The key to successful epilepsy surgery is the localization of the epileptogenic focus. Although temporal lobe epilepsy is the type of epilepsy with the highest rate of surgical success, the accuracy of focal localization remains a major factor in surgical failure in temporal lobe epilepsy without an imaging lesion.
In patients with temporal lobe epilepsy without imaging foci, seizures almost always begin in the hippocampus and amygdala, the central structures of the temporal lobe. In recent years, a small number of patients have been found to have seizures starting in the temporal cortex, known as temporal lobe cortical epilepsy. Although temporal cortical epilepsy accounts for a small percentage of temporal lobe epilepsy patients, standard anteromedial temporal lobectomy is less effective than central temporal lobe epilepsy, and temporal cortical epilepsy generally requires only noninvasive monitoring to identify the epileptogenic focus. Therefore, it is meaningful to differentiate temporal lobe cortical from central temporal lobe epilepsy as early as possible during preoperative localization.
I. Central temporal lobe epilepsy:
(i) Central temporal lobe epilepsy is diagnosed based on deep electrodes showing seizures starting in the hippocampus and amygdala or a cessation or significant reduction in seizures after standard anterior temporal lobectomy. Sclerosis of the hippocampus or central temporal lobe is the epileptogenic cause and the most common pathological change in the vast majority of patients with central temporal lobe epilepsy and, therefore, a hallmark of central temporal lobe epilepsy.
French and Williamson found that the clinical course of central temporal lobe epilepsy was characterized by the following features: (1) 81% of patients had a history of convulsions in infancy and early childhood; (2) most patients began to have seizures before the age of 15 years; (3) complex partial seizures were the predominant form of seizures, with secondary generalized tonic-clonic seizures; and (4) spasticity. (3) complex partial seizures are the main form of seizures, and there may be secondary generalized tonic-clonic seizures; (4) spastic seizure continuity is uncommon; (5) aura, especially abdominal visceral sensory aura, is very common; (6) there is often a period of resting seizures during the course of the disease; (7) in most patients, the spike and sharp-slow wave foci in the interictal EEG are located in the anterior temporal lobe (F7/F8); about half of the patients show unilateral; the other half are bilateral, but usually The other half are bilateral, but usually one side is predominant. Spikes of 5-10 Hz can be recorded within 30 seconds after the seizure in 80% of patients. 67% of patients have post-ictal slow wave activity. All these abnormal waves have reliable localization value. ⑨ Sodium amytal test showed that the majority of patients had memory loss or impairment on the focal side. Oundtead also suggested that if the onset of the disease was in infancy, patients who experienced a period of quiescence followed by a seizure were more likely to have postoperative pathology of central temporal lobe sclerosis compared to patients without quiescence. Although the above clinical features cannot be used as a basis for the diagnosis of central temporal lobe epilepsy, seizures that differ significantly from it are mostly not central temporal lobe epilepsy.
(ii) Hippocampal imaging:
Many publications have shown that hippocampal changes revealed by MRI are closely correlated with pathologic findings of sclerotic hippocampal lesions and have very high sensitivity and specificity, and predict good surgical outcomes. Therefore, in temporal lobe epilepsy, asymmetry or atrophy of the hippocampus on MRI can be a marker of the epileptogenic temporal lobe.
A coronal scan image parallel to the long axis of the brainstem or perpendicular to the lateral fissure is required to show both sides of the hippocampus, and the long axis of the hippocampus is transected at a right angle close to 90°. Individual variation in hippocampal size is large in normal subjects, but the shape and size of the hippocampus is symmetrical on both sides. The vertical and horizontal diameters of the hippocampal cross-section on the atrophic side are reduced. Therefore, hippocampal atrophy can be confirmed by comparing both sides of the hippocampus. Some authors use the method of calculating the difference between the two hippocampal volumes: R-L < -0.2 cm for right hippocampal atrophy; R-L > 0.6 cm for left hippocampal atrophy; some use the method of left and right hippocampal volume ratio: (L (R)/(L+R); symmetry of both sides is the ratio = 0, positive value is right-sided atrophy; negative value is left-sided atrophy; the average volume ratio of normal adults is 0.002; ratio >±2 standard deviation is a significant difference.
In addition, the T2 image signal of the normal hippocampus was slightly higher than that of the temporal cortex and significantly higher than that of the white matter. The signal of the sclerotic hippocampus was higher than that of the normal hippocampus. Bilateral hippocampal abnormalities can be identified based on hippocampal high signal.
Many papers have shown that hippocampal sclerosis is often associated with amygdala lesions. Hudson performed quantitative measurements of amygdala neuroblast density and astrogliosis in eight patients with amygdala sclerosis with hippocampal sclerosis and eight patients without hippocampal sclerosis. A comparison was made with the postmortem amygdala of patients who died of non-neurological diseases. Significant neuronal reduction and gliosis were found in the amygdala of patients with or without hippocampal sclerosis. There was no significant difference between the two groups of patients. Clinical findings: patients with hippocampal sclerosis often had a history of early brain injury, whereas patients with amygdala sclerosis only did not. Neuropsychological tests showed that patients with hippocampal sclerosis had significantly higher levels of memory impairment than patients without hippocampal sclerosis. Therefore, it can be concluded that amygdala sclerosis can occur in isolation and can be a separate type. It is characterized by the absence of a history of early brain injury, is not associated with a history of early convulsions, and has less impaired memory than patients with hippocampal sclerosis.
(iii) Dual hippocampal lesions
Cendes et al. analyzed 167 patients with partial epilepsy with temporal lobe or non-temporal lobe lesions and found that 15% of patients had a combination of hippocampal atrophy on one side. Some lesions, such as neuronal migration disorders, penetrating malformations, and reactive gliosis, are more likely to be associated with hippocampal atrophy, with an incidence of about 25%; whether these lesions are combined with hippocampal atrophy is not related to the distance of the lesion from the hippocampus. Other lesions, such as benign tumors and vascular lesions, have a low incidence of hippocampal atrophy, at 2% and 9%, respectively, and only when the lesion is close to the hippocampus.
Injuries sustained during critical periods of brain development can lead to abnormal brain development and structural malformations. Such malformations can range from small, inconspicuous foci of cortical dysplasia to large foci of gray matter heterotopia or malformed tumors. Raymond analyzed 100 patients with hippocampal sclerosis using MRI volumetric measurements and found that 15% of them had obscure cortical dysplasia, mainly in the form of gray matter heterotopia around the occipital and temporal horns of the ventricles. The mechanism of occurrence of structural lesions with central temporal lobe sclerosis may be the same as that of lesions occurring during embryonic proliferation or early development. Current imaging techniques may miss the mild cortical dysplasia that accompanies patients with hippocampal sclerosis.
Second, temporal lobe cortical epilepsy
It is generally accepted that temporal lobe cortical epilepsy is diagnosed when seizures begin in the temporal cortex outside the lateral paracentral sulcus. In the absence of an affective lesion, the intracranial record must demonstrate that the seizure focus is located in the lateral temporal lobe region, not the central base, and does not contain a non-temporal lobe cortex. Temporal lobe cortical epilepsy may also be considered if no isolated foci of seizure onset are found, if hippocampal seizures can be excluded and if seizures disappear or are very significantly reduced after anterior temporal resection, including a larger temporal lobe cortical resection. The absence of significant abnormal hippocampal pathology may provide further evidence. Epilepsy in which seizure onset involves both central and lateral temporal lobe structures should be analyzed separately for its possible extent. Differentiation of central versus cortical epilepsy must be performed with both deep central temporal lobe electrodes and subdural electrodes; using only pterygoid electrodes and subdural electrodes is not reliable.
Less is currently known about temporal lobe cortical epilepsy. Ebner suggests that epigastric aura symptoms are more common in central epilepsy and less common in cortical epilepsy, whereas Saygi suggests that epigastric aura symptoms are almost as common in both patients as auditory and vertigo aura, which are considered to be cortical foci in the International Classification. Saygi, Gil-Nagel and Foldvary made three separate investigations, all of which concluded that behavioral manifestations during seizures (ipsilateral limbic automatisms, contralateral atonic postures, and oro-digestive organ automatisms) were significantly more common in central epilepsy.Marks et al [10], after analyzing the seizure characteristics and sites of seizures in a group of 38 patients with secondary epilepsy after central nervous system infection found that:post-meningitis epilepsy was mostly central temporal lobe epileptic foci while post-encephalitis epilepsy was mostly cortical epileptic foci (temporal and parietal lobes). However, five patients with encephalitis had central temporal lobe sclerosis, and all five patients had encephalitis before the age of 4 years. All of the patients with late onset encephalitis had cortical foci. Thus, age at onset is more important than the nature of the infection in predicting central temporal lobe sclerosis or cortical foci.
Ebersole and Pacia found that: in scalp EEG, the presence of seizure waves above 5 Hz confined to one side early in the seizure is highly predictive of central temporal lobe epilepsy, while seizure waves below 5 Hz with variable morphology or clinical seizures without significant EEG discharges but only interrupted normal background waves, and unilateral or widespread irregular slow waves are more likely to be temporal lobe cortical epilepsy.
It is generally believed that the presence or absence of MRI hippocampal atrophy can differentiate central from cortical epilepsy. However, several studies have shown that patients with foci of temporal lobe cortical damage can also have MRI hippocampal atrophy, the so-called “dual lesion”.
To date, studies of temporal lobe cortical epilepsy have been based on the observation of patients with foci of temporal lobe cortical damage. In these patients, the seizure area may be located in the cortex surrounding the temporal lobe lesion, but the clinical and other features of patients with focal lesions do not necessarily apply to patients with temporal lobe cortical epilepsy without focal lesions.
Why are features of temporal lobe cortical epilepsy difficult to detect? Several studies have found that the experience of an aura must be attended by excitation of the interconnected pathways between the base of the temporal lobe and the cortex. This makes it unlikely that the aura is the basis for differentiation. This extensive connection between the central temporal lobe and the cortex means that the propagation of seizure discharges is rapid in both regions. And, a seizure focus in one area can have a profound effect on the function of the other area. If this inference is correct, it would be difficult to distinguish between the two types on the basis of electrophysiological and functional studies. Many phenomena suggest that the above inference is correct. As already mentioned, central temporal lobe sclerosis is occasionally found in patients with cortical epilepsy, and careful analysis of the hippocampus removed from patients with temporal lobe cortical lesions reveals a small but significant cytopenia in most patients. In addition, intracranial electrodes in patients with cortical epilepsy have shown that sometimes the lateral cortex is involved at the same time as the hippocampus at seizure onset. Should this condition be diagnosed as central epilepsy or cortical epilepsy? Or what about overlapping epilepsy of both? Thus, purely surgically confirmed cases of temporal lobe cortical epilepsy are more difficult to pick than any symptomatic epilepsy. Hoch also lacks significance.
Third, the definitive lateral value of clinical seizure presentation
Fakheury et al. reported that aura, especially epigastric sensory aura, occurs more often in right temporal lobe epilepsy. However, according to Palimini, although empirical aura, such as complex visual and auditory hallucinations, are more frequent in right temporal lobe epilepsy, there is no statistically significant localization value of the aura.
Déjà vu is a once-recognized illusion that is unrelated to the person’s cognitive response to an unfamiliar external stimulus. Weinand et al. studied eight patients with refractory temporal lobe epilepsy who had déjà vu during the seizure phase by long-time EEG monitoring with subdural strip electrodes. The epileptogenic foci were found to be located in the central region of the temporal lobe, and all were located in the dominant hemisphere of the non-sharpshooter. In six of the patients with right sharpshooter, the epileptic foci were located in the right temporal lobe where both language and sharpshooter were non-dominant hemispheres; in two patients with left sharpshooter, the epileptic foci were located in the left temporal lobe where the non-sharpshooter was dominant. Thus, the non-sharpshooter had a more stable lateralization relationship with déjà vu than the non-verbal dominant hemisphere.
Efron suggests that the temporal lobe and first frontal gyrus of the language-dominant hemisphere are associated with temporal discrimination, whereas déjà vu is thought to be a disorder of the temporal marking mechanism. The nondominant hemisphere receives time-related entity sensations, which are then transmitted to the linguistic dominant hemisphere. He argues that clinical and experimental evidence shows that right-handed and mostly left-handed individuals discriminate time in the left hemisphere, so that temporal lobe damage in the nondominant hemisphere slows down the transmission of information to the dominant hemisphere that discriminates time, so that the same information is received twice by the dominant hemisphere, and the almost simultaneous double-layered images caused by the slowed transmission due to temporal lobe dysfunction in the nondominant hemisphere cause the newly perceived The newly perceived entity appears “familiar”.
Fakheury suggests that the mechanism may be the result of epileptic discharges in the dominant temporal lobe or sensory language center. Postictal language occurs in 59% of complex partial seizures in left temporal lobe epilepsy and not in right temporal lobe epilepsy. Many papers reported that the epileptic source of postictal aphasia is the dominant hemisphere. privitera concluded that: postictal language test is more accurate than any other non-invasive examination in identifying the epileptic focus. Language disorganization with reading delay occurs almost exclusively in complex partial seizures in left temporal lobe epilepsy.
Gabr found significant localization value in all language manifestations only for postictal aphasia versus normal language during seizures: in 92% of patients with postictal aphasia, the epileptic focus was located in the temporal lobe of the dominant hemisphere; in 83% of patients with normal language during seizures, the epileptic focus was located in the nondominant hemisphere.
Automatism: Fakheury also found that: automatism was mostly seen in right temporal lobe epilepsy, and there was no significant difference between the various automatisms (motor, oro-digestive organ, non-verbal articulatory). The two patients with spitting and coughing automatism were both with right temporal lobe epilepsy, which may be related to ictal vomiting, which occurs mainly in right temporal lobe epilepsy. Upper and lower limb automatism had a definite lateral value only when accompanied by tonic or sluggish posture of the contralateral limb. Bilateral limb automatisms have no definite lateral value.
Tonic head deflection with limited spasticity, secondary to generalized seizures, is mostly seen in left temporal lobe epilepsy. Tonic head deflection is mainly shifted to the contralateral side of the epileptic focus. Many authors believe that head deflection has no definite lateral significance. However, in clonic or tonic head-eye deflection resulting in forced unnatural head-eye posture, the seizure source is located in the contralateral hemisphere.Fakheury et al. found that: tonic head-eye deflection occurs mostly before secondary generalized seizures, 90% of which occur contralateral to the seizure source.
Unilateral dystonic posturing (Dystonic posturing) is defined as a passive, unnatural posture of a single limb (one upper or lower limb) that may be flexed or extended. It usually has a rotational component at either the distal or proximal end of the limb. Many patients also have a tachycardia and tremor component, which is distinctly different from a tonic posture. The latter is manifested by extension or flexion of the limb only, without rotation or unnatural posture.
In Kotagal’s study of 41 complex partial seizures in 18 patients, it was found that the dystonic posture of one limb occurred on the opposite side of the foci of discharge. In 39 of these complex partial seizures, autonomic symptoms occurred on the contralateral limb of the dystonia. 11 of the complex partial seizures showed significant and sustained head and eye deflection to the dystonic side, which occurred after the dystonic posture. 7 seizures were recorded as starting in the medial temporal lobe and base. At the onset of the dystonic posture, the maximum discharge activity was located at the base of the temporal lobe, with the least amount of cerebral convexity. Unilateral dystonic posture is commonly associated with complex partial temporal lobe seizures and is a highly specific localization sign. It always occurs before head-eye deflection and may be one of the earliest clinical manifestations of the spread of temporal lobe discharges to other parts of the brain.
Postictal state: The return to the preictal response state within 1 minute after the end of the seizure (disappearance of the seizure discharge) is a rapid recovery, and this rapid recovery is seen only in right temporal lobe seizures. The reason for this may be related to the primary role of the dominant hemisphere in maintaining consciousness. As Serafetinides found, injection of sodium amytal into the dominant hemisphere tends to cause loss of consciousness with prolonged recovery of consciousness. When stroke, head trauma involves the dominant hemisphere, it tends to lead to impaired consciousness.
IV. Temporal lobectomy
Currently, the method and extent of temporal lobectomy varies greatly among different treatment centers. Awad and Katz proposed a method to assess the extent of temporal lobectomy based on postoperative MRI. The method was: MRI coronal scan with a level thickness of <1 cm was performed 3 months after surgery.
Temporal lobe zonation and resection index: two basic sections were made at the anterior and posterior margins of the midbrain, and then two equal sections were made at the anterior and paramedian temporal lobes of the midbrain, respectively. This produced five layers of continuous coronal views from anterior to posterior. Each layer was then divided into four quadrants: superior lateral (SL), inferior lateral (IL), base (B), and central (M). This divided the temporal lobe into 20 subdivisions on MRI.
The resection index for each partition was divided into: 0-unresected; 1-partially resected; and 2-completely resected. The resection index for each quadrant was the sum of the resection indices in the same quadrant in 5 dimensions (0-10). The total resection index for the temporal lobe was the sum of the resection indexes in the four quadrants (0-40). The medial-superior half of levels 2-5 is considered to contain amygdala and hippocampal structures. Resection of these structures was analyzed separately from the resection index.
The authors found that the postoperative seizure cessation rate was significantly higher in 40 temporal lobe resection patients with resection index >15 than in patients with resection index <15. This was particularly significant for larger resections at the base of the temporal lobe and the inferior lateral aspect. In patients with central base temporal lobe lesions, resection of the amygdala and hippocampal structures is particularly important in seizure control. The authors concluded that as many medial, basal, and inferior lateral temporal lobe structures should be resected as possible.
By assessing the extent of temporal lobe resection as described above, Katz and Awad et al. also analyzed the correlation between the extent of temporal lobe resection and postoperative memory and visual field abnormalities in 20 patients with intractable epilepsy. It was concluded that unilateral temporal lobectomy did not result in significant alterations in memory. After left temporal lobe resection, there was only a slight decrease in near memory. After resection of one temporal lobe, the contralateral temporal lobe could compensate for its role and generally did not show significant memory deficits.