tissue elasticity by OCE. Further studies are needed to help separate the effects due



to corneal thickness alone and those due to corneal hydration, which may affect both elastic and viscous components of the observed viscoelastic response.

Average central corneal thickness and elasticity of whole fresh rabbit eyes (n=10) over time. Deepithelialized corneas were treated with 0.9% saline every 5 minutes for 60 minutes, then with 20% dextran. Central corneal thickness was measured using OCT and elasticity was quantified as the dynamic natural frequency from optical cohernce elastograhy imaging.Commercial Relationships: Michael D. Twa; Srilatha Vantipalli, None; Manmohan Singh, None; Jiasong Li, None; Kirill Larin, NoneSupport: NIH/NEI R01-EY022362

Program Number: 2374 Poster Board Number: A0103Presentation Time: 3:45 PM–5:30 PMAdhesion Strength of Descemet Membrane in Healthy and Diabetic Donor CorneasChaid Schwarz1, Benjamin T. Aldrich2, 3, Kimberlee Burckart3, Gregory Schmidt3, Cynthia Reed3, Mark A. Greiner2, 3, Edward A. Sander1. 1Biomedical Engineering, The University of Iowa, Iowa City, IA; 2Department of Ophthalmology and Visual Sciences, The University of Iowa, Iowa City, IA; 3Iowa Lions Eye Bank, Coralville, IA.Purpose: Descemet membrane endothelial keratoplasty (DMEK) is an increasingly popular surgical procedure for treating diseases involving the corneal endothelium. Anecdotal evidence suggests that differences in the force required to peel the Endothelium-Descemet membrane complex (EDM) during DMEK surgical preparation may contribute to the greater incidence of preparation failures observed in diabetic compared to non-diabetic donor tissues (Greiner et al., 2014). Here, we report on the development of a mechanical testing method for quantifying the peel strength of human donor corneas, and our progress on quantitatively assessing the putative differences between non-diabetic and diabetic tissues.Methods: Corneoscleral tissues stored less than 14 days from procurement from control non-diabetic, non-severe diabetic, and severe diabetic donors 50-75 years old were used in this study. To reflect loading conditions used in surgical preparation, we applied cantilever bending to rectangular sections of corneal EDM during a continuous peel test. Using a thin Nitinol bar anchored to peripheral tissue with a calibrated force-displacement relationship, a displacement of 0.25 mm/s was applied to the EDM in order to separate EDM from the stromal layer over the length of the

tissue section. Video tracking and Matlab were used to analyze bar displacement and calculate the force required to separate EDM from stroma over the course of the peel.Results: No significant difference was found in the average peel force/width (g/mm) necessary to separate EDM form stroma in non-diabetic (0.082 ± 0.011) and non-severe diabetic (0.079 ± 0.004) donor tissue (p = 0.907). In contrast, severe diabetic donor tissue required significantly greater force/width to peel EDM (0.103 ± 0.028) compared to both non-diabetic (p = 0.038) and non-severe diabetic (p = 0.029) tissue samples.Conclusions: We have for the first time quantified the adhesion strength between Descemet membrane and the underlying stroma in human corneas. Differences in adhesion strength amongst non-diabetic, non-severe diabetic, and severe diabetic tissues suggest that severe diabetes mellitus may be a risk factor influencing donor preparation failures.Commercial Relationships: Chaid Schwarz, None; Benjamin T. Aldrich, None; Kimberlee Burckart, None; Gregory Schmidt, None; Cynthia Reed, None; Mark A. Greiner, None; Edward A. Sander, None

Program Number: 2375 Poster Board Number: A0104Presentation Time: 3:45 PM–5:30 PMDecreased corneal biomechanical properties in children with osteogenesis imperfectaBrenda L. Bohnsack, Lisa Lagrou, Sayoko E. Moroi. University of Michigan, Ann Arbor, MI.Purpose: Osteogenesis imperfecta (OI) is a genetic disease due to type I collagen defects. In adults, OI is associated with degenerative ocular pathologies, however, eye findings in children are not well-described.Methods: A prospective, comparative, cross-sectional study of pediatric patients (2-19 years) diagnosed with OI. Thirteen patients with OI and five control patients underwent complete ophthalmic examinations. Additional tests included ocular response analyzer, pachymetry, axial length measurement, automated perimetry, and, retinal nerve fiber layer (RNFL) thickness. Statistically significant differences between OI and control groups were determined using independent-samples t test.Results: We found significantly (p<0.05) decreased corneal hysteresisResults: We found significantly (p<0.05) decreased corneal hysteresis (8.2±1.4mmHg vs. 10.5±1.3), and central corneal thickness (CCT; 451.9 ± 29.1mm vs. 550.5±30.7 mm) in OI patients compared to controls. There was no significant difference in corneal resistance factor (9.9±1.3 mmHg vs. 10.5 ±0.8 mmHg). None of these differences correlated with age, race, or gender. IOP corrected for corneal thickness and hysteresis was significantly higher in OI patients than controls (19.1 ± 3.0 mmHg vs. 13.7 ± 0.7 mmHg). There were no differences in keratometry, axial length and RNFL thickness in OI patients compared to controls.Conclusions: Strength, flexibility and organization of type-I-collagen are critical for corneal structure and function. In our study, we found that children with OI had decreased corneal hysteresis and CCT, which resulted in artificially low IOP readings. In adults, OI has been associated with primary open angle glaucoma and keratoconus. Our findings suggest that in OI, alterations in corneal biomechanical properties are present at a young age and affected individuals should be routinely screened for glaucoma and corneal pathologies.Commercial Relationships: Brenda L. Bohnsack, None; Lisa Lagrou, None; Sayoko E. Moroi, None



Support: NIH Grant K08EY022912-01; Research to Prevent Blindness Career Development Award; Edward Mallinckrodt Jr Foundation Grant; Paul Lichter, M.D. Research Discovery Fund; Alcon Research Institute Young Investigators Grant

Program Number: 2376 Poster Board Number: A0105Presentation Time: 3:45 PM–5:30 PMPatient-Specific Computational Analysis of Laser in situ Keratomileusis: A clinical validation studyIbrahim Seven, Ali Vahdati, William J. Dupps. Ophthalmology, Cleveland Clinic Cole Eye Institute, Mayfield Hts, OH.Purpose: To develop a microstructural based, 3D, patient specific, finite element model (FEM) of LASIK and compare the outcomes of the simulations to actual post-LASIK follow-up tomographies.Methods: Patient-specific tomographic data from 20 eyes of 12 patients undergoing sphero-cylindrical LASIK treatment were meshed including epithelium, flap, wound, and residual stromal bed using a custom meshing software. Case-specific treatment settings were simulated. Simulated keratometry values (SimK) including the average curvature of the steep (K1), flat (K2) meridians, and central 3mm circular region (Kmean) were obtained from the anterior surfaces of both the actual tomographies and their generated FEMs. Mean difference (MD) and mean absolute difference (MAD) between these values were calculated to assess prediction error. Pearson correlation values (R) between individual errors in predicted Kmean and age, corneal hysteresis (CH), and corneal resistance factor (CRF) were calculated.Results: MD and MAD between simulated and actual post-LASIK cases were -0.13 ± 0.36 D and 0.28 ± 0.25 D respectively. The differences between simulated and actual SimK values were not statistically significant (p values for Kmean=0.1, K1=0.4, K2=0.5). R values between the individual differences in Kmean and CH, CRF, and age were 0.63, 0.53, and -0.5, respectively (p = 0.004, = 0.01, and = 0.02, respectively).Conclusions: The investigated computational modeling approach demonstrated low prediction errors and may be useful for clinical guidance in planning LASIK. Clinical biomechanical metrics and surrogates explain some of the variance in prediction error and may be useful to tailor generic material properties into patient specific material properties to increase prediction accuracy.Commercial Relationships: Ibrahim Seven, None; Ali Vahdati, None; William J. Dupps, Cleveland Clinic/Optoquest (P)Support: R01 EY023381, Ohio Third Frontier Innovation Platform Award TECH 13-059, Unrestricted Grant from Research to Prevent Blindness to the Dept. of Ophthalmology of the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University.

Program Number: 2377 Poster Board Number: A0106Presentation Time: 3:45 PM–5:30 PMWST11-D and Near Infra Red (NIR) light induces ex-vivo and in-vivo corneal stiffening using low energy and short irradiation timeJurriaan Brekelmans1, 2, Alexandra Goz2, 3, Alexander Brandis4, Mor Dickman1, Rudy M. Nuijts1, Avigdor Scherz2, Arie Marcovich2, 3. 1Ophthalmology, University Eye Clinic Maastricht, Maastricht, Netherlands; 2Plant and Environmental Sciences, Weizmann Institute of Science, Rehovot, Israel; 3Ophthalmology, Kaplan Medical Center, Rehovot, Israel; 4Biological Services, Weizmann Institute of Science, Rehovot, Israel.Purpose: Controversy exists regarding the efficacy of shortened corneal cross-linking (CXL) protocols. We previously described the safety and efficacy of WST11-D/NIR in corneal stiffening using parameters similar to the Dresden protocol for Riboflavin/UVA CXL.

This ex-vivo and in-vivo laboratory study sets out to determine the stiffening effect of shortened WST11-D/NIR protocols with 1, 5 and 30 minutes of NIR irradiation in a rabbit model.Methods: Corneas of 3 months old, male, New Zealand White rabbits were treated in-vivo (n=24) and ex-vivo (n=54) using three irradiation times and biomechanically tested immediately (ex-vivo) or one month after treatment (in-vivo). One cornea of each pair was mechanically de-epithelialized and topically impregnated with 2.5 mg/ml WST11 (Steba-Biotech, France) combined with dextran-500 (WST11-D) for 20 minutes. Next, irradiation with NIR light at 10mW/cm2 was applied for either 30 minutes (nin=3, nex=11), 5 minutes (nin=6, nex=5) or 1 minute (nin=3, nex=11) using a diode laser at 755nm (CeramOptics, Israel). Untreated fellow eyes served as controls. Corneal strips, 4±0.2mm in width, underwent biomechanical stress-strain measurements using a biomaterial tester (Minimat, Rheometric Scientific GmbH, Germany) to obtain the tangent modulus. Results were compared by paired Student-T tests and linear mixed modelling using SPSS (IBM Corp., Armonk, NY).Results: Ex-vivo treatment resulted in a similar significant increase (>100%) in Young’s modulus (p<0.001) for all three irradiation protocols, figure 1. One month after in-vivo treatment Young’s modulus increased significantly after 30 and 5 minutes of irradiation (125% and 35% respectively, p<0.05) but did not increase significantly after 1 minute of irradiation, figure 2.Conclusions: Shortening NIR irradiation time to 5 minutes without increasing total delivered energy results in significant corneal stiffening both ex-vivo and in-vivo. Differences between ex-vivo and in-vivo stiffening may result from living tissue behavior, making 1 minute irradiation in-vivo not sufficient. This novel ultra-fast treatment with reduced irradiation time and overall irradiation energy may provide a safe alternative to Riboflavin/UVA CXL for the treatment of corneal ectasia.

Fig. 1 Young’s Modulus, ex-vivo experiments.



Fig. 2 Young’s Modulus, in-vivo experiments.Commercial Relationships: Jurriaan Brekelmans, None; Alexandra Goz, None; Alexander Brandis, Int. publication Number: WO 2013/027222 Al (P); Mor Dickman, None; Rudy M. Nuijts, None; Avigdor Scherz, Int. publication Number: WO 2013/027222 Al (P); Arie Marcovich, Int. publication Number: WO 2013/027222 Al (P)

Program Number: 2378 Poster Board Number: A0107Presentation Time: 3:45 PM–5:30 PMRemodeling of the Cornea Induced by a Shape-Changing Hydrogel InlayAlan J. Lang1, Arturo Chayet2, Enrique Barragán-Garza3. 1R&D, ReVision Optics, Lake Forest, CA; 2Codet Vision Institute, Tijuana, Mexico; 3Laser Ocular Hidalgo, Monterrey, Mexico.Purpose: Assess epithelial and stromal remodeling anterior to a hydrogel inlay placed within the cornea.Methods: The analysis was performed on subjects (N = 30) enrolled in a prospective clinical investigation of the Raindrop® Near Vision Inlay, implanted in corneal of the non-dominate eye of emmetropic presbyopic subjects. The inlay had a meniscus shape with a 2 mm diameter and a 34 micron central thickness. The inlay was placed beneath a targeted 150 microns corneal flap (femtosecond laser), over the center of the light constricted pupil. The change to the anterior corneal surface profile (ΔC) was measured using postoperative minus preoperative wavefront measurements. The change to the epithelial thickness profile (ΔE) was measured using Fourier domain optical coherence tomography (OCT). Changes to the anterior stroma (flap) were reflected in alteration of Bowman’s layer shape (ΔBow) and visible OCT images, but this changes was more accurately calculated from the above two measurements; ΔBow = ΔC - ΔE.Results: When placed on the flap bed, the inlay’s volume displaced stroma anterior to the inlay, with 91% of the inlay volume reflected in the change to Bowman’s layer directly above the inlay’s diameter. The 8% loss is approximately accounted for by a slight increase in stromal volume just outside the inlay diameter. Bowman’s layer rose centrally by 28 ± 8.5 microns, 85% of the inlay’s central thickness.

The stromal rise thinned the overlying epithelium by -18 ± 5.1 microns, equivalent to 35% of the preoperative epithelial thickness. The volume of the epithelium thinned was redistributed peripheral to the inlay diameter (2 mm), extending the change at the anterior corneal surface to about twice the inlay diameter. The central anterior corneal surface rose by 10 ± 3.2 microns, creating a progressive Add power profile with a central peak of -5 D.Conclusions: When a meniscus shaped corneal inlay is placed beneath a corneal flap, epithelial remodeling redistributed most of the inlay’s volume, with a much smaller biomechanical remodeling of the stroma anterior to the inlay.*CAUTION: Investigational device. Limited by Federal (United States) law to investigational use.Commercial Relationships: Alan J. Lang, ReVision Optics, Inc.; Arturo Chayet, ReVision Optics, Inc. (C); Enrique Barragán-Garza, ReVision Optics, Inc. (C)

Program Number: 2379 Poster Board Number: A0108Presentation Time: 3:45 PM–5:30 PMRetroprosthetic Membrane Formation in Boston Keratoprothesis: A Case Control Matched Comparison of Titanium versus PMMA Back PlateJoelle Hallak1, Rushi K. Talati2, Jose De la Cruz1, Maria S. Cortina1. 1Ophthalmology, University of Illinois at Chicago, Chicago, IL; 2Ophthalmology, Northwestern University, Chicago, IL.Purpose: Retroprosthetic membrane (RPM) formation is the most common complication following Boston keratoprosthesis (Kpro) implantation. While design changes have been made to improve tissue compatibility, little evidence exists comparing the frequency of RPM formation in various Kpro models. This retrospective, case-control matched study compared the frequency of RPM formation in patients implanted with a titanium back plate Kpro versus patients implanted with a polymethyl methacrylate (PMMA) back plate Kpro.Methods: Twenty-one eyes from 21 consecutive patients implanted with a titanium back plate Kpro were case-matched with 21 eyes from 21 patients implanted with a PMMA backplate Kpro by a single surgeon at a large academic center. Case-matching was based on preoperative diagnosis, type of surgery (e.g. aphakic vs pseudophakic, Kpro only vs Kpro combined with vitrectomy and/or glaucoma shunt) and age. Formation of visually significant RPM requiring YAG laser treatment or surgical membranectomy was the primary outcome measure. Initial visual acuity (VA), best corrected visual acuity (BCVA), and VA at follow-up were secondary outcome measures. Paired and unpaired t-tests were performed for statistical significance.Results: Ten patients implanted with a titanium Kpro developed visually significant RPM compared to 7 patients implanted with a PMMA Kpro (p=0.38). Time from surgery to YAG treatment was not significantly different between the titanium and PMMA groups (227.10 ± 44.58 days vs 275.86 ± 75.50 days). Initial VA prior to surgical intervention was better for titanium Kpro patients (14.3% vs 0% had VA of 20/200 or better; p=0.03 for mean VA difference). Following surgery, patients implanted with a titanium Kpro achieved BCVA earlier (63.19 ± 16.44 days vs 173.90 ± 34.16 days; p=0.01), however the BCVA was not found to be statistically different (p=0.19). VA at most recent follow-up was not significantly different between the two groups (p=0.92).Conclusions: Titanium Kpro back plates do not seem to significantly reduce RPM formation compared to PMMA Kpro back plates when patients are controlled for preoperative diagnosis. However, a larger prospective study would be necessary to confirm these findings. While differences in visual outcomes exist between the two Kpro



models in the short-term, they are temporal and insignificant in the long-term.Commercial Relationships: Joelle Hallak; Rushi K. Talati, None; Jose De la Cruz, Alcon (C); Maria S. Cortina, NoneSupport: Research to Prevent Blindness Departmental Support

Program Number: 2380 Poster Board Number: A0109Presentation Time: 3:45 PM–5:30 PMThe Resistance to Collagenase Dissolution of Cross-Linked Cryopreserved Donor CorneasDarrell R. Lewis, Mark Seamone, Christopher Seamone. Ophthalmology, Dalhousie University, Halifax, NS, Canada.Purpose: To determine if UV-A riboflavin cross-linking (CXL) of donor corneas after cryopreservation confers resistance to keratolysis as has been described for fresh donor corneas.Methods: Three methods of donor cornea cryopreservation were compared to donor corneas stored in Optisol-GS at 4°C for less than 2 weeks (control). The cryopreservation arms consisted of corneas stored in Optisol-GS at -15°C, Optisol-GS at -70°C, and ≥99% anhydrous glycerol at -80°C. Within each cryopreservation method, corneas were randomized to either receive CXL or not. The primary outcome measure was the time to complete corneal dissolution in a collagenase-A solution.Donor corneas cryopreserved in Optisol-GS at -15°C (n=4), Optisol-GS at -70°C (n=4), and ≥99% glycerol at -80°C (n=4) were thawed for 30 minutes in a room temperature balanced salt solution. Corneas were mounted on artificial anterior chambers and de-epithelialized. Pre-treatment with 0.1% riboflavin in dextran solution every 2 minutes for 15 minutes was undertaken. High fluence, accelerated CXL (9 mW/cm2, 10 min) was then performed. As controls, 4 donor corneas matching each of the cryopreservation methods were thawed but did not undergo CXL. Additionally, 4 fresh donor corneas did not undergo CXL. Corneas were punched with an 8mm Baron trephine. These 8 mm corneal buttons were agitated in a 0.3% collagenase-A solution at 37°C. Corneas were examined every 30 minutes until complete dissolution occurred.Data was analysed via Mann-Whitney U-test and Kruskal-Wallis analysis where appropriate.Results: Donor corneas cryopreserved in ≥99% glycerol at -80°C required a statistically significantly longer time for complete corneal dissolution than fresh donor corneas (p=0.03). Dissolution times for donor corneas cryopreserved in Optisol-GS at either -15°C or -70°C did not statistically significantly differ from fresh donor corneas, p=0.13 and p=0.26 respectively.Conclusions: Cross-linking donor corneas cryopreserved in ≥99% glycerol at -80°C increases their resistance to keratolysis. Donor corneas cryopreserved in Optisol-GS have a similar resistance to keratolysis as fresh donor corneas.The ability to enhance cryopreserved donor corneas via CXL may prove advantageous to reduce graft failure rates. Cryopreserved tissue may be appropriate for deep anterior lamellar keratoplasty, glaucoma tube shunt coverage, Boston keratoprosthesis carriers, and tectonic grafting for melts.Commercial Relationships: Darrell R. Lewis, None; Mark Seamone, None; Christopher Seamone, None

Program Number: 2381 Poster Board Number: A0110Presentation Time: 3:45 PM–5:30 PMBiomechanical properties of the cornea measurement using Scheimpflug noncontact tonometry in aniridia patientsBogumil Wowra, Dariusz Dobrowolski, Dorota Tarnawska, Edward Wylegala. Railway Hospital Katowice. II School of Medicine with the Division of Dentistry in Zabrze Medical University of Silesia, Katowice, Poland.Purpose: To evaluate biomechanical properties of the cornea in patients with congenital aniridia.Methods: 26 patients (16 women, 10 men) with congenital aniridia aged from 6 to 55 underwent corneal examination using Scheimpflug noncontact tonometry. CCT, deformation amplitude, first applanation time, second applanation time, highest concavity and IOP were evaluated.Results: CCT ranged from 496 μm to 718 μm (mean 623 μm). Deformation amplitude range from 0.36 to 1.01 mm (mean 0.62 mm) corelated with IOP result range from 15.6 mmHg to 42.8 mmHg (mean 22.82 mmHg).Conclusions: Evaluation of corneal biomechanical properties is useful in estimation real IOP in aniridia patients.Commercial Relationships: Bogumil Wowra; Dariusz Dobrowolski, None; Dorota Tarnawska, None; Edward Wylegala, None

Program Number: 2382 Poster Board Number: A0111Presentation Time: 3:45 PM–5:30 PMCross-linking biomechanical effect in human corneas by same energy, different UV-A fluence: an enzymatic digestion comparative evaluationGeorge Asimellis1, 2, A. J. Kanellopoulos2, 3. 1Kentucky College of Optometry, PikeVille, KY; 2LaserVision.gr Clinical & Research Eye Institute, Athens, Greece; 3Ophthalmology, NY University, NY, NY.Purpose: Collagenase has been known to contribute to break-down of collagen in the corneal stroma. This breakdown is a vigorous biochemical process used as an indirect metric of corneal biomechanical properties. The stabilizing biochemical effect of corneal cross-linking (CXL) may be thus reflected by an increased resistance to collagenase digestion. Clinical results suggesting comparable effectiveness of Higher-fluence (irradiance) CXL variations to the conventional protocol (3 mW/cm2) in stabilizing keratoconus progression have been reported. The enzymatic degradation resistance modulations achieved via CXL application at different UV-A irradiances has not been studied in human corneas. The purpose of this study is to comparatively evaluate ex-vivo these enzymatic degradation resistance differences.Methods: The study involved twenty-five human donor corneas, randomly allocated to 5 groups (n=5 each). CXL was applied with UV-A irradiance of 3-mW/cm2, 9-mW/cm2, 18-mW/cm2, 30-mW/cm2, and 45-mW/cm2, maintaining equal cumulative energy dose of 5.4 J/cm2. UV-A was delivered on half of the cornea. The non-irradiated halves served as controls. Specimens were subjected to collagenase-A enzymatic digestion. The time to complete dissolution in each specimen was recorded.Results: Time to dissolution in group-A (3-mW/cm2 for 30’) was 321±13.4 minutes (range: 300 to 330) compared to 171±8.2 (range: 165 to 180) for their control. In group-B (9-mW/cm2 for 10’) 282±19.6 minutes (range: 270 to 315) compared to 177±6.7 (165 to 180) for their control. In group-C (18 mW/cm2 for 5’) 267±19.6 minutes (range: 240 to 285) compared to 177±7.7 (range 165 to 180) for their control. In group-D (30-mW/cm2 for 3‘) 252±12.5 minutes (range: 240 to 270) compared to 180±10.6 minutes (range: 165 to 195) for their control. In group-E (45-mW/cm2 for 2‘) 204±17.1



minutes (range: 180 to 225) compared to 186±8.2 minutes (range: 180 to 195) for their control.Conclusions: The data in this ex-vivo human corneal study indicate that the biomechanical effect of CXL studied by resistance to enzymatic digestion in human corneas is comparable between irradiance of 9, 18 and 30-mW/cm2, and appears to be reduced at fluence of 45mW/cm2.

Design and execution of the patterned half-area illumination of the treated corneas.Commercial Relationships: George Asimellis, None; A. J. Kanellopoulos, i-Optics (C), ISP Surgical (C), Keramed (C), Avedro (C), Allergan (C), Alcon (C)

Program Number: 2383 Poster Board Number: A0112Presentation Time: 3:45 PM–5:30 PMPenetrating keratoplasty versus deep anterior lamellar keratoplasty: which technique establishes healthy corneal biomechanics?HANS R. VELLARA, Noor Q. Ali, Akilesh Gokul, Jason Turuwhenua, Charles N. McGhee, Dipika Patel. Ophthalmology, University of Auckland, Auckland, New Zealand.Purpose: The cornea is biomechanically compromised in keratoconus. Ideal management options would restore the corneal biomechanics to that of a healthy cornea. This study analysed the corneal biomechanical properties of keratoconic corneas which had undergone either a penetrating keratoplasty (PKP) or a deep anterior lamellar keratoplasty (DALK) and compared them to those of healthy control corneasMethods: All eyes were examined by slit-lamp biomicroscopy, Pentacam tomography, and the CorVis ST (CST). The CST is a non-contact tonometer coupled with an ultra-high speed Scheimpflug camera which provides descriptors of ocular biomechanics. Furthermore, specialised software was designed in house to remove the collateral globe displacement to the air-pulse to isolate the maximum corneal deformation (MCD). A multiple regression analysis was used to compare CST outputs and additionally derived parameters between groups after controlling for patient age, intra-ocular pressure (IOP), central corneal thickness (CCT), maximum simulated keratometry (KMAX), suture number, and time from surgery.Results: This study included 42 PKP, 27 DALK, and a 152 healthy control eyes. There were no significant differences in patient age, IOP, time from surgery, KMAX, donor size, and donor age, between treated groups (p>0.05). CCT was significantly higher in DALK-

treated corneas (590 ± 52µm, p<0.001) (compared to PKP-treated (542 ± 50µm), and healthy control corneas (544 ± 34µm)). Furthermore, there were no significant differences in patient age, IOP, and KMAX between treated and control groups (p>0.05). However, multiple regression analysis revealed a significantly higher MCD in PKP-treated corneas (0.86 ± 0.02mm) than in DALK (0.80 ± 0.02 mm, p=0.03) and healthy controls (0.79 ± 0.01mm, p=0.02). Additionally, the PKP-treated corneas had a greater number of parameters significantly different to control corneas than DALK-treated corneas.Conclusions: DALK-treated corneas were biomechanically more similar to healthy controls than PKP-treated corneas. This might be due to the scar tissue interface in DALK-treated corneas redirecting the force applied to the peripheral cornea and the scar interface in PKP-treated corneas resulting in an incomplete transfer of energy to the peripheral cornea.Commercial Relationships: HANS R. VELLARA, None; Noor Q. Ali, None; Akilesh Gokul, None; Jason Turuwhenua, None; Charles N. McGhee, None; Dipika Patel, None

Program Number: 2384 Poster Board Number: A0113Presentation Time: 3:45 PM–5:30 PMThe Mechanics of Corneal Deformation and Rupture for Penetrating Injury in the Human EyeAndrew Rau3, Scott Lovald3, Steven Nissman1, John McNulty2, Jorge Ochoa3, Michael Baldwinson4. 1University of Pennsylvania Perelman School of Medicine, Scheie Eye Institute, Philadelphia, PA; 2Materials and Corrosion Engineering, Exponent, Inc., Menlo Park, CA; 3Biomedical Engineering Practice, Exponent, Inc., Philadelphia, PA; 4Google[X], Mountain View, CA.Purpose: A penetrating eye injury is a surgical emergency with a guarded visual prognosis. The purpose of this study was to determine the force required to rupture the cornea with a penetrating object, and to study how this force is affected by the object geometry.Methods: Thirty-six human cadaveric eye specimens were used for the study. Spherical indenters of three different diameters (1.0, 1.5, and 2.0 mm) were pressed into the apex of the cornea at two displacement rates (1.0 mm/s and 5.0 mm/s) until rupture of the specimen occurred. Low strain stiffness (between 3-5 mm indenter displacement), high strain stiffness (between 80%-90% of the maximum failure force), indenter displacement at failure, and the force at failure were determined from the test data and used to characterize the mechanical tissue response. Multi-variable regression analysis was performed on the output parameters in order to determine associations of the input variables (indenter size, test speed, tissue postmortem time, and specimen geometry) on the mechanics of the tissue response.Results: Twenty-nine of the 36 specimens failed at the location where the indenter contacted the cornea, four specimens failed at the limbus, and three specimens failed in the sclera near sites of muscle attachment. The average force at failure caused by the 1.0 mm, 1.5 mm, and 2.0 mm diameter indenters increased from 30.5±5.5 N to 40.5±8.3 N to 58.2±14.5 N, respectively (p<0.002) (Table 1, Figure 1). The force at failure was also determined to be associated with the donor age (p<0.001), and globe diameter (p<0.041), but was not associated with intraocular pressure, tissue postmortem time, axial length, or indenter speed.Conclusions: This study has quantified the force-displacement response of a large series of human cadaveric eyes subjected to penetrating indentation loads on the cornea. The results provide useful data for characterizing the relationship between corneal rupture and the geometry of a penetrating object.



The outcome measures for all test groups. Note: results presented as Avg. ± St. Dev.

Force (N) vs. displacement (mm) curves for the 1.0 mm indenter tests (specimens 25-36). Specimens that failed at the contact location in the cornea are indicated with a circle at the failure point, while specimens that failed in other regions of the eye are indicated with an X at the failure point.Commercial Relationships: Andrew Rau, Google[X] (F); Scott Lovald, Google[X] (F); Steven Nissman, None; John McNulty, Google[X] (F); Jorge Ochoa, Google[X] (F); Michael Baldwinson, Google[X] Support: Partial support for this study was provided by Google[X]

Program Number: 2385 Poster Board Number: A0114Presentation Time: 3:45 PM–5:30 PMComparison between the diagnostic ability of corneal Tomography and novel biomechanical parameters in keratoconusVishal Jhanji1, Tommy Chan1, Marco Yu2. 1Ophthalmology, Chinese Univ of Hong Kong, Kowloon, Hong Kong; 2Hang Seng Management College, Kowloon, Hong Kong.Purpose: Early and accurate diagnosis of keratoconus is of paramount importance for effective management of these cases. Corneal tomography as well as biomechanical evaluation are being increasingly employed for this purpose. This study compared the diagnostic ability of tomographic and novel biomechanical parameters in cases with keratoconusMethods: Forty-two eyes of 42 participants were included (21 normal, 21 keratoconus). Corneal tomographic parameters were measured using Scheimpflug imaging (Pentacam, Oculus, Wetzlar, Germany). Biomechanical parameters were obtained from a Scheimpflug camera (Corvis, Oculus, Wetzlar, Germany). The area under receiver operating curve (AUC) and partial AUC for specificity ≥ 80% for each parameter was calculated to assess the

discrimination ability and compared between devices. Correlation analysis was performed between parameters obtained from both devices. A p-value < 0.05 was considered statistically significant. In addition, deformation amplitude (DA) ratio 1 and 2, a ratio between deformation amplitude at apex and at 1mm and 2 mm from the centre of the corneal apex respectively, was proposed and validated on Corvis.Results: Corneal parameters were selected for comparison based on their diagnostic performance. Belin/Ambrosio Enhanced Ectasia Display (BAD) showed the highest AUC and partial AUC followed by Ambrosio’s relational thickness maximum (ARTmax) with AUC ≥ 0.95 and partial AUC ≥ 0.18. DA ratio 1 and 2 also demonstrated the highest AUC and partial AUC amongst other Corvis parameters. There was no significant difference for AUC and partial AUC of BAD and ARTmax compared to those of DA Ratio 1 and 2 (p>0.186). Significant correlation was found between BAD and DA Ratio 1 and 2 (p<0.003) as well as ARTmax and DA Ratio 1 and 2 (p<0.002) for eyes with keratoconus.Conclusions: Novel biomechanical parameters were able to provide discriminative ability between normal and keratoconic eyes comparable to tomographic analysis using Scheimpflug imaging. Corneal biomechanics may have the potential to effectively diagnose keratoconus.Commercial Relationships: Vishal Jhanji, None; Tommy Chan, None; Marco Yu, None

Program Number: 2386 Poster Board Number: A0115Presentation Time: 3:45 PM–5:30 PMPhysical properties of electron-beam irradiated corneas stored in recombinant human serum albuminKhoa D. Tran1, Mark A. Terry2, Christopher Stoeger1. 1Lions VisionGift, Portland, OR; 2Devers Eye Institute, Portland, OR.Purpose: While electron-beam (e-beam) irradiated cornea tissue have been successfully used in ophthalmic procedures, structural changes within the tissue as a result of e-beam irradiation and the resulting functional outcomes have not been described in detail. Here, we report preliminary studies comparing the physical properties of fresh versus e-beam treated corneas.Methods: The clarity, structural properties, and surgical handling of e-beam irradiated corneas were examined and compared to fresh donor corneas. Dark field microscopy and computer-aided analysis using Matlab were performed to determine corneal clarity. Differential scanning calorimetry and tissue histology were used to examine the corneal collagen matrix organization. The rigidity (elastic modulus) of the cornea were determined using Brillouin optical microscopy. Tissue and suture handling evaluations were performed by two different fellowship-trained cornea surgeons who were masked to tissue treatment conditions prior to handling.Results: The clarity of fresh and e-beam irradiated corneas were 92.4 ± 3.5% and 89.7 ± 2.7%, respectively (N=18, P=0.009). Differential scanning calorimetry revealed that the denaturing temperature of e-beam irradiated tissue is approximately 2°C lower than fresh corneas (N=5), although histological sections comparing fresh and e-beam treated tissue revealed minimal changes in the collagen matrix (N=8). Brillouin optical microscopy and analysis suggests that irradiated corneas are slightly more rigid than fresh corneas, although this change did not affect the tissue’s ability to handle external forces as no statistically significant differences were detected in the shear moduli of the two groups (7.3 kPa vs. 6.2 kPa, N=6). Likewise, cornea surgeons evaluating e-beam treated tissue did not note major differences in the tissue’s elasticity or ability to handle sutures.Conclusions: Our preliminary results suggest that e-beam treated corneas possess similar properties to fresh donor corneas. Although



there is a 2.7% decrease in clarity and slight changes to the structure of the corneal tissue after irradiation, these changes do not appear to have adverse effects on their clinical and functional properties. Additional studies are required to strengthen our current understanding of e-beam irradiation on corneal tissue, including examination of e-beam treated tissues after prolonged storage (>1 year).Commercial Relationships: Khoa D. Tran, None; Mark A. Terry, None; Christopher Stoeger, None

Program Number: 2387 Poster Board Number: A0116Presentation Time: 3:45 PM–5:30 PMHierarchical changes in corneal collagen structure under loadJames S. Bell1, Sally Hayes1, Charles Whitford2, Ahmed Elsheikh2, Keith M. Meek1. 1Cardiff University, Cardiff, United Kingdom; 2University of Liverpool, Liverpool, United Kingdom.Purpose: Current understanding of corneal biomechanics is based upon hierarchical maps of collagen distribution, but these maps are obtained from corneas at rest. To quantify the mechanical and microstructural response to load, we carried out mechanical tests of corneal samples in an X-ray beam.Methods: 14 human corneas with scleral rims were obtained from Bristol Eye Bank and transported in organ culture medium to Diamond Light Source, where they were deswelled in 15% dextran solution for one day. Full thickness super-inferior tensile strips 3mm in width were sectioned from the corneas and attached to a custom-built extensometer that could be mounted on the stages of the I02 and I22 beamlines.X-ray scatter images were obtained in the wide angle (WAXS – I02) and small angle (SAXS – I22) configurations, allowing changes in the collagen structure and fibrillar architecture to be measured at length scales ranging from angstroms (e.g. the helical period of tropocollagen molecules) up to tens of nanometres (e.g. the diameter of collagen fibrils). Variations across the cornea were assessed by scanning the X-ray beam across the sample.Results: X-ray scattering images were taken following static strains of 1.4%, 2.8%, 5% and 8%. The direction of preferential molecular and fibrillar orientation shifted from naso-temporal to superior-inferior following the 1.4% uniaxial strain application and the degree of anisotropy increased with strain. The largest change in molecular architecture took place between 2.8% and 5% uniaxial strain increments, after which the total collagen aligned in the superior-inferior direction was 24±10% greater than in other directions and the ratio of collagen aligned in the superior-inferior direction to that in the naso-temporal direction changed from 0.86±0.04 at rest to 1.90±0.40. Beyond 5% uniaxial strain the rate of reorientation diminished.Conclusions: Quantifying changes in the structure and fibrillar architecture of the cornea under load provides a means to relate findings from mechanical tests with changes in microstructure. We find here that in healthy human corneas there is a reorientation “cut-off” at around 5% tensile strain, after which resistance to load is mediated by other means, such as fibril slipping. The ability of the corneal collagen network to reorientate informs us about binding and interconnectivity, and the ability of the cornea to adapt in what is a variable mechanical environment.Commercial Relationships: James S. Bell, None; Sally Hayes, None; Charles Whitford, None; Ahmed Elsheikh, None; Keith M. Meek, NoneSupport: MRC Programme Grant 503626

Program Number: 2388 Poster Board Number: A0117Presentation Time: 3:45 PM–5:30 PM3D Inflation Strains in Porcine CorneasKeyton Clayson1, 2, Elias Pavlatos1, Jun Liu1. 1Biomedical Engineering, The Ohio State University, Columbus, OH; 2Biophysics, The Ohio State University, Columbus, OH.Purpose: To measure the 3D deformation of the cornea using high-resolution ultrasound speckle tracking, and to evaluate the presence of shear during inflations of the cornea.Methods: Seven porcine globes were tested within 72 hours post-mortem. Whole globes were immersed in a 10% dextran solution for 1 hour to reduce swelling, then secured to a custom-built holder and immersed in 0.9% saline. The anterior chamber was infused with Optisol GS (Bausch and Lomb) via a column system to control intraocular pressure (IOP). The globes were preconditioned with 25 pressure cycles from 10 to 12 mmHg, then equilibrated at 10 mmHg for 1 hour. Testing involved raising the pressure to 11 mmHg and 12mmHg consecutively, with a 15 minute equilibration time at each pressure. For each step, a 55 MHz ultrasound probe with an imaging window of 5.5 mm was aligned along the nasal-temporal direction of the cornea, and consecutive 2D frames of radiofrequency data were acquired at 14 µm steps scanned over 2 mm in the superior-inferior direction to form a volume centered on the corneal apex. The 3D principal strains (ε1, ε2, ε3), max shear, and volume ratio were calculated using an ultrasound speckle tracking technique (Cruz Perez et al, ABME, 2015). Volumetric strain maps and vector plots were visualized in Paraview (Kitware Inc).Results: The average principal strains, max shear, and volume ratio in porcine corneas are summarized in Table 1. Representative maps of the principal strains and principal vector directions are shown in Figure 1. During corneal inflation, the principal strain of largest magnitude was through-thickness compression (ε3), while the other two principals (ε1, ε2) were tensile and largely in-plane. Significant shear with a magnitude larger than all principals was observed. The volume ratios close to one confirmed minimum swelling and near-incompressibility during inflation.Conclusions: 3D ultrasound speckle tracking showed that the cornea experiences compressive, tensile, and shear strains during IOP increase. The predominance of shear deformation during pressure increase suggests that weakness in resisting shear may be involved in corneal ectasia and warrants future investigations.

Table 1: Mean ± SD of principal strains and maximum shear during inflation (n=7).

Figure 1: Maps of principal vector directions (top) and principal strains (bottom) in a representative cornea (A-P: anterior-posterior, S-I: superior-inferior, N-T: nasal-temporal).Commercial Relationships: Keyton Clayson, None; Elias Pavlatos; Jun Liu, NoneSupport: NIH RO1EY020929



Program Number: 2389 Poster Board Number: A0118Presentation Time: 3:45 PM–5:30 PMA Multicenter Study of Ocular Response Analyzer Custom Variables in Disqualified and Candidate Refractive Surgery Screening PatientsVinicius S. De Stefano1, 2, Ibrahim Seven1, James B. Randleman3, William J. Dupps1. 1Cole Eye Institute, Cleveland Clinic, Cleveland, OH; 2Ophthalmology, Federal University of Sao Paulo, Sao Paulo, Brazil; 3Emory Eye Center, Emory University, Atlanta, GA.Purpose: To assess the ability of air-puff derived measures of biomechanical behavior to predict clinical decisions on refractive surgery candidacy.Methods: Data were retrospectively collected from consecutive refractive surgery screening exams by two surgeons from different academic centers (Cleveland Clinic Cole Eye Institute (CC) and Emory Eye Center (EEC)). Only right eyes were analyzed. The disqualified subject groups consisted of 19 (CC) and 28 (EEC) eyes and included subjects who were disqualified as candidates based on clinical impression after reviewing available clinical, topo/tomographic (Placido and Scheimpflug), and biomechanical data (Ocular Response Analyzer (ORA)). Control groups consisted of eyes deemed clinically to be safe candidates for LASIK (n=26 (CC) and 23 (EEC)). Keratometric and pachymetric data from the Pentacam were compared for each group. While standard ORA variables were available to surgeons at the time of the decision, the 3 investigator-derived variables (Table 1, previously described by Hallahan et al, Ophthalmol 2014) were calculated retrospectively and surgeons were effectively masked to them. Comparisons were performed using t-tests for independent samples.Results: Anterior curvature and corneal thickness data were significantly different between groups (Table 1). Hysteresis loop area (HLA) was the only biomechanical variable that was significantly different between disqualified and control subjects across both centers (Table 2).Conclusions: Across two clinical sites, a custom biomechanical variable, HLA, was predictive of which refractive surgery screening patients were selected for LASIK and which were disqualified. Certain measures of the corneal dynamic response to an air puff may serve as objective correlates to clinically perceived risk of ectasia.

Table 1. Disqualified patients group and control group baseline and pre-operative corneal features.

Table 2. Pre-operative variables derived from the ORA for disqualified and control (candidate) patient groups.Commercial Relationships: Vinicius S. De Stefano, None; Ibrahim Seven, None; James B. Randleman, None; William J. DuppsSupport: CAPES Grant PDSE - 99999.007333/2015-03; R01 EY023381; Ohio Third Frontier Innovation Platform Award TECH 13-059; Unrestricted Grant from RPB to the Dept. of Ophthalmology of the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University; Unrestricted Grant from RPB to the Dept. of Ophthalmology of the Emory University

Program Number: 2390 Poster Board Number: A0119Presentation Time: 3:45 PM–5:30 PMA New Stiffness Parameter in Air Puff Induced Corneal Deformation AnalysisCynthia J. Roberts1, Ashraf M. Mahmoud1, Jeffrey Bons4, Arif Hossain4, Ahmed Elsheikh5, Riccardo Vinciguerra6, Paolo Vinciguerra7, 8, Renato Ambrosio2, 3. 1Ophthalmology & VIsual Science; and Biomedical Engineering, The Ohio State University, Columbus, OH; 2Rio de Janeiro Corneal Tomography and Biomechanics Study Group, Rio de Janeiro, Brazil; 3Ophthalmology, Federal University of São Paulo, Rio de Janeiro, Brazil; 4Mechanical & Aerospace Engineering, The Ohio State University, Columbus, OH; 5School of Engineering, University of Liverpool, Liverpool, United Kingdom; 6Division of Ophthalmology, University of Insubria, Varese, Italy; 7Eye Center, Humanitas Clinical and Research Center, Rozzano, Italy; 8Vincieye Clinic, Milano, Italy.Purpose: To investigate a new stiffness parameter in corneal deformation analysis and compare responses in normal (NL) and keratoconic (KC) subjects, matched for intraocular pressure (IOP).Methods: A new stiffness parameter (SP) is defined as the resultant pressure at inward applanation, divided by corneal deflection amplitude at highest concavity (HCDeflectAmp). The spatial and temporal profiles of the Corvis ST air puff (Oculus, Wetzlar, Germany) were characterized using hot wire anemometry from 0 to 16mm from the nozzle. Measured velocity was correlated with the pressure profile exported by the Corvis ST, measured within the nozzle. The z position of the cornea at the time of inward applanation was used to calculate an adjusted air pressure value (adjAP1) at the time and position of first applanation. An algorithm to correct IOP estimation based on finite element modeling, termed IOPfem, was used for the equation: SP = (adjAP1 - IOPfem)/ HCDeflectAmp. Linear regression analyses between corneal deformation parameters



(CDP’s) and SP were performed on a retrospective dataset of 258 KC eyes and 460 NL eyes. CDP’s from a subset of 119 eyes of 119 subjects in each group were matched for IOPfem and compared using t-tests. Significance threshold was p < 0.05.Results: All CDP’s evaluated showed a significant difference between NL and KC, except peak distance. (Table 1) The KC group had lower SP values, thinner pachymetry, shorter applanation lengths, greater absolute values of applanation velocities, earlier first applanation times and later second applanation times, greater HC deformation and HC deflection amplitudes, and lower HC radius of concave curvature (greater concave curvature). All CDP’s evaluated showed a significant relationship with SP in both groups. Stiffer eyes were associated with greater pachymetry, longer applanation lengths, lower absolute value of applanation velocities, later first applanation times, earlier second applanation times, lower HC deformation and HC deflection amplitudes, shorter peak distances, greater HC radius of concave curvatures (flatter), and higher values of IOPfem.Conclusions: Keratoconic eyes demonstrated less resistance to deformation than normal eyes with similar IOP. All of the deformation parameters investigated showed a significant relationship with the new stiffness parameter. This may be useful in future biomechanical studies comparing populations without having to match IOP.

Commercial Relationships: Cynthia J. Roberts, Oculus (C), Oculus (F); Ashraf M. Mahmoud, None; Jeffrey Bons, None; Arif Hossain, None; Ahmed Elsheikh, Oculus (F); Riccardo Vinciguerra, None; paolo vinciguerra; Renato Ambrosio, Oculus (C)Support: OCULUS Optikgeräte GmbH

Program Number: 2391 Poster Board Number: A0120Presentation Time: 3:45 PM–5:30 PMAssessment of biomechanical properties of porcine corneas with air-puff swept source OCTEwa C. Maczynska1, Karol M. Karnowski1, Bartlomiej Kaluzny2, Ireneusz Grulkowski1, Maciej D. Wojtkowski1. 1Faculty of Physics, Astronomy and Informatics, Nicolaus Copernicus University, Torun, Poland; 2Collegium Medicum, Nicolaus Copernicus University, Bydgoszcz, Poland.Purpose: A novel approach to assess Intraocular Pressure (IOP) and/or specific biomechanical properties of the cornea were proposed. Biomechanical properties of the porcine corneas ex vivo at well-controlled levels of the intraocular pressure, and the impact of cross-linking (CXL) on corneal biomechanics were studied.Methods: Thirty fresh, enucleated porcine eyes were measured with high-speed swept source Optical Coherence Tomography (SS-OCT) instrument combined with the air-puff chamber for non-contact generation of corneal deformation. Data acquired from

corneal apex during deformation-recovery process together with the reference pressure value enabled generation of corneal hysteresis representing viscoelastic properties of the cornea (Figs.1A-B). The corneal response to the air pulse was determined for ascending (5 to 35 mmHg) and descending (from 35 to 5 mmHg) IOP levels. Later on, the corneal response to the air pulse before and after CXL procedure under constant IOP condition was measured. In all cases, the parameters of recorded hysteresis loops (applied pressure vs. deformation amplitude) were analyzed.Results: Air-puff SS-OCT measurements enabled to calculate corneal hysteresis loop parameters such as the hysteresis curve area or maximum deformation amplitude. The corneal hysteresis area became lower with ascending IOP value (Fig. 1B). Additionally, viscoelastic behavior of ex vivo porcine corneas was manifested as a difference in the hysteresis area measured for ascending and descending IOP values (Fig. 1C). There was a statistically significant correlation between hysteresis area and maximum deformation amplitude (R=0.99, p=0.001) (Fig. 1D). The central corneal thickness drops immediately after CXL procedure (p=0.05) (Fig. 2A). CXL made the corneas more stiff resulting in lower hysteresis area values (Figs. 2B-C).Conclusions: OCT combined with the air-puff gives additional information about IOP and biomechanical properties of pig eye ex vivo. Further development of this method has potential of being used as a tool for minimally invasive assessment of IOP and central biomechanical properties in healthy and diseased eyes in humans.



Commercial Relationships: Ewa C. Maczynska, None; Karol M. Karnowski, None; Bartlomiej Kaluzny, None; Ireneusz Grulkowski, None; Maciej D. WojtkowskiSupport: Foundation for Polish Science (TEAM; #TEAM/2011-8/8), Polish Ministry of Science and Higher Education (IUVENTUS PLUS; #IP2014 014073)

Program Number: 2392 Poster Board Number: A0121Presentation Time: 3:45 PM–5:30 PMTwo-photon fluorescence microscopy (TFM) and UVA transmission of corneas with HPMC and dextran solutions prior to corneal crosslinking (CXL)Theo G Seiler1, 2, Tobias Ehmke3, Isaak Fischinger1, 2, Alexander Heisterkamp4, Beatrice E. Frueh1. 1Universitätsklinik für Augenheilkunde, Inselspital Bern, Bern, Switzerland; 2IROC AG, Zuerich, Switzerland; 3Laserzentrum Hannover e.V., Hanover, Germany; 4Institut für Quantenoptik, Universität Hannover, Hanover, Germany.Purpose: To evaluate the corneal stroma riboflavin gradient and its implications to safety prior to CXL in commercially available dextran/HPMC solutions at different time points.Methods: Four different groups of porcine corneas, 5 each, were compared regarding the riboflavin concentration in the anterior stroma. Before all experiments, stable corneal hydration conditions were established for the corresponding solution. Two groups (A and B) were treated with 0.1% riboflavin in 20% dextran and 2 groups with 0.1% riboflavin in 1.1% HPMC in 1.1% NaCl (C and D). After imbibition, two-photon microscopy was used to determine fluorescence intensity after 10 minutes (group A and C) and 30 minutes (group B and D). For signal attenuation and concentration determination corneas were saturated and measured a second time by TFM. Additionally, the distribution was simulated mathematically and compared to the experimental results. UVA transmission was determined in all 4 groups after the imbibition.Results: Riboflavin concentration is decreasing with depth for all solutions. Analysis of the fit functions leads to diffusion coefficients of D = 7.67x10-7 for the HPMC solution for 10 minutes and D= 1.31x10-7cm2/s for the dextran solution for 30 minutes. The riboflavin distribution after 10 minutes of imbibition is similar to the one after 30 minutes of 20% dextran (Fig.1). Transmission values of the HPMC solutions were 0.16±0.01% after 10 and 0.05±0.01 after 30 minutes, respectively 0.26±0.02% and 0.19±0.01% in the dextran group. Physiologically hydrated corneas imbibed with 20% dextran after 30 minutes lead to a transmission of 0.16±0.01%.Conclusions: The riboflavin distribution after 10 minutes of imbibition of HPMC is comparable to the distribution after 30

minutes of dextran and, therefore, preferable in clinical use. Lower viscosity of imbibition solution as well as higher hydration of corneal stroma could be identified as factors for accelerating the diffusion of riboflavin. With correct settings two-photon microscopy is a precise way to determine the concentration distribution of riboflavin in corneal stroma.

Commercial Relationships: Theo G Seiler; Tobias Ehmke, None; Isaak Fischinger, None; Alexander Heisterkamp, None; Beatrice E. Frueh, None

Program Number: 2393 Poster Board Number: A0122Presentation Time: 3:45 PM–5:30 PMMultivariate analyses of in vivo anisotropic, hyperelastic biomechanical properties of the human cornea with age, IOP and corneal thicknessAbhijit Sinha Roy, Mathew Francis, Sushma Tejwani, Jyoti Matalia, Himanshu Matalia, Rohit Shetty. Narayana Nethralaya, Bangalore, India.Purpose: We performed multivariate analyses of anisotropic, hyperelastic biomechanical properties of the human cornea using in vivo air-puff applanation and inverse finite element technique.Methods: 50 eyes of 25 normal subjects, ranging in age from 11-50 years, were imaged with Corvis-ST and Pentacam (OCULUS Optikgerate Gmbh, Germany). Inclusion criteria were spherical equivalent less than -3D, no prior ocular surgery, no inflammation and no corneal degenerations. Cornea deformation from Corvis-ST and 3-D geometry from Pentacam was used to construct patient-specific inverse finite element models based on a recent study (Sinha Roy et al., J Mech Behav Biomed Mat 2015; 48:173-82). The recent study quantified the properties using a fiber and depth dependent hyperelastic model with 4 material constants (C10, C20, k1 and k2). C10 and C20 quantified the elastic response of the cellular matrix. k1 and k2 quantified the non-linear stiffening effect of the collagen fibers, arranged in a lamellar network with depth-dependent crosslinks. Multivariate analyses of covariance of material constants was performed with age, intraocular pressure (IOP) and central corneal thickness (CCT) as covariates. Association between the eyes of a subject was also studied in the multivariate model. Mean±standard error of the mean is reported here.Results: Mean age, IOP and CCT was 22.4±1.26 years, 534.6±3.8 μm and 15.5±0.4 mmHg, respectively. Mean C10, C20, k1 and k2 was 87.41±6.3 kPa, 5.1±0.25 MPa, 35±0.6 kPa and 154.4±23.5, respectively. C10 (p=0.62), C20 (p=0.42) and k1 (p=0.20) were similar for the left and right eye of the same subject. However, k2



of the left eye was significantly higher than the same of the right eye (p=0.01). k2 was also positively correlated with age (p=0.02) indicating a signficant age related stiffening effect in the human cornea. C10 was positively correlated with CCT (p=0.01). The other constants did not correlate with either age, IOP or CCT (p>0.05).Conclusions: We studied the in vivo anisotropic properties of the human cornea as a function of IOP, CCT and age for the first time. Increase in k2 with age indicated some age related stiffening effect in the human cornea. CCT may also influence the biomechanical property of the cellular matrix (C10). Further study with a larger sample size and older subjects is needed to confirm the trends.Commercial Relationships: Abhijit Sinha Roy, Carl Zeiss (F), Carl Zeiss (R), Bioptigen (F), Avedro (F), Topcon (F), Cleveland Clinic Innovations (P), Narayana Nethralaya Foundation (P); Mathew Francis; Sushma Tejwani, None; Jyoti Matalia, None; himanshu Matalia, None; rohit shetty, Carl Zeiss (F), Narayana Nethralaya (P), Allergan (F)

Program Number: 2394 Poster Board Number: A0123Presentation Time: 3:45 PM–5:30 PMEffect of cataract surgery on the biomechanical property of the corneaYoshiaki Kiuchi1, Yoshitaka Nakao1, Noriko Ihara1, Shunsuke Nakakura2, Ryo Asaoka3, Makoto Aihara3. 1Ophthalmology, Hiroshima University, Minami-Ku, Japan; 2２) Department of Ophthalmology, Tsukazaki Hospital, Himeji, Japan; 3２) Department of Ophthalmology, The University of Tokyo, Tokyo, Japan.Purpose: Cataract surgery may change the ocular biomechanical property. Corneal visualization scheimpflug technology (Corvis ST Tonometry: CST) provides several parameters that represent the corneal biomechanical property. The purpose of our study was to investigate the effect of phacoemulsification and intraocular lens implantation on the CST parameters.Methods: Data from 59 normal eyes of 59 subjects (mean age=74.0 years old) were prospectively acquired at Tsukazaki Hospital and Hiroshima University Hospital between February 2014 and December 2014. The intraocular pressure (IOP) with Goldmann applanation tonometry (IOP-G), central corneal thickness, axial length, corneal curvature, and CST parameters were measured before and one week, one month and 3 months after cataract surgery. Measurements with CST and IOP-G were carried out three times per patient and the averages of all three measurements were used in statistical analyses (linear mixed effect model).Results: IOP-G significantly decreased from 15.2±4.3 mmHg to 11.8±3.1 mmHg at 3 months after cataract surgery (p<0.05). The maximum deformation amplitude and peak distance increased significantly. The changes after cataract surgery of A1 time, A2 time, A1 length and A2 length were not significant. Although in normal subjects A2 velocity decreases with IOP reduction (Asaoka, PLOS ONE 2015), A2 velocity in our subjects increased significantly after cataract surgery (p<0.05).Conclusions: The corneal biomechanical property might be changed by cataract surgery.Commercial Relationships: Yoshiaki Kiuchi, None; Yoshitaka Nakao, None; Noriko Ihara; Shunsuke Nakakura, None; Ryo Asaoka, None; Makoto Aihara, NoneSupport: Research Grant 26462688 (to YK) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Program Number: 2395 Poster Board Number: A0124Presentation Time: 3:45 PM–5:30 PMSmall incision lenticule extraction (SMILE) vs LASIK: an ex vivo biomechanical evaluation of low and high myopic correctionsBriana C. Gapsis1, Adam Ahlquist1, Henry Perry1, 4, George Asimellis2, A. J. Kanellopoulos2, 3. 1Department of Ophthalmology, Nassau University Medical Center, East Meadow, NY; 2Laservision.gr Clinical & Research Eye Institute, Athens, Greece; 3Department of Ophthalmology, New York University Medical School, New York, NY; 4Ophthalmic Consultants of Long Island, East Meadow, NY.Purpose: The small-incision lenticule extraction (SMILE) procedure and laser in situ keratomileusis (LASIK) are both safe, effective, and predictable surgical techniques for the correction of myopia. SMILE presents an alternative to LASIK, which currently may be considered the most established form of laser vision correction. The purpose of this study is to evaluate corneal biomechanical changes associated with low and high myopic correction performed with small incision lenticule extraction (SMILE, employing the Visumax laser, Carl Zeiss Meditec, Jena Germany) compared to Femtosecond-laser assisted LASIK (FS200 & EX500 lasers, Alcon Surgical, Ft. Worth, TX)Methods: Thirty human donor corneas were subjected to myopic SMILE or LASIK. These corneas were randomly allocated to one of four investigative groups (n=5 each), subjected to the following treatments: group-A -3.00 diopters (D) SMILE; group-B -8.00D SMILE; group-C -3.00D LASIK; group-D -8.00D LASIK. Additionally, two control groups (n=5 each) were formed, one for each procedure: group-E SMILE and group-F LASIK. The corneas in these control groups were subjected to the corresponding femtosecond-laser lamellar cuts but not to tissue removal. Biomechanical evaluation of tensile strength was conducted by biaxial force application. Primary outcome measures were stress at 10% and 15% strain, and Young’s modulus at 10% and 15% strain.Results: In SMILE, the average relative difference (Δ) of the four metrics evaluated was -35.6% between the -3.00D correction and control and -50.2% between the -8.00D correction and control. In LASIK, average Δ was -21.2% between the -3.00D correction and control, and -50.4% between the -8.00D correction and control. When evaluating the same degree of myopic correction, SMILE, compared to LASIK, appears to result in greater biomechanical reduction for the -3.00D correction (-25.9%) while the difference at -8.00D correction is not statistically significant.Conclusions: Biomechanical tensile strength is reduced with increasing amounts of myopia corrected in both procedures. LASIK results in less strength reduction in smaller myopes while, in higher myopes, SMILE and LASIK appear to result in similar corneal tensile strength reduction.Commercial Relationships: Briana C. Gapsis, None; Adam Ahlquist, None; Henry Perry, Blephex (C), Omidria (C), Ophthalmic Consultants of Long Island, Allergan (C), Alcon (C); George Asimellis, None; A. J. Kanellopoulos, Alcon/WaveLight (C), i-Optics (C), Avedo (C), ISP Surgical (C), Keramed (C), Allergan (C)



Program Number: 2396 Poster Board Number: A0125Presentation Time: 3:45 PM–5:30 PMBlood thyroid hormone levels and corneal tomographical, topographical and biomechanical propertiesZsuzsa Flaskó1, 2, Elena Zemova2, Timo Eppig3, László Módis4, Achim Langenbucher3, Zisis Gatzioufas5, Nóra Szentmáry2, 6, Berthold Seitz2. 1Department of Ophthalmology, Kenézy Hospital, Debrecen, Hungary; 2Department of Ophthalmology, Saarland University Medical Center, Homburg/Saar, Germany; 3Experimental Ophthalmology, Saarland University, Homburg/Saar, Germany; 4Department of Ophthalmology, Debrecen Medical University Center, Debrecen, Hungary; 5Moorfields Eye Hospital, London, United Kingdom; 6Department of Ophthalmology, Semmelweis University, Budapest, Hungary.Purpose: Hypothyroidism could be associated with keratoconus and may be one of its causing factors. Our aim was to determine the potential impact of blood thyroid hormone levels on corneal topography, tomography and biomechanics.Methods: 757 eyes (376 left eyes, 500 males) were included in our study. Patients were classified as euthyroid (n=563) or hypothyroid (n=194) based on blood thyroid hormone status (TSH, FT3, FT4). We further subdivided hypothyroid patients in those with (n=122) and without (n=72) medication. Patients were examined using TMS-5 topographer (Tomey, Tennenlohe, Germany)(keratoconus index: KCI; keratoconus severity index: KSI; keratoconus prediction index: KPI), Pentacam HR tomographer (Oculus Optikgeräte GmbH, Wetzlar, Germany)(keratoconus index: KI; central keratoconus index: CKI) and Ocular Response Analyzer (ORA, Reichert Ophthalmic Instruments, Buffalo, NY) (corneal resistance factor: CFR; corneal hysteresis: CH; keratoconus match index: KMI).Results: Using Mann-Whitney U test, KCI, KSI, KPI of TMS-5 and KI, CKI of Pentacam were significantly lower in hypothyroid patients with medication compared to euthyoids or hypothyroid patients without medication (p<0.001 for all). In addition, KMI, CH and CRF of ORA were significantly greater in hypothyreosis patients with medication compared to euthyroids or hypothyroid patients without medication (p<0.001 for all). Using a generalised linear model, KMI depended on blood FT3 level (p=0.029), however, other parameters of TMS-5, Pentacam or ORA were not affected significantly by blood TSH, FT3 or FT4 levels (p>0.136).Conclusions: Blood TSH, FT3 and FT4 levels may affect keratoconus specific indices of corneal topography and tomography, and corneal biomechanical properties. Unlike in keratoconus, an abnormally increased KMI index of ORA may be indicative for reduced FT3 blood level and hypothyreosis.Commercial Relationships: Zsuzsa Flaskó, None; Elena Zemova, None; Timo Eppig, None; László Módis, None; Achim Langenbucher, None; Zisis Gatzioufas, None; Nóra Szentmáry, None; Berthold Seitz, NoneSupport: HARVO (Hungarian Association for Research in Vision and Ophthalmology) Travel Grant

Program Number: 2397 Poster Board Number: A0126Presentation Time: 3:45 PM–5:30 PMNew dynamic corneal biomechanical parameters derived from curvature versus pressure analysis of Ocular Response Analyzer (ORA) dataDavid Luce. Ophthalmology, Luce Technical Consulting, Clarence Center, NY.Purpose: To present calculation and visualization methods for new and existing corneal biomechanical properties using curvature versus pressure analysis of Ocular Response deformation data.

Methods: Corneal deformation, induced by the ORA air jet, is detected optically, This signal undergoes sign changes, inversion and calibration to provide corneal curvature that is then plotted versus the applied eye pressure.. Energy conversion (to heat) and the dynamic viscoelastic moduli during the 30 msec deformation are explicitly visible in the “hysteresis” loop that is created. Numerical results are calculated from the loop’s geometric properties, especially areas of the sub-loops of positive and negative curvature. A Corneal Bending Modulus (CBM) is determined from the slope of the initial curvature following first applanation (zero curvature).Retrospective ORA data from 34 eyes of 17 subjects from a regular clinical practice and with varied ocular conditions were analyzed.Results: We report here on statistics for a single parameter, the Corneal Bending Modulus (CBM), the average slope of the pressure/curvature function in a region starting at first applanation and ending before peak pressure and any iris interference. The KC and 4 norm.diverse population’s mean/std dev / age is 47 ± 9 yrs. and includes normal, keratoconus, LASIK, INTACS, PRK, PKP subjects. The CV for the CBM between right and left eyes is 26% and 9.`% for unaltered eyes, normal and KC, N=7.Conclusions: In all cases in the 34 diverse eyes normal eyes have lower compliance than the KC eyes (N=7). In all cases the populations examined here are small. Despite ourconclusions are that there are indications that the technique at using curvature shows significant promise. Muc that justifies continued work. It should be pointed out that the mathematical treatment taken here differs fom the current ORA in that the inversion of the raw ORA data emphasizes the low signal levels regions, the opposite of the current ORA data, The new biomechanical method method extracts informartion during the initinal negative curvature, a unique balance point. In addition any significant interference/with the iris\lens is cleaarly visible and detected.

Diverse CBM images



Figure 2 - Summary of CBM trial resultsCommercial Relationships: David Luce

Program Number: 2398 Poster Board Number: A0127Presentation Time: 3:45 PM–5:30 PMBiomechanical properties and IOP reconstruction from air-puff corneal deformation imaging: validations in model and porcine eyesNandor Bekesi, Carlos Dorronsoro, Susana Marcos. Institute of Optics, Consejo Superior de Investig Cientif CSIC, Madrid, Spain.Purpose: Keratoconus is a disease that alters the mechanical behavior of the cornea. Air-puff corneal dynamic deformation depends on the corneal biomechanical properties, corneal thickness and Intraocular Pressure (IOP). The aim of the study is to estimate both mechanical properties and IOP by means of inverse Finite Element modeling.Methods: Corneal deformation of artificial hydrogel corneal models and enucleated porcine eyes (n = 8) to an air puff were measured by high speed Scheimpflug imaging (Corvis ST) at different IOPs (15 to 45 mmHg). A computational optimization incorporating inverse modeling was utilized in order to obtain mechanical properties from Corvis deformation images. Corneal Finite Element Models were programmed in ANSYS software, using Scheimpflug-based corneal geometry measurements, and literature values for the sclera. Air-puff pressure dynamics was measured in a previous study (Kling et al PLOS One 2014). The parameters of the hyperelastic material model and the IOP were changed in an iterative simulation to fit the simulated deformed shape and the deformation history of the corneas with the measured ones.Results: The maximum deformation amplitudes (DA) of the artificial corneal model were 0.69, 0.52 and 0.47 mm at IOP of 15, 30 and 45 mmHg, respectively. Simulations resulted in Young’s moduli of 0.475, 0.465 and 0.478 MPa for the three IOP levels respectively. DA of porcine corneas were 1.25, 0.869 and 0.624 mm at IOP of 15, 30 and 45 mmHg, respectively. Resulting Young’s moduli of the cornea were 0.462, 0.436 and 0.454 MPa. The simulated IOP values matched the nominal IOP values within the step size (1 mmHg) of the optimization.Conclusions: High speed imaging together with numerical optimization allows simultaneous accurate reconstruction of both corneal mechanical properties and IOP, as validated with models ex vivo. This study extends previous work where the reconstructed mechanical parameters were obtained ex vivo with fixed, known IOP. The proposed method can be useful to monitor corneal mechanical

parameters in patients (i.e. keratoconus and cross-linking treatment). Also, it allows absolute estimates of IOP, independent of mechanical parameters, of high relevance to monitoring of glaucoma.Commercial Relationships: Nandor Bekesi; Carlos Dorronsoro; Susana Marcos, NoneSupport: Comunidad de Madrid & EU Marie Curie (FP7/2007-2013/REA 291820), ERC Advanced Grant 294099, Spanish Government Grant FIS2014-56643-R

Program Number: 2399 Poster Board Number: A0128Presentation Time: 3:45 PM–5:30 PMFinite Element Analysis of Penetrating Injury to the Human EyeScott Lovald1, Andrew Rau1, Steven Nissman2, Nicoli Ames3, John McNulty4, Jorge Ochoa1, Michael Baldwinson5. 1Biomedical Engineering, Exponent, Menlo Park, CA; 2Scheie Eye Institute, University of Pennsylvania Perelman School of Medicine, Philadelphia, PA; 3Mechanical Engineering, Exponent, Boulder, CO; 4Materials and Corrosion Engineering, Exponent, Menlo Park, CA; 5Google[x], Mountain View, CA.Purpose: Penetrating injuries to the eye are among the most frequent causes of permanent visual impairment. The purpose of this study was to determine the strain at which rupture occurs in the cornea due to a penetrating object.Methods: For the experiment, probes of varying diameters (1.0, 1.5, and 2.0 mm) were pressed into the apex of the cornea in 36 human cadaveric globes until perforation of the specimen. To match the experiment, an axisymmetric finite element model of the idealized human globe was created in Abaqus 6.14 (Figure 1). The cornea and sclera were modeled as isotropic nonlinear hyperelastic materials. To evaluate model sensitivity, three separate models for the sclera were constructed (minimum, medium, and maximum stiffness) within the range of experimentally-observed material behavior. In addition, two separate internal pressure conditions were implemented: 1) a sealed fluid cavity with an initial pressure, and 2) a constant pressure applied directly to the cavity surfaces. The model was used to map the force-displacement response of the experiments and quantitatively determine a peak strain at which the eye ruptures.Results: For the experiments, the average force at failure increased from 30.5±5.5 N (1.0 mm probe) to 40.5±8.3 N (1.5 mm probe) to 58.2±14.5 N (2.0 mm probe) as the probe size increased (p<0.002). The force-displacement responses of the finite element models of all three probe sizes bounded and tracked the experimental data. The peak strain at failure in the cornea was located on its posterior surface. This strain was in the range of 29% to 33% for all models analyzed. Figure 2 shows strain contours for the 1 mm probe.Conclusions: The current study has developed a validated finite element model of the human eye for analysis of penetrating injury to the cornea. The results have determined an objective failure strain of corneal tissue, which is consistent between sensitivity studies of varying material models, pressure conditions, and penetrating objects sizes. These results provide critical, quantitative information for understanding the risk of penetrating eye injuries.



Finite element model set-up for globe indentation. Material sections are indicated in the middle image. The figure on the right shows the indenter probe displaced into the cornea.

The maximum principal strain in the cornea is shown for each 1.0 mm probe analysis at the step corresponding to the average experimental failure force.Commercial Relationships: Scott Lovald; Andrew Rau, Google[x] (F); Steven Nissman, None; Nicoli Ames, Google[x] (F); John McNulty, Google[x] (F); Jorge Ochoa, Google[x] (F); Michael Baldwinson, Google[x] Support: Partial financial support for this study was provided by Google[X]

Program Number: 2400 Poster Board Number: A0129Presentation Time: 3:45 PM–5:30 PMShort-term orthokeratology on corneal tangent elastic modulus - a pilot studyAndrew K. Lam1, Ying Hon1, Angel Wong1, Jimmy Tse1, Guo-Zhen Chen2, Shu-Hao Lu2, David C. C. Lam2. 1School of Optometry, The Hong Kong Polytechnic University, Hong Kong, China; 2Department of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, Hong Kong, China.

Purpose: To investigate changes in corneal tangent elastic modulus and other corneal biomechanics following short-term orthokeratology.Methods: Seventeen Chinese subjects (age, 18-25 years) were fitted with orthokeratology (ortho-k) lenses for 1 month in both eyes. Their baseline spherical equivalent was between -2.13 and -4.38D with a corneal toricity within 2D. Subjective refraction and ocular measurements were monitored in 4 visits, namely at the baseline and after the first overnight wear, the first week, and first month of ortho-k treatment. Corneal topography and pachymetry were performed. The corneal tangent elastic modulus (E) was measured using a novel corneal indentation device. Corneal hysteresis (CH) and the corneal resistance factor (CRF) were measured using the Ocular Response Analyzer. All measurements were conducted in the morning. One eye from each subject with a better uncorrected visual acuity (UCVA) after 1 month of ortho-k treatment was used for analysis.Results: The mean spherical equivalent decreased from -3.15 +/- 0.72D at the baseline to -0.11 +/- 0.22D at 1 month, final UCVA -0.07 +/- 0.07. Significant corneal flattening (both the steepest and flattest meridians) was observed as quickly as after 1 night of ortho-k treatment (Friedman test, p<0.001). Similar rapid central corneal thinning was demonstrated (Repeated measures analysis of variance, RMANOVA, p<0.001). Compared with the baseline, the CRF decreased significantly throughout the ortho-k treatment (RMANOVA, p=0.001) (from 10.4 to 9.8mmHg, a reduction of 5.8%). No significant change was observed in CH (RMANOVA, p=0.066). A significant increase in E was observed (RMANOVA, p=0.001) (from 0.456 to 0.509MPa, an increase of 11.8%). No correlation was found between initial corneal tangent elastic modulus and the speed of myopia reduction.Conclusions: Similar to previous studies, short-term ortho-k treatment altered corneal biomechanics in terms of the CRF with no change in CH. The increase in corneal tangent elastic modulus revealed a stiffer cornea from ortho-k treatment. Should corneal tangent elastic modulus be a more suitable parameter for evaluating the corneal response to ortho-k treatment requires more studies.

Corneal hysteresis and corneal resistance factor in ortho-k treatment. Each error bar indicates 1 standard deviation.



Corneal tangent elastic modulus in ortho-k treatment. Each error bar indicates 1 standard deviation.Commercial Relationships: Andrew K. Lam, None; Ying Hon, None; Angel Wong, None; Jimmy Tse, None; Guo-Zhen Chen, None; Shu-Hao Lu, None; David C. C. Lam

ARVO 2 Annual Meeting Abstracts 273 Corneal Extracellular ... · Program Number: 2357 Poster Board Number: A0086 Presentation Time: 3:45 PM–5:30 PM p38/TGFβ-signaling pathway involved

Documents