Predictive Value of Vestibular Evoked Myogenic Potentials in the Diagnosis of Menière's Disease and Vestibular Migraine.

Abstract

To explore the usefulness of vestibular evoked myogenic potentials (VEMPs) in the diagnosis of Menière's disease (MD) and vestibular migraine (VM). Retrospective cohort. Multidisciplinary neurotology clinic. Definite MD and definite VM patients between January, 2015 and May, 2017, as well as healthy volunteers. Cervical and ocular VEMP (cVEMP and oVEMP) testing. VEMP response, amplitude, and latency. Twenty five definite MD, 34 definite VM, and 13 control subjects were analyzed. MD affected ears had significantly lower cVEMP (p = 0.007) and oVEMP (p < 0.001) amplitudes than control ears. VM ears had significantly lower oVEMP (p = 0.001), but not cVEMP (p = 0.198) amplitudes than control ears. MD affected ears, but not VM ears, had significantly more absent cVEMP (25.9% versus 0%, p = 0.005) and oVEMP responses (40.7% versus 0%, p < 0.001) than control ears. oVEMP latency was significantly shorter for both MD affected (p < 0.001) and VM ears (p < 0.001) than control ears. Sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) were calculated using a neurotology clinic MD prevalence of 7.9% in dizzy patients. A present cVEMP or oVEMP both have more than 93% chance of ruling MD out in the tested ear, while a cVEMP amplitude more than 54.9 μV or oVEMP amplitude more than 5.1 μV both have more than 94% of ruling out MD in the tested ear. Despite some overlap in VEMP results between MD and VM, when the diagnosis is uncertain between the two disorders due to symptomatic overlap and nonspecific audiometric data, VEMPs can be helpful in guiding treatment toward one disease entity or the other until more evidence points to a definitive diagnosis.