Identification of predictors and development of a screening protocol for cerebello-pontine lesions in patients presenting with audio-vestibular dysfunction

Hear me out: rethinking internal auditory meatus magnetic resonance imaging in primary care. A cohort evaluation

Background

Imaging detects acoustic neuroma, a rare pathology associated with asymmetric sensorineural hearing loss and tinnitus, that is mostly managed conservatively. Scanning indication is debatable, without evaluation in primary care, despite the high burden of audiovestibular symptoms and commissioning of general practitioner imaging.

Method

Cohort evaluation of two years' internal auditory meatus magnetic resonance imaging in primary care.

Results

Of 200 scans requested by 77 general practitioners, only 33 per cent conformed to guideline indications. Most were referred to specialists, regardless of result. Only 10.5 per cent were appropriately imaged to rule out neuroma without specialist referral. One neuroma was detected (diagnostic yield 0.5 per cent) in a patient already referred. Incidental findings were shown in 44.5 per cent, triggering low-value cascades in 18 per cent. Whilst fewer than 1 in a 1000 imaged patients may improve through surgery, 1 in 5 can suffer negative imaging cascades.

Conclusion

Considering the bi-directional relationship between distress and audio-vestibular symptoms, anxiety-provoking imaging overuse should be minimised. In low-prevalence primary care, retrocochlear imaging could be limited to those with asymmetric sensorineural hearing loss. Alternatively, assessment and imaging could be shifted to audiologist-led settings, with a wider therapeutic offer, likely more beneficial and cost-effective than conventional surgical pathways.

Development of a diagnostic model to identify patients at high risk for cerebellopontine angle lesions

Purpose

To develop a diagnostic model to identify patients at high risk of a CPA lesion.

Methods

A consecutive cohort of patients with AAD referred by a general practitioner, who underwent their first MRI examination of the CPA between 2005 and 2015 was included. Demographics, symptoms, findings during physical examination, and pure-tone audiometry were used as potential predictors. The presence of a CPA lesion was used as outcome.

Results

We analyzed data of 2,214 patients, detecting 73 CPA lesions in 69 (3.1%) patients. The final model contained eleven variables, namely gender [male] [OR 1.055 (95% CI 0.885–1.905)], sudden onset of hearing loss [OR 0.768 (95% CI 0.318–0.992)], gradual onset of hearing loss [OR 1.069 (95% CI 0.500–1.450)], unilateral tinnitus [OR 0.682 (95% CI 0.374–0.999)], complaints of unilateral aural fullness [OR 1.006 (95% CI 0.783–2.155)], instability [OR 1.006 (95% CI 0.580–2.121)], headache [OR 0.959 (95% CI 0.059–1.090)], facial numbness [OR 2.746 (95% CI 0.548–11.085)], facial nerve dysfunction during physical examination [OR 1.024 (95% CI 0.280–3.702)], and asymmetry in BC at 1 kHz [OR 1.013 (95% CI 1.000–1.027)] and 4 kHz [OR 1.008 (95% CI 1.000–1.026)].

Conclusion

The proposed diagnostic model is a first step in selecting patients with a high risk of a CPA lesion among those with AAD. It needs to be externally validated prior to its implementation in clinical practice.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00405-021-06778-6.