[Visualization of endolymphatic hydrops in 3D-FLAIR MRI after intratympanic Gd-DTPA administration in Meniere's disease patients].

Objective: To study the feasibility of endolymphatic visualization and the diagnosis of Meniere's disease by applying intratympanic gadolinium administration through the tympanic membrance and three-dimensional fluid-attenuated inversion recovery magnetic resonance imaging (3D-FLAIR MRI). To study the relationship between the endolymphatic hydrops visualized by MRI and audio-vestibular functional tests, such as pure tone audiometry (PTA), electrocochleography (EcoG), caloric test and vestibular evoked myogenic potential (VEMP).

Methods: With a three Tesla magnetic resonance imaging (MRI) unit, 3D-FLAIR imaging was performed 24 hours after intratympanic gadolinium through the tympanic membrance in 32 patients with clinically diagnosed unilateral Meniere's Disease. We visualized the enhanced imaging of perilymphatic space in bilateral cochlea, vestibular and (or) canal, scoring scala tympani and scala vestibule of bilateral cochlear basal turn respectively and measuring the developing area of bilateral vestibule and the signal intensity ratio (SIR) between the vestibule and the brain stem subjectively. PTA, EcoG, caloric test and VEMP were performed. The relationship between the endolymphatic hydrops visualized by MRI and audio-vestibular functional tests were studied.

Results: The gadolinium appeared in almost all parts of the perilymph in cochlea, vestibular and (or) canals in all 32 patients' inner ears, so the endolymphatic space was clearly shown on 3D-FLAIR imaging. The scala vestibuli score value between the affected side and the healthy side were statistically significant (Z = 4.309, P < 0.05) . The developing vestibular area between the affected side and the healthy side [(6.04 ± 2.89) mm(2), (8.28 ± 3.04)mm(2)] were statistically significant (t = 3.322, P < 0.05) . Abnormal vestibular evoked myogenic potentials were significantly correlated with the developing vestibular area of the affected side (F = 11.96, P < 0.05) . Abnormal electrocochleography were significantly correlated with scala vestibuli score value of cochlear basal turn in the affected side (Z = 3.17, P < 0.05) . No significant correlation was found between the scala vestibuli score value or the developing vestibular area and caloric test or PTA findings.

Conclusions: 3D-FLAIR MRI with intratympanic gadolinium injection through the tympanic membrance can discriminate the border between the perilymph and the endolymph and show endolymphatic hydrops. This method may provide radiographic reference for the diagnosis of Meniere's disease. The results of VEMP and electrocochleography might have appropriate correlation with degree of vestibular and cochlear hydrops.