Review
Epilepsy surgery
Department of Clinical and
Experimental Epilepsy, University
College London Institute of
Neurology, London, UK
Correspondence to
Dr Fergus Rugg-­Gunn, Dept
of Clinical and Experimental
Epilepsy, University College
London Institute of Neurology,
London WC1N 3BG, UK; ​f.​rugg-​
gunn@​ucl.​ac.u​ k
Accepted 4 June 2019
Published Online First
16 August 2019
© Author(s) (or their
employer(s)) 2020. No
commercial re-­use. See rights
and permissions. Published
by BMJ.
To cite: Rugg-­Gunn F,
Miserocchi A, McEvoy A.
Pract Neurol 2020;20:4–14.
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Fergus Rugg-­Gunn, Anna Miserocchi, Andrew McEvoy
Abstract appropriately chosen and used anti-­
Epilepsy surgery offers the chance of seizure epileptic drug schedules, either as mono-
remission for the 30%–40% of patients with therapy or in combination.1 Surgical
focal epilepsy whose seizures continue despite treatment of refractory temporal lobe
anti-­epileptic medications. Epilepsy surgery epilepsy improves both the seizure
encompasses curative resective procedures, outcome (58% seizure free) and the quality
palliative techniques such as corpus callosotomy of life, compared with optimal medical
and implantation of stimulation devices. management (8% seizure free).2 Attaining
Pre-­surgical evaluation aims to identify the
seizure freedom is also associated with
epileptogenic zone and to prevent post-­ reduced mortality, for example, from
operative neurological and cognitive deficits. sudden unexpected death in epilepsy.3
This entails optimal imaging, prolonged video-­
About 1.5% of people newly diagnosed
electroencephalogram (EEG) recordings, and with epilepsy may eventually require
neuropsychological and psychiatric assessments; epilepsy surgery. In the UK, this equates
some patients may then require nuclear medicine to approximately 450 patients per year;
imaging and intracranial EEG recording. The best the number of operations performed
outcomes are in those with an electro-­clinically equals the number of emergent new cases.
concordant structural lesion on MRI (60%–70% However, there are also many people
seizure freedom). Lower rates of seizure freedom
with drug-­refractory epilepsy who might
are expected in people with extra-­temporal
benefit from surgery but are not consid-
lobe foci, focal-­to-­bilateral tonic-­clonic seizures,
ered or assessed.
normal structural imaging, psychiatric co-­
The delay between onset of focal
morbidity and learning disability. Nevertheless,
epilepsy and epilepsy surgery is still 15–20
surgery for epilepsy is under-­used and should be
years,4 5 indicating delay and under-­
considered for all patients with refractory focal
referral for this potentially life-­changing
epilepsy in whom two or three anti-­epileptic
treatment.
medications have been ineffective.
Most patients with refractory focal
epilepsy can be appropriately considered
for surgery. Over half of those referred
Introduction will not be suitable for current resective
Approximately 60%–70% of people with options, but they may be suitable for a
focal epilepsy become seizure free with palliative procedure or implantation of a
medication. However, in those 30%–40% stimulator.
whose seizures continue despite medi- In general, about 50% of those who
cation, clinicians should consider other undergo initial, non-­invasive, pre-­surgical
options, such as epilepsy surgery, vagus investigations do not proceed further;
nerve stimulation or ketogenic diet.
25%–40% are offered a resection without
Surgical success may be defined as the
needing further investigations, and
complete cessation of seizures without
10%–30% require intracranial electroen-
post-­operative cognitive, psychiatric or
cephalogram (EEG) recordings.
neurological dysfunction. The outcome
of surgery typically represents a balance
between seizure control and post-­ Who should be considered/referred for
operative deficit. The pre-­surgical inves- epilepsy surgery?
tigative pathway aims to inform this ►► All adults with refractory focal epilepsy
(adequate trials of two or more appropri-
process by localising the epileptogenic
zone, eloquent cortex, and major white ately selected anti-­epileptic medications)
matter tracts. should be considered for resective/curative
Refractory epilepsy is the failure surgery
to achieve sustained seizure freedom ►► All adults with refractory focal or gener-
despite adequate trials of two tolerated, alised epilepsy who are not candidates
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192 1 of 13
Review

completely disconnected) to give seizure freedom. No

Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
single pre-­operative investigation can characterise the
epileptogenic zone completely reliably, and even when
combining various investigative modalities there may
still be variable concordance.(figure 1)
When pre-­ operative non-­ invasive investigations
have a high degree of congruence between these
zones, it may be possible to recommend surgery with
predictable levels of benefit and risk. However, if
non-­invasive investigations are discordant, proceeding
directly to resective surgery may be rejected in favour
of establishing more definitive localising data using,
for example, invasive EEG recordings (figure 2).

Identifying the seizure focus: non-invasive techniques

Figure 1 Sir Victor Horsley (far left) in the operating theatre at
Identifying the epileptogenic zone starts with a
Queen Square, 1906. Accompanied by Dr Powell, anaesthetist
and Professor Theodor Kocher from Berne (far right), a strong
detailed elucidation of the nature and tempo of the
advocate of asepsis during surgery. seizure semiological characteristics. Initial investiga-
tions include detailed imaging and interictal and ictal
EEG recordings (figure 2). High-­resolution anatom-
for resective surgery should be considered for palliative ical MRI with an epilepsy-­dedicated protocol should
procedures, such as vagus nerve stimulation be performed (ideally on a 3-­Tesla MRI scanner) and
►► A low IQ or memory impairment is not a contraindica- evaluated by an expert neuroradiologist. Up to 86% of
tion to resective surgery
►► Older patients should be considered for surgery but the
risk of complications is higher
►► A history of long-­ term psychiatric disorder does not
exclude a patient from resective surgery but close psychi-
atric supervision post-­operatively would be mandatory.
►► Bilateral interictal epileptiform activity is not a contrain-
dication to resective surgery; unilateral onset seizures
often have bilateral interictal epileptiform activity
►► People with refractory focal epilepsy and a normal
structural MRI scan of brain should be considered for
surgery; other investigations may identify a single epilep-
togenic zone amenable to surgical resection
►► Multiple or diffuse lesions on MRI are not a contrain-
dication to surgery; seizures may arise from only one of
the visible abnormalities or from a part of the lesion.
►► People whose seizure semiology suggests involvement
of primary eloquent cortex should be considered for
surgery; essential functions can be localised and the
symptomatic zone may be distinct from the epilepto-
genic zone due to seizure propagation.
Note that all patients with refractory focal epilepsy
and an apparently clinically concordant lesion on
initial imaging still need comprehensive pre-­surgical
evaluation; a lesion’s epileptogenicity must always be
confirmed before resection. The exception is when
the primary rationale for surgery is tumour resection,
rather than seizure control.

Pre-surgical evaluation
The principal aim of pre-­ surgical evaluation is to
determine the epileptogenic zone and its relationship
to eloquent areas of the brain. The epileptogenic zone
is a theoretical construct, defined as the minimum
amount of cortex that must be resected (inactivated or
Figure 2 Epilepsy surgery assessment (reproduced from
Duncan JS and colleagues by permission of The Lancet).34 EEG,
electroencephalogram; ¹⁸F-­FDG, ¹⁸F-­fluorodeoxyglucose; fMRI,
functional MRI; MEG, magnetoencephalography; PET, positron
emission tomography; SPECT, single-­photon emission CT.

2 of 13 Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192


Figure 3 1.5T (left) and 3T (right) axial T2-­weighted MR
images showing left frontal focal cortical dysplasia. The lesion is
more readily identifiable on the 3T-­derived images.
cases of hippocampal sclerosis may remain undetected
using a standard MRI sequence reported by a non-­
expert radiologist.6 A 3-­Tesla MR scanner improves
the identification of structural lesions by up to 20%
compared with a 1.5-­Tesla scanner (figure 3).7
Prolonged video-­ EEG telemetry is mandatory,
often with anti-­epileptic drug reduction to increase
the number of seizures recorded within a reasonable
time frame. During and immediately after seizures,
it is important to perform neurocognitive testing
to establish functional deficit, to aid localisation.8 It
is not appropriate to construct a surgical hypothesis
using only interictal data as this only poorly defines
the epileptogenic zone.
Figure 4 ¹⁸FDG PET scan showing left temporal
hypometabolism in a 32-­year-­old man with left temporal
lobe epilepsy and a normal MRI scan of brain (A, axial slice;
B, coronal slice). (C) Surface rendered left lateral projection
of glucose uptake. (D) Results of a statistical voxel-­based
comparison of surface-­rendered glucose uptake in the patient,
compared with a set of control data (using Neurostat-­3D SSP
software). There is an area of hypometabolism (green) in the
left temporal lobe (D). The patient became seizure free after left
temporal lobe resection (adapted from Rathore and colleagues
by permission of Elsevier35). FDG-­PET, fluorodeoxyglucose-­
positron emission tomography.
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192
Review
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Figure 5 Ictal SPECT scan in patient with temporal lobe
epilepsy and previous middle cerebral artery territory infarct
with hemiparesis. Coronal T1-­weighted MR image showing
areas of post-­infarction encephalomalacia (left) and with
superimposed subtraction SPECT perfusion image (middle). This
shows an area of ictal hyperperfusion in the left temporal lobe,
when the ictal perfusion scan is subtracted from the interictal
SPECT scan (right). SPECT, single-­photon emission CT.
If the initial investigations are discordant or, if the
MRI scan of brain is normal or non-­definitive, nuclear
medicine studies such as fluorodeoxyglucose-­positron
emission tomography (figure 4) and ictal single-­
photon emission CT (figure 5) are used to generate
a hypothesis that may then be tested with intracra-
nial EEG recordings. Additionally, more advanced
developmental MRI or neurophysiological tech-
niques—such as magnetoencephalography (figure 6)
or electrical source imaging—may help to localise the
seizure focus.
Figure 6 MEG informing the placement of intracranial EEG
electrode placement. (A) axial T2-­weighted MR image of
30-­year-­old man with pharmaco-­resistant frontal lobe epilepsy
and normal structural imaging. MEG dipoles reconstructed
from interictal epileptiform activity detected with a 275-­channel
whole-­head magnetometer (E) are superimposed on axial
(B), sagittal (C) and coronal (D) T1-­weighted MR images and
localise to the anterior cingulate. This was evaluated with a
mesial frontal strip of intracranial EEG electrodes (F) and seizure
onset (red circles) co-­localised to the MEG dipoles. The patient
subsequently underwent resection (post-­operative axial T2-­
weighted MR image (G) and remains seizure free for 5 years.
Histopathological evaluation identified focal cortical dysplasia.
EEG, electroencephalogram; MEG, magnetoencephalography.
3 of 13
Review
Figure 7 Three-­dimensional surface-­reconstructed MR image
(left) and perioperative photograph (right) showing placement
of depth electrodes as part of a stereo electroencephalogram
study.
Identification of the seizure focus: invasive techniques
Up to 20%–30% of surgical candidates need intracra-
nial EEG recordings to define the epileptogenic zone.
The aim of invasive EEG recording is to acquire neuro-
physiological data to support or disprove a hypoth-
esis regarding the site of seizure onset. Typically, this
is required for non-­lesional focal epilepsy or if non-­
invasive investigations are non-­localising or discordant
(figure 2).
The type of intracranial recording depends on the
suspected pathophysiological substrate of the epilepsy
and its location. Intracranial EEG recording mainly
involves two techniques:
►► Depth electrode implantation uses multiple electrodes
stereotactically implanted into the brain parenchyma
via small screws fixed to the skull (figures 7–9). This
stereo-­EEG technique allows recording from both deep
Figure 8 Three-­dimensional surface-­reconstructed MR image
with superimposed vasculature derived from digital subtraction
or CT angiography and pre-­operative depth electrode EEG
placement strategy for stereo EEG study using neuronavigation
software. Each EEG electrode is depicted in a different colour
(courtesy of Vejay N Vakharia, UCL Institute of Neurology). EEG,
electroencephalogram.
4 of 13
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Figure 9 Axial (top left), sagittal (top right) and coronal
(bottom left) T1-­weighted MR images and three-­dimensional
surface reconstructed image (bottom right) with superimposed
depth electrode electroencephalogram proposal planned with
neuronavigation software, before surgery. Electrode entry
point, trajectory, depth and target localising information can
be uploaded to the surgical navigation software during surgery
to guide accurate and safe placement (courtesy of Vejay N
Vakharia, UCL Institute of Neurology).
and superficial areas. Its overall morbidity is about
1.3%, equating to a risk of 1 in 287 electrodes, including
a 1% risk of haemorrhage.9 The implantation can be
performed with a frame-­based stereotactic approach or
using frameless image-­guidance systems that promise to
simplify the pre-­ surgical planning of electrode place-
ment (figures 8 and 9). Both frame-­based and frameless
approaches can be robotically assisted.
►► Subdural electrodes (strips and grids) are placed directly
on to the brain surface (figures 10–12). Subdural strips
can be placed through simple burr holes, whereas grids
require a craniotomy and can record from a larger area
of contiguous cortex. They are frequently used when
epileptogenic lesions are adjacent to eloquent cortex.
Subdural grids enable detailed extra-­ operative direct
cortical stimulation, facilitating the mapping of eloquent
cortex. The main advantage of this technique is a more
comprehensive cortical stimulation study, compared
Figure 10 Intra-­operative photographs showing placement of
subdural electroencephalogram recording strips and grids.
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192

Figure 11 Intra-­operative photographs and reconstructed
three-­dimensional images: before placement of subdural
electroencephalogram grid (left) and after placement (right).
with stereo-­
EEG, where cortical sampling is spatially
more limited.
The duration of invasive monitoring depends on
the seizure frequency, the success of any planned stim-
ulation and patient compliance. It may need several
weeks of depth electrode recording to characterise a
patient’s seizure disorder fully. In contrast, subdural
grid recordings seldom extend beyond 10–14 days, as
these patients often have a higher seizure frequency,
and the procedure carries a higher risk of infection.
Figure 12 Post-­operative MRI scan of brain coronal T1-­
weighted images (top row) and corresponding skull radiograph
(bottom row) showing placement of subdural intracranial
electroencephalogram recording electrodes.
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192
Review
Invasive monitoring may be stopped at any stage if
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
there is a clinically significant adverse event, such as
intracranial haematoma (<5% of cases) or infection
resulting from the wires passing through the scalp (2%
of studies).10 These risks can be reduced by careful
intra-­operative technique, appropriate post-­operative
nursing care and prophylactic antibiotics. Invasive
intracranial EEG studies are time-­consuming, expen-
sive, have an inherent risk of complications, and
require numerous personnel and access numerous
allied investigations. This limits the number of neuro-
science centres that can support a comprehensive
epilepsy surgery programme.
About 40% of patients who undergo invasive
recording are deemed unsuitable for resective surgery,
for three main reasons: the epileptogenic zone cannot
be satisfactorily determined, there are multiple poten-
tial seizure foci or the epileptogenic zone is situated in
eloquent cortex.
Preservation of cognitive and neurological function
Neuropsychological testing can help to predict
the post-­ operative cognitive outcome and seizure
control.11 Whereas bilateral hippocampal resection
results in profound anterograde amnesia, unilateral
temporal lobe resections may result in material-­specific
memory dysfunction. Approximately 30% of patients
undergoing dominant temporal lobe resection develop
difficulties in verbal memory processing and word-­
retrieval; a similar proportion of patients undergoing
non-­dominant temporal lobe resection develop diffi-
culties with non-­verbal or visual memory processing.
A decline in verbal memory is generally more disabling
than a decline in visual memory. Those at greatest
risk of a troublesome decline in language and verbal
memory are high-­ functioning people who undergo
an anterior temporal lobe resection in the speech-­
dominant hemisphere. A proportion of people note
an improved memory after temporal lobe resection,
particularly if the resection is on the non-­dominant
side.
Pre-­operative neuropsychological scores, in conjunc-
tion with MRI and other clinical data, can be used to
predict post-­ operative neuropsychological change
using logistic regression techniques.12 Those at high
risk of a significant memory decline can be advised
pre-­operatively and can be trained in compensatory
strategies before the surgery.
Functional MRI (fMRI) can lateralise and localise
cerebral areas involved in language function. Language
lateralisation assessed using fMRI language tasks
correlates well with that assessed using the carotid
amytal test.13 Thus in most epilepsy surgery centres,
language fMRI has largely replaced the amytal test.
Resections close to eloquent language cortex require
a more detailed and accurate assessment of the anatom-
ical relationship between seizure focus and language
areas than can be assumed from an fMRI activation
5 of 13
 Review
Figure 13 Intra-­operative photographs showing the use of
cortical stimulation (left) to identify eloquent cortex and (right)
post-­operative surgical field. Note the preservation of branching
vasculature.
pattern; it may be necessary to perform electro-­cortical
stimulation or an awake resection (figure 13). Awake
craniotomy poses significant challenges to both
surgeon and anaesthetic team; seizures may occur
during the procedure or patients may become agitated
or distressed. There is neither immediate control of
blood pressure nor the possibility of hyperventilation
to reduce intracranial pressure.
Tractography derives from diffusion-­tensor imaging;
it allows identification of nerve fibre tracts within
the brain, and demonstrates the structural basis of
connectivity between brain regions. Visual pathway
tractography can predict the likelihood and extent of
a visual field defect following anterior temporal lobe
resection. Typically, up to 10% of patients undergoing
anterior temporal lobe resection develop a significant
superior quadrantanopic field defect post-­operatively,
which can prevent restoration of driving eligibility
Figure 14 Post-­operative coronal (top left) and sagittal
(top right) T1-­weighted MR images with superimposed pre-­
operative Meyer’s loop optic radiation derived from diffusion
tensor imaging tractography. There is anterior extension of
Meyer’s loop into the temporal pole, which resulted in a
quadrantanopic visual field defect, a recognised complication
of anterior temporal lobe resection. Intra-­operative
photographs (bottom row) through the surgical microscope
with optic radiation tractographic information (bottom right)
superimposed to inform the surgeon and reduce the risk of
post-­operative field defect.
6 of 13
(figure 14). Tractography data made available to the
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
surgeon through a visual overlay when using the oper-
ating microscope has reduced the incidence of post-­
operative visual field deficit.14
Epilepsy surgery multidisciplinary meeting
A multidisciplinary and systematic approach to inves-
tigations is essential to a surgical pathway (figure 2).
During the epilepsy surgery multidisciplinary team
meeting—attended by epileptologists, neurosur-
gical team, neurophysiologists, neuropsychologists,
neuropsychiatrists and epilepsy specialist nurses—the
seizure history is presented and then each investigation
discussed in detail. Having reached consensus on a
potentially curative or palliative surgical approach, the
team formulates a detailed management plan. Subse-
quently, this risk:benefit analysis is discussed in detail
with the patient and along with written information
(Box 1). Patients must be given realistic expectations
of what may be achieved, and what may be the nega-
tive consequences of epilepsy surgery(Box 2).
Surgical Procedures: curative
Lesionectomy
The increased anatomical resolution of modern MRI
has led to the identification of many more cortically
based lesions. Small lesions such as cavernomas, focal
areas of cortical dysplasia (figure 15) and indolent
tumours such as dysembryoplastic neuroepithelial
tumours are highly epileptogenic, and their resection
is associated with a high rate of seizure freedom.
Interventional MRI allows documentation of
complete lesion resection before completing the
surgical procedure. It also enables the surgical naviga-
tion software to be recalibrated during the operation,
improving its accuracy. Thus, interventional MRI can
potentially improve the rate of seizure freedom from
surgery, and reduce the risk of neurological deficit
(figures 16 and 17).
Lobectomy
Temporal lobe
Anterior temporal lobe resection, including of the
mesial temporal lobe structures, accounts for about
half of operative procedures performed in specialist
epilepsy centres. Contemporary approaches attempt
to limit the size of the neocortical resection to mini-
mise neurocognitive sequelae, using either the
method described by Spencer15 or selective amygda-
lohippocampectomy. Anterior temporal lobe resection
that includes removal of up to 4.5 cm of neocortex
shows a trend towards improved seizure outcomes
compared with selective amygdalohippocampectomy,
and with minimal differences in neuropsychological
outcome. Despite this, there is still controversy about
the different approaches. Having familiarity with
a specific approach or technique is associated with
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192
 Review

Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Box 1. Examples of pre-­surgical discussions with patients

Patients with refractory focal epilepsy and electroclinically concordant hippocampal sclerosis
►► ‘We have now had a chance to review the results of all of your investigations in the surgery multidisciplinary
meeting. Your MRI brain scan has shown evidence of scarring in the inner aspect of the left temporal lobe, called
hippocampal sclerosis, and this is a common finding in people with difficult-­to-­control seizures. The brain-­wave EEG
recordings have confirmed that your seizures start from this area and the memory and language tests performed by
the neuropsychologists showed that your memory for words is less efficient than your memory for pictures and places.
Again, this suggests that the left temporal lobe is the region of your brain from which your seizures arise.
►► It was the consensus of the meeting that we can offer surgery to help your epilepsy. Specifically, we propose removing
part of the temporal lobe including the area of scarring. We estimate that by doing so, there is a 60%–70% chance of
stopping your seizures with an additional 15% chance of significantly improving your seizures but without stopping
them completely. There is a 1 in 100 chance of your seizures worsening after surgery.
►► Of course, no operation on the brain can take place without risk. I would estimate that there is a 1 in 100 chance of
developing a serious new problem, such as weakness of an arm or leg or difficulty with speech and this may not be
recoverable. The risk of a less severe but noticeable problem such as loss of vision in the top right hand corner of both
eyes that would prevent you from driving even if you became seizure free is about 1 in 10. There is a 1 in 3 chance that
your memory for words may not be as good following surgery but if the seizures are stopped we would expect this to
level out over time. Also, about 1 in 3 people undergoing surgery of this kind experience mood disturbance afterwards
but this is typically temporary and improves after 3–6 months. Whilst the risks may seem worrying, it is important to
consider that the risks of surgery are very similar to the risks you run from your epilepsy, as it currently is, over a 2-­year
period. Furthermore, medications alone are associated with a less than 5% chance of stopping your seizures, and if
seizure continue, it is likely that your memory will worsen over the coming years’.
Patients with refractory focal epilepsy and normal optimal imaging requiring an intracranial EEG study.
►► ‘We have now had a chance to review the results of all of your investigations in the surgery multidisciplinary meeting.
Your MRI brain scan appears entirely normal but the PET scan, which looks at the amount of sugar the brain uses,
suggests that your seizures may be arising from the front of the brain on the right hand side. The brain-­wave EEG
recordings suggest that your seizures start from this area and the memory and language tests performed by the
neuropsychologists are consistent with this.
►► It was the consensus of the meeting that we are not able to proceed directly with surgery to help your epilepsy but that
further investigations are required to see whether this will be possible in due course. Specifically, we propose performing
a surgical operation to place EEG recording electrodes inside the brain, through small holes in the skull. This is called
intracranial EEG. The aim of this is to find out very accurately which area of the front of the brain gives rise to the
seizures and to see if an operation to remove this area would stop your seizures. We estimate that there is a 50%–60%
chance of the intracranial EEG test identifying an area of the brain from where your seizures arise. If this is removed at
an operation afterwards, we estimate that there is a 50% chance that your seizures would stop. This means that if we
proceed at this point, you have a 30% chance overall of becoming seizure free with surgery.
►► Of course, no operation on the brain can take place without risk. I would estimate that as a result of the operation to
place electrodes inside the brain, there is a 2–3 in 100 chance of developing a serious new problem, such as a bleed
inside the brain. This may lead to symptoms similar to a stroke including weakness of an arm or leg and a second
emergency operation might be needed. There is also a 2–3 in 100 risk of infection. If one of these major complications of
surgery arose the EEG recordings might need to be stopped early, even if useful information about your seizures had not
yet been obtained.
►► Should you go on to have surgery to remove a part of the brain that we think is giving rise to your seizures, I would
estimate that there is a 1 in 100 chance of developing a serious new problem, such as weakness of an arm or leg or
difficulty with speech and this may not be recoverable. There may be other risks of developing a new problem depending
on exactly what operation is performed and this will be discussed with you in detail before proceeding.
►► Whilst the risks may seem worrying, it is important to consider that the risks of surgery are very similar to the risks you
run from your epilepsy, as it currently is, over a 2-­year period. Furthermore, medications alone are associated with only a
5% chance of stopping your seizures’.

improved seizure outcome and a lower morbidity, and
this should therefore influence the surgical strategy.
The initial seizure-­free rate following resection of
hippocampal sclerosis (figure 18) is approximately
75%–80%, and approximately 70%–75% for resection
of other temporal lobe lesions.4 There is gradual attri-
tion of seizure freedom over subsequent years so that
40%–50% can expect to have remained totally seizure
free after 20 years. The failure to achieve seizure
freedom may be due to insufficient resection of the

Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192 7 of 13

 Review

Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Box 2. Benefits/risks of surgery

Potential benefits
2 25
►► Seizure freedom
36
►► Reduced seizure severity
37
►► Reduced medication load
►► Cognitive gains from reduction of both medication
load and seizure activity38
►► Reduced risk of sudden unexpected death in epilepsy
(SUDEP) and injury3 39
40
►► Possible improved long-­term psychiatric outcomes
41
►► Improved quality of life

Risks
26
►► Perioperative mortality and morbidity
►► Post-­operative neurological and cognitive deficits42 43 Figure 16 Interventional MRI operating suite. The red line
particularly if seizures continue post-­operatively42 represents the 5-­gauss static magnetic field strength around the
►► Possible short-­term and de novo long-­term psychiatric scanner.
complications44 45

structural lesion, the location of eloquent cortex as

epileptogenic mesial temporal structures, seizures
arising from the contralateral mesial temporal lobe,
lateral temporal neocortical epilepsy, dual pathology,
secondary epileptogenesis and extra-­ temporal lobe
epilepsy mimicking temporal lobe epilepsy, as with
insula or posterior cingulate foci.
Extra-temporal lobe:
Seizure outcome is typically less good for extra-­
temporal resections than with temporal lobe resec-
tions, particularly in non-­lesional cases4 but may be as
high as 50%–60% following careful evaluation with
stereo EEG.16 Depending on the pathology, it may be
necessary to make large resections of the epileptogenic
zone, thus putting eloquent cortex at risk. Neverthe-
less, pre-­
operative discussion may have determined
a patient preference for risking neurological deficit
rather than having persistent seizures. Certain post-­
operative neurological and neuropsychological deficits
are often predictable, such as a hemianopic field deficit
following occipital lobe resections, or a Gerstmann’s
syndrome or hemisensory impairment following
parietal lobe surgery. The extent and location of the
resection depends upon the presence and extent of a
defined by fMRI or electrical stimulation studies, the
position and trajectory of white matter tracts, and the
location of major blood vessels.
Hemispherotomy
Hemispherotomy inevitably leads to profound neuro-
logical deficit, including hemiplegia and hemianopia;
this is therefore most appropriate for those with pre-­
existing deficit (figure 19). If patients can walk pre-­
operatively, most remain able to do so. Typically, there
is loss of fine motor skills in the contralateral upper
and lower limbs, but cognitive function is stable. Hemi-
spherectomy used to cause long-­term complications in
up to one-­third of patients, including superficial cere-
bral haemosiderosis. As a result, alternative techniques
were developed, including a functional hemispherec-
tomy, in which the temporal lobe and central cortex
are removed and the corpus callosum and frontal and
occipital cortex disconnected. The seizure outcomes

Figure 17 Interventional MRI. Intra-­operative MR images

Figure 15 Coronal fluid-­attenuated inversion recovery MR
images showing right frontal lobe focal cortical dysplasia with
typical T2 hyperintense tail towards the ventricle seen in focal
cortical dysplasia type IIb.
showing desired area of resection (purple outline) and
corticospinal tracts derived from diffusion tensor imaging
tractography (orange outline) superimposed on intra-­operative
MR images. The pre-­operative (left cluster) and perioperative
(right cluster) investigations show the extent of resection and
post-­operative preservation of corticospinal tracts.

8 of 13 Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192


Figure 18 Coronal T1-­weighted (left) and fluid-­attenuated
inversion recovery (right) MR images showing left mesial
temporal lobe sclerosis.
remain unchanged, but the complication rate has
significantly improved.
The success of hemispherectomy depends on the
underlying pathology, with excellent outcomes and
seizure freedom rates approaching 75%–85% for
pathologies such as Rasmussen’s encephalitis and focal
infarcts, but with a poorer outcome for patients with
hemi-­megalencephaly.17
Surgical procedures: Palliative
The objective of these functional procedures is to
palliate rather than to cure the epilepsy. They should
be offered only if resective surgery is considered to be
inappropriate or too risky.
Corpus callosotomy
The primary indication for corpus callosotomy
is atonic drop attacks, although it is effective for
other seizure types. The disconnection slows inter-­
hemispheric seizure propagation and provides patients
with a warning or disrupts a seizure, the expression
of which relies on synchrony . About 74% of people
have favourable outcomes with corpus callosotomy,
including 39% who stop having drop attacks.18 The
operation may cause either immediate or delayed
symptoms of disconnection. In order to minimise this
risk in adults, the callosotomy is usually carried out in
Figure 19 Axial fluid-­attenuated inversion recovery (left)
showing left-­sided Rasmussen’s encephalitis and coronal T1
(right) MR images showing left hemispherectomy.
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192
Review
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Figure 20 Coronal T1-­weighted (left) and axial diffusion
tensor imaging tractography (right) MR images showing
anterior corpus callosotomy with interruption of anterior but
not posterior callosal tracts.
two stages, with the anterior two-­thirds of the corpus
callosum being divided first and the posterior third
divided later (figure 20).
Stimulation techniques
Vagus nerve stimulation
The pathophysiological basis of periodic vagus nerve
stimulation has not been fully elucidated but may
involve autonomic nervous pathways and augmented
function of neurotransmitters such as gamma-­
aminobutyric acid (GABA). Besides intermittent
stimulation, the patient or companion can also effect
on-­demand stimulation. The newest devices also stim-
ulate on detection of ictal tachycardia.
The left vagus nerve is used to avoid cardiac side
effects, and the electrode placed on the nerve in the
neck between the common carotid artery and the
internal jugular vein. Side effects include hoarseness
and coughing during stimulation and neck discomfort.
The beneficial effect of vagus nerve stimulation may
take up to 2 years to emerge. Long-­term studies have
shown that although very few people become seizure
free, up to 43%–64% have their seizure frequency
reduced by 50% or more.19
Intracranial stimulation
Traditional ‘open-­loop’ deep-­brain stimulation tech-
niques use continuous or scheduled stimulation, and so
do not depend on the presence of epileptiform activity.
In one study of people with focal epilepsy, bilateral
stimulation of the anterior nuclei of the thalamus was
associated with an immediate mean decrease in seizure
frequency of 29%, and of 56% at 2 years. The proce-
dure was generally well tolerated without symptom-
atic haemorrhage or infection, but the treatment group
developed more depression and memory difficulties.20
‘Closed-­
loop’ or responsive cortical stimulation
has been an important development. Here, the stim-
ulation is provided only having detected abnormal
ictal or interictal epileptiform discharges. The
Neuropace responsive neurostimulator (RNS) delivers
9 of 13
Review
a pre-­symptomatic short train of electrical pulses to
the brain through implanted leads, on detection of
abnormal electrical activity via an implanted strip
electrode on the brain surface. A multicentre, double-­
blind, randomised control trial showed a 38% reduc-
tion in mean seizure frequency in the treatment group
compared with a 17% reduction in the sham group.21
As with vagus nerve stimulation, the efficacy improved
over time, with a 66% seizure reduction at 6 years.
Mild adverse events, such as implant site pain, head-
ache and dysaesthesia, were common in both the
treated and sham groups.
Other implantable responsive devices currently in
development use optogenetics, local cooling or drug-­
delivery systems; trials of these in humans will begin in
the next few years.
Outcome
Seizure control
The outcome from epilepsy surgery is based on several
facets including seizure control, neuropsychological
development, neurological deficit, quality of life and
psychosocial adjustment. Of these, seizure control
is the one most commonly ascertained. In one large
cohort study of almost exclusively curative procedures,
the average post-­operative seizure remission rate was
52% at 5 years and 47% at 10 years, with a range of
between 40% for extra-­ temporal lesionectomy and
64% for hemispherectomy.4
Early seizure recurrence predicted a worse seizure
outcome, in line with other studies.22
Patients with an identified epileptogenic lesion are
two to three times more likely to become seizure free
post-­operatively than patients with normal imaging.23
The factors associated with an increased risk of
seizure recurrence post-­ operatively include normal
MRI scan of brain, a history of focal-­to-­bilateral tonic-­
clonic seizures, a psychiatric history, extra-­temporal
rather than temporal lobe surgery, older age, and
having tried a higher number of medications before
surgery (figure 21).24
Figure 21 Probability of seizure freedom in selected groups.
The group with the best chance of seizure freedom (HS), no
SGTCS, temporal surgery, no psychiatric history and no learning
disability) compared with single significant prognostic features.
(Reprinted with permission.24 (2017). BMJ. All rights reserved.)
HS, hippocampal sclerosis; SGTCS, secondarily generalised
tonic-­clonic seizures.
10 of 13
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Figure 22 Coronal T1 MR image (left) showing right
inferior temporal lobe encephalocoele (circled) with intra-­
operative photograph of defect in floor of middle cranial
fossa (middle). Post-­operative coronal T1-­weighted MR image
showing evidence of right temporal lobe resection and two
complications of surgery—a right-­sided subgaleal cerebrospinal
fluid collection and a left-­sided septated subdural haemorrhage.
Complications
The neurological complications of epilepsy surgery
depend largely on the location and extent of the
surgical resection. The overall complication rate is
around 7%–8%,25 though this is higher in people aged
over 50 years at 6%–25%. Most neurological deficits
are predictable (such as visual field deficit); the risk
of new, long-­term, unexpected neurological complica-
tions is low, at less than 5%. Transient complications
such as infection or cerebrospinal fluid collections are
more common (figure 22).26
Psychiatric disorders are common in people with
epilepsy. Having a prior or current history of psychi-
atric disorders is associated with a lower chance of
seizure freedom following surgery, but this is not a
contraindication for surgery. Following successful
surgery, there may be short-­ term worsening of
psychiatric symptoms, in particular anxiety, but
this is often followed by long-­term improvement.27
Nevertheless, de novo psychiatric disorders, such as
depression, anxiety or psychosis may develop in up
to 26% of people after temporal lobe surgery. Careful
post-­operative psychiatric supervision is therefore
important.27 Interestingly, new onset of dissociative
(non-­epileptic) seizures may develop in 4%–8% of
people following surgery.28
Medication withdrawal
Patients who become seizure free following surgery
may wish to consider subsequent medication with-
drawal. However, the risk of seizure recurrence, the
nature of prognostic factors, and the timing and rate of
medication withdrawal remain unclear. Some studies
suggest a high rate of recurrence, with failure to
recapture seizure freedom on subsequently restarting
medication,29 whereas others report that medication
withdrawal is not clearly associated with an increased
risk.30 In general, about one-­third of people withdraw
medication completely after surgery.4 31 However,
with no clear data or guidelines, we prefer to maintain
pre-­operative medication for at least 12 months after
surgery, before gradually reducing it to monotherapy
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192

Figure 23 MRI-­guided stereotactic laser
amygdalohippocampectomy. Depiction of irreversible damage
zone estimate at completion of ablation in the axial (A) and
sagittal (D) planes. Post-­ablation FLAIR image in the axial plane
(B) and axial T1 image acquisition with gadolinium contrast
enhancement (E). The post-­operative FLAIR image with the
estimated irreversible damage zone superimposed (C). Arrows
indicate ablation damage zone estimate; arrowheads point
to location of the implanted laser fibre assembly (reproduced
from Wicks RT and colleagues by permission of Congress of
Neurological Surgeons).32 FLAIR, fluid-­attenuated inversion
recovery.
during the second post-­operative year, provided there
is ongoing seizure freedom.
Recent developments
There is increasing interest in minimally invasive
lesioning epilepsy surgery techniques. The philosophy
of these treatments is that it may be possible to destroy
or modulate a deep epilepsy focus without needing to
sacrifice or damage potentially normally functioning
surrounding brain tissue. They offer the possibility of
stopping seizures yet with improved neurocognitive
outcomes.
One such technique is thermal ablation, also called
laser interstitial thermal therapy.32 The low-­ voltage
laser is introduced via an optic fibre using MRI guid-
ance and a coagulative necrotic lesion is created using
MRI thermal maps in real time (figure 23). This is
undertaken as a day case procedure and is becoming
widely adopted for hippocampal sclerosis, hypotha-
lamic hamartomas and cortical dysplasia. Compared
with gamma-­knife radiotherapy, there are no issues
with long-­term radiation risks and the region of abla-
tion is better demarcated.
In terms of outcome, real-­time MR stereotactic laser
amygdalohippocampectomy generally gives lower
seizure free rates (53%–80% at 1–3 years) than open
surgery in the small series completed to date. Hippo-
campal sclerosis on pre-­operative imaging predicts a
better outcome. Neurological complications, such
as visual field defects, intracranial haemorrhage and
cranial nerve defects occur at a rate comparable with
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192
Review
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
Key points
►► All patients with refractory focal epilepsy should be
considered for resective epilepsy surgery.
►► Pre-­surgical evaluation—to characterise the
epileptogenic zone and to prevent post-­operative
neurological, psychiatric and cognitive deficit—
requires a multidisciplinary approach with a
comprehensive investigative pathway.
►► Surgery carries a risk of permanent neurocognitive or
psychological deficit but this needs to be balanced
against the risks of ongoing seizure activity.
►► New techniques, such as laser ablation and responsive
stimulation, aim to reduce epileptic activity while
preserving neurocognitive function better than with
standard resective surgery.
open surgery, but neuropsychological decline (partic-
ularly of verbal memory) is less common.33 We await
further assessment of the long-­term seizure outcome
and procedural risk from this and other lesioning tech-
niques (such as high-­frequency ultrasound ablation)
and data on their potential use for more challenging
surgical targets such as periventricular nodular hetero-
topia or corpus callosotomy.
Acknowledgements We are very grateful to Professor John
Duncan for reviewing this manuscript.
Contributors FR-­G drafted and revised the manuscript. AME
and AM provided figures and revised the manuscript.
Funding The authors have not declared a specific grant for this
research from any funding agency in the public, commercial or
not-­for-­profit sectors.
Competing interests FR-­G has received a speaker's honorarium
from LivaNova. AME has received honoraria from Baxter,
UCB and Integra, sponsorship to attend meetings from
Leksell, Medtronic, Brainlab, Modus V and hospitality from
Livanova. AM has received sponsorship to attend meetings
from Medtronic and Modus V and received hospitality from
Livanova. Research funding via Wellcome has Medtronic
named as the preferred commercial partner.
Patient consent for publication Not required.
Provenance and peer review Commissioned. Externally peer
reviewed by Khalid Hamandi, Cardiff, UK.
References
1. Kwan P, Arzimanoglou A, Berg AT, et al. Definition of drug
resistant epilepsy: consensus proposal by the ad hoc Task Force
of the ILAE Commission on therapeutic strategies. Epilepsia
2010;51:1069–77.
2. Wiebe S, Blume WT, Girvin JP, et al. Effectiveness, efficiency
of surgery for temporal lobe epilepsy study G. A randomized,
controlled trial of surgery for temporal-­lobe epilepsy. N Engl J
Med 2001;345:311–8.
3. Bell GS, Sinha S, Tisi Jd, et al. Premature mortality in
refractory partial epilepsy: does surgical treatment make a
difference? Journal of Neurology, Neurosurgery & Psychiatry
2010;81:716–8.
11 of 13
 Review
4. de Tisi J, Bell GS, Peacock JL, et al. The long-­term outcome
of adult epilepsy surgery, patterns of seizure remission, and
relapse: a cohort study. The Lancet 2011;378:1388–95.
5. Vakharia VN, Duncan JS, Witt J-­A, et al. Getting the best
outcomes from epilepsy surgery. Ann Neurol 2018;83:676–90.
6. von Oertzen Jet al. Standard magnetic resonance imaging is
inadequate for patients with refractory focal epilepsy. J Neurol
Neurosurg Psychiatry 2002;73:643–7.
7. Winston GP, Micallef C, Kendell BE, et al. The value of repeat
neuroimaging for epilepsy at a tertiary referral centre: 16 years
of experience. Epilepsy Research 2013;105:349–55.
8. Beniczky S, Neufeld M, Diehl B, et al. Testing patients during
seizures: A European consensus procedure developed by a joint
taskforce of the ILAE - Commission on European Affairs and
the European Epilepsy Monitoring Unit Association. Epilepsia
2016;57:1363–8.
9. Mullin JP, Shriver M, Alomar S, et al. Is SEEG safe?
A systematic review and meta-­analysis of stereo-­
electroencephalography-­related complications. Epilepsia
2016;57:386–401.
10. Arya R, Mangano FT, Horn PS, et al. Adverse events related
to extraoperative invasive EEG monitoring with subdural grid
electrodes: a systematic review and meta-­analysis. Epilepsia
2013;54:828–39.
11. Baxendale S. Neuropsychological assessment in epilepsy. Pract
Neurol 2018;18:43–8.
12. Baxendale S, Thompson P, Harkness W, et al. Predicting
memory decline following epilepsy surgery: a multivariate
approach. Epilepsia 2006;47:1887–94.
13. Bauer PR, Reitsma JB, Houweling BM, et al. Can fMRI
safely replace the Wada test for preoperative assessment
of language lateralisation? A meta-­analysis and systematic
review. Journal of Neurology, Neurosurgery & Psychiatry
2014;85:581–8.
14. Winston GP, Daga P, White MJ, et al. Preventing visual field
deficits from neurosurgery. Neurology 2014;83:604–11.
15. Spencer DD, Spencer SS, Mattson RH, et al. Access to
the posterior medial temporal lobe structures in the
surgical treatment of temporal lobe epilepsy. Neurosurgery
1984;15:667–71.
16. Bonini F, McGonigal A, Scavarda D, et al. Predictive factors
of surgical outcome in frontal lobe epilepsy explored with
Stereoelectroencephalography. Neurosurgery 2017.
17. Schmeiser B, Zentner J, Steinhoff BJ, et al. Functional
hemispherectomy is safe and effective in adult patients with
epilepsy. Epilepsy & Behavior 2017;77:19–25.
18. Tanriverdi T, Olivier A, Poulin N, et al. Long-­term seizure
outcome after corpus callosotomy: a retrospective analysis of
95 patients. JNS 2009;110:332–42.
19. Elliott RE, Morsi A, Kalhorn SP, et al. Vagus nerve stimulation
in 436 consecutive patients with treatment-­resistant epilepsy:
long-­term outcomes and predictors of response. Epilepsy &
Behavior 2011;20:57–63.
20. Fisher R, Salanova V, Witt T, et al. Electrical stimulation of
the anterior nucleus of thalamus for treatment of refractory
epilepsy. Epilepsia 2010;51:899–908.
21. Morrell MJ, RNS System in Epilepsy Study Group. Responsive
cortical stimulation for the treatment of medically intractable
partial epilepsy. Neurology 2011;77:1295–304.
22. Kelley K, Theodore WH. Prognosis 30 years after temporal
lobectomy. Neurology 2005;64:1974–6.
23. Téllez-­Zenteno JF, Ronquillo LH, Moien-­Afshari F, et al.
Surgical outcomes in lesional and non-­lesional epilepsy:
12 of 13
a systematic review and meta-­analysis. Epilepsy Research
Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
2010;89:310–8.
24. Bell GS, de Tisi J, Gonzalez-­Fraile JC, et al. Factors affecting
seizure outcome after epilepsy surgery: an observational series.
J Neurol Neurosurg Psychiatry 2017;88:933–40.
25. West S, Nolan SJ, Cotton J, et al. Surgery for epilepsy.
Cochrane Database Syst Rev 2015;101.
26. Gooneratne IK, Mannan S, de Tisi J, et al. Somatic
complications of epilepsy surgery over 25 years at a single
center. Epilepsy Research 2017;132:70–7.
27. Cleary RA, Baxendale SA, Thompson PJ, et al. Predicting and
preventing psychopathology following temporal lobe epilepsy
surgery. Epilepsy & Behavior 2013;26:322–34.
28. Asadi-­Pooya AA, Asadollahi M, Tinker J, et al. Post-­epilepsy
surgery psychogenic nonepileptic seizures. Epilepsia
2016;57:1691–6.
29. Menon R, Rathore C, Sarma SP, et al. Feasibility of
antiepileptic drug withdrawal following extratemporal
resective epilepsy surgery. Neurology 2012;79:770–6.
30. Kerling F, Pauli E, Lorber B, et al. Drug withdrawal after
successful epilepsy surgery: how safe is it? Epilepsy & Behavior
2009;15:476–80.
31. Park K-­I, Lee SK, Chu K, et al. Withdrawal of antiepileptic
drugs after neocortical epilepsy surgery. Ann Neurol.
2010;67:230–8.
32. Wicks RT, Jermakowicz WJ, Jagid JR, et al. Laser interstitial
thermal therapy for mesial temporal lobe epilepsy.
Neurosurgery 2016;79(Suppl 1):S83–S91.
33. Gross RE, Stern MA, Willie JT, et al. Stereotactic laser
amygdalohippocampotomy for mesial temporal lobe epilepsy.
Ann Neurol. 2018;83:575–87.
34. Duncan JS, Winston GP, Koepp MJ, et al. Brain imaging in
the assessment for epilepsy surgery. The Lancet Neurology
2016;15:420–33.
35. Rathore C, Dickson JC, Teotónio R, et al. The utility of
18F-­fluorodeoxyglucose PET (FDG PET) in epilepsy surgery.
Epilepsy Research 2014;108:1306–14.
36. Fairclough S, O'Keeffe AG, de Tisi J, et al. Auras and the
risk of seizures with impaired consciousness following
epilepsy surgery: implications for driving. J Neurol Neurosurg
Psychiatry 2017.
37. Ladino LD, Hernández-­Ronquillo L, Téllez-­Zenteno JF.
Management of antiepileptic drugs following epilepsy surgery:
a meta-­analysis. Epilepsy Research 2014;108:765–74.
38. Baxendale S, Thompson PJ, Duncan JS. Improvements in
memory function following anterior temporal lobe resection
for epilepsy. Neurology 2008;71:1319–25.
39. Sperling MR, Barshow S, Nei M, et al. A reappraisal of
mortality after epilepsy surgery. Neurology 2016;86:1938–44.
40. Ramos-­Perdigués S, Baillés E, Mané A, et al. A prospective
study contrasting the psychiatric outcome in drug-­resistant
epilepsy between patients who underwent surgery and a
control group. Epilepsia 2016;57:1680–90.
41. Téllez-­Zenteno JF, Wiebe S. Long-­term seizure and
psychosocial outcomes of epilepsy surgery. Curr Treat Options
Neurol 2008;10:253–9.
42. Baxendale S, Thompson P. Red flags in epilepsy surgery:
identifying the patients who pay a high cognitive price for
an unsuccessful surgical outcome. Epilepsy & Behavior
2018;78:269–72.
43. Tebo CC, Evins AI, Christos PJ, et al. Evolution of cranial
epilepsy surgery complication rates: a 32-­year systematic
review and meta-­analysis. JNS 2014;120:1415–27.
Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192
 Review
44. Blumer D, Wakhlu S, Davies K, et al. Psychiatric outcome of 45. Iranzo-­Tatay C, Rubio-­Granero T, Gutierrez A, et al.

Pract Neurol: first published as 10.1136/practneurol-2019-002192 on 16 August 2019. Downloaded from http://pn.bmj.com/ on August 12, 2022 by guest. Protected by copyright.
temporal lobectomy for epilepsy: incidence and treatment of Psychiatric symptoms after temporal epilepsy surgery. A one-­
psychiatric complications. Epilepsia 1998;39:478–86. year follow-­up study. Epilepsy & Behavior 2017;70:154–60.

Rugg-­Gunn F, et al. Pract Neurol 2020;20:4–14. doi:10.1136/practneurol-2019-002192 13 of 13