Development and evaluation of a novel porcine desiccation-induced dry eye model for investigation of the aetiology and treatment for dry eye : implications for the ageing eye

Abstract

Dry eye syndrome is multi-factorial in origin and has become a worldwide ocular problem, especially affecting the elderly. There have been several investigations into the aetiology and treatment of dry eye. Apart from clinical studies, several animal models of dry eye have been studied. However, due to the species difference and the complexity of dry eye, no single animal model can represent completely the human dry eye condition. Porcine organs are better substitutes for human organs; however, porcine eyes have never been used to study dry eye. Moreover, among all developed animal models, none can be used to study the effect of blinking in dry eye. Therefore, the aims of this study were to: * Determine the possibility of using porcine eye to establish an animal dry eye model (DEM) * Develop a novel porcine dry eye model (pDEM) with adjustable 'lacrimation' and 'blink' rates * Ascertain the reproducibility of the pDEM system * Simulate different severities of desiccation-induced dry eye in the pDEM * Investigate the effect of different 'blink' rates on desiccation-induced dry eye simulated by the pDEM * Investigate the effect of different artificial tears on desiccation-induced dry eye using the pDEM * Investigate the effect of different viscosities of solutions on desiccation-induced dry eye using the pDEM. Baseline data on the number of dead cells on the porcine corneal epithelial surface were obtained using the trypan blue exclusion technique. The medians (range) of the number of dead cells in the central and peripheral porcine corneal epithelium (standardized area of 0.25 mm2) were 136 (112-173) and 96 (82-107) cells, respectively when stained immediately after enucleation. The number of dead cells in the central region was significantly greater (p < 0.05) than in the periphery in all groups. After exposure to air, the corneal cells were maintained by regular administration of saline solution but there was a significant increase (p < 0.05) in the number of dead cells in those corneas without desiccation protection. A pDEM was developed and evaluated so that different severities of desiccation-induced dry eye could be simulated by adjusting the 'lacrimation-blink' intervals. The integrity of the porcine cornea and the number of dead corneal epithelial cells were assessed by sodium fluorescein and the trypan blue exclusion technique, respectively. There were significant differences (p < 0.01) between the final fluorescein grades in the corneas when the 'lacrimation-blink' intervals set at 20 seconds, 40 seconds and 60 seconds and the medians were grade 1, grade 2.5 and grade 4, respectively. There were also significant increases (p < 0.05) in the numbers of dead cells in the central cornea, when the 'lacrimation-blink' interval increased. The effect of different 'blink' rates on the desiccation-induced dry eye was studied by keeping the same amount of 'lacrimation' but varying the 'blink' rate. Increasing the 'blink' rate from 3 blinks/minute to 5 blinks/minute significantly reduced the amount of corneal damage, however, a further increase to 10 blinks/minute did not result in a further significant reduction in the corneal damage. When studying the effect of different artificial tear formulations, it was found that both Bion(R)Tears and Vismed(R)are more effective than Senju(R) and DPBS for protecting the cornea against desiccation. This result might relate to the viscosity of the artificial tear formulations as Bion(R)Tears and Vismed(R) have higher viscosities than Senju(R). The effect of different viscosities was studied by applying 0.3%, 1.5% and 2.0% methylcellulose solutions with viscosity of 1.5 cP, 6.2 cP and 9.2 cP, respectively, to the corneal surface in the desiccation-induced dry eye model. The increase in viscosity from 1.5 cP to 6.2 cP significantly reduced (p = 0.008) corneal damage. However, a further increase in the viscosity from 6.2 cP to 9.2 cP did not further reduce (p > 0.05) corneal damage. In conclusion, we have developed and evaluated a novel porcine desiccation-induced dry eye model, incorporating 'lacrimation' and 'blinking' systems, that is capable of simulating different severities of dry eye and which can be used to study the aetiology and treatment of desiccation-induced dry eye.