Exploratory study on factors associated with poor prognosis despite positive stapedial reflex in peripheral facial...

The stapedial reflex (SR) is widely used to predict prognosis in patients with viral peripheral facial nerve palsy (VPFNP), including Bell's palsy, zoster sine herpete (ZSH), and Ramsay Hunt syndrome (Hunt syndrome). While SR positivity is generally associated with cure, it does not guarantee this outcome in all patients. This study aimed to identify predictors of poor prognosis in SR-positive VPFNP patients, focusing on clinical data obtainable during the early stage of the disease.

A retrospective exploratory study was conducted on 230 SR-positive VPFNP patients treated at a tertiary hospital between 2013 and 2022. Prognostic factors included clinical information at the initial visit: pain, the severity of facial palsy assessed using the Yanagihara facial grading system (Y-system), diagnosis (Bell's palsy/ZSH or Hunt syndrome), clinical presentation (typical or atypical), and the timing of SR testing. Typical presentations were defined as cases with a clear onset date, isolated facial nerve palsy, and no cranial nerve involvement beyond the vestibulocochlear nerve. Atypical cases included unclear onset date, recurrent palsy, or additional cranial and other nerve deficits. Cure was defined as a score of Yanagihara facial grading system (Y-score) ≥ 36 within six months of onset, while non-cure was defined as Y-score < 36. Electroneuronography (ENoG) and systemic comorbidities, such as diabetes mellitus (DM) and white matter lesions (WMLs) detected on MRI, which represent systemic microvascular insufficiency, were also evaluated.

Of the 230 patients, 222 (96.5%) achieved cure, while 8 (3.5%) were classified as non-cured. Significant predictors of poor prognosis included pain at the initial visit (odds ratio [OR]: ∞; 95% confidence interval [CI]: 5.21-∞), an initial Y-score ≤ 10 (OR: 8.25; 95% CI: 2.25-29.35), and atypical clinical presentations (OR: 9.83; 95% CI: 1.56-54.44). Importantly, all patients without pain were cured. ENoG score ≤ 10% strongly predicted poor outcomes (OR: 255; 95% CI: 32.51-1548), although ENoG should be conducted seven to ten days after onset for optimal accuracy. The timing of SR testing did not significantly affect prognostic accuracy, and systemic comorbidities, such as DM and WMLs, showed no significant association with prognosis.

This study highlights that, despite SR positivity, VPFNP patients with pain, severe palsy (Y-score ≤ 10), or atypical presentations are at higher risk for non-cure. Combining SR results with early-stage clinical indicators based on simple patient evaluation enables accurate prognostic predictions, improving patient counseling and treatment planning.