Adhesion Failures Determine the Pattern of Choroidal Neovascularization in the Eye: A Computer Simulation Study
Figure 4
Sub-RPE CNV dependence on adhesion.
3D plot of the regression-inferred average MW using 10 simulation replicas for each adhesion scenario in the 3D parameter space obtained by setting RRp = RRl and RBp = RBl. The average MW shows the stalk cell locus even when CNV fails to initiate, so a region prone to ET1 CNV develops ET1 CNV only if CNV initiates. Red corresponds to MW = 1 and purple corresponds to MW = 0. The black region at the top-left corner indicates the locus of normal adhesion. MW = 1.0 for RPE-RPE junctional adhesion normal, RPE-BrM junctional adhesion severely impaired (weak) and RPE-POS labile adhesion normal. The three isosurfaces correspond to MW = 0.25 (back), 0.5 (middle) and 0.90 (front). The five adhesion parameters and their (multi)linear combinations account for 93% of the observed variance in average MW for all 108 adhesion scenarios (adjusted R2 = 0.89). Severe impairment of RPE-POS labile adhesion greatly reduces the MW, so ET1 CNV can only occur when RPE-POS labile adhesion is near normal. Scenarios with severe impairment of RPE-BrM junctional adhesion (RBp = RBl = 1), and normal RPE-POS labile adhesion are prone to ET1 CNV for a wide range of RPE-RPE junctional adhesion impairment (MW>0.95 for RRp = RRl>1.5). The red region with MW>0.9 has Pinit>0.8. We have rotated the axes from their orientation in Figure 3 to show the regions in the parameter space prone to ET1 CNV. To show the structure of the isosurfaces, we have rotated the axes relative to Figure 3.