Nomogram Using Optical Coherence Tomography and Visual Field Parameters to Predict Brain Lesions in Patients with Bitemporal Hemianopia

Abstract

PURPOSE: This study aims to develop a nomogram to predict brain lesions in patients with complete or incomplete bitemporal hemianopia by combining results from optical coherence tomography (OCT) and visual field (VF) testing.

MATERIAL AND METHODS: We reviewed the medical records of patients who underwent magnetic resonance imaging (MRI) to identify brain lesions due to bitemporal hemianopia between January 2010 and March 2017, retrospectively. The patients were divided into two groups based on MRI findings: brain-lesion (+) group that had brain lesions on MRI (n = 63), and brain-lesion (-) group without brain lesions on MRI (n = 16). We compared OCT and VF findings between the two groups to find factors that could predict a brain lesion. Multiple logistic regression analysis was performed to select prognostic factors, and we constructed a nomogram to predict brain lesions on MRI.

RESULT: The VF mean deviation was lower (p = 0.011) and all sectors of peripapillary retinal nerve fiber layer thickness except the temporal region were thicker in the brain-lesion (+) group. However, there was no statistically significant difference in macular ganglion cell-inner plexiform layer between the two groups. The area under the receiver operating characteristic curve of the nomogram for predicting brain lesions on MRI was 0.916.

CONCLUSION: We developed a nomogram using VF and OCT examinations as a novel and accurate screening method to predict brain lesions in patients with bitemporal hemianopia and aid ophthalmologists and other clinicians in deciding whether to further evaluate a patient by MRI.