Establishment And Clinical Application Of In Vivo Corneal Biomechanical Measurement Methods And Preliminary Study Of Genipin Cross-linking On Rabbit Cornea

Objective：To assess the repeatability and reproducibility of IOP measurements and thebiomechanical properties of cornea in eyes with keratoconus and glaucoma usingCorvis ST. A novel system which combined optical coherence tomography (OCT)with air jet indentation technology is developed to assess the corneal biomechanicalproperties. The repeatability and reproducibility of the system, and feasibility ofclinical application are investigated through corneal phantom experiment andclinical trials, respectively. To evaluate histologic and biomechanical change inrabbit cornea after collagen cross-linking (CXL), using Genipin with0.2%concentration, Rapid-CXL (30mW/cm2in3min) and Standard-CXL (3mW/cm2in30min).Methods:1. A comparison control study on. The subjects included normal cornea volunteer,keratoconus patients and patients with glaucoma. All patients completed thesystematic ophthalmology examination, and the biomechanical measurement usingCorvis ST. Evaluation the repeatability and reproducibility of IOP measurement, andcompared with the NCT and DCT measurement. Comparison the cornealbiomechanical properties between normal and keratocon eyes and eyes withglaucoma, to evaluate the diagnosis efficiency of the biomechanical parameters.2. A corneal mold was designed to make eight different stiffness phantoms with thermal forming method. The IOP of phantom are regulated through saline-filledpressure-controlling manometer and measured by pressure sensor. Three consecutivemeasurements at each set pressure were obtained by two operators each to assess therepeatability and reproducibility of the system. Strip extensiometry tests wereperformed to determine the corneal phantom Young’s modulus and compare with theresults from OCT based air jet indentation measurement. In addition,10normal eyesfrom10people were recruited to investigative the clinical application feasibility ofthe OCT based air jet indentation system.3. Fifty-seven healthy Chinese rabbits were cross-linked using Genipin, Rapid-CXLand Standard-CXL with the removal of corneal epithelium. The right eye waschosen as experimental group, and left eye as the control group. At1day,7days and1month after CXL, the corneas were evaluated using hematoxylin-eosin staining(HE) and immunostaining with markers of epithelial junction protein (-catenin),stromal keratocytes (CD34), apoptosis (TUNEL) and collagen fibers (Massonstaining). At1day and3months after CXL, strip extensiometry test were performedto evaluate corneal biomechanical properties of the rabbit cornea with different CXLmethods.Results:1. The measurements of IOP using Corvis ST had a good repeatability, with anintraclass correlation coefficient (ICC) of0.83. Bland-Altman consistency analysisshowed, the reproducibility between the IOPCorand IOPNCTwas good, within thegroup deviation of (-0.5±1.8) mmHg, but poor reproducibility with IOPDCT, thatthe group deviation was (-3.9±2.1) mmHg.2. Most of the biomechanical parameters of the keratoconic eyes were significantlydifferent from those of the normal eyes. The deformation amplitude (DA) was thebest predictive parameter, with an area under the curve of0.865. With the severity ofkeratoconus upgrading, the diagnostic efficiency of deformation amplitudeincreased. In both the normal and keratoconus groups, the DA was negativelycorrelated with IOP, CCT, and corneal volume at3mm. 3. During the first and second applanations, normal eyes flattened faster and had asmaller radius of curvature and a shorter peak distance at the highest concavity thandid glaucomatous eyes. DA was lower in the POAG and PACG groups than in thecontrol group (P <0.05for both), but no difference was found between the POAGand PACG groups. DA was negatively correlated with IOPGATin all three groups4. The central thickness of the corneal phantoms were504.12±17.04μm and theYoung’s moduli of the corneal phantoms measured by the strip extensiometry testrange from90to1,400kPa. The ICC of the measurement in corneal phantom andhuman corneas were0.9926and0.9695, respectively. The Bland-Altman plotshowed that the measurements by different operator are consistent with a differenceless than1.1μm. Results showed that the corneal phantom stiffness measured by thecurrent system correlated well with the corresponding Young’s modulus obtained bystrip extensiometry test (r=0.99, P <0.001).5. The-catenin expression levels in corneal epithelial cells found to be decreasedafter CXL and gradually recovered over time in all the three groups, and resumedback to normal level1month after CXL. There were no difference in-cateninexpression levels and percentage of apoptotic cells in epithelial cells during therecovery process among the three groups. The analysis of stroma componentsindicated a loss of keratocytes in supper stroma in all the three CXL methods, withthe most severe in Genipin group; similarly, collagen fibers appeared to bedisorganized in all groups while their pattern resumed back to normal after1month.Corneal edema and inflammatory cell infiltration were observed using HE stainingin all three groups. The Genipin group had the most severe inflammation and theslowest rate of recovery. After CXL, the expression of CD34decreased and was stillhigher than normal, as compared with1month after CXL. Three CXL methodsshowed no corneal endothelial cell injury at any time points.6. All of the CXL methods increased the postoperative corneal biomechanical strength.The corneal Young’s modulus at8%strain measured at7days after treatment inGenipin group showed an increase of278.65%, as compared with the control group. There was no significant difference of modulus between Rapid-CXL andStandard-CXL at7days and3month. The corneal biomechanical strength wasobserved to be decreased at the3months after CXL compared with7days in all thethree groups, and the decrease in Genipin group was about35%to41%.Conclusions:1. Corvis ST offers an alternative method for measuring corneal biomechanicalproperties. The possibility of classifying keratoconus based on deformationamplitude deserves clinical attention. Several parameters including DA exhibitedsignificant differences between glaucomatous eyes and normal control eyes.Assessment with the Corvis ST may allow clinicians to detect glaucoma.2. Preliminary experimental results showed that OCT based air jet indentation systemhas high reproducibility and repeatability. This measurement system provides anovel non-invasive and non-contact means for the assessment of cornealbiomechanical properties for using in medical diagnosis and monitoring ofbiomechanical properties of the cornea.3. Histologic and immunohistochemical findings on rabbit corneas showed that thebiochemical and morphologic alterations in the epithelium and stroma aresignificantly improved1month after CXL. All the three CXL methods caneffectively strengthen the postoperative corneal biomechanical strength, with theGenipin group increased the most. There was no significant difference of modulusbetween Rapid-CXL and Standard-CXL, and the corneal biomechanical strengthwas found to be decreased over time.