Acute facial nerve palsy is one of the most commonly encountered cranial neuropathies characterized by varying degrees of facial muscle paralysis. Bell’s palsy, an acute facial neuropathy with idiopathic origins, is the most common etiology of acute facial nerve palsy. The symptoms can begin suddenly and worsen over a few hours to days, varying from mild difficulty in making facial muscle movements and facial expressions to inability to close the eye or mouth or even social and psychological distress [1]. The neurologic and psychological symptoms are more prominent when the patient cannot voluntarily move their facial muscles.

Numerous studies have evaluated the degree of nerve degeneration in facial nerve palsy using electroneuronography (ENoG), introduced by Esslen [2] in 1973. The nerve excitability test, maximal stimulation test, ENoG, electromyography, and blink reflex are special tests that can assess the electrophysiological extent of facial nerve damage. In particular, ENoG, by determining action potential amplitude, provides the degree of nerve fiber degeneration with great accuracy [3].

Since it takes about 72 hours for Wallerian degeneration to reach the extratemporal nerve distal to the stylomastoid foramen, which is the area for electrostimulation for ENoG in idiopathic facial palsy, ENoG can be normal in the first 72 hours after symptom onset. The response to ENoG is quickly lost during the next 5–6 days and would not change after 21 days when the nerve degeneration is complete [3,4]. This means that performing ENoG within 14 days after symptom onset may be too early for the test to reflect the full effect of the neurodegeneration. Thus, it seems necessary to determine the predictive usefulness of ENoG performed around 21 days after symptom onset, which is slightly later than the timeframe in traditional studies.

Several factors appear to be related to the prognosis of facial palsy, such as age, sex, and severity of paralysis, and the degree of nerve degeneration obtained by ENoG is the most reliable prognostic factor [5]. However, prior studies on ENoG findings as a prognostic factor did not consider the clinical severity of the paralysis [6,7]. Since facial palsies with more severe axonal damage are associated with longer treatment times and greater variance in treatment results, setting an appropriate treatment duration according to each patient’s treatment potential is necessary for severe cases of facial palsy.

This study aimed to evaluate the cutoff values of electroneuronographic data obtained within 2–4 weeks after symptom onset in severe facial palsy cases to provide an objective basis for predicting functional presentation within 2 months post-symptom onset.

It is established that ENoG amplitude has a prognostic value in facial palsy. Our study showed that ENoG amplitude also has predictive value in severe facial palsy. In our study, statistically significant differences were found between patients with severe facial palsy in the good and poor recovery groups concerning the facial nerve degeneration rate of each of the four facial muscles (frontalis, nasalis, orbicularis oculi, and orbicularis oris) and their mean values and trimmed means (p<0.005).

The ENoG amplitudes of the four facial muscles showed significant differences between the two groups; however, only that of the nasalis muscle had prognostic value for severe facial palsy. This finding is concurrent with those of several other studies. Kim et al. [15] studied the prognostic value of ENoG of Bell’s palsy at the orbicularis oculi and nasolabial fold in 81 patients and found that the ENoG of the nasolabial fold had more prognostic value than that of the orbicularis oculi. The facial nerve has five extracranial terminal branches: temporal, zygomatic, buccal, marginal mandibular, and cervical. According to most literature and anatomy textbooks, the frontalis muscle is innervated by the temporal branch, the orbicularis oculi muscle is double-innervated by the temporal and zygomatic branches, the buccal branch innervates the nasalis muscle, and the orbicularis oris muscle is double-innervated by the buccal and mandibular branches [16,17]. In our study, we adopted the muscles representative of each branch in standardized settings; temporal-frontalis, zygomatic-orbicularis oculi, buccal-nasalis, and mandibular-orbicularis oris. Martínez Pascual et al. [18] studied the extracranial course of the facial nerve in 23 Caucasian adult cadavers and found that the buccal branch has the most varied origin. It can be inferred that muscles located in the central part of the face are more significant since they are dominated by the facial nerve plexus [15]. Since ENoG amplitude reflects nerve fibers reacting to electrical stimulation and their synchronicity [19], the ENoG of muscles that are mostly innervated by the facial nerve plexus, such as central muscles of the face, e.g., the nasalis (or nasolabial fold as in Kim et al. [15]), is more significant than that of their counterparts for predicting recovery from facial nerve insult. Similarly, Engström et al. [20] reported that initial (mean, day 11) ENoG measurements of the nasalis and/or mentalis were the best single recordings for predicting favorable outcomes in patients with Bell’s palsy.

Sex has a predictive value for severe facial palsy. Although it seems controversial whether sex is a risk factor for facial palsy, several studies have suggested a male predominance in Bell’s palsy. According to the National Health Insurance Service National Sample Cohort data of Korea from 2006 to 2015, the male sex is a significant risk factor for Bell’s palsy [21]. Another Korean population-based study reported similar epidemiologic results [22]. In our study, the male sex appeared to have a slight predominance in the overall incidence of severe facial palsy (64.9%). Male patients exhibited better prognoses in the 2-month post-onset period in another previous Korean study. The retrospective cohort study, which aimed to establish whether clinical prognostic factors differ according to H-B grade, found that among patients with severe facial palsy (initial H-B grades, 5–6), male had a more favorable outcome than female (odds ratio, 1.12) [23].

Many prior studies have shown that a <90% degeneration rate of the affected facial nerve has a favorable outcome [5-7]. Our ROC analysis showed that the best cutoff values of preserved ENoG amplitude (%) for sensitivity and specificity were 72.0%, 92.5%, 78.5%, and 89.5% in the frontalis, nasalis, orbicularis oculi, and orbicularis oris muscles, respectively. The nasalis and orbicularis oris muscles displayed similar outcomes to those in previous studies, whereas the frontalis and orbicularis oculi muscles showed stricter cutoff values (72.0% and 78.5%, respectively). It seems that the more severe the nerve damage, the more variable the regenerative potential for each branch of the facial nerve, resulting in heterogeneous regeneration patterns. Since prior studies did not consider facial palsy severity, a 90% prognostic threshold cannot be uniformly applied to severe facial palsy.

The pathophysiology of idiopathic facial palsy remains unclear; it is considered that an inflammatory mechanism is a major contributor to the disease pathogenesis. Cayir et al. [11] reported that higher CAR and NLR levels were associated with poor prognoses in patients with Bell’s palsy. Our study did not find significant differences in laboratory data (CAR, NLR, and WBC count) between the two groups in the univariate analysis or logistic regression model. These inflammatory markers can identify levels of inflammation at the acute phase; however, their responses to strong anti-inflammatory medications (such as corticosteroids) vary among individuals. Additionally, there were differences in the time when anti-inflammatory medication was administered, the time when the blood test was performed, and the pathophysiology of the patients. Eliminating this variable response rate could yield a better probability of predicting disease progression; this may be achieved by controlling the blood sampling time before treatment.

This study has some limitations. First, in excluding patients who underwent electrodiagnosis other than 2–4 weeks from symptom onset, 66 patients (64.1%) with missing values were detected, resulting in bias. However, Wallerian degeneration in the peripheral facial nerve takes 72 h after symptom onset to reflect in electrodiagnostic tests and approximately 21 days for complete degeneration [3,4]. Thus, a strength of this study is the inclusion of only electrodiagnostic data collected after this period (at least 2 weeks after the onset of symptoms). Second, the laboratory data of five patients (17.2%) in the good recovery group were unavailable, and the mean imputation method was used to obtain the missing values. Third, the follow-up period was limited to 2 months. In most cases of facial palsy, the complete recovery rate reaches a plateau after 3 months from onset [24]. However, in some cases, recovery occurs after 1 year; therefore, a longer follow-up period would have yielded ENoG amplitude values that can predict the final recovery state.

In conclusion, this study suggests that the female sex and lower ENoG of the nasalis muscle acquired 2–4 weeks from symptom onset have negative prognostic values for the 2-month functional outcome of severe facial palsy.