Asymmetric epileptic spasms after corpus callosotomy in children with West syndrome may be a good indicator for unilateral epileptic focus and subsequent resective surgery

2.1 Selection of patients

This study included 16 patients with medically refractory ES who were successfully treated with a combination of CC and subsequent resective surgery in the National Nagasaki Medical Center from 2006 to 2021. Resective surgeries also included multilobar disconnective surgeries and hemispherotomy.¹² Clinical details are summarized in Table 1. Their age at epilepsy onset ranged from 2 to 22 months (median 5.5 months). Patients initially diagnosed with West syndrome and treated with appropriate medical treatments including adrenocorticotropic hormone and conventional antiseizure medicines, were referred to our center due to the inefficacy of those treatments. One patient had developed tonic seizures in addition to ES, and both seizures caused sudden head drop or fall (case 3). A preoperative comprehensive evaluation including video-electroencephalography (EEG) monitoring, MRI, PET and/or single-photon emission computed tomography (SPECT), and developmental and cognitive tests was done. Three patients had a specific MRI lesion (cases 2, 6, and 7), which comprised an indiscrete abnormality such as multilobar pachygyria with the blurring of the cortical-subcortical boundary in two patients (cases 2 and 7) and a transmantle sign in the white matter in one (case 6). Additionally, seven patients had unilateral (cases 10–15) or bilateral (case 8) nonspecific lesions, including mild lobar or hemispheric volume loss and subtle signal changes in the white matter. EEG and PET/SPECT also showed laterality in some patients. Although these might be unilateral lesions suggesting a seizure focus, CC was firstly performed because a resectable epileptic focus was unidentified by ictal behavior and EEG.^{10, 12}

Furthermore, ES ceased or reduced in frequency immediately after CC but recurred during the follow-up period. One patient with tonic seizures had only ES after CC (case 3). Therefore, medical treatments were modified in some patients, but ES could not be controlled. Long-term video-EEG monitoring was repeated as scheduled and showed asymmetric ES and lateralized epileptiform activities, suggesting explicitly localized or lateralized epileptic focus. Additionally, hypometabolism on PET and/or hyperperfusion on ictal SPECT coincided with the responsible region or hemisphere for those epileptic activities on EEG. Based on those findings and neurological conditions, resective or disconnective surgery was performed to control ES. Frontal, frontotemporal, or parietal lobe surgery was performed in six cases based on predominant epileptic activities in those lobes. Posterior quadrantectomy¹⁸ was performed in two cases based on predominant epileptic activities in the posterior cortex, and hemispherotomy was performed in five cases who developed diffuse hemispheric epileptic activities with contralateral hemiparesis. Multilobar disconnective surgery preserving the central cortex (subtotal hemispherotomy)⁵ was performed in three cases who developed diffuse hemispheric epileptic activities without hemiparesis. An interval from CC to subsequent resective surgery ranged from 5 to 91 months (median 14.5 months). No unexpected postoperative complications occurred except those predicted before surgery such as hemiparesis and hemianopia in cases with multilobar disconnective surgeries and hemispherotomy. After surgery, all the patients accomplished freedom from ES through antiseizure medicines, indicating that the responsible hemisphere for their residual ES after CC was precisely identified.

2.2 Description of ES before and after CC

Video-EEG monitoring was performed repeatedly for at least 12 hours in each patient. This study analyzed video-EEG data collected just before CC and resective surgery. The EEG was recorded using NicVue (Natus Medical Incorporated) with electrodes placed according to the international 10–20 system. In some patients, electromyography (EMG) was simultaneously referred from the bilateral deltoid muscle.

To investigate the correlation between clinical expression of ES and the responsible hemisphere, we visually reviewed the recorded video-EEG. Whereas we analyzed electroclinical ES with ictal behaviors on video-EEG or EMG. An independent reviewer (DU), who was blinded to the patients' treatment, documented the behavioral ES without any clinical information about those patients. If the behaviors could not be evaluated due to poor recording resolution or patient's posture, eg lateral position on the bed, those were scored as “unknown” and excluded from the analysis. When ictal movements were subtle and laterality was difficult to assess, senior physicians (TO, RH, and YW) evaluated the characteristic of the ES and a consensus decision of all three senior physicians was taken into consideration. If even one of the three did not agree, those ES were excluded from the analysis. In particular, NF, MCU, and MCL were described by focusing on their laterality, ie “symmetric” or “asymmetric,” and “left” or “right.” We counted the number of NF, MCU, and MCL, but all those expressions were not always accompanied simultaneously in individual ES. For example, some seizures comprised only NF, while others could only display MCU and/or MCL. Therefore, the number of individual ictal movements is not the same even in the same patient.

Asymmetric NF was characterized by neck flexion with the head turning or falling to one side (Figures 1 and 2). Asymmetric MCU and MCL were characterized by the unilaterally predominant muscular contractions involving the shoulder, arm, and leg (Figures 1 and 2). Supplemental video data are also available online for those ictal movements (Video S1 and S2). The spasms characteristics vary depending on the muscle groups' involvement and body postures.² In fact, individual or group muscular contractions could not always be documented in detail. Usually, when analyzed by video, spasms captured in the front sitting or supine position were the easiest to score the symmetry/asymmetry of the movements. In the upper extremity, the abduction or flexion (forward or lateral movement) of the shoulder joint and flexion of the elbow joint were most frequently seen. In the lower extremity, the hip and knee joints were usually flexed. Thus, most of the analyzed spasms were flexor spasms. However, we observed some spasms accompanied movements in which the shoulder joints were flexed while the elbow joints were kept extended. In such cases, it might be flexor + extensor spasms, but it is impossible to make a clear classification.

ES was also confirmed by correlates of EEG including high voltage positive slow-wave, fast activity, and/or decremental background attenuation^{19, 20} (Figures 1 and 2). However, the laterality of those ictal correlates was not always documented in this study. We counted the number of all electroclinical ES in each patient. However, subtle ES including motionless staring²⁰ and rapid eye movements were excluded from the observation in this study since those are difficult to be described precisely due to the limited resolution of our video data. In some patients, ES appeared repetitively and developed to larger movements later in an event. We defined them as a cluster if each of them was comprised not less than five ES in this study. We observed up to twenty consecutive ES from the onset of visible movement for each cluster and included the first ten ES for statistical analysis to reduce the disparity of total counts of ES per patient.

All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University),²¹ a graphical user interface of R (The R Foundation for Statistical Computing). The chi-square test or Fisher's exact test was used for statistical analysis. Statistical significance was defined at P < 0.001. In the contingency table, basic statistics are shown, including odds ratios and 95% confidence intervals (CI).

2.3 Scoring symmetric and asymmetric expressions

Neck flexion, MCU, and MCL in all ES were scored in each patient focusing on their symmetry/asymmetry and laterality. The direction of NF was defined as the direction to which the neck flexed or the head turned during ES. If the direction of NF deviated toward the left or right side, NF was scored as asymmetrically right or left, respectively (ASR and ASL, respectively). If it was straight to the middle, NF was scored as symmetric (S). For MCU and MCL, the predominant side was defined as the side in which the unilateral extremity was involved (the shoulder and arm for MCU and the leg for MCL). If the predominant side was right or left, MCU and MCL were scored as asymmetrically right or left, respectively (ASR and ASL, respectively). Even when the upper or lower extremities moved bilaterally, if the momentum on one side was larger than the other, we scored it as asymmetric. MCU and MCL were scored as symmetric (S) if they were symmetrically involved.

Then, the number of symmetric movements (N_S) and asymmetrically right- and left-predominant movements (N_ASR and N_ASL, respectively) were counted separately before and after surgery. Finally, the relative frequency of symmetric movements and laterality index of asymmetric movements were calculated as follows,

$\left(\text{Relative frequency of symmetric movements}\right)=\frac{\left({\mathrm{N}}_{\mathrm{S}}\right)}{\left({\mathrm{N}}_{\mathrm{S}}+{\mathrm{N}}_{\mathrm{ASL}}+{\mathrm{N}}_{\mathrm{ASR}}\right)}$

$\left(\text{Laterality index of asymmetric movements}\right)=\frac{\left({\mathrm{N}}_{\mathrm{ASR}}-{\mathrm{N}}_{\mathrm{ASL}}\right)}{\left({\mathrm{N}}_{\mathrm{S}}+{\mathrm{N}}_{\mathrm{ASL}}+{\mathrm{N}}_{\mathrm{ASR}}\right)}$

To visually identify the changes in symmetry or asymmetry for each movement before and after CC, these calculated values were plotted on the graph as follows, (Figure 3A–C)

X = (Relative laterality index of asymmetric movements)

Y = (Relative frequency of symmetric movements)

Here, it means that all movements were seen on the right side if (X, Y) = (1, 0) and on the left side if (X, Y) = (−1, 0). Further, if (X, Y) = (0, 1), it means that all movements are symmetrical (Figure 3). It was defined as significant lateralization if the absolute value of the laterality index is above 0.5.

3.1 Neck flexion (NF)

In terms of the overall number of ES, the percentage of asymmetric NF was significantly higher after CC than before CC (82.9% vs. 20.1%, respectively; odds ratio, 19.3; 95% CI, 10.9–34.2; P < 0.001; Table 2).

The relative frequency of symmetric NF decreased and the absolute value of the laterality index of asymmetric NF increased significantly after CC in 12 patients (Table 3, Figure 3A). The direction of asymmetric NF was ipsilateral to the side of the responsible hemisphere that was the resected side at the subsequent surgery with significant lateralization in 9 of 12 patients (75%, Table 3).

3.2 Muscular contraction of the upper extremity (MCU)

In terms of the total number of ES, the percentage of asymmetric MCU was significantly higher after CC than before CC (81% vs. 39.3% respectively; odds ratio 6.6; 95% CI 4.1–10.7; P < 0.001; Table 2).

The relative frequency of symmetric MCU decreased and the absolute value of the laterality index of asymmetric MCU increased significantly after CC in 11 patients (Table 3, Figure 3B). The predominant side of MCU was contralateral to the side of the responsible hemisphere with significant lateralization in all 11 patients (100%, Table 3).

3.3 Muscular contraction of the lower extremities (MCL)

In terms of the total number of ES, the percentage of asymmetric MCL was significantly higher after CC than before CC (77.6% vs. 29.9%, respectively; odds ratio 8.1; 95% CI 4.5–14.6; P < 0.001; Table 2).

The relative frequency of symmetric MCL decreased and the absolute value of the laterality index of asymmetric MCL increased significantly after CC in nine patients (Table 3, Figure 3C). The predominant side of MCL was contralateral to the side of the responsible hemisphere with significant lateralization in seven patients (77.8%, Table 3).

3.4 ES occurred in a cluster

Asymmetric NF was seen during a cluster in four patients before CC (data not shown) and seven patients after CC (Table S2). The direction of NF was consistently ipsilateral to the side of the responsible hemisphere throughout a cluster only in two patients (cases 13 and 15). In other patients, symmetry of NF gradually disappeared or developed in each cluster (cases 1, 4, 5, 6, and 14). However, initial or developed laterality during a cluster was consistent with the responsible hemisphere in four patients.

Asymmetric MCU was seen during a cluster in seven patients (data not shown) before CC and in nine patients after CC (Table S2). The predominant side of MCU was consistently contralateral to the side of the responsible hemisphere throughout a cluster in five patients (cases 2, 10, and 14–16). The MCU symmetry gradually disappeared or developed in other four patients (cases 1,4, 6, and 8) and the predominant side of MCU was indicative of the responsible hemisphere in one patient in whom the laterality did not switch during a cluster.

Asymmetric MCL was identified during a cluster in four patients (data not shown) before CC and in six patients after CC (Table S2). Once asymmetric MCL appeared, the predominant side was consistently contralateral to the responsible hemisphere in only one patient. Asymmetric features often disappeared or developed in the other five patients (cases 2, 6, 8, 15, and 16), and the predominant side of MCL was indicative of the responsible hemisphere in three patients in whom the laterality did not reverse during a cluster.

3.5 Laterality on MRI and pre-CC EEG

Preoperative MRI showed unilateral malformation of cortical development (MCD) in three patients (cases 2, 6, and 7; Table 1). Most ictal movements were bilaterally symmetric before CC (87.5% for NF and MCU and 78.3% for MCL; Table 2). However, after CC majority of those showed asymmetric expression (91.2% for NF, 80.4% for MCU, and 73.7% for MCL; Table 2). The responsible hemisphere for post-CC ES was consistent with the side of the lesion in two of the three patients (cases 2 and 7).

Six patients also had a unilateral nonspecific lesion in single or multiple lobes (cases 10–15; Table 1). EEG discharges were also bilateral in all patients but unilaterally predominant in five (cases 10 and 12–15; Table 1). However, its laterality was not always consistent with the side of the MRI nonspecific lesion. Many pre-CC ictal movements were symmetric (66.7% for NF, 64% for MCU, and 63.3% for MCL; Table 2). But after CC, a higher rate of asymmetric ictal movements (89.8% for NF, 97.1% for MCU, and 88% for MCL; Table 2) were observed. The responsible hemisphere for post-CC ES was consistent with the side of the lesion in all patients. However, the predominant side of pre-CC epileptiform discharges was paradoxically contralateral to the side of the lesion in two of five patients (cases 14 and 15; Table 1).

There were eight patients with and without laterality in epileptiform discharges on pre-CC EEG (Table 1). Even if there was laterality in pre-CC epileptiform discharges, many ictal movements were still symmetric (62.5% for NF, 64.9% for MCU, and 62.6% for MCL; Table 2). Again, most post-CC ictal movements showed asymmetric expressions (85.7% for NF, 83.8% for MCU, and 72.9% for MCL; Table 2). The responsible hemisphere for post-CC ES was not consistent with the predominant side of pre-CC epileptiform discharges in four of eight patients (cases 1, 3, 14, and 15; Table 1).

Preoperative functional neuroimaging was complicated to be interpreted and was not described in detail here due to differences in the modality (SPECT or PET), the distribution (focal or diffuse), and the patterns (hyper- or hypo-accumulation). However, as with MRI and EEG, the pre-CC laterality in their findings was not associated with the asymmetry of pre-CC ictal movements that was finally clarified after CC (data not shown). Of seven patients who showed the laterality on functional neuroimaging before CC, the side of hypo-accumulation on SPECT or PET was consistent with the responsible side of post-CC ictal movements in five cases (cases 1, 2, 9, 13, and 15; Table 1).