Background: To report a linear risk score obtained using clock-hour optical coherence tomography (OCT) data from papilledema and pseudopapilledema nerves that differentiates between the 2 diagnoses with high sensitivity and specificity.
Methods: Patients presenting to a single neuro-ophthalmologist with papilledema or pseudopapilledema were included for a retrospective review. The absolute consecutive difference in OCT retinal nerve fiber layer (RNFL) thickness between adjacent clock hours and the mean magnitude of thickness for clock hours 1-12 were compared between the 2 groups using mixed-effect models adjusting for age and clock hour with a random intercept for subjects and eyes (nested within subject). The area under the curve (AUC) for the receiver operating characteristics curve and a separate calibration curve was used to evaluate potential clinical usage.
Results: Forty-four eyes with papilledema and 72 eyes with pseudopapilledema, 36 of whom had optic nerve drusen met criteria. The papilledema group had a higher mean RNFL thickness (papilledema = 163 ± 68 µm, pseudopapilledema = 82 ± 22 µm, P < 0.001). The papilledema groups also had more variability between consecutive clock hours (papilledema = 57 ± 20 µm, pseudopapilledema = 26 ± 11 µm, P < 0.001). A linear combination of each patient's averaged values separated the 2 groups with an AUC of 98.4% (95% CI 95.5%-100%) with an optimized sensitivity of 88.9% and specificity of 95.5% as well as good calibration (mean absolute error = 0.015).
Conclusions: Patients with papilledema have higher intrinsic variability and magnitude within their OCT, and this finding reliably distinguishes them from those with pseudopapilledema.
Copyright © 2020 by North American Neuro-Ophthalmology Society.