Purchase this article with an account.
M. R. Lesk, J. Wang, I. A. Sigal; Clinical Visual Field Defects Correlate With Numerical Predictions of Biomechanical Strain in the Optic Nerve Head. Invest. Ophthalmol. Vis. Sci. 2009;50(13):5218.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Ocular hypoperfusion and IOP-induced mechanical strains likely play a synergistic role in glaucomatous axonal damage. Our objective was to determine if there is a relationship between strains predicted using finite element (FE) models and visual field mean defect (MD), in vasospastic and non-vasospastic patients.
Normal (n=53), ocular hypertensives (n=51), glaucoma suspects (n=51) and advanced open-angle glaucoma (n=106) patients were recruited (total n=261). IOP, axial length, central corneal thickness (CCT), corneal hysteresis (CH) and ONH surface morphology (using an HRT) were measured. Also recorded were maximum historic IOP, H24-2 MD and vasospasticity (according to a history of cold-hands). The measurements were then transformed into the input factors of an FE model (Sigal et al, IOVS 2005, 46(11)) to obtain predictions of strains for each eye. When an FE-model input factor was not directly available from the measurements it was either left at the default level (e.g. lamina cribrosa thickness), or estimated from the measurements (e.g. scleral thickness and stiffness estimated from CCT and CH respectively). Correlation analysis was carried out between the FE-predicted strains and MD for the patient population as a whole, and divided into vasospastic (n=51) and non-vasospastic (n=171; 39 unassigned).
MD correlated significantly with all modes of strain (tensile and compressive in the LC and in the pre-laminar neural tissue): R= -0.34 to -0.44 (p < 0.001). When eye-specific surrogate scleral stiffness estimates were replaced by a default value the correlations were still significant but weak (R= -0.19 to -0.31, p<0.001). When the cohort was divided into vasospastic and non-vasospastic groups the correlations were stronger in the vasospastic group: R = -0.61 to -0.70 (p<0.001) vs. R = -0.19 to -0.36 (p<0.016).
IOP-induced strains within the ONH predicted by the FE model correlated significantly with visual field defects. In all correlations higher strains were associated with increased visual field defect. The correlations were stronger in vasospastic patients and when scleral stiffness estimates were included. Although the results are encouraging, the assumptions implicit in the FE models and the transformations of clinical measurements into input factors are substantial and should be studied further.
This PDF is available to Subscribers Only