Epilepsy surgery assessment and testing

Question 1

Refractory (Drug Resistant, Intractable, Pharmacoresistant) Epilepsy

Answer 1

The latest definition by the international league against epilepsy (ILAE) task force defines refractory epilepsy as failure of “adequate” trials of two antiseizure medications (ASMs) either as monotherapy or as combination (poly-) therapy, to control seizures [1]. In order to meet this cri- terion, it is critical to make sure that the ASMs have been given enough chance (“adequate trial”)—i.e., they must have been used at the maximum tolerated dose (with no severe side effects)—and given enough time (determine sei- zure reduction after a follow-up period 3 times the longest interseizure interval, or one year, whichever longer). Therefore, with proper man- agement, in most cases the diagnosis of refrac- tory epilepsy should be possible to make within 1–2 years of the start of the seizures.

Question 2

Treatment Options for Refractory Partial Epilepsy

Answer 2

Besides trying different combinations of ASMs, a variety of treatment options are available, including established methods of surgical resec- tion of the seizure focus, or surgical “non-resection” options such as vagus nerve stimulation (VNS), responsive neurostimulation (RNS), or multiple subpial transections (MST). Currently there are also several investigational surgical treatments available such as deep brain stimulation (DBS, which has been approved as a treatment in Europe), transcranial magnetic stimulation (TMS), trigeminal nerve stimulation (TGNS), external VNS, or transcranial direct current stimulation (tDCS).
In particular circumstances, other non-surgical methods, e.g., ketogenic diet in young children, can be of therapeutic value. Currently, research for newer ASMs as well as novel potential therapies such as gene therapy, cell transplanta- tion, or vaccination is underway.

Question 3

Presurgical evaluation summary

Answer 3

The purpose of presurgical evaluation is to char- acterize the seizure type and to lateralize and localize the seizure onset focus in patients with refractory partial epilepsy. Therefore, patients should be admitted to a properly equipped epi- lepsy monitoring unit (EMU) for continuous video-EEG monitoring. The ASMs are usually tapered down in order to record electrographic and clinical seizures for the above purposes. Patients also need appropriate neuroimaging studies including high-resolution epilepsy proto- col MRI and other available imaging technologies as indicted or available such as MR spectroscopy (MRS), positron emission tomography (PET), single-photon emission computed tomography (SPECT), or magneto-encephalography (MEG).
In case the scalp recording is of limited value, patients who are considered potential candidates for surgical treatment may need invasive recordings using strips, grids, or depth electrodes and possibly cortical mapping before an appro- priate surgical procedure could be planned. After lateralization and localization of the seizure onset focus, potential surgical candidates should undergo a series of tests for further evaluation to define their final candidacy for an appropriate surgical treatment including neuropsychological testing and functional MRI or intracarotid amo- barbital procedure (IAP, Wada test)

Question 4

Neuropsychology

Answer 4

The main questions that are addressed by a neu- ropsychological test include determining parts of the brain that are impaired. In particular, higher cognitive functions such as verbal and visual memory and language are studied. The test also helps establish a baseline for future comparison. This helps predict potential postsurgical deficits. Studies have shown that the best predictor of postoperative adequacy is the preoperative cog- nitive and psychosocial status; i.e., the lower the preoperative cognitive and psychosocial status, the lower the risk for further decline [2, 3].

Question 5

Intracarotid Amobarbital Procedure (Wada Test)

Answer 5

This test “imitates” the prospective temporal lobectomy by temporarily inhibiting unilateral brain functions using a drug. Therefore, it helps lateralizing language dominance and memory function. The Wada test evaluates memory function of each temporal lobe separately to determine whether the nonepileptic side would be capable of handling memory function by itself after the affected temporal lobe is surgically removed. The test also can assist with seizure onset side since there is typically concordance between the seizure onset side and poor memory function on that side (upon contralateral injection).

Question 6

Epilepsy surgery - efficacy and safety RCTs

Answer 6

Randomized, controlled trials (RCT) to assess the efficacy and safety of epilepsy surgery were missing till 2001. In the first such study [4], 80 patients with temporal lobe epilepsy (TLE) were randomly assigned to surgery (n = 40) or treat- ment with ASMs for one year (n = 40). The primary outcome was seizure freedom and sec- ondary outcome included seizure frequency and severity, quality of life (QOL), disability, and death.
At 1 year, the cumulative proportion of patients who were seizure free was 58% in the surgical group versus 8% in the medical group (P < 0.001). Patients in the surgical group had fewer complex-partial seizures (CPS) and sig- nificantly better quality of life (P < 0.001 for both comparisons) than the patients in the med- ical group. Four patients (10%) had adverse effects of surgery (mainly the expected mild language and memory-related problems such as word finding and short-term memory difficulties) while one patient in the medical group died. This study confirmed that in TLE, surgery is superior to prolonged medical therapy. This RCT also showed that randomized trials of surgery for epilepsy are feasible and appear to yield precise estimates of treatment effects.

Question 7

Early Randomized Surgical Epilepsy Trial (ERSET)

Answer 7

It has been well established that years of active epilepsy predict cognitive impairment in children and adolescents [5, 6]. Therefore, in order to investigate the effects of early surgery, i.e., whether it would be superior to continued med- ical management, Engel et al., conducted a multicenter, parallel-group RCT soon after the failure of 2 ASM trials in patients with temporal lobe epilepsy [7]. Thirty-eight patients (18 M/20 F; age 12 years) with MTS and refractory MTLE who were within 2 consecutive years of adequate trials of 2 ASMs were randomized (1) continued ASM (n = 23), or (2) anterior mesial temporal lobectomy (AMTR) plus ASM treatment (n = 15) and were followed for two years. The primary outcome was seizure freedom during the second year of follow-up, and the secondary outcome was health-related quality of life (QOL), cognitive function, and social adaptation.
Seizure freedom during the second year of follow-up was reported in 11 of 15 patients in the surgical group versus none of the 23 in the medical group (P < 0.001). Also, improvement of QOL was higher in the surgical group (P = 0.01). Memory decline occurred in 4 patients (36%) after surgery. Adverse events included one stroke in a surgical case versus 3 cases of status epilepticus in the medical group. It was concluded that resective surgery plus ASM in patients with new refractory MTLE results in lower probability of seizures during second year of follow-up than continued ASM treatment alone.

Question 8

Epilepsy surgical methods

Answer 8

There are different surgical methods depending on the patient and seizure type and other char- acteristics including:
Temporal lobe surgery
Lobectomy
Resection of the epileptogenic zone Lesionectomy
Corpus Callosotomy Hemispherectomy
Multiple subpial transections (MST)

Question 9

Types of temporal lobectomies

Answer 9

There are different methods to remove the seizure focus in the temporal lobe. Anterior temporal lobectomy is the classic and most commonly used type of surgery, but it may be done using different approaches including:
– Standard (en bloc) anterior temporal lobec- tomy (ATL) including 3–6 cm of anterior temporal neocortex and 1–3 cm of mesial structures (amygdala and hippocampus)
– Modified (Yale group) and limited neocorti- cal resection (3.5 cm from temporal pole) sparing superior temporal gyrus, to address language deficits
– Selective amygdalohippocampectomy
– Stereotactic radiosurgery [8, 9]

Question 10

Selective Amygdalohippocampectomy (SAH)

Answer 10

This method was introduced by Niemeyer in 1958 in an attempt to preserve the lateral tem- poral cortex out of concern for language deficits. The technique includes accessing temporal horn to selectively resecting mesial temporal struc- tures through a small incision in the middle temporal gyrus while preserving the neocortical area. Other approaches to selective amygdalo- hippocampectomy include:
– Transsylvian approach [10, 11].
– Subtemporal approach [12].
– Other variants of the transcortical approach
[13].

Question 11

Long-term outcomes following surgery - MTLE, MTS

Answer 11

Temporal lobectomy provides continued long-term seizure control but risk of seizure recurrence >= 2 years after surgery is present. In one report, 50 consecutive post-temporal lobec- tomy patients with MTS (mean follow-up 5.8 years, range 2–9.2) seizure-free rates were 82% at 12 months, 76% at 24 months, and 64% at 63 months [14]. Complete, or better, seizure outcome was associated with significantly better long-term QOL, and risk factor for seizure recurrence was the reduction in ASM intake—or absorption—in 5 of 17 patients (29%), including 3 of 5 with a first seizure recurrence within 24 months.

Question 12

Outcomes after standard ATL

Answer 12

In another study, 116 patients with MTS, MTLE, and post-anterior temporal lobectomy with amygdalohippocampectomy (ATL-AH) were studied (follow-up period: 6.7 years) [15]. Complete seizure freedom was seen in 103 patients (89%) and Engel Class I or II outcome in 109 patients (94%). The highest concordance (i.e., test consistent with the side of eventual surgery) was seen with video-EEG (100%), PET (100%), MRI (99.0%), and Wada test (90.4%). The lowest concordance was seen with SPECT (84.6%) and neuropsychological testing (82.5%). A strong Wada memory lateralization appearedto be the predictor of excellent long-term seizure control, while less disparity in the memory score between the sides was the predictor of persistent seizures.

Question 13

Temporal Lobectomy—Inferior Temporal Approach

Answer 13

Inferior temporal gyrus approach to mesial tem- poral lobe resection is safe and effective with low morbidity and mortality. One study reviewed 483 patients with AMTL resection via inferior tem- poral gyrus approach for TLE [16]. Thirteen complications (2.7%) (3 months post-op) were reported including eight delayed SDH (1.6%), two superficial wound infections (0.4%), one delayed ICH (0.2%), one small lacunar stroke (0.2%), and one transient frontalis nerve palsy (0.2%). There were no deaths or severe neuro- logical impairments. Complications were more common among older patients.

Question 14

Selective Amygdalohippocampectomy

Answer 14

SAH in TLE patients with MTS results in seizure-free outcomes comparable to procedures with more extensive temporal neocortical resec- tions [17]. Although this method was introduced to minimize the neurocognitive side effects of temporal lobectomy, interestingly, at this point there is more controversy regarding postopera- tive neuropsychological outcomes, rather than seizure-free outcome, when compared to stan- dard ATL. Some studies have suggested that SAH results in better cognitive function com- pared to ATL [10], while others have shown no evidence of a clear neurocognitive benefit and in fact SAH might cause significant verbal memory deficits in dominant temporal lobe resection [18, 19].
In another study, 76 adult patients with SAH for MTLE via the trans-middle temporal gyrus approach reported 92% Engel Class I or II with very low surgical morbidity and no mortality. Postoperative neuropsychological testing showed verbal memory decline in the left SAH group, but no memory decline in the right SAH group was seen while some even showed improvement [20].

Question 15

Neurocognitive Deficits and Risk Factors Following ATL

Answer 15

Cognitive impairment is very common in epilepsy patients and may be negatively or pos- itively affected by surgery. Larger temporal lobe resections are associated with better seizure control, but at the same time resecting more functional tissues carries higher risk of cognitive outcome [21].
Comparison of the changes in cognitive per- formance in relation to the extent of resection of mesial and lateral temporal structures (1–2 cm and >2 cm for mesial, and 4 or 4 cm for neocortical) in 47 right-handed patients with left temporal lobectomy for MTLE showed no dif- ference in cognitive outcome between the groups. However, there was a negative correla- tion with patient age at seizure onset [22].

Question 16

Standard vs selective ATL

Answer 16

A meta-analysis of standard anterior temporal lobectomy (ATL) versus selective SAH for sei- zure control in TLE included 11 studies (1203 patients) and concluded that ATL is more likely to achieve an Engel Class I outcome compared with SAH (p < 0.01). Standard ATL confers better chance of achieving freedom from dis- abling seizures in patients with TLE [23].

Question 17

RTL vs LTL

Answer 17

Comparison of neuropsychological outcome following RTL versus LTL shows postoperative decline in verbal memory after LTL, perfor- mance intelligence decline after LTL (depending on infero-lateral and basal region removal), and visuospatial memory outcome after RTL (depending on basal and hippocampal region removal). More resection is associated with worse functioning and vice versa [24].

Question 18

Outcome Following ATL
in Nonlesional TLE Surgery

Answer 18

A normal MRI is not against surgery in patients with TLE. Sixty-four adult patients with refrac- tory TLE but normal MRI who had underwent TLE surgery (1996–2009) were followed for 1– 14.5 years (mean 4.1). Standard anterior tempo- ral lobectomy was done in 84% and an unre- markable pathology was reported in 45% of the patients. Complete seizure freedom rates are given as follows:
– 1 year: 76% (Engel Class 1: 81%)
– 2 years: 66% (Engel Class 1: 76%)
– 7 years: 47% (Engel Class 1: 69%)
The negative predictors (risk factors) were (1) higher baseline seizure frequency and (2) preoperative generalized tonic–clonic sei- zures. Memory decline was reported with domi- nant hippocampus resections [26].

Answer 18

Surgical outcome following surgery in MRI-negative (nonlesional) patients with refrac- tory partial epilepsy can result in favorable out- come. A review of 399 patients has shown positive long-term outcome after 0.5–15.7 year (mean 6.2) follow-up period [25]. The seizure types included focal dyscognitive seizures complex-partial): 237 (59%), GTCS: 119 (30%), focal without alteration of awareness (simple-partial seizures): 26 (6%), and mixed: 17 (4%). Of these 372 (93%) had temporal lobe, and 27 (7%) extra-temporal lobe resections. The pathology showed MTS in 113 patients (28%), gliosis in 237 (59%), and normal tissue in 49 (12%). The overall Engel Class I outcome is given as follows:
– 81% at 6 months – 78% at 1 year
– 76% at 2 years
– 74% at 5 years
– 72% at 10 years
Almost all seizures occurred during the first year after surgery. The positive predictive factor was seizure control during the first follow-up year. A Class I outcome at first year indicated 92% probability of seizure remission at 10 years. Negative risk factors included (1) extra-temporal seizure focus (p < 0.001), (2) previous surgery (p < 0.001), (3) male gender (p = 0.035), and (4) normal tissue in pathology (p = 0.038).

Question 18

Outcome Following Frontal Lobe Epilepsy Surgery

Answer 18

Patients with frontal lobe epilepsy (FLE) and an identifiable focal lesion are more likely to achieve seizure freedom than those with poorly defined seizure focus. A review and meta-analysis of 21 studies (total of 1199 patients) with FLE surgery including studies with >= 10 patient and follow-up period >= 48 months showed seizure freedom Engel Class I outcome) in 45.1%. Significant predictors of long-term seizure freedom included lesional origin, abnormal MRI, and localized frontal resection (vs. more extensive lobectomy). In patients with lesional FLE, improved outcome was more likely to be achieved after gross-total resection rather than subtotal lesionectomy [27].
In another study, 25 patients with history of resective surgery after intracranial EEG record- ing were reviewed. A seizure-free (Engel Class 1) outcome was seen in 15 patients (60%), while 10 patients (40%) continued to have seizures (Classes II, III, and IV). Risk factors for lack of seizure freedom included:
– Left frontal lobe epilepsy surgery;
– Dominant hemisphere;
– Patients without aura;
– Interictal epileptiform discharges in scalp
EEG;
– Intracranial EEG widespread (>2 cm) in
contrast to focal seizure onset;
– Shorter latency to onset of seizure spread; and
– Ictal involvement beyond frontal lobe.
Lack of seizure freedom is likely because of widespread epileptogenicity (as indicated by rapid spread of ictal activity). Early resection may improve seizure outcomes of FLE surgery, particularly in children [28, 29].

Question 19

SMA seizures

Answer 19

These seizures presented with tonic posturing of the extremities, usually bilateral, and may appear as “fencer posturing.” Awareness is typically retained during these seizures. The primary epileptogenic zone is usually outside the SMA with rapid spread to the SMA, hence the semi- ology. The interictal, and even ictal, EEG is often unrevealing. If the seizure onset focus is outside the SMA, resection of epileptogenic zone alone, leaving the SMA intact, might be enough [30]. Synchronous interictal/ictal discharges in SMA and primary cortex with a time lag of 25/100 ms. have been reported [31]. Resecting the EEG onset zone within the SMA while sparing pri- mary motor cortex may result in >90% seizure reduction [32]. Following the SMA resection, while preserving primary motor cortex, a transi- tory paresis or severe deficits without permanent loss of motor or speech functions may be seen (typically lasting 24 h), but favorable surgical outcome is common [33].

Question 20

MST

Answer 20

This technique was introduced to spare the eloquent cortex in patients in whom the epilep- togenic zone lies in eloquent cortex. The NST is based on the notion that epileptogenic dis- charges require side-to-side (horizontal) interac- tion of cortical neurons while the major functional properties of cortical tissue depend upon the vertical fibers. Therefore, severing the tangential intracortical fibers in the seizure focus using a small blade is performed, while vertical fiber connections and blood vessels are pre- served [34].
In a report of 21 patients (18 intractable epi- lepsy and 3 Landau–Kleffner syndrome (LKS)) who underwent either resection plus MST (12) or MST alone, in precentral and postcentral regions (follow-up: *1–5 years), significant seizure reduction was seen in 11 of 18 patients (61%) and 3 LKS patients. The latter group who were mute before operation showed significant speech recovery. There were no chronic neurological deficits. Other studies have reported up to 56% seizure freedom and 95% seizure reduction in patients with intractable epilepsy arising from eloquent cortex following combined resection and MST versus no seizure freedom and >50% seizure reduction in those treated with MST alone. Predictor of complete seizure freedom appears to be the disappearance of epileptiform discharges in the post-op EEG. Subtle, but per manent deficits in about one-third of patients with MST performed in eloquent cortex. There- fore, MST surrounding a lesionectomy may minimize the excised volume and improve sei- zure control [35].
However, a meta-analysis of data from 6 major epilepsy centers (211 patients, 53 with MST alone) reported similar results between the MST plus resection and MST alone procedures: The MST plus resection resulted in >95% seizure reduction (GTCS 87%, CPS 68%, SPS 68%), compared to >95% seizure reduction (GTCS 71%, CPS 62%, SPS 63%) in MST alone group.
The outcome seemed to be independent of factors such as EEG localization, MST location, age at onset, or duration of epilepsy. These results suggest that MST may be efficient by itself, with minimal neurologic compromise, and should be investigated as a stand-alone procedure [36].

Question 21

Overall seizure free outcome

Answer 21

The overall seizure-free rates following different types of surgery in different brain regions are reported as follows:
Temporal lobectomy: 55–80% Frontal lobe resections: 5–18% Frontal lobectomy: 23–68% Parietal lobe resections: 45% Occipital resections: 46–88% Hemispherotomy: 60%

Question 22

Epilepsy surgery - long-term outcomes (>=5 years)

Answer 22

Excellent short-term results of resective epilepsy surgery have been well established. Therefore, review and meta-analysis of long-term outcomes of largest case series of patients of any age after resective or non-resective epilepsy surgery have been attempted. After a mean follow-up of >= 5 years, resective surgery resulted in the fol- lowing seizure freedom rates:
Temporal lobe resections: 66%
Occipital and parietal resections: 46%
Frontal lobe resections: 27%
Multiple subpial transections: 16%
Callosotomy (free of most disabling seizures): 35%
Therefore, long-term seizure-free rate follow- ing temporal lobe resective surgery appears to be favorable similar to that of short-term-controlled studies, but it is consistently lower after extra-temporal or palliative surgeries [37].

Question 23

Failed epilepsy surgery and re-operation

Answer 23

Reoperation following a failed previous surgery can be an efficacious and reasonably safe approach. Successful reoperation has been reported in patients with concordance between their postsurgical imaging and electroclinical findings, and no brain trauma or infection before their seizure onset. A review of 15 case series including 402 adult patients reoperated 2– 5.5 years later (reoperation rate: 3.8–14%; follow-up of 6 months–4 years) post-reoperation seizure freedom was reported at 36.6% and complications rate at 13.5% [38].
It is also safe to use subdural grid electrodes in patients with prior craniotomy with favorable long-term seizure-free outcomes. In these patients, ictal onset at the edge of original sur- gical bed (more with lesional epilepsy) seems to be a predictor of seizure freedom [39].

Question 24

Parietal lobe epilepsy

Answer 24

Auras are commonly present in these seizure; 94% of patients report somatosensory auras (painful dysesthesias), vertigo, aphasia, or disturbances of one’s body image. The ictal propagation to the SMA may result in hypermotor manifestations while propagation to the temporo-limbic regions may result in complex visual or auditory halluci- nations and automatisms. Scalp ictal EEG is rarely localizing. Parietal lobe seizures have more variable scatter of interictal EEG discharges and less localizing ictal discharges compared to temporal or frontal lobe seizures. Overall, the semiology is of less value in these patients. High-frequency oscillations (HFOs) may be useful for localization as they more concentrated in seizure focus.
Postoperative sensory deficits such as tempo- rary partial hemisensory or Gerstmann syndrome may be seen when corticectomy involves post-central gyrus. However, resective surgery can result in seizure freedom or significant sei- zure reduction especially when a lesion is pre- sent. The most common pathologies include low-grade tumors, cortical dysplasia, gliotic scars, or cavernous vascular malformations.
Complete or nearly complete seizure freedom has been reported in 65–67.5% of patients with favorable outcome factor being the absence of post-resection epileptiform discharges on the EEG [40–43].

Question 25

Occipital lobe epilepsy

Answer 25

Auras are reported in 88% of these patients. The auras consist of elementary visual hallucina- tions, ictal amaurosis, eye movement sensations, early forced blinking or eyelid flutter, and con- tralateral visual field deficits. There is often eye/head deviation (usually contralateral to the side of seizure origin), loss of awareness, vari- ous types of automatism, fumbling (typical for temporal lobe seizures), and at times asymmet- rical tonic or focal clonic motor patterns (char- acteristic of frontal lobe seizures). Medial or lobar lesions are more likely to cause visual field defects. The scalp EEG is rarely localizing [44, 45]. Intracranial EEG recording correctly iden- tifies occipital lobe seizure origin in most, but not all of these patients. The variability in semiology depends on seizure spread patterns, i.e., medially, laterally, above/below the sylvian fissure, both ipsilateral and contralateral to the seizure origin. fter, focal resection seizure freedom is seen in 46–88% of patients. The most common pathologies include dysplasia, tumors, and glio- sis. Following resection, about 50% of patients will not experience any new visual deficits while new quadrantanopia or hemianopia has been reported in 17%. Tailored resections (e.g., in lateral occipital lesions) may help preserve intact vision in about 38% of patient [46].

Question 26

Insular-opercular seizures

Answer 26

These seizures usually present as nocturnal complex motor seizures. The auras include vis- cerosensitive or somatosensory symptoms. Ictal semiology consists of asymmetric tonic–dystonic posturing and/or hyperkinetic automatisms (bimanual/bipedal activity and ballistic move- ments). Simultaneous insular and opercular ictal discharges are present. Complex motor mani- festations are seen when the seizure spreads to frontomesial regions (cingulum, superior frontal gyrus, and SMA) and/or mesial and neocortical temporal lobe structures.
Favorable outcome can be achieved with insular-opercular cortical resections. The most common underlying pathology is focal cortical dysplasia [47].

Question 27

Epilepsy surgery in children

Answer 27

Epilepsy surgery is commonly performed in children. The long-term outcome (5–21 years) studied in 47 children with age at surgery ranging from 0.5 to 18.7 years (mean 8) reported 49% (23/47) seizure freedom and >75% seizure reduction in 13% (6/47). All of these children were assessed for cognitive function pre- and postsurgery and at follow-up. Twenty-one patients required a reoperation to achieve satis- factory seizure outcomes with low complications rate and no increase in seizures. Cognitive function was well preserved as 76% (34/47) followed their expected cognitive trajectory. Patients who were seizure free showed significant and long-term improvement in their cognitive processing speed, in particular those who were on no ASM [48].

Question 28

Extra-temporal Epilepsy Surgery in Children

Answer 28

In general, surgical outcomes for extra-temporal lobe epilepsy (ETLE) are worse than those for TLE. A meta-analysis of the available literature (17 studies, 95 patients) reported that pathology (cortical dysplasia) and seizure type (CPS) were the positive outcome predictors. Factors con- tributing to less favorable outcome seem to be diffuse nature of pathology involved in ETLE, difficulty localizing the seizure focus in young children, and involvement of “eloquent” cortex [49].

Question 29

Ictal onset high frequency oscillations

Answer 29

Retrospective review of high-frequency oscilla- tions (HFOs > 80 Hz; sampling rate: 2000 Hz) recorded in intracranial EEG in pediatric patients suggest a high prevalence of ictal HFO zones in 93% of patients. Complete resection of ictal HFOs, regardless of the frequency bands, is highly correlated with a favorable surgical outcome. In one series, complete resection resulted in 82% seizure freedom versus 21% after incomplete resection. The most common pathology in these patients was cortical dysplasia [50].