ORIGINAL RESEARCH

published: 17 May 2021

doi: 10.3389/fneur.2021.654668

Surgical Outcomes and EEG

Prognostic Factors After Stereotactic
Laser Amygdalohippocampectomy
for Mesial Temporal Lobe Epilepsy
Shasha Wu 1*† , Naoum P. Issa 1† , Maureen Lacy 2 , David Satzer 3 , Sandra L. Rose 1 ,
Carina W. Yang 4 , John M. Collins 4 , Xi Liu 5 , Taixin Sun 6 , Vernon L. Towle 1 ,
Douglas R. Nordli Jr. 7 , Peter C. Warnke 3 and James X. Tao 1
1
Department of Neurology, The University of Chicago, Chicago, IL, United States, 2 Department of Psychiatry, The University
of Chicago, Chicago, IL, United States, 3 Department of Neurosurgery, The University of Chicago, Chicago, IL, United States,
4
Department of Radiology, The University of Chicago, Chicago, IL, United States, 5 Department of Neurology, Wuhan
University, Wuhan, China, 6 Department of Neurology, Beijing Electric Power Hospital, Beijing, China, 7 Department of
Pediatric Neurology, The University of Chicago, Chicago, IL, United States

Edited by: Objective: To assess the seizure outcomes of stereotactic laser

Emma Losito,
APHP, France amygdalohippocampectomy (SLAH) in consecutive patients with mesial temporal
Reviewed by: lobe epilepsy (mTLE) in a single center and identify scalp EEG and imaging factors in the
Kais Gadhoumi, presurgical evaluation that correlate with post-surgical seizure recurrence.
Duke University, United States
Chengyuan Wu, Methods: We retrospectively reviewed the medical and EEG records of 30 patients with
Thomas Jefferson University, drug-resistant mTLE who underwent SLAH and had at least 1 year of follow-up. Surgical
United States
outcomes were classified using the Engel scale. Univariate hazard ratios were used to
*Correspondence:
Shasha Wu
evaluate the risk factors associated with seizure recurrence after SLAH.
swu4@neurology.bsd.uchicago.edu
Results: The overall Engel class I outcome after SLAH was 13/30 (43%), with a mean
† These authors have contributed postoperative follow-up of 48.9 ± 17.6 months. Scalp EEG findings of interictal regional
equally to this work and share first
slow activity (IRSA) on the side of surgery (HR = 4.05, p = 0.005) and non-lateralizing or
authorship
contra-lateralizing seizure onset (HR = 4.31, p = 0.006) were negatively correlated with
Specialty section: postsurgical seizure freedom. Scalp EEG with either one of the above features strongly
This article was submitted to predicted seizure recurrence after surgery (HR = 7.13, p < 0.001) with 100% sensitivity
Epilepsy,
a section of the journal and 71% specificity.
Frontiers in Neurology
Significance: Understanding the factors associated with good or poor surgical
Received: 17 January 2021
outcomes can help choose the best candidates for SLAH. Of the variables assessed,
Accepted: 20 April 2021
Published: 17 May 2021 scalp EEG findings were the most clearly associated with seizure outcomes after SLAH.
Citation: Keywords: epilepsy surgery, medication resistant epilepsy, anterior temporal lobectomy, laser ablation, selective
Wu S, Issa NP, Lacy M, Satzer D, amygdalohippocampectomy, minimally invasive epilepsy surgery
Rose SL, Yang CW, Collins JM, Liu X,
Sun T, Towle VL, Nordli DR Jr,
Warnke PC and Tao JX (2021)
Surgical Outcomes and EEG
INTRODUCTION
Prognostic Factors After Stereotactic
Laser Amygdalohippocampectomy for
Temporal lobe epilepsy (TLE) is the most common type of medication-resistant epilepsy. Despite
Mesial Temporal Lobe Epilepsy. the development of new generations of anti-seizure medications (ASMs), 30–40% of patients
Front. Neurol. 12:654668. become resistant to ASM treatment (1). Anterior temporal lobectomy (ATL) and selective
doi: 10.3389/fneur.2021.654668 amygdalohippocampectomy (SAH) have been the gold standard surgical interventions for these

Frontiers in Neurology | www.frontiersin.org 1 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

patients, achieving 60–80% seizure freedom, though in highly emission tomography (FDG-PET) (except patients 1, 2, and
selected patient cohorts (2–5). However, patients are often 28), neuropsychological evaluation and functional MRI (fMRI).
hesitant or unwilling to consider open surgery due to the Scalp video EEG was recorded using Xltek, Natus Medical
fear of associated morbidity of a craniotomy, the concern Incorporated (Pleasanton, California, USA) at a sampling rate
for potential neurological deficits, and the risk of significant of 512 Hz. Electrodes were placed based on the international 10–
cognitive decline (6). 20 arrangement with supplementary sub-temporal electrodes F9,
In recent years, there has been a shift from open T9, F10, T10 from the 10–10 system, and mastoid electrodes M1,
resection to minimally invasive epilepsy surgery to M2. Scalp EEG filter settings are 1–70 Hz with notch filter on. No
minimize the complications associated with craniotomy automated artifact rejection was used.
and resection. Emerging data have shown that stereotactic Ictal and interictal electrographic patterns were tabulated.
laser amygdalohippocampectomy (SLAH) performed using Temporal intermittent rhythmic delta activity (TIRDA) is
MR-guided laser interstitial thermal therapy (MRgLITT) is a defined as short bursts of repetitive, rhythmic, 1–4 Hz activity
safe and effective alternative to open surgery for patients with of 50–100 mV in amplitude, predominantly running over the
mesial temporal lobe epilepsy (mTLE) (7, 8). In a recent study anterior temporal regions (13). Interictal regional slow activity
of 234 patients from 11 epilepsy centers, 58% of patients who (IRSA) is defined as delta activity over the temporal region, either
underwent SLAH achieved Engel I outcome after 1 and 2 years of continuous and polymorphic or intermittent and rhythmic on
postoperative follow-up (9). Compared to open resection, SLAH the same side of surgery which is present for more than 50%
is more tolerable and offers superior neurocognitive outcomes of the recording (14). An ictal onset EEG pattern in which the
by sparing the lateral temporal neocortex (10). SLAH has been side of onset cannot be determined in one or more seizures
adopted as the first-line surgical option in many epilepsy centers was categorized as a “non-lateralizing” seizure onset. Patient
in the United States for patients with mTLE with or without 7 had seizures recorded independently from both right and
mesial temporal sclerosis (MTS). left temporal regions on scalp EEG and was categorized as
Although early data showed that the seizure freedom rate “contra-lateralizing” seizure onset because a minority of seizures
of SLAH is close to or slightly inferior to that of traditional lateralized to the contralateral (non-operative) side.
open surgery (11, 12), its long-term seizure outcome has Brain imaging patterns were classified for each subject.
not been determined. Prognostic factors for identifying ideal Mesial temporal sclerosis (MTS) was defined as the presence of
candidates for SLAH in a not highly selected group have not MRI T2 /FLAIR signal hyperintensity with reduced hippocampal
been established. In this retrospective study, we aim to assess volume or loss of hippocampal internal architecture (15). 18F-
the seizure outcomes of SLAH in patients with at least 1 year FDG PET area of hypometabolism was classified as ipsilateral
of postoperative follow-up and to identify prognostic factors if it was on the same side of seizure onset and subsequent
from the presurgical scalp EEG and imaging that influence the SLAH, bilateral if hypometabolism was detected in both
surgical outcomes. temporal regions, and multifocal if hypometabolism involved
the ipsilateral or bilateral frontal, parietal, or occipital lobe in
MATERIALS AND METHODS addition to the temporal lobe.
Intracranial EEG recording with depth (stereo-
Patients and History electroencephalography, SEEG) and/or subdural electrodes
Thirty consecutive patients with drug-resistant TLE who was performed in 24 of 30 patients. Intracranial recording was
underwent SLAH by MRgLITT at the University of Chicago indicated when patients had a normal brain MRI, when there was
Medical Center from January 2014 to December 2019 with concern for lateral temporal neocortical onset, extratemporal
at least 1 year follow-up were included in the study. Patients onset, or if bilateral temporal onset could not be ruled out during
with bilateral interictal activity or bilateral seizures with one non-invasive evaluation. Simultaneous scalp and invasive EEG
side clearly predominating were also included making this were recorded using the above-mentioned recording system
cohort more heterogeneous than previously reported (11, 12). at a sampling rate of 1,024 Hz. Intracranial EEG filter settings
In patients who underwent a second SLAH due to persistent are 1–100 Hz with notch filter off. Eleven of 24 patients had
seizures, outcomes were reported with respect to reoperation. bitemporal intracranial recording due to concerns for bi-mesial
Exclusion criteria were prior open temporal lobe surgery, temporal onset. The methodology of intracranial recording was
ablation of structures outside the mesial temporal lobe at the detailed in our previous publication (16).
same time as SLAH, and postoperative psychogenic non-epileptic
seizures that precluded Engel classification. Patients’ charts,
imaging and EEG data were retrospectively reviewed. Stereotactic Laser
Amygdalohippocampectomy (SLAH)
Presurgical Evaluation Patients were eligible for SLAH when their intracranial EEG
All patients underwent a comprehensive neurological history onset localized to mesial temporal structures or predominantly
interview and examination, inpatient video-EEG monitoring, from one mesial temporal lobe in patients with bilateral mesial
brain MRI with volumetric 1.0 mm-section T1 coronal temporal onset seizures (patient 7). SLAH was performed in
and fluid-attenuated inversion recovery (FLAIR) images to all subjects using the Visualase system (Visualase Medtronic,
assess hippocampal volume, 18-fluorodeoxyglucose positron MN USA) by the same neurosurgeon (PCW). The location and

Frontiers in Neurology | www.frontiersin.org 2 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

volume of the ablation were determined either based on the RESULTS

results of intracranial EEG monitoring (the seizure onset zone
and immediate spreading zone were ablated) or, in the absence Demographic Data and Surgical Outcomes
of intracranial electrophysiological data, with the intention of Clinical data from 30 patients (17 female) with medically
ablating the AHC. Typically, three to five lesions were generated intractable TLE were reviewed and summarized in Table 1.
along the longitudinal axis of AHC. The detailed surgical The mean duration of epilepsy was 21.5 ± 16.9 years
technique used for MRI-guided SLAH has been previously (mean ± standard deviation; range 1–57 years). The
described (16). Seven of 30 patients had a second SLAH due to mean age at surgery was 43.6 ± 14.9 years (range 20–69
recurrent seizures (Table 1). The second SLAH was designed to years). The mean postoperative follow-up was 48.9 ± 17.6
target the residual AHC; the remnant mesial amygdala was the months (range 12–79 months).
target of re-ablation in four of seven patients. At the most recent follow-up, 13/30 (43%) remained seizure-
free after the most recent SLAH. The seizure freedom rate
was 18/30 (60%) at 1 year and 3/10 (30%) at 5 years. The
Assessment of Seizure Outcomes time from surgery to the first seizure varied between 1 and
The current Engel class seizure status and the time from SLAH 36 months. In ten of 17 patients with seizure recurrence,
to the first seizure recurrence were utilized as the endpoints seizures recurred within 6 months after surgery and all had
in the analysis. Engel classification is defined as the following: Engel class III or IV outcomes. Seven patients with recurrent
class I, free of seizures (patient may have aura); class II, rare seizures underwent a second SLAH to ablate residual mesial
disabling seizures; class III, worthwhile improvement; class IV, temporal tissue (Table 1). Four of those seven were seizure-free
no worthwhile improvement (17). For patients who had a second for more than 1 year after the second SLAH. The remaining
SLAH, the postoperative outcome was based on the second SLAH three continued experiencing seizures. Postoperative supervised
if there were more than 1 year of follow-up after the second reduction of at least one ASM was conducted in 11 of the 30
SLAH. Acute postoperative seizures that occurred in the first patients (Table 1).
week after surgery were not counted as recurrent (18). Patients
with seizures occurring after unsupervised ASM withdrawal (e.g.,
missed doses) with subsequent seizure-freedom for more than 2
Factors Associated With Surgery Failure
Five parameters were assessed using a univariate Cox
years after resuming the medications were classified as having an
proportional hazards model to identify which, if any, associated
Engel class I outcome with seizure freedom (17). Preoperative
with failure of surgery to produce seizure freedom. Hazard ratios
ASMs were maintained at least for 6 months after SLAH and in
>1.0 imply patients were likely to have recurrent seizures after
some cases were reduced thereafter if patients remained seizure-
surgery, and hazard ratios <1.0 imply patients were unlikely
free or adjusted if the seizures were not controlled. The surgical
to have recurrent seizures after surgery. A power analysis
complication rate from this cohort has been previously reported
suggests that an effect size of 2.7–3 (Hazard Ratio) would
(16), and surgical complications are not considered endpoints as
likely be detectable with 27–30 total subjects and the observed
we assume that they occur sporadically.
distribution of characteristics (unbalanced sampling). While
several other parameters are likely associated with outcome,
Data Analysis analysis was limited to five parameters because of the small
The goal of the study was to establish prognostic factors in pre- overall sample size (30 subjects with 17 that had postoperative
surgical evaluation for the effect of SLAH on seizures. Parameters seizure recurrence). The calculated hazard ratios are listed in
from the pre-surgical evaluation were tested as independent Table 2.
variables in a univariate Cox proportional hazard model with the The absence of unilateral MTS on MRI or of 18F-FDG PET
outcome being the duration of postsurgical seizure freedom. The hypometabolism restricted to the unilateral temporal appeared
parameters tested were: (1) Unilateral MTS on MRI. (2) 18F-FDG to be associated with a higher risk of seizure recurrence, but these
PET showed unilateral temporal hypometabolism. (3) Presence relationships did not reach statistical significance with the small
of interictal regional slow activity (IRSA) ipsilateral to the side of number of subjects available (Table 2). Three patients had MRI
surgery. (4) Presence of non-temporal interictal findings in the findings other than MTS. Patient 19 had diffuse band heterotopia
form of contralateral TIRDA, contralateral spikes or sharp waves, in bifrontal, parietal and temporal regions. Patient 28 had
or extratemporal interictal epileptiform activity. (5) Presence of ventricular enlargement due to normal pressure hydrocephalus.
non- or contra-lateralizing seizure onset. Stata software was used Patient 1 had right internal capsule gliosis.
for data analysis. A power analysis was performed to estimate Interictal scalp EEG patterns found outside of the surgical
effect size (Hazard Ratio) from the sample numbers available temporal lobe, consisting of bilateral or extratemporal sharp
using a long-rank test (Freedman method; power = 0.8, alpha waves, or contralateral TIRDA, also appeared to be associated
= 0.05) and an unbalanced design (the number of subjects with a higher risk of seizure recurrence. However, this
with a particular characteristic was different from the number relationship did not reach statistical significance (Table 2).
of subjects without that characteristic). A Bonferroni correction Sixteen of 30 patients had such interictal findings, and 12 of
for multiple comparisons was made; an initial p-value of 0.05 these 16 patients (75%) had postoperative seizures, on average
was considered significant, which is reduced to 0.01 because five 7.1 months after surgery. By comparison, only five of 14 patients
different parameters were tested. (36%) with unilateral temporal lobe discharges had postoperative

Frontiers in Neurology | www.frontiersin.org 3 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

TABLE 1 | Demographic information arranged by surgical outcomes.

ID Sex Age Epi duration MTS PET Non TL EEG IRSA Non-lat or contra sz Side Postop (months) Engel outcome Time to sz ASM reduction 2nd surg

11 F 41 4 No Yes Yes No No R 63 I NA No No
22 M 56 55 No No No No No R 44 I NA No No
23 F 51 5 No Yes No No No R 42 I NA Yes No
26 F 22 5 No Yes No No No R 37 I NA Yes No
4 F 42 26 Yes Yes Yes No No L 52 I NA* No Yes
6 M 25 15 Yes Yes No No No R 64 I NA Yes No
9 M 42 41 Yes Yes No No No L 28 I NA# No Yes
10 F 46 30 Yes No No No No R 66 I NA No No
14 F 46 23 Yes Yes Yes No No L 38 I NA@ Yes Yes
18 F 32 27 Yes Yes No No No R 49 I NA Yes No
20 F 61 34 Yes Yes Yes No No L 39 I NA∧ No Yes
29 M 67 50 Yes Yes No No No L 18 I NA No No
31 M 69 3 Yes Yes No No No L 12 I NA No No
2 F 53 3 No NA No Yes No R 78 II 17 Yes No
15 F 58 11 Yes No Yes Yes No R 59 II 10 No No
17 M 32 23 Yes Yes Yes Yes No L 50 II 24 Yes No
3 M 65 57 Yes No Yes No Yes L 74 II 18 Yes No
19 M 32 22 No No No No No R 48 II 18 Yes No
25 F 46 45 Yes Yes Yes No No R 39 II 12 Yes No
21 F 50 43 No No Yes Yes No L 45 III 2 Yes No
12 M 29 25 Yes Yes Yes Yes No L 51 III 2 No Yes
13 M 32 11 No Yes No No Yes L 60 III 1 No No
24 F 60 13 No No Yes No Yes R 42 III 3 No No
5 F 36 7 No Yes Yes No No L 66 III 1 No No
8 F 50 6 No Yes No No No R 69 III 1 No No
30 F 21 6 Yes Yes Yes No No R 14 III 8 No No
16 F 41 33 No No Yes Yes Yes R 41 IV 1 No Yes
7 M 20 5 No No Yes Yes Yes R 62 IV 3 No Yes
28 M 63 1 Yes NA No Yes No R 37 IV 1 No No
1 M 20 17 No NA Yes No Yes L 79 IV 1 No No

For patients who had a second surgery, time to seizure and postoperative follow-up duration are measured from the time of the second surgery.
Epi duration, duration of the epilepsy; MTS, ipsilateral mesial temporal lobe sclerosis; PET, ipsilateral temporal hypometabolism on positron emission tomography; non-TL EEG, non-
temporal lobe interictal EEG findings; IRSA, interictal regional slow activity; Non-lat or contra sz, non-lateralizing or contra-lateralizing seizure onset; side, side of surgery; Post-op, post-op
follow-up months; time to sz, time from surgery to first postoperative seizure; ASM; anti-seizure medications; 2nd surg, second surgery.
* Patient 4 had a seizure 11 months after the first ablation and was seizure free at 52 months after the second ablation. # Patient 9 had a seizure 36 months after the first ablation and was

seizure free at 28 months after the second ablation. @ Patient 14 had a seizure 7 months after the first ablation and was seizure free at 38 months after the second ablation. ∧ Patient
20 had a seizure 1 month after the first ablation and was seizure free at 39 months after the second ablation.

seizures, on average 7.6 months after surgery (HR 2.80, CI 0.97– hypometabolism in the lateral posterior temporal neocortex
8.03, p = 0.056). (patients 7, 12, and 15), with hypometabolism extending beyond
One additional interictal scalp EEG parameter was found to the temporal cortex into the occipital cortex in two (patients
be associated with seizure recurrence after surgery. All patients 7 and 15). In the other three there was either no clear
had focal slowing ipsilateral to the surgical side. Twenty-two hypometabolism (patient 16), exclusively anterior temporal lobe
patients had TIRDA and eight of 30 patients were noted to hypometabolism (patient 17), or bilateral anterior temporal lobe
have prominent, near-continuous focal slowing over the surgical hypometabolism (patient 21). For the 22 patients who did not
temporal lobe (Figure 1), similar to the interictal regional slow have IRSA, only nine (41%) had seizures after surgery, on average
activity (IRSA) described by Koutroumanidis et al. (14) All eight 7.0 months after surgery. The resulting hazard ratio for IRSA
patients with IRSA had recurrent seizures (100%, three Engel was 4.05 (CI 1.51–10.86, p = 0.005, which is significant after
class II and five class III and IV), with an average of 7.5 months Bonferroni correction for multiple comparisons; Table 2).
to first seizure after surgery. Koutroumanidis et al. (14) suggested Of the parameters assessed, ictal scalp EEG patterns had the
that temporal IRSA was often associated with hypometabolism closest association with surgery failure. Five patients had scalp
in the lateral posterior temporal lobe. A PET scan was available ictal EEGs that were non-lateralizing for one or more seizures
for six of the eight subjects with IRSA; in three there was (Table 1; Figure 2). One of five had a class II outcome and the

Frontiers in Neurology | www.frontiersin.org 4 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

TABLE 2 | Hazard ratios associated with characteristics identified during presurgical planning.

# #
Presurgical HR p 95% with/without postop sz Months to 1st sz
characteristic—univariate Confidence characteristic with/without with/without
analysis interval characteristic characteristic

No unilateral MTS 2.22 0.108 0.84–5.87 14/16 10/7 4.8/10.7

No unilateral PET 2.34 0.115 0.81–6.72 9/18 7/7 7.9/7.0
hypometabolism
Non-temporal lobe interictal 2.80 0.056 0.97–8.03 16/14 12/5 7.1/7.6
findings
Ipsilateral temporal lRSA 4.05 0.005 1.51–10.86 8/22 8/9 7.5/7.0
Non- or contra-lateralizing 4.31 0.006 1.51–12.34 6/24 6/11 4.5/8.7
seizure onset
Non- or contra-lateralizing 7.13 <0.001 2.41–21.07 12/18 12/5 6.9/8.0
seizure onset or
Ipsilateral temporal lRSA

Results of a univariate analysis, with each listed parameter tested as an independent variable. MTS, mesial temporal sclerosis on MRI. IRSA, interictal regional slow activity. HR, hazard
ratio for recurrent seizures after surgery, calculated with the Cox proportional hazards model. # with/without characteristic: the number of subjects in the cohort who had the defining
characteristic/the number of subjects in the cohort who did not have the defining characteristic; there were 30 subjects in the cohort. # postop sz with/without characteristic: the number
of subjects in the group with the characteristic that had a postoperative seizure/the number of subjects in the cohort who did not have the defining characteristic that had a postoperative
seizure. Months to 1st sz: the average number of months between surgery and the first seizure in subjects who had a seizure.

FIGURE 1 | Scalp EEG showing nearly continuous interictal regional slow activity (IRSA) on the surgical side (right hemisphere) in patient 2.

remaining four had class III or IV outcomes. Patient 7 had underwent right hippocampal SLAH and had a class IV outcome
independent bilateral ictal onset on scalp EEG recording with despite two SLAH on the right side. None of the six patients
most of the seizures starting over the right temporal region with either non-lateralizing or contra-lateralizing seizure onset
(R:L = 7:1). His intracranial recording showed independent were seizure-free (100% recurrence) compared to 11 of 24 (46%)
bitemporal interictal epileptiform discharges (IEDs), but all seizure-recurrence in patients who had only unilateral seizure
seizures were recorded from the right hippocampus. The patient onset. The average time from surgery to first seizure was 4.5

Frontiers in Neurology | www.frontiersin.org 5 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

FIGURE 2 | Scalp EEG showing the unilateral seizure onset (right temporal seizure onset) in patient 18 (top) and non-lateralizing seizure onset in patient 1 (bottom).
The orange lines indicate EEG seizure onset.

months in patients with non- or contra-lateraling seizure onset onset was associated with a hazard ratio of 7.13 (p < 0.001)
compared to 8.7 months in patients with unilateral seizure onset for recurrent seizures, with 100% sensitivity and 71% specificity.
(HR 4.31, CI 1.51–12.34, p = 0.006, which is significant after The positive predictive value (PPV) was 0.72, with 13 of 18
Bonferroni correction for multiple comparisons). patients without either of these patterns becoming seizure-free
Of the twelve patients who had either non- or contra- after surgery. The negative predictive value was 1.0, with all 12
lateralizing seizure onset or ipsilateral IRSA, none were seizure- patients with either IRSA or non- or contra-lateralizing ictal
free after surgery (100%) (Table 2). The seizures occurred, on onset having a postsurgical seizure. A Kaplan-Meier survival
average, 6.9 months after surgery. By comparison, only five curve shows the longer seizure freedom for the group of patients
of 18 patients (28%) with neither of these EEG features had without either IRSA or non- or contra-lateralizing ictal onset
a postoperative seizure, with an average latency of 8 months. compared to the group of patients with either EEG patterns
The presence of either IRSA or non- or contra-lateralizing ictal (Figure 3).

Frontiers in Neurology | www.frontiersin.org 6 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

DISCUSSION
Prognostic factors have been studied extensively for traditional
ATL (19, 20). This study analyzed the association between
presurgical characteristics and surgery outcome after SLAH in a
cohort of patients that included patients with either unilateral or
bilateral EEG findings. This approach was chosen following the
findings that in lesional cases even bilateral EEG-changes did not
preclude seizure freedom (21). A finding of either non- or contra-
lateralizing ictal EEG onset or interictal ipsilateral temporal IRSA
associated with the failure to achieve seizure freedom after SLAH
with a sensitivity of 100% and a specificity of 71%.
Patients with bitemporal seizures are poor candidates for
unilateral resection (22). However, the significance of non-
lateralizing ictal EEG regarding surgery outcome is not known.
We found that a non-lateralizing EEG or contra-lateralizing
seizure onset for even a minority of captured seizures is FIGURE 3 | Kaplan-Meier survival curve showing longer seizure freedom for
significantly negatively associated with seizure outcome. The the group of patients without either IRSA or non- or contra-lateralizing seizure
onset (dashed lines) than for the group of patients with either IRSA or non- or
presence of a non-lateralizing ictal EEG could suggest fast contra-lateralizing seizure onset (solid line). Two patients had both IRSA and a
bilateral synchronization or a contralateral seizure onset and lead non- or contra-lateralizing seizure onset on EEG. IRSA, interictal regional slow
to unfavorable surgical outcomes (23, 24). activity; SLAH, stereotactic laser amygdalohippocampectomy.
Temporal intermittent focal slowing (TIRDA) is a well-known
pattern that is considered an interictal marker for TLE (25, 26).
By contrast, continuous focal or regional slowing (IRSA) usually
suggests an underlying “structural” abnormality and is not were not statistically significant during the long-term follow-
considered epileptiform (27). In the current data set, however, up. Similar findings were reported from other groups recently
IRSA on the surgical side was associated with recurrent seizures (8, 36). Again, the observed hazard ratios were smaller than the
after SLAH (HR 4.05). The presence of IRSA could imply an detectable HR based on a power analysis (potential type I error),
extended epileptogenic zone beyond the mesial temporal region. so a larger data set would clarify the relationship between the
This is supported by the imaging findings of Koutroumanidis et imaging parameters and postsurgical seizure freedom.
al. (14) in which IRSA was associated with hypometabolism in While the use of SLAH has been increasing, the long-term
the lateral temporal neocortex. However, only half of the subjects outcome after SLAH is not yet well-understood. We found
that both IRSA and a PET scan showed lateral posterior temporal that the seizure-freedom rate after SLAH decreases gradually
hypometabolism, suggesting that IRSA, even independent of PET during the first 5 years of postoperative follow-up in this cohort
findings, may be a negative predictor of surgical outcome. To our with bilateral EEG changes and additional IRSA. The long-term
knowledge, this is the first description of IRSA in association with seizure outcome of SLAH appears to be marginally lower than
surgery outcomes in TLE. that of ATL, in which complete seizure freedom was seen in
Bi-temporal IEDs and postoperative contralateral TIRDA 55.3% of patients at 2 years and in 47.7% at 5 years after surgery
have been associated with poor surgical outcome while but the ATL cohorts are more selective (5). One possible reason
presurgical unilateral IEDs are good prognostic factors in for the difference in seizure freedom between ATL and SLAH
surgery for TLE (28–30). Our findings suggested that interictal could be due to incomplete ablation of extrahippocampal mesial
EEG findings outside the affected temporal lobe (bitemporal, tissues, as suggested by the recent study of Wu et al. (9) With
extratemporal IEDs and contralateral TIRDA) were not SLAH, structures other than the amygdalohippocampal complex,
significant independent factors to predict surgical outcome (HR including piriform, entorhinal, perirhinal, and parahippocampal
2.80, p = 0.06). However, based on a power analysis, the observed cortices, are not typically targeted and are inconsistently ablated.
HR of 2.8 would be in the borderline detectable effect size for the These extra-hippocampal mesial structures can be involved in
small number of subjects (potential type I error), and a larger seizure generation and propagation in mTLE and are typically
data set would be needed to determine whether this feature is a resected during ATL and SAH (37–40). Therefore, ablation of
significant predictor of surgical failure. additional mesial temporal structures, perhaps using two laser
Drug-resistant TLE associated with MTS has the best outcome probes targeting both longitudinal AHC and the mesial part of
after traditional ATL or SAH (31–33). Similar findings were also the amygdala, may be necessary to improve long-term seizure
reported after SLAH from our previous study and other epilepsy freedom. In a parallel study at our center, an analysis of ablated
centers (11, 12, 16, 34). Temporal PET hypometabolism has also regions suggests that parahippocampal ablation is associated with
been associated with a higher rate of seizure freedom after ATL seizure freedom in mesial temporal lobe epilepsy.
(35). While the lack of unilateral MTS or ipsilateral PET appeared Another important factor in explaining the inferior seizure
to be associated with a higher risk of seizure recurrence and a outcome after SLAH is the patient’s preference. Because of the
shorter time between surgery and first seizure, the relationships low risk of complications and much faster recovery time, almost

Frontiers in Neurology | www.frontiersin.org 7 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

all patients chose minimally invasive SLAH over traditional ATL seizure onset are strong negative markers of prognosis after
or SAH when we offered these two options to treat temporal lobe SLAH. The seizure-freedom rate after SLAH gradually decreases
epilepsy. In some patients, SLAH was offered as a “palliative” over the course of the initial 5-year postoperative follow-up.
procedure due to significant seizure burden and comorbidity Understanding the factors associated with good or poor surgical
despite the presurgical evaluation suggesting the seizure focus outcomes can help the selection of the best candidates for
was more extensive than a unilateral mesial temporal region. SLAH and help predict the outcome before surgery. Multi-
For example, palliative surgery without expectation for seizure center and long-term follow-up studies are warranted to
freedom was offered to patient 7, who had bitemporal onset clarify the long-term safety and efficacy of SLAH for patients
seizures on scalp EEG, and patient 19 who had double cortex on with mTLE.
MRI brain.
DATA AVAILABILITY STATEMENT
STUDY LIMITATIONS
The original contributions presented in the study are included
There are two major limitations to this study. First, its single- in the article/Supplementary Material, further inquiries can be
center retrospective design limits generalizability since the directed to the corresponding author.
patient population and indications for surgery may differ at
other surgical centers. The more significant limitation is the ETHICS STATEMENT
small sample size. There are many other potential predictors
of surgical outcome, for example the duration of epilepsy, The studies involving human participants were reviewed
seizure type, presence of non-MTS lesions, neuropsychological and approved by The University of Chicago Biological
findings, the extent of hypometabolism, but with only 30 subjects Sciences Division/University of Chicago Medical Center
it is not statistically appropriate to test all possibilities. We AURA Institutional Review Board. The patients/participants
therefore limited our analysis to five parameters related to provided their written informed consent to participate in
imaging and scalp EEG. However, to allow subsequent meta- this study.
analyses of other parameters, we have included additional data
in a Supplementary Table. Similarly, the power analysis suggests AUTHOR CONTRIBUTIONS
that only large effects could be detected with 30 subjects (Hazard
Ratios of ∼3). With 17 events, defined as recurrent seizures after SW and NI contributed to the study design, data collection,
completed ablation, a multivariate analysis was not possible. This and manuscript preparation. XL and TS contributed to the data
study is therefore preliminary, identifying two parameters that collection. PW performed the surgeries. DS, ML, SR, CY, JC, VT,
related to surgical failure with SLAH that should inform future DN, and JT contributed to the study design and data collection.
larger prospective studies. All authors approved the final manuscript.

CONCLUSION SUPPLEMENTARY MATERIAL

In conclusion, of the variables assessed, scalp EEG findings were The Supplementary Material for this article can be found
the most clearly associated with seizure outcomes after SLAH. online at: https://www.frontiersin.org/articles/10.3389/fneur.
Interictal regional slow activity and a non- or contra-lateralizing 2021.654668/full#supplementary-material

REFERENCES late recurrence and risks for seizure recurrence. Brain. (2004) 127(Pt
9):2018–30. doi: 10.1093/brain/awh221
1. Chen Z, Brodie MJ, Liew D, Kwan P. Treatment outcomes in patients with 6. Rausch R, Kraemer S, Pietras CJ, Le M, Vickrey BG, Passaro EA.
newly diagnosed epilepsy treated with established and new antiepileptic Early and late cognitive changes following temporal lobe surgery for
drugs: a 30-year longitudinal cohort study. JAMA Neurol. (2018) 75:279– epilepsy. Neurology. (2003) 60:951–9. doi: 10.1212/01.WNL.0000048203.23
86. doi: 10.1001/jamaneurol.2017.3949 766.A1
2. Jutila L, Immonen A, Mervaala E, Partanen J, Partanen K, Puranen M, 7. Wicks RT, Jermakowicz WJ, Jagid JR, Couture DE, Willie
et al. Long term outcome of temporal lobe epilepsy surgery: analyses of JT, Laxton AW, et al. Laser interstitial thermal therapy for
140 consecutive patients. J Neurol Neurosurg Psychiatry. (2002) 73:486– mesial temporal lobe epilepsy. Neurosurgery. (2016) 79(Suppl.
94. doi: 10.1136/jnnp.73.5.486 1):S83–S91. doi: 10.1227/NEU.0000000000001439
3. Wiebe S, Blume WT, Girvin JP, Eliasziw M. Effectiveness, efficiency of 8. Donos C, Breier J, Friedman E, Rollo P, Johnson J, Moss L, et al. Laser
surgery for temporal lobe epilepsy study G. A randomized, controlled trial ablation for mesial temporal lobe epilepsy: surgical and cognitive outcomes
of surgery for temporal-lobe epilepsy. N Engl J Med. (2001) 345:311– with and without mesial temporal sclerosis. Epilepsia. (2018) 59:1421–
8. doi: 10.1056/NEJM200108023450501 32. doi: 10.1111/epi.14443
4. Sperling MR, O’Connor MJ, Saykin AJ, Plummer C. Temporal lobectomy for 9. Wu C, Jermakowicz WJ, Chakravorti S, Cajigas I, Sharan AD, Jagid JR, et
refractory epilepsy. JAMA. (1996) 276:470–5. doi: 10.1001/jama.276.6.470 al. Effects of surgical targeting in laser interstitial thermal therapy for mesial
5. McIntosh AM, Kalnins RM, Mitchell LA, Fabinyi GC, Briellmann temporal lobe epilepsy: a multicenter study of 234 patients. Epilepsia. (2019)
RS, Berkovic SF. Temporal lobectomy: long-term seizure outcome, 60:1171–83. doi: 10.1111/epi.15565

Frontiers in Neurology | www.frontiersin.org 8 May 2021 | Volume 12 | Article 654668

Wu et al. Prognostic Factors With SLAH

10. Drane DL. MRI-guided stereotactic laser ablation for epilepsy surgery: 28. Burkholder DB, Sulc V, Hoffman EM, Cascino GD, Britton
promising preliminary results for cognitive outcome. Epilepsy Res. (2018) JW, So EL, et al. Interictal scalp electroencephalography and
142:170–5. doi: 10.1016/j.eplepsyres.2017.09.016 intraoperative electrocorticography in magnetic resonance imaging-
11. Kang JY, Wu C, Tracy J, Lorenzo M, Evans J, Nei M, et al. Laser interstitial negative temporal lobe epilepsy surgery. JAMA Neurol. (2014)
thermal therapy for medically intractable mesial temporal lobe epilepsy. 71:702–9. doi: 10.1001/jamaneurol.2014.585
Epilepsia. (2016) 57:325–34. doi: 10.1111/epi.13284 29. Schulz R, Luders HO, Hoppe M, Tuxhorn I, May T, Ebner A. Interictal
12. Willie JT, Laxpati NG, Drane DL, Gowda A, Appin C, Hao C, et al. Real-time EEG and ictal scalp EEG propagation are highly predictive of surgical
magnetic resonance-guided stereotactic laser amygdalohippocampotomy for outcome in mesial temporal lobe epilepsy. Epilepsia. (2000) 41:564–
mesial temporal lobe epilepsy. Neurosurgery. (2014) 74:569–84; discussion 70. doi: 10.1111/j.1528-1157.2000.tb00210.x
84–5. doi: 10.1227/NEU.0000000000000343 30. Tatum WO, Thottempudi N, Gupta V, Feyissa AM, Grewal SS,
13. Reiher J, Beaudry M, Leduc CP. Temporal intermittent rhythmic delta Wharen RE, et al. De novo temporal intermittent rhythmic delta
activity (TIRDA) in the diagnosis of complex partial epilepsy: sensitivity, activity after laser interstitial thermal therapy for mesial temporal
specificity and predictive value. Can J Neurol Sci. (1989) 16:398– lobe epilepsy predicts poor seizure outcome. Clin Neurophysiol. (2019)
401. doi: 10.1017/S0317167100029450 130:122–7. doi: 10.1016/j.clinph.2018.11.012
14. Koutroumanidis M, Binnie CD, Elwes RD, Polkey CE, Seed P, Alarcon 31. Josephson CB, Dykeman J, Fiest KM, Liu X, Sadler RM, Jette
G, et al. Interictal regional slow activity in temporal lobe epilepsy N, et al. Systematic review and meta-analysis of standard vs
correlates with lateral temporal hypometabolism as imaged with 18FDG selective temporal lobe epilepsy surgery. Neurology. (2013)
PET: neurophysiological and metabolic implications. J Neurol Neurosurg 80:1669–76. doi: 10.1212/WNL.0b013e3182904f82
Psychiatry. (1998) 65:170–6. doi: 10.1136/jnnp.65.2.170 32. Jeong SW, Lee SK, Hong KS, Kim KK, Chung CK, Kim H. Prognostic
15. Malmgren K, Thom M. Hippocampal sclerosis–origins and imaging. factors for the surgery for mesial temporal lobe epilepsy: longitudinal analysis.
Epilepsia. (2012) 53(Suppl. 4):19–33. doi: 10.1111/j.1528-1167.2012.03610.x Epilepsia. (2005) 46:1273–9. doi: 10.1111/j.1528-1167.2005.33504.x
16. Tao JX, Wu S, Lacy M, Rose S, Issa NP, Yang CW, et al. 33. Tellez-Zenteno JF, Hernandez Ronquillo L, Moien-Afshari F, Wiebe S. Surgical
Stereotactic EEG-guided laser interstitial thermal therapy for mesial outcomes in lesional and non-lesional epilepsy: a systematic review and meta-
temporal lobe epilepsy. J Neurol Neurosurg Psychiatry. (2018) analysis. Epilepsy Res. (2010) 89:310–8. doi: 10.1016/j.eplepsyres.2010.02.007
89:542–8. doi: 10.1136/jnnp-2017-316833 34. Waseem H, Osborn KE, Schoenberg MR, Kelley V, Bozorg A, Cabello D, et
17. Engel Jr J, Van Ness PC, Rasmussen TB, Ojemann LM. Outcome with respect al. Laser ablation therapy: an alternative treatment for medically resistant
to epileptic seizures. In: Engel Jr J, editor. Surgical Treatment of the Epilepsies. mesial temporal lobe epilepsy after age 50. Epilepsy Behav. (2015) 51:152–
New York, NY: Raven Press (1993). p. 609–21. 7. doi: 10.1016/j.yebeh.2015.07.022
18. Tigaran S, Cascino GD, McClelland RL, So EL, Richard Marsh 35. Manno EM, Sperling MR, Ding X, Jaggi J, Alavi A, O’Connor MJ, et al.
W. Acute postoperative seizures after frontal lobe cortical Predictors of outcome after anterior temporal lobectomy: positron emission
resection for intractable partial epilepsy. Epilepsia. (2003) tomography. Neurology. (1994) 44:2331–6. doi: 10.1212/WNL.44.12.2321
44:831–5. doi: 10.1046/j.1528-1157.2003.56402.x 36. Youngerman BE, Oh JY, Anbarasan D, Billakota S, Casadei CH, Corrigan
19. Meador KJ. Predictors of temporal lobe epilepsy surgery outcomes. Epilepsy EK, et al. Laser ablation is effective for temporal lobe epilepsy with
Curr. (2003) 3:125–6. doi: 10.1046/j.1535-7597.2003.03404.x and without mesial temporal sclerosis if hippocampal seizure onsets
20. Mariani V, Revay M, D’Orio P, Rizzi M, Pelliccia V, Nichelatti M, et al. are localized by stereoelectroencephalography. Epilepsia. (2018) 59:595–
Prognostic factors of postoperative seizure outcome in patients with temporal 606. doi: 10.1111/epi.14004
lobe epilepsy and normal magnetic resonance imaging. J Neurol. (2019) 37. Vismer MS, Forcelli PA, Skopin MD, Gale K, Koubeissi MZ. The piriform,
266:2144–56. doi: 10.1007/s00415-019-09394-x perirhinal, and entorhinal cortex in seizure generation. Front Neural Circuits.
21. Gupta A, Chirla A, Wyllie E, Lachhwani DK, Kotagal P, Bingaman (2015) 9:27. doi: 10.3389/fncir.2015.00027
WE. Pediatric epilepsy surgery in focal lesions and generalized 38. Bartolomei F, Khalil M, Wendling F, Sontheimer A, Regis J, Ranjeva
electroencephalogram abnormalities. Pediatr Neurol. (2007) JP, et al. Entorhinal cortex involvement in human mesial temporal lobe
37:8–15. doi: 10.1016/j.pediatrneurol.2007.03.004 epilepsy: an electrophysiologic and volumetric study. Epilepsia. (2005) 46:677–
22. Didato G, Chiesa V, Villani F, Pelliccia V, Deleo F, Gozzo F, 87. doi: 10.1111/j.1528-1167.2005.43804.x
et al. Bitemporal epilepsy: a specific anatomo-electro-clinical 39. Karunakaran S, Rollo MJ, Kim K, Johnson JA, Kalamangalam GP, Aazhang B,
phenotype in the temporal lobe epilepsy spectrum. Seizure. (2015) et al. The interictal mesial temporal lobe epilepsy network. Epilepsia. (2018)
31:112–9. doi: 10.1016/j.seizure.2015.07.013 59:244–58. doi: 10.1111/epi.13959
23. Carrette E, Vonck K, De Herdt V, Van Dycke A, El Tahry R, Meurs A, et 40. Al-Otaibi F, Baeesa SS, Parrent AG, Girvin JP, Steven D. Surgical techniques
al. Predictive factors for outcome of invasive video-EEG monitoring and for the treatment of temporal lobe epilepsy. Epilepsy Res Treat. (2012)
subsequent resective surgery in patients with refractory epilepsy. Clin Neurol 2012:374848. doi: 10.1155/2012/374848
Neurosurg. (2010) 112:118–26. doi: 10.1016/j.clineuro.2009.10.017
24. Malter MP, Bahrenberg C, Niehusmann P, Elger CE, Surges R. Features of Conflict of Interest: SW, NI, ML, SR, JC, PW, and JT are investigators in the
scalp EEG in unilateral mesial temporal lobe epilepsy due to hippocampal Stereotactic Laser Ablation for Temporal Lobe Epilepsy (SLATE) trial funded
sclerosis: determining factors and predictive value for epilepsy surgery. Clin by Medtronic.
Neurophysiol. (2016) 127:1081–7. doi: 10.1016/j.clinph.2015.06.035
25. Geyer JD, Bilir E, Faught RE, Kuzniecky R, Gilliam F. Significance of The remaining authors declare that the research was conducted in the absence of
interictal temporal lobe delta activity for localization of the primary any commercial or financial relationships that could be construed as a potential
epileptogenic region. Neurology. (1999) 52:202–5. doi: 10.1212/WNL.5 conflict of interest.
2.1.202
26. Tao JX, Chen XJ, Baldwin M, Yung I, Rose S, Frim D, et al. Interictal Copyright © 2021 Wu, Issa, Lacy, Satzer, Rose, Yang, Collins, Liu, Sun, Towle,
regional delta slowing is an EEG marker of epileptic network in temporal Nordli, Warnke and Tao. This is an open-access article distributed under the terms
lobe epilepsy. Epilepsia. (2011) 52:467–76. doi: 10.1111/j.1528-1167.2010.0 of the Creative Commons Attribution License (CC BY). The use, distribution or
2918.x reproduction in other forums is permitted, provided the original author(s) and the
27. Jan MM, Sadler M, Rahey SR. Electroencephalographic features copyright owner(s) are credited and that the original publication in this journal
of temporal lobe epilepsy. Can J Neurol Sci. (2010) 37:439– is cited, in accordance with accepted academic practice. No use, distribution or
48. doi: 10.1017/S0317167100010441 reproduction is permitted which does not comply with these terms.

Frontiers in Neurology | www.frontiersin.org 9 May 2021 | Volume 12 | Article 654668