Purpose:
Our goal was to estimate the range of acute IOP-induced lamina cribrosa displacement (LCD) and scleral canal expansion (SCE), over populations with varying optic nerve head (ONH) characteristics.
Methods:
We generated two populations of 100,000 cases each with controlled distributions, either uniform (U) or Gaussian (G), of eye radius, scleral thickness and modulus (stiffness), lamina cribrosa radius, position and modulus, neural tissue modulus and pre-laminar tissue compressibility. The characteristics varied and their ranges were based on the literature [Sigal IA, et al. IOVS 2005;46(11) 4189-99]. Although the distributions of the ONH characteristics are still unknown, G is ostensibly more realistic because it acknowledges the intuitive idea that more eyes are midrange than extreme. The LCD and SCE induced by a 10 mmHg increase in IOP were then predicted for each case using surrogate models of published finite element models of the ONH [Sigal IA. IOVS 2009;50(6) 2785-95].
Results:
Most cases had small LCD ranging from -10 to 10 µm: 82.5% of G and 61.7% of U. Almost all had less than 10 µm SCE: 99.8% of G and 95.3% of U. LCD and SCE were distributed differently in each population. The normal region (Figure) was 79% smaller in G than in U (57 vs. 274 µm2). The response distributions had positive skewness, more so in G than in U: 0.66 vs. 0.35 for LCD, and 1.87 vs. 1.65 for SCE. The response means were lower in G than in U: mean LCD was 7.9 µm in G and 8.7 µm in U, mean SCE was 3.7 µm in G and 4.4 µm in U.
Conclusions:
The small range of LCD and SCE, especially in G, suggests that distinguishing a normal from an abnormal response is a challenge given the resolution of current imaging technologies. Further work is needed to determine if this may be possible at elevated IOPs. It also means that the response of the ONH to small increases in IOP was robust, producing similar responses despite wide variations in tissue geometry and properties (for example, a 9-fold increase in scleral stiffness). While some results seem natural, the asymmetry in the responses and differences in means would have been difficult to predict without modeling because they arise from the complex nonlinear sensitivity of LCD and SCE on the ONH characteristics.
Keywords: computational modeling • lamina cribrosa • intraocular pressure