Efficacy and safety of combined UV-light corneal crosslinking and fine-needle diathermy to regress pathological murine corneal (lymph)angiogenesis in vivo.

PURPOSE: To compare safety and efficacy of isolated and combined UV-light corneal crosslinking (CXL) and fine-needle diathermy (FND) to regress pathological corneal vessels in vivo. METHODS: Mice with inflamed and pathologically vascularized corneas received CXL or FND as monotherapy or a combination of both treatments. Corneal pathological blood and lymphatic vessels, immune cells and the morphology of anterior segment structures were evaluated. RESULTS: All three approaches were able to regress blood and lymphatic vessels in mice. A comparative analysis of the three methods revealed that the FND monotherapy and the CXL + FND combination were significantly more effective than the CXL monotherapy, one and 2 weeks after therapy and especially in regressing lymphatic vessels. Furthermore, the combination therapy induced significantly less immune cell recruitment compared to the monotherapies. All three methods were safe to use in regards of corneal integrity. CONCLUSIONS: A combination of FND and CXL led to regression of pathological corneal lymphatic and blood vessels and reduced the infiltration of immune cells into inflamed murine corneas. This approach offers a new effective, safe and clinically usable strategy to treat eyes with mature pathological blood vessels and even more so for lymphatic vessels, for example prior to high-risk corneal transplantation.

The role of the osmosensitive transcription factor NFAT5 in corneal edema resorption after injury.

The osmosensitive transcription factor nuclear factor of activated T cells 5 (NFAT5; or tonicity-responsive enhancer binding protein; TonEBP) plays a key role in macrophage-driven regulation of cutaneous salt and water balance. In the immune-privileged and transparent cornea, disturbances in fluid balance and pathological edema result in corneal transparency loss, which is one of the main causes of blindness worldwide. The role of NFAT5 in the cornea has not yet been investigated. We analyzed the expression and function of NFAT5 in naive corneas and in an established mouse model of perforating corneal injury (PCI), which causes acute corneal edema and transparency loss. In uninjured corneas, NFAT5 was mainly expressed in corneal fibroblasts. In contrast, after PCI, NFAT5 expression was highly upregulated in recruited corneal macrophages. NFAT5 deficiency did not alter corneal thickness in steady state; however, loss of NFAT5 led to accelerated resorption of corneal edema after PCI. Mechanistically, we found that myeloid cell-derived NFAT5 is crucial for controlling corneal edema, as edema resorption after PCI was significantly enhanced in mice with conditional loss of NFAT5 in the myeloid cell lineage, presumably due to increased pinocytosis of corneal macrophages. Collectively, we uncovered a suppressive role for NFAT5 in corneal edema resorption, thereby identifying a novel therapeutic target to combat edema-induced corneal blindness.

Cystathionine ß-synthase as novel endogenous regulator of lymphangiogenesis via modulating VEGF receptor 2 and 3.

Lymphangiogenesis is a key player in several diseases such as tumor metastasis, obesity, and graft rejection. Endogenous regulation of lymphangiogenesis is only partly understood. Here we use the normally avascular cornea as a model to identify endogenous regulators of lymphangiogenesis. Quantitative trait locus analysis of a large low-lymphangiogenic BALB/cN x high-lymphangiogenic C57BL/6 N intercross and prioritization by whole-transcriptome sequencing identify a novel gene responsible for differences in lymphatic vessel architecture on chromosome 17, the cystathionine ß-synthase (Cbs). Inhibition of CBS in lymphatic endothelial cells results in reduce proliferation, migration, altered tube-formation, and decrease expression of vascular endothelial growth factor (VEGF) receptor 2 (VEGF-R2) and VEGF-R3, but not their ligands VEGF-C and VEGF-D. Also in vivo inflammation-induced lymphangiogenesis is significantly reduce in C57BL/6 N mice after pharmacological inhibition of CBS. The results confirm CBS as a novel endogenous regulator of lymphangiogenesis acting via VEGF receptor 2 and 3-regulation and open new treatment avenues in diseases associated with pathologic lymphangiogenesis.

Segmentation and Evaluation of Corneal Nerves and Dendritic Cells From In Vivo Confocal Microscopy Images Using Deep Learning.

Purpose: Segmentation and evaluation of in vivo confocal microscopy (IVCM) images requires manual intervention, which is time consuming, laborious, and non-reproducible. The aim of this research was to develop and validate deep learning-based methods that could automatically segment and evaluate corneal nerve fibers (CNFs) and dendritic cells (DCs) in IVCM images, thereby reducing processing time to analyze larger volumes of clinical images. Methods: CNF and DC segmentation models were developed based on U-Net and Mask R-CNN architectures, respectively; 10-fold cross-validation was used to evaluate both models. The CNF model was trained and tested using 1097 and 122 images, and the DC model was trained and tested using 679 and 75 images, respectively, at each fold. The CNF morphology, number of nerves, number of branching points, nerve length, and tortuosity were analyzed; for DCs, number, size, and immature-mature cells were analyzed. Python-based software was written for model training, testing, and automatic morphometric parameters evaluation. Results: The CNF model achieved on average 86.1% sensitivity and 90.1% specificity, and the DC model achieved on average 89.37% precision, 94.43% recall, and 91.83% F1 score. The interclass correlation coefficient (ICC) between manual annotation and automatic segmentation were 0.85, 0.87, 0.95, and 0.88 for CNF number, length, branching points, and tortuosity, respectively, and the ICC for DC number and size were 0.95 and 0.92, respectively. Conclusions: Our proposed methods demonstrated reliable consistency between manual annotation and automatic segmentation of CNF and DC with rapid speed. The results showed that these approaches have the potential to be implemented into clinical practice in IVCM images. Translational Relevance: The deep learning-based automatic segmentation and quantification algorithm significantly increases the efficiency of evaluating IVCM images, thereby supporting and potentially improving the diagnosis and treatment of ocular surface disease associated with corneal nerves and dendritic cells.

Dynamic Contrast Microscopic Optical Coherence Tomography As a Novel Method for Assessing Corneal Epithelium During Exposure to Benzalkonium Chloride.

Purpose: Microscopic optical coherence tomography (mOCT) has an imaging resolution of 1 µm in all voxel dimensions, but individual epithelial cells are difficult to resolve due to lack of scattering contrast. Adding dynamic contrast processing to mOCT (dmOCT) results in color images that enable visualization of individual cells and possibly give information on cellular function via the calculation of a motility coefficient. We propose this technique as a novel method of evaluating the ocular surface after exposure to a toxic chemical, benzalkonium chloride (BAK). Methods: Ex vivo cross-section images were acquired with a custom-built, frequency-domain mOCT system. Eyes were explanted from healthy adult C57BL/6 mice and imaged every 30 minutes with five sets of dmOCT scans at each imaging time. Total epithelium and stroma thicknesses were measured from a single mOCT B-scan, and measures of color changes (hue) and the motility coefficient were acquired from dmOCT scans. Results: After 30-minute exposures to 0.005% BAK, local motility decreased and total epithelium thickness increased compared to controls. For basal epithelium cells, local motility decreased after 60-minute exposures, and the hue shifted red after 90-minute exposures. Stroma thickness did not significantly swell until 150-minute exposures to BAK. Conclusions: dmOCT allows us to view the behavior of the cornea epithelium under toxic stress due to BAK, revealing parallel swelling of the extracellular matrix and changes in local subcellular motion. Translational Relevance: The evaluation of the cornea epithelium using dmOCT is helpful to our understanding of the toxic effects of BAK.

Longitudinal In Vivo Changes in Radial Peripapillary Capillaries and Optic Nerve Head Structure in Non-Human Primates With Early Experimental Glaucoma.

Purpose: There is conflicting evidence regarding whether a loss of radial peripapillary capillaries (RPCs) precedes neuronal loss in glaucoma. We examined the time course of in vivo changes in RPCs, optic nerve head (ONH) structure, and retinal nerve fiber layer thickness (RNFLT) in experimental glaucoma (EG). Methods: Spectral domain optical coherence tomography images were acquired before and approximately every two weeks after inducing unilateral EG in nine rhesus monkeys to quantify mean anterior lamina cribrosa surface depth (ALCSD), minimum rim width (MRW), and RNFLT. Perfused RPC density was measured from adaptive optics scanning laser ophthalmoscope images acquired on the temporal half of the ONH. The time of first significant change was quantified as when values fell and remained outside of the 95% confidence interval established from control eyes. Results: Mean ALCSD and/or MRW were the first parameters to change in eight EG eyes. RPC density changed first in the ninth. At their first points of change, mean ALCSD posteriorly deformed by 100.2 ± 101.2 µm, MRW thinned by 82.3 ± 65.9 µm, RNFLT decreased by 25 ± 14 µm, and RPC density decreased by 4.5 ± 2.1%. RPC density decreased before RNFL thinning in 5 EG eyes. RNFLT decreased before RPC density decreased in two EG eyes, whereas two EG eyes had simultaneous changes. Conclusions: In most EG eyes, RPC density decreased before (or simultaneous with) a change in RNFLT, suggesting that vascular factors may play a role in axonal loss in some eyes in early glaucoma.

Topical application of cannabinoid-ligands ameliorates experimental dry-eye disease.

PURPOSE: Dry eye disease (DED) is a multifactorial disease, with limitations regarding efficacy and tolerability of applied substances. Among several candidates, the endocannabinoid system with its receptors (CB1R and CB2R) were reported to modulate inflammation, wound healing and pain, which are also core DED pathomechanisms. This study is to investigate the therapeutic responses of Δ-9 tetrahydrocannabinol (a non-selective agonist) and two selective antagonists, SR141716A (CB1R antagonist) and SR144528 (CB2R antagonist), as a topical application using a DED mouse model. METHOD: Experimental DED was induced in naïve C57BL/6 mice. Expression of CBR at the ocular surface of naïve and DED mice was determined by qPCR and in-situ hybridization. Either THC or CBR antagonists were compounded in an aqueous solution and dosed during the induction of DED. Tear production, cornea sensitivity, and cornea fluorescence staining were tested. At the end of each experiment, corneas were stained with ß3-tubulin for analysis of corneal nerve morphology. Conjunctiva was analyzed for CD4+ and CD8+ infiltration. RESULTS: CB1R and CB2R are present at the ocular surface, and desiccating stress increased CBR expressions (p < 0.05). After 10 days of DED induction, treated groups demonstrated a reduced CBR expression in the cornea, which was concurrent with improvements in the DED phenotype including fluorescence staining & inflammation. Applying THC protected corneal nerve morphology, thus maintained corneal sensitivity and reduced CD4+ T-cell infiltration. The CB1R antagonist maintained cornea sensitivity without changing nerve morphology. CONCLUSIONS: Endocannabinoid receptor modulation presents a potential multi-functional therapeutic approach for DED.

Ocular Graft-versus-Host Disease in a Chemotherapy-Based Minor-Mismatch Mouse Model Features Corneal (Lymph-) Angiogenesis.

Ocular graft-versus-host disease (oGVHD) is a fast progressing, autoimmunological disease following hematopoietic stem cell transplantation, leading to severe inflammation of the eye and destruction of the lacrimal functional unit with consecutive sight-threatening consequences. The therapeutic "window of opportunity" is narrow, and current treatment options are limited and often insufficient. To achieve new insights into the pathogenesis and to develop new therapeutic approaches, clinically relevant models of oGVHD are desirable. In this study, the ocular phenotype was described in a murine, chemotherapy-based, minor-mismatch GVHD model mimicking early-onset chronic oGVHD, with corneal epitheliopathy, inflammation of the lacrimal glands, and blepharitis. Additionally, corneal lymphangiogenesis was observed as part of oGVHD pathogenesis for the first time, thus opening up the investigation of lymphangiogenesis as a potential therapeutic and diagnostic tool.

Automatic Segmentation of Retinal Capillaries in Adaptive Optics Scanning Laser Ophthalmoscope Perfusion Images Using a Convolutional Neural Network.

Purpose: Adaptive optics scanning laser ophthalmoscope (AOSLO) capillary perfusion images can possess large variations in contrast, intensity, and background signal, thereby limiting the use of global or adaptive thresholding techniques for automatic segmentation. We sought to develop an automated approach to segment perfused capillaries in AOSLO images. Methods: 12,979 image patches were extracted from manually segmented AOSLO montages from 14 eyes and used to train a convolutional neural network (CNN) that classified pixels as capillaries, large vessels, background, or image canvas. 1764 patches were extracted from AOSLO montages of four separate subjects, and were segmented manually by two raters (ground truth) and automatically by the CNN, an Otsu's approach, and a Frangi approach. A modified Dice coefficient was created to account for slight spatial differences between the same manually and CNN-segmented capillaries. Results: CNN capillary segmentation had an accuracy (0.94), a Dice coefficient (0.67), and a modified Dice coefficient (0.90) that were significantly higher than other automated approaches (P < 0.05). There were no significant differences in capillary density and mean segment length between manual ground-truth and CNN segmentations (P > 0.05). Conclusions: Close agreement between the CNN and manual segmentations enables robust and objective quantification of perfused capillary metrics. The developed CNN is time and computationally efficient, and distinguishes capillaries from areas containing diffuse background signal and larger underlying vessels. Translational Relevance: This automatic segmentation algorithm greatly increases the efficiency of quantifying AOSLO capillary perfusion images.

In vivo assessment of foveal geometry and cone photoreceptor density and spacing in children.

The fovea undergoes significant developmental changes from birth into adolescence. However, there is limited data examining cone photoreceptor density, foveal pit shape, and foveal avascular zone (FAZ) size in children. The purpose of this study was to determine whether overall foveal structure differs as a function of age and refractive status in children. Forty-eight healthy children (ages 5.8 to 15.8 years) underwent optical coherence tomography imaging to quantify foveal point thickness and foveal pit diameter, depth, and slope. Adaptive optics scanning laser ophthalmoscope (AOSLO) images of foveal capillaries and cone photoreceptors were acquired in a subset of children to quantify FAZ metrics and cone densities at 0.2, 0.3, and 0.5 mm eccentricities. Results show that foveal pit and FAZ metrics were not related to age, axial length, or refractive status. However, linear cone density was lower in myopic versus non-myopic children at eccentricities of 0.2 mm (mean ± SD = 50,022 ± 5,878 cones/mm2 vs 58,989 ± 4,822 cones/mm2, P < 0.001) and 0.3 mm (43,944 ± 5,547 cones/mm2 vs 48,622 ± 3,538 cones/mm2, P < 0.001). These results suggest FAZ and foveal pit metrics do not systematically differ with age in children, while myopic eyes have decreased linear cone density near the foveal center. Significance Statement: The development of the fovea begins prior to birth and continues through the early teenage years until it reaches adult-like properties. Although the majority of changes during childhood are related to the maturation and migration of cone photoreceptors, in vivo data describing cone packing in children is limited. We assessed overall foveal structure in children as young as 5.8 years old by quantifying cone density and spacing, foveal avascular zone size, and foveal pit morphometry to investigate potential structural differences as a function of age and refractive status. While foveal avascular zone and foveal pit metrics did not significantly differ with age, results indicate that myopic children have lower linear cone densities close to the foveal center compared to non-myopic children.

The optic nerve head, lamina cribrosa, and nerve fiber layer in non-myopic and myopic children.

The purpose of this study was to evaluate the optic nerve head, lamina cribrosa, retina, and choroid in school age children using spectral domain optical coherence tomography (SD-OCT) and to assess these structural parameters in relation to age, axial length, and refractive error. Healthy children, ages 11.15 ± 2.62 years (range 6-15 years, n = 53), underwent cycloplegic autorefraction, biometry, and SD-OCT imaging in both eyes. Images were analyzed using custom written programs in MATLAB, after adjustment for lateral magnification. Peripapillary retinal nerve fiber layer (RNFL) thickness, retinal and choroidal thicknesses, Bruch's membrane opening (BMO) area, minimum rim width (MRW), and anterior lamina cribrosa surface depth (ALCSD) were determined and analyzed with age, axial length, and refraction. Results show that axial length increased and refractive error became more myopic with increasing age (R2 = 0.25, ß = 0.18, P < 0.0001 and R2 = 0.27, ß = -0.37, P < 0.0001, respectively). Minimum foveal thickness and central 1 mm retinal thickness increased with increasing age (R2 = 0.15, ß = 2.38, P < 0.01 and R2 = 0.11, ß = 3.16, P < 0.05, respectively). Age-adjusted raw values for peripapillary RNFL thickness decreased with increasing axial length (R2 = 0.11, ß = -3.18, P < 0.05); however, this relationship was not present when image magnification was corrected (R2 = 0.07, ß = 2.72, P = 0.09). BMO area increased with myopic refractive error (R2 = 0.16, ß = -0.10, P < 0.01). Age-adjusted vertical cup-to-disc ratio decreased with increasing axial length and myopic refractive error (R2 = 0.12, ß = -0.05, P < 0.05 and R2 = 0.11, ß = 0.03, P = 0.05, respectively). Mean MRW, mean ALCSD, and peripapillary choroidal thickness were not associated with age, axial length, or refraction. Mean MRW was significantly thinner in eyes with deeper ALCS (R2 = 0.41, ß = -0.83, P < 0.0001). These findings provide normal values for retinal and optic nerve head parameters in school age children, and also suggest that ocular remodeling occurs in some structures in school age children with normal eye growth and during early stages of myopia development.