” With the development of modern imaging techniques, especially intracranial electrode monitoring widely used to localize epileptic foci, more and more patients with limited epilepsy can be treated by minimally invasive surgery. The surgical outcome is good, especially for patients with epileptic foci located in the medial temporal lobe. The preoperative evaluation and the various different surgical approaches are briefly described.
Epilepsy is one of the common neurological disorders with an incidence of 0.5-1.0% and an estimated cumulative lifetime incidence of 3%. Among patients with a known cause, cerebrovascular disease is the most common cause (10.9% of new cases), followed by congenital disorders (8%), trauma (5.5%), tumors (4.1%), degenerative lesions (3.5%), and infections (2.5%).
Approximately 40% of patients have generalized seizures and 60% have focal (partial) seizures. Approximately 55% of focal epilepsies originating in brain regions are temporal lobe epilepsies. The remaining 45% are frontal, parietal, and occipital lobe epilepsies. This difference in etiology and clinical presentation is also reflected in the different treatment options, including antiepileptic medication and surgical treatment.
Control of seizures and improvement of quality of life are the main goals of any treatment. Pharmacological antiepileptic treatment is the first step. However, only 33% or less of patients have complete control after one year on a single medication, and only 10-20% of patients who fail the above treatment take 2 medications for complete seizure control.
Interestingly, despite the continued availability of new antiepileptic drugs, the ratio of patients with drug-resistant epilepsy has not decreased significantly. Overall, approximately 30-40% of patients with epilepsy progress to intractable epilepsy and 50% of these are candidates for surgical treatment. In the United States, there are at least 100,000 patients suitable for surgical treatment, and the number is increasing by 5,000-10,000 cases per year. Although a large number of patients are suitable for surgery, only 2,000 surgeries are performed in the United States each year. This discrepancy between supply and demand may be due to.
1. the neglect by many internists of advances in the safety and effectiveness of surgical treatment of epilepsy.
2. the fact that many patients would rather have seizures and their side effects than undergo surgery
3. the fact that although some studies have confirmed the effectiveness of surgical treatment, one third of patients cannot accept the cost of expensive preoperative evaluation and surgical treatment.
However, the fact that surgical treatment of epilepsy has been accepted in the last decade has led to a large number of epileptic patients around the world being treated surgically. This is attributed to the development of preoperative evaluation and surgical techniques. Current preoperative evaluation tools include: monitoring of non-invasive and invasive long-range EEG, quantitative magnetic resonance imaging (MRI), functional MRI (f-MRI), positron emission tomography (PET), single electron emission tomography (SPECT), spectral MRI (MRS), and magnetoencephalography (MEG).
New surgical techniques have made epilepsy surgery safer and more effective. For example, the introduction of minimally invasive techniques allows for more selective procedures, such as selective amygdala hippocampal resection [145, 255]. In addition, palliative surgical approaches such as multiple subxiphoid transection (MST) and the formerly commonly used corpus callosotomy have received renewed attention [130, 138, 140, 179, 180, 209, 281]. Some centers have also started some new approaches such as radiosurgery for the evaluation of treatment of partial epilepsy [170-173, 227].
The best evidence that the development of epilepsy surgery has changed the daily routine of neurologists is the so-called medial temporal lobe epilepsy (the most common form of partial epilepsy). In this type, seizures originate from the amygdala and hippocampus. For a long time, medial temporal lobe epilepsy was so intractable that patients had to live with its presence. But now anterior 2/3 temporal lobe resection or selective amygdala hippocampal resection (specific removal of these two internal temporal lobe structures that cause epilepsy) has freed 80% of patients from seizures after surgery [54, 145, 200, 212, 255,]. Having obtained good surgical results decades ago, it is now clearly recognized that refractory temporal lobe epilepsy that has failed to respond to pharmacological treatment should be operated on as early as possible.
Surgical treatment has greatly simplified the treatment of medial temporal lobe epilepsy, yet drug-resistant extratemporal lobe epilepsy remains a challenge for epileptologists and neurosurgeons. Until 10 years ago, most treatment for these patients was unsatisfactory, with only 40-50% of patients being seizure-free postoperatively, but advances in imaging technology have simplified the precise localization of epilepsy in extratemporal lobe epilepsy. In addition, advances in neurosurgery have improved surgical outcomes, so surgery has become a treatment option for drug-resistant extratemporal lobe epilepsy.
Interictal and interictal activity in abnormal brain regions
When considering surgery, preoperative work must identify the different brain regions that produce seizures or have epileptiform activity during interictal and ictal periods. luders proposed the concept of six different brain regions (Table 1) [121, 122, 184]. This allows the neurosurgeon as well as the neurologist to not only choose the surgical option but also to predict the functional deficits that will result from the surgery. Thus, a preoperative evaluation including various diagnostic studies with different means becomes clear.
Current indications for surgical treatment
Before surgical treatment options for epilepsy can be determined, those hypothetical diagnoses must be confirmed. This may sound banal, but it has an important role because it determines further diagnostic steps. In epileptology, history is a very important diagnostic tool because accurate observation can distinguish seizures from epileptiform seizures and an accurate history can provide important clues to their origin. The seizure characteristics of various forms of limited epilepsy are listed in Table 2 [12, 25, 55, 96, 186, 188-190, 201, 202, 260, 263, 268-270].
In refractory epilepsy, when the diagnostic hypothesis of partial epilepsy and preparation for surgical treatment is presented based on the history, the patient should be evaluated for the feasibility of surgery based on the criteria proposed by Walker [249].
1. progressive neurological diseases (e.g., malignant brain tumors, cerebrovascular disease, multiple sclerosis, etc.) should be excluded. Although patients with malignant tumors can also have seizures, this type of surgery is not epilepsy surgery because its main purpose is not to treat epilepsy but the tumor.
2. It must be determined that they are resistant to the drug. It is very important to determine that the original antiepileptic drug used has been used to the extreme (individual toxic effect). This means not only that the antiepileptic drug has been used to the extreme of the therapeutic range, but also that the blood concentration of the drug may be below the therapeutic level but that side effects have occurred due to individual differences in metabolism (which may occur only at levels several times greater than the extreme level).
3. A minimum medical history of 1-2 years. However, exceptions can be made (especially for epilepsy caused by structural lesions or early diagnosis of medial temporal lobe epilepsy).
4. The patient has a disability due to the seizure, although this may vary individually. Even if there are only seizures with sensory phenomena without altered consciousness, their surgical treatment should be carefully considered when seizures are very frequent.
5. Patients should undergo preoperative examination before surgery and must be informed that they still need to continue taking antiepileptic drugs after surgery.
6. IQ below 70 indicates the presence of diffuse brain lesions, and the chances of successful surgery are very small, so surgery should be performed with great caution.
7. Psychiatric disorders are not suitable for epilepsy surgery.
Preoperative evaluation steps
In the absence of the previously mentioned unsuitable conditions for surgery, the patient usually requires further evaluation. The preoperative evaluation of the patient to be operated on consists of two stages: a noninvasive evaluation (Stage I), and an invasive evaluation (Stage II). Stage I includes detailed history taking, review of previous treatment protocols and neurological examination, neuropsychological testing, psychiatric and psychological assessment, interictal and ictal EEG (including video EEG monitoring), MEG, CT, MRI, MRS, fMRI, SPECT, PET, etc.
The first stage of the preoperative evaluation is to screen patients for operable treatment. If epileptogenic lesions can be identified during this phase, surgery can be performed directly. However, for some patients, the first stage does not accurately localize the epileptogenic lesion, and an invasive evaluation (stage II) is required. This includes deep video EEG, video EEG monitoring with subdural and/or epidural electrodes. Further intra-arterial pentobarbital neuropsychological control testing is required if needed. The second stage may determine if the patient is surgically treatable or if intraoperative EEG monitoring is required (stage III). (Table 3) [40].
Non-invasive assessment (stage I)
Scalp electroencephalography
The assessment of seizures by scalp EEG and video EEG monitoring in the awake and sleep conditions is an important examination in this phase. Video EEG monitoring not only monitors the seizure but also records the EEG to lateralize and localize the origin of the seizure.
Many studies have shown that the brain areas with repetitive epileptiform discharges in multiple EEGs during the interictal period correspond to the location of the discharges during the seizure [5, 213]. Thus, interictal epileptiform discharges can be detected in 50% of patients during routine awake EEG [103, 157, 163, 164, 197, 252]. Sleep EEG recordings have increased the accuracy of temporal lobe epilepsy diagnosis to 90% [29, 197, 252].
In medial temporal lobe epilepsy, typical epileptiform potentials are seen in the frontal and temporal lobe base [260]. This potential change is very pronounced in temporal lobe neocortical epilepsy. The interictal EEG in temporal lobe epilepsy typically shows sinusoidal activity of high amplitude and slow frequency [256].
Interictal and interictal scalp EEG is valuable for localization and lateralization of the origin of medial temporal lobe epilepsy. However, its value in non-temporal lobe epilepsy is uncertain. In frontal lobe epilepsy interictal and ictal EEG often shows non-specific changes or no changes, i.e. abnormal changes are seen in only 10% of patients [96, 163, 270]. In fact, some frontal lobe epilepsies can have very odd motor signs EEG is often undiagnostic and cannot even distinguish between hysterical seizures, epileptiform seizures and epileptic seizures.
In contrast to frontal lobe epilepsy, occipital and parietal lobe epilepsy can often be seen with epileptiform potentials in interictal and interictal EEG [11, 202]. Localized epileptiform electrical activity is found in approximately 65% of patients with parietal lobe epilepsy and 79-97% of patients with occipital lobe epilepsy. Only 10% of patients with parietal lobe epilepsy and 14% of patients with occipital lobe do not have abnormal EEG changes [11, 202].
Magnetoencephalography tracing (MEG)
MEG is a method of measuring the magnetic field generated by cortical pyramidal cells [3, 48, 136, 181, 193]. Unlike EEG, MEG measures magnetic fields rather than electrical activity.
Many studies have shown that MEG in refractory epilepsy reveals epileptiform activity during interictal and ictal phases [3,4,49,106,182,226,229]. Many studies have compared the ability to localize epileptic foci with MEG and conventional EEG and MRI [50,106,207].Smith et al. performed MEG on 50 patients with possible surgery [207]. Their results were 56% consistent with conventional methods of localization, 12% partially consistent, 10% inconsistent, 16% no spikes in MEG, and 6% insufficient data [207]. This study suggests that MEG is more effective in localizing epileptic foci on the cortical surface than deeper ones. Other studies have yielded similar results especially in non-temporal lobe epilepsy [104,106,134]. Therefore, MEG is still effective for the detection of interictal and interictal epileptiform discharges.
MEG is used clinically mainly in neocortical epilepsy with normal MRI presentation and MRI presentation with large regional abnormalities.MEG allows the use of less traumatic EEG monitoring, thus reducing treatment costs and mortality. However, the use of MEG in the localization of the origin of seizures is limited by the high cost of MEG and the reluctance of insurance companies to cover this cost.
Structural imaging
The advent of MRI has revolutionized the evaluation of patients with epilepsy. It provides detailed anatomical detail and can show large structural lesions very sensitively. Its higher sensitivity and specificity compared to CT has made MRI the examination technique of choice in the preoperative evaluation of refractory epilepsy [221].
Usually, high-resolution MRI is only used for the examination of specific types of epilepsy [9,22,34]. In the last decade, this technique has started to be used for the examination of small, even microscopic epileptogenic foci. Thus, it is possible to detect very small abnormal changes due to migration disorders (cortical dysplasia, multiple microcephalic gyrus, malformations, etc.) in addition to tumors and vascular malformations, which are common causes of epilepsy. In addition to these structural abnormalities, MRI often reveals medial temporal lobe sclerosis in patients with medial temporal lobe epilepsy, as evidenced by hippocampal atrophy on T1 and enhanced hippocampal signal on T2 [8,58,92,95,113,127,242]. Although medial temporal lobe sclerosis can often be diagnosed by MRI, modern quantitative volume-based MRI techniques measuring hippocampal node structure would further improve the sensitivity of the diagnosis [23,90,114].
Tumors (ganglioglioma, low-grade astrocytoma, oligodendroglioma, etc.) can be detected in 30-40% of patients with non-temporal lobe epilepsy [27]. The same prevalence is found in disorders of migratory disorders. Rare structural lesions are Stureg-Weber syndrome, postnatal injuries (ventricular penetrating cysts), scarring, etc. These intracranial lesions almost always coincide with epileptogenic foci and such patients can be treated directly surgically if further noninvasive findings can be explained by lesions [37,105,231]. However, in about 20% of patients the most advanced MRI techniques fail to detect any lesion. Although these patients are difficult to treat medically, surgical treatment may yield satisfactory results [205].
New MRI techniques such as fluid-attenuated inversion recovery sequences (FLAIR) and diffusion and perfusion MRI can help in further localization of the origin of the attack. substantial lesions on FLAIR show high signal while CSF shows bottom signal [162]. Therefore, FLAIR can identify a hippocampal sclerosis and lesions rather than ectopic structures [7,265]. Diffusion-weighted was initially used for the identification of acute cerebral infarction, but it has also been reported to be sufficiently sensitive for the demonstration of epileptogenic foci [42,111,266]. Although diffusion and perfusion MRI have high sensitivity, their role in preoperative evaluation remains to be further validated.
Functional MRI
MRI techniques that have been used more frequently in recent years include fMRI and MRS [152,191,192]. It is a high-resolution, noninvasive method for detecting neural activity by blood oxygen levels [63,64,87,93,239,250]. Thus the neurosurgeon can decide which cortical areas must be avoided based on fMRI when designing the surgical plan. If these areas are involved in surgery, there is a risk of neurological deficits after surgery. Although some of the data are very encouraging, fMRI has not yet replaced the role of intracranial electrode function tracing in the preoperative evaluation. Triggered EEG may facilitate the use of this new technique [107,230,250].
Nuclear magnetic resonance spectroscopy
It is the only noninvasive test that can detect chemical signals in the body.MRS reflects the chemical changes in the examined area according to the difference in the resonance of each different nucleus in a magnetic field of different frequencies. In epileptic patients, MRS mainly detects 31P and 1H nuclei in the body. Generally MRS software can acquire both short time echo and long time echo signals. The latter spectrum can be obtained for methionine (NAA), choline-containing compounds (Cho), phosphocreatine and creatine (Cr) and lactate [33,191]. Short-term echoes include the above metabolites and inositol, glutamine, glutamate, phenylalanine, glucose, cyanocrab inositol/taurine, proteins and lipids [10,28,86,137,161].
NAA is mainly located in neurons and precursor cells, and its reduction indicates neuronal loss or abnormal function. In contrast, Cho and Cr are present in neurons and glial cells. It is now found that 60-90% of MRS in patients with temporal lobe epilepsy have reduced hippocampal NAA concentrations [112]. The localization value of MRS in non-temporal lobe epilepsy has been less studied, and Stanley et al. reported that 20 patients with non-temporal lobe epilepsy who were routinely examined by MRS had significantly reduced NAA-related resonance intensities (e.g., NAA/Cr, NAA/Cr+Cho), with a predominance of the seizure origin area [225]. However, the role of MRS in non-temporal lobe epilepsy remains to be determined.
Functional imaging
In addition to structural imaging, functional imaging is also a useful method for preoperative evaluation to detect brain lesions. (hypermetabolism on PET) [21,72,75,76].
The commonly used tracers for interictal and ictal SPECT are 99mTc-hexamethylpropylene diamine oxime (HMPAO) and 99mTc-diethylcysteine (ECD). Seizure-phase SPECT requires intravenous administration prior to a clinical/EEG seizure. This test requires a nurse or physician to inject the tracer and observe the video EEG [232]. Although complex and time-consuming, it is very useful, especially for non-temporal lobe epilepsy.
The sensitivity of interictal SPECT is about 50% in temporal lobe epilepsy [72,232] and even lower in non-temporal lobe epilepsy [91,126], which limits its use in preoperative evaluation. In contrast, seizure SPECT is a very useful tool for preoperative evaluation. Its sensitivity rate for temporal lobe epilepsy is 90-97% [41,84,217], and the accuracy of localization for non-temporal lobe epilepsy depends on the timing of tracer injection. Most studies have reported its sensitivity of 81-90% for non-temporal lobe epilepsy [45-47,77,83,117,125,126]. The fusion of interictal and ictal SPECT images can significantly improve their localization value and can be fused with the patient’s MRI images to obtain “ictal differential images” [117,147].
Interictal and interictal (rarely) PET is often performed with 18F-deoxyglucose (18F-FDG) and 11C-flumazenil [82,110,233,234]. 60-90% of temporal lobe epilepsies show hypometabolism in the interictal phase, so it can be used for preoperative evaluation [82]. However, PET has a low sensitivity in non-temporal lobe epilepsy (approximately 45C60%) and provides little useful information about epileptogenesis [82]. Its role in localizing MRI-negative frontal lobe epilepsy has not been determined. Although PET can show local hypometabolism in the frontal lobe, it is possible that it is not consistent with the epileptogenic focus. Areas of hypometabolism are also found in regions outside the frontal lobe [82].
Neuropsychological assessment
Neuropsychological examination should be performed preoperatively in both temporal and non-temporal lobe epilepsy patients. These include: memory (the primary function of the temporal lobe), learning, IQ, verbal orientation, motor skills, visual perception and visual reconstruction functions, attention and concentration, and verbal and non-verbal fluency [102,154,168]. Although the tests vary from center to center, some of them are accepted and established, such as the WechslerAdultIntelligenceScale for IQ tests [97,98,251]. Some other tests may detect cognitive impairment associated with seizures. Frontal lobe function assessments include the WisconsinCardSorting, DesignFluency, Stroop, tower, trail-making, finger-tapping, Purdue-GroovedPegboardtest [19,43,69,102,118, 135,158,175,224,228,235]. the Rey-Osterrieth complex graphics test and somatosensory tests such as two-point discrimination perception can be used for parietal function assessment [71,98,118,153,177]. Memory assessment includes the learning and memorization of verbal and non-verbal information. The most commonly used tests are the original and corrected WechslerMemoryScale.In addition to using word collocations, story retellings, and often learning tests such as the California or ReyAuditoryverballearningtests.Language proficiency assessments include WesternAphasiaBattery, BostonDiagnosticAphasiaExamination, and BostonNamingtests.The last one is also used for functional assessment of temporal lobe neocortex [98-100,102,236].
Traumatic assessment (Stage II)
If the characteristics of the seizure are consistent with the results of the stage I assessment then the patient can be treated directly with surgery. This is common in patients with temporal lobe epilepsy [37,105,231]. However, traumatic EEG is required when the epileptogenic focus cannot be accurately localized in the noninvasive preoperative evaluation. In addition, isopentobarbital (Wada) testing is required if the site of origin of the seizure overlaps with areas of language and/or memory function.
Traumatic EEG recording
The proportion of patients undergoing intracranial electrode placement varies among epilepsy centers [148, 183, 216, 279]. The need for traumatic EEG monitoring is seen in the following: no obvious epileptogenic lesions on MRI; suspicion of multiple epileptogenic foci; multiple epileptiform discharges found on scalp EEG or no epileptiform discharges in the interictal period, scalp EEG that does not identify the origin of the seizure or points to multiple sites, inconsistent with the results of the stage I assessment; non-invasive methods that reveal diffuse origins or proximity to important functional areas (e.g. perirolandic Wernicke’s area, or Broca’s area). For these reasons, traumatic EEG recordings were performed in 5-20% of patients with temporal lobe epilepsy and 40-70% of patients with non-temporal lobe epilepsy [218].
If there is reason to suspect that seizures originate from medial temporal lobe structures, intracranial EEG can be recorded by semi-traumatic disc-shaped electrode placement [257, 259, 261]. However, if the results of the stage I evaluation point to a non-temporal lobe or temporal lobe neocortex, strip and/or grid electrodes are placed and deep electrodes are rarely applied. Strip and/or grid electrodes are placed subdurally through a small craniotomy, and deep electrodes are placed through a craniotomy under a stereotactic technique [120, 148, 149, 216].
The accuracy of traumatic electrode monitoring depends not only on the examination modality but also on the cause of the epilepsy and the origin of the seizure.Spencer and Lee compared the accuracy of different electrodes through 53 cases with different types and etiologies of epilepsy (Table 4) [219]. It was shown that grid electrodes were more accurate than strip electrodes in cases with lesions and neocortex (especially frontal lobe). However, the accuracy of the latter was higher than that of the former in cases without lesions and intermediate structures. Other studies have yielded the same results [176, 215, 216, 222].
The mortality rate with intracranial electrodes is low (1-2%) and there is very little wound infection or bleeding. Due to the use of intracranial subdural electrodes and deep electrodes, the origin of seizures can be determined in 70-80% of patients with non-temporal lobe epilepsy. Traumatic assessment is still useful after a failed Stage I assessment [204].
Isopentobarbital (Wada) test
Since 1960, intracarotid isopentobarbital to test the dominant fixation side of language and memory function (Wada test) has been part of the preoperative evaluation [247]. Transarterial injection of sodium isopentobarbital into the carotid artery (overall Wadatest) or selective injection into the anterior choroidal artery (selective or superselective Wadatest) can further assess the abnormal functional hemisphere to determine the risk of postoperative memory deficits [223, 247, 262]. dominant side), but is less useful in non-temporal lobe epilepsy [101].
Surgical treatment of refractory epilepsy (stage III)
Patients are indicated for surgical treatment of epilepsy if they meet the above mentioned criteria after preoperative evaluation. The best surgical option should be discussed in consultation with epileptologists, neuropsychologists, psychiatrists, and neurosurgeons. Surgical approaches are classified according to indications, pathology, method and extent of resection as follows.
1. etiologic treatment and palliative treatment. Etiologic treatment (e.g., resection of the anterior 2/3 of the temporal lobe or resection of cavernous vascular malformations) is the removal of the epileptic focus to obtain a seizure-free outcome. Palliative treatment (e.g., corpus callosotomy) is done by blocking the seizure spread pathway or removing the seizure onset secondary to epilepsy. It is conceivable that its surgical results are less effective than etiologic treatment.
2. According to the radiological findings, there are lesioned and non-lesioned types. The former refers to radiological findings of lesions (i.e., resection of so-called tumors and cavernous vascular malformations, etc.), while the latter is seen in patients with normal MRI or only nonspecific pathological changes.
3. In addition, it can be divided into resection and non-excision (excision). In resection, the mutated brain tissue is removed (e.g. lesion excision or amygdala hippocampal resection), while non-excisional procedures (dissection) include corpus callosotomy and submural transection. Implantation of a vagus nerve stimulator is also a non-excisional procedure.
Finally, the extent of resection can be divided into resection of the epileptogenic focus only (resection according to the individual case such as frontal lobe partial cortical resection) and expanded standard resection according to the size of the epileptogenic focus (e.g., standard resection of the anterior 2/3 of temporal lobe). Obviously, there is a preference for removing only a small area of brain tissue, even though some patients may require a second surgery because of unsatisfactory results. The common surgical approaches and their results are described below.
Temporal lobectomy
Excision of the anterior temporal lobe is the most common and effective surgical procedure. One of the most common procedures is the standard anterior temporal lobe resection. However, there are small differences in this procedure from center to center. Some surgeons prefer the Falconer procedure (with the eagle hunter), which removes the entire temporal lobe including the amygdala and hippocampus (usually 4.5-6.5 cm from the temporal pole, avoiding resection of the dominant hemispheric language area) [35, 59, 241, 248] (Figure 1). Some surgeons have also decided on the extent of temporal lobe resection based on intraoperative cortical EEG findings [68, 237, 238].
Standard temporal lobe resection with strict control of surgical indications can result in an 80% seizure control rate [51, 52, 94, 95, 210, 212]. Approximately 30% of patients with temporal lobe neocortical lesions have surgical outcomes complicated by the presence of dual pathological alterations [24, 116, 119]. Complications are rare with standard temporal lobectomy. Common complications include visual field defects (greater than 50%), hemianopia (2-4%), temporary or permanent hemiparesis (4% and 12%, respectively), infection (meningitis or abscess) and epidural hematoma (both less than 0.5%), transient III and IV cranial nerve palsy (less than 0.1%), temporary aphasia of 4-7 days (greater than 20%), permanent speech impairment (1-3%), memory deficits (1%) and temporary mental status or depressive manifestations (2-20%). Their mortality rate is less than 1% [159, 160].
Selective amygdala hippocampal resection
The most common focal epilepsy, medial temporal lobe seizures, has abnormal discharges originating from the amygdala, hippocampus, and parahippocampal gyrus [256]. This fact suggested that restrictive resection of these findings could eliminate seizures, and in 1958 Paolo Niemeyer reported the first transcortical selective resection of the amygdala and hippocampus – selective medial temporal lobe structural resection [145]. Although it was effective, it was gradually forgotten due to the development of other techniques.
In 1975 neurosurgeon Yasargil and epileptologist Bemoulli modified the selective amygdala-hippocampal resection after Wieser [255] (Figure 1). After removal of the amygdala, the anterior hippocampus and part of the parahippocampal gyrus were also removed. In addition to these structures, the most important afferent pathway (entorhinalarea) and the seizure diffusion pathway (hook bundle, anterior commissure) were blocked. These are associated with good postoperative outcomes. In a retrospective study of 369 patients with selective amygdala hippocampal resection at least 12 months follow-up (mean age 85.2), 67% had no seizures or only aura seizures. 11% had no more than 1 or 2 seizures per year and 15% had at least a 90% reduction in seizure frequency; however, 8% of patients had no substantial improvement [264]. Other studies on the structural comparison of this procedure with other procedures have reached similar conclusions [85, 108, 150, 155, 156, 210, 278, 282]. The efficiency of surgery is related to the presence of structural lesions (especially severe hippocampal sclerosis), the history of febrile convulsions, the extent of hippocampal and especially parahippocampal gyrus resection, the presence of contralateral epileptogenic foci, whether seizures are transient (especially originating from the anterior part of the central structures), and whether seizures are transmitted contralaterally to the contralateral hemisphere [199, 200, 259].
Extratemporal lobectomy
In contrast to the standard temporal lobectomy mentioned above, complete resection of the frontal, parietal or occipital lobes is now rarely used (Figure 1). The current surgical treatment of non-temporal lobe epilepsy is mainly local lobectomy limited to the epileptogenic focus [166].
In 75 cases of frontal lobe epilepsy surgery, 64% were seizure-free after surgery, 12% had rare seizures, 16% had a satisfactory decrease in seizure frequency, and 12% had no significant results [195]. The 26 patients with tumors (2 malignant) in the group had the best postoperative results (81% seizure-free). Only 50% of lesion-free patients were seizure-free postoperatively. This result is in agreement with other centers reported [61, 62, 144, 187, 208, 253, 283].
Of the 39 parietal lobe epilepsy surgeries, 52% were seizure-free postoperatively and 30% had a decrease in seizure frequency of more than 90% [151]. Complications included temporary sensory-motor impairment or mild aphasic symptoms (20%), permanent sensory-motor impairment (12%) and exacerbation of preoperative sensory impairment (15%) [189, 190]. Less common complications were temporary lower quadrant visual field defects, left-right loss of recognition and local Gerstmann’s syndrome [189, 190].
Of 30 occipital lobe epilepsy s