Prognostic Value of Early Magnetic Resonance Imaging Patterns in Sudden Hearing Loss

Introduction: Sudden sensorineural hearing loss (SSHL) is a relatively frequent disease, but a sensitive marker or a reliable test to identify the underlying cause is still unavailable. Neuroradiology appears to offer the most promising tools, especially magnetic resonance imaging (MRI). In a recent study from our group, we explored the ability of MRI to detect subtle changes in the inner ear compartments by means of a 3D-fluid-attenuated inversion recovery sequence, aiming at identifying 3 distinct MRI patterns (haemorrhagic, inflammatory, brain-labyrinth barrier breakdown). In the present study, we contrasted the MRI patterns at onset with relevant prognostic factors, with the audiological features of each patient’s SSHL and with treatment outcomes. Methods: In this retrospective study, we enrolled 50 adult subjects (54.61 ± 18.26 years) with SSHL. They underwent an MRI within 72 h from admission, and 5 audiological evaluations: at admission, on the 5th day after the start of medical therapy, at the end of the first cycle of hyperbaric oxygen therapy, then 1 and 6 months later. Results: Abnormalities of the MRI signal and/or post-contrast enhancement asymmetry of the cochlea (“pattern+ MRI”) correlated with worse audiological outcomes at 1 month, but the different MRI patterns were not correlated with any specific prognostic model, despite rigid protocol settings. However, a significant difference was found for low-tone SSHL, which were always “pattern” negative at MRI (p = 0.01), and for profound SSHL which demonstrated a pattern+ MRI in 80% (p = 0.04). At the onset of SSHL, a pattern+ MRI was found in 29/50 cases (58.0%) and was related with lesser degree of recovery of pure-tone average at 1 month and lesser chance to retain the hearing threshold benefit in the long term. Given the limited numbers of patients enrolled so far, the relative impact of comorbidities on each MRI pattern remains uncertain. At 6 months, we observed a trend of greater and more stable recovery (p = 0.023) and less frequent recurrence of SSHL in patients with a normal MRI. Conclusions: The 3 observed MRI patterns did not correlate consistently with specific audio-vestibular features or any peculiar aspect of the patient’s clinical history. Larger series of patients with SSHL are needed, possibly from multicentric studies.