Evaluation of MR-Tractography Findings in Hemifacial Spasm Patients Injected with Botulinum Neurotoxin
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.344602
Source of Support: None, Conflict of Interest: None
Keywords: Axonal retrograde transport, botulinum neurotoxin, diffusion tensor imaging, hemifacial spasm, tractography
Botulinum neurotoxin (BoNT) is a potent biological toxin extracted from Clostridium Botulinum bacteria. It blocks nerve conduction between peripheral nerve endings and muscle fibers accomplished by catalyzing the proteolysis of soluble N-ethylmaleimide-sensitive factor activating protein receptor (SNARE) proteins, which provide acetylcholine release from the presynaptic region. This mechanism results in the weakening or paralysis of the skeletal muscles.
BoNT injection is mainly used for medical purposes; it is frequently used for cosmetic purposes as well. The hypothesis that frequent application of this treatment modality may also affect the central nervous system constitutes the subject of our study. Many studies and methods have shown the spread of this neurotoxin. Single-fiber EMG provided strong evidence on the dissemination of BoNT from injected area to distant muscles. BoNT cannot cross the blood–brain barrier in healthy individuals. For this reason, hematogenous dissemination to the central nervous system is not expected. The direct effect of BoNT in the central nervous system has not been demonstrated in humans. However, it is known that BoNT has a rapid retrograde spread to the anterior horn cell after intramuscular injection into the extremity muscles. It may spread to the contralateral anterior horn cell by transsynaptic pathway and show presynaptic effect in distant muscles with the anterograde spread. Considering this spread pattern of BoNT, we aimed to research the BoNT effect by comparing the FA and ADC values of the injected and non-injected sides of MRI-tractography findings in patients with hemifacial spasm after injection.
The involuntary irregular clonic and tonic movements of the facial muscles stimulated by the seventh cranial nerve are defined as “hemifacial spasm.” Spasms usually begin as a subtle eyelid fasciculation and then follow a course involving the periocular, peri-oral, and platysma muscles. We evaluated the data obtained using diffusion tensor imaging (DTI) after BoNT was injected into these muscles. DTI measurements were performed on neuroanatomical regions determined in the central nervous system based on the retrograde propagation hypothesis along the seventh cranial nerve trace. DTI is a crucial MRI technique that enables the quantitative data of brain white matter pathways and the non-invasive in vivo orientation. The general name of calculations of the construction of white matter with DTI information is tractography. MRI-tractography is the only clinically appropriate form of imaging that allows non-invasive imaging of myelin-coated white matter axons. DTI measures the apparent diffusion coefficient (ADC) and fractional anisotropy (FA) values [Figure 1]. Measure observed in the environment, where molecules move in each direction.
ADC measurement is insufficient in an anisotropic environment such as white matter where the apparent diffusion depends on the tissue alignment. In this case, the apparent diffusion tensor is defined by the tensor application with at least six or more ADC measurements. FA is a measure of the tendency of diffusion of water molecules to travel predominantly toward a particular direction. In practice, the approximation of the FA numerical value as “zero” indicates that the myelin structure deteriorates and that of the ADC numerical value as “one” indicates that the liquid medium is increasing. This study compared the ADC and FA numerical values of the injected side and the uninjected side. It was examined whether there was a significant difference.
BoNT is widely used in hyperkinetic movement disorders such as hemifacial spasm and blepharospasm and for cosmetic purposes. In this study, we aimed to demonstrate the possible central effects of BoNT in hemifacial spasm patients.
Patients diagnosed clinically as “idiopathic hemifacial spasm” and treated with Botox® at the University of Kocaeli, Department of Neurology and Movement Disorders Clinic between 2010 and 2017 were included in this study. MRI-tractography imaging plans were made after the written approval was admitted. Patients injected on both sides of the face, with hemifacial spasms due to structural causes, and with a contraindication to MRI were excluded from the study.
The number of injections of BoNT per patient ranges from 2 to 30. Patients were divided into two groups of 2–10 injections and 11 or more [Figure 2].
The essential clinical characteristics of the patients, the side on which the injection was made, and the total number of injections of BoNT were recorded. Total dose amounts were not used in the study as they could not be obtained from the records in some patients in terms of the International Unit (IU). The exit point of the facial nerve from the brain stem, motor cortex, centrum semiovale, and internal capsule was determined as measurement points associated with the facial nerve, and bilateral measurements of these regions were performed. Both ADC and FA values of each region were obtained.
This study was approved by the Clinical Research Ethics Committee of Kocaeli, Faculty of Medicine.
MRI data analysis
Imaging was performed using the 3T Philips MRI unit with the 16-channel head coil while the patient was supine (Philips Achieva Intera Release Eindhoven, Netherlands, Philips Medical Systems Achieva 1 2008-07-18 Release, Software 22).
Diffusion tensor images were taken on the axial plane. The imaging parameters are indicated in [Figure 3].
Postprocessing of the acquired images was performed on the 3T Philips workstation (Release 126.96.36.199 2007-12-03, Philips Medical Systems, Netherlands BV).
DTI and tractography images were processed on the workstation. FA and ADC values obtained from the corresponding regions were recorded.
Statistical analysis was performed using IBM SPSS 20.0 (SPSS Inc., Chicago, IL, USA). The normality test was assessed using the Kolmogorov–Smirnov test. Numerical variables with normal distribution were calculated and studied as mean ± standard deviation, numerical variables without normal distribution as median (25th–75th percentile), and categorical variables as frequency (percentages). Differences between groups were determined by Student's t test for numerical variables with normal distribution and Mann–Whitney U test for numerical variables without normal distribution. Differences between recurrent surveys were analyzed by Wilcoxon's t test. P < 0.05 was considered statistically significant.
Seventeen female and 13 male patients aged between 36 and 71 years (mean: 57.5 ± 9.1) treated with BoNT at least twice on one side of their face were included in the study. Eight of 17 female and four of 13 male patients to the right side and nine of 17 female and nine of 13 male patients to the left side were injected with BoNT. The number of injections applied to female patients ranged from 2 to 23, with a mean number of 9.8 (±6.6), whereas that applied to male patients ranged from 2 to 30, with an average number of 8.2 (±7.3). Naming was used for ease of identification. The left motor cortex was named as Fiber 1, right motor cortex as Fiber 2, left centrum semiovale as Fiber 3, right centrum semiovale as Fiber 4, left internal capsule as Fiber 5, right internal capsule as Fiber 6, left exit point of the facial nerve from the brain stem as Fiber 7, and right exit point of the facial nerve from the brain stem as Fiber 8.
Bilateral FA and ADC values of right-sided (n = 12) and left-sided (n = 18) BoNT injected patients obtained from the motor cortex, centrum semiovale, internal capsule, and exit point of the facial nerve from the brain stem were compared respectively with the paired t test. A significant difference was found for the ADC values between the right motor cortex and the left motor cortex in right-sided BoNT injected patients (P = 0.005). FA values between the right motor cortex and left motor cortex (P = 0.014) and ADC values between the right internal capsule and the left internal capsule (P = 0.012) in left-sided BoNT injected patients were statistically significant [Figure 4] and [Figure 5].
We also investigated the correlation between BoNT injection frequencies and FA-ADC values. The data were analyzed by dividing the total number of injections per person (2–10 and 11 or more) into two groups as injection frequencies. There was no significant difference between the groups in the examined areas. However, it was determined that the ADC values of the left internal capsule area were positively correlated with the frequency of injections without dividing into two groups [Figure 6].
We started on the hypothesis that Bont's central effect may be through retrograde transport of the toxin. The axonal retrograde transport of BoNT has been demonstrated in past studies. In an in vitro rat hippocampal neuron culture study, the heavy chain of BoNT and its carrier have been shown to flow from the peripheral nerve end to the soma. The importance of increased presynaptic activity has been pointed out. In another study, BoNT marked with radioactive “Technetium 99m” was injected into rat detrusor muscle. Then, radiolabeled BoNT A was observed for up to 6 weeks in the dorsal root ganglia. Significant findings were obtained by retrograde propagation in another research, where BoNT-A was injected into the muscle, which provides the mustache movement in the rat. Three days later, marked BoNT-associated SNAP-25 molecules were detected by Western blot analysis in the microdissection of the ipsilateral facial nerve nucleus. One of the critical findings of this study is that these molecules are not detected in non-neuronal-related regions. This finding indicates that the toxin does not spread passively.
The amount of injected BoNT doses is suggested to be effective on retrograde transport. However, the use of picogram-level toxin in animal experiments supports the retrograde spread of nanogram-level toxin in clinical conditions such as torticollis and dystonia. For this reason, one of the restrictions of our study may be that the dose amounts were not recorded, which may make us suppose that this effect is negligible. In contrast, a positive correlation between injection frequency and the ADC values in the left internal capsule may indicate that the dose amounts have an important role in retrograde transport.
We found that the left motor cortex of the patients with BoNT injection into the right side of the face and the right motor cortex of the patients with BoNT injection into the left side of the face have more pathological FA and ADC values. However, contrary to this negative effect found in our study, similar studies did not reach a similar result. Patients with writer cramps and BoNT injections were studied in a PET study in which the effect of BoNT on cortical motor fields was investigated. Writer cramp symptoms were reduced by BoNT treatment, and PET activity increased in the parietal cortex and caudal supplementary motor area. However, this treatment did not affect impaired activation of the primary motor cortex and premotor cortex.
FA and ADC values of the internal capsule areas were compared in the patient group in which BoNT was injected to the left side of the face, and the mean ADC value of the left side was found to be closer to the pathologic level than the right. Unlike the previous findings, the pathologic finding was found on the side where the BoNT was injected. Neuroanatomy of the facial nerve may be an important factor if the presence of axonal retrograde transport of toxin throughout the facial nerve tract is considered. The motor efferents of the facial nerve originate from the motor nucleus located in the cauda of the pons. The neurons that innervate the muscles of the upper half of the face come out of the rostral part of the nucleus, and the supranuclear innervation of this area is bilateral. However, the innervation of the facial muscles in the lower half of the face is from the caudal part of the nucleus, and the innervation is unilateral. Because BoNT injection is applied unilaterally to our patients' lower and upper muscles, it is not possible to specify to which side the toxin is likely to go through.
Central effects of BoNT studies by MRI-tractography are very limited. In a study, FA values were assessed by tractography after BoNT injection in patients with cerebral palsy. It has been pointed out that FA indicates the degree of integrity and stability of functional connection and white matter. Two groups were randomized after 6 months. FA values were compared only after physiotherapy in the first group and after physiotherapy plus BoNT injection in the second group. Both groups showed a significant increase in FA values of both motor and sensory fibers according to baseline imaging. This result was interpreted as significant in terms of post-treatment plasticity. However, there was no difference between the first group and the second group in comparison with the FA values. In MR-tractography studies in which dystonia was assessed and mean diffusivity (MD) values were assumed to be equivalent to FA, and ADC were investigated, dystonic individuals had more pathological FA and MD values than normal individuals. However, the result of Blood et al.'s  study that the pathological FA and MD values were not observed after the treatment of BoNT and that the asymmetry disappears when compared with the other hemisphere does not coincide with the pathological values obtained after the BoNT injection in our study. The fact that the results we obtained are relative and do not have pre-work values may have led to the inability to achieve results in the same direction as previous studies. It is thought that if all these studies, including ours, reach a consistent negative or positive result, it will contribute to revealing the central effect of BoNT.
DTI is a highly sensitive method of showing the microstructural structure of white matter modulated by neural activity. Characteristics of this structure include activity changes dependent on collateral sprouting, myelination changes, activity-dependent changes in fast axonal transport, and activity-dependent microstructural changes such as cell swelling. FA and ADC values measured after BoNT injection reached normal limits, suggesting that secondary toxin changes are more compatible with reversible pathological processes than degenerative processes. Toxin-dependent changes are more compatible with reversible pathological processes than degenerative processes. They may explain the observation that FA and ADC values measured after BoNT injection are within the previous studies' normal limits.
In this study, we compared right and left neuroanatomical regions in hemifacial spasm, a lateralized disorder. We hypothesized that the possible retrograde transport neurotoxin molecules could act opposite as the administration was performed unilaterally. Therefore, the opposite region should not be affected. However, the fact that the upper facial region has bilateral innervation is a limitation. Also, the lack of baseline MR tractography before Bont injection is perhaps the biggest limitation. To study the effect, two groups were formed in terms of injection frequency (2–10 times and 11 and above). However, it may be a more accurate method to use the dose amount in International Units (IU) instead of the injection frequency. It also contributes to observing the cumulative dose effect. Total dose amounts (IU) were not used in the study as they could not be obtained from the records in some patients.
Botulinum neurotoxin administration in patients with hemifacial spasms may cause some changes in the central nervous system as well as peripheral effects. It is unclear whether this effect is direct as axonal retrograde transport of the toxin, as demonstrated in animal experiments, or indirectly as blockade of the afferent fiber-mediated conduction. We need more detailed studies to support the theory that the central effect is caused by the retrograde transport of toxins and to evaluate the effects of BoNT doses on central motor pathways.
Declaration of patient consent
Full and detailed consent from the patient/guardian has been taken. The patient's identity has been adequately anonymized. If anything related to the patient's identity is shown, adequate consent has been taken from the patient/relative/guardian. The journal will not be responsible for any medico-legal issues arising out of issues related to the patient's identity.
The main author would like to acknowledge Mr. Barış Özgür for his assistance with this research.
Financial support and sponsorship
This study was supported by the Scientific Research Projects Board of the University of Kocaeli.
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]