Vascular reaction parameters correlate with severe nerve fibre thickness loss and visual field defects in glaucoma.

PURPOSE: Vascular risk factors and ocular perfusion abnormalities are key elements in the pathogenesis of glaucoma. The retinal vessel analyser (RVA; IMEDOS Systems, Germany) enables non-invasive assessment of dynamic changes in retinal vessel diameters in response to light stimulation. In this pilot study, we explored whether parameters of vascular regulation correlate with visual field defects and retinal nerve fibre layer (RNFL) thickness (SD-OCT, Spectralis) in glaucoma patients. METHODS: A cross-sectional observational study was conducted involving 34 eyes from 34 patients with advanced visual field defects associated with primary open-angle glaucoma (POAG). Following pharmacological pupil dilation, RVA measurements were performed according to a standardised protocol including stimulation with monochromatic flicker light. The resulting vascular response parameters were then analysed for correlations with both global and hemispheric visual field defects, as well as global and corresponding hemispheric RNFL thickness. RESULTS: Maximal venous dilatation (r = 0.39, p < 0.02) as well as arterial (r = 0.41, p < 0.01) and venous amplitude of vessel reaction (r = 0.31, p < 0.04) were significantly correlated to overall visual field defect mean deviation (MD). Venous maximal dilatation (r = 0.47, p < 0.004), the amplitude of vessel reaction (r = 0.37, p < 0.01) and the amplitude of arterial vessel reaction (r = 0.33, p < 0.02) were significantly correlated to overall RNFL thickness. Time until maximal dilatation in the arteries was significantly correlated to the corresponding hemispheric RNFL thickness (r = - 0.43, p < 0.007). CONCLUSION: Vascular reaction parameters show significant correlations with structural and functional impairment in advanced stages of glaucoma. These findings support the hypothesis that disturbed ocular blood flow and autoregulatory mechanisms may contribute to disease severity.