This retrospective study collected data by reviewing medical records of 144 consecutive patients who underwent microvascular decompression (MVD) for HFS performed by a single surgeon (Y.Z.) between November 2019 and May 2021. Two of 144 were excluded due to lack of imaging data. This study was approved by the ethical committees of the First Affiliated Hospital of Chongqing Medical University. Informed consent was obtained from all the individual participants included in the study. Our research collected detailed preoperative information, including gender, affected side, hypertension, diabetes, age of symptom onset, duration of spasms, severity grading of spasms, number of offending vessels, whether the offending vessels includes the vertebral artery. To record disease progression, we adopted a severity grading system based on Lee et al. which divided the severity of HFS into four grade: Grade I - limited to periorbital region on one side; Grade II - involving other muscle groups on the same side of the face, such as the orbicularis oris muscle or zygomaticus muscle; Grade III - affecting vision due to frequent spasms; Grade IV - affecting vision and causing bilateral facial asymmetry (6, 7). However, there is currently no unified standard for calculating the progression rate of HFS patients. The average time taken to progress one grade among 142 patients was 1.61 years/grade in our research. Therefore, we defined rapid progression group as those with a progression rate of <1.61 years/grade, and slow progression group as those with a progression rate more than 1.61 years/grade.

Patients who met the clinical presentation of HFS, had a duration longer than 3 months, excluded from secondary lesions by CT or MRI scans, and showed offending vessels on preoperative magnetic resonance tomography angiography (MRTA) were included.

Patients who had a duration shorter than 3 months, a previous diagnosis of psychiatric disorders, stroke, encephalitis, dementia, multiple sclerosis, or secondary hemifacial spasm were excluded.

All analyses were performed using SPSS 27.0 software. Continuous data were presented as mean ± standard deviation (x ± s) and were analyzed by using independent samples t-test. Categorical data were expressed as proportions (%) and were analyzed by using Pearson's chi-square test, with R × C contingency table method employed (Fisher's exact test used when the expected count was below 5 in more than 1/5 of the cells or when the expected count was <1). A correlation analysis was performed to identify the factors correlated with progression rate of facial spasm. A significance level of α = 0.05 was used for all tests. Drawing in this study was performed using GraphPad Prism8.

Within rapid progression group, 36 patients (40.0%) were male, while in slow progression group, there were 20 male patients (38.5%). Similarly, hypertension was identified in 11 patients (12.2%) in rapid progression group and in 15 patients (28.8%) in slow progression group. The occurrence of diabetes was observed in 4 patients (4.4%) in rapid progression group and 4 patients (7.7%) in slow progression group. Among rapid progression group, 41 patients (45.5%) experienced affliction on the left side. In slow progression group, 32 patients (61.5%) were affected on the left side. Specifically, in rapid progression group, 29 patients (32.2%) had offending vessels involving the vertebral artery, while in slow progression group, 21 patients (40.4%) had offending vessels involving the vertebral artery. In terms of brainstem compression in the REZ area, rapid progression group had 35 patients (38.9%) presenting brainstem compression, while slow progression group had only 5 (9.7%). In rapid progression group, 55 patients (61.1%) experienced an onset age of >45 years. In slow progression group, 21 patients (40.4%) were >45 years old when they initiated facial spasm. Furthermore, in rapid progression group, 51 patients (56.7%) had a facial nerve angle ≤ 96.50 degrees. In slow progression group, 20 patients (38.5%) had a facial nerve angle ≤ 96.50 degrees. Rapid progression group had an average of 1.40 ± 0.54 offending vessels, while slow progression group exhibited an average of 1.35 ± 0.48 offending vessels. Notably, rapid progression group exhibited positive APOE ε4 expression in 59 patients (65.6%), whereas in slow progression group, positive expression was observed in 11 patients (21.2%) (Table 1).

In the univariate analysis, there were significant statistical differences between the two groups in terms of age of onset (P = 0.021 <0.05), facial nerve angle size (P < 0.01), hypertension (P = 0.01 <0.05), presence of APOE ε4 expression (P < 0.01), and different degrees of brainstem compression in the REZ (P < 0.01). However, there were no significant statistical differences between the two groups in terms of gender, affected side, involvement of the vertebral artery, number of offending vessels, or diabetes. In multivariate logistic regression analysis result showed that patients with the age of symptom onset being >45 years was positively correlated with the progression rate (P < 0.01). Among patients with hemifacial spasm, those who expressed the APOE ε4 genotype had a fivefold higher progression rate compared to non-expressers (P < 0.01). Furthermore, more severe brainstem compression had a faster progression rate (P = 0.006 <0.05). When the facial nerve angle was ≤ 96.5°, the disease progressed faster (P < 0.01) (Table 2 and Figure 5).

Furthermore, we found no difference between the severity of facial spasms and the progression rate of the disease (t = 2.47, P = 0.12 > 0.05) (Figure 6).

Patients with an age of symptom onset >45 years had a faster progression rate. Conte et al. (15) showed that the spread latency correlated with disease duration and inversely correlated with age of onset. They explained that this result implied that age negatively impacts the course of the disease and that aged facial nerve fibers were more prone to develop spread. Jannetta et al. (16) had identified aging as a factor associated with a higher risk of elongated arteries, and brain aging, ultimately increasing the risk of neurovascular compression. Miki Ohta et al. (12) indicated that arterial changes associated with atherosclerosis in elderly patients may contribute to the development of hemifacial spasm. Based on the relevant researchers, we speculate that this may be due to two points: (1) aging facial nerves are more prone to demyelinating changes and faster spread. (2) age-related arterial stiffness leads to more severe atherosclerosis in blood vessels. Therefore, age plays a crucial role in the progression rate of hemifacial spasm.

In patients with a facial nerve angle ≤ 96.5°, the disease progressed faster. Smaller angle indicates a potentially more severe compression of the facial nerve root by offending vessel. Zhu et al. (17) observed that patients with hemifacial spasm had significantly smaller facial nerve angles compared to the unaffected side. And patients who experienced recurrence after microvascular decompression surgery had smaller facial nerve angles compared to the non-recurrence group. This can be explained by the fact that a smaller facial nerve angle increases the likelihood of vascular compression on the facial nerve in the REZ. According to the “demyelination theory” proposed as a possible mechanism for the development of hemifacial spasm (18–21), it is hypothesized that a smaller facial nerve angle leads to more severe compression of offending vessels on the facial nerve in the REZ area, that ultimately results in a faster progression of hemifacial spasm.

The APOE gene consists of 4 exons and 3 introns, totaling 3597 nucleotides. APOE has 3 alleles (APOE ε2, APOE ε3, and APOE ε4) and 6 phenotypes (ε2/ε2, ε2/ε3, ε2/ε4, ε3/ε3, ε3/ε4, and ε4/ε4) (22). Among them, APOE ε3 is the most common wild-type allele in the population. APOE ε2 and ε4 carry mutations and, with ε4 being considered a negative regulatory factor in many diseases (23, 24). Additionally, ε4 may promote the release of inflammatory factors (25) and play a role in the pathological processes of cerebrovascular and demyelinating diseases (26–28). Previous research has found that people carrying APOE ε4 have a higher incidence rate of HFS (11, 29). Research has found that APOE ε4 can accelerate the pathological process by interfering with neuronal and myelin regeneration (23, 24). Pathological changes such as demyelination, axonal loss, and Schwann cell proliferation can occur in HFS patients. We speculate that the repair of myelin in the compressed area may be disrupted in APOE ε4 gene carriers, making abnormal conduction of electrical signals more likely to occur. Moreover, APOE ε4 allele promotes the secretion of inflammatory factors and aggravates neuroinflammation. Thus, APOE ε4 allele interferes with the repair of myelin sheath in the compression area and results in electrical signals being more prone to ectopic conduction. It is reasonable to explain that patients with APOE ε4 positive expression have a faster rate of disease progression.

Through our data analysis, we found that the more severe the compression of the brainstem in the REZ area, the faster the progression rate of hemifacial spasm. Studies have found that primary hemifacial spasm is triggered by offending vessels, which lead to increased excitability of facial nerve in some focal areas through mechanical factors (indirect stretching) (30). The increased excitability can spread from one branch of nerve to another branch and result in faster excitation conduction (31). At the same time, it has been found that the facial nerve root may increase the expression of sodium ion channel Nav1.8 on the neuron membrane when it is pulled by the instrument, and the resting membrane potential also depends on sodium conductance, especially through the persistent sodium channel. Other sodium channels are also involved in the generation of action potentials (transient sodium channels). In the pressurized facial nerve in primary HFS rat model, one of the sodium ion channels, Nav1.8 channel, was found to be overexpressed, which affected the resting potential of neurons and increased the excitability of the facial nerve nucleus (30). We hypothesize that, mechanical compression by offending vessels lead to overexpression of sodium channels and result in faster conduction, and the more severe brainstem compression in the REZ region, the faster symptom progression will be.

In our data analysis, we found that gender and the affected side did not have a significant impact on the progression rate of hemifacial spasm in the two patient groups with differing rates of progression. Nurminen et al. (32) conducted a statistical analysis and reported a higher likelihood of hemifacial spasm in females, with a higher occurrence of left-sided hemifacial spasm. However, they did not provide a deeper explanation for his observations, and the relationship between hemifacial spasm and gender or affected side remains inconclusive. Some studies have suggested that postoperative complications were more common happened in cases of hemifacial spasm involving the vertebral artery in the offending vessels and those patients also exhibited a higher rate of delayed recovery after surgery (33, 34). However, there is no definitive evidence to suggest that the number of the offending vessels correlates with the severity of compression in the REZ area. Therefore, the presence of the vertebral artery in the offending vessels, or an increased number of the offending vessels, does not establish a clear relationship with the progression of hemifacial spasm. While there is established research indicating that a long history of diabetes may lead to atherosclerosis, a study by Ohta et al. (12) found no significant correlation between atherosclerosis and the occurrence of hemifacial spasm. Currently, there is also no evidence to suggest a relationship between diabetes and the incidence of hemifacial spasm.

We did not find a significant correlation between the rate of progression and the severity of hemifacial spasm (HFS) in our study. We initially thought that areas surrounding the facial nerve with faster progression may exhibit higher levels of pro-inflammatory cytokine release and slower repair of demyelination, leading to more severe symptoms in patients (29). Additionally, compression or stretching of the offending vessels on the facial nerve may not only stimulate excessive expression of the Nav1.8 channel but also cause hyperexcitability and “cross-excitation” between axons, resulting in more severe symptoms of facial spasm (30). However, our study did not find any differences between the two groups, which may be related to the limited number of cases collected.

Our study has certain limitations. Firstly, this study is a single-center retrospective and small sample study. Facial spasm in different areas might have different incidence rate and progression rate. More possible factors related to progression rate might be missed. Therefore, more factors need to be included in the future, such as smoking, drinking, preoperative carbamazepine intake history, botulinum toxin injection history etc. Secondly, some patients with long medical histories might have experienced recall bias during the data collection process. A more detailed assessment scale of progression rate and a more reasonable rating of the severity of hemifacial spasm are needed in the future research.