Optical coherence tomography shape analysis of the peripapillary retinal pigment epithelium layer in presumed optic nerve sheath meningiomas

Patrick Sibony,Matthew Strachovsky,Robert Honkanen,Mark J Kupersmith
DOI: https://doi.org/10.1097/WNO.0000000000000107
Abstract:Background: Geometric morphometrics (GM) was used to compare the shape of the peripapillary retinal pigment epithelium-Bruch's membrane (ppRPE) layer imaged on spectral domain optical coherence tomography (SD-OCT) of patients with presumed optic nerve sheath meningiomas (pONSM) and normal subjects. Methods: We compared 2 groups: 30 normals to 10 patients (11 eyes) with pONSM. We digitized 20 equidistant semi-landmarks on OCT images of the ppRPE-layer, spanning 2500 μm on each side of the neural canal opening (NCO). Data were analyzed using standard GM techniques including a generalized least squares Procrustes superimposition, principal component analysis (PCA), thin-plate spline, and permutation statistical analysis to evaluate differences in shape. We also analyzed other variables with respect to shape including tumor size-proximity to the globe, age, retinal nerve fiber layer, and optic disc height. Results: All pONSM patients were female (age 37-66 years); 10 had unilateral and 1 had bilateral optic nerve involvement. Ten of the eyes had optic disc edema at presentation, 4 went on to develop shunt vessels, and 4 had optic atrophy. The ppRPE-layer bordering the NCO in normals is V-shaped pointing away from the vitreous; the ppRPE-layer in pONSM is indented causing an inverted-U shaped deformation skewed nasally toward the vitreous. PCA showed a significant difference between normals and pONSM (permutation, n = 10,000, P = 0.001). The size and proximity of the tumor to the globe correlates with the shape of the ppRPE-layer (r = 0.75, P = 0.04). Correlation between shape variables and RNFL thickening (r = 0.51), optic disc height (r = 0.67), and age (r = 0.67) were not statistically significant. Conclusion: The shape of the RPE layer in pONSM is characterized by an inverted-U shape or indentation that differs significantly from normals. It is indistinguishable from the shape we previously reported in papilledema and is not caused by disc edema. The mechanism in pONSM is unknown but may involve a change in the compliance of the nerve and/or localized sequestration of cerebrospinal fluid in the distal optic nerve sheath.
What problem does this paper attempt to address?